A cross-disease human microglial framework identifies disease-enriched subsets and tool compounds for microglial polarization

Abstract

Human microglia play a pivotal role in neurological diseases, but few targeted therapies that directly modulate microglial state or function exist due to an incomplete understanding of microglial heterogeneity. We use single-cell RNA sequencing to profile live human microglia from autopsies or surgical resections across diverse neurological diseases and central nervous system regions. We observe a central divide between oxidative and heterocyclic metabolism and identify subsets associated with antigen presentation, motility, and proliferation. Specific subsets are enriched in susceptibility genes for neurodegenerative diseases or the disease-associated microglial signature. We validate subtypes in situ with an RNAscope-immunofluorescence pipeline and leverage our dataset as a classification resource, finding that iPSC model systems recapitulate substantial in vivo heterogeneity. Finally, we identify and validate candidates for chemically inducing subtype-specific states in vitro, showing that Camptothecin downregulates the transcriptional signature of disease-enriched subsets and upregulates a signature previously shown to be depleted in Alzheimer's.

Benchmarking single-cell hashtag oligo demultiplexing methods

“…Each demultiplexing method (including the genetic demultiplexing) can return one of 10 assignments for a cell: singlet, corresponding to one of the eight unique samples; doublet; or negative. We compare six HTO demultiplexing methods: BFF (Boggy et al 2022); deMULTIplex (McGinnis et al 2019); demuxmix (Tuddenham et al 2022); GMM-Demux (Xin et al 2020); hashedDrops (Lun, Riesenfeld, et al 2019) and HTODemux (Stoeckius et al 2018). BFF has two modes, BFF raw and BFF cluster , and we present the output of both.…”

“…We explore the effect of varying the confidence threshold set in Figure S5 in the supplementary materials. demuxmix demuxmix [15] is similar to GMM-Demux but uses a negative binomial mixture model on the untransformed HTO counts, rather than a mixed Gaussian on the CLR-transformed counts. For each HTO, all cells are clustered into positive and negative clusters using 𝑘-means clustering.…”

Cellular dynamics across aged human brains uncover a multicellular cascade leading to Alzheimer’s disease

“…6). Previous snRNA-seq studies have associated cells and pathways to AD [4][5][6][7][8][9][10][11][12][13][14][15] , but were limited to a case-control study design and thus could neither infer cellular dynamics nor decouple AD and alternative aging. Because we use a sample of the older population and a prospective cohort study design, we leverage the full diversity of the older brain, allowing us to reconstitute the different trajectories of brain aging in a data-driven manner by applying BEYOND.…”

“…However, bulk profiling of the brain parenchyma loses much of the information embedded in the intricate cellular architecture of the brain structures affected by AD. Single-cell and single-nucleus RNA-seq [4][5][6][7][8][9][10][11][12][13][14][15][16] have begun to offer a different perspective, underscoring the cellular perturbations that are part of the sequence of events leading to AD 17 . Currently, it appears that [4][5][6][7][8][9][10][11][12][13][14][15] : (1) changes in expression programs of multiple cell types, including glial, neuronal and vascular cells, are associated with AD, (2) within each cell type only specific subpopulations are likely to be actively involved in AD pathogenesis, and (3) AD-related expression programs changes are coordinated across multicellular communities of various cell subpopulations 4 .…”

Towards elucidating disease-relevant states of neurons and glia by CRISPR-based functional genomics

“…In both of these studies, follow-up of hits from the primary screens with CROP-seq, a technique for combining CRISPRbased gene perturbation with single-cell RNA-seq [64], unveiled distinct cell states that overlapped with diseaseassociated cell states and, also importantly, genes and cellular pathways that drive these cell states. For example, in Dräger et al [89], CROP-seq uncovered a range of cell states in the hiPSC-derived microglia at baseline, which align with microglial states observed in microglia isolated from human brains [94] and mouse models of AD [95]. These states include a state marked by high expression of interferon-responsive genes, a state marked by high expression of chemokines, and a state marked by expression of SPP1 [89].…”