Melanoma is an aggressive malignancy with poor prognosis. Eradication of tumor cells requires an effective interaction between melanoma cells and different players of the immune system. As the most potent professional antigen-presenting cells, dendritic cells (DCs) play a pivotal role in mounting a specific immune response where their intratumoral and peritumoral density as well as their functional status are correlated with clinical staging of the disease and with patients’ survival. Under steady-state conditions, internalization of apoptotic cells by immature DCs designates a state of tolerance to self-antigens. Nevertheless, pathogens and necrotic cells interacting with pattern recognition receptors trigger downstream signaling pathways that evoke maturation of DCs, leading to the production of pro-inflammatory cytokines. These mature DCs are essential for T-cell priming and subsequent development of a specific immune response. Altered functions of DCs have an impact on the development of various disorders including autoimmune diseases and cancers. Herein, we focus on the checkpoints created throughout DCs antigen capturing and presentation to T cells, with subsequent development of either tolerance or immune response, with an emphasis on the role played by DCs in melanoma tumorigenesis and their therapeutic potential.
Melanocytes are cells derived from the neural crest that migrate to the epidermis where they reside in the basal layer, forming the epidermal melanin unit. Abnormally proliferating melanocytes give rise to benign nevi, dysplastic nevi and melanomas. The latter are presented by highly infiltrating cells with a strong potential to rapidly metastasize. Immune eradication of the tumor cells depends on a complex interaction between melanoma cells and various candidate cell types of the immune system including dendritic cells (DCs), effector T cells and regulatory T (Treg) cells. In this regard, DCs, as potent antigen presenting cells (APC), play a key role in mounting a specific T-cell immune response or in inducing a state of tolerance to tumor antigens. This review conveys insights into the characterization of different subsets of DCs and their role in melanoma.
The origin and classification of the dendritic cells
Dendritic cells differentiate from lymphoid or myeloid progenitors which are descendants of CD34+ haematopoietic stem cells (Hussein, 2005). In the steady state, FLT3-ligand enhances the generation of both myeloid and plasmacytoid DCs in vivo (Maraskovsky et al., 1996; Pulendran et al., 2000) and in vitro (Blom et al., 2000; Chen et al., 2004). Lymphoid progenitors give rise to plasmacytoid DCs (pDCs) which characteristically express high levels of CD123+ (interleukin-3 receptor), CD4 and CD62 ligand (CD62L), but lack expression of myeloid markers such as CD11c, CD13, CD14 and CD33 (Bendriss-Vermare et al., 2001; Briere et al., 2002). In addition, pDCs secrete large amounts of type I interferon (IFN) in response to viral infection (Grouard et al., 1997; Siegal et al., 1999). On the other hand, myeloid DCs (mDCs) originate from haematopoietic myeloid progenitor cells and differentiate into both a CD11c+ CD1a+ subset (Langerhans cells), which resides in the skin epidermis, and a CD11c+ CD1a – subset that is widely distributed in the skin dermis, blood, afferent lymph and interstitial spaces of different organs. Notably, under physiological stress, activated mDCs secrete large amounts of inflammatory chemokines such as IL-8, monocyte chemoattractant protein (MCP) and macrophage inflammatory protein (MIP), which attracts CD14+ monocytes. The latter differentiate into mDCs CD11c+ CD1a – that ensure the continuity of local antigen sampling (Jaksits et al., 1999).
According to the intensity of DC surface antigen expression and cytokine profile, mDCs are subdivided into immature, semi-immature and mature DCs (Table 1). Immature DCs are characterized by low surface expression of CD80/CD86 and MHC class II molecules, and by the absence of cytokine production (Rutella et al., 2006). However, semi-immature DCs are characterized by an up-regulation of these surface molecules with unique production of IL-10. Lastly, mature DCs differ from the previous two subtypes by the copious secretion of IL-12 and TNF-α and discontinued IL-10 production, up-regulation of CD40, CD80, CD86 and MHC class II molecules (Caux et al., 1994; Rutella et al., 2006). In addition, mature DCs lose their phagocytic capacity and their adhesive potential with a reorganization of the cytoskeleton that renders them highly motile (Trombetta and Mellman, 2005).
