Sonic hedgehog signals to multiple prostate stromal stem cells that replenish distinct stromal subtypes during regeneration

Abstract

The adult mouse prostate has a seemingly endless capacity for regeneration, and sonic hedgehog (SHH) signaling has been implicated in this stem cell-driven process. However, it is not clear whether SHH acts on the epithelium or stromal cells that secrete factors required for epithelial expansion. Because little is known about stromal stem cells compared with their epithelial counterparts, we used in vivo mouse genetics tools to characterize four prostate stromal subtypes and their stem cells. Using knockin reporter alleles, we uncovered that SHH signals from prostate basal epithelial cells to adjacent stromal cells. Furthermore, the SHH target gene Gli1 is preferentially expressed in subepithelial fibroblast-like cells, one of four prostate stromal subtypes and the subtype closest to the epithelial source of SHH. Using Genetic Inducible Fate Mapping to mark adult Gli1- or Smooth muscle actin-expressing cells and follow their fate during regeneration, we uncovered that Gli1-expressing cells exhibit long-term self-renewal capacity during multiple rounds of androgen-mediated regeneration after castration-induced involution, and depleted smooth muscle cells are mainly replenished by preexisting smooth muscle cells. Based on our Genetic Inducible Fate Mapping studies, we propose a model where SHH signals to multiple stromal stem cells, which are largely unipotent in vivo.

Keywords: Gli1 expression; genetic fate mapping; mesenchymal lineage analysis; smooth muscle.

Fig. 1.

Shh is expressed by prostate basal cells and signals to stromal cells. Expression of components of the HH signaling pathway was determined in the adult prostate using lacZ knockin mouse lines. (A, C, E, and G) X-GAL staining of DP sections of adult mice showing expression of bGAL in the indicated mouse lines. (B, D, F, and H) FIHC staining for bGAL (red) and CK5 (green). White arrows indicate bGAL+, CK5+ cells. Black arrows indicate bGAL+, CK5− cells. (Scale bars: X-GAL, 50 μm; FIHC, 20 μm.)

Fig. 2.

Gli1 expression is enriched in prostate ductal subepithelial cells. (A) FIHC staining of DP sections for SMA (green) and CD34 (red) showing four stromal subtypes: subepithelial cell (Sub), SMC, wrapping cell (Wrap), and interstitial fibroblast (IF). (B) FIHC staining of Gli1^lacZ/+ DP sections for SMA (green) and bGAL (red) showing that Gli1 is expressed in Sub, SMC, and Wrap. (C and D) Percentage of Gli1+ cells in the three ductal stromal subtypes compared with the normal distribution of each subtype as a percentage of CD34+ cells; 10,159 CD34+ cells (including Gli1+ cells) were counted from three mice. Data are presented as mean ± SEM. **P < 0.01; ***P < 0.001 by ANOVA. (Scale bar: 20 μm.)

Fig. 3.

Gli1-GIFM–labeled dorsal prostate stromal cells expand after one round of I/R. (A) Schematic showing experimental design; 5 d non-I/R and 5 wk non-I/R indicate non-I/R mice that were killed 5 d and 5 wk after being administered Tamoxifen, respectively. (B–E) X-GAL staining of Gli1^CreER/+; R26^lacZ/lacZ DP sections from mice after (B) 5 d non-I/R, (C) 5 wk non-I/R, (D) one round of I/R (I/Rx1), and (E) three rounds of I/R (I/Rx3). (F) Quantification of the percentage of ductal stromal area that is X-GAL+ in Gli1^CreER/+; R26^lacZ/lacZ DPs. *P < 0.05; **P < 0.01 by ANOVA. Data are presented as mean ± SEM. Each data point represents the average of three sections from one animal. (Scale bar: 100 μm.)

Fig. 4.

Prostate SMCs marked by Gli1-GIFM expand in the nonproximal regions after I/R. (A–C) FIHC staining of Gli1^CreER/+; R26^Yfp/Yfp DP sections for SMA (green) and YFP (red). Representative pictures showing (A) YFP+ Sub, (B) SMC, and (C) Wrap marked by Gli1-GIFM. (D and F) Quantification of the number of cells of each YFP+ subtype normalized to the SMA+ area (millimeters squared) in (D) proximal and (F) nonproximal ductal regions in mice after 5 wk non-I/R, mice 14 d after castration (end of I), and I/Rx1 mice. (E and G) Quantification of each YFP+ subtype as a percentage of the total YFP+ cells in (E) proximal and (G) nonproximal ductal regions after 5 wk non-I/R, at the end of I, and after I/Rx1. Data are presented as mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001 by ANOVA; 4,599 YFP+ cells were counted from three mice. Each data point represents one animal.

Fig. 5.

Smooth muscle cells marked by Sma-GIFM persist after six rounds of I/R. X-GAL staining of DP sections of Sma^CreER/+; R26^lacZ/+ mice (A) administered corn oil (control) or after 5 wk (B) non-I/R, (C) I/Rx1, or (D) I/Rx6. TM, Tamoxifen. (Scale bar: 100 μm.)

Fig. 6.

Summary of SHH signaling in the adult prostate and model of stromal regeneration. Schematic showing that Shh and Gli3 are coexpressed in the basal cells (BCs), that Gli1 expression is enriched in Subs, and that Gli1, Gli2, and Gli3 are expressed in all four stromal subtypes: Subs, SMCs, Wraps, and IFs. Curved arrows indicate our proposed model that SHH signals to four lineage-restricted unipotent stromal stem/progenitor cells, three of which are shown. Based on our Sma-GIFM results, SMCs are replenished by preexisting SMCs (red arrow). By extrapolation and the spatial separation of the three subtypes, we propose that Subs and Wraps are replenished by distinct stem/progenitor cells (white arrows). BM, basement membrane; LC, luminal cell.