Induction of Thymic Stromal Lymphopoietin Expression in Keratinocytes Is Necessary for Generating an Atopic Dermatitis upon Application of the Active Vitamin D3 Analogue MC903 on Mouse Skin

atopic dermatitis lymph node T helper 2 thymic stromal lymphopoietin wild-type Thymic stromal lymphopoietin (TSLP) appears to be a master switch for T helper (Th)2 allergic inflammation such as atopic dermatitis (AD) and asthma (Liu, 2006Liu Y.J. Thymic stromal lymphopoietin: master switch for allergic inflammation.J Exp Med. 2006; 203: 269-273Google Scholar). TSLP expression is increased in keratinocytes in skin lesions of patients with AD (Soumelis et al., 2002Soumelis V. Reche P.A. Kanzler H. Yuan W. Edward G. Homey B. et al.Human epithelial cells trigger dendritic cell mediated allergic inflammation by producing TSLP.Nat Immunol. 2002; 3: 673-680Google Scholar), and various mouse models have established a link between TSLP expression and the pathogenesis of AD (Li et al., 2005Li M. Messaddeq N. Teletin M. Pasquali J.L. Metzger D. Chambon P. Retinoid X receptor ablation in adult mouse keratinocytes generates an atopic dermatitis triggered by thymic stromal lymphopoietin.Proc Natl Acad Sci USA. 2005; 102: 14795-14800Google Scholar, Li et al., 2006Li M. Hener P. Zhang Z. Kato S. Metzger D. Chambon P. Topical vitamin D3 and low-calcemic analogs induce thymic stromal lymphopoietin in mouse keratinocytes and trigger an atopic dermatitis.Proc Natl Acad Sci USA. 2006; 103: 11736-11741Google Scholar; Yoo et al., 2005Yoo J. Omori M. Gyarmati D. Zhou B. Aye T. Brewer A. et al.Spontaneous atopic dermatitis in mice expressing an inducible thymic stromal lymphopoietin transgene specifically in the skin.J Exp Med. 2005; 202: 541-549Google Scholar). Transgenic mice overexpressing TSLP in keratinocytes have been shown to develop a human AD-like syndrome (Li et al., 2005Li M. Messaddeq N. Teletin M. Pasquali J.L. Metzger D. Chambon P. Retinoid X receptor ablation in adult mouse keratinocytes generates an atopic dermatitis triggered by thymic stromal lymphopoietin.Proc Natl Acad Sci USA. 2005; 102: 14795-14800Google Scholar; Yoo et al., 2005Yoo J. Omori M. Gyarmati D. Zhou B. Aye T. Brewer A. et al.Spontaneous atopic dermatitis in mice expressing an inducible thymic stromal lymphopoietin transgene specifically in the skin.J Exp Med. 2005; 202: 541-549Google Scholar), indicating that TSLP expression is sufficient to initiate AD-like skin inflammatory responses. We have also reported that topical application of either active vitamin D3 or the low-calcemic analogue MC903 (calcipotriol) triggers a mouse AD-like syndrome, characterized by a red, scaly, and lesioned skin, accompanied by an epidermal hyperplasia and a dermal infiltration of CD4+ lymphocytes, eosinophils, dendritic cells, and mast cells, as well as by an increase of Th2 cytokine in skin, and an elevated serum IgE and blood eosinophilia (Li et al., 2006Li M. Hener P. Zhang Z. Kato S. Metzger D. Chambon P. Topical vitamin D3 and low-calcemic analogs induce thymic stromal lymphopoietin in mouse keratinocytes and trigger an atopic dermatitis.Proc Natl Acad Sci USA. 2006; 103: 11736-11741Google Scholar). Importantly, we demonstrated that MC903 treatment induced TSLP expression in mouse keratinocytes and that this induction was mediated through keratinocytic vitamin D receptor, as upon MC903 treatment, VDRep-/- mice in which vitamin D receptor was selectively ablated in keratinocytes neither expressed TSLP, nor developed an AD-like syndrome (Li et al., 2006Li M. Hener P. Zhang Z. Kato S. Metzger D. Chambon P. Topical vitamin D3 and low-calcemic analogs induce thymic stromal lymphopoietin in mouse keratinocytes and trigger an atopic dermatitis.Proc Natl Acad Sci USA. 2006; 103: 11736-11741Google Scholar). However, it has not yet been demonstrated that TSLP expression in keratinocytes is necessary for the generation of an AD-like phenotype. It is indeed known that vitamin D3 has pleiotropic actions (Lin and White, 2004Lin R. White J.H. The pleiotropic actions of vitamin D.Bioessays. 