Award Winners and Abstracts of the 33rd Annual Symposium of The Protein Society; Seattle, Washington, June 30‐July 3, 2019

Abstract Topics

1 – Amyloid and Aggregation

2 – Bioinformatics

3 – Chaperones

4 – Chemical Biology

5 – Computational Modeling/Simulation

6 – Design/Engineering

7 – Dynamics and Allostery

8 – Enzymology

9 – Evolution

10 – Folding

11 – Intrinsically Disordered Proteins

12 – Membrane Proteins

13 – Metabolic Engineering/Energy Applications

14 – Motors & Machines

15 – Peptides

16 – Protein Interactions and Assemblies

17 – Proteins in Cells

18 – Proteomics

19 – Proteostasis and Quality Control

20 – Single Molecule Studies

21 – Structure (X‐Ray/NMR/EM)

22 – Synthetic Biology

23 – Systems Biology

24 – Therapeutics and Antibodies

25 – Transcription/Translation/Post‐Translational Modifications

26 – Other

21. Structure (x‐ray/NMR/EM): 21. Structure (x‐ray/NMR/EM)

ABS002

STRUCTURAL AND BIOPHYSICAL CHARACTERIZATION OF ACYL‐CO‐A BINDING PROTEINS OF LEISHMANIA MAJOR

Shalini Verma ¹

¹National Institute of Immunology (New Delhi, India)

Acyl‐Co‐A‐binding proteins (ACBPs) are small alpha helical proteins, ubiquitously present in all living systems. ACBP's have been reported to be involved in membrane biosynthesis, regulation of gene expression, and various enzyme activities related to the lipid metabolism. ACBP's participate in acyl‐Co‐A transport, and functions in the intracellular acyl CoA pool formation. ACBP gene has been reported to be essential in the blood stream form of Trypanosoma brucei, a close relative of Leishmania. As the amastigote stage of Leishmania utilizes fatty acids as one of the source for energy production, these proteins might have importance in the parasite survival. In Leishmania major, we have identified six Acyl‐Co‐A binding proteins. The function of these proteins in Leishmania is largely unknown. In this study, we have structurally and functionally characterized the free standing ACBP's of Leishmania major. X‐ray diffraction of the protein crystals show the classical topology i.e. four alpha helical bundle structure, typical of ACBPs. We have found some structural differences in the loop regions and ligand binding regions of free standing ACBPs. NMR assignment was performed to understand the binding mechanism and dynamics of the proteins. Binding studies showed that Lmj ACBPs have high affinities and specificities towards the medium to long chain acyl‐Co‐A esters. Crystal structures, chemical shift assignments and mutational studies were helpful in the identification of the binding site residues and their interaction with the ligands. The structural data on ACBPs can be further employed in using them as drug targets against Leishmania major.

10. Folding

ABS004

FAST PRESSURE JUMP ALL‐ATOM SIMULATIONS AND EXPERIMENTS REVEAL SITE‐SPECIFIC PROTEIN DEHYDRATION‐FOLDING DYNAMICS

Taras Pogorelov ¹, Maxim Prigozhin², Yi Zhang¹, Klaus Schulten¹, Martin Gruebele¹

¹University of Illinois at Urbana‐Champaign (Urbana, United States); ²Stanford (Stanford, United States)

As theory and experiment have shown, protein dehydration is a major contributor to protein folding. Dehydration upon folding can be characterized directly by all‐atom simulations of fast pressure drops, which create desolvated pockets inside the nascent hydrophobic core. Here we study pressure‐drop refolding of three lambda repressor fragment (6‐85) mutants computationally and experimentally. The three mutants report on tertiary structure formation via different fluorescent helix‐helix contact pairs. All‐atom simulations of pressure drops capture refolding and unfolding of all three mutants by a similar mechanism, thus validating the non‐perturbative nature of the fluorescent contact probes. Analysis of simulated inter‐probe distances shows that the ‐helix pair 1‐3 distance displays a slower characteristic timescale than the 1‐2 or 3‐2 pairs. To see whether slow packing of ‐helices 1 and 3 is eflected in the rate‐limiting folding step, fast pressure‐drop relaxation experiments captured refolding on a millisecond timescale. These experiments reveal that refolding monitored by 1‐3 contact formation indeed is much slower than when monitored by 1‐2 or 3‐2 contact formation. Unlike the case of the two‐state folder 3D, whose drying and core formation proceed in concert, 6‐85 repeatedly dries and rewets different local tertiary contacts before finally forming a solvent‐excluded core, explaining the non‐two‐state behavior observed during refolding in molecular dynamics simulations. This work demonstrates that proteins can explore desolvated pockets and dry globular states numerous times prior to reaching the native conformation.

1. Amyloid and aggregation: 16. Protein interactions and assemblies

ABS008

TRANSTHYRETIN DISASSEMBLY MECHANISM AND METAL‐INDUCED OXIDATION DEGRADATION PATHWAY STUDIED VIA NATIVE MASS SPECTROMETRY AND SURFACE‐INDUCED DISSOCIATION

Mehdi Shirzadeh ¹

¹Texas A&M University (College Station, United States)

In first study, surface‐induced dissociation (SID) was used to investigate disassembly mechanism of transthyretin (TTR) using subunit exchange (SUE) experiment. Previous studies have shown that SID generates gas‐phase dissociation products resembling solution phase structure. For TTR as a dimer of dimer, SID of TTR yields different dimers corresponding to different topologies of TTR products from SUE (Figure 1A). Our results have confirmed model proposed by Kelly et al. (stepwise dissociation of tetramers to dimers and dimers to monomers). Also, the effect of temperature and ionic strength on TTR disassembly kinetics was revisited.

In second study, a home‐built Fourier transform ion mobility‐mass spectrometry (IM‐MS) was used to address whether oxidation promote or inhibit aggregation. With presence several endogenous metals (Zn, Cu, Cr and Ni), peaks corresponding to one and two Zn(II) bound to TTR were observed in MS. Stepwise mass shifts of 64 Da resembling third and fourth zinc binding were observed with longer data acquisition; however, SID of TTR revealed that oxidation of Cys‐10 and Met‐13 (two and one respectively) is the origin of mass increase which eventually leads to tetramer backbone fragmentation (b10). Using developed high‐resolution IM‐MS, two additional extended conformations were observed for oxidized TTR indicative of TTR unfolding which results in oligomer formation (Figure 1B). Interestingly, copper and chromium addition accelerated TTR oxidation indicating a metal‐induced oxidation phenomenon.

16. Protein interactions and assemblies: 21. Structure (x‐ray/NMR/EM)

ABS009

UNLOCKING THE MECHANISM OF HIV‐1 VIRAL ASSEMBLY NUCLEATION WITH NATIVE MASS SPECTROMETRY

Samantha Sarni ¹, Erik Olson¹, Shuohui Liu², Karin Musier‐Forsyth², Vicki Wysocki³

¹Ohio State Biochemistry Program, The Ohio State University (Columbus, United States); ²Chemistry and Biochemistry, The Ohio State University (Columbus, United States); ³NIH P41 Resource for Native Mass Spectrometry Guided Structural BIology, The Ohio State University (Columbus, United States)

The HIV‐1 pandemic has claimed over 36 million lives, warranting investigation of the mechanism of virion assembly to guide the development of novel therapeutics. Of the numerous drugs available to treat individuals, none of are aimed at disrupting the assembly of the immature viral particle. This work aims to probe the mechanism by which the virus nucleates immature particle assembly around a single dimeric RNA genome (gRNA) using native mass spectrometry (MS). Virion assembly requires Gag, the primary viral structural protein that interacts with the 5‐untranslated region (5UTR) of the viral gRNA. Gag specifically recognizes and binds to a packaging signal sequence (Psi) within the 5UTR, yet promiscuously interacts with any nucleic acid. Gag is an intrinsically dynamic precursor protein with RNA binding sites at each terminus. It is postulated that Gag binds Psi using one RNA binding domain in an extended conformation, while using both ends to bind non‐Psi RNA in a compact conformation. The size, heterogeneity, and dynamic nature of this system has precluded traditional structural biology techniques. Here, MS was used to determine the stoichiometry of Gag:RNA interactions. Gag bound to Psi with 1:1 and 2:1 stoichiometry but bound to non‐Psi RNA with exclusively 1:1 stoichiometry. These data suggest either that two independent binding sites are present on Psi or that Psi can induce Gag dimerization under conditions where non‐Psi RNA does not. These data are consistent with Gag conformational differences that depend on the identity of the bound RNA.

16. Protein interactions and assemblies: 10. Folding

ABS011

ROTENONE INTERACTIONS REMODEL PROTEIN IN‐TO CYTOTOXIC CONFORMERS

Shweta Devi ¹, Tulika Srivastava², Minal Chaturvedi², Smriti Priya³

¹Academy of Scientific & Innovative Research (AcSIR), India (Lucknow, India); ²Academy of Scientific & Innovative Research (AcSIR), India (Lucknow, India); ³Systems Toxicology and Health Risk Assessment Group, CSIR‐Indian Institute of Toxicology Research, Lucknow (Lucknow, India)

The fundamental mechanism for nucleation during protein misfolding and aggregation under the unrestricted influence of environmental tormentors, especially pesticides is poorly explored. These toxicants could play a major role in the genesis of several pathogenic disorders. Certain reports have established the effect of environmental toxicants on misfolding of Intrinsically Disordered Proteins (IDPs) but their role on non‐IDPs or metabolic proteins and their gain of toxicity is still elusive. To fill this lacuna, we hypothesized that the pesticides (e.g. Rotenone) as environmental toxicants may interact with metabolic proteins and lead to the generation of toxic conformers.

Docking studies confirmed the binding position of Rotenone to Malic Dehydrogenase (MDH), which was interestingly overlapping with the co‐factor binding site. Further, Rotenone did not affect the native activity of MDH however, prominent structural changes were observed. The hydrophobic changes in secondary structure of MDH were analyzed with Bis‐ANS. Circular dichroism showed structural alterations accompanied by the complete loss of alpha helices with an increase in beta sheets. Rotenone also destabilized the protein and reduced the melting temperature by 3°C as detected by thermal unfolding analysis. LC‐MS analysis confirmed stable Rotenone binding to MDH leading to misfolded conformers. Rotenone also interferes with spontaneous refolding of MDH, leading to the formation of toxic seeds. The cytotoxic effect of the seeds was established by the decrease in viability of SH‐SY5Y cells (60% viability for 0.25M seeds). The study has confirmed the binding of Rotenone to MDH and the generation of misfolded conformers which are cytotoxic in nature.

21. Structure (x‐ray/NMR/EM): 4. Chemical biology

ABS013

SNAPSHOTS OF IRREVERSIBLE FGFR1 INHIBITION

Maria Kalyukina ¹, Yuliana Yosaatmadja², Adam Patterson³, Jeff Smaill³, Christopher Squire¹

¹School of Biological Science, The University of Auckland; Maurice Wilkins Centre for Molecular Biodiscovery (Auckland, New Zealand); ²School of Biological Science, The University of Oxford; The University of Auckland (Oxford, United Kingdom); ³Auckland Cancer Society Research Centre, The University of Auckland (Auckland, New Zealand)

Oncogenic aberrations of fibroblast growth factor (FGF) signalling pathways promote the development of human malignancies1 including lung cancer, with non‐small cell lung cancer (NSCLC) accounting for more than 80% of all cases2. Therefore, fibroblast growth factor receptors (FGFRs) present attractive molecular targets for drug development3,4. The FGFR1‐4 family are receptor tyrosine kinases for which a renewed research and clinical focus has been their targeting by irreversible inhibitors that covalently modify the protein.

We have investigated a series of novel irreversible FGFR inhibitors and some clinical trial compounds including the potent and FGFR‐selective TAS‐1205 that is currently under phase I/II clinical trials (ClinicalTrials.gov identifier NCT02052778). We have focused on understanding the structural basis of irreversible inhibition, specifically, how the reactive centre of the inhibitor is presented to the target cysteine located in the ATP‐binding loop (P‐loop) of the kinase domain.

The structural basis of the exceptional TAS‐120 reactivity is revealed by a sequence of three crystal structures free ligand, reversible FGFR1, and the first reported irreversible FGFR1‐adduct structure6. We hypothesize that within our own series of compounds that inherently lower reactivity correlates with substitution at the acrylamide group, a potential advantage in a drug candidate in limiting off‐target toxicity. We show that structural flexibility within inhibitors as well as in the protein P‐loop, are critical molecular mechanisms in determining both reactivity and selectivity in FGFR targeting.

21. Structure (x‐ray/NMR/EM): 12. Membrane proteins

ABS014

STRUCTURE OF THE INFLUENZA B VIRUS M2 PROTON CHANNEL IN LIPID BILAYERS FROM SOLID‐STATE NMR

Venkata S. Mandala ¹, Martin D. Gelenter¹, Shu‐Yu Liao¹, Alex R. Loftis¹, Alexander A. Shcherbakov¹, Bradley L. Pentelute¹, Mei Hong¹

¹Massachusetts Institute of Technology (Cambridge, United States)

Influenza A and B viruses cause seasonal flu epidemics. The influenza B M2 protein (BM2) is a membrane‐embedded acid‐activated proton channel that is essential for the viral lifecycle. BM2 shares only 24% sequence identity with AM2 for the transmembrane (TM) domain, its structure is much less characterized compared to AM2, and no antiviral drug is so far available to inhibit BM2.

We have now used solid‐state NMR to determine the structure of BM2(151) in phospholipid bilayers. Using 2D and 3D correlation experiments, we have assigned the chemical shifts of the majority of the protein, yielding backbone (, ) torsion angles. Residues 628 form a well‐ordered ‐helix, whereas residues 15 and 2951 predominantly display chemical shifts indicative of random coil conformation. To elucidate how the channel opens in response to acidic pH, we have measured the orientation of the TM helix in lipid bilayers using rotationally averaged 1H‐15N dipolar couplings. Measurement of inter‐helical distances using 13C‐19F REDOR allowed us to determine the tetramer packing. These results show that the BM2‐TM domain forms a four‐helix bundle that is similar to the AM2‐TM structure despite the markedly different amino acid sequences, but that differs from the coiled‐coil structure reported for detergent micelle bound BM2. Thus, AM2 inhibitors do not target BM2 because of the polar nature of the BM2 pore rather than different three‐dimensional structures. Differences between the high‐pH and low‐pH structures give insights into the mechanism of proton conduction across this polar‐residue lined channel.

13. Metabolic engineering/energy applications: 8. Enzymology

ABS015

UNCOVERING THE IMPORTANT ENZYMES INVOLVED IN THE BIOSYNTHETIC PATHWAY OF BIOACTIVE POLYACETYLENES IN BIDENS PILOSA USING INTEGRATIVE OMICS APPROACHES

Lie‐Fen Shyur ¹, Hieng‐Ming Ting², Hsiao‐Hang Chung³, Wei‐Hsi Wang¹, Ya‐Ting Chao¹, Yi‐Chang Sung¹, Shih‐Shun Lin⁴

¹Agricultural Biotechnology Research Center, Academia Sinica (Taipei, Taiwan); ²Institute of Plant Biology, National Taiwan University (Taipei, Taiwan); ³Department of Horticulture, National Ilan University (Ilan, Taiwan); ⁴Institute of Biotechnology, National Taiwan University (Taipei, Taiwan)

Bidens pilosa L. (BP) (Asteraceae) is an herbal medicine commonly used for treating various disorders, such as inflammation, enteritis, dysentery, diabetes and hypertension. We have identified polyacetylenes (PA) compounds as the key bioactive compounds in BP plant attributed to its anti‐diabetes and anti‐angiogenesis effects. PAs are the type of compounds with carbon‐carbon triple bonds or alkynyl functional groups which might be derived from fatty acid or polyketide precursors; however, the information for the biosynthesis of bioactive polyacetylenes in this pharmacologically important plant is scarce. This study aims to uncover the important enzymes and regulatory factors which are involved in the PA biosynthetic pathway in BP plant. Integrative transcriptomics, functional genomics and metabolomics approaches in couple with various stress treatments, such as physical wounding, phytohormone (PH) treatments were used, aiming to differentiate and correlate the PAs production with the respective genes/enzymes involved in PAs biosynthesis in BP plant. Further, ContigViews, a bioinformatics tool was adapted to in silico analysis of gene correlation and networking assisting candidate genes selection. We have identified a couple of enzymes belonging to acetylenase and desaturase, which are likely involved in PA biosynthesis as validated by RNAi gene‐knock down as well as virus mediated gene overexpression approaches. The biochemical properties of both enzymes are on the way of characterization. This is the first study to discover the enzymes involved in PA biosynthesis in the important medicinal plant Bidens pilosa.

16. Protein interactions and assemblies: 23. Systems biology

ABS016

KINETIC TRAPPING AND ROBUSTNESS IN PROTEASOME ASSEMBLY

Anupama Kante ¹, Pushpa Itagi¹, Eric Deeds²

¹University of Kansas (Los Angeles, United States); ²University of California, Los Angeles (Los Angeles, United States)

Many macromolecular machines are involved in cellular functions like protein homeostasis, cellular transport, cell fate determination, etc. These machines are synthesized as a set of subunits that must be assembled in order for the complex to be functional. Kinetic trapping, where large intermediates are formed that cannot proceed to the fully‐assembled structure, can drastically reduce assembly yields. It is thus likely that macromolecular machines have evolved assembly pathways that avoid kinetic trapping. In this study we have chosen the 20S proteasome Core Particle (CP) from the actinomycete Rhodococcus erythropolis as our model system to characterize kinetic trapping, since its subunits can be purified and they readily self‐assemble into functional CPs in vitro. The proteasome CP is made up of four stacked rings of either seven or seven subunits; these rings are arranged in an 7777 order. Using mathematical models and experimental data, we have shown that CP assembly in Rhodococcus likely involves a trade‐off between assembly speed and robustness to kinetic trapping.These findings not only provide insight into the evolutionary pressures on macromolecular assemblies but also can be used to develop novel drugs that inhibit CP assembly (particularly for the treatment of Mycobacterium tuberculosis infections). Our results are also important for the engineering of protein nano machines that self‐assemble quickly and efficiently.

20. Single molecule studies: 16. Protein interactions and assemblies

ABS018

NSF‐MEDIATED DISASSEMBLY OF ON‐ AND OFF PATHWAY SNARE COMPLEXES AND INHIBITION BY COMPLEXIN

Ucheor Choi ¹, Minglei Zhao², Ian White¹, Axel Brunger¹

¹Stanford University (Stanford, United States); ²University of Chicago (Chicago, United States)

SNARE complex disassembly by the ATPase NSF is essential for neurotransmitter release and other membrane trafficking processes. We developed a single molecule FRET assay to monitor repeated rounds of NSF‐mediated disassembly and reassembly of individual SNARE complexes. For ternary neuronal SNARE complexes, disassembly proceeds in a single step within 100 msec. We observed short‐ (< 0.32 sec) and long‐lived ( 0.32 sec) disassembled states. The long‐lived states represent fully disassembled SNARE complex, while the short‐lived states correspond to failed disassembly or immediate re‐assembly. Either high ionic strength or decreased SNAP concentration reduces the disassembly rate while increasing the frequency of short‐lived states. NSF is also capable of disassembling anti‐parallel ternary SNARE complexes, suggesting a role in quality control. Finally, complexin‐1 competes with SNAP binding to the SNARE complex; addition of complexin‐1 has an effect similar to that of decreasing the SNAP concentration, suggesting a possible regulatory role in disassembly.

1. Amyloid and aggregation: 21. Structure (x‐ray/NMR/EM)

ABS019

A NOVEL AMYLOID FIBRIL STRUCTURE FORMED BY THE PEPTIDE HORMONE GLUCAGON

Martin Gelenter ¹, Katelyn Smith², Shu‐Yu Liao¹, Venkata Mandala¹, Aurelio Dregni¹, Matthew Lamm², Yu Tian³, Wei Xu², Darrin Pochan³, Thomas Tucker², Yongchao Su², Mei Hong¹

¹Massachusetts Institute of Technology (Cambridge, United States); ²Merck & Co., Inc. (Kenilworth, United States); ³University of Delaware (Newark, United States)

Under suitable conditions, many proteins aggregate into cross‐ amyloid fibrils stabilized by sidechain interactions and intermolecular hydrogen bonds. Glucagon and insulin are peptide hormones that maintain blood glucose homeostasis and are used to treat hypo‐ and hyperglycemia, respectively. Although insulin is stable for weeks in its solution formulation, glucagon fibrillizes rapidly, thus requiring formulation as a lyophilized powder that is reconstituted in solution immediately prior to use.

We have now determined a 1.5 å structure of glucagon fibrils formed at low pH using solid‐state NMR spectroscopy. Unexpectedly, two sets of chemical shifts with distinct water accessibilities and intermolecular contacts are observed, indicating the coexistence of two molecular conformations in the fibril. These two conformations alternate in a pair of antiparallel ‐sheets, which contain symmetric homodimer cross sections. In contrast to previously solved amyloid fibril structures, which contain both rigid and dynamic domains, the entire glucagon fibril peptide is rigid and its ‐sheet spans an unprecedented length of 27 residues, or nearly 10 nm. This novel amyloid structure is stabilized by numerous aromatic, cation‐, hydrophobic, and polar interactions. Determination of the low‐pH glucagon fibril structure opens the path for the rational design of glucagon analogs that resist fibril formation, increase its therapeutic efficacy, and enable the design of dual insulin/glucagon pumps as artificial pancreases for diabetic patients.

12. Membrane proteins: 21. Structure (x‐ray/NMR/EM)

ABS020

NATIVE‐STATE PROLYL ISOMERIZATION IS INVOLVED IN THE ACTIVATION OF A CNG CHANNEL

Philipp Schmidpeter ¹, Crina Nimigean¹

¹Weill Cornell Medicine (New York, United States)

The cyclic nucleotide‐gated channel SthK activates biphasically with cAMP application, with the slow phase reminiscent of the cAMP‐induced activation of eukaryotic HCN channels. The mechanistic underpinning for this effect is elusive. Here we show that SthK employs regulatory prolyl cis/trans isomerization in the cyclic‐nucleotide binding domain to slow down cAMP‐induced activation kinetics and fine‐tune activity. Substitution of a single Pro in SthK by Ala abolishes the slow activation phase and increases the apparent affinity of SthK for cAMP four‐fold, as measured in stopped‐flow assays. The same effects are observed for WT SthK in the presence of prolyl isomerases (PPIases), in a PPIase concentration‐dependent way. Neither the P‐A mutation nor application of PPIases affect the steady‐state single‐channel characteristics in planar lipid‐bilayer recordings. This suggests a mechanism where two channel conformations differentiated by a Pro in cis or trans configuration exist in equilibrium: while cis Pro is favored in the apo‐state, addition of cAMP shifts the equilibrium towards trans Pro in the open state. Activation of these two SthK conformations with different rates can explain the biphasic activation kinetics. Removal of the cis species in P‐A SthK or addition of PPIases that help to rapidly shift the equilibrium towards trans Pro in WT, will both lead to the disappearance of the slow phase. The cryoEM structure of P‐A SthK revealed subtle differences from the WT structure, suggesting that the mutant indeed adopts a pre‐active conformation. We propose that prolyl isomerization functions as molecular pacemaker for SthK that can be modulated by PPIases.

ABS021

IDENTIFICATION OF NOVEL SMALL MOLECULE RAS MODULATORS: A NEW PATH IN CANCER DRUG DISCOVERY

Patrick DePaolo ¹, Michael Sabio¹, William Windsor¹, Peter Tolias¹

¹Stevens Institute of Technology (Hoboken, United States)

Ras is a small GTPase enzyme that binds guanine diphosphate (GDP), a nucleotide substrate, to its guanine‐nucleotide binding pocket. Ras activation occurs via SOS, a guanine nucleotide exchange factor, which catalyzes the substitution of GDP for guanine triphosphate (GTP), rendering the Ras protein in an active, on state. Oncogenic Ras mutants are constitutively active, propagating proliferative cell signaling, leading to aberrant cell growth and metastatic tumors that are present in 30% of all human cancers. Discovery of a potent, small molecule Ras inhibitor has been challenging. We have begun to characterize a class of compounds that should inhibit Ras. These inhibitors were discovered using Ras‐SOS molecular modeling and computational docking studies focused on identifying potential compound binding to the Ras‐SOS complex interface. Several in‐vitro biochemical assays were performed to elucidate the structural significance, binding kinetics, and nucleotide exchange effects of these small molecule binders. A fluorescence nucleotide exchange assay was optimized to discover compounds that had an atypical nucleotide exchange rate (off‐state to on‐state) in the kinetic assay. Compounds discovered in the assay were then analyzed by Surface Plasmon Resonance (SPR) to further elucidate details on the binding kinetics between Ras and compound. Of the examined compounds, several exhibited novel activities including activation and inhibition of the Ras‐SOS catalytic reaction. This discovery of both inhibitors and activator compounds provides a promising new direction for the development of potent Ras‐implicated cancer drugs.

20. Single molecule studies: 7. Dynamics and allostery

ABS022

MICRO‐SECOND X‐RAY SINGLE MOLECULE DYNAMICS OF FUNCTIONAL PROTEINS USING SR AND LAB X‐RAY SOURCE

Yuji Sasaki¹, Yuji Sasaki ¹, Masahiro Kuramochi², Masaki Ishihara², Shoko Fujimura³, Kazuhiro Mio³

¹The University of Tokyo (Kashiwa, Japan); ²Graduate School of Frontier Sciences, The Univ. Tokyo (Kasiwa, Japan); ³Operand OIL, National Institute of Advanced Industrial Science and Technology (Kashiwa, Japan)

Diffracted X‐ray Tracking (DXT) using normal synchrotron orbital radiation source has been developed for obtaining the information of the 3D internal motions of single protein molecules with both high time‐resolution (micro‐seconds) and high precision (nm/1000). DXT is a method to obtain three‐dimensional (3D) dynamics through trajectories of the Laue diffraction spots from the labelled individual gold nanocrystal. Until now, we observed Brownian motions of individual DNA molecules, functional protein membranes (KcsA, nAChR, PAM‐a7AChR, TRPV, GPCR), antigen‐antibody interactions, and a super‐weak force (pN) that comes from x‐ray radiation pressure. Recently, we use DXT to monitor single molecule allosteric dynamics of trout Hemoglobin.

DXT makes effective use of pink beam. When using monochromatic X‐rays, it is impossible to track all motions of diffraction spots. However, we detected a clear blinking of diffracted X‐ray intensity due to motion of gold nanocrystals labeled with protein using monochromatic X‐ray. Now, we call Diffracted X‐ray Blinking (DXB). The observed X‐ray blinking intensity from the labeled and moving gold nanocrystals correlated with the velocity of the diffraction spots by autocorrelation function (ACF). Additionally, the standard deviation of the X‐ray blinking intensity distribution shifts to the larger side than when those stopped. Recently, we developed this technique to observe the molecular dynamics using laboratory X‐ray source. We have recently succeeded in being able to measure in vivo single molecular observations using living cells and Caenorhabditis elegans (C‐elegans).

3. Chaperones: 10. Folding

ABS023

TRANSIENT SPLITTING OF HSP104 HEXAMERIC RING AND ITS IMPLICATION IN PROTEIN DISAGGREGATION

Masafumi Yohda ¹, Yosuke Inoue¹, Yuya Hanazono², Kentaro Noi³, Akihiro Kawamoto⁴, Kazuki Takeda², Keiichi Noguchi¹, Keiichi Namba⁴, Teru Ogura³, Kunio Miki², Kyosuke Shinohara¹

¹Tokyo University of Agriculture & Technology (Koganei, Japan); ²Kyoto University (Kyoto, Japan); ³Kumamoto University (Kumamoto, Japan); ⁴Osaka University (Osaka, Japan)

Hsp104 and its bacterial homolog ClpB are AAA+ family proteins that rescue damaged proteins from aggregate state with the Hsp70 system. Since previous structural studies of Hsp104 and ClpB showed that Hsp104/ClpB forms the hexameric closed flat ring structure with a central channel, the substrate polypeptide is thought to be threaded through the central channel. On the other hand, the recent high resolution cryo‐EM analysis has shown the spiral architecture of Hsp104 oligomer. It is proposed that Hsp104 takes two conformations, Closed spiral with open pore before ATP hydrolysis and Closed flat ring with closed pore during ATP hydrolysis‐driven protein disaggregation. Here, we report the third conformation, Split spiral conformation. We have determined the crystal structure of Hsp104 from the thermophilic fungus, Chaetomium thermophilum (CtHsp104) in complex with ADP at 2.7 å resolution. CtHsp104 assembles into an oligomer of infinite‐bound spiral filament and the hexameric unit is in a split spiral conformation. In cryoEM and AFM observation, CtHsp104 mainly exists as the closed spiral conformation. It changes to Closed flat ring" in the presence of ATPgammaS. Strikingly, high‐speed AFM observation in the presence of ATP reveals transient splitting of the hexameric ring for approximately 0.7 sec when substrate physically interacts with the CtHsp104. We speculates that the splitting of the ring is important for capturing a substrate polypeptide in the protein aggregation.

5. Computational modeling/simulation: 7. Dynamics and allostery

ABS025

ADVANCED CLUSTERING, MACHINE LEARNING AND CONFORMATIONAL CHANGE: NEW METHODS AND SOME APPLICATIONS

Freddie Salsbury ¹, Ryan Melvin¹, Jiajie Xiao¹

¹Wake forest university (WINSTON SALEM, United States)

We will present recent work focused on improvements in statistical analysis of molecular dynamics simulations with applications to allostery and conformational change. First, we will present non‐parametric uses of two modern clustering techniques (HDBSCAN, iMWK‐Means) for clustering analysis, which can be used for structural analysis, ensemble analysis and Markov model analysis. Being non‐parametric, these methods require neither prior knowledge nor parameter tuning. We will also show how these methods can be used to understand correlated motions with the creation of "dynamical domains." Second, we will present a novel application of supervised learning, i.e. the decision tree learning on hydrogen bonds, to identify predicted key hydrogen bonds that distinguish the responses to different perturbations. For both these methods, we will focus on applications to large complexes, an MMR protein bound to damaged DNA, and to a smaller protein, thrombin.

19. Proteostasis and quality control: 25. Transcription/translation/post‐translational modifications

ABS026

SPONTANEOUS ISOMERIZATION IN LONG‐LIVED PROTEINS IS THE KEY TO UNDERSTANDING WHY ALZHEIMERS DISEASE COULD BE A LYSOSOMAL STORAGE DISORDER

Ryan Julian ¹

¹UC Riverside (Riverside, United States)

Alzheimers disease (AD) afflicts millions of people and has drawn significant attention from researchers, but the underlying cause remains elusive, as do effective treatment options. Amyloid aggregates have been associated with AD since its discovery, but ubiquitous commonalities with lysosomal storage disorders have been largely overlooked. Herein, we detail that isomerization of amino acids in long‐lived proteins, which occurs spontaneously over time, prevents proteolysis by lysosomal proteases both in isolation and in microglial cells. Amyloid‐ is highly isomerized in neuritic plaques, and we demonstrate that the rate of isomerization is sufficiently rapid to precede normal clearance. Gradual accrual of isomerized peptide fragments in the lysosome is hypothesized to lead to lysosomal storage and eventually cell death and neurodegeneration.

ABS027

AN ANALYTICAL REVOLUTION: INTRODUCING THE NEXT GENERATION OPTIMA AUC

Chad Schwartz¹, Olga Di Guida ²

¹Ex Beckman Coulter Employee (Indianapolis, United States); ²Beckman Coulter Life Sciences (Brea, United States)

The Optima AUC is the latest innovation from Beckman Coulter and offers significant advantages as an analytical ultracentrifugation platform over the predecessor ProteomeLab. Besides the plethora of advancements in user experience on the new platform, the Optima AUCs optical, thermal, and drive systems have also been upgraded significantly, creating an overall better experience to the end‐user in terms of data quality and ease of use. A battery of applications from proteins and nanomaterials to viral vectors and lipid‐based nanoparticles have already been assayed on the new instrument and the results are very positive in terms of data quality, reproducibility, and capabilities, especially in the case of multi‐wavelength experiments. Here, insights are provided to highlight the advantages of the new instrument as well as proof‐of‐concept data on the aforementioned applications.

24. Therapeutics and antibodies: 25. Transcription/translation/post‐translational modifications

ABS028

HIGH THROUGHPUT GLYCAN PROFILING FOR IMPROVED QUALITY CONTROL OF THERAPEUTIC GLYCOPROTEINS

Baolin Zhang ¹, Lei Zhang¹, Shen Luo¹

¹Food and Drug Administration (Silver Spring, United States)

The quality of therapeutic glycoproteins, including monoclonal antibodies (mAbs) and other biologics, is dictated by their carbohydrate or glycan profiles. However, the complexity of protein glycosylation poses a daunting analytical challenge. The most widely used methods, including HPLC coupled with mass spectrometry (MS), involve laborious enzymatic digestion procedures and analyses of the resulting free glycans and naked proteins. Lectin microarray technology has the potential to address the challenge of glycan analysis by harnessing lectins (natural carbohydrate‐binding proteins) to decipher glycan structures. This presentation describes a lectin microarray platform that directly measures glycans attached to intact proteins without the need of clipping glycans from the protein backbone. We tested a variety of samples from therapeutic mAbs to plasma proteins using lectin chips printed with a set of 45 different lectins, which selectively recognize glycan epitopes that are found in recombinant glycoproteins (e.g., fucose, sialic acids, mannose, and N‐acetylglucosamine). For all the samples tested, the glycan profiles derived from lectin‐binding signals are consistent with their known glycan profiles. Of importance, the lectin microarray is sensitive to detect alterations in the terminal glycan species such as galactose versus sialic acid epitopes. Upon optimization of lectin chips, lectin microarray platform could be adopted as a complementary tool for high throughput screening of glycan profiles of therapeutic glycoproteins.

12. Membrane proteins: 21. Structure (x‐ray/NMR/EM)

ABS029

A COMPREHENSIVE STUDY ON A MONOTOPIC MEMBRANE PROTEIN (S)‐MANDELATE DEHYDROGENASE AND ITS CHIMERAS

Narayanasami Sukumar ¹, Bharati Mitra², Sahana Sukumar³, Scott Mathews⁴

¹Cornell University (Argonne, United States); ²Wayne State University School of Medicine (Detroit, United States); ³Metea Valley High School (Aurora, United States); ⁴Washington University School of Medicine (St Louis, United States)

(S)‐ Mandelate Dehydrogenase (MDH) is a membrane associated, FMN‐dependent ‐hydroxy acid oxidizing enzyme. The x‐ray structure of (S)‐mandelate dehydrogenase (MDH) from Pseudomonas putida at 2.2å resolution have been solved recently. As the wtMDH was notoriously difficult to crystallize for a long time, to make it amenable for crystallization, the 39 residue internal binding segment of MDH was replaced with a 20 residue segment derived from one of its soluble homologue, Glycolate Oxidase (GOX). This resulted in a soluble chimera of MDH, called MDH‐GOX2, which retained full catalytic activity of wild type MDH. The x‐ray structure of MDH‐GOX2, G81A mutant of MDH‐GOX2 and in complex with its slow substrates were previously reported.

Based on the structure of MDH, it is possible to propose that the interaction of MDH with the membrane is stabilized by coplanar electrostatic interaction, along with three ‐helixes providing additional stability by inserting into the membrane. These ‐helixes may play a role in enzymes mobility and one of these ‐helixes could act as a gate for an electron acceptor by opening and closing the channel leading to its cofactor FMN. The putative membrane binding surface of MDH has a concentric circular ridge, which projects away from the protein surface by ~4å; this is unique and not observed in other monotopic membrane proteins to our knowledge. A detailed comparison between the MDH and its chimera MDH‐GOX2 and also chimera mutant G81A of MDH‐GOX2 has been made. The results would be discussed.

4. Chemical biology: 16. Protein interactions and assemblies

ABS030

GENETICALLY ENCODED PHOTOCAGED QUINONE METHIDE FOR PHOTO‐CONTROLLED CHEMICAL CROSSLINKING

Jun Liu ¹, Shanshan Li¹, Nayyar A. Aslam², Feng Zheng², Bing Yang¹, Rujin Cheng³, Nanxi Wang¹, Sharon Rozovsky³, Peng G. Wang⁴, Qian Wang², Lei Wang¹

¹University of California, San Francisco (San Francisco, United States); ²Chinese Academy of Sciences (Hangzhou, China); ³University of Delaware (Newark, United States); ⁴Georgia State University (Georgia, United States)

Quinone methide (QM) is a highly reactive short‐lived intermediate that continuously draw researchers attention for organic synthesis. QM‐derived small molecular chemical probes have also been developed and utilized for probing protein‐protein interactions. However, it has not been site specifically incorporated into proteins. To overcome this limitation, we have synthesized a photocaged para‐QM precursor (QMP) unnatural amino acid and genetically encoded it into proteins. The QM can be generated in situ in a photo‐controlled fashion for protein crosslinking with surrounding nucleophile within minutes. We demonstrated the robust crosslinking efficiency and diverse substrate specificity of QM both in vitro and in vivo. We envisage this genetically encoded QM should greatly expand the toolbox for protein crosslinkings to probe protein‐protein interactions in the future.

20. Single molecule studies: 25. Transcription/translation/post‐translational modifications

ABS031

PROTEIN SYNTHESIS STUDIES ON SINGLE MOLECULE LEVEL

Joerg Fitter ¹, Henning Hoefig¹, Alexandros Katranidis¹

¹Research Centre Juelich (Juelich, Germany)

Protein synthesis is a fundamental cellular process by which ribosomes decode genetic information and convert it into an amino acid sequence. The use of cell‐free protein synthesis (CFPS) systems allowed us to follow some essential steps on single molecule level. By monitoring the synthesis of green fluorescence proteins (GFPs) two color colocalization imaging and confocal two‐color coincidence detection (TCCD) was employed to quantify the fraction of active ribosomes and to measured their productivity [1,2]. Based on a further development of the TCCD method we monitored the subunit dissociation of 70S ribosomes for translation initiation in a cell free transcription/translation assay and unravel a previously undetermined second mechanism in bacterial protein synthesis initiation. On the contrary to initiation of canonical mRNA which requires dissociated 30S and 50S ribosomal subunits before translation, we show that protein synthesis is also initiated directly by 70S complexes, i.e. 30S and 50S subunits remain associated [3]. In another approach we aim to follow co‐translational folding of growing nascent polypeptide chains while they are still tethered to the ribosome. In a first step towards this goal we accomplished an direct incorporation of two different fluorescent dyes into the sequence [4]. We made use of the incorporated dyes to perform single molecule FRET and traced different protein conformations.

[1] Katranidis et al., Angew. Chem. Int. Edit., 48, 1758‐1761 (2009)

[2] Kempf et al., Scientific Reports, 7, 46753 (2017)

[3] Höfig et al., Communications Biology, under review (2019)

[4] Sadoine et al., ACS Synth. Biol., 7, 405‐411 (2018)

6. Design/engineering: 12. Membrane proteins

ABS032

QTY DESIGNED HEAT RESISTANT SOLUBLE TRANSMEMBRANE PROTEINS RECEPTOR WITH TUNABLE LIGAND AFFINITY

Rui Qing ¹, Qiuyi Han², Myriam Badr³, Haeyoon Chung¹, Michael Skuhersky¹, Thomas Schubert⁴, Shuguang Zhang¹

¹Massachusetts Institute of Technology (Cambridge, United States); ²Fudan University (Shanghai, China); ³NanoTemper Technologies (Cambridge, United States); ⁴2bind GmbH (Regensburg, Germany)

Chemokine receptors are one type of 7‐TM proteins members of G protein‐coupled receptors critical in human immunology and diseases. We previously reported QTY Code engineered detergent‐free chemokine receptors expressed in SF9 insect cells that binds to their native ligands. The route is expensive and has a low yield. Herein we present a low cost and simple route of e.coli expression, refolding and purification of 5 QTY designed chemokine receptors CCR5QTY, CCR10QTY, CXCR4QTY, CXCR5QTY and CXCR7QTY with throughput around 10mg per liter in LB media. As‐synthesized receptors bind their respective chemokines and exhibit affinity in nanomolar (nM) range, similar to values from native and SF9 produced variants. With arginine additive, QTY proteins exhibited exceptional heat resistance by retaining ligand binding activity after treatment at elevated temperature as high as 100 °C. Functionality and affinity towards specific ligands for these receptors can be regulated by hybridization of extracellular (EC) loops from different receptors. This controllability not only helps us to understand more about native GPCR binding mechanism, but also enables a feasible pathway of producing water‐soluble membrane proteins with tunable functionality for in vitro and in vivo applications.

20. Single molecule studies: 7. Dynamics and allostery

ABS033

MOLECULAR ORIGIN OF DISEASE MUTATIONS IN CAMP‐DEPENDENT PROTEIN KINASE A

Amy Chau ¹, Yuxin Hao¹, Clare Canavan¹, Susan Taylor², Rodrigo Maillard¹

¹Department of Chemistry, Georgetown University (Washington, United States); ²Department of Pharmacology, University of California, San Diego (La Jolla, CA, United States)

cAMP‐dependent protein kinase (PKA) is a biomedically important protein that regulates diverse cellular signaling pathways via allosteric regulation. PKA consists of the catalytic subunit that performs phosphorylation on protein substrates and the regulatory subunit that modulates the catalytic subunits activity. The regulatory subunit has two cyclic nucleotide binding (CNB‐A/B) domains, where the binding of cAMP initiates the propagation of signals from one domain to the other and finally releases the catalytic subunit. How the proteins communication is modulated by ligand binding remains poorly understood. Here, we use isothermal titration calorimetry and optical tweezers to investigate how ligand binding affects one domain in the presence of the other. We find that CNB‐A domain has a lower cAMP affinity than CNB‐B domain. The binding affinity of both CNB domains increases when the other domain is present. This result highlights the influence of one CNB domain to the other is bi‐directional and asymmetric. Additionally, we investigate W260 from CNB‐B that serves as the capping residue on cAMP in CNB‐A to probe the driving forces behind interdomain communication. The mutation W260A exhibits a decoupling effect between the CNB domains and destabilizes the conformation integrity in the bound state. This suggests W260 has important contributions to the CNB domains ligand binding cooperativity and structural stability. This study allows us to dissect how ligands induce domain communications in allosteric proteins. Moreover, this study enabled the identification and characterization of aberrant allosteric events associated with mutations observed in the disease state.

11. Intrinsically disordered proteins: 16. Protein interactions and assemblies

ABS034

RECRUITMENT OF AMYLOID‐ OLIGOMERS BY THE PRION PROTEIN

Priyanka Madhu ¹, Samrat Mukhopadhyay¹

¹Indian Institute of Science Education and Research, Mohali (MOHALI, India)

Amyloid‐ (A) oligomers are known to cause synaptic dysfunction in Alzheimers disease (AD) and have been shown to mediate the downstream cellular toxicity by binding to one of the cell‐surface receptors, the prion protein (PrP). However, the mechanism of interaction between conformationally distinct A oligomers and PrP is poorly understood. In this work, we dissect the mechanism of intermolecular association between conformationally distinct, prefibrillar (A11‐positive) and fibrillar (OC‐positive) A42 oligomers and PrP by using an array of biophysical and biochemical tools. We created a number of Cys mutational sites along PrP to obtain the site‐specific interaction of PrP with A oligomers. Our site‐specific binding titrations using fluorescence polarization, quenching, and intermolecular Förster resonance energy transfer measurements demonstrate that the heterotypic association of A oligomers and PrP primarily occurs via the N‐terminal intrinsically disordered region of PrP. Our fluorescence polarization experiments varying the ionic strength indicate that the intermolecular association is primarily driven by electrostatic interactions. We determine the binding affinities of OC‐positive fibrillar oligomers having an in‐register parallel ‐sheet structure and A11‐positive prefibrillar oligomers possessing an anti‐parallel ‐sheet structure with PrP. Our studies also demonstrate the toxic effects of A oligomers on binding PrP in mammalian cells. Taken together, our results revealed the electrostatic interactions between the intrinsically disordered region of PrP and drive the preferential recruitment mediating the deleterious effects of OC‐positive oligomers. Our studies also underscore the importance of designing the therapeutic strategies that target the interaction between OC‐positive oligomers and PrP.

16. Protein interactions and assemblies: 17. Proteins in cells

ABS035

TARGETING FTSZ ALONG WITH LAMA IS AN EFFECTIVE ANTIBACTERIAL STRATEGY AGAINST MYCOBACTERIUM SPECIES

RISHU TIWARI ¹, DULAL PANDA¹

¹INDIAN INSTITUTE OF TECHNOLOGY BOMBAY (MUMBAI, India)

Mycobacterium has a unique characteristic of asymmetric cell division which reduces the homogenous response of drugs and helps a subpopulation of the pathogen to evade the effect of drugs. LamA (loss of asymmetry) is responsible for this asymmetric division of Mycobacterium. The depletion of LamA reduces the asymmetric division in Mycobacterium. FtsZ polymerizes to form a cytokinetic Z‐ring that engineers the partitioning of bacterial cells. PC190723 inhibits the growth of Mycobacterium smegmatis with a minimal inhibitory concentration of 5 μM. The compound disassembles the Z‐ring in bacteria and also diffuses the localization of accessory cell division proteins like LamA and Wag31. The effect of PC190723 on the Z‐ring dynamics of M. smegmatis was studied using Fluorescence correlation spectroscopy (FCS). The effect of PC190723 on the diffusion coefficient (D) and translational correlation time () of FtsZ in live M. smegmatis cells was determined. FCS analysis showed that PC190723 dampened the dynamics of Z‐ring in live M. smegmatis cells where the D reduced from 73 ± 7.3 to 18 ± 3.8 μm2 s‐1 and the increased from 0.32 ± 0.03 to 1.35 ± 0.28 msec in the presence of PC190723. The effect of PC190723 on the growth of M. smegmatis was also checked in the presence of varying levels of LamA. The results suggested that LamA antagonized the antibacterial effect of PC190723. Thus, targeting FtsZ along with LamA can be an interesting strategy to inhibit the growth of Mycobacterium.

8. Enzymology: 7. Dynamics and allostery

ABS036

INTERSUBUNIT INTERACTIONS INVOLVING A LARGE SURFACE LOOP SHAPE THE CATALYTIC PROPERTIES AND STABILITY OF AN ALKALINE PHOSPHATASE

Jens Hjörleifsson ¹, Elena Papaleo², Bjarni Ásgeirsson¹

¹Science Institute University of Iceland (Reykjavik, Iceland); ²Danish Cancer Society Research Center (Copenhagen, Denmark)

Cold‐adapted enzymes have attracted research interest due to their high catalytic efficiency at low temperatures compared with enzymes from mesophilic or thermophilic organisms. The role played by remote residues away from the catalytic site of cold‐adapted enzymes is still not well understood. Here, we focused on the major loop at the interface between the two monomers of a cold‐adapted alkaline phosphatase by molecular dynamics calculations, x‐ray crystallography, experimental mutagenesis, kinetic analyses and thermal stability assays. We designed several different variants that were predicted to affect intermolecular interactions between the monomers of the enzyme. Computer simulations showed high flexibility of the major loop, where several hydrogen bonds and salt bridges were asymmetrically populated in the dimeric structure but absent in the crystal structure due to large loop transitions. By disrupting hydrogen bonds between a loop‐residing Arg residue and Ser residues in the opposite monomer across the interface, we showed conclusively the importance of the structural communication in this region and the capability of mutagenesis to modulate over a long range the conformational states of the residues in the catalytic site. In summary, our results imply a functional role for the unstructured major loop where increased local flexibility is not a parameter which relates to higher catalytic efficiency. Instead, a well‐organized and coherent set of interactions is needed to promote a conformational change to release product. This is achieved by dynamics of cooperating surface loops with cross‐talk from both monomers near the active site.

16. Protein interactions and assemblies: 8. Enzymology

ABS038

THE CRITICAL ROLE OF TYROSINE KINASE SEQUENCE SPECIFICITY IN T CELL ACTIVATION

Neel Shah ¹, Wan‐Lin Lo², Arthur Weiss², John Kuriyan³

¹Columbia University, Department of Chemistry (New York, United States); ²University of California, San Francisco (San Francisco, CA, United States); ³University of California, Berkeley (Berkeley, CA, United States)

T cells respond to foreign antigens with remarkable selectivity. The fidelity of T cells cannot be fully explained by differences in the affinities of cognate and non‐cognate antigens for the T cell antigen receptor (TCR). Thus, this fidelity must be encoded in downstream signaling steps, many of which are driven by cytoplasmic tyrosine kinases. In this study, we examined the role of tyrosine kinase specificity in controlling T cell activation. We developed a high‐throughput platform that combines bacterial surface display of peptide libraries with cell sorting and deep sequencing to profile tyrosine kinase specificity. Using this platform, we profiled the T cell tyrosine kinases Lck and ZAP‐70 and discovered an electrostatic substrate selection mechanism that enforces tight control in the initiating steps of TCR signaling. The specificity screens also revealed an unexpected kinetic bottleneck in the TCR pathway, defined by the slow phosphorylation of a single site on the scaffold protein LAT by ZAP‐70. Guided by our specificity profiling results, we engineered mutations near this critical phosphosite in LAT that increased the rate of signaling through this step. These mutations also increased T cell sensitivity toward weak agonists and self‐antigens. Our results strongly support a kinetic proofreading model for T cell antigen selectivity, whereby a slow commitment step downstream of the TCR enables the discrimination of TCR‐antigen complexes with modestly different lifetimes. These results also establish tyrosine kinase sequence specificity as a critical mechanism for signaling control in this system.

16. Protein interactions and assemblies: 17. Proteins in cells

ABS039

EMERGING ROLES FOR THE ACTIN BINDING PROTEIN PALLADIN IN REGULATION OF HIGHLY MOTILE CELLS

Moriah Beck ¹, Ritu Gurung¹, Sharifah Albariki¹, Aaron Dhanda², Carol Otey³, Wayne Vogl⁴, Julian Guttman²

¹Wichita State University (Wichita, United States); ²Simon Fraser University (Burnaby, Canada); ³University of North Carolina at Chapel Hill (Chapel Hill, United States); ⁴University of British Columbia (Vancouver, Canada)

Palladin is a recently discovered actin binding protein that plays a key role in both normal cell migration and invasive cell motility, yet its precise function in organizing the actin cytoskeleton is unknown. The majority of our previous research has focused on the scaffolding activity of palladin, whereas our results here add a new dimension to the relationship between palladin and actin by highlighting the direct role in actin assembly. Here we show that the C‐terminal immunoglobulin‐like domain of palladin, which is directly responsible for actin binding and bundling, also stimulates actin polymerization in vitro. Palladin eliminates the lag phase that is characteristic of the slow nucleation step of actin polymerization and dramatically reduced depolymerization. Total internal reflection fluorescence microscopy was employed to visualize the organization and dynamics of palladin‐induced actin polymerization. Using Listeria infections, we show that palladin is recruited by the bacteria during cellular entry and intracellular motility. Moreover, depletion of palladin results in shorter and misshapen actin comet tails. We also show that overexpression of palladin can reverse the loss of Listeria comet tail‐based motility in the presence of the Arp2/3 inhibitor, CK‐666, and in Arp2/3 complex depleted cells. Furthermore, Listeria actin clouds and comet tails can be generated by palladin in the absence of the Arp2/3 complex in reconstitution assays. Collectively, our results demonstrate that palladin stimulates actin polymerization and indicate that palladin is part of an actin organization and nucleation complex that can functionally replace the Arp2/3 complex in actin‐rich structures.

1. Amyloid and aggregation: 17. Proteins in cells

ABS040

REAL‐TIME MONITORING OF ‐SYNUCLEIN‐INDUCED CELL MEMBRANE DISRUPTION IN PARKINSONS DISEASE

Jacob Parres‐Gold ¹, Stephanie Wong Su¹, Andy Chieng¹, Yixian Wang¹

¹California State University, Los Angeles (Van Nuys, United States)

Parkinsons disease (PD) affects more than one million people in the United States alone, making it the second most common neurodegenerative disorder in the US. In PD, aggregates of the protein alpha‐synuclein (‐Syn) are known to induce the death of dopamine‐producing cells in the substantia nigra, ultimately hindering cognitive function. Although monomeric ‐Syn likely regulates vesicle trafficking, mutations and environmental conditions can promote its aggregation into toxic oligomers and, eventually, long fibrils. These aggregates have been observed to increase membrane permeability and promote membrane damage. However, the distinct nature and propagation of this disruption have yet to be characterized. This work used scanning ion conductance microscopy (SICM) to image in real‐time the topographies of SH‐SY5Y neuroblastoma cell membranes after the addition of ‐Syn aggregates. The ‐Syn aggregates were produced in‐vitro and characterized by atomic force microscopy (AFM) and circular dichroism (CD) spectroscopy. Significant damage to the cell membrane was observed, with both large defects and small pore‐like structures appearing across the membrane over time. This suggested that the ‐Syn aggregates may themselves be binding to and disrupting the membrane. Overall, this work has used scanning ion conductance microscopy as a novel method for monitoring in real‐time damage to neuroblastoma cell membranes and has suggested membrane binding as a mechanism of ‐Syn‐induced cell death.

1. Amyloid and aggregation: 16. Protein interactions and assemblies

ABS041

RUTIN INTERACTS WEAKLY WITH ‐SYNUCLEIN AND SUPPRESSES ITS AGGREGATION BY MODULATING ITS FIBRILLATION PATHWAY

Geetika Verma ¹, Rajiv Bhat¹

¹Jawaharlal Nehru University (New Delhi, India)

‐Synuclein (‐Syn) is known to contribute to Parkinsons Disease (PD) pathogenesis via protein aggregation that leads to neuronal death by targeting synaptic function. One promising strategy for PD treatment is to either inhibit fibril formation or disaggregate them into non‐toxic species. It has been reported that several flavonoids with 3,4‐dihydroxyl groups are found to have inhibitory effect on fibril formation. An array of potential amyloid polyphenolic inhibitors including Rutin containing these groups exhibit antiamyloidogenic activity by binding to hydrophobic ‐sheets and simultaneously disrupt the hydrogen bond formation, thereby preventing aggregation. Besides this, Rutin is also observed to have neuroprotective effect, reduce oxidative stress, and modulate immune and inflammatory cell functions. We have, thus, investigated the interaction of Rutin with ‐Syn and observed it to exert concentration dependent suppression of ‐Syn aggregation with substantial suppression at 200 μM that leads to the formation of alpha‐helix rich, smaller aggregates as shown by thioflavin T assay, CD spectroscopy, TEM and AFM. These species are found to be less cytotoxic as compared to the amyloid fibrils as observed by MTT assay performed on human neuroblastoma cell line. The binding isotherm of Rutin with ‐Syn is exothermic in nature with a binding constant of ~ 10^4 M‐1 suggesting maximum hydrogen bonding in the NAC and the C‐terminal region of ‐Syn as observed by calorimetric and docking studies. The study signifies the effect of Rutin on the modulation of ‐Syn aggregation into less toxic species that could make it a potential therapeutic agent for PD.

3. Chaperones: 16. Protein interactions and assemblies

ABS042

GENERATION OF RECOMBINANT LIGAND‐BINDING FRAGMENTS OF LOW‐DENSITY LIPOPROTEIN RECEPTOR‐RELATED PROTEIN 1 USING CO‐EXPRESSION WITH ITS CHAPERONE RECEPTOR‐ASSOCIATED PROTEIN

Ekaterina Marakasova ¹, Gabriela Uceda‐Cortez¹, Svetlana Shestopal¹, Timothy K Lee¹, Andrey Sarafanov¹

¹Division of Plasma Protein Therapeutics; Office of Tissues and Advanced Therapies; Center for Biologics Evaluation and Research; U. S. Food and Drug Administration (Silver Spring, United States)

Introduction: Low‐density lipoprotein receptor‐related protein 1 (LRP) is an endocytic receptor, which is expressed in many tissues and has numerous ligands. The ligand‐binding moiety of LRP is presented by complement‐type repeats (CRs), grouped in clusters. The clusters II, III and IV are known to bind all ligands of the receptor. Among these ligands, a prominent role plays a molecular chaperone of LRP, termed receptor‐associated protein (RAP). Previously, production of recombinant CR‐fragments resulted in low yields because most of protein was expressed in non‐functional misfolded multimeric form. In this study, we aimed to generate recombinant clusters II‐IV of LRP using co‐expression with RAP.

Methods: The fragments of LRP (Q07954) were co‐expressed with RAP (P30533) in a baculovirus system. The proteins were purified using metal‐affinity chromatography followed by size‐exclusion chromatography. The protein binding properties were studied by surface plasmon resonance.

Results: Co‐expression with RAP resulted in significantly higher yield of cluster II monomeric form, compared to expression without RAP. Using RAP with an insect endoplasmic reticulum retention signal HTEL resulted in further increase of the yield. Clusters III and IV were then expressed using the same approach, also resulted in high yields. Functional properties of generated clusters II‐IV were confirmed by testing their binding with RAP and blood coagulation factor VIII (an LRP‐ligand).

Conclusion: We developed a method of generation of recombinant CR‐fragments of LRP at high yields. The proteins were expressed in predominantly monomeric form, preserving the binding properties of the receptor and, by this, indicating correct protein folding.

6. Design/engineering: 16. Protein interactions and assemblies

ABS043

CONSTRUCTION OF PROTEIN SUPRAMOLECULES BASED ON DOMAIN‐SWAPPING MECHANISM

Masaru Yamanaka ¹, Satoshi Nagao¹, Chunguang Ren¹, Mohan Zhang¹, Akiya Oda¹, Yoshiki Higuchi²

¹Graduate School of Science and Technology Division of Materials Science, Nara Institute of Science and Technology (Ikoma, Japan); ²Graduate School of Life Science, University of Hyogo (Kamigori‐cho, Japan)

Natural protein supramolecules perform sophisticated functions in vivo. If we can freely build a protein supramolecule, it would lead innovation in various fields, including pharmacy, engineering, and so on. Methods to construct artificial protein supramolecules have been developing but are still limited. Thus, we have utilized domain swapping as a novel tool for construction of protein supramolecules. Domain swapping is a protein supramolecule formation mechanism, exchanging a protein structural region between molecules; in principle, it can occur in any protein. A hyperstable protein formed a thermally and pH stable dimer by domain swapping. An engineered protein having an elongated hinge loop, which is the loop with different conformations in the native monomer and domain‐swapped structures, formed a nanoring‐structured oligomer by domain swapping. Two chimeric proteins formed a heterogeneous domain‐swapped dimer. A circular permutated and helix linker‐inserted protein formed a cyclic domain‐swapped trimer. A gas sensor protein which exhibits CO responsive dimermonomer transition formed a non‐native domain‐swapped dimer. Interestingly, the gas sensor domain‐swapped dimers assembled into oligomers in the absence of CO, whereas the oligomers disassembled to the domain‐swapped dimers in the presence of CO. The interactions in the native gas sensor protein were maintained in the domain‐swapped oligomer, possessing the original ligand‐binding sensitive conformational change. These results demonstrate that domain swapping is useful for construction of unique protein supramolecules possessing unique structure and/or conformational change.

1. Amyloid and aggregation: 16. Protein interactions and assemblies

ABS044

MOLECULAR MECHANISMS OF PEPTIDE SELF‐ASSEMBLY IN HYDROGEL FORMATION

Gabriel Braun ¹, Sara Linse², Karin åkerfeldt¹

¹Department of Chemistry, Haverford College (Haverford, United States); ²Department of Biochemistry and Structural Biology, Lund University (Lund, Sweden)

The mechanism of fibril formation for a hydrogel‐forming peptide, the KD peptide, was investigated through fluorescence aggregation kinetic assays using thioflavin T as a reporter dye. These assays showed that the rate of self‐assembly is dependent on monomer concentration up to, but not above, 50 μM, suggesting the presence of a saturating aggregation mechanism. The fitting of data below this saturation point to kinetic models derived from various molecular mechanisms of self‐assembly indicated that the rate of self‐assembly for this system is consistent with a secondary nucleation dominated mechanism. Analysis of the monomer concentration dependence of aggregation using the half‐time plot suggested that the secondary nucleation mechanism (by which the nuclei of new fibrils are formed) is responsible for observed saturation in the rate of self‐assembly. Moving forward, seeded kinetic assays could help verify the mechanism of fibril formation as well as confirm which mechanism saturates at high monomer concentration.

2. Bioinformatics: 18. Proteomics

ABS045

UNIPROT: A UNIVERSAL HUB OF PROTEIN KNOWLEDGE

Alex Bateman ¹

¹EMBL‐EBI (Cambridge, United Kingdom)

In this presentation I will discuss the UniProt KnowledgeBase which has been providing data for protein sequence and function for over 30 years. I will discuss how we are dealing with the deluge of sequence data to organise it in ways to make our users lives easier and more scalable. I will show with examples how our expert curators gather knowledge from the literature, evaluate it and add value through structuring it with a variety of ontologies. Does expert curation scale with the vast biomedical literature? I will address this important point, firstly through recent experiments to estimate the fraction of the literature we curate and secondly, through automated systems to annotate proteins that have never been experimentally characterised. Finally I will discuss how UniProt helps to connect and integrate the worldwide community of biological resources and their data

2. Bioinformatics: 9. Evolution

ABS046

EVOLUTIONARY ANALYSIS OF ROSSMANN‐LIKE FOLD PROTEINS

Kirill Medvedev ¹, Lisa Kinch², Nick Grishin³

¹Department of Biophysics, University of Texas Southwestern Medical Center, Dallas, Texas, United States of America (Dallas, United States); ²Howard Hughes Medical Institute, University of Texas Southwestern Medical Center, Dallas, Texas, United States of America (Dallas, United States); ³1. Departments of Biophysics and Biochemistry, University of Texas Southwestern Medical Center, Dallas, Texas, United States of America. 2. Howard Hughes Medical Institute, University of Texas Southwestern Medical Center, Dallas, Texas, United States of America (Dallas, United States)

By now, genomes of more than 25,000 species have been sequenced and more than 130,000 protein structures have been determined. However, the majority of them remain to be investigated and their function studied. We analyzed the most numerous group of enzymes that constitute about 20% of all known protein structures, namely the Rossmann‐like fold. The classical Rossmann fold, also known as a doubly‐wound three layer/sandwich, consists of two—units (321456 topology) that form a single parallel sheet flanked by ‐helices on both sides and contain a characteristic crossover between strands 3 and 4. We defined its core minimal Rossmann‐like motif (RLM) unit of three ‐strands flanked by two ‐helices and found all known protein structures containing the RLM. We show that RLM enzymes function predominantly in metabolism, covering 38% of reference metabolic pathways. We find that closely related RLM enzyme families can catalyze different reaction chemistries using similar folds. Alternatively, different RLM folds can converge on catalyzing the same reactions. We showed that RLM enzymes utilize ligands from 20 chemical superclasses of organic and inorganic compounds. Homologous RLM domains can exhibit diverging active sites that accommodate alternate ligands, but with similar binding modes. The Rossmann fold is considered one of the most ancient folds, utilizing iron‐sulfur clusters as cofactors and being the part of ancient energy metabolism, the Wood‐Ljungdahl pathway, used by LUCA. Our data suggests that the top three disease categories with mutations in RLM proteins are diseases of endocrine system, nervous system and developmental anomalies.

ABS047

CRYO‐EM STRUCTURE OF THE APO FORM OF HUMAN PRMT5:MEP50 COMPLEX

Wei Zhou ¹, Gaya Yadav², Qiu‐Xing Jiang², Chenglong Li³

¹Department of Biochemistry and Molecular Biology, College of Medicine, University of Florida, USA (Gainesville, United States); ²Department of Microbiology and Cell Science, University of Florida, USA (Gainesville, United States); ³Department of Medicinal Chemistry, College of Pharmacy, University of Florida, USA (Gainesville, United States)

Protein arginine methyltransferase 5 (PRMT5) is an important epigenetic enzyme involved in cancer development. It carries out symmetric arginine dimethylation on histone and non‐histone proteins by utilizing S‐Adenosyl‐L‐methionine (SAM) as its cofactor. Overexpression of PRMT5 in a variety of cancers, especially in methylthioadenosine phosphorylase (MTAP)‐deleted cancers, makes it a compelling target for anticancer therapeutics. PRMT5 was found to bind with a regulatory partner methylosome associated protein 50 (MEP50) in a 1:1 ratio both in vitro and in vivo. Four copies of the PRMT5:MEP50 heterodimer formed into an octamer to be functional in catalysis. So far, all crystal structures of human PRMT5:MEP50 have been resolved in complex with its cofactor SAM or a cofactor analog. Here, we report an apo structure of the human PRMT5:MEP50 complex determined by single particle cryo‐electron microscopy (cryo‐EM) at a resolution of 3.4 å. The overall assembly of the apo form is in agreement with previously reported crystal structures, while a critical loop at the active site is missing in the apo structure. Our new finding suggests that the loop region in the active site is intrinsically flexible and cofactor binding at the active site induces a conformational change, which explains known difficulty in obtaining crystal structures of the apo human PRMT5:MEP50. Structural comparison of the apo human PRMT5 to structures in different binding states gives us a better understanding of the catalytic mechanism of Type II PRMTs and provides meaningful structural insights in design and development of small molecule inhibitors against PRMT5.

5. Computational modeling/simulation: 11. Intrinsically disordered proteins

ABS048

OPTIMIZED MOLECULAR DYNAMICS FORCE FIELD REVEALS ATOMISTIC PATHWAYS OF SPONTANEOUS DISORDER‐TO‐ORDER PEPTIDE‐PROTEIN BINDING

Lei Yu ¹, Da‐Wei Li², Rafael Brüschweiler³

¹Department of Chemistry and Biochemistry, The Ohio State University (Columbus, United States); ²Campus Chemical Instrumental Center, The Ohio State University (Columbus, United States); ³Department of Chemistry and Biochemistry, Campus Chemical Instrumental Center, Department of Biological Chemistry and Pharmacology, The Ohio State University (Columbus, United States)

Intrinsically disordered proteins (IDPs), which exist in fully or partially unfolded forms under native conditions, are widely encoded in human genes. IDPs play crucial biological roles, such as in molecular recognition and post‐translational protein modification, where their specific behavior upon interactions with other proteins dictates their function. However, current protein force fields used for molecular dynamics (MD) simulations are inadequate to capture such behavior in a quantitative manner. To better reproduce experimental structural propensities and dynamics of IDPs, we optimized our prior protein force field in a residue‐specific manner based on dihedral angle distributions taken from an experimental coil library. Various nuclear magnetic resonance (NMR) parameters were then back‐calculated from MD trajectories and compared with experimental NMR data. Results showed substantial improvement in the agreement of scalar J‐coupling constants for IDPs, such as alpha‐synuclein, as well as NMR S2 order parameters of flexible loop regions of globular proteins that probe dynamics on the psns timescale. Globular proteins, such as ubiquitin, remained folded in the simulations and well reproduced experimental residual dipolar coupling data. Furthermore, our optimized force field, which is capable of modelling both structured and unstructured protein regions, revealed pathways for the disorder‐to‐order transition in the spontaneous binding between p53TAD1 and MDM2. In sum, this new protein force field provides an increasingly realistic ensemble view of IDPs and flexible regions of folded proteins to help decipher their function.

16. Protein interactions and assemblies: 26. Other

ABS049

EFFECT OF ETHANOL AND PROTEIN CONTENT ON THE GELATION OF ALMOND, LENTIL, AND PEA PROTEINS

Nahla kreidly¹, Nahla kreidly ¹, Graciela Padua², Hakime Yavuz³

¹University of illinois at urbana champaign (champaign, United States); ²Research Professor (Champaign, United States); ³Masters Student (Champaign, United States)

Plant protein based gels are currently of high interest due to consumers preference over the use of animal proteins. Protein gelation is often achieved by heat treatment or the addition of gelling agents. However, heat treatment may damage nutrients, while the addition of gel inducing agents may be objectionable in certain foods. In this study, we investigated the effect of a novel ethanol‐induced gelation process that is carried out at room temperature.

Dynamic rheological measurements using oscillation tests were employed to characterize the viscoelastic properties of ethanol‐induced gels from almond, lentil, and pea proteins and to investigate the effect of protein content (10‐20% w/v) and ethanol (30‐80% v/v) on the gelling ability of those plant proteins in binary solvent systems of ethanol‐water. All proteins formed instantaneous gels, with a storage modulus (G) greater than the loss modulus (G), immediately upon contact with ethanol. The stiffest gels were observed at 20% protein content in 40% ethanol for all proteins. The microstructure of protein gels dried at room temperature was examined by scanning electron microscopy (SEM) and atomic force microscopy (AFM). Dried protein gels were porous, in contrast with the granular structure of untreated proteins, which revealed a protein network structure potentially able to hold oil droplets and prevent their phase separation. Vegetable protein networks are promising in the food industry as high nutritional value structural components and stabilizers for vitamins and bioactive compounds.

8. Enzymology: 13. Metabolic engineering/energy applications

ABS052

FUNCTIONAL ANALYSIS OF THERMOSTABLE PQQ‐DEPENDENT GLUCOSE DEHYDROGENASE

Tsutomu Mikawa ¹, Ako Kagawa¹, Kayo Kitaura¹, Takanori Kigawa¹

¹RIKEN Center for Biosystems Dynamics Research (Yokohama, Japan)

Pyrroloquinoline quinone‐dependent glucose dehydrogenases (PQQ‐GDHs) extract two electrons when they oxidize D‐glucose to D‐glucono‐1,5‐lactone. Since PQQ‐GDHs have very high reaction rate, they are often used for the negative electrode of biofuel cells and the glucose sensor for monitoring blood glucose level. At present, PQQ‐GDHs derived from mesophilic bacteria are usually used for these applications. However, more stable PQQ‐GDHs are considered useful for them. In this study, we prepared PQQ‐GDH derived from a thermophilic bacterium, Thermus Thermophilus HB8 (ttGDH), as a possible candidate and analyzed its function.

We constructed the expression vector for ttGDH, expressed ttGDH in Escherichia coli and succeeded in obtaining large amount of ttGDH as an apoenzyme. Since PQQ‐GDHs are known to exist in the periplasmic space, we analyzed N‐terminal sequence of the obtained ttGDH and confirmed that the N‐terminal portion was appropriately removed. Then, an excess amount of coenzyme PQQ was added to the apo‐ttGDH and incubated for an hour at 70, the optimum growth temperature of T. thermophiles. Thus, we succeeded in obtaining active ttGDH (holo‐ttGDH). Meanwhile, active ttGDH was not obtained when the incubation was carried out on ice or at room temperature. holo‐ttGDH showed higher activity as temperature increased and kept its activity even over 80. The Tm value of holo‐ttGDH determined by circular dichroism spectroscopy was more than 95, whereas usual PQQ‐GDHs began to degenerate around 45. These results indicate that holo‐ttGDH is very stable.

Finally, we used holo‐ttGDH for a biofuel cell and confirmed that the biofuel cell works stably at 60.

1. Amyloid and aggregation: 2. Bioinformatics

ABS053

TANDEM DOMAIN SWAPPING AND THE LINK TO PROTEIN AGGREGATION

Aleix Lafita ¹, Pengfei Tian², Robert Best², Alex Bateman¹

¹European Bioinformatics Institute EMBL‐EBI, Wellcome Genome Campus, Hinxton, Cambridge, UK (Hinxton, United Kingdom); ²Laboratory of Chemical Physics, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD, USA (Bethesda, United States)

Recent experiments with proteins formed of tandem identical domains have measured increased rates of misfolding and aggregation. Further biophysical characterization has identified the presence of metastable conformations involving swapping of secondary structure elements between adjacent domains, named tandem domain swaps, as one of the possible determinants of protein aggregation.

In order to study tandem domain swapping, we have developed a computational tool named TAndem DOmain Swap Stability predictor (TADOSS). TADOSS is based on a simple approximation observed in previous molecular dynamics simulations to estimate the free energy difference between native and domain swapped conformations. The method consists in three steps illustrated in the figure below: 1) calculate energy of native residue interactions of the domain, 2) calculate the energy contributions of forming a hinge loop and connecting the domain termini with a linker, and 3) build a free‐energy profile along the domain sequence by systematically evaluating all possible hinge loop positions, where higher free energy differences correspond to more stable domain swap conformations. Hence, TADOSS is able to discriminate domains susceptible to domain swapping and to identify structural regions with high propensity to form hinge loops.

Our efforts are now focused on identifying general sequence and structural determinants of tandem domain swapping and their implication in protein aggregation from large scale TADOSS predictions. TADOSS is open‐source and freely distributed under an MIT software license, hosted on GitHub (https://github.com/lafita/tadoss).

21. Structure (x‐ray/NMR/EM): 15. Peptides

ABS054

TWO‐STEP BAK ACTIVATION INITIATES MITOCHONDRIAL APOPTOSIS

Geetika Singh ¹, Siva Vaithiyalingam¹, Jaeki Min¹, Brett Waddell¹, Cristina Guibao¹, Dan McNamara¹, Zoran Rankovic¹, Seetharaman Jayaraman¹

¹St. Jude Children's Research Hospital (Memphis, United States)

The B‐cell lymphoma 2 (BCL‐2) family proteins regulate mitochondrial apoptosis. This pathway is triggered when the BCL‐2 family effector protein BCL2 antagonist/killer 1 (BAK) porates mitochondria to release cytochrome c and initiate the caspase cascade leading to cell demise. Previously our lab has shown that BH3 interacting‐domain death agonist (BID) activates BAK through binding of BID hydrophobic residues (H0‐H5) to six pockets (P0‐P5) of BAK hydrophobic groove. Here we probe the mechanism of BAK activation with BID peptides engineered to incorporate larger amino acids at one or more of the H0‐H5 positions. We identified peptides that have higher affinity for BAK than WT BID peptides and do not compromise direct BAK activation at low concentrations. At high concentrations these peptides inhibit BAK activation supporting the putative hit‐and‐run mechanism. We obtained high‐resolution crystal structures of these peptides in complex with BAK, which provided molecular insights into BAK activation. In addition, we demonstrate that BAK auto‐activates through a domino mechanism. We identified mutants that block auto‐activation but not direct activation by BID. Our work redefines the mechanism of BAK activation in apoptosis.

17. Proteins in cells: 7. Dynamics and allostery

ABS055

BINDING AND MOLECULAR DYNAMICS SIMULATION STUDIES OF BRYOSTATIN 1 AND MUNC13‐1 C1 DOMAIN INTERACTION IN THE PRESENCE OF PHOSPHOLIPID

Youngki You ¹, Francisco Blanco¹, Agnes Czikora², Noemi Kedei², Peter Blumberg², Joydip Das¹

¹College of Pharmacy, University of Houston (Houston, United States); ²National Cancer Institute, NIH (Bethesda, United States)

Munc13‐1 is an essential presynaptic active zone protein involved in synaptic vesicle priming in neurons. Munc13‐1 contains a C1 domain where the activators DAG/phorbol ester bind. Bryostatin 1, a macrocyclic lactone derived from marine sources has been implicated in neuroprotection and possesses the therapeutic potential to treat Alzheimers disease. It activates protein kinase C (PKC) by binding to its C1 domain. Here we studied the interaction between Munc13‐1 C1 and bryostatin 1 using radio‐ligand binding assay and Molecular Dynamics (MD) simulation studies. Bryostatin‐1 binds to the Munc13‐1 C1 domain with high affinity (Ki = 0.45 nM) and translocates Munc13‐1 from cytosol to the plasma membrane in hippocampal‐derived HT22 cells. Molecular docking showed that bryostatin 1 interacts with C1 domain by forming two hydrogen bonds to I590 and R592 and forms an additional hydrogen bond during 100 ns MD simulation in presence of the lipid membrane. The protein‐ligand complex maintained its position in the membrane by interacting with it during the simulation. Whereas the RMSD plot of the C1 domain with bryostatin 1 remains steady at around 2.5 å, the phorbol ester bound C1 domain shows high fluctuation in the range of 2 to 7 å during 100 ns. In conclusion, the C1 domain maintained its interaction to bryostatin 1 and the complex kept its structural stability by embedding deeply into the lipid membrane. This study is significant for designing Munc13‐1 inhibitors for CNS disorders.

6. Design/engineering: 10. Folding

ABS056

OBSERVATIONS OF AN UNFOLDING INTERMEDIATE DURING THE THERMAL UNFOLDING OF THE PROTEINASE K LIKE SERINE PROTEINASE VPR

Kristinn Ragnar Óskarsson ¹, Magnús Már Kristjánsson¹

¹University of Iceland (Reykjavík, Iceland)

Unfolding processes of proteins and identification of initiation sites of unfolding and unfolding intermediates are a rich source of important information that can be utilized in the rational design of proteins and for protein stabilization approaches. Here we demonstrate the possible identification of an unfolding intermediate during the thermal unfolding of VPR a cold active proteinase K‐like serine proteinase. By modification of residues near the N‐terminus of VPR, the stability of the structure was modulated to produce variants of differing stabilities. With these modifications inserted by single point mutations and by varying conditions, such as pH and calcium concentration, an unfolding intermediate was identified by differential scanning calorimetry (DSC) and circular dichroism (CD). This intermediate apparently has a low secondary structure content but may be prevented from fully unfolding by interactions at the N‐terminus and one of its calcium binding sites. These results give insight into the protein unfolding pathway of the proteinase and how VPR is kept in its kinetically stable and active form in nature.

1. Amyloid and aggregation: 4. Chemical biology

ABS057

CHARACTERIZATION AND INHIBITION OF INSULIN AMYLOID FORMATION AT PHYSIOLOGICAL PH

Sinem Apaydın ¹, Chris D. Tran², Anne Kokel², Béla Török², Marianna Török²

¹University of Massachusetts Boston, College of Science and Mathematics, Department of Chemistry and Integrative Biosciences (IB) Program (Boston, United States); ²University of Massachusetts Boston, Department of Chemistry (Boston, United States)

Protein aggregation in ‐sheet‐rich amyloid fibrils is associated with more than 40 human diseases, such as neurodegenerative diseases and diabetes. Type II diabetes doubles the risk for developing Alzheimers disease and a link was suggested between the oligomeric states of insulin and Alzheimers amyloid‐beta (A) peptide. Insulin amyloidosis is also found to occur at insulin injection sites.

Insulin is quite prone to form fibrils under acidic conditions and at high temperatures in vitro. Our goal is to optimize growing conditions at physiological pH for reproducible production of insulin fibrils and compare the specificity of small molecule inhibitors for the self‐assembly of A and insulin, respectively. Racemic trifluoromethyl‐indolyl‐acetic acid ethyl ester derivatives, which inhibited amyloid formation of A in our earlier studies, are tested under optimized physiological conditions for their potential inhibitory activities in the amyloid formation of insulin. The effects of these organofluorine inhibitors are investigated by Thioflavin‐T Fluorescence Spectroscopy, Atomic Force Microscopy, Circular Dichroism, and High Resolution Mass Spectrometry.

Our results indicate that the morphology of fibrils grown at physiological pH is different as compared to that of acidic fibrils. We have also observed consistency of the fibrils within respective parallel experiments in terms of structural similarity at both neutral and acidic pH. The addition of the A amyloid inhibitor compounds has significant effect on the self‐assembly of insulin, as well. HR‐ESI‐MS data indicate that the halogen substituents in the 5‐position of indole strengthens the small molecule‐peptide complex formation commonly favoring 1:2 stoichiometry.

18. Proteomics: 8. Enzymology

ABS059

EXPRESSION AND CHARACTERIZATION OF A LIPID PEROXIDASE FROM NITROSOMONAS EUROPEAE STRUCTURALLY RELATED TO PROSTAGLANDIN H2 SYNTHASE

Alecia Cunniff¹, Barry Selinsky ¹, Rebecca Skaf², Virginia Butchy²

¹Villanova University (Villanova, United States); ²Chemistry Department, Villanova University (Villanova, United States)

Prostaglandin‐endoperoxide synthase (PGHS; EC 1.14.99.1) catalyzes the two step synthesis of prostaglandin H2, an important precursor in the synthesis of prostaglandins and eicosanoids. The first step of the reaction involves the addition of two molecules of oxygen to arachidonic acid to form prostaglandin G2. In the second step, the C15 hydroperoxide in prostaglandin G2 is reduced to an alcohol, generating prostaglandin H2. PGHS has been isolated and characterized from many different eukaryotes (both vertebrates and invertebrates) and has recently been expressed from an open reading frame in the red algae Gracilaria vermiculophylla. A number of putative PGHS orthologs have been identified in bacterial genomes. In this study, a putative PGHS protein from Nitrosomosas europeae has been cloned, expressed, and characterized. Soluble Nitrosomonas protein was isolated to greater than 90% purity exploiting an N‐terminal his‐6 tag. The expressed protein exists in solution as an oligomer, likely a tetramer, and binds heme at a ratio of approximately 0.40 heme molecules per protein monomer. The expressed protein is incapable of adding oxygen to unsaturated lipid substrates, but is a capable peroxidase using both hydrogen peroxide and 15(S)‐hydroperoxy eicosapentaenoic acid as substrates. This result contrasts experiments using a similar protein previously expressed from the genome of Nostoc punctiforme, which exhibits lipoxygenase, but not peroxidase, activity. Structural modeling of the Nostoc and Nitrosomonas proteins yields superimposable structures. Mutagenesis studies of the Nostoc protein are in progress to generate peroxidase activity in the protein product.

2. Bioinformatics: 6. Design/engineering

ABS060

MAPPING STRUCTURE AND INTERACTION IN BETA TURNS

Nicholas Newell¹, Nicholas Newell ¹

¹Newell (Reading, United States)

Beta turns constitute more than 20% of all residues in proteins and play crucial roles in structure and function. They are commonly classified by dihedral angles into a small set of types that provides only a low‐resolution picture of turn backbone geometries, and more than a quarter of turns remain unclassified. Furthermore, the systematic treatment of side‐chain motifs in turns has been limited to the tabulation of single‐position amino acid propensities supplemented by visual surveys and counts, and the interactions between turns and the structure in their N‐ and C‐terminal neighborhoods have not been systematically characterized.

In this work, a three‐stage, cartesian‐space clustering algorithm is applied, first to generate a fine‐scale partitioning and 3D mapping of the backbone distribution of beta turns, then to map, for each backbone geometry, the distributions of side‐chain and rotamer structures for motifs involving one, two, or three amino acids in the turns and their immediate N‐ and C‐terminal neighborhoods, and finally to map the important N‐ and C‐terminal structural contexts of each motif in each backbone geometry.

This work demonstrates that the combination of turn backbone, side‐chain, and context clustering, along with statistical motif detection and 3D conformational mapping, is a powerful tool for structural analyses. The results expand the existing picture of beta turns by providing a comprehensive, unified, and high‐precision treatment of the backbone and side‐chain structure of all beta turns, including associated interactions and contexts, and should prove useful in protein design, structure prediction, and in assessing the structural consequences of disease‐associated mutations.

21. Structure (x‐ray/NMR/EM): 26. Other

ABS061

NORTHEASTERN COLLABORATIVE ACCESS TEAM (NE‐CAT) CRYSTALLOGRAPHY BEAM LINES FOR CHALLENGING STRUCTURAL BIOLOGY RESEARCH

Igor Kourinov ¹, Malcolm Capel¹, Surajit Banerjee¹, Ed Lynch¹, Frank Murphy¹, David Neau¹, Kay Perry¹, Kanagalaghatta Rajashankar¹, Cynthia Salbego¹, Jonathan Schuermann¹, Narayanasami Sukumar¹, James Withrow¹, Steven E Ealick¹

¹Cornell University (Argonne, United States)

The NorthEastern Collaborative Access Team (NE‐CAT) focuses on the design and operation of synchrotron X‐ray beamlines for the solution of technically challenging structural biology problems and provides an important resource for the national and international research community. Currently NE‐CAT operates two undulator beamlines: 24ID‐C fully tunable in the energy range from 6 to 21keV and 24ID‐E fixed energy at 12662eV (optimized for Se SAD experiments).

Both beamlines are equipped with state‐of‐the‐art instrumentation. MD2 microdiffractometers installed at both beamlines provide very clean beams down to 5 microns in diameter and are capable of visualizing micron‐sized crystals. Large area detectors (including Pilatus‐6MF), not only provide the best diffraction data, but also make possible to resolve large unit cells. Both beamlines are equipped with ALS style automatic sample mounters. Towards improving the diffraction from low‐resolution crystals, NE‐CAT has installed humidity controlled device, in the 24ID‐E hutch. Locally developed software suite RAPD provides data collection strategies, quasi‐real time data integration and scaling and simple automated MR/SAD pipeline through 384 core computing cluster. Users of the beamlines are supported by experienced resident crystallographers.

To meet the needs of technically challenging crystallographic projects, cutting‐edge hardware and software ideas are implemented. A summary of beamline capabilities, technology, scientific highlights, future developments and details of availability will be presented.

The NE‐CAT facility is open to the whole crystallography community via APS General User Program. Detailed descriptions of the beamlines can be found at http://necat.chem.cornell.edu.

NE‐CAT funding is provided by P30 grant from NIGMS and from member institutions.

7. Dynamics and allostery: 8. Enzymology

ABS063

INVESTIGATION OF STRUCTURAL FACTORS CONTROLLING LOOP DYNAMICS IN ACYL PROTEIN THIOESTERASES

R. Jeremy Johnson ¹, Asif Hossain¹

¹Butler University (Indianapolis, United States)

FTT258 is a close bacterial relative of human acyl protein thioesterases (APTs). Human APTs catalyze the depalmitoylation of key signaling proteins attached to the plasma membrane. For FTT258, catalysis was hypothesized to be regulated by the large‐scale movement of an essential dynamic loop, which transitioned the protein between active and inactive states. However, the causes for the movement of this loop are unknown. Herein, I investigated factors that trigger loop movement in APTs like FTT258. In particular, I exposed FTT258 to inhibitors structurally similar to long chain lipids and to membrane‐mimicking surfactants to determine if these factors induce loop movement. Loop movement was detected by measuring the change in intrinsic tryptophan fluorescence of a central tryptophan on the dynamic loop before and after exposure to these conditions. Although inhibitors and substrate mimics induced minor changes in loop dynamics, larger shifts were measured with various membrane‐mimicking surfactants, suggesting that APT loop dynamics are mainly regulated by proximity to the plasma membrane. Ultimately, understanding the factors controlling APT function could provide novel mechanisms for controlling cell signaling.

