An intracisternal A-particle sequence codes for an antigen recognized by syngeneic cytolytic T lymphocytes on a mouse spontaneous leukemia

A CTL clone recognizes an IAP-encoded antigen

Eur. J. Immunol. 1994. 24: 2203-2212

Vinciane de BergeyckO, Etienne De Plaen, Patrick Chomez, Thierry Boon and Aline Van Pel Ludwig Institute for Cancer Research, Brussels Branch, Brussels and Cellular Genetics Unit, UniversitC Catholique de Louvain. Brussels

An intracisternal A-particle sequence codes for an antigen recognized by syngeneic cytolytic T lymphocytes on a mouse spontaneous leukemia* Cytolytic T lymphocyte (CTL) clones directed against spontaneous mouse leukemia LEC have been obtained. By transfecting a cosmid library into cells which were then tested for their ability to stimulate the CTL, we identified the gene coding for the antigen recognized by one of these CTL clones. It is the gag gene of an endogenous defective retrovirus that belongs to the intracisternal A particle (IAP) family. A gag-encoded nonapeptide presented by the H-2 Dk molecule caused recognition by the anti-LEC CTL clone. Southern blot and polymerase chain reaction analyses indicated that the expression of the antigen by the LEC tumor cell line resulted from the transposition of an IAP sequence into a new genomic location.

1 Introduction In the mouse, virus-induced tumors and tumors induced by high doses of chemical carcinogens or ultraviolet irradiation express antigens that can elicit an immune rejection response [l-71. This could be demonstrated by the resistance against tumor challenge observed in syngeneic animals previously inoculated with the same tumor, either as irradiated cells or as living cells later removed by surgery. On the contrary, no evidence of such protection was observed for tumors that had arisen spontaneously [8]. Nevertheless, in vitro mutagen treatment of two such spontaneous leukemias of CBA/Ht mice generated immunogenic variants that were rejected by syngeneic mice, and conferred protection against the parental tumor. This immune protection was specific, for no protection was observed when mice were challenged with the other spontaneous leukemia or with a syngeneic radioinduced thymic leukemia [9]. This indicated that spontaneous tumors carry antigens that can be targets for syngeneic rejection responses even though they are not immunogenic on their own.These antigens are not tissue culture artifacts since protection after immunization with tum- variants was observed not only against a challenge with parental tumor cell lines but also against the original transplantable tumors ~91. [I 131251

*

2203

This work was partially supported by the Belgian program on InteruniversityPoles of attraction initiated by the Belgian State, Prime Minister’s Office, Office for Science, Technology and Culture (OSTC).The scientific responsibility is assumed by the authors. This research was partially subsidized by the Fonds Maisin, the Dr. Steiner prize (Switzerland), and by the Association contre le Cancer (Belgium). Supported by the Fonds National de la Recherche Scientifique,

Brussels, Belgium. Correspondence: Aline Van Pel, Ludwig Institute for Cancer Research, Brussels Branch, 74 avenue Hippocrate, B-1200 Brussels, Belgium (Fax: 32-2-7629405)

Abbreviations: IAP: Intracisternal A particle LTR: Long terminal repeat Key words: Cytolytic T cell epitope / Intracisternal A particle / Tumor antigen 0 VCH Verlagsgesellschaft mbH, D-69451 Weinheim, 1994

In several tumor systems, the tumor rejection antigens induce a cytolytic T lymphocyte (CTL) response, and highly specific CTL clones have been obtained [ 10-121. For mastocytoma P815, it was possible to demonstrate that the antigen recognized by a CTL clone was relevant to the immune rejection of the tumor cells by syngeneic mice, because tumor cells that escaped partial immune rejection in vivo were found to be resistant to lysis by this CTL [13]. The gene that codes for this antigen was identified by a procedure involving the transfection of a cosmid library, and the detection of antigen-expressing transfectants by their ability to stimulate CTL proliferation [14]. This gene named P1A is expressed in several mouse mastocytomas but not in normal tissues with the exception of testis and placenta. Immunizations of syngeneic CBA/Ht mice (H-2k) with tum- variants derived from spontaneous leukemia LEC generated CTL that lysed the parental tumor and did not lyse other syngeneic leukemias [9]. Stable CTL clones specific for leukemia LEC were obtained. We report here the isolation and characterization of the gene that codes for the antigen recognized by one of these CTL clones.

2 Materials and methods 2.1 Cell lines and culture conditions Spontaneous leukemia LEC was isolated from CBA/Ht mice by H. Hewitt, adapted to culture and cloned [9]. Antigen-loss variants were obtained by co-cultivating approximately 3 x lo6 cells of an azaguanine-resistant subclone and 5 x lo6 to 8 x 106 CTL-LEC1: 5.The surviving cells were allowed to expand, cloned and tested for resistance to lysis by CTL-LEC1 :5. P1.HTR.KkDk,P1.HTR.Kk and P1.HTR.Dk are the recipient cell lines used for transfections. They were obtained by co-transfecting 5 x 106 P1.HTR.tk- cells [15] with 3 pg of DNA from the pSVtkneob-selective plasmid [16], 4 pgof DNA from a pBR322 plasmid containing the H-2 Kk gene [17] and 4 pg of DNA from the 8.5-kb EcoRI-Hind111 fragment of cosmid k12.1 coding for H-2 Dk [18]. The transfectants were submitted to G418 selection (1.5 mg/ml) 0014-2980/94/0909-2203$10.00+ .25/0

2204

V. de Bergeyck, E. De Plaen, P. Chomez et al.

and cloned. The clones were screened for the expression of H-2kclass I molecules in a visual lysis assay using an anti-Dk CTL clone and an anti-Kk CTL clone (see Sect. 2.2). LEC cells were cultured in Dulbecco's modified Eagle medium (DMEM) supplemented with L-arginine (5.5 x 10-4 M), L-asparagine (2.4 x M), L-glutamine (1.5 x 10-3 M), glucose (3.5 g/I), Hepes (lop2 M), 2-mercaptoethanol ( 5 x M) and 10 YO (v/v) fetal calf serum (FCS). P815 cells were cultured in DMEM supplemented with 10 YO FCS. All cells were incubated in water-saturated air containing 8 % COz at 37°C. 2.2 CTL clones The derivation and culture conditions of clone CTLLECl :5 have been described previously [9]. Two to four days before the stimulation assays, cultures were divided in 2-4 parts and medium containing 50 YOof supernatant from secondary mixed lymphocyte cultures (MLC) was added up to 1 ml. Anti-Dk CTL clone was isolated from C57BL/6 mice (H-2b) immunized with cllD fibroblasts derived from L cells [19]. Anti-Kk CTL clone was isolated from DBA/2 mice (H-2d) injected with P1.HTR cells expressing Kk. Every 3-4 days, 6 x lo4 to lo5 allogeneic CTL were transferred to 1-ml cultures containing 5 x lo6 irradiated (3000 rad) CBA/Ht spleen cells and 50 YO of MLC supernatant. 2.3 DNA-mediated gene transfer and selection of transfectants The calcium phosphate precipitation method described by wolfel et 1201 was used to transfect groups Of I! lo6 P1.HTR'KkDkcells with the DNA Of interest (30-60 kg Of DNA from groups pg Of DNA from individual cosmids, plasmids or purified fragments) and the

