Classical DCs (cDC) are a broad category of LN/spleen resident DC that have a high turnover and are constantly replaced by blood-derived precursors. cDCs develop from a hematopoietic lineage distinct from other leukocytes requiring the transcription factor zbtb46 and the cytokine Flt3L [53–56]. Importantly, cDC can be divided into two groups that were initially demarcated as “lymphoid” and “myeloid” DC lineages but are now best described as IRF8 dependent and IRF4 dependent cDC. IRF8 cDC require the transcriptional factor Basic leucine zipper transcription factor ATF-like (BATF3), interferon regulatory factor 8 (IRF8), and inhibitor of DNA protein 2 (ID2) (Table 9.2) [57–59]. They express high levels of Flt3, proliferate after administration of Flt3L and are nearly absent in Flt3L−/− mice [55, 60]. They can be best identified based on the expression of CD8α but not CD8β and are often referred to as CD8+ cDC. They also express no or low levels of integrin CD11b as well as other macrophage defining markers [61] and represent 20–40 % of the secondary lymphoid organ resident DC [62, 63]. IRF4 cDC similarly proliferate in response to Flt3L and are reduced in Flt3- and Flt3L-deficient mice, but to lower levels in comparison to IRF8 cDC [55, 60]. They depend on transcriptional factor IRF4 and zbtb46, but not on BATF3, IRF8 or ID2 for their development [57, 58, 64]. They can be identified based on their expression of integrin CD11b + and the absence of CD8α.

The best defined DC subsets in the dermis can be classified as CD103+ dermal DC (dDC) or CD11b + dDC. CD103+ dDC are homologues of the IRF8/CD8+ cDC. They express the integrin CD103 that binds e-cadherin expressed by epithelial cells but lack expression of CD8a [58]. CD103+ cDC lack the macrophage markers CD11b, CD115, CD172a, F4/80, and CX3CR1 [65]. They also depend on BATF3, IRF8 and ID2 [57–59]. In addition to Flt3L, CD103+ cDC also require Csf-2 for their development [66, 67]. This DC subset is fairly infrequent in the dermis but they migrate and are replenished from bone marrow precursors at a high rate [68].

CD11b + dDC comprise the majority of DC in the dermis but are less well studied than CD103+ dDC [68]. They are homologous to IRF4 cDC and can be identified based on expression of CD11b and the absence of CD103. They require cytokine Csf-1, IRF4 and Flt-3 for development [51, 66, 69]. Importantly, these dDC are often confused with macrophages that are also abundant in the dermis and express CD11c, MHC-II and CD11b. CD11b + dDC can be distinguished based on the absence of FcγRI (CD64) and MerTK expression [70].

Langerhans cells constitute the sole APC population in the epidermis under steady-state conditions. Murine LCs are uniformly CD11b + F4/80+ and lack CX3CR1 expression [51]. They express the C-type lectin Langerin (CD207), which is involved in the formation of Birbeck’s granules, a pathognomonic marker for LC [71]. LC account for 3–5 % of epidermal cells and stand apart from the other DC subsets through their unique ontogeny and homeostatic properties [72]. During ontogeny, LC precursors seed the epidermis first from hematopoetic precursors in the yolk-sac and then from the fetal liver [73, 74]. In contrast to most classical DC, LC develop independently of Flt3 and Flt3L and require keratinocyte-derived IL-34 signaling on Csf-1R for their development. CSF1R signaling is also required for macrophage development suggesting that LC may be more closely related to macrophages than other DC subsets [75–78]. In addition, LC require Runx3, PU.1, ID2 and BMP7 for their differentiation and autocrine TGF-β for their epidermal maintenance and homeostasis [79–84]. In the adult, LC form a self renewing population that can be replenished from blood derived monocytes after strong inflammatory stimuli such as UV light [85–90]. LC precursors emigrate into the epidermis via the hair follicle [91]. In the absence of strong inflammatory stimuli, LC will remain of host origin after bone marrow transplantation

