Mosaic loss of Y chromosome and mortality after coronary angiography

Abstract

Background and aims: Acquired somatic mutations emerged as important drivers of adverse cardiovascular disease outcomes. Recently, mosaic loss of Y chromosome (LOY) in haematopoietic cells was identified to induce diffuse cardiac fibrosis in male mice. The aim of the present study was to determine the association between LOY and cardiovascular mortality in patients undergoing coronary angiography.

Methods: LOY was quantified in 1698 male participants of the LURIC study, who underwent coronary angiography, and its association with all-cause and cardiovascular mortality was determined. Furthermore, the interaction between LOY and inherited genetic susceptibility for cardiac fibrosis was assessed.

Results: The frequency of LOY steeply increased in male participants of LURIC at the age of 60 years. Loss of Y chromosome > 17% was associated with significantly higher all-cause [hazard ratio (HR) 1.41, 95% confidence interval (CI) 1.09-1.82] and cardiovascular mortality (HR 1.49, 95% CI 1.09-2.03), which was driven by a higher risk for fatal myocardial infarction (HR 2.65, 95% CI 1.46-4.81). Loss of Y chromosome > 17% was associated with a profibrotic and proinflammatory plasma protein expression profile as characterized by higher plasma levels of osteoprotegerin, matrix metalloproteinase-12, growth differentiation factor 15, heparin-binding EGF-like growth factor, and resistin. Genetic predisposition for lower myocardial fibrosis attenuated the association between LOY and cardiovascular mortality. Genome-wide methylation analyses identified differential methylation in 298 genes including ACTB, RPS5, WDR1, CD151, and ARAP1. Single-cell RNA sequencing further confirmed differential gene expression of 37 of these genes in LOY in peripheral blood mononuclear cells comprising a set of fibrosis-regulating genes including RPS5. RPS5 silencing in macrophages induced a paracrine induction of collagen expression in cardiac fibroblasts documenting a functional role in vitro.

Conclusions: LOY represents an important independent risk factor for cardiovascular mortality in male patients with coronary artery disease. Targeting LOY may represent a sex-specific personalized medicine approach.

Keywords: Coronary artery disease; Fibrosis; Loss of Y chromosome.

Structured graphical abstract

Frequency of loss of Y chromosome (LOY) in leukocytes increases with age, associates with a profibrotic protein signature, interstitial myocardial fibrosis, and leads to higher cardiovascular mortality.

Figure 1

Association between loss of Y chromosome and baseline characteristics. (A) Restricted cubic spline plot of the association between age and loss of Y chromosome. The line indicates loss of Y chromosome (%) with the 95% confidence interval (grey-shaded area). Histogram shows the age distribution (right Y axis). (B) Multivariate adjusted least square means and the corresponding 95% confidence interval of loss of Y chromosome as determined by using linear regression models according to age (divided at the median), smoking status, diabetes, and myocardial infarction (all analyses are shown in Supplementary data online, Table S1). Analyses are adjusted for age, testosterone, triglycerides, LDL-C, glycated haemoglobin, Friesinger score, troponin T, diabetes, hypertension, coronary artery disease, smoking, myocardial infarction, lipid-lowering therapy, smoking, and high-sensitivity C-reactive protein

Figure 2

Association between loss of Y chromosome and mortality. (A) Unadjusted all-cause mortality rates in females, males with loss of Y chromosome ≤ 17%, and males with loss of Y chromosome > 17% during follow-up. (B) Plot of the survival function according to loss of Y chromosome. Analysis is adjusted for age, age, smoking status, body mass index, LDL cholesterol, high-sensitivity C-reactive protein, troponin T, diabetes, hypertension, myocardial infarction, and coronary artery disease. The line indicates mortality with the 95% confidence interval (grey-shaded area). Histogram shows the distribution of loss of Y chromosome (right Y axis). (C) Survival curve showing the association between loss of Y chromosome and all-cause mortality as determined by Cox regression analyses adjusted for age, age, smoking status, body mass index, LDL cholesterol, high-sensitivity C-reactive protein, troponin T, diabetes, hypertension, myocardial infarction, and coronary artery disease. For the plot, covariates were set to their mean value. HR, hazard ratio

Figure 3

Association between loss of Y chromosome and specific causes of death. Cumulative incidence curves showing the association between loss of Y chromosome and (A) cardiovascular mortality and (B) death due to fatal myocardial infarction as determined by competing risk regression analyses adjusted for age, age, smoking status, body mass index, LDL cholesterol, high-sensitivity C-reactive protein, troponin T, diabetes, hypertension, myocardial infarction, and coronary artery disease. SHR, subhazard ratio

Figure 4

Association between loss of Y chromosome and plasma protein profile. (A) Volcano plot showing the effect size (B) as well as the significance level of the normalized protein levels of 80 plasma proteins quantified using the OLINK platform in patients with loss of Y chromosome > 17% as compared with those with loss of Y chromosome ≤ 17%. Analyses were adjusted for age, smoking status, body mass index, LDL cholesterol, high-sensitivity C-reactive protein, troponin T, diabetes, hypertension, myocardial infarction, coronary artery disease, and number of NPX values below detection across all proteins. To account for multiple testing, P-values were calculated at FDR of .05 (P-value*rank/number of comparisons, i.e. N = 80). (B) Heat map showing protein plasma levels (OLINK normalized protein expression levels transformed as Z-scores) according to loss of Y chromosome ≤ 17% and >17%

Figure 5

Modulation of the effects of loss of Y chromosome on mortality by myocardial fibrosis. (A) Survival curve showing the association between loss of Y chromosome and all-cause mortality and (B) cumulative incidence curve showing the association between loss of Y chromosome and cardiovascular mortality according to the weighted genetic risk score for myocardial fibrosis as determined by Cox regression analysis and competing risk regression analysis adjusted for age, age, smoking status, body mass index, LDL cholesterol, high-sensitivity C-reactive protein, troponin T, diabetes, hypertension, myocardial infarction, and coronary artery disease. SHR, subhazard ratio

Figure 6

Loss of Y chromosome alters methylation and fibrosis-regulating gene expression. (A) Circos plot. Green band corresponds to -log₁₀(P) for association of loss of Y chromosome with DNA methylation with Y axis truncated at 30 by chromosomal position. Red band and red gene labels correspond to -log₁₀(P) for higher gene expression and blue band and blue gene labels to -log₁₀(P) for lower gene expression in loss of Y chromosome peripheral blood mononuclear cells as compared with Y-positive peripheral blood mononuclear cells of the 298 differentially methylated genes as determined by single-cell RNA sequencing

Figure 7

Knockdown of the loss of Y chromosome target gene RPS5 induces a profibrotic response. (A) Experimental outline. (B) Gene expression in THP1-derived macrophages after small interfering RNA-mediated knockdown of the loss of Y chromosome target genes EEF1D, ZNF138, and RPS5 to determine knockdown efficiency. (C) Quantification of the COL1A1-positive area normalized to DAPI-positive area in cardiac fibroblasts incubated for 72 h with supernatant of THP1-derived macrophages after knockdown of EEF1D, ZNF138, or RPS5, respectively, or transforming growth factor-β (50 ng/mL) as positive control. (D) Representative confocal images of cardiac fibroblasts. SN, supernatant