Plasma Proteomics to Identify Drug Targets for Ischemic Heart Disease

Background Integrated analyses of plasma proteomic and genetic markers in prospective studies can clarify the causal relevance of proteins and discover novel targets for ischemic heart disease (IHD) and other diseases. Objectives The purpose of this study was to examine associations of proteomics and genetics data with IHD in population studies to discover novel preventive treatments. Methods We conducted a nested case-cohort study in the China Kadoorie Biobank (CKB) involving 1,971 incident IHD cases and 2,001 subcohort participants who were genotyped and free of prior cardiovascular disease. We measured 1,463 proteins in the stored baseline samples using the OLINK EXPLORE panel. Cox regression yielded adjusted HRs for IHD associated with individual proteins after accounting for multiple testing. Moreover, cis-protein quantitative loci (pQTLs) identified for proteins in genome-wide association studies of CKB and of UK Biobank were used as instrumental variables in separate 2-sample Mendelian randomization (MR) studies involving global CARDIOGRAM+C4D consortium (210,842 IHD cases and 1,378,170 controls). Results Overall 361 proteins were significantly associated at false discovery rate <0.05 with risk of IHD (349 positively, 12 inversely) in CKB, including N-terminal prohormone of brain natriuretic peptide and proprotein convertase subtilisin/kexin type 9. Of these 361 proteins, 212 had cis-pQTLs in CKB, and MR analyses of 198 variants in CARDIOGRAM+C4D identified 13 proteins that showed potentially causal associations with IHD. Independent MR analyses of 307 cis-pQTLs identified in Europeans replicated associations for 4 proteins (FURIN, proteinase-activated receptor-1, Asialoglycoprotein receptor-1, and matrix metalloproteinase-3). Further downstream analyses showed that FURIN, which is highly expressed in endothelial cells, is a potential novel target and matrix metalloproteinase-3 a potential repurposing target for IHD. Conclusions Integrated analyses of proteomic and genetic data in Chinese and European adults provided causal support for FURIN and multiple other proteins as potential novel drug targets for treatment of IHD.

I schemic heart disease (IHD) accounts for about 9 million annual deaths worldwide, and the incidence rates are increasing in many populations, including China. 1 Tobacco smoking, hypertension, dyslipidemia, and adiposity are established risk factors for IHD, but other risk factors remain to be discovered.Recently, genome-wide association studies (GWAS) of IHD in primarily European ancestry populations have identified >240 common genetic variants, 2,3 but the mechanisms underlying many of these associations remain to be elucidated. 2Most drug targets are enzymes, transport, or structural proteins, and novel assays that measure plasma levels of large numbers of proteins, particularly when integrated with genetic data, could discover novel risk factors and potential therapeutic targets for IHD and improve risk prediction and early detection of IHD. 4 Previous studies of plasma proteins and IHD have highlighted the roles of apolipoproteins, 5 inflammatory proteins (eg, interleukin [IL]-6), 6 or early markers of disease (troponins or N-terminal prohormone of brain natriuretic peptide [NT-proBNP]). 7vances in proteomic assays now enable measurement of thousands of proteins in blood, 8 and their application in prospective studies have identified novel biomarkers for IHD. 9 However, previous proteomics studies of IHD have been confined to European populations, 10 and many lacked concomitant genetic data to assess and clarify the causal relevance of any observed associations, 4,9,11 greatly limiting the translational potential of such study findings.
We present integrated analyses of observational and genetic data for 1,463 proteins with risk of incident IHD in a nested case-cohort study within the prospective China Kadoorie Biobank (CKB).The study aimed to: 1) identify proteins associated with risk of IHD independent of established cardiovascular disease (CVD) risk factors; 2) use cis-protein quantitative loci (pQTLs) identified in GWAS of CKB and UK Biobank (UKB) for these proteins to assess their causal relevance for IHD in 2-sample Mendelian randomization (MR) analyses in a global IHD GWAS consortium; and 3) explore mechanisms linking such proteins with IHD using enrichment analyses, tissue expression, and single-gene knockout (KO) models.

