Mapping the energetic and allosteric landscapes of protein binding domains

Deep mutational scanning quantifies DNA binding and predicts clinical outcomes of PAX6 variants

Nonsense and missense mutations in the transcription factor PAX6 cause a wide range of eye development defects, including aniridia, microphthalmia and coloboma. To understand how changes of PAX6:DNA binding cause these phenotypes, we combined saturation mutagenesis of the paired domain of PAX6 with a yeast one-hybrid (Y1H) assay in which expression of a PAX6-GAL4 fusion gene drives antibiotic resistance. We quantified binding of more than 2700 single amino-acid variants to two DNA sequence elements. Mutations in DNA-facing residues of the N-terminal subdomain and linker region were most detrimental, as were mutations to prolines and to negatively charged residues. Many variants caused sequence-specific molecular gain-of-function effects, including variants in position 71 that increased binding to the LE9 enhancer but decreased binding to a SELEX-derived binding site. In the absence of antibiotic selection, variants that retained DNA binding slowed yeast growth, likely because such variants perturbed the yeast transcriptome. Benchmarking against known patient variants and applying ACMG/AMP guidelines to variant classification, we obtained supporting-to-moderate evidence that 977 variants are likely pathogenic and 1306 are likely benign. Our analysis shows that most pathogenic mutations in the paired domain of PAX6 can be explained simply by the effects of these mutations on PAX6:DNA association, and establishes Y1H as a generalisable assay for the interpretation of variant effects in transcription factors.

Discovering mechanisms of human genetic variation and controlling cell states at scale

Population-scale sequencing efforts have catalogued substantial genetic variation in humans such that variant discovery dramatically outpaces interpretation. We discuss how single-cell sequencing is poised to reveal genetic mechanisms at a rate that may soon approach that of variant discovery. The functional genomics toolkit is sufficiently modular to systematically profile almost any type of variation within increasingly diverse contexts and with molecularly comprehensive and unbiased readouts. As a result, we can construct deep phenotypic atlases of variant effects that span the entire regulatory cascade. The same conceptual approach to interpreting genetic variation should be applied to engineering therapeutic cell states. In this way, variant mechanism discovery and cell state engineering will become reciprocating and iterative processes towards genomic medicine.

Gauge fixing for sequence-function relationships

Quantitative models of sequence-function relationships are ubiquitous in computational biology, e.g., for modeling the DNA binding of transcription factors or the fitness landscapes of proteins. Interpreting these models, however, is complicated by the fact that the values of model parameters can often be changed without affecting model predictions. Before the values of model parameters can be meaningfully interpreted, one must remove these degrees of freedom (called "gauge freedoms" in physics) by imposing additional constraints (a process called "fixing the gauge"). However, strategies for fixing the gauge of sequence-function relationships have received little attention. Here we derive an analytically tractable family of gauges for a large class of sequence-function relationships. These gauges are derived in the context of models with all-order interactions, but an important subset of these gauges can be applied to diverse types of models, including additive models, pairwise-interaction models, and models with higher-order interactions. Many commonly used gauges are special cases of gauges within this family. We demonstrate the utility of this family of gauges by showing how different choices of gauge can be used both to explore complex activity landscapes and to reveal simplified models that are approximately correct within localized regions of sequence space. The results provide practical gauge-fixing strategies and demonstrate the utility of gauge-fixing for model exploration and interpretation.

Optimization of a deep mutational scanning workflow to improve quantification of mutation effects on protein-protein interactions

Deep Mutational Scanning (DMS) assays are powerful tools to study sequence-function relationships by measuring the effects of thousands of sequence variants on protein function. During a DMS experiment, several technical artefacts might distort non-linearly the functional score obtained, potentially biasing the interpretation of the results. We therefore tested several technical parameters in the deepPCA workflow, a DMS assay for protein-protein interactions, in order to identify technical sources of non-linearities. We found that parameters common to many DMS assays such as amount of transformed DNA, timepoint of harvest and library composition can cause non-linearities in the data. Designing experiments in a way to minimize these non-linear effects will improve the quantification and interpretation of mutation effects.

Analysis of 10,478 cancer genomes identifies candidate driver genes and opportunities for precision oncology

Tumor genomic profiling is increasingly seen as a prerequisite to guide the treatment of patients with cancer. To explore the value of whole-genome sequencing (WGS) in broadening the scope of cancers potentially amenable to a precision therapy, we analysed whole-genome sequencing data on 10,478 patients spanning 35 cancer types recruited to the UK 100,000 Genomes Project. We identified 330 candidate driver genes, including 74 that are new to any cancer. We estimate that approximately 55% of patients studied harbor at least one clinically relevant mutation, predicting either sensitivity or resistance to certain treatments or clinical trial eligibility. By performing computational chemogenomic analysis of cancer mutations we identify additional targets for compounds that represent attractive candidates for future clinical trials. This study represents one of the most comprehensive efforts thus far to identify cancer driver genes in the real world setting and assess their impact on informing precision oncology.

Engineering of Halide Methyltransferase BxHMT through Dynamic Cross-Correlation Network Analysis

Halide methyltransferases (HMTs) provide an effective way to regenerate S-adenosyl methionine (SAM) from S-adenosyl homocysteine and reactive electrophiles, such as methyl iodide (MeI) and methyl toluene sulfonate (MeOTs). As compared with MeI, the cost-effective unnatural substrate MeOTs can be accessed directly from cheap and abundant alcohols, but shows only limited reactivity in SAM production. In this study, we developed a dynamic cross-correlation network analysis (DCCNA) strategy for quickly identifying hot spots influencing the catalytic efficiency of the enzyme, and applied it to the evolution of HMT from Paraburkholderia xenovorans. Finally, the optimal mutant, M4 (V55T/C125S/L127T/L129P), exhibited remarkable improvement, with a specific activity of 4.08 U/mg towards MeOTs, representing an 82-fold increase as compared to the wild-type (WT) enzyme. Notably, M4 also demonstrated a positive impact on the catalytic ability with other methyl donors. The structural mechanism behind the enhanced enzyme activity was uncovered by molecular dynamics simulations. Our work not only contributes a promising biocatalyst for the regeneration of SAM, but also offers a strategy for efficient enzyme engineering.

