Sex-chromosome mechanisms in cardiac development and disease

Many human diseases, including cardiovascular disease, show differences between men and women in pathology and treatment outcomes. In the case of cardiac disease, sex differences are exemplified by differences in the frequency of specific types of congenital and adult-onset heart disease. Clinical studies have suggested that gonadal hormones are a factor in sex bias. However, recent research has shown that gene and protein networks under non-hormonal control also account for cardiac sex differences. In this Review, we describe the sex-chromosome pathways that lead to sex differences in the development and function of the heart and highlight how these findings affect future care and treatment of cardiac disease. Conlon and Arnold discuss the sex differences in cardiac physiology and pathology, the sex-chromosome pathways underlying such differences, and future studies that are needed to assess how cardiac sex differences are maintained and propagated.

Males and females differ in the anatomy and physiology of the heart 1-7 (Table 1). These dissimilarities appear at the earliest stages of development. Males and females are born with an equal number of cardiomyocytes 8 . However, female cardiomyocytes have a significantly lower rate of programmed cell death during early cardiac development than do male cardiomyocytes 9 . This difference in programmed cell death leads to female hearts having a higher number of cardiomyocytes 9 but a smaller heart and significantly reduced epicardial artery size compared with males, even after adjusting for body size and mass.
Sex differences occur not just in cell number and epicardial tissue, but also in cardiomyocyte physiology (Table 1). Single isolated male cardiomyocytes contract more intensely and at a more rapid rate than female cardiomyocytes [10][11][12] . Furthermore, males and females display differences in the relative abundance of ion-channel subunits 12,13 . In accordance with these observations, isolated cardiomyocytes from females exhibit substantial sex differences in calcium excitation-contraction and more significant fractional shortening than do those from males [10][11][12] . Collectively, these anatomical and physiological differences lead to females, relative to males, having an elevated resting heart rate, shorter depolarization, more minor contraction of the ventricles, that is, the QRS duration on an electrocardiogram (EKG), and a shorter interval between ventricular depolarization and repolarization, that is, the ST amplitude on an EKG 10, [14][15][16] .
Sex differences during development continue to birth and throughout aging [17][18][19][20] . For example, as the heart ages in both sexes, there is an increase in collagen level and collagen crosslinking 21 . However, males have differences in the content and crosslinking of the extracellular matrix at a significantly earlier age than in females 14,22 . Ultimately, these variations make male hearts more fibrotic than female hearts 20,23,24 . Thus, there are fundamental sex differences in the anatomy, physiology, and function of the heart from formation and throughout life, with profound implications for normal heart health and cardiac disease.

Sex differences in heart disease
Men and women differ in prevalence, treatment outcome, and survival of cardiovascular disease (Table 2). These findings align with differences in frequency between the sexes for specific types of congenital heart disease 7,17,[25][26][27][28][29] ; for example, women have a higher prevalence of atrial septal defects, whereas men exhibit a higher prevalence of ventricular septal defects [30][31][32] .
Discordance is accentuated in patient outcomes [33][34][35][36][37][38][39][40][41][42][43][44] . For example, women have a 60% higher one-year mortality rate for myocardial infarction (MI) than men, even after accounting for the severity of infarction 45 . In a more extreme case, even with adjustments for risk factors, women who undergo aortic-valve replacement have a 30-day mortality rate that is four times higher than that of men who receive the same procedure 46 .
Clinical studies have shown that menopause is an influencing factor in differing outcomes. Women have a higher mortality rate yet a lower incidence of cardiac disease until they reach menopause; by contrast, the rate in postmenopausal women is almost equivalent to that in men 47 .
To directly measure protein changes between male and female adult heart tissue, Shi et al. conducted quantitative mass spectrometry (MS) with tandem mass tags (TMT) 13,77 of the same sample set used in the adult mouse transcriptomic analyses. Ninety-five cardiac proteins had differential abundances in males versus females, and 76 showed enrichment in males and 19 in females. Surprisingly, there was almost a complete discordance between RNA sex differences and protein sex differences; of all the proteins that showed a male-female difference, 97.5% did not exhibit any significant difference at the mRNA level. Moreover, cardiac proteins that had sex differences were associated with biological processes that were disparate from the processes observed for differentially expressed mRNAs. Males showed enrichment in cell cycle and heart processes, and females showed enrichment in membrane and epithelial-related pathways 13 . In sum, these findings suggest that sex differences in cardiac protein expression are established at least partly by post-transcriptional mechanisms.