21. Structure (x‐ray/NMR/EM): 25. Transcription/translation/post‐translational modifications

ABS065

INNER WORKINGS OF A COMPASS: CRYSTAL STRUCTURE OF THE COMPASS H3K4 METHYLTRANSFERASE CATALYTIC MODULE

Peter Hsu ¹, Heng Li¹, Ho‐Tak Lau¹, Calvin Leonen¹, Abhinav Dhall¹, Shao‐en Ong¹, Champak Chatterjee¹, Ning Zheng¹

¹University of Washington (Seattle, United States)

The SET1/MLL family of histone methyltransferases are conserved in eukaryotes and regulate transcription by catalyzing histone H3K4 mono‐, di‐, and tri‐methylation. These enzymes form a common five‐subunit catalytic core, whose assembly is critical for their basal and regulated enzymatic activities through unknown mechanisms. Here we present the crystal structure of the intact yeast COMPASS histone methyltransferase catalytic module, consisting of Swd1, Swd3, Bre2, Sdc1, and Set1. The complex is organized by Swd1, whose conserved C‐terminal tail not only nucleates Swd3 and a Bre2‐Sdc1 subcomplex, but also joins Set1 to construct a regulatory pocket next to the catalytic site. This inter‐subunit pocket is targeted by a previously unrecognized enzyme‐modulating motif in Swd3 and features a doorstop‐style mechanism dictating substrate selectivity among SET1/MLL family members. By spatially mapping the functional components of COMPASS, our results provide a structural framework for understanding the multifaceted functions and regulation of the H3K4 methyltransferase family.

17. Proteins in cells: 12. Membrane proteins

ABS066

VASOINHIBIN INHIBITS THROMBIN‐INDUCED ANGIOGENESIS, VASOPERMEABILITY, PLATELET AGGREGATION, AND CANCER INVASION

Juan Pablo Robles ¹, Magdalena Zamora¹, Gonzalo Martínez de la Escalera¹, Carmen Clapp¹

¹Instituto de Neurobiología, UNAM (Universidad Nacional Autónoma de México) (Querétaro, Mexico)

Besides its role in hemostasis, thrombin is a potent inducer of angiogenesis, vasopermeability, tumor growth, and metastasis. Thrombin signals through protease activated receptors (PARs), a family of G‐protein coupled receptors, that signal through G‐proteins (G12/13, Gq11, Gi, and G) to activate several pathways (RhoGEFs‐Rho‐Rac, PLC‐IP3‐Ca2+,PLC‐DAG‐PKC‐Raf1‐MAPK, and PI3K‐Akt) that lead to increased mobility, adhesion, survival, and proliferation of endothelial cells, platelets, myocytes, epithelial cells, and neurons. Vasoinhibin is a negative regulator of angiogenesis and vasopermeability that acts on endothelial cells to antagonize signaling pathways (Ras‐Tiam1‐Rac‐Pac1, PLC‐IP3‐Ca2+, Ras‐Raf1‐MAPK; and PI3K‐Akt) known to be activated by thrombin. Here, we evaluated whether vasoinhibin inhibits thrombin‐induced biological effects. We found that vasoinhibin inhibited thrombin‐induced angiogenesis, assessed by the proliferation, migration, and survival of bovine pulmonary artery endothelial cells (CPAE) in culture. Furthermore, vasoinhibin blocked thrombin‐induced vasopermeability revealed by the inhibiton of the transient decrease in endothelial cell monolayer resistance in response to thrombin. Finally, we found that vasoinhibin antagonized thrombin effects on non‐endothelial cells. Vasoinhibin prevented thrombin‐induced platelet aggregation in a test‐tube and inhibited the in vitro proliferation, migration, invasion, and viability of breast and prostate cancer cells stimulated by thrombin. In summary, vasoinhibin can counteract thrombin effects, implying that vasoinhibin may regulate tissue repair and hemostasis, and protect against tumor growth and metastasis in both, antiangiogenesis‐dependent and ‐independent manners. Furthermore, vasoinhibin represent one of the few inhibitors of PARs and current studies are addressing this cellular signaling pathway.

Research was supported by CONACYT grant 289568 and A1‐S‐9620.

16. Protein interactions and assemblies: 17. Proteins in cells

ABS068

PHYSIOLOGICALLY‐RELEVANT CROWDING EFFECTS ON THE SH3‐SON OF SEVENLESS INTERACTION

Samantha Stadmiller ¹, Jhoan Sebastian Aguilar², Gary Pielak²

¹University of North Carolina at Chapel Hill Chemistry Department (Chapel Hill, United States); ²University of North Carolina at Chapel Hill, Chemistry Department (Chapel Hill, United States)

Nearly all biological processes, including tightly regulated protein‐protein interactions implicated in cell signaling, occur inside living cells where the concentration of macromolecules can exceed 300 g/L. One such interaction is between a globular 7‐kDa SH3 domain and a 25‐kDa intrinsically disordered region of Son of Sevenless (SOS). Despite its importance as a mediator in the mitogen‐activated protein kinase signaling pathway present in all eukaryotic cells, most biophysical characterizations of this complex are performed in dilute buffered solutions where cosolute concentrations rarely exceed 10 g/L. We are investigating the effects of crowding, or high concentrations of physiologically‐relevant cosolutes, on the equilibrium thermodynamics and kinetics of binding between SH3 and SOS. The SH3 domain is labeled with a fluorine atom on its sole tryptophan, allowing the interaction to be monitored with 19F NMR. Subsequent NMR lineshape analysis allows quantification of both rate constants, and hence the equilibrium constant. Two systems are used. The first system comprises the SH3 domain and a SOS‐derived peptide. Experiments using a range of biologically‐relevant molecules reveal that the kinetics are affected more than the stability of the complex. The origins of these effects are likely increased viscosity and electrostatic interactions with crowder molecules. The second system is that between SH3 and the full‐length disordered region of SOS. Crowding effects are expected to be larger in magnitude than those observed for the peptide studies based on hypotheses from traditional crowding theory. These experiments have immediate implications for understanding the effects of the cellular environment on protein‐protein interactions.

19. Proteostasis and quality control: 24. Therapeutics and antibodies

ABS069

THE ROLE OF KLHDC2 IN RECOGNIZING DIGLYCINE C‐END DEGRON AND ITS THERAPEUTIC POTENTIAL

Domnita Valeria Rusnac ¹, Daniele Canzani², Hsiu‐Chuan Lin³, Karena Tien⁴, Thomas R. Hinds⁴, Ashley F. Tsue⁴, Hsueh‐Chi S. Yen³, Matthew F. Bush², Jie Fan⁵, Ning Zheng⁴

¹University of Washington (Seattle, United States); ²Department of Chemistry, University of Washington (Seattle, United States); ³Institute of Molecular Biology, Academia Sinica and Genome and Systems Biology Degree Program, National Taiwan University and Academia Sinica (Taipei, Taiwan); ⁴Howard Hughes Medical Institute, Department of Pharmacology, University of Washington (Seattle, United States); ⁵Accutar Biotech (New York, United States)

Recent studies have shown that the stability of select proteins or polypeptides is controlled by their extreme C‐terminal sequences, which are also known as C‐end degrons. These motifs are specifically recognized by the understudied BC‐Box substrate receptors, which are part of the multi‐subunit CUL2‐RING ubiquitin ligase (CRL2) family. I have determined the structure of KLHDC2, an evolutionarily conserved BC‐Box protein, in complex with multiple C‐terminal degron peptides, which are recognized by the E3 with nanomolar affinities. These 8‐amino acid C‐terminal degrons contain a characteristic diglycine motif at their extreme C‐terminus and degenerate upstream sequences. In my presentation, I will not only summarize the structural mechanism by which KLHDC2 recognizes these degenerate C‐terminal degrons, but also highlight our recent data on direct identification of cellular C‐terminal degron‐like peptides copurified with BC‐box proteins by native mass spectrometry analysis. Due to its cellular localization, tissue distribution and structural properties, KLHDC2 has emerged as an attractive E3 for functional and therapeutic repurposing. I will present our ongoing endeavor of using artificial intelligence‐aided virtual screen to search for small molecules that can dock to the deep degron‐binging pocket of KLHDC2. These compounds hold the promise to serve as a new class of E3‐reprogramming agents for targeted protein degradation.

24. Therapeutics and antibodies: 21. Structure (x‐ray/NMR/EM)

ABS070

TRANSLESION SYNTHESIS PATHWAY AS TARGET FOR ANTI‐CANCER DRUG DESIGN

Dmitry Korzhnev ¹

¹Department of Molecular Biology & Biophysics, University of Connecticut Health Center (Farmington, United States)

Translesion synthesis (TLS) is a cellular pathway that helps rescue replication following DNA damage caused by environmental and endogenous carcinogens. This pathway employs low‐fidelity TLS DNA polymerases that can copy over replication blocking DNA lesions while temporarily leaving them unrepaired, preventing cell death at the expense of increasing rate of mutations. TLS enzymes are important for normal cell survival after DNA damage; however, they also enable cancers to survive genotoxic chemotherapy. Furthermore, TLS contributes to increased mutagenesis in tumors promoting acquired chemoresistance. Owing to its central role in mutagenesis and cell survival after DNA damage, TLS pathway has emerged as a potential target for the development of anti‐cancer agents. This presentation will describe how the detailed structural knowledge of key protein‐protein interactions underlying TLS regulation is beginning to translate into the development of small molecule TLS inhibitors that target these interactions as anti‐cancer therapeutics.

References

1. A Pozhidaeva, Y Pustovalova, S D'Souza, I Bezsonova, GC Walker, DM Korzhnev, Biochemistry, 2012, 51, 5506‐5520

2. Y Pustovalova, MTQ Magalhães, S DSouza, A.A. Rizzo, G Korza, GC Walker, DM Korzhnev, Biochemistry, 2016, 55, 2043‐2053

3. Y Pustovalova, I Bezsonova, DM Korzhnev, FEBS Letters, 2012, 586, 30513056

4. DM Korzhnev, MK Hadden, J. Med. Chem., 2016, 59, 9321‐9336.

5. V.Sail, AA Rizzo, N Chatterjee, RC Dash, Z Ozen, GC Walker, DM Korzhnev, MK Hadden. ACS Chem. Biol., 2017, 12, 1903‐1912.

6. Z Ozen, RC Dash, KR McCarthy, SA Chow, AA Rizzo, DM Korzhnev, MK Hadden, Bioorg Med Chem. 2018, 26, 4301‐4309.

16. Protein interactions and assemblies: 4. Chemical biology

ABS071

INHIBITION OF TRANSCRIPTIONAL ACTIVATOR‐COACTIVATOR PROTEIN‐PROTEIN INTERACTIONS WITH NATURAL PRODUCTS

Meghan Breen ¹, Stephen Joy¹, Matthew Beyersdorf¹, Matthew Henley¹, Samantha De Salle¹, Anna Mapp¹

¹University of Michigan (Ann Arbor, United States)

Dysregulated transcription is observed in nearly all human diseases, and the protein‐protein interactions (PPI) regulating transcriptional activation have emerged as exciting opportunities for the development of new therapeutics. For example, the PPI between the transcription factor Myb and the KIX domain of the master coactivator CREB binding protein (CBP) is required for the progression of acute myeloid leukemia, making this a promising target. However, the conformational lability of the KIX domain and the lack of well‐defined pockets pose major challenges for the development of selective small molecule inhibitors. Using a series of fluorescence polarization screens, we identified garcinolic acid as an inhibitor of the CBP KIX domain. Garcinolic acid is the most potent small molecule CBP KIX ligand (Kd = 1 uM) reported to date and shows 80‐fold selectivity for CBP KIX over homologous KIX domains. A suite of NMR studies demonstrated that garcinolic acid inhibits transcription factor binding through orthosteric and allosteric mechanisms. Garcinolic acid also engages its target in leukemia cells as shown through inhibition of co‐immunoprecipitation of CBP by Myb in cell lysate and downregulation of Myb‐dependent genes. These results validate the utility of our multiple screening approach for identifying modulators of conformationally dynamic targets, and garcinolic acid will be a valuable tool for probing critical KIX‐activator interactions.

6. Design/engineering: 5. Computational modeling/simulation

ABS072

HARNESSING NEW AND EMERGING COMPUTATIONAL TECHNOLOGIES TO ADVANCE DESIGN OF FOLDING PROTEIN‐LIKE HETEROPOLYMERS

Vikram Mulligan ¹

¹Center for Computational Biology, Flatiron Instititue (New York, United States)

Proteins possess the remarkable ability to fold spontaneously into precise and intricate three‐dimensional structures that are responsible for their functions, with the amino acid sequence of a given protein uniquely defining its fold and function. Computational protein design and structure prediction tools, such as the Rosetta software suite, have enabled the creation of new protein folds, new binding proteins, new enzymes, and new high‐order macromolecular assemblies. However, most past protein design efforts have relied on knowledge gleaned from databases of known protein structures. The ultimate test of our understanding of the mechanisms of protein folding is the creation of new, non‐protein heteropolymers that are able to fold as proteins do. In this presentation, I will describe the generalization of Rosettas design and structure prediction methods, using unbiased sampling approaches that are not dependent on prior structural knowledge, in order to permit the creation of exotic, synthetic folding heteropolymers. I will present experimentally‐solved structures of designed molecules with secondary structures, tertiary structures, and symmetries inaccessible to natural proteins, as well as druglike molecules that fold into shapes perfectly complementary to binding pockets in proteins of therapeutic interest. Finally, I will describe ongoing work to harness machine learning approaches to enable prior knowledge‐independent structure prediction of arbitrary heteropolymers, and new approaches using current‐generation quantum computers to advance heteropolymer design.

1. Amyloid and aggregation: 16. Protein interactions and assemblies

ABS073

INFLUENCE OF POROUS MATERIALS ON AMYLOID‐BETA PROTEIN AGGREGATION

Michael Lucas ¹, Benjamin Keitz¹

¹University of Texas at Austin (Austin, United States)

The aggregation of amyloidogenic proteins is associated with a variety of neurodegenerative diseases. In particular, amyloid fibrils and oligomers formed from Amyloid‐Beta (A‐Beta), play a key role in the progression of Alzheimers Disease. However, the aggregation pathway of A‐Beta is extremely complex, resulting in a structurally diverse population of fibrils, oligomers, and other aggregates. A better understanding of this mechanism in addition to the ability to direct amyloid formation is necessary for effective therapeutic design. Exogenous materials, such as hydrophobic surfaces, nanoparticles, and polymers, have been shown to influence amyloid formation in vitro and in vivo. To better understand how exogenous materials could direct amyloid formation, A‐Beta(1‐40) was exposed to porous materials, specifically zeolites and mesoporous silica, allowing for us to study the combined effects of surface chemistry and porosity on the aggregation of A‐Beta. Through Thioflavin T fluorescence, we found that zeolite Y containing Na+, Mg2+, and Fe3+ ions increased the primary nucleation rate of A‐Beta. On the other hand, CuY and ZnY accelerated the formation of oligomeric intermediates, but overall inhibited the formation of fibrils, as evidenced by Western Blot and TEM. Additionally, seeding studies confirmed that ZnY and CuY oligomers were kinetically trapped in an inactive state. These findings were supported through measurements of binding affinity, in which ZnY and CuY exhibited higher association constants. We also examined the effect of SBA‐15, which has pores large enough to encapsulate A‐Beta monomers and larger structures. Overall, our results show that porous materials can impact amyloid aggregation and speciation.

3. Chaperones: 10. Folding

ABS074

THE CHAPERONIN TRIC/CCT ASSOCIATES WITH PREFOLDIN THROUGH A CONSERVED ELECTROSTATIC INTERFACE ESSENTIAL FOR CELLULAR PROTEOSTASIS

Daniel Gestaut ¹, Soung Hun Roh², Boxue Ma³, Grigore Pintile³, Lukasz Joachimiak⁴, Alexander Leitner⁵, Thomas Walzotheini⁶, Ruedi Aebersold⁵, Wah Chiu⁷, Judith Frydman⁸

¹Stanford University (Stanford, United States); ²Dept of Biological Science, Seoul National University (Seoul, South Korea); ³Baylor College of Medicine (Houston, United States); ⁴Dept of Biochemistry, UTSouthwestern, North Campus (Dallas, United States); ⁵Institute of Molecular Systems Biology, Dept of Biology, ETH (Zurich, Switzerland); ⁶Institute of Computational Biology, Helmholtz Zentrum München (Neuherberg, Germany); ⁷Dept of Bioengineering, Stanford University (Stanford, United States); ⁸Dept of Biology and Genetics, Stanford University (Stanford, United States)

Maintaining proteostasis in eukaryotic protein folding involves cooperation of distinct chaperone systems. To understand how the essential ring‐shaped chaperonin TRiC/CCT cooperates with the chaperone Prefoldin/GIMc (PFD) we integrated cryoEM, crosslinking‐mass‐spectrometry and biochemical and cellular approaches to elucidate the structural and functional interplay between TRiC/CCT and PFD. We find these hetero‐oligomeric chaperones associate in a defined architecture, through a conserved interface of electrostatic contacts that serves as a pivot point for a TRiC‐PFD conformational cycle. PFD alternates between an open latched conformation and a closed engaged conformation that aligns the PFD‐TRiC substrate binding chambers. PFD can act after TRiC bound its substrates to enhance the rate and yield of the folding reaction, suppressing non‐productive reaction cycles. Disrupting the TRiC‐PFD interaction in vivo is strongly deleterious, leading to accumulation of amyloid aggregates. The supra‐chaperone assembly formed by PFD and TRiC is essential to prevent toxic conformations and ensure effective cellular proteostasis.

ABS075

ENGINEERING HALOGEN BONDS TO AFFECT PROTEIN STABILITY, ACTIVITY, AND RECOGNITION

P. Shing Ho ¹, Rhianon Kay Hartje¹, Anna‐Carin Carlsson¹

¹Colorado State University (Fort Collins, United States)

Halogen bonds are non‐covalent interactions that are being increasingly applied as substitutes for classical hydrogen bonds in materials, solid‐state, and medicinal chemistries (1). Our laboratory has been applying this distinct class of interaction to control the structure and stability of biomolecular systems (2). For proteins in particular, we have recently shown that introduction of a chlorotyrosine (a marker for immuno‐responses and aging) increases the thermal stability and the activity at elevated temperatures of the model T4 lysozyme enzyme (3). The effects come from an intramolecular hydrogen bond increasing the halogen bonding potential of the chlorine substituent, resulting in a synergistic interaction that we have called a hydrogen bond enhanced halogen bond (or HBeXB for short). Our group is currently engineering halogen bonds or hydrogen bonds to provide unique recognition interfaces to, for example assemble specific heterotrimeric coiled‐coil complexes for molecular design. Finally, we are engineering halogen bonds and HBeXBs to stabilize metastable proteins and, in that way, potentially affecting their functions in cells. The results of the studies provide principles to allow halogen bonds to be exploited for protein engineering and drug/inhibitor design.

1. Cavallo, G., Metrangolo, P., Milani, R., Pilati, T., Priimagi, A., Resnati, G., Terraneo, G. (2016) The Halogen Bond. Chem Rev. 116: 2478‐2601.

2. Scholfield, M.R., Zanden, C.M., Carter, M., Ho, P.S. (2013) Halogen bonding (X‐bonding): a biological perspective. Protein Sci. 22: 139‐152.

3. Carlsson, A.C., Scholfield, M.R., Rowe, R.K., Ford, M.C., Alexander, A.T., Mehl, R.A., Ho, P.S. (2018) Increasing Enzyme Stability and Activity through Hydrogen Bond‐Enhanced Halogen Bonds. Biochemistry, 57: 4135‐4147.

21. Structure (x‐ray/NMR/EM): 24. Therapeutics and antibodies

ABS076

STRUCTURAL AND MECHANISTIC STUDIES OF THE IMMUNE RESPONSE TO THE BLOOD COAGULATION FACTOR VIII C1 DOMAIN

Shaun Peters ¹, Steven Reese², Cris Mitchell³, Joseph Gish⁴

¹Western Washington University (Bellingham, United States); ²Undergraduate (Bellingham, United States); ³Undergraduate (Bellingham, United States); ⁴Graduate Student (Bellingham, United States)

Blood coagulation factor VIII functions as a cofactor in the blood coagulation cascade for proteolytic activation of factor X. During coagulation, factor VIII is proteolytically activated by thrombin and binds to phosphatidylserine of activated platelet surfaces in coordination with activated factor IX to form the intrinsic Xase complex. The C1 and C2 domains of factor VIII are responsible for binding to activated platelet membranes.

During replacement therapy for hemophilia A patients, approximately 30% of severe hemophiliacs develop an inhibitory immune response to the therapeutic factor VIII. These inhibitory antibodies inhibit coagulation activity of the therapeutic protein and basal factor VIII. Several of the inhibitory antibodies recognize epitopes on the C1 and C2 domains of factor VIII. Previous structural and phospholipid binding studies indicated the classical monoclonal anti‐C2 antibody disrupts factor VIII C2 domain phospholipid binding while the nonclassical monoclonal anti‐C2 antibody disrupts the dissociation of factor VIII from von Willebrand factor. No structural or mechanistic data have been collected for the C1 domain.

Here, we report a novel protein expression for the isolated human factor VIII C1 domain and a C1/C2 domain construct. Sedimentation assays for various C1/mAb complexes incubated with unilamellar liposomes were developed to investigate the effect of C1 antibody association on phospholipid binding. Our future studies entail structural characterization of the isolated C1 or C1/C2 domain constructs bound to anti‐C1 domain inhibitory antibodies in order to categorize these immunogenic epitopes and to investigate the phospholipid binding capability of FVIII C2 domain mutants.

24. Therapeutics and antibodies: 16. Protein interactions and assemblies

ABS077

COUPLING OF STABILITY AND SELF‐ASSOCIATION OF A THERAPEUTIC PROTEIN

Natalia Markova¹, Muneera Beach ¹, Matthew McGann¹, Erik Noprdling², Vilhelm Ek², Sergi Kuprin²

¹Malvern Panalytical (Westborough, United States); ²Swedish Orphan Biovitrum AB (Stockholm, Sweden)

Construct 5o is a small 2‐domain recombinant protein candidate drug for which limited conclusions could be drawn based in traditional optical methods in size exclusion chromatography (SEC) elution was dependent on the column load, indicating self‐association. Additional studies using differential scanning calorimetry (DSC) and isothermal titration calorimetry (ITC) were initiated to better understand the behavior and stability of this molecule.

DSC revealed two distinct thermal transitions, Tm1 ~27‐37 °C and Tm2 ~5‐70 °C. Parameters varied with ionic strength of the buffer. The first transition was strongly dependent on protein concentration, while the second showed less response to concentration. The first transition could be coupled to self association of the protein, which is favored at low temperatures and low ionic strength. Twelve other genetic variants were studied with constructs varying at one or more amino acid positions and yielded significantly different DSC thermograms. For all variants, two transitions were identified, and most variability was observed for the first transition. In ITC experiments concentrated protein was diluted in a stepwise manner into corresponding buffer. The data confirmed temperature and buffer dependent oligomerization of the protein, correlating with SEC results.

The results of these experiments indicated variants with maximum structure linked oligomerisation, enabling improved candidate selection and guidance for formulation development. ITC data also provided measure of equilibrium constants for oligomerisation, providing assessment of drug behavior and state in solution.

1. Amyloid and aggregation: 11. Intrinsically disordered proteins

ABS078

THE STRUCTURE OF ALPHA‐SYNUCLEIN SECONDARY NUCLEI IS DOMINATED BY THE SOLUTION CONDITIONS RATHER THAN THE SEED FIBRIL STRAIN

Alessia Peduzzo ¹, Sara Linse², Alexander Büll¹

¹Institute of Physical Biology, Heinrich Heine Universität Düsseldorf (Düsseldorf, Germany); ²Department of Biochemistry and Structural Biology, Centre for Molecular Protein Science, Lund University (Lund, Sweden)

Alpha‐synuclein (a‐syn) is a natively unfolded protein predominantly localized in the presynaptic terminals of neurons. It has been shown that a‐syn fibrils are the major component of abnormal neuronal aggregates known as Lewy bodies, the characteristic hallmark of Parkinson's disease. Amyloid fibrils arise through primary nucleation from soluble monomers, which in the case of a‐syn requires suitable surfaces with an affinity for the proteins, followed by the elongation of the nuclei by monomer addition. Secondary nucleation, on the other hand, involves the binding of monomers onto the surface of the pre‐existing fibrils, leading to the formation of new fibrils. While it is well‐established that the newly added monomer in the process of fibril elongation adopts the conformation of the monomers in the seed, often called templating, it is still unclear under which conditions fibrils formed through secondary nucleation of monomers on the surface of fibrils copy the structure of the parent. Here we show by biochemical and microscopical methods that the secondary nucleation of a‐syn, induced by mildly acidic pH, leads to fibrils that structurally resemble more closely those formed de novo under the same conditions, rather than the seeds. This result has important implications for the mechanistic understanding of the secondary nucleation of amyloid fibrils and its role in the propagation of aggregate pathology in protein misfolding diseases.

19. Proteostasis and quality control: 21. Structure (x‐ray/NMR/EM)

ABS079

CONFORMATIONAL CHANGES RESPONSIBLE FOR ACTIVATION OF PARKIN, AN E3 UBIQUITIN LIGASE INVOLVED IN PARKINSON'S DISEASE

Kalle Gehring ¹

¹McGill University (Montreal, Canada)

Mutations in the ubiquitin ligase parkin are responsible for a familial form of Parkinsons disease. Together, a protein kinase PINK1 and parkin constitute a mitochondrial quality control system required for the long‐term health of dopaminergic neurons. Parkin is activated by PINK1 through a two‐step mechanism involving binding of phosphorylated ubiquitin and parkin phosphorylation. We have used X‐ray crystallography and hydrogen‐deuterium exchange mass spectrometry to characterize the large conformational changes responsible for parkin activation (Sauvé et al, Nat Struct Mol Biol, 2018). In the absence of mitochondrial damage, parkin is cytosolic and adopts an autoinhibited confirmation (Trempe et al, Science, 2013). Upon activation, the parkin structure opens to expose the E2‐binding and catalytic sites (Sauvé et al, EMBO J, 2015). The structure explains the presence of internal linkers that allow large domain movements in parkin. Ubiquitination and mitochondrial recruitment assays reveal the importance of the different protein‐protein interactions. The results rationalize parkin's specificity as a ubiquitin ligase and previously unexplained Parkinsons disease mutations.

16. Protein interactions and assemblies: 12. Membrane proteins

ABS080

CELLULAR SIGNALING THROUGH CYSTEINE PHOSPHORYLATION

Kalle Gehring ¹

¹McGill University (Montreal, Canada)

PRLs are a family of highly oncogenic protein phosphatases that function as pseudophosphatases through direct binding of CNNM magnesium transporters (Saha, S. et al., Science 294, 1343‐6, 2001; Funato, Y. et al., J Clin Invest 124, 5398‐410. 2014). PRLs inhibit CNNM‐dependent magnesium efflux and the PRL‐CNNM interaction is essential for PRL oncogenicity. Structurally, PRLs bear similarity to the dual specificity phosphatases with a conserved, catalytic cysteine that acts as a nucleophile in catalysis (Kozlov, G. et al., J Biol Chem 279, 11882‐9, 2004). The enzymatic cycle generates a transient intermediate phosphocysteine from dephosphorylation of the phosphatase substrate. Unique to the PRL family, this intermediate is extremely long‐lived and along with cysteine oxidation regulates the interaction between PRLs and CNNM magnesium transporters (Zhang, H. et al., Sci Rep 7, 48, 2017). PRLs are endogenously phosphorylated on cysteine in cultured cells and the levels of phosphorylation changes in response to the availability of magnesium (Gulerez, I. et al. EMBO reports 17, 1890‐900, 2016). These studies reveal a new paradigm in cell signaling and are relevant for the design of therapeutics against PRLs and metastatic cancers.

6. Design/engineering: 22. Synthetic biology

ABS081

DE NOVO DESIGN OF BIOACTIVE PROTEIN SWITCHES

ROBERT LANGAN ¹, Scott Boyken¹, Andrew Ng², Jennifer Samson³, Galen Dods², Taylor Nguyen², Alexandra Westbrook², Marc Lajoie¹, Zibo Chen¹, Stephanie Berger¹, Vikram Mulligan¹, John Dueber³, Walter Novak⁴, Hana El‐Samad², David Baker¹

¹University of Washington (SEATTLE, United States); ²UCSF (San Francisco, United States); ³UC Berkeley (Berkeley, United States); ⁴Wabash College (Crawfordsville, United States)

Allosteric regulation of protein function is widespread in biology, but challenging for de novo protein design as it requires explicit design of multiple states with comparable free energies. We explore the possibility of de novo designing switchable protein systems through modulation of competing inter and intra‐molecular interactions. We design a static, five‐helix Cage with a single interface that can interact either intra‐molecularly with a terminal Latch helix or inter‐molecularly with a peptide Key. Encoded on the Latch are functional motifs for binding, degradation, or nuclear export that function only when the Key displaces the Latch from the Cage. We describe orthogonal Cage‐Key systems that function in vitro, in yeast and in mammalian cells with up to 300‐fold activation of function by Key. The design of switchable protein function controlled by induced conformational change is a milestone for de novo protein design and opens up new avenues for synthetic biology and cellular engineering.

16. Protein interactions and assemblies: 4. Chemical biology

ABS082

WANTED: SMALL MOLECULES TO INHIBIT PNT DOMAIN‐MEDIATED POLYMERIZATION OF ETV6 CHIMERIC ONCOPROTEINS

Chloe Gerak ¹, Sophia Cho¹, Mark Okon¹, Richard Sessions², Michel Roberge¹, Lawrence McIntosh¹

¹University of British Columbia (Vancouver, Canada); ²University of Bristol (Bristol, United Kingdom)

ETV6 is a modular transcriptional repressor for which head‐to‐tail polymerization of its PNT (or SAM) domain facilitates cooperative binding to tandem DNA sites by its ETS domain. Chromosomal translocations frequently fuse the ETV6 PNT domain to one of several protein tyrosine kinases. The resulting chimeric oncoproteins undergo ligand‐independent self‐association, autophosphorylation, and aberrant stimulation of downstream signaling pathways leading to a variety of cancers. Accordingly, our research goal is to identify and characterize small molecule inhibitors of ETV6 PNT domain polymerization to prevent the constitutive activation of these oncoproteins. Protein‐protein interactions are challenging to disrupt therapeutically, and thus we are following a multi‐pronged approach for lead compound discovery. We have conducted alanine scanning mutagenesis and amide hydrogen exchange experiments to obtain structural, dynamic and thermodynamic insights into the PNT domain polymerization interfaces. In parallel, we have used in silico virtual docking and molecular dynamic simulations to identify and predict candidate inhibitors that bind these interfaces. We are beginning to test these candidates in vitro using NMR spectroscopy and in vivo using a high‐throughput split‐luciferase protein complementation assay that we developed to detect the disruption of PNT domain self‐association.

16. Protein interactions and assemblies: 12. Membrane proteins

ABS083

MODULATION OF LIGAND AND RECEPTOR STATE IN DDR‐COLLAGEN INTERACTIONS

Gunjan Agarwal ¹

¹Ohio State University (Columbus, United States)

Discoidin Domain Receptors (DDR1 and DDR2) are receptor tyrosine kinases that signal in response to collagen, including collagen type 1. We had previously shown that interactions of DDRs with collagen inhibits collagen fibrillogenesis and alters the fibril structure. The objectives of this study were to evaluate (i) how the oligomeric state of the receptor influences collagen fibrillogenesis and (ii) how collagen impacts the oligomeric state of the receptor and ensuing receptor phosphorylation.

Studies were conducted using purified recombinant DDR1 and DDR2 ectodomain and MC3T3‐E1 cells expressing full‐length DDR2‐GFP or DDR1b‐YFP. The receptor‐ligand interactions were evaluated using solid‐phase binding assays, collagen turbidity measurements, atomic force microscopy, electron microscopy and fluorescence microscopy.

Our results show that the oligomeric form of both DDR1 and DDR2 ECD displayed enhanced binding to collagen and inhibition of fibrillogenesis. However, DDR1 and DDR2 significantly differed in their ability to undergo ligand‐induced clustering. While both the ectodomain and full length DDR1 clustered in response to collagen, DDR2 showed no such effect. However, after prolonged collagen stimulation, both DDR1b‐YFP and DDR2‐GFP formed filamentous structures which co‐localized with collagen fibrils. Receptor phosphorylation was primarily located on the clusters and filamentous structures formed by DDR1b and/or DDR2. Our results uncover key differences and similarities in the clustering abilities and spatial distribution of DDR1b and DDR2 and provide novel insights into activation of these collagen receptors.

7. Dynamics and allostery: 21. Structure (x‐ray/NMR/EM)

ABS085

ACTIVATION DYNAMICS OF USP7 DEUBIQUITINASE

IRINA BEZSONOVA ¹, Gabrielle Valles¹, Dmitry Korzhnev¹

¹UCONN Health (Farmington, United States)

Ubiquitin‐specific protease 7 (USP7) is a deubiquitinating enzyme (DUB) that plays a pivotal role in multiple oncogenic pathways and therefore is a desirable target for new anti‐cancer therapies. It is a cysteine protease that belongs to the USP family of DUBs. USP7 is unique among other USPs in that its active site is catalytically incompetent in the apo‐state and can rearrange into a productive conformation upon substrate binding. Such substrate‐induced structural rearrangements likely play a role of a safety switch that turns the enzyme on only when it is engaged with a specific ubiquitinated substrate.

Although USP7 has been crystallized in both apo‐ and ubiquitin‐bound forms that highlighted the structural differences between the active and inactive protein conformations, these structures represent static snapshots of the enzyme. Whether USP7 samples these conformations in solution and the role USP7 conformational dynamics play in its function remain unknown.

Using the latest developments in Nuclear Magnetic Resonance (NMR) spectroscopy we have observed and characterized the conformational dynamics of deubiquitinating enzyme USP7 in solution. Our data suggest that the USP7 has distinct sites of conformational exchange located in key functional sites of the enzyme.

16. Protein interactions and assemblies: 26. Other

ABS086

SELECT IONIC RESIDUES IN THE C‐TERMINAL DOMAIN OF HUMAN APOLIPOPROTEIN A‐I REGULATE SELF‐ASSOCIATION

John Burdick ¹, Rohin Basi¹, Kaitlyn Burns¹, Paul Weers¹

¹California State University Long Beach (Long Beach, United States)

Human apolipoprotein A‐I (apoA‐I) is the main protein of high‐density lipoprotein. ApoA‐I is comprised of two domains; a highly ordered N‐terminal domain and less structured C‐terminal domain. The C‐terminal domain is responsible for the initiation of lipid binding; a core function of apoA‐I. This domain is also the site of oligomerization and home to six lysine residues at positions 195/206/208 of helix 8 and 226/238/239 of helix 10. Previous studies have found that substitution of all six lysine residues with glutamine resulted in a predominantly monomeric form of the protein. This indicates that ionic bonds play a role in self‐association. To investigate the role each individual lysine plays in self‐association of the C‐terminal domain, site‐directed mutagenesis was implemented to generate an array of mutant variants. Six single and two triple mutant proteins were engineered and expressed in E. coli. Dimethyl suberimidate crosslinking demonstrated a decrease in self‐association of the mutant apoA‐I variants compared to the wild‐type. Based on the SDS‐PAGE crosslinking pattern of the 238 and 239 single mutant and 226/238/239 triple mutant the contribution to self‐association by the helix 10 residues is additive. These results were confirmed using size‐exclusion fast protein liquid chromatography. Both helix 8 and 10 triple mutants exhibited longer elution times and compared to the wild‐type. These results suggest the lysines of helix 10 are critical to self‐association. This also indicates that no single lysine side‐chain is entirely responsible for self‐association; instead the degree of self‐association is cumulatively decreased with removal of each ionic residue.

6. Design/engineering: 22. Synthetic biology

ABS088

COMPUTATIONAL DESIGN AND FUNCTIONALIZATION OF POROUS PROTEINS

Chunfu Xu ¹, Peilong Lu¹, Tamer Gamal El‐Din¹, Xue‐Yuan Pei², Matthew Johnson¹, Atsuko Uyeda³, Matt Bick¹, Michael Luciano⁴, Venu Bandi⁴, Martin Schnermann⁴, Tomoaki Matsuura³, Ben Luisi², William Catterall¹, David Baker¹

¹University of Washington (Seattle, United States); ²University of Cambridge (Cambridge, United Kingdom); ³Osaka University (Osaka, Japan); ⁴National Cancer Institute (Frederick, United States)

The ability to custom design structurally well‐defined transmembrane channels would have broad applications in areas ranging from nanopore sequencing to small molecule filtration and sensing, but remains an outstanding challenge. Here, we report the computational design of protein pores formed by two concentric rings of ‐helices that are stable and mono‐disperse in both water‐soluble and membrane protein forms. Crystal structures of the soluble forms of a 12‐helical pore with an inner diameter at the narrowest constriction of 4 å, and of a 16‐helical pore with a constriction radius of 11 å, are very close to the design models (0.84 and 2.51 å RMSD respectively). Patch clamp electrophysiology experiments show that the transmembrane form of the 12‐helix pore expressed in insect cells allows passage of ions across the membrane with selectivity for potassium over sodium. The transmembrane form of the 16‐helix pore, but not the 12‐helix pore, allows passage of biotinylated Alexa Fluor 488 when incorporated into liposomes using in vitro protein synthesis. The ability to produce structurally well‐defined transmembrane channels both in cells and in vitro opens the door to the creation of designer pores for a wide variety of applications. A custom‐designed 10‐helical bundle is repurposed to bind light‐responsive small molecules, all‐trans retinal and merocyanine retinal, by Schiff‐base bond formation with an activated lysine. The binding of merocyanine retinal turns on bright near‐infrared fluorescence. Incorporation of the designed fluorescence protein into cell membrane and investigation of its membrane voltage sensitivity are currently undergoing.

ABS089

EXPRESSION OF BAND 3, A MEMBRANE PROTEIN

Chun‐Fu Chen¹, Chao‐Lin LIU ¹, Chia‐Rui Shen²

¹Ming Chi University of Technology (New Taipei, Taiwan); ²Department of Medical Biotechnology and Laboratory, Chang Gung University (Taoyuan, Taiwan)

Band 3 is the anion exchanger 1 on the cell membrane of red blood cell. It mediates the exchange of the cellular HCO3‐ with CI‐ in plasma. It is reported Band 3 is the target of T help dependent autoantibody in New Zealand Black mice. The immune response results in spontaneous development of autoimmune hemolytic anemia. Band 3 is subjected in different expression system. However, most of the recombinant Band 3 is the truncated form.

In bacteria expression system, the mass production, time as well as cost saving are the advantages. The folding and modification defects are the main drawback. In the project, the Band 3 is subcloned and expressed in pET system, the bacteria expression system. After induction with IPTG in large scale expression, the harvest was collected and 6M urea was applied during the purification process. Finally, the recombinant Band 3 was obtained and identified with mass spectrometry.

20. Single molecule studies: 26. Other

ABS091

FUNCTIONAL ANALYSIS OF ACA 01, A NOVEL CHEMOKINE‐BINDING TICK EVASIN

Sayeeda Chowdhury ¹, Ram Bhusal²

¹Monash University (Clayton, Australia); ²Infection and Immunity Program, Monash Biomedicine Discovery Institute, and Department of Biochemistry and Molecular Biology, Monash University, Clayton, VIC 3800, Australia. (Clayton, Australia)

Chemokines are mammalian proteins that regulate the migration of leukocytes, a central feature of inflammatory responses. Ticks are hematophagous arachnids that live on mammalian hosts and can transmit many viral and bacterial infections between their hosts. As a strategy to suppress the inflammatory responses of their mammalian hosts and thereby prolong their feeding and residence times, ticks have evolved the ability to produce salivary proteins, known as evasins, which bind to host chemokines, blocking activation of chemokine receptors and preventing leukocyte migration. We have identified evasins from a variety of tick species. ACA‐01 is an evasin found in the tick species Amblyomma cajanennense. An N‐terminal SUMO tag was used for the soluble expression of ACA‐01 in an E. coli expression system. The purified protein was analysed with ESI LC MS to predict the formation of four intramolecular disulfide bonds. The data was further validated by Circular Dichroism to predict secondary structure. Binding data obtained using a competitive fluorescence anisotropy assay, showed that ACA‐01 expressed in E. coli binds to some chemokines with dissociation equilibrium constants (KD) in the range 10‐70 nM. To identify the features of ACA‐01 controlling chemokine‐binding affinity and selectivity, we designed truncations and mutations of the N‐terminal region, the first ‐sheet and the C‐terminal region of ACA‐01. We report here the effects of these mutations on binding to a range of human chemokines. These studies will establish a foundation for developing the anti‐inflammatory therapeutic potential of evasin proteins.

12. Membrane proteins: 24. Therapeutics and antibodies

ABS092

DISCOVERING NOVEL ANTIBODIES AGAINST PEPTIDISC STABILIZED MEMBRANE PROTEINS

James Saville ¹, Franck Duong², Katherine Zhao³

¹University of British Columbia (Vancouver, Canada); ²Professor of Biochemistry at The University of British Columbia (Vancouver, Canada); ³PhD Candidate at The University of British Columbia (Vancouver, Canada)

Membrane proteins constitute the majority (~60%) of drug targets due to their accessibility on the outside of our cells and central roles in fundamental cellular processes. The exploitation of these important drug targets is complicated due to their inherent instability in aqueous environments; a prerequisite for nearly all current biochemical analyses. Here, we present an application of the peptidisc technology, a novel membrane mimetic that facilitates the facile capture of diverse membrane proteins in their natively folded state. The broad applicability and minimal optimization required for peptidisc reconstitution aligns well with high throughput screening processes, such as those involved in antibody development. It is now well established that antibodies have numerous potential health benefits when developed as therapeutics. Current approaches to developing antibodies against membrane proteins involve using denaturing detergents or truncated portions of membrane protein targets. The resulting antibodies, developed under non‐native conditions, often do not translate well back into the native cellular context. We evaluate the technical challenges of raising antibodies against whole and natively folded membrane‐spanning proteins while integrating the peptidisc technology into established antibody screening methods. Methods such as classical animal immunization and phage and yeast display are examined and preliminary data suggesting the peptidisc as a potentially useful tool for the discovery of therapeutic antibodies against correctly folded membrane proteins is presented.

18. Proteomics: 2. Bioinformatics

ABS093

COINCIDENCE MAPS OF PROTEOLYTIC CLEAVAGE, SECONDARY STRUCTURE, AND EXON ORIGINS FOR THE SOLUBLE HUMAN PROTEIN HORMONE PROTEOME: FUNCTIONAL ASSOCIATIONS?

Kenneth L Campbell ¹, Nurit Haspel², Naomi Stuffers³

¹Univ. of Mass. Boston (Boston, United States); ²Dept. of Computer Science, Univ. of Massachusetts Boston (Boston, United States); ³Dept. of Biology, Univ. of Mass. Boston, 100 Morrissey Blvd., Boston, MA 02125 (Boston, United States)

Protein hormones seem too complex for single‐message use. Their transcripts sometimes code multiple (even antagonistic) hormones or pro‐proteins and some proteolytic fragments have secondary functions. Nesting secondary functions for release during processing in the synthetic cell, circulation, or target cells, makes biochemical sense. To explore our catalog of 2013 soluble human protein hormone transcripts and transcript products we arranged them using multiple sequence alignment (MAFFT) and published dendrograms, then co‐mapped: secondary structure, predicted cleavages (PROSPER; 24 proteases: aspartate, cysteine, metalloprotease, serine families), and known exon boundaries (e!Ensembl). Analysis of the 458 canonical transcripts in the list shows proteolysis in 457 cases with residual peptides of >10 residues (number/transcript = 9+/‐7 m+/‐SD, range 068; length/peptide, 18+/‐8 residues; range, 1087); only 1.29% of fragments had amino acid repeats of >4 residues. Cleavage patterns are retained across hormonal subfamilies, transcript isoforms, and transcript fragments down to 10‐20 residues. Coincidence is visually clear in subfamilies for proteolysis and secondary structure, less robust across full super families. Exon boundaries are conserved for subfamilies, but not super families. Early analysis of 37 random canonical transcripts for coincidence of proteolysis and exon boundaries show cleavage rates >4.4 times higher (% possible sites; p <0.01) within 3 (vs >3) residues of exon boundaries; computations and statistical analysis of coincidences of proteolysis, exon boundaries, and secondary structure changes for all transcripts are underway. Results suggest evolutionary cleavage pattern conservation that may allow physiological access to secondary nested signals within hormonal transcripts.

6. Design/engineering: 16. Protein interactions and assemblies

ABS094

DE NOVO DESIGN OF SELF‐ASSEMBLING HELICAL PROTEIN FILAMENTS

Hao Shen ¹

¹Institute for Protein Design, University of Washington (Seattle, United States)

There has been some success in designing stable peptide filaments; however, mimicking the controllable and reversible assembly of many natural protein filaments is challenging. We devised a general computational approach to designing self‐assembling helical filaments from monomeric proteins and use this approach to design proteins (34 out of 124 by negative stain electron micrograph) that assemble into micrometer‐scale filaments with a wide range of geometries in vivo and in vitro. Cryo‐electron microscopy structures of six designs are close to the computational design models. The filament diameter can be tuned by varying the number of repeats in the monomer. Anchor and capping units, built from monomers that lack an interaction interface, can be used to control assembly and disassembly. The filaments provide new phenomena through interactions with mammalian cells including the tendency to wrap around the cell nucleus. We also extended the building blocks to multi‐component, pH‐responsive and potentially more functional proteins to design more controllable filaments. The ability to generate dynamic, highly ordered structures that span micrometers from protein monomers opens up possibilities for the fabrication of new multiscale metamaterials.

3. Chaperones: 7. Dynamics and allostery

ABS095

DISTINCT PATHWAYS OF ACTIVATION OF HUMAN SMALL HEAT SHOCK PROTEIN HSPB5 BY DIFFERENT STRESS FACTORS

Maria Janowska ¹, Rachel Klevit¹

¹University of Washington (Seattle, United States)

Small heat shock proteins (sHsp) are a cells first stress response units. Their function is to maintain the solubility of proteins exposed to stress and is associated with changes in sHsp oligomer size and architecture under stress conditions. Multiple stress factors can lead to sHsp activation and include temperature, acidosis and ion imbalance. We have focused our efforts on the ubiquitously expressed human small heat shock protein HSPB5 that is upregulated under stress. Our research builds on our earlier discovery that at least one highly conserved histidine residue is important for the activation of HSPB5 and seeks to uncover the activation mechanism in detail. In the search for important residues we have focused on histidines as they are likely responsive to at least two stress factors: pH and changes in metal ion concentration. Our results show two distinct mechanisms of activation of HSPB5s structured core domain via pH change or interactions with metal ions. We observe that activation of HSPB5 via different triggers leads to different structural rearrangement in the core domain. Furthermore, changes in HSPB5 induced by either pH decrease or zinc binding are not restricted to the core domain, but are propagated to disordered regions of the protein. Our findings shed light on the modes of action of sHsp by identifying their important regulatory regions to provide a a path to better understanding of the role this family of proteins plays in the development of diseases such as cancer, cataracts, and neurodegenerative diseases.

16. Protein interactions and assemblies: 25. Transcription/translation/post‐translational modifications

ABS097

DETERMINING THE MECHANISM OF ACTION OF THE ANTIBIOTIC ARGYRIN B

Chris Swanson ¹, Riley Roberts¹, Jessica Mantchev¹, Catie Shelton¹, Justin Walter¹, Bassam Haddad¹, P. Clint Spiegel¹

¹Western Washington University (Bellingham, United States)

Elongation factor G (EF‐G) is a key bacterial translation factor and GTPase that is highly conserved in bacteria, with homologs in many other organisms. The peptide argyrin B, derived from myxobacteria and actinomycetes targets EF‐G specifically. Argyrin B has shown antibiotic activity against the bacteria Pseudomonas aeruginosa and membrane compromised strains of other gram negative bacteria. In order to elucidate the mechanism of action for argyrin B we performed in vitro experiments with the E. coli 70s bacterial ribosomes and different EF‐G variants. We initially observed the results of argyrin B on EF‐G to 70S binding and the GTPase activity that this binding facilitates. We also examined the binding affinity for substrate analogues. This work then investigated the effects on the two key functions of EF‐G, translocation and ribosome recycling. We ascertained from our experimentation that argyrin B increases binding to the ribosome, while having only a marginal effect on GTPase activity and substrate binding. Furthermore argyrin B did not present a change in ribosomal translocation but the drug does appear to inhibit ribosome dissociation, a key step in ribosome recycling. These experiments will narrow the possibilities of the mechanism of argyrin B and work to further the understanding of a unique antibiotic.

16. Protein interactions and assemblies: 12. Membrane proteins

ABS099

IMPACT OF OXIDATIVE STRESS ON THE STRUCTURAL CONFORMATION AND CHEMICAL INTEGRITY OF SOLUBLE CLIC1

Olga Faerch ¹, Stoyan Stoychev², Heini Dirr¹

¹PSFRU, School of Molecular and Cell Biology, Faculty of Science, University of the Witwatersrand, Johannesburg (Johannesburg, South Africa); ²Biosciences, Council for Scientific and Industrial Research (Pretoria, South Africa)

Chloride intracellular channel protein 1 (CLIC1) is soluble in the cytoplasm and capable of unassisted and reversible membrane insertion. CLIC1 is involved in oxidative stress responses, including cell cycle control (related to cancer) and A‐induced neurotoxicity (related to Alzheimers disease). Under oxidising conditions, Cys24 in the transmembrane domain experiences redox activation. This causes intra Cys24‐S‐S‐Cys59 bond formation, which exposes flat monomeric hydrophobic faces to dimerisation. This study aims to establish the impact of H2O2‐induced oxidative stress on the structural and chemical nature of soluble CLIC1. Purified CLIC1 was either reduced with DTT or oxidised with H2O2. The oligomeric state of reduced CLIC1 is solely monomeric, whereas the oxidised form is separated into dimer and monomer. Far‐UV CD demonstrated that the reduced form and both oxidised forms are mainly alpha‐helical, with no significant variations between samples. Fluorescence spectroscopy established that the overall tertiary structure remains unaffected by oxidation. Thermal melt studies indicated that the monomeric form (reduced and oxidised) has greater stability than the oxidised dimer. The effect of oxidation on susceptible residues was investigated using LC‐MS/MS. In addition to a Bradford assay and a fluorescence time dependent study, Trp35 remained unaffected following oxidation. However, Met32 appears to form a sulfoxide side‐chain upon oxidation. Using tandem IAA and NEM Cys‐labelling prior to LC‐MS/MS, Cys24 and Cys59 were observed to form a disulfide bond, whereas the remaining four Cys residues did not. Further analysis of LC‐MS/MS data will elucidate more concerning the chemical modifications induced by H2O2‐mediated oxidation on CLIC1.

16. Protein interactions and assemblies: 6. Design/engineering

ABS100

STRUCTURAL BASIS FOR ‐35 ELEMENT RECOGNITION BY SIGMA 4‐CHIMERA PROTEINS AND THEIR INTERACTIONS WITH PMRA RESPONSE REGULATOR

Chinpan Chen ¹, Yuan‐Chao Lou¹

¹Institute of Biomedical Sciences, Academia Sinica (Taipei, Taiwan)

In class II transcription activation, the transcription factor normally binds to the promoter near the 35 position and contacts the domain 4 of sigma factors (sigma4) to activate transcription. However, sigma4 of sigma‐70 factor appears to be poorly folded on its own. Here, by fusing sigma4 with the RNA polymerase beta‐flap‐tip‐helix, we constructed 2 sigma4 chimera proteins, one from sigma‐70 factor (sigma4‐70) and another from sigma‐S factor (sigma4‐S) of Klebsiella pneumoniae. The 2 chimera proteins well folded into a monomeric form with strong binding affinities for 35 element DNA. Determining the crystal structure of sigma4‐S in complex with 35 element DNA revealed that sigma4‐S adopts a similar structure as sigma4 in the E. coli RNAP/sigma‐S holoenzyme and recognizes 35 element DNA specifically by several conserved residues from the helix‐turn‐helix motif. By using NMR, sigma4‐70 was demonstrated to recognize 35 element DNA similar to sigma4‐S, but DNA binding reduced slow dynamics on sigma4‐70 only. Finally, sigma4‐70 was shown to interact with the response regulator PmrA and its promoter DNA. The chimera proteins are capable of 35 element DNA recognition and can be used for study with transcription factors or other factors that interact with domain 4 of sigma factors.

16. Protein interactions and assemblies: 21. Structure (x‐ray/NMR/EM)

ABS104

THE STRUCTURE OF THE INTERLEUKIN 11 SIGNALLING COMPLEX

Riley D. Metcalfe ¹, Kahenia Aizel¹, Courtney O. Zlatic¹, Paul M. Nguyen², Paul R. Gooley¹, Tracy L. Putoczki², Michael D.W. Griffin¹

¹Department of Biochemistry and Molecular Biology, Bio21 Molecular Science and Biotechnology Institute, University of Melbourne, Parkville, Victoria 3010, Australia (Melbourne, Australia); ²Inflammation Division, Walter and Eliza Hall Institute of Medical Research, Parkville, Victoria 3052, Australia (Melbourne, Australia)

The multifunctional cytokine interleukin (IL) 11 has recently been shown to have roles in several cancers, including gastrointestinal cancers. IL‐11 forms a signaling complex on the cell surface with two receptors, the IL‐11 specific receptor IL‐11R and the shared receptor glycoprotein (gp) 130. This results in the activation of intracellular signalling pathways, such as the JAK‐STAT pathway. Despite the potential therapeutic significance of IL‐11 signaling, little is known about the structural biology of IL‐11, or the IL‐11 signaling complex, hindering the development of novel therapies targeting IL‐11.

We have recently solved the crystal structure of the IL‐11 signalling complex. The structure is hexameric, consisting of two copies each of IL‐11, IL‐11R and gp130, with three copies of the hexamer being present in the asymmetric unit. We have confirmed this stoichiometry though extensive analytical ultracentrifugation (AUC), electron microscopy, and small‐angle X‐ray scattering (SAXS) characterization of the signalling complex. In addition, we have used isothermal titration calorimetry, to provide a thermodynamic understanding of complex formation. We have also studied several biologically active, antagonistic mutants of IL‐11, which prevent correct formation of the IL‐11 signalling complex. We have shown that these function by altering binding to IL‐11R, and also abolishing formation of the complete signalling hexamer. These antagonistic mutants are active in vivo, and are highly potent.

This work provides a structural and mechanistic basis for the formation of the IL‐11 signalling complex. This provides a scaffold for the future development of therapies that target IL‐11 signalling, which will be of therapeutic benefit.

21. Structure (x‐ray/NMR/EM): 16. Protein interactions and assemblies

ABS106

CRYO‐EM OF THE MALARIA PARASITE PA28/20S PROTEASOME COMPLEX REVEALS AN UNUSUAL ACTIVATION MECHANISM WITH IMPLICATIONS FOR ARTEMISININ SENSITIVITY

Stanley Xie¹, Michael Griffin ¹, Riley Metcalfe², Eric Hanssen², Tuo Yang², David Gilley², Andrew Leis³, Craig Morton², Michael Kuiper⁴, Michael Parker², Natalie Spillman², Wilson Wong⁵, Christopher Tsu⁶, Lawrence Dick⁶, Leann Tilley²

¹University of Melbourne (Parkville, Australia); ²Department of Biochemistry and Molecular Biology, Bio21 Molecular Science and Biotechnology Institute, The University of Melbourne, Melbourne, VIC 3010, Australia (Parkville, Australia); ³Advanced Microscopy Facility, Bio21 Molecular Science and Biotechnology Institute, The University of Melbourne, Melbourne, VIC 3010, Australia (Parkville, Australia); ⁴Data61 CSIRO, Docklands VIC 8012 Australia, Australian Cancer Research Foundation Rational Drug Discovery Centre, St Vincent's Institute of Medical Research, Fitzroy, Victoria 3065, Australia (Fitzroy, Australia); ⁵Infection and Immunity Division, The Walter and Eliza Hall Institute, Parkville, VIC 3052, Australia (Parkville, Australia); ⁶Oncology Clinical R&D, Takeda Pharmaceuticals International Co., Cambridge, Massachusetts 02139, USA (Cambridge, United States)

The proteasome is a multi‐subunit enzyme complex that is responsible for most of the non‐lysosomal proteolysis in eukaryotic cells. The malaria parasite, Plasmodium falciparum (Pf) is highly reliant on its protein turnover machinery, thus the proteasome is a drug target in the treatment of malaria. The activity of the 20S proteasome is regulated by protein complexes, such as the 19S complex. The PA28 regulator (also 11S or REG) has been shown to stimulate 20S proteasome peptidase activity in vitro, but its role in vivo remains unclear.

We show that genetic deletion of PfPA28 renders parasites more sensitive to anti‐malarial drugs, consistent with a role for PA28 in responding to proteotoxic stress. The crystal structure of PfPA28 reveals a bell‐shaped structure with a highly charged central channel, and large dynamic loops bordering the apical pore. We solved the structure of Pf20S in complex with one and two PfPA28 caps using single‐particle cryo‐EM, revealing the binding and activation mechanism of PfPA28 and providing evidence that PfPA28 employs a mechanism of 20S activation distinct from other 11S activators and the 19S complex. Cryo‐EM data also showed that PfPA28 and Pf20S form a dynamic complex, with PfPA28 undergoing large rigid‐body motions on Pf20S. We propose that the large loops of PfPA28 control entry of substrate to the Pf20S lumen and that the dynamic motions of the activator permit lateral transfer of proteasome products through the PfPA28/Pf20S interface as an alternative mechanism of substrate egress, avoiding the need for products to traverse the PfPA28 pore.

6. Design/engineering: 16. Protein interactions and assemblies

ABS107

TESTING PROTEIN DESIGN IN MASSIVE THROUGHPUT USING HIGH‐DENSITY PEPTIDE ARRAYS

Oana‐Nicoleta Antonescu ¹, Kristoffer Enøe Johansson¹, Jakob Rahr Winther¹

¹Linderstrøm‐Lang Centre for Protein Science, University of Copenhagen (Copenhagen, Denmark)

The ability to accurately design functional proteins can help us get a better understanding of the physics governing protein folding and interaction, as well as aid development of protein products with limitless applications. While computational protein design is capable of generating a large number of sequences for a specific fold, the practicalities of the current methods for testing the designs in cells (cloning, sequencing, expression, purification) become overwhelming. To overcome these challenges, we have produced a truncated Green Fluorescent Protein (tGFP) that lacks strand 10 (s10) in its beta‐barrel, tGFP presents a significant loss of function compared to the full‐length protein, but addition of a synthetic s10 peptide results in complete recovery of fluorescence. By immobilizing thousands of s10 variants on solid support in a peptide array setting, we could study in high‐throughput how different designs of s10 are capable of recovering the function of GFP, i.e. green fluorescence. To compare functional complementation versus non‐functional binding of the designed peptides, we have DyLight633‐labeled tGFP and investigated the binding at the array surface, by imaging it in the red channel. The results from a library of 15,301 s10 designs in 12‐fold replica show that the assay is capable of resolving the binding as well as the functional landscape of the split GFP interaction. It also offers the possibility of studying cooperativity between multiple mutations in the peptide sequence. To summarise, we have established and validated a novel platform that, in high‐throughput and with rapid turnover, can test computational peptide design experimentally.

16. Protein interactions and assemblies: 12. Membrane proteins

ABS108

THE INTERACTION OF IAA‐94 WITH THE SOLUBLE CONFORMATION OF THE CLIC1 PROTEIN AND ITS STRUCTURAL HOMOLOG HGSTP1‐1

Roland Worth ¹, Heinrich Dirr¹

¹University of the Witwatersrand (Johannesburg, South Africa)

Unlike most membrane proteins, the chloride intracellular channel 1 (CLIC1) protein is unique in that it can reversibly transit between a soluble and membrane‐bound conformation. Unfortunately, however, the only known structure of CLIC1 is for its soluble conformation, which conforms to the canonical topology of the glutathione transferase (GST) superfamily, but unlike GSTs which are homodimers, soluble CLIC1 is monomeric. Interestingly, a known inhibitor towards the chloride ion channel activity of CLIC1, compound indanyloxyaceticacid‐94 (IAA‐94) is noted to be a structural analog of ethacrynic acid (EA), an inhibitor of GSTs such as hGSTP1‐1. Therefore, in view of the structural homology between soluble CLIC1 and GST proteins, it remains uncertain if the constraint in the chloride conductance of CLIC1 is due to IAA‐94 binding to its soluble conformation to prevent its structural transition into a membrane competent state and/or IAA‐94 binding to the membrane‐bound conformation of CLIC1. To investigate the potential of soluble CLIC1, as well as is structural homolog hGSTP1‐1, to bind to IAA‐94, the thermoanalytical technique isothermal titration calorimetry (ITC) was used to study the physical basis of this molecular interaction. We have been able to show that IAA‐94 does not in fact bind to soluble CLIC1, but instead it can bind to hGSTP1‐1, suggesting that a dimeric interface would be required for the binding of IAA‐94 to soluble CLIC1. The inhibition in the ion channel activity of CLIC1 is therefore due to IAA‐94 binding to its membrane‐bound conformation and/or the dimeric intermediate species formed prior to membrane insertion.

4. Chemical biology: 21. Structure (x‐ray/NMR/EM)

ABS110

DISCOVERY AND CHARACTERIZATION OF SMALL MOLECULE INHIBITORS OF THE BROMODOMAIN CONTAINING PROTEINS BRD9 AND BRD7THE TARGETABLE SUBUNITS OF SWI/SNF CHROMATIN REMODELING COMPLEXES

Rezaul Karim ¹, Alice Chan², Ernst Schönbrunn²

¹USF Health Morsani College of Medicine, University of South Florida; Department of Drug Discovery, Moffitt Cancer Center (Tampa, United States); ²Department of Drug Discovery, Moffitt Cancer Center (Tampa, United States)

Inhibition of the bromodomain (BRD) containing protein 9 (BRD9) with small molecules has emerged as an attractive strategy to target SWI/SNF (SWItch/Sucrose Non‐Fermenting) chromatin remodeling complexes as it is mutated in 20% of all cancers. However, current BRD9 inhibitors also inhibit bromodomain containing protein 7 (BRD7), a close homolog of BRD9 but mediates different biological functions. The structural basis of inhibitor binding to BRD7 remained unknown. In this study, our objectives were to discover a novel BRD9 inhibitor and to elucidate the binding mode of small molecule inhibitors to BRD7. We used differential scanning fluorimetry (DSF) and isothermal titration calorimetry (ITC) as biophysical assays to determine inhibitory potential and utilized X‐ray crystallography to determine high resolution 3D structures of a series of BRD7/9‐inhibitor complexes. We have discovered that Cdc‐2 like kinase inhibitor TG003 also inhibits the bromodomains of both BRD9 and BRD7. This is the first report of a kinase inhibitor acting on bromodomains outside the BET (bromodomain and extra‐terminal) family. Combined, our findings provide a new structural framework for the rational design of dual bromodomain‐kinase inhibitors specifically targeting BRD9 and/or BRD7.

16. Protein interactions and assemblies: 23. Systems biology

ABS111

CALCIUM BINDING PROTEINS AND THE REGULATION OF THE VISUAL SENSORY SYSTEM: FROM MOLECULES TO NETWORKS

Daniele Dell'Orco ¹

¹University of Verona (Verona, Italy)

Light sensitivity in photereceptors is finely regulated by Ca2+ as the main second messenger. The homeostasis of Ca2+ is strictly connected to that of another second messenger, guanosine 3',5'‐cyclic monophosphate (cGMP). The drop of cytoplasmic Ca2+‐concentration following light absorption is detected by GCAPs, a group of neuronal calcium sensor proteins that control the activity of membrane bound guanylate cyclases by switching conformation in a Ca2+‐dependent manner. GCAPs contribute to shaping the response of cells to light, allowing the signaling cascade to make gradual responses to small changes in [Ca2+].

Single point mutations in GCAP1 are known to cause severe disturbance of their Ca2+‐sensing properties resulting in the onset of retinal dystrophies. Alterations in protein structure/function relationships can be characterized by biochemical and biophysical approaches, however in order to fully understand the dynamic changes of the homeostasis of both cGMP and Ca2+ in a photoreceptor cell under disease‐associated conditions, a more comprehensive kinetic description of the phototransduction cascade is necessary. Such system‐level kinetic description of the biochemical cascade based on experimentally determined parameters allows the simulation of the photoresponse in mouse rods by several illumination stimuli.

Recent results on the characterization of native and altered GCAP1 variants will be presented, which range from fine biophysical analyses focused on the intra‐molecular communication pathways connecting individual Ca2+ binding sites in the GCAP1 molecule to the putative residues at the interface with the guanylate cyclase to the system‐level properties of the rod outer segment, that trigger the electrical response of the cell to light.

12. Membrane proteins: 2. Bioinformatics

ABS112

IS WZA THE ONLY BACTERIAL OUTER‐MEMBRANE PROTEIN WITH HELICAL TRANSMEMBRANE SEGMENTS?

Sajith Jayasinghe ¹, Simon Keng¹, Ekta Priyam¹

¹California State University San Marcos (San Marcos, United States)

Almost all integral bacterial outer‐membrane proteins (OMPs) form closed ‐barrels composed of anti‐parallel transmembrane ‐strands containing alternating polar and non‐polar residues. One OMP, the capsular polysaccharide transporter Wza, has been shown to contain a c‐terminal ‐helical transmembrane (TM) segment which participates in the formation of an ‐helical barrel structure in the bacterial outer‐membrane. It is curious that to date only one OMP with a helical TM segment has been discovered, and the presence of a helical TM segment in an OMP is unexpected since it is generally thought that all OMPs have ‐strand TM segments to prevent their accidental insertion into the inner‐membrane by the SecYEG translocon. From a sequence analysis of an initial set of 28 OMPs proteins with unknown structure, 11 were predicted to contain a single c‐terminal helical TM segment (similar to Wza). Of these 11, five were proteins with ~ 40‐80% sequence similarity to Wza, indicatng the possibility that a helical TM may be a feature common to all outer‐membrane polysaccharide export lipoproteins. The remaining six lipoproteins appear to be unrelated to Wza and suggests the possibility that helical TM segments may be more widely distributed within bacterial OMPs.

24. Therapeutics and antibodies: 21. Structure (x‐ray/NMR/EM)

ABS113

THE USE OF SMALL‐ANGLE SCATTERING FOR STUDYING EXCIPIENT MODULATED PHYSICAL STABILITY AND VISCOSITY OF MONOCLONAL ANTIBODY FORMULATIONS

Joseph Curtis ¹, Amy Xu², Monica Castellanos³, Kevin Mattison⁴

¹NIST (Gaithersburg, United States); ²University of Maryland, College Park (Gaithersburg, United States); ³GlaxoSmithKline (Rockville, United States); ⁴Malvern Instruments (Westborough, United States)

Excipients are used in therapeutic protein formulations to maintain physical stability and to reduce solution viscosity. Undesirable protein‐protein interactions (PPI) can lead to protein aggregation and high solution viscosity, therefore, understanding the effects of excipients on PPI is important for optimized formulation design. In this study, we used NIST monoclonal antibody (NISTmAb) reference material as a model protein to examine the physical stability and viscosity in the presence of various excipients, varying in pH, salt composition and the presence of co‐solutes. Small‐angle scattering (SAS) using X‐rays (SAXS) or neutrons (SANS) together with other biophysical measurements were performed to obtain various experimental parameters to characterize excipient modulated PPI and bulk solution viscosities. In particular, good correlation was found between SAS and dynamic light scattering (DLS), suggesting the validity of using DLS for predicting protein colloidal stability in concentrated solutions. Detailed analysis of structure factor S(q) measured by SAS enabled the dissection of PPI into its attractive and repulsive components. The use of ionic excipients such as pH and salts lead to increased short‐range attraction, while the nonionic excipients including sugars, amino acids, and polysorbate surfactants lead to increased repulsive PPI with increasing protein concentration. The use of S(q), kD B22 (osmotic second virial coefficient) as viscosity predictors were compared to predicted results with the measured viscosities. Although B22 and S(q) appeared to be better viscosity predictors than kD, disagreement between predicted and measured results suggests other factors apart from PPI contribute to the bulk rheological properties of concentrated protein solutions.

26. Other: 26. Other

ABS114

TAKING A MAGIC LEAP INTO AUGMENTED REALITY PROTEIN STRUCTURE VISUALIZATION

Sajith Jayasinghe ¹, Byron Dehlavi¹, Lei Tang¹

¹California State University San Marcos (San Marcos, United States)

Protein function is intimately tied to a protein's three‐dimensional (3D) structure. From a pedagogical perspective it is critical that students of biochemistry understand the protein structure‐function paradigm so that they are prepared to tackle the research questions that follow. Most, if not all, students learn protein structure using two‐dimensional representations in a textbook, and they have difficulty appreciating the intricate 3D nature of proteins. The emergent technologies of virtual reality (VR) and augmented reality (AR) hold the promise of providing students a much richer and more interactive learning experience than has been possible so far. While VR technologies remove the user from the real world, augmented reality technology uses head mounted displays to superimpose computer generated 3D images on to the user's real world. Users can interact with the 3D AR images using their hands or via hand held controllers. Since the user remains aware of their surroundings AR technologies can be deployed in the classroom to provide students information while they listen to a lecture, or while they read their textbooks. We have developed a workflow to process Protein Data Bank coordinate files to generate models that can be loaded on to the Magic Leap One augment reality hardware suite, and for the user to interact with the protein model to investigate interesting characteristics. We are developing interactive AR tutorials that can be leveraged to enhance student learning of protein structure in the biochemistry curriculum.

2. Bioinformatics: 8. Enzymology

ABS115

EVALUATING THE MOLECULAR FUNCTION FAMILIES OF PHOSDUCINS USING MULTI‐ITERATIVE SEQUENCE SEARCHING TECHNIQUE

Sarah Hosler ¹, Jacquelyn Fetrow¹

¹Albright College (Andover, United States)

Our long‐term goal is to understand the functional determinants that specify different mechanisms within protein superfamilies. The specific goal of the ongoing project is to identify the functionally relevant clusters within the phosducin superfamily and to identify the functional determinants that distinguish the protein functional sites in each of the clusters.. The method used to identify functionally relevant clusters is MISST (Multi‐level Iterative Sequence Searching Technique), a method based on active site profiling. The phosducin superfamily was chosen after a thorough literature review of the thioredoxin fold suprafamily. Phosducin proteins are involved in G‐protein coupling and potentially impacts canine retinal atrophy. The first step in applying the MISST process is selection of key, or functionally important residues, identified by determining the conservative residues throughout the five phosducin proteins that are identified in the RCSB PDB database. Based on research of the thioredoxin suprafamily, it was apparent that the conserved cysteine and proline are part of the functionally important residues. The third residue was determined using Chimera, to structurally view the closeness of the residues to the proline and cysteine. We chose the leucine that is nearby in structure, with its side chain facing towards the other key residues, but is not close in sequence, this is leucine 196 in 3evi. MISST will be applied to the phosducin superfamily, from which key mechanistic determinants will be identified for each functionally relevant group. These results of the MISST experiment will then be compared to that of previous studies on the phosducin superfamily.

8. Enzymology: 22. Synthetic biology

ABS116

A NOVEL AND PROMISING MULTI‐ENZYME CO‐EMBEDDED ORGANICINORGANIC HYBRID NANOFLOWER WITH ENHANCED STABILITY AND CATALYTIC ACTIVITY

Duygu Aydemir ¹, Firdevs Gecili², Nalan Ozdemir², Nuriye Nuray Ulusu¹

¹Koc University School of Medicine Department of Medical Biochemistry (Istanbul, Turkey); ²Biochemistry Division, Chemistry Department, Faculty of Science, Erciyes University (Kayseri, Turkey)

Nanoflowers are newly developed flower‐shaped nano‐particles in the range of nano or micro scales such as copper, zinc, magnesium, calcium or copper‐capsular nanoflowers. Among them organic‐inorganic hybrid nanoflowers have been spotlighted, since they can overcome enzyme‐related limitations including low stability, high production cost, substrate/product inhibition and difficult recovery. Thus, enzyme immobilization enables to overcome these limitations and hybrid nanoflowers has become an effective method for enzyme immobilization. In this study, for the first time, multi‐enzyme co‐embedded organicinorganic hybrid nanoflower was synthesized using a three enzymes mixture (‐amylase, lipase and protease) as the organic components and Cu2+ ions as the inorganic component. In this way, all these three enzymes were immobilized in the same hybrid structures. Synthesized multi‐enzyme hybrid nanoflowers (‐amylase‐lipase‐protease‐ Cu2+ hybrid nanoflowers, ALP‐ihNFs) were characterized by their morphology and chemical point of view by using different techniques such as SEM, FTIR, EDX, and XRD. Afterwards we compared enzyme activity and stability of nanoflower to the each free enzyme including lipase, amylase and protease at different pH and temperature spectrophotometrically. Our data have revealed that nanoflower‐bound enzyme activities were significantly higher compared to the free enzyme and reusability of nanoflowers have been proven. In conclusion, fist time we showed a triple enzyme co‐embedded organicinorganic hybrid nanoflower with enhanced catalytic stability and reusability in the literature. This hybrid nanoflower can be used for treatment of wastewater, biosensors, biocatalysts, and bio‐related devices.

16. Protein interactions and assemblies: 10. Folding

ABS117

COMPARISON OF CALORIMETRY MEASUREMENTS OF BINDING OF A STREPTOMYCES TRYPSIN INHIBITOR TO THE THERMOSTABLE SUBTILASE AQUALYSIN I AND ITS COLD ADAPTED HOMOLOGUE, VPR

Sveinn Bjarnason ¹, Kristinn Ragnar Óskarsson¹, Magnús Már Kristjánsson¹

¹University of Iceland (Hafnarfjörður, Iceland)

The study of temperature adaptation of proteins has many facets, one of which is a comparison of binding affinities. We used a model system consisting of two structural homologous subtilisin‐like serine proteinases, the thermostable aqualyisin I (AQUI) and the cold adaptive VPR in an attempt to gauge different binding attributes of the enzymes. To this end, a protease inhibitor, STI, was cloned from S.lividans and overexpressed in E.coli. Protein‐protein binding was monitored using isothermal titration calorimetry (ITC) and differential scanning calorimetry (DSC). Thermograms were produced for each protein separately and with the subtilases with bound inhibitor at different molar ratios. Inhibitor/subtilase molar ratios ranged from 0 to 2 for the thermograms. Binding of the inhibitor might be too tight KD < 10‐9 M for correctly assessing binding constants with ITC, but H could easily be determined. The difference in H of binding for the subtilases was minimal. In the presence of excess inhibitor in the DSC thermograms, an exothermic event was observed in the middle of endothermic protein denaturation peaks. Comparing the energy in this exothermic peak to the ITC binding experiments, indicated that the H calculated from ITC binding corresponded to the energy in the exothermic peak in DSC thermograms. We attribute this exothermic peak to inhibitor binding to a newly released subtilase, where a bound inhibitor appears to denature at a lower temperature then the native homodimer form. To our knowledge this is the first reported binding observed during a thermal denaturation of a subtilase inhibitor complex.

16. Protein interactions and assemblies: 26. Other

ABS119

A HUMAN ACIDIC FIBROBLAST VARIANT WITH INCREASED STABILITY AND ENHANCED CELL PROLIFERATION ACTIVITY

Chynna Denham ¹, Shilpi Agrawal¹, T.K.S. Kumar¹

¹University of Arkansas (Hartford, United States)

Human acidic fibroblast growth factor‐1 (hFGF‐1) is a powerful and crucial protein involved in angiogenesis, cell proliferation, cell differentiation, and wound healing. hFGF‐1 is known to have relatively short half‐life and poor thermal stability. hFGF‐1 is known to exist in a partially unfolded state at physiological pH and temperature (37°C). However, in the body, hFGF‐1s positively charged region binds to a negatively charged polysaccharide called heparin. This stabilizes the protein and increases its half‐life. In this context, we introduced site‐specific mutations in hFGF‐1 at two particular sites within the heparin‐binding region (K126N and R136E). Characterization experiments, such as thermal and chemical denaturation, enzyme digestion, isothermal calorimetry (ITC), and fluorescence studies, were performed, and the K126N/R136E mutant and wild‐type proteins were compared in the presence and absence of heparin. The thermal denaturation results showed that K126N/R136E exhibits higher inherent stability, with a Tm of 49.3°C, than the wild‐type. The chemical denaturation characterization showed similar results, with K126N/R136E and wild‐type having Cm values of 2.52 M urea and 1.46 M urea, respectively. Enzyme digestion studies showed that K126N/R136E is more resistant to proteolytic cleavage than the wild‐type protein. ITC results confirmed that K126N/R136E has lost its heparin binding affinity. Overall, the results of this study implicate the possibility of using this hFGF‐1 mutant in pharmaceutical wound healing medications that can be used globally, including countries that do not have access to the cold storage that is currently needed to store and preserve the protein.

2. Bioinformatics: 5. Computational modeling/simulation

ABS120

PROPERTIES FOR PREDICTING PROTEIN FUNCTION

Caitlyn L. Mills¹, Mary Jo Ondrechen ¹, Lydia A. Ruffner¹, Penny J. Beuning¹

¹Northeastern University (Boston, United States)

There are now over 14,000 Structural Genomics (SG) protein structures deposited in the Protein Data Bank (PDB); most of these are of unknown or uncertain biochemical function. Here we present a powerful approach based on computed chemical properties of the individual residues in a protein structure. Local arrays of predicted active residues for sets of proteins of known function are matched with those of SG proteins. For instance, a superfamily consists of proteins with similar structure but multiple different kinds of biochemical function, including different types of reactivity as well as different substrate specificities. Each functional type has a different set of active amino acids. Typically active residues include those in the first layer that make direct contact with the substrate molecule(s) and also some residues in the second and third layers that play supporting roles in the catalytic process. Graph Representation of Active Sites for Prediction of Function (GRASP‐Func) establishes local arrays of predicted residues that are common to proteins of the same function. Predicted local arrays for each SG protein are aligned against those of the known members of each functional family. Cases of predicted misannotation, where our prediction differs from the originally assigned function, are especially interesting. Experimental testing is performed by direct biochemical assays to test our predictions. For instance, we show that our predictions are confirmed that RV0760c from Mycobacterium tuberculosis has ketosteroid isomerase activity, whereas NP_103587.1 from Mesorhizobium loti does not.

Acknowledgments: NSF grant CHE‐1305655, MathWorks, Inc. and PhRMA Foundation Fellowship (CLM).

5. Computational modeling/simulation: 7. Dynamics and allostery

ABS122

NETWORK‐LEVEL ANALYSIS OF MOLECULAR DYNAMICS SIMULATIONS REVEALS ALLOSTERIC PROPERTIES OF CALCIUM SENSOR PROTEINS

Valerio Marino ¹, Daniele Dell'Orco¹

¹University of Verona (Verona, Italy)

Neuronal Calcium Sensors (NCS) are highly conserved proteins specifically expressed in neurons. Calcium (Ca2+) binding to their EF‐hand motifs results in a conformational change fundamental for the regulation of a specific target and the downstream biological process. Here we present a comprehensive analysis of the allosteric communication between metal‐binding EF‐hands and the target interfaces of NCS1, Recoverin (Rec) and GCAP1. In detail, all three proteins were investigated in different metal‐loading states and in complex with specific target peptides. For each protein state, a Protein Structure Network (PSN) accounting for persistent nonbonded interactions was built based on exhaustive 1 μs Molecular Dynamics simulations. Results for NCS1 and Rec highlighted allosteric inter‐domain communication between EF‐hands and residues of the target interface, while GCAP1 N‐terminal myristoylation was found to mediate the long‐range communication between EF4 and both EF2 and the GCAP1/GC interface. PSN analysis identified key residues involved in the highest number of interactions (high‐degree hubs) and helped unveil their role in allosteric and pathological mechanisms, as some GCAP1 hubs are the target of retinal dystrophy mutations. Finally, we suggest an evolution‐driven correlation between high‐degree hubs and their conservation among homologous NCS proteins. Such analysis is virtually extendable to any protein/target complex for which structural information is available.

7. Dynamics and allostery: 21. Structure (x‐ray/NMR/EM)

ABS123

CALCIUM AND INTEGRIN BINDING PROTEIN 2 (CIB2): AN ATYPICAL METAL SENSOR PROTEIN

Giuditta Dal Cortivo¹, Giuditta Dal Cortivo ¹, Rosario Vallone², Valerio Marino¹, Mariapina D'Onofrio¹, Daniele Dell'Orco¹

¹University of Verona (Verona, Italy); ²Strada Le Grazie, 8 (Verona, Italy)

CIB2 is a 22‐kDa protein with two functional EF‐hands, EF3 and EF4, which coordinate Ca2+ and Mg2+. Binding of metal cations leads to a conformational change from a molten globule state to a well‐defined tertiary structure. The biological role of CIB2 remains unclear, as it is supposed to be involved in several biological processes including the regulation of oncogenic signaling in ovarian cancer, contribution to HIV‐1 viral entry and mechanotransduction. Recently, five point mutations in CIB2 have been associated with genetic diseases affecting hearing. We specifically studied the Glu64Asp variant by a combination of biochemical and biophysical techniques. While SEC and DLS suggest that CIB2 forms a functional dimer, mass spectrometry clearly shows that CIB2 is a monomer. Interestingly, despite its name, CIB2 shows a very low affinity for Ca2+, while it is expected to bind magnesium under physiological conditions. Surprisingly, NMR spectroscopy revealed long‐range allosteric communication between Glu64, located at the N‐terminal domain, and the metal cation binding site EF3, located at the C‐terminal domain. By breaking inter‐domain communication, the conservative mutation impairs the ability of CIB2 to switch to its Mg2+‐bound form. We further investigated the binding of CIB2 to peptides from two integrins: IIB, a common target of CIB family members, and 7B, which seems to be specific for CIB2. For both the tested peptides we found a moderate affinity with fast dissociation kinetics. The data presented here show that the somewhat atypical features of CIB2 as a metal sensor may open interesting biological scenarios.

8. Enzymology: 6. Design/engineering

ABS124

A SINGLE MUTATION ASP98SER, WHICH IMPROVES THE CATALYTIC PROPERTIES OF THE THERMOSTABLE SUBTILASE AQUALYSIN I, INCREASES FLEXIBILITY AT ITS ACTIVE SITE

Arnor Saevarsson¹, Magnus Kristjanssosn ¹, Brynjar Ellertsson¹

¹University of Iceland (Reykjavik, Iceland)

Aqualysin I (AQUI), a highly thermostable subtilisin‐like serine proteinase (subtilase) from the thermophile Thermus aquaticus and VPR, from a psychrophilic Vibrio sp., are structural homologues, but differ greatly with respect to stability and catalytic properties. It has been postulated that the difference observed in the catalytic properties of cold‐ and heat adapted enzymes may reflect a difference in their molecular flexibilities. Thus, lower catalytic activity of thermophilic enzymes, compared to lower temperature homologues, would reflect their structural rigidity, which would also contribute to their higher thermal stability. A variant of AQUI containing a single mutation Asp98Ser (AQUI_D98S), as present in VPR, resulted in over 2 fold increase in catalytic efficiency (kcat/Km) compared to wild type (AQUI_wt). The D98S mutation did not affect the thermal stability of the enzyme. Asp98 is connected by hydrogen bonds to residues of the substrate binding site of AQUI. Some of these hydrogen bonds are lost in AQUI_D98S, which may lead to increased molecular flexibility at the active site, resulting in higher catalytic activity. To test this possible relationship between molecular flexibility and catalytic activity we used site‐directed spin labeling (SDSL) in the active site of AQUI_wt and the AQUI_D98S variant. The motion of the active site spin label was monitored by electron paramagnetic (EPR) spectroscopy. The EPR line shapes suggested considerably more motion of the active site spin label in the more active AQUI_D98S compared to AQUI_wt, suggesting higher flexibility of its active site, coinciding with its higher catalytic activity.

16. Protein interactions and assemblies: 4. Chemical biology

ABS125

GENERATION OF 13 FULL LENGTH PROTEINS OF THE CGAS‐STING PATHWAY FOR DRUG TRACTABILITY ASSESSMENT

Yong Jiang¹, Yong Jiang ¹

¹GSK (Collegeville,, United States)

In early drug discovery, selecting a tractable protein target within a well‐characterized, clinically relevant pathway has been shown to increase the success rate for clinical translation. Recently, Affinity Selection Mass Spectrometry (AS‐MS) has become a powerful tool to assess protein target tractability. The generation of high quality, full‐length protein has proven to be one of the biggest challenge using this platform. Here, we focus on the expression and purification of effector poteins within the cGAS‐STING pathway, as cGAS‐STING plays a critical role in the activation of autophagy, as well as induction of interferons and inflammatory cytokines, which is prevelant in many disease states. Full‐length proteins including cGAS, STING, TBK1, IRF3, IRF7, IKK, IKK, IKK, NFkB, INF, INF, INF and MDA5 were generated and deemed fit for purpose for ASMS. Successful QC criteria (>70% purity, no major contaminant, nonaggregate) were achieved for all samples. As a result of the success of protein purification, ASMS can be enabled to help identify primitive tool molecules to probe the targets validation in early drug discovery.

ABS127

A GENERATIVE ALGORITHM FOR PROTEINS FROM THE NTF2‐LIKE SUPERFAMILY

Benjamin Basanta ¹, Mathew Bick², Ted Baughman³, Philip Leung³, Eric Nalefski³, David Baker⁴

¹Institute for Protein Design (Seattle, United States); ²University of Washington (Seattle, United States); ³Global Good (Seattle, United States); ⁴Howard Hughes Medical Institute (Seattle, United States)

Proteins from the NTF2‐like superfamily exhibit a wide range of pocket shapes and sizes in a relatively small scaffold. This structural diversity is based on the configuration of a curved beta‐sheet combined with three alpha helices, providing support for a wide range of functions. Systematic generation of pocket structural diversity remains an outstanding challenge in de novo protein design. We aimed to adapt the simple system that gives rise to wide structural diversity in NTF2‐like proteins to de novo design. Here we developed and tested a generative algorithm for proteins of the NTF2‐like superfamily. By experimentally testing the stability of thousands of proteins and analyzing these results using machine learning techniques, we refined the algorithm to produce more diverse and stable proteins. The designs generated by the algorithm cover a significant part of the native NTF2‐like structural space, as well as structural space not sampled by nature. As a proof of concept, we used the models generated by the algorithm as scaffolds to design small‐molecule binding proteins. One of these designs is able to bind a mycotoxin of general interest public health, aflatoxin B1, with modest affinity. Systematic generation of pocket structural diversity should provide protein designers with custom‐tailored scaffolds for a wide range of binding and active sites.

