Non-Canonical and Sexually Dimorphic X Dosage Compensation States in the Mouse and Human Germline

Abstract

Somatic X dosage compensation requires two mechanisms: X inactivation balances X gene output between males (XY) and females (XX), while X upregulation, hypothesized by Ohno and documented in vivo, balances X gene with autosomal gene output. Whether X dosage compensation occurs in germ cells is unclear. We show that mouse and human germ cells exhibit non-canonical X dosage states that differ from the soma and between the sexes. Prior to genome-wide reprogramming, X upregulation is present, consistent with Ohno's hypothesis. Subsequently, however, it is erased. In females, erasure follows loss of X inactivation, causing X dosage excess. Conversely, in males, erasure leads to permanent X dosage decompensation. Sex chromosomally abnormal models exhibit a "sex-reversed" X dosage state: XX males, like XX females, develop X dosage excess, while XO females, like XY males, develop X dosage decompensation. Thus, germline X dosage compensation states are determined by X chromosome number, not phenotypic sex. These unexpected differences in X dosage compensation states between germline and soma offer unique perspectives on sex chromosome infertility.

Keywords: X inactivation; X upregulation; dosage compensation; genome-wide reprogramming; germline development; sex chromosomes.

Figure 1

Transcriptome Profiling of Mouse Germ Cells (A) Unsupervised hierarchical clustering of all dataset samples. The dendrogram shown is based on Jensen-Shannon distances between conditions. Dark orange, epiblast and early germ cells; dark blue, late germ cells; green, somatic cells; gray, E12.5 XX male germ cells. (B) MDS plot of gene expression in all replicates within the dataset. (C) Time course of relative expression (FPKM) for pluripotency genes in XY males and XX females. (D) Time course of relative expression (FPKM) for meiosis genes in XY males and XX females.

Figure 2

Analysis of X:A Ratios in Somatic Cells (A) Bootstrapped X:A ratios from E14.5 liver and tail and gonadal somatic cell populations (E9.5–P11; E9.5 refers to caudal embryo somatic cells) using different lower FPKM thresholds, and focusing on all genes (i.e., “non-ubiquitous genes”). (B) Definition of a ubiquitous gene set by addition of sequential samples of germ cells and somatic cells. The number of genes expressed in “all samples” is predictably high when the number of samples included is low and decreases as more samples are added before plateauing at a stable gene set. (C) Gene ontology (GO) enrichment analysis, defined using Database for Annotation, Visualization and Integrated Discovery (DAVID), of ubiquitously expressed genes (5,656 on the autosomes and 155 on the X chromosome). Benjamini-Hochberg-corrected p values were plotted against the top ten “biological processes” GO terms. (D) X:A ratios for the same samples as in (A), but considering only ubiquitous genes. (E) Density plot of log₂ FPKM ratios of expression in E14.5 XY male gonadal somatic cells and E14.5 XY male liver for the X chromosome versus the autosomes. (F) Density plot of log₂ FPKM ratios of expression in E14.5 XY male gonadal somatic cells and E14.5 XY male liver for the X chromosome versus the autosomes, after imposition of an upper FPKM expression threshold. (G) X:A ratios for the same samples as in (A) and (D) considering only ubiquitous genes together with an upper FPKM threshold.

Figure 3

X:A Ratios in XX and XO Female Mouse Germ Cells (A) X:A ratios of ubiquitously expressed genes in XX females using an FPKM ≥1 and an upper expression threshold. (B) X:A ratios of ubiquitously expressed genes in XX females using an FPKM ≥1 and no upper expression threshold. (C) X:A ratios of ubiquitously expressed genes in XO females using an FPKM ≥1 and an upper expression threshold. (D) X:A ratios of ubiquitously expressed genes in XO females using an FPKM ≥1 and no upper expression threshold. Dark orange, epiblast and early germ cells; dark blue, late germ cells.

Figure 4

X:A Ratios in XY and XX Sry Male Mouse Germ Cells (A) X:A ratios of ubiquitously expressed genes in XY males using an FPKM ≥1 and an upper expression threshold. (B) X:A ratios of ubiquitously expressed genes in XY males using an FPKM ≥1 and no upper expression threshold. (C) X:A ratios of ubiquitously expressed genes in XX Sry males using an FPKM ≥1 and an upper expression threshold. (D) X:A ratios of ubiquitously expressed genes in XX Sry males using an FPKM ≥1 and no upper expression threshold. Dark orange, epiblast and early germ cells; dark blue, late germ cells.

Figure 5

X:A Ratios in Human XY ESCs, Gonadal Somatic Cells, and Germ Cells (A) X:A ratios of ubiquitously expressed genes in XY ESCs, and week 7 male gonadal somatic cells and female gonadal somatic cells using an FPKM ≥1 and the presence or absence of an upper FPKM threshold. (B) X:A ratios of ubiquitously expressed genes in XX female germ cells using an FPKM ≥1 and an upper expression threshold. Data from Tang et al. (2015) are shown with closed circles; data from Gkountela et al. (2015) are shown with open circles. (C) X:A ratios of ubiquitously expressed genes in XX female germ cells using an FPKM ≥1 and no upper expression threshold. (D) X:A ratios of ubiquitously expressed genes in XY male germ cells using an FPKM ≥1 and an upper expression threshold. (E) X:A ratios of ubiquitously expressed genes in XY male germ cells using an FPKM ≥1 and no upper expression threshold. Dark orange, early germ cells; dark blue, late germ cells.

Figure 6

A Model for X Dosage Compensation Patterns A model for X dosage compensation patterns in the XX Female (A), XO female (B), XY male (C), and XX male (D) mammalian germline. Prior to reprogramming, XX female and XX male germ cells have one upregulated X chromosome (Xa, orange) and one inactive X chromosome (Xi, gray), while XY males and XO female cells have a single upregulated X chromosome (Xa, orange). In all four cases, the dosage of X genes is balanced with that of the autosomes (A, blue), i.e., the X:A ratio is 1. During reprogramming, upregulation of the active X chromosome is maintained in all four genotypes, but in XX females and XX males, the inactive X chromosome begins to reactivate (gray becomes blue). The outcome is X:A ratios greater than 1 in XX females and in XX males. Later, X upregulation is lost, reinstating X:A ratios in XX females to 1, but leaving XY males and XO females X dosage decompensated, with X:A ratios less than 1. XX males show germ cell loss from E15.5, which may be due to high X:A ratios seen.