Laboratory/Clinical Translational Research

Oncology 2005;68:163–170DOI: 10.1159/000086770

Dendritic Cells Transduced with Tumor-Associated Antigen Gene Elicit PotentTherapeutic Antitumor Immunity: Comparisonwith Immunodominant Peptide-Pulsed DCs

Masaki Nakamura Makoto Iwahashi Mikihito Nakamori Kentaro UedaToshiyasu Ojima Teiji Naka Koichiro Ishida Hiroki Yamaue

Second Department of Surgery, Wakayama Medical University, School of Medicine, Wakayama, Japan

Received: July 28, 2003Accepted after revision: June 2, 2004Published online: July 4, 2005

Makoto IwahashiSecond Department of Surgery, Wakayama Medical University, School of Medicine811-1 Kimiidera, Wakayama 641-8510 (Japan)Tel. +81 73 441 0613, Fax +81 73 446 6566E-Mail makoto@wakayama-med.ac.jp

ABCFax + 41 61 306 12 34E-Mail karger@karger.chwww.karger.com

© 2005 S. Karger AG, Basel0030–2414/05/0683–0163$22.00/0

Accessible online at:www.karger.com/ocl

Supported by Grant-in-Aid No. 12671170 from the Ministry of Edu-cation, Culture, Sports, Science and Technology, Japan.

Key WordsDendritic cells W Adenovirus vector W gp70 W

Immunodominant peptide W CD4+T cell response W CCR7

AbstractSeveral studies have shown that vaccine therapy usingdendritic cells (DCs) pulsed with specific tumor antigenpeptides can effectively induce antitumor immunity.Peptide-pulsed DC therapy is reported to be effectiveagainst melanoma, while it is still not sufficient to showthe antitumor therapeutic effect against epithelial solidtumors such as gastrointestinal malignancies. Recently,it has been reported that vaccine therapy using DCstransduced with a surrogate tumor antigen gene can elic-it a potent therapeutic antitumor immunity. In this study,we investigated the efficacy of vaccine therapy usingDCs transduced with the natural tumor antigen in com-parison with peptide-pulsed DCs. DCs derived from mu-rine bone marrow were adenovirally transduced withmurine endogenous tumor antigen gp70 gene, which isexpressed in CT26 cells, or DCs were pulsed with theimmunodominant peptide AH-1 derived from gp70. We

compared these two cancer vaccines in terms of induc-tion of antigen-specific cytotoxic T lymphocyte (CTL)responses, CD4+ T cell response against tumor cells,migratory capacity of DCs and therapeutic immunity invivo. The cytotoxic activity of splenocytes against CT26and Meth-A pulsed with AH-1 in mice immunized withgp70 gene-transduced DCs was higher than that withAH-1-pulsed DCs. CD4+ T cells induced from mice immu-nized with gp70 gene-transduced DCs produced higherlevels of IFN-Á by stimulation with CT26 than those frommice immunized with AH-1-pulsed DCs (p ! 0.0001), andit was suggested that DCs transduced with tumor-associ-ated antigen (TAA) gene induced tumor-specific CD4+ Tcells, and those CD4+ T cells played a critical role in thepriming phase of the CD8+ T cell response for the induc-tion of CD8+ CTL. Furthermore, DCs adenovirally trans-duced with TAA gene showed an enhancement of ex-pression of CC chemokine receptor 7 and improved themigratory capacity to draining lymph nodes. In subcuta-neous models, the vaccination using gp70 gene-trans-duced DCs provided a remarkably higher therapeuticefficacy than that using AH-1-pulsed DCs. These resultssuggested that vaccine therapy using DCs adenovirallytransduced with TAA gene can elicit potent antitumorimmunity, and may be useful for clinical application.

Copyright © 2005 S. Karger AG, Basel

164 Oncology 2005;68:163–170 Nakamura/Iwahashi/Nakamori/Ueda/Ojima/Naka/Ishida/Yamaue

Introduction

Several groups have identified tumor-associated anti-gens (TAAs) such as MAGE and MART-1, recognized bytumor-specific CD8 cytotoxic T lymphocytes (CTLs), aswell as cloned genes encoding these tumor-associatedantigens [1–3]. We have also identified CTL epitope pep-tides of HER2/neu and p53 restricted by HLA-A24,which is one of the most common alleles in Japanese andCaucasians [4, 5]. On the other hand, dendritic cells(DCs) are the most potent antigen-presenting cells [6, 7],and therefore, DCs pulsed with tumor antigen peptidesmight offer an attractive solution. In fact, it has beenreported that when patients with advanced melanomawere immunized with peptide- or tumor lysate-pulsedDCs, an objective response was recognized in 5 of 16patients with regression of metastatic lesions in variousorgans [8]. However, in epithelial solid tumors such asgastrointestinal cancers or prostate cancers, althoughsome patients showed antigen-specific T cell responses invitro by vaccination using peptide-pulsed DCs, tumorregressions were not observed at all [9, 10]. We also per-formed a phase I trial of cancer vaccine therapy using DCspulsed with carcinoembryonic antigen-specific peptide(CEA 652), which was bound by HLA-A*2402, for pa-tients with colorectal cancer. The peripheral blood mono-nuclear cells from 4 of 7 patients after vaccinationsshowed HLA-A24-restricted and carcinoembryonic anti-gen-specific cytotoxic activity. However, no clinical re-sponse was recognized [11].