5. Computational modeling/simulation: 16. Protein interactions and assemblies

ABS128

INFLUENCE OF PULLING GEOMETRY ON MECHANICAL STABILITY OF PROTEIN‐PEPTIDE COMPLEXES

Maksim Kouza ¹, Andrzej Kolinski², Irina Buhimschi¹, Andrzej Kloczkowski¹

¹Nationwide Childrens Hospital (Columbus, United States); ²Faculty of Chemistry, University of Warsaw (Warsaw, Poland)

Protein‐peptide interactions play essential roles in many cellular processes and their structural characterization is the major focus of current experimental and theoretical research. Two decades ago, it was proposed to employ the steered molecular dynamics to assess the strength of protein‐peptide interactions (1). The idea behind using steered molecular dynamics simulations is that the mechanical stability can be used as an efficient alternative to computationally highly demanding estimation of binding affinity and aggregation rate(2). However, mechanical stability defined as a peak in force‐extension profile depends on the choice of the pulling direction. Here we propose an uncommon choice of the pulling direction along resultant dipole moment vector, which has not been explored in simulations so far. Using explicit solvent all‐atom MD simulations(3), we apply steered molecular dynamics technique to probe mechanical resistance of protein‐peptide system pulled along two different vectors(4). A novel pulling direction, along the resultant dipole moment vector, results in stronger forces compared to commonly used peptide unbinding along center of masses vector. Our results demonstrate that resultant dipole moment is one of the factors influencing the mechanical stability of protein‐peptide complex.

• 1) H. Grubmuller et al., Science 271, 997999(1996)
• 2) M. Kouza et al., J Chem Phys 148, 215106 (2018); M. Kouza et al. J Chem Phys 146, 135101 (2017)
• 3) M. Kouza et al., Computational Methods to Study the Structure and Dynamics of Biomolecules and Biomolecular Processes: From Bioinformatics to Molecular Quantum Mechanics, 541‐558 (2019)
• 4) M. Kouza, A. Banerji, A. Kolinski, I. Buhimschi, A. Kloczkowski, Molecules 23, 1995 (2018)

21. Structure (x‐ray/NMR/EM): 16. Protein interactions and assemblies

ABS129

CRYSTAL STRUCTURES OF THE COMPLEX OF A KALLIKREIN INHIBITOR FROM BAUHINIA BAUHINIOIDES WITH TRYPSIN AND MODELING OF KALLIKREIN COMPLEXES

Mi Li ¹

¹Leidos Biomedical Research, Inc (Frederick, United States)

Structures of a recombinant Kunitz‐type serine protease inhibitor from Bauhinia bauhinioides (BbKI) complexed with bovine trypsin were determined in two crystal forms. The crystal structure with the L55R mutant of BbKI was determined in space group P64 at 1.94 å resolution and that with native BbKI in the monoclinic space group P21 at 3.95 å resolution. The asymmetric unit of the later crystals contained 44 independent complexes, thus representing one of the largest numbers of independent objects deposited in the Protein Data Bank. Additionally, the structure of the complex with native BbKI was determined at 2.0 å resolution from P64 crystals isomorphous to those of the mutant. Since BbKI has previously been found to be a potent inhibitor of the trypsin‐like plasma kallikrein, it was also tested against several tissue kallikreins. It was found that BbKI is a potent inhibitor of human tissue kallikrein 4 (KLK4) and the chymotrypsin‐like human tissue kallikrein 7 (KLK7). Structures of BbKI complexed with the catalytic domain of human plasma kallikrein were modeled, as well as those with KLK4 and KLK7, and the structures were analyzed in order to identify the interactions that are responsible for inhibitory potency.

12. Membrane proteins: 21. Structure (x‐ray/NMR/EM)

ABS130

DNA‐CORRALLED NANODISCS FOR THE STRUCTURAL AND FUNCTIONAL CHARACTERIZATION OF MEMBRANE PROTEIN AND VIRAL ENTRY

Gerhard Wagner¹, Mahmoud Nasr ², William Shih³, Meng Zhang¹, Krishna Das¹

¹Harvard Medical School (Boston, United States); ²Harvad Medical School (Boston, United States); ³Wyss Institute (Boston, United States)

We have developed a modular method to manufacture large sized nanodiscs using DNA origami Corrals as scaffold. Large sized nanodisc can be produced after adding extra lipids to DNA corrals decorated with small lipid bilayer nanodiscs. We can achieve high efficiency of large sized DNA‐corralled nanodisc (DCND) reconstitution, including 75‐nm, 40‐nm and 20‐nm sized nanodisc. We demonstrated the reconstitution of densely packed/crystalline membrane protein array within these large nanodiscs for the structure determination. Furthermore, we used these nanodiscs as a model membrane to study poliovirus entry process.

18. Proteomics: 23. Systems biology

ABS133

MEASURING THE FUNCTIONAL EFFECT OF AMINO ACID SUBSTITUTIONS PROTEOME‐WIDE USING MISTRANSLATION

Stephanie Zimmerman ¹, Ricard Rodriguez‐Mias¹, Kyle Hess¹, Judit Villen¹, Stanley Fields¹

¹University of Washington Department of Genome Sciences (Seattle, United States)

Experimentally measuring the effect of mutations in diverse proteins would improve how accurately we can predict the consequences of genetic variation. To this end, we are developing a method that measures the functional effects of amino acid substitutions in the entire proteome in a single experiment. This method uses mistranslation to create proteome‐wide libraries of variants. We then impose a biochemical selection on the pool of variants, such as increased temperature which causes some proteins to denature. We use mass spectrometry to quantify each variant relative to its wild type counterpart before and after the selection, identifying variants that change the physical property in question. To generate variants, we engineered several mutant tRNAs that mistranslate serine at levels of 5‐10% at specific non‐serine codons in yeast. Next, we used the thermal proteome profiling method to measure the stability of thousands of variant proteins that contain one of six different amino acids instead of serine. In parallel, we also measured the effects of these variants on protein solubility. Although most substitutions had no effect on thermal stability or solubility, we identified many variants that destabilize proteins, including examples in homologs of clinically‐relevant proteins. We are developing selections that interrogate other aspects of protein function, such as protein‐protein interactions. Ultimately, our goal is to measure the effects of many different substitutions at all sites in the proteome to predict clinically and biologically relevant variation.

1. Amyloid and aggregation: 16. Protein interactions and assemblies

ABS134

DYNAMICS OF AMYLOID FIBRILS PLAY A ROLE IN SEEDING AND PROPAGATING THE AGGREGATION OF ‐SYNUCLEIN

Jonathan Williams ¹, Xue Yang¹, Jean Baum¹

¹Rutgers University (Piscataway, United States)

Neurodegenerative disease progression has been increasingly linked to seeding and propagation of fibril formation throughout the brain. Current understanding of the mechanism by which amyloid fibrils seed the templated misfolding of endogenous monomeric N‐terminally acetylated ‐synuclein in Parkinsons disease is limited. It has been proposed that fibril polymorphism plays a role in recruitment of monomeric ‐synuclein to aggregate, however our work suggests that fibril dynamics also play a critical role in this process. Our data show that ‐synuclein fibrils that are formed by co‐incubation with the neuroprotective intrinsically disordered protein ‐synuclein (co‐S/S fibrils) are less cytotoxic, exhibit reduced cell seeding capacity and are more resistant to fibril shedding compared to ‐synuclein fibrils alone. Using solid‐state NMR, and we show that these co‐S/S fibrils exhibit increased dynamics of residues located in the N‐terminal imperfect KTKEGV repeats. Measurements of water accessibility by solid‐state NMR also indicate that these residues are more water accessible in the co‐S/S fibrils versus the ‐synuclein fibrils alone. Furthermore, proteinase K digestion assays show that the co‐S/S fibrils are degraded to a larger extent than ‐synuclein fibrils. This work brings a new dimension to our understanding of the possible mechanisms of fibril seeding and propagation in the cell, and highlights the importance of fibril dynamics in disease.

16. Protein interactions and assemblies: 7. Dynamics and allostery

ABS135

THE MECHANISM OF CAMKII REGULATION: FROM FERTILIZATION TO ENCODING LONG‐TERM MEMORY

Margaret Stratton ¹

¹University of Massachusetts, Amherst (amherst, United States)

Ca2+‐calmodulin dependent protein kinase II (CaMKII) is a crucial oligomeric enzyme in neuronal and cardiac signaling, fertilization and immunity. Work in the Stratton lab is focused on understanding the role of this fascinating enzyme in different tissues. We have used RNA sequencing to determine which transcripts are present and we are characterizing these different variants in terms of their activation profiles. To facilitate our studies in cells, we have developed a novel, substrate‐based, genetically‐encoded sensor for CaMKII activity, FRESCA (FRET‐based Sensor for CaMKII Activity), which has allowed us to monitor CaMKII activity in live cells under various conditions. We hope that by gaining an understanding of CaMKII in vitro and in cells, we will be able to better understand medical conditions in which it is implicated, such as memory deficiencies and infertility.

21. Structure (x‐ray/NMR/EM): 12. Membrane proteins

ABS136

VISUALIZING CONFORMATIONAL CHANGES OF THE MAGNESIUM CHANNEL CORA USING SYNTHETIC ANTIBODIES

Satchal Erramilli ¹, Piotr Tokarz¹, Kamil Nosol¹, Przemyslaw Dutka¹, Blazej Skrobek¹, Pawel Dominik¹, Somnath Mukherjee¹, Anthony Kossiakoff¹

¹The University of Chicago (Chicago, United States)

Synthetic antibodies (sAB), based on a Fab framework (50 kDa) and generated using phage display, have demonstrated structural biology utility. sABs form stable and rigid complexes with target proteins to facilitate crystallization, reduce heterogeneity, and act as fiducials for orientation in cryo‐EM. in vitro phage display selections can be tailored to produce region‐ and conformation‐specific binders for desired protein functional states.

Here we demonstrate that sABs can be utilized to visualize protein conformational changes by EM. The structure of the pentameric ion channel, CorA has been solved in inactive Mg2+‐bound and active Mg2+‐free states. Using phage display, we generated a cohort of sABs for CorA to probe structure‐function relationships. One sAB, C100, binds one copy per monomer and can be used as a fiducial in cryo‐EM. By monitoring C100 translations in 2D projections, we can measure relative Fab positions to deduce conformational changes leading from channel activation (low [Mg2+]) to inactivation (high [Mg2+]). Beyond 2D projections, we can characterize the populations of states at any given [Mg2+].

We additionally obtained several conformation‐specific sABs that are sensitive to [Mg2+]. Using SPR, we can probe the energetic contributions of magnesium binding by measuring sAB affinity as a function of [Mg2+]. We subsequently determine how these sABs bind to CorA using EM, allowing us to link structural transitions with energetic contributions of magnesium binding. Thus, we are able to demonstrate the biological responses of CorA to environmental magnesium changes. We believe this approach provides a foundation for annotating these pathways in other systems.

16. Protein interactions and assemblies: 23. Systems biology

ABS137

HIGH‐THROUGHPUT IDENTIFICATION OF DOMINANT NEGATIVE POLYPEPTIDES IN YEAST

Michael Dorrity ¹, Michael Dorrity², Christine Queitsch², Stanley Fields²

¹University of Washington (Seattle, United States); ²University of Washington, Genome Sciences (Seattle, United States)

Dominant negative polypeptides can inhibit protein function by binding to the wild type version or by titrating a ligand. Here, we use high‐throughput sequencing of libraries composed of fragments of yeast genes to identify dominant negative polypeptides based on their depletion during cell growth. The method can uncover numerous inhibitory polypeptides for a protein and thereby define these fragments with exquisite resolution, even pinpointing individual residues with critical functional roles. Remarkably, some proteins contain multiple minimal inhibitory regions distant in primary sequence but nearby in three‐dimensional structure. These minima correspond to internal structural features rather than ligand interaction, suggesting that the inhibition may instead relate to protein folding. Furthermore, many dominant negative fragments showed temperature‐dependent effects, suggesting that the protein‐folding environment of cells is important in determining whether gene truncations have deleterious effects.

ABS138

EPR REVEALS DIFFERENT CONFORMATIONS OF LCRG

Pallavi Guha Biswas ¹, Pavanjeet Kaur², Andrew McShan¹, Kawaljit Kaur¹, Likai Song², Roberto De Guzman¹

¹University of Kansas (Lawrence, United States); ²Florida State University (Tallahasse, United States)

Yersinia pestis is the causative agent of bubonic plague. Yersinia assembles a protein nanoinjector of the Type III Secretion System (T3SS) to inject virulence effector proteins into its target host cells. The T3SS consists of the needle apparatus, chaperones, and effector proteins. The Yersinia LcrG protein functions as a chaperone to the LcrV tip protein. Previous results by NMR and CD spectroscopy showed that LcrG lacks a tertiary structure but contains partially folded alpha helices. Results by others suggest that the two alpha helices of LcrG may interact with each other. Here, we used Electron Paramagnetic Resonance (EPR) spectroscopy to determine the conformations of LcrG in the presence and absence of its binding partner, the tip protein LcrV. For EPR, we engineered cysteine mutations in LcrG and LcrV for attachment of spin labels. EPR results suggest that when LcrG is not bound to the tip protein LcrV, there may be a population of LcrG that has transient tertiary structures where the two helices are in close contact with each other. When LcrG is bound to the tip protein, LcrV, the two helices of LcrG do not interact, suggesting that LcrG exists in an extended form. EPR provides additional insight into different conformations of LcrG in the free and the bound form with its binding partner.

5. Computational modeling/simulation: 2. Bioinformatics

ABS139

BETTER TOGETHER: 20+ YEARS OF SCIENTIFIC SOFTWARE DEVELOPMENT IN THE ROSETTA MACROMOLECULAR MODELING SUITE

Julia Koehler Leman ¹, Brian Weitzner², Douglas Renfrew¹, Richard Bonneau¹

¹Simons Foundation/NYU (New York, United States); ²University of Washington (Seattle, United States)

Many scientific disciplines rely on computational methods for data analysis, model generation and prediction. Implementation of these methods is often accomplished by researchers within specific domains, without formal training in software engineering or computer science. Perhaps unsurprisingly, sustainability and maintainability of tools developed for basic science research are often underappreciated in academic environments. Our experience developing the Rosetta scientific software suite, one of the largest software suites for macromolecular modeling, with 3 million lines of code and many state‐of‐the‐art applications, has provided us with more than two decades of experience in how to effectively develop advanced scientific software in a global community with hundreds of developers. Rosetta is developed collaboratively by the RosettaCommons, a community of developers from over 60 laboratories worldwide. Since the mid 1990s, Rosetta has been developed mostly by academics with diverse backgrounds including chemistry, biology, physiology, physics, engineering, mathematics, and computer science. We present lessons learned and address technical aspects like version control, licensing, testing, documentation, maintenance and a variety of community‐building strategies such as conferences, training, hackathons, user interaction, as well as outreach and diversity efforts.

5. Computational modeling/simulation: 22. Synthetic biology

ABS140

MECHANICAL PROPERTIES OF DESIGNED PROTEIN FIBERS

Neville Bethel ¹, Matt Bick¹, David Baker¹

¹Institute for Protein Design, University of Washington (Seattle, United States)

Protein fibers like silk are stronger than steel and more flexible than nylon. Traditional biological approaches have yielded little understanding for how protein based materials can have these remarkable properties. On a fundamental level, it is not clear how protein geometry, stability and intermolecular interactions affect the mechanical response of protein fibers or networks. One issue is that protein based materials found in nature tend to be complex and have features not directly responsible for their mechanical properties, such as enzymatic activity. Here, we aim to design simplified

protein fibers de novo, and we plan to examine a series of these proteins using simulation, microscopy and single molecule experiments to deconstruct the microscopic interactions responsible for their macroscopic properties. An additional goal is to develop and validate a computational model that can predict the mechanical properties of a designed protein from its sequence. Early protein design models and electron microscopy data will be discussed.

21. Structure (x‐ray/NMR/EM): 16. Protein interactions and assemblies

ABS141

BIOPHYSICAL STUDIES OF MINOR TRANSLOCON IPAC OF THE TYPE III SECRETION SYSTEM IN SHIGELLA

Amritangshu Chakravarty ¹, Helen Peng¹, Dr Roberto N De Guzman¹

¹University of Kansas (Lawrence, United States)

Pathogenic Gram‐negative bacteria such as Salmonella, Shigella and Burkholderia inject virulence factors into eukaryotic cells by assembling the Type III Secretion System (T3SS). The structural component of the T3SS is the needle apparatus, which consists of a base, an extracellular needle, a tip complex and the translocon. The translocon proteins are classified into the major and minor translocon proteins based on their molecular weights. The atomic structure of the minor translocon protein is currently unknown. The minor translocons from various bacteria are predicted to have one transmembrane domain, which is important for insertion into membranes. The minor translocon protein of Shigella is IpaC. We have subcloned, expressed and purified the full‐length as well the N‐terminal and the C‐terminal domains of IpaC, and studied their interactions with the chaperone IpgC, the tip protein IpaD and the major translocon protein IpaB by NMR spectroscopy. We have shown that full‐length IpaC and its N‐terminal and C‐terminal domains bind to IpgC, IpaD and IpaB. We have also used NMR to demonstrate that IpaC undergoes conformational changes upon insertion into micelles. The results will help us better understand the role of IpaC in the pathogenesis of these bacteria.

6. Design/engineering: 5. Computational modeling/simulation

ABS142

COMPUTATIONAL PROTEIN DESIGN WITH MULTISITE LAMBDA DYNAMICS

Ryan Hayes ¹, Jonah Vilseck¹, Charles Brooks III¹

¹University of Michigan (Ann Arbor, United States)

Most protein design problems are free energy optimization problems, often focused on folding or binding free energies. Conventional computational protein design methods have enjoyed a great deal of success, but sometimes fail due to insufficient accuracy. Alchemical free energy methods have found wide use in the related field of computer aided drug design, and offer rigorous free energy predictions that can complement existing protein design methods. Among alchemical methods, multisite lambda dynamics (MSLD) is an emerging method that is uniquely suited to protein design because of its ability to characterize the combinatorial sequence spaces encountered in protein design. Studies of T4 lysozyme reveal unprecedented agreement with experimental measurements, (with Pearson correlations of 0.9), that exceed the accuracy of other methods. New developments enable spaces of over 30,000 sequences to be sampled robustly within a single simulation. Encouraged by these results demonstrating the promise of MSLD in protein design, we have begun to apply MSLD to design‐like problems in systems including ribonuclease H.

12. Membrane proteins: 18. Proteomics

ABS143

PROFILING THE E. COLI MEMBRANE INTERACTOME CAPTURED IN PEPTIDISC LIBRARIES

Irvinder Wason ¹, Irvinder S. Wason², Greg Stacey³, John Young², Michael Carlson², Zhiyu Zhao², David G. Rattray³, Nichollas Scott³, Craig Kerr³, Mohan Babu⁴, Leonard J. Foster³, Franck Duong²

¹University of British Columbia (Vancouver, Canada); ²Department of Biochemistry and Molecular Biology, Faculty of Medicine, Life Sciences Institute, University of British Columbia, Vancouver, British Columbia, Canada (Vancouver, Canada); ³Department of Biochemistry and Molecular Biology, Faculty of Medicine, Michael Smith Laboratory, University of British Columbia, Vancouver, British Columbia, Canada. (Vancouver, Canada); ⁴Department of Biochemistry, Faculty of Science, University of Regina, Canada. (Regina, Canada)

Protein‐correlation‐profiling (PCP), in combination with quantitative proteomics, has emerged as a high‐throughput method for the rapid identification of dynamic protein complexes in their native conditions. While PCP has been successfully applied to soluble proteomes, characterization of the membrane interactome has lagged partly due to use of detergents necessary to maintain protein solubility. Here, we apply the peptidisc, a one‐size fits all membrane mimetic, for capture and high‐resolution fractionation of the Escherichia coli cell envelope proteome without the presence of detergent. Analysis of the SILAC labeled peptidisc library via PCP allows generation of over 4900 possible binary interactions. Using the Sec translocon and MetNI ABC transporter as validation targets, we show that our dataset is a useful resource for identification of protein interactions not otherwise detected by standard affinity purification. The application of the peptidisc to the proteomics workflow is emerging as a promising novel approach for rapid characterization of membrane protein interactions under native expression conditions without genetic manipulation.

16. Protein interactions and assemblies: 23. Systems biology

ABS144

SYSTEMS STRUCTURAL BIOLOGY OF THE HEART: IMPACT OF LYSINE ACETYLATION ON PROTEIN CONFORMATIONS AND INTERACTIONS

Juan Chavez ¹, Matthew Walker¹, Arianne Caudal¹, Bo Zhou¹, Andrew Keller¹, Rong Tian¹, James Bruce¹

¹University of Washington (Seattle, United States)

Introduction: Heart disease is a leading cause of death worldwide. Mitochondrial dysfunction is intimately tied with cardiomyopathies. Changes in the redox balance of nicotinamide adenine dinucleotide (NADH/NAD+) ratio and protein acetylation are molecular features underling mitochondrial function/dysfunction. An elevated NADH/NAD+ ratio results in hyperacetylation due to decreased activity of the sirtuin NAD+ dependant deacetylases, contributing to the development of heart disease. Normalization of the NADH/NAD+ ratio benefits mitochondrial and heart function. Current proteomics techniques enable the identification and quantitation of thousands of acetylated sites on proteins. However, the structural impact of acetylation at these sites remains largely unknown. Here we combine acetylome analysis with chemical cross‐linking and mass spectrometry to gain insight into the structural/interaction impacts of acetylation.

Methods: Hearts were harvested from wild type and sirtuin 3 knockout mice. Isolated hearts were perfused for 40 min prior to chemical cross‐linking. Proteins were extracted and digested with trypsin. Cross‐linked peptides and acetylated peptides were enriched and analyzed by liquid chromatography‐mass spectrometry.

Results: The combined acetylome/cross‐linked peptide pair analysis resulted in the identification of 1165 non‐redundant acetylated peptides from 430 proteins and 1544 cross‐linked peptide pairs representing 275 protein pairs. Fifty‐six percent of the cross‐linked proteins were also identified as acetylated. Increased acetylation was observed in the Sir3KO mice, including well known Sir3 targets such as ATP synthase subunit O and the alpha subunit of the mitochondrial trifunctional protein.

Conclusion: Integration of acetylation and cross‐linked site information was mapped onto structural models to reveal acetylation driven structural changes to these enzymes.

6. Design/engineering: 24. Therapeutics and antibodies

ABS146

GENERATION OF POLYMERIC RECOMBINANT HEMOGLOBIN USING PEG‐AZIDE DENDRIMERS AND DBCO‐MODIFIED HEMOGLOBIN

Dedeepya Gudipati ¹, Johann Sigurjonsson¹, Leah Huey¹, Spencer Anthony‐Cahill¹

¹Western Washington University (Bellingham, United States)

Studies have shown that increasing the size of cell‐free hemoglobin (Hb) molecules through chemical crosslinking with glutaraldehyde or addition of polyethylene glycol (PEG) chains reduces vasoactivity and the associated undesirable side‐effects following intravenous injection of cell‐free Hb in animals (1). The long‐term goal of our lab is to develop a monodisperse polymeric hemoglobin‐based oxygen carrier (HBOC), with defined molecular weight. To achieve this goal, Hb molecules will be linked to an azide‐dendrimer scaffold through click chemistry, which occurs between an azide group and a cyclooctyne group of another molecule. Previous work has established that we can produce recombinant Hb conjugated to a dibenzoocycloocytyne (DBCO) using sortase enzyme from S. aureus. The recombinant hemoglobin containing a C‐terminal sortase recognition motif (LPETG) was linked to azido‐FITC as a proof of concept. The successful conjugation of the Hb to FITC was confirmed through SDS‐page electrophoresis, fluorescence detection and ESMS analysis. The next step in making polymeric Hb is to react this DBCO‐Hb with a molecule that contains multiple azide groups. Initial attempts at linking DBCO‐Hb to two‐arm and four‐arm azido‐PEG crosslinkers yielded no detectable poly‐Hbs. Current work is focused on optimizing conditions which will allow the modified hemoglobin molecules to react with azido‐PEG dendrimers and other polyazides.

(1) A.F. Palmer and M. Intaglietta, Blood Substitutes, Ann. Rev. Biomed. Eng. 16: 77‐101 (2014); C.L. Varnado, T.L. Mollan, I. Birukou, B.J. Smith, D.P. Henderson and J.S. Olson, Development of recombinant hemoglobin‐based oxygen carriers, Antioxid. Redox. Signal. 18: 2314‐2328 (2013).

9. Evolution: 21. Structure (x‐ray/NMR/EM)

ABS149

INDIRECT SEXUAL SELECTION DRIVES RAPID EVOLUTION OF AN INTRINSICALLY DISORDERED SPERM PROTEIN

Damien Wilburn ¹, Lisa Tuttle¹, Rachel Klevit¹, Willie Swanson¹

¹University of Washington (Seattle, United States)

In all animals, genes that encode the sperm proteome often evolve faster than the rest of the genome. Sexual selection can create coevolutionary chases and rapid evolution of sperm proteins that recognize egg receptors, but previously not explained accelerated evolution of other sperm proteins. The marine mollusk abalone is a classic model of reproduction where the coevolving proteins sperm lysin and egg VERL interact to mediate species‐specific fertilization. For sperm to successfully compete against their brethren and win the fertilization race, they secrete enormous quantities of lysin that are packaged at intracellular concentrations of 0.1‐1.0 M. Here we describe how another rapidly evolving sperm protein enables tight lysin packaging via the novel mechanism of forming Fuzzy Interacting Transient Zwitterion (FITZ) complexes. This protein, the FITZ Anionic Partner (FITZAP), is an intrinsically disordered protein that interacts with lysin at its VERL binding interface to form fast exchanging heterodimers. FITZAP contains an N‐terminal string of aspartates that impart a negative charge and complement the high positive charge of lysin. Under physiological salt conditions, the heterodimeric FITZ complexes electrostatically associate, but upon secretion into high ionic strength sea water, these electrostatic interactions are disrupted, the FITZ heterodimers dissociate, and lysin is liberated to interact with the egg. As lysin adapts to maintain interactions with VERL via direct sexual selection, and FITZAP is recognizing this same interface, VERL is imposing indirect sexual selection on FITZAP. This tethered molecular arms race may provide a framework for understanding the widespread pattern of rapid sperm protein evolution.

11. Intrinsically disordered proteins: 16. Protein interactions and assemblies

ABS150

STRUCTURAL BASIS FOR BINDING OF AMOTL1 TO THE WW DOMAIN PROTEINS, YES‐ASSOCIATED PROTEIN AND KIBRA

Amber Vogel ¹, Ethiene Kwok¹, Diego Rodriguez¹, Afua Nyarko¹

¹Oregon State University (Corvallis, United States)

Kibra (kidney and brain expressed protein) and AmotL1 (Angiomotin‐like 1) are scaffolding proteins linked to cell polarity, memory performance, and cognition. These two proteins also regulate cell proliferation, directly or indirectly, by targeting Yes‐associated protein (Yap), a transcriptional coactivator linked to cell proliferation and tumor development. Yap and Kibra each contain two tandem WW domains, a 40‐residue interaction module composed of three antiparallel ‐strands and two conserved tryptophan residues, which bind (L/P)PxY motifs, where L is leucine, P is proline, Y is tyrosine, and x is any amino acid. AmotL1 contains three (L/P)PxY motifs, and while its interaction with Kibra and Yap has been reported in cell culture models, the structural bases of the interactions are not well understood. Using circular dichroism (CD), isothermal titration calorimetry (ITC) and analytical size exclusion chromatography (SEC), we show that Kibra and Yap bind adjacent segments in a primarily disordered region of AmotL1 with micromolar binding affinities, and that the three proteins form a ternary complex in solution. We use NMR to identify specific residues mediating each set of interactions, and to probe the conformational changes associated with complex formation. Our findings provide novel, molecular‐level insights that extend understanding of the diverse functions of Kibra and AmotL1.

21. Structure (x‐ray/NMR/EM): 5. Computational modeling/simulation

ABS151

EXAMINATION OF SUBSTRATE BINDING AND SPECIFICITY IN A PEPX FROM L. HELVETICUS

Nicholas Bratt¹, Deanna Ojennus ¹, Tersa Almaw¹, Kent Jones¹, Douglas Juers²

¹Whitworth University (Spokane, United States); ²Whitman College (Walla Walla, United States)

Prolyl aminodipeptidases are serine peptidases that hydrolyze peptide bonds at the N‐terminus of substrates, liberating an Xaa‐Pro dipeptide. The structure of a prolyl aminodipeptidase (PEPX) from Lactobacillus helveticus was solved by X‐ray crystallography at 2.0 å resolution. The structure was found to be very similar to the known PEPX homolog from Lactococcus lactis with four distinct domains including the catalytic domain which exhibits a canonical alpha/beta hydrolase fold. All the residues at the active site that potentially interact with substrate are strictly conserved between the two enzymes with the exception of a single phenylalanine which is found to be a cysteine in the L. helveticus PEPX. This small difference in binding site residues may explain reported differences in specificity between the L. lactis and L. helveticus enzymes which varies with the identity of the N‐terminal amino acid on the substrate. To examine the interactions that must occur for substrate binding, peptides were docked to the L. helveticus PEPX structure and the observed binding modes compared to known structures of human dipeptidyl peptidase‐4 (DPP‐IV) bound to substrates, mimics, and inhibitors.

11. Intrinsically disordered proteins: 16. Protein interactions and assemblies

ABS152

MUTUAL COMMUNICATION BETWEEN THE KIBRA WW DOMAINS MODULATES INTERACTIONS WITH LATS1

Kasie Baker ¹, Ethiene Kwok¹, Diego Rodriguez¹, Afua Nyarko¹

¹Oregon State University (Corvallis, United States)

Large tumor suppressor 1 (LATS1) kinase and kidney and brain expressed protein (Kibra) are members of the Hippo signaling pathway an evolutionarily conserved regulator of cell and organ growth. Kibra contains two WW domains, small globular domains that bind PPxY motifs (P = Proline, Y = Tyrosine, X = any amino acid), and LATS1 contains two PPPY motifs. Although it is well established that Kibra and LATS1 form a WW domain‐PPPY mediated complex, how binding is coordinated is unclear. Here we use constructs which contain all putative binding domains/motifs in isothermal titration calorimetry (ITC) studies to show that the isolated Kibra WW domains can bind the LATS1 PPPY motifs with weak affinity but mutual communication between the domains in the tandem WW domain construct enhance binding. Nuclear magnetic resonance (NMR) experiments reveal that chemical shift changes in LATS1 upon binding Kibra are for the most part restricted to the PPPY motifs and the eight flanking residues. These results show that the KibraLATS1 interaction is highly specific, does not involve interactions with the inter‐PPPY motif linker segment, and is fine‐tuned by mutual communications between the Kibra WW domains.

19. Proteostasis and quality control: 10. Folding

ABS153

EVALUATING BIOPHYSICAL CONSTRAINTS ON THE SEQUENCE OF RHODOPSIN BY DEEP MUTATIONAL SCANNING

Wesley Penn ¹, Andrew McKee¹, Veronica Nash¹, Charles Kuntz¹, Timmothy Gruenhagan¹, Hope Hicks², Francis Roushar¹, Mahesh Chandak¹, Christopher Hemmermich¹, Douglas Rusch¹, Jens Meiler², Jonathan Schlebach¹

¹Indiana University (Bloomington, United States); ²Vanderbilt University (Nashville, United States)

The selection pressures that shape protein structure and function are sometimes at odds with the biophysical constraints of protein folding. In addition to those associated with protein folding, membrane proteins are also constrained by the energetics associated with their solvation within the lipid bilayer. We recently found the expression and maturation of rhodopsin to be limited by the hydrophobicity of its seventh transmembrane domain (TM7), which contains several polar residues that are essential for function. The topological preferences of TM7 essentially mediate the propensity of this receptor to achieve its correct topology within the membrane and to form its native binding pocket. Based on these observations, we suspected the efficiency of rhodopsin biosynthesis should be more sensitive to mutations within TM7 relative to those that fall within a more hydrophobic domain like TM2. To compare biosynthetic sequence constraints within TMs 2 & 7, we employed Deep Mutational Scanning to measure the expression levels of ~3,000 single‐codon variants bearing mutations within these domains. Our results confirm that a higher proportion of mutations within TM7 (50%) decrease the yield of mature rhodopsin at the plasma membrane relative to those in TM2 (34%). Moreover, we find that most TM7 variants cannot be stabilized by rhodopsins retinal cofactor, which likely reflects the propensity of mutations within TM7 to promote topological defects. Taken together, these results suggest solvation energetics may significantly restrict the evolutionary sequence space that is accessible to polar TM domains. Our findings also suggest mechanistic considerations for the design of pharmacological chaperones.

19. Proteostasis and quality control: 17. Proteins in cells

ABS154

RAPID PHARMACOLOGICAL PROFILING OF GENETIC VARIANTS BY DEEP MUTATIONAL SCANNING

Francis Roushar ¹, Wesley Penn², Jonathan Schlebach²

¹Indiana University (Bloomington, United States); ²Indiana University (Bloomington, United States)

Several hundred mutations within the G‐Protein Coupled Receptor rhodopsin are known to cause retinitis pigmentosa (RP), a common form of inherited blindness. Drug discovery efforts have primarily focused on development of pharmacological chaperones that bind and stabilize the misfolded P23H rhodopsin variant, which is responsible for ~7% of all RP cases. However, these compounds are typically only effective against a subset of other RP‐causing mutations. To comprehensively assess which additional RP variants respond to currently available compounds, we developed a deep mutational scanning platform that utilized surface immunostaining to measure the effects of small molecules on the expression of 138 known RP variants in parallel. Preliminary characterization of this library suggests that it contains two classes of variants when stably expressed in HEK293T cells, including those that induce misfolding and mistrafficking (type II) and those that instead likely compromise protein function. Furthermore, we show the proportion of cells expressing properly trafficked rhodopsin variants roughly doubles in response to the investigational pharmacochaperone 9‐cis‐retinal, which confirms that only a subset of the misfolded variants can be rescued by this compound. We will use this approach to measure the effects of a panel of pre‐clinical pharmacochaperones and proteostasis regulators on the trafficking of the entire spectrum of known RP mutants. Additional efforts are needed to determine the effects of these compounds on the functional properties of RP variants. Nevertheless, this study will advance our understanding of the molecular basis of RP and provide new tools to address current challenges in precision medicine.

ABS155

BIOTIN BINDER DESIGN USING DE NOVO PROTEIN SCAFFOLDS

Gyu Rie Lee ¹, Anastassia Vorobieva¹, Brian Weitzner¹, Benjamin Basanta¹, David Baker¹

¹University of Washington (Seattle, United States)

Recently there has been a major breakthrough in designing a de novo functional protein. A fluorescence‐activating beta barrel was designed using Rosetta. Rosetta is a biomolecular modeling software which has been used in successful cases of designing novel proteins with different folds and function. The distinguishing feature of the recent functional protein design was that it was based on a de novo beta barrel scaffold. However, challenges still remain in extending the computational method to robustly design proteins that can specifically bind to other small molecules. One of our approaches to tackle this problem is designing diverse de novo scaffolds. We have designed diverse beta barrel scaffolds in the hope of sampling scaffolds which can accommodate small molecules of our interest. Thousands of backbone structures were modeled and the sequences were designed using Rosetta. As a test case, we aimed to design a biotin binding protein using the generated scaffolds. The computational process to design a functional protein roughly followed that of designing the fluorescence‐activating beta barrel. De novo protein scaffolds which can have local interactions with the target small molecule, biotin, were initially screened using a computational method called RIF Dock. The side chains backing up the ligand‐interacting residues were re‐designed using Rosetta. The synthesized genes of the selected designs were expressed in yeast and/or E.coli. They were further experimentally tested for binding of biotin by flow cytometric analysis and fluorescence polarization.

11. Intrinsically disordered proteins: 10. Folding

ABS156

AN EXAMINATION OF THE SURFACE OF THE INTRINSICALLY DISORDERED PROTEIN ALPHA SYNUCLEIN

María Rocío Rial Hawila¹, José María Delfino ¹, Gabriela Elena Gómez¹

¹University of Buenos Aires (Buenos Aires, Argentina)

The nature and size of the accessible surface area (SASA) of the polypeptide chain plays a pivotal role in protein folding and complex formation. To investigate SASA, we employ diazirine (DZN), a minute precursor of the extremely reactive methylene carbene (:CH2). Methylation signatures left on the polypeptide provide telltale clues on conformation and interactions. The extent of methylation (EM) metric derived straightforwardly from mass spectra (ESI‐MS or MALDI‐TOF) discriminates between native and alternate states. The archetypical IDP human alpha synuclein (AS) aggregates into amyloid fibrils, constituents of Lewy bodies, a hallmark of Parkinson disease. DZN labeling proves particularly fit to analyzing the conformational plasticity inherent to IDPs, an elusive goal by other classical biophysical methods. Unlike well‐structured proteins where the methylation signal differs strikingly between native and unfolded states, AS in buffer or equilibrated in 6 M GdmCl display a similarly enhanced EM value, pointing to the high solvent exposure of AS under physiological conditions. Interestingly, compaction of monomeric AS due to calcium binding yields a somewhat decreased EM value. Remarkably, AS fibrils cause a larger fall in the EM value, a consequence of surface occlusion at an interface. Tryptic fragmentation of AS coupled to MALDI‐TOF analysis is revealing methylation patterns at increased resolution. The peptide coverage achieved so far allows to construct a solvent‐accessibility map of the different sequential domains of AS. This new information illuminates the role played by the constituent parts in the monomeric ensemble as well as the changes observed en route to the fibrillar aggregate.

11. Intrinsically disordered proteins: 26. Other

ABS158

COMBINING SMFRET AND DEER DISTANCE MEASUREMENTS TO CHARACTERIZE DISORDERED PROTEINS

Tatyana Smirnova¹, Tatyana Smirnova ¹, Keith Weninger¹, Hugo Sanabria²

¹North Carolina State University (Raleigh, United States); ²Clemson University (Clemson, United States)

Intrinsically disordered proteins (IDPs) are fascinating biomolecules that lack well‐defined structures in their free states under native‐like conditions. Here we report on combining spin‐labeling Double Electron‐Electron Resonance (DEER) experiments and smFRET measurements to characterize two IDPs. While neuronal SNARE protein SNAP‐25 is a highly disordered protein in isolation, it folds into a stable ‐helix bundle upon forming the SNARE complex. Results of smFRET and DEER measurements of distances and distance distributions, from identical labeling positions, agree well once the difference in size of the labels is considered. While for the disordered SNAP‐25 DEER experiments reveal two distance distributions, smFRET detects only one population with the mean distance of 4.3 nm. It is suggested that the two population ensembles are not detected by sm TIRFM FRET because of a fast interconversion of the two forms that is frozen out in low temperature DEER. Measurements using FRET in a Multiparameter Fluorescence Detection (MFD) mode confirmed ms interconversion in S25. In ID C‐terminal domain of the NR2B subunit of the NMDA receptor (N2B), DEER and sm‐FRET data report on one long distance conformation centered at 7.3 nm and another conformation characterized by a broader distribution at shorter distances with TIRFM FRET showing stochastic switching between two distinct disordered conformational ensembles on a second timescale. FRET in MFD mode on the N2B showed a complex dynamic exchange among at least three states.

In summary, a combination of smFRET and DEER offers complimentary data on conformations of IDPs.

16. Protein interactions and assemblies: 21. Structure (x‐ray/NMR/EM)

ABS160

MAINTENANCE OF ALPHA‐HELICES IN NON‐IDEAL DIMERIC PLASMINOGEN‐BINDING GROUP A STREPTOCOCCAL M‐PROTEINS DETERMINES THEIR TIGHT BINDINGS TO HUMAN PLASMINOGEN

Cunjia Qiu ¹, Yue Yuan², Zhong Liang², Shaun Lee³, Victoria Ploplis², Francis Castellino²

¹Department of Chemistry and Biochemistry, University of Notre Dame (Notre Dame, United States); ²W. M. Keck Center for Transgene Research, University of Notre Dame (Notre Dame, United States); ³Department of Biological Sciences, University of Notre Dame (Notre Dame, United States)

Group A Streptococcus (GAS) is one leading infectious cause of human mortality. Skin‐tropic pattern D GAS strains express a special virulence factor, plasminogen‐binding group A Streptococcal M‐protein (PAM) that specifically recruits human plasminogen (hPg) from host. In this investigation, we combine several biophysical techniques, especially CD, AUC, NMR and SPR, to dissect the PAM structure and determine how structural changes influence the binding affinity of PAM to hPg.

We herein constructed seven PAMs originally expressed from different pattern D GAS strains. All the recombinant PAMs consist of hypervariable region, A‐, B‐, C‐, D‐domains, and a Pro/Gly‐rich region. We have shown that each PAM forms a dimer in solution at 25 °C, and c‐repeats in the C‐domain constitute the most important part in PAM dimerization. Solution structures of peptides that contain A‐ and B‐domains demonstrated that both domains are helix‐destabilizing and thus contribute little to dimerization. We accordingly established a crucible‐tongs structural model to depict non‐ideal dimeric PAMs. However, these PAMs dissociate into unstructured monomers at 37 °C, to different extent. Although all PAMs tightly bind to hPg at a nM range at 25 °C, dissociation occurring at 37 °C makes Class II PAMs containing only a2‐repeat in the A‐domain bind to hPg ~1,000 weaker. But it impacts little on the hPg‐binding capacity of Class I/III PAMs containing intact a1a2‐repeats. We conclude that existence of the a1‐repeat protects alpha‐helical contents in the a2‐repeat, making binding residues orientate correctly, while this protection disappears in Class II PAMs missing the a1‐repeat.

21. Structure (x‐ray/NMR/EM): 7. Dynamics and allostery

ABS162

THE ROLE OF CONFORMATIONAL DYNAMICS IN SHEAR‐ENHANCED FIMH‐MEDIATED BACTERIAL ADHESION

Pearl Magala ¹, Dagmara Kisiela¹, Angelo Ramos¹, Wendy Thomas¹, Evgeni Sokurenko¹, Rachel Klevit¹

¹University of Washington (Seattle, United States)

Pathogenic bacteria, such as Escherichia coli (E. coli) establish infections in the intestinal and urinary tracts by adhering to host epithelial cells via the adhesion protein FimH. However, the intestinal and urinary tracts constantly undergo flow conditions that generate shear forces, which could serve to pull bacteria off the host cell. Instead, bacterial adhesion strengthens in the presence of shear forces to prevent elimination from the host. FimH is located at the tip of the bacterial pili and mediates the adhesion of bacteria to host cells by specifically recognizing and binding to mannose ligands in host glycoprotein receptors the first step in establishing infections. In contrast to most biomolecular interactions that weaken when pulled on by force, interactions between FimH and mannose strengthen under shear force conditions. FimH binds ligands via its lectin domain and is anchored to the rest of the bacterial pili by its pilin domain. Ligand binding by FimH is accompanied by substantial structural rearrangements, in which its two domains dissociate from each other and its binding pocket closes around the ligand. Existing crystal structures of FimH do not provide clarity regarding how the two conformational alterations are coupled allosterically. We are using solution Nuclear Magnetic Resonance spectroscopy to examine the conformational dynamics and solutions structures of FimH to define the pathways and intermediate states that the adhesin adopts to promote ligand binding and shear enhanced FimH mediated bacterial adhesion.

2. Bioinformatics: 26. Other

ABS163

ENGAGING THE PROTEIN SCIENCE COMMUNITY TO EXPAND PROTEIN LITERATURE REPRESENTATION AND ANNOTATIONS IN UNIPROT

Cecilia Arighi ¹, Hongzhan Huang², Yongxing Chen³, Qinghua Wang³, Peter McGarvey⁴, Cathy Wu³, UniProt Consortium⁵, UniProt Consortium⁶, UniProt Consortium⁷

¹Protein Information Resource, University of Delaware (Newark, United States); ²Protein Information Resource, University of Delaware (Newark, United States); ³Protein Information Resource, University of Delaware (Newark, United States); ⁴Protein Information Resource, Georgetown University Medical Center (Washington DC, United States); ⁵EMBL‐EBI (Hinxton, United Kingdom); ⁶SIB (Geneva, Switzerland); ⁷Protein Information Resource (Newark/Washington DC, United States)

UniProt Knowledgebase is a publicly available database with access to a vast amount of protein sequence and function information. Expert curation at UniProt includes a critical review of experimental data from the literature and predicted data from a range of sequence analysis tools. A representative set of literature articles is selected for annotation. Thus, many articles with potentially useful content may not be included. Also, UniProt expert curation focuses on selected species with proteins from many organisms not being actively annotated.

To facilitate access to more comprehensive literature related to entries, UniProt compiles and organizes publications from external biological databases and text mining results. For a better user experience, this bibliography is classified, via a neural network‐based method or based on the source databases, into the different topics in the entry, similar to the curated references. These publications are available under Computationally mapped in the publication section of the protein entry.

Still, many experts request articles and annotations to be added to protein entries. To respond to this need, we are developing a Community section where researchers are able to add directly the articles that they deem relevant to an entry, along with performing optional annotation tasks. ORCIDs will be used to validate and give credit to the contributors.

With the community expert contributions, UniProt will enable access to a more comprehensive set of articles and annotations, enabling discovery and benefitting the wider protein science community.

24. Therapeutics and antibodies: 7. Dynamics and allostery

ABS164

DISCOVERY OF A NOVEL SMALL‐MOLECULE ACTIVATOR THAT CORRECTS G6PD DEFICIENCY

SUNHEE HWANG ¹

¹STANFORD UNIVERSITY (STANFORD, United States)

Glucose‐6‐phosphate dehydrogenase (G6PD) deficiency, one of the most common human genetic disorders, is caused by over 160 different point mutations and results in a variety of health problems, including hemolytic anemia and neurological damage. As G6PD is a major source of protective anti‐oxidant through NADPH production, G6PD deficiency likely contributes to the severity of many other acute and chronic diseases associated with increased oxidative stress. However, no medications are available to treat G6PD deficiency directly; thus, we sought to identify a small molecule that increases catalysis of G6PD mutant enzymes. Crystal structure and mutagenesis analyses of one common mutant (Canton, R459L) provided insight into the functional defect of the enzyme. Using high‐throughput screening, we identified AG1, a small molecule that increases the activity and/or stability of the wild‐type, the Canton mutant and several other common human G6PD mutants as well. AG1 reduced oxidative stress and increased NADPH levels in vivo, in zebrafish, and prevented chloroquine‐induced hemolysis in human erythrocytes. Our study suggests that a pharmacological agent, of which AG1 may be a lead, will likely alleviate the clinical problems associated with multiple variants of G6PD deficiency.

19. Proteostasis and quality control: 12. Membrane proteins

ABS165

MOLECULAR BASIS FOR THE EVOLVED INSTABILITY OF A HUMAN G‐PROTEIN COUPLED RECEPTOR

Laura Chamness ¹, Charles Kuntz¹, Wesley Penn¹, Jens Meiler², Jonathan Schlebach¹

¹Indiana University (Bloomington, United States); ²Vanderbilt University (Nashville TN, United States)

Despite millions of years of evolution, many eukaryotic membrane proteins remain predisposed to misfolding and degradation within the endoplasmic reticulum (ER). One extreme example is the gonadotropin‐releasing hormone receptor (GnRHR), which is a G‐protein coupled receptor involved in reproductive steroidogenesis. Though non‐mammalian forms of GnRHR exhibit robust trafficking to the plasma membrane, mammalian GnRHRs tend to misfold and accumulate within the ER. There is significant evidence to suggest the inefficient folding and trafficking of mammalian GnRHRs is a product of evolutionary selection. Nevertheless, the mechanistic basis for this evolved instability of mammalian GnRHRs has yet to be explored. To evaluate the evolutionary modifications that coincided with this proteostatic divergence, we compiled and analyzed the sequences of known type I GnRHRs. Multiple sequence alignments reveal that the transmembrane (TM) helices of mammalian GnRHRs are considerably more polar than those of non‐mammalian GnRHRs. We show that this enhanced polarity compromises the translocon‐mediated membrane integration of TMs 2 and 6, which suggests this increase in polarity induces cotranslational misfolding. Structural models of these receptors suggest two of the key side chains that increase the polarity of TM6 are projected into the membrane core, and are therefore unlikely to be involved in native conformational transitions. Moreover, mutations that re‐introduce the hydrophobic side chains to these positions are capable of partially restoring the plasma membrane expression of human GnRHR. Together, these findings suggest that the effects of mutations on the fidelity of cotranslational folding contributed to evolutionary modifications of the plasma membrane expression of GnRHRs.

6. Design/engineering: 16. Protein interactions and assemblies

ABS166

PROGRAMMABLE DESIGN OF ORTHOGONAL PROTEIN HETERODIMERS

Zibo Chen ¹, Scott Boyken¹, Mengxuan Jia², Florian Busch², David Flores‐Solis³, Matthew Bick¹, Peilong Lu¹, Zachary VanAernum², Aniruddha Sahasrabuddhe², Robert Langan¹, Sherry Bermeo¹, T. J Brunette¹, Vikram Mulligan¹, Vicki Wysocki², Frank DiMaio¹, David Baker¹

¹University of Washington (Seattle, United States); ²The Ohio State University (Columbus, United States); ³University of California Santa Cruz (Santa Cruz, United States)

Specificity of interactions between two DNA strands, or between protein and DNA, is often achieved by varying bases or side chains coming off the DNA or protein backbone. By contrast, specificity of proteinprotein interactions usually involves backbone shape complementarity, which is less modular and hence harder to generalize. Coiled‐coil heterodimers are an exception, but the restricted geometry of interactions across the heterodimer interface limits the number of orthogonal pairs that can be created simply by varying side‐chain interactions. Here we show that proteinprotein interaction specificity can be achieved using extensive and modular side‐chain hydrogen‐bond networks. We used the Crick generating equations to produce millions of four‐helix backbones with varying degrees of supercoiling around a central axis, identified those accommodating extensive hydrogen‐bond networks, and used Rosetta to connect pairs of helices with short loops and to optimize the remainder of the sequence. Of 97 such designs expressed in Escherichia coli, 65 formed constitutive heterodimers, and the crystal structures of four designs were in close agreement with the computational models and confirmed the designed hydrogen‐bond networks. In cells, six heterodimers were fully orthogonal, and in vitrofollowing mixing of 32 chains from 16 heterodimer designs, denaturation in 5 M guanidine hydrochloride and reannealingalmost all of the interactions observed by native mass spectrometry were between the designed cognate pairs. The ability to design orthogonal protein heterodimers should enable sophisticated protein‐based control logic for synthetic biology, and illustrates that nature has not fully explored the possibilities for programmable biomolecular interaction modalities.

8. Enzymology: 22. Synthetic biology

ABS167

DESIGN OF CATALYTIC DE NOVO PROTEINS

Shane Caldwell ¹, Shane Caldwell¹, Susana Vazquez Torres¹, Cathleen Zeymer², Ian Haydon¹, Don Hilvert², David Baker¹

¹University of Washington (Shoreline, United States); ²ETH Zurich (Zurich, Switzerland)

Despite many recent successes in the field of de novo protein design, catalytic proteins ‐ enzymes ‐ are not yet easily produced. Some of this challenge has resulted from the fact that catalytic residues often cannot be introduced into a de novo protein scaffold without disrupting non‐catalytic properties such as domain stability and folding kinetics. However, an increasing array of de novo scaffolds available for design has dramatically expanded the protein architectures in which we may engineer catalytic function. By combining multiple de novo protein domains into multi‐domain proteins with close packing between domains, we can produce clefts and active sites without compromising the independent folding and stability of the domains. Extending this logic forward, the modular combination of domains can then be mixed and matched to custom‐build a catalytic site for any reaction of interest.

12. Membrane proteins: 15. Peptides

ABS168

NANOPORE‐CONFINED LIPID BILAYERS FOR ORIENTED SAMPLE EPR AND NMR STUDIES OF MEMBRANE PROTEINS

Sergey Milikisiyants¹, ALEX SMIRNOV ², Melanie Chestnut², Morteza Jafarabadi², Antonin Marek², Maxim Voinov², Alexander Nevzorov²

¹NCSU (RALEIGH, United States); ²North Carolina State University (Raleigh, United States)

Among many membrane mimetic systems currently in use, fully hydrated lipid bilayers represent an optimal environment for membrane proteins to fold and function. Previously, we have developed methods for forming self‐assembled nanotubular bilayers inside cylindrical nanopores of anodic aluminum oxide (AAO). Such hybrid nanostructures, named lipid nanotube arrays, represent a new type of substrate‐supported lipid bilayers that are minimally perturbed but aligned macroscopically to a rather high degree. Here we demonstrate a dramatic improvement in this method by developing AAO substrates with exceptionally uniform high‐density nanopore structure and optimizing lipid loading protocols. We then employed oriented sample (OS) solid state NMR for an examination of lipid‐induced conformational changes of Pf1 coat protein over exceptionally broad range of environmental conditions, including temperature, pH, and lipid composition. We also extended the oriented sample (OS) lipid nanotube technology to spin labeling EPR including W‐band (94.3 GHz) (HF EPR) spectroscopy. For the latter, a new photonic band‐gap W‐band resonator to accommodate planar AAO substrates has been developed and tested. Such a resonator/AAO system provides additional resolution for studies of conformational changes of model ion channels by the spin labeling EPR. We also provide an initial demonstration of long‐range distance measurements between spin‐labeled sites by carrying out DEER from a membrane protein aligned by lipid nanotube array method.

6. Design/engineering: 16. Protein interactions and assemblies

ABS171

DESIGN OF NOVEL LECTINS BY COMPUTER‐AIDED DIRECTED EVOLUTION

Prerna Sharma ¹, Ismail C. Kazan², Sefika B. Ozkan², Giovanna Ghirlanda³

¹Arizona State University (Tempe, United States); ²Centre for Biological Physics, Department of Physics, Arizona State University (Tempe, United States); ³School of Molecular Sciences, Arizona State University (Tempe, United States)

Glycosylation is one of the most prevalent post‐translational modifications, mediating complex biological processes. Despite its importance, a comprehensive mapping and understanding of the glycome faces tremendous challenges due to inherent difficulties in identifying and tracking glycosylation. To address these challenges, we employed computer‐guided directed evolution to design and characterize novel lectins with improved binding affinities.

Our starting point is a small 11kDa lectin, cyanovirin‐N (CVN), which exhibit potent antiviral activity by binding to highly branched oligomannosides (Man9) of gp120. CVN comprises two glycan‐binding domains, and undergoes domain‐swapped dimerization. First, we designed a monomeric version, P51Gm4, which binds dimannose with Kd =114 μM through a single carbohydrate‐binding pocket. To identify remote positions participating in binding site conformational dynamics, we subjected P51Gm4 to dynamic coupling analysis (DCI) and multiple sequence alignment (MSA) to identify co‐evolved sites. Then analysis was coupled to flexible docking tool (Adaptive BP‐Dock) which screened mutants based on changes in binding energies. Experimental characterization of several mutants identified I34Y (located far from binding site), binding dimannose with increased affinity as shown by ITC measurements (Kd =57 μM). The crystal structure of I34Y is superimposable to P51Gm4, however, molecular dynamic simulations identified a change in the frequency of opening and closing of the binding site. The improved affinity indicates that I34 is dynamically coupled to the binding site and mediates spatial contacts to distal residues within tertiary structure of the protein. This initial finding can help generate a mutational landscape for designing lectins with better affinity.

7. Dynamics and allostery: 21. Structure (x‐ray/NMR/EM)

ABS173

STAPHYLOCOCCUS AUREUS ISDH: THE CHEMICAL AND DYNAMIC BASIS OF HEME EXTRACTION FROM HUMAN HEMOGLOBIN

Ken Ellis‐Guardiola ¹, Joseph Clayton², Brendan Mahoney¹, Clarissa Pham¹, Sinan Sabuncu³, Jeff Wereszczynski², Pierre Moenne‐Loccoz³, Robert Clubb¹

¹UCLA (LOS ANGELES, United States); ²Illinois Institute of Technology (Chicago, IL, United States); ³Oregon Health & Science University (Portland, OR, United States)

Iron is a versatile enzyme metal cofactor that is used in a wide‐range of essential cellular processes. During infections bacterial pathogens acquire iron from human hemoglobin (Hb), which contains the majority of the body's total iron in the form of heme. We have gained insight into how pathogenic Staphylococcus aureus acquires heme from Hb, using NMR, crystallography, quantitative stopped‐flow heme transfer measurements and computational methods. Heme capture is mediated by the IsdH receptor, which engages Hb via a tri‐domain unit containing N2 and N3 NEAT domains that are separated by a helical linker segment. Our isothermal titration calorimetry experiments and crystal structures of the IsdH‐Hb complex reveal two energetically distinct protein‐protein interfaces that work together to promote heme release. A high‐affinity receptor‐Hb(A‐helix) interface contributes 95% of the total binding standard free energy, enabling much weaker receptor interactions with Hb's F‐helix. Using a receptor mutant that only binds to the ‐subunit of Hb and a stopped‐flow transfer assay, we rigorously determined the energetics and micro‐rate constants of heme extraction. The receptor accelerates heme release ~13,400‐fold and is limited primarily by an enthalpic barrier. Interestingly, NMR and computational studies of the 175 kDa Hb‐receptor complex reveal that the extraction interface forms transiently to dislodge heme. A unifying computational model of the transfer pathway consistent with experimental data is presented. Our results shed light on to the molecular mechanism used by bacteria to capture heme and could lead to new anti‐infective agents that work by limiting microbial access to iron.

7. Dynamics and allostery: 1. Amyloid and aggregation

ABS174

ALTERED DYNAMICS OF CATARACTS‐ASSOCIATED S‐CRYSTALLIN MUTANTS MEASURED BY NMR

Heather Forsythe ¹, Kayla Jara², Calvin Vetter³, Patrick Reardon⁴, Elisar Barbar², Kirsten Lampi³

¹Oregon State University (Corvallis, United States); ²Biochemistry & Biophysics, Oregon State University (Corvallis, OR, United States); ³Integrative Biosciences and Biochemistry & Molecular Biology, Oregon Health & Science University (Portland, OR, United States); ⁴Nuclear Magnetic Resonance Facility, Oregon State University (Corvallis, OR, United States)

Cataract formation, resulting from crystallin protein aggregation is still a poorly understood process. In the aging lens, crystallins are known to undergo significant levels of post‐translational modification including, deamidation. S‐crystallin is the most highly expressed crystallin in the cortex of the lens and is one of the most highly conserved protein sequences across mammalian species. In the cataracts‐prone lens, key residues of S‐crystallin have been identified as having elevated rates of deamidation, but how deamidation results in aggregation is unclear. Using nuclear magnetic resonance (NMR), differences in chemical shifts, 15N backbone relaxation, and the hydrogen‐deuterium exchange rates between Asn‐to‐Asp mutants and WT‐S have been characterized. Additionally, spin‐relaxation data was analyzed further to probe protein dynamics behavior from pico‐to‐millisecond time scales. Compared to WT, mutants show significantly altered flexibility and solvent accessibility (representative mutant, Figure 1). Furthermore, mutants are found to have altered backbone conformational entropy profiles. Residue‐specific differences appear both near and distal to the sites of mutations, suggesting that deamidation causes global changes to S stability. Results support a model that deamidation drives changes in conformational entropy, backbone dynamics, and solvent accessibility. These changes may have related effects, allowing for elevated incidents of interaction between nearby crystallin monomers, and ultimately resulting in protein aggregation.

7. Dynamics and allostery: 20. Single molecule studies

ABS175

FORCE‐DEPENDENT ALLOSTERIC ENHANCEMENT OF AE‐CATENIN BINDING TO F‐ACTIN BY VINCULIN

Nicolas Bax ¹, Derek Huang², Sabine Pokutta¹, Alexander Dunn³, William Weis¹

¹Structural Biology, Stanford University (Palo Alto, United States); ²Biophysics, Stanford University (Stanford, United States); ³Chemical Engineering, Stanford University (94301, United States)

The linkage between cadherin‐based transmembrane adhesions and the actomyosin cytoskeleton is a fundamental feature of metazoan tissues. Mechanical tension on the minimal cadherin‐catenin complex increases the actin‐binding lifetime of alpha‐catenin (termed catch‐bond behavior) and reveals a cryptic binding site in aE‐catenin for the actin‐binding protein vinculin. While vinculin is thought to reinforce the cadherin‐catenin/actin linkage, the mechanistic details of this process remain poorly understood. Although the actin‐binding activity of vinculin likely contributes to junctional strengthening, we find that that complex formation with an aE‐catenin binding domain from vinculin led to longer‐lived actin binding, but in a manner that was enhanced under force. Computational modeling suggests that this force‐dependent strengthening of individual bonds results in adhesions with increased resilience to fluctuating loads and higher energetic efficiency in force transmission at cell‐cell junctions. Our results demonstrate a form of force‐dependent allosteric regulation that may enhance the ability of cells to form robust connections and sense mechanical cues at cell‐cell contacts.

21. Structure (x‐ray/NMR/EM): 12. Membrane proteins

ABS177

THE CYCLIC NUCLEOTIDEBINDING HOMOLOGY DOMAIN OF THE INTEGRAL MEMBRANE PROTEIN CNNM MEDIATES DIMERIZATION AND IS REQUIRED FOR MG2+ EFFLUX ACTIVITY

Yu Seby Chen ¹, Guennadi Kozlov¹, Rayan Fakih¹, Yosuke Funato², Hiroaki Miki², Kalle Gehring¹

¹McGill University (Montreal, Canada); ²Osaka University (Osaka, Japan)

Proteins of the cyclin M family (CNNMs; also called ancient conserved domain proteins, or ACDPs) are represented by four integral membrane proteins that have been proposed to function as Mg2+ transporters. CNNMs are associated with a number of genetic diseases affecting ion movement and cancer via their association with highly oncogenic phosphatases of regenerating liver (PRLs). Structurally, CNNMs contain an N‐terminal extracellular domain, a transmembrane domain (DUF21), and a large cytosolic region containing a cystathionine‐‐synthase (CBS) domain and a putative cyclic nucleotidebinding homology (CNBH) domain. Although the CBS domain has been extensively characterized, little is known about the CNBH domain. Here, we determined the first crystal structures of the CNBH domains of CNNM2 and CNNM3 at 2.6 and 1.9 å resolutions. Contrary to expectation, these domains did not bind cyclic nucleotides, but mediated dimerization both in crystals and in solution. Analytical ultracentrifugation experiments revealed an inverse correlation between the propensity of the CNBH domains to dimerize and the ability of CNNMs to mediate Mg2+ efflux. CNBH domains from active family members were observed as both dimers and monomers, whereas the inactive member, CNNM3, was observed only as a dimer. Mutational analysis revealed that the CNBH domain was required for Mg2+ efflux activity of CNNM4. This work provides a structural basis for understanding the function of CNNM proteins in Mg2+ transport and associated diseases.

6. Design/engineering: 12. Membrane proteins

ABS180

COMPUTATIONAL DESIGN OF MULTIPASS TRANSMEMBRANE PROTEINS

Peilong Lu ¹, Chunfu Xu¹, Duyoung Min², Frank Dimaio¹, Tamer El‐Din¹, Lance Stewart¹, Justin Kollman¹, Tomoaki Matsuura¹, William A. Catterall¹, James U. Bowie¹, David Baker¹

¹University of Washington ( Seattle, United States); ²UCLA (Los Angeles, United States)

In recent years, soluble protein design has achieved successes such as artificial enzymes and large protein cages. Membrane proteins present a considerable design challenge, but here too there have been advances, including the design of a zinc‐transporting tetramer. Here, we report the design of stable transmembrane monomers, homodimers, trimers, tetramers, hexamers, and octamers with up to 16 membrane‐spanning regions in an oligomer. The designed proteins adopted the target oligomerization state and localized to the predicted cellular membranes, and crystal structures of the designed dimer and tetramer reflected the design models. Patch clamp electrophysiology experiments show that the transmembrane form of the 12‐helix hexamer pore expressed in insect cells allows passage of ions across the membrane with selectivity for potassium over sodium. The transmembrane form of the 16‐helix octamer pore, but not the 12‐helix hexamer, allows passage of biotinylated Alexa Fluor 488 when incorporated into liposomes using in vitro protein synthesis. The ability to produce structurally well‐defined transmembrane channels both in cells and in vitro opens the door to the creation of designer pores for a wide variety of applications.

11. Intrinsically disordered proteins: 16. Protein interactions and assemblies

ABS181

COMING TOGETHER IN THE DNA DAMAGE RESPONSE: INTERACTIONS WITH THE INTRINSICALLY DISORDERED REGION OF BRCA1

Christine Hurd¹, Mikaela Stewart ², Brian Morote‐Costas¹

¹Texas Christian University (Fort Worth, United States); ²Texas Christian University Biology Dept. (Fort Worth, United States)

The intrinsically disordered region of breast cancer 1 early onset protein (BRCA1) binds to the partner and localizer of BRCA2 (PALB2) in order to recruit both PALB2 and BRCA2 to areas of DNA damage. Mutations in the gene coding for PALB2 disrupt this interaction causing defects in DNA damage repair and increased risk of breast cancer. To better understand the molecular details of this important interaction, we have developed a system to study the interaction of BRCA1 and PALB2 in vitro. We find that minimized binding regions of BRCA1 and PALB2 proteins participate in a high‐affinity interaction in vitro. Our preliminary circular dichroism and nuclear magnetic resonance data conflict with the previously predicted model of BRCA1 forming a coiled coil upon interaction with PALB2. We are also using this in vitro system to mimic BRCA1 phosphorylation that occurs upon DNA damage to gauge its effect on the structure of BRCA1 and interaction with PALB2. We will present our structural findings and new insights regarding this important DNA damage response recruiting event.

21. Structure (x‐ray/NMR/EM): 5. Computational modeling/simulation

ABS182

FIVE REASONS TO PAY ATTENTION TO CASP‐SAXS

Susan Tsutakawa ¹, Greg Hura¹, Andry Kryshtafovych², Krzysztof Fidelis², John Tainer³

¹Lawrence Berkeley National Laboratory (Berkeley, United States); ²University of California, Davis (Davis, United States); ³MD Anderson (Houston, United States)

Atomic structures are critical for medical science from understanding how proteins function to drug design. Structure prediction algorithms could provide predicted structures for these purposes in the near future. Yet, there are gaps in prediction algorithms for identifying which predictions are accurate and for predicting accurately structures of large multi‐domain proteins, protein complexes, and flexible proteins. Experimental data from Small Angle X‐ray Scattering (SAXS) has the potential to validate structure predictions and to improve the accuracy of the predictions. As SAXS is a high throughput method, data can be collected at a fraction of the cost, time, and labor of other techniques (Reason 1). SAXS measures all electron pair distances within a protein molecule in solution, which provides distance information for the protein core, multimerization state, and flexibility (Reason 2). This distance information can be used as part of scoring functions and/or as restraints for structure prediction algorithms (Reason 3). To facilitate linking SAXS to structure prediction algorithms, we collaborate with the Critical Assessment of protein Structure Prediction (CASP). We found that participants made more effective use of SAXS data than in the 2016 CASP12 and were better able to improve the overall shape of their prediction (Reason 4). We also observed improved starting models and better ranking of their models and in two cases, there was significant improvement in fold (Reason 5). Although CASP13 supported the potential of SAXS for integration with protein structure prediction algorithms, it also revealed the barriers and challenges in its use.

6. Design/engineering: 22. Synthetic biology

ABS183

THE LIMITS OF TETHERING IN KINASE SIGNALING REACTIONS

Elizabeth Speltz ¹, TJ Brunette¹, Fabio Parmeggiani¹, David Baker¹, Jesse Zalatan¹

¹University of Washington (Seattle, United States)

In cell signaling pathways, scaffold proteins coordinate the assembly of multi‐protein complexes that co‐localize kinases and their substrates. To understand the relationship between structure and function in these systems, we have engineered a series of designed protein scaffolds that physically tether a kinase to its substrate with tunable control over the distance between the reacting proteins. In a simple model scaffold with flexibility, we observe significant rate enhancements that depend on scaffold structure. Surprisingly, however, the kinetic data strongly suggest that the intended target structure with 1:1:1 stoichiometry between scaffold, kinase, and substrate is unreactive; instead, a higher‐order multi‐protein assembly is likely to be the reactive species. This result has important implications for designing scaffold proteins in synthetic biology and for understanding the functions of natural scaffold proteins.

23. Systems biology: 21. Structure (x‐ray/NMR/EM)

ABS185

SYSTEMATIC IDENTIFICATION OF RECOGNITION MOTIFS FOR THE HUB PROTEIN LC8

Aidan Estelle ¹, Nathan Jespersen¹, David Hendrix¹, Ylva Ivarsson², Norman Davey³

¹Department of Biochemistry and Biophysics, Oregon State University (Corvallis, United States); ²Department of Chemistry BMC, Uppsala University (Uppsala, Sweden); ³Conway Institute of Biomolecular and Biomedical Sciences, University College Dublin (Dublin, Ireland)

Hub proteins participate in cellular regulation by dynamic binding of multiple proteins within interaction networks. The hub protein LC8 reversibly interacts with more than 100 partners through a flexible pocket at its dimer interface. Specifically, LC8 recognizes a TQT anchor motif in disordered regions of binding partners. However, despite the large number of LC8 binders, predicting new LC8 binding proteins is exceptionally challenging. To explore the diversity of the LC8 partner pool, and to improve our understanding of LC8 binding, we screened for LC8 binding partners using a proteomic peptide phage display library composed of peptides from the human proteome, which had no bias towards a known LC8 motif. Of the identified hits, we validated binding of 26 anchor‐containing peptides, 16 of which were entirely novel, using isothermal titration calorimetry. Strikingly, numerous peptides containing the TQT anchor do not bind LC8, indicating that residues outside of the anchor facilitate LC8 interactions. Using both LC8‐binding and non‐binding peptides containing the motif anchor, we developed the LC8Pred algorithm that highlights the importance of the residues flanking the anchor, and parses random sequences to predict LC8 binding sequences with ~78% accuracy. To explore the effectiveness of LC8Pred, we tested it on known LC8‐binding protein Chica, as well as the previously‐unidentified LC8‐binding protein Kank1. Our findings significantly advance the comprehensive overview of the LC8 hub interactome, and we have made a database LC8Hub of LC8‐binding interactions publicly available, along with LC8Pred.

6. Design/engineering: 5. Computational modeling/simulation

ABS188

PROTEIN SHAPE SCULPTING USING RIGID HELICAL JUNCTIONS

TJ Brunette ¹, Matthew Bick¹, Jesse Hansen¹, David Baker¹

¹University of Washington (Seattle, United States)

The ability to quickly assemble large proteins with arbitrary shapes would enable the design of protein binders that wrap around their target and the positioning of multiple functional sites in specified orientations. We describe a general method for generating rigid fusions between proteins with terminal helices, and use it to design 75k structurally unique junctions between de‐novo designed repeat proteins, homo‐oligomers and hetero‐oligomers. Of 34 junction designs that were experimentally characterized, 28 have circular dichroism and solution x‐ray scattering profiles consistent with the design models and are stable at 95 °C. Crystal structures of 4 designed junctions were in close agreement with RMSDs ranging from 0.93 to 1.58 å. Electron microscopic images of extended tetrameric structures and ~10nm diameter L and V shapes generated using the designed junctions were close to the design models, demonstrating the control the rigid junctions provide for protein shape sculpting over multiple nanometer length scales.

16. Protein interactions and assemblies: 21. Structure (x‐ray/NMR/EM)

ABS190

ANALYSIS OF COMPUTATIONALLY DESIGNED COOPERATIVE PROTEIN‐PROTEIN INTERACTIONS BY NATIVE MASS SPECTROMETRY

Florian Busch ¹, Zachary VanAerum¹, Mengxuan Jia², Zibo Chen³, David Baker³, Vicki Wysocki²

¹The Ohio State University (Columbus, United States); ²The Ohio State University (Columbus, United States); ³University of Washington (Seattle, United States)

Native mass spectrometry (nMS) has become an important tool for the analysis of biomolecular interactions due to its high sensitivity and its ability to simultaneously differentiate and detect multiple species within a mixture. Here, we used nMS to investigate proteins computationally designed to display binding cooperativity. Specifically, proteins were intended to function as logic AND gates, where the colocalization of two proteins of interest requires the presence of one or several additional proteins (= linkers). The designed proteins were mixed, subjected to a denaturation/refolding procedure, and measured by nMS. Detected proteins and protein complexes were quantified using Intact Mass (Protein Metrics). The dependency of complex formation on a particular linker was determined by analyzing mixtures with varying linker concentrations. Complex formation was studied as a function of linker concentration for a total of three designed logic AND‐gates with a common Alinker (monomer/heterodimer/heterotimer)B structure. Concentration dependent complex formation was observed with only negligible amounts of partial complexes being formed even in the presence of a six‐fold excess of linker over A and B subunits, demonstrating the successful design of cooperative protein‐protein interactions.

16. Protein interactions and assemblies: 7. Dynamics and allostery

ABS191

REAL‐TIME MONITORING OF CLOCK CONTROLLED SIGNAL TRANSDUCTION PATHWAY

Archana G. Chavan ¹, Andy LiWang¹, Joel Heisler¹, Yonggang Chang¹

¹University of California, Merced (Merced, United States)

STUDY OBJECTIVE: The goal is to figure out how a circadian clock system generates oscillating signals with a ~24‐hr period, and how those signals are transmitted to control gene expression.

METHODS: The circadian clock system of cyanobacteria was reconstituted in vitro. This system includes the core oscillator, signal transduction enzymes, a clock‐controlled transcription factor, promoter DNA, and ATP. Each component was labeled one at a time with a fluorophore so that circadian rhythms of protein‐protein and protein‐DNA interactions could be measured in real‐time over several days.

RESULTS: Real‐time in vitro fluorescence measurements revealed that the core oscillator opens two temporally distinct windows to transiently activate mutually antagonistic signal transduction pathways, one late in the day and the other at night, that result in rhythmic promoter DNA binding by the transcription factor. Also, one signal transduction enzyme lengthens clock period through interactions with the core oscillator, whereas the other enzyme shortens the period. Together they tune the clock system to match the day/night cycle of the earth.

CONCLUSIONS: The ability to reconstitute in vitro the cyanobacterial circadian clock system allows highly precise measurements of every clock component in real time, bringing to light the succession of transient interactions separated in time optimized to regulate gene expression according to time of day.

Now that the methodology is established, we will investigate the mechanism of temperature compensation: clock‐controlled gene expression rhythms are insensitive to changes in ambient temperature.

3. Chaperones: 16. Protein interactions and assemblies

ABS192

STRUCTURAL ANALYSIS OF DNAJ PROTEIN ERDJ6 AND NON‐NATIVE PROINSULIN

Lindsay Hammack ¹, Mary Clay¹, Charalampos Kalodimos¹

¹St. Jude Children's Research Hospital (Memphis, United States)

DnaJ molecular chaperones function as co‐chaperones with Hsp70 proteins to maintain cellular homeostasis. DnaJs prevent protein aggregation by binding non‐native, damaged, or misfolded proteins. DnaJ recognition and binding is mediated though exposed hydrophobic regions on non‐native proteins. Despite the central importance of chaperone binding, the structural basis of their substrate interaction remains poorly understood. To understand DnaJ anti‐aggregation mechanisms, we have elected to investigate ERdj6, a key DnaJ chaperone located in the ER. Using an integrative structural biology approach, spearheaded by nuclear magnetic resonance (NMR) spectroscopy, we aim to solve an atomic resolution structure of ERdj6 in complex with a non‐native substrate. Our results include a preliminary NMR structure of apo‐ERdj6. Additionally we have identified non‐native substrates, including endogenous substrate, non‐native proinsulin. Moreover, ERdj6 binding sites have been identified on non‐native proinsulin via chemical shift perturbations. We anticipate our ERdj6‐proinsulin structure will illuminate how DnaJs recognize non‐native proteins, promote folding, and prevent aggregation.

24. Therapeutics and antibodies: 6. Design/engineering

ABS194

EXPLORING COMPOSITION OF PEPTIDE LINKER TO ENHANCE STABILITY OF ANTIBODY FRAGMENTS FOR CANCER THERAPEUTICS

Jeong Min Han ¹, Thomas Magliery¹

¹Department of Chemistry and Biochemistry, The Ohio State University (Columbus, United States)

Single‐chain variable fragment (scFv) is the smallest functional unit of monoclonal antibody with a non‐native peptide linker holding variable domain together. Its small size allows bacterial expression making engineering easy, and further gains pharmacokinetic advantages such as better tumor penetration and reduced immunogenicity. However, despite the great therapeutic potential, scFvs have not made it into the clinic. Part of the problem is that nearly all scFvs have been engineered with flexible (Gly4Ser)n linkers, which are susceptible to proteolysis leading scFvs to poor stability. Yet there has been no comprehensive study on compositional effects of the linker.

Here, we explored the linker composition to enhance stability of 3E8, a Tumor‐Associated Glycoprotein (TAG)‐72 specific scFv; TAG‐72 is present in 80% of adenocarcinomas.

First, sets of linker libraries in varying compositions were constructed and characterized to determine favorable linker trends in scFvs. The results showed stabilizing effects in conformationally constrained linkers. Then, a much broader sequence space was explored via combinatorial linker library and phage display, a high‐throughput screening method. After the fifth round of enrichment, the library was deep sequenced, and the hits appearing in repeat were further characterized to confirm stability enhancement. The results showed significant increase in expression level and stability with high frequency of constrained amino acids. Moving forward, we are currently working on a high‐resolution structure to further rationalize at the molecular level.

Taken together, this study provides invaluable insight into scFv engineering that is currently lacking in order to stabilize scFvs and expand its utility as cancer therapeutics.

1. Amyloid and aggregation: 3. Chaperones

ABS196

THE BACTERIAL CURLI ACCESSORY PROTEIN CSGF INFLUENCES THE AGGREGATION OF HUMAN ISLET AMYLOID POLYPEPTIDE

Sajith Jayasinghe¹, Osmar Meza‐Barajas ¹, Allison Newel¹, Ashwag Binmahfooz¹

¹California State University San Marcos (San Marcos, United States)

Gram‐negative bacteria, such as E.coli and Salmonella, contain proteinaceous, hair‐like, cell surface filaments known as curli. Curli serve to facilitate cell‐cell interactions and are essential for host cell colonization. Curli assembly involves six proteins, CsgA, CsgB, CsgC, CsgE, CsgF, and CsgG. CsgE and CsgF are thought to act as chaperones to help prevent the premature aggregation of CsgA and/or CsgB, and to help transport these proteins, through the outer‐membrane protein CsgG, to the cell surface where they assemble to form curli. It has been observed that CsgF is able to inhibit the aggregation of CsgA, the major protein component of Curli. To investigate the possible interaction of CsgF with the human islet amyloid polypeptide (hIAPP). In the presence of CsgF no increase in Thioflavin T fluorescence was observed for freshly solubilized hIAPP monitored as a function of time, suggesting that CsgF prevents the aggregation of hIAPP during the time period of observation. The nature of the CsgF‐hIAPP interaction was probed using a series of single cysteine mutants of CsgF labeled via the individual cysteine side chains with the fluorophore IAEDANS. In the presence of hIAPP the fluorophore at position of 23 of CsgF was found to be less exposed the quencher KI, suggesting that the N‐terminal region of CsgF may be involved in mediating the interaction of CsgF with hIAPP.

16. Protein interactions and assemblies: 21. Structure (x‐ray/NMR/EM)

ABS197

COMPARISON OF STRATEGIES FOR THE ENRICHMENT OF CROSS‐LINKED PEPTIDES

Andrew Norris ¹, Florian Busch¹, Vicki Wysocki¹

¹The Ohio State University (Columbus, United States)

The mass spectrometric analysis of cross‐linked peptides from proteins and protein complexes can provide information on intra‐ and inter‐molecular residue‐residue distances, which can then be used as constraints to model protein tertiary and quaternary structures. A fundamental and challenging aspect of this technique is to detect and identify the cross‐linked peptides in the presence of an excess of non‐cross‐linked peptides. In this work, we examined strategies used to enrich for cross‐linked peptides prior to LC‐MS and found that employing multiple strategies can help to retain more information on cross‐links present within a protein/protein complex. We cross‐linked homo‐dimeric enolase from Saccharomyces cerevisiae and proteins from the tryptophan synthesis pathway with disuccinimidyl sulfoxide (DSSO) and either enriched for inter‐molecular cross‐links by separating oligomers via SDS‐PAGE prior to protein digestion, or enriched for inter‐ and intra‐molecular cross‐links after in‐solution digestion. For the latter, we used size‐exclusion chromatography (SEC), mixed‐mode strong cation‐exchange (MCX), and capillary electrophoresis (CE) prior to LC‐MS as well as field asymmetric ion mobility spectrometry (FAIMS) as in‐line gas phase separation technique. All data were collected on an Orbitrap Fusion mass spectrometer and analyzed with XlinkX. From the five enrichment strategies preformed on enolase, over two hundred cross‐links were identified containing over one hundred unique cross‐links with less than twenty cross‐links common across all strategies. The results indicate that there is a variability in cross‐links identified depending on the enrichment strategy and demonstrate the utility of combining enrichment strategies to allow for a high coverage of cross‐links across the protein complex.

1. Amyloid and aggregation: 5. Computational modeling/simulation

ABS198

COMPUTATIONAL STUDY ON AGGREGATION AND DISRUPTION OF AMYLOID FIBRILS

Seokmin Shin ¹, Kyunghee Lee², MinJun Lee¹, Jeseong Yoon¹

¹Seoul National University (Seoul, South Korea); ²Sejong University (Seoul, South Korea)

Computational studies can provide basic understanding concerning the detailed mechanism of amyloid formation. The formation of amyloid fibrils and prefibrilar aggregates of misfolded proteins and peptides is very important to elucidate possible causes for various neurodegenerative disorders. Systematic molecular dynamics simulations have been performed on the formation of the oligomers and protofibrils of A peptide and peptides from ‐synuclein. Amyloid deposits of A protein in neuronal cells are known to be a major symptom of Alzheimers disease. Recent solid state nuclear magnetic resonance and cryogenic electron microscopy experiments identified that A42 shows an S‐shaped triple‐ structure, in contrast to the previously suggested U‐shaped ‐arch structure. We performed several different simulations in order to explain the conformational stability and aggregation mechanisms of this triple‐ motif. Our results provide an explanation of the formation and stability of the triple‐ structure‐containing A42 and its consequent high toxicity. Amyloid formation may be disrupted in several different ways. We investigated the effect of urea on the dis‐aggregation of protofibrils of A42 by extensive simulations. Recent experiments suggested that graphene quantum dots (GQDs) inhibit fibrillization of ‐synuclein and interact directly with mature fibrils, triggering their dis‐aggregation. We performed simulations in order to elucidate the mechanism of disruption of fibrils from ‐synuclein and A peptide by graphene quantum dots (GQDs).

12. Membrane proteins: 10. Folding

ABS199

MAPPING ACCESSIBILITY IN THE BACTERIAL MECHANOSENSITIVE CHANNEL MSCS TO A SMALL PHOTOREACTIVE PROBE

Gabriela Elena Gómez¹, José María Delfino ¹, Yan Wang², Andriy Anishkin³, Sergei Sukharev⁴

¹University of Buenos Aires (Buenos Aires, Argentina); ²Proteomics Core Facility. University of Maryland (College Park, MD 20742, United States); ³Biology Department. University of Maryland (College Park, MD 20742, United States); ⁴Biology Department and IPST. University of Maryland (College Park, MD 20742, United States)

Diazirine (DZN), a water‐sized photo reagent yielding highly reactive methylene carbene, probes protein regions accessibility and packing. Here we target the homo heptameric bacterial mechanosensitive channel MscS playing a critical role in osmotic shock‐driven osmolyte release. Three transmembrane helices (TM1‐TM3) from all of the subunits form the barrel with a TM3‐lined pore. Each subunit contributes to a large hollow cytoplasmic vestibule (cage) with seven side portals permeable to ions and small osmolytes. The available structural data raise questions about perturbations of the protein‐lipid boundary and the solvent occupancy of the conductive pore forming a hydrophobic constriction (gate). MscS embedded in the inner membrane of Escherichia coli was labeled with DZN. After purification of the tagged protein and tryptic digestion, fragments were subjected to Orbitrap MS. Thus, methylation probabilities at single‐residue resolution were mapped onto two crystal structures and a model of the resting state of MscS. Patterns reveal clear hot spots, likely dictated by the specific chemical environment around DZN. Protein surfaces surrounded by lipids (TM1) are entirely protected from labeling, a likely consequence of the strong carbene‐scavenging capacity of aliphatic chains. By contrast, interhelical crevices and many internal small voids appear labeled. Interestingly, some hydrophilic segments on the outer side of the cytoplasmic cage are reproducibly protected from labeling, suggesting their involvement in tight interactions with some cytoplasmic components. DZN labels side chains in the hydrophobic gate constriction. The rest of the pore interior is spared from labeling, supporting the notion that the pore is largely desolvated (vapor plugged).

