Direct reprogramming of fibroblasts into embryonic Sertoli-like cells by defined factors

Abstract

Sertoli cells are considered the "supporting cells" of the testis that play an essential role in sex determination during embryogenesis and in spermatogenesis during adulthood. Their essential roles in male fertility along with their immunosuppressive and neurotrophic properties make them an attractive cell type for therapeutic applications. Here we demonstrate the generation of induced embryonic Sertoli-like cells (ieSCs) by ectopic expression of five transcription factors. We characterize the role of specific transcription factor combinations in the transition from fibroblasts to ieSCs and identify key steps in the process. Initially, transduced fibroblasts underwent a mesenchymal to epithelial transition and then acquired the ability to aggregate, formed tubular-like structures, and expressed embryonic Sertoli-specific markers. These Sertoli-like cells facilitated neuronal differentiation and self-renewal of neural progenitor cells (NPCs), supported the survival of germ cells in culture, and cooperated with endogenous embryonic Sertoli and primordial germ cells in the generation of testicular cords in the fetal gonad.

Figure 1. Nr5a1, Wt1 and Dmrt1 promote proliferation and induce mesenchymal to epithelial transition

(A) Schematic representation of the strategy to test candidates that induce mesenchymal to epithelial transition (MET). (B) Systematic approach to discover the candidates that are responsible for the formation of epithelial foci. In each infection different transcription factor was removed from the pool of the nine factors. The formation of the epithelial foci was examined 1–2 weeks post infection. (C) Immunostaining of Nr5a1, Wt1 and Dmrt1 (red) in MEFs (MEFs^NWD) 3 weeks post infection. (D) Bright field images of MEFs and tail tip fibroblasts (TTFs) from both sexes that were infected with Nr5a1, Wt1 and Dmrt1 and were cultured for 3 weeks. (E) Left-schematic representation of the key factors that block the MET process. Right-A Cluster of genes with known role in MET inhibition that were downregulated in MEFs^NWD, immature Sertoli and mature Sertoli cells. Log2 relative gene expression is visualized as shades of red (higher than MEFs) and shades of green (lower than MEFs) in the indicated cells. (F) qRT-PCR of the indicated genes normalized to the Hprt housekeeping gene in MEFs, ieSCs, immature and mature Sertoli cells. (G) Immunostaining of Vim and Cdh1 (green) in MEFs and MEFs^NWD. (H) Colony formation assay of MEFs and MEFs^NWD. Cells (number as indicated) were cultured for the indicated time periods. The colonies were stained with crystal violet and imaged. (I) FACS analyses demonstrate the percentage of cells that are actively synthesizing DNA after 30 min of BrdU pulse. See also Figures S1 and S2.

Figure 2. Nr5a1, Wt1 and Sox9 induce cell aggregation

(A) Schematic representation of the strategy to test candidates that induce cell aggregation. (B) Systematic approach to discover the candidates that are responsible for cell aggregation. In each infection different factor was removed from the pool of the nine factors. After 3 weeks, cells were seeded as single cells and tested for aggregates 24h later. (C) Formation of cell aggregates in the indicated cells following the introduction of Nr5a1, Wt1 and Sox9 for 3 weeks. (D) qRT-PCR of the indicated genes normalized to the Hprt housekeeping gene in MEFs and MEFs^NWS.

