Distinct dosage compensations of ploidy-sensitive and -insensitive X chromosome genes during development and in diseases

Abstract

The active X chromosome in mammals is upregulated to balance its dosage to autosomes during evolution. However, it is elusive why the known dosage compensation machinery showed uneven and small influence on X genes. Here, based on >20,000 transcriptomes, we identified two X gene groups (ploidy-sensitive [PSX] and ploidy-insensitive [PIX]), showing distinct but evolutionarily conserved dosage compensations (termed XAR). We demonstrated that XAR-PIX was downregulated whereas XAR-PSX upregulated at both RNA and protein levels across cancer types, in contrast with their trends during stem cell differentiation. XAR-PIX, but not XAR-PSX, was lower and correlated with autoantibodies and inflammation in patients of lupus, suggesting that insufficient dosage of PIX genes contribute to lupus pathogenesis. We further identified and experimentally validated two XAR regulators, TP53 and ATRX. Collectively, we provided insights into X dosage compensation in mammals and demonstrated different regulation of PSX and PIX and their pathophysiological roles in human diseases.

Keywords: Biological sciences; Genetics; Molecular physiology.

Graphical abstract

Figure 1

Identification and characterization of PSX and PIX genes on X chromosome (A) Classification X chromosome genes based on absolute values of Pearson correlations (xaxis) compared to the average absolute correlation value (blue vertical line). The illustration of calculation of ploidy-versus-expression correlation was shown on the right. (B and C) Cumulative distribution of expression levels of PSX (B) and PIX (C) genes on X chromosome, compared to autosomal genes, using samples from the TCGA project or the GTEx project. A total of 100 samples were randomly selected and their median expression levels for genes were shown for visualization. Vertical lines indicate medians. (D) PIX and PSX genes distribution on X chromosome. PIX enriched regions were marked by arrows. (E–G) Comparing gene length (E), gene sequence conservation (F), and mRNA half-life (two-sided t-test, G; data are represented as median +/− IQR) between PIX and PSX genes. PIX: ploidy-insensitive X chromosome genes, PSX: ploidy-sensitive X chromosome genes. See also Figure S1.

Figure 2

Differential dosages of PSX and PIX genes across normal tissues and cancer types (A) XAR-PSX and XAR-PIX across 27 normal tissues from GTEx were close to one, except for blood and testis. (B) XAR-PSX and XAR-PIX across 33 cancer types from TCGA. (C) XAR-PSX was higher in cancers than in matched normal tissues. Effect sizes were indicated by Cohen’SD statistics in blue (negligible[N]<0.2, 0.2<=small[S]<0.5, 0.5<=moderate[M]<0.8, large[L]≥0.8). (D) XAR-PIX were lower in cancers than in matched normal tissues. Effect sizes were indicated by Cohen’SD statistics in blue. (C and D) Two-sided Wilcoxon test p-value significance: ns >0.05, ∗ ≤ 0.05, ∗∗ ≤0.01, ∗∗∗ ≤ 0.001, ∗∗∗∗ ≤ 0.0001. (A–D) Data are represented as median +/− IQR. See also Figures S2 and S3. Matching of tumor types to normal tissues were as follows: ACC, Adrenal Gland; BRCA, Breast; COAD, Colon; ESCA, Esophagus; GBM, Brain; HNSC, Salivary Gland; KICH, Kidney; KIRC, Kidney; KIRP, Kidney; LGG, Brain; LIHC, Liver; LUAD, Lung; LUSC, Lung; OV, Ovary; PAAD, Pancreas; PRAD, Prostate; PCPG, Adrenal Gland; READ, Colon; SARC, Adipose Tissue; SKCM, Skin; STAD, Stomach; TGCT, Testis; THCA, Thyroid; UCS, Uterus; UCEC, Uterus. XAR: X-over-autosome dosage ratio, ACC: Adrenocortical carcinoma, BLCA: Bladder urothelial Carcinoma, BRCA: Breast invasive carcinoma, CESC: Cervical squamous cell carcinoma and endocervical adenocarcinoma, CHOL: Cholangiocarcinoma, COAD: Colon adenocarcinoma, CCRCC: Clear cell renal cell carcinoma, DLBC: Lymphoid neoplasm diffuse large B-cell lymphoma, ESCA: Esophageal carcinoma, GBM: Glioblastoma multiforme, HNSC: Head and neck squamous cell carcinoma, KICH: Kidney chromophobe, KIRC: Kidney renal clear cell carcinoma, KIRP: Kidney renal papillary cell carcinoma, LAML: Acute myeloid leukemia, LGG: Brain lower grade glioma, LIHC: Liver hepatocellular carcinoma, LUAD: Lung adenocarcinoma, LUSC: Lung squamous cell carcinoma, MESO: Mesothelioma, OV: Ovarian serous cystadenocarcinoma, PAAD: Pancreatic adenocarcinoma, PCPG: Pheochromocytoma and paraganglioma, PDA: Pancreatic ductal adenocarcinoma, PRAD: Prostate adenocarcinoma, READ: Rectum adenocarcinoma, SARC: Sarcoma, SKCM: Skin Cutaneous melanoma, STAD: Stomach adenocarcinoma, TGCT: Testicular germ cell tumors, THCA: Thyroid carcinoma, THYM: Thymoma, UCEC: Uterine corpus endometrial carcinoma, UCS: Uterine carcinosarcoma, UVM: Uveal melanoma.

