Genome-wide association analysis and Mendelian randomization proteomics identify drug targets for heart failure

Abstract

We conduct a large-scale meta-analysis of heart failure genome-wide association studies (GWAS) consisting of over 90,000 heart failure cases and more than 1 million control individuals of European ancestry to uncover novel genetic determinants for heart failure. Using the GWAS results and blood protein quantitative loci, we perform Mendelian randomization and colocalization analyses on human proteins to provide putative causal evidence for the role of druggable proteins in the genesis of heart failure. We identify 39 genome-wide significant heart failure risk variants, of which 18 are previously unreported. Using a combination of Mendelian randomization proteomics and genetic cis-only colocalization analyses, we identify 10 additional putatively causal genes for heart failure. Findings from GWAS and Mendelian randomization-proteomics identify seven (CAMK2D, PRKD1, PRKD3, MAPK3, TNFSF12, APOC3 and NAE1) proteins as potential targets for interventions to be used in primary prevention of heart failure.

Fig. 1. Schematic diagram of the datasets and analyses.

HF heart failure, MVP Million Veteran Program cohort, GWAS genome-wide association study, pQTL protein quantitative trait loci, PheWAS phenome-wide association study, MR Mendelian randomization, FDR false discovery rate, PP.H4 posterior probability of H₄.

Fig. 2. Manhattan plots showing associations with HF from a GWAS meta-analysis on n = 1,266,315 individuals and b MR-wide proteomics.

a Manhattan plot showing the −log₁₀(P value) of association for each SNP from the GWAS meta-analysis plotted on the y-axis against genomic position on the x-axis. The red dotted line corresponds to the genome-wide significance threshold. The summary statistics of independent lead SNPs are noted in Supplementary Data 1. b Manhattan plot showing the −log₁₀-transformed FDR-adjusted P value of association for each gene plotted against genomic position on the x-axis. All tests were two-sided and adjusted for multiple comparisons. The blue line corresponds to an FDR threshold of 5% and points are color-coded by drug tractability information based on data provided by OpenTargets; green for druggable genes. FDR false discovery rate.

Fig. 3. Plots showing a genetic association of 18 HF loci against risk factors for HF and b MR and colocalization estimates of MR-proteomic genes-hits against HF risk factors.

a The color of the bubble corresponds to the beta coefficient of the genetic association between the loci (x-axis) and trait (y-axis). Blue corresponds to a negative and red corresponds to a positive beta coefficient. The size of each bubble corresponds to the negative logarithm of the association p value; larger size corresponds to lower p values. Loci are grouped by druggable and non-druggable genes. All tests were two-sided without adjustment for multiple comparisons. Associations which passed the p value threshold (p < 1 × 10⁻⁴) are denoted by a yellow diamond. b This bubble plot shows MR estimates for which p < 1 × 10⁻⁴. The size of each bubble corresponds to the posterior probability for hypothesis 4 derived from colocalization. The color of the bubble corresponds to the beta coefficient derived from MR. Blue corresponds to a negative association and red corresponds to a positive association; note that a positive $\beta$ indicates either an increase in protein levels corresponding to an increase in HF risk or a decrease in protein levels corresponding to a decrease in HF risk, while a negative $\beta$ indicates either a decrease in protein levels corresponding to an increase in HF risk or an increase in protein levels corresponding to a decrease in HF risk. The intensity of the color corresponds to −log₁₀(P value) for the strength of association in the MR. All tests were two-sided without adjustment for multiple comparisons. Loci are grouped by druggable and non-druggable genes. TNXB, SIRPA, and ENPEP genes are not included as these had no MR estimates on HF risk factors that pass the p < 1 × 10⁻⁴ threshold. $\beta$, Beta coefficient, AC alcohol consumption, AF atrial fibrillation, BMI body mass index, CAD coronary artery disease, COPD chronic obstructive pulmonary disease, DBP diastolic blood pressure, eGFR estimated glomerular filtration rate, HDL-C high-density lipoprotein cholesterol, IL-6 Interleukin-6, LDL-C low-density lipoprotein cholesterol, NT-proBNP N-terminal proBNP, SBP systolic blood pressure, SMK smoking, T2D type-2 diabetes, TRP troponin I cardiac muscle, PP.H4 posterior probability of H₄.