METHODS
STUDY POPULATION AND DESIGN.Details of the study design, methods, and participants in CKB have been previously reported. 12In brief, CKB recruited >512,000 adults aged 30 for long-term storage in liquid nitrogen.Prior international, national, and local ethical approvals for the study were obtained and all participants provided written informed consent.After enrollment, the vital status and occurrence of specific diseases was monitored based on linkage to death and disease registers and to the National Health Insurance system for all hospital admissions. 12All causes of death or incident diseases were coded using ICD-10 codes by trained health workers, blinded to baseline information and integrated centrally.PROTEOMICS ASSAYS.Plasma levels of 1,463 proteins were measured using the OLINK EXPLORE platform in OLINK (Uppsala, Sweden), which included 4 similar-sized panels (ie, cardiometabolic, inflammation, neurology, and oncology). 13Details of the proteomics assays are provided in the Supplemental Methods, Supplemental Tables 1 and 2, and Supplemental Figure 1.
STATISTICAL METHODS.For observational analyses of proteins with IHD, Cox regression models were used to estimate adjusted HRs (and 95% CIs) for IHD using the Prentice pseudo-partial likelihood for casecohort designs. 14All analyses were stratified based on sex and study area, and adjusted sequentially in different models for additional CVD risk factors (Supplemental Methods).For genetic analyses involving 1-and 2-sample MR analyses, we used cis-pQTLs derived from GWAS of protein levels in Chinese (CKB) and European (UKB) adults to assess the causal relevance of proteins identified in CKB observational analyses. 15r proteins with cis-pQTLs identified in CKB, we undertook 1-sample MR analyses in the case-cohort sample (n ¼ 1,937 IHD cases) and in the overall CKB genotyping data set (n ¼ 6,499 IHD cases). 16We further conducted 2-sample MR analyses of the proteins identified in CKB observational analyses using: 1) cis-pQTLs obtained from CKB, with lookups in the CARDIOGRAMþC4D (CC4D) Consortium of European and Japanese adults that included 210,842 IHD cases and 1,378,170 controls, 2 and also separately in Biobank Japan 17 ; and 2) cis-pQTLs from UKB 15 with lookups in the CC4D of European adults only (n ¼ 181,522 cases). 2 To explore genetic associations of cis-pQTLs with CVD traits, we searched PhenoScanner, which included GWAS results for >100 traits or disease outcomes, 18 using a P value threshold of <10 Â 5 -8 .The study design involved observational and genetic analyses in the China Kadoorie Biobank (CKB) and UK Biobank (UKB) studies.The observational associations of proteomics with ischemic heart disease (IHD) were replicated using Mendelian randomization (MR) approaches using instrumental variables discovered in CKB and replicated in UKB.The downstream analyses included phenome-wide associations (PheWAS), knockout models (KO), assessment of enrichment, and functional genome-wide association study analyses (FGWAS), respectively.
Additional downstream analyses were also undertaken of any proteins causally associated with IHD in genetic analyses, including: 1) KO models; 2) enrichment analyses using the DAVID (Database for Annotation, Visualisation and Integrated Discovery) bioinformatics resource 6.8; 19 3) enrichment in chromatin in human umbilical endothelial cells tissues using functional GWAS; 20   Proteomic Targets for Ischemic Heart Disease likely to be men (62.1% vs 37.9%), to be less welleducated, and to smoke than subcohort participants.
IHD cases also had a higher prevalence of hypertension and diabetes and higher mean levels of systolic blood pressure (SBP), adiposity, and random blood glucose (Table 1).The median time from enrollment to diagnosis of IHD cases was 6.8 years (IQR: 4.5 years).3 and 4, and 3).These 361 proteins were only moderately correlated with each other, with 99.8% of protein pairs having correlation coefficients ranging from À0.5 to þ0.5.After applying a more stringent Bonferroni-correction threshold, 93 proteins remained significantly associated with IHD (Supplemental Table 5).
The associations were typically log-linear, with no evidence of a threshold beyond which protein levels were no longer associated with IHD (Supplemental Figure 3).The results were unaltered in sensitivity analyses (Supplemental proteins overlapping with those identified in the main analyses (Supplemental Table 7, Supplemental Figure 4).

