From Bioorganic Models to Cells.

I endeavor to share how various choices-some deliberate, some unconscious-and the unmistakable influence of many others shaped my scientific pursuits. I am fascinated by how two long-term, major streams of my research, DNA replication and purine biosynthesis, have merged with unexpected interconnections. If I have imparted to many of the talented individuals who have passed through my lab a degree of my passion for uncloaking the mysteries hidden in scientific research and an understanding of the honesty and rigor it demands and its impact on the world community, then my mentorship has been successful.

Moving Through Barriers in Science and Life.

This first serious attempt at an autobiographical accounting has forced me to sit still long enough to compile my thoughts about a long personal and scientific journey. I especially hope that my trajectory will be of interest and perhaps beneficial to much younger women who are just getting started in their careers. To paraphrase from Virginia Woolf's writings in A Room of One's Own at the beginning of the 20th century, "for most of history Anonymous was a Woman." However, Ms. Woolf is also quoted as saying "nothing has really happened until it has been described," a harbinger of the enormous historical changes that were about to be enacted and recorded by women in the sciences and other disciplines. The progress in my chosen field of study-the chemical basis of enzyme action-has also been remarkable, from the first description of an enzyme's 3D structure to a growing and deep understanding of the origins of enzyme catalysis.

Metabolite-Enzyme Coevolution: From Single Enzymes to Metabolic Pathways and Networks.

How individual enzymes evolved is relatively well understood. However, individual enzymes rarely confer a physiological advantage on their own. Judging by its current state, the emergence of metabolism seemingly demanded the simultaneous emergence of many enzymes. Indeed, how multicomponent interlocked systems, like metabolic pathways, evolved is largely an open question. This complexity can be unlocked if we assume that survival of the fittest applies not only to genes and enzymes but also to the metabolites they produce. This review develops our current knowledge of enzyme evolution into a wider hypothesis of pathway and network evolution. We describe the current models for pathway evolution and offer an integrative metabolite-enzyme coevolution hypothesis. Our hypothesis addresses the origins of new metabolites and of new enzymes and the order of their recruitment. We aim to not only survey established knowledge but also present open questions and potential ways of addressing them.

Protein Evolution and Design.

This article introduces the Protein Evolution and Design theme of the Annual Review of Biochemistry Volume 87.

A Rich Man, Poor Man Story of S-Adenosylmethionine and Cobalamin Revisited.

S-adenosylmethionine (AdoMet) has been referred to as both "a poor man's adenosylcobalamin (AdoCbl)" and "a rich man's AdoCbl," but today, with the ever-increasing number of functions attributed to each cofactor, both appear equally rich and surprising. The recent characterization of an organometallic species in an AdoMet radical enzyme suggests that the line that differentiates them in nature will be constantly challenged. Here, we compare and contrast AdoMet and cobalamin (Cbl) and consider why Cbl-dependent AdoMet radical enzymes require two cofactors that are so similar in their reactivity. We further carry out structural comparisons employing the recently determined crystal structure of oxetanocin-A biosynthetic enzyme OxsB, the first three-dimensional structural data on a Cbl-dependent AdoMet radical enzyme. We find that the structural motifs responsible for housing the AdoMet radical machinery are largely conserved, whereas the motifs responsible for binding additional cofactors are much more varied.

Directed Evolution of Protein Catalysts.

Directed evolution is a powerful technique for generating tailor-made enzymes for a wide range of biocatalytic applications. Following the principles of natural evolution, iterative cycles of mutagenesis and screening or selection are applied to modify protein properties, enhance catalytic activities, or develop completely new protein catalysts for non-natural chemical transformations. This review briefly surveys the experimental methods used to generate genetic diversity and screen or select for improved enzyme variants. Emphasis is placed on a key challenge, namely how to generate novel catalytic activities that expand the scope of natural reactions. Two particularly effective strategies, exploiting catalytic promiscuity and rational design, are illustrated by representative examples of successfully evolved enzymes. Opportunities for extending these approaches to more complex biocatalytic systems are also considered.

