- Email: [email protected]
Self-Antigen Presentation by Keratinocytes in the Inflamed Adult Skin Modulates T-Cell Auto-Reactivity
Accepted Article Preview: Published ahead of advance online publication www.jidonline.org
Self-Antigen Presentation by Keratinocytes in the Inflamed Adult Skin Modulates T Cell Auto-Reactivity Michael Meister, Amel Tounsi, Evelyn Gaffal, Tobias Bald, Maria Papatriantafyllou, Julia Ludwig, Georg Pougialis, Felix Bestvater, Luisa Klotz, Gerhard Moldenhauer, Thomas Tu¨ting, Gu¨nter J Ha¨mmerling, Bernd Arnold, Thilo Oelert
Cite this article as: Michael Meister, Amel Tounsi, Evelyn Gaffal, Tobias Bald, Maria Papatriantafyllou, Julia Ludwig, Georg Pougialis, Felix Bestvater, Luisa Klotz, Gerhard Moldenhauer, Thomas Tu¨ting, Gu¨nter J Ha¨mmerling, Bernd Arnold, Thilo Oelert, Self-Antigen Presentation by Keratinocytes in the Inflamed Adult Skin Modulates T Cell Auto-Reactivity, Journal of Investigative Dermatology accepted article preview 2 April 2015; doi: 10.1038/jid.2015.130. This is a PDF file of an unedited peer-reviewed manuscript that has been accepted for publication. NPG are providing this early version of the manuscript as a service to our customers. The manuscript will undergo copyediting, typesetting and a proof review before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers apply.
Received 15 October 2014; revised 16 March 2015; accepted 19 March 2015; Accepted article preview online 2 April 2015
© 2015 The Society for Investigative Dermatology
Self-antigen presentation by keratinocytes in the inflamed adult skin modulates T cell auto-reactivity
Michael Meister1, Amel Tounsi1, Evelyn Gaffal3, Tobias Bald3, Maria Papatriantafyllou1, Julia Ludwig1, Georg Pougialis1, Felix Bestvater2, Luisa Klotz4, Gerhard Moldenhauer1, Thomas Tüting3, Günter J. Hämmerling1, Bernd Arnold1, Thilo Oelert1
1
Department of Molecular Immunology and
2
Core Facility Light Microscopy, German Cancer
Research Center (DKFZ), 69120 Heidelberg, Germany 3
Laboratory of Experimental Dermatology, Department of Dermatology and Allergy, University of
Bonn, 53105 Bonn, Germany 4
Department of Neurology, Clinic for Neurology, University Hospital Münster, 48149 Münster,
Germany
Address correspondence to: Bernd Arnold, Department of Molecular Immunology, German Cancer Research Center, 69120 Heidelberg, Germany. Phone: 49.6221.42.3728; Fax: 49.6221.42.3761; Email: [email protected]
Short title: Self-antigen presentation by keratinocytes
1
© 2015 The Society for Investigative Dermatology
Abstract Keratinocytes play a pivotal role in the regulation of immune responses but the impact of antigen-presentation by these cells is still poorly understood, particularly in a situation where the antigen will be presented only in adult life. Here, we generated a transgenic mouse model in which keratinocytes exclusively present a myelin basic protein (MBP) peptide covalently linked to the MHC class II β-chain, solely under inflammatory conditions. In these mice, inflammation caused by epicutaneous contact sensitizer treatment resulted in keratinocyte-mediated expansion of MBP-specific CD4+ T cells in the skin. Moreover, repeated contact sensitizer application preceding a systemic MBP immunization reduced the reactivity of the respective CD4+ T cells and lowered the symptoms of the resulting experimental autoimmune encephalomyelitis. This down-regulation was CD4+ T cell mediated and dependent on the presence of the immune modulator Dickkopf-3. Thus, presentation of a neo self-antigen by keratinocytes in the inflamed, adult skin can modulate CD4+ T cell autoaggression at a distal organ.
2
© 2015 The Society for Investigative Dermatology
Introduction The observation that the type and strength of immune responses depend on the local microenvironment of a given tissue (Alferink et al., 1999; Matzinger, 2007) has fostered investigations on the contribution of various tissue-specific components to the resulting immune response. In the skin the prominent role of Langerhans cells (Mayerova et al., 2004; Nagao et al., 2012; Stoitzner et al., 2006) and dermal dendritic cells (Allan et al., 2006; Idoyaga et al., 2013; Itano et al., 2003) in inducing antigen specific immunity or tolerance is well established. Less is known about the role of epidermal keratinocytes in the activation or suppression of T cell effector functions. For example, keratinocyte specific expression of a nominal antigen, such as Ovalbumin (OVA) led to proliferation of OVA specific CD4+ T cells and was able to cause symptoms of an inflammatory dermatitis (Rosenblum et al., 2011). However, in this transgenic system, OVA-derived peptides were also presented by professional antigen presenting cells (APC) in the skin draining lymph nodes, which prevented to identify the contribution of antigen presentation by keratinocytes to the resulting T cell response. In a similar study, mice expressing a model antigen in keratinocytes developed a graft versus host like reaction after intradermal injection of CD8+ T cells carrying the respective specificity (Kim et al., 2009). In this case, antigen presentation by Langerhans cells and dendritic cells was experimentally excluded. Thus, it was proposed that keratinocytes can directly prime naïve CD8+ T cells in the adult skin (Kim et al., 2009). We have previously shown that CD8+ T cells with a transgenic MHC class I Kb-specific T cell receptor (TCR) become tolerant following encounter of the Kb antigen on keratinocytes (Alferink et al., 1998). As in this animal model the respective TCR recognized only the intact MHC molecule, it was proposed that T cell stimulation occurred through direct contact with keratinocytes. However, this contact mainly occurs during the neonatal phase when extensive trafficking of T cells through tissues including the skin is observed (Kimpton et al., 1995). In contrast, T cell migration into the healthy, adult skin is restricted (Mackay, 1993)
3
© 2015 The Society for Investigative Dermatology
and T cells can only infiltrate inflammatory sites (Hou and Doherty, 1993; Tietz et al., 1998; Wang et al., 2010). In parallel it was also reported that CD4+ T cells can be directly activated by keratinocytes specifically presenting peptide–MHC class II conjugates. Inflammatory skin disease was observed in 20-40% of mice expressing such a transgenic MHC class II I-Ab β-chain under the keratin14 promotor and a corresponding auto-reactive TCR on CD4+ T cells. In this study, keratinocytes were shown to activate and are the primary targets of auto-reactive CD4+ T cells (Fan et al., 2003). Here, we investigated whether keratinocytes that present a sessile self-antigen in the inflamed, adult skin can prime CD4+ T cells in the absence of antigen presentation by professional APCs and whether this activation can modulate a distant, auto-reactive T cell response. Therefore, we generated a transgenic mouse model, in which keratinocytes express a silent antigenic determinant, namely an encephalitogenic myelin basic protein (MBP1-10) peptide, covalently linked to the MHC class II IAu β-chain (Ker-MBP mice). Cell surface presentation of this transgene product requires co-expression of the endogenous MHC II α-chain, which needs to be induced by inflammatory stimuli like for example Interferon-γ (Nickoloff and Turka, 1994). This allowed us to study the contribution of keratinocytes to T cell activation in the inflamed skin and to the associated auto-reactivity at a distal site in the context of experimental autoimmune encephalomyelitis (EAE). In the Ker-MBP mouse model, transferred MBP-specific Tg4 CD4+ T cells proliferated only in sensitized skin but not in the draining lymph nodes. Repeated skin sensitization induced regulatory T cells, which limited MBP-specific CD4+ T cell reactivity and lowered EAE symptoms after MBP immunization. Moreover, the recently described immune modulator Dickkopf-3 (DKK3) (Papatriantafyllou et al., 2012) was found to contribute to the downmodulation of MBP-specific CD4+ T cell reactivity. Together, these findings suggest that selfantigen presentation by keratinocytes under inflammatory conditions can modulate autoreactive CD4+ T cell responses at a distal site.
4
© 2015 The Society for Investigative Dermatology
Results Generation and characterization of transgenic Ker-MBP mice To generate a mouse model for conditional induction of a self-antigen in the inflamed skin of adult mice, we fused a cDNA encoding the MBP1-10 peptide and a glycine-serine linker to the genomic DNA of the I-Au β-chain. The central nervous system (CNS) derived peptide MBP1-10 was chosen as a model antigen because auto-antigens of the brain can also be found in the skin under physiological conditions (Kormos et al., 2011). This construct was put under the control of a keratinIV promoter element that is known to direct expression exclusively to keratinocytes of hair follicles (Alferink et al., 1998) (Ker-MBP, Figure 1a). The transgenic MBP-IAu -chain is constitutively expressed in resting keratinocytes. However, cell surface presentation of the transgene requires co-expression of the endogenous MHC class II chain which is absent in keratinocytes in the healthy skin but induced by inflammatory stimuli like IFN-γ (Nickoloff and Turka, 1994) (Figure 1b). In order to ensure keratinocyte restricted MBP-IAu MHC class II -chain production, we initially confirmed tissue-specific expression of the transgene by RT-PCR using RNA extracted from the respective tissues (Figure 1c). Moreover,
MBP-IAu
-chain
expression
in
keratinocytes
was
assessed
by
immunofluorescence staining in the skin with a MBP1-10-specific rabbit anti-serum (Figure 1d). To clarify whether surface presentation of the MBP-MHC II complex on keratinocytes is inducible and could be recognized by T cells, TCR-transgenic MBP-specific Tg4 T cells were labelled with CFSE and transferred into Ker-MBP mice one day before application of the contact sensitizers oxazolone (OXA) or diphenylcycloprenone (DPCP). In particular, contact sensitizers were used to induce inflammation in the skin, as they were shown to be potent inducers of MHC class II expression in keratinocytes (Stringer et al., 1991). 5 days after transfer, T cells were isolated from draining lymph nodes and treated skin and analysed by flow cytometry. Proliferating Tg4 T cells could be observed in the inflamed skin of Ker-MBP mice, but were undetectable in draining lymph nodes (Figure 1e). As a control, untreated Ker-MBP mice were analysed. There, transferred Tg4 cells did not proliferate in lymph nodes
5
© 2015 The Society for Investigative Dermatology
and were not detectable in skin of these mice (Figure 1e, lower panel). However, strong proliferation was observed in draining lymph nodes when MBP-peptide loaded DCs were coinjected with the transferred Tg4 T cells (Figure 1f). Thus, these results reveal that the KerMBP mouse model allows us to the study the consequences of keratinocyte restricted selfantigen presentation in the skin, in the absence of antigen presentation by professional APC in the draining lymph nodes.
