Comparing the roles of sex chromosome-encoded protein homologs in gene regulation

In this review, Lavorando et al. discuss the roles of five pairs of sex chromosome-encoded protein homologs (X–Y paired proteins; namely, SRY/SOX3, ZFX/ZFY, KDM5C/KDM5D, UTX/UTY, and TBL1X/TBL1Y) in the regulation of chromatin and transcription. The authors highlight the nonhormonal contributions of these X–Y paired proteins to sexually dimorphic gene expression and sex biases in disease.

Abstract

The X and Y chromosomes play important roles outside of human reproduction; namely, their potential contribution to human sex biases in physiology and disease. While sex biases are often thought to be an effect of hormones and environmental exposures, genes encoded on the sex chromosomes also play a role. Seventeen homologous gene pairs exist on the X and Y chromosomes whose proteins have critical functions in biology, from direct regulation of transcription and translation to intercellular signaling and formation of extracellular structures. In this review, we cover the current understanding of several of these sex chromosome-encoded protein homologs that are involved in transcription and chromatin regulation: SRY/SOX3, ZFX/ZFY, KDM5C/KDM5D, UTX/UTY, and TBL1X/TBL1Y. Their mechanisms of gene regulation are discussed, including any redundancies or divergent roles of the X- and Y-chromosome homologs. Additionally, we discuss associated diseases related to these proteins and any sex biases that exist therein in an effort to drive further research into how these pairs contribute to sexually dimorphic gene regulation in health and disease.

Outside of the reproductive organs, sex-dependent protein expression was first observed in the rat liver >40 years ago (Gustafsson et al. 1983; Roy and Chatterjee 1983). After these early advances, the study of sexually dimorphic gene expression exploded with the rise in affordable next-generation sequencing. Depending on the analysis performed, different studies have reported a wide range of genes that are expressed in a sex-biased manner, spanning from ∼1000 genes (Melé et al. 2015) to >13,000 genes (Oliva et al. 2020). Despite this, many studies have come to the conclusion that sex-biased gene expression is driven by both hormonal and nonhormonal factors (Melé et al. 2015; Naqvi et al. 2019; Lopes-Ramos et al. 2020; Oliva et al. 2020). This same conclusion can be drawn from mouse studies using the “four core genotypes,” in which wild-type XX and XY mice are compared with XY mice lacking Sry (so-called “XY females” that develop ovaries) and XX mice expressing Sry from an autosome (so-called “XX males” that develop testes) (De Vries et al. 2002). These studies highlight the importance of considering the nonhormonal sex determinants when looking for the underlying causes of sexual dimorphism in health and disease, particularly the sex chromosomes and the genes contained therein.

Sex chromosomes were first discovered in 1905 by Nettie Stevens (1905), with human sex chromosomes being discovered 18 years later in 1923 (Painter 1923). All humans possess at least one X chromosome, though there are a great number of survivable X-chromosome aneuploidies, with up to five X chromosomes being reported (49; XXXXX) (Linden et al. 1995). Roughly half the population also possess the much smaller Y chromosome that contains, among other genes, the aforementioned gene SRY, which was found to be the determinant of traditionally “male” sexual development in 1990 (Berta et al. 1990; Gubbay et al. 1990; Sinclair et al. 1990) after a nearly 30 year search (Ford et al. 1959; Jacobs and Strong 1959; Welshons and Russell 1959; Vergnaud et al. 1986). The evolution of SRY was a key event in the evolution of the sex chromosomes as they exist today (Fig. 1). The modern sex chromosomes began divergence from their autosomal ancestors between 148 million and 180 million years ago with the evolution of SRY on the autosomal ancestor of the Y chromosome (Wallis et al. 2008). Following the development of SRY, the Y chromosome further diverged from the X, losing 97% of its original genes and enriching in genes thought to be involved in male sexual development (Lahn et al. 2001). In recent years, interest in the remaining protein-coding genes on the Y chromosome (particularly those with homologs on the X chromosome) and their roles in health and disease has greatly expanded (Bellott et al. 2014; Cortez et al. 2014; Godfrey et al. 2020; Nguyen et al. 2020; Shi et al. 2021; Owens et al. 2023).

