Skip to main content
NCBI home page
ACS Medicinal Chemistry Letters logoLink to ACS Medicinal Chemistry Letters
. 2024 Jun 17;15(7):1071–1079. doi: 10.1021/acsmedchemlett.4c00154

Multiplexed Native Mass Spectrometry Determination of Ligand Selectivity for Fatty Acid-Binding Proteins

Michelle Q Phan †, Indu R Chandrashekaran †,‡, Naureen Akhtar †,‡, Evgenia Konstantinidou †,‡, Shane M Devine †, Bradley C Doak †,‡, Thomas Nebl ⊥, Darren J Creek §, Martin J Scanlon †,‡, Raymond S Norton †,‡,*
PMCID: PMC11247632  PMID: 39015264

Abstract

graphic file with name ml4c00154_0007.jpg

Although multiple approaches for characterizing protein–ligand interactions are available in target-based drug discovery, their throughput for determining selectivity is quite limited. Herein, we describe the application of native mass spectrometry for rapid, multiplexed screening of the selectivity of eight small-molecule ligands for five fatty acid-binding protein isoforms. Using high-resolution mass spectrometry, we were able to identify and quantify up to 20 different protein species in a single spectrum. We show that selectivity profiles generated by native mass spectrometry are in good agreement with those of traditional solution-phase techniques such as isothermal titration calorimetry and fluorescence polarization. Furthermore, we propose strategies for effective investigation of selectivity by native mass spectrometry, thus highlighting the potential of this technique to be used as an orthogonal method to traditional biophysical approaches for rapid, multiplexed screening of protein–ligand complexes.

Keywords: Native Mass Spectrometry, Ligand Selectivity, Fatty Acid-Binding Protein, Drug Discovery

Native mass spectrometry (nMS) is a rapid, sensitive and high-throughput technique most commonly used to probe noncovalent interactions of biomolecules, particularly the binding stoichiometries and affinities of folded intact proteins and their small-molecule ligands.1−3 While traditional biophysical techniques such as nuclear magnetic resonance (NMR) spectroscopy, protein X-ray crystallography, isothermal titration calorimetry (ITC) and surface plasmon resonance (SPR) remain effective in target-based drug discovery,4 limitations associated with these methods have encouraged increasing applications of nMS as an orthogonal approach for ligand screening.5−9 To elaborate, NMR requires milligram quantities of protein and is challenging for larger proteins; X-ray crystallography can be time-consuming and is often limited by the need for crystallization, ITC is low-throughput and requires moderate quantities of protein, and SPR requires immobilization of one of the binding partners, which may affect the binding site. In contrast, nMS utilizes only picomole quantities of protein and compound, is less restricted by analyte size, and does not require labeling, crystallization or immobilization.10,11 Furthermore, nMS possesses a wide dynamic range, with reported KD values in the low nanomolar and high millimolar range demonstrating good agreement with those generated by SPR and ITC.5−7 While throughput was historically limited owing to the need for manual sample preparation and data acquisition, advances in instrumentation and technology have enabled nMS workflows to be automated, facilitating the use of nMS for fragment screening purposes.8,12,13

Despite the emergence of nMS as a powerful quantitative screening tool in target-based drug discovery, the technique continues to be underutilized. Given that biophysical techniques such as ITC and SPR remain well-established for measuring dissociation constants, nMS is perhaps more beneficial as a complementary method when it is applied in a qualitative manner. Although nMS can preserve weakly bound protein–ligand complexes in the gas phase, significant optimization of instrument conditions is often required to maintain and accurately quantify the affinity of such interactions. Unlike most other biophysical approaches, however, nMS offers the capability to directly visualize and resolve individual analytes and complexes within heterogeneous mixtures with relative ease. Indeed, recent technical advances have enabled increasing applications of multiplexed nMS-based screening in which multiple ligands are evaluated simultaneously against a single protein target, ranging from low complexity samples containing only a few small-molecule binders to highly complex natural product extracts.14−18

Although multiplexed screening of ligands is not exclusive to nMS,19,20 analysis of protein mixtures via current approaches remains challenging. This is especially relevant for the investigation of target selectivity, where selectivity profiles are typically developed by generating affinity data separately for each protein of interest, which is often both time-consuming and resource-intensive. While implementation of nMS for the investigation of ligand selectivity against multiple protein targets has yet to be reported, nMS has shown potential as a target identification approach, with a recent study successfully identifying the protein partner of a single ligand from a protein mixture.21 Thus, the capability of nMS to achieve rapid, multiplexed screening while simultaneously identifying hits with favorable binding affinity and selectivity profiles highlights the potential for screening campaigns to be accelerated.

