Convergent evolution of epigenome recruited DNA repair across the Tree of Life

Reviewer #1 (Public review):

Summary:

Previous studies have shown that the MSH6 family of mismatch repair proteins contains an unstructured N-terminal domain that contains either a PWWP domain, a Tudor domain or neither and that the interaction of the histone reader domains with the appropriate histone H3 modification enhances mismatch repair, and hence reduces mutation rates in coding regions to some extent. However, the elimination of the MSH6-histone modification probably does not completely eliminate mismatch repair, although the published papers on this point do not seem definitive.

In this study, the authors perform a details phylogenetic analysis of the presence of the PWWP and Tudor domains in MSH6 proteins across the tree of life. They observe that there are basically three classes of organisms that contain either a PWWP domain, a Tudor domain, or neither. On the basis of their analysis, they suggest that this represents convergent evolution of the independent acquisition of histone reader domains and that key amino acid residues in the reader domains are selected for.

Strengths:

The phylogenetic aspects of the work seem well done and the basic evolutionary conclusions of the work are well supported. The basic evolutionary conclusions are interesting and there is little to criticize from my perspective.

Weaknesses:

A major concern about this paper is that the authors fail to put their work into the proper context of what is already known about the N-terminus of MSH6. Further, their structural studies, which are really structural illustrations, are misleading, often incorrect, and not always helpful in addition to having been published before.

Reviewer #2 (Public review):

Summary:

In this work, Monroe JG and colleagues show a compelling case of convergent evolution in the fusion between an important mismatch repair protein (MSH6) and histone reader domains across the tree of life. These fused MSH6 readers have been shown to be important for the recruitment of MSH6 to exon-rich genome locations, therefore improving the efficiency of reducing mutation rates in coding regions.

Comparative genomic analyses here performed revealed independent instances of MSH6 fusion with histone readers in plants and metazoa with several instances of putative loss (or gain) across the phylogeny. The work also unveiled instances of MSH6 fusion putatively interesting domains in fungi which might be worth exploring in the future.

The authors also show potential signatures of purifying selection in functional amino acids MSH6 histone readers.

Overall the approach is adequate for the questions proposed to be answered, the analyses are rigorous and support the authors' claims.

DNA repair genes are essential to maintain genome stability and fidelity, and alterations in these pathways have been associated with hypermutation phenotypes in the context for instance of cancer in humans, with sometimes implications in treatment resistance. This is an important work that contributes to our understanding of the evolutionary consequences of the evolution of epigenome-targeted DNA repair.

Strengths:

The methods used are adequate for the questions and support the results. The search for MSH6 fusions was rigorous and conservative, which strengthens the significance of the claims on the evolutionary history of these fusion events.

Weaknesses:

I did not identify any major weaknesses, but please see my suggestions/recommendations.

Reviewer #3 (Public review):

Summary:

In the manuscript entitled "Convergent evolution of epigenome recruited DNA repair across the Tree of Life", Monroe et al. investigate bioinformatically how some important mechanisms of epigenome-targeted DNA repair evolved at the tree of life scale. They provide a clear example of convergent evolution of these mechanisms between animals and plants, investigating more than 4000 eukaryotic genomes, and uncovering a significant association between gain/retention of such mechanisms with genome size and high intron content, that at least partially explains the evolutionary patterns observed within major eukaryotic lineages.

Strengths:

The manuscript is well written, clear, and understandable, and has potentially broad interest. It provides a thorough analysis of the evolution of MSH6-related DNA repair mechanisms using more than 4000 eukaryotic genomes, a pretty impressive number allowing to identify both large-scale (i.e. kingdoms) as well as shorter-scale (i.e. phyla, orders) evolutionary patterns. Moreover, despite providing no experimental validation, it investigates with a sufficient degree of depth, a potential relationship between gain/retention of epigenome recruited DNA repair mediated by MSH6 and genomic, as well as life-history (population size, body mass, lifespan), traits. In particular, it provides convincing evidence for a causative effect between genome size/intron content and the presence/absence of this mechanism. Moreover, it stimulates further scientific investigation and biological questions to be addressed, such as the conservation of epigenomes across the tree of life, the existence of potential trade-offs in gain/retention vs. loss of such mechanisms, and the relationship between these processes, mutation rate heterogeneity, and evolvability.

