{"id":"33e71876-d210-4812-b6a4-dfd12513a71d","arxiv_id":"2505.00600","paper_version":2,"verdict":"UNVERDICTED","confidence":"HIGH","novelty_score":0.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Local frustration in proteins sits at a sweet spot that enables the motions needed for enzyme catalysis; this review gathers the supporting evidence.","lead":"This paper reviews how a property called local frustration in proteins helps control the motions that make enzymes work. It argues that evolution tunes this frustration to a sweet spot that lets enzymes move and catalyze reactions efficiently.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claim that evolution tunes local frustration “to near optimal values” is asserted but never tested: the cited evidence is correlational and relies on an unvalidated coarse-grained proxy, with no fitness or phylogenetic null.","rationale":"The reader's weakest assumption correctly flags the frustratometeR proxy as fragile, and I agree that the quantitative backbone of the review depends on it. However, I would push one step further: even granting the proxy, the abstract's “near optimal values” clause is an evolutionary-optimality claim that the review never tests. The cited evidence is correlational or limited to individual systems, and the paper itself acknowledges the absence of a definitive frustration measure and of a quantitative general mechanism. The load-bearing gap is therefore not an internal inconsistency but an unsupported inference from local enrichment and case-study correlations to selection for optimal frustration. A concrete computational check using existing deep-mutational scanning or the cited enzyme datasets could settle whether wild-type frustration sits at a fitness peak; if it does not, or if the correlation disappears once stability is controlled, the headline claim should be tempered. Since this is a review article, the appropriate adjustment is conditional acceptance on revising the abstract and highlights to match the evidence, rather than outright rejection.","tokens_in":12241,"tokens_out":8987,"duration_ms":99732,"concrete_test":"On a large enzyme deep-mutational scanning dataset (or on the Burns et al. enzyme I and TEV protease datasets already cited), compute per-mutant local frustration with frustratometeR on the same structures or MD ensembles used for dynamics, then regress measured kcat/KM or activity on predicted frustration change. Fit a quadratic and test whether wild-type frustration lies within the 95% CI of the fitted optimum, while also controlling for folding stability. If wild-type is not at or near the peak, or if frustration change is not predictive after controlling for stability, the “near-optimal evolutionary tuning” claim is contradicted.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The abstract's central claim is that “the biologically relevant dynamics is tuned by the evolution of protein sequences that modulate the local frustration patterns to near optimal values.” Three conditions must hold: (1) the frustratometeR index used in Figure 3 and throughout measures the physical frustration that actually controls catalytic dynamics, not just folding-decoy energetics; (2) sequence changes alter this index in a way that tracks catalytic rate; and (3) observed patterns reflect selection for optimal frustration rather than a byproduct of catalytic-site conservation, stability constraints, or mutational robustness. The paper itself concedes “there is no definitive way to determine and quantify the local frustration” and “we are still lacking a quantitative general mechanism.” The cited studies are correlational or system-specific: Freiberger et al. show active-site enrichment, Hou et al. show stability shells, Weinreb at al. show mutations in strained regions affect flexibility, but none compares a fitness/activity landscape against a neutral phylogenetic null. Thus the strongest clause—“near optimal values”—is an evolutionary claim the review never actually tests.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript is a review article on the role of local frustration in protein dynamics and enzyme catalysis. The authors synthesize theory, computation, and experiment to argue that local frustration—residual energetic conflicts in folded proteins—creates ruggedness in energy landscapes that underlies functionally important motions (FIMs) and thereby tunes catalytic power. They discuss correlative surveys (Freiberger et al., Hou et al.), mutation studies (Weinreb et al., Ojeda-May, Stelzl et al.), steric/substrate-product frustration in adenylate kinase, redox modulation in nDsbD, and design implications from ProteinMPNN redesigns. The central claim, stated in the abstract, is that evolution tunes local frustration patterns to near optimal values to control biologically relevant dynamics. The manuscript contains no new experimental or computational data; it is a narrative synthesis with illustrative figures and a forward-looking design perspective.","tokens_in":12485,"tokens_out":4400,"duration_ms":49853,"significance":"If the central claim is correct, local frustration would constitute a quantifiable, evolutionarily selected feature that links sequence, dynamics, and catalysis, with practical implications for enzyme design and inhibitor development. The review's strength is its integration of independent lines of evidence—bioinformatic