Sampling Strategies for Experimentally Mapping Molecular Fitness Landscapes Using High-Throughput Methods

Empirical studies of genotype-phenotype-fitness maps of proteins are fundamental to understanding the evolutionary process, in elucidating the space of possible genotypes accessible through mutations in a landscape of phenotypes and fitness effects. Yet, comprehensively mapping molecular fitness landscapes remains challenging since all possible combinations of amino acid substitutions for even a few protein sites are encoded by an enormous genotype space. High-throughput mapping of genotype space can be achieved using large-scale screening experiments known as multiplexed assays of variant effect (MAVEs). However, to accommodate such multi-mutational studies, the size of MAVEs has grown to the point where a priori determination of sampling requirements is needed. To address this problem, we propose calculations and simulation methods to approximate minimum sampling requirements for multi-mutational MAVEs, which we combine with a new library construction protocol to experimentally validate our approximation approaches. Analysis of our simulated data reveals how sampling trajectories differ between simulations of nucleotide versus amino acid variants and among mutagenesis schemes. For this, we show quantitatively that marginal gains in sampling efficiency demand increasingly greater sampling effort when sampling for nucleotide sequences over their encoded amino acid equivalents. We present a new library construction protocol that efficiently maximizes sequence variation, and demonstrate using ultradeep sequencing that the library encodes virtually all possible combinations of mutations within the experimental design. Insights learned from our analyses together with the methodological advances reported herein are immediately applicable toward pooled experimental screens of arbitrary design, enabling further assay upscaling and expanded testing of genotype space.

Aptamer-Based DNA Allosteric Switch for Regulation of Protein Activity

Allostery is a fundamental way to regulate the function of biomolecules playing crucial roles in cell metabolism and proliferation and is deemed the second secret of life. Given the limited understanding of the structure of natural allosteric molecules, the development of artificial allosteric molecules brings a huge opportunity to transform the allosteric mechanism into practical applications. In this study, the concept of bionics is introduced into the design of artificial allosteric molecules and an allosteric DNA switch with an activity site and an allosteric site based on two aptamers for selective inhibition of thrombin activity. Compared with the single aptamer, the allosteric switch possesses a significantly enhanced inhibition ability, which can be precisely regulated by converting the switch states. Moreover, the dynamic allosteric switch is further subjected to the control of the DNA threshold circuit for realizing automatic concentration determination and activity inhibition of thrombin. These compelling results confirm that this allosteric switch equipped with self-sensing and information-processing modules puts a new slant on the research of allosteric mechanisms and further application of allosteric tactics in chemical and biomedical fields.

Discovery and characterization of novel FGFR1 inhibitors in triple-negative breast cancer via hybrid virtual screening and molecular dynamics simulations

The overexpression of FGFR1 is thought to significantly contribute to the progression of triple-negative breast cancer (TNBC), impacting aspects such as tumorigenesis, growth, metastasis, and drug resistance. Consequently, the pursuit of effective inhibitors for FGFR1 is a key area of research interest. In response to this need, our study developed a hybrid virtual screening method. Utilizing KarmaDock, an innovative algorithm that blends deep learning with molecular docking, alongside Schrödinger's Residue Scanning. This strategy led us to identify compound 6, which demonstrated promising FGFR1 inhibitory activity, evidenced by an IC50 value of approximately 0.24 nM in the HTRF bioassay. Further evaluation revealed that this compound also inhibits the FGFR1 V561M variant with an IC50 value around 1.24 nM. Our subsequent investigations demonstrate that Compound 6 robustly suppresses the migration and invasion capacities of TNBC cell lines, through the downregulation of p-FGFR1 and modulation of EMT markers, highlighting its promise as a potent anti-metastatic therapeutic agent. Additionally, our use of molecular dynamics simulations provided a deeper understanding of the compound's specific binding interactions with FGFR1.

A parameterized two-domain thermodynamic model explains diverse mutational effects on protein allostery

New experimental findings continue to challenge our understanding of protein allostery. Recent deep mutational scanning study showed that allosteric hotspots in the tetracycline repressor (TetR) and its homologous transcriptional factors are broadly distributed rather than spanning well-defined structural pathways as often assumed. Moreover, hotspot mutation-induced allostery loss was rescued by distributed additional mutations in a degenerate fashion. Here, we develop a two-domain thermodynamic model for TetR, which readily rationalizes these intriguing observations. The model accurately captures the in vivo activities of various mutants with changes in physically transparent parameters, allowing the data-based quantification of mutational effects using statistical inference. Our analysis reveals the intrinsic connection of intra- and inter-domain properties for allosteric regulation and illustrate epistatic interactions that are consistent with structural features of the protein. The insights gained from this study into the nature of two-domain allostery are expected to have broader implications for other multi-domain allosteric proteins.

Predicted mechanistic impacts of human protein missense variants

Genome sequencing efforts have led to the discovery of tens of millions of protein missense variants found in the human population with the majority of these having no annotated role and some likely contributing to trait variation and disease. Sequence-based artificial intelligence approaches have become highly accurate at predicting variants that are detrimental to the function of proteins but they do not inform on mechanisms of disruption. Here we combined sequence and structure-based methods to perform proteome-wide prediction of deleterious variants with information on their impact on protein stability, protein-protein interactions and small-molecule binding pockets. AlphaFold2 structures were used to predict approximately 100,000 small-molecule binding pockets and stability changes for over 200 million variants. To inform on protein-protein interfaces we used AlphaFold2 to predict structures for nearly 500,000 protein complexes. We illustrate the value of mechanism-aware variant effect predictions to study the relation between protein stability and abundance and the structural properties of interfaces underlying trans protein quantitative trait loci (pQTLs). We characterised the distribution of mechanistic impacts of protein variants found in patients and experimentally studied example disease linked variants in FGFR1.

Rosace: a robust deep mutational scanning analysis framework employing position and mean-variance shrinkage

Deep mutational scanning (DMS) measures the effects of thousands of genetic variants in a protein simultaneously. The small sample size renders classical statistical methods ineffective. For example, p-values cannot be correctly calibrated when treating variants independently. We propose Rosace, a Bayesian framework for analyzing growth-based DMS data. Rosace leverages amino acid position information to increase power and control the false discovery rate by sharing information across parameters via shrinkage. We also developed Rosette for simulating the distributional properties of DMS. We show that Rosace is robust to the violation of model assumptions and is more powerful than existing tools.

