Ontogeny and function of epidermal murine Langerhans cells

Abstract

Langerhans cells (LC) are epidermal resident antigen-presenting cells that share a common ontogeny with macrophages but function as dendritic cells (DC). Their development, recruitment and retention in the epidermis is orchestrated by interactions with keratinocytes through multiple mechanisms. LC and dermal DC subsets often show functional redundancy but LC are required for specific types of adaptive immune responses when antigen is concentrated in the epidermis. This review will focus on those developmental and functional properties that are unique to LC.

Langerhans cells (LC) are the only MHC class II expressing antigen-presenting cells (APC) under steady-state conditions in the epidermis, the outermost cellular layer of the skin. They were initially observed by Paul Langerhans in 1868 and thought to function as part of the peripheral nervous system but landmark studies 40 years ago placed them firmly within the hematopoietic system ^1–4. Early work by Ralph Steinman and others approximately 30 year ago found that immature LC and dendritic cells (DC) efficiently acquired and processed antigen^5,6. Exposure of LC and DC to activating stimuli allowed for highly efficient activation of naive T cells in mixed lymphocyte reactions. Inflammatory stimuli also greatly enhanced migration of LC out of the epidermis and into regional lymph nodes (LN). Thus, LC were considered the prototypical migratory DC envisioned by the DC-paradigm leading also to coinage of the term “LC paradigm”⁷. A corollary to the DC-paradigm is that presentation of self antigen by DC or LC in the absence of inflammatory stimuli deletes or silences autoreactive T cell clones thereby providing a basis for peripheral self-tolerance⁸. The location of LC at a barrier surface provides them with access to skin pathogens, commensal organisms, allergens, contact sensitizers and epidermal self-antigens. Thus, LC were assumed to mediate initiation of adaptive immunity against foreign antigens and tolerance to self-antigens found in the skin.

More recently, there has been considerable progress investigating skin DC. Notably, several subsets of dermal DC were identified and have been shown to be required for many of the functions originally ascribed to LC. The phenotypes (Table 1) and functions of skin APC subsets have been reviewed recently⁹. In addition, LC were found to be closely related to macrophages based on a shared ontogeny^10,11. Thus, LC are turning out to be a rather unique cell type. This review will explore the unique aspects of murine LC biology and the contribution these cells provide to the establishment and regulation of cutaneous immune responses.

Table 1.

Mouse antigen-presenting phenotypes

Name		LC	Dermal cDC1	Dermal cDC2
Other Names			XCR1+ dDC	CD11b+ dDC
			CD103+ dDC	IRF4 dDC
			IRF8+ dDC
Location		Epidermis	Dermis	Dermis
Transcription Factors	ID2	+	+	−
	PU.1	+	−
	Runx3	+	−	−
	Batf3	−	+	−
	IRF4	−	−	+
	IRF8	+	+	−
Surface Markers	CD8a	−	−	−
	CD103	−	+	−
	XCR1	−	+	−
	Clec9A	−	+	−
	CD11b	+	−	+
	CD207 (Langerin)	+	+	−
	CD301b	−/+	−	+
	CD172	−	−	+
	CD64	−	−	−
	MERTK	−	−	−
	CCR2	−	−	−
	F4/80	+	−	−
Soluble
Factors/Receptors	Flt3	−	+	+
	CSF-1R	+	−	+
	CSF-2R	−	+	+
	IL-34	+	−	−
	TGF-β	+	−	−

Development and maintenance of the epidermal LC network

During ontogeny, primitive myeloid LC progenitors initially from the yolk-sac and later from the fetal liver seed the skin (Fig. 1) ¹¹. On day 2 after birth, these cells undergo a 10–20-fold expansion during which they assume a dendritic morphology and begin to express the surface markers MHC class II and Langerin in a step-wise manner with ultimate establishment of the adult LC network by 3 weeks of age^12,13. LC differentiation requires several transcription factors related to Transforming growth factor-β (TGF-β) signaling including Runx3 and ID2 as well as engagement of the receptor CSF1R (MCSFR) by KC-derived IL-34^10,14–17. Once formed, adult LC form a self-renewing, radio-resistant population within the epidermis¹⁸.

Figure 1. Antigen presenting cells in the skin.

