Abstract 17754: Characterizing the Human Coronary Aorta Vascular Endothelial Cell Transcriptome With RNA Sequencing
Introduction: Transcriptome profiling of cardiovascular disease (CVD) relevant tissues/cells may provide molecular functional insights into the results from CVD genetic and genomic studies. In addition to characterization the role of protein-coding gene transcripts in the pathophysiology of CVD, novel regulators may be identified including long intergenic non-coding RNAs (lincRNAs). To this end, we performed a RNA-seq experiment to characterize mRNAs and lincRNAs in human aorta vascular endothelial cells.
Methods: We conducted a paired-end deep RNA sequencing on eight human primary endothelial cell samples with Illumina HiSeq 2000. We aligned reads with TopHat to the Human Reference Genome (hg19), and used Cufflinks to assemble transcripts and estimate abundances. Results were compared with Illumina’s Human BodyMap 2.0 RNASeq data to discover novel and endothelial-specific genes and lincRNAs.
Results: We generated 2.14 billion reads of 50bp with an average of 134 million paired-end reads per sample. Among them, 97% of reads were mapped uniquely to the genome. For a total of 25,593 known genes, 75% were expressed at a very low level (FPKM < 8). Although sequencing was performed at a relatively high coverage, 26.8% of genes were undetectable in these endothelial cells, including the liver-specific gene, PCSK9. However, more than 100 genes were highly expressed (FPKM > 1000), including several genes associated with CVD or CVD risk factors, e.g., CDKN2A, COL4A1/A2, LDLR, SORT1, STX2, and VWF. For 8,267 known lincRNAs, >75% of them were undetectable, and only 13 lincRNAs were moderately expressed (FPKM > 20), including CDKN2B-AS1, a functional candidate at the well-known 9p21 CVD risk locus. When compared with Human BodyMap data, several lincRNAs were novel or expressed specifically in these cells, e.g., linc-ZDHHC17-1 on chr12q21.
Conclusion: By pooling extremely deep sequenced data from 8 endothelial cell samples, we generate a landscape of transcription in these cells. Many CVD related genes are highly expressed or expressed specifically in these cells. We find several lincRNAs that appear to be expressed specifically in these cells, identifying a subset of lincRNAs that may play important functional roles in human aorta vascular endothelial cells.
- © 2013 by American Heart Association, Inc.