Chemical Regulation of the Protein Quality Control E3 Ubiquitin Ligase C-terminus of Hsc70 Interacting Protein (CHIP)

Abstract

The ubiquitin ligase C-terminus of Hsc70 interacting protein (CHIP) is an important regulator of proteostasis. Despite playing an important role in maintaining proteostasis, little progress has been made in developing small molecules that regulate ubiquitin transfer by CHIP. Here we used differential scanning fluorimetry to identify compounds that bound CHIP. Compounds that bound CHIP were then analyzed by quantitative ubiquitination assays to identify those that altered CHIP function. One compound, MS.001, inhibited both the chaperone binding and ubiquitin ligase activity of CHIP at low micromolar concentrations. Interestingly, we found that MS.001 did not have activity against isolated U-box or tetratricopeptide (TPR) domains, but instead only inhibited full-length CHIP. Using in silico docking we identified a potential MS.001 binding site on the linker domain of CHIP and mutation of this site rendered CHIP resistant to MS.001. Together our data identify an inhibitor of the E3 ligase CHIP and provides insight into the development of compounds that regulate CHIP activity.

Graphical Abstract

Identification of a CHIP inhibitor:

CHIP is a protein quality control E3 ubiquitin-ligase that has been implicated in numerous diseases. Here we identified an inhibitor of CHIP and revealed that allosteric regulation of CHIP may be one way to regulate its activity.

Introduction

Many neurodegenerative diseases are characterized by the accumulation of misfolded proteins.^[1] To prevent the accumulation of misfolded proteins, cells utilize an array of protein quality control pathways. Among these pathways are molecular chaperones and the ubiquitin-proteasome system. The E3 ubiquitin ligase C-terminus of Hsc70-interacting protein (CHIP) sits at the interface of these two pathways and is critical for preventing the accumulation of misfolded proteins. CHIP binds chaperones through an N-terminal tetratricopeptide repeat (TPR) domain and recruits E2 ubiquitin-conjugating enzymes via a C-terminal U-box domain. This allows CHIP to mediate the ubiquitination of chaperone-bound misfolded proteins, targeting them for proteasomal degradation.^[2]

CHIP has been implicated as a neuroprotective protein in several neurodegenerative diseases including Alzheimer’s disease, Parkinson’s disease, amyotrophic lateral sclerosis (ALS), and the polyglutamine diseases.^[3] Overexpression of CHIP improves disease progression in models of neurodegeneration.^{[3q, 3ae, 3ah, 4]} Conversely, decreasing levels of CHIP exacerbates neurodegenerative disease and decreases organismal longevity.^{[3ah, 5]} Therefore, in these diseases increasing ubiquitin transfer by CHIP would be expected to be therapeutic.

In other diseases, CHIP-mediated degradation of immaturely folded proteins contributes to disease progression. For example, in both Cystic Fibrosis (CF) and Niemann-Pick Disease type C (NPC), the most common disease-causing mutations to cystic fibrosis transmembrane conductance regulator (CFTR^Δ508) and NPC Intracellular Cholesterol Transporter 1 (NPC1^I1061T), respectively, are ubiquitinated in a CHIP-dependent manner^[6]. Importantly, both CFTR^Δ508 and NPC1^I1061T mutants are functional, and increasing their levels is sufficient to alleviate symptoms of disease.^[7] Therefore, the degradation of both CFTR^Δ508 and NPC1^I1061T mediated by CHIP contributes to disease progression by preventing sufficient levels of functional protein from reaching cell membranes. In addition, in Spinocerebellar ataxia autosomal recessive type 16 (SCAR16), a disease caused by recessive mutations in CHIP, further decreasing the activity of mutant CHIP is predicted to lessen disease severity.^[8] Therefore, the development of CHIP inhibitors may be useful for the treatment of some diseases.

