Identification of an epitope encoded in the env gene of friend murine leukemia virus recognized by anti-friend virus cytotoxic T lymphocytes

VIROLOGY

181, 51-t66

Identification

(1951)

of an Epitope Encoded in the env Gene of Friend Murine Recognized by Anti-Friend Virus Cytotoxic T Lymphocytes

Leukemia

Virus

KE-SAN RUAN AND FRANK LILLY’ Department

of Molecular

Genetics, Albert Einstein College of Medicine,

Received June 14, 1990; accepted

November

Bronx, New York 1046 1

14, 1990

We have previously shown that strong epitopes recognized by anti-Friend virus (FV) cytotoxic T lymphocytes (CTL) in H-P mice are encoded in both the env and gag&o/ regions of the helper friend murine leukemia virus genome. Two approaches have been used to identify these epitopes. At the nucleic acid level, we have constructed env genes with either of two in-frame deletions: pKR2, an env gene with a 681 -bp deletion in the gp70 region and inserted into the pSV2-gpt-1 expression vector; and pKR1, an env gene with an 81 -bp deletion in the pl5E region and inserted into pSV2-gpt-1. Cell clones were established by transfecting Fisher rat embryo cells with pDb (the H-2Db restriction element), pNE0 (for G418 selection) and either pKR1 or pKR2. Db and env gene expression was monitored by immunoprecipitation with polyclonal antibodies or by detection of viral RNA on Northern blots. Expressor cell clones were tested for susceptibility to lysis by polyclonal anti-FV/Db CTL in “Cr-release assays. Whereas cells expressing pKR1 were lysed to the same extent as cells expressing the intact env gene, cells expressing pKR2 were resistant to lysis, suggesting that all detectable env epitopes are encoded within the 681-bp deletion. Polypeptides representing the two most likely candidate epitopes encoded in this segment were synthesized and tested for their abilities to sensitize FRE cells expressing Db alone for lysis by the CTL. One 17-mer polypeptide, AGTGNRCCNCYEGAYEA, functioned as a strong CTL epitope in this assay, but the other 18-mer polypeptide was inactive. Studies of the role of this epitope in the immune response to candidate viral vaccines are in progress. o is% Academic PWSS, inc.

INTRODUCTION

genes (Gotch et al., 1987a). In several cases, specific oligopeptide sequences of viral proteins have been demonstrated to function as epitopes recognized by populations of virus-specific CTL when presented by particular class I molecules (Townsend et a/., 1985; Gotch et al., 198713). Previous studies in our laboratory have shown that immunization of mice carrying the H-P MHC haplotype with syngeneic spleen cells infected with Friend erythroleukemia virus (FV) generates a strong CTL response that is restricted exclusively by the Db class I MHC molecule (Plata and Lilly, 1979). Viral epitopes recognized by these CTL populations were found to be encoded in both of the viruses that together comprise FV: the defective, pathogenic spleen focus-forming virus (SFFV) and its native helper, Friend murine leukemia virus (FMuLV) (Holt eta/., 1986). Furthermore, both the gag and env genes of the FMuLV genome encode strong CTL epitopes (Holt et al., 1986). We now report studies that identify a peptide sequence in the FMuLV envgene that, in the presence of the Db molecule, acts as a strong epitope recognized by anti-FV/Db CTL.

Virus-specific cytotoxic T lymphocytes (CTL) lyse infected target cells based on their ability to recognize processed viral antigens in the context of class I molecules encoded in the major histocompatibility complex (MHC) of the cells (Zinkernagel and Doherty, 1979). The recent crystallographic resolution of the structure of class I MHC molecules has elucidated the basis for this phenomenon, termed MHC restriction, by defining a groove on the extracellular surfaces of these molecules into which stereochemically appropriate peptide fragments of viral gene products may be fitted (Bjorkman et a/., 1987). Intracellular degradation of viral proteins synthesized in infected cells produces peptide fragments that are available for loading into the grooves of any class I MHC molecules that can accept them (Townsend et a/., 1986b). The class I molecules thus armed with a protein fragment, or epitope, are translocated to the cell surface for presentation to epitape/class I molecule-specific CTL. Viral epitopes have been mapped to a number of specific viral genes, including those for the influenza hemagglutinin (Bennink et a/., 1984; Braciale et al., 1984; Townsend et a/., 1984), nucleoprotein genes (Townsend et al., 1984; Yewdell et a/., 1985), and the matrix and polymerase

MATERIALS

AND METHODS

Mice and cells Inbred BALB.B mice (H-26) were bred in our own colony. Fisher rat embryo (FRE) clone 2 fibroblasts were

’ To whom requests for reprints should be addressed. 91

0042.6822/91

$3.00

CopyrIght 0 1991 by Academic Press, Inc. All rights of reproduction I” any form resewed.

