Elevated expression of urokinase plasminogen activator in rodent models and patients with cerebral amyloid angiopathy

Aims The aim of this work is to study the association of urokinase plasminogen activator (uPA) with development and progression of cerebral amyloid angiopathy (CAA). Materials and methods We studied the expression of uPA mRNA by quantitative polymerase chain reaction (qPCR) and co‐localisation of uPA with amyloid‐β (Aβ) using immunohistochemistry in the cerebral vasculature of rTg‐DI rats compared with wild‐type (WT) rats and in a sporadic CAA (sCAA) patient and control subject using immunohistochemistry. Cerebrospinal fluid (CSF) uPA levels were measured in rTg‐DI and WT rats and in two separate cohorts of sCAA and Dutch‐type hereditary CAA (D‐CAA) patients and controls, using enzyme‐linked immunosorbent assays (ELISA). Results The presence of uPA was clearly detected in the cerebral vasculature of rTg‐DI rats and an sCAA patient but not in WT rats or a non‐CAA human control. uPA expression was highly co‐localised with microvascular Aβ deposits. In rTg‐DI rats, uPA mRNA expression was highly elevated at 3 months of age (coinciding with the emergence of microvascular Aβ deposition) and sustained up to 12 months of age (with severe microvascular CAA deposition) compared with WT rats. CSF uPA levels were elevated in rTg‐DI rats compared with WT rats (p = 0.03), and in sCAA patients compared with controls (after adjustment for age of subjects, p = 0.05 and p = 0.03). No differences in CSF uPA levels were found between asymptomatic and symptomatic D‐CAA patients and their respective controls (after age‐adjustment, p = 0.09 and p = 0.44). Increased cerebrovascular expression of uPA in CAA correlates with increased quantities of CSF uPA in rTg‐DI rats and human CAA patients, suggesting that uPA could serve as a biomarker for CAA. Rodent models and patients suffering from cerebral amyloid angiopathy (CAA) show overexpression of urokinase plasminogen activator (uPA) in cerebrovascular tissue, which furthermore strongly co‐localises with vascular amyloid β. uPA levels in cerebrospinal fluid appear elevated in sporadic CAA patients, but not in hereditary Dutch‐CAA patients, after analysis in two separate cohorts. uPA expression appears to be related to presence of CAA pathology and therefore should be considered as a mechanistic influence in CAA pathophysiology and potential biomarker for sporadic CAA.