Identifying cell type-specific autonomous and non-autonomous interactions in AD

ABSTRACT A central question in Alzheimer's Disease (AD) is the role of cell type-specific changes associated with the onset and progression of the disease. Genome-wide association studies have identified genes linked to AD, and recent large-scale transcriptomic and proteomic studies of post-mortem brain tissue have highlighted candidate genes and proteins with significant associations to pathological and clinical manifestations of AD. Together, these genetic and molecular studies have implicated neurons, astrocytes, and microglia as key players in AD pathogenesis and cognitive decline. However, these genetic and post-mortem profiling studies cannot resolve the degree to which a change in cell type expression profile is cell autonomous, as opposed to being a downstream effect of changes in other cell types. To tackle this question, we propose an innovative mixed in vitro co-culture approach to understand the effects of cell types from donors with variations in AD pathology and/or AD dementia on cell types derived from individuals with or without cognitive decline or pathology. We use induced pluripotent stem cell (iPSC) lines from members of the Religious Order Study/Memory and Aging Project (ROSMAP), for whom we have detailed pathological, clinical, whole- genome sequencing, and post-mortem molecular data. Our preliminary data on monocultures from these donor lines shows a high degree of concordance between in vitro and post-mortem transcriptomic, proteomic, and proteinopathic data. This allows us to anchor our in vitro findings directly to patient phenotypes and analyses of post-mortem brain tissue data. Based on this anchoring, our approach is to “mix and match” cell type-specific cultures from different iPSC lines, thereby allowing us to disentangle which cell type-specific changes in vitro are likely cell type autonomous, and which are induced non-autonomously by cells from a donor with a different genetic background, pathology, and ante-mortem cognitive trajectory. Importantly, we look for conserved, polygenic AD-associated signals in vitro that are present across donors with varied genetic backgrounds, as opposed to directly characterizing the effect of a specific variant on cell type. Given the highly multifactorial nature of AD genetic risk variants, we believe this approach allows for robust identification of convergent cell type-specific effects associated with the disease.