Fast evolutionary turnover and overlapping variances of sex-biased gene expression patterns defy a simple binary sex-classification of somatic tissues

  1. Chen Xie
  2. Sven Künzel
  3. Diethard Tautz

  • Curated by eLife

    eLife logo

    eLife Assessment

    This study presents data on sex differences in gene expression across organs of four mice taxa. The authors have generated a unique and convincing dataset that fills a gap left by previous studies. They claim that sex-biased expression in the soma can overlap between genetic males and females, and that the relevant patterns both turn over quickly over short evolutionary times and do so faster in somatic than gonadal tissues. These conclusions could largely have been predicted by extrapolating from previous findings in the field, but nevertheless demonstrating them directly is a fundamental advance.

    [Editorial note: The work was originally assessed by colleagues who are active in the field of evolution of sex differences or in areas adjacent to this field (see initial assessment at https://doi.org/10.7554/eLife.99602.2). The appeals process involved consultation with experts working in other areas of evolutionary biology. The above assessment synthesises the opinions of both sets of reviewers.]

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Abstract

The phenotypic differences between the sexes are generated by genes with sex-biased expression. These range from a few major regulators to large numbers of organ-specific effector genes in sexually mature individuals. We explore the variation and micro-evolutionary patterns of these genes in a large dataset from natural populations of sub-species and species of mice, with a particular focus on somatic organs. Within these short phylogenetic distances, we find a faster evolutionary turnover of sex-biased gene expression in somatic tissues, but not in the gonads, when compared to the turnover of non-sex-biased genes. We show that somatically expressed sex-biased genes occur mostly only in a subset of the co-expression modules of each organ and the turnover of genes between the taxa occurs often within the main modules. Given the fast evolution of somatic sex-biased expression patterns, we were interested to study the within-group variances and their evolutionary turnover. To visualize the individual variances, we have developed a sex-biased gene expression index (SBI) that represents the cumulative expression of all sex-biased genes for each individual in each organ. We find that SBI distributions range from binary patterns in the gonads to overlapping patterns in the somatic organs. They do not correlate between organs of the same individuals, thus supporting a mosaic model of somatic sex-determination of individuals. Comparison with data from humans shows fewer sex-biased genes compared to mice and strongly overlapping SBI distributions between the somatic organs of the sexes. We conclude that adult individuals are composed of a mosaic spectrum of sex characteristics in their somatic tissues that should not be cumulated into a simple binary classification.

Article activity feed

  1. eLife

    eLife Assessment

    This study presents data on sex differences in gene expression across organs of four mice taxa. The authors have generated a unique and convincing dataset that fills a gap left by previous studies. They claim that sex-biased expression in the soma can overlap between genetic males and females, and that the relevant patterns both turn over quickly over short evolutionary times and do so faster in somatic than gonadal tissues. These conclusions could largely have been predicted by extrapolating from previous findings in the field, but nevertheless demonstrating them directly is a fundamental advance.

    [Editorial note: The work was originally assessed by colleagues who are active in the field of evolution of sex differences or in areas adjacent to this field (see initial assessment at https://doi.org/10.7554/eLife.99602.2). The appeals process involved consultation with experts working in other areas of evolutionary biology. The above assessment synthesises the opinions of both sets of reviewers.]

  2. eLife

    Reviewer #4 (Public review):

    The paper by Xie et al. investigates the micro-evolutionary dynamics of sex-biased gene expression across somatic and gonadal tissues in four mouse taxa, with comparative analyses in humans. The study introduces a new metric, the Sex-Bias Index (SBI), to quantify individual-level variation in sex-biased gene expression, and explores the evolutionary turnover, variance, and adaptive evolution of these genes.

    These strengths of the paper are not in dispute:

    Novelty: The study is among the first to systematically analyze sex-biased gene expression at a micro-evolutionary scale in outbred animals, using closely related mouse taxa. This contrasts with most previous work, which focused on macro-evolutionary comparisons between distant species.

    Controlled Sampling: The use of age-matched, outbred individuals raised under standardized conditions minimizes environmental confounders, allowing for robust within- and between-taxon comparisons.

    Somatic vs. Gonadal Focus: Unlike many earlier studies that emphasized gonadal tissues, this work provides a detailed analysis of somatic organs, revealing rapid evolutionary turnover and mosaicism in sex-biased gene expression.

    Sex-Bias Index (SBI): The SBI offers a cumulative, individual-level measure of sex-biased gene expression, facilitating visualization of variance and overlap between sexes within tissues. While one can argue about whether a new metric is necessary (as the authors argue), the combination of fold-change cutoffs, non-parametric Wilcoxon tests, and FDR correction reduces false positives, addressing concerns raised in the field about inflated detection of sex-biased genes.

    Evolutionary implications: The study demonstrates that sex-biased gene expression in somatic tissues evolves more rapidly than in gonads, and that this turnover is often accompanied by signatures of adaptive protein evolution. The lack of correlation in SBI across tissues within individuals supports a mosaic model of sex-biased gene expression, challenging binary models of sexual differentiation.

    The weaknesses are already listed by previous rounds of review but I will add one more: in an attempt to be comprehensive, the writing is quite dry and the main conclusions sort of get hidden within the less important observations.

    Since the debate is mostly about what words to use to describe the importance and the strength of evidence, I thought it would be useful to directly compare this study to other studies that address the same topic:

    Naqvi et al. Science 2019 (David Page lab): Conservation, acquisition, and functional impact of sex-biased gene expression in mammals

    Oliva et al. Science 2020 (Stranger lab): The impact of sex on gene expression across human tissues

    Rodríguez-Montes et al. Science 2023 (Kaessman, Cardoso-Moreira labs)

    Let's start with the fact that all three peer studies have had a major impact. Second, although Naqvi et al. (2019) and Oliva et al. (2020) provided foundational cross-species and cross-tissue analyses of sex-biased gene expression, but did not address micro-evolutionary turnover or individual-level variance. Third, Rodríguez-Montes et al. (2023) focused on developmental and evolutionary patterns of sex-biased expression, but at a broader phylogenetic scale and without the individual-level or module-based analyses presented here. None of the peer studies addressed the possibility of mosaicism within individuals, none of them addressed the relations between expression bias and adaptive evolution. So the comparison is really a bit of an apples to oranges comparison: the peer studies are about patterns in deep phylogeny, whereas the present study is an amazing (to me) analysis of inter-individual mosaicism, which is at the heart of this kind of variation, which would totally be missed or worse misinterpreted in deep phylogenetic analyses. Having said that, in my subjective opinion, all three related papers are better written than the present one, but to me there is no question this belongs in the same pedestal as all of them.

  3. eLife

    Reviewer #5 (Public review):

    Xie et al. present a data set of impressive size to study changes in sex-biased gene expression. A clear strength that sets the study apart from previous work is the use of age-matched outbred individuals raised in the same environment, which minimizes non-genetic variance, and the comparison of closely related taxa. Also in contrast to many previous studies, while gonads, which have often been the focus of sex-biased gene expression studies, are not ignored, multiple gonadal tissues are being compared to an array of somatic tissues. The study design therefore can offer a particularly rich and nuanced view of how sex differences change across tissues and over short evolutionary times.

    I liked the idea of summarizing over the mean expression of gene sets, instead of just using numbers of DEGs for comparisons, even though the introduction of the term "Sex-Biased Index (SBI)" seems somewhat of an overkill. The summary analyses are definitely useful to visualize variability in sex-biased gene expression programs. The authors find that the expression patterns of sex-biased genes change faster than those of non-sex-biased genes - but only in somatic tissues. They also provide some evidence that this correlates with higher rates of potentially adaptive coding sequence changes in the taxa where expression is sex-biased, with the proviso that a stronger modeling framework would have made these inferences more robust.

    I was most surprised by the finding that the fast change in expression patterns is linked to different gene expression modules becoming sex-biased in the different taxa studied. This is in my eyes a remarkable observation that could not have been predicted from previous knowledge.

    The use of human GTEx and patient scRNA-seq data is a nice addition, although there are known confounding issues with these resources, given that these are not random samples and environmental conditions are uncontrolled. Nevertheless, as the human data echo the trends seen with the much more rigorous mouse data set, I do not have principal objections to this addition. Furthermore, the human data do allow the authors to conclude that only very few genes with sex-biased expression are shared in the soma of mice and humans.

    In summary, I believe that this contribution has the potential to fundamentally change how we see sex-biased gene expression differences in vertebrates, given that the author's conclusions are grounded in a data set of compelling quality and size.

  4. eLife

    Author response:

    The following is the authors’ response to the previous reviews

    Reviewer #2 (Public review):

    Summary:

    The manuscript by Xie and colleagues presents transcriptomic experiments that measure gene expression in eight different tissues taken from adult female and male mice from four species. These data are used to make inferences regarding the evolution of sex-biased gene expression across these taxa.

    Strengths:

    The experimental methods and data analysis appear appropriate. The authors promote their study as unprecedented in its size and technical precision.

    We do not understand the statement "the authors promote" as if there was a doubt about this. If there is a doubt, we welcome to see it specified.

    Weaknesses:

    The manuscript does not present a clear set of novel evolutionary conclusions. The major findings recapitulate many previous comparative transcriptomics studies - gene expression variation is prevalent between individuals, sexes, and species; and genes with sex-biased expression evolve more rapidly than genes with unbiased expression - but it is not clear how the study extends our understanding of gene expression or its evolution.

