|
|
||||||||
Department of Clinical Oncology, Leiden University Medical Center, Leiden, The Netherlands
| Abstract |
|---|
|
|
|---|
-chains, but different complementarity-determining region 3 loops of the TCR
-chains. Introduction of the TCRs into proper recipient cells should reveal whether the different complementarity-determining region 3
loops are important for tumor cell recognition. | Introduction |
|---|
|
|
|---|
Recently, CD4+ T cells recognizing MHC class II-restricted epitopes encoded by cancer/testis Ags like NY-ESO-1 (10, 11), MAGE-6 (12), and CAMEL (13) have been detected in peripheral blood of cancer patients, strongly suggesting that these tumor Ags are processed and presented in the MHC class II pathway in vivo. In cancer patients, however, the tumor Ag-specific CD4+ T cells often belong to the Th2 subtype, producing cytokines such as IL-4, IL-5, and IL-13 (12, 13), whereas IL-2- and IFN-
-producing Th1 cells are believed to be required for proper activation of cytotoxic CD8+ T cells. It has been suggested that tumor Ag-specific CD4+ Th2 cells in cancer patients are induced by malfunctioning DC, expressing low levels of MHC and costimulatory molecules (14, 15, 16, 17), whereas mature DC, expressing high levels of MHC and costimulatory molecules, are required for the induction of Th1 immunity and potent CD8+ T cell-mediated antitumor responses.
CAMEL (109 aa) is a tumor-specific protein translated from the LAGE-1 gene in an alternative open reading frame (ORF) starting at a second ATG start codon located 40 bp downstream of the first ATG start site (18). LAGE-1 shows 94% homology to tumor Ag NY-ESO-1 and both genes are located on chromosome Xp28 (19, 20). The LAGE-1 and NY-ESO-1 genes are frequently coexpressed in a wide variety of tumors, like melanoma, breast carcinoma, prostate and bladder cancers, but silent in normal tissues except for testis (19, 20, 21). Like LAGE-1, the NY-ESO-1 gene is translated in an alternative ORF, resulting in a 58 aa NY-ESO-ORF2 protein from which the N-terminal 54 aa are identical with CAMEL.
In this study, we investigated whether CD4+ T cell responses could be induced in vitro by mature DC pulsed with recombinant CAMEL protein. The data show efficient induction of CAMEL-specific CD4+ T cells expressing both Th1 and Th2 cytokines in two healthy donors. Because CAMEL and NY-ESO-ORF2 are not expressed in normal tissues, these CAMEL-specific CD4+ T cells resulted from in vitro priming of naive CD4+ T cells. The isolation of CD4+ T cell clones from the T cell cultures of both donors led to the identification of four naturally processed HLA-DR-binding CAMEL epitopes: CAMEL120, CAMEL1433, CAMEL4665, and CAMEL81102. Two of these peptides (CAMEL120 and CAMEL1433) also contain previously identified HLA class I-binding CD8+ T cell epitopes.
| Materials and Methods |
|---|
|
|
|---|
Melanoma cell lines 518A2, 90.07, 93.08, 93.11, 93.15.2, and 94.03 were established in our laboratory. Line 518A2/IL-2.14 was obtained by transfecting the 518A2 cell line with IL-2 cDNA (22). Melanoma cell lines FM3 and FM6 were kindly provided by J. Zeuthen (Danish Cancer Society, Copenhagen, Denmark). All melanoma cell lines were cultured in DMEM (Invitrogen, Breda, The Netherlands) supplemented with 8% heat-inactivated FCS, 4 mM L-glutamine, 50 µg/ml penicillin and 50 µg/ml streptomycin. EBV-transformed B-lymphoblastic cell lines (EBV-B cells) were generated from PBMC of healthy donors and maintained in RPMI 1640 (Invitrogen) with FCS, L-glutamine and antibiotics. All melanoma cell lines and EBV-B cells were HLA typed using standard serological methods.
Peptides and recombinant protein
Peptides were synthesized by solid-phase methods, using an automated multiple peptide synthesizer (Abimed AMS 422; Abimed Analyes-Technik, Langenfeld, Germany) and Fmoc chemistry. After reversed phase HPLC analysis, peptides were dissolved in DMSO at 50 mg/ml and stored at −70°C. From this stock solution, peptide was diluted in PBS to a final concentration of 1 mg/ml and stored at −20°C.
Recombinant CAMEL protein was produced in a bacterial expression system. Briefly, a 349 bp LAGE-1 cDNA fragment was amplified using a sense primer containing a NdeI site (5'-GAAGAACATATGCTGATGGCCCAGGAGGC-3') and an antisense primer containing a BglII site (5'-TTAAAGATCTCAGAACCGCCCCTGGTCG-3'). The LAGE-1 cDNA fragment only encodes CAMEL due to deletion of the first ATG start site. The cDNA fragment was cloned into the NdeI and BamHI sites of the pET19b vector (Novagen, Madison, WI) in frame with an N-terminal (His)10-tag. Expression of pET19b-CAMEL in Escherichia coli strain BL21(DE3) (Stratagene, Amsterdam, The Netherlands) was induced by addition of 1 mM isopropyl-
-D-thiogalactoside (Sigma-Aldrich, Zwijndrecht, The Netherlands). After 4 h, rCAMEL was purified from bacterial lysate by nickel-chelate affinity chromatography using Ni-NTA agarose according to manufacturers instructions (Westburg, Leusden, The Netherlands). Recombinant CAMEL was eluted in 8 M urea, 100 mM NaH2PO4, and 10 mM Tris, pH 4.5 and dialyzed in distilled water. After lyophilization, recombinant CAMEL was dissolved in distilled water at 1 mg/ml and stored at −70°C.
Generation of DCs
PBMC were isolated by Ficoll gradient centrifugation from heparinized blood from two healthy donors. Donors were HLA class II-typed as HLA-DR3, -DR7, -DR52, -DR53, -DQ2 (donor 1) and HLA-DR3, -DR4, -DR52, -DR53, -DQ2, -DQ8 (donor 2) using standard serological techniques. PBMC were depleted for T and B cells using CD2- and CD19-dynabeads (Dynal Biotech, Oslo, Norway) and cultured at 2 x 106 cells/well in 6-well plates in AIM-V (Invitrogen) with 116 µg/ml L-arginine, 36 µg/ml L-asparagine, 215 µg/ml L-glutamine (AAG), and antibiotics, supplemented with 500 U/ml IL-4 (PeproTech, Rocky Hill, NJ) and 800 U/ml GM-CSF (Behringwerke, Marburg, Germany) for 6 days. Immature DC were pulsed with 10 µg/ml recombinant CAMEL for 4 h at 37°C and subsequently matured for 2 days in AIM-V containing 1 µg/ml activating anti-CD40 Ab (Cymbus Biotechnology, Chandlers Ford, U.K.), 500 U/ml IL-4 and 800 U/ml GM-CSF. For peptide loading, immature DC were cultured for 2 days in AIM-V containing anti-CD40 Ab, IL-4 and GM-CSF as previously described and subsequently pulsed with 10 µg/ml peptide for 2 h at 37°C.
Generation of CAMEL-specific CD4+ T cells
For the induction of CAMEL-specific CD4+ T cells, 20 x 106 nonadherent PBMC were incubated with autologous, mature DC pulsed with recombinant CAMEL at a ratio of 10:1 at 2 x 106 PBMC/well in 24-well plates in IMDM (Invitrogen) with 5% pooled human AB serum (HS), AAG and antibiotics, supplemented with 10 ng/ml IL-7 (PeproTech) and 100 pg/ml IL-12 (Sigma-Aldrich). T cell cultures were weekly restimulated as previously described in IMDM with 5% HS, AAG and antibiotics, supplemented with 150 U/ml IL-2.
To obtain CAMEL-specific CD4+ T cell clones, T cell cultures were seeded at 1 cell/well in 96-well U-bottom plates, each well containing 5 x 103 irradiated, autologous DC pulsed with recombinant CAMEL, 105 irradiated allogeneic PBMC, 5 x 103 irradiated allogeneic EBV-B cells and 1 µg/ml leucoagglutinin (Sigma-Aldrich) and 150 U/ml IL-2. Because many T cell clones isolated from donor 1 were shown to be specific for bacterial Ags, the T cell clones from donor 2 were stimulated with autologous DC pulsed with CAMEL peptide mix 1 and mix 2 instead of recombinant CAMEL protein. Growing clones were weekly restimulated as previously described.