8. Enzymology: 16. Protein interactions and assemblies

ABS200

IONIC STRENGTH MODULATES DIMERIZATION AND ENZYME ACTIVITY OF A ‐GLYCOSIDASE

Felipe Akihiro Otsuka ¹, Rafael Chagas¹, Vitor Almeida¹, Maiara Frutoso¹, Sandro Roberto Marana¹

¹University of São Paulo (São Paulo, Brazil)

Enzymes that catalyze the hydrolysis of O‐glycoside bond in beta conformation releasing a monosaccharide from the non‐reducing end from glycoconjugates are called ‐glycosidases. These enzymes, part of the Glycoside Hydrolase (GH) families, possess a characteristic tertiary structure ((/)8 barrel) and a conserved catalytic mechanism. Here we are studying a ‐glycosidase, classified in the GH1 family, from the organism Spodoptera frugiperda (Sfgly) aiming to understand the implications of its quaternary structure on the catalytic activity. Using SEC‐MALS (size exclusion chromatography coupled to multiangle light scattering), we showed that the Sfgly coexists as dimer and monomer in solution. Moreover, the dimer/monomer (D/M) population ratio is highly influenced by the ionic strength of the buffer, when in low ionic strength (10 mM potassium phosphate buffer pH 6) the D/M is lower than in high ionic strength (100 mM potassium phosphate buffer pH 6). A proper separation of the dimer and monomer led us to study their enzyme kinetic separately. The dimer state was found to be 5 times more active than the monomer upon two different synthetic substrates, p‐nitrophenyl‐‐glucopyranoside and p‐nitrophenyl‐‐fucopyranoside. Besides that, Sfgly mutants have been designed to stabilize and to impair the dimeric interface to fully understand the dimerization process and its role in the catalytic activity. For now, it is presumed that drastic conformational changes may be involved in the dimerization in a way that such changes affect the active site.

6. Design/engineering: 13. Metabolic engineering/energy applications

ABS201

DESIGN OF LIGHT HARVESTING PROTEINS FOR PHOTOSYNTHESIS

Nathan Ennist ¹, Adam Moyer¹, Derrick Hicks¹, Chunfu Xu¹, TJ Brunette¹, David Baker¹

¹University of Washington (Seattle, United States)

Photosynthetic antenna proteins bind networks of chromophores that efficiently absorb and transfer solar energy to reaction centers where electron transfer reactions take place. De novo design of light harvesting complexes may lead to new proteins for energy conversion or other light‐driven chemistry. We have designed a set of chlorophyll binding proteins with the goal of producing delocalized excited states similar to those observed in light harvesting LH1 and LH2 proteins from purple bacteria. Chlorophyll binding is verified by UV/vis spectroscopy, fluorescence spectroscopy, native mass spectrometry, and circular dichroism. Binding of chlorophyll oligomers alters excited state dynamics, as shown by fluorescence lifetime and fluorescence quantum yield measurements. Future work will aim to efficiently harvest light at wavelengths complementary to natural antenna complexes and develop reaction centers that can accept energy from designed light harvesting complexes, separate charge, and produce fuel.

15. Peptides: 20. Single molecule studies

ABS202

THROMBIN CLEAVES PROLACTIN INTO NOVEL VASOINHIBIN ISOFORMS

Magdalena Zamora Corona ¹, Juan Pablo Robles¹, Manuel Benigno Aguilar¹, Livia Lenke², Thomas Bertsch², Gonzalo Martínez de la Escalera¹, Jakob Triebel², Carmen Clapp¹

¹Instituto de Neurobiología, Universidad Nacional Autónoma de México (UNAM) (Queretaro, Mexico); ²Institute for Clinical Chemistry, Laboratory Medicine and Transfusion Medicine, Nuremberg General Hospital & Paracelsus Medical University (Nuremberg, Germany)

Vasoinhibin belongs to a family of antiangiogenic proteins with molecular masses that range from 9 to 18 kDa generated by the proteolytic cleavage of the hormone prolactin (PRL). Vasoinhibin has therapeutic implications in diabetic retinopathy, peripartum cardiomyopathy, and cancer. Furthermore, vasoinhibin has profibrinolytic effects which may impact thrombotic disorders. Thrombin is a protease that cleaves fibrinogen to the essential clot component, fibrin. Here, we studied whether thrombin cleaves PRL to vasoinhibin as a putative mechanism regulating hemostasis and angiogenesis in response to tissue injury. PRL was incubated with thrombin and the resulting proteolytic products were evaluated by Western blot, mass spectrometry, HPLC purification, and bioactivity. The consensus cleavage site for thrombin is Arg‐Gly and thrombin cleaved PRL at Arg48‐Gly49 and Arg103‐Gly104 to generate N‐terminal PRL fragments of 6 and 11 kDa, respectively, that inhibited the proliferation and migration of endothelial cells in culture. These findings extend the vasoinhibin family to proteins smaller than 9 kDa and add thrombin to the list of PRL cleaving enzymes that generate vasoinhibin. Addressing the physiological relevance of these novel proteins warrants further research.

The study was supported by the National Council of Science and Technology of Mexico (CONACYT) grants 289568 and A1‐S‐9620.

8. Enzymology: 10. Folding

ABS203

STRUCTURAL AND BIOCHEMICAL STUDIES OF MTB L‐ASPARAGINASE REVEAL SURVIVAL MECHANISM OF MYCOBACTERIUM TUBERCULOSIS INSIDE MACROPHAGES

Arti Kataria ¹, Bishwajit Kundu²

¹Kusuma School of Biological Sciences, IIT D, India (NEW DELHI, India); ²Kusuma School of Biological Sciences, IIT Delhi, India (Delhi, India)

Resistance to anti‐TB drugs has made Mycobacterium tuberculosis (Mtb) an escalating global health crisis enforcing the search for newer drug targets and treatment methods. Herein, we report Mtb L‐asparaginase (MtA) implicated in acidification arrest and survival of Mtb inside the phagosome of human macrophages as a new druggable target. Structural details of recombinant MtA through small angle x‐ray scattering (SAXS), size‐exclusion and mass‐spectrometry identified its functional unit as a dimer. Further biophysical analysis showed striking differences in optimum pH (8), temperature (50 oC) and enzyme kinetics (Km and Vmax of 4.35 mM and 2.2 U/ml respectively) of MtA with other mesophilic asparaginases. High‐resolution crystallographic data affirmed a critical Tyr to Val replacement in catalytic site I and suggested active site ‐hairpin remodeling upon substrate binding, indicating an alternate catalysis mechanism. Further, a targeted docking‐based screening of a combinatorial library of 15 million compounds against MtA identified M3, M26, M29 and M31 as novel, candidate molecules. Surface plasmon resonance (SPR) yielded strong micromolar affinities of these molecules that was corroborated by five‐fold reduction in MtA enzymatic activity. Finally, testing of these compounds on mycobacterial cultures showed promising inhibitory concentrations (IC50) of less than 50 μM. Overall, in lieu of its major structural differences with human asparaginase, MtA is presented as a new, druggable metabolic target against drug resistant TB.

3. Chaperones: 19. Proteostasis and quality control

ABS204

REMOVAL OF ZINC GIVES INSIGHTS INTO THE EFFECT OF THIS METAL ON THE STABILITY AND FUNCTION OF THE ZINC‐BINDING CO‐CHAPERONE YDJ1

Jemmyson Jesus ¹, Annelize Aragão¹, Marco Arruda¹, Carlos Ramos¹

¹University of Campinas (Campinas, Brazil)

Ydj1 (Hsp40) in yeast, has two zinc finger domains in each monomer and plays crucial roles in cells, such binding partially folded proteins and assisting the Hsp70 chaperone family in protein folding. So far, little is known about how Zn binding affects the structure and function of this chaperone. This work investigates the influence of Zn‐binding on both the structure and functional activity of Hsp40. Insights into the effect of zinc on the structure and function of Ydj1 without having to modify its primary structure, a method was developed to quantify and remove the zinc from the protein. Ydj1 was produced and purified, and its zinc content was determined by ICP‐MS. The result showed that two zinc atoms were bound per monomer of protein. To Zn removal, variations on chelating agent (EDTA, EGTA, 1,10‐phenanthroline), chelator concentration, reaction time, pH, and temperature were tested. These procedures had no effect on the overall secondary structure of the protein since no significant changes in the circular dichroism spectrum were observed. Significant Zn removal (91±2%, n=3) was achieved using 1,10‐phenanthroline (1x10‐3 mol L‐1) at 37°C with a pH 8.5 for 24 h. Zinc removal affected the thermal stability of the protein since the middle of the transition (Tm) decreased from 63±1°C to 60±1°C after Zn extraction. The effect on the ability of Ydj1 to protect a model protein (luciferase) against aggregation was completely abolished after the Zn removal procedure. As conclusion, zinc plays an important role in the stability and activity of Ydj1.

7. Dynamics and allostery: 21. Structure (x‐ray/NMR/EM)

ABS205

SELECTIVE INHIBITION OF CALCINEURIN ACTIVITY IN PATHOGENIC FUNGI

Ronald Venters ¹, Sophie Gobeil², Leonard Spicer³, Benjamin Bobay¹

¹Duke University NMR Center (Durham, United States); ²Duke University Department of Biochemistry (Durham, NC, United States); ³Duke University Department of Biochemistry and Radiology (Durham, NC, United States)

Invasive fungal infections are a leading cause of death in immunocompromised patients and remain difficult to treat since fungal pathogens, like mammals, are eukaryotes and share many orthologous proteins. As a result, current antifungal drugs have limited clinical value, are sometimes toxic, can adversely affect human reaction pathways and are increasingly ineffective due to emerging resistance. Current research in our laboratory is focused on targeting the calcineurin signaling pathway that has been shown to be required for fungal pathogenesis. One potential antifungal drug, FK506, establishes a ternary complex between the phosphatase, calcineurin, and the 12‐kDa peptidyl‐prolyl isomerase FK506‐binding protein, FKBP12. It has been well established that calcineurin is inhibited when in complex with FK506/FKBP12. Unfortunately, FK506 is also immunosuppressive in humans, precluding its usage as an antifungal drug, especially in immunocompromised patients. In order to improve therapeutic efficacy, we have undertaken a unique effort that utilizes structural biology, molecular dynamics and molecular mycology aimed at identifying biophysical features of these complexes that might confer fungal specificity for inhibiting calcineurin activity with the objective of designing novel drugs to treat invasive fungal diseases with reduced human immunosuppression. Here we describe the NMR assignment, inhibitor binding and dynamics data that are the first steps towards defining, at a residue specific level, the impacts of FK506 binding to fungal and mammalian FKBP12 proteins. Our data highlight differences between the human and fungal FKBP12s that could lead to the design of more selective anti‐fungal drugs that are not immunosuppressive to the human host.

6. Design/engineering: 4. Chemical biology

ABS206

TUNING ZINC BINDING ABILITY OF CALPROTECTIN

Aslin Rodriguez Nassif ¹, Walter Chazin¹

¹Vanderbilt University (Nashville, United States)

Pathogens need transition metals to satisfy their physiological needs. In infections, the pathogen must obtain transition metals from the host in order to grow and reproduce. S100 proteins, specifically the S100A8/S100A9 heterodimer commonly known as calprotectin (CP) acts as a defender against several pathogens via a mechanism termed nutritional immunity. CP is released in large quantities at the sites of infection and due to its high affinity Zn (II) and Mn (II), is able to sequester these essential transition metals away from the pathogen. The asymmetric nature of the CP heterodimer leads to the formation of two transition metal binding sites. Site one (S1) coordinates ions using six histidine residues and has a high affinity for Zn (II) and Mn (II). Site two (S2) coordinates ions with three histidine and one aspartate residue and binds Zn (II) with high affinity but not Mn (II). Remarkably, certain pathogens have evolved to bypass the nutritional immunity defense mechanism, stealing Zn (II) back from CP by secreting small siderophores or binding to transporters. To counter zinc piracy and understand more about the host‐pathogen struggle, we are working to develop a 'super Zn (II) binding' CP variant, focusing initially on S2. Initial substitution of the histidine residues for cysteine creates CP variants with higher affinity for Zn (II). Here, we will present our latest results to re‐engineer CP to achieve this goal, as well as plans to create a CP variant that binds Mn (II) but not Zn (II) with high affinity.

3. Chaperones: 19. Proteostasis and quality control

ABS207

ALTERNATIVE FORMS OF ENERGY MODULATE GROUP II CHAPERONIN ACTIVITY

Kevin Goncalves ¹, Tom Lopez¹, Judith Frydman¹

¹Stanford University (Stanford, United States)

The protein folding machinery of cells dictates protein homeostasis in response to stressors such as heat shock, protein misfolding, and protein aggregation. Chaperonins are 1‐MDa, oligomeric rings that have been shown to mediate protein aggregation suppression and protein folding in an adenosine triphosphate (ATP) ‐dependent manner. In addition, it has been shown that key regulators of protein translation, cell cycle, and cell signaling are regulated by guanosine triphosphate (GTP) and GTPase‐exchange factors. It had not yet been determined whether chaperonins are tightly regulated in the same fashion. We have demonstrated that a model group II chaperonin, archaeal M. Maripaludis (mmCpn), experiences dramatic conformational changes linked specifically to the utilization of GTP. Furthermore, these differences are influenced by structural modalities of the ATP‐binding pocket that were previously described as the phosphate and nucleotide sensing loops. Nucleotide hydrolysis assays reveal that GTP hydrolysis drops significantly relative to ATP hydrolysis. The decrease in GTP hydrolysis is coupled with a substantial reduction in non‐native protein folding capability of the chaperonin. Lastly, the conformational change induced by GTP hydrolysis is extensive enough to drive the disassembly of the chaperonin. In conclusion, we show that group II chaperonins utilize energy sources differently and is inherently linked to their protein quality control activity.

8. Enzymology: 25. Transcription/translation/post‐translational modifications

ABS208

MOLECULAR MECHANISM OF UBIQUITIN E2 ENZYME ACTIVATION IN ERAD

Tobias Ritterhoff ¹, Christian Lips², Thomas Sommer², Rachel Klevit¹

¹University of Washington (Seattle, United States); ²Max‐Delbrück‐Center for Molecular Medicine (Berlin, Germany)

Endoplasmic reticulum‐associated degradation (ERAD) is an essential eukaryotic protein quality control process; it recognizes misfolded ER proteins and degrades them via the Ubiquitin/proteasome system. In yeast, ERAD ubiquitylation is carried out by two E2 enzymes, Ubc6 and Ubc7, that display different biochemichal profiles. Like most E2s, Ubc7 attaches Ub to the amino group on a lysine sidechain. Ubc6 has an unusual activity in which it can ubiquitylate target proteins on hydroxy residues; it is a mono‐ubiquitylating E2 that is thought to put the first Ubiquitin onto a target protein (priming E2). Ubc7 efficiently builds Ubiquitin chains linked via the Lys48 residue, thereby marking targets for degradation. Target protein specificity is provided by the ERAD E3 ligases, Hrd1 and Doa10, which are able to activate either of the two E2 enzymes. We used in vitro activity assays, NMR, and yeast cell culture to compare these two functionally distinct, yet collaborating E2 enzymes and the two ERAD E3 ligases. We find that, compared to the canonical Ubc7, Ubc6 is stimulated by the two E3s through a different and so far novel molecular mechanism. We also find that Ubc7 can act as a priming E2. In contrast to its priming activity, we find that chain elongation by Ubc7 shows little dependency on activation by an E3 ligase. In addition to shedding light on the mechanism of ERAD ubiquitylation, our study puts into sharp focus the view that E2 enzymes are at the center of the ubiquitylation reaction by determining product formation.

7. Dynamics and allostery: 6. Design/engineering

ABS209

PROFILING LATENT AND ENGINEERABLE ALLOSTERY IN ION CHANNELS THROUGH SYSTEMATIC DOMAIN INSERTION

Daniel Schmidt ¹, Willow Coyote‐Maestas¹, David Nedrud¹, Yungui He¹, Chad Myers¹

¹University of Minnesota (Minneapolis, United States)

Ion channels evolved prior to emergence of multicellular life. They organize into families with distinct structural features and domain topologies. We are focused on understanding how these features and topologies enabled, and directed, the evolution of allosteric regulation that affects channel gating.

Recently, we developed an experimental method differential domain insertion profiling that measures the degree by which in‐frame inserted domains with different biophysical properties disrupt ion channel structure and function. A theoretical framework based on the Ensemble Allostery Model allows us to relate this experimental measure to regional protein dynamics and, ultimately, site‐specific allosteric capacity.

We found that in the inward rectifier K+ channel 2.1 (Kir2.1) differential permissibility reports not only the allosteric sites harnessed in Kir2.1 itself, but also sites that are equivalent to those in close homologs (e.g., Kir3.x or Kir6.x). Inserting light‐switchable domains into these putative allosteric sites allowed us to control Kir2.1 function with light.

In aggregate, these data suggest that allosteric regulation might have evolved by leveraging a pool of allosteric capacity that is intrinsic to the Kir family. Such exaptation of latent allostery solves a central problem in Darwinian evolution: how allosteric modulators binding can become selected unless a compatible pathway to transduce this signal is already present in the modulated protein.

I will present our work on Kir2.1 and discuss our findings in the context of protein dynamics, protein evolution and protein engineering. I will also discuss ongoing work on greatly improved enabling methods, such as generating unbiased domain insertion libraries.

ABS210

APPLICATIONS OF ULTRACENTRIFUGATION IN PURIFICATION AND CHARACTERIZATION OF BIOMOLECULES

Akash Bhattacharya¹, Nicholas Harrington ¹, Ross Verheul¹, Eric Von Seggern¹, Stephen Otts¹

¹Beckman Coulter Life Sciences (Loveland, United States)

Ultracentrifuges spin samples with centrifugal forces typically spanning 100,000 1,000,000 x g. At these high forces, the constituent molecules in the sample separate based on their physical properties (e.g., size, mass, density, anisotropy). Accordingly, ultracentrifugation is commonly used to purify, as well as characterize, low‐molecular weight polymers up to multi‐megaDalton protein complexes and organelles.

12. Membrane proteins: 17. Proteins in cells

ABS211

CELLULAR ABUNDANCE MEASUREMENTS OF THOUSANDS OF VARIANTS OF VITAMIN K EPOXIDE REDUCTASE (VKOR) RESOLVES TOPOLOGY AND MECHANISMS OF DRUG RESISTANCE

Melissa Chiasson ¹, Katherine Sitko¹, Jason Stephany¹, Allan Rettie², Douglas Fowler¹

¹Department of Genome Sciences, University of Washington (Seattle, United States); ²Department of Medicinal Chemistry, University of Washington (Seattle, United States)

Proteins must be expressed at sufficient levels to function properly. We developed a method, variant abundance by massively parallel sequencing (VAMP‐seq), which measures the abundance of thousands of variants of a protein simultaneously using fluorescent reporters, flow cytometry, and high‐throughput sequencing. We applied VAMP‐seq to VKOR, an ER‐resident transmembrane protein and the target of the anticoagulant warfarin. VKORs structure, topology, and the mechanism by which some variants lead to warfarin resistance are still poorly understood. Using VAMP‐seq, we measured abundance scores for 2,695 of the 3,078 possible single amino acid variants of VKOR. Windowed averaging of abundance scores revealed four distinct regions where charged and polar residues profoundly reduced abundance, supporting a four transmembrane domain topology. Furthermore, we observed a unique trimodal distribution of abundance scores, which segregate by domain, suggesting that membrane proteins have a distinct mutational profile compared to cytoplasmic proteins. Here, residues comprising the internal core of VKOR are mutationally intolerant, residues that interact with lipid bilayer are more mutationally tolerant and ER/cytoplasmic loops are highly mutationally tolerant. Finally, only one of 22 warfarin resistance mutations, A26T, has increased abundance, illustrating that this is not a prevalent mechanism of warfarin resistance. In sum, we show that abundance data can resolve both structural and mechanistic questions and anticipate that with more data we will be able to build a comprehensive atlas of membrane protein‐specific mutational patterns.

6. Design/engineering: 21. Structure (x‐ray/NMR/EM)

ABS212

MODULAR AND EXPANDABLE PROTEIN‐DNA CO‐CRYSTAL SCAFFOLDS TO ASSIST IN X‐RAY DIFFRACTION OF DNA‐BINDING MACROMOLECULES

Abigail Ward ¹, Christopher Snow¹

¹Colorado State University (Fort Collins, United States)

Isoreticular Co‐crystals (ICC) are a novel class of designed protein‐DNA co‐crystals. In each ICC, the DNA stacks end‐to‐end. Furthermore, the crystal symmetry allows expansion of the DNA‐DNA interaction without breaking protein‐protein contacts, hence providing larger solvent channels for guest diffusion. Due to canonical base‐pairing, the DNA inserted for the expansion is modular, providing an interchangeable DNA sequence for scaffold assisted x‐ray diffraction studies. The ICC scaffold starts from existing co‐crystals of protein and DNA, each selected with custom Python scripts. Candidate structures were narrowed down using PyMOL visualization and experimental convenience to a working list of 20 candidate structures. To experimentally validate ICC scaffolds, molecular biology techniques were utilized: cloning, expression, purification and crystallization. To date, we have grown expanded ICC crystals using sitting and hanging drop vapor diffusion. The crystals grown have a specific binding sequence exposed to the pores for a guest protein of interest. The research presented offers a new approach to scaffold assisted x‐ray crystallography. In principle, any macromolecule that binds tightly to a specific DNA sequence may be revealed in the x‐ray diffraction pattern of the co‐crystal scaffold.

1. Amyloid and aggregation: 10. Folding

ABS213

AMYLOID FORMATION BY THE RNA RECOGNITION MOTIFS OF DISEASE‐LINKED RNA‐BINDING PROTEINS

Sashank Agrawal ¹, Woei‐Chyn Chu², Hanna S. Yuan³

¹1Institute of Molecular Biology, Academia Sinica, Taipei, Taiwan 11529, ROC. 2Molecular and Cell Biology Program, Taiwan International Graduate Program, Academia Sinica, Taipei, Taiwan 11529, ROC. 3Graduate Institute of Life Sciences, National Defense Medical Center, Taipei, Taiwan 11490, ROC. (Taipei, Taiwan); ²Department of Biomedical Engineering, National Yang‐Ming University, Taipei, Taiwan, ROC (Taipei, Taiwan); ³Institute of Molecular Biology, Academia Sinica, Taipei, Taiwan 11529, ROC. (Taipei, Taiwan)

Aberrant aggregation of a group of RNA‐binding proteins, including TDP‐43, FUS and RBM45, are associated with the neurodegenerative diseases, including amyotrophic lateral sclerosis (ALS) and fronto‐temporal lobar dementia (FTLD). These three disease‐linked RNA‐binding proteins all contain at least one RNA recognition motif (RRM), however, it is not clear if these RRMs contribute to their aggregation‐prone properties. Here, we compared the biophysical and fibril formation properties of five RRMs from the three disease‐linked RNA‐binding proteins and five RRMs from non‐diseased proteins to determine if disease‐linked RRMs share common features making them prone to self‐assembly. We found that the disease‐linked RRMs have reversible thermal unfolding‐refolding properties and have a slightly lower average thermal melting point compared to non‐diseased RRMs. TDP‐43 RRM1, FUS RRM, as well as three beta2 peptides from these two RRMs, are prone to fibril formation in vitro, and all these fibrils of RRMs and their peptides are amyloids as confirmed by FT‐IR analysis and X‐ray diffraction. Our results suggest that some disease‐linked RRMs indeed play important roles in amyloid formation and could possibly drive protein self‐assembly. Our results thus provide an important basis for further studies of RRMs and RRM‐containing proteins in protein aggregation and neurodegeneration.

14. Motors & machines: 5. Computational modeling/simulation

ABS214

MOLECULAR MECHANISMS OF THE INTERHEAD COORDINATION BY INTERHEAD TENSION IN CYTOPLASMIC DYNEINS

Qian Wang ¹, Biman Jana², Michael Diehl³, Margaret Cheung⁴, Anatoly Kolomeisky¹, José Onuchic¹

¹Rice University, Center for Theoretical Biological Physics (Houston, United States); ²Indian Association for the Cultivation of Science (Kolkata, India); ³Rice University, department of chemistry (Houston, United States); ⁴University of Houston, department of physics (Houston, United States)

Cytoplasmic dyneins play a major role in retrograde cellular transport by moving vesicles and organelles along microtubule filaments. Dyneins are multidomain motor proteins with two heads that coordinate their motion via their interhead tension. Compared with the leading head, the trailing head has a higher detachment rate from microtubules, facilitating the movement. However, the molecular mechanism of such coordination is unknown. To elucidate this mechanism, we performed molecular dynamics simulations

on a cytoplasmic dynein with a structure‐based coarse‐grained model that probes the effect of the interhead tension on the structure. The tension creates a torque that influences the head rotating about its stalk. The conformation of the stalk switches from the registry to the registry during the rotation, weakening the binding affinity to microtubules. The directions of the tension and the torque of the leading head are opposite to those of the trailing head, breaking the structural symmetry between the heads. The leading head transitions less often to the registry than the trailing head. The former thus has a greater binding affinity to the microtubule than the latter. We measured the moment arm of the torque from a dynein structure in the simulations to develop a phenomenological model that captures the influence of the head rotating about its stalk on the differential detachment rates of the two heads. Our study provides a consistent molecular picture for interhead coordination via interhead tension.

12. Membrane proteins: 21. Structure (x‐ray/NMR/EM)

ABS216

FUNCTION/STRUCTURE IN OPA1‐MEDIATED MITOCHONDRIAL INNER‐MEMBRANE FUSION

Luke Chao ¹, Yifan Ge¹, Sivakumar Boopathy¹, Julie McDonald¹

¹MGH (Boston, United States)

Mitochondrial membrane dynamics underlies the organelle's functional diversity. OPA1 (mutated in dominant atrophy) is the mitochondrial inner‐membrane fusogen. A member of the dynamin family GTPases, OPA1 catalyzes membrane fusion through a series of sequential nucleotide‐hydrolysis coupled conformational rearrangements. We present a mechanism of OPA1 regulation from an in vitro reconstituted membrane fusion assay that distinguishes lipid mixing and content release. We describe preliminary electron cryo‐microscopy (cryo‐EM) reconstructions of OPA1 and their implications for self‐assembly. Finally, we will discuss the relationship between OPA1 protein conformational state and fusogenic membrane transitions.

ABS217

STRUCTURAL MODELING AND IN SILICO SCREENING OF POTENTIAL SMALL MOLECULE ALLOSTERIC AGONISTS OF GLP‐1 RECEPTOR

Zhijun Li ¹, Tejashree Redij¹, Rajan Chaudhari¹, Zhiyu Li¹, Xianxin Hua²

¹University of the Sciences in Philadelphia (Philadelphia, United States); ²University of Pennsylvania (Philadelphia, United States)

The Glucagon‐like peptide 1 receptor (GLP‐1R) belongs to the pharmaceutically important Class B family of G‐protein coupled receptors (GPCRs) and its incretin peptide ligand GLP‐1 analogs are adopted drugs for the treatment of type 2 diabetes. Despite remarkable anti‐diabetic effects, GLP‐1 peptide‐based drugs have several shortcomings. On the other hand, developing nonpeptidic small molecule drugs targeting GLP‐1R remains elusive. Here, we first constructed a three‐dimensional structure model of the transmembrane (TM) domain of human GLP‐1R using homology modeling and conformational sampling techniques. Next, a potential allosteric binding site on the TM domain was predicted computationally. In silico screening of drug‐like compounds against this predicted allosteric site has identified nine compounds as potential GLP‐1R agonists. The independent agonistic activity of two compounds was subsequently confirmed using a cAMP response element (CRE)‐based luciferase reporting system. One compound was also shown to stimulate insulin secretion through in vitro assay. In addition, this compound synergized with GLP‐1 to activate human GLP‐1R. These results demonstrated that allosteric regulation potentially exists in GLP‐1R and can be exploited for developing small molecule agonists. The success of this work will help pave the way for small molecule drug discovery targeting other Class B GPCRs through allosteric regulations.

21. Structure (x‐ray/NMR/EM): 4. Chemical biology

ABS218

PROTEIN, LIGAND AND WATER CHARACTERIZATION BY MULTIPLE SOLVENT CRYSTAL STRUCTURES

Sorabh Agarwal ¹, Mychal Smith², Miriam Segura‐Totten³, Carla Mattos⁴

¹Northeastern University (Boston, United States); ²Department of Chemistry, Virginia commonwealth University (Richmond, United States); ³Department of Biology, University of North Georgia (Dahlonega, United States); ⁴Department of Chemistry and Chemical Biology, Northeastern University (Boston, United States)

Drug development depends in part on screening libraries of compounds for targeted ligands with high affinity. We have further developed the multiple solvent crystal structures (MSCS) method to analyze ligand binding hot spots by soaking protein crystals in organic solvents. MSCS allows for the exploration of a larger chemical space than traditional screening methods while concurrently overcoming practical concentration limitations. While rich in information, MSCS structural data sets are laborious to obtain. We present an MSCS pipeline with an expanded solvent set that includes representative drug‐like functional groups. This pipeline includes tests for assessing feasibility of MSCS for a protein of interest and generation of a highly conserved water molecule map. The DRoP program is used to select conserved water molecules, such that a more advanced stage of refinement is reached, allowing placement of organics more accurately, converging on high quality maps. Structures allow for identification of protein‐ligand interactions, hot spots and conserved water binding sites. We focused on three proteins to develop our improved MSCS protocol. We tested our DRoP conserved water pipeline on the DNA‐binding protein barrier‐to‐autointegration factor (BAF) and found a highly conserved water network that mediates the interaction of BAF to DNA. Analysis of the sugar binding protein concanavalin A identified hot spots with multiple organic molecules bound. The membrane protein mPGES was used to optimize a variety of crosslinking approaches to promote MSCS characterization of membrane proteins. MSCS allows for more complete characterization of protein‐ligand binding and water analysis compared to standard crystallographic approaches.

21. Structure (x‐ray/NMR/EM): 8. Enzymology

ABS221

A STRUCTURAL AND FUNCTIONAL ANALYSIS OF BSHA: INSIGHTS INTO THE CATALYTIC MECHANISM AND FEEDBACK INHIBITION BY BACILLITHIOL

Paul Cook ¹, Christopher Royer¹, Kelsey Winchell¹

¹Grand Valley State University (Allendale, United States)

Bacillithiol is a low molecular weight thiol produced by many pathogenic gram‐positive bacteria such as Bacillus anthracis and Staphylococcus aureus. The compound is involved in the maintenance of redox homeostasis, detoxification of xenobiotic agents, and resistance to the FDA‐approved antibiotic fosfomycin. It is produced via a pathway utilizing the enzymes BshA (a glycosyltransferase), BshB (a zinc‐dependent deacetylase), and BshC (a cysteine ligase). Here we will discuss the ligand‐bound X‐ray crystallographic structures and steady‐state kinetics results of BshA. Our BshA structures complexed with UDP and N‐acetylglucosamine corroborate the SNi‐like mechanism that it and other GT‐B retaining glycosyltransferases are hypothesized to utilize. Additionally, we demonstrate that bacillithiol is a competitive inhibitor of the BshA substrate UDP‐N‐acetylglucosamine, suggesting a mechanism of feedback inhibition. The structures and bound ligands give insight into the function of BshA and may provide routes for potential therapies to circumvent fosfomycin resistance mechanisms.

24. Therapeutics and antibodies: 21. Structure (x‐ray/NMR/EM)

ABS222

A TWO‐PRONG APPROACH TO DEVELOPING AN INHIBITOR SCREENING METHOD FOR COMPOUNDS AGAINST CRYPTOSPORIDIUM PARVUM N‐MYRISTOYLTRANSFERASE

Alexandra Reers ¹, Yi Liu², Matt Hulverson², Alexis Kaushansky³, Peter Myler³, Wesley Van Voorhis², Erkang Fan², Bart Staker¹

¹Seattle Children's Research Institute (SEATTLE, United States); ²University of Washington (SEATTLE, United States); ³Seattle Children's Research Institute/University of Washington (SEATTLE, United States)

The parasite Cryptosporidium spp. is largely responsible for diarrheal infections in infants in many South East Asian countries‐ resulting in multi‐day moderate‐to severe diarrhea episodes. Waterborne transmission is the most predominant route of exposure: the parasite accounted for up to 60 % of waterborne outbreaks between 2004 and 2010. The parasite is resilient to treatment with various anti‐parasitic therapeutic agents. To this day there are no generally effective vaccine, prophylaxis or drugs against Cryptosporidium spp. The piggy‐back approach, which takes advantage of classes of molecules that have already been identified as a hit or lead and explores their efficacy on a different target species‐ have become an interesting alternative in the hopes of cutting cost and time in the drug discovery process.

We decided to evaluate the potency of three molecule scaffolds ‐ previously identified by GSK as potent inhibitors against Plasmodium falciparum N‐myristoyltransferase (NMT) in a compound library screen ‐ against Cryptosporidium parvum NMT. NMT catalyzes the transfer of 14‐carbon fatty acid, myristate, to the N‐terminus of substrate proteins after methionine removal. Generally, it is held that myristylation of substrate proteins are essential for protein‐protein and protein‐membrane interaction in the cell.

With this study we assessed the potential of the three‐compound series with a two‐prong approach: Firstly, we compared the functional activity of recombinant, purified CpNMT with and without the inhibitory compounds using a fluorescent based activity assay. Secondly, we identified the structure of (wildtype and mutant) CpNMT in ternary complex with MyristateCoA and compound by X‐ray crystallography.

6. Design/engineering: 8. Enzymology

ABS223

ENGINEERING SORTASE A; ACTIVITY AND SELECTIVITY OF NEW HYBRID AND ANCESTRAL VARIANTS OF SORTASE A

Sarah Struyvenberg ¹, Jordan Valgardson², Nick Horvath², John Antos², Jeanine Amacher²

¹Western Washington University (Bellingham, United States); ²Western Washington University (Bellingham, United States)

Bacterial sortase enzymes are a beneficial tool in innovative mechanisms of protein engineering. However, important limitations to utilization of sortases for engineered purposes exist; namely, that sortase A (SrtA) is a relatively poor enzyme and very specific for the substrate containing LPATXG motif. Exciting previous work from our collaborators reveals that sortases from different species recognize different sequences and that activity can vary. Therefore, we wanted to create and investigate hybrid sortase enzymes between SrtA from S. aureus and Streptococcus pneumoniae, wherein we swapped a substrate‐interacting loop between the E and F strands. Our hypothesis is that these residues are responsible for the differing specificities of these two enzymes and that our loop‐swapped S. aureus enzyme will show S. pneumoniae sequence selectivity, and vice versa. In addition to the creation of loop swapped complexes, we have also used ancestral sequence reconstruction (ASR) to investigate the specificity and activity of ancestral sortase sequences in gram negative bacteria. We have engineered two ancestral sortase complexes, ancestral SrtAstaph and SrtAstrep using ASR of sequences obtained from NCBI. Our hypothesis is that SrtAstaph and SrtAstrep will be more active, promiscuous enzymes than their extant relatives. Here, we present activity and selectivity data for our loop swapped variants, as well as our ancestral enzymes, in comparison to the two native S. aureus and S. pneumoniae SrtA enzymes. The discovery or development of a more promiscuous sortase enzyme could lead to more efficient mechanisms of protein engineering then are currently available to researchers.

12. Membrane proteins: 21. Structure (x‐ray/NMR/EM)

ABS224

USE OF SPIN‐LABELED NANODISCS TO IMPROVE STRUCTURAL DETERMINATION OF MEMBRANE PROTEINS BY ESR

Chieh‐Chin Li ¹, Yun‐Wei Chiang¹

¹Department of Chemistry, National Tsing Hua University (Hsinchu City, Taiwan)

Determining structures and conformational changes of membrane proteins (MPs) is important to understanding protein activity. Site‐directed spin labeling technique in combination with double electron electron resonance (DEER) is a powerful tool to explore conformational changes of MPs. However, the MP studies by DEER suer from relatively weak dipolar signals due to their complex nature in membrane environments. To resolve the challenge in the MP study, we report the use of lipid nanodiscs (NDs) to improve the distance resolution resolved by DEER. Two genetically engineered membrane scaffold protein (MSP) mutants are introduced, each of which is shown to form double‐labeled ND efficiently and with high homogeneity. Strikingly, the resultant interspin distance distribution is featured by a small distribution width, suggesting high resolution. Also, a small but distinct changes in ND geometry with MP incorporation can be detected, indicating the great sensitivity of DEER measurement using double‐labeled NDs. When DEER is performed on a binary mixture of the doubled‐labeled ND devoid of MPs and the unlabeled ND with incorporated double‐labeled MPs, the overall amplitude of dipolar signals is increased, leading to a critical enhancement of the distance resolution. With this approach, the determination of MP structures can be studied at high resolution in NDs.

7. Dynamics and allostery: 4. Chemical biology

ABS225

PROTEIN LOCAL DYNAMICS AND ITS COUPLING TO SOLVENT

Yun‐Hsuan Kuo ¹, Yun‐Wei Chiang¹

¹National Tsing Hua University (Hsinchu, Taiwan)

Solvent plays an important role in protein dynamics and function, but how it regulates the dynamics remains debated. Here, saturation transfer electron spin resonance (ST‐ESR) is employed to explore the issue and characterize the dynamics on a longer (from s to s) time scale than has been extensively studied. First, we utilize ST‐ESR with small spin molecule, TEMPOL, to monitor dynamics about solvents in glycerol/water system. We found that dynamics of solvent enters into a millisecond timescale at subfreezing temperatures (180240 K) and reveals dynamical changeovers agreeing to liquidliquid transition (LLT) in the state diagram of the glycerol/water system. We then investigate dynamics of T4 lysozyme (T4L) in the same temperature range. By combining site‐directed spin labeling with four different probes, we systematically map out the variation in local (site‐specific) dynamics around a protein surface in glycerol/water mixtures. We found that protein dynamics depends strongly on the local environment in a protein. At highly exposed sites, protein and solvent dynamics are coupled, whereas they deviate from each other when temperature is greater than LLT temperature (190 K) of the solvent. At less exposed sites, protein however retains a dynamical motion, which is distinctly different from the bulk solvent, throughout the temperature range studied. This study not only reveals the hierarchy of the protein dynamics associated with the protein side‐chain motions, but also provides quantitative descriptions for the dynamic components observed in the ST‐ESR results of the fully hydrated T4L.

10. Folding: 7. Dynamics and allostery

ABS226

UNFOLDING EVENTS OF BID PROTEIN DURING THERMAL DENATURATION BY ESR ABSORPTION‐MODE SPECTROSCOPY

Chien‐Lun Hung ¹, Yun‐Wei Chiang¹

¹Department of Chemistry, National Tsing Hua University (Hsinchu city, Taiwan)

Bid, a BH3‐only protein of Bcl‐2 protein family, plays a critical role in connecting death receptor pathway to mitochondria‐mediated apoptosis. Its structure has been solved by NMR more than a decade ago, but the structure‐function relation remains largely unexplored. Here we study the thermostability of Bid protein and explore how the thermally‐induced unfolding affects the death‐promoting function of Bid. We demonstrate by circular dichroism (CD) spectroscopy that Bid is a highly stable protein and remains partially folded even at temperatures greater than 350 K. We also employ pulsed ESR dipolar spectroscopy and show that Bid structure is basically unchanged even in a solution containing 3 M GdnHCl. More than 30 single‐labeled mutants covering all of eight helices were studied by cw‐ESR in the temperature range of 300‐345 K. The onset of local structural disruption in response to the thermal/chemical denaturation can be clearly determined using the analysis based on the ESR absorption‐mode spectra. As such, local thermostability over Bid structure is mapped out, and a new interface between helices 3, 6, and 8 is identified as the beginning of unfolding event. This study verifies that the death‐promoting function of Bid is not lost upon the thermally‐induced unfolding and aggregation. The newly developed ESR absorption peak‐height method, introduced here, is demonstrated as a powerful tool for studying temperature dependence of protein unfolding/stability.

21. Structure (x‐ray/NMR/EM): 8. Enzymology

ABS227

STRUCTURAL INSIGHTS INTO CHLORAMPHENICOL‐METABOLIZING ENZYME FROM METAGENOME

Sang‐Hoon Kim¹, Sangkee Rhee ¹

¹Seoul National University (SEOUL, South Korea)

Metagenomes have gained considerable attention for use in biotechnological applications, because novel biological activities are often identified in the genomes. Recently, metagenome‐derived EstDL136 was found to possess chloramphenicol (Cm)‐metabolizing features. Sequence analysis showed EstDL136 to be a member of the hormone‐sensitive lipase (HSL) family with an Asp‐His‐Ser catalytic triad and a notable substrate specificity. Recently, we determined the crystal structures of EstDL136 and in a complex with Cm. Consistent with the high sequence similarity, the structure of EstDL136 is homologous to that of the HSL family. The active site of EstDL136 is a relatively shallow pocket that could accommodate Cm as a substrate as opposed to the long acyl chain substrates typical of the HSL family. Mutational analyses further suggested that several residues in the vicinity of the active site play roles in the Cm‐binding of EstDL136. These results provide structural and functional insights into a metagenome‐derived EstDL136.

16. Protein interactions and assemblies: 17. Proteins in cells

ABS228

BIOCHEMICAL ANALYSIS OF PRIA HELICASES FROM GRAM‐POSITIVE BACTERIA REVEALS DISTINCT DNA UNWINDING ACTIVITY IN DNA REPLICATION RESTART

Chwan‐Deng Hsiao ¹, Min‐Guan Lin¹, Yi‐Ching Li¹

¹Institute of Molecular Biology, Academia Sinica (Taipei, Taiwan)

DNA replication forks often encounter template DNA lesions, which can stall progression of the replication forks. In gram‐positive bacteria, the PriA‐dependent pathway is the major replication restart mechanism and it requires several primosome proteins. Among them, the PriA proteina 3 to 5 superfamily‐2 DNA helicaseis the key factor recognizing DNA lesions and it further recruits other proteins. Here, we investigated the PriA ATPase and helicase activities of Streptococcus pneumoniae (SpPriA) and Geobacillus stearothermophilus (GstPriA) through biochemical and kinetic analyses. Using various DNA duplexes with unstructured single‐stranded regions, we observed distinct unwinding abilities for GstPriA and SpPriA according to increased GC content of duplex DNA substrate. In contrast to the unwinding activity of GstPriA helicase, we found that SpPriA is unable to unwind duplex DNA substrate with high GC content. However, the presence of DnaD loader protein allows SpPriA to unwind such duplex DNA. Our findings suggest that SpPriA possesses limited unwinding ability in high GC DNA and that DnaD acts as an accessory protein to destabilize stalled replication forks through its single‐stranded DNA binding ability, thereby enhancing PriA unwinding activity.

8. Enzymology: 6. Design/engineering

ABS229

EFFECTS OF REACTIVE GLUTAMINES AND A BINDING SITE REGION ON THE FACTOR XIII SUBSTRATE SPECIFICITY FOR FIBRINOGEN C (233‐425)

Muriel Maurer ¹, Mohammed Hindi², Francis D.O. Ablan², Chad Stephens², Kelly Mouapi²

¹University of Louisville (Louisville, United States); ²Chemistry Department, University of Louisville (Louisville, United States)

Factor XIII (FXIII) and fibrin are critical proteins in the final stages of blood coagulation. FXIII‐catalyzed crosslinks between side chains of specific fibrin glutamines and lysines help to generate a clot that is resistant to fibrinolysis. The C‐terminal fibrinogen C region (233‐425) contains three reactive glutamines (Q237, Q328, and Q366) and a FXIII binding site (C 389‐403) with E396 serving as a key anchor point. Glutamine reactivities within WT and mutant Fbg C (233‐425) were investigated using a MALDI‐TOF mass spectrometry assay employing the lysine mimic glycine ethyl ester. With WT C, the FXIIIA reactivity ranking is Q237 >> Q366 Q328. The C E396A mutant did not prevent the transglutaminase function of FXIII A2 or A2B2. However, Q237 reactivity was modestly hindered suggesting a role of the binding site in controlling Q237 access. Additionally, FXIIIA could maintain transglutaminase function with truncated C (233‐388), missing the 389‐403 anchoring site. With Fbg C (233‐425) Q237N, the Q328 and Q366 reactivities became hindered suggesting native Q237 crosslinking contributes to FXIIIA substrate access. By contrast with C (233‐425) Q366N, the crosslinking of Q237 was not affected while Q328 became more reactive. The genetic variant C Q328P (Seoul II) influences fibrin clot structure and has been reported to increase myocardial infarcts. Minimal effects on Q237 and Q366 reactivites suggest Q328P does not cause a major C structural change. In summary, FXIIIA utilizes different Q environments and a fibrinogen C anchoring site to control substrate specificity.

1. Amyloid and aggregation: 4. Chemical biology

ABS230

IDENTIFY THE AMYLOIDOGENIC PEPTIDES AND CREATE PHOTOCONTROLLABLE PROBES FOR NEURODEGENERATIVE DISEASE

Jen‐Tse Huang ¹

¹Institute of Chemistry, Academia Sinica (Taipei, Taiwan)

Despite hyper‐phosphorylated TDP‐43 has been confirmed as one of the major components in the inclusion bodies of patients with Amyotrophic Lateral Sclerosis (ALS) and Frontotemporal lobar degeneration (FTLD), the underlying disease mechanism has not yet been fully elucidated. Our group has synthesized numerous TDP‐43 peptides and illustrated these fragments could form fibrillar aggregates, sharing similar morphology with neuronal cytoplasmic inclusions in ALS patients from 2010. Later on, we further identified the amyloidogenic core sequence of TDP‐43 and characterized their prion‐like properties both in test tube and cellular model. Based on these findings, we establish a photocontrollable platform to trigger amyloidogenesis to recapitulate the pathogenesis of amyotrophic lateral sclerosis (ALS) by applying a chemically engineered probe as a switch in live cells. The photocontrollable probe can self‐assemble into a spherical vesicle but rapidly develops massive nanofibrils with amyloid properties upon photoactivation. In cellular experiments, this cell‐penetrable vesicle was retained in the cytoplasm, seeded the mislocalized endogenous TDP‐43 into aggregates upon irradiation, and consequently initiated apoptosis. In addition, this photocontrollable vesicle interfered with nucleocytoplasmic protein transport (importin and Ran) and triggered cortical neuron degeneration. Our developed strategy provides in vitro and in vivo spatiotemporal control of neurotoxic fibrillar aggregate formation, which can be applied in the studies of protein misfolding and amyloid‐induced pathogenesis.

6. Design/engineering: 16. Protein interactions and assemblies

ABS231

THERMAL RECONSTRUCTION OF PROTEIN NANO‐BUILDING BLOCK COMPLEXES USING AN ULTRA‐STABLE DE NOVO PROTEIN DOMAIN

Ryoichi Arai ¹, Naoya Kimura¹, Naoya Kobayashi¹

¹Shinshu University (Ueda, Japan)

Research on molecular design and assembly of artificial protein complexes is an important step in protein design and engineering. We designed and created a protein nanobuilding block (PN‐Block), WA20‐foldon, by fusing an intermolecularly folded dimeric de novo protein WA20 and a trimeric foldon domain from bacteriophage T4 fibritin (Kobayashi et al., JACS 2015). The WA20‐foldon, as a simple and versatile nanobuilding block, self‐assembled into several oligomeric nano‐architectures in multiples of 6‐mer. We also designed and created de novo extender protein nanobuilding blocks (ePN‐Blocks), by fusing tandemly two WA20 with various linkers, to construct self‐assembling cyclized and extended chain‐like nanostructure complexes (Kobayashi et al., ACS Synth. Biol. 2018). In this study, to stabilize the de novo protein WA20 for application of PN‐Blocks in nanotechnology, we designed and developed a WA20 variant, called SUWA (Super WA20), by several mutations to stabilize helices and hydrophobic cores. Thermal denaturation experiment shows denaturation midpoint temperature (Tm) for SUWA is extremely high, 122°C. We also constructed a PN‐Block, SUWA‐foldon, and its native PAGE and SAXS analyses suggest that the SUWA‐foldon self‐assembled into several homooligomeric complexes. Partial thermal denaturation and reconstruction experiments demonstrated selective reassembly of the SUWA‐foldon supramolecular complexes. These results suggest that the PN‐Block strategy is useful for reconstructing self‐assembling protein complex nanoarchitectures.

ABS232

REVERSIBLE AND ORTHOGONAL FOUR HELIX BUNDLE HETERODIMERS

Ajasja Ljubetic ¹, Ryan Kibler², Zibo Chen², Sherry Bermeo², Roman Jerala³, David Baker⁴

¹Department of Biochemistry, UW; Institute for Protein Design, UW; Department of Synthetic Biology and Immunology, National Institute of Chemistry, Ljubljana, SI (Seattle, United States); ²Department of Biochemistry, UW; Institute for Protein Design, UW (Seattle, United States); ³Department of Synthetic Biology and Immunology, National Institute of Chemistry, Ljubljana, Slovenia. (Ljubljana, Slovenia); ⁴Department of Biochemistry, UW; Institute for Protein Design, UW; Howard Hughes Medical Institute, UW (Seattle, United States)

Reversible and dynamic control of protein‐protein interactions is a key goal of synthetic biology. Orthogonal four helix bundles (4HB) containing buried hydrogen bond networks have recently been designed. Such 4HB building blocks offer a large number of different interactions that could be controlled at the same time. However, the 4HB heterodimers are not well behaved when each part is expressed separately (usually homodimerization or aggregation is observed).

Here we show an approach to design 4HB building blocks, that are well behaved as monomers and therefore also reversible.

By comparing native and designed interface we noticed that native interfaces contain a larger fraction of polar residues. Therefore, we have designed 4HB heterodimers with a larger proportion of polar buried hydrogen bond networks.

Such building blocks have many applications. We have included 4HB building blocks into a coiled‐coil (CC) protein origami tetrahedral cage. 4HBs enable new topologies to be constructed, including a cage that can be completely opened by opening a single segment at the edge. Ultimately, we hope to include 4HBs which will act as switches, assembling and disassembling in response to external stimuli, for example change in pH or ligand binding.

Figure 1: A) Four helix bundles enable new topologies. Here a cage can be opened or closed with a single segment, which was not possible before. B) All atom model, with secondary structure shown as helices and a transparent surface.

21. Structure (x‐ray/NMR/EM): 8. Enzymology

ABS233

CRYSTAL STRUCTURE OF THE UDP‐GLUCOSE PYROPHOSPHORYLASE FROM YERSINIA PESTIS, AN ANTI‐PLAGUE DRUG TARGET

George Lountos ¹, Morgan Gibbs², Rajesh Gumpena², David Waugh²

¹Basic Science Program, Frederick National Laboratory for Cancer Research (Frederick, United States); ²Macromolecular Crystallography Laboratory, National Cancer Institute (Frederick, MD, United States)

Yersinia pestis, the etiological agent of bubonic plague, continues to be classified as a high‐threat, category A bioterrorism agent by the Center for Disease Control. Currently, only a small number of antibiotics are used as treatment options, and drug‐resistant isolates of Y. pestis have appeared in various geographical regions. Given the ongoing threat of its misuse as a bioterrorism agent, the discovery of novel therapeutic agents against Y. pestis warrants continued investigation. The UDP‐glucose pyrophosphorylase (UGP) from Y. pestis is a metabolic enzyme that has been shown to be important in the survival of Y. pestis in mouse macrophages. We have solved the crystal structure of the UDP‐glucose pyrophosphorylase from Y. pestis at 2.17 å resolution which reveals high structural homology with other reported bacterial UGPs. Given the shared sequence conservation of the active site between bacterial UGP homologs and their low sequence conservation with the eukaryotic UGP counterparts, it may be possible to develop broad‐spectrum inhibitors that target bacterial UGPs while avoiding off‐target effects in the eukaryotic host. The structure reported here provides the first‐step in the rational design of novel UGP inhibitors. Additionally, we will discuss how this project was designed and developed as a training module for an undergraduate student in protein expression and purification and structure determination by X‐ray crystallography for the National Institutes of Health Summer Internship Program in Biomedical Sciences.

7. Dynamics and allostery: 16. Protein interactions and assemblies

ABS235

POPULATION SHIFTS FROM ALLOSTERIC COUPLING OF RNA AND TRYPTOPHAN IN THE GENE‐REGULATING RING‐SHAPED PROTEIN TRAP

Melody Holmquist¹, Mark Foster ¹, Elihu Ihms², Weicheng Li¹, Cameron Jamshidi¹, Vicki Wysocki¹, Paul Gollnick¹

¹The Ohio State University (Columbus, United States); ²Vaccine Research Center, NIAID/NIH (Bethesda, United States)

Heterotropic allostery pervades macromolecular function, providing a means for regulating molecular properties. The un‐decameric (11‐mer) protein TRAP from Bacillus spp. participates in a feedback regulatory mechanism in which excess free tryptophan (Trp) activates TRAP to bind a specific mRNA sequence, resulting in attenuated expression of the trp operon. Thus, Trp and RNA are heterotropically coupled through their mutual interaction with TRAP.

Understanding the mechanism by which the Trp ligand regulates the RNA binding activity of TRAP requires quantifying the structural and thermodynamic coupling between the bound Trp ligands (up to 11), and between the Trp ligands and RNA at the microscopic level. However, this goal is complicated because allosteric effects can distort the proportionality between populations and traditional experimental observables like NMR chemical shifts (Fig 1, left). On the other hand, the mass‐shift in native mass spectrometry (nMS) depends only on the number of bound ligands, making it ideal for monitoring populations of liganded states.

We applied a nearest neighbor (NN) statistical thermodynamic model to obtain microscopic thermodynamic parameters from ITC data.1 Population distributions predicted from this model are observed directly by high‐resolution nMS (Figure 1, center). We also use nMS to quantify how heterotropic coupling of RNA to Trp causes redistribution of Trp ligands to TRAP rings with bound RNA. These findings illustrate how allostery is achieved by population shifts, and demonstrate the utility of nMS for describing complex allosteric behavior of regulatory macromolecules.

1. Ihms, et al. 2017, Biophysical Journal, 112(7): 132838. https://doi.org/10.1016/j.bpj.2017.02.031

5. Computational modeling/simulation: 20. Single molecule studies

ABS237

THE STRUCTURAL AND FUNCTIONAL ROLES OF DISULPHIDE BRIDGES IN THE SOLANUM TUBEROSUM PLANT SPECIFIC INSERT, A SAPOSIN‐LIKE PROTEIN

John H. Dupuis ¹, Rickey Y. Yada¹

¹Food, Nutrition, and Health Program, Faculty of Land and Food Systems, The University of British Columbia (Vancouver, Canada)

The plant specific insert (PSI) is found in nearly all plants, and has sequential and/or structural homologues across all kingdoms of life. In plants, it has a defensive role with the requirements of acidic pH and anionic lipids. The PSI contains a set of three highly conserved disulphide linkages that bridge together the proteins helical domains. To elucidate the role of disulphide linkages in maintaining overall protein tertiary structure, this work examined the effects of their sequential removal, both individually and in combination, using in silico methods. The effects of pH on disulphide removal were explored using all‐atom molecular dynamics (MD) and coarse‐grained MD (CGMD). Tertiary structure was found to remain stable at both acidic (active) and neutral (inactive) pH despite removal of disulphide linkages, with root‐mean‐square deviation and radius of gyration being remarkably stable over the 100 ns simulation. Negligible changes were seen in the intra‐ and inter‐monomer associations based on an analysis of estimated salt bridges, hydrogen bonds, and hydrophobic interactions. Localized unfolding from helical to unstructured was observed in the dimer during CGMD, allowing for enhanced membrane interactions. Where linkages were reduced, increased total interactions between the dimer and the membranes were observed. Interactions in helix 1 (monomer A), and helix 2 and the loop portion (monomer B) were greatly enhanced, particularly with the anionic lipid component. These results suggest that the role of the disulphide linkages is to regulate interactions with target membranes, a key property for intracellular antimicrobial peptides.

16. Protein interactions and assemblies: 1. Amyloid and aggregation

ABS238

NOVEL WAY TO STUDY THE FUNCTION OF NATIVE PROTEINS IN SOLUTION

Gabriella Kiss ¹, Matthias Langhorst¹, Gavin Young², Daniel Cole¹, Phillipp Kukura²

¹Refeyn Ltd, 33 George St, Oxford OX1 2AY, UK (BEAVERTON, United States); ²Department of Chemistry, Physical and Theoretical Chemistry Laboratory, University of Oxford (Oxford, United Kingdom)

The cellular processes underpinning life are orchestrated by proteins and the interactions they make with themselves and other biomolecules. A range of techniques has been developed to characterise these associations, but structural and dynamic heterogeneity remain a fundamental challenge. Mass photometry is based on interferometric scattering microscopy and can mass‐image single biomolecules in solution. Thereby, we can resolve oligomeric distributions at high dynamic range, mass‐measure polypeptides, glyco‐ and lipoproteins. These capabilities enable us to quantify the molecular mechanisms of processes as diverse as homo‐ and hetero‐oligomeric protein assembly, amyloidogenic protein aggregation and more [1]. The label free imaging and solution‐based measurement offers optimal settings for studying protein‐protein interactions, protein‐nucleic acid interactions or protein interaction with any molecules causing mass change. This ability to investigate biomolecules in their native state with high mass accuracy and resolution provides critical, complementary information to static structural techniques in the context of protein function and regulation.

It also enables us to investigate native conditions for the different biomolecules and optimise various buffer, salt and factor conditions necessary for their activity. Single molecule mass imaging provides access to protein dynamics and interactions and thus introduces a third, light‐based approach to measuring mass in addition to the historical mechanical and spectrometric methodologies and widens our capabilities to study biological processes.

1. Young et al., Quantitative mass imaging of single biological macromolecules. Science 2018 360: 423‐327

8. Enzymology: 21. Structure (x‐ray/NMR/EM)

ABS239

HYSTERESIS AND ALLOSTERY IN HUMAN UDP‐GLUCOSE DEHYDROGENASE REQUIRE A FLEXIBLE PROTEIN CORE

Nathaniel Beattie ¹, Brittany Pioso¹, Andrew Sidlo¹, Nicholas Keul¹, Zachary Wood¹

¹University of Georgia (Athens, United States)

Human UDP‐glucose dehydrogenase (hUGDH) oxidizes UDP‐glucose to UDP‐glucuronic acid, an essential substrate in the phase II metabolism of drugs. The activity of hUGDH is regulated by the conformation of a buried allosteric switch (T131 loop/6 helix). Substrate binding induces the allosteric switch to slowly isomerize from an inactive E* conformation to the active E state, which can be observed as enzyme hysteresis. When the feedback inhibitor UDP‐xylose binds, the allosteric switch and surrounding residues in the protein core repack, converting the hexamer into an inactive, horseshoe‐shaped complex (E). This allosteric transition is facilitated by large cavities and declivities in the protein core that provide the space required to accommodate the alternate packing arrangements. Here, we have used the A104L substitution to fill a cavity in the E state and sterically prevent repacking of the core into the E state. Steady state analysis shows that hUGDHA104L binds UDP‐xylose with lower affinity and that the inhibition is no longer cooperative. This means that the allosteric transition to the high‐UDP‐xylose affinity E state is blocked by the substitution. The crystal structures of hUGDHA104L show that the allosteric switch still adopts the E and E* states, albeit with a more rigid protein core. However, the progress curves of hUGDHA104L do not show hysteresis, which suggests that the E* and E states are now in rapid equilibrium. Our data suggest that hysteresis in native hUGDH originates from the conformational entropy of the E* state protein core.

17. Proteins in cells: 5. Computational modeling/simulation

ABS240

THE POLYDISPERSITY PROBLEM: INVESTIGATING THE EFFECT OF CROWDING AGENT POLYDISPERSITY ON PROTEIN STABILITY

Alan van Giessen ¹, Anastasia Osti¹

¹Mount Holyoke College (South Hadlet, United States)

The dense, heterogeneous cellular environment is known to affect protein stability through interactions with other biomacromolecules. The effect of excluded volume due to these biomolecules, also known as crowding agents, has long been known to increase the stability of a test protein. The cellular environment is heterogeneous not only in terms of its chemical composition, but also in terms of the sizes of the biomacromolecules present. It has been shown experimentally that the presence of polydisperse or mixed crowding agents has a non‐additive effect, i.e. that there is an optimal mixing ratio where the effect of the crowding agents is larger than that of monodisperse systems of each crowder. Here we investigate the role of polydisperse crowding on a small 15‐residue helical test protein through computer simulation. Our crowding agents are either spheres or helical proteins; however, the crowding agents only interact with the test protein through excluded volume interactions. For each shape of crowding agent, simulations were run with two sizes of crowders at different mixing ratios, ranging from all small crowders to all larger crowders at a constant density of 300 mg/mL. We relate the thermal stability and thermodynamic stability, as measured by the change in folding temperature and the change in the free energy of unfolding respectively, to the mixing ratio and to the shape of the crowding agent. In particular, we relate the non‐additivity to the excluded volume of the crowding agents. In addition, we discuss the impact of crowder polydispersity on the free energy landscapes.

16. Protein interactions and assemblies: 21. Structure (x‐ray/NMR/EM)

ABS241

THE ROLE OF NON‐MOTIF SELECTIVITY DETERMINANTS IN PDZ DOMAIN‐BINDING INTERACTIONS

Melody Gao ¹, Nick Pederson¹, Sarah Struyvenberg¹, Iain Mackley¹, Jeanine Amacher¹

¹Western Washington University (Bellingham, United States)

An important class of protein‐protein interactions in the cell involve recognition of short linear motifs (SLiMs) or peptides which often have a relatively weak affinity and are transient. Dysregulation of SLiM‐binding domains and target interactions are implicated in a number of human diseases. Due to the recognition of only a couple of amino acid positions by SLiM‐binding domains, interaction networks overlap and specific targeting is hard to achieve. The PDZ domain was used as a model system in order to define the role of non‐motif selectivity determinants in SLiM‐ or peptide‐mediated interactions. They recognize the extreme C‐terminus of target proteins and binding motifs are based on only two residues. Dissecting the binding networks of over 250 PDZ domains using only motif preferences is impossible. Likewise, two sequences can bind drastically different numbers of PDZ domains. For example, the C‐terminal sequence of the human papillomavirus E6 oncoprotein (HPV16 E6) interacts with over a dozen PDZ domain‐containing proteins, while the C‐terminal sequence of cystic fibrosis transmembrane conductance regulator (CFTR) interacts with less than five, yet both contain identical motif residues. We reveal that single‐residue substitution of peptides and structural biology allow us to determine non‐motif selectivity determinants for multiple PDZ domains that target the HPV16 E6 oncoprotein. In addition, we present the first known crystal structure of a PDZ domain from choanoflagellates, our closest non‐metazoan ancestor, and describe non‐motif selectivity determinants in these unique organisms. We show that non‐motif specificity is critical in order to characterize SLiM interactions networks in the cell.

18. Proteomics: 26. Other

ABS242

PROTEOME COMPARISON OF DIFFERENT HONEYS USING ELECTROPHORESIS AND MASS SPECTROMETRY

Tyler Thornton ¹, Taylor Anderson¹, Casey Harding¹, Rawlings Lyle¹, Clayton Rawson¹, Austin Sherwin¹, Craig Thulin¹

¹Utah Valley University (Orem, United States)

Honey is an important agricultural product, and elucidating its biochemistry can help us in understanding its sources and characteristics. We previously identified high‐abundance honey proteins from tryptic digests of total mixed protein precipitated with TCA by utilizing LCMSMS and data‐dependant acquisition (DDA) methodologies. We now report electrophoretic comparisons of the proteomes of various honeys and subsequent in‐gel digestion to identify compared proteins. Electrophoretic results demonstrate the presence of 4‐8 predominant bands and many less‐abundant proteins in each honey sample. Three of the prominent bands are the same across all honeys, and the rest appear at varying apparent molecular weights. The abundant proteins are Major Royal Jelly Proteins (MJRP) and have been identified in digests of TCA precipitates subjected to DDA LCMSMS analyses. The electrophoresis of proteins from different honey samples has demonstrated the diversity of lesser abundant proteins within the honey proteome. These less‐abundant proteins have not been previously characterized. We will report identifications of honey proteins found through in‐gel digestion coupled with LCMSMS.

16. Protein interactions and assemblies: 17. Proteins in cells

ABS246

ACTIVATION OF THE EXOCYST TETHERING COMPLEX FOR SNARE COMPLEX REGULATION AND MEMBRANE FUSION

Mary Munson ¹, Dante Lepore¹, Michael Feyder¹, Lillian Kenner², Leonora Martinez‐Nunez¹, Adam Frost²

¹University of Massachusetts Med School (Worcester, United States); ²University of California, San Francisco (San Francisco, United States)

A major challenge for a molecular understanding of membrane trafficking has been the elucidation of high resolution structures of large, multisubunit tethering complexes (MTCs) that spatially and temporally control intracellular membrane fusion. Exocyst is a hetero‐octameric protein complex, proposed to tether secretory vesicles at the plasma membrane to provide quality control of SNARE‐mediated membrane fusion. Breakthroughs in methodologies, including sample preparation, biochemical characterization, fluorescence and single‐particle cryoEM, are providing critical insights into the structure and function of the exocyst. We are investigating how the yeast exocyst interacts with SNARE proteins to control SNARE complex formation and membrane fusion. Intriguingly, fully assembled exocyst interacts weakly with the individual SNAREs and SNARE complexes, despite previously observed robust SNARE binding with the recombinant proteins. Using an auxin‐inducible degradation system and mutant yeast strains, we purified exocyst subcomplexes and mutant complexes and showed that several have increased affinities for the different SNAREs. Negative stain EM, combined with GFP tagging, was used to visualize and identify individual subunits within an exocyst subcomplex. Comparison of the negative stain images to the cryoEM structure of fully assembled exocyst revealed that several subunits become more dynamic and accessible for SNARE interactions. We propose that exocyst becomes activated and undergoes a conformational change, in order to efficiently bind SNAREs to stimulate fusion.

21. Structure (x‐ray/NMR/EM): 6. Design/engineering

ABS247

FIXED TARGET DELIVERY FOR SERIAL FEMTOSECOND CRYSTALLOGRAPHY OF WEAKLY‐DIFFRACTING OBJECTS

Megan Shelby ¹, Deepshika Gilbile², Thomas Grant³, Carolin Seuring⁴, Brent Segelke¹, Wei He¹, Angela Evans², Tim Pakendorf⁴, Pontus Fischer⁴, Mark Hunter⁵, Alex Batyuk⁵, Miriam Bathelmess⁴, Alke Meents⁴, Tonya Kuhl², Matthew Coleman¹, Matthias Frank¹

¹Lawrence Livermore National Lab (Livermore, United States); ²UC Davis (Davis, CA, United States); ³University of Buffalo (Buffalo, NY, United States); ⁴Center for Free‐Electron Laser Science (Hamburg, Germany); ⁵SLAC National Accelerator Lab (Menlo Park, CA, United States)

X‐ray Free Electron Lasers (XFEL) are enjoying an increasing impact as techniques utilizing their ultrafast, intense, and highly coherent pulses for structural biology are developed. These include serial femtosecond crystallography (SFX), which entails collection of nearly damage‐free diffraction patterns for crystalline and semi‐crystalline samples.

The goal of this work is to demonstrate the use of polymer thin films and graphene for sandwiching and support of various weakly diffracting objects on fixed targets as a generally applicable method for high‐throughput and high‐resolution biological imaging at room temperature. Rapid scanning using the Roadrunner fixed target system has recently been implemented at LCLS (Linac Coherent Light Source), at the MFX endstation in humidified helium and at the CXI endstation in vacuum. Our aim has been to adopt a rapid‐scanning approach in the both the vacuum environment of CXI and the humidified environment at MFX using polymer thin‐films and graphene as support materials to minimize the background for weakly diffracting samples. Our initial studies focus on 1) 2D crystals of Steptavidin (SA) and 3D microcrystals of 2) Rapid encystment protein (Rep24) to provide a benchmark for polymer/graphene sandwich performance and 3) nanolipoprotein (NLP) particles, or nanodiscs, weakly diffracting objects for which these constitute novel structural studies.

4. Chemical biology: 17. Proteins in cells

ABS248

OXIDIZED DOPAMINE CAUSES NEURONAL CELL DEATH BY IMPAIRING PROTEIN FUNCTION AND FOLDING

Dennis Özcelik ¹, Eduardo Felipe Alves Fernandes¹, Dominik Johann Essig¹

¹University of Copenhagen (Copenhagen, Denmark)

Parkinsons disease (PD) is a neurodegenerative disorder posing major challenges to aging society in the Western world. The loss of dopaminergic neurons in the brain is a hallmark of PD. Very recent studies suggest that oxidation of the neurotransmitter dopamine is implicated in disease development by generating oxidative stress; however, the molecular details are not fully understood.

We chemically modified dopamine and related compounds in order to elucidate the chemophysical properties of these molecules, and to monitor their role in biological systems. Our approach provides unique insights into the reactivity of dopamine‐related quinones. Further, via a combination of Huisgen cycloaddition chemistry (or click‐chemistry) and fluorescence microscopy, we found that these highly reactive quinones cause widespread protein modification under a large range of conditions in isolated proteins, cell lysates from human neuroblastoma cell lines and primary neurons from rodents. Next, we investigated the impact of dopamine‐related quinones on cell viability and demonstrated in vitro that this protein modification causes substantial neuronal cell death. We continued to identify intracellular targets of dopamine oxidation using a mass spectrometry approach. Our analysis produced a large number of protein targets including many cellular chaperones. Subsequent validation using functional in vitro assays demonstrates that dopamine oxidation inactivates major chaperone systems.

Our study points to a mechanism of neuronal cell death in the PD model that links oxidative stress and protein unfolding to dopamine oxidation and protein modification in dopaminergic neurons. Our finding suggests that there is another unexplored but important layer of neurochemistry contributing to PD development.

21. Structure (x‐ray/NMR/EM): 16. Protein interactions and assemblies

ABS249

STRUCTURAL ELUCIDATION OF A NOVEL, TANDEM DEUBIQUITINASE/UBIQUITIN‐BINDING DOMAIN FROM THE PATHOGENIC BACTERIUM, ORIENTIA TSUTSUGAMUSHI

Christopher Lim ¹, Jason Berk¹, Yong Xiong¹, Mark Hochstrasser¹

¹Yale University (New Haven, United States)

The obligate intracellular bacterium Orientia tsutsugamushi is the causative agent of scrub typhus, an acute, febrile disease endemic to India and many other South East Asian countries. Although scrub typhus is the cause of approximately 1 million infections and over 10,000 deaths annually, very little is known about the pathogenesis and biochemical mechanisms underpinning this neglected tropical disease. Through extensive sequence alignment, we have identified a protein expressed during Orientia infection (the OtDUB) that harbors a putative deubiquitinase (DUB) domain, a cryptic ubiquitin‐binding domain (UBD), as well as an extensive C‐terminal accessory domain of unknown function. In vitro, the cryptic UBD binds ubiquitin (Ub) with low nanomolar affinity as measured by (ITC), and is the highest affinity Ub interaction yet discovered.

Here, we present the 2.5 å crystal structure of the OtDUB in complex with ubiquitin. To our surprise, each molecule of the OtDUB is bound to three molecules of Ub, each with a unique binding site. The crystal structure explains the preference for cleavage of K63‐linked Ub in vitro, and bolsters the notion that the function of the DUB during infection is to counteract its own ubiquitination by host E3 ubiquitin ligases. Furthermore, the UBD interacts with ubiquitin using the canonical Ile44 patch, yet also employs non‐canonical electrostatic contacts that together span an extensive buried surface area of nearly 800 å2. Taken together, our biochemical and structural findings expand our understanding of the already diverse collection of ubiquitin‐binding domains, and may motivate discovery of new, high‐affinity UBDs.

3. Chaperones: 1. Amyloid and aggregation

ABS250

TRANSIENT INTERACTIONS INVOLVING A DISORDERED REGION OF HSPB1 DRIVE CHAPERONE ACTIVITY TOWARD TAU

Hannah Baughman ¹, Amanda Clouser², Rachel Klevit², Abhinav Nath³

¹University of Washington (Seattle, United States); ²University of Washington Department of Biochemistry (Seattle, United States); ³University of Washington Department of Medicinal Chemistry (Seattle, United States)

Molecular chaperones are vital players in maintaining protein solubility and preventing aberrant protein aggregation, which has been linked to numerous diseases. Small heat shock proteins (sHSPs) are a class of molecular chaperones that function in an ATP‐independent manner to engage destabilized or aggregation‐prone proteins and delay aggregation. All sHSPs consist of a core alpha‐crystallin domain flanked by highly variable, disordered N‐ and C‐terminal regions. These disordered terminal regions promote the formation of highly dynamic, polydisperse oligomeric ensembles that have frustrated attempts at structural characterization of sHSPs. Due to experimental challenges posed by such systems, understanding of the mechanisms of sHSP function remains rudimentary. In this work, we characterized interactions between the human small heat shock protein HspB1 and the microtubule‐associated protein tau, which has been implicated in multiple dementias including Alzheimers disease. Using NMR techniques and fluorescence assays, we have shown that HspB1 is able to delay the onset of tau aggregation. This chaperone activity is achieved through transient interactions between the disordered N‐terminal region of HspB1 and two aggregation‐prone motifs in tau. Tau has also been shown to bind the alpha‐crystallin domain of HspB1; however, we found that these interactions are not only insufficient for chaperone activity but are actually counterproductive. This work deepens our understanding of the strategies employed by cells to prevent pathological tau aggregation and sheds light on the mechanisms by which sHSPs achieve their function.

19. Proteostasis and quality control: 1. Amyloid and aggregation

ABS251

EXAMINING THE EFFECTS OF MUTATION ON THE AGGREGATION AND DEGRADATION OF AN ALS‐ASSOCIATED PROTEIN

Mikaela Elder ¹, Sean Cascarina¹, Lindsey Brookbank¹, Eric Ross¹

¹Colorado State University (Fort Collins, United States)

Protein aggregation is frequently deleterious, so cells possess extensive machinery designed to resolubilize or degrade aggregation‐prone proteins. We are using the human protein hnRNPA2 as a model to examine how a protein's amino acid sequence affects both its intrinsic aggregation propensity and the ability of the cellular proteostasis machinery to prevent this aggregation. Mutations in hnRNPA2 have been linked to degenerative diseases, including ALS. hnRNPA2 contains an aggregation‐prone prion‐like domain, and disease‐associated mutations in this domain increase the proteins aggregation propensity. To better understand how amino acid sequence affects this prion‐like domain, we randomly mutated a segment of hnRNPA2 and screened mutants for their degradation and aggregation propensity. Most amino acids that increased aggregation propensity increased protein degradation, suggesting that the cell is recognizing and degrading aggregation‐prone proteins. However, aromatic amino acids increased protein aggregation without increasing degradation propensity. This suggests that aromatic amino acids may have the unique ability to cause aggregation, without triggering recognition by the anti‐aggregation machinery. To confirm this result, we sequentially inserted aromatic amino acids into the aggregation‐prone segments of hnRNPA2 and analyzed the aggregation and degradation propensities of these mutants. As predicted, mutants with more aromatic amino acids showed increased aggregation, without increasing degradation. Future experiments will examine whether these trends apply to other aggregation‐prone proteins, potentially allowing for improved methods of predicting the effects of mutations.

3. Chaperones: 4. Chemical biology

ABS252

SMALL MOLECULE MODULATION OF HSP60/10 CHAPERONIN SYSTEMS: MORE COMMON THAN PREVIOUSLY THOUGHT?

Mckayla Stevens ¹, Sanofar Abdeen², Nilshad Salim², Anne‐Marie Ray², Alex Washburn², Siddhi Chitre², Jared Sivinski³, Yangshin Park², Quyen Hoang², Eli Chapman³, Steven Johnson²

¹Indiana University School of Medicine (Indianapolis, United States); ²IUSM (Indianapolis, United States); ³UA College of Pharmacy (Tucson, United States)

All living organisms contain a unique class of molecular chaperones called 60 kilodalton heat shock proteins (HSP60, also known as GroEL in bacteria). While some organisms contain more than one HSP60 or GroEL isoform, one has consistently proven essential in the micro‐organisms thus far evaluated. Accordingly, we have been investigating targeting HSP60 and GroEL chaperonin systems as an antibiotic strategy. Our initial studies focused on applying this antibiotic strategy for treating African Sleeping Sickness (caused by Trypanosoma brucei), drug‐resistant bacterial infections (particularly Methicillin‐resistant Staphylococcus aureus MRSA), and tuberculosis. Intriguingly, our studies found that three known antibiotics suramin, closantel, and rafoxanide were potent inhibitors of bacterial GroEL and human HSP60. These findings prompted us to explore what other known drugs, natural products, and bioactive molecules might also inhibit these chaperonin systems. Initial high‐throughput screening of 3,680 known drugs, natural products, and bioactive molecules identified 161 hit inhibitors of the Escherichia coli GroEL chaperonin system (4.3% hit rate). From a purchased subset of 60 hits, 29 compounds (48%) re‐confirmed as selective GroEL inhibitors in our assays, all of which were nearly equipotent against human HSP60. These findings illuminate the notion that targeting chaperonin systems might be a more common occurrence than originally appreciated. Future studies to determine the extent to which GroEL and HSP60 inhibition contributes to the function of these known drugs, natural products, and bioactive molecules in vivo are required.

24. Therapeutics and antibodies: 6. Design/engineering

ABS253

INFLUENCE OF THE ENDOPLASMIC RETICULUM LOCALIZATION SEQUENCE ON THE CYTOTOXICITY OF PSEUDOMONAS EXOTOXIN A‐BASED RECOMBINANT IMMUNOTOXINS

Jillian Baker ¹, John Weldon¹

¹Towson University (Baltimore, United States)

Recombinant immunotoxins (RITs) are chimeric proteins that are engineered to join an antibody to a protein toxin. The antibody allows for receptor‐specific targeting of cells while the toxin is internalized and subsequently induces cell death. One of the most commonly utilized RITs is based on Pseudomonas exotoxin A (PE), a bacterial toxin secreted by Pseudomonas aeruginosa. An important feature of the toxin is a C‐terminal endoplasmic reticulum (ER) retention sequence, REDLK, which is analogous to the canonical KDEL found at the C‐terminus of ER resident proteins. It has been shown that replacing the REDLK sequence of PE with KDEL enhances the cytotoxicity of PE‐based RITs. The goal of this project is to assess how mutations to the ER localization sequence affect cytotoxicity when different forms of PE are used. Both a 38‐kDA (PE38) and a 24‐kDa (PE‐LR or PE24) fragment of PE have been used to develop RITs. Preliminary data has shown that an anti‐transferrin receptor/PE24 RIT does not show enhanced cytotoxicity on HEK293 cells when the wild‐type REDLK sequence is replaced by KDEL. Conversely, deleting REDLK diminishes its cytotoxicity. RITs based on PE38 are being prepared for further testing.

1. Amyloid and aggregation: 2. Bioinformatics

ABS254

NATURAL AND PATHOGENIC PROTEIN SEQUENCE VARIATION AFFECTING PRION‐LIKE DOMAINS WITHIN AND ACROSS HUMAN PROTEOMES

Sean Cascarina ¹, Eric Ross¹

¹Colorado State University (Fort Collins, United States)

Protein aggregation is involved in a variety of muscular and neurodegenerative disorders. For many of these disorders, current models suggest a prion‐like molecular mechanism of disease, whereby protein aggregates structurally convert soluble protein to the misfolded form and spread to neighboring cells in an infectious manner. A variety of proteins with prion‐like domains (PrLDs) have recently been linked to these disorders. The development of prion prediction algorithms has facilitated the large‐scale identification of PrLDs among reference proteomes for various organisms. While initial identification of PrLDs from reference proteomes has resulted in important discoveries related to prion‐like proteins, the degree to which intraspecies protein sequence diversity influences predicted prion propensity has not been systematically examined. Here, we explore protein sequence variation introduced at the genetic, post‐transcriptional, and post‐translational levels, and its influence on predicted aggregation propensity for human PrLDs. We find that sequence variation is relatively common among PrLDs and in some cases can result in relatively large differences in predicted prion propensity. Analysis of a clinical variant database (consisting of single amino acid substitutions associated with human diseases) reveals a number of mutations within PrLDs that are predicted to increase prion propensity. Our results expand the list of candidate human PrLDs, estimate the effects of sequence variation on the aggregation propensity of PrLDs, and suggest the involvement prion‐like mechanisms in additional human diseases.

ABS256

DEVELOPMENT OF LOCKR‐ACTIVATING LOGIC CIRCUITS

Ryan Kibler ¹, Zibo Chen¹, Marc Lajoie¹, Bobby Langan¹, David Baker¹

¹University of Washington (Seattle, United States)

Customizing and strictly controlling cellular behavior is a major goal of synthetic biology. This is predominantly accomplished via genetic circuits or the rewiring of protein signaling pathways by exchanging protein domains. These strategies, while shown to be powerful and effective, are limited by the availability of natural parts and are made less predictable by poorly understood crosstalk in signaling pathways and stochastic gene expression in genetic circuits. Synthetic signaling pathways built from de novo proteins could avoid these issues by utilizing a large catalog of designed proteins with well‐defined interactions. The recent developments of Latching Orthogonal Cage‐Key pRoteins (LOCKR) and Cooperatively‐Inducible Protein HeterodimeR (CIPHR) logic enables the construction of such synthetic signaling proteins. This poster presents work toward building de novo signaling modules which function by coupling CIPHR logic operation to LOCKR activation. Currently, this system achieves a weak 2‐5 fold activation of LOCKR through flexible genetic fusions of the LOCKR switch and key to CIPHR gates. Imposing rigid structure on the fusion junctions should improve the fold activation by positioning the switch and key in the optimal conformation for LOCKR activation. Work on the sequestration of Designed HeteroDimers (DHDs), the inputs to CIPHR logic, in LOCKR will also be presented. This will enable signal propagation between CIPHR logic gates and make them composable.

21. Structure (x‐ray/NMR/EM): 24. Therapeutics and antibodies

ABS258

ACCELERATING INFECTIOUS DISEASE RESEARCH THROUGH STRUCTURAL GENOMICS ‐ THE SEATTLE STRUCTURAL GENOMICS CENTER FOR INFECTIOUS DISEASES (SSGCID)

Garry W. Buchko ¹, Thomas E. Edwards², Donald Lorimer², Bart L. Staker³, Robin Stacy³, David Veesler⁴, Gabriele Varani⁴, Lance J. Stewart⁴, Wesley C. Myler⁴, Peter J. Myler³

¹Pacific Northwest National Laboratory (Richland, United States); ²UCB (Bainbridge Island, United States); ³Seattle Children's Research Institute (Seattle, United States); ⁴University of Washington (Seattle, United States)

The Seattle Structural Genomics Center for Infectious Disease (http://ssgcid.org) is one of two Structural Genomics centers established by NIAID in 2007 to solve protein structures from organisms causing infectious disease. Both Centers actively engage with infectious disease researchers to select Community Request (CR) targets for entry in their structure determination pipelines and to collaboratively interpret and publish results from successful structure determinations. SSGCID target selection focuses on essential enzymes, virulence factors, drug targets and vaccine candidates from numerous bacterial, eukaryotic, viral pathogens. In general, target genes are PCR amplified, cloned and screened for soluble expression in Escherichia coli. Proteins are then purified in milligram amounts, screened for crystallization, and analyzed by X‐ray diffraction using an in‐house source or off‐site synchrotron beam‐line. To address solving structures for proteins that fail to crystallize, proteins <25 kDa in molecular weight are queued for structure determination by NMR, and recently, selected targets > ~100 kDa are queued for cryo‐EM. Since project inception in late 2007, over 6,000 targets have entered the SSGCID structure determination pipeline, of which >4000 are CRs, resulting in deposition of nearly 1,200 protein structures in the Protein Data Bank (PDB). In addition, more than 8000 expression clones and 4000 purified proteins are publicly available free of charge. An overview of SSGCID will be provided, highlighted with examples of recently solved structures.

16. Protein interactions and assemblies: 11. Intrinsically disordered proteins

ABS259

AMELOGENIN ‐ A MULTI‐PRONGED APPROACH TO IDENTIFY STRUCTURAL FEATURES GUIDING ENAMEL FORMATION

Garry W. Buchko ¹, Jinhui Tao¹, Rajith J. Arachchige¹, Sarah D. Burton¹, Yongsoon Shin¹, Bojana Ginovska¹, Barbara J. Tarasevich¹, Wendy J. Shaw¹

¹Pacific Northwest National Laboratory (Richland, United States)

The weaving of hydroxyapatite into enamel, one of natures hardest materials, depends on an ~180‐residue protein, amelogenin, the predominant enamel matrix protein at the initial stages of amelogenesis. While the molecular mechanism amelogenin uses in synthesizing the unique architecture of enamel is poorly understood, it is expected that structural features of the protein is crucial. To address the impact of amelogenins structure in orchestrating biomineralization we are employing a suite of diverse, non‐conventional techniques to characterize the structure of amelogenin in both the solution‐ and solid‐state. Quantitative high resolution, in situ atomic force microscopy (AFM) employed to study the energetics of murine amelogenin (wild‐type, two naturally occurring single‐amino acid variants associated with amelogenis imperfecta (T21I and P41T), and a model variant (P71T)) binding onto single crystal hydroxyapatite (100) in real time identified differences that may be responsible for disrupting normal hydroxyapatite growth and enzymatic degradation. Dipolar Assisted Rotational Resonance solid‐state NMR experiments applied to study the structure of wildtype and mutant amelogenin containing 13C‐ and 15N‐residue specific labels identified beta‐sheet secondary structure in defined regions of the protein when bound to hydroxyapatite crystals. High‐resolution solution‐state NMR studies to survey the pH‐induced self‐assembly of amelogenin provided residue‐specific details of the process highlighted by the presence of amide resonances for the C‐terminal region (S157‐D180) over the entire pH range. These are all powerful tools to be further exploited towards understanding the mechanisms of biomineralization in general and towards predictively synthesizing functional materials tailored by macromolecular design.

6. Design/engineering: 22. Synthetic biology

ABS260

HARNESSING BACKBONE STRAIN TO DESIGN BETA‐BARREL PROTEINS DE NOVO: FROM FIRST PRINCIPLES TO APPLICATION

Anastassia Vorobieva ¹, Jiayi Dou², William Sheffler¹, Binchen Mao¹, Matthew Bick¹, Lindsey Doyle³, Jason Klima¹, Lauren Gagnon¹, Yakov Kipnis¹, Barry Stoddard³, David Baker¹

¹University of Washington (Seattle, United States); ²Stanford University (Stanford, United States); ³Fred Hutchinson Cancer Research Center (Seattle, United States)

The up‐and‐down beta‐barrel an arrangement of antiparallel beta‐strands into a single beta‐sheet that twists to close on itself is an excellent protein fold for engineering new activity such as small‐molecule binding and catalysis. It is therefore a target of choice for de novo design, which requires an accurate model of the protein backbone as a starting point for sequence design calculations. Despite the existence of many parametric models that describe the beta‐barrel fold, none have been successfully applied to design a protein that folds into a beta‐barrel, to our knowledge. Our own attempts of de novo design with one of such parametric models produced mostly insoluble proteins. The parametric model assumed constant twist between the beta‐strands that comprise the beta‐barrel. Using a rational approach coupled with computational modelling, we found that this assumption results in strained protein backbones to achieve favourable hydrogen‐bonding geometries through the beta‐sheet. The strain can be relieved by strategic placement of glycine residues (glycine kinks) and of beta‐bulges into the beta‐strands. After incorporating these torsional irregularities in the protein backbone, we could design a sequence that folds into a thermostable beta‐barrel with a new shape not observed in native proteins. We will show that the described principles can be used to generate proteins robust enough to support small‐molecule binding and catalysis in addition to tolerating a wide range of insertions into the beta‐turns. Finally, we will discuss how these principles can be applied to the design of transmembrane beta‐barrels.