Weaknesses:

Despite the interesting and necessary insights provided on (1) the evolution of DNA repair mechanisms, and (2) the convergent evolution of molecular mechanisms, this bioinformatic study emanates from studies in humans and Arabidopsis already showing signs of potential convergent evolution in aspects of epigenome-recruited DNA repair. For this, this study, although bioinformatically remarkably thorough, does not come as a surprise, potentially lowering its novelty.

What could have increased further its impact, interest, and novelty could have been a more comprehensive understanding of the causative processes leading to gain/retention vs. loss of MSH6-related epigenetic recruitment mechanisms. The authors provide interesting associations with life-history traits (yet not significant), and significant links with genome size and intron content only at the theoretical level. For the first aspect, the analyses could have expanded toward other life-history traits. For the second, maybe it could have been even possible to tackle experimentally some of the generated questions, functionally in some models, or deepened using specific case studies.

Author response:

Public Reviews:

Reviewer #1 (Public review):

Summary:

Previous studies have shown that the MSH6 family of mismatch repair proteins contains an unstructured N-terminal domain that contains either a PWWP domain, a Tudor domain or neither and that the interaction of the histone reader domains with the appropriate histone H3 modification enhances mismatch repair, and hence reduces mutation rates in coding regions to some extent. However, the elimination of the MSH6-histone modification probably does not completely eliminate mismatch repair, although the published papers on this point do not seem definitive.

In this study, the authors perform a details phylogenetic analysis of the presence of the PWWP and Tudor domains in MSH6 proteins across the tree of life. They observe that there are basically three classes of organisms that contain either a PWWP domain, a Tudor domain, or neither. On the basis of their analysis, they suggest that this represents convergent evolution of the independent acquisition of histone reader domains and that key amino acid residues in the reader domains are selected for.

Strengths:

The phylogenetic aspects of the work seem well done and the basic evolutionary conclusions of the work are well supported. The basic evolutionary conclusions are interesting and there is little to criticize from my perspective.

Thank you for the positive evaluation. We appreciate your interest and review.

Weaknesses:

A major concern about this paper is that the authors fail to put their work into the proper context of what is already known about the N-terminus of MSH6. Further, their structural studies, which are really structural illustrations, are misleading, often incorrect, and not always helpful in addition to having been published before.

Thank you for the helpful suggestions on this front. We agree that some of the structural visualizations were over simplified and apologize for the lack of clarity. Notably, we did not annotate the presence of putative or known short PCNA-interacting protein (PIP) motifs which have been found at the linker disordered N-terminus of MSH6 proteins. Indeed, while not direct to our investigation of the origins of histone readers, the PIP motifs are an interesting and functionally important feature of MSH6 structural biology, especially because they may facilitate DNA repair processes more generally. In the revised manuscript, we aim to improve the scholarship on this topic and clarify the presence/importance of this motif for MSH6 function, as well as what is known about the structural biology of the MSH6 N-terminus more broadly. We will add annotations of the PIP motif and will also improve structural prediction by visualizing MSH6 structure in its dimerized form with MSH2, for a more accurate estimate of its folding in vivo. We hope that these in addition to other valuable suggested improvements will enhance the revised manuscript.

Reviewer #2 (Public review):

Summary:

In this work, Monroe JG and colleagues show a compelling case of convergent evolution in the fusion between an important mismatch repair protein (MSH6) and histone reader domains across the tree of life. These fused MSH6 readers have been shown to be important for the recruitment of MSH6 to exon-rich genome locations, therefore improving the efficiency of reducing mutation rates in coding regions.

Comparative genomic analyses here performed revealed independent instances of MSH6 fusion with histone readers in plants and metazoa with several instances of putative loss (or gain) across the phylogeny. The work also unveiled instances of MSH6 fusion putatively interesting domains in fungi which might be worth exploring in the future.

The authors also show potential signatures of purifying selection in functional amino acids MSH6 histone readers.

Overall the approach is adequate for the questions proposed to be answered, the analyses are rigorous and support the authors' claims.

