Cyclooxygenase-2 Inhibition Promotes Enhancement of ... - Core

ORIGINAL ARTICLE

Cyclooxygenase-2 Inhibition Promotes Enhancement of Antitumor Responses by Transcutaneous Vaccination with Cytosine-Phosphate-GuanosineOligodeoxynucleotides and Model Tumor Antigen Joe Inoue1 and Yukihiko Aramaki1 One of the principal goals in tumor immune prophylaxis and tumor therapy is the induction of antitumor responses by generating sufficient numbers of tumor antigen-specific helper T (Th)1 cells and cytotoxic T lymphocytes (CTLs). We have demonstrated that the administration of cytosine-phosphate-guanosineoligodeoxynucleotide (CpG-ODN) through tape-stripped skin induced a Th1-type immune response and suggested that the skin is a potential site for vaccination. CpG-ODN induces the expression of cyclooxygenase (COX)-2, and its product prostaglandin (PG) E2 underlies an immunosuppressive network, therefore it is a simple strategy to use a COX-2 inhibitor for tumor vaccination with CpG-ODN. In this study, we examined whether a COX-2 inhibitor enhances the antitumor immune response induced by CpG-ODN with model tumor antigen, ovalbumin (OVA), applied to tape-stripped skin in mice. The COX-2 inhibitor remarkably enhanced antigen-specific Th1-type immune responses and generation of CTLs induced by transcutaneous vaccination with CpG-ODN and OVA. PGE2 and IL-10 levels in the skin were significantly decreased and production of IL-12 was enhanced. This vaccination also induces an effective antitumor immunity in tumor-challenged mice. These results suggested that transcutaneous vaccination with a COX-2 inhibitor, CpG-ODN, and tumor antigen is a very simple and cost-effective strategy for tumor vaccine and may be readily achievable. Journal of Investigative Dermatology (2007) 127, 614–621. doi:10.1038/sj.jid.5700656; published online 7 December 2006

INTRODUCTION CD4 þ helper T (Th) cells are subdivided into at least two subsets, Th1 and Th2, on the basis of the cytokines they secrete (Fiorentino et al., 1989; Mosmann and Coffman, 1989). Th1-like immune responses enhance cellular immunity in association with increased levels of Th1-like cytokines such as IL-2 and IFN-g and upregulated activity of CTLs, indicating that the suppression of Th1 cells and CTLs can result in an acceleration of tumor growth (Tuttle et al., 1993). One of the principal goals in tumor immune prophylaxis and tumor therapy is the induction of antitumor responses by generating sufficient numbers of tumor antigen-specific Th1 cells and cytotoxic T lymphocytes (CTLs). Therefore, it is very 1

School of Pharmacy, Tokyo University of Pharmacy and Life Science, Hachioji, Tokyo, Japan Correspondence: Dr Yukihiko Aramaki, School of Pharmacy, Tokyo University of Pharmacy and Life Science, 1432-1 Horinouchi, Hachioji, Tokyo 192-0392, Japan. E-mail: [email protected] Abbreviations: APC, antigen-presenting cell; CpG-ODN, cytosinephosphate-guanosine-oligodeoxynucleotide; COX, cyclooxygenase; CTL, cytotoxic T lymphocyte; DC, dendritic cell; LC, langerhans cell; mRNA, messenger RNA; NK, natural killer; PBS, phosphate-buffered saline; Th, helper T Received 17 February 2006; revised 12 September 2006; accepted 11 October 2006; published online 7 December 2006

614

Journal of Investigative Dermatology (2007), Volume 127

important for tumor immune therapy to establish efficient Th1 and CTL priming strategies. Unmethylated CpG dinucleotides flanked by certain bases (CpG-motif), which are present in bacterial DNA, have been shown to be immunostimulatory (Krieg, 2002). Previous studies have found that both the bacterial CpG-motif and synthetic oligodeoxynucleotides containing a CpG-motif (CpG-ODN) activate cells such as B cells, macrophages, and dendritic cells (DC), through toll-like receptor-9 (Krieg, 2002; Agrawal and Kandimalla, 2003). Signaling through toll-like receptor-9 leads to the secretion of large amounts of proinflammatory cytokines such as type-I IFN, IL-1b, IL-6, tumor necrosis factor-a, and IL-12 (Hemmi et al., 2003; Spies et al., 2003). IL-12 acts on T and natural killer (NK) cells inducing the production of cytokines, primarily IFN-g, enhancing NK cell cytotoxic activity. CpG-ODN also induces the production of great numbers of CTLs which have antitumor effects (Davis et al., 1998; Miconnet et al., 2002). Consequently, CpG-ODN could be effective as an adjuvant for cellular and humoral immunities (Klinman et al., 2004). Cyclooxygenase (COX)-2 and its product prostaglandin (PG) E2 underlie an immunosuppressive network that is important in tumor therapy and vaccination (Howe et al., 2005; Sharma et al., 2005). CpG-ODN induces COX-2 expression and PGE2 production in DCs and macrophages & 2006 The Society for Investigative Dermatology

