¹Centre for Asthma and Respiratory Disease (CARD), School of Biomedical Sciences, Faculty of Health, University of Newcastle and Hunter Medical Research Institute, Callaghan, New South Wales 2300, Australia. E-mails: Paul.Foster@newcastle.edu.au or Joerg.Mattes@newcastle.edu.au

Despite intensive research, the pivotal factors that link abnormalities in permeability of the dermal barrier with inflammation and the generation of skin lesions in allergic dermatitis (AD) remain unknown. In a landmark paper published by Jin et al.¹ in the Journal of Clinical Investigation, IL-21 is identified as a critical regulator of the processes that lead to sensitization and allergic inflammation of the skin.

Allergic dermatitis is an inflammatory disease of the skin of unknown aetiology.^{2, 3} AD clinically presents early in life (under 5 years of age) and is characterized by eczematous lesions, drying and thickening of the skin, and pruritus (severe itching).² Susceptibility may be associated with a history of asthma and other atopic conditions characterized by elevated levels of IgE (for example, allergic reactions to foods, rhinitis and urticaria). The incidence of AD is rising in the Western world and the environmental factors driving increased susceptibility are thought to be associated with the global 'atopic march'. In general, two paradigms have emerged to describe the aetiology of AD: (1) that a defect in immunological responsiveness to antigens encountered on the skin surface (stratum corneum) underpins the development of lesions, and (2) that inherited and acquired abnormalities in permeability of the barrier are the key drivers of disease initiation and progression.^{2, 3, 4}

Like in the case of asthma, the CD4 T helper-1 and -2 paradigm has been widely used to explain the aetiology of AD. Early in life, skin lesions are associated with an abnormal Th2 response to antigens and numerous studies have implicated Th2 cytokines in pathological events. However, like in the case of asthma, the Th2 paradigm does not fully account for the heterogeneity or chronicity of disease. As AD progresses, the T-cell response may switch or is skewed to a 'more' Th1 phenotype, probably because of chronic infections at the site of skin lesions. Interestingly, asthma may also evolve towards a more neutrophilic and Th1 dominant phenotype that is linked to chronic respiratory infections and is largely unresponsive to steroid therapy. Indeed, although very distinct disorders, a number of parallels can be drawn between aetiology and disease progression in AD and asthma.

The second paradigm suggests that inherent abnormalities in pathways that regulate skin homeostasis predispose to broad barrier failure. This alters permeability and antimicrobial functions that lead to inflammation and the development of lesions.⁴ Inherited barrier abnormalities in AD have been associated with a defective stratum corneum (outer layers of the skin). In particular, loss-of-function mutations in filaggrin, which regulates the structural and functional integrity of the epidermal cytoskeleton, can enhance allergic sensitization.⁵ Although there is open debate on the aetiology of disease and critical susceptibility factors for AD, there is clear evidence that the interactions between immunological pathways and processes that maintain barrier integrity are central to the perpetuation of disease (Figure 1).

Figure 1.

Mononuclear cell enters skin in response to tissue injury and secretes IL-21. This promotes the activation of MMP2 and MMP9 in skin cells (most likely Langerhans cells, dermal DCs, keratinocytes and fibroblasts) and migration of DCs from the skin to local LNs. DC migration in the skin is known to be regulated by activation of these molecules. Migration of DCs to regional draining LNs is under the regulation of CCR7 on the DC and CCL19 and CCL21 in the node compartment. DCs engage naive T cells, resulting in expansion of Th2 cells and the induction of hallmark features of allergic inflammation and AD.

