| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Surgery Branch, National Cancer Institute, Bethesda, MD 20892
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We previously isolated a gene encoding gp75/TRP-1 as a tumor Ag recognized by TIL586, which was associated with antitumor activity in vivo (18). T cell clones were then established from the TIL586 cell line by the limiting dilution method. Some of the clones were capable of recognizing A31 + normal melanocytes, 586mel tumor cells, and 586EBV B cells pulsed with the peptide ORF3P (MSLQRQFLR), derived from the alternative open reading frame of the TRP-1 gene (24). Several T cell clones, however, did not appear to recognize the TRP-1 gene and its ORF3P peptide, although they were capable of recognizing 586mel tumor cells and HLA-A31+ melanocytes. This led to the identification of TRP-2 as a tumor Ag recognized by CTL clones (19). Interestingly, murine TRP-2 has been identified as a tumor Ag for B16 melanoma recognized by CTL generated by immunization of murine splenocytes with B16 tumor cells (25). This represents the first example of both human and mouse homologous proteins of TRP-2 acting as tumor Ags that are recognized by their corresponding CTLs. Therefore, the murine counterpart may be an ideal Ag for evaluating the role of differentiation Ags in antitumor responses in mice.
While a number of tumor Ags have been shown to be recognized by melanoma-reactive T cells, there are only limited data demonstrating direct recognition of tumor-shared Ags expressed in different types of tumor by CTL. The MAGE family of gene has been shown to be expressed in many types of tumor, but data demonstrating recognition of these tumor types by CTL have not been presented. Several reports have indicated that it is possible to generate CTL using peptides from HER-2/neu which is overexpressed in 20 to 30% of breast tumors (26, 27, 28). However, HER-2/neu is expressed in normal tissues as well. In this report, we demonstrate that CTL clones, or cloids, derived from the tumor-reactive TIL586 line recognize a breast and melanoma-shared tumor Ag encoded by the cancer ag gene 3 (CAG-3), identical to NY-ESO-1independently isolated by using the serum derived from a patient with esophageal cancer (29). Strikingly, we found that CTL clones were capable of recognizing HLA-A31-positive fresh and cultured breast tumor cells. Antigenic peptides from the normal open reading frame as well as an alternative open reading frame of the NY-ESO-1 gene were identified. This represents the first example that two different gene products translated from the same gene can be recognized by CTL clones derived from a TIL with antitumor activity in vivo. Therefore, NY-ESO-1 could be an important tumor Ag for use in the immunotherapy of patients with breast cancer, melanoma, and other types of cancer.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following chemicals and reagents were purchased from the sources indicated: RPMI 1640, AIM-V medium, Lipofectamine, and G418 from Life Technologies (Gaithersburg, MD); the eukaryotic expression vector pcDNA3.1 from Invitrogen (San Diego, CA); anti-HLA-A31 mAb from One Lambda (Canoga Park, CA); and anti-IgM Ab conjugated with FITC from Vector Laboratories (Burlingame, CA).
Cell lines and T cell clones
Melanoma cell lines 397mel, 397mel/A31, 586mel, 624mel, 624mel/A31, and EBV-transformed B cell lines 586EBV and 1510EBV were established in our laboratory and cultured in RPMI 1640 medium containing 10% FCS. Normal cultured melanocytes derived from infant foreskin (NHEM680 purchased from Clonetics, San Diego, CA) were cultured in melanocyte growth medium (MGM; Clonetics). 1295Br and 1315Br, fresh cryopreserved breast tumor digests, were cleaned with Ficoll gradient before use in T cell assays; 1295 fibroblast were cultured cells from the autologous patient for 22 to 64 days. 1315Br, culture A, were breast tumor cells grown in immunodeficient mice and then cultured in keratinocyte-SFM/2% FCS medium (Life Technologies), and 1315Br, culture B, were grown in Hams F12/5% FCS for 77 to 80 passages; these cells were kindly provided by Dr. Stephen Ethier, University of Michigan (Ann Arbor, MI). 1398Br was a human papillomavirus (HPV) E6/E7-immortalized breast tumor line established in the Surgery Branch, National Cancer Institute. Prostate tumor lines 1535Pro, 1542Pro, and 1510 fibroblast were HPV E6/E7-immortalized cell lines.