Here, we utilized differential scanning fluorimetry (DSF) to identify small molecules that alter the melting temperature (T_M) of CHIP. These compounds were then tested in a time-resolved fluorescence resonance energy transfer (TR-FRET) ubiquitination assay to identify molecules that alter ubiquitin transfer by CHIP. Using this assay we identified a compound, MS.001, that inhibited the ability to stimulate ubiquitin chain formation by CHIP. Surprisingly, MS.001 had no effect on ubiquitin chain formation mediated by the U-box of CHIP. We also assessed the effect MS.001 has on the ability of CHIP to bind chaperones using fluorescence polarization (FP) and found that MS.001 inhibited the ability of full-length CHIP to bind chaperones. Similar to our results with the U-box domain of CHIP, MS.001 had minimal effect on the ability of the TPR domain of CHIP to bind chaperones. To identify how MS.001 inhibits full-length CHIP we utilized in silico docking predictions to identify potential MS.001 binding sites on the surface of CHIP. Using site-directed mutagenesis we mutated a predicted MS.001 binding site and found this CHIP mutant was insensitive to MS.001 mediated inhibition. Finally, we performed structure-activity relationship (SAR) studies to define structural features of MS.001 important for modulation of CHIP. Together, these studies identified a novel CHIP inhibitor and suggest that allosteric regulation of CHIP is a viable method for modulating its activity.

Results and Discussion

Identification of small molecules that modulate CHIP activity

A vast amount of literature suggests that small molecules that regulate CHIP activity may be beneficial for the treatment of several diseases. To begin identifying compounds that regulate CHIP function, we utilized DSF as a primary screen to identify small molecules that bound CHIP. Initially, we performed DSF with CHIP alone to measure the T_M of CHIP. Consistent with previous results, we measured the T_M of CHIP at 39 ± 0.4°C (Figure 1a).^[8b] Once the baseline T_M for CHIP had been determined, we performed DSF with CHIP and individual compounds from the NIH Diversity Set V library (1594 compounds). From this screen we identified 73 small molecules (4.58%) that changed the T_M of CHIP by more than two standard deviations (Figure 1b). Utilizing these hits, we next performed TR-FRET based ubiquitination assays to identify compounds that altered ubiquitin transfer by CHIP. Of the 73 compounds that were hits in our primary screen, 16 (1%) inhibited CHIP at a concentration of 50μM (Figure 1c). Of these compounds, we selected MS.001 for further analysis (Figure 1d).

Figure 1: Identification of small molecules that modulate the melting temperature of CHIP.

A. The melting temperature of CHIP was determined by DSF. Recombinant CHIP was incubated with SYPRO Orange and incubated in a thermocycler at increasing temperatures. Fluorescence was measured and plotted versus temperature. B. Identification of compounds that alter the T_M of CHIP. CHIP was incubated with 100 μM of each small molecule from NIH Diversity Set V for 30 minutes prior to analysis by DSF. The red lines indicate two standard deviations from the mean. Compounds more than two deviations from the mean are considered hits. C. Identification of compounds that alter ubiquitin transfer by CHIP. TR-FRET-based ubiquitination assays were performed using compounds identified in B. The red lines indicate two standard deviations from the mean. Compounds more than two deviations from the mean are considered hits. D. Two-dimensional structure of MS.001.

MS.001 selectively inhibits CHIP

To validate that MS.001 inhibits CHIP, we independently resynthesized MS.001 (rMS.001) and performed ubiquitination assays at 33 μM and 100 μM.^[9] In this assay, both MS.001 and rMS.001 efficiently inhibited ubiquitin transfer by CHIP, validating MS.001 as a CHIP inhibitor (Figure 2a). In addition to confirming that MS.001 inhibited CHIP, we also wanted to confirm that this inhibition was specific to CHIP and not general to all proteins containing U-box domains. To test this, we assessed the ability of MS.001 to inhibit E4B’s U-box domain. Consistent with MS.001 selectively inhibiting CHIP and not being a general U-box inhibitor, we found that MS.001 had no effect on ubiquitin chain formation mediated by E4B’s U-box domain at 33 μM and 100 μM (Figure 2b). Together, these data are consistent with MS.001 selectively inhibiting ubiquitin transfer by CHIP.

Figure 2: MS.001 selectively inhibits CHIP.

