2001;61:7925-7933. Cancer Res   Cho-Lea Tso, Amnon Zisman, Allan Pantuck, et al.   Stimulating FactorProtein Consisting of G250 and Granulocyte/Monocyte-ColonyActivity against Renal Cell Carcinoma Using a Chimeric Fusion Induction of G250-targeted and T-Cell-mediated Antitumor

  Updated version

  http://cancerres.aacrjournals.org/content/61/21/7925

Access the most recent version of this article at:

   

   

  Cited Articles

  http://cancerres.aacrjournals.org/content/61/21/7925.full.html#ref-list-1

This article cites by 47 articles, 21 of which you can access for free at:

  Citing articles

  http://cancerres.aacrjournals.org/content/61/21/7925.full.html#related-urls

This article has been cited by 8 HighWire-hosted articles. Access the articles at:

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  .permissions@aacr.orgDepartment at

To request permission to re-use all or part of this article, contact the AACR Publications

Research. on August 17, 2014. © 2001 American Association for Cancercancerres.aacrjournals.org Downloaded from

Research. on August 17, 2014. © 2001 American Association for Cancercancerres.aacrjournals.org Downloaded from

[CANCER RESEARCH 61, 7925–7933, November 1, 2001]

Induction of G250-targeted and T-Cell-mediated Antitumor Activity against RenalCell Carcinoma Using a Chimeric Fusion Protein Consisting of G250 andGranulocyte/Monocyte-Colony Stimulating Factor

Cho-Lea Tso, Amnon Zisman, Allan Pantuck, Randy Calilliw, Jose M. Hernandez, Sun Paik, David Nguyen,Barbara Gitlitz, Peter I. Shintaku, Jean de Kernion, Robert Figlin, and Arie Belldegrun1

Departments of Urology [C-L. T., A. Z., A. P., R. C., J. M. H., S. P., D. N., B. G., J. d. K., R. F., A. B.], Medicine [R. F.], and Pathology [P. I. S.] and Jonsson ComprehensiveCancer Center [C-L. T., A. Z., A. P., R. C., J. M. H., S. P., D. N., B. G., P. I. S., J. d. K., R. F., A. B.], UCLA Kidney Cancer Program, University of California Los Angeles, LosAngeles, California 90095

ABSTRACT

Immunotherapy targeting for the induction of a T-cell-mediated anti-tumor response in patients with renal cell carcinoma (RCC) appears tohold significant promise. Here we describe a novel RCC vaccine strategythat allows for the concomitant delivery of dual immune activators: G250,a widely expressed RCC associated antigen; and granulocyte/macro-phage-colony stimulating factor (GM-CSF), an immunomodulatory factorfor antigen-presenting cells. The G250-GM-CSF fusion gene was con-structed and expressed in Sf9 cells using a baculovirus expression vectorsystem. The Mr 66,000 fusion protein (FP) was subsequently purifiedthrough a 6xHis-Ni2�-NTA affinity column and SP Sepharose/fast proteinliquid chromatography. The purified FP retains GM-CSF bioactivity,which is comparable, on a molar basis, to that of recombinant GM-CSFwhen tested in a GM-CSF-dependent cell line. When combined withinterleukin 4 (IL-4; 1000 units/ml), FP (0.34 �g/ml) induces differentiationof monocytes (CD14�) into dendritic cells (DCs) expressing surface mark-ers characteristic for antigen-presenting cells. Up-regulation of matureDCs (CD83�CD19�; 17% versus 6%) with enhanced expression of HLAclass I and class II antigens was detected in FP-cultured DCs as comparedwith DCs cultured with recombinant GM-CSF. Treatment of peripheralblood mononuclear cells (PBMCs) with FP alone (2.7 �g/107 cells) aug-ments both T-cell helper 1 (Th1) and Th2 cytokine mRNA expression(IL-2, IL-4, GM-CSF, IFN-�, and tumor necrosis factor-�). Comparisonof various immune manipulation strategies in parallel, bulk PBMCstreated with FP (0.34 �g/ml) plus IL-4 (1000 units/ml) for 1 week andrestimulated weekly with FP plus IL-2 (20 IU/ml) induced maximalgrowth expansion of active T cells expressing the T-cell receptor andspecific anti-RCC cytotoxicity, which could be blocked by the addition ofanti-HLA class I, anti-CD3, or anti-CD8 antibodies. In one tested patient,an augmented cytotoxicity against lymph node-derived RCC target wasdetermined as compared with that against primary tumor targets, whichcorresponded to an 8-fold higher G250 mRNA expression in lymph nodetumor as compared with primary tumor. The replacement of FP withrecombinant GM-CSF as an immunostimulant completely abrogated theselection of RCC-specific killer cells in peripheral blood mononuclear cellcultures. All FP-modulated peripheral blood mononuclear cell cultureswith antitumor activity showed an up-regulated CD3�CD4� cell popula-tion. These results suggest that GM-CSF-G250 FP is a potent immunos-timulant with the capacity for activating immunomodulatory DCs andinducing a T-helper cell-supported, G250-targeted, and CD8�-mediatedantitumor response. These findings may have important implications forthe use of GM-CSF-G250 FP as a tumor vaccine for the treatment ofpatients with advanced kidney cancer.

INTRODUCTION

Metastatic RCC2 poses a therapeutic challenge because of its re-sistance to conventional modes of therapy, such as chemotherapy andradiation (1). Advances in the treatment of metastatic RCC haveevolved significantly in the last decade since the Food and DrugAdministration’s approval of IL-2 in 1992. It has become clear thatimmunotherapy is capable of producing durable remissions in selectedRCC patients; yet, the overall response rates of immunotherapy re-main 25% at best (2), at the cost of considerable toxicities to thepatient. The recent identification of MHC-restricted TAAs and theunderstanding of the critical role of immunomodulatory DCs haveprovided the rationale for the development of tumor vaccines forcancer therapy (3, 4). Many cancer vaccine strategies have beendesigned and tested in both animal models and clinical trials withencouraging results. These include peptide-based (5, 6), DC-based(7–11), recombinant viruses/DNA/RNA-based (12–14), and genemodified tumor cell-based (15–17) vaccines. Despite the fact thatRCC is thought to be a relatively immunogenic tumor, no RCC-associated antigens have been identified and characterized in associ-ation with a significant rationale for the development of a kidneycancer-targeted tumor vaccine (18–20).

The first widely expressed RCC TAA that contains HLA-A2-restricted CTL epitopes has been recently identified and cloned froma RCC cell line (21, 22). This RCC-associated transmembrane protein,designated as G250, has been proven to be identical to MN/CAIX, acell adhesion molecule that was first identified in cervical cancer andthat contains carbonic anhydrase activity (23, 24). Immunohistochem-ical staining with monoclonal antibody G250 revealed that �75% ofprimary and metastatic RCCs expressed G250, whereas little to noexpression was detected in normal kidney (21). G250 expression isfound in nearly all clear cell carcinomas of the kidney, the mostcommon RCC variant, which provides further basis for the use ofG250 as a significant immune target for anticancer immunotherapy.Antigen presentation is a crucial and initial step for vaccine-basedimmunotherapy to be achieved. We therefore hypothesized that achimeric protein consisting of G250 and GM-CSF, an immunomodu-latory factor for the generation of functional DCs, would augmentvaccine potency as compared with the use of either agent alone.Several chimeric FPs containing GM-CSF have been reported andshown to exert a variety of complex biological effects dependent ontheir components (25–28). GM-CSF has been well characterized as agrowth factor that induces the proliferation and maturation of myeloidprogenitor cells (29). It enhances macrophage and granulocyte naturalcytotoxicity against tumor cells (30). The function of GM-CSF as a

Received 3/8/01; accepted 9/4/01.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1 To whom requests for reprints should be addressed, at Department of Urology,UCLA School of Medicine, CHS 66-118, Los Angeles, CA 90095-1738. Phone: (310)206-1434; Fax: (310) 206-5343; E-mail: abelldegrun@mednet.ucla.edu.

2 The abbreviations used are: RCC, renal cell carcinoma; IL, interleukin; DC, dendriticcell; GM-CSF, granulocyte/macrophage-colony stimulating factor; FP, fusion protein;FACS, fluorescence-activated cell sorting; PBMC, peripheral blood mononuclear cell;LN, lymph node; LU, lytic unit(s); APC, antigen-presenting cell; TAA, tumor-associatedantigen; FPLC, fast protein liquid chromatography; PE, phycoerythrin; RT-PCR, reversetranscription-PCR.

7925

Research. on August 17, 2014. © 2001 American Association for Cancercancerres.aacrjournals.org Downloaded from

key factor for the differentiation of DCs further substantiates its use inimmune-based vaccine therapy (31). Evidence for the immunoadju-vant impact of GM-CSF in vaccine-based immunotherapy has beendemonstrated in animal models. Cutaneous immunization with tumorpeptide at sites containing epidermal DCs newly recruited by pretreat-ment with DNA encoding GM-CSF elicited an antigen-specific T-cellresponse, whereas peptide immunization of control skin sites showedno immune response (32). Likewise, treatment of established tumorwith a cellular vaccine consisting of GM-CSF gene-modified DCshybridized with malignant melanoma cells showed superior therapeu-tic efficacy to the treatment with the same vaccine generated withnonmodified DCs (33). A Phase I trial further demonstrated thatsystemic injection of GM-CSF and IL-4 was capable of inducingtumor regression or stable disease response in patients with advancedRCC and prostate cancer (34). Similarly, vaccination of patients withirradiated autologous RCCs or melanoma cells engineered to secretehuman GM-CSF also induced potent antitumor immunity (16, 17).

In this study, we describe our strategy to generate a FP consistingof G250 and GM-CSF. We tested its feasibility, bioactivity, andspecificity as a nonviral, noncellular RCC tumor vaccine and defineits role as an immunostimulant of a G250-targeted, antitumor responsein PBMC cultures derived from patients with advanced kidney cancer.

MATERIALS AND METHODS

Cloning of GM-CSF-G250 Fusion Gene in pVL1393 Vector. Plasmidp91023(B)-GM-CSF (kindly provided by Dr. Judith Gasson, UCLA School ofMedicine) was digested with EcoRI, and the 0.8-kb fragment containing thefull-length GM-CSF cDNA was used to generate the 0.4-kb GM-CSF frag-ment, containing the functional epitope flanked by the EcoRI site on the 5� sideand NotI site on the 3� side replacing the GM-CSF stop codon, by DNA PCR.The GM-CSF fragment from the PCR product was subcloned into the EcoRIand BglII sites of the polyhedrin gene locus-based baculovirus transfer vectorpVL1393 (PharMingen, San Diego, CA). Similarly, pBM20CMVG250 (kindlyprovided by Dr. Egbert Oosterwijk, University Hospital Nijmegen, Nijmegen,the Netherlands) was used to amplify the full-length G250 cDNA (1.6-kb)containing the NotI site, followed by a six-nucleotide linker coding for twoarginines by PCR in the 5�-flanking region of the G250 after the removal of itsstart codon. The 3�-flanking region of G250 was designed to encode sixhistidines followed by the stop codon and BglII site. The G250 fragment wasgel purified. Both the vector pVL1393 already containing the GM-CSF and theG250 PCR amplified fragments were cut out with NotI and BglII. The G250fragment and the vector were ligated for 3 h at 16°C and later transformed andplated on LB plates. The colonies containing the correct plasmid were purifiedby cesium chloride buoyant ultracentrifugation. The plasmid was cut with a setof different restriction enzymes to verify the plasmids. The plasmid cloneswere further verified for histidine tag using an Amplicycle Sequencing kit(Perkin Elmer).

Generation and Purification of FP. The recombinant baculovirus con-taining the His-tagged GM-CSF-G250 fusion gene was generated by cotrans-fection of 0.5 �g of BaculoGold DNA (modified AcNPV baculovirus DNA;PharMingen) and 5 �g of pVL1393/GM-CSF-G250 in Sf9 cells (Spodopterafrugiperda). Viruses were further amplified at a low multiplicity of infection(�1) in adherent Sf9 insect cells, and the titers of the virus were determinedby plaque assay. Expression of GM-CSF-G250 FP in Sf9 cells was determinedby immunocytochemical analysis using anti-G250 monoclonal antibody (kind-ly provided by Dr. Egbert Oosterwijk), anti-GM-CSF antibody (Genzyme,Cambridge, MA), and irrelevant control antibody. Sf9 cells infected withpVL1392-XylE recombinant virus (PharMingen) and uninfected Sf9 cells wereused as negative controls for FP expression and analysis. The viruses used forprotein production were isolated and amplified from a single plaque. Celllysate was prepared from the Sf9 cells infected with viruses at a multiplicity ofinfection of 5 for 3 days with insect cell lysis buffer containing a proteaseinhibitor mixture (PharMingen). Filtered lysate (0.22 �m filter) was applied toa Ni2�-NTA agarose column with high affinity for 6xHis (Qiagen, SantaClarita, CA). After extensive washing of the column (50 mM sodium phos-

phate, 300 mM NaCl, and 10% glycerol, pH 8.0), the FP was stepwise elutedin a column (50 mM sodium phosphate, 300 mM NaCl, and 10% glycerol, pH6.0) at 4°C using increasing concentrations of imidazole from 0.1 M up to 0.5M. Fractions were analyzed with Western blot using anti-GM-CSF antibody.The peak fractions were combined, dialyzed, and resubmitted to Ni2�-NTAagarose column. Fractions containing FP were pooled, dialyzed, and furtherapplied to a FPLC column containing SP Sepharose (Amersham PharmaciaBiotech, Piscataway, NJ). FP was eluted with an increasing salt gradient from50 mM to 1 M NaCl in buffer X (20 mM Tris, 1 mM EDTA, and 10% glycerol),and the fractions containing FP were pooled, dialyzed, and sterilized through0.2 �m filter. The Coomassie blue and silver stains were used to analyze thepurity of GM-CSF-G250 FP. The protein concentration was determined byBio-Rad DC Protein assay (Bio-Rad, Hercules, CA).

GM-CSF Bioassay. The biological activity of the GM-CSF component ofthe FP was determined by measuring the proliferation of GM-CSF-dependentTF-1 cells (35) in the presence of FP. TF-1 cells (2 � 104 cells/well) seededin triplicate in 96-well plates in culture medium (RPMI 1640 � 10% FBS)contained titrated concentrations of FP or the corresponding molar concentra-tions of recombinant human GM-CSF (rh-GM-CSF; Genetics Institute, Bos-ton, MA). Cultures were incubated for 5 days, and [3H]thymidine (0.1 �Ci/well) was added 12 h before harvest. The incorporated [3H]thymidine wasmeasured by scintillation counting with a beta counter.

Phenotypic Analysis of DCs by FACS. The phenotype of DCs generatedfrom both adherent and bulk PBMCs was determined by two-color immuno-fluorescence staining as described in our previous report (35). Both adherentPBMCs and nonfractionated bulk PBMCs were cultured with IL-4 (1000units/ml; R&D Systems, Minneapolis, MN) plus either GM-CSF (800 units/ml; Genetics Institute, Boston, MA) or FP (0.34 �g/ml; equivalent to 800units/ml GM-CSF) for 7 days, and the identity of the DCs was determined. Cellcultures (1 � 105 cells) were resuspended in 50 �l of FACS buffer (PBS, 2%new born calf serum, and 0.1% sodium azide) and incubated with 10 �l of theappropriate FITC- or PE-labeled monoclonal antibodies for 30 min at 4°C.After staining, cells were washed twice with PBS and resuspended in 200 �lof FACS buffer plus 200 �l of 2% paraformaldehyde. Five to 10,000 events/sample were acquired on a Becton-Dickinson FACScan II flow cytometer thatsimultaneously acquires forward and side scatter, as well as FL1 (FITC) andFL2 (PE) data, and analyzed utilizing the CellQuest Software (Becton-Dick-inson, San Jose, CA). Settings for all parameters were optimized at theinitiation of the study and were maintained constant throughout all subsequentanalyses. The DC population in bulk PBMC culture was gated based on theirsize and granularity (36). The voltage on the photomultiplier tubes wasdecreased when the DC population was analyzed, because large cells tend tohave greater autofluorescence. In all samples, the position of quadrant cursorswas determined by setting them on samples stained with the appropriateisotype control antibody. The following antibodies were used for character-ization of the DC phenotype: anti-CD86 (B7-2; PharMingen); anti-CD40(Caltag, Burlingame, CA); anti-HLA class I (W6/32, ATCC HB95); anti-HLA-DR (Immunocytometry System; Becton Dickinson, Mountain View,CA); anti-CD14 (Catlag Laboratories, San Francisco, CA); and isotype controlIgG1/IgG2a (Beckton Dickinson). The CD83� surface marker was used todelineate the maturation of DCs. To discriminate DCs (CD83�CD19�) fromactivated B cells (CD83�CD19�), dual color staining using CD19FITC andCD83PE (Immunotech, Marseille, France) was performed.

Semiquantitative RT-PCR Analysis of Cytokine Profile in PBMCs.Total RNA was extracted from PBMCs treated with FP (2.7 �g/107 cells) forvarious time intervals up to 24 h at 37°C, using acid guanidine isothiocyanate-phenol-chloroform extraction. Reverse transcription of messenger RNA intocDNA was carried out by incubating titrated RNA with avian myeloblastosisvirus reverse transcriptase, primer oligo(dT), deoxynucleotide triphosphate,and RNAse inhibitor at 42°C for 1 h. One �l of each cDNA sample wasamplified using PCR in a total volume of 25 �l (30 ng of [32P]5�-oligonu-cleotide, 100 ng of 3�-oligonucleotide primer, 2.5 �l of modified 10� PCRbuffer, 1.25 units of Taq polymerase, and autoclaved double distilled water toa volume of 25 �l). The PCR mixture was amplified for 25 cycles in a DNAThermocycler (Perkin-Elmer, Norwalk, CT). Each cycle consisted of denatur-ation at 94°C for 1 min and annealing/extension at 65°C for 2 min. The32P-labeled PCR products were then visualized directly via acrylamide gelelectrophoresis and autoradiography and then quantitated by excision of bandsand subsequent scintillation counting. The signal intensity of each amplified

7926

GENERATION OF RENAL CELL CARCINOMA TUMOR VACCINE

Research. on August 17, 2014. © 2001 American Association for Cancercancerres.aacrjournals.org Downloaded from

product was calibrated to its corresponding �-actin mRNA expression as aninternal control for quantitation of expression levels. In addition, quantitativeanalysis was further elucidated by a serial dilution of mRNA (1:3, 1:10, 1:30,and 1:300) and coamplification of �-actin and GM-CSF mRNA. The sequences ofthe oligonucleotide primer pairs were as follows: �-actin: 5�-CAACTCCATCAT-GAAGTGTGAC, 3�-CCACACGGAGTACTTGCGCTC; GM-CSF: 5�-CCAT-GATGGCCAGCCACTAC, 3�-CTTGTTTCATGAGAGAGCAGC, TNF-�:5�-TCTCGAACCCCGAGTGACAA, 3�-TACGACGGCAAGGATTACATC;IFN-�: 5�-ATGAAATATACAAGTTATATCTTGGCTTT, 3�-ATGCTCTT-CGACCTCGAAACAGCAT; IL-2: 5�-GGAATTAATAATTACAAGAAT-CCC, 3�-GTTTCAGATCCCCTTTAGTTCCAG; IL-4: 5�-CTTCCCCCTCT-GTTCTTCCT, 3�-TTCCTGTCGAGCCGTTTCAG.

Immunomodulation of PBMCs with FP. Fresh isolated PBMCs frompatients with RCC expressing G250 were cultured in RPMI 1640 supple-mented with 10% autologous serum. Various schedules of immunomodulatoryprotocols of PBMC cultures with FP were carried out as described in Table 1and Fig. 5. The growth of PBMCs was determined by cell count, and thecytolytic activity of PBMCs was assayed for different targets in a prolonged18-h 51Cr release assay to enhance the detection of T-cell activity (37). Fivethousand 51Cr-labeled target cells/well were seeded in a 96-well microtiterplate (Costar, Cambridge, MA) and mixed with PBMCs (E:T ratios 40:1, 20:1,10:1, and 5:1). Cytotoxicity was expressed as LU per 106 effector cells, withthe LU being defined as the number of effector cells that induce 30% lysis.T-cell-mediated and RCC-specific cytotoxicity was confirmed by blockingassays in which targeted autologous tumor cells were pretreated with anti-HLAclass I, class II, or PBMCs were pretreated anti-CD3, anti-CD4, anti-CD8, orisotype control antibody (Becton Dickinson) for 30 min at 4°C before theaddition of cells to cytotoxicity culture plates. Spontaneous release of alltargets was �20% of the corresponding maximal release of 51Cr release. Thefollowing target cells were used: autologous normal kidney cells (G250�),autologous RCC tumor cells (G250�), allogeneic RCC cells (G250�), R11(allogenic RCC with minimal G250 expression), allogeneic prostate cells(CL-1), and human fibroblasts (hFb).

RESULTS

Generation of GM-CSF-G250 FP from Baculovirus-infectedSf9 Cells. Baculovirus expression technology and the 6xHis affinitypurification system were used to generate GM-CSF-G250 FP asdescribed in “Materials and Methods.” Evidence for successful genecloning and generation of recombinant baculovirus was provided byimmunohistochemical staining of infected Sf9 cells using anti-GM-CSF and anti-G250. Abundant G250 and GM-CSF protein expressionwere detected in Sf9 cells that were infected with GM-CSF-G250recombinant baculovirus (Fig. 1A), whereas expression of neitherGM-CSF or G250 was detected in noninfected cells (Fig. 1A) or cellsinfected with pVL1392-XylE recombinant viruses (data not shown).Western blot analysis was used to evaluate the efficiency of the 6xHisaffinity tag in FP for Ni2�-NTA agarose. The expected Mr 66,000band that was detected with anti-GM-CSF appeared in fractionsnumbered 5–25, with a peak concentration in fractions numbered15–19 (Fig. 1B). The protein purity was further improved by a rerunof positive fractions through the Ni2�-NTA agarose column, and thenthe proteins were subjected to FPLC using SP Sepharose column. A

major single Mr 66,000 band was detected in SDS-PAGE analysisstained with Coomassie blue (Fig. 1C).

Purified GM-CSF-G250 FP Retained GM-CSF Bioactivity. Todetermine whether the bioactivity of the GM-CSF was retained in thepurified FP, the FP was analyzed for its ability to support the prolif-eration of a GM-CSF-dependent cell line, TF-1. Serial dilutions of FPwere performed to span the effective concentration range. The resultsfrom the [3H]thymidine incorporation assay demonstrated that the FPcould support TF-1 cell growth in a biphasic dose-dependent manner(Fig. 2). When compared with recombinant GM-CSF (Fig. 2), com-parable bioactivity was obtained with the FP and equivalent molarconcentrations of recombinant GM-CSF in the range between 0 and6.71 ng/ml (0–30.2 ng/ml FP). However, there was a �1 log differ-ence in the specific activities/mol of GM-CSF and FP (favoring FP)when half-maximal activities were used for comparison. In the pres-ence of FP concentrations �30.2 ng/ml, the growth of TF-1 exceededrecombinant GM-CSF-mediated growth by 1.3-fold.

Immunomodulatory Effect of FP on APCs in PBMC Culture.To study how the FP could affect the development of DCs, PBMCsderived from patients with RCC were cultured in the presence of FP(0.34 �g/ml) plus IL-4 (1000 units/ml) for 7 days and compared withthat cultured in GM-CSF (800 units/ml) plus IL-4. FACS analysisrevealed a high percentage of large granulocytes expressing B7-2�,CD40�, and HLA-DR� in both conditions, whereas CD14� cellswere negligible (Fig. 3A). However, when compared with DCs thatwere cultured with recombinant cytokines, an enhanced expression ofboth HLA class I (mean relative linear fluorescence intensity, 4830versus 3215) and HLA class II (6890 versus 6290) was detected in theFP-modulated DC cultures (Fig. 3B). Moreover, a 3-fold increase inthe percentage of mature DCs (CD83�CD19�) was detected in FP-modulated DC cultures (Fig. 3C). This observation was consistent inseveral bulk PBMC cultures derived from three RCC patients and twohealthy donors. A similar FP-mediated immunomodulatory profilewas also obtained from the conventional adherent DC cultures treatedwith IL-4 and FP (data not shown). A lesser efficiency of DCdifferentiation was observed when DCs were cultured in the presenceof FP alone without IL-4; a mix of the CD14� and CD14�B7-2� cellpopulation was determined on day 7 (data not shown).

FP Induces Activation of Cytokine Genes in PBMCs. To iden-tify whether the FP has a direct effect on the regulation of cytokinegenes in PBMCs, freshly isolated PBMCs that were derived fromRCC patients were treated with FP alone (2.7 �g/107 cells) andcompared with those that were treated with GM-CSF alone. Thekinetics of cytokine gene activation was followed by analysis ofmultiple cytokine mRNA expression through the time course indi-cated in Fig. 4. Treatment of uncultured PBMCs with FP graduallyenhanced GM-CSF, tumor necrosis factor-�, IFN-�, IL-4, and IL-2mRNA expression with the peak level at 24 h after treatment, exceptIL-4. Peak of IL-4 mRNA expression was detected at 6 h after thetreatment (Fig. 4). A similar cytokine gene activation profile was alsodetected in cells that were treated with recombinant GM-CSF. How-

Table 1 Phenotypic modulation of bulk PBMCs by fusion protein (FP)

Phenotype (%) Precultured IL-2 IL-2 � GM-CSF IL-2 � FPIL-2 � IL4 � FPa

2 IL-2 � FPFP 2 IL-2 � FP

FP � IL-42 IL-2 � FP

CD56�CD3� 25 13 25 11 4 9 1CD56�CD3� 46 47 67 70 84 88 96CD4�CD8� 31 28 58 39 42 66 46CD4�CD8� 3 10 10 20 29 4 28CD4�CD8� 22 24 16 31 24 22 25CD3�TcR� 40 45 60 72 69 79 96CD3�CD25� 19 43 60 61 54 17 86

a 40 IU/ml IL-2; 1000 IU/ml IL-4; 0.34 �g/ml FP.

7927

GENERATION OF RENAL CELL CARCINOMA TUMOR VACCINE

Research. on August 17, 2014. © 2001 American Association for Cancercancerres.aacrjournals.org Downloaded from

ever, the peak expression level was detected between 1 and 6 h oftreatment.

FP Induces T-Cell-mediated, G250-targeted Immune Responsein PBMC Cultures. Six immunomodulatory protocols with andwithout FP were tested and compared in PBMC cultures. Theseculture conditions included: (a) IL-2 alone (40 IU/ml); (b) IL-2 � GM-CSF (800 units/ml); (c) IL-2 � FP (0.34 �g/ml; restimulatedweekly); (d) IL-2 � IL-4 (1000 units/ml) � FP for 1 week and thenrestimulated with FP� IL-2; (e) FP alone for 1 week and then restimu-lated with FP � IL-2; and ( f ) FP � IL-4 for 1 week and thenrestimulated with IL-2 � FP. As indicated in Fig. 5 (patient 1), amongvarious immunomodulatory conditions tested, pretreatment of PBMCswith FP plus IL-4 for 1 week and subsequent restimulation with IL-2(40 IU/ml) and FP weekly showed the highest growth expansion(6.0�; Fig. 5A). A similar enhanced growth profile with this particularcondition was determined in three PBMC cultures derived fromadditional patients with RCC. Notably, an additive growth effect ofGM-CSF with IL-2 was detected when compared with IL-2 alone(Fig. 5). GM-CSF alone, however, failed to propagate PBMCs inlong-term culture. Patient 1 had RCC metastasis in a LN. Enhancedcytotoxicity against LN-derived tumor target was determined in allfour FP-modulated PBMC cultures (three cycles of restimulation)when compared with the cytotoxicity against the primary tumor target(Fig. 5B). Remarkably, this enhanced killing activity correspondedwith an 8-fold increase in G250 mRNA expression in the LN metas-tasis over the primary tumor, as determined by a semiquantitative

Fig. 1. Expression and purification of GM-CSF-G250 FP. A, immunohistochemical staining for G250 and GM-CSF expression with anti-G250 and anti-GM-CSF antibodies in Sf9cells infected with and without fusion gene recombinant baculovirus. �100. B, Western blot analysis of 6xHis-tagged GM-CSF-G250 FP eluted from the Ni2�-NTA affinity columnusing anti-GM-CSF antibody. L, loading; BT, break through; W, wash. C, Coomassie blue-stained SDS-PAGE of fusion protein eluted from Ni2�-NTA affinity column (Lane 1) andfurther purified with SP Sepharose/FPLC (Lanes 2 and 3; two batches of final products). MW, molecular weight.

Fig. 2. Comparison of functional activity of recombinant GM-CSF and purifiedGM-CSF-G250 FP. GM-CSF activity was measured using the GM-CSF-dependent humancell line, TF-1. The TF-1 cells (2 � 104/well/ml) were cultured in the presence of aserially diluted amount of recombinant GM-CSF or purified GM-CSF-G250 FP asindicated. After a 5-day incubation, the cultures were pulsed with 0.1 �Ci of tritiatedthymidine for an additional 12 h. The cultures were then harvested, and the incorporatedthymidine was measured by scintillation counting.

7928

GENERATION OF RENAL CELL CARCINOMA TUMOR VACCINE

Research. on August 17, 2014. © 2001 American Association for Cancercancerres.aacrjournals.org Downloaded from

RT-PCR (Fig. 5D). By pretreating the LN tumor target cells withanti-HLA class I (77%) or, alternatively, by pretreating the effectorcells with anti-CD3 (66%) or anti-CD8 (55%) prior to the cytotoxicityassay, RCC-targeted cytotoxicity was markedly reduced. On the otherhand, anti-HLA class II (33%) or anti-CD4 (33%) treatment led onlyto a lesser inhibition of cytotoxicity (Fig. 5C). Although poor growthexpansion (1.8�) was detected in the condition in which PBMCs werepretreated with FP alone for 1 week and then restimulated with IL-2plus FP, the highest cytotoxicity against both primary and LN-derived

RCC targets was detected when compared with other tested conditions(Fig. 5B). High killing activity against K562 was detected in mostearly cultures of FP-modulated PBMCs, which was noted to graduallydecrease, whereas the killing activity against autologous RCC wasmaintained. In contrast, both non-FP-stimulated cultures, IL-2, andIL-2 plus GM-CSF showed a nonspecific killing activity with rela-tively high cytotoxicity against K562 (Fig. 5B).

To identify the phenotypic identity of FP-modulated PBMCs thatpossess antitumor activity, phenotypic analysis was performed on the

Fig. 3. Immunomodulatory effects of FP on DCs. A, double-color flow cytometric analysis of DCs grown in GM-CSF (800 units/ml) plus IL-4 (1000 units/ml) or FP plus IL-4 for7 days. Cells were labeled with FITC- and PE-conjugated antibodies against cell surface markers of DCs, as indicated. Cells that were larger than lymphocytes were selectively gated.Negative controls correspond to labeling with an isotype-matched control antibody (Ig�1/Ig�2a). This analysis is representative of five DC cultures. B, flow cytometric analysis of HLAantigens of DCs cultured in GM-CSF plus IL-4 or FP plus IL-4. Cells are labeled with primary antibody (HLA class I or class II) and FITC-conjugated secondary antibody. This analysisis representative of four different DCs derived from four RCC patients. C, double-color flow cytometric analysis of DCs expressed CD83�CD19� that were cultured in the conditionas indicated. Data, means of triplicate; bars, SD. This analysis is representative of four different DC cultures obtained from four RCC patients.

Fig. 4. Time course of cytokine mRNA expres-sion in PBMCs that were treated with GM-CSF-G250 FP (2.7 �g/107 cells) for various time periodsas indicated and then harvested for semiquantita-tive RT-PCR analysis. The 32P-labeled PCR prod-ucts were separated by electrophoresis through a7% acrylamide gel. Gels were dried and subjectedto autoradiography. Titrated standard was preparedfrom diluted RNA samples extracted from PBMCstreated with FP for 24 h.

7929

GENERATION OF RENAL CELL CARCINOMA TUMOR VACCINE

Research. on August 17, 2014. © 2001 American Association for Cancercancerres.aacrjournals.org Downloaded from

day when cytotoxicity was determined. A markedly increased T-cellpopulation (70–94%) expressing T-cell receptor (72–96%) was de-tected in all FP-stimulated PBMC cultures, when compared withprecultured (40%) or cytokine-cultured PBMCs (IL-2, 45%; IL-2 � GM-CSF, 60%; Table 1). Notably, the T-cell population express-ing the most IL-2 receptor (CD3�CD25�; 86%) was found to occurafter pretreatment with FP plus IL-4 followed by restimulation withIL-2 and FP. This also corresponded with the greatest T-cell growthexpansion with this specific condition compared with all other testedimmunomodulatory protocols (Fig. 5A). Correspondingly, thosePBMCs pretreated with FP for 1 week demonstrated a minimal T-cellpopulation expressing IL-2 receptor (19%) with the lowest growthexpansion (Table 1 and Fig. 5A). In contrast, PBMCs cultured in the

absence of FP (IL-2 � GM-CSF) expressed a higher percentage ofCD56� (13%, 20%), which corresponded with a higher natural killeractivity against K562 when compared with FP-modulated cultures(Fig. 5B).

Replacement of FP with GM-CSF Abrogates the Selection ofRCC-targeted Cytotoxic T Cells. To confirm that the component ofG250 in the FP is the determinant for the growth selection of CTLsagainst RCC, a cytotoxicity assay was performed with PBMCs thatwere cultured in the presence of GM-CSF and IL-4 for 1 week andthen continuously restimulated with IL-2 and GM-CSF (800 units/ml).Minimal cytotoxicity against autologous RCC was determined in alltested PBMC cultures without FP stimulation, whereas the corre-sponding PBMC cultures stimulated with FP showed an MHC-restricted, T-cell-mediated cytotoxicity against autologous RCC (day35; Fig. 6A) that expressed a high level of G250 (data not shown). The

Fig. 5. Growth and cytotoxicity profiles of patient-derived PBMCs stimulated withGM-CSF-G250 FP. A, growth expansion of PBMCs (patient 1) induced by variousimmunomodulatory strategies as indicated. Cell cultures were stimulated with FP on days0, 6, 12, and 18. Culture medium was changed weekly but maintained in a constantvolume. Cell counts were performed on day 20. Expansion fold was calculated by thedivision of final cell counts/ml with cell counts/ml seeded on day 0 (3 � 105 cells/ml).Data, means of triplicate; bars, SD. This analysis is representative of four different PBMCcultures derived from four RCC patients showing a similar growth profile. B, cytotoxicityof PBMCs (patient 1) against autologous normal kidney cells, primary tumor cells, andLN-derived tumor cells. Cytotoxicity was determined by 18-h 51Cr-release assay on day21. Killing activity was expressed as LU per 106 effector cells. LU is defined as thenumber of effector cells capable of inducing 30% lysis. Spontaneous release for tumortarget was �20% of the corresponding maximal release. Data, means of triplicate; bars,SD. C, inhibition of cytotoxicity against autologous LN tumor cells by antibodies specificto T-cell markers and HLA antigens. Tumor target cells or PBMCs were pretreated withthe respective antibody as indicated prior to the cytotoxicity assay. Data, means oftriplicate; bars, SD. D, semiquantitative RT-PCR analysis of G250 mRNA expression bynormal kidney, primary tumor, and LN metastasis obtained from patient 1. Tu, tumor.

Fig. 6. FP-induced, G250-targeted, and MHC-restricted T-cell immunity. A–C, cyto-toxicity of PBMCs against autologous and allogenic tumor targets as indicated. PBMCcultures were pretreated with IL-4 (1000 units/ml) and FP (0.34 �g/ml) or IL-4 andGM-CSF (800 units/ml) for 1 week and then restimulated with IL-2 (40 IU/ml) and FP orIL-2 and GM-CSF weekly. Cytotoxicity was determined by 18-h 51Cr-release assay onday 35. Cytotoxicity against autologous tumor target was measured in the presence ofisotype control antibody or antibodies specific to HLA class I, HLA class II, CD3, CD4,or CD8. Data, means of triplicate; bars, SD. D, flow cytometric analysis of FP-modulatedPBMC phenotypes that expressed antitumor activity (day 35). Bars, SD.

7930

GENERATION OF RENAL CELL CARCINOMA TUMOR VACCINE

Research. on August 17, 2014. © 2001 American Association for Cancercancerres.aacrjournals.org Downloaded from

cytotoxic killing capacity against natural killer-sensitive (K562) cellswas negligible (patient 2, 3.5 LU; patient 3, 2.5 LU; patient 4, 2.0LU). A predominant CD3�CD4� cell population was detected in allthree FP-modulated PBMC cultures (68, 74, and 66%) that expressedantitumor activity when compared with the CD3�CD8� cell popula-tion (25, 22, and 30%; Fig. 6B). Moreover, both Th1 and Th2 cytokinemRNA were detected in these FP-modulated PBMC cultures, whichincluded GM-CSF, tumor necrosis factor-�, IFN-�, IL-2, and IL-4(data not shown).

DISCUSSION

RCC is responsive to immunotherapy. However, no documentedimmune-based treatment protocol is available to date that wouldeffectively eradicate tumor lesions in the majority of patients. Themeans of immune strategy, the type of immune activators used, thegiven dose, the method of administration, and the pretreatment im-mune status of patients all could influence the ultimate immuneresponse in cancer patients that are treated with immune-based ther-apy. Therefore, a critical issue for an effective cancer vaccine is thedevelopment of a potent adjuvant that can facilitate both induction andaugmentation of an immune response with antitumor activity. Towork toward this goal, we proposed a chimeric construct consisting ofG250 and GM-CSF. The demonstration of G250 expression in Sf9cells and GM-CSF bioactivity in the purified Mr 66,000 protein bandconfirmed the efficacy of the gene construct and the effectiveness ofthe protein purification process used. When compared with recombi-nant GM-CSF, the enhanced effectiveness and biphasic behavior ofFP in the stimulation of TF-1 cell growth may be attributable in partto the stabilizing effect of larger molecules that lead to a longerhalf-life, as well as to a reduction in the binding coefficient of thelarger molecule to the GM-CSF receptor. This may also explain inpart the delayed cytokine gene activation in FP-modulated PBMCs.

Antigen presentation by DCs is crucial not only for the induction ofprimary immune responses but is also important for the regulation ofthe type of T-cell-mediated immune response (38). We recently de-veloped a nonfractionated bulk PBMC culture system for the study ofthe maturation and immunomodulatory function of CD14�-derivedDCs and the interaction between the DCs and cocultured lymphocytes(36). Using this system, antigen loading can be performed during theearly culture period of PBMCs in the presence of GM-CSF and IL-4,when immature DC/monocytes can take up and process tumor anti-gen. We have demonstrated previously that DC-modulated, cocul-tured lymphocytes in bulk PBMC culture can be further expanded toCTLs by repetitive stimulation with low-dose IL-2 and RCC tumorlysate (36). Likewise, direct treatment of bulk PBMCs with IL-4 andFP not only induced the differentiation of CD14� cells into DCs butalso increased the maturation of DCs when compared with DCsgenerated in IL-4 and GM-CSF. This suggests that the signalingpathway in DC maturation can be induced by FP stimulation (39).Moreover, when compared with recombinant GM-CSF, up-regulationof HLA antigen expression was determined on FP-modulated DCs,indirectly suggesting that DCs are capable of internalizing, process-ing, and presenting FP determinants. Whether the G250 was taken upby APCs through the GM-CSF component/GM-CSF receptor inter-nalization or via the G250 component or both remains to be deter-mined.

It appears that preincubation of PBMCs with FP and IL-4 prior toexposing IL-2 is favorable to achieve a better effector expansion. Thismay be explained if exogenous IL-4 could synergize the FP for themobilization of DC differentiation and maturation and subsequentlypresent the antigen peptides to the surrounding immune cells. Al-though a successful CTL selection also could be achieved by other

tested immunomodulatory protocols with FP, the growth expansion ofCTLs was not favorable. This may be partly associated with a “de-layed” DC differentiation under the suboptimal concentration of IL-4(FP can induce IL-4 secretion by PBMCs). High expression of IL-2Rand T-cell receptor in FP-modulated immune cells indicates that theywere highly activated and proliferative. In contrast, preexposure ofIL-2 to nonantigen-stimulated PBMCs usually results in the expansionof nonspecific, lymphokine-activated killer cells with short-term kill-ing activity (40). To rule out the possibility of other common TAAsparticipating in the RCC-targeted cytotoxicity, several other TAAswere screened, including RAGE, gp100, PRAME, and NY-ESO-1.Among the tested TAAs, G250 was the only common TAA that wasexpressed by all lysed RCC specimens in the study. The availabilityof purified G250 protein and G250 gene-transduced autologous nor-mal cells would further verify the mechanism of FP-mediated, G250-targeted, and MHC-restricted cytolytic activity. Three of the “non-responding” PBMC cultures were determined. One of them showedlow activity of killing capacity against autologous RCC (day 35),whereas the other two showed a nonspecific killing activity against alltested tumor targets that included K562 and non-RCC tumor targets.Therefore, the optimal use of FP, in terms of dose for the activation ofG250-targeted antitumor response remains to be further tested andcharacterized. The dose of FP used in this study was based on themolar concentration of GM-CSF in the FP as the optimal dose forthe development of DCs we have tested. However, this has limited theability to adjust doses of the G250. Recently, Huang et al. (41)demonstrated that even immunogenic tumors, such as those modifiedto express costimulatory molecules, fail to stimulate the immunesystem unless functional APCs are available to process and presentthe antigens. It thus appears that the most effective anti-cancer vaccinestrategy should target the activation of T cells through the manipula-tion of APCs in patients.

The replacement of FP with an equivalent dose of recombinantGM-CSF abrogated the selection and propagation of RCC-specificCTLs, suggesting that activation and propagation of CTLs is antigen(G250) dependent and that the GM-CSF moiety of the FP has servedas an effective adjuvant for antigen presentation, amplification ofT-cell activity, and induction of cytokine response (42–44). AlthoughFP-induced, G250-targeted antitumor activity is mainly mediated byCD8� T cells, a predominant up-regulation of CD4� T cells wasdetected in most cultures. The FP-mediated Th1 and Th2 cytokinerelease and enhancement of HLA class II expression in DC cellsfurther suggests that FP-mediated antitumor immunity may involvepriming of both CD4� and CD8� T cells specific for G250. In thisresponse, CD4� T helper cells may have a role in providing regula-tory signals required for the priming of MHC class I-restricted CD8�

CTLs (15). Conversely, RCC-specific cytotoxic CD4� cells havebeen reported (45). Therefore, partial blocking of cytotoxicity byanti-CD4 and anti-HLA class II may indicate that the CD4 cellpopulation also participates in lysis of tumor cells.

Studies comparing the efficacy of various formulations of tumorvaccines in parallel demonstrated that the use of DCs transfected withDNA coding for TAAs is superior to peptide-pulsed DCs and nakedDNA-based vaccine for eliciting both antigen-specific CD8� andCD4� T-cell responses (46). This observation indicates that com-puter-predicted peptides might not be naturally processed by DCs,because of their typical expression of immunoproteasomes that pos-sess cleavage specificity for production of immunodominant peptides(47, 48). Thus, some in vitro peptide-reactive T cells could only lysepeptide-pulsed cell targets but not tumor cells expressing the entiretumor antigen (20, 49). Likewise, a peptide-based vaccine couldeffectively elicit expansion of vaccine-specific T cells in PBMCs ofcancer patients, but such response does not necessarily correspond

7931

GENERATION OF RENAL CELL CARCINOMA TUMOR VACCINE

Research. on August 17, 2014. © 2001 American Association for Cancercancerres.aacrjournals.org Downloaded from

with a clinical tumor regression (50). Therefore, immunization withthe current construct of whole G250 antigen may have an advantageover a peptide-pulsed approach by the presentation of multiple orunidentified epitopes in association with MHC class I and class IImolecules by APCs. On the basis of the potency and specificity in theactivation of G250-reactive T cells with antitumor activity, GM-CSF-G250 FP shows excellent promise as a RCC cancer vaccine to justifya Phase I clinical trial for the treatment of patients with advancedkidney cancer. To further explore an alternative G250-targeted anti-cancer strategy, generation of DCs transduced with FP gene has beeninitiated. Using this strategy, we will be able to further investigate theimmune mechanism manipulated by G250.

REFERENCES

1. Figlin, R. A. Renal cell carcinoma: management of advanced disease. J. Urol., 161:381–387, 1999.

2. Fisher, R. I., Rosenburg, S. A., Sznol, M., Parkinson, D. R., and Fyfe, G. High-dosealdesleukin in renal cell carcinoma: long-term survival update. Cancer J. Sci. Am., 3:S70, 1997.

3. Wang, R. F. Human tumor antigen: implications for cancer vaccine development. J.Mol. Med., 77: 640–655, 1999.

4. Xu, M., Qiu, G., Jiang, A., von Hofe, E., and Humphreys, R. E. Genetic modificationof tumor antigen presentation. Trends Biotechnol., 18: 167–172, 2000.

5. Rosenberg, S. A., Yang, J. C., Schwartzentruber, D. J., Hwu, P., Marincola, F. M.,Topalian, S. L., Restifo, N. P., Sznol, M., Schwarz, S. L., Spiess, P. J., Wunderlich,J. R., Seipp, C. A., Einhorn, J. H., Rogers-Freezer, L., and White, D. E. Impact ofcytokine administration on the generation of antitumor reactivity in patients withmetastatic melanoma receiving a peptide vaccine. J. Immunol., 163: 1690–1705,1999.

6. Parkhurst, M. R., Salgaller, M. L., Southwood, S., Robbins, P. F., Sette, A.,Rosenberg, S. A., and Kawakami, Y. Improved induction of melanoma-reactive CTLwith peptides from the melanoma antigen gp100 modified at HLA-A*0201-bindingresidues. J. Immunol., 157: 2539–2548, 1996.

7. Yang, S., Kittlesen, D., Slingluff, C. L., Jr., Vervaert, C. E., Seigler, H. F., andDarrow, T. L. Dendritic cells infected with a vaccinia vector carrying the humangp100 gene simultaneously present multiple specificities and elicit high-affinity Tcells reactive to multiple epitopes and restricted by HLA-A2 and -A3. J. Immunol.,164: 4204–4211, 2000.

8. Condon, C., Watkins, S. C., Celluzzi, C. M., Thompson, K., and Falo, L. D., Jr.DNA-based immunization by in vivo transfection of dendritic cells. Nat. Med., 2:1122–1128, 1996.

9. Zhou, W. Z., Hoon, D. S., Huang, S. K., Fuji, S., Hashimoto, K., Morishita, R., andKaneda, Y. RNA melanoma vaccine: induction of antitumor immunity by humanglycoprotein 100 mRNA immunization. Hum. Gene Ther., 10: 2719–2724, 1996.

10. Nestle, F. O., Alijagic, S., Gilliet, M., Sun, Y., Grabbe, S., Dummer, R., Burg, G., andSchadendorf, D. Vaccination of melanoma patients with peptide- or tumor lysate-pulsed dendritic cells [see comments]. Nat. Med., 4: 328–332, 1998.

11. Mulders, P., Tso, C. L., Gitlitz, B., Kaboo, R., Hinkel, A., Frand, S., Kiertscher, S.,Roth, M. D., de Kernion, J., Figlin, R., and Belldegrun, A. Presentation of renal tumorantigens by human dendritic cells activates tumor-infiltrating lymphocytes againstautologous tumor: implications for live kidney cancer vaccines. Clin. Cancer Res., 5:445–454, 1999.

12. Ulmer, J. B., Fu, T. M., Deck, R. R., Friedman, A., Guan, L., DeWitt, C., Liu, X.,Wang, S., Liu, M. A., Donnelly, J. J., and Caulfield, M. J. Protective CD4� andCD8� T cells against influenza virus induced by vaccination with nucleoproteinDNA. J. Virol., 72: 5648–5653, 1998.

13. Ying, H., Zaks, T. Z., Wang, R. F., Irvine, K. R., Kammula, U. S., Marincola, F. M.,Leitner, W. W., and Restifo, N. P. Cancer therapy using a self-replicating RNAvaccine. Nat. Med., 5: 823–827, 1999.

14. Liu, M. A. Vaccine developments. Nat. Med., 4: 515–519, 1998.15. Mach, N., Gillessen, S., Wilson, S. B., Sheehan, C., Mihm, M., and Dranoff, G.

Differences in dendritic cells stimulated in vivo by tumor engineered to secretgranulocyte-macrophage colony-stimulating factor or Flit3-ligan. Cancer Res., 60:3239–3246, 1999.

16. Simons, J. W., Jaffee, E. M., Weber, C. E., Levitsky, H. I., Nelson, W. G., Carducci,M. A., Lazenby, A. J., Cohen, L. K., Finn, C. C., Clift, S. M., Hauda, K. M., Beck,L. A., Leiferman, K. M., Owens, A. H., Jr., Piantadosi, S., Dranoff, G., Mulligan,R. C., Pardoll, D. M., and Marshall, F. F. Bioactivity of autologous irradiated renalcell carcinoma vaccines generated by ex vivo granulocyte-macrophage colony-stim-ulating factor gene transfer. Cancer Res., 57: 1537–1546, 1997.

17. Soiffer, R., Lynch, T., Mihm, M., Jung, K., Rhuda, C., Schmollinger, J. C., Hodi,F. S., Liebster, L., Lam, P., Mentzer, S., Singer, S., Tanabe, K. K., Cosimi, A. B.,Duda, R., Sober, A., Bhan, A., Daley, J., Neugerg, D., Parry, G., Rokovich, J.,Richards, L., Drayer, J., Berns, A., Clift, S., Cohen, L. K., Mulligan, R. C., andDranoff, G. Vaccination with irradiated autologous melanoma cells engineered tosecrete human granulocyte-macrophage colony-stimulating factor generates potentantitumor immunity in patients with metastatic melanoma. Proc. Natl. Acad. Sci.USA, 95: 13141–13146, 1998.

18. Gaugler, B., Brouwenstijn, N., Vantomme, V., Szikora, J. P., Van der Spek, C. W.,Patard, J. J., Boon, T., Schrier, P., and Van den Eynde, B. J. A new gene coding foran antigen recognized by autologous cytolytic T lymphocytes on a human renalcarcinoma. Immunogenetics, 44: 323–330, 1996.

19. Brandle, D., Brasseur, F., Weynants, P., Boon, T., and Van den Eynde, B. A mutatedHLA-A2 molecule recognized by autologous cytotoxic T lymphocytes on a humanrenal cell carcinoma. J. Exp. Med., 183: 2501–2508, 1996.

20. Brossart, P., Stuhler, G., Flad, T., Stevanovic, S., Rammensee, H. G., Kanz, L., andBrugger, W. Her-2/neu-derived peptides are tumor-associated antigens expressed byhuman renal cell and colon carcinoma lines and are recognized by in vitro inducedspecific cytotoxic T lymphocytes. Cancer Res., 58: 732–736, 1998.

21. Grabmaier, K., Vissers, J. L., De Weijert, M. C., Oosterwijk-Wakka, J. C., VanBokhoven, A., Brakenhoff, R. H., Noessner, E., Mulders, P. A., Merkx, G., Figdor,C. G., Adema, G. J., and Oosterwijk, E. Molecular cloning and immunogenicity ofrenal cell carcinoma-associated antigen G250. Int. J. Cancer, 85: 865–870, 2000.

22. Vissers, J. L., De Vries, I. J., Schreurs, M. W., Engelen, L. P., Oosterwijk, E., Figdor,C. G., and Adema, G. J. The renal cell carcinoma-associated antigen G250 encodesa human leukocyte antigen (HLA)-A2.1-restricted epitope recognized by cytotoxic Tlymphocytes. Cancer Res., 59: 5554–5559, 1999.

23. Opavsky, R., Pastorekova, S., Zelnık, V., Gibadulinova, A., Stanbridge, E. J., Zavada,J., Kettmann, R., and Pastorek, J. Human MN/CA9 gene, a novel member of thecarbonic anhydrase family: structure and exon to protein domain relationships.Genomics, 33: 480–487, 1996.

24. Zavada, J., Zavadova, Z., Pastorek, J., Biesova, Z., Jezek, J., and Velek, J. Humantumour-associated cell adhesion protein MN/CA IX: identification of M75 epitopeand of the region mediating cell adhesion. Br. J. Cancer, 82: 1808–1813, 2000.

25. Hall, P. D., Willingham, M. C., Kreitman, R. J., and Frankel, A. E. DT388-GM-CSF,a novel fusion toxin consisting of a truncated diphtheria toxin fused to humangranulocyte-macrophage colony-stimulating factor, prolongs host survival in a SCIDmouse model of acute myeloid leukemia. Leukemia (Baltimore), 13: 629–633, 1999.

26. Tripathi, P. K., Qin, H., Bhattacharya-Chatterjee, M., Ceriani, R. L., Foon, K. A., andChatterjee, S. K. Construction and characterization of a chimeric fusion proteinconsisting of an anti-idiotype antibody mimicking a breast cancer-associated antigenand the cytokine GM-CSF. Hybridoma, 18: 193–202, 1999.

27. Battaglia, A., Fattorossi, A., Pierelli, L., Bonanno, G., Marone, M., Ranelletti, F. O.,Coscarella, A., De Santis, R., Bach, S., and Mancuso, S. The fusion protein MEN11303 (granulocyte-macrophage colony stimulating factor/erythropoietin) acts as apotent inducer of erythropoiesis. Exp. Hematol., 28: 490–498, 2000.

28. Batova, A., Kamps, A., Gillies, S. D., Reisfeld, R. A., and Yu, A. L. The Ch14.18-GM-CSF fusion protein is effective at mediating antibody-dependent cellular cyto-toxicity and complement-dependent cytotoxicity in vitro. Clin. Cancer Res., 5:4259–4263, 1999.

29. Hill, A. D., Naama, H. A., Calvano, S. E., and Daly, J. M. The effect of granulocyte-macrophage colony-stimulating factor on myeloid cells and its clinical applications.J. Leukocyte Biol., 58: 634–642, 1995.

30. Parhar, R. S., Ernst, P., Sheth, K. V., and al-Sedairy, S. T. Anti-tumor cytotoxicpotential and effect on human bone marrow GM-CFU of human LAK cells generatedin response to various cytokines. Eur. Cytokine Network, 3: 299–306, 1992.

31. Jonuleit, H., Knop, J., and Enk, A. H. Cytokines and their effects on maturation,differentiation and migration of dendritic cells. Arch. Dermatol. Res., 289: 1–8, 1996.

32. Bowne, W. B., Wolchok, J. D., Hawkins, W. G., Srinivasan, R., Gregor, P., Blachere,N. E., Moroi, Y., Engelhorn, M. E., Houghton, A. N., and Lewis, J. J. Injection ofDNA encoding granulocyte-macrophage colony-stimulating factor recruits dendriticcells for immune adjuvant effects. Cytokines Cell. Mol. Ther., 5: 217–225, 1999.

33. Cao, X., Zhang, W., Wang, J., Zhang, M., Huang, X., Hamada, H., and Chen, W.Therapy of established tumour with a hybrid cellular vaccine generated by usinggranulocyte-macrophage colony-stimulating factor genetically modified dendriticcells. Immunology, 97: 616–625, 1999.

34. Roth, M. D., Gitlitz, B. J., Kiertscher, S. M., Park, A. N., Mendenhall, M., Moldawer,N., and Figlin, R. A. Granulocyte macrophage colony-stimulating factor and inter-leukin 4 enhance the number and antigen-presenting activity of circulating CD14�and CD83� cells in cancer patients. Cancer Res., 60: 1934–1941, 2000.

35. Kitamura, T., Tange, T., Terasawa, T., Chiba, S., Kuwaki, T., Miyagawa, K., Piao,Y. F., Miyazono, K., Urabe, A., and Takaku, F. Establishment and characterization ofa unique human cell line that proliferates dependently on GM-CSF, IL-3, or eryth-ropoietin. J. Cell. Physiol., 140: 323–334, 1989.

36. Hinkel, A., Tso, C. L., Gitlitz, B. J., Neagos, N., Schmid, I., Paik, S. H., de Kernion,J., Figlin, R., and Belldegrun, A. Immunomodulatory dendritic cells generated fromnon-fractionated bulk peripheral blood mononuclear cell cultures induce growth ofcytotoxic T cells against renal cell carcinoma. J. Immunother., 23: 83–93, 2000.

37. Kriangkum, J., Xu, B., Gervais, C., Paquette, D., Jacobs, F. A., Martin, L., andSuresh, M. R. Development and characterization of a bispecific single-chain antibodydirected against T cells and ovarian carcinoma. Hybridoma, 19: 33–41, 2000.

38. Banchereau, J., Briere, F., Caux, C., Davoust, J., Lebecque, S., Liu, Y. J., Pulendran,B., and Palucka, K. Immunobiology of dendritic cells. Annu. Rev. Immunol., 18:767–811, 2000.

39. Rescigno, M., Martino, M., Sutherland, C. L., Gold, M. R., and Ricciardi-Castagnoli,P. Dendritic cell survival and maturation are regulated by different signaling path-ways. J. Exp. Med., 188: 2175–2180, 1998.

40. Roussel, E., Gerrard, J. M., and Greenberg, A. H. Long-term cultures of humanperipheral blood lymphocytes with recombinant human interleukin-2 generate apopulation of virtually pure CD3� CD16� CD56� large granular lymphocyte LAKcells. Clin. Exp. Immunol., 82: 416–421, 1990.

7932

GENERATION OF RENAL CELL CARCINOMA TUMOR VACCINE

Research. on August 17, 2014. © 2001 American Association for Cancercancerres.aacrjournals.org Downloaded from

41. Huang, A. Y., Golumbek, P., Ahmadzadeh, M., Jaffee, E., Pardoll, D., and Levitsky,H. Role of bone marrow-derived cells in presenting MHC class I-restricted tumorantigens. Science (Wash. DC), 264: 961–965, 1994.

42. Disis, M. L., Bernhard, H., Shiota, F. M., Hand, S. L., Gralow, J. R., Huseby, E. S.,Gillis, S., and Cheever, M. A. Granulocyte-macrophage colony-stimulating factor: aneffective adjuvant for protein and peptide-based vaccines. Blood, 88: 202–210, 1996.

43. Bendandi, M., Gocke, C. D., Kobrin, C. B., Benko, F. A., Sternas, L. A., Pennington,R., Watson, T. M., Reynolds, C. W., Gause, B. L., Duffey, P. L., Jaffe, E. S.,Creekmore, S. P., Longo, D. L., and Kwak, L. W. Complete molecular remissionsinduced by patient-specific vaccination plus granulocyte-monocyte colony-stimulat-ing factor against lymphoma. Nat. Med., 5: 1171–1177, 1999.

44. Kass, E., Panicali, D. L., Mazzara, G., Schlom, J., and Greiner, J. W. Granulocyte/macrophage-colony stimulating factor produced by recombinant avian poxvirusesenriches the regional lymph nodes with antigen-presenting cells and acts as animmunoadjuvant. Cancer Res., 61: 206–214, 2001.

45. Belldegrun, A., Muul, L. M., and Rosenberg, S. A. Interleukin 2 expanded tumor-infiltrating lymphocytes in human renal cell cancer: isolation, characterization, andantitumor activity. Cancer Res., 48: 206–214, 1988.

46. Yang, S., Vervaert, C. E., Burch, J., Jr., Grichnik, J., Seigler, H. F., and Darrow, T. L.Murine dendritic cells transfected with human GP100 elicit both antigen-specificCD8(�) and CD4(�) T-cell responses and are more effective than DNA vaccines atgenerating anti-tumor immunity. Int. J. Cancer, 83: 532–540, 1999.

47. Van den Eynde, B. J., and Morel, S. Differential processing of class-I-restrictedepitopes by the standard proteasome and the immunoproteasome. Curr. Opin. Immu-nol., 13: 147–153, 2001.

48. Cascio, P., Hilton, C., Kisselev, A. F., Rock, K. L., and Goldberg, A. L. 26Sproteasomes and immunoproteasomes produce mainly N-extended versions of anantigenic peptide. EMBO J., 20: 2357–2366, 2001.

49. Rammensee, H. G., Falk, K., and Rotzschke, O. Peptides naturally presented by MHCclass I molecules. Annu. Rev. Immunol., 11: 213–244, 1993.

50. Lee, K. H., Wang, E., Nielsen, M. B., Wunderlich, J., Migueles, S., Connors, M.,Steinberg, S. M., Rosenberg, S. A., and Marincola, F. M. Increased vaccine-specificT cell frequency after peptide-based vaccination correlates with increased suscepti-bility to in vitro stimulation but does not lead to tumor regression. J. Immunol., 163:6292–6300, 1999.

7933

GENERATION OF RENAL CELL CARCINOMA TUMOR VACCINE

Research. on August 17, 2014. © 2001 American Association for Cancercancerres.aacrjournals.org Downloaded from