Next Article in Journal
Current Treatment Methods for Charcot–Marie–Tooth Diseases
Previous Article in Journal
Attenuation of PM2.5-Induced Lung Injury by 4-Phenylbutyric Acid: Maintenance of [Ca2+]i Stability between Endoplasmic Reticulum and Mitochondria
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Biomolecules
Volume 14
Issue 9
10.3390/biom14091136
biomolecules-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Application of CoLD-CoP to Detecting Competitively and Cooperatively Binding Ligands

by
Shiva V. Patnala
,
Roberto Robles
and
David A. Snyder
*
Department of Chemistry, College of Science and Health, William Paterson University, 300 Pompton Road, Wayne, NJ 07470, USA
*
Author to whom correspondence should be addressed.
Biomolecules 2024, 14(9), 1136; https://doi.org/10.3390/biom14091136
Submission received: 30 June 2024 / Revised: 21 August 2024 / Accepted: 23 August 2024 / Published: 9 September 2024
Browse Figures
Review Reports Versions Notes

Abstract

:
NMR utilization in fragment-based drug discovery requires techniques to detect weakly binding fragments and to subsequently identify cooperatively binding fragments. Such cooperatively binding fragments can then be optimized or linked in order to develop viable drug candidates. Similarly, ligands or substrates that bind macromolecules (including enzymes) in competition with the endogenous ligand or substrate are valuable probes of macromolecular chemistry and function. The lengthy and costly process of identifying competitive or cooperative binding can be streamlined by coupling computational biochemistry and spectroscopy tools. The Clustering of Ligand Diffusion Coefficient Pairs (CoLD-CoP) method, previously developed by Snyder and co-workers, detects weakly binding ligands by analyzing pairs of diffusion spectra, obtained in the absence and the presence of a protein. We extended the CoLD-CoP method to analyze spectra pairs (each in the presence of a protein) with or without a critical ligand, to detect both competitive and cooperative binding.

1. Introduction

For at least two decades, a “target-rich”, “lead-poor” pipeline has characterized modern drug development [1]. Over the last 20 years, the amount by which pharmaceutical companies reallocate their net revenues into R&D has increased by about 20%, and by 2021, pharmaceutical companies collectively spent a third of their total budget on preclinical phase R&D [2]. Fragment-Based Drug Discovery (FBDD) addresses the lead generation problem by combinatorial assembly of (perhaps weakly) binding fragments [3]: the FBDD process identifies multiple, small “fragment” molecules, which are then assembled, like pieces of a jigsaw puzzle, and optimized to increase target hit affinity up to 100–1000-fold [4,5].
A key step in FBDD is, therefore, figuring out which (weakly) binding fragments are linkable, i.e., bind cooperatively so that linking those fragments would likely significantly enhance binding affinity [5]. On the other hand, a compound (full size or fragment) that binds competitively with the endogenous substrate or ligand of a given enzyme or receptor is a good lead for developing inhibitors of that enzyme or receptor [6]. Furthermore, competitive inhibitors are useful probes for better understanding enzyme active sites, chemistry, and even enzyme flexibility and dynamics [7].
Nuclear Magnetic Resonance (NMR) is a powerful tool with many applications in drug development, including FBDD pipelines [4,8]. NMR-based techniques, such as Diffusion Ordered Spectroscopy (DOSY), which identifies ligands by their change in diffusion coefficient upon target binding, can identify ligands with KD values ranging from 100 pM to 10 mM [9,10,11]. However, promising fragment molecules may very well bind with a KD up to 30 mM [12], and experiments to detect fragment binding typically require protein concentrations of one-thousandth or, depending on the type of experiment performed, up to one-tenth of the ligand concentration [13]. Challenges in the use of DOSY to identify ligand binding include the large uncertainty in diffusion coefficients inferred from DOSY data [14,15] as well as the inherent low sensitivity of NMR experiments. Ensuring enough protein is available to form protein–ligand complexes at concentrations sufficient to identify even the weakest-binding promising fragments typically requires around 10 mM ligand concentration and hence 10 μM protein concentration.
This paper demonstrates the application of the Clustering of Ligand Diffusion Coefficient Pairs (CoLD-CoP) method [16], which uses statistical techniques to overcome the large uncertainty in diffusion coefficients inferred from DOSY data, to identify competitively and cooperatively binding ligands. We demonstrate competitive binding on a well-studied system: 4-hydroxy-α-cinnamic acid (4-HCCA) and salicylic acid are each known to competitively inhibit (mushroom) tyrosinase [17,18] and hence are expected to compete with each other in binding that enzyme. On the other hand, lysozyme, a well-studied enzyme catalyzing the hydrolysis of a polymeric substance, has a large active site that can accommodate a macromolecular substrate. Hence, it would not be surprising to discover competitive inhibitors of that enzyme which cooperatively bind it. Imidazole [19,20], tris [21,22], and N-acetyl-glucosamine (GlcNAc) [23] are known competitive inhibitors of lysozyme. In particular, the substrate of lysozyme, peptidoglycan, is a co-polymer of GlcNAc and N-acetylmuramic acid [24]. We demonstrate that we can detect both competitive and cooperative binding—the former at a protein concentration of <5 μM and the latter being an expected result validated with Molecular Mechanics Generalized Born Surface Area (MM-GBSA) [25] calculations—using CoLD-CoP.

