Coronary Microvascular Dysfunction Is Associated with a Pro-Inflammatory Circulating Transcriptome in Patients with Nonobstructive Coronary Arteries

Graphical Abstract

Background:

Coronary microvascular dysfunction (CMD) is a mediator of ischemia with non-obstructive coronary arteries (INOCA), but pathogenic mechanisms remain unknown.¹ Heightened inflammation has been reported in small cohorts of patients with CMD.^2-4 Whole blood gene expression profiling is an unbiased method to investigate molecular and cellular pathways associated with disease.⁵ We analyzed whole blood expression profiles in patients with and without CMD to provide insights into INOCA mechanisms.

Methods:

De-identified data that support the study findings are available from the corresponding author upon reasonable request. We prospectively enrolled women undergoing clinically indicated invasive coronary angiography with ischemic symptoms, myocardial ischemia by stress testing, or both, into a single center diagnostic, cross-sectional study of CMD (NCT03537586). The study was approved by the NYU Grossman School of Medicine Institutional Review Board, and all subjects gave informed consent before participation. All patients with non-obstructive coronary arteries, defined as <50% stenosis in all major epicardial coronary arteries, were included. Patients with reduced left ventricular ejection fraction <40%, clinical heart failure, or with known hypertrophic cardiomyopathy were excluded. Coronary physiology was evaluated at the time of index invasive coronary angiography using coronary thermodilution techniques to measure coronary flow in the left anterior descending coronary artery before and after the administration of adenosine, an endothelium-independent coronary vasodilator. Coronary flow reserve (CFR) was calculated as the ratio of hyperemic to basal coronary flow measured by thermodilution. CMD was defined as a CFR <2.5.¹

Peripheral whole blood was collected into PAXgene Blood RNA tubes (Qiagen) immediately prior to invasive coronary angiography. RNA was isolated and sequenced with Illumina HiSeq4000. Quality control was assessed with FastQC and MultiQC software. Reads were mapped to human reference genome (hg38) using STAR, duplicates removed, and read count tables generated yielding library sizes between 0.5 and 12 million reads (mean 5.4 million). Normalization and age adjusted differential expression analyses were performed with R package DESeq2 to identify differentially expressed transcripts in patients with and without CMD, and by CFR as a continuous function. Canonical pathways were obtained with Ingenuity Pathway Analysis (IPA). Nominal P-values <0.05 were used to identify relevant genes and pathways.

Results:

Among 28 women enrolled with INOCA, 7 had CMD (mean 58.3 ± 9.6 years, 71% white race, 43% Hispanic/Latino) and 21 did not (mean age 60.5 ± 8.1 years, 86% white race, 29% Hispanic/Latino). 1,430 transcripts were differentially expressed when patients were stratified by CMD status (P<0.05) (Panels A and B). Unsupervised hierarchical clustering of the whole blood transcriptome demonstrated perfect separation between women with and without CMD. In IPA stratified by CMD status as a dichotomous variable, IL-4 signaling (Z-score 2.7), Th2 pathways (2.6), T Cell receptor signaling (2.4), regulation of IL-2 expression in T-lymphocytes (2.3), Oxidative phosphorylation (2.2), macrophage alternative activation signaling pathways (1.4), NF-κB Signaling (1.3), and interferon signaling (1.1), were upregulated in patients with CMD. While the inflammatory transcriptional signature was significantly different in women with versus without CMD, there were no differences in blood counts of major cell types, including leukocyte subtypes, red blood cells and platelets between groups.

When CFR (median 3.2 [IQR 2.3-4.2]) was evaluated as a continuous variable, IPA revealed lower CFR values were significantly associated with upregulated inflammatory pathways, including S100 signaling pathways (Z-score 5.2), cytokine storm signaling pathways (4.6), neutrophil extracellular trap signaling (4.3), oxidative phosphorylation (4.3), macrophage activation (4.1), interferon (3.6), IL-8 (2.6), IL-15 (2.5) and IL-6 (2.0) signaling, Fcγ receptor-mediated phagocytosis (3.3), pathways related to reactive oxygen species production (2.6), and inflammasome signaling (2.2) (Panel C).

Discussion:

In a cohort of women with INOCA, individuals with CMD defined by a low CFR had a pro-inflammatory whole blood transcriptional signature, despite no observed differences in the circulating cell counts at the time of RNA collection. These findings support the hypothesis that inflammation may underlie CMD and contribute to ischemic symptoms in INOCA. In a series of 1471 women with INOCA who underwent non-invasive transthoracic doppler imaging for CMD and large-scale proteomic analyses, CMD was strongly associated with pathways involving inflammation.⁴ In a Cardiovascular Inflammation Reduction Trial sub-study, lower CFR correlated with increased IL-6 and a trend toward higher hsCRP.³ The current transcriptomic analyses extend the findings of previous biomarker-based analyses.

Although this is the first reported evaluation of whole blood RNA sequencing in patients with INOCA, there are some important limitations. First, the number of participants was small, and coronary spasm testing was not systematically performed. Second, only transcriptional profiles of circulating cells were captured in this analysis, and endothelial and smooth muscle transcriptional profiles were not assessed. Single-cell RNA sequencing was not performed, limiting cell-specific insights. Third, we did not include men in this cohort, precluding evaluation of sex-specific differences in transcriptional profiles. Still, our analyses demonstrate the potential of circulating RNA to provide critical insights into the molecular mechanisms associated with CMD in patients with INOCA.

Conclusion:

Among women with INOCA undergoing invasive testing, pro-inflammatory signaling pathways were upregulated with the prevalence and severity of CMD. Additional investigation is warranted to explore inflammation in CMD to identify noninvasive diagnostic and potential therapeutic targets.

Figure: Differentially expressed genes and canonical pathways in patients with CMD.

Panel A: Heatmap indicating transcript intensity (row Z-score of normalized counts) for top 282 differentially expressed transcripts (p<0.01) between patients with (n=7) and without (n=21) CMD. Panel B: Volcano plot of differentially expressed genes (p<0.05) between patients with and without CMD. Panel C: Selected pathways upregulated in patients with lower CFR and corresponding Z-scores by Ingenuity Pathway Analysis (p<0.05).