RUAN AND LILLY

92

used as the parental cell line for the establishment of target cells for CTL (Scolnick et a/., 1976). DF-45 is a clone derived from FRE cells expressing complete FV (SFFV and FMuLV) and the ob restriction element; N34 is an FRE clone expressing the FMuLV env gene and ob; and B2 is a clone of FRE cells expressing ob only (Holt et al., 1986). FRE cells and derivatives were maintained in Dulbecco’s minimal essential medium (DMEM) supplemented with 10% heat-inactivated fetal calf serum (FCS), 2 mM L-glutamine and 20 U/ml penicillin-streptomycin. Antisera Expression of mouse H-2 and FMuLV pl5E proteins were detected using monoclonal antibodies (ascites) from hybridoma cell lines 28-l 4-8 (anti-Db) (Ozato and Sachs, 1981) and 273 (anti-p1 5E) (Chesebro et al., 1981). Goat anti-Rauscher murine leukemia virus gp70 serum was supplied through the Viral Oncology Division, National Cancer Institute (Bethesda, MD). Antisera were heat-inactivated at 56” for 30 min. Lymphocyte

cultures

Spleen cells from BALB.B mice infected 1O-l 4 days earlier with a high dose of FV were fixed by brief exposure to glutaraldehyde (Bubbers and Henney, 1975). Treated cells (2.5 X 107) were inoculated ip into 6- to 8-week-old female BALB.B recipients; 4 weeks later the mice received a second immunization. Immune spleen cells were harvested 2 weeks later, and syngeneic mixed lymphocyte-tumor cell cultures were established by restimulating 2.5 X 1O7 responder cells in vitro with 1O6 irradiated (5000 rad) FV-infected BALB.B spleen cells in RPMI culture medium (GIBCO) with 10% FCS. After 6 days at 37”, responder immune cells were harvested and used in assays for cytolytic activity. Construction

of viral expression

vectors

Plasmid pFe1 1 was constructed by subcloning the Hindlll fragment (essentially the FMuLV env gene) of plasmid 2-l al c (FMuLV) into pBR322 (see Fig. 1). This HindIll fragment contains two BstEll restriction sites spanning 681 bp in the gp70 coding region of the env gene and two Stul sites spanning 81 bp in the p15E region. Deletion of either the BstEll or Stul fragment maintains the open reading frame of the env gene. After double digestion of pFe1 1 with HindIll and Stul, the 2.4- and 1.7-kb restriction fragments of the env gene insert were isolated, ligated, and then linearized by redigestion with Hindlll. Plasmid pKR1 was constructed by inserting this fragment into the HindIll site of the expression vector pSV;Z/gpt (Mulligan and Berg,

1980). Using the same strategy, vector pKR2 was constructed by double digestion of pFe1 1 with HindIll and BstEll and recovery of the resulting 1.4- and 2.1-kb fragments of the insert. Establishment

of target cell lines for CTL assays

FRE cells were recipients for cotransfection with viral and restriction element DNAs using a calcium phosphate transfection method (Chen and Okayama, 1988). Exponentially growing cells were trypsinized, seeded at a density of 5 X 1O5 cells in 10 ml of growth medium in T75 flasks, and incubated at 37” overnight. Intact plasmid DNAs [lo fig pKR1 or pKR2, 10 pg pDb (Allen et a/., 1984) and 1 pg of the dominant selection marker pSV2neo (Southern and Berg, 1982) dissolved in 50 ~1 H,O] were coprecipitated by sequential addition of 0.5 ml of 0.25 M CaCI, followed by 0.5 ml of 2X BBS (50 mM Iv-/II-bis(2-hydroxyethyl)2-amino-ethanesulfonic acid, 280 mM NaCI, 0.5 mM Na,HPO,, adjusted to pH 6.96 with 1 N NaOH at room temperature); the mixture was left at room temperature for 20 min. All solutions were passed through 0.45~pm filters (Nalge). The calcium phosphate-DNA mixture was added dropwise to a single recipient cell culture with gentle mixing, and the culture was incubated overnight at 37”. The culture was then split at a ratio of 1:5 and incubated in medium containing 0.8 mg/ml of G418 (GIBCO). After 2 weeks, individual colonies of G418resistant cells were visualized and isolated with cloning cylinders. Transformation frequency was approximately 2-3 X 1O4 clones/pg pSV2neo. lmmunoprecipitation Subconfluent cell cultures in T75 flasks were incubated for 30 min at 37” in 2 ml leucine-free MEM containing 100 &i of [3H]leucine (Amersham) and 1% dialyzed FCS. Labeled cells were washed with cold phosphate-buffered saline and lysed with 2 ml extraction buffer (10 mM sodium phosphate buffer, pH 7.6, 1 mM EDTA, 100 mM NaCI, 1% Triton X-100, 0.5% deoxycholate, and 0.1% SDS) for 1 hr at 4”. Supernatants from the lysates were precleared by incubating with normal goat serum (5 @g/ml) on ice for 30 min; 50 ~1 of 10% Staph A (BRL) was added, and the mixture was shaken at 4” for 30 min. lmmunoprecipitation was carried out by incubating 1 ml of supernatant with 10 ~1 of antiserum for 30 min at 4’; 50 pg of Staph A was added and the mixture shaken for 30 min at 4”. After centrifugation, the pellets were washed three times with cold extraction buffer, then resuspended and boiled for 2 min in electrophoresis buffer [50 m&‘TrisHCI, pH 6.8, 1O/oSDS, 1% 2-mercaptoethanol, 20%