24. Therapeutics and antibodies: 16. Protein interactions and assemblies

ABS261

NON‐IDEALITY OF PROTEIN‐BASED THERAPEUTICS IN BIOLOGICAL ENVIRONMENTS

Hayli Larsen ¹

¹University of Washington Department of Medicinal Chemistry (Seattle, United States)

Biologics, also known as protein‐based therapeutics, show great potential in the treatment of a wide variety of diseases ranging from asthma, Crohn's disease, diabetes, to some of the most aggressive cancers. Despite showing great promise, biologic drug development is difficult, expensive, and takes, on average, over 10 years from initial screening to approval. Our understanding of factors controlling pharmacokinetics (PK) and clearance of biologics is extremely limited, and falls far behind our understanding in corresponding areas of small‐molecule drugs. Biologics interact with their targets and receptors in complex, crowded, non‐ideal environments such as serum and the endosomal lumen. The effects of these complex biological fluids could very well be inadequately represented in traditional, ideal buffer systems used for in vitro characterization. One approach to understanding these effects is to investigate how non‐ideal environments affect the biophysical properties of biologics and model proteins. We hypothesize that non‐ideal behaviors of biologics in serum can ultimately be predicted, given accurate characterization and using relevant model systems. We are currently addressing this hypothesis through the use of fluorescence correlation spectroscopy, FCS, a biophysical technique that measures the diffusive properties of low concentration fluorescent molecules as they move through a well‐defined detection volume. FCS allows us to characterize the specific and non‐specific interactions of biologics and model antibodies with serum constituents and other binding partners. Utilizing this technique, we have observed non‐ideality in three antibodies through second virial coefficient calculations that suggest attractive interactions between these antibodies (at nanomolar concentrations) and serum components.

6. Design/engineering: 5. Computational modeling/simulation

ABS262

STRUCTURAL STUDIES OF ENGINEERED ADENO‐ASSOCIATED VIRUS CAPSIDS THAT CROSS BLOOD‐BRAIN BARRIER EFFICIENTLY

Xiaozhe Ding ¹, Sripriya Kumar², Andrey Malyutin², Viviana Gradinaru²

¹California Institute of Technology (Pasadena, United States); ²Caltech (Pasadena, United States)

Adeno‐associated virus (AAV) is a small (25nm diameter), non‐pathogenic Parvovirus. Because of its broad tropism and documented safety profile, recombinant AAV has become one of the most widely‐used gene delivery vectors in gene therapy trials and neuroscience research. Through directed evolution, we have engineered AAV capsid variants that allowed gene delivery to central and peripheral nervous systems of animals upon noninvasive systemic delivery. However, the biochemical mechanisms for the variants that cross the blood‐brain barrier (BBB) efficiently remain elusive.

Using single‐particle CryoEM, we obtained atomic structures of several AAV variants including PHP.eB, an AAV9 variant with the highest CNS transduction efficiency among the variants. Surprisingly, the structural changes caused by the PHP.eB peptide insertion are restricted to the site of modification while the rest of the capsid protein has an identical structure to that of AAV9. This makes computational modeling of the structure of these loop insertions with atomic accuracy feasible.

To look for structural features that differentiate BBB‐crossing AAV variants from the other AAV variants, we developed an automatic pipeline to obtain structure models of a larger number of engineered AAV variants by computational simulation. This pipeline allowed us to model loop structures of hundreds of AAV variants in less than a week using Rosetta protocols such as comparative modeling (Song et al., 2013) and remodeling on Amazon cloud computing services. These experimental (by CryoEM) and computational (by Rosetta) structures will lay the foundation for a mechanistic understanding of the BBB traversal capability of PHP.eB‐like AAV variants.

8. Enzymology: 4. Chemical biology

ABS263

OS9BGLU31 TRANSGLUCOSIDASE VARIANTS WITH HIGH AND PROMISCUOUS ACTIVITY

James R Ketudat Cairns ¹, Linh Tran², Sunaree Choknud³, Vincent Blay Roger⁴, Robert C. Robinson²

¹Suranaree University of Technology (Nakhon Ratchasima Muang District, Thailand); ²Research Institute for Interdisciplinary Science, Okayama University (Okayama, Japan); ³Center for Biomolecular Structure, Function and Application and School of Chemistry, Suranaree University of Technology (Nakhon Ratchasima, Thailand); ⁴The Ohio State University (Columbus, OH, United States)

Glycoside hydrolase family 1 (GH1) primarily contains beta‐glucosidases and related hydrolase enzymes, but in the last 10 years it has been shown to also contain transglycosidases with little hydrolase activity. One such transglycosidase is Os9BGlu31, which transfers glucose between 1‐O‐beta‐glucosyl phenolic esters and their corresponding free acids, along with certain aromatic alcohols and fatty acids. Previously, we showed that mutation of single amino acids near the acceptor binding position from hydrophobic to hydrophilic residues had relatively little effect on the ratio of transglycosylation versus hydrolysis, while mutation the W243N increased both hydrolysis and transglycosylation. Mutation of the W243 residue to all remaining common amino acids identified W243L as another variant with high activity and different preferences compared to W243N. Furthermore, combining the previously described low activity mutationL241D with the W243N mutation resulted in a variant with the highest hydrolysis rate and higher rates for transglycosylation of certain phenolic acids and the flavonoid luteolin. This epistasis in the combined mutations likely reflects the close proximity of these two residues in the active site. Since homology modeling and docking could not adequately explain the relative activities of the variants, we have recently crystallized Os9BGlu31 W243L, the structure of which will provide more precision for explaining enzyme‐substrate interactions. This work has generated a useful tool set for generating glycoconjugates of interest, including phytohormone glucoconjugates under study in our laboratory. Further structural information should enable us to improve these tools for efficient production of more diverse glucoconjugates.

8. Enzymology: 25. Transcription/translation/post‐translational modifications

ABS264

GENETIC SELECTIONS FOR THE DISCOVERY OF NEW REDUCTASES AND OXIDASES OF METHIONINE

Bruno Manta ¹, Mehmet Berkmen¹

¹New England Biolabs (Ipswich, United States)

Oxidation of methionine to methionine sulfoxide (MSO) is considered an oxidative damage that can lead to protein dysfunction or degradation. However, oxidized methionine can also be repaired by enzymes called methionine sulfoxide reductases (MSR). MSRs are conserved from prokaryotes to mammals and represent one of the largest groups of reductases devoted to revert oxidative damage. Considering that only a few chemical oxidants target methionine, particularly hydrogen peroxide and hypochlorous acid, a major question in the field is how methionine is oxidized in vivo. A major breakthrough was the discovery of vertebrate‐specific enzymes MICALs, large multidomain proteins that oxidize conserved methionines on actin affecting its polymerization. The discovery MICALs indicated the existence of enzymes that can oxidize methionine. The objective of our work is to identify new oxidases of methionine in prokaryotic genetic selections and screen libraries of environmental and microbiome genomic DNA. Selection was done on an E. coli strain lacking five MSRs (5) that cant grow on minimal media when MSO is provided instead of methionine. When libraries are transformed on that strain, a growth recovery on MSO will capture any functional MSR capable of reducing MSO to methionine. To screen for methionine oxidases, we developed a synthetic lethality screen forcing 5 to retain a plasmid carrying an MSRs when a potential oxidase is expressed from plasmid. The hits we obtained will allow us to identify the building blocks of methionine sulfoxide‐based signaling and to speculate on the conservation and evolution of methionine redox signaling.

21. Structure (x‐ray/NMR/EM): 1. Amyloid and aggregation

ABS265

STRUCTURE AND DYNAMICS OF TAU AMYLOID FIBRILS INVESTIGATED BY SOLID‐STATE NMR SPECTROSCOPY

Aurelio Dregni ¹, Venkata S. Mandala¹, Haifan Wu², Matthew R. Elkins³, William F. DeGrado², Mei Hong³

¹Department of Chemistry, Massachusetts Institute of Technology, 170 Albany Street, Cambridge, MA 02139 (Cambridge, United States); ²Department of Pharmaceutical Chemistry, University of California, San Francisco, San Francisco, CA 94158 (San Francisco, United States); ³Department of Chemistry, Massachusetts Institute of Technology, 170 Albany Street, Cambridge, MA 02139 (Cambridge, United States)

Tau is an intrinsically disordered protein that stabilizes microtubules in neurons. However, tau can pathologically misfold to form fibrillar aggregates in many neurodegenerative diseases such as Alzheimers disease (AD). Understanding the structure and misfolding pathways of tau amyloid fibrils is therefore essential for elucidating the molecular basis of these diseases and for rational design of aggregation inhibitors. Tau fibrils extracted from the brains of patients with AD and Picks disease show disease‐specific molecular conformations. In comparison, a recent study of recombinant 2N4R tau fibrillized in vitro using heparin found polymorphic fibril morphologies, with at least three distinct molecular conformations observed.

We have now investigated the three‐dimensional fold and dynamics of heparin‐fibrillized full‐length 0N4R tau using solid‐state NMR spectroscopy. A large number of two‐ and three‐dimensional 15N‐13C and 13C‐13C correlation NMR spectra reveal that 0N4R tau fibrils adopt a unique ‐sheet conformation, which differs from previously observed structures. The tau fibrils exhibit heterogeneous dynamics: in addition to the rigid ‐sheet core, partially mobile and nearly isotropically mobile domains are observed and distinguished using various solid‐state NMR experiments. These results have implications for how the dynamics of the disordered domain of tau shapes the structure of the ‐sheet core.

17. Proteins in cells: 10. Folding

ABS267

CRITICAL PHENOMENA IN THE TEMPERATURE‐PRESSURE‐CROWDING PHASE DIAGRAM OF A PROTEIN

Margaret Cheung ¹, Andrei Gasic¹, Mayank Boob², Maxim Prigozhin³, Dirar Homouz⁴, Anna Wirth², Caleb Daugherty¹, Martin Gruebele²

¹University of Houston (Houston, United States); ²University of Illinois, Urbana‐Champaign (Champaign, United States); ³Stanford (Stanford, United States); ⁴Khalifa University of Science and Technology (Abu Dhabi, United Arab Emirates)

In the cell, proteins fold and perform complex functions through global structural rearrangements. Function requires a protein to be at the brink of stability to be susceptible to small environmental fluctuations, yet stable enough to maintain structural integrity. These apparently conflicting behaviors are exhibited by systems near a critical point, where distinct phases merge ‐ a concept contrary to previous studies indicating proteins have a well‐defined folded/unfolded phase boundary in the pressure‐temperature plane. Here, by modeling the protein phosphoglycerate kinase (PGK) on the temperature (T), pressure (P), and crowding volume‐fraction () phase diagram, we demonstrate a critical transition where new phases merge, and PGK exhibits large structural fluctuations. Above the critical point, intermediate conformations between folded and unfolded phases disappear. When increases, this point moves to lower Tc. We verify the calculations with experiments mapping the T‐P‐ space, which likewise reveal a critical point at 305 K and 170 MPa that moves to lower Tc as increases. Crowding places PGK near a critical line in its natural parameter space, where large conformational changes can occur without costly free energy barriers. Specific structures are proposed for each phase based on simulation.

12. Membrane proteins: 4. Chemical biology

ABS269

MODULATION OF ROD OPSIN STABILITY, FUNCTION AND MEMBRANE SUPRAMOLECULAR ORGANIZATION BY FLAVONOIDS

Joseph T Ortega ¹, Tanu Parmar¹, Beata Jastrzebska¹

¹Case Western Reserve University (Cleveland, United States)

Rhodopsin (Rho), a visual G protein‐coupled receptor is expressed in rod photoreceptors, where it transmits a light signal into the cell and mediates its conversion into a nervous impulse. Over a hundred mutations in Rho are linked to various ocular impairments. Thus, efforts are directed towards developing novel effective, ligands targeting Rho, improving its folding and stability. Modulation of Rho structure by flavonoids and their beneficial effect in several eye diseases have been reported. Thus, these natural compounds could provide a viable approach to drug discovery efforts. However, the underlying mechanism of their action is not fully understood. In this study, we used molecular docking, thermal stability, and functional assays, including Gt activation and Meta II decay to clarify the effects of two common bioactive flavonoids, quercetin and myricetin on rod opsin stability and function. We also utilized mammalian cell expression systems, high content fluorescent imaging, and bioluminescence resonance energy transfer to evaluate the effect of quercetin and myricetin on membrane trafficking and supramolecular organization within the phospholipid bilayer. We found that through direct interaction with ligand‐free opsin, flavonoids modulate its conformation, allowing faster entry of the retinal chromophore into the binding pocket. Moreover, flavonoids enhanced opsin stability, most likely through introducing structural rigidity and promoting receptor self‐association within biological membranes. The binding of flavonoids to retinitis pigmentosa‐associated P23H rod opsin mutant partially rescued its membrane trafficking defects. Together, our results suggest possibilities to utilize flavonoids as lead compounds to discover novel non‐retinoid therapeutics for the treatment of Rho‐related retinopathies.

24. Therapeutics and antibodies: 17. Proteins in cells

ABS270

MODULATION OF IGG BLOOD‐BRAIN BARRIER PERMEABILITY VIA FAB GLYCAN SIALYLATION

JOHN FINKE ¹, Emily Swanson², Lewis Samantha², Ayres Kari², Emily Wing³, William Banks³

¹University of Washington (University Place, United States); ²University of Washington Tacoma (Tacoma, United States); ³VA Puget Sound Health Care System (Seattle, United States)

A major limitation in treatments for central nervous system diseases is the inability to deliver drugs to the target in the brain. To determine a role for protein glycosylation in blood‐brain barrier transcytosis, we investigated whether sialylation of Fab glycans alters permeability of IgG antibodies through the blood‐brain barrier. This was accomplished using monoclonal antibody 4G8 with a single sialylated Fab glycan in the VL region and minimal sialic acid on the Fc glycan. Using an in vitro BBB model, sialylation of 4G8 did not increase influx into the brain but, instead, reduced efflux from the brain. The Fc glycan had no impact on 4G8 BBB permeability but its removal abrogated the sialic acid‐mediated efflux reduction. Studies with human IgG showed that sialic acid reduced BBB permeability in both the influx and efflux directions. In summary, sialylation appears to be an important regulatory of transcytosis in the neurovascular unit.

24. Therapeutics and antibodies: 25. Transcription/translation/post‐translational modifications

ABS271

MASS SPECTROMETRY PROFILING OF N‐LINKED GLYCANS THAT MODULATE IGG BLOOD‐BRAIN BARRIER PERMEABILITY

JOHN FINKE¹, Emily Swanson ¹, Samantha Lewis², Abigail Deleon², Emily Wing³, William Banks³

¹University of Washington (Tacoma, United States); ²University of Washington Tacoma (Tacoma, United States); ³VA Puget Sound Health Care System (Tacoma, United States)

Glycan modification may provide a means to improve delivery of antibodies and other biopharmaceutics into the central nervous system. However, the inherent complexity of glycans makes a straightforward modification strategy unclear. Here, we used LC/MS to characterize the full glycan profile of two antibody preparations with different BBB permeability behavior: (1) Monoclonal antibody 4G8 and (2) human IgG from pooled serum. It was found that the 4G8 profile is highly processed and homogeneous, with >90% of sialylated Fab glycans in the G3FS1 configuration, with a high fraction of glycolated sialic acid, and > 75% of asialylated Fab glycans in the G4F configuration. By contrast, human serum IgG glycans were highly diverse, with less processing in Fab glycans and more sialylation in Fc glycans. This information will be used to guide future glycan modification efforts to improve brain penetration of IgG drugs

25. Transcription/translation/post‐translational modifications: 16. Protein interactions and assemblies

ABS275

BIOPHYSICAL AND STRUCTURAL ANALYSIS OF ABDOMINAL A AND ABDOMINAL B HOMEODOMAIN TRANSCRIPTION FACTORS

Rylee Simons ¹, Donald Spratt¹, Rachel Orlomoski¹, Jaqueline Dresch¹, Robert Drewell²

¹Clark University (Worcester, United States); ²Clark University (Worcester, United States)

Homeodomain transcription factors (HDs) regulate gene expression by binding to cis‐regulatory modules (CRMs) that control many important pathways and biological functions, particularly during early embryo development. These proteins are evolutionarily conserved and contain a 60 amino acid sequence called the homeobox. This highly‐basic region of the HDs is responsible for binding to DNA. Two homeodomain transcription factors in Drosophila melanogaster are Abdominal A (AbdA) and Abdominal B (AbdB). These proteins are part of the HOM‐C complex in Drosophila and are related to the human HOX genes. Both proteins regulate the abdominal segments of the fly and mutations result in improper development and disease. Additionally, AbdA has been shown to regulate muscle and nervous system development and AbdB is needed to form correct genitalia and left/right symmetry. To examine how AbdA and AbdB bind and recognize specific DNA sequences, the homeoboxes of AbdA and AbdB were overexpressed in E. coli and purified for biophysical and structural studies. Using electrophoretic mobility shift assays (EMSAs) and nuclear magnetic resonance (NMR) spectroscopy, we demonstrate that the isolated proteins are well folded and functional. We have also employed isothermal titration calorimetry (ITC) to measure the binding affinity of AbdA and AbdB with different DNA sequences. These cumulative results demonstrate that 1) the homeobox of AbdA and AbdB can be successfully purified; 2) they exhibit different binding affinities depending on the DNA sequence presented; and 3) that our future 3D NMR structural studies are feasible.

11. Intrinsically disordered proteins: 5. Computational modeling/simulation

ABS277

EXPLOITING AUTOINHIBITION MECHANISM FOR SCREENING OF SMALL MOLECULES THAT MODULATE DNA BINDING TO ETS TRANSCRIPTION FACTORS

Jennifer Bui ¹, Cecilia Borajero¹, Lawrence McIntosh¹, Joerg Gsponer¹

¹University of British Columbia (Vancouver, Canada)

Autoinhibition is a wide‐spread regulatory strategy for fine‐tuning activities of many signaling and regulatory proteins. The Ets1 proto‐oncoprotein is autoinhibited by intrinsically disordered regions (IDR) adjacent to its DNA binding domain (DBD). To delineate molecular mechanisms that allow Ets1 switch between an active, DNA binding state and an inhibited state, we used a combination of Rosetta and Molecular Dynamic modellings as well as NMR studies. We found IDR of Ets‐1 mimics a major grove of DNA backbone. We also discovered an allosteric pocket that forms by residues of IDR which then used for screening of small molecules. From our multi‐screening protocol and NMR titrations, we found SMR000368692 mimics the binding of IDR to DBD. It activates the same network of interacting residues as also observed in the inhibited phosphorylated state. Our approach presents an opportunity for inhibitor discoveries targeting transcription factors that contain long stretches of IDRs.

6. Design/engineering: 5. Computational modeling/simulation

ABS278

DESIGNING BUTTRESSED LOOPS TO DIVERSIFY THE FUNCTIONALITYOF DE NOVO PROTEIN SCAFFOLDS

Hanlun Jiang ¹

¹University of Washington (Seattle, United States)

Despite of their crucial roles in protein functions, loops longer than five residues are rarely present in de novo proteins. This is due to the enormous conformational space of loops and the lack of rules describing the structural and sequence features that can guide the design. A common strategy used by native proteins to stabilize long loops is to form extensive interactions between loops and their neighboring residues, known as buttressing. Inspired by the native proteins and the recent advances in loop design methods, we propose to develop new protocols for building buttressed long loops on two types of protein scaffolds: designed helical repeat proteins and de novo TIM barrels. Computationally and experimentally validated designs will potentially lead to not only the functionalization of the de novo protein scaffolds but also the new rules that guide the loop design in future.

7. Dynamics and allostery: 8. Enzymology

ABS279

UNDERSTANDING THE INTERFACE: EXPLORING MALATE DEHYDROGENASE USING COMPUTATIONAL AND EXPERIMENTAL APPROACHES

Ellis Bell ¹, James Burnett², Michael Schwabe¹, Jessica Bell¹

¹University of San Diego (San Diego, United States); ²University of Pittsburgh (Pittsburgh, United States)

Dimeric Malate Dehydrogenase exhibits properties attributed to subunit interactions. The dimer interface comprises 47 residues, clustered in four groupings in the sequence, 15 residues are conserved in eukaryata, with 7 more functionally conserved. Structures of watermelon glyoxysomal MDH with or without the allosteric ligand Citrate bound to one subunit, were examined to explore the nature of subunit contacts (HINT). In addition, to examine second sphere residues with potential roles in catalysis, and to establish differences in conserved crystallographic water molecules we used POOL and DRoP respectively. Intra‐ and inter‐molecular HINT analysis with no ligands bound versus the dimer with Citrate bound to one subunit indicates that D87 forms multiple hydrogen bonds within the interfacial 266‐270 loop region, some having increased intensity with Citrate bound, (mobile loop closed) as compared to no ligands bound, (mobile loop open). Further analysis suggests R196 and T268 lose favorable interactions with D87 on the opposite subunit, while E256 loses unfavorable interactions with D90 upon citrate binding which draws S266 further into the active site causing T268 to shift away from D87 and closer to Q58. This affects L269‐Q58 interaction across the interface. S266A and L269A mutants show loss of citrate inhibition and binding, and diminished substrate inhibition. T268D and I88A show little impact on cofactor binding although I88A becomes monomeric as shown by SEC and cross‐linking. Coupled with changes in S266, T268 and L269 interactions across the interface, it appears that subunit interactions are triggered by cofactor induced changes in L269‐Q58 interactions between subunits.

7. Dynamics and allostery: 21. Structure (x‐ray/NMR/EM)

ABS280

TURNING UP THE HEAT ON DYNAMIC PROTEINS: OBSERVING MOLECULAR MOTION IN REAL TIME WITH TEMPERATURE‐JUMP X‐RAY CRYSTALLOGRAPHY

Michael Thompson ¹, Alexander Wolff¹, Eriko Nango², Minoru Kubo³, Iris Young¹, Takanori Nakane⁴, Michihiro Sugahara², Rie Tanaka², Kazutaka Ito⁵, Aaron Brewster⁶, Shigeki Owada², Fumiaki Yumoto⁷, Nicholas Sauter⁶, Kensuke Tono⁸, So Iwata⁹, James Fraser¹

¹University of California, San Francisco (San Francisco, United States); ²RIKEN SPring‐8 center (Hyogo, Japan); ³University of Hyogo (Hyogo, Japan); ⁴MRC Laboratory of Molecular Biology (Cambridge, United Kingdom); ⁵Asahei‐Kasei Pharmaceutical Company (Tokyo, Japan); ⁶Lawrence Berkeley National Laboratory (Berkeley, United States); ⁷KEK High‐Energy Accelerator Research Organization (Tokyo, Japan); ⁸Japan Synchrotron Radiation Institute (Hyogo, Japan); ⁹University of Kyoto (Kyoto, Japan)

The importance of dynamics for protein function is widely appreciated; however, it remains challenging to understand, in atomic detail, how a molecules biological activity is enabled by the physical coupling of its conformational fluctuations across varied length and time scales. Toward this end, time‐resolved X‐ray crystallography, in which fast X‐ray pulses are used to measure structural changes of macromolecules in real time, is a powerful tool for studying dynamics because it provides simultaneous structural and kinetic data. Unfortunately, its widespread use in structural biology has been limited by a major technical hurdle: in order to observe dynamic behavior from an ensemble‐averaged experimental measurement, it is necessary to synchronize conformational changes for a significant fraction of molecules in the sample. Recently, we have been able to overcome this challenge by exploiting the temperature‐sensitivity of protein conformational ensembles. Specifically, we performed temperature‐jump crystallography, in which a pulsed infrared laser was used to rapidly heat crystallized proteins, initiating conformational dynamics that were subsequently monitored in real time using ultrafast X‐ray pulses from a free‐electron laser. Our experiments captured signatures of functional motions in lysozyme, which offer new insight into the catalytic cycle of this well‐studied enzyme and validate the use of temperature‐jump as a universal perturbation method for time‐resolved crystallographic studies of protein dynamics.

8. Enzymology: 7. Dynamics and allostery

ABS281

A COMMUNITY BASED CURE PROJECT TO EXPLORE STRUCTURE‐FUNCTION RELATIONSHIPS IN MALATE DEHYDROGENASE

Jessica Bell ¹, Joseph Provost¹, Ellis Bell¹

¹University of San Diego (San Diego, United States)

The Malate Dehydrogenase CUREs Community (MCC) project, funded by a grant from the National Science Foundation, involves proteincentric, Course‐based Undergraduate Research Experiences, (CUREs), focusing on Malate Dehydrogenase. These are being conducted in introductory to advanced level courses in diverse institutions, ranging from Community Colleges to Research intensive institutions. MCC has developed, and made available, a variety of written and physical resources to facilitate incorporation of these CUREs into the curriculum. Research projects integrate foundational concepts of protein structure and function with basic research into a variety of aspects of malate dehydrogenase, organized around three cluster themes, protein conformation, cellular biochemistry, & mechanism. The projects frequently extend beyond the classroom into longer term, student‐centered research projects, and have led to presentations at national meetings and publications. Recent projects, combining bioinformatics, computational approaches, and wet lab, have elucidated a mechanism for subunit interactions, explored the role of the flexible loop involved in substrate and effector binding, probed the formation of metabolons with Citrate Synthase, and investigated the role of second sphere residues around the active site. All CUREs implemented by MCC contain the same elements: Scientific Background, Hypothesis Development, Proposal, Experiments/Teamwork to test hypothesis, Data Analysis and Conclusions, and Presentation. Several of the projects include collaboration between institutions. In addition to the scientific questions being addressed, MCC is exploring several pedagogical questions concerning elements that make CUREs effective high impact teaching practices, including duration (whole semester versus 6‐7 week mini CUREs) and scientific collaboration. NSF ‐MCB‐0448905 & NSF‐1726932 EHR‐IUSE

7. Dynamics and allostery: 16. Protein interactions and assemblies

ABS282

THE ROLE OF DYNAMICS IN TRANSCRIPTION FACTOR DNA‐BINDING SPECIFICITY

Karlton Scheu ¹, Soymya De², Lawrence McIntosh¹

¹University of British Columbia (Burnaby, Canada); ²Indian Institute of Technology (Kharagpur, India)

A biological specificity conundrum challenges our understanding of the DNA‐binding mechanisms exhibited by many transcription factor families. Although sharing a highly conserved DNA‐binding domain that recognizes a consensus DNA sequence, individual members often regulate the transcription of a distinct set of target genes. This has been corroborated with an increasing number of genome‐wide DNA‐protein interaction studies showing that static X‐ray crystallographic structures of isolated DNA‐binding domains with consensus DNA oligonucleotides do not fully explain the different in vivo specificities exhibited by members of the same transcription factor families. This scenario is well‐exemplified by the ETS family with 28 human paralogs; here different family members can redundantly regulate the expression of common housekeeping genes. However, one or a very few highly related ETS family members are also able to bind non‐consensus DNA sequences to control the expression of genes with more tissue specific functions. Although many factors dictate transcription factor function, we propose that the dynamic properties of individual family members modulate their specificities for binding a continuum of DNA sequences. To test this hypothesis, I have introduced cavity forming mutations into the ETS domain of ETV6 that are distal from its DNA‐binding interface. These mutations increase the msec‐usec time scale motions of the protein as evidenced by relaxation dispersion NMR spectroscopy. I am now implementing an unbiased Bind‐N‐Seq approach with Next‐Gen sequencing to examine how the mutations alter relative affinities of ETV6 for DNAs ranging from those with a consensus ETS binding site to those with completely "non‐specific" sequences.

16. Protein interactions and assemblies: 12. Membrane proteins

ABS283

ELECTROSTATICS GOVERN MEMBRANE INTERACTIONS OF THE HSV‐1 NUCLEAR EGRESS COMPLEX

Mike Thorsen ¹, David Hoogerheide², Janna Bigalke³, Elizabeth Draganova³, Ekaterina Heldwein³

¹Tufts Sackler School (Boston, United States); ²National Institute of Standards and Technology (Gaithersburg, United States); ³Tufts University School of Medicine (Boston, United States)

During replication of herpesviruses, capsids escape from the nucleus into the cytoplasm by an unusual mechanism where they bud at the inner nuclear membrane to form enveloped capsids that subsequently fuse with the outer nuclear membrane. This process requires the virally encoded nuclear egress complex (NEC), a heterodimer of proteins UL31 and UL34. Using in‐vitro model systems, our lab previously discovered that the NEC is a complete budding machine by demonstrating that purified recombinant NEC vesiculates synthetic membranes in the absence of any other factors. What remained unclear, however, is how the NEC generates negative membrane curvature a prerequisite for budding. Proteins generate positive membrane curvature by either scaffolding or membrane insertion. Using neutron reflectometry, we found that the NEC does not penetrate the lipid bilayer beyond the polar headgroups and thus likely acts as a scaffold, which is consistent with its ability to oligomerize into hexagonal arrays on membranes. We also found that the in‐vitro budding activity requires electrostatic interactions between the NEC and the membranes. We narrowed the membrane‐interacting region down to a 10‐residue stretch within UL31 that contains two important di‐basic motifs at the N and C termini, respectively. A single di‐basic motif at either terminus was insufficient for budding; however, when relocated to the middle, it enabled near wild‐type budding. Our data suggest that the NEC generates negative curvature by scaffolding the membrane and that both the size and the location of the positive charge within the membrane‐interacting regions are essential for NEC‐mediated membrane budding.

21. Structure (x‐ray/NMR/EM): 25. Transcription/translation/post‐translational modifications

ABS284

STRUCTURE OF TRP REPRESSOR AND ITS COMPLEXES FROM FRANCISELLA TULARENSIS SHOWS PRESERVATION OF KEY WATER MOLECULE

Youngchang Kim¹, Andrzej Joachimiak ¹, Natalia Maltseva²

¹Argonne National Laboratory (Argonne, United States); ²Structural Genomics of Infectious Diseases, the University of Chicago (Chicago, United States)

The hydration pattern of the trp repressor/operator complex from Escherichia coli, showed that fixed water sites mediate important interactions between the repressor and the bases that specify the operators identity. We report determination of high‐resolution structures of trp repressor and its complexes from Francisella tularensis including: aporepressor, complexes with L‐Trp (activator), indole propionic acid (IPA) (inhibitor) and with specific trp operator. Structure analysis shows conservation of ligand binding and specific repressor‐DNA interactions, including solvent mediated and suggests that the majority of these interactions may be preserved across trp repressor family. The trp operator of F. tularensis shows several base substitutions but preserves the key A N7‐bound water molecule that mediates contact with the repressor. This interaction involves base pairs that specify the identity elements of the operator, suggesting that the repressor does, indeed, recognize the hydrated N7 as part of the operators discriminatory binding surface. This water is also preserved in the complex of repressor‐operator‐IPA. The uncomplexed repressor and aporepressor have also a firmly fixed water molecule at this position, therefore being a part of the repressor hydration shell. Stably bound water molecules are likely underrepresented in models of the interface of many specific biological complexes either because of spatial disorder in the crystal or inadequate resolution. The trp repressor/operator complex is an example of water‐mediated specific interactions and it highlights the important role of water molecules at fixed sites in the chemistry of molecular recognition.

Funding: NIH NIAID contracts HHSN272201200026C, HHSN272201700060C and DOE BER contract DE‐AC02‐06CH11357.

16. Protein interactions and assemblies: 20. Single molecule studies

ABS285

DESIGNING FRET BASED ASSAYS TO STUDY THE BINDING OF FIBROBLAST GROWTH FACTOR TO ITS RECEPTOR

Mamello Mohale ¹, ASHLEY HOWARD², MUSAAB HABEEB ALI AL‐AMMEER², RAVI KUMAR GUNDAMPATI², T.K.S KUMAR², COLIN HEYES²

¹University of Arkansas (Fayetteville, United States); ²University of Arkansas, Department of Biochemistry and Chemistry, Fayetteville 72703, AR, USA (Fayetteville, United States)

Fibroblast Growth Factor (FGF) binding to its receptor (FGFR) plays a critical role in triggering physiological processes during angiogenesis. New blood vessels are formed by signaling pathways such as MAPK following the interaction of FGF/FGFR. However, when the signaling process is not at homeostasis, excessive proliferation of cells can occur and lead to cancer. Therefore, there is a need to understand the binding mechanism of FGF/FGFR to aid the development of cancer therapies. Fluorescence Resonance Energy Transfer (FRET) has proven to be a powerful technique for investigating kinetic, thermodynamic and structural properties underlying protein interactions. Furthermore, FRET can be performed at the single molecule level, which provides mechanistic heterogeneity information. The distant dependence of FRET (1‐10nm) means that it relies on site‐specific fluorescence labelling of the proteins in order to quantitatively interpret the data. It is imperative for the fluorescent dye to maintain the structural integrity of the proteins and exhibit high labelling efficiencies and quantum yields. The main objective of this study is to investigate which dye labels and protein sites meet these criteria. We have added cysteine resides to several mutants of FGF and FGFR using site‐directed mutagenesis and labelled them with a range of maleimide‐functionalized fluorescent dyes. Our results indicated that Alexa dyes had high labelling efficiencies but low quantum yield although they did maintain protein structural integrity. Cyanine and iFluor dyes showed moderate to high labelling efficiencies and quantum yields. FRET studies will be performed on the successfully labeled proteins to determine their binding parameters.

12. Membrane proteins: 4. Chemical biology

ABS286

THE UNIQUE AMINO ACID COMPOSITION OF THE CHROMOPHORE‐BINDING POCKET CONTRIBUTES TO THE RETINAL BINDING SPECIFICITY IN HUMAN CONE OPSINS

Beata Jastrzebska ¹, Joseph Ortega¹, Kota Katayama², Sahil Gulati³, Krzysztof Pakczewski³

¹Case Western Reserve University (Cleveland, United States); ²Case Wester Reserve University (Cleveland, United States); ³University of California‐Irvine (Irvine, United States)

Both rod and cone vertebrate photoreceptors utilize common 11‐cis‐retinal chromophore to capture light photons, which then isomerizes to its all‐trans configuration to initiate phototransduction. Despite this similarity these receptors possess distinct photochemical properties associated with differences in the amino acid composition of the retinal‐binding pocket. In contrast to rhodopsin (Rh), the crystal structures of cone opsins remain to be solved. Thus, to understand the differences in the photochemistry of vision between rod and cone opsins we applied the binding analyses of 11‐cis‐6‐membered‐retinal (11‐cis‐6‐mr‐retinal) analog to human blue, green, and red cone opsins and compared to rod opsin. Cone opsins were expressed in Sf9 cells and rod opsin was isolated from bovine retinas and incubated with 11‐cis‐6‐mr‐retinal, followed by their immunoafinity purification. UV‐visible absorbance spectra of purified cone opsins revealed that among the three cone opsins only blue cone opsin, likewise rod opsin could accommodate the 11‐cis‐6mr‐retinal in its chromophore‐binding pocket, resulting in the formation of a synthetic blue pigment (B6mr). By using a combination of primary sequence alignment, molecular modeling, and mutagenesis experiments we found the specific amino acid residue 6.48 (Tyr262 in blue cone opsins and Trp281 in green and red cone opsins) as a chromophore binding selectivity filter in human cone opsins. Together, these studies revealed that the unique amino acid composition of the chromophore‐binding pocket contributes to the molecular mechanisms of the cone opsins regeneration with 11‐cis‐6‐mr‐retinal.

21. Structure (x‐ray/NMR/EM): 26. Other

ABS287

THE STRUCTURE OF A HIGHLY CONSERVED PICOCYANOBACTERIAL PROTEIN REVEALS A TUDOR DOMAIN WITH A NOVEL TRNA BINDING FUNCTION

Katherine Bauer ¹, Rose Dicovitsky², Maria Pellegrini³, Olga Zhaxybayeva⁴, Michael Ragusa⁵

¹Department of Biochemistry & Cell Biology (Hanover, United States); ²Department of Chemistry, Dartmouth College (Hanover, United States); ³Department of Chemistry, Dartmouth College (Hanover, United States); ⁴Departments of Biological Sciences & Computer Science, Dartmouth College (Hanover, United States); ⁵Departments of Biochemistry & Cell Biology and Chemistry (Hanover, United States)

Cyanobacteria from the genera Prochlorococcus and Synechococcus are the most abundant microbes in the global ocean. Due to their numbers, these bacteria play an important role in the global carbon cycle. While both Prochlorococcus and Syenchococcus have small genomes, the gene pool diversity is large within the species. Prochlorococcus alone is estimated to contain 80,000 genes within its gene pool. With such extreme gene diversity, it is striking to find a 62 amino acid protein with almost 100% sequence conservation across all sampled genomes. This protein, Prochlorococcus/Synechococcus Hyper Conserved Protein (PSHCP), has an unknown function, though its conservation suggests that this function is essential. The objective of this study was to gain insight into the function of PSHCP. Using Nuclear Magnetic Resonance spectroscopy (NMR), we determined that PSHCP adopts a Tudor domain fold. Surprisingly, the protein lacks the canonical residues necessary to form an aromatic cage, making it unlikely that PSHCP binds methylated proteins. Another, less documented, function of Tudor domains is binding nucleic acids. In addition, the PSHCP gene is flanked both 5 and 3 by genes encoding transfer RNA (tRNA). As genes in bacteria are often clustered based on their part in similar pathways, we hypothesized that PSHCP interacts with tRNA. Using both NMR and Isothermal Titration Calorimetry (ITC), we show that PSHCP does bind tRNA with a low micromolar affinity. This study provides evidence for a novel Tudor domain function, that of tRNA binding.

21. Structure (x‐ray/NMR/EM): 8. Enzymology

ABS288

STRUCTURAL CHARACTERIZATION OF ACINETOBACTER‐DERIVED CEPHALOSPORINASE‐7 IN COMPLEX WITH CEFTAZIDIME AND ITS TRANSITION STATE ANALOG

Brandy Curtis ¹, Emilia Caselli², Magdalena Taracila², Robert Bonomo³, Fabio Prati², Rachel Powers¹, Brad Wallar¹

¹Department of Chemistry, Grand Valley State University (Allendale, United States); ²Department of Life Sciences, University of Modena and Reggio Emilia (Modena, Italy); ³Research Service, Louis Stokes Cleveland Department of Veterans Affairs Medical Center; Departments of Medicine, Pharmacology, Molecular Biology and Microbiology, Case Western Reserve University School of Medicine (Cleveland, United States)

One of the most prevalent mechanisms of resistance in multidrug resistant bacteria is the production of ‐lactamase enzymes that inactivate ‐lactam antibiotics. The third generation cephalosporins are widely used ‐lactams designed to be resistant to ‐lactamase hydrolysis. However, extended‐spectrum enzymes have evolved to rapidly inactivate this particular class of antibiotics. To better understand the turnover of third generation cephalosporins by ‐lactamases, we determined the X‐ray crystal structure of ADC‐7, the class C cephalosporinase from A. baumannii, in an acyl‐enzyme complex with the third‐generation cephalosporin ceftazidime (2.40 å), as well as in complex with a boronic acid transition state analog inhibitor that contains the R1 side chain of ceftazidime (1.67 å). In the acyl‐enzyme complex, the carbonyl oxygen is situated in the oxyanion hole where it makes key stabilizing interactions with the main chain nitrogens of S64 and S315. The boronic acid O1 hydroxyl group is similarly positioned in this area. Additionally, conserved residues N152 and Q120 hydrogen bond with the amide group of the R1 side chain in both complexes. Comparison of these structures with related class C ‐lactamase complexes provides insight into possible reasons for the slow turnover of ceftazidime. These complexes provide the first snapshots of two steps in the hydrolysis of third‐generation cephalosporins by ADC‐7 and offer insight into the catalytic mechanism of ADC‐7. In addition, the transition state analog inhibitor shows a potential interaction to R340, a residue unique to ADC enzymes, offering the potential for optimization efforts of this novel class of inhibitors.

21. Structure (x‐ray/NMR/EM): 8. Enzymology

ABS289

CHARACTERIZATION OF NOVEL TRIAZOLE‐CONTAINING BORONIC ACID TRANSITION STATE INHIBITORS (BATSIS) OF ACINETOBACTER‐DERIVED CEPHALOSPORINASE (ADC‐7)

Erin Fish¹, Erin Fish ¹, Emilia Caselli², Magdalena Taracila², Robert Bonomo³, Fabio Prati², Rachel Powers¹, Brad Wallar¹

¹Department of Chemistry at Grand Valley State University (Allendale, United States); ²Department of Life Sciences at University of Modena and Reggio Emilia (Modena, Italy); ³Louis Stokes Cleveland Department of Veterans Affairs Medical Center; Departments of Medicine, Pharmacology, Molecular Biology and Microbiology, Case Western Reserve University School of Medicine (Cleveland, United States)

Much of the resistance to ‐lactam antibiotics in the multidrug resistant bacterium Acinetobacter baumannii is attributed to its production of ‐lactamase enzymes that deactivate ‐lactams by hydrolyzing the amide of the defining lactam ring. In order to develop more effective treatment against multidrug resistant pathogens like Acinetobacter baumannii, we have characterized a series of six novel Boronic Acid Transition State Inhibitors (BATSIs) for activity against the class C ‐lactamase, Acinetobacter‐derived cephalosporinase‐7 (ADC‐7). This series of BATSIs is characterized by the presence of a triazole ring that allows for easier synthesis and modification than previously studied BATSIs. The X‐ray crystal structures of ADC‐7 in complexes with each of these inhibitors were determined to resolutions ranging from 1.7 to 2.0 å. In all six complexes, hydrogen bonding is observed between the triazole and conserved residues Q120 and N152, suggesting that the triazole may act as a mimic of the R1 amide group common to ‐lactams. Beyond the triazole group, several of the BATSIs participate in unique interactions with ADC‐7 and help to explain the differences in binding affinities (Ki ~ 0.09 1.75 μM). Overall, our structure‐function analysis suggests the triazole BATSIs provide a convenient and intriguing template for further modification to increase their binding affinity for ADC‐7.

ABS290

DISCOVERY AND CHARACTERIZATION OF A PAP248‐286/LIPID CO‐ASSEMBLY: THE MESSICLE STORY

Eleanor W Vane ¹, Abhinav Nath¹

¹University of Washington Department of Medicinal Chemistry (Seattle, United States)

Protein/lipid co‐assemblies are understudied, yet important, to the function of antimicrobial peptides as well as the pathological effects of amyloid. We have chosen to study the formation of such aggregates with an amyloidogenic peptide fragment of prostatic acid phosphatase, PAP248‐286. PAP248‐286 was identified as an important component of seminal fluid that forms amyloid, termed semen‐derived enhancer of viral infection (SEVI), that enhances the viral infectivity of HIV. SEVI amyloid also displays antimicrobial activities through a bacterial agglutination mechanism. Using a variety of biophysical techniques, including microscopy, fluorescence, and amyloid kinetic studies, we show that in addition to forming amyloid, PAP248‐286 also has the ability to assemble with lipid vesicles into large, yet distinct, amyloid‐like peptide/lipid co‐aggregates. Formation of this co‐aggregate, which we have termed messicle, is controlled by the peptide:lipid (P:L) ratio, as well as lipid composition. Due to the relative stability of messicles to changes in P:L ratio upon formation, we propose that messicles are formed by polyvalent PAP248‐286 forming multiple, weak interactions with itself and the lipid vesicles, bridging and enmeshing them. Formation is therefore dependent on the binding affinity of PAP248‐286 for the membrane. We believe it is possible that some or all of the biological activities assigned to SEVI amyloid could be attributed to this newly uncovered state of PAP248‐286 in a peptide/lipid co‐aggregate. More broadly speaking, this work also provides a potential framework for finding and characterizing the formation of peptide/lipid co‐aggregates by other amyloid‐forming proteins and antimicrobial peptides.

16. Protein interactions and assemblies: 10. Folding

ABS291

TWO CALCIUM SENSORS, ONE TARGET: PRP40 INTERACTIONS WITH BOTH CALMODULIN AND CENTRIN

Adalberto Diaz‐Casas ¹, Walter Chazin²

¹Vanderbilt University (Nashville, United States); ²Vanderbilt University (Nashville, United States)

Prp40 is a splicing factor that potentially couples defects in huntingtin (htt) protein to defective RNA splicing leading to the pathological polyglutamine repeat associated with Huntingtons Disease. Human Prp40 homologs A and B contain two WW domains and six FF (FF1‐FF6) domains, and both are known to engage the htt scaffold. Whereas interactions via the WW domain have been mapped and characterized, little is known about the roles of the FF domain‐mediated interactions. Based on data from a yeast two‐hybrid screen, we investigated the interaction of Prp40A with the EF‐hand calcium sensor centrin 2 (Cen2). Isothermal titration calorimetry (ITC) and two‐dimensional infrared correlation spectroscopy were used to show that Cen2 binds to a consensus motif in the FF3 domain of human Prp40A. Remarkably, a different group showed that yeast Prp40 interacts with another EF‐hand calcium sensor, calmodulin (CaM). CaM has also been shown to interact with htt. Here we progress report on our efforts to elucidate the molecular details of the regulatory mechanism of Prp40 activity. The affinity of the Prp40A FF3 domain for Cen2 was measured at 25 °C and 37 °C using ITC; tighter binding was observed at the higher temperature. While initially proposed to interact with a WW domain, we found that CaM binds to two different hPrp40A FF domains. ITC was used to show that the hPrp40A FF1 domain binds significantly more tightly to CaM than the FF3 domain. Together, these results will provide new insights to the interplay between Cen2 and CaM‐mediated calcium signal transduction.

1. Amyloid and aggregation: 11. Intrinsically disordered proteins

ABS292

DESIGN OF DUAL‐ACTION LIPID‐NANODISCS IN CONTROLLING AMYLIN AGGREGATION INVOLVED IN TYPE‐2 DIABETES

BIKASH SAHOO ¹, Takuya Genjo¹, Takahiro Watanabe‐Nakayama², Toshio Ando², Ayyalusamy Ramamoorthy¹

¹Biophysics Program, Department of Chemistry, University of Michigan, Ann Arbor, MI 48109 USA (ANN ARBOR, United States); ²Bio‐AFM Frontier Research Center, Kanazawa University, Kanazawa 920‐1192, Japan (Kanazawa, Japan)

The human islet‐amyloid polypeptide (hIAPP) comprising of 37‐residues is a histopathological hallmark of type 2 diabetes (T2D). Biological membrane plays a central role in catalyzing its aggregation and severity of pathological conditions. Here, we demonstrate the bimodal application of lipid‐nanodiscs in studying the hIAPP aggregation pathway and controlling its amyloid aggregation. Our results show that apolipoprotein‐A‐I mimetic (4F) peptide‐encapsulated‐nanodiscs (~10 nm) provides real‐time monitoring of membrane‐dependent hIAPP interaction and aggregation using fluorescence, NMR and molecular dynamics (MD) simulations. In contrast, a cationic polymer‐encapsulated (PMAQA)‐lipid‐nanodisc (~10 nm) showed the opposite action. As an example, DMPC/DMPG (7:3) 4F‐nanodiscs substantially accelerate hIAPP aggregation, whereas PMAQA‐nanodiscs inhibits hIAPP aggregation. An interesting discovery is that the lipid composition and concentration could be tuned to control hIAPPs aggregation kinetics only in PMAQA‐nanodiscs. High‐speed AFM was used to monitor the inhibition of hIAPP aggregation in real‐time by PMAQA‐nanodiscs and membrane thinning was ascertained. NMR and CD revealed a partially folded helical hIAPP trapped in the PMAQA‐nanodiscs. Interaction analysis of hIAPP with free 4F‐peptide or PMAQA highlighted a potent inhibitory action by the latter, while the former does not affect its aggregation. The morphological and pathological phenotypes of lipid‐nanodiscs associated IAPP species were revealed using TEM and in‐vitro cell assay. We further developed the parameterization of PMAQA in Martinii‐force‐field and designed PMAQA‐ and 4F‐nanodiscs to provide mechanistic insights into the bimodal activities in multi‐microseconds time‐scale at the atomic‐level. The lipid‐nanodiscs would also aid in the design of new multifunctional materials for future studies of amyloid‐related diseases.

16. Protein interactions and assemblies: 21. Structure (x‐ray/NMR/EM)

ABS293

FUNCTIONAL AND STRUCTURAL STUDY ON HERC4

Young Sun Lee ¹, Donald Spratt¹

¹Clark University (Worcester, United States)

HECT domain and RCC1‐like domain‐containing protein 4 (HERC4) is a member of the Homologous to E6AP C‐terminus (HECT) E3 ubiquitin ligase family that covalently attaches ubiquitin to its substrates for proteasomal degradation. Previous studies have shown that HERC4 builds a polyubiquitin chain on to C‐terminus of Salvador (Sav), a protein involved for organ size control, and regulates the expression and activity of Sav. HERC4 is also overexpressed in lung, breast, liver, and cervical cancer and plays a critical role in metastasis, migration, and proliferation of cancer. In mice, the absence of HERC4 reduced male fertility by creating immature sperm with angular flagella. The goal of this study was to characterize the protein to discover the way ubiquitin binds and interacts with HERC4. The wildtype and mutant (H1023A, T1024V, T1024S, C1025A) C‐terminus of HERC4 (943‐1057 residues), the region with catalytical function, were overexpressed in E. coli and purified. Circular dichroism spectra show that wildtype and substituted HERC4 proteins are mainly ‐helical secondary structure and stable at high temperatures. Ubiquitin activity assays of wildtype and substituted proteins, with the exception of C1025A (catalytic cysteine) were able to form a thioester bond with ubiquitin. Analytical gel filtration analysis demonstrated that HERC4 wildtype and mutants exist as monomers. 2D and 3D nuclear magnetic resonance (NMR) spectroscopy of wildtype HERC4 have been collected in preparation for structural determination. This study provides the first structural and biochemical insights into HERC4‐dependent ubiquitylation and will help us learn how HERC4s role in cancer metastasis and spermatogenesis.

1. Amyloid and aggregation: 24. Therapeutics and antibodies

ABS294

SUMO‐DERIVED PEPTIDES AS INHIBITORS OF ‐SYNUCLEIN AGGREGATION

Zhaohui LIANG ¹, Junqing YANG¹, Ho Yin Edwin CHAN¹, Ming Ming Marianne LEE¹, Michael Kenneth CHAN¹

¹The Chinese university of Hong Kong (HONG KONG, China)

The misfolding and aggregation of ‐synuclein protein results in the formation of amyloid fibrils in brain, and in turn synucleinopathies that include the second most common neurodegenerative disease ‐ Parkinsons disease (PD). Previous studies have shown that sumoylation of ‐synuclein occurs naturally in the brain, and this sumoylated ‐synuclein can suppress ‐synuclein aggregation. Here we first report that variants of small ubiquitin‐like modifier 1 (SUMO1) protein can directly suppress ‐synuclein aggregation in vitro with an efficiency that appears comparable to that reported for sumoylated‐‐synuclein.

After systematically preparing and testing a series of SUMO1 variants, we developed a minimal peptide based on the core region of SUMO1 as a potential therapeutic lead for in vivo studies. This peptide exhibited improved suppression activity relative to SUMO1 itself in in vitro studies. Also, the peptide reduced cytotoxicity and blocked the uptake of aggregated ‐synuclein in SH‐SY5Y cells. Larval feeding of the peptide significantly ameliorated retinal degeneration and the disease symptoms in Drosophila PD models by suppressing the loss of dopaminergic neurons. To elucidate the binding region of this interaction, we have employed crosslinking reaction followed by Nano liquid chromatography‐Fourier transform ion cyclotron resonance mass spectrometry (NanoLC‐FTICR‐MS) and microscale thermophoresis to test the predicted regions. These findings suggest a new avenue for developing treatments for PD and/or other synucleinopathies.

15. Peptides: 9. Evolution

ABS295

NOVEL PORE‐FORMING PEPTIDES ASSEMBLING IN LIPOSOME MEMBRANES SELECTED BY COMBINING CDNA DISPLAY METHOD WITH CELL SORTER SYSTEM

Naoto Nemoto ¹, Toshiki Miyajima¹, Takeru Yoshinobu¹, Yusuke Sekiya², Ryuji Kawano²

¹Saitama University (Saitama, Japan); ²Tokyo University of Agriculture and Technology (Tokyo, Japan)

In the previous study, we have selected some liposome‐binding peptides by in vitro selection method using cDNA display [1]. cDNA display method is a genotype‐phenotype linking method with a cell‐free translation system by fusing a cDNA with its coding polypeptide via a puromycin molecule (i.e. an antibiotic) covalently. It is also more stable and robust than mRNA display and ribosome display, because cell‐based selections can be performed with this technology just like phage display [2]. In this study, we performed an in vitro selection of pore‐forming peptides which act functionally like antimicrobial peptides in liposome membranes, by using cDNA display with a Fluorescence Activated Cell SortingFACSsystem. Interestingly, the selected peptides form channels with ranging from 2 to 5 nm diameter in the liposome membrane by analyzing high‐throughput lipid bilayer system [3].

[1] Kobayashi, S.; Terai, T.; Yoshikawa, Y.; Ohkawa, R.; Ebihara, M.; Hayashi, M.; Takiguchi, K.; Nemoto, N, (2017) Chem Commun, 63, 3458‐3461.

[2] Ueno, S.; Yoshida, S.; Mondal, A.; Nishina, K.; Koyama, K.; Sakata, I.; Miura, K.; Hayashi, Y.; Nemoto, N.; Nishigaki, K.; Sakai, T. (2012) Proc Natl Acad Sci U S A. 109, 11121‐11126.

[3] Watanabe, H.; Gubbiotti, A.; Chinappi, M.; Takai, N.; Tanaka, K.; Tsumoto, T.; Kawano, R. (2017) Anal. Chem. 89, 11269‐11277.

16. Protein interactions and assemblies: 17. Proteins in cells

ABS296

EXPRESSION AND CHARACTERIZATION OF THE DROSOPHILA MELANOGASTER (DM)IKK: COMPLEX

Samantha Cohen ¹, Sheri Wu¹, Tom Huxford¹

¹San Diego State University (San Diego, United States)

In Drosophila, the IMD pathway is indispensable for proper innate immune responses. Infection by gram‐negative bacteria elicits a signaling cascade culminating in the rapid induction of antimicrobial peptide gene expression by the NF‐B transcription factor Relish. Signal‐dependent activation of Relish is an essential component in the IMD pathway and is regulated by the Drosophila melanogaster IB Kinase (DmIKK) complex. DmIKK is composed of two subunits: the catalytic subunit IKK and non‐catalytic subunit IKK. Like many protein kinases, DmIKK must be activated for proper catalysis. Although there has been extensive research into the signaling pathways of the Drosophila innate immune response, little is known about the molecular mechanisms that lead to DmIKK activation. Here, we report the optimization of large‐scale expression and purification of the DmIKK complex. Utilizing the baculovirus expression system in insect cells, we have isolated highly‐pure recombinant DmIKK complex (~2 mg per 100 mL of cultured Sf9 cells) for structural and biophysical studies aimed at understanding its mechanism of activation.

20. Single molecule studies: 16. Protein interactions and assemblies

ABS298

ENTEROPATHOGENIC E. COLI HIJACKS PROGRAMMED HOST‐CELL DEATH PATHWAYS BY INTERFERING WITH THE HIGHER ORDER OLIGOMERIZATION OF IMMUNE SYSTEM PROTEINS

Yann GAMBIN ¹, Ana Monserrat‐Martinez², Emma SIERECKI²

¹EMBL Australia (Randwick, Australia); ²EMBL Australia (Randwick, Australia)

During an infection, pathogens operate by hijacking crucial intracellular pathways within their hosts. In the case of Enteropathogenic Escherichia coli (EPEC) infection, bacterial proteins interact with host proteins to subvert programmed cell death pathways. Recently, it has been found that many proteins from the immune system polymerise to transduce these death signals, forming higher order assembly structures.

Here we used single‐molecule techniques to study how EPEC effector proteins interfere with the higher order assembly of immune proteins involved in signal transduction in inflammation, apoptosis and cell survival pathways.

A selection of immune system proteins involved in signal transduction in different cell pathways was chosen to perform an in vitro screen against EPEC effector proteins by confocal spectroscopy. It was observed that espF (a bacterial effector protein) was able to change TRAF2 oligomerization/aggregation propensity. TRAF2 is a cytosolic protein involved in signal transduction in the cell death pathway. In case of infection or stress, the protein forms higher order assembly structures that eventually lead the cell to apoptosis. Further experiments performed in HeLa cells transfected with TRAF2 showed an increase in these structures in stressed cells when compared with the control group. When espF and TRAF2 were co‐transfected, co‐localization of both proteins was observed in distinct punctate intracellular structures reminiscent of higher order assemblies.

Understanding precisely this host‐pathogen interaction could give us information useful for future drug development against EPEC infection.

1. Amyloid and aggregation: 20. Single molecule studies

ABS299

MODULATION OF INTERACTOME BY PROTEIN SELF‐ASSEMBLY: THE CASE OF ALPHA‐SYNUCLEIN

Emma Sierecki ¹, Andre Leitao¹, James Brown¹, Alex Chappard¹, Yann Gambin¹

¹UNSW (Randwick, Australia)

Protein‐protein interactions modulate protein function and activity. In advanced proteomics analysis, the identify of binding partners is used to infer biological functions and enable classifications, for example to signalling pathways. However, many proteins also oligomerise and self‐assembly into higher‐order structures, which in turn has major impacts on function and the repertoire of interaction partners. What is less well understood is how protein‐protein interactions impact on protein self‐assembly. Because binding partners alter valency, the number of binding sites, affinity, even transient interactions can accelerate and inhibit protein self‐assembly or results in different higher‐order structures.Overall, a co‐dependency of protein self‐assembly and protein interactions should exist.

‐Synuclein is an intrinsically disordered protein (IDP) that has been proposed to adopt different conformations upon binding to different partners. However, validated binders are rare and there is no structure of ‐Synuclein bound to a protein partner. Using a proximity assay (AlphaScreen) and single molecule experiments, we systematically acquired an extensive dataset of interactions with 70 proteins from the Lewy bodies. This shows that despite its small size and absence of structure, ‐Synuclein binds specifically to different partners, and that interactions are very different for the different species formed along the aggregation pathway. Our data show a clear selectivity and specificity for different ‐Synuclein species (monomer, oligomer or amyloid form) and highlight different biological functions for the different species. We also identified protein partners that affect ‐Synuclein's aggregation pathway.

21. Structure (x‐ray/NMR/EM): 16. Protein interactions and assemblies

ABS300

CRYSTAL STRUCTURES OF SNX11 REVEAL THE MEMBRANE BINDING MECHANISM

Xu Tingting ¹, Liu Jinsong¹

¹Guangzhou Institutes of Biomedicine and Health, CAS (Guangzhou, China)

Sorting Nexin (SNX) protein are involved in the cargo transporting, endosome homeostasis, protein degradation etc. This family member proteins can specifically target to the membrane enriched with phosphorylated‐phosphatidylinositols (PIPs) through their PX domain. Therefore, the PX domain is critical for the function of SNXs. Up to date, only two structures reported the recognition details of PX bound to PI3P, whereas many studies show that SNX can bind to di‐ or triphosphorylated‐ PIs. Our previous studies revealed that SNX11 can bind to both PI3P and PI35P2. In addition, SNX11 possesses a PX‐extended domain with two more helices at the C terminus of classical PX domain, which is required for its function. In this study, we obtain additional structures of SNX11‐PX, and reveal conformational change that will shed light on the membrane binding mechanism.

8. Enzymology: 4. Chemical biology

ABS301

CATALYTIC MECHANISM OF TIAS5 IN TIACUMICIN B BIOSYNTHESIS PATHWAY

Yongzhi Lu ¹, Yan Dong¹, Mingze Sun¹, Jinsong Liu¹

¹Guangzhou Institutes of Biomedicine and Health, Chinese Academy Sciences (Guangzhou, China)

Tiacumicin B is a narrow‐spectrum antibiotic that shows activities against various Gram‐positive pathogenic bacteria and breast cancer cells. In 2011, Tiacumicin B was approved for the treatment of Clostridium difficile infection (CDI), which has become a serious disease among hospitalizations. Methylation in Tiacumicin B biosynthesis, catalyzed by TiaS5, evidently improves the activity of Tiacumicin B. Till now, the catalytic mechanisms of TiaS5 is unclear. In order to investigate the catalytic mechanisms of TiaS5, we determined the apo and SAH bound structure of TiaS5. Structural analysis reveals that the dimerization is prerequisite for macrocyclic substrate binding and methyltransfer activity. The macrocyclic substrate binding pocket is consisted of C‐terminus of one monomer and the Rossmann domain of the other monomer.

21. Structure (x‐ray/NMR/EM): 4. Chemical biology

ABS302

STRUCTURE OF THE TRIMETHYLAMINE METHYLTRANSFERASE REVEALS A DISTINCT ENVIRONMENT FOR METHYLAMINE ACTIVATION BY PYRROLYSINE

Jiaxin Li ¹, Patrick T. Kang², Jodie Y. Lee³, Ruisheng Jiang³, Jitesh A. Soares³, Joseph A. Krzycki⁴, Michael K. Chan¹

¹School of Life Science and Center of Novel Biomaterials, The Chinese University of Hong Kong (Hong Kong, China); ²Ohio State University Biochemistry Program, The Ohio State University (Columbus, OH, United States); ³Department of Microbiology, The Ohio State University (Columbus, OH, United States); ⁴Ohio State University Biochemistry Program and Department of Microbiology, The Ohio State University (Columbus, OH, United States)

L‐pyrrolysine, the 22nd genetically‐encoded amino acid, plays a unique role in the methyl group activation of mono‐, di‐, and tri‐methylamines in methanogens. Crystal structures of the Methanosarcina barkeri monomethylamine methyltransferase (MtmB) first revealed pyrrolysine and key features of its role in methylamine binding and activation. Herein, we report the second structure of a pyrrolysine containing protein, the M. barkeri trimethylamine methyltransferase (MttB).

By utilizing X‐ray crystallography, we got the structure of both native and SO3‐bound MttB. As expected, MttB adopts a TIM barrel fold characteristic of substrate:corrinoid methyltransferase enzymes with the pyrroline ring of the pyrrolysine amino acid located at the center of the barrel, just like the MtmB. However, there are important differences. Consistent with the distinct location of the pyrrolysine amber codon in mttB and mtmB the pyrrolysine amino acid originates from opposite sides of the TIM barrel fold of the two proteins. The mode of activation of the imine bond of pyrrolysine in MttB for amine binding is also different involving hydrogen bonding from a water that in turn is in hydrogen bonding distance to an amide carbonyl and the hydroxyl group of tyrosine. Also unlike MtmB, only a single conformation is observed both for the native structure and for the sulfite complex. In summary, the structure of MttB provides the second example of a pyrrolysine‐containing protein and reveals distinct differences in the possible mechanism of protein‐assisted activation of the pyrroline ring, and in the conformational properties of the pyrrolysine ring during turnover.

8. Enzymology: 4. Chemical biology

ABS303

STRUCTURAL AND BIOCHEMICAL ANALYSIS OF THE HECT E3 UBIQUITIN LIGASE HECW2

Justine Bohl ¹, Donald Spratt¹

¹Clark University (Worcester, United States)

HECT, C2, and WW domain‐containing protein 2 (HECW2) is a member of the Homologous to E6AP C‐Terminus (HECT) E3 Ubiquitin ligase family of proteins involved in the ubiquitylation of substrates via the ubiquitin‐signaling pathway. By attaching ubiquitin to substrate molecules such as lamin A, lamin B1, and proliferating cell nuclear antigen (PCNA), HECW2 can regulate their expression and effect the formation of nuclear lamina. It has been suggested that HECW2 plays a crucial role in neurodegenerative diseases caused by lamina deformation by ubiquitylating lamin protein substrates to target them for proteasomal degradation. The C‐terminal lobe of the HECT domain of HECW2 (residues 1237‐1572), the region containing the catalytic cysteine (C1540) responsible for creating the thioester bond with ubiquitin, was overexpressed in E. coli and purified for biochemical analysis. Circular dichroism demonstrated that the protein is predominantly alpha‐helical and is thermally stable. Ubiquitin activity assays showed that specific residues in the C‐lobe of HECW2 are required for ubiquitin transfer and that the isolated C‐lobe can be charged with ubiquitin in an E2‐independent manner. Studies of substituted residues in the HECW2 C‐lobe show subtle structural and functional differences. This work provides a new insight into the catalytic mechanism of human HECW2 and how its dysfunction may be linked to laminopathies.

21. Structure (x‐ray/NMR/EM): 26. Other

ABS304

CRYSTAL STRUCTURE OF THE RED C‐TERMINAL DOMAIN IN COMPLEX WITH EXONUCLEASE REVEALS AN UNEXPECTED HOMOLOGY WITH ORF AND AN INTERACTION WITH ESCHERICHIA COLI SINGLE STRANDED DNA BINDING PROTEIN

Brian Caldwell ¹, Ekaterina Zakharova², Gabriel Filsinger³, Timothy Wannier⁴, Jordan Hempfling¹, Lee Chun‐Der⁵, Dehua Pei⁵, George Church⁴, Charles Bell²

¹Ohio State Biochemistry Program (Columbus, United States); ²Ohio State University department of Biochemistry and Pharmacology (Columbus, United States); ³Harvard Medical School department of Systems Biology (Cambridge, United States); ⁴Harvard Medical School department of Genetics (Cambridge, United States); ⁵Ohio State University department of Chemistry and Biochemistry (Columbus, United States)

Many bacteriophages encode a simple (synaptase‐exonuclease) SynExo recombination system for the repair of double‐stranded DNA (dsDNA) breaks by a single strand annealing (SSA) mechanism. The best studied of these is the Red system from bacteriophage, which consists of two proteins: a 5‐3 exonuclease, Exo, that binds to DNA ends and resects the 5 strand, and a synaptase, Red, that anneals the resulting 3 overhang to a complementary strand from another chromosome. Interest in Exo and Red stems in part from their distant homologies with the mammalian recombination proteins Dna2 and Rad52, respectively. The proteins have also been exploited in powerful methods for bacterial genome engineering, most notably recombineering and Multiplex Automated Genome Engineering (MAGE). Interestingly, Exo and Red bind to one another to form a complex that is required for a successful annealing reaction in vivo. The exact role of this complex is unknown, but it may serve to integrate the two steps of the SSA reaction by physically loading Red onto the 3 overhang as it is generated by Exo. Recently, our group has solved a crystal structure of a minimal complex that captures the essential features of the Exo ‐ Red interaction. Using this structure as a platform for rational mutagenesis, we have performed in vitro pull down and in vivo recombination assays to show that disrupting the interaction prevents SSA. Interestingly, these results have also shown that an interaction between single‐stranded DNA binding protein (SSB) and Red is also required for recombination.

8. Enzymology: 24. Therapeutics and antibodies

ABS305

THE NEGLECTED HIGH MOLECULAR WEIGHT ENZYMES OF SNAKE VENOM: CANDIDATE TARGETS FOR TREATING TISSUE NECROSIS BY SNAKEBITE ENVENOMING

I‐Jin Lin ¹, Chun‐Lin Long², Yue‐Hu Wang³, Wen‐guey Wu¹

¹Institute of Bioinformatics and Structural Biology, National Tsing Hua University, Hsinchu, Taiwan (Hsinchu, Taiwan); ²College of Life and Environmental Sciences, Minzu University of China (Beijing, China); ³Key Laboratory of Economic Plants and Biotechnology and the Yunnan Key Laboratory for Wild Plant Resources, Kunming Institute of Botany, Chinese Academy of Sciences (Kunming, China)

High molecular weight enzymes (HMWs) of snake venom have been neglected in the snakebite envenoming research, despite the fact that they are ubiquitous in most snake venom. According to the X‐ray crystal structures presented in this study, two HMWs from Taiwan cobra venom, phosphodiesterase (PDE) and 5‐nucleotidase (5NT), are structurally homologous with human ENPP1 and CD73, respectively. Hence, we propose the hypothesis that while cardiotoxins (CTXs) from cobra venom result in cytotoxicity, HMWs exacerbate the CTXs‐induced tissue necrosis through dysfunction of the immune system near the bitten area. Due to the lack of effective treatment for severe tissue necrosis by cobra bites, over 50 compounds from medicinal plants have been purified, characterized, and their inhibitory effects have been tested on CTXs, PDE and 5NT. In the hemolytical assay, eight polyphenols were observed to suppress CTX3 from Taiwan Cobra. Additionally, 15 flavonoid derivatives were found to inhibit activities of PDE and 5NT. The fluorescence experiment and the crystal complex structure provided evidence of direct binding between inhibitors and PDE. Furthermore, one chalcone was identified as the common inhibitor for both CTX3 and PDE, offering a promising candidate treatment for tissue necrosis by snakebite envenoming. Given that various herbal compounds have passed clinical trials and can be repurposed to treat snakebite envenoming, our strategy of applying effective small molecules to inhibit CTXs as well as HMWs provides an avenue for fighting the neglected tropical disease of snakebites.

8. Enzymology: 18. Proteomics

ABS306

STRUCTURAL AND MECHANISTIC CHARACTERIZATION OF HERC2 E3 UBIQUITIN LIGASE WITH IMPLICATIONS IN CANCER, PRADER‐WILLI SYNDROME, AND EYE COLOR

Kayla Rich ¹, Noah Schwaegerle¹, Donald Spratt¹

¹Clark University (Worcester, United States)

HERC2 (HECT domain containing and RCC1‐like domain containing E3 ubiquitin ligase 2) is a HECT E3 ubiquitin protein ligase that is implicated with breast cancer, Prader‐Willi syndrome, and eye color. HECT E3 ubiquitin ligases covalently attach ubiquitin onto substrates. Ubiquitin acts as a signal for various cellular processes such as proteasomal degradation, DNA damage repair, cell cycle and transcription regulation, and trafficking. HERC2 has proposed implications with DNA damage repair and proteasomal degradation. The HERC2 gene has been linked to the neurodevelopmental disorder Prader‐Willi syndrome and is part of the cis‐regulatory module for OCA2 (oculocutaneous albinism 2). Only a few substrates of HERC2 have been identified, one includes BRCA1 (breast cancer 1). Determining the structural and functional properties of HERC2 is important for implications in diseases. Utilizing biochemical and biophysical techniques, such as circular dichroism and NMR spectroscopy, the HERC2 C‐lobe structure was characterized. These studies have shown that HERC2 is mainly ‐helical with a flexible C‐terminal tail extension. Mechanistic activity assays have shown the C‐terminal tail extension is important for polyubiquitin chain formation. The truncation of C‐terminal tail decreases the stability and function of HERC2. Structural and mechanistic studies on HERC2 C‐lobe are important for elucidating HERC2s cellular role along with its implications with disease. Understanding the structure and function of HERC2 is essential for drug therapy.

4. Chemical biology: 7. Dynamics and allostery

ABS310

BIOPHYSICAL AND STRUCTURAL ANALYSIS OF DROSOPHILA TRANSCRIPTION FACTORS

Aaron Bogle ¹, Rachel Orlomoski², Robert Drewell², Jacqueline Dresch², Donald Spratt²

¹1996 (worcester, United States); ²Clark University (Worcester, United States)

Gene expression is regulated by transcription factors (TFs) that bind to specific DNA sequences. While some of the DNA sequences that these TFs bind to are known, there are many unanswered questions about how and why these TF‐DNA complexes form. How do the residues in the TF contribute to nucleotide specificity and recognition? Is the binding affinity of a TF‐DNA complex a good predictor for gene activation/repression? To answer these questions, the DNA‐binding domains of TFs were overexpressed in E. coli, purified, and biochemically characterized. In this study, the homeodomain of FUSHI TARAZU (FTZ‐HD) and the zinc‐finger domain of KRUPPEL (KR‐ZFD) were examined due to their importance in Drosophila melanogaster embryo development. Electrophoretic Mobility Shift Assays (EMSAs) were performed in the presence of different DNA sequences to see how slight changes in the consensus sequence affected binding affinity. Nuclear Magnetic Resonance (NMR) spectroscopy was used to examine the 3D structure of FTZ‐HD and to identify specific amino acids in FTZ‐HD involved in TF‐DNA complex formation. Isothermal Titration Calorimetry (ITC) experiments are currently being performed to determine the dissociation constant (K¬D) for these TF‐DNA complexes with different target DNA sequences. We have observed significant differences in TF‐DNA binding with even one nucleotide change demonstrating that these TFs show high specificity in DNA sequence recognition.

10. Folding: 16. Protein interactions and assemblies

ABS311

THE MEASUREMENT OF VOLUME CHANGE BY CAPILLARY DILATOMETRY

Peter Kahn ¹

¹Rutgers University (New Brunswick, United States)

Capillary dilatometry enables direct measurement of changes in volume, an extensive thermodynamic property. The results provide insight into the changes in hydration that occur upon protein folding, ligand binding and the interactions of proteins with nucleic acids and other cellular components. Often the entropy change arising from release of hydrating solvent provides the main driving force of a binding reaction. For technical reasons, though, capillary dilatometry has not been as widely used in protein biochemistry and biophysics as other methods such as calorimetry. Described here are simple apparatus and simple methods which bring the technique within the capacity of any laboratory. Even very simple results such as changes in hydration of potassium and chloride ions upon dilution are shown to have implications for macromolecular based phenomena. Examples of applications to protein folding are shown. These include estimations of the numbers of water molecules expelled from hydrophobic surfaces. For ribonuclease A ~240 waters are involved. For the molten globule transition of cytochrome c an upper limit of ~600 are expelled. Such estimations are essential input to statistical mechanical analyses of the contribution of these waters to the entropy of folding.

12. Membrane proteins: 16. Protein interactions and assemblies

ABS312

OLIGOMERIZATION OF LIPID MEMBRANE BOUND CYTOCHROME P450

Nirupama Sumangala ¹, Thirupathi Ravula², Ayyalusamy Ramamoorthy²

¹Biophysics, University of Michigan, Ann Arbor, MI, 48109 USA. (Ann Arbor, United States); ²Biophysics and Department of Chemistry, University of Michigan, Ann Arbor, MI, 48109 USA. (Ann Arbor, United States)

Cytochrome P450 (P450) are a remarkable class of heme‐containing, membrane‐bound enzymes that metabolize approximately 75% of pharmaceutical drugs on the market. For the enzymatic catalysis, cytochrome P450 require two electrons from its membrane‐bound redox partners, cytochrome P450 reductase (CPR) and cytochrome b5. Studies on cytochrome P450 metabolon have primarily been limited to the monomerized, truncated soluble domains of P450 due to many challenges posed by the membrane interaction of cytochrome P450. Recent NMR and biophysical studies have demonstrated the importance of the lipid membrane environment for the structural and functional stabilization of cytochrome P450 and its complex with redox partners. Considering the huge discrepancies between the concentration of P450s and CPR (5‐fold difference) in‐vivo, cytochrome P450s are likely to have a cross talk in the lipid‐membrane through oligomerization. In this study, we emphasize the importance of studying full‐length oligomeric P450s in native‐like membrane environment. We reconstituted P450 oligomers in lipid nanodiscs and characterized the spin‐state equilibrium in the presence of substrate and redox partners. By studying the shift in the hemes spin multiplicity from low to high spin upon the introduction of a substrate, our results suggest that P450 oligomerization influences the drug metabolism. Therefore, elucidation of factors that affect the cytochrome P450 enzyme catalysis in a physiologically relevant environment is critical for the development of better therapeutic drugs.

12. Membrane proteins: 7. Dynamics and allostery

ABS313

CARDIOLIPIN TRIGGERS CYTOCHROME‐C PEROXIDASE ACTIVITY VIA DYNAMIC CHANGES TO MEDIATE MITOCHONDRIAL APOPTOSIS

Mingyue Li ¹, Abhishek Mandal¹, Vladimir Tyurin², Maria DeLucia¹, Jinwoo Ahn¹, Valerian Kagan², Patrick van der Wel³

¹Department of Structural Biology, University of Pittsburgh (Pittsburgh, United States); ²Department of Environmental and Occupational Health, University of Pittsburgh (Pittsburgh, United States); ³Zernike Institute for Advanced Materials, University of Groningen (Groningen, Netherlands)

Mitochondrial apoptosis is a shared cell death mechanism, which is highly regulated by protein‐protein and protein‐membrane interactions. A molecular understanding of apoptosis is important for the development of therapeutics for neurogenerative disorders and cancers. The peroxidation of cardiolipins (CL) by reactive oxygen species (ROS), regulated and enhanced by cytochrome c (cyt c), is a pivotal signaling event. The oxidized CL species generated upon interactions with cyt c represent eat‐me signals on the mitochondrial surface. The structural details of cyt‐c and CL interaction may guide the development of novel inhibitors that target apoptosis.

Using mass spectrometry based lipidomics, we observed in vitro the preferential oxidation of polyunsaturated CL and identified key CL oxygenation products, which are potential apoptotic signaling molecules. Solid state NMR spectroscopy provides high‐resolution structural and dynamics insights into the CL‐bound cyt‐c and the membrane. Cyt‐c is localized at the hydrophilic membrane surface and forms a complex with CL nanodomains. In this complex, CL acts as a regulator of cyt c dynamics. Notably, the protein retains almost the same structure when it binds to different anionic phospholipids, which cause a lower peroxidase activation. While it has been suggested that partial unfolding of cyt c is necessary for its peroxidase activity, our studies reveal a modest amount of structural and dynamic changes in the redox‐active protein‐lipid complex, leading to a deeper understanding of the catalytic mechanism of pro‐apoptotic CL peroxidation.

Reference

Li., M et al., Structure 2019. DOI: https://doi.org/10.1016/j.str.2019.02.007

25. Transcription/translation/post‐translational modifications: 18. Proteomics

ABS314

IPTMNET: AN INTEGRATED RESOURCE FOR PROTEIN POST‐TRANSLATIONAL MODIFICATION NETWORK DISCOVERY

Cecilia Arighi ¹, Hongzhan Huang², Karen Ross³, JIa Ren², Julie Cowart², Sachin Gavali², Qinghua Wang², K Vijay‐Shanker², Cathy Wu²

¹CBCB, University of Delaware (Newark, United States); ²University of Delaware (Newark, United States); ³Georgetown University Medical Center (District of Columbia, United States)

Protein post‐translational modifications (PTMs) play a pivotal role in numerous biological processes by modulating regulation of protein function. We have developed iPTMnet for PTM knowledge discovery, employing an integrative bioinformatics approachcombining text mining, data mining, and ontological representation to capture rich PTM information, including PTM enzyme‐substrate‐site relationships, PTM‐specific protein‐protein interactions (PPIs) and PTM conservation across species. iPTMnet encompasses data from i) our PTM‐focused text mining tools, RLIMS‐P and eFIP, which extract phosphorylation information from full‐scale mining of PubMed abstracts and full‐length articles including phosphorylation dependent protein‐protein interactions (PPIs); ii) a set of curated databases with experimentally observed PTMs; and iii) Protein Ontology (PRO) that organizes proteins and PTM proteoforms (observed protein forms derived from a single gene, including isoforms and combinations of PTMs), enabling their representation, annotation and comparison within and across species. Presently covering eight major PTM types (phosphorylation, ubiquitination, acetylation, methylation, glycosylation, S‐nitrosylation, sumoylation and myristoylation), iPTMnet has several unique features: i) rich up‐to‐date integrated PTM information from the scientific literature and knowledgebases; ii) representation of PTM proteins, PTM enzymes and their relations at the proteoform level; iii) score for the quality of the underlying data; iv) network visualization of PTM enzyme‐substrate‐site and PPI relations; and v) sequence alignment visualization of singly‐modified, multiply‐modified and/or overlapping PTM forms/sites within and across species. iPTMnet connects PTM proteoforms with their modifying enzymes and multiple coordinated PTMs across taxa, thereby unifying fragmented PTM information into a biologically meaningful context for visual and systematic PTM knowledge discovery.

1. Amyloid and aggregation: 21. Structure (x‐ray/NMR/EM)

ABS315

FIBRIL FORMATION, PHASE TRANSITION, AND INTERACTORS OF ORB2, A PROTEIN IMPORTANT IN LONG‐TERM MEMORY

Connor Hurd¹, Ansgar Siemer ¹, Connor Hurd², Silvia Cervantes², Alexander Falk², Maria Soria², Samridhi Garg², Ansgar Siemer²

¹University of Southern California (Los Angeles, United States); ²Univeristy of Southern California (Los Angeles, United States)

Orb2 is a key regulator of long‐term memory (LTM) in Drosophila. It can form amyloid‐like aggregates in vivo and in vitro. Mutants that impede aggregation also affect LTM. Orb2 has low complexity sequences at its N‐terminus and RNA recognition motifs at its C‐terminus reminiscent of proteins that are known to undergo liquid‐liquid phase separation.

One focus of our research has been the N‐terminal amphipathic domain unique to the low abundant isoform A (Orb2A). Mutations of this region can effectively block LTM. Using solid‐state NMR, EPR, and other biophysical techniques, we found that this region can form amyloid fibrils on its own. In addition, we described multiple binding partners that interfere with the aggregation of this region. In vivo, the presence of Orb2A is essential for the aggregation of the more common isoform Orb2B. To determine the interaction of these two isoforms, we studied the behavior of full‐length Orb2A and Orb2B. We found that Orb2 can undergo a phase separation and that phase separation has an influence on its ability to form fibrils. Finally, we will present solid‐state NMR data that compare fibrils formed by the N‐termini of Orb2B and Orb2A.

2. Bioinformatics: 10. Folding

ABS318

USING DEEP LEARNING NEURAL NETWORKS FOR INVERSE PROTEIN FOLDING PREDICTIONS

Alyssa La Fleur ¹, Tersa Almaw¹, Deanna Ojennus¹, Kent Jones¹

¹Whitworth University (Spokane, United States)

A high accuracy Deep Learning Neural Network (DNN) for predicting residue position amino acid probabilities in a known protein structure was replicated with a non‐supercomputer training regime and improved through structure and feature engineering. DNNs like this could be useful for generating solutions to the Inverse Protein Folding Problem (IPFP) by generating a maximum likelihood sequence of amino acids that fold to a desired structure. Network improvements included adding batch normalization and dropout regularization layers, and the engineering of two new features: the number of amino acids in spheres of increasing radii around the position being predicted, and hydrophobicity scores of neighboring amino acid side chains. Two networks were created using these improvements, a backbone DNN which added the radii feature and a backbone/sequence DNN which added both new features. Five‐fold cross‐validation was used to train the replicated, backbone, and backbone/sequence DNNs resulting in 38.2%, 40.9% and 44.1% accuracy, respectively. All networks were used to compare predicted probabilities with known stability profiles for mutants of T4 lysozyme and also with activity data for mutants of a prolyl aminodipeptidase (PEPX ). Both improved DNNs assigned the highest probability T4 Lysozyme amino acid identity as the naturally occurring residue with a correlation between decreasing probability and stability for each mutant examined. PEPX prediction distributions were found to be heavily dependent on minor structural changes. Additional work is needed to optimize network parameters for the improved network structure, and to classify differences between predicted amino acids and desired enzyme activity and/or stability.

11. Intrinsically disordered proteins: 21. Structure (x‐ray/NMR/EM)

ABS320

THE HUMAN ZINC‐ AND IRON‐REGULATED TRANSPORT PROTEIN 4 INTRACELLULAR LOOP REMAINS DISORDERED UPON HIGH‐AFFINITY ZINC BINDING

Elizabeth Bafaro ¹, Mark Maciejewski², Jeffrey Hoch², Robert Dempski¹

¹Worcester Polytechnic Institute (Worcester, United States); ²UConn Health, University of Connecticut (Farmington, United States)

The human zinc and ironregulated transport protein 4 (hZIP4) is the main transporter responsible for dietary zinc uptake and is vital for maintaining cellular zinc homeostasis. hZIP4 plasma membrane levels respond to intracellular zinc concentrations via a mechanism that involves post‐translational modification and a histidine‐rich region on its large intracellular loop (ICL2). To gain insight into the zinc sensing mechanism by ICL2, we examined the structural characteristics of the isolated ICL2 protein region, both in the absence and presence of zinc, using nuclear magnetic resonance (NMR) spectroscopy. NMR chemical shifts of the isolated ICL2 protein were consistent with an intrinsically disordered protein. The ICL2 protein remained disordered upon zinc binding, although it exhibits nanomolar affinity for zinc. Zinc coordination resulted in exchange broadening of residues within the histidine‐rich region and supports a model of flexible zinc binding sites. Chemical shift displacement of the single, distal histidine, located near sites for post‐translational modification, provides the first evidence for metal interactions outside the histidine‐rich region. Transient tertiary interactions were probed using paramagnetic relaxation enhancement experiments and indicated contacts between the histidinerich region and predicted posttranslational modification sites in the disordered ensemble. These residuespecific data support a model whereby hZIP4 ICL2 acts as a zinc sensor to maintain cellular zinc homeostasis. Furthermore, zinc sensing by the ICL2 protein demonstrates that highaffinity binding can be congruent with conformational disorder.

5. Computational modeling/simulation: 11. Intrinsically disordered proteins

ABS321

THE RATIONAL DISCOVERY AND DESIGN OF DISORDERED PROTEIN LIGANDS

David Baggett ¹, Abhinav Nath¹

¹University of Washington (Seattle, United States)

Intrinsically disordered proteins (IDPs) are core components of many biological processes and are central players in several pathologies. Despite being important drug targets, attempts to design small molecule ligands that would help understand and attenuate their behavior are frustrated by the structural diversity exhibited by these flexible proteins. To accommodate the dynamic nature of IDPs, we have developed a procedure that efficiently identifies active small‐molecule ligands for disordered proteins. This method utilizes enhanced sampling molecular dynamics and adapts conformational clustering algorithms to identify persistent local structures. Ligands are then screened using a combination of computational docking and machine learning to identify promising lead compounds.