surveys, single-molecule/nano-rheology experiments, MD simulations, and protein design—into a coherent energy-landscape framework. The paper is clearly written and the figures are helpful. However, the strongest claim (evolutionary tuning to near-optimal values) is not directly supported by any study cited, and the operational measure of frustration (frustratometeR) is a coarse-grained proxy whose connection to the physical frustration that governs catalysis is not established within the manuscript. The review would be more convincing if it explicitly distinguished established findings from the authors' interpretive framework and if it offered concrete criteria for falsifying the optimization claim.","major_comments":[{"comment":"The abstract asserts that 'the biologically relevant dynamics is tuned by the evolution of protein sequences that modulate the local frustration patterns to near optimal values.' This is the paper's strongest and most novel claim, but no cited study tests it as an evolutionary optimization hypothesis. The evidence presented is correlational: Freiberger et al. show active-site enrichment in frustrated contacts, Hou et al. show stability-weakness shells, Weinreb et al. show mutations in strained regions affect flexibility and activity, and Burns et al. show temperature-sensitive loop contacts. None of these compares an activity or fitness landscape against a neutral phylogenetic null, and the manuscript itself concedes (in 'Local frustration tunes enzymes' internal dynamics') that 'there is no definitive way to determine and quantify the local frustration' and (in 'Design Implications and Future Perspectives') that 'we are still lacking a quantitative general mechanism.' As written, the phrase 'near optimal values' overstates the support. Please reframe it as an explicit, testable hypothesis or add a dedicated section discussing what evidence would be required (e.g., selection statistics on frustration indices, phylogenetic contrasts) and what the current literature does and does not show.","section":"Abstract and Section 'Local frustration tunes enzymes' internal dynamics'"},{"comment":"The review consistently treats the frustratometeR configurational frustration index (ref 18, Figure 3) as a direct measure of the physical frustration that governs functional dynamics. However, the index is computed from folding-decoy energetics, and the paper states that no definitive quantification of local frustration exists. The case studies of Guanylate Kinase, Shikimate Kinase, and nDsbD rely on this proxy to identify 'frustrated' regions, yet the manuscript does not validate that the decoy-based index reports the same conflicts that produce functional strain, viscoelasticity, or loop opening in vivo. This is load-bearing because the entire synthesis depends on equating a structural bioinformatics quantity with a dynamical mechanism. Please state the proxy's validation status explicitly, and, where possible, cross-check it against experimentally measured strain or flexibility (e.g., Weinreb et al.'s nano-rheology) in at least one system so that the reader can assess how much of the argument rests on the proxy.","section":"Figure 3 and case studies in 'Functionally important motions drive catalysis'"},{"comment":"The manuscript moves from correlation to causation in its treatment of mutation studies. For example, the text says that mutations in high-strain regions 'can significantly dampen both protein flexibility and catalytic activity' (Weinreb et al.) and that in Shikimate Kinase increased local frustration 'correlated with experimentally observed reductions in catalytic efficiency.' These observations are consistent with a role for frustration, but they could also arise from off-target effects on global stability, folding cooperativity, or active-site geometry, and the cited papers may not control for all such alternatives. The review does not discuss which controls or comparisons in these studies justify the causal interpretation that local frustration itself tunes catalysis. Please temper the causal wording or, alternatively, explicitly identify the controls in the cited experiments that make the causal reading credible.","section":"Section 'Functionally important motions drive catalysis'"}],"minor_comments":[{"comment":"Several references are preprints or recently published items without full bibliographic details (e.g., ref 29 is a bioRxiv preprint from 2025; ref 26 is in press). Please confirm the journal's policy on citing preprints and mark them consistently.","section":"References"},{"comment":"The caption orders the right panels as 'top: sweet spot, middle: too much frustration, bottom: too little frustration,' and the main text repeats this ordering. This is internally consistent, but the visual distinction between 'too much' and 'too little' would be clearer if the figure included representative free-energy curves or population labels rather than only schematic cartoons.","section":"Figure 2"},{"comment":"The final sentence of the Acknowledgements draws attention to the erosion of Argentina's scientific tradition. This is a legitimate political statement, but it is not part of the scientific content; consider moving it to a cover letter or a separate statement to adhere to journal formatting conventions.","section":"Acknowledgements"},{"comment":"The term 'local frustration' is used in two senses: as a structural descriptor (computed from protein structures) and as a dynamic mechanism (modulating conformational substates). The connection between these senses is implicit in the energy-landscape framework, but the transition is abrupt. Define the relationship explicitly at the first occurrence in 'Local frustration tunes enzymes' internal dynamics'.