Recently, it has been demonstrated that vaccinationusing DCs transduced with the TAA gene can elicit apotent therapeutic antitumor immunity [12–14]. Thisstrategy has potential advantages as follows. First, DCstransduced with the entire TAA gene may present multi-ple epitopes including previously unknown epitopes asso-ciated with different MHC class I molecules [12]. Second,TAA gene-transduced DCs may possibly present helperepitopes associated with MHC class II molecules [15, 16].Third, DCs are provided with a renewable supply of theantigen for presentation by transduction with the gene, asopposed to a single pulse of peptide [14]. However, surro-gate tumor antigen genes such as ovalbumin, ß-galactosi-dase, and human TAA were used in the previous studies[12–14], and therefore, the data reported in these studiesmight possibly be artificial to some extent.

In the previous study, we used CT26 murine colorectalcancer cells, which endogenously express gp70, an enve-lope protein of an endogenous ecotropic murine leukemiavirus, and showed that specific CTL responses against

CT26 and therapeutic immunity could be induced inmice immunized with DCs transduced with gp70 gene[17].

The nonamer peptide AH-1 derived from gp70 hasbeen proved to be an immunodominant peptide [18].Therefore, this model is quite adequate for accuratelyevaluating the ability of natural TAA gene-transducedDCs, compared with the immunodominant peptide-pulsed DCs. In the present study, we examined the effica-cy of vaccination using DCs transduced with TAA gene interms of induction of CTLs activity, CD4+ T cell responseagainst tumor cells and the in vivo tumor suppression,compared with peptide-pulsed DCs. Then, we clarifiedthe possibility of clinical application and the problems ofthis strategy by comparison with peptide-pulsed DC ther-apy, which has already been applied in many clinicalstudies.

Materials and Methods

Mice and Cell LinesSix- to eight-week-old female BALB/c (H-2d) mice were pur-

chased from Japan SLC (Hamamatsu, Japan).Syngeneic tumor cell line CT26 (H-2d), an undifferentiated low-

immunogenic murine colorectal adenocarcinoma cell line (kindlyprovided by Hoffmann-La Roche, Kamakura, Japan) and humanembryonic kidney cell line 293 (American Type Culture Collection,Rockville, Md., USA) were grown in DMEM (Nissui PharmaceuticalCo., Tokyo, Japan) supplemented with 10% fetal bovine serum(FBS) (Gibco, New York, N.Y., USA), 2 mM L-glutamine (Gibco),100 U/ml penicillin and 100 Ìg/ml streptomycin (Gibco). Meth-A(H-2d), a methylcholanthrene-induced sarcoma cell line (Riken GeneBank, Ibaraki, Japan) was grown in RPMI 1640 (Nissui Pharmaceu-tical Co.) supplemented with 10% FBS, 2 mM L-glutamine, 100 U/ml penicillin and 100 Ìg/ml streptomycin.

Generation of DCs in vitro from Bone MarrowDCs were generated from murine bone marrow precursors as

described previously [17, 19]. In brief, murine bone marrow cellswere seeded at 2 ! 106 per dish in 10 ml RPMI 1640 medium sup-plemented with 10% FBS, 2 mM L-glutamine, 50 ÌM 2-mercapto-ethanol (Gibco), 100 U/ml penicillin and 100 Ìg/ml streptomycincontaining 200 U/ml recombinant murine GM-CSF (rmGM-CSF)(kindly provided by Kirin Brewery, Tokyo, Japan). On day 3 of cul-ture, 10 ml of fresh medium containing 200 U/ml rmGM-CSF wasadded to the dishes. On days 6 and 8, half of the culture supernatantwas replaced with fresh complete medium containing 200 U/ml ofrmGM-CSF. On day 10, nonadherent cells were collected and usedfor the experiments. More than 99% of the cells showed high expres-sion of CD11c, CD80, MHC class I antigen and MHC class II anti-gen, and around 30% showed low expression of CD86 by flow cytom-etric analysis using FACScan (Becton Dickinson, Mountain View,Calif., USA).

Effective Immunotherapy Using DCsTransduced with TAA Gene

Oncology 2005;68:163–170 165

Recombinant Ad Vectors and Transduction of DCs with Ad VectorThe recombinant Axgp70 and AxLacZ were generated by the

COS-TPC method as previously described [17, 20, 21]. AxLacZexpressing a LacZ reporter gene was used as mock virus. DCs weretransfected with Axgp70 or AxLacZ using a centrifugal method [17].DCs were mixed with Ad vector at a multiplicity of infection (moi) of100 in a final volume of 1 ml of phosphate-buffered saline (PBS)containing 1% FBS (FBS/PBS), and centrifuged at 2000 g at 37°Cfor 2 h. DCs were then washed twice with PBS and cultured in com-plete medium containing 200 U/ml rmGM-CSF for 24 h, and wereused for the experiments.