We tested our strategy by targeting the microtubule binding region of the disordered protein tau, and successfully identified novel tau ligands. By exploring the chemical space around these ligands, we have been able to identify factors critical for activity and affinity and use this information to identify more active compounds.

With this method we identified novel ligands that expand our toolkit of protein aggregation inhibitors into new areas of chemical space. Notably, the discovery of this new family of disordered protein ligands was achieved more quickly and with less expense than conventional high‐throughput screening or docking alone would have allowed.

5. Computational modeling/simulation: 7. Dynamics and allostery

ABS322

ANALYSIS OF LOOP MOTIONS IN 1 μS SIMULATIONS OF OXA‐66 REVEALS STRIKING DIFFERENCES IN FLEXIBILITY BETWEEN MUTANTS

Joshua Grey ¹, David Leonard², Agnieszka Szarecka³

¹Grand Valley State University (Allendale, United States); ²Grand Valley State University, Department of Chemistry (Allendale, United States); ³Grand Valley State University, Department of Cell and Molecular Biology (Allendale, United States)

Antibiotic resistant bacteria are a leading cause of nosocomial infections and use ‐lactamases as their primary mechanism of resistance against ‐lactams. Mutants of OXA‐66 ‐lactamase isolated from drug‐resistant strains of A. baumannii display alarming potential for activity against carbapenems. Previous simulations of OXA‐66 mutants indicate a significant modulation of loop dynamics surrounding the active site, including greater flexibility and enlargement of the binding pocket. However, it is not clear if the timescale of our approximately 250 ns simulations allows sufficient sampling of loop conformations. Here we present data from 1 μs timescale molecular dynamics simulations of OXA‐66 P130A, I129L/L167V, and W222L mutants. We performed C‐RMSD, C‐RMSF, and principal component analyses focusing on binding pocket loops in addition to the entire protein. Our data suggest that the 24 amino acid long P loop undergoes large conformational changes in P130A and W222L mutants, but is less flexible in I129L/L167V. In contrast, another important structural segment, the ‐loop, appears to exhibit more flexibility in I129L/L167V while remaining rigid in P130A and W222L mutants. Additionally, we find conformational changes in the 711 loop and motional correlation to P‐loop across all mutants. Our data indicate that 1 μs is sufficient for sampling of both P and ‐loops and reveal patterns of correlated motions in P, 56, and 711 loops.

3. Chaperones: 10. Folding

ABS323

MEASURING THE UNFOLDING AND LIGAND‐BINDING OF CUSF, A COPPER CHAPERONE

Isabel Zecua ¹, Blake Gillespie²

¹CSU Channel Islands (Camarillo, United States); ²Undergraduate Independent Research Professor (Camarilo, United States)

Although a protein's stability derives from its structure, ligands can also affect structure and stability. The bacterial copper chaperone CusF is stabilized via ligand binding of silver or copper ion, showing increased stability to thermal and chemical denaturation. We are investigating the thermodynamic mechanism of this stabilization through tryptophan fluorescence measurements of CusF unfolding behavior in the absence (apo) and presence (holo) of its ligand Ag(I). Stability curve analysis shows the temperature‐ and ligand dependence of enthalpy, entropy, and heat capacity of unfolding. We show that ligand binding increases CusF's unfolding enthalpy (H) by approximately 5kcal‐mol, its heat capacity (Cp) by 0.5kcal‐mol, and its entropy (S) by 30 cal‐mol. Though the enthalpy change is the most obvious determinant of CusF's ligand‐dependent stabilization, the mixed mechanism ‐ particularly the heat capacity change ‐ leaves open the possibility that CusF might show residual structure in the unfolded state of its holo‐structure. Such residual structure would narrow the difference between folded and unfolded state heat capacities, resulting in a smaller Cp and a wider range of apparent thermal stability.

25. Transcription/translation/post‐translational modifications: 16. Protein interactions and assemblies

ABS324

BIOPHYSICAL AND STRUCTURAL ANALYSIS OF ANTENNAPEDIA AND ULTRABITHORAX HOMEODOMAIN TRANSCRIPTION FACTOR‐DNA BINDING AFFINITIES

Jeanmarie W. Loss ¹, Rachel J. Orlomoski¹, Jacqueline M. Dresch¹, Robert A. Drewell¹, Donald E. Spratt²

¹Clark University (Worcester, United States); ²Clark University (Worcester, MA, United States)

Transcription factors (TFs) are proteins that assist in controlling gene expression and bind at cis‐regulatory modules (CRMs) upstream of genes. The activity of these TFs can both activate and repress the transcription of a given gene associated with the CRM, and in Drosophila melanogaster (fruit fly), the development of the embryo is controlled by a cascade of TF activity. Currently, there is a poor understanding of how these TFs recognize and bind to their respective DNA targets. To answer this important question, two homeodomain (HD) TFs, Antennapedia (Antp) and Ultrabithorax (Ubx), were overexpressed in E. coli and purified for electrophoretic mobility shift assays (EMSAs) and isothermal titration calorimetry (ITC) analysis. Our collective results refute the notion that the most frequently found DNA binding sequence by SELEX data is also the tightest binding affinity, suggesting that HD‐DNA binding is specific and that the sequences being bound to may be interdependent. Nuclear magnetic resonance (NMR) spectra have also been collected and analyzed for Ubx and our results confirm the presence of the helix‐turn‐helix motif that has previously been found within other HD‐TFs. Future studies will include the structural characterization of these HD‐DNA complex, allowing for further clarification on specific contact points between the HD and the DNA sequence.

16. Protein interactions and assemblies: 21. Structure (x‐ray/NMR/EM)

ABS325

BIOPHYSICAL EXAMINATION OF UBIQUITIN E3 LIGASE, HECTD1: AN IMPORTANT REGULATOR IN NEUROLOGICAL DEVELOPMENT

Misa Mai ¹, Donald Spratt¹

¹Clark University (Worcester, United States)

Homologous to the E6AP Carboxyl Terminus (HECT) Domain E3 Ubiquitin Protein Ligase 1 (HECTD1) is an E3 ubiquitin ligase involved in the post‐translational attachment of ubiquitin to a target protein for proteasomal degradation. Recent studies have shown that HECTD1 dysfunction can cause glioblastoma and exencephaly through dysregulation of head mesenchyme formation. Currently there is not 3D structure available for the catalytic core of HECTD1. In this study, we set about examining the C‐terminal lobe of the HECTD1 catalytic domain to better understand how this region of the protein orchestrates ubiquitin transfer on to a substrate protein. This required the overexpression of the protein in E. coli using 13C/15N‐labeling strategies for 2D/3D nuclear magnetic resonance (NMR) spectra. The backbone resonances of the HECTD1 C‐lobe have been unambiguously assigned and secondary structural elements in the domain have been predicted using the chemical shift index. The biochemical activity of HECTD1 substituted proteins were also assessed by in vitro ubiquitylation assays. Our cumulative results indicate that HECTD1 C‐lobe is a stable, monomeric protein that is capable of building polyubiquitin chains. These studies provide initial insight into HECTD1 E3 ubiquitin ligase mechanism and lay the groundwork for future therapeutic development.

16. Protein interactions and assemblies: 18. Proteomics

ABS326

STRUCTURAL EXAMINATION OF THE HECT E3 LIGASE AREL1 AND ITS IMPLICATIONS IN APOPTOSIS

Emily Ladda ¹, Donald Spratt²

¹Clark University (Worcester, United States); ²Clark University (Worcester, United States)

Apoptosis Resistant E3 Ligase 1 (Arel1) is a member of the HECT (Homologous to E6AP Carboxyl Terminus) E3 ubiquitin ligase family. All HECT E3 ligases contain a conserved HECT domain which facilitates the covalent attachment of ubiquitin onto target substrates. Ubiquitin signaling is involved in numerous cell signaling pathways that result in proteasomal degradation, DNA damage repair, and cell trafficking, among others. The biological function of Arel1 is to inhibit intrinsic apoptosis by ubiquitination of pro‐apoptotic IAP (inhibitor of apoptosis) antagonist proteins, SMAC, ARTS, and HtrA2 for proteasomal degradation. Interestingly, previous studies have found that Arel1 is the only HECT E3 ubiquitin ligase identified to date, that is capable of building polyubiquitin chains via lysine 33 linkages. In order to better understand the structure and elucidate the molecular function of Arel1 and the formation of Lys33 polyubiquitin chains, circular dichroism (CD), in vitro ubiquitin activity assays, and western blots were employed to study both the C‐lobe and HECT domain of Arel1. Both C‐lobe and HECT Arel1 were found to be E2 independent, however, this function was much diminished in HECT Arel1. Additionally, HECT Arel1 was capable of forming ubiquitin chains as visualized via ubiquitin activity assays. This project provides insight into the structure and mechanism of Arel1s HECT domain and Lys33 chain formation.

21. Structure (x‐ray/NMR/EM): 8. Enzymology

ABS328

STRUCTURE AND INTRINSIC HYDROLYSIS OF NRAS Q61 MUTANTS

Derion Reid ¹, Spiro Pavlopoulos¹, Carla Mattos²

¹Northeastern University (Boston, United States); ²Northeastern University (Boston, United States)

Ras GTPases are molecular switches involved in diverse signal transduction pathways that regulate cellular proliferation, survival, migration, and apoptosis. Ras is activated by guanine nucleotide exchange factors (GEFs) that facilitate the exchange of guanosine diphosphate (GDP) for guanosine triphosphate (GTP) and deactivated with the aid of GTPase activating proteins (GAP). More recently intrinsic hydrolysis has been proposed to play a role in the context of the effector Raf, where conformational states of the switch regions are modulated by binding of Ca2+ at a remote allosteric site. While HRas and KRas have been extensively studied, the structural biology and biochemistry of NRas have been largely overlooked. NRasQ61 mutants are known as primary drivers of melanoma tumors, yet there are no inhibitors that target these proteins. The goal of this work is to determine how NRas Q61 mutants affect its structure and hydrolytic activity. Our lab solved the structure of NRas bound to the GTP analogue GppNHp, which highlighted biochemical and structural differences compared to the other Ras isoforms. Furthermore, the NMR spectrum of NRas suggests open and disordered switch regions compared to HRas and KRas. Here we present structures of the common oncogenic mutants NRasQ61H, NRasQ61L, and NRasQ61R, accompanied by hydrolysis data showing that each mutant lowers the hydrolysis rate constant of NRas in unique ways. The structures of the NRas mutants show stabilization of the switch regions in an anti‐catalytic conformation. Understanding these conformational preferences will help further influence the development of novel therapeutic targets of NRas mutant melanoma.

21. Structure (x‐ray/NMR/EM): 10. Folding

ABS331

STRUCTURE‐ACTIVITY RELATIONSHIPS IN THE METAMORPHIC, ANTIMICROBIAL PROTEIN XCL1

Acacia Dishman ¹, Michelle Lee², Gerard Wong², Brian Volkman¹

¹Medical College of Wisconsin (Milwaukee, United States); ²UCLA (Los Angeles, United States)

With the number of multi‐drug resistant pathogens on the rise, it is necessary to develop new antimicrobial therapeutics. Antimicrobial peptides (AMPs) are innate immune molecules that act as potent, broad spectrum antimicrobial agents, and can provide novel design ideas for the development of new drugs. Metamorphic proteins are another unique protein family that can stimulate ideas for new, switchable drugs. These proteins adopt two or more different native structures, expanding the traditional protein folding paradigm. XCL1 is a metamorphic, antimicrobial protein which populates two distinct native structures. One native structure has broad‐spectrum antibacterial activity, while the other is a significantly less potent bacterial killer. Here, we aim to better understand XCL1s antimicrobial behavior from a unique structural perspective. In order for AMPs to kill microbes via membrane disruption, they must induce negative gaussian curvature (NGC) in the organisms membrane. This behavior can be measured via small angle X‐Ray scattering (SAXS). Here, SAXS was used to assess the ability of XCL1, and a panel of XCL1 structural variants, to induce NGC in various artificial membrane compositions. XCL1 induced NGC in E. Coli‐like membrane compositions with the same structural dependence seen in previous bactericidal assays. Unexpectedly, XCL1 was also able to induce NGC in candida‐like membranes. XCL1s anti‐candida activity was confirmed in vitro, uncovering a new antimicrobial function in XCL1. Future work will include exploration of structural dependence of XCL1s anti‐candida action. In all, a better understanding of XCL1s antimicrobial structure‐activity relationships can inform the future development of switchable antimicrobial agents.

2. Bioinformatics: 8. Enzymology

ABS333

FUNCTIONALLY RELEVANT CLUSTERING OF THE ARSENATE REDUCTASE (ARSC) SUPERFAMILY

Mikaela Rosen ¹, Jacquelyn Fetrow², Carol Parish¹, Janelle Leuthaeuser¹

¹University of Richmond (Richmond, United States); ²Albright College (Reading, United States)

We are applying computational methods that can expand our understanding of the atomic‐level detail of protein active sites, focusing on the arsenate reductase (ArsC) superfamily. Global arsenic concentrations have been rising, especially due to increased pollutants. ArsC proteins are vital within arsenic redox microorganisms that lower the concentrations of arsenate and are, therefore, important for the bioremediation strategies that prevent arsenic from reaching alarming levels in the environment. Our research utilizes Multi‐level Iterative Sequence Searching Technique (MISST), a method which uses the functional site profiles from structurally known proteins to search Genbank and to identify functionally relevant groups that comprise protein superfamilies. Our results suggest that the ArsC superfamily, previously thought of as two or three groups, should in fact be six isofunctional groups. We have evaluated each of these groups, in turn, providing details on active site features, Genbank function annotations and phylogenetic distributions. Several of these groups were found to have stark similarities to protein tyrosine phosphatases (PTP) and transcriptional regulators. In some cases, active site features were additionally evaluated using molecular dynamics and electrostatics calculations to identify potential mechanistic determinants for the redox mechanism in each group. These results provide insight into distinct arsenate reductase mechanisms and present hypotheses about enzyme mechanism that can now be tested experimentally.

11. Intrinsically disordered proteins: 21. Structure (x‐ray/NMR/EM)

ABS334

DEFINING GP41‐1 EXTEIN SPLICE JUNCTION

Carla Madrid ¹, Thuy Nguyen², Kimberly Reynolds², Kendra Frederick²

¹UT Southwestern Medical Center (Dallas, United States); ²UT Southwestern Medical Center (Dallas, United States)

Split inteins are an intervening polypeptide sequence that excises from a precursor protein with concomitant splicing of the flanking sequence. This process is considered to be a post‐translational modification. Split inteins undergoes protein splicing which is an autocatalytic process. Split Inteins can be found in coding regions of prokaryotes, unicellular eukaryotes and organelles. The chemistry involved in the protein splicing is important and is dependent on the amino acids on the C terminal of the precursor protein. Which is why we have set out to determine what amino acid sequence is valid for the chemistry to occur. We are looking at the N and C terminal region of the native protein to help us better understand the chemistry. This is important for us because we can develop techniques that allows us design split intein within the protein sequence without disrupting the sequence of a native protein.

21. Structure (x‐ray/NMR/EM): 5. Computational modeling/simulation

ABS335

ACHIEVING BETTER‐THAN‐2‐å RESOLUTION BY SINGLE‐PARTICLE CRYO‐EM AT 200 KEV

Mark Herzik ¹, Mengyu Wu², Gabriel Lander²

¹University of California, San Diego (La Jolla, United States); ²The Scripps Research Institute (La Jolla, United States)

Nearly all single‐particle cryo‐EM structures resolved to better than 4‐å resolution have been determined using 300‐keV transmission electron microscopes (TEMs). We demonstrate that it is possible to obtain reconstructions of macromolecular complexes of different sizes to better than 2‐å resolution using a 200‐keV TEM. These structures are of sufficient quality to unambiguously assign amino acid rotameric conformations, identify ordered water molecules and bound ligands, as well as resolve "holes" within cyclic side chains. These unprecedented results push the limits of conventional underfocus single‐particle cryo‐EM closer to the theoretical limits.

16. Protein interactions and assemblies: 21. Structure (x‐ray/NMR/EM)

ABS336

INVESTIGATING BACTERIAL SORTASE SUBSTRATE SELECTIVITY USING ANCESTRAL PROTEIN RECONSTRUCTION AND SEQUENCE NETWORK ANALYSIS

Jordan Valgardson¹, Jordan Valgardson ¹, Sarah Struyvenberg¹, Zach Sailer², Jeanine Amacher¹

¹Western Washington University (Bellingham, United States); ²University of Oregon (Eugene, United States)

Sortase proteins are bacterial transpeptidases that attach proteins to the bacterial cell wall in gram‐positive bacteria. The sortase A class of proteins see use in a laboratory setting as a protein engineering tool, allowing for the combination of two protein products localized at a LPXTG sorting motif on one of the two proteins. While sortase A proteins have seen success for protein engineering, they are limited by their high substrate specificity and low efficiency. To understand the bio‐structural basis of substrate specificity in sortase enzymes, as well as, the evolution of their substrate specificity, we developed an ancestral sequence reconstruction pipeline for the sortase A protein family. Utilizing the NCBI non‐redundant sequence database for sortase we first generated a sequence similarity network with 3383 nodes. From this sequence similarity network, we were able to unambiguously distinguish sortase A sequences from the other classes of sortase enzyme. With 371 sortase A sequences in hand, we were able to generate a high confidence phylogenetic tree for sortase A. The phylogenetic tree and corresponding multisequence alignment show that the terminal loop region connecting the catalytic region and the subsequent beta sheet varied with different substrate specificity, and likely plays a role in substrate specificity. Additionally, we have generated two ancestral proteins that are ancestors to the Staphylococcus aureus and Streptococcus pneumonia sortase A families. Moreover, the network analysis of the sortase protein family indicates many more sortase classes than the 6 that are reported in the literature.

10. Folding: 20. Single molecule studies

ABS337

EFFECT OF EXPERIMENTAL PARAMETERS OF OPTICAL TRAPS ON FOLDING/UNFOLDING DYNAMICS OF FAST FOLDING PROTEINS

Rama Reddy Goluguri ¹, Mourad Sadqi¹, Victor Munoz¹

¹UC Merced (Merced, United States)

Optical traps have been used to study mechanical unfolding of single protein molecule tethered between two beads. DNA handles are conjugated on either side of the protein to minimize bead protein interaction and to minimize non‐ specific bead‐bead interactions. Theoretical studies have suggested that the stiffness of the trap and length of the DNA molecular handles affect the dynamics of the protein being studied. Mechanical force was shown to induce huge barriers in a small fast folding protein gpW, which was shown to have a marginal barrier between folded and unfolded state. We have aimed to study the effect of experimental parameters of optical tweezer experiments on folding/unfolding dynamics of fast folding proteins. Preliminary results on the mechanical folding/unfolding studies on fast folding protein Engrailed homeodomain will be discussed.

10. Folding: 6. Design/engineering

ABS340

OPTIMIZATION OF PROTEIN STRUCTURE AND FUNCTION: THE IMPORTANCE OF LOOP LENGTH

Neha Nandwani ¹, Praneeth Reddy², Manjula Ramu², Jayant Udgaonkar³, Shachi Gosavi²

¹Stanford University (Stanford, United States); ²National Centre for Biological Sciences (Bengaluru, India); ³Indian Institute of Science Education and Research, Pune (Pune, India)

The folding energy landscape of a protein is modulated by the function of the protein. Functional residues (often found in loops) are not evolutionarily optimized for folding and can therefore result in the introduction of conformational frustration in the protein. Although such strain has been shown to facilitate function in several proteins, it also results in a rugged folding landscape and a conformational switch can often dissipate such strain. Function therefore renders a protein structurally vulnerable, and it is important to ask how proteins balance function and structure. We have recently demonstrated that the five‐residue motif, QVVAG, when engineered in a surface loop in a monomeric protein drives domain swapping. Here, using several mutants of the monomeric plant protein monellin, we show that QVVAG is a frustrating sequence, which increases the potential energy of the loop into which it is engineered, resulting in domain swapping. The conformational switch to the domain‐swapped state relieves the frustration in the loop because of the change in the dihedral angles of the loop residues, and is also expected to result in the loss of function. We observe that the increase in potential energy and correspondingly the propensity to undergo domain swapping is inversely correlated to the length of the target loop. Hence, the length of the loop containing the functional residues of a protein is critical and its optimization over evolution could be one factor that helps in the maintenance of the fine balance between a proteins function and structure.

17. Proteins in cells: 16. Protein interactions and assemblies

ABS341

BIOPHYSICAL AND BIOCHEMICAL CHARACTERIZATION OF HACE1, A HECT E3 UBIQUITIN LIGASE IMPLICATED IN CANCER AND HUNTINGTONS DISEASE

Diana Argiles Castillo ¹, Donald Spratt¹

¹Clark University (Worcester, United States)

HACE1 (HECT domain and Ankyrin repeat‐Containing E3 ubiquitin protein ligase 1) is a Homologous to E6AP C‐terminal (HECT) E3 ubiquitin ligase involved in the ubiquitylation‐signaling pathway. HECT E3 ligases catalyze the covalent attachment of ubiquitin to substrate proteins to control signal transduction, DNA repair, cell cycle progression and gene expression. HACE1 is downregulated in various cancers such as Wilms Tumor, the most common childhood kidney cancer. HACE1 has also been linked to the antioxidative stress response pathway and has been found to be downregulated in Huntingtons disease. To learn how this important enzyme works at the atomic level, the HACE1 HECT domain C‐lobe, the region of the enzyme responsible for ubiquitin transfer activity, was overexpressed in E. coli and purified. The isolated protein was then structurally and biochemically examined using analytical size exclusion chromatography, circular dichroism, 2D/3D nuclear magnetic resonance (NMR) spectroscopy, and ubiquitin activity assays. We demonstrate that the HACE1 C‐lobe can form a stable disulfide complex with ubiquitin to mimic the short‐lived thioester HACE1~ubiquitin intermediate. Ubiquitin activity assays revealed that the HACE1 C‐lobe can build unanchored polyubiquitin chains, a characteristic linked to Huntingtons disease. NMR resonance assignments for the backbone of HACE1 C‐lobe have been completed in preparation for future structural calculations. These mechanistic and structural studies of HACE1 will lay the groundwork for therapeutic development.

ABS342

ROOM TEMPERATURE CRYSTALLOGRAPHY OF RETINAL PROTEINS: INVESTIGATING THE RETINAL ISOMERIZATION MECHANISM

Gebhard F.X. Schertler ¹

¹Paul Scherrer Institut / ETH Zurich D‐BIOL (Villigen PSI, Switzerland)

Free‐electron lasers are X‐ray sources with unprecedented peak brilliance and time structure. We are applying them to explore serial crystallography for membrane proteins using bacteriorhodopsin or rhodopsin as a model system. Our team was able to trigger their photo cycles in lipidic cubic phase crystals of the retinal proteins. In bacteriorhodopsin, it was possible to observe for the first time excited state dynamics immediately after photon absorption. In addition, two time points immediately after the isomerization event show the relaxation to the first photo intermediates. We have carried out similar studies with bovine rhodopsin crystals and we obtained a series of time points for rhodopsin: one picosecond, 100 picoseconds, 16 nanoseconds, 2 microseconds and 200 microseconds after light illumination. In both retinal systems, our goal is a better understanding of quantum efficiency and stereo selectivity of the isomerization reactions in the retinal protein and how the changes in the bound ligand is converted into a conformational change of the protein. Free electron lasers for the first time enable us to obtain direct structural data on an enzyme reaction from femtoseconds to microseconds. In bacteriorhodopsin changes in the water mediated hydrogen bond network around the chromophore lead to important insights on the regulation of the pK of the retinylidene Schiff base, which is at the center of the proton pump mechanism. In rhodopsin, the reorganization of residues close to the beta ionone ring is important and the early changes of carboxyl groups close to the Schiff base can be studied. The comparison of different retinal proteins will also deepen our understanding how the protein environment is modulating the stereo selectivity of these systems.

6. Design/engineering: 22. Synthetic biology

ABS343

DESIGNER PROTEINS ‐ FROM FOLD TO APPLICATIONS

Eva‐Maria Strauch ¹, David Baker²

¹University of Georgia (Athens, United States); ²University of Washington (Seattle, United States)

Many viral surface glycoproteins and cell surface receptors are homo‐oligomers, and hence can potentially be targeted by geometrically matched homo‐oligomers that engage all subunits simultaneously to attain high avidity and/or lock subunits together. The adaptive immune system cannot generally employ this strategy since the individual antibody binding sites are not arranged with appropriate geometry to simultaneously engage multiple sites in a single target homo‐oligomer. Secondly, to provide the protein‐scaffolding framework, we will introduce a computational platform that enables us to efficiently sample and design any given topologies with high structural diversity to serve as new structural building block, guide future design efforts and help our general understanding of stability. Nature only samples a small fraction in sequence space, yet many more amino acid combinations can fold into stable proteins. Using a high‐throughput stability screen, we evaluated 45,000 of 9 topologies designed with our new pipeline and derived stability prediction models using machine learning algorithms.

16. Protein interactions and assemblies: 4. Chemical biology

ABS344

IDENTIFYING HOT SPOT RESIDUES AT THE ETV6 PNT DOMAIN POLYMERIZATION INTERFACE

Sophia Cho ¹, Chloe Gerak¹, Michel Roberge¹, Lawrence McIntosh¹

¹University of British Columbia (Vancouver, Canada)

ETV6 (or TEL) chimeric oncoproteins are linked directly to many human cancers. These constitutively active oncoproteins are frequently formed by chromosomal translocations that fuse gene fragments encoding the self‐associating PNT (or SAM) domain of this ETS family transcriptional repressor to those encoding the kinase domain from one of many protein tyrosine kinases or to the DNA‐binding domain from one of several transcription factors. We hypothesize that inhibition of PNT domain polymerization could be a viable therapeutic strategy for treating many ETV6‐driven cancers. However, the PNT domain polymerizes via the high‐affinity head‐to‐tail association of two relatively flat interfaces that are likely difficult to target with small molecule inhibitors. Our research objective is to dissect the structural and thermodynamic contributions of interfacial residues towards PNT domain polymerization. To this end, we have used alanine scanning mutagenesis combined with surface plasmon resonance binding studies to identify "hot spot" residues for polymerization. The identification of these critical residues will aid on‐going in silico drug design efforts to discover lead compounds that disrupt the PNT domain‐mediated self‐association of ETV6 oncoproteins.

5. Computational modeling/simulation: 25. Transcription/translation/post‐translational modifications

ABS345

DYNAMICS OF OPENING AND CLOSING MOTIONS OF THE CLAMP OF BACTERIAL RNA POLYMERASE

Ilona Christy Unarta ¹, Kubo Shintaroh², Wei Wang¹, Xuhui Huang³, Shoji Takada², Xuhui Huang¹

¹Hong Kong University of Science and Technology (Hong Kong, China); ²Kyoto University (Kyoto, Japan); ³Hong Kong University of Science and Technology (Hong Kong, China)

RNA Polymerase (RNAP) is an enzyme responsible of transcription in the central dogma. RNAP structure resembles a pair of crab claw. The clamp domain of RNAP is the flexible domain, which can open and close (Fig. 1) depending on the transcription step. Clamp domain motion is known to be important in other steps of transcription, i.e.: initiation step and paused state. To understand how RNAP regulates the clamp domain motion in atomistic resolution, we employed extensive molecular dynamics (MD) simulation and Markov State Model (MSM). Using the structures from cumulated 60 s AA‐MD simulations, we built a reliable MSM to obtain the timescale of the clamp motion and the metastable intermediate states between open and closed states. We observed that clamp opens in ~200 s timescale. A notable metastable state between open and closed state is the swiveled state, which is a motion perpendicular to the clamp swinging and is also an important motion of RNAP. In addition, the clamp opening seems to be highly correlated with the change of secondary structure in the switch regions, regions at the base of the clamp domain. With this observation, we may be able to look further into the specific residues that regulates the clamp domain.

9. Evolution: 8. Enzymology

ABS347

RESURRECTION OF ANCESTRAL EFFECTOR CASPASES IDENTIFIES NOVEL NETWORKS FOR EVOLUTION OF SUBSTRATE SPECIFICITY

Clay Clark¹, Clay Clark ¹, Robert Grinshpon², Suman Shrestha³, James Titus‐McQuillan³, Paul Hamilton², Paul Swartz²

¹University of Texas at Arlington (Arlington, United States); ²NC State University (Raleigh, United States); ³UT Arlington (Arlington, United States)

Apoptotic caspases evolved from a common ancestor (CA) more than 650 million years ago, and the CA provided the caspase‐hemoglobinase scaffold found in extant caspases. A series of gene duplications resulted in two subfamilies consisting of initiator and effector caspases, and the effector caspase genes (‐3, ‐6, and ‐7) were fixed into the Chordata phylum before the emergence of ray‐finned fish and have persisted throughout mammalian evolution. All caspases require an aspartate residue at the P1 position of substrates, so each caspase evolved discrete cellular roles through changes in substrate recognition at the P4 position combined with allosteric regulation. We examined the evolution of substrate specificity in caspase‐6, which prefers valine at the P4 residue, compared to caspases‐3 and ‐7, which prefer aspartate, by reconstructing the CA of effector caspases (AncCP‐Ef) and the CA of caspase‐6 (AncCP‐6An). We show that AncCP‐Ef is a promiscuous enzyme with little distinction between Asp, Val, or Leu at P4. The specificity of caspase‐6 was defined early in its evolution, where AncCP‐6An demonstrates preference for Val over Asp at P4. Structures of AncCP‐Ef and of AncCP‐6An show a network of charged amino acids near the S4 pocket that, when combined with repositioning a flexible active site loop, resulted in a more hydrophobic binding pocket in AncCP‐6An. The ancestral protein reconstructions show that the caspase‐hemoglobinase fold has been conserved for over 650 million years and that only three substitutions in the scaffold are necessary to shift substrate selection toward Val over Asp.

16. Protein interactions and assemblies: 21. Structure (x‐ray/NMR/EM)

ABS348

STRUCTURAL BIOLOGY OF THE MIZ‐1/C‐MYC INTERACTION TO ACCELERATE THE DEVELOPMENT OF C‐MYC INHIBITORS

Jean‐Michel Moreau ¹, Martin Montagne¹, Danny Létourneau¹, Mikaël Bédard¹, Pierre Lavigne¹

¹Université de Sherbrooke (Sherbrooke, Canada)

c‐Myc is a b‐HLH‐LZ transcription factor that plays a central role in cellular growth and proliferation by activating and repressing a plethora of genes implicated in ribogenesis, apoptosis and cell cycle. c‐Myc levels are deregulated and increased beyond physiological levels by the translocation of its gene or the activation of driver oncogenes in a wide variety of cancers. Tumor cells overexpressing c‐Myc become addicted to its enhanced transcriptional activity and this addiction has been shown to be their Achilles heel. In fact, inhibition of c‐Myc transcriptional activities with a dominant negative (i.e. Omomyc) in mice models of cancers leads to tumor regression and has established c‐Myc as a prime target for treatment. Currently, many efforts are deployed to develop small molecules that can inhibit c‐Mycs transcriptional activities by preventing the heterodimerization of its b‐HLH‐LZ with that of Max (its obligate partner) and subsequent DNA binding to transcriptional start site. However, no such inhibitors have made it to the clinic. We have recently discovered that a region of Miz‐1 (Myc Interacting Zinc finger protein‐1), dubbed Mid2, can specifically bind the b‐HLH‐LZ c‐Myc and prevent its heterodimerization with Max. In this poster, we show, using Circular Dichroism, solution‐state NMR and EMSA, that a sub‐region of the Mid2, called Mid2sb, forms a more thermodynamically stable complex than the c‐Myc/Max heterodimeric b‐HLH‐LZ and prevents c‐Myc from binding to DNA. Our result supports the notion that the Mid2sb possess structural information that can accelerate the development of a new generation of c‐Myc inhibitors.

21. Structure (x‐ray/NMR/EM): 16. Protein interactions and assemblies

ABS349

PARABS, CHROMOSOME PARTITIONING SYSTEM

Yuh Ju Sun ¹

¹Institute of Bioinformatics and Structural Biology, National Tsing Hua University (Taipei, Taiwan)

ParABS, an important DNA partitioning process in chromosome segregation. ParABS consists of three major components: ParA (an ATPase), ParB (a parS binding protein) and parS (a centromere‐like DNA). The homologous proteins of ParA and ParB in Helicobacter pylori are HpSoj and HpSpo0J, respectively. We determined the crystal structures of HpSoj and HpSpo0J with DNA complexes. In the HpSoj‐DNA complex, HpSoj nonspecifically binds DNA through a continuous basic binding patch formed. In the HpSpo0J‐DNA complex, HpSpo0J folds into an elongated structure with a flexible N‐terminal domain for protein‐protein interaction and a conserved DNA‐binding domain for parS DNA binding. Furthermore, we detected the HpSpo0J‐HpSoj‐DNA complex, the nucleoid adaptor complex (NAC), by electron microscopy. NAC formation is promoted by HpSoj participation and specific parS DNA facilitation.

21. Structure (x‐ray/NMR/EM): 16. Protein interactions and assemblies

ABS350

ELUCIDATING THE ROLE OF PROTEIN PARTNERSHIPS IN MODULATING DNA BINDING SPECIFICITY OF TRANSCRIPTION FACTORS

Bidisha Acharya ¹, Snigdha Maiti¹, Aditya Jyoti Basak¹, Soumya De¹

¹Indian Institute of Technology, Kharagpur (Kharagpur, India)

Transcription factors (TFs) are proteins that bind DNA and control gene expression. Due to their extensive biological roles, they are under tight regulation, loss of which leads to diseases. In Drosophila, TFs of the HOX family regulate the development of different body regions. These TFs have highly conserved DNA binding domains, which recognize very similar DNA sequences in vitro. However, in vivo these proteins have highly specific biological functions, as they bind distinct DNA sequences. This altered specificity is driven by the interactions of the HOX TFs to Extradenticle (EXD), another homeodomain containing TF. In this study, we have investigated the structural basis of the change in DNA binding affinity and specificity of HOX and EXD TFs in the presence of the partner protein.

Sex combs reduced (SCR) and Deformed (DFD) are Drosophila HOX proteins that interact with EXD through a conserved YPWM motif. Our EMSA‐based binding studies show that these HOX proteins have tight nanomolar binding affinity for their respective promoter sequences, which is further enhanced in the presence of EXD. Interestingly, EXD, whose DNA binding homeodomain has strong structural and sequence homology to HOX homeodomain, has weak micromolar affinity for the same sequences. We are using solution NMR spectroscopy to investigate the thousand‐fold difference in the binding affinity of the HOX and EXD homeodomains. We show significant structural changes occur in EXD in the pairwise EXD‐HOX and EXD‐DNA interactions, which have a synergistic effect in the formation of very tight HOX‐DNA‐EXD ternary complex.

21. Structure (x‐ray/NMR/EM): 17. Proteins in cells

ABS351

HOMODIMER INTERFACE MUTATIONS OF HUMAN GALECTIN‐7 ALTER ITS BIOLOGICAL ACTIVITY

Ngoc Thu Hang Pham ¹, Myriam Létourneau¹, Marlène Fortier¹, Carolina Perusquía Hernández¹, Marie‐Aude Pinoteau¹, Jacinthe Gagnon¹, Philippe Egesborg¹, David Chatenet¹, Yves St‐Pierre¹, Charles Calmettes¹, Nicolas Doucet¹

¹Centre Armand‐Frappier Santé Biotechnologie, Institut National de la Recherche Scientifique (INRS), Université du Québec (Laval, Canada)

Human galectins are beta‐galactoside binding proteins subdivided into three groups in accordance with their structural organization: tandem repeat, chimera, and prototype. Among these galectins, prototype galectin‐7 (GAL‐7), characterized by a homodimeric molecular organization of its carbohydrate recognition domain (CRD), is involved in different types of cancer, including carcinomas, lymphomas and melanomas. Its overexpression in tumor cells not only confers resistance to cell death stimuli, but extracellular GAL‐7 also induces apoptosis of lymphocytes, abrogating immune system response against tumor antigens. Consequently, GAL‐7 is a promising target for cancer therapy. To this day, the development of GAL‐7 modulators has almost exclusively focused on small‐molecule Glycan Binding Site (GBS) inhibitors aimed at perturbation of glycoreceptor interactions. However, due to high GBS similarity among different galectin homologs, this remains a high risk strategy because of unwanted off‐target effects on other beneficial anti‐tumor galectins. Furthermore, GBS inhibitors are ineffective at targeting glycan‐independent function of GAL‐7. New approaches are thus required to develop effective and highly specific GAL‐7 inhibitors. Prior structural investigations of ancestral galectins have suggested that stabilization of their oligomeric state through evolutionary pressure improves ligand affinity and biological function. Since destabilization of GAL‐7 architecture could potentially alter its affinity towards glycoproteins and biological function, our main research objective is to dissect the molecular importance of GAL‐7 homodimer formation in celullar function. In this study, we will present the impact of homodimer interface mutations on protein stability and induction of Jurkat T‐cell apoptosis.

21. Structure (x‐ray/NMR/EM): 24. Therapeutics and antibodies

ABS352

THE STRUCTURAL BASIS OF ADHESION REGULATION BY THE CADHERIN‐CATENIN COMPLEX

Allison Maker ¹, Brad Hammerson², David Dranow³, Richard Mangio², Leslayann Schecterson², Bart Staker², Barry Gumbiner¹

¹University of Washington (Seattle, United States); ²Seattle Children's Research Institute (Seattle, United States); ³UCB Pharma (Seattle, United States)

E‐cadherin (CDH1) is a transmembrane receptor that mediates cell‐cell adhesion. Previously, we developed functional antibodies that bind to E‐cadherin and activate its adhesion state. We now seek to structurally characterize the unified cadherin‐catenin‐complex and observe its interaction with functional antibodies. The complex of full‐length human E‐cadherin and human p1204a‐, alpha‐, and beta‐catenin has been reconstituted in glyco‐diosgenin (GDN) detergent and MSP1D1 nanodiscs. We are optimizing the isolated complex in both forms for structural determination with cryo‐electron microscopy (cryo‐EM). Additionally, an activating antibody fragment (Fab), 19A11, was successfully co‐crystallized in complex with the first two extracellular domains of human E‐Cadherin (hEC1‐2). E‐cadherin forms adhesive dimers mediated by the Trp2 residue in the adhesive tip. In our structure, 19A11 binds to EC1, inducing a slight shift inward in Trp2. Interestingly, 19A11 may also block the interface of a hypothesized dimerization intermediate known as the X‐dimer. Ongoing work to discern this activation mechanism involves co‐crystallizing 19A11 with hEC1‐2 mutants that are unable to form fully adhesive dimers but can form X‐dimers. In conclusion, an activating functional antibody to E‐cadherin was crystallized with its epitope region, and the complete cadherin‐catenin complex has been reconstituted. Future work aims to connect these two projects by analyzing functional and neutral antibody binding to the full complex with cryo‐EM and examining structural changes.

21. Structure (x‐ray/NMR/EM): 12. Membrane proteins

ABS353

SOLUTION STATE STRUCTURAL AND DYNAMIC STUDIES OF MOUSE BTNL2, AN ORPHAN T CELL COINHIBITORY MOLECULE

ADITYA JYOTI BASAK ¹, Snigdha Maiti², Anita Hansda³, Dhrubajyoti Mahata², Woonghee Lee⁴, Gayatri Mukherjee³, Soumya De², Dibyendu Samanta²

¹INDIAN INSTITUTE OF TECHNOLOGY KHARAGPUR (KHARAGPUR, India); ²School of Bioscience, Indian Institute of Technology Kharagpur, India (Kharagpur, India); ³School of Medical Science and Technology, Indian Institute of Technology Kharagpur, India (Kharagpur, India); ⁴National Magnetic Resonance Facility At Madison (NMRFAM), and Biochemistry Department, University of Wisconsin‐Madison (Wisconsin Madison, United States)

BTNL2, a T‐cell inhibitory protein belonging to the immunoglobulin (Ig) super family of proteins, modulates regulatory T‐cell differentiation. Widely expressed in gut epithelial tissue, it plays important role in the pathogenesis of various chronic inflammatory diseases. The therapeutic potential of this inhibitory molecule must be further understood by studying the structural basis of such function.

Here we report the solution NMR structure of the N‐terminal IgV domain of mouse BTNL2 (Asp 28 to Ala 143), which shares 65% sequence identity with human BTNL2. This protein was expressed in E. coli as inclusion bodies, which was refolded, purified and found to be functionally active as demonstrated by its binding with CD4+ T cell.

In solution BTNL2 D28‐A143 exists as a monomer, as assessed by size‐exclusion chromatography and NMR dynamics experiments. Also, a well‐dispersed 15N‐HSQC spectrum indicates a properly folded domain. Backbone and side chains were assigned by triple resonance three‐dimensional experiments using 13C, 15N labelled proteins. The NOESY cross‐peaks were assigned by the AUDANA algorithm and structure was calculated by PONDEROSA C/S. A final ensemble of 20 structures has backbone and heavy atom RMSD of 0.45 å and 0.78 å respectively, for all rigid residues. BTNL2D28‐A143 adopts a typical Ig‐fold consisting of a beta sandwich with nine strands arranged in an antiparallel fashion. A cysteine disulfide bond between strands B and F holds the two sheets together. Heteronuclear 15N NOE, R1 and R2 data indicate that the protein is fairly rigid except for loop regions, which show greater flexibility.

18. Proteomics: 8. Enzymology

ABS354

CONTRIBUTION OF CLEAVAGE AND MHCII BINDING EVENTS TO THE GENERATION OF HEMAGGLUTININ IMMUNODOMINANT PEPTIDES

Tynan Becker ¹, Thomas Kuhn²

¹Dept. of Biology and Wildlife University of Alaska Fairbanks (Fairbanks, United States); ²Dept. of Chemistry and Biochemistry University of Alaska Fairbanks (Fairbanks, United States)

Antigen processing leading to the production of immunodominant peptides remains unclear. We examined the contribution of enzymatic cleavage to the generation of peptides by digesting hemagglutinin (HA) from the A/New Caledonia/20/99 H1N1 strain of influenza in vitro using three cysteine proteases.

Enzymatic digestion of HA was performed at pH 6.4 and 5.0 using cathepsins B, H and S at various time points over a 24 hour time period. Each digest was separated using SDS‐PAGE, tryptic in‐gel digestion was performed, and peptide sequences were determined using mass spectrometry (MS). The remainder of the in‐solution cathepsin digests was also sequenced with MS. Data analyses were performed using MaxQuant.

Immunodominant peptides were identified after 24 hours of digestion at both pH 6.4 and 5.0, but differed by pH. The kinetics of the digestion also differed by pH. Early time points at pH 5.0 are similar in peptide coverage as late time points at pH 6.4.

Some immunodominant and subdominant peptides are resistant to enzymatic digestion, but resistant species are pH specific. Digestion at pH 6.4 appeared slower overall than at pH 5.0. This provides confirmation that the kinetics and peptide repertoire change as the endosome matures.

Addition of MHCII to the digestion reaction will allow us to elucidate the contribution of binding events to the origination of immunodominant peptides. MHCII will be allowed to pre‐bind the full HA protein or cleaved peptides. Following binding, the peptide‐MHCII will be immunoprecipitated and eluted peptides sequenced with mass spec.

6. Design/engineering: 5. Computational modeling/simulation

ABS355

HIGH‐THROUGHPUT DE NOVO DESIGN OF STABLE AND HIGH‐AFFINITY BINDERS

Nihal Korkmaz ¹, TJ Brunette¹, David Baker¹

¹Institute for Protein Design, University of Washington (Seattle, United States)

De novo protein design using the Rosetta software has been very successful for creating hyper stable proteins with clefts that can accommodate small peptide therapeutic targets. Although we can successfully design binders for various targets, we get very few binders for each target. To increase both the number of binders and the binding affinities of the binders while preserving/increasing protein stability, we have switched to a high‐throughput workflow.

The pursued massively parallel approach extends from designing the backbones and binding pockets, to manufacturing the synthetic genes and experimental screening. We have integrated large‐scale computational design, oligonucleotide synthesis, yeast display screening and next‐generation sequencing (NGS) to allow this high‐throughput workflow. So far, we designed and tested 60,000 proteins (90‐119 residues long) for stability and binding to various peptide targets.

Using the previously developed protease screening assay, NGS and machine learning techniques, we have identified the common features that contribute the most to protein stability of the libraries ordered. We were also able to evaluate the effect of customizing the interfaces for binding, on the protein stability compared to stable scaffold library. We are aiming to increase both the stability and binding affinities of the binder libraries, as well as the number of binders we get.

21. Structure (x‐ray/NMR/EM): 16. Protein interactions and assemblies

ABS356

THE SUPRAMOLECULAR STRUCTURE OF THE BACTERIAL STRESSOSOME REVEALED BY CRYO‐EM UNVEILS ITS MECHANISM OF ACTIVATION

Allison Williams ¹

¹Institut Pasteur, Department of Microbiology (Paris, France)

The stressosome is the epicenter of the stress response in bacteria, and one of the largest bacterial nanomachines. How the stressosome integrates and transmits stress signals from the environment has remained elusive. The stressosome consists of multiple copies of three proteins RsbR, RsbS and RsbT, a kinase that is important for its activation. Here using cryo‐electron microscopy, we determined the atomic organization of the 1.8 megadalton Listeria monocytogenes stressosome at 3.38å resolution. The structure shows that RsbR and RsbS are organized in a 60 protomers truncated icosahedron. Two phosphorylation sites on RsbR (T175, T209) and one on RsbS (S56) are arranged on a straight line that is interrupted by a 13 amino acid flexible loop in RsbR. RsbR T175 and RsbS S56 are accessible on the surface and are phosphorylated under normal stress conditions. Access to T209 is partially hidden by the RsbR flexible loop, whose open or closed position could modulate stressosome activation. Modification of the flexible loop or of residues involved in the RsbR/RsbS interaction, generates protein assemblies that are toxic when introduced in wild type bacteria, suggesting that a properly assembled stressosome is important for Listeria monocytogenes survival. Our data provide the first atomic model of the stressosome core assembly, and describe the discovery of a loop that is important for stressosome activation, paving the way towards elucidating the structural basis of stressosome function in bacteria.

8. Enzymology: 21. Structure (x‐ray/NMR/EM)

ABS358

IDENTIFICATION AND CHARACTERIZATION OF AN OXALYL COA‐SYNTHETASE FROM GRASS PEA (LATHYRUS SATIVUS L.)

Moshe Goldsmith ¹, Shiri Barad², Orly Dym³, Shira Albeck³, Yoav Peleg³, Ziv Reich²

¹Dept. of Biomolecular Sciences (Rehovot, Israel); ²Department of Biomolecular Sciences, Weizmann Institute of Science, Rehovot, Israel (Rehovot, Israel); ³Department of Life Science Core Facilities, Weizmann Institute of Science, Rehovot, Israel (Rehovot, Israel)

Grass pea (Lathyrus sativus L.) is grown in many parts of the world as an agricultural crop for human and animal consumption. Although it is advantageous as a drought resistant legume crop that can withstand harsh environmental conditions, it also produces a neurotoxic compound, beta‐N‐oxalyl‐L‐alpha,beta‐diaminopropionic acid (beta‐ODAP) that may cause a neurodegenerative syndrome, lathyrism, if consumed as a main diet component. The importance of developing a grass pea cultivar that is defective in the biosynthesis of beta‐ODAP led us to isolate and characterize a key enzyme implicated in its biosynthesis, oxalyl CoA‐synthetase (OCS). The principal activity of OCS is to promote the catabolism of oxalate. Oxalate is a secondary metabolite involved in a number of physiological roles such as the regulation of calcium levels, the prevention of aluminum toxicity and plant defense from insects and herbivores. However, high levels of oxalate are toxic to cells and require tight regulation using catabolic enzymes such as OCS. We cloned the gene encoding OCS from L. sativus (LsOCS), expressed and purified it from E. coli cells, and analyzed its kinetic efficiency with several di‐carboxylic acids. We found its highest CoA‐ligation activity with oxalate, but also promiscuous activities with other acids such as malonate and succinate. In addition, we crystalized the protein in its AMP‐bound form and show that it had adopted a thioester‐forming conformation. Our findings reveal a high degree of similarity in sequence, structure and catalytic efficiency between LsOCS and its homolog from A. thaliana.

21. Structure (x‐ray/NMR/EM): 8. Enzymology

ABS359

CRYO‐EM STRUCTURE OF THE DIHYDROLIPOAMIDE SUCCINYLTRANSFERASE (E2) COMPONENT OF THE HUMAN ALPHA‐KETOGLUTARATE DEHYDROGENASE COMPLEX

Balint Nagy¹, Attila Ambrus ¹, Zsofia Zambo¹, Agnes Hubert¹, Martin Polak², Eszter Szabo¹, Jirí Novacek², Frank Jordan³, Vera Adam‐Vizi¹

¹Department of Medical Biochemistry, Semmelweis University (Budapest, Hungary); ²CEITEC‐Central European Institute of Technology, Masaryk University (Brno, Czech Republic); ³Department of Chemistry, Rutgers University (Newark, United States)

The alpha‐ketoglutarate dehydrogenase complex (KGDHc) plays a key role in the energy balance of cells by catalyzing a rate‐limiting step in the citric acid cycle. The dihydrolipoamide succinyltransferase component (E2) forms the structural core of the enzyme complex also bearing a catalytic function that transfers a succinyl group to Coenzyme A.

We determined the cryo‐EM structure of the human KGDHc‐E2 component. The protein was expressed in a pET‐52b(+)/BL21(DE3) E. coli expression system with a double Strep affinity tag fused to the N‐terminus and purified in a single step by affinity chromatography. The cryo‐EM analysis of the protein complex did not permit examination of the flexible domains of E2, hence the final structure model only represents residues 151 to 386. The resolution of the structure is 2.9 å. The 24 copies of E2 assembled into a highly symmetrical cubic structure, which is built up from trimers. The hKGDHc‐E2 subunit was found to be highly similar in overall structure and active site geometry to both its prokaryotic (E. coli) analogue and the E2 subunit of the human pyruvate dehydrogenase complex. Structure analysis of this component still continues in our laboratory, especially because the hKGDHc‐E1‐E2 subcomplex recently proved to be capable of generating substantial amounts of superoxide, a reactive oxygen species, under certain pathologically relevant in vitro conditions.

6. Design/engineering: 16. Protein interactions and assemblies

ABS360

AUTOMATED DESIGN OF INTERFACE STRUCTURE FOR TARGETED BINDERS

YU ZHAO¹, YU ZHAO ², Gevorg Grigoryan¹

¹Dartmouth college (Hanover, United States); ²Dartmouth (Hanover, United States)

We have developed an array of protein‐design techniques based on the fundamental observation that the 3D protein structure space consists of frequently recurrent local tertiary motifs; we refer to these as TERMs (for tertiary motifs). Excitingly, this knowledge, coupled with appropriate statistical techniques, appears sufficient to enable the robust design of novel structures and sequences. We have established this through a number of design projects, ranging from designing novel binding peptides to de‐novo mini proteins. In this study, we consider the problem of binder designi.e., the task of identifying a novel structural strategy for binding to an arbitrary target and designing a sequence that successfully executes such a binding strategy. Specifically, we consider this application in the context of TERMs and how the apparently digital nature of native protein structure can be used to simplify the task.

8. Enzymology: 21. Structure (x‐ray/NMR/EM)

ABS361

NEW CRISPR/CAS9 CHARACTERIZATION BROADENS THE PROTOSPACER‐ADJACENT MOTIF RECOGNITION

Trung Thach ¹, Nam Hyeong Kim², Junho Hur³, Sang‐Seob Lee⁴, Yong Ho Kim²

¹Sungkyunkwan University (Suwon, South Korea); ²SKKU Advanced Institute of Nanotechnology (SAINT), Sungkyunkwan University (Suwon, South Korea); ³School of Medicine, Kyung Hee University (Seoul, South Korea); ⁴Department of Biological Engineering, Kyonggi University (Suwon, South Korea)

A major limitation to expand RNA‐guided CRISPR/Cas9 toolkit in genome editing is a sequence‐specific recognition of the protospacer adjacent motif (PAM) at the target DNA site by the Cas9 protein. Exploration and characterization of new CRISPR/Cas9 system binding distinct PAM sequences would beneficially expand the biological applications of the toolkit. In this study, we screened and identified a new CRISPR/Cas9, yet previously characterized, from our genome library of non‐pathogen microorganisms. To the end, we reveal a distinct PAM sequence recognition by the Cas9 ortholog. Further biochemical characterization shows that Arg‐1083 and Arg‐1116 residues are key determinants for PAM recognition. Our results collectively provide a powerful platform to discover new CRISPR/Cas9 systems that expands the target space for CRISPR/Cas9 toolkit applications.

ABS363

THE YEAST SUP35 PROTEIN FORMS A LARGE NUMBER OF INFECTIOUS STRUCTURES

Yu‐Wen Huang ¹, Chih‐Yen King¹

¹Institute of Molecular Biology, Academia Sinica (Taipei, Taiwan)

How many different amyloid structures can a protein adopt? Recent high‐resolution studies revealed that a single sequence was capable of forming several distinct structures, but they did not give clear indications about how large a maximal number could be. Here we report the isolation of 23 strains of the [PSI+] prion, the infectious amyloid conformation of the yeast Sup35 protein. Plasmids expressing mutant Sup35 are introduced into yeast to observe strain‐specific phenotype responses. Prion aggregates of different strain type are further distinguished with a set of Sup35‐GFP fusion proteins. We show that each strain is unique and elementary, i.e. not a distinct mixture of some basic structures. By adding prion particles to solutions of a purified Sup35 N‐terminal fragment, specific infectivity is generated de novo for most [PSI+] strains, and unlike the case of mammalian prion strains, the nucleation process does not require accessory conformational cofactors. Different strain types thus must be specified by as many protein‐only structures. Polypeptides in an amyloid fiber adopt a quasi‐planar folding pattern and stack up periodically. Folding in reduced dimensions may permit multiple ways to match hydrophobic patches and opposite charges to generate a large number of stable structures, in great contrast to the compact 3‐dimensional fit observed in a globular domain.

21. Structure (x‐ray/NMR/EM): 25. Transcription/translation/post‐translational modifications

ABS364

EIF2B‐CATALYZED NUCLEOTIDE EXCHANGE AND PHOSPHOREGULATION BY THE INTEGRATED STRESS RESPONSE

Lillian Kenner¹, Lillian Kenner ¹

¹UCSF (san francisco, United States)

The integrated stress response (ISR) tunes the rate of protein synthesis. Control is exerted by phosphorylation of the general translation initiation factor eIF2. eIF2 is a guanosine triphosphatase that becomes activated by eIF2B, a two‐fold symmetric and heterodecameric complex that functions as eIF2s dedicated nucleotide exchange factor. Phosphorylation converts eIF2 from a substrate into an inhibitor of eIF2B. We report cryoelectron microscopy structures of eIF2 bound to eIF2B in the dephosphorylated state. The structures reveal that the eIF2B decamer is a static platform upon which one or two flexible eIF2 trimers bind and align with eIF2Bs bipartite catalytic centers to catalyze nucleotide exchange. Phosphorylation refolds eIF2, allowing it to contact eIF2B at a different interface and, we surmise, thereby sequestering it into a nonproductive complex.

8. Enzymology: 7. Dynamics and allostery

ABS365

THE MOLECULAR MECHANISM OF DISEASE MUTATIONS IN HUMAN GLUTAMINE SYNTHETASE AND COMPENSATORY RESCUE BY SECONDARY MUTATIONS

Erin Thompson ¹, Avi Samelson¹, Martin Kampmann¹, James Fraser¹

¹UCSF (San Francisco, United States)

Protein function depends on dynamically transitioning between folded conformational substates. A prime example of this is the homodecameric metabolic enzyme human glutamine synthetase. Disease mutations in this enzyme lead to a rare recessive disease, glutamine deficiency. Specifically, the R341C disease mutation is hypothesized to be due to local unfolding of a helix necessary for a conformational transition during catalysis. We have assessed the GS disease mutants affect on stability, kinetics, and oligomeric assembly using small angle x‐ray scattering, negative stain electron microscopy, and differential scanning flourimetry and found distinct mechanisms for individual mutants. Interestingly, point mutations that cause disease in humans can be found as the wild type residue in other species. We have identified species where this is true and wanted to test whether human disease mutations can be compensated for in a way not sampled by evolution through a deep mutational scan of human glutamine synthetase. These experiments will define the often elusive mechanism whereby specific defects can often be rescued by the addition of compensatory mutations. Our experimental analysis will reveal how epistasis constrains the conformational ensemble of glutamine synthetase.

5. Computational modeling/simulation: 2. Bioinformatics

ABS366

DYNAMIC DOCKING BETWEEN AN ENZYME AND ITS INHIBITOR USING MULTICANONICAL MD SIMULATIONS

Narutoshi Kamiya ¹, Gert‐Jan Bekker²

¹University of Hyogo (Kobe, Japan); ²Osaka University (Suita, Japan)

We have performed dynamic docking simulations between a flexible receptor and a highly flexible ligand by employing multicanonical molecular dynamics (McMD) [1‐2]. We have applied our method to predict the native binding configuration and sample the intermediary binding structures between the enzyme ‐secretase 1 with a wide binding pocket and its medium‐sized inhibitor 3MR [3]. Representative structures located at free energy minima obtained from McMD were taken and subjected to canonical simulations to refine and validate them, reproducing the native complex structure in high agreement with the experimental data. In addition, the binding free energy was estimated by umbrella sampling (US) simulations along representative pathways obtained from the McMD ensemble, followed by weighted histogram analysis to estimate the affinity, which also reproduced the experimental inhibitory affinity. The sampled ensemble by the US simulations smoothly connected the bound and unbound states, refining the binding pathway while staying true to the McMD ensemble. Interestingly, the loss of interactions between the two molecules along the pathway were clearly shown in the free energy landscape, reiterating the fundamental importance of atomistic interactions to the binding affinity between receptor and drug.

[1] N. Kamiya et al. Proteins 70, 41‐53 (2008).

[2] G.‐J. Bekker et al. J. Chem. Theory Comput. 13, 2389‐2399 (2017).

[3] G.‐J. Bekker et al. J. Phys. Chem. B 123, 2479‐2490 (2019).

18. Proteomics: 2. Bioinformatics

ABS367

CYTOTOXIC ACTIVITY OF NON‐SPECIFIC LIPID TRANSFER PROTEIN (NSLTP) FROM FENNEL (FOENICULUM VULGARE) SEEDS

Mekdes Megeressa ¹, Yamna Khurshid², Aftab Ahmed³

¹Chapman University (Aliso Viejo, United States); ²Chapman University, PhD Scholar (Irvine, United States); ³Chapman University, PhD, Associate Professor (Supervisor) (Irvine, United States)

Fennel (Foeniculum Vulgare), is a flowering medicinal plant that belongs to the family Umbelliferae (Apiaceae). It is native to southern Europe and Mediterranean region with long history of use by humans as a spice, medicine, and fresh vegetable. There are limited scientific studies about the structure and functions of proteins isolated from the seeds and their pharmacological activities. This project explores the complete structure of Non‐Specific Lipid transfer protein (nsLTP) isolated from fennel seeds along with their cytotoxic activities. The protein was extracted using Tris/HCl pH 8 buffer and purified by the combination gel filtration and reverse phase HPLC. The purity was confirmed by SDS‐PAGE gel electrophoresis and intact mass analysis by MALDI‐TOF mass spectrometry. The purified nsLTP‐protein was modified by 4‐vinyl pyridine followed by digestion with trypsin. The tryptic digest was then separated by reversed phase HPLC. The primary structure of nsLTP was established by combining N‐terminal amino acid sequence of intact chain and tryptic peptides by Edman degradation in automated protein sequencing system. The data suggest that Fennel NsLTP is a monomeric protein of 10kDa based on SDS‐PAGE electrophoresis. Multiple sequence alignment performed using Clustal Omega depicted a sequence similarity with Daucus carota and Actinidia chinensis. The MTT assay result revealed that Fennel NsLTP led to dose dependent cytotoxic effect in MCF‐7 cells with an IC50 value of 90.1μg/ml. Besides, mRNA expression of the genes Bax, EGFR, VEGF, Caspase‐3, Survivin, MMP, and Bcl‐2 were modulated as assessed by real time PCR.

4. Chemical biology: 18. Proteomics

ABS369

DIFFERENTIAL LOCALIZATION OF AN ENGINEERED RAS RHEOSTAT REVEALS UNIQUE RAS‐ERK SIGNALING DYNAMICS

Emily M Dieter ¹, John Rose¹, Dustin Maly¹

¹University of Washington (Seattle, United States)

Engineered protein switches that can be controlled through user‐defined inputs are vital tools for understanding and controlling cellular dynamics. Historically, many approaches rely on multiprotein intermolecular regulation, which are only applicable to proteins that can be separated from their site of function. Our lab has computationally designed and developed a genetically encoded RAS rheostat that solely relies on intramolecular allosteric regulation. This system, termed Chemically Inducible Activator of RAS (CIAR), is comprised of the catalytic domain of Son of Sevenless (SOS), a RAS activator, whose catalytic site is gated by BCL‐xL/BH3 protein‐peptide interactions. The BCL‐xL/BH3 interaction can be disrupted using commercially available small‐molecule BCL‐xL inhibitors, enabling tunable activation of endogenous RAS. Using CIAR with an HRAS localization sequence, we demonstrated that focal RAS activation produced linear and sustained RAS/ERK signaling dynamics, a stark contrast to the transient activation yielded through EGF stimulation. These differences extend to different cell types, and were also observed when profiling was done using quantitative phosphoproteomics. Currently we have generated CIAR switches with localization sequences for four RAS isoforms: KRAS4A, KRAS4B, HRAS, and NRAS. Comparison of RAS/ERK signaling dynamics by examining phosphorylation of individual pathway members, as well as looking at global changes through proteomics allows us to explore these systems on both a single cell and population level. Further utilization of the CIAR system will help us to disentangle RAS isoform signaling, resulting in a better understanding of RAS biology.

21. Structure (x‐ray/NMR/EM): 19. Proteostasis and quality control

ABS370

EXPLORING THE NOVEL STRUCTURE OF HUMAN MYELOID‐DERIVED GROWTH FACTOR

Valeriu Bortnov ¹, Marco Tonelli¹, Woonghee Lee¹, John Markley¹, Deane Mosher¹

¹University of Wisconsin‐Madison (Madison, United States)

Human myeloid‐derived growth factor (MYDGF) is the prototype of the widely distributed MYDGF family of proteins found in organisms as distant as single‐celled eukaryotes. Human MYDGF comprises a signal sequence followed by a 142‐residue mature protein ending in a C‐terminus endoplasmic reticulum (ER) retention sequence (ERS), suggestive of functions in the secretory pathway. Interestingly, the only publication that addresses function concluded that MYDGF secreted from monocytes/macrophages promotes tissue repair in a murine model of myocardial infarction and has therapeutic potential. The structure of MYDGF‐family proteins, which shares no similarity to any other protein family, has not been investigated. We now report the structure of 13C,15N‐labeled human MYDGF solved at pH 6 by nuclear magnetic resonance spectroscopy (Figure 1). The structure consists 10 ‐strands forming 3 ‐sheets and a short ‐helix. The two larger ‐sheets, which are held together by a disulfide bridge, are capped at one side by the smaller ‐sheet and ‐helix. The global fold is stable from pH 4.0 to 7.5 as analyzed by circular dichroism spectroscopy. Conserved residues among 87 unique MYDGF homologs map throughout the structure, with greatest conservation in the two cysteines, the ERS, and loops on the face opposite the ERS. This structure can guide future functional studies of the MYDGF family of proteins, including its role as a healing factor in stressed tissues such as infarcted myocardium.

5. Computational modeling/simulation: 6. Design/engineering

ABS371

ENGINEERING ORTHOGONALITY INTO THE CHEMOKINE‐GPCR INTERFACE USING ROSETTA

Michael Wedemeyer ¹, Benjamin K. Mueller², Jens Meiler², Brian Volkman¹

¹Medical College of Wisconsin (Milwaukee, United States); ²Vanderbilt University (Nashville, United States)

It is critical for multicellular organisms to organize and control the migration of its resident cells. In vertebrates this essential function is carried out by chemokines; a family of 50 secreted proteins that provoke directed cell migration through interactions with 20 chemokine receptive G‐protein coupled receptors (GPCRs). The selectively‐promiscuous network of interactions between these proteins play major roles in immune surveillance, inflammation, and cancers, and have proven to be a challenging yet medically important drug target. Recently, the structures of several chemokine‐GPCR complexes have revealed the intimate details underpinning the binding mechanism of three unique chemokine‐GPCR pairs. Though this data covers less than 3% of the known chemokine interaction network, the structural information can be harnessed to give valuable insight into ligand specificity and support structure‐based drug discovery. To achieve this goal, we have developed detailed homology models of chemokine‐GPCR complexes using Rosetta hybridization of template fragments from both NMR and crystallographic structures. Final models were selected according to clustering of root‐mean‐square deviation of atomic positions, Rosetta energy score contributions, and novel spatial metrics, then guided rational mutation of key interacting residues designed to interrupt binding to native proteins yet retain affinity for corresponding engineered proteins. Functional readouts are assessing potential orthogonal pairs, and additional designs can improve selectivity using an iterative process of guided empiricism. Ultimately, this work has improved our understanding of chemokine‐GPCR interactions, developed tools to investigate chemokine specificity, and built toward an orthogonal chemotactic pair that may be used to investigate in vivo chemokine functionality.

21. Structure (x‐ray/NMR/EM): 7. Dynamics and allostery

ABS372

HYDROPHOBIC LIGANDS INFLUENCE THE STRUCTURE, STABILITY, AND PROCESSING OF THE MAJOR COCKROACH ALLERGEN BLA G 1

Alexander Foo ¹, Peter Thompson¹, Lalith Perera¹, Simrat Arora¹, Eugene DeRose¹, Jason Williams¹, Geoffrey Mueller¹

¹National Institute of Environmental Health Sciences (Research Triangle Park, United States)

RATIONAL: The cockroach allergen Bla g 1 forms a novel fold featuring an exceptionally large 3758 å3 hydrophobic cavity enclosed by 12 amphipathic alpha‐helices. This cavity allows Bla g 1 to bind a range of hydrophobic ligands. Given the ability of such ligands to enhance sensitization against numerous other allergens, understanding the structural basis of this interaction and its implications for Bla g 1 structure, stability, and processing could yield insights into the molecular determinants of allergenicity.

METHODS: The effect of various fatty‐acid, and diacyl cargoes on Bla g 1 structure and thermostability was assessed using circular dichroism coupled with solution‐NMR, and molecular modeling, while proteolytic assays were used to probe for changes in its susceptibility to cleavage by the endosomal protease cathepsin S, a key player in the antigen processing pathway.

RESULTS: The Bla g 1 cavity accommodates an exceptionally high occupancy of 8 fatty‐acid chains, binding of which significantly enhances the stability of Bla g 1. This effect was dependent on alkyl chain length, with longer‐chain cargoes providing up to a ~20C increase in melting temperature while simultaneously eliminating cathepsin cleavage; the latter of which was found to correlate to a decreased production of known T‐cell epitopes. Diacyl chain cargoes provided similar enhancements to thermostability, though yielded only a modest (0‐2 fold) reduction in proteolysis.

CONCLUSIONS: Binding of lipid cargoes can enhance the stability and proteolytic resistance of Bla g 1. This may hinder endosomal processing and antigen presentation, skewing the TH1/TH2 response to favour allergy.

6. Design/engineering: 1. Amyloid and aggregation

ABS373

CAPTURING THE STRUCTURAL FLEXIBILITY OF SINGLE‐LAYER BETA‐SHEET WITHIN ISOMORPHOUS CRYSTALS REVEALED BY COMPREHENSIVE STRUCTURE DETERMINATIONS

Koki Makabe ¹, Hideki Fujiwara¹, Kenta Hongo², Yuki Hori¹, Norio Yoshida³

¹Yamagata university (Yonezawa, Japan); ²JAIST (Nomi, Japan); ³Kyushu university (Fukuoka, Japan)

Amyloids are formed from ‐rich self‐assemblies from protein and peptides. Despite its importance, their insoluble and heterogeneous nature hamper the detailed structural analysis. To avoid this problem, we have designed a peptide self‐assembly mimic (PSAM) that consists of a central single‐layer ‐sheet (SLB) capped by terminal domains. We found that a PSAM variant which has a grafting ‐strand of an amyloid forming chameleon sequence showed two distinct structures within isomorphous crystals. Thus, we gained an interest in the structural differences of PSAMs within isomorphous crystals. Herein, we report the structural variations trapped within isomorphous crystals by comprehensive structure determinations. Additionally, we demonstrate the structural plasticity of PSAM‐VLGDV1 via molecular dynamics simulations. Hydration structure analysis revealed that water molecule locates on the ‐sheet surface is important for its plasticity. Our findings suggest that a marginal structural difference can be trapped at the time of crystal core formation, which propagates during crystal growth. Our results suggest that structural plasticity of ‐sheet would play an important role in the macroscopic shape formation of peptide self‐assemblies.

4. Chemical biology: 7. Dynamics and allostery

ABS374

CYSTEINE SCANNING MASS SPECTROMETRY TOWARDS ELUCIDATION OF INTRAMOLECULAR STRUCTURE‐FUNCTION RELATIONSHIPS IN MULTI‐DOMAIN KINASES

Zachary E. Potter ¹, Dr. Dustin Maly¹

¹Department of Chemistry, University of Washington (Seattle, United States)

There are many well‐documented examples of multi‐domain kinases with regulatory domains that exert allosteric control over kinase activity by disrupting the alignment of key active site residues. Despite several decades of biochemical and structural characterization of Src Family Kinases and in particular, Src kinase, details about the structural and functional relationships between Srcs N‐terminal SH4 domain and catalytic domain remain unclear. Recent evidence suggests that the SH4 domain contributes to the auto‐inhibition of Src kinase activity by making physical contact with the catalytic domain at an orphan ligand binding pocket called the F pocket. This intramolecular protein‐protein interaction contributes to a closed global conformation equilibrium and diminished catalytic activity. To help probe this interaction more directly, and to study the intramolecular protein‐protein interactions of Src in general, we have developed a cysteine scanning method coupled with mass spectrometry. A library of 82 individual cysteine point mutants were made in full‐length Src constructs to generate a pool of protein for downstream experiments. A complementary cleavable cysteine reactive small molecule was employed to enrich mutant cysteines. We conclude that this method could be applied to study the intramolecular allosteric regulatory mechanisms of other multi‐domain kinases.

18. Proteomics: 2. Bioinformatics

ABS375

PROTEOMIC AND CYTOTOXIC CHARACTERIZATION OF PROTEINS FROM CUSCUTA (DODDER) TENDRILS

Umaima Akhtar ¹, Mekdes Megeressa², Basir Syed², Ishtiaq Ahmed Khan³, Keykavous Parang², Aftab Ahmed⁴

¹Chapman University (Irvine, United States); ²Department of Biomedical and Pharmaceutical Sciences, Chapman University, School of Pharmacy, Irvine, CA 92618 (Irvine, United States); ³Panjwani Centre for Molecular Medicine and Drug Research, International Centre for Chemical and Biological Sciences, University of Karachi, Pakistan (Karachi, Pakistan); ⁴PhD Associate Professor (Supervisor) Department of Biomedical and Pharmaceutical Sciences, Chapman University, School of Pharmacy, Irvine, CA 92618 (Irvine, United States)

Cuscuta species (Dodders) are parasitic plants, which affect a wide range of hosts. Cuscuta upon attachment to host binds to its vasculature and withdraw water, nutrients, and macromolecule such as protein, carbohydrates and lipids from their host. These plants contain different chemical constituents such as flavonoid, terpenoids, fatty acids, alkaloids, and carbohydrates. Phytochemicals from Cuscuta were reported to possess hepatoprotective, antimicrobial, antiviral, and anxiolytic activities. In this study, we aim to extract proteins from Cuscuta tendrils for comprehensive identification of proteins and investigate their anticancer potential. Cuscuta tendrils were grinded and defatted with hexane. Proteins were extracted in 25mM Sodium phosphate pH6.5 and precipitated in 80% ammonium sulfate. The gel filtration chromatography was performed on a Superdex 75 column. The crude protein and eluted fractions were analyzed by electrophoresis using 12% Tris/tricine gels. The crude protein was modified with carboimidomethyl and digested with TPCK treated trypsin. Mass spectrometry analysis of tryptic digest was performed by QTOF LC‐MS/MS and data generated was searched against NCBInr database using PEAKS‐studio software. The crude protein extract and fractions were assessed for their anticancer potential on MCF‐7 breast cancer cell line. The mass spectrometry data by Peaks Studio‐X revealed the presence of 561 proteins belonging to 96 proteins groups. Cuscuta crude protein extract and fractions F‐1 and F3 showed 70%, 77% and 78% percent inhibition respectively against MCF‐7 cell lines at 100μg/ml. This study will increase knowledge about Cuscuta proteome and will be utilized for further study on the mechanistic effect of active proteins.

4. Chemical biology: 18. Proteomics

ABS376

A CHEMOPROTEOMIC METHOD FOR CHARACTERIZING KINASE COMPLEXES

Linglan Fang ¹

¹University of Washington (Seattle, United States)

Small molecule inhibitors often only block a subset of the cellular functions of their protein targets and, in many cases, lead to phenotypic effects that are not well understood. This is especially true for inhibitors of multi‐domain kinases, which possess numerous interaction sites and have been demonstrated to take part in important phosphotransferase‐independent functions in the cell. To systematically characterize how inhibited protein kinases influence cellular behavior, we developed a chemoproteomic method for interrogating the cellular localization and interactomes of inhibitor‐bound kinases. By developing a set of selective inhibitors of Src kinase that contain a trans‐cyclooctene (TCO) click handle, we leveraged this moietys rapid and mild reaction with tetrazines to enrich and characterize the proteins bound to this multi‐domain signaling enzyme. We show that Srcs cellular interactions are highly dependent on the intermolecular accessibility of its regulatory domains, which can be allosterically modulated through its ATP‐binding site with inhibitors. Furthermore, we find that the signaling status of the cell also largely affects Srcs interactome. Finally, we demonstrate that our TCO‐conjugated probes can also be used as a part of a proximity ligation assay to study Srcs localization and interactions in situ. Together, our chemoproteomic platform represents a comprehensive method for studying the localization and interactomes of inhibitor‐bound kinases and potentially other druggable protein targets.

7. Dynamics and allostery: 21. Structure (x‐ray/NMR/EM)

ABS377

OPTOALLOSTERY: AN EXPERIMENTAL STUDY OF THE MECHANISM OF LIGHT‐INDUCED ALLOSTERIC CONTROL OF ENGINEERED RHO FAMILY GTPASES

Abha Jain ¹, Nikolay Dokholyan², Andrew Lee³

¹University of North Carolina at Chapel Hill (Chapel Hill, United States); ²United States Departments of Pharmacology and Biochemistry and Molecular Biology, Penn State College of Medicine, PA 17033 (Hershey, United States); ³Division of Chemical Biology and Medicinal Chemistry, UNC Eshelman School of Pharmacy, University of North Carolina at Chapel Hill, Chapel Hill, NC (Chapel Hill, United States)

GTPases play a vital role in many cellular signaling processes. Malfunction of their activity can lead to uncontrolled signaling, resulting in tumorigenesis, neurodegenerative diseases, and cirrhosis, etc. In previous work by Dagliyan et. al. (Science, 2016, 354, 1441), GTPases were engineered to respond allosterically to light via insertion of a LOV domain. Upon blue light illumination, the LOV domain undergoes a large structural change that is thought to be translated to the GTPase, resulting in light dependent signaling. To better understand the allosteric mechanism, we are using structural and biochemical approaches to characterize how these proteins respond to light stimulation. Specifically, NMR spectroscopy will be used for detecting differences in structure and dynamics for dark and lit states, and an in vitro binding assay will be used to quantify allosteric control. Currently, work is on two light controlled GTPases, Cdc42 and Rac1, and TROSY‐HSQC spectra of the wild type, as well as for the constructs that mimic dark and lit states, have been collected. The WT spectra exhibit good chemical shift dispersion indicative of stably folded proteins, and dramatic differences are observed between dark and lit states. To better characterize allosteric function, an in vitro assay was developed by measuring binding affinity with the effector protein, PAK1. Preliminary ITC data show stronger binding by WT (or dark) relative to the lit states. Finally, we anticipate making chemical shift assignments of the backbone amides, that will enable comparison of dark vs. lit states, providing detailed information about the light‐induced allosteric mechanism.

21. Structure (x‐ray/NMR/EM): 26. Other

ABS378

MOLECULAR MECHANISMS GIVING RISE TO HUMAN DIHYDROLIPOAMIDE DEHYDROGENASE DEFICIENCY ‐ STRUCTURAL ANALYSIS OF SEVEN DISEASE‐RELEVANT ENZYME VARIANTS

Eszter Szabó ¹, Piotr Wilk², Bálint Nagy¹, Réka Mizsei¹, Zsófia Zámbó¹, Dávid Bui¹, Andrzej Weichsel³, Palaniappa Arjunan⁴, Beáta Törőcsik¹, Ágnes Hubert¹, William Furey⁴, William Monfort³, Frank Jordan⁵, Manfred Weiss², Vera Ádám‐Vizi¹, Attila Ambrus¹

¹Department of Medical Biochemistry, MTA‐SE Laboratory for Neurobiochemistry, Semmelweis University (Budapest, Hungary); ²Macromolecular Crystallography, Helmholtz‐Zentrum Berlin (Berlin, Germany); ³Department of Chemistry and Biochemistry, University of Arizona (Tucson, United States); ⁴Department of Pharmacology and Chemical Biology, University of Pittsburgh; Biocrystallography Laboratory, Veterans Affairs Medical Center (Pittsburgh, United States); ⁵Department of Chemistry, Rutgers, The State University of New Jersey (Newark, United States)

Dihydrolipoamide dehydrogenase (LADH, E3) is a homodimeric flavin‐disulfide oxidoreductase that catalyzes the oxidation of dihydrolipoamide using NAD+ as a co‐substrate. As a common third subunit to the mitochondrial alpha‐keto acid dehydrogenase complexes and part of the glycine cleavage system, human (h) E3 plays pivotal roles in metabolism. Pathogenic mutations to the hE3 coding DLD gene result in inactive or partially inactive protein variants affecting several central metabolic pathways simultaneously and leading to hE3 deficiency. Clinical manifestations of hE3 deficiency are versatile, generally very severe, and not correlating well with the loss in hE3 activity or the localization of the pathogenic substitution. This implies that other auxiliary biochemical mechanisms, presumably the elevated reactive oxygen species (ROS) generating activities of certain pathogenic variants and/or impaired interactions among the subunits of the relevant multienzyme complexes, might also contribute to the pathogenesis.

We determined the high‐resolution crystal structures of the wild type hE3 and seven of its disease‐causing variants at resolutions ranging from 1.44 to 2.34 å to analyze the structural bases of hE3 deficiency. The investigated substitutions reside in either the active site (P453L), the cofactor‐binding region (G194C), or the dimer interface (G426E, D444V, I445M, R447G, and R460G). Based on the analyses of the crystal structures, we present the first comprehensive study that links the three‐dimensional structure of hE3 as well as the residual LADH activities, altered capacities for ROS generation, and compromised affinities for multienzyme complexes upon pathogenic mutations to the severity of the clinical symptoms in hE3 deficiency.

ABS379

CHARACTERIZATION OF PROTEINS BY MICROFLUIDIC CE‐SDS

April Blodgett ¹

¹PerkinElmer (Hopkinton, United States)

Microfluidic capillary electrophoresis (CE‐SDS) is a routinely used method for monitoring and documenting critical quality attributes of biotherapeutic agents due to its automated operation, on‐column detection, great resolving power, protein quantification capability, and speed of analysis. CE‐SDS offers an automated alternative to traditional SDS‐PAGE by streamlining the multiple, manual steps of slab gel electrophoresis while providing the throughput and data quality essential to biotherapeutic workflows. Such characterization can be used upstream to monitor protein expression during cell line development, downstream to assess protein recovery and purity during process development, during formulation and forced degradation studies, and again during manufacturing as a QC/release assay under 21 CFR 11 compliance. Here, the characterization of a reference monoclonal antibody from the National Institute of Standards and Technology (NIST) using the LabChip® GXII Touch protein characterization system is compared to the characterization published by NIST using the SCIEX® PA‐800 Plus system.

ABS380

BIOCHEMICAL, BIOPHYSICAL AND STRUCTURAL CHARACTERIZATION OF ISONIAZID RESISTANCE KATG VARIANTS FROM MYCOBACTERIUM TUBERCULOSIS

Brenda Uribe ¹, Xavier Soberon¹, Humberto Flores¹

¹Biotechnology Institute, UNAM (Cuernavaca, Mexico)

Introduction. Tuberculosis (TB) is an infectious disease caused by Mycobacterium tuberculosis bacteria. Isoniazid (INH) is the most used antibiotic to treat TB world‐wide. INH is a pro‐drug, converted inside the cell by KatG into a free radical, then it binds to NAD+ forming the INH‐NAD adduct. INH‐NAD binds to the InhA enzyme inhibiting its activity (cell wall biosynthesis). INH‐resistance is the result of mutations that alter the function of KatG protein, impairing the adduct formation while preserving functionality that is necessary for cell survival (~70% of resistant isolates have the S315T mutation).

The objective of this study is the characterization of new and known KatG variants, found in INH‐resistance samples.

Methodology. The genes of selected variants were constructed by site‐directed mutagenesis, a 6‐histidine tag was added at the amino‐terminus. Proteins were purified by Ni‐NTA affinity column. The catalase, peroxidase and INH‐NAD formation activities were measured. UV‐Visible spectra were routinely recorded for all variants at room temperature. To monitor de thermal unfolding of WT and S315T, they were incubated with temperatures increasing from 20 to 90°C and the Circular Dichroism spectrum were measured.

Results. Plasmid pKK‐KatG allowed high levels expression of soluble recombinant proteins. Enzymatic assays confirmed that the variants have reduced adduct formation compared with the KatG‐WT; this is according to the observed phenotype resistance (the KatG‐L333V variant is an exception). WT and S315T variants showed similar Circular Dichroism spectrum indicating that the secondary structure is not lost at high temperatures, with different behaviors after refolding (activity and heme content).

Conclusions. A KatG expression system in E. coli was constructed. Mutant S315T production of adduct is lower than WT, as previously reported. Variants have lower INH‐NAD formation activity compared to the WT, which also correlates with the reported resistance. We hypothesize that the stability effect of the variants may have implications in INH‐resistance.

ABS382

INCREASING USER CAPABILITIES AT THE GM/CA@APS X‐RAY CRYSTALLOGRAPHY USER FACILITY AT THE ADVANCED PHOTON SOURCE

Michael Becker ¹, Stephen Corcoran¹, Dale Ferguson¹, Mark Hilgart¹, David Kissick¹, Oleg Makarov¹, Craig Ogata¹, Ruslan Sanishvili¹, Sergey Stepanov¹, Nagarajan Venugopalan¹, Qingping Xu¹, Shenglan Xu¹, Robert Fischetti¹, Janet Smith²

¹Argonne National Laboratory (Lemont, United States); ²University of Michigan (Ann Arbor, United States)

The General Medical Sciences and Cancer Institutes structural biology facility at the Advanced Photon Source (GM/CA @ APS) operates a national user facility for crystallographic structure determination of biological macromolecules, including two undulator beamlines ‐ beamline 23ID‐B and beamline 23ID‐D ‐ with an emphasis on challenging, high‐impact projects. GM/CA beamlines provide stable, intense X‐ray beams of user‐selectable size down to 5‐micron diameter, an intuitive user interface for experiment control, and an automated pipeline for data processing. This poster gives an overview of recent developments, including advances in the implementation of SONICC (Second Order Nonlinear Imaging of Chiral Crystals) for visualizing challenging crystals, such as in mesophase, and advances in the implementation of serial crystallography experiments, including proof‐of‐principle experiments. Researchers who seek to pursue novel structures via serial crystallography with viscous‐jet (RFF) or fixed‐target (CMO) mounting systems at GM/CA@APS are encouraged to contact indicated staff members to help enable their experiments. These efforts align well with long‐term aims relevant to the planned APS upgrade (APS‐U), which will also be presented.

Acknowledgment: GM/CA@APS has been funded in whole or in part with Federal funds from the National Cancer Institute (ACB‐12002) and the National Institute of General Medical Sciences (AGM‐12006). This research used resources of the Advanced Photon Source, a U.S. Department of Energy (DOE) Office of Science User Facility operated for the DOE Office of Science by Argonne National Laboratory under Contract No. DE‐AC02‐06CH11357. The Eiger 16M detector was funded by an NIHOffice of Research Infrastructure Programs, High‐End Instrumentation Grant (1S10OD012289‐01A1).

ABS383

A PSR DOMAIN IN ATG32 IS REQUIRED FOR MITOPHAGY

Xue Xia ¹, Sarah Katzenell¹, Erin Reinhart¹, Katherine Bauer¹, Maria Pellegrini¹, Michael Ragusa¹

¹Dartmouth College (Hanover, United States)

Mitochondria are targeted for degradation by mitophagy, a selective form of autophagy. In Saccharomyces cerevisiae, mitophagy is dependent on the autophagy receptor, Atg32, an outer mitochondrial membrane protein. Once activated, Atg32 recruits the autophagy machinery to mitochondria, facilitating mitochondrial capture in phagophores, the precursors to autophagosomes. However, the mechanism of Atg32 activation remains poorly understood. To investigate this crucial step in mitophagy regulation, we examined the structure of Atg32. We have identified a structured domain in Atg32 that is essential for the initiation of mitophagy, as it is required for the proteolysis of the C‐terminal domain of Atg32 and the subsequent recruitment of Atg11. The solution structure of this domain was determined by NMR spectroscopy, revealing that Atg32 contains a previously undescribed pseudo‐receiver (PsR) domain. Our data suggest that the PsR domain of Atg32regulates Atg32 activation and the initiation of mitophagy.