Separating effects of sex chromosomes from effects of gonadal hormones
Because the sex chromosomes are the only inherited factors known to differ between male and female zygotes, all sex differences in subsequent development must be attributed to the differential action of these chromosomes, including effects on downstream pathways composed predominantly of autosomal genes 78 . The testis-determining Y gene Sry, which also inhibits differentiation of the ovaries 79 , is the first gene in the cascade that causes sex differences in the type of gonad. Because the ovaries and testes secrete different levels of hormones, which have potent effects throughout the body, gonadal hormones comprise a major class of factors that directly cause sex differences in diverse tissues. These hormones differ from the other major class of agents that cause sexual differentiation, which are gene products, and possibly nongenic effects, of the X and Y chromosomes. A major unresolved challenge is identifying the X and Y genes that cause sex differences in tissues without mediation by gonadal hormones 80 . During embryonic life, before the gonads differentiate, all sex differences in XX versus XY tissues can be attributed to sex-chromosome effects. However, after gonadal differentiation, the sex-chromosome and gonadal-hormone effects are usually confounded and challenging to separate when studying wild-type males and females.
Researchers have long studied the effects of gonadal hormones using various methods to manipulate the level of the hormones, or their synthesis or receptors, to determine how and where the hormones act. These approaches have become more sophisticated with the advent of cell-specific methods to alter gonadal-hormone receptors and synthesis in molecularly defined cell types 81 , to recognize cell-specific hormonal effects in complex tissue systems that control phenotypes of interest. Classically, the effects of gonadal hormones have been divided into two groups: 'activational' effects that are transient and disappear when the hormonal effect is removed (for example, by gonadectomy of either sex), and 'organizational' effects that are permanent, remaining even when the hormone is absent or declines 82 . Examples of activational effects include effects of estrogens that are reduced after ovariectomy or menopause or when estrogen levels decrease during the estrous or menstrual cycle. Examples of organizational effects include the permanent prenatal masculinization by testosterone of the external genitals (penis, scrotum) and specific brain regions of males.
The most useful method to detect sex chromosome effects involves the study of 'Four Core Genotypes' (FCG) mice ( Jackson Laboratory strain 10905) 58,83,84 (Fig. 1). In FCG mice, the type of sex chromosome is unrelated to gonadal type, so sex chromosome effects can be detected as differences in XX versus XY mice with the same type of gonad. The Y chromosome is deleted for Sry, producing the hormone-replacement therapy has little effect on the incidence of heart disease [53][54][55][56] . Furthermore, a sex-specific program controlled genetically by the X or Y chromosome can function outside the sex organs 13,57-60 . Thus, although hormones clearly have a critical influence in cardiac disease, the mechanisms that lead to sex differences in cardiac homeostasis and disease are largely unknown.

Regulation of cardiac sex differences
Despite the significant sex differences in heart development and disease, knowledge of the mechanisms involved in establishing and maintaining male-female cardiac differences remains deficient. The inequalities remain despite efforts and proposals to change the sex bias in research by the US National Institute of Health and the Office of Research on Women's Health [61][62][63][64] . Our knowledge is compromised by the absence of the X or Y chromosome in most genomewide association studies, as well as by sex biases in clinical trials and treatment 30,41,44,62,[65][66][67][68][69][70] . Moreover, few systematic studies have been conducted to assess transcriptional or protein differences between healthy mammalian male and female cardiac tissue.
Microarray and meta-analyses of human organ tissue have identified differentially expressed cardiac genes that vary between the sexes [71][72][73][74][75] . In one of the more extensive studies, InanlooRahatloo et al. performed transcriptomic analyses of human left ventricle tissue derived from 46 healthy individuals; they found that 126 genes were upregulated in females, and 54 genes were upregulated in males 76 . Pathway analyses showed that females had enriched expression of genes involved in the inflammatory pathway and cell movement, whereas males showed higher expression levels of genes involved in cell morphology. By contrast, in adult mouse (C57BL/6J) hearts 13 , of the 223 genes that showed differential sex expression, 47% were enriched in males and were involved in processes that are primarily associated Larger heart and epicardial artery Smaller heart and epicardial artery 9 Single isolated CM contract more strongly and more rapidly Single isolated CM contract less strongly and less rapidly 10-12 Greater relative abundance of ion-channel subunits in the heart Decreased expression of ionchannel subunits in the heart 12,13 Lower resting heart rate Elevated resting heart rate 10,14-16

