Genetically predicted expression of cuproptosis and copper-transport genes and coronary artery disease risk: a cis-Mendelian randomization and colocalization study

Background. Cuproptosis, a copper-dependent form of regulated cell death, has been repeatedly reported to be transcriptionally activated in atherosclerotic plaque, and cuproptosis- and copper-transport genes have been nominated as candidate biomarkers and therapeutic targets on the basis of descriptive expression studies. Whether genetically determined expression of these genes is causally related to atherosclerotic cardiovascular disease has not been tested. Methods. We first confirmed, across two independent carotid plaque transcriptomic cohorts (GSE43292 and GSE28829), that a panel of copper-metabolism and cuproptosis genes is concordantly dysregulated in plaque and that the copper importers track the degree of inflammatory-cell infiltration. We then performed two-sample cis-Mendelian randomization (MR) and Bayesian colocalization to test the causal effect of the genetically predicted expression of 11 cuproptosis/copper genes on coronary artery disease (CAD). Instruments were cis-eQTLs from whole-blood (eQTLGen, n = 31,684) and tibial artery (GTEx v8, n = 584); the outcome was CAD (CARDIoGRAMplusC4D/Aragam 2022; 181,522 cases, 1,165,690 participants, European ancestry). Large-artery atherosclerotic stroke (GIGASTROKE) was examined as a secondary outcome. LPL served as a positive control. The positive control was intended to validate the direction and harmonization of the MR pipeline, not to demonstrate that single-causal-variant colocalization must succeed at a known multi-signal locus such as LPL. Results. In plaque, copper-import genes (SLC31A2, SLC31A1) were upregulated and correlated strongly with myeloid/inflammatory infiltration (SLC31A2 vs. M1-macrophage r = +0.87), whereas the exporter ATP7B correlated inversely (r = -0.73). The positive control LPL showed the expected robust protective effect on CAD (IVW OR 0.90 per SD higher expression, 95% CI 0.87-0.93, P = 4.1 x 10-12), validating the pipeline. None of the 11 cuproptosis/copper genes showed colocalization-supported causal evidence for CAD. Two genes reached nominal MR significance in single-instrument analyses (MTF1 OR 1.16, P = 2.7 x 10-3, DLAT OR 1.41, P = 2.2 x 10-3,) but were not supported by colocalization (MTF1 PP.H4 = 4 x 10-6 with PP.H3 = 0.9998; DLAT PP.H4 = 0.29), indicating linkage-disequilibrium confounding rather than causality. For well-instrumented genes (e.g. ATP7B, SLC31A2, FDX1), estimates were null and precise enough to exclude odds ratios larger than {approx}1.05 per SD. Conclusions. Common cis-regulatory variation determining the baseline expression of cuproptosis/copper-transport genes in available eQTL tissues did not support a detectable causal effect on CAD risk. This does not exclude locally acquired plaque cuproptosis; rather, the reported upregulation of these genes in plaque is most consistent with a consequence or marker of established disease rather than a germline-encoded causal driver, and cautions against target nomination based on descriptive transcriptomics alone. Keywords: cuproptosis; atherosclerosis; coronary artery disease; Mendelian randomization; colocalization