LC are the only MHC class II antigen presenting cells in uninflamed epidermis. They arise from embryonic monocytic precursors and in the adult can be easily identified based on their epidermal location and surface markers (Table 1). The dermis is populated by two major subsets of DC that arise from dedicated pre-DC precursors in the bone-marrow. Dermal cDC1 are closely related to cDC1 in secondary lymphoid tissues. Dermal cDC1 are often referred to based on their surface marker expression and transcription factor dependence as XCR1⁺ dDC, CD103⁺ dDC or IRF8⁺ dDC. Dermal cDC2 are closely related to cDC2 in secondary lymphoid tissue and are often termed CD11b⁺ dDC or IRF4 dDC. A less well characterized DC subset termed double-negative DC (DNDC) based on the absent expression of CD103 and CD11b as well as macrophages and monocyte-derived DC (moDC) also reside in the dermis. During inflammation, monocyte-derived LC are recruited into the epidermis at the follicular isthmus and infundibulum but are excluded from the bulge region.

Notably, LC ontogeny is clearly distinct from classic DC development. DC arise from bone-marrow precursors, require the cytokine Flt3L, have a short half-life and do not self-renew. LC development more closely resembles that of other tissue macrophages, particularly microglia. Like microglia, LC arise from primitive myeloid progenitors and require tissue-derived IL-34^16,19. This is further supported by the fact that macrophage populations can self-renew in peripheral tissues. Thus, it has been proposed that LC should be considered a subset of tissue macrophages akin to microglia in the brain, alveolar macrophages in the lung and Kupffer cells in the liver ^20,21. While this makes sense based on ontogeny, it neglects the fact that LC, unlike tissue macrophages have the capacity to migrate into regional lymph nodes. The LC gene expression profile matches that of other migratory DC populations and they can efficiently prime naive T cells^22,23. This is a critical function not shared with tissue macrophages. The dichotomy between ontogeny and function likely results from tissue programming of precursors by the epidermis as has been demonstrated with macrophages by adoptive transfer between tissues^24,25,26. Thus, while LC in the skin share many features with tissue macrophages and have been speculated to have macrophage-like functions while skin-resident, they also clearly function as DC. Thus, LC are a unique, hybrid cell type that are probably best considered to be a specialized form of DC.

The cytokine TGF-β1 is particularly important for the development of the LC network. In vitro cultures of hematopoietic stem cell (HSC) precursors yield LC in the presence of TGF-β1. Mice lacking the transcription factors ID2, Runx3, and Pu.1 as well as Axl that are all involved with TGFβ1-responses, lack or have reduced LC numbers^14,15,27,28. BMP7, a member of the TGF-β superfamily, is required for optimal LC development²⁹. Finally, Tgfb2^−/− mice lack LC³⁰. Interestingly, TGF-β1 signaling is also required to maintain the LC network after it has developed. When mice in which Tgfbr1, Tgfbr2, or genes such as Lamtor2 in the TGF-β pathway are conditionally ablated from LC they lose the capacity to remain in the epidermis and spontaneously migrate into regional lymph nodes^31–33. Similarly, ablation of Tgfb1 from differentiated LC results in spontaneous homeostatic LC migration^32,34. Thus, despite many sources of TGF-β1 in the epidermis (e.g. keratinocytes, γδ T cells and LC), LC depend on autocrine and/or paracrine TGF-β1 for epidermal residence. TGF-β1 signaling is also sufficient to prevent homeostatic LC migration as mice in which LC express a mutated, constitutively active TGF-βRI fail to migrate to regional lymph under steady-state conditions (Fig. 2) ³⁵. TGF-β1 is secreted as an inactive, latent form associated with LAP and in the epidermis requires activation by the integrins avβ6 or avβ8 that are expressed by non-overlapping subsets of keratinocytes (avβ6 in the interfollicular regions and avβ8 near the hair follicles) ^35,36. Thus, transactivation of LC-derived TGF-β1 by integrins expressed by keratinocytes is required to maintain the epidermal residence of LC under non-inflammatory conditions. TGF-β1 signaling is required for expression of Axl that has anti-inflammatory effects and may act on LC as well as KC to inhibit migration ²⁸. From this, the inference is reasonably made that keratinocyte expression of avβ6 or avβ8 likely in conjunction with additional signals may be a required event for homeostatic LC igration.

Figure 2. Keratinocytes and TGF-β control LC migration.