IDENTIFICATION OF A FRIEND MLV env GENE EPITOPE

glycerol, and 0.001% bromphenol blue]. The solutions were electrophoresed on 7 or 9% NaDodSO,-polyacrylamide gels with 3.59/ostacking gels. 14C-labeled protein markers (BRL) were electrophoresed simultaneously. Gels were fixed and dried, and autoradiograms were made using Kodak XAR-5 film at -70”. Detection of cell-surface expression of gp70 antigens Surface expression of antigens detected by mouse monoclonal anti-gp70 No. 55 (Chesebro et a/., 1981) followed by fluoresceinated goat anti-mouse IgG Fc serum (Cappel, West Chester, PA) was studied on cells of clones KRl-2 and KR2-3 using a Becton-Dickinson FACSCAN cell sorter. Positive and negative control cells were from clone N34 (FMuLV env and ob) and clone B2 (ob only), respectively (Holt et a/., 1986). RNA hybridization RNA isolated from cell cultures by the guanidium method (Ausubel et a/., 1987) was loaded at 5 fig/lane on 1O/O agarose gels containing 8% formaldehyde, 20 mll/l morpholinosulfonic acid, 1 mM Na,EDTA, and 0.5 pg ethidium bromide. After electrophoresis, RNA was transferred to a Nitron membrane (Schleicher & Schuell). DNA probes were labeled with [a-32P]dCTP by the multiprimer DNA labeling system (Amersham) and applied to the membrane at a concentration of 2 X IO6 cpmlml in a buffer containing 50% v/v formamide at a final concentration of 3X SET (20X SET = 3 M NaCI, 0.02 n/r EDTA, 1 1\/1Tris-HCI, pH 7.9) 0.1 AII sodium phosphate, pH 7.0, 5X BFP solution (100X BFP = 2% bovine serum albumin, 2% Ficoll, 2% polyvinylpyrrolidone), and 100 pg/ml of denatured salmon sperm DNA for 16-20 hr at 50”. Prehybridization was carried out at 37” for 2-4 hr in the same buffer. The membrane was washed three times for 30 min at 37” with 2X SET buffer containing 0.1 n/l sodium phosphate buffer and 0.2% NaDodSO,. Finally the membrane was rinsed in 1x SET buffer, dried, and exposed for 24-48 hr to x-ray film (Kodak XK5) at -70” with an intensifying screen. Reverse transcription and polymerase chain reaction (RT-PCR) amplification of viral mRNA sequences and DNA sequencing Total RNAs from FRE cells and from clones KRl-2 and KR2-3 were prepared as described above. Two 20-mer DNA oligonucleotides were synthesized as primers for amplification of viral env sequences: Pl (5’CGGTGCAACACTGCCTGGAA-3’) and P2 (5’-CGAACCATCCTTGGCTCGAC-3’) (see Fig. 5). RT-PCR was

93

performed by the method of Kawasaki (1990). cDNAs were made by treating 5 pg of RNA with RNAse-free DNAse and annealing with 10 ng of primer P2, and 5 units of AMV reverse transcriptase was added in a total volume of 20 ~1at 50” for 20 min. Primer P2 was used together with AMV reverse transcriptase for making cDNAs, and primers Pl and P2 were used to carry out 30 cycles of the PCR reaction on each cDNA preparation (Kawasaki, 1990). Each cycle consisted of 1-min denaturation (94”) followed by 2-min annealing (55”) and 2-min extension (72”) steps. DNA products were extracted once with chloroform, precipitated with 0.1 vol of 3 M sodium acetate and 1 vol of isopropanol at room temperature, washed once with 70% ethanol, and dissolved in 20 ~1of TE buffer (10 rnM Tris, 1 mM EDTA, pH 7.8). PCR-amplified DNA from clone KR2-3 was sequenced using an internal 5’ oligonucleotide primer [P3 (5’-ACTAGCCAGGCTGTCCAGGT-3’)] (see Fig. 5) based on the chain termination sequencing reaction (Sanger et al., 1977) using Sequinase version 2.0 (USB Corp., Cleveland). Cytotoxicity assay Cytolytic activity was determined using a 51Cr-release assay (Holt et al., 1986). Briefly, 1O45’Cr-tabeled target cells were mixed and incubated for 6 hr at 37” with varying numbers of CTL in a final volume of 0.2 ml Iscove’s modified DMEM (GIBCO) containing antibiotics, L-glutamine, and 10% FCS. 51Crrelease for each point was calculated from the mean radioactivity detected in four replicate samples, and the data are presented using the formula 4/o51Crrelease = 100 x [(experimental release - spontaneous release)/(maximum release - spontaneous release)], where the spontaneous release was obtained from target cells incubated in medium alone, and the maximum release was from target cells incubated in 1 N NaOH. Peptide synthesis and epitope analysis Two candidate CTL peptide epitopes encoded within the 681-bp deletion of pKR2 were synthesized on an Applied Biosystems 430A instrument in the Laboratot-yfor Macromolecular Analysis (Albert Einstein College of Medicine). The peptides were cleaved from the resin with hydrogen fluoride and partially purified on a Sephadex G-l 0 column. Peptide compositions were verified by amino acid analysis and by mass spectrometry on a Finnegan MAT 90 instrument. The peptides were further purified by high-pressure liquid chromatography on a C8-300A Dynamax column. For determination of their capacities to sensitize target cells for lysis by CTL populations (Townsend et al., 1986b; Chi-

RUAN AND LILLY

94

mini et a/., 1989), purified peptides were lyophilized, dissolved in PBS, and sterilized by filtration. 5’Cr-labeled B2 target cells were incubated in 0.2 ml of RPM/ 10 containing varying concentrations of peptide, and CTL were added at an effector-to-target cell ratio of 100: 1, 33: 1, or 1 1: 1. After incubation for 6 hr at 37”, 5’Cr releases, based on the means of four replicate samples for each point, were calculated as described above. Minimum releases were determined by incubating B2 cells in a high concentration of peptide in the absence of CTL, and maximum releases were determined from B2 cells incubated in 1 N NaOH. RESULTS Expression genes

vectors for partially

deleted FMuLV env

Plasmid 2-l al c (Holt et a/., 1986) derived from the circularly permuted FMuLV,, genome in plasmid 2-l al (Oliff et al., 1980), contains a nonpermuted copy of the FMuLV genome with two long terminal repeats (LTR) and flanking viral sequences, cloned into pBR322 (Fig. 1). Plasmid pFe1 1 was obtained by subcloning the HindIll fragment of plasmid 2-lalc into pBR322; the insert begins near the 3’ end of the pal gene and includes the entire env gene followed by a complete 3’ LTR. A previous study in our laboratory showed that the FMuLV env gene encodes at least one strong epitope recognized by FV-specific CTL generated in BALB.B mice (H-2b) by immunization with syngeneic FV-infected spleen cells (Holt et al., 1986). In this study we established a cell clone, N34, by cotransfection of cells of the FRE line with both pSV2Fe/gpt-1 (the FMuLV HindlIt fragment in pFe1 1 subcloned into the pSV2/gpt eukaryotic expression vector), pDb [a genomic clone of the H-2ob gene (Allen et a/., 1984)], and pSV2neo [a dominant selection marker (Southern and Berg, 1982)]. N34 cells were highly susceptible to lysis by BALB.B anti-FV/Db CTL, confirming that the H-2Db molecule is the restriction element used for presentation of FV Tcell epitopes in these mice. We then wished to define more precisely the location of this FV epitope within the FMuLV env gene. Inspection of the restriction map of the pFe1 1 insert revealed that two restriction enzymes, BstEll and Stul, recognize sites represented twice each within the gp70 and pl5E regions, respectively, of the env gene; in each case deletion of the intervening DNA would maintain the original reading frame. Plasmids pKR1 and pKR2 were constructed by deleting either the Stul fragment (8 1 bp located in the 5’ portion of the pl5E region) or the BstEll fragment (681 bp encoding most of the 3’ half of the

gp70 region), respectively, from the pFe1 1 insert and introducing the deleted variants into the Hindlll site of pSV2/gpt (Fig. 1). Characterization

of transfected

cell clones

Two cell clones obtained from the pKRl/pDb transfection and 10 from the pKR2/pDb transfection were metabolically radiolabeled. Expression of the Db molecule was tested by immunoprecipitation of cell extracts using anti-Db monoclonal antibody 28-l 4-8 followed by autoradiography of the electrophoretically separated proteins. Figure 2 shows that the Db protein was expressed at relatively high levels in 3 of the 12 clones: KRl-2, KR2-3, and KR2-8. By immunoprecipitation with a polyclonal anti-gp70 serum, a metabolically labeled extract from clone KRl-2 produced a strong band representing the slightly truncated env gene precursor, gPr85 (Fig. 3). However, no cell-surface expression was detected on KRl-2 cells by anti-gp70 serum in a fluorescence-activated cell sorter study, even though control cell clone N34, transfected with the intact parental envgene construct, was strongly positive in the same experiment (data not shown). No env protein was detected by anti-gp70 in any clone of the KR2 series either by immunoprecipitation (Fig. 3) or by FACS studies. Presumably the large deletion in the gp70 coding region of pKR2 either eliminated most antibody binding sites in the molecule or caused the translation product to be degraded too rapidly for detection. While immunoprecipitation of an extract from clone KR2-3 with a monoclonal anti-p1 5E antibody produced a weak band of the expected 21.8-kDa size in a markedly overloaded lane (data not shown), this finding was only suggestive evidence for expression of the truncated env gene. To obtain further evidence for env gene expression in clones of the KR2 series, we examined the cells for expression of viral RNA. RNA was extracted and analyzed by Northern blotting, using as a probe the 32P-labeled Xbal-Hindlll fragment from pFe1 1 that includes the entire FMuLV env gene and 3’ LTR (Fig. 1). Three clones, KR2-3, KR2-8, and KR2-10, expressed significant levels of RNA of the 2.75-kb size expected for the pKR2 transcript, but each also showed significant levels of RNAs 3.05-and 5.5-kb in size, presumably representing (i) mostly vector pSV2/gpt sequences transcribed from the 3’ LTR of the viral insert and (ii) the entire length of the plasmid from the SV40 promotor to the poly(A) site of the vector Eco-gpt gene, respectively (Fig. 4). To confirm the accuracy of RNA corresponding to the deleted env genes in the KRl-2 and KR2-3

IDENTIFICATION

OF A FRIEND MLV env GENE EPITOPE

95

FMuLV(P-lalc)