The full spectrum of SLC22 OCT1 mutations illuminates the bridge between drug transporter biophysics and pharmacogenomics

Mutations in transporters can impact an individual's response to drugs and cause many diseases. Few variants in transporters have been evaluated for their functional impact. Here, we combine saturation mutagenesis and multi-phenotypic screening to dissect the impact of 11,213 missense single-amino-acid deletions, and synonymous variants across the 554 residues of OCT1, a key liver xenobiotic transporter. By quantifying in parallel expression and substrate uptake, we find that most variants exert their primary effect on protein abundance, a phenotype not commonly measured alongside function. Using our mutagenesis results combined with structure prediction and molecular dynamic simulations, we develop accurate structure-function models of the entire transport cycle, providing biophysical characterization of all known and possible human OCT1 polymorphisms. This work provides a complete functional map of OCT1 variants along with a framework for integrating functional genomics, biophysical modeling, and human genetics to predict variant effects on disease and drug efficacy.

An extension of the Walsh-Hadamard transform to calculate and model epistasis in genetic landscapes of arbitrary shape and complexity

Accurate models describing the relationship between genotype and phenotype are necessary in order to understand and predict how mutations to biological sequences affect the fitness and evolution of living organisms. The apparent abundance of epistasis (genetic interactions), both between and within genes, complicates this task and how to build mechanistic models that incorporate epistatic coefficients (genetic interaction terms) is an open question. The Walsh-Hadamard transform represents a rigorous computational framework for calculating and modeling epistatic interactions at the level of individual genotypic values (known as genetical, biological or physiological epistasis), and can therefore be used to address fundamental questions related to sequence-to-function encodings. However, one of its main limitations is that it can only accommodate two alleles (amino acid or nucleotide states) per sequence position. In this paper we provide an extension of the Walsh-Hadamard transform that allows the calculation and modeling of background-averaged epistasis (also known as ensemble epistasis) in genetic landscapes with an arbitrary number of states per position (20 for amino acids, 4 for nucleotides, etc.). We also provide a recursive formula for the inverse matrix and then derive formulae to directly extract any element of either matrix without having to rely on the computationally intensive task of constructing or inverting large matrices. Finally, we demonstrate the utility of our theory by using it to model epistasis within both simulated and empirical multiallelic fitness landscapes, revealing that both pairwise and higher-order genetic interactions are enriched between physically interacting positions.

Integrated multiplexed assays of variant effect reveal determinants of catechol-O-methyltransferase gene expression

Multiplexed assays of variant effect are powerful methods to profile the consequences of rare variants on gene expression and organismal fitness. Yet, few studies have integrated several multiplexed assays to map variant effects on gene expression in coding sequences. Here, we pioneered a multiplexed assay based on polysome profiling to measure variant effects on translation at scale, uncovering single-nucleotide variants that increase or decrease ribosome load. By combining high-throughput ribosome load data with multiplexed mRNA and protein abundance readouts, we mapped the cis-regulatory landscape of thousands of catechol-O-methyltransferase (COMT) variants from RNA to protein and found numerous coding variants that alter COMT expression. Finally, we trained machine learning models to map signatures of variant effects on COMT gene expression and uncovered both directional and divergent impacts across expression layers. Our analyses reveal expression phenotypes for thousands of variants in COMT and highlight variant effects on both single and multiple layers of expression. Our findings prompt future studies that integrate several multiplexed assays for the readout of gene expression.

Multiplexed, multimodal profiling of the intracellular activity, interactions, and druggability of protein variants using LABEL-seq

Multiplexed assays of variant effect are powerful tools for assessing the impact of protein sequence variation, but are limited to measuring a single protein property and often rely on indirect readouts of intracellular protein function. Here, we developed LAbeling with Barcodes and Enrichment for biochemicaL analysis by sequencing (LABEL-seq), a platform for the multimodal profiling of thousands of protein variants in cultured human cells. Multimodal measurement of ~20,000 variant effects for ~1,600 BRaf variants using LABEL-seq revealed that variation at positions that are frequently mutated in cancer had minimal effects on folding and intracellular abundance but could dramatically alter activity, protein-protein interactions, and druggability. Integrative analysis of our multimodal measurements identified networks of positions with similar roles in regulating BRaf's signaling properties and enabled predictive modeling of variant effects on complex processes such as cell proliferation and small molecule-promoted degradation. LABEL-seq provides a scalable approach for the direct measurement of multiple biochemical effects of protein variants in their native cellular context, yielding insight into protein function, disease mechanisms, and druggability.

A complete map of specificity encoding for a partially fuzzy protein interaction

Thousands of human proteins function by binding short linear motifs embedded in intrinsically disordered regions. How affinity and specificity are encoded in these binding domains and the motifs themselves is not well understood. The evolvability of binding specificity - how rapidly and extensively it can change upon mutation - is also largely unexplored, as is the contribution of 'fuzzy' dynamic residues to affinity and specificity in protein-protein interactions. Here we report the first complete map of specificity encoding for a globular protein domain. Quantifying >200,000 energetic interactions between a PDZ domain and its ligand identifies 20 major energetically coupled pairs of sites that control specificity. These are organized into six modules, with most mutations in each module reprogramming specificity for a single position in the ligand. Nine of the major energetic couplings controlling specificity are between structural contacts and 11 have an allosteric mechanism of action. The dynamic tail of the ligand is more robust to mutation than the structured residues but contributes additively to binding affinity and communicates with structured residues to enable changes in specificity. Our results quantify the binding specificities of >1,800 globular proteins to reveal how specificity is encoded and provide a direct comparison of the encoding of affinity and specificity in structured and dynamic molecular recognition.