Review article
https://doi.org/10.1038/s44161-023-00256-4 Ychromosome ( Fig. 1). A Sry transgene is then inserted into chromosome 3, so that XY -(SryTG+) mice have functional testes. When these FCG males are mated to XX females, the four core genotypes are produced: XX and XYmice with ovaries that lack Sry (XXF and XYF), and XX(SryTG+) and XY -(SryTG+) mice with Sry and testes (XXM and XYM) ( Fig. 1). This model can be used to answer three main questions: (1) does the phenotype of interest differ in XX and XY mice with the same type of gonads (a sex chromosome effect)? This question is answered twice: in XXF versus XYF gonadal females that lack Sry, and in XXM versus XYM gonadal males with Sry.
(2) Does the phenotype of interest differ in mice with different gonads (a gonadal-hormone effect) but with the same sex chromosomes? Again, this question is answered twice, by comparing XXF versus XXM and XYF versus XYM. (3) Is there an interaction between the effects of the two factors? For example, does the level of gonadal hormone influence the magnitude of a sex chromosome effect? To date, the FCG model has been used to discover sex-chromosome effects in many different tissues and disease conditions, including mouse models of autoimmune disease, metabolism, Alzheimer's disease, neural-tube closure defects, bladder cancer, and diverse behavioral and neural sex differences 80,[84][85][86][87] . Using the FCG model, Shi et al. identified 519 proteins that segregated with ovaries and testes. These proteins were hormonally controlled and segregated with the presence or absence of Sry (that is, they were Sry-dependent), whereas 159 proteins co-segregated with sex chromosomes (XX versus XY) 13 . This finding confirmed the existence of the sex-chromosome complement pathway in male and female cardiac tissues.
Experiments in FCG models have also uncovered evidence of sex differences in cardiovascular diseases attributed in part to sex-chromosome effects ( Fig. 1) 88 . In a mouse model of ischemia-reperfusion injury, mice with two X chromosomes have worse disease than do mice with one X chromosome, irrespective of gonadal sex 89 . Pulmonary hypertension influences females more than males, because of the protective effect of the Y chromosome gene Uty 90 . In several models of atherosclerosis (mice lacking the low-density lipoprotein receptor gene or the ApoE gene, or in mice given an atherogenic diet), XX mice have much greater atherosclerosis than XY mice 91 . This sex-chromosome effect is associated with hyperlipidemia in the XX mice. The distribution and attributes of aortic aneurysms differs in XX and XY FCG mice, and the difference is attributed to the number of X chromosomes based on studies of XY* mice 92 . Higher levels of androgens in males synergize with XY sex-chromosome complement to promote greater incidence of aneurysms. Drug-induced bradycardia is different in XX and XY FCG mice 93 . Angiotensin-II induced systemic hypertension was greater in XX than XY FCG mice 94 . Sex differences in stroke in aged mice are caused by sex-chromosome complement, specifically the number of X chromosomes, and the X-chromosome gene Kdm6a has been suggested to contribute to the sex-chromosome effect 95,96 . Finally, the X gene Kdm5c causes greater adiposity and hyperlipidemia in XX than in XY mice 7 , two physiological effects that could well influence cardiovascular function. In aggregate, these studies suggest that sex-chromosome effects have a direct role in cardiovascular disease.

