Sertoli cell-only phenotype and scRNA-seq define PRAMEF12 as a factor essential for spermatogenesis in mice

Abstract

Spermatogonial stem cells (SSCs) have the dual capacity to self-renew and differentiate into progenitor spermatogonia that develop into mature spermatozoa. Here, we document that preferentially expressed antigen of melanoma family member 12 (PRAMEF12) plays a key role in maintenance of the spermatogenic lineage. In male mice, genetic ablation of Pramef12 arrests spermatogenesis and results in sterility which can be rescued by transgenic expression of Pramef12. Pramef12 deficiency globally decreases expression of spermatogenic-related genes, and single-cell transcriptional analysis of post-natal male germline cells identifies four spermatogonial states. In the absence of Pramef12 expression, there are fewer spermatogonial stem cells which exhibit lower expression of SSC maintenance-related genes and are defective in their ability to differentiate. The disruption of the first wave of spermatogenesis in juvenile mice results in agametic seminiferous tubules. These observations mimic a Sertoli cell-only syndrome in humans and may have translational implications for reproductive medicine.

Fig. 1

Pramef12 is essential for spermatogenesis and male fertility. a The hierarchy of different cell types of spermatogonia during SSC self-renewal and differentiation. The A_s spermatogonia are heterogeneous with SSCs and A_s progenitors. The A_s and A_pr progenitors have the potential to become SSCs. A_s spermatogonia produce chains of A_pr A_al(4), A_al(8), and A_al(16) undifferentiated spermatogonia that are connected by cytoplasmic bridges and are precursors of differentiated spermatogonia. b Fertility of > 3 pairs of Pramef12^Null and Pramef12^Het male and female mice mated 1:1. Mean litter sizes ± s.d. are shown with indicated genotypes. c Testes of P90 adult Pramef12^Null and Pramef12^Het mice. Scale bar, 1 mm. d Ratios of testis to body weight of Pramef12^Null and Pramef12^Het mice shown in c. Mean ± s.d, n = 6 biologically independent testes from six different animals. *P = 1.44E-08 by two-tailed Student’s t test. e Adult testis sections from Pramef12^Null and Pramef12^Het mice stained with periodic acid-Schiff (PAS) and hematoxylin. Pramef12^Null testes are agametic with a Sertoli cell-only phenotype. Scale bar, 50 μm. f Immunofluorescence of P90 adult testes from Pramef12^Null and Pramef12^Het mice after co-staining with antibodies to DDX4 (germ cells) and WT1 (Sertoli cells) as well as Hoechst 33342 (DNA). Scale bar, 50 μm. g Immunohistochemistry of P90 adult testes from Pramef12^Null and Pramef12^Het mice after staining with antibodies to cyclin D1 and PLZF. Arrowheads indicate positive-staining spermatogonia. Scale bar, 50 μm. h Quantification of cyclin D1-positive and PLZF-positive spermatogonia in Pramef12^Null and Pramef12^Het testes. Mean ± s.d, n = 3 biologically independent testes from three different animals. *P = 1.44E-28 (cyclin D1) and *P = 3.19E-26 (PLZF) by two-tailed Student’s t test. i Same as e, but of cauda epididymides. Representative of n = 6 c, n = 3 e, f, g, i independent biological replicates with similar results per condition

Fig. 2

Pramef12 is required for the first wave of spermatogenesis. a Morphology of testes from P14, P21, and P35 Pramef12^Null and Pramef12^Het mice. Scale bar, 1 mm. b Comparison of testes weight of P7, P14, P21, and P35 Pramef12^Null and Pramef12^Het mice. Mean ± s.d, n = 5 biologically independent testes from five different animals. *P = 3.81E-03 (P14), *P = 2.38E-06 (P21) and *P = 4.84E-08 (P35) by two-tailed Student’s t test. Immunohistochemistry c and immunofluorescence d staining of DDX4 in the testicular sections from P2 and P7 Pramef12^Null and Pramef12^Het mice. The merge in d is with Hoechst 33342 stained DNA. Scale bar, 50 μm. e Quantification of DDX4-positive cells per tubules in Pramef12^Null and Pramef12^Het testes at P2 and P7. Mean ± s.d, n = 3 biologically independent testes from three different animals. *P = 2.71E-07 by two-tailed Student’s t test. f Testicular sections of P10 and P14 Pramef12^Null and Pramef12^Het mice were stained with periodic acid-Schiff (PAS) and hematoxylin. Scale bar, 50 μm. g Immunofluorescence staining of γH2AX in testicular sections from P14 Pramef12^Null and Pramef12^Het mice. DNA was stained with Hoechst 33342. Arrows, pachytene spermatocytes (XY body); asterisks, Sertoli cell-only tubules. Scale bar, 50 μm. h, i PAS staining of P21 and P35 testes in Pramef12^Null and Pramef12^Het mice. The reduction in germ cell number becomes notable with the appearance of more agametic tubules (asterisks). Spermatogonia, 1; spermatocytes, 2; spermatids, 3; and elongated spermatozoa, 4; asterisks, agametic tubules. Scale bar, 50 μm. j Same as i, but of cauda epididymides. k Percentage of agametic tubules in Pramef12^Null testes from P7 to P35. Pramef12^Het control testes contain no empty tubules. Mean ± s.d, n = 3 biologically independent testes from 3 different animals at each age point. Representative of n = 5 a, n = 3 c, d, f–j independent biological replicates with similar results per condition