2004; 26: 21-28Google Scholar), and it has been shown that vitamin D3 can have a direct effect on naive CD4+ T cells to enhance the development of Th2 cells (Boonstra et al., 2001Boonstra A. Barrat F.J. Crain C. Heath V.L. Savelkoul H.F. O'Garra A. 1alpha,25-Dihydroxyvitamin d3 has a direct effect on naive CD4(+) T cells to enhance the development of Th2 cells.J Immunol. 2001; 167: 4974-4980Google Scholar), and also act directly on antigen-presenting cells (such as macrophages and dendritic cells) (Lin and White, 2004Lin R. White J.H. The pleiotropic actions of vitamin D.Bioessays. 2004; 26: 21-28Google Scholar). With the aim of demonstrating the indispensability of keratinocytic TSLP in the generation of MC903-induced AD, a loxP-flanked (“floxed”) TSLP mouse line in which exons 2, 3, and 4 of both alleles of the TSLP gene are floxed (TSLPL2/L2 mice) was produced (Figure 1a; Supplementary Methods), and crossed with K14-Cre transgenic mice (Li et al., 2001Li M. Chiba H. Warot X. Messaddeq N. Gerard C. Chambon P. et al.RXR-alpha ablation in skin keratinocytes results in alopecia and epidermal alterations.Development. 2001; 128: 675-688Google Scholar) to obtain keratinocyte-selective TSLP mutant mice (TSLPep-/-), or with cytomegalovirus-Cre transgenic mice (Dupé et al., 1997Dupé V. Davenne M. Brocard J. Dollé P. Mark M. Dierich A. et al.In vivo functional analysis of the Hoxa-1 3′ retinoic acid response element (3′RARE).Development. 1997; 124: 399-410Google Scholar) to generate germline null mutant mice (TSLP-/-). Quantitative PCR analyses of L- and L2 alleles (Supplementary Methods) of various tissues from 6-week-old TSLPep-/- mice showed, as expected, that the excision of TSLP selectively occurred in the epidermis, as well as in salivary gland, tongue, and thymus (Figure 1b; data not shown). TSLP RNA level was severely diminished in TSLPep-/- skin when compared to age- and sex-matched control wild-type (WT) littermate mice (Figure 2k). Accordingly, when 6-week-old female TSLPep-/- and WT mice were daily topically treated with MC903 on both ears (2nmol per ear) for 3 days, and ears were sampled at day 4 (D4), TSLP was detected by immunohistochemistry in epidermal keratinocytes of MC903-treated WT ears, whereas no signal could be seen in MC903-treated TSLPep-/- ears (Figure 1c). Consistently, MC903-treated WT mice exhibited a high TSLP serum level (>2500pg/ml) at D4 (Li et al., 2006Li M. Hener P. Zhang Z. Kato S. Metzger D. Chambon P. Topical vitamin D3 and low-calcemic analogs induce thymic stromal lymphopoietin in mouse keratinocytes and trigger an atopic dermatitis.Proc Natl Acad Sci USA. 2006; 103: 11736-11741Google Scholar), whereas in TSLPep-/- mice, MC903 failed to induce serum TSLP (<16pg/ml; Figure 1d). Download .pdf (.07 MB) Help with pdf files Supplementary MethodsFigure 2Induction of TSLP expression in keratinocytes is necessary for generating an AD-like syndrome upon application of the active vitamin D3 analogue MC903 on mouse skin. (a) Appearance and (b) hematoxylin/eosin-stained sections of ears sampled at day 16 (D16) from wild-type (WT) and TSLPep-/- mice, topically ear-treated with ethanol (EtOH, as vehicle control) or MC903 (1nmol per ear) daily for 15 days. Arrows point to dermal/epidermal junction. Scale bar, 