J Inoue and Y Aramaki COX-2 Inhibitor Enhances Antitumor Immunity by CpG-ODN

(Chen et al., 2001; Ghosh et al., 2001). PGE2 is a potent inhibitor of Th1-type immune responses (Katamura et al., 1995), inhibiting IFN-g production as well as IL-12 and IL-12R expression (Abe et al., 1997; Wu et al., 1998), and is also a potent inducer of production of the immunosuppressive cytokine IL-10 (Harizi et al., 2002). Furthermore, IL-10 suppresses the production IL-12 by antigen-presenting cells (APCs), therefore APCs from IL-10 knockout mice are induced to produce a large quantity of IL-12 by CpG-ODN (Yi et al., 2002). Thus it is very simple strategy to use a COX-2 inhibitor for tumor vaccination with CpG-ODN. The skin has recently become a target site in the development of non-invasive vaccine technologies (Beignon et al., 2002; Klimuk et al., 2004). We have reported that the administration of CpG-ODN through tape-stripped skin induced a Th1-type immune response and suggested that the skin is a potential site for vaccination (Inoue et al., 2005a, b). It is also favorable that APCs, especially langerhans cells (LCs) and dermal DCs, are common in the skin and easy to target. In this study, we examined whether a COX-2 inhibitor enhances antitumor immune responses induced by CpGODN with model tumor antigen ovalbumin (OVA), applied to tape-stripped skin in mice. The COX-2 inhibitor remarkably enhanced the Th1-type immune response and generation of CTLs, and a transcutaneous vaccination with the inhibitor, CpG-ODN and OVA was effective in inducing antigen-specific antitumor immunity in vivo. These results suggested that transcutaneous vaccination with a COX-2 inhibitor, CpG-ODN, and tumor antigen is an effective approach. RESULTS COX-2 inhibitor enhances CpG-ODN-induced Th1-dominant immune responses on transcutaneous vaccination

CpG-ODN induces COX-2 expression and its product PGE2 underlies an immunosuppressive network (Howe et al., 2005;

Sharma et al., 2005). Therefore it is a very simple strategy to use a COX-2 inhibitor to enhance the effect of CpG-ODN. First we examined whether the COX-2 inhibitor NS3-398, enhances the antigen-specific Th1-dominant immune response induced by CpG-ODN, applied to tape-stripped skin in mice. Spleen cells were prepared on day 7 from the last vaccination, and were incubated in the presence of 2 mg/ml of OVA for 24 hours. Supernatants were collected and concentrations of IFN-g were evaluated by ELISA. As shown in Figure 1a, the production of IFN-g increased on the coapplication of OVA and CpG-ODN, and upregulation of IFNg production was observed in mice, which were treated with the COX-2 inhibitor (NS-398). The co-application of OVA and NS-398 did not upregulate the IFN-g production. Furthermore, the vaccination with control antigen (BSA) did not induce the IFN-g production (Figure 1b). These results suggested that antigen-specific immune responses are induced and antigen-nonspecific immune responses, such as NK and NK T cell-mediated IFN-g production, are not shown in this experiment. Next we examined the production of OVA-specific IgG2a antibody, which is considered to be a Th1-like Ig isotype. OVA-specific IgG2a production increased drastically when NS-398 was applied with CpGODN from the skin (Figure 1c). These findings suggested that co-application of CpG-ODN with a COX-2 inhibitor to tapestripped skin leads to enhanced Th1-dominant immune responses. Effect of vaccination route on IFN-c production

Vaccines consisting of CpG-ODN and OVA, with or without NS-398, were administered by different routes, intravenously (i.v.), intraperitoneally (i.p.), or transcutaneously, and the level of production of IFN-g was determined. Seven days after the last vaccination, the spleen cells were prepared, and IFN-g concentrations in the culture supernatant were evaluated by ELISA. In each case, levels of IFN-g were

2,500

2,500

1.25

2,000 1,500 1,000

Abs (450 nm)

c 1.50

IFN- (pg/ml)

b 3,000

2,000 1,500 1,000

0.75 0.50

CpG+BSA +NS398

CpG+BSA

BSA

0

Control

0

CpG+OVA +NS398

0

CpG+OVA

0.25

OVA+NS398

500 OVA

cont OVA CpG+OVA+buffer CpG+OVA NS398+CpG+OVA

1.00

500 Control

IFN- (pg/ml)

a 3,000

7

8

9

10

11

12 (n )

Dilution (1/2n )

Figure 1. A COX-2 inhibitor enhances CpG-ODN-induced Th1-dominant immune responses following a transcutaneous vaccination. (a, b) NS-398 (25 mg), CpG-ODN (50 mg), and OVA or BSA (50 mg) were applied to shaved abdominal skin after barrier disruption by tape stripping (six times). BSA was used for control antigen. Spleen cells (5 105 per well) were prepared 7 days after the last vaccination, and incubated in the presence of OVA (2 mg/ml) for 24 hours. The concentration of IFN-g in the culture supernatant was evaluated by ELISA. OVA (50 mg) (m), NS-398 (25 mg) plus OVA (50 mg) plus CpG-ODN (50 mg) (J), buffer plus OVA (50 mg) plus CpG-ODN (50 mg) (K), and OVA (50 mg) plus CpG-ODN (50 mg) (’) were applied to the tape-stripped skin. Sera were collected 2 weeks after the last vaccination and the OVA-specific IgG2a titer was evaluated by ELISA (c). Each value represents the mean7SD for five mice. *Po0.05 compared with OVA plus CpG-ODN. #Po0.05 compared with OVA.

www.jidonline.org

615


increased in CpG-ODN plus OVA-treated mice, and enormous amount of IFN-g were detected following an i.v. vaccination (Figure 2). On the other hand, in the case of vaccination with NS-398, upregulation of IFN-g production was observed when the vaccine was administered from the skin but not i.p. and i.v., and the level of IFN-g was the same as on the i.v. administration of CpG-ODN and OVA (Figure 2). No enhancement of IFN-g production was observed when NS398 was administered i.p. or i.v. These results suggest that skin is the potentially the best site for the induction of a Th1dominant immune response by a COX-2 inhibitor. Effect of COX-2 inhibitor on cytokine production in the skin

Next we examined the mechanisms of the enhancement of Th1-type immune responses induced by the transcutaneous vaccination with the COX-2 inhibitor. It is reported that CpGODN promotes Th1-like immune responses with the production of PGE2 and IL-10, which are considered to be suppressive 3,000

NS398+CpG+OVA CpG+OVA

IFN- (pg/ml)