    There have been no "previous comparative transcriptomics studies" at a micro- evolutionary scale in animals, hence, we do not "replicate" these. And our contrast between somatic and gonadal patterns reveals insights that have not been recognized before, namely that gonadal sex-specific expression turnover is actually not faster that the corresponding non-sex-specific truover. We have now further clarified this distinction throughout the text and have also adapted the title of the paper accordingly.

    We agree with the overall statement that "gene expression variation is prevalent between individuals, sexes, and species" but the aspect of "sex-biased gene expression between individuals" has not been systematically analysed before in such a context.

    Concerning the statement that "genes with sex-biased expression evolve more rapidly than genes with unbiased expression", we note that this is mostly derived from gonadal data and that there is no study that has quantified this so far at a population level and between subspecies in comparison to somatic data.

    Our results show further that previous assumptions of a substantial set of genes with sex- biased expression conserved between mice and humans are due to underestimating the convergence issues when there is an extremly fast turnover of sex-biased gene expression. This has a major implication for using mice as a model for gender-speficic medicine questions in humans.

    Many gene expression differences between individual animals are selectively neutral, because these differences in mRNA concentration are buffered at the level of translation, or differences in protein abundance have no effect on cellular or organismal function. The hypothesis that sex-biased genes are enriched for selectively neutral expression differences is supported by the excess of inter-individual expression variance and inter-specific expression differences in sex-biased genes.

    This statement repeats a statement from the first round of reviews. We had added new data and extensive discussion on this topic. We do not understand why this has not been taken into account. In fact, a major strength of our paper is that it shows that most sex- biased gene expression differences are not neutral!

    There are two major issues here: to identify sex-biased gene expression in the first place, we (and all other papers in the field) use the neutral model as null-hypothesis. Genes that are not compatible with this null-hypothesis are considered sex-biased. In contrast to most previous papers, we have the possibility to take into account the variances between individuals to add an additional significance test. Hence, we can apply a much more rigorous two-step process: first a ratio-cutoff plus a Wilcoxon rank sum test with correction for multiple testing to identify significant deviations from the null-hypothesis. We have added some additional statements in the Results and Discussion sections to emphasize this.Second, by focusing on the genes that are not following a neutral model, the variance and divergences data support the action of selection, rather than neutral drift.

    A higher rate of adaptive coding evolution is inferred among sex-biased genes as a group, but it is not clear whether this signal is driven by many sex-biased genes experiencing a little positive selection, or a few sex-biased genes experiencing a lot of positive selection, so the relationship between expression and protein-coding evolution remains unclear.

    Again, there are two major issues here. First, the distribution of alpha-values shown in Figure 3B are rather homogeneous, i.e. there is not support for a scenario that the average is driven by only a few genes.

    Second, it seems that the referee wants to see an analysis where dn/ds ratios are broken down for every single gene. This has been done in previous papers, but it is now understood that this procedure is fraught with error because of the demographic contingencies inherent to natural populations that can yield wrong results for individual loci. We have added some statements to the text to clarify this further.

    It is likely that only a subset of the gene expression differences detected here will have phenotypic effects relevant for fitness or medicine, but without some idea of how many or which genes comprise this subset, it is difficult to interpret the results in this context.

    It is the basic underlying assumption for the whole research field that significantly sex- biased genes are phenotypically relevant for fitness, since they would otherwise not be sex- biased in the first place.

    Throughout the paper the concepts of sexual selection and sexually antagonistic selection are conflated; while both modes of selection can drive the evolution of sexually dimorphic gene expression, the conditions promoting and consequence of both kinds of selection are different, and the manuscript is not clear about the significance of the results for either mode of selection.

    We had explained in our previous response that our data collection was not designed to distinguish between these two processes. But given that the issue is being brought up again, we have now added some discussion on this issue.

    The manuscript's conclusion that "most of the genetic underpinnings of sex-differences show no long-term evolutionary stability" is not supported by the data, which measured gene expression phenotypes but did not investigate the underlying genetic variation causing these differences between individuals, sexes, or species.

    We agree that - under a strict definition - our use of the term "genetic underpinning" in this conclusion sentence can be criticized. The most correct term would be "transcriptional underpinnings", but of course, given that it is the current practice of the whole field to assume that "transcriptional" is part of the overall genetics, we do not consider our initial statement as incorrect. Still, we have changed the term accordingly.

    Furthermore, most of the gene expression differences are observed between sex-specific organs such as testes and ovaries, which are downstream of the sex-determination pathway that is conserved in these four mouse species, so these conclusions are limited to gene expression phenotypes in somatic organs shared by the sexes.

    Yes - correct. But the whole focus of the paper is on somatic expression, i.e. organs that share the same cell compositions. Of course, the comparison between gonadal organs is conflated by being composed of different cell types. We have extended the discussion of this point.

    The differences between sex-biased expression in mice and humans are attributed to differences in the two species effective population sizes; but the human samples have significantly more environmental variation than the mouse samples taken from age-matched animals reared in controlled conditions, which could also explain the observed pattern.

    These are indeed the two alternative explanations that we had discussed (last paragraph of the discussion section, now the penultimate paragraph).

    The smoothed density plots in Figure 5 are confusing and misleading. Examining the individual SBI values in Table S9 reveals that all of the female and male SBI values for each species and organ are non-overlapping, with the exception of the heart in domesticus and mammary gland in musculus, where one male and one female individual fall within the range of the other sex. The smoothed plots therefore exaggerate the overlap between the sexes;

    Smoothing across discrete values is an entirely standard procedure for continuous variables. It allows to visualize the inherent data trends that cannot easily be glanced from simple inspection of the actual values. This is a mathematical procedure, not an "exaggeration". We used the same smoothening procedure for all the comparisons, and it is clear that the distributions between females and males of the sex organs and a few somatic organs are well separated (non-overlapping), which serves as a control.

    in particular, the extreme variation shown in the SBI in the mammary glands in spretus females and spicilegus males is hard to understand given the normalized values in Table S3. The R code used to generate the smoothed plots is not included in the Github repository, so it is not possible to independently recreate those plots from the underlying data.

    We apologize that there was indeed an error in the Figure - the columns for SPR and SPI were accidentally interchanged. We have corrected this figure. Generally, the smoothened patterns we show are easily verified by looking up the respective primary values. We apologize that the code lines for the plots were accidentally omitted. We have used a standard function from ggplot2: geom_density, with "adjust=3, alpha=0.5" for all plots and included this description in the Methods. We have now added this to the R code in the GitHub repository.

    The correlations provided in Table S9 are confusing - most of the reported correlations are 1.0, which are not recovered when using the SBI values in Table S9, and which does not support the manuscript's assertion that sex-biased gene expression can vary between organs within an individual. Indeed, using the SBI values in Table S9, many correlations across organs are negative, which is expected given the description of the result in the text.

    There is a misunderstanding here. The tables do not report correlations, but only p-values for correlations, the raw ones and the ones after corrections for multiple testing. P = 1.0 means no significant correlation. We have adjusted the caption of this table to clarify this further.

    Reviewer #3 (Public review):

    This manuscript reports interesting data on sex differences in expression across several somatic and reproductive tissues among 4 mice species or subspecies. The focus is on sex- biased expression in the somatic tissues, where the authors report high rates of turnover such that the majority of sex-biased genes are only sex-biased in one or two taxa. The authors show sex-biased genes have higher expression variance than unbiased genes but also provide some evidence that sex-bias is likely to evolve from genes with higher expression variance. The authors find that sex-biased genes (both female- and male-biased) experience more adaptive evolution (i.e., higher alpha values) than unbiased genes. The authors develop a summary statistic (Sex-Bias Index, SBI) of each individual's degree of sex- bias for a given tissue. They show that the distribution of SBI values often overlap considerably for somatic (but not reproductive) tissues and that SBI values are not correlated across tissues, which they interpret as indicating an individual can be relatively "male-like" in one tissue and relatively "female-like" in another tissue.

    This is a good summary of the data, but we are puzzled that it does not include the completely new module analysis and the finding of extremely fast evolution of sex-biased somatic gene expression compared to the gonadal one.

    Though the data are interesting, there are some disappointing aspects to how the authors have chosen to present the work. For example, their criteria for sex-bias requires an expression ratio of one sex to the other of 1.25. A reasonably large fraction of the "sex- biased genes" have ratios just beyond this cut-off (Fig. S1). A gene which has a ratio of 1.27 in taxa 1 can be declared as "sex-biased" but which has a ratio of 1.23 in taxa 2 will not be declared as "sex-biased". It is impossible to know from how the data are presented in the main text the extent to which the supposed very high turnover represents substantial changes in dimorphic expression. A simple plot of the expression sex ratio of taxa 1 vs taxa 2 would be illuminating but the authors declined this suggestion.

    Choosing a cutoff is the standard practice when dealing with continuously distributed data. As we have pointed out, we looked at various cutoff options and decided to use the present one, based on the observed data distributions. Note that some studies have used even lower ones (e.g. 1.1). To visualize the data distribution, we had provided the overall distribution of ratios, because one would have to look at many more plots otherwise. But we have now also added individual plots as Figure 1, Figure supplement 2, as requested. They confirm what is also evident from the overall plots, namely that most ratio changes are larger than the incremental values suggested by the reviewer. Note that the original data are of course also available for inspection.

