Blood plasma proteome-wide association study implicates novel proteins in the pathogenesis of multiple cardiovascular diseases.

Background: Cardiovascular diseases (CVD) are the leading cause of global mortality, yet current treatments benefit only a subset of patients. To identify new potential treatment targets, we conducted the first proteome wide association study (PWAS) for 26 CVDs using plasma proteomics data from the largest cohort to date (53,022 individuals in the UK Biobank Pharma Proteomics Project (UKB-PPP)).

Methods And Results: We calculated single nucleotide polymorphism (SNP)-protein weights using the UKB-PPP dataset and integrated these weights with genome-wide association study (GWAS) summary statistics for 26 CVDs across three categories (16 cardiac, 5 venous, and 5 cerebrovascular diseases) in up to 1,308,460 individuals. PWAS was performed using the Functional Summary-based Imputation (FUSION) framework to identify protein-disease associations. Replication was conducted in two independent human plasma proteomic datasets (comprising 7213 and 3301 participants, respectively). We identified 155 proteins associated with CVDs and further Mendelian randomization analysis revealed 72 proteins with evidence of a causal association. Of these, 26 out of 35 available proteins were validated. Notably, 33 of the 72 proteins were not previously implicated in GWAS of CVDs. For example, PROC was found to be associated with venous thromboembolism (P = 6.32 × 10). We further conducted longitudinal analyses using plasma proteomics data and peripheral blood mononuclear cells single cell RNA-seq data. The results showed that 90.63% (29/32) of the detected proteins exhibited stable plasma expression, and 18 genes displayed stable expression in at least one cell type, particularly in CD14+ monocytes. We also utilized these proteins to construct disease diagnostic models, and notably, models for 14 out of 18 diseases achieved an area under the curve (AUC) exceeding 0.8, indicating promising diagnostic potential.

Conclusions: We identified 72 proteins that causally influence CVD risk, providing new mechanistic insights into CVD and may prove to be promising targets as CVD therapeutics.