The researchers first highlight a key challenge: stabilizing protein redesigns often reduce catalytic performance. They therefore ask whether a highly stabilized—but kinetically impaired—redesign can still outperform the wild-type (WT) enzyme during evolution. Their starting test case is ProteinMPNN redesign D4, which initially expresses well but cleaves substrate far more slowly than WT (about 20-fold lower rate), even though it shows a higher melting temperature.

To extend the comparison beyond a single redesign, the team also selects PROSS1, a top-performing alternative redesign designed to improve both stability and catalysis. Parallel PACE campaigns are then launched from D4 and PROSS1, using the same substrate panel and the same phage flow-rate schedule previously applied to WT and another redesign (D3), enabling a like-for-like assessment of how each redesigned “starting point” alters evolutionary trajectories.

Across 44 independent evolutions on multiple devices and replicates, both D4- and PROSS1-initiated populations clear selection on three substrates spanning low, intermediate, and high difficulty. Sequencing reveals that the consensus genotypes that emerge are consistently distinct from those evolved from WT or D3 backgrounds, indicating that redesign changes the fitness landscape enough to redirect evolution.

A key outcome is that the redesign-derived proteases remain more active than WT-evolved variants across all substrates. D3-evolved and PROSS1-evolved enzymes achieve similar performance, while D4-evolved variants match or exceed the other redesigns despite their slower starting catalysis—evidence that redesign can increase evolvability even when the initial enzyme is functionally handicapped.

The authors further perform a comprehensive “grafting” experiment: evolved genotypes are tested in each alternative protease background. Redesign-to-redesign compatibility is largely high—many variants remain functional when transplanted—yet D4 shows notable variability, with some genotypes becoming highly active only within their original background. They suggest this reflects a trade-off between robustness to destabilizing mutations and sensitivity to slow-catalysis-associated changes.

In contrast, grafting redesign-evolved mutations into the WT protease yields dramatically lower activity, including near-zero function on the easiest substrate and functional variants only for a minority of cases on harder substrates. Activity correlations between redesign backgrounds and WT are poor, implying that redesign enables new evolutionary solutions that WT cannot access efficiently.

Finally, the team maps genotype activity onto an embedding space from a protein language model (ESM-C). In WT, the fitness landscape is sparse: high-function solutions cluster far from the starting sequence, while local single-point mutants show little to no activity. Redesign expands the landscape—raising average fitness and allowing function in sequences closer to the starting point, including single-mutant neighborhoods—consistent with the idea that redesign accelerates adaptation by expanding the set of viable paths.

Subject of Research: AI-redesigned starting points and their effect on protein evolvability and fitness landscapes during PACE evolution

Article Title: AI-redesigned starting points and outcomes enhance protein evolution

Article References: Krasnow, N.A., Xu, J.A., Zhang, E. et al. AI-redesigned starting points and outcomes enhance protein evolution. Nature (2026). https://doi.org/10.1038/s41586-026-10820-0

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s41586-026-10820-0

Keywords: Protein engineering; AI redesign; PACE; fitness landscape; protease evolution; language model embeddings

At the heart of their findings lies a striking observation regarding the enzyme’s anatomy: the core regions — the densely packed interiors — are far more conserved than the exposed surfaces. This discovery aligns with a fundamental principle of protein stability, where internal residues maintain the structural integrity necessary for proper folding and function. By leveraging high-resolution structural predictions and meticulous conservation ratios, the study charts a nuanced map revealing that binding sites—the functional hotspots where substrates engage—exhibit the highest levels of conservation across orthologous enzyme groups.

Intriguingly, the research identifies exceptions that challenge broad generalizations. For instance, within the fatty acid synthetase subunit Fas2p, binding sites display greater variability than even the overall protein structure. This finding hints at specialized evolutionary pressures or functional adaptations unique to certain enzyme classes, underscoring the complex tapestry of molecular evolution beyond canonical expectations.