    I was particularly intrigued by the authors' inference of the proportion of adaptive substitutions ("alpha") in different gene sets. The show alpha is higher for sex-biased than unbiased genes and nicely shows that the genes that are unbiased in focal taxa but sex- biased in the sister taxa also have low alpha. It would be even stronger that sex-bias is associated with adaptive evolution to estimate alpha for only those genes that are sex- biased in the focal taxa but not in the sister taxa (the current version estimates alpha on all sex-biased genes within the focal taxa, both those that are sex-biased and those that are unbiased in the sister taxa).

    We have added the respective values in the results section, but since fewer genes are involved, they are less comparable to the other sets of genes. Still, the tendencies remain.

    The author's Sex Bias Index is measured in an individual sample as: SBI = median(TPM of female-biased genes) - median(TPM of male-biased genes). This index has some strange properties when one works through some toy examples (though any summary statistic will have limitations). The authors do little to jointly discuss the merits and limitations of this metric. It would have been interesting to examine their two key points (degree of overlapping distributions between sexes and correlation across tissues) using other individual measures of sex-bias.

    We had responded to this comment before (including the explanation that it has no strange properties when one applies the normalization that is now implemented) and we have added a whole section devoted to the discussion of the merits of the SBI. We do not know which other "individual measures of sex-bias" this should be compared to. Still, we have now added a paragraph in the discussion about using PCA as an alternative to show that this would result in similar conclusions, but is technically less suitable for this purpose.

    Figure 5 shows symmetric gaussian-looking distributions of SBI but it makes me wonder to what extent this is the magic of model fitting software as there are only 9 data points underlying each distribution. Whereas Figure 5 shows many broadly overlapping distributions for SBI, Figure 6 seems to suggest the sexes are quite well separated for SBI (e.g., brain in MUS, heart in DOM).

    We use a standard fitting function in R (see above), which tries to fit a normalized distribution, but this function can also add an additional peak when the data are too heterogeneous (e.g. Mammary in Figure 7).

    Fig. S1 should be shown as the log(F/M) ratio so it is easier to see the symmetry, or lack thereof, of female and male-biased genes.

    The log will work differently for values <1, compared to values >1 when used in a single plot. We have now generated combined plots with symmetric values to allow a better comparability.

    It is important to note that for the variance analysis that IQR/median was calculated for each gene within each sex for each tissue. This is a key piece of information that should be in the methods or legend of the main figure (not buried in Supplemental Table 17).

    ​We have now moved these descriptions into the Methods section.

  5. Version published to 10.7554/elife.99602.3 on eLife
  6. Version published to 10.7554/elife.99602 on eLife
  7. Version published to 10.7554/elife.99602.2 on eLife
  8. eLife

    eLife Assessment

    This study presents useful data on sex differences in gene expression across organs of four mice taxa. While the methods and analysis are largely sound, the strength of evidence is solid only in parts and the conclusions drawn from the results are not always appropriate.

  9. eLife

    Reviewer #2 (Public review):

    Summary:

    The manuscript by Xie and colleagues presents transcriptomic experiments that measure gene expression in eight different tissues taken from adult female and male mice from four species. These data are used to make inferences regarding the evolution of sex-biased gene expression across these taxa.

    Strengths:

    The experimental methods and data analysis appear appropriate. The authors promote their study as unprecedented in its size and technical precision.

    Weaknesses:

    The manuscript does not present a clear set of novel evolutionary conclusions. The major findings recapitulate many previous comparative transcriptomics studies - gene expression variation is prevalent between individuals, sexes, and species; and genes with sex-biased expression evolve more rapidly than genes with unbiased expression - but it is not clear how the study extends our understanding of gene expression or its evolution.

    Many gene expression differences between individual animals are selectively neutral, because these differences in mRNA concentration are buffered at the level of translation, or differences in protein abundance have no effect on cellular or organismal function. The hypothesis that sex-biased genes are enriched for selectively neutral expression differences is supported by the excess of inter-individual expression variance and inter-specific expression differences in sex-biased genes. A higher rate of adaptive coding evolution is inferred among sex-biased genes as a group, but it is not clear whether this signal is driven by many sex-biased genes experiencing a little positive selection, or a few sex-biased genes experiencing a lot of positive selection, so the relationship between expression and protein-coding evolution remains unclear. It is likely that only a subset of the gene expression differences detected here will have phenotypic effects relevant for fitness or medicine, but without some idea of how many or which genes comprise this subset, it is difficult to interpret the results in this context.

    Throughout the paper the concepts of sexual selection and sexually antagonistic selection are conflated; while both modes of selection can drive the evolution of sexually dimorphic gene expression, the conditions promoting and consequence of both kinds of selection are different, and the manuscript is not clear about the significance of the results for either mode of selection.

    The manuscript's conclusion that "most of the genetic underpinnings of sex-differences show no long-term evolutionary stability" is not supported by the data, which measured gene expression phenotypes but did not investigate the underlying genetic variation causing these differences between individuals, sexes, or species. Furthermore, most of the gene expression differences are observed between sex-specific organs such as testes and ovaries, which are downstream of the sex-determination pathway that is conserved in these four mouse species, so these conclusions are limited to gene expression phenotypes in somatic organs shared by the sexes.

    The differences between sex-biased expression in mice and humans are attributed to differences in the two species effective population sizes; but the human samples have significantly more environmental variation than the mouse samples taken from age-matched animals reared in controlled conditions, which could also explain the observed pattern.

    The smoothed density plots in Figure 5 are confusing and misleading. Examining the individual SBI values in Table S9 reveals that all of the female and male SBI values for each species and organ are non-overlapping, with the exception of the heart in domesticus and mammary gland in musculus, where one male and one female individual fall within the range of the other sex. The smoothed plots therefore exaggerate the overlap between the sexes; in particular, the extreme variation shown in the SBI in the mammary glands in spretus females and spicilegus males is hard to understand given the normalized values in Table S3. The R code used to generate the smoothed plots is not included in the Github repository, so it is not possible to independently recreate those plots from the underlying data.

    The correlations provided in Table S9 are confusing - most of the reported correlations are 1.0, which are not recovered when using the SBI values in Table S9, and which does not support the manuscript's assertion that sex-biased gene expression can vary between organs within an individual. Indeed, using the SBI values in Table S9, many correlations across organs are negative, which is expected given the description of the result in the text.

  10. eLife

    Reviewer #3 (Public review):

    This manuscript reports interesting data on sex differences in expression across several somatic and reproductive tissues among 4 mice species or subspecies. The focus is on sex-biased expression in the somatic tissues, where the authors report high rates of turnover such that the majority of sex-biased genes are only sex-biased in one or two taxa. The authors show sex-biased genes have higher expression variance than unbiased genes but also provide some evidence that sex-bias is likely to evolve from genes with higher expression variance. The authors find that sex-biased genes (both female- and male-biased) experience more adaptive evolution (i.e., higher alpha values) than unbiased genes. The authors develop a summary statistic (Sex-Bias Index, SBI) of each individual's degree of sex-bias for a given tissue. They show that the distribution of SBI values often overlap considerably for somatic (but not reproductive) tissues and that SBI values are not correlated across tissues, which they interpret as indicating an individual can be relatively "male-like" in one tissue and relatively "female-like" in another tissue.

    Though the data are interesting, there are some disappointing aspects to how the authors have chosen to present the work. For example, their criteria for sex-bias requires an expression ratio of one sex to the other of 1.25. A reasonably large fraction of the "sex-biased genes" have ratios just beyond this cut-off (Fig. S1). A gene which has a ratio of 1.27 in taxa 1 can be declared as "sex-biased" but which has a ratio of 1.23 in taxa 2 will not be declared as "sex-biased". It is impossible to know from how the data are presented in the main text the extent to which the supposed very high turnover represents substantial changes in dimorphic expression. A simple plot of the expression sex ratio of taxa 1 vs taxa 2 would be illuminating but the authors declined this suggestion.

    I was particularly intrigued by the authors' inference of the proportion of adaptive substitutions ("alpha") in different gene sets. The show alpha is higher for sex-biased than unbiased genes and nicely shows that the genes that are unbiased in focal taxa but sex-biased in the sister taxa also have low alpha. It would be even stronger that sex-bias is associated with adaptive evolution to estimate alpha for only those genes that are sex-biased in the focal taxa but not in the sister taxa (the current version estimates alpha on all sex-biased genes within the focal taxa, both those that are sex-biased and those that are unbiased in the sister taxa).

    The author's Sex Bias Index is measured in an individual sample as: SBI = median(TPM of female-biased genes) - median(TPM of male-biased genes). This index has some strange properties when one works through some toy examples (though any summary statistic will have limitations). The authors do little to jointly discuss the merits and limitations of this metric. It would have been interesting to examine their two key points (degree of overlapping distributions between sexes and correlation across tissues) using other individual measures of sex-bias.