DNA Of pHMR272 (1-3 as marker [21]. After 48 h, selection of transfected cells was carried Out in DMEM supplementedwith '% FCS and hygromy(350 pg/ml) and a colony assay was performed to estimate the number of independent transfectants. Five to six days later, the antibiotic-resistant transfectants were separated from dead cells by density centrifugation with Ficoll-Paque (Pharmacia Fine Chemicals, Piscataway, NJ). 2.4 Screening of transfectants with CTL 2.4.1 Stimulation assay Detection of transfectants expressing antigen LEC-A was carried out on pools of 20-30 transfectants using the proliferation assay described by Wolfel et al. [20]. The number of pools was adapted in order to test each transfectant approximately five times. After5 days at 37 "C, each microculture (containing approximately 3.5 x lOj cells) was duplicated. Replicates were kept at 30°C. One day later, each master culture was resuspended in 200 pI of LEC culture medium containing 2000-3000 CTL-LEC1 :5, l@ irradiated syngeneic feeder spleen cells, HAT, 20 U/ml of human rIL-2 (a gift from Biogen, Geneva, Switzerland)

J. Immunol. 1994. 24: 2203-2212

and 10 U/ml of murine rIL-4 (0.1 % of supernatant from baculovirus-infected Sf9 cells, a gift from J.Van Snick). Six to eight days later, the lytic activity of each microculture was measured by transferring 100 pl to another well containing 1500-6000 51Cr-labeledLEC cells. 51Crrelease was measured after 4 h. The duplicates corresponding to the microcultures showing CTL activity were cloned. Clones (100-250) were screened with CTL-LEC1 :5 in a visual lysis assay as described in [22].

2.4.2 Assay for cytolytic activity Briefly, CTL were incubated with 2000 51Cr-labeledcells at various lymphocyte/target ratios in 96-well conical-bottom microplates in a final volume of 200 pl. W r release in the supernatant was measured after 4-6 h of incubation at 37°C as described in [23]. 2.4.3 TNF production assay Transfectants were plated in 96-well microplates at 2000 cells/well. After 3-4 days at 37"C, the plates were centrifuged, the medium was discarded and microcultures were resuspended in 50 pl of medium. CTL-LECl :5 (2000) were added in 50 p1 of CTL culture medium supplemented with 20 U/ml of human rIL-2 and 10 U/ml of murine rIL-4. After 20 h, the supernatant was collected and itsTNFcontent was determined by its cytolytic effect on WEHI-164 clone 13 cells [24] as previously described [25]. 2.5 Restriction map of the 9-kb BamHI fragment The 9-kb BamHI fragment was cloned in pTZ18R (pharmacia). The recombinant DNA was linearized at the ScaI site of the plasmid and then partially digested with EcoRI, SmaI, XbaI or SalI restriction enzymes. Restriction fragments were separated on a o.5 agarose gel, blotted on nitrocellulose (Schleicher & Schuell [BA85], Dassel, FRG) and hybridized with 32P-labeIed or universal primers. Hybridization and washing conditions were as in [26]. The distances between EcoRI, SmaI, XbaI or SalI sites and the ScaI site were determined. 2.6 DNA sequencing and homology search DNA sequencing was performed on inserts obtained by digestion with restriction enzymes or by deletions with exonuclease I11 (Erase-a-base, Promega, Leiden, The Netherlands) and cloned in pTZ18R. All sequencing reactions were performed using the dideoxy chain termination method (Pharmacia T7 Sequencing Kit) and synthetic oligonucleotides. The computer search for sequence homology was done with program BLAST. Genbank database release 82 (April 1994) was used. 2.7 Polymerase chain reactions (PCR) PCR were carried out using recombinant Taq DNA polymerase (Perkin-Elmer-Cetus, Emeryville, CA) in a TRIOThermoblock (Biometra, Gottingen, FRG).

Eur. J. Immunol. 1994.24: 2203-2212

A CTL clone recognizes an IAP-encoded antigen

2205

6om

PCR fragments were obtained by denaturing 100 ng of the 268-bp XbaI-EcoRI fragment at 94°C for 4min and 50 performing amplifications for 20 cycles (94°C 1 min; 64 "C 2 min; 72 "C 2 min) with specific primers. BamHI or EcoRI restriction sites and ATG or STOP codons were included in the primers (lowercase). Fragments corresponding to positions 699-797, 762-875 and 852-953 of the LEC-A IAP sequence were generated respectively with oligonucleotides 5'-aaggatccatgGATGAAATTCAGGACAGGACAAGCTATCAGAAG-3' and 5'-aagaattcctaGGGTTCAAGACCCTTGGAAAGG-3', with oligonucleotides 5'-aaggatccatg.01 .03 .1 .3 1 lymphocyteltarget ratio GGTAAGTATACAGGCCTTTCCAAG-3' and 5'-aagaattctcaATCTTTTTTCITCTCCCITIlTCCTC-3'and with oligonucleotides 5'-aaggatccatgGGAAAAAGGGAGA- Figure 1. Cytolytic activity of clone CTL-LEC1:5 on S'Cr-labeled AGAAAAAAGATCG-3' and 5'-aagaattcttaATCTGCT- target cells measured after 6 h. TCAGAACI'ACTAAGAGCT. PCR products were cloned in the BamHI and EcoFU sites of the pcDNAI/Amp 50 expression vector (Invitrogen Corporation, Oxon, GB) and then subcloned to the KpnI and EcoRI sites of pSVK6.4. This plasmid is a modified pSVK3 expression vector L .o 30 (Pharmacia): the EcoRI site of pSVK3 was removed by 0 2 0 filling in with Klenow enzyme, and an EcoRI linker E: (Pharmacia) was introduced in the XhoI site after treat- L 10 0 ment with Klenow enzyme. LEC 24x10 24x10 . plate C plate D plate A pate D PCR on genomic DNA were performed with oligonucleotides 5'-TGGTTAACGGGTCAGACATGA-3'(Fig. 6 positions -22 to -2 sense), 5'-GGAAAGGCCI'GTATACTTACCT-3' (positions 782-760 anti-sense), 5'-CGTGGGAATCCTCTGAAGGG-3' (positions 2796-2815 sense) and 5 '-CAATCAGCACCATGAG GACAGC-3' (positions 4473-4452 anti-sense). DNA (100 ng) was first denatured at 94 "C for 5 min. Amplifications were then performed for 30 cycles (1 min 94°C; 2 min at 60°C for oligonucleotides -22 to -2 and 782-760, at 65°C for oligonucleotides 2796-2815 and 4473-4452, and at 62°C for oligonucleotides -22 to -2 and 4473-4452; 3 min 72°C).

