| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Immunohematology and Blood Transfusion, Leiden University Medical Center, Leiden, The Netherlands
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
- and ß-chains, which proved to be identical. Our
research shows that all sources of RBL-5, RMA, RMA-S, MBL-2, and EL-4
tumors are derived from a single tumor line, most likely
EL-4. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CTL epitopes presented by murine tumors that are induced by viruses or viral oncogenes are generally virus encoded (4). Next to these immunogenic epitopes four additional classes of tumor-specific T cell epitopes have been described. First, several carcinogen-induced murine tumors were shown to express unique CTL epitopes that arose from point mutations in cellular genes. The critical event was not a change in the expression of these genes, but substitution of an amino acid in a peptide that resulted in increased MHC binding and/or T cell recognition. Examples of such CTL epitopes have been found at the surface of the murine mastocytoma line P815 and its mutagenized variants (5). Second, experiments with the P815 system revealed that ectopic expression of cellular genes could lead to presentation of epitopes derived from these genes (P815A/B Ag) (6). By now, many other members of this so-called testis cancer family of tumor Ags have been identified in human tumors (7). A third class of tumor-specific CTL epitopes commonly found in murine tumors is encoded by (re)activated endogenous retroviral sequences that are integrated in the cellular genome. For instance, the widely used EL-4 lymphoma expresses a CTL epitope derived from an endogenous mouse mammary tumor virus (MMTV)2 envelope protein (8), whereas the C26 colon carcinoma was found to express an epitope encoded by an endogenous murine leukemia virus (MuLV) (9). Induced expression of these sequences appears to be the result of DNA demethylation (10, 11). A fourth category of tumor-associated epitopes, which has originally been defined in human melanomas, concerns tissue lineage-specific Ags such as the melanocyte-specific Ags (12). A murine counterpart has been found in the B16 melanoma (13).
MuLV are naturally occurring retroviruses that induce hematological tumors in mice. The crucial role of CTL as well as Th cells in the rejection of MuLV-induced tumors has been firmly established (14). Widely used tumor cell lines transformed by antigenically related MuLV are the Friend MuLV-induced erythroleukemia FBL-3, the Moloney MuLV-induced T cell lymphoma MBL-2, and the Rauscher MuLV-induced T cell lymphoma RBL-5 and its derivatives RMA and RMA-S. For these tumors, two CTL epitopes and one Th epitope were molecularly identified and were all found to be of viral origin (15, 16, 17). Interestingly, the CTL response against the MBL-2 tumor was shown to include an additional specificity directed to a nonviral Ag (18). Although the existence of this nonviral tumor epitope was postulated in the eighties, its identity remained an enigma despite the fact that several laboratories have since attempted to identify this epitope. In the present paper we describe the molecular identification of this epitope that is selectively expressed by certain T cell lymphomas. In addition, we provide evidence that these tumor cell lines originate from the same clonal cell line.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All cell lines used in this study were derived from C57BL/6 (B6, H-2b) mice. RMA and RMA-S cell lines are T cell lymphomas derived from the Rauscher MuLV-induced RBL-5 cell line (19). MBL-2 and FBL-3 are isolated from a Moloney and a Friend MuLV inoculated C57BL/6 mouse, respectively (20). EL-4 is a dimethylbenzanthracene-induced thymoma cell line (21). Transformed mouse embryo cells are primary embryonic cells transfected with plasmids encoding activated oncogenes (22). B16 melanoma cells and MCF1233 MuLV-induced B cell lymphomas 771 and 786 have been described previously (13, 23). The Moloney MuLV-induced pro-B cell lymphomas 33H2 and 33A3 were provided by Dr. M. Schilham (Leiden University Medical Center, Leiden, The Netherlands). FRE.Db and FRE.Kb are stable transfectants of the Fisher rat embryo cell line. HeLa.Db and T2.Db cells are stable H-2Db transfectants of the human cervical carcinoma cell line HeLa and T cell hybrid T2, respectively. Freshly isolated sarcomas were obtained by i.m. injections of the Moloney murine sarcoma and leukemia virus complex as previously described (24). All cell lines were cultured in IMEM (BioWhittaker Europe, Verviers, Belgium) supplemented with 8% heat-inactivated FCS (Life Technologies, Breda, The Netherlands), 2 mM L-glutamine (ICN Biomedicals, Costa Mesa, CA), 100 IU/ml penicillin (Yamanouchi Pharma, Leiderdorp, The Netherlands), and 30 µM 2-ME (Merck, Darmstadt, Germany) at 37°C in humidified air with 5% CO2.