ABS384

STRUCTURAL ANALYSIS OF CYTOCHROME BM3 FOR SYNTHESIS OF CAFFEIC ACID

Jorge Jimenez Niebla ¹, Gloria Saab Rincón²

¹IBT, UNAM (Cuernavaca, Mexico); ²Instituto de Biotecnología, U.N.A.M. (Cuernavaca, Mexico)

Caffeic acid is an important metabolite in plants that is part of the phenylpropanoid pathway. Its important applications in the pharmaceutical industry and its rather low content in plant tissues has boosted the search of suitable ways to obtain this molecule. Its precursor, p‐coumaric acid is a widely distributed compound in plant tissues as such or as conjugated form. It is a relatively inexpensive reagent that is commercially available. The conversion of p‐coumaric acid to caffeic acid depends on a hydroxylation of the aromatic ring at C3; This reaction is catalyzed by cytochrome p450 in plants: However, eukaryotic cytochromes are anchored to a membrane, so their expression in other organisms is difficult, besides they depend on the activity of a reductase to regenerate its catalytic site.

Cytochrome P450 from Bacillus megaterium (CYT BM3) is a soluble and self‐sufficient bacterial cytochrome that uses straight‐chain fatty acids as substrates. CYT BM3 has been widely used as a biocatalyst for a wide variety of substrates with different chemical properties; however, it is not able to use small organic molecules naturally. We seek to create an intelligent library based on structural analysis and bioinformatic tools to generate a limited number of variants and a high throughput screening method in conjunction with techniques of HPLC and RMN to select and characterize variants capable to form caffeic acid from p‐coumaric acid.

ABS385

INFLUENCE OF MECHANICAL FORCE ON THE LIFETIME OF ACTIVATED FIMH‐MANNOSE BONDS

Laura Carlucci ¹, Wendy Thomas¹

¹University of Washington, Department of Bioengineering (Seattle, United States)

Adhesive interactions must withstand mechanical forces occurring in an organism. Forces have influenced some proteins such as integrins, clotting factors, and bacterial adhesive proteins to develop catch bond properties, in which the lifetime between a biomolecule and ligand increases with applied force. One of the most well studied catch bond behaviors is found between the E. coli adhesive protein, FimH, and its ligand, mannose. FimH has two distinct conformations with significantly different mannose affinities. The low affinity conformation predominates the initial FimH‐mannose interactions, as this state, characterized with an open binding pocket, facilitates fast kinetics. A force pulling on the FimH‐mannose interaction elongates FimH, inducing a transition into the high mannose affinity state, which includes a closed mannose binding pocket.

While in this force‐stabilized high affinity conformation, FimH has a remarkably long‐lived lifetime that has never been fully measured before. We suspect the closed binding pocket is responsible for this lifetime, and mannose dissociation occurs when the pocket spontaneously opens. If the opening of this pocket occurs independent of the strength of the applied force, mannose unbinding would thus represent a force independent, or ideal bond, interaction. To explore this possibility, we measured the lifetimes of the mannose‐FimH interaction at various forces with FimH in the high affinity state.

We utilized a magnetic tweezer apparatus, to apply various constant forces to mannose‐coated magnetic beads bound to immobilized FimH. While the FimH‐mannose bond does not appear force‐independent, the interaction does not fit a classical force‐dependent interaction. We suspect fluctuations in the ligand‐binding pocket result in a combination of force‐dependent and force‐independent behavior. Understanding the relationship between force and FimHs incredibly stable binding pocket, would grant further insight into the catch bond mechanism associated with FimH.

ABS386

CHARACTERIZATION OF E. COLI LPXA INHIBITORS TARGETING VARIOUS ENZYMATIC STATES BY NMR SPECTROSCOPY

Feng Wang ¹, Andreas Frank¹, Andreas Lingel¹, Alexandra Frommlet¹, Wooseok Han¹, Xiaolei Ma¹, Alun Bermingham¹, Barbara Chie Leon¹, Chi‐Min Ho¹, Patrick Lee¹, Min Li¹, Jacob Shaul¹, Charles Wartchow¹, Tsuyoshi Uehara¹

¹Novartis Institutes for BioMedical Research (Emeryville, United States)

LpxA that catalyzes the first step of lipopolysaccharide biosynthesis is conserved in Gram‐negative bacteria and essential for the bacterial growth and structural integrity, and thus, it is an attractive therapeutic target for multi‐drug resistance mechanisms in Gramnegative pathogens. Escherichia coli LpxA transfers R‐3‐hydroxymyristate from the acyl carrier protein to UDP‐N‐acetylglucosamine. The catalytic process of LpxA involves multiple states and the association with the substrates and products. At Novartis, small molecular LpxA inhibitors were identified from phenotypic screens. Using NMR spectroscopy and the MILVAT selectively labeled protein, we can monitor different states of this large homotrimer enzyme in action. Our NMR characterization shows that three compounds possess distinct mechanisms of inhibition, targeting different dynamic states of LpxA. Three compounds bind and inhibit the apo, product‐bound, and ACP‐bound LpxA respectively. Two co‐crystal structures of these compounds had been solved, and structure‐based hit‐to‐lead optimizations were conducted to improve compounds targeting the more desirable mechanism of action.

ABS387

INCREASED ANTIGEN BINDING AFFINITY AND DECREASED THERMAL STABILITY OF AN ANTI‐(4‐HYDROXY‐3‐NITROPHENYL)ACETYL ANTIBODY POSSESSING A GLYCINE RESIDUE AT POSITION 95 OF THE HEAVY CHAIN

Masayuki Oda ¹, Takachika Azuma²

¹Kyoto Prefectural University (Kyoto, Japan); ²Antibody Engineering Research Center (Noda, Japan)

Anti‐(4‐hydroxy‐3‐nitrophenyl)acetyl (NP) antibodies could be classified into at least two groups, Tyr95H‐ and Gly95H‐type, based on the amino acid residue at position 95 of the heavy chain, which corresponds to the junction of VH and D segments. The Gly95H‐type antibodies appeared at an early stage of immunization, while the Gly95H‐type antibodies appeared at a late stage. In this study, we selected two representative anti‐NP antibodies, N1G9 (Tyr95H‐type) and C6 (Gly95H‐type) and prepared single‐chain Fv (scFv) antibodies, N1G9‐scFv and C6‐scFv, respectively. The antigen binding affinities of N1G9‐scFv and C6‐scFv analyzed using surface plasmon resonance (SPR) biosensor and isothermal titration calorimetry (ITC) were comparable to those of parent N1G9 and C6 antibodies, respectively. The antigen binding affinity of C6‐scFv was about 40‐times higher than that of N1G9‐scFv. The thermal stabilities of N1G9‐scFv and C6‐scFv in the absence or presence of antigen were analyzed using circular dichroism (CD) and differential scanning calorimetry (DSC). The apparent melting temperature of N1G9‐scFv was 66°C, while that of C6‐scFv was 49°C, indicating that the latter was highly unstable. However, C6‐scFv greatly gained thermal stability upon the antigen binding. The difference in stability can explain the prevalence of Tyr95H‐type at an early stage of immunization. At a late stage of immunization, under conditions of antigen excess, B‐cells possessing B‐cell receptors of Gly95H‐type could gain stability by antigen binding and differentiate into plasma cells that secreted high affinity Gly95H‐type antibodies.

ABS388

CHARACTERIZING THE INTRINSICALLY DISORDERED DOMAIN OF LIAT1

Akshaya Arva ¹, Christopher Brower¹, Yasar Kasu¹

¹Texas Woman's University (Denton, United States)

Liquid phase separation facilitates the formation of membrane‐less intracellular compartments that function to regulate enzymatic reactions, particularly those involving RNA. Some well‐known examples of compartments formed by phase separation include cytosolic stress granules and P‐bodies as well as nuclear Cajal bodies and the nucleolus. Phase separation is driven by proteins harboring intrinsically disordered domains that are often interspersed with regions of low sequence complexity. The Ligand of ATE1 (LIAT1) protein contains an evolutionarily conserved central domain through which it interacts with Arginyltransferase 1 (ATE1), an enzyme involved in the N‐degron pathway of protein degradation. Here we show that the N‐terminal half of LIAT1 is intrinsically disordered, and contains conserved regions of low complexity. Using bimolecular fluorescence complementation and immunocytochemistry, we found that LIAT1 self‐assembles both in the cytosol and in the nucleus. Reminiscent of proteins involved in phase separation, LIAT1 participates in the nucleolus and forms cytosolic granular structures with ATE1. We also identified sequence determinants within LIAT1s intrinsically disordered domain important for these activities. In all, this study supports a role for LIAT1 in liquid‐phase separation both in the nucleolus and in the cytosol.

ABS389

NOVEL ALPHA‐SHEET SECONDARY STRUCTURE DRIVES AGGREGATION AND TOXICITY IN ALZHEIMER'S DISEASE

Dylan Shea ¹, Valerie Daggett²

¹University of Washington (Seatte, United States); ²PI (Seattle, United States)

Alzheimers disease (AD) is characterized by the deposition of ‐sheetrich, insoluble amyloid ‐peptide (A) plaques; however, plaque burden is not correlated with cognitive impairment in AD patients; instead, it is correlated with the presence of toxic soluble oligomers. Here, we show, by a variety of different techniques, that these A oligomers adopt a nonstandard secondary structure, termed ‐sheet. These oligomers form in the lag phase of aggregation, when A‐associated cytotoxicity peaks, en route to forming nontoxic ‐sheet fibrils. De novo‐designed ‐sheet peptides specifically and tightly bind the toxic oligomers over monomeric and fibrillar forms of A, leading to inhibition of aggregation in vitro and neurotoxicity in neuroblastoma cells. Based on this specific binding, a soluble oligomer‐binding assay (SOBA) was developed as an indirect probe of ‐sheet content. Combined SOBA and toxicity experiments demonstrate a strong correlation between ‐sheet content and toxicity. The designed ‐sheet peptides are also active in vivo where they inhibit A‐induced paralysis in a transgenic A Caenorhabditis elegans model and specifically target and clear soluble, toxic oligomers in a transgenic APPsw mouse model. The ‐sheet hypothesis has profound implications for further understanding the mechanism behind AD pathogenesis.

ABS390

SOLUTION STRUCTURE OF THE IWP‐051‐BOUND H‐NOX FROM SHEWANELLA WOODYI REVEALS A CONSERVED BINDING POCKET FOR SOLUBLE GUANYLYL CYCLASE STIMULATORS

Cheng‐Yu Chen ¹, Woonghee Lee², William Montfort¹

¹University of Arizona, Department of Chemistry and Biochemistry (Tucson, United States); ²National Magnetic Resonance Facility at Madison, Biochemistry Department, University of Wisconsin‐Madison (Madison, United States)

Soluble guanylyl cyclase (sGC) plays key roles in regulating nitric oxide signaling. Unlike traditional therapy using NO donors, the sGC stimulators stimulate cyclase activity up to several hundred fold without having side effects such as tolerance or NO toxicity. Although having promising clinical results, the molecular mechanisms on how and where the stimulators bind to sGC remain unclear. Here by using solution NMR spectroscopy, we determined the structures of the native and IWP‐051‐bound H‐NOX from Shewanella woodyi. The IWP‐051‐bound H‐NOX structure provides the first structure evidence for the stimulator binding to the H‐NOX domains. By comparing the native and the IWP‐051‐bound H‐NOX structures, we demonstrated a conformational change on the distal subdomain upon binding of the stimulator. We also showed that the binding site pockets are conserved in sequence and share similar surface properties among bacteria H‐NOX proteins. Finally, we demonstrated protein backbone dynamics change at or near the binding site residues upon binding of the stimulator. Together these data provide possible mechanism on how the stimulators stimulate sGC activity and help future drug development.

ABS391

FUNCTIONAL EVOLUTION OF TRPM2 CHANNELS

Iordan Iordanov ¹, Balázs Tóth¹, Andras Szollosi¹, László Csanády¹

¹Department of Medical Biochemistry and MTA‐SE Lendület Ion Channel Research Group, Semmelweis University (Budapest, Hungary)

The ion channel Transient Receptor Potential Melastatin 2 (TRPM2), a member of the family of TRP proteins, is found encoded in the genomes of various organisms, with evolutionary span from colonial single‐cell eukaryotes to mammals. In humans, TRPM2 (hsTRPM2) plays a role in the immune response and insulin secretion, and is activated by cytosolic ADP‐ribose (ADPR), Ca2+ and phosphatidylinositol bisphosphate (PIP2). The channel contains a nudix‐type motif 9 (NUDT9)‐homology (NUDT9‐H) domain similar to ADPR phosphohydrolases (ADPRases), such as the soluble monomeric enzyme NUDT9 in mitochondria. NUDT9 splits ADPR into AMP and ribose‐5‐phosphate, and it was suspected that the putative enzymatic activity of the NUDT9‐H domain of hsTRPM2 may be strictly or loosely coupled to channel gating. It was later found that most likely ADPR gates TRPM2 via a different binding site, and that the NUDT9‐H domain is catalytically inactive due to mutations in the conserved Nudix box sequence. However, multiple sequence alignment suggested enzymatically active (canonical) Nudix motifs in TRPM2 channels of invertebrates and inactive (vestigial) motifs in vertebrates. For example, TRPM2 of the choanoflagellate Salpingoeca rosetta (srTRPM2) and the cnidarian Nematostella vectensis (nvTRPM2) are active ADPRases. Critical mutations that inactivate the enzymatic activity of nvTRPM2 do not affect its channel currents, suggesting that catalysis is uncoupled from gating. In addition, variations in the channels pore sequence responsible for irreversible inactivation (rundown) of hsTRPM2 also appeared in vertebrates. For example, TRPM2 of Danio rerio (zebrafish, drTRPM2) and hsTRPM2 channels inactivate, but srTRPM2 and nvTRPM2 currents are stable. Therefore, enzymatic activity and pore stability were simultaneously lost in vertebrate TRPM2 channels, possibly substituting ligand hydrolysis with channel rundown as the main mechanism for channel inactivation.

ABS392

THE EFFECTS OF A SMALL MOLECULE INHIBITOR ON CDC42, ITS MUTANT AND ITS INTERACTION WITH EFFECTOR PROTEINS

Djamali Muhoza ¹, Paul Adams¹

¹University of Arkansas (Fayetteville, United States)

The Ras superfamily of G‐proteins is important in cell proliferation, inhibition of cell death, and cell transformation. Abnormal expression and function of these proteins is observed in many diseases including several forms of cancer. Cdc42 is a Ras related protein involved in cell adhesion and cell cycle regulation. Ras protein abnormalities have proven difficult to target and inhibit but recent studies have highlighted potential inhibitors for Cdc42. Studies have found small molecules and miRNAs that can reverse the effects of Cdc42 in cells. However, much is still not understood about the effects of these inhibitors on the protein structure, stability and function. In this study, a novel Cdc42 inhibitor, ZCL278, was investigated. The binding affinity of ZCL278 to Cdc42 was calculated using different experimental methods. Effects of the ZCL278 on protein conformation and stability of the protein were also studied. A Cdc42 mutant in the region thought to bind the inhibitor was used for comparison. Results from this study show a considerable binding affinity of the drug to the protein. They show that ZCL278 makes the protein more stable and resistant to denaturation. Potential amino acids involved in the binding were also identified. These studies further highlight ZCL278 as a potential Cdc42 inhibitor that can be used to block Cdc42 abnormal activity and its effects in the cell.

ABS393

SEQUENCE SPECIFICITY FOR PEPTIDE SUBSTRATES IN THIOETHER CROSSLINKING REACTION CATALYZED BY RADICAL SAM ENZYME QHPD

Toshinori Oozeki ¹, Kazuki Kozakai¹, Tadashi Nakai², Katsuyuki Tanizawa¹, Toshihide Okajima¹

¹Osaka university (Ibaraki, Japan); ²Hiroshima Institute of Technology (Hiroshima, Japan)

QhpD is a radical S‐adenosyl‐L‐methionine (SAM) enzyme that involved in posttranslational modification in the active‐site subunit QhpC of quinohemo protein amine dehydrogenase [1]. QhpD catalyzes the formation of three intrapeptidyl thioether bonds between the SH group of a Cys residue and the methylene carbon atom of a Glu or Asp residue within a proteinous substrate, QhpC. In this study, using QhpD and QhpC from Paracoccus denitrificans, we elucidated the ability of QhpD to form interpeptidyl crosslinking in peptide substrates. Especially we focus the sequence dependency of the loop‐out region to be formed in QhpC. Although Cys and Asp/Glu residues of the crosslinking loop constitute the thioether bond and are conserved in QhpC, the dependency of the intervening sequence between the two residues of the crosslinking loop was unknown so far. First, to elucidate sequence specificity of the loop region, we constructed a C‐terminally shortened QhpC variant (sQhpC) containing the leader peptide and only the first crosslinking site (CTSFDPGWE: Crosslinked residues are both ends). sQhpC was successfully crosslinked under anaerobic conditions by adding SAM and a reducing agent (Na2S2O4), resulting in the 7‐residues cyclic region (TSFDPGW). Furthermore, we analyzed the crosslinking formation by QhpD using sQhpC mutants containing various sequences between Cys and Glu residues. As a result, we found that various sequences and length are allowed for QhpD reaction. These results suggest that QhpD is a useful tool to introduce multi‐cyclic region in the peptide.

1. T. Nakai, et al. (2015) J. Biol. Chem., 290, 11144‐66.

ABS394

STRUCTURAL BASIS FOR CONFORMATIONAL CHANGE OF THE TOPAQUINONE COFACTOR DURING THE CATALYTIC REACTION OF BACTERIAL COPPER AMINE OXIDASE

Toshihide Okajima ¹, Takeshi Murakawa², Seiki Baba³, Satoshi Kanagawa¹, Hideyuki Hayashi², Takato Yano², Takashi Kumasaka³, Katsuyuki Tanizawa¹

¹Osaka University (Ibaraki, Japan); ²Osaka Medical College (Takatsuki, Japan); ³Japan Synchrotron Radiation Research Institute (Sayo‐gun, Japan)

Copper amine oxidases catalyze oxidative deamination of primary amines and produce the corresponding aldehydes. Its Tyr‐derived cofactor, topa quinone (TPQ), is reduced to an aminoresorcinol form (TPQamr) in the initial half‐reaction with substrate amine. TPQamr is in equilibrium with the TPQ semiquinone radical (TPQsq) by intramolecular electron transfer to Cu2+. We have demonstrated that TPQsq takes an on‐copper conformation, in which the 4‐OH group ligated directly to Cu2+, by X‐ray crystallography of the enzyme from Arthrobacter globiformis (AGAO). In the present study, to access structural basis for the TPQsq formation requiring the conformational change, we elucidated TPQamr/TPQsq equilibrium in both crystal and solution and determined the crystal structures reduced with amine substrates including ethylamine, phenylethylamine, and serotonin.

Thermodynamic analyses in non‐cryocooled crystals and solution of AGAO demonstrated that enthalpy and entropy changes for the transition from TPQamr to TPQsq are positive, indicating enthalpy‐driven formation of TPQsq. In the active‐site structure, the TPQ ring in the off‐copper conformation of TPQamr were restricted in its motion and tethered to Tyr284 with a strong hydrogen bond, whereas the TPQ ring of the on‐copper TPQsq had sufficient freedom. These differences well explained the thermodynamic parameter changes. In addition, binding of product aldehydes or substrate amines at the next cycle may facilitate equilibrium shift to TPQsq. We speculates that these are a part of the driving force of TPQsq formation.

References

1. T. Murakawa et al. (2015) J. Biol. Chem., 290, 23094

2. T. Murakawa et al. (2019) Proc. Natl. Acad. Sci. U.S.A., 116, 135

ABS395

REGULATORY ROLE OF 5‐AMP IN CAMP SIGNALOSOME DYNAMICS

Nikhil Tulsian ¹, Abhijeet Ghode²

¹National University of Singapore (Singapore, Singapore); ²Dept. of Biological Sciences, National University of Singapore (Singapore, Singapore)

The termination phase of cyclic AMP signal transduction in cells leads to increased levels of the product 5‐AMP, generated through the action of cAMP phosphodiesterases (PDEs). cAMP bound to its high affinity receptors such as regulatory subunit of Protein Kinase A (PKAR), is rapidly hydrolyzed through the action of colocalized PDEs. In these complexes, the cAMP binding site in PKA is merged with the PDE catalytic site to enable rapid turnover of cAMP through a process of substrate channeling. Here, we describe the functioning of this dynamic hydrolase‐receptor‐ligand ternary complex, which constitutes a basic unit of the cAMP signalosome. Our results reveal the PDE‐PKAR subunit interface and highlight the interplay of cAMP and 5‐AMP in maintaining and dissociating the interactions between elements of the cAMP signalosome. Our results indicate that PKAR toggles between a PKAC subunit‐bound and a PDE‐bound conformation governed by relative concentrations of cAMP and 5‐AMP. Our results with amide hydrogen/deuterium exchange mass spectrometry (HDXMS) specifically highlighted the transient interaction dynamics of signaling elements in cAMP signalosome complexes. A burst of cAMP initiates the activation response. Over time, levels of AMP rise to promote resetting of the cAMP‐PDE‐PKA signalosome to the basal state. This is the first instance highlighting an important and direct role of 5‐AMP in cAMP signalosome dynamics. Our results further reveal that oscillations in cAMP and 5‐AMP levels control output response of cAMP signaling, with implications in spatio‐temporal regulation mediated by the signalosomes. Our results also reveal a novel AMP‐bound conformation of PKAR critical for cAMP signal termination.

ABS396

DISRUPTION OF HOMOPHILIC PROTEIN‐PROTEIN INTERACTION OF P‐CADHERIN BY A FRAGMENT COMPOUND AS A TRIGGER TO INHIBIT CELL ADHESION

Akinobu Senoo ¹, Satoru Nagatoishi², Kouhei Yoshida³, Sho Ito⁴, Go Ueno⁴, Takumi Tashima¹, Shota Kudou¹, Kouhei Tsumoto³

¹Department of Chemistry and Biotechnology, School of Engineering, The university of Tokyo (Bunkyo‐ku, Japan); ²Institute of Medical Science, The University of Tokyo (Tokyo‐to, Japan); ³Department of Bioengineering, School of Engineering, The University of Tokyo (Tokyo‐to, Japan); ⁴RIKEN, SPring‐8 Center (Hyogo‐ken, Japan)

P‐cadherin is a protein responsible for the calcium‐dependent cell adhesion. P‐cadherin is overexpressed in several kinds of cancer cells and related to cancer metastasis. Therefore, P‐cadherin is one of the principal targets of drug discovery. P‐cadherin works in homophilic protein‐protein interaction (PPI), so an inhibitor of that homodimerization can become a promising candidate for a novel anticancer drug.

A part of the extracellular domain called EC12 and its mutants were prepared using E. coli expression system. Using the recombinant proteins, SPR‐based fragment screening was firstly performed to select small molecule ligands that bind to P‐cadherin and inhibit the homodimerization. Then, the inhibitory activity in the cellular level was checked by cell aggregation assay using CHO cells stably expressing P‐cadherin of full length. The small molecule ligand obtained in the SPR screening inhibited the cell aggregation dose‐dependently; that is, the compound inhibited the actual cell adhesion. To elucidate its inhibitory mechanism, hydrogen‐deuterium exchange mass spectrometry (HDX‐MS) was adopted, by which the exposure of PPI interface to the solvent was monitored. In the presence of the hit compound, one of the interfaces of homophilic PPI was more exposed to the solvent than in the absence of the compound, which suggested that the hit compound disrupted the homophilic dimerization of P‐cadherin. Our approach realized the selection of the hit compound for the inhibition of cell adhesion with only a mM order affinity to the target protein.

ABS397

HOW NATURE HARNESSES ENTROPY TO TUNE PROTEIN FUNCTION

Zachary Wood ¹, Nick Keul¹, Krishnadev Oruganty², Elizabeth Schaper Bergman³, Nathan Beattie¹, Weston McDonald¹, Renuka Kadirvelraj¹, Michael Gross³, Robert Phillips¹, Stephen Harvey⁴

¹University of Georgia (Athens, United States); ²University of Michigan (Ann Arbor, United States); ³Washington University (St. Louis, United States); ⁴University of Pennsylvania (Philadelphia, United States)

How evolution shapes the conformational landscape of a protein to tune a specific function is poorly understood. Protein evolution is constrained by the stability of the folded, native state. Despite this, many proteins contain intrinsically disordered (ID) peptide segments. In fact, 44% of human proteins contain ID segments >30 residues in length. The majority of these segments have no known function and are often removed to facilitate structural studies. Here we show that an ID segment can enhance the affinity of an effector binding site by modifying the dynamics of an allosteric network. The enhanced affinity does not depend on the sequence or charge of the ID segment. Instead, changes in effector binding affinity can be accurately predicted based on segment length alone. Using a combination of transient state kinetics, hydrogen‐deuterium exchange mass spectrometry, thermal denaturation studies, computer simulation and crystal structure analysis, we show that the ID segment alters the energy landscape of a folded protein to favor the allosteric response. Our evidence shows that the ID segment generates an entropic force that can rectify the conformational ensemble of a protein to favor a specific functional state. Thus, the persistence of intrinsic disorder in the proteome may reflect the evolution of low complexity structural elements that can tune a specific protein function.

ABS398

STRUCTURAL MODELING OF ANTIMICROBIAL PEPTIDES IN THE DATABASE OF ANTIMICROBIAL ACTIVITY AND STRUCTURE OF PEPTIDES

Anthony Armstrong ¹, Phil Cruz¹, Andrei Gabrielian¹, Mindia Chubinidze², Malak Pirtskhalava², Darrell Hurt¹, Alex Rosenthal¹, Mike Tartakovsky¹

¹Office of Cyber Infrastructure & Computational Biology, National Institute of Allergy and Infectious Diseases, National Institutes of Health (Bethesda, United States); ²Ivane Beritashvili Center of Experimental Biomedicine (Tbilisi, Georgia)

Antimicrobial peptides (AMPs) comprise a structurally and functionally diverse class of agents conferring native immunity in plants, animals, and bacteria. AMPs have been recognized for their potential therapeutic value, particularly in light of continually evolving resistance of pathogens against many traditional small molecule antibiotics. The Database of Antimicrobial Activity and Structure of Peptides (DBAASP) comprises manually‐curated, empirically‐determined activities for over 12,800 natural and synthetic AMPs, provides links to available experimentally determined structures, and endeavors to provide structural information for AMPs lacking experimental structures. A deeper understanding of AMP structure‐function relationships and successful AMP engineering efforts require atomic‐level structural information. To facilitate model generation for short peptides of fewer than thirty amino acids, an automated molecular dynamics (MD) pipeline has been developed and implemented on the high‐performance computing cluster of the National Institute of Allergy and Infectious Diseases (NIAID NIH). Starting from an extended conformation, peptide sequences retrieved from the DBAASP undergo 400 ns of MD simulation in explicit solvent. A representative structure generated from a central trajectory frame is made available for viewing or download in addition to the simulation trajectory, an analysis of secondary structure, and a self‐consistency heatmap suggesting conformational propensities. Peptides are processed in 1.8 days on average using a single GPU device. At present the DBAASP contains structural data associated with over 850 peptides. A detailed description of pipeline implementation and its output will be discussed.

ABS399

MODULATING ENZYME ACTIVITY BETWEEN SUGAR HYDROLYSIS AND SUGAR TRANSFER USING AN EVOLUTIONARY APPROACH

Rodrigo Arreola‐Barroso ¹, Wendy Xolalpa‐Villanueva¹, Leticia Olvera‐Rodriguez¹, Gloria Saab‐Rincón²

¹Institute of Biotechnology, UNAM (Cuernavaca, Mexico); ²Institute of Biotechnology (Cuernavaca, Mexico)

Many biologically active compounds are glyco‐compounds. The glycosidic moiety can not only improve their pharmacokinetic properties but also be crucial for their function. Their synthesis requires adding a sugar to a molecule, a process that usually involves multiple‐step reaction pathways to ensure a single product. On the other hand, most synthetic enzymes require a nucleotide‐activated sugar as substrate, resulting in a very costly alternative. To engineer tools capable of glycosylating molecules from cheap starting materials, one would require a transferase, an enzyme that transfers sugar units from a previously existing oligo‐ or polysaccharide.

The alpha‐amylase is a family of enzymes that preferentially hydrolyze ‐1,4‐glucosydic bonds in starch, however, some amylases are also able to transfer glucose units to acceptors different from water. This difference in reactivity gives rise to two groups in the alpha‐amylase family: the hydrolases and the transferases. In this study, we compared the inter‐residue contacts between these two groups, and identified residues outside the catalytic pocket whose substitution, by the amino acid suggested by our analysis, modified the transfer/hydrolysis ratio of the enzymes under study. We conclude that by analyzing the inter‐residue contacts that have been explored in nature for the alpha‐amylase family members it is possible to identify target sites for mutagenesis outside the catalytic site aimed to modify the transfer/hydrolysis ratio

ABS401

IMPACTING RESPIRATORY THERAPEUTIC PROGRAMS THROUGH PROTEIN BIOPHYSICS AND STRUCTURAL BIOLOGY

Michael Eddins ¹, Hua‐poo Su², Xiao Xiao³, Jennifer Shipman², Srivanya Tummala², John Reid², Yacob Gomez Llorente², Zhifeng Chen³, Eberhard Durr³, James Cook³, Kerim Babaoglu², Stephen Soisson², Lan Zhang³, Kalpit Vora³, Alexei Brooun²

¹Merck and Co., Inc., Computational and Structural Chemistry (West Point, United States); ²Merck and Co., Inc., Computational and Structural Chemistry (West Point, United States); ³Merck and Co., Inc., Department of Infectious Diseases and Vaccines Research (West, United States)

Respiratory tract infections due to respiratory syncytial virus (RSV) are a severe risk for infants, elderly, and immunocompromised adults. Infections of vulnerable populations are a major burden on the healthcare system with no vaccines available and limited therapeutic options. In efforts to identify/design vaccines and therapeutic antibodies, the ability to screen designs quickly and efficiently is crucial. In addition, production and supply of well characterized antigens and antigen/Fab complexes is important to enable structure guided efforts in a timely manner. Structurally mapping antibody‐epitope interactions helps understand how these antibodies function and provides insight into how we can design novel therapeutics to target these diseases. We present work flows that are utilized to enable antibody and antigen design screening, and structure enabled epitope mapping from gene to structure.

ABS402

EVALUATION OF THE EFFECT OF HEAT CAPACITY ON THE CATALYSIS OF A DIMERIC ENZYME

Ekaterina Jalomo Khayrova ¹, Christopher Bahl², Gloria Saab Rincón¹

¹Instituto de Biotecnología, Universidad Nacional Autónoma de México (Cuernavaca, Mexico); ²Institute for Protein Desing, University of Washington (Seattle, United States)

Enzymatic catalysis has a strong dependence on temperature. The optimum temperature is the temperature at which they reach the maximum activity. Above this, a decay is observed in the speed of reaction that has been attributed to the denaturation of the enzyme. Previously the enzymatic behavior with respect to temperature was explained by the Arrhenius and Eyring equation, in this the entropic component was despised because they considered that above the optimal temperature the enzyme began its denaturing process. However, careful analyzes have shown that not all the enzymes begin the process of denaturation in some degrees above the optimum temperature. Therefore, the theory of macromolecular velocity has been proposed, in which the dependence of enzymatic catalysis on temperature is explained, with a mathematical rearrangement of the Arrhenius and Eyring equation taking into account the change in the heat capacity and the consideration that the activation enthalpy and entropy parameters are dependent on the temperature. The change in heat capacity (Cp) is a thermodynamic property that describes the difference between the basal state of the enzyme‐substrate complex and the enzyme‐transition state complex. The heat capacity of an enzyme, in part, is given by its vibrational modes, which are intimately related to the dynamics of the enzyme. As a proof of principle, in the present work we propose to modify the vibrational movements of an enzyme without modifying its catalytic site to evaluate the effect of the change in the calorific capacity on enzymatic catalysis.

ABS403

DISTINCT STRUCTURAL FEATURES OF THE LON PROTEASE DRIVE CONSERVED HAND‐OVER‐HAND SUBSTRATE TRANSLOCATION

Mia Shin ¹, Ananya Asmita², Cristina Puchades¹, Eric Adjei², R. Luke Wiseman¹, A. Wali Karzai², Gabriel C. Lander¹

¹The Scripps Research Institute (La Jolla, United States); ²Stony Brook University (Stony Brook, United States)

Hand‐over‐hand translocation is emerging as the conserved mechanism by which ATP hydrolysis drives substrate translocation within the classical clade of AAA+ proteins. However, the operating principles of the distantly related HCLR (HslUV, ClpX, Lon, RuvB) clade remains poorly defined. We determined a cryo‐electron microscopy structure of Yersinia pestis Lon trapped in the act of processing substrate. This structure revealed that sequential ATP hydrolysis and hand‐over‐hand substrate translocation are conserved in this AAA+ protease. However, Lon processes substrates through a distinct molecular mechanism involving structural features unique to the HCLR clade. Our findings define a previously unobserved translocation mechanism that is likely conserved across HCLR proteins and reveal how fundamentally distinct structural configurations of distantly‐related AAA+ enzymes can power hand‐over‐hand substrate translocation.

ABS404

MODULATING RECEPTOR SIGNALING USING VARIABODY; A NOVEL BISPECIFIC ANTIBODY FORMAT ENABLES ONE‐POT SYNTHESIS OF FAB‐DIMER LIBRARY

Yasuhisa Shiraishi ¹, Akifumi Kato¹, Munetake Shimabe², Jun Taneo³, Minako Oogi³, Kaname Kimura¹, Shigeyuki Yokoyama⁴, Kensaku Sakamoto⁵

¹Antibody & Biologics Research Labs, Research Functions Unit, R&D Division, Kyowa Hakko Kirin Co., Ltd. (Machida‐shi, Japan); ²Clinical Sciences Research Labs, Translational Research Unit, R&D Division, Kyowa Hakko Kirin Co., Ltd. (Tokyo, Japan); ³Immunology & Allery Research Labs, Immunology & Allery R&D Unit, R&D Division, Kyowa Hakko Kirin Co., Ltd. (Tokyo, Japan); ⁴RIKEN Cluster for Science, Technology and Innovation Hub (Yokohama, Japan); ⁵Laboratory for Nonnatural Amino Acid Technology, RIKEN Center for Biosystems Dynamics Research (Yokohama, Japan)

Structural changes of cell membrane receptors such as dimerization are general mechanisms controlling signal transduction. Recently, the use of engineered mimetics and antibody‐based molecules including bispecific antibodies that induce receptor dimerization and signaling, has attracted a great deal of interest related to wide‐ranging biological research and therapeutic use. Although various types of bispecific formats have been reported, they are mainly based on recombinant technology which possesses format dependent structural limitations and difficulty in rapid construction of various library. These obstacles restrict the applicability of bispecific antibodies, and therefore, an alternative bispecific format with high variability in their structure and compatibility in library construction is required to facilitate the efficient screening for potent agonists.

Here we describe a novel bispecific antibody format Variabody, where two Fabs containing non‐natural amino acids with orthogonal chemical reactivity were conjugated one another via a chemical linker (Fig. 1). The relative orientation and distance between two Fabs could be controlled to achieve optimal dimerization and signal transduction of receptors. We developed the method, which enables one‐pot synthesis of Variabody library with multi‐well reaction plates. Some of the created Variabodies based on anti‐Her2 Fabs exhibited agonistic activity in cultured cells as opposed to the antagonistic nature of antibodies. In another case, Fab‐dimers engaging CD32B, inhibitory Fc receptor, to the B cell receptor complex triggered potent CD32B‐mediated inhibitory signaling. Variabody technology allows rapid generation of bispecific antibody libraries and provides promising platform for the efficient development of research tools as well as novel therapeutics with agonistic activity.

ABS405

CALCULATING POTENTIAL OF MEAN FORCE (PMF) WITH UMBRELLA SAMPLING PREDICTS STRUCTURE‐ACTIVITY RELATIONSHIPS (SAR) FOR POTENTIAL ABL TYROSINE KINASE INHIBITORS DERIVED FROM 2‐PYRAZOLINYL‐1‐CARBOTHIOAMIDE

Beom Soo Kim¹, Wookyung Yu ¹, Sangho Ji¹, Sang Won Jung²

¹Department of Brain and Cognitive Sciences, Daegu Gyeongbuk Institute of Science and Technology (DGIST) (Dalsung‐gun, South Korea); ²Center for Supercomputing and Big Data, Daegu Gyeongbuk Institute of Science and Technology (DGIST) (Taegu, South Korea)

The ABL 1 is proto‐oncogene encoding non‐receptor tyrosine kinase that is one of proteins cause chronic myeloid leukemia (CML). Many Abl tyrosine kinase inhibitors was developed and marketed like imatinib, dasatinib, and so on. However, the issue of their side effect still is discussed. Our previous study reported that derivatives of 2‐pyrazolinyl‐1‐carbothioamide (2PC) synthesized from chalcones showed inhibitory effects to Abl tyrosine kinase and their structure and activity relationships (SAR) were also analyzed by QSAR. Therefore, we suggested that it is needed to investigate correlations between ligand based‐study and structure‐based study. Here, to describe between SAR and protein‐ligand interactions, we defined reaction coordinate of protein‐ligand dissociation using steered molecular dynamics (SMD) simulations. And, their potential of mean force (PMF) was calculated using umbrella sampling (US) to evaluate their free energy landscape during ligand dissociation events. In previous study, it is observed cytotoxicity differences between 2‐methoxy derivatives and 3,4,5‐trimethoxy derivatives with QSAR. So, we selected 3 derivatives, 2PC‐1, 2PC‐2, 2PC‐3. Interestingly, differences in the free energy between derivatives were appeared in PMF results and ligand binding affinities was described using free energy in these data. So, this study is expected not only to develop novel Abl kinase inhibitors with low side‐effects but also propose one of the approaches to discover novel drugs from natural products.

ABS406

FUNCTIONAL NETWORKS STUDY OF G USING COEVOLUTION ANALYSIS

MINJAE SEO¹, Wookyung Yu ¹

¹DGIST (Daegu Gyeongbuk Institute of Science and Technology) (DAEGU, South Korea)

Heterotrimeric G proteins coupling with G‐protein coupled receptors (GPCR) conduct major signal transduction in diverse cell environments including neuron. Ligand binding affects the conformational change of GPCR, which transduces a signal to the heteromeric G protein. Then G protein is dissociated into G and G‐G complex to trigger various biological pathways. Even though binding and dissociation between GPCR and G protein are the most important step in signal transduction, it is not clearly solved whether which residues are important. In this study, we investigated G protein, which has direct interaction with GPCR, from the evolutionary view using statistical coupling analysis (SCA). SCA captured the co‐evolving residues among various species and organized them into functional subunits, so‐called sectors, of G proteins. We rearranged the positions of these sectors to more clearly distinguish networks of correlated residues and mapped on the structure of GPCR‐G protein complex. Through these processes, we found the distinct sectors that compose structure‐to‐functional networks of G protein. Furthermore, this result is consistent with previously known evolutionary determinants of G protein.

ABS407

BIFURCATED H‐BONDING IN MEMBRANE PROTEINS

Esther S Brielle ¹, Isaiah T Arkin²

¹The Alexander Grass Center for Bioengineering. The Hebrew University of Jerusalem, Edmond J. Safra Campus, Jerusalem, 91904, Israel. (Jerusalem, Israel); ²The Alexander Silberman Institute of Life Sciences. Department of Biological Chemistry. The Hebrew University of Jerusalem, Edmond J. Safra Campus, Jerusalem, 91904, Israel. (Jerusalem, Israel)

H‐bonds are prominent stabilizing forces in proteins. The canonical H‐bonds responsible for protein secondary structure are composed of a single donor and a single acceptor. Yet often, H‐bond configurations involve multiple donors and/or acceptors. Our research investigates the energetics of such bonds and in particular how polar side chains in membrane proteins form bifurcated H‐bonds with an over‐coordinated proton acceptor.

The canonical H‐bonding pattern of ‐helices involves the backbone NH of residue i and the backbone carbonyl of residue i‐4. When residue i is a serine or threonine, the sidechain prefers the non‐ideal ‐60 rotamer in order to backbond to the backbone carbonyl of residue i‐4, which becomes an overcoordinated proton acceptor. 53% of serines and 68% of threonines in transmembrane ‐helices form i‐4 bifurcated H‐bonds. Using fourier transform infrared spectroscopy (FTIR) combined with density functional theory (DFT) quantum simulations, we measure the strength of this additional H‐bond as 50‐60% of canonical ‐helix H bonds [1].

Serine and threonine side‐chains can backbond to the backbone carbonyl at position i‐3 as well. This type of bifurcated H‐bond forms in 35% of serines and 38% of threonines in transmembrane ‐helices. In fact, these side chains are sometimes over‐coordinated to both the i‐3 and the i‐4 backbone carbonyls simultaneously, creating a multiplexed H‐bonding system with multiple proton donors and acceptors. We find that the i‐3 bifurcated H‐bond has similar energy to that of the i‐4 bifrucated H‐bond, but it adopts the +60 rotamer and forms in more polar environments.

[1] Esther S. Feldblum and Isaiah T. Arkin. Strength of a bifurcated h bond. Proceedings of the National Academy of Sciences, 111(11):4085‐4090, 2014.

ABS408

ENGINEERING THERAPEUTICS FOR THE TREATMENT OF ANEMIA IN ONCOLOGY PATIENTS

Ning Yang ¹, Marina Chemerovski‐Glikman², Lia Cardarelli², Jarrett Adams², Sachdev Sidhu¹

¹Donnelly Center for Cellular and Biomolecular Research, University of Toronto (Toronto, Canada); ²Toronto Recombinant Antibody Centre, University of Toronto (Toronto, Canada)

Erythropietin (EPO) is a natural hormone involved in the regulation of erythropoiesis. While recombinant human EPO has been widely used to treat anemia, a common complication that affects ~90% of cancer patients receiving chemotherapy, a growing number of studies have demonstrated that EPO treatment can compromise patient overall survival. However, the underlying mechanism remains unclear and new therapeutic strategies for the treatment of chemotherapy‐induced anemia are required.

In this study, we deciphered the mechanisms by which EPO exerted stimulatory effects in cancer cells. We demonstrated that EPO acts through the non‐canonical Ephrin type‐B receptor 4 (EPHB4) to mediate tumor cell migration. To further elucidate this interaction, we isolated two specific EPHB4 antibodies able to abolish EPO‐mediated tumor migration in vitro. These antibodies therefore could potentially serve to protect patients from the adverse effects of EPO therapy mediated through EPHB4 without affecting the beneficial effects mediated through the EPO receptor (EPOR). We also generated highly specific synthetic agonists for EPO receptor (EPOR), which is responsible for EPO‐mediated erythropoiesis. From a phage‐displayed diabody library, we obtained several EPOR‐specific agonists. Our preliminary data show that our engineered diabodies bind to EPOR with high affinity and act as EPOR agonists by triggering its downstream signaling pathways, thereby stimulating the proliferation of erythroid cells. Our studies may lead to the development of improved therapeutics with a goal of providing effective and safe treatment of anemia in cancer patients.

ABS409

THE RETAINING GLYCOSIDE HYDROLASE T26H MUTANT OF T4 LYSOZYME UTILIZES A REVERSE PROTONATION CATALYTIC MECHANISM

Jacob Brockerman ¹, Mark Okon¹, Stephen Withers¹, Lawrence McIntosh¹

¹University of British Columbia (Vancouver, Canada)

T4 phage lysozyme (T4L) is an inverting glycoside hydrolase that cleaves bacterial cell wall peptidoglycan. Remarkably, the single amino acid replacement of Thr26 to His (T26H) changes its catalytic mechanism from inverting to retaining and also imparts transglycosylase ability (Kuroki et al., PNAS 1999; 96:8949). T26HT4L is proposed to utilize a Koshland double displacement mechanism with His26 acting as a nucleophile to form a glycosyl‐enzyme intermediate and Glu11 serving as a general acid to protonate the leaving aglycone. To gain further insights into this or alternative catalytic mechanisms, we used NMR spectroscopy to determine the protonation states and, when possible, the acid dissociation (pKa) values of most ionizable residues (Asp, Glu, His, and Arg) in the mutant lysozyme. Of particular interest, the pKa value of the general acid Glu11 is 4.7 ± 0.1, whereas that of the putative nucleophile His26 is 6.8 ± 0.1. If the proposed double displacement mechanism holds true, then T26HT4L follows an underappreciated reverse protonation pathway in which only a small population of the free enzyme is ever in its catalytically‐competent ionization state with the nucleophile His26 deprotonated and the general acid Glu11 protonated. Our NMR measurements also confirm that all arginines in T26HT4L, including the active site Arg145, are positively charged under neutral pH conditions.

ABS410

DISCOVERY AND INTERROGATION OF THE REGULATORY ROLE OF THE SH4 DOMAIN OF SRC FAMILY KINASES

Sujata Chakraborty ¹, Ethan Ahler², Linglan Fang³, Emily Dieter³

¹University of Washington (Seattle, United States); ²Doctor (Seattle, United States); ³Graduate Student (Seattle, United States)

Protein kinases are dynamic molecular switches that participate in a plethora of signal transduction pathways. Composed of a bilobed catalytic domain (CD), all kinases possess an ATP binding cleft, where a specific substrate gets phosphorylated. In order to facilitate this phosphotransferase event with great fidelity kinases are tightly regulated in cells. About half of the known 530 eukaryotic protein kinases, possess at least one regulatory domain that facilitates kinase regulation in cells via intra or intermolecular protein‐protein interactions. Misregulation of phosphotransferase event causes hyperactivation of signaling cascades ultimately leading to several disease forms including cancer, inflammation etc. Therefore, understanding kinase regulation is of utmost importance. Here we have used Deep mutational Scanning (DMS) to investigate one of the best studied multi‐domain protein kinases, Src. Src is a non‐receptor protein kinase composed of several accessory domains, such as the SH4, unique, SH3 and SH2 domains in addition to its SH1 or CD. While Src regulation via its SH3 and SH2 domains are well documented in literature, the regulatory role of SH4 has so far been neglected. Our DMS study revealed several hyperactivating clusters within the CD of Src that can be sites for regulatory domain association. Our data helped identify a unique patch in the C‐lobe of Srcs CD that associates with its SH4 domain. Disassociation of this interaction can lead to hyperactivation of the kinase, membrane re‐localization and unique cellular morphology that is independent of phosphotransferase role of the kinase. We think DMS is a very useful tool that can be utilized in investigating new regulatory mechanism in several other less understood protein kinases and shed light towards their structure, function and regulation in cells.

ABS411

HOMOTYPIC AND HETEROTYPIC INTERACTIONS OF PLEXIN AND NEUROPILIN TM DOMAINS

Shaun Christie ¹, Soon‐Jeung Kim², Paul Toth¹, Jeannine Muller‐Greven², Matthias Buck², Adam Smith¹

¹The University of Akron (Akron, United States); ²Case Western Reserve University (Cleveland, United States)

Plexin family receptors and their co‐receptor neuropilin‐1 (Nrp1) allow for signaling by semaphorin ligands and have broad expression among tissue types. These receptors induce many functions including axon branching, vasculature guidance and cell migration. Direct molecular interaction between these receptors is not completely understood, but receptor complexes have been shown to regulate signaling in some cases. The transmembrane (TM) domains of many proteins interact with each other to promote dimerization and orientation of proximal soluble domains, however, the role of the TM domains specifically in these interactions is poorly understood. In order to study these TM interactions we produced constructs containing the TM domain and small portions of the ECD and ICD fused to a fluorescent protein. This allows the use of pulsed interleaved excitation fluorescence cross correlation spectroscopy to resolve mobility, proximity, and the degree of oligomerization. Due to the unique sequence characteristics of the TM domains of Plexin B1, B2, D1, and Nrp1 our study explores how these TM domains interact with each other in the live cell membrane environment. As a positive control, we include the well‐studied interaction of the Glycophorin A TM domain, which dimerizes at a GxxxG motif. In addition, we used combinations of TM and full length protein to resolve the strength of the TM interaction. The results presented here show that the TM domains can work in concert with soluble domains to influence oligomerization and can be used to guide potential development of peptides for disease related studies.

ABS412

ATOMISTIC VIEW OF AN UNFOLDING PATHWAY IN A SEVEN‐HELICAL MEMBRANE PROTEIN

Peng Xiao¹, Leonid Brown ¹, David Bolton¹, Vladimir Ladizhansky¹

¹University of Guelph (Guelph, Canada)

Membrane protein folding, structure, and function strongly depend on a cell membrane environment, yet structural characterization of membrane protein folding within a lipid bilayer is challenging. Here we combine hydrogen‐deuterium exchange and solid‐state NMR to obtain an atomistic description of the thermally induced reversible unfolding pathway of a seven‐helical trimeric photoreceptor Anabaena Sensory Rhodopsin in lipids. The pathway is characterized by progressive outward displacement of the cytoplasmic half of the seventh helix, which expands the internal water‐filled cavity towards the cytoplasmic surface. The trimeric interface and a stable core of residues clustered around the retinal cofactor remain intact until the irreversible unfolding stage, and contribute to the protein thermal stability. The approach employed here opens new ways to probe membrane protein unfolding in lipid bilayers.

ABS413

MECHANISM OF MICROTUBULE NUCLEATION IN THE PCM

SHIOU‐LAN LIN ¹

¹Institute of Bioinformatics and Structural Biology, National Tsing‐Hua University (Hsin‐chu, Taiwan)

The centrosome is the major microtubule‐organizing center (MTOC) that contains more than 200 proteins. It consists of two orthogonally arranged centriole, surrounded by pericentriolar material (PCM), where a core microtubule nucleating factor, the ‐tubulin ring complex (TuRC), resides to initiate new filament formation. However, recent studies showed that even in the absence of TuRC, PCM phase separation itself is capable to trigger microtubule nucleation. To elucidate the underlying mechanism of microtubule nucleation by PCM‐regulated phase separation, we analyzed centrosomal proteins that contain coiled‐coil domains and found sequential poly‐K/R in sort regions. We hypothesized that microtubule nucleation is attributed by the charge‐charge interaction between poly‐K/R regions in centrosomal proteins and poly‐D/E C‐terminal tails of tubulins in the PCM condensates. To test this model, we exploited multivalent protein systems as a proxy for the crowded PCM environment in the absence of TuRC and poly‐K/R sequences with increasing length were linked to the multivalent proteins genetically. The microtubule nucleation ability of the poly‐K/R containing protein condensates was correlated with the length of poly‐K/R sequences. Natural poly‐K/R motifs from centrosomal proteins also promote microtubule nucleation. These data suggested the charge‐charge interaction between the highly conserved tubulin tails and the poly‐K/R regions in centrosomal proteins contributes to microtubule growth in the centrosome.

ABS414

PURIFICATION OF RECOMBINANT ADIPOSE TRIGLYCERIDE LIPASE (ATGL) FOR BIOCHEMICAL, BIOPHYSICAL AND MECHANISTIC STUDIES

Suman Shanker ¹, Kim Fennell¹, Nicole Caspers¹, Yang Cong¹, Erik Ralph¹, Jessica Calloway¹, Jemy Gutierrez¹, Benjamin Reidich¹, Cecile Vernochet¹, Francis Rajamohan¹

¹Pfizer (Groton, United States)

Adipose triglyceride lipase (ATGL) catalyzes the rate‐limiting step of triacylglycerol (TG) hydrolysis in multiple tissues. It is mainly localized to TG‐rich intracellular lipid droplets. The catalytic site of ATGL consists of an unusual dyad comprising Ser47 and Asp166 located within the patatin domain at the N‐terminus of the protein. The nucleophilic Ser47 is located within a highly conserved GXSXG motif as found in other lipases of the / hydrolase fold family. The C‐terminal domain is highly variable and has been reported to have a role in targeting to lipid droplets for some family members. Similar to other lipases, ATGL activity is regulated by comparative gene identification‐58 (CGI‐58 or ABHD5), a cofactor that directly binds to and activates ATGL. In humans, ATGL deficiency causes neutral lipid storage disease with myopathy (NLSDM) characterized by a systemic TG accumulation. A recent animal study showed that the absence of ATGL in mice fed with a high‐fat diet (HFD) leads to reduced food intake, decreased lipid synthesis, and obesity resistance. These results highlight a previously unrecognized interdependence between functional lipolysis and effective lipid biogenesis and suggest that ATGL inhibition in adipose tissue may ameliorate HFD‐induced obesity as well as insulin resistance and type‐2 diabetes. Most importantly, a pharmacological approach using the ATGL inhibitor, Atglistatin®, has recapitulated many phenotypes of genetic mouse models. However, the impact of ATGL in regulating plasma FA concentrations is less clear in humans. To date, few functional characterizations of human ATGL were attempted due to the difficulties involved in expression and purification of recombinant ATGL protein. One of our major objectives at the outset was to generate high quality protein reagents for biochemical, biophysical and mechanistic studies. In this study, we have expressed and purified high quality recombinant ATGL protein from insect cells and studied its biophysical and biochemical properties

ABS415

STRUCTURAL BASIS OF OLA1 ACTIVATION BY BARD1 BRCT

Ting Chen ¹

¹National Tsing Hua University (Hsinchu City, Taiwan)

Centrosomes, the major microtubule organizing center in mammalian cells, play a critical role in cell division and spindle formation. The breast cancer associated gene 1 (BRCA1) is a tumor suppressor of breast and ovarian cancers. It forms a heterodimer with BRCA1‐associated RING domain protein (BARD1), and acts as an E3 ubiquitin ligase in DNA repair, transcription, ubiquitination and centrosome regulation. Importantly, BRCA1 silencing and mutations result in abnormal centrosome amplification. In addition, an Obg‐like ATPase OLA1 binds to the N‐terminal region of BRCA1 and C‐terminal region of BARD1. These interactions are critical for correct centrosome number. Here, we determined the molecular interplay between BARD1 and OLA1 by a series of biophysical and biochemical analyses. Based on the NMR and enzymatic studies, ATP‐bound OLA1 binds to the BARD1 BRCT domain better than the nucleotide‐free OLA1, and this interaction enhances the ATPase activity of OLA1, whose turnover number (kcat) is increased in the presence of BARD1 BRCT. Structural analyses in combination with the mutagenesis study indicated that a highly conserved region on the BARD1 BRCT domain contribute to the binding of OLA1. Moreover, a cancer associated mutation on the BRCT significantly reduces its to activate OLA1. Altogether, these data identify a functional surface on the BARD1 BRCT domains that contribute to centrosome regulation by modulating the activity of OLA1.

ABS416

CATALYTIC BIOSCAVENGER WITH IMPROVED STABILITY AND REDUCED SUSCEPTIBILITY TO OXIDATION TO TREAT ACUTE POISONING WITH NEUROTOXIC ORGANOPHOSPHOROUS COMPOUNDS (OPS)

Laura Job ¹, Anja Köhler², Benjamin Escher¹, Franz Worek³, Arne Skerra¹

¹Chair of Biological Chemistry, Technical University Munich (Freising, Germany); ²Chair of Biological Chemistry, Technical University Munich, Freising; Bundeswehr Institute of Pharmacology and Toxicology (Munich, Germany); ³Bundeswehr Institute of Pharmacology and Toxicology (Munich, Germany)

Organophosphorous nerve agents pose a severe toxicological threat, both at disposal, during use as pesticides, in military conflicts and if considering potential acts of terrorism. The organophosphates of the V‐type (VX, VR, CVX) are highly problematic due to their persistence, low vapour pressure and the slow biodistribution following contact via skin or lung while satisfying treatments for detoxification are unavailable so far. Hydrolytic enzymes, which may be administered into the blood stream by injection and decompose the circulating nerve agent into less toxic compounds in vivo could provide a solution. Indeed, nature offers some enzymes that can hydrolyze OPs, although with low catalytic activities; for the phosphotriesterase found in the bacterium Brevundimonas diminuta (BdPTE), mutants with improved catalytic efficiencies have been described; yet, their biochemical properties are insufficient for therapeutic application. Hence, the goal of our research project is the development of novel variants of BdPTE using rational protein design. The mutant enzymes are produced as recombinant proteins in Escherichia coli, purified to homogeneity and their structural, biochemical and enzymological properties are assessed. Engineered BdPTEs with broad substrate spectrum, including pesticides as well as V‐type nerve agents, and higher stability were identified. Some candidates show enhanced thermal stability and are less susceptible to oxidation, as demonstrated by mass spectrometry measurements. Moreover, we have applied PASylation technology in order to effect extended half‐life of this enzyme in circulation. These mutants of BdPTE may show promise for therapeutic application in the treatment of nerve agent as well as pesticide intoxications.

ABS417

MULTI‐SUBUNIT E.COLI EXPRESSION SYSTEM APPLIED TO THE X‐RAY CRYSTALLOGRAPHIC ANALYSIS OF S. POMBE MEDIATOR COMPLEX

Kayo Nozawa ¹, Thomas Schneider², Patrick Cramer³

¹The University of Tokyo, Institute for Quantitative Biosciences (Bunkyo‐ku, Japan); ²Deutsches Elektronen Synchrotron (Hamburg, Germany); ³Max Planck Institute for Biophysical Chemistry (Göttingen, Germany)

High‐resolution structural information of biological macromolecules allows to understand the biochemical mechanisms underlying life. In the process of transcription of DNA into messenger RNA, Mediator is the central multi‐subunit complex directing RNA polymerase II to transcription start sites. The yeast Mediator complex, comprising 25 subunits with a total mass of 1.4 MDa, is required for the majority of protein coding gene expression. In the past, the assay and structural studies had been hampered by difficulties in obtaining pure, ~mg order Mediator proteins. In a long‐term effort to solve the high‐resolution Mediator structure, we co‐expressed Mediator subunits in Escherichia coli and determined crystal structures of essential core Mediator (cMed) comprises 15 subunits (Nozawa et al., Nature 2017). We prepared recombinant cMed from the fission yeast S. pombe, and improved protein solubility and yield by co‐expression of Med1. To obtain crystals, we removed the flexible Med14 residues 581879 and five C‐terminal residues of Med11. Crystal dehydration improved the diffraction limit from ~8 to ~4 å resolution. Complete diffraction data to 3.4 å resolution were obtained with a collimated synchrotron beam. Phases were derived from a heavy‐metal cluster derivative, incorporated selenomethionine, and the head module structure. Interpretation of the electron density was facilitated by sequence markers and led to a refined structure with excellent stereochemistry. The structure reveals the head and middle modules, and shows how the two modules interact. These results provide a framework for understanding Mediator function in the transcription pre‐initiation complex.

ABS418

THE EFFECTS OF PROTONATION OF A PHOSPHORYLATED AMINO ACID ON THE MOLECULAR RECOGNITION: COMPARATIVE STUDIES OF GENERIC PROTEINS AND AN ANTIBODY

Raiji Kawade ¹, Daisuke Kuroda¹, Jose Caaveiro², Hiroki Akiba³, Shigeru Okumura⁴, Toshiaki Maruyama⁴, Kevin Entzminger⁴, Kouhei Tsumoto¹

¹The University of Tokyo (Bunkyo‐ku, Japan); ²Kyushu University (Kyushu, Japan); ³National Institute of Health and Nutrition (Osaka, Japan); ⁴Abwiz Bio Inc. (San Diego, United States)

Since phosphorylation of an amino acid plays important roles in various cellular processes, how phosphorylated amino acids are recognized has been widely discussed. However, many previous studies assumed that the protonation state of a phosphate group was (PO32) despite the fact that the phosphate group (pKa ~7) would exist as an equilibrium mixture of the unprotonated state (PO32‐) and the singly protonated state (PO3H) in physiological condition. To analyze effects of a protonation state on protein dynamics, we performed MD simulations of 4 different proteinphosphorylated peptide complexes in various biological processes in which each peptide contained a phosphoserine residue in the unprotonated or singly protonated state. Our result showed that the (PO32‐) was more preferable to (PO3H‐) in the interactions due to the larger mobility of the phosphate group in the (PO32‐) state. Furthermore, we also obtained a monoclonal antibody toward a phosphorylated peptide by phage display, and X‐ray crystallography, thermodynamic analysis, mutagenesis, and MD simulations showed this antibody captures both protonation states equally well. Putting together, our results suggest that even a single protonation could have a large effect on molecular recognition of a phosphate group.

ABS419

ILLUMINATING THE EVOLUTION OF BEETLE BIOLUMINESCENCE WITH FATTY ACYL‐COA SYNTHETASES

Spencer Adams Jr. ¹, Stephen Miller¹

¹University of Massachusetts Medical School (Worcester, United States)

Fireflies are beetles that evolved the ability to emit light using the biochemical reaction known as bioluminescence. The chemistry of firefly bioluminescence requires that an enzyme known as a luciferase oxidize the small‐molecule substrate, D‐luciferin. Firefly luciferase evolved from an ancestral fatty acyl‐CoA synthetase enzyme and still retains this activity. Conversely, fatty acyl‐CoA synthetases from non‐luminescent insects, such as the fruit fly enzyme CG6178, can also function as luciferases if they are supplied synthetic luciferins. This latent luciferase activity in insect fatty acyl‐CoA synthetases is not observed with the natural beetle substrate, D‐luciferin. To better understand its latent luciferase activity, as well as the molecular basis of its unique substrate selectivity, we used x‐ray crystallography to solve the structure of CG6178. This structure reveals differences in the active site relative to firefly luciferase that we targeted for mutagenesis. These mutants improve the luciferase activity of CG6178 up to 20‐fold and even allow it to catalyze light emission from D‐luciferin. With this success in mind, we created chimeric luciferase enzymes that combine one of the two domains from firefly luciferase with the other domain from three similar insect fatty acyl‐CoA synthetases. These chimeras emit robust light sufficient for bioluminescent imaging in live mice, with each exhibiting unique substrate selectivity. Therefore, fatty acyl‐CoA synthetases provide a novel platform for engineering luciferases with unique substrate selectivity robust enough for in vivo use.

ABS420

NMR NOE ASSIGNMENTS AND BUILDUP MEASUREMENTS OF IM7 IN SOLUTION: TOWARDS INTERNAL DYNAMICS CHARACTERIZATION FROM NOES

Xinyao Xiang ¹, Chunhua Yuan², Alexandar Hansen², Lei Brüschweiler‐Li², Rafael Brüschweiler³

¹Department of Chemistry and Biochemistry, The Ohio State University (Columbus, United States); ²Campus Chemical Instrumental Center, The Ohio State University (Columbus, United States); ³Department of Chemistry and Biochemistry, & Campus Chemical Instrumental Center, & Department of Biological Chemistry and Pharmacology, The Ohio State University (Columbus, United States)

The homonuclear nuclear Overhauser effects (NOEs) between proton pairs are conventionally used as semi‐quantitative restraints for structure determination of biomacromolecules by nuclear magnetic resonance (NMR) spectroscopy. The observed experimental NOEs are averages over all the conformations sampled by a biomolecule during the measurement time; therefore, they also contain dynamics information. However, those dynamics are usually hidden in the experimental errors of NOE measurements. The recently developed exact NOE method by Vögeli et al. [1] allows for an accurate determination of NOE rates through analyzing the NOE buildups from a set of multidimensional NOESY spectra with various mixing times from which internal flexibility is derived through ensemble generation. Here, we apply this method on the colicin E7 immunity protein 7 (Im7). We have acquired a near complete resonance assignment of Im7 with stereospecifically assigned HBs and performed NOE assignments on 3D 15N‐edited NOESY and 13C‐edited NOESY experiments with 100 ms mixing time. The NOE restraints collected are used as input for the CNS software to calculate the structure in a traditional way, and the generated ensemble converges well and shows overall good agreement among the NOEs. We have recorded the buildups of 15N‐edited HSQC‐NOESYs with mixing times ranging from 15 ms to 200 ms at 850 MHz with high resolution and high quality. The accurate NOE rates extracted are being used to further refine the conformational ensemble to reveal internal motions at atomic resolution.

[1] Vogeli, B.; Segawa, T. F.; Leitz, D.; Sobol, A.; Choutko, A.; Trzesniak, D.; van Gunsteren, W.; Riek, R. Exact distances and internal dynamics of perdeuterated ubiquitin from NOE buildups. J. Am. Chem. Soc. 2009, 131 (47), 17215‐17225.

ABS421

STRUCTURAL AND ENZYMATIC CHARACTERIZATION OF A PENICILLIN BINDING PROTEIN FROM LEPTOSPIRA INTERROGANS

Jademilson Celestino dos Santos ¹, Sumit Handa², Luis Guilherme Virgilio Fernandes³, Partho Ghosh², Ana Lucia Tabet Oller Nascimento³

¹UCSD/Insituto Butantan (San Diego / Sao Paulo, United States); ²UCSD (San Diego, United States); ³Instituto Butantan (Sao Paulo, Brazil)

Leptospirosis is a widespread zoonotic disease caused by pathogenic spirochaetes of the genus Leptospira; this bacteria has a high prevalence in tropical and subtropical regions significantly impacting public health. The transmission of leptospirosis has been associated with exposure of individuals to wild or farm animals. Recently, the disease became prevalent in cities with sanitation problems and large populations of urban rodents, the latter of which contaminate the environment through their urine. Historically, antibiotics have been used to treat leptospirosis: penicillin was the first drug used to fight leptospirosis. Other antibiotics like doxycycline, cephalosporins, chloramphenicol, and azithromycin have also been tested in clinical trials. Despite the use of antibiotics in the treatment of leptospirosis, their role is still not completely clear due to the lack of effective clinical trials, particularly for severe cases of the disease. Here we present the crystal structure from Lsa45, a previously uncharacterized protein, which revealed a bifunctional enzyme that has two domains: a large / domain and a small ‐helix domain. The structure allowed us to understand the functionality of Lsa45; the protein was first characterized as a Penicillin Binding Protein (PBP) with weak affinity to ‐lactams. Additionally, an esterase functionality was characterized on Lsa45 with a greater affinity for the p‐nitrophenyl acetate substrate. These findings are important to understanding the mechanism of action antibiotics in the treatment of leptospirosis.

ABS422

STRUCTURE DETERMINATION OF NONTUBERCULOSIS MYCOBACTERIA DIHYDROFOLATE REDUCTASE TO INFORM STRUCTURE‐GUIDED DRUG DISCOVERY

Rachael Zigweid ¹, Brad Hammerson¹, Abe Shim¹, Stephen Mayclin², Jan Abendroth², Bart Staker¹, Peter Myler¹

¹Seattle Children's Research Institute (Seattle, United States); ²UCB Seattle (Bainbridge Island, United States)

Dihydrofolate reductase (DHFR) catalyzes the last step in the biosynthesis of tetrahydrofolate (THF), an essential coenzyme for the biosynthesis of DNA, RNA, and proteins. Inhibition of DHFR has been used to treat cancer and infectious disease, making it a promising target against pathogenic mycobacterial species. Mycobacterias unique cell wall composition protects against long exposure to acids, alkalis, detergents, oxidative bursts, and lysis by complement. It also provides resistance to antibiotics such as penicillin, which disrupt cell wall biosynthesis, requiring alternative treatments to be developed.

The structure of Mycobacterium tuberculosis (Mtb) DHFR (MtDHFR), the pathogen that causes Tuberculosis (Tb), has been determined in complex with several inhibitors, but these are not currently used to treat Tb due to weak activity and increased drug resistance. However, there are over 190 species of nontuberculosis mycobacteria (NTM), including M. abscessus and M. ulcerans, which cause chronic lung infection and severe skin lesions, respectively. The Seattle Structural Genomics Center for Infectious Disease (SSGCID) at Seattle Childrens Research Institute (SCRI) has produced 3 surface‐entropy mutants each of M. abscessus and M. ulcerans DHFR, and structural data of M. ulcerans in complex with P218 has been obtained with a resolution of 0.92AA. P218 is a derivative of WR99210, a specific inhibitor of Plasmodium falciparum DHFR with promising minimum inhibitory concentrations (MIC) on Mtb.

While P218 has poor activity against Mtb, synthetic analogs show decreased MIC and increased activity in Mtb‐infected macrophages. Structural data from NTMs can be compared to Mtb to inform structure‐guided drug discovery to develop stronger inhibitors against mycobacterial DHFR, leading to more effective treatment of Tb and other mycobacterial infections.

ABS423

DEVELOPING ORTHOGONAL SINGLE‐MOLECULE CONSTRUCTS USING HUH ENDONUCLEASES AND DNA HANDLE SELF‐ASSEMBLY

Andrew Nelson ¹, Kassidy Tompkins¹, Blake Everett¹, Klaus Lovendahl², Wendy Gordon¹

¹University of Minnesota‐ Twin Cities (Minneapolis, United States); ²University of Washington (Seattle, United States)

When studying proteins in single‐molecule force experiments, the quality of the data is often dependent on the merit of the experimental setup. Regardless of assay or means, observing proteins on the femtomolar scale requires constructs which are rigid, uniform, and orientation‐specific. While advances in protein manipulation through DNA nanostructures have greatly improved the efficiency of single‐molecule force studies, difficulty still lies in effectively linking protein to DNA. Current conjugation methods using chemical crosslinking or small molecule interactions are either nonspecific, weak, or costly, and all require chemical DNA modification before protein binding. We identify a more effective approach in the use of HUH endonucleases, dubbed HUH‐tags. Native to many bacterial and viral species, HUH‐tags are small protein domains capable of forming covalent adducts with DNA in a sequence‐specific manner. Because of this, we propose HUH‐tags as a novel protein fusion tag for isolating chimeric proteins of interest in DNA handles. Combined with current methods in DNA handle self‐assembly, HUH‐tags introduce easily mutable, high‐throughput protein immobilization constructs. Through magnetic tweezer force microscopy, we demonstrate the utility of HUH‐tags as both glass surface tethers and fusion tags for protein immobilization, with the future promise of observing protein‐protein interactions.

ABS424

PROTEIN DISORDER IN RECONSTITUTED DYNEIN CARGO ATTACHMENT SUBCOMPLEX

Kayla Jara ¹, Cat Hoang¹, Sanjana Saravanan¹, Elisar Barbar¹

¹Oregon State University (Corvallis, United States)

Cytoplasmic dynein is a 1.6‐MDa motor protein complex composed of heavy, intermediate, light intermediate, and light chain subunits. The motor domain of dynein has been well characterized, however the cargo attachment domain contains 300 amino acids of disorder that remain missing from resolved dynein structures. This disorder is in dyneins intermediate chain (IC), the key protein in dyneins cargo attachment domain, responsible for loading cargo, maintaining stability, and modulating dynein activity. The N‐terminus of IC (N‐IC) interacts with dynein light chain subunits; Tctex, LC8, and LC7 to form a subcomplex that acts as a polybivalent scaffold. The assembly of monomeric IC and the homodimeric light chains is unique in that binding induces tight packing at protein interfaces, leaving linker regions completely disordered. In this work, we reconstitute the IC and light chain subcomplex from Chaetomium thermophilum, a thermophilic filamentous fungus and a novel choice for dynein studies. We have recombinantly expressed and purified each component and subsequently purified the ~150 kDa subcomplex by size exclusion chromatography (SEC). Initial characterization by SEC and analytical ultracentrifugation (AUC) indicate that the subcomplex binds with moderate affinity that is enhanced by addition of a non‐dynein regulator such as the nuclear distribution protein (NudE) or the p150Glued subunit of dynactin. Further, nuclear magnetic resonance (NMR) data confirm remaining disorder upon subcomplex assembly. This retained intrinsic disorder of IC contributes to the cargo attachment domain of dynein that is often deemed too flexible for informative structural studies, and may be the key to dyneins regulation.

ABS425

BIASING BINDING ORIENTATION OF THE C‐TERMINAL STRAND EXCHANGE LIMITS CHAPERONE FUNCTION IN HUMAN ALPHA‐B CRYSTALLIN

James Hebda ¹, Derrick Draeger¹, Khan Nguyen¹, Anna Nevels¹

¹Austin College (Sherman, United States)

Alpha‐B Crystallin (aBX) is a key small heat shock protein (sHSP) that is expressed in various tissue in the body, including the lens, brain, heart, and skeletal muscle. This protein plays an important role in limiting protein aggregation. In the lens of the eye, Alpha‐B and Alpha‐A Crystallin function to delay the onset of protein aggregation that leads to cataract formation. aBXs chaperone function has also been linked many other conditions including neurodegenerative diseases and cancer. aBX, like most sHSPs, forms heterogeneous large order oligomers. The dynamic nature of these oligomers has been linked to chaperone function. Studies seeking to determine of the molecular interactions governing oligomerization and chaperone function have implicated the N‐terminal domain, the crystallin domain, as well as the dynamic C‐terminal domain. The C‐terminal strand takes part in a highly dynamic strand exchange within one of the known dimer interfaces. A unique feature of the C‐terminal domain is a nine residue palindromic region that may allow for dynamic, bi‐directional strand exchange within dimers. This structural plasticity has been linked to both chaperone function and oligomer formation. Several novel mutants have here been designed to interfere with expected electrostatic interactions inferred from known crystal structures of aBX. These variants of aBX were designed to bias one orientation over the other or equally disfavor strand exchange. These mutants have been characterized to determine their chaperone function using insulin as a model aggregating protein. Results demonstrate a decrease in chaperone function when limiting exchange binding to one orientation while no significant effect when the mutation disrupts both binding orientations equally. These data support the hypothesis that the C‐terminal strand exchange plasticity plays an important role in aBX chaperone function.

ABS426

PH DEPENDENCE ON BINDING AND RELEASE OF FOLATE BY FOLIC ACID RECEPTOR

Thomas Paul ¹, Hedieh Torabifard¹, Jonah Vilseck¹, Ryan Hayes¹, Charles Brooks²

¹Department of Chemistry, University of Michigan (Ann Arbor, United States); ²Department of Chemistry and Biophysics Program, University of Michigan (Ann Arbor, United States)

Folate is a basic component of cellular metabolism, is used for DNA synthesis and repair, and is essential for rapidly dividing cancer cells. Folic acid receptor (FR) is a membrane‐bound protein held to the cells surface through a glycophosphatidylinositol anchor, which has high affinity for binding and transporting folate into cells. It has restricted expression in normal cells but is highly expressed in various nonmucinous tumors of epithelial origin. This over‐expression is thought to be a possible target for anticancer drugs and this observation is supported by the current use of antifolates in cancer chemotherapies. FR transports folate across the cell membrane through endocytosis, going from neutral to acidic pH, and delivers it to an endosome. However, the pH dependent binding and release of folate from FR is not well understood. Furthermore identification of key amino acid residues within FR responsible for the pH dependent conformational changes observed within the apo and holo protein structures will further our efforts in designing new antifolates. To that end, we have studied pKa shifts for both apo and holo forms of FR using explicit solvent constant pH molecular dynamics (CPHMDMSD). Our key findings highlight three histidine residues that have downward pKa shifts that contribute significantly to destabilizing the FR:folate complex at pH values consistent with an endosomal environment. Additionally, pH dependent binding profiles of all ionizable amino acid residues have been constructed, which elucidates the delicate balance between stabilizing and destabilizing residues within FR.

ABS427

AMP REGULATION OF BIFUNCTIONAL ADP‐DEPENDENT SUGAR KINASES FROM ARCHAEA: EVOLUTIONARY HISTORY AND KINETIC CHARACTERIZATION

Gabriel Vallejos¹, Victoria Guixe ¹, Sixto M Herrera¹, Victor Castro‐Fernandez¹

¹Departamento de Biología, Facultad de Ciencias, Universidad de Chile (Santiago, Chile)

Glucokinase (GK) and phosphofructokinase (PFK) catalyzed reactions are major points for the allosteric control in the glycolytic pathway. In Euryarchaeota, glycolysis presents unique modifications, such as the ADP‐dependence of GK and PFK activities, although in Methanococcales only one bifunctional enzyme performs both activities. To date, no allosteric properties toward classical effectors of bacterial and eukaryotic enzymes have been reported for these archaeal enzymes. Nonetheless, we found that the bifunctional enzyme from Methanococcus maripaludis (MmPFK/GK) is activated by its reaction product AMP.

To address the evolutionary history and the kinetic mechanism of the activation by AMP, we perform a comparative kinetic study that includes extant homologs enzymes from other phylogenetic branches of this family, along with four resurrected ancestors. Through steady‐state kinetics, we determine that MmPFK/GK has an ordered sequential mechanism where MgADP is the first substrate to bind to the enzyme and AMP the last product to leave. Both activities are activated by AMP and this occurs through an increase in the affinity for the sugar substrate. This also leads to an increased substrate‐inhibition through sugar binding to the free enzyme, generating a non‐productive complex. AMP activation is correlated with bifunctionality and is evolutionarily conserved in phylogenetic branches leading to bifunctional enzymes. However, this feature was lost during the evolutionary trajectory towards specific GK or PFK enzymes. The results support that AMP activation is an ancestral trait, key for the regulation of bifunctional enzymes, but not when these activities are present in separate enzymes (Fondecyt 1191321).

ABS428

CHARACTERIZING CATCH BOND CLUSTERS USING DNA ORIGAMI AND ATOMIC FORCE MICROSCOPY

Molly Mollica ¹, Olga Yakovenko¹, Nathan Sniadecki¹, Wendy Thomas¹

¹University of Washington (Seattle, United States)

Catch bonds are bonds that become longer lived when subjected to increased tensile force. Many receptors that are normally exposed to mechanical force in vivo form catch bonds, such as P‐selectin on white blood cells, glycoprotein Ib on platelets, and FimH on bacteria. Catch bonds are therefore involved in immunity, thrombosis, and infection. Knowing the function of a catch bond‐forming receptor is not sufficient to tell us the functional significance of its catch behavior. To understand the functional significance of catch bonds, it would help to understand the behaviors mediated by clusters of catch bonds. We hypothesize that clusters of catch bonds can exhibit all types of force‐dependent and independent behavior depending on cluster geometry. We also hypothesize that bond strength depends on cluster size and geometry. In order to test this hypothesis, we program clusters on a DNA origami nanostructure (Figure 1A.) This nanostructure has 40 potential cluster binding sites, which allows for varying cluster size (Figure 1C vs. 1D), geometry (Figure 1C vs. 1E), and composition (Figure 1D vs. 1F) on a given nanostructure. While nanostructures present programmed clusters of interest, an atomic force microscope (AFM) is used to measure the rupture force and bond lifetime of the clusters. Nanostructure folding conditions were optimized using agarose gel electrophoresis (AGE) and nanostructures of expected shape and size were visualized using AFM (Figure 1B). Nanostructures were purified via polyethylene glycol (PEG) purification and successful nanostructure protein functionalization and purification was observed via AGE and a Typhoon gel imager. Preliminary AFM rupture strength data support the notion that DNA origami nanotechnology is a feasible system to conduct mechanical strength measurements on programmed catch bond clusters.

ABS429

DESIGN AND VALIDATION OF DE NOVO DESIGNED PROTEIN MINI BINDERS OF RIBOSOMAL RNA SMALL SUBUNIT METHYLTRANSFERASE A FROM BURKHOLDERIA PSEUDOMALLEI

Bradley Hammerson ¹, Longxing Cao², Brian Coventry², Matt Clifton³, Jan Abendroth³, Banumathi Sankaran³, Lance Stewart², Bart Staker⁴, David Baker², Peter Myler⁴

¹Seattle Childrens Research Institute Center for Global Infectious Disease Research (Seattle, United States); ²University Of Washington (Seattle, United States); ³UCB (Bainbridge Island, United States); ⁴Seattle Children's Research Institute ‐ Center for Global Infectious Disease Research (Seattle, United States)

The bacterial pathogen Burkholderia pseudomallei is the causative agent of melioidosis a frequently fatal disease in humans occurring in tropical climates like those in Southeast Asia and Australia. Here we present the 1.75 å structure of the Ribosomal RNA small subunit methyltransferase A, loss of function mutations of which have been previously reported to confer Doxycycline (DOX) resistance to this pathogen. A collaborative project between the Institute for Protein Design at the University of Washington and the Seattle Structural Genomics Center for Infectious Disease (SSGCID) has been undertaken to utilize the structural data generated by the SSGCID to further development of de novo designed protein mini binders to anti‐microbial resistance targets. This structure has been used along with the Rosetta Macromolecular Modeling Suite, Rifdock and other related computational tools developed by the institute for protein design to guide the initial design, and subsequent structural modifications to generate improved designs with improved affinity and temperature stability. This ultimately yielded approximately 40 designed mini protein binders with sub‐micromolar affinity as measured in a Fluorescent Activated Cell Sorting (FACS) based binding assay measuring binding of a biotinylated target protein to the designs expressed on the surface of yeast cells. Assessment of the expression and feasibility of purification of a selection of the designed mini‐binders is currently ongoing. Future efforts will focus on further improvement of the affinity of the molecule which could ultimately be useful in diagnostic mass spectrometry‐based investigations of DOX resistance in B. pseudomallei and related organisms.

ABS430

NOVEL STRUCTURE OF FLAVOHEMOGLOBIN FROM MALASSEZIA YAMATOENSIS DETERMINED BY SAD PHASING

Madison Bolejack ¹, Jan Abendroth¹, David Fox III¹, Thomas Edwards¹, Peter Myler², Stephen Mayclin¹

¹SSGCID/UCB (Bainbridge Island, United States); ²SSGCID (Seattle, United States)

We have determined the first flavohemoglobin structure from the Malassezia yamatoensis using single wavelength anomalous dispersion (SAD) phases from the heme‐iron atom buried within the protein. Members of the Malassezia fungal genus are responsible for a variety of skin diseases such as pityriasis versicolor and seborrheic dermatitis. Some strains of Malassezia are resistant to an array of azole‐based compounds, the primary course of treatment for fungal infections. Flavohemoglobins play a role in defense against nitric oxide (NO) toxicity. The flavin moiety functions as a dioxygenase, shuttling electrons from NADH to the heme and reducing the concentration of free NO. The binding of compounds, particularly azoles, to the flavin moiety has been shown to disrupt electron transfer, leaving the organism vulnerable to NO‐mediated toxicity. The crystal structure of M. yamatoensis flavohemoglobin, as well as potential future structures in the presence of azole‐based compounds, may shed light on the resistance mechanism of Malassezia species, and aid in development of improved fungal treatments. In this experiment, the protein in complex with its heme cofactor was crystallized in a polyethelyene glycol (PEG)‐based solution at 14C at a 1:1 ratio of protein to crystallant, resulting in red‐colored crystals that diffracted to 1.70 å resolution. The structure was phased using anomalous signal originating from the iron atom in the heme cofactor, refined in PHENIX, and deposited into the Protein Databank (PDB) with PDB code 6O0A.

ABS431

DISCERNING THE BIOCHEMICAL FUNCTION FOR THE CATALYTIC DOMAIN OF THE PLASMODIUM BEM46‐LIKE PROTEIN (PBLP)

Anna Groat Carmona¹, Koryn Aguon ¹, Koryn Aguon¹, Misaki Seto¹

¹University of Washington Tacoma (Tacoma, United States)

Malaria is a mosquito‐borne disease caused by the Plasmodium parasite, which infects approximately 216 million people worldwide and kills an estimated 445,000 people each year. Although malaria can be treated, the rise in drug resistance has created a need for new intervention strategies, including the identification of new drug targets like the Plasmodium BEM46‐like protein (PBLP). PBLP is a highly conserved, uncharacterized /‐hydrolase that plays an important role in modulating the formation of invasive‐stage parasites and is expressed throughout the parasite life cycle. PBLP shares structural homology and amino acid identity with other BEM46‐like proteins in the /‐hydrolase superfamily. Previously, a cDNA copy of the pblp coding sequence was amplified and then mutated using overlap‐extension PCR to disrupt the putative catalytic active site (Ser153, Asp229, and His258), which was predicted using bioinformatic analyses. PBLP is membrane bound in liver‐stage and blood‐stage parasites so the hydrophobic transmembrane domain within the PBLP protein sequence was excluded during cloning so that subsequent protein purification procedures could be done in solution. Ultimately, our goal is to express this parasitic protein in a bacterial system by successfully ligating the wild‐type and mutant pblp sequences into a bacterial protein expression vector so that it can be purified, and its catalytic activity can be characterized using functional enzymatic assays. These results will enable a better characterization of PBLP and potentially lead to the development of novel drug targets for the prevention and treatment of malaria.

ABS432

CHARACTERIZATION OF MECHANISMS INVOLVED IN SBP1 REVERSIBLE PROTEIN AGGREGATION ON SACCHAROMYCES CEREVISIAE

Jesus Ruiz Flores ¹, Francisco Torres Quiroz¹

¹National Autonomous University of Mexico (Tlahuac, Mexico)

The cell has developed different mechanisms for facing adverse conditions during their life cycle. One of these strategies is the formation of ribonucleoprotein (RNP) complexes or granules, composed by mRNA and proteins that helps the cell to repress translation and protect those vital mRNAs until more favourable conditions. These granules can be found in both the nucleus and the cytoplasm of the cell. In yeast, Saccharomyces cerevisiae, one of such stressful condition for the cells is the nutrient depravation, in response of this; the cell forms some RNPs known as processing bodies (P‐bodies). The P‐bodies are aggregates of translationally repressed mRNA and are composed of several proteins related to mRNA decay machinery. One of this proteins is the single‐stranded nucleic acid‐binding protein (Sbp1). It has been demonstrated that Sbp1 remains soluble in the nucleus and cytoplasm during exponential growth, and under stressful conditions, such as lack of glucose, it aggregates in cytoplasmic RNPs. In this work we observed that in S. cerevisiae, some amino acids in Sbp1 regulate the aggregation of this protein. When such amino acids were mutated, the cells showed more granules during glucose deprivation.

This work was supported by grant IN209219 from PAPIIT‐DGAPA, UNAM.

