A potential role of progestin-induced laminin-5/α6-integrin signaling in the formation of side branches in the mammary gland

Abstract

Mammary organoids from adult mice produce tubules, analogous to mammary ducts in vivo, in response to hepatocyte growth factor (HGF) when cultured in collagen gels. The combination of HGF plus progestin (R5020) causes reduced tubule number and length. We hypothesized that the inhibitory effect on tubulogenesis was due to progestin-mediated alteration of HGF/c-Met signaling. Using molecular inhibitors and short hairpin RNA, it was determined that HGF activation of Ras-related C3 botulinum toxin substrate (Rac1) was required for the formation of cytoplasmic extensions, the first step of tubulogenesis, and that Rac1 activity was Src kinase (Src) and focal adhesion kinase (FAK) dependent. The highly novel finding was that R5020 reduced tubulogenesis by up-regulating and increasing extracellular laminin and α6-integrin ligation to reduce activation of the Src, focal adhesion kinase, and Rac1 pathway. Receptor activator of nuclear factor-κB ligand, another progesterone-induced paracrine factor, did not replicate this effect of R5020. The inhibitory effect of R5020 on tubulogenesis was likely mediated through progesterone receptor (PR) isoform A (PRA), because PRA is the predominant PR isoform expressed in the organoids, and the progestin-induced effect was prevented by the PR antagonist RU486. These results provide a plausible mechanism that explains progestin/PRA-mediated blunting of HGF-induced tubulogenesis in vitro and is proposed to be relevant to progesterone/PRA-induced side-branching in vivo during pregnancy.

Fig. 1.

Time course of morphological changes induced by HGF. Organoids plated in Col I gels were treated with BM control (B) or HGF (H) (50 ng/ml) for up to 24 h. Panel A, Representative phase-contrast micrographs of organoids at 0, 4, 12, and 24 h. White arrowheads indicate extensions, white arrows indicate chains, and black arrowheads indicate tubules. Scale bar, 0.05 mm. Panel B, Quantitation of extensions, chains, and tubules at 4 or 24 h treatment. a–c, P < 0.05 that HGF treatment increased indicated structures compared with BM (n = 10). Panel C, Analysis of active Rac1 at 4 h treatment with HGF. Rac1-GTP levels were analyzed by Rac1-GTP pull-down and immunoblot. For densitometry, Rac1-GTP was normalized to total Rac1, and fold change relative to BM was calculated. *, P < 0.05, HGF Rac1-GTP greater than BM Rac1-GTP (n = 4).

Fig. 2.

Rac1 shRNA blocks HGF-induced tubulogenesis. Organoids were infected with control virus (GFP⁺ vector and GFP⁺ scramble) or virus carrying shRNA against Rac1 (GFP⁺ Rac1 shRNA1 and GFP⁺ Rac1 shRNA2) and plated in Col I gels. At 48 h after infection, organoids were treated with HGF for 24 h. A, Representative phase-contrast and GFP overlay images of organoids at 4 and 24 h. White arrowheads indicate extensions, white arrows indicate chains, and black arrowheads indicate tubules. Insets are enlargements of GFP⁺ and GFP⁻ extensions. B, Quantitation of extensions, chains, and tubules at 4 or 24 h. a–c, P < 0.05 that the numbers of indicated structures are reduced by Rac1 shRNA compared with vector control. A total of 30–50 organoids were analyzed per treatment from two separate experiments. C, Immunoblot of total Rac1 at 4 h after treatment with BM or HGF. For densitometry, Rac1 was normalized to total Erk1/2.

Fig. 3.

Src or FAK inhibition blocks HGF-induced tubulogenesis. Organoids plated in Col I gels were treated with BM (B) or HGF (H) in the presence or absence of the pp60-c Src inhibitor 2 (PP2; 20 μm) or the FAK inhibitor (FAKi-14; 10 μm) for 24 h. Panel A, Representative phase-contrast micrographs of organoids at 24 h. White arrow indicates a chain, and black arrowhead indicates a tubule. Scale bar, 0.05 mm. Panel B, Quantitation of extensions, chains, and tubules after 4 and 24 h treatment with the Src or FAK inhibitors. a–f, P < 0.05 that Src or FAK inhibitors decreased the numbers of indicated structures compared with controls (n = 3). Panel C, Rac1-GTP levels were analyzed by Rac1-GTP pull-down and immunoblot. Phospho-Src levels were analyzed by immunoblot; phospho-FAK levels were analyzed by immunoprecipitation and immunoblot. For densitometry, Rac1-GTP was normalized to total Rac1, and fold change relative to BM was calculated. *, P < 0.05, HGF Rac1-GTP greater than BM Rac1-GTP; #, P < 0.05 that Src inhibitor or FAK inhibitor reduced HGF Rac1-GTP (n = 3).

