Autocrine/paracrine actions of growth hormone in human melanoma cell lines

Abstract

Melanoma is the most aggressive skin cancer. Its aggressiveness is most commonly attributed to ERK pathway mutations leading to constitutive signaling. Though initial tumor regression results from targeting this pathway, resistance often emerges. Interestingly, interrogation of the NCI-60 database indicates high growth hormone receptor (GHR) expression in melanoma cell lines. To further characterize melanoma, we tested responsiveness to human growth hormone (GH). GH treatment resulted in GHR signaling and increased invasion and migration, which was inhibited by a GHR monoclonal antibody (mAb) antagonist in WM35, SK-MEL 5, SK-MEL 28 and SK-MEL 119 cell lines. We also detected GH in the conditioned medium (CM) of human melanoma cell lines. GHR, JAK2 and STAT5 were basally phosphorylated in these cell lines, consistent with autocrine/paracrine GH production. Together, our results suggest that melanomas are enriched in GHR and produce GH that acts in an autocrine/paracrine manner. We suggest that GHR may constitute a therapeutic target in melanoma.

Keywords: Autocrine; Growth hormone (GH); Growth hormone receptor (GHR); Melanoma; Paracrine.

Fig. 1

GH Induces Phosphorylation of GHR/JAK2/STAT5 in Melanoma A-F - Time course of GH-induced phosphorylation of GHR, JAK2, and STAT5 in WM35 cells. Serum-starved cells were treated with vehicle or GH (500 ng/mL) for the indicated periods. Detergent cell extracts were immunoprecipitated with anti-GHR (A) or anti-JAK2 (C) vs NI control serum. Eluates were resolved by SDS-PAGE and serially immunoblotted with anti-pY and anti-GHR (A) or anti-pY and anti-JAK2 (C). Detergent cell extracts were electrophoresed and serially immunoblotted with anti-pSTAT5 and anti-STAT5 (E). Data from three separate experiments, performed as in A,C, and E, were used to densitometrically estimate indicated relative phosphorylation levels. For each experiment, maximum phosphorylation achieved in response to GH was considered 100%. Data are expressed as mean ± SE for each GH treatment duration. IP, immunoprecipitation; NI, nonimmune; WB, Western blot. G-J, GH concentration dependence of STAT5 and GHR phosphorylation. Serum-starved WM35 and SK-MEL 5 cells were treated with vehicle or the indicated concentrations of GH for 10 min. Detergent cell extracts were either resolved by SDS-PAGE without immunoprecipitation and serially immunoblotted with anti-pSTAT5 and anti-STAT5 (G,I) or immunoprecipitated with anti-GHR (H, J) vs NI control serum. Eluates were resolved by SDS-PAGE and serially immunoblotted with anti-pY and anti-GHR (H, J).

Fig. 2

GH induces invasion and migration in WM35 cells A, B - Melanoma cells (2 × 10⁴ cells/500 μl + 10% FBS-containing medium) were placed in the upper chamber of Boyden chamber containing buffer (BB) or 500 ng/mL of GH. The lower chamber contained 750 μl of medium supplemented with 10% FBS and 10 μg/mL of Fibronectin. After 16 h, invaded cells on the lower surface of the membranes were fixed with chilled paraformaldehyde and stained with crystal violet. A representative image from eight independent experiments is shown (A). The invaded cells were counted in eight randomly selected microscopic fields on the membrane and the results are summarized and expressed as the mean number of invaded cells ± SEM per microscopic field (B). Significant difference versus control group, P = .0004. C, D - WM35 cells were treated with BB or GH (25 ng/mL) or co-treated with anti-GHR_cyt-mAb and GH or anti-GHR_ext-mAb and GH for 16 h. Control compared to GH, P = .002. GH compared to GHR_ext-mAb, P = .02. Anti-GHR_cyt-mAb compared to anti-GHR_ext-mAb, P = .05. E,F- WM35 cells were subjected to a scratch wound and treated with GH (500 ng/mL) or BB for 6 h. A representative image is shown in (E). Migration was estimated as the relative fold change of wound closure, P = .015 (F).

Fig. 3

Autocrine GH production and action in melanoma cells. A - Melanoma Conditioned Medium (CM) promotes survival of 32D-GHR cells Serum-starved 32D-GHR cells were exposed to the indicated concentrations of hGH or CM from WM35 or SK-MEL 5 cells (1:2 dilution) for 48 h and cell viability was measured by MTT assay, as detailed in Materials and Methods. Data (mean ± SEM of quadruplicate determinations) for this experiment are plotted as the fold increase in OD₅₆₇ relative to the value determined when no GH was added (BB). The experiment shown is representative of three such experiments. Control compared to WM35, P = .03. Control compared to SK-MEL 5, P = .02.B-G - Anti-GHR_ext-mAbinhibits melanoma invasion in absent added GH in SK-MEL 5, SK-MEL 119, and SK-MEL 28 cells. Melanoma cells (2 × 10⁴ cells/500 μl + 10%FBS containing medium) were placed in the upper chamber of Boyden chamber containing anti-GHR_ext-mAb or anti-GHR_cyt-mAb (20  μg/mL each). The lower chamber contained 750 μl of medium supplemented with 10% FBS and 10 μg/mL of Fibronectin. Invasion was measured as in Fig. 2. Representative images and quantitation of multiple experiments using SK-MEL5 (B,C; n = 8), SK-MEL 119 (D,E; n = 6), and SK-MEL 28 (F,G; n = 3) are shown. P-values compared to control mAb are as indicated.

Fig. 4

Anti-GHR_ext-mAbinhibits melanoma migration in the absence of GH in SK-MEL 5, WM35 and SK-MEL 28 and SK-MEL 119 cells. A,B – SK-MEL 5. Boyden Chamber migration assay was performed as in Methods. Representative images (A) and quantitation of n = 4 experiments (B) are shown. C–H - Scratch migration assays were performed as in Methods. Representative images (C,E,G) and quantitation of multiple experiments (D,F, H) are shown for WM35 (C, D; n = 4), SK-MEL 28 (E, F; n = 6), and SK-MEL 119 (G,H; n = 4) cells. P-values compared to controls are as indicated.