Elevated cyclic AMP levels promote BRAFCA/Pten-/- mouse melanoma growth but pCREB is negatively correlated with human melanoma progression

Abstract

Melanocyte development and differentiation are regulated by cAMP, which is produced by the adenylate cyclase (AC) enzyme upon activation of the melanocortin-1-receptor (MC1R). Individuals carrying single amino acid substitution variants of MC1R have impaired cAMP signaling and higher risk of melanoma. However, the contribution of AC to this risk is not clear. Downstream of AC, the phosphorylated transcription factor, cyclic AMP Responsive Element Binding Protein (pCREB), which is activated by protein kinase A, regulates the expression of several genes including the melanocyte master regulator MITF. The roles of AC and CREB in melanoma development and growth are not well understood. Here, we investigated the effect of topical application of AC inhibitor on Braf^CA/Pten^-/- mouse melanoma development. We show that AC inhibitor delays melanoma growth independent of MAPK pathway activity and melanin content. Next, employing a primary melanoma tissue microarray and quantitative immunohistochemistry, we show that pCREB levels are positively correlated with the proliferative status of melanoma, but low pCREB expression is associated with tumor aggressiveness and metastatic recurrence. These data suggest that low cAMP signaling inhibits tumor growth but is a predictor of melanoma aggressiveness.

Keywords: Adenylate cyclase (AC); BRAF/PTEN melanoma; CREB; Cyclic AMP (cAMP); Melanoma progression; Melanoma tissue microarray.

Figure 1. Inhibition of adenylate cyclase (AC) decreases melanoma tumor growth

(A) Growth kinetics of individual tumors and (B) end-point tumor volume. Melanoma tumors were induced with topical application of 5mM 4-hydroxitamoxifen followed by 14 days of topical application of either AC inhibitor SQ22,536 or vehicle DMSO. Data show volume calculated using the formula for a prolate ellipsoid, (length × width²)/2. DMSO: n = 5 (females; 17 tumors); SQ22,536: n = 6 (1 females, 5 males; 19 tumors). Data (mean ± SD) analyzed by unpaired Student’s t test are shown. P values: *** indicates P ≤0.001. (C) Tissue melanin in tumors excised from treated sites of Braf^CA/Pten^−/− mice. Each data point shows pooled values from three representative areas of a single tumor. Graphs show data (mean ± SD) for tumor tissue melanin estimated by interpolating values to a standard curve generated using synthetic melanin. DMSO: n = 10 tumors from 3 mice; SQ22,536: n = 13 tumors from 4 mice. (D) Western blot shows pCREB and pERK levels in 3 representative melanoma tumors (left panel) and quantification of pCREB/CREB or pERK/ERK ratio for each treatment (right panel). Data (mean ± SD) for 3 different tumors were analyzed by unpaired Student’s t test.

Figure 2. Histological and immunohistochemical analysis of Braf^CA/Pten^−/− mouse tumors

(A) H&E and immunohistochemical staining with antibodies to melanoma marker S100A4 and proliferative marker Ki67. Representative sections of tumors excised from mice treated with topical application of DMSO or adenylate cyclase (AC) inhibitor SQ22,536. Magnification: 20×, scale bar = 200µm.

Figure 3. Effect of AC inhibition on Braf^CA/Pten^−/− mouse melanoma tumor-derived and human melanoma cells

MTT assay showing the effect of treatment with 50µM of the AC inhibitor SQ22,536 on growth of (A) melanoma cells isolated and cultured from vehicle-treated mouse tumor and (B) human primary melanoma cells, or (C) human metastatic melanoma cells. SQ22,536 in the medium was replenished every 48 h throughout the all the experiments. DMSO: Control; SQ22,536: AC inhibitor. All data were analyzed by unpaired Student’s t test; *** p ≤0.001.

Figure 4. Characterization of AC inhibition in melanocytes and melanoma cells

(A) and (B) Effect of treatment with 100µM SQ22,536 for 8 h on [cAMP] of human melanocytes NHM and metastatic melanoma cell lines as measured by ELISA. CT – cholera toxin; IBMX - 3-Isobutyl-1-methylxanthine; TPA - 12-O-Tetradecanoylphorbol-13-acetate. Data (mean ± SD) show 3 replicates normalized to a DMSO and analyzed by unpaired Student’s t test; * p = 0.0126. (C) Effect of 5 h treatment with 100µM SQ22,536 on the cAMP Responsive Element (CRE) promoter measured by luciferase reporter activity in human melanoma cell lines. Data (mean ± SD) show at least 3 replicates normalized to DMSO controls.

Figure 5. Effect of AC inhibition in MAPK pathway inhibition

(A) Western blot analysis shows the effect of 8 h treatment with 100µM SQ22,536 on pCREB and pERK of human melanoma cell lines. (B) Survival measured by MTT assay after 72 h treatment with increasing concentrations of MEK inhibitor AZD6244 alone or in combination with the AC inhibitor SQ22,536. Data (mean ± SD) from 5–6 replicates normalized to a DMSO control are shown. (C) Effect of AC inhibition on survival of melanoma cell lines was measured using MTT assay. Cells plated in 96 wells were treated with increasing concentrations of SQ22,536 for 72 h. Data (mean ± SD) from 5–6 replicates normalized to DMSO control are shown.

Figure 6. Relationship between Ki67 and pCREB in cutaneous nevi and primary melanoma

A and B show linear relationship between Ki67 and pCREB in: (A) nevi and (B) primary melanomas. The staining intensity obtained from automated multispectral image analyses for pCREB and Ki67 are shown.

Figure 7. Relationship of Ki67 and pCREB expression to melanoma clinico-pathological characteristics

(A) Ki67 levels in nevi and primary melanoma. Data (±SD) were analyzed by unpaired Student’s t test. **** = P < 0.0001. (B) pCREB levels in nevi and primary melanoma. Data were analyzed by unpaired Student’s t test. Kaplan–Meier Survival analysis (C) and (D) show disease (recurrence)-free survival of patients with primary melanoma tumors from all stages: (C) Nuclear pCREB; Low = optical density of ≤0.037 (D) Cytoplasmic pCREB; Low = optical density of ≤0.008.