Evolutionary Convergence of Pathway-Specific Enzyme Expression Stoichiometry.

Coexpression of proteins in response to pathway-inducing signals is the founding paradigm of gene regulation. However, it remains unexplored whether the relative abundance of co-regulated proteins requires precise tuning. Here, we present large-scale analyses of protein stoichiometry and corresponding regulatory strategies for 21 pathways and 67-224 operons in divergent bacteria separated by 0.6-2 billion years. Using end-enriched RNA-sequencing (Rend-seq) with single-nucleotide resolution, we found that many bacterial gene clusters encoding conserved pathways have undergone massive divergence in transcript abundance and architectures via remodeling of internal promoters and terminators. Remarkably, these evolutionary changes are compensated post-transcriptionally to maintain preferred stoichiometry of protein synthesis rates. Even more strikingly, in eukaryotic budding yeast, functionally analogous proteins that arose independently from bacterial counterparts also evolved to convergent in-pathway expression. The broad requirement for exact protein stoichiometries despite regulatory divergence provides an unexpected principle for building biological pathways both in nature and for synthetic activities.

Expanding chemistry through in vitro and in vivo biocatalysis.

Living systems contain a vast network of metabolic reactions, providing a wealth of enzymes and cells as potential biocatalysts for chemical processes. The properties of protein and cell biocatalysts-high selectivity, the ability to control reaction sequence and operation in environmentally benign conditions-offer approaches to produce molecules at high efficiency while lowering the cost and environmental impact of industrial chemistry. Furthermore, biocatalysis offers the opportunity to generate chemical structures and functions that may be inaccessible to chemical synthesis. Here we consider developments in enzymes, biosynthetic pathways and cellular engineering that enable their use in catalysis for new chemistry and beyond.

Boron catalysis in a designer enzyme.

Enzymes play an increasingly important role in improving the benignity and efficiency of chemical production, yet the diversity of their applications lags heavily behind chemical catalysts as a result of the relatively narrow range of reaction mechanisms of enzymes. The creation of enzymes containing non-biological functionalities facilitates reaction mechanisms outside nature's canon and paves the way towards fully programmable biocatalysis1-3. Here we present a completely genetically encoded boronic-acid-containing designer enzyme with organocatalytic reactivity not achievable with natural or engineered biocatalysts4,5. This boron enzyme catalyses the kinetic resolution of hydroxyketones by oxime formation, in which crucial interactions with the protein scaffold assist in the catalysis. A directed evolution campaign led to a variant with natural-enzyme-like enantioselectivities for several different substrates. The unique activation mode of the boron enzyme was confirmed using X-ray crystallography, high-resolution mass spectrometry (HRMS) and 11B NMR spectroscopy. Our study demonstrates that genetic-code expansion can be used to create evolvable enantioselective enzymes that rely on xenobiotic catalytic moieties such as boronic acids and access reaction mechanisms not reachable through catalytic promiscuity of natural or engineered enzymes.

Design of protein catalysts.

Diverse engineering strategies have been developed to create enzymes with novel catalytic activities. Among these, computational approaches hold particular promise. Enzymes have been computationally designed to promote several nonbiological reactions, including a Diels-Alder cycloaddition, proton transfer, multistep retroaldol transformations, and metal-dependent hydrolysis of phosphotriesters. Although their efficiencies (kcat/KM = 0.1-100 M(-1) s(-1)) are typically low compared with those of the best natural enzymes (10(6)-10(8) M(-1) s(-1)), these catalysts are excellent starting points for laboratory evolution. This review surveys recent progress in combining computational and evolutionary approaches to enzyme design, together with insights into enzyme function gained from studies of the engineered catalysts.

Hydrogen tunneling links protein dynamics to enzyme catalysis.