Fatty-acid binding proteins (FABPs) are a family of intracellular lipid chaperones that regulate the uptake and intracellular distribution of lipophilic ligands such as fatty acids.22 FABPs play a central role in coordinating lipid responses and signaling in cells, but the mechanisms of action of these proteins in metabolic signaling remain unclear. As FABPs are important mediators of metabolic processes and play key roles in metabolic diseases linked to aberrant lipid utilization, such as diabetes and cancer, pharmacological modulation of FABPs may provide opportunities to target signaling pathways in a tissue-specific manner and facilitate the development of novel therapeutics for metabolic diseases.23

There are 10 isoforms of FABPs, each of which is expressed abundantly in different tissues involved in active lipid metabolism. Among the FABPs, isoforms 1–5 are the most well-characterized. Although these FABP isoforms show only around 20–70% sequence similarity, they share a highly conserved tertiary structure and exhibit shared ligand specificity.24 Compared to other FABPs, FABP1 possesses a larger solvent accessible binding cavity and is unique in its ability to bind two fatty acids per protein molecule.25 In contrast, FABP2 has a small solvent accessible binding cavity. FABP5 has six conserved cysteine residues, with Cys120 and Cys127 forming a disulfide bond, a feature that is unique to this isoform.26 The FABP isoforms also possess both unique and overlapping functions in specific tissues. For example, FABP4 and FABP5 are coexpressed in adipocytes and reduction of FABP4 levels in adipocytes is compensated by overexpression of FABP5. Structure–function studies indicate that conformational changes driven by specific ligand binding promote protein–protein interactions that dictate the specific function of each FABP isoform.27

Development of selective inhibitors for these proteins has been challenging, and no small-molecule inhibitors of FABPs have entered the clinic.28 Several known drugs bind to FABPs, but exhibit poor isoform selectivity and often show similar binding specificity to nuclear hormone receptors, thereby limiting their utility in selectively targeting FABPs. Higher throughput approaches for studying selectivity are therefore necessary to facilitate the development of compounds that display significant selectivity for various FABP isoforms. Such compounds will serve as valuable tools with which to dissect the cellular functions of FABPs.

In this study, we demonstrate the application of nMS as a rapid and multiplexed technique for investigating the selectivity of known FABP ligands. We successfully identified and quantified up to 20 different protein species in mixtures containing a single ligand and up to five FABP isoforms, thus enabling the selectivity profiles of eight different ligands to be determined. These selectivity profiles were consistent with those generated by traditional solution-phase approaches, encompassing a wide dynamic range. This study shows that nMS can be used as an orthogonal method to traditional biophysical approaches for multiplexed and higher-throughput screening of selective ligands.

Ionization and Quantification of FABP Mixtures by nMS

FABP isoforms 1–5 were visible by nMS, with low charge states indicative of folded protein (Figure 1A).2 Identical charge state distributions were observed, likely owing to the similar masses and highly conserved tertiary structure of the isoforms, with all five proteins ranging between 14 and 15 kDa. Measured masses were consistent with expected values for all isoforms (Supporting Table S1). Additional peaks corresponding to noncovalent binding interactions with acetic acid were also present with variable intensities for FABPs 2–5. Although these interactions appear to be of moderate affinity, especially for FABP5, in which approximately half of its total protein signal is in complex, these spectral intensities are in fact the manifestation of a low affinity interaction (mM) caused by the high concentration of acetate in solution.29 Given that the binding of small-molecule ligands was not affected by the presence of acetic acid (Figure 2), quantification of each isoform was therefore calculated as the total abundance of both unbound and acetic acid-bound species. Despite all isoforms being equimolar (10 μM) in their respective solutions, as verified via a bicinchoninic (BCA) assay, the total signal intensity of each isoform was found to vary in the gas phase (Figure 1A), presumably owing to differences in protein ionization efficiency. As the FABPs possess similar properties, efforts to optimize instrument parameters yielded similar instrument conditions for all isoforms. Specifically, application of a +0.7–1.0 kV potential to emitter tips was used and a capillary temperature of 100 °C was preferred. S-lens RF level and trapping gas pressure were typically set at 200 and 1.0, respectively, to enhance ion transmission and minimize in-source dissociation.

Figure 1.

Figure 1

Open in a new tab

Native mass spectra of 10 μM FABP isoforms 1–5 with unbound and acetic acid-bound protein species visible (A). Equimolar mixture of 10 μM FABPs 1–5 (B). All isoforms were observed with charge states +6, +7 and +8. Peaks corresponding to the unbound proteins (circles) and their respective acetic acid-bound species (stars) are as indicated. Quantification of each isoform in the spectrum as a fraction of total protein (%) is shown on the right.

Figure 2.

Figure 2

Open in a new tab

Molecular structure of WY14643 (A) and its affinities for each FABP isoform as determined in a fluorescence polarization (FP) assay (B) are shown. Representative mass spectra of 10 μM FABPs 1–5 with (C) 10 μM and (D) 50 μM of WY14643. Peaks corresponding to the unbound proteins, acetic acid-bound and ligand-bound species are as indicated. Quantification of each ligand-bound isoform as a fraction of the total individual isoform signal (%) is shown on the right.