Figure 1.

Evolution of the modern human sex chromosomes from their autosomal ancestors.

Although the Y chromosome is relatively gene-poor compared with the X chromosome, the human X and Y possess 17 homologous genes (Bellott et al. 2014; Cortez et al. 2014). Many of the X-linked homologs of these X–Y pairs have been implicated in the regulation of gene expression at the transcriptional, epigenetic, translational, and post-translational levels (Bellott et al. 2014; Cortez et al. 2014; Owens et al. 2024). In contrast to the well-appreciated roles of the X-linked homologs, the Y-linked homologs are only recently being studied (Walport et al. 2014; Nguyen et al. 2020; Gong et al. 2021; Shen et al. 2022; Owens et al. 2023; Yanas et al. 2023). This relative paucity of information was likely due to the assumption that the protein expression of Y-linked homologs was relegated to the reproductive system, an assumption that was challenged by mass spectrometry studies that found the Y-linked homologs in tissues throughout the body (Godfrey et al. 2020). These findings suggest that individuals with only X chromosomes (XO, XX, XXX, etc.) express only the X-linked homologs, while individuals with X and Y chromosomes (XY, XXY, XYY, etc.) express both the X-linked homologs and the Y-linked homologs in certain tissues. This raises the possibility that some degree of sex-biased gene expression may be due to differences between these X–Y paired proteins in either their activity, expression level, or expression pattern.

Below, we discuss five X–Y protein pairs that are known to play a role in epigenetic and transcriptional regulation (SOX3/SRY, ZFX/ZFY, KDM5C/KDM5D, UTX/UTY, and TBL1X/TBL1Y) (Fig. 2), focusing on the different genes and regulatory networks affected by each pair. Where applicable, we discuss any diseases related to the dysregulation of these proteins and any sex biases observed in these diseases. Additionally, we highlight instances where different X–Y pairs may be functionally redundant and instances where they may be functionally divergent, hypothesizing as to how these differences may contribute to sex-biased gene expression in health and in disease.

Figure 2.

Key interactions of the X–Y paired proteins discussed in this review. Names and relative positions of each X–Y paired gene are shown in the middle of the figure. Genes whose proteins are discussed in this review are shown in bold. (*) TSPY is duplicated ∼40 times at this position. (**) RBMY is duplicated at this position. (A) SRY, in complex with SF1, binds the TESCO element upstream of SOX9 and activates its transcription. (B) ZFX or ZFY bind to GGCCT repeats throughout the genome using its final three zinc finger domains. (C) KDM5C, in complex with the CoREST complex, demethylates H3K4me3/2 to repress transcription. (D) UTX, in its noncatalytic role, recruits p300 and MLL4 to activate enhancers. In a catalytic role, UTX demethylates H3K27me3 to activate HOX gene expression. (E) TBL1X scaffolds the NCoR/SMRT complex, which represses the transcription of target genes.

SRY and SOX3

The gene SRY (sex-determining region Y) is the major determinant of testicular development (the testes-determining factor [TDF]) (Berta et al. 1990; Gubbay et al. 1990; Sinclair et al. 1990). During fetal development in XY therian mammals (eutherians and marsupials), expression of the SRY protein drives the differentiation of gonadal precursor cells into testes (Koopman et al. 1991; Hacker et al. 1995; Wilhelm et al. 2007; Wallis et al. 2008). Since its identification as the TDF in 1990, numerous mechanisms of action have been hypothesized for SRY, ranging from repressing a negative regulator of testicular development (McElreavey et al. 1993) to mediating protein–DNA interactions and chromatin structures that govern sex determination (Pontiggia et al. 1994) to affecting pre-mRNA splicing (Ohe et al. 2002) to acting as a transcriptional activator of essential male-specific target genes (Thevenet et al. 2005). This last hypothesis was supported by the discovery that Sry is critical for the expression of the Sox9 gene (Sekido and Lovell-Badge 2008).