Under steady-state conditions, integrins avβ6 and avβ8 transactivate LC-derived TGF-β-LAP. a) Tonic TGF-β signaling in LC as well as LC-KC structural interactions are required for their epidermal retention. b) Migratory signals such as UV light reduce KC expression of avβ6 and avβ8 reducing the availability of active TGF-β. The absence of active TGF-β likely in conjunction with still unknown factors results in LC migration. Inflammatory cytokines including IL-1β and TNF from KC and dermal infiltrates also promote LC migration but likely act indirectly on KC.

LC self-renew and remain of host origin in murine bone marrow transplantation models^18,37,38. LCs can repair DNA damage through the action of the cyclin-dependent kinase inhibitor, CDKN1A, which permits cell cycle arrest, providing protection against ionizing radiation³⁹. However, strong inflammatory stimuli such as UV light can deplete LC¹⁰. In this context, CCR2-dependent GR1^hi monocytes are recruited into the epidermis to replace LC that have migrated (Fig. 1) ¹⁰. Recruitment of monocyte precursors into the epidermis occurs at the hair follicle and requires the chemokine receptors CCR2 and CCR6⁴⁰. The ligand for CCR2, CCL2, is expressed at the follicular isthmus and the ligand for CCR6, CCL20 is expressed at the follicular infundibulum. LC are excluded from the immune-privileged bulge region containing keratinocyte stem cells by CCL8 and CCR8. Interestingly, LC derived from GR1^hi monocytes arise independently of TGF-β and are short lived in the epidermis. They are replaced by a second wave of steady-state-derived long-term LC⁴¹. The biological significance of these transient monocyte-derived LC remains unclear through there is evidence to suggest they participate in inflammatory cytokine circuits in psoriatic skin^42,43.

Activation-induced LC migration

Exposure to UV light and haptens are the best studied inflammatory stimuli that lead to LC activation and migration. Activated LC reduce expression of E-cadherin that forms a structural tether with E-cadherin-expressing keratinocytes thereby allowing egress from the epidermis⁴⁴. Migration out of the epidermis is facilitated by CXCR4 and the adhesion molecular EpCAM^45–47. Activated LC begin to express CCR7 and once in the dermis follow a CCL19 and CCL21 chemokine gradient through the dermal lymphatics and into the paracortex of the skin-draining LN⁴⁸.

LC migration is triggered by the coordinated action of IL-1β, IL-18 and TNF since administration of blocking antibodies to these inflammatory cytokines and Casp1^−/−, Il1b^{− /−} and Tnfr2^−/− mice show decreased hapten-induced migration^49–53. It remains unclear, however, whether these cytokines act on LC directly or indirectly via KC (Fig. 2). LC migration in response to hapten and Candida albicans infection is unaffected in mice with a selective LC ablation of MyD88 that renders them insensitive to IL-1β and IL-18 as well as Toll-like receptor 2 (TLR2) ligands found in the C. albicans cell wall⁵⁴. In contrast, mice with a KC-specific ablation of MyD88 have reduced levels of LC migration in response to protein immunization⁵⁵. In addition, UV exposure reduces the capacity of keratinocytes to provide active TGF-β to LC and constitutive TGF-β signaling in LC inhibits both UV-induced expression of CCR7 and LC migration³⁵. Thus, keratinocytes play an important role in facilitating LC migration⁵⁶ and, in at least some contexts, regulated TGF-β transactivation by keratinocytes may act as a trigger for LC migration.

Langerhans cells and priming adaptive T cell responses

Despite intense focus by many laboratories, the precise function of LC remains controversial largely due to the use of varying techniques. Early work focused on LC in culture. The identification, however, that LC selectively express Langerin, a c-type lectin receptor, was a major breakthrough⁵⁷. Using Langerin as a LC marker greatly facilitated analysis of LC ex vivo and allowed for the engineering of 3 independently derived mouse lines that efficiently ablated LC (Table 2). The primate diphtheria toxin receptor (DTR) was introduced into the endogenous Langerin locus to create two lines of murine Langerin-DTR (muLangerin-DTR) mice^58,59. Human genomic BAC DNA containing the Langerin locus that had been modified to express active Diphtheria toxin or DTR was used to create human Langerin-DTA (huLangerin-DTA) and huLangerin-DTR transgenic mice^60,61. It was later discovered that dermal cDC1 (also known as CD103⁺ dDC) and cDC1 (also known as LN and spleen resident CD8⁺ DC) in on the C57BL/6 genetic backgrounds also express Langerin. This clearly complicated interpretation of the early work utilizing muLangerin-DTR mice and analyzing LC based on Langerin expression. HuLangerin-DTA and huLangerin-DTR mice selectively target LC since transgene expression recapitulates the human pattern of Langerin expression in LC but not cDC1.