Molecular basis of proton-sensing by G protein-coupled receptors

Three proton-sensing G protein-coupled receptors (GPCRs), GPR4, GPR65, and GPR68, respond to changes in extracellular pH to regulate diverse physiology and are implicated in a wide range of diseases. A central challenge in determining how protons activate these receptors is identifying the set of residues that bind protons. Here, we determine structures of each receptor to understand the spatial arrangement of putative proton sensing residues in the active state. With a newly developed deep mutational scanning approach, we determined the functional importance of every residue in proton activation for GPR68 by generating ~9,500 mutants and measuring effects on signaling and surface expression. This unbiased screen revealed that, unlike other proton-sensitive cell surface channels and receptors, no single site is critical for proton recognition in GPR68. Instead, a network of titratable residues extend from the extracellular surface to the transmembrane region and converge on canonical class A GPCR activation motifs to activate proton-sensing GPCRs. More broadly, our approach integrating structure and unbiased functional interrogation defines a new framework for understanding the rich complexity of GPCR signaling.

The genetic landscape of a metabolic interaction

While much prior work has explored the constraints on protein sequence and evolution induced by physical protein-protein interactions, the sequence-level constraints emerging from non-binding functional interactions in metabolism remain unclear. To quantify how variation in the activity of one enzyme constrains the biochemical parameters and sequence of another, we focus on dihydrofolate reductase (DHFR) and thymidylate synthase (TYMS), a pair of enzymes catalyzing consecutive reactions in folate metabolism. We use deep mutational scanning to quantify the growth rate effect of 2696 DHFR single mutations in 3 TYMS backgrounds under conditions selected to emphasize biochemical epistasis. Our data are well-described by a relatively simple enzyme velocity to growth rate model that quantifies how metabolic context tunes enzyme mutational tolerance. Together our results reveal the structural distribution of epistasis in a metabolic enzyme and establish a foundation for the design of multi-enzyme systems.

Characterizing glucokinase variant mechanisms using a multiplexed abundance assay

BACKGROUND

Amino acid substitutions can perturb protein activity in multiple ways. Understanding their mechanistic basis may pinpoint how residues contribute to protein function. Here, we characterize the mechanisms underlying variant effects in human glucokinase (GCK) variants, building on our previous comprehensive study on GCK variant activity.

RESULTS

Using a yeast growth-based assay, we score the abundance of 95% of GCK missense and nonsense variants. When combining the abundance scores with our previously determined activity scores, we find that 43% of hypoactive variants also decrease cellular protein abundance. The low-abundance variants are enriched in the large domain, while residues in the small domain are tolerant to mutations with respect to abundance. Instead, many variants in the small domain perturb GCK conformational dynamics which are essential for appropriate activity.

CONCLUSIONS

In this study, we identify residues important for GCK metabolic stability and conformational dynamics. These residues could be targeted to modulate GCK activity, and thereby affect glucose homeostasis.

A map of the rubisco biochemical landscape

Rubisco is the primary CO2 fixing enzyme of the biosphere yet has slow kinetics. The roles of evolution and chemical mechanism in constraining the sequence landscape of rubisco remain debated. In order to map sequence to function, we developed a massively parallel assay for rubisco using an engineered E. coli where enzyme function is coupled to growth. By assaying >99% of single amino acid mutants across CO2 concentrations, we inferred enzyme velocity and CO2 affinity for thousands of substitutions. We identified many highly conserved positions that tolerate mutation and rare mutations that improve CO2 affinity. These data suggest that non-trivial kinetic improvements are readily accessible and provide a comprehensive sequence-to-function mapping for enzyme engineering efforts.

Machine learning approaches in predicting allosteric sites

Allosteric regulation is a fundamental biological mechanism that can control critical cellular processes via allosteric modulator binding to protein distal functional sites. The advantages of allosteric modulators over orthosteric ones have sparked the development of numerous computational approaches, such as the identification of allosteric binding sites, to facilitate allosteric drug discovery. Building on the success of machine learning (ML) models for solving complex problems in biology and chemistry, several ML models for predicting allosteric sites have been developed. In this review, we provide an overview of these models and discuss future perspectives powered by the field of artificial intelligence such as protein language models.

Cryptic genetic variation shapes the fate of gene duplicates in a protein interaction network

Paralogous genes are often redundant for long periods of time before they diverge in function. While their functions are preserved, paralogous proteins can accumulate mutations that, through epistasis, could impact their fate in the future. By quantifying the impact of all single-amino acid substitutions on the binding of two myosin proteins to their interaction partners, we find that the future evolution of these proteins is highly contingent on their regulatory divergence and the mutations that have silently accumulated in their protein binding domains. Differences in the promoter strength of the two paralogs amplify the impact of mutations on binding in the lowly expressed one. While some mutations would be sufficient to non-functionalize one paralog, they would have minimal impact on the other. Our results reveal how functionally equivalent protein domains could be destined to specific fates by regulatory and cryptic coding sequence changes that currently have little to no functional impact.

Mutational scanning pinpoints distinct binding sites of key ATGL regulators in lipolysis

ATGL is a key enzyme in intracellular lipolysis and plays an important role in metabolic and cardiovascular diseases. ATGL is tightly regulated by a known set of protein-protein interaction partners with activating or inhibiting functions in the control of lipolysis. Here, we use deep mutational protein interaction perturbation scanning and generate comprehensive profiles of single amino acid variants that affect the interactions of ATGL with its regulatory partners: CGI-58, G0S2, PLIN1, PLIN5 and CIDEC. Twenty-three ATGL amino acid variants yield a specific interaction perturbation pattern when validated in co-immunoprecipitation experiments in mammalian cells. We identify and characterize eleven highly selective ATGL switch mutations which affect the interaction of one of the five partners without affecting the others. Switch mutations thus provide distinct interaction determinants for ATGL's key regulatory proteins at an amino acid resolution. When we test triglyceride hydrolase activity in vitro and lipolysis in cells, the activity patterns of the ATGL switch variants trace to their protein interaction profile. In the context of structural data, the integration of variant binding and activity profiles provides insights into the regulation of lipolysis and the impact of mutations in human disease.

Biophysics-based protein language models for protein engineering

Protein language models trained on evolutionary data have emerged as powerful tools for predictive problems involving protein sequence, structure, and function. However, these models overlook decades of research into biophysical factors governing protein function. We propose Mutational Effect Transfer Learning (METL), a protein language model framework that unites advanced machine learning and biophysical modeling. Using the METL framework, we pretrain transformer-based neural networks on biophysical simulation data to capture fundamental relationships between protein sequence, structure, and energetics. We finetune METL on experimental sequence-function data to harness these biophysical signals and apply them when predicting protein properties like thermostability, catalytic activity, and fluorescence. METL excels in challenging protein engineering tasks like generalizing from small training sets and position extrapolation, although existing methods that train on evolutionary signals remain powerful for many types of experimental assays. We demonstrate METL's ability to design functional green fluorescent protein variants when trained on only 64 examples, showcasing the potential of biophysics-based protein language models for protein engineering.

