An integrated single-nucleus and spatial transcriptomics atlas reveals the molecular landscape of the human hippocampus

Abstract

The hippocampus contains many unique cell types, which serve the structure's specialized functions, including learning, memory and cognition. These cells have distinct spatial topography, morphology, physiology, and connectivity, highlighting the need for transcriptome-wide profiling strategies that retain cytoarchitectural organization. Here, we generated spatially-resolved transcriptomics (SRT) and single-nucleus RNA-sequencing (snRNA-seq) data from adjacent tissue sections of the anterior human hippocampus across ten adult neurotypical donors. We defined molecular profiles for hippocampal cell types and spatial domains. Using non-negative matrix factorization and transfer learning, we integrated these data to define gene expression patterns within the snRNA-seq data and infer the expression of these patterns in the SRT data. Using this approach, we leveraged existing rodent datasets that feature information on circuit connectivity and neural activity induction to make predictions about axonal projection targets and likelihood of ensemble recruitment in spatially-defined cellular populations of the human hippocampus. Finally, we integrated genome-wide association studies with transcriptomic data to identify enrichment of genetic components for neurodevelopmental, neuropsychiatric, and neurodegenerative disorders across cell types, spatial domains, and gene expression patterns of the human hippocampus. To make this comprehensive molecular atlas accessible to the scientific community, both raw and processed data are freely available, including through interactive web applications.

Spatiotemporal analysis of gene expression in the human dentate gyrus reveals age-associated changes in cellular maturation and neuroinflammation

“…Nmf patterns generated from one dataset can predict the presence and distribution of these same latent factors in another dataset via transfer learning 58 . This strategy was recently used to integrate snRNA-seq and Visium data from adult neurotypical donors in HPC 59 . Specifically, 100 nmf patterns were identified in snRNA-seq data from which we identified three patterns (nmf26, nmf5, and nmf14) that were enriched in GC populations with a sequential overlap of weight gradients where nmf26 partially overlaps with nmf5, which partially overlaps with nmf14 ( Figure 4a ).…”

“…Specifically, 100 nmf patterns were identified in snRNA-seq data from which we identified three patterns (nmf26, nmf5, and nmf14) that were enriched in GC populations with a sequential overlap of weight gradients where nmf26 partially overlaps with nmf5, which partially overlaps with nmf14 ( Figure 4a ). Projecting these three patterns onto paired SRT data from the same donors 59 revealed that, although all three patterns were restricted to the GCL ( Figure 4b ), nmf26 weights were more sparsely distributed. We thus compared the top 10 marker genes for these three patterns to better understand the biological relevance ( Figure 4c ).…”

“…( a ) UMAP of human adult HPC snRNA-seq dataset 59 with superfine cell class annotations and nuclei are colored by the weight of three nmf patterns associated selectively with the GC superfine cell class. RGB scale converted to represent all three nmf patterns simultaneously.…”

“…Overlapping gradients of weights from different nmf patterns have a blended color for a given nucleus. (b) Representative spot plots from the adult HPC SRT dataset 59 . Spot outlines are colored black to delineate GCL, CA3&4, and CA1, other spatial domains are outlined in grey.…”

“…As depicted in the UMAP, nmf26 (infant-associated) and nmf14 (non-infant-associated) are enriched in different GC clusters (GC.3 and GC.4, respectively) ( Figure 4a ). To identify genes that distinguish between these two cell types, we used the pseudobulked DE results taken from Nelson et al, 2024 59 which calculated DEGs by comparing one cluster to all others. We then compared GC.3 and GC.4 to identify gene uniquely enriched in either GC.3 or GC.4 ( Figure S24a ).…”