TIL586 were isolated from the tumor specimen of a patient with metastatic melanoma and grown in medium containing IL-2 (6000 IU/ml) (Chiron, Emeryville, CA) for 32 to 60 days as previously described (30). TIL586 and TIL1244 were predominantly CD8+ T cells. TIL1244 recognized the TRP-2 peptide in the context of HLA-A31 and -A33 (31). The T cell clones or cloids were generated by limiting dilution methods (at 1 cell/well) from the TIL586 cell line, using allogeneic PBL (1 x 103 cells/well) as feeder cells in RPMI 1640 containing 10% human AB sera and 500 IU IL-2. After 12 days, the T cell clones were expanded in AIM-V medium containing 6000 IU/ml IL-2. To obtain an optimal expansion, we used the OKT3 expansion method, described by S. Riddell (32). Briefly, on day 0, 5 x 104 to 5 x 105 T cells were cocultured with HLA-A31+ PBL (500:1, PBL:T cell ratio) and 586EBV B cells (100:1, EBV:T cell ratio) in 25 ml of RPMI 1640 containing 11% human sera, 30 ng/ml OKT3 Ab, and antibiotics. On day 1, IL-2 was added at a final concentration of 180 IU/ml. The medium was changed with fresh medium containing 11% human sera and 180 IU/ml IL-2 on day 5. The medium was then changed every 3 days. On days 12 through 14, T cells were harvested, counted, and cryopreserved.
cDNA library construction
Total RNA was extracted from 586mel using Trizol reagent (Life
Technologies). Poly(A) RNA was purified from total RNA by the
polyATtract system (Promega, Madison, WI) and converted to cDNA using a
cDNA construction kit (Life Technologies) with an oligo(dT) primer
containing a NotI site. The cDNA was ligated to
BstXI adaptors and digested with NotI, then
ligated to the expression vector pcDNA3.1. The cDNA library was
electroporated into DH10B cells (Life Technologies). Plasmid DNA pools,
each consisting of 
100 cDNA clones, were prepared from bacteria.
cDNA library screening and granulocyte/macrophage-CSF (GM-CSF) secretion assay
DNA transfection and GM-CSF assays were performed as previously
described (18). Briefly, 200 ng of cDNA pools and 50 ng of the
HLA-A31 DNA were mixed with 2 µl of Lipofectamine in 100
µl of serum-free DMEM for 15 to 45 min. The DNA/Lipofectamine mixture
was then added to the COS-7 (5 x 104) cells and
incubated overnight. The following day, cells were washed twice with
AIM-V medium. CTL clones 5 or 10 were added at a concentration of
5 x 104 cells/well in AIM-V medium containing 120
IU/ml of IL-2. After 18 to 24 h of incubation, 100 µl of
supernatant was collected and GM-CSF concentration was measured in a
standard ELISA assay (R&D Systems, Minneapolis, MN). For testing
peptide recognition, 586EBV, 1510EBV, or T2 cells were incubated with
peptides at 37°C for 90 min, and then washed three times with AIM-V
medium containing 120 IU/ml of IL-2. T cells were added and incubated
for an additional 18 to 24 h; 100 µl of supernatant was
collected for the GM-CSF assay. In some experiments, IFN-
release
assays were done using a standard ELISA kit (Endogen, Woburn, MA).
Northern blot analysis
Total RNA was isolated by the guanidine isothiocyanate/cesium
chloride centrifugation method. Total RNA from normal human tissue was
purchased from Clontech (Palo Alto, CA). Twenty micrograms of total RNA
was subjected to electrophoresis in a 1.2% formaldehyde agarose gel
and transferred to a nylon membrane. An 0.8-kb DNA fragment of the
NY-ESO-1 gene was labeled with
[
-32P]CTP by the random priming method.
Prehybridization and hybridization were performed according to the
QuickHyb protocol (Stratagene, La Jolla, CA). Membranes were washed
twice with 2x SSC/0.1% SDS at room temperature for 15 min and twice
with 0.1x SSC/0.1% SDS at 60°C for 30 min. The autoradiography was
performed at -70°C.
Reverse transcriptase-PCR
Total RNA was extracted from tumor cell lines as described above. Five hundred nanograms of total RNA was used for conversion of RNA to cDNA by avian myeloblastosis virus (AMV) reverse transcriptase. cDNA was then amplified by PCR using the ESO-P2 (5'-GCGGCTTCAGGGCTGAATGGATG) and ESO-P5 (5'-AAGCCGTCCTCCTCCAGCGACA) primers and One-Step RT-PCR system (Life Technologies). PCR products were amplified under denaturation conditions at 94° for 30 s, annealing at 55° for 30 s, extension at 72° for 3 min for 40 cycles, and final elongation at 72° for 10 min. PCR products were analyzed on a 3% agarose gel.