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The review is essentially a synthesis of the authors' own frustration framework with independent experimental studies. The framework is influential, and the review is well suited to a journal that publishes authoritative perspectives, but the manuscript would be substantially stronger if it engaged more explicitly with skeptical positions, particularly the Warshel school's arguments that conformational dynamics do not generally catalyze the chemical step (refs 32 and 39 are cited but not discussed). The central claim about evolutionary optimization is presented as a conclusion rather than a hypothesis; the authors should either provide a clear falsifiable formulation or soften the abstract. There is no evidence of misconduct or missing data, and the limitations are partially acknowledged in the text, which supports a major-revision rather than a reject decision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a review, not a research paper, so the standard accept/reject tests don't directly apply. What the authors do well is pull together a scattershot recent literature — Weinreb's viscoelastic enzymes, Burns's temperature-sensitive loops, Stelzl's redox switch, Li/Kong's steric frustration in AdK, Sumida's ProteinMPNN rescue — and organize it around a simple, teachable image: a \"sweet spot\" of local frustration where functionally important motions (FIMs) can emerge, with too little freezing the protein and too much dissolving it into biologically unimportant motions. The cartoons in Figures 1 and 2 are genuinely helpful. They also concede the two biggest weaknesses right in the text: \"there is no definitive way to determine and quantify the local frustration,\" and \"we are still lacking a quantitative general mechanism.\" That honesty is to their credit.\n\nThe soft spots are real but mostly proportioned. The abstract's closing claim — that evolution tunes local frustration \"to near optimal values\" — goes beyond any evidence in the review. The cited studies are correlational or system-specific, and none compares a fitness/activity landscape against a neutral phylogenetic null. The frustratometeR index used in Figure 3 is a coarse-grained proxy based on decoy energetics, and the entire synthesis inherits its assumptions. The sweet spot itself is qualitative; there is no quantitative mapping from frustration levels to catalytic rates. And the review cites the stability-activity tradeoff literature without fully disentangling frustration from those harder constraints. One small style snag: several reference annotations (refs 20, 21, 29, 33, 36, 37, 40) editorialize with asterisked commentary, which is unusual in a review.\n\nStill, the paper is a legitimate contribution as a review. It is written by the people who built the local frustration framework, and they cite independent work alongside their own — none of the central claims reduce to their own calculations. It is coherent on its own terms. I would send it to a serious referee: a good reviewer should push for the \"near optimal values\" language to be framed as a hypothesis, check the selection of case studies for fairness, and see whether the Discussion does more to separate frustration from stability-activity tradeoffs. I wouldn't desk-reject it. If I worked on enzyme engineering or protein dynamics, I'd cite it for the conceptual framing; for a reading group, it's a reasonable overview but not a must-read primary paper.\n\nNet: worth reviewing, with minor revisions expected.","headline":"A useful and honest narrative review of local frustration in enzyme catalysis; the 'near optimal values' evolutionary claim is overreach, but the synthesis is worth sending to a referee.","tokens_in":12886,"tokens_out":3503,"would_cite":true,"duration_ms":35343,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Evolution tunes protein frustration to power enzyme catalysis","keywords":["local frustration","protein dynamics","enzyme catalysis","energy landscapes","conformational substates","enzyme evolution","protein design","frustration conservation"],"falsifier":"Measure catalytic efficiency and microsecond-to-millisecond dynamics for a library of single-point mutants in and around predicted frustrated patches, and check whether changes in the frustration index predict changes in activity. If a mutation that leaves the index unchanged nonetheless abolishes catalysis, or one that raises frustration speeds catalysis in a system where the index predicts slowing, the proposed causal link between local frustration, dynamics, and catalytic power would fail.","tokens_in":12100,"feed_emoji":"🔬","tokens_out":7625,"duration_ms":78927,"temperature":0.7,"pith_summary":"This review argues that local frustration—the minority of interactions in a folded protein whose energies are not optimized for stability—is not a flaw but a functional feature. It claims evolution adjusts protein sequences so that frustration sits at a near-optimal level: enough to create the rugged energy landscape whose conformational substates drive functionally