I. AD models induced by epicutaneous sensitization
I.1 An animal model of AD induced by skin injury and epicutaneous sensitization with allergen
Our laboratory has developed a mouse model of AD induced by repeated epicutaneous (EC) sensitization of tape-stripped skin with ovalbumin (OVA) (Spergel et al., 1998
). This model operates in all five strains of mice tested to date including BALB/c and C57BL/6 mouse strains (Spergel et al., 1999
). The back skin of mice is shaved and tape stripped 6 times with 3M tape, mimicking skin injury inflicted by scratching in patients with AD. One hundred μg of OVA in 100 μl of normal saline or 100 μl of normal saline is placed on a 1 × 1 cm patch of sterile gauze, which is secured to the skin with a transparent bioocclusive dressing. This ensures that the antigen is not accessible to licking. Each mouse has a total of three one-week exposures to the patch at the same site that is separated from each other by 2 week intervals ().
Figure 1 EC sensitization protocol. Mice were sensitized with OVA (100 μg) or saline applied in 100 μl to a sterile patch. The patch was placed for a one-week period, then removed. Two weeks later, an identical patch was reapplied to the same skin (more ...)
EC sensitized mice develop increased scratching behavior and their skin develops lesions characterized by epidermal and dermal thickening, infiltration of CD4+
T cells, and eosinophils () and upregulated expression of the Th2 cytokines IL-4, IL-5 and IL-13 () with little or no change in the expression IFN-γ. There is enhanced expression of eotaxin and TARC, the chemokines that respectively attract CCR3+
eosinophils and skin homing CCR4+
T cells. There is also increased deposition of collagen. Systemically, serum OVA-specific IgG1, IgE and IgG2a are elevated, and splenocytes from OVA-sensitized mice produce increased level of IL-4, IL-5, IL-13 and IFN-γ in response to OVA re-stimulation (Spergel et al., 1998
). The fact that antigen specific IFN-γ producing cells are present in the spleen, with no detectable upregulation of IFN-γ expression in sensitized skin sites suggest that local factors at the site of sensitization promote selectively the activation of Th2 cells. In this respect, TSLP promotes the secretion of Th2 cytokines with no detectable effect on the secretion of Th1 cytokines by TCR-OVA transgenic T cells stimulated in vitro
with OVA peptide (He, R. unpublished observations). In addition OVA-sensitized mice develop increased airway hyper-responsiveness (AHR) following inhalation challenge with OVA, a feature observed in asthmatic patients with AD history (Spergel et al., 1998
). Decreasing the cycles of sensitization from three to two or compressing the duration of the sensitization protocol by decreasing the interval between the cycles of sensitization leads to suboptimal development of allergic skin inflammation. The requirement for a seven-week protocol of EC sensitization, although cumbersome, appears to mimic the exacerbation of AD over time. Withdrawal of antigen sensitization at the end of the 7 week protocol results in decreased skin inflammation with IL-4 mRNA levels returning to baseline within 7-10 days; but IL-13 mRNA levels decrease over a longer period of time.
Figure 2 Histological features of OVA and saline-sensitized skin sites in BALB/c mice. Skin sections were stained with H&E and examined at 200× and 400× magnification. There is marked hyperplasia of the epidermis, a dermal infiltrate and (more ...)