Fig. 3

Further evaluation of spermatogenic phenotype of Pramef12^Null mice. a Testicular sections of P48 Pramef12^Null and Pramef12^Het mice were stained with periodic acid-Schiff (PAS) and hematoxylin. Asterisks, agametic tubules. Scale bar, 50 μm. b Same as a, but of P60 testes. c Same as a, but of P75 testes. d Statistical analysis of empty tubules from P48 to P90 Pramef12^Null mice. Mean ± s.d, n = 3 biologically independent testes from three different animals at each age point. Representative of n = 3 a–c independent biological replicates with similar results per condition

Fig. 4

PRAMEF12 is specifically expressed in spermatogonia. a–c Abundance of Pramef12 transcripts in different adult tissues and P7 testes a, b and E17.5-P56 testes c was determined by RT-PCR a or qRT-PCR b, c using β-actin as a control. Molecular mass, left in a. The highest expression level of Pramef12 relative to β-actin was set to 1 in b, c. Mean ± s.d, n = 3 biologically independent samples per condition. d–f Analyses of the abundance of Pramef12 transcripts in 11 testicular cell types d, e and in cluster 1 of re-clustered germ cells f based on previously reported scRNA-seq data in adult mouse testes. Numbers in d and the Y axis in e reflect the averaged UMIs detected in each individual cell. After re-clustering germ cells, Id4, Plzf, and Sall4 serve as positive controls and, like Sall4, Pramef12 was primarily enriched in the spermatogonial cluster 1, which corresponds to undifferentiated spermatogonia f. P value in f was obtained from the published scRNA-seq data. Avg diff, log₂-scale fold change. Percent of cells expressing gene in cluster 1 (Pct.1) or in non-cluster 1 cells (Pct.2). Exp and Exp + reflect mean expression of the marker in all cells of this cluster and in the marker-positive cells of this cluster, respectively. g Schematic representation of transgene expressing ^HAPramef12^mCherry cDNA driven by the Pramef12 promoter (3.2 kb). h Whole-mount staining of P2, P7 and adult Pramef12^HA/mCherry testes with antibodies to mCherry (left), PLZF (middle) and merged (right). Scale bar, 50 μm. Examples of A_s, A_pr, and A_al undifferentiated spermatogonia in adult testes (dotted circles) and enlarged (right panels). Arrowheads, high-light cytoplasmic location. Scale bar, 20 μm. i Same as h for P7 and adult testes, but with antibodies to mCherry (left), KIT (middle) and merged (right). Arrows, high-light co-localization of the two markers. j The expression profile of intracellular molecular markers for male germline cells. Except for NANOS2 (underlined), PRAMEF12 (underlined, red) is unique in its cell type specificity and is largely enriched in undifferentiated spermatogonia. Representative of n = 3 a, h, i independent biological replicates with similar results per condition

Fig. 5

Essential role of PRAMEF12 in early spermatogenic maintenance. a Immunofluorescence of whole-mount testes from P2, P4, and P7 Pramef12^Null (left group) and Pramef12^Het (right group) mice after staining with antibodies to DDX4 (left) or PLZF (middle) and merged (right). The number of PLZF-positive spermatogonia was comparable in P2 Pramef12^Null and Pramef12^Het tubules and significantly decreased in Pramef12^Null tubules as early as at P4, although still present at P7. Scale bar, 50 μm. b Same as a, but with antibodies to GFRA1 (co-receptor for GDNF) at P4. Asterisks, agametic tubules. c Same as a, but with antibodies to KI67 (marker of mitosis) and PLZF using whole-mount testes from Pramef12^Null (top) and Pramef12^Het (bottom) mice at P7. d The percent of KI67⁺; PLZF⁺ double-positive among the total PLZF⁺ single-positive spermatogonia in P7 Pramef12^Null and Pramef12^Het tubules. At least six random regions of individual testis were counted from at least three different mice. Mean ± s.d, n = 3 biologically independent samples per condition. *P = 5.89E-23 by two-tailed Student’s t test. Representative of n = 3 a–c independent biological replicates with similar results per condition

Fig. 6

Essential role of PRAMEF12 in spermatogonial differentiation. a Immunofluorescence of whole-mount testes from P2, P4, and P7 Pramef12^Null and Pramef12^Het mice after staining with antibodies to DDX4 (left) or KIT (middle) and merged (right). KIT-positive spermatogonia were rarely detected at P2 in either Pramef12^Null or Pramef12^Het tubules but were present at P4 and increased at P7 in Pramef12^Het tubules. Few KIT-positive spermatogonia (arrow) were observed in P7 Pramef12^Null tubules. Scale bar, 50 μm. b Immunohistochemistry of sections from P7 Pramef12^Null (top panels) and Pramef12^Het (bottom panels) testes after staining with antibodies to DDX4 (left) or KIT (middle) and merged with Hoechst 33342 to stain DNA (right). Arrowheads, KIT-positive cells in Pramef12^Null testes. Scale bar, 50 μm. c Quantitative real-time RT-PCR of Kit mRNA abundance in P7 Pramef12^Null and Pramef12^Het testes using β-actin as an internal load control and setting the abundance in Pramef12^Het testis to 1. Mean ± s.d, n = 3 biologically independent samples per condition. *P = 1.43E-05 by two-tailed Student’s t test. d Immunoblot of KIT protein in P7 Pramef12^Null and Pramef12^Het testes using α-tubulin as a load control. e Same as a, but with antibodies to SOHLH1, SOHLH2, and PLZF. The number of spermatogonia positive for SOHLH1 and SOHLH2 was significantly reduced in P7 Pramef12^Null tubules (top panels). Representative of n = 3 a, b, d, e independent biological replicates with similar results per condition

Fig. 7

Pramef12 deficiency alters expression patterns of spermatogenic genes. a MA plot (log ratio RNA abundance versus abundance) of RNA-seq data from Pramef12^Null and Pramef12^Het testes at P2. 11 genes and 70 genes were upregulated and downregulated, respectively, in Pramef12^Null testes using adjusted P < 0.1 as the cut off. b Gene ontology of downregulated genes in P2 Pramef12^Null testes. Genes that support stem cell population maintenance indicated in red text fly-out. GO terms were obtained by GOrilla analysis. P value was obtained from the data by GOrilla analysis. c Genes related to stem cell maintenance and differentiation that were significantly downregulated (log₂-fold change) in P2 Pramef12^Null testes. d Same as a, but at P7. In all, 1599 genes and 2147 genes were upregulated and downregulated in Pramef12^Null testes, respectively. e Same as b, but at P7. P value was obtained from the data by GOrilla analysis. f Venn diagram depicting the overlap of downregulated genes determined at two developmental time points, P2 and P7. g RNA-seq results of selected transcripts (log₂-fold change) related to SSC maintenance, differentiation and spermatogenesis that were significantly downregulated in P7 Pramef12^Null testes. h Quantitative RT-PCR validation of downregulated genes involved in SSC maintenance, differentiation and spermatogenesis in Pramef12^Null testes at P7. For comparison, the abundance (relative to β-actin) of each gene in Pramef12^Het control mice was set to 1. Data are presented as mean ± s.d for n = 3 biologically independent samples per condition

Fig. 8

Single-cell transcriptome of Pramef12^Het and Pramef12^Null testicular cells. a Schematic illustration of the workflow for scRNA-seq analysis. b tSNE and clustering analysis of combined single-cell transcriptome data from P7 Pramef12^Het and Pramef12^Null testicular cells. Each dot represents a single-cell and cell clusters are distinguished by colors. c Dotplot for expression of selected marker genes across all identified cell types. d Gene expression patterns of selected somatic cell marker genes projected on the tSNE plots. e Expression patterns of marker genes for germ cells visualized in tSNE plots. Pramef12 is exclusively expressed in germ cell population but not in somatic cell populations

Fig. 9

Transcriptome-wide signatures of germline stem cell development. a Focused analysis (UMAP clustering and pseudotime ordering) of combined (left), Pramef12^Het (middle), and Pramef12^Null (right) germ cells documented four biological subtypes following the order of SPG1 to SPG4 (state 1 to state 4). b Gene expression patterns of selected marker genes corresponding to each cellular state on the UMAP plots. c Summary schematic depicting the percentage of spermatogonia in each cellular state in Pramef12^Null and Pramef12^Het testes. Pramef12 deficiency significantly impairs germline stem cell development as determined by the decreased percentage of differentiated cells in SPG3 and SPG4. d Differentially expressed genes (DEGs) between Pramef12^Null and Pramef12^Het cells in SPG2, SPG3, and SPG4. Log₂FC, log₂-fold change. Null and Het indicate the percentage of cells in which the gene was detected in Pramef12^Null and Pramef12^Het samples, respectively. e UMAP plots of the expression patterns of selected DEGs in SPG2 subtype in Pramef12^Null and Pramef12^Het samples

Fig. 10

Spermatogonial stem cell fates in Pramef12^Null male mice. a UMAP clustering of the combined germ cells including P0, P3, P6, P7 Pramef12^Het, and Pramef12^Null samples documented six biological subtypes following a developmental trajectory of SPG1 to SPG6 (upper left). Five additional UMAP plots specific to each individual sample (right and below). b Gene expression patterns of selected spermatogonial marker genes corresponding to each cellular state on the UMAP plots: SPG1 (Id4, Etv5, Ret); SPG2 (Id4, Etv5, Lhx1, Ret); SPG3 (Id4, Etv5, Lhx1, Ret); SPG4 (Id4, Etv5, Lhx1, Ret); SPG5 (Neurog3, Sox3); and SPG6 (Kit, Stra8). c Dotplot representation of average scaled expression and the percentage of cells within each cluster for selected marker genes across all ages and cell types. d Table of the distribution and the percentage of the cells within each cluster and each sample