Esya AD360: A multi‐parametric platform for establishing a novel lysosome‐based biomarker for Alzheimer’s Disease

Abstract

AbstractBackgroundBiomarker development in Alzheimer’s Disease (AD) relied on techniques such as neuroimaging, and proteomics to investigate tauopathy and amyloidosis (Musiek & Holtzman, 2015). However, multiple studies show that lysosomal dysfunction is associated with Alzheimer’s disease (AD) by regulating both the formation and degradation of Aβ (van Weering & Scheper, 2019). To leverage this paradigm shift in our understanding of AD pathology, we describe a method using lysosomal two ion‐mapping (2‐IM) technology in addition to RNA panel for early detection of AD.MethodEsya’s diagnostic (Esya AD360) is a multiparameter platform incorporating 1) 2‐IM probes that ratiometrically measure concentrations of ions such as H+ and Ca2+ with single lysosome resolution 2) ApoE genotyping and, 3) a panel comprising 12 RNA targets identified through RNA‐Seq. The diagnostic utility of Esya AD360 was assessed through application to fibroblasts from a clinically diagnosed cohort comprising healthy individuals and patients diagnosed with AD, frontotemporal dementia (FTD), Parkinson’s, and Huntington’s disease. A model probability score of predicting AD was calculated using a logistic binary regression.ResultThe diagnostic power of Esya AD360 is improved by addition of a custom RNA panel and ApoE genotyping to our pre‐existing 2‐IM model. Logistic regression demonstrates that the assay is >96% accurate in differentiating AD from HI (specificity of >99% and sensitivity of 92.3%). This was further validated using a small validation sample set comprising 12 autopsy confirmed AD samples. Esya AD360 successfully identifies all 12 samples as AD patients, independently validating assay accuracy. After combining the validation data, we find the specificity of Esya AD360 to be >99%, and 96% sensitive. We also demonstrate that Esya AD360 can successfully differentiate between AD and non‐AD dementias (FTD, Huntington’s and Parkinson’s), with an AUC of 0.99 between HI and AD, and an AUC of 0.99 between AD and non‐AD dementias.ConclusionThe incorporation of three independent features has resulted in an accurate, specific, and sensitive assay at identifying AD and differentiating it from healthy and non‐AD dementias. This technology can be further optimised and applied to other neurological disorders and will allow Esya to develop a more accurate platform for personalized medicine in neurology.