Loss of expression of the apoptosis protease activator protein-1 (APAF-1) in human melanoma is thought to promote resistance to programmed cell death by preventing caspase-9 activation. However, the role of the APAF-1-dependent pathway in apoptosis activated by cellular stress and/or DNA damage has been recently questioned. We investigated APAF-1 expression in a large panel of human melanomas and assessed cellular response to several proapoptotic agents in tumors expressing or lacking APAF-1 protein. In two melanomas with wild-type p53 but with differential expression of APAF-1, treatment with camptothecin, celecoxib, or an nitric oxide synthase inhibitor (1400W) significantly modulated expression of 36 of 96 genes in an apoptosis-specific cDNA macroarray, but APAF-1 mRNA levels were not induced (in APAF-1 ؊ cells) nor upregulated (in APAF-1 ؉ cells), a finding confirmed at the protein level. Treatment with cisplatin, camptothecin, etoposide, betulinic acid, celecoxib, 1400W, and staurosporine promoted enzymatic activity not only of caspases -2, -8, and -3 but also of caspase-9 in both APAF-1 ϩ and APAF-1 ؊ tumor cells. Moreover, drug-induced caspase-9 enzymatic activity could be not only partially but significantly reduced by caspase-2, -3, and -8 -specific inhibitors in both APAF-1 ϩ and APAF-1 ؊ tumor cells. In response to 1 to 100 mol/L of cisplatin, camptothecin, or celecoxib, APAF-1 ؉ melanomas (n ‫؍‬ 12) did not show significantly increased levels of apoptosis compared with APAF-1 ؊ tumors (n ‫؍‬ 7), with the exception of enhanced apoptosis in response to a very high dose (100 mol/L) of etoposide. These results suggest that the response of human melanoma cells to different proapoptotic agents may be independent of their APAF-1 phenotype.

Purpose To assess the role of Apollon in melanoma resistance to intrinsic and extrinsic pathways of apoptosis and to identify strategies to reduce its expression. Experimental Design Apollon expression was assessed in melanoma cells in vitro and in vivo. Apollon modulation and melanoma apoptosis were evaluated by Western blot and/or flow cytometry in response to cytotoxic drugs, mitogen-activated protein/extracellular signal–regulated kinase (MEK)-, BRAFV600E-, and mTOR-specific inhibitors, TRAIL and anti-HLA class II monoclonal antibodies (mAb). Mitochondrial depolarization, caspase activation, apoptosis assays, and gene expression profiling were used to test effects of Apollon silencing, by siRNA, on melanoma response to antitumor agents. Results Apollon was constitutively expressed by melanoma cells, in vitro and in vivo, and at higher levels than in benign melanocytic lesions. Melanoma apoptosis correlated significantly with Apollon protein downmodulation in response to cytotoxic drugs, MEK, or BRAFV600E-specific inhibitors. Combinatorial treatment with MEK and mTOR inhibitors and HLA class II ligation, by a specific mAb, promoted Apollon downmodulation and enhanced melanoma apoptosis. Apollon downmodulation induced by antitumor agents was caspase independent, but proteasome dependent. Knockdown of Apollon, by siRNA, triggered apoptosis and/or significantly enhanced melanoma cell death in response to cytotoxic drugs, MEK- and BRAFV600E-specific inhibitors, and soluble or membrane-bound TRAIL. Apollon silencing promoted mitochondrial depolarization and caspase-2, caspase-8, caspase-9, and caspase-3 activation in response to different antitumor agents and altered the profile of genes modulated by MEK or BRAFV600E-specific inhibitors. Conclusions Targeting of Apollon may significantly improve melanoma cell death in response to antitumor agents that trigger the intrinsic or the extrinsic apoptosis pathways.

Neoplastic cells are thought to have defective expression of costimulatory molecules. However, in this study, we show that human melanoma cells express LIGHT/TNFSF14, a ligand of herpesvirus entry mediator on T cells and of lymphotoxin beta receptor on stromal cells. In vitro, melanoma cells stained for LIGHT in the intracellular compartment, with weak or negative cell surface expression. However, LIGHT was expressed on tumor-derived microvesicles released from melanoma cells. In vivo, LIGHT was found in metastatic lesions, and the extent of lymphotoxin beta receptor expression on the stromal cells was significantly associated with a "brisk" T-cell infiltrate in the neoplastic tissue. In the lesions with a brisk T-cell infiltrate, stromal cells surrounding the tumor also stained for the T-cell attractant chemokine CCL21. The intratumoral T lymphocytes frequently expressed herpesvirus entry mediator and were characterized by a differentiated phenotype. Coculture of lymphocytes with LIGHT(+) melanoma-derived microvesicles or even with LIGHT(+) melanoma cells in the presence of interleukin-2 costimulated LIGHT-dependent CD3(+)CD8(+) T-cell proliferation. However, lymphocyte coculture with LIGHT(+) microvesicles in the presence of interleukin-2 was also associated with an apoptotic response as documented by increased binding of Annexin V by CD3(+)CD8(+) T cells. These data suggest that LIGHT constitutively expressed in human melanoma cells and microvesicles may contribute to regulate T-cell responses to tumor cells.