The use of flow cytometric and gene profiling methods have allowed for refined characterization of DC in the human skin (Table 9.2). This is an active area of research but the current data suggests that human DC subsets are homologous to mice DC subsets but express a different set of identifying markers. DC in humans are defined as lacking lineage markers CD3, CD19, CD14, CD20, CD56 and glycophorin A [92]. Conventional human DC also express MHC-II and CD11c, but do not express markers such as CD303 (BDCA-2) and CD304 (BDCA-4) that are exhibited on plasmacytoid DC. Circulating pre-DC differentiate into conventional DC in peripheral tissue in mice while human skin CD1c + (BDCA-1) and CD141+ (BDCA-3) are found and arise in circulating blood. CD1c + DC represent the major APC population in the human dermis as well as circulation and are similar to the mouse CD11b + conventional DC [93, 94]. CD141+ DC are the only DC to express XCR1 and make up a small population, representing the same lineages as the mouse IRF8 and cD103+ DC [94]. The human dermis also contains CD14+ monocyte derived DC and macrophages. Langerhans cells are the only APC population in the human epidermis, expressing CD1a, and E-Cadherin. In contrast to the mice DC subsets in which LC and IRF8 DC both express Langerin, human LC are the only skin DC to express Langerin [95].

An important function of DC is the acquisition of foreign antigen that is then processed and cross-presented in the context of MHC-I in order to activate CD8+ cytotoxic killers (CTL). CD8+ CTLs provide immunity to viral and intracellular bacteria as well as many varieties of neoplasia. Identifying which DC subsets cross-present antigen is of particular importance for designing effective vaccines. Numerous in vivo experiments in mice have identified the IRF8 cDC and CD103+ dDC and the primary DC subset responsible for cross-presenation [96–99]. Mice with a selective deficiency of these subsets (i.e. Batf3^−/− mice) are unable to mount CD8+ T cell response against subcutaneous infection with West Nile virus, epicutaneous infection with C. albicans, and are unable to reject fibrosarcomas [57, 100, 101]. In addition, CD103+ dDC cross present keratinocyte derived antigens to CD8+ T cells that could be important in promoting cross-tolerance [68]. IRF8 cDC and CD103+ dDCs express MHC-I related genes [102, 103] and are a source of IL-12 and IL-15, cytokines that drive differentiation of cytotoxic CD8+ T cells [104]. Moreover, these DC uniquely express the chemokine receptor XCR1. XCL1, the ligand for XCR1, is rapidly produced by CD8+ T cells upon antigen presentation and promotes CTL differentiation [105].

Other DC subsets may also have the capacity to cross-present antigen. CD11b + dDC can present antigen to CD8+ T cells. Whether this occurs in only specific circumstances and whether this is presentation of antigen expressed by the DC itself or represents true cross-presentation is unclear [106–109]. LC grown in vitro or isolated from human or mouse skin explants efficiently cross-present antigen to CD8 T cells in vitro. In contrast, LC were unable to generate CD8+ T cell expansion in response to skin infection with HSV-1 or C. albicans in vivo [101, 110]. It is unclear whether these conflicting results represent disparate functions of human vs. mouse LC, derive from differences in experimental technique, or reflect differences in DC isolated from skin vs lymph node [111]. Interestingly, all DC subsets isolated from human skin explants can prime CD8+ T cells to some extent in vitro but CD141+ DC, the human homolog of IRF8+ and CD103+ dDC, are significantly more efficient than other subsets [94, 100, 112, 113].