2.8 Peptide sensitization assay

k k P1 H T R K D

.

.

eX1o4

6x10~

group 3

group 2

Figure 2. Detection of transfectants expressing antigen LEC-A. (A) Reconstruction experiments with microcultures seeded with 2500 CTL-LECI :5 and either no stimulating cells, LEC cells, a mixture of LEC cells and P1.HTR.KkDkcells, or PI.HTR.KkDk cells. After 6 days, the lytic activity of the microcultures was tested on 6OOO 51Cr-labeledLEC cells. Each point represents the lytic activity of a single microculture. (B) Detection of a genomic transfectant-expressingantigen LEC-A. Pools of 30 hygromycinresistant cells were tested for their ability to stimulate proliferation of 3000 CTL-LEC1:5.The lytic activity of the microcultures was measured 8 days later on 3000 S'Cr-labeled LEC cells. Each point represents the lytic activity of a single microculture. (C) Detection of a transfectant-expressing antigen LEC-A among cells transfected with DNA from cosmid group 36. Stimulation assay was performed on pools of 25 transfectants with 2500 CTL. Six days later, the lytic activity of the microcultures was tested on 6OOO SICr-labeledLEC cells. Each point represents the lytic activity of a single microculture.

The peptides were synthesized by C. Servis (Ludwig Institute for Cancer Research, Lausanne Branch, Institute of Biochemistry, University of Lausanne) using the F-moc, t-Bu strategy as described in [27].Various concentrations of the peptides were incubated with 2000 51Cr-labeled To clone the gene coding for antigen LEC-A, we resorted to P1.HTR.KkDk for 30min at 37" before the addition of a gene transfection approach involving the detection of CTL-LEC1: 5 at a lymphocyte/target ratio of 10/1. 51Cr transfectants by their ability to stimulate the proliferation release was measured after 4 h at 37°C. of the relevant CTL. LEC variants selected for the loss of antigen LEC-A could not be used as DNA recipients, because LEC cells integrate exogenous DNA at very low 3 Results frequencies (< lo-'). Suitable frequencies (10-3-10-4) had been obtained with the highly transfectable P1.HTR line [15], but this H-2d cell line could not be used for t h e 3.1 Transfection with genomic DNA cloning of antigen LEC-A which is presented by an H-2k Several CTL clones recognizing tumor LEC have been molecule. Therefore, we transfected it with the Kk and the obtained [9]. Clone LECl: 5 lyses the LEC cells and shows Dk genes. A mixture of LEC and P1.HTR.KkDkcells at a little or no lysis on syngeneic leukemias LEB [9] and T H 1:25 ratio was able to induce proliferation of CTL-LECl :5 [28] (Fig. l).The antigen recognized by this CTL clone was (Fig. 2A), suggesting that the identification of named LEC-A. CTL-LEC1 :5 proliferated significantly P1.HTR.KkDk transfectants expressing antigen LEC-A better in the presence of stimulator cells expressing this might be achieved by testing pools of approximately 25 antigen than in the absence of such cells. This proliferation transfected cells in a CTL stimulation assay. could be measured either by counting the lymphocytes after a few days of culture or by measuring the total lytic activity A total of 24 groups of 5 x lo6 P1.HTR.KkDkcells were co-transfected with genomic DNA from a LEC subclone of the culture (Fig. 2A).

J. Immunol. 1994. 24: 2203-2212

V. de Bergeyck. E. De Plaen. F? Chomez et al.

2206

[30]. This library, which contained about 10 equivalents of the diploid genome, was divided into 40 groups containing each 42 OOO independent cosmids. The DNA of each group was extracted and co-transfected with pHMR272 into P1.HTR.KkDk cells. An average of 5500 independent hygromycin-resistant transfectants per group were obtained. The transfectants were tested for their ability to stimulate the proliferation of CTL-LEC1:5. Transfectants expressing antigen LEC-A were obtained with 4 out of the 40 cosmid groups (Fig. 2C). The DNA of these cosmid groups was transfected again. Two additional transfectants were obtained. The lysis of two transfectants by CTLLEC1:5 is shown in Fig. 3.

and with cosmid pHMR272 which confers resistance to hygromycin B. After antibiotic selection, the transfectants were divided in pools of 30 cells which were allowed to multiply for 5 days. The microcultures were then duplicated, and tested for their ability to stimulate CTLLECl :5. The proliferation was evaluated visually and confirmed by measuring the lytic activity of the microcultures on 51Cr-labeledLEC cells (Fig. 2B). A few positive microcultures were obtained. Their duplicates were subcloned and each clone was tested for its sensitivity to lysis by the CTL-LEC1: 5 in a visual lysis assay. Clones that were lysed by the CTL were obtained with 4 of the 24 transfection groups. Their lysis by CTL-LEC1: 5 was comparable with that of LEC (Fig. 3). Thus, 4 independent transfectants expressing antigen LEC-A were obtained from a total of 61 500 hygromycin-resistant transfectants. On the basis of previous estimates that transfectants incorporate approximately lo00 kb of DNA [29], it follows that the 4 transfectants expressing the antigen were obtained by transfecting the equivalent of ten diploid genomes.

To isolate a cosmid carrying the gene encoding antigen LEC-A, 1 pg of DNA extracted from the cosmid transfectants was packaged directly into lambda phage components [30,31]. Five of the six transfectantsproduced cosmids.The pools of cosmids rescued from each transfectant were amplified and their DNA was transfected into P1.HTR.KkDk.Only the cosmids that were obtained from transfectant Pl.LEC.TC36.1 were found to transfer the expression of LEC-A. Because this transfectant produced a large number of different cosmids, a sib selection was performed to isolate the cosmid coding for antigen LEC-A. Cosmid P5.C3 was found to confer the expression of LEC-A with high efficiency (Fig. 4).

We concluded that the gene coding for antigen LEC-A could be transfected efficiently. Because the transfectants obtained with genomic DNA did not allow us to retrieve the relevant DNA sequence, we repeated the procedure with a cosmid library.

3.2 Isolation of a cosmid coding for antigen LEC-A A cosmid library was constructed in vector c2RB with the DNA of a LEC subclone, using the procedure described in LEC ' 60-

-

PI.HTR.K~$

O

PI.LEC.TCI 1.1 I

PI.LEC.TZ

PI .LEC.TC36.1

1

u)

-%

U)

50-

-t 0

40-

P

0.