Figure 3

XAR-PSX and XAR-PIX were inversely associated with stemness (A) Cancer stemness (RNAss) was positively correlated with XAR-PSX (top) and negatively with XAR-PIX (bottom). Colors indicate-log10 p-values, and sizes of rectangles represent Pearson correlation. Correlations with FDR<0.1 were marked by circles. (B) Comparison of XAR-PIX (top) and XAR-PSX (bottom) between ESCs and iPSCs from females or males. (C) Comparison of XAR-PIX (top) and XAR-PSX (bottom) between sexes from ESCs or iPSCs. (D) XAR-PSX dynamics of embryonic stem cells (grouped by sex) during preimplantation from day 3 (E3) to 7 (E7). (E) Strong negative correlation of XAR-PSX with differentiation culturing time in both human (top) and mouse (bottom). (F) Strong positive correlation of XAR-PIX with differentiation culturing time in both human (top) and mouse (bottom). (B–D) Two-sided Wilcoxon test. Data are represented as median +/− IQR.

Figure 4

XAR in autoimmune disease (A) XAR-PIX (top) and XAR-PSX (bottom) in SLE patients compared to healthy controls. (B) XAR-PIX (top) and XAR-PSX (bottom) in SLE patients with anti-Ro autoantibodies compared to those without anti-Ro. (C) XAR-PIX (top) and XAR-PSX (bottom) in SLE patients with more active interferon response (ISM) compared to those with less ISM. (A–C) Two-sided Wilcoxon test.Data are represented as median +/− IQR.

Figure 5

TP53 and ATRX were two regulators of XAR (A) Significance of XAR-PIX difference (yaxis, two-sided t-test) between wildtype and mutated genes across cancers in male. TP53 and ATRX were marked by arrows. For better visualization, only genes with FDR<1 were shown as points and those with FDR<1e-4 labeled. (B and C) Cumulative distribution of PIX and autosomal gene expression levels for HT1080 (B) and MCF7 (C), before and after TP53-KO. Vertical lines indicate median expression used to calculate XAR-PIX dosages. (D and E) Cumulative distribution of PSX and autosomal gene expression levels for HT1080 (D) and MCF7 (E) before and after TP53-KO. Vertical lines indicate median expression used to calculate XAR-PSX dosages. (F) Gene expression changes of MSL complex subunits KAT8, MSL1, and MSL2 in MCF7 (left) and HT1080 (right) cells, before and after TP53-KO. Data are represented as mean ±SD. (G) ATRX knockout (KO) in mouse induced pluripotent stem cells (iPSCs) significantly reduced both XAR-PIX (middle) and XAR-PSX (right). (H) ATRX knockout (KO) in human GBM cells significantly reduced both XAR-PIX (middle) and XAR-PSX (right). (G and H) Data are represented as mean ± SD (barplot) or median +/− IQR (boxplot). (F–H) Two-sided t-test, ∗P<= 0.05, ∗∗ ≤0.01, ∗∗∗ ≤ 0.001, ∗∗∗∗ ≤ 0.0001. See also Figures S4–S6.

Figure 6

Major findings of this study Differential dosages of two groups of X chromosome genes (PIX and PSX) were observed in human and in animals at both mRNA and protein levels. PIX and PSX dosages were converging during stem cell differentiation but diverging in cancer. PIX dosage was lower in blood cells of SLE autoimmune patients than in healthy individuals. TP53 and ATRX were identified and validated as two XAR regulators. XAR: X-over-autosome dosage ratio, PIX: ploidy-insensitive X genes, PSX: ploidy-sensitive X genes.