Despite the recent report linking
PHACTR1-rs9349379 with
EDN1 in ECs [
12],
PHACTR1 itself remains a strong causal candidate gene at this CAD-associated locus because of the eQTL effect in hCA [
5,
6]. Further, it was recently shown in an in vitro cell culture system that PHACTR1 contributes to VSMC calcification, a hallmark of CAD [
18]. Using two different methods, we confirmed that the human
PHACTR1 gene encodes six main transcripts that are differentially expressed. Although our results are largely consistent with a study by Reschen et al. [
13], we note three important differences: First, we confirmed that there are six, and not three,
PHACTR1 transcripts. Indeed, whereas Reschen and colleagues reported only one intermediate transcript (transcript B+ in Fig.
1), we found that four intermediate transcripts (with or without exons 7.8 and 10.11) are co-expressed in the heart (Fig.
2). Whether these four PHACTR1 isoforms, of 650, 581, 557 and 488 amino acids, have the same functions is currently unknown. Second, we confirmed using our own custom polyclonal antibody against PHACTR1 that the protein is present in VSMCs. In contrast, we have not been able to detect by immunoblotting the protein encoded by the short transcript. And third, our eQTL analyses in 158 hCA validated an effect of genotypes at rs9349379 on the expression of intermediate transcripts A+ and B+, but not the short immune-specific transcript. One possible interpretation of this result is that rs9349379 mediates in part its effect on CAD risk by regulating the expression or splicing of these
PHACTR1 intermediate transcripts in ECs and/or VSMCs within hCA, and not through the short immune-specific
PHACTR1 transcript. In support of this hypothesis, we also know that rs9349379 is associated with fibromuscular dysplasia [
4], a disease without an inflammatory component. Precisely how, at the molecular level, genotypes at rs9349379 modulate cell type-specific transcriptional expression and/or splicing remains to be determined.
Our study presents with some limitations. As for every transcriptomic profiling experiments, weakly expressed transcripts might have missed detection. We tried to account for this by combining two complementary methods and profiling PHACTR1 transcripts in high quality mRNA prepared from many tissues and cell types. The number of available hCA samples for our eQTL analyses was small (N = 36). To increase statistical power, we also analyzed RNA-seq data from the GTEx Project (N = 122). Because short-reads cannot be used unambiguously to reconstruct specific transcripts, we decided to analyze the expression levels of specific exons. This analysis generated results that were highly concordant with our transcript-specific qPCR results. Finally, although we did not detect an eQTL effect for the short PHACTR1 transcript in hCA, it does not rule out a potential role for PHACTR1 in immune cells in the context of atherosclerosis progression and CAD. Indeed, the short PHACTR1 protein might play a critical atherosclerotic function in circulating monocytes, which would not have been captured in hCA samples. However, it is critical to confirm that the short PHACTR1 transcript can be translated into an active protein in human cells.