Turner and Klinefelter syndromes
Potential functions for non-hormonal pathways in controlling cardiac protein expression initially came from observing associations between sex-chromosome aneuploidies and heart disease 97,98 . These data included the two most frequent sex-chromosome aneuploidies, Turner syndrome and Klinefelter syndrome.
Turner syndrome is defined as the loss of part or all of one X chromosome (45, X) in females [97][98][99][100][101][102] . Cardiovascular disease accounts for 99% of the loss of Turner syndrome fetuses and more than 50% of premature Turner syndrome postnatal deaths 100,103 . Turner syndrome-associated deaths are due to a range of cardiac diseases, including congenital heart disease, aortic dilation and ischemic heart disease 100,104 . These data would imply that a gene or set of genes on the X chromosome is required in a double dose for normal heart development and function. However, because women with Turner syndrome present with complete or partial loss of a copy of the X chromosome, and there is little correlation between the karyotype of people with the disease and their cardiac phenotype, it is unclear which gene(s) lost from the X chromosome in Turner syndrome are causative for heart disease [98][99][100]103 . Interestingly, protection from Turner syndrome is also afforded by the presence of a Y chromosome in XY individuals, which implies that the Y chromosome and second X chromosome have similar protective factors.
Klinefelter syndrome is the most common sex-chromosome abnormality in males and is caused by the presence of a second X chromosome (47, XXY) 105,106 . Those with Klinefelter syndrome, like those with Turner syndrome, have a significant increase in mortality caused by cardiac abnormalities. However, those with Klinefelter syndrome experience impaired cardiopulmonary performance, impaired left ventricular systolic and diastolic function and an increased risk of congenital heart disease [107][108][109][110][111] . Although Klinefelter syndrome phenotypes may be attributed to a reduction in testosterone, hormone therapy has no effect on Klinefelter syndrome-associated cardiovascular disease 108 . These findings suggest that the second copy of some X-linked gene(s) disrupts essential cardiac networks in XXY individuals. However, the genes on the X chromosome that are causative to the Klinefelter phenotype are yet to be described. Several methods have been used to compare mice with one or two X chromosomes while holding other important characteristics constant (Fig. 2). Comparing XX and XO mice is a model of human Turner Syndrome (45, X), in which the two groups are both gonadal females that differ in the number of X chromosomes (two versus one). Comparing XXY and XY mice is a model of Klinefelter syndrome (47, XXY), in which gonadal males have two X chromosomes instead of one. We have used the XY* model (Fig. 2), available on a C57BL/6J background (MMRRC:043694-UCD), which, by using a simple cross, produces both Turner and Klinefelter model mice in the same litter 83 . The XY* males possess a Y* chromosome with an aberrant pseudoautosomal region (PAR), which recombines abnormally with the X chromosome, producing four types of offspring. Two types of progeny are XY* X and XX gonadal females, which are very similar to XO and XX mice, respectively 83,85 . These groups can be compared to test the effects of one versus two X chromosomes in gonadal females, that is, the Turner syndrome test. The other two groups, XY* and XX Y* gonadal males, are very similar to XY and XXY, respectively; thus, these groups can be used to assess the effects of one versus two X chromosomes in mice with a Y chromosome and testes, that is, the Klinefelter test 83 . The XY* model is also a test of the effects of the Y chromosome, comparing XO versus XY mice, or XX versus XXY mice. The simplicity of the XY* model accounts for its increasing popularity in measuring X-and Y-chromosome effects in mice, especially in conjunction with the FCG model 89,90 . Tandem use of FCG and XY* models is powerful because the methods test for sex-chromosome effects separately, enabling independent replication of results found in the other model and establishing whether sexchromosome effects in FCG mice are caused by X or Y genes 83 .

X-linked proteins' control of male-female differences
To determine whether X-linked genes directly regulate male-female differences in cardiac protein expression, we compared cardiac protein abundances in heart tissue derived from XX females versus Turner syndrome females (XO). Because it was possible that reducing the number of X chromosomes could lead to an alteration in gonadal hormones, candidate proteins were cross-referenced between the Turner syndrome data and data from the sex-chromosome pathway. Analyses revealed that 30% of the cardiac proteins in the sex-chromosome pathway showed altered expression in response to the loss of X-linked genes. Thus, a gene or set of genes on the X chromosome regulates cardiac protein expression by a dosage-sensitive mechanism 13 .
In the same study, we quantified protein expression in heart tissue derived from XY males versus Klinefelter syndrome males (XXY). Cross-referencing this dataset with the FCG data analyses revealed that 18% of the cardiac proteins that showed an alteration in expression between XY males and XXY males were in the sex-chromosome complement pathway. Importantly, we cross-referenced the proteins that showed differences in XO versus XX animals with proteins that showed differences in XY versus XXY mice; this work revealed that 26% of proteins that showed effects of X chromosome number occurred in both gonadal females and gonadal males 13 . Taken together, these findings suggest that genes on the X-chromosome function outside the gonad in a dose-dependent manner to regulate protein expression in the heart.

