BackgroundThe objective of this study is to estimate the causal relationship between plasma proteins and myocardial infarction (MI) through Mendelian randomization (MR), predict potential target-mediated side effects associated with protein interventions, and ensure a comprehensive assessment of clinical safety.MethodsFrom 3 proteome genome-wide association studies (GWASs) involving 9775 European participants, 331 unique blood proteins were screened and chosed. The summary data related to MI were derived from a GWAS meta-analysis, incorporating approximately 61,000 cases and 577,000 controls. The assessment of associations between blood proteins and MI was conducted through MR analyses. A phenome-wide MR (Phe-MR) analysis was subsequently employed to determine the potential on-target side effects of protein interventions.ResultsCausal mediators for MI were identified, encompassing cardiotrophin-1 (CT-1) (odds ratio [OR] per SD increase: 1.16; 95% confidence interval [CI]: 1.13-1.18; p = 1.29 __-mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"-__× __-mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"-__10 - 31 ), Selenoprotein S (SELENOS) (OR: 1.16; 95% CI: 1.13-1.20; p = 4.73 __-mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"-__× __-mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"-__10 - 24 ), killer cell immunoglobulin-like receptor 2DS2 (KIR2DS2) (OR: 0.93; 95% CI: 0.90-0.96; p = 1.08 __-mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"-__× __-mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"-__10 - 5 ), vacuolar protein sorting-associated protein 29 (VPS29) (OR: 0.92; 95% CI: 0.90-0.94; p = 8.05 __-mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"-__× __-mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"-__10 - 13 ), and histo-blood group ABO system transferase (NAGAT) (OR: 1.05; 95% CI: 1.03-1.07; p = 1.41 __-mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"-__× __-mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"-__10 - 5 ). In the Phe-MR analysis, memory loss risk was mediated by CT-1, VPS29 exhibited favorable effects on the risk of 5 diseases, and KIR2DS2 showed no predicted detrimental side effects.ConclusionsElevated genetic predictions of KIR2DS2 and VPS29 appear to be linked to a reduced risk of MI, whereas an increased risk is associated with CT-1, SELENOS, and NAGAT. The characterization of side effect profiles aids in the prioritization of drug targets. Notably, KIR2DS2 emerges as a potentially promising target for preventing and treating MI, devoid of predicted detrimental side effects.