A key event in allergic contact dermatitis (ACD) is the priming of hapten-specific naïve CD4+ and CD8+ T cells in the regional lymph node (see Chap. 23) [114]. Although haptens can drain to the lymph node via lymphatic flow this does not lead to a productive T cell response [115, 116]. Instead, antigen presentation in the lymph node by migratory DCs is absolutely essential for the generation of responses to peripheral antigen [115–118]. Most studies examining the contribution of individual DC subsets have been performed in mice with selective ablations of individual DC subsets using contact hypersensitivity (CHS) assays to small haptens such as DNFB. In one series of studies, mice in which LC are selectively ablated develop enhanced contact hypersensitivity responses to various contact allergens suggesting that LC suppress the development of CHS. The suppressive function of LC occurs during the initial priming step and the absence of LC during the effector phase does not affect CHS [119–121]. LC suppression of CHS responses depends on cognate interaction with CD4+ T cells and LC derived IL-10 [121]. LC have been also demonstrated to suppress CHS by tolerizing CD8+ T cells and activating regulatory T cells [122]. CD103+ dDC appear to participate in the development of CHS since some mice with a conditional depletion of CD103+ dDC have reduced CHS [123, 124]. This suggests that CD103+ dermal DC, rather than Langerhans cells, promote the development of contact hypersensitivity. Other data, however, particularly using low doses of hapten find that LC are required for the development of CHS [125–127]. Moreover, mice constitutively lacking C103+ dDC and IRF8 cDC develop CHS responses normally [114]. In addition, transplant of hapten primed CD11b + dermal DC can transfer CHS in vivo [128]. Thus, the relative importance of individual DC subsets for the induction of CHS remains unresolved. Untangling the functions of skin-resident DC during CHS is hindered by the intrinsic experimental variability of the assay and the difficulty in analyzing hapten specific T cells responses.

In the setting of skin infection, distinct skin resident DC subsets have diverging functions in mounting T cell differentiation. In the setting of C. albicans skin infection, CD103+ dDC are required for the development of CD8+ T cell and Th1 responses through a mechanism that likely involves CD103+ dDC-derived IL-12 [101]. In contrast, Langerhans cells are specialized to drive Th17 differentiation in the setting of an epicutaneous C. albicans infection and produce significant amounts of Th17 differentiating cytokines such as TGF-β, IL-1β and IL-6. The function of skin DC may vary with the pathogen used since mice lacking CD103+ dDC and IRF8+ DC are able to mount protective responses to West Nile Virus and cutaneous Leishmania major infection [57]. In the setting of N. brasiliensis helminth infection, CD11b + and IRF4+ DC are important in inducing Th2 immunity [35, 36]. Similarly, dermal immunization with papain and epicutaneous immunization with FITC promote Th2 responses that required CD11b + dDC [36, 128]. Th2 induction by CD11b + dDC likely involves DC-derived thymic stromal lymphopoetin (TSLP) [129, 130]. LC also express the receptor for TSLP and been shown to initiate epicutaneous sensitization with protein antigens and induce Th2-type immune responses via TSLP signaling [129, 131].

Immunological tolerance describes a state of T cell unresponsiveness. A failure of tolerance to self antigens results in autoimmune disease. As discussed above, “semi-matured” DC present self antigen and participate in tolerance to self antigen. Dendritic cells that are “semi-matured” by repeated injections of TNFα suppress autoimmunity [132]. Few but not all pan-DC depletion models demonstrate the generation of spontaneous autoimmunity [133–135]. It is important to note that while the role of individual DC subsets have been shown to be important in peripheral tolerance, no single constitutive DC subset deficiency models have been demonstrated to lead to autoimmunity [55, 57, 58, 91]. Individual DC subsets do suppress inflammatory responses. LC suppress irritant responses in hapten challenge settings and in mouse model of acrodermatitis entropathica [119, 136]. In addition, LC have been shown to suppress immune responses to L. major via the activation of regulatory T cells and promote tolerance to minor-mismatched skin grafts [137, 138]. Receptor activator of NF-kB ligand (RANKL) expression on keratinocytes mediates LC directed Treg activation and regulates UV induced immunosuppression [139]. In addition, CD11b + dDCs are thought to have a superior ability in inducing peripheral Treg differentiation due to their unique expression of aldehyde dehydrogenase (ALDH), an enzyme that metabolizes exogenous vitamin A into retinoic acid that helps regulatory T cell differentiation [140–143]. Steady state targeting of antigens to LC and CD103+ dDC can induce Treg cells [38, 39, 144] demonstrating that most DC are capable of inducing Tregs in vivo. Thus, it appears likely, that the ability to suppress effector responses is not limited to an individual “suppressive” DC subset.

