Turning high-throughput structural biology into predictive inhibitor design

A common challenge in drug design pertains to finding chemical modifications to a ligand that increases its affinity to the target protein. An underutilised advance is the increase in structural biology throughput, which has progressed from an artisanal endeavour to a monthly throughput of up to 100 different ligands against a protein in modern synchrotrons. However, the missing piece is a framework that turns high throughput crystallography data into predictive models for ligand design. Here we designed a simple machine learning approach that predicts protein-ligand affinity from experimental structures of diverse ligands against a single protein paired with biochemical measurements. Our key insight is using physics-based energy descriptors to represent protein-ligand complexes, and a learning-to-rank approach that infers the relevant differences between binding modes. We ran a high throughput crystallography campaign against the SARS-CoV-2 Main Protease (MPro), obtaining parallel measurements of over 200 protein-ligand complexes and the binding activity. This allows us to design a one-step library syntheses which improved the potency of two distinct micromolar hits by over 10-fold, arriving at a non-covalent and non-peptidomimetic inhibitor with 120 nM antiviral efficacy. Crucially, our approach successfully extends ligands to unexplored regions of the binding pocket, executing large and fruitful moves in chemical space with simple chemistry.

Download Full-text

COVID Moonshot: Open Science Discovery of SARS-CoV-2 Main Protease Inhibitors by Combining Crowdsourcing, High-Throughput Experiments, Computational Simulations, and Machine Learning

10.1101/2020.10.29.339317 ◽

2020 ◽

Cited By ~ 1

Author(s):

◽

Hagit Achdout ◽

Anthony Aimon ◽

Elad Bar-David ◽

Haim Barr ◽

...

Keyword(s):

Machine Learning ◽

High Throughput ◽

Open Science ◽

Viral Inhibition ◽

Enzymatic Assays ◽

Main Protease ◽

Covalent Inhibitors ◽

Ic50 Values ◽

High Throughput Experiments ◽

High Throughput Crystallography

AbstractHerein we provide a living summary of the data generated during the COVID Moonshot project focused on the development of SARS-CoV-2 main protease (Mpro) inhibitors. Our approach uniquely combines crowdsourced medicinal chemistry insights with high throughput crystallography, exascale computational chemistry infrastructure for simulations, and machine learning in triaging designs and predicting synthetic routes. This manuscript describes our methodologies leading to both covalent and non-covalent inhibitors displaying protease IC50 values under 150 nM and viral inhibition under 5 uM in multiple different viral replication assays. Furthermore, we provide over 200 crystal structures of fragment-like and lead-like molecules in complex with the main protease. Over 1000 synthesized and ordered compounds are also reported with the corresponding activity in Mpro enzymatic assays using two different experimental setups. The data referenced in this document will be continually updated to reflect the current experimental progress of the COVID Moonshot project, and serves as a citable reference for ensuing publications. All of the generated data is open to other researchers who may find it of use.

Download Full-text

Metabolic intervention targeting 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3 (PFKFB3) using a structure-based design.

Journal of Clinical Oncology ◽

10.1200/jco.2012.30.15_suppl.e13518 ◽

2012 ◽

Vol 30 (15_suppl) ◽

pp. e13518-e13518

Author(s):

Lars Ährlund-Richter ◽

Katarina Färnegårdh ◽

Elisee Wiita ◽

Mattias Jönsson ◽

Carina Norström ◽

...

Keyword(s):

Cell Growth ◽

High Throughput ◽

High Throughput Screening ◽

Structural Information ◽

Human Cancer ◽

Isothermal Calorimetry ◽

Binding Pocket ◽

Binding Modes ◽

X Ray ◽

X Ray Crystallography

e13518 Background: By producing fructose-2,6-bisphosphate, PFKFB3 functions as an activator of anaerobic glycolysis. PFKFB3 is both over expressed and over activated in many of the types of human cancer. Specific inhibition of the PFKFB3 enzyme results in a reduction in metabolism and cell growth in oxygen-deficient cancer environments. Methods: High-throughput screening. Medicinal Chemistry. Structure-Based Drug Design, X-ray Crystallography. NMR. Isothermal Calorimetry. Dynamic Light Scatttering. ADME. Results: A high-throughput screening of 50.000 selected compounds, by means of a biochemical assay, generated 105 hits including both ATP-and non-ATP competitive hits as identified by NMR binding experiments. The latter type was prioritized and two hits with a similar “ring-linker-ring structure” were selected for further expansions. Interestingly, although structurally similar, the two hits were found by means of X-ray crystallography to exhibit different binding modes within the fructose pocket. Based on their respective binding mode, two chemical series were developed displaying different ADME properties and PFKFB isoenzyme selectivity. Calorimetry verified a reversible strong enthalpy driven, direct binding for both chemical series. A third chemical series was developed towards yet another unoccupied binding pocket within the fructose-site, yielding a 5-fold increase in potency. Strong interactions within the new pocket were confirmed using X-ray crystallography. Our PFKFB3 inhibitors were shown to reduce tumor cell growth in vitro and to exhibit combinatory effects with Cisplatin. Conclusions: We have targeted the fructose-binding pocket of PFKFB3, developed compounds with nM binding potency and have gained a detailed understanding of SAR via structural information. The structure-based analysis has provided a good understanding of the molecular interactions, which is important for further biological/clinical positioning: e.g., combination with chemotherapy, optimization of PK properties and proof of principle in vivo.