To determine an appropriate method for quantification of protein mixtures, FABPs 1–5 were buffer exchanged separately and combined in an equimolar manner (Figure 1B). As observed previously with spectra of individual isoforms, the fraction of each isoform, calculated as the total abundance of the isoform divided by the abundance of total protein, also showed variation in the mixture. Further optimization of instrument parameters showed similar variation in isoform ratios, suggesting that apparent protein abundances may be a consequence of both inherent ionization efficiency and ion suppression by other isoforms in the mixture.21,30 It is worth noting that one of the main assumptions applied in quantification of single protein–ligand binding by nMS is that the ionization efficiency of the unbound and bound states of a given protein is identical.3,10 We have also adopted this assumption, thus deeming these differences in isoform abundance immaterial to our selectivity measurements.

nMS for Rapid and Multiplexed Measurement of Selectivity

WY14643 is an agonist of the nuclear receptor peroxisome proliferator-activated receptor α (PPARα)31 and was found to bind promiscuously to the FABPs in a fluorescence polarization (FP) assay (our unpublished results) (Figure 2B). Multiplexed analysis of FABPs 1–5 incubated with WY14643 displayed distinct variability of the protein–ligand complex formation across all isoforms (Figure 2). Although the total abundance of each isoform was again found to be variable, the ionization efficiencies of individual isoforms and their respective bound complexes were assumed to be identical.1 Notably, up to 20 binding events were identified by nMS, corresponding to the unbound, acetic acid-bound, ligand-bound, and acetic acid-ligand-bound states of all five isoforms, exemplifying the high-resolution capabilities of the Orbitrap mass analyzer. In addition, the improved resolution offered by use of submicron emitters was critical for deconvoluting protein species of similar mass,16,17 such as the FABP3-acetic acid and FABP4-WY14643 complexes, which differed by only 8 Da (Figure 2C).

In an equimolar mixture of FABPs 1–5 and WY14643 (10 μM of each FABP and 10 μM ligand), binding interactions were detected for only four isoforms (Figure 2C). FABP3 exhibited the highest fraction of bound complex at 23%, followed by FABP4 (18%), FABP1 (7%) and FABP5 (7%), which is consistent with known affinity values (Figure 2B). No formation of ligand-protein complex was observed for FABP2, suggesting that there was either no binding of WY14643 to the isoform or interactions were too weak at that ligand concentration to be detected in the gas-phase. Further analysis with an equal concentration of ligand to the total protein concentration (50 μM) produced a different selectivity ranking, with WY14643 no longer most selective for FABP3 (Figure 2D). Nevertheless, we regard the limiting ligand concentration (10 μM) to reflect the ‘true’ selectivity, as the ligand typically exhibits preferential binding to the isoform for which it has the highest affinity. In contrast, although higher ligand concentrations enable weaker binding interactions to be detected, there is increased potential for nonspecific binding in the gas phase, which may affect selectivity measurements. Indeed, at 50 μM WY14643, fractions of bound protein were increased across all isoforms, enabling the detection of even the weak affinity (millimolar) interactions of FABP2, thus demonstrating the ability of nMS to detect complexes over a wide dynamic range in a single spectrum.

The selectivities of FABP inhibitors BMS309403, compound 77, compound 95(32) and PPARα agonist GW7647 (Supporting Figure S1) were also investigated by nMS and found to be in good agreement with expected selectivity rankings (Supporting Table S2). In comparison to WY14643, which exhibited moderate to weak affinity across the different FABPs, these ligands have affinities in the nanomolar range. Interestingly, at a ligand concentration of 10 μM, all four compounds achieved almost complete saturation of the most selective isoform. Further incubation with 50 μM ligand resulted in saturation of additional isoforms possessing comparatively weaker nanomolar affinity interactions (Supporting Table S2). Exceptions were observed at excess ligand concentrations for BMS309403 and GW7647, which are poorly soluble and hence showed binding only to their respective preferred isoforms, owing to compound precipitation.

In contrast to WY14643, which exhibited micromolar affinity across most isoforms and thus showed binding to multiple proteins at the limiting ligand concentration, nanomolar affinity ligands such as compounds 77 and 95 typically achieved saturation of the isoform for which they showed the greatest selectivity, suggesting a higher degree of selectivity. However, given that saturation of other isoforms was observed for compounds 77 and 95 at excess ligand concentrations, it appears that the degree of selectivity as measured by nMS is also dependent on the affinity of binding interactions, where higher fractions of bound protein may indicate a high affinity interaction rather than high selectivity. Therefore, the fold-change in the fraction of bound protein across different isoforms is not a quantitative measure of the degree of selectivity. Instead, nMS provides a qualitative indication of selectivity, with rankings indicating the most selective isoform.