No abstract

Human melanomas can express unique tumor antigens, resulting from mutated proteins, and shared epitopes encoded for by normal genes, but these two classes of antigens have not been previously compared for immunogenicity and retention in metastatic cells. Here, we identified a new unique antigen generated by a point mutation in the peroxiredoxin 5 (Prdx5) gene in an HLA-A*0201+ human metastatic melanoma lacking the wild-type allele. An antioxidant assay, with recombinant Prdx5 proteins, and evaluation of peroxide accumulation in transiently transfected cells, indicated that the mutant protein retained its enzymatic activity. The mutation in the Prdx5 protein did not generate a new HLA agretope but yielded an HLA-A*0201–restricted T cell epitope (Prdx5110-119). By HLA-tetramer analysis, in a tumor-invaded lymph node, >50% of mutant Prdx5-specific CD8+ T cells (frequency 0.37%/CD8+) showed a CCR7+/− CD45RA− “TCM” or “TEM” phenotype, as found in Melan-A/MART-1–specific T cells (frequency 0.68%/CD8+) in the same tissue. In agreement with their memory phenotype, the Prdx5-specific T cells readily expanded in vitro in mixed lymphocyte-tumor culture, as did the Melan-/MART-1–specific T cells. By immunohistochemistry of the invaded lymph node, the mutant Prdx5 protein was expressed in all neoplastic cells, in contrast with the heterogeneous expression of shared antigens as Melan-A/MART-1, gp100 and tyrosinase. Thus, a unique tumor antigen can be as immunogenic as the melanoma differentiation antigens but, in contrast to the latter, may be retained in all metastatic cells possibly as result of the relevant cellular function exerted by the mutated protein.

6 Figures from Constitutive Expression and Costimulatory Function of LIGHT/TNFSF14 on Human Melanoma Cells and Melanoma-Derived Microvesicles

<p>PDF file, 1009KB.</p>

<div>Abstract<p><b>Purpose:</b> To assess the role of Apollon in melanoma resistance to intrinsic and extrinsic pathways of apoptosis and to identify strategies to reduce its expression.</p><p><b>Experimental Design:</b> Apollon expression was assessed in melanoma cells <i>in vitro</i> and <i>in vivo</i>. Apollon modulation and melanoma apoptosis were evaluated by Western blot and/or flow cytometry in response to cytotoxic drugs, mitogen-activated protein/extracellular signal–regulated kinase (MEK)-, BRAF^V600E-, and mTOR-specific inhibitors, TRAIL and anti-HLA class II monoclonal antibodies (mAb). Mitochondrial depolarization, caspase activation, apoptosis assays, and gene expression profiling were used to test effects of Apollon silencing, by siRNA, on melanoma response to antitumor agents.</p><p><b>Results:</b> Apollon was constitutively expressed by melanoma cells, <i>in vitro</i> and <i>in vivo</i>, and at higher levels than in benign melanocytic lesions. Melanoma apoptosis correlated significantly with Apollon protein downmodulation in response to cytotoxic drugs, MEK, or BRAF^V600E-specific inhibitors. Combinatorial treatment with MEK and mTOR inhibitors and HLA class II ligation, by a specific mAb, promoted Apollon downmodulation and enhanced melanoma apoptosis. Apollon downmodulation induced by antitumor agents was caspase independent, but proteasome dependent. Knockdown of Apollon, by siRNA, triggered apoptosis and/or significantly enhanced melanoma cell death in response to cytotoxic drugs, MEK- and BRAF^V600E-specific inhibitors, and soluble or membrane-bound TRAIL. Apollon silencing promoted mitochondrial depolarization and caspase-2, caspase-8, caspase-9, and caspase-3 activation in response to different antitumor agents and altered the profile of genes modulated by MEK or BRAF^V600E-specific inhibitors.</p><p><b>Conclusions:</b> Targeting of Apollon may significantly improve melanoma cell death in response to antitumor agents that trigger the intrinsic or the extrinsic apoptosis pathways. <i>Clin Cancer Res; 18(12); 3316–27. ©2012 AACR</i>.</p></div>