Download Full-text

High-throughput Screening and Experimental Identification of Potent Drugs Targeting SARS-CoV-2 Main Protease

10.21203/rs.3.rs-40014/v1 ◽

2020 ◽

Author(s):

Yan Li ◽

Jinyong Zhang ◽

Ning Wang ◽

Yanjing Zhang ◽

Yongjun Yang ◽

...

Keyword(s):

High Throughput ◽

High Throughput Screening ◽

Drug Targets ◽

Binding Pocket ◽

Substrate Binding Pocket ◽

Stable Conformation ◽

Cross Docking ◽

Main Protease ◽

Computer Based ◽

Old Drugs

Abstract The main protease (Mpro) is one of the best-characterized drug targets among coronaviruses. In the current study, we adopted a multiple cross-docking strategy against different crystal structures of SARS-CoV-2 Mpro to perform computer-based high-throughput virtual screening of possible inhibitors from a drug database using Autodock Vina and SeeSAR software, combined with our in-house automatic processing scripts. The KDs between screened candidates and Mpro were determined using Biacore. Seven drugs were found to fit the substrate-binding pocket of Mpro with a stable conformation, showing high KDs that ranged from 6.79E-7 M to 5.20E-5 M. Finally, mutagenesis studies confirmed that these drugs interact with Mpro specifically, suggesting that our method was reliable and convincing. Given the safety of these old drugs, they may serve as promising candidates to treat the infection of SARS-CoV-2. Our results also provide rational explanations for the behaviour of five drugs evaluated in clinical trials.

Download Full-text

COVID Moonshot: Open Science Discovery of SARS-CoV-2 Main Protease Inhibitors by Combining Crowdsourcing, High-Throughput Experiments, Computational Simulations, and Machine Learning

10.26434/chemrxiv.13158218 ◽

2020 ◽

Author(s):

The COVID Moonshot Consortium ◽

John Chodera ◽

Alpha Lee ◽

Nir London ◽

Frank von Delft

Keyword(s):

Machine Learning ◽

High Throughput ◽

Open Science ◽

Viral Inhibition ◽

Enzymatic Assays ◽

Main Protease ◽

Covalent Inhibitors ◽

Ic50 Values ◽

High Throughput Experiments ◽

High Throughput Crystallography

<div><div><div><p>Herein we provide a living summary of the data generated during the COVID Moonshot project focused on the development of SARS-CoV-2 main protease (Mpro) inhibitors. Our approach uniquely combines crowdsourced medicinal chemistry insights with high throughput crystallography, exascale computational chemistry infrastructure for simulations, and machine learning in triaging designs and predicting synthetic routes. This manuscript describes our methodologies leading to both covalent and non-covalent inhibitors displaying protease IC50 values under 150 nM and viral inhibition under 5 uM in multiple different viral replication assays. Furthermore, we provide over 200 crystal structures of fragment-like and lead-like molecules in complex with the main protease. Over 1000 synthesized and ordered compounds are also reported with the corresponding activity in Mpro enzymatic assays using two different experimental setups. The data referenced in this document will be continually updated to reflect the current experimental progress of the COVID Moonshot project, and serves as a citable reference for ensuing publications. All of the generated data is open to other researchers who may find it of use.<br></p></div></div></div>

Download Full-text

Structural biology and bioinformatics in drug design: opportunities and challenges for target identification and lead discovery

Philosophical Transactions of the Royal Society B Biological Sciences ◽

10.1098/rstb.2005.1800 ◽

2006 ◽

Vol 361 (1467) ◽

pp. 413-423 ◽

Cited By ~ 94

Author(s):

Tom L Blundell ◽

Bancinyane L Sibanda ◽

Rinaldo Wander Montalvão ◽

Suzanne Brewerton ◽

Vijayalakshmi Chelliah ◽

...

