Interleukin-12 Requires Initial CD80-Mediated T-Cell Activation to Support Immune Responses Toward Human Breast and Ovarian Carcinoma

Home , Interleukin 12

Interleukin-12 requires initial CD80-mediated T-cell activation to support immune responses toward human breast and ovarian carcinoma

Brigitte Gu¨ckel,1,5 Geeske C. Meyer,2 Wolfgang Rudy,3,4 Richard Batrla,3,4 Stefan C. Meuer,3 Gunter Bastert,5 Diethelm Wallwiener,1 and Ulrich Moebius4,6 1Department of Obstetrics and Gynecology, University of Tu¨bingen, Tu¨bingen, Germany; 2Gene Therapy of Tumors Research Group, German Cancer Research Center, Heidelberg, Germany; 3Institute for Immunology, and Departments of 4Urology and 5Obstetrics and Gynecology, University of Heidelberg, Heidelberg, Germany; and 6MediGene, Martiensried, Germany.

One possible reason for the poor immunogenicity of tumors is the induction of peripheral tolerance by tumor cells that fail to deliver costimulatory signals. Furthermore, T cells stimulated with wild-type tumor cells often fail to secrete cytokines. The present study has been undertaken to identify cytokines that cooperate with CD80 in T-cell activation in vitro toward human breast and ovarian carcinoma cell lines. Tumor cell-mediated T-lymphocyte activation was analyzed directly in allogeneic mixed lymphocyte/tumor cell cultures as proliferation and effector functions were assessed in cytotoxic T-cell assays. Interleukin-7 (IL-7) ampliﬁed the proliferative response toward CD80-transfected breast and ovarian carcinomas and stimulated predominantly CD4ϩ T lymphocytes. IL-12 represses the proliferative response of naive T cells but cooperates with CD80-mediated activation during secondary stimulations. In long-term T-cell cultures, IL-12 synergizes with CD80 expression to stimulate cytolytic CD8ϩ T-cell lines, which recognize a breast carcinoma line in a human histocompatibility leukocyte antigen-restricted manner. These studies illustrate that costimulation is necessary for tumor cells to function as alloantigen-presenting cells. Furthermore, when added after the priming of T cells with CD80-transfected tumor cells, IL-12 could be helpful in propagating sufﬁcient T-cell numbers to be used in adoptive transfers during cellular immunotherapy.

Key words: CD80; interleukin-12; tumor immunotherapy; costimulation; lymphocyte activation; cytokines.

he signaling events initiated when T-cell receptor CD80 (B7-1) and CD86 (B7-2), which are expressed Tbinds peptide ligand in the context of the major on dendritic cells, activated macrophages, and B cells, histocompatibility complex (MHC) define the specificity seem to be the most important molecules in delivering of an immune response and enable the development of costimulatory signals.4–7 The interaction of CD80 and/or T-cell functional ability. Depending upon the nature of CD86 with its receptor, CD28, leads to the activation of the stimulus, T lymphocytes can be fully activated, resting T cells by an augmentation of lymphokine secre- partially activated, or become tolerant. The delivery of tion.8–10 Furthermore, antibody (Ab)-mediated signal- “signal 1” via the T-cell receptor without an accompa- ing via the CD28 receptor was shown to prevent the nying costimulatory signal (termed “signal 2”) results in induction of anergy.10,11 In contrast to CD28, CTLA-4, 1–3 a state referred to as anergy. Costimulatory signals the second ligand for CD80/CD86, has been shown to be are provided with the interaction of several T-lympho- a negative regulator of T-cell activation.12–14 cyte adhesion molecules with their respective ligands on 1–4 Because carcinomas commonly express MHC class I antigen (Ag)-presenting cells (APCs). They lead to molecules, and because several tumor-specific Ags have downstream events that control the expression of genes, been defined that could be targeted by cytotoxic T including those for interleukin-12 (IL-12), and eventu- lymphocytes (CTLs) (reviewed in Ref. 15), a possible ally other cytokines. reason for the poor immunogenicity of tumors is the induction of peripheral tolerance by tumor cells that fail to deliver costimulatory signals. We and others have Received December 30, 1997; accepted August 23, 1998. demonstrated that supplying missing costimulatory Address correspondence and reprint requests to Dr. Brigitte Gu¨ckel, signals by the transfection of CD80 into tumor cells Department of Obstetrics and Gynecology, University of Tu¨bingen, facilitates the generation of tumor-reactive human CTLs Schleichstrasse 4, 72076 Tu¨bingen, Germany. in vitro16–19 and leads to the suppression of tumor © 1999 Stockton Press 0929-1903/99/$12.00/ϩ0 growth in several animal models.20–23 However, these

228 Cancer Gene Therapy, Vol 6, No 3, 1999: pp 228–237 GU¨ CKEL, MEYER, RUDY, ET AL: IL-12 AND CD80 SYNERGISTICALLY INDUCE T-CELL ACTIVATION 229 experiments also indicated that CD80 transfection alone line GG was derived from a malignant ascites effusion of an may not be sufficient to activate CTLs and suggested that ovarian cancer patient. Both tumor cell lines were established additional molecules may be required. The optimal in our laboratory and were grown adherently in tissue culture triggering of tumor-specific CTL responses requires flasks in Dulbecco’s modified Eagle’s medium (Life Technol- T-cell help that could be provided by the expression of ogies, Eggenstein, Germany) supplemented with 10% fetal calf serum (BioWhittaker, Walkersville, Md), 2 mM L-glutamine, MHC class II molecules either by transfection or, in and 1% penicillin/streptomycin (Life Technologies). The cases in which MHC class II expression on tumor cells CD80-transfected variants KS24 and GG8 have been described ␥ ␥ 19 could be up-regulated, by interferon- (IFN- ) treat- previously and were maintained as their nontransfected ment and/or tumor necrosis factor-␣ (TNF-␣) treat- parental counterparts, with the exception of G418, which was ment. The coexpression of CD80 and MHC class II supplemented at 0.5 mg/mL. To enhance the expression of resulted in the effective induction of MHC class I-re- MHC class I and class II and several adhesion molecules with stricted tumor immunity, which was mediated in part by costimulatory functions, tumor cells were cultured for 48 hours ϩ ␣ ␥ CD4 T helper (Th) lymphocytes.18,19,24 Alternatively, in complete medium containing TNF- (5 ng/mL) and IFN- (500 U/mL) (both provided by Boehringer Mannheim, Mann- the requirement for helper cells and APCs could be 19 bypassed by the addition of appropriate cytokines.25–31 heim, Germany) as described previously. The NK-sensitive erythroblastoid cell line K562 and the One consequence of the poor immunogenicity of Epstein-Barr virus-transformed B lymphoid cell line Laz509 tumors is that specifically reacting T lymphocytes often were maintained in RPMI 1640. fail to secrete cytokines, which are important for their Peripheral blood mononuclear cells were isolated from the differentiation and propagation. The direct modification heparinized blood of healthy donors by Ficoll-Hypaque (Bio- of tumor cells using cytokine-DNA including IL-2, IL-4, chrom, Berlin, Germany) density centrifugation. T cells were IL-7, IL-12, TNF-␣, IFN-␥, and GM-CSF as a strategy to purified after an adherence step on plastic dishes and subse- support host immunity has been studied intensively in quent rosetting with sheep erythrocytes (ICN, Meckenheim, several animal models.25–31 A variety of specific and Germany). Preactivated T-cell blasts were obtained by incubating resting T cells with phytohemagglutinin (PHA)-L unspecific immune reactions involving T lymphocytes, ␮ natural killer (NK) cells, and macrophages could be (Boehringer Mannheim) at a final concentration of 1 g/mL. Mitogen-activated T lymphocytes were harvested for stimula- involved in tumor rejection. tion experiments and washed intensively after 7 days of PHA In this study we focused on two cytokines, IL-12 and stimulation to minimize the background proliferation of T-cell IL-7, because they were described to be of particular blasts collected at an earlier timepoint. importance in the context of costimulation: IL-12 was originally described as a factor promoting NK cell and 32,33 Human histocompatibility leukocyte Ag (HLA) typing CTL activity. More recently, IL-12 has been found to of tumors and T lymphocytes modulate IFN-␥ secretion and to induce the maturation of Th type 1 (Th1) cells.34,35 Furthermore, IL-12 was Effector T cells of healthy donors and stimulator cells were demonstrated to synergize with CD80/CD28 interactions HLA-A2-matched. Expression of the HLA-A2 allele was de- in inducing efficient proliferation and cytokine produc- termined by immunofluorescence using the monoclonal Ab 36–39 (mAb) BB7.2 (American Type Culture Collection (ATCC), tion of murine and human T lymphocytes. IL-7 was Manassas, Va). The A0201* subtype was confirmed in the initially described as a growth factor for B-cell precur- Department of Immunohematology and Blood Bank (Univer- sors in vitro but subsequently has been demonstrated to ϩ sity Hospital, Leiden, The Netherlands) using specific poly- promote tumor rejection via the expansion of CD8 T merase chain reaction primers. The HLA phenotypes of the lymphocytes and the activation of CTL effector func- tumor cell lines were determined by the Blood Bank, Univer- tions.40–43 IL-7 is also described as a potent costimulus sity of Heidelberg: KS (HLA-A2, HLA-B35,60, HLA-C3,4, involving the CD28 activation pathway and shows syn- HLA-DRB1*15,1302, HLA-DQB1*0602,0604, and HLA- ergistic effects with IL-12 on human T-cell activation.44–46 DQA1*0102,0102), GG (HLA-A2,32, HLA-B7,27, HLA-C1,7, The aim of the present study was the identification of HLA-DRB1*08,11,HLA-DQB1*04,0301, and HLA- cytokines that cooperate with CD80 in T-lymphocyte DQA1*0401,0501). activation toward human breast and ovarian carcinoma cells. We showed previously that CD80 expression is Quantitative flow cytometry necessary for tumor cells to function as alloantigen- Immunofluorescence analysis was performed on a FACScan presenting cells. In the present report, we tried to cytometer (Becton Dickinson, Heidelberg, Germany) as de- develop culture conditions that would allow us to mod- scribed previously.18,19 All immunofluorescence reagents were ulate the immune responses toward genetically modified pretitrated. The following unconjugated mAbs were used in Ј tumor cells by IL-7 and IL-12, thus leading to the combination with goat F(ab )2 anti-mouse immunoglobulin (G propagation of tumor-reactive T lymphocytes. plus L)-fluorescein isothiocyanate (Tago, Camarillo, Calif) as a second-stage reagent: anti-HLA-A2 (BB7.2, ATCC), anti- CD56 (T199, Dianova, Hamburg, Germany), and anti-CD80 (MAB104, Dianova). The secondary Ab alone was used as MATERIALS AND METHODS negative control. Anti-CD3 (UCHT-1), anti-CD4 (13B8.2), Cell lines and culture conditions anti-CD8 (B9.11), anti-CD25 (B1.49.9), and anti-CD28 (CD28.2) (all from Dianova) were used as phycoerythrin and The breast carcinoma cell line KS was established from a fluorescein isothiocyanate conjugates, respectively, in two- pleural effusion of a female patient. The ovarian carcinoma cell color immunofluorescence.

Cancer Gene Therapy, Vol 6, No 3, 1999 230 GU¨ CKEL, MEYER, RUDY, ET AL: IL-12 AND CD80 SYNERGISTICALLY INDUCE T-CELL ACTIVATION

Proliferation assay T cells were cultured in RPMI 1640 containing 10% human sera (BioWhittaker) at 1 ϫ 105 cells/well in 96-well, ﬂat-bottom microtiter plates (Nunc, Roskilde, Denmark) together with 1 ϫ 104 ␥-irradiated (200 Gy) stimulator cells (200 ␮L volume). The following cytokines were added exogenously at various concentrations and timepoints as indicated: recombinant IL-2 (rIL-2) (Cetus Immune, Emeryville, Calif), rIL-7, and rIL-12 (both obtained from R&D Systems, Minneapolis, Minn). Cells were incubated for 5 days and pulsed for an additional 18 hours with 37 kBq of [3H]thymidine (74.0 GBq/mmol, New England Nuclear, Boston, Mass) per well. The incorporation of [3H]thymidine deoxyribose (TdR) was measured by liquid scintillation counting; the results shown are the average of triplicate cultures Ϯ SEM. Stimulator cell background was subtracted from all values.

Propagation of T lymphocytes T cells of healthy donors (1 ϫ 106 cells/well) were cocultured with irradiated allogeneous KS or KS24 breast carcinoma cells (1 ϫ 105 cells/well) in 24-well culture plates (2 mL/well) in complete RPMI 1640 containing 10% human sera (BioWhit- taker). T lymphocytes and stimulator cells were HLA-A2- matched. Cultures were restimulated at weekly intervals at a lymphocyte to tumor cell ratio of 10:1. The cultures contained different cytokines as indicated: rIL-7 was added at 1 ng/mL from the beginning of the stimulations; rIL-12 was added on day 7 (start of the second stimulation) at a ﬁnal concentration of 0.1 ng/mL. On day 3 of the second stimulation (day 10), all cultures were adjusted to 30 U/mL rIL-2.

Cytotoxicity assay Cell-mediated cytotoxicity was tested in a standardized 51Cr release assay using 1 ϫ 103 target cells/well. The 51Cr-labeled target cells were incubated with effector cells at different effector to target (E:T) ratios in V-bottom microtiter plates (Nunc). In blocking studies, target cells were preincubated at 4°C for 15 minutes with anti-MHC class I mAb (W6/32, 10 ␮g/mL, ATCC). Cold target inhibition experiments were performed by adding a 50-fold excess of unlabeled K562. Micro- cultures were set up in triplicate and were spun down following a 3-hour incubation. Subsequently, supernatants were collected and counted in a gamma counter. Specific lysis was calculated according to the following formula % lysis ϭ ([test release Ϫ spontaneous release]/[maximum release Ϫ sponta- Figure 1. IL-12 synergizes with CD80 in inducing proliferation of ϫ preactivated T-cell blasts. A: Resting allogeneic T cells were cul- neous release]) 100. ‚ Spontaneous and maximum releases were determined by tured in medium alone ( ) or with parental KS or GG tumor cells that had been either untreated (Ⅺ) or cytokine-pretreated (f), and with incubating targets without effector cells in medium alone and E F with 1% sodium dodecyl sulfate, respectively. the corresponding CD80 transfectants ( , ) in the presence of the indicated IL-12 concentrations. Proliferation was assessed by To facilitate a comparative analysis of cytolytic activity, 3 results were expressed as lytic activity (LA). One LA was [ H]TdR incorporation after 5 days of culture. B: PHA-preactivated defined as percent lysis at an E:T ratio of 30:1 multiplied by the T-cell blasts of the same donor were used instead of naive T cells total number of effector cells generated. and cultured under the same conditions. The results shown are from one experiment with cells of the same donor and are representative of three independent experiments with T cells from different donors. RESULTS to nontransfected tumor cell lines, stimulated the pro- IL-12 represses the proliferative response of resting liferation of allogeneic resting T lymphocytes in vitro; peripheral blood mononuclear cells toward CD80- proliferation was enhanced further by pretreating tumor transfected breast and ovarian tumor variants in cells with TNF-␣ and/or IFN-␥.19 The influence of IL-12 primary stimulations on the activation of the T lymphocytes induced by ϩ We demonstrated previously that CD80-expressing CD80 KS24 (breast carcinoma) and GG8 (ovarian breast and ovarian tumor cell transfectants, in contrast carcinoma) variants is shown in Figure 1A. Increasing

Cancer Gene Therapy, Vol 6, No 3, 1999 GU¨ CKEL, MEYER, RUDY, ET AL: IL-12 AND CD80 SYNERGISTICALLY INDUCE T-CELL ACTIVATION 231

Figure 2. Kinetics of suppressive (A) and synergistic (B) effects of IL-12. Allogenic effector cells were cultured in medium alone (open bars), cocultured with KS tumor cells (striped bars), or cocultured with the CD80-expressing variant KS24 (ﬁlled bars). Cultures were supplemented with 1 ng/mL IL-12 on day 0, 1, or 3 of the cultivation.

concentrations of IL-12 did not support a proliferative proliferation rate (Fig 2A). Similarly, the augmentation response of resting allogenic T lymphocytes against of the response of preactivated T cells by exogenous unmodified and cytokine-treated KS and GG tumor cells IL-12 was maximal when added during the first 24 hours as measured by [3H]TdR uptake. Surprisingly, IL-12 of stimulation (day 0) (Fig 2B). repressed the T-cell response toward CD80-transfected and cytokine-treated KS as well as GG cells. In the case In contrast to primary stimulations, IL-12 augments of KS cells, in which T-cell proliferation was already T-cell responses and synergizes with CD80-mediated observed following CD80 expression, IL-12 also exerted activation in secondary stimulations an inhibitory effect. Because a cooperative effect between CD28/CD80- To determine whether IL-12 induces a long-lasting mediated costimulation with preactivated T cells has unresponsiveness of naive T lymphocytes, we performed been observed previously,36,39,47 we performed similar rechallenge experiments (Fig 3). As illustrated previ- experiments with PHA-activated T-cell blasts. As shown ously, IL-12 inhibited the proliferative response of rest- in Figure 1B, increasing concentrations of IL-12 induced ing T lymphocytes toward stimulator cells during the a modest proliferation of PHA-preactivated T-cell blasts primary stimulation (Fig 3A). T cells primed separately that was not enhanced by the untransfected parental cell with KS and KS24 in either the absence or presence of lines KS and GG. In case of the CD80-expressing KS24 IL-12 (1000 pg/mL) were then restimulated after 5 days tumor variant, IL-12 augmented T-cell proliferation in a with the same Ag (Fig 3B). Only KS24-primed T lym- dose-dependent fashion up to a 15-fold higher magni- phocytes could be reactivated. Independent of the pres- tude. IFN-␥/TNF-␣ preincubation of KS24 transfectants ence of IL-12 in the primary stimulation, increasing further increased their stimulating potential in compar- IL-12 concentrations in restimulation experiments aug- ison with untreated KS24 cells. IL-12 also augmented mented T-cell proliferation in a dose-dependent man- T-lymphocyte proliferation toward the CD80-trans- ner. The results shown in Figure 3 are representative of fected GG8 tumor variant. However, GG8 tumor cells three independent experiments with T cells from three only induced proliferation following treatment with different donors. Comparable results were obtained with IFN-␥ and/or TNF-␣, as shown previously.19 Our exper- the ovarian tumor cell line GG and the corresponding iments indicate that a cooperative effect between IL-12 CD80-expressing variant (data not shown). These exper- and CD80-expressing tumor transfectants was only ob- iments indicate that IL-12 may cooperate with CD80 in served with preactivated T lymphocytes, but not with the generation of tumor-reactive T-cell lines when used naive T cells. under appropriate conditions. Kinetic experiments, as shown in Figure 2, demonstrated that IL-12 strongly suppressed the proliferation IL-7 promotes proliferation of T lymphocytes of naive T cells toward the KS24 transfectant only when preactivated by CD80-transfected tumor cells added during the early induction phase of the immune response. Addition of the cytokine at a later time (day 1 Figure 4 demonstrates that IL-7 amplifies the prolifera- or day 3) showed only slight suppressive effects on the tive response of naive T cells toward the KS24 breast

Cancer Gene Therapy, Vol 6, No 3, 1999 232 GU¨ CKEL, MEYER, RUDY, ET AL: IL-12 AND CD80 SYNERGISTICALLY INDUCE T-CELL ACTIVATION

Figure 3. IL-12 synergizes with CD80-mediated activation upon Ag challenge. Allogeneic resting T cells were incubated in primary cultures with KS breast carcinoma cells (E), with the CD80-expressing KS24 variant (Ⅺ), or without stimulator cells (‚) in the absence or presence of the indicated IL-12 concentrations. The proliferative responses were determined after 5 days in a primary MLTC. After 5 days of primary culture, primed T cells were rechallenged with KS (E, F) or KS24 (Ⅺ, f) cells in the presence or absence of IL-12. The proliferation rates were measured after 4 days of culture by [3H]TdR uptake (2. MLTC). To determine the residual proliferative activity of T cells without rechallenge, primed effectors were further cultured without stimulator cells in the presence of various IL-12 concentrations (‚, Œ, छ, ࡗ). carcinoma variant by Յ1.5-fold (Fig 4A) and toward under different culture conditions, we performed long- cytokine-preincubated GG8 ovarian carcinoma transfec- term mixed lymphocyte tumor cell cultures (MLTCs) tants by Յ6-fold (Fig 4B). However, we also observed that were stimulated separately with IFN-␥/TNF-␣-pre- that T cells, cultivated either alone or together with treated KS and KS24 cells. The growth kinetics of T nontransfected tumor cells, responded to IL-7 in a lymphocytes of one representative healthy donor are dose-dependent fashion. This ﬁnding indicates that IL-7 shown in Figure 5. The addition of IL-7 or IL-12 supports T-cell proliferation independent of the pres- (beginning on days 0 and 7, respectively) to cultures ence of Ag or CD80, but can also augment the response stimulated with parental IFN-␥/TNF-␣-pretreated KS toward CD80-expressing tumor variants in an additive cells did not lead to long-term T-cell lines. Importantly, manner. the expansion of T lymphocytes stimulated with KS24 was clearly supported by IL-12 or IL-7 up to two IL-12 and IL-7 synergize with CD80-mediated logarithmic units, demonstrating that both ILs cooperate costimulation in inducing efﬁcient proliferation of with CD80 costimulation in long-term T-cell propaga- cytolytic long-term T-cell cultures tion. Similar results were obtained with two different To assess the proliferative and functional activity as well T-cell donors; however, one donor gave rise to stable as the phenotype of T cells that respond to KS variants T-cell lines only following stimulation with KS24 in

Cancer Gene Therapy, Vol 6, No 3, 1999 GU¨ CKEL, MEYER, RUDY, ET AL: IL-12 AND CD80 SYNERGISTICALLY INDUCE T-CELL ACTIVATION 233

Figure 4. Additive/synergistic effects of IL-7 on CD80-mediated activation of T lymphocytes. Allo- geneic T cells were cultured with medium alone (‚), with cytokine-pretreated or untreated parental KS or GG tumor cells (f, Ⅺ), and with the corresponding CD80 transfectants (F, E). IL-7 was added at the indicated concentrations at the start of the culture. Proliferation was assessed by [3H]TdR incorporation after 5 days of culture. combination with IL-7 or IL-12, but not without cyto- decreased toward later timepoints (after day 35) (data kines (see Table 1). not shown). All cultures generated under the conditions described Table 1 also summarizes the cytolytic activity of ϩ above contained predominantly CD3 T cells (93–99%) T-lymphocyte cultures generated by stimulation with KS ϩ and only small amounts of CD56 cells (0.2–6%) that or KS24 cells in the presence of IL-7 and IL-12, respec- were preferentially detected in cultures containing IL- tively. T cells propagated with the parental KS cell line 12. IL-7 supported the propagation and activation of showed either no LA or weak LA toward KS target cells, ϩ CD4 T cells, whereas IL-12 promoted the development especially when an excess of K562 cells was added to ϩ of CD8 T-cell lines. The CD4/CD8 ratios of the T-cell block NK activity. Stimulation with KS24 cells in the lines generated from all three donors are presented in absence of IL-7 or IL-12 gave rise to CTLs in two of Table 1. T cells activated with KS24 cells expressed three donors. IL-7 increased the potential of KS24 cells CD28 during the initial phase of the stimulation, which to induce CTLs to a variable degree. However, the

Figure 5. Growth effects of IL-7 and IL-12 on long-term T-cell cultures. Long- term MLTCs were stimulated separately with IFN-␥/TNF-␣-pretreated KS (open symbols) and KS24 (closed symbols) cells. In all stimulation protocols, the cultures were adjusted to 30 U/mL IL-2 on day 7. IL-7 was added on day 0 of culture, whereas IL-12 was given after T cell priming for 7 days with tumor cells and IL-2 alone. T lymphocytes were restimulated and counted every 7 days. Representative growth kinetics of lymphocytes from one of three healthy donors are shown.

Cancer Gene Therapy, Vol 6, No 3, 1999 234 GU¨ CKEL, MEYER, RUDY, ET AL: IL-12 AND CD80 SYNERGISTICALLY INDUCE T-CELL ACTIVATION

Table 1. Development and Cytotoxicity of T Lymphocytes Stimulated with KS Cells and CD80؉ KS Variant

Lysis of KS under K562 competition CD4/CD8 ratio* (LA ϫ 106)†

Donor Stimulus‡ ϾDay 70 Day 27 Day 64

1 KS/C§ Ϫ¶2Ϫ KS/C ϩ IL-7 4.6 Negative 2 KS/C ϩ IL-12 Ϫ 47 Ϫ KS24/C 0.4 Negative 80 KS24/C ϩ IL-7 2.3 122.5 20,962 KS24/C ϩ IL-12 0.02 1232 56,545 2 KS/C Ϫ Negative Ϫ KS/C ϩ IL-7 Ϫ 10 Ϫ KS/C ϩ IL-12 Ϫ 19 Ϫ KS24/C Ϫ Negative Ϫ KS24/C ϩ IL-7 99.0 40 1093 KS24/C ϩ IL-12 0.12 425 9242 3 KS/C Ϫ 93 Ϫ KS/C ϩ IL-7 Ϫ 356 Ϫ KS/C ϩ IL-12 Ϫ 317 477 KS24/C 0.04 224 22,669 KS24/C ϩ IL-7 1.42 269 55,718 KS24/C ϩ IL-12 0.05 815 117,858 *Phenotypic analysis of permanent T cell lines was performed on day 70 by immunofluorescence. †The cytolytic response of T cells was tested against KS targets under “cold” target competition with unlabeled K562 cells at an inhibitor to target ratio of 50:1. Lysis of targets was assessed in a 51Cr release assay at four different E:T ratios. The results presented here were calculated and converted in LA as described in Materials and Methods. ‡Allogeneic T lymphocytes were stimulated weekly with CD80-transfected (KS24) or nontransfected (KS) breast tumor cells as described in Figure 5. §Stimulator cells were pretreated with a combination of TNF-␣ (25 ng/mL) and IFN-␥ (500 U/mL) for 48 hours. ¶Insufficient expansion of T cells, analysis not possible. strongest LA toward KS cells and the lowest NK activity proliferation, the response of activated T cells was (cytolysis of K562, data not shown) were observed when augmented by IL-12 when rechallenged with the same KS24 cells were used to stimulate T cells in the presence CD80-transfected tumor cells. A synergy between of IL-12. This finding was more obvious when cells were CD80-mediated costimulation and IL-12 on preacti- analyzed on day 64. To examine whether this reactivity vated T cells has been demonstrated previously with against KS tumor cells was MHC-restricted, blocking murine Th1 clones and human PHA-activated T-cell experiments using mAbs were performed. As shown in Figure 6, the cytolysis of KS tumor cells was inhibited by an MHC class I-specific mAb (84% inhibition, data shown for donor 1).

DISCUSSION

The modification of malignant cells with genes coding for costimulatory molecules such as CD80 is aimed at endowing tumor cells with professional Ag-presenting competence. Current studies have demonstrated that CD80 expression on tumors is already necessary, yet insufficient to elicit T-cell responses, and that the addition of appropriate cytokines still improves the generation of tumor-reactive T lymphocytes. Thus, IL-12 and Figure 6. CD80 costimulation together with IL-12 promotes the IL-7 were reported to enhance the therapeutic effect of activation of MHC class I-restricted T lymphocytes. The cytolytic 28,37,38,43,44,48 responses of T cells (representative data of donor 1, see also Fig 5 murine CD80-transfected tumor vaccines. and Table 1) stimulated with cytokine-pretreated KS24 cells in the By analyzing the proliferative response as a parameter presence of IL-12 were tested on day 64 against parental KS cells for T-cell activation in allogeneic MLTCs, we observed (with and without cold target competition) and against K562 cells at that IL-12 exerted a bimodal function on the stimulation an E:T ratio of 25:1. An MHC class I-specific mAb (W6/32) was used of T cells by CD80-transfected breast and ovarian cancer to inhibit the cytolysis of KS target cells. For details see Materials cell lines. Although IL-12 inhibited the primary T-cell and Methods.

Cancer Gene Therapy, Vol 6, No 3, 1999 GU¨ CKEL, MEYER, RUDY, ET AL: IL-12 AND CD80 SYNERGISTICALLY INDUCE T-CELL ACTIVATION 235 blasts.36,39 An inhibition of primary T-cell activation was tumor Ags can be generated by stimulation with HLA- also observed in a study performed with CD80-trans- matched allogeneic tumor cell lines as demonstrated in fected melanoma cells.47 Interestingly, this inhibitory the case of melanoma.54,55 Our results suggest that the effect was only observed with CD80, but not CD86- coapplication of IL-12 with CD80-transfected tumor transfected melanoma cells, which is in concordance cells in a clinical protocol could be harmful and inhibit with a functional difference of both molecules. Prelimi- tumor-specific responses. In this regard, the use of nary results indicate that the inhibition of primary T-cell CD86-expressing tumor cell vaccines may have an ad- activation induced by CD80-transfected breast and ovar- vantage over CD80-transfected lines.48 IL-12, when used ian cancer cells is mediated through IFN-␥. Confirming under appropriate conditions in vitro, can act as a T-cell earlier observations, IFN-␥ does not always function in growth factor. Here, we demonstrated that IL-12 allows the same direction (i.e., seems to be an early activation the continuous in vitro expansion of T cells to obtain signal for T lymphocytes and a late signal to switch off sufficient cell numbers to be used in adoptive transfer this response through the induction of apoptosis).49,50 during cellular immunotherapy. These contrasting effects seemed to be directly linked to ␥ ␤ the expression of the inhibitory IFN- -receptor -chain ACKNOWLEDGMENTS and to the presence of accessory cells.50–53 In addition to IL-12, IL-7 was shown to act synergis- We thank Jutta Funk and Maria Szekelyi for expert technical assistance and the physicians in the Department of Gyneco- tically with CD80-mediated costimulation on the prolif- 45,46 logic Oncology for supplying us with tumor tissue and blood eration of T lymphocytes. Furthermore, IL-7 was samples. Special thanks go to Anke Meyer and Bruno Kyewski shown to enhance the immunogenicity of murine CD80- for critically reading the manuscript. This work was supported transfected tumor cells, even if those tumors failed to by grants from the Tumorzentrum Heidelberg/Mannheim. elicit a sufficient antitumor response.44 Possible mechanisms might be the induction of several cytokines (IL-2, REFERENCES IL-4, IL-12, TNF, GM-CSF) augmenting tumor Ag presentation by APCs, and an up-regulation of cytokine 44–46 1. Mueller DL, Jenkins MK, Schwartz RH. Clonal expansion receptors on the corresponding T cells. Our results versus functional clonal inactivation: a costimulatory sig- from primary and long-term stimulations of T lympho- nalling pathway determines the outcome of T cell antigen cytes with CD80-transfected breast and ovarian carci- receptor occupancy. Annu Rev Immunol. 1989;7:445–480. noma cells support the cooperative effect of IL-7 and 2. Schwartz RH. A culture model for T lymphocyte clonal CD80 costimulation. anergy. Science. 1990;248:1349–1356. To analyze T-cell effector functions, we established 3. Jenkins MK, Burrel E, Ashwell JD. Antigen presentation long-term MLTCs by adding IL-12 after T-cell priming by resting B cells: effectiveness at inducing T cell prolifer- with tumor cells alone. In all cases, IL-12 was only able ation is determined by costimulatory signals, not TCR to promote proliferation in cooperation with CD80- occupancy. J Immunol. 1990;144:1585–1589. mediated costimulation. The observation that the timing 4. Freeman GJ, Freedman AS, Segil JM, et al. B7, a member of the Ig superfamily with unique expression on activated of the addition of IL-12 to MLTCs is critical (almost no and neoplastic B cells. J Immunol. 1989;143:2714–2722. inhibitory effect was observed if added at day 3 or later 5. Azuma M, Cayabyab M, Buck D, et al. CD28 interaction or in secondary stimulations) may be explained by the with B7 costimulates primary allogenic proliferative re- finding that preactivated T cells produced less IFN-␥ sponses and cytotoxicity mediated by small, resting T when restimulated with CD80-transfected tumor cells lymphocytes. J Exp Med. 1992;175:353–360. and did not respond to exogenous IFN-␥.47 IL-12 was 6. Linsley PS, Green JL, Tan P, et al. Coexpression and reported to activate NK activity predominantly and to functional cooperation of CTLA-4 and CD28 on activated augment CTL development.32,33 Indeed, the NK activity T lymphocytes. J Exp Med. 1992;176:1595–1604. ϩ of T cells expanded with CD80 KS24 cells in the 7. June CH, Bluestone JA, Nadler LM, et al. The B7 and presence of IL-12 was higher during the first four CD28 receptor families. Immunol Today. 1994;15:321–331. 8. Lindsten T, June CH, Ledbetter JA, et al. Regulation of stimulation cycles compared with T lymphocytes stimu- ϩ lymphokine messenger RNA stability by a surface-medi- lated with CD80 KS24 cells alone. However, at later ated T-cell activation pathway. Science. 1989;244:339–342. timepoints, increased cytolysis of parental KS cells by ϩ 9. Fraser JD, Irving BA, Crabtree GR, et al. Regulation of predominantly CD8 T cells was observed. In contrast, interleukin-2 gene enhancer activity by the T cell accessory the T-cell lines generated in the presence of IL-7 turned molecule CD28. Science. 1991;251:313–316. ϩ out to consist mainly of CD4 T cells exhibiting less LA 10. Gimmi CD, Freeman GJ, Gribben JG, et al. B-cell surface toward the parental tumor lines. antigen B7 provides a costimulatory signal that induces T Several clinical studies were initiated to investigate cells to proliferate and secrete IL-2. Proc Natl Acad Sci the therapeutic potential of CD80-transfected tumor cell USA. 1991;88:6575–6579. vaccines in combination with different cytokines. To 11. Gimmi CD, Freeman GJ, Gribben JG, et al. Human T-cell clonal anergy is induced by antigen presentation in the circumvent the requirement for generating individual absence of B7 costimulation. Proc Natl Acad Sci USA. vaccines for each patient, allogeneic cell lines with 1993;90:6586–6590. shared HLA alleles inducing responses against common 12. Walunas TL, Lenschow TJ, Bakker CY, et al. CTLA-4 can tumor Ags are under investigation. These studies are function as a negative regulator of T cell activation. based on the finding that T cells specific for shared Immunity. 1994;1:405–413.

Cancer Gene Therapy, Vol 6, No 3, 1999 236 GU¨ CKEL, MEYER, RUDY, ET AL: IL-12 AND CD80 SYNERGISTICALLY INDUCE T-CELL ACTIVATION

13. Waterhouse P, Penninger JM, Timms E, et al. Lympho- irradiated tumor cells engineered to secrete murine gran- proliferative disorders with early lethality in mice deficient ulocyte-macrophage colony-stimulating factor stimulates in CTLA-4. Science. 1995;270:985–988. potent, specific, and long-lasting anti-tumor immunity. 14. Leach DR, Krummel MP, Allison JP. Enhancement of Proc Natl Acad Sci USA. 1993;90:3539–3543. antitumor immunity by CTLA-4 blockade. Science. 1996;271: 32. Chehimi J, Starr S, Frank I, et al. Natural killer (NK) cell 1734–1736. stimulatory factor increases the cytotoxic activity of NK 15. Boon T, van der Bruggen P. Human tumor antigens recog- cells from both healthy donors and human immunodefi- nized by T lymphocytes. J Exp Med. 1996;183:725–729. ciency virus-infected patients. J Exp Med. 1992;175:789– 16. Baskar S, Ostrand-Rosenberg S, Nabavi N, et al. Consti- 796. tutive expression of B7 restores immunogenicity of tumor 33. Gately MK, Wolitzky AG, Quinn PM, et al. Regulation of cells expressing truncated MHC class II molecules. Proc human cytolytic lymphocyte responses by interleukin-12. Natl Acad Sci USA. 1993;90:5687–5690. Cell Immunol. 1992;143:127–142. 17. Do¨hring C, Angman L, Spagnoli G, et al. T-helper- and 34. Chan SH, Kobayashi M, Santoli D, et al. Mechanisms of accessory-cell-independent cytotoxic responses to human IFN-␥ induction by natural killer cell stimulatory factor tumor cells transfected with a B7 retroviral vector. Int J (NKSF/IL-12): role of transcription and mRNA stability in Cancer. 1994;57:754–759. the synergistic interaction between NKSF and IL-2. J Im- 18. Habicht A, Lindauer M, Galmbacher P, et al. Develop- munol. 1992;148:92–98. ment of immunogenic colorectal cancer cell lines for 35. Manetti R, Parronchi P, Guidizi MG, et al. Natural killer vaccination: expression of CD80 (B7.1) is not sufficient to cell stimulatory factor interleukin 12 (IL-12) induces T restore impaired primary T cell activation in vitro. Eur J helper type 1 (Th1)-specific immune responses and inhib- Cancer. 1995;31:2396–202. its the development of IL-4-producing Th cells. J Exp Med. 19. Gu¨ckel B, Lindauer M, Rudy W, et al. CD80-transfected 1993;177:1199–1204. human breast and ovarian tumor cell lines: improved 36. Murphy EE, Terres G, Macatonia SE, et al. B7 and ϩ immunogenicity and induction of cytolytic CD8 T lym- interleukin 12 cooperate for proliferation and interferon ␥ phocytes. Cytokines Mol Ther. 1995;1:211–221. production by mouse T helper clones that are unrespon- 20. Chen L, Ashe S, Brady WA, et al. Costimulation of sive to B7 costimulation. J Exp Med. 1994;180:223–231. anti-tumor immunity by the B7 counter receptor for the T 37. Zitvogel L, Robbins PD, Storkus WJ, et al. Interleukin-12 lymphocyte molecules CD28 and CTLA4. Cell. 1992;71: and B7.1 co-stimulation cooperate in the induction of 1093–1102. effective antitumor immunity and therapy of established 21. Townsend S, Allison JP. Tumor rejection after direct tumors. Eur J Immunol. 1996;26:1335–1341. ϩ costimulation of CD8 T cells by B7-transfected mela- 38. Coughlin CM, Wysocka M, Kurzawa HL, et al. B7-1 and noma cells. Science. 1993;259:368–370. interleukin 12 synergistically induce antitumor immunity. 22. Hodge JW, Abrams S, Schlom J, et al. Induction of Cancer Res. 1995;55:4980–4987. antitumor immunity by recombinant vaccinia viruses ex- 39. Kubin M, Kamoun M, Trinchieri G. Interleukin 12 syner- pressing B7-1 or B7-2 costimulatory molecules. Cancer gizes with B7/CD28 interaction in inducing efficient pro- Res. 1994;54:5552–5555. liferation and cytokine production of human T cells. J Exp 23. Yang G, Hellstro¨m KE, Hellstro¨m I, et al. Antitumor Med. 1994;180:211–222. immunity elicited by tumor cells transfected with B7-2, a 40. Namen EA, Schmierer AE, March CJ, et al. B cell second ligand for CD28/CTLA-4 costimulatory molecules. precursor growth-promoting activity: purification and J Immunol. 1995;154:2794–2800. characterization of a growth factor active on lymphocyte 24. Lindauer M, Rudy W, Habicht A, et al. Requirement of precursors. J Exp Med. 1988;16:988–1002. CD54, CD80, and class II MHC expression of a human 41. Chazen DG, Pereira MBG, Legros G, et al. Interleukin-7 colorectal cancer cell line to induce primary T cell activa- is a T cell growth factor. Proc Natl Acad Sci USA. tion and class I MHC-restricted cytolytic T lymphocytes. 1989;86:5923–5927. Immunology. 1998;93:390–397. 42. Alderson MR, Sassenfeld HM, Widmer MB. Interleukin-7 25. Fearon ER, Pardoll DM, Itaya T, et al. Interleukin-2 enhances cytolytic T lymphocyte generation and induces production by tumor cells bypasses T helper function in lymphokine-activated killer cells from human peripheral the generation of an antitumor response. Cell. 1990;60: blood. J Exp Med. 1990;172:577–587. 397–403. 43. Lynch DH, Miller RE. Interleukin 7 promotes long-term 26. Tepper RI, Pattengale PK, Leder P. Murine interleukin-4 in vitro growth of antitumor cytotoxic T lymphocytes with displays potent anti-tumor activity in vitro. Cell. 1989;57: immunotherapeutic efficacy in vivo. J Exp Med. 1994;179: 503–512. 31–42. 27. Jicha DL, Schwarz S, Mule JJ, et al. Interleukin-7 mediates 44. Cayeux S, Beck C, Aicher A, et al. Tumor cells cotrans- the generation and expansion of murine allosensitized and fected with interleukin-7 and B7.1 induce CD25 and CD28 antitumor CTL. Cell Immunol. 1992;14:71–83. on tumor-infiltrating T lymphocytes and are strong vac- 28. Brunda MJ, Luistro L, Warrier RR, et al. Antitumor and cines. Eur J Immunol. 1995;25:2325–2331. antimetastatic activity of interleukin 12 against murine 45. Costello R, Brailly H, Mallet F, et al. Interleukin-7 is a tumors. J Exp Med. 1993;178:1223–1230. potent co-stimulus of the adhesion pathway involving CD2 29. Blankenstein T, Qin Z, U¨ berla K, et al. Tumor suppression and CD28 molecules. Immunology. 1993;80:451–457. after tumor cell-targeted tumor necrosis factor ␣ gene 46. Mehrotra PT, Grant AJ, Siegel JP. Synergistic effects of transfer. J Exp Med. 1991;173:1047–1052. IL-7 and IL-12 on human T cell activation. J Immunol. 30. Gansbacher B, Bannerji R, Daniels B, et al. Retroviral 1995;154:5093–5102. vector-mediated ␥-interferon gene transfer into tumor 47. Norton SD, Zuckerman Z, Urdahl KB, et al. The CD28 cells generates potent and long-lasting antitumor immu- ligand, B7, enhances IL-2 production by providing a nity. Cancer Res. 1990;50:7820–7825. costimulatory signal to T cells. J Immunol. 1992;149:1556– 31. Dranoff G, Jaffee E, Lanzenby A, et al. Vaccination with 1561.

Cancer Gene Therapy, Vol 6, No 3, 1999 GU¨ CKEL, MEYER, RUDY, ET AL: IL-12 AND CD80 SYNERGISTICALLY INDUCE T-CELL ACTIVATION 237

48. Rudy W, Gu¨ckel B, Siebels M, et al. Differential function 52. Pernis A, Gupta S, Gollob KJ, et al. Lack of interferon ␥ of CD80- and CD86-transfected human melanoma cells in receptor ␤ chain and the prevention of interferon ␥ the presence of IL-12 and IFN-␥. Int Immunol. 1997;9: signalling in TH1 cells. Science. 1995;269:245–247. 853–860. 53. Novelli F, Bernabei P, Ozmen L, et al. Switching on the 49. Nastala CL, Edington HD, McKinney TG, et al. Recom- proliferation or apoptosis of activated human T lympho- binant IL-12 administration induces tumor regression in cytes by IFN-␥ is correlated with the differential expression association with IFN-␥ production. J Immunol. 1994;153: of the ␣- and ␤-chains of its receptor. J Immunol. 1996; 1697–1706. 157:1935–1943. 50. Liu Y, Janeway CA. Interferon ␥ plays a critical role in 54. Stevens EJ, Jacknin L, Robbins PF, et al. Generation of inducing cell death of effector T cells: a possible third tumor-specific CTLs from melanoma patients by using mechanism of self tolerance. J Exp Med. 1990;172:1735– peripheral blood stimulated with allogeneic melanoma 1740. tumor cell lines. J Immunol. 1995;154:762–771. 51. Novelli F, Di Pierro F, Franca di Celle P, et al. Environ- 55. Yang S, Darrow TL, Seigler HF. Generation of primary mental signals influencing expression of the IFN-␥ re- tumor-specific cytotoxic T lymphocytes from autologous ceptor on human T cells control whether IFN-␥ pro- and human lymphocyte antigen class I-matched allogeneic motes proliferation or apoptosis. J Immunol. 1994;152: peripheral blood lymphocytes by B7 gene-modified mela- 496–504. noma cells. Cancer Res. 1997;57:1561–1568.

Cancer Gene Therapy, Vol 6, No 3, 1999