    Figure 5 shows symmetric gaussian-looking distributions of SBI but it makes me wonder to what extent this is the magic of model fitting software as there are only 9 data points underlying each distribution. Whereas Figure 5 shows many broadly overlapping distributions for SBI, Figure 6 seems to suggest the sexes are quite well separated for SBI (e.g., brain in MUS, heart in DOM).

    Fig. S1 should be shown as the log(F/M) ratio so it is easier to see the symmetry, or lack thereof, of female and male-biased genes.
    It is important to note that for the variance analysis that IQR/median was calculated for each gene within each sex for each tissue. This is a key piece of information that should be in the methods or legend of the main figure (not buried in Supplemental Table 17).

  11. eLife

    Author response:

    The following is the authors’ response to the current reviews.

    We are disappointed that the reviewers do not acknowledge that our data constitute a major step forward for the field. We will prepare a revised version that takes care of the remaining small issues concerning the technical descriptions and a detailed response to the current round of comments. We will also add a summary of the major new findings of our study.

    The following is the authors’ response to the original reviews.

    We appreciate the time of the reviewers and their detailed comments, which have helped to improve the manuscript.

    Our study presents the largest systematic dataset so far on the evolution of sex-biased gene expression in animals. It is also the first that explores the patterns of individual variation in sex-biased gene expression and the SBI is an entirely new procedure to directly visulize these variance patterns in an intuitive way.

    Also, we should like to point out that our study contradicts recent conclusions that had suggested that a substantial set of sex-biased genes has conserved functions between humans and mice and that mice can therefore be informative for gender-specific medicine studies. Our data suggest that only a very small set of genes are conserved in their sex-biased expression between mice and humans in more than one organ.

    In the revised version we have made the following major updates:

    - added a rate comparison of gene regulation turnover between sex-biased and non-sex-biased genes

    - added additional statistics to the variance comparisons and selection tests

    - added a regulatory module analysis that shows that much of the gene turnover happens within modules

    - added a mosaic pattern analysis that shows the individual complexity of sex-biased patterns

    - extended introduction and discussion

    Reviewer #1 (Public Review):
    The authors describe a comprehensive analysis of sex-biased expression across multiple tissues and species of mouse. Their results are broadly consistent with previous work, and their methods are robust, as the large volume of work in this area has converged toward a standardized approach.

    I have a few quibbles with the findings, and the main novelty here is the rapid evolution of sex-biased expression over shorter evolutionary intervals than previously documented, although this is not statistically supported. The other main findings, detailed below, are somewhat overstated.

    (1) In the introduction, the authors conflate gametic sex, which is indeed largely binary (with small sperm, large eggs, no intermediate gametic form, and no overlap in size) with somatic sexual dimorphism, which can be bimodal (though sometimes is even more complicated), with a large variance in either sex and generally with a great deal of overlap between males and females. A good appraisal of this distinction is at . This distinction in gene expression has been recognized for at least 20 years, with observations that sex-biased expression in the soma is far less than in the gonad.

    For example, the authors frame their work with the following statement:

    "The different organs show a large individual variation in sex-biased gene expression, making it impossible to classify individuals in simple binary terms. Hence, the seemingly strong conservation of binary sex-states does not find an equivalent underpinning when one looks at the gene-expression makeup of the sexes"

    The authors use this conflation to set up a straw man argument, perhaps in part due to recent political discussions on this topic. They seem to be implying one of two things. a) That previous studies of sex-biased expression of the soma claim a binary classification. I know of no such claim, and many have clearly shown quite the opposite, particularly studies of intra-sexual variation, which are common - see https://doi.org/10.1093/molbev/msx293, https://doi.org/10.1371/journal.pgen.1003697, https://doi.org/10.1111/mec.14408, https://doi.org/10.1111/mec.13919, https://doi.org/10.1111/j.1558-5646.2010.01106.x for just a few examples. Or b) They are the first to observe this non-binary pattern for the soma, but again, many have observed this. For example, many have noted that reproductive or gonad transcriptome data cluster first by sex, but somatic tissue clusters first by species or tissue, then by sex (https://doi.org/10.1073/pnas.1501339112, https://doi.org/10.7554/eLife.67485)

    Figure 4 illustrates the conceptual difference between bimodal and binary sexual conceptions. This figure makes it clear that males and females have different means, but in all cases the distributions are bimodal.

    I would suggest that the authors heavily revise the paper with this more nuanced understanding of the literature and sex differences in their paper, and place their findings in the context of previous work.

    We are sorry that our introduction seems to have been too short to make our points sufficiently clear. Of course, overlapping somatic variation has been shown for morphological characters, but we were aiming to assess this at the sex-biased transcriptome level. Previous studies looking at sex-biased genes were usually limited by the techniques that were available at their times, resulting in a focus on gonads in most studies and almost all have too few individuals included to study within-group variation. We detail this below for the papers that are mentioned by the referee. In view of this, we cite them now as examples for the prevalent focus on gonadal comparisons in most studies. Only Scharmann et al. 2021 on plant leaf dimorphism is indeed relevant for our study with respect to its general findings and we make now extensive reference to it. In addition, we have generally modified the introduction and substantially extended the discussion to make our points clear.

    Snell-Rood 2010: the paper focuses on sex-specific morphological structures in beetles. It samples six somatic tissues for four individuals each of each class. Analysis is done via microarray hybridizations. While categorial differences were traced, variability between individuals was not discussed. By today´s standards, microarrays have anyway too much technical variability to even consider such a discussion.

    Pointer et al. 2013: this paper studies three sexual phenotypes in a bird species, females, dominant males and subordinate males. Tissues include telencephalon, spleen and left gonad. The focus of the analysis is on the gonads, since only few sex-biased genes were found in spleen and brain (according to suppl. Table S1, 0 for the spleen and 2 for the brain). No inferences could be made on somatic variation.

    Harrison 2015: this paper focuses on gonads plus spleen in six bird species with between 2-6 individuals for each sex collected. In the spleen, only one female biased gene and no male biased gene was detected. Hence, the data do not allow to infer patterns of somatic variation.

    Dean et al. 2016: this paper compares four categories of fish caught around nests, with four to seven individuals per category. Only gonads were analyzed, hence no inferences could be made about somatic variability between individuals.

    Cardoso et al. 2017: this paper test categories of fish with alternative reproductive tactics based on brain transcriptomes. While it uses 9-10 individuals per category, it uses pools for sequencing with two pools per category. This does not allow to make any inference on individual variation.

    Todd et al 2017: this paper focuses on three categories of a fish species, females and dominant and sneaker males. It uses brain and gonads as samples with five individuals each for each category. For the brain, more different genes were found between the two types of males, rather than between females and males (3 and 9 respectively). The paper focuses on individual gene descriptions and does not mention somatic variation.

    Scharmann 2021: the paper focuses on 10 species of plants with sexually dimorphic leafs. 5-6 individuals were sampled per sex. The major finding is that sex-biased gene expression does not correlate with the degree of sexual dimorphism of the leafes. The study shows also a fast evolution of sex-biased expression and states that signatures of adaptive evolution are weak. But it does not discuss variance patterns within populations.

    (2) The authors also claim that "sexual conflict is one of the major drivers of evolutionary divergence already at the early species divergence level." However, making the connection between sex-biased genes and sexual conflict remains fraught. Although it is tempting to use sex-biased gene expression (or any form of phenotypic dimorphism) as an indicator of sexual conflict, resolved or not, as many have pointed out, one needs measures of sex-specific selection, ideally fitness, to make this case (https://doi.org/10.1086/595841, 10.1101/cshperspect.a017632). In many cases, sexual dimorphism can arise in one sex only without conflict (e.g. 10.1098/rspb.2010.2220). As such, sex-biased genes alone are not sufficient to discriminate between ongoing and resolved conflict.

    We imply sexual conflict as a driver of genomic divergence patterns in a similar way as it has been done by many authors before (e.g. Mank 2017a, Price et al. 2023, Tosto et al. 2023). While we fully appreciate the point of the referee, we do not really see where we deviate from the standard wording that is used in the context of genomic data. In such data, it is of course usually assumed that they represent solved conflicts (Figure 1D in Cox and Calsbeek) where selection differentials would not be measurable anyway. (Please note also that the phylogenetic approach used in Oliver and Monteiro 2010 becomes rather problematic in view of introgressive hybridization patterns in butterflies), We have extended the discussion to address this.

    (3) To make the case that sex-biased genes are under selection, the authors report alpha values in Figure 3B. Alpha value comparisons like this over large numbers of genes often have high variance. Are any of the values for male- female- and un-biased genes significantly different from one another? This is needed to make the claim of positive selection.

    Sorry, we had accidentally not included the statistics in the final version of the figure. We have added this now in the supplementary table but have also generally changed the statistical approach and the design of the figure.

    Reviewer #2 (Public Review):

    The manuscript by Xie and colleagues presents transcriptomic experiments that measure gene expression in eight different tissues taken from adult female and male mice from four species. These data are used to make inferences regarding the evolution of sex-biased gene expression across these taxa. The experimental methods and data analysis are appropriate; however, most of the conclusions drawn in the manuscript have either been previously reported in the literature or are not fully supported by the data.

    We are not aware of any study that has analyzed somatic sex-biased expression in such a large and taxonomically well resolved closely related taxa of animals. Only the study by Scharman et al. 2021 on plant leaves comes close to it, but even this did not specifically analyze the intragroup variation aspects. Of course, some of our results confirm previous conclusions, but we should still like to point out that they go far beyond them.

