Sojin Song a 1, Jong Uk Lee a b 1, Myeong Jin Jeon a, Soohyun Kim a, Chan-Nyoung Lee c, Sang Jun Sim a *
aDepartment of Chemical and Biological Engineering, Korea University, Seoul, 02841, Republic of Korea
bDepartment of Chemical Engineering, Sunchon National University, Jeollanam-do, 57922, Republic of Korea
cKorea University Anam Hospital, Seoul, 02841, Republic of Korea
1These authors contributed equally to this work.
*Corresponding author: correspondence to Sang Jun Sim
Alzheimer's disease (AD) is a neurodegenerative disease of complex pathogenesis, with overt symptoms following disease progression. Early AD diagnosis is challenging due to the lack of robust biomarkers and limited patient access to diagnostics via neuroimaging and cerebrospinal fluid (CSF) tests. Exosomes present in body fluids are attracting attention as diagnostic biomarkers that directly reflect neuropathological features within the brain. In particular, exosomal miRNAs (exomiRs) signatures are involved in AD pathogenesis, showing a different expression between patients and the healthy controls (HCs). However, low yield and high homologous nature impede the accuracy and reproducibility of exosome blood-based AD diagnostics. Here, we developed a programmable curved plasmonic nanoarchitecture-based biosensor to analyze exomiRs in clinical serum samples for accurate AD diagnosis. To allow the detection of exomiRs in serum at attomolar levels, nanospaces (e.g., nanocrevice and nanocavity) were introduced into the nanostructures to dramatically increase the spectral sensitivity by adjusting the bending angle of the plasmonic nanostructure through sodium chloride concentration control. The developed biosensor classifies individuals into AD, mild cognitive impairment (MCI) patients, and HCs through profiling and quantifying exomiRs. Furthermore, integrating analysis expression patterns of multiple exosomal biomarkers improved serum-based diagnostic performance (average accuracy of 98.22%). Therefore, precise, highly sensitive serum-derived exosomal biomarker detection-based plasmonic biosensor has a robust capacity to predict the molecular pathologic of neurodegenerative disease, progression of cognitive decline, MCI/AD conversion, as well as early diagnosis and treatment.