Explore the vast range of titles available.

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.

The ability to routinely generate efficient protein catalysts of bond-forming reactions chosen by researchers, rather than nature, is a long-standing goal of the molecular life sciences. Here, we describe a directed evolution strategy for enzymes that catalyze, in principle, any bond-forming reaction. The system integrates yeast display, enzyme-mediated bioconjugation, and fluorescence-activated cell sorting to isolate cells expressing proteins that catalyze the coupling of two substrates chosen by the researcher. We validated the system using model screens for Staphylococcus aureus sortase A—catalyzed transpeptidation activity, resulting in enrichment factors of 6,000-fold after a single round of screening. We applied the system to evolve sortase A for improved catalytic activity. After eight rounds of screening, we isolated variants of sortase A with up to a 140-fold increase in LPETG-coupling activity compared with the starting wild-type enzyme. An evolved sortase variant enabled much more efficient labeling of LPETG-tagged human CD154 expressed on the surface of HeLa cells compared with wild-type sortase. Because the method developed here does not rely on any particular screenable or selectable property of the substrates or product, it represents a powerful alternative to existing enzyme evolution methods.

To reduce experimental effort associated with directed protein evolution and to explore the sequence space encoded by mutating multiple positions simultaneously, we incorporate machine learning into the directed evolution workflow. Combinatorial sequence space can be quite expensive to sample experimentally, but machine-learning models trained on tested variants provide a fast method for testing sequence space computationally. We validated this approach on a large published empirical fitness landscape for human GB1 binding protein, demonstrating that machine learning-guided directed evolution finds variants with higher fitness than those found by other directed evolution approaches. We then provide an example application in evolving an enzyme to produce each of the two possible product enantiomers (i.e., stereodivergence) of a new-to-nature carbene Si–H insertion reaction. The approach predicted libraries enriched in functional enzymes and fixed seven mutations in two rounds of evolution to identify variants for selective catalysis with 93% and 79% ee (enantiomeric excess). By greatly increasing throughput with in silico modeling, machine learning enhances the quality and diversity of sequence solutions for a protein engineering problem.

Continuous directed evolution methods allow the key steps of evolution—gene diversification, selection, and replication—to proceed in the laboratory with minimal researcher intervention. As a result, continuous evolution can find solutions much more quickly than traditional discrete evolution methods. Continuous evolution also enables the exploration of longer and more numerous evolutionary trajectories, increasing the likelihood of accessing solutions that require many steps through sequence space and greatly facilitating the iterative refinement of selection conditions and targeted mutagenesis strategies. Here we review the historical advances that have expanded continuous evolution from its earliest days as an experimental curiosity to its present state as a powerful and surprisingly general strategy for generating tailor-made biomolecules, and discuss more recent improvements with an eye to the future. This historical Perspective on continuous directed evolution focuses on laboratory approaches that enable greater understanding of evolving molecular populations and offer investigators tools to guide the emergence of new biomolecular systems.

The use of CRISPR-Cas9 as a therapeutic reagent is hampered by its off-target effects. Although rationally designed S. pyogenes Cas9 (SpCas9) variants that display higher specificities than the wild-type SpCas9 protein are available, these attenuated Cas9 variants are often poorly efficient in human cells. Here, we develop a directed evolution approach in E. coli to obtain Sniper-Cas9, which shows high specificities without killing on-target activities in human cells. Unlike other engineered Cas9 variants, Sniper-Cas9 shows WT-level on-target activities with extended or truncated sgRNAs with further reduced off-target activities and works well in a preassembled ribonucleoprotein (RNP) format to allow DNA-free genome editing. Undesired off-target effects can hamper the use of CRISPR-Cas9 in therapeutic applications. Here the authors use a directed evolution approach to develop Sniper-Cas9 which combines high specificity with no loss of on-target activity.

