Expression of human GITRL on myeloid dendritic cells enhances their ...

Expression of human GITRL on myeloid dendritic cells enhances their immunostimulatory function but does not abrogate the suppressive effect of CD4⫹CD25⫹ regulatory T cells Sandra Tuyaerts,*,1 Sonja Van Meirvenne,*,1 Aude Bonehill,* Carlo Heirman,* Jurgen Corthals,* Herman Waldmann,† Karine Breckpot,* Kris Thielemans,* and Joeri L. Aerts*,2 *Laboratory of Molecular and Cellular Therapy, Department of Physiology and Immunology, Medical School of the Vrije Universiteit Brussel, Brussels, Belgium; and †Sir William Dunn School of Pathology, University of Oxford, Oxford, United Kingdom

Abstract: CD4ⴙCD25ⴙ regulatory T cells (Treg) have been described as an important hurdle for immunotherapy. Engagement of glucocorticoid-induced TNF receptor-related protein (GITR) has emerged recently as an important mechanism to control the suppression of CD4ⴙCD25ⴙ Treg. Furthermore, it has been documented extensively that GITR ligation is costimulatory for naive and activated T cells in the murine setting. However, little is known about the role of the human GITR ligand (huGITRL). We wanted to explore whether huGITRL could enhance antigen-specific T cell priming by dendritic cells (DC). First, we confirmed the endogenous expression of GITRL on HUVEC. We also detected GITRL expression on EBV-B cell lines, whereas no GITRL expression was observed on human monocyte-derived DC. Electroporation of GITRL mRNA in monocyte-derived DC resulted in a strong and long-lasting surface expression of GITRL. In contrast to data obtained in mice, no significant abrogation of Treg suppression by GITRL-expressing human DC was observed. Consistent with our mouse data, we showed that huGITRL is costimulatory for responder T cells. Furthermore, we found that GITRL-expressing DC primed increased numbers of Melan-A-specific CD8ⴙ T cells. We conclude that although huGITRL is not capable of alleviating Treg suppression of responder T cells, huGITRL overexpression on monocyte-derived DC enhances their capacity to induce antigen-specific T cell responses. Thus, GITRL incorporation in DC might improve the antitumor immune response after vaccination. J. Leukoc. Biol. 82: 93–105; 2007. Key Words: GITR 䡠 vaccination 䡠 costimulation 䡠 immunotherapy

later also applied to T cell activation [2]. This model states that for productive activation of naive T cells, it is necessary, not only to provide a signal to the TCR in the form of peptide-MHC complex (signal 1), but also, additional signals need to be provided through cognate receptors on the T cell and the APC (signal 2). During the last few years, the number of potentially costimulatory molecules has increased dramatically, and the function of each of these has been refined. It appears that these costimulatory molecules not only provide activation signals to naive T cells but also modulate the properties of the T cells, such as the expression pattern of specific surface markers and cytokine production. However, some of these surface markers have been described to exert inhibitory effects on T cells to down-regulate T cell responses. Thus, distinct combinations of costimulatory/inhibitory molecules expressed on APC reflect the stimulus these cells have received through their encounter with pathogens. In turn, these stimuli can be imprinted upon the T cells, ultimately resulting in an adequately controlled immune response [3–9]. The costimulatory molecules can broadly be divided in two groups according to their structural features. One group consists of the B7-CD28 family, based on structural homology of its members with the main costimulatory pair B7.1 (or CD80) and B7.2 (or CD86) on APC and CD28 on T cells. Other members of this family include ICOS ligand, B7-H1, and B7-dendritic cells (DC), which are expressed on APC and ICOS, CTLA-4, and programmed death-1 (PD-1) expressed on T cells. Most costimulatory molecules are members of the TNF superfamily, however. They include CD40, CD70, 4-1BBL, CD30L, OX40 ligand (OX40L), and lymphotoxin-related inducible ligand that competes for glycoprotein D binding to herpes virus entry mediator on T cells (LIGHT) on APC and CD40 ligand (CD40L; or CD154), CD27, 4-1BB (or CD137), CD30, OX40,

1

These authors contributed equally to this work. Correspondence: Laboratory of Molecular and Cellular Therapy, Department of Physiology and Immunology, Medical School of the VUB, Laarbeeklaan 103/E, 1090 Brussels, Belgium. E-mail: [email protected] Received September 14, 2006; revised March 14, 2007; accepted March 30, 2007. doi: 10.1189/jlb.0906568 2

INTRODUCTION A generally accepted theory in immunology is the two-signal hypothesis, originally developed for B cell activation [1] but 0741-5400/07/0082-0093 © Society for Leukocyte Biology

Journal of Leukocyte Biology Volume 82, July 2007 93

and herpesvirus entry mediator on T cells. Recently, glucocorticoid-induced TNF receptor (HVEM)-related protein (GITR) was identified as a new member of this family and was found to be constitutively expressed on CD4⫹CD25⫹ T regulatory cells (Treg) and on CD4⫹CDC25– T cells upon activation [10 –13]. Functionally, it was shown that engagement of GITR abrogates the suppressive function of Treg in mice [13–17], although it is unclear whether this system works directly on Treg [14, 15] or through a mechanism where responder T cells are protected against subsequent Treg inhibition by GITR cross-linking [16]. GITR ligation was also demonstrated to be costimulatory for responder T cells [18 –21] and NKT cells [22]. Moreover, GITR triggering in vivo resulted in a marked tumor reduction in tumor-bearing mice [23], enhanced virus-specific T cell immunity [24, 25], attenuation of graft-versus-host disease [26], and exacerbation of autoimmune responses [14, 27–30] and was a successful stimulant of antigen-specific vaccination [25, 31]. Furthermore, GITR triggering can provoke or protect from apoptosis, depending on the strength of the stimulus [10, 19, 27, 32–35]. Recently, a role for human GITR (huGITR) in the costimulation of NK cells by GITR ligand (GITRL)-expressing, activated plasmacytoid DC was described [36]. Based on these observations, we postulated that GITRL could enhance the efficacy of DC to activate immune responses. In the murine system, GITRL was shown to be expressed on CD11c⫹ DC [19, 37]. Although the role of GITRL in Treg suppression and costimulation of responder T cells has been described abundantly in the mouse system, little data are available for the expression and function of GITRL in human cells. Therefore, we addressed this issue through a model in which we overexpress GITRL on human monocyte-derived DC. In this report, we confirm that huGITRL and mouse GITRL (moGITRL) have significantly different expression patterns. Furthermore, we show that comparable with the murine system, huGITRL is costimulatory for CD4⫹ and CD8⫹ T cells. In contrast to data obtained in mice, however, we could not observe abrogation of Treg suppression by huGITRL. In addition, we demonstrate that introduction of huGITRL into monocyte-derived DC enhances their capacity to prime antigenspecific CD8⫹ T cells, as was also described for its murine equivalent [31, 38].

MATERIALS AND METHODS Mice, cell lines, and tumor-infiltrating lymphocytes (TIL) Six- to 8-week-old female C57BL/6 (H-2b) mice were purchased from Harlan (Horst, The Netherlands). Animals were maintained and treated according to the institutional guidelines. 293T and EA·hy 926 (HUVEC line, ref. [39], kind gift of Dr. Ivan Van Riet, Department of Hematology and Immunology, Vrije Universiteit Brussel, Brussels, Belgium) were maintained in DMEM (Cambrex, Verviers, Belgium), supplemented with 10% FCS (Harlan Sera-Lab, Sussex, UK), 100 U/ml penicillin and 100 ␮g/ml streptomycin (PS; Cambrex), and 0.24 mM Lasparagine, 0.55 mM L-arginine, and 2 mM L-glutamine (AAG; Cambrex). T2 cells (TAP-deficient, HLA-A2⫹) were cultured in IMDM containing 10% FCS, PS, and AAG. The P815 mouse mastocytoma cell line, melanoma cell lines 624-mel, 938-mel, and 1087-mel, and EBV-B cell lines 888-EBV, 1087-EBV, and 1088-EBV were grown in RPMI 1640 containing 10% FCS, PS, and AAG

94

Journal of Leukocyte Biology Volume 82, July 2007

[melanoma and EBV-B cells were a kind gift from Dr. Suzanne Topalian, Surgery Branch, National Cancer Institute (NCI), National Institutes of Health (NIH), Bethesda, MD, USA]. The HLA-A2-restricted TIL, TIL 1941 (MART-1 specific) and TIL L2D8 (gp100-specific), were kind gifts from Drs. John Wunderlich and Mark Dudley (Surgery Branch, NCI, NIH), respectively, and were maintained in RPMI 1640 containing 10% heat-inactivated human AB serum (PAA Laboratories, Linz, Austria), PS, AAG, and 6000 IU/ml IL-2 (Chiron, Emeryville, CA, USA). The tumor cell line EL4 was cultured in DMEM containing 5% heat-inactivated FCS (Harlan), L-glutamine, PS, and 50 ␮M 2-ME.

cDNA cloning The huGITRL cDNA was cloned from HUVEC mRNA, based on expression data from the literature [12]. Full-length huGITRL was amplified by PCR and cloned into a pCR2.1 vector by TA cloning; a Kozak sequence was introduced. Plasmids obtained from positive colonies were verified by restriction and sequence analysis. huGITRL was excised as an XbaI-EcoRI fragment and ligated in pGEM4Z-A64. Cloning of enhanced green fluorescent protein (eGFP) and truncated nerve growth factor receptor (tNGFR) in pGEM4Z-A64 was described previously [40]. The production of in vitro-transcribed mRNA in endotoxin-free conditions was described previously [40]. Integrity and purity were tested routinely for in vitro-synthesized mRNA.

Lentiviral transduction of EL4 cells The pWPT-eGFP and pWPT-moGITRL lentiviral vectors were kindly provided by Dr. Randolph Noelle (Department of Microbiology and Immunology, Dartmouth Medical School and the Norris Cotton Cancer Center, Lebanon, NH, USA). Lentiviral particles were generated, concentrated, and titrated as described [41]. For transduction, 106 EL4 cells were resuspended in 100 ␮l Opti-MEM (Invitrogen, Ghent, Belgium) containing eGFP- or moGITRL-encoding lentiviral particles at a multiplicity of infection of, respectively, 15 or 5 in the presence of 10 ␮g/ml protamine sulfate. After a 2-h incubation at 37°C, DMEM containing supplements was added to reach a final concentration of 106 cells/ml. After expansion, eGFP- and moGITRL-positive cells were sorted by flow cytometry, which resulted in stable eGFP- and moGITRL-expressing EL4 cell lines.

Evaluation of huGITRL and human forkhead box p3 (Foxp3) expression by quantitative RT-PCR Total RNA was extracted using the SV Total RNA isolation system (Promega, Madison, WI, USA), and cDNA was prepared using the Superscript first-strand preamplification kit (Invitrogen). For the PCR reaction, primers and probes (always 5⬘ FAM- and 3⬘ TAMRA-labeled) for huGITRL [forward-primer (FP): agt tgg cta atc ttt att ttt ctc caa, reverse-primer (RP): aga aga tgc cat ttg cca ttt t, Taqman probe (TP): ccc tgt atg gct aag ttt gga cca tta ccc] and human Foxp3 (FP: ggc act cct cca gga cag, RP: gct gat cat ggc tgg gct ct, TP: att tca tgc acc agc tct caa cgg) were developed. As a control gene, ␤-glucuronidase was used as described previously [42]. All primers and probes were synthesized by Thermo Electron Corp. (Ulm, Germany). PCR conditions were as follows: 50°C for 2 min and 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. PCR reactions were performed on an ABI Prism 7700 (Applied Biosystems, Warrington, UK), and analysis was done by Sequence Detection System software (Applied Biosystems). All primersets were intron-spanning to avoid amplification of contaminating genomic DNA.

Purification of Treg Human CD4⫹CD25⫹ Treg were purified from PBMC of cancer patients as described previously [43]. Briefly, PBMC were negatively enriched for CD4⫹ T cells using the CD4⫹ T cell isolation kit II (Miltenyi Biotec, Bergisch Gladbach, Germany), after which CD4⫹CD25⫹ T cells were positively selected through the combined use of CD25-PE antibody (PharMingen, San Jose, CA, USA) and anti-PE microbeads (Miltenyi Biotec) and selection of the positive cells using an LS and MS column (Miltenyi Biotec) subsequently. Pure CD4⫹CD25– responder T cells were obtained by depleting the residual CD25⫹ cells in the CD4⫹CD25– fraction obtained during Treg selection over an LD column (Miltenyi Biotec).

http://www.jleukbio.org

Murine CD4⫹CD25⫹ Treg and CD4⫹CD25– responder T cells were purified from spleen cell suspensions using the CD4⫹CD25⫹ regulatory T cell isolation kit purchased from Miltenyi Biotec. Briefly, CD4⫹ T cells were pre-enriched by depleting CD8⫹ T cells, NK cells, B cells, macrophages, and RBC. Subsequently, CD25⫹ cells were positively selected from the enriched CD4⫹ T cell fraction using a CD25-PE antibody and anti-PE microbeads.

In vitro generation of human monocyte-derived DC PBMC were used as a source of DC precursors and isolated from leukapheresis products of healthy donors. The generation and the cryopreservation of immature DC (iDC) and mature DC (mDC) were described previously [44]. On Day 5, iDC were harvested and matured or cryopreserved in liquid nitrogen until further use. iDC were matured for 24 h at a cell density of 2.5 ⫻ 105 DC/ml by addition of an inflammatory cytokine cocktail as described previously [44].

Electroporation of cells iDC or mDC, K562 cells, P815 cells, or T2 cells were electroporated with RNA using a protocol optimized previously [40, 45]. Briefly, 8 ⫻ 106 cells were electroporated with 20 ␮g RNA; immediately after electroporation, cells were transferred into serum-containing medium for further use.

Synthetic peptides and peptide pulsing The Melan-A/MART-1-derived peptide corresponding to the optimized, immunodominant epitope (amino acids 26 –35; ELAGIGILTV) [46] and the gp100 (209-2M) peptide (IMDQVPFSV) [47], both HLA-A*0201-restricted, were purchased from Thermo Electron. The HLA-A2-restricted gag peptide (gag-A2 peptide, HXB2 gag peptide-complete set, NIH, AIDS Research and Reference Reagent Program, McKesson BioServices Corp., Rockville, MD, USA) was used as a negative control. DC or T2 cells were loaded with peptide as described before [45].

Flow cytometry Human monocyte-derived DC were stained using mAb to CCR7 (unlabeled), CD11c-allophycocyanin, CD25-PE, CD40-PE, CD80-PE, CD83-PE, CD86PE, HLA-ABC-FITC (all from PharMingen), and HLA-DR-biotin (Clone L243, prepared in-house). huGITRL was detected using a mouse anti-huGITRL mAb (Clone 109114, R&D Systems, Oxford, UK) in conjugation with a PE-labeled goat anti-mouse secondary antibody (Serotec, Oxford, UK). Human T cells were phenotyped with mAb to CD4-FITC/PerCP-Cy5.5, CD8-PE-Cy5.5, CD25-PE, CD27-allophycocyanin, CD28-allophycocyanin, CD45 receptor A (CD45RA)-biotin, CD62L-FITC, CTLA-4-PE-Cy5.5 (all from PharMingen), GITR-allophycocyanin (R&D Systems), CCR7 (unlabeled), and HLA-DR-biotin. Biotinylated antibodies were detected with streptavidin conjugated with FITC, PE, PE-Cy5.5, or allophycocyanin, depending on the experiment. CCR7 was detected by rat anti-mouse IgM-biotin mAb (PharMingen), followed by streptavidin-PE-Cy5.5. Nonreactive isotype-matched mAb (PharMingen) were used as controls. The phenotypic profile of murine CD4⫹ T cells was determined using anti-CD4-FITC mAb (PharMingen), anti-CD62L-FITC mAb (ImmunoTools GmbH, Friesoythe, Germany), and anti-CCR7-biotin mAb (eBioscience, San Diego, CA, USA), detected by streptavidin-allophycocyanin (PharMingen), and anti-GITR mAb (YGITR 765, unlabeled), followed by FITC-labeled anti-rat Igtot antibody (Harlan Sera-Lab, Loughborough, UK). Intracytoplasmic Foxp3 expression was assessed using allophycocyanin-labeled anti-Foxp3 mAb (eBioscience) following the manufacturer’s protocol. Expression of moGITRL by EL4 cells was analyzed using the YGL 386 mAb and detected with PE-labeled goat anti-rat Igtot antibody. Data were collected using a FACSCanto flow cytometer (Becton Dickinson, San Jose, CA, USA) and analyzed using FACSDiva and CellQuest software (Becton Dickinson). Cells were gated electronically according to light-scatter properties to exclude dead and contaminating cells.

Human proliferation assays Nonadherent cells or purified CD4⫹ or CD8⫹ T cells (using CD4 and CD8 microbeads, Miltenyi Biotec) were used as responder T cells (1⫻105), which

were cocultured with stimulator cells (DC or irradiated P815 cells, as specified) at the indicated ratios in round-bottom 96 wells. Cells were cocultured for 5 days, during the last 16 –18 h of which 1 ␮Ci/well 3H-methyl-thymidine (Amersham, Little Chalfont, UK) was added, and thymidine uptake was quantified by liquid scintillation counting (Microbeta, Wallac, Turku, Finland).

Murine in vitro proliferation assays For costimulation assays, 5 ⫻ 104 CD4⫹CD25– T cells were cocultured in triplicate in round-bottom 96 wells with 5 ⫻ 104 mitomycin C-treated spleen cells in the presence of 1 ␮g/ml anti-CD3ε (145-2C11). Titrated numbers of EL4-eGFP- or EL4-GITRL-expressing cells were added at indicated effector T cell (Teff):EL4 ratios. For Treg assays, purified Treg were added to the same conditions, as described above at a Treg:Teff ratio of 1:1. In both experiments, cells were pulsed overnight with 1 ␮Ci/well 3H-methyl-thymidine after 3 days of culture at 37°C, 5% CO2, and proliferation was then measured.

Induction of a naive CD4⫹ T cell response by human monocyte-derived DC Naive cord-blood PBMC were obtained by Ficoll density gradient centrifugation, and CD4⫹ T cells were isolated by immunomagnetic selection using CD4 microbeads (Miltenyi Biotec). CD4⫹ T cells were consistently ⬎90% pure and ⬎85% CD45RA-positive. Next, 5 ⫻ 104 naive CD4⫹ T cells were cocultured with tNGFR- or GITRL-electroporated DC at a ratio of one DC:five T cells in 200 ␮l, in round-bottom 96 wells, and in IMDM containing 1% AB serum, PS, and AAG. After 6 days, stimulated T cells were harvested, resuspended at a density of 1 ⫻ 106 T cells/ml in the presence of 4.7 ⫻ 104 CD3/CD28 T cell expander beads (Dynal, Invitrogen), and plated at 200 ␮l per well in roundbottom 96 wells. After incubation at 37°C for 24 h, the supernatant was harvested and assayed for IFN-␥ (Human IFN-␥ CytosetTM, BioSource International, Camarillo, CA, USA), IL-4 (Pierce Biotechnology, Aalst, Belgium), and IL-10 (R&D Systems) content by commercially available ELISA kits.

Induction of Melan-A-specific CD8⫹ T lymphocytes T cells and DC were obtained from HLA-A*0201⫹ healthy donors. DC were tNGFR- or GITRL-electroporated and afterward, pulsed with Melan-A-A2 peptide for 2 h. After washing, peptide-pulsed mRNA electroporated DC were cocultured with autologous CD8⫹ T cells at a DC:T ratio of 1:10 in six wells at a cell density of 2 ⫻ 106 cells/ml in IMDM containing 1% AB, PS, AAG, and 50 ␮M 2-ME. Two days later, 25 U/ml IL-2 and 10 ng/ml IL-7 were added. CD8⫹ T cells were restimulated weekly with the same stimulator DC as used in the primary stimulation. After three rounds of stimulation, CD8⫹ T cells were harvested, and their antigen specificity and function were determined.

Tetramer staining T cells were stained with 10 nM PE-labeled HLA-A2 tetramers containing the Melan-A (ELAGIGILTV) or MAGE-A3 (FLWGPRALV) peptides. Tetramers were prepared in-house. Subsequently, cells were stained with a FITC-labeled anti-CD8 mAb, and 1 ⫻ 105 cells were analyzed by flow cytometry.

IFN-␥ secretion assay The amount of IFN-␥ secreted by T cells after restimulation was assessed by coculturing 5 ⫻ 103 T cells with 2 ⫻ 104 stimulator cells, pulsed with the indicated peptides in 200 ␮l, in round-bottom 96 wells, and in the presence of 25 U/ml IL-2. After 20 h of coculture, the supernatant was tested for IFN-␥ content by ELISA using a commercially available kit (Endogen, Woburg, MA, USA).

Intracellular cytokine production The ability of Melan-A-primed CD8⫹ T cells to produce cytokines upon specific restimulation was investigated using intracellular staining for IFN-␥, IL-2, and TNF-␣ according to the manufacturer’s instructions (Becton Dickinson). T2 cells pulsed with the indicated peptides were cocultured with primed CD8⫹ T cells at a responder:stimulator (R:S) ratio of 10:1 for 2–3 h at 37°C. Golgi-Plug (brefeldin A, Becton Dickinson) was then added to block cytokine secretion, and cells were incubated further for 12 h at 37°C. CD8⫹ T

Tuyaerts et al. Human GITRL enhances DC stimulatory capacity

95

cells were then stained with PE-Cy5.5-conjugated anti-CD8, washed, permeabilized, and stained intracellularly with IL-2-FITC/IFN-␥-PE and TNF-␣FITC/IFN-␥-PE. Cytokine production by CD8⫹ T cells was measured by flow cytometry.

CD107a mobilization assay Primed CD8⫹ T cells (1⫻105) were restimulated with 4 ⫻ 104 Melan-A-A2 or gag-A2 peptide-loaded T2 cells in the presence of Golgi-Stop (monensin, Becton Dickinson) and PE-Cy5.5-labeled anti-CD107a mAb or an irrelevant isotype control. After 6 h incubation, cells were harvested and stained with FITC-labeled anti-CD8 mAb, 1 ⫻ 105 cells were analyzed by flow cytometry, and the percentage of CD8⫹CD107a⫹ T cells was calculated.

Statistical analysis Comparison of results obtained for Treg and responder T cells was analyzed using the Mann-Whitney U test. Results from antigen presentation by eGFPversus GITRL-electroporated cells to TIL were compared using the two-tailed, unpaired t-test. All analyses were performed using Prism software (GraphPad, San Diego, CA, USA). Differences were considered significant when P ⬍ 0.05.

RESULTS Endogenous expression of huGITRL As little is known about the endogenous expression of GITRL in the human setting, we performed an extensive screening of

different types of primary cells and cultured cell lines by flow cytometry and real-time PCR. We were unable to detect human GITRL expression on different PBMC subsets (CD3⫹, CD14⫹, CD19⫹, CD56⫹ cells), several melanoma cell lines (624-mel, 938-mel, 1087-mel), and monocyte-derived myeloid iDC or mDC (data not shown and Fig. 1A). Consistent with results published previously [10 –12, 36], huGITRL could be detected on the HUVEC line EA.Hy926. We also found huGITRL expression on the EBV-transformed B cell lines 888-EBV, 1087-EBV, and 1088-EBV (Fig. 1B).

Overexpression of huGITRL and moGITRL Based on murine data, we postulated that huGITRL expression on DC might be beneficial for the induction of T cell responses. As in contrast to published data on mouse DC [19], human myeloid DC do not endogenously express huGITRL in the immature state or upon maturation, we decided to investigate the impact of huGITRL overexpression on the function of these cells in the context of T cell activation. To obtain huGITRL expression on myeloid DC, huGITRL mRNA was electroporated into monocyte-derived DC. Figure 1C shows abundant expression of huGITRL after mRNA electroporation of K562 cells or monocyte-derived DC. huGITRL expression was strong and long-lasting, as we could detect

Fig. 1. huGITRL and moGITRL expression on different cell types. (A) Endogenous huGITRL expression was evaluated by flow cytometry on iDC and DC matured for 24 h with a cytokine mixture as described in Materials and Methods. CD25 and CD83 expression is shown as a control for maturation. Shaded histograms represent isotype control staining, and open histograms show huGITRL staining. Data are representative of three independent experiments. (B) Expression of huGITRL on the HUVEC line EA.Hy926, and the EBV-transformed B cell lines 888-EBV, 1087-EBV, and 1088-EBV were examined by flow cytometry. Shaded histograms represent isotype control staining, and open histograms show huGITRL staining. Data are representative of several independent experiments. (C) K562 cells and mDC were electroporated with eGFP or huGITRL mRNA, and transgene expression was assessed by flow cytometry 4 h and 24 h after mRNA electroporation. Shaded histograms represent isotype control staining, and open histograms show huGITRL staining. Data are representative of multiple independent experiments. (D) Murine EL4 cells were transduced lentivirally with eGFP or moGITRL and enriched further by FACS sorting. Persistence of eGFP or moGITRL expression was monitored by flow cytometry. Data are representative of multiple independent experiments.

96



huGITRL on the DC surface for up to 96 h (Fig. 1C and data not shown). Upon electroporation with huGITRL mRNA, monocyte-derived DC were slightly activated, as shown by a small upregulation of CD25, CD40, CD80, CD86, and CCR7. Furthermore, huGITRL-expressing monocyte-derived DC tended to secrete higher amounts of IFN-␣, IL-6, IL-10, IL-12p70, and TNF-␣, whereas no difference in IL-8 secretion was detected between tNGFR- or huGITRL-electroporated DC (data not shown). For experiments with moGITRL, EL4 cells were transduced lentivirally, resulting in sustained eGFP and moGITRL expression. Upon FACS sorting, more than 90% pure eGFP- and moGITRL-expressing cells were obtained (Fig. 1D).

huGITRL does not abrogate the inhibitory function of Treg on Teff In the murine system, it has been documented extensively that GITR is expressed preferentially by CD4⫹CD25⫹ Treg and that GITR ligation abrogates their suppressive function [13– 15, 17]. As little is known about GITR ligation and Treg function in humans, we decided to set up an experiment in which responder T cells were cultured with tNGFR- or huGITRL-electroporated DC in the absence and in the presence of Treg. As the frequency of Treg is higher in cancer patients compared with healthy individuals, we decided to isolate these cells from the blood of cancer patients. Human Treg and responder T cells were separated according to a protocol described by Ahmadzadeh and Rosenberg [43], and purity of both populations, as determined by CD4 and CD25 expression, is shown in Figure 2A. On average, CD4⫹CD25– responder T cells were 98 ⫾ 2.5% pure, and purity of CD4⫹CD25⫹ Treg was 74 ⫾ 23%. When comparing T cell phenotypes, Treg were found to express consistently higher amounts of CTLA-4, GITR, and HLA-DR. CCR7 and CD45RA were down-regulated on Treg, whereas no differences could be found regarding the expression of CD27, CD28, and CD62L between Treg and responder T cells (Fig. 2B). Differences were statistically significant for CTLA-4 (P⬍0.0001), GITR (P⫽0.004), CCR7 (P⫽0.003), and HLA-DR (P⫽0.008). When analyzing Foxp3 expression by quantitative RT-PCR, we show nearly undetectable levels in CD4⫹CD25– responder T cells, whereas CD4⫹CD25⫹ Treg expressed high amounts of this molecule (Fig. 2C). To examine the influence of GITRL overexpression on Treg inhibition, human CD4⫹CD25– responder T cells were cocultured with eGFP- or huGITRL-expressing DC at a ratio of one DC:30 T cells in the presence and absence of Treg at a 1:1 Treg:Teff ratio. We consistently obtained pronounced inhibition of responder T cell proliferation when Treg were added to the DC-T cell cocultures, but we did not observe significant abrogation of Treg activity in the presence of huGITRL-electroporated DC, as compared with eGFP-expressing DC (Fig. 2D). Even at lower Treg:Teff ratios, where suppression is supposed to be reduced, GITRL could not alleviate Treg suppression (data not shown). Furthermore, suppression exerted by four different antigen-specific Treg clones (kindly provided by Dr. Pierre van der Bruggen,

Ludwig Institute for Cancer Research, Brussels, Belgium) on responder CD4⫹ T cell proliferation was not reduced when irradiated K562 cells overexpressing huGITRL were added to the assay (data not shown). As these results contrast with findings described previously in mice, we wanted to see whether we could confirm these data in our system. Murine Treg were isolated from C57BL/6 mice. When looking at the phenotype of the cells after purification, higher expression of GITR and CCR7 and lower expression of CD62L were observed on CD4⫹CD25⫹ Treg compared with CD4⫹CD25– responder T cells. Purified CD4⫹CD25⫹ Treg were all positive for Foxp3, whereas in the CD4⫹CD25– fraction, only a minor percentage of Foxp3-positive cells was observed (Fig. 3A). Given that the bulk of published mouse data has been obtained using the agonistic anti-GITR antibody DTA-1 for GITR triggering, we first showed that we could successfully abolish mouse Treg-mediated suppression of responder T cell proliferation by adding anti-moGITR antibodies (DTA-1 and YGITR were equally effective; Fig. 3B). As a more direct comparison with the human experiments, we used moGITRL-overexpressing cells to analyze the effect of the natural ligand for GITR on Treg inhibition. Hence, mitomycin C-treated EL4 cells expressing moGITRL could reverse Treg suppression completely and even gave rise to a strong, additional costimulatory effect, at various R:S ratios (Fig. 3C). Thus, in contrast to its murine counterpart, huGITRL was not able to remove Treg suppression in our system.

huGITRL is costimulatory for polyclonal, naive T cells In the murine system, costimulation of responder T cells was described as a second major function of GITR triggering [18 –21, 33]. In an attempt to address this issue in the human setting, we first used a model, where no additional costimulatory signals are present [48]. For this purpose, P815 cells were electroporated with huGITRL or tNGFR, and their capacity to costimulate an anti-CD3-mediated T cell response was evaluated. huGITRL-expressing P815 cells showed an enhanced capacity to stimulate T cell proliferation (Fig. 4A), demonstrating that huGITRL can indeed provide costimulatory signals for CD4⫹ and CD8⫹ responder T cells. As DC already express a significant amount of other costimulatory molecules, we tested whether huGITRL could provide an additional costimulatory signal using DC as stimulator cells. Therefore, we cocultured immature huGITRL- or tNGFR-electroporated DC with allogeneic T cells at different DC:T cell ratios and measured T cell proliferation after 5 days. As shown in Figure 4B, we observed a higher T cell proliferative capacity after stimulation with huGITRL-expressing iDC. When using mDC as stimulators, the effect was less clear but still measurable (data not shown). We next compared tNGFR- and huGITRL-electroporated DC for their capacity to activate a naive CD4⫹ T cell response. Naive cord-blood CD4⫹CD45RA⫹ T cells were stimulated with allogeneic tNGFR- or huGITRL-electroporated DC, and after 6 days of coculture, T cells were


97

98



Fig. 3. Murine Treg isolation and suppression. (A) Murine CD4⫹CD25⫹ Treg and CD4⫹CD25– responder T cells were purified from splenocyte (SC) C57BL/6 mice as described. Purity and phenotype were assessed by flow cytometry. Data are representative of four independent experiments. (B) Murine CD4⫹CD25– responder T cells from C57BL/6 mice were cultured in the presence of 1 or 5 ␮g/ml anti-CD3 mAb and mitomycin C-treated spleen cells with addition of 5 ␮g/ml DTA-1, YGITR 765, or an isotype control (IC) antibody as indicated. When specified, Treg were added at a 1:1 Treg:Teff ratio. After 3 days, proliferation was measured by [3H]thymidine incorporation. Data shown are representative of four independent experiments. (C) Murine CD4⫹CD25– responder T cells from C57BL/6 mice were cultured in the presence of 1 ␮g/ml anti-CD3 mAb and mitomycin C-treated spleen cells with addition of eGFP- or moGITRL-expressing mitomycin C-treated EL4 cells at different ratios. Where indicated, Treg were added at a 1:1 Treg:Teff ratio. After 3 days, proliferation was measured by [3H]thymidine incorporation. Data shown are representative of four independent experiments.

restimulated overnight with anti-CD3/anti-CD28-coated beads, and the supernatant was assayed for IFN-␥, IL-4, and IL-10 production. No differences could be observed in IL-4 and IL-10 secretion induced by tNGFR- or huGITRL-electroporated DC, but huGITRL-electroporated DC consistently induced a higher IFN-␥ production by naive CD4⫹ T cells (Fig. 4C). These results indicate that huGITRL enhances the capacity of monocyte-derived DC to activate naı¨ve CD4⫹ T cells. Here, we also wanted to compare the data obtained for huGITRL with data from the mouse system. Therefore,

CD4⫹CD25– responder T cells from C57BL/6 mice were cocultured with eGFP- or moGITRL-expressing, mitomycin C-treated EL4 cells in the presence of 1 or 5 ␮g/ml anti-CD3 antibody and mitomycin C-treated spleen cells, and proliferation was measured after 3 days. A strong, costimulatory effect of moGITRL on responder T cells was observed for all responder:EL4 ratios (Fig. 5). Similar results were obtained when 3T3 fibroblasts, transduced retrovirally with moGITRL, were used for GITR stimulation (data not shown). In conclusion, huGITRL and moGITRL provide a costimulatory signal to T cells.

4 Fig. 2. Human Treg isolation and suppression. (A) Human CD4⫹CD25⫹ Treg and CD4⫹CD25– responder T cells were purified from nonadherent fraction (NAF) as described, and purity was assessed by flow cytometry. Data are representative of 12 independent experiments. (B) The phenotype of purified human CD4⫹CD25⫹ Treg and CD4⫹CD25– responder T cells was determined by flow cytometry. Each dot represents one individual experiment, and the median is indicated by a horizontal line. Closed symbols show CD4⫹CD25– T cells, whereas open symbols represent CD4⫹CD25⫹ Treg. Significant differences are indicated by asterisks (**, P⬍0.01; ***, P⬍0.0001). (C) Total RNA was extracted from human CD4⫹CD25– and CD4⫹CD25⫹ T cells, and cDNA was synthesized. Foxp3 expression was assessed by real-time quantitative PCR. Data are representative of two independent experiments. (D) Human CD4⫹CD25– responder T cells were cocultured with eGFP- or huGITRL-electroporated DC at a DC:T cell ratio of 1:30. Where indicated, Treg were added at a 1:1 Treg:Teff ratio. After 5 days, proliferation was measured by [3H]thymidine incorporation. Data shown are representative of five independent experiments.


99

Fig. 4. Costimulatory effect of huGITRL. (A) P815 cells were electroporated with tNGFR or huGITRL mRNA, irradiated (10,000 rad), and cocultured with human CD4⫹ and CD8⫹ T cells at 1:1 and 1:2 ratios, respectively, in the presence of 2 ␮g/ml OKT3. After 5 days, proliferation was measured by [3H]thymidine incorporation. Data shown are representative of three independent experiments. (B) iDC were tNGFR or huGITRL mRNA-electroporated and subsequently cultured at the indicated ratios with allogeneic T cells. After 5 days, proliferation was measured by [3H]thymidine incorporation. Data shown are representative of six independent experiments. (C) tNGFR- and huGITRL-electroporated DC were used to stimulate allogeneic CD4⫹ CD45RA⫹ T cells for 6 days. After restimulation with anti-CD3/anti-CD28-coated beads for 24 h, IFN-␥, IL-4, and IL-10 were measured by ELISA. Data shown are representative of three independent experiments.

huGITRL-mediated costimulation of antigenexperienced and -naive T cells upon antigenic stimulation In the previous experiments, we showed that huGITRL is costimulatory for T cells in an antigen-nonspecific setting. To determine whether huGITRL was also costimulatory in an antigen-specific model, T2 cells were electroporated with eGFP or huGITRL, pulsed with varying amounts of Melan-A/ MART-1 or gp100 HLA-A*0201-binding peptides, and subsequently, cocultured with Melan-A/MART-1 or gp100-specific TIL clones, respectively. Twenty-four hours after the start

Fig. 5. Costimulatory effect of moGITRL. eGFP- or moGITRL-expressing, mitomycin C-treated EL4 cells were cocultured with moCD4⫹CD25– responder T cells from C57BL/6 mice at different ratios in the presence of 1 ␮g/ml anti-CD3 mAb and mitomycin C-treated spleen cells. After 3 days, proliferation was measured by [3H]thymidine incorporation. Data shown are representative of four independent experiments.

100


of coculture, IFN-␥ secretion was measured. As shown in Figure 6A, huGITRL-expressing cells induced a higher IFN-␥ secretion consistently for Melan-A/MART-1 and gp100 T cell clones. We next explored whether huGITRL incorporation in DC would also have an effect on the priming of Melan-A/MART1-specific CD8⫹ T cells. To this end, DC from a HLA-A*0201positive donor were electroporated with huGITRL or tNGFR mRNA, pulsed with the immunodominant Melan-A/MART-1 peptide, and cocultured with autologous CD8⫹ T cells. After 3 weekly stimulations, tetramer-positive CD8⫹ T cells were measured, and as shown in Table 1, stimulations with huGITRLexpressing DC induced higher numbers of antigen-specific CD8⫹ T cells consistently. These Melan-A/MART-1-specific T cells proved to be fully functional, as assessed by IFN-␥ secretion (Table 1 and Fig. 6B), intracellular cytokine production (Fig. 6C), and CD107a mobilization as a marker for lytic activity (Fig. 6D). The primed CD8⫹ tetramer⫹ T cells appeared to be CD45RA– CD45RO⫹ CD27⫹ CD28⫹, together with a variable expression of CD62L and CCR7, which is indicative of a central memory phenotype (data not shown). We thus showed that huGITRL-expressing DC improved antigen-specific CD8⫹ T cell priming.

DISCUSSION The data presented here show that huGITRL and moGITRL can provide a costimulatory signal to T cells, leading to http://www.jleukbio.org

enhanced proliferation and cytokine production. This confirms mouse data published previously, where GITRL was shown to be costimulatory for CD4⫹ and CD8⫹ T cells [18 –21, 33]. It is surprising that in contrast to the murine system, where Treg inhibition is the best-described property of this molecule [13, 14, 16, 17], no alleviation of Treg inhibition could be found for huGITRL. Nevertheless, our results indicate that GITRL might potentiate DC-based immunotherapy approaches further. When we set out with our study, virtually no information was available about the endogenous expression of huGITRL, cloned originally from HUVEC [10, 12], which we also used as a source to clone the huGITRL cDNA. Only sporadic reports have emerged since then about the endogenous expression of huGITRL. One study shows protein expression in the retinal pigment epithelium and photoreceptors [49]. It was not until the recent systematic expression study by Hanabuchi et al. [36] that huGITRL was shown to be expressed exclusively in activated plasmacytoid DC, whereas no expression could be observed in other immunological cell types, including myeloid DC, T cells, and B cells. This contrasts with mouse studies, where moGITRL was shown to be expressed widely on B-1 B cells, myeloid DC, macrophages, plasmacytoid DC, and a whole range of other immunological cell types [16, 19]. Our data are consistent with previous studies, as we could not find huGITRL expression on monocyte-derived iDC or mDC. We also did not find expression on other cells present in PBMC such as T cells, B cells, NK cells, and monocytes. It is interesting that we found huGITRL expression on EBV-transformed B cells. In mice, GITR is preferentially expressed on CD4⫹CD25⫹ Treg and on activated T cells, but low levels can also be detected on B cells, macrophages, DC, and NK cells [14, 38]. Expression of GITR in the human system seemed to be more heterogeneous, and the highest expression was found in NK cells; moderate expression on memory T cells, Treg, and neutrophils; and low expression on resting T cells, B cells, monocytes, and myeloid and plasmacytoid DC [10 –12, 36, 50]. In the mouse system, a well-characterized, agonistic GITR antibody (DTA-1) has been described [14]. It was shown that addition of this antibody to cocultures of responder T cells and Treg could remove the inhibition of the Treg completely, which we could confirm in this study. We also showed the same effect for another anti-moGITR antibody (YGITR 765). Moreover, addition of recombinant moGITRL [17, 51] or cells overexpressing moGITRL on their surface [16, 51] in coculture with responder T cells and Treg yields similar results as the DTA-1 antibody in the mouse system, which we confirmed in this study. This inspired us to investigate whether the same was true for the human system. As no agonistic GITR antibody is known for huGITR, we decided to overexpress the natural ligand of huGITR on different cell types. After cloning huGITRL into an mRNA production vector, we were able to overexpress huGITRL on several different cell types using mRNA electroporation. Expression levels were high and durable (more than 72 h). We observed a modest activation of huGITRL-

electroporated DC, which might be a result of secondary mRNA structures in the huGITRL mRNA used for electroporation [52, 53]. In mice, GITRL was shown repeatedly to function as an inhibitor of Treg-mediated suppression of responder T cell proliferation. However, we could not reverse the inhibitory effect of polyclonal human Treg by huGITR ligation. Moreover, K562 cells overexpressing huGITRL could not block suppression of several Treg clones in a suppression assay. In contrast, we could abolish mouse Treg suppression completely by addition of anti-moGITR antibodies or moGITRLexpressing cells. Thus, huGITRL does not seem to share the property of Treg inhibition of its murine equivalent in our system. In support of these results, Baecher-Allan et al. [54] showed previously in a human setting that blocking huGITR by a mAb provided costimulation to responder T cells but only slightly reduced Treg inhibition. In addition, Levings et al. [55] report that addition of the same anti-huGITR mAb or recombinant huGITRL could not reverse Treg suppression, which was also shown by Valencia et al [56]. In contrast to these results, Cardona et al. [57] show that superantigenmediated inhibition of Treg activity could be restored partly by blocking with an anti-GITRL antibody. In a recent paper, Ito et al. [58] report the inhibition of Treg1 cells by human OX40L, which was not observed with huGITRL [58]. In addition to its function in removing Treg suppression, an important costimulatory activity has been demonstrated for moGITRL [18 –21, 24, 25, 33]. As a confirmation, we could show a strong costimulatory effect of EL4 cells overexpressing moGITRL on responder T cell proliferation, even in the presence of Treg. We used several different settings to investigate whether this costimulatory function can also be attributed to huGITRL. First, in analogy to the group of Watts et al. [48], we designed a system where no other costimulatory factors are present and where at the same time, strong TCR stimulation can occur. We thus used mouse P815 mastocytoma cells containing high levels of FcRs, rendering them capable of capturing and presenting anti-CD3 antibody (OKT-3) on their surface, thus providing the TCR activation signal. After electroporation with huGITRL or irrelevant mRNA, these cells were used in an in vitro proliferation assay. We consistently found increased proliferation for the huGITRL-expressing P815 cells, when using CD4⫹ T cells and CD8⫹ T cells. In the next step, we investigated whether the costimulatory effect of huGITRL was also measurable when other costimulatory molecules were present in the system. Indeed, we could show enhanced T cell proliferation induced by huGITRL-electroporated DC, compared with tNGFR-electroporated DC. Furthermore, we observed that huGITRL incorporation improved the capacity of monocyte-derived DC to activate naı¨ve CD4⫹ T cell responses. Another set of experiments was designed to examine if huGITRL also exerted its costimulatory function in an antigenspecific setting. We could indeed show that huGITRL-expressing stimulator cells induce greater activation of Melan-A/ MART-1 and gp100 T cell clones. Finally, we showed that DC overexpressing huGITRL can induce increased numbers of antigen-specific CD8⫹ T cells. In addition, more cytokine-


101

producing cells are generated, and a greater number of CD8⫹ T cells with a cytolytic capacity could be detected. These findings have also been observed for other TNFR family members. Dannull et al. [59] reported that introduction of OX40L

102


into DC-augmented, allogeneic and HLA Class II epitopespecific CD4⫹ T cell responses improved the stimulation of antigen-specific CTL in vitro and facilitated Th1 polarization of naive CD4⫹ T cell responses. Grunebach et al. [60] described


TABLE 1.

Screening of CD8⫹ T Cells Stimulated with Melan-A/MART-1 Peptide-Loaded DCa tNGFR

DC type Donor 1 Donor 2 Donor 3b Donor 4 Donor 5 Donor 6

mDC mDC mDC iDC mDCc iDCc iDCc iDCc mDCc

⫹

⫹

% CD8 tetramer

GITRL IFN-␥ (pg/ml)

9.2 1.4 0.5 1.8 7.2 0.8 0.5 0.1 0.4

105 98 12 82 66 31 0 ND ND

⫹

⫹

% CD8 tetramer 22 2.2 1.5 25 14 7.1 4.5 1.5 1

IFN-␥ (pg/ml) 652 112 69 1928 106 84 134 ND ND

HLA-A0201⫹ CD8⫹ T cells from four different healthy donors were stimulated three times weekly with Melan-A/MART-1 peptide-loaded DC, which were tNGFR- or GITRL-electroporated. After three consecutive stimulations, the percentage of Melan-A/MART-1-specific CD8⫹ T cells was determined by tetramer staining, and the amount of IFN-␥ secreted after antigen-specific restimulation was determined by ELISA. bEqual results have been obtained in the absence of exogenous IL-2 and IL-7. cStimulation was performed without the addition of exogenous IL-2 and IL-7. ND, not determined. a

that 4-1BBL-transfected DC showed an improved ability to elicit primary CTL responses. Morel et al. [61] showed that engagement of LIGHT and CD40L resulted in potentiation of allogeneic T cell proliferation and enhanced priming of CTL responses. In conclusion, we found that important differences exist between huGITRL and moGITRL both functionally and with regard to their expression pattern. One of the most important discrepancies we observed is that no reduction in Treg inhibition was found for huGITRL, whereas moGITRL alleviated Treg suppression clearly. huGITRL-mediated costimulation of NK cells was described previously [36], and we show here costimulation of CD4⫹ and CD8⫹ T cells by huGITRL in a naive and an antigen-experienced setting. Furthermore, huGITRL introduction enhances the capacity of myeloid DC to prime Melan-A/MART-1-specific CD8⫹ T cells. This discrepancy between huGITRL and moGITRL in its potential to alleviate Treg-mediated suppression in vitro might have important implications for further studies, which want to explore the use of GITR cross-linking to remove Treg inhibition in a clinical setting, e.g., in tumor vaccination trials, as GITR cross-linking in mice was shown to not only de-repress Teff in vitro but also in vivo. Thus, moGITRL might not be a representative model for its human equivalent as far as its role in the removal of Treg suppression is concerned. Nevertheless, also in mice, GITR triggering was shown recently to enhance vaccine-induced CD8⫹ T cell responses and even break tolerance, independent of its function on Treg [24 –26, 31, 38]. Altogether, from these recent mouse data, the concept emerges that although the main in vitro function of GITR is Treg

inhibition, the in vivo focus shifts toward its costimulatory effect. Thus, although in vitro GITR triggering in the human system does not remove Treg inhibition, it would still be useful to investigate whether GITR costimulation in vivo can also enhance CD8⫹ T cell responses as it does in the murine system, e.g., by vaccination with GITRL-expressing, in vitrogenerated, monocyte-derived DC.

ACKNOWLEDGMENTS This work was supported by grants to K. T. from the Fund for Scientific Research Flanders (FWO Vlaanderen), the ministry of Science (IUAP/PAI V), ‘De Belgische Federatie voor Kankerbestrijding’, and a Network of Excellence sponsored by the EU. J. L. A. was supported by a return grant from the federal government (BELSPO). S. T. was supported by a grant from the ‘Vlaamse Liga tegen Kanker’, and K. B. is a postdoctoral fellow of the FWO Vlaanderen. We thank Erna Borms, Christine Huysmans, and Hilde Lambrecht for their help with DC cultures, Elsy Vaeremans and Peggy Verbuyst for mRNA preparation, Roger Andries and Jos Theunissen for technical assistance, Maja Debulpaep for providing us with tetramers, Sabine Allard for help with intracellular cytokine stainings, Pierre van der Bruggen and Sabrina Ottaviani for their generous gift of Treg clones, and Mojgan Ahmadzadeh for help with the Treg purification and useful discussions.

4 Fig. 6. huGITRL-mediated costimulation of antigen-experienced and -naive T cells upon antigenic stimulation. (A) T2 cells were eGFP- or huGITRLelectroporated and pulsed with various concentrations of the Melan-A/MART-1 or gp100 peptide. Gag peptide was used as a negative control. Cells were then cocultured with Melan-A/MART-1 or gp100-specific TIL clones for 24 h, and IFN-␥ was measured by ELISA. Significant differences are indicated by asterisks (**, P⬍0.01; *, P⬍0.05). Data are representative of three experiments. (B) Melan-A/MART-1-primed CD8⫹ T cells were cocultured with T2 cells pulsed with gag or Melan-A/MART-1 HLA-A2-binding peptides. After 24 h, the supernatant was harvested, and IFN-␥ production was assayed by ELISA. Data are representative of four experiments. (C) Intracellular IFN-␥/IL-2 and IFN-␥/TNF-␣ production by Melan-A/MART-1-primed CD8⫹ T cells was measured by flow cytometry. T2 cells, pulsed with gag or Melan-A/MART-1 peptide, were used as stimulators. T cells were gated on forward-scatter (FSC)/side-scatter (SSC) characteristics and CD8 positivity. Data are representative of four experiments. (D) Cytolytic activity of Melan-A/MART-1-specific T cells was determined by a CD107a mobilization assay. Primed T cells were restimulated with T2 cells pulsed with the indicated peptides in the presence of anti-CD107-PE-Cy5 mAb and monensin. After 6 h, cells were harvested, stained with anti-CD8-FITC, and analyzed by flow cytometry. Cells were gated by FSC/SSC characteristics, and CD107a expression was measured within the CD8⫹ population. Data are representative of four independent experiments.


103

REFERENCES 1. Bretscher, P., Cohn, M. (1970) A theory of self-nonself discrimination. Science 169, 1042–1049. 2. Lafferty, K. J., Gill, R. G. (1993) The maintenance of self-tolerance. Immunol. Cell Biol. 71, 209 –214. 3. Croft, M. (2003) Co-stimulatory members of the TNFR family: keys to effective T-cell immunity? Nat. Rev. Immunol. 3, 609 – 620. 4. Grohmann, U., Orabona, C., Fallarino, F., Vacca, C., Calcinaro, F., Falorni, A., Candeloro, P., Belladonna, M. L., Bianchi, R., Fioretti, M. C., Puccetti, P. (2002) CTLA-4-Ig regulates tryptophan catabolism in vivo. Nat. Immunol. 3, 1097–1101. 5. O’Sullivan, B., Thomas, R. (2003) CD40 and dendritic cell function. Crit. Rev. Immunol. 23, 83–107. 6. Subudhi, S. K., Alegre, M. L., Fu, Y. X. (2005) The balance of immune responses: costimulation versus coinhibition. J. Mol. Med. 83, 193–202. 7. Suh, W. K., Gajewska, B. U., Okada, H., Gronski, M. A., Bertram, E. M., Dawicki, W., Duncan, G. S., Bukczynski, J., Plyte, S., Elia, A., et al. (2003) The B7 family member B7–H3 preferentially down-regulates T helper type 1-mediated immune responses. Nat. Immunol. 4, 899 –906. 8. Liang, L., Sha, W. C. (2002) The right place at the right time: novel B7 family members regulate effector T cell responses. Curr. Opin. Immunol. 14, 384 –390. 9. Watts, T. H. (2005) TNF/TNFR family members in costimulation of T cell responses. Annu. Rev. Immunol. 23, 23– 68. 10. Gurney, A. L., Marsters, S. A., Huang, R. M., Pitti, R. M., Mark, D. T., Baldwin, D. T., Gray, A. M., Dowd, A. D., Brush, A. D., Heldens, A. D., Schow, A. D., Goddard, A. D., Wood, W. I., Baker, K. P., Godowski, P. J., Ashkenazi, A. (1999) Identification of a new member of the tumor necrosis factor family and its receptor, a human ortholog of mouse GITR. Curr. Biol. 9, 215–218. 11. Kwon, B., Youn, B. S., Kwon, B. S. (1999) Functions of newly identified members of the tumor necrosis factor receptor/ligand superfamilies in lymphocytes. Curr. Opin. Immunol. 11, 340 –345. 12. Kwon, B., Yu, K. Y., Ni, J., Yu, G. L., Jang, I. K., Kim, Y. J., Xing, L., Liu, D., Wang, S. X., Kwon, B. S. (1999) Identification of a novel activationinducible protein of the tumor necrosis factor receptor superfamily and its ligand. J. Biol. Chem. 274, 6056 – 6061. 13. Nocentini, G., Riccardi, C. (2005) GITR: a multifaceted regulator of immunity belonging to the tumor necrosis factor receptor superfamily. Eur. J. Immunol. 35, 1016 –1022. 14. Shimizu, J., Yamazaki, S., Takahashi, T., Ishida, Y., Sakaguchi, S. (2002) Stimulation of CD25(⫹)CD4(⫹) regulatory T cells through GITR breaks immunological self-tolerance. Nat. Immunol. 3, 135–142. 15. McHugh, R. S., Whitters, M. J., Piccirillo, C. A., Young, D. A., Shevach, E. M., Collins, M., Byrne, M. C. (2002) CD4(⫹)CD25(⫹) immunoregulatory T cells: gene expression analysis reveals a functional role for the glucocorticoid-induced TNF receptor. Immunity 16, 311–323. 16. Stephens, G. L., McHugh, R. S., Whitters, M. J., Young, D. A., Luxenberg, D., Carreno, B. M., Collins, M., Shevach, E. M. (2004) Engagement of glucocorticoid-induced TNFR family-related receptor on effector T cells by its ligand mediates resistance to suppression by CD4⫹CD25⫹ T cells. J. Immunol. 173, 5008 –5020. 17. Ji, H. B., Liao, G., Faubion, W. A., Abadia-Molina, A. C., Cozzo, C., Laroux, F. S., Caton, A., Terhorst, C. (2004) Cutting edge: the natural ligand for glucocorticoid-induced TNF receptor-related protein abrogates regulatory T cell suppression. J. Immunol. 172, 5823–5827. 18. Esparza, E. M., Arch, R. H. (2005) Glucocorticoid-induced TNF receptor, a costimulatory receptor on naive and activated T cells, uses TNF receptor-associated factor 2 in a novel fashion as an inhibitor of NF-{␬}B activation. J. Immunol. 174, 7875–7882. 19. Tone, M., Tone, Y., Adams, E., Yates, S. F., Frewin, M. R., Cobbold, S. P., Waldmann, H. (2003) Mouse glucocorticoid-induced tumor necrosis factor receptor ligand is costimulatory for T cells. Proc. Natl. Acad. Sci. USA 100, 15059 –15064. 20. Esparza, E. M., Arch, R. H. (2005) Glucocorticoid-induced TNF receptor functions as a costimulatory receptor that promotes survival in early phases of T cell activation. J. Immunol. 174, 7869 –7874. 21. Kanamaru, F., Youngnak, P., Hashiguchi, M., Nishioka, T., Takahashi, T., Sakaguchi, S., Ishikawa, I., Azuma, M. (2004) Costimulation via glucocorticoid-induced TNF receptor in both conventional and CD25⫹ regulatory CD4⫹ T cells. J. Immunol. 172, 7306 –7314. 22. Kim, H. J., Kim, H. Y., Kim, B. K., Kim, S., Chung, D. H. (2006) Engagement of glucocorticoid-induced TNF receptor costimulates NKT cell activation in vitro and in vivo. J. Immunol. 176, 3507–3515. 23. Ko, K., Yamazaki, S., Nakamura, K., Nishioka, T., Hirota, K., Yamaguchi, T., Shimizu, J., Nomura, T., Chiba, T., Sakaguchi, S. (2005) Treatment of advanced tumors with agonistic anti-GITR mAb and its effects on tumor-

104


24. 25.

26.

27.

28.

29.

30.

31.

32.

33.

34. 35. 36.

37.

38.

39. 40.

41.

42.

infiltrating Foxp3⫹CD25⫹CD4⫹ regulatory T cells. J. Exp. Med. 202, 885– 891. La, S., Kim, E., Kwon, B. (2005) In vivo ligation of glucocorticoid-induced TNF receptor enhances the T-cell immunity to herpes simplex virus type 1. Exp. Mol. Med. 37, 193–198. Stone, G. W., Barzee, S., Snarsky, V., Kee, K., Spina, C. A., Yu, X. F., Kornbluth, R. S. (2006) Multimeric soluble CD40 ligand and GITR ligand as adjuvants for human immunodeficiency virus DNA vaccines. J. Virol. 80, 1762–1772. Kim, J., Choi, W. S., Kang, H., Kim, H. J., Suh, J. H., Sakaguchi, S., Kwon, B. (2006) Conversion of alloantigen-specific CD8⫹ T cell anergy to CD8⫹ T cell priming through in vivo ligation of glucocorticoid-induced TNF receptor. J. Immunol. 176, 5223–5231. Muriglan, S. J., Ramirez-Montagut, T., Alpdogan, O., Van Huystee, T. W., Eng, J. M., Hubbard, V. M., Kochman, A. A., Tjoe, K. H., Riccardi, C., Pandolfi, P. P., Sakaguchi, S., Houghton, A. N., Van Den Brink, M. R. (2004) GITR activation induces an opposite effect on alloreactive CD4(⫹) and CD8(⫹) T cells in graft-versus-host disease. J. Exp. Med. 200, 149 –157. Uraushihara, K., Kanai, T., Ko, K., Totsuka, T., Makita, S., Iiyama, R., Nakamura, T., Watanabe, M. (2003) Regulation of murine inflammatory bowel disease by CD25⫹ and CD25– CD4⫹ glucocorticoid-induced TNF receptor family-related gene⫹ regulatory T cells. J. Immunol. 171, 708 –716. Suri, A., Shimizu, J., Katz, J. D., Sakaguchi, S., Unanue, E. R., Kanagawa, O. (2004) Regulation of autoimmune diabetes by non-islet-specific T cells—a role for the glucocorticoid-induced TNF receptor. Eur. J. Immunol. 34, 447– 454. Kohm, A. P., Williams, J. S., Miller, S. D. (2004) Cutting edge: ligation of the glucocorticoid-induced TNF receptor enhances autoreactive CD4⫹ T cell activation and experimental autoimmune encephalomyelitis. J. Immunol. 172, 4686 – 4690. Cohen, A. D., Diab, A., Perales, M. A., Wolchok, J. D., Rizzuto, G., Merghoub, T., Huggins, D., Liu, C., Turk, M. J., Restifo, N. P., Sakaguchi, S., Houghton, A. N. (2006) Agonist anti-GITR antibody enhances vaccineinduced CD8(⫹) T-cell responses and tumor immunity. Cancer Res. 66, 4904 – 4912. Nocentini, G., Giunchi, L., Ronchetti, S., Krausz, L. T., Bartoli, A., Moraca, R., Migliorati, G., Riccardi, C. (1997) A new member of the tumor necrosis factor/nerve growth factor receptor family inhibits T cell receptorinduced apoptosis. Proc. Natl. Acad. Sci. USA 94, 6216 – 6221. Ronchetti, S., Zollo, O., Bruscoli, S., Agostini, M., Bianchini, R., Nocentini, G., Ayroldi, E., Riccardi, C. (2004) GITR, a member of the TNF receptor superfamily, is costimulatory to mouse T lymphocyte subpopulations. Eur. J. Immunol. 34, 613– 622. Ronchetti, S., Nocentini, G., Riccardi, C., Pandolfi, P. P. (2002) Role of GITR in activation response of T lymphocytes. Blood 100, 350 –352. Spinicelli, S., Nocentini, G., Ronchetti, S., Krausz, L. T., Bianchini, R., Riccardi, C. (2002) GITR interacts with the pro-apoptotic protein Siva and induces apoptosis. Cell Death Differ. 9, 1382–1384. Hanabuchi, S., Watanabe, N., Wang, Y. H., Ito, T., Shaw, J., Cao, W., Qin, F. X., Liu, Y. J. (2006) Human plasmacytoid predendritic cells activate NK cells through glucocorticoid-induced tumor necrosis factor receptorligand (GITRL). Blood 107, 3617–3623. Yu, K. Y., Kim, H. S., Song, S. Y., Min, S. S., Jeong, J. J., Youn, B. S. (2003) Identification of a ligand for glucocorticoid-induced tumor necrosis factor receptor constitutively expressed in dendritic cells. Biochem. Biophys. Res. Commun. 310, 433– 438. Ramirez-Montagut, T., Chow, A., Hirschhorn-Cymerman, D., Terwey, T. H., Kochman, A. A., Lu, S., Miles, R. C., Sakaguchi, S., Houghton, A. N., van den Brink, M. R. (2006) Glucocorticoid-induced TNF receptor family related gene activation overcomes tolerance/ignorance to melanoma differentiation antigens and enhances antitumor immunity. J. Immunol. 176, 6434 – 6442. Edgell, C. J., McDonald, C. C., Graham, J. B. (1983) Permanent cell line expressing human factor VIII-related antigen established by hybridization. Proc. Natl. Acad. Sci. USA 80, 3734 –3737. Tuyaerts, S., Michiels, A., Corthals, J., Bonehill, A., Heirman, C., de Greef, C., Noppe, S. M., Thielemans, K. (2003) Induction of influenza matrix protein 1 and MelanA-specific T lymphocytes in vitro using mRNA-electroporated dendritic cells. Cancer Gene Ther. 10, 696 –706. Dullaers, M., Van Meirvenne, S., Heirman, C., Straetman, L., Bonehill, A., Aerts, J. L., Thielemans, K., Breckpot, K. (2006) Induction of effective therapeutic antitumor immunity by direct in vivo administration of lentiviral vectors. Gene Ther. 13, 630 – 640. Beillard, E., Pallisgaard, N., van der Velden, V. H., Bi, W., Dee, R., van der Schoot, E., Delabesse, E., Macintyre, E., Gottardi, E., Saglio, G., Watzinger, F., Lion, T., van Dongen, J. J., Hokland, P., Gabert, J. (2003)


43. 44.

45.

46.

47.

48.

49.

50.

51.

Evaluation of candidate control genes for diagnosis and residual disease detection in leukemic patients using ‘real-time’ quantitative reversetranscriptase polymerase chain reaction (RQ-PCR)—a Europe against cancer program. Leukemia 17, 2474 –2486. Ahmadzadeh, M., Rosenberg, S. A. (2006) IL-2 administration increases CD4⫹ CD25(hi) Foxp3⫹ regulatory T cells in cancer patients. Blood 107, 2409 –2414. Tuyaerts, S., Noppe, S. M., Corthals, J., Breckpot, K., Heirman, C., De Greef, C., Van Riet, I., Thielemans, K. (2002) Generation of large numbers of dendritic cells in a closed system using cell factories. J. Immunol. Methods 264, 135–151. Michiels, A., Tuyaerts, S., Bonehill, A., Corthals, J., Breckpot, K., Heirman, C., Van Meirvenne, S., Dullaers, M., Allard, S., Brasseur, F., van der Bruggen, P., Thielemans, K. (2005) Electroporation of immature and mature dendritic cells: implications for dendritic cell-based vaccines. Gene Ther. 12, 772–782. Valmori, D., Fonteneau, J. F., Lizana, C. M., Gervois, N., Lienard, D., Rimoldi, D., Jongeneel, V., Jotereau, F., Cerottini, J. C., Romero, P. (1998) Enhanced generation of specific tumor-reactive CTL in vitro by selected Melan-A/MART-1 immunodominant peptide analogues. J. Immunol. 160, 1750 –1758. Parkhurst, M. R., Salgaller, M. L., Southwood, S., Robbins, P. F., Sette, A., Rosenberg, S. A., Kawakami, Y. (1996) Improved induction of melanomareactive CTL with peptides from the melanoma antigen gp100 modified at HLA-A*0201-binding residues. J. Immunol. 157, 2539 –2548. Wen, T., Bukczynski, J., Watts, T. H. (2002) 4-1BB ligand-mediated costimulation of human T cells induces CD4 and CD8 T cell expansion, cytokine production, and the development of cytolytic effector function. J. Immunol. 168, 4897– 4906. Kim, B. J., Li, Z., Fariss, R. N., Shen, D. F., Mahesh, S. P., Egwuagu, C., Yu, C. R., Nagineni, C. N., Chan, C. C., Nussenblatt, R. B. (2004) Constitutive and cytokine-induced GITR ligand expression on human retinal pigment epithelium and photoreceptors. Invest. Ophthalmol. Vis. Sci. 45, 3170 –3176. Li, Z., Mahesh, S. P., Kim, B. J., Buggage, R. R., Nussenblatt, R. B. (2003) Expression of glucocorticoid induced TNF receptor family related protein (GITR) on peripheral T cells from normal human donors and patients with non-infectious uveitis. J. Autoimmun. 21, 83–92. Calmels, B., Paul, S., Futin, N., Ledoux, C., Stoeckel, F., Acres, B. (2005) Bypassing tumor-associated immune suppression with recombinant ade-

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

novirus constructs expressing membrane bound or secreted GITR-L. Cancer Gene Ther. 12, 198 –205. Kariko, K., Ni, H., Capodici, J., Lamphier, M., Weissman, D. (2004) mRNA is an endogenous ligand for Toll-like receptor 3. J. Biol. Chem. 279, 12542–12550. Ceppi, M., Ruggli, N., Tache, V., Gerber, H., McCullough, K. C., Summerfield, A. (2005) Double-stranded secondary structures on mRNA induce type I interferon (IFN ␣/␤) production and maturation of mRNAtransfected monocyte-derived dendritic cells. J. Gene Med. 7, 452– 465. Baecher-Allan, C., Viglietta, V., Hafler, D. A. (2002) Inhibition of human CD4(⫹)CD25(⫹high) regulatory T cell function. J. Immunol. 169, 6210 – 6217. Levings, M. K., Sangregorio, R., Sartirana, C., Moschin, A. L., Battaglia, M., Orban, P. C., Roncarolo, M. G. (2002) Human CD25⫹CD4⫹ T suppressor cell clones produce transforming growth factor ␤, but not interleukin 10, and are distinct from type 1 T regulatory cells. J. Exp. Med. 196, 1335–1346. Valencia, X., Stephens, G., Goldbach-Mansky, R., Wilson, M., Shevach, E. M., Lipsky, P. E. (2006) TNF downmodulates the function of human CD4⫹CD25hi T-regulatory cells. Blood 108, 253–261. Cardona, I. D., Goleva, E., Ou, L. S., Leung, D. Y. (2006) Staphylococcal enterotoxin B inhibits regulatory T cells by inducing glucocorticoidinduced TNF receptor-related protein ligand on monocytes. J. Allergy Clin. Immunol. 117, 688 – 695. Ito, T., Wang, Y. H., Duramad, O., Hanabuchi, S., Perng, O. A., Gilliet, M., Qin, F. X., Liu, Y. J. (2006) OX40 ligand shuts down IL-10-producing regulatory T cells. Proc. Natl. Acad. Sci. USA 103, 13138 –13143. Dannull, J., Nair, S., Su, Z., Boczkowski, D., DeBeck, C., Yang, B., Gilboa, E., Vieweg, J. (2005) Enhancing the immunostimulatory function of dendritic cells by transfection with mRNA encoding OX40 ligand. Blood 105, 3206 –3213. Grunebach, F., Kayser, K., Weck, M. M., Muller, M. R., Appel, S., Brossart, P. (2005) Cotransfection of dendritic cells with RNA coding for HER-2/neu and 4-1BBL increases the induction of tumor antigen specific cytotoxic T lymphocytes. Cancer Gene Ther. 12, 749 –756. Morel, Y., Truneh, A., Sweet, R. W., Olive, D., Costello, R. T. (2001) The TNF superfamily members LIGHT and CD154 (CD40 ligand) costimulate induction of dendritic cell maturation and elicit specific CTL activity. J. Immunol. 167, 2479 –2486.


105