2. Materials and Methods

Scheme 1 outlines the steps of our approach. Briefly, our approach involves making three different mixtures: one containing only the putative ligands (the ligand-only mixture: Step 1.A), one with putative ligands and a protein of interest (the ligand+protein mixture: Step I.B), and one with the putative ligands, a protein of interest, and a known ligand binding the protein of interest (the ligand+protein+known ligand mixture: Step I.C). Following the acquisition and processing of DOSY data for the ligand-only, ligand+protein, and ligand+protein+known ligand mixtures (Step 2), the DOSY peak lists are compared using CoLD-CoP [16].
As described in Scheme 1, substituting the ligand+protein+known ligand peak list for the ligand-only peak list typically used in running CoLD-CoP (c.f. Step 3.B.i) identifies ligands diffusing more slowly in the absence of the known ligand (Step 3.B.ii), whereas substituting the ligand+protein peak list for the ligand-only peak list and the ligand+protein+known ligand peak list for the ligand+protein peak list identifies ligands diffusing more slowly in the presence of the known ligand (Step 3.B.iii). As illustrated in Scheme 2, ligands that diffuse more slowly in the absence of the known ligand bind the protein of interest more effectively without the known ligand and hence compete with it for protein binding. Conversely, ligands that diffuse more slowly in the presence of a known ligand bind the protein of interest better with the known ligand, indicating cooperative binding.
Table 1 describes the components of the ligand-only, ligand+protein, and ligand+protein+known ligand mixtures used to validate the approach described in Scheme 1. Briefly, solution TL is the ligand mixture used for probing competitive/cooperative binding of mushroom tyrosinase, with TLP being the corresponding ligand+protein mixture and TLPK being the corresponding ligand+protein+known ligand mixture. TLP was prepared by adding mushroom tyrosinase (to a concentration of 4 μM) to a portion of mixture TL; in the process of preparing mixture TLP, some salicylic acid precipitated, resulting in a concentration of only 1 mM in mixtures TLP and TLPK as opposed to the 4 mM concentration in mixture TL. TLPK was prepared by mixing a portion of mixture TLP with 4-HCCA.
Similarly, the ligand-only, ligand+protein, and ligand+protein+known ligand mixtures for probing competitive/cooperative lysozyme binding are labeled LL, LLP, and LLPK. LLP was prepared by adding lysozyme (to a concentration of 1 mM) to a portion of mixture LL, and LLPK was prepared by adding GlcNAc to mixture LLP. The solvent for all solutions was D2O. Solutions LL, LLP, and LLPK were adjusted to pH = 6 using NaOH. The pH of solutions TL, TLP, and TLPK was not adjusted.
Data acquisition was performed on a 400 MHz Bruker AvanceTM III using the ledbpgp2s pulse sequence with a diffusion time of 90 ms and a gradient pulse length of 1.9 ms. Initial processing (in the direct dimension) utilized NMRPipe [26], and DOSY processing proceeded using the DOSY Toolbox [27] along with the Covariance Toolbox [28], resulting in the DOSY spectra shown in Figure 1. Both DOSY processing and CoLD-CoP analysis were performed in MATLAB (various versions used including R2015b, R2022a, R2022b, R2023a, R2023b) [29]. iGEMDOCK [30] docked tris, imidazole, and GlcNAc to lysozyme (PDB ID 2D4K15). The ternary complex (tris-GlcNAc-lysozyme) was manually adjusted using the Schrödinger [31] interface. Complexes were energy minimized in DESMOND [32] using the Schrödinger interface and default minimization parameters. LigPlot+ [33,34] analyzed and illustrated ligand/protein interfaces. Structures were rendered in MacPYMOL [35]. MM-GBSA and MM-PBSA calculations were performed using the FastDRH webserver [36].

3. Results

3.1. Detecting Competitive Binding to Tyrosinase

Using the approach described in Scheme 1, CoLD-CoP identifies two clusters of diffusion coefficient pairs as diffusing significantly slower in the absence of 4-HCCA than in its presence, indicating competitive binding (Figure 2). Based on the resonance assignments for mixture TL (Table S1; Tables S2 and S3 provide resonance assignments for mixtures TLP and TLPK, respectively), both of these clusters are assigned to salicylic acid. Thus, CoLD-CoP correctly identifies that 4-HCCA and salicylic acid competitively bind tyrosinase. Note, however, that due to their similar molecular masses and hence diffusion coefficients, the clustering routines in CoLD-CoP cannot distinguish between tris and tartaric acid; salicylate (which also has a similar molecular mass and hence diffusion coefficient) is distinguished (albeit split into two clusters) due to its competitive binding changing its diffusion coefficient in the presence of 4-HCCA.

3.2. Detecting Cooperative Binding to Lysozyme

CoLD-CoP identifies that two components of the ligand mixture LL with diffusion coefficients and chemical shifts corresponding to tris and imidazole diffuse slower in the presence of GlcNAc than in its absence (Figure 3B,C), indicating that tris and imidazole each bind cooperatively with GlcNAc. As shown in Figure 3A, even though tris and imidazole are each known competitive inhibitors of lysozyme, they do not compete with GlcNAc for binding lysozyme.

