Abstract Body (Do not enter title and authors here): Introduction: The association of inflammatory, metalloproteinases, and cell adhesion proteins with vulnerable atherosclerotic plaques is well known. However, translating proteins into tissue biomarkers has remained challenging. Moreover, the varying contributions of different subregions (media, necrotic core, and fibrous cap) are often overlooked. Hence, we hypothesize that plaque subregions are varyingly associated with plaque vulnerability, and subregion-specific proteomic analysis can reveal tissue biomarkers for plaque vulnerability.

Methods: We designed a balanced cohort (AHA-classification, sex, etc.) of stable and unstable plaques from carotid endarterectomy patients (n=112) and microdissected the histological subregions. The proteomes of 336 samples were acquired using an ultrasensitive ion-mobility mass spectrometer. LIMMA and GSEA were used for the differential and functional analyses.

Results: We quantified 4082 proteins, with ECM and matrisome proteins significantly over-represented (FDR<0.001). The necrotic core proteomes were notably distinct and enriched in metalloproteinase, immunoregulation, and inflammatory proteins. Differential analysis identified 132 proteins associated with vulnerability in the necrotic core, 80 in the fibrous cap, and none in the media (FDR<0.05). Functional validation via siRNA-mediated knockdown of the most statistically associated protein (PCSK9) in oxidized phospholipid-activated vascular smooth muscle cells showed downregulation of TNF-α signaling and upregulation of TGF-β and JAK-STAT signaling, linking plaque vulnerability to decreased proliferation and increased inflammatory signaling.

Our consensus clustering identified five molecular subtypes of plaques. We used a genetic algorithm-based approach to develop a protein panel with the highest predictive power for plaque vulnerability across subtypes. The optimal panel consisted of proteins from the necrotic core and fibrous cap and had a predictive accuracy of 0.87 (p<0.001) in the holdout dataset. Immunohistochemistry validation in a matching cohort confirmed the panel’s association with plaque vulnerability.

Conclusion: Our findings suggest that the necrotic core and fibrous cap have the strongest association with carotid plaque vulnerability, and the proposed proteins hold the potential to predict plaque vulnerability. This study highlights the potential of histology-guided proteomics in developing biomarkers for predicting clinical outcomes in atherosclerosis.