    There are two ways the manuscript could be modified to better strengthen the conclusions.

    First, some of the observed differences in gene expression have very little to no effect on other phenotypes, and are not relevant to medicine or fitness. Selectively neutral gene expression differences have been inferred in previous studies, and consistent with that work, sex-biased and between-species expression differences in this study may also be enriched for selectively neutral expression differences. This idea is supported by the analysis of expression variance, which indicates that genes that show sex-biased expression also tend to show more inter-individual variation. This perspective is also supported by the MK analysis of molecular evolution, which suggests that positive selection is more prevalent among genes that are sex-biased in both mus and dom, and genes that switch sex-biased expression are under less selection at the level of both protein-coding sequence and gene expression.

    We have now revisited these points by additional statistical analysis of the variance patterns and an extended discussion under the heading "Neutral or adaptive?".

    As an aside, I was confused by (line 176): "implying that the enhanced positive selection pressure is triggered by their status of being sex-biased in either taxon." - don't the MK values suggest an excess of positive selection on genes that are sex-biased in both taxa?

    There are different sets of genes that are sex-biased in these two taxa - hence this observation is actually a strong argument for selection on these genes. We have changed the correspondiung text to make this clearer.

    Without an estimate of the proportion of differentially expressed genes that might be relevant for broader physiological or organismal phenotypes, it is difficult to assess the accuracy and relevance of the manuscript's conclusions. One (crude) approach would be to analyze subsets of genes stratified by the magnitude of expression differences; while there is a weak relationship between expression differences and fitness effects, on average large gene expression differences are more likely to affect additional phenotypes than small expression differences.

    We agree that it remains a challenge to show functional effects for the sex-biased genes. The argument that they should have a function is laid out above (and stated in many reviews on the topic). To use the expression level as a proxy of function does not seem justified, given the current literature. For example, genes that are highly conected in modules are not necessrily highly expressed (e.g. transcription factors). Also, genes may be highly expressed in a rare cell type of an organ and have an important funtion there, but this would not show up across the RNA of the whole organ. The most direct functional relationship between sex-biased expression and phenotype comes from the human data in Naqvi et al. 2019 - which we had cited.

    Another perspective would be to compare the within-species variance to the between-species variance to identify genes with an excess of the latter relative to the former (similar logic to an MK test of amino acid substitutions).

    Such an analysis was actually our intial motivation for this study. However, the new (and surprising!) result is that the status of being sex-biased shows such a high turnover that not many genes are left per organ where one could even try to make such a test. However, we have extended the variance analysis with reciprocal gene sets (as we had done it for the MK test) and extended the discussion on the topic, including citation of our prior work on these questions.

    Second, the analysis could be more informative if it distinguished between genes that are expressed across multiple tissues in both sexes that may show greater expression in one sex than the other, versus genes with specialized function expressed solely in (usually) reproductive tissues of one sex (e.g. ovary-specific genes). One approach to quantify this distinction would be metrics like those used defined by [Yanai I, et al. 2005. Genome-wide midrange transcription profiles reveal expression-level relationships in human tissue specification. Bioinformatics 21:650-659.] These approaches can be used to separate out groups of genes by the extent to which they are expressed in both sexes versus genes that are primarily expressed in sex-specific tissue such as testes or ovaries. This more fine-grained analysis would also potentially inform the section describing the evolution/conservation of sex-biased expression: I expect there must be genes with conserved expression specifically in ovaries or testes (these are ancient animal structures!) but these may have been excluded by the requirement that genes be sex-biased and expressed in at least two organs.

    Given that our study focuses on somatic sex-biased genes, we refrain from a comparative analysis of genes that are only expressed in the sex-organs in this paper. With respect to sharing of sex-biased gene expresssion between the somatic tissues, we show in Figure 8 that there are only very few of them (8 female-biased and 3 male-biased). A separate statistical treatment is not possible for this small set of genes.

    There are at least three examples of statements in the discussion that at the moment misinterpret the experimental results.

    The discussion frames the results in the context of sexual selection and sexually antagonistic selection, but these concepts are not synonymous. Sexual selection can shape phenotypes that are specific to one sex, causing no antagonism; and fitness differences between males and females resulting from sexually antagonistic variation in somatic phenotypes may not be acted on by sexual selection. Furthermore, the conditions promoting and consequence of both kinds of selection can be different, so they should be treated separately for the purposes of this discussion.

    We cannot make such a distinction for gene expression patterns - and we are not aware that this was done before in the literature (except gene expression was directly linked to a morphological structure). We have updated this discussion accordingly.

    The discussion claims that "Our data show that sex-biased gene expression evolves extremely fast" but a comparison or expectation for the rate of evolution is not provided. Many other studies have used comparative transcriptomics to estimate rates of gene expression evolution between species, including mice; are the results here substantially and significantly different from those previous studies? Furthermore, the experimental design does not distinguish between those gene expression phenotypes that are fixed between species as compared to those that are polymorphic within one or more species which prevents straightforward interpretation of differences in gene expression as interspecific differences.

    Our statement was in relation to the comparison between somatic and gondadal gene turnover, as well as the comparison to humans. We have now included an additional analysis for a direct comparison with non-sex-biased genes in the same populations (Figure 2B). Note that gene expression variances cannot get fixed anyway, they can only become different in average and magnitude.

    The conclusion that "Our results show that most of the genetic underpinnings of sex differences show no long-term evolutionary stability, which is in strong contrast to the perceived evolutionary stability of two sexes" - seems beyond the scope of this study. This manuscript does not address the genetic underpinnings of sex differences (this would involve eQTL or the like), rather it looks at sex differences in gene expression phenotypes.

    This comes back to the points discussed above about the validity to infer function from sex-biased expression. We have updated the text to clarify this.

    Simply addressing the question of phenotypic evolutionary stability would be more informative if genes expressed specifically in reproductive tissues were separated from somatic sex-biased genes to determine if they show similar patterns of expression evolution.

    Our study is generally focused on somatic gene expression. The comparison with reproductive tissues serves merely as a reference. Since they are of course very different tissues, they should not be compared with each other in the same way. We have now specifically addressed this point in the discussion.

    Reviewer #3 (Public Review):

    This manuscript reports some interesting and important patterns. The results on sex-bias in different tissues and across four taxa would benefit from alternative (or additional) presentation styles. In my view, the most important results are with respect to alpha (fraction of beneficial amino acid changes) in relation to sex-bias (though the authors have made this as a somewhat minor point in this version).

    The part that the authors emphasize I don't find very interesting (i.e., the sexes have overlapping expression profiles in many nongonadal tissues), nor do I believe they have the appropriate data necessary to convincingly demonstrate this (which would require multiple measures from the same individual).

    This is the first study that reports such overlaps and we show that this is not always the case (e.g. liver and kidney data in mice). We are not aware of any preditions of how such patterns would look like and how they would evolve - why should such a new finding not be interesting? Concerning the appropriateness of the data we do not agree with the point the referee makes - see response below.

    This study reports several interesting patterns with respect to sex differences in gene expression across organs of four mice taxa. An alternative presentation of the data would yield a clearer and more convincing case that the patterns the authors claim are legitimate.

    I recommend that the authors clarify what qualifies as "sex-bias".

    This is defined by the statistical criteria that we have applied, following the general standard of papers on this topic.

    Recommendations for the authors:

    Reviewer #1 (Recommendations For The Authors):

    (1) "However, already Darwin has pointed out that the phenotypes of the sexes should evolve fast". I think the authors mean that Darwin was quick to point out that sex-specific phenotypes evolve quickly".

    We have modified this text part.

    (2) Non-gonadal is more often referred to as somatic. I would encourage the authors to use this more common term for accessibility.

    We have adopted this term

    (3) Figure 5 is interesting, however, it is difficult to know whether the decreased bimodality in humans compared to mice is biological or technical due to the differences in the underlying data. For example, the mouse samples tightly controlled age and environmental conditions within each species. It is not possible to do that with human samples, and there are very good reasons to think that these factors will affect variance in both sexes.

    Yes, this is certainly true and we know this also from other comparative data between mice and humans. Still, this is human reality vs mouse artificialness. We pick this now up in the discussion.

    (4) Line 273. The large numbers of cells needed for single-cell analysis require that most studies pool multiple samples, however these pools are helpful in themselves. This approach was used by https://doi.org/10.1093/evlett/qrad013 to quantify the degree of sex-bias within cell types across multiple tissues and to compare how bulk and single-cell sex-bias measures compare. Sex-bias in some somatic cell types was very high, even when bulk sex-bias in those tissues was not. This suggests that the bulk data the authors use in this study may in fact obscure the pattern of sex-bias.

    Yes, we agree, and this is exactly how we did the analysis and interpretation, based on the cited paper.

    (5)- Line 379 "Total RNAs were" should be "Total RNA was"

    Corrected

    References cited in this review and which should be included in the manuscript :

    Sam L Sharpe, Andrew P Anderson, Idelle Cooper, Timothy Y James, Alexandra E Kralick, Hans Lindahl, Sara E Lipshutz, J F McLaughlin, Banu Subramaniam, Alicia Roth Weigel, A Kelsey Lewis, Sex and Biology: Broader Impacts Beyond the Binary, Integrative, and Comparative Biology, Volume 63, Issue 4, October 2023, Pages 960-967.