DC have been shown to be important for B cell mediated antibody responses. Langerhans cells project their dendrites through the epidermis and capture antigens, leading to production of significant IgG1 production [145]. Steady state antigen targeting to DC to various antigen uptake receptors have shown the induction of robust humoral immunity. Specifically, targeting antigen to IRF8+ and CD103+ DC induce the differentiation of antigen specific T follicular helper cells and germinal center B cells [146, 147]. DC induction of antibody response is likely an important contribution to skin autoimmune disease such as pemphigus vulgaris as well as for vaccine therapeutics.

In addition to subsets of dendritic cells, other subsets of MHC-II antigen presenting cells are present or can be recruited into the skin. Dermal macrophages do not migrate to the draining lymph nodes in mice and form a self renewing population similar to Langerhans cells [148]. Dermal macrophages have a lower capacity to present antigen compared to dermal DC and function in scavenging and killing microorganisms [148]. They also likely participate in presentation of antigen to T cells that are recruited into the skin [93]. In stark contrast to dermal DC, macrophages as well as monocyte derived DC in the human and mouse dermis express high levels of IL-10 transcript, suggesting a possible anti-inflammatory role [70]. Some dermal macrophages express CD4 and surround post-capillary venules and produce chemokines that promote extravasation of neutrophils into the infected dermis [149].

In healthy mouse skin, dDC and monocyte-derived DC have a fast turnover whereas dermal macrophages have a slower turnover and a longer life. Most tissues macrophages are thought to be derived from blood LY6Chi monocytes while some macrophages are established prenatally, and derive from self replicating yolk sac progenitors [73, 150, 151]. In addition, there is a small population of blood derived CD16-expressing monocytes that display an advanced stage of differentiation with effector functions related to antigen processing and presentation [152, 153]. They arise in inflammatory conditions and represent the main producers of inflammatory cytokines such as TNFα and have a high capacity to stimulate antigen-independent T-cell responses. Monocyte derived DC can be further subdivided into 6-sulfo LacNAc (slan) -positive and -negative monocytes that both produce pro-inflammatory cytokines. The expression of slan was initially identified on an inflammatory human DC subset found in the blood and skin [154]. In steady-state, murine classical monocytes continuously extravasate and transport tissue antigens to the LN without differentiating into DC or macrophages [155]. Despite high levels of antigen capture and expression of MHC-II, monocyte derived DC are relatively poor antigen presenters [156, 157]. Alternative mechanisms such as cytokine production or antigen transfer with productive DC and APC may be their role in adaptive immunity.

Inflammatory DCs refer to populations of DCs that are transiently formed in response to various inflammatory stimuli and disappear when the stimuli is resolved. Inflammatory DC development and function remain poorly understood, with the lack of surface markers to identify them. The phenotype of inflammatory DCs is influenced by kinetics and the nature of the stimuli. Inflammatory DCs that accumulate in the LN in response to lipopolysaccharide (LPS) administration express DC-specific transcription factor zbtb46 and fail to accumulate in Flt3L−/− mice, validating them as DC [53, 56]. Another subset of inflammatory DCs identified in mice infected with L. monocytogenes was termed TNF-α/iNOS-producing DCs (TipDC) because of their ability to produce high levels of TNF-α and iNOS. TipDC are also found abundantly in psoriatic skin and are believed to be a key effector population [158]. However, in contrast to LPS-induced DC, TipDCs lack zbtb46, suggesting that they are distinct from true DC [53, 56]. The full understanding of inflammatory DC function is lacking, as are subset specific depletion models.

