Premature ovarian insufficiency is associated with global alterations in the regulatory landscape and gene expression in balanced X-autosome translocations

Abstract

Background: Patients with balanced X-autosome translocations and premature ovarian insufficiency (POI) constitute an interesting paradigm to study the effect of chromosome repositioning. Their breakpoints are clustered within cytobands Xq13-Xq21, 80% of them in Xq21, and usually, no gene disruption can be associated with POI phenotype. As deletions within Xq21 do not cause POI, and since different breakpoints and translocations with different autosomes lead to this same gonadal phenotype, a "position effect" is hypothesized as a possible mechanism underlying POI pathogenesis.

Objective and methods: To study the effect of the balanced X-autosome translocations that result in POI, we fine-mapped the breakpoints in six patients with POI and balanced X-autosome translocations and addressed gene expression and chromatin accessibility changes in four of them.

Results: We observed differential expression in 85 coding genes, associated with protein regulation, multicellular regulation, integrin signaling, and immune response pathways, and 120 differential peaks for the three interrogated histone marks, most of which were mapped in high-activity chromatin state regions. The integrative analysis between transcriptome and chromatin data pointed to 12 peaks mapped less than 2 Mb from 11 differentially expressed genes in genomic regions not related to the patients' chromosomal rearrangement, suggesting that translocations have broad effects on the chromatin structure.

Conclusion: Since a wide impact on gene regulation was observed in patients, our results observed in this study support the hypothesis of position effect as a pathogenic mechanism for premature ovarian insufficiency associated with X-autosome translocations. This work emphasizes the relevance of chromatin changes in structural variation, since it advances our knowledge of the impact of perturbations in the regulatory landscape within interphase nuclei, resulting in the position effect pathogenicity.

Keywords: Chromatin structure; Position effect; RNA sequencing; X-autosome translocation.

Fig. 1

Study workflow. a Patients’ selection criteria. b Blood collection for DNA extraction and cell culture. c Breakpoint mapping by whole genome sequencing searching for chimeric reads and inserts. d Prediction of TAD disruption and screening of position effect candidate genes. e LCL establishment for chromatin crosslink and RNA extraction. f Transcriptome profiling using RNA-seq to leverage differentially expressed genes (DEGs). g Histone modification screening by ChIP-seq for epigenetic landscape assessment. BP breakpoint

Fig. 2

Prediction of TADs in the POI2 region in the X chromosome, position of patients’ breakpoints, POI candidate genes, and gene expression levels in the POI2 region. Overview of cell type-invariant TADs and chromatin states in POI2 region and relative position of patients’ breakpoints (shown by patients’ number). Below, the candidate genes for ovarian function (black bars) harbored by disrupted TADs (FOXO4, POF1B, and DIAPH2) or neighboring TAD (FGF16). The arrow direction (up or down) indicates the direction of effect from the gene expression comparison between patients and controls (upregulated or downregulated)

Fig. 3

IGV visualization of differential peaks at 17p12 and expression levels of DEGs within this region. IGV tracks: encode GM12878 H3K27ac in orange, and RefSeq genes. a H3K27ac peak in patients (blue) and controls (pink) at the ARHGAP44 gene body. Note that H3K27ac is decreased in patients. b Two H3K27ac peaks in patients (blue) and controls (pink), one at the promoter region and one at the HS3ST3B1 gene body. Note that both peaks are decreased in patients. c Significant expression difference between patients and controls in FPKM levels of ARHGAP44 and HS3ST3B1 genes, respectively