Full figure and legend (256K)

Recently, Jin et al.¹ identified for the first time that protein production of IL-21 and its receptor (IL-21R) was up-regulated in skin lesions of patients suffering from active AD, but not normal controls. IL-21 was expressed in mononuclear leukocytes that infiltrate the dermis, and the IL-21R was expressed by epidermal keratinocytes and along the dermal–epidermal junction. IL-21 is a pleiotropic cytokine derived from activated CD4 T cells.⁶ This raised the question—Is IL-21 a critical regulator of inflammation and barrier function in AD? To advance our understanding of the potential role of IL-21 in the pathogenesis of AD, a mouse model of tape stripping of saved skin was used to mimic tissue trauma induced by scratching dry, itchy skin that underlies human AD. Tissue injury induced by tape stripping resulted in the up-regulation of IL-21 and IL-21R mRNA. Epicutaneous application of the archetypical Th2 antigen ovalbumin to tape-stripped skin resulted in allergic inflammation at the site of exposure that mimicked aspects of the inflammatory response and tissue lesions in AD. In IL-21R deficient mice epicutaneous allergic sensitization was profoundly impaired. In the absence of IL-21R the expression of CCL17/TARC (recruits skin-homing T cells), Th2 cytokines (IL-4 and IL-13), eotaxin (chemotactic signal for eosinophils) and IFN were not increased, and isolated T cells did not proliferate in response to Ova exposure. IL-21 signalling was also shown to be essential for the development of cutaneous inflammation in a Th1 cell model of contact hypersensitivity, highlighting the pivotal role of this pathway in promoting inflammation in the skin. The impaired systemic response to epicutaneous allergic sensitization was not related to an intrinsic defect in dendritic cell (DC) function, as IL-21-deficient DCs effectively induced proliferation of Ova-specific DO11.10 transgenic Th2 and Th1 cells in culture and at steady state and the same frequency of CCR7+CD11c+ DCs were found in peripheral tissues. However, recruitment of adoptively transferred naive Ova-specific T cells to the site of Ova challenge after tape stripping was significantly impaired, and this correlated with defective DC migration to the lymphatic ducts of tape-stripped skin (in the absence of Ova treatment). The authors did not demonstrate that DC migration was decreased after Ova exposure to tape-stripped skin, which would have strengthened their proposed mechanism and this could have been achieved by using FITC-labelled Ova. However, DC migration to draining lymph nodes (DLNs) was also impaired in a model of cutaneous hapten-induced hypersensitivity. Thus, signals for the migration of T cells and DCs to the DLN were significantly attenuated in the absence of IL-21 signalling.

Next, the investigators undertook a detailed and elegant analysis of DC phenotype and migration in the absence of IL-21R, which included bone marrow chimeric experiments in irradiated mice. IL-21R-deficient DCs were found to have an intrinsic defect in their ability to migrate from the skin to DLNs. Studies also suggested a contribution from non-haematopoietic cells to the mechanism limiting DC trafficking. This conclusion is tempered by the failure to examine DC trafficking in IL-21R-deficient recipient chimeras because of breeding difficulties. However, impaired mobilization was also associated with decreased expression of CCR7 by skin DCs and lack of responsiveness of those cells to the CCR7 ligands CCL19 and CCL21. These chemokines are expressed by high endothelial venules and stromal cells in LNs and signal through CCR7 to promote migration of DCs.⁷ In this study, reduced DC migration was linked indirectly to IL-21 deficiency by in vitro experiments, in which activation of the enzymic activity of metalloproteinases MMP2 and MMP6 in skin cells was shown to be IL-21 dependant. MMP2 and MMP6 are expressed by skin Langerhans cells, dermal DCs, keratinocytes and fibroblasts, and migration of DCs from the skin to local LNs is known to be regulated by activation of these molecules.⁸ Thus, in the absence of IL-21 signalling, the critical signals that regulate the recruitment of DCs from the skin to the DLNs are absent and the downstream events that lead to sensitization, T-cell recruitment, expansion and subsequent inflammation cannot be initiated. Jin et al.¹ conclude their investigation by directly demonstrating the central importance of the IL-21 pathway in their model system by administering a soluble IL–21R–IgG2aFc fusion protein to wild-type mice, which significantly inhibited the development of allergic inflammation of the skin.

Although mouse models do not reflect all of the complexities of human disease,⁹ the data presented by Jin et al.¹ are important first steps in identifying a new pathway that may have potential therapeutic implications for the treatment of AD. Future studies will require confirmation of the up-regulation of IL-21/IL-21R signalling system in extended populations of AD patients and the subsequent development of topical anti-IL-21 treatments.