Plasmacytoid DCs (pDCs) represent a small subset of DCs that share a similar origin to DC but a distinct life cycle. They circulate in blood and lymphoid tissues and can be found in the skin and other peripheral tissues under inflammatory conditions. They express lower levels of MHC-II and costimulatory molecules in the steady state and display a narrow range of PRRs that include Toll-like receptors 7 and 9. Upon recognition of pathogenic nucleic acids, they produce massive amounts of type I IFNs that are important for resistance to viral infections but also participate in the development of autoimmune conditions such as psoriasis and systemic lupus erythematosus [159–163].

There are three known genetic DC deficiencies in humans. DCML deficiency syndrome is caused by a mutations of GATA-binding factor 2 (GATA2) and demonstrates a complete absence of blood DCs, pDCs, tissue cDCs, circulating monocytes, B cells and NK lymphoid cells [164]. IRF8 null mutations are found in humans and lead to the absence of monocytes, pDC, cDCs, and dermal DCs and defective IL-12 production and intact epidermal LC, resembling the mice knockout of IRF8 [165]. Mutation of adenylate kinase 2, a phosphotransferase required for nucleotide homeostasis causes a form of severe combined immunodeficiency known as reticular dysgenesis. It is associated with impaired formation of all nucleated blood cells, including neutrophils, lymphocytes, monocytes, and cDCs as well as LCs [166]. Patients with these defects all have increased risk of infections, but it is difficult to pinpoint the specific function of individual DC since the defects affect a broad range of cell types.

Psoriasis is also another condition in which DC are implicated for pathogenesis. Repeated application of the TLR7 ligand imiquimod to murine skin induces an inflammatory response that recapitulates some similarities to that of human psoriasis, including hyperproliferation and differentiation of keratinocytes and thickening of the epidermis [167, 168]. Infiltrates of neutrophils in the epidermis and DC, macrophages and T cells in the dermis is found after imiquimod induced psoriasis an other genetic mouse models of psoriasis. In the imiquimod model of psoriasis, IL-23 production by DC elicits dermal gamma delta T cells to secrete IL-17 and IL-22 which recruit neutrophils and drive the proliferation of keratinocytes, respectively [169–173]. TLR signaling in DC seems to be necessary and sufficient for the formation of IL-23 mediated psoriasis [174]. While pDC are found in the skin of psoriatic patients and mouse models of psoriasis, conditional ablation of pDC does not lead to attenuation of disease [174]. More recently, it has been suggested that sensory nociception leads to the activation of DC to produce IL-23 after the application of imiquimod and cutaneous denervation leads to attenuation of disease in transgenic mice models of psoriasis [175, 176]. However, which DC or monocyte or macrophage subset is responsible for IL-23 secretion remains controversial [172, 174].

Efficient vaccination requires acquisition of antigen by DC. The density and easy access to DC in the skin makes them an attractive target for vaccination strategies.

Interestingly, the smallpox vaccine, which was the first vaccine developed, is administered intradermally and remains one of the most effective vaccines developed [55, 74, 119, 177]. The density of skin DC is thought to explain the increased efficiency of skin immunization compared with the intramuscular route [93]. Most vaccines, however, bypass the APC rich epidermis and the dermis and target the connective and adipose tissue rich hypodermis. More recently, newly developed delivery systems are becoming more common in order to target the more superficial layers of the skin. For example, microneedle vaccination has been used to introduce encapsulated influenza virus vaccine into the dermis leading to robust cellular and humoral immune responses [178]. Laser-generated micropores have been used for intradermal immunizations to induce potent immune responses [179]. Combinations of novel delivery systems with epitopes coupled to antibodies that specifically target cell surface endocytic receptors (e.g. DEC-205 or Langerin) expressed by APC populations are becoming more popular for researchers to study DC functions as well as for vaccination trials against cancers and pathogens [180, 181]. For example, targeting vaccine antigens directed against receptor Clec9A to CD103+ or CD8+ cDC enables robust cross-presentation and CD8+ T cell response [135, 146, 147]. In addition, antibody responses are generated via the production of T follicular helper cells [147]. Given the potential for DC specific differences in function, it is plausible that specific DC may be targeted for a tailored cellular immune response, such as tolerance in the setting of autoimmunity [38, 39, 144].