30-

u)

z

2010-

- 0 1 .3

1

3

10

1

3

1 3 1 0

.1.3

1 3 1 0

lymphocyteltarget

cosmid P5C3

.1

3

1

3

10

20 kb Scal fragment

1

3

1

3

10

pTZI8R

.1

.3

+ 9kb BamHl

1

3

10

lymphocyteltarget

/+

-.1

.3

1

3

10

.1

.3

1

3

10

ratio

clone Exolll 17b

.1

.3

ratio

1

3 10

Figure 3. Sensitivity of LEC, P1 .HTR.KkDk,genomic transfectant (P1 .LEC.T2) and two cosmid transfectants (Pl.LEC.TC11.11 and TC36.1) to lysis by CTL-LECl:5. "Cr release was measured after 4 h.

pHMR272

Figure 4. Expression of antigen LECA by cells transfected with cosmid PS.C3orfragmentsof thiscosmid.The DNA was co-transfected with pHMR272 into Pl.HTR.KLDk cells. The hygromycin-resistant cells were slCr-labeled and incubated with CTLLECl : S at various lymphocyteltarget ratios. The 51Crrelease was measured after 4 h.

--/ .1

.3

1

3

10

Eur. J. Immunol. 1994. 24: 2203-2212

A CTL clone recognizes an IAP-encoded antigen

3.3 Characterization of the sequence encoding LEC-A antigen Because the ability of cosmid P5.C3 to transfer the expression of LEC-A was not abolished by digestion with ScaI, fragments obtained with this restriction enzyme were isolated and transfected. A 20-kb fragment conferred the expression of the antigen (Fig. 4). This fragment was ligated to EcoRI adaptors, cloned into phage EMBL3A and digested with BamHI. A 9-kb BamHI fragment cloned into pTZ18R produced transfectants that were recognized by CTL-LEC1:5 (Fig. 4). Expression of antigen LEC-A was considerably reduced by the removal of the 1.6-kb SalI fragment located on one end of the 9-kb fragment (Fig. 5A), suggesting that a sequence located at this end was important for antigen expression. We therefore made progressive deletions from the other end of the 9-kb fragment with exonuclease 111. Partially deleted subclones were transfected (Fig. 5B). Clone ExoIII17b (4.38 kb) was the smallest fragment capable of transferring the expression of the antigen (Fig. 4). The sequence of clone ExoIII17b was determined. Data bank search revealed a very strong similarity with the genome of endogenous defective retroviruses that belong to the murine intracisternal A particle (IAP) family. The complete nucleotide sequence of the IAP element encoding antigen LEC-A (LEC-A IAP) is shown in Fig. 6. This sequence is a class I retrotransposon with a central region flanked by long terminal repeats (LTR) .The transcriptional regulatory elements of the 3’LTR are required to transfer efficiently LEC-A antigen expression (data not shown).

EcoRl Smal

.c

.t.t-

c

c

Xbal

-

Sdll

BdmHl

The 4344-bp LEC-A sequence was compared with the 7095-bp sequence of MIA14, a full-length provirus whose genetic structure has been described (accession number M17551/locus name MUSFTIAP [32]). This comparison showed that in the LEC-A IAP element a 2753-bp deletion removed the entire pol region. The major ORF of the LEC-A IAP sequence is located within the gag region. The 599-amino acid protein encoded by this ORF was compared to the 827-amino acid gag protein encoded by the MIA14 IAP element. Amino acids 41 to 564 of the LEC-A protein are 93% identical to positions 1 to 532 of the MIA14 protein.

3.4 Presentation of antigen LEC-A by Dk To identify the H-2k class I molecule presenting antigen LEC-A, cosmid P5.C3 was co-transfected with pHMR272 into P1.HTR cells that express either H-2 Dk or Kk gene. The hygromycin-resistant transfectants were tested for recognition by CTL-LEC1 :5 in a 51Cr-release assay. Only the transfectants expressing the Dk molecule were specifically lysed (Fig. 7).

3.5 Antigenic peptides It has been established that the expression of antigens recognized by CTL on P815 can be transferred by small gene fragments containing the sequence coding for the antigenic peptide [33]. Accordingly, we co-transfected P1 .HTR.KkDkwith the DNA of pHMR272 and restriction

+ +

t

EcoR,

Hindlll

0

I

7.8 kb

clone Exolll3b

6.9 kb

clone Exolll7a

4.45 kb

clone Exollll Ib

4.38 kb

clone Exollll7b

3.05 kb clone Exolllla

3.8 kb

2

3

4

EoRl

+

.c

Smal

t .tt

.c

probe

r

2207

Xbdr

Sail

BamHl

LECIA-IAP

S,LTR

5

RUR

coding region

6

I

7

e m region

8

3,LTR

9 kb

I

Xbel -Xbdl

I .98 kb Smal - Smdl

4.59 bp EcoRl - EcoRl 268 bp Xbal -&&I

I0S bp KR fragment

Figure 5. (A) Restriction map of the 9-kb BamHI fragment. The position and the structure of the IAP sequence transferring the expression of antigen LEC-A are shown. (B) The plasmid carrying the 9-kb BamHI fragment was digested with KpnI and SpeI restriction enzymes. The SpeI ends of the resulting fragments were progressively deleted by the exonuclease 111 method. Deleted fragments were circularized generating plasmids containing shortened inserts. The DNA of cloned fragments was co-transfected with pHMR272 into P1 .HTR.KkDkcells. Hygromycin-resistant transfectants were tested in a CTL stimulation assay performed on pools of 20 independent transfectants. The fragments shown in grey transferred the expression of antigen LEC-A. (C) Obtention of a small gene fragment transferring the expression of antigen LEC-A.The DNA of cloned fragments was cotransfected with pHMR272 into P1.HTR.KkDkcells. Antigen expression was analyzed by a TNF production assay performed on pools of 2000 independent transfectants.

2208

A,

J. Immunol. 1994. 24: 2203-2212

V. de Bergeyck, E. De Plaen, F! Chomez e t al.

o

cttcatgaag qtagtcctgq tgaaataqt qqcaqaaata cagagtgtca aqctcaqaat tttattttqa ttttttcttg ga~~~ccattttata ,~?~.-.~taatqaca

G'XA-

CCCACATICG CCGTTACAAG ATGGCCCTGA CAGCXTGTT CTAAGXGTA AACAAATAAT CIWGCATGT GCCAACGGTA TClTA'TGACT ACTIGPGCFC

120

CWCGCCGAT GGGCTGCAGC CAATcMjGGA GTGACACGTC CGACGCGPAG GAGAATGCTC CTTAAGAGGG ACCGGGITPC

240 360

CACXCXGCT CCTGAAGATG TAI\GMATAA A G X T K C C G CAGAAGATTC T G T C K T G TGTFCTFCLT GGCCGGPCOP GAGAACCXGT C G M T

C-

GATPCAACTG TA

?CGGGACTGG W3AGGAAG ATCCCTCATT CCAGAACCAG AACTGCGXT CGCGGTAATA AAGGTPXCG T

GAACTITXA GcIy;GGGI\AC

n WlTAAGAAG

400

GAGAGTACCA GTCAGTACAG CT'ITACGAGG TAAW'TGAT ClTGAACllT CTAACGMAT TCAAGACXl CTATCAGAAG

600

ECORI U

N

S

E

L

F

S

W

G

T

R

V

P

V

S

T

A

T R M C T U ; M bTGTlTWJ2 CMCMATTT TTCTGGTGTT AGGAGCCCTT TII;TICC"T

L

R

G

K

S

D

L

E

L

S

N

TCACAWlTA m G T G G T T AAGATAGGGA G

E

G

A

I

Q

D

S

L S E V 3 7 720

B TGAAATTCAG GRCMGCTAT

XbaI K W N

U

F

G

L

E

I

F

L

V

L

G

A

L

L

F

L

F

T

C

Y

Q

V

V

K

I

G

R

I

L

D

E

I

Q

D K L S 7 7 040

CAGAAGTAAA GCGGAGAGAG AGAGTAGGAA CARGGAGAAA ~ - + - ~ ~ l ' & A ~ , ~ - , ~ C A A G G G XTGAACCCGAG T GAAAAGTTAA GGTTAGGTAG G M T A C C F X AGAGAGATTA

