Scientist Stories: David Page, Sex in Health and Disease

Welcome to today's webinar, Why Sex Matters in Health and Disease. I'm Jonathan Weissman, a member here at the Whitehead Institute, where my lab studies a range of problems from how tumors evolve to how proteins are made, folded, and disposed of in cells. I'm excited to be your host and moderator today. Today's webinar is part of a series of talks that we've organized to share Whitehead Institute science and research with the wider community. In a few moments, we'll hear from our presenter, fellow Whitehead Institute member David Page. First, I'd like to let you know that this webinar is being recorded for future viewing. Also, we'll be answering questions at the end of today's presentation, but you can submit a question at any time using the Q&A box at the bottom of your screen. With that, I'd like to introduce our speaker today, David Page. David earned his undergraduate degree in chemistry at Swarthmore College in 1978, and then went on to earn his MD from Harvard Medical School and the Harvard-MIT Health Sciences and Technology Program while simultaneously conducting human genetics research in the laboratory of David Botstein at MIT. In 1984, he became the first Whitehead Fellow at the then-new Whitehead Institute, He then joined the faculty of Whitehead in 1988 and founded in 1992 the Whitehead Task Force on Genetics and Public Policy. David served as the director of the Whitehead Institute from 2004 to 2020. Currently, David is a member of the— at the Whitehead Institute and also a professor of biology at MIT and an investigator at the Howard Hughes Medical Institute. Additionally, David serves on the selection committee of the March of Dimes Prize in Developmental Biology and the Taubman prize in translational medical science. He serves as the chair of the visiting committee of Harvard Medical School and Harvard School of Dental Medicine and on the board of directors of PepsiCo. At the Whitehead, David's research focuses on understanding where male and female cells, tissues, and organs are essentially the same and where they're fundamentally different. Today, he will discuss his lab's latest findings on the genetic differences between males and females and on the biological and medical ramifications of these differences. David, welcome and thank you.