We have used this mouse model to determine the critical cellular players involved in allergic skin inflammation. Using RAG2-/-
mice, which lack both T and B cells, B-cell–deficient IgH-/-
mice, T-cell receptor β-/-
mice and CD40 deficient mice, we demonstrated that TCRαβ+
T cells, but not γδ+
T cells, B cells, or CD40L-CD40 interactions, are critical for skin inflammation and the Th2 response in AD (Woodward et al., 2001
). The study of mast cell deficient mice indicated that mast cells are not important for the development of Th2 mediated skin inflammation; however they regulate IFN-γ expression in the skin (Alenius et al., 2002
) (I have a question on IFN-g expression. In this paper, Alenius showed IFN-g upregulation in the skin, whereas most of time we don't. We had said in a previous sentence that IFN-g is not changed at the peripheral skin site. Do we need to be consistent regarding to IFN-g expression in the skin?). This is important, given the role of IFN-γ in upregulating Fas expression on keratinocytes targeting them for killing by activated FasL+
T cells (Trautmann et al., 2000
) and given the role of IgE-mediated reactions in exacerbating AD (Milgrom, 2002
). A recent study demonstrated that iNKT cells are not required for allergic skin inflammation in this model. Skin infiltration by eosinophils and CD4+
cells and expression of mRNA encoding IL-4 and IL-13 in OVA-sensitized skin were similar in WT and CD1d-/-
mice. No significant increase in iNKT cells was detectable in epicutaneously sensitized skin. In contrast, iNKT cells were found in the bronchoalveolar lavage (BAL) fluid from OVA-challenged epicutaneously sensitized WT mice, but not CD1d-/-
mice, and EC sensitized CD1d-/-
mice had decreased expression of IL-4, IL-5, and IL-13 mRNA in the lung and impaired AHR in response to airway challenge with OVA (ElKhal et al., 2006
We have used the EC sensitization model to examine the role of a number of molecules cytokines, chemokines and molecules of innate immunity in the development of allergic skin inflammation elicited by EC exposure to allergens (Kawamoto et al., 2004
; Laouini et al., 2005
; Ma et al., 2002
; Spergel et al., 1999
). Both the Th2 cytokines IL-4 and IL-5 and the Th1 cytokine IFN-γ play important roles in the inflammation and hypertrophy of the skin in AD. Eosinophils are virtually absent in OVA-sensitized skin sites of IL-5-/-
mice, OVA-sensitized skin sites of IL-4-/-
mice have increased inflammatory cells but decreased eosinophils, and those of IFN-γ -/-
mice have decreased thickening of the dermal layer (Spergel et al., 1999
IL-10 plays an important role in the Th2 response to antigen and in the development of skin eosinophilia in our model (Laouini et al., 2003a
). Skin infiltration by eosinophils and expression of eotaxin, IL-4, and IL-5 mRNA in OVA-sensitized skin sites were all severely diminished in IL-10-/-
mice. Following in vitro
re-stimulation with OVA, splenocytes from EC-sensitized IL-10-/-
mice secreted significantly less IL-4, but significantly more IFN-γ than splenocytes from WT controls. IL-10-/-
APCs skewed the in vitro
response of OVA T cell receptor (TCR) transgenic T cells towards Th1. Examination of the Th response of WT and IL-10-/-
mice immunized with OVA-pulsed WT or IL-10-/-
DCs revealed that both DCs and T cells participate in IL-10-mediated skewing of the Th2 response in vivo. Current experiments are addressing the hypothesis that IL-10 released by keratinocytes following mechanical injury might promote the Th2 response to EC sensitization through polarizing skin DCs to support Th2 differentiation.
The CC chemokine receptor 3 (CCR3) is expressed by eosinophils, mast cells, and Th2 cells. Recruitment of eosinophils to OVA-sensitized skin was severely impaired in CCR3-/-
mice. These mice also have impaired recruitment of eosinophils in their lung parenchyma and BAL fluid and fail to develop AHR to methacholine following antigen inhalation. These results suggest that CCR3 plays an essential role in eosinophil recruitment to the skin and the lung and in the development of AHR (Ma et al., 2002
). Skin homing T cells express the chemokine receptor CCR4. The CCR4 ligand TARC is highly expressed in AD skin lesions. Experiments with CCR4-/-
mice have revealed decreased CD4+
cell infiltration in OVA sensitized sites as well as decreased expression of IL-4 and IL-13 mRNA levels (our unpublished observations). CCR10 is also expressed on a subset of skin homing cells. Anti-CCR10 was reported to inhibit skin inflammation in response to EC sensitization with OVA (Homey et al., 2002
). However, we have found that CCR10-/-
mice develop normal allergic skin inflammation (our unpublished observations).