A. Confirmation that MS.001 inhibits CHIP. MS.001 was resynthesized (rMS.001) and its composition was confirmed by mass spectrometry and NMR. rMS.001 was then compared to MS.001. Samples were analyzed by SDS-PAGE and Western blotting. B. MS.001 does not inhibit E4B’s U-box domain. In vitro ubiquitination assays were carried out with increasing concentrations of MS.001. Samples were prepared by SDS-PAGE and Western blotting.

MS.001 inhibits ubiquitin chain formation and chaperone binding by CHIP

CHIP bridges the ubiquitin and chaperone systems by recruiting E2 ubiquitin-conjugating enzymes via its U-box domain and binding molecular chaperones via its TPR domain. Based on our screen, MS.001 inhibited the E3 ligase activity of CHIP at a concentration of 50 μM. To determine how potently MS.001 inhibited ubiquitin transfer by CHIP we performed TR-FRET based ubiquitination assays with increasing concentrations of MS.001. From this assay we determined that MS.001 inhibits CHIP with an IC₅₀ of 3.3 ± 0.6 μM (Figure 3a). In addition to its E3 ligase activity, CHIP also acts as a co-chaperone by binding the C-terminal -EEVD motif of chaperones and stimulating the ubiquitination of chaperone-bound client proteins. To determine if MS.001 inhibits the CHIP/chaperone interaction, we utilized a rhodamine-labeled peptide that mimics the C-terminal -SSGPTIEEVD motif of Hsc70 in fluorescence polarization assays to monitor the CHIP/chaperone interaction.^[8b] Interestingly, MS.001 inhibited the binding of the peptide by CHIP with an IC₅₀ of 3.8 ± 0.8 μM (Figure 3b). Together this data indicates that MS.001 inhibits both ubiquitin transfer and chaperone binding by CHIP with similar IC₅₀ values.

Figure 3: MS.001 inhibits both the E3 ligase activity and chaperone binding at low micromolar concentrations.

A. MS.001 inhibits CHIP at low micromolar concentrations. TR-FRET ubiquitination assays were performed with increasing concentrations of MS.001. The hillslope was calculated as −1.053 with an R² of 0.9379. B. MS.001 inhibits the CHIP/chaperone interaction at a low micromolar concentration. FP assays were performed with a rhodamine-labelled peptide derived from the C-terminus of Hsc70 (r-SSGPTIEEVD) and increasing concentrations of MS.001.

Identification of the MS.001 binding site on CHIP

Because MS.001 did not inhibit either the U-box or TPR domains of CHIP in isolation (Figure 4), we reasoned that MS.001 likely bound the linker domain of CHIP, or a binding pocket that is only present on full-length CHIP. To determine where MS.001 bound CHIP, we attempted to solve the structure of MS.001 bound to CHIP using x-ray crystallography. Unfortunately, we were unable to obtain CHIP crystals that diffracted well enough to enable a high-resolution structure (data not shown). Alternatively, we utilized in silico docking simulations to predict potential binding pockets on CHIP. Of the potential binding sites predicted for MS.001, one binding pocket predicted a hydrogen bond between the MS.001 sulfoxide and the S159 hydroxyl group in the linker region of the asymmetrical CHIP dimer (Figure 5a, b). This site was particularly interesting because it is not located in either the TPR or U-box domains of CHIP that were not inhibited by MS.001 (Figure 4). To test if this pocket was a potential binding site for MS.001 on CHIP, we utilized a CHIP^S159A mutant to perform in vitro ubiquitination assays. Here, the CHIP^S159A mutant demonstrated wild-type levels of ubiquitination in our TR-FRET based ubiquitination assay (Figure 5c). However, unlike wild-type CHIP, CHIP^S159A was not inhibited by MS.001 (Figure 5c). Together, these data suggest that MS.001 allosterically inhibits CHIP by binding the linker domain of CHIP and locking it in an inactive conformation.

Figure 4. MS.001 does not inhibit the isolated U-box or TPR domains of CHIP.