GENETIC ASSOCIATIONS OF PROTEINS WITH IHD.
Among the 361 significant proteins identified in CKB observational analyses, cis-pQTL variants were identified in CKB GWAS for 212.In 1-sample MR analyses, none of these associations were significant.
In 2-sample MR analyses involving CC4D and Biobank Japan, 13 proteins (FURIN, Sortilin [SORT1], placenta growth factor [PGF], Asialoglyco- proteinase-activated receptor 1 [F2R], OBP2B, TFPI,  The volcano plots show the associations of proteins with IHD stratified by OLINK panels in CKB after adjustment for confounding factors.All models were stratified by sex and region and adjusted for age, age 2 , fasting time, fasting time 2 , ambient temperature, ambient temperature 2 , plate identification, education, smoking, alcohol consumption, physical activity, systolic blood pressure (SBP), type 2 diabetes, ApoB/ApoA, and body mass index (BMI).Red, blue, and gray dots denote positive significant, inverse significant, and nonsignificant associations, respectively.Time in study was used as time scale in all models.Apo ¼ apolipoprotein; other abbreviations as in Figure 1.Mazidi et al Proteomic Targets for Ischemic Heart Disease (Supplemental Figure 6).Moreover, in 2-sample MR analyses solely involving Biobank Japan (n ¼ 29,319 cases), 4 proteins (SORT, PGF, OBP2B, and ANGPTL1) were replicated (Table 2).The Pearson correlation coefficients of these 13 proteins with each other were all #0.75.

DOWNSTREAM ANALYSES OF NOVEL PROTEINS AS
POTENTIAL DRUG TARGETS FOR IHD.In phenomewide association analyses of these 13 proteins, cis-pQTLs for FURIN were associated with several CVD traits, including high blood pressure and other CVD type traits, whereas SORT1 cis-pQTL was associated with 2 CVD types, lipids, and 7 additional traits (Supplemental Table 8).Likewise, cis-pQTLs for PGF, F2R, TFPI, and EFNA1 were related to CVD outcomes (PGF) and CVD risk factors.Within CKB, TFPI, PGF, F2R, and ASGR1 were each significantly associated with all 4 CVD risk factors (smoking, SBP, adiposity, and type 2 diabetes).Moreover, FURIN, ASGR1, SORT1, TFPI, PGF, F2R, and EFNA1 were also associated with SBP, adiposity, and type 2 diabetes.
Analysis of Open Targets and other databases indicated evidence of drug development, including phase II-IV trials, for 6 proteins (PGF, ASGR1, F2R, TFPI, tenascin, and MMP3), with 3 related to CVD outcomes or traits (F2R for MI, TFPI for ICH, and ASGR1 for familial hypercholesterolemia) (Table 2).
However, there were no reports of drug targets for the 7 remaining proteins, including FURIN, CCL17, and SORT1.Drug target development for Tclin outcomes revealed PGF and F2R each having 1 "active drug," whereas Tchem demonstrated "active ligands" for the majority of proteins including FURIN (Table 2).

Combined analyses of proteomic and genomic data
in Chinese adults provided strong support for the causal relevance of 13 proteins for IHD, with 4 of these proteins (ie, FURIN, F2R, ASGR1, and MMP3) further replicated in European populations.Further downstream analyses such as phenome-wide associations and KO models confirmed the importance of several of these 13 proteins for CVD or CVD-related traits and their potentials as drug targets for IHD.Among these 13 proteins, there was, however, no evidence of drug development for 7, while for the remaining 6 proteins only 1 was associated with drug development for IHD.Moreover, none of them were identified in previous CC4D using large-scale GWAS data alone. 26][27] Indeed, NT-proBNP was the single protein most strongly associated with IHD in almost all studies (including the present study), but few previous studies assessed the causal relevance of these associations.The present study involving both CKB and UKB showed that cis-pQTLs for NT-proBNP were unrelated with IHD.This suggested that NT-proBNP is not a causal risk factor for IHD, but a marker of