The relationship between protein dynamics and function is a subject of considerable contemporary interest. Although protein motions are frequently observed during ligand binding and release steps, the contribution of protein motions to the catalysis of bond making/breaking processes is more difficult to probe and verify. Here, we show how the quantum mechanical hydrogen tunneling associated with enzymatic C-H bond cleavage provides a unique window into the necessity of protein dynamics for achieving optimal catalysis. Experimental findings support a hierarchy of thermodynamically equilibrated motions that control the H-donor and -acceptor distance and active-site electrostatics, creating an ensemble of conformations suitable for H-tunneling. A possible extension of this view to methyl transfer and other catalyzed reactions is also presented. The impact of understanding these dynamics on the conceptual framework for enzyme activity, inhibitor/drug design, and biomimetic catalyst design is likely to be substantial.

A designed photoenzyme for enantioselective [2+2] cycloadditions.

The ability to program new modes of catalysis into proteins would allow the development of enzyme families with functions beyond those found in nature. To this end, genetic code expansion methodology holds particular promise, as it allows the site-selective introduction of new functional elements into proteins as noncanonical amino acid side chains1-4. Here we exploit an expanded genetic code to develop a photoenzyme that operates by means of triplet energy transfer (EnT) catalysis, a versatile mode of reactivity in organic synthesis that is not accessible to biocatalysis at present5-12. Installation of a genetically encoded photosensitizer into the beta-propeller scaffold of DA_20_00 (ref. 13) converts a de novo Diels-Alderase into a photoenzyme for [2+2] cycloadditions (EnT1.0). Subsequent development and implementation of a platform for photoenzyme evolution afforded an efficient and enantioselective enzyme (EnT1.3, up to 99% enantiomeric excess (e.e.)) that can promote intramolecular and bimolecular cycloadditions, including transformations that have proved challenging to achieve selectively with small-molecule catalysts. EnT1.3 performs >300 turnovers and, in contrast to small-molecule photocatalysts, can operate effectively under aerobic conditions and at ambient temperatures. An X-ray crystal structure of an EnT1.3-product complex shows how multiple functional components work in synergy to promote efficient and selective photocatalysis. This study opens up a wealth of new excited-state chemistry in protein active sites and establishes the framework for developing a new generation of enantioselective photocatalysts.

Benchmark reaction rates, the stability of biological molecules in water, and the evolution of catalytic power in enzymes.

The rates of enzyme reactions fall within a relatively narrow range. To estimate the rate enhancements produced by enzymes, and their expected affinities for transition state analog inhibitors, it is necessary to measure the rates of the corresponding reactions in water in the absence of a catalyst. This review describes the spontaneous cleavages of C-C, C-H, C-N, C-O, P-O, and S-O bonds in biological molecules, as well as the uncatalyzed reactions that correspond to phosphoryl transfer reactions catalyzed by kinases and to peptidyl transfer in the ribosome. The rates of these reactions, some with half-lives in excess of one million years, span an overall range of 10¹9-fold. Moreover, the slowest reactions tend to be most sensitive to temperature, with rates that increase as much as 107-fold when the temperature is raised from 25° to 100°C. That tendency collapses, by many orders of magnitude, the time that would have been required for chemical evolution on a warm earth. If the catalytic effect of primitive enzymes, like that of modern enzymes and many nonenzymatic catalysts, were mainly to reduce a reaction's enthalpy of activation, then the resulting rate enhancement would have increased automatically as the surroundings cooled. By reducing the time required for early chemical evolution in a warm environment, these findings counter the view that not enough time has passed for terrestrial life to have evolved to its present level of complexity.

Enzymatic transition states, transition-state analogs, dynamics, thermodynamics, and lifetimes.

