UV Exposure Boosts Transcutaneous Immunization and Improves Tumor Immunity: Cytotoxic T-Cell Priming through the Skin

ORIGINAL ARTICLE

UV Exposure Boosts Transcutaneous Immunization and Improves Tumor Immunity: Cytotoxic T-Cell Priming through the Skin Pamela Stein1, Gerd Rechtsteiner1,4, Tobias Warger1,5, Tobias Bopp1, Thorsten Fuhr1, Steve Pru¨fer1, Hans-Christian Probst1, Michael Stassen1, Peter Langguth2, Hansjo¨rg Schild1 and Markus P. Radsak1,3 Immunologic approaches to combat cancer aim at the induction of tumor-reactive immune responses to achieve long-term protection. In this context, we recently developed a transcutaneous immunization (TCI) method using the Toll-like receptor (TLR) 7 agonist imiquimod and a peptide epitope. Application onto intact skin induces potent cytotoxic T lymphocyte (CTL) responses and protection against transplanted tumors. The purpose of this study was to explore the effects of UV irradiation on imiquimod-based TCI. Here we show that skin exposure to low-dose UV light before TCI with imiquimod strongly boosts specific CTL responses leading to memory formation and enhanced tumor protection. Toward the mechanisms, we show that the activation of bone-marrow-derived dermal dendritic cells (DCs), but not Langerin-expressing DCs, is responsible for enhanced CTL activation. We describe an optimized TCI method that mediates enhanced CTL and antitumor responses by a DC- and TLR-dependent mechanism. These data may provide the basis for the future development of advanced vaccination protocols against tumors and persistent virus infections. Journal of Investigative Dermatology (2011) 131, 211–219; doi:10.1038/jid.2010.254; published online 26 August 2010

INTRODUCTION UV light may have a variety of acute and chronic effects on the skin: acute exposure directly induces inflammatory responses resulting in skin irritations or sunburn (Cooper et al., 2003), whereas chronic or low-dose UV exposure causes no apparent inflammatory reactions, but leads to immunosuppression and the development of skin cancers (Halliday and Lyons, 2008). UV-induced immunosuppression is mediated by T regulatory (Treg) cells as clearly shown in skin grafting models (Fisher and Kripke, 1982) and more recently in delayed-type hypersensitivity (DTH) reactions (Rana et al., 2008). In the latter models, IL-10-producing CD4 þ CD25 þ FoxP3 þ Treg cells have been identified to be 1

Institute of Immunology, Johannes Gutenberg University Medical Center, Mainz, Germany; 2Institute of Pharmacy, Johannes Gutenberg University, Mainz, Germany and 3Third Department of Medicine, Johannes Gutenberg University Medical Center, Mainz, Germany

4

Current address: Fresenius Medical Care, Bad Homburg, Germany.

5

Current address: Department of Immunology, Schering-Plough, Zurich, Switzerland. Correspondence: Markus P. Radsak, Institute of Immunology, Johannes Gutenberg-University Medical Center, Obere Zahlbacher Street 67, D-55131 Mainz, Germany. E-mail: [email protected] Abbreviations: APC, antigen-presenting cell; CTL, cytotoxic T lymphocyte; DC, dendritic cell; DTH, delayed-type hypersensitivity; LC, Langerhans cell; TLR, Toll-like receptor

Received 21 October 2009; revised 24 June 2010; accepted 12 July 2010; published online 26 August 2010

& 2011 The Society for Investigative Dermatology

responsible for the diminished response against haptens (Schwarz et al., 2004; Maeda et al., 2008). In contrast, the imidazoquinoline derivative imiquimod is a synthetic Toll-like receptor (TLR) 7 agonist (Hemmi et al., 2002) and induces inflammatory responses when applied to the skin (Palamara et al., 2004). Besides this, imiquimod is also effective in patients with HPV-associated or malignant skin diseases, when administered topically (Navi and Huntley, 2004). In this context, we and others have shown that the concurrent application of a cytotoxic T lymphocyte (CTL) epitope with imiquimod onto the skin (transcutaneous immunization, TCI) elicits potent primary CTL responses (Thomsen et al., 2004; Rechtsteiner et al., 2005). Induction of specific immune responses by TCI with imiquimod takes advantage of the direct access to skin-resident antigenpresenting cells (APC) to initiate adaptive immunity (Romani et al., 1989; Stoitzner et al., 2008). However, TCI with imiquimod alone merely mediates a delay in growth of transplanted EG.7 tumors, as shown by us (Warger et al., 2007) and also by others in B16 melanoma models (Itoh and Celis, 2005; Stoitzner et al., 2008). Thatcher et al. (2006) have investigated the concurrent use of imiquimod with UV exposure. In a DTH model, they showed that imiquimod treatment of the skin before UV exposure prevents the establishment of UV-induced hapten tolerance by an IL-12p70-dependent mechanism. Interestingly, they found a synergistic activation of CD11c þ APC in the draining lymph nodes on imiquimod and UV treatment. www.jidonline.org

