Gene Transfer of the Co-Stimulatory Molecules B7-1 and B7-2 Enhances the Immunogenicity of Human Renal Cell Carcinoma to a Different Extent

Gene Transfer of the Co-Stimulatory Molecules B7-1 and B7-2Enhances the Immunogenicity of Human Renal Cell Carcinomato a Different Extent

D. JUNG*, C. HILMES*, A. KNUTH†, E. JAEGER†, C. HUBER* & B. SELIGER*

*Johannes Gutenberg-Universita¨t, III. Medizinische Klinik, Langenbeckstr. 1, 55101 Mainz and†Nord West Krankenhaus,II. Medizinische Klinik, 60488 Frankfurt

(Received 22 December 1998; Accepted in revised form 29 April 1999)

Jung D, Hilmes C, Knuth A, Jaeger E, Huber C, Seliger B. Gene Transfer of the Co-Stimulatory MoleculesB7-1 and B7-2 Enhances the Immunogenicity of Human Renal Cell Carcinoma to a Different Extent. ScandJ Immunol 1999;50:242–249

Stimulation of a specific antitumour immune response with recruitment and induction of T-cell effectorfunctions represents an attractive concept in human cancer therapy. Different cytokines and the B7 co-stimulatory molecules are both able to provide proliferation and activation signals for T cells. In the presentstudy, we first demonstrated the absence of both B7-1 and B7-2 expression in human renal cell carcinoma(RCC) cell lines. The lack of B7 expression was associated with a low or absent proliferative response ofallogeneic and autologous T cells upon stimulation with tumour cells. In order to investigate the role of B7-1and B7-2, the human RCC cell line, MZ1257RC, which expresses normal levels of adhesion molecules andmajor histocompatibility complex (MHC) class I surface antigens, was transfected with B7-1 and B7-2expression vectors, respectively. The B7-1- and B7-2-transduced MZ1257RC cells were potent stimulators ofallogeneic and autologous T-cell proliferation. B7-2 transfectants were approximately two- to threefold moreeffective in the induction of primary T-cell activation than B7-1-transduced cells. Interleukin (IL)-12synergized with the B7/CD28 interaction to enhance allogeneic T-cell proliferation, independently of the B7molecule transduced. In contrast, IL-2 only co-operatively increased T-cell activation in the presence of B7-2.Our results suggest the following: first, that co-stimulatory molecules are required for efficient T-cellresponses directed against RCC; second, that B7-2 appears to be a more potent stimulator of tumour immunityas compared to B7-1; and third, that B7 molecules selectively co-operate with different T-cell stimulatorycytokines. The different activity of B7-1 and B7-2 molecules on the immunogenicity of RCC will haveimplications for the development and optimization of RCC-specific cancer vaccines.

Barbara Seliger, PhD, Johannes Gutenberg Universita¨t, III. Medizinische Klinik, Langenbeckstr. 1, 55101Mainz, Germany

INTRODUCTION

A prominent goal of immunotherapy is the induction of a specificantitumour T-cell response, which results in the eradication ofthe tumour. The physiological stimulation of naive T cells iscontrolled by several parameters [1–3]. Two key factors areinvolved in T-cell activation. The first is provided by theinteraction of the T-cell receptor (TCR) with antigenic peptidepresented by the major histocompatibility complex (MHC)expressed on antigen-presenting cells (APC). However, cross-linking of the TCR by antigen alone is not sufficient to induce a

productive immune response, thereby leading to long-term anti-gen-specific unresponsiveness [4]. Effective activation and differ-entiation of T cells requires at least a second antigen-independentco-stimulatory signal provided by different accessory moleculeson the surface of APC. Although many ligand–receptor pairshave been shown to function as co-stimulators, the interaction ofCD28/CTLA-4 on the T cell with B7 on APC represents the best-characterized pathway for co-stimulation [1–3,5]. B7-1 and B7-2, members of the immunoglobulin (Ig) supergene family, areencoded by separate genes and provide co-stimulatory function,both in vitro and in vivo, for APC-dependent T-cell activation

Scand. J. Immunol.50, 242–249, 1999

q 1999 Blackwell Science Ltd

[6–9]. At the present time it is controversial as to whether theco-stimulatory signals provided by B7-1 and B7-2 are func-tionally distinct or overlapping [10–12]. Several lines ofevidence indicate that CD28 is able to deliver biochemicalsignals that function in synergy with TCR-mediated signallingto initiate and maintain T-cell responses [13,14]. The B7-mediated co-stimulation is sufficient to prevent the induction ofalloantigen-specific anergy and involves the activation ofTCR-associated, specific protein tyrosine kinase-mediatedpathways [15–17]. In the absence of co-stimulation, cytokinescan also prevent the induction of T-cell anergy [18]. Inter-leukin (IL)-2 is capable of inducing the generation of cytotoxicT lymphocytes (CTL) and stimulating T-cell proliferation.The heterodimeric cytokine IL-12, comprising one 40 000MW (p40) and one 35 000 MW (p35) subunit, exerts multipleeffects on T cells and natural killer (NK) cells. Its mainactivities are the induction of interferon-g (IFN-g) productionby NK cells and T cells, promotion of the development of Thelper 1 (Th1) responses and stimulation of T-cell prolifera-tion. However, the effect of IL-12 is modest and dependenton the stage of T-cell activation [19].

