Neurofilaments in neurodegenerative diseases - Focus on axonal injury cerebrospinal fluid biomarkers

Abstract

Neurofilaments are structural proteins from the intermediate filament family type IV that comprises neurofilament light, medium, heavy (NfL, NfM, NfH), and α-internexin (AINX). Under pathological conditions, neurofilaments are released into the surrounding fluids and are associated with the neurodegenerative process in many neurological diseases, including Alzheimer’s disease (AD), frontotemporal dementia (FTD), Huntington’s disease (HD), multiple sclerosis (MS), and corticobasal degeneration (CBD). In these diseases, NfL has so far been the most studied and relevant neurofilament, however, AINX has a unique trait of being a central nervous system (CNS)-specific protein. This provides a unique opportunity to complement one of the flaws of NfL as a biomarker, i.e., that it represents neurodegeneration irrespective of its origin. The work in this thesis pursued deeper understanding of neurofilament proteins by developing novel assays for detection of oligomeric forms of NfL (oNfL) and CNS-specific AINX. In Papers I and II we aimed to investigate if cerebrospinal fluid (CSF) NfL was detected by immunoassays in oligomeric or monomeric form. Therefore, in Paper I we developed a CSF NfL ELISA for detection of oNfL, using the same antibody for both capture and detection in the sandwich setup. Results showed that in both CSF and the assay calibrator, detection was consistent with dimerized NfL. Also, in two clinical cohorts (controls, AD, and FTD subtypes), direct comparison between oNfL and “regular” NfL concentrations, suggested that “regular” NfL immunoassays might mostly detect oNfL. In Paper II, through western blot and immunoprecipitation experiments, we confirmed the presence of oNfL in CSF. By subjecting CSF pools to reducing conditions, we identified 3 to 4-fold lower molecular weight NfL species than with CSF in non-reducing conditions, indicating that oligomerization is dependent on disulfide bonds. In Papers III-V, we intended to initiate the investigation of AINX in CSF across neurological diseases and compare its performance with NfL, starting with the development and validation of a novel immunoassay for CSF AINX on the ultrasensitive Single molecule array (Simoa) platform in Paper III. Biomarker potential was assessed in clinical cohorts of AD and diverse neurological diseases (HD, CBD, MS, dementia with Lewy bodies, and semantic dementia). Here, AINX and NfL showed similar levels across diseases and a significant positive correlation. In Paper IV, AINX was quantified in a clinical cohort of MS. Again, AINX correlated well with NfL and markers of inflammation and was significantly increased in MS patients compared with healthy controls. In Paper V, we found AINX increases in AD and FTD, and biomarker correlations showed significance between AINX, NfL and typical AD biomarkers within the AD group. In the FTD group, correlations were only seen with NfL, suggesting that AINX is released into CSF independently of the typical AD biomarkers and could therefore reflect the neurodegenerative processes of both diseases. In conclusion, the outcomes of this thesis were two novel and robust neurofilament assays and insights from diverse clinical cohorts. These findings highlight the importance of full characterization of NfL and the potential for AINX to be a CNS-specific axonal injury biomarker.