Table 1. Dendritic cell maturation phenotypes
Low CD80 Low CD86 Low MHC class II molecules
CD80 CD86 MHC class II molecules
CD40 High CD80 High CD86 High MHC class II molecules
DCs and immune tolerance
In addition to mounting an immune response, DCs play an essential role in maintaining immune tolerance in the thymus and secondary lymphoid organs. In the thymus, binding of self-antigens to DC receptors and subsequent presentation to developing immune cells result in receptor editing, rearrangement of TCR-α chain, that results in the selection of non-auto-reactive immune cells (Kouskoff and Nemazee, 2001; Singh and Schwartz, 2006). Failure of receptor editing may be compensated by deletion of thymocytes with high affinity for self-antigen (Zhang et al., 2005).
In spite of the multiple mechanisms protecting against autoimmunity, a non-negligible number of potentially auto-reactive lymphocytes escape to the periphery. To prevent auto-reactivity in secondary lymphoid organs, T-cell tolerance is attained by a mechanism known as receptor revision, which implicates secondary rearrangement of TCR-β (Kouskoff and Nemazee, 2001). Again, other mechanisms like T-cell deletion and T-cell anergy are implicated in the induction of peripheral tolerance (Powell, 2006; Romagnani, 2006). DCs render T-cell deletion possible by several means. For instance, signaling through CD95 (Fas-Apo-I) on T cells upon interaction with Fas ligand (Fas-L) on a subclass of DCs induces apoptosis of auto-reactive T-cell clones (Kurts et al., 1998; Suss and Shortman, 1996). In accordance with these data, mutations in the Fas receptor or Fas-L genes have been reported to be associated with defective apoptosis and subsequent development of autoimmune diseases in both human (Oliveira and Fleisher, 2004) and experimental mouse models (Stranges et al., 2007). Induction of apoptosis after T-cell priming is likely a limiting factor for T-cell activation and a measure for maintaining self-tolerance and preventing autoimmunity (Chen et al., 2006). The expression of indolamine 2,3 dioxygenase (IDO) enzyme by DCs is another contributing factor for T-cell deletion. Degradation of tryptophan by this enzyme results in metabolites, like kynurenin, which exert a cytotoxic action on T cells (Mellor and Munn, 2004; Munn et al., 2002). The role of DCs in the induction of T-cell anergy could be attributed to the absence of co-stimulatory signals provided by DCs and/or by the secretion of inhibitory cytokines like IL-10 (Lechler et al., 2001).
Sequences of targeting immature dendritic cells
Under steady-state conditions, DCs are implicated in capturing apoptotic cells. Notably, apoptosis is a programmed mode of cell death that is not accompanied by an inflammatory process. Consequently, DCs are kept in situ without maturation (Schwartz, 2003; Schwartz et al., 1989). The expression of certain DC receptors has been reported to play a key role in mediating apoptotic cell-induced inhibition of DCs. In this regard, the recognition of oxidized phosphatidylserine and phosphatidylcholine on the surface of apoptotic cells (Williams et al., 2005) by CD36 molecule on the DCs was shown to enhance the IL-10 secretion and to reduce the production of TNF-α and IL-1β, thus contributing to peripheral tolerance and impairment of DC maturation (Pluddemann et al., 2007). In addition, interaction of the receptor tyrosine kinase Mer (MerTK) expressed on DCs with its ligand Gas-6 (growth-arrest-specific gene) on apoptotic cells results in the synthesis of anti-inflammatory cytokines such as IL-10 and TGF-β by DCs, with subsequent immunosuppressive effect (Wallet et al., 2008).
In addition, DCs are kept in the immature state owing to the phagosomal pH which, in contrast to macrophages, is neutral or even alkaline (Gatti and Pierre, 2003). Such pH is unfavorable for activating DC proteases, resulting in reduced antigen processing and slow degradation of the MHC invariant chain (Ii-chain), an event that interferes with antigen presentation with failure of effector T-cell activation (Pierre and Mellman, 1998; Trombetta et al., 2003).