Several murine and human tumours have been shown to evadeimmune surveillance by suppression of the MHC class I antigen-processing machinery [20,21]. Reintroduction of MHC class Igenes, and of other molecules critical for antigen processing,such as the peptide transporter, TAP, and the proteasome subunits,or of cytokines known as potent up-regulators of the differentcomponents of the MHC class I antigen-processing pathway, hasthe capacity to induce T-cell-dependent antitumour responsesand rejection of tumours [20,22,23]. However, tumour cells alsofail to stimulate allogeneic T-cell proliferation, despite appar-ently normal levels of MHC class I surface expression, adeficiency that appears to be caused by lack of expression ofco-stimulatory molecules [24].

An increased knowledge of the B7/CD28 molecules, and ofthe mechanisms involved in co-stimulation of T cells, pave theway for manipulation of T-cell activation [2,3,14]. Recentstudies have demonstrated that B7 expression of tumourcells results in increased tumour immunogenicity, enhancedgeneration of allogeneic and autologous tumour-reactive CTL,protection against subsequent challenge with wild-type B7-negative tumour cells and active treatment of established disease[25–30].

There are several variables influencing the efficacy of B7-mediated immunotherapy, e.g. the origin of the tumour cell, itsinherent immunogenicity and the amount of transduced B7molecules expressed by the tumour cell. In highly immunogenictumours, such as melanoma, transduction of B7-1 alone was ableto induce tumour immunity in mice [29]. The impressive resultsobtained in these animal studies led to approval for the firstclinical trials of B7-1 immunotherapy in patients with melanoma[31,32]. The existence of large amounts of lymphocytic infil-trates in tumoursin situ,the high spontaneous regression rate andthe relatively high response rate of immunotherapy with cyto-kines suggest a specific T-cell response in renal cell carcinoma

(RCC) [33,34]. However, in contrast to melanoma, only a fewRCC-specific CTL clones have been established [35–39]. Thepurpose of this study was to elucidate whether the immunogeni-city of RCC could be enhanced by B7-1 or B7-2 gene transferalone or in combination with cytokine treatment. B7-1 and B7-2expression vectors were developed and effectively transducedinto different RCC lines by electroporation. The B7-1 and B7-2transfectants were investigated for their ability to induce pro-liferative responses of allogeneic and autologous T cells in thepresence and absence of IL-2 or IL-12.

MATERIALS AND METHODS

Cells and cell lines. Peripheral blood mononuclear cells (PBMC) wereisolated from fresh blood or buffy coat from healthy, nonmatched humanleucocyte antigen (HLA)-A2þ donors, and from the respective RCCpatients, by Ficoll–Hypaque density-gradient separation. Monocyteswere depleted by adherence to plastic tissue-culture flasks prior to theenrichment of T cells by E-rosetting with neuraminidase-treated redblood cells. T cells were cultured overnight in RPMI-1640 mediumsupplemented with 10% fetal calf serum (FCS),L-glutamine (2 mM),penicillin (10 IU/ml) and streptomycin (100mg/ml) and were eitheremployed immediately or cryopreserved for future use.

The RCC cell lines, designated as ‘MZ’-RC, were established in thelaboratories of A. Knuth or H. Gabbert (Institute of Pathology, Du¨sseldorf,Germany) from fresh tumour specimens of patients with spontaneousRCC. The Epstein–Barr virus (EBV)-transformed B lymphocytes of aRCC patient (1257-EBV) were established by Knuth and co-workers.The different RCC cell lines (MZ1257RC, MZ1851RC, MZ1940RC andMZ1973RC) were cultured in Dulbecco’s modified Eagle’s minimalessential medium (DMEM, Gibco, Karlsruhe, Germany) containing10 mM HEPES buffer,L-arginine (84 mg/l), L-glutamine (584 mg/l),penicillin (10 IU/ml), streptomycin (100mg/ml) and 10% FCS. The B-lymphoblastoid cell line, T2 [40], used as a positive control for B7expression, and the 1257-EBV transformed B lymphocytes were main-tained in RPMI-1640 medium supplemented with 10 mM HEPES buffer,L-arginine (242 mg/l),L-asparagine (50 mg/ml),L-glutamine (300 mg/l),penicillin (10 IU/ml), streptomycin (100mg/ml), 1% nonessential aminoacids and 10% FCS.

B7 gene transfer. RCC cells were stably transfected using a square pulseelectroporator (Fischer, Heidelberg, Germany) with the plasmid vectorspCMVIrNeo (P46; an empty vector containing the internal ribosomalentry site–neomycin resistance [IRES–neor] cassette), pCMVB7-1Ir-Neo (P70; human B7-1 cDNA under the control of the cytomegalovirus(CMV) promoter), and pBCMVB7-2Ir-Neo (P56; human B7-2 cDNAunder the control of the CMV promoter; Fig. 1). Both, the human B7-1and B7-2 cDNA were cloned after reverse transcription–polymerasechain reaction (RT–PCR) from mRNA of T2 cells (D. Jung & B. Seliger,personal communication). Briefly, 2.5× 106 cells in 500ml of DMEMþ 1% FCS were mixed gently with 10mg linearized vector. Thesuspension was pulsed once using 450 V, 2 msec and 960 F. After theelectroporation procedure, RCC cells were immediately plated out instandard medium containing 10% FCS. Twenty four h later, transfectantswere subjected to selection medium containing 500mg/ml G-418(Gibco) until neor colonies formed.neor bulk cultures, representinga mixture of more than 100 B7-expressing single-cell clones,were cultured and maintained in selective long-term culture mediumcontaining 300mg/ml active G418.