SOX9 is an autosomal gene essential for the differentiation of Sertoli cells in mouse and human development (Huang et al. 1999; Chaboissier et al. 2004). Sry, in a complex containing steroidogenic factor 1 (Sf1), binds to an enhancer region 1.4 kb upstream of the Sox9 transcription start site that is sufficient to cause Sox9 expression (testis-specific enhancer core of Sox9 [TESCO]) (Fig. 2A; Sekido and Lovell-Badge 2008). Interestingly, SOX9 binds to TESCO as well, suggesting that it regulates its own expression in the testis (Sekido and Lovell-Badge 2008). CRISPR/Cas9 deletion of TESCO reduced Sox9 expression in XY fetal gonads to ∼50% but did not induce sex reversal, indicating that other enhancers exist to maintain Sox9 expression once activated (Gonen et al. 2017). One such element, Enh13, which lies upstream of TESCO, is essential for upregulation of Sox9 and the subsequent initiation of mouse testis development, as its deletion results in XY mice developing ovaries (Gonen et al. 2018). Outside of its role in Sox9 upregulation, SRY has other target genes important for its role in sex determination, such as Tcf21 and Ntf3, which have functions in testicular morphogenesis (Bhandari et al. 2012; Li et al. 2014).

SRY is believed to have evolved from its X-chromosome-linked homolog, SOX3 (Stevanović et al. 1993). Although SRY and SOX3 form one of the most divergent pairs of X–Y paired proteins in amino acid sequence (∼44% identical), both proteins possess a high-mobility group (HMG)-box DNA binding domain (the defining feature of the SOX family of transcription factors) that is more homologous than the proteins as a whole (∼71% identical). Although significantly less well studied than SRY, SOX3 has some capacity to act as a transcription factor, as its upregulation in mice and humans can cause XX male sex reversal by upregulating the expression of Sox9 in a mechanism like that of Sry (Sutton et al. 2011). When co-overexpressed in mice with Sf1, Sox3 drove activation of TESCO to a twofold higher level than Sf1 alone, suggesting that Sox3 can (at least partially) perform the functions of Sry when overexpressed (Sutton et al. 2011). Multiple studies have indicated a role for SOX3 in proper gonadal function, such as in oocyte development, testis differentiation, and spermatogenesis (Weiss et al. 2003; McAninch et al. 2020). Using ChIP-seq, Sox3 was found to bind 778 putative genomic sites in the mouse testis, including promoter/enhancer regions for Ngn3 and the histone genes H3f4, Hist1h2aa, and Hist1h2ab (McAninch et al. 2020).

Beyond the reproductive tract, Sox3, as well as Sox1 and Sox2, are key regulators of neuroprogenitor (NP) cells throughout the nervous system (Bylund et al. 2003; Rizzoti et al. 2004; Rogers et al. 2013). The dysregulation of Sox3 in NP cells results in abnormal expression of at least 19 genes, including the homeobox gene Dbx1 (Rogers et al. 2014). Likewise, Sry is also expressed in nonreproductive tissues including heart, liver, kidney, and brain (Clépet et al. 1993; Mayer et al. 1998). It is unknown whether SRY is redundant to SOX3 in nonreproductive tissues or whether it performs divergent functions. Differential activity of SOX3 and SRY during neurodevelopment may underlie the nonhormonal sexual dimorphism of the brain (Sekido 2014), though further studies are required to investigate this hypothesis.

