THE IMPACT OF SEX CHROMOSOME COMPOSITION ON GENE EXPRESSION AND REGULATORY DIMORPHISMS IN MOUSE EMBRYONIC STEM CELLS

Abstract

The variation between males and females constitutes the largest phenotypic dimorphism in any given species and, in humans, contributes to differences seen in the risk, incidence and response to treatment for a number of diseases. A primary point of divergence is driven by variation in the sex hormone composition between males and females. However, we hypothesize that the sex chromosome composition alone drives dramatic events to which subsequent hormonal exposure either ameliorates or potentiates these differences. Methods to study early development are often limited to in vitro systems or in model organisms, such as in mouse. Human embryonic stem (ES) cell lines are available but are of limited access, are of high passages, and require more stringent growth conditions. Direct comparisons between male and female lines are extremely difficult, due to the abundance of multiple X chromosome statuses as well as non-random X chromosome inactivation. In the absence of more refined culture methods for human ES cells, we opted to use mouse ES cell lines derived in-house as this would enable us to determine allelic expression patterns as well as more easily maintain pluripotency and random X chromosome inactivation with established culture parameters. As such, to test our hypothesis, we derived an extensive panel of low passage mouse embryonic stem (ES) cell lines from reciprocal crosses between the C57BL/6 and CAST/EiJ mice. In total, I had access to over 20 unique mouse ES lines, including two of which show aneuploidy with loss of one of the sex chromosomes. The addition of these two lines to the experimental design grants us the ability to tease apart the individual contributions of the X and the Y chromosome in early development. Additionally, sex chromosome aneuploidies have yet to be evaluated in terms of their effects on the epigenome as well as their influence on directed differentiation in vitro. To set the foundation for our studies, we first performed a series of RNA-seq analyses in which we expanded the number and variation of sex-specific differences from previous reports using microarray. We then interrogated the contribution from each of the sex chromosome complements on gene ontology. Additionally, we identified and validated sex-specific alternative splicing events, for which there is very limited reporting. With an emphasis on genome-wide regulatory patterns, we then performed an unbiased weighted gene co-expression network analysis (WGCNA) for which we identified a key sex-specific expression module. The main driver of this module was the gene encoding Prdm14, a pivotal transcription factor involved in pluripotency. Luciferase assays with a known Prdm14-responsive enhancer showed higher expression when transfected into female than in male ES cells. Because Prdm14 is more abundant in female ES cells, this suggests that the dosage of this transcription factor is a key factor of its capacity to activate gene expression. This is the first ever documented sex-specific differential enhancer activity and further underscores the need to not only evaluate expression but functionality of the protein product within biological systems. Prdm14 has the dual capability of promoting and repressing transcription depending on its binding partners. Evaluation of histone modifications overlapping with known Prdm14 binding motifs in the promoters of the co-expressed genes revealed a unique signature between the male and female mouse ES lines. Based on our analyses, we hypothesize that the higher Prdm14 abundance in XX ES cells can activate gene expression even if genes harbor a repressive histone modification in their promoters. We propose that the lower abundance of Prdm14 in male ES cells can only activate genes that do not exhibit repressive histone modifications. With this information in hand, we then performed a directed differentiation assay to the cardiomyocyte lineage. From these experiments, we identified an XX-specific impairment to differentiate without chromosome loss. Additionally, the 39,X lines exhibited dysregulation of cardiac-related genes, potentially correlating with the defects seen in Turner syndrome patients. Overall these findings help to expand upon an underrepresented field in the basic sciences, namely the underlying contributions of the sex chromosome complement on gene expression and regulatory dimorphisms.