important motions, but not so much that motions become noisy or so little that they freeze. Seen this way, enzyme catalysis depends on frustration patterns near active sites and in distal regions that couple dynamically to them, which is why active sites are consistently enriched in frustrated contacts. The payoff is a unifying explanation for why stability-optimized protein design often destroys activity, and a rationale for treating frustration as an evolutionary fingerprint of function.","feed_headline":"Evolution tunes protein frustration to power enzyme catalysis","feed_subtitle":"Enzymatic speed comes from a sweet spot of internal energetic conflict, not just active-site chemistry.","key_machinery":"The load-bearing object is the local frustration index, a coarse-grained measure that classifies each residue–residue contact as minimally frustrated, neutral, or highly frustrated by comparing the contact's native energy with the distribution of energies in structurally randomized decoys. This index is what lets the authors map frustration onto known enzymes and connect it to dynamics: frustrated patches mark functional regions, minimally frustrated contacts mark stability and foldability, and neutral contacts fill the rest. Its role in the argument is to translate energy-landscape theory into a per-residue observable that can be surveyed across enzyme families, mutated in silico, and compared with experimental measures of motion and catalysis.","core_discovery":"The central claim is that the biologically relevant dynamics of enzymes is tuned by evolution through the modulation of local frustration patterns to near-optimal values. Frustration roughens an otherwise smooth folding funnel: highly frustrated contacts are concentrated at catalytic sites, surrounded by weakly frustrated shells, and these energetic signatures are more conserved across enzyme families than the underlying sequences. Too little frustration freezes the conformational substates that carry out the reaction cycle; too much diffuses the motion into biologically unimportant fluctuations; the functional sweet spot in between supports the functionally important motions that bind, align, and release substrates and products. The review assembles evidence from enzyme surveys, nano-rheology, molecular dynamics, redox-dependent loop switching, and protein design to show that perturbing frustration—even far from the active site—changes dynamics and activity in predictable directions.","pith_inferences":["A testable extension: comparing orthologous enzymes from organisms adapted to different temperatures, one would predict that the optimal frustration level shifts so that conformational substates remain populated at the physiological temperature; thermophilic enzymes should show quantitatively different frustration patterns than mesophilic ones.","The review's logic implies that frustration-aware sequence design, rather than stability-aware design, should be used when generating novel biocatalysts; catalytically active designed proteins may need to be deliberately destabilized locally to recreate functional substates.","If the frustration index is validated against direct measurements of conformational entropy or millisecond dynamics across many mutants, it could become a practical predictive screen for the functional effects of distal mutations.","One open question the authors do not resolve: whether the frustration index derived from static structures or short simulations reports the same conflicts that operate along the full catalytic cycle, so linking frustration to time-resolved experimental observables would strengthen the synthesis."],"forward_implications":["Active-site frustration is a conserved feature across enzyme classes, so any model that claims to capture the origins of catalysis should reproduce the enrichment of frustrated contacts at catalytic residues.","Engineering an enzyme for higher stability alone can silence it, because stability-optimizing substitutions remove the very frustrated interactions that enable conformational transitions; preserving conserved or frustrated residues should be part of design protocols.","Mutations far from the active site can change catalytic efficiency by altering long-range dynamical coupling through frustrated regions, expanding the search space for enzyme engineering and for understanding disease variants.","If product release is rate-limiting in multi-substrate enzymes, substrate-induced steric or electrostatic frustration can be a general mechanism to accelerate turnover, not a quirk of adenylate kinase.","Frustration conservation across protein families can be used to identify functionally essential residues, complementing sequence conservation in predicting sites where mutations will matter."],"supporting_citations":[{"why":"Introduces the local frustration analysis that maps native contacts into minimally frustrated, neutral, and highly frustrated classes.","marker":"[17]"},{"why":"Provides the computational tool used to compute the local frustration index for the enzymes shown and discussed.","marker":"[18]"},{"why":"Survey of annotated active sites showing catalytic sites are enriched in highly frustrated interactions across enzyme classes.","marker":"[19]"},{"why":"Independent statistical-potential analysis showing catalytic residues sit in stability weaknesses surrounded by stabilizing shells.","marker":"[21]"},{"why":"Nano-rheology