Pulsing of DCs with the Immunodominant Peptide, AH-1Peptide AH-1, nonamer SPSYVYHQF, derived from the enve-

lope protein (gp70) of an endogenous ecotropic murine leukemiavirus (MuLV) was synthesized by Takara to a purity of 190% asdetermined by HPLC and amino acid analysis.

DCs (2 ! 106) were cultured in 1 ml of RPMI 1640 mediumsupplemented with 10% FBS, 2 mM L-glutamine, 50 ÌM 2-mercap-toethanol, 100 U/ml penicillin and 100 Ìg/ml streptomycin contain-ing 100 Ìg/ml of AH-1 at 37°C for 3 h (with gentle shaking every30 min). Cells were then washed twice in PBS and used for the exper-iments.

Induction of Antigen-Specific CTL and Cytotoxicity AssayTo evaluate levels of CTL activity, BALB/c mice were immu-

nized once by subcutaneous injection of 5 ! 105 genetically modifiedDCs or AH-1-pulsed DCs. Spleens were removed 14 days after DCimmunization, and then the in vivo primed splenocytes were cocul-tured (4 ! 106/ml) with irradiated (10,000 rad) CT26 cells (4 ! 105/ml) in a 6 well-plate in complete medium containing IL-2 at 50 units/ml. After 5 days of coculture, the in vivo restimulated splenocyteswere assayed in a 4-hour 51Cr-release assay as previously described[22, 23]. CT26, Meth-A and Meth-A pulsed with AH-1 (100 Ìg/2 !106) (Meth-A/AH-1) were used as target cells. The maximum amountof 51Cr incorporated was determined by adding 1 N HCl to the targetcells. The percentage cytotoxicity was calculated as follows: % lysis =[(sample cpm – spontaneous cpm)/(maximum cpm – spontaneouscpm)] !100.

Murine IFN-Á Release AssayWe examined the tumor antigen-specific CD4+ and CD8+ T cell

responses using a murine IFN-Á (mIFN-Á) release assay following themethod described previously [15]. BALB/c mice were immunizedonce by subcutaneous injection of 5 ! 105 genetically modified DCsor AH-1-pulsed DCs. Spleens were removed 14 days after DC immu-nization, and then CD4+ and CD8+ T cells were isolated from the invivo primed splenocytes using a magnetic cell sorting (MACS) sys-tem (Miltenyi Biotec, Bergisch Gladbach, Germany). Splenocyteswere magnetically labeled with CD8a (Ly-2) or CD4 (L3T4) Micro-Beads (Miltenyi Biotec) and passed through a separation column,which was placed in the magnetic field of a MACS separator. Themagnetically labeled CD8+ T cells or CD4+ T cells were retained inthe column while the unlabeled cells ran through. After the removalof the column from the magnetic field, the magnetically retained cellscould be eluted as the positively selected cell fraction. The purity ofthe positively selected cells, tested by flow cytometry, varied between90 and 95% (data not shown). CD4+ T cells and CD8+ T cells (2 !106 per well in a 12-well plate) were cultured in the presence of irra-diated (10,000 rad) CT26 (lymphocytes: CT26 ratio, 10:1) in com-

plete medium containing IL-2 at 50 units/ml. After 5 days of cocul-ture, responding lymphocytes (5 ! 104) were stimulated with CT26(5 ! 103), in a total volume of 200 Ìl of complete medium in a 96-well round-bottomed plate. Supernatants collected 24 h later weretested for mIFN-Á release. The mIFN-Á release assay was performedwith an ELISA kit (Endogen, Woburn, Mass., USA) according to themanufacturer’s instructions.

Trafficking Study of Subcutaneously Injected TAA Gene-Transduced DCs or Peptide-Pulsed DCsWe examined whether DCs adenovirally transduced with gp70

gene or DCs pulsed with AH-1 could migrate to regional lymph nodes(LNs), following the method described previously [17]. Briefly, DCswere labeled with the fluorescent dye PKH67 (ZYNAXIS, Malvern,Pa., USA), and were injected subcutaneously into the mice in thelower abdomen. After 72 h, the mice were sacrificed, and the drain-ing inguinal LNs were removed. All the harvested LNs were gentlycrushed, suspended in PBS, and these suspensions were subjected toflow cytometric analysis to detect fluorescent-positive cells withinthe LN preparation.

Analysis of mRNA Expression of Murine CC ChemokineReceptor 7 (CCR7) Using Reverse Transcriptase-PolymeraseChain ReactionTotal RNA was isolated from DCs adenovirally transduced with

gp70 gene or DCs pulsed with AH-1, and reverse transcription-poly-merase chain reaction was then performed using a pair of specificprimers as previously described [17].