50μm. (c–g) Immunohistochemical (IHC) staining of CD3 (c), CD4 (d), CD11b (e), GR1 (f), and CD11c (g) on ear sections from MC903 topically treated WT (left panels) and TSLPep-/- (right panels) mice at D16. White dashed lines indicate the dermal/epidermal junction. Yellow corresponds to staining of antibodies, whereas blue corresponds to DAPI staining of nuclei. White arrows point to one of the specific antibody-stained cells in the dermis, and white arrowheads to one of the specific antibody-stained cells in the epidermis. Scale bars, 50μm. (h) Toluidine blue (TB)-stained sections. Red arrows point to one of the mast cells with intense blue in the dermis. Scale bar, 50μm. (i) Serum IgE levels of MC903-treated WT and TSLPep-/- mice at D0, D8, and D16. (j) Eosinophil counts in blood of MC903-treated WT and TSLPep-/- mice at D0 and D16. (k, l) Cytokine and chemokine expression in ears (k) and ear-draining lymph nodes (l) of EtOH or MC903-treated WT and TSLPep-/- mice at D16. *P<0.05. Error bars indicate s.d.View Large Image Figure ViewerDownload (PPT)Figure 2Induction of TSLP expression in keratinocytes is necessary for generating an AD-like syndrome upon application of the active vitamin D3 analogue MC903 on mouse skin. (a) Appearance and (b) hematoxylin/eosin-stained sections of ears sampled at day 16 (D16) from wild-type (WT) and TSLPep-/- mice, topically ear-treated with ethanol (EtOH, as vehicle control) or MC903 (1nmol per ear) daily for 15 days. Arrows point to dermal/epidermal junction. Scale bar, 50μm. (c–g) Immunohistochemical (IHC) staining of CD3 (c), CD4 (d), CD11b (e), GR1 (f), and CD11c (g) on ear sections from MC903 topically treated WT (left panels) and TSLPep-/- (right panels) mice at D16. White dashed lines indicate the dermal/epidermal junction. Yellow corresponds to staining of antibodies, whereas blue corresponds to DAPI staining of nuclei. White arrows point to one of the specific antibody-stained cells in the dermis, and white arrowheads to one of the specific antibody-stained cells in the epidermis. Scale bars, 50μm. (h) Toluidine blue (TB)-stained sections. Red arrows point to one of the mast cells with intense blue in the dermis. Scale bar, 50μm. (i) Serum IgE levels of MC903-treated WT and TSLPep-/- mice at D0, D8, and D16. (j) Eosinophil counts in blood of MC903-treated WT and TSLPep-/- mice at D0 and D16. (k, l) Cytokine and chemokine expression in ears (k) and ear-draining lymph nodes (l) of EtOH or MC903-treated WT and TSLPep-/- mice at D16. *P<0.05. Error bars indicate s.d.View Large Image Figure ViewerDownload (PPT) TSLPep-/- pups grew normally and did not exhibit any apparent abnormalities with time (3–50 weeks), and histological analyses in the skin did not reveal any obvious differences between WT and TSLPep-/- mice (Figure 2b; data not shown). MC903 (1nmol per ear) was applied daily on 8-week-old WT or TSLPep-/- ears for 15 days to generate the AD-like syndrome (Li et al., 2006Li M. Hener P. Zhang Z. Kato S. Metzger D. Chambon P. Topical vitamin D3 and low-calcemic analogs induce thymic stromal lymphopoietin in mouse keratinocytes and trigger an atopic dermatitis.Proc Natl Acad Sci USA. 2006; 103: 11736-11741Google Scholar). At D16, ethanol (as vehicle control) treatment had no effect on WT and TSLPep-/- ears (Figure 2a, upper panel), whereas MC903-treated WT ears were red, scaly, swollen, lesioned, and crusted (Figure 2a, lower left panel). In contrast, MC903-treated TSLPep-/- ears appeared dry and scaly, but were not red or inflamed, and no lesion developed (Figure 2a, lower right panel). Histologically, at D16, MC903-treated WT skin exhibited an epidermal hyperplasia and a heavy dermal cell