    Included

    Masculinization of Gene Expression Is Associated with Exaggeration of Male Sexual Dimorphism Pointer MA, Harrison PW, Wright AE, Mank JE (2013) Masculinization of Gene Expression Is Associated with Exaggeration of Male Sexual Dimorphism. PLOS Genetics 9(8): e1003697.

    Included

    Erica V Todd, Hui Liu, Melissa S Lamm, Jodi T Thomas, Kim Rutherford, Kelly C Thompson, John R Godwin, Neil J Gemmell, Female Mimicry by Sneaker Males Has a Transcriptomic Signature in Both the Brain and the Gonad in a Sex-Changing Fish, Molecular Biology and Evolution, Volume 35, Issue 1, January 2018, Pages 225-241.

    Included

    Cardoso SD, Gonçalves D, Goesmann A, Canário AVM, Oliveira RF. Temporal variation in brain transcriptome is associated with the expression of female mimicry as a sequential male alternative reproductive tactic in fish. Mol Ecol. 2018; 27: 789-803.

    Included

    Dean, R., Wright, A.E., Marsh-Rollo, S.E., Nugent, B.M., Alonzo, S.H. and Mank, J.E. (2017), Sperm competition shapes gene expression and sequence evolution in the ocellated wrasse. Mol Ecol, 26: 505-518.

    Included

    Emilie C. Snell‐Rood, Amy Cash, Mira V. Han, Teiya Kijimoto, Justen Andrews, Armin P. Moczek, DEVELOPMENTAL DECOUPLING OF ALTERNATIVE PHENOTYPES: INSIGHTS FROM THE TRANSCRIPTOMES OF HORN‐POLYPHENIC BEETLES, Evolution, Volume 65, Issue 1, 1 January 2011.

    Not included, since its technical approach is not really comparable

    Harrison PW, Wright AE, Zimmer F, Dean R, Montgomery SH, Pointer MA, Mank JE (2015) Sexual selection drives evolution and rapid turnover of male gene expression. Proceedings of the National Academy of Sciences, USA 112: 4393-4398.

    Included

    Mathias Scharmann, Anthony G Rebelo, John R Pannell (2021) High rates of evolution preceded shifts to sex-biased gene expression in Leucadendron, the most sexually dimorphic angiosperms eLife 10:e67485.

    Included

    Sexually Antagonistic Selection, Sexual Dimorphism, and the Resolution of Intralocus Sexual Conflict. Robert M. Cox and Ryan Calsbeek , The American Naturalist 2009 173:2, 176-187.

    Included

    Ingleby FC, Flis I, Morrow EH. Sex-biased gene expression and sexual conflict throughout development. Cold Spring Harb Perspect Biol. 2014 Nov 6;7(1):a017632.

    Included

    Oliver JC, Monteiro A 2011. On the origins of sexual dimorphism in butterflies. Proc Biol Sci 278: 1981-1988.

    Included

    Iulia Darolti, Judith E Mank, Sex-biased gene expression at single-cell resolution: cause and consequence of sexual dimorphism, Evolution Letters, Volume 7, Issue 3, June 2023, Pages 148-156.

    Included

    Reviewer #2 (Recommendations For The Authors):

    I am concerned the smoothed density plots in Figure 4 may be providing a misleading sense of the distributions since each distribution is inferred from only 9 values. A boxplot might better represent the data to the reader.

    Boxplots with 9 values are much more difficult to interpret for a reader, this is the very reason why one tends to smoothen them. In this way, they also become similar to the standard plots that are used for showing morphological variation between the sexes. Note that the original data are availble for the individual values, if these are of special interest in some cases. In addition, our new “mosaic” analysis (Figure 6) provides another presentation for readers.

    Line 235: "the overall numbers are lower" I assume this is the number of genes included in the analyses, but this should be explicitly stated.

    Clarified in the text

    The analysis of gene expression from different brain regions in control individuals from the Alzheimer's study (line 273) suffers from low power and it is not clear to me how much taking samples from different brain regions eliminates the issue of different cell types within a sample (the stated motivation for this analysis). While I support publishing negative results, this section does not feel like it adds much to the manuscript and could be cut in my opinion.

    This is actually a study on single cell types, differentiating each of them. We are sorry that the text was apparently unclear about this. Given that there are studies that show the importance of looking at single cell data, we still think that is a suitable analysis. We have updated the text to make it clearer.

    It might be useful to separate out X-linked genes from autosomal genes to see if they show consistent patterns with regard to sex-bias.

    We have added this information in suppl. Table S2 and include some description in the text.

    Reviewer #3 (Recommendations For The Authors):

    Comments follow the order of the Results section:

    (1) The latter half of this line in the Methods is too vague to be helpful: "We have explored a range of cutoffs and found that a sex-bias ratio of 1.25-fold difference of MEDIAN expression values combined with a Wilcoxon rank sum test and Benjamini-Hochberg FDR correction (using FDR <0.1 as cutoff) (Benjamini & Hochberg, 1995) yields the best compromise between sensitivity and specificity". What precisely is meant by "the best compromise between sensitivity and specificity"?

    We explain now that this was based on pre-tests with comparing randomized with actual data. However, we agree that this is in the end a subjective decision, but there is no single standard used in the literature, especially when somatic organs are included. We consider our criteria as rather stringent.

    (2) The 1.25 number for sex bias is, ultimately, an arbitrary cut-off. It is common in this literature to choose some arbitrary level and, in this sense, the authors are following common practice. The choice of 1.25 should be stated in the main text as it is a lower (but not reasonable) value than has been used in many other papers.

    It is not only the cutoff, but also the Wilcoxon test and FDR correction that defines the threshold. See also comment above.

    (3) In truth, dimorphism is continuous rather than discrete (i.e, greater or less than 1.25 fold different). Thus, where possible it would be useful to present results in a fashion that allows readers to see the continuous range of ratios rather than having to worry about whether the patterns are due to the rather arbitrary choices of how genes were binned into sex-bias categories.

    It is necessary to work with cutoffs in such cases - and this is the usual practice for any such paper. But we provide now in Figure 1 Figure supplement 1 plots with the female/male ratio distributions.

    a) Number of genes that are female- / male-biased. I would like to be able to see a version of Figure 1 showing the full distribution of TPM ratios rather than bar graphs of the numbers of (arbitrarily defined) female- and male-biased genes. This will be, of course, a larger figure (a full distribution rather than 2 bars for each species for each organ) and so could be relegated to Supplementary Material (assuming the message of that figure is the same as the current Figure 1).

    This is a very unusual request, given that no other paper has done this either. It would indeed result in a non-managable figure size, or many separate figures that would be difficult to scrutinize. Note that there would be one plot of two (female and male) TPM distributions for each sex-biased gene in each organ and each taxon, leading to hundreds of thousands of plots. We think that by providing the general distributions as plots (see above), and the original data as supplements is sufficient.

    b) Turnover of genes with sex bias. This important issue is addressed in Figure 2. First, it is not precisely clear what "percentages of sums of shared genes for any pairwise comparison" in Figure 2 legend means and no further detail is given in the Methods; this must be made clearer or the info in Figure 2 is meaningless. Regardless, this approach again relies heavily on the arbitrary criterion of defining sex-bias. Thus, I would like to see correlation plots of the log(TPM ratio) between taxa as done in the classic multispecies fly paper of Zhang et al. 2007. In Figure 2 it is quite clear that male-biased genes evolve with respect to sex bias more rapidly than female-biased genes.

    We have provided a better explanation of this analysis. Note that the Zhang et al. 2007 paper was not focussing on somatic expression and covers a much broader evolutionary spectrum. Hence, the results are not comparable. Also, we doubt that it would be so helpful to generate a huge figure with all these plots.

    (4) Is there a simpler explanation for the results in the "Variance patterns" section? The total variance for any variable can be decomposed into the variance within and among "groups". If we use "sex" as the group, then there are genes - labelled sex-biased genes - that were identified as such, in essence, because they have high among-group variance. Given that we then know a priori at the start of this section of sex-biased genes have high among-group variance, is it at all surprising that they have higher total variance than the unbiased genes (which we know a priori have low among-group variance)? Perhaps I misunderstood the point of this section. Maybe it would be more meaningful to examine the WITHIN-SEX variance (averaged across the two sexes) instead.

    We did calculate IQR/median (“normalized variance”) with the nine mice for each gene and each sex in each organ, hence sex is not a variance factor in this calculation. The algorithm steps are outlined in suppl. Table S17. We have now also added a variance calculation for reciprocal gene sets and added an extended discussion of these results.

    (5) Analysis of alpha for sex-biased genes. This was the most interesting part of this manuscript to me.

    (a) More information about what SNVs were used is required.

    i. Were only sites where SPR was fixed used? (If not, how was polarization done?)

    ii. Were sites only considered diverged if they were fixed for different bases in DOM and MUS? (If not, what was the criteria?)

    iii. Using, say, DOM as the focal species, a site must be polymorphic in DOM. But did its status (polymorphic/fixed) in MUS matter?