Table 2.

LC depleter mouse lines

Model	Specificity	Caveats
muLangerin-DTR DT Day -1	LC Dermal cDC1 and cDC1 in C57BL/6 mice	LC and dermal cDC1 ablated
muLangerin-DTR DT Day -7/13	LC Dermal cDC1 have largely repopulated	Dermal cDC1 numbers reduced
WT→muLangerin-DTR Bone marrow chimera	LC	Reconstitution may not be complete. DETC may not reconstitute
huLangerin-DTA	LC	Constitutive absence of LC and possible compensatory effects Potential additional genes on BAC transgenic
huLangerin-DTR	LC	Potential additional genes on BAC transgenic

These early technical problems have been largely overcome. The phenotype of LC is now well characterized and they can be easily and reliably isolated from lymph node using several markers including CD11c⁺, MHC class II⁺, Langerin⁺ (CD207⁺), CD11b⁺, CD103⁻ (Table 1). To selectively ablate LC in muLangerin-DTR mice, DT can be administered 1–2 weeks prior to experimentation (Table 2) ⁶². Since LC repopulate the epidermis after ablation more slowly than dermal cDC1 under steady-state conditions, there is a window after DT administration where LC are absent but dermal cDC1 have largely recovered. Using muLangerin-DTR mice as recipients in bone marrow chimera experiments is also commonly employed. Since LC are radio-resistant and self-renew, they remain of host origin in chimeras¹⁸. Thus, administering DT to Wild-type → muLangerin-DTR chimeras targets LC and not dermal cDC1 that have repopulated from donor origin. Although these approaches all have their caveats, there is now a large body of literature that accurately assays LC requirement in vivo across many contexts (Fig. 3) that will be explored below.

Figure 3. Murine skin DC subsets drive distinct T cell phenotypes.

LC, dermal cDC1, and dermal cDC2 promote distinct but overlapping T cell phenotypes. In most contexts, the same antigen is present in both the epidermis and dermis resulting in a redundancy between LC and dermal DC subsets. A requirement for LC is observed primarily when antigens or adjuvants are confined to the epidermis. Strong evidence is shown by red arrows and suggestive evidence is shown in blue.

LC in cross presentation

Presentation of exogenous antigen by DC requires specialized processing and cross-presentation, a function critical for cytotoxic T cell responses against viruses, intracellular pathogens and tumors. Targeting antigen to the correct DC subsets for cross-presentation is an important goal for effective vaccine design. In general, mouse and human dermal cDC1 have a superior ability compared with LC to present particulate or necrotic cell-derived antigens for cross-priming to CD8⁺ T cells, particularly in the context of TLR3 co-stimulation^63–65. In the context of cutaneous infections, Batf3^−/− mice deficient in cDC1 but not mice lacking LC are unable to mount CD8⁺ T cell response against epicutaneous infection with C. albicans or herpes simplex virus, priming of commensal specific IL-17 secreting CD8⁺ T cells and rejection of syngeneic tumors^66–69. Moreover, dermal cDC1 more efficiently cross prime keratinocyte derived self-antigens to CD8⁺ T cells suggesting a role in cross-tolerance^67,70.

Despite being less efficient than cDC1, there is evidence that LC are capable of cross presenting antigen to CD8 T cells in certain contexts. In vitro, antigen-pulsed LC can cross present antigen⁷¹. LC deficient mice develop reduced CTL mediated immunity following cutaneous immunization with antigen conjugated nanoparticles⁷². LC targeted in vivo with foreign antigen using an anti-Langerin antibody conjugated to antigen in the presence of certain adjuvants resulted in CD8 proliferation but functional tolerance (e.g. cross-tolerance), not cross priming⁷³. Finally, like most DC, LC can present endogenous antigen to CD8 T cells resulting in direct-tolerance or direct-priming in the presence of an adjuvant (Stoecklinger in press). In the context of graft vs. host disease where LC are a source of alloantigens, LC within the skin are required to license pathogenic CD8 effectors⁷⁴[ Santos e Sousa NI 2017 under review—ZOLTAN what is the status??].