Semi-immature dendritic cells and maintenance of the immune tolerance
When immature DCs are subjected to endogenous or exogenous stimuli, the expression of MHC class II and co-stimulatory molecules like CD80/CD86 is enhanced and then semi-immature DCs are likely to develop. Nevertheless, semi-immature DCs are important players for keeping tolerance. This function is attributable, in part, to lack of pro-inflammatory cytokine secretion by DCs and to up-regulation of inhibitory molecules like B7-H3 and B7-H4 molecules on their surface. These molecules are members of programmed death ligand family (PDL) (Kryczek et al., 2006b; Mahnke et al., 2007b) that interfere with Th1 cell response and induction of immune tolerance (Kryczek et al., 2006a; Suh et al., 2003). Furthermore, semi-immature DCs up-regulate lymphocyte activation gene-3 (LAG-3), an MHC class II binding CD4 homologue. Engagement of LAG-3 molecules interferes with the differentiation of fully competent DCs, thus limiting the magnitude of an ongoing immune response (Buisson and Triebel, 2005; Workman and Vignali, 2005). Contrary to immature DCs, semi-immature DCs are mobile cells whose migration is favored by enhanced expression of CCR7 molecules. The latter, directs DCs to T-cell-dependent zone in the draining lymph nodes (LN) being guided by CCL19 and CCL21 chemokines (Forster et al., 2008). Nevertheless, in the steady-state conditions, tolerance to self-antigen is ensured by the presence of another partner, namely, T regulatory (Treg) cells. In this regard, the constitutive expression of the transcription factor Foxp3 characterizes the naturally occurring thymic-derived CD4+ CD25+ T regulatory (Treg) cell subset, while the expression of the immunosuppressive cytokines IL-10 and TGF-β identifies the antigen-derived Treg cells (Yamazaki et al., 2003). The direct interaction between DCs and Treg cells via CD80/CD86 and CTLA4, respectively, results in Treg cell proliferation in an antigen-dependent manner and secretion of IL-10 and TGF-β that mediates T-cell suppression (Houot et al., 2006).
Indeed, a mutual relationship exists between Treg cells and semi-immature DCs where the presence of Treg cells is crucial for IL-10 induction and surface inhibitory molecules expression on semi-immature DCs (Mahnke et al., 2007a; Wei et al., 2006). The dual presence of semi-immature DCs and Treg cells interfere with T-cell priming and maintains tolerance. For instance, induction of oral tolerance in mice was reported to be associated with increased DC numbers in Peyer’s patches, augmented IL-10 production, and induction of CD4+ CD25+ Treg cells. Interestingly, adoptive transfer of these tolerogenic DCs resulted in the suppression of severe arthritis (Min et al., 2006).
Mature DCs and induction of the specific immune response
The induction of antigen-specific T cells is dependent on the level of antigen presentation, the expression of co-stimulatory molecules such as CD80 and CD86, and on cytokine secretion by activated DCs (Sporri and Reis e Sousa, 2005). The latter cells express a variety of receptors that serve in antigen capturing, activation and migration to regional LN. In the context of receptor-induced signal delivery, expression of Toll-like receptors (TLRs) by DCs is particularly important. These are pattern recognition receptors (PRR) constitutively expressed on the plasma membrane as well as the phagosome/endosome compartment of DCs (Villadangos et al., 2005). Their activation is influenced by pathogen-associated molecular patterns (PAMPs). TLR downstream signaling leads to enhanced intra-nuclear translocation of NF-κB and secretion of pro-inflammatory cytokines such as IL-6, IL-1β, IL-12p40/p70 and TNF-α (Banchereau and Steinman, 1998). Furthermore, TLR activation is accompanied by several events like actin polymerization (Lindquist et al., 2004), expression of the actin-bundling protein fascin which is involved in antigen presentation by mature DCs (Al-Alwan et al., 2001; Ross et al., 2000) and phagosome maturation and fusion with the lysosomal compartment (Blander and Medzhitov, 2006) where antigen degradation is accelerated by enhanced lysosomal acidification (Trombetta et al., 2003). Concomitantly, activated DC proteases act on the Ii-MHC complex, leading to CLIP (class II invariant chain-associated peptide) dissociation and peptide-MHC class II complex formation. This complex is delivered to the plasma membrane at the level of the immunologic synapse, accompanied by co-stimulatory molecules like CD86 (Barois et al., 2002; Kleijmeer et al., 2001; Trombetta et al., 2003).
