The adult human testis transcriptional cell atlas

Abstract

Human adult spermatogenesis balances spermatogonial stem cell (SSC) self-renewal and differentiation, alongside complex germ cell-niche interactions, to ensure long-term fertility and faithful genome propagation. Here, we performed single-cell RNA sequencing of ~6500 testicular cells from young adults. We found five niche/somatic cell types (Leydig, myoid, Sertoli, endothelial, macrophage), and observed germline-niche interactions and key human-mouse differences. Spermatogenesis, including meiosis, was reconstructed computationally, revealing sequential coding, non-coding, and repeat-element transcriptional signatures. Interestingly, we identified five discrete transcriptional/developmental spermatogonial states, including a novel early SSC state, termed State 0. Epigenetic features and nascent transcription analyses suggested developmental plasticity within spermatogonial States. To understand the origin of State 0, we profiled testicular cells from infants, and identified distinct similarities between adult State 0 and infant SSCs. Overall, our datasets describe key transcriptional and epigenetic signatures of the normal adult human testis, and provide new insights into germ cell developmental transitions and plasticity.

Fig. 1

Single cell transcriptome profiling from healthy adult whole testes. a Schematic illustration of the experimental workflow. b tSNE and clustering analysis of combined single-cell transcriptome data from human testes (n = 6490). Each dot represents a single cell and is colored according to its cluster identity as indicated on the figure key. The 13 cluster identities were assigned based on marker gene expression shown in Fig. 1d and Supplementary information, Fig. S2. tSNE: t-distributed stochastic neighbor embedding. Note: the 40 µm filtering step likely limits capture of the large Sertoli cells. c tSNE plot of single cell transcriptome data with cells colored based on their donors of origin, as indicated on the figure key. d Expression patterns of selected markers projected on the tSNE plot. Red indicates high expression and gray indicates low or no expression, as shown on the figure key. Top row represents somatic/niche cell markers; bottom three rows are representative germ cell markers. For each cell type, we show one marker in the main figures and a gallery in supplementary information, Fig. S2

Fig. 2

Expression patterns of representative genes marking niche cells, and Niche-Germline interactions. a Expression patterns (violin plot) of macrophage-specific genes across the 13 different Clusters (Clusters 1–8 = germ cells; Clusters 9–13 = Niche/somatic cells). b Expression patterns (violin plot) of endothelial cell-specific genes across the different clusters. c Expression patterns (violin plot) of myoid cell-specific genes across the different clusters. d Expression patterns (violin plot) of Sertoli cell specific genes across different clusters. e Expression patterns (violin plot) of Leydig cell specific genes across different clusters. f Relative expression levels of representative genes from different key signaling pathways projected onto the tSNE plot from Fig. 1b. Stage-specific expression is highlighted by blue dotted circles

Fig. 3

Gene expression dynamics during spermatogenesis. a Pseudotime analysis on germ cells (Clusters 1–8). Cluster 1 represents the start of pseudotime, with Cluster 8 at the end. b K-means clustering of genes exhibiting differential expression (n = 8485) across germ cell populations. Note: each row represents a gene, and each column represents a single cell, with columns/cells placed in pseudotime order as defined in Fig. 3a and depicted by a thick colored line (top, color code as in Fig. 3a). Differential gene expression levels utilize a Z score, which represents the variance from the mean, as defined on the color key in the right top corner. c Differentially-expressed genes and associated GO terms (using DAVID v6.7) characteristic of germ cell developmental transitions, based on the 8 germ cell Clusters represented in Fig. 2a. The 5 most significant up-regulated GO terms are annotated in pink boxes, and down-regulated GO terms in green boxes. GO: gene ontology

Fig. 4

Gene Expression Dynamics during Meiosis. a Focused analysis (tSNE, clustering and pseudotime ordering) of the cells from Clusters 3 and 4 (from Figs. 1b and 3a) reveals developmental progression during meiosis I. b Expression patterns of known meiotic markers projected onto the tSNE plot. Red indicates high expression and gray indicates low or no expression, key on figure. c K-means clustering of genes exhibiting differential expression (n = 4594) during meiosis I. Note: each row represents a gene, and each column represents a single cell, with columns/cells placed in pseudotime order as defined in a and depicted by a thick colored line (top, color code as in a). Gene expression levels utilize a Z score, which depicts variance from the mean, as defined on the color key in the right top corner. d Expression levels of representative genes during meiosis progression. x-axis represents pseudotime (as defined on a), and y-axis represents gene expression levels. e Expression patterns of key transcription factors during meiosis, with their expression projected onto the tSNE plot (a)

Fig. 5

Identification of five discrete transcriptional states for SSCs. a Focused analysis (tSNE, clustering and pseudotime ordering) of Clusters 1 and 2 (from Fig. 1b and 3a) reveals five discrete cellular states (States 0 to 4) during SSC development. b K-means clustering (k = 6) of genes exhibiting differential gene expression in States 0–4. Six gene clusters (termed S1-S6) were identified. Gene ontology associated with each gene block is shown on the right. Note: each row represents a gene, and each column represents a single cell, with columns/cells placed in pseudotime order (depicted by different colors on the top of the figure) as defined in a. Gene expression levels utilize a Z score, which depicts variance from the mean, as defined on the color key in the right top corner. c Relative expression levels of selected SSC markers projected on the tSNE plot represented in a. d Violin plots representing the expression levels of the selected markers shown in c in States 0–4 (x-axis). e Relative expression levels of selective State 0-specific markers projected on the tSNE plot represented in a. f Violin plots representing the expression levels of selective State 0-specific markers shown in e in States 0–4 (x-axis)

Fig. 6

Computational and molecular examination of spermatogonial plasticity. a Visualization of the RNA velocity analysis results on the tSNE plot of SSCs (see main text for details on vectors). b Heatmap and hierarchical clustering of ATAC-seq data from KIT⁺ spermatogonia (two replicates), SSEA4⁺ SSCs (four replicates) and ESCs (two replicates). Note: SSEA4⁺ SSC and ESC data are from ref. c Schematic summarizing the combinatorial gene expression programs and cellular events promoting five distinct SSC states (States 0–4) and depiction of the proposed spermatogonial dynamics/kinetics and behavioral plasticity of States with main cellular events and molecular pathways. Dotted arrows are speculative

Fig. 7

Single cell RNA profiling from infant testis and comparison to adult scRNA-seq data. a tSNE and clustering analysis of single-cell transcriptome from infant testis (n = 1340). b Expression patterns of representative markers to help assign cell identities. c tSNE and pseudotime analysis of infant germ cells and adult spermatogonia. d Expression patterns (violin plot) of representative genes in infant germ cells and adult spermatogonia

Fig. 8

RNA and protein staining to validate state 0. a Sequential RNA FISH of SSC markers on tubular sections. Two SSCs are highlighted as representative examples. Blue is the DAPI signal; red detects RNA FISH signal for gene as indicated on the figure. White dashed line circles represent the cell boundaries. Scale bar: 5 μm. b Expression patterns of FGFR3 (marks State 0 and 1), TSPAN33 (marks State 0) and SSEA4 (marks State 1) in testicular cells via flow cytometry. Human single testicular cells were used for staining with the markers indicated; non-stained cells were used as control for gating purposes. Left: FGFR3⁺ cells were identified and used for analysis in the middle and right panels. Middle: co-staining pattern between FGFR3 (x-axis) and TSPAN33 (y-axis), with blue as TSPAN33⁺ and red and gray as TSPAN33⁻. Right: co-staining pattern between FGFR3 (x-axis) and SSEA4 (y-axis), with black as SSEA4⁺ and gray as SSEA4⁻. c Immunolocalization of UTF1 (State 0 marker, in blue), GFRA1 (State 1 marker, in red) and FGFR3 or TCF3 or MKI67 (in green). Each combination of single or triple antigen (named in green on the left side) is represented by 5 panels. Scale bar: 10 μm. d Immunolocalization of UTF1 (State 0 marker, in blue), GFRA1 (State 1 marker, in red) and 4 new candidate markers (in green). Each antigen (named in green on the left side) is represented by 5 panels. Scale bar: 10 μm