The skin immune system
The skin, being the body’s most exposed barrier, has particular innate and adaptive immune mechanisms. The skin immune system is a complex interplay, as approximately half of the cell types in the skin are involved in the immune response. The major site of immunological activity resides in the papillary dermis, where T cells, monocytes and dendritic cells (DCs) are preferentially distributed (Bos and Luiten, 2009). Keratinocytes, mast cells, macrophages, Langerhans cells (LCs) and dermal DCs are the first barrier of the skin immune system. Keratinocytes have the capacity to produce a broad range of immune-stimulating and downregulating cytokines [interleukin (IL)-1, IL-6, IL-10, IL-12, regulated upon activation, normal T cell expressed and secreted, interferon (IFN)-α/β, macrophage colony-stimulating factor (M-CSF), tumor necrosis factor (TNF)-α, granulocyte macrophage colony-stimulating factor, granulocyte colony-stimulating factor, transforming growth factor (TGF)-β, β-defensin]. Various stimuli trigger keratinocytes to produce some of these molecules capable of controlling cell proliferation, migration of cells and activation of lymphocytes, DCs, mast cells and macrophages (Kupper and Fuhlbrigge, 2004; Woods et al., 2005). Besides protection against parasites, mast cells play a complex role in the innate and adaptive immune system. The initiating activity of mast cells on the innate immune response is suggested by the expression of multiple pathogen recognition receptors, such as Toll-like receptors (TLR) (Herschko and Rivera, 2010).
The TLR family is considered an important defense mechanism against various infectious pathogens and is a major player in the pathophysiology of inflammatory and neoplastic disorders of the skin. Toll-like receptor expression has been reported on keratinocytes, immune cells (T lymphocytes, DCs, LCs, mast cells, monocytes, eosinophils, neutrophils), vascular endothelial cells, fibroblasts and adipocytes in the skin. Binding of the TLRs by factors such as pathogen-associated molecular patterns (PAMPs), which are shared by a large group of pathogens, results in the activation of the nuclear factor (NF)-κB signaling pathways. In addition, TLRs can sense endogenous molecular signatures of tissue damage, also known as ‘danger-associated molecular patterns’ (DAMPs), which includes high-mobility group box-1 (HMGB1), extracellular ATP, heat shock proteins and extracellular matrix breakdown products. NF-κB is a major inducer of the inflammatory response and links the innate to the adaptive immune system. Activation of NF-κB leads to the production of inflammatory cytokines, chemokines, antimicrobial peptides, matrix metalloproteinases, nitric oxide synthase and the upregulation of adhesion molecules [E-selectin, P-selectin, intercellular adhesion molecule 1 (ICAM-1) and vascular cell adhesion molecule 1 (VCAM1)]. Besides a direct cytotoxicity to pathogens or aberrant cells, both innate and adaptive immune cells are recruited to the site of inflammation (Kupper and Fuhlbrigge, 2004; Miller and Modlin, 2007).