211

P Stein et al. UV Enhances Transcutaneous Immunization

UV light may be suppressive on local inflammation of the skin (Maeda et al., 2008; Rana et al., 2008) and suppress T cell-mediated immune responses (Ghoreishi and Dutz, 2006). This is prevented by the TLR7 agonist imiquimod (Thatcher et al., 2006). To this end, we investigated the effects of UV exposure in a DTH model in the presence or absence of imiquimod. The mice were sensitized with dinitrofluorobenzene (DNFB) on the shaved back skin and challenged 4 days later on the ear. As depicted in Figure 1a and consistent with others (Thatcher et al., 2006), local UV exposure prevented ear swelling, whereas the response was restored after the additional treatment with imiquimod. To address directly how this kind of UV pretreatment affects peptide-specific CTL responses elicited by TCI with imiquimod, we applied imiquimod (50 mg containing 2.5 mg imiquimod per day on 2 consecutive days) together with the major histocompatibility complex (MHC) class I-restricted T cell epitope OVA257–264 after low-dose UV-B exposure. We detected a greatly increased frequency of peptide-specific CD8 T cells after UV-B exposure compared with the robust CTL activation detected after TCI with imiquimod alone (Figure 1b). Notably, the combination of UV exposure with imiquimod-based TCI not only restored, but greatly enhanced T cell activation as indicated by an enormous increase in cytolytic activity (Figure 1c). To further characterize UV-enhanced TCI with imiquimod, we pretreated the animals 24 hours before with UV light of the A, B, or C spectra. The CTL response was analyzed 7 days later by the frequency of specific CTL and an in vivo cytotoxicity assay. The pretreatment of the skin with UV-B was superior to UV-C in the induction of specific CD8 T cells (Supplementary Figure S1A online) with similar cytolytic activity (Supplementary Figure S1B online). In contrast, the pretreatment with UV-A did not increase the CTL frequency or in vivo cytolytic activity compared with TCI with imiquimod alone. These results indicated that preconditioning of the skin with UV-B or C light enhances the TCI with imiquimod-induced T cell response in a synergistic way. 212

Journal of Investigative Dermatology (2011), Volume 131

**

100 80

*

60 40 20 0 sens. chall.

+ +

+ +

+ +

UV-B

Imiquimod

UV-B imiquimod

Specific CD8+ T cells (%)

UV strongly augments TCI with imiquimod-induced primary CTL responses

120

Specific lysis (%)

RESULTS

NS

140 Ear swelling (%)

To evaluate whether this observation is relevant in the context of transcutaneous vaccinations, we investigated the effects of UV exposure on CTL responses induced by TCI with imiquimod. We found that low-dose UV exposure not only reverted UV-mediated suppression, but actually boosted specific CTL responses leading to an enhanced primary CTL response, efficient memory CTL formation and improved tumor protection. Toward the mechanisms, we show that UV/TCI with imiquimod is accompanied by an increased recruitment of dermal CD11c þ dendritic cell (dermal DC) to the draining lymph nodes. Taken together, in this work we provide evidence that the combination of an immunosuppressive with an immunostimulatory treatment of the skin boosts a specific CTL response. These results may provide the basis for advanced therapeutic vaccination strategies against tumors and persistent infections.

10

+ +

*

*

– +

*

8 6 4 2 0 Untreated UV-B

100

*

TCI

*

UV-B TCI

*

80 60 40 20 0 Untreated UV-B

TCI

UV-B TCI

Figure 1. Modulation of T-cell responses in the skin by UV exposure and imiquimod. (a) Mice were sensitized (0.5% dinitrofluorobenzene (DNFB)) after UV-B exposure (90 mJ cm2) or imiquimod (2.5 mg) as indicated. Antigen challenge was performed 4 days later on the ears (0.2% DNFB). (b, c) Mice were immunized by transcutaneous immunization (TCI) with imiquimod (2.5 mg) using the peptide SIINFEKL (100 mg) in 50 mg cream per application without or after UV-B exposure (90 mJ cm2). The frequency of specific CD8 þ T cells after 7 days in the spleen was assessed by flow cytometry (b) or by in vivo cytolytic activity (c). A cumulative analysis of four independent experiments, each performed with three mice per group, is shown. *Significant difference to sens./chall., by unpaired Student’s t-test. **Significant difference to UV-B by unpaired Student’s t-test.

UV-B-enhanced TCI with imiquimod induces improved memory CTL formation

We have previously shown that TCI with imiquimod induces a potent primary CTL response, but fails to induce significant memory T cell formation due to inefficient co-stimulation. This may be overcome by CD40 ligation (Warger et al., 2007). As we found the most prominent primary CTL response after TCI with imiquimod post-UV-B exposure, we focused on UV-B-mediated effects in the following experiments. We asked whether this treatment regimen also has an impact on memory formation and analyzed the peptide-specific CTL for frequency, phenotype and in vivo

P Stein et al. UV Enhances Transcutaneous Immunization

CTL response, but also by creating an effective memory response as important prerequisite for long-term protective immune responses. Importantly, the observed memory response was almost comparable to TCI combined with CD40 ligation that induces a functionally relevant memory response as measured by long-term tumor protection (Warger et al., 2007).

Untreated

Vehicle

UV/vehicle

TCI

* * *

UV/TCI

UV treatment synergizes with TCI with imiquimod in a dose- and time-dependent manner

*

*

*

CD40/TCI 0

*

*

10 20 Specific CD8+ T cells (%)

Specific lysis (%)

100

*

80

*

60 40 20 0 Untreated

TCI

UV-B vehicle

UV-B TCI

Anti-CD40 TCI

Figure 2. UV-B exposure before transcutaneous immunization (TCI) with imiquimod induces a cytotoxic T lymphocyte (CTL) memory response. (a) C57BL/6 mice were immunized by TCI with imiquimod in the absence or presence of UV-B light (90 mJ cm2) or after intravenous anti-CD40-Ab (100 mg) injection (days 3 to 0). The frequency of peptide-specific CD8 þ T cells was monitored after 7, 21, and 40 days. (b) After 30 days the in vivo cytolytic activity was assessed after adoptive transfer of peptide loaded and carboxyfluorescein diacetate succinimidyl ester-labeled target cells without prior restimulation. The depicted results are summarized data from at least three independent experiments with three mice per group. *Significant difference by unpaired Student’s t-test.

cytolytic activity over time, and in the memory phase after immunization. The number of specific CD8 T cells decreased over time after reaching a maximum after 1 week postimmunization treatments (Figure 2a). After 40 days, peptidespecific CTLs were detectable only in the group treated with UV or CD40 ligation in addition to TCI with imiquimod. The majority of these peptide-specific CTLs were antigen experienced indicated by high expression of CD44 (anti-CD40/TCI: 95.4±2.1%; UV/TCI: 89.9±6.9%). Additional coexpression of CD62Lhi compatible with a central memory phenotype was found in: 34.7±13.9% in the anti-CD40/TCI group and in 16.3±4.5% in the UV/TCI group, whereas the majority in both groups had a CD44hiCD62Llo phenotype, consistent with an effector phenotype (Seder and Ahmed, 2003). Consistently, significant cytolytic activity was observed only in these groups (Figure 2b). These results suggest that the combination of low-dose UV-B exposure with imiquimod-based TCI with imiquimod is truly synergistic not only by inducing a superadditive primary

UV irradiation may induce skin irritation in a dose-dependent manner. To determine the dose of UV-B required for the synergistic effect on TCI with imiquimod, we treated mice as before titrating the time of UV exposure from 30 to 300 mJ cm2 one day before TCI with imiquimod. As measure of skin irritation we determined the minimal erythema dose after UV treatment, which was at 300 mJ cm2 (Supplementary Figure S2 online) and consistent with others (Maeda et al., 2008; Rana et al., 2008). There was a dose-dependent enhancement of the TCI with imiquimod-induced CTL response that was not further increased beyond 90 mJ cm2 of UV-B (Figure 3a). Interestingly, the pretreatment with UV light still augmented the induced CTL response at a UV dose below the minimal erythema dose indicating that the activation of skin-resident cells by UV treatment still synergizes with TCI with imiquimod at a dosage below an amount that causes visible irritation of the skin. Next, we determined the time interval required between UV and imiquimod treatment to induce the optimal CTL response. Therefore, the interval between UV exposure and TCI with imiquimod was varied to obtain the time sequence for optimal CTL induction. We found the maximal CTL response when UV treatment was performed 24 hours before TCI with imiquimod, as observed by the frequency of peptide-specific CD8 þ T cells and target cell lysis (Figure 3b and c). These results suggest that the activation of skin-resident cells by UV irradiation sensitizes these cells for TLR-mediated activation and enhanced peptide presentation to induce a superior CTL response within a narrow time frame of 24–48 hours. Defining the spectra, timing, and dose of UV exposure before TCI with imiquimod, we established a treatment protocol resulting in a remarkably enhanced CTL response without any additional skin irritation. Interestingly, the optimal dose and timing of UV exposure before TCI with imiquimod coincides with reported regimens for the generation of UV-induced tolerance (Fisher and Kripke, 1982; Schwarz et al., 2004; Maeda et al., 2008; Rana et al., 2008). UV/TCI with imiquimod increases protection against EG.7 tumors

After characterizing an improved CTL induction upon TCI with imiquimod post-UV-B exposure, we wanted to evaluate this vaccination protocol in terms of tumor protection. Therefore we inoculated EG.7 thymoma cells expressing chicken ovalbumin as a surrogate tumor antigen. Tumors were allowed to grow until palpable before the immunization treatments were initiated: mice were either immunized with TCI with imiquimod or with additional UV exposure at www.jidonline.org

213

P Stein et al. UV Enhances Transcutaneous Immunization

Specific lysis (%)

100

*

*

*

*

150

300

80 60 40 20 0 No UV

30

45

60

75 90 120 UV-B / mJ cm–2

18 16 14 12 10 8 6 4 2 0

100

* *

*

Specific lysis (%)

Specific CD8+ T cells (%)

TCI

80 60

*

*

40 20 0

–48 h Untreated

–24 h

+24 h +24/48 h UV-B TCI

–48 h –24 h +24 h +24/48 h Untreated UV-B TCI

Figure 3. Low dose of UVB before transcutaneous immunization (TCI) with imiquimod leads to increased immune responses. C57BL/6 mice were immunized and UV-B irradiated as indicated. (a) Dose of UV light was titrated from 0 to 300 mJ cm2 and cytolytic activity was assessed. To determine the optimal timing of UV treatment, we irradiated mice 48 or 24 hours before and 24 or 24 þ 48 hours after immunization. Specific CD8 þ T cells (b) and cytolytic activity (c) were quantified by flow cytometric analysis. One representative experiment out of two performed with n ¼ 3 mice per group is shown. *A significant difference (Po0.05) by unpaired Student’s t-test compared with the no-UV group in a or the 24 hour group in b and c.

weekly intervals. UV exposure alone had no impact on tumor growth compared with untreated animals (Figure 4a, upper panel) and resulted in a median death after 22 days (range 19–28). TCI with imiquimod significantly delayed tumor growth and induced durable tumor regression in 8 of 37 mice (22%). Animals treated with UV exposure before TCI with imiquimod all showed an initial response resulting in significantly smaller tumors compared with the group without UV treatment (Figure 4a, lower panel). Durable protection was observed in 14 of 33 animals (42%) and significantly higher compared with the group treated with TCI alone (Figure 4b). Analysis of the relapsed tumors revealed that tumor growth was not because of loss of the target antigen. Reisolated tumor cells from tumor-bearing mice still expressed the target epitope SIINFEKL in the context of H2Kb measured with a specific antibody (data not shown). Together, these results indicate that the improved activation of CTL induced by UV and TCI with imiquimod leads to enhanced tumor protection. Langerhans cells and dermal DC mediate CTL activation on imiquimod-based TCI and UV/TCI

For the activation of CD8 T cells the interaction with professional APC is crucial (Banchereau and Steinman, 214

Journal of Investigative Dermatology (2011), Volume 131

1998). Using our imiquimod-based TCI approach, it was found that T cells need to interact with skin-resident APC such as Langerhans cells (LCs) or dermal DCs for priming (Romani et al., 1989; Stoitzner et al., 2008). In this context, a number of DC populations have been described besides epidermal LCs: the CD11c þ dermal DC as well as bonemarrow-derived Langerin þ dermal DC populations (Bennett et al., 2005; Bursch et al., 2007; Ginhoux et al., 2007; Poulin et al., 2007). To address the contributions of distinct DC subsets, we analyzed the number and phenotype of DC populations in the draining lymph nodes after imiquimodbased TCI or UV/TCI and FITC treatment of the skin for the identification of skin-derived DC populations. After the immunization treatments, the number of skin-derived FITC þ and CD11c þ DCs in the draining lymph nodes (Figure 5a), but not Langerin þ DCs (Figure 5b), was increased with a maximum after 2 days and additional UV exposure. Furthermore, we found the increased induction of IL-12 in the draining lymph nodes of the UV/TCI with imiquimodtreated group compared with TCI with imiquimod alone or the non-draining lymph nodes (Figure 5c). The majority of FITC þ DCs were CD11c þ CD11b þ Langerin and had high CD40 and CD86 expression (Figure 5d), compatible with an activated dermal DC phenotype (Heath and Carbone, 2009).

P Stein et al. UV Enhances Transcutaneous Immunization

Tumor size (mm2)

600

600

Untreated

400

400

200

200

0

UV alone NS

0 0

10

20

30

0

40

10

20

30

40

Tumor size (mm2)

600 TCI with imiquimod 400

600

UV/TCI with imiquimod

400

* 200

200

0

0 0

10 20 30 40 Days after tumor inoculation

0

30 40 10 20 Days after tumor inoculation

Immunizations None

100

UV Survival (%)

TCI UV-TCI 50

*

0 0

20 40 Days after tumor inoculation

60

Figure 4. Delayed tumor growth and superior tumor protection after UV-B exposure before transcutaneous immunization (TCI). EG.7 cells (4  105) were injected subcutaneously into the flank of mice. Immunization treatments were started after 10–12 days after inoculation when tumors became palpable. The mean tumor size with SEM (a) and survival (b) were monitored. Depicted are the cumulative results of four independent experiments with n ¼ 37 mice in the TCI group, n ¼ 33 in the UV/TCI group, n ¼ 33 in the UV group, and n ¼ 30 mice in the untreated control group. *A significant difference (Po0.05) between TCI and UV/TCI by Mantel–Cox test for survival analysis or by paired Student’s t-test for comparison of tumor growth; NS, not significant.

To address whether this CD11c þ DC population is relevant for CTL priming on TCI with imiquimod, we took advantage of mice transgenic for the diphtheria toxin receptor (DTR): one mouse strain expressing the DTR specifically in LC under the control of the Langerin promoter (Langerin-DTREGFP) or another mouse strain expressing the DTR in CD11c þ DCs (CD11c-DTR) under the control of the CD11c promoter. By the administration of diphtheria toxin (DT) the respective DC populations can be specifically ablated. Langerin-DTREGFP mice were immunized as before by imiquimod-based TCI with or without UV exposure. DT treatment resulted in a near-total depletion of epidermal Langerin þ cells as shown by immunohistology staining of epidermal sheets (not shown). The CTL activity was reduced

to about 50% after imiquimod-based TCI with or without UV exposure, indicating a significant contribution of these cells to the induced CTL response (Figure 5d). To examine the contribution of bone-marrow-derived Langerin þ or CD11c þ DC subsets more closely, we generated chimeric mice in which C57BL/6 wild-type mice were lethally irradiated and reconstituted with bone marrow cells from LangerinDTREGFP or CD11c-DTR transgenic mice. In these animals, radio-resistant LCs remain of host origin and not sensitive to DT depletion whereas radio-sensitive dermal DC populations are replaced by the transplanted DTR transgenic cells (Merad et al., 2002; Ginhoux et al., 2007). Subsequently the chimeric animals were treated with imiquimod-based TCI or UV/TCI after depletion of the respective DC population. In this www.jidonline.org

215

P Stein et al. UV Enhances Transcutaneous Immunization

CD11c+ cells in LN (abs. number)

*

*

100,000

50,000

0

Control TCI UV-B TCI

6,000

4,000

2,000

0

Control TCI UV-B TCI

Draining LN

Fold IL-12 induction

FITC+Langerin+ cells + Langerin cells in LN (abs. number)

FITC+CD11c+ DC

150,000

Nondraining LN

*

Draining LN

Draining LN Nondraining LN

*

3

Control TCI UV-B TCI

Control TCI UV-B TCI

Nondraining LN

Control

* 2

TCI

1

UV-B TCI

0 TCI

UV-B TCI

Langerin-DTREGFP Specific lysis (%)

80 60

**

40 20 0 Untreated TCI

Specific lysis (%)

100

100

Not depleted

TCI

60 40 20

Depleted

100

CD86

CD11c-DTR→ C57BL/6

80 60 40 20 0

0 UV-B TCI

CD40

Langerin-DTREGFP → C57BL/6

80

** UV-B TCI

CD11b

Specific lysis (%)

Control

Untreated TCI

UV-B TCI

Not depleted

TCI

UV-B TCI

Depleted

Untreated TCI

UV-B TCI

Not depleted

**

**

TCI

UV-B TCI

Depleted

Figure 5. CD11c þ , but not Langerin þ dendritic cells (DCs), are essential for cytotoxic T lymphocyte (CTL) responses induced by transcutaneous immunization (TCI) with imiquimod. (a, b) Mice were immunized as indicated after FITC painting of the back skin. Lymph nodes were collected after 48 hours and analyzed by flow cytometry. CD11c þ MHC II þ cells (median, quartiles, range) or Langerin þ MHC II þ cells are shown (gated on PICD19CD90 cells). (c) IL-12p35 mRNA induction by qRT-PCR in the draining (filled bars) and nondraining (open bars) lymph nodes after immunizations is shown. (d) The DC phenotype in the draining lymph node 48 hours after the indicated treatments is shown (gated on FITC þ CD11c þ MHC II þ cells). Langerin-DTREGFP mice (e), chimeric mice reconstituted with bone marrow from Langerin-DTREGFP (Langerin-DTREGFP-C57BL/6; f) or CD11c-DTR mice (CD11c-DTR-C57BL/6; g) were used for immunizations. The depicted in vivo cytolytic activity was assessed 7 days later. ‘‘Depleted’’ indicates intraperitoneal diphtheria toxin (DT) injections as described in the Materials and Methods section. The results are representative of at least two independent experiments performed with n ¼ 3 mice per group. *A difference (Po0.05) by Mann–Whitney U-test between the draining versus the nondraining lymph nodes. **A difference (Po0.05) by unpaired Student’s t-test between groups treated with DT or not.

situation, the depletion of bone-marrow-derived Langerin þ cells did not lead to any alteration of the CTL responses induced by either immunization protocol (Figure 5f), whereas depletion of CD11c þ cells resulted in the complete abrogation of the imiquimod-based TCI or UV/TCI-induced CTL activity (Figure 5g). These data suggest that the CTL activation on TCI with imiquimod is mediated primarily by bonemarrow-derived dermal DCs and that epidermal LCs may have an auxiliary role, but they are not essential for CTL priming in this setting. Other DC populations such as bonemarrow-derived Langerin þ DCs (Bursch et al., 2007) obviously have no functional role in TCI with imiquimodinduced CTL activation. DISCUSSION TCI might be an attractive way to deliver antigens for the induction of curative immune responses against malignancies or persistent virus infections by directly targeting skinresident APC. By combining the topical application of a TLR agonist with a CTL epitope, potent primary T cell 216

Journal of Investigative Dermatology (2011), Volume 131

responses are initiated (Rechtsteiner et al., 2005). However, we and others experienced that TCI with imiquimod-induced CTL responses rapidly fade away resulting in poor memory formation and only partial tumor protection (Itoh and Celis, 2005; Warger et al., 2007; Stoitzner et al., 2008). Because low-dose UV exposure has been shown to have a synergistic effect on TLR7-mediated activation of APC in vivo, we explored our imiquimod-based TCI protocol toward UV-induced additional effects. A previous report found no effect of UV exposure on cholera-toxin/tapestripping-based TCI with peptide (Ghoreishi and Dutz, 2006). However, we found that UV exposure before TCI with imiquimod not only prevented UV-mediated tolerance as suggested by previous results in a DTH model (Thatcher et al., 2006), but actually boosted the peptide-specific CTL response and also created a considerable memory response. Toward the identification of the APC populations involved and in line with a previous report (Thatcher et al., 2006), we detected an increased number of CD11c þ , but not Langerin þ , DCs in the draining lymph nodes on additional

P Stein et al. UV Enhances Transcutaneous Immunization

UV after TCI with imiquimod. To address the functional contribution this DC population to UV/TCI with imiquimod, we used transgenic mice that can be specifically ablated of distinct DC subsets and bone marrow chimeras of these to evaluate their origin. Using this powerful technique we can now attribute an important role for a dermal CD11c þ DC population for CTL priming after TCI, especially after additional UV exposure. We can exclude a significant contribution of radio-sensitive, bone-marrow-derived Langerin þ cells (Bursch et al., 2007; Ginhoux et al., 2007) that seem important in cholera-toxin-based TCI (Chang et al., 2008). The contribution of bone-marrow-derived CD11c þ dermal DCs is well known and distinct from the role of LCs in cross-priming and skin-associated immunity (Jung et al., 2002; Kissenpfennig et al., 2005). However, the activation of dermal DCs through TLR7 alone cannot explain the CTL superactivation upon additional UV exposure. Thatcher et al. (2006) show an increased production of IL12p70 of CD11c þ DC on imiquimod/UVB treatment. Consistent with this observation, we find an increased number of dermal DC and the induction of IL-12 in the draining lymph nodes on UV-B/imiquimod treatment. Therefore, we suggest that imiquimod/UV treatment induces the improved activation and migration of dermal DC to the regional lymph nodes where this converts to superior CTL responses, possibly mediated by IL-12. However, the exact molecular mechanisms how this superior DC activation is mediated remain to be elucidated. The biologic and biochemical processes after UV exposure are complex: potentially after UV exposure a mediator is released that synergizes with the TLR7 ligands in terms of DC activation, that is, urocanic acid (Walterscheid et al., 2006). This might also occur indirectly with a role for mast cells (Heib et al., 2007; Biggs et al., 2010) or Treg cells that have a well-described role in UV-induced immunosuppression (Schwarz et al., 2004; Maeda et al., 2008). It might be that Tregs directly or indirectly activated by UV become functionally inactive in the presence of the inflammatory milieu created by the TLR7 agonist and UV allowing an improved CTL activation. Taken together, we show the TLR agonist imiquimod can turn UV-induced immunosuppression into a robust and longlasting CTL response. This enhanced CTL response is able to reject established tumors expressing a target antigen. Therefore future vaccination protocols may want to incorporate not only activating adjuvants such as TLR agonist, but should also include means to inhibit or modify immunosuppressive factors. Our results provide some of the underlying mechanisms of TCI with imiquimod-based CTL responses and further extend the basis for new TCI approaches to treat cancer or persistent infections.

et al., 2005). All animal procedures were conducted in accordance with the institutional guidelines.

DTH responses C57BL/6 mice were sensitized by painting of DNFB (Sigma-Aldrich, Taufkirchen, Germany) solution (50 ml; 0.5% in acetone/olive oil, 4:1) on the shaved back. After 4 days, 20 ml of 0.2% DNFB was applied onto the left ear for elicitation. Ear swelling was quantified with a caliper 24 hours after challenge. DTH was determined as the amount of swelling of the hapten-challenged ear compared with the thickness of the vehicle-treated ear and expressed in centimeters  103 (mean±SD).

Transcutaneous immunizations Transcutaneous immunizations with imiquimod were described previously (Rechtsteiner et al., 2005). Briefly, dorsal hair was removed with electric clippers. Peptides (100 mg in DMSO from Sigma-Aldrich) were added to 50 mg cream containing 5% imiquimod (Aldara) or vehicle cream (both from 3M Pharmaceuticals, St Paul, MN) and applied on the dorsal skin of anesthetized mice on days 0 and 1. Synthetic peptide OVA257–264 (SIINFEKL) from chicken ovalbumin was kindly provided by Dr Stevanovic (Department Immunology, Institute for Cell Biology, University of Tu¨bingen, Germany). Where indicated the anti-CD40 (clone FGK-45, a gift from Dr R Offringa, Leiden, the Netherlands) antibody was additionally applied by daily i.v. injection of 100 mg antibody at days 0, 1, and 2 (Warger et al., 2007).

UV exposure In some experiments the animals were exposed to UV light at day 1 or as indicated. The shaved mice were anesthetized with ketamine/ xylazine (71.2:0.2 mg per mouse) and placed under the respective UV source at a distance of 25 cm from back skin. A narrow band UV lamp (365 nm, 8 W for UV-A; 312 nm, 8 W for UV-B; 254 nm, 8 W for UV-C, all from Peqlab, Erlangen, Germany) was used. UV-B lamp stably emitted light with 0.39 mW cm2 measured with a UV radiometer (UV34; PCE Group, Meschede, Germany). UVB doses from 30 to 300 mJ cm2 were calculated by the exposure times ranging from 10 to 300 seconds with respect to the respective light source.

Tumor rejection assay Tumors were implanted by s.c. injection of 4  105 EG.7 thymoma cells (from ATCC) as described previously (Warger et al., 2007). Tumors were allowed to grow until palpable (10–12 days after inoculation) before treatments were initiated. Tumor size was monitored three times per week with a caliper measuring the diameter in two dimensions. Mice were killed when tumor size exceeded 20 mm in both diameters.

Flow cytometric analyses and in vivo cytotoxicity assay

MATERIALS AND METHODS Mice C57BL/6 mice at 6–8 weeks were obtained from the local animal facility of the University of Mainz. CD11c-DTR mice are on C57BL/6 background and were provided by S Jung (Jung et al., 2002). Langerin-DTREGFP and Langerin-EGFP mice are on C57BL/6 background and were provided by B Malissen (Kissenpfennig

The following mAbs were used for analyses by flow cytometry: APC-Cy7-conjugated anti-CD8 (clone 53-6.7) and APC-conjugated CD44 (clone IM7) (both from eBioscience, Frankfurt, Germany), FITC-conjugated CD44 and PE-conjugated CD62L (clone MEL-14) (from ImmunoTools, Friesoythe, Germany), and APC-conjugated CD62L (from BD Pharmingen, Hamburg, Germany). Blood samples were collected after tail vein incision and incubated on ice with www.jidonline.org

217

P Stein et al. UV Enhances Transcutaneous Immunization

specific mAbs as indicated after a hypotonic lysis step. H2-KbSIINFEKL-specific T cells were detected by H2-Kb tetramer labeling as described previously (Warger et al., 2006). The assessment of in vivo cytolytic activity has been described (Rechtsteiner et al., 2005). For the evaluation of the cytolytic activity in the memory phase, target cells were transferred 30 days after immunization without prior restimulation and splenocytes were analyzed by flow cytometry 48 hours later. The expression of the target antigen on EG.7 cells was assessed by flow cytometry labeling using a mAb recognizing the OVA257–264 epitope in the context of H2-Kb (clone 25D1.16; Porgador et al., 1997). All analyses were performed with a FACSCanto or a LSRII Flow Cytometer and FACSDiva software (BD Pharmingen).

Preparation of lymph nodes Langerin-EGFP transgenic mice or C57BL/6 mice after FITC painting (400 ml, 5 mg ml1 in acetone/dibutyl phthalate (1:1) with 5% DMSO; all from Sigma-Aldrich) were used for these experiments as indicated. After the specified treatments, cervical, axillar, brachial, and inguinal lymph nodes were collected, picked, and digested for 45 minutes with type 2 collagenase (50 mg ml1 from Worthington Biochemicals, Lakewood, NJ). The reaction was stopped by adding 10 mM EDTA (from Sigma-Aldrich). The cells were labeled with the following mAbs: APC-Cy7-conjugated CD11c (clone N418), PE-conjugated CD11b (clone M1/70) and CD8 (clone 53-6.7), PE-Cy7-conjugated CD205 (clone 205yekta), FITC-conjugated B220 (clone RA3-6B2), APC-conjugated CD86 (clone GL-1) and CD40 (clone 1C10) (all from eBioscience); Pacific Blue MHC class II (clone M5/114.15.2) and PE-Cy7-conjugated CD4 (clone RM4-5) (both from BD Pharmingen). The cells were finally resuspended in Trucount tubes (from BD Pharmingen) allowing the absolute quantification according to the manufacturer’s instructions.

Detection of mRNA by qRT-PCR mRNA was extracted from lymph node cells from draining or nondraining lymph nodes 48 hours after the indicated immunization treatments using Trizol (Invitrogen, Karlsruhe, Germany). cDNA was synthesized with RevertAid M-MuLV reverse transcriptase according to the manufacturer’s instructions (MBI Fermentas, St Leon-Rot, Germany). Specific mRNA levels were quantified by RT–PCR using the following oligonucleotide primers: IL-12p35 (forward) GTCAAT CACGCTACCTCCTC; IL-12p35 (reverse) CTGCACAGCTCATCGAT GGC; HGPRT (forward) GTTGGATACAGGCCAGACTTTGTTG; HG PRT (reverse) GAGGGTAGGCTGGCCTATAGGCT. Oligonucleotides were chosen to span at least one intron at the level of genomic DNA. qRT-PCR analyses to quantify the expression of IL-12 and murine HGPRT were performed in triplicates on an iCycler (Bio-Rad, Munich, Germany) using the IQ SYBR Green Supermix (Bio-Rad). After normalization of the data according to the expression of HGPRT mRNA, the relative expression level IL-12 mRNA was calculated.

Generation of bone marrow chimeras Bone marrow chimeras were generated by lethal irradiation of recipient C57BL/6 mice with 9.5 Gy from Cs137 source (OB58-BA; Buchler, Braunschweig, Germany) and i.v. transfer of 5  105 bone marrow cells from the indicated donor strain. The animals were maintained under specific pathogen-free conditions for 8 weeks 218

Journal of Investigative Dermatology (2011), Volume 131

before use for the immunization experiments. Where indicated, CD11c þ cells from bone marrow chimeras or Langerin þ cells from Langerin-DTREGFP were depleted by i.p. injection of DT (300 ng; Sigma-Aldrich). Chimeras were injected from day 3 to day 4 every other day whereas Langerin-DTREGFP mice were injected at day 2.

Statistical analysis Different groups were analyzed by a two-tailed Student’s t-test or Mann-Whitney U-test for comparison between two groups as indicated. Survival analysis was performed by the Mantel–Cox test. For all analyses, Po0.05 was considered statistically significant. CONFLICT OF INTEREST The authors state no conflict of interest.

ACKNOWLEDGMENTS We thank Annekatrin Meinl for excellent technical assistance. This work was supported by the Deutsche Forschungsgemeinschaft (research grants KFO183 TP7 (to MPR and HS) and Sta 984/1-1 (to MS)). SUPPLEMENTARY MATERIAL Supplementary material is linked to the online version of the paper at http:// www.nature.com/jid

REFERENCES Banchereau J, Steinman RM (1998) Dendritic cells and the control of immunity. Nature 392:245–52 Bennett CL, van Rijn E, Jung S et al. (2005) Inducible ablation of mouse Langerhans cells diminishes but fails to abrogate contact hypersensitivity. J Cell Biol 169:569–76 Biggs L, Yu C, Fedoric B et al. (2010) Evidence that vitamin D3 promotes mast cell-dependent reduction of chronic UVB-induced skin pathology in mice. J Exp Med 207:455–63 Bursch LS, Wang L, Igyarto B et al. (2007) Identification of a novel population of Langerin+ dendritic cells. J Exp Med 204:3147–56 Chang SY, Cha HR, Igarashi O et al. (2008) Cutting edge: Langerin+ dendritic cells in the mesenteric lymph node set the stage for skin and gut immune system cross-talk. J Immunol 180:4361–5 Cooper KD, Baron ED, Matsui MS (2003) Implications of UV-induced inflammation and immunomodulation. Cutis 72:11–5 Fisher MS, Kripke ML (1982) Suppressor T lymphocytes control the development of primary skin cancers in ultraviolet-irradiated mice. Science 216:1133–4 Ghoreishi M, Dutz JP (2006) Tolerance induction by transcutaneous immunization through ultraviolet-irradiated skin is transferable through CD4+CD25+ T regulatory cells and is dependent on host-derived IL-10. J Immunol 176:2635–44 Ginhoux F, Collin MP, Bogunovic M et al. (2007) Blood-derived dermal Langerin+ dendritic cells survey the skin in the steady state. J Exp Med 204:3133–46 Halliday GM, Lyons JG (2008) Inflammatory doses of UV may not be necessary for skin carcinogenesis. Photochem Photobiol 84:272–83 Heath WR, Carbone FR (2009) Dendritic cell subsets in primary and secondary T cell responses at body surfaces. Nat Immunol 10:1237–44 Heib V, Becker M, Warger T et al. (2007) Mast cells are crucial for early inflammation, migration of Langerhans cells, and CTL responses following topical application of TLR7 ligand in mice. Blood 110:946–53 Hemmi H, Kaisho T, Takeuchi O et al. (2002) Small anti-viral compounds activate immune cells via the TLR7 MyD88-dependent signaling pathway. Nat Immunol 3:196–200

P Stein et al. UV Enhances Transcutaneous Immunization

Itoh T, Celis E (2005) Transcutaneous immunization with cytotoxic T-cell peptide epitopes provides effective antitumor immunity in mice. J Immunother 28:430–7

Rechtsteiner G, Warger T, Osterloh P et al. (2005) Cutting edge: priming of CTL by transcutaneous peptide immunization with imiquimod. J Immunol 174:2476–80

Jung S, Unutmaz D, Wong P et al. (2002) In vivo depletion of CD11c(+) dendritic cells abrogates priming of CD8(+) T cells by exogenous cell-associated antigens. Immunity 17:211–20

Romani N, Koide S, Crowley M et al. (1989) Presentation of exogenous protein antigens by dendritic cells to T cell clones. Intact protein is presented best by immature, epidermal Langerhans cells. J Exp Med 169:1169–78

Kissenpfennig A, Henri S, Dubois B et al. (2005) Dynamics and function of Langerhans cells in vivo: dermal dendritic cells colonize lymph node areas distinct from slower migrating Langerhans cells. Immunity 22:643–54

Schwarz A, Maeda A, Wild MK et al. (2004) Ultraviolet radiation-induced regulatory T cells not only inhibit the induction but can suppress the effector phase of contact hypersensitivity. J Immunol 172:1036–43

Maeda A, Beissert S, Schwarz T et al. (2008) Phenotypic and functional characterization of ultraviolet radiation-induced regulatory T cells. J Immunol 180:3065–71

Seder RA, Ahmed R (2003) Similarities and differences in CD4+ and CD8+ effector and memory T cell generation. Nat Immunol 4:835–42

Merad M, Manz MG, Karsunky H et al. (2002) Langerhans cells renew in the skin throughout life under steady-state conditions. Nat Immunol 3:1135–41 Navi D, Huntley A (2004) Imiquimod 5 percent cream and the treatment of cutaneous malignancy. Dermatol Online J 10:4 Palamara F, Meindl S, Holcmann M et al. (2004) Identification and characterization of pDC-like cells in normal mouse skin and melanomas treated with imiquimod. J Immunol 173:3051–61 Porgador A, Yewdell JW, Deng Y et al. (1997) Localization, quantitation, and in situ detection of specific peptide-MHC class I complexes using a monoclonal antibody. Immunity 6:715–26 Poulin LF, Henri S, de Bovis B et al. (2007) The dermis contains langerin+ dendritic cells that develop and function independently of epidermal Langerhans cells. J Exp Med 204:3119–31 Rana S, Byrne SN, MacDonald LJ et al. (2008) Ultraviolet B suppresses immunity by inhibiting effector and memory T cells. Am J Pathol 172:993–1004

Stoitzner P, Green LK, Jung JY et al. (2008) Tumor immunotherapy by epicutaneous immunization requires Langerhans cells. J Immunol 180:1991–8 Thatcher TH, Luzina I, Fishelevich R et al. (2006) Topical imiquimod treatment prevents UV-light induced loss of contact hypersensitivity and immune tolerance. J Invest Dermatol 126:821–31 Thomsen LL, Topley P, Daly MG et al. (2004) Imiquimod and resiquimod in a mouse model: adjuvants for DNA vaccination by particle-mediated immunotherapeutic delivery. Vaccine 22:1799–809 Walterscheid JP, Nghiem DX, Kazimi N et al. (2006) cis-Urocanic acid, a sunlight-induced immunosuppressive factor, activates immune suppression via the 5-HT2A receptor. Proc Natl Acad Sci USA 103:17420–5 Warger T, Hilf N, Rechtsteiner G et al. (2006) Interaction of TLR2 and TLR4 ligands with the N-terminal domain of Gp96 amplifies innate and adaptive immune responses. J Biol Chem 281:22545–53 Warger T, Rechtsteiner G, Schmid B et al. (2007) Transcutaneous immunization with imiquimod is amplified by CD40 ligation and results in sustained cytotoxic T-lymphocyte activation and tumor protection. Clin Rev Allergy Immunol 32:57–66

www.jidonline.org

219