Mazidi et al
Proteomic Targets for Ischemic Heart Disease N O V E M B E R 1 4 , 2 0 2 3 : 1 9 0 6 -1 9 2 0 subclinical atherosclerosis, which could inform risk prediction but not treatment.For PCSK9, we did not find significant genetic associations with IHD using leading cis-pQTL in CKB.However, separate analyses of >100,000 genotyped participants in CKB demonstrated that higher levels of PCSK9 genetic scores were significantly and positively associated with risk of CVD (manuscript under review).
Recently, a genetic study of 304 proteins measured using mass spectrometry among 2,410 Chinese adults reported significant associations of cis-QTLs for 3 proteins (apolipoprotein a, apolipoprotein E, and haptoglobin) with CAD in 2-sample MR analyses involving Biobank Japan.However, none of these proteins overlapped with those evaluated in CKB.
Moreover, unlike in the present report, that Chinese study did not include observational analyses of proteins with CVD outcomes. 28In a previous report from the DECODE study of 35,559 individuals, 1,287 of 4,907 proteins measured using SomaScan platform were significantly associated with myocardial infarction (MI) (n ¼ 3,457 cases), with the top 3 proteins being NPPB (HR: 1.64 per 1-SD higher concentration), MMP-12 (1.55), and CTHRC1 (1.50). 9The present study also showed significant associations of NPPB and MMP-12 with IHD (1.41 and 1.35, respectively), but did not assess CTHRC1.Likewise, one of the proteins that was most strongly and inversely associated with IHD The proteins shown in bold were significantly associated with IHD in both Europeans and Chinese.a Beta and SE were calculated using Wald ratio formula.b Additional replication in the Biobank Japan (29,319 IHD cases) and directionally consistent with observational analysed in CKB.c Target has at least 1 approved drug.d Target has at least 1 ChEMBL compound with an activity cutoff of <30 nmol/L.
JACC VOL.82, NO. 20, 2023   Mazidi et al Proteomic Targets for Ischemic Heart Disease (contactin-5) in the present study was also reported by DECODE (HR: 0.78 vs 0.77).However, DECODE did not evaluate the causal relevance of these associations using MR analyses. 9To date, no large studies have directly compared associations of overlapping proteins measured using different assay platforms (eg, OLINK EXPLORE vs SomaScan) with IHD in the same individuals.Nevertheless, it has been reported that the available OLINK panels include more targeted selection of proteins, whereas SomaScan panels include greater analytical breath. 29e integrated analyses of proteomic and genomic data in the present study demonstrated that FURIN (also known as PCSK3) was the most strongly (and causally) IHD-associated protein.The FURIN gene encodes the subtilisin-like proprotein convertase (PCSK3).FURIN plays a regulatory role in inflammation and atherosclerosis. 30FURIN is also involved in activation of proBNP, PCSK9, and ANGPTL, which each contribute to the development of atherosclerosis. 31In addition, FURIN has been implicated in the regulation of blood pressure by controlling activation of prorenin receptors, proliferation of vascular smooth muscle, and activation of epithelial sodium channels.In experimental KO models, irreversible inhibition of FURIN (via a1-antitrypsin) attenuates the progression of atherosclerosis. 32Previous studies reported that higher levels of FURIN were associated with MCP-1, leading to activation of inflammation. 33spite the latter associations with FURIN, there has  For the 3 remaining proteins (ie, F2R, ASGR1, and MMP3), we found evidence of drug development.F2R is a G protein-coupled receptor that is irreversibly activated by MMP proteins. 36Previous studies have reported higher levels of F2R expression in human atherosclerotic lesions. 37Inhibition of F2R receptors by vorapaxar, which reduces thrombin-induced platelet activation after coronary revascularization in patients with acute occlusion of coronary arteries, 38 has been approved for treatment of IHD, stroke, and other CVD outcomes.ASGR1 is associated with low-density lipoprotein-cholesterol and is a target for treatment of familial hypercholesterolemia. 39 A previous genetic study in Europeans (42,524   cases and 249,414 controls) reported that ASGR1 insufficiency was associated with lower levels of lowdensity lipoprotein-cholesterol and lower risks of IHD, 40 consistent with present study findings.MMP proteins are highly expressed in endothelial and vascular smooth muscle cells and have been implicated in angiogenesis and atherosclerosis. 41In the DECODE study of 35,559 Icelanders with 3,457 incident MI cases, plasma levels of MMP3, assayed using the SomaScan platform, were inversely associated with risk of MI (HR: 0.86), but the cause-effect nature of association was not evaluated. 9In contrast, we demonstrated positive associations of MMP3 with risks of IHD in both observational and genetic analyses.It is unclear whether the discrepant results for MMP3 between the DECODE and CKB studies reflect assay differences or other ethnic differences between 2 populations.Importantly, MMP3 is currently being investigated as a novel protein target for treatment of lung and breast cancer, highlighting potential opportunities for drug-repurposing for IHD.
Among the remaining 9 proteins causally associated with IHD in Chinese but not in Europeans, there was no evidence of drug development for SORT1, OBP2B, ANGPTL1, CCL17, REG1B, and EFNA1, whereas none of the remaining 3 proteins (PGF, TFPI, and TNC) were associated with drug development for IHD.
Both ANGPTL and angiopoietin-1 proteins have been implicated in angiogenesis, inflammation, lipids, and glucose metabolism. 42ANGPT1 is an established drug target for the treatment of cancer, but we did not identify any evidence of drug development related to IHD. 43 Hence, given the present study findings, ANGPT1 should also be considered as a protein target for IHD.SORT1 augments PCSK9 secretion, 44 and has been implicated in development of atherosclerosis. 45evious studies have provided genetic support, consistent with present study findings, for the causal relevance of SORT1 for MI and other CVD outcomes. 46wever, we found no evidence of drug development for SORT1 for any CVD outcomes.CCL17 is a chemokine that is selectively expressed in human macrophages and plays a pivotal role in atherosclerosis by promotion of leukocyte infiltration into endothelial cells. 47We found somewhat stronger genetic associations for CCL17 with IHD in CKB than in UKB (1.10 vs 1.04), which may reflect differences in the lead variants used or their MAFs (UKB 0.08 vs CKB 0.50) between the 2 populations.In animal experiments, KO of CCL17 genes delayed progression of atherosclerosis and attenuated ischemia-reperfusion injury. 482 The present study identified 13 potential novel protein targets for drug treatment of IHD that had not been previously discovered using large-scale genomic data alone. 2In addition to discovery of potential novel drug targets, large-scale proteomics and genetic data in diverse populations should improve risk prediction, early detection, and prevention strategies of IHD and many other diseases.
ACKNOWLEDGMENTS The chief acknowledgment is to the participants, the project staff, and the China Center for Disease Control and its regional offices for assisting with the fieldwork.
to 79 years from 10 diverse (5 urban, 5 rural) regions in China during 2004-2008.The baseline survey included a laptop-based questionnaire covering sociodemographic factors, lifestyle, medical history and medication use, and physical measurements (eg, height, weight, waist circumference, blood pressure, and lung function).Study participants provided a 10-mL nonfasting blood sample (with time since last meal recorded)