Delving deeper into the secondary structural elements, the study contrasts α-helices and β-sheets, revealing a surprising trend: helical segments showcase more variation compared to extended β-sheet regions. Given that α-helices are typically associated with dynamic flexibility, this observation contradicts the presupposed idea that flexible regions evolve more rapidly. The authors postulate that the presence of conserved turns within random coil regions may contribute to this phenomenon, suggesting a refined understanding of secondary structure evolution that transcends simple categorizations.

The research also extends its scope to encompass larger protein folds, examining evolutionary conservation at the architectural level. Utilizing data from The Encyclopedia of Domains database, the study finds that amino acids embedded within defined protein folds, such as the Rossmann fold and the TIM barrel, are markedly conserved compared to non-folded regions. These dominant folds are not only structurally significant but have stood the test of time to preserve crucial biochemical functions, signaling their evolutionary robustness and indispensable roles.

Beyond mere static conservation, the study probes the dynamic evolutionary forces acting upon enzyme residues by estimating non-synonymous to synonymous substitution rates (dN/dS) across both full proteins and individual sites. Surface residues show a higher likelihood of neutral drift, reflecting their increased evolutionary tolerance, while specific instances reveal positive selection acting on residues implicated in protein-protein interactions or substrate binding. These adaptive changes highlight the nuanced balance between structural preservation and evolutionary innovation.

One of the most compelling revelations concerns the spatial organization of conserved residues within enzymes. Rather than being randomly scattered, fully conserved amino acids tend to cluster within distinct regions. By constructing residue-interaction networks, the researchers identify clusters that correspond not only to known substrate-binding and allosteric sites but also to symmetric protein-protein interaction interfaces. In pyruvate kinase, for example, these clusters delineate critical functional domains responsible for enzymatic activity and regulation, underscoring their biological significance.

In quantifying the overlap between conserved clusters and functionally annotated binding sites, the study reports that over 90% of such sites coincide with at least one conserved cluster. However, not all clusters correlate with recognized binding regions, suggesting the existence of unexplored interaction sites or structural elements pivotal to enzyme function. This finding prompts a call for comprehensive annotations and deeper molecular characterizations.

Advancing beyond descriptive analyses, the research team employs machine learning techniques, specifically histogram-based gradient boosting classifiers, to predict the likelihood of conserved clusters corresponding to functional binding sites based on physicochemical properties. The model achieves performance metrics notably superior to random baselines, reinforcing the predictive value of conserved residue networks in identifying functionally critical enzyme regions.

Further analyses reveal that conserved clusters extend their relevance to protein-protein interaction sites, albeit to a lesser extent than small-molecule binding sites. Enzymes engaged in numerous physical interactions or integrated within protein complexes exhibit higher overall conservation, aligning with the concept that structural conservation is often driven by the necessity to maintain essential biological interactions.

Interestingly, the study observes that certain known small-molecule binding sites evade inclusion within conserved clusters. This discrepancy may be attributable to the diminutive size of some ligand interaction sites—such as metal coordination centers—or gaps in current functional annotations. Such nuances emphasize the complexity inherent in protein structure-function relationships and advocate for ongoing refinement of biochemical databases.

Altogether, this extensive body of work delineates a clear hierarchy in enzyme structural evolution, where conservation intricately mirrors functional imperatives, ranging from buried cores to ligand-binding pockets and intermolecular interfaces. The integration of evolutionary metrics, structural biology, and computational prediction frameworks paves the way for future explorations into enzyme adaptability, paving new avenues for bioengineering and drug design.

By harnessing vast evolutionary timelines and diverse analytical approaches, this study stands as a testament to the profound influence of metabolism in proteomic evolution, elucidating enduring molecular patterns that have shaped life’s biochemical toolkit for hundreds of millions of years.

Subject of Research: Evolutionary conservation and structural adaptation of enzymes over 400 million years, focusing on the relationship between metabolic function and enzyme structural features.

Article Title: The role of metabolism in shaping enzyme structures over 400 million years.

Article References:
Lemke, O., Heineike, B.M., Viknander, S. et al. The role of metabolism in shaping enzyme structures over 400 million years. Nature (2025). https://doi.org/10.1038/s41586-025-09205-6

Image Credits: AI Generated