Computational and functional strategies to decipher lncRNAs in human atherosclerosis

点击翻译按钮获取中文摘要

This proposal addresses knowledge gaps in cell-specific function and disease causation of long non-coding RNAs (lncRNAs) in human atherosclerosis. Despite prominent examples of functional lncRNAs in cardiovascular diseases (CVD), their lack of conservation and cell-specificity have limited our understanding of their role in CVD. These challenges are particularly problematic in human atherosclerosis which is characterized by complex multi- cellular lesions. Further, recent single cell (sc)RNAseq data including our preliminary studies suggest that, relative to mRNAs, lncRNA expression in primary human cells may be restricted to key cell subpopulations. An overarching hypothesis is that many human lncRNAs modulate atherosclerosis and CVD risk via their discrete expression and function in specific lesion cell subpopulations. Thus, more precise knowledge of lncRNA cell-specific relationship to human atherosclerosis is required to drive mechanism-based clinical translation. Here we address key questions for lncRNAs in human atherosclerosis and CVD risk. First, which lncRNAs are expressed in human lesions and associate with clinical CVD? Second, for lncRNAs expressed in human lesions, in which specific lesion cell subpopulation are they functional? In Aim 1, we will address the first issue by analyzing differential expression of lncRNAs through deep RNAseq of a large nested case-control (n=260 with “symptomatic/unstable” and n=260 with “asymptomatic/stable” plaques) study of carotid atherosclerosis from the Munich Vascular Biobank (MVB). We will also determine whether lncRNAs demonstrate differential allele specific expression (ASE) between symptomatic/unstable vs. asymptomatic/stable plaques and if cis-eQTL variants for lncRNAs with differential ASE are associated with coronary heart disease (CHD) in large public genetic datasets. Prioritized lncRNAs will undergo cell-specific functional genomic follow-up in human vascular cells including our human induced pluripotent stem cell (hIPSC) vascular models. In Aim 2, we propose to use a novel deconvolution algorithm and integration of large-scale bulk RNAseq data from Aim 1 with selective single cell (sc)RNAseq of fresh lesions (n=60) to identify subpopulations and their lncRNAs that associate with symptomatic/unstable plaques and have causal genetic relationships to CHD. ScRNAseq of the fresh carotid lesions will be used to cluster cells and identify lesion subpopulations. Result from this analysis will permit computational deconvolution of the cell subpopulation composition of all MVB bulk RNAseq lesions (n=520) and assignment of subpopulation-specific lncRNA expression and relationship to symptomatic/unstable plaques. Subpopulation-specific lncRNA cis-eQTLs also will be identified and used to determine their causal relationship to CHD in genetic datasets. These findings, coupled to subpopulation-specific functional studies, will define subpopulation-specific lncRNA functions in human atherosclerosis. Our proposal leverages unique genomic data, innovative computational methods and functional genomics, and interdisciplinary expertise to direct in vivo translation and precision therapeutic targeting of cell-specific vascular lncRNA functions in atherosclerotic CVD.