3.3. Validation of Cooperative Binding

The cooperative binding of tris and GlcNAc, though unexpected, is not surprising given that lysozyme’s active site is large enough to accommodate a polymeric substrate. Modeling the ternary complex (coordinates in PDB format available as supporting information) of lysozyme, tris, and GlcNAc (Figure 4 and Figure 5) illustrates that in this complex, tris and GlcNAc hydrogen bond, providing a potential mechanism for the cooperative binding observed in our CoLD-CoP-based assay. MM-GBSA calculations further validate cooperative binding; note that, at least for parameter sets correctly identifying GlcNAc as a better binder than tris, the binding energy for GlcNAc and tris together is more negative than the sum of the binding energies for each (Table 2).
The dimerization of lysozyme [20,37] explains the cooperative binding of imidazole (which can bind lysozyme allosterically [20]) and GlcNAc: each bind to monomeric lysozyme, and GlcNAc binding facilitates imidazole binding by shifting lysozyme’s dimerization equilibrium toward the monomer (Scheme 3).

4. Discussion

A putative ligand identified as binding competitively with a lead fragment (in this example, 4-HCCA and GlcNAc) would be a poor choice to assemble with the lead fragment in the context of FBDD. However, the identification of competitive binders, which may themselves be agonists, to endogenous ligands/substrates provides protein chemists with probes of protein structure and function [38]. Specifically, the competitive binding of 4-HCCA and salicylic acid to Tyrosinase validates CoLD-CoP’s novel application to identifying competitive binders, showcasing the method’s potential to provide insights into enzyme activity and protein-ligand interactions, thereby enhancing the understanding and development of therapeutic agents.
In the context of FBDD, the discovery that tris and imidazole bind cooperatively with GlcNAc is tantamount to identifying synergistically acting fragments. Linkage of these fragments may generate a viable lead compound for further chemical optimization. In a system with a known protein structure, computational and theoretical modeling can guide the assembly of fragments, identified as cooperative binders via CoLD-CoP, into a lead compound. In the absence of a protein structure, differences in chemical shifts between the pair of DOSY spectra input into CoLD-CoP can shed light on how the ligands identified by CoLD-CoP actually bind a target protein.

5. Conclusions

The application of CoLD-CoP to identify multiple cooperative and competitive binding interactions with a single pair of experiments, with protein concentrations as low as 4 μM and ligand concentrations as low as 1 mM, facilitates follow-up studies in a FBDD process. This study demonstrates the ability of CoLD-CoP to identify cooperative binders by demonstrating the cooperative binding of GlcNAc and tris to lysozyme. The CoLD-CoP method is particularly suited to FBDD as CoLD-CoP-based analyses can utilize spectra acquired at lower field strengths and on samples with minimal amounts of protein. The methods utilized in this paper exemplify the use of CoLD-CoP in identifying fragments that cooperatively bind a given target to assist in the process of piecing together a promising lead compound based on the results of fragment screening. In general, the application of CoLD-CoP to identify cooperative and competitive binding ligands will facilitate the discovery of compounds modifying protein function, which is critical in both drug development and chemical biology.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/biom14091136/s1, Table S1: Assigned Peaks in DOSY Spectrum of Ligand Mixture (with neither tyrosinase nor α-cyano-4-hydroxycinnamic acid present). Table S2: Assigned Peaks in DOSY Spectrum of Ligand+Protein Mixture (with tyrosinase but not α-cyano-4-hydroxycinnamic acid present). Table S3: Assigned Peaks in DOSY Spectrum of Ligand+Protein Mixture (with both tyrosinase and α-cyano-4-hydroxycinnamic acid present). Also included is a list of abbreviations used in the tables.

Author Contributions

Conceptualization, S.V.P. and D.A.S.; methodology, S.V.P. and D.A.S.; software, D.A.S.; validation, S.V.P., R.R. and D.A.S.; formal analysis, S.V.P., R.R., and D.A.S.; investigation, S.V.P., R.R., and D.A.S.; resources, D.A.S.; data curation, D.A.S.; writing—original draft preparation, S.V.P. and D.A.S.; writing—review and editing, S.V.P., R.R., and D.A.S.; visualization, S.V.P. and D.A.S.; supervision, D.A.S.; project administration, D.A.S. All authors have read and agreed to the published version of the manuscript.

Funding

The research presented in this paper was facilitated by the award of Assigned Release Time (ART) for research to D.A.S. from the William Paterson University Office of the Provost. This research received no external funding.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The original data presented in the study, along with scripts used in the analysis presented herein, are openly available in Zenodo at DOI: 10.5281/zenodo.12170717.