Experimental Design of DC Vaccine Therapy in SubcutaneousTumor ModelsBALB/c mice (6-8 weeks old) were inoculated subcutaneously

into the right flank with 1 ! 106 CT26 cells. Five days later, tumor-bearing mice (n = 5/group) were treated with subcutaneous injectionin the opposite flank with 5 ! 105 of genetically modified DCs (DC-AxLacZ, DC-Axgp70) or DC pulsed with the immunodominant pep-tide AH-1 (DC/AH-1) (n = 5/group). The development of tumors wasmonitored every day. Twenty days after tumor implantation, themice were killed and the subcutaneous tumor was estimated usingthe following formula: (short diameter)2 ! long diameter ! 0.52.

Statistical AnalysisQuantitative results are expressed as mean B standard deviation.

Statistical analysis was performed by ANOVA and Fisher’s test usingthe Statview 5.0 software (Abacus Concepts, Berkeley, Calif., USA).A p value !0.05 was considered to indicate statistical significance.

Results

Tumor-Specific CTL Responses Induced byImmunization with TAA Gene-Transduced DCs orPeptide-Pulsed DCsAntigen-specific CTL responses induced by immuni-

zation with DCs transduced with murine TAA gp70 gene(DC-Axgp70) were first compared to DC/AH-1. The cyto-toxic activities against CT26 in spleen cells of mice immu-

166 Oncology 2005;68:163–170 Nakamura/Iwahashi/Nakamori/Ueda/Ojima/Naka/Ishida/Yamaue

Fig. 1. CTL activity induced by immunization using TAA gene-transduced DCs or peptide-pulsed DCs. BALB/c mice were immu-nized once by subcutaneous injection of DCs transfected withAxgp70: DC-Axgp70 ($), DCs pulsed with AH-1: DC/AH-1 (P) orDCs transfected with AxLacZ: DC-AxLacZ (d). Spleen cells wereisolated at 14 days after immunization, and restimulated in vitro for5 days with irradiated CT26 cells. CT26 (a), Meth-A (b), or Meth-A

pulsed with AH-1: Meth-A/AH-1 (c) were used as targets to deter-mine the cytotoxic activity of effector cells. Results are shown as themean B SD (n = 5 for each group). a ** Significantly different fromDC-AxLacZ (p ! 0.001) and from DC/AH-1 (E/T: 50, p ! 0.001).## Significantly different from DC-AxLacZ (p ! 0.005). c §§ Signifi-cantly different from DC-AxLacZ (p ! 0.001).

nized with DC-Axgp70 were significantly higher thanthose in the mice immunized with DC/AH-1 (E/T: 50, p !0.001). These two groups showed higher cytotoxicityagainst CT26 than the DC-AxLacZ-immunized group(p ! 0.005; fig. 1a). On the other hand, the cytotoxic activ-ities against Meth-A cells, which do not express gp70,were less than 5% in all groups (fig. 1b). The cytotoxicactivities against Meth-A/AH-1 in the spleen cells of miceimmunized with DC-Axgp70 or with DC/AH-1 were sig-nificantly higher than those in the mice immunized withDC-AxLacZ (p ! 0.001), although there were no differ-ences in the cytotoxic activity between the mice immu-nized with DC-Axgp70 and those immunized with DC/AH-1 (fig. 1c).

Tumor Antigen-Specific CD8+ and CD4+ T CellResponses Induced by Immunization with TAA Gene-Transduced DCs or Peptide-Pulsed DCsTo compare tumor antigen-specific CD8+ and CD4+

T cell responses between the mice immunized with TAAgene-transduced DCs and peptide-pulsed DCs, their abil-ity to synthesize mIFN-Á upon specific restimulation wastested. IFN-Á production of CD8+ T cells from the mice

immunized with DC-Axgp70 or DC/AH-1 was signifi-cantly higher than that from the control mice (p !

0.0001). Moreover, that from the mice immunized withDC-Axgp70 was higher than that from the mice immu-nized with DC/AH-1 (p ! 0.0001; fig. 2a). On the otherhand, IFN-Á production of CD4+ T cells from the miceimmunized with DC-Axgp70 was significantly higherthan that from the mice immunized with DC/AH-1 (p !0.0001), but there was no difference between the miceimmunized with DC/AH-1 and the control group(fig. 2b).

Enhancement of Migratory Capacity by TAAGene-Transduced DCs in vivoTo investigate why the transduction of DCs with TAA

gene enhances their capacity to induce tumor-specificCTL, we compared migratory capacity to regional LNs ofTAA gene-transduced DCs with peptide-pulsed DCs.