2,500 2,000 1,500 1,000 5,00 0 Control

Skin

I.p

I.v

Figure 2. The COX-2 inhibitor enhances the effect of the transcutaneous vaccination but not the i.p. or i.v. vaccination. NS-398 (25 mg), OVA (50 mg), and CpG-ODN (50 mg) were applied transcutaneously, i.p. or i.v. Spleen cells (5 105 per well) were prepared 7 days after the last vaccination, and incubated in the presence of OVA (2 mg/ml) for 24 hours. The concentration of IFN-g in the culture supernatant was evaluated by ELISA. Each value represents the mean7SD for three mice. *Po0.05, **Po0.01 compared with OVA plus CpG-ODN. Similar results were obtained in three different experiments.

a 2,000

b 2,000

1,500

1,500

mediators in vivo (Chen et al., 2001). Thus, we first examined the changes in cytokine levels in the skin 12 hours after the transcutaneous vaccination. IL-12, IL-10, and PGE2 levels increased significantly when OVA was co-applied with CpGODN (Figure 3). On the other hand, the application of NS-398 drastically increased IL-12 production and decreased significantly IL-10 and PGE2 production in the skin (Figure 3). These findings suggest that NS-398 suppresses the production of PGE2 and IL-10 in the skin and this suppression induces the production of a large quantity of IL-12. Effect of COX-2 inhibitor on migration of APC and mRNA expression in draining lymph nodes

It is reported that the number of allergen-bearing LCs in the LN was significantly greater in IL-10 knockout mice than in wild-type mice, and these mice demonstrated a greater increase in production of tumor necrosis factor-a and IL-1b which induce the migration of LCs to the LN in the allergenexposed epidermis (Wang et al., 1999). So we next examined the effect of a COX-2 inhibitor on the migration of APCs to draining LNs. NS-398, CpG-ODN, and Alexa-488-conjugated OVA were applied to tape-stripped skin, and the draining LNs were removed 24 hours later. LN single cells were treated with anti-major histocompatibility complex (MHC) class II mAb or anti-CD11c mAb and then doublepositive cells were analyzed on FACScan. As shown in Figure 4a, application of CpG-ODN to the tape-stripped skin significantly enhanced the migration of Alexa-488 and MHC class II double-positive cells, but any significant differences between CpG-ODN and CpG-ODN plus NS398 was not observed. In case of CD11c analysis, the migration of Alexa-488 and CD11c double-positive cells to draining LNs were significantly enhanced when NS-398 was co-applied (Figure 4a). These results suggest that the coapplication with NS-398 especially enhances the migration of antigen-bearing-CD11c-positive cells to the draining LNs. Next we examined the effect of these APCs on T-cell differentiation in the draining LN. NS-398, CpG-ODN and OVA were applied to the tape-stripped skin, then draining LNs were removed, and the cytokine mRNA expression in the

c 30

1,000 500

PEG2 (ng /ml)

IL-10 (pg/ml)

IL-12 (pg /ml)

25

1,000

20 15 10

500

CpG+OVA +NS398

CpG+OVA

OVA

CpG+OVA +NS398

CpG+OVA

0 OVA

0

Control

CpG+OVA +NS398

CpG+OVA

OVA

Control

0

Control

5

Figure 3. Effect of the COX-2 inhibitor on IL-12, IL-10, and PGE2 production in the skin. NS-398 (25 mg), OVA (50 mg), and CpG-ODN (50 mg) were applied to the skin, and the levels of cytokines in the skin were determined 12 hours later. IL-12, IL-10, and PGE2 concentrations were evaluated by ELISA. Each value represents the mean7SD for three mice. *Po0.05 compared with OVA plus CpG-ODN. #Po0.05 compared with OVA.

616



a

b

IFN-

Control IL-12 OVA IL-10 CpG+OVA -Actin

CpG+OVA +NS398 15

Relative intensity/-actin

OVA CpG+OVA CpG+OVA +NS398

0

0.5

1

1.5

Alexa488+ CD11+ cells (%)

2

IFN-

0 1

IL-12

0 1

IL-10

0 Control

Control

1

CpG+OVA +NS398

10

CpG+OVA

5

Alexa488+ MHC-classII+ cells (%)

OVA

0

Figure 4. Effect of the COX-2 inhibitor on skin APC migration and mRNA expression in draining lymph nodes. (a) Twenty-four hours after the application of NS-398 CpG-ODN, Alexa-488-conjugated OVA, or PBS (control) to the skin, the draining lymph nodes were removed. Lymph node single cells (106) were incubated with phycoerythrin-conjugated anti-MHC class II mAb or phycoerythrin-conjugated anti-CD11c mAb and two-color fluorescence was analyzed with FACScan. Following the application of NS-398, CpG-ODN, and OVA to the skin, the draining lymph nodes were removed 5 days after the last vaccination. The cytokine mRNA expression in lymph nodes was determined by RT-PCR and normalized by b-actin levels (b). Each value represents the mean7SD for three mice. *Po0.05 compared with OVA plus CpG-ODN. #Po0.05 compared with OVA. Similar results were obtained in three different experiments.

LN was determined by RT-PCR. As shown in Figure 4b, IL-10 mRNA expression induced by CpG-ODN treatment was clearly reduced by NS-398. On the other hand, IL-12 mRNA was clearly expressed in the LNs of NS-398-treated mice (Figure 4b). These findings suggest that the migration of LCs and dermal DCs which acquired the ability to produce IL-12 by NS-398 and CpG-ODN treatment, or direct delivery of these drugs induce highly expression of IFN-g mRNA in the draining LN. It is reported that IL-12 induces Th1 differentiation (Hsieh et al., 1993; Trinchieri, 1993), and our experiment shows that the IFN-g mRNA expression induced mainly by Th1 in the LN was upregulated on the application of NS398 (Figure 4b). COX-2 inhibitor enhances CTL induction by CpG-ODN

It is reported that treatment with CpG-ODN generates great numbers of CTLs which have antitumor effects (Davis et al., 1998; Miconnet et al., 2002). So we next examined whether the COX-2 inhibitor NS-398 enhances the generation and activity of CTLs induced by CpG-ODN. Splenocytes from vaccinated mice were stimulated with mitomycin C-treated E.G7-OVA cells and cultured for 5 days. CTLs were collected and incubated with E.G7-OVA cells for 4 hours and the amount of lactate dehydrogenase released was measured. As shown in Figure 5, specific cell lysis was observed in splenocytes prepared from CpG-ODN-treated mice and this CTL activity was significantly increased in CpG-ODN plus NS-398-treated mice. These results suggest that NS-398

enhances not only Th1-type immune responses but also CTLs activities. Effect of COX-2 inhibitor on tumor growth and survival

It is very important for tumor immune prophylaxis and tumor therapy to induce antitumor responses by generating sufficient numbers of tumor antigen-specific Th1 cells and CTLs. We examined the effect of a COX-2 inhibitor on the generation of antigen-specific Th1 cells and CTLs by CpGODN and suggested that vaccination with such an inhibitor from the skin increased the numbers and activity of these cells. Subsequently, we examined the tumor immune prophylaxis on transcutaneous vaccination with the COX-2 inhibitor. Following the transcutaneous vaccination, mice were inoculated s.c. with E.G7-OVA tumor cells, and subsequent tumor growth was assessed by measuring tumor volume. Survival rate was calculated by counting surviving mice. As shown in Figure 6a, tumor growth was inhibited in CpG-ODN-treated mice and this inhibitory effect was significantly enhanced when mice were treated with NS398. Furthermore, 60% of NS-398-treated mice survived 100 days, but only 10% of the mice treated with CpG-ODN only survived (Figure 6b). DISCUSSION Vaccination is one of the most cost-effective ways to prevent and control several diseases and there is still a great need to develop a new generation of safer vaccines that can www.jidonline.org

617


be effectively administered by simple, economical, and practical immunization procedures. Recently, there has been a lot of interest in the potential of non-invasive routes, such as via the skin, for tumor vaccine delivery (Seo et al., 2000; Takigawa et al., 2001; Bins et al., 2005). Because the skin-associated lymphoid tissue, comprised of powerful APCs such as LCs and dermal DCs, re-circulating T cells, and the regional LNs, ensures the efficient presentation of tumor antigen to immunocompetent cells and induction of strong immune responses. We have recently reported that the administration of CpG-ODN through tape-stripped skin induced an antigen-specific Th1-type immune 80

cont OVA CpG+OVA NS398+CpG+OVA

Lysis (%)

60 40 20 0 0 40 20 Ratio (effector/target) Figure 5. The COX-2 inhibitor enhances the effect of CpG-ODN on CTLs. NS-398 (25 mg), OVA (50 mg), and CpG-ODN (50 mg) were applied to the skin, and spleen cells were prepared 7 days after the last vaccination. To generate CTLs, splenocytes were stimulated with mitomycin C-treated E.G7-OVA cells and cultured for 5 days. Then, CTLs and E.G7-OVA cells were incubated for 4 hours at 371C. Supernatants were collected and the amount of lactate dehydrogenase released was measured in triplicates. The specific lysis of target cells was quantified using the formula (experimentaleffector spontaneoustarget spontaneous)/(target maximumtarget spontaneous) 100. Each value represents the mean7SD for three mice. *Po0.05 compared with OVA plus CpG-ODN. #Po0.05 compared with OVA. Similar results were obtained in three different experiments.

a

response and suggested that the skin is a potential site for vaccination (Inoue et al., 2005a, b). In this study, we tried to establish efficient Th1 and CTL priming strategies against tumors, using CpG-ODN and a COX-2 inhibitor applied to the skin. Our examination showed that the skin is the best route of administration for efficient Th1 and CTL activity following the application of CpG-ODN and the COX-2 inhibitor, because the production of IFN-g by splenocytes was remarkably enhanced following application to the skin but not after i.p. and i.v. injections (Figure 2). In our examination, the COX-2 inhibitor strongly suppressed T-cell function in a dosedependent manner, but enhanced the production of IL-12 by APCs stimulated with CpG-ODN in vitro (data not shown). It is suggested that the administration of a COX-2 inhibitor i.p. and i.v. made it hard to deliver the drug to APCs. Thus, no enhancement of IFN-g production was observed when the COX-2 inhibitor was administrated i.p. and i.v. (Figure 2). On the other hand, on transcutaneous application, the COX-2 inhibitor may be localized to the skin because of the hydrophobicity of the drug. It is favorable that LCs and dermal DCs exist in the skin and the drug may easily interact with these cells. Indeed, APCs in the skin, especially LCs and dermal DCs were known to produce IL-12 in response to CpGODN, and this ability may be upregulated when NS-398 is applied with CpG-ODN (Figure 3). Furthermore, the vaccination with NS-398 enhances the migration of antigen-bearingCD11c-positive cells to the draining LNs (Figure 4a). This response may be explained by the direct activation of CD11cpositive cells, especially LCs, by CpG-ODN with NS-398. And this upregulation may be the one of the mechanisms of the effective induction and generation of Th1 and CTLs by transcuraneous vaccination with NS-398. Thus, transcutaneous vaccination may be useful to bring out the potential for vaccination. However, the enhancement of the migration of antigen-bearing-MHC class II-positive cells was not observed

b 120 100 3,000

Survival (%)

Tumor volume (mm3)

4,000

2,000

80 60 40

1,000 20 0 0

5 10 15 20 25 30 35 40 Days post tumor injection Cont OVA

0 20

40 60 80 100 Days post tumor injection

120

CpG+OVA NS398+CpG+OVA

Figure 6. Effect of the COX-2 inhibitor on tumor growth and survival in mice vaccinated with CpG-ODN. (a, b) OVA (50 mg) (m), NS-398 (25 mg) plus OVA (50 mg) plus CpG-ODN (50 mg) (J), and OVA (50 mg) plus CpG-ODN (50 mg) (K) were applied to the skin. One week after the last vaccination, mice were inoculated s.c. with E.G7-OVA tumor cells (5 106 cells/mouse), and (a) subsequent tumor growth was assessed by taking an average of tumor diameters measured with calipers and calculating tumor volume with the formula: tumor volume ¼ (length) (width)2/2. (b) The rate of survival was determined by counting surviving mice. Each value represents the mean7SD for ten mice. *Po0.05 compared with OVA plus CpG-ODN. Similar results were obtained in two different experiments.

618



even when vaccinated with NS-398. This response may come from the difference in susceptibility of these cell populations to NS-398. Our findings suggested that transcutaneous vaccination with a COX-2 inhibitor induces efficient Th1 priming. Th1dominant immunity plays a critical role in the induction of antitumor immunity multiple helper functions such as the capacity to enhance CTL priming of APCs and to promote migration, proliferation, and cytotoxicity of tumor-specific CTLs (Ridge et al., 1998; Marzo et al., 2000; Giuntoli et al., 2002; Yu et al., 2003). Chamoto et al. (2004) reported that adaptive transfer of Th1 induces effective antitumor immunity. It is also reported that LCs and dermal DCs have an effect on antigen cross-presentation and cross-priming to generate CTLs (Ackerman and Cresswell, 2004; Heath et al., 2004; Matsuo et al., 2004), and CpG-ODN enhances this effect dependent on IFN-g, type-I IFN production, and MHC class-I expression in these APCs (Cho et al., 2002; Kuchtey et al., 2005). PGE2 exhibited dose-dependent inhibition of these functions of APCs, therefore we expected an enhancement of CTL induction on transcutaneous vaccination with a COX-2 inhibitor which induces strong Th1 immune responses. Transcutaneous vaccination with CpG-ODN induced antigen-specific tumor cell lysis and the COX-2 inhibitor remarkably enhanced this effect (Figure 5). These results suggested that LCs and dermal DCs in COX-2 inhibitor-treated mice changed the production of PGE2 and cytokines, and also MHC class-I expression. These APCs migrated the LN and may assist CD8 þ T cells to tumor antigen-specific CTLs. Furthermore, it is very important for tumor immune prophylaxis and tumor therapy to induce antitumor responses by tumor antigen-specific Th1 cells and CTLs. We demonstrated here that a COX-2 inhibitor enhances the generation of antigen-specific Th1 cells and CTLs by CpG-ODN, and transcutaneous vaccination with the COX-2 inhibitor was effective in terms of tumor immune prophylaxis, tumor growth inhibition, and survival (Figure 6). We are now investigating the effect of this transcutaneous vaccination for tumor therapy model, vaccinated after the tumor inoculation, focus on the effect of co-application with NS-398 in the generation and activation of Th1 cells, CTLs, and NK cells in tumor-bearing mice. CpG-ODNs are known to stimulate strong NK cell responses, but in our experiments, NK cell-mediated IFN-g production was not observed (Figure 1). It is because that the activation of NK cell by CpG-ODNs were very early responses. In fact, systemically injection of CpG-ODNs induces the activation of NK and NK T cell in 4 hours after the injection (Suzuki et al., 2004). On the other hands, it is reported that activation of NK and NK T cell by CpG-ODN was an indirect effect mediated by activated APCs, and this response was shown in a week after application of CpG-ODN. And then, the contribution of NK and NK T cells in enhancement of antitumor immunity by CpG-ODN is not yet elucidated, and additional studies are needed. Repeated CpG-ODN administration through the i.p. route induces strongly reduced primary humoral immune

responses and immunoglobulin class switching with multifocal liver necrosis and hemorrhagic ascites as a side effect (Heikenwalder et al., 2004). All these side effects were strictly dependent on the CpG-motif and toll-like receptor-9, as neither the CpG-ODN treatment of toll-like receptor-9 knockout mice nor the repetitive challenge of wild-type mice with non-stimulatory ODN was immunotoxic or hepatotoxic. We succeeded in inducing strong antitumor immunity with a transcutaneous vaccination. We expect the transcutaneous route to facilitate the targeting of APCs and reduce the distribution of drugs in the whole body compared to the i.p. or i.v. route. Therefore, we consider that the skin is the most effective site for vaccination with CpG-ODN despite these adverse effects, but additional study is needed. All these findings suggest that transcutaneous vaccination with a COX-2 inhibitor is a very simple and effective way to induce antitumor immunity and that such a vaccine is noninvasive and effective. In addition, COX-2 inhibitors are commonly used in clinical medicine, therefore this strategy may be readily achievable when the use of CpG-ODN is authorized. MATERIALS AND METHODS Animals and tumors Specific pathogen-free C57BL/6 mice (male, 8–12 weeks old) were purchased from SLC Co. Ltd. (Shizuoka, Japan). E.G7-OVA cells (transfected with EL4 to express OVA) (Carbone and Bevan, 1990) were obtained from the ATCC (Manassas, VA) and cultured in complete medium consisting of Rosewell Park Memorial Institute1640 medium supplemented with 10% fetal calf serum, 100 U/ml penicillin, 100 mg/ml streptomycin, 2 mM L-glutamine, and 5 105 M 2-mercaptoethanol.

Reagents The sequence for CpG-ODN was 50 -TCCATGACGTTCCTGATG CT-30 , and HPLC-purified phosphorothioate ODN was obtained from Sigma Genosys (Tokyo, Japan). The COX-2 inhibitor (NS-398) was obtained from Sigma (Tokyo, Japan) and stocked in DMSO (5 mg/ml).

Animal experiment and cytotoxicity assay Animal use and relevant experimental procedures were approved by the Tokyo University of Pharmacy and Life Science Committee on the Care and Use of Laboratory Animals. Mice were anesthetized with an i.p. injection of Nembutal (1.5 mg/mouse), the abdominal skin was shaved, and then the stratum corneum was removed by stripping with adhesive tape (Nichiban, Japan) six times. NS-398 (25 mg/10 ml; (dissolved in DMSO:PBS (phosphate-buffered saline) ¼ 1:1)) was applied onto the tape-stripped skin, and after 15 minutes CpG-ODN (50 mg) and OVA (50 mg) in 50 ml of PBS solution were applied. To boost the response, the same procedure was applied to the tape-stripped skin 1 week later. Mice were sacrificed at day 14. CTL activity was evaluated by the method used to generate CTL. In brief, 5 106/ml of splenocytes depleted of red blood cells were stimulated with mitomycin C (Sigma, Japan)treated E.G7-OVA cells (2.5 105/ml), and cultured in complete medium for 5 days. Then, CTLs and E.G7-OVA cells (2 104) www.jidonline.org

619


were incubated for 4 hours at 371C. Supernatants were collected and the amount of lactate dehydrogenase released was measured with a CytoTox 96 cytotoxicity assay kit (Promega, Japan). The specific lysis of target cells was quantified using the formula (experimentaleffector spontaneoustarget spontaneous)/(target maximumtarget spontaneous) 100.

24 hours after the first vaccination. Lymph node single cells (106) were incubated for 30 minutes on ice with phycoerythrin-conjugated anti-MHC class II mAb or phycoerythrin-conjugated anti-CD11c mAb (obtained from Pharmingen). Two-color fluorescence was analyzed on FACScan (Becton Dickinson, Franklin Lakes, NJ).

Preparation of mouse splenic cells

RT-PCR

Following the application of NS-398, CpG-ODN, and OVA via different routes, tape-stripped skin, i.p., or i.v., the spleen was removed 7 days after the last vaccination, and splenic single cell suspensions were prepared (Inoue et al., 2005a, b). To examine the production of cytokines, spleen cells (5 105/well) were incubated in Rosewell Park Memorial Institute-1640 medium supplemented with 10% fetal calf serum, 100 U/ml penicillin, 100 mg/ml streptomycin, 2 mM L-glutamine, and 5 105 M 2-mercaptoethanol for 24 hours in the presence of OVA (2 mg/ml).

Following the application of NS-398, CpG-ODN, and OVA to the tape-stripped skin, the draining lymph nodes were removed 5 days after the last vaccination. The cytokine mRNA expression in lymph nodes was determined by RT-PCR and normalized by b-actin levels. Total RNA (2 mg) was isolated with Isogen solution (Nippon Gene, Toyama, Japan) as reported previously (Inoue et al., 2005a, b). Complementary DNA was synthesized using SuperScript III. Then, complementary DNAs were amplified with primers specific for each cytokine (IL-12 p40, forward, 50 -CAGAAGCTAACCATCTCCTGG TTTG-30 , reverse, 50 -TCCGGAGTAATTTGGTGGTTCACAC-30 ; IL10, forward, 50 -GGACAACATACTGCTAACCGACTC-30 , reverse, 50 AAAATCACTCTTCACCTGCTCCAC-30 ; IFN-g, forward, 50 -CTCAA GTGGCATAGATGT-30 , reverse, 50 -GAGATAATCTGGCTCTGCAG GATT-30 ; b-actin, forward, 50 -GCACCACACCTTCTACAATGAG-30 , reverse, 50 -TTGGCATAGAGGTCTTTACGGA-30 ). The primers for b-actin were used as an internal control. PCR cycles were as follows: IL-12 p40 and IL-10 for 30 cycles, IFN-g and b-actin for 28 cycles. Primers designed based on the mouse sequences were obtained from Sigma Genosys (Tokyo, Japan).

Determination of cytokine levels The level of IFN-g in the supernatant of splenic cells was routinely determined with a sandwich ELISA using pairs of purified capture and biotinylated detection mAbs (all obtained from Pharmingen, San Diego, CA) recognizing murine IFN-g, according to the manufacturer’s directions.

Changes in OVA specific-antibody levels Blood samples from mice treated with NS-398, CpG-ODN, and OVA were collected 14 days after the last vaccination from the retroorbital plexus and sera were pooled. In brief, a 2-fold dilution of mouse sera (50 ml) was poured into an OVA-coated ELISA plate, and stood overnight to confirm that an immune reaction took place. The plate was then washed four times with PBS containing 0.05% Tween-20 (PBS-T), biotin-conjugated anti-mouse IgG2a mAb (obtained from Pharmingen) was added, and the plate was stood at 41C overnight. After a wash with PBS-T, avidin-conjugated peroxidase was added and the absorbance at 450 nm was determined as described previously (Inoue et al., 2005a, b).

Tumor growth and survival For tumor development, mice were immunized twice as stated above. One week after the last vaccination, mice were inoculated s.c. with E.G7-OVA tumor cells (5 106 cells/mouse), and subsequent tumor growth was assessed by taking an average of tumor diameters measured with calipers and calculating tumor volume using the formula: tumor volume ¼ (length) (width)2/2. The rate of survival was determined by counting surviving mice. In both experimental protocols, control mice were treated with an application of saline to intact skin or tape-stripped skin.