Acknowledgments

The authors thank Manar Doughouz, Christina Kohn, Melissa Synsmir, and Rebecca Vera for many stimulating discussions concerning applications of the CoLD-CoP method and the Department of Chemistry for materials used in this project.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Rees, D.C.; Congreve, M.; Murray, C.W.; Carr, R. Fragment-Based Lead Discovery. Nat. Rev. Drug Discov. 2004, 3, 660–672. [Google Scholar] [CrossRef] [PubMed]
  2. Austin, D.; Hayford, T. Research and Development in the Pharmaceutical Industry; Congressional Budget Office: Washington, DC, USA, 2021. [Google Scholar]
  3. Lamoree, B.; Hubbard, R.E. Current Perspectives in Fragment-Based Lead Discovery (FBLD). Essays Biochem. 2017, 61, 453–464. [Google Scholar]
  4. Emwas, A.-H.; Szczepski, K.; Poulson, B.G.; Chandra, K.; McKay, R.T.; Dhahri, M.; Alahmari, F.; Jaremko, L.; Lachowicz, J.I.; Jaremko, M. NMR as a “Gold Standard” Method in Drug Design and Discovery. Molecules 2020, 25, 4597. [Google Scholar] [CrossRef] [PubMed]
  5. Kirsch, P.; Hartman, A.M.; Hirsch, A.K.; Empting, M. Concepts and Core Principles of Fragment-Based Drug Design. Molecules 2019, 24, 4309. [Google Scholar] [CrossRef]
  6. Ouertani, A.; Neifar, M.; Ouertani, R.; Masmoudi, A.S.; Mosbah, A.; Cherif, A. Effectiveness of Enzyme Inhibitors in Biomedicine and Pharmacotherapy. Adv. Tissue Eng. Regen. Med. Open Access 2019, 5, 85–90. [Google Scholar] [CrossRef]
  7. Anderson, J.W.; Vaisar, D.; Jones, D.N.; Pegram, L.M.; Vigers, G.P.; Chen, H.; Moffat, J.G.; Ahn, N.G. Conformation Selection by ATP-Competitive Inhibitors and Allosteric Communication in ERK2. eLife 2024, 12, RP91507. [Google Scholar] [CrossRef]
  8. Kerber, P.J.; Nuñez, R.; Jensen, D.R.; Zhou, A.L.; Peterson, F.C.; Hill, R.B.; Volkman, B.F.; Smith, B.C. Chapter Nine-Fragment-Based Screening by Protein-Detected NMR Spectroscopy. In Methods in Enzymology; Lloyd, M., Ed.; Academic Press: Cambridge, MA, USA, 2023; Volume 690, pp. 285–310. ISBN 0076-6879. [Google Scholar]
  9. Fielding, L. NMR Methods for the Determination of Protein–Ligand Dissociation Constants. Prog. Nucl. Magn. Reson. Spectrosc. 2007, 51, 219–242. [Google Scholar] [CrossRef]
  10. Jayalakshmi, V.; Krishna, N.R. CORCEMA Refinement of the Bound Ligand Conformation within the Protein Binding Pocket in Reversibly Forming Weak Complexes Using STD-NMR Intensities. J. Magn. Reson. 2004, 168, 36–45. [Google Scholar] [CrossRef]
  11. Meyer, B.; Peters, T. NMR Spectroscopy Techniques for Screening and Identifying Ligand Binding to Protein Receptors. Angew. Chem. Int. Ed. 2003, 42, 864–890. [Google Scholar] [CrossRef]
  12. Carr, R.A.; Congreve, M.; Murray, C.W.; Rees, D.C. Fragment-Based Lead Discovery: Leads by Design. Drug Discov. Today 2005, 10, 987–992. [Google Scholar] [CrossRef]
  13. Di Carluccio, C.; Forgione, M.C.; Martini, S.; Berti, F.; Molinaro, A.; Marchetti, R.; Silipo, A. Investigation of Protein-Ligand Complexes by Ligand-Based NMR Methods. Carbohydr. Res. 2021, 503, 108313. [Google Scholar] [CrossRef] [PubMed]
  14. Delsuc, M.; Malliavin, T. Maximum Entropy Processing of DOSY NMR Spectra. Anal. Chem. 1998, 70, 2146–2148. [Google Scholar] [CrossRef]
  15. Guest, J.; Kiraly, P.; Nilsson, M.; Morris, G.A. Signal-to-Noise Ratio in Diffusion-Ordered Spectroscopy: How Good Is Good Enough? Magn Reson 2021, 2, 733–739. [Google Scholar] [CrossRef]
  16. Snyder, D.A.; Chantova, M.; Chaudhry, S. Analysis of Ligand–Protein Exchange by Clustering of Ligand Diffusion Coefficient Pairs (CoLD-CoP). J. Magn. Reson. 2015, 255, 44–50. [Google Scholar] [CrossRef]
  17. Qiu, L.; Chen, Q.-H.; Zhuang, J.-X.; Zhong, X.; Zhou, J.-J.; Guo, Y.-J.; Chen, Q.-X. Inhibitory Effects of α-Cyano-4-Hydroxycinnamic Acid on the Activity of Mushroom Tyrosinase. Food Chem. 2009, 112, 609–613. [Google Scholar] [CrossRef]
  18. Zhang, J.-P.; Chen, Q.-X.; Song, K.-K.; Xie, J.-J. Inhibitory Effects of Salicylic Acid Family Compounds on the Diphenolase Activity of Mushroom Tyrosinase. Food Chem. 2006, 95, 579–584. [Google Scholar] [CrossRef]
  19. Swan, I.D. The Inhibition of Hen Egg-White Lysozyme by Imidazole and Indole Derivatives. J. Mol. Biol. 1972, 65, 59–62. [Google Scholar] [CrossRef] [PubMed]
  20. Hampe, O.G.; Tondo, C.V.; Hasson-Voloch, A. A Biophysical Model of Lysozyme Self-Association. Biophys. J. 1982, 40, 77–82. [Google Scholar] [CrossRef]
  21. Kang, T.; Hao, W.; Niu, Y.; Luo, Z.; Jin, G. Kinetic Analysis of the Weak Affinity Interaction between Tris and Lysozyme. Biochem. Biophys. Res. Commun. 2015, 457, 659–663. [Google Scholar] [CrossRef]
  22. Quan, L.; Wei, D.; Jiang, X.; Liu, Y.; Li, Z.; Li, N.; Li, K.; Liu, F.; Lai, L. Resurveying the Tris Buffer Solution: The Specific Interaction between Tris (Hydroxymethyl) Aminomethane and Lysozyme. Anal. Biochem. 2008, 378, 144–150. [Google Scholar] [CrossRef]
  23. Von Dreele, R. Binding of N-Acetylglucosamine to Chicken Egg Lysozyme: A Powder Diffraction Study. Acta Crystallogr. Sect. D Biol. Crystallogr. 2001, 57, 1836–1842. [Google Scholar] [CrossRef] [PubMed]
  24. Manchenko, G.P. Handbook of Detection of Enzymes on Electrophoretic Gels; CRC Press: Boca Raton, FL, USA, 2002; ISBN 0-429-12800-2. [Google Scholar]
  25. Tuccinardi, T. What Is the Current Value of MM/PBSA and MM/GBSA Methods in Drug Discovery? Expert Opin. Drug Discov. 2021, 16, 1233–1237. [Google Scholar] [CrossRef] [PubMed]
  26. Delaglio, F.; Grzesiek, S.; Vuister, G.W.; Zhu, G.; Pfeifer, J.; Bax, A. NMRPipe: A Multidimensional Spectral Processing System Based on UNIX Pipes. J. Biomol. NMR 1995, 6, 277–293. [Google Scholar] [CrossRef] [PubMed]
  27. Nilsson, M. The DOSY Toolbox: A New Tool for Processing PFG NMR Diffusion Data. J. Magn. Reson. 2009, 200, 296–302. [Google Scholar] [CrossRef] [PubMed]
  28. Short, T.; Alzapiedi, L.; Brüschweiler, R.; Snyder, D. A Covariance NMR Toolbox for MATLAB and OCTAVE. J. Magn. Reson. 2011, 209, 75–78. [Google Scholar] [CrossRef]
  29. The MathWorks. Matlab; The MathWorks: Natick, MA, USA, 2012. [Google Scholar]
  30. Hsu, K.-C.; Chen, Y.-F.; Lin, S.-R.; Yang, J.-M. iGEMDOCK: A Graphical Environment of Enhancing GEMDOCK Using Pharmacological Interactions and Post-Screening Analysis. BMC Bioinform. 2011, 12, S33. [Google Scholar] [CrossRef]
  31. Schrodinger, L. Schrodinger Software Suite; Schrödinger, LLC.: New York, NY, USA, 2011; Volume 670. [Google Scholar]
  32. Bowers, K.J.; Chow, E.; Xu, H.; Dror, R.O.; Eastwood, M.P.; Gregersen, B.A.; Klepeis, J.L.; Kolossvary, I.; Moraes, M.A.; Sacerdoti, F.D. Scalable Algorithms for Molecular Dynamics Simulations on Commodity Clusters. In SC ‘06: Proceedings of the 2006 ACM/IEEE conference on Supercomputing; Association for Computing Machinery: New York, NY, USA, 2006; p. 84-es. [Google Scholar]
  33. Laskowski, R.A.; Swindells, M.B. LigPlot+: Multiple Ligand–Protein Interaction Diagrams for Drug Discovery. J. Chem. Inf. Model. 2011, 51, 2778–2786. [Google Scholar] [CrossRef]
  34. Wallace, A.C.; Laskowski, R.A.; Thornton, J.M. LIGPLOT: A Program to Generate Schematic Diagrams of Protein-Ligand Interactions. Protein Eng. Des. Sel. 1995, 8, 127–134. [Google Scholar] [CrossRef]
  35. DeLano, W. MacPyMOL: A PyMOL-Based Molecular Graphics Application for MacOS X; DeLano Scientific LLC.: Palo Alto, CA, USA, 2007. [Google Scholar]
  36. Wang, Z.; Pan, H.; Sun, H.; Kang, Y.; Liu, H.; Cao, D.; Hou, T. fastDRH: A Webserver to Predict and Analyze Protein–Ligand Complexes Based on Molecular Docking and MM/PB (GB) SA Computation. Brief. Bioinform. 2022, 23, bbac201. [Google Scholar] [CrossRef]
  37. Onuma, K.; Inaka, K. Lysozyme Dimer Association: Similarities and Differences Compared with Lysozyme Monomer Association. J. Cryst. Growth 2008, 310, 1174–1181. [Google Scholar] [CrossRef]
  38. Klein Herenbrink, C.; Sykes, D.A.; Donthamsetti, P.; Canals, M.; Coudrat, T.; Shonberg, J.; Scammells, P.J.; Capuano, B.; Sexton, P.M.; Charlton, S.J. The Role of Kinetic Context in Apparent Biased Agonism at GPCRs. Nat. Commun. 2016, 7, 10842. [Google Scholar] [CrossRef] [PubMed]
Biomolecules 14 01136 sch001
Scheme 1. Detection of cooperative and competitive binding using CoLD-CoP.
Scheme 1. Detection of cooperative and competitive binding using CoLD-CoP.
Biomolecules 14 01136 sch001
Biomolecules 14 01136 sch002
Scheme 2. Use of CoLD-CoP to detect competitive and cooperative binding. Interactions identified by CoLD-CoP map to the yellow-highlighted portion of the plots. (A) Traditional CoLD-CoP (Scheme 1, Step #3.B.i): Diffusion coefficients measured in the presence and absence of protein (ligand+protein vs. ligand-only). Ligand–protein binding is identified by slower diffusion in the presence of the protein. (B) Competitive binding (Scheme 1, Step #3.B.ii): diffusion coefficients measured in the absence and presence of a known ligand (ligand+protein vs. ligand+protein+known ligand). Competitive binding is identified by slower diffusion in the absence of the known ligand. (C) Cooperative binding (Scheme 1, Step #3.B.iii): diffusion coefficients measured in the presence and absence of a known ligand (ligand+protein+known ligand vs. ligand+protein). Cooperative binding is identified by slower diffusion in the presence of the known ligand.
Scheme 2. Use of CoLD-CoP to detect competitive and cooperative binding. Interactions identified by CoLD-CoP map to the yellow-highlighted portion of the plots. (A) Traditional CoLD-CoP (Scheme 1, Step #3.B.i): Diffusion coefficients measured in the presence and absence of protein (ligand+protein vs. ligand-only). Ligand–protein binding is identified by slower diffusion in the presence of the protein. (B) Competitive binding (Scheme 1, Step #3.B.ii): diffusion coefficients measured in the absence and presence of a known ligand (ligand+protein vs. ligand+protein+known ligand). Competitive binding is identified by slower diffusion in the absence of the known ligand. (C) Cooperative binding (Scheme 1, Step #3.B.iii): diffusion coefficients measured in the presence and absence of a known ligand (ligand+protein+known ligand vs. ligand+protein). Cooperative binding is identified by slower diffusion in the presence of the known ligand.
Biomolecules 14 01136 sch002
Biomolecules 14 01136 g001
Figure 1. DOSY spectra for mixtures (as labeled in Table 1) (A) TL, (B) TLP, (C) TLPK, (D) LL, (E) LLP, and (F) LLPK, obtained using the tyrosinase (A–C) and lysozyme systems (D–F). Vertical arrows of color-coordinated (to match the color scheme used in Figure 2 and Figure 3) compounds show peak correspondences between panels. Horizontal line styles correspond to diffusion coefficients in related mixtures. Solid lines indicate the average (by eye) diffusion coefficient for a given compound in the ligand+only (A,D), ligand+protein (B,E), and ligand+protein+known ligand (C,F) mixtures. The dashed lines indicate the comparable diffusion coefficients in the ligand+only (B,E) and ligand+protein (C,F) mixtures. The blue solid line in panel (A) and blue dashed line in panel (B) indicate the diffusion coefficient of salicylic acid in the absence of protein. The red solid line in panel (B) and red dashed line in panel (C) indicate the diffusion coefficient of salicylic acid in the presence of tyrosinase (but without 4-HCCA present). The muted red solid line in panel (C) indicates the diffusion coefficient of salicylic acid in the presence of both tyrosinase and 4-HCCA. Note that the muted red line is below the dashed red line in panel (C), indicating faster diffusion in the presence of 4-HCCA than in its absence, and hence that salicylic acid and 4-HCCA bind tyrosinase competitively. By way of comparison, the diffusion coefficients of tris and tartaric acid, indicated by the cyan annotations, do not change much between panels (B,C), as indicated by the proximity of the dashed cyan line and the muted cyan line in panel (C). The proximity of the yellow dashed line and yellow-orange solid line in panel (E) indicates the extremely weak binding of imidazole in the absence of GlcNAc. The larger gap between the solid and dashed lines in panel (F) indicates the cooperative binding of GlcNAc and imidazole to lysozyme. The near coincidence of the dashed and solid green lines in panel (E) indicates that tris binds lysozyme so weakly that its binding cannot be detected even by CoLD-CoP. However, the gap between the solid green line and the dashed mint green line in panel (F) indicates that GlcNAc and tris cooperatively bind lysozyme.
Figure 1. DOSY spectra for mixtures (as labeled in Table 1) (A) TL, (B) TLP, (C) TLPK, (D) LL, (E) LLP, and (F) LLPK, obtained using the tyrosinase (A–C) and lysozyme systems (D–F). Vertical arrows of color-coordinated (to match the color scheme used in Figure 2 and Figure 3) compounds show peak correspondences between panels. Horizontal line styles correspond to diffusion coefficients in related mixtures. Solid lines indicate the average (by eye) diffusion coefficient for a given compound in the ligand+only (A,D), ligand+protein (B,E), and ligand+protein+known ligand (C,F) mixtures. The dashed lines indicate the comparable diffusion coefficients in the ligand+only (B,E) and ligand+protein (C,F) mixtures. The blue solid line in panel (A) and blue dashed line in panel (B) indicate the diffusion coefficient of salicylic acid in the absence of protein. The red solid line in panel (B) and red dashed line in panel (C) indicate the diffusion coefficient of salicylic acid in the presence of tyrosinase (but without 4-HCCA present). The muted red solid line in panel (C) indicates the diffusion coefficient of salicylic acid in the presence of both tyrosinase and 4-HCCA. Note that the muted red line is below the dashed red line in panel (C), indicating faster diffusion in the presence of 4-HCCA than in its absence, and hence that salicylic acid and 4-HCCA bind tyrosinase competitively. By way of comparison, the diffusion coefficients of tris and tartaric acid, indicated by the cyan annotations, do not change much between panels (B,C), as indicated by the proximity of the dashed cyan line and the muted cyan line in panel (C). The proximity of the yellow dashed line and yellow-orange solid line in panel (E) indicates the extremely weak binding of imidazole in the absence of GlcNAc. The larger gap between the solid and dashed lines in panel (F) indicates the cooperative binding of GlcNAc and imidazole to lysozyme. The near coincidence of the dashed and solid green lines in panel (E) indicates that tris binds lysozyme so weakly that its binding cannot be detected even by CoLD-CoP. However, the gap between the solid green line and the dashed mint green line in panel (F) indicates that GlcNAc and tris cooperatively bind lysozyme.
Biomolecules 14 01136 g001
Biomolecules 14 01136 g002