ABS433

STRUCTURAL INSIGHTS INTO THE EVOLUTION OF THE CAZY GT8 GLYCOSYLTRANSFERASE GLYCOGENIN

Hyun Woo Kim ¹, Msano Mandalasi¹, Zachary Wood¹, Christopher West¹

¹Department of Biochemistry and Molecular biology, University of Georgia (Athens, United States)

Protein glycosylation is a post‐translational modification which mediates diverse biological functions including protein folding, protein‐protein interactions, and cell signaling. Glycan synthesis is carried out by glycosyltransferases (GTs), a class of enzymes that transfer a sugar moiety onto an acceptor substrate. Despite the extensive documentation of GTs functions, structural elements by which GTs recognize a sugar‐nucleotide donor or an acceptor substrate is poorly understood for many. Here, we examine the architecture of a sugar‐nucleotide binding pocket and its associated acceptor glycan channel in an evolutionary context. The CAZy GT8 family GT Gat1 is an 3‐galactosyltransferase that modifies a terminal ‐glucose residue of a tetrasaccharide on the E3(SCF)ubiquitin ligase subunit Skp1 in protists, and is an apparent evolutionary predecessor of glycogenin, an 4‐glucosyltransferase that modifies ‐glucan termini of glycogen in yeast and animals. Our new crystal structure of Gat1 and existing structures of glycogenin were examined to compare the binding mode of their substrates. While both GTs share a remarkable similarity in structure and have conserved modes of binding the UDP moiety of their sugar‐nucleotide donor, differences in the sugar binding pockets provide specificity for either glucose or galactose. The acceptor binding for both GTs depends on a groove leading into the active site, and the sidechains in this region provide complementary interactions for binding their respective glycan acceptors. This example proposes a minimal set of amino acid substitutions in the active site, in the context of a common enzyme scaffold, that can mediate the diversification of GTs and glycans.

ABS434

THE PROTONATION STATE OF AN EVOLUTIONARILY CONSERVED HISTIDINE MODULATES DOMAIN SWAPPING STABILITY OF THE DNABINDING DOMAIN OF HUMAN FOXP1

Exequiel Medina¹, Jorge Babul ¹, Ricardo Coñuecar¹, Cesar A. Ramírez‐Sarmiento²

¹Departamento de Biología, Facultad de Ciencias, Universidad de Chile (Santiago, Chile); ²Institute for Biological and Medical Engineering, Pontificia Universidad Católica de Chile (Santiago, Chile)

Forkhead box P (FoxP) proteins are versatile Fox transcription factors that control the spatiotemporality of expression of multiple genes during cell development, immunity and tissue homeostasis. Different from other Fox proteins, FoxP members can form domain‐swapped dimers through their DNA‐binding domains, allowing the spatial organization of distant chromosome elements by bridging two DNA molecules together. Furthermore, FoxP proteins contain a histidine residue in helix H3 (H59) that is only conserved in FoxM/O/P subfamilies, possibly responsible for structural changes modulated by pH. Here, we explore the consequences of folding and dimerization of the forkhead domain of human FoxP1 due to protonation changes in H59. Using dissociation kinetics and equilibrium unfolding experiments, we demonstrated that the protonation of this residue leads to destabilization of the domain‐swapped dimer by increasing the free energy difference between the monomeric and the transition state of the dimerization process. Anisotropy changes, using singlecysteine mutants within helix H3 and helix H5 to attach a fluorescent dye, showed that pH changes only have an impact on the stability of helix H3, highlighting the relevance of H59 in the dynamics and local stability of this secondary structure element. Finally, molecular dynamics simulations suggested that the local impact in the helix stabilization is due to a hydrogen bond formed between N4 of H59 and a neighboring backbone CO of the helix. These findings will be relevant to determine the evolutionaryconserved side chains that contribute to the modulation of domain swapping dimerization in FoxP proteins. Funding: Fondecyt 1170701.

ABS435

STRUCTURE‐BASED DESIGN OF SELECTIVE INHIBITORS AGAINST THE BACE PROTEIN FAMILY

Emma Lendy ¹, Emilio Cardenas², Yu‐Chen Yen³, Arun Ghosh², Andrew Mesecar¹

¹Department of Biochemistry, Purdue University (West Lafayette, United States); ²Department of Chemistry, Purdue University (West Lafayette, United States); ³Department of Biology, Purdue University (West Lafayette, United States)

The two members of the BACE (Beta Amyloid precursor protein Cleaving Enzyme) protein family have high sequence homology yet contribute to two very different diseases. BACE1 contributes to the development of Alzheimers Disease, while BACE2 can exacerbate the progression of Type II Diabetes. Knockout studies in mice show that inhibiting both BACE1 and BACE2 leads to serious developmental delays and increased mortality in mice. Therefore, it is imperative that drug molecules being designed to treat Alzheimers or Diabetes be selective for either BACE1 or BACE2. X‐ray crystallography, synthetic medicinal chemistry, enzyme kinetics and cell‐based studies are being used to design selective BACE1 and BACE2 inhibitors.

Selectivity for BACE1 or BACE2 is being evaluated through a unique work flow process that was developed in our lab for inhibitor evaluation that considers tight binding kinetics. Over 100 compounds have been analyzed throughout this study, many of which possess sub‐micromolar affinity for either BACE1 or BACE2. New compounds show strong inhibition of both BACE1 and BACE2. Conversely, other compounds have been shown to have a 63‐fold specificity for BACE1, or 220‐fold specificity for BACE2. These data are being used to develop structure‐activity relationship (SAR) to guide the synthesis of new inhibitors for both BACE1 and BACE2. Altogether, this research aids in the development of potent and specific inhibitors for the treatment of Alzheimers Disease and Type II Diabetes.

ABS436

TOWARDS SYNTHETIC ALLOSTERIC TRANSCRIPTIONAL MODULATORS: DEFINING THE ROLE OF CONFORMATIONAL ENTROPY IN COACTIVATOR COMPLEXES

Amanda Peiffer ¹, Charles Brooks III¹, Anna Mapp¹

¹University of Michigan (Ann Arbor, United States)

Much of the assembly process of the transcriptional machinery is governed by transient and dynamic protein‐protein interactions (PPIs) that defy standard characterization strategies. Transcriptional coactivators are the hubs of this process, interacting with transcriptional activators, epigenetic modulators, and other coactivators to assemble the transcriptional machinery. Coactivators are also a molecular recognition enigma, as the diverse array of binding partners is recognized using a much smaller number of conformationally dynamic domains. The mechanistic signatures that underpin this multi‐partner yet selective binding profile are not fully understood. Furthermore, despite the important functional role coactivators play in healthy and diseased organisms, they are often deemed undruggable, with few synthetic probes having been validated for mechanistic studies. My research addresses the urgent need for a deeper understanding of the molecular recognition surrounding the assembly of dynamic PPIs through a computational and experimental platform to ultimately define the role of conformational entropy in allostery regulation of transcriptional coactivators. Specifically, I utilize molecular dynamics simulations to dissect the thermodynamic properties governing allosteric communication in the activator binding domain of KIX, a domain of the master coactivator protein CBP. Through these studies, I identify conformational microstates of KIX that preclude favorable interactions with native binding partners including MLL, c‐Myb, and CREB, thus opening an avenue for targeted drug discovery through exploitation of native allostery.

ABS437

ANALYSIS OF CALMODULIN‐INTERACTING PROTEINS CAPTURED IN LIVE CELLS BY PHOTOACTIVATED CROSS‐LINKING: EVIDENCE FOR AN ACTIVE CA2+ SIGNALING MICRODOMAIN

Anthony Persechini ¹, DJ Black¹, Quang‐Kim Tran², Andrew Keightley¹, Ameya Chinawalker¹, Cole McMullin¹

¹University of Missouri at Kansas City (Kansas City, United States); ²Des Moines University Osteopathic Medical Center (Des Moines, United States)

We have a developed a new method for real time evaluation of signaling through calmodulin (CaM) interactomes in live cells. Proteins interacting with expressed 6‐His/FLAG tagged CaM in cells metabolically labeled with a photoreactive methionine analog are captured by rapid (t1/2 ~7 sec) photoactivated cross‐linking. Adducts with tagged CaM are stringently enriched, followed by identification and relative quantification of captured proteins based on mass spectrometry. A set of 511 proteins that includes 31 known CaM binding proteins (CBPs) was thus derived from a human cell line stably expressing tagged CaM. The estimated mole ratios of tagged CaM to the sums of known CBPs (11.7±1.2) or all captured proteins (1.2±0.1) are consistent with hundreds of background interactors, which appear to be captured in lower fractional amounts than known CBPs. A CBP enriched group of 108 proteins captured in fractional amounts 8‐fold or more above the median value was analyzed further. Capture of 11 proteins is decreased by brief chelation of extracellular Ca2+, which does not affect the cytosolic free Ca2+ concentration. Capture of these proteins is unchanged at the peak free Ca2+ concentration produced by ionomycin, unless preceded by chelation, whereupon it restores capture to untreated or higher levels. Eight of these proteins are known CBPs, including IP3 receptors (ITPR1, 2, 3), wolframin (WFS1), PMCA pumps (ATP2B1, 4), CaM kinase I (CAMKI) and calcineurin (PPP3CA). The identities and capture dependencies of these proteins suggest that they reside in an active Ca2+ signaling microdomain within an ER‐PM junction.

ABS438

THE DEVELOPMENT OF E. COLI EXPRESSION SYSTEM FOR G PROTEIN‐COUPLED RECEPTORS

Nanao Suzuki ¹, Yuuki Takamuku¹, Chika Yoshida¹, Takeshi Murata¹

¹Graduate School of Science, Chiba University (Chiba, Japan)

G protein‐coupled receptors (GPCRs) are one of the most important families of drug target. However, many GPCRs are unstable and have difficulty with expression and purification. To increase the thermal stability of GPCRs, three strategies are adopted: 1. deletion of N terminus, C terminus and long loops, 2. insertion of a fusion protein like apocytochrome b562RIL, T4 lysozyme, or the catalytic domain of Pyrococcus abyssi glycogen synthase into the third intracellular loop, and 3. addition of thermal‐stabilized mutations. These modifications are needed for many GPCRs and require much effort. If these experiments are performed quickly and easily, less effort is required. We focus on developing Escherichia coli (E. coli) expression system for GPCRs to determine the stability of GPCR mutants. E. coli expression has three benefits, quick, easy, and low costs, while it suffers from the disadvantage of low expression of unstable membrane proteins. We tried to utilize this unfavorable property of E. coli expression system for screening of high expressed and thermal stabilized GPCR constructs.

In this research, we selected the human adenosine A2A receptor (A2AR) and human serotonin 2A receptor (5‐HT2AR) as model target GPCRs. We modified E. coli expression pET‐vector for GPCRs. We expressed A2AR and 5‐HT2AR that red fluorescent protein (RFP) was fused with at the C‐terminus using the modified vector, and optimized the expression conditions by monitoring RFP fluorescence. We measured the expression level changes with/without fusion protein, mutations, and ligands. In addition, we applied thermal stabilization evaluation using modified CBB clear‐native PAGE [1] to A2AR‐RFP expressed in E. coli.

[1] N. Suzuki et al., Anal Biochem., 2018, 548, 7‐14.

ABS439

INTRA‐MELANOSOMAL DOMAINS OF HUMAN RECOMBINANT TYROSINASES PRONE TO PROTEIN AGGREGATION AT PHYSIOLOGICAL TEMPERATURES

Monika Dolinska ¹, Claudia Kassouf¹, Paul Wingfield², Yuri Sergeev¹

¹National Eye Institute/NIH (Bethesda, United States); ²National Institute of Arthritis and Musculoskeletal and Skin Diseases/NIH (Bethesda, United States)

Tyrosinases are melanocyte‐specific enzymes involved in melanin synthesis pathway. Tyrosinase (Tyr) and Tyrosinase Related Protein 1 (Tyrp1) exhibit significant sequence homology and structural similarity of the active domains, maintaining conserved binding sites essential for their catalytic activity. Tyr catalyzes the hydroxylation of L‐tyrosine to L‐DOPA and the oxidation of L‐DOPA to dopaquinone, as well as the oxidation of 5,6‐dihydroxyindole to indole 5,6‐quinone, whereas both, Tyr and Tyrp1 catalyze the oxidation of 5,6‐dihydroxyindole‐2‐carboxylic acid (DHICA) into indole‐5,6‐quinone‐2‐carboxylic acid, thereby playing an important role in pigment formation. Recently, we showed that the Tyrp1‐mediated protection of human tyrosinase activity does not involve stable interactions of tyrosinase domains. Here, the recombinant intra‐melanosomal domains of Tyr and Tyrp1 were studied in vitro at different temperatures to define their oligomeric states and thermostability. Proteins were expressed in T. ni larvae and purified. The tyrosinases properties were studied using gel filtration (GF), dynamic light scattering (DLS), and sedimentation velocity (SV) methods. In addition, we used molecular modeling to better elucidate the mechanism of human TYRP1 and DHICA binding. In the experiments in vitro, both tyrosinases present the highest enzymatic activity at 37°C. In silico, Tyrp1 demonstrates the best DHICA binding, as well. However, GF, DLS, and SV show increased protein aggregation above 35°C. This suggests that recombinant intra‐melanosomal tyrosinase domains have decreased stability at this condition. In human melanosome, posttranslational modifications, the melanosomal lipid membranes, or protein‐protein interactions might play a role in tyrosinases protein stabilities and the resistance to tyrosinases aggregation at physiological temperatures in vivo.

ABS440

ENGINEERING SCFV ANTIBODY AGAINST CONSERVED REGIONS OF DENGUE VIRUS ENVELOPE PROTEIN

Abhishek S Rathore ¹, Rinkoo D Gupta²

¹Colorado State University (FORT COLLINS, United States); ²South Asian University (New Delhi, India)

Dengue is one of the deadliest mosquito‐borne viral diseases of the tropical regions of the world and is caused by dengue virus (DV). Till date, there is lack of an effective vaccine against DV due to a phenomenon called antibody‐dependent enhancement (ADE) of disease. Our study focuses on development of a scFv antibody that can bind and neutralize all four serotypes of DV. For this, two highly conserved regions in domain II of dengue envelope protein, Fusion loop and Bc loop were taken as a target for scFv antibody development. Fusion and Bc loop were used to make a small antigenic protein FuBc. The structures of scFv and FuBc proteins were created by homology modelling, and the interaction between them was checked using ZDock. The interaction of these two proteins were good, however, these interactions were further improved by creating an in silico scFv mutant library where CDR regions of scFv were mutated. The best mutations that enhance the binding of scFv and stabilizes scFv‐FuBc complex were further evaluated by molecular dynamics simulations. scFv mutant 1 was best mutant with mutation energy of ‐9.573. To validate the scFv‐FuBc interaction, these proteins were expressed in a bacterial expression system. Interaction of the scFv‐FuBc proteins was then analysed by surface plasmon resonance, which gave a KD value of 2.03 μM showing significant interaction between scFv‐FuBc proteins. However, for the development of a neutralizing antibody scFv mutant 1 could be used as a candidate to improve scFv binding efficiency even further.

ABS441

A RATIONALLY DESIGNED PLANT‐PRODUCED IGA HAS IMPROVED YIELD AND EXHIBITS CROSS SEROTYPE PROTECTION AGAINST ENTEROHEMORRHAGIC ESCHERICHIA COLI

Adam Chin‐Fatt ¹, Rima Menassa¹

¹Western University (London, Canada)

The seven most prevalent strains of Enterohemorrhagic Escherichia coli (EHEC) collectively comprise more than 95% of the disease burden globally, affecting an estimated 2.8 million people annually. Although a plant production system is well established as a useful platform for enabling the post‐translational modifications necessary for IgA folding, yield continues to be the most significant hurdle preventing transitioning of these therapeutics to market. We identified a series of single domain antibodies that can enable cross‐serotype protection against EHEC. Considering the modular nature of the IgA, we rationalized that we could engineer the Fc component for improved yield of Fc fusions with these single domain antibodies without impacting the folding of these binders. We have successfully engineered a more stable Fc, by supercharging and introducing de novo disulfide bonds, that shows higher yield in planta by three to four fold. Using immunofluorescent labelling, we also have demonstrated that the VHH is still able to exhibit cross‐serotype binding and neutralization across four of the seven strains. Co‐immunoprecipitation experiments indicate that the rationally designed mutations have also not impacted the Fcs ability to structurally assemble with other subunits into its secretory form. Overall, this study provides a proof of concept that stability engineering of plant‐produced IgA biologics is a viable strategy for overcoming the yield hurdle of plant‐produced antibodies and related therapeutics.

ABS442

STRUCTURE OF THE PR DOMAIN FROM PRDM3 AND ITS FUNCTION IN ACUTE MYELOID LEUKEMIA

Sharon Loa ¹, Tung‐Chung Mou², Kelly McGlynn³, Archibald Perkins³, Stephen Sprang², Klara Briknarova¹

¹Department of Chemistry and Biochemistry, University of Montana (Missoula, United States); ²Center for Biomolecular Structure and Dynamics, University of Montana (Missoula, United States); ³Department of Pathology and Laboratory Medicine, University of Rochester Medical Center (Rochester, United States)

PRDM3 is a transcription factor that is essential for long‐term hematopoietic stem cell function and also for transformation and sustenance of a subset of acute myeloid leukemias (AML). In its N‐terminal region, PRDM3 contains a derivative of a SET domain termed the PR domain. SET domains are characterized by histone lysine methyltransferase activity, but the enzymatic activity of the PR domain from PRDM3 is controversial. To provide framework for our functional studies, we determined the structure of the PR domain from PRDM3 by X‐ray crystallography. We then used serial replating assays to investigate the role of the PR domain in leukemogenesis, and we showed that PRDM3 proteins containing Y111A or Y175A mutations are ineffective at oncogenic transformation. These two residues are located in a deep pocket of the PR domain where the histone substrates are expected to bind. Our results demonstrate that the PR domain is critical for the leukemogenic function of PRDM3 and is, therefore, a novel target for AML therapeutic intervention. In addition, we investigated the interactions between the PR domain and its previously reported substrates, S‐adenosyl‐L‐methionine (SAM) and peptides from histone H3, using NMR spectroscopy. Only small peak shifts were observed in NMR spectra of the PR domain upon addition of SAM and H3 peptides, indicating that the interactions are very weak and/or nonspecific and, hence, may not be physiologically relevant. Our results suggest the PR domain from PRDM3 may require additional binding partners for enzymatic activity and/or may function through a different molecular mechanism that remains to be elucidated.

ABS444

MUTATION‐BASED TUNING OF THE RICE CYCLOPHILIN LRT2

Nathan Korson ¹, Lucila Andrea Acevedo¹, Linda Nicholson¹

¹Cornell University (Ithaca, United States)

The objective of this study was to quantify the activity and specificity of the rice cyclophilin LRT2 and of various LRT2 mutants to lay the groundwork for tuning the activity of this enzyme and observing the resulting phenotype in the rice plant. The natural LRT2 substrate OsIAA11, a transcription repressor protein that regulates genes involved in lateral root initiation, was employed for these studies. Although prolyl isomerase enzymes (such as cyclophilins) and the cis‐trans proline switches they act on are conserved from bacteria to humans, specific in vivo functions of this class of molecular switch have only recently come to light. Still missing are quantitative characterizations of the impact of specific cis‐trans isomerization rates on biological processes in living cells, and on phenotypes in whole organisms. LRT2 is essential for lateral root initiation in rice, thereby providing an opportunity for establishing correlations between LRT2 activity and phenotype. Nine site‐specific LRT2 mutants were generated, fluorescence spectroscopy was employed to identify the most thermally stable mutants, and NMR spectroscopy was applied to measure the cis‐trans isomerization rates for five stable mutants. Substrate affinity was measured for three mutants of particular interest. NMR lineshape analysis was applied to determine the four‐state reaction cycle parameters for wild type LRT2 and its P125K mutant. These studies identify several thermally stable LRT2 mutants with reduced activity. These mutants can now be incorporated into the endogenous LRT2 gene using CRISPR/Cas9 technology to investigate the impact of this isomerization activity on lateral root development.

ABS445

CONFORMATIONAL DYNAMICS OF DEUBIQUITINASE A IN REGULATION AND SUBSTRATE SPECIFICITY

Ying Li ¹, Ashish Kabra¹, Efsita Rumpa¹

¹University of Louisville (Louisville, United States)

Ubiquitination is a prevalent form of post‐translational modification that controls many cellular processes in eukaryotes. Deubiquitinase A (DUBA) is a protease responsible for deconjugation of poly‐ubiquitin chains from target proteins. It plays functional roles in innate immune responses and T cell regulation in humans. DUBA is a member of the deubiquitinase (DUB) family, which has been implicated in a wide range of human diseases, including cancer, neurodegenerative diseases, inflammatory and autoimmune diseases. The functional roles of conformational dynamics in DUB family are not well defined, though previous studies have hinted at the importance of dynamics in both regulation of enzyme activity and recognition of poly‐ubiquitin chains of specific linkages. We have combined solution NMR spectroscopy, biochemical assays and molecular dynamics simulations to define the structural and dynamic basis of DUBA activation by phosphorylation and substrate specificity. The NMR data suggest that the conformational energy landscape of DUBA is altered by phosphorylation despite the minimal structural changes. Biochemical assays suggest that the enhancement in the catalytic turnover rate is responsible for the activation of DUBA rather than the increase in substrate affinities. NMR studies and molecular dynamics simulations have been carried out on the enzyme‐substrate complex to understand the differences in dynamic properties between the active and inactive forms of DUBA. Overall, our data suggest that modulation of dynamic properties underlies the activation of DUBA and motions in the enzyme‐substrate complex are essential for selection of substrates.

ABS446

SINGLE‐MOLECULE FORCE SPECTROSCOPY REVEALS COOPERATIVE INTERFAICAL METAL SITES IN HUMAN ANTIBACTERIAL PROTEIN S100A12 HOMODIMER

peng zheng ¹

¹nanjing university (Nanjing, China)

S100A12 (A12) is a type of human antibacterial protein with high metal chelating ability, which is a critical player in the so‐called "nutritional immunity". A12 homodimer binds tightly with zinc ions using its two interfacial AspHis3 metal binding sites that scavenges the zinc ion from bacterial growth environment and starve bacterial to death. And this homodimer form with two metal binding sites is the basic structural and functional unit for this unique anti‐bacterial protein. Here, we apply AFM‐based single molecule force microscopy (SMFS), to study the strength of these two metal sites. By engineering a suitable A12 homodimer sample, we ruptured the two zinc binding sites during the protein homodimer dissociation process and measured their strength. We discover that the break of the two sites only shows one force peak with a rupture force of ~90 pN in the WT construct. Moreover, the measurements on two A12 mutants without the internal or the external metal site, respectively, showed two different stabilities. The mutant with only the external metal site shows a high force of 70 pN while the one with the internal site is much weaker, of ~30 pN. Thus, the two metal sites in S100A12 stabilize each other and behavior as cooperative functional sites.

ABS447

ALTERING THE CONFORMATIONAL SPECIFICITY OF DNA BINDING PROTEINS

Seul Ki Lee ¹, Chan Yang Park¹, Chaehee Park², Hee‐Jung Choi², Yang‐Gyun Kim¹

¹Sungkyunkwan University (Suwon, South Korea); ²Seoul National University (Seoul, South Korea)

A family of proteins contains a domain(s) that binds to left‐handed Z‐DNA with a high affinity, termed Z domains. The structures of all Z domains solved so far have the winged helix‐turn‐helix (wHTH) motif to bind and stabilize Z‐DNA. This is astounding that many B‐DNA binding proteins also share a very similar wHTH motif. In this presentation, the globular domain of Histone H5 (GH5), a canonical B‐DNA binding protein, was engineered to be a Z‐DNA binding protein. The structure of the globular domain of histone H5 (GH5), one of wHTH motif domains, exhibits a high homology to that of hZADAR in the three‐dimensional structure, although there is no similarity in amino acid sequence. The C‐terminal region of GH5 was first extensively mutagenized to resemble that of hZADAR. Subsequently, the number of amino acid residues originated from hZADAR was gradually reduced from the mutations. A series of GH5 mutants were created to bind Z‐DNA. Among them, the mutant having only 8 amino acids different from the wild‐type GH5 almost equals hZADAR in B‐to‐Z conversion activity as well as Z‐DNA binding. Its crystal structure with Z‐DNA clearly demonstrates that key interactions observed in the structure of hZADAR/Z‐DNA complex still play crucial roles for Z‐DNA recognition and stabilization. In conclusion, the wHTH fold of GH5 could offer structural grounds for creating a novel conformational‐specific DNA binding protein. [This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2018R1D1A1B07048910).]

ABS448

PROTEIN‐INDUCED STRUCTURAL DEVIATIONS OF Z‐DNA

Hyuk Won ¹, Chaehee Park², Ji‐Ye Yun¹, Young Eun Won¹, Hee‐Jung Choi², Yang‐Gyun Kim¹

¹Sungkyunkwan University (Suwon, South Korea); ²Seoul National University (Seoul, South Korea)

Unlike B‐DNA, Z‐DNA has a left‐handed double helix and a unique zig‐zag sugar phosphate backbone by alternating anti‐ and syn‐ conformation of the nucleotides. Most Z‐DNAs tend to have repetitive purine‐pyrimidine sequences, but non‐purine‐pyrimidine structures have also been reported. Although Z‐DNA is thermodynamically unstable compared to B‐DNA, but can be stabilized through Z‐DNA binding proteins under physiological conditions. For example, hZADAR1, a Z‐DNA binding domain from human ADAR1, interacts with the Z‐DNA backbone in a conformation‐specific manner. Here, we present the crystal structure of d(CCCGGG)2, another non‐purine/pyrimidine repeat in Z‐form, complexed with hZADAR1. The 3.4 resolution structure was solved by molecular replacement method using the hZADAR1 domain structure as a search model (1QBJ). An initial map clearly showed the Z‐conformation of d(CCCGGG)2. In this complex structure, we observed a remarkable structural change in Z‐DNA structure and unidentified new interactions between Z‐DNA backbone and Z domain mediated by His159. The results of circular dichroism (CD) and microscale thermophoresis (MST) support the role of His 159 for Z‐DNA stabilization. [This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2018R1D1A1B07048910).]

ABS449

COARSE‐GRAINED PROTEIN MODELLING WITH SURPASS

Dominik Gront ¹, Justyna Kryś¹

¹University of Warsaw Faculty of Chemistry (Warsaw, Poland)

The recently published SURPASS (Single United Residue per Pre‐Averaged Secondary Structure fragment) model has been proposed to facilitate fast modeling of large proteins [1,2]. The design of the model is unique and strongly supported by the statistical analysis of structural regularities characteristic for protein systems. Coarse‐graining of protein chain structures assumes a single center of interactions per residue and accounts for preaveraged effects of four adjacent residue fragments. Knowledge‐based statistical potentials encode complex interaction patterns of these fragments. Using the Replica Exchange Monte Carlo sampling scheme and a generic version of the SURPASS force field we performed test simulations of a representative set of single‐domain globular proteins. Hierarchical clustering of produced structures shows [3], that the method samples a significant part of conformational space and reproduces protein structures, including native‐like, with surprisingly good accuracy.

[1] Dawid, A., Gront, D., Koliski, A. "SURPASS Low‐Resolution Coarse‐Grained Protein Modeling", Journal of Chemical Theory and Computation 13(11), 5766‐5779, 2017.

[2] Dawid, A., Gront, D., Koliski, A. "Coarse‐Grained Modeling of the Interplay between Secondary Structure Propensities and Protein Fold Assembly", Journal of Chemical Theory and Computation 14 (4), 22772287, 2018.

[3] Gront, D., Hansmann UHE., Koliski A. "Exploring protein energy landscapes with hierarchical clustering", International Journal of Quantum Chemistry, 105:826830, 2005

ABS450

AN EVOLUTIONARILY CONSERVED MECHANISM OF AMYLIN MISFOLDING IN TYPE 2 DIABETES

Caitlyn Fields ¹, Justin Lomont¹, Kacie Rich¹, Sidney Dicke¹, Megan Petti¹, Martin Zanni¹

¹University of Wisconsin, Madison (Madison, United States)

Islet amyloid polypeptide (IAPP) is a 37 residue peptide that misfolds and aggregates into fibrils. These fibrils are deposited as plaques in pancreatic islet tissue in 90‐95% of patients with type 2 diabetes. Species dependent sequence differences in residues 20‐29 (the FGAIL region) appear to dictate the tendency to aggregate in vitro and the potential to contract type 2 diabetes. We use 2D IR spectroscopy and isotope labeling to study the aggregation mechanism of IAPP from seven different animal species. For the species susceptible to type 2 diabetes, we observe a transient oligomeric intermediate on route to fiber formation. This oligomeric structure is not observed in IAPP of species that do not contract amyloid‐associated type 2 diabetes. We present 2D IR data probing the structure of this toxic intermediate along with toxicity studies that correlate the presence of this intermediate with beta‐cell misfunction. The data supports the hypothesis that this intermediate is the source of cytotoxicity in amylin aggregation.

ABS451

INVESTIGATING THE FITNESS EFFECT OF MUTATIONS IN THE DIPHTHAMIDE HISTIDINE OF HUMAN ELONGATION FACTOR 2

John Weldon ¹, Brian Masters¹, Nadim Alkharouf¹, Benjamin Atha¹, Jack Sanford¹, Lauren Russell¹, John Cyprien¹

¹Towson University (Towson, United States)

Elongation factor 2 (EF2) and its bacterial homolog EFG are GTPases that are required during translation for the movement of ribosomes along mRNA. Both eukaryotic and archaeal EF2, but not EFG, contain a unique post‐translational modification called diphthamide that is not found in any other protein. The diphthamide modification is highly conserved, occurs only at a specific histidine (His715 in mammals), and requires a complex three‐step synthesis pathway involving at least seven different proteins. It is also the target of several bacterial toxins (diphtheria toxin, Pseudomonas exotoxin A, and cholix toxin), which block its function to inhibit protein synthesis. Circumstantial evidence surrounding the diphthamide suggests that it plays an important role in translation, but its precise function remains elusive. We have generated a library of mutations at the diphthamide histidine, stably expressed them in HEK293 cells, and evaluated the influence of those mutations on the survival of the transfected cells using Illumina high‐throughput sequencing. We observed large variation in synonymous codon preference, but no significant correlation with human codon usage. A single codon for alanine (GCT) dominated the final population of cells. There was a significant negative correlation between the volume of the mutant amino acid and survival. Work is underway to evaluate the mutation library in additional cell lines.

ABS452

AN INACTIVATED DIMERIC VIBRIO ALKALINE PHOSPHATASE CONVERTS TO A STATE WITH A DIFFERENT PROMISCUOUS ACTIVITY

Jens Gudmundur Hjörleifsson¹, Bjarni Ásgeirsson ¹, Kristófer Arnar Eiríksson¹, Bjarni Ásgeirsson¹

¹University of Iceland (Reykjavik, Iceland)

Alkaline phosphatases (AP) of the phoA phenotype are all functional as homodimeric enzymes. Vibrio splendidus AP (VAP), a cold‐active variant has been shown to adopt an inactive dimeric state in terms of para‐nitrophenyl phosphate (pNPP) hydrolysis, induced by low heat (room temperature) or low urea concentrations (1.0 M). The native and active dimeric state of the enzyme is stabilized by anions such as chloride, presumably by binding to one of the active site Zn2+. The inactive dimeric state resembled closely the native state, having similar CD and Trp fluorescence spectra. However, when a fluorescent probe (bimane) was attached to selected sites at the dimer interface, a subtle change in solvent exposure was detected. This indicated that the inactive dimeric form might have looser packing at the two‐fold symmetry axis. A great acceleration in promiscuous hydrolysis of para‐nitrophenyl sulfate (pNPS) was observed as the enzyme made its transition into the inactive phosphatase state. This property of rate enhancement and discrimination between these two substrates via a conformational change has not been observed in other APs such as the E. coli AP variant.

ABS453

NONLINEAR PARTIAL CORRELATION TO IDENTIFY CONTRIBUTION OF RESIDUES IN GLOBAL CONFORMATIONAL DYNAMICS

Amitava Roy ¹, Michael Bender¹, Rong Yang¹, Paula Lei¹, KC Cheng¹, Frank Arnold¹

¹NIH (Hamilton, United States)

Correlated motions are inherent to proteins and can be essential for proteins to carry out their function. Characterizing the dynamics of multidomain proteins in terms of positional fluctuations and correlated motions using molecular dynamics (MD) simulation is a powerful and often practiced first step toward elucidating molecular behavior and function and allostery. We assess the ability of distance correlation (Dcor) for detecting long‐range concerted motion in proteins. We further investigate contribution of each residue involved in the low frequency concerted motions in proteins using partial distance correlation (pDcor). We demonstrate the usefulness of Dcor for assessing the conformational changes in an antigen‐binding (Fab) fragment of an antibody and successfully identify the residues contributing to the conformational changes using pDcor. Our in‐silico results explains the conformational changes of different Fab constructs studied using size exclusion chromatography.

ABS454

HX TO MEASURE MEMBRANE PROTEIN ELECTROSTATICS

Esther Brielle ¹, Isaiah T Arkin²

¹The Alexander Grass Center for Bioengineering. The Hebrew University of Jerusalem, Edmond J. Safra Campus, Jerusalem, 91904, Israel. (Jerusalem, Israel); ²The Alexander Silberman Institute of Life Sciences. Department of Biological Chemistry. The Hebrew University of Jerusalem, Edmond J. Safra Campus, Jerusalem, 91904, Israel. (Jerusalem, Israel)

Proteins are composed of a heterogenous electrostatic environment. Of particular note are amphipathic membrane protein helices that arrange to form a polar channel lumen and a nonpolar membrane interface, which allow for transporting polar substances across the hydrophobic membrane bilayer. Being able to measure the local polarity of a given site along the transmembrane protein is of particular interest for investigating protein structure and function. Here we develop a protocol to measure the local environment electrostatic field along the transmembrane helix backbone within its native membrane environment [1].

We conduct hydrogen exchange (Hx) of all of the polar hydrogens of a transmembrane protein in organic solvent under acidic conditions. In this way, we are able to study any site using Hx even if the site is not accessible in the proteins native state. We then introduce the protein into its native environment and employ isotopic labeling to achieve site‐specific information at a desired location within the protein. Fourier transform infrared (FTIR) spectroscopy and molecular dynamics simulations allow us to correlate between the Hx FTIR amide I mode shift of the labeled carbonyl and the electrostatic field of the local environment. We find good correlation between Hx amide I mode shifts and the local electrostatic field at that site along the protein backbone.

[1] Esther S. Brielle and Isaiah T. Arkin. Site‐Specific Hydrogen Exchange in a Membrane Environment Analyzed by Infrared Spectroscopy. The Journal of Physical Chemistry Letters, 9(14):4059‐4065, 2018.

ABS455

STUDIES ON HUMAN EPIDERMAL GROWTH FACTOR RECEPTOR 2/4 (HER2/4) INHIBITORS THAT CAUSE CHANGES IN PROTEIN EXPRESSION LEVEL OF PROTOZOAN PARASITE, TOXOPLASMA GONDII

Won‐Kyu Lee ¹, Hye‐Jin Ahn², Jaehui Park¹, Seul gi Oh¹, Hyeweon Kang¹, Myung‐Ho Sohn¹, Hojin Yoo¹, Hye‐Jung Kim¹, Saehae Choi¹, Dae Young Kim¹, Jurang Woo¹, Ho‐Woo Nam²

¹New Drug Development Center, OSONG Biomedical Innovation Foundation (Cheongju, South Korea); ²Department of Parasitology, College of Medicine, The Catholic Univeristy of Korea (Seoul, South Korea)

Toxoplasma gondii, a ubiquitous, intracellular parasite of the phylum Apicomplexa, infects an estimated one‐third of the human population as well as a broad range of warm‐blooded animals. We have observed in previous reports that some tyrosine kinase inhibitors supressed the growth of T. gondii infected with host ARPE‐10 cells. These results suggested that inhibitors of receptor tyrosine kinase such as HER2/4 may be used as a therapeutic agent for inhibiting parasite growth with minimal adverse effects on host cells. In this report, we conducted a proteomic analysis to observe changes in host proteins that were altered via infection with T. gondii and the treatment of HER2/4 inhibitors. As a result, the expression level of PTPK, SEMA7A, PPP2R2A, RUS1 and COLM1 protein were significantly changed, and the results were confirmed by western blot analysis. Changes in various host proteins, including these five proteins, can be used as a basis for explaining the effects of T. gondii infections and HER2/4 inhibitors.

ABS456

COMPARISON OF KINETICS OF PURIFIED ANTIBODY AND ANIMAL CELL EXPRESSION SUP USING BLI SYSTEM

Myung ho Sohn ¹, Myung ho Sohn¹, HOJIN Yoo¹, Won‐Kyu Lee¹, Sora Park¹, Saehae Choi¹, So‐Young Choi¹

¹New Drug Developmnet Center, Osong Medical Innovation Foundation (Cheongju‐si, South Korea)

Real‐time and label‐free antibody screening systems are becoming more popular because of the increasing output of purified antibodies and antibody supernatant from many antibody discovery platforms.

One of the biggest advantages of BLI(Bio‐layer interferometry) system in measuring KInetics is high througput.

It is relatively time‐consuming and labor‐intensive to measure kinetic by purifying the candidate antibody group that has been excised and purifying the candidate antibody group. If kinetics can be directly measured in the crude sample without purification, more efficient antibody screening is possible. In this study, we investigated the binding affinity between purified antibody and crude antibody on purified HEK293F cell, Expi293F, and ExpiCHO. These results show that high throughput screening can be performed through the BLI in the selection and affinity measurement of antibodies in crude samples.

ABS458

CHROMATIN AS VIEWED BY UBIQUITIN WRITERS: DETERMINANTS OF H2A SITE SPECIFICITY BY RING UBIQUITIN E3 LIGASES, BRCA1/BARD1 AND RING1B/BMI1

Sam Witus ¹, Alex Zelter¹, Evie Henry¹, Mikaela Stewart¹, Trisha Davis¹, Rachel Klevit¹

¹University of Washington School of Medicine (Seattle, United States)

Deleterious mutations in the RING domain of the E3 ligase BRCA1 have long been implicated in familial breast and ovarian cancers. More recently, mutations in its heterodimeric RING partner, BARD1, have also been identified, further implicating the ubiquitin ligase activity of BRCA1/BARD1 in tumor suppression. We recently reported that while cancer‐associated missense mutations in the BRCA1 RING are ligase‐dead to all substrates due to the loss of E2 binding, cancer‐associated mutations in the BARD1 RING show specific loss of function towards nucleosomal histone H2A (Stewart et al. PNAS. 2018). BRCA1/BARD1 selectively monoubiquitylates lysine residues on the extreme C‐terminal tail of histone H2A in nucleosomes. This epigenetic modification is important in DNA repair and transcriptional regulation. The mechanism by which BRCA1/BARD1 engages the nucleosome to confer specific modification remains unknown, as attempts to crystallize the complex have proven unsuccessful. Here we present a biochemical characterization of the BRCA1/BARD1 nucleosome complex. We are using techniques such as crosslinking/mass spectrometry, NMR, and mutagenesis to map the binding surface for BRCA1/BARD1 on the nucleosome. These approaches identify residues on the nucleosome that are critical for H2A ubiquitylation. Our work provides insight into the underlying mechanisms of specificity for a growing number of E3s that monoubiquitylate distinct residues on the nucleosome in E3 ligase‐mediated chromatin regulation.

ABS459

ENGINEERED VIRUS‐LIKE‐PARTICLES FOR GPCR SPECIFIC THERAPEUTIC ANTIBODY DISCOVERY

Mart Ustav ¹, Jarrett Adams¹, Sachdev Sidhu¹

¹University of Toronto (Toronto, Canada)

G‐protein‐coupled‐receptors (GPCRs) compose one of the largest class of drug‐targeted molecules, but a majority of the GPCR proteome has still remained undrugged. Monoclonal antibodies (mAbs) targeting GPCRs have several advantages over small molecule drugs, but the complex nature of GPCRs hinders progress in mAbs discovery. When removed from the lipid bilayer, GPCRs are often unstable, and their overexpression in mammalian cells is difficult to establish. Therefore, adequate expression yields of functional GPCRs is a key challenge for mAbs discovery and development. To address this bottleneck we have established a robust method for the expression of GPCRs and other target membrane proteins in high density on HIV‐1 gag and Ebola VP40 Virus‐Like‐Particles (VLPs).

Several viral proteins interact with the intracellular C‐terminal domain of viral envelope glycoproteins, thereby enabling efficient incorporation of glycoproteins on the VLP surface. Inspired by this mechanism, we engineered the gag protein of HIV‐1 and VP40 protein of Ebola virus to enable a tight interaction with a short peptide fused to the C‐terminal cytoplasmatic tail of therapeutically relevant membrane proteins. Such interaction allowed us to efficiently incorporate membrane proteins on the surface of VLPs, enabling high expression yields of target proteins. We used these engineered VLPs for the isolation of synthetic mAbs from large phage‐ libraries and were able to identify mAbs binding therapeutically relevant GPCRs. In summary we have established a robust method suitable for GPCR expression on VLPs and utilize this system to target the therapeutically relevant GPCRome with mAbs.

ABS460

PREDICTING PROTEIN‐PROTEIN INTERFACE DOMAINS USING MULTIPLE SCALE ANALYSIS

Ben Tribelhorn¹, Victor Hsu ², Mike Bailey³

¹University of Portland (Portland, United States); ²Oregon State University (Corvallis, United States); ³Oregon State University, School of Electrical Engineering and Computer Science (Corvallis, United States)

Interactions between proteins are fundamental to many of lifes processes, and yet little is known about how a protein recognizes a specific binding domain on its partner protein. Current methods of classifying protein‐protein interactions on the basis of amino acid boundaries introduce an a priori bias to interface prediction, which results in the over‐segmentation of the actual interface. This bias can be reduced (or eliminated) by introducing a stochastic methodology to data segmentation prior to feature extraction. In the present study we use both computer vision and machine learning algorithms to recast the intrinsically three‐dimensional protein‐protein interface prediction problem into a feature extraction and classification problem. The advantage of this computational approach is that it attempts to eliminate conceptual bias, and the results can be used to inform existing docking software by identifying probable interface domains on a protein surface.

We used a Random Forest classifier in part because the protein interface prediction problem is characterized by weak predictive features. Also, the limited amount of available structural data for training restricted the number of trees we could implement before observing data overfitting. Each of the fourteen features included in our system were assessed using the leave‐one‐out validation method. Our analysis demonstrated that, perhaps unsurprisingly, hydrogen bond donors were the most useful features for defining an interface domain, and that the features were not independent. Our results allow for predicting critical components of interface domains, which may improve subsequent protein‐protein docking simulations.

ABS461

CRYO‐EM STRUCTURE OF THE GENE THERAPY VECTOR, ADENO‐ASSOCIATED VIRUS, WITH ITS CELL RECEPTOR, AAVR

Nancy Meyer ¹, Guiqing Hu², Omar Davulcu³, Qing Xie³, Alex Noble⁴, Craig Yoshioka⁵, Drew Gingerich⁵, Andrew Trzynka³, Larry David³, Scott Stagg⁶, Michael Chapman⁷

¹Oregon Health and Science University (Portland, United States); ²Inst. Molecular Biophysics, Florida State Univ., Tallahassee, FL 32306 (Tallahassee, United States); ³Dept. Biochem. & Mol. Biol., Oregon Health & Sci. Univ., Portland, OR 97239 (Portland, United States); ⁴New York Structural Biology Center, New York, NY 10027 (New York, United States); ⁵OHSU Center for Spatial Systems Biomedicine, Portland OR 97201 (Portland, United States); ⁶Inst. Mol. Biophys. and Dept. Chem. & Biochem., Florida State Univ., Tallahassee, FL 32306 (Tallahassee, United States); ⁷Department of Biochemistry, University of Missouri, Columbia, MO 65211 (Columbia, United States)

Adeno‐associated virus (AAV) vectors are preeminent in emerging clinical gene therapies. The modulation of their cell specificities and optimization of immune‐escape will be required to suit applications beyond the most tractable genetic diseases. The interactions of AAV with its cellular receptor, AAVR, are key to understanding cell‐entry and trafficking with the rigor needed to engineer tissue‐specific vectors. Here, we demonstrate cryo‐electron microscopys (cryoEM) utility in structure determination of flexible molecules, using a hybrid, divide‐and‐conquer strategy. AAVR is an integral transmembrane protein receptor, whose ectodomain contains five polycystic kidney disease domains (PKDs). Cryo‐electron tomography with a five‐PKD construct shows ordered binding of part of the flexible receptor to the viral surface, with distal PKDs in multiple conformations. Regions of the virus and receptor in close physical proximity can be identified by cross‐linking/mass spectrometry (x‐MS), which helps in determining overall receptor orientation, and permits simple modeling of tomographic density with the aid of x‐MS distance constraints. Single particle cryo‐EM of virus in complex with a two‐PKD receptor fragment further reveals the interactions at 2.4 å resolution and allows fitting of atomic models to reveal residue‐level interactions. We see AAVR binds between AAV‐2s spikes on a plateau that is conserved throughout serotypes, except in one clade whose structure is AAVR‐incompatible. Interestingly, AAVRs footprint overlaps the epitopes of several neutralizing antibodies, prompting a re‐evaluation of neutralization mechanisms. The structure provides a roadmap for experimental probing and manipulation of AAVs capsid‐receptor interactions, and informs strategies for determining flexible receptor binding interactions more broadly.

ABS462

MAF1B FROM TOXOPLASMA GONDII INTERACTS WITH HUMAN RALGAP1, POTENTIALLY ALTERING HOST IMMUNE SIGNALLING

Cameron Powell ¹, Matthew Blank², Reece Hoffman¹, John Burke¹, John Boyle², Martin Boulanger¹

¹University of Victoria (Victoria, Canada); ²University of Pittsburgh (Pittsburgh, United States)

Parasites interact intimately with their hosts, often subverting host processes to promote parasite growth and survival. The common human parasite Toxoplasma gondii replicates exclusively in a vacuole within a host cell, and alters host cell function via secreted proteins. One of these secreted proteins, MAF1b, acts to concentrate mitochondria around the parasitophorous vacuole, altering the host immune response. While the T. gondii protein mediating recruitment of host mitochondria is known, the identities of host partners that mediate this interaction are not well established, and the mechanism by which MAF1b mediates changes in host immune signalling is unknown. Here, we show that T. gondii MAF1b binds to human Ral GTPase Accelerating Protein (RalGAP), providing a basis for modulation of the host immune response by MAF1b. Subsequent biophysical analyses isolate the specific MAF1b surface residues required for RalGAP binding. This work provides new insight into a key host‐pathogen interaction and identifies possible targets for future therapeutic intervention as well as a deeper understanding of crucial parasite biology.

ABS463

TARGETED MUTATIONAL PERTURBATIONS OF THE SMALL GTPASE RAN REVEAL HOW PLEIOTROPY IS ENCODED IN A MODEL MOLECULAR SWITCH

Christopher Mathy ¹, Tina Perica¹, Yang Zhang¹, Jiewei Xu¹, Gwendolyn Jang¹, Danielle Swaney¹, Nevan Krogan¹, Tanja Kortemme¹

¹University of California, San Francisco (San Francisco, United States)

The small GTPase Ran is an essential, conserved, and pleiotropic signaling hub. Ran is known to regulate many cellular processes through physical interactions with numerous proteins, but the mechanism by which it can simultaneously regulate distinct processes is poorly understood. In this work, we made genomic point mutations in structurally‐characterized interaction interfaces of Ran to perturb its function in S. cerevisiae, and measured the phenotype of each mutant using high‐throughput quantitative genetic interaction profiling. The mutants showed rich and distinct phenotypes, but interestingly did not group according to the interaction interface they perturbed. Instead, we find the phenotype of each mutation is best explained by the effect it has on the GTPase cycle, which is regulated by a GAP and GEF pair. Biochemical characterization identified groups of mutants that primarily perturbed either the nucleotide exchange or the GTP hydrolysis part of the GTPase cycle. Analysis of the conformational equilibrium of GTP‐bound Ran using 31P NMR revealed allosteric mutants that disfavor the hydrolytically competent state, demonstrating that coupling between distant interface residues and the active site of the switch impacts cycle kinetics. Combined analysis of the cycle kinetic parameters of the mutants and their corresponding phenotypes revealed how distinct cellular pathways differentially sense parts of the GTPase cycle. The analysis identified three groups of yeast genes. Genes that correlate with the GEF‐perturbed mutants appear to be most sensitive to overall levels of RanGTP in the cell, those that correlate with the GAP‐perturbed mutants appear to be most sensitive to the dynamics of turning off RanGTP signaling, and those that correlate with both groups appear to require complete cycling of RanGTP for their function. Therefore, we propose that Ran pleiotropy is encoded by the biochemical rates of each step in the GTPase cycle which themselves are differentially sensed by distinct cellular pathways.

ABS464

CIS‐ACTING GLYCAN DRIVES PROTEIN‐PROTEIN INTERACTIONS OF SKP1 IN DICTYOSTELIUM AND TOXOPLASMA

Hyunwoo Kim¹, Christopher West ¹, Alexander Eletsky¹, James Prestegard¹

¹University of Georgia (Athens, United States)

Glycosylation diversifies proteins beyond the scope of the 20 amino acids. The importance of glycosylation is widely acknowledged, but molecular mechanisms by which it modulates protein structure and function remain poorly understood. Here we describe a glycan that regulates protein/protein interactions involving Skp1, an essential subunit of E3(SCF) ubiquitin ligases. Skp1 recruits F‐box proteins (FBPs) and their target substrates to the ligase for polyubiquitination and subsequent degradation in the 26S proteasome. Subsite‐2 of the F‐box binding domain of Skp1 undergoes a coil‐to‐helix transition during complex assembly. A hydroxyproline residue within this region can be modified by a highly conserved linear pentasaccharide in protists, and the resulting glycan promotes formation of Skp1/FBP subcomplexes in vitro and in vivo. Our NMR studies on Skp1 structure and glycan dynamics, together with molecular dynamics trajectories that satisfy NMR constraints, suggest a cis‐acting mechanism by which the glycan encourages local helicity in addition to promoting an open conformation susceptible to interact with an F‐box. However, NMR data were collected at a concentration where Skp1 is a dimer while MD simulations were conducted on a monomer. The solution structure of a truncated Skp1 dimer buries subsite‐1 of the F‐box binding site, but its 2.5 μM Kd, observed using analytical ultracentrifugation, suggests that the monomer is the functional entity in cells. We introduced mutations at the predicted dimer interface to generate a stable monomer that is competent for glycosylation, which will be useful for confirming predicted interactions between the glycan and neighboring amino acids that are hypothesized to mediate its conformational effects.

ABS465

EFFORTS TO ENHANCE THE EXPRESSION OF FUNCTIONALLY ACTIVE MEMBRANE PROTEINS USING BACMAM EXPRESSION SYSTEM

Srivanya Tummala ¹, Noel Byrne¹, Jennifer Shipman¹, James Kostas¹, Richard Edwards¹, Kaspar Hollenstein¹, Harini Krishnamurthy¹, Alexei Brooun¹, Stephen Soisson¹

¹Merck & Co. (North Wales, United States)

Membrane proteins control basic cellular activities which represent a third of the total proteins in human proteome and more importantly constitute about 60% of drug targets. Obtaining structural information of membrane proteins with ligands to enable Structure Based Drug Design (SBDD) is extremely challenging due to their low expression, poor protein stability, low ligand occupancy after extraction from membrane and in turn diminished functional activity of purified protein. Here we present our efforts to explore improvements in expression level of membrane proteins along with their functional activity for biophysical characterization and structural studies using BacMam expression system. We will present side by side comparison with of BacMam mediated expression with transient transfection mammalian expression protocol for several GPCRs and Ion Channels to produce functional integral membrane proteins.

ABS466

ENGINEERING THE NEXT GENERATION OF SH2 SUPERBINDERS TO PROBE THE PHOSPHOPROTEOME AND ANTAGONIZE CANCER CELL SIGNALLING

Greg Martyn ¹, Gianluca Veggiani¹, Sachdev Sidhu¹

¹Department of Molecular Genetics, University of Toronto (Toronto, Canada)

Cancer cell proliferation is dependent on tyrosine kinase signal transduction pathways. Phosphorylated tyrosine (pY) marks are laid down by tyrosine kinases, erased by phosphatases, and read by, among others, the Src Homology 2 (SH2) domain. The Sidhu group has previously engineered the Src SH2 domain to bind its pY ligands with a high affinity (superSrc). SuperSrc was shown to antagonize cancer cell proliferation by acting as a competitive inhibitor of pY marks essential to signal transduction pathways. SuperSrc has also been used to probe the phosphoproteome of cancer cells. Therefore, engineering the 119 other SH2 domains in humans into superbinders will allow us to probe the phosphoproteome and antagonize cancer cell proliferation much more comprehensively. Using the superbinding motifs found in superSrc and superFes (Sidhu Lab ‐ unpublished) I have demonstrated that we can engineer other SH2 domains into superbinders by grafting the superbinder mutations discovered in superSrc and superFes into other SH2 domains. I then demonstrated that increased hydrogen bond formation is responsible for the superbinder properties by solving the structure of Src and Fes superbinder variants and performing alanine scanning on residues that form the pY binding pocket. Finally, using mass spectrometry, I have demonstrated that SH2 superbinders bind the same peptides as their wild‐type (wt) counterparts (albeit with a higher affinity) and therefore retain specificity. I have established a new cohort of SH2 superbinders that I will use to probe the phosphoproteome of cancer cell lines in greater depth. In addition to this, I will assess the ability of these new SH2 superbinders to shut down signal transduction pathways essential to cancer cell growth and survival.

ABS467

INCREASING SURFACE CHARGE CONVERTS SPY INTO A MORE EFFICIENT CHAPERONE

Wei He¹, Shu Quan ¹, Jiayin Zhang¹

¹East China University of Science and Technology (Shanghai, China)

Chaperones are essential components of the protein homeostasis network. Recently, it was shown that molecular engineering could be used to optimize chaperone function. Here, we show that the anti‐aggregation activity of the chaperone Spy can be improved by altering the electrostatic potential of its concave surface. This strategy is more efficient than enhancing the hydrophobicity of Spys surface, challenging the traditional notion that hydrophobic interactions are the driving force guiding chaperone‐substrate binding. Instead, our data support the notion that electrostatic forces predominate in the binding interaction and hydrophobic forces predominate in the dissociation reaction. Our results reveal that increasing short‐range hydrophobicity interactions deleteriously alter Spy's ability to capture substrates thus reducing its in vitro chaperone activity. Our results deepen the understanding of the mechanistic basis of Spy's chaperone function and illustrate how new surface‐based mutational strategies can facilitate the rational improvement of molecular chaperones.

ABS468

THE MULTIPLE CONFORMER STORY: CHARACTERIZATION OF 10E8 ANTIBODY CONSTRUCTS BY SEC, HIC AND MOLECULAR DYNAMICS SIMULATIONS

MICHAEL BENDER ¹, Amitava Roy², Sylvie Yang¹, Yile Li¹, Xiangchun Wang¹, Frank Arnold¹, Paula Lei¹

¹Vaccine Production Program, VRC, NIAID, NIH (Gaithersburg, United States); ²Rocky Mountain Labs, NIAID, NIH (Hamilton, United States)

10E8 is a highly potent broadly neutralizing antibody (bNAb) that targets the membrane‐proximal external region (MPER) of the HIV‐1 envelope. During the pre‐clinical analytical development stage, 10E8 exhibited a multi‐monomer peak profile when run on Size‐Exclusion Chromatography (SEC). The cause of its secondary interactions with the SEC column matrix was found to be the multiple states of hydrophobicity of this bNAb. The bNAb was run on a hydrophobic interaction chromatography (HIC) column, and had a chromatogram exhibiting the similar multi‐peak profile that strongly correlated with the SEC chromatogram. Point mutations made in the hypothesized binding pocket showed dramatically different chromatograms: including the creation of a disulfide bond between the CDRH3 and CDRL1 resulting in a single monomeric species when run by SEC in stark contrast to the original multi‐peak profile. The experimental results coupled with the in‐silico molecular dynamics simulations begins to explain the conformational changes of the 10E8 antibody occurring in solution.

ABS469

CONSTRUCTION OF CHIMERIC CALBINDIN D9K PROTEINS SHOWING A CA2+ INDUCED ‐CONFORMATIONAL CHANGE

Emma Liliana Arévalo‐Salina ¹, Humberto Flores‐Soto¹, Joel Osuna‐Quintero¹, Gloria Saab‐Rincón¹

¹Instituto de Biotecnología, UNAM (Cuernavaca, Mexico)

The EF‐hand superfamily are Ca2+‐binding proteins that use a helix‐loop‐helix motif to bind Ca2+ (EF‐hand). This group of proteins is divided into two classes according to their response to Ca2+: calcium sensors and calcium modulators, being a large calcium‐dependent conformational change in sensor proteins, that distinguishes them. In this work we induced a Ca2+‐dependent conformational change in a modulator EF‐hand motif (ClbN of Calbindin D9k), by replacing some secondary structure elements from the sensor motif (SCIII of Troponin C). To detect the induced conformational change, the modules were fused to a reporter system based on the activity of the enzyme prephenate dehydrogenase (TyrA) from E. coli. This system allowed the determination of apparent Ca2+‐binding constants. The individual chimeric modules were then reinserted into the Calbindin D9k protein to determine if they could induce this behavior in the whole protein. By means of circular dichroism and extrinsic fluorescence it was determined that the ClbD9k‐SCIII and ClbD9k‐HIClbNSCIII variants show an increase in their helical structure and a change in fluorescence in the presence of Ca2+. This suggests that either, the insertion of a sensor module to the Calbindin D9k modulator protein, or the exchange of both, the Ca2+‐binding loop and the helix II by the corresponding elements of the SCIII module are sufficient to induce a conformational change in Calbindin D9k. By comparing the behavior of the different chimeras in both systems we observed that the behavior of an individual motif is strongly influenced by the context in which it is located.

ABS470

MACHINE‐LEARNING‐GUIDED MUTAGENESIS FOR DIRECTED EVOLUTION OF FLUORESCENT PROTEINS

Tomoshi Kameda ¹, Yutaka Saito¹, Misaki Oikawa², Hikaru Nakazawa², Teppei Niide², Koji Tsuda³, Mitsuo Umetsu²

¹Artificial Intelligence Research Center, AIST (Koto, Japan); ²Depart. Biomolecular Engineerring, Tohoku Univ. (Sendai, Japan); ³Depart. Computational Biology and Medical Sciences, Univ. Tokyo (Kashiwa, Japan)

Molecular evolution based on mutagenesis is widely used in protein engineering. However, optimal proteins are often difficult to obtain due to a large sequence space. Here, we propose a novel approach that combines molecular evolution with machine learning. In this approach, we conduct two rounds of mutagenesis where an initial library of protein variants is used to train a machine‐learning model to guide mutagenesis for the second‐round library. This enables to prepare a small library suited for screening experiments with high enrichment of functional proteins. We demonstrated a proof‐of‐concept of our approach by altering the reference green fluorescent protein (GFP) so that its fluorescence is changed into yellow. We successfully obtained a number of proteins showing yellow fluorescence, 12 of which had longer wavelengths than the reference yellow fluorescent protein (YFP). These results show the potential of our approach as a powerful method for directed evolution of fluorescent proteins.

ABS471

SUBSTRATE‐BASED ALLOSTERIC REGULATION OF A HOMODIMERIC ENZYME

Christopher Di Pietrantonio¹, Keith Taverner ¹, Pedram Mehrabi², Tae Hun Kim³, Adnan Sljoka⁴, Christopher Ing³, Regis Pomes³, Emil F. Pai⁵, R. Scott Prosser¹

¹Department of Chemistry, University of Toronto (Mississauga, Canada); ²Max‐Planck‐Institute for Structure and Dynamics of Matter, Department for Atomically Resolved Dynamics (Hamburg, Germany); ³The Hospital for Sick Children (Toronto, Canada); ⁴CREST, Japan Science and Technology Agency (JST), Department of Informatics, School of Science and Technology, Kwansei Gakuin University, Japan (Nishinomiya, Japan); ⁵Department of Biochemistry, University of Toronto (Toronto, Canada)

Approximately twenty percent of enzymes exhibit substrate inhibition. Here, a homodimeric enzyme, fluoroacetate dehalogenase utilizes a regulatory cap domain to allow only one protomer at a time to bind substrate. At high substrate concentrations, a second substrate binds to a site along the binding channel of the occupied protomer, giving rise to catalytic inhibition. While a single mutation (K152I) in the allosteric pocket abrogates second site binding and removes inhibitory effects, it also precipitously lowers the maximum catalytic rate, implying a role for the allosteric pocket at low substrate concentrations. NMR and docking experiments show that the allosteric pocket first desolvates the substrate, allowing it is to be deposited in the active site. This triggers the empty second substrate pocket to then serve as an allosteric conduit between the protomers, enabling sampling of activation states needed for catalysis. These results illustrate the role of dynamics along allosteric networks in facilitating chemical function.

ABS472

ASSESSING THE RESILIENCE OF PROTEINS TO THE EFFECTS OF DRUGS

Tess Thackray ¹, Bodi Van Roy¹, Filip Jagodzinski¹

¹Western Washington University (Bellingham, United States)

The biological and chemical interactions between a protein and a ligand can provide insights about the efficacy of a drug. To study how a protein might be resilient to the effects of a drug, we generate, in silico, all possible mutants with single amino acid substitutions for a dozen protein ligand complexes. We use our custom rMutant‐2 software, which relies on a combination of homology modeling along with energy minimization to account for steric clashes that are introduced when mutating a small residue to a larger one. We employ a variety of combinatorial methods, and record the change of the all‐atom energy of the protein ligand complex during energy minimization, to assess the extent of the effect of the ligand. We create a drug resilience profile (DRP) using data for all of the mutants, and strive to characterize all protein ligand complexes so as to provide biochemical insights about the effects of the ligand. Our results indicate that the degree to which a protein is resilient to a drug, as assessed using computation methods and our DRP, is a combination of multiple factors, including the number of hydrogen bonds and hydrophobic interactions that the ligand forms with the protein, as well as the atomic makeup of the ligand. This work has the potential to further validate our software used to assess the impact of drugs, and to supplement experiments done in the wet lab, which can be time and cost prohibitive. (edited)

ABS473

A SCALABLE COMPUTE FRAMEWORK FOR GENERATING AND ASSESSING PROTEIN MUTANTS

Dylan Carpenter ¹, Michael Albert¹, Sam Herr¹, Filip Jagodzinski¹

¹Western Washington University (Bellingham, United States)

A mutation of a single protein residue has the potential to affect the chemical interactions between a protein and a ligand, which may alter the intended effect of the drug. In order to aid efforts to discover potentially viable drugs, we aim to analyze protein‐ligand complexes in which the protein is mutated via an in silico approach; the ligand is left unaltered. Our approach integrates our custom ProMute‐2 software which uses off‐the‐shelf homology modeling techniques to perform in silico single and multiple amino acid substitutions. Our approach, which is currently being extended to be employed on a compute cluster environment, generates protein mutant‐ligand complexes, and analyzes the effect of the ligand using a mix of combinatorial approaches along with energetics terms from short runs of energy minimization. We present benchmark timing results and show that it is possible to generate a data set of exhaustive mutants with single amino acid substitutions for a 100 residue protein in as little as one hour, using a single quad‐core personal computer. We also showcase a variety of plots of energetics metrics to reveal computational requirements for generating exhaustive mutants for a variety of protein‐ligand complexes. Our proof‐of‐concept results demonstrate that exhaustive mutations of proteins can be efficiently generated through in silico methods, and we envision that the rich datasets that we make available can inform a variety of pharmacology studies.

ABS474

DESIGNED PROTEIN LOGIC FOR ULTRA‐SPECIFIC CELL TARGETING

Marc Lajoie¹, Jilliane Bruffey ², Scott Boyken¹, Alexander Salter³, Anusha Rajan³, Robert Langan¹, Audrey Olshefsky⁴, Vishaka Muhunthan³, Mesfin Gewe⁵, Alfredo Quijano Rubio⁴, Colin Correnti⁵, Stanley Riddell⁶, David Baker⁷

¹Institute for Protein Design, University of Washington, Seattle, WA, USA; Department of Biochemistry, University of Washington, Seattle, WA, USA (Seattle, United States); ²Institute for Protein Design, University of Washington, Seattle, WA, USA; Department of Biochemistry, University of Washington, Seattle, WA, USA; Graduate Program in Molecular and Cellular Biology, University of Washington, Seattle, WA, USA (Seattle, United States); ³Program in Immunology, Fred Hutchinson Cancer Research Center, Seattle, WA, USA; Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, WA, USA (Seattle, United States); ⁴Institute for Protein Design, University of Washington, Seattle, WA, USA; Department of Bioengineering, University of Washington, Seattle, WA, USA (Seattle, United States); ⁵Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, WA, USA (Seattle, United States); ⁶Program in Immunology, Fred Hutchinson Cancer Research Center, Seattle, WA, USA; Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, WA, USA; Department of Medicine, University of Washington School of Medicine, Seattle, WA, USA (Seattle, United States); ⁷Institute for Protein Design, University of Washington, Seattle, WA, USA; Department of Biochemistry, University of Washington, Seattle, WA, USA; Howard Hughes Medical Institute, University of Washington, Seattle, WA, USA (Seattle, United States)

Targeted immunotherapies typically rely on the recognition of a single cell‐surface marker, and as such, their specificity is limited and prone to incidentally harming healthy cells. This is problematic for developing cancer therapies, where tumor‐specific antigens are usually expressed at lower levels on healthy tissues.

We have developed de novo designed protein switches capable of performing Boolean logic on the surface of cells so as to enable ultra‐specific targeting of a given cell type within a complex mixture of cells. These protein switches are activated only on the surface of cells that have precise combinations of antigens that satisfy the logical operation. Upon activation, a bioactive peptide then recruits the immunotherapy to the cell. These logic operations can be used to enhance tumor selectivity (AND logic), to enable flexible targeting of heterogeneous tumors (OR logic), and to avoid harming off‐target tissues that share similar expression profiles with the cancer cells (NOT logic). Furthermore, the ability to combine these operators to perform complex logic opens the door for unprecedented control over therapeutic targeting.

We demonstrate the use of protein switches to perform simple AND as well as complex AND/OR and AND/NOT logic and apply them to redirect T cell specificity against tumor cells expressing certain combinations of surface antigens while sparing cells expressing single antigens. Furthermore, this system is modular and highly tunable it can be adapted for the recruitment of a wide array of effectors to diverse target cell types.

ABS475

INTEGRATING THE INFLUENCE OF PH IN NMR CHEMICAL SHIFT PREDICTION METHODS

Efrosini Artikis ¹, Charles Brooks III²

¹Graduate Student (Ann Arbor, United States); ²Department of Chemistry and Biophysics, University of Michigan (Ann Arbor, United States)

Due to the unique ability for NMR chemical shifts (CSs) to comprehensively characterize protein structure and dynamics, there has been significant effort in recent decades to establish a theoretical framework for the prediction of these observables. However, the reliance of machine learning algorithms on large quantities of experimental data has prohibited the parameterization of pH in most prediction models. Recent work on model peptides, however, has elucidated the role of through‐bond and through‐space contributions such that a description of pH can now be incorporated in a physics‐based chemical shift prediction model. In this work, we incorporate inductive, electrostatic and conformational descriptions of pH in a classical physics‐based model for which parameters are fit using least‐squares regression. Constant pH molecular dynamics simulations on well‐studied proteins allow for ensemble averaged parameters to be computed thereby recapitulating aspects of the measured NMR ensemble. The incorporation of pH in NMR chemical shift prediction is essential for the accurate description of the chemical environment and plays an important role in understanding protein chemistry.

ABS476

LARGE‐SCALE CHARACTERIZATION OF PTEN MISSENSE VARIANTS THAT DIFFERENTIALLY AFFECT INTRACELLULAR PROTEIN ABUNDANCE AND PHOSPHATASE ACTIVITY

Kenneth Matreyek ¹, Douglas Fowler¹, Jason Stephany¹

¹University of Washington (Seattle, United States)

PTEN is a multi‐functional tumor suppressor that uses its phosphatase activity to directly counteract pro‐oncogenic signaling, though it also has additional roles altering immune signaling, neuronal function, and genome stability. Accordingly, germline PTEN variants underlie a spectrum of developmental abnormalities ranging from intellectual disability and autism to increased cancer predisposition, while PTEN somatic variants are found in diverse tumors where it can be a cancer driver. Experimental characterization is needed to interpret the biochemical properties and associated disease severity of each PTEN variant, but traditional approaches are unable to keep pace in comprehensively characterizing the glut of clinically observed PTEN variants.

Deep mutational scans can address this shortcoming by assessing thousands of variants in a single experiment. Recently, we characterized 4,113 PTEN missense variants for intracellular abundance, while a separate group assessed the activities of 7,254 PTEN missense variants for phosphatase activity. Here, we used an independent library to characterize 970 additional PTEN variants for abundance. The combined abundance dataset contains 168 positions with 16 or more substitutions, revealing patterns in PTEN mutational tolerance for its folding and accumulation within cells. We also compare how some variants differentially alter PTEN abundance and activity, such as stable but inactive variants that are likely dominant negatives. We observe enrichment of uncharacterized, stable but inactive variants in cancer genomics data, supporting their likely dominant negative effects, and highlighting the unique ability of orthogonal deep mutational scans to identify clinically relevant subtypes of PTEN loss‐of‐function missense variants at scale.

ABS478

REGULATORY MECHANISMS OF THE DEUBIQUITINASE BAP1

Maxime Uriarte ¹, Salima Daou², Haithem Barbour¹, Oumamai Ahmed³, Louis Masclef¹, Caroline Baril⁴, Nadine Sen Nkwe¹, Eric Bonneil⁵, Derek Ceccarelli⁶, Jean‐Yves Masson⁷, Pierre Thibault⁵, Frank Sicheri², Haining Yang⁸, Michele Carbone⁸, Marc Therrien⁹, El Bachir Affar¹

¹Maisonneuve‐Rosemont Hospital Research Center and Department of Medicine, University of Montréal (Montreal, Canada); ²Lunenfeld‐Tanenbaum Research Institute, Sinai Health System (Toronto, Canada); ³Maisonneuve‐Rosemont Hospital Research Center and Department of Medicine, University of Montréal, Montréal (Montreal, Canada); ⁴Institute for Research in Immunology and Cancer, Laboratory of Intracellular Signaling, University of Montréal (Montreal, Canada); ⁵Institute for Research in Immunology and Cancer, Laboratory of Proteomics and Bioanalytical Mass Spectrometry, University of Montréal (Montreal, Canada); ⁶Lunenfeld‐Tanenbaum Research Institute, Sinai Health System, Toronto (Toronto, Canada); ⁷CHU de Quebec Research Center (Oncology Axis), Laval University Cancer Research Center (Quebec, Canada); ⁸University of Hawaii Cancer Center, University of Hawaii, Honolulu (Honolulu, United States); ⁹Département de pathologie et biologie cellulaire, University of Montréal, Montréal (Montreal, Canada)

The tumor suppressor BAP1 is a major deubiquitinase (DUB) for histone H2AK119, a modification involved in the coordination of transcription and DNA repair. We established that BAP1 is subjected to several quality control mechanisms ensuring proper targeting and deubiquitination of its substrates in the nucleus. First, the ubiquitin‐conjugating and ‐ligase hybrid UBE2O multi‐monoubiquitinates the nuclear localization signal of BAP1, thereby inducing its cytoplasmic sequestration. Importantly, intramolecular interactions, involving multiple domains, ensure BAP1 proper conformation and endow this DUB with the ability to autodeubiquitinate, thus counteracting UBE2O action. Significantly, we identified cancer‐derived BAP1 mutations that abrogate autodeubiquitination and promote its cytoplasmic retention. In the nucleus, BAP1 assembles DUB complexes with the transcription regulators Additional Sex Combs‐Like (ASXLs). ASXLs use their DEUBiquitinase ADaptor (DEUBAD) domain to stimulate BAP1 activity. Interestingly, we found that the DEUBAD domain is monoubiquitinated, resulting in an increased stability of ASXLs, which in turn stimulates BAP1 DUB activity. ASXLs monoubiquitination is directly catalyzed by UBE2E family of ubiquitin‐conjugating enzymes, which ensure a rapid turnover of free ASXLs while stabilizing ASXLs integrated within the BAP1 complexes, hence tightly regulating the dosage of ASXLs and DUB activity. Excess of free ASXLs proteins might be quickly primed by UBE2Es for degradation to prevent potential unwanted effects of orphan ASXLs. As we demonstrated that monoubiquitination of DEUBAD is promoted by the proper folding of BAP1, we also concluded that monoubiquitination of DEUBAD could provide a quality control mechanism for the assembly of a DUB activity competent complex.

ABS479

STRUCTURAL CHARACTERIZATION OF THE LC8‐RAVP COMPLEX REVEALS A NEW ROLE FOR LC8 IN LYSSAVIRUS PHOSPHOPROTEINS

Nathan Jespersen ¹, Cedric Leyrat², Francine Gérard³, Jean‐Marie Bourhis³, Danielle Blondel⁴, Marc Jamin³, Elisar Barbar¹

¹Oregon State University (Corvallis, United States); ²Institut de Génomique Fonctionnelle (Montpellier, France); ³Université Grenoble Alpes (Grenoble, France); ⁴Institut de biologie intégrative de la cellule (Gif‐Sur‐Yvette, France)

The rabies and Ebola viruses recruit the highly conserved host protein LC8 for their own reproductive success. in vivo knockouts of the LC8 recognition motif within the rabies virus phosphoprotein (RavP) result in completely non‐lethal viral infections. In this work, we provide a model for the molecular role LC8 plays in viral lethality. We show that RavP and LC8 co‐localize in rabies infected cells, and that LC8 interactions are essential for efficient viral polymerase activity. NMR, SAXS, and molecular modeling demonstrate that LC8 binding to a disordered linker adjacent to an endogenous dimerization domain results in restrictions in RavP domain orientations. The resulting dynamic RavP‐LC8 tetrameric complex is similar to that of a related phosphoprotein that does not bind LC8, suggesting a regulatory role for the LC8‐induced structure. The high conservation of the LC8 motif in Lyssavirus phosphoproteins evinces a broader purpose for LC8 in regulating vital downstream phosphoprotein functions.

ABS480

HARNESSING THE NATURAL PROPERTIES OF HUH‐ENDONUCLEASES FOR COVALENT PROTEIN‐DNA LINKAGE TECHNOLOGIES

Kassidy Tompkins ¹, Andrew Nelson¹, Blake Everet¹, Andrew Lemmex¹, Lidia Swanson¹, Wendy Gordon¹

¹University of Minnesota (Minneapolis, United States)

In nature, HUH‐endonucleases (histidine‐hydrophobic‐histidine) create nicks in hairpin loops at the origin of replication during viral rolling circle replication and bacterial plasmid conjugation. Intriguingly, these metal‐dependent domains can form rapid, covalent, and specific linkages with single‐stranded DNA (ssDNA) under physiologic conditions. Referred to as HUH‐tags, they are becoming valuable tools in technologies such as live cell imaging, force spectroscopy, genome editing, and cell targeting. In our study, we compared naturally useful properties of HUH‐tags derived from circovirus, nanovirus, geminivirus, and bacteria. We demonstrate how fast kinetics, programmable target sequence specificity, and human cell line expression of the viral HUH‐tags are useful to these technologies, rivaling the commercially available SNAP‐tag. We compared the HUH‐tag cleavage kinetics using molecular beacon stopped‐flow experiments, which reveal exceptionally fast conjugation of the viral HUH‐tags displaying cleavage rate constants up to 0.53 s‐1. Though viral HUH‐tags attack a highly conserved in nature, we explored the sequence specificity of the viral HUH‐tags using a library of DNA oligonucleotide variants. Interestingly, the viral HUH‐tags have promiscuous sequence specificity, yet we also discovered DNA variants with orthogonal target sequences that will allow for multiplexing. Previous observations suggest that most HUH‐tags do not express well in the cytoplasm or cell surface of mammalian cells. Thus, we screened for active HUH‐tags in human cell lines that express either intracellular or cell surface HUH‐tag constructs and found several that are functional in this context. Viral HUH‐tags are emerging as an optimal protein‐DNA linkage tool with exceptional natural properties and tunable characteristics.

ABS481

DETERMINATION OF PROTEIN COMPLEX ARCHITECTURE GUIDED BY LOW‐RESOLUTION CRYO‐ELECTRON MICROSCOPY DENSITY

Daniel Farrell ¹, Frank DiMaio¹

¹University of Washington (seattle, United States)

Recent advances in cryo‐electron microscopy (cryo‐EM) have revolutionized the world of structural biology, however 54% of all released maps are confined to the resolution range of 5‐20å. At these low resolutions we are unable to build models de‐novo. However, utilizing a combination of ab initio and homology modeling, and Rosettas physically realistic forcefield we are able to correctly model many simulated, and experimental cryo‐EM complexes. Our application utilizes Rosettas fast protein interface optimization, a low‐resolution optimized protein‐density docking protocol, and experimental data (if applicable) in order to, on average, place ~75% of the protein complex with an approximate error rate of 20%. Although our error rate is high, errors are often obvious and a combination of manual inspection and iterative complex assembly nearly always yields the correct result except in regions of extremely reduced density. With this application, we were able to solve the structure of the pseudo‐symmetric Bardet‐Biedl syndrome (BBS) protein complex at 5å. Combined, our data suggests that our work will be an invaluable resource for cryo‐EM complexes that struggle to breach the 5A resolution barrier.

ABS482

STRUCTURAL AND FUNCTIONAL CHARACTERIZATION OF A TRISTETRAPROLIN FAMILY TANDEM ZINC FINGER PROTEIN

Stephanie Hicks ¹, Ronald Venters², Wi Lai¹, Monica Pillon¹, Perry Blackshear³

¹Signal Transduction Laboratory, National Institute of Environmental Health Sciences (RTP, United States); ²Duke University NMR Center (Durham, United States); ³Signal Transduction Laboratory, National Institute of Environmental Health Sciences; Departments of Medicine and Biochemistry, Duke University Medical Center (RTP, United States)

Tristetraprolin (TTP) family members contain tandem zinc finger (TZF) domains that bind to mRNAs containing AU‐rich elements (AREs) to regulate mRNA turnover. With limited structural and biochemical data available, the mechanism of TTP‐mediated mRNA decay is not well understood. Using a combination of techniques, including NMR and small‐angle X‐ray scattering (SAXS), we are investigating the structural and functional determinants for RNA binding and mRNA decay using the single TTP family member found in the spikemoss, Selaginella moellendorffii. The spikemoss protein represents the first full‐length TTP family member that we have been able to purify to homogeneity in its native state. Although the spikemoss protein is the smallest TTP family member identified to date, it maintains both a highly conserved TZF domain and a C‐terminal binding site for the CNOT1 subunit of the CCR4‐NOT1 deadenylase complex, making it an ideal candidate for functional and structural studies. To date, our findings with the spikemoss protein include: 1) the TZF domain contains the molecular determinants necessary for tight binding to RNAs containing a human TTP (hTTP) target sequence; 2) upon RNA binding, structural changes are induced primarily within the TZF domain; 3) the full‐length protein interacts with the CNOT1 subunit of the CCR4‐NOT1 deadenylase complex in a manner similar to the equivalent interaction between human CNOT1 and TTP; and 4) the spikemoss protein can facilitate mRNA decay in transfected mammalian cells that is dependent upon the presence of the C‐terminal CNOT1 interacting domain.

ABS483

COMPUTATIONAL DESIGN OF A DE NOVO, MODULAR MINIPROTEIN TARGETING PD‐1

Cassie Bryan ¹, Gabriel Rocklin², David Baker¹

¹University of Washington (Seattle, United States); ²Northwestern (Chicago, United States)

Programmed Cell Death Protein 1 (PD‐1), expressed on activated T cells, inhibits T cell function and proliferation to prevent an excessive immune response. Tumor cells often take advantage of this pathway by over‐expressing a ligand of PD‐1, PD‐L1, to evade immune destruction. Monoclonal antibodies that target PD‐1 and PD‐L1 are clinically useful for the treatment of some solid tumors and confirm PD‐1 as an important cancer immunotherapy target. However, due to their large size, these antibodies have inferior tumor penetrability and perform poorly against many large solid tumors. Additionally, owing to their complex folding pathway, antibodies are not readily amenable to cell‐based technologies and poly‐specific molecule platforms that show promise as the next generation of immunotherapies. Using a combination of computational design and experimental approaches, we have developed a de novo miniprotein that specifically binds murine and human PD‐1 at the ligand interface. The 5.6 kDa protein contains three disulfide bonds, making it highly stable to thermal and chemical denaturation as well as proteolysis, and can be secreted solubly from mammalian cells. We have genetically and chemically attached this protein to a range of designed homo‐oligomers, including an icosahedral protein nanocage, to demonstrate its use as a modular binding domain. This small, hyperstable PD‐1 binding protein has ideal properties for application in a variety of targeted and cell‐based immunotherapy platforms.

ABS484

ANOTHER REASON WHY SOLVING LOTS OF PROTEIN STRUCTURES IS USEFUL: STRUCTURAL DIVERSITY IN THE MYCOBACTERIA DUF3349 FAMILY

Garry W. Buchko ¹, Jan Abendroth², John I. Robinson², Isabelle Phan³, Wesley C. Van Voorhis⁴, Peter J. Myler³, Thomas E. Edwards²

¹Pacific Northwest National Laboratory (Richland, United States); ²UCB (Bainbridge Island, United States); ³Seattle Children's Research Institute (Seattle, United States); ⁴University of Washington (Seattle, United States)

A superfamily of proteins with a "Domain of Unknown Function", DUF3349, is present predominately in Mycobacterium and Rhodococcus bacterial species suggesting that these proteins may have a biological function unique to these bacteria. Following upon our inaugural structure of a DUF3349 superfamily member, Mycobacterium tuberculosis Rv0543c, we determined the structures of three additional DUF3349 proteins: Mycobacterium abscessus MAB_3403c, Mycobacterium smegmatis MSMEG_1063, and MSMEG_1066. Like Rv0543c, the NMR solution structure of MSMEG_1063 revealed a monomeric five alpha‐helix bundle with a similar overall topology. Surprisingly, the crystal structure of MSMEG_1066 showed a five alpha‐helix protein with a strikingly different topology and a tetrameric quaternary structure. The tetrameric structure is not a by‐product of crystal packing, as size exclusion chromatography indicates MSMEG_1066 is also a tetramer in solution. The NMR solution structure of MAB_3403c revealed a monomeric alpha‐helical protein with a folding topology similar to the three C‐terminal helices in the protomer of the MSMEG_1066 tetramer, but, it is missing 18 N‐terminal residues necessary for tetramer formation. These structures, together with a more detailed bioinformatics analysis of the DUF3349 primary amino acid sequences, suggest two subfamilies within the DUF3349 family that each contain five ‐helices, but, adopt different folds. Moreover, one fold initiates the assembly of a unique tetrameric quaternary structure. The division of the DUF3349 into two distinct subfamilies would have been lost if structure solution had stopped with the first structure in the DUF3349 family, highlighting the insights generated by solving many structures in a protein family.

ABS485

SUBUNIT MASS ANALYSIS FOR MONITORING MULTIPLE ATTRIBUTES OF MONOCLONAL ANTIBODIES

Peiran Liu ¹

¹Bristol‐Myers Squibb (West Windsor Township, United States)

Multi‐Attribute Methods (MAMs) are appealing due to their ability to provide data on multiple molecular attributes from a single assay. If fully realized, such tests could reduce the number of assays required to support a product control strategy while providing equivalent or greater product understanding relative to the conventional approach. In doing so, MAMs have the potential to decrease development and manufacture costs by reducing the number of tests in a release panel. Among the many techniques used in the development of MAMs, mass spectrometry (MS) has been particularly effective due to the fact that comprehensive information can be collected at both peptide and protein levels. In this work, we report a MAM which is based on subunit mass analysis. The MAM assay is shown to be suitable for use as a combined method for identity testing, glycan profiling, and protein ratio determination for co‐formulated monoclonal antibody (mAb) drugs. This is achieved by taking advantage of the high mass accuracy and relative quantification capabilities of intact mass analysis using quadrupole time of flight mass spectrometry (Q‐TOF). Protein identification is achieved by comparing the measured masses of light chain (LC) and heavy chain (HC) of mAbs against their theoretical values. Specificity is based on instrument mass accuracy. Glycan profiling and relative protein ratios are determined by the relative peak intensities of the protein HC glycoforms and LC respectively. Results for these relative quantifications agree well with those obtained by the conventional hydrophilic interaction liquid chromatography (HILIC) and reversed phase LC methods. The suitability of this MAM method for use in a quality control setting is demonstrated through assessment specificity for mAb identity, and accuracy, precision, linearity and robustness for glycan profiling and ratio determination.

ABS486

INTERACTION STUDIES OF CALMODULIN‐LIKE PROTEIN 19 (CML19), THE CENTRIN 2 OF ARABIDOPSIS THALIANA, WITH RAD4 AND SAC3B TARGET PEPTIDES

Matteo Trande ¹, Marco Pedretti¹, Paola Dominici¹, Alessandra Astegno¹

¹University of Verona (Verona, Italy)

Arabidopsis centrin 2 (AtCEN2), also known as calmodulin‐like protein 19 (CML19), is a member of the EF‐hand superfamily of calcium (Ca2+)‐binding proteins. In addition to the notion that AtCEN2 participates in DNA nucleotide excision repair (NER) via its Ca2+ dependent interaction with RAD4, the Arabidopsis homolog of human xeroderma pigmentosum group C protein (XPC) [1], ATCEN2 was found to be part of TREX‐2 mRNA export complex, via direct interaction with SAC3B protein [2].

Starting from the hypothesis of a direct physical interaction of AtCEN2 with SAC3B and RAD4, herein we focused on a sequence‐based identification of the SAC3B and RAD4 binding sites to AtCEN2 and on a fragment‐based peptide design to investigate the thermodynamic and structural characterization of the intermolecular interactions, using a combination of NMR, CD and fluorescence spectroscopy as well as native page and isothermal titration calorimetry.

These complementary techniques showed that AtCEN2 binds to RAD4 peptide in a strictly Ca2+‐dependent manner, while SAC3B peptide interacts with both apo‐ and Ca2+‐bound AtCEN2 in vitro. Both interactions occur with high affinity, forming a protein‐peptide complex with a 1:1 stoichiometry. It was further suggested that the interaction of RAD4 and SAC3B peptides with AtCEN2 is mediated solely by the C‐terminal half of the protein.

Our data clearly demonstrated that RAD4 and SAC3B peptides interact with AtCEN2 in vitro, supporting the double functional role of this protein in NER and mRNA export systems.

[1] Lu Liang et al., 2006, Plant Mol Biol, 61, 345‐56

[2] Qing Lu et al., 2010, The Plant Journal, 61, 259270