In vitro methods have enabled the rapid and efficient evolution of proteins and successful generation of novel and highly functional proteins. However, the available methods consider only globular proteins (e.g., antibodies, enzymes) and not membrane proteins despite the biological and pharmaceutical importance of the latter. In this study, we report the development of a method called liposome display that can evolve the properties of membrane proteins entirely in vitro. This method, which involves in vitro protein synthesis inside liposomes, which are cell-sized phospholipid vesicles, was applied to the pore-forming activity of α-hemolysin, a membrane protein derived from Staphylococcus aureus . The obtained α-hemolysin mutant possessed only two point mutations but exhibited a 30-fold increase in its pore-forming activity compared with the WT. Given the ability to synthesize various membrane proteins and modify protein synthesis and functional screening conditions, this method will allow for the rapid and efficient evolution of a wide range of membrane proteins.

Improving crop traits requires genetic diversity, which allows breeders to select advantageous alleles of key genes. In species or loci that lack sufficient genetic diversity, synthetic directed evolution (SDE) can supplement natural variation, thus expanding the possibilities for trait engineering. In this review, we explore recent advances and applications of SDE for crop improvement, highlighting potential targets (coding sequences and cis -regulatory elements) and computational tools to enhance crop resilience and performance across diverse environments. Recent advancements in SDE approaches have streamlined the generation of variants and the selection processes; by leveraging these advanced technologies and principles, we can minimize concerns about host fitness and unintended effects, thus opening promising avenues for effectively enhancing crop traits.

The RNA-based organisms from which modern life is thought to have descended would have depended on an RNA polymerase ribozyme to copy functional RNA molecules, including copying the polymerase itself. Such a polymerase must have been capable of copying structured RNAs with high efficiency and high fidelity to maintain genetic information across successive generations. Here the class I RNA polymerase ribozyme was evolved in vitro for the ability to synthesize functional ribozymes, resulting in the markedly improved ability to synthesize complex RNAs using nucleoside 5′-triphosphate (NTP) substrates. The polymerase is descended from the class I ligase, which contains the same catalytic core as the polymerase. The class I ligase can be synthesized by the improved polymerase as three separate RNA strands that assemble to form a functional ligase. The polymerase also can synthesize the complement of each of these three strands. Despite this remarkable level of activity, only a very small fraction of the assembled ligases retain catalytic activity due to the presence of disabling mutations. Thus, the fidelity of RNA polymerization should be considered a major impediment to the construction of a self-sustained, RNA-based evolving system. The propagation of heritable information requires both efficient and accurate synthesis of genetic molecules, a requirement relevant to both laboratory systems and the early history of life on Earth.

Directed evolution involves generating diverse sequence variants of a gene of interest to produce a desirable trait under selective pressure. CRISPR-Cas9 (clustered regularly interspaced short palindromic repeats-CRISPR-associated protein 9) systems can be programmed to target any genomic locus and perform targeted directed evolution. Here, we discuss the opportunities and challenges of this emerging platform for targeted crop improvement.

Directed evolution (DE) is a powerful tool to optimize protein fitness for a specific application. However, DE can be inefficient when mutations exhibit non-additive, or epistatic, behavior. Here, we present Active Learning-assisted Directed Evolution (ALDE), an iterative machine learning-assisted DE workflow that leverages uncertainty quantification to explore the search space of proteins more efficiently than current DE methods. We apply ALDE to an engineering landscape that is challenging for DE: optimization of five epistatic residues in the active site of an enzyme. In three rounds of wet-lab experimentation, we improve the yield of a desired product of a non-native cyclopropanation reaction from 12% to 93%. We also perform computational simulations on existing protein sequence-fitness datasets to support our argument that ALDE can be more effective than DE. Overall, ALDE is a practical and broadly applicable strategy to unlock improved protein engineering outcomes. Directed evolution is a powerful method to optimize protein fitness. Here, authors develop an active learning workflow using machine learning to more efficiently explore the design space of proteins.