Figure 2. CoLD-CoP comparisons: circles and asterisks are colored (automatically by the MATLAB toolbox implementing CoLD-CoP) based on the cluster to which CoLD-CoP assigns them. Ideally, each cluster represents one and only one molecule, but molecules of similar size whose diffusion coefficient does not change between the solutions being compared may not be distinguishable based on paired diffusion coefficients. Open circles indicate diffusion coefficients that do not significantly change between the compared solutions while (filled) asterisks indicate a significant result. (A) CoLD-CoP comparison of solutions TL and TLP. In this comparison, the chemical shifts in the cluster denoted by blue asterisks are assignable to salicylate; that they are asterisks (rather than open circles) indicates that it binds tyrosinase; (B) comparison of solutions TLP and TLPK to detect competitive binding as described in Scheme 1; (C) Same as (B), but zoomed in to show competitive binding more clearly: asterisks indicate clusters with significantly higher diffusion coefficients in the presence of 4-HCCA, indicating competitive binding. Each significant cluster (yellow and red asterisks) is associated with chemical shifts assigned to salicylate. Note that salicylate, tartrate, and tris all have similar molecular masses and hence similar diffusion coefficients. This approach cannot distinguish between tris and tartrate (empty cyan circles), but does distinguish (albeit dividing it into two clusters) salicylate as competing with 4-HCCA for tyrosinase binding.
Figure 2. CoLD-CoP comparisons: circles and asterisks are colored (automatically by the MATLAB toolbox implementing CoLD-CoP) based on the cluster to which CoLD-CoP assigns them. Ideally, each cluster represents one and only one molecule, but molecules of similar size whose diffusion coefficient does not change between the solutions being compared may not be distinguishable based on paired diffusion coefficients. Open circles indicate diffusion coefficients that do not significantly change between the compared solutions while (filled) asterisks indicate a significant result. (A) CoLD-CoP comparison of solutions TL and TLP. In this comparison, the chemical shifts in the cluster denoted by blue asterisks are assignable to salicylate; that they are asterisks (rather than open circles) indicates that it binds tyrosinase; (B) comparison of solutions TLP and TLPK to detect competitive binding as described in Scheme 1; (C) Same as (B), but zoomed in to show competitive binding more clearly: asterisks indicate clusters with significantly higher diffusion coefficients in the presence of 4-HCCA, indicating competitive binding. Each significant cluster (yellow and red asterisks) is associated with chemical shifts assigned to salicylate. Note that salicylate, tartrate, and tris all have similar molecular masses and hence similar diffusion coefficients. This approach cannot distinguish between tris and tartrate (empty cyan circles), but does distinguish (albeit dividing it into two clusters) salicylate as competing with 4-HCCA for tyrosinase binding.
Biomolecules 14 01136 g002
Biomolecules 14 01136 g003
Figure 3. (A) The purpose of this comparison, performed as described in Scheme 1, Step 3.B.ii, is to detect any ligands binding competitively with GlcNAc via CoLD-CoP analysis of solution LLP (Lysozyme Ligands: Caffeine, Citrate, Imidazole, Tartrate, Tris, Tryptamine) against LLPK (Lysozyme +Ligands+Known Ligand: GlcNAc). CoLD-CoP does identify any ligand as competing with GlcNAc for lysozyme binding. (B) A similar comparison performed as described in Scheme 1, Step 3.B.iii. (C) Zoomed-in portion of panel (B). Points are colored by cluster as identified by CoLD-CoP, which identifies two diffusion coefficient pairs (closed circles) as diffusing significantly slower in the presence of (i.e., cooperatively binding with) GlcNAc: the closed orange circle corresponds (by chemical shift) to imidazole, and the closed green circle corresponds to tris.
Figure 3. (A) The purpose of this comparison, performed as described in Scheme 1, Step 3.B.ii, is to detect any ligands binding competitively with GlcNAc via CoLD-CoP analysis of solution LLP (Lysozyme Ligands: Caffeine, Citrate, Imidazole, Tartrate, Tris, Tryptamine) against LLPK (Lysozyme +Ligands+Known Ligand: GlcNAc). CoLD-CoP does identify any ligand as competing with GlcNAc for lysozyme binding. (B) A similar comparison performed as described in Scheme 1, Step 3.B.iii. (C) Zoomed-in portion of panel (B). Points are colored by cluster as identified by CoLD-CoP, which identifies two diffusion coefficient pairs (closed circles) as diffusing significantly slower in the presence of (i.e., cooperatively binding with) GlcNAc: the closed orange circle corresponds (by chemical shift) to imidazole, and the closed green circle corresponds to tris.
Biomolecules 14 01136 g003
Biomolecules 14 01136 g004
Figure 4. Model of the lysozyme/GlcNAc/tris ternary complex, illustrating the accommodation of both GlcNAc and tris in lysozyme’s active site. The binding surface of lysozyme is colored by charge (positive, blue; negative, red). Hydrogen bonds between GlcNAc (blue) and tris (green) are depicted as dotted red lines.