Cytokine determination in the skin Following the application of NS-398, CpG-ODN, and OVA to the tape-stripped skin, the skin was removed 12 hours after the first vaccination. The skin was excised and homogenated with the Polytron homogenizer in 1 ml of saline containing protease inhibitors (0.17 mg/ml phenylmethylsulphonyl fluoride, 0.02 mg/ml leupeptin, and 0.01 mg/ml aprotinin). The homogenate was centrifuged at 9,000 g for 10 minutes and the supernatant was collected. The levels of IL-10, IL-12 p70, and PGE2 in the skin were routinely determined with a sandwich ELISA using pairs of purified capture and biotinylated detection mAbs (IL-10 and IL-12 p70 were obtained from Pharmingen, and the PGE2 Enzyme Immunoassay Kit was obtained from Oxford Biomedical Research, Oxford, MI) recognizing murine IL-10, IL-12 p70, and PGE2 according to the manufacturer’s directions.

Assay for transcutaneous vaccination induced APC migration Following the application of NS-398, CpG-ODN, and Alexa-488conjugated OVA (obtained from Molecular Probes, Eugene, OR) to the tape-stripped skin, the draining lymph nodes were removed 620


Statistical analysis The paired Student’s t-test was used to compare paired groups. An analysis of variance was used for multi-group analysis. Values of P40.05 were considered to indicate a lack of significance. CONFLICT OF INTEREST The authors state no conflict of interest.

ACKNOWLEDGMENTS We are grateful to Miss T. Yoshida, Miss H. Kumamoto, Miss M. Ishida, Mr K. Kunii, and Miss Y. Saigo for technical assistance.

REFERENCES Abe N, Katamura K, Shintaku N, Fukui T, Kiyomasu T, Iio J et al. (1997) Prostaglandin E2 and IL-4 provide naive CD4+ T cells with distinct inhibitory signals for the priming of IFN-gamma production. Cell Immunol 181:86–9 Ackerman AL, Cresswell P (2004) Cellular mechanisms governing crosspresentation of exogenous antigens. Nat Immunol 5:678–84


Agrawal S, Kandimalla ER (2003) Modulation of toll-like receptor 9 responses through synthetic immunostimulatory motifs of DNA. Ann NY Acad Sci 1002:30–42 Beignon AS, Briand JP, Muller S, Partidos CD (2002) Immunization onto bare skin with synthetic peptides: immunomodulation with a CpG-containing oligodeoxynucleotide and effective priming of influenza virus-specific CD4+ T cells. Immunology 105:204–12 Bins AD, Jorritsma A, Wolkers MC, Hung CF, Wu TC, Schumacher TN et al. (2005) A rapid and potent DNA vaccination strategy defined by in vivo monitoring of antigen expression. Nat Med 11:899–904 Carbone FR, Bevan MJ (1990) Class I-restricted processing and presentation of exogenous cell-associated antigen in vivo. J Exp Med 171:377–87 Chamoto K, Tsuji T, Funamoto H, Kosaka A, Matsuzaki J, Sato T et al. (2004) Potentiation of tumor eradication by adoptive immunotherapy with Tcell receptor gene-transduced T-helper type 1 cells. Cancer Res 64:386–90 Chen Y, Zhang J, Moore SA, Ballas ZK, Portanova JP, Krieg AM et al. (2001) CpG DNA induces cyclooxygenase-2 expression and prostaglandin production. Int Immunol 13:1013–20 Cho HJ, Hayashi T, Datta SK, Takabayashi K, Van Uden JH, Horner A et al. (2002) IFN-a beta promote priming of antigen-specific CD8+ and CD4+ T lymphocytes by immunostimulatory DNA-based vaccines. J Immunol 168:4907–13 Davis HL, Weeratna R, Waldschmidt TJ, Tygrett L, Schorr J, Krieg AM (1998) CpG DNA is a potent enhancer of specific immunity in mice immunized with recombinant hepatitis B surface antigen. J Immunol 160:870–6 Fiorentino DF, Bond MW, Mosmann TR (1989) Two types of mouse T helper cell. IV. Th2 clones secrete a factor that inhibits cytokine production by Th1 clones. J Exp Med 170:2081–95 Ghosh DK, Misukonis MA, Reich C, Pisetsky DS, Weinberg JB (2001) Host response to infection: the role of CpG DNA in induction of cyclooxygenase 2 and nitric oxide synthase 2 in murine macrophages. Infect Immun 69:7703–10

acquisition of ability to produce IFN-gamma and IL-2, but not IL-4 and IL-5. J Immunol 155:4604–12 Klimuk SK, Najar HM, Semple SC, Aslanian S, Dutz JP (2004) Epicutaneous application of CpG oligodeoxynucleotides with peptide or protein antigen promotes the generation of CTL. J Invest Dermatol 122:1042–9 Klinman DM, Currie D, Gursel I, Verthelyi D (2004) Use of CpG oligodeoxynucleotides as immune adjuvants. Immunol Rev 199:201–16 Krieg AM (2002) CpG motifs in bacterial DNA and their immune effects. Annu Rev Immunol 20:709–60 Kuchtey J, Chefalo PJ, Gray RC, Ramachandra L, Harding CV (2005) Enhancement of dendritic cell antigen cross-presentation by CpG DNA involves type I IFN and stabilization of class I MHC mRNA. J Immunol 175:2244–51 Marzo AL, Kinnear BF, Lake RA, Frelinger JJ, Collins EJ, Robinson BW et al. (2000) Tumor-specific CD4+ T cells have a major ‘‘post-licensing’’ role in CTL mediated anti-tumor immunity. J Immunol 165:6047–55 Matsuo M, Nagata Y, Sato E, Atanackovic D, Valmori D, Chen YT et al. (2004) IFN-gamma enables cross-presentation of exogenous protein antigen in human Langerhans cells by potentiating maturation. Proc Natl Acad Sci USA 101:14467–72 Miconnet I, Koenig S, Speiser D, Krieg A, Guillaume P, Cerottini JC et al. (2002) CpG are efficient adjuvants for specific CTL induction against tumor antigen-derived peptide. J Immunol 168:1212–8 Mosmann TR, Coffman RL (1989) TH1 and TH2 cells: different patterns of lymphokine secretion lead to different functional properties. Annu Rev Immunol 7:145–73 Ridge JP, Di Rosa F, Matzinger P (1998) A conditioned dendritic cell can be a temporal bridge between a CD4+ T-helper and a T-killer cell. Nature 393:474–8 Seo N, Tokura Y, Nishijima T, Hashizume H, Furukawa F, Takigawa M (2000) Percutaneous peptide immunization via corneum barrier-disrupted murine skin for experimental tumor immunoprophylaxis. Proc Natl Acad Sci USA 97:371–6

Giuntoli RL II, Lu J, Kobayashi H, Kennedy R, Celis E (2002) Direct costimulation of tumor-reactive CTL by helper T cells potentiate their proliferation, survival, and effector function. Clin Cancer Res 8:922–31

Sharma S, Yang SC, Zhu L, Reckamp K, Gardner B, Baratelli F et al. (2005) Tumor cyclooxygenase-2/prostaglandin E2-dependent promotion of FOXP3 expression and CD4+ CD25+ T regulatory cell activities in lung cancer. Cancer Res 65:5211–20

Harizi H, Juzan M, Pitard V, Moreau JF, Gualde N (2002) Cyclooxygenase2-issued prostaglandin e(2) enhances the production of endogenous IL10, which down-regulates dendritic cell functions. J Immunol 168: 2255–63

Spies B, Hochrein H, Vabulas M, Huster K, Busch DH, Schmitz F et al. (2003) Vaccination with plasmid DNA activates dendritic cells via toll-like receptor 9 (TLR9) but functions in TLR9-deficient mice. J Immunol 171:5908–12

Heath WR, Belz GT, Behrens GM, Smith CM, Forehan SP, Parish IA et al. (2004) Cross-presentation, dendritic cell subsets, and the generation of immunity to cellular antigens. Immunol Rev 199:9–26

Suzuki Y, Wakita D, Chamoto K, Narita Y, Tsuji T, Takeshima T et al. (2004) Liposome-encapsulated CpG oligodeoxynucleotides as a potent adjuvant for inducing type 1 innate immunity. Cancer Res 64: 8754–60

Heikenwalder M, Polymenidou M, Junt T, Sigurdson C, Wagner H, Akira S et al. (2004) Lymphoid follicle destruction and immunosuppression after repeated CpG oligodeoxynucleotide administration. Nat Med 10:187–92 Hemmi H, Kaisho T, Takeda K, Akira S (2003) The roles of toll-like receptor 9, MyD88, and DNA-dependent protein kinase catalytic subunit in the effects of two distinct CpG DNAs on dendritic cell subsets. J Immunol 170:3059–64 Howe LR, Chang SH, Tolle KC, Dillon R, Young LJ, Cardiff RD et al. (2005) HER2/neu-induced mammary tumorigenesis and angiogenesis are reduced in cyclooxygenase-2 knockout mice. Cancer Res 65: 10113–9

Takigawa M, Tokura Y, Hashizume H, Yagi H, Seo N (2001) Percutaneous peptide immunization via corneum barrier-disrupted murine skin for experimental tumor immunoprophylaxis. Ann NY Acad Sci 941:139–46 Trinchieri G (1993) Interleukin-12 and its role in the generation of TH1 cells. Immunol Today 14:335–8 Tuttle TM, McCrady CW, Inge TH, Salour M, Bear HD (1993) gammaInterferon plays a key role in T-cell-induced tumor regression. Cancer Res 53:833–9 Wang B, Zhuang L, Fujisawa H, Shinder GA, Feliciani C, Shivji GM et al. (1999) Enhanced epidermal Langerhans cell migration in IL-10 knockout mice. J Immunol 162:277–83

Hsieh CS, Macatonia SE, Tripp CS, Wolf SF, O’Garra A, Murphy KM (1993) Development of TH1 CD4+ T cells through IL-12 produced by Listeriainduced macrophages. Science 260:547–9

Wu CY, Wang K, McDyer JF, Seder RA (1998) Prostaglandin E2 and dexamethasone inhibit IL-12 receptor expression and IL-12 responsiveness. J Immunol 161:2723–30

Inoue J, Yotsumoto S, Sakamoto T, Tsuchiya S, Aramaki Y (2005a) Changes in immune responses to antigen applied to tape-stripped skin with CpGoligodeoxynucleotide in mice. J Control Release 108:294–305

Yi AK, Yoon JG, Yeo SJ, Hong SC, English BK, Krieg AM (2002) Role of mitogen-activated protein kinases in CpG DNA-mediated IL-10 and IL12 production: central role of extracellular signal-regulated kinase in the negative feedback loop of the CpG DNA-mediated Th1 response. J Immunol 168:4711–20

Inoue J, Yotsumoto S, Sakamoto T, Tsuchiya S, Aramaki Y (2005b) Changes in immune responses to antigen applied to tape-stripped skin with CpGoligodeoxynucleotide in NC/Nga mice. Pharm Res 22:1627–33 Katamura K, Shintaku N, Yamauchi Y, Fukui T, Ohshima Y, Mayumi M et al. (1995) Prostaglandin E2 at priming of naive CD4+ T cells inhibits

Yu P, Spiotto MT, Lee Y, Schreiber H, Fu YX (2003) Complementary role of CD4+ T cells and secondary lymphoid tissues for cross-presentation of tumor antigen to CD8+ T cells. J Exp Med 197:985–95

www.jidonline.org

621