/

\

/

pFell

,

/

/

/

/

/

\

IL1

+

-

L??b-

k

-

-

-

1 1

gp70

XI

H

\

++ - -

POl

H

\

peplId@ A ppllti B

2-F

_ -

pl5E

e

,

k-------------I -

-

gq

1

1

pBR322 H

ss

e

-

LTR

-

,fi

2.1 kb -

-

L7E

_

_

,

Eco-gpl

s

S

FIG. 1. Construction of FMuLV env gene deletion vectors. Deletions of either the Stul fragment (81 bp) or the BstEll fragment (681 bp) were made in the HindIll fragment of 2-l al c (FMuLV). The resulting truncated fragments were inserted into the HindIll site of pSVZ/gpt to produce the pKR1 and pKR2 expression vectors, respectively. C, C/al; R, EcoRI; H, HindIll; X. Xbal; B, BstEll; S, Sful; P, Pvull.

clones, cellular RNA was subjected to reverse transcription followed by PCR amplification to synthesize DNA fragments spanning the sites of the two envgene deletions. The PCR products showed the expected sizes of 1.35 kb (KRl-2) and 0.75 kb (KR2-3), which correspond to the nucleotide sequences between the primers used (Fig. 5). Treatment of the two PCR products with Sacll further confirmed their identities by producing subfragments of the predicted sizes (Fig. 5). Finally, DNA sequencing across the deletion site in KR2-3 confirmed that the RNA accurately reflected the desired deletion (data not shown). Given the abundant presence of the correct viral RNA in KR2-3 cells, it seems likely that the markedly truncated envgene was translated, but that the protein product was rapidly de-

graded in these cells. After transient expression, protein degradation would have been by the same intracellular mechanisms by which the normal protein is processed. Thus any epitopes remaining in the truncated protein would be available for presentation by the H-2Db molecule to T lymphocytes. Identification

of a CTL epitope within

FMuLV gp70

FRE clones expressing pKR1 (KRl-2) or pKR2 (KR23, KR2-8, and KR2-10) were examined in a 5’Cr-release assay for their sensitivity to lysis by BALB.B antiFV/Db CTL. Cells of clone KRl-2, which express a truncated FMuLV env gene in their cytoplasm but not on their cell surfaces, were as susceptible to lysis as

RUAN AND LILLY

96

pKR2 pKRl I II 8 std I 2 3 4 5 6 7 8 9 IO I 2

~PKR+J 3

81OE

4

FIG. 2. Expression of the H-2Db protein in transfected FRE cell clones. Cells from 2 clones of the KRl series and 10 of the KR2 series were metabolically radiolabeled, and extracts were immunoprecipitated with anti-Db monoclonal antibody 28-l 4-8. Precipitates were separated by gel electrophoresis and visualized by autoradiography. Bands representing the 45.kDa product were detected In clones KRl-2, KR2-3, and KR228.

the control line, DF-45, an FRE clone expressing complete FV (FMuLV and SFFV) plus the Db restriction element. In contrast, clones KRZ-3, KR2-8, and KR2-10 were not detectably lysed by the CTL (Fig. 6). This finding suggests that all significant FMuLV env gene epitopes recognized by anti-FV/Db CTL are encoded in the 681 bp of the gp70 region that were deleted from pKR2. Analysis by the algorithms of both Berzofsky (DeLisi and Berzofsky, 1985; Margalit et al., 1987) and Rothbard (Rothbard and Taylor, 1988) of the amino acid sequence encoded in the 681-bp deletion of pKR2 suggested two candidate polypeptide sequences that might serve as T-cell epitopes. These two sequences, a 17-mer (peptide A) representing amino acids 318-

pKR2 pKRl std I I 2 3 4 5 6 7 8 9 IO , I 2 ,

FIG. 4. Expressron of FMuLV env gene mRNA in clones of the KR2 series. For Northern blotting, RNA from cells of the KR2 series of clones was separated on a 1% agarose-formaldehyde denaturing gel, transferred to a Nitron membrane, and probed with the “P-labeled Xbal-HindIll fragment from pFe1 1 that includes the entire FMuLV env gene and 3’ LTR. The 2.75-kb band represents the transcript of the env gene containing a 681 -bp BstEll deletion. The 3.05.kb band consists mostly of vector pSV2/gpt sequences transcribed from the 3’ LTR of the vrral insert The 5.5.kb band corresponds to the entrre length of the plasmrd from the SV40 promotor to the poly(A) site of the Eco-gpt gene In the vector.

334 of the env gene product (AGTGDRLLNLVQGAYQA) and an 18-mer representing amino acids 398-415 (TVPKTHQALCNTTLKIDK), were synthesized. B2 cells, a clone of FRE transfected with pDb only, were incubated in the presence of various concentrations of either peptide and tested for susceptibility to lysis by BALB.B anti-FV/Db CTL. Peptide A strongly sensitized the cells in a concentration-dependent manner for lysis, but peptide B was ineffective in this assay (Fig. 7).

DISCUSSION We showed previously that anti-FV CTL generated in FIG. 3. Expression of gp70 in transfected FRE cell clones. Radiolabeled cell extracts, as in Fig. 2, were rmmunoprecipitated with a polyclonal anti-gp70 serum. After gel electrophoresis and autoradiography, a band representing gPr85 was seen only in clone KRl-2; no clone of the KR2 series expressed a detectable gp70-related product.