Untreated DC, DC-AxLacZ, DC-Axgp70 and DC/AH-1 were stably labeled with a fluorescent dye (PKH67) andsubcutaneously injected into mice. After 3 days, twodraining LNs were removed from each mouse. Interest-ingly, remarkable swelling of LNs in the mice immunized

Effective Immunotherapy Using DCsTransduced with TAA Gene

Oncology 2005;68:163–170 167

Fig. 2. Tumor antigen-specific CD8+ and CD4+ T cells responseinduced by immunization with TAA gene-transduced DCs or pep-tide-pulsed DCs. BALB/c mice were immunized once by subcuta-neous injection of DC-Axgp70, DC/AH-1 or DC-AxLacZ. Spleencells were isolated at 14 days after immunization, and then CD8+and CD4+ T cells were isolated from the in vivo primed splenocytes.CD8+ T cells (a) and CD4+ T cells (b) were cultured in the presence

of irradiated CT26 for 5 days, and responding lymphocytes were re-stimulated with CT26 in a 96-well round-bottomed plate. Superna-tants collected 24 h later were tested for mIFN-Á release. Results areshown as the mean B SD (n = 5 for each group). a ** Significantlydifferent from DC/AH-1 and control groups (p ! 0.0001). ## Signifi-cantly different from control groups (p ! 0.0001). b ** Significantlydifferent from DC/AH-1 and control groups (p ! 0.0001).

Fig. 3. Migratory capacity of TAA gene-transduced DCs or peptide-pulsed DCs in vivo. Untreated DCs, DC-AxLacZ, DC-Axgp70, DC/AH-1 were labeled with the fluorescent dye PKH67 and injected subcutaneously into thelower abdomens of mice. After 72 h, single-cell suspensions from the draining LNs were analyzed for labeled cells byflow cytometry.

with DC-Axgp70 was recognized (around 2.6 mm indiameter), as compared to mice immunized with DC/AH-1 (around 1.6 mm). Single-cell suspensions from drainingLNs were analyzed for labeled cells by flow cytometry.The population of PKH67-positive cells in the regionalLNs of mice immunized with untreated DC, DC-Axgp70,DC/AH-1 and DC-AxLacZ was 1.99, 10.21, 3.81 and

8.06%, respectively, and it was small in the LNs of themice immunized with DC/AH-1, but was obviously in-creased in those of the mice immunized with DC-Axgp70.In contrast, hardly any fluorescent-positive cells could befound in the LNs of the mice injected with untreated DCs(fig. 3).

168 Oncology 2005;68:163–170 Nakamura/Iwahashi/Nakamori/Ueda/Ojima/Naka/Ishida/Yamaue

Murine CCR7 mRNA Expression in TAA Gene-Transduced DCs or Peptide-Pulsed DCsTo investigate the influence of the adenoviral gene

transduction or peptide pulse on DCs, we examined theexpression of CCR7 on DCs, which could consequentlymodify the migratory capacity of DCs to regional LNs.CCR7 mRNA expression was not recognized in untreatedDCs, but was markedly induced by adenoviral transduc-tion with gp70 gene. In contrast, it was barely induced inpeptide-pulsed DCs (fig. 4).

Fig. 4. Murine CCR7 mRNA expression in TAA gene-transducedDCs or peptide-pulsed DCs. DCs were transfected with AxLacZ,Axgp70 or pulsed with AH-1. After 48 h, total RNA was extractedfrom these DCs and the expression of mCCR7 mRNA was assessedby RT-PCR.

Fig. 5. Tumor growth suppression in themice immunized with TAA gene-transducedDCs or peptide-pulsed DCs in a subcuta-neous tumor model. BALB/c mice were inoc-ulated subcutaneously in the right flank with1 ! 106 CT26. Five days later, tumor-bear-ing mice (n = 5/group) were treated with sub-cutaneous injection in the opposite flankwith 5 ! 105 DC-AxLacZ (P), DC-Axgp70(j), DC/AH-1 (o), or PBS ()). Results arepresented as mean tumor volumes of micethat developed tumors in each group. ** Sig-nificantly different from PBS, DC-AxLacZand DC/AH-1 (p ! 0.005).

Therapeutic Efficacy of the Vaccination Using TAAGene-Transduced DCs or Peptide-Pulsed DCs inSubcutaneous Tumor ModelsWe compared the therapeutic efficacy of vaccination

therapy using TAA gene-transduced DCs with peptide-pulsed DCs in vivo. The therapeutic efficacy of a singlevaccination using gp70 gene-transduced DCs was greaterin this pre-existing tumor model than that using AH-1-pulsed DCs. On day 20 after tumor implantation, quanti-tative analysis of the tumor volume showed significantdifferences between the DC-Axgp70 group and the con-trol groups or even the DC/AH-1 group (p ! 0.005). Therewas a difference between the DC/AH-1 group and thecontrol groups, but not a significant difference (fig. 5).

Discussion

Among the many different immunotherapeutic strate-gies, we provide here an important comparison of twoDC-based vaccines using TAA gene or tumor antigen pep-tide. Our in vitro cytotoxicity studies showed that thecytotoxic activity against Meth-A/AH-1 from mice immu-nized with DC-Axgp70 was almost as high as that of miceimmunized with DC/AH-1, suggesting that DC-Axgp70and DC/AH-1 were capable of inducing gp70-specificCTLs. On the other hand, the cytotoxic activity againstCT26, which endogenously expressed gp70 from the miceimmunized with DC-Axgp70, was significantly higher

Effective Immunotherapy Using DCsTransduced with TAA Gene

Oncology 2005;68:163–170 169

than that of the mice immunized with DC/AH-1. Theseresults suggested that tumor cell lines pulsed with antigenpeptide, which were usually used as target cells in studiesof peptide-pulsed DCs, could not reflect accurately thecytotoxic activity against natural tumor cells, which en-dogenously express tumor antigen [8].

The cytotoxic activity against CT26 by the immuniza-tion of DCs transduced with TAA gene was almost com-pletely blocked by anti-CD8 monoclonal antibody andanti-MHC-class I monoclonal antibody, but not blockedby anti-CD4 monoclonal antibody, suggesting the pheno-type of the effector cells was MHC-class I restricted CD8+T cells (data not shown). One of the reasons why potentCD8+ CTL activities were induced by the immunizationusing DCs transduced with TAA gene may be that TAAgene-transduced DCs might continuously present multi-ple epitopes in association with MHC class I molecules[12, 13]. However, there seems to be other reasons assome investigators have reported [15, 24]. Tumor-specificCD4+ T cells could also play important roles in terms offacilitating the induction of CD8+ antitumor CTL re-sponse. In this context, we examined the antigen-specificresponses of CD4+ T cells and CD8+ T cells separatelyusing an IFN-Á release assay. Our results showed that theCD8+ T cell response to CT26 induced from the miceimmunized with DC-Axgp70 was significantly strongerthan that from the mice immunized with DC/AH-1, andthese data were exactly parallel with the CTL activities.Noteworthily, high IFN-Á production from CD4+ T cellswas recognized in the mice immunized with DC-Axgp70,while the level of IFN-Á production from CD4+ T cells inthe mice immunized with DC/AH-1 was almost the sameas that of the controls. These results suggested that DCstransduced with TAA gene could endogenously expresshelper epitopes associated with MHC class II molecules,inducing tumor-specific CD4+ T cells, and those CD4+ Tcells played a critical role in the priming phase of CD8+ Tcell response for the induction of CD8+ CTLs, resulting ina remarkably potent CTL response.

It is also very important to examine the effect of adeno-viral gene transduction on DC functions comparing withpeptide pulsing. We have focused on CCR7 [17], which isrelated to the migratory capacity of DCs to regional LNs[25, 26]. Our results in the present study suggested thepossibility that adenoviral gene transduction would en-hance the expression of CCR7 on DCs, leading to anenhanced migratory capacity of DCs to draining LNs inwhich DCs could stimulate T cells. Consequently, a highlyeffective antitumor immune response would be inducedin vivo. In fact, it has been reported that CCR7 mRNA

expression on DCs progressively increases with their mat-uration [27], and that DCs are activated by adenoviralgene-transduction and produce IL-12 in significantamounts [28]. Therefore, this vaccine therapy seems to bean ideal strategy for a DC-based immunotherapy.

In the in vivo experiments, our results showed that asingle treatment with gp70 gene-transduced DCs provideda more marked effect in terms of suppression of tumorgrowth than that with DC/AH-1. These are not onlybecause of potent CD8+ CTL activities, but also antigen-specific CD4+ helper T cell responses were induced by theimmunization with TAA gene-transduced DCs. It has alsobeen reported that mRNA expression of tumor antigen inDCs by adenoviral transduction of the TAA gene is main-tained for at least 5 days in vitro [13], and therefore a singlesubcutaneous immunization of genetically modified DCswould elicit long-lasting antigen-specific CTL immunity(300–400 days) [29]. There is another benefit associatedwith vaccine therapy using TAA gene-transduced DCs.The half-life of most MHC-restricted epitopes is only 5–10h [30]. To achieve effective immune responses by singlepeptide-pulsed DCs, multiple injections of peptide-pulsedDCs may be required, while TAA gene-transduced DCscan continuously produce and load antigen peptides onMHC molecules. A single immunization with TAA gene-transduced DCs could thus elicit a potent tumor-specificimmune response.

In the present study, no mice immunized with TAAgene-transduced DCs showed complete remission of theestablished tumor. To improve the antitumor effect invivo, it may be necessary to increase the number of gene-modified DCs for a vaccination or the frequency of thevaccination. Actually, however, this is difficult in clinicaluse. Different ideas for enhancing the antitumor effect arenecessary for the future clinical application. In this con-text, cotransduction of DCs with a Toll-like receptor gene,which is associated with the maturation of DCs [31], orimmunostimulatory cytokine genes such as IL-12 orCD40-ligand [32] is expected to be the subject of furtherstudy. Quite recently, we have reported that GM-CSFcotransduction of DCs transduced with gp70 can signifi-cantly enhance antigen-specific CTL responses and thera-peutic antitumor immunity [17].

In conclusion, vaccination therapy using DCs trans-duced with the entire gene encoding endogenous TAAelicited more efficient therapeutic immune responsesagainst tumors than vaccination using DCs pulsed withthe immunodominant peptides. Further studies to en-hance the in vivo antitumor effect will be necessary beforeclinical application.

170 Oncology 2005;68:163–170 Nakamura/Iwahashi/Nakamori/Ueda/Ojima/Naka/Ishida/Yamaue

References

1 van der Bruggen P, Traversari C, Chomez P,Lurquin C, De Plaen E, Van den Eynde B,Knuth A, Boon T: A gene encoding an antigenrecognized by cytolytic T lymphocytes on ahuman melanoma. Science 1991;254:1643–1647.

2 Kawakami Y, Eliyahu S, Sakaguchi K, RobbinsPF, Rivoltini L, Yannelli JR, Appella E, Rosen-berg SA: Identification of the immunodomi-nant peptides of the MART-1 human melano-ma antigen recognized by the majority of HLA-A2-restricted tumor infiltrating lymphocytes. JExp Med 1994;180:347–352.

3 Nukaya I, Yasumoto M, Iwasaki T, Ideno M,Sette A, Celis E, Takesako K, Kato I: Identifi-cation of HLA-A24 epitope peptides of carci-noembryonic antigen which induce tumor-reactive cytotoxic T lymphocyte. Int J Cancer1999;80:92–97.

4 Tanaka H, Tsunoda T, Nukaya I, Sette A, Mas-tuda K, Umano Y, Yamaue H, Takesako K,Tanimura H: Mapping the HLA-A24-re-stricted T-cell epitope peptide from a tumor-associated antigen HER2/neu: Possible immu-notherapy for colorectal carcinomas. Br J Can-cer 2001;84:94–99.

5 Umano Y, Tsunoda T, Tanaka H, Matsuda K,Yamaue H, Tanimura H: Generation of cyto-toxic T cell responses to an HLA-A24 restrictedepitope peptide derived from wild-type p53. BrJ Cancer 2001;84:1052–1057.

6 Steinman RM: The dendritic cell system andits role in immunogenicity. Annu Rev Immu-nol 1991;9:271–296.

7 Hart DNJ: Dendritic cells: Unique leukocytepopulations which control the primary im-mune response. Blood 1997;90:3245–3287.

8 Nestle FO, Alijagic S, Gilliet M, Sun Y, GrabbeS, Dummer R, Burg G, Schadendorf D: Vacci-nation of melanoma patients with peptide- ortumor lysate-pulsed dendritic cells. Nat Med1998;4:328–332.

9 Morse MA, Deng Y, Coleman D, Hull S, Ki-trell-Fisher E, Nair S, Schlom J, Ryback ME,Lyerly HK: A phase I study of active immuno-therapy with carcinoembryonic antigen pep-tide (CAP-1)-pulsed, autologous human cul-tured dendritic cells in patients with metastaticmalignancies expressing carcinoembryonic an-tigen. Clin Cancer Res 1999;5:1331–1338.

10 Burch PA, Breen JK, Buckner JC, GastineauDA, Kaur JA, Laus RL, Padley DJ, PeshwaMV, Pitot HC, Richardson RL, Smits BJ, So-papan P, Strang G, Valone FH, Vuk-PavlovicS: Priming tissue-specific cellular immunity ina phase I trial of autologous dendritic cells forprostate cancer. Clin Cancer Res 2000;6:2175–2182.

11 Matsuda K, Tsunoda T, Tanaka H, Umano Y,Tanimura H, Nukaya I, Takesako K, YamaueH: Enhancement of cytotoxic T-lymphocyte re-sponses in patients with gastrointestinal malig-nancies following vaccination with CEA pep-tide-pulsed dendritic cells. Cancer ImmunolImmunother 2004;53:609–616.

12 Song W, Kong HL, Carpenter H, Torii H,Granstein R, Rafii S, Moore MAS, Crystal RG:Dendritic cells genetically modified with anadenovirus vector encoding the cDNA for amodel antigen induce protective and thera-peutic antitumor immunity. J Exp Med 1997;186:1247–1256.

13 Ribas A, Butterfield LH, McBride WH, JilaniSM, Bui LA, Vollmer CM, Lau R, Dissette VB,Hu B, Chen AY, Glaspy JA, Economou JS:Genetic immunization for the melanoma anti-gen MART-1/Melan-A using recombinant ade-novirus-transduced murine dendritic cells.Cancer Res 1997;57:2865–2869.

14 Kaplan JM, Yu Q, Piraino ST, Pennington SE,Shankara S, Woodworth LA, Roberts BL: In-duction of antitumor immunity with dendriticcells transduced with adenovirus vector-encod-ing endogenous tumor-associated antigens. JImmunol 1999;163:699–707.

15 Pérez-Dı́ez A, Butterfield LH, Li L, Chakrabor-ty NG, Economou JS, Mukherji B: Generationof CD8+ and CD4+ T-cell response to dendriticcells genetically engineered to express theMART-1/Melan-A gene. Cancer Res 1998;58:5305–5309.

16 Diebold SS, Cotton M, Koch N, Zenke M:MHC class II presentation of endogenously ex-pressed antigens by transfected dendritic cells.Gene Ther 2001;8:487–493.

17 Nakamura M, Iwahashi M, Nakamori M, UedaK, Matsuura I, Noguchi K, Yamaue H: Den-dritic cells genetically engineered to simulta-neously express endogenous tumor antigen andgranulocyte macrophage colony-stimulatingfactor elicit potent therapeutic antitumor im-munity. Clin Cancer Res 2002;8:2742–2749.

18 Huang AYC, Gulden PH, Woods AS, ThomasMC, Tong CD, Wang W, Engelhard VH, Pas-ternack G, Cotter R, Hunt D, Pardoll DM, Jaf-fee EM: The immunodominant major histo-compatibility complex class I-restricted anti-gen of a murine colon tumor derives from anendogenous retroviral gene product. Proc NatlAcad Sci USA 1996;93:9730–9735.

19 Lutz MB, Kukutsch N, Ogilvie ALJ, Rössner S,Koch F, Romani N, Schuler G: An advancedculture method for generating large quantitiesof highly pure dendritic cells from mouse bonemarrow. J Immunol Methods 1999;223:77–92.

20 Ueda K, Iwahashi M, Nakamori M, NakamuraM, Yamaue H, Tanimura H: Enhanced se-lective gene expression by adenovirus vectorusing Cre/loxP regulation system for humancarcinoembryonic antigen-producing carcino-ma. Oncology 2000;59:255–265.

21 Ueda K, Iwahashi M, Nakamori M, NakamuraM, Matsuura I, Yamaue H, Tanimura H: Car-cinoembryonic antigen-specific suicide genetherapy of cytosine deaminase/5-fluorocyto-sine enhanced by the Cre/loxP system in theorthotopic gastric carcinoma model. CancerRes 2001;61:6158–6162.

22 Iwahashi M, Tanimura H, Yamaue H, Tsuno-da T, Tani M, Tamai M, Noguchi K, Hotta T:Defective autologous mixed lymphocyte reac-tion (AMLR) and killer activity generated inthe AMLR in cancer patients. Int J Cancer1992;51:67–71.

23 Terasawa H, Tanimura H, Nakamori M, Tsu-noda T, Iwahashi M, Tani M, Yamaue H: Anti-tumor effects of interleukin-2 gene-modifiedfibroblasts in an orthotopic colon cancer mod-el. Jpn J Cancer Res 1999;90:1000–1006.

24 Steitz J, Bruck J, Knop J, Tuting T: Adenovi-rus-transduced dendritic cells stimulate cellu-lar immunity to melanoma via a CD4+ T cell-dependent mechanism. Gene Ther 2001;8:1255–1263.

25 Yanagihara S, Komura E, Nagafune J, WataraiH, Yamaguchi Y: EBI1/CCR7 is a new mem-ber of dendritic cell chemokine receptor that isup-regulated upon maturation. J Immunol1998;161:3096–3102.

26 Kellermann SA, Hudak S, Oldham ER, Liu YJ,McEvoy LM: The CC chemokine receptor-7ligands 6Ckine and macrophage inflammatoryprotein-3ß are potent chemoattractants for invitro- and in vivo-derived dendritic cells. JImmunol 1999;162:3859–3864.

27 Chan VWF, Kothakota S, Rohan MC, Panga-niban-Lustan L, Gardner JP, Wachowicz MS,Winter JA, Williams LT: Secondary lymphoid-tissue chemokine (SLC) is chemotactic for ma-ture dendritic cells. Blood 1999;93:3610–3616.

28 Hirschowitz EA, Weaver JD, Hidalgo GE, Do-herty DE: Murine dendritic cells infected withadenovirus vectors show signs of activation.Gene Ther 2000;7:1112–1120.

29 Song W, Tong Y, Carpenter H, Kong HL, Crys-tal RG: Persistent, antigen-specific, therapeuticantitumor immunity by dendritic cells geneti-cally modified with an adenoviral vector toexpress a model tumor antigen. Gene Ther2000;7:2080–2086.

30 Eberl G, Widmann C, Corradin G: The func-tional half-life of H-2Kd-restricted T cell epi-topes on living cells. Eur J Immunol 1996;26:1993–1999.

31 Termeer C, Benedix F, Sleeman J, Fieber C,Voith U, Ahrens T, Miyake K, Freudenberg M,Galanos C, Simon JC: Oligosaccharides ofHyaluronan activate dendritic cells via toll-likereceptor 4. J Exp Med 2002;195:99–111.

32 Klein C, Bueler H, Mulligan RC: Comparativeanalysis of genetically modified dendritic cellsand tumor cells as therapeutic cancer vaccines.J Exp Med 2000;191:1699–1708.