A three-level regulatory mechanism of the aldo-keto reductase subfamily AKR12D

Modulation of protein function through allosteric regulation is central in biology, but biomacromolecular systems involving multiple subunits and ligands may exhibit complex regulatory mechanisms at different levels, which remain poorly understood. Here, we discover an aldo-keto reductase termed AKRtyl and present its three-level regulatory mechanism. Specifically, by combining steady-state and transient kinetics, X-ray crystallography and molecular dynamics simulation, we demonstrate that AKRtyl exhibits a positive synergy mediated by an unusual Monod-Wyman-Changeux (MWC) paradigm of allosteric regulation at low concentrations of the cofactor NADPH, but an inhibitory effect at high concentrations is observed. While the substrate tylosin binds at a remote allosteric site with positive cooperativity. We further reveal that these regulatory mechanisms are conserved in AKR12D subfamily, and that substrate cooperativity is common in AKRs across three kingdoms of life. This work provides an intriguing example for understanding complex allosteric regulatory networks.

Interpreting cis-regulatory mechanisms from genomic deep neural networks using surrogate models

Deep neural networks (DNNs) have greatly advanced the ability to predict genome function from sequence. Interpreting genomic DNNs in terms of biological mechanisms, however, remains difficult. Here we introduce SQUID, a genomic DNN interpretability framework based on surrogate modeling. SQUID approximates genomic DNNs in user-specified regions of sequence space using surrogate models, i.e., simpler models that are mechanistically interpretable. Importantly, SQUID removes the confounding effects that nonlinearities and heteroscedastic noise in functional genomics data can have on model interpretation. Benchmarking analysis on multiple genomic DNNs shows that SQUID, when compared to established interpretability methods, identifies motifs that are more consistent across genomic loci and yields improved single-nucleotide variant-effect predictions. SQUID also supports surrogate models that quantify epistatic interactions within and between cis-regulatory elements. SQUID thus advances the ability to mechanistically interpret genomic DNNs.

Opportunities and Challenges for Machine Learning-Assisted Enzyme Engineering

Enzymes can be engineered at the level of their amino acid sequences to optimize key properties such as expression, stability, substrate range, and catalytic efficiency-or even to unlock new catalytic activities not found in nature. Because the search space of possible proteins is vast, enzyme engineering usually involves discovering an enzyme starting point that has some level of the desired activity followed by directed evolution to improve its "fitness" for a desired application. Recently, machine learning (ML) has emerged as a powerful tool to complement this empirical process. ML models can contribute to (1) starting point discovery by functional annotation of known protein sequences or generating novel protein sequences with desired functions and (2) navigating protein fitness landscapes for fitness optimization by learning mappings between protein sequences and their associated fitness values. In this Outlook, we explain how ML complements enzyme engineering and discuss its future potential to unlock improved engineering outcomes.

Protein design using structure-based residue preferences

Recent developments in protein design rely on large neural networks with up to 100s of millions of parameters, yet it is unclear which residue dependencies are critical for determining protein function. Here, we show that amino acid preferences at individual residues-without accounting for mutation interactions-explain much and sometimes virtually all of the combinatorial mutation effects across 8 datasets (R2 ~ 78-98%). Hence, few observations (~100 times the number of mutated residues) enable accurate prediction of held-out variant effects (Pearson r > 0.80). We hypothesized that the local structural contexts around a residue could be sufficient to predict mutation preferences, and develop an unsupervised approach termed CoVES (Combinatorial Variant Effects from Structure). Our results suggest that CoVES outperforms not just model-free methods but also similarly to complex models for creating functional and diverse protein variants. CoVES offers an effective alternative to complicated models for identifying functional protein mutations.

Allosteric regulation in SARS-CoV-2 spike protein

Allosteric regulation is common in protein-protein interactions and is thus promising in drug design. Previous experimental and simulation work supported the presence of allosteric regulation in the SARS-CoV-2 spike protein. Here the route of allosteric regulation in SARS-CoV-2 spike protein is examined by all-atom explicit solvent molecular dynamics simulations, contrastive machine learning, and the Ohm approach. It was found that peptide binding to the polybasic cleavage sites, especially the one at the first subunit of the trimeric spike protein, activates the fluctuation of the spike protein's backbone, which eventually propagates to the receptor-binding domain on the third subunit that binds to ACE2. Remarkably, the allosteric regulation routes starting from the polybasic cleavage sites share a high fraction (39-67%) of the critical amino acids with the routes starting from the nitrogen-terminal domains, suggesting the presence of an allosteric regulation network in the spike protein. Our study paves the way for the rational design of allosteric antibody inhibitors.

Perspectives on Computational Enzyme Modeling: From Mechanisms to Design and Drug Development

Understanding enzyme mechanisms is essential for unraveling the complex molecular machinery of life. In this review, we survey the field of computational enzymology, highlighting key principles governing enzyme mechanisms and discussing ongoing challenges and promising advances. Over the years, computer simulations have become indispensable in the study of enzyme mechanisms, with the integration of experimental and computational exploration now established as a holistic approach to gain deep insights into enzymatic catalysis. Numerous studies have demonstrated the power of computer simulations in characterizing reaction pathways, transition states, substrate selectivity, product distribution, and dynamic conformational changes for various enzymes. Nevertheless, significant challenges remain in investigating the mechanisms of complex multistep reactions, large-scale conformational changes, and allosteric regulation. Beyond mechanistic studies, computational enzyme modeling has emerged as an essential tool for computer-aided enzyme design and the rational discovery of covalent drugs for targeted therapies. Overall, enzyme design/engineering and covalent drug development can greatly benefit from our understanding of the detailed mechanisms of enzymes, such as protein dynamics, entropy contributions, and allostery, as revealed by computational studies. Such a convergence of different research approaches is expected to continue, creating synergies in enzyme research. This review, by outlining the ever-expanding field of enzyme research, aims to provide guidance for future research directions and facilitate new developments in this important and evolving field.