C57BL/6 mice were bred and obtained from the TNO-PG breeding facility (Leiden, The Netherlands). C57BL/6 nu/nu mice were obtained from IFFA Credo (Zeist, The Netherlands). All mice were kept under specific pathogen-free conditions in the animal facility of Leiden University Medical Center.
Generation and analysis of CTL clones
The env-specific CTL clone 10B6 was generated by immunization of
a B6.CH-2bm13 mouse with Moloney MuLV; it
recognizes the SSWDFITV epitope presented by
H-2Kb (env; aa 189–196) as described previously
(15). CTL clones specific for the tumor epitope and the
CCLCLTVFL epitope (gag leader; aa 75–83), presented by
H-2Db, were derived from spleen cells of C57BL/6
(B6) mice immunized with irradiated RMA tumor cells. Bulk cultures were
restimulated weekly with irradiated RMA cells and irradiated naive
spleen cells as feeders in complete culture medium supplemented with
2% (v/v) conditioned medium from Con A- and PMA-stimulated rat
splenocytes. After 3–5 wk the conditioned medium was gradually
replaced by 10 Cetus units of recombinant human IL-2 (Eurocetus,
Amsterdam, The Netherlands). CTL clones were obtained by standard
limiting dilution and were weekly restimulated with RMA cells, feeders,
and IL-2. All obtained CTL clones expressed ß-TCR and CD8 as
tested by flow cytometry.
The cytolytic activity of CTL clones was measured by means of a
chromium (51Cr) release assay, as described
previously (14). The mean percent specific lysis of
triplicate wells was calculated as follows: % specific lysis =
([cpm experimental release - cpm spontaneous release]/[cpm
maximum release - cpm spontaneous release]) x 100.
Measurement of secreted TNF- by stimulated CTL was performed with a
bioassay, using WEHI 164 clone 13 cells, as described previously
(25). The percentage of TNF- released in triplicate
wells was calculated as follows: % TNF- release =
[(A550–650 experimental wells
- A550–650 wells containing medium
only)/(A550–650 wells containing 500
pg/ml TNF- - A550–650 wells
containing medium only)] x 100.
Moloney virus infection
In vitro virus infections were performed with Abelson Moloney virus collected from supernatant of NIH-3T3 Abelson virus nonproducer cells (ANN-1), which were productively infected with cloned Moloney MuLV (26). Batches of virus-containing supernatants were collected after 24 h. Virus batches were stored at -80°C. FRE transfectants were infected with Abelson Moloney MuLV by culturing the cells for 3–4 days in complete culture medium containing 50% (v/v) virus supernatant from ANN-1 cells with 10 µg/ml polybrene (Sigma-Aldrich, Zwyndrecht, The Netherlands).
MHC class I-peptide binding assay
The binding capacity of peptides to H-2Db was determined using the RMA-S binding assay as previously described (27). Briefly, RMA-S cells were cultured for 36 h at 26°C and were added to serial dilutions of peptide. After 4 h of incubation at 37°C, cells were washed and stained with the mAb 28.14.8S specific for H-2Db and FITC-labeled goat-anti-mouse Ig. Fluorescence was determined using a FACScan cytometer (Becton Dickinson, Mountain View, CA). The fluorescence index was calculated as follows: FI = (mean fluorescence with peptide/mean fluorescence without peptide).
RT-PCR of TCR genes
Total RNA from 107 tumor cells was
isolated using TRIzol according to the manufacturer’s recommendation
(Life Technologies). cDNA was generated by oligo(T)-primed RNA using
AMV reverse transcriptase (Promega, Madison, WI). The reaction was heat
inactivated, diluted in water, and stored at 20°C until usage. PCR
incubation temperatures were 95, 58, and 72°C subsequently (1 min for
all steps, 30 cycles). Primer sequences for determination of TCR V
and Vß gene usage were previously published (28). The
downstream primer used for the TCR C region was 5'-TGG CGT TGGTCT
CTT TGA AG-3', resulting in a product of 
400 bp, and the primer for
the Cß region was 5'-CTT GGG TGG AGT CAC ATT TCT C-3', resulting in a
product of 200 bp. PCR products were directly cloned using a TOPO TA
cloning kit (Invitrogen, San Diego, CA) and were sequenced using
standard procedures.
Purification of eluted peptides
Peptides were eluted out of purified H-2Db or H-2Kb molecules as previously described (29). Briefly, MHC class I molecules were purified by affinity chromatography with 28-14-8S-coupled (Db-specific Ab) or B8-24-3-coupled (Kb-specific Ab) cyanogen bromide-activated Sepharose 4B beads (Pharmacia LKB, Uppsala, Sweden). Peptides were eluted by acid treatment and were separated from the heavy chains and ß2-microglobulin by filtration over a 10-kDa cutoff Centricon filter (Amicon, Lexington, MA). Peptides were fractionated using reverse phase micro C2C18 HPLC (Smart System, Pharmacia). Buffer A was 0.1% trifluoroacetic acid in water; buffer B was 0.1% trifluoroacetic acid in acetonitrile.
Mass spectrometry
Electrospray ionization mass spectrometry was performed on a
hybrid quadrupole time-of-flight mass spectrometer (Q-TOF, Micromass,
Manchester, U.K.) equipped with an on-line nanoelectrospray interface
(capillary tip, 20 µm internal diameter x 90 µm outer
diameter) with an approximate flow rate of 250 nl/min. This flow was
obtained by splitting of the 0.4 ml/min flow of a conventional high
pressure gradient system using an Acurate flow splitter AC-400-VAR
(LC-Packings, Amsterdam, The Netherlands). Injections were made with a
dedicated micro/nano HPLC autosampler (FAMOS, LC-Packings). The
analytical HPLC column was packed with PEPMAP (15 cm x 75 µm,
5-µm particle size; LC-Packings). The gradient went from 10% B to
90% B in 30 min (A: 95/5/1, v/v/v, water/methanol/acetic acid; B:
10/90/1, v/v/v, water/methanol/acetic acid). Mass spectra were recorded
from a mass of 50–2000 Da every second with a resolution of 5000 FWHM.
The resolution allows direct determination of the monoisotopic mass,
also from multiple charged ions. In the MS/MS mode, ions were selected
with a window of 2 Da with the first quadrupole, and fragments were
collected with high efficiency with the orthogonal time-of-flight mass
spectrometer. The collision gas applied was argon (4 x
105 mbar), and the collision voltage was 30 V
(for similar procedures, see Ref. 29, 30).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The CTL response against MuLV-induced lymphomas of the Friend,
Moloney, and Rauscher (FMR) types in C57BL/6
(H-2b) is predominantly
H-2Db restricted (15, 24). A peptide
derived from the gag leader sequence was identified as the major target
epitope (gagL75–83)(16). In
addition, a subdominant H-2Kb-restricted response
was described toward a viral env-derived peptide,
env189–196 (15). By analysis of
several independent T cell cultures from RMA-immunized mice we isolated
CTL lines with the previously described
gagL75–83 specificity (Fig. 1
, A and B).
However, some RMA-specific cultures failed to recognize this epitope
and the env189–196 epitope (Fig. 1, C
and D). To determine whether these CTL recognized other
virus-encoded peptides we tested Moloney MuLV-infected
FRE.Db and FRE.Kb cells.
GagL75–83- and
env189–196-reactive CTL specifically recognized
MuLV-infected FRE.Db and
FRE.Kb cells, respectively (Fig. 1, F
and G), indicating proper MuLV infection as well as MHC
class I processing and presentation. In contrast, CTL that did not
recognize one of the defined viral epitopes (Fig. 1, C and
D) also failed to recognize Moloney MuLV-infected cells
(Fig. 1E), whereas RMA tumor cells were efficiently
recognized. In the remainder of this manuscript we will refer to these
CTL as anti-tumor CTL. Also, transiently expressed cDNAs encoding
viral gag, pol, or env genes in
HeLa.Db and HeLa.Kb cells
consistently failed to sensitize these cells for recognition by these
CTL, while the transfected cells were stimulatory for
anti-env189–196 and
anti-gagL75–83 CTL (data not shown). Taken
together, our data suggested that the anti-tumor CTL are directed
against a nonviral epitope expressed by RMA cells. Several groups have
reported CTL with comparable, unknown specificity (16, 31, 32), and the involved epitope has long been searched for. We
therefore set out to identify the cognate peptide of these
CTL.
|
Several CTL clones were derived from bulk cultures displaying the
anti-tumor reactivity. A large panel of syngeneic murine tumor
cells of different tissue origin was examined for recognition by these
clones in cytotoxicity assays and cytokine release assays. CTL clones
specific for the viral gagL75–83 peptide lysed
all cell lines containing the FMR MuLV, because this peptide sequence
is conserved in this MuLV type. This recognition was independent of the
lymphoid lineage of the transformed cell lines (Fig. 2B). MuLV-induced T cell
lymphomas (RMA and MBL-2), MuLV-induced pro-B cell lymphomas (33A3 and
33H2), and an MuLV-induced erythroleukemia (FBL-3) were all lysed,
whereas chemically induced EL-4 cells that do not express FMR MuLV Ags
were not lysed (Fig. 2B). In contrast, anti-tumor CTL
clones only recognized the T cell lymphomas RMA, MBL-2 and EL-4 (Fig. 2A), as well as the RBL-5 cell line, the parental cell of
RMA (data not shown). The B cell lymphomas and erythroleukemia cells
were not lysed (Fig. 2A). This recognition pattern suggested
that anti-tumor CTL recognize a nonviral Ag that is selectively
expressed by T cell lymphomas. In accordance with this idea, mouse
embryo cells transformed by a variety of oncogenes, B16 melanoma cells,
and MCF MuLV-induced B cell lymphomas 771 and 786 were not recognized
by the anti-tumor CTL (not shown). In addition, freshly isolated
sarcomas that were induced by i.m. inoculation of the Moloney
MuLV/Moloney sarcoma virus complex did not express the tumor epitope,
whereas these sarcomas did express the Moloney virus encoded epitopes
(data not shown). Finally, we tested syngeneic Con A-activated T cell
blasts as a source of nontransformed T cells. These cells were not
recognized by our anti-tumor CTL (data not shown). In conclusion,
our anti-tumor CTL recognize an epitope that is selectively
expressed on the widely used T cell lymphomas RMA/RBL-5, MBL-2, and
EL-4 and that is most likely of nonviral origin.
|
We set out to identify the epitope of the anti-tumor CTL
through biochemical purification and sequencing by mass spectrometry.
Peptides were eluted from immunopurified Db and
Kb molecules from 20 x 109
MBL-2 cells and fractionated by reverse phase HPLC. This showed that a
Db-binding peptide, present in a fraction eluting
from the column at a low acetonitrile concentration (15%), was
capable of sensitizing target cells for lysis by anti-tumor CTL
(data not shown). Subsequently, Db-binding
peptides from 120 x 109 RMA cells were
separated by reverse phase HPLC. Anti-tumor CTL selectively recognized
four fractions (Fig. 3), which were
pooled and subsequently separated by reverse phase HPLC using a
methanol gradient. A single fraction sensitized target cells for lysis
by the CTL (data not shown). In a third HPLC run, the active fraction
was further separated on a nano-LC column that was connected to a mass
spectrometer. Twelve fractions were collected in a 96-well plate, and
only one fraction was recognized by anti-tumor CTL. Nevertheless,
five detectable peptide masses were present in this fraction, and
collision spectra (MS/MS) of all these were recorded (data not shown).
Interpretation of one of the MS/MS spectra (Fig. 4), with an ion mass of 472.8
(2+), yielded the partial six-amino acid
C-terminal sequence (D/E)NA(K/Q)A(I/L). However, ambiguities were left
for the sequence of the amino acids at the N-terminus. Peptide mixes
were synthesized containing the sequence XXXENAKA(I/L) and
XXXENAQA(I/L), where several different amino acids were placed on the
first three positions (X). Testing these mixes with anti-tumor CTL
revealed that only a glutamine (Q) at position 7 resulted in strong
recognition. Further testing of peptides with various amino acids on
position 4 containing random amino acids on the first three positions
showed by far the strongest CTL response to the peptides with a
glutamic acid (E) on this position. In contrast, the order and
character of the first three amino acids did not have any impact on the
CTL recognition of the peptide. Comparison between the MS/MS spectra of
several active synthetic candidates and the MS/MS spectrum of the
eluted peptide determined in the active fraction (Fig. 4) led to
the following four coeluting peptides: NKGENAQAI, NKGENAQAL,
KNGENAQAI, and KNGENAQAL. These peptides comply very well with the
published binding motif for H-2Db
(33). Importantly, we found no matches for either of the
candidate sequences in the different available protein and DNA
databases. This indicates that the epitope recognized by our
anti-tumor CTL is most likely derived from an as yet unknown
gene.
|
|
Because biochemical analysis did not provide further indications
in favor of one of the peptide candidates, all four were tested in CTL
recognition and MHC class I binding assays. At high peptide
concentrations (high nanomolar range) no differences in lysis were
observed among the four peptides by anti-tumor CTL clones (Fig. 5A). However, when the
peptides were titrated, the recognition of the peptides carrying a
C-terminal leucine (L) decreased markedly faster than that of the
peptides with an isoleucine (I) at this position. The concentration
needed for half-maximal lysis of the L peptides was at least 10-fold
higher than the concentration needed for the I-containing peptides
(Fig. 5A). The variation at the N-terminus (NK/KN)GENAQAI or
(NK/KN)GENAQAL of the peptide did not influence recognition by the CTL.
Peptide binding studies with the four peptides were performed with a
widely used assay that measures the stabilization of MHC class I
molecules at the surface of TAP-deficient RMA-S cells by exogenous
loading of synthetic peptides (27). Fig. 5B
shows that all four peptides can be considered as intermediate to
strong Db-binding peptides compared with several
known Db-binding CTL epitopes that have been
tested in our laboratory (27). However, a reproducible
5-fold difference was observed between the peptides with isoleucine or
leucine as C-terminal residue (Fig. 5B). The peptides ending
with isoleucine showed stronger binding, suggesting that this accounts
for the fact that they are better targets for CTL recognition.
|
The efficacy and specificity of the CTL response against the newly
identified epitope were further analyzed in vivo. First, we tested the
efficacy of adoptively transferred CTL against RMA in tumor-bearing
mice. Nude mice were injected i.p. with RMA tumor cells, leading to
progressive tumor burden within 3 wk. Long-lasting tumor protection was
observed for all mice that were treated with the
gagL75–83-specific CTL and for approximately
half the mice that received the anti-tumor CTL (Fig. 6). No difference in tumor protection was
detected between mice receiving the CTL i.v. (Fig. 6) or i.p. (not
shown). All mice receiving IL-2 only or saline developed progressively
growing tumors.
|
), the NKGENAQAI and NKGENAQAL peptides
as well as the KNGENAQAI and KNGENAQAL peptides were combined in one
vaccine. Injection of these peptides in combination with the helper
peptide resulted in more efficient tumor protection than vaccination
with the helper peptide alone. In contrast, addition of the subdominant
env187–196 CTL epitope did not improve the tumor
protective capacity. Taken together, these data show that
anti-tumor CTL exhibit in vivo tumoricidal activity in both
peptide-vaccinated mice and an adoptive immunotherapy setting.
Moreover, CTL that are induced by the different synthetic peptides are
capable of recognizing the naturally presented peptide on the surface
of RMA. The slightly stronger protective effect of vaccination with the
NKGENAQA(I/L) peptides suggests, but does not prove, that the
asparagine residue represents the N-terminus of the natural CTL
epitope.
|
We found the tumor-derived CTL epitope to be expressed by three
different T cell lymphomas (RMA, MBL-2, and EL-4). These tumors are
widely considered to be independently derived, distinct cell lines.
Therefore, we considered this novel epitope to be a
transformation-associated Ag that is exclusively expressed in the T
cell lineage. By coincidence, we were informed by C. G. Brooks
(Newcastle, U.K.) about the fact that these tumor cells expressed the
same TCR Vß region (Vß12), including fully identical CDR3 regions
(43) (TCR ß-chain sequence accession no. AF020206). This
prompted us to analyze the different tumor cells in our laboratory for
their TCR ß- and -chains. By TCR-specific PCR analysis and
subsequent DNA sequencing we confirmed the identical Vß usage in
these cell lines. Furthermore, we found that all recognized tumor cell
lines also shared the usage of one TCR V gene (V10), as shown in
Fig. 7 for RMA, MBL-2, and EL-4. By
sequencing the V-J-C region we established that these tumors
shared identical junctional regions. This TCR sequence is filed in
the GenBank database (accession no. AF218247). We excluded the
possibility of cross-contamination of cell lines within our laboratory,
because we confirmed these findings with cell samples obtained from
several other laboratories (not shown, see Fig. 7). The common origin
of these cell lines is most likely the result of a cross-contamination
during in vitro culture or in vivo passage many years ago, as described
for other widely used cell lines (34, 35, 36). Importantly,
these findings also imply that the epitope described here represents a
unique epitope rather than a T-cell-lineage-specific tumor Ag. The
sequence of this novel CTL epitope is not comprised within the
clonotypic TCR or ß sequences or by MuLV sequences, suggesting
that it is encoded by an as yet unknown cellular gene.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
- and ß-chains
have unequivocally revealed that these widely used cell lines have a
common origin. Cross-contamination during in vitro culture or in vivo
passage of these cell lines most likely lies at the basis of this
phenomenon. Early analyses of the immune responses to these tumors have
determined both virus-directed and tumor cell-directed CTL
(18). Recently, an important virus-derived CTL epitope was
identified (16). In the present paper we describe
identification of the tumor-associated CTL epitope as determined by
mass spectrometry. The nonamer peptide NKGENAQAI most likely represents
the natural epitope sequence, although we cannot exclude that the order
of the first two residues is reversed and/or that the C-terminal
residue is leucine instead of isoleucine.
The newly identified CTL epitope was selectively expressed on the
transformed T cells RBL-5/RMA, MBL-2, and EL-4. B cell lymphomas and
murine tumors of other origin were not recognized by these CTL. The
fact that this epitope was presented by T cell lymphomas regardless of
the transforming agent (MuLV induced as well as chemically induced; see
Fig. 2
A) suggested that this epitope originated from a T
cell differentiation Ag. However, in close collaboration with the
laboratory of C. G. Brooks (Newcastle, U.K.) we discovered that
all cell lines recognized by our CTL express identically rearranged TCR
and ß genes. This indicates that these cell lines, which have
widely been considered as distinct and independently isolated cell
lines, represent cross-contaminants of the same cell line. Importantly,
this cross-contamination was demonstrated in two independent
laboratories, which never exchanged cell lines. Furthermore, the same
TCR profiles were found in cell samples obtained from yet other
laboratories. The cell lines that are involved in this widespread
cross-contamination are MBL-2, EL-4, and RBL-5 together with its
derivatives RMA and RMA-S. Importantly, several of these cell lines
display very distinct phenotypes, such as the TAP deficiency in RMA-S
cells that leads to low expression of MHC class I on the cell surface
and the absence of FMR type MuLV proteins from EL-4.
In itself, the fact that such widely used cell lines have been cross-contaminated is not surprising, since this has been described for other commonly used human tumor cell lines such as the HeLa cervical carcinoma cell line (34, 37). Nevertheless, this report together with that of Brooks and co-workers (see Footnote 3) presents the first case of such widespread cross-contamination of murine cell lines, which is even more striking because these cell lines are among the most widely employed murine tumor lines for immunological research. Although it is very difficult to make a reliable reconstruction of the events that resulted in the cross-contamination, there are two basic possibilities. First, all cell lines may be derived from EL-4, a chemically induced cell line generated in 1945 (21). Secondary infection with Rauscher or Moloney MuLV of EL-4 could have resulted in RBL-5 and MBL-2. Transduction of such retroviruses can readily occur in laboratories where such retroviruses and cell lines are propagated in parallel. Virus infection could also have occurred by in vivo passage of these cells, which was a commonly used procedure to propagate tumors several years ago. Hence, we have observed that Rauscher and Moloney MuLV, the reported transforming agents of RMA and MBL-2, respectively, have strong sequence homology in immunologically relevant regions. It has never been carefully checked whether the integrated MuLV gene copies in these tumor cell lines display clear differences. Alternatively, all cell lines may be derived from the same Rauscher/Moloney MuLV-induced tumor, with EL-4 being a virus-loss variant. This latter possibility we consider unlikely (see below). Importantly, screening of cell lines from several laboratories confirmed that this cross-contamination is widespread, implicating that it took place many years ago. The fact that the RBL-5 cell line, which gave rise to RMA and RMA-S in the eighties, is also involved in the cross-contamination establishes the idea that the cross-contamination must have taken place before the mid-eighties. Of note, the Friend MuLV-induced erythroleukemia FBL-3 is a distinct cell line and has no relationship to the EL-4/RBL-5/MBL-2 tumor lines, because it obviously does not express TCR genes.
The finding that all tumor cell lines that present the novel
tumor-specific CTL epitope are all derivatives of one common tumor cell
line implies that this epitope is uniquely expressed by one tumor
rather than a lineage-specific Ag. Similar CTL epitopes were shown to
be derived from point-mutated cellular genes and have especially been
found in murine tumors that were induced by potent carcinogens
(5, 38, 39). In this respect it is of interest to note
that EL-4 cells are obtained from a dimethylbenzanthracene-treated
mouse, a compound known to efficiently induce DNA mutations (40, 41). The previously identified MMTV-derived CTL epitope
expressed on EL-4 was also shown to differ at one critical amino acid
position compared with known MMTV sequences. Interestingly, this
epitope was found to be shared by EL-4 and RMA (8). These
considerations are in favor of the hypothesis that RBL-5 and MBL-2 are
derivatives of EL-4, rather than the reverse (see above). Homology
searches in the available databases for our peptide sequence, allowing
minor amino acid variation, did not yield a gene of interest. We have
excluded the joining region of the TCR - or ß-chain as expressed
on the tumor cells and known (endogenous) viral sequences, e.g., MuLV
and MMTV as the possible genes that might encode this peptide.
Furthermore, our attempts to isolate the gene encoding this CTL epitope
through expression screening of cDNA libraries failed. Recently, we
successfully employed this approach to clone a novel CTL epitope
expressed on certain murine tumor cells (25). Extensive
screening of >60,000 cDNA clones from a size-selected cDNA library did
not result in identification of a positive clone. In summary, this CTL
epitope is most likely generated by a point mutation in an unknown
cellular gene.
Our in vivo experiments have shown that the CTL against the newly identified epitope play a clear role in the protective anti-tumor response. Nevertheless, the CTL response against the viral gagL75–83 epitope appears to be more efficient in tumor eradication in prophylactic peptide vaccinations as well as in therapeutic adoptive transfer setting. One possible explanation for this is the fact that RMA shows high expression of its MuLV genes and therefore presents virus-derived MHC class I peptides to a higher extent than peptides derived from cellular genes. Alternatively, the amounts of CTL precursors specific for the involved peptides may differ. Peptide vaccination with the helper peptide only can result in complete protection against this MHC class II-negative tumor (14); however, in these experiments this protective capacity was only sufficient for a significant delay of tumor growth. These results indicate that the CTL response to the novel epitope contributes markedly to the total CTL response against these tumors.
In conclusion, the Db-binding peptide NKGENAQAI
or a very similar sequence is the RMA-specific CTL epitope that has
long been searched for. The identification of this peptide together
with the expression of identical TCR and ß rearrangements
revealed that RBL-5/RMA, MBL-2, and EL-4 have a common origin.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Abbreviations used in this paper: MMTV, mouse mammary tumor virus; MuLV, murine leukemia virus; B6, C57BL/6; FMR, Friend-, Moloney-, and Rauscher-type MuLV; FRE, Fischer rat embryo; MS, mass spectrometry.
Received for publication February 28, 2000. Accepted for publication April 25, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  W. W. Leitner, M. C. Baker, T. L. Berenberg, M. C. Lu, P. J. Yannie, and M. C. Udey Enhancement of DNA tumor vaccine efficacy by gene gun-mediated codelivery of threshold amounts of plasmid-encoded helper antigen Blood, January 1, 2009; 113(1): 37 - 45. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. van Hall, S. Laban, D. Koppers-Lalic, J. Koch, C. Precup, P. Asmawidjaja, R. Offringa, and E. J. H. J. Wiertz The Varicellovirus-Encoded TAP Inhibitor UL49.5 Regulates the Presentation of CTL Epitopes by Qa-1b1 J. Immunol., January 15, 2007; 178(2): 657 - 662. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Ohta, E. Gorelik, S. J. Prasad, F. Ronchese, D. Lukashev, M. K. K. Wong, X. Huang, S. Caldwell, K. Liu, P. Smith, et al. A2A adenosine receptor protects tumors from antitumor T cells PNAS, August 29, 2006; 103(35): 13132 - 13137. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Kessler, X. Hong, J. Petrovic, A. Borodovsky, N. P. Dantuma, M. Bogyo, H. S. Overkleeft, H. Ploegh, and R. Glas Pathways Accessory to Proteasomal Proteolysis Are Less Efficient in Major Histocompatibility Complex Class I Antigen Production J. Biol. Chem., March 14, 2003; 278(12): 10013 - 10021. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Kelly, K. Takeda, P. K. Darcy, H. Yagita, and M. J. Smyth A Role for IFN-{gamma} in Primary and Secondary Immunity Generated by NK Cell-Sensitive Tumor-Expressing CD80 In Vivo J. Immunol., May 1, 2002; 168(9): 4472 - 4479. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Iwanami, A. Niwa, Y. Yasutomi, N. Tabata, and M. Miyazawa Role of Natural Killer Cells in Resistance against Friend Retrovirus-Induced Leukemia J. Virol., April 1, 2001; 75(7): 3152 - 3163. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
,
Specific lysis by anti-tumor CTL (ln17) as indicated on the
left y-axis. In this experiment duplicate wells were
incubated with an E:T cell ratio of 7.5:1. The specific lysis of RMA
cells was 55%. Each test well contained peptide material corresponding
to 4 x 106 RMA cells. The right
y-axis shows the percentage of acetonitrile in each HPLC
fraction, as indicated by the solid line.