    We have added a more detailed description on this in the Methods section. For the direct answers of the three questions: (i) yes; (ii) yes; (iii) no, considering that DOM and MUS are two subspecies of Mus musculus separating recently, a variant might occur before separating and there might be gene flow between them.

    (b) A particularly interesting part of the analysis is the investigation of alpha for genes that are NOT sex-biased in one taxa but are sex-biased in the other. At the moment (as I understand it), alpha is only calculated for these genes in the taxa where they are NOT sex-biased (and this alpha value can be compared to the alpha of sex-biased genes and of unbiased genes in that taxa). I would like to see both sets of genes (set 1: those sex-biased in MUS and not in DOM; set 2: those sex-biased DOM and not in MUS) analyzed in each of the 2 species, with results presented in a 2x2 table.

    By definition of these categories, these genes are sex-biased in the respective other taxon, hence the values are already in the table. They are named as “reciprocal”.

    (c) No confidence intervals are given for the alpha values, despite the legend of Figure 3 referring to them.

    These were accidentally omitted - we now included the full table in suppl. Table S6; Figure 3 was modified to show violin plots of the bootstrap distributions

    The author's creation and use of a "sex-bias index" (SBI). My greatest skepticism of this manuscript is with respect to the value of their manufactured index, SBI. Of course, it is possible to create such an index but does this literature really need this index or does this just add to the "clutter" in the literature for this field? Is it helping to illuminate important patterns? This index is presumably some attempt to quantify how "male-like" or "female-like" overall expression is for a given individual (for a given organ). It is calculated as SBI = (MEDIAN of all female-biased tpm) - (MEDIAN of all male-biased tpm).

    (6) A main result that comes from this is that the sexes tend to overlap for these values for most nongonad tissues but are clearly distinct for gonadal tissues. I do not think this result would come as a surprise to almost anyone and I'm far from convinced that this metric is a good way to quantify that point. Let's consider testes vs. ovaries. Compared to non-gonadal tissues, I am reasonably certain that not only are there many more genes that are classified as "sex-biased" in gonads but also the magnitude of sex-bias among these genes is typically much greater than it is for the so-called sex-biased genes in nongonadal tissue (density plots requested in #3a would make this clear). In other words, males and females are, on average, very different with respect to expression in gonads so even allowing for variation within each sex will still result in a clear separation of all individuals of the two sexes. In contrast, males and females are, on average, much less different in, say, heart so when we consider the variation within each sex, there is overlap. One could imagine a variety of different metrics which could be used to make this point. The merits of "SBI" are unclear. It is a novel metric and its properties are poorly understood. (A simple alternative would be looking at individual scores along the axis separating mean/median males and females; almost certainly, for gonads, this would be very similar to PC scores for PC1.)

    As throughout the text, we use gonadal comparisons only as general reference, not as the main result. The main result that we are stressing is the fast turnover of these patterns, including from binary to overlapping for kidney and liver in mouse. We consider this as a new finding. If it comes "not to a surprise to anyone", isn´t it great that one does not have to guess anymore but has finally real data on this?

    We have now also added a mosaic analysis to show that the SBI can be used as summary measure in different presentations.

    The use of a single PC axis is no good alternative, since it throws away the information from the other axis.

    We have now included an explicit discussion on the usefulness of the SBI.

    (7) For simplicity, let's assume all males are identical and all females are identical. Let's imagine that heart and kidney have the exact same set of sex-biased genes. There are 20 female-biased genes; they all happen to be identical in expression level (within tissue) and look like this:

    Female TPM Male TPM TPM ratio (F:M)

    Heart 4 2 2

    Kidney 40 20 2

    And there are 20 male-biased genes that look like this:

    Female TPM Male TPM TPM ratio (F:M)

    Heart 1 3 1/3

    Kidney 10 30 1/3

    Most people would describe these two tissues as equally sex-biased.

    However, the SBIs would be:

    Female SBI Male SBI Sex difference (F - M)

    Heart 4-1 = 3 2 - 3 = -1 4

    Kidney 40-10 =30 20-30 = -10 40

    Is it a desirable property that by this metric these two tissues have wildly different SBI values for each sex as well as for the difference between sexes? (At the very least, shouldn't you make readers aware of these strange properties of SBI so they can decide how much value they put into them?)

    Actually, in this example the simple ratio between the expression levels has a strange property, since it does not reflect a much higher expression of the relevant genes in the kidney. The SBI is actually more suitable for making such cases clear. Of course, this is under the assumption that expression level has a meaning for the phenotype, but this is the general assumption for all RNA-Seq experiment comparisons.

    (8) With respect to Figure 4, why do females often have mean SBI values close to zero or even negative (e.g., kidney, mammary glands)? Is this simply because the female-biased genes tend to have lower TPM than the male-biased genes? It seems that the value zero for this metric is really not very biologically meaningful because this metric is a difference of two things that are not necessarily expected to be equal.

    This is the extra information about the expression levels that is gained via the SBI values (see comment above). However, we noticed that people can get confused about this. We have now added a re-scaling step to focus completely on the variance information in these plots.

    (9) Interpreting variances. A substantial fraction of the latter half of the manuscript focuses on interpreting variances among individual samples. This is problematic because there is no replication within individuals (i.e.., "repeatability"), thus it is impossible to infer the extent of observed variance among individuals of a given group (e.g., among females) is due to true biological differences among individuals or is simply due to noise (i.e., "measurement error" in the broad sense). Is the larger variance for mammary glands than liver or gonads just due to measurement error? What is the evidence?

    This point was of course a major issue during the times where microarrays were used for transcriptome studies. However, the first systematic RNA-Seq studies showed already that the technical replicability is so high, that technical replicates are not required. In fact, practically all RNA-Seq studies are done without technical replicates for this reason.

    (10) Because I have little confidence in the SBI metric (#7-8) and in interpreting within sex variances (#9), I found little value in the human results and how SBI distributions (and degree of overlap between sexes) compare between humans and mice.

    We disagree - the current published status is that there are thousands of sex-biased gene in humans and this has implications for gender-specific medicine (Oliva et al. 2020). Our results show a much more nuanced picture in this respect.

    (11) I found even less value in the single-cell data. It too suffers from the issues above. Further, as the authors more or less state, the data are too limited to say much of value here. It is impossible to tell to what extent the results are simply due to data limitations.

    We have pointed out that it is still valuable to have them. They are good enough to exclude the possibility that only a small set of cells drives the overall pattern across an organ. We have further clarified this in the text.

    (12) The code for data analysis should be posted on GitHub or some other repository.

    The code for the sex-biased gene detection and analysis has been posted on GitHub (see Code availability in the manuscript).

  12. Arcadia Science

    Nine age-matched adult females and adult males each were chosen from each of the four taxa, 72 individuals are included in total in the overall analysis. As somatic organs we included brain (whole brain), heart, liver (left medial lobe), kidney (right) and mammary gland (fourth, right). Note that the mammary glands in mice have similar sizes in both sexes before lactation and are therefore directly comparable.

    There's some evidence of sex-specific cell type heterogeneity in organs (e.g. https://pmc.ncbi.nlm.nih.gov/articles/PMC10210449/; https://pmc.ncbi.nlm.nih.gov/articles/PMC7615307/#S1). It seems possible that consistent sex-specific organ heterogeneity might be another explanation for the patterns you see and, if present, could change interpretations/conclusions. E.g., sex-biased differences could arise from cell number variation rather than intrinsic transcriptional differences. How much of a concern is that here?

  13. Version published to 10.7554/elife.99602.1 on eLife
  14. eLife

    eLife assessment

    This is a useful study on sex differences in gene expression across organs of four mice taxa, although there are some shortcomings in the data analyses and interpretations that should to be better placed in the broader context of the current literature. Hence, the evidence in the current form is incomplete, with several overstated key conclusions.

  15. eLife

    Reviewer #1 (Public Review):

    The authors describe a comprehensive analysis of sex-biased expression across multiple tissues and species of mouse. Their results are broadly consistent with previous work, and their methods are robust, as the large volume of work in this area has converged toward a standardized approach.

    I have a few quibbles with the findings, and the main novelty here is the rapid evolution of sex-biased expression over shorter evolutionary intervals than previously documented, although this is not statistically supported. The other main findings, detailed below, are somewhat overstated.

    (1) In the introduction, the authors conflate gametic sex, which is indeed largely binary (with small sperm, large eggs, no intermediate gametic form, and no overlap in size) with somatic sexual dimorphism, which can be bimodal (though sometimes is even more complicated), with a large variance in either sex and generally with a great deal of overlap between males and females. A good appraisal of this distinction is at https://doi.org/10.1093/icb/icad113. This distinction in gene expression has been recognized for at least 20 years, with observations that sex-biased expression in the soma is far less than in the gonad.