E

V

K

R

R

E

R

V

G

T

R

R

K

G

K

Y

T

G

L

S

K

G

L

E

P

E

E

K

L

R

L

G

R

N

T

W

R E I R 1 1 7

GAAGMAMG AGGAAlUAGG GAGAAGAAAA AAGATCGATT AGTGGI\GGTC TCTAGGAGAT ACTCGTCACT GGATGAGCTC AGGAAGCCAG C " T A G T A G TPCTGAAGCA G

A

e

960

T

EcoRI R

K

R

S

S

E

G

K

R

E

K

K

K

D

R

L

V

E

V

S

R

R

Y

W

S

GGAAACAGAC TGGGAGGAAG AAGCAGCCCA TTACCAGCCA GCTAA-T E

T

D

W

E

E

E

A

A

H

' I Q P

A

N

S

S

L

D

E

L

R

K

P

A

L

S

S

Q

F

A

CAAGAAAAAA GCCMAAGCG GCTWCGAAG

R

K

K

P

K

A

A

G

E

G

S

E

A

O E E F 1 5 7 1080

TGATI'GGCCT cM;GGcMTc D

W

P

Q G N R 1 9 7 1200

GGCl'ACCAGG TGCACTCCCG CCCTATGCGG AGXCCCGCC CTGCGTAGTG CGTCAGCCCG TADPGCGTCA GCAATiXGCA GAGAGGCAGT GCGCAGAGAG GCAGTGCGCA GACTCATICA L

P

G

A

L

P

P

Y

A

E

S

P

P

C

V

V

R

Q

P

V

V

R

Q

Q

C

A

E

R

Q

C

A

E

R

Q

C

A

D S F I 3 3 7 1320

TKCCAGAGA GGAACAAAGG MAATACARC AGGCATITCC G G T C m G M GGAGCCGAGG GTGGGCGTGT CCACGCFCCG GTAGAATACT TACRMTTAA GGAAATTGCC GAGKGGWC P

R

E

E

Q

R

K

I

Q

Q

A

F

P

V

F

E

G

A

E

G

G

R

V

H

A

P

V

E

Y

L

Q

I

K

E

I

A

E S V R 3 7 7 1440

GTAAATACU: AACUATGCT AATTITACCT XGTGCAGTT AGACAGGCTC CCTtXCATCG CACTAACTCC TGCTGACTGG CAMCGGTFG TAAAAGCCGC KTCCCTAGT T-A K

Y

G

ATA-TG

T

N

A

A

L

R

F

T

L

V

Q

L

D

R

L

A

G

U

A

L

T

P

A D W

Q

T

V

V

K

A

A

L

P

S

U G K Y 3 1 7 1560

TGWACGAAG CTGCACAAGC GCAGGCCCGA GCAAACGCAG CTGCTlTwIC TCCAGAGCAG AGAGATKGA CPTPIC.ACPT GTTAAC'XGT

GAGA-

M E W

N

W

H

E

A

A

Q

A

P

A

R

A

N

A

A

A

L

T

P

E

Q

R

D

W

T

F

D

L

L

T

G

Q G A Y 3 5 7 1600

ATTCKCTGA TCAGACRMC TACCATIYiGG GAGCTTATGC C C A G A m t T TCCACCJXTA TTAGGGCCTG GAAGGCGCTC TCTYXAGCAG Gl'GAAACCAC TGGTCAGTTA ACAAAGATAA S

A

D

Q

T

N

Y H W G

A

Y

A

Q

I

S

S

T

A

I

R

A

W

K

A

L

S

R

A

G

E

T

T

G

Q

L

T K I I 3 9 7 1000

TCCAGGGACC ?CAGGAATCC TICTCAGATT TIGTOXCAG AAXACAGAG GCAGCAGAGC GTATITITGG AGAGTCAGAG CARGC'XIXC CTCTGATAGA ACAGCTAATC TATGAGCMG Q

G

P

Q

E

S

F

S

D

F

V

A

R

U

T

E

A

A

E

R

I

F

G

E

S

E

Q

A

A

P

L

I

E

Q

L

I

Y E Q A 4 3 7 1920

CCACAAAGGA GTGCCGAGCG CCCATAGCCC CAAGAAAGM CAMGGCTCA CARGACPGGC TCAGGGTCTG VXAGAGCTT GGu;GRCCTC TCACCAATGC AGGClTAGCG GCCGCCATCC T

K

E

C

R

A

A

I

A

P

R

K

N

K

G

L

Q

D

W

L

R

V

C

R

E

L

G

G

P

L

T

N

A

G

L

A

A A I L 4 7 7 2040

TCCAAWWA GAACCGCTCC ATGAGCAGAA ATGAXAGAG G A C A T G m T AATTGCGGAA AGCCn%GCA TTTTAAGMA GATTGCAGAG CTCCAGATAA ACAGGGAGGG AClKXCACTC

Q

S

Q

N

R

S

U

S

R

N

D

Q

R

T

C

F

N

C

G

K

P

G

H

F

K

K

D

C

R

A

P

D

K

Q

G

G

T L T L 5 1 7 2160

TTTXTCTAA GTGTGGCAAG GGTTATCATA GACCTGACCA GTGTCGCXT GWXGGGATA TAAAGGGCAG AGTCCTTCCC CCACCTGATA GTCAATCAGC TTATGTGCCA AAAAACGGGT