Tailoring Protein Mixtures for Unambiguous Deconvolution

Owing to the narrow mass range of the different FABP isoforms, incubation of several ligands yielded mixtures containing species of similar mass. Given that the FABP isoforms exhibit identical charge state distributions, these species would be indistinguishable by nMS because of overlap of m/z peaks arising from isobaric and glycosylated variants.33 This was observed with the nonsteroidal anti-inflammatory drug, niflumic acid (Figure 3A), where the mass of its bound complex with FABP3 differs by only 3 Da from unbound FABP2. Indeed, titration of niflumic acid with individual isoforms indicated the potential m/z peak overlap of these species (Figure 3C). To determine whether multiplexed analysis for all five isoforms was feasible, mixtures of FABPs 1–5 with and without niflumic acid were compared (Supporting Figure S2). An apparent increase in the fraction of total protein of FABP2 was observed after ligand incubation, from 27% to 38%. Conversely, there was an apparent decrease in the total protein fraction of FABP3, from 22% to 15% before and after the addition of niflumic acid, respectively. Moreover, deconvolution of the apo FABP2 m/z peak showed a mass increase of 1.2 Da, thus confirming overlap with the niflumic acid-bound FABP3 m/z peak.

Figure 3.

Figure 3

Open in a new tab

Molecular structure of niflumic acid (A) and its affinities for each FABP isoform as determined by ITC (B). Mass spectral overlay of individual FABP isoforms (10 μM) with 10 μM niflumic acid (C), peaks of apo FABP2 and ligand-bound FABP3 are highlighted. Representative mass spectra of FABPs 1,2,4,5 (D) and FABPs 1,3,4,5 (E) with 50 μM niflumic acid. Peaks corresponding to the unbound proteins, acetic acid-bound and ligand-bound species are as indicated. Quantification of each ligand-bound isoform as a fraction of the total individual isoform signal (%) is shown on the right.

In order to facilitate unambiguous identification and quantification of all protein species, selectivity measurements of niflumic acid were obtained with different FABP mixtures, omitting either FABP3 (Figure 3D) or FABP2 (Figure 3E). Omission of FABP3 from the mixture produced a selectivity profile consistent with both ITC affinities (Figure 3B) and individual isoform data (Figure 3C), with niflumic acid being most selective for FABP4 (37% complex), followed by FABP1 (30%), FABP5 (8%) and FABP2 (0%) respectively (Figure 3D). An identical selectivity ranking was observed with omission of FABP2 from the mixture (Figure 3E), but with niflumic acid being most selective for FABP3 (41% bound protein), for which it has the strongest binding affinity by ITC (Figure 3B). Importantly, since niflumic acid was present in excess (50 μM), replacement of FABP2 with FABP3 did not appear to affect its binding to FABP1, FABP4 and FABP5. The fractions of bound protein were identical for these three isoforms across both mixtures, demonstrating the reproducibility of selectivity measurements by nMS across different protein mixtures.

Another strategy for ensuring unambiguous assignment is to replace proteins of similar mass with a tagged construct. Given that FABP3 tended to form complexes with mass similar to those of other protein species in the same mixture, we substituted the protein with a His6-tagged construct. Selectivity measurements with niflumic acid using this mixture produced an identical selectivity ranking to the previous approach, with niflumic acid most selective for FABP3 followed by FABP4 and FABP1 at bound protein fractions of 17%, 14% and 12%, respectively (Figure 4). Although an FABP5 complex was not detected in this spectrum, the overall selectivity profile is consistent with ITC data. In particular, addition of a His6-tag did not appear to hinder ligand binding to FABP3, thus demonstrating that selectivity measurements can be reproduced with tagged proteins.

Figure 4.

Figure 4

Open in a new tab

Representative mass spectra of FABPs 1–5 (FABP3 substituted with His6-FABP3) with 10 μM niflumic acid. Peaks corresponding to the unbound proteins, acetic acid-bound and ligand-bound species are as indicated. Quantification of each ligand-bound isoform as a fraction of the total individual isoform signal (%) is shown on the right.

In-Source Dissociation of FABP-Ligand Complexes Influences Apparent Selectivity by nMS

Optimization of instrument parameters for minimal dissociation of gaseous ion complexes is a crucial process for quantifying protein–ligand interactions. However, the extent to which intact complexes dissociate in the ion source is dependent on the affinity, the nature of the binding interactions and the protein itself.34,35 To determine whether multiplexed selectivity measurements by nMS were influenced by the in-source dissociation of individual FABP isoforms,36 a series of ligand titrations was acquired for niflumic acid (Figure 5). Electrospray ionization measurements were performed using a capillary temperature of 100 °C and without collisional activation to minimize in-source dissociation. Saturation of various isoforms was achieved with a ligand concentration of 100 μM.