DNA repair genes are essential to maintain genome stability and fidelity, and alterations in these pathways have been associated with hypermutation phenotypes in the context for instance of cancer in humans, with sometimes implications in treatment resistance. This is an important work that contributes to our understanding of the evolutionary consequences of the evolution of epigenome-targeted DNA repair.

Strengths:

The methods used are adequate for the questions and support the results. The search for MSH6 fusions was rigorous and conservative, which strengthens the significance of the claims on the evolutionary history of these fusion events.

Thank you for the positive evaluation. We appreciate your interest and review.

Weaknesses:

I did not identify any major weaknesses, but please see my suggestions/recommendations.

Thank you, we will also address your suggestions, which provide valuable recommendations for improving the revised manuscript.

Reviewer #3 (Public review):

Summary:

In the manuscript entitled "Convergent evolution of epigenome recruited DNA repair across the Tree of Life", Monroe et al. investigate bioinformatically how some important mechanisms of epigenome-targeted DNA repair evolved at the tree of life scale. They provide a clear example of convergent evolution of these mechanisms between animals and plants, investigating more than 4000 eukaryotic genomes, and uncovering a significant association between gain/retention of such mechanisms with genome size and high intron content, that at least partially explains the evolutionary patterns observed within major eukaryotic lineages.

Strengths:

The manuscript is well written, clear, and understandable, and has potentially broad interest. It provides a thorough analysis of the evolution of MSH6-related DNA repair mechanisms using more than 4000 eukaryotic genomes, a pretty impressive number allowing to identify both large-scale (i.e. kingdoms) as well as shorter-scale (i.e. phyla, orders) evolutionary patterns. Moreover, despite providing no experimental validation, it investigates with a sufficient degree of depth, a potential relationship between gain/retention of epigenome recruited DNA repair mediated by MSH6 and genomic, as well as life-history (population size, body mass, lifespan), traits. In particular, it provides convincing evidence for a causative effect between genome size/intron content and the presence/absence of this mechanism. Moreover, it stimulates further scientific investigation and biological questions to be addressed, such as the conservation of epigenomes across the tree of life, the existence of potential trade-offs in gain/retention vs. loss of such mechanisms, and the relationship between these processes, mutation rate heterogeneity, and evolvability.

Thank you for the positive evaluation. We appreciate your interest and review.

Weaknesses:

Despite the interesting and necessary insights provided on (1) the evolution of DNA repair mechanisms, and (2) the convergent evolution of molecular mechanisms, this bioinformatic study emanates from studies in humans and Arabidopsis already showing signs of potential convergent evolution in aspects of epigenome-recruited DNA repair. For this, this study, although bioinformatically remarkably thorough, does not come as a surprise, potentially lowering its novelty.

What could have increased further its impact, interest, and novelty could have been a more comprehensive understanding of the causative processes leading to gain/retention vs. loss of MSH6-related epigenetic recruitment mechanisms. The authors provide interesting associations with life-history traits (yet not significant), and significant links with genome size and intron content only at the theoretical level. For the first aspect, the analyses could have expanded toward other life-history traits. For the second, maybe it could have been even possible to tackle experimentally some of the generated questions, functionally in some models, or deepened using specific case studies.

We agree that this work expands on recent experimental work in humans and Arabidopsis on the function of histone readers in MSH6, PWWP and Tudor, respectively. However, the evolution of these fusions remained a significant knowledge gap, limiting the degree to which functional work could be translated to other organisms. This study definitively characterized the evolutionary history of MHS6 histone readers and lays the groundwork for future investigations in diverse species. We agree that more causal inference would be valuable to understand the evolutionary pressures acting on MSH6 histone reader presence/absence. Indeed, we prioritized the conservative approach of testing hypotheses with strict phylogenetically constrained contrasts. While we observed highly significant associations between histone readers and genomic traits like intron content, associations with life history traits were only significant before accounting for phylogeny. It is possible that this is due to a lack of power because such traits are only available in limited taxa. In the revised manuscript, we aim to clarify potential causes, outline future experimental work beyond the scope of this individual study, and argue that this work highlights the need to catalog trait diversity at broader phylogenetic scales. We also address other valuable suggestions in the revised manuscript.