Suppression of the proinflammatory response of metastatic melanoma cells increases TRAIL-induced apoptosis

Abstract

Melanoma is the most lethal form of human skin cancer. However, only limited chemotherapy is currently available for the metastatic stage of the disease. Since chemotherapy, radiation and sodium arsenite treatment operate mainly through induction of the intrinsic mitochondrial pathway, a strongly decreased mitochondrial function in metastatic melanoma cells, could be responsible for low efficacy of the conventional therapy of melanoma. Another feature of metastatic melanoma cells is their proinflammatory phenotype, linked to endogenous expression of the inflammatory cytokines, such as TNFα IL6 and IL8, their receptors, and constitutive NF-κB- and STAT3-dependent gene expression, including cyclooxygenase-2 (PTGS2/COX2). In the present study, we treated melanoma cells with immunological (monoclonal antibody against TNFα or IL6), pharmacological (small molecular inhibitors of IKKβ-NF-κB and JAK2-STAT3) or genetic (specific RNAi for COX-2) agents that suppressed the inflammatory response in combination with induction of apoptosis via TRAIL. As a result of these combined treatments, exogenous TRAIL via interactions with TRAIL-R2/R1 strongly increased levels of apoptosis in resistant melanoma cells. The present study provides new understanding of the regulation of TRAIL-mediated apoptosis in melanoma and will serve as the foundation for the potential development of a novel approach for a therapy of resistant melanomas.

Fig. 1

Mitochondrial (mt) DNA content, mitochondrial respiration and AKT levels in human melanoma cell lines. A: Content of mitochondrial (mt) DNA was determined using the real-time PCR as described in the Materials and Methods Section. Number of mutations in the D-loop of mtDNA is indicated. All of these mutations represent previously reported polymorphisms. B: The rate of oxygen consumption for normal human fibroblasts TIG-3, human melanocytes and melanoma lines was determined using an oxygen electrode unit. C: Western blot analysis of total and phospho-AKT levels in melanocytes, TIG-3 fibroblasts and indicated melanoma lines. D: Surface levels of EGFR and NGFR1 p75 in human melanoma cells were determined by immunostaining with anti-EGFR-PE and anti-NGFR1-PE mAbs using FACS analysis; n.s., nonspecific staining.

Fig. 2

The mitochondrial death pathway in human melanoma cell lines. A,B: Cell cycle-apoptosis analysis of melanocytes, FEMX, LU1205, and WM9 melanoma cells, which were treated sodium arsenite (5 µM) in the presence or absence of either NGF-β (100 ng/ml) or EGF (50 ng/ml) for the next 24 h; non-treated control cells were designed as (Con). Caspase inhibitors, zVAD, LEHD, and IETD (50 µM), were additionally used. After treatment, cells were stained with PI and analyzed by FACS assay. Error bars represent mean ± SD (Student's t-test, P < 0.05). C: Effects of sodium arsenite on oxygen consumption of melanocytes (M-cytes), FEMX and LU1205 melanoma cells 24h after treatment. D: Clonogenic survival assay of melanoma cell lines 12 days after sodium arsenite treatment.

Fig. 3

A role of NF-κB for expression and secretion of TNFα and IL6 and regulation of TRAIL-induced apoptosis in human melanoma cell lines. A: Effect of an IKK-NF-κB inhibitor BMS-345541 (10 µM) on the NF-κB DNA-binding activity determined by EMSA. Position of two main DNA-binding complexes is shown. DNA-binding activity of the ubiquitous transcription factor NF-Y was used as an internal control. Effect of BMS-345541 on expression levels of indicated proteins in LU1205 cells was determined by Western blot analysis. B: Effect of an IKK-NF-κB inhibitor BMS-345541 (10 µM) on 2×NF-κB-Luc reporter activity in transiently transfected melanoma cell lines. C,D: Effect of BMS-345541 (10 µM) on TNFα and IL6 secretion by human melanocytes (MC) and melanoma cell lines determined by ELISA. E: Surface expression of TNFR1 was determined by immunostaining with anti-TNFR1-PE mAb and FACS analysis; n.s., nonspecific staining. F: Effects TRAIL (30ng/ml) alone or in combination with BMS-345541 (10 µM) on apoptosis (24 and 48 h after exposure) in indicated melanoma lines. Apoptosis levels were determined, as described above. Error bars represent mean ± SD (Student's t-test, P < 0.05). G: Effects of TRAIL (30 ng/ml) and BMS-345541 (10 µM) on clonogenic survival melanoma cell lines.

Fig. 4

Effects of STAT3 inhibition on melanoma cell apoptosis. A: Dose-dependent effect of STAT3 inhibitor-6 (S3I-201) on STAT-dependent luciferase activity. WM9 cells were transiently transfected with 3xSTAT-Luc reporter and then treated for 6 h using S3I-201. Western blot analysis of phospho-Tyr705-STAT3 and COX-2 levels after 6-h exposure with S3I-201. Actin was used as a loading control. B: Surface expression levels of DR5 were determined for indicated cell lines using with anti-DR5-PE mAb and flow cytometry. C: Effects of S3I-201 (50 µM) alone or in combination with TRAIL on cell cycle-apoptosis in LU1205 and WM9 melanoma cells. Results of a typical experiment are presented (one from three). D: Clonogenic survival assay of LU1205 and WM9 cells 12 days after treatment with S3I-201 (50 µM), TRAIL (30 ng/ml) or their combination. Double star indicates zero survival of WM9 cells after combined treatment. Error bars represent mean ± SD (Student's t-test, P < 0.05).

Fig. 5

Double inhibition of NF-κB and STAT3: effects on TRAIL-induced apoptosis. A: Cells were treated with BMS-345541 (10 µM), S3I-201 (50 µM), TRAIL (30 ng/ml) individually or in combination. Apoptotic levels were determined using PI staining of DNAand flow cytometry. B,C: Effects of anti-TNFα (aTNF) (5 µg/ml), anti-IL6 (aIL6) (5 µg/ml) mAbs or their combination on TRAIL-induced apoptosis in melanoma cell lines. Error bars represent mean ± SD (Student's t-test, P <0.05)

Fig. 6

Effects of downregulation of COX-2 on TRAIL-induced apoptosis. A: Effects of NS398 (50 µM), an enzymatic inhibitor of COX-2, on TRAIL (30ng/ml) or TRAIL+ CHX (1 µg/ml) induced apoptosis in fibroblasts and indicated melanoma lines. B: Effects of COX-2 suppression by specific COX-2 RNAi on COX-2 and cFLIP-L protein expression levels in WM9 and LU1205 melanoma cells. Cell cultures stably transfected by control or COX-2 RNAi were previously described. C: TRAIL-induced apoptosis in control and COX-2 deficient cell cultures. Apoptotic levels were determined 48 h after treatment using PI staining and the flow cytometry. Error bars represent mean ± SD (Student's t test, P <0.05).

Fig. 7

Resveratrol (RSV), a natural inhibitor of NF-κB and STAT3 activation, negatively regulates secretion of TNFα and IL6, COX-2 expression and increases TRAIL-induced apoptosis in melanoma lines. A,B: Effects of RSV (25–100 µM) on TNFα and IL6 secretion. C: Western blot analysis of COX-2 protein levels following RSV exposure. D: RSV (50 µM) suppressed oxygen consumption in melanoma cells. E: Effects of RSV, TRAIL, and their combination on clonogenic survival of melanoma cell lines. Error bars represent mean ± SD (Student's t-test, P < 0.05).