experiments showing mutations in high-strain regions far from the active site dampen flexibility and catalysis.","marker":"[26]"},{"why":"Simulations showing steric frustration between incoming substrate and bound product accelerates product release in adenylate kinase.","marker":"[30]"},{"why":"Temperature-sensitive contacts in disordered catalytic loops tune enzyme I activity and can be modulated by mutation.","marker":"[33]"},{"why":"Shows redox-dependent local frustration controls loop opening and conformational selection in nDsbD.","marker":"[34]"},{"why":"Shows stability-optimized sequence redesign destroys TEV protease activity unless conserved functional residues are fixed.","marker":"[37]"},{"why":"Uses frustration analysis to steer structure-prediction models toward functional conformational states, connecting frustration to dynamics.","marker":"[40]"}],"fun_headline_variants":["Evolution tunes frustration to optimize enzyme dynamics","Protein frustration sweet spot powers enzyme catalysis","Evolution refines frustration for enzyme function","Frustration balance explains enzyme speed","Enzymes work best when protein frustration is just right"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole synthesis rests on the assumption that the computer measure of frustration—comparing a protein's native contacts to randomized alternative structures—captures the physical conflicts that truly govern enzyme motion and catalysis under working conditions.","fun_headline_variants_meta":{"raw":{"variants":["Evolution tunes frustration to optimize enzyme dynamics","Protein frustration sweet spot powers enzyme catalysis","Evolution refines frustration for enzyme function","Frustration balance explains enzyme speed","Enzymes work best when protein frustration is just right"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000492,"raw_usage":{"total_tokens":2337,"prompt_tokens":786,"completion_tokens":1551,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":402,"completion_tokens_details":{"reasoning_tokens":1486}},"tokens_in":402,"tokens_out":1551,"duration_ms":11873,"temperature":1.0,"reasoning_tokens":1486,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T04:37:19.387142+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure catalytic efficiency and microsecond-to-millisecond dynamics for a library of single-point mutants in and around predicted frustrated patches, and check whether changes in the frustration index predict changes in activity. If a mutation that leaves the index unchanged nonetheless abolishes catalysis, or one that raises frustration speeds catalysis in a system where the index predicts slowing, the proposed causal link between local frustration, dynamics, and catalytic power would fail.","supporting_citations":[{"cited_title":"Localizing frustration in native proteins and protein assemblies","cited_arxiv_id":null,"evidence_quote":"Introduces the local frustration analysis that maps native contacts into minimally frustrated, neutral, and highly frustrated classes."},{"cited_title":"FrustratometeR: an R-package to compute local frustration in protein structures, point mutants and MD simulations","cited_arxiv_id":null,"evidence_quote":"Provides the computational tool used to compute the local frustration index for the enzymes shown and discussed."},{"cited_title":"Local frustration around enzyme active sites","cited_arxiv_id":null,"evidence_quote":"Survey of annotated active sites showing catalytic sites are enriched in highly frustrated interactions across enzyme classes."},{"cited_title":"Enzyme Stability-Activity Trade-Off: New Insights from Protein Stability Weaknesses and Evolutionary Conservation","cited_arxiv_id":null,"evidence_quote":"Independent statistical-potential analysis showing catalytic residues sit in stability weaknesses surrounded by stabilizing shells."},{"cited_title":"Enzymes as viscoelastic catalytic machines","cited_arxiv_id":null,"evidence_quote":"Nano-rheology experiments showing mutations in high-strain regions far from the active site dampen flexibility and catalysis."},{"cited_title":"Overcoming the Bottleneck of the Enzymatic Cycle by Steric Frustration","cited_arxiv_id":null,"evidence_quote":"Simulations showing steric frustration between incoming substrate and bound product accelerates product release in adenylate kinase."},{"cited_title":"Temperature-sensitive contacts in disordered loops tune enzyme I activity","cited_arxiv_id":null,"evidence_quote":"Temperature-sensitive contacts in disordered catalytic loops tune enzyme I activity and can be modulated by mutation."},{"cited_title":"Local frustration determines loop opening during the catalytic cycle of an oxidoreductase","cited_arxiv_id":null,"evidence_quote":"Shows redox-dependent local frustration controls loop opening and conformational selection in nDsbD."},{"cited_title":"Improving Protein Expression, Stability, and Function with ProteinMPNN","cited_arxiv_id":null,"evidence_quote":"Shows stability-optimized sequence redesign destroys TEV protease activity unless conserved functional residues are fixed."},{"cited_title":"Predicting protein conformational motions using energetic frustration analysis and AlphaFold2","cited_arxiv_id":null,"evidence_quote":"Uses frustration analysis to steer structure-prediction models toward functional conformational states, connecting frustration to dynamics."}],"review_version":1}