The primary adaptive skin immune surveillance consists of antigen-presenting cells (epidermal LCs, dermal DCs, macrophages) which phagocytose foreign antigens and present them to lymphocytes in the skin-draining lymph nodes. Dendritic cells are key regulators of the skin immune system and form a bridge between the innate and adaptive immune system (Bos and Luiten, 2009). Langerhans cells, a subtype of DCs residing in the epidermis, are considered potent inducers of T-cell responses. Their maturity state is divided into three stages that represent the phases needed to capture and present antigens, also called the ‘LC paradigm’. Immature LCs form an organized network in the suprabasal layer of the epidermis to detect pathogens invading the epidermis. Immature LCs are characterized by a low major histocompatibility complex (MHC)-II surface expression and no co-stimulatory molecules (e.g. CD80, CD86) and have endocytotic, migratory and antigen-presenting capacities. In a second stage, LCs become activated due to microbial products and inflammatory cytokines (such as TNF-α) largely derived from surrounding keratinocytes. Subsequently, LCs lose E-cadherin expression, express chemokine receptors (e.g. CCR7) and secrete metalloproteinases, allowing migration to the lymph node (Igyarto and Kaplan, 2010). During the migration process, LCs become functionally mature by acquiring co-stimulatory molecules and producing inflammatory cytokines. Mature LCs are potent primers of naïve T cells, which acquire subsequently a skin-homing capacity, proliferate clonally and mature into effectory and memory T cells (Kupper and Fuhlbrigge, 2004). Immature LCs presenting (self-)antigens in the lymph node without co-stimuli (such as CD80, CD86, MHC–peptide complexes, chemokine receptors) induce peripheral tolerance (Hawiger et al., 2001). In the dermis, DCs and macrophages have a remarkable heterogeneity and functionality, ranging from DCs expressing BDCA-1 (CD1c) and DC-SIGN (CD209) to macrophages expressing CD163 (Nestle and Nickoloff, 2007). The exact function of the recently identified subset of langerin expressing dermal DCs which accounts for up to 30–60% of the Langerin+ DCs in the skin-draining lymph nodes of mice, remains to be elucidated (Romani et al., 2010). Dermal DCs and LCs not only have an antigen-presenting function but also skew the effector immune reaction by sensing ‘danger signals’ and are therefore considered cutaneous sentinels (Girardi, 2007). Plasmacytoid dendritic cells (pDCs) are a subtype of DCs, present in the blood, lymphoid tissues and inflammatory skin. ‘Resting’ pDCs are poor antigen-presenting cells and are even considered to be tolerogenic (Conrad et al., 2009; Swiecki and Colonna, 2010). However, pDCs are specialized in sensing nucleic acids from viral pathogens through endosomal TLRs such as TLR3, 7, 8 and 9 and induces the production of massive amounts of type-I interferons. Interestingly, although endogenous DNA cannot by itself cause pDC activation, damage to keratinocytes results in the release of the antimicrobial peptide LL37 (Cathelicidin antimicrobial peptide). When complexed and aggregated with LL37, self-DNA released by damaged cells can trigger TLR9 and produce a robust pDC-derived IFN-α response. This mechanism is thought to underly the pathological pDC-dependent auto-immune activation of T cells within psoriatic lesions (Lande et al., 2007). This is an illustration of how a DC subset can translate innate danger signals into an (inappropriate) adaptive immune response.
In the secondary immune surveillance phase, T lymphocytes in the draining lymph node become activated, express the appropriate homing receptors [such as cutaneous lymphocyte-associated antigen (CLA), lymphocyte function-associated antigen-1, very late activation antigen-4 (VLA-4) and chemokine receptors CXCR3, CCR4 and CCR10] and bind the skin-specific receptors on post-capillary venules [E-selectin, CC-chemokine ligand 17 (CCL17) and ICAM-1]. Once at the target site, the T lymphocytes become functionally active due to the inflammatory environment and influence the nature and activation state of the infiltrate by modifying the chemokine balance and producing inflammatory cytokines (e.g. IFN-γ, TNF-α). In contrast, a subpopulation of T cells, the so-called regulatory T cells (Tregs), dampen the inflammation and induce tolerance. In the tertiary phase, the antigen-specific memory cells migrate to other lymph nodes, providing a rapid immune response in case of second exposure to the same antigen (Kupper and Fuhlbrigge, 2004).
The melanocyte interplay with the immune system
Melanocytes, originally derived from the neural crest, are situated just above the basal membrane. The best known function of melanocytes is the production of melanin pigment. The dendritic structure of melanocytes, creating extensive contacts, especially with keratinocytes, suggests a complex signaling network (=the melanocyte-epidermal unit) (Lu et al., 2002).