Jonathan, thank you for that very generous introduction. I'm just going to take and I want to offer my greetings to all of you out there, friends and supporters of Whitehead Institute. I want to talk today about my lab's mission and encourage you to join me in thinking about how best to pursue that mission. And let me start by sharing my screen. So our task today is to consider why sex matters in health and disease. And I'm going to, as a subtitle, offer the chromosomal basis of sex differences in human health and disease. So with that, let me turn to my lab's mission. And that is to discover the molecular bridges that we infer must connect the sex chromosomes to sex differences in health and disease across the body. And I'm going to assert today that all biologically based differences, and by biologically based differences, I mean those that are not due to social or environmental determinants. I'm going to assert that all biologically based sex differences trace their origins to the sex chromosomes. Before I start, I'm going to say that there are, for the scientists out there among you, and I think there are many, There are two special cases where molecular bridges from the sex chromosomes to sex differential biology are already well described. The first special case is the reproductive tract, where the roles of the Y chromosome gene SRY and downstream actors, including the sex steroids, have been described literally over the past century. The second special case are the X-linked recessive disorders like colorblindness and hemophilia, all of which are more common in males because they lack a second copy of the gene in question. But I would assert that apart from these two special cases, the X-linked recessives and the reproductive tract, I would assert that apart from these two special cases, there is no molecular bridge known between the sex chromosomes on the left side of this diagram and sex differential biology in humans. Our mission today and going forward is to discover the bridges that we infer must exist. In the biomedical community, there is growing interest in describing and cataloging sex differences in health and disease, but there is no empirical or conceptual framework with which to connect these differences back to the X and Y chromosomes. So we aspire to build that empirical and conceptual framework. But what do I mean by sex differences? Well, let's look first to the reproductive tract. In the reproductive tract, sex differences are binary. Pretty straightforward. Egg or sperm, ovary or testis. But outside the reproductive tract and across the body, sex differences are not binary, but instead involve continuous variables with overlapping distributions. Overlapping but shifted distributions in females and males. Consider human height, where the distribution observed in females is shifted 13 centimeters or 5 inches from the distribution observed in males. In fact, the world's largest genome-wide association, or GWAS studies, encompassing 5 million research subjects, GWAS studies of height implicate 1,000 or more genes in variation in height. It turns out it's the same 1,000 genes that have been implicated in females and males, but with no explanation for the 13-centimeter shift between the sexes. Diseases display sex differences. For example, the autoimmune disorder lupus is 6 times as common in females while autism spectrum disorder is diagnosed 4 times as often in males as in females. In many disorders, even when females and males are affected at similar frequency, the severity or the consequences are often greater in one sex than another. Consider, for example, dilated cardiomyopathy, where heterozygous mutations in autosomally encoded components of the sarcomere, the contractile machinery of the heart. Heterozygous mutations in autosomally encoded sarcomere components cause a dangerous ballooning of the heart. In this plot, which is from Cricket and John Seidman's lab, we see the survival curve for females with this disorder. Males with equivalent autosomal genotypes die at a much younger age. There is today no explanation for what is generally a 10-year difference in the occurrence of very severe, if not lethal events. For individuals with dilated cardiomyopathy between males and females. So how then to discover the molecular bridges from such sex differences in health and disease back to the sex chromosomes? That is our quest. We believe that a critical intermediate step is actually sex-biased expression of autosomal genes, genes that otherwise function identically in females and males. Is there evidence of such an intermediate step? Yes, as we found in a multi-species survey of gene expression. Here I'm going to tell you about the work of my student Sahin Naqvi, currently a postdoc at Stanford, but while he was in my lab, Sahin and colleagues looked for sex differences across humans, macaques, mice, rats, and dogs. He examined the 12 tissues shown here that are present in both females and males. These are tissues that at first glance we would consider to be identical. In males and females. And in humans, Song-Hyun analyzed publicly available so-called GTEx consortium data, and in the other 4 species, he generated the RNA sequence data anew, setting the stage for— setting the stage for the first and to date only serious cross-species analysis of sex-biased gene expression. One question that arises is, how common are male or female biases in autosomal gene expression? Well, of 12,000 genes where orthologs, where essentially the same gene could be matched across the species, Sahin found that of those 12,000 genes, 85% of them show sex bias in one or more tissues in one or more of the 5 species. Now, 12 tissues times 12,000 genes equals 144,000 gene-tissue combinations that Sahin was examining across these 5 species. And what he found was that 40% of these gene-tissue combinations showed sex-biased expression in one or more of the species. Now, are these rampant sex biases in gene expression conserved across the 5 species? Well, the answer is mostly not. Only 7% of the sex-biased gene-tissue combinations show that bias in the same direction across at least 4 of the 5 species. So most of the sex-biased gene expression is not conserved. Nonetheless, in each tissue, Sahin found hundreds of genes that showed conserved sex-biased gene expression. The question then arises, does any of this matter? Does sex-biased autosomal gene expression explain sex differences in a trait? So to take on this question, Sahin examined an intensely studied trait with bountiful publicly available genetic data. That is the example of human height that I've already introduced. And it was a brilliant insight on Sahin's part towards the end of his thesis work to address this question of height. And what Sahin found, first of all, he was inspired by his findings with a transcription factor that is named L-CORL. It was previously identified previously genetically associated with height and body size, not only in humans but also in cattle and horses. And it turns out