Cytotoxicity assays
Cytolytic assays were performed as previously described (24). Briefly, the target cells were labeled with 51Cr for 90 min. After washing three times, the cells were incubated with peptides at a concentration of 1 µg/ml for 90 min. The cells were washed again, counted, and then mixed with CTL clone 5 at the indicated E:T ratio. Chromium release was measured after a 4-h incubation. For titration of the peptides recognized by CTL clone 5, 586EBV B cells were incubated with various concentrations of the purified peptides. Percentage of specific lysis was determined from the equation: (A - B)/(C - B) x 100, where A is the lysis of 586EBV B cells by CTL clone 5 in the presence of a peptide, B is the spontaneous release from 586EBV B cells in the presence of the same peptide but in the absence of effector cells, and C is the maximum chromium release.
The peptides were synthesized by a solid phase method using a peptide synthesizer (Model AMS 422, Gilson, Worthington, OH). Some peptides were purified by HPLC and had >98% purity. The mass of some peptides was confirmed by mass spectrometry analysis.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The TIL586 cell line with antitumor activity in vivo had
previously been shown to recognize the tumor Ag gp75/TRP-1 (18).
Tumor-reactive CTL clones isolated from TIL586 by limiting dilution
(24) appeared to recognize several distinct Ags. The first group of CTL
clones recognized the ORF3P peptide derived from a gene product
translated from an alternative open reading frame of the
TRP-1/gp75 gene (24). The second group of CTL clones
recognized TRP-2 as a tumor Ag in the context of HLA-A31 (19). A third
group of 14 CTL clones, recently isolated, recognized 586mel tumor
cells, but did not recognize normal HLA-A31+ melanocytes
nor COS-7 cells transfected with either TRP-1 or
TRP-2 (see below) (Fig. 1
).
These results suggested that these T cell clones recognized additional
tumor Ags expressed in 586mel tumor cells. Four CTL clones (clones 2,
5, 10, and 14) were expanded, and CTL clones 5 and 10 were used to
screen cDNA libraries.
|
To determine the HLA molecule responsible for presenting Ag to CTL
clones 5 and 10, we transfected HLA-A31 cDNA into HLA-A31-negative
tumor lines such as 397mel and 624mel and tested for T cell
recognition. Transfectants of 397mel and 624mel expressing HLA-A31 were
strongly recognized by CTL clones 5 and 10 (Table I), suggesting that T cell recognition
was restricted by HLA-A31.
|
Isolation of a gene encoding a tumor Ag
To identify the new tumor Ag(s) recognized by CTL clones 5 and 10,
we made a cDNA library derived from 586mel. Each cDNA pool consisted of
100 cDNA clones. After screening a total of 2.5 x
105 cDNA clones, we identified 15 positive cDNA pools that
conferred T cell recognition by CTL 5 or 10 when cotransfected into
COS-7 along with HLA-A31 cDNA. The positive clones were then
tested for recognition by both CTL clones 5 and 10. It was found that
both CTL clones recognized the same cDNA pools. Individual colonies
were isolated from each positive pool and tested for T cell reactivity.
Representative data are shown in Figure 2. CTL clone 5 recognized COS-7
cotransfected with cDNA clone 1 or 2 and HLA-A31, but not
COS-7 alone, COS-7 transfected with cDNA clone 1 or 2, or transfected
only with the HLA-A31 cDNA (Fig. 2). DNA sequencing analysis
indicated that all 10 of the cDNA-positive clones from different
positive pools overlapped; the DNA and the amino acid sequence of these
clones are shown in Figure 3. A search of
all available databases revealed that the coding region of this gene,
which we named cancer antigen gene 3 (CAG-3; GenBank
accession no. AF038567), was identical to NY-ESO-1, which
was recently reported to be an Ag recognized by a serum Ab derived from
a patient with esophageal cancer (29). Our longest cDNA clone contained
an additional 37 nucleotides upstream of the previously reported 5'-end
untranslated region. Two other proteins in the databases were found to
contain homologous sequences in a limited region. The gene product of
NY-ESO-1 (we have used NY-ESO-1 for CAG-3 in the following text) has
52% similarity to the tegument protein (UL36) of herpes simplex virus
type 1 in the 64-aa segment and 47% similarity to enterobactin
synthetase component F (serine-activating enzyme) in the 48-amino acid
region (Fig. 3B).