infiltrate (Figure 2b, lower left panel) that included numerous eosinophils (as shown by hematoxylin/eosin staining (Figure 2b, inset), and by CD11b or GR1 immunohistochemistry that stains granulocytes), CD3+ T lymphocytes mainly comprising CD4+ helper T cells, CD11c+ dendritic cells, and mast cells (toluidine blue staining) (Figure 2c–h; Supplementary Methods), as previously reported (Li et al., 2006Li M. Hener P. Zhang Z. Kato S. Metzger D. Chambon P. Topical vitamin D3 and low-calcemic analogs induce thymic stromal lymphopoietin in mouse keratinocytes and trigger an atopic dermatitis.Proc Natl Acad Sci USA. 2006; 103: 11736-11741Google Scholar). In contrast, dermal infiltration was largely abrogated in MC903-treated TSLPep-/- ears, showing only few eosinophils, CD4+ helper T cells, dendritic cells, and mast cells (Figure 2c–h). However, some epidermal hyperplasia was observed (Figure 2b, lower right panel), suggesting that MC903 application could induce some epidermal hyperplasia in a TSLP-independent manner. A systemic increase in serum IgE levels, which was lower in TSLPep-/- than in WT mice at D16, was observed upon MC903 treatment (Figure 2i). Furthermore, an increase in eosinophil and lymphocyte counts was observed in WT but not in TSLPep-/- blood (Figure 2j; data not shown). Taken together, these results indicated that selective ablation of TSLP in epidermal keratinocytes precludes the generation of an AD-like syndrome upon MC903 topical application. Not unexpectedly, similar observations were made with TSLP germline mutants (TSLP-/-) and with mutants in which TSLP was selectively ablated in keratinocytes of adult mice by using the tamoxifen-inducible CreERT2 recombinase (Metzger et al., 2003Metzger D. Indra A.K. Li M. Chapellier B. Calleja C. Ghyselinck N.B. et al.Targeted conditional somatic mutagenesis in the mouse: temporally-controlled knock out of retinoid receptors in epidermal keratinocytes.Methods Enzymol. 2003; 364: 379-408Google Scholar; data not shown). To further characterize the involvement of TSLP-dependent signaling in AD generation, we next examined the pattern of skin cytokine expression upon MC903 treatment (Figure 2k; Supplementary Methods). TSLP expression was highly induced by MC903 treatment in WT, but not in TSLPep-/- skin. However, a slight increase of TSLP in TSLPep-/- skin was noted upon MC903 treatment, most probably reflecting an induction of TSLP by MC903 in nonkeratinocytic cells (for example, mast cells and basophils that also express TSLP; Soumelis and Liu, 2004Soumelis V. Liu Y.J. Human thymic stromal lymphopoietin: a novel epithelial cell-derived cytokine and a potential key player in the induction of allergic inflammation.Springer Semin Immunopathol. 2004; 25: 325-333Google Scholar; Sokol et al., 2008Sokol C.L. Barton G.M. Farr A.G. Medzhitov R. A mechanism for the initiation of allergen-induced T helper type 2 responses.Nat Immunol. 2008; 9: 310-318Google Scholar), as this increase was not observed in TSLP-/- null mutant mice (Figure S1). Th2 cytokines (IL4, IL13, IL10, IL6, and IL31) and eosinophil-attractant chemokines (eotaxin-2 and monocyte chemotactic protein-2) were highly induced by MC903 in WT skin, whereas this induction was much weaker in MC903-treated TSLPep-/- skin, although not abolished, when compared with vehicle (ethanol) treatment. To examine whether this weak induction could reflect a keratinocytic TSLP-independent regulation, cytokine expression was also analyzed in TSLP-/- null mutant skin (Figure S1). IL4, IL13, IL31, and eotaxin-2 were not induced in TSLP-/- skin upon MC903 treatment, indicating that the MC903-induced expression of these cytokines is