In addition to PRR signaling, DC activation is also mediated by tissue-derived environmental factors and a variety of cells such as NK, the invariant natural killer T cells (iNKT) and γδ T cells (Hayday and Tigelaar, 2003; Reis e Sousa, 2004). Following activation, DCs generate signals that influence the differentiation of antigen-specific T cells. Accordingly, naïve CD4+ T cells can be differentiated into IFN-γ secreting Th1 cells, into Th2 cells characterized by the secretion of IL-4, IL-5 and IL-13, and into Th17 cells identified by the production of IL-17 (Diebold, 2008; Weaver et al., 2006). In addition to peptide-MHC-class II complex interaction with TCR, peptides sorting in the DC cytosol can prime CD8+ T cells together with MHC-class I peptides. This cross-presentation is a unique function of DCs that requires efficient proteasome activity and functional TAP (transfer-associated with antigen presentation) (Dudziak et al., 2007; Rodriguez et al., 1999). The ongoing immune reaction is controlled by a variety of regulatory T cells, such as natural CD25+ regulatory cells, Th3 cells secreting high levels of TGF-β, and Treg cells producing IL-10 (Yates et al., 2007). Of note, IL-2, IL-4, IL-7 and IL-15 were proven to maintain the optimal regulatory function of human CD4+ CD25+ T cells in a PI3K-dependent manner (Valencia et al., 2006). Nevertheless, binding of TNF-α to its receptor TNFR11 constitutively expressed by Treg cells will lead to a decrease in Foxp3 expression and reversal of Treg cells’ suppressive potential (Chen et al., 2007).
Dendritic cell subsets in human skin
Langerhans cells (LGC) constitute a subset of DCs that reside in their immature state in the skin epidermis. LGC are CD1a+ high, Langerin+ (CD207) and DC-LAMP+ (CD208). However, three APC classes exist in the healthy human dermis, namely, CD14+ and CD1a+ dermal APCs, and migratory dermal LGC (Angel et al., 2007).
Dermal Langerin+ DCs migrate from the skin to the LN after inflammation and in the steady state, and represent the majority of Langerin+ DCs in skin draining (LN) (Bursch et al., 2007; Poulin et al., 2007). Migration of LGC to the draining LN is controlled by their secretion of collagenase, the expression of certain chemokine receptors like CCR6 that interacts with CCL20 secreted by epithelial cells (Le Borgne et al., 2006) CCR7 interacting with CCL21 secreted by the endothelium of lymph vessels, and CCL19 highly expressed in the lymph node deep cortical zone (Forster et al., 2008).
Reaching the deep cortical zone of the LN, LGC are referred to as interdigitating DCs that play a central role in T cell priming and in eliciting a specific cellular immune response (Henrickson et al., 2008; Le Borgne et al., 2006). Nevertheless, dermal DCs preferentially migrate in proximity to B cell follicles where they function as helper cells for B cell through the secretion of cytokines like IL-6 and IL-12 and by supporting the development of a subpopulation of CD4+ T cells known as T follicle helper cells that support immunoglobulin production from B cells (Caux et al., 2000; Kissenpfennig et al., 2005). A summary of events is provided in Fig. 1.
DCs in melanoma
In melanoma, the density and distribution of mature DCs in tumor tissue and their relation with infiltrating T lymphocytes are of prognostic value. Immuno-histological examination of melanoma skin lesions revealed that immature DCs are distributed in the melanoma cell nests as well as in the surrounding stroma, while DCs expressing maturation markers such as DC-LAMP (CD208) and CD83 are confined to the peritumoral areas, associated with T lymphocytes (Ladanyi et al., 2007; Ueno et al., 2007). Indeed, the peritumoral density of DCs and the state of T lymphocyte activation are correlated with melanoma tumor thickness and patient’s survival, thereby favoring the use of these parameters as a predictor of treatment response in patients who receive immunotherapy (Kobayashi et al., 2007; Simonetti et al., 2007). Of note, melanoma is associated with an important anomaly in the blood vessels within the tumor boundaries. Such vessels do not express the leukocyte homing receptors E-selectin (CD62E), P-selectin (CD62P) and intercellular adhesion molecule-1 (ICAM-1; CD54). This could explain, in part, the defective recruitment of effector T cells to the melanoma tumor and impaired tumor-specific CTL response (Weishaupt et al., 2007).
