Conditional gene regulation models demonstrate a pro-proliferative role for growth hormone receptor in prostate cancer

Abstract

Background: Humans with inactivating mutations in growth hormone receptor (GHR) have lower rates of cancer, including prostate cancer. Similarly, mice with inactivating Ghr mutations are protected from prostatic intraepithelial neoplasia in the C3(1)/TAg prostate cancer model. However, gaps in clinical relevance in those models persist. The current study addresses these gaps and the ongoing role of Ghr in prostate cancer using loss-of-function and gain-of-function models.

Methods: Conditional Ghr inactivation was achieved in the C3(1)/TAg model by employing a tamoxifen-inducible Cre and a prostate-specific Cre. In parallel, a transgenic GH antagonist was also used. Pathology, proliferation, and gene expression of 6-month old mouse prostates were assessed. Analysis of The Cancer Genome Atlas data was conducted to identify GHR overexpression in a subset of human prostate cancers. Ghr overexpression was modeled in PTEN-P2 and TRAMP-C2 mouse prostate cancer cells using stable transfectants. The growth, proliferation, and gene expression effects of Ghr overexpression was assessed in vitro and in vivo.

Results: Loss-of-function for Ghr globally or in prostatic epithelial cells reduced proliferation and stratification of the prostatic epithelium in the C3(1)/TAg model. Genes and gene sets involved in the immune system and tumorigenesis, for example, were dysregulated upon global Ghr disruption. Analysis of The Cancer Genome Atlas revealed higher GHR expression in human prostate cancers with ERG-fusion genes or ETV1-fusion genes. Modeling the GHR overexpression observed in these human prostate cancers by overexpressing Ghr in mouse prostate cancer cells with mutant Pten or T-antigen driver genes increased proliferation of prostate cancer cells in vitro and in vivo. Ghr overexpression regulated the expression of multiple genes oppositely to Ghr loss-of-function models.

Conclusions: Loss-of-function and gain-of-function Ghr models, including prostatic epithelial cell specific alterations in Ghr, altered proliferation, and gene expression. These data suggest that changes in GHR activity in human prostatic epithelial cells play a role in proliferation and gene regulation in prostate cancer, suggesting the potential for disrupting GH signaling, for example by the FDA approved GH antagonist pegvisomant, may be beneficial in treating prostate cancer.

Keywords: GH; GHR; IGF-1; Pten; RNA-seq; T-antigen.

Figure 1.

A Kaplan-Meier curve is shown for genotypes Ghr^flox/flox;Rosa-Cre-ER^T2;C3(1)/TAg^+/0 (n=36), Ghr^flox/flox;C3(1)/TAg^+/0 (n=18), Ghr^flox/flox;Pb-Cre^+/0;C3(1)/TAg^+/0 (n=15), and GHA^+/0;C3(1)/TAg^+/0 (n=20). A Mantel-Cox log-rank text reveals no significant difference in survival curves between groups.

Figure 2.

RT-qPCR was used to measure Ghr expression in the DLP (A) and livers (B) of mice. Expression is presented relative to the expression of the housekeeping gene Tbp. Serum IGF-1 levels were measured by ELISA (C). Serum testosterone was quantified by LC-MS/MS (D). N = 3–7 per group. Statistically significant differences from vehicle control determined by ANOVA with Tukey’s multiple comparison’s test are indicated; *p<0.05, **p<0.01.

Figure 3.

Representative images of H&E-stained (A-E) and Ki67-stained (F-J) sections are shown for vehicle control (A, F), tamoxifen control (B, G), global Ghr deletion (C, H) prostate-specific Ghr deletion (D, I) and GHA (E, J) mouse DLPs. Scale bar for each image = 50 μm. Dotted borders in A-E highlight examples of areas identified as epithelial stratification for quantification purposes. Arrows in F-J point to examples of Ki67 positive cells (brown stain). (K) Area of stratification was assessed by an observer blinded to group and taken as a percentage of total prostatic area with a multi-layered epithelium (n=3–6 per group). (L) Ki67 labeling index was determined by counting Ki67 positive cells and total cells based in nuclear hematoxylin stain by an observer blinded to group, (n=3–5 per group). Statistically significant differences determined by ANOVA and Tukey’s multiple comparisons test are indicated, *p<0.05, **p<0.01.

Figure 4.

A volcano plot of differentially expressed genes in the DLP of C3(1)/TAg (A) and WT (B) mice with tamoxifen-induced Ghr deletion compared tamoxifen treated controls with log₂ fold change on the x-axis and the negative log₁₀ of the adjusted p-value on the y-axis. Thresholds for significance are −log₁₀p-value > 2 and log₂FoldChange > 1 or < −1. Genes upregulated in response to Ghr deletion are colored red, and genes downregulated in response to Ghr deletion are colored cyan.

Figure 5.

TCGA project data for GHR expression in human prostate cancers separated by molecular signature are shown (A) (data visualization and statistics via the UALCAN data portal, p<1×10⁻¹² for normal vs ERG-fusion, p=1.16×10⁻⁴ for normal vs ETV1-fusion). The UALCAN data portal was used to compare GHR (vertical axis) and ERG (horizontal axis) expression in TCGA prostate tumors (B). Pearson Correlation coefficient: 0.58.

Figure 6.

PTEN-P2 or TRAMP-C2 cells stably transfected with a plasmid containing either mouse Ghr (_mGhr) or an empty vector control (_EV) suspended in rat collagen pellets (3.5 × 10⁵ cells per graft) were surgically grafted under the kidney capsules of 10-week-old male Balb/C nu/nu mice. Xenografts were grown for 5 weeks. Examples of resultant xenografts are shown for PTEN-P2_EV (A), PTEN-P2_mGhr (B), TRAMP-C2_EV (C) and TRAMP-C2_mGhr (D). Photos in A-D show kidneys (dark red) with xenografts (lighter areas on kidney surface inside the dashed lines) and a ruler for scale. (E) Graft masses are shown for empty vector controls (black bars, n=5 per group) and Ghr upregulated (grey bars, n=4 per group). Statistically significant differences from control are indicated *p<0.05; Student’s t-test).

Figure 7.

Representative images of Ki67-stained sections are shown for PTEN-P2_EV (A), PTEN-P2_mGhr (B), TRAMP-C2_EV (C) and TRAMP-C2_mGhr (D). Scale bar for each image = 50 μm. Arrows point to examples of Ki67 positive cells (brown stain). (E) Ki67 labeling index was determined by counting Ki67 positive cells and total cells based in nuclear hematoxylin stain by an observer blinded to group, n=4 per group. Statistically significant differences determined by ANOVA are indicated, *p<0.05, **p<0.01.

Figure 8.

RNA extracted from DLPs of global Ghr deletion or tamoxifen groups in C3(1)/TAg and WT mice were used to confirm expression changes via RT-qPCR in genes first identified by RNA-seq (A). Gene expression as measured by qPCR in PTEN-P2 and TRAMP-C2 grafts with overexpression of Ghr (mGhr) compared to empty vector (EV) controls for genes identified as Ghr regulated in this study (B) and genes previously implicated by disruption of the GH/IGF-1 axis with pegvisomant treatment (28) (C). Gene expression is shown relative to the expression of the housekeeping gene, Tbp, and normalized to control = 1. Statistically significant differences determined by ANOVA with Tukey’s multiple comparison’s test are indicated, *p<0.05, **p<0.01; n = 3–4 per group.