A parametrized two-domain thermodynamic model explains diverse mutational effects on protein allostery

New experimental findings continue to challenge our understanding of protein allostery. Recent deep mutational scanning study showed that allosteric hotspots in the tetracycline repressor (TetR) and its homologous transcriptional factors are broadly distributed rather than spanning well-defined structural pathways as often assumed. Moreover, hotspot mutation-induced allostery loss was rescued by distributed additional mutations in a degenerate fashion. Here, we develop a two-domain thermodynamic model for TetR, which readily rationalizes these intriguing observations. The model accurately captures the in vivo activities of various mutants with changes in physically transparent parameters, allowing the data-based quantification of mutational effects using statistical inference. Our analysis reveals the intrinsic connection of intra- and inter-domain properties for allosteric regulation and illustrate epistatic interactions that are consistent with structural features of the protein. The insights gained from this study into the nature of two-domain allostery are expected to have broader implications for other multidomain allosteric proteins.

Inference of annealed protein fitness landscapes with AnnealDCA

The design of proteins with specific tasks is a major challenge in molecular biology with important diagnostic and therapeutic applications. High-throughput screening methods have been developed to systematically evaluate protein activity, but only a small fraction of possible protein variants can be tested using these techniques. Computational models that explore the sequence space in-silico to identify the fittest molecules for a given function are needed to overcome this limitation. In this article, we propose AnnealDCA, a machine-learning framework to learn the protein fitness landscape from sequencing data derived from a broad range of experiments that use selection and sequencing to quantify protein activity. We demonstrate the effectiveness of our method by applying it to antibody Rep-Seq data of immunized mice and screening experiments, assessing the quality of the fitness landscape reconstructions. Our method can be applied to several experimental cases where a population of protein variants undergoes various rounds of selection and sequencing, without relying on the computation of variants enrichment ratios, and thus can be used even in cases of disjoint sequence samples.

Allosteric drugs: New principles and design approaches

Focusing on an important biomedical implication of allostery - design of allosteric drugs, we describe characteristics of allosteric sites, effectors, and their modes of actions distinguishing them from the orthosteric counterparts and calling for new principles and protocols in the quests for allosteric drugs. We show the importance of considering both binding affinity and allosteric signaling in establishing the structure-activity relationships (SARs) toward design of allosteric effectors, arguing that pairs of allosteric sites and their effector ligands - the site-effector pairs - should be generated and adjusted simultaneously in the framework of what we call directed design protocol. Key ideas and approaches for designing allosteric effectors including reverse perturbation, targeted and agnostic analysis are also discussed here. Several promising computational approaches are highlighted, along with the need for and potential advantages of utilizing generative models to facilitate discovery/design of new allosteric drugs.

The energetic and allosteric landscape for KRAS inhibition

Thousands of proteins have been validated genetically as therapeutic targets for human diseases1. However, very few have been successfully targeted, and many are considered 'undruggable'. This is particularly true for proteins that function via protein-protein interactions-direct inhibition of binding interfaces is difficult and requires the identification of allosteric sites. However, most proteins have no known allosteric sites, and a comprehensive allosteric map does not exist for any protein. Here we address this shortcoming by charting multiple global atlases of inhibitory allosteric communication in KRAS. We quantified the effects of more than 26,000 mutations on the folding of KRAS and its binding to six interaction partners. Genetic interactions in double mutants enabled us to perform biophysical measurements at scale, inferring more than 22,000 causal free energy changes. These energy landscapes quantify how mutations tune the binding specificity of a signalling protein and map the inhibitory allosteric sites for an important therapeutic target. Allosteric propagation is particularly effective across the central β-sheet of KRAS, and multiple surface pockets are genetically validated as allosterically active, including a distal pocket in the C-terminal lobe of the protein. Allosteric mutations typically inhibit binding to all tested effectors, but they can also change the binding specificity, revealing the regulatory, evolutionary and therapeutic potential to tune pathway activation. Using the approach described here, it should be possible to rapidly and comprehensively identify allosteric target sites in many proteins.

Modulation of Allostery with Multiple Mechanisms by Hotspot Mutations in TetR

Modulating allosteric coupling offers unique opportunities for biomedical applications. Such efforts can benefit from efficient prediction and evaluation of allostery hotspot residues that dictate the degree of cooperativity between distant sites. We demonstrate that effects of allostery hotspot mutations can be evaluated qualitatively and semiquantitatively by molecular dynamics simulations in a bacterial tetracycline repressor (TetR). The simulations recapitulate the effects of these mutations on abolishing the induction function of TetR and provide a rationale for the different rescuabilities observed to restore allosteric coupling of the hotspot mutations. We demonstrate that the same noninducible phenotype could be the result of perturbations in distinct structural and energetic properties of TetR. Our work underscores the value of explicitly computing the functional free energy landscapes to effectively evaluate and rank hotspot mutations despite the prevalence of compensatory interactions and therefore provides quantitative guidance to allostery modulation for therapeutic and engineering applications.

ASD2023: towards the integrating landscapes of allosteric knowledgebase

Allosteric regulation, induced by perturbations at an allosteric site topographically distinct from the orthosteric site, is one of the most direct and efficient ways to fine-tune macromolecular function. The Allosteric Database (ASD; accessible online at http://mdl.shsmu.edu.cn/ASD) has been systematically developed since 2009 to provide comprehensive information on allosteric regulation. In recent years, allostery has seen sustained growth and wide-ranging applications in life sciences, from basic research to new therapeutics development, while also elucidating emerging obstacles across allosteric research stages. To overcome these challenges and maintain high-quality data center services, novel features were curated in the ASD2023 update: (i) 66 589 potential allosteric sites, covering > 80% of the human proteome and constituting the human allosteric pocketome; (ii) 748 allosteric protein-protein interaction (PPI) modulators with clear mechanisms, aiding protein machine studies and PPI-targeted drug discovery; (iii) 'Allosteric Hit-to-Lead,' a pioneering dataset providing panoramic views from 87 well-defined allosteric hits to 6565 leads and (iv) 456 dualsteric modulators for exploring the simultaneous regulation of allosteric and orthosteric sites. Meanwhile, ASD2023 maintains a significant growth of foundational allosteric data. Based on these efforts, the allosteric knowledgebase is progressively evolving towards an integrated landscape, facilitating advancements in allosteric target identification, mechanistic exploration and drug discovery.