that previous studies had revealed the direction of L-CORL's effect. That is, in both humans and horses, more expression of L-CORL means decreased height. It's inversely related to height. What Sahin identified was a subtle but conserved female bias in L-CORL expression, a subtle but conserved female bias that had previously gone unnoticed. Now, in the pituitary, as you can see here, L-CORL tends to be expressed more highly in females in all 5 species examined. This strongly suggests that this conserved female bias existed in the common ancestor of all 5 species on the order of 100 million years ago, and it was preserved in all 5 lineages to modern times. Then Sahin asked whether this striking example of L-CORL, was this an isolated example or was it a harbinger of something more general? Now, what Sian found was that by comparing his own data on sex bias with other investigators' data on height, he found that other autosomal genes with conserved female bias tended to decrease height, while autosomal genes with conserved male bias tended to increase height. I'll let that sink in, let you think about the polarity of those effects. Now, summing these effects across hundreds of implicated genes, it turns out these two effects decrease mean height in females and increase it in males, accounting for 12% of the 13-centimeter difference that had been well documented between male and female distributions. Also illustrating not just this point with respect to height, but potentially establishing a new paradigm. That is the conserved sex biases in expression of autosomal genes, genes that otherwise operate equivalently in females and males, such conserved sex biases could contribute to sex differences in health across human populations. I want to just take a moment to reflect on how this differs from our understanding of sex differences in the reproductive tract. Having worked on the reproductive tract for decades, I know that research there has been guided by the principle of looking for male-specific or female-specific pathways and actors. What I'm emphasizing here, arising from Songyun's work, is the concept that pathways that operate identically in males and females can yield sex-biased phenotypic distributions purely and simply as a result of sex bias in the expression of those genes. Not on/off, not on/off switches between the sexes, but different dial settings between males and females with respect to those genes. This is a concept that the lab and I hope to refine and extend in the coming years. This was the first example to illustrate this principle. But in a very exciting development, a year after Sahin's study of height, Steve McCarroll and colleagues at Harvard Medical School published work on an entirely different subject, but with echoes of Sahin's findings. So it turns out that lupus, as I mentioned, is more common in females, as is another autoimmune disorder, Sjögren's syndrome. Both of these disorders show their strongest genetic autosomal association with something called C4, a complement protein. Steve showed that higher concentrations of C4 protein are protective. It turns out that males have modestly higher C4 concentrations, than genotypically matched females, protecting them against both disorders. And similar to Sahin's observations with height, what this illustrates conceptually is that sex-biased expression of autosomal genes can shift the distribution. In this case, the distribution of disease risk can shift that distribution between the sexes. Now, Steve and colleagues do not yet know why C4 protein levels are higher in males. It turns out that these sex differences may arise post-transcriptionally in the case of C4. So, as with height, the molecular bridges to the sex chromosomes remain to be built. So, let's talk about my favorite pair of chromosomes, the sex chromosomes, and how we have been coming to think about them over the course of the last year in particular. Humans nominally have sex chromosomes of two kinds. Up at the top, the green bit, the pseudoautosomal region, or PAR, is kept identical on X and Y by crossing over during male meiosis. The rest of the X chromosome is the non-pseudoautosomal region, or NPX, and the rest of the Y is the non-pseudoautosomal NPY. Now, while diverged in DNA sequence and gene content, NPX and NPY share 17 gene pairs. These NPX-NPY gene pairs are not pseudoautosomal. I'll just say again, those genes are not pseudoautosomal. They're not identical, they're not swapped between the X and the Y at male meiosis. But while geneticists say, we always say that humans have two kinds of sex chromosomes, epigenetically speaking, I will argue that human somatic cells have sex chromosomes of three kinds. XA, the active, so-called active X, XI, the so-called inactive X, Y. And the amazing thing is that in human beings, these can coexist. These chromosomes can coexist in the same somatic cell in various combinations. And I'll just point out that though we often look for clues and cues and parallels, there is no parallel to this situation. In flies, which have at most two kinds of sex chromosomes in any given cell, or worms, that is C. elegans, with one kind of sex chromosome in a given cell. So let's consider the implications of the fact that human somatic cells carry sex chromosomes of three epigenetic sorts. It turns out then that euploid human somatic cells come in two varieties, either typically 1XA and 1XI, or 1XA plus 1Y. Now, 1XA is present in euploid somatic cells of both sexes. After extensive genetic characterization, our lab can find no differences between the XAs in female and male cells. That is to say that the XAs appear to be epigenetically uniform. I'll say that while the X is often referred to both by scientists and non-scientists alike, the X is often referred to as a female chromosome, The XA is no more female than any autosome. And so the chromosomes that differ between the sexes, the chromosomes that differ between females and males, are XY versus Y. Now this might seem like a small thing, but we geneticists have for a century said that females are XX and males are XY. But what I'm asserting now is that it may be more helpful to say that females are XI and males are Y. In other words, the other 45 chromosomes, all 44 autosomes, 22 pairs, plus the first sex chromosome, the XA. The first 45 chromosomes are genetically and epigenetically equivalent in females and males, and only the 46th chromosome differs. This allows us to restate that mission that I laid out at the very beginning. It allows us to restate this mission in a more focused way. A more focused proposition is that our task is to connect the 46th chromosome, be it Xi or Y, to the sex-to-sex differences in health and disease, probably through this intermediate of sex-biased autosomal gene expression. Now, having spent my career studying the Y, we have now turned our attention to the human X, Y. I want to offer first, as I introduce you to my new love, the X, Y, I want to offer first an essential piece of background. That is something