C

S

K

C

G

K

G

Y

H

R

A

D

O

C

R

S

V

R

D

I

K

G

R

CATCGGCCC TCGGTCCCAG GCCCCTCAAA GATATGGGM CCGGTlTGTC AGGACCCAGG AAGCA-G S

G

P

R S Q

G

P

Q

R

Y

G

N

R

F

V

R

T

Q

E

A

V

P

P

P

D

S

Q

S

A

Y

V

P

K N G S 5 5 7

E

A

T

Q

E

D

P

Q G W

T

C

V

2200

P P P T 5 9 7

ACCATXGCC TTATTCPCGG C O X G G T X A CTCACCTTGC AGGUTTAGT

SmaI Y

L

AGAGGCGACC CAGGAAGACC CACAAGGGTG GACCTGCGTG CCGCCPCCGA R

CTICCTATTA A%CCWAAA TGAGTATNX GCCGGTGCCA G F X X X C T A TACCAKCTT S

V

I

2400

599 1520 2640

2760 2000 3000 3120 3240

3360 3400 3600 3720 3840 3960 4000

# 'c,

TTAWXCTAC m;n;CrCn; CCTFXCCGT GACGTCAACT CGGCCGATGG GCTGCAGCCA ATCAGGGAGT GACACGTCCG AGGCW\AOGA GAATGClYCT TAAGAGGGAC GGGGTTKGT

4100

CTcTpGcTl'C TIWWlWT TXCTGAAGA TGTAAGAATA MGCTTECC GCAGAAGATT cpU;rrTGTG GTGTICTICC

4320

tttaaaagct aaqaaagtat aaqgatttaa tatttaaaga ggatagcaag catattctqt cttqgacttt ggcaaaqcag

4440

mKTcTcp C K T P X l W

TTXWWlT GC-

rrr;Car;ra;%AGAAECG

-atcta

atgtgagaaa t ~ t ~ ~ ~ ~ ~ C ~ ~ ? , g ~ s -&atcQct ~ ~ j t t atg ce t ca da BamBI XbaI S a l I

B.

-

tgqttaacgq gtcagacatg aaqaacaaat ctatttaaaa gctaagaaaq tataagqatt taatatttaa agaggatagc aagcatattc tgtcttggac tttggcaaag cagatgtqag aaatgctgtc ctcatggtgc tqattg

4499

Eur. J. Immunol. 1994. 24: 2203-2212 k k

P1.HTR.K D

.1

.3

1

3 10

P1.HTR.K

.1

.3

1

A CTL clone recognizes an IAP-encoded antigen

k

3 10

iyrnphocyte/target

acid peptide gag.86-99 sensitized the cells to lysis when used at very high concentrations (Fig. 8). Shorter peptides were prepared. Efficient lysis of P1.HTR.KkDk cells was observed with decapeptide gag.88-97 and nonapeptide gag.88-96 (Fig. 8). For those peptides, a concentration of 100 n M produced 50% of the maximal lysis, which was reached at concentrations above 1 PM. Both the N-terminal arginine and the C-terminal leucine of the nonapeptide (RRKGKYTGL) appeared to be essential, as shown by the absence of activity of peptides gag.89-98 and gag.86-95.

PI.HTR.D'

.1

.3

1

3

10

ratio

Figure 7. LEC-A antigen is presented by H-2 Dk. P1.HTR cells, expressing one or both of the Kk and Dk genes were transfected with the DNA of cosmid P5.C3 and pHMR272 (W) or with pHMR272 alone ( 0 ) The . hygromycin-resistanttransfectants were tested for recognition by CTZ-LEC1:5 in a 4-h 51Cr release assay.

fragments of the 4344-bp LEC-A IAP sequence cloned into expression vector pSVK3. The hygromycin-resistant transfectants were screened for their ability to induce TNF production by CTL-LEC1:5. The 3.8-kb XbaI, 1.98-kb SmaI and 0.47-kb EcoRI fragments were found to confer the expression of antigen LEC-A (Fig. 5C). As expected, the 268-bp XbaI-EcoRI fragment contained in all these fragments was also positive. PCR amplification was used to produce smaller overlapping regions of the 268-bp fragment, which were cloned into expression vector pSVK6.4. A fragment ranging from nucleotide 699 to 797 of the LEC-A IAP element transferred the expression of antigen LEC-A. Four overlapping peptides encoded by this fragment were synthesized. P1.HTR.KkDk cells were incubated with these peptides and tested for recognition by CTL-LEC1 :5. The 14-amino

gag 8 6 - 9 9

gag 8 6 - 9 5 gag 8 8 - 9 6 gag 8 8 - 9 7

qaq 8 9 - 9 8

-

P 1 -

I CGG IAG4I GRG IAGR 1 GTAI G I 3 I ACA IA f f i I AG4I -1GGT

2209

3.6 Transposition of LEC-A IAP sequence Genomic DNA extracted from LEC, LEB and T H leukemia cells and from syngeneic CBA/Ht normal kidneys were digested by BamHI. A Southern blot of the digested DNA was hybridized with a 1.75-kb HindIII-EcoRI fragment corresponding to a sequence located upstream of the LEC-A IAP sequence (Fig. 5A). Two bands were observed with the LEC tumor:a 6-kb and a 10.5-kb fragment (Fig. 9A). Only the 6-kb band was observed in CBA/Ht kidney tissue and in syngeneic tumors LEB and T H which are not recognized by CTL-LEC1 :5 (Fig. 9A). This suggested that the antigen is encoded by the 10.5-kb BamHI fragment. The size of this fragment is compatible with the insertion of the 4344-bp LEC-A IAP in a 6-kb region bounded by two BamHI sites. To confirm the transposition of LEC-A IAP in the LEC tumor, PCR amplifications were performed on DNA extracted from LEC, LEB, TH and kidney tissue, with pairs of primers comprising one oligonucleotide located in the IAP sequence and another located in the adjacent upstream or downstream sequence (Fig. 6). These primers ensured the amplification of the sequences surrounding either the 5' or the 3' integration site. PCR products were obtained only with LEC DNA (Fig. 9B). The pair of primers located upstream and downstream from the LEC-A IAP element enabled amplification of a sequence found in all cells (Fig. 9C). The

> 10 pk-4

no lysis

a t 10 pk-4

100 I1M 100 nM

no lysis at 10 pk-4

IAAGl TAT IACA ICU: IClTl l'CC IMG IGGT ICIT [ GAR I CCCI GRG IGAR IA4G

Figure 8. Amino acid and nucleotide sequence of the region coding for the antigenic peptide. Synthetic peptides were tested for their ability to render P1.HTR.KkDkcells sensitive to lysis by CTL-LECl: 5. Concentrations producing 50 % of the lysis obtained at saturating concentrations of peptides are indicated. Position 1 corresponds to the first amino acid of the longest ORF encoded by the IAP sequence.

4 Figure 6. (A)Nucleotide sequence of the IAP element encoding antigen LEC-A (uppercase) and adjacent sequences (lower case). Each LTR is flanked by 4-bp inverted repeats S'TGTTIACAA3' (+).The 6-bp direct repeats (AACAAA +) at each end of the provirus are shown.The position of the deletion spanning the pol region is indicated (V).The amino acid sequence correspondingto t h e longest O W of the complete provirus is indicated.The sequence of the nonapeptide recognized by CI'L-LEC1: 5 is boxed. Oligonucleotidesused in PCR performed to confirm the IAP transposition (-. .-. .-) are shown. (B) Sequence of the 146-bp PCR product obtained from LEC genomic DNA with primers located upstream and downstream of the LEC-A IAP genome (Fig. 9C).