Fig. 4.

R5020 reduces HGF-induced tubulogenesis. Organoids plated in Col I gels were treated with BM (B), HGF (H), or HGF+R5020 (HR) for 4 h. Panel A, Representative phase-contrast micrographs of organoids at 4 h. White arrowhead indicates extensions. Scale bar, 0.05 mm. Panel B, Quantitation of extensions, chains, and tubules after 4 h treatment. a, P < 0.05 that HGF+R5020 reduced the numbers of indicated structures compared with HGF (n = 7). Panel C, The effect of R5020 on HGF-induced Rac1 activation. Rac1-GTP levels at 4 h were determined by Rac1-GTP pull-down and immunoblot. For densitometry, Rac1-GTP was normalized to total Rac1, and fold change relative to BM was calculated. *, P < 0.05, HGF Rac1-GTP greater than BM Rac1-GTP; #, P < 0.05 that R5020 reduced HGF Rac1-GTP (n = 3). Panel D, Representative immunoblots of the R5020 effect on phospho-Src and phospho-FAK 4 h after treatment.

Fig. 5.

R5020 increases Lmγ2 expression in mammary organoids, and Lm reduces HGF-induced Rac1-GTP. A, Organoids were cultured as a monolayer on Col I or Lm-1 and treated with BM or HGF for 4 h. Rac1-GTP levels were analyzed by Rac1-GTP pull-down and immunoblot. For densitometry, Rac1-GTP was normalized to total Rac1, and fold change relative to BM was calculated. B, Confocal immunofluorescence images of organoids plated on Col I gels at 4 h after in situ double-antibody labeling with anti-cytokeratin 18 (CK18) (green) and anti-Lmγ2 (red). Note increased intensity of Lmγ2 (red) staining in R5020- and HGF+R5020-treated organoids. HGF+R5020 treatment increased Lmγ2 (red) staining that was localized to blunted tubules (white arrows). There was a loss of staining upon RU486 addition to HGF+R5020-treated organoids. Scale bar, 0.05 mm.

Fig. 6.

Lmγ2 shRNA reverses the effect of R5020 on tubulogenesis. Organoids were infected with control virus (GFP⁺ scramble) or virus carrying shRNA against Lmγ2 (GFP⁺ Lmγ2 shRNA) and plated in Col I gels. At 24 h after infection, organoids were treated with HGF (H) or HGF+R5020 (HR) for 24 h. A, Representative phase-contrast and GFP overlay images of organoids at 24 h. White arrows indicate chains, and black arrowheads indicate tubules. B, Quantitation of extensions, chains, and tubules at 4 or 24 h. a, P < 0.05 that Lmγ2 shRNA increased the number of tubules compared with HR controls. A total of 48–67 organoids were analyzed per treatment from two separate experiments. C, Quantitation of the percentage of organoids producing long tubules (≥50 pixels); 134–141 organoids were analyzed per treatment from three separate experiments. *, P < 0.05 that Lmγ2 shRNA increased the percentage of organoids with long tubules. D and E, The effect of Lmγ2 shRNA on Rac1 activation (D) and phospho-Src (E). Rac1-GTP levels were analyzed by Rac1-GTP pull-down and immunoblot. For densitometry, Rac1-GTP was normalized to total Rac1. Phospho-Src was assessed by immunoblot 4 h after treatment with HGF+R5020.

Fig. 7.

Effect of antibody blocking of α6-integrin on R5020 inhibition of tubulogenesis. Organoids plated in Col I gels were treated with HGF (H) or HGF+R5020 (HR) in the presence or absence of the α6-integrin blocking antibody, GoH3 (10 μg/ml) for 24 h. A, Representative phase-contrast micrographs of organoids at 24 h. White arrows indicate chains, and black arrowheads indicate tubules. Scale bar, 0.05 mm. B, Quantitation of extensions, chains, and tubules after 4 and 24 h treatment. A total of 54–69 organoids were analyzed per treatment from two separate experiments. C, Percentage of organoids forming long tubules (≥140 pixels) after 24 h. D, The effect of α6-integrin blocking antibody on Rac1 activity and phospho-Src. A representative experiment shows Rac1-GTP and phospho-Src levels at 4 h analyzed by Rac1-GTP pull-down and immunoblot. For densitometry, Rac1-GTP was normalized to total Rac1.