Cardiac sex differences and congenital heart disease
Congenital heart disease (CHD) is the most common congenital malformation 112,113 , and a significant portion of CHD states are sexbiased and developmental in origin 2,114 . Because the heart in humans and mice undergoes early organogenesis prior to gonad formation and sexual differentiation [115][116][117] , it has been suggested that differences between male and female hearts are in part unrelated to gonadal hormones (Fig. 3). Indeed, Deegan et al. 118 defined transcriptional differences between male and female mouse embryonic stem cell (ESC)-induced cardiomyocytes (iCM). The sex differences were temporally regulated and occurred at a period corresponding to mouse embryonic day 8.5

Review article
https://doi.org/10.1038/s44161-023-00256-4 (E8.5)-E9.5 development. In accordance, quantitative proteomic and transcriptional profiling of mouse embryonic hearts has revealed that sex-biased RNA expression and protein expression appear as early as E9. 5 (ref. 13); this stage is significantly earlier than the gonadal expression of the testis-determining gene Sry, and it precedes mammalian gonadal development. Interestingly, a significant number of the sexbiased proteins are associated with types of congenital heart disease that display sex differences 13 . Given that Sry is not expressed in the embryo before E11, these studies suggest that sex differences at E9.5 are regulated by Sry-independent pathways that function before gonadal development.
Similar to studies of adult heart tissue, transcriptional profiling and quantitative proteomic-based approaches with embryonic tissue have shown a gross disparity between genes that show sex differences versus proteins that show sex differences. Only 4% of the mRNAs that displayed at least a 1.5-fold change between males and females exhibited corresponding changes at the protein level in the adult heart, and only three mRNAs at E9.5 that displayed a change between males and females also showed differences in protein level 13 . Thus, it appears that post-transcriptional mechanisms are the principal contributors to male-female cardiac differences during embryogenesis and in adults.

Cardiac sex differences and aging
Intriguingly, sex-biased gene and protein expression are temporally regulated 13,20,118,119 . Transcriptomic analyses have identified more sex differences in male-female transcriptomes in sexually immature juvenile mice and reproductively incompetent aged females than in normal adult females 18,20 . In agreement with these findings, Han et al. have detailed male-female age-related changes in cardiac splicing and exon usage 120 . These findings are consistent with the observation that, at E9.5 and in the adult heart, most cardiac sex-differentiated proteins displayed temporal differences 13,20 . In aggregate, these findings In XY* mice, an aberrant pseudoautosomal region (PAR) on the Y* chromosome leads to unusual recombination with the X chromosome, producing four genotypes on a C57BL/6J background. The minute Y* X chromosome is nearly equivalent to a PAR, containing very few X genes. Two genotypes (XO, XY) have one X chromosome, and two genotypes (XX, XXY) have two X chromosomes. The level of expression of X-escapee genes is a function of the number of X chromosomes. When used with the FCG model, the XY* model can determine whether a sex-chromosome effect is caused by X or Y genes. The model has been used fruitfully in the study of ischemia-reperfusion injury 146 , pulmonary hypertension 148 , aortopathies such as aneurysms, stroke in aged mice 95 and metabolism and adiposity 149

Fig. 3 | In mice
, cardiac tissue is specified and determined, and forms a beating heart, before the differentiation of bipotential gonads into a testis or ovary. Relative time scale of heart and gonadal development in mice. The heart is formed from two populations of cells that give rise to defined structures in the adult heart: the first heart field (FHF), which gives rise to the left ventricle (LV), and the second heart field (SHF), which gives rise to the right ventricle (RV) and outflow tract (OFT). CNC, cardiac neural crest cell; Epi, epicardium.

Review article
https://doi.org/10.1038/s44161-023-00256-4 imply that a subset of male-female differences in cardiac gene and protein expression are controlled by genes on the X chromosome or Y chromosome and thus by a sex-chromosome pathway(s) and not by sex hormones. Heart failure is an epidemic: it affects more than 18 million people worldwide, including more than 6.2 million in the United States 121 . The primary cause of systolic heart failure is dysfunction from massive loss of cardiac muscle, often the result of acute MI. As mammalian hearts have limited natural regeneration in response to injury, a more thorough understanding of heart failure is essential to treatment. Recent studies have demonstrated sex differences in heart failure, including that MI events are less common in premenopausal females 122 . In an effort to begin to address these issues, Pullen et al. conducted a comprehensive study of male and female mice in response to permanent coronary ligation 123 . Collectively, the findings showed improved survival and more complete recovery in female versus male mice. Moreover, male mice demonstrated an increase in markers and physiology of non-resolving inflammation 123 . Together, these findings may provide a predictive signature for response to MI treatment in men and women.

Current status
A key goal is to define the gene or group of genes on the X chromosome that initiates cardiac sex disparities. Relatively few proteins in the heart and liver 124 that display sex differences map to sex chromosomes. In the heart, a potential 39 genes on the X chromosome, functioning alone or in combination, appear to be at least partly responsible for male-female sex differences in adult mouse hearts (Fig. 4).
Recent studies have demonstrated that the Y chromosome is required in bone marrow cells for cardiac function, and that loss of the Y chromosome was associated with an increase in fibrosis 125 . However, none of the genes or proteins identified in the sex-chromosome pathways in adult or embryonic heart tissue mapped to the Y chromosome 13 . Thus, it appears that gene(s) on the Y chromosome do not directly regulate the sex-chromosome pathway but may indirectly influence cardiac function.
On the basis of the observation that X-linked genes act in a dosagespecific manner to initiate cardiac sex differences (Box 1), we suggest that the candidate genes might escape X-inactivation (although other mechanisms are possible, for example, ref. 80). In accordance with this proposal, Aquado et al. 126 have recently demonstrated a correlation between genes that escape X-chromosome inactivation, such as Bmx and Sts, and sex myofibroblast activation 126 .
Although it has been established that a defined set of genes on the X chromosome in mice and humans escapes X-inactivation, X-inactivation occurs in a tissue-and cell-type-specific and age-specific manner [127][128][129][130] . Critically, no studies have demonstrated which of the X-chromosome genes escape inactivation in the embryonic heart or in a specific cardiac cell type. Understanding how X-linked genes function in a tissue-specific manner depends on identifying the genes that escape X-inactivation in cardiac tissue during development and identifying the genes on the X chromosome that initiate sex disparities in a cardiac-specific manner.
Collectively, it appears that post-transcriptional mechanisms are at least partially involved in propagating and maintaining protein sex differences (Fig. 5). One possibility for dissimilar male-female protein expression could be a set of male-female alternatively spliced cardiac mRNAs. Han et al. have recently defined a set of cardiac transcripts that show sex-specific alternative exon usage and alternative splice forms between 4-month-old and 20-month-old mice 120 , thus raising the possibility that a set of splicing factors may lead to a sex differential protein stability.
A second possibility is that transcripts that encode the same protein are different in the untranslated regions (UTRs) in males versus females. Differences in the 5′-UTRs of mRNAs, including cardiac genes 131 , can alter translational efficiency 132,133 . The critical nature of this variation is emphasized by the large number of disease-causing mutations that map to 5′-UTRs 132 . Similarly, alternative transcriptional start sites or alternative polyadenylation of a transcript can alter mRNA translational efficiency and stability 134 . In addition, alternative transcriptional start sites or alternative polyadenylation of any single gene can be regulated in mammals in a cell-, tissue-, or sex-specific manner (for example, refs. 131,135). This regulation is exemplified by people with cardiac disease, among whom there is a direct association between changes in the length of polyadenylation and severity of disease 136 Fig. 4 | Genes or genes coding for proteins on the X chromosome that display male-female differences in cardiac expression. Genes reported to show malefemale differential expression at the mRNA or protein level in adult mouse hearts mapped onto the human X chromosome 13,18,120 .