Figure 5.

Figure 5

Open in a new tab

Ligand titration of niflumic acid (0.4 to 100 μM) with 10 μM FABP1 (A), FABP2 (B), FABP3 (C), FABP4 (D) or FABP5 (E). Solid points represent the fractional abundances of bound protein measured by nMS, and the solid curve is the best fit of these data. The red dashed curve shows the expected curve using the KD values calculated from the best fit of the data and corrected for in-source dissociation (F).

Since FABP1 (Figure 5A), FABP2 (Figure 5B) and FABP5 (Figure 5E) possess weaker affinities for niflumic acid, saturation was not achieved, particularly for FABP2 which exhibited very weak binding. In contrast, near saturation was observed for FABP3 (Figure 5C) and FABP4 (Figure 5D), with plotted curves approaching limiting values of 59% and 74%, respectively. This indicates that the FABP3 complexes are more susceptible to in-source dissociation compared to FABP4. Although FABP1 and FABP5 complexes also show dissociation comparable to that of FABP3, with limiting values of 52% and 48%, respectively, these are likely attributed to weaker binding interactions leading to greater dissociation, even in identical instrument conditions. Interestingly, despite varying degrees of in-source dissociation, selectivity rankings were consistent across all ligand concentrations (Supporting Table S4).

KD values for each isoform were also calculated from the ligand titration experiments. Without correcting for in-source dissociation, the apparent affinities of niflumic acid were almost identical for FABP3 and FABP4 (7.0 and 7.3 μM, respectively). However, after considering the different dissociation factors, niflumic acid was found to be 1.4-fold more selective for FABP3 (Figure 5F), compared to ∼5-fold by ITC. This suggests that the similar fractions of bound protein for the two isoforms could be caused by a combination of in-source dissociation and reduced selectivity in the gas phase. Furthermore, affinities measured by nMS for niflumic acid were overall found to be weaker for all isoforms compared with ITC, indicating that binding interactions are attenuated in the gas phase, even after correcting for in-source dissociation.

In contrast to niflumic acid, other nanomolar affinity compounds were found to achieve near complete saturation with negligible in-source dissociation. Indeed, ligand titrations of FABP4 and FABP5 with compound 95 both approach limiting values of ∼0.95, suggesting that the nanomolar affinity interactions with these isoforms were not affected by in-source dissociation (Figure 6). These compounds exhibited high affinity interactions predominantly with FABP1, FABP4 or FABP5 (Supporting Table S2), implying that FABP3 is most affected by in-source dissociation. Nevertheless, it is important to note that, despite the decrease in selectivity observed for FABP3, the overall selectivity profile generated via nMS for niflumic acid remains consistent with that measured by traditional biophysical approaches.

Figure 6.

Figure 6

Open in a new tab

Ligand titration of compound 95 (0.1 to 50 μM) with 10 μM FABP4 (A) and FABP5 (B). Solid points represent the fractional abundances of bound protein measured by nMS, and the solid curve is the best fit of these data. The red dashed curve shows the expected curve after correction for in-source dissociation. The structure of compound 95 is shown (C).

In addition to selectivity, it is possible to quantify the fractions of bound proteins for affinity determination. However, it is worth noting that the fraction of bound protein should not be used solely for affinity determination. Since proteins in a mixture are essentially competing for ligand binding, the fraction of a given complex can be influenced by the affinities of other complexes. Furthermore, since ionization efficiency differs across the isoforms, it is challenging to quantify the total amount of bound and unbound ligand in a protein mixture.

Conclusions

Our investigation into the selectivity of eight known FABP ligands showed that profiles generated by nMS were consistent with the expected profiles generated by traditional solution-phase methods. This work therefore highlights the potential of nMS to serve as a valuable complementary method to traditional approaches for fragment-based drug design.4 While there is increasing literature evidence of the merits of nMS as a high-throughput and sensitive method for quantifying binding interactions, it is important to note that well-established techniques such as NMR, ITC and SPR remain compatible with such determinations, despite various associated limitations. Instead, our results indicate that the strength of nMS lies in multiplexed investigations of the ligand selectivity against multiple protein targets. Although affinity determination using this multiplexed approach is also possible, absolute quantification of protein–ligand complexes in protein mixtures was challenging.

For effective selectivity measurements, instrument parameters should be optimized for different protein mixtures, ideally using submicron emitters to minimize variation in the gas phase and improve resolution. Prior screening of protein–ligand mixtures should be conducted to ensure the masses of all species differ sufficiently (>5 Da) for accurate quantification and assignment. In-source dissociation of different proteins should also be considered as proteins more significantly affected by dissociation may produce false negatives. These considerations will be beneficial for enabling the identification of selective ligands by nMS to be progressed for further characterization and elaboration.

Acknowledgments

M.Q.P. was supported by an Australian Government Research Training Program scholarship. Technical support was provided by Dr Ghizal Siddiqui at the Drug Target Identification Node of the Monash Proteomics and Metabolomics Facility.

Glossary

ABBREVIATIONS

BCA

bicinchoninic acid

DMSO

dimethyl sulfoxide

FABP

fatty acid-binding protein

FP

fluorescence polarization

ITC

isothermal titration calorimetry

NMR

nuclear magnetic resonance

nMS

native mass spectrometry

PPARα

peroxisome proliferator-activated receptor α

SPR

surface plasmon resonance.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsmedchemlett.4c00154.

  • FABP isoform masses, additional ligand information and properties, additional nMS selectivity and ligand titration data, experimental details and procedures, NMR spectra and LCMS data for compounds 77 and 95 (PDF)

Author Present Address

The Walter and Eliza Hall Institute of Medical Research and Department of Medical Biology, University of Melbourne, Parkville, VIC 3052, Australia

Author Contributions

R.S.N., D.J.C. and M.J.S. conceived the project. M.Q.P. designed and performed all mass spectrometry experiments. I.R.C. conducted all biophysical characterization. N.A. recombinantly expressed all FABP isoforms, E.K. synthesized compounds 77 and 95. S.M.D. and B.C.D. advised on general experimental design. T.N. advised on the mass spectrometry and provided essential resources for native experiments. M.Q.P. wrote the manuscript with input from all authors.

The authors declare no competing financial interest.

Notes

Safety Statement: No unexpected or unusually high safety hazards were encountered when carrying out the reported work.

Supplementary Material

ml4c00154_si_001.pdf (1.3MB, pdf)

References

  1. Loo J. A. Studying noncovalent protein complexes by electrospray ionization mass spectrometry. Mass Spectrom. Rev. 1997, 16, 1–23. . [DOI] [PubMed] [Google Scholar]
  2. Hofstadler S. A.; Sannes-Lowery K. A. Applications of ESI-MS in Drug Discovery: Interrogation of Noncovalent Complexes. Nat. Rev. Drug. Discov. 2006, 5, 585–595. 10.1038/nrd2083. [DOI] [PubMed] [Google Scholar]
  3. Bennett J. L.; Nguyen G. T. H.; Donald W. A. Protein–Small Molecule Interactions in Native Mass Spectrometry. Chem. Rev. 2022, 122, 7327–7385. 10.1021/acs.chemrev.1c00293. [DOI] [PubMed] [Google Scholar]
  4. Doak B. C.; Norton R. S.; Scanlon M. J. The Ways and Means of Fragment-Based Drug Design. Pharmacol. Ther. 2016, 167, 28–37. 10.1016/j.pharmthera.2016.07.003. [DOI] [PubMed] [Google Scholar]
  5. Maple H. J.; Garlish R. A.; Rigau-Roca L.; Porter J.; Whitcombe I.; Prosser C. E.; Kennedy J.; Henry A. J.; Taylor R. J.; Crump M. P.; Crosby J. Automated Protein–Ligand Interaction Screening by Mass Spectrometry. J. Med. Chem. 2012, 55, 837–851. 10.1021/jm201347k. [DOI] [PubMed] [Google Scholar]
  6. Chrysanthopoulos P. K.; Mujumdar P.; Woods L. A.; Dolezal O.; Ren B.; Peat T. S.; Poulsen S.-A. Identification of a New Zinc Binding Chemotype by Fragment Screening. J. Med. Chem. 2017, 60, 7333–7349. 10.1021/acs.jmedchem.7b00606. [DOI] [PubMed] [Google Scholar]
  7. Gavriilidou A. F. M.; Holding F. P.; Coyle J. E.; Zenobi R. Application of Native ESI-MS to Characterize Interactions between Compounds Derived from Fragment-Based Discovery Campaigns and Two Pharmaceutically Relevant Proteins. SLAS Discov. 2018, 23, 951–959. 10.1177/2472555218775921. [DOI] [PubMed] [Google Scholar]
  8. Gavriilidou A. F. M.; Sokratous K.; Yen H.-Y.; De Colibus L. High-Throughput Native Mass Spectrometry Screening in Drug Discovery. Front. Mol. Biosci. 2022, 9, 837901. 10.3389/fmolb.2022.837901. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Vaaltyn M. C.; Mateos-Jimenez M.; Müller R.; Mackay C. L.; Edkins A. L.; Clarke D. J.; Veale C. G. L. Native Mass Spectrometry-Guided Screening Identifies Hit Fragments for HOP-HSP90 PPI Inhibition. ChemBioChem 2022, 23, e202200322. 10.1002/cbic.202200322. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Pedro L.; Quinn R. J. Native Mass Spectrometry in Fragment-Based Drug Discovery. Molecules 2016, 21, 984. 10.3390/molecules21080984. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Leney A. C.; Heck A. J. R. Native Mass Spectrometry: What is in the Name?. J. Am. Soc. Mass. Spectrom. 2017, 28, 5–13. 10.1007/s13361-016-1545-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. VanAernum Z. L.; Busch F.; Jones B. J.; Jia M.; Chen Z.; Boyken S. E.; Sahasrabuddhe A.; Baker D.; Wysocki V. H. Rapid Online Buffer Exchange for Screening of Proteins, Protein Complexes and Cell Lysates by Native Mass Spectrometry. Nat. Protoc. 2020, 15, 1132–1157. 10.1038/s41596-019-0281-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Maple H.; Scheibner O.; Baumert M.; Allen M.; Taylor R.; Garlish R.; Bromirski M.; Burnley R. Application of the Exactive Plus EMR for Automated Protein–Ligand Screening by Non-Covalent Mass Spectrometry. Rapid Commun. Mass. Spectrom. 2014, 28, 1561. 10.1002/rcm.6925. [DOI] [PubMed] [Google Scholar]
  14. Kitova E. N.; El-Hawiet A.; Klassen J. S. Screening Carbohydrate Libraries for Protein Interactions Using the Direct ESI-MS Assay. Applications to Libraries of Unknown Concentration. J. Am. Soc. Mass. Spectrom. 2014, 25, 1908–1916. 10.1007/s13361-014-0964-2. [DOI] [PubMed] [Google Scholar]
  15. Poulsen S.-A. Fragment Screening by Native State Mass Spectrometry. Aust. J. Chem. 2013, 66, 1495–1501. 10.1071/CH13190. [DOI] [Google Scholar]
  16. Nguyen G. T. H.; Tran T. N.; Podgorski M. N.; Bell S. G.; Supuran C. T.; Donald W. A. Nanoscale Ion Emitters in Native Mass Spectrometry for Measuring Ligand–Protein Binding Affinities. ACS Cent Sci 2019, 5, 308–318. 10.1021/acscentsci.8b00787. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Nguyen G. T. H.; Nocentini A.; Angeli A.; Gratteri P.; Supuran C. T.; Donald W. A. Perfluoroalkyl Substances of Significant Environmental Concern can Strongly Inhibit Human Carbonic Anhydrase Isozymes. Anal. Chem. 2020, 92, 4614–4622. 10.1021/acs.analchem.0c00163. [DOI] [PubMed] [Google Scholar]
  18. Nguyen G. T. H.; Bennett J. L.; Liu S.; Hancock S. E.; Winter D. L.; Glover D. J.; Donald W. A. Multiplexed Screening of Thousands of Natural Products for Protein–Ligand Binding in Native Mass Spectrometry. J. Am. Chem. Soc. 2021, 143, 21379–21387. 10.1021/jacs.1c10408. [DOI] [PubMed] [Google Scholar]