A,B. MS.001 does not efficiently inhibit the U-box domain of CHIP. Ubiquitination reactions were carried out in the presence and absence of MS.001 for the times indicated. Samples were analyzed by SDS-PAGE and Western blotting (A) or by TR-FRET ubiquitination assays (B). C. MS.001 does not inhibit the TPR domain of CHIP from binding chaperones. FP assays were performed with a rhodamine-labelled peptide derived from the C-terminus of Hsc70 (R-SSGPTIEEVD) and either full-length CHIP or the TPR domain of CHIP in the presence of increasing concentrations of MS.001.

Figure 5: Computational modeling identifies a potential MS.001 binding site.

A. In silico identification of a potential MS.001 (yellow) binding site on the crystal structure of CHIP (2C2L). The extended protomer and bent protomer of the CHIP dimer are shown adjacent to the crystal structure of the E2 UbcH5 bound to the U-box domain of CHIP (2OXQ) (U-box-orange, UbcH5-yellow). Glide software was utilized to predict where on the surface of CHIP MS.001 binds. The TPR domain of CHIP is grey, its linker domain is blue, and its U-box is orange. UbcH5 is red. B. Representation of the predicted interactions between the CHIP S159 binding pocket and MS.001. Dashed lines represent hydrogen bonds between CHIP and MS.001. C. Mutation of the predicted MS.001 interacting residue, CHIP^S159A, renders CHIP insensitive to MS.001. TR-FRET based ubiquitination assays were performed for the time points indicated with either no CHIP, CHIP, or CHIP^S159A in the presence or absence of MS.001. D-F. Suggested model of CHIP inhibition by MS.001. Structural overlays of full-length CHIP (2C2L) and the U-box of CHIP bound to UbcH5 (2OXQ) were produced. As indicated in A and B, MS.001 is predicted to bind in the pocket present only in the bent protomer of CHIP. In this conformation the TPR domain of CHIP would interfere with the binding of an E2. To better visualize this individual protomers bound to UbcH5 are shown in the active (E) and inactive (F) conformation. In the inactive confirmation the TPR domain (grey) of CHIP clashes with the E2 UbcH5 (red, 50% transparent).

SAR by catalog identifies features of MS.001 required for CHIP inactivation

To gain a better understanding of the structural relationship between CHIP and MS.001, we performed structure-activity relationship studies using commercially available small molecules with similar chemical structures as MS.001 (Figure 6a). The five commercially available compounds used for SAR studies (MS.002-MS.006) shared a similar benzofurazan scaffold with differential modification of reactive groups (Figure 6a). To test the ability of MS.002-MS.006 to inhibit the activity of CHIP, we used fluorescence polarization and TR-FRET-based ubiquitination assays. In these assays, only MS.002 was able to significantly inhibit either the chaperone binding or ubiquitin chain formation activities of CHIP (Figure 6b,c). Of note, MS.004 and MS.005 caused an increase in signal in the TR-FRET ubiquitination assay but were not observed to stimulate ubiquitin chain formation by western blot analysis (Supplemental Figure 1a,b). Together, our SAR data supports the hypothesis of a selective interaction between MS.001 and the linker domain of CHIP and suggests that MS.001’s oxygen atom interaction with S159 of CHIP is important for binding.

Figure 6: SAR by catalog is consistent with the predicted MS.001 binding site.

A. Structures of commercially available small molecule analogs of MS.001 utilized in this study. B. Only MS.001 and MS.002 inhibit CHIP/chaperone interaction. Fluorescence polarization assays were performed with the indicated concentrations of each MS.001 analog to determine if they inhibit the CHIP/chaperone interaction. C. Only MS.001 and MS.002 inhibited the E3 ligase activity of CHIP when TR-FRET based ubiquitination assays were performed with increasing concentrations of the indicated MS.001 analog.

