Machine Learning for Prioritization of Thermostabilizing Mutations for G-Protein Coupled Receptors

The application of artificial intelligence to accelerate G protein-coupled receptor drug discovery

The application of artificial intelligence (AI) approaches to drug discovery for G protein-coupled receptors (GPCRs) is a rapidly expanding area. Artificial intelligence can be used at multiple stages during the drug discovery process, from aiding our understanding of the fundamental actions of GPCRs to the discovery of new ligand-GPCR interactions or the prediction of clinical responses. Here, we provide an overview of the concepts behind artificial intelligence, including the subfields of machine learning and deep learning. We summarise the published applications of artificial intelligence to different stages of the GPCR drug discovery process. Finally, we reflect on the benefits and limitations of artificial intelligence and share our vision for the exciting potential for further development of applications to aid GPCR drug discovery. In addition to making the drug discovery process "faster, smarter and cheaper," we anticipate that the application of artificial intelligence will create exciting new opportunities for GPCR drug discovery. LINKED ARTICLES: This article is part of a themed issue Therapeutic Targeting of G Protein-Coupled Receptors: hot topics from the Australasian Society of Clinical and Experimental Pharmacologists and Toxicologists 2021 Virtual Annual Scientific Meeting. To view the other articles in this section visit http://onlinelibrary.wiley.com/doi/10.1111/bph.v181.14/issuetoc.

AI-driven GPCR analysis, engineering, and targeting

This article investigates the role of recent advances in Artificial Intelligence (AI) to revolutionise the study of G protein-coupled receptors (GPCRs). AI has been applied to many areas of GPCR research, including the application of machine learning (ML) in GPCR classification, prediction of GPCR activation levels, modelling GPCR 3D structures and interactions, understanding G-protein selectivity, aiding elucidation of GPCRs structures, and drug design. Despite progress, challenges in predicting GPCR structures and addressing the complex nature of GPCRs remain, providing avenues for future research and development.

Incorporating physics to overcome data scarcity in predictive modeling of protein function: A case study of BK channels

Machine learning has played transformative roles in numerous chemical and biophysical problems such as protein folding where large amount of data exists. Nonetheless, many important problems remain challenging for data-driven machine learning approaches due to the limitation of data scarcity. One approach to overcome data scarcity is to incorporate physical principles such as through molecular modeling and simulation. Here, we focus on the big potassium (BK) channels that play important roles in cardiovascular and neural systems. Many mutants of BK channel are associated with various neurological and cardiovascular diseases, but the molecular effects are unknown. The voltage gating properties of BK channels have been characterized for 473 site-specific mutations experimentally over the last three decades; yet, these functional data by themselves remain far too sparse to derive a predictive model of BK channel voltage gating. Using physics-based modeling, we quantify the energetic effects of all single mutations on both open and closed states of the channel. Together with dynamic properties derived from atomistic simulations, these physical descriptors allow the training of random forest models that could reproduce unseen experimentally measured shifts in gating voltage, ∆V1/2, with a RMSE ~ 32 mV and correlation coefficient of R ~ 0.7. Importantly, the model appears capable of uncovering nontrivial physical principles underlying the gating of the channel, including a central role of hydrophobic gating. The model was further evaluated using four novel mutations of L235 and V236 on the S5 helix, mutations of which are predicted to have opposing effects on V1/2 and suggest a key role of S5 in mediating voltage sensor-pore coupling. The measured ∆V1/2 agree quantitatively with prediction for all four mutations, with a high correlation of R = 0.92 and RMSE = 18 mV. Therefore, the model can capture nontrivial voltage gating properties in regions where few mutations are known. The success of predictive modeling of BK voltage gating demonstrates the potential of combining physics and statistical learning for overcoming data scarcity in nontrivial protein function prediction.