2210

V. de Bergeyck, E. De Plaen, €? Chomez et al.

J. Imrnunol. 1994. 24: 2203-2212

a lower level is found in most normal mouse tissues. There are about 1000 IAP copies per haploid genome of Mus musculus [34]. Roughly 700 of these copies are full size. They contain two LTR, as well as gag, pol and env-related regions. The other 300 IAP sequences carry deletions, like the element that codes for antigen LEC-A. They are classified into distinct subclasses according to the size and the location of the deletion [34].The LEC-A IAP element belongs to the A 3 subclass which lacks the pol region. L31, another A 3 IAP element, contains almost exactly the same deletion as the LEC-A IAP element [35].

Figure9. Evidence for the transcription of the 4344-bp IAP sequence encoding antigen LEC-A. (A) Genomic DNA (20 pg) was digested with BamHI. Restriction fragments were separated o n 0.5 % agarose gels, transferredto nitrocellulose and hybridized with the "P-labeled HindIII-EcoRI probe shown in Fig. 5 . DNA transfer. hybridization and washing conditions were as in [26]. (B) PCR amplifications were performed on genomic DNA with oligonucleotides flanking either the 5' integration site or the 3' integration site giving PCR products of 805 bp and 1,678 bp respectively.For each template, the two PCR products were pooled and an aliquot was size-fractionated on a 1.5 % agarose gel stained with ethidium bromide. (C) PCR amplifications were performed on genomic DNA with the same primers located upstream and downstream of the LEC-A IAP genome. An aliquot of the PCR product was size-fractionated on a 3 % low melting agarose gel stained with ethidium bromide.

comparison of this sequence (Fig. 6B) with sequences flanking the LEC-A IAP element indicated that the insertion of the LEC-A IAP element caused a 6-bp duplication (S'AACAAA3'). Three independent LEC antigen-loss variants had been obtained by selecting in v i m for resistance to CTL directed against antigen LEC-A. No 10.5-kb band was visible on the Southern blot prepared with BamH1-digested DNA of these variants and no PCR product was obtained with the pairs of primers flanking the 5' and the 3' LEC-A IAP integration sites (data not shown). This demonstrates that the transposed IAP underwent a deletion in these variants, and that it directs the expression of antigen LEC-A.

4 Discussion We have shown that antigen LEC-A,which is defined as the target of CTL clone LECl :5 directed against spontaneous leukemia LEC, is encoded by a defective endogenous retroviral sequence of the murine IAP family. A-type particles encoded by these proviruses are assembled on the membrane of the endoplasmic reticulum (ER) and bud into the cisternae, hence their name intracisternal A particles (IAP). A high level of IAP transcripts is found in a wide range of tumors and in pre-implantation embryos, whereas

The ability to delineate small gene fragments that transferred the expression of antigen LEC-A led to the identification of the antigenic peptide recognized by CTL-LEC1:5. This peptide is encoded by the gag gene and is presented by the Dk class I molecule. The CTL clone directed against antigen LEC-A recognized at similar concentrations a decapeptide and a nonapeptide. This is in agreement with the observation that naturally occurring class I-bound peptides are 8-10 amino acids long [36]. To our best knowledge, this is the first identification of a H-2 Dk binding peptide. It has been shown that both mutations and gene activation can account for the appearance of new antigens that are encoded by the cellular genome and can be recognized by CTL. Mutations can generate new antigenic peptides capable of binding to class I molecules [26] or they can generate a new epitope for CTL recognition on peptides that are already capable of binding [37]. Gene activation is responsible for the expression of antigens P815A and B on mastocytoma P815 [14].The relevant gene is identical to its counterpart from normal syngeneic kidney. It is highly expressed in the P815 tumor and in other mastocytomas but it is silent in normal mast cells and other normal tissues except testis and placenta (Van den Eynde, unpublished results). On human tumors, the activation of genes MAGE1 and MAGE-3 is also responsible for the expression of antigens recognized by autologous CTL [38, 391. To understand the process leading to the expression of antigen LEC-A, we compared genomic DNA extracted from LEC, from two syngeneic tumors that were not recognized by CTL-LEC1: 5, and from syngeneic normal kidney cells. This revealed that in the LEC tumor the LEC-A IAP element had undergone a transposition. The transposed provirus was found to be deleted in three independent LEC subclones selected for loss of expression of antigen LEC-A, demonstrating that the antigenic peptide is indeed produced by the transposed LEC-A IAP sequence. Thus we can conclude that antigen LEC-A resulted from the transposition of a IAP sequence. But whether the antigen was produced by a mutation that occurred in the course of the transposition or by the transcriptional activation of the IAP as a result of the transposition is less clear. The transposition of retroelements occurs in two steps :a transcript is converted by reverse transcription into a duplex DNA copy, which is then integrated into a new genomic locus [40]. This process is highly mutagenic, with mutations almost certainly occurring at the reverse transcription step [40]. Most of them are located in the LTR. Mutations in coding regions could also be generated during

Eur. J. Immunol. 1994. 24: 2203-2212

reverse transcription, resulting in the production of new antigenic peptides. On the other hand, IAP expression could also be increased either by mutations affecting the LTR or by integration in a highly transcribed region of the genome. This would also result in the production of antigenic peptides. One observation suggests that LEC-A antigen may have resulted from the transcriptional activation of the IAP element rather than from a mutation: a number of transfectants obtained with cosmids were recognized by antiLEC-A CTL even though they did not carry an IAP element surrounded by the same sequence as the LEC-A IAF! This was established by PCR analysis with primers corresponding to the LEC-A IAP sequence and to the surrounding sequences. Our interpretation is that other IAP sequences present in the LEC genome may produce the same antigenic peptide when they are activated as a result of transfection. This suggests that the antigenic peptide of LEC-A is encoded by several IAP sequences and is therefore probably not produced by a point mutation. But the argument is not conclusive. IAP elements that are able to confer antigen LEC-A expression are not transcribed in many cells. Several H-2k cell lines were tested for CTL recognition, including two other syngeneic tumors (one leukemia and one adenocarcinorna), one lymphoma, one myeloma, one melanoma and six fibroblastic cell lines. None of these cells were recognized by the CTL. Transpositions of IAP elements are occasionally associated with tumoral transformation because they can activate the genes located near the transposition site. Such gene activation has been reported for several proto-oncogenes and for genes encoding growth factors, growth-factor receptors, cytokines, and cytokine receptors [34]. For instance, bone marrow-derived mast cell cultures that are 1L-3 dependent can occasionally form autocrine IL-3secreting tumors when they are transfected with a rus gene [41]. For one tumor, the insertion of a IAP sequence in front of the IL-3 gene was shown to enhance transcription [42]. It is therefore possible that the transposition of the LEC-A IAP element may at the same time have contributed to tumor formation and caused tumor antigenicity. Endogenous retroviral elements make up at least 0.1 to 0.6 % of the human genome and are transcribed in various tissues and cell lines. In some cases the expression of these elements appears to be associated with neoplasia [43].This suggests that endogenous retroelements could also encode antigen recognized by CTL on human tumors. We thank M . Swinarska and C. Vervaetfor their excellent technical assistance.