Conclusion

Developing compounds that regulate ubiquitin ligases such as CHIP is an attractive way to treat many diseases. Here, we developed a pipeline to identify small molecules that alter CHIP activity (Figure 1). We identified MS.001 as a small molecule that inhibits CHIP (Figure 1, 2). We found that MS.001 inhibits both ubiquitin transfer and chaperone binding by CHIP with low micromolar affinity (Figure 3). Interestingly, while MS.001 inhibited full-length CHIP, it did not inhibit the U-box domain of CHIP from stimulating ubiquitin chain formation or the TPR domain of CHIP from binding chaperones (Figure 4). Together, these data suggest that MS.001 may bind to CHIP outside of these domains. To identify potential sites where MS.001 binds CHIP, we utilized in silico docking simulations. Using these simulations, we identified a potential MS.001 binding site in a pocket present only in the inactive protomer of CHIP. Mutation of S159, a predicted MS.001 interacting residue in this pocket, yielded a CHIP mutant with WT activity but rendered CHIP insensitive to MS.001 mediated inhibition (Figure 5). SAR studies also provided insights into the structural aspects of MS.001 important for the inhibition of CHIP (Figure 6). Importantly our study suggests that the allosteric regulation of CHIP using small molecules can be an effective method for modulating CHIP activity.

Allosteric regulation of CHIP

Our data suggest that MS.001 inhibits CHIP by binding a pocket in the linker domain that exists when CHIP is in an inactive state, unable to bind E2 ubiquitin-conjugating enzymes (Figure 5 and 6). This is interesting because structural analysis of CHIP has offered differing views about the conformation of the linker domain of CHIP. One crystal structure that contains the TPR, helical linker, and U-box domains of CHIP shows CHIP as an asymmetrical dimer.^[10] In this structure, one protomer has an extended helical linker with the TPR and U-box domains accessible for the recruitment of chaperones and E2 ubiquitin-conjugating enzymes, respectively. However, the helical linker domain of the opposing protomer is interrupted, resulting in repositioning of the TPR domain to abut the U-box domain. This repositioning sterically hinders the recruitment of E2~Ub conjugates to its U-box domain and renders this protomer inactive.^[10] In a second structure of CHIP that contains only the helical linker and U-box domains of CHIP, both of helical linkers are present in an extended conformation like the active CHIP protomer of the asymmetrical dimer. In this structure, it would be expected that both CHIP protomers would be active and capable of recruiting E2~Ub conjugates to ubiquitinate substrates.^[11] Finally, a third set of structural studies were carried out using amide hydrogen exchange mass spectrometry (HX-MS). In this study, it was suggested that CHIP likely exists predominantly as a symmetrical homodimer. However, unlike the previous structure of a symmetrical dimer, the data suggested that both protomers of CHIP would share the interrupted helical linker of the inactive version of CHIP.^[11–12] Together, existing structural data are consistent with the hypothesis that CHIP protomers are dynamic and capable of alternating between the two CHIP conformations observed in structural studies.^[10–12] Consistent with our data, one possibility is that MS.001 inhibits CHIP by locking both CHIP protomers in an inactive, interrupted helical conformation. While our data are consistent with CHIP binding this pocket, we were unable to obtain structural data that clearly delineates how MS.001 binds CHIP. In the future structural data that delineates the details of the CHIP/MS.001 interaction will be helpful for future development of CHIP inhibitors that may be useful for treating some rare diseases.

MS.001 inhibits the CHIP/chaperone interaction

One surprising observation was that in addition to inhibiting the E3 ubiquitin ligase activity of CHIP, MS.001 also prevents the CHIP/chaperone interaction (Figure 3b, 4c). This was unexpected as both the active and inactive CHIP conformers in the crystal structure of CHIP bind the C-terminal peptide of Hsp90.^[10] Despite the C-terminal peptide of Hsp90 binding both the active and inactive conformer of CHIP, it is worth noting that in the inactive conformer CHIP’s U-box domain is positioned within angstroms of the N-terminus of the peptide. In our FP experiments, we utilized a peptide that corresponds to the C-terminus of Hsc70. This peptide is both longer than the peptide used in the co-crystal structure and contains a N-terminal rhodamine. The addition of additional amino acids and the N-terminal fluorophore add increased size to this peptide and may make it inaccessible to the TPR domain of CHIP due the proximity of the U-box domain in this conformer.

Small molecules that activate CHIP may also be beneficial.

While CHIP Inhibitors may be beneficial for patients with a limited number of rare diseases, CHIP activating compounds would likely be useful in a wide array of diseases, including neurodegenerative diseases. In multiple model systems, CHIP is protective against proteins that aggregate and contribute to the development of neurodegenerative diseases. In these diseases, CHIP targets misfolded proteins for degradation, reducing protein aggregation and providing a protective effect. Increasing the transfer of ubiquitin by CHIP would be expected to accelerate degradation of misfolded proteins, providing a therapeutic effect in diseases of protein misfolding. In the future, development of small molecules that bind selectively to the active confirmation of CHIP may be one way to develop CHIP activating compounds. This would be highly relevant as activating ubiquitin transfer by CHIP would be expected to be protective in many neurodegenerative diseases.

Inhibition of CHIP may be beneficial for some rare diseases

For several rare diseases, the inhibition of CHIP could attenuate the aberrant degradation of functional, disease-related proteins. In the case of NPC, CHIP stimulates clearance of a mutant form of NPC1, NPC1^I1061T.^[6a] Importantly for NPC treatment, NPC1^I1061T retains activity but is degraded by CHIP prior to reaching the lysosomal membrane.^[7a] Therefore, stabilizing NPC1^I1061T levels through the inhibition of CHIP is a potential therapeutic option for the treatment of NPC. Analogous to the NPC1^I1061T mutant, CFTR^Δ508 degradation is stimulated in part by CHIP-mediated ubiquitination.^[7b] Thus, inhibition of CHIP is expected to increase CFTR^Δ508 levels, and this increase in CFTR^Δ508 levels would be therapeutic.^[13] Lastly, mutation of CHIP itself results in diseases including Spinocerebellar ataxia autosomal recessive type 16 (SCAR16) and Spinocerebellar ataxia 48 (SCAR48).^[14] Most mutations that cause these spinocerebellar ataxias result in destabilization of CHIP, and linear models of SCAR16 as a function of the biochemical properties of CHIP suggest that further inhibition of mutant CHIP activity may attenuate disease phenotypes.^[8] Together, these studies suggest that CHIP inhibitors may be therapeutic in specific diseases where CHIP stimulates the degradation of functional proteins or in cases where a mutation in CHIP itself alters CHIP function and directly contributes to disease progression.

Experimental Section

Constructs

CHIP was cloned into pGEX6p-1 as previously described.^[15] UbcH5c WT pET28a was obtained from Addgene (Addgene 12643).^[16] Hsp70 was cloned into pMCSG7 as previously described^[3ab]. Mutations were introduced into wild-type CHIP in a pGEX6P-1 vector via QuikChange Lightning Mutagenesis Kit (Stratagene).

Protein Purification

E. coli BL21(DE3) containing pET28 UbcH5c was grown at 37 °C and induced with 1.0mM IPTG for 24 hours at 16 °C. Cell pellets were resuspended in 30mM MES pH 6.0 and disrupted by sonication (3 cycles of 30 second pulses). The extracts were cleared by centrifugation and applied to SP Fast Flow (GE). UbcH5 was eluted with a 0–0.5M NaCl gradient in 30mM MES pH 6.0 with 2mM DTT and 0.5mM PMSF. Fractions containing UbcH5 were purified by size-exclusion chromatography in 25mM sodium phosphate pH 7.0, 2mM DTT.

E. coli BL21(DE3) containing pGEX6p-1 CHIP was grown at 37 °C and induced with 1.0mM IPTG for 24 hours at 16 °C. Cell pellets were resuspended in NETN buffer, clarified, and applied to glutathione beads. CHIP was cleaved from glutathione beads with H3c protease and purified by size-exclusion chromatography in 50mM Tris, 50mM KCl, pH 7.5.

E. coli BL21(DE3) containing pMCSG7 Hsp72 was grown at 37 °C and induced with 1.0mM IPTG for 24 hours at 16 °C. Cell pellets were resuspended in 50mM Tris pH 7.5, 1M NaCl, 20mM imidazole. Cell pellets were disrupted by sonication (3 cycles of 30 second pulses). The extracts were cleared by centrifugation and applied to nickel agarose (Gold Biotechnology) before elution with a 0–0.5M Imidazole gradient. Fractions containing Hsp72 were then purified by size-exclusion chromatography in 50mM Tris, 50mM KCl, pH 7.5.

Ubiquitination Assays

Ubiquitination reactions were performed as previously described.^{[3ab, 8b, 17]} Reactions were stopped by addition of 4× Laemmli buffer and boiling. Ubiquitination reactions were then visualized by SDS-PAGE, and Western blotting with appropriate antibodies.

Western Blotting

For in vitro ubiquitination assays, reactions were stopped by addition of 4× Laemmli buffer and boiled for 3 minutes at 100 °C. In vitro ubiquitination assays were resolved by SDS-PAGE and visualized by Western blotting with the indicated antibodies: Anti-Hsp/Hsc-70 SC33575 (Santa Cruz), Anti-CHIP AB10000 (Millipore), or Anti-Ubiquitin P4D1 (Cell Signaling).

Thermal Shift Assay

Proteins were diluted to 5μM in PBS and placed at 4 °C. The SYPRO orange protein stain was diluted to 200× from a 5000× stock in PBS. In a 96 well plate, 45μL of protein was combined with 5μL of SYPRO orange solution in triplicate. The mixture was briefly spun down to avoid bubbles before fluorescent readings were taken. Beginning at 25 °C each successive one-minute cycle increased the temperature by one-degree Celsius, ending at 95 °C. The melting curve for each protein was obtained by Gaussian Curve regression set to fit points between the minimum and the maximum intensity read for each sample. The melting temperature was defined as the temperature halfway between the minimum and maximum value.^[18]

Fluorescence Polarization

Fluorescence polarization experiments were carried out as previously described.^[19] Briefly polarization was determined by plotting FP readings as a function of CHIP protein concentration after incubation with 20nM Rhodamine-B-labelled Hsc70 peptide diluted in assay buffer (50mM HEPES, 75mM NaCl, and 0.01% v/v Triton X-100 pH 7.4). Following the addition of peptide to CHIP dilutions, the plates were covered and incubated at either room temperature or 37 °C for 30 minutes to attain binding equilibrium. Polarization values in millipolarization units (mP) were measured using a Tecan Spark plate reader at excitation/emission of 544/612 nm. Fluorescence polarization experiments were performed using 384-well, black, low-volume, flat-bottom plates. All experiments were performed at least two independent times in triplicate shown as an average. Error is shown as standard deviation. Experimental data were analyzed using GraphPad Prism.

TR-FRET ubiquitination assays

In vitro ubiquitination assays were performed using a ubiquitination mix containing 2mM ATP, 4mM MgCl₂, 50nM E1, 200nM UbcH5c, 500nM Fluorescein-labelled ubiquitin (U-580 BostonBiochem), 50nM Terbium-labelled ubiquitin PV4375 Invitrogen), 50nM Hsp70, and 1μM CHIP variant. Reactions were performed for the times indicated at 37°C and quenched with 20mM EDTA. TR-FRET readings were taken on a Tecan Spark plate reader at excitation of 340 nm and emission at 495/520 nm following a 100-microsecond delay after excitation. TR-FRET results are presented as a ratio of 520/495 nm fluorescence readings.

In silico docking simulations

The CHIP crystal structure (PDB ID: 2CL2) was imported into the Schrödinger Maestro suite (2017–3). Water molecules further than 5 Å from the protein surface were removed, hydrogen bonds were optimized at pH 7.0, and the structure was minimized in the OPLS2005 forcefield using the protein preparation wizard. Three-dimensional structures of MS.001 at pH 7.0 ± 2.0 were generated using the LigPrep module within Maestro. Prepared MS.001 structures were then docked into a 60 × 60 × 60 Å grid centered on the centroid of the CHIP structure using Glide in Extra-Precision (XP) mode, and output poses were ranked according to XP GlideScore.^[20]