Experimental analysis of enzymatic transition-state structures uses kinetic isotope effects (KIEs) to report on bonding and geometry differences between reactants and the transition state. Computational correlation of experimental values with chemical models permits three-dimensional geometric and electrostatic assignment of transition states formed at enzymatic catalytic sites. The combination of experimental and computational access to transition-state information permits (a) the design of transition-state analogs as powerful enzymatic inhibitors, (b) exploration of protein features linked to transition-state structure, (c) analysis of ensemble atomic motions involved in achieving the transition state, (d) transition-state lifetimes, and (e) separation of ground-state (Michaelis complexes) from transition-state effects. Transition-state analogs with picomolar dissociation constants have been achieved for several enzymatic targets. Transition states of closely related isozymes indicate that the protein's dynamic architecture is linked to transition-state structure. Fast dynamic motions in catalytic sites are linked to transition-state generation. Enzymatic transition states have lifetimes of femtoseconds, the lifetime of bond vibrations. Binding isotope effects (BIEs) reveal relative reactant and transition-state analog binding distortion for comparison with actual transition states.

GloEC: a hierarchical-aware global model for predicting enzyme function.

The annotation of enzyme function is a fundamental challenge in industrial biotechnology and pathologies. Numerous computational methods have been proposed to predict enzyme function by annotating enzyme labels with Enzyme Commission number. However, the existing methods face difficulties in modelling the hierarchical structure of enzyme label in a global view. Moreover, they haven't gone entirely to leverage the mutual interactions between different levels of enzyme label. In this paper, we formulate the hierarchy of enzyme label as a directed enzyme graph and propose a hierarchy-GCN (Graph Convolutional Network) encoder to globally model enzyme label dependency on the enzyme graph. Based on the enzyme hierarchy encoder, we develop an end-to-end hierarchical-aware global model named GloEC to predict enzyme function. GloEC learns hierarchical-aware enzyme label embeddings via the hierarchy-GCN encoder and conducts deductive fusion of label-aware enzyme features to predict enzyme labels. Meanwhile, our hierarchy-GCN encoder is designed to bidirectionally compute to investigate the enzyme label correlation information in both bottom-up and top-down manners, which has not been explored in enzyme function prediction. Comparative experiments on three benchmark datasets show that GloEC achieves better predictive performance as compared to the existing methods. The case studies also demonstrate that GloEC is capable of effectively predicting the function of isoenzyme. GloEC is available at: https://github.com/hyr0771/GloEC.

MPEK: a multitask deep learning framework based on pretrained language models for enzymatic reaction kinetic parameters prediction.

Enzymatic reaction kinetics are central in analyzing enzymatic reaction mechanisms and target-enzyme optimization, and thus in biomanufacturing and other industries. The enzyme turnover number (kcat) and Michaelis constant (Km), key kinetic parameters for measuring enzyme catalytic efficiency, are crucial for analyzing enzymatic reaction mechanisms and the directed evolution of target enzymes. Experimental determination of kcat and Km is costly in terms of time, labor, and cost. To consider the intrinsic connection between kcat and Km and further improve the prediction performance, we propose a universal pretrained multitask deep learning model, MPEK, to predict these parameters simultaneously while considering pH, temperature, and organismal information. Through testing on the same kcat and Km test datasets, MPEK demonstrated superior prediction performance over the previous models. Specifically, MPEK achieved the Pearson coefficient of 0.808 for predicting kcat, improving ca. 14.6% and 7.6% compared to the DLKcat and UniKP models, and it achieved the Pearson coefficient of 0.777 for predicting Km, improving ca. 34.9% and 53.3% compared to the Kroll_model and UniKP models. More importantly, MPEK was able to reveal enzyme promiscuity and was sensitive to slight changes in the mutant enzyme sequence. In addition, in three case studies, it was shown that MPEK has the potential for assisted enzyme mining and directed evolution. To facilitate in silico evaluation of enzyme catalytic efficiency, we have established a web server implementing this model, which can be accessed at http://mathtc.nscc-tj.cn/mpek.

ifDEEPre: large protein language-based deep learning enables interpretable and fast predictions of enzyme commission numbers.

Accurate understanding of the biological functions of enzymes is vital for various tasks in both pathologies and industrial biotechnology. However, the existing methods are usually not fast enough and lack explanations on the prediction results, which severely limits their real-world applications. Following our previous work, DEEPre, we propose a new interpretable and fast version (ifDEEPre) by designing novel self-guided attention and incorporating biological knowledge learned via large protein language models to accurately predict the commission numbers of enzymes and confirm their functions. Novel self-guided attention is designed to optimize the unique contributions of representations, automatically detecting key protein motifs to provide meaningful interpretations. Representations learned from raw protein sequences are strictly screened to improve the running speed of the framework, 50 times faster than DEEPre while requiring 12.89 times smaller storage space. Large language modules are incorporated to learn physical properties from hundreds of millions of proteins, extending biological knowledge of the whole network. Extensive experiments indicate that ifDEEPre outperforms all the current methods, achieving more than 14.22% larger F1-score on the NEW dataset. Furthermore, the trained ifDEEPre models accurately capture multi-level protein biological patterns and infer evolutionary trends of enzymes by taking only raw sequences without label information. Meanwhile, ifDEEPre predicts the evolutionary relationships between different yeast sub-species, which are highly consistent with the ground truth. Case studies indicate that ifDEEPre can detect key amino acid motifs, which have important implications for designing novel enzymes. A web server running ifDEEPre is available at https://proj.cse.cuhk.edu.hk/aihlab/ifdeepre/ to provide convenient services to the public. Meanwhile, ifDEEPre is freely available on GitHub at https://github.com/ml4bio/ifDEEPre/.

Noncanonical Amino Acids in Biocatalysis.

In recent years, powerful genetic code reprogramming methods have emerged that allow new functional components to be embedded into proteins as noncanonical amino acid (ncAA) side chains. In this review, we will illustrate how the availability of an expanded set of amino acid building blocks has opened a wealth of new opportunities in enzymology and biocatalysis research. Genetic code reprogramming has provided new insights into enzyme mechanisms by allowing introduction of new spectroscopic probes and the targeted replacement of individual atoms or functional groups. NcAAs have also been used to develop engineered biocatalysts with improved activity, selectivity, and stability, as well as enzymes with artificial regulatory elements that are responsive to external stimuli. Perhaps most ambitiously, the combination of genetic code reprogramming and laboratory evolution has given rise to new classes of enzymes that use ncAAs as key catalytic elements. With the framework for developing ncAA-containing biocatalysts now firmly established, we are optimistic that genetic code reprogramming will become a progressively more powerful tool in the armory of enzyme designers and engineers in the coming years.

Light-driven post-translational installation of reactive protein side chains.

Post-translational modifications (PTMs) greatly expand the structures and functions of proteins in nature1,2. Although synthetic protein functionalization strategies allow mimicry of PTMs3,4, as well as formation of unnatural protein variants with diverse potential functions, including drug carrying5, tracking, imaging6 and partner crosslinking7, the range of functional groups that can be introduced remains limited. Here we describe the visible-light-driven installation of side chains at dehydroalanine residues in proteins through the formation of carbon-centred radicals that allow C-C bond formation in water. Control of the reaction redox allows site-selective modification with good conversions and reduced protein damage. In situ generation of boronic acid catechol ester derivatives generates RH2C• radicals that form the native (ß-CH2-γ-CH2) linkage of natural residues and PTMs, whereas in situ potentiation of pyridylsulfonyl derivatives by Fe(II) generates RF2C• radicals that form equivalent ß-CH2-γ-CF2 linkages bearing difluoromethylene labels. These reactions are chemically tolerant and incorporate a wide range of functionalities (more than 50 unique residues/side chains) into diverse protein scaffolds and sites. Initiation can be applied chemoselectively in the presence of sensitive groups in the radical precursors, enabling installation of previously incompatible side chains. The resulting protein function and reactivity are used to install radical precursors for homolytic on-protein radical generation; to study enzyme function with natural, unnatural and CF2-labelled post-translationally modified protein substrates via simultaneous sensing of both chemo- and stereoselectivity; and to create generalized 'alkylator proteins' with a spectrum of heterolytic covalent-bond-forming activity (that is, reacting diversely with small molecules at one extreme or selectively with protein targets through good mimicry at the other). Post-translational access to such reactions and chemical groups on proteins could be useful in both revealing and creating protein function.