H-P mice are strongly lytic when tested on target cells of the FRE line expressing transfected constructs encoding the intact FMuLV env gene and the Db class I restriction element (Holt et al., 1986). We have now constructed an env gene vector from which we deleted

IDENTIFICATION

OF A FRIEND MLV env GENE EPITOPE

681 bp

FMuLV-env

6169

6391

I_ - - - - - -gp,o

BstE II

- - - - - -t

sac n

--

StuI

7600

- - -plSE

- - -1

FIG. 5. FMuLV env gene-specific DNAs obtained by reverse transcription and PCR amplification of cellular RNAfrom FRE cells (lanes 4 and 5) and their derivative clones KR2-3 (lanes 2 and 3) and KRl-2 (lanes 6 and 7) revealed by ethidium bromide in a 1.5% agarose gel. Primer P2 was used for reverse transcription, and primers Pl and P2 were used for PCR amplification; primer P3 was used for DNA sequencing (not shown). The first lane of each pair shows the intact PCR DNA products, and the second lane shows these products after restriction with Sacll. In lane 8, the small, 242-bp Sacll fragment of the KRI-2 DNA product was lost from the gel, as were the molecular weight markers smaller than 3 10 bp (lane 1). The sizes of the bands are consistent with predictions based on the restriction map of the gene, as shown.

97

gesting that all CTL epitopes of the env gene were encoded within the deleted region. Analysis of the 227-amino-acid sequence encoded in the deleted BstEll fragment by two different algorithms to identify candidate T-cell epitopes suggested two likely sequences for consideration. We synthesized these polypeptides, of which one (peptide A) proved to possess the ability to sensitize cells expressing the Db protein for lysis by anti-FV/Db CTL at the same high level as a control FRE target cell clone expressing Db plus complete FV (FMuLV and SFFV). Given the strength of the epitope represented by this peptide, it may prove to be the only one recognized by these CTL in the FMuLV env gene. We also studied a second env gene construct containing a much smaller in-frame deletion in the p15E region; it retained both the peptide A-encoding sequences and a strong capacity to sensitize FRE cells for CTL lysis. Anti-gp70 antibodies readily precipitated a cytoplasmic env gene product from these transfected cells, but immunofluorescence studies failed to detect its expression on cell surfaces. This finding is consistent with other studies indicating that T-cell rec-

.-.

DF45 (pasHlvm conlml)

l -9

pKRl-2

x-x

pKR2-3

0-D

pKR2.8

o -0

pKR2.10

1 30 1

a large BstEll fragment encoding 227 amino acids in the 3’ portion of gp70e”” for use in an analogous study. Given the large size of this deletion, which comprises about half of the gp70-encoding sequences, it is not surprising that a polyclonal goat anti-gp70 antiserum did not precipitate a gene product, even though 3 of 10 FRE cell clones cotransfected with this deletion construct and the Db protein expressed relatively high levels of env gene-specific RNA accurately reflecting the desired deletion. Since the pl5E-encoding region of the env gene construct remained intact and in the proper reading frame, we used an anti-p1 5E antibody in an attempt to demonstrate the presence of this protein in the three RNA expressor cell clones. One heavily overloaded lane (clone KR2-3) showed a weak band corresponding to the calculated weight of pl5E, thereby providing suggestive evidence for transient expression of the truncated envgene product. Anti-FV/Db CTL had no Iytic effect on FRE clones expressing both the truncated env gene RNA and the Db protein, sug-

. .

. 20-

.

/

>; . lo0

:,<,

#

8

33

100

I 11

Effector : Target Cell Ratio FIG. 6. Recognition by polyclonal anti-FV/Db CTL of cells of the KRl and KR2 clones. CTL were generated by immunizing BALBB mice twice with syngeneic, FV-infected spleen cells. Spleen cells from these mice were restimulated in culture for 6 days and assayed for lytic activity against “0-labeled cells of clones of the KRl and KR2 series. Each point in the figure represents the mean of the specific 51Cr release from four replicate samples.

WAN

98

AND LILLY

B 25mM

pep

l

5mM

pep

.

/:

1mM

pep

Etfector:

Effector:

0.2mM

Target Cell Ratio

pep

.-.

pep/L

O-0

pep6

Target = 100: 1

FIG. 7. Lysis by anti-FV/Db CTL of FRE-Db cells in the presence of gp70 pepttdes A or 8. (A) “Cr-labeled 82 cells (a clone of FRE transfected only with the f?’ gene) were incubated in the presence of various concentrations of either peptide A or peptide B for 6 hr. Anti-FV/Db CTL were added at an effector:target cell ratio of 100: 1. “Cr releases were the mean values of four replicate samples. The positive control was cell clone DF45 (expressing Db, FMuLV, and SFFV) in the absence of peptrdes. The negative control was 62 cells plus 25 mM peptide but with no CTL added. (f?) The same experiment as in A carried out at different effector:target cell ratios.

ognition does not depend on cell-surface expression of intact antigen molecules. A noteworthy observation in our earlier studies was that FRE cells expressing SFFV and Db but not FMuLV were highly sensitive to lysis by CTL generated in H-2’ mice immunized with ceils expressing either complete FV (FMuLV plus SFFV) or FMuLV alone, a finding that indicated the existence of a common CTL epitope encoded in both viral genomes (Holt et a/., 1986). A search for the amino acid sequence of the peptide A l7-mer in the SFFV genome (Clark and Mak, 1983) shows that a 17-amino-acid sequence encoded in the SFFV env gene differs from peptide A in only its aminoterminal residue (Fig. 8). Furthermore, a sequence differing from peptide A in only 3 of its 17 residues is

present in the env gene of the Moloney strain of MuLV (Shinnick et a/., 1981) suggesting that the cross-reactivity of anti-FV/Db CTL on leukemias induced by this virus in H-P mice (Plata and Lilly, 1979) might be due to recognition by the CTL of this same epitope. These same sequences are present in the mink cell focus-inducing virus isolates derived from FMuLV and MMuLV, respectively, so that cells of H-2“ leukemias induced by these viruses may also be sensitive to lysis by anti-FV/ Db CTL. Thus, this epitope may be a major factor that defines the classical FMR subdivision of the murine retrovirus family with respect to the cellular immune response. The CTL epitope identified in these studies is recognized only in the context of the Db class I molecule on

IDENTIFICATION psptlde

OF A FRIEND MLV env GENE EPITOPE

99

R

~TGDRLLNLUQGFI!”

FIluLU

276:

PIPTQPPP-

LNLTNPDKTQECULCLUSGPPYYEGUAULGTYSHTSAPANCSURSQH

SFFU

270;

G-RppSQQp-----------------------------------------------N-----LKEKCCF~RD

tlHuLU

3,,2;

LS---L-----EN------D---------S------------R-----------------------------

FIG. 8. Comparison of amino acid sequences encoded in the FMuLV env gene (Koch et a/., 1984) with related sequences and Mak, 1983) and Moloney MuLV (Shinnick ef a/., 1981) envgenes. Numbers are those of the first amino acid listed, starting leader methionine. The 17-mer peptide A sequence is indicated.

H-P target cells. Our earlier work indicates that the companion Kb class I molecule of the H-26 haplotype does not present this or any other epitope of FV. It remains to be determined whether FV-specific CTL generated in mice of other H-2 haplotypes might recognize the peptide A epitope as presented by different murine class I restriction elements.

ACKNOWLEDGMENTS We thank Drs. 1. A. Berzofsky and J. B. Rothbard for their help in the identification of candidate epitopes, Dr. R. Angelotti for help in the synthesis of the identified peptides, and David Polsky for assistance in the immunological assays. This work was supported by Grant CA1 9931 from the National Institutes of Health. K.-S.R. is supported by a fellowship from the Cancer Research Institute, Inc., and F.L. is an American Cancer Society Research Professor.

REFERENCES ALLEN, H., WRAITH, D., PALA, P.. ASKONAS, B., and FLAVELL, R. A. (1984). Domain interactions of H-Z class I antigens alter cytotoxic T-cell recognition sites. Nature (London) 309, 279-281. AUSUBEL, K. M., BRENT. R., KINGSTON, R. E., MOORE, D. D., SMITH, 1. A., SEIDMAN, 1. G., and STRUHL, K., Eds. (1987). Guanldinium method for total RNA preparation. In “Current Protocols in Molecular Biology,” pp. 4.2.1-4.2.5. Wiley, New York. BENNINK,J. D., YEWDELL,J. W., SMITH, G. L.. MOLLER, C., and Moss, B. (1984). Recombinant vaccinia virus primes and stimulates influenza haemagglutinin-specific cytotoxic T cells. Nature (London) 311, 578-579. BJORKMAN, P. J., SAPER, M. A., SAMRAOUI, B., BENNEIT, W. S.. STROMINGER,J. L., and WILEY, D. C. (1987). The foreign antigen binding site and T cell recognition regions of class I histocompatibility antigens. Nature (London) 329,5 12-518. BRACIALE,T. J., BRACIALE,V. L., HENKEL,T. J., SAMBROOK,J., and GETHING, M.-J. (1984). Cytotoxic T lymphocyte recognition of the influenza haemagglutinin gene product expressed by DNA-mediated gene transfer. /. Exp. Med. 159, 341-354. BUBBERS,J. E., and HENNEY, C. S. (1975). Studies on the synthetic capacity and antigenic expression of glutaraldehyde-fixed target cells. J. Immunol. 114, 1126-l 131. CHEN, C. A., and OKAYAMA, H. (1988). Calcium phosphate-mediated gene transfer: A highly efficient transfection system for stably transforming cells with plasmid DNA. Biofechniques 6, 632-638.

of the SFFV (Clark from the envgene

CHESEBRO,B., WEHRLY, K., CLOYD, M., BRIIT, W., PORTIS,J., COLLINS, J., and NISHIO, J. (1981). Characterization of mouse monoclonal antibodies specific for Friend leukemia virus-induced erythroleukemia cells: Friend-specific and FMR-specific antigens. virology 112,131-144. CHIMINI, G., PALA, P., SIRE, J., JORDAN, B. R., and MARYANSKI. J. L. (1989). Recognition of oligonucleotide-encoded T cell epitopes introduced into a gene unrelated to the original antigen. J. Exp. Med. 169,297-302. CLARK, S. P., and MAK, T. W. (1983). Complete nucleotide sequence of an infectious clone of Friend spleen focus-forming provirus: Gp55 is an envelope fusion glycoprotein. Proc. Nat/. Acad. SC;. USA 80, 5037-5041. DELISI, C., and BERZOFSKY,J. A. (1985). T-cell antigenic sites tend to be amphipathic structures. Proc. Nat/. Acad. Sci. USA 82, 70487052. GOTCH, F. M., MCMICHAEL, A. J., SMITH, G. L., and Moss, B. (1987a). Identification of viral molecules recognized by influenza-specific human cytotoxic T lymphocytes. J. Exp. Med. 165, 408-416. GOTCH, F., ROTHBARD, J., HOWLAND, K., TOWNSEND, A., and McMICHAEL, A. (1987b). Cytotoxic T lymphocytes recognize a fragment of influenza virus matrix protein in association with HLA-A2. Nature (London) 326, 881-882. HOLT, C. A., OSORIO, K.. and LILLY, F. (1986). Friend virus-specific cytotoxic T lymphocytes recognize both gag and env gene-encoded specificities. J. Exp. Med. 164, 21 l-226. KAWASAKI,E. S. (1990). Amplification of RNA. In “PCR Protocols: A Guide to Methods and Applications” (M. A. Innis, D. H. Gelfand, J. J. Sninsky, and T. J. White, Eds.), pp. 21-27. Academic Press, San Diego. KOCH, W., ZIMMERMANN, W., OLIFF, A., and FRIEDRICH,R. (1984). Molecular analysis of the envelope gene and long terminal repeat of Friend mink cell focusinducing virus: Implication for the function of these sequences. J. Viral. 49, 828-840. MARGALIT. H., SPROUGE,J. L.. CORNE~E, 1. L., CEASE, K. B., DELISI, C., and BEWOFSKY,1. A. (1987). Prediction of immunodominant helper T cell antigenic sites from the primary sequence. 1. Immunol. 138, 2213-2229. MULLIGAN, R. C., and BERG, P. (1980). Expression of a bacterial gene in mammalian cells. Science 209, 1422-l 427. OLIFF, A., HAGER, G. L., CHANG, E. H., SCOLNICK, E. M., CHAN, H. W., and Lowv, D. R. (1980). Transfection of molecularly cloned Friend leukemia virus DNA yields a highly leukemogenic helper-independent type C virus. J. Viral. 33, 475-486. OZATO. K., and SACHS, D. H. (1981). Monoclonal antibodies to mouse MHC antigens. III. Hybridoma antibodies reacting to anti-

100

RUAN AND LILLY

gen of the H-2b haplotype reveal genetic control of isotype expresslon. J. Immunol. 126, 317-32 1. PLATA, F., and LILLY, F. (1979). Viral specificity of H-2 restricted T killer cells directed against syngeneic tumors induced by Gross, Friend, or Rauscher leukemia virus. J. hp. Med. 150, 1174-l 186. ROTHBARD,1. B., and TAYLOR, W. R. (1988). A sequence pattern common to T cell epitopes. EMBO J. 7, 93-100. SANGER, F., NICKLEN, S., and COULSON. A. R. (1977). DNA sequencing with chain-terminating inhibitors. Proc. Nat/. Acad. Sci. USA 74, 5463-5467. SCOLNICK. E. M., WILLIAMS, D., MARYAK, J., VASS, W., GOLDBERG,R. J., and PARKS,W. P. (1976). Type C particle-positive and type C partcle-negative rat cell lines: Characterization of the coding capacity of endogenous sarcoma vlrus-speclfic RNA. J. Viral. 20, 570-582. SHINNICK. T. M., LERNER, R. A., and SUTCLIFFE.J. G. (1981). Nucleotide sequence of Moloney murlne leukemia virus. Nature (London) 293, 543-548. SOUTHERN, P. J., and BERG, P. (1982). Transformation of mammaljan cells to antibiotic resistance with a bacterial gene under control of the SV40 early region promoter. J. Mol. Appl. Gen. 1, 327-341. TOWNSEND, A. R. M., BASTIN, J., GOULD, K.. and BROWNLEE, G. G.

(1986a). Cytotoxic T lymphocyte recognize influenza haemagglutinin that lacks a signal sequence. Nature (London) 324, 575-577. TOWNSEND, A. R. M., GOTCH, F. M., and DAVEY, J. (1985). Cytotoxic T cells recognize fragments of the Influenza nucleoprotein. Cell 42, 457-467. TOWNSEND, A. R. M., MCMICHAEL, A. J., CARTER, M. P., HUDDLESTON, 1. A., and BROWNLEE,G. G. (1984). Cytotoxic T cell recognition of the Influenza nucleoprotein and hemagglutinln expressed in trans fected mouse L cells. Ce// 39, 13-25. TOWNSEND, A. R. M., ROTHBARD, J., GOTCH, F M., BARHADUR, G., WRAITH, D.. and MCMICHAEL, A. (1986b). The epitopes of influenza nucleoprotein recognized by cytotoxic T lymphocytes can be defined with short synthetic peptides. Cell 44, 959-968. YEWDELL, J. W., BENNINK, J. R., SMITH, G. L., and Moss, 6. (1985). Influenza A virus nucleoprotein IS a major target antigen for cross reactive anti-influenza A virus cytotoxlc T lymphocites. Proc. Nat/. Acad. Sci. USA 82, 1785-l 789. ZINKERNAGEL,R. M.. and DOHERTY,P. C. (1979). MHC-restricted cytotoxic T cells: Studies on the bIologIcal role of polymorphic major transplantation antigens determlnlng T-cell restriction-specificity, function and responsiveness. Adv. Immunoi. 27, 51-l 77.