    For example, the authors frame their work with the following statement:
    "The different organs show a large individual variation in sex-biased gene expression, making it impossible to classify individuals in simple binary terms. Hence, the seemingly strong conservation of binary sex-states does not find an equivalent underpinning when one looks at the gene-expression makeup of the sexes"

    The authors use this conflation to set up a straw man argument, perhaps in part due to recent political discussions on this topic. They seem to be implying one of two things. a) That previous studies of sex-biased expression of the soma claim a binary classification. I know of no such claim, and many have clearly shown quite the opposite, particularly studies of intra-sexual variation, which are common - see https://doi.org/10.1093/molbev/msx293, https://doi.org/10.1371/journal.pgen.1003697, https://doi.org/10.1111/mec.14408, https://doi.org/10.1111/mec.13919, https://doi.org/10.1111/j.1558-5646.2010.01106.x for just a few examples. Or b) They are the first to observe this non-binary pattern for the soma, but again, many have observed this. For example, many have noted that reproductive or gonad transcriptome data cluster first by sex, but somatic tissue clusters first by species or tissue, then by sex (https://doi.org/10.1073/pnas.1501339112, https://doi.org/10.7554/eLife.67485)
    Figure 4 illustrates the conceptual difference between bimodal and binary sexual conceptions. This figure makes it clear that males and females have different means, but in all cases the distributions are bimodal.

    I would suggest that the authors heavily revise the paper with this more nuanced understanding of the literature and sex differences in their paper, and place their findings in the context of previous work.

    (2) The authors also claim that "sexual conflict is one of the major drivers of evolutionary divergence already at the early species divergence level." However, making the connection between sex-biased genes and sexual conflict remains fraught. Although it is tempting to use sex-biased gene expression (or any form of phenotypic dimorphism) as an indicator of sexual conflict, resolved or not, as many have pointed out, one needs measures of sex-specific selection, ideally fitness, to make this case (https://doi.org/10.1086/595841, 10.1101/cshperspect.a017632). In many cases, sexual dimorphism can arise in one sex only without conflict (e.g. 10.1098/rspb.2010.2220). As such, sex-biased genes alone are not sufficient to discriminate between ongoing and resolved conflict.

    (3) To make the case that sex-biased genes are under selection, the authors report alpha values in Figure 3B. Alpha value comparisons like this over large numbers of genes often have high variance. Are any of the values for male- female- and un-biased genes significantly different from one another? This is needed to make the claim of positive selection.

  16. eLife

    Reviewer #2 (Public Review):

    The manuscript by Xie and colleagues presents transcriptomic experiments that measure gene expression in eight different tissues taken from adult female and male mice from four species. These data are used to make inferences regarding the evolution of sex-biased gene expression across these taxa. The experimental methods and data analysis are appropriate; however, most of the conclusions drawn in the manuscript have either been previously reported in the literature or are not fully supported by the data.

    There are two ways the manuscript could be modified to better strengthen the conclusions.

    First, some of the observed differences in gene expression have very little to no effect on other phenotypes, and are not relevant to medicine or fitness. Selectively neutral gene expression differences have been inferred in previous studies, and consistent with that work, sex-biased and between-species expression differences in this study may also be enriched for selectively neutral expression differences. This idea is supported by the analysis of expression variance, which indicates that genes that show sex-biased expression also tend to show more inter-individual variation. This perspective is also supported by the MK analysis of molecular evolution, which suggests that positive selection is more prevalent among genes that are sex-biased in both mus and dom, and genes that switch sex-biased expression are under less selection at the level of both protein-coding sequence and gene expression.

    As an aside, I was confused by (line 176): "implying that the enhanced positive selection pressure is triggered by their status of being sex-biased in either taxon." - don't the MK values suggest an excess of positive selection on genes that are sex-biased in both taxa?

    Without an estimate of the proportion of differentially expressed genes that might be relevant for broader physiological or organismal phenotypes, it is difficult to assess the accuracy and relevance of the manuscript's conclusions. One (crude) approach would be to analyze subsets of genes stratified by the magnitude of expression differences; while there is a weak relationship between expression differences and fitness effects, on average large gene expression differences are more likely to affect additional phenotypes than small expression differences. Another perspective would be to compare the within-species variance to the between-species variance to identify genes with an excess of the latter relative to the former (similar logic to an MK test of amino acid substitutions).

    Second, the analysis could be more informative if it distinguished between genes that are expressed across multiple tissues in both sexes that may show greater expression in one sex than the other, versus genes with specialized function expressed solely in (usually) reproductive tissues of one sex (e.g. ovary-specific genes). One approach to quantify this distinction would be metrics like those used defined by [Yanai I, et al. 2005. Genome-wide midrange transcription profiles reveal expression-level relationships in human tissue specification. Bioinformatics 21:650-659.] These approaches can be used to separate out groups of genes by the extent to which they are expressed in both sexes versus genes that are primarily expressed in sex-specific tissue such as testes or ovaries. This more fine-grained analysis would also potentially inform the section describing the evolution/conservation of sex-biased expression: I expect there must be genes with conserved expression specifically in ovaries or testes (these are ancient animal structures!) but these may have been excluded by the requirement that genes be sex-biased and expressed in at least two organs.

    There are at least three examples of statements in the discussion that at the moment misinterpret the experimental results.

    The discussion frames the results in the context of sexual selection and sexually antagonistic selection, but these concepts are not synonymous. Sexual selection can shape phenotypes that are specific to one sex, causing no antagonism; and fitness differences between males and females resulting from sexually antagonistic variation in somatic phenotypes may not be acted on by sexual selection. Furthermore, the conditions promoting and consequence of both kinds of selection can be different, so they should be treated separately for the purposes of this discussion.

    The discussion claims that "Our data show that sex-biased gene expression evolves extremely fast" but a comparison or expectation for the rate of evolution is not provided. Many other studies have used comparative transcriptomics to estimate rates of gene expression evolution between species, including mice; are the results here substantially and significantly different from those previous studies? Furthermore, the experimental design does not distinguish between those gene expression phenotypes that are fixed between species as compared to those that are polymorphic within one or more species which prevents straightforward interpretation of differences in gene expression as interspecific differences.

    The conclusion that "Our results show that most of the genetic underpinnings of sex differences show no long-term evolutionary stability, which is in strong contrast to the perceived evolutionary stability of two sexes" - seems beyond the scope of this study. This manuscript does not address the genetic underpinnings of sex differences (this would involve eQTL or the like), rather it looks at sex differences in gene expression phenotypes. Simply addressing the question of phenotypic evolutionary stability would be more informative if genes expressed specifically in reproductive tissues were separated from somatic sex-biased genes to determine if they show similar patterns of expression evolution.

  17. eLife

    Reviewer #3 (Public Review):

    This manuscript reports some interesting and important patterns. The results on sex-bias in different tissues and across four taxa would benefit from alternative (or additional) presentation styles. In my view, the most important results are with respect to alpha (fraction of beneficial amino acid changes) in relation to sex-bias (though the authors have made this as a somewhat minor point in this version).

    The part that the authors emphasize I don't find very interesting (i.e., the sexes have overlapping expression profiles in many nongonadal tissues), nor do I believe they have the appropriate data necessary to convincingly demonstrate this (which would require multiple measures from the same individual).

    This study reports several interesting patterns with respect to sex differences in gene expression across organs of four mice taxa. An alternative presentation of the data would yield a clearer and more convincing case that the patterns the authors claim are legitimate.

    I recommend that the authors clarify what qualifies as "sex-bias".

  18. eLife

    Author response:

    We appreciate the time of the reviewers and their detailed comments, which will help to improve the manuscript.

    We are sorry that at least one reviewer seems to have had the impression that we have conflated issues about gonadal and non-gonadal sex phenotypes. This referee suggests that we should use Sharpe et al. (2023) to develop our concepts. However, what is discussed in Sharpe et al. was already the guiding principle for our study (without knowing this paper before). In our paper, we introduce the gonadal binary sex (which is self-evidently also the basis for creating the dataset in the first place, because we needed to separate males from females) and go then on to the question of (adult) sex phenotypes for the rest of the paper. The gonadal data are included only as comparison for contrasting the patterns in the non-gonadal tissues.

    Our study presents the largest systematic dataset so far on the evolution of sex-biased gene expression. It is also the first that explores the patterns of individual variation in sex-biased gene expression and the SBI is an entirely new procedure to directly visualize these variance patterns in an intuitive way (note that the relative position of the distributions along the X-axis is indeed not relevant). The results are actually quite nuanced (e.g. the rather dynamv changes seen in mouse kidney and liver comparisons) and go certainly beyond what would have been predictable based on the current literature.

    Also, we should like to point out that our study contradicts recent conclusions that were published in high profile journals, that had suggested that a substantial set of sex-biased genes has conserved functions between humans and mice and that mice can therefore be informative for gender-specific medicine studies. Our data suggest that that only a very small set of genes are conserved in their sex-biased expression. These are epigenetic regulator genes and it will therefore be interesting in the future to focus on their roles in generating the differences between sexual phenotypes in given species.

    We will be happy to use the referee comments to clarify all of these points in a revised version. But we do not think that our "evidence is incomplete" and that there are several "overstated key conclusions". We have used all canonical statistical analyses that are typically used in papers of sex-biased gene expression, as acknowledged by reviewers 1 and 2. The additional statistical analyses that are requested are not within the scope of such papers, but could be subject to separate general studies, independent of the sex-bias analysis (e.g. the role of highly expressed genes versus low expressed genes, or the analysis of the fraction of neutrally evolving loci).

    Finally, it is unclear why the overall rating of the paper is at the lowest possible category ("useful study"), given that it adds a substantial amount of data and new insights into the exploration of the non-binary nature of sexual phenotypes.

  19. Version published to 10.1101/2024.05.22.595301 on bioRxiv