Fig. 5: Sex-dependent differences in aging mouse HSCs according to scRNA-seq analysis.

a Main clusters in the aggregate of all samples (3- and 28-month-old males and females). Overall, 8 clusters were assigned to the aggregates of all the samples, with the majority of cells forming continuous clusters, except for cluster 6, which was a minor cluster distal to the main SP population and likely more differentiated cell contaminants. Cluster 6 was removed from the view but can be found in Supplementary Fig. 5. b, c Split view of the main SP clusters in separate samples according to sex and age (b) and the percentage of each cluster (c). YM, young male; OM, old male; YF, young female; OF, old female. d, e Cluster 2 was reclustered into two subclusters (d) that exhibited different distributions in the aggregated UMAP (e). Clusters were assigned to HSPCs according to key gene expression features as follows: cluster 0 for LT-HSCs that were negative for the surface markers CD48 and Flt3; cluster 1 for ST-HSCs that closely shared many signature genes with cluster 0 but had an increase in Flt3 and CD34 expression; cluster 2 had stronger activation of Flt3 and CD34; the lower part of cluster 2 (mostly subcluster 0) was assigned to MPP and the upper part (mostly subcluster 1) was assigned to CMP; clusters 3 and 4 were assigned to MkP and MEP, respectively, as both enriched for platelet genes, including Itga2b and Gp1bb; however, cluster 4 was less cycling and had lower Flt3; cluster 5 for GMP that was enriched for the marker genes Mpo, Ctsg, Elane, Prtn3 and Ms4a3; and cluster 7 for common lymphoid progenitors (CLPs) that had a UMAP position close to clusters 0 and 1 but carried its signature genes regulating B and T lymphoid cell development, including Erg1, Fos, Nr4a1, Nr4a2 and Nfkbia. In contrast, the MPP and CMP cells were less well defined by gene signatures.