Cancer Immunol Immunother (2007) 56:1565–1575 DOI 10.1007/s00262-007-0300-9
O RI G I NAL ART I C LE
A HLA-DQ5 restricted Melan-A/MART-1 epitope presented by melanoma tumor cells to CD4+ T lymphocytes Pierre Larrieu · Laure-Hélène Ouisse · Yannick Guilloux · Francine Jotereau · Jean-François Fonteneau
Received: 5 October 2006 / Accepted: 8 February 2007 / Published online: 23 February 2007 © Springer-Verlag 2007
Abstract Melan-A/MART1 is a melanocytic diVerentiation antigen expressed by tumor cells of the majority of melanoma patients and, as such, is considered as a good target for melanoma immunotherapy. Nonetheless, the number of class I and II restricted Melan-A epitopes identiWed so far remains limited. Here we describe a new Melan-A/MART-1 epitope recognized in the context of HLA-DQa1*0101 and HLADQb1*0501, -DQb1*0502 or -DQb1*0504 molecules by a CD4+ T cell clone. This clone was obtained by in vitro stimulation of PBMC from a healthy donor by the Melan-A51–73 peptide previously reported to contain a HLA-DR4 epitope. The Melan-A51–73 peptide, therefore contains both HLA-DR4 and HLA-DQ5 restricted epitope. We further show that Melan-A51–63 is the minimal peptide optimally recognized by the HLA-DQ5 restricted CD4+ clone. Importantly, this clone speciWcally recognizes and kills tumor cell lines expressing Melan-A and either HLA-DQb1*0501, -DQb1*0504 or -DQb1*0502 molecules. Moreover, we could detect CD4+ T cells secreting IFN- in response to Melan-A51–63 and Melan-A51–73 peptides among tumor inWltrating and blood lymphocytes from HLADQ5+ patients. This suggests that spontaneous CD4+ T cell responses against this HLA-DQ5 epitope occur in vivo. Together these data signiWcantly increase the
P. Larrieu · L.-H. Ouisse · Y. Guilloux · F. Jotereau · J.-F. Fonteneau (&) INSERM U601, Institut de biologie, 9 quai moncousu, 44093 Nantes cedex, France e-mail:
[email protected] Y. Guilloux · F. Jotereau Université de Nantes, 44322 Nantes, France
fraction of melanoma patients susceptible to beneWt from a Melan-A class II restricted vaccine approach. Keywords CD4+ T cells · Cancer · Melanoma · Melan-A/MART1 · HLA-DQ5 Abbreviation list B-LCL B Lymphocyte cell line pHS Pooled human serum TIL Tumor inWltrating lymphocytes
Introduction CD8+ T cells are critical components in immune responses to infections and tumors. They diVerentiate into cytotoxic T lymphocytes (CTLs) and acquire the capacity to lyse virus infected or tumor antigen expressing cells. However, CD8+ T cell diVerentiation or responses often rely on CD4+ T cell help. Indeed, tumor regression in the mouse often depends on both helper T cells and CTL functions [26, 32]. Some studies further suggested that CD4 T cells might contribute to the induction and development of CTL response through various mechanisms, such as by increasing the CTL priming capacity of DC, through CD40/CD40-L interaction [4, 27, 28] and by producing soluble factors such as IL-2 [10, 23]. It is assumed that CD4+ T cells may also contribute directly to anti-tumor responses by killing antigen expressing tumor cells [1, 7, 25, 35]. Therefore, identiWcation of tumor antigens and tumor epitopes recognized by CD4+ T cells is probably an important step towards the design of anti-tumoral immunotherapy. Among melanoma associated antigens recognized by CD8+ and CD4+ T cells, Melan-A/MART-1 is one
123
1566
of the most consistently expressed by both a high fraction of melanoma tumor cells and melanoma tumor samples [8, 20]. Melan-A epitopes recognized by CD8+ T cells and CD4¡ T cells have been described [3, 5, 8, 12, 16, 35]. Furthermore, we and others found that it is the most frequently recognized antigen by CD8+ TIL in the HLA-A2 context [3, 15, 21, 30]. Since HLA-A2 is the most frequently expressed HLA class I allele within the Caucasian population, vaccines and adoptive therapy trials targeting Melan-A speciWc CTL could be developed for a signiWcant fraction of patients. IdentiWcation of Melan-A epitopes recognized by CD4+ T cells would probably allow the design of immunotherapeutic treatments to simultaneously develop CD8+ and CD4+ T cell responses against Melan-A in these patients. It is therefore critical to identify others Melan-A HLA class II epitopes for the design of anti-tumoral vaccine using this antigen. In this study, we identiWed a new Melan-A epitope (Melan-A51–73) recognized by a CD4+ T cell clone in the HLA-DQa1*0101/HLA-DQb1*0501 context. This epitope is located in the Melan-A51–73 regions already shown to contain an HLA-DR4 restricted epitope [35]. In addition, we show that HLA-DQ5+ melanoma cell lines expressing Melan-A naturally present this epitope and that speciWc CD4+ T cell responses can be detected directly or after a single in vitro peptide stimulation among tumor inWltrating and blood lymphocytes of HLA-DQ5+ melanoma patients. These data increase the fraction of melanoma patients in which both CD8+ and CD4+ Melan-A speciWc responses might be simulated.
Materials and methods Culture medium Culture medium RPMI1640 (Gibco BLR, Gaithersburg, MD) was supplemented with penicillin-streptomycin (10 g/ml) and L-Glutamine (2 mM) (Life Technologies, Cergy-Pontoise, France) and either with 1% human plasma, 8% pooled human serum (pHS) or 10% fetal calf serum (FCS, Eurobio, Les Ulis, France). Cell lines B-lymphocyte cell lines (BLCLs): HOM2, GRE, Adam, Ward, don12, Henr, Lepe, DEM, Bign, Lebr, Boleth and Hour, were generated by culturing CD14 negative cells from PBMCs with a supernatant of the EBV-producing B95.8 marmoset cell line in RPMI 1640 containing 20% FCS and 1 mg/ml cyclosporine
123
Cancer Immunol Immunother (2007) 56:1565–1575
A. Melanoma cell lines M... were established in our laboratory. Melanoma cell lines FM25 and FM29 are a kind gift from Pr. J. Zeuthen. All these cell lines were maintained with RPMI 1640 containing 10% FCS. Synthetic peptides Melan-A/MART151–73 (RNGYRALMDKSLHVGTQCALTRR) and NYESO-1157–170 (SLLMWITQCFLPVF) peptides were purchased from Epytop (Nimes, Frances). Melan-A/MART152–73, 53–73, 51–72, 51–71, 51–70, 51–69, 51–68, 51–67, 51–66, 51–65, 51–64, 51–63, 51–62, 51–61, 51–60, 51–59 and 51–58 were purchased from Genepep (Montpellier, France). All peptides were at least 80% pure. Lyophilized peptides were diluted in DMSO and stored at ¡80°C. Induction of Melan-A/MART1 speciWc CD4+ T cells Induction of Melan-A/MART1 speciWc CD4+ T cells was performed as described [11] with minor modiWcations. Blood from HLA-DRb1*0401+ or HLADPb1*0401+ healthy donors was obtained from the “Nantes Etablissement Français du Sang”. Peripheral blood mononuclear cells (PBMC) were puriWed by Ficoll separation. Monocytes were separated of lymphocytes by elutriation (Aventi J-20, Beckman Coulter). CD25 negative lymphocytes were puriWed from Lymphocytes-enriched fraction by negative depletion using CD25 magnetic microbeads (miltenyi biotec, Paris, France) according to the manufacturer’s instructions, and were then frozen. DCs were generated by culturing the monocyte-enriched fraction in six well plates at 3 £ 106 cells/well in 3 ml of RPMI containing 1% human plasma, 100 IU/ml GM-CSF (abcys) and 300 U/ml IL-4 (abcys). The cytokines were added to the culture at day 0 and 3. On day 5, 100 g/ml PolyI/C (Sigma) and 10 ng/ml TNF- (abcys) were added for 48 h to induce DC maturation. Mature DCs were collected on day 7. A fraction of them was pulsed with peptide and used to stimulate CD25- lymphocytes. The remaining DCs were cryopreserved for future T cell restimulation. Mature DCs were incubated with 10 M of Melan-A51–73 or NYESO-1157–170 peptide at 37°C for 2 h and then washed twice. 2 £ 105 peptide pulsed DCs were cultured with 2 £ 106 CD25- lymphocytes in 2 ml of RPMI 1640 containing 8% pHS, 1,000 U/ml IL-6 (R&D systems) and 5 ng/ml IL-12 (R&D systems) in 10 wells of 24 well culture plates. T cell cultures were restimulated weekly with 2 £ 105 peptide pulsed DCs in presence of 10 U/ml IL-2 (Proleukin, Chiron Corp.)
Cancer Immunol Immunother (2007) 56:1565–1575
and 5 ng/ml IL-7 (R&D systems). Cultures were maintained at less than 1.5 £ 106 T cells/ml. Six days after the third stimulations, an aliquot of each T cell culture was used to evaluate the percentage of peptide speciWc T cells by IFN- intracytoplasmic staining. Cloning of T cells Cells from polyclonal cultures containing speciWc T cells were cloned by limiting dilution as previously described [11]. BrieXy, T cells were plated in U-bottom 96 well plates with irradiated feeder cells, at concentrations of 10, 1 or 0.3 T cells/well. Irradiated (35 gray) feeder cells consisted of 1 £ 105 allogenic PBMCs and 2 £ 104 BLCL cells/well. Stimulatory medium consisted of RPMI1640 containing 5% PHS, 150 U/ml IL-2 and 1 g/ml Phytohemagglutinin-L (PHA-L, Sigma). After 5 days, half of the volume of medium (75 l/well) was replaced with culture medium containing 5% PHS and 150 U/ml IL-2 without PHA-L. After 2 weeks, aliquots of each clone were tested for peptide speciWcity by a TNF production assay. SpeciWc clones were maintained in culture by periodic restimulation. For all experiments, clones were used at least 14 days after the last stimulation. IFN- intracellular staining Hundred microliters aliquots of polyclonal T cells cultures or 1 £ 105 T cell clones were cultured with or without peptide (Melan-A51–73 or NYESO-1157–170), with 1 £ 10 5 unpulsed or peptide (10 M) pulsed B-LCL or with 1 £ 105 unpulsed, peptide (10 M) pulsed or soluble Melan-A/MART-1 protein (50 g/ml) pulsed DC in the presence of 10 g/ml of Brefeldin A (Sigma) for 6 h at 37°C. In some experiments, 10 g/ml of mAb against HLA-DP (clone B.7.21, produced in our laboratory), HLA-DQ (clone SK-10, BD) or HLA-DR (clone L243, BD) were added to the culture. The Melan-A/MART-1 protein, kindly provided by Dr. G. Ritter, was a puriWed, laboratory grade, Escherichia coli-derived recombinant fulllength Melan-A protein (His-tagged) prepared at the Ludwig Institute for Cancer Research Institute/Cornell University Bioprocess Development Research Laboratory (Ithaca, NY, USA). Finally, cells were Wxed for 10 min at room temperature with PBS containing 4% paraformaldehyde (Sigma). In some experiments, cells were stained with FITC conjugated mAb speciWc for CD4 molecule before Wxation. IFN- intracellular staining of Wxed cells was performed as described by Jung et al. [14]. BrieXy, cells were stained with 5 g/ml of PE conjugated mAb speciWc for IFN-
1567
(clone 4S.B3, BD Pharmingen) and then washed. Antibody dilutions and washes were made with PBS containing 0.1% BSA and 0.1% saponin at room temperature. After staining, cells were resuspended in PBS and analyzed on a FACScan (Beckton&Dickinson, Franklin Lake, NJ) with a gate set on T cells on the FSC-SSC dot plot. TNF production assay 1 £ 104 T cell clones were cultured with or without antigenic peptides, with 3 £ 104 untreated or IFN- treated (48 h, 500 IU/ml) melanoma cell lines, or with 3 £ 104 unpulsed or antigenic peptide pulsed transfected U293 human embryonic kidney cells for 6 h at 37°C. Then culture supernatants were collected and TNF released by T cells was measured with a biological assay using cytotoxicity against WEHI 164 clone 13 [13]. TNF production was considered positive when more than 30 pg/ml of TNF was released. Cloning of cDNA encoding HLA-DQb1*0501, HLADQa1*0101 or HLA-DQa1*0303 in pcDNA3 plasmids cDNA clones encoding the HLA-DQb1*0501, -DQa1*0101 and -DQa1*0303 chains were obtained as follows. RNA prepared from PBMCs used to obtain Mel18 CD4+ T cell clone was converted to cDNA with Moloney murine leukemia virus reverse transcriptase (Boehringer Mannheim) using an oligo-dT primer. For PCR HLA-DQa ampliWcation, PCR assays were performed with primers 5⬘-CGGAATTCGCCGCCATG ATCCTAAACAAAGCTCTG-3⬘ and 5⬘-CTAGTCT AGATCACAA TGG CCC TTG GTGT-3⬘ for 35 cycles (1 min at 94°C, 1 min at 58°C, 1 min at 72°C). For PCR HLA-DQb ampliWcation, PCR assays were performed with primers 5⬘-CGGAATTCGCCGCC ATGTCTTGGAAGAAGTCCTTTG-3⬘ and 5⬘-CTA GTCTAGATCAGTGCAGAAGCCCTTTC-3⬘ for 35 cycles (1 min at 94°C, 1 mn at 56°C, 1 min at 72°C). The PCR product was puriWed using the QIAquick PCR puriWcation kit (Qiagen, Hilden, Germany), digested with EcoRI and Xba I and ligated into expression vector pcDNA3 (Invitrogen). Constructs were transfected by electroporation into E. coli DH5, and plasmid DNA extracted from several independent colonies was sequenced. Transfection of U293 human embryonic Kidney cells 3.5 £ 104/well U 293 human embryonic Kidney cells were plated in Xat-bottomed 96 wells plates 24 h before transfection. Cells were washed with optimem
123
1568
medium (Invitrogen) and were cotransfected with 1 l of plus reagent (Invitrogen), 0.5 l of Lipofectamine reagent (Invitrogen), 12 ng of plasmid pcDNA3 containing the HLA-DRb1*0501 cDNA, 12 ng of plasmid pcDNA3 containing the HLA-DRa1*0101 cDNA or HLA-DRa1*0303, 24 ng of plasmid pcDNA3 containing an Ii cDNA, and 12 ng of plasmid EBO-76pl containing a CIITA cDNA in 70 l of optimem. Each transfection conditions were performed in duplicate. After 24 hrs, cells were washed and used for experiments. 51
Cr release assay
Target cells (melanoma cell lines) were pulsed for 1 h with Na51 2 CrO4 (NEN life science, Paris, France) and then washed. 104 target cells/well were mixed with eVector cells (Mel18 CD4+ T cell clone) at diVerent eVector to target ratio of 5/1, 10/1 and 20/1. After 4 h incubation at 37°C, 25 l of supernatant were harvested, and added to 100 l scintillation cocktail (optiphase supermix, Wallac, UK) before liquid scintillation counting. The percent of speciWc lysis was calculated as follows: (sample release—spontaneous release/maximum release—spontaneous release) £ 100. The spontaneous release was calculated from targets incubated with culture medium, and the maximum release from targets incubated with 1% Triton X-100. Detection of Melan-A51–63 and Melan-A51–73 CD4+ T cell responses among TIL and PBMC of melanoma patients Blood from patient M253 was obtained from Department of Dermatology of Hotel Dieu of Nantes upon patient’s informed consent. PBMCs were puriWed by Ficoll separation. Then, CD4+ PBMCs and CD4PBMCs were puriWed using CD4+ magnetic microbeads (miltenyi biotec) according to the manufacturer’s instructions. 105 CD4+ PBMCs were cultured with 105 CD4- irradiated PBMCs (35 gray) and 10 M of Melan-A51–63 or Melan-A51–73 in 200 l of RPMI medium containing 8% pHS, 10 ng/ml IL-7 and 10 IU/ml IL-2, in U-bottom 96 well plates. After 2 weeks, CD4+ T cell responses to Melan-A51–63 or Melan-A51–73 were assessed by IFN- intracellular and CD4 surface staining. TIL from melanoma patients were obtained as previously described [2]. TIL were cultured with or without 10 M Melan-A51–63 peptide in presence of brefeldin-A. After 6 h, cells were Wxed and IFN- production by CD4+ T cells was assessed by IFN- intracytoplasmic and CD4 surface staining.
123
Cancer Immunol Immunother (2007) 56:1565–1575
Results Induction of Melan-A51–73 or NYESO-1157–170 speciWc T cells To generate CD4+ T cells recognizing the Melan-A51–73 peptide on HLA-DRb1*0401 [35] and the NYESO1157–170 peptide on HLA-DPb1*0401 [36], we used sorted CD25 negative PBMC from an HLADRb1*0401+ and a HLA-DPb1*0401+ healthy donor, respectively. These CD25 negative PBMC were stimulated with peptide pulsed autologous mature DCs as previously described [11]. CD25+ T cell depletion was performed since these cells have been reported to limit the in vitro induction of antigen speciWc CD4+ T cells [9]. Six days after the third stimulation, each culture wells was tested for the presence of peptide speciWc T cells by intracellular IFN- secretion, in response to the stimulating peptide in an “autopresentation” assay. In both stimulated cultures, we found peptide speciWc T cells populations containing 3.34% of Melan-A51–73 and 3.23% of NYESO-1157–170 reactive T cells (Fig. 1a). To determine HLA class II restriction of these responses, we used HLA-DRb1*0401 and HLADPb1*0401 B lymphocyte cell lines (B-LCL) as peptide presenting cells. As expected NYESO-1157–170 stimulated T cells responded to peptide pulsed HLADPb1*0401 B-LCL. In contrast, Melan-A51–73 stimulated T cells did not respond to the peptide pulsed HLA-DRb1*0401 B-LCL, suggesting that the response observed in the autopresentation assay was not restricted by the HLA-DRb1*0401 molecule, but by another HLA class II allele from the donor (Fig. 1b). HLA-DQa1*0101/HLA-DQb1*0501 molecules present Melan-A51–73 peptide to speciWc CD4+ T cells To identify the HLA class II restriction of Melan-A51–73 reactive T cells, we cloned under limiting dilution the positive population as previously described [11]. We obtained 139 clones from 600 wells at a seeding dilution of 0.5% cells/well. 22 clones produced TNF in response to the Melan-A51–73 peptide. Only 9 out of these 22 clones could be expanded. The clone Mel18 was further expanded to characterize its HLA class II restriction. We also derived a NYESO-1157–170 speciWc T cell clone NY67. We conWrmed that these two clones were CD4+ CD8¡ (data not shown), and by autopresentation and IFN- intracytoplasmic staining, that both had the expected peptide speciWcity (Fig. 2a).
Cancer Immunol Immunother (2007) 56:1565–1575
Fig. 1 Polyclonal T cells lines speciWc for Melan-A51–73 or NYESO-1157–170: DCs and T cells were prepared from PBMCs of HLA-DRb1*0401+ or HLA-DPb1*0401+ healthy donors. Melan-A51–73 pulsed HLA-DRb1*0401+ or NYESO-1157–170 pulsed HLA-DPb1*0401+ DC were plated with CD25- autologous T cells in presence of IL-6 and -12. T cells were then restimulated weekly with peptide pulsed DCs in presence of IL-2 and -7. a Six days after the third stimulation an aliquot of each T cell cultures was cultured with or without the antigenic peptide in presence of Brefeldin A at 37°C during 6 h. Cells were then Wxed, permeabilized and stained for intracytoplasmic IFN-. Fluorescence was analyzed by Xow cytometry with a gate set on T cells. Histograms from the two positive cultures are shown. b Six days after the third stimulations an aliquot of T cell cultures stimulated by the Melan-A peptide was cultured with unpulsed or MelanA51–73 pulsed HLA-DRb1*0401+ B-LCL, and an aliquot of T cell cultures stimulated by the NYESO-1 peptide was cultured with unpulsed or NYESO-1157–170 pulsed HLA-DPb1*0401+ B-LCL in presence of Brefeldin A at 37°C during 6 h. Cells were then Wxed, permeabilized and stained for intracytoplasmic IFN-. Fluorescence was analyzed by Xow cytometry with a gate set on T cells. Histograms from the two positive cultures are shown
In this assay, we then added inhibitory anti-HLADP, anti-HLA-DQ or anti-HLA-DR mAb, to identify the isotype restriction of Mel18. Only the anti-HLADQ mAb inhibited response of Mel18 (Fig. 2b), whereas, as expected, only the anti-HLA-DP mAb inhibited response of NY67. Therefore, the Mel18
1569
clone recognize Melan-A51–73 in the context of one of the donor HLA-DQ alleles (DQb1*0301, DQb1*0501). To identify which HLA-DQb molecule presents the Melan-A51–73 peptide to clone Mel18, we compared peptide presentation by B-LCL expressing the HLADQb1*0501 but not the HLA-DQb1*0301 and vice versa. The HLA-DQb1*0501 expressing B-LCL (HOM2, GRE, ADAM, Ward, Don12 and Henr), but not the HLA-DQb1*0301 expressing B-LCL, did present the peptide to clone Mel18 (Fig. 2c). Furthermore, the B-LCL HOM2, homozygous for HLA-DQb1*0501 presented the peptide more eYciently than the other HLA-DQb1*0501 B-LCL, which express only one copy of this allele. We also observed in this assay that HLA-DQb1*0502 expressing B-LCL (Lepe, Bign and DEM) were able to present Melan-A51–73 peptide to clone Mel18, although with a lower eYciency than HLA-DQb1*0501 expressing B-LCL. As both a and b chains of HLA-DQ molecules participate in the polymorphism of peptide binding, we wanted to determine whether both HLA-DQa chains of the donor (-DQa1*0101 and -DQa1*0303), or only one contributed to the presentation of Melan-A51–73 peptide to clone Mel18. We cloned from the donor PBMC, the cDNA encoding HLA-DQb1*0501, -DQa1*0101 and -DQa1*0303 in pcDNA3 plasmids. We transfected human embryonic kidney cells U937 with plasmids encoding the transactivator CIITA, the invariant chain Ii, HLA-DQb1*0501 and either the HLA-DQa1*0101 or the HLA-DQa1*0303 molecule. Transfected cells were then pulsed or not with 10 M of Melan-A51–73 peptide and tested for their capacity to induce TNF secretion by the clone Mel18. Only peptide pulsed U937 cells transfected with HLADQa1*0101 molecules, but not with HLA-DQa1*0303, were able to stimulate TNF production from the clone Mel18 (Fig. 2d). This result demonstrates that the Melan-A51–73 peptide is presented by HLADQa1*0101/HLA-DQb1*0501 molecules, a result in accordance with the fact that there is a preferential linkage between this two alleles. Mapping of the minimal Melan-A peptide presented by HLA-DQa1*0101/HLA-DQb1*0501 dimer To determine the minimal peptide sequence able to stimulate the Mel18 clone, we tested a panel of overlapping synthetic peptides located in the Melan-A51–73 regions for their capacity to induce Mel18 TNF secretion in an autopresentation assay. Removal of the N-terminal arginine 51 abrogated the Mel18 response (Fig. 3). In contrast, removal of C-terminal residues until histidine 63 did not aVect signiWcantly the Mel18
123
1570
Cancer Immunol Immunother (2007) 56:1565–1575
B
% IFN-g positive cells
A
100
no peptide peptide peptide + anti HLA-DP peptide + anti HLA-DQ peptide + anti HLA-DR
80
60
40
20
0
NY67
D
600
100 80 60 40 20 0 0
0,01
0,03
0,1
0,3
1
3
HOM2 GRE Adam Ward Don12 Henr
HLA-DQb1*0501+
Lepe DEM Bign
HLA-DQb1*0502+
Lebr Boleth Hour
HLA-DQb1*0501/02-
10
TNF (pg/ml)
500
C % IFN-γ positive cells
MEL18
HLA-DQa1*0101 HLA-DQa1*0303
400 300 200 100 0
- peptide
+10µM peptide
peptide concentration (µM)
Fig. 2 The Melan-A51–73 speciWc CD4+ T cell clone, Mel18, is restricted by the HLA-DQa1*0101/HLA-DQb1*0501 molecules. a Mel18 and NY67 CD4+ T cell clones were cultured alone or with diVerent concentrations of Melan-A51–73 or NYESO-1157–170 peptides respectively in presence of Brefeldin A at 37°C during 6 h. Cells were then Wxed, permeabilized and stained for intracytoplasmic IFN-. Fluorescence was analyzed by Xow cytometry. b Mel18 and NY67 CD4+ T cell clones were cultured alone or with 10 M of Melan-A51–73 or NYESO-1157–170 peptides respectively in presence of Brefeldin A at 37°C during 6 h. Inhibitory mAb against HLA-DR, -DP or -DQ molecules were added to the cultures. Cells were then Wxed, permeabilized and stained for intracytoplasmic IFN-. Fluorescence was analyzed by Xow
cytometry. c Mel18 CD4+ T cell clone was cultured alone or with diVerent concentrations of Melan-A51–73 peptide pulsed HLADQb1*0501+, -DQb1*0502+ or -DQb1*0501/02- B-LCL in presence of Brefeldin A at 37°C during 6 h. Cells were then stained for surface CD4, Wxed, permeabilized and stained for intracytoplasmic IFN-. Fluorescence was analyzed by Xow cytometry with a gate on CD4+ cells. d U293 human embryonic kidney cells were transfected with plasmids encoding the transactivator CIITA, the invariant chain Ii, HLA-DQb1*0501 and -DQa1*0101 or -DQa1*0303 molecule. 48 h after transfection, U293 cells were pulsed or not with 10 M Melan-A51–73 peptide and washed. Then, they were cocultured with Mel18 CD4+ T cell clone. After 6 h, TNF was measured in co-culture supernatants
response. Removal of histidine 63 induced a decrease in Mel18 response, while shortening up to serine 61 abrogated the response. These results show that the 13-mer Melan-A51–63 peptide is the minimal peptide optimally recognized by the Mel18 CD4+ T cell clone. However, other Melan-A peptides contained within the 51–73 region represent potential epitopes for this clone.
The Melan-A epitope recognized by Mel18 CD4+ T cell clone is presented by Melan-A+ HLA-DQ5+ melanoma tumor cells
123
We then asked whether Melan-A, HLA-DQ5 expressing melanoma tumor cells (M253, HLA-DQb1*0501; M77, HLA-DQb1*0502; M6, HLA-DQb1*0504) were able to present the Melan-A epitope recognized by the
Cancer Immunol Immunother (2007) 56:1565–1575
A
1600 1400 tum cel IFN-γ tum cel tum cel + Mel18 IFN-γ tum cel + Mel 18 IFN-γ tum cel + peptide IFN-γ tum cel + peptide + Mel18
1200
TNF (pg/ml)
RNGYRALMDKSLHVGTQCALTRR NGYRALMDKSLHVGTQCALTRR GYRALMDKSLHVGTQCALTRR RNGYRALMDKSLHVGT RNGYRALMDKSLHVG RNGYRALMDKSLHV RNGYRALMDKSLH RNGYRALMDKSL RNGYRALMDKS RNGYRALMDK
1000 800 600 400
100
200
80
0
60
DQb1*0501+ DQb1*0504+ DQb1*0502+
M253
M6
M77
M96
M171
M47
M45
DQb1*05-
B
40 20 0
0
0.003 0.01 0.03 0.1
0.3
1
3
10
peptide concentration (µM)
Fig. 3 Melan-A51–63 peptide is the minimal epitope optimally recognized by Mel18 CD4+ T cell clone: HLA-DQb1*0501+ B-LCL GRE was pulsed with diVerent concentrations of a panel of truncated peptides located in Melan-A51–73 region during 2 h and washed. Then, Mel18 CD4+ T cell clone was co-cultured with the peptide pulsed B-LCL GRE in presence of brefeldin-A. After 6 h, IFN- secretion by the Mel18 T cell clone was assessed by IFN- intracytoplasmic and CD4 surface staining
CD4+ T cell clone Mel18. We measured the TNF response of clone Mel18 to these cell lines and to HLA-DQ5 negative cell lines (M45, M47, M96 and M171) previously treated or not by IFN- and previously pulsed or not with the Melan-A51–73 peptide. We observed a Mel18 response to untreated M6 and M253, and to a lower extent untreated M77 melanoma cell lines. IFN- treatment, which increases HLA class II molecules expression by melanoma cell lines (data not shown), increased the Mel18 response to M253 cell line. This response was further increased by MelanA51–73 peptide addition to M253 cell line. Therefore, a Melan-A peptide recognized by Mel18, is presented spontaneously by HLA-DQb1*0501+, -DQb1*0502+ and -DQb1*0504+ melanoma cell lines. The HLADQ5 negative cell lines were not recognized by the Mel18 clone. Since CD4+ T cells can be cytotoxic, we checked whether this was the case for the Mel18 clone. We observed that both M6 and M253 melanoma cell lines were killed at a signiWcant level by this clone (Fig. 4b). This activity was rather low and similar on both cell lines at an E/T ratio of 5. It was signiWcantly increased at higher E/T ratio and by IFN- treatment of melanoma cell lines. The HLA-DQ5 negative melanoma cell lines M199 and FM25 were not killed in these conditions.
% Lysis
% IFN-γ positive cells
(51-73) (52-73) (52-73) (51-66) (51-65) (51-64) (51-63) (51-62) (51-61) (51-60)
1571
45 40 35 30 25 20 15 10 5 0
M6 IFN-γ M6 M253 IFN-γ M253 M199 IFN-γ M199 FM25 IFN-γ FM25
5
10
20
Effector/Target Ratio
Fig. 4 The Melan-A epitope recognized by Mel18 CD4+ T cell clone is presented by Melan-A+ HLA-DQ5+ melanoma tumor cells. a Mel18 CD4+ T cell clone was cultured with diVerent untreated, IFN- treated melanoma tumor cell lines or Melan-A51– 73 peptide pulsed IFN- treated melanoma tumor cell lines. After 6 h, TNF was measured in culture supernatants. b Cytolytic activity of Mel18 CD4+ T cell clone against untreated or IFN- treated melanoma tumor cell lines assessed by 4 h 51Cr release assay
The Melan-A epitope recognized by Mel18 CD4+ T cell clone is not presented by soluble Melan-A protein pulsed HLA-DQ5+ DC We then assessed the ability of DC to process and present this HLA-DQ5 restricted Melan-A epitope from an exogenous supply of the Melan-A protein. HLADQb1*0501+ DC were incubated with soluble MelanA protein in the presence of TNF- and PolyI/C during 24 h, washed and co-cultured with clone Mel18 in presence of Brefeldin-A. These cells were unable to activate clone Mel18, whereas they eYciently induced IFN- production by this clone after pulsing with Melan-A51–73 peptide (Fig. 5a). The lack of presentation by Melan-A protein pulsed DC was not due to a defect of Melan-A protein internalization since we observed a strong Melan-A intracellular staining of DC incubated with the Melan-A protein (Fig. 5b). Presence of spontaneous CD4+ T cell responses to Melan-A51–63 and Melan-A51–73 among the TIL and PBMC of HLA-DQ5+ melanoma patients Finally, we looked for the presence of CD4+ T cells speciWc for Melan-A51–63 and Melan-A51–73 peptide among
123
1572
Cancer Immunol Immunother (2007) 56:1565–1575
Fig. 5 The Melan-A epitope recognized by Mel18 CD4+ T cell clone is not presented by soluble Melan-A protein pulsed DC. a HLA-DQb1*0501+ DC were cultured alone or with 50 g/ml of soluble Melan-A protein during 24 h in presence of a maturation stimulus (TNF-, PolyI/C). Some of DCs cultured alone were pulsed during 1 h with 10 M Melan-A51–73 peptide. Then, DC were washed and co-cultured with the Mel18 clone in presence of
brefeldin-A. After 6 h, IFN- production by Mel18 clone was assessed by IFN- intracytoplasmic staining. b HLA-DQb1*0501+ DC were cultured alone or with 50 g/ml of soluble Melan-A protein during 24 h in presence of a maturation stimulus (TNF-, PolyI/C). Then, DC were washed, Wxed and Melan-A MART-1 internalization was assessed by Melan-A intracellular staining
tumor inWltrating lymphocytes (TIL) populations of 6 HLA-DQ5+ patients: M6, M17, M74, M77, M125 and M253. We observed that 0.33% of the CD4+ TIL from the HLA-DQb1*0502+ patient M17 responded to the Melan-A51–63 peptide (Fig. 6a). We failed to detect any CD4+ T cell response against this peptide in the TIL populations of the Wve other HLA-DQ5+ patients. We then assessed the presence of HLA-DQ5 restricted Melan-A speciWc T cells in the blood of a melanoma patient. To this end, we Wrst stimulated CD4+ sorted PBMC obtained from HLA-DQb1*0501+ patient M253 by CD4- autologous PBMC and MelanA51–63 or Melan-A51–73 peptides. Two weeks later, we checked the capacity of stimulated PBMC to respond to these peptides. In two out of Wfty cultures stimulated by the Melan-A51–73 peptide, a fraction of CD4+ T cells (0.5 and 1.3%) responded to Melan-A51–63 and MelanA51–73 peptide (Fig. 6b). Lower but potentially signiWcant fractions of IFN- producing CD4 T cells were also detected among PBMC stimulated the Melan-A51–63 peptide (data not shown).
the blood of an healthy donor, we obtained an HLADQa1*0101/HLA-DQb1*0501 restricted Melan-A speciWc CD4+ T cell clone. Using this clone, we identiWed a number of potential HLA-DQ5 restricted Melan-A epitopes located in the region 51–73 and showed that at least one of these is presented spontaneously by HLADQ5+ melanoma cell lines. Interestingly, we also report data supporting the existence of CD4+ T cell responses against this epitope in some HLA-DQ5+ patients. Our results, together with those of Zarour et al. [35], show that the Melan-A51–73 peptides can be presented to CD4+ T cells in at least two diVerent HLA class II contexts: HLA-DR4 and HLA-DQ5. Melan-A51–73 peptide is therefore, another example of a CD4+ T cell promiscuous epitope presentation, as reported for several other tumor associated antigens: HER2/neu [19, 31], MAGE-A3 [18, 29], NYESO-1 [24], CEA [6] and gp100 [17]. HLA-DQa1*0101/HLA-DQb1*0501 alleles are expressed by about 10% of the French Caucasian populations (http://www.Allelefrequencies.net). Together with the high frequency of expression of HLADRb1*0401 [35], the present data suggests that MelanA51–73 could be used to stimulate CD4+ tumor reactive T cells responses from a signiWcant fraction of Caucasian patients. Importantly, melanoma tumor cells expressing MelanA and HLA-DQb1*0501, -DQb1*0502 or -DQb1*0504
Discussion In this study, while trying to obtain CD4+ T cells speciWc for Melan-A51–73 in the HLA-DRb1*0401 context from
123
Cancer Immunol Immunother (2007) 56:1565–1575
1573
Fig. 6 Presence of spontaneous CD4+ T cell responses to MelanA51–63 and Melan-A51–73 in TIL and PBMC of HLA-DQ5+ melanoma patients. a TIL from patient M17 were cultured with or without 10 M Melan-A51–63 in presence of brefeldin-A. After 6 h, IFN- production by CD4+ T cells was assessed by IFN- intracytoplasmic and CD4 surface staining. b CD4+ PBMC from
patient M253 were cultured with irradiated CD4- PBMC and 10 M of Melan-A51–73 in presence of IL-2 and -7. After 15 days, the cells were cultured with or without 10 M Melan-A51–63 or Melan-A51–73 in presence of brefeldin-A. After 6 h, IFN- production by CD4+ T cells was assessed by IFN- intracytoplasmic and CD4 surface staining
were recognized by the Melan-A51–73 speciWc CD4+ T cell clone. This result demonstrates that this epitope is naturally processed and eYciently presented by these tumor cells on HLA-DQ5 molecules, as it is also the case for tumor cells which present Melan-A51–73 in the HLA-DR4 context [35]. However, HLA-DQ5+ DC incubated with recombinant Melan-A protein failed to present this epitope to the clone, despite the fact that the Melan-A protein was internalized. Recently, Godefroy et al. [12] described two new Melan-A epitopes recognized by CD4+ T cell clones in the HLA-DR11 and -DR52 context. The HLA-DR52 restricted epitope was presented by HLA-DR52+ DC after incubation of DC with soluble Melan-A protein,
whereas the HLA-DR11 restricted epitope was not presented by HLA-DR11+ DC. HLA-DQb1*0501+ DC were not able to present the Melan-A51–73 epitope to the Mel18 CD4+ T cell clone after incubation with soluble Melan-A protein, suggesting that this epitope behaves like the HLA-DR11 restricted epitope described by Godefroy et al. [12]. The fact that the HLA-DQ5+ Melan-A CD4 epitope is eYciently presented by melanoma cell lines and not by DC suggests either that processing of this epitope depends on melanoma cell speciWc properties, like the addressing of melanosomal protein to the lysosmal compartment [22], or that the DC antigen processing pathway leads to the destruction or poor production of the
123
1574
HLA-DQ5 Melan-A CD4 epitope. However, we could not exclude that HLA-DQ5+ DC might be able to present this epitope after incubation with soluble Melan-A protein, but the Mel18 clone is not avid enough to respond to the few epitopes presented by DC. Our data also show that a spontaneous CD4+T cell response against this epitope may occur among TIL of HLA-DQ5+ melanoma patients, as we were able to detect such a response in TIL of one out of six patients. Furthermore, we were also able to detect such a CD4+ T cell response in one HLA-DQ5+ melanoma patient after a short-term stimulation of PBMC with the Melan-A51–73 peptide. Similar observations were also reported by Zarour et al. [35] and, recently, by Bioley et al. [5]. Zarour et al. [35] were able to detect spontaneous response against Melan-A51–73 in the blood of six out of ten HLA-DR4+ patient by ELISPOT, whereas they did not detect such responses in the blood of healthy donors. Bioley et al. [5] reported CD4+ T cell responses against several Melan-A epitopes presented in diVerent HLA classe II context after a short-term peptide stimulation of melanoma patients blood PBMC. Our results together with the results of Zarour et al. and Bioley et al., clearly show that spontaneous CD4+ T cell responses against diVerent Melan-A epitopes occur in vivo in multiple HLA class II context. Vaccination strategies consisting in the induction or ampliWcation of CD4+ T cell responses against this epitope are attractive, since we show that the Melan-A51–73 speciWc CD4+ T clone is able to lyse melanoma tumor cells, which presents this epitope. In addition, such CD4+ T cells response might provide help for the development of anti-tumour CTL responses, as it has been reported in several mouse models of tumor regression [26, 32]. However, such strategies have to be designed carefully to avoid the induction of CD4+ regulatory T cells response against this epitope that would be detrimental to the patients. Indeed, Wang et al. showed that tumor antigens LAGE-1 and ARTC1 can be recognized by CD4+ regulatory T cells [34], but conditions to avoid induction of such CD4+ regulatory T cells are not yet fully understood [33]. In conclusion, we have reported here that Melan-A/ MART1 contains a promiscuous CD4+ T cell epitope within the region 51–73 eYciently presented by HLADQb1*0501+, HLA-DQb1*0504+ and to a lower extent HLA-DQb1*0502+ melanoma tumor cells. Characterization of such epitopes will probably be helpful to design and to evaluate Melan-A/MART1 based anti-tumoral vaccination strategies. Acknowledgments This work was supported by grants from the « Ligue Nationale contre le Cancer » (labelisation 2002–2005).
123
Cancer Immunol Immunother (2007) 56:1565–1575
References 1. Adhikary D, Behrends U, Moosmann A, Witter K, Bornkamm GW, Mautner J (2006) Control of Epstein-Barr virus infection in vitro by T helper cells speciWc for virion glycoproteins. J Exp Med 203(4):995–1006 2. Benlalam H, Labarriere N, Linard B, Derre L, Diez E, PandolWno MC, Bonneville M, Jotereau F (2001) Comprehensive analysis of the frequency of recognition of melanoma-associated antigen (MAA) by CD8 melanoma inWltrating lymphocytes (TIL): implications for immunotherapy. Eur J Immunol 31(7):2007–2015 3. Benlalam H, Linard B, Guilloux Y, Moreau-Aubry A, Derre L, Diez E, Dreno B, Jotereau F, Labarriere N (2003) IdentiWcation of Wve new HLA-B*3501-restricted epitopes derived from common melanoma-associated antigens, spontaneously recognized by tumor-inWltrating lymphocytes. J Immunol 171(11):6283–6289 4. Bennett SR, Carbone FR, Karamalis F, Flavell RA, Miller JF, Heath WR (1998) Help for cytotoxic-T-cell responses is mediated by CD40 signalling. Nature 393(6684):478–480 5. Bioley G, Jandus C, Tuyaerts S, Rimoldi D, Kwok WW, Speiser DE, Tiercy JM, Thielemans K, Cerottini JC, Romero P (2006). Melan-A/MART-1-speciWc CD4 T cells in melanoma patients: identiWcation of new épitopes and ex vivo visualization of speciWc T cells by MHC class II tetramers. J Immunol 177(10):6769–6779 6. Campi G, Crosti M, Consogno G, Facchinetti V, Conti-Fine BM, Longhi R, Casorati G, Dellabona P, Protti MP (2003) CD4(+) T cells from healthy subjects and colon cancer patients recognize a carcinoembryonic antigen-speciWc immunodominant epitope. Cancer Res 63(23):8481–8486 7. Chiari R, Hames G, Stroobant V, Texier C, Maillere B, Boon T, Coulie PG (2000) IdentiWcation of a tumor-speciWc shared antigen derived from an Eph receptor and presented to CD4 T cells on HLA class II molecules. Cancer Res 60(17):4855–4863 8. Coulie PG, Brichard V, Van Pel A, Wolfel T, Schneider J, Traversari C, Mattei S, De Plaen E, Lurquin C, Szikora JP, Renauld JC, Boon T (1994) A new gene coding for a diVerentiation antigen recognized by autologous cytolytic T lymphocytes on HLA-A2 melanomas. J Exp Med 180(1):35–42 9. Danke NA, Koelle DM, Yee C, Beheray S, Kwok WW (2004) Autoreactive T cells in healthy individuals. J Immunol 172(10):5967–5972 10. Fearon ER, Pardoll DM, Itaya T, Golumbek P, Levitsky HI, Simons JW, Karasuyama H, Vogelstein B, Frost P (1990) Interleukin-2 production by tumor cells bypasses T helper function in the generation of an antitumor response. Cell 60(3):397–403 11. Fonteneau JF, Larsson M, Somersan S, Sanders C, Munz C, Kwok WW, Bhardwaj N, Jotereau F (2001) Generation of high quantities of viral and tumor-speciWc human CD4+ and CD8+ T-cell clones using peptide pulsed mature dendritic cells. J Immunol Methods 258(1–2):111–126 12. Godefroy E, Scotto L, Souleimanian NE, Ritter G, Old LJ, Jotereau F, Valmori D, Ayyoub M (2006) IdentiWcation of two Melan-A CD4(+) T cell epitopes presented by frequently expressed MHC class II alleles. Clin Immunol 121(1):54–62 13. Hansen MB, Nielsen SE, Berg K (1989) Re-examination and further development of a precise and rapid dye method for measuring cell growth/cell kill. J Immunol Methods 119(2):203–210 14. Jung T, Schauer U, Heusser C, Neumann C, Rieger C (1993) Detection of intracellular cytokines by Xow cytometry. J Immunol Methods 159(1–2):197–207
Cancer Immunol Immunother (2007) 56:1565–1575 15. Kawakami Y, Dang N, Wang X, Tupesis J, Robbins PF, Wang RF, Wunderlich JR, Yannelli JR, Rosenberg SA (2000) Recognition of shared melanoma antigens in association with major HLA-A alleles by tumor inWltrating T lymphocytes from 123 patients with melanoma. J Immunother 23(1):17–27 16. Kawakami Y, Eliyahu S, Sakaguchi K, Robbins PF, Rivoltini L, Yannelli JR, Appella E, Rosenberg SA (1994) IdentiWcation of the immunodominant peptides of the MART-1 human melanoma antigen recognized by the majority of HLA-A2restricted tumor inWltrating lymphocytes. J Exp Med 180(1):347–352 17. Kobayashi H, Lu J, Celis E (2001) IdentiWcation of helper T-cell epitopes that encompass or lie proximal to cytotoxic T-cell epitopes in the gp100 melanoma tumor antigen. Cancer Res 61(20):7577–7584 18. Kobayashi H, Song Y, Hoon DS, Appella E, Celis E (2001) Tumor-reactive T helper lymphocytes recognize a promiscuous MAGE-A3 epitope presented by various major histocompatibility complex class II alleles. Cancer Res 61(12):4773–4778 19. Kobayashi H, Wood M, Song Y, Appella E, Celis E (2000) DeWning promiscuous MHC class II helper T-cell epitopes for the HER2/neu tumor antigen. Cancer Res 60(18):5228–5236 20. Labarriere N, Diez E, PandolWno MC, Viret C, Guilloux Y, Le Guiner S, Fonteneau JF, Dreno B, Jotereau F (1997) Optimal T cell activation by melanoma cells depends on a minimal level of antigen transcription. J Immunol 158(3):1238–1245 21. Labarriere N, PandolWno MC, Raingeard D, Le Guiner S, Diez E, Le Drean E, Dreno B, Jotereau F (1998) Frequency and relative fraction of tumor antigen-speciWc T cells among lymphocytes from melanoma-invaded lymph nodes. Int J Cancer 78(2):209–215 22. Levy F, Muehlethaler K, Salvi S, Peitrequin AL, Lindholm CK, Cerottini JC, Rimoldi D (2005) Ubiquitylation of a melanosomal protein by HECT-E3 ligases serves as sorting signal for lysosomal degradation. Mol Biol Cell 16(4):1777–1787 23. Lu Z, Yuan L, Zhou X, Sotomayor E, Levitsky HI, Pardoll DM (2000) CD40-independent pathways of T cell help for priming of CD8(+) cytotoxic T lymphocytes. J Exp Med 191(3):541–550 24. Mandic M, Castelli F, Janjic B, Almunia C, Andrade P, Gillet D, Brusic V, Kirkwood JM, Maillere B, Zarour HM (2005) One NY-ESO-1-derived epitope that promiscuously binds to multiple HLA-DR and HLA-DP4 molecules and stimulates autologous CD4+ T cells from patients with NY-ESO-1expressing melanoma. J Immunol 174(3):1751–1759 25. Manici S, Sturniolo T, Imro MA, Hammer J, Sinigaglia F, Noppen C, Spagnoli G, Mazzi B, Bellone M, Dellabona P, Protti MP (1999) Melanoma cells present a MAGE-3 epitope
1575
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
to CD4(+) cytotoxic T cells in association with histocompatibility leukocyte antigen DR11. J Exp Med 189(5):871–876 Ossendorp F, Mengede E, Camps M, Filius R, Melief C J (1998) SpeciWc T helper cell requirement for optimal induction of cytotoxic T lymphocytes against major histocompatibility complex class II negative tumors. J Exp Med 187(5):693–702 Ridge JP, Di Rosa F, Matzinger P (1998) A conditioned dendritic cell can be a temporal bridge between a CD4+ T-helper and a T-killer cell. Nature 393(6684):474–478 Schoenberger SP, Toes RE, van der Voort EI, OVringa R, Melief CJ (1998) T-cell help for cytotoxic T lymphocytes is mediated by CD40-CD40L interactions. Nature 393(6684):480– 483 Schultz ES, Schuler-Thurner B, Stroobant V, Jenne L, Berger TG, Thielemanns K, van der Bruggen P, Schuler G (2004) Functional analysis of tumor-speciWc Th cell responses detected in melanoma patients after dendritic cell-based immunotherapy. J Immunol 172(2):1304–1310 Seiter S, Monsurro V, Nielsen MB, Wang E, Provenzano M, Wunderlich JR, Rosenberg SA, Marincola FM (2002) Frequency of MART-1/MelanA and gp100/PMel17-speciWc T cells in tumor metastases and cultured tumor-inWltrating lymphocytes. J Immunother 25(3):252–263 Sotiriadou R, Perez SA, Gritzapis AD, Sotiropoulou PA, Echner H, Heinzel S, Mamalaki A, Pawelec G, Voelter W, Baxevanis CN, Papamichail M (2001) Peptide HER2(776– 788) represents a naturally processed broad MHC class II-restricted T cell epitope. Br J Cancer 85(10):1527–1534 Toes RE, Ossendorp F, OVringa R, Melief CJ (1999) CD4 T cells and their role in antitumor immune responses. J Exp Med 189(5):753–756 Wang RF (2006) Immune suppression by tumor-speciWc CD4+ regulatory T-cells in cancer. Semin Cancer Biol 16(1):73–79 Wang HY, Lee DA, Peng G, Guo Z, Li Y, Kiniwa Y, Shevach EM, Wang RF (2004) Tumor-speciWc human CD4+ regulatory T cells and their ligands: implications for immunotherapy. Immunity 20(1):107–118 Zarour HM, Kirkwood JM, Kierstead LS, Herr W, Brusic V, SlingluV CL Jr, Sidney J, Sette A, Storkus WJ (2000) MelanA/MART-1(51–73) represents an immunogenic HLA-DR4restricted epitope recognized by melanoma-reactive CD4(+) T cells. Proc Natl Acad Sci U S A 97(1):400–405 Zeng G, Wang X, Robbins PF, Rosenberg SA, Wang RF (2001) CD4(+) T cell recognition of MHC class II-restricted epitopes from NY-ESO-1 presented by a prevalent HLA DP4 allele: association with NY-ESO-1 antibody production. Proc Natl Acad Sci U S A 98(7):3964–3969
123