Functional presentation of the transgenic MBP-IAu MHC II ß-chain in the skin of Ker-MBP mice only occurs upon contact sensitizer treatment. In order to increase the low endogenous level of MBP-specific CD4+ T cells found in the polyclonal TCR repertoire, we crossed Ker-MBP mice to the MBP-specific TCR transgenic Tg4 mice (Liu et al., 1995). This allowed us to study the functionality of the MBP-MHC class II complex presentation by keratinocytes in vivo in more detail. Furthermore, it enabled us to exclude accidental thymic or peripheral presentation of the transgene, which might induce tolerance against the MBP antigen (Alferink et al., 1998; Kurschus et al., 2006). Analysis of the thymus revealed comparable proportions of thymocyte subsets in Tg4 and Tg4xKer-MBP mice (Figure 2a) indicating a normal T cell development in both mouse strains. Furthermore, CD4+ T cell proportions as well as their memory differentiation status, identified by surface expression of CD62L and CD44, were similar in in Tg4 and Tg4xKer-MBP mice (Figure 2b), suggesting that there is no induction of peripheral tolerance. To asses this on a functional level, Ker-MBP positive and negative Tg4 mice were challenged with MBP peptide in CFA. Both types of mice showed similar levels of EAE (Figure 2c) revealing that there is no influence of the Ker-MBP transgene on the functional readout system of our mouse model per se. Importantly, contact sensitizer treatment resulted in the accumulation of CD4+ T cells in the skin of Tg4xKer-MBP mice but not in the skin of Tg4 mice, 8 days after one single application. However, this was not accompanied by severe skin disease like scaling or
6
© 2015 The Society for Investigative Dermatology
alopecia. As well, we did not observe interface dermatitis or an eczematous reaction in the double transgenic mice (Figure 2d).
Repeated skin sensitization limits EAE symptoms in Ker-MBP mice. Next, we asked whether treatment of Ker-MBP mice with contact sensitizers would have a functional impact on a following, active immunization with MBP. Untreated Ker-MBP and wt mice developed comparable EAE symptoms upon MBP1-10 immunization (Figure 3a). As repeated exposure of T cells to self-antigens in tissues has been reported to attenuate the severity of autoimmune responses (Lara-Corrales and Pope, 2010; Rosenblum et al., 2011), we decided to perform repetitive skin treatments with a contact sensitizer before EAE induction. And indeed, 4 repetitive pre-treatments with DPCP reduced the severity of EAE symptoms in Ker-MBP mice compared to littermate controls (Figure 3b). However, this difference was apparently not due to a generally impaired expression of the endogenous MHC class II α-chain in the skin of Ker-MBP mice (Figure 3c). Reduction of the anti-MBP T cell response in Ker-MBP mice was not based on a general DPCP-induced immune suppression in the skin, as pre-treatment did not affect a delayed-type hypersensitivity response to another contact sensitizer, oxazolone (Figure 3d). Thus, DPCP-mediated chronic inflammation in the skin of Ker-MBP mice results in continuous cell surface presentation of the MBP peptide on keratinocytes, which in turn down-modulates a following MBP-specific T cell response in the brain.
CD4+ T cells mediate the down-modulation of T cell reactivity in Ker-MBP mice To clarify the mechanisms that limit the MBP-induced autoimmunity shown in Figure 3b, it was initially investigated whether this was associated with a dominant form of immuneregulation. EAE induced by the transfer of MBP-activated splenocytes resulted in significantly less severe EAE symptoms in DPCP pre-treated Ker-MBP mice compared to non-transgenic
7
© 2015 The Society for Investigative Dermatology
littermates (Figure 4a). Furthermore, CD4+ T cells were isolated from spleens of DPCP pretreated wt and Ker-MBP mice and transferred into wt mice. Upon MBP immunization, lower EAE scores were observed in recipients of T cells isolated from pre-treated Ker-MBP mice compared to animals that received T cells from pre-treated non-transgenic littermates (Figure 4b). These findings suggest that the down-modulation of MBP-specific T cell reactivity in pretreated Ker-MBP mice was mediated by regulatory T cells. However, we could neither observe changes in the frequency of CD4+CD25+Foxp3+ regulatory T cells in the lymph nodes and spleens of DPCP treated Ker-MBP mice nor could detect altered expression of forkhead box P3 (FOXP3) in skin of these mice (Figure 5c, d). Next, we analysed the expression levels of several immune modulators like programmed cell death 1 ligand (PD-L1) that was shown to be an important regulator of CD8+ T cell effector function in a similar mouse model (Okiyama and Katz, 2014). However, no differential expression of interleukin10 (IL-10), transforming growth factor β (TGFβ), cytotoxic T lymphocyte associated antigen 4 (CTLA4) and PD-L1 could be detected in untreated or DPCP treated skin between wt and Ker-MBP groups (Figure 4e). We have previously reported that the secreted protein Dickkopf-3 (DKK3), which is also expressed in hair follicles (Ohyama et al., 2006), can limit CD8+ T cell reactivity in vitro and in vivo (Papatriantafyllou et al., 2012). Analysis of DKK3 expression revealed a significant up-regulation in DPCP pre-treated skin of Ker-MBP mice compared to non-transgenic littermates (Figure 4e). However, this up-regulation was not accompanied by an increase of the systemic DKK3 level in the serum (Figure 4f). Moreover, no DKK3 expression in CD4+ T cells isolated from spleen and lymph nodes of untreated and DPCP pre-treated Ker-MBP and wt mice (Figure 4g).
Limitation of T cell reactivity in Ker-MBP mice depends on the presence of Dickkopf-3. In order to clarify whether DKK3 contributes to the limitation of CD4+ T cell reactivity in pretreated transgenic animals, Ker-MBP mice deficient for DKK3 were established and fourtimes pre-treated with DPCP. In contrast to DKK3-competent mice, pre-treatment did not result in a reduction of EAE symptoms in DKK3 deficient Ker-MBP mice in comparison to
8
© 2015 The Society for Investigative Dermatology
non-transgenic Dkk3-/- littermates (Figure 5a). Additionally, we blocked DKK3 function in KerMBP and wt mice by administration of a neutralizing DKK3-specific antibody at the beginning of the DPCP treatment. Similar to the DKK3-deficient situation, we did not observe a reduction of MBP-triggered EAE as a result of DPCP pre-treatment in Ker-MBP mice compared to wt controls (Figure 5b). Absence of EAE down-modulation was again not due to altered endogenous MHC class II expression (Figure 5c). Finally, we investigated whether the observed changes in EAE severity were linked with the effector function of the respective MBP-specific CD4+ T cells. Wild-type and Ker-MBP mice were immunized with MBP1-10 and seven days later splenic T cells were re-stimulated in vitro with the respective peptide. The percentages of MBP-specific CD4+ T cells producing interleukin-2 (IL-2), tumour necrosis factor (TNF) and interleukin-17 (IL-17) were comparable in the two experimental groups (Figure 5d). However, the percentages of IL-2, TNF and IL-17 producing MBP-specific T cells were significantly reduced among in vitro re-stimulated CD4+ T cells that had been isolated from Ker-MBP mice following four-times DPCP treatment. By contrast, the effector function of MBP-specific CD4+ T cells from DPCP-treated KerMBP.Dkk3-/- mice was not compromised in vitro compared to untreated controls (Figure 5d). Thus, chronic skin inflammation, as modelled by repeated DPCP treatment and the associated priming of MBP-specific CD4+ T cells by keratinocytes in the skin can reduce systemic CD4+ T cell reactivity in an antigen-dependent manner through a mechanism that involves DKK3 expression.
9
© 2015 The Society for Investigative Dermatology
Discussion The skin provides a first line of defence against physical insults, toxins and pathogens. Therefore, immune surveillance of this barrier has to be finely balanced to ensure its integrity. Here, we asked how T cells respond to cognate auto-antigens that are presented exclusively by keratinocytes of the inflamed skin, and whether this presentation can influence T cell reactivity against the respective antigen in a distant organ. Our data suggest that (i) keratinocytes can present self-antigen to stimulate CD4+ T cells under inflammatory conditions; (ii) repeated antigen stimulation in the inflamed skin leads to down-modulation of T cell responses; (iii) this down-modulation is CD4+ T cell mediated and depends on the expression of Dickkopf-3. In our Ker-MBP mouse model, a MBP peptide that is covalently linked to the MHC class II A u β-chain is exclusively and constitutively expressed in keratinocytes of hair follicles. In naïve mice carrying the Ker-MBP transgene, MBP-specific CD4+ T cells do not seem to be aware of their cognate antigen, suggesting that the MBP-peptide is not presented under steady state conditions. Upon skin sensitization, MBP presentation on the cell surface of keratinocytes is enabled. If professional APCs would take up the MBP peptide in this situation one would expect MBPspecific T cell proliferation in the draining lymph nodes after migration of CD11c+ cells was induced by contact sensitizers. However, proliferation of MBP-specific T cells was only observed in the skin and was absent from draining lymph nodes. These observations support the view that in the Ker-MBP mouse model the MBP1-10 peptide represents a truly sessile antigen in the skin that can be presented exclusively by keratinocytes and only upon inflammation. The cutaneous inflammation that was induced by a single treatment with the contact sensitizer oxazolone was more potent in Tg4xKer-MBP mice than in Tg4 mice. However, severe skin disease, such as scaling and alopecia was not observed in the double transgenic mice. This is in contrast to what was observed in a transgenic mouse model where a nominal antigen, such as Ovalbumin (OVA), was inducibly expressed in keratinocytes (Rosenblum et 10
© 2015 The Society for Investigative Dermatology
al., 2011). In these mice, OVA-specific T cells caused pronounced inflammatory dermatitis that spontaneously resolved after 20-30 days (Rosenblum et al., 2011). The different results obtained in the two models might reflect the fact that OVA can be presented to DO11 T cells by professional APCs in the Ker-OVA mouse model, whereas MBP is presented to Tg4 T cells exclusively by keratinocytes in the Ker-MBP system. There is increasing evidence that repeated exposure of T cells to self-antigens in tissues attenuates the severity of autoimmune responses (Lara-Corrales and Pope, 2010; Rosenblum et al., 2011). Accordingly, in our mouse model repeated applications of DPCP that allowed for continuous availability of the MBP-MHC class II complex on keratinocytes limited the severity of subsequent EAE. Interestingly, DPCP is used clinically for the treatment of alopecia areata because of its proposed immune-modulatory capacity (Hoffmann et al., 1994). It was reported that epicutaneous immunization with the auto-antigenic MBP peptide was able to inhibit the induction of EAE (Bynoe et al., 2003). This protection was based on CD4+ T cell regulation, but was independent of CD4+CD25+ regulatory T cells and was not mediated by immune modulators such as IL-4, IL-10 or TGFβ. In this study, the MBP peptide was shown to be presented by professional APCs in the draining lymph nodes. It is interesting that repeated tolerogenic epicutaneous applications of MBP peptide have similar consequences as continuous MBP peptide presentation by keratinocytes under inflammatory conditions in our mouse model. DKK3 was found to be significantly up-regulated in the skin of DPCP-treated Ker-MBP mice compared to littermate controls. We have recently reported that DKK3 can limit CD8+ T cell mediated tumor and skin graft rejection (Papatriantafyllou et al., 2012). DKK3 is mainly expressed at sites that have been traditionally referred to as immune privileged including the hair follicle (Barrantes Idel et al., 2006; Hermann et al., 2007; Ohyama et al., 2006). It is therefore, tempting to speculate that DKK3 may contribute to the immunosuppressive milieu of such tissues. Here we show that abrogation of DKK3 function was sufficient to reverse the reduction of T cell-mediated EAE symptoms in DPCP-treated Ker-MBP mice. Furthermore,
11
© 2015 The Society for Investigative Dermatology
expression of DKK3 in the skin was associated with decreased cytokine production by MBPspecific CD4+T cells. Whether DKK3 acts directly or indirectly on CD4+T cells in this in vivo setting is presently unknown. Neutralizing DKK3-specific antibodies could block the downmodulation of CD4+T cell reactivity when administered at the beginning of the DPCP treatment. This finding suggests that DKK3 is involved in the induction of the observed T cell regulation rather than directly suppressing CD4+T cell reactivity in the skin. Further investigations are needed to clarify the mechanism through which DKK3 limits CD4+T cell reactivity. In summary, based on our newly established mouse model we present evidence that the induction and presentation of a self-antigen by keratinocytes during skin inflammation is sufficient to modulate T cell auto-reactivity at a distal site. Keratinocytes could activate MBPspecific T cells in the skin independent from professional APCs. Furthermore, persistent stimulation of MBP-specific CD4+ T cells by keratinocytes in the chronically inflamed skin led to a reduction of EAE symptoms. Thus, repeated antigen-presentation by keratinocytes during skin inflammation may dampen the respective T cell reactivity at a distal site.
12
© 2015 The Society for Investigative Dermatology
Material and Methods Mice B10.PL mice were obtained from Jackson Laboratories. TCR-transgenic Tg4 mice have been described (Acha-Orbea et al., 1988). Dkk3 -/- mice (Barrantes Idel et al., 2006) were crossed to a H-2KU background (B10.PL). Mice were maintained under specific pathogen free conditions at the German Cancer Research Center. All animal experiments were approved by the local government authorities (Regierungspräsidium Karlsruhe, Germany).
Generation of transgenic Ker-MBP mice For keratinocyte-specific expression the 2.4 kb fragment of the Keratin IV promoter was used (Alferink et al., 1998). The introduction of the signal sequence and the MBP1-10-peptide (ASQKRPSQRS) into the I-Au β-chain cDNA has been described (Kurschus et al., 2006). To assure appropriate cleavage of the signal sequence the amino acid sequence GDS was introduced N-terminal to the MBP-sequence. As a substitute of natural acetylation a glycine residue was inserted between the GDS sequence and the MBP1-10 sequence. We replaced the cDNA after exon 2 with the genomic DNA of the I-Ab β-chain (obtained from RZPD; identical with the β-chain I-Au coding sequence). The DNA construct was linearized and injected into C57BL/6 x DBA/2 F2 oocytes. Transgenic mice were backcrossed more than 10 times to B10.PL and were crossed to Tg4 mice.
Histology Skin tissue was fixed in a zinc-based fixative (Dako), embedded in paraffin, and routinely stained with H&E. Furthermore, sections were stained with rat anti-mouse CD4 or rat antimouse MAC-3 followed by a biotinylated rabbit anti-rat IgG and development with the LSABAP kit (Dako).
13
© 2015 The Society for Investigative Dermatology
Immunofluorescence Skin tissue was fixed with phosphate-buffered 4% formaldehyde. Five-micrometer paraffin sections were de-paraffinized and rehydrated, before staining with rabbit polyclonal Abs specific for MBP1-10 or Keratin-15 followed by fluorochrome–conjugated secondary Abs. Counterstaining was performed with DAPI. Pictures were generated on a Cell observer microscope (Zeiss).
Flow cytometry Surface antigen staining was performed according to standard procedures. 2.5x106 splenocytes were re-stimulated with 50µg/ml MBP1-10 in the presence of Golgi Plug (BD Bioscience) for 6 h and intracellular cytokine staining was performed with Cytofix/Cytoperm reagents (BD Bioscience). Fluorochrome labeled mAbs were purchased from BD Bioscience. Flow cytometric analyses were done on a FACSCanto (BD Bioscience)
EAE induction EAE was induced by injecting 200 µg N-terminal acetylated MBP1-10 peptide AcASQKRPSQRS (Ac1-10) emulsified in 100µl complete Freund´s adjuvant containing 200 µg Mycobacterium tuberculosis (Thomas Geyer) s.c.. At day 1 and 2 after immunization each mouse received 200ng pertussis toxin i.p. (Merck Biosciences). Clinical symptoms were scored as follows: 0, normal; 1, limp tail or hind limb weakness; 2, limp tail and hind limb weakness; 3, partial hind limp paralysis; 4, complete hind limb paralysis; 5, dead or moribund, killed by investigator.
Contact sensitization Oxazolone (4-ethoxymethylene-2-phenyloxazol-5-one) and DPCP (2,3-Diphenylcycloprop-2en-1-one) were dissolved in acetone and applied to mouse abdominal mouse skin one day after shaving. For single treatment, 100µl of a 3% oxazolone or 2% DPCP (w/v) solution was applied. For repeated DPCP treatment 20µl of a 2% solution, followed by 100µl of a 0.1%,
14
© 2015 The Society for Investigative Dermatology
0.25% or 0.5% (w/v) solution, each one week apart, was applied. Experimental procedures were pursued one week after last application.
Anti-Dkk3 antibody treatment In parallel to the DPCP skin treatments, 0.5 mg of the anti-Dkk3 4.22 mAb (Papatriantafyllou et al., 2012) in 200µl PBS was applied i.p., 3 times, each one week apart.
T cell proliferation in vivo Splenocytes were incubated with 5 µM CFSE (Molecular Probes) in DPBS for 15 min at 37°C. 5x107 Cells were injected into the tail vein and five days later, proliferation of T cells in lymph nodes and skin was analysed by flow cytometry.
EAE transfer 10 days after B10.PL mice were immunized with MBP1-!0 in CFA, spleens were taken out and single cell suspensions using prepared. Splenocytes were cultured for 3 days in RPMI containing MBP1-10 peptide (20µg/ml) and IL-12 (20ng/ml) at 37°C and 5% CO2. 2x107 of these cells were injected i.v. into the respective recipient mice followed by two applications of pertussis toxin i.p. (200ng).
T cell transfer CD4+ T cells were isolated from spleen and lymph nodes of DPCP pre-treated mice using the Dynabeads Untouched CD4+ T cell isolation Kit (Life Technologies) according to manufacturers instructions. 5x105 of theses CD4+ T cells were transferred i.v. into wt mice followed by induction of EAE via MBP1-10/CFA immunization and i.p. pertussis toxin (200ng) application.
15
© 2015 The Society for Investigative Dermatology
qRT-PCR Total RNA was extracted from skin the RNeasy kit (Qiagen). RNA was reverse transcribed using the SuperScript II kit (Invitrogen). Quantitative real-time PCR was performed on an AB 7500 RT-PCR System using Absolute qPCR SYBR Green ROX Mix (Thermo Scientific). The oligonucleotide primer sequences were: Actb-Fp: 5´-TGACAGGATGCAGAAGGAGATTA-3`; Actb-Rp: 5´- AGCCACCGATCCACACAGA-3´; Foxp3-Fp: 5´-CACAACCTGAGCCTGCACAA3´;
Foxp3-Rp:
5´-TCAAATTCATCTACGGTCCACACT-3´;
Tgfb1-Fp:
5´-
TGGAGCAACATGTGGAACTC-3´; Tgfb1-Rp: 5´- CAGCAGCCGGTTACCAAG-3´; Il10-Fp: 5´- CCAGCCTGAGCCTAGAATTCA-3´; Il10-Rp: 5´- TGGCTTCAAACCACACATAGGA-3´; Pdl1-Fp:
5´-
TGGACAAACAGTGACCACCAA-3´;
Pdl1-Rp:
5´-
TGTCCGGGAAGTGGTGACA-3´; Ctla4-Fp: 5´- ACTCATGTACCCACCGCCATA-3´; Ctla4Rp: 5´- TGGTTCTGGATCAATGACATAAAT-3´; Dkk3-Fp: 5´- TCCCATTGCCACCTTTGG-3´; Dkk3-Rp:
5´-
CCAGTTCTCCAGCTTCAAGTACAC-3´;
H2-Aa-Fp:
TCTCCCTCCTGTGATCAACA-3´; H2-Aa-Rp: 5´- CACCGTCTGCGACTGACTTG-3.
5´Data
were calculated relative to the housekeeping gene Actb by using the 2-ΔΔCT method.
ELISA Mouse serum was added to anti-Dkk3 4.22 mAb coated flexible assay plates (Greiner). Biotinylated goat anti-mouse Dkk3 antibody (R&D Systems) was applied as detection antibody followed by streptavidin-peroxidase (Jackson ImmunoResearch). After washing, orthophenylene diamine (OPD) substrate was applied. Reaction was stopped with sulphuric acid. Optical densities were analysed with a Viktor 2 photometer (Perkin Elmer) at 490 nm.
16
© 2015 The Society for Investigative Dermatology
Statistical analysis Two tailed Student’s t-test was used unless indicated otherwise to test significant numerical differences between groups. Differences in EAE progression in between two groups was analyzed using the repeated measures ANOVA test. Dead mice due to EAE were scored 5 until the end of the study. P values of less than 0.05 were regarded as statistically significant (*P <0.05; **P < 0.01; ***P < 0.001).
Conflict of Interest The authors state no conflict of interest.
Acknowledgements We thank G. Hollmann and G. Küblbeck for expert technical assistance. Tg4 mice were generously provided by D. Wraith. This work was supported by the Deutsche Forschungsgemeinschaft (SFB 938) and by the 6th EC-RTD framework program through the RISET Integrated Project.
17
© 2015 The Society for Investigative Dermatology
References Acha-Orbea H, Mitchell DJ, Timmermann L, et al. (1988) Limited heterogeneity of T cell receptors from lymphocytes mediating autoimmune encephalomyelitis allows specific immune intervention. Cell 54:263-73. Alferink J, Aigner S, Reibke R, et al. (1999) Peripheral T-cell tolerance: the contribution of permissive T-cell migration into parenchymal tissues of the neonate. Immunol Rev 169:255-61. Alferink J, Tafuri A, Vestweber D, et al. (1998) Control of neonatal tolerance to tissue antigens by peripheral T cell trafficking. Science 282:1338-41. Allan RS, Waithman J, Bedoui S, et al. (2006) Migratory dendritic cells transfer antigen to a lymph node-resident dendritic cell population for efficient CTL priming. Immunity 25:153-62. Barrantes Idel B, Montero-Pedrazuela A, Guadano-Ferraz A, et al. (2006) Generation and characterization of dickkopf3 mutant mice. Mol Cell Biol 26:2317-26. Bynoe MS, Evans JT, Viret C, et al. (2003) Epicutaneous immunization with autoantigenic peptides induces T suppressor cells that prevent experimental allergic encephalomyelitis. Immunity 19:317-28. Fan L, Busser BW, Lifsted TQ, et al. (2003) Antigen presentation by keratinocytes directs autoimmune skin disease. Proc Natl Acad Sci U S A 100:3386-91. Hermann M, Pirkebner D, Draxl A, et al. (2007) Dickkopf-3 is expressed in a subset of adult human pancreatic beta cells. Histochemistry and cell biology 127:513-21. Hoffmann R, Wenzel E, Huth A, et al. (1994) Cytokine mRNA levels in Alopecia areata before and after treatment with the contact allergen diphenylcyclopropenone. J Invest Dermatol 103:530-3. Hou S, Doherty PC (1993) Partitioning of responder CD8+ T cells in lymph node and lung of mice with Sendai virus pneumonia by LECAM-1 and CD45RB phenotype. J Immunol 150:5494-500. Idoyaga J, Fiorese C, Zbytnuik L, et al. (2013) Specialized role of migratory dendritic cells in peripheral tolerance induction. J Clin Invest 123:844-54. Itano AA, McSorley SJ, Reinhardt RL, et al. (2003) Distinct dendritic cell populations sequentially present antigen to CD4 T cells and stimulate different aspects of cell-mediated immunity. Immunity 19:47-57. Kim BS, Miyagawa F, Cho YH, et al. (2009) Keratinocytes function as accessory cells for presentation of endogenous antigen expressed in the epidermis. J Invest Dermatol 129:2805-17. Kimpton WG, Washington EA, Cahill RN (1995) Virgin alpha beta and gamma delta T cells recirculate extensively through peripheral tissues and skin during normal development of the fetal immune system. Int Immunol 7:1567-77. Kormos B, Belso N, Bebes A, et al. (2011) In vitro dedifferentiation of melanocytes from adult epidermis. PLoS One 6:e17197. Kurschus FC, Oelert T, Liliensiek B, et al. (2006) Experimental autoimmune encephalomyelitis in mice expressing the autoantigen MBP 1-10 covalently bound to the MHC class II molecule I-Au. Int Immunol 18:151-62.
18
© 2015 The Society for Investigative Dermatology
Lara-Corrales I, Pope E (2010) Autoimmune blistering diseases in children. Semin Cutan Med Surg 29:85-91. Liu GY, Fairchild PJ, Smith RM, et al. (1995) Low avidity recognition of self-antigen by T cells permits escape from central tolerance. Immunity 3:407-15. Mackay CR (1993) Homing of naive, memory and effector lymphocytes. Curr Opin Immunol 5:423-7. Matzinger P (2007) Friendly and dangerous signals: is the tissue in control? In: Nat Immunol (11-3. Mayerova D, Parke EA, Bursch LS, et al. (2004) Langerhans cells activate naive self-antigen-specific CD8 T cells in the steady state. Immunity 21:391-400. Nagao K, Kobayashi T, Moro K, et al. (2012) Stress-induced production of chemokines by hair follicles regulates the trafficking of dendritic cells in skin. Nat Immunol 13:744-52. Nickoloff BJ, Turka LA (1994) Immunological functions of non-professional antigen-presenting cells: new insights from studies of T-cell interactions with keratinocytes. Immunol Today 15:464-9. Ohyama M, Terunuma A, Tock CL, et al. (2006) Characterization and isolation of stem cell-enriched human hair follicle bulge cells. J Clin Invest 116:249-60. Okiyama N, Katz SI (2014) Programmed cell death 1 (PD-1) regulates the effector function of CD8 T cells via PD-L1 expressed on target keratinocytes. Journal of autoimmunity 53:1-9. Papatriantafyllou M, Moldenhauer G, Ludwig J, et al. (2012) Dickkopf-3, an immune modulator in peripheral CD8 T-cell tolerance. Proc Natl Acad Sci U S A 109:1631-6. Rosenblum MD, Gratz IK, Paw JS, et al. (2011) Response to self antigen imprints regulatory memory in tissues. Nature 480:538-42. Stoitzner P, Tripp CH, Eberhart A, et al. (2006) Langerhans cells cross-present antigen derived from skin. Proc Natl Acad Sci U S A 103:7783-8. Stringer CP, Hicks R, Botham PA (1991) The expression of MHC class II (Ia) antigens on mouse keratinocytes following epicutaneous application of contact sensitizers and irritants. Br J Dermatol 125:521-8. Tietz W, Allemand Y, Borges E, et al. (1998) CD4+ T cells migrate into inflamed skin only if they express ligands for E- and P-selectin. J Immunol 161:963-70. Wang X, Fujita M, Prado R, et al. (2010) Visualizing CD4 T-cell migration into inflamed skin and its inhibition by CCR4/CCR10 blockades using in vivo imaging model. Br J Dermatol 162:487-96.
19
© 2015 The Society for Investigative Dermatology
Figure Legends Figure 1. MBP-IAu β-chain transgene expression in Ker-MBP mice is skin restricted. (a) MBP-IAu -chain transgenic construct. (b) Schematic illustration of the Ker-MBP Mouse model. (c) Semi-quantitative RT-PCR of the MBP-IAu β-chain in indicated tissues of Ker-MBP and wt mice. (d) Representative immunofluorescence stains for MBP1-10 (green), keratin15 (red) and nuclei (blue) in skin of Ker-MBP and control mice (Bar: 200µm). (e) 5x107 CFSElabelled splenocytes from Tg4 mice were transferred i.v. into Ker-MBP mice. The next day shaved abdomen of was treated with oxazolone (3%) or DPCP (2%) or was left untreated. 5 days later, CFSE profiles of Tg4 CD4+ T cells isolated from draining lymph nodes (LN) and skin were analysed (representative experiment; n.d.=not detected). (f) 5x107 CFSE labelled Tg4 splenocytes were co-transferred with MBP1-10-peptide loaded DCs into Ker-MBP mice. Representative CFSE dilution of Tg4 T cells from LN 5 days after transfer.
Figure 2. Functional MBP-MHC class II complex presentation in Ker-MBP mice only occurs upon contact sensitizer treatment (a) Representative dot plots of CD8 and Vβ8.2 vs. CD4 expression on thymocytes of Tg4xKer-MBP and Tg4 control mice. (b) Representative dot plots of Vβ8.2 vs. CD4 expression and histograms of CD44 and CD62L expression on CD4 T cells from lymph nodes of Tg4xKer-MBP and Tg4 control mice. (c) Tg4xKer-MBP and Tg4 were immunized by MBP1-10/CFA s.c. injection (t0) and EAE symptoms were monitored. Mean clinical score ± SEM is plotted over time. (n=8, n.s.). (d) Representative histology of the skin of Tg4xKerMBP and Tg4 mice 8 days after single application of oxazolone (3%). Infiltration of leukocyte populations was analyzed using hematoxylin/eosin (Bar:100µm), anti-MAC-3 (Bar:100µm) or anti-CD4 (red) staining (Bar: 25µm).
20
© 2015 The Society for Investigative Dermatology
Figure 3. Repeated induction of MBP-MHC class II complex presentation limits severity of subsequent EAE in Ker-MBP mice. (a) EAE was induced in Ker-MBP and wt mice (t0) and symptoms were monitored. Mean clinical score ± SEM is plotted over time (data pooled from two individual experiments, n=12). (b) Skin of Ker-MBP and wt mice was 4 times pre-treated with DPCP (each one week apart). 7 days later, EAE was induced (t0). Mean clinical EAE score ± SEM is plotted over time (data pooled from three individual experiments, n=18). (c) QRT-PCR of H2-Aa in skin of Ker-MBP and wt mice 7 days after a 4 DPCP times pre-treatment. Expression levels are calculated relative to Actb (n=6). (d) 7 days after a 4 times DPCP pre-treatment, 3% Oxazolone was applied to back skin of Ker-MBP and wt mice. One week later, the right ear was challenged by 1% oxazolone and ear thickness was monitored. As a control, ears of untreated wt mice were challenged. Ear thickness ± SEM is plotted over time (data pooled from 2 individual experiments, n=10).
Figure 4. DPCP pre-treatment of Ker-MBP mice induces a dominant form of tolerance. (a) 10 days after MBP1-10/CFA immunization of B10.PL mice, splenocytes were taken out and re-stimulated in vitro for 3 days in the presence of MBP1-10 peptide and IL-12. Subsequently, activated splenocytes were transferred i.v. into Ker-MBP or wt mice (t0) which were 4 times DPCP pre-treated. Mean clinical EAE score ± SEM is plotted over time (n=10). (b) Ker-MBP and wt mice were 4 times DPCP pre-treated. One week later, CD4+ T cells from spleen and lymph nodes were transferred into wt recipients. EAE was subsequently induced (t0) and clinical EAE score was monitored (data pooled from three individual experiments, n=18). (c) Percentages of CD25+FOXP3+CD4+ T cells in spleen and lymph nodes of Ker-MBP and wt mice, 4 times pre-treated with DPCP (n=6). (d) QRT-PCR of Foxp3 in skin of Ker-MBP and wt mice pre-treated like in (c). Expression levels are relative to Actb (n=6). (e) QRT-PCR on RNA isolated from total skin of DPCP pre-treated or untreated Ker-MBP and wt mice. Expression levels of indicated genes are relative to Actb (n=6-12). (f) Dkk3 concentration in serum of untreated and 4 times DPCP pre-treated Ker-MBP and wt mice, measured by
21
© 2015 The Society for Investigative Dermatology
ELISA (n=4). (g) QRT-PCR of Dkk3 in CD4+ T cells isolated from untreated and 4 times DPCP pre-treated Ker-MBP and wt mice. As negative control, CD4+ T cells isolated from untreated Dkk3-/- mice were used. As positive control, total skin of 4 times DPCP pre-treated mice was used (n=3).
Figure 5. Reduction of the MBP specific CD4+ T cell response in Ker-MBP mice depends on Dickkopf-3. (a) Ker-MBPxDkk3-/-, Dkk3-/-, Ker-MBP and wt mice were 4 times DPCP pre-treated before induction of EAE (t0). Mean clinical score ± SEM is plotted over time (data pooled from 2 individual experiments, n=12). (b) Ker-MBP and wt mice were 4 times pre-treated. In both groups, 500µg of anti-Dkk3 mAb clone 4.22 was applied i.p. along with the first three DPCP applications. One week after 4th application EAE was induced (t0) and symptoms were monitored. Mean clinical score ± SEM is plotted over time (data pooled from 2 individual experiments, n=16). (c) QRT-PCR of H2-Aa in skin of Ker-MBP.Dkk3-/- and Dkk3-/- mice 7 days after a 4 times DPCP pre-treatment. Expression levels were calculated relative to Actb (n=6). (d) Ker-MBP and Ker-MBP.Dkk3-/- mice were 4 times DPCP pre-treated or left untreated and EAE was induced. 7 days later splenocytes were in vitro re-stimulated with MBP1-10 peptide and analyzed by flow cytometry. Depicted are the percentages of IL-2, TNF and IL-17 producing CD4+ T cells (data pooled from three individual experiments, n=6).