Figure 4. Model of the lysozyme/GlcNAc/tris ternary complex, illustrating the accommodation of both GlcNAc and tris in lysozyme’s active site. The binding surface of lysozyme is colored by charge (positive, blue; negative, red). Hydrogen bonds between GlcNAc (blue) and tris (green) are depicted as dotted red lines.
Biomolecules 14 01136 g004
Biomolecules 14 01136 g005
Figure 5. Interactions in the (A) lysozyme/GlcNAc complex, (B) lysozyme/tris complex, and (C,D) lysozyme/GlcNAc/tris complex highlighting hydrogen bonds with (C) GlcNAc and (D) tris. Lysozyme residues hydrogen-bonded to a ligand are shown in green. The presence of tris does not affect the hydrogen-bonding between GlcNAc and lysozyme. Two hydrogen bonds between tris and GlcNAc compensate for the loss of hydrogen bonds to Asn59 and Gln57 in the ternary complex and suggest a mechanism for cooperative lysozyme binding by tris and GlcNAc.
Figure 5. Interactions in the (A) lysozyme/GlcNAc complex, (B) lysozyme/tris complex, and (C,D) lysozyme/GlcNAc/tris complex highlighting hydrogen bonds with (C) GlcNAc and (D) tris. Lysozyme residues hydrogen-bonded to a ligand are shown in green. The presence of tris does not affect the hydrogen-bonding between GlcNAc and lysozyme. Two hydrogen bonds between tris and GlcNAc compensate for the loss of hydrogen bonds to Asn59 and Gln57 in the ternary complex and suggest a mechanism for cooperative lysozyme binding by tris and GlcNAc.
Biomolecules 14 01136 g005
Biomolecules 14 01136 sch003
Scheme 3. Model for cooperative GlcNAc and imidazole binding. (A) Lysozyme monomer (active), (B) Lysozyme dimer (inactive), (C) Lysozyme monomer + imidazole (competitively inhibited), (D) Lysozyme monomer + GlcNAc (competitively inhibited), (E) Lysozyme monomer + 2 imidazole, (F) Lysozyme monomer + GlcNAc + imidazole. GlcNAc and the imidazole at the dimer interface are shown where they dock to lysozyme. GlcNAc and imidazole structures are highlighted by purple hexagons and yellow pentagons, respectively, for clarity.
Scheme 3. Model for cooperative GlcNAc and imidazole binding. (A) Lysozyme monomer (active), (B) Lysozyme dimer (inactive), (C) Lysozyme monomer + imidazole (competitively inhibited), (D) Lysozyme monomer + GlcNAc (competitively inhibited), (E) Lysozyme monomer + 2 imidazole, (F) Lysozyme monomer + GlcNAc + imidazole. GlcNAc and the imidazole at the dimer interface are shown where they dock to lysozyme. GlcNAc and imidazole structures are highlighted by purple hexagons and yellow pentagons, respectively, for clarity.
Biomolecules 14 01136 sch003
Table 1. Mixtures used in this study.
Table 1. Mixtures used in this study.
MixtureCompoundConcentration (mM 1)
Tyrosinase Ligand OnlyGlcNAc2
(TL)Salicylic Acid3
Tartrate2
Thiamine2
Tris 210
Tryptamine3
ligand+protein TL—Salicylic Acid 3(as above)
(TLP)Salicylic Acid1
Mushroom Tyrosinase4 μM
ligand+protein+known ligandTLP(as above)
(TLPK)4-HCCA 48
Lysozyme ligand onlyCaffeine50
(LL)Citrate50
Imidazole100
Tartrate50
Tris 250
Tryptamine50
ligand+protein LL(as above)
(LLP)Hen Egg White Lysozyme1
ligand+protein+known ligandLLP(as above)
(LLPK)GlcNAc300
1 Unless otherwise indicated. 2 2-Amino-2-(hydroxymethyl)propane-1,3-diol, as Sigma Trizma® Pre-Set Crystals. 3 During preparation of solution TLP, some salicylic acid precipitated, resulting in a lower concentration of salicylic acid in this mixture. 4 4-hydroxy-α-cyanocinnamic acid.
Table 2. Results of MM-PBSA and MM-GBSA calculations for Lysozyme complexes.
Table 2. Results of MM-PBSA and MM-GBSA calculations for Lysozyme complexes.
ParametersGlcNAc 1Tris 1Ternary Complex 1Ternary Complex—
(GlcNAc + Tris) 1
PB15.41−5.32−5.34−5.43
PB3−6.89−16.99−27.33−3.45
PB4−17.16−25.67−42.350.48
GB1−23.13−28.59−49.642.08
GB2−25.42−27.1−55.25−2.73
GB5 2−29−28.05−59.96−2.91
GB6−20.25−26.19−50.8−4.36
GB7 2−34.67−26.31−68.03−7.05
GB8 2,3−21.77−20.27−45.05−3.01
1 energies in units of kcal/mol. 2 Results in bold correctly predict a lower free binding energy for GlcNAc than for tris. Note that all approaches that correctly predict GlcNAc binds better than tris—interestingly these all use the Generalized Born (GB) solvent model—also indicate cooperative binding between tris and GlcNAc. 3 Best GB parameter set as described in the FastDRH paper [36].
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Patnala, S.V.; Robles, R.; Snyder, D.A. Application of CoLD-CoP to Detecting Competitively and Cooperatively Binding Ligands. Biomolecules 2024, 14, 1136. https://doi.org/10.3390/biom14091136

AMA Style

Patnala SV, Robles R, Snyder DA. Application of CoLD-CoP to Detecting Competitively and Cooperatively Binding Ligands. Biomolecules. 2024; 14(9):1136. https://doi.org/10.3390/biom14091136

Chicago/Turabian Style

Patnala, Shiva V., Roberto Robles, and David A. Snyder. 2024. "Application of CoLD-CoP to Detecting Competitively and Cooperatively Binding Ligands" Biomolecules 14, no. 9: 1136. https://doi.org/10.3390/biom14091136

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop