henhaobio
Sign in

Cell ReportsDevelopmental BiologyPeer-reviewed journal article

Deciphering cell lineage specification during male sex determination with single-cell RNA sequencing

Isabelle StévantiD, Yasmine NeirijnckiD, Christelle BoreliD, Jessica EscoffieriD, Lee B. SmithiD, Stylianos E. AntonarakisiD, Emmanouil T. DermitzakisiD, Serge NefiD

PDFDOIPreprint
In-depth full-text digest available · Jump to digest ↓

Abstract

The gonad is a unique biological system for studying cell fate decisions. However, major questions remain regarding the identity of somatic progenitor cells and the transcriptional events driving cell differentiation. Using time course single cell RNA sequencing on XY mouse gonads during sex determination, we identified a single population of somatic progenitor cells prior sex determination. A subset of these progenitors differentiate into Sertoli cells, a process characterized by a highly dynamic genetic program consisting of sequential waves of gene expression. Another subset of multipotent cells maintains their progenitor state but undergo significant transcriptional changes that restrict their competence towards a steroidogenic fate required for the differentiation of fetal Leydig cells. These results question the dogma of the existence of two distinct somatic cell lineages at the onset of sex determination and propose a new model of lineage specification from a unique progenitor cell population.

In-depth summary

Claude Opus 5.5 · xhigh effort · 8 Nov 2026 · from the full text of version 1 (18 Sept 2017) · CC BY 4.0

Time-course single-cell RNA-seq of NR5A1-GFP+ mouse XY gonad cells points to one multipotent somatic progenitor pool that yields Sertoli cells and, progressively, steroidogenic Leydig precursors.

It was unclear whether Sertoli and fetal Leydig cells come from one progenitor or from two predefined supporting and steroidogenic lineages, because precursor-specific markers were lacking and earlier time-course transcriptomes relied on pooled, non-synchronous sorted cells. Profiling single NR5A1+ cells across sex determination resolves precursor heterogeneity and orders cells in pseudotime to catalogue lineage-specific expression dynamics.

Key results

  1. ShownUnsupervised clustering resolved six somatic populations. All E10.5 cells fell into one transcriptionally homogeneous progenitor cluster that persists to E13.5, with no separate supporting and steroidogenic precursor groups.

    400 of 435 libraries passed QC (median 5628 genes per cell). HCPC gave six clusters with 2802 DE genes (q<0.05). E10.5 cells occurred only in C2 (183 cells, E10.5 to E13.5). Other clusters: interstitial progenitors 106, pre-Sertoli 31, Sertoli 70, fetal Leydig 7 and endothelial 3 cells. · Somatic cell purification and single-cell…; Single-cell transcriptomics identifies six…

  2. SuggestedPseudotime reconstruction traced the interstitial progenitor, Sertoli and fetal Leydig lineages to a shared E10.5 origin. Progenitor-lineage genes rise gradually and are often already expressed, while Sertoli genes switch on abruptly as progenitor genes go down.

    Diffusion map plus Slingshot gave three lineages rooted in E10.5 cells. Of interstitial progenitor-enriched genes, 45% (162 of 357) were already expressed in common progenitors (e.g., Gpc3, Sfrp1), vs 22% (75 of 340) of Sertoli-enriched genes (e.g., Wt1, Bex4, Lsr). The remaining Sertoli genes rose sharply and strongly right at differentiation (P13 to P15). · Pseudotime reconstruction identifies cell…

  3. ShownAlong the progenitor trajectory, stem-cell and Sry-cofactor genes fade after E11.5 while markers of fetal Leydig precursors build up from E12.5. This fits a gradual narrowing toward a steroidogenic fate.

    1734 pseudotime-dependent genes (q<0.05) fell into nine profiles. P2 to P4 (E10.5 to E11.5) include Sall1, Lin28a, Trim71, Foxp1, Wt1 and Cbx2. P6 to P9 (from E12.5) include Arx, Pdgfra, Gli2, Tcf21 and Ptch1, with GO terms for extracellular matrix, angiogenesis and cell migration. · Progenitor cells gradually acquire a…; Discussion

  4. ShownSertoli differentiation runs as successive waves of transcription: short, transient gene peaks at onset are followed by lasting activation of factors that may maintain Sertoli identity.

    2319 pseudotime-dependent genes (q<0.05) fell into 13 profiles. 1217 genes (P7 to P13) mark differentiation from E11.5, including Sry, Sox9, Amh and Dhh. 221 genes (P8) follow the narrow Sry peak (e.g., Kdm3a, Nr0b1). P9, P10 and P13 show sustained activation (Nr5a1, Dmrt1, Gata4). · Sertoli cell differentiation program is…; Discussion

  5. ShownThe Tg(Nr5a1-GFP) reporter reliably labels gonadal somatic cells and not germ cells, which supports the sorting strategy.

    GFP and endogenous Nr5a1 transcript levels correlated at r=0.82 (Spearman). At E16.5, GFP co-localized with NR2F2, HSD3B and SOX9 but not with the germ cell marker DDX4. Across all cells, 16 459 protein-coding genes were detected (RPKM>0). · Somatic cell purification and single-cell…; Figure legends

Methods, data and reproducibility

Nr5a1-GFP+ somatic cells were FACS-sorted from XY Tg(Nr5a1-GFP) mouse gonads at five stages (E10.5 to E16.5) and profiled on the Fluidigm C1 with Nextera XT libraries. Cells were clustered (HCPC on principal components of variable genes), ordered along a diffusion-map/Slingshot pseudotime, and tested for pseudotime-dependent expression with Monocle2.

  • FACS-sorted NR5A1-GFP+ somatic cells from XY Tg(Nr5a1-GFP) embryonic gonads at E10.5, E11.5, E12.5, E13.5 and E16.5, profiled by Fluidigm C1 scRNA-seq · mouse · n = 435 libraries sequenced; 400 cells passed QC
  • Co-immunofluorescence of GFP with NR2F2, HSD3B, SOX9 and DDX4 on E16.5 Tg(Nr5a1-GFP) gonad sections (confocal imaging) · mouse
  • Mouse Genome Informatics (MGI) reproductive phenotype annotations used to flag transcription factors in the Sertoli expression profiles · mouse
Methods & tools
Fluidigm C1 Autoprep System, Illumina Nextera XT DNA Sample Preparation kit, Fluorescence-activated cell sorting (FACS), HCPC (hierarchical clustering on principal components), t-SNE, Monocle2, Destiny (diffusion maps), Slingshot, k-means clustering of gene expression profiles, GO enrichment analysis, ZEN software (confocal imaging), Mouse Genome Informatics (MGI)
Reagents & environment
R version 3.4.0. HCPC, Monocle2, Destiny and Slingshot were used, but their package versions are not stated.
Experimental conditions
Run at the Vital-IT high-performance computing centre of the SIB Swiss Institute of Bioinformatics (http://www.vital-it.ch). Hardware and runtime are not specified.
Reproducibility gaps
The text gives no sequencing data accession and no code repository. QC thresholds and analysis parameters are left to supplemental procedures that are not included here.
Limitations (5)
  • Only 7 fetal Leydig cells were captured, too few to reconstruct their differentiation program at high resolution. · stated by the authors
  • Additional rare NR5A1+ cell types cannot be ruled out, and the absence of lineage-specific fates before sex determination is supported only at the transcriptional level. · stated by the authors
  • What drives the choice between Sertoli and progenitor fate is unresolved. The proposed repression program (for example Sox11, Sfrp1) is speculative. · stated by the authors
  • Lineage relationships come from pseudotime inference on 400 Fluidigm C1-captured cells, with no in vivo lineage tracing or perturbation. The proposed loss of Sertoli competence through Wt1 and Cbx2 downregulation is therefore correlative. · noted in this summary
  • Only XY gonads were analyzed (E10.5 XX controls were excluded), and animals were pooled across litters. So it is untested whether the shared progenitor model holds in XX gonads or varies between individual embryos. · noted in this summary

Read this if you study gonadal sex determination or the origins of Sertoli and Leydig cells and want a single-cell, pseudotime-resolved catalogue of gene expression in mouse fetal testis somatic cells.

Citations
169
FWCI
12.88
Field percentile
top 1%

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISStévant '18Lardenois '23Mayère '21Liao '18Chen '26Mayère '22
85% semantic matchDevelopmental Biology
Isabelle Stévant, Françoise Kühne, Andy Greenfield+3 more·bioRxiv·Sep 7, 2018· 5 citations