22
© 2015 The Society for Investigative Dermatology
© 2015 The Society for Investigative Dermatology
© 2015 The Society for Investigative Dermatology
© 2015 The Society for Investigative Der
© 2015 The Society for Investigative Dermatology
© 2015 The Society for Investigative Dermatology
Self-Antigen Presentation by Keratinocytes in the Inflamed Adult Skin Modulates T Cell Auto-Reactivity Michael Meister, Amel Tounsi, Evelyn Gaffal, Tobias Bald, Maria Papatriantafyllou, Julia Ludwig, Georg Pougialis, Felix Bestvater, Luisa Klotz, Gerhard Moldenhauer, Thomas Tu¨ting, Gu¨nter J Ha¨mmerling, Bernd Arnold, Thilo Oelert
Cite this article as: Michael Meister, Amel Tounsi, Evelyn Gaffal, Tobias Bald, Maria Papatriantafyllou, Julia Ludwig, Georg Pougialis, Felix Bestvater, Luisa Klotz, Gerhard Moldenhauer, Thomas Tu¨ting, Gu¨nter J Ha¨mmerling, Bernd Arnold, Thilo Oelert, Self-Antigen Presentation by Keratinocytes in the Inflamed Adult Skin Modulates T Cell Auto-Reactivity, Journal of Investigative Dermatology accepted article preview 2 April 2015; doi: 10.1038/jid.2015.130. This is a PDF file of an unedited peer-reviewed manuscript that has been accepted for publication. NPG are providing this early version of the manuscript as a service to our customers. The manuscript will undergo copyediting, typesetting and a proof review before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers apply.
Received 15 October 2014; revised 16 March 2015; accepted 19 March 2015; Accepted article preview online 2 April 2015
© 2015 The Society for Investigative Dermatology
Self-antigen presentation by keratinocytes in the inflamed adult skin modulates T cell auto-reactivity
Michael Meister1, Amel Tounsi1, Evelyn Gaffal3, Tobias Bald3, Maria Papatriantafyllou1, Julia Ludwig1, Georg Pougialis1, Felix Bestvater2, Luisa Klotz4, Gerhard Moldenhauer1, Thomas Tüting3, Günter J. Hämmerling1, Bernd Arnold1, Thilo Oelert1
1
Department of Molecular Immunology and
2
Core Facility Light Microscopy, German Cancer
Research Center (DKFZ), 69120 Heidelberg, Germany 3
Laboratory of Experimental Dermatology, Department of Dermatology and Allergy, University of
Bonn, 53105 Bonn, Germany 4
Department of Neurology, Clinic for Neurology, University Hospital Münster, 48149 Münster,
Germany
Address correspondence to: Bernd Arnold, Department of Molecular Immunology, German Cancer Research Center, 69120 Heidelberg, Germany. Phone: 49.6221.42.3728; Fax: 49.6221.42.3761; Email: [email protected]
Short title: Self-antigen presentation by keratinocytes
1
© 2015 The Society for Investigative Dermatology
Abstract Keratinocytes play a pivotal role in the regulation of immune responses but the impact of antigen-presentation by these cells is still poorly understood, particularly in a situation where the antigen will be presented only in adult life. Here, we generated a transgenic mouse model in which keratinocytes exclusively present a myelin basic protein (MBP) peptide covalently linked to the MHC class II β-chain, solely under inflammatory conditions. In these mice, inflammation caused by epicutaneous contact sensitizer treatment resulted in keratinocyte-mediated expansion of MBP-specific CD4+ T cells in the skin. Moreover, repeated contact sensitizer application preceding a systemic MBP immunization reduced the reactivity of the respective CD4+ T cells and lowered the symptoms of the resulting experimental autoimmune encephalomyelitis. This down-regulation was CD4+ T cell mediated and dependent on the presence of the immune modulator Dickkopf-3. Thus, presentation of a neo self-antigen by keratinocytes in the inflamed, adult skin can modulate CD4+ T cell autoaggression at a distal organ.
2
© 2015 The Society for Investigative Dermatology
Introduction The observation that the type and strength of immune responses depend on the local microenvironment of a given tissue (Alferink et al., 1999; Matzinger, 2007) has fostered investigations on the contribution of various tissue-specific components to the resulting immune response. In the skin the prominent role of Langerhans cells (Mayerova et al., 2004; Nagao et al., 2012; Stoitzner et al., 2006) and dermal dendritic cells (Allan et al., 2006; Idoyaga et al., 2013; Itano et al., 2003) in inducing antigen specific immunity or tolerance is well established. Less is known about the role of epidermal keratinocytes in the activation or suppression of T cell effector functions. For example, keratinocyte specific expression of a nominal antigen, such as Ovalbumin (OVA) led to proliferation of OVA specific CD4+ T cells and was able to cause symptoms of an inflammatory dermatitis (Rosenblum et al., 2011). However, in this transgenic system, OVA-derived peptides were also presented by professional antigen presenting cells (APC) in the skin draining lymph nodes, which prevented to identify the contribution of antigen presentation by keratinocytes to the resulting T cell response. In a similar study, mice expressing a model antigen in keratinocytes developed a graft versus host like reaction after intradermal injection of CD8+ T cells carrying the respective specificity (Kim et al., 2009). In this case, antigen presentation by Langerhans cells and dendritic cells was experimentally excluded. Thus, it was proposed that keratinocytes can directly prime naïve CD8+ T cells in the adult skin (Kim et al., 2009). We have previously shown that CD8+ T cells with a transgenic MHC class I Kb-specific T cell receptor (TCR) become tolerant following encounter of the Kb antigen on keratinocytes (Alferink et al., 1998). As in this animal model the respective TCR recognized only the intact MHC molecule, it was proposed that T cell stimulation occurred through direct contact with keratinocytes. However, this contact mainly occurs during the neonatal phase when extensive trafficking of T cells through tissues including the skin is observed (Kimpton et al., 1995). In contrast, T cell migration into the healthy, adult skin is restricted (Mackay, 1993)
3
© 2015 The Society for Investigative Dermatology
and T cells can only infiltrate inflammatory sites (Hou and Doherty, 1993; Tietz et al., 1998; Wang et al., 2010). In parallel it was also reported that CD4+ T cells can be directly activated by keratinocytes specifically presenting peptide–MHC class II conjugates. Inflammatory skin disease was observed in 20-40% of mice expressing such a transgenic MHC class II I-Ab β-chain under the keratin14 promotor and a corresponding auto-reactive TCR on CD4+ T cells. In this study, keratinocytes were shown to activate and are the primary targets of auto-reactive CD4+ T cells (Fan et al., 2003). Here, we investigated whether keratinocytes that present a sessile self-antigen in the inflamed, adult skin can prime CD4+ T cells in the absence of antigen presentation by professional APCs and whether this activation can modulate a distant, auto-reactive T cell response. Therefore, we generated a transgenic mouse model, in which keratinocytes express a silent antigenic determinant, namely an encephalitogenic myelin basic protein (MBP1-10) peptide, covalently linked to the MHC class II IAu β-chain (Ker-MBP mice). Cell surface presentation of this transgene product requires co-expression of the endogenous MHC II α-chain, which needs to be induced by inflammatory stimuli like for example Interferon-γ (Nickoloff and Turka, 1994). This allowed us to study the contribution of keratinocytes to T cell activation in the inflamed skin and to the associated auto-reactivity at a distal site in the context of experimental autoimmune encephalomyelitis (EAE). In the Ker-MBP mouse model, transferred MBP-specific Tg4 CD4+ T cells proliferated only in sensitized skin but not in the draining lymph nodes. Repeated skin sensitization induced regulatory T cells, which limited MBP-specific CD4+ T cell reactivity and lowered EAE symptoms after MBP immunization. Moreover, the recently described immune modulator Dickkopf-3 (DKK3) (Papatriantafyllou et al., 2012) was found to contribute to the downmodulation of MBP-specific CD4+ T cell reactivity. Together, these findings suggest that selfantigen presentation by keratinocytes under inflammatory conditions can modulate autoreactive CD4+ T cell responses at a distal site.
4
© 2015 The Society for Investigative Dermatology
Results Generation and characterization of transgenic Ker-MBP mice To generate a mouse model for conditional induction of a self-antigen in the inflamed skin of adult mice, we fused a cDNA encoding the MBP1-10 peptide and a glycine-serine linker to the genomic DNA of the I-Au β-chain. The central nervous system (CNS) derived peptide MBP1-10 was chosen as a model antigen because auto-antigens of the brain can also be found in the skin under physiological conditions (Kormos et al., 2011). This construct was put under the control of a keratinIV promoter element that is known to direct expression exclusively to keratinocytes of hair follicles (Alferink et al., 1998) (Ker-MBP, Figure 1a). The transgenic MBP-IAu -chain is constitutively expressed in resting keratinocytes. However, cell surface presentation of the transgene requires co-expression of the endogenous MHC class II chain which is absent in keratinocytes in the healthy skin but induced by inflammatory stimuli like IFN-γ (Nickoloff and Turka, 1994) (Figure 1b). In order to ensure keratinocyte restricted MBP-IAu MHC class II -chain production, we initially confirmed tissue-specific expression of the transgene by RT-PCR using RNA extracted from the respective tissues (Figure 1c). Moreover,
MBP-IAu
-chain
expression
in
keratinocytes
was
assessed
by
immunofluorescence staining in the skin with a MBP1-10-specific rabbit anti-serum (Figure 1d). To clarify whether surface presentation of the MBP-MHC II complex on keratinocytes is inducible and could be recognized by T cells, TCR-transgenic MBP-specific Tg4 T cells were labelled with CFSE and transferred into Ker-MBP mice one day before application of the contact sensitizers oxazolone (OXA) or diphenylcycloprenone (DPCP). In particular, contact sensitizers were used to induce inflammation in the skin, as they were shown to be potent inducers of MHC class II expression in keratinocytes (Stringer et al., 1991). 5 days after transfer, T cells were isolated from draining lymph nodes and treated skin and analysed by flow cytometry. Proliferating Tg4 T cells could be observed in the inflamed skin of Ker-MBP mice, but were undetectable in draining lymph nodes (Figure 1e). As a control, untreated Ker-MBP mice were analysed. There, transferred Tg4 cells did not proliferate in lymph nodes
5
© 2015 The Society for Investigative Dermatology
and were not detectable in skin of these mice (Figure 1e, lower panel). However, strong proliferation was observed in draining lymph nodes when MBP-peptide loaded DCs were coinjected with the transferred Tg4 T cells (Figure 1f). Thus, these results reveal that the KerMBP mouse model allows us to the study the consequences of keratinocyte restricted selfantigen presentation in the skin, in the absence of antigen presentation by professional APC in the draining lymph nodes.
Functional presentation of the transgenic MBP-IAu MHC II ß-chain in the skin of Ker-MBP mice only occurs upon contact sensitizer treatment. In order to increase the low endogenous level of MBP-specific CD4+ T cells found in the polyclonal TCR repertoire, we crossed Ker-MBP mice to the MBP-specific TCR transgenic Tg4 mice (Liu et al., 1995). This allowed us to study the functionality of the MBP-MHC class II complex presentation by keratinocytes in vivo in more detail. Furthermore, it enabled us to exclude accidental thymic or peripheral presentation of the transgene, which might induce tolerance against the MBP antigen (Alferink et al., 1998; Kurschus et al., 2006). Analysis of the thymus revealed comparable proportions of thymocyte subsets in Tg4 and Tg4xKer-MBP mice (Figure 2a) indicating a normal T cell development in both mouse strains. Furthermore, CD4+ T cell proportions as well as their memory differentiation status, identified by surface expression of CD62L and CD44, were similar in in Tg4 and Tg4xKer-MBP mice (Figure 2b), suggesting that there is no induction of peripheral tolerance. To asses this on a functional level, Ker-MBP positive and negative Tg4 mice were challenged with MBP peptide in CFA. Both types of mice showed similar levels of EAE (Figure 2c) revealing that there is no influence of the Ker-MBP transgene on the functional readout system of our mouse model per se. Importantly, contact sensitizer treatment resulted in the accumulation of CD4+ T cells in the skin of Tg4xKer-MBP mice but not in the skin of Tg4 mice, 8 days after one single application. However, this was not accompanied by severe skin disease like scaling or
6
© 2015 The Society for Investigative Dermatology
alopecia. As well, we did not observe interface dermatitis or an eczematous reaction in the double transgenic mice (Figure 2d).
Repeated skin sensitization limits EAE symptoms in Ker-MBP mice. Next, we asked whether treatment of Ker-MBP mice with contact sensitizers would have a functional impact on a following, active immunization with MBP. Untreated Ker-MBP and wt mice developed comparable EAE symptoms upon MBP1-10 immunization (Figure 3a). As repeated exposure of T cells to self-antigens in tissues has been reported to attenuate the severity of autoimmune responses (Lara-Corrales and Pope, 2010; Rosenblum et al., 2011), we decided to perform repetitive skin treatments with a contact sensitizer before EAE induction. And indeed, 4 repetitive pre-treatments with DPCP reduced the severity of EAE symptoms in Ker-MBP mice compared to littermate controls (Figure 3b). However, this difference was apparently not due to a generally impaired expression of the endogenous MHC class II α-chain in the skin of Ker-MBP mice (Figure 3c). Reduction of the anti-MBP T cell response in Ker-MBP mice was not based on a general DPCP-induced immune suppression in the skin, as pre-treatment did not affect a delayed-type hypersensitivity response to another contact sensitizer, oxazolone (Figure 3d). Thus, DPCP-mediated chronic inflammation in the skin of Ker-MBP mice results in continuous cell surface presentation of the MBP peptide on keratinocytes, which in turn down-modulates a following MBP-specific T cell response in the brain.
CD4+ T cells mediate the down-modulation of T cell reactivity in Ker-MBP mice To clarify the mechanisms that limit the MBP-induced autoimmunity shown in Figure 3b, it was initially investigated whether this was associated with a dominant form of immuneregulation. EAE induced by the transfer of MBP-activated splenocytes resulted in significantly less severe EAE symptoms in DPCP pre-treated Ker-MBP mice compared to non-transgenic
7
© 2015 The Society for Investigative Dermatology
littermates (Figure 4a). Furthermore, CD4+ T cells were isolated from spleens of DPCP pretreated wt and Ker-MBP mice and transferred into wt mice. Upon MBP immunization, lower EAE scores were observed in recipients of T cells isolated from pre-treated Ker-MBP mice compared to animals that received T cells from pre-treated non-transgenic littermates (Figure 4b). These findings suggest that the down-modulation of MBP-specific T cell reactivity in pretreated Ker-MBP mice was mediated by regulatory T cells. However, we could neither observe changes in the frequency of CD4+CD25+Foxp3+ regulatory T cells in the lymph nodes and spleens of DPCP treated Ker-MBP mice nor could detect altered expression of forkhead box P3 (FOXP3) in skin of these mice (Figure 5c, d). Next, we analysed the expression levels of several immune modulators like programmed cell death 1 ligand (PD-L1) that was shown to be an important regulator of CD8+ T cell effector function in a similar mouse model (Okiyama and Katz, 2014). However, no differential expression of interleukin10 (IL-10), transforming growth factor β (TGFβ), cytotoxic T lymphocyte associated antigen 4 (CTLA4) and PD-L1 could be detected in untreated or DPCP treated skin between wt and Ker-MBP groups (Figure 4e). We have previously reported that the secreted protein Dickkopf-3 (DKK3), which is also expressed in hair follicles (Ohyama et al., 2006), can limit CD8+ T cell reactivity in vitro and in vivo (Papatriantafyllou et al., 2012). Analysis of DKK3 expression revealed a significant up-regulation in DPCP pre-treated skin of Ker-MBP mice compared to non-transgenic littermates (Figure 4e). However, this up-regulation was not accompanied by an increase of the systemic DKK3 level in the serum (Figure 4f). Moreover, no DKK3 expression in CD4+ T cells isolated from spleen and lymph nodes of untreated and DPCP pre-treated Ker-MBP and wt mice (Figure 4g).
Limitation of T cell reactivity in Ker-MBP mice depends on the presence of Dickkopf-3. In order to clarify whether DKK3 contributes to the limitation of CD4+ T cell reactivity in pretreated transgenic animals, Ker-MBP mice deficient for DKK3 were established and fourtimes pre-treated with DPCP. In contrast to DKK3-competent mice, pre-treatment did not result in a reduction of EAE symptoms in DKK3 deficient Ker-MBP mice in comparison to
8
© 2015 The Society for Investigative Dermatology
non-transgenic Dkk3-/- littermates (Figure 5a). Additionally, we blocked DKK3 function in KerMBP and wt mice by administration of a neutralizing DKK3-specific antibody at the beginning of the DPCP treatment. Similar to the DKK3-deficient situation, we did not observe a reduction of MBP-triggered EAE as a result of DPCP pre-treatment in Ker-MBP mice compared to wt controls (Figure 5b). Absence of EAE down-modulation was again not due to altered endogenous MHC class II expression (Figure 5c). Finally, we investigated whether the observed changes in EAE severity were linked with the effector function of the respective MBP-specific CD4+ T cells. Wild-type and Ker-MBP mice were immunized with MBP1-10 and seven days later splenic T cells were re-stimulated in vitro with the respective peptide. The percentages of MBP-specific CD4+ T cells producing interleukin-2 (IL-2), tumour necrosis factor (TNF) and interleukin-17 (IL-17) were comparable in the two experimental groups (Figure 5d). However, the percentages of IL-2, TNF and IL-17 producing MBP-specific T cells were significantly reduced among in vitro re-stimulated CD4+ T cells that had been isolated from Ker-MBP mice following four-times DPCP treatment. By contrast, the effector function of MBP-specific CD4+ T cells from DPCP-treated KerMBP.Dkk3-/- mice was not compromised in vitro compared to untreated controls (Figure 5d). Thus, chronic skin inflammation, as modelled by repeated DPCP treatment and the associated priming of MBP-specific CD4+ T cells by keratinocytes in the skin can reduce systemic CD4+ T cell reactivity in an antigen-dependent manner through a mechanism that involves DKK3 expression.
9
© 2015 The Society for Investigative Dermatology
Discussion The skin provides a first line of defence against physical insults, toxins and pathogens. Therefore, immune surveillance of this barrier has to be finely balanced to ensure its integrity. Here, we asked how T cells respond to cognate auto-antigens that are presented exclusively by keratinocytes of the inflamed skin, and whether this presentation can influence T cell reactivity against the respective antigen in a distant organ. Our data suggest that (i) keratinocytes can present self-antigen to stimulate CD4+ T cells under inflammatory conditions; (ii) repeated antigen stimulation in the inflamed skin leads to down-modulation of T cell responses; (iii) this down-modulation is CD4+ T cell mediated and depends on the expression of Dickkopf-3. In our Ker-MBP mouse model, a MBP peptide that is covalently linked to the MHC class II A u β-chain is exclusively and constitutively expressed in keratinocytes of hair follicles. In naïve mice carrying the Ker-MBP transgene, MBP-specific CD4+ T cells do not seem to be aware of their cognate antigen, suggesting that the MBP-peptide is not presented under steady state conditions. Upon skin sensitization, MBP presentation on the cell surface of keratinocytes is enabled. If professional APCs would take up the MBP peptide in this situation one would expect MBPspecific T cell proliferation in the draining lymph nodes after migration of CD11c+ cells was induced by contact sensitizers. However, proliferation of MBP-specific T cells was only observed in the skin and was absent from draining lymph nodes. These observations support the view that in the Ker-MBP mouse model the MBP1-10 peptide represents a truly sessile antigen in the skin that can be presented exclusively by keratinocytes and only upon inflammation. The cutaneous inflammation that was induced by a single treatment with the contact sensitizer oxazolone was more potent in Tg4xKer-MBP mice than in Tg4 mice. However, severe skin disease, such as scaling and alopecia was not observed in the double transgenic mice. This is in contrast to what was observed in a transgenic mouse model where a nominal antigen, such as Ovalbumin (OVA), was inducibly expressed in keratinocytes (Rosenblum et 10
© 2015 The Society for Investigative Dermatology
al., 2011). In these mice, OVA-specific T cells caused pronounced inflammatory dermatitis that spontaneously resolved after 20-30 days (Rosenblum et al., 2011). The different results obtained in the two models might reflect the fact that OVA can be presented to DO11 T cells by professional APCs in the Ker-OVA mouse model, whereas MBP is presented to Tg4 T cells exclusively by keratinocytes in the Ker-MBP system. There is increasing evidence that repeated exposure of T cells to self-antigens in tissues attenuates the severity of autoimmune responses (Lara-Corrales and Pope, 2010; Rosenblum et al., 2011). Accordingly, in our mouse model repeated applications of DPCP that allowed for continuous availability of the MBP-MHC class II complex on keratinocytes limited the severity of subsequent EAE. Interestingly, DPCP is used clinically for the treatment of alopecia areata because of its proposed immune-modulatory capacity (Hoffmann et al., 1994). It was reported that epicutaneous immunization with the auto-antigenic MBP peptide was able to inhibit the induction of EAE (Bynoe et al., 2003). This protection was based on CD4+ T cell regulation, but was independent of CD4+CD25+ regulatory T cells and was not mediated by immune modulators such as IL-4, IL-10 or TGFβ. In this study, the MBP peptide was shown to be presented by professional APCs in the draining lymph nodes. It is interesting that repeated tolerogenic epicutaneous applications of MBP peptide have similar consequences as continuous MBP peptide presentation by keratinocytes under inflammatory conditions in our mouse model. DKK3 was found to be significantly up-regulated in the skin of DPCP-treated Ker-MBP mice compared to littermate controls. We have recently reported that DKK3 can limit CD8+ T cell mediated tumor and skin graft rejection (Papatriantafyllou et al., 2012). DKK3 is mainly expressed at sites that have been traditionally referred to as immune privileged including the hair follicle (Barrantes Idel et al., 2006; Hermann et al., 2007; Ohyama et al., 2006). It is therefore, tempting to speculate that DKK3 may contribute to the immunosuppressive milieu of such tissues. Here we show that abrogation of DKK3 function was sufficient to reverse the reduction of T cell-mediated EAE symptoms in DPCP-treated Ker-MBP mice. Furthermore,
11
© 2015 The Society for Investigative Dermatology
expression of DKK3 in the skin was associated with decreased cytokine production by MBPspecific CD4+T cells. Whether DKK3 acts directly or indirectly on CD4+T cells in this in vivo setting is presently unknown. Neutralizing DKK3-specific antibodies could block the downmodulation of CD4+T cell reactivity when administered at the beginning of the DPCP treatment. This finding suggests that DKK3 is involved in the induction of the observed T cell regulation rather than directly suppressing CD4+T cell reactivity in the skin. Further investigations are needed to clarify the mechanism through which DKK3 limits CD4+T cell reactivity. In summary, based on our newly established mouse model we present evidence that the induction and presentation of a self-antigen by keratinocytes during skin inflammation is sufficient to modulate T cell auto-reactivity at a distal site. Keratinocytes could activate MBPspecific T cells in the skin independent from professional APCs. Furthermore, persistent stimulation of MBP-specific CD4+ T cells by keratinocytes in the chronically inflamed skin led to a reduction of EAE symptoms. Thus, repeated antigen-presentation by keratinocytes during skin inflammation may dampen the respective T cell reactivity at a distal site.