9 PGF
The CKB complies with all required ethical standards for medical research on human subjects.Ethical approvals were obtained and have been maintained by the relevant institutional ethical research committees in the United Kingdom and China.All participants provided written informed consent.The present study involved a case-cohort design (Figure 1, Central Illustration) comprising 1,937 incident cases of IHD (ICD-10 codes: I20, I22-I25) that were accrued during a 12-year follow-up before January 1, 2019 and a subcohort of 2,001 participants.The IHD cases were a random sample of incident IHD cases that had GWAS data and no prior history of CVD, and excluded statin users at enrollment.The subcohort participants were randomly selected from a population subset of 69,353 genotyped participants, but were genetically SEE PAGE 1932 A B B R E V I A T I O N S A N D A C R O N Y M = placenta growth factor pQTLs = protein quantitative loci SBP = systolic blood pressure UKB = UK Biobank The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors' institutions and Food and Drug Administration guidelines, including patient consent where appropriate.For more information, visit the Author Center.Manuscript received May 19, 2023; revised manuscript received August 11, 2023, accepted September 5, 2023.JACC VOL.82, NO. 20, 2023 Mazidi et al N O V E M B E R 1 4 , 2 0 2 3 : 1 9 0 6 -1 9 2 0 Proteomic Targets for Ischemic Heart Disease unrelated to each other and had no prior history of CVD or statin use at baseline.

Figure 1
Figure1provides an overview of the main analytic approaches and key findings, including observational, genetic, and relevant additional downstream analyses.OBSERVATIONAL ASSOCIATIONS OF PROTEINSWITH CONVENTIONAL CVD RISK FACTORS AND WITH IHD.A total of 1,009 proteins were significantly associated (at FDR <0.05) with body mass index, 928 with SBP, 833 with diabetes, and 329 with smoking, and 169 proteins with all of these risk factors (Supplemental Figure2).In Model 2, 558 proteins were significantly associated at FDR <0.05 with risk of IHD (548 positively, 10 inversely) (Supplemental Table3),

FIGURE 3
FIGURE 3 Adjusted HRs of IHD Associated With Selected Leading Proteins in CKB Forest plot shows adjusted HRs of IHD for proteins most strongly associated with IHD for cardiometabolic, inflammation, neurology, and oncology panels.The numbers of significantly associated proteins on each panel are shown and results for individual proteins are displayed.The models were adjusted for covariates as in Figure2.The boxes are HRs and the horizontal lines are 95% CIs.The area of each box is inversely proportional to the variance of the log HR.Abbreviations as in Figure1.Continued on the next pageMazidi et alJ A C C V O L .8 2 , N O . 2 0 , 2 0 2 3 Proteomic Targets for Ischemic Heart Disease N O V E M B E R 1 4 , 2 0 2 3 : 1 9 0 6 -1 9 2 0 tenascin, EFNA1, REG1B, and matrix metalloproteinase [MMP]3) were significantly associated at FDR <0.05 with IHD (Table 2, Supplemental Figure 5).Of these 13 proteins, 3 (FURIN, SORT1, and PGF) had similar HRs in both observational and MR analyses, whereas the HRs for the remaining 10 proteins were less extreme in MR analyses albeit directionally concordant with the observational analyses FIGURE 3 Continued