Immune evasion of melanoma cells
Regression of melanoma requires a complex interaction between tumor cells and players of the immune system, including DCs and T cells (Table 2). However, melanoma cells often escape the tumor-specific T-cell response by down-regulation of HLA class I and class II expression (Al-Batran et al., 2005; Belli et al., 2002) and loss of melanoma-associated antigens (Maeurer et al., 1996; Trefzer et al., 2004). The expression by tumor cells of B7-H1 molecules (Dong et al., 2002) and Fas-L (Shukuwa et al., 2002) is a promoting factor for apoptosis of infiltrating T lymphocytes and contributes to the immune evasion of tumor cells. Other inhibitory molecules, B7-H3 and B7-H4, expressed on melanoma cells (Zang and Allison, 2007) are additional factors contributing to impaired T-cell functions. The secretion by the tumor cells of inhibitory cytokines like TGF-β (Rodeck et al., 1999; Teicher, 2001), IL-10 (Chen et al., 1994) also favors tumor escape. In this regard, tumor-derived TGF-β was reported to induce apoptosis of CD4+ T cells and DCs in the sentinel LN, an event that precedes metastasis of tumor cells to the primary draining LN (Ito et al., 2006). In accordance with this finding, over-expression of Bcl-2 and Bcl-xL was reported to oppose the apoptotic effect of TGF-β in advanced melanoma (Esche et al., 2001).
Enhanced production of deleterious metabolites such as lactic acid
Prevalence of immature phenotype
Retention of captured antigens,
Defective inflammatory cytokine secretion
Cytotoxic T cells
↓ T, NK and LAK cell killing activity
CD4+ T cells
Polarization towards Th2 with ↑ secretion of IL-4 and IL-13
Dominant Treg functions with ↑ secretion of IL-10
Malignant cells produce massive amounts of TGF-β and become autocrine and/or paracrine growth-stimulated by this cytokine (Gold, 1999; Javelaud et al., 2008). In experimental melanoma, TGF-β promotes osteolytic bone metastases by stimulating the expression of prometastatic factors such as parathyroid hormone-related protein (PTHrP), interleukin-11 (IL-11), the chemotactic receptor CXCR4, and osteopontin. Inhibition of TGF-β signaling via stable over expression of Smad7 in melanoma cells impaired bone metastasis (Javelaud et al., 2007). Aside from direct effects on tumor cells, increased expression of TGF-β is also associated with the suppression of cytotoxic immune cells, leading to selective advantage for tumor cell survival. More specifically, high levels of TGF-β have been associated with impaired functions lymphocyte-activated killer (LAK) cells (Hsiao et al., 2004) and inhibition of NK cell lytic activity by selective down-modulation of activating receptors, such as NKG2D (Lee et al., 2004).
Metastatic melanoma often secretes large amounts of IL-10. This cytokine polarizes T cells towards Th2 type (Enk et al., 1997) and down-regulates the expression of CD1a molecules on DCs, thus interfering with DC capacity for antigen presentation (Gerlini et al., 2004). It is worthy to mention that tumor growth is characterized by high production of lactic acid because of enhanced glycolysis. Melanoma-derived lactic acid is another important factor contributing to tumor escape mechanisms through modulation of DC phenotype (Gottfried et al., 2006) and suppression of proliferation, cytokine production and cytotoxic activity of human T lymphocytes (Fischer et al., 2007). Finally, the presence in the microenvironment of metabolic enzymes such as IDO (Hwu et al., 2000; Mellor and Munn, 2004), arginase (Bronte and Zanovello, 2005; Zea et al., 2005) and inducible nitric oxide synthase (Blesson et al., 2002) leads to deleterious effect on T-cell functions.
In the context of tumor evasion, angiogenesis is one of the processes that enhance migration and tumor metastasis. In melanoma, high expression levels of vascular endothelial growth factor (VEGF) enhance the proliferation and survival of endothelial cells by differential expression of anti-apoptotic genes (bcl-2) and activation of matrix metalloproteinase-3 (MMP3) (Mahabeleshwar and Byzova, 2007; Tas et al., 2008). Furthermore, the negative association between elevated levels of VEGF and DC differentiation (Osada et al., 2008) might explain, in part, the impaired anticancer immunity in advanced malignancies.
DCs and melanoma progression
It is important to note that peritumoral DCs are immature. This status may be attributed to several factors including the absence of inflammatory stimuli that accompany the earlier phase of tumor growth and retention of captured melanoma antigens. The latter explanation is based on a recent observation in breast cancer whereby the mucin-1 antigen captured by DCs mannose receptors is retained in the early endosomes and poorly expressed on the surface of DCs, thereby impeding the generation of a specific immune response (Hiltbold et al., 2000; Vlad et al., 2004).