|
|
Northern blot analyses were performed using NY-ESO-1cDNA as a probe to evaluate the expression pattern in different
tissues. Testis tissue was shown to be the only positive in the
expression of NY-ESO-1 among the normal human tissues
tested. NY-ESO-1 was found to be expressed in several types
of cancers including melanoma and breast cancer (data not shown). These
results were consistent with previously reported data (29). To
determine whether the melanoma-reactive CTLs would also recognize other
tumor cells, we tested T cell reactivity against HLA-A31-positive
breast and prostate tumor cells by CTL clone 5. We used IFN- to
monitor T cell recognition in these experiments, because prostate tumor
cells alone secret GM-CSF, but not IFN-. As shown in Table II, CTL clone 5 was capable of
recognizing HLA-A31-positive 1295Br and 1315Br fresh breast tumor
cells, but neither HLA-A31-negative 1405Br and 1411Br fresh breast
tumor cells nor HLA-A31-positive 1295 fibroblast cells derived from the
autologous patient 1295. In addition, CTL clone 5 recognized the
cultured HLA-A31-positive 1315Br (culture A and B) cells (Table II),
but did not respond to the cultured HLA-A31-negative cultured 1398Br
breast cancer cells nor to the cultured HLA-A31-positive 1510
fibroblast. Although CTL clone 5 somehow did not respond to the
cultured HLA-A31-positive 1315Br (culture A and B) cells in experiment
1, additional experiments showed T cell recognition of the cultured
HLA-A31-positive 1315BR (culture A and B) cells (data not shown). The
expression of HLA-A31 and NY-ESO-1 in the fresh 1315Br and 1295Br tumor
cells was confirmed by FACS analysis and RT-PCR analysis (Fig. 4). CTL clone 5 recognized neither the
HLA-A31+ 1535 prostate tumor cells, because of lack of
expression of NY-ESO-1, nor the HLA-A31- 1542 prostate
tumor cells (Table II). These results strongly suggest that an
antigenic peptide of NY-ESO-1 was expressed at sufficient levels on the
surface of breast tumor cells to be recognized by T cells. Therefore,
NY-ESO-1 may serve as an immune target for the immunotherapy of
patients with breast cancer.
|
|
To determine the antigenic epitopes recognized by
NY-ESO-1-reactive CTL, we made a series of synthetic peptides based on
the peptide-binding motif for HLA-A31, which consists of hydrophobic
residues at position 2 and positively charged residues at position 9
(33). These peptides were pulsed onto HLA-A31-positive 1510 EBV B cells
and tested for their ability to stimulate cytokine release by CTL clone
5. As shown in Table III, the 10-mer
peptide ESO10–53 (ASGPGGGAPR), starting at position 53 of the NY-ESO-1
protein, was strongly recognized by CTL clone 5, while the overlapping
9-mer peptides, ESO9-54 as well as ESO10–127, were weakly
recognized when pulsed onto 1510EBV B cells. CTL clone 10 recognized
the same peptide as CTL clone 5 (data not shown). Interestingly, CTL
clone 2 did not recognize any of these peptides (Table III) even though
it recognized 586mel and COS-7 transfected with NY-ESO-1 (see below).
The reactivity of CTL clone 5 was undetectable when either the ESO9-54
or the ESO10–127 peptides were used at concentrations below 100 nM to
sensitize EBV cells. To examine whether CTL clone 5 also recognized
peptides that contained the core amino acid sequence with the extension
of amino acid residues at either the N or C terminus, we made
overlapping 11-mer, 12-mer, 13-mer, 14-mer, and 15-mer peptides, as
well as several peptides containing substitutions at either position 1,
2, or 10 of ESO10–53 (Table IV and
Figure 3
). CTL clone 5 was capable of recognizing 11-mer, 12-mer,
13-mer, 14-mer, and 15-mer peptides with amino acid extensions at the N
terminus of the ESO-53 core peptide, although the longer peptides
appeared to stimulate significantly less GM-CSF secretion than did the
ESO10–53 10-mer peptide (Table IV). However, an extension of only a
single amino acid residue at the C terminus of ESO10–53 abrogated its
ability to stimulate T cells (Table IV). CTL clone 5 did not recognize
the 8-mer peptide.
|
|
A). In addition to measuring
cytokine release stimulated by ESO10–53, the ability of this peptide
to sensitize target cells for lysis by CTL clone 5 was examined. CTL
clone 5 was capable of lysing HLA-A31-positive tumor 586mel, but not
the HLA-A31-negative and NY-ESO-1-positive tumor 397mel (Fig. 5B). CTL clone 5 lysed >60% of either 586EBV or 1510EBV B
cells that had been incubated with the ESO10–53 peptide at an E:T
ratio of 40:1, and 5 to 10% lysis of target cells was observed at
an E:T ratio of 2.5:1. CTL clone did not lyse either 586EBV or 1510EBV
B cells alone or pulsed with an irrelevant peptide, nor did it lyse the
HLA-A31-negative T2 cells pulsed with the ESO10–53 peptide (Fig. 5C).
|
). The modified 10-mer peptides with a substitution at
position 2 derived from the wild-type ASGPGGGAPR were still recognized
by CTL clone 5 when pulsed on 586EBV B cells. The reactivity of
peptides containing a substitution of either Ala, Ile, Leu, or Val at
position 2 was lower than that of the wild-type peptide, while one
peptide containing a substitution of Thr for Ser at position 2 resulted
in a slightly higher reactivity than the wild-type ESO10–53 peptide.
In contrast, peptides containing substitutions of Arg with Lys or His
completely lost their ability to stimulate T cells, suggesting that the
Arg at the C terminus of the ESO10–53 peptide represents a critical
anchor residue. Peptides with a substitution at position 1 were
recognized poorly or not recognized at all by CTL clone 5 (Table IV).
These results indicate that the ESO-53 peptide, ASGPGGGAPR, represents
the best peptide for T cell recognition. Antigenic peptides derived from an alternative open reading frame
Two additional CTL clones, clones 2 and 14, appeared to recognize
586mel as well as COS-7 cells transfected with NY-ESO-1 and HLA-A31
cDNA, but failed to recognize the ESO10–53 peptide (Fig. 6). CTL clone 5 and TIL1244 were used for
the specificity controls. Additional experiments showed that CTL clone
2 did not respond to any of 19 other peptides containing the HLA-A31
binding motif derived from the normal open reading frame of NY-ESO-1
(Table III). To test the hypothesis that CTL may recognize a peptide
from a gene product translated from an alternative open reading frame
of the same gene, we made synthetic peptides with HLA-A31 binding motif
on the basis of an amino acid sequence predicted from the second open
reading frames (ORF2) (Fig. 3). Strikingly, CTL clone 2 recognized
ESORF2-9–19 (AAQERRVPR) as well as the overlapping ESORF2-10–18
(LAAQERRVPR) peptides when pulsed onto 1510EBV B cells. Representative
data for CTL clone 2 are shown in Figure 7A. CTL clone 14 recognized
the same peptides as CTL clone 2 (data not shown). These results
suggest that CTL clones 2 and 14 recognized an antigenic peptide
derived from the ORF2 (Fig. 3). A protein database search revealed that
the 58-aa protein of ORF2 has a 52% similarity to the chain A
of glutamate dehydrogenase in a 25-aa region (34). Peptide titration
experiments demonstrated that CTL clone 2 was capable of lysing 1510EBV
pulsed with ESORF2-10–18 (LAAQERRVPR) at relatively low concentrations
of peptide, but failed to lyse 1510EBV pulsed with ESO10–53 or
HLA-A31-negative 1102EBV pulsed with ESORF2-10–18 (Fig. 7B). In addition, CTL clone 2 also recognized overlapping
11-mer, 12-mer, and 13-mer peptides with amino acid extensions at the N
terminus of the ESORF2-10–18 peptide at relatively high concentrations
(data not shown).
|
|
C, the recognition pattern of CTL clone
2 was similar to that of CTL clone 5 on tumor cells. CTL clone 2
recognized the HLA-A31-positive fresh 1315Br and 1295Br breast tumor as
well as 586mel and 1388mel, but did not recognize HLA-A31-negative
fresh 1411Br breast tumor, 397mel, nor the HLA-A31-negative 1295
fibroblast. Although 1353mel expresses HLA-A31, neither CTL clone 2 nor
clone 5 responded to 1353mel because 1353mel is an NY-ESO-1-negative
tumor. As previously demonstrated, CTL clone 5 recognized the ESO10–53
ASGPGGGAPR peptide and CTL clone 2 recognized the ORF2-10–18
LAAQERRVPR peptide derived from the ORF2 following incubation with
1510EBV B cells (Fig. 7C). These results strongly suggest
that the ORF2 gene product was translated, processed, and presented in
melanoma as well as breast tumors. Therefore, we concluded that
NY-ESO-1 encode two different proteins: a protein with 180
aa (ORF1) recognized by CTL clones 5 and 10; and a protein with 58 aa
(ORF2) recognized by CTL clones 2 and 14. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Five differentiation Ags including tyrosinase, MART-1, gp100, TRP-1/gp75, and TRP-2 have been identified as melanoma Ags recognized by T cells derived from TILs (13, 14, 15, 16, 17, 18, 19), which had been shown to be associated with antitumor reactivity in vivo. Thus far, no TILs available in the Surgery Branch at National Cancer Institute recognize MAGE-1, BAGE, and GAGE, which are expressed only in the testis and in cancer cells. Here we show that NY-ESO-1, another cancer-shared Ag, is a tumor Ag recognized by HLA-A31-restricted T cells. NY-ESO-1 was independently isolated using the autologous serum from a patient with esophageal cancer (29), suggesting that the NY-ESO-1 gene product was an immune target for Ab-mediated immunity. Results from this study provide evidence that NY-ESO-1 is also an immune target recognized by T cells. This is further supported by a recent report that HLA-A2-restricted CTL established in a melanoma patient recognized NY-ESO-1 (35). Several tumor Ags, including MAGE-1, tyrosinase, and TRP-1, recognized by CTL have been found to be reactive with Ab as well (36, 37). Since NY-ESO-1 is not expressed in normal human tissues except the testis, which does not express MHC class I molecules and is considered as an immunologically privileged site, this gene product may constitute a safe immune target for the immunotherapy of patients with cancer.
NY-ESO-1 is a breast cancer Ag
Based on its gene expression pattern, NY-ESO-1 belongs to a
member of an expanding family of Ags including MAGE-1, MAGE-3, BAGE,
GAGE, and HOM-MEL-40 (9, 10, 11, 12, 30). However, NY-ESO-1 is highly expressed
in a significant proportion of breast, prostate, and bladder cancers
(29) as compared with MAGE, BAGE, and GAGE. More importantly, CTL
clones recognized two HLA-A31-positive fresh and cultured breast cancer
cells (Table II and Fig. 7C). To our knowledge, this is the
first demonstration of CTL recognition of NY-ESO-1 positive breast
cancer cells. Although the expression of MAGE-1, MAGE-3, and others
were reported to be detected by RT-PCR in breast tumors at a low
frequency (<5–10%), CTL recognition of these breast tumors by the
Ag-specific CTL has not been documented. It has been difficult to
generate breast-reactive CTLs from PBL in vitro, although
MHC-restricted T cells that recognized HER-2/neu peptides on breast
cancer cells have been reported (26, 27, 28). Identification of NY-ESO-1
peptides presented on the cell surface of breast cancers is important
for the development of Ag-specific cancer vaccines for the treatment of
patients with breast cancer. The CTL cloning approach described here
represents a strategy for the isolation of breast cancer Ags.
Translation of different ORFs as a mechanism for generating T cell epitopes
To define antigenic peptides, we found the 10-mer ASGPGGGAPR derived from the NY-ESO-1 protein as the best antigenic peptide recognized by CTL clone 5 although the 9-mer, 11-mer, 12-mer, 13-mer, 14-mer, and 15-mer peptides were also recognized. This reactivity may be due to the presence of two proline residues in these peptides. Proline residues in the core peptide sequence may allow the peptides to bulge out of the MHC binding pocket, and thus the anchored residues in the longer peptides can still fit into the HLA-A31 molecule. An alternative explanation is that the longer peptides may be processed to the shorter peptides by extracellular or serum proteases (38). However, when the longer peptides were pulsed onto 586EBV B cells in serum-free conditions, they were still recognized by CTL clone 5 (data not shown). This experiment, however, does not exclude the possibility that these longer peptides were processed by extracellular proteases. The modified peptide, ATGPGGGAPR, with a substitution of Thr for Ser appeared to slightly improve T cell recognition. It has been shown that modified peptides could improve binding affinity to the MHC class I molecule, thus enhancing the immunogenicity of peptides (39). It is not clear why CTL clone 5 recognized the unrelated ESO10–127 peptide, but its recognition was weak and could only be detected at a relatively high peptide concentration.
Interestingly, two additional CTL clones recognized COS-7 transfected
with NY-ESO-1 plus HLA-A31 cDNA, but not the ESO10–53 peptide, which
was derived from ORF1 of the NY-ESO-1 gene (Fig. 6). Further analysis
showed that CTL clone 2 recognized a peptide from the gene product
translated from an alternative open reading frame (ORF2) of NY-ESO-1.