LC and T_H17 cell differentiation

T_H17 cells protect against extracellular fungal and bacterial pathogens and can be pathogenic in autoimmune diseases such as psoriasis. LC are required for the development of T_H17 responses in response to C. albicans epicutaneous skin infection and provide protection to subsequent cutaneous but not systemic C. albicans infections^66,75,76. In this context, LC engaging with Dectin-1 ligands expressed by C. albicans yeast were a non-redundant source of IL-6, a key cytokine required for T_H17 differentiation. The absence of LC had no effect on T_H1 differentiation in this model. Mice with a keratinocyte specific ablation of the protease ADAM17 develop spontaneous cutaneous dysbiosis with overgrowth of Staphylococcus aureus. In this model, LCs were required for the generation of IL-17-producing CD4⁺ and γδ T cells in the skin⁷⁷. The capacity to promote T_H17 is not unique to LC as cDC2 are required for T_H17 generation against bacterial and fungal pathogens in other tissues^78,79. T_H17 differentiation in response to epicutaneous C. albicans infection depends on LC and not dermal cDC2 because the ligand for Dectin-1 is available primarily on C. albicans yeast forms found in the epidermis and not hyphal forms found in the dermis. The same is likely true in ADAM17-deficient mice where dysbiotic S. aureus remain superficial on the skin. Thus, LC are required for T_H17 differentiation particularly when the antigen is concentrated in the epidermis.

LC and T_FH and T_H2 cell differentiation

The development of humoral immunity to cutaneous antigen involves the induction of CD4+ T follicular helper (T_FH) cells that promote plasma cell and germinal center (GC) development. LC can acquire epicutaneously applied antigen and were required for the production of protective IgG1 in a model of Staphylococcus scalded skin syndrome^80,81. LC-deficient mice have decreased GC and T_FH cell responses after intradermal immunization and infection with Leishmania major^82,83. Foreign antigen targeted selectively to LC expanded T_FH cells and promoted GC formation resulting in antigen-specific IgG1 thus indicating that antigen presentation by LC is sufficient to promote a humoral response⁸⁴. The ability to promote humoral responses, however, is not unique to LC. Foreign but not self-antigen targeted to cDC1 even in the absence of adjuvant induced T_FH cell expansion, GC formation and protective antibody responses^85,86. In addition, T_FH cells and GC responses after an intradermal nanoparticle immunization required DC migration but were only partially abrogated by the absence of LC while GC formation in response to hapten that penetrates deep into skin was independent of LC⁸³. Thus, it appears that LC like other subsets of DC have the capacity to promote humoral responses.

LC may have a special function for the development of IgE. Epicutaneous immunization with OVA results in antigen-specific IgE that is reduced in mice lacking LC and mice with LC-specific ablation of the receptor for the cytokine TSLP ⁸⁷. In addition, LC-deficient mice have reduced steady-state levels of serum IgE. In this model, LC are not required for T cell proliferation but are required for optimal IL-4 expression in skin draining LN suggesting a possible role in T_H2 differentiation. T_H2 differentiation to epicutaneous house dust mice, however, is clearly reduced in Irf4^fl/flItgax (CD11c)-Cre mice indicating a requirement for dermal cDC2 but not LC⁸⁸. Antigen-specific IgE and IgG1 are reduced after dermal injection of antigen in Irf4^fl/flItgax Itgax-Cre mice but this was not observed after dermal infection with the helminth Nippostrongylus brasilliensis in MGL2-DTR mice that target cDC2 as well as macrophage populations^89,90. T_H2 differentiation in the intradermal papain model was not affected by the absence of LC and LC ablation of STAT5 that is required for TSLP signaling did not reduce TH2 differentiation in response to epicutaneous challenge with the hapten FITC^90,91. Thus, LC can promote IgE production in response to epicutaneous protein immunizations but their capacity to promote T_H2 differentiation appears to be limited.

