A First-In-Class Antibody Enabling Detection of Altered Peripheral Non-Phosphorylated Clusterin With Translational...

Clinical diagnosis of dementia, with Alzheimer's disease (AD) as the major form, relies heavily on memory tests and the caregiver descriptions, both of which are subjective. The discovery of reliable biomarkers may facilitate objective diagnosis. The protein clusterin (CLU), encoded by a well-established AD risk gene, is consistently elevated in AD patients, but its biomarker utility is limited by high interindividual variability. As CLU is produced in most organs and tissues, quantifying CLU secreted specifically from the brain into the bloodstream may help the development of new diagnostic methods. CLU in blood can be phosphorylated at T393-S394 and/or S396, but these phosphorylations are absent in brain parenchyma. Peripherally administered non-phosphorylated CLU has been shown to reduce neuroinflammation and AD pathology in mouse models. A monoclonal antibody, 3D3F10, targeting non-phosphorylated CLU at T393-S394 or S396 was generated. This antibody showed an ability to distinguish serum samples of dementia and non-dementia (p < 0.001, AUC = 0.897, sensitivity: 81.8%, specificity: 95.5%, Youden index: 0.77), but did not distinguish Parkinson's disease. The direction of the change contradicted our initial hypothesis, and further analyses suggested that phospho-CLU in dementia patients is unlikely to originate from the brain. These results established 3D3F10 as a novel tool for modification-specific CLU detection, indicated a potential of non-phospho-CLU as a biomarker for dementia, and peripheral phospho-CLU might play a role in pathogenesis.