Understanding the human brain's intricate structure and function has been a long-standing goal in neuroscience. Early insights, such as the case of Phineas Gage, highlighted the importance of specific brain regions for particular functions. Advances in single-cell technologies have revolutionized our ability to characterize brain cell types and their functions at an unprecedented level of detail. However, most existing single-cell studies are limited in scope, often focusing on a single brain region, a small number of cells, or a specific disease. This limitation hinders our understanding of the overall heterogeneity of brain cells across different regions and developmental stages, and in both health and disease. The Human Cell Atlas (HCA) initiative and other consortia have generated extensive single-cell datasets, but integrating these diverse datasets remains a major challenge. This paper addresses this challenge by creating the Brain Cell Atlas, a unified resource that integrates data from multiple studies to provide a more complete picture of brain cell diversity and heterogeneity.

The authors reviewed the existing literature on single-cell studies of the human brain. They noted that while several studies have profiled specific brain regions or cell types, a comprehensive, integrated atlas encompassing multiple regions, developmental stages, and disease states was lacking. They cited several key studies that provided foundational data but also highlighted the limitations of these individual studies in capturing the full complexity of brain cell heterogeneity. These studies, while valuable on their own, often lacked the scale and scope to fully explore rare cell types or regional variations. The authors emphasized the potential of integrating multiple datasets to uncover novel insights into brain development, function, and disease.

The Brain Cell Atlas was constructed by integrating data from 70 human and 103 mouse single-cell RNA sequencing (scRNA-seq) and single-nucleus RNA sequencing (snRNA-seq) studies. Data were obtained from various public repositories, including GEO, UCSC Genome Browser, ArrayExpress, Allen Brain Map, and Synapse. The data were meticulously curated, ensuring consistency in naming conventions and metadata. Quality control steps were implemented to remove low-quality cells and doublets. The researchers then used a combination of seven established reference-based machine learning methods and a novel hierarchical annotation workflow (scAnnot) to achieve a consensus cell type annotation across the entire dataset. scAnnot, based on the Variational Autoencoder model from scANVI, allowed for multigranularity cell type annotation, overcoming the limitations of existing reference-based methods. Data integration was performed using scANVI to infer a shared latent space, minimizing batch effects. The Silhouette score was used to evaluate the quality of data integration. To identify putative neural progenitor cells (NPCs) in the adult human hippocampus, the researchers integrated data from multiple studies across different developmental stages and used a combination of marker gene expression, trajectory inference, and cross-species comparison with mouse data. To identify and characterize a novel PCDH9high microglia population, they integrated data from multiple studies focusing on the prefrontal cortex and hippocampus. Differential gene expression analysis, gene ontology (GO) enrichment analysis, and gene set enrichment analysis (GSEA) were used to characterize the differences between this subpopulation and other microglia. Finally, CellChat was used to explore cell-cell communication networks within neuronal niches.

The Brain Cell Atlas comprises 11.3 million human cells and 15 million mouse cells from various brain regions. The human data include cells from adult, fetal, organoid, and tumor samples, covering a wide age range (6 gestational weeks to over 80 years old) and various disease states, including Alzheimer's disease, epilepsy, and gliomas. The atlas identified putative neural progenitor cells (NPCs) in the adult human hippocampus, a finding that has been debated in the literature. The researchers used several lines of evidence to support this finding, including marker gene expression, trajectory inference (pseudotime and RNA velocity), and cross-species comparison with mouse data. Furthermore, immunostaining confirmed the presence of proliferating NPCs in the adult human hippocampus. The atlas also identified a novel population of microglia with high PCDH9 expression. This PCDH9high microglia subpopulation showed distinct gene expression patterns compared to other microglia, suggesting a specialized role. The study found that this subpopulation is enriched for immune-related genes and exhibits regional heterogeneity, with different gene expression profiles in the hippocampus and prefrontal cortex. The researchers further investigated the cell-cell communication network of this subpopulation, revealing specific interactions with other cell types in both regions, particularly glutamatergic neurons in the hippocampus.

The findings of this study significantly advance our understanding of brain cell diversity and heterogeneity. The identification of putative NPCs in the adult human hippocampus provides valuable insights into adult neurogenesis and has important implications for the development of therapies for neurological injuries and neurodegenerative diseases. The characterization of the PCDH9high microglia subpopulation highlights the functional complexity of microglia and suggests that they may play specialized roles in different brain regions and in response to different stimuli. The comprehensive nature of the Brain Cell Atlas allows for a deeper exploration of regional heterogeneity and cell-cell communication networks, which are crucial for understanding brain function and disease pathogenesis.

The Brain Cell Atlas is a comprehensive resource that significantly advances our understanding of brain cell types and functions. The identification of putative NPCs in the adult human hippocampus and the characterization of PCDH9high microglia represent significant contributions. Future research could leverage this atlas to investigate the roles of these and other cell types in brain development, aging, and disease, ultimately leading to a more complete understanding of brain function and the development of novel therapies for brain disorders.

While the Brain Cell Atlas represents a substantial advancement in our understanding of brain cell diversity, some limitations exist. Batch effects, despite efforts to mitigate them, could still influence the results. The reliance on publicly available data may limit the ability to control for all experimental variables. Furthermore, the identification of putative NPCs requires further validation using independent methods. Finally, functional studies are needed to fully elucidate the roles of the identified cell types, particularly the PCDH9high microglia.