Acute phase markers in CSF reveal inflammatory changes in Alzheimer’s disease that are impacted by APOE ε4, sex and age but not pathology

Abstract

AbstractBackgroundNeuroinflammation may accelerate Alzheimer’s disease (AD) progression. We profiled acute phase response (APR) proteins (stimulated by IL‐6 released from pro‐inflammatory glia) in CSF of people along the clinical and pathology (CSF t‐tau/Aβ42) spectrum of AD to investigate the association between disease outcomes and inflammatory changes.MethodsAPR proteins in CSF from 920 subjects in the Swedish BioFINDER study were quantitively analyzed using multiple reaction modelling mass spectrometry: α1‐antitrypsin, α1‐antichymotrypsin, ceruloplasmin, complement C3, α‐fibrinogen, β‐fibrinogen, γ‐fibrinogen, haptoglobin, and hemopexin. Ferritin was measured using multiplex ELISA. CSF t‐tau and Aβ42 was measured using Euroimmun ELISA and the cohort was dichotomized for AD pathology using the t‐tau/Aβ42 ratio. Subjects were also stratified by the following clinical diagnoses: cognitively normal (CN); subjective cognitive decline (SCD); mild cognitive impairment (MCI); and AD dementia. N per group was: CN t‐tau/Aβ42‐ve=335; CN t‐tau/Aβ42+ve=104; SCD t‐tau/Aβ42‐ve=105; CN t‐tau/Aβ42+ve=44; MCI t‐tau/Aβ42‐ve=78; MCI t‐tau/Aβ42+ve=133; AD t‐tau/Aβ42+ve=121.ResultsIn separate multiple regressions (adjusting for diagnosis, age, sex, APOE‐ε4), CSF t‐tau/Aβ42 status did not affect APR protein levels in CN and SCD subjects (Figure 1). Rather, the APR elevated with symptomatic progression: MCI t‐tau/Aβ42+ve subjects had 4/10 proteins elevated compared to CN tau/Aβ42‐ve subjects (α1 antichymotrypsin, ferritin, α‐fibrinogen, β‐fibrinogen; 0.001<P<0.027), and all 10 proteins were elevated in AD (1.6x10‐8<P<0.017). Surprisingly, the APR was more robustly activated in people with cognitive changes but with low t‐tau/Aβ42 levels (7/10 and 5/10 proteins elevated in SCD and MCI respectively). APOE ε4 status did not affect levels of any APR proteins in CSF, while these were elevated in females and with increased age. The performance of the APR in predicting clinical diagnosis was greater in females, in younger subjects, and those without APOE ε4 (Figure 2).ConclusionsThe APR in CSF reveals inflammatory changes that occur with clinical and not pathological dimensions of AD. The APR has a complex interaction with clinical variables: the extent of the APR is affected by sex and age, not APOE ε4, yet the APR has more influence in non‐ε4 carriers, younger people and females (Figure 3). These data provide new insight into neuroinflammation in AD and how this varies according to demographics.