Functionally, immature DCs have been reported to limit T-cell activation and thus contribute positively to tumor progression. In this regard, the interaction of DCs with PD-1 and CTLA-4 molecules expressed on T cells results in suppression of CTL tumor-specific response (Probst et al., 2005). In addition, tumor antigens cross-presented to CD8+ cells via immature DCs lead to improper CTL priming with subsequent uncontrolled tumor growth (Fuchs and Matzinger, 1996; van Mierlo et al., 2004). Furthermore, deficient IL-12 secretion by immature DCs tends to polarize T cells towards Th2 type with dominant secretion of IL-4 and IL-13 that participate in tumor growth and proliferation (Aspord et al., 2007; Kukreja et al., 2006). The expression of LAP (latency-associated peptide) by immature DCs was reported to bind TGF-β, thus, limiting T-cell activation (Gandhi et al., 2007). Indeed, immune tolerance by immature DCs could be also mediated by the induction of suppressor Treg cells characterized by the expression of CD4, CD25 and Foxp3 (Shevach et al., 2006). Finally, defective apoptosis of tumor cells could be attributed, in part, to down-regulation of TNFR-apoptosis-induced ligand (TRAIL) on DCs (Kemp et al., 2003).
Treatment of melanoma with an emphasis on DC-based immunotherapy
Malignant melanoma is an aggressive skin disease with high incidence mortality. Management of cutaneous melanoma requires careful clinical examination, skin biopsies and precise immuno-histological analysis. The latter is essential for tumor staging and for the subsequent decision of either medical and/or surgical intervention. In this context, excision margin of primary melanoma is dependent on the thickness of the lesion. Different comparative studies had recommended that the excision margin of 1 cm is adequate for lesions of thickness <1 mm (Veronesi and Cascinelli, 1991). However, for lesions up to 4 mm in thickness, an excision margin of 2 cm is recommended (Khayat et al., 2003; Ringborg et al., 1996). For lesions > 4 mm thick, a 3 cm margin excision is acceptable (Jack et al., 2006). Concerning metastatic melanoma, the most frequent sites for metastasis include skin, LN, lung, brain and the gastro-intestinal tract (Allen and Coit, 2002; Wong and Coit, 2004). Surgical resection of metastatic lesions has been found to improve the patient survival for up to 1 yr-disease-free interval (Hersey et al., 1983). This is attributed to decrease of the immunosuppressive effect of the tumor (Hsueh et al., 2000) and enhancement of the specific humoral and cellular immune response (Cormier et al., 2005).
It is worthy to mention that sentinel lymph node (SLN) biopsy is recommended for all patients with primary tumor thickness above 1 mm, and for patients with ulcerative lesions. Such biopsy is useful for detecting occult metastasis in the draining lymph node, which is indicative of lymph adenectomy (Gershenwald et al., 1999; Thomas and Clark, 2004). However, SLN biopsy might carry the risk of in-transit-metastasis where melanoma cells might present in the tissues underlying the epidermis (Estourgie et al., 2004).
Aside from these conventional approaches for treating melanoma, DCs have been thought to have a promising therapeutic potential. In this context, DC-based vaccines are prepared by in vitro culture of mobilized CD34+ hematopoietic progenitors in the presence of GM-CSF, FLt3 and TNF-α (Banchereau et al., 2001; Caux et al., 1997) or by culturing blood-derived monocytes CD14+ in the presence of GM-CSF and IL-4 (Nestle et al., 1998; Thurner et al., 1999). Maturation of DCs is achieved by the addition to culture media of pro-inflammatory cytokines like TNF-α, IL-1β and prostaglandin E (de Vries et al., 2003). DC activation could also be achieved by CD40 ligation or TLR triggering by LPS or CpG with a positive impact on the generation of CD8+ CTL cells (Matasic et al., 2001). Such generated DCs are then loaded with melanoma antigens. The latter are grouped into three classes of molecules:
• Differentiation antigens which are proteins expressed on melanocytic cells. They include MART-1/Melan-A, Tyrosinase, Tyrosinase-related protein 1(gp75) and Tyrosinase-related protein 2 (gp100).