Keyword(s):

Drug Discovery ◽

Structural Biology ◽

Genome Sequencing ◽

High Throughput ◽

Structure Determination ◽

Target Identification ◽

Three Dimensional ◽

Central Place ◽

Lead Discovery ◽

High Throughput Crystallography

Impressive progress in genome sequencing, protein expression and high-throughput crystallography and NMR has radically transformed the opportunities to use protein three-dimensional structures to accelerate drug discovery, but the quantity and complexity of the data have ensured a central place for informatics. Structural biology and bioinformatics have assisted in lead optimization and target identification where they have well established roles; they can now contribute to lead discovery, exploiting high-throughput methods of structure determination that provide powerful approaches to screening of fragment binding.

Download Full-text

High-Throughput Crystallography Reveals Boron-Containing Inhibitors of a Penicillin-Binding Protein with Di- and Tricovalent Binding Modes

Journal of Medicinal Chemistry ◽

10.1021/acs.jmedchem.1c00717 ◽

2021 ◽

Author(s):

Hector Newman ◽

Alen Krajnc ◽

Dom Bellini ◽

Charles J. Eyermann ◽

Grant A. Boyle ◽

...

Keyword(s):

High Throughput ◽

Binding Protein ◽

Binding Modes ◽

Penicillin Binding Protein ◽

High Throughput Crystallography

Download Full-text

COVID Moonshot: Open Science Discovery of SARS-CoV-2 Main Protease Inhibitors by Combining Crowdsourcing, High-Throughput Experiments, Computational Simulations, and Machine Learning

10.26434/chemrxiv.13158218.v1 ◽

2020 ◽

Author(s):

The COVID Moonshot Consortium ◽

John Chodera ◽

Alpha Lee ◽

Nir London ◽

Frank von Delft

Keyword(s):

Machine Learning ◽

High Throughput ◽

Open Science ◽

Viral Inhibition ◽

Enzymatic Assays ◽

Main Protease ◽

Covalent Inhibitors ◽

Ic50 Values ◽

High Throughput Experiments ◽

High Throughput Crystallography

Download Full-text

Application of Molecular Docking for the Development of Improved HIV-1 Reverse Transcriptase Inhibitors

Current Computer - Aided Drug Design ◽

10.2174/1573409916666200628103359 ◽

2020 ◽

Vol 16 ◽

Author(s):

Arash Soltani ◽

Seyed Isaac Hashemy ◽

Farnaz Zahedi Avval ◽

Houshang Rafatpanah ◽

Seyed Abdolrahim Rezaee ◽

...

Keyword(s):

Reverse Transcriptase ◽

Binding Capacity ◽

Binding Pocket ◽

Ligand Docking ◽

Binding Modes ◽

Wild Type ◽

Parent Molecule ◽

Discovery Studio ◽

Approved Drugs ◽

Hiv 1

Introoduction: Inhibition of the reverse transcriptase (RT) enzyme of human immunodeficiency virus (HIV) by low molecular weight inhibitors is still an active area of research. Here, protein-ligand interactions and possible binding modes of novel compounds with the HIV-1 RT binding pocket (the wild-type as well as Y181C and K103N mutants) were obtained and discussed. Methods: A molecular fragment-based approach using FDA-approved drugs were followed to design novel chemical derivatives using delavirdine, efavirenz, etravirine and rilpivirine as the scaffolds. The drug-likeliness of the derivatives was evaluated using Swiss-ADME. Then the parent molecule and derivatives were docked into the binding pocket of related crystal structures (PDB ID: 4G1Q, 1IKW, 1KLM and 3MEC). Genetic Optimization for Ligand Docking (GOLD) Suite 5.2.2 software was used for docking and the results analyzed in the Discovery Studio Visualizer 4. A derivative was chosen for further analysis, if it passed drug-likeliness and the docked energy was more favorable than that of its parent molecule. Out of the fifty-seven derivatives, forty-eight failed in druglikeness screening by Swiss-ADME or in docking stage. Results: The final results showed that the selected compounds had higher predicted binding affinities than their parent scaffolds in both wild-type and the mutants. Binding energy improvement was higher for the structures designed based on second-generation NNRTIs (etravirine and rilpivirine) than the first-generation NNRTIs (delavirdine and efavirenz). For example, while the docked energy for rilpivirine was -51 KJ/mol, it was improved for its derivatives RPV01 and RPV15 up to -58.3 and -54.5 KJ/mol, respectively. Conclusion: In this study, we have identified and proposed some novel molecules with improved binding capacity for HIV RT using fragment-based approach.

Download Full-text