LC and tolerance induction

The concept that antigen presentation by immature DC promotes anergic or immunosuppressive T cells to support peripheral tolerance has been well established by classic antigen targeting experiments^92–94. This is likely true for LC since resting human LC activate and induce proliferation of skin-resident regulatory T (T_reg) cells in vitro whereas activated LC preferentially induced proliferation of effector memory T cells⁹⁵. These experiments analyzed skin-resident memory cells and it is less clear whether LC have a special capacity to induce tolerance. Targeting antigen to LC in vivo promotes T_reg cell proliferation but this appears to be restricted to self and not foreign antigen^84,96. In the specific context of L. major infection, LC suppress anti-Leishmania responses possibly through T_reg cell expansion⁹⁷. In mouse models of allergic contact dermatitis, pretreatment with a precise dose of the innocuous hapten DNTB can tolerize to subsequent sensitization with DNFB, a strong sensitizer, through a mechanism that require LC for CD8 T cell tolerance and activation of T_reg cell ⁹⁸. Thus, there are specific situations where LC appear to promote tolerance that may be related to antigen restriction to the epidermis. The promotion of tolerance, however, is not a special attribute of LC since all migratory DC share a similar immune-suppressive gene expression profile and subsets of dermal DC also promote tolerance^99–102. Moreover, mice with constitutive ablation of LC or other individual DC subsets have not been reported to develop autoimmunity. Thus, individual skin DC subsets are likely sufficient but redundant for peripheral tolerance in most contexts.

LC and contact hypersensitivity

Contact hypersensitivity (CHS) is a mouse model for human allergic contact dermatitis¹⁰³. In general, small molecule sensitizing haptens that penetrate the skin are used to immunize mice and the effector response is measured after application of the same hapten at a distant site. Selective ablation of LC in muLangerin-DTR mice using either the delayed immunization or bone marrow chimera techniques (Table 2 and discussed above) did not affect CHS responses^62,104. A requirement for LC, however, was observed at low doses of hapten that may concentrate antigen in the epidermis¹⁰⁵. Urushiol, the active sensitizer in poison ivy, is presented to T cells on CD1a – an antigen-presenting molecule that is strongly expressed by human and not mouse LC¹⁰⁶. Transgenic mice expressing human CD1a on LC have greatly increased CHS responses to urushiol ¹⁰⁷. Thus, human LC are likely the key antigen presenting cell in allergic contact dermatitis to antigens presented through CD1a but in mice, their role is limited to contexts in which antigen is primarily epidermal.

In contrast to data using muLangerin-DTR models of LC ablation, CHS to several haptens at a range of doses is reliably increased in huLangerin-DTA mice that have a constitutive and selective absence of LC^60,108. Acute ablation of LC in huLangerin-DTR mice prior to sensitization also increases CHS responses, though the effect is less pronounced and consistent than with huLangerin-DTA mice⁶¹. Cells isolated from lymph nodes of huLangerin-DTA mice can adoptively transfer exaggerated CHS responses. Mice with a constitutive LC-specific ablation of MHC class II or IL-10 but not MyD88 also have increased CHS responses^54,108. Notably, delayed type hypersensitivity (DTH) responses to C. albicans is also increased in huLangerin-DTA mice⁶⁶. A similar exaggerated response is observed after intradermal injection of C. albicans in naive huLangerin-DTA mice¹⁰⁹. This occurs only in the context of constitutive ablation or long-term depletion (DHK, unpublished observation) of LC and can be adoptively transferred to wild type mice by liver resident type I innate lymphoid cells (ILC1). These data raise the possibility that LC may indirectly suppress cutaneous immune responsiveness or the long-term absence of LC may promote an unidentified compensatory inflammatory response.

Concluding thoughts and future perspective

The past several years has seen significant progress in the understanding of murine LC biology. The recognition that LC are only one of several antigen-presenting-cells in the skin and the development of tools to accurately identify and target LC have allowed for a more detailed and nuanced study of LC. It is now appreciated that LC are a unique cell type that have a close ontogenetic relationship with macrophages and a close functional relationship with DC. To date, most of the functional analysis of LC has focused on their capacity to drive antigen specific T cell responses. LC are clearly involved in T_H17 and T_FH cell differentiation. There is also support for their involvement in T_reg cell, CD8 T cell and perhaps T_H2 responses. They appear to have little involvement in T_H1 immunity. LC share many of these functions with other DC subsets and are non-redundant primarily in contexts in which antigen is confined or concentrated to the epidermis. This fits well with the concept that an individual DC subset has the potential to promote several but not all T helper phenotypes and that the ultimate T cell response is dictated by a combination of antigen location, DC subset presenting antigen and the microbial-associated molecular pattern and/or cytokine environment. Despite progress, there are several aspects of LC biology that remains poorly defined. The relationship of LC with immune effectors within the epidermis, the interaction of LC with cells of the innate immune system including KC, the effect of CD1a presentation of lipid molecules by LC, and whether LC share a functional along with an ontogenetic relationship with macrophages all remain to be fully explored. These represent major frontiers for future exploration.