ZFY and ZFX

ZFY encodes a transcription factor with 13 C-terminal zinc finger DNA binding domains, a central nuclear localization sequence, and an N-terminal acidic transactivation domain. This, together with the fact that the ZFY gene is located in close proximity to the gene for SRY in both humans and mice (Fig. 2; Page et al. 1987; Mardon and Page 1989), drove hypotheses that ZFY was the TDF. When first identified, it was noted that the X chromosome had a gene that was highly homologous to ZFY that was later identified as ZFX (Schneider-Gädicke et al. 1989; Mardon et al. 1990). Due to their high sequence identity (92% identical in overall amino acid sequence; 97% identical across the zinc finger domains), ZFX and ZFY are often proposed to be redundant (Mardon et al. 1990; Palmer et al. 1990; North et al. 1991; Ni et al. 2020). Both ZFX and ZFY share a consensus binding motif in the genome (GGCCT, often in CpG-rich promotor regions) and activate the transcription of many of the same genes (Fig. 2B; Taylor-Harris et al. 1995; Rhie et al. 2018; Ni et al. 2020; San Roman et al. 2024). Recent work has identified that promotor binding by ZFX facilitates Pol II elongation by recruiting protein factors necessary for the acetylation of histone H4, such as the COMPASS-like complex, ZNF593, and TIP60 (Hsu et al. 2024). It is likely that ZFY can recruit many of the same protein partners, given their high homology.

This functional redundancy may be necessary for maintaining gene expression parity between XX and XY individuals. Using patient-derived cell lines with an array of sex chromosome ploidies (ranging from one to four Xs and zero to four Ys), San Roman et al. (2024) found that the inactive X chromosome and the Y chromosome drove similar gene expression patterns across several autosomal genes. These gene expression patterns are in large part driven by the activities of ZFY and ZFX (which escapes X inactivation) (San Roman et al. 2023). The investigators of this study proposed that the significant overlap in targets between ZFX and ZFY may be an evolutionary holdover from the autosomal ancestors of the sex chromosomes, which, being essentially identical, would naturally drive similar programs of gene expression.

Beyond driving this gene expression program, ZFX has been implicated as an oncogene in glioma (Zhou et al. 2011), renal cancer (Fang et al. 2014), and prostate cancer (along with ZFY) (Tricoli and Bracken 1993; Cai et al. 2018). Missense mutations of ZFX are often found in parathyroid adenoma (Soong and Arnold 2014) and have been recently found in cases of neurodevelopmental disorders (Shepherdson et al. 2024). These mutations are almost always located in or around zinc finger 12, in line with the fact that only the last three zinc finger domains (11–13) are required for DNA binding (Grants et al. 2010). Interestingly, no ZFY mutants have been found in parathyroid adenoma (Romano et al. 2017), and the majority of ZFX-related neurodevelopmental disorder cases were male (Shepherdson et al. 2024), suggesting that ZFY may be insufficient to compensate for loss of ZFX in these tissues, possibly due to differences in protein expression level.

KDM5C (SMCX/JARID1C) and KDM5D (SMCY/JARID1D)

The post-translational modification of histones, particularly their methylation, is a key epigenetic regulator of gene expression. Histone modifications are controlled by the activities of “writer” enzymes that deposit modifications and “eraser” enzymes that remove them. One such eraser, lysine-specific demethylase 5C (KDM5C, also known as SMCX or JARID1C), is an X-chromosome-encoded enzyme that erases trimethylation and dimethylation of lysine 4 on histone H3 (H3K4) to monomethylation (Iwase et al. 2007). The effects of KDM5C-mediated demethylation are somewhat context-dependent: At promoters, KDM5C restricts transcription; however, at enhancers, it stimulates activity (Outchkourov et al. 2013). In the nucleus, KDM5C forms a complex with other histone-modifying enzymes (such as histone deacetylases HDAC1/2 and the H3K9 methyltransferase G9a) and the transcriptional repressor RE1 silencing factor (REST, also known as neuron-restrictive silencer factor [NRSF]) (Fig. 2C; Tahiliani et al. 2007).