Figure 3. ieSCs are epithelial cells with high proliferative capability

(A) Immunostaining of Nr5a1, Wt1, Dmrt1, Gata4 and Sox9 (red) in MEFs (MEFs^NWDG4S) 3 weeks post infection. (B) Bright field of epithelial focus generated from transduced MEFs and TTFs from both sexes. (C) Bright field of cells from the upper panel after 2 weeks cultivation on matrigel. (D) H2b-GFP ieSCs were co-cultured with 2-day old testis suspension for 48 hr. Cells were imaged using the bright field and GFP channel. White arrow marks endogenous Sertoli cell and green arrow marks ieSC. (E) qRT-PCR of the indicated genes normalized to the Hprt housekeeping gene in MEFs, MEFs^NWD, ieSCs, immature and mature Sertoli cells. (F) Colony forming assay of the indicated cells. 500 cells were seeded and cultured for a week. The colonies were stained with crystal violet and imaged. (G) Proliferation curve of MEFs and ieSCs during 8 days in culture as indicated by cell number (cells were trypsinized and counted on the indicated days). (H) FACS analyses demonstrate the percentage of cells that are actively synthesizing DNA after 30 min of BrdU pulse. (I) qRT-PCR of the indicated genes normalized to the Hprt housekeeping gene in MEFs, ieSCs, immature and mature Sertoli cells. (J) Western blot analysis of the indicated proteins in MEFs, ieSCs, immature and mature Sertoli cells. See also Figures S3 and S4 and movies S1–S4.

Figure 4. ieSCs exhibit normal karyotype and express embryonic Sertoli specific transcription profile

(A) Hierarchical clustering of gene expression array profiles as measured by mouse SurePrint G3 Gene Expression Microarrays-8×60K (Agilent). Complete linkage hierarchical clustering analysis was performed using Pearson’s correlation metric. The dendrogram includes individual samples from MEFs, TTFs, ieSCs (MEFs), ieSCs (TTFs female), 14.5 dpc male gonad, immature and mature Sertoli cells. (B) Hierarchical clustering dendrogram using embryonic Sertoli specific profile (n=200) generated mostly from two independent studies (Bouma et al., 2010; Boyer et al., 2004). Samples and genes were clustered using Cluster 3.0. Shades of red show the pairwise Pearson correlations of gene expression. Clustering tree is shown on top of the heatmap. (C) Scatter plots of the indicated samples (legend on the top) on the principle component 1 (PCA1) and PCA2 axes derived from principle component analysis (PCA) of the samples. (D) Scatter plots of gene expression values (means of the indicated samples) in log2. 2 fold (1 log2) lines were drawn. Y-axis-high genes were drawn in red, while X-axis-high genes were drawn in green. (E and F) qRT-PCR of the indicated genes normalized to the Hprt housekeeping gene in MEFs, 14.5 dpc female gonad, ieSCs (derived from female TTFs), ieSCs (derived from male MEFs), 14.5 dpc male gonad, immature and mature Sertoli cells. (G) Immunostaining of Amh (green) in ieSC culture. (H and I) qRT-PCR of the indicated genes normalized to the Hprt housekeeping gene in MEFs, 14.5 dpc female gonad, ieSCs (derived from female TTFs), ieSCs (derived from male MEFs), 14.5 dpc male gonad, immature and mature Sertoli cells. (J) Cytogenetic analysis of ieSCs shows normal karyotype after 17 passages. See also Figure S4.