ACIDES: on-line monitoring of forward genetic screens for protein engineering

Forward genetic screens of mutated variants are a versatile strategy for protein engineering and investigation, which has been successfully applied to various studies like directed evolution (DE) and deep mutational scanning (DMS). While next-generation sequencing can track millions of variants during the screening rounds, the vast and noisy nature of the sequencing data impedes the estimation of the performance of individual variants. Here, we propose ACIDES that combines statistical inference and in-silico simulations to improve performance estimation in the library selection process by attributing accurate statistical scores to individual variants. We tested ACIDES first on a random-peptide-insertion experiment and then on multiple public datasets from DE and DMS studies. ACIDES allows experimentalists to reliably estimate variant performance on the fly and can aid protein engineering and research pipelines in a range of applications, including gene therapy.

Local energetic frustration conservation in protein families and superfamilies

Energetic local frustration offers a biophysical perspective to interpret the effects of sequence variability on protein families. Here we present a methodology to analyze local frustration patterns within protein families and superfamilies that allows us to uncover constraints related to stability and function, and identify differential frustration patterns in families with a common ancestry. We analyze these signals in very well studied protein families such as PDZ, SH3, ɑ and β globins and RAS families. Recent advances in protein structure prediction make it possible to analyze a vast majority of the protein space. An automatic and unsupervised proteome-wide analysis on the SARS-CoV-2 virus demonstrates the potential of our approach to enhance our understanding of the natural phenotypic diversity of protein families beyond single protein instances. We apply our method to modify biophysical properties of natural proteins based on their family properties, as well as perform unsupervised analysis of large datasets to shed light on the physicochemical signatures of poorly characterized proteins such as the ones belonging to emergent pathogens.

Modulation of Allostery with Multiple Mechanisms by Hotspot Mutations in TetR

Modulating allosteric coupling offers unique opportunities for biomedical applications. Such efforts can benefit from efficient prediction and evaluation of allostery hotspot residues that dictate the degree of co-operativity between distant sites. We demonstrate that effects of allostery hotspot mutations can be evaluated qualitatively and semi-quantitatively by molecular dynamics simulations in a bacterial tetracycline repressor (TetR). The simulations recapitulate the effects of these mutations on abolishing the induction function of TetR and provide a rationale for the different degrees of rescuability observed to restore allosteric coupling of the hotspot mutations. We demonstrate that the same non-inducible phenotype could be the result of perturbations in distinct structural and energetic properties of TetR. Our work underscore the value of explicitly computing the functional free energy landscapes to effectively evaluate and rank hotspot mutations despite the prevalence of compensatory interactions, and therefore provide quantitative guidance to allostery modulation for therapeutic and engineering applications.

ProteinGym: Large-Scale Benchmarks for Protein Design and Fitness Prediction

Predicting the effects of mutations in proteins is critical to many applications, from understanding genetic disease to designing novel proteins that can address our most pressing challenges in climate, agriculture and healthcare. Despite a surge in machine learning-based protein models to tackle these questions, an assessment of their respective benefits is challenging due to the use of distinct, often contrived, experimental datasets, and the variable performance of models across different protein families. Addressing these challenges requires scale. To that end we introduce ProteinGym, a large-scale and holistic set of benchmarks specifically designed for protein fitness prediction and design. It encompasses both a broad collection of over 250 standardized deep mutational scanning assays, spanning millions of mutated sequences, as well as curated clinical datasets providing high-quality expert annotations about mutation effects. We devise a robust evaluation framework that combines metrics for both fitness prediction and design, factors in known limitations of the underlying experimental methods, and covers both zero-shot and supervised settings. We report the performance of a diverse set of over 70 high-performing models from various subfields (eg., alignment-based, inverse folding) into a unified benchmark suite. We open source the corresponding codebase, datasets, MSAs, structures, model predictions and develop a user-friendly website that facilitates data access and analysis.

Interpretable neural architecture search and transfer learning for understanding CRISPR-Cas9 off-target enzymatic reactions

Finely tuned enzymatic pathways control cellular processes, and their dysregulation can lead to disease. Developing predictive and interpretable models for these pathways is challenging because of the complexity of the pathways and of the cellular and genomic contexts. Here we introduce Elektrum, a deep learning framework that addresses these challenges with data-driven and biophysically interpretable models for determining the kinetics of biochemical systems. First, it uses in vitro kinetic assays to rapidly hypothesize an ensemble of high-quality kinetically interpretable neural networks (KINNs) that predict reaction rates. It then employs a transfer learning step, where the KINNs are inserted as intermediary layers into deeper convolutional neural networks, fine-tuning the predictions for reaction-dependent in vivo outcomes. We apply Elektrum to predict CRISPR-Cas9 off-target editing probabilities and demonstrate that Elektrum achieves improved performance, regularizes neural network architectures and maintains physical interpretability.

A rugged yet easily navigable fitness landscape

Fitness landscape theory predicts that rugged landscapes with multiple peaks impair Darwinian evolution, but experimental evidence is limited. In this study, we used genome editing to map the fitness of >260,000 genotypes of the key metabolic enzyme dihydrofolate reductase in the presence of the antibiotic trimethoprim, which targets this enzyme. The resulting landscape is highly rugged and harbors 514 fitness peaks. However, its highest peaks are accessible to evolving populations via abundant fitness-increasing paths. Different peaks share large basins of attraction that render the outcome of adaptive evolution highly contingent on chance events. Our work shows that ruggedness need not be an obstacle to Darwinian evolution but can reduce its predictability. If true in general, the complexity of optimization problems on realistic landscapes may require reappraisal.

Integrated multiplexed assays of variant effect reveal cis-regulatory determinants of catechol-O-methyltransferase gene expression

Multiplexed assays of variant effect are powerful methods to profile the consequences of rare variants on gene expression and organismal fitness. Yet, few studies have integrated several multiplexed assays to map variant effects on gene expression in coding sequences. Here, we pioneered a multiplexed assay based on polysome profiling to measure variant effects on translation at scale, uncovering single-nucleotide variants that increase and decrease ribosome load. By combining high-throughput ribosome load data with multiplexed mRNA and protein abundance readouts, we mapped the cis-regulatory landscape of thousands of catechol-O-methyltransferase (COMT) variants from RNA to protein and found numerous coding variants that alter COMT expression. Finally, we trained machine learning models to map signatures of variant effects on COMT gene expression and uncovered both directional and divergent impacts across expression layers. Our analyses reveal expression phenotypes for thousands of variants in COMT and highlight variant effects on both single and multiple layers of expression. Our findings prompt future studies that integrate several multiplexed assays for the readout of gene expression.