fully dependent on TSLP. In contrast, some induction of IL10, IL6, and monocyte chemotactic protein-2 was observed in TSLP-/- skin upon MC903 treatment, suggesting that their overall induction may be contributed by both TSLP-dependent and -independent signaling. On the other hand, upregulation of the Th1 cytokine IFNγ by MC903 similarly occurred in WT and TSLPep-/- skin, indicating that its induction was independent of TSLP. Download .pdf (.24 MB) Help with pdf files Supplementary Figure S1Supplementary Fig. S1 Comparison of cytokine and chemokine expression in ears of ethanol or MC903-treated wildtype (WT) and TSLP-/- null (germ line mutation of TSLP) mice. Ears were topically administrated either with ethanol (EtOH, as vehicle control) or MC903 (1 nmole per ear) daily for 15 days, and sampled for analyses at day 16 (D16). *, P<0.05. Error bars indicate s.d. Cytokine expression was also analyzed in skin-draining lymph nodes (LN; Figure 2l). Upon MC903 treatment, the induction of Th2 cytokines (IL4, IL13, IL10, IL6, and IL31) in WT LNs was not observed in TSLPep-/- LNs, indicating that the MC903-triggered Th2 response is abolished in TSLPep-/- mice. No change was seen for IFNγ transcript levels in WT or TSLPep-/- LNs, from either MC903- or ethanol-treated mice, suggesting that MC903 application does not involve a Th1 response in LNs. In conclusion, we demonstrate unequivocally here that TSLP produced by keratinocytes is absolutely required in pathogenesis of AD triggered by topical application of the vitamin D3 analogue MC903. We also show that in this AD model, induction of IL4, IL13, IL31, and eotaxin-2 is fully TSLP dependent, whereas that of IL6, IL10, and monocyte chemotactic protein-2 is only partially TSLP dependent, and that of IFNγ is TSLP independent. Finally, our floxed TSLP mice will be helpful for selective ablation of TSLP in other cell types (Soumelis and Liu, 2004Soumelis V. Liu Y.J. Human thymic stromal lymphopoietin: a novel epithelial cell-derived cytokine and a potential key player in the induction of allergic inflammation.Springer Semin Immunopathol. 2004; 25: 325-333Google Scholar; Ziegler and Liu, 2006Ziegler S.F. Liu Y.J. Thymic stromal lymphopoietin in normal and pathogenic T cell development and function.Nat Immunol. 2006; 7: 709-714Google Scholar; Holgate, 2007Holgate S.T. The epithelium takes centre stage in asthma and atopic dermatitis.Trends Immunol. 2007; 28: 248-251Google Scholar; Sokol et al., 2008Sokol C.L. Barton G.M. Farr A.G. Medzhitov R. A mechanism for the initiation of allergen-induced T helper type 2 responses.Nat Immunol. 2008; 9: 310-318Google Scholar), and therefore to further elucidate the physiological and pathological function of this cytokine. We thank the staff of the mouse knockout, histopathology, hematology, and animal facilities of the Institut de Génétique et de Biologie Moléculaire et Cellulaire (IGBMC) and Institut Clinique de la Souris for their kind help; Dr N Rochel-Guiberteau (IGBMC) and Dr D Moras (IGBMC) for vitamin D3 analogue MC903 from LEO Pharmaceutical Products (Denmark); and Dr SM Dymecki (Harvard Medical School, Boston) for FLP deleter (ACTB:FLPe) mice. This work was supported by funds from the Centre National de la Recherche Scientifique, the Institut National de la Santé et de la Recherche Médicale, the Collège de France, the Ministère de l'Enseignement supérieur et de la Recherche, the Association pour la Recherche à l'IGBMC (ARI), and l'Agence Nationale de la Recherche. Supplementary Methods Figure S1. Comparison of cytokine and chemokine expression in ears of ethanol or MC903-treated wild-type (WT) and TSLP-/- null (germline mutation of TSLP) mice.