Growth hormone is a cellular senescence target in pituitary and nonpituitary cells

Abstract

Premature proliferative arrest in benign or early-stage tumors induced by oncoproteins, chromosomal instability, or DNA damage is associated with p53/p21 activation, culminating in either senescence or apoptosis, depending on cell context. Growth hormone (GH) elicits direct peripheral metabolic actions as well as growth effects mediated by insulin-like growth factor 1 (IGF1). Locally produced peripheral tissue GH, in contrast to circulating pituitary-derived endocrine GH, has been proposed to be both proapoptotic and prooncogenic. Pituitary adenomas expressing and secreting GH are invariably benign and exhibit DNA damage and a senescent phenotype. We therefore tested effects of nutlin-induced p53-mediated senescence in rat and human pituitary cells. We show that DNA damage senescence induced by nutlin triggers the p53/p21 senescent pathway, with subsequent marked induction of intracellular pituitary GH in vitro. In contrast, GH is not induced in cells devoid of p53. Furthermore we show that p53 binds specific GH promoter motifs and enhances GH transcription and secretion in senescent pituitary adenoma cells and also in nonpituitary (human breast and colon) cells. In vivo, treatment with nutlin results in up-regulation of both p53 and GH in the pituitary gland, as well as increased GH expression in nonpituitary tissues (lung and liver). Intracrine GH acts in pituitary cells as an apoptosis switch for p53-mediated senescence, likely protecting the pituitary adenoma from progression to malignancy. Unlike in the pituitary, in nonpituitary cells GH exerts antiapoptotic properties. Thus, the results show that GH is a direct p53 transcriptional target and fulfills criteria as a p53 target gene. Induced GH is a readily measurable cell marker for p53-mediated cellular senescence.

Keywords: acromegaly; pituitary hormone.

Fig. 1.

Senescence induces pituitary cell GH expression. (A) SA-β-galactosidase activity in GC and rat primary pituitary cells treated with 7 µM Nutlin 3 (nutlin) or DMSO (control) for 48 h. (B) GC cell proliferation rates and GH secretion measured by RIA after 3 µM nutlin for 72 h. GH secretion is normalized to cell number. (C) Western blot analysis of GC and primary rat pituitary cells treated with 7 µM nutlin for 48 h. (D) GC cells were transfected with pGL4-luc or pGL4-rGH luc reporter plasmids and treated with DMSO or 7 µM nutlin for 48 h. Results were normalized to cotransfected Renilla control reporter vector for transfection efficiency and expressed as mean ± SE calculated from triplicate assays, and experiments repeated three times with similar results. Results of a representative experiment are shown; *P < 0.05.

Fig. 2.

Senescence induces GH expression in nonpituitary cells. (A) SA-β-galactosidase activity in MCF7, HT116, and hPCC cells. MCF7 and HCT116 cells were treated with 7 µM nutlin for 72 and 96 h, respectively; hPCC cells were treated with 3 µM nutlin for 48 h. (B and C) Western blot analysis of (B) MCF7 and HCT116 cells and (C) hPCCs treated with nutlin. (D) human antibody array showing medium GH derived from MCF7 and HCT116 cells treated with nutlin.

Fig. 3.

p53 activates GH transcription in senescent cells. Western blot analysis of (A) GC cells infected with lentiviral particles expressing rat MDM2 shRNA. (B) GC cells transiently transfected with pCMV or pCMV-rp53 (p53). (C) GC cells were plated in triplicate wells, transiently transfected with pCMV or pCMV-rp53 for 48 h, culture medium collected, and GH measured by RIA. (D and E) Western blot analysis of (D) HCT116 cells transiently transfected with pcDNA of pcDNA-hp53 (p53) and (E) GC cells transfected with scrambled siRNA or rat sip53 for 24 h, followed by indicated nutlin doses for 48 h. (F) Western blot analysis of HCT116 cells with or without p53 expression treated with 7 µM nutlin for 96 h. All experiments were repeated two times, and representative results are shown.

