Rizzo Lab Research

OVERVIEW

Under the broad umbrella of Computational Structural Biology, research in the Rizzo lab involves Development, Validation, and Application of improved atomic-level computational modeling procedures and protocols for ranking and prioritizing compounds complexed with a biological target in order to better quantify, understand, and predict molecular recognition. Improved computational methods have great potential to save billions of dollars in drug development costs and reduce the time associated with bringing clinically useful medicines to market. Computational techniques for which we have expertise include docking (virtual screening), de novo design, molecular dynamics simulations, free energy calculations, SAR and associated analysis (energy decomposition, molecular footprinting, fold-resistance characterization). As outlined below, our Contributions to Science can be arranged into the following groups.

DOCK Method Development

A long-term goal of our group involves development and validation of improved computational procedures for predicting molecular recognition. Many of our application projects have a virtual screening component, thus a substantial effort has been undertaken to evaluate and improve sampling and scoring procedures in the program DOCK6, for which we are key developers. Among our accomplishments, we have: (i) spearheaded the last six DOCK6 releases (version 6.4, 6.5, 6.6, 6.7, 6.8, 6.9) graciously assisted by S. Brozell, D. Case group Rutgers, (ii) provided numerous code enhancements including growth trees, bug fixes, improved ligand internal energy, RMSD tether, torsion pre-minimizer, database filter, footprint similarity scoring, pharmacophore matching, multi-grid options, anchor selection options, SASA code, symmetry-corrected RMSD (Hungarian algorithm), (iii) constructed large validation databases, which allows us and others to develop and optimize new docking protocols, and (iv) provided educational resources for the community including a suite of online tutorials, fielding questions posted by the online community to the DOCK-fans listserv, and teaching a hands-on computer lab on molecule modeling, docking, and molecular dynamics. Websites reflecting some of our research contributions to the community include: (a) the DOCK6 Website, dock.compbio.ucsf.edu/DOCK_6/index.htm, (b) Rizzo Group Online Validation Databases, ringo.ams.sunysb.edu/index.php/Rizzo_Lab_Downloads, and (c) Rizzo Group Online Tutorials, ringo.ams.sunysb.edu/index.php/DOCK_Tutorials. Importantly, our software is free for academics and users receive all source code which, together with the educational activities outlined above, facilitates transparency, rigor, and reproducibility of the calculations. Representative publications include:

• Mukherjee, S.; Balius, T.E.; Rizzo, R. C. Docking Validation Resources: Protein Family and Ligand Flexibility Experiments. J. Chem. Inf. Model, 2010, 50, 1986-2000 PMCID: PMC3058392
• Brozell, S. R.; Mukherjee, S.; Balius, T. E.; Roe, D. R.; Case, D. A.; Rizzo, R. C. Evaluation of DOCK 6 as a Pose Generation and Database Enrichment Tool, J. Comput-Aided Mol. Des., 2012, 26, 749-773 PMCID: PMC3902891
• Allen, W.J; Balius, T. E.; Mukherjee, S.; Brozell, S. R.; Moustakas, D. T.; Lang, P. T.; Case, D. A.; Kuntz, I. D.; Rizzo, R. C. DOCK 6: Impact of New Features and Current Docking Performance, J. Comput. Chem., 2015, 36, 1132–1156 PMCID: PMC4469538
• Allen, W. J.; Fochtman, B. C.; Balius, T. E.; Rizzo, R. C. Customizable de novo Design Strategies for DOCK: Application to HIVgp41 and Other Therapeutic Targets, J. Comput. Chem., 2017, 38, 2641–2663 PMCID: PMC5659719

Scoring Functions for Virtual Screening and De novo Design

In addition to sampling, a critical component of molecular modeling software involves scoring functions. In particular, scoring functions are used for clustering, pruning, and rank-ordering of partially and fully-grown conformers (virtual screening) or newly constructed molecules (de novo design, genetic algorithms) so that the most promising candidates can be prioritized. We have developed, validated, and used in real word applications powerful new scoring functions, implemented into DOCK6, that enables users to search for candidates related to a known reference inhibitor, substrate, collection of protein hotspot residues, and more. Examples include footprint similarity (FPS) score, Hungarian matching similarity (HMS) score, and pharmacophore matching similarity (FMS) score, all of which can be used alone, or in combination, with or without standard energy-based terms. Representative publications include:

• Balius, T. E.; Mukherjee, S.; Rizzo, R. C. Implementation and Evaluation of a Docking-rescoring Method using Molecular Footprint Comparisons. J. Comput. Chem., 2011, 32, 2273-2289 PMCID: PMC3181325
• Balius, T. E.; Allen, W. J.; Mukherjee, S.; Rizzo, R. C. Grid-based Molecular Footprint Comparison Method for Docking and De Novo Design: Application to HIVgp41, J. Comput. Chem., 2013, 34, 1226-1240 PMCID: PMC4016043
• Allen, W. J.; Rizzo, R. C. Implementation of the Hungarian Algorithm to Account for Ligand Symmetry and Similarity in Structure-based Design, J. Chem. Inf. Model., 2014, 54, 518-529 PMCID: PMC3958141
• Jiang, L.; Rizzo, R. C. Pharmacophore-based Similarity Scoring for DOCK, J. Phys. Chem. B, 2015, 119, 1083-1102 PMCID: PMC4306494

Mechanisms of Specificity and Resistance:

In addition to inhibitor discovery we also employ atomic-level modeling to characterize mechanisms of binding, specificity, and in particular "resistance". As examples, we were the first lab to provide quantitative evidence that van der Waals interactions drive C-peptide binding to HIVgp41, supporting the hypothesis that a conserved hydrophobic pocket on gp41 is an important drug target site. And, we constructed and validated the first complete structural binding model for the fusion inhibitor T20 (Fuzeon) with gp41, subsequently verified by experiment (Buzeon et al, PLoS Pathog 2010). Another study quantified the specific ligand interactions which govern molecular recognition and resistance for the protein neuraminidase from influenza. Water mediated interactions are increasingly being appreciated as important for molecular recognition. Importantly, our work has elucidated the role of water in cancer-causing and acquired resistance mutations affecting specificity for several FDA-approved compounds and experimental inhibitors affecting ErbB-family receptor tyrosine kinases. Other breakthroughs, made in collaboration with experimental groups, include identification of residues most likely to be involved in substrate binding and catalysis for thioesterases (Wilson lab), and uncovering the molecular determinates of ceramide specificity and sphingolipid discriminations (Hannun lab). Representative publications include:

• Chachra, R.; Rizzo, R. C. Origins of Resistance Conferred by the R292K Neuraminidase Mutation via Molecular Dynamics and Free Energy Calculations. J. Chem. Theory Comput., 2008, 4, 1526-1540 PMID: 26621436
• Balius, T.E.; Rizzo, R. C. Quantitative Prediction of Fold Resistance for Inhibitors of EGFR. Biochemistry, 2009, 48, 8435-8448 PMCID: PMC2741091
• McGillick, B. E.; Balius, T.E.; Mukherjee, S.; Rizzo, R. C. Origins of Resistance to the HIVgp41 Viral Entry Inhibitor T20. Biochemistry, 2010, 49, 3575-3592 PMCID: PMC2867330
• Huang, Y.; Rizzo, R. C. A Water-based Mechanism of Specificity and Resistance for Lapatinib with ErbB Family Kinases, Biochemistry, 2012, 51, 2390-2406 PMID: 22352796

Inhibitor Discovery

Application projects in our group, in collaboration with several experimental labs (Jacobs, Deutsch, Ojima, Subramanyam, Miller), have identified experimentally-verified inhibitors targeting Zika glycoprotein E, Ebola GP2, HIVgp41, botulinum neurotoxin subtypes E and A, fatty acid binding protein (FABP), and HER2. Importantly, these successes are a direct result of our extensive method development efforts that have led to several innovative screening approaches, including those that leverage the wealth of energetic and structural information inherent to atomic-level molecular footprints – per-residue interaction patterns (maps) within targetable pockets on proteins – to rationally identify small molecules for compatibility with each target. Representative publications include:

• Zhou, Y.; McGillick, B. E.; Teng, Y. H.; Haranahalli, K; Ojima, I; Subramanyam, S.; Rizzo, R. C. Identification of Small Molecule Inhibitors of Botulinum Neurotoxin Serotype E via Footprint Similarity, Bioorg. Med. Chem., 2016, 24, 4875–4889 PMID: 27543389
• Guo, J.; Collins, S.; Miller, W.T.; Rizzo, R. C. Identification of a water-coordinating HER2 inhibitor by virtual screening using similarity-based scoring, Biochemistry, 2018, 57, 4934-4951, PMCID: PMC6110523
• Zhou, Y; Elmes, M. W.; Sweeney, J. M.; Joseph, O. M.; Che, J.; Hsu, H. C.; Li, H.; Deutsch, D. G.; Ojima, I; Kaczocha, M; Rizzo, R. C. Identification of Fatty Acid Binding Protein 5 Inhibitors Through Similarity-based Screening, Biochemistry, 2019, 58, 4304−4316, PMCID: PMC6812325
• Singleton, C. S.; Humbly, M.S.; Yi, H. A.; Rizzo, R. C.; Jacobs, A. Identification of Ebola Virus Inhibitors Targeting GP2 using Principles of Molecular Mimicry, Journal of Virology, 2019, 93, e00676-19 PMID: 31092576