B7-Transduced RCC Lines Enhance T-Cell Responses243

q 1999 Blackwell Science Ltd,Scandinavian Journal of Immunology, 50, 242–249

PCR analyses. The primers used for PCR were purchased fromPharmacia Biotech (Freiburg, Germany) or MWG (Mu¨nchen, Germany).The amplifications were performed in a thermocycler (Biometra, Go¨ttin-gen, Germany) usingTaq DNA polymerase (Boehringer Mannheim,Mannheim, Germany) at a concentration of 40 U/ml in a volume of

50ml. For genomic PCR analysis, extracts from 1 to 3×103 cells wereprepared by lysis (10 min at 958C in 20ml of water) and proteinase Ktreatment (400mg/ml, for 60 min at 558C followed by heat inactivationfor 10 min at 958C) and subjected to hot-start PCR using primers specificfor the vector P46 (sense primer: 50-ACGACTCACTATAGGGAGACC-30), and for B7-1 (antisense primer: 50-GCTCTTGCTTGGTTGTATTC-30) and B7-2 (antisense primer: 50-GAAACAGACAAGCTGATGGA-30), respectively.

Flow cytometry. The following antibodies were used in flow cyto-metric analyses: the hybridoma supernatants of anti-HLA class I (W6/32), recognizing assembled HLA-A, -B, and -C locus products com-plexed tob2-microglobulin (b2-m) [41], and of anti-HLA-DR (HB55);the monoclonal antibodies (MoAbs) MsIgG (negative control), B7-1(CD80; clone MAB104), lymphocyte function-associated antigen-3(LFA-3) (CD58; clone AICD58), CD40 (clone MoAb89), intracellularadhesion molecule (ICAM)-3 (CD50; clone HP2/19), phycoerythrin(PE)-conjugated CD4, and fluorescein isothiocyanate (FITC)-conjugatedMoAbs CD8 and ICAM-1 (CD54; clone 84H10) purchased fromCoulter-Immunotech Diagnostics (Hamburg, Germany), ICAM-2(CD102; clone B-T1) from Natutech (Frankfurt, Germany) and B7-2(CD86; clone 2331) from Pharmingen (San Diego, CA, USA).

For flow cytometric analysis, 5×105 parental or transfected tumourcells or isolated T cells were incubated with saturating amounts of therespective MoAbs for 30 min at 48C, then washed with ice-cold phos-phate-buffered saline (PBS), and when no directly conjugated antibodieswere employed were stained with an FITC-labelled goat antimouse Igantibody (FITC-GAM Ig; Coulter-Immunotech Diagnostics) for an addi-tional 30 min at 48C. After washing twice with PBS, stained cells wereanalysed on an EPICS XL-MCL (Coulter Electronics, Krefeld, Germany).Immunofluoresence analyses were performed at least three times. In thefluorescence-activated cell sorter (FACScan) histograms, the verticalaxis represents the relative cell number, whereas the horizontal axisrepresents log fluorescence intensity.

Quantitative determination of cell-surface antigens was performedwith the DAKO QIFIKIT (Dako, Hamburg, Germany) according to themanufacturer’s manual. It is based on the parallel incubation of a seriesof calibration beads coated with well-defined quantities of mouse anti-human CD5 MoAb from clone ST1. The calibration beads were analysedby flow cytometry and the results obtained were used for the constructionof a calibration curve expressing mean fluorescence intensity (MFI)against antibody-binding capacity. The different cell specimenswere analysed in parallel on the flow cytometer, as described above,and the antigen density was calculated by interpolation on the calibrationcurve.

T-cell proliferation assay. Gamma-irradiated (20 000 rad) tumour cells(1× 104 cells/well) were co-cultured with T cells (T cell to tumour cellratio: 10 : 1) in complete RPMI (200ml/well), at 378C for 5 days, in96-well flat-bottom microtitre plates, in the presence or absence ofanti-CD28 MoAb (0.5mg/ml, clone Kolt4; HISS Diagnostics, Freiburg,Germany), recombinant human IL-2 (Pharmingen) or human recombi-nant human IL-12 (Pharmingen). On day 5, cultures were pulsed with[3H]-thymidine (1mCi/well; Du Pont, Bad Homburg, Germany) for 18 h,harvested onto glass-fibre filters (Filtermate 196 Micro Cell Harvester,Packard, Dreieich, Germany) and assayed for [3H]-thymidine incorpora-tion using a liquid scintillation counter (Betaplate 1205, Berthold, BadWildbad, Germany). All cultures were set up in triplicate.

Measurement of soluble cytokines. Supernatants from cells werecollected and the amount of IL-10 (Per Septive Diagnostics, Framingham,MA, USA) and transforming growth factor-b (TGF-b; Genzyme, Cam-bridge, MA, USA) released into the culture medium was determined by

244 D. Jung et al.


Fig. 1. Vectors employed for B7 gene transfer. (A) P46, a controlvector carrying theneor gene. (B) P70, containing the human (hu)B7-1 cDNA and the internal ribosomal entry site (IRES)–neor

cassette. C. P56, containing the hu B7-2 cDNA and the IRES–neor

cassette.

commercially available enzyme-linked immunosorbent assays (ELISAs),according to the manufacturers’s instructions.

RESULTS

Lack of B7 expression in RCC lines is associated withimpaired T-cell stimulation

The expression of a number of immune modulatory molecules,such as B7 co-stimulatory molecules, MHC antigens and adhe-sion molecules, was examined in four RCC lines (MZ1257RC,MZ1851RC, MZ1940RC and MZ1973RC) using flow cytometry.The four RCC cell lines analysed expressed marginal levels ofB7-1, B7-2, MHC class II, ICAM-2 or ICAM-3 on their cellsurface (Table 1). In contrast, ICAM-1 was highly expressed onall RCC lines tested, whereas the expression of the MHC class Iantigens, CD40 and LFA-3 varied between the different RCClines. The RCC lines MZ1257RC, MZ1851RC and MZ1973RCwere investigated for their capacity to activate autologous and/orallogeneic T lymphocytesin vitro. The proliferative response ofisolated T cells from three healthy HLA-matched volunteers andfrom the autologous RCC patients stimulated with irradiated,unmodified RCC cell lines was determined after 6 days of cultureby measuring [3H]-thymidine incorporation. Representative resultsare shown in Fig. 2, demonstrating that MZ1257RC, MZ1851RCand MZ1973RC cells were poor stimulators of allogeneic and/orautologous mixed lymphocyte reaction when compared withHLA-A2þ B7-expressing lymphoblastoid cells. The inability ofthe different RCC lines to effectively induce T-cell proliferationwas independent of the donor type employed (data not shown),

and was not caused by secretion of immunosuppressive factorsby the tumour cells because only low levels of IL-10 and TGF-b

were detected in the supernatant of the different RCC lines(Table 1).

Stable transfection of RCC lines with B7-expression vectors

In order to test whether the lack of expression of co-stimulatorysignals was responsible for the failure of RCC lines to stimulateallogeneic and autologous T cells, MZ1257RC cells were chosenfor gene-transfer experiments. This cell line expresses relativelyhigh levels of MHC class I surface antigens and adhesionmolecules compared with MZ1851RC, MZ1940RC andMZ1973RC cells, as well as to its autologous EBV-transformedB lymphocytes (Table 1). Using an optimized electroporationprotocol, MZ1257RC cells were stably transfected with plasmidvectors containing the selectable markerneor and either the humanB7-1 (P70) or B7-2 gene (P56) under the control of the CMVpromoter (Fig. 1). A vector carrying theneor gene (P46) alonewas used as a control. Cells were propagated in G-418-containingselection medium for 4 weeks beforeneor batch cultures wereanalysed for the integration, as well as the expression, of therespective transgene(s). Using vector-specific primers, the inte-gration of B7-1 or B7-2 was demonstrated in the respectiveneor

batch cultures (data not shown). The cell-surface expression andantigen density of the human B7-1 and B7-2 genes on therespective MZ1257RC transfectants was determined by flowcytometry using the B7-specific MoAbs, CD80 and CD86.Again, P46-transfected MZ1257RC cells were used as a control.As shown in Fig. 3, the B7-transduced MZ1257RC cells



Table 1. Expression analysis of proteins involved in T-cell activation in different renal cell carcinoma(RCC) cell lines using flow cytometry or enzyme-linked immunosorbent assay (ELISA)

MZ1257RC MZ1851RC MZ1940RC MZ1973RC 1257-EBV

MHC class I 125.1 24.6 78.8 198.0 565.4MHC class II 1.2 1.1 1.0 1.0 8.5B7-1 1.3 1.4 1.1 1.0 26.4B7-2 1.6 1.0 1.5 1.6 68.4ICAM-1 46.5 15.8 38.5 37.6 19.9ICAM-2 1.3 1.1 1.1 0.9 9.3ICAM-3 1.9 1.0 1.3 1.2 118.3LFA-3 10.5 4.2 3.8 4.4 NDCD40 6.6 ND ND ND 103.8TGF-b 0 0 0 0 NDIL-10 0 0 0 0 ND

The data are expressed as mean specific fluorescence intensity (MFI) values in relation to theimmunoglobulin G (IgG) negative control (MFI¼ 1).

ELISA data are expressed as absolute values.Two melanoma cell lines secreting interleukin (IL)-10 as well as transforming growth factor-b (TGF-b)

were used as positive controls.ICAM, intracellular adhesion molecule; LFA-3, lymphocyte function-associated antigen-3; MHC, major

histocompatibility complex; ND, not determined.

expressed similar levels of B7-1 or B7-2 molecules on the cellsurface, whereas MZ1257RC cells transfected with the ‘empty’neor vector and parental tumour cells showed no measurable B7surface expression. Quantitative determination of the antigendensity demonstrated that B7-1-transfected MZ1257RC batch

cultures expressed 30397 B7-1 binding sites/cell, B7-2-trans-fected MZ1257 RC batch cultures 36026 B7-2 binding sites/cell,whereas the MHC class I surface expression of the B7-1-transfected MZ1257RC batch culture was lower than that ofB7-2 transfectants.

B7-1- and B7-2-transfected RCC cells are potent stimulatorsof the allogeneic T-cell response

B7-1- and B7-2-transfected MZ1257RC batch cultures, repre-senting a bulk of more than 100 B7-1- or B7-2-expressingindividual colonies, were used to stimulate autologous as wellas HLA-A2-matched allogeneic T cells of healthy donors (n¼ 3)for 5 days before [3H]-thymidine uptake was determined. Incontrast to parental and P46-transfected cells, both B7-1- and B7-2-transfected MZ1257RC batch cultures were found to induceallogeneic and autologous T-cell proliferation, but the B7 mole-cules affected T-cell activation to a different extent (Fig. 4): thestimulatory capacity of B7-2 transfectants was, in all experi-ments, 1.5- to twofold higher when compared with B7-1-trans-duced MZ1257RC cultures. Similar results were obtained withB7-1- and B7-2-transfected MZ1257RC clones (data not shown).This heterogeneous effect could neither be explained by thelevels of B7-1 or B7-2 surface expression, as a result of similarantigen density of the molecules on the transfectants, nor byphenotypic alterations following transfection, because there wasno change in the expression of CD40, LFA-3 and ICAM-1 (datanot shown).

To confirm that the proliferation was induced by the B7/CD28co-stimulatory pathway, all reactions were also performed inparallel in the presence of anti-CD28 MoAb. The proliferativeresponse of both isolated autologous and allogeneic T cellsstimulated by B7 transfectants was totally blocked by anti-CD28 MoAb, as shown representatively for allogeneic T cells(Fig. 4), but not by anti-ICAM-1, used as a control antibody (datanot shown). The background proliferative response of parentaland P46-transfected MZ1257RC was not blocked by anti-CD28MoAb (Fig. 4). These data suggest that the growth promotion ofT cells induced by B7 transfectants was caused by a specificCD28-dependent co-stimulation.

Differential effects of cytokines on B7-mediated T-cellstimulation by RCC

B7 and cytokines, such as IL-2, IL-7 and IL-12, have been shownto synergistically induce activation and proliferation of T cellsinvitro and tumour immunityin vivo [42,43]. To further investigatewhether the proliferative response of allogeneic T cells could beenhanced by cytokines, we compared the stimulatory capacity ofB7-1 and B7-2 MZ1257RC bulk cultures for allogeneic purifiedT cells in the presence of recombinant IL-2 and IL-12, respec-tively (Fig. 5). IL-12 alone did not induce T-cell proliferation,but enhanced both the B7-1- and B7-2-mediated activation ofallogeneic T cells. As expected, IL-2 was capable of stimulating

246 D. Jung et al.


40000

35000

30000

25000

20000

15000

10000

5000

0MZ1257RC MZ1851RC MZ1973RC EBV

3 H-t

hym

idin

e

Fig. 2. Proliferative response of T cells against human renal cellcarcinoma (RCC) cell lines. The proliferative capacity of allogeneicand/or autologous T cells against three RCC cell lines with distincthuman leucocyte antigen (HLA) phenotypes was determined usingmixed lymphocyte tumour culture (MLTC). T-cell proliferation assaysusing 1257-Epstein–Barr virus (EBV) transformed B cells were usedas controls. The experiments were performed as described in theMaterials and methods. Results are expressed as mean [3H]-thymidineincorporation of two independent experiments consisting of triplicatecultures. Black bars represent results of MLTC using allogeneic Tcells; white bars represent data employing autologous T cells.

140

120

100

80

60

40

20

0

Mea

n s

pec

ific

flu

ore

scen

ce

MZ1257RC 1257/neo 1257/B7-1 1257/B7-2

Fig. 3. B7 and major histocompatibility complex (MHC) class Isurface expression of transfectants. The parental MZ1257RC cells, andmock-transfected (1257/neo) and B7-transfected batch cultures weresubjected to flow cytometry using the respective monoclonalantibodies (MoAb), as described. The results represent mean specificfluoresence intensity of at least three independent experiments. Blackbars, MHC I; white bars, B7-1; grey bars, B7-2.

T-cell proliferation in mixed lymphocyte tumour culture (MLTC)reactions using parental and P46-transfected MZ1257RC cells,which could be further enhanced by B7-2, but not by B7-1,expression. No difference in the phenotype of T cells wasobserved after IL-2 and IL-12 treatment of MLTC cultures usingB7-1 and B7-2 as stimulators. Independently of the stimulatorsand cytokine, the percentage of CD4þ T cells ranged, in allMLTC reactions analysed, from 33 to 37%, and that of CD8þ

T cells from 53 to 58%. Thus, alterations in the pattern ofcytokine secretion is not expected. These data suggest:

(1) that B7-1 and B7-2 may trigger the immune system indifferent ways; and

(2) that the cytokines IL-12 and IL-2 co-operatively enhancethe function of the B7 co-stimulatory molecules by induction ofimmunogenicity of RCC, employing complementary T-cell-directed pathways.

DISCUSSION

A role for the cellular immune system in the control of melanomaand RCC is underscored by the relatively high rate of sponta-neous regression of these malignancies, their response to immuno-therapy and the existence of large T-cell infiltrations in thetumours. There exist different mechanisms of how tumour cellscould evade immune surveillance, such as production of immuno-suppressive cytokines, down-regulation of components of theMHC class I antigen-presentation machinery and impairedexpression of co-stimulatory signals. Several studies have exploredthe use of B7-transfected tumour cells as a strategy for enhance-ment of T-cell responsesin vitro and for generation of systemicantitumour immunity in the mouse, whereas only minimal infor-mation is available employing B7-modified tumour cells inclinical trials [2,25–30,44]. Although it has been demonstratedthat B7-1 and B7-2 transfection can provide B7-negative tumourcells with co-stimulatory function, the activity of both B7 mole-cules in human malignant cells has not been examined in parallel.We demonstrate here that the lack of B7/CD28 co-stimulationseems to be one mechanism whereby human RCC lines expressnormal levels of MHC class I antigens to escape CTL-mediatedrecognition. Unmodified RCC lines of different HLA phenotypeswere not able to effectively stimulate autologous and allogeneicT lymphocytes, as demonstrated by MLTC reactions (Fig. 2).The observed minimal proliferation of T cells could be causedby other receptor/counter-receptor interactions known to affectT-cell activation [1,5], as the RCC cells employed expressed theaccessory molecules ICAM-1, LFA-3 and CD40 (Table 1).Indeed, such molecules may work co-operatively to augmentT-cell expansion [1,5,45].

Transfection of both B7 molecules into MZ1257RC cellsstrongly induced autologous and allogeneic T-cell responses,but B7-2 was more effective in T-cell stimulation than B7-1.These data were confirmed by analysis of other RCC cell lineswith B7-1 and B7-2 molecules (data not shown). Thus, both B7molecules can function as co-stimulatory ligands in RCC toenhance their immunogenicity. The results concerning the B7-1-modified RCC cells were consistent with results of other groups[26,30,31], also demonstrating that a proliferative activation ofallogeneic and autologous T cells could be induced by B7-1-modified tumour cells. Our study extends these findings to B7-2gene-modified RCC cells. The different efficacy of T-cellactivation by co-cultivation with irradiated B7-1- or B7-2-expressing RCC cells could not be explained by quantitativedifferences in the amounts of B7-1 or B7-2 surface expression



12000

10000

8000

6000

4000

2000

0MZ1257RC 1257/neo 1257/B7-1 1257/B7-2

3 H-t

hym

idin

e

Fig. 4. Induction of allogeneic and autologous T-cell responses usingB7-transfected MZ1257RC cells as stimulators. Stimulation ofallogeneic (black bars) and autologous (white bars) T-cell proliferationby B7-transfected MZ1257RC cells in the absence (black/white bars)and presence (grey bars) of anti-CD28 monoclonal antibody (MoAb).The results concerning anti-CD28 MoAb are only shown for theproliferation of allogeneic T cells as they are similar to that ofautologous T cells. The results represent the mean value of at leasttwo independent experiments using triplicate cultures.

35000

30000

25000

20000

15000

10000

5000

0T-cells alone IL-2 IL-12

3 H-t

hym

idin

e

Fig. 5. Effect of interleukin (IL)-2 and IL-12 on the stimulation ofallogeneic T-cell proliferation by parental and B7-expressingMZ1257RC cells. Proliferative response of allogeneic T lymphocytestowards B7-transduced and mock-transfected (1257/neo) tumour cellswere performed, as described in the Materials and methods, in thepresence or absence of 10 U/ml IL-2 or 20 U/ml IL-12. Onerepresentative of three independent experiments is shown. Black bars,1257/neo; white bars, 1257/B7-1; grey bars, 1257/B7-2.

in these transfectants (Fig. 4), but seemed to be based ondistinct functions of the B7 co-stimulatory molecules. Althoughseveral lines of evidence suggest that the signals provided byB7-1 and B7-2 may be at least, in part, redundant [7,9,46], ourfindings emphasize intrinsic differences in the binding andsignalling capacity of these two molecules, thereby confirmingthe involvement of B7-2 in the primary immune response: B7-2appears to be the dominant and more effective co-stimulatoryligand during primary immune responses, whereas B7-1 seemedto be more critical in prolonging primary or co-stimulatorysecondary T-cell responses [13]. Our data suggest that transfec-tion of human RCC cells with B7-2, rather than with B7-1,would be an effective method forin vitro generation of CTL.The distinct and sometimes conflicting roles of the B7-1- andB7-2-mediated co-stimulation have been described during anti-gen presentation by APC for both humoral and cellularresponses [47–49]. In addition, some studies have demonstratedan independent regulation of B7-1 and B7-2 expression byvarious stimuli [10]. However, further analysis of the B7-1-and B7-2-mediated signals and their co-stimulatory functionwill be critical to an overall understanding of the roles of bothmolecules in T-cell activation and especially in the developmentof antitumour immunity.

In a number of murine tumour models, B7-mediated T-cellco-stimulation induced tumour rejection, but the vaccine potencyof gene-modified tumour cells has still to be improved. This wasthe motivation for various studies combining cytokines withB7. Indeed, different cytokines, such as IL-7 and IL-12, havebeen shown to co-operate with B7-1 by inducing an antitumourresponse [42,43]. We extend these results by demonstratingdifferential effects of IL-2 and IL-12 on the B7-1/B7-2-mediatedco-stimulation. The conflicting result regarding the lack of addi-tional activity of IL-2 on the B7-1-transfected MZ1257RC cellsmay be explained by their lower levels of MHC class I surfaceexpression when compared with B7-2 transfectants (Fig. 4). Thesynergistic effect of IL-12 on B7-1- and B7-2-mediated co-stimulation could be mediated either directly, by enhancing thedevelopment of immune responses, or indirectly by the inductionof CD40-ligand (CD40L) on T cells [50], which could interactwith CD40 on MZ1257RC cells. Thus, B7 transfer, alone or incombination with cytokines, into RCC cells might be a usefultool for in vitro CTL generation. We are currently addressing thispossibility. As new treatment modalities of RCC patients areurgently needed, these data might contribute to the establishmentof a therapeutic vaccine trial of human RCC patients using B7-modified RCC cells.

ACKNOWLEDGMENTS

We thank Dr Gabbert, Institute of Pathology, Du¨sseldorf forproviding the MZ1257RC cells, Mrs U. Petrat-Benzaim forexcellent technical assistance and I. Schmidt for help in prepar-ing the manuscript. This work was supported by grants from theMildred Scheel Stiftung W111/94 (to B.S.) and the BoehringerIngelheim Funds (to B.S. and C.H.).

REFERENCES

1 Dubey C, Croft M. Accessory molecule regulation of naive CD4T cell activation. Immunol Res 1996;15:114–25.

2 Linsley PS, Ledbetter JA. The role of the CD28 receptor during T cellresponses to antigen. Annu Rev Immunol 1993;11:191–212.

3 Lenschow D, Walunas T, Bluestone J. CD28/B 7 system of T cellcostimulation. Annu Rev Immunol 1996;14:233–58.

4 Mueller DL, Jenkins MK, Schwartz RH. Clonal expansion versusfunctional clonal inactivation: a costimulatory signalling pathwaydetermines the outcome of T cell antigen receptor occupancy. AnnuRev Immunol 1989;7:445–80.

5 Jenkins MK, Johnson JG. Molecules involved in T-cell costimula-tion. Curr Opin Immunol 1993;5:361–7.

6 Azuma M, Ito D, Yagita Het al. B70 antigen is a second ligand forCTLA-4 and CD28. Nature 1993;366:76–9.

7 Freeman GJ, Gribben JG, Boussiotis VAet al. Cloning of B7–2: aCTLA-4 counter-receptor that costimulates human T cell prolifera-tion. Science 1993;262:909–11.

8 Chen L, Ashe S, Brady WAet al. Co-stimulation of anti-tumorimmunity by the B7 counterreceptor for the T-lymphocyte moleculesCD28 and CTLA-4. Cell 1992;71:1093–102.

9 Linsley PS, Brady W, Grosmaire Let al. Binding of the B cellactivation antigen B7 to CD28 costimulates T cell proliferation andinterleukin-2 mRNA accumulation. J Exp Med 1991;173:721–30.

10 Hathcock KS, Laszlo G, Pucillo Cet al. Comparative analysis of B7-1 and B7-2 costimulatory ligands: expression and function. J ExpMed 1994;180:631–40.

11 Kuchroo VK, Das MP, Brown JAet al. B7-1 and B7-2 costimulatorymolecules activate differentially the Th1/ Th2 developmental path-ways: application to autoimmune disease therapy. Cell 1995;80:707–18.

12 Lanier LL, O’Fallon S, Somoza Cet al. CD80 (B7) and CD86 (B7.2)provide similar co-stimulatory signals for T-cell proliferation, cyto-kine production and generation of CTL. J Immunol 1995;1:97–105.

13 Tan P, Anasetti C, Hansen JAet al. Induction of alloantigen-specifichyporesponsiveness in human T lymphocytes by blocking interactionof CD28 with its natural ligand B7/BB1. J Exp Med 1993;177:165–73.

14 Bluestone JA. New perspectives of CD28–B7-mediated T cell co-stimulation. Immunity 1995;2:555–9.

15 Harding FA, McArthur JG, Gross JAet al. CD28-mediated signal-ling co-stimulates murine T cells and prevents induction of anergy inT-cell clones. Nature (London) 1992;356:607–9.

16 Gimmi CD, Freeman GJ, Gribben JGet al. Human T-cell clonalanergy is induced by antigen presentation in the absence of B7costimulation. Proc Natl Acad Sci USA 1993;90:6586–90.

17 Boussiotis VA, Barber DL, Lee BJet al. Differential association ofprotein tyrosine kinases with the T cell receptor is linked to theinduction of anergy and its prevention by B7 family-mediated co-stimulation. J Exp Med 1996;184:365–76.

18 Essery G, Feldmann M, Lamp J. Interleukin-2 can prevent andreverse antigen-induced unresponsiveness in cloned human Tlymphocytes. Immunology 1988;64:413–7.

19 Gajewski TF, Renauld JC, van Pel Aet al. Costimulation with B7-1,IL-2 and IL-12 is sufficient for primary generation of murine anti-tumor cytolytic T lymphocytesin vitro. J Immunol 1995;154:5637–48.

20 Seliger B, Maeurer M, Ferrrone S. TAP off – tumors on. ImmunolToday 1997;18:292–9.

248 D. Jung et al.


21 Restifo NP, Esquivel F, Kawakami Yet al. Identification ofhuman cancers deficient in antigen processing. J Exp Med 1993;177:265–72.

22 Attaya M, Jameson S, Martinez CKet al. Ham-2 corrects the class Iantigen-processing defect in RMA-S cells. Nature 1992;355:647–9.

23 Restifo NP, Spiess PJ, Karp SEet al. A non-immunogenic sarcomatransduced with the cDNA for interferon-gamma elicits CD8þ

T cells against the wild-type tumor: correlation with antigen presen-tation capacity. J Exp Med 1992;175:1423–31.

24 Talmage DW, Woolnough JA, Hemmingsen Het al. Activation ofcytotoxic T cells by non-stimulating tumor cells and spleen factor(s).Proc Natl Acad Sci USA 1977;74:4610–4.

25 Baskar S, Ostrand-Rosenberg S, Nabavi Net al. Constitutiveexpression of B7 restores immunogenicity of tumor cells expressingtruncated major-histocompatibility-complex-class-II molecules. ProcNatl Acad Sci USA 1993;90:5687–90.

26 Chen L, McGowan P, Ashe Set al. Tumor immunogenicity deter-mines the effect of B7 co-stimulation on T-cell-mediated tumorimmunity. J Exp Med 1994;179:523–32.

27 Ramarathinam L, Castle M, Wu Yet al. T-cell costimulation by B7/BB1 induces CD8 T-cell-dependent tumor rejection: an importantrole of B7/BB1 in the induction, recruitment and effector function ofanti-tumor cells. J Exp Med 1994;179:1205–14.

28 Chamberlain RS, Carroll MW, Bronte Vet al. Costimulation enhancesthe active immunotherapy effect of recombinant anticancer vaccines.Cancer Res 1996;56:2832–6.

29 Townsend SE, Allison JP. Tumor rejection after direct co-stimula-tion of CD8þ T cells by B7-transfected melanoma cells. Science1993;259:368–70.

30 Yang S, Darrow TL, Seigler HF. Generation of primary tumor-specificcytotoxic T lymphocytes from autologous and human lymphocyteantigen class I matched allogeneic peripheral blood lymphocytes byB7 gene-modified melanoma cells. Cancer Res 1997;57:1561–8.

31 Sule-Suso J, Arienti F, Melani Cet al. A B7–1-transfected humanmelanoma line stimulates proliferation and cytotoxicity of autologousand allogeneic lymphocytes. Eur J Immunol 1995;25:2737–42.

32 Fenton RT, Sznol M, Luster DGet al. A phase I trial of B7-transfected or parental lethally irradiated allogenic melanoma celllines to induce cell-mediated immunity against tumor-associatedantigen-presented by HLA-A2 or HLA-A1 in patients with stage IVmelanoma. Human Gene Ther 1995;6:87–106.

33 Wang Q, Redovan C, Tubbs Ret al. Selective cytokine geneexpression in renal cell carcinoma tumor cells and tumor-infiltratinglymphocytes. Int J Cancer 1995;61:780–5.

34 Rosenberg SA, Lotze MT, Muul LMet al. A progress report on thetreatment of 157 patients with advanced cancer using lymphokine-activated killer cells and interleukin-2 or high-dose interleukin-2alone. N Engl J Med 1987;316:879–97.

35 Schendel DJ, Gansbacher B, Oberneder Ret al. Tumor-specific lysisof human renal-cell carcinomas by tumor-infiltrating lymphocytes. I.HLA-A2-restricted recognition of autologous and allogeneic tumorlines. J Immunol 1993;151:4209–20.

36 Finke JH, Rayman P, Edinger Met al. Characterization of a humanrenal cell carcinoma specific cytotoxic CD8þ T-cell line. J Immuno-ther 1992;11:1–11.

37 Koo AS, Tso CL, Shimbukuro Tet al. Autologous tumor-specificcytotoxicity of tumor infiltrating lymphocytes derived from humanrenal cell carcinoma. J Immunother 1991;10:347–54.

38 Bernhard H, Karbach J, Wo¨lfel T et al. Cellular immune response tohuman renal cell carcinomas: definition of a common antigenrecognized by HLA-A2-restricted cytotoxic T lymphocytes. Int JCancer 1994;59:1–6.

39 Brouwenstijn N, Gaugler B, Kru¨se KM et al. Renal cell carcinoma-specific lysis by cytotoxic T-lymphocyte clones isolated fromperipheral blood lymphocytes and tumor-infiltrating lymphocytes.Int J Cancer 1996;68:177–82.

40 Riberdy JM, Cresswell P. The antigen processing mutant T2 suggestsa role for MHC-linked genes in class II antigen presentation. JImmunol 1992;148:2586–90.

41 Brodsky FM, Parham P, Barn CJet al. Monoclonal antibodies foranalysis of the HLA system. Immunol Rev 1979;47:342–67.

42 Cayeux S, Beck C, Aicher Aet al. Tumor cells cotransfected withinterleukin-7 and B7.1 genes induce CD25 and CD28 on tumor-infiltrating T lymphocytes and are strong vaccines. Eur J Immunol1995;24:2325–31.

43 Kubin M, Kamoun M, Trinchieri G. Interleukin-12 synergizes withB7/CD28 interaction in inducing efficient proliferation and cytokineproduction of human T cells. J Exp Med 1994;180:211–22.

44 Bain C, Merrouche Y, Puisieux Iet al. B7-1 gene transduction ofhuman renal cell carcinoma cell lines restores the proliferativeresponse and cytotoxic function of allogeneic T cell. Int J Cancer1996;67:769–76.

45 Damle NK, Klussman K, Linsley PS, Aruffo A, Ledbetter JA.Differential regulatory effects of intracellular adhesion molecule-1on costimulation by the CD28 counter-receptor B7. J Immunol1992;149:2541–8.

46 Ghiotto-Rageneau M, Battifora M, Truneh A, Waterfield MD,Olive D. Comparison of CD28-B7.1 and B7.2 functional interac-tion in resting human T-cells: phosphatidylinositol 3-kinase associ-ation to CD28 and cytokine production. Eur J Immunol 1996;26:34–41.

47 Freeman GJ, Boussiotis VA, Anumanthan Aet al. B7-1 and B7-2 donot deliver identical costimulatory signals, since B7-2 but not B7-1preferentially costimulates the initial production of IL-4. Immunity1995;2:523–32.

48 Racke MK, Scott DE, Quigley Let al. Distinct roles for B7-1 (CD80)and B7-2 (CD86) in the initiation of experimental allergic enceph-alomyelitis. J Clin Invest 1995;96:2195–203.

49 Borriello F, Sethna MP, Boyd Det al. B7-1 and B7-2 have over-lapping, critical roles in immunoglobulin class switching and germi-nal center formation. Immunity 1997;6:303–13.

50 Peng X, Remacle JE, Kasran A, Huylebroeck D, Ceuppens JL. IL-12up-regulates CD40 ligand (CD154) expresson on human T cells. JImmunol 1998;160:1166–72.



Documents

Gene Transfer of the Co-Stimulatory Molecules B7-1 and B7-2 Enhances the Immunogenicity of Human Renal Cell Carcinoma to a Different Extent