Molecular bases of sex differences
• Cardiac mRNAs and cardiac proteins differ in abundance between males and females. • Cardiac sex disparities are controlled not only by gonadal hormones, but also by sex-chromosome mechanisms. • Genes encoding proteins in the sex-chromosome pathway are both X-linked and autosomal, although none that have been are associated with the Y chromosome. • Genes on the X chromosome are both necessary and sufficient to drive the sex-chromosome pathway. • Genes on the X chromosome function in a dose-dependent manner. • Male-female cardiomyocyte differences begin at or shortly after the time cardiomyocyte fate is determined, which is before the gonads differentiate and produce hormones. • Most male-female differences are established posttranscriptionally.

Review article
https://doi.org/10.1038/s44161-023-00256-4 sites or alternative polyadenylation 137 . Defining sex-specific posttranscriptional events requires specific chemistries and approaches for each modification. For example, identification of alternative promoters or alternative 5′ transcription start sites requires approaches such as CAGE technology 138 , and defining alternative polyadenylation requires approaches such as Poly(A)-ClickSeq 139,140 . To date, there are no reports of sex-specific PolyA or UTR processing of mRNAs in the heart, and systematic studies to investigate this possibility are needed. A third prospect is that males and females differ in cardiac protein turnover, a possibility that has not been reported on. However, many proteins encoded by the X chromosome are expressed in the heart and function in the F-box-Cullen pathway, for example, CUL4B and UBE2A. Hence, defining the post-transcriptional events that propagate and maintain male-female cardiac protein differences will require an array of integrated, system-based studies.

Conclusions
Transcriptomic analyses have identified sex-specific profiles associated with human disease and animal models of heart failure 11,75,141,142 . However, as there is a discordance between mRNA and protein abundance, even the limited data on sex-chromosome proteins will extend potential disease candidates, aiding in the diagnosis of congenital and adult-onset heart disease. The data so far have been mostly correlative. Therefore, knowledge of components of the sex-chromosome pathway may be used to prevent inappropriate therapy in individuals who have sex-chromosome-associated defects and are therefore unresponsive to hormonal therapy. The ultimate goal from these studies is to build on current knowledge to define potential therapies for congenital and adult heart disease.

Perspectives
Heart failure is an epidemic that affects more than 18 million people worldwide, including more than 6.2 million in the United States 121 . The primary cause of systolic heart failure is dysfunction from massive loss of cardiac muscle, often the result of acute MI. People show highly varied responses to MI, ranging from full recovery to heart failure that requires organ transplantation. This variability makes it difficult to predict responses and outcomes, and to personalize treatments, problems that could be overcome by better biomarkers of recovery. Recent studies have demonstrated that sex differences are a contributing factor, and that MI events are less frequent in premenopausal females 122 . In an effort to begin to address these issues, Pullen et al. conducted a comprehensive study of male and female mice in response to permanent coronary ligation 123 . Collectively, their findings showed improved survival and more complete recovery in female versus male mice. Moreover, male mice demonstrated an increase in markers and physiology of non-resolving inflammation 123 . Together, these findings may provide a signature for male and female patients that is predictive of response to MI treatment. A problem that remains in characterizing cardiac sex differences and their treatment is the sex disparities in clinical trials. This issue is compounded by the absence of sex chromosomes in most genome-wide association studies. In an attempt to address these deficiencies, the US National Institutes of Health (NIH) set in place in 1993 the NIH Revitalization Act, and in 2016, the NIH sought to have all grant applications report sex as a biological variable 143 . To date, these policies have found varying success 62,144 . For example, Geller et al. found that less than a third of NIH-funded trials reported in 14 medical journals included sex as a variable 145 . The NIH has taken further actions to correct oversights (reviewed in Arnegard et al. 144 ). These initiatives are complemented by proposals that have been put forth by clinicians

Review article
https://doi.org/10.1038/s44161-023-00256-4 and scientists in a number of editorials 30,62,68 . Suggestions include having all funding bodies, as well as all journals, require appropriate sex representation and requiring upfront justification for all single-sex studies, for example 62 . A more complex issue is how to enlist more women in clinical trials and how to translate laboratory studies back to the clinic. Although proposals have been presented (for example, ref. 62), these problems remain a work in progress. It will be essential to address these issues to uncover potential targets for drug therapies for sex-biased disease states, including possible differences in treatment regimens for men and women.