Differential Proteomics of Cardiovascular Risk and Coronary Artery Disease in Humans

BACKGROUND: Proteomics of atypical phenotypes may help unravel cardiovascular disease mechanisms. AIM: We aimed to prospectively screen the proteome of four types of individuals: with or without coronary artery disease (CAD), each with or without multiple risk factors. Associations with individual risk factors and circulating biomarkers were also tested to provide a functional context to the protein hits. MATERIALS AND METHODS: The CAPIRE study (ClinicalTrials.gov Identifier: NCT02157662) is a cross-sectional study aimed at identifying possible new mechanisms promoting or protecting against atherothrombosis. Quantification (by aptamer technology), ranking (using partial least squares), and correlations (by multivariate regression) of ~5000 plasma proteins were performed in consecutive individuals aged 45–75 years, without previous cardiovascular disease, undergoing computed tomography angiography for suspected CAD, showing either >5/16 atherosclerotic segments (CAD(+)) or completely clean arteries (CAD(−)) and either ≤ 1 risk factor (RF(+)) or ≥3 risk factors (RF(−)) (based on history, blood pressure, glycemia, lipids, and smoking). RESULTS: Of 544 individuals, 39% were atypical (93 CAD(+)/RF(−); 120 CAD(−)/RF(+)) and 61% typical (102 CAD(+)/RF(+); 229 CAD(−)/RF(−)). In the comparison with CAD(+)/RF(−) adjusted for sex and age, CAD(−)/RF(+) was associated with increased atrial myosin regulatory light chain 2 (MYO) and C-C motif chemokine-22 (C-C-22), and reduced protein shisa-3 homolog (PS-3) and platelet-activating factor acetylhydrolase (PAF-AH). Extending the analysis to the entire cohort, an additional 8 proteins were independently associated with CAD or RF; by logistic regression, the 12-protein panel alone discriminated the four groups with AUC(ROC)'s of 0.72–0.81 (overall p = 1.0e(−38)). Among them, insulin-like growth factor binding protein-3 is positively associated with RF, lower BMI, and HDL-cholesterol, renin with CAD higher glycated hemoglobin HbA(1c), and smoking. CONCLUSIONS: In a CCTA-based cohort, four proteins, involved in opposing vascular processes (healing vs. adverse remodeling), are specifically associated with low CAD burden in high CV-risk individuals (high MYO and C-C-22) and high CAD burden in low-risk subjects (high PS-3 and PAF-AH), in interaction with BMI, smoking, diabetes, HDL-cholesterol, and HbA(1c). These findings could contribute to a deeper understanding of the atherosclerotic process beyond traditional risk profile assessment and potentially constitute new treatment targets.