Serum Proteomics Reveals Biomarkers for Diagnosis, Stratification, and Mechanistic Insights into Cerebral Microbleeds

Wu-meng Yin^1,2Liu-chang He¹Guang-yi Han¹An-ming Li¹Hang-hang Zhu¹Yuan CAO¹Xinli Xue³Lei Zhang³Chang-he Shi¹Yuming Xu^1,2*Yunchao Wang^1*

• ¹Department of Neurology, First Affiliated Hospital of Zhengzhou University, Zhengzhou, China
• ²NHC Key Laboratory of Prevention and treatment of Cerebrovascular Diseases, Zhengzhou, China
• ³Clinical Systems Biology Laboratories, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, China

Abstract Objective: Cerebral microbleeds (CMBs) are small vascular lesions detectable on MRI and are associated with increased stroke risk and cognitive decline. However, imaging-based diagnosis is limited by cost and accessibility. This study aimed to identify serum protein biomarkers for early CMB diagnosis and to elucidate molecular mechanisms underlying CMB subtypes. Methods: We enrolled 43 patients with MRI confirmed CMBs and 38 healthy controls. Serum proteomic profiling used high performance liquid chromatography coupled with tandem mass spectrometry. Differential protein expression and pathway enrichment analyses were performed. Biomarkers were selected using LASSO, support vector machine recursive feature elimination, and random forest algorithms. Validation employed enzyme linked immunosorbent assay (n = 60) and Western blotting (n = 8). Results: We identified 151 proteins that differed between CMB and control groups. Altered pathways involved inflammation, extracellular matrix remodelling, and lipid metabolism. Five proteins emerged as candidate biomarkers: MMP3, EFEMP1, TIMP1, UMOD, and UBA52. MMP3, EFEMP1, TIMP1, and UMOD showed robust validation performance with AUC values greater than 0.7, and EFEMP1 positively correlated with CMB burden. Subtype analysis distinguished lobar from deep CMBs, with RCN1, NEO1, and APLP1 effectively discriminating subtypes