Oxidative stress biomarkers and longitudinal changes in human brain imaging across the Alzheimer’s disease continuum

Abstract

AbstractBackgroundOxidative stress is critical for Alzheimer’s disease (AD) pathophysiology. Brain mitochondrial function, proteins, and lipids are susceptible to oxidative damage caused by reactive oxygen species. In turn, this damage leads to key AD‐like changes in the brain, such as impaired glucose metabolism (i.e., lower fluorodeoxyglucose, or FDG‐PET) and gray matter volume atrophy in AD‐sensitive regions. It is unknown if oxidative stress biomarkers like superoxide dismutase (SOD) proteins track these changes using longitudinal brain imaging data, and what mechanisms may be at work (Figure 1).MethodUsing Alzheimer’s Disease Neuroimaging Initiative (ADNI) data, 192 aged adults (75 unimpaired, 48 MCI, 69 AD) had annual CSF and FDG data available for over 6 years. A two‐phase MRM targeted proteomics panel was used to gauge CSF SOD1 levels from serum. Standard assays in CSF determined Aβ, p‐tau‐181, and total tau. Brain glucose uptake was determined using FDG SUVR calibrated to the pons. Brain volume was examined longitudinally using FreeSurfer.ResultLongitudinally, among SOD1 quartiles (Figure 2), the lowest SOD1 group showed marked hypometabolism among older adults at baseline (R2=0.528, p<.001, “blue line”), but 6 years later instead predicted hypermetabolism (R2=0.389, p<.001). The highest SOD1 group showed no association at baseline but at 6 years later predicted hypometabolism among older adults (R2=0.151, p<.05, “dotted red line”). CSF Aβ and p‐tau‐181 fully mediated these effects (ps<.01), such that lower Aβ and higher p‐tau‐181 over time resulted in higher SOD1 predicting less glucose uptake. Similarly, higher SOD1 was not related to atrophy in AD‐sensitive regions like parahippocampus at baseline, but 5 years later was strongly related to more brain atrophy (R2=0.207,p<.001) (Figure 3), which was mediated by higher CSF total tau (p<.001).ConclusionLike previous reports, we found that SOD1 initially showed neutral or ameliorative patterns, reflecting its classical role in reducing oxidative stress. With increased Aβ or p‐tau‐181 deposition over time, however, SOD1 became a marker of the degree of oxidative damage that may have occurred in neural tissue due to Aβ or tau.