The presence of KDM5C is crucial for REST-mediated silencing in neurons, as KDM5C and REST co-occupy the promoters of REST target genes, and depletion of KDM5C significantly increased expression of REST target genes (Tahiliani et al. 2007). In the context of neurodevelopment, KDM5C directly targets genes involved in WNT signaling through acting on their promoter regions, including CTNNB1, encoding β-catenin, and the gene FOXG1, which is a master regulator of brain development (Hou et al. 2020; Karwacki-Neisius et al. 2024). In mice, the knockout of Kdm5c leads to spurious transcription of germline-restricted genes and repression of genes involved in neuronal maturation (Iwase et al. 2016). Beyond its role in the nervous system, KDM5C is essential for the repression of germline genes (Scandaglia et al. 2017) and the promotion of proper DNA replication (Rondinelli et al. 2015) and additionally plays a role in upregulating the long noncoding RNA Xist (Samanta et al. 2022).

The Y-chromosome homolog of KDM5C is KDM5D (also known as SMCY or JARID1D), which is the least studied member of the KDM5 family of proteins. KDM5D also harbors H3K4me3/2 demethylase activities, though it is less active than KDM5C (Iwase et al. 2007; Lee et al. 2007a). Although further research needs to be performed to understand the breadth of KDM5D target genes, KDM5C/KDM5D may have distinct targets (Mizukami et al. 2019). For example, in contrast to KDM5C, KDM5D acts in a complex with the protein Ring6a/MBLR, which is a component of the polycomb complex PRC1, where MBLR regulates its enzymatic activity (Lee et al. 2007a). These proteins directly associate and regulate the expression of the human Engrailed 2 gene through modulating its H3K4 methylation level (Lee et al. 2007a).

Although KDM5D may have divergent functions from KDM5C, both proteins are known tumor suppressors (Niu et al. 2012; Rondinelli et al. 2015; Wang et al. 2015; Li et al. 2016; Komura et al. 2018). Additionally, KDM5D may be able to compensate for KDM5C loss during cardiac development (Kosugi et al. 2020) despite its diminished catalytic activity (Iwase et al. 2007), suggesting either that KDM5D is active enough to support development during KDM5C loss or that KDM5D and KDM5C share noncatalytic functions. However, KDM5D may not be able to compensate for loss of function of KDM5C in neurodevelopment, as loss-of-function mutations in KDM5C cause the X-linked neurodevelopmental disorder Claes–Jensen syndrome, which appears more frequently in males (Claes et al. 2000; Jensen et al. 2005; Rujirabanjerd et al. 2010). As many of these mutations disrupt KDM5C's demethylase activity (Tahiliani et al. 2007), it is likely that proper neurodevelopment relies on KDM5C-mediated demethylation and that the reduced activity of KDM5D is insufficient. In support of this notion, a study examining Kdm5c KO male mice (Kdm5x^−/Y) and Kdm5c HET female mice (Kdm5x^–/+) identified increased dysregulation of gene expression in the forebrains of the Kdm5c KO males compared with the heterozygous females, suggesting that KDM5D expression is unable to completely compensate for loss of KDM5C (Bonefas et al. 2023). This study also identified sex-specific gene expression changes in these two mouse lines compared with their wild-type controls, further highlighting the sexual dimorphism of KDM5C deficiency, though more work is required to tease out the precise mechanisms by which this X–Y pair influences neurodevelopment in a sex-specific manner.

UTX (KDM6A) and UTY (KDM6C)

In addition to KDM5C, another X–Y paired histone lysine demethylase, H3K27me3/2 demethylase UTX (KDM6A), is encoded by the X chromosome (Agger et al. 2007; Hong et al. 2007; Lee et al. 2007b). UTX's demethylase activity is crucial for its regulation of HOX gene clusters (which are critical for development) (Agger et al. 2007; Lan et al. 2007; Lee et al. 2007b; Shah and Sukumar 2010). Additionally, UTX-mediated demethylation is necessary for the proper development of invariant natural killer T cells (Beyaz et al. 2017). Beyond demethylation, however, UTX is well appreciated for playing numerous noncatalytic gene regulatory roles. Catalytically inactive UTX is able to rescue differentiation of embryonic stem cells along the mesoderm and ectoderm lineages (Wang et al. 2012; Morales Torres et al. 2013). Additionally, knock-in of catalytically dead Utx allows for viable and fertile male and female Utx^−/− mice, suggesting that the enzymatic activity is dispensable for normal mouse development (Faralli et al. 2016).