12
© 2015 The Society for Investigative Dermatology
Material and Methods Mice B10.PL mice were obtained from Jackson Laboratories. TCR-transgenic Tg4 mice have been described (Acha-Orbea et al., 1988). Dkk3 -/- mice (Barrantes Idel et al., 2006) were crossed to a H-2KU background (B10.PL). Mice were maintained under specific pathogen free conditions at the German Cancer Research Center. All animal experiments were approved by the local government authorities (Regierungspräsidium Karlsruhe, Germany).
Generation of transgenic Ker-MBP mice For keratinocyte-specific expression the 2.4 kb fragment of the Keratin IV promoter was used (Alferink et al., 1998). The introduction of the signal sequence and the MBP1-10-peptide (ASQKRPSQRS) into the I-Au β-chain cDNA has been described (Kurschus et al., 2006). To assure appropriate cleavage of the signal sequence the amino acid sequence GDS was introduced N-terminal to the MBP-sequence. As a substitute of natural acetylation a glycine residue was inserted between the GDS sequence and the MBP1-10 sequence. We replaced the cDNA after exon 2 with the genomic DNA of the I-Ab β-chain (obtained from RZPD; identical with the β-chain I-Au coding sequence). The DNA construct was linearized and injected into C57BL/6 x DBA/2 F2 oocytes. Transgenic mice were backcrossed more than 10 times to B10.PL and were crossed to Tg4 mice.
Histology Skin tissue was fixed in a zinc-based fixative (Dako), embedded in paraffin, and routinely stained with H&E. Furthermore, sections were stained with rat anti-mouse CD4 or rat antimouse MAC-3 followed by a biotinylated rabbit anti-rat IgG and development with the LSABAP kit (Dako).
13
© 2015 The Society for Investigative Dermatology
Immunofluorescence Skin tissue was fixed with phosphate-buffered 4% formaldehyde. Five-micrometer paraffin sections were de-paraffinized and rehydrated, before staining with rabbit polyclonal Abs specific for MBP1-10 or Keratin-15 followed by fluorochrome–conjugated secondary Abs. Counterstaining was performed with DAPI. Pictures were generated on a Cell observer microscope (Zeiss).
Flow cytometry Surface antigen staining was performed according to standard procedures. 2.5x106 splenocytes were re-stimulated with 50µg/ml MBP1-10 in the presence of Golgi Plug (BD Bioscience) for 6 h and intracellular cytokine staining was performed with Cytofix/Cytoperm reagents (BD Bioscience). Fluorochrome labeled mAbs were purchased from BD Bioscience. Flow cytometric analyses were done on a FACSCanto (BD Bioscience)
EAE induction EAE was induced by injecting 200 µg N-terminal acetylated MBP1-10 peptide AcASQKRPSQRS (Ac1-10) emulsified in 100µl complete Freund´s adjuvant containing 200 µg Mycobacterium tuberculosis (Thomas Geyer) s.c.. At day 1 and 2 after immunization each mouse received 200ng pertussis toxin i.p. (Merck Biosciences). Clinical symptoms were scored as follows: 0, normal; 1, limp tail or hind limb weakness; 2, limp tail and hind limb weakness; 3, partial hind limp paralysis; 4, complete hind limb paralysis; 5, dead or moribund, killed by investigator.
Contact sensitization Oxazolone (4-ethoxymethylene-2-phenyloxazol-5-one) and DPCP (2,3-Diphenylcycloprop-2en-1-one) were dissolved in acetone and applied to mouse abdominal mouse skin one day after shaving. For single treatment, 100µl of a 3% oxazolone or 2% DPCP (w/v) solution was applied. For repeated DPCP treatment 20µl of a 2% solution, followed by 100µl of a 0.1%,
14
© 2015 The Society for Investigative Dermatology
0.25% or 0.5% (w/v) solution, each one week apart, was applied. Experimental procedures were pursued one week after last application.
Anti-Dkk3 antibody treatment In parallel to the DPCP skin treatments, 0.5 mg of the anti-Dkk3 4.22 mAb (Papatriantafyllou et al., 2012) in 200µl PBS was applied i.p., 3 times, each one week apart.
T cell proliferation in vivo Splenocytes were incubated with 5 µM CFSE (Molecular Probes) in DPBS for 15 min at 37°C. 5x107 Cells were injected into the tail vein and five days later, proliferation of T cells in lymph nodes and skin was analysed by flow cytometry.
EAE transfer 10 days after B10.PL mice were immunized with MBP1-!0 in CFA, spleens were taken out and single cell suspensions using prepared. Splenocytes were cultured for 3 days in RPMI containing MBP1-10 peptide (20µg/ml) and IL-12 (20ng/ml) at 37°C and 5% CO2. 2x107 of these cells were injected i.v. into the respective recipient mice followed by two applications of pertussis toxin i.p. (200ng).
T cell transfer CD4+ T cells were isolated from spleen and lymph nodes of DPCP pre-treated mice using the Dynabeads Untouched CD4+ T cell isolation Kit (Life Technologies) according to manufacturers instructions. 5x105 of theses CD4+ T cells were transferred i.v. into wt mice followed by induction of EAE via MBP1-10/CFA immunization and i.p. pertussis toxin (200ng) application.
15
© 2015 The Society for Investigative Dermatology
qRT-PCR Total RNA was extracted from skin the RNeasy kit (Qiagen). RNA was reverse transcribed using the SuperScript II kit (Invitrogen). Quantitative real-time PCR was performed on an AB 7500 RT-PCR System using Absolute qPCR SYBR Green ROX Mix (Thermo Scientific). The oligonucleotide primer sequences were: Actb-Fp: 5´-TGACAGGATGCAGAAGGAGATTA-3`; Actb-Rp: 5´- AGCCACCGATCCACACAGA-3´; Foxp3-Fp: 5´-CACAACCTGAGCCTGCACAA3´;
Foxp3-Rp:
5´-TCAAATTCATCTACGGTCCACACT-3´;
Tgfb1-Fp:
5´-
TGGAGCAACATGTGGAACTC-3´; Tgfb1-Rp: 5´- CAGCAGCCGGTTACCAAG-3´; Il10-Fp: 5´- CCAGCCTGAGCCTAGAATTCA-3´; Il10-Rp: 5´- TGGCTTCAAACCACACATAGGA-3´; Pdl1-Fp:
5´-
TGGACAAACAGTGACCACCAA-3´;
Pdl1-Rp:
5´-
TGTCCGGGAAGTGGTGACA-3´; Ctla4-Fp: 5´- ACTCATGTACCCACCGCCATA-3´; Ctla4Rp: 5´- TGGTTCTGGATCAATGACATAAAT-3´; Dkk3-Fp: 5´- TCCCATTGCCACCTTTGG-3´; Dkk3-Rp:
5´-
CCAGTTCTCCAGCTTCAAGTACAC-3´;
H2-Aa-Fp:
TCTCCCTCCTGTGATCAACA-3´; H2-Aa-Rp: 5´- CACCGTCTGCGACTGACTTG-3.
5´Data
were calculated relative to the housekeeping gene Actb by using the 2-ΔΔCT method.
ELISA Mouse serum was added to anti-Dkk3 4.22 mAb coated flexible assay plates (Greiner). Biotinylated goat anti-mouse Dkk3 antibody (R&D Systems) was applied as detection antibody followed by streptavidin-peroxidase (Jackson ImmunoResearch). After washing, orthophenylene diamine (OPD) substrate was applied. Reaction was stopped with sulphuric acid. Optical densities were analysed with a Viktor 2 photometer (Perkin Elmer) at 490 nm.
16
© 2015 The Society for Investigative Dermatology
Statistical analysis Two tailed Student’s t-test was used unless indicated otherwise to test significant numerical differences between groups. Differences in EAE progression in between two groups was analyzed using the repeated measures ANOVA test. Dead mice due to EAE were scored 5 until the end of the study. P values of less than 0.05 were regarded as statistically significant (*P <0.05; **P < 0.01; ***P < 0.001).
Conflict of Interest The authors state no conflict of interest.
Acknowledgements We thank G. Hollmann and G. Küblbeck for expert technical assistance. Tg4 mice were generously provided by D. Wraith. This work was supported by the Deutsche Forschungsgemeinschaft (SFB 938) and by the 6th EC-RTD framework program through the RISET Integrated Project.
