Abstract
Genome-wide association studies (GWAS) have validated a strong association of atherosclerosis with the CDKN2A/B locus, a locus harboring three tumor suppressor genes: p14ARF , p15INK4b , and p16INK4a . Post-GWAS functional analysis reveals that CUX is a transcriptional activator of p16INK4a via its specific binding to a functional SNP (fSNP) rs1537371 on the atherosclerosis-associated CDKN2A/B locus, regulating endothelial senescence. In this work, we characterize SATB2, another transcription factor that specifically binds to rs1537371. We demonstrate that even though both CUX1 and SATB2 are the homeodomain transcription factors, unlike CUX1, SATB2 is a transcriptional suppressor of p16INK4a and overexpression of SATB2 competes with CUX1 for its binding to rs1537371, which inhibits p16INK4a and p16INK4a -dependent cellular senescence in human endothelial cells (ECs). Surprisingly, we discovered that SATB2 expression is transcriptionally repressed by CUX1. Therefore, upregulation of CUX1 inhibits SATB2 expression, which enhances the binding of CUX1 to rs1537371 and subsequently fine-tunes p16INK4a expression. Remarkably, we also demonstrate that IL-1β, a senescence-associated secretory phenotype (SASP) gene itself and a biomarker for atherosclerosis, induces cellular senescence also by upregulating CUX1 and/or downregulating SATB2 in human ECs. A model is proposed to reconcile our findings showing how both primary and secondary senescence are activated via the atherosclerosis-associated p16INK4a expression.