• Germ cell proteins, expressed on other types of tumors. They are represented by MAGE-1, MAGE-3, GAGE-1, GAGE-3, NY-ESO-1, NY-ESO-3 and BAGE.
• Gangliosides, surface glycolipid antigens highly expressed on melanoma cells. They include GM2, GD3 and GD2 (Chapman, 2007).
These tumor antigens are derived from either autologous (Thurner et al., 1999) or allogeneic tumor cell lysates (Chang et al., 2002; Mahdian et al., 2006). DCs loaded with allogeneic shared melanoma antigens like MAGE-3, gp100, tyrosinase and MART-1 have been shown to elicit cross-priming of naïve CD8+ T cells against autologous tumors (Saleh et al., 2005). Other strategies include direct peptide antigen binding to DCs, without the need for antigen processing by DCs (Berard et al., 2000), or transfection of DCs with cDNA encoding tumor antigens like MART-1 (Lee et al., 1999) or transfection with autologous tumor mRNA (Gordon et al., 2004). Finally, loaded DCs are injected either in situ intradermally or subcutaneously (Kyte et al., 2006), or via intra-lymphatic route where DCs present antigens to T cell in the context of MHC with the generation of CTL-tumor specific immune response (Candido et al., 2001).
The results of DC-based vaccines rely on several critical factors such as the maturation and activation status of DCs, the level of inhibitory cytokines and/or cell-mediated suppression as well as tumor mass. In this respect, DC-based vaccines carried out in advanced melanoma showed regression of individual tumor masses, however, complete remission was not achieved (Grover et al., 2006).
Recently, autologous DCs pulsed with HLA-DP4-restricted peptide epitope of the melanoma-associated antigen NY-ESO-1 were shown to be helpful for in vitro generation of autologous antigen-specific CD4+ T-cell clones. The infusion of the expanded clone into a patient with advanced melanoma resulted in the induction of melanoma-specific T-cell responses that mediated durable and complete tumor regression (Hunder et al., 2008).
Also, recently, a combination therapy with CpG (a DC activator) and JS1-124 (a STAT3 inhibitor) in melanoma-bearing mice generated synergistic anti-tumor effects and led to significant reduction of immuno-suppression, as evidenced by lower intra-tumoral level of VEGF and TGF-β and decreased number of Treg cells CD4+ CD25+ Foxp3+ in the regional LN (Molavi et al., 2008).
Interestingly, other recent studies reported that intraperitoneal injection of polyinosinic-polycytidilic acid (polyI:C), a Toll-like receptor 3 agonist (Jiang et al., 2008) or the combined administration of imatinib mesylate and IL-2 (Mignot et al., 2008) have led to regression of murine melanoma. The observed antitumor efficacy was attributed, in part, to the expansion of a unique NK cell subset described as IFN-producing killer dendritic cells. In addition, other trials are emerging, with the aim to target the tumor antigen-specific regulatory T cells (Tregs). Depletion of the Treg population has proven to break tolerance against metastatic melanoma in humans (Powell et al., 2007; Rasku et al., 2008).
In conclusion, DCs are professional APC with a potential in determining the fate of antigen specific T cells and ultimate immune response. The presence of an adequate number of activated DCs in the tumor area is a critical determinant for eradication of malignant cells by limiting their immune evasion. Yet, although DC-based vaccines seemed to have an encouraging therapeutic potential with proven antitumor efficacy in animal models, several hurdles remain in the human setting. Clearly, the outcome of DC-based therapeutic approaches relies strongly on the stage of melanoma, as complete remission has never been achieved in advanced stages of the disease. Nevertheless, optimization of the host immune response could be enforced by a multimodal immunotherapy whereby DC-based vaccine could be combined with another therapeutic arm, one of them being the depletion of suppressor Treg cells. The recent evidence that autologous CD4+ primed in vitro by melanoma antigen-pulsed DCs resulted in complete remission in a patient with advanced melanoma, underlies the fruitful cooperation between the different players of the immune system and the tumor cells. Such finding paves the way for further trials capitalizing on such synergy. Other approaches to overcome clinical obstacles facing DC-based immunotherapies may also rely on combinatorial treatments with cytokines or immuno-modulatory agents such as CTLA-4 or PD-1 inhibitors.