Incorporating physics to overcome data scarcity in predictive modeling of protein function: a case study of BK channels

Machine learning has played transformative roles in numerous chemical and biophysical problems such as protein folding where large amount of data exists. Nonetheless, many important problems remain challenging for data-driven machine learning approaches due to the limitation of data scarcity. One approach to overcome data scarcity is to incorporate physical principles such as through molecular modeling and simulation. Here, we focus on the big potassium (BK) channels that play important roles in cardiovascular and neural systems. Many mutants of BK channel are associated with various neurological and cardiovascular diseases, but the molecular effects are unknown. The voltage gating properties of BK channels have been characterized for 473 site-specific mutations experimentally over the last three decades; yet, these functional data by themselves remain far too sparse to derive a predictive model of BK channel voltage gating. Using physics-based modeling, we quantify the energetic effects of all single mutations on both open and closed states of the channel. Together with dynamic properties derived from atomistic simulations, these physical descriptors allow the training of random forest models that could reproduce unseen experimentally measured shifts in gating voltage, ΔV 1/2 , with a RMSE ∼ 32 mV and correlation coefficient of R ∼ 0.7. Importantly, the model appears capable of uncovering nontrivial physical principles underlying the gating of the channel, including a central role of hydrophobic gating. The model was further evaluated using four novel mutations of L235 and V236 on the S5 helix, mutations of which are predicted to have opposing effects on V 1/2 and suggest a key role of S5 in mediating voltage sensor-pore coupling. The measured ΔV 1/2 agree quantitatively with prediction for all four mutations, with a high correlation of R = 0.92 and RMSE = 18 mV. Therefore, the model can capture nontrivial voltage gating properties in regions where few mutations are known. The success of predictive modeling of BK voltage gating demonstrates the potential of combining physics and statistical learning for overcoming data scarcity in nontrivial protein function prediction.

Author Summary

Deep machine learning has brought many exciting breakthroughs in chemistry, physics and biology. These models require large amount of training data and struggle when the data is scarce. The latter is true for predictive modeling of the function of complex proteins such as ion channels, where only hundreds of mutational data may be available. Using the big potassium (BK) channel as a biologically important model system, we demonstrate that a reliable predictive model of its voltage gating property could be derived from only 473 mutational data by incorporating physics-derived features, which include dynamic properties from molecular dynamics simulations and energetic quantities from Rosetta mutation calculations. We show that the final random forest model captures key trends and hotspots in mutational effects of BK voltage gating, such as the important role of pore hydrophobicity. A particularly curious prediction is that mutations of two adjacent residues on the S5 helix would always have opposite effects on the gating voltage, which was confirmed by experimental characterization of four novel mutations. The current work demonstrates the importance and effectiveness of incorporating physics in predictive modeling of protein function with scarce data.

Fluorescence-Detection Size-Exclusion Chromatography-Based Thermostability Assay for Membrane Proteins

Green fluorescent proteins (GFPs) have lightened up almost every aspect of biological research including protein sciences. In the field of membrane protein structural biology, GFPs have been used widely to monitor membrane protein localization, expression level, the purification process and yield, and the stability inside the cells and in the test tube. Of particular interest is the fluorescence-detector size-exclusion chromatography-based thermostability assay (FSEC-TS). By simple heating and FSEC, the generally applicable method allows rapid assessment of the thermostability of GFP-fused membrane proteins without purification. Here we describe the experimental details and some typical results for the FSEC-TS method.

Towards generalizable predictions for G protein-coupled receptor variant expression

Missense mutations that compromise the plasma membrane expression (PME) of integral membrane proteins are the root cause of numerous genetic diseases. Differentiation of this class of mutations from those that specifically modify the activity of the folded protein has proven useful for the development and targeting of precision therapeutics. Nevertheless, it remains challenging to predict the effects of mutations on the stability and/ or expression of membrane proteins. In this work, we utilize deep mutational scanning data to train a series of artificial neural networks to predict the PME of transmembrane domain variants of G protein-coupled receptors from structural and/ or evolutionary features. We show that our best-performing network, which we term the PME predictor, can recapitulate mutagenic trends within rhodopsin and can differentiate pathogenic transmembrane domain variants that cause it to misfold from those that compromise its signaling. This network also generates statistically significant predictions for the relative PME of transmembrane domain variants for another class A G protein-coupled receptor (β₂ adrenergic receptor) but not for an unrelated voltage-gated potassium channel (KCNQ1). Notably, our analyses of these networks suggest structural features alone are generally sufficient to recapitulate the observed mutagenic trends. Moreover, our findings imply that networks trained in this manner may be generalizable to proteins that share a common fold. Implications of our findings for the design of mechanistically specific genetic predictors are discussed.

Sequence coevolution and structure stabilization modulate olfactory receptor expression

Olfactory receptors (ORs) belong to class A G-protein coupled receptors (GPCRs) and are activated by a variety of odorants. To date, there is no three-dimensional structure of an OR available. One of the major bottlenecks in obtaining purified protein for structural studies of ORs is their poor expression in heterologous cells. To design mutants that enhance expression and thereby enable protein purification, we first identified computable physical properties that recapitulate OR and class A GPCR expression and further conducted an iterative computational prediction-experimental test cycle and generated human OR mutants that express as high as biogenic amine receptors for which structures have been solved. In the process of developing the computational method to recapitulate the expression of ORs in membranes, we identified properties, such as amino acid sequence coevolution, and the strength of the interactions between intracellular loop 1 (ICL1) and the helix 8 region of ORs, to enhance their heterologous expression. We identified mutations that are directly located in these regions as well as other mutations not located in these regions but allosterically strengthen the ICL1-helix 8 enhance expression. These mutants also showed functional responses to known odorants. This method to enhance heterologous expression of mammalian ORs will facilitate high-throughput "deorphanization" of ORs, and enable OR purification for biochemical and structural studies to understand odorant-OR interactions.

Computational design of a cutinase for plastic biodegradation by mining molecular dynamics simulations trajectories

Polyethylene terephthalate (PET) has caused serious environmental concerns but could be degraded at high temperature. Previous studies show that cutinase from Thermobifida fusca KW3 (TfCut2) is capable of degrading and upcycling PET but is limited by its thermal stability. Nowadays, Popular protein stability modification methods rely mostly on the crystal structures, but ignore the fact that the actual conformation of protein is complex and constantly changing. To solve these problems, we developed a computational approach to design variants with enhanced protein thermal stability by mining Molecular Dynamics simulation trajectories using Machine Learning methods (MDL). The optimal classification accuracy and the optimal Pearson correlation coefficient of MDL model were 0.780 and 0.716, respectively. And we successfully designed variants with high ΔT_m values using MDL method. The optimal variant S121P/D174S/D204P had the highest ΔT_m value of 9.3 °C, and the PET degradation ratio increased by 46.42-fold at 70℃, compared with that of wild type TfCut2. These results deepen our understanding on the complex conformations of proteins and may enhance the plastic recycling and sustainability at glass transition temperature.

An online GPCR structure analysis platform

We present an online, interactive platform for comparative analysis of all available G-protein coupled receptor (GPCR) structures while correlating to functional data. The comprehensive platform encompasses structure similarity, secondary structure, protein backbone packing and movement, residue-residue contact networks, amino acid properties and prospective design of experimental mutagenesis studies. This lets any researcher tap the potential of sophisticated structural analyses enabling a plethora of basic and applied receptor research studies.

IMPROvER: the Integral Membrane Protein Stability Selector

Identifying stabilising variants of membrane protein targets is often required for structure determination. Our new computational pipeline, the Integral Membrane Protein Stability Selector (IMPROvER) provides a rational approach to variant selection by employing three independent approaches: deep-sequence, model-based and data-driven. In silico tests using known stability data, and in vitro tests using three membrane protein targets with 7, 11 and 16 transmembrane helices provided measures of success. In vitro, individual approaches alone all identified stabilising variants at a rate better than expected by random selection. Low numbers of overlapping predictions between approaches meant a greater success rate was achieved (fourfold better than random) when approaches were combined and selections restricted to the highest ranked sites. The mix of information IMPROvER uses can be extracted for any helical membrane protein. We have developed the first general-purpose tool for selecting stabilising variants of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\upalpha$$\end{document}α-helical membrane proteins, increasing efficiency and reducing workload. IMPROvER can be accessed at http://improver.ddns.net/IMPROvER/.

Expression and purification of recombinant G protein-coupled receptors: A review

Given their extensive role in cell signalling, GPCRs are significant drug targets; despite this, many of these receptors have limited or no available prophylaxis. Novel drug design and discovery significantly rely on structure determination, of which GPCRs are typically elusive. Progress has been made thus far to produce sufficient quantity and quality of protein for downstream analysis. As such, this review highlights the systems available for recombinant GPCR expression, with consideration of their advantages and disadvantages, as well as examples of receptors successfully expressed in these systems. Additionally, an overview is given on the use of detergents and the styrene maleic acid (SMA) co-polymer for membrane solubilisation, as well as purification techniques.