Knockout of the murine ortholog to the human 9p21 coronary artery disease locus leads to SMC proliferation, vascular calcification, and advanced atherosclerosis

Genome-wide association studies (GWAS) have unequivocally identified genetic variation at the chromosome 9p21 locus as the most important commonly inherited risk factor for coronary artery disease (CAD). This locus potentiates risk independently of all known traditional risk factors, highlighting its potential as a ‘precision’ translational target. However, despite more than a decade of study, the mechanism by which the 9p21 locus promotes disease remains enigmatic because: 1. The risk haplotype contains no protein coding genes; 2. Debate persists about whether 9p21 regulates plaque burden or plaque vulnerability; and 3. No atherosclerotic mouse model exists which recapitulates the human risk allele^1,2.

Herein, we performed detailed histopathological analyses of 180 human autopsy specimens and confirmed that carriers of the 9p21 risk allele (‘C’ at rs1333049) had significantly more plaque burden, but not more features of plaque vulnerability, according to the Virmani classification status³. Risk allele status was not associated with lesional macrophage burden, cap thickness, or necrotic core size (not shown), but was strongly correlated with lesional calcification, even after controlling for total lesion size (Fig A).

Figure.

A. Human carriers of the 9p21 risk allele develop more calcified, but not more vulnerable atherosclerotic plaques. Numbers in the graph indicate the number of coronary specimens analyzed. B. Knockout of the 9p21 locus leads to a significant increase in atherosclerotic burden after 13 or 20 weeks of high fat diet. C. Knockout of the 9p21 locus does not increase intraplaque hemorrhage (IH) in the tandem-stenosis plaque rupture model (IH positive example at right). D. Von Kossa staining confirms that knockout of the 9p21 locus promotes lesional calcification. E. MTT assays reveal that primary-cultured SMCs from chr4^{Δ70kb/Δ70kb} mice are hyperproliferative. F. Knockout of the 9p21 locus induces a calcification phenotype in vivo and vitro, indicated by upregulation of Runx2 and enhanced Alizarin red staining after exposure to 7 days of high-phosphate (left) or osteogenic (right, β-glycerophosphate, ascorbic acid, and dexamethasone) medium. G. Palbociclib decreases osteogenic genes in human (left) and mouse (right) SMCs, with disproportionate benefit in cells from chr4^{Δ70kb/Δ70kb} mice. In vitro experiments were repeated at least 3 times. Data are expressed as mean±SEM. *P< 0.05, †P< 0.01, ‡P< 0.005, §P< 0.001 via one-way ANOVA or Kruskal-Wallis test, or two-tailed Student’s t or Mann-Whitney U test.

To extend these studies into an experimental model of vascular disease, we generated an atheroprone mouse deficient in the murine ortholog of the 9p21 risk locus (chr4^{Δ70kb /Δ70kb}, ApoE^−/− on 129 background)⁴. In keeping with widely-replicated human epidemiologic data, these animals demonstrated no changes in blood pressure, lipid levels, body weight or fasting glucose (not shown). Conversely, both male and female mice deficient in the syntenic region developed larger atherosclerotic plaques when exposed to high fat Western diet, and demonstrated a clear gene dosage effect (Fig B). Consistent with our human data, these mice did not demonstrate any susceptibility towards lesion vulnerability when subjected to a surgical plaque rupture model (Fig C), and similarly had no increase in macrophage content or necrotic core burden (data not shown). On the other hand, these mice did demonstrate increased rates of vascular calcification, reflected by significantly higher lesional Von Kossa content after 20 weeks of high fat diet (Fig D).

To explore the mechanism by which this locus might increase plaque size and calcification, we analyzed the behavior of primary cultured aortic smooth muscle cells (SMCs) from mice deficient in the 9p21 ortholog. Ethics approvals were provided by Stanford University (protocol 27279). As previously reported⁴, these cells had a significant growth advantage (Fig E). Moreover, we found these SMCs to be prone towards de-differentiation, exemplified by their propensity to: 1. Downregulate their ‘SMC-specific’ contractile machinery (e.g. Tagln); 2. Upregulate genes known to accelerate calcification (including the master regulator, Runx2); and 3. Undergo enhanced ossification when exposed to osteogenic medium (quantified by Alizarin Red staining, Fig F). Because carriers of the 9p21 risk allele (and mice deficient in the 9p21 syntenic locus) are known to have pathologically reduced expression of cyclin dependent kinase (CDK) inhibitors, we treated primary SMCs with Palbociclib (a specific inhibitor of CDK4/6), and found that this agent prevented the upregulation of calcification markers, especially in chr4^{Δ70kb /Δ70kb} SMCs (Fig G).

Taken together, this brief report provides several new findings related to the top genetic CAD locus. First, this study uses detailed pathologic analyses to reveal that carriers of the 9p21 risk allele have larger but not more vulnerable lesions, potentially explaining why these individuals do not have increased rates of myocardial infarction after correcting for their increased plaque burden⁵. Second, we confirm that atheroprone mice deficient in the 9p21 ortholog do develop disease reflective of the human condition, and that these animals likely can be coupled with deep-sequencing and gene-editing efforts in human SMCs to further model the biology underlying this key genetic risk locus. Finally, we provide evidence that 9p21 may promote lesion expansion by inducing the proliferation of de-differentiated and osteogenic SMCs, and highlight this pathway as a potential pharmacogenomic target for those known to harbor the risk allele (~50% of non-African populations).