Figure 5. ieSCs migrate, aggregate and form tubular-like structures

(A) Migration assay of H2b-GFP MEFs, ieSCs and immature Sertoli cells during 24 hr of incubation using Boyden chamber (8μM pore). 3×10⁵ cells were seeded in the top of the insert in serum-free medium, while serum was placed in the well below. After 24 hr the cells that migrated to the other side of the membrane were trypsinized and counted. Left panel-representative fields of MEFs, ieSCs and immature Sertoli cells that had migrated through the membrane. Right panel-graph depicts the average number of cells that were migrated through the other side of the membrane taken from ten independent fields. (B) Comparison between MEFs and ieSCs in their aggregation potential. High dense single cell suspension (5×10⁶ cells) of MEFs and ieSCs were culture for 3 hours and imaged in the indicated time points. Right-a graph summarizes the average number of aggregates from ten fields from five independent experiments. (C) Representative images of the generated aggregates 24 hr post plating of freshly isolated (24 hrs) H2b-GFP endogenous immature sertoli cells (isolated from 3 day-old mouse testis) and H2b-GFP ieSCs. (D) Immunostaining of Col4a and Lama1 (green) in MEFs and ieSCs. (E, F) Comparison between MEFs, ieSCs, immature and mature Sertoli cells in their capability to form cord-like structures and tubulogenesis when cultured with 10%FBS or 2%FBS medium containing dox on matrigel for 72–96 hr. (G) Comparison between MEFs and ieSCs in their capability to form ring-like structures when seeded sparsely. (H) Spontaneously formed three-dimensional tubule-like structure by ieSCs that was generated in 1–2 weeks culture without passaging in dox containing medium (upper panel) and endogenous seminiferous tubule isolated from 3 week-old mouse testis. The tubules were fixed and stained for Gata4 and Sox9 and subjected to confocal microscopy (I) Generation of tubule-like structure in vivo by ieSCs. H2b-GFP ieSCs or H2b-GFP MEFs were mixed with 5 day-old mouse testis suspension and centrifuged for 5 min to obtain a solid pellet (each pellet contains 2×10⁶ MEFs or ieSCs and one testis). The pellets were then transplanted under the skin of the back of NOD/SCID mice. Three weeks post injection; the transplants were removed, fixed and stained for GFP. Black broken line marks tubule’s borders. Black arrow marks GFP-postive ieSC. White arrow marks endogenous Sertoli cell. Asterisks indicate statistical significance of differences in the mean of the number of positive cells between the indicated groups (*- p-value <0.05, **- p-value <0.01, ***- p-value <0.0001, Mann-Whitney U test). See also Figures S5 and S6.

Figure 6. ieSCs facilitate germ cell survival in vitro

(A) Immunostaining of Ddx4 (red) after 7 days of coculture of testicular suspensions from 1-day-old postnatal pups with MEFs or ieSCs. During the entire experiment the cells were grown in mouse ES medium. The ieSCs-testicular suspension co-culture was grown with or without dox. (B) A graph depicts the average number of Ddx4-positive cells per field in 5 independent fields in the indicated groups. (C) qRT-PCR of Gdnf normalized to the Hprt housekeeping gene in MEFs, testis alone, ieSCs and ieSCs that were co-cultured in a dox-free medium. (D) Immunostaining of Tubb3 (red), Pax6 (green) and Gfap (green) after 8 days of embryonic cortical progenitor differentiation. The cortical progenitor cells received conditioned medium from MEFs, ieSCs or fresh basic differentiating medium on a daily basis. Two representative fields from Tubb3 are depicted showing long and branched axons. (E) A graph depicts the average number of Tubb3-positve cells per field in 5 independent fields. (F) qRT-PCR of the indicated genes normalized to the Hprt housekeeping gene in MEFs and ieSCs. Asterisks indicate statistical significance of differences in the mean of the number of positive cells between the indicated groups (*- p-value <0.05, **- p-value <0.01, ***- p-value <0.0001, Mann-Whitney U test).

Figure 7. ieSCs cooperate with endogenous embryonic Sertoli cells and PGCs in the formation of the testicular cords

(A) Schematic representation of the technique to inject ieSCs into the gonad. Left- H2b-GFP spheres formation after 1 week on matrigel, middle- injection of the spheres into 12.5/13.5 dpc XY gonads (magnification of ieSCs H2b-GFP sphere 24 hr after injection to the gonad), right-culturing for 4 days with doxycycline, then gonad staining (Sox9-red, Ddx4-blue, ieSCs-green), sectioning and confocal microscopy analysis. (B) Representative z-axis confocal scans for two testicular cords 4 days post injection to 13.5 XY gonads. (C) Representative z-axis confocal scans from two testicular cords 4 days post injection into 12.5 XY gonads. Sections with number indicate sequential z-axis confocal scan through one testicular cord. White arrow marks ieSC. (D) A scheme of the developing testis and the various process eSCs undergo during differentiation (modified from (Skinner and Griswold, 2005)). Key factors that were demonstrated to facilitate these processes in vitro are shown. See also Figure S7 and movie S5.