Integrating multi-modal omics to identify therapeutic atherosclerosis pathways for coronary heart disease

Abstract

Background: Urinary metabolism breakdown products reflect metabolic changes in atherosclerosis-relevant tissues and may contain relevant therapeutic leads. We integrated data on urinary metabolism breakdown products, plasma proteins, atherosclerotic plaque tissue, and single-cell expression to identify druggable metabolic pathways for coronary heart disease (CHD). Methods: Mendelian randomisation was employed to interrogate findings from independent genome-wide association studies on 954 urinary metabolism breakdown products, 1562 unique proteins, and 181,522 CHD cases, establishing directionally concordant associations. Using the Athero-Express Biobank, concordant plasma proteins were linked to plaque vulnerability using protein and mRNA expression in plaque. Single-cell RNA sequencing data obtained from carotid plaque samples were used to test for differential expression of concordant proteins across plaque cell types. Findings: In total, 29 urinary metabolism breakdown products associated with CHD, predominantly originating from amino acid metabolism (n = 12) or unclassified origin (n = 9). We identified 113 plasma proteins with directionally concordant associations with these urinary metabolism breakdown products and CHD. Of the 110 proteins available in plaque, 16 were associated with plaque vulnerability. This included positive control proteins targeted by drugs indicated for CHD, such IL6R (targeted by tocilizumab) and AT1B2 (targeted by digoxin), as well as a potential repurposing opportunity C1S (targeted by sutimlimab). Interpretation: We have identified amino acid metabolism as an important contributing pathway to CHD risk. These metabolism pathways were linked to 16 prioritised proteins relevant for CHD with involvement in atherosclerotic plaques, providing important insights for drug development. Funding: SR and SP are supported by a VIDI Fellowship (project number 09150172010050) from the Dutch Organisation for Health Research and Development (ZonMW) awarded to SP. AFS is supported by BHF grant PG/22/10989, the UCL BHF Research Accelerator AA/18/6/34223, the UCL BHF Centre of Research Excellence RE/24/130013, MR/V033867/1, the National Institute for Health and Care Research University College London Hospitals Biomedical Research Centre, and the EU Horizon scheme (AI4HF 101080430 and DataTools4Heart 101057849). MV is supported by a postdoc talent grant from the Amsterdam Cardiovascular Sciences. This work was funded by UK Research and Innovation (UKRI) under the UK government's Horizon Europe funding guarantee EP/Z000211/1, and by the Rosetrees CF-2-2023-M-2/122. This publication is part of the project “Computational medicine for cardiac disease” with file number 2023.022 of the research programme “Computing Time on National Computer Facilities” which is (partly) financed by the Dutch Research Council (NWO).