IL-15 is involved in lymphocyte homeostasis. To investigate the role of IL-15 in the skin in vivo, mice were generated that overexpress IL-15 in keratinocytes, resulting in increased IL-15 protein levels in the skin but not elevated IL-15 serum concentrations. Keratin 14 (K14)-IL-15 transgenic (tg) mice showed increased contact hypersensitivity (CHS) responses. Transfer of primed wild-type (wt) and tg T cells into naive wt or tg recipients indicated that skin-derived IL-15 enhanced the induction but not the elicitation phase of CHS. Tg mice could be sensitized even by suboptimal hapten concentrations. Accordingly, Langerhans cells (LC) from tg skin were identified as potent allostimulators, suggesting the involvement of IL-15-stimulated LC in the induction of adaptive immunity. Overexpression of IL-15 also strengthened innate immunity since tg mice infected with human HSV type I developed significantly smaller HSV skin lesions. In addition, tg mice resisted re-infection with HSV more effectively than wt mice did, which was associated with an elevated anti-HSV Ab production. Accordingly, injection of serum from re-infected tg mice protected naive recipients significantly from epicutaneous HSV infection, indicating that anti-HSV Ab produced by tg mice play an important role in resistance in vivo. Together, our results show that overexpression of IL-15 in the epidermis enhances both innate and adaptive cutaneous immunity.
IL-15 is a pleiotropic cytokine that is important for immune cell homeostasis as well as peripheral immune functions 1, 2. IL-15 and IL-2 are structurally related. The heterotrimeric receptors for IL-15 and IL-2 share both the IL-2R/IL-15Rβ (CD122) and the IL-2R/IL-15R common γ (γc) chains. The α chain of the IL-15R (IL-15Rα) and CD122 control binding, whereas CD122 and γc elicit signal transduction. In vitro and in vivo studies have demonstrated a critical role for IL-15 in the development and survival of the NK cell lineage 1. Other studies in mutant mice have shown that IL-15 is a selective growth factor for memory CD8+ rather than CD4+ T cells as well as for certain subsets of intraepithelial lymphocytes 3. Ubiquitous transgenic overexpression of IL-15 under the control of an MHC class I promoter led to the initial expansion of NK cells and memory-phenotype CD8+ T cells and then to the development of a fatal leukemia, resulting in premature death of these mutant mice 4. Therefore, it is difficult to investigate the long-term in vivo effects of IL-15 on other immunocompetent cell types.
IL-15 is broadly expressed in multiple cell types and tissues. Accordingly, several reports show abundant IL-15 mRNA transcripts in many tissues and cell types 1. However, IL-15 expression is tightly controlled at the levels of transcription, translation, and intracellular transport 5, 6. In particular, the IL-15 protein is posttranscriptionally regulated by several controlling elements, e.g. 12 AUG sequences in the 5′ UTR that impede translation, two insufficient signal peptides, and a negative regulator at the C terminus of the precursor protein. Thus, investigations on the in vivo function of IL-15 are greatly impaired.
Within the skin, the primary cellular source of IL-15 is keratinocytes. These cells express IL-15 mRNA but produce the IL-15 protein only upon stimulation such as with ultraviolet radiation or wounding, and in psoriatic lesions 7–9. IL-15 transcripts have also been detected in freshly isolated human Langerhans cells (LC) 7. IL-15 was even reported to be involved in the in vitro generation of human LC from monocytes 8. Additionally, IL-15 affects dendritic epidermal T cells of the murine skin, which are a small population of intraepithelial γ δ T cells 10, 11. Thus, there is growing evidence that skin-derived IL-15 might play a role in the regulation of immune responses.
To decipher the in vivo effects of keratinocyte-derived IL-15 in immunity and in governing T cell responses, we have generated transgenic mice that overexpress IL-15 in basal keratinocytes under control of the skin-specific keratin 14 (K14) promoter; these are termed K14-IL-15 transgenic (tg) mice. In the following, evidence is provided for a previously unappreciated in vivo role of IL-15 in bridging innate and adaptive cutaneous immune reactions.
2.1 Generation and phenotype of tg mice
To investigate the role of keratinocyte-derived IL-15 in vivo in immune responses, IL-15 expression was targeted to the epidermis using the K14 expression cassette. Several regulatory posttranscriptional mechanisms have been identified that control IL-15 translation and secretion 5, 6. Therefore, a modified mouse IL-15 cDNA was cloned that lacks the posttranscriptional checkpoints controlling endogenous IL-15 production, thus yielding optimal IL-15 overexpression. Modifications of the transgenic construct are depicted in Fig. 1A and included removing upstream AUG sequences that impair translation, replacing the insufficiently translated and secreted endogenous IL-15 signal peptide with the CD33 signal peptide, and stabilizing the C terminus of the IL-15 protein with a FLAG epitope tag. This FLAG epitope tag also enabled differentiation between endogenous and tg IL-15 expression in the skin.
To show correct expression of the transgene, immunohistochemistry and Western blot analysis was performed on tg skin tissue using anti-FLAG and anti-IL-15 Ab. As demonstrated in Fig. 1B and C, a strong and uniform expression of IL-15/FLAG was detected in the epidermis of tg animals. Expression of IL-15 in the epidermis did not lead to increased serum concentrations of IL-15 protein (Fig. 1C). Furthermore, tg mice showed normal numbers of CD4+, CD8+, NK1.1+ cells and CD122+ T cells in the spleen and lymph nodes compared to controls (data not shown). These findings suggest that targeted overexpression of IL-15 in the epidermis did not result in detectable systemic changes in the NK and T cell compartment. Transgene expression followed the expected pattern in other tissues, with the exception that IL-15 expression was below detectable levels in the thymus, where K14 expression has been documented in some transgenic strains before 12. The transgene insertion locus or possibly the transgene copy number may account for the lack of IL-15 expression in the thymus. Homozygous tg mice breed well and are healthy.
2.2 Enhanced acquired immune responses to contact allergens in tg mice
To study adaptive cutaneous immune responses in vivo, contact hypersensitivity (CHS) experiments were performed. Wild-type (wt) and tg mice were sensitized with 2,4,-dinitrofluorobenzene (DNFB) 13. Subsequently, all groups were ear-challenged with DNFB and ear swelling was assessed as a measure of CHS. Wt mice developed an ear swelling response that resolved after 120 h (Fig. 2A). In contrast, tg animals mounted a significantly increased and prolonged CHS response. Tg mice still presented with a marked ear swelling response even 120 h after challenge, which was similar to the maximal CHS response observed in wt mice at 48 h (Fig. 2A). As shown in Fig. 2B, CHS responses were Ag-specific since DNFB-sensitized mice failed to develop ear swelling after challenge with an irrelevant Ag (oxazolone). These findings indicate that overexpression of IL-15 by keratinocytes leads to a strong enhancement of adaptive immune responses.
Next we analyzed the development of CHS memory responses in vivo, since IL-15 has been reported to mediate T cell survival 14. Therefore, wt and tg animals, which had been sensitized and challenged with DNFB, were re-challenged with the same hapten after a six-week interval. In contrast to wt controls, tg mice mounted a significantly enhanced CHS response after re-challenge (Fig. 3).
2.3 Increased LC antigen-presenting function in tg mice
To investigate whether skin-derived IL-15 enhances the induction or the elicitation phase of CHS, adoptive transfer of bulk CD3+ T cells from DNFB-sensitized wt or tg mice into naive wt or tg recipients was performed. When naive wt recipients received primed T cells from tg mice, significantly increased CHS responses were elicited, suggesting that augmented T cell priming had occurred in tg mice (Fig. 4). However, if primed wt T cells were injected into naive tg or wt mice, similar CHS responses were measured. Hence, overexpression of IL-15 in the skin of the recipients does not result in an increased ear swelling, indicating that transgene expression enhances the induction but not the elicitation phase. Therefore, the augmented CHS responses may not be caused by an elevated skin-reactivity induced by IL-15, but may be caused by an enhanced sensitization.
Since LC are involved in the induction of CHS responses, the number and antigen-presenting function of LC were investigated. Immunofluorescence staining of epidermal ear sheets from tg and wt mice with anti-Langerin as well as with anti-IA Ab to visualize LC revealed normal numbers of LC in tg skin compared to wt skin (data not shown). To investigate the antigen-presenting function of LC, epidermal cell suspensions are usually prepared, enriched for I-A+ LC, and used in the mixed epidermal-cell–lymphocyte reaction. However, for a yet-unknown reason it is impossible to obtain epidermal cell suspensions from tg skin using trypsin with or without dispase digestion. Whether this inability to harvest epidermal cell suspensions by trypsinization of skin is linked to the increased resistance to apoptosis by IL-15-treated keratinocytes is currently under investigation 15.
To circumvent this obstacle and to still be able to study LC antigen-presenting function, we prepared 6-mm punch biopsies from tg and wt ears and mechanically split the epidermis from the dermis. Subsequently allogeneic T cells (H-2d) were overlayed with these wt and tg epidermal sheets to allow LC to migrate out of the epidermis to induce T cell proliferation. One-fifth of the total cells that emigrated from the epidermis and 50% of CD11c+ dendritic cells were Langerin+ (see supplemental Fig. 1A of Supporting Information). In control groups, T cells were either overlayed with LC-depleted tg epidermis or co-cultured with high concentrations of IL-15. LC from tg skin were significantly superior T cell stimulators than LC from wt skin were (Fig. 4B). The increased T cell stimulation was primarily induced by LC, because when T cells were overlaid with LC-depleted tg epidermis, it did not result in enhanced T cell activation (Fig. 4B). The LC depletion procedure did not affect cutaneous transgene expression (see supplemental Fig. 2 of Supporting Information). Furthermore, LC that had emigrated from tg epidermis showed enhanced signs of activation as evidenced firstly by the increased expression of B7 (CD80/ CD86) depicted in Fig. 4C and secondly by reduced CD205 (DEC205) expression (see supplemental Fig. 1B of Supporting Information).
To study whether this increased hyperreactivity of LC from tg skin in vitro was of functional relevance in vivo, tg and wt mice were sensitized with suboptimal DNFB concentrations. Wt mice were unable to mount CHS responses upon challenge when suboptimal sensitizing doses were used (Fig. 4D). In contrast, tg mice were still able to develop significant CHS responses even when DNFB doses as low as 0.0005% were used for epicutaneous sensitization. These findings indicate that the antigen-presenting function of LC in vivo is strongly enhanced by keratinocyte-derived IL-15.
2.4 Enhanced innate immune responses against HSV infection in tg mice
Several lines of evidence indicate that IL-15 is involved in the innate immune response to microbial pathogens 1, 16, 17. In agreement, we found, using RT-PCR, an up-regulation of IL-15 transcripts in the skin of normal mice infected with HSV type 1 (Fig. 5A). To investigate the role of IL-15 in cutaneous innate antiviral immunity in vivo, we epicutaneously infected tg and wt mice with HSV. This infection regimen has been successfully shown to produce cutaneous HSV lesions 18. HSV infection induced erythematous and blistering skin lesions in wt mice (Fig. 5B). However, the HSV-induced skin manifestations in tg mice were significantly smaller and better controlled, since the lesion size was minor compared with wt mice (Fig. 5B, C). Upon histopathology, HSV lesions in wt skin showed significantly more cellular inflammation containing neutrophils, lymphocytes and eosinophils compared with lesional tg skin (see supplemental Fig. 3 of Supporting Information). Additionally, staining of HSV-infected skin specimens with anti-HSV Ab revealed a markedly reduced expression of viral proteins in tg skin and subcutaneous tissue in comparison with infected wt skin (Fig. 5D). Thus, overexpression of IL-15 by murine keratinocytes is able to induce strong protective antiviral innate immunity against HSV infection.
2.5 Increased Th cell function in vivo upon HSV re-infection in tg mice mediates antiviral immune responses
To determine the effects of IL-15 on cellular defense responses, we employed tg and wt mice that had been exposed to HSV and re-infected them after 4 weeks. Although some resistance was apparent in the wt mice — the animals were inoculated with a 10-fold higher dose but lesions were smaller than following the primary infection — the tg mice developed significantly smaller HSV-induced skin lesions compared with wt mice, suggesting that a protective antiviral effect was caused by cutaneous overexpression of IL-15 (Fig. 6A).
Therefore, splenic and lymph node T cells were obtained after HSV re-infection and analyzed. Tg mice showed increased numbers of NK1.1+ cells after re-challenge compared with wt animals, whereas activation of CD8+ T cells was similar in both groups of mice (Fig. 6B). In addition, enhanced numbers of activated CD4+CD44+ T cells were detectable in HSV-re-infected tg mice (Fig. 6B). These CD4+ T cells produced, upon CD3/CD28 stimulation, significantly more IFN-γ but not IL-4 compared with CD4+ T cells obtained from controls, suggesting differentiation into Th1 by CD4+ T cells from tg mice cells upon HSV infection (Fig. 6C). These findings indicate that local IL-15 expression by keratinocytes is able to stimulate NK1.1+ cells and CD4+ T cell responses, the latter being probably involved in mediating adaptive immune responses.
2.6 Adoptive transfer of serum from HSV-reinfected tg mice protects naive recipients from epicutaneous HSV challenge
HSV infection leads to the development of anti-HSV Ab responses in mice and humans. Therefore Ig serum levels were measured in HSV-re-infected tg and wt mice. Untreated animals served as negative controls. Wt controls developed increased concentrations of IgG3 upon viral re-infection. Highly increased concentrations of IgG2a and IgG3 were detectable in tg serum, suggesting marked induction of Ig production upon HSV infection in tg mice (data not shown). To determine if this Ab response was directed against HSV, serum from re-infected tg and control mice was allowed to bind to HSV antigens utilizing Western blot analysis. The results shown in Fig. 7A demonstrate a marked antiviral Ab production in re-infected tg mice compared with wt mice; this Ab was specific for even lower molecular weight HSV Ag.
To investigate whether these antibodies are protective in vivo, an adoptive transfer model was employed. Serum from HSV-re-infected tg mice and controls was harvested after the second viral exposure. Subsequently, serum from tg or wt mice was injected intravenously into naive wt mice. Twenty-four hours after the transfer, recipient mice were challenged by epicutaneous application of viable HSV. Animals that were treated with serum from wt controls developed skin lesions similar to the PBS recipients. In contrast, wt mice that had received serum from tg mice showed significantly smaller lesions after HSV infection (Fig. 7B).
These data support the concept that IL-15 not only plays an important role in the activation of innate and adaptive immunity but also links both types of responses during antiviral defense.
To better understand the in vivo effects of IL-15, we generated a transgenic mouse model that overexpresses IL-15 in the epidermis. We decided to express the transgene locally in the skin for the following reasons. Ubiquitous transgene overexpression of IL-15 under control of a MHC class I promoter led to increased IL-15 serum levels and caused the development of a fatal leukemia, resulting in premature death of these transgenic animals 4. Thus, this model does not allow the investigation of long-term in vivo effects of IL-15 on immune cells. Weanticipated to overcome this problem by confining the overexpression of IL-15 to a particular organ and we selected the skin since it has been demonstrated that keratinocytes can function as a natural source of IL-15 and that the secretion of IL-15 by these cells is inducible 7, 9.
The epidermis is constantly exposed to potential noxious influences and thus represents an organ in which innate and adaptive immune responses are frequently induced. CHS responses to topically applied haptens were found to be significantly increased in tg mice. We can exclude that the enhanced ear swelling response is due to a decreased inflammatory threshold as a consequence of the IL-15 overexpression since the inflammatory response against irritants was not affected (data not shown). Since IL-15 is involved in the stimulation of T cells, which elicit CHS responses, it was necessary to dissect the effects of cutaneous IL-15 expression on the elicitation of CHS, by adoptive transfer experiments 14, 19. Our data argue against a significant effect of transgene expression on the elicitation of CHS.
These results suggest that the enhanced CHS responses appear to be due to a more efficient priming of naive T cells by IL-15-stimulated LC. Indeed, LC from tg skin showed an increased antigen-presenting function in vitro and in vivo since LC from tg skin were potent allostimulators and tg mice could be even sensitized with suboptimal hapten doses. In vitro studies from other groups have shown that culture of human monocytes in GM-CSF plus IL-15 induced LC-like dendritic cells, suggesting that IL-15 might play an important role in the development of LC precursors 8. In tg mice, however, normal numbers of epidermal I-A+ LC were detected by in situ staining of epidermal sheets. Furthermore, migration experiments using FITC as a hapten revealed no enhanced migration of hapten-laden LC out of the epidermis into the draining lymph nodes since equal numbers of FITC+Langerin+ LC were detectable inthese lymph nodes (data not shown).
Herpes simplex is one of the most common viral infections of the skin. Thus, we were interested in the impact of overexpression of IL-15 on this type of infection. Our observation that tg micecoped much better with primary infection against HSV implies that overexpression of IL-15 might enhance an innate immune response. Accordingly, the analysis of spleen cells from HSV-re-infected tg mice revealed an enhanced proportion of NKT cells and activated CD4+ T cells. The increased expansion of NK1.1+ T cells may contribute to the enhanced innate immune response.
The increase of CD44-expressing CD4+ T cells upon HSV re-infection is indicative of Th cell activation in tg mice. On the basis of this finding and of the observation that the tg micecoped even better with the HSV re-infection, we surmised that an adaptive immune response might have been induced in tg mice. The activated CD4+ T cells may belong to the Th1 type since, upon stimulation with anti-CD3 and anti-CD28, these cells released increased amounts of IFN-γ but no IL-4. In addition, tg mice revealed significantly enhanced Ig levels of all subtypes. These Abappear to be protective since HSV lesions were significantly reduced in wt mice that had received serum from HSV-re-infected tg mice.
At this point, a direct effect of IL-15 on B cells cannot be completely ruled out since a previous report showed that IL-15 co-stimulated pre-activated B cells for Ig production 20. Alternatively, IL-15 might activate B cells via induction of T cell help. Accordingly, a marked activation of CD4+CD44+ (Fig. 6B) and CD4+CD69+ (data not shown) T cells was seen in tg mice after re-infection with HSV. Induction of CD4+ T cells might be a direct effect of IL-15 or indirectly mediated by IL-15-stimulatedLC. However, the functional relevance of LC in inducing antiviral immunity appears to be somewhat controversial since it was shown that CD8α+ dendritic cells were superior when compared with LC in stimulating MHC class-I restricted immune responses against HSV in a transgenic model 21. Therefore, we can currently not completely exclude the possibility that CD8α+ or dermal dendritic cells also participate in the antiviral defense.
In summary, the present study demonstrates that local overexpression of IL-15 in the epidermis protects mice from cutaneous HSV infection. The fact that naive tg mice show resistance to primary HSV infection indicates that overexpression of IL-15 enhances an innate antiviral immune response. In addition, upon re-infection tg mice are even better protected and produce Ab that confer protection, indicating that an adaptive immune response has been induced. The latter may be due to enhanced activation of APC by IL-15, as demonstrated by the enhanced sensitization phase of CHS. Therefore, this model supports a crucial role of IL-15 in bridging innate and adaptive immunity. This may be of practical relevance for the future treatment of virus-induced cutaneous diseases. The topicalimmunomodulator imiquimod has turned out to be highly beneficial in the treatment of human papilloma virus-induced warts and other viral cutaneous disorders 22. Its efficacy is based on the fact that imiquimod stimulates both the innate immune response and the cellular arm of acquired immunity, indicating the advantage of utilizing both pathways. Drugs inducing the release of IL-15may follow this therapeutic route.
4 Materials and methods
4.1 Generation of K14-IL-15-transgenic mice
The gene for murine IL-15 was placed under the control of the human K14 promoter using standard methods. The BamHI cloning site was modified by ligating a polylinker into this site, resulting in a multiple cloning site containing the restriction enzyme sites SalI, BglII, BamHI and XbaI and resulting in the plasmid pAMM11. The approximately 500 bp BglII/XbaI fragment of pEGFP-c2 CD33 IL-15 FLAG was cloned into the BglII/XbaI site of pAMM11 to create pAMM77.
Plasmid DNA to be used for microinjections was first purified through CsCl gradient centrifugation. The expression cassette containing the IL-15 gene under the K14 promotor was released from the plasmid and purified through 0.7% agarose gel electrophoresis following digestion with SphI and SmaI, extracted from the gel (PCR purification Kit; Roche, Mannheim, Germany), resuspended in TE* buffer (10 mM Tris, pH 7.4 / 0.1 mM EDTA) and used for microinjection at a concentration of 2 ng/μl into mouse C57BL/6/C3H/HeN F1 × C57BL/6 and FVB/N oocytes. Two founder lines with similar transgene expression were identified by PCR (AM28, CAATGATATACACTGTTTGAGATGA; AM65, CGTGTTGATGAACATTTGGACAA; cycling profile: 95°C for 3 min; [95°C for 1 min; 54°C for 1 min; 72°C for 1 min × 35; 72°C for 5 min]) and Southern blotting. Experiments were performed with tg mice on a C57BL/6/C3H/HeN background. Mice were housed under specific pathogen-free conditions and experiments were performed according to institutional regulations.
4.2 Reverse-transcription PCR
Mouse tissues were snap-frozen before RNA isolation and reverse transcription. RNA was extracted using RNAeasy columns (Qiagen, Hilden, Germany) according to the manufacturer's instructions. The cDNA was synthesized from 1 μg of total RNA using random hexanucleotide primers and the Reverse Transcription System (Promega, Madison, WI, USA) Primers used were: IL-15 forward, 5;prime;TTTCATCCCAGTTGCAAAGTT-3′ and IL-15 reverse, 5′-ACATTCCTTGCAGCCAGATT-3′; β-actin forward, 5′-GTGGGGCGCCCCAGGCACCA-3′ and β-actin reverse, 5′-CTCCTTAATGTCACGCACGATTTC-3′. Cycling profile: 94°C for 4 min; [94°C for 1 min; 56°C for 1 min; 72°C for 1 min] × 35; 72°C for 5 min. Aliquots of PCR products were electrophoresed on 1.5% agarose gelsand visualized by ethidium bromide staining.
The CHS experiments were performed as described in detail elsewhere 13. Re-challenge was performed 6 weeks after the first challenge. For sensitization with suboptimal doses, mice were painted with 100 μl of 0.05, 0.005 and 0.0005% DNFB on their shaved backs. Challenge was performed by painting 12 μl of 0.3% DNFB on both sides of the left ear. The degree of earswelling was determined at the indicated time points after re-challenge in comparison with the non-rechallenged right ear.
4.4 Adoptive transfer of immune response
Donor mice were sensitized by painting 100 μl DNFB solution (0.5% in acetone / olive oil, 4 / 1) on the shaved back on day 0. On day 5, the spleen and regional lymph nodes were removed, single-cell suspensions were prepared as described before and 12×106 T cells were injected i.v. into each recipient mouse 23. After 24 h, recipients were challenged and ear swelling was evaluated as above.
4.5 HSV infection
Briefly, the backs of sex- and age-matched wt and tg mice were shaved, tape-stripped and epicutaneously infected with 2.5×104 plaque-forming units (PFU) of a wild-type HSV type 1isolate from a clinical case (a kind gift from Prof Mary Norval, Medical Microbiology, University of Edinburgh, GB) 18. The virus strain was prepared and titered on Vero cells using standard techniques. After infection, skin lesions were measured in two dimensions using a Venier caliber (Mitutoyo, Japan).
For re-challenge experiments, mice were re-infected 4 weeks after the first HSV exposure using 2.5×105 PFU. Four days after re-infection, the spleen of the mice was prepared and lymphocytes analyzed by flow cytometry and for cytokine production following stimulation in vitro. Peripheral blood was obtained from anesthetized mice by cardiac puncture. The serum was collected and stored until antibody analysis or for use in adoptive transfer. For adoptive transfer, groups of naive wt mice were i.v. injected with 500 μl serum obtained from HSV-re-infected wt or tg mice 96 h and 24 h before these differentially treated mice were epicutaneously infected with 25,000 PFU HSV.
4.6 Mixed epidermal-cell–lymphocyte reaction
Proliferation of T cells was assessed by [3H]thymidine incorporation. Cells (1×106/ml) were cultured in triplicates in 96-well-round-bottom plates, in a final volume of 200 μl, and one 6-mm punch biopsy piece of an epidermal sheet was added to each well. Epidermal sheets were prepared by mechanically separating the epidermis from the dermis using forceps. For removal of epidermal LC, ears of mice were topically treated on four consecutive days with mometason furoat (Ecural®, Essex Pharma, Munich, Germany) and biopsied one week later. This treatment did not significantly impair transgene expression (see supplemental Fig. 2 of Supporting Information). Other groups of T cells were cultured in 50 ng/ml IL-15 (R&D Systems, Wiesbaden, Germany).
4.7 Cell preparation and flow cytometry
Single-cell suspensions of spleens were prepared as described before 23. For harvesting, LC epidermal sheets were prepared and LC were allowed to migrate out of the epidermis into culture medium for 3 days. Expression of cell surface markers was analyzed by standard four-color flow cytometry on a FACScaliburTM flow cytometer (BD PharMingen, Heidelberg, Germany) with CELLQuestTM software (BD PharMingen). Cells were stained for FACS® analysis in PBS containing 1% FCS with the following mouse monoclonal antibodies: FITC-conjugated anti-CD43 (clone 1B11) and anti-CD44 (clone IM7); PE-conjugated anti-CD122 (clone TM-β1), anti-NK1.1 (clone PK136) and anti-TCR β-chain (clone H57-597); peridinin chlorophyll protein-conjugated anti-CD3 (clone 145-2C11); allophycocyanin-conjugated anti-CD4 (clone RM4-5), anti-CD8 (clone 53-6.7) and anti-Langerin/CD207 (clone 929F3, kindly provided by Dr S. Saeland, Schering-Plough, Dardilly, France). Isotype-matched control antibodies were included in each staining. Unless indicated, all antibodies as well as isotype-matched controls were obtained from BD PharMingen.
4.8 Histology, immunohistology and immunofluorescence
Tissue specimens were fixed and staining was performed by standard methods 24. Epidermal sheets were stained as described previously 24. In brief, ears were mechanically split into dorsal and ventral sides, incubated in 2 mM EDTA, washed with PBS and fixed in acetone. Sheets were incubated in 1% FCS in PBS and stained with the antibody overnight (anti-mouse-CD3 and anti-mouse-Thy1.2; BD PharMingen). Sheets were then incubated with a Oregon-Green- or Texas-Red-coupled secondary antibody (Molecular Probes), mounted onto slides and examined using an Olympus BX61 microscope and the MetaMorph software (Visitron Systems, Puchheim, Germany). Positive cells in randomly selected visual fields were counted using an ocular grid.
4.9 Western blot analysis
Serum of tg mice and wt controls as well as skin and thymus lysates or HSV proteins were loaded directly onto 15% polyacrylamide gels. Recombinant mIL-15 protein was used as a positive control. Proteins were electrophoresed under denaturing conditions and electroblotted onto nitrocellulose membranes (Amersham Biosciences Europe, Freiburg, Germany) at 150 mA for 1 h. Membranes were blocked for 2 h with 5% non-fat dry milk in TBS plus 0.5% Tween-20 (TBST) and then incubated overnight with the appropriate antibody (anti-mouse-IL-15, clone M49, a kind gift from Genmab, Utrecht, The Netherlands and Amgen Inc., Thousand Oaks, CA, USA) diluted 1:500 in TBST plus 1% non-fat dry milk. Membranes were washed with TBST and incubated for 2 h with horseradish-peroxidase-conjugated secondary antibody diluted 1:1000 in TBST. Proteins were detected using enhanced chemoluminescence reagents (Amersham Pharmacia Biosciences, Europe).
4.10 Cytometric bead assay
The cytokine activity in culture supernatants of spleen cells (2×106 cells/ml) from tg and wt mice was assayed by a cytometric bead assay (CBA; BD PharMingen) according to the manufacturer's instructions. For detection and quantification of Ig, serum from HSV-re-infected wt and tg mice was collected and assayed by a murine Ig CBA kit.
4.11 Statistical analysis
The significance of differences between the mean values obtained was assessed by the Student's t-test for unpaired data. p values <0.05 were regarded as being significant.
The excellent technical assistance of Maik Voskort and Joachim Windau is gratefully acknowledged. This work was supported by the German Research Associationgrant DFG BE 1580/6–1 (S. B.), Interdisciplinary Center for Clinical Research (IZKF) grant Lo2/065/04 (K. L.; S. B.), IMF grant (S. B.) from the University of Muenster, and NIDDK grants DK 33506, 43351, 51003 and 54427 (H.-C. R).