Recently, we found that EC sensitization with OVA drives the generation of IL-17-producing T cells in draining lymph nodes and spleen and a local and systemic Th17 response (He et al., 2007
). OVA inhalation by EC-sensitized mice induced IL-17 and CXCL2 expression and neutrophil influx in the lung along with bronchial hyperreactivity, which were reversed by IL-17 blockade. This is in contrast to the eosinophil dominated response to airway challenge of intraperitonealy immunized mice. Although IL-17 was expressed in EC sensitized skin, there was little expression of CXCL2 and little infiltration of neutrophils at EC sensitized skin sites. However, mechanical injury upregulated the expression of IL-6 and IL-23 in skin. IL-6, like TGFβ is an inducer of Th17 cells (Veldhoen et al., 2006
), while IL-23 promotes the growth of these cells (Langrish et al., 2005
). DCs trafficking from skin to lymph nodes expressed more IL-23 and induced more IL-17 secretion by naïve T cells than splenic dendritic cells. This was inhibited by neutralizing IL-23 in vitro
and by intradermal injection of anti-TGFβ neutralizing antibody in vivo
. These findings suggest that initial cutaneous exposure to antigens in patients with AD may selectively induce the generation of IL-17 producing cells. Upon antigen inhalation these cells are recruited to the lungs where they are activate to secrete IL-17, which drives a neutrophil rich inflammation in the airways. These findings should prompt a search for airway and lung neutrophils in AD patients who develop asthma in response to inhalation of EC sensitizers. This would have become important therapeutic implications.
We have identified a number of negative regulators of allergic skin inflammation in our model. C3aR-/-
mice exhibited an exaggerated Th2 response to EC sensitization with OVA. Presentation of OVA peptide by C3aR-/-
APCs caused significantly more IL-4 and IL-5 secretion by T cells from TCR-OVA DO11.10 transgenic mice compared with presentation by WT APCs. C3a inhibited the ability of splenocytes, but not of highly purified T cells, to secrete Th2 cytokines in response to TCR ligation. This inhibition was mediated by IL-12 secreted by APCs in response to C3a. These results suggest that C3a-C3aR interactions inhibit the ability of APCs to drive Th2 cell differentiation in response to epicutaneously introduced antigen (Kawamoto et al., 2004
). COX-2 was also shown to limit the Th2 response to EC sensitization. Infiltration by eosinophils and expression of IL-4 mRNA in ovalbumin-sensitized skin sites, OVA specific IgE and IgG1 antibody responses, and IL-4 secretion by splenocytes after OVA re-stimulation were all significantly increased in EC mice that received NS-398, a COX-2 inhibitor. In contrast, OVA specific IgG2a antibody response and IFN-γ secretion by splenocytes after OVA re-stimulation were significantly decreased in these mice. COX-2-deficient mice also exhibited an enhanced systemic Th2 response to EC sensitization. These findings are important as they suggest that COX inhibitors may worsen allergic skin inflammation in patients with AD (Laouini et al., 2005
). Complement component C3 is synthesized by keratinocytes and is activated after skin injury. Skin Infiltration by eosinophils and expression of Th2 cytokines in OVA-sensitized skin sites was impaired in C3-/-
mice. Splenocytes from epicutaneously sensitized C3-/-
mice secreted less IL-4, IL-5, IL-13, and IFN-γ in response to OVA re-stimulation than splenocytes from WT control animals. C3-/-
mice also had impaired IgG1, IgG2a, and IgE antibody responses after both epicutaneous immunization. These results suggest that C3 plays an important role in both the Th1 and Th2 response to antigen in AD. (Yalcindag et al., 2006
). The opposing consequences of C3aR and C3 deficiency in our model, suggests that C3 degradation products other than C3a may promote allergic skin inflammation. C3b is a good candidate as its receptor is expressed on DCs.
Mechanical injury is critical in our model, because application of OVA to the skin of hairless mice does not result in the development of an immune response to OVA. Recently we have begun to test the hypothesis that mechanical injury allows not only the breaching of the skin barrier and the entry of antigen which is then captured by skin DCs, but also releases mediators that may play critical roles in polarizing the DCs to drive the differentiation of Th2 cells in draining lymph nodes (DLN). Gene array analysis of mouse skin 12 hrs after skin injury reveals the upregulation of a number of cytokines with a remarkable increase in IL-6, a cytokine which is important for both Th2 and Th17 differentiation differentiation. There is also increase in IL-23, IL-1 and IL-10 gene expression. In addition a number of chemokine genes, as well as genes for metalloproteinases and kallikeins are highly upregulated. These injury-induced molecules are likely to play an important role in determining the polarity of the immune response to EC sensitization. In this regard blocking IL-23 blocks the Th17 response (Chen et al., 2006
; Langrish et al., 2005
) and blocking IL-10 impairs the Th2 response (Oh et al., 2002
). We are using FITC painting of shaved versus shaved and tape stripped skin to track DCs that have emigrated from skin to DLN in order to test the hypothesis that this polarization effect is exerted at the level of the DCs that carry antigen from skin to DLN (). Preliminary data suggests that FITChi
DCs isolated from DLN of shaved tape stripped skin induce significantly more Th2 cytokine secretion in TCR-OVA transgenic D011.10 cells than FITChi
DCs isolated from DLN of shaved skin that has not been tape stripped. Comparative analysis of the genes differentially expressed by these two populations of DCs should help elucidate the nature of the “danger signal” elicited by mechanical skin injury that results in the generation of a predominantly Th2 response to EC sensitization.
Scheme for testing the effect of tape stripping on DC polarity.
I.3. Hapten induced mouse models of AD
Haptens such as oxazolone (Ox) and trinitrochlorobenzene (TNCB) are commonly used to induce allergic contact dermatitis and have been thought to evoke primarily a Th1 dominated response. However, it has been recently reported that multiple challenges with oxazolone or TNCB to the skin of hairless mice over a extended period causes the skin inflammation to shift from a typical Th1 dominated delayed-type hypersensitivity response to a chronic Th2 dominated inflammatory response that is similar to human AD.(Matsumoto et al., 2004
) (Man et al., 2007
). Indeed, nine to ten challenges with Ox to hairless mice produced a chronic Th2-like skin inflammation. The inflammation was characterized by dermal infiltration of Th2 lymphocytes that express the PGD2 receptor CRTH, mast cells and eosinophils, increased expression of IL-4 in the dermis and highly elevated IgE levels. Repeated challenge with Ox led to increased epidermal hyperplasia and decreased expression of the skin differentiation proteins filaggrin, loricrin, and involucrin. A skin barrier abnormality became evident and was associated with decreased stratum corneum ceramide content, decreased stratum corneum hydration, transepidermal water loss, and impaired lamellar body secretion, resulting in reduced lamellar membranes, as observed in AD patients. Furthermore, as in human AD, epidermal serine protease activity in SC (define) increased and expression of two lamellar body-derived antimicrobial peptides, CRAMP and mBD3, declined after Ox challenges, paralleling the decrease of their human homologues in AD skin lesions. These changes were not observed after a single challenge with hapten, the classical way to elicit hapten delayed hypersensitivity reaction.
Although the hapten repeated sensitization model is not a genetically driven model, many of its aspects may be applicable to extrinsic allergen driven AD. Indeed, it particularly illustrates the notion that once allergen is introduced via a breach in the barrier, the resulting allergen driven inflammation further damages the skin barrier. This amplificatory cycle may play an important role in the perpetuation and exacerbation of human AD. This model needs to be compared head to head with a protein (OVA and HDM) repeated sensitization model. Because of its reproducibility, predictability, low cost, and relative rapidity, the hapten repeated sensitization model could prove useful for evaluating pathogenic mechanisms and potential therapies for AD.