  19. Norton R. S.; Leung E. W. W.; Chandrashekaran I. R.; MacRaild C. A. Applications of 19F-NMR in Fragment-Based Drug Discovery. Molecules 2016, 21, 860. 10.3390/molecules21070860. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Gee C. T.; Arntson K. E.; Urick A. K.; Mishra N. K.; Hawk L. M. L.; Wisniewski A. J.; Pomerantz W. C. K. Protein-Observed 19F-NMR for Fragment Screening, Affinity Quantification and Druggability Assessment. Nat. Protoc. 2016, 11, 1414–1427. 10.1038/nprot.2016.079. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Liu M.; Van Voorhis W. C.; Quinn R. J. Development of a Target Identification Approach using Native Mass Spectrometry. Sci. Rep. 2021, 11, 2387. 10.1038/s41598-021-81859-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Furuhashi M.; Hotamisligil G. S. Fatty Acid-Binding Proteins: Role in Metabolic Diseases and Potential as Drug Targets. Nat. Rev. Drug. Discov. 2008, 7, 489–503. 10.1038/nrd2589. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Furuhashi M.; Tuncman G.; Görgün C.; Makowski L.; Atsumi G.; Vaillancourt E.; Kono K.; Babaev V.; Fazio S.; Linton M.; Sulsky R.; Robl J.; Parker R.; Hotamisligil G. Treatment of Diabetes and Atherosclerosis by Inhibiting Fatty-Acid-Binding Protein aP2. Nature 2007, 447, 959–65. 10.1038/nature05844. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Chmurzynska A. The Multigene Family of Fatty Acid-Binding Proteins (FABPs): Function, Structure and Polymorphism. J. Appl. Genet. 2006, 47, 39–48. 10.1007/BF03194597. [DOI] [PubMed] [Google Scholar]
  25. Thompson J.; Ory J.; Reese-Wagoner A.; Banaszak L. The Liver Fatty Acid Binding Protein - Comparison of Cavity Properties of Intracellular Lipid-Binding Proteins. Mol. Cell. Biochem. 1999, 192, 9–16. 10.1023/A:1006806616963. [DOI] [PubMed] [Google Scholar]
  26. Odani S.; Namba Y.; Ishii A.; Ono T.; Fujii H. Disulfide Bonds in Rat Cutaneous Fatty Acid-Binding Protein. J. Biochem. 2000, 128, 355–361. 10.1093/oxfordjournals.jbchem.a022761. [DOI] [PubMed] [Google Scholar]
  27. Patil R.; Mohanty B.; Liu B.; Chandrashekaran I. R.; Headey S. J.; Williams M. L.; Clements C. S.; Ilyichova O.; Doak B. C.; Genissel P.; Weaver R. J.; Vuillard L.; Halls M. L.; Porter C. J. H.; Scanlon M. J. A Ligand-Induced Structural Change in Fatty Acid-Binding Protein 1 is Associated with Potentiation of Peroxisome Proliferator-Activated Receptor α Agonists. J. Biol. Chem. 2019, 294, 3720–3734. 10.1074/jbc.RA118.006848. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Floresta G.; Patamia V.; Zagni C.; Rescifina A. Adipocyte fatty acid binding protein 4 (FABP4) inhibitors. An update from 2017 to early 2022. Eur. J. Med. Chem. 2022, 240 (240), 114604. 10.1016/j.ejmech.2022.114604. [DOI] [PubMed] [Google Scholar]
  29. Benkestock K.; Van Pelt C. K.; Åkerud T.; Sterling A.; Edlund P.-O.; Roeraade J. Automated Nano-Electrospray Mass Spectrometry for Protein-Ligand Screening by Noncovalent Interaction Applied to Human H-FABP and A-FABP. J. Biomol. Screen. 2003, 8, 247–256. 10.1177/1087057103008003002. [DOI] [PubMed] [Google Scholar]
  30. Root K.; Wittwer Y.; Barylyuk K.; Anders U.; Zenobi R. Insight into Signal Response of Protein Ions in Native ESI-MS from the Analysis of Model Mixtures of Covalently Linked Protein Oligomers. J. Am. Soc. Mass. Spectrom. 2017, 28, 1863–1875. 10.1007/s13361-017-1690-3. [DOI] [PubMed] [Google Scholar]
  31. Yang R.; Wang P.; Chen Z.; Hu W.; Gong Y.; Zhang W.; Huang C. WY-14643, a Selective Agonist of Peroxisome Proliferator-Activated Receptor-α, Ameliorates Lipopolysaccharide-Induced Depressive-Like Behaviors by Preventing Neuroinflammation and Oxido-Nitrosative Stress in Mice. Pharmacol., Biochem. Behav. 2017, 153, 97–104. 10.1016/j.pbb.2016.12.010. [DOI] [PubMed] [Google Scholar]
  32. Buettelmann B.; Ceccarelli S. M.; Kuehne H.; Kuhn B.; Neidhart W.; Obst Sander U.; Richter H.. New Bicyclic Thiophenylamide Compounds. Patent No. WO2013189841A1, December 27, 2013.
  33. Tamara S.; den Boer M. A.; Heck A. J. R. High-Resolution Native Mass Spectrometry. Chem. Rev. 2022, 122, 7269–7326. 10.1021/acs.chemrev.1c00212. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Robinson C. V.; Chung E. W.; Kragelund B. B.; Knudsen J.; Aplin R. T.; Poulsen F. M.; Dobson C. M. Probing the Nature of Noncovalent Interactions by Mass Spectrometry. A Study of Protein–CoA Ligand Binding and Assembly. J. Am. Chem. Soc. 1996, 118, 8646–8653. 10.1021/ja960211x. [DOI] [Google Scholar]
  35. Bich C.; Baer S.; Jecklin M. C.; Zenobi R. Probing the Hydrophobic Effect of Noncovalent Complexes by Mass Spectrometry. J. Am. Soc. Mass. Spectrom. 2010, 21, 286–289. 10.1016/j.jasms.2009.10.012. [DOI] [PubMed] [Google Scholar]
  36. Báez Bolivar E. G.; Bui D. T.; Kitova E. N.; Han L.; Zheng R. B.; Luber E. J.; Sayed S. Y.; Mahal L. K.; Klassen J. S. Submicron Emitters Enable Reliable Quantification of Weak Protein–Glycan Interactions by ESI-MS. Anal. Chem. 2021, 93, 4231–4239. 10.1021/acs.analchem.0c05003. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ml4c00154_si_001.pdf (1.3MB, pdf)

Articles from ACS Medicinal Chemistry Letters are provided here courtesy of American Chemical Society

RESOURCES