Post by Stephanie Williams
What's the science?
Hyperactive neurons have been shown to contribute to the accumulation and spread of proteins (tau and amyloid-beta) known to aggregate in the brain in Alzheimer’s disease (AD). Hyperactivity in the hippocampus in particular is correlated with amyloid accumulation and has been found in asymptomatic individuals at risk for developing AD. Recently, contradictory evidence has arisen from previous studies investigating the link between tau accumulation and hippocampal activity. As a result, the relationship between hyperactivity in the hippocampus and the proteins that aggregate in AD remains unclear. This week in the Journal of Neuroscience, Huijbers and colleagues investigated the link between memory-related brain activity in older adults and tau and amyloid accumulation.
How did they do it?
The authors used neuroimaging to measure neural activity and protein accumulation related to AD in a large sample of cognitively normal, older adults. Specifically, they measured 1) brain activity using functional magnetic resonance imaging (fMRI), 2) levels of amyloid-beta using positron emission tomography (PET) and 3) tau protein accumulation using PET. The authors measured tau accumulation in specific brain regions known to be vulnerable to molecular changes in AD — the entorhinal cortex and inferior temporal cortex. During the fMRI session, participants were instructed to remember faces they saw on a screen. Participants saw a series of unfamiliar faces (the encoding period), which they needed to remember. After a brief delay period, another set of faces appeared on the screen; these faces were either new, famous faces or previously presented faces. Participants reported whether they recognized the face by pressing a button. The authors recorded the participants’ response time and used their responses to track which of the faces had been successfully encoded.
The authors assessed the correlation between the three brain measures (fMRI activity, levels of tau, and levels of neocortical amyloid) and different behavioral measurements (eg. hit-rate, false alarm rate, response times) from the facial recognition task. They assessed which brain areas were engaged in successful encoding of the faces. To characterize the amount of amyloid in the neocortex, the authors first compared thirty subjects with the highest amount of amyloid and thirty subjects with the lowest amount of amyloid, to determine regional differences. Tau PET data was analyzed in a similar manner— comparisons of brain regions between a group of thirty adults with the highest amount of tau and thirty adults with the lowest amount of tau were performed to create maps of the pattern of tau accumulation. The authors then assessed the relationship between the three sets of data (fMRI, amyloid PET, tau PET) to investigate whether hippocampal activity was associated with protein accumulation. They fit linear models to the data to confirm the relationship between tau and encoding success.
What did they find?
The authors confirmed the results of a recent study that had suggested a link between tau accumulation and increased hippocampal activity. Using the fMRI data, the authors identified brain areas (visual cortex, fusiform gyrus, parahippocampus and hippocampus) that were engaged during successful encoding while performing the face recognition task, as well as areas that were engaged during unsuccessful encoding (posteriomedial cortex, anterior cingulate cortex, angular gyrus, lateral temporal cortex). They found that increased activity was not associated with better encoding success. Using the PET data, the authors identified several areas where the amount of amyloid was different between the high (overall amyloid) and low (overall amyloid) groups (anterior cingulate cortex, angular gyrus, lateral temporal regions). Similarly, they found that there was a significant difference in the amount of tau in the high and low groups in the temporal lobe, which was consistent with previous findings.
The authors observed that increased hippocampal activity was related to tau accumulation in the inferior temporal cortex, but not with amyloid levels in the neocortex, or tau accumulation in entorhinal cortex. This was an important finding, as tau is found in the entorhinal cortex during advanced aging, while tau spreads to the inferior temporal cortex in AD. This suggests that hippocampal hyperactivity is related to the spread of tau pathology from the entorhinal cortex to the neocortex, prior to clinical symptoms of AD. The authors confirmed the relationship between tau and hippocampal activity by fitting a general linear model to characterize the relationship between fMRI activity and tau accumulation. They found that tau and amyloid accumulation had opposite effects on hippocampal activity.
What's the impact?
Using a large sample size, the authors confirmed hippocampal hyperactivity begins in preclinical stages and suggest it is associated with tau accumulation. They showed a novel, opposing relationship between amyloid and tau relative to hippocampal activity during a task. This study highlights an important unanswered question in AD research — whether increased hippocampal activity is reflective of abnormal processing or plays a compensatory role. These findings work to resolve the seemingly contradictory evidence from previous studies in which some normal adults were found to have normal hippocampal activity while adults in other studies were found to have hippocampal hyperactivity.
Huijbers et al. Tau accumulation in clinically normal older adults is associated with hippocampal hyperactivity. The Journal of Neuroscience (2018). Access the original scientific publication here.