17
© 2015 The Society for Investigative Dermatology
References Acha-Orbea H, Mitchell DJ, Timmermann L, et al. (1988) Limited heterogeneity of T cell receptors from lymphocytes mediating autoimmune encephalomyelitis allows specific immune intervention. Cell 54:263-73. Alferink J, Aigner S, Reibke R, et al. (1999) Peripheral T-cell tolerance: the contribution of permissive T-cell migration into parenchymal tissues of the neonate. Immunol Rev 169:255-61. Alferink J, Tafuri A, Vestweber D, et al. (1998) Control of neonatal tolerance to tissue antigens by peripheral T cell trafficking. Science 282:1338-41. Allan RS, Waithman J, Bedoui S, et al. (2006) Migratory dendritic cells transfer antigen to a lymph node-resident dendritic cell population for efficient CTL priming. Immunity 25:153-62. Barrantes Idel B, Montero-Pedrazuela A, Guadano-Ferraz A, et al. (2006) Generation and characterization of dickkopf3 mutant mice. Mol Cell Biol 26:2317-26. Bynoe MS, Evans JT, Viret C, et al. (2003) Epicutaneous immunization with autoantigenic peptides induces T suppressor cells that prevent experimental allergic encephalomyelitis. Immunity 19:317-28. Fan L, Busser BW, Lifsted TQ, et al. (2003) Antigen presentation by keratinocytes directs autoimmune skin disease. Proc Natl Acad Sci U S A 100:3386-91. Hermann M, Pirkebner D, Draxl A, et al. (2007) Dickkopf-3 is expressed in a subset of adult human pancreatic beta cells. Histochemistry and cell biology 127:513-21. Hoffmann R, Wenzel E, Huth A, et al. (1994) Cytokine mRNA levels in Alopecia areata before and after treatment with the contact allergen diphenylcyclopropenone. J Invest Dermatol 103:530-3. Hou S, Doherty PC (1993) Partitioning of responder CD8+ T cells in lymph node and lung of mice with Sendai virus pneumonia by LECAM-1 and CD45RB phenotype. J Immunol 150:5494-500. Idoyaga J, Fiorese C, Zbytnuik L, et al. (2013) Specialized role of migratory dendritic cells in peripheral tolerance induction. J Clin Invest 123:844-54. Itano AA, McSorley SJ, Reinhardt RL, et al. (2003) Distinct dendritic cell populations sequentially present antigen to CD4 T cells and stimulate different aspects of cell-mediated immunity. Immunity 19:47-57. Kim BS, Miyagawa F, Cho YH, et al. (2009) Keratinocytes function as accessory cells for presentation of endogenous antigen expressed in the epidermis. J Invest Dermatol 129:2805-17. Kimpton WG, Washington EA, Cahill RN (1995) Virgin alpha beta and gamma delta T cells recirculate extensively through peripheral tissues and skin during normal development of the fetal immune system. Int Immunol 7:1567-77. Kormos B, Belso N, Bebes A, et al. (2011) In vitro dedifferentiation of melanocytes from adult epidermis. PLoS One 6:e17197. Kurschus FC, Oelert T, Liliensiek B, et al. (2006) Experimental autoimmune encephalomyelitis in mice expressing the autoantigen MBP 1-10 covalently bound to the MHC class II molecule I-Au. Int Immunol 18:151-62.
18
© 2015 The Society for Investigative Dermatology
Lara-Corrales I, Pope E (2010) Autoimmune blistering diseases in children. Semin Cutan Med Surg 29:85-91. Liu GY, Fairchild PJ, Smith RM, et al. (1995) Low avidity recognition of self-antigen by T cells permits escape from central tolerance. Immunity 3:407-15. Mackay CR (1993) Homing of naive, memory and effector lymphocytes. Curr Opin Immunol 5:423-7. Matzinger P (2007) Friendly and dangerous signals: is the tissue in control? In: Nat Immunol (11-3. Mayerova D, Parke EA, Bursch LS, et al. (2004) Langerhans cells activate naive self-antigen-specific CD8 T cells in the steady state. Immunity 21:391-400. Nagao K, Kobayashi T, Moro K, et al. (2012) Stress-induced production of chemokines by hair follicles regulates the trafficking of dendritic cells in skin. Nat Immunol 13:744-52. Nickoloff BJ, Turka LA (1994) Immunological functions of non-professional antigen-presenting cells: new insights from studies of T-cell interactions with keratinocytes. Immunol Today 15:464-9. Ohyama M, Terunuma A, Tock CL, et al. (2006) Characterization and isolation of stem cell-enriched human hair follicle bulge cells. J Clin Invest 116:249-60. Okiyama N, Katz SI (2014) Programmed cell death 1 (PD-1) regulates the effector function of CD8 T cells via PD-L1 expressed on target keratinocytes. Journal of autoimmunity 53:1-9. Papatriantafyllou M, Moldenhauer G, Ludwig J, et al. (2012) Dickkopf-3, an immune modulator in peripheral CD8 T-cell tolerance. Proc Natl Acad Sci U S A 109:1631-6. Rosenblum MD, Gratz IK, Paw JS, et al. (2011) Response to self antigen imprints regulatory memory in tissues. Nature 480:538-42. Stoitzner P, Tripp CH, Eberhart A, et al. (2006) Langerhans cells cross-present antigen derived from skin. Proc Natl Acad Sci U S A 103:7783-8. Stringer CP, Hicks R, Botham PA (1991) The expression of MHC class II (Ia) antigens on mouse keratinocytes following epicutaneous application of contact sensitizers and irritants. Br J Dermatol 125:521-8. Tietz W, Allemand Y, Borges E, et al. (1998) CD4+ T cells migrate into inflamed skin only if they express ligands for E- and P-selectin. J Immunol 161:963-70. Wang X, Fujita M, Prado R, et al. (2010) Visualizing CD4 T-cell migration into inflamed skin and its inhibition by CCR4/CCR10 blockades using in vivo imaging model. Br J Dermatol 162:487-96.
19
© 2015 The Society for Investigative Dermatology
Figure Legends Figure 1. MBP-IAu β-chain transgene expression in Ker-MBP mice is skin restricted. (a) MBP-IAu -chain transgenic construct. (b) Schematic illustration of the Ker-MBP Mouse model. (c) Semi-quantitative RT-PCR of the MBP-IAu β-chain in indicated tissues of Ker-MBP and wt mice. (d) Representative immunofluorescence stains for MBP1-10 (green), keratin15 (red) and nuclei (blue) in skin of Ker-MBP and control mice (Bar: 200µm). (e) 5x107 CFSElabelled splenocytes from Tg4 mice were transferred i.v. into Ker-MBP mice. The next day shaved abdomen of was treated with oxazolone (3%) or DPCP (2%) or was left untreated. 5 days later, CFSE profiles of Tg4 CD4+ T cells isolated from draining lymph nodes (LN) and skin were analysed (representative experiment; n.d.=not detected). (f) 5x107 CFSE labelled Tg4 splenocytes were co-transferred with MBP1-10-peptide loaded DCs into Ker-MBP mice. Representative CFSE dilution of Tg4 T cells from LN 5 days after transfer.
Figure 2. Functional MBP-MHC class II complex presentation in Ker-MBP mice only occurs upon contact sensitizer treatment (a) Representative dot plots of CD8 and Vβ8.2 vs. CD4 expression on thymocytes of Tg4xKer-MBP and Tg4 control mice. (b) Representative dot plots of Vβ8.2 vs. CD4 expression and histograms of CD44 and CD62L expression on CD4 T cells from lymph nodes of Tg4xKer-MBP and Tg4 control mice. (c) Tg4xKer-MBP and Tg4 were immunized by MBP1-10/CFA s.c. injection (t0) and EAE symptoms were monitored. Mean clinical score ± SEM is plotted over time. (n=8, n.s.). (d) Representative histology of the skin of Tg4xKerMBP and Tg4 mice 8 days after single application of oxazolone (3%). Infiltration of leukocyte populations was analyzed using hematoxylin/eosin (Bar:100µm), anti-MAC-3 (Bar:100µm) or anti-CD4 (red) staining (Bar: 25µm).
20
© 2015 The Society for Investigative Dermatology
Figure 3. Repeated induction of MBP-MHC class II complex presentation limits severity of subsequent EAE in Ker-MBP mice. (a) EAE was induced in Ker-MBP and wt mice (t0) and symptoms were monitored. Mean clinical score ± SEM is plotted over time (data pooled from two individual experiments, n=12). (b) Skin of Ker-MBP and wt mice was 4 times pre-treated with DPCP (each one week apart). 7 days later, EAE was induced (t0). Mean clinical EAE score ± SEM is plotted over time (data pooled from three individual experiments, n=18). (c) QRT-PCR of H2-Aa in skin of Ker-MBP and wt mice 7 days after a 4 DPCP times pre-treatment. Expression levels are calculated relative to Actb (n=6). (d) 7 days after a 4 times DPCP pre-treatment, 3% Oxazolone was applied to back skin of Ker-MBP and wt mice. One week later, the right ear was challenged by 1% oxazolone and ear thickness was monitored. As a control, ears of untreated wt mice were challenged. Ear thickness ± SEM is plotted over time (data pooled from 2 individual experiments, n=10).
Figure 4. DPCP pre-treatment of Ker-MBP mice induces a dominant form of tolerance. (a) 10 days after MBP1-10/CFA immunization of B10.PL mice, splenocytes were taken out and re-stimulated in vitro for 3 days in the presence of MBP1-10 peptide and IL-12. Subsequently, activated splenocytes were transferred i.v. into Ker-MBP or wt mice (t0) which were 4 times DPCP pre-treated. Mean clinical EAE score ± SEM is plotted over time (n=10). (b) Ker-MBP and wt mice were 4 times DPCP pre-treated. One week later, CD4+ T cells from spleen and lymph nodes were transferred into wt recipients. EAE was subsequently induced (t0) and clinical EAE score was monitored (data pooled from three individual experiments, n=18). (c) Percentages of CD25+FOXP3+CD4+ T cells in spleen and lymph nodes of Ker-MBP and wt mice, 4 times pre-treated with DPCP (n=6). (d) QRT-PCR of Foxp3 in skin of Ker-MBP and wt mice pre-treated like in (c). Expression levels are relative to Actb (n=6). (e) QRT-PCR on RNA isolated from total skin of DPCP pre-treated or untreated Ker-MBP and wt mice. Expression levels of indicated genes are relative to Actb (n=6-12). (f) Dkk3 concentration in serum of untreated and 4 times DPCP pre-treated Ker-MBP and wt mice, measured by
21
© 2015 The Society for Investigative Dermatology
ELISA (n=4). (g) QRT-PCR of Dkk3 in CD4+ T cells isolated from untreated and 4 times DPCP pre-treated Ker-MBP and wt mice. As negative control, CD4+ T cells isolated from untreated Dkk3-/- mice were used. As positive control, total skin of 4 times DPCP pre-treated mice was used (n=3).
Figure 5. Reduction of the MBP specific CD4+ T cell response in Ker-MBP mice depends on Dickkopf-3. (a) Ker-MBPxDkk3-/-, Dkk3-/-, Ker-MBP and wt mice were 4 times DPCP pre-treated before induction of EAE (t0). Mean clinical score ± SEM is plotted over time (data pooled from 2 individual experiments, n=12). (b) Ker-MBP and wt mice were 4 times pre-treated. In both groups, 500µg of anti-Dkk3 mAb clone 4.22 was applied i.p. along with the first three DPCP applications. One week after 4th application EAE was induced (t0) and symptoms were monitored. Mean clinical score ± SEM is plotted over time (data pooled from 2 individual experiments, n=16). (c) QRT-PCR of H2-Aa in skin of Ker-MBP.Dkk3-/- and Dkk3-/- mice 7 days after a 4 times DPCP pre-treatment. Expression levels were calculated relative to Actb (n=6). (d) Ker-MBP and Ker-MBP.Dkk3-/- mice were 4 times DPCP pre-treated or left untreated and EAE was induced. 7 days later splenocytes were in vitro re-stimulated with MBP1-10 peptide and analyzed by flow cytometry. Depicted are the percentages of IL-2, TNF and IL-17 producing CD4+ T cells (data pooled from three individual experiments, n=6).
22
© 2015 The Society for Investigative Dermatology
© 2015 The Society for Investigative Dermatology
© 2015 The Society for Investigative Dermatology
© 2015 The Society for Investigative Der
© 2015 The Society for Investigative Dermatology
© 2015 The Society for Investigative Dermatology