Received May 18, 1994; accepted June 14, 1994.

5 References 1 Gross, L., Cancer Res. 1943. 3: 326. 2 Prehn, R. T. and Main, J. M., J. Natl. Cancer Inst. 1957. 18: 769.

A CTL clone recognizes an IAP-encoded antigen

221 1

3 Klein, G., Sjogren, H., Klein, E. and Hellstrom, K. E., Cancer Res. 1960. 20: 1561. 4 Kripke, M. L. and Fisher, M. S., J. Natl. Cancer Inst. 1976.57: 211. 5 Sjogren, H. 0.. Hellstrom, I. and Klein, G., Exp. Cell Rex 1961. 23: 204. 6 Khera, K. S., Ashkenasi, A., Rapp, F. and Melnick, J. L.. J. Immunol. 1963. 91: 604. 7 Klein, E . and Klein, G., J. Natl. Cancer Inst. 1964. 32: 547. 8 Hewitt, H., Blake, E. and Walder, A., Br. J. Cancer 1976.33: 241. 9 Van Pel, A.,Vessi&re, F. and Boon,T., J. Exp. Med. 1983. 157: 1992. 10 Brunner, K. T., MacDonald, H. R. and Cerottini, J.-C.. J. Immunol. 1980. 124: 1627. 11 Maryanski, J. L.,Van Snick, J., Cerottini. J. C. and Boon, T.. Eur. J. Immunol. 1982. 12: 401. 12 Palladino, M. A., Srivastava, P. K., Oettgen. H . F. and DeLeo. A. B., Cancer Res. 1987. 47: 5074. 13 Uyttenhove, C., Maryanski, J. L. and Bo0n.T.. J. Exp. Med. 1983. 157: 1040. 14 Van den Eynde, B., LethC, B.,Van Pel, A.. De Plaen, E. and Boon, T., J. Exp. Med. 1991. 173: 1373. 15 Van Pel, A., De Plaen, E . and Boon, T., Somat. Cell Mol. Genet. 1985. 11: 467. 16 Nicolas, J. F. and Berg, I?, CSH Conferences Cell ProIiJ 1983. 10: 469. 17 Bernd, A., Burgert, H. G., Archibald, A. L. and Kvist. S.. Nucleic Acids Res. 1984. 12: 9473. 18 Stephan, D., Sun, H., Fischer Lindahl, K., Meyer, E., Hammerling, G., Hood, L. and Steinmetz, M., J. Exp. Med. 1986. 163: 1227. 19 Corsaro, C. M. and Pearson, M. L., Somat. Cell Mol. Genet. 1981. 7: 603. 20 Wolfel,T.,Van Pel, A., De Plaen, E., Lurquin, C., Maryanski, J. L. and Boon, T., Immunogenetics. 1987. 26: 178. 21 Bernard, H.-U., Krammer, G. and Rowekamp, W. G., Exp. Cell. Biol. 1985. 158: 237. 22 Maryanski, J. and Boon, T., Eur. J. Immunol. 1982. 12: 406. 23 Boon, T., Van Snick, J., Van Pel, A.. Uyttenhove. C. and Marchand, M., J. Exp. Med. 1980. 152: 1184. 24 Espevik, T. and Nissen-Meyer. J., J. Immunol. Mefhods 1986. 95: 99. 25 Traversari, C., van der Bruggen, P., Van den Eynde. B.. Hainaut, P., Lemoine, C., Ohta, N., Old, L. and Boon. T.. Immunogenetics. 1992. 35: 145. 26 Lurquin, C.,Van Pel, A., Mariame, B., De Plaen, E.. Szikora. J.-P., Janssens, C., Reddehase. M. J., Lejeune. J. and Bo0n.T.. Cell. 1989. 58: 293. 27 Atherton, E., Logan, C. J. and Sheppard, R. C., J. Chem. Soc. Lond. Perkin Trans. 1981. I: 538. 28 Van Pel, A. and Boon,T., Proc. Natl. Acad. Sci. USA 1982. 79: 4718. 29 Perucho, M., Hanahan, D. and Wigler, M., Cell. 1980.22: 309. 30 De Plaen, E., Lurquin, C.,Van Pel, A., Mariame. B., Szikora, J.-P.,Wolfel,T., Sibille, C., Chomez, P. and B0on.T.. Proc. Natl. Acad. Sci. USA 1988. 85: 2274. 31 Lau,Y. F. and Kan,Y. W., Proc. Natl. Acad. Sci. USA 1983.80: 5225. 32 Mietz, J. A., Grossman, Z., Lueders, K. K. and Kuff, E. L.. J. Virol. 1987. 61: 3020. 33 Chomez, P., De Plaen, E.,Van Pel, A., De Smet, C., Szikora. J.-P., Lurquin, C., Lebacq-Verheyden. A.-M. and Boon. T.. Immunogenetics. 1992. 35: 241. 34 Kuff,E. L. andLueders, K. K..Adv. CancerRes. 1988.51: 183. 35 Aota, S., Gojobori,T., Shigesada, K.. Ozeki. H. and Ikemura. T., Gene 1987. 56: 1. 36 Falk, K.. Rotzschke, 0..Stevanovic, S., Jung, G. and Rammensee, H.-G., Nature 1991. 351: 290. 37 Sibille, C., Chomez, €?,Wildmann, C. ,Van Pel. A.. De Plaen. E., Maryanski, J. L., de Bergeyck,V. and Boon,T., J. Exp. Med. 1990. 172: 35.

2212

V. de Bergeyck, E. De Plaen, P. Chomez et al.

38 van der Bruggen, P..Traversari, C., Chornez, P., Lurquin, C., De Plaen, E.,Van den Eynde, B., Knuth, A. and Boon, T., Science 1991. 254: 1643. 39 Gaugler, B. ,Van den Eynde, B.,van der Bruggen, P., Romero, P., Gaforio, J. J., De Plaen, E., Leth6, B., Brasseur, F. and Boon,T., J. Exp. Med. 1994. 179: 921. 40 Heidmann, 0. and Heidmann, T., Cell. 1991. 64: 159.

J. Imrnunol. 1994.24: 2203-2212 41 Nair, A. P. K., Diamantis, I. D., Conscience, J.-E, Kindler,V., Hofer, P. and Moroni, C., Mol. Cell. Biol. 1989. 9: 1183. 42 Hirsch, H. H., Nair, A. P. K. and Moroni, C., J. Exp. Med. 1993. 178: 403. 43 Leib-Mosch, C., Brack-Werner, R.,Werner,T., Bachmann, M., Faff, O., Erfle, V. and Hehlmann. R., Cancer Res. (Suppl.) 1990. 50: 5636.