Common acquired melanocytic nevi are neoplasms arising mostly during the first 3 decades of life, whereas congenital nevi are present from birth (Gelfer and Rivers, 2007). Despite their usually benign nature, expression of melanoma-associated antigens occurs (Cui and Willingham, 2004). Melanoma, the malignant counterpart, expresses a broad range of tumor-associated antigens roughly divided into four classes: cancer/testis-antigens [e.g. melanocyte-associated antigen (MAGE), B melanoma antigen (BAGE), G-antigen (GAGE), New York esophageal squamous cell carcinoma 1 (NY-ESO-1)], melanocyte differentiation antigens [e.g. tyrosinase, Melan A/melanoma antigen recognized by T cells (MART-1), tyrosinase-related protein 1 and 2 (TYRP-1 and TYRP-2), glycoprotein 100], tumor-specific mutated proteins (e.g. cycline-dependent kinase 4, β-catenin, fibronectin) and a group of aberrantly translated intronic sequences [e.g. melanoma-associated antigen (mutated)-1 (MUM-1), p15, N-acetylglucosaminyltransferase V] (Dranoff, 2009; Ram and Shoenfeld, 2007).
Spontaneous immune-mediated regression of nevi is particularly common in childhood and adolescence and could be a reflection of an active immune system that eliminates normal and neoplastic melanocytes and prevents tumor development (Cui and Willingham, 2004). Immunodeficiency is associated with a higher incidence of melanocytic nevi (Baron and Krol, 2005). During immunosuppression, numerous eruptive nevi may develop, accompanied by an increased lymphocytic infiltration (Piaserico et al., 2006). Involution of such nevi in transplant patients after suspension of immunosuppressive therapy has been reported. This illustrates the skin immunosurveillance against melanocytic proliferation (Zattra et al., 2009). Moreover, a decrease in the number of nevi has been reported in patients with chronic-graft-versus-host disease, suggesting that an increased inflammation in the skin may stimulate the immune reaction against nevi (Andreani et al., 2002). A 4- to 7-fold increase in melanoma risk has been reported in the immunosuppressed population compared to age- and gender-matched controls (Jensen et al., 1999).
The hurdles of immunotherapy for melanoma
Melanoma is considered a highly immunogenic tumor, accounting for 11% of spontaneously regressing tumors while representing only 1.8% of the total cancer burden (Everson and Cole, 1966). Spontaneous regression has been reported in all progression stages. A higher frequency of tumor-infiltrating lymphocytes (TILs) is associated with a better prognosis in melanoma (Oble et al., 2009). However, although melanoma-specific T-cell responses (e.g. Melan-A/MART-1 and gp100) have been detected in melanoma lesions, complete spontaneous or therapy-induced regression in advanced disease is a rare phenomenon.
Increasing evidence indicates that the tumoral cells skew the immune system to a more immunosuppressed state and acquire some mechanisms to escape an efficient immune response. Tumors are able to recruit and stimulate Tregs under the influence of certain, yet unknown factors (Serafini et al., 2006). These cells are important in controlling autoimmunity in physiological conditions, but they also appear to suppress a cytotoxic T-cell-mediated response against autoantigens on the tumor (Viguier et al., 2004). Several immunomodulating proteins can be expressed at the tumoral level. Tumoral overexpression of Programmed Death Ligand-1 (PD-L1) and, to a smaller degree, Programmed Death Ligand-2 (PD-L2) can diminish the antigen-receptor signalling by binding to the receptors on T and B lymphocytes (Freeman et al., 2000). Tumoral PD-L1 expression has recently been linked to a worse prognosis in melanoma (Hino et al., 2010). Loss of MHC-I in melanoma is associated with disease progression (Chang et al., 2005). Melanoma cells are also able to express MHC-II, but the costimulating molecules CD80 and CD86 are often absent, which induces tolerance (Denfeld et al., 1995). Ectopic expression of HLA-G, a nonclassic MHC-I molecule, can protect melanoma cells against natural killer (NK)-cells (Paul et al., 1998). Expression and secretion of Fas ligand by melanoma cells leads to apoptosis of immune cells (Andreola et al., 2002). In more advanced melanoma stages, the absence of P-selectin expression could play a role in inhibiting an efficient antitumoral response (Nooijen et al., 1998).
In spite of the existence of different immunotherapeutic interventions in melanoma, significant clinical responses are only observed in a subset of patients. Currently, these responders cannot be identified in advance. The pathogenesis of regression in melanocytic lesions is a complex phenomenon. Insights into the immune responses to different melanocytic lesions may help to identify the mechanisms and triggers of an efficient immunological elimination of melanocytes.