Sex differences in fetal development and sensitivity to intrauterine perturbations are well documented, however, there remains a paucity of information on specific molecular and/or genetic mechanisms responsible for divergent male and female development and health outcomes. We previously showed that OGT, a protein encoded by an X-linked gene, contributes to female fetal protection from prenatal stress, in part through its regulation of the H3K27 methyltransferase, EZH2 31,32,34. Increased OGT is associated with higher levels of H3K27me3 in the female placenta, in both mouse and human tissue 34. However, less is understood as to the sex-specific roles of the X- and Y-linked H3K27 demethylases, UTX and UTY. Comparisons of UTX and UTY amino acid sequences demonstrate extensive sequence homology, reaching 88% homology, both within and outside of the catalytic domain, yet the function of UTY remains less clear 16,17. Important to sex-differences in H3K27me3-mediated transcriptional regulation in the placenta, XY tissues naturally have reduced OGT in addition to the presence of UTY, potentially culminating in overall reduced levels of H3K27me3 and reduced transcriptional repressive control in tissues such as the placenta 34. As such, we hypothesized that these differences may account for an increased vulnerability for males in utero, especially for neurodevelopmental risk 4–6, 56. We previously reported no sex difference for placental Utx expression suggesting that Utx, unlike Ogt, may not escape X inactivation in the placenta 31,34. Therefore, we hypothesized that if UTY indeed had transcriptional influence, either via potential demethylation of H3K27me3 and/or other as of yet unidentified activity, it would serve as a molecular mechanism that could direct male-specific fetal development. Therefore, we generated a Uty-overexpressing transgenic mouse and utilized a transcriptomic approach to assess potential changes in placental and fetal brain development.
Initial phenotypic assessment found no overt developmental changes or outcomes in the XX + Uty mice. As Uty is not a master regulator (i.e., not known to be at the top of a gene regulation hierarchy) and limited protein-protein interactions have been described for UTY, we did not anticipate that over-expression of Uty in a wild-type female mouse would produce profound phenotypic changes. We examined global health outcomes in the XX + Uty mice and confirmed that they showed no signs of developmental abnormalities or health consequences resulting from Uty expression. Significant differences in gene expression based on conservative corrections for multiple comparisons and distinct differences in clustering assessed by cluster dendrogram and principal components analysis were not observed in the placenta or hypothalamus. We interpret these findings as an indication that we have not caused significant, potentially detrimental changes to the genome by the random insertion approach. Furthermore, the lack of baseline changes in body weight and functionality of the stress-axis (HPA) in XX + Uty compared to XX animals indicates that growth, metabolism, and neuroendocrine regulation were relatively normal, under control housing and feeding conditions.
Utilizing our transcriptomic results as a hypothesis generating dataset, we adjusted our statistical parameters to examine relationships and patterns based on unadjusted p-values and in all detectable transcripts from our list of differentially expressed genes. We first compared gene sets between tissues to identify those that were similarly regulated in both. In the placenta and developing hypothalamus of XX + Uty animals, we found 206 genes that were enriched for biological processes including autophosphorylation and kinase activity. Autophosphorylation of protein kinases is a common process by which the enzyme adds a phosphate group to itself, altering its catalytic activity. Examples of such tyrosine kinases include epidermal growth factor receptor, insulin receptors, and SRC kinases that are important for proliferation, differentiation, and metabolism 57,58. In addition to the shared biological processes, we also identified the gene for centrosome-associated protein 350 (Cep350) as a similarly downregulated gene in XX + Uty placenta and hypothalamus. CEP350 is required for the anchoring of microtubules and recruitment of some nuclear receptors, such as peroxisome proliferator-activated receptor alpha (PPARα), a major regulator of lipid metabolism. Centrosomal genes, including Cep350, play a pivotal structural role in primary cilia that are antenna-like sensory organelles found in the placenta and hypothalamus to coordinate developmental processes and metabolic homeostasis 59–61. Furthermore, defective cilia have been implicated in the pathogenesis of preeclampsia and metabolic disorders 60,61. Dysregulation of Cep350 associated with Uty expression likely deserves further interrogation as a male-specific source of vulnerability for pregnancy complications and metabolic dysregulation. These overlapping gene sets in the placenta and hypothalamus of XX + Uty animals consistently suggest an association between Uty expression and regulation of genes important for metabolism.
Unique to the placenta, almost 50% of the 14,230 detected transcripts in the placenta showed differential expression profiles that were completely distinct to XX + Uty animals. Gene ontology (GO) identified biological processes related to transcription, translation, and post-translational protein modifications, suggesting that UTY plays a regulatory role in gene expression and protein synthesis and modification. Not surprisingly, we saw a ~ 20% overlap in gene regulation between XX + Uty and XX females, but we focused on the potential masculinizing effects of UTY on gene regulation, (i.e., pattern similarities between XX + Uty and XY mice), as evidence of novel UTY function. We found that 16.4% of detectable transcripts were male-biased in their expression pattern in the placenta where biological processes vital to cellular signaling, metabolism, and inflammation were enriched, including protein ubiquitination 50.