FIGURE 4
FIGURE 4 Replication of Observational Analyses in CKB in Genetic Analyses in CARDIOGRAMþC4D

Study design Discovery in observational analyses Genetic instruments 2-sample MR analyses Identification of causal proteins Downstream analyses of mechanisms and drug targets CKB:
Summary of Study Design, Analytic Approaches, and Key Findings FIGURE 1 Plasma Proteomics Identifies Novel Drug Targets for Ischemic Heart Disease RESULTSOverall, patients with IHD were older (mean baseline age: 63.8 AE 9.2 years vs 51.2 AE 10.3 years) and more CENTRAL ILLUSTRATION Mazidi M, et al.J Am Coll Cardiol.2023;82(20):1906-1920.

Table 6
) or additional analyses without censoring any subcohort participants who developed incident IHD during follow-up (332 significant proteins in the final model, at FDR <0.05).In an internal replication analyses of subcohort participants, 236 proteins were significantly associated with risk of IHD (n ¼ 232 cases) in the final model, with 147 (62%)

TABLE 1
Baseline Characteristics of IHD Cases and Subcohort Participants in CKB Values are mean AE SD or %.All values are adjusted for age, sex, and region, where appropriate.a Medical history was based on self-report of physician-diagnosed conditions, with use of medication restricted to those reporting relevant conditions.BMI ¼ body mass index; DBP ¼ diastolic blood pressure; MET ¼ metabolic equivalent of task h/d; RBG ¼ random blood glucose; SBP ¼ systolic blood pressure; WC ¼ waist circumference.

TABLE 2
Genetic Information and Key Findings in CC4D and Downstream Analyses for Proteins Associated With IHD in Europeans and Chinese

HR Per 1-SD Higher Protein Concentration
For assisting with conduct and organization of the study, the authors thank Judith Mackay in Hong Kong; Yu Wang, Gonghuan Yang, Zhengfu Qiang, Lin Feng, Maigeng Zhou, Wenhua Zhao, and Yan Zhang at the China Center for Disease Control; and Lingzhi Kong, Xiucheng Yu, and Kun Li at the Chinese Ministry of Health.The authors thank Dr Anuj Goel for assistance in evaluation of enrichment analysis.The China Kadoorie Biobank (CKB) is a global resource for the investigation of lifestyle, environmental, blood biochemical, and genetic factors as determinants of common diseases.CKB is committed to making the cohort data available to the scientific community.Details of the data are currently available to the open access users and how to apply for it are provided on the CKB study website.CKB research tracker no.: 2022-0037; CKB data release: 17.Code availability: Custom code was used for statistical analyses and will be made available to external researchers on reasonable request to bona fide researchers.FUNDING SUPPORT AND AUTHOR DISCLOSURES The CKB baseline survey and the first resurvey were supported by the Kadoorie Charitable Foundation in Hong Kong.The long-term follow-up and subsequent resurveys have been supported by Wellcome grants to Oxford University (212946/Z/18/Z, 202922/Z/16/ Z, 104085/Z/14/Z, 088158/Z/09/Z) and grants from the National Natural Science Foundation of China (82192901, 82192904, 82192900) and from the National Key Research and Development Program of China (2016YFC0900500).The UK Medical Research Council (MC_UU_00017/1, MC_UU_12026/2, MC_U137686851), Cancer Research UK (C16077/A29186, C500/A16896), and British Heart Foundation (CH/1996001/9454) provide core funding to the Clinical Trial Service Unit and Epidemiological Studies Unit, Oxford University for the project.The proteomic assays were supported by a BHF Intermediate Clinical Research Fellowship to MVH (FS/18/23/ 33512), Novo Nordisk, and OLINK.DNA extraction and genotyping were supported by GlaxoSmithKline and the UK Medical Research Council (MC-PC-13049, MC-PC-14135).Dr Holmes is currently employed by 23andMe (and owns stock in 23andMe, Inc).Dr Howson is a full-time employee of Novo Nordisk Research Centre Oxford Limited and owns shares in Novo Nordisk.All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
TRANSLATIONAL OUTLOOK: Future research will expand the array of potential drug targets that can be evaluated in clinical studies of a diversity of patients with IHD.