Dominance vs epistasis: the biophysical origins and plasticity of genetic interactions within and between alleles

An important challenge in genetics, evolution and biotechnology is to understand and predict how mutations combine to alter phenotypes, including molecular activities, fitness and disease. In diploids, mutations in a gene can combine on the same chromosome or on different chromosomes as a "heteroallelic combination". However, a direct comparison of the extent, sign, and stability of the genetic interactions between variants within and between alleles is lacking. Here we use thermodynamic models of protein folding and ligand-binding to show that interactions between mutations within and between alleles are expected in even very simple biophysical systems. Protein folding alone generates within-allele interactions and a single molecular interaction is sufficient to cause between-allele interactions and dominance. These interactions change differently, quantitatively and qualitatively as a system becomes more complex. Altering the concentration of a ligand can, for example, switch alleles from dominant to recessive. Our results show that intra-molecular epistasis and dominance should be widely expected in even the simplest biological systems but also reinforce the view that they are plastic system properties and so a formidable challenge to predict. Accurate prediction of both intra-molecular epistasis and dominance will require either detailed mechanistic understanding and experimental parameterization or brute-force measurement and learning.

The Genetic Landscape of a Metabolic Interaction

Enzyme abundance, catalytic activity, and ultimately sequence are all shaped by the need of growing cells to maintain metabolic flux while minimizing accumulation of deleterious intermediates. While much prior work has explored the constraints on protein sequence and evolution induced by physical protein-protein interactions, the sequence-level constraints emerging from non-binding functional interactions in metabolism remain unclear. To quantify how variation in the activity of one enzyme constrains the biochemical parameters and sequence of another, we focused on dihydrofolate reductase (DHFR) and thymidylate synthase (TYMS), a pair of enzymes catalyzing consecutive reactions in folate metabolism. We used deep mutational scanning to quantify the growth rate effect of 2,696 DHFR single mutations in 3 TYMS backgrounds under conditions selected to emphasize biochemical epistasis. Our data are well-described by a relatively simple enzyme velocity to growth rate model that quantifies how metabolic context tunes enzyme mutational tolerance. Together our results reveal the structural distribution of epistasis in a metabolic enzyme and establish a foundation for the design of multi-enzyme systems.

Jointly modeling deep mutational scans identifies shifted mutational effects among SARS-CoV-2 spike homologs

Deep mutational scanning (DMS) is a high-throughput experimental technique that measures the effects of thousands of mutations to a protein. These experiments can be performed on multiple homologs of a protein or on the same protein selected under multiple conditions. It is often of biological interest to identify mutations with shifted effects across homologs or conditions. However, it is challenging to determine if observed shifts arise from biological signal or experimental noise. Here, we describe a method for jointly inferring mutational effects across multiple DMS experiments while also identifying mutations that have shifted in their effects among experiments. A key aspect of our method is to regularize the inferred shifts, so that they are nonzero only when strongly supported by the data. We apply this method to DMS experiments that measure how mutations to spike proteins from SARS-CoV-2 variants (Delta, Omicron BA.1, and Omicron BA.2) affect cell entry. Most mutational effects are conserved between these spike homologs, but a fraction have markedly shifted. We experimentally validate a subset of the mutations inferred to have shifted effects, and confirm differences of > 1,000-fold in the impact of the same mutation on spike-mediated viral infection across spikes from different SARS-CoV-2 variants. Overall, our work establishes a general approach for comparing sets of DMS experiments to identify biologically important shifts in mutational effects.

Missense variant interaction scanning reveals a critical role of the FERM domain for tumor suppressor protein NF2 conformation and function

NF2 (moesin-ezrin-radixin-like [MERLIN] tumor suppressor) is frequently inactivated in cancer, where its NF2 tumor suppressor functionality is tightly coupled to protein conformation. How NF2 conformation is regulated and how NF2 conformation influences tumor suppressor activity is a largely open question. Here, we systematically characterized three NF2 conformation-dependent protein interactions utilizing deep mutational scanning interaction perturbation analyses. We identified two regions in NF2 with clustered mutations which affected conformation-dependent protein interactions. NF2 variants in the F2-F3 subdomain and the α3H helix region substantially modulated NF2 conformation and homomerization. Mutations in the F2-F3 subdomain altered proliferation in three cell lines and matched patterns of disease mutations in NF2 related-schwannomatosis. This study highlights the power of systematic mutational interaction perturbation analysis to identify missense variants impacting NF2 conformation and provides insight into NF2 tumor suppressor function.

Discovering functionally important sites in proteins

Proteins play important roles in biology, biotechnology and pharmacology, and missense variants are a common cause of disease. Discovering functionally important sites in proteins is a central but difficult problem because of the lack of large, systematic data sets. Sequence conservation can highlight residues that are functionally important but is often convoluted with a signal for preserving structural stability. We here present a machine learning method to predict functional sites by combining statistical models for protein sequences with biophysical models of stability. We train the model using multiplexed experimental data on variant effects and validate it broadly. We show how the model can be used to discover active sites, as well as regulatory and binding sites. We illustrate the utility of the model by prospective prediction and subsequent experimental validation on the functional consequences of missense variants in HPRT1 which may cause Lesch-Nyhan syndrome, and pinpoint the molecular mechanisms by which they cause disease.

SUNi mutagenesis: Scalable and uniform nicking for efficient generation of variant libraries

Multiplexed assays of variant effects (MAVEs) have made possible the functional assessment of all possible mutations to genes and regulatory sequences. A core pillar of the approach is generation of variant libraries, but current methods are either difficult to scale or not uniform enough to enable MAVEs at the scale of gene families or beyond. We present an improved method called Scalable and Uniform Nicking (SUNi) mutagenesis that combines massive scalability with high uniformity to enable cost-effective MAVEs of gene families and eventually genomes.