Urokinase plasminogen activator (uPA) as a novel biomarker for cerebral amyloid angiopathy

Abstract

Background Diagnosis of the highly prevalent disease cerebral amyloid angiopathy (CAA) is mainly based on radiological identification of cerebral micro‐ or macrobleeds (according to the diagnostically acclaimed Boston Criteria) in the end‐stage of disease. Body fluid biomarkers (e.g. in cerebrospinal fluid) could potentially act as alternatives and allow for earlier identification of disease in a minimally invasive manner. We have explored the involvement of potential CAA‐biomarker urokinase plasminogen activator (uPA) in CAA development and progression in human patients and rodent disease models (rTg‐DI). Additionally, we analyzed uPA cerebrospinal fluid (CSF) levels in rTg‐DI models and human CAA patients versus controls to assess diagnostic accuracy of uPA as a biomarker for CAA. Method PLAU gene expression and uPA localization were studied in cerebrovascular tissue of a rTg‐DI models and wild‐type rats, alongside a human sporadic CAA patient and a control subject, using rt‐qPCR and immunohistochemistry respectively. Additionally, CSF Aβ40 and uPA levels were determined in rTg‐DI rats and in human patients with possible or probable CAA (according to the Boston Criteria; n=28), and control subjects (n=40), using ELISA. Result Immunohistochemistry showed strong overexpression of uPA in brain tissue of rTg‐DI models and human CAA patients. Additionally, uPA showed a strong colocalization with Aβ peptides, restricted to the vasculature of rTg‐DI rats and human CAA patients. Expression and localization of both uPA and Aβ peptides were negligible in controls. Additionally, CAA rats displayed significant overexpression of the PLAU gene in brain vasculature as well as significant elevation of uPA in their CSF. In humans, CSF Aβ40 levels were reduced in CAA patients (7.66 ± 3.36 ng/mL) compared to controls (9.58 ± 3.88 ng/mL, p=0.04). CSF uPA levels were higher in CAA patients (median (IQR): 92.0 (76.1 ‐ 109.0) pg/mL compared to human controls (68.5 (56.6 ‐ 78.4) pg/mL; p<0.001). CSF Aβ40 and uPA combined yielded the highest AUC (0.87) to distinguish CAA from controls. Conclusion uPA was overexpressed in rTg‐DI rats and its levels were significantly elevated in both rodent and human CAA. CSF Aβ40 and uPA serve as excellent biomarkers to discriminate CAA from controls, especially when combined.