UTX has several protein partners crucial for its noncatalytic activities. For example, UTX acts as a scaffold for the MLL3–MLL4 COMPASS complex (Fig. 2D; Lee et al. 2007b; Shpargel et al. 2012). In a catalytic-independent manner, UTX recruits MLL4 and p300 to the complex, which deposit H3K4me1 and H3K27ac in enhancer regions of the genome, thus activating them (Wang et al. 2017). In addition to the COMPASS complex, UTX interacts with RNA Pol II (Wang et al. 2013), the lysine acetyltransferase CREBBP (Wang et al. 2017), the chromatin remodeling proteins CHD4 and SMARCA4, and BRG1/BRM-containing SWI/SNF complexes (Miller et al. 2010; Lee et al. 2012; Tie et al. 2012; Gozdecka et al. 2018).

UTX shares many of these noncatalytic activities with its homolog, UTY (KDM6C). Although UTY is significantly less active than UTX (human UTY retains ∼2% the activity of UTX, and mouse Uty appears fully inactive) (Shpargel et al. 2012; Walport et al. 2014), UTY interacts with many of the same chromatin remodeling factors as UTX (Shpargel et al. 2012). Furthermore, mouse Uty rescues the loss of Utx in mouse embryos and supports mouse survival up until birth (Wang et al. 2012), and in a mouse model with hematopoietic-specific loss of Utx, hemizygous mice (Utx^−/Y, lacking Utx but expressing Uty) showed no differences in hematological phenotype or survival when compared with control (Utx^+/Y) mice, suggesting that Uty can perform many of the noncatalytic functions of Utx in hematopoietic stem and progenitor cells (Gozdecka et al. 2018).

Despite the above-noted overlap in functions, UTY appears to fall short as a tumor suppressor compared with UTX. While UTX is a well-appreciated tumor suppressor (Dunford et al. 2017), UTY is far less effective in this regard (Andricovich et al. 2018; Shi et al. 2021). In recent work, this disparity was traced back to sequence differences between both proteins’ core intrinsically disordered regions (cIDRs). Both UTX and UTY undergo liquid–liquid phase separation (LLPS) driven by their respective cIDRs, and this phase separation is crucial for UTX's organization of histone-modifying partner proteins (Shi et al. 2021). These UTX condensates facilitate its tumor suppression and prodifferentiation activities. As such, a UTX nonsense mutation that truncates the protein before the cIDR (555*) is the most common UTX mutation in many cancers (Shi et al. 2021). The UTY cIDR, however, causes it to form less dynamic condensates than those of UTX, diminishing its tumor-suppressive capabilities (Shi et al. 2021). Beyond the cIDR, UTX mutations have also been found associated with the neurodevelopmental disorder Kabuki syndrome (Miyake et al. 2013; Van Laarhoven et al. 2015; Piunti and Shilatifard 2016). UTY mutations are not reported in this syndrome, and male individuals are more severely affected than females, suggesting that UTY may not be able to compensate for UTX loss in human neurodevelopment (Bögershausen et al. 2016; Gažová et al. 2019).

TBL1X and TBL1Y

Transducin β-like 1, X-linked (TBL1X) often functions as part of multiprotein complexes, including the nuclear receptor corepressor (NCoR) and silencing mediator for tetanoid and thyroid hormone receptor (SMRT) transcription repression complexes (Fig. 2E; Guenther et al. 2000; Zhang et al. 2002). In addition to TBL1X, these complexes contain components such as histone-modifying enzyme HDAC3 and the intracellular signaling protein GPS2, in addition to TBL1X's autosomal homolog, TBLR1 (Guenther et al. 2000; Li et al. 2000; Zhang et al. 2002; Yoon et al. 2003). TBL1X-containing complexes are typically transcription repressors, as is the case with NCoR/SMRT, which is necessary for the repression of thyroid hormone receptor target genes (Ishizuka and Lazar 2003). Interestingly, TBL1X also plays a role in derepression of NCoR/SMRT targets. TBL1X and TBLR1 are necessary for ligand-dependent recruitment of the ubiquitin/19S proteasome to and subsequent degradation of the NCoR/SMRT corepressors, allowing for target gene activation (Perissi et al. 2004).

In addition to NCoR/SMRT, TBL1X and TBLR1 also form a well-studied interaction with methyl-CpG binding protein 2 (MeCP2), a reader of DNA 5-methylcytidine (5mC) with high expression in the brain (Shahbazian et al. 2002). Mutations in MeCP2 cause a severe pediatric neurological disorder called Rett syndrome (RTT) (Bienvenu et al. 2000). Interestingly, several RTT-related MeCP2 mutations disrupt the binding of MeCP2 to TBL1X/TBLR1, suggesting that inability to recruit TBL1X-related gene silencing machinery may be one possible pathological mechanism of Rett syndrome (Lyst and Bird 2015; Tillotson and Bird 2020).

In contrast to the wealth of literature surrounding TBL1X, its Y-linked homolog, TBL1Y, is considerably understudied. When expressed ectopically in HEK293T cells, TBL1Y coimmunoprecipitates with SMRT and leads to the repression of a luciferase reporter construct under the control of a known SMRT target promotor, indicating that it may be able to act in the same transcription repression complexes as TBL1X (Di Stazio et al. 2019). However, TBL1Y was found to be unable to trigger transcription repression in an earlier study using a GAL4-TBL1 fusion system (Yan et al. 2005), suggesting that TBL1Y-mediated repression may be impaired relative to TBL1X or that TBL1Y-mediated repression may be cell type-specific. Further studies are required to determine whether (and when/where) TBL1Y can perform the transcription repression functions of TBL1X or whether it has potentially divergent functions.

Much like the other X–Y paired proteins, TBL1X is linked to a myriad of diseases. For example, in the context of breast cancer, TBL1X has been found to cooperate with the transcription factor ZEB1 to repress CDH1 (encoding E-cadherin) expression and activate ZEB1, which contributes to epithelial-to-mesenchymal transition (Rivero et al. 2019). Loss-of-function mutations have been associated with diseases like hypothyroidism (Heinen et al. 2016) and sensorineural deafness (Bassi et al. 1999). Interestingly, mutations in TBL1Y have also been linked to hereditary hearing loss in males, suggesting that the presence of TBL1X alone cannot compensate (Di Stazio et al. 2019). In support of this, individuals with Turner syndrome (X0 karyotype) show increased occurrence of late-onset sensorineural hearing loss (Elsheikh et al. 2002; Morimoto et al. 2006; Di Stazio et al. 2019).

Conclusion

In recent years, studies of the mechanisms and implications of sex biases in human health and disease have greatly expanded. This expansion was due to a number of factors, ranging from federal policy changes (see the National Institutes of Health's policy on sex as a biological variable [https://orwh.od.nih.gov/sex-gender/orwh-mission-area-sex-gender-in-research/nih-policy-on-sex-as-biological-variable]) to technological advancements in sequencing (The GTEx Consortium 2013) and mass spectrometry (Godfrey et al. 2020; Gelfand and Ambati 2023). One key finding of these studies is that biological sex differences do not begin and end with the sex hormones and that the nonhormonal contributions of the sex chromosomes must be considered for a more complete picture of human health and disease (McCarthy and Arnold 2011). In particular, the identification of 17 X–Y paired proteins (Bellott et al. 2014; Cortez et al. 2014) has proven fertile ground for the discovery of potential mechanisms of biological sex differences. In this review, we have highlighted five of these X–Y paired proteins that play roles in the regulation of gene transcription. Although these pairs can certainly perform overlapping functions, such as the shared gene expression program activated by both ZFX and ZFY (San Roman et al. 2024), one member of the pair may be insufficient to perform the functions of the other, such as in the case of SOX3's inability to activate expression of SOX9 (Sutton et al. 2011) or UTY's insufficiency as a tumor suppressor relative to UTX (Shi et al. 2021).

A major commonality between all five of these protein pairs is their link to neurodevelopment. Sox3 is required for proper regulation of NP cells (Rogers et al. 2013); ZFX mutations have recently been linked with neurodevelopmental disorder (Shepherdson et al. 2024); mutations in both KDM5C and UTX are causative of Claes–Jensen syndrome and type II Kabuki syndrome (Claes et al. 2000; Jensen et al. 2005; Rujirabanjerd et al. 2010; Miyake et al. 2013; Van Laarhoven et al. 2015; Piunti and Shilatifard 2016), respectively; and mutations in TBL1X and TBL1Y are associated with sensorineural deafness (Fig. 3A; Bassi et al. 1999; Di Stazio et al. 2019). Beyond the pairs discussed here, other X-linked homologs, such as DDX3X (Lennox et al. 2020), NLGN4X (Nguyen et al. 2020), and USP9X (Homan et al. 2014), are implicated in neurodevelopmental disorders as well. Although striking, this commonality is not entirely surprising, as the brain is often cited as one of the more sexually dimorphic nonreproductive organs (Arnold and Burgoyne 2004; McCarthy and Arnold 2011; Naqvi et al. 2019; Oliva et al. 2020). The findings outlined in this review raise the hypothesis that many of the sex differences observed in the brain may be due to differences between X–Y protein pairs.

Figure 3.

Contribution of X–Y paired proteins to sex-biased diseases. (A) Summary of neurodevelopmental disorders related to the X–Y paired proteins discussed. (*) ZFX-related neurodevelopmental disorder was reported in 2024 (Shepherdson et al. 2024), and more patient information is needed to accurately describe the potential sex bias. (B) Y-linked homologs may be less active enzymes than their X-linked homologs, as is the case with UTX and UTY. (C) X- and Y-linked homologs may be expressed in different brain tissues/regions or may be expressed in the same region at different developmental stages (as noted by the clock icons next to the proteins). (D) Y-linked homologs may be expressed to a lower level than X-linked homologs.

When considering that many of the neurodevelopmental disorders listed above either occur more frequently in males or are more severe in males, it may be the case that the Y-chromosome-encoded homolog of many of these pairs is insufficient to make up for the loss of the X-chromosome-encoded homolog. This insufficiency may take several forms, such as (1) the Y homolog is catalytically weaker than the X (in the case of enzymes) (Fig. 3B); (2) the Y homolog is not expressed in the same region of the brain or during the same developmental state as the X homolog, meaning the Y homolog cannot make up for loss of the X homolog (Fig. 3C); and (3) the Y homolog is expressed at the same time and in the same place as the X homolog but at too low a level to make up for the loss of the X homolog (Fig. 3D). Although Y-linked homologs appear uniformly less capable of supporting brain development than X-linked homologs, they are often just as capable of supporting the development of other tissues (Gozdecka et al. 2018; Kosugi et al. 2020). As these protein pairs become more and more well studied, it will be essential to parse when and where Y-linked homologs are sufficient or insufficient to support tissue development and function. Such an undertaking will require the combined efforts of genetic, biochemical, and developmental studies to gather a complete picture of how the X–Y paired proteins contribute to human health and disease.

Competing interest statement

The authors declare no competing interests.