called the N-1 rule of the X chromosome, which has been in place since 1961. The N-1 rule says that No matter how many X chromosomes a diploid human somatic cell has, only one is an XA and all others are XIs. We've exploited the N-1 rule to characterize XI, and we've done this in a project led by Adriana San Roman, who is now— actually, next week will be entering the academic job market. Searching for her own independent faculty position. In a project led by Adriana and with the help of clinical scientists at NIH, at the Focus Foundation, and elsewhere, Adriana and colleagues have studied about 100 individuals with sex chromosome aneuploidy. That is individuals who instead of having two Xs or an X and a Y have another have some other sex chromosome constitution. Ed Roden and colleagues have generated lymphoblastoid cell lines, primary fibroblast cultures, and conducted RNA sequencing. Studying individuals with between 1 and 4 X chromosomes, that is 1 XA and 1, 2, or 3 XIs, And it turns out that this enables, for the mathematically inclined among you, this enables linear modeling of Xis effect on transcript levels for each chromosome X genes. This is an amazing resource that nature has delivered for study. And, and we're going to turn, if you look for a moment at this equation. We're going to turn to beta sub 0 and beta sub x in a minute. But first, let's inspect the raw data for 3 genes to get a feel for what this looks like. I'm going to show you Adriana and colleagues' data for 3 genes. Let's start with PRPS2. This is an X-linked gene that is expressed by XA. It's expressed only by active X chromosomes, not by XI. On this plot, each dot represents the level of PRPS2 transcripts in one individual. Each of these dots is one person. The individuals have 1-4 X chromosomes as outlined here on the x-axis. No pun intended, but it works quite well, the x-axis here. We see some variation among individuals, but as expected, there is no change in general in expression as Xis are added to the single Xa present in each individual. Perhaps more interesting is the gene exist. It is the long non-coding RNA that is expressed by Xi and not by Xa, and it plays a central role in the process of X inactivation. It is expressed in individuals with one— sorry, it is not expressed in individuals with one X chromosome. You can see that the dots here are beautifully at the baseline, but its expression rises as you add Xis. But perhaps even more interesting to us, are genes, X-linked genes like KDM5C, that are expressed by all X chromosomes and whose expression marches upward as one adds XIs. Very tidy, very modular, almost algebraic, with each dot again representing a different unrelated human individual. Many with a Y chromosome, many without, sorted only by the number of X chromosomes. And I would offer here that this is so modular and so algebraic that if, if all you knew was the level of expression of KDM5C in lymphoblastoid cells from an individual, you could estimate with reasonable confidence the number of X chromosomes that individual carries. And the absolute beauty, algebraic beauty of this data inspired Adriana to measure the slope and intercept of these points and devise a metric to express the effect of adding an Xi. So for each gene, one can, by linear modeling, estimate the slope Beta sub x, which is the change per Xi, and also the intercept beta 0, which is expression from XA, and normalize across chromosome X genes by dividing the slope by the intercept to arrive at a new metric that we call Delta E sub x. Now genes like PRPS2 in the upper left-hand corner, Genes like PRPS2 have a slope of 0 and hence a delta E sub X of 0, while genes like CADIUM5C have a positive slope and hence a positive delta E sub X value. Looking systematically across the whole chromosome, Adriana found that each chromosome X gene has a characteristic delta E sub X value. Now on the horizontal axis here are the delta E sub X are plotted the values of Delta E sub X for 357 chromosome X genes that are expressed in LCLs. On the vertical axis is the statistical strength of departure from the null hypothesis of a Delta E sub X of 0. As expected, most genes are like PRPS2 with Delta E sub X values of approximately 0. But others range from modest negative values to positive values as high as 1. Note here again KDM5C and EXIST, which I've shown you the raw data for. All told, 38% of expressed chromosome X genes have a significantly non-zero delta E sub X value. Which is about twice the 15-20% of genes that have been reported in the literature to, quote, escape X inactivation. A sizable number of genes actually have negative Delta E sub X values. Their transcript levels fall in the presence of one or more Xis. It's a pretty amazing finding. That as you add for these genes, as you add X chromosomes, the expression of that gene declines. Now, this was such an unexpected finding that we had to ask ourselves, is this an artifact? Well, we absolutely think not. On a gene-by-gene basis, the Delta E sub X values are generally conserved between transformed lymphoblastoid cells the Delta E sub X values plotted on the horizontal axis here versus primary fibroblasts in cultures with the Delta E sub X values plotted on the vertical axis. The question arises, is expression from Xi The only way that non-zero delta E sub x values can arise. Well, and how to explain the negative delta E sub x values? Well, what if XA and XI are not independent and additive contributors to expression as has been universally, if implicitly, assumed by the X chromosome field for decades. We needed to compare Delta E sub X with a more direct measure of Xi expression. Adriana conducted allele-specific analyses calculating the allelic ratio, or AR, that is the ratio of transcripts from Xi and Xa, and compared AR with delta E sub X. Now it turns out if XA and XI expression are independent and additive, as has been universally assumed, then the way these values are calculated, AR should equal delta E sub X, which is represented— that expectation is represented by the diagonal here. Indeed, most chromosome X genes behave as expected. They fall near the diagonal with similar AR and delta E sub X values. But other genes don't behave as expected. They don't follow these rules. Consider MPP1. This is an X-linked gene whose values fall far below the diagonal. MPP1's delta E sub X value is 0.24. That is to say that each Xi increases transcript levels by 24%, but its allelic ratio is 0, demonstrating that Xi does not express MPP1. Only XA expresses MPP1. We conclude that Xi upregulates MPP1 in trans on Xa, which is the only chromosome that expresses this gene. Now let's contrast this positively modulated rogue MPP1. Let's contrast it with the negatively modulated rogue DDX3X, whose values rise well above the diagonal with a delta E sub x value of 0.26, which means that each Xi increases transcript levels about 26%. Virtually identical. Note the delta E sub x value for DDX3X is virtually identical to that of NPP1. But DDX3X's allelic ratio is much higher at 0.55. Meaning that in cells with 2 X chromosomes, transcripts from Xi are 55% as abundant as those from Xa. We conclude that Xi expresses DDX3X robustly, but also negatively modulates DDX3X transcripts that derive from Xa. Now I will say here that DDX3X is the most elaborately regulated gene that we've identified to date on the human X chromosome, and we are studying it intensely now and going forward. Summing it all up, it looks as though XI, what had been called the inactive X, is actually modulating transcript levels of at least 121 genes on XA. A substantial fraction of the genes on XA are modulated by Xis activity. Put together, these results reveal the existence of two uniquely X chromosome modes of gene regulation in humans. One is transcriptional attenuation of nearly all genes on Xi in cis. This is what is traditionally referred to as exon activation. In addition, there is modulation positive or negative of individual XA genes by Xi, and this occurs in trans. The new metric, Delta E sub X, to which I've introduced you, this new metric captures both cis and trans effects, These differ widely in magnitude and even in polarity among chromosome X genes. Well, let's bring this all back to the question of the differences between males and females in many healthy and diseased states. What Adriana then did was to use this knowledge to set out to identify the genes on the X chromosome that are most likely, we think, to drive sex differences in health and disease as judged by two criteria. Adriana considered that first these genes— she required first that these genes should— their expression should change upon adding an X-Y. And they should be exquisitely sensitive to both underexpression and overexpression, what we call Goldilocks dosage sensitivity. And these are characteristics that the sensitivity to under and overexpression are characteristics that Adriana assessed through reanalysis of human population and evolutionary genetic databases. Looking systematically, across the X chromosome, she identified a treasure map of just 10 NPX genes that we consider most likely to drive the somatic differences between XX females and XY males. Each of the genes, and they're listed here, displays a delta E sub X value of at least 0.1 and high dosage sensitivity as judged by population and evolutionary genetic metrics. Of these 10 treasure map genes, 5 have homologs on the Y chromosome, NPY homologs. Speaking to their unusual genetics, really unusual genetics, These mutations in 7 of these 10 genes cause childhood syndromes, 5 of which are listed in the human geneticists' favorite reference, Online Mendelian Inheritance in Man, or OMIM. 5 of these are listed as X-linked dominant. That is, the syndromes appear in heterozygous females. So I'm saying there is a whole new genetics of the X chromosome that goes beyond the textbook X-linked recessive model to encompass these treasure map genes, which are highly enriched in X-linked dominant phenotypes. And these genes on the X and their NPY homologues will be at the forefront of our efforts to discern the molecular bridges that link Xi versus Y to sex differences in health and disease. And I should also point out, I note the presence here of DDX3X, one of the modulated genes that I told you about. Let me close by mentioning some related projects, most of them, almost all of them collaborative, that are in their early stages, projects that build upon and extend the work that I've just described for you in detail. First of all, with Boston IVF, we are examining sex differences in gene expression in human preimplantation embryos before sex steroids make their first appearance. Then there are two projects that examine the postnatal effects of sex steroids across the body. First, with Boston Children's Hospital and their transgender clinic, we are examining the effects of transitional hormone therapy, gender-affirming hormone therapy, the effects— its effects on gene expression and peripheral blood cells. And also across the body, we're examining the impact of changes in sex hormones that accompany menopause. And with leading experts in specific tissues and cell types, we're taking deep dives into the origins of sex differences in cardiac metabolism, in thermoregulation, immunometabolism, and adipose tissue. And in glioblastoma, autism, and microglia. In all of these and other areas, we're applying the emerging understandings of the sex chromosomes that I've detailed for you. Now, before I turn to questions, again, I want to thank and highlight the people who have conducted all the work that I've described. And I also want to recognize those whose support has made this work possible. So first, let me say that the broad initiative to understand sex differences in health and disease that we've been undertaking here at Whitehead Institute has been conducted by an absolutely spectacular crew. I highlight here present members of the lab and some alumni of ours and other labs that Whitehead that have contributed. I put in bold the names of the individuals who have contributed directly to the projects that I've described in detail today. Others are working on projects I've not had time to detail today. I will especially again highlight the team-leading efforts of Sahin Naqvi and Adriana San Roman, who helped direct and oversee the, the two projects that I described. And I would point out that actually both Sahin and Adriana are currently on the job market, and I think they're spectacular candidates who will become outstanding colleagues of faculties at universities and medical schools lucky enough to recruit them. Let me say that this work is not limited to Whitehead Institute. As I've emphasized, we are collaborating with a growing network of scientists and physicians at other institutions, and I list some of those institutions and the individuals and actually some of the family and support groups that are engaged as well. And finally, none of this work would be possible, of course, without financial resources. And I'd like to close by thanking all of these foundations and institutions that have provided support that's, that for specific individuals, or specific projects within the lab. But I also want to especially highlight those who have provided unrestricted support. It's— the work that we're undertaking is rather non-conventional. We're taking a very non-conventional approach and what we think is a rather bold approach to this question of sex differences. In health and disease. And for that, the unrestricted support allows us to be maximally bold. And here I want to highlight the individual— the, the contributions of unrestricted support that have come from Howard Hughes Whitehead Institute, from the Darbola Center on Women's Health, and from all the forward-looking individuals whose names I list here, all providers of unrestricted support that have made the most innovative work possible. And with that, I will stop sharing my screen, and I would look forward to taking questions if, if time permits. Thank you all very much.

Thank you, David, for a wonderful seminar. We have a few questions. Maybe I'll take the prerogative and start with one, which is, do you have any sense of the machinery that controls the expression from Xi, either in magnitude or in the ratios? And do you see genetic differences in those that then manifest as if you have an extra X chromosome, for example?

Yeah, okay. So let me, let's see, paraphrase the question and say, you know, what's the molecular mechanism underlying Xi versus Xa? And First, I guess what I'm— we don't— so I should say, there are enormous— there's an enormous legacy of study of the molecular process of X inactivation, a process that is laid down during early mammalian embryogenesis. And that we think is then maintained and propagated in a stable way as a few hundred cells become the entire— give rise to the trillions of the many trillions of cells that make up the entire body. So what we're doing here is studying not that mechanism of X inactivation, but the product, the Xi and Xa products of that process. Now, what we have, I think we're coming to have a feel for the Xi. Sometimes people have talked about how biologists get to develop a feel for their organism. I think we're beginning to develop a feel for the XI. And I think what I'm struck by is that the XI needs to be taken on its own terms. I think historically the XI has been viewed as the weak sibling of the XA. Again, carrying the term inactive around its collar does not lead to a strong sense of self-esteem. But I'm familiar with having rescued the Y chromosome from the trash heap, intellectually speaking, and I feel that we're in the beginnings of doing the same for Xi. What I really want to emphasize is that I think the human X chromosome can come in two very distinct epigenetic states. They are XA and XI. Each of them demands explanation. I think in not too many years' time, we're going to have a much richer understanding of how the XI versus the Y is driving those male-female differences. I think it'll be rooted in the understanding of the molecular processes of X inactivation that are occurring during embryogenesis and how they're maintained as the body develops. But, you know, actually maybe embedded in the question was also something about variation. And I think the thing we've been struck by most so far is how actually invariant the Xi appears to be, how modular, how profoundly modular it appears to be among unrelated individuals. We had no right to observe such cleanly algebraic formations of data points. And so I'm really struck by how constant and modular and invariant to a first approximation the Xi is among individuals.

Terrific. We'll move on to some of the audience's questions now. Does your lab conduct any research about intersex individuals, cases where the reproductive tract is not binary?

It's a great question. We have over the years, our lab has done a lot of studies on the role of the sex chromosomes in the development of male and female reproductive anatomy. I think that the thing that we're working on at present that perhaps comes closest, at least to the spirit of the question, is our really intense collaboration with physicians and scientists in the transgender clinic and the gender diverse clinic at Boston Children's Hospital. So we are deeply engaged in these studies, both from the point of view of trying to understand the biology and trying to understand how trans therapy, which involves changing the sex hormonal status of an individual, how those therapies, those gender-affirming therapies, actually alter gene expression. We're deeply involved in studies of this sort. Thank you.

Following on that a little bit, can we see brain or structural hormonal similarities between a transgender brain versus the sex of the brain that they identify with?

That's an area I'm not necessarily an expert on those questions, but I think that with regard to trans, I think there have been at most some debated and somewhat controversial findings with respect to brain differences. But I'd be happy to follow up with anybody. That's not a subject on which we have actually conducted research ourselves. And it's a rather— I think, as our audience will be well aware, the whole subject, the whole subject of trans therapy and of course research on trans individuals has become quite a politically fraught one of late. And so, but we're proud to be involved in these studies. Great.

This is maybe more of a statement, but I think it can be turned into a question. I'm interested in understanding sex-based metabolic differences in human iPSC-derived cardiomyocytes.

We're intensely interested in the possibility of metabolic differences between males and females, including in the heart. Our studies of the heart, which are very much ongoing, are a bit preliminary at this time. Our studies of the heart have actually focused primarily on samples taken directly from the body. That is in either humans or mice. And we have been— we have actually not done much with iPS-derived cardiomyocytes as yet. And part of the reasoning is that across the body, the sex differences that we observe at a transcriptomic level and at a metabolic level, these sex differences are quite subtle. And as those who do a lot of work with IPS differentiation are well aware, epigenetic subtleties in differentiation schemes can, can be significant. And so we have, we have to date been focusing on, on cells and tissues taken directly from the body. I think we, once we become convinced that we're on to something there, we would be in a good position to try to reproduce those things in an iPS setting, but we're not starting within the iPS setting.

Terrific, thanks. Another question: what do these 10 special X genes do to the Y chromosome gene expression?

Ah, that's a great question. Well, so it turns out these, as I mentioned, the 10 treasure map genes include 5 of the 10 actually have counterparts on the Y chromosome. So we could almost say, what are the effects of the 10 treasure map genes on the X together with their counterparts in 5 genes on the Y chromosome? It's a really good question. We have not seen much impact of the— I think if the question was related to the effect of the X on the Y, we've not seen— I'd have to think carefully about whether we actually have leverage on this question. I think the question which we have most leverage right now is what is the impact of Xi on XA? We have a number of particular biological circumstances make that a favorable and leveraged question. We're also beginning to look at the question of how do XA, XI, and Y impact autosomal gene expression? That is— and could that be driving some of the sex-biased gene expression that we think will be responsible for sex-biased for sex differences in outcomes. So I think there are, I mean, essentially we can frame a whole series of questions about who's acting upon whom. And so it's a great question. I think I need to think more deeply about whether we have leverage on that specific angle. Yeah.

Terrific. Thanks, David. Let's, we have time for maybe one more question. And this is one that's specifically focusing on thyroid conditions as a very prominent example of gender differences that may lead to your findings.

Yeah, so, um, wow, you know, it's a great question. Many, many thyroid conditions, um, uh, are female— are strongly female biased, um, including thyroid cancer and some non-cancerous thyroid conditions. It's— the thyroid is a tissue that I would love to add to our shortlist of targets. We are not currently focused on discovering sex differences in the thyroid, but I would love to go there. So, if there's someone who'd like to follow up on that front, I'd be happy to discuss it with you. We've been trying hard. There's sex biases all across the body and in all disease categories, and we've been carefully considering where to place our bets. But the thyroid is in the on-deck circle for conditions. Thank you for bringing that up. Great question. Terrific.

Well, I think we've run out of time, but thank you, David, for a most enjoyable and enlightening hour.

Thank you, Jonathan, and thanks all who attended and participated and asked questions. Thank you all very much.

Yeah, thanks. Bye, guys.