IFN-
ELISPOT assay
IFN-
ELISPOT assays were performed as previously described with small modifications (23). Ninety six-well nylon Silent Screen plates (Nalge Nunc International, Rochester, NY) were coated with 100 µl of a mouse mAb against human IFN-
(1-D1K; Mabtech, Nacka, Sweden) diluted to 5 µg/ml in PBS overnight at 4°C. Wells were washed with PBS and blocked with 50 µl IMDM with 5% HS for 1 h at 37°C. T cells (15 x 104) were incubated together with target cells (12 x 104 cells/well) in duplicate. For blocking studies, target cells were preincubated with Abs against HLA class I (w6/32), HLA-DR (B8.11.2), HLA-DQ (SPV-L3), or HLA-DQ (B7.21) (24, 25) for 30 min at 37°C. After overnight incubation at 37°C, wells were washed in PBS/0.05% Tween 20 and incubated with 100 µl of a biotinylated Ab against human IFN-
(7-B6-1-biotin; Mabtech) diluted at 0.3 µg/ml in PBS for 2 h at room temperature. After several washes, wells were incubated with complexes between avidin and biotinylated HRP (Vectastain ABC system; Vector Laboratories, Burlingame, CA) for 1 h at 37°C and subsequently with 100 µl substrate solution containing 3-amino-9-ethylcarbazole (Sigma-Aldrich). Alternatively, wells were incubated with 100 µl streptavidin-alkaline phosphatase (Mabtech) in PBS/0.5% FCS for 1 h at room temperature and subsequently with substrate solution (alkaline phosphatase-conjugate substrate kit; Bio-Rad, Hercules, CA). Colorimetric reactions were stopped under running tap water and spots were counted by computer-associated video image analysis using KS ELISPOT software release 4.1 (Carl Zeiss Vision, Hallbergmoos, Germany).
IL-13 ELISPOT assay
The IL-13 ELISPOT assay was performed as previously described (13). Briefly, 96-well ELISA plates (Greiner, Alphen aan den Rijn, The Netherlands) were coated with 100 µl of an Ab against human IL-13 (QS-13; U-CyTech, Utrecht, The Netherlands) diluted to 10 µg/ml in PBS overnight at 4°C. Plates were washed several times with PBS/0.05% Tween 20 and blocked with 50 µl/well PBS/1% BSA for 1 h at 37°C. T cells (15 x 104 cells/well) were seeded together with target cells (12 x 104 cells/well) in duplicate. For blocking studies, target cells were preincubated with Abs against HLA class I (w6/32), HLA-DR (B8.11.2), HLA-DQ (SPV-L3), or HLA-DP (B7.21) (24, 25). After 5 h at 37°C, wells were incubated with 200 µl ice-cold deionized water on melting ice for 10 min, washed several times and subsequently incubated with 100 µl of a biotinylated polyclonal Ab against human IL-13 (U-CyTech) diluted in PBS/1% BSA overnight at 4°C. After several washes, wells were incubated with 50 µl of a gold-labeled anti-biotin Ab for 1 h at 37°C. Wells were washed several times and incubated with 30 µl activator mix (U-CyTech) until silver salt precipitates were formed at the site of gold clusters. Plates were washed and spots were counted using light microscopy.
Cloning of TCR
- and
-chain
TCR
-chain and
-chain were identified as described before (26). The complete coding sequences of the TCR
- and
-chains were amplified using the following specific primers containing the ATG start site: V
3 ATG, 5'-CCACCATGGAAACTCTCCTGGGAG-3' in combination with C
, 5'-GATGGCGGAGGCAGTCTCTG-3' for the TCR
-chain and V
5 ATG, 5'-TGCCATGGGCTCCAGGCTGC-3' in combination with C
1, 5'-TCCAGGGCTGCCTTCAGAAATCC-3' for the TCR
-chains. PCR products were cloned into TA-cloning vector pCR2.1-topo (Invitrogen) and sequenced (BaseClear, Leiden, The Netherlands).
| Results |
|---|
|
|
|---|
(Th1) and IL-13 (Th2) ELISPOT assays. Peptide mix 1 consisted of three 20-mer peptides (CAMEL120, CAMEL1433, and CAMEL4665) with predicted binding motifs for several HLA-DR alleles (data not shown), whereas peptide mixes 2 and 3 contained four 22-mer peptides covering aa 2172 and aa 61109 of the CAMEL protein sequence, respectively. As shown in Fig. 1, the T cell culture of donor 1 recognized autologous DC pulsed with CAMEL peptide mix 3, whereas autologous DC pulsed with CAMEL peptide mix 1 and to a lesser extent CAMEL peptide mix 2 were recognized by the T cell culture of donor 2. Fig. 1 shows the number of spots produced in IFN-
ELISPOT assays, but similar results were obtained in IL-13 ELISPOT assays (data not shown). For further analysis of the recognized CAMEL epitope(s), several CD4+ T cell clones were isolated from the T cell cultures of donor 1 and donor 2 by limiting dilution. Two CD4+ T cell clones (clones 1/21 and 1/53) of donor 1 and four CD4+ T cell clones (clones 2/12, 2/33, 2/52, and 2/54) of donor 2 were isolated.
|
as well as IL-13, as illustrated for clone 1/21 in Fig. 2A. CAMEL81102 is a naturally processed epitope because the CD4+ T cell clones 1/21 and 1/53 also recognized DC pulsed with recombinant CAMEL protein. Recognition of the CAMEL81102 peptide could be blocked by Abs against HLA-DR, whereas Abs against HLA-DQ, HLA-DP, and HLA class I had no effect (Fig. 2B). Pulsing CAMEL81102 on EBV-B cells sharing one or more HLA-DR alleles with donor 1 demonstrated that the CD4+ T cell clones 1/21 and 1/53 both recognize the CAMEL81102 epitope in the context of HLA-DR7 (illustrated for clone 1/21 in Fig. 3). Whether the CAMEL81102 epitope is also presented by tumor cells could not be investigated because HLA-DR7-positive, CAMEL-expressing tumor cells were not available.
|
|
as well as IL-13, whereas CD4+ T cell clone 2/33 only produced IL-13 and not or hardly IFN-
(Fig. 4A). For all four CD4+ T cell clones, peptide recognition was inhibited by Abs against HLA-DR, but not by Abs against HLA-DQ, HLA-DP, or HLA class I (Fig. 4B), indicating that the CAMEL epitopes are presented in HLA-DR. For further analysis of the restriction elements, CD4+ T cell clones 2/52 and 2/54 were tested against a panel of CAMEL1433-pulsed EBV-B cells sharing one or more HLA-DR alleles with donor 2. Fig. 5 shows that CD4+ T cell clone 2/52 as well as clone 2/54 specifically produced IFN-
upon stimulation with the CAMEL1433 peptide in the context of HLA-DR3. The CD4+ T cell clones 2/12 and 2/33 failed to grow and thus could not be further analyzed for their restriction elements.
|
|
upon stimulation with melanoma cell lines that expressed HLA-DR3 and CAMEL (Fig. 6), nor granzyme B, TNF-
, IL-4, and IL-13 (data not shown).
|
release by CD4+ T cell clones 2/52 and 2/54. As shown in Fig. 7B, CD4+ T cell clones 2/52 and 2/54 recognize the same minimal CAMEL epitope: AAQERRVPR (aa 1927). Moreover, substitution of amino acids in the minimal epitope with alanine residues had similar effects on the recognition by CD4+ T cell clones 2/52 and 2/54 (Fig. 7C), indicating that the same amino acids are contacted by the TCRs of both clones. Analysis of surface expression of CD3 and other molecules involved in T cell activation (CD2, CD4, CD25, CD28, LFA-1) did not reveal any difference between CD4+ T cell clones 2/52 and 2/54 (data not shown). In addition, expression of CTLA-4, CD94, and killer cell Ig-like receptors could not be detected (data not shown).
|
expression was analyzed using a panel of Abs recognizing 24 different TCRV
. FACS analysis demonstrated that CD4+ T cell clones 2/52 and 2/54 both expressed TCRV
5.1 (data not shown). Sequencing of the TCR
- and
-chains demonstrated that the CD4+ T cell clones 2/52 and 2/54 express identical TCR
-chain, but revealed expression of different TCR
-chain (Table I). In the TCR
-chain of clone 2/52, the AV3S1 gene is combined with the AJ39 gene, whereas in the TCR
-chain of clone 2/54 the AJ-48 gene is incorporated, indicating that CD4+ T cell clones 2/52 and 2/54 express TCRs that differ in 10 aa in the complementarity-determining region (CDR)3
loop of the TCR
-chain.
|
| Discussion |
|---|
|
|
|---|
and IL-2 are required for the efficient activation and maintenance of CD8+ T cell-mediated immune responses (1, 2, 3, 4, 5, 6, 7, 8, 9). Therefore, induction of tumor-specific CD4+ Th1 cells should be addressed in the development of immunotherapy for cancer. In this study, CD4+ T cells specific for the alternatively translated tumor Ag CAMEL were induced in vitro by mature DC pulsed with bacterially expressed recombinant CAMEL protein in two healthy donors. Because CAMEL is not expressed in normal tissues, these CAMEL-specific CD4+ T cells could be considered as induced by in vitro priming of naive CD4+ T cells by DC pulsed with recombinant CAMEL protein. It has been reported by Chaux et al. (27) that stimulation with DC pulsed with bacterially expressed recombinant MAGE-3 protein resulted in strong activation of T cells specific for bacterial contaminants. In our experiments, T cells specific for bacterial components were also stimulated by DC pulsed with recombinant CAMEL protein, as demonstrated by recognition of a bacterially expressed control protein (data not shown). However, the T cell cultures of both donors were also shown to contain T cells specific for CAMEL peptides, demonstrating that CAMEL-specific T cells can still be generated even in the presence of T cells reactive with bacterial Ags.
The isolation of CD4+ T cell clones from the T cell cultures stimulated with DC pulsed with recombinant CAMEL protein led to the identification of four naturally processed, HLA-DR-binding CAMEL epitopes: CAMEL120, CAMEL1433, CAMEL4665, and CAMEL81102. The HLA-DR restriction elements of the CAMEL1433- and CAMEL81102-specific CD4+ T cell clones were HLA-DR3 and HLA-DR7, respectively. These HLA-DR alleles are expressed in 1525% (HLA-DR3) and 2530% (HLA-DR7) of the study population. We recently demonstrated that the CAMEL81102 peptide is recognized in the context of HLA-DR11 and HLA-DR12 by CD4+ T cell clones isolated from peripheral blood of a melanoma patient (13). These HLA-DR11 and HLA-DR12-restricted CD4+ T cell clones and the HLA-DR7-restricted CD4+ T cell clone described in this study, both recognize the same minimal CAMEL8693 (PWKRSWSA) epitope (Ref. 13 and data not shown), indicating that CD4+ T cells can be stimulated by the CAMEL8693 epitope in the context of various HLA-DR alleles.
The HLA-DR11- and HLA-DR12-restricted, CAMEL81102-specific CD4+ T cell clones isolated from a melanoma patient have been reported to belong to the Th2 subtype, producing high levels of IL-4, IL-5, and IL-13 (13). Tumor Ag-specific CD4+ Th2 cells are probably not sufficient for tumor eradication because Th2-type immunity has been described to correlate with active disease in melanoma and renal cell carcinoma patients, whereas Th1 or mixed Th1/Th2 immunity was displayed in disease-free patients (12). It can be speculated that in cancer patients tumor Ag-specific Th2 cells are induced by DC impaired in their maturation, whereas fully mature DC are required for Th1 immunity and CD8+ T cell-mediated antitumor responses. In our study, five of the six CD4+ T cell clones produce Th1 as well as Th2 cytokines. The generation of CD4+ T cell clones that produce Th1 cytokines might be attributed to the fully mature DC that were used for induction.
Because melanomas have been described to express MHC class II molecules (28), CD4+ T cells may directly recognize epitopes processed and presented in the MHC class II pathway of tumor cells. Therefore, the CAMEL1433-specific CD4+ T cell clones 2/52 and 2/54 were both tested for recognition of HLA-DR3-positive, CAMEL-expressing tumor cells. Whereas CD4+ T cell clones 2/52 and 2/54 were both shown to recognize the exogenously pulsed CAMEL1433 peptide with similar affinities, only CD4+ T cell clone 2/54, and not clone 2/52, recognized HLA-DR3-positive CAMEL-expressing tumor cells. Lack of tumor cell recognition by CD4+ T cell clone 2/52 could not be explained by expression of inhibitory NK receptors as described by others (29). A difference in cell surface expression of costimulatory and adhesion molecules between the CD4+ T cell clones could not be demonstrated. However, despite similar expression, differential engagement of these molecules in tumor cell recognition by CD4+ T cell clones 2/52 and 2/54 cannot be excluded. Interestingly, like our CD4+ T cell clones 2/52 and 2/54, CD4+ T cells specific for the immunodominant epitope of hen-egg white lysozyme (HEL) have been described to differ in recognition of endogenously processed HEL epitopes, whereas exogenously pulsed HEL epitopes were similarly recognized (30). In this study, however, the difference in recognition of endogenously processed epitopes could be explained by TCRs that make different contacts with residues of the HEL epitope (31), whereas the TCRs of CD4+ T cell clones 2/52 and 2/54 were shown to contact similar residues of the minimal CAMEL1927 epitope.
TCR analysis demonstrated that CD4+ T cell clones 2/52 and 2/54 express identical TCR
-chains, but different TCR
-chains with different incorporated J gene segments. The junction of the V gene and J gene segment encodes the CDR3 of the TCR
-chain (32). Crystal structures of TCR-MHC peptide complexes revealed that CDR3 loops play a role in the interaction of the TCR with antigenic peptide bound to the MHC molecule (33, 34, 35). Therefore, the difference in tumor cell recognition between CD4+ T cell clones 2/52 and 2/54 might be explained by the 10 different amino acids in the CDR3 loop of the TCR
-chain. Introduction of the TCRs of CD4+ T cell clones 2/52 and 2/54 into a proper recipient cell line should reveal whether the different CDR3
loops are important for tumor cell recognition.
Thus far, two HLA class I-restricted epitopes that are shared by the alternatively translated CAMEL and NY-ESO-ORF2 proteins have been identified, i.e., the HLA-A*0201-restricted CAMEL111 epitope (18) and the HLA-A31-restricted CAMEL1827 epitope (36). Therefore, the CAMEL120 and CAMEL1433 peptides not only contain HLA-DR-binding CD4+ T cell epitopes, as described in this study, but also HLA class I-binding CD8+ T cell epitopes. Because several studies have shown that synthetic peptides comprising overlapping Th and CTL epitopes efficiently induce tumor Ag-specific CD4+ T cell responses and powerful, long-lasting tumor Ag-specific CD8+ T cell responses (37, 38, 39), the CAMEL/NY-ESO-ORF2 peptides harboring CD4+ and CD8+ T cell epitopes are interesting peptides to investigate for future immunotherapy of cancer.
| Acknowledgments |
|---|
| Footnotes |
|---|
2 Abbreviations used in this paper: DC, dendritic cell; ORF, open reading frame; HS, human serum; HEL, hen-egg white lysozyme; CDR, complementarity-determining region. ![]()
Received for publication July 14, 2003. Accepted for publication February 6, 2004.
| References |
|---|
|
|
|---|
-ELISPOT assay. Int. J. Cancer 93:549.[Medline]

T-cell receptor, influenza HA peptide and MHC class II molecule, HLA-DR1. EMBO J. 19:5611.[Medline]
This article has been cited by other articles:
![]() |
O. Ho and W. R. Green Alternative Translational Products and Cryptic T Cell Epitopes: Expecting the Unexpected J. Immunol., December 15, 2006; 177(12): 8283 - 8289. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Lucchese, J. Willers, A. Mittelman, D. Kanduc, and R. Dummer Proteomic Scan for Tyrosinase Peptide Antigenic Pattern in Vitiligo and Melanoma: Role of Sequence Similarity and HLA-DR1 Affinity J. Immunol., November 15, 2005; 175(10): 7009 - 7020. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |