Studies of postmortem brain tissue have yielded a wealth of information about the pathways and proteins associated with cognitive decline in late-onset Alzheimer’s disease (AD). However, the chain of events can’t be established definitively from tissue recovered at the end stage of the disease.
Tracy Young-Pearse, PhD, associate chair of Neuroscience Research in the Department of Neurology at Brigham and Women’s Hospital, Valentina Lagomarsino, a PhD Candidate, Richard V. Pearse II, PhD, senior scientist, and colleagues have created an experimental system that fills the gap between these autopsy studies and genetic studies. They’ve shown that neurons derived from induced pluripotent stem cells (iPSCs) reflect molecular mechanisms underlying late-onset AD—and could be helpful in evaluating current and potential therapies. Their report appears in Neuron.
Comparing iPSC-derived Neurons and Brain Tissue
The researchers had access to cryopreserved peripheral blood mononuclear cells—along with rich clinical, pathologic and genomic/proteomic detail—on participants in the Religious Order Study (ROS) and Rush Memory and Aging Project (MAP), two longitudinal cohort studies of aging and AD.
They selected 53 deceased individuals who spanned the clinical and neuropathological spectrum of aging. Sixteen had clinical and neuropathological diagnoses of AD and Alzheimer’s dementia, 17 had a pathologic diagnosis of AD but no clinical diagnosis, and 20 had neither a clinical nor pathologic diagnosis of AD.
The researchers derived iPSC lines from the blood cells, then reprogrammed them into functional induced neurons (iNs). They looked for concordance between the RNA sequencing expression profiles and proteomic profiles of the iNs and the frontal cortex of the individual from whom they were derived. All brain-to-iN gene correlations were positive and statistically significant.
Amyloid-Beta and Tau
Levels of amyloid-beta (Aβ) and tau measured in iNs were correlated with the level of Aβ plaque burden and tau tangle burden in the brain of the same individual.
Furthermore, specific forms of Aβ and tau present in iNs were significantly associated with the individual’s rate of cognitive decline. The strongest association was the ratio of Aβ42 (the longest and most amyloidogenic AD-related peptide) to Aβ37 (the shortest and least amyloidogenic).
Remarkably, the cognitive trajectory was more strongly related to Aβ measurements in iNs than to Aβ plaque burden or tau tangle burden.
Polygenic Risk Score
No individual genetic variant for late-onset AD was significantly associated with Aβ42/ Aβ37 or tau in iNs. Therefore, the researchers calculated a polygenic risk score for each donor based on recent large genome-wide association studies of late-onset AD.
The score was significantly higher in individuals with an AD diagnosis, both in the iPSC cohort and in the larger ROS/MAP cohort. After exclusion of the APOE gene, which is the strongest single genetic risk factor for AD[PT1] , the polygenic risk score remained associated with the Aβ42/Aβ37 ratio in iNs.
Thus, that Aβ ratio is affected at least in part by numerous genetic variants of small effect size. These data regarding the polygenic nature of Ad suggests that there are many different genetic “roads” that may lead to AD. Thus, treatments customized to these different genetic causes of late-onset AD may be much more effective than current strategies for treatment.
Protein Phosphatase 1
A high polygenic risk score also correlated with low levels of certain proteins, including protein phosphatase 1 (PP1), which dephosphorylates tau and is low in the brains of patients with AD. As another example of the potential of this new experimental system for studying AD, the researchers investigated PPI activity in iNs.
Pharmacologic reduction of Aβ42/Aβ37 in iNs increased PP1 levels, while the elevation of Aβ42/Aβ37 reduced the levels of PP1. Inhibiting PPI had no effect on the Aβ42/Aβ37 ratio, but it decreased tau phosphorylation. It seems PP1 is both a driver downstream of AD genetic risk and a mechanistic link between AD-associated Aβ and tau.
This unique set of iPSC lines may ultimately be useful across many central nervous system cell types for investigating the mechanisms of neuropathology and cognitive decline across individuals. These studies are ongoing at the Brigham.
The data in this study suggests the intriguing possibility that Aβ37 has a direct protective role in neutralizing longer Aβ peptides. If future research is confirmatory, this observation will have important implications for designing immunotherapeutic agents that target Aβ—and evaluating which patients will respond to this type of therapy.