Introduction: The specific protein composition of stroke-causing emboli is unknown. Because ischemic stroke has a varied etiology, it is possible that the composition of the thrombus from which an embolus originated will have distinctive molecular characteristics reflective of the underlying pathophysiology. In this study, we used mass spectrometry to evaluate the protein composition of retrieved emboli from patients with differing stroke etiologies and correlated the protein levels to serum predictors of atherosclerosis.
Methods: Emboli from 20 consecutive acute stroke patients were retrieved by embolectomy during routine stroke care. Thrombus proteins were extracted, digested and multidimentional fractionation of peptides was performed. Fractionated peptides underwent nano-liquid chromatography with tandem mass spectrometry. Spectra were searched using Mascot software in which results with p<0.05 (95% confidence interval) were considered significant and indicating identity. The results were correlated to A1C, LDL, and ESR taken on admission.
Results: Demographics: 55% had atrial fibrillation, 20% had significant proximal vessel atherosclerosis, 10% were cryptogenic, and three had other identified etiologies (left ventricular thrombus, dissection, endocarditis). Eighty-one common proteins (eg hemoglobin, fibrin, actin) were found in all 20 emboli. Serum LDL levels correlated with Septin-2 (rs=0.78, p=0.028), Phosphoglycerate Kinase 1 (rs =0.75, p=0.036), Integrin Alpha-M (rs=0.68, p=0.033) and Glucose-6-phosphate dehydrogenase (rs =0.63, p=0.05). Septin-7 levels inversely correlated to ESR (rs = -0.84, p=0.01). No significant protein correlations to A1C were found.
Conclusion: To our knowledge, this study is the first mass spectrometry analysis of thrombi retrieved from acute stroke patients and the first to correlate the thrombus proteome to clinical features of the patient. Notably, we found proteins associated with inflammation (eg Integrin Alpha–M) in emboli from patients with high LDL. Although these findings are tempered by a small sample size, we provide preliminary support for the feasibility of utilizing proteomic analysis of emboli to discover proteins that may be used as markers for stroke etiology.