Although there are several examples that two different proteins are
translated from different ORFs from a single viral mRNA (40), very few
cases have been reported in single eukaryotic mRNA. Two examples have
been reported: TRP-1/gp75, which encodes two different
proteins, gp75 recognized by sera from a patient with melanoma and a
24-aa gene product recognized by CTLs (24, 37); and 43 kDa-Ag
recognized by squamous cell carcinoma-reactive CTL (41). Interestingly,
the cyclin-dependent kinase inhibitor p16 gene also encodes
two gene products, p16 and p19ARF. However, the
alternative ORF gene product, p19ARF, is
translated from an alternatively spliced transcript that shares
an identical sequence with the p16 transcript except the first
exon (41). Recent studies showed that the alternative reading frame
product p19ARF plays a role in cell cycle
regulation and tumor suppression (42). The NY-ESO-1 gene
reported here encodes two different proteins: a 180-aa protein encoded
by ORF1 and a 58-aa protein encoded by ORF2. The ORF2 product is
located inside ORF1. Strikingly, CTL clones 5 and 10 recognized
antigenic peptides derived from ORF1, while CTL clones 2 and 14
recognized an antigenic peptide from ORF2. Both CTL clones 2 and 5
recognized several HLA-A31+ melanomas as well as fresh
breast tumors (Table II and Fig. 7C), suggesting that the
translation of the alternative ORF2 may serve as a general mechanism in
vivo for generating T cell epitopes.
The mechanism by which the alternative ORF is translated is currently not known. However, there are several potential explanations for the production of alternative reading frames. The leaky-scanning model is one of them: ribosomes occasionally bypass the first AUG with a poor KOZAK consensus sequence and initiates translation at a downstream AUG (43, 44). In the case of NY-ESO-1, the start codon for translation of ORF1 is not in the optimal context. There are four potential start codons that could be used to translate the ORF2. In our previous report, a CTL epitope was translated from an alternative reading frame of gp75/TRP-1 (24). Recognition of the epitope of ORF3 by CTL was affected by the presence of the first ATG codon used for the translation of the gp75 protein and was completely abolished when the internal ATG preceding the T cell epitope (ORF3P) was changed to ATC, suggesting that ribosomal scanning may be a possible mechanism. In the studies of the influenza nucleoprotein, CTL epitopes were produced via a ribosomal scanning mechanism (44). However, in a similar study, ribosomal frameshifting was suggested as a mechanism for the production of T cell epitopes (45). Other mechanisms have also been proposed for the production of T cell epitopes derived from different genes. For example, several T cell epitopes were identified from gene products translated from a cryptic initiation codon of the 5' untranslated region (46, 47), from a non-ATG-defined open reading frame (48), and from introns of transcript (22, 49, 50). Since both the alternative open reading frames of gp75/TRP-1 and NY-ESO-1 are located within the primary open reading frame, it is of particular interest to understand the underlying mechanism.
Although there are only a few examples of the usage of the alternative open reading frames in eukaryotes reported in the literature, we believe that more examples will be reported in the future when tumor-reactive CTL and autoantibodies are available and used to identify target proteins or peptides. Therefore, it is important to understand the biologic significance of the gene products translated from alternative open reading frames. One possibility is that these gene products serve as antigenic targets of the Ag processing machinery to increase the efficiency and capacity of the immune surveillance. Identification of T cell epitopes derived from different open reading frames of NY-ESO-1 suggests that the identity of immunogenic peptides for cancer vaccines may not be limited only to peptides derived from the primary open reading frame. It is not clear at the present time whether the ORF2 gene product of NY-ESO-1 and the ORF3 gene product of gp75/TRP-1 have biologic functions in addition to immune responses of T cells. Further studies are needed to characterize the biologic function for both the ORF1 and ORF2 products of NY-ESO-1.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 S.L.J. and G.Z. contributed equally to this work.
3 Abbreviations used in this paper: TIL, tumor-infiltrating lymphocyte; CAG-3, cancer Ag gene 3; ORF, open reading frame; GM-CSF, granulocyte/macrophage-CSF; mel, melanoma tumor cell; TRP, tyrosinase-related protein.
Received for publication April 2, 1998. Accepted for publication June 2, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  I. D. Davis, W. Chen, H. Jackson, P. Parente, M. Shackleton, W. Hopkins, Q. Chen, N. Dimopoulos, T. Luke, R. Murphy, et al. Recombinant NY-ESO-1 protein with ISCOMATRIX adjuvant induces broad integrated antibody and CD4+ and CD8+ T cell responses in humans PNAS, July 20, 2004; 101(29): 10697 - 10702. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. S. Voo, T. Fu, H. Y. Wang, J. Tellam, H. E. Heslop, M. K. Brenner, C. M. Rooney, and R.-F. Wang Evidence for the Presentation of Major Histocompatibility Complex Class I-restricted Epstein-Barr Virus Nuclear Antigen 1 Peptides to CD8+ T Lymphocytes J. Exp. Med., February 17, 2004; 199(4): 459 - 470. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. S. Voo, T. Fu, H. E. Heslop, M. K. Brenner, C. M. Rooney, and R.-F. Wang Identification of HLA-DP3-restricted Peptides from EBNA1 Recognized by CD4+ T Cells Cancer Res., December 15, 2002; 62(24): 7195 - 7199. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Y. Wang, J. Zhou, K. Zhu, A. I. Riker, F. M. Marincola, and R.-F. Wang Identification of a Mutated Fibronectin As a Tumor Antigen Recognized by CD4+T Cells: Its Role in Extracellular Matrix Formation and Tumor Metastasis J. Exp. Med., June 3, 2002; 195(11): 1397 - 1406. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. A. Rosenberg, P. Tong-On, Y. Li, J. P. Riley, M. El-Gamil, M. R. Parkhurst, and P. F. Robbins Identification of BING-4 Cancer Antigen Translated From an Alternative Open Reading Frame of a Gene in the Extended MHC Class II Region Using Lymphocytes From a Patient With a Durable Complete Regression Following Immunotherapy J. Immunol., March 1, 2002; 168(5): 2402 - 2407. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Probst-Kepper, V. Stroobant, R. Kridel, B. Gaugler, C. Landry, F. Brasseur, J.-P. Cosyns, B. Weynand, T. Boon, and B. J. Van den Eynde An Alternative Open Reading Frame of the Human Macrophage Colony-Stimulating Factor Gene Is Independently Translated and Codes for an Antigenic Peptide of 14 Amino Acids Recognized by Tumor-Infiltrating Cd8 T Lymphocytes J. Exp. Med., May 21, 2001; 193(10): 1189 - 1198. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Kawakami, X. Wang, T. Shofuda, H. Sumimoto, J. P. Tupesis, E. Fitzgerald, and S. A. Rosenberg Isolation of a New Melanoma Antigen, MART-2, Containing a Mutated Epitope Recognized by Autologous Tumor-Infiltrating T Lymphocytes J. Immunol., February 15, 2001; 166(4): 2871 - 2877. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Valmori, V. Dutoit, D. Liénard, D. Rimoldi, M. J. Pittet, P. Champagne, K. Ellefsen, U. Sahin, D. Speiser, F. Lejeune, et al. Naturally Occurring Human Lymphocyte Antigen-A2 Restricted CD8+ T-Cell Response to the Cancer Testis Antigen NY-ESO-1 in Melanoma Patients Cancer Res., August 1, 2000; 60(16): 4499 - 4506. [Abstract] [Full Text]
|
|
|
|
|
|
G. Zeng, C. E. Touloukian, X. Wang, N. P. Restifo, S. A. Rosenberg, and R.-F. Wang Identification of CD4+ T Cell Epitopes from NY-ESO-1 Presented by HLA-DR Molecules J. Immunol., July 15, 2000; 165(2): 1153 - 1159. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. J. Van den Eynde, B. Gaugler, M. Probst-Kepper, L. Michaux, O. Devuyst, F. Lorge, P. Weynants, and T. Boon A New Antigen Recognized by Cytolytic T Lymphocytes on a Human Kidney Tumor Results from Reverse Strand Transcription J. Exp. Med., December 20, 1999; 190(12): 1793 - 1800. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. van Baren, F. Brasseur, D. Godelaine, G. Hames, A. Ferrant, F. Lehmann, M. Andre, C. Ravoet, C. Doyen, G. C. Spagnoli, et al. Genes Encoding Tumor-Specific Antigens Are Expressed in Human Myeloma Cells Blood, August 15, 1999; 94(4): 1156 - 1164. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L.-Q. Huang, F. Brasseur, A. Serrano, E. De Plaen, P. van der Bruggen, T. Boon, and A. Van Pel Cytolytic T Lymphocytes Recognize an Antigen Encoded by MAGE-A10 on a Human Melanoma J. Immunol., June 1, 1999; 162(11): 6849 - 6854. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Wang, X. Wang, A. C. Atwood, S. L. Topalian, and S. A. Rosenberg Cloning Genes Encoding MHC Class II-Restricted Antigens: Mutated CDC27 as a Tumor Antigen Science, May 21, 1999; 284(5418): 1351 - 1354. [Abstract] [Full Text]
|
|
|
|
|
|
R.-F. Wang, X. Wang, and S. A. Rosenberg Identification of a Novel Major Histocompatibility Complex Class II-restricted Tumor Antigen Resulting from a Chromosomal Rearrangement Recognized by CD4+ T Cells J. Exp. Med., May 17, 1999; 189(10): 1659 - 1668. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Zeng, X. Wang, P. F. Robbins, S. A. Rosenberg, and R.-F. Wang 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 PNAS, March 27, 2001; 98(7): 3964 - 3969. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