Ubiquitination is a common post-translational protein modification that tags proteins for sorting, localization, trafficking, and degradation. Therefore, protein ubiquitination is essential to both intracellular and extracellular communication that can impact diverse cellular processes such as DNA transcription, cell cycle, ribosome biogenesis, and inflammation. For example, new evidence suggests that ubiquitinated proteins are directed toward extracellular secretion via membrane-bound vesicles known as extracellular vesicles (EVs). EVs serve as carriers for a variety of proteins, microRNAs, mRNAs, and lipids transporting these biological molecules to neighboring cells and mediating physiological processes involved in development and homeostasis 62. While the physiological consequences of altered protein ubiquitination associated with Uty expression are impossible to discern without further studies, it may be one underlying mechanism that contributes to several other biological processes enriched in our masculinized XX + Uty female dataset, including vesicle transport, cholesterol homeostasis, and inflammation.
Regulation of immune tolerance is essential for successful pregnancy and relies on coordination and communication between neighboring cells. The semi-allogeneic placenta derives half of its genetic material from the mother and half from the father which leads to the expression of foreign, paternal, proteins in the intrauterine environment 63,64. Therefore, several mechanisms are in place to protect the placenta from being targeted by the maternal immune system 65. We observed alterations in the regulation of pathways associated with inflammatory and immune responses in placentas with Uty expression that raises some questions regarding the role of UTY in chronic placental inflammation, a phenotype more often observed with male fetuses and associated with premature delivery 66,67. Other studies have reported associations between experimentally reduced Uty expression and changes in immune response, although there are inconsistencies in the directionality of immune dysregulation 68,69.
Disruption of cholesterol homeostasis also deserves further interrogation as a pathway that is partially regulated by UTY in the placenta. The placenta acquires cholesterol from maternal circulation and synthesizes cholesterol that is subsequently transferred to the fetus or used by the placenta to make hormones. Cholesterol levels in healthy pregnancies increase over time. However, too much cholesterol is associated with increases in lipid peroxidation, reactive oxygen species, and inflammation, as well as pregnancy complications including preeclampsia and fetal growth restriction 70. These similarly regulated pathways between XX + Uty and XY placentas provide novel evidence that UTY might play a role in male-specific vulnerability to inflammation and pregnancy complications in utero.
Unique to the hypothalamus, 14,553 transcripts were detected. Compared to the placenta, there were fewer genes distinct to the XX + Uty females (36.5%), and more genes with similar expression profiles between XX + Uty and XY (24.8%). We also detected a greater number of genes (39) overlap between the masculinized XX + Uty gene list and previously published male-biased genes in the developing mouse hypothalamus, including estrogen receptor alpha (Esr1), and several estrogen responsive genes, insulin-like growth factor 1 (Igf1), oxytocin (Oxt), and protein kinase C delta (Prkcd) 34, 51–55. In the developing brain, testosterone aromatized to estradiol masculinizes the organization of neural circuits 71–76. Thus, it is interesting to suggest that UTY may have a more ‘masculinizing’ effect on the hypothalamic transcriptome compared to the placenta. However, the biological pathways enriched in both the XX + Uty unique gene list and masculinized XX + Uty gene list were less specialized in the brain and more broadly related to the regulation of gene expression, protein phosphorylation, and RNA splicing and processing that may reflect cell type specific and/or timing of analysis outcomes. Our findings, along with recently reported evidence that UTY plays a role in male-specific neural stem cell differentiation, demonstrate that UTY may be a contributing molecular mechanism by which sex differences in brain organization arise 23.
The capacity for UTY to function as an H3K27me3 demethylase, like its X-linked homologue UTX, has been scrutinized for decades 15,16,18,77. Utilizing previously published data from our lab, we integrated corresponding genes that were identified by H3K27me3 ChIP-Seq with our masculinized XX + Uty placenta dataset 34. More specifically, we limited the placental ChIP-Seq data set so that we were only examining genes with transcriptional start sites that had counts lower in males compared to XX females, with the assumption that these genes were demethylated in a more male-specific direction. Interestingly, a small percentage of the genes from the masculinized XX + Uty group (9.5%) overlapped with the filtered ChIP-Seq gene list, however, that small percentage accounted for nearly 50% of the total masculinized XX + Uty genes. From these data we are not able to determine whether UTY may participate in demethylation of H3K27me3 in a direct or indirect manner, however, these findings suggest that a substantial portion of the masculinized XX + Uty genes have a reduction in this transcriptional repressive mark in the presence of UTY.
In examination of adult phenotypes for effects of UTY, while adult XX + Uty females showed no significant differences from XX females under normal control conditions, they presented with surprising alterations in body weight and glucose tolerance following a 5-week calorically dense dietary challenge. While we hypothesized that a detectable phenotype would resemble wild-type males, we observed that XX + Uty females did not present in a masculinized manner. Surprisingly, XX + Uty females at the end of the HFD exposure weighed less and had a greater glucose clearance compared to either wild-type males or females. As this phenotype was unique to XX + Uty females, we interpret this finding as a potential role for UTY in metabolic regulation that may be related to developmental changes or current adult activity of UTY and may be distinct from XY and XX mice as a result of conflicting or compounding effects of UTY on an XX background. In examination of food consumption as an explanation for body weight differences, we found that XX + Uty females in fact consumed fewer calories than males on the HFD. However, no significant differences were detected between XX + Uty females and XX females suggesting again unique interactions of UTY on an XX background that are impacting metabolic processes. Future investigations into the mechanisms and hormonal regulation driving this Uty-associated phenotype could provide important insight about sex differences in metabolism and risk of metabolic disorders.