Fig. 4.

p53 binds to and activates GH promoter. (A) ChIP assay showing specific p53 binding to the rat GH promoter. GC cells were transfected with pCMV-rp53 and chromatin immunoprecipitated with anti-p53 or control IgG antibody as indicated. GH promoter fragments were detected by PCR analysis using four primer sets. Enrichment of specific GH promoter sequences was obtained with primer sets 1, 2, and 3 but not with set 4 or with IgG antibody. The experiment was repeated twice, and results of a representative assay shown. (B) ChiP assay showing specific endogenous p53 binding to the rat GH promoter. GC cells were treated with 7 µM nutlin for 48 h and chromatin immunoprecipitated with anti-p53 or control IgG antibody as indicated. GH promoter fragments were detected by PCR analysis using three primer sets. Enrichment of specific GH promoter sequences was obtained with primer set 2 but not with set 1 or 3 or with IgG antibody. The experiment was repeated twice, and results of a representative assay shown. (C) Representative EMSA shows binding of GH nuclear extract (NE) p53 with probe 1(−1154/−1144 bp upstream from the GH TSS), corresponding to putative p53 binding sites on GH promoter. Competition assays were performed using a 100-fold excess of cold probe 1 (lane 4) and unlabeled p53 consensus oligos (lane 5). Labeled p53 consensus oligos were used as positive control, showing DNA–protein complex of the same size (lane 6). Addition of anti-p53 antibodies (Ab) markedly decreased p53 complexing with labeled probe 1 (lane 3) and with labeled p53 consensus oligos (lane 7), indicating the specificity of p53 binding to GH promoter. (D) GC cells were transfected with pGL4-luc or pGL4-rGH luc reporter plasmids and cotransfected with pcDNA3 expressing vector, or pcDNA3-hp53; cells were harvested after 24 h, assayed for luciferase, and results normalized to cotrasfected Renilla reporter vector for transfection efficiency. Results are expressed as mean ± SE from triplicate assays, and experiments were repeated three times with similar results. Results of a representative experiment are shown. *P < 0.05.

Fig. 5.

GH–secreting human pituitary adenomas are senescent, and senescence further induces GH expression and secretion. (A–C) Confocal image of three different human GH-secreting adenoma specimens labeled with GH or β-galactosidase (both green). Nuclei are counterstained with DNA-specific dye DAPI (blue). Blood appears pink. Both proteins are similarly distributed in pituitary adenoma. Primary GH cell adenoma cells were plated in duplicate wells, treated with 5 µM nutlin for 48 h. (D) Cells were collected and Western blot analysis performed. The experiments were performed in cultured cells derived from three GH-secreting adenomas with similar results, and the result of a representative experiment is shown. PRL expression served as loading control. (E) Culture medium from three cases of GH-secreting adenoma cells was collected, and GH and PRL measured by ELISA.

Fig. 6.

Senescence induces GH expression in vivo. Mice were injected with 40 mg/kg body weight nutlin (N) or DMSO (C) i.p. every 2 d (six doses). (A) Fold induction of tissue GH expression detected by real-time PCR (mean ± SD, n = 3). Respective tissue GH expression in vehicle-treated mice was normalized to 1. Pit, pituitary. (B) Western blot analysis of p53, GH. and PRL expression in two pooled pituitary gland extracts.

Fig. 7.

Induced GH protects GC cells from apoptosis and promotes apoptosis in HCT116 cells. Western blot analysis for GH and apoptotic markers. (A) GC and HCT116 cells were transfected with pIRES2-ZsGreen1 (V), rGH-IRES2-ZSGreen1 (rGH), or hGH-IRES2-ZSGreen1(hGH) and collected 48 h later. (B) GC cells were transfected with 5 µM rat siGH or scramble siRNA for 12 h and then treated with 7 µM nutlin for the indicated times. (C) GC cells were transfected with pIRES2-ZsGreen1 or rGH-pIRES2-ZsGreen1 plasmids for 12 h and treated with 7 μM nutlin for the indicated times. (D) HCT116 cells were treated with 5 µM nutlin for 48 h, then were cotreated with scramble (sc) or siGH RNA for another 24 or 48 h. Experiments were repeated twice, and representative blots are shown.