SV40 large tumor antigen associated synthetic peptides define native antigenic determinants and induce protective tumor immunity in mice

Moleculur Pergamon

Immunolog.y,

Vol. 31, No. 14, pp. 1077-1087, 1994

Copyright $> 1994 Elsevier Science Ltd Printed in Great Britain. All rights reserved

0161-5890(94)00083-2

0161-5890194 $7.00 + 0.00

SV40 LARGE TUMOR ANTIGEN ASSOCIATED SYNTHETIC PEPTIDES DEFINE NATIVE ANTIGENIC DETERMINANTS AND INDUCE PROTECTIVE TUMOR IMMUNITY IN MICE R. K. BRIGHT,?

M. H. SHEARER*

and R. C. KENNEDY*t$

*Department of Virology and Immunology, Southwest Foundation for Biomedical Research, San Antonio, TX 78228, U.S.A.; and tDepartment of Microbiology, The University of Texas Health Science Center, San Antonio, TX 78284, U.S.A. (First received 29 November

1993; accepted in revised form

3 May

1994)

Abstract-Synthetic peptides were utilized to define antigenic determinants on simian virus 40 (SV40) large tumor antigen (T-ag). Six synthetic peptides representing predicted B-cell epitopes on SV40 T-ag were used to immunize mice to compare the humoral immune responses and ascertain the ability of the peptide preparations to induce protective tumor immunity in vivo. Anti-peptide antibodies from BALB/c and C57BL/6 mice were examined for reactivity with SV40 T-ag by various immunologic assays. Antibodies from both strains to four of the peptides recognized recombinant SV40 T-ag by ELISA. However, T-ag recognition by anti-peptide antibodies differed when assessed by Western blot. Antibodies induced by the same four peptides in BALB/c mice recognized T-ag, whereas only three of the six peptides induced antibodies in C57BL/6 mice capable of recognizing SV40 T-ag by Western blot. Flow cytometric analysis revealed that antibodies to peptides corresponding to T-ag amino acid residues 632-652 and 690-708 from BALB/c mice were able to recognize the surface of SV40 transformed cells, whereas five of the six peptides induced surface reactive antibodies in C57BL/6 mice. More important, peptides 632-652 and 690-708 elicited a protective immune response in BALB/c mice subsequently challenged with a lethal dose of syngeneic SV40 transformed cells. However, this tumor immunity was incomplete as only 50% of the mice survived the tumor challenge. These data indicate that antibodies induced by synthetic peptides corresponding to predicted B-cell epitopes on SV40 T-ag are capable of recognizing native and denatured determinants on T-ag. Furthermore, immune responses elicited by selected peptides partially protected BALB/c mice from a lethal tumor challenge. Key words: SV40 large tumor

antigen,

synthetic

INTRODUCTION

Synthetic peptides represent tools to analyse protein structure and have been utilized for studying the epitope specificity of cellular and humoral immune responses induced by protein antigens. This technology has also been employed to assist in the development of active immunotherapies, including vaccines to prevent and treat a variety of infectious viral, bacterial, and protozoal diseases (reviewed in Rothbard, 1992). Synthetically generated peptides designed to emulate immunogenic epitopes on native protein antigens provide beneficial tools for analysing the immune responses to such patho-

$Author to whom correspondence should be addressed. Abbreviations: BBS, borate buffered saline; DMEM, Dulbecco’s minimal essential medium; EDTA, ethylenediaminetetraacetic acid; GAM, goat anti-mouse IgG Fc specific reagent; HBsAg, hepatitis B surface antigen; HRP, horseradish peroxidase; KLH, keyhole limpet hemocyanin; NGS, normal goat serum; PBS-T, phosphate buffered saline containing Tween-20; SV40, simian virus 40; T-ag, large tumor antigen.

peptides,

tumor

vaccines.

gens and have also been utilized to examine specific immune responses to tumor-associated and tumorspecific antigens. This information can be used to develop peptide based strategies for the development of possible therapies for the treatment of such diseases. SV40 is a member of the papovavirus family, which includes the human viruses, BKV and JCV. This group of viruses is capable of inducing tumors in rodents and transforming cells of both human and rodent origin in vitro (Butel et al., 1972). In addition, studies have presented evidence demonstrating the presence of BKV and JCV associated with numerous human tumors (Dorries et al., 1987; Fiori and de Mayorca, 1976; Krieg et al., 1981). A common protein antigen among these viruses is the large tumor antigen (T-ag). The T-ag share functional identity, as well as overall amino acid sequence homology (72%) (Frisque et al., 1984), and are immunologically cross-reactive. Monoclonal antibodies generated to SV40 T-ag are capable of recognizing T-ag from BKV and JCV (Gurney et al., 1980; Harlow et al., 1981), and a recent study demonstrated that SV40 T-ag specific cytotoxic T-cell clones recognize common epitopes on JCV T-ag (Deckhut et al., 1991). Rodent cells 1077

1078

R. K. BRIGHT

transformed by super infection with SV40 in ritro express SV40 T-ag in the cell nucleus, the cytoplasm, and to a small extent (5%) intact at the cell surface (Butel. 1986; Butel and Jarvis, 1986: Soule et al., 1980). Mice immunized with recombinant SV40 T-ag or inoculated with inactivated syngeneic SV40 transformed cells elicit SV40 T-ag specific immune responses and are protected from tumor challenge with live SV40 transformed cells (Tevethia, 1980; Shearer rf (11.. 1993). Therefore, SV40 T-ag represents a virally encoded tumor-specific antigen that is involved in the induction of protective tumor immunity. Previous studies examined SV40 T-ag induced immune responses in multiple inbred strains of mice and concluded that the observed protective tumor immunity was likely to be mediated by MHC restricted CTL (reviewed in Tevethia, 1990). Mice possessing the H-2’ MHC haplotype (C57BL/6) were reported as demonstrating the greatest SV40 T-ag specific CTL activity, whereas BALB/c mice (H-2“) produced little to no detectable CTL activity. Additionally, the only SV40 transformed tumorigenic cell line available for in riro tumor challenge experiments is of BALB/c origin. Thus, the utilization of CTL mechanisms for SV40 T-ag specific protective tumor immunity in z+o is an inference from these in vitro experiments. This reported difference in SV40 T-ag CTL activity between BALB/c and C57BL/6 mouse strains prompted us to examine and compare the humoral immune responses to SV40 T-ag induced in these two inbred strains. We generated synthetic peptides that were designed to emulate six distinct B-cell epitopes on the native structure of SV40 T-ag using computer algorithms and examined antibodies induced by immunization with these peptides for reactivity with native SV40 T-ag. In addition, we assessed the ability of the various SV40 T-ag peptide preparations to induce protective tumor immunity in riro. In the present study, we demonstrate that immunization of both BALB/c and C57BL/6 mice with synthetic peptides corresponding to six putative epitopes on SV40 T-ag induced antibodies capable of reacting with SV40 T-ag to various degrees by ELISA, Western blot, and by flow cytometric analysis of SV40 transformed cell surfaces. Moreover, immunization with two of the six SV40 T-ag synthetic peptides, which define distinct epitopes on SV40 T-ag, induced partially protective tumor immunity in BALB/c mice. Together. these data indicate that synthetic peptides corresponding to predicted B-cell epitopes on SV40 T-ag are capable of eliciting antibodies in both inbred strains of mice that recognize the denatured tumor antigen by Western blot analysis, as well as native tumor antigen expressed on the surface of SV40 transformed cells by flow cytometry. In addition, two of the synthetic peptide preparations were capable of inducing partial protective tumor immunity in BALBjc mice. These synthetic peptide preparations may be useful in determining which SV40 T-ag epitopes are important in antibody based tumor immunity, and in dissecting the mechanism(s) of protective immunity induced in BALB/c mice by SV40 T-ag.

et cd

MATERIALS

AND METHODS

Mice Female BALB/c and C57BL/6 mice, 68 weeks of age (Jackson Labs, Bar Harbor, ME), were used for immunization with either baculovirus derived recombinant SV40 T-ag or keyhole limpet hemocyanin (KLH) coupled synthetic peptides. In addition, female BALB/c mice were used for in t+o tumor protection experiments. Cells and medium The tumorigenic SV40 transformed BALB/c kidney fibroblast cell line designated mKSA has been described in detail elsewhere (Kit et al., 1969). The mKSA cell line was cultured in Dulbecco’s minimal essential medium (DMEM) supplemented with glutamine (2 mM), sodium pyruvate (1 mM), nonessential amino acids (0.1 mM), HEPES buffer (10 mM), penicillin (500 U/ml), streptomycin (500 mg/ml), and 10% (v/v) heat inactivated FCS. Cells were passaged by incubation of the monolayer with a buffered solution containing 0.5 mg/ml trypsin and 0.2 mg/ml ethylenediaminetetraacetic acid (EDTA) (Sigma Chemical Co., St Louis, MO). Antigens Recombinant SV40 T-ag was generated in the Sf9 insect cell line using a baculovirus (Autographu calijornica nuclear polyhedrosis virus) expression vector system (Lanford, 1988) and immunoaffinity purified by methods previously described (Simanis and Lane, 1985). Recombinant SV40 T-ag purity was assessed by SDS-PAGE (Shearer et al., 1990), and the concentration was determined using an extinction coefficient of 1.14 for a 0. I % solution at 280 nm. Plasma purified hepatitis B surface antigen (HBsAg) served as a control antigen. Computer

algorithms

to predict

SV40

T-ag epitopes

To identify immunogenic regions of SV40 T-ag for the synthesis of corresponding peptides, the complete amino acid sequence of SV40 T-ag was analysed by computer algorithm for regions of hydrophilicity and beta turn potential, as previously described (Pauletti et a/., 1985). Each amino acid in the sequence of the protein was assigned its individual hydrophilicity value. Hydrophilicity values were repeatedly averaged for groups of seven amino acids along the length of the entire SV40 T-ag polypeptide chain and were subsequently plotted for heptapeptide regions from highest to lowest for the entire SV40 T-ag amino acid sequence. Those regions demonstrating the highest hydrophilicity values were selected for determination of beta turn potential. Regions of SV40 T-ag with beta turn potential were selected from the computer generated predicted secondary structure of SV40 T-ag. Synthetic

peptides

Synthetic peptides were derived from the amino acid sequence of SV40 T-ag based on the known nucleic acid sequence. Selection of the predicted B-cell epitopes on SV40 T-ag for the design of the synthetic peptides was

Synthetic peptide induced tumor immunity based on the following criteria: (i) hydrophilicity; (ii) overall amino acid content to facilitate synthesis; and (iii) the potential for beta turns to impart secondary structure. Hydrophilicity and beta turn potential have been reported previously as useful methods for predicting putative B-cell epitopes within a protein antigen (Pauletti et al., 1985). All peptides were synthesized with standard solid phase technology and F-mot chemistry and subsequently purified by reverse phase HPLC utilizing methods previously described (Attanasio et al., 1990). Amino acid sequence and content of each peptide was confirmed by amino acid analysis and mass spectroscopy. Amino acid sequences of the six SV40 T-ag peptides utilized in this study are described in Table 1. Prior to immunization, all peptides were coupled to KLH through the carboxyl-terminal cysteine residues using the heterobifunctional cross-linker m-maleimidobenzoyl-N-hydroxysuccinimide. The coupling ratio of peptide to carrier protein was determined to be approximately 1: 1, as assessed by amino acid analysis of the conjugates. The control peptide corresponds to amino acid residues 304 to 321 of the V3 loop of gpl20 from the human immunodeficiency virus type 1 (HIV-l), IIIB isolate. Immunizations Individual groups of five mice (BALB/c and C57BL/6) were each immunized i.p. with either an alum precipitate of immunoaffinity purified recombinant SV40 T-ag, or a Freund’s adjuvant emulsion of KLH-coupled synthetic peptides. Sera collected prior to immunization served as preimmune controls. Mice were primed with 2Opg of recombinant SV40 T-ag as an alum precipitate, or 50 pg of KLH-peptide conjugate in complete Freund’s adjuvant. All subsequent injections were given at 1Cday intervals. Three additional injections of 10 pg each were given to the SV40 T-ag groups, and five additional injections of 25 pg in incomplete Freund’s adjuvant to each of the KLH-peptide conjugate groups. Serum was collected and pooled following the second, third, and fourth injections for SV40 T-ag immunized mice, and the fourth, fifth, and sixth injections for peptide immunized mice. The serum was pooled from mice with similar antibody titers prior to performing the various assays to ensure the availability of sufficient quantities of immune serum for future experiments. In addition, a group of

1079

five mice was immunized with the KLH coupled HIV-l gp 120 peptide (designated 5077) to serve as a control for the preliminary in vivo tumor protection studies. ELISA

for the detection

qf antibodies to SV40 T-ag

All binding curves were generated by an ELISA. To initially examine SV40 T-ag reactivity of the immune sera, 50 ~1 of purified recombinant SV40 T-ag or plasma purified HBsAg, at concentrations of 4,ugg/ml in BBS (50 mM borate, pH 8.2) were used to coat 96-well flat bottom microtiter plates overnight at 4-C. Nonspecific binding was blocked by the addition of 200~1 of 10% (w/v) normal goat serum (NGS) in BBS for 1 hr at 37°C. Fifty microliters of a 1:50 dilution (in NGS/BBS) of pooled sera from mice immunized with the various synthetic peptides was added to individual wells in duplicate and titrated by serial four-fold dilutions, then incubated at 37°C for 1 hr. Finally, 50 ~1 of a horseradish peroxidase (HRP) conjugated goat anti-mouse IgG Fc specific reagent (GAM) (Sigma), diluted in 10% NGS/BBS, was added to each well and incubated for 1 hr at 37°C. Each plate was washed three to five times with PBS containing 0.02% (w/v) Tween-20 (PBS-T) between each incubation step. Assays were developed with 2,2’-azino-di(3-ethylbenzthiazoline sulfonic acid) containing 0.01% (v/v) hydrogen peroxide. Enzyme substrate reactions were terminated by the addition of 100,ul of 5% (w/v) SDS. Optical density at 410 nm (OD,,,) was determined using an automatic ELISA plate reader (Dynatech Labs Inc., Alexandria, VA). Endpoint titers were established as three times the OD,,, value obtained for a 1: 50 dilution of pooled preimmune sera. ELISA for the detection of antibodies to synthetic peptides To examine specificity of the anti-peptide sera, ELISA assays were performed as described above. Briefly, 50 ~1 of the unconjugated SV40 T-ag peptide at a concentration of 5 pg/ml (in BBS) was coated onto individual 96-well flat bottom microtiter plates and incubated overnight at 4°C. Nonspecific binding was blocked with 10% (w/v) NGS/BBS. Fifty microliters of a 1: 50 dilution of the pooled anti-peptide sera was added to individual wells in duplicate, and four-fold serial dilutions were performed to determine endpoint titers, as described above. Endpoint titers were calculated as three times the OD,,O value obtained for a I:50 dilution of pooled preimmune sera.

Table 1. SV40 T-ag synthetic peptide amino acid sequences

Western blot analysis

.4mino acid number

To examine the ability of the anti-SV40 T-ag and anti-peptide antibodies to bind native denatured SV40 T-ag from SV40 transformed cells, Western blot assays were performed using mKSA whole cell lysates. The mKSA cells were grown to confluency in media (DMEM supplemented as described in Materials and Methods) with incubation at 37C in 5% CO, in air. Approximately 5 x IO’ cells were harvested with trypsin/EDTA, washed three times with 2 ml of PBS, and lysed with 1 ml of 50 mM Tris-HCl buffer containing 100 mM NaCl, 100 p M aprotinin, 100 PM leupeptin, and 1% (v/v)

3656 121-141

472-490 632-652 660-679 690-708

Amino acid sequence” CGKKSKEFHPDKGGDEEKMKKMN CGQHSTPPKKKRKVEDPKDFPSE CFEDVKGTGGESRDLPSGQG CGSDDDDEDSQENADKNEDGGEK CGETGIDSQSQGSFQAPQSSNS CRGFTSFKKPPTPPPEPET _

“Underlined amino acids were added to facilitate coupling to carrier proteins (CG), or to replace cysteine residues to prevent intrachain disulfide bond formation (S).

1080

R. K. BRIGHT efal.

NP-40 (pH 9.0). After agitation at 5 min intervals for 30 min, the whole cell lysate was clarified of insoluble material by centrifugation at 4000 rpm for 20 min. The supernatant was collected, transferred to a fresh microcentrifuge tube, and further clarified by centrifugation at 13,000 rpm for 20 min. A 100 ~1 aliquot of the final supernatant was electrophoresed on a discontinuous (4% to 12%) SDS-PAGE gel, followed by electrolytic transfer to nitrocellulose for immunoblotting. Nitrocellulose sheets were blocked from nonspecific binding by incubation in BLOTTO (25.0 g/l nonfat dry milk, 25.0ml/l liquid gelatin, and 0.02% Tween-20 (w/v) in ddIH,O) overnight at 4°C. Pooled sera were diluted 1: 100 in PBS-T containing 1% (v/v) NGS and incubated with individual 0.5 cm wide nitrocellulose strips on a rocker overnight at 4°C. Antibody binding was detected by the addition of GAM-HRP (Sigma) diluted in PBST/NGS and incubation with rocking for approximately 2 hr at room temperature. Individual strips were developed with the peroxidase sensitive substrate, 3,3’-diaminobenzidine (DAB) (Sigma) containing 0.03% (v/v) H,O,. Enzyme substrate reactions were terminated by rinsing the strips with distilled deionized water. Each strip was washed three times for 5 min with PBS-T between each incubation step. High molecular weight markers were run simultaneously with the lysates and stained for total protein (Amid0 Black) after transfer to nitrocellulose to establish approximate molecular weights of the proteins from the lysate. Flow cytometry Prior to conducting in oiuo tumor protection assays, it was of interest to determine the ability of the immune sera to recognize the surface of SV40 transformed (mKSA) cells. The mKSA cell line was cultured as described above. Cells were harvested by incubating the monolayer in PBS containing 1 mM EDTA in the absence of trypsin to preserve surface expressed proteins. The cell pellet was washed twice in cold PBS then resuspended in PBS containing 5% (v/v) FCS and 0.01% (w/v) sodium azide (PBS-BA). Cells were aliquoted (4 x lO’j20 ~1) into individual polystyrene culture tubes and incubated with a 1: 100 dilution (in PBS-BA) of pooled sera for 1 hr at 37’C. Each tube was washed three times with 1 ml of PBS-BA. Following the final wash, cell pellets were resuspended in 20 ~1 of PBS-BA containing a 1: 40 dilution (based on titration experiments) of a goat anti-mouse IgG whole molecule specific reagent, conjugated with FITC (Sigma), and incubated for 30min at 4°C in the dark. Finally, the individual cell pellets were washed with 2 ml of PBS-BA, then fixed with 500 ~1 of Hematall isotonic diluent containing 2% (v/v) of 37% USP methanol free formaldehyde. Fluorescence intensity was determined within 24 hr using a Becton-Dickinson FACScan and the Consort 30 software application (Fig. 4). A sample containing cells only was used to set the negative marker population. The data are presented in a bar graph format as relative fluorescence intensity, based on fluorescence intensity values obtained from individual histograms (Figs 3A, B).

In vivo tumor protection

assay

To determine if the six SV40 T-ag associated synthetic peptides were capable of eliciting protective immune responses, an in vice tumor challenge experiment was conducted. Groups of BALB/c mice were immunized four times with recombinant SV40 T-ag or six times with SV40 T-ag associated synthetic peptides (as described above). After resting the mice for approximately 30 days, a final booster injection was given i.p. Approximately 14 days following the booster injection, mice were challenged with an i.p. inoculation of five LD,, (2.5 x 10’) (Mernaugh et al., 1992) of mKSA cells in PBS (pH 7.2). Mice immunized with the HIV-l gp120 peptide or alum alone served as controls. Protection data are reported as the mean survival time in days and percentage survival greater than 60 days post tumor challenge. Stutistical

analysis

The tumor challenge data were analysed with a two tailed Student’s r-test to determine whether significant differences existed between the survival times of SV40 T-ag and SV40 T-ag synthetic peptide immunized mice compared to control peptide immunized mice. The Student’s t-test employed the mean survival time (days) for independent samples of unequal variances (Rosner. 1990).

RESULTS Anti-peptide responses peptides defining SV40

in mice imunized T-ag epitopes

with synthetic

Mice from two inbred strains were individually immunized with each of the six synthetic peptides described in Table 1. Pooled anti-peptide sera were screened for specificity by examining cross-reactivity with each of the other SV40 T-ag peptides, as well as with a control HIV-l gp120 peptide. BALB/c antibodies generated by immunization with five of the SV40 T-ag synthetic peptides corresponding to amino acid residues 3656, 121-141, 632-652, 660-679 and 690-708 (reciprocal endpoint titers ranged from 50@41,600) demonstrated reactivity with their eliciting peptides and failed to significantly recognize any of the other five SV40 T-ag synthetic peptides by ELISA (Table 2). BALB/c antibodies generated to the sixth SV40 T-ag synthetic peptide analogous to amino acid residues 472-490 bound its respective peptide with a reciprocal endpoint titer of 41,600 and failed to recognize four of the five distinct SV40 T-ag synthetic peptides by ELISA. However. antibodies to peptide 472490 demonstrated reactivity with SV40 T-ag peptide 690-708 (reciprocal endpoint titer of 3200). This antibody cross-reactivity may reflect chemical similarities between peptides 472-490 and 690-708 that may impart a possible epitope identity between the stretch of amino acids at position 5 to 9 (KGTGG) in peptide 472490. and the amino acid sequence at position 9 to 14 (KPPTPP) in peptide 690-708. The reason for the apparent inability of BALB/c antibodies with specificity for peptide 69G-708

Synthetic peptide induced tumor immunity

1081

Table 2. Anti-peptide responses in mice immunized with synthetic peptides defining SV40 T-ag epitopes Inbred strain

Immunization

BALB/c

None 36 121 472 632 660 690 None

C57BL/6

36 121 472 632 660 690

36 _h 500” _ _ 800 _

SV40 T-ag epitopes defined by synthetic peptides 121 472 632 660 690 5077” 22,400 _ _ _ 800 _

41,600 _ _ _ _ 2600 _

800 _ _ _ _ 3200 _

22,400 _ _ _ _ 3200 _

_ _ 3200 11,600 _ _ _ 10,400

_ _ _ _ _ -

OControl HIV-l gp120 synthetic peptide. ‘Denotes undetectable antibody levels.
to recognize peptide 472-490 is unknown.

In C57BL/6 mice, antibodies generated by immunization with the six individual SV40 T-ag synthetic peptides each demonstrated specificity for their respective peptide by failing to recognize any of the other five SV40 T-ag synthetic peptides (Table 2). Reciprocal endpoint titers for C57BL/6 anti-peptide responses following immunization with the six SV40 T-ag peptides ranged from 800 to 10,400. In addition, anti-peptide sera from both inbred mouse strains failed to recognize a synthetic peptide corresponding to gp120 of HIV-l. Sera collected from both strains prior to immunization also failed to bind any of the synthetic peptides examined. These data demonstrate the peptide reactivity and specificity of the antibody responses generated by immunization of

Reciprcxxl

Dilution

BALB/c and C57BL/6 mice with six distinct SV40 T-ag synthetic peptides. SV40 T-ag reactivity of anti-peptide responses Anti-peptide sera from both inbred strains were examined for the ability to recognize recombinant SV40 T-ag by ELISA. BALB/c antibodies with specificity for SV40 T-ag synthetic peptides analogous to residues 472-490, 632-652, 660679 and 690-708 recognized SV40 T-ag with reciprocal endpoint titers of 3200, 51,200, 12,800 and 12,800, respectively (Fig. IA). Similarly, antipeptide responses induced in C57BL/6 mice by peptides 472490, 632-652, 660-679 and 690-708 demonstrated binding to SV40 T-ag, with reciprocal endpoint titers of 800, 3200, 800 and 51,200, respectively (Fig. 1B). The

Of %Nm

Fig. 1. SV40 T-ag binding curves from mice immunized with the six SV40 T-ag synthetic peptides. (A) BALB/c anti-peptide sera binding SV40 T-ag by ELISA: Preimmune (+), 36-56 (x), 121-141 (*), 472-490 (A), 632-652 (@), 660-679 (O), and 690-708 (m). (B) C57BL/6 anti-peptide sera binding SV40 T-ag by ELISA: Preimmune (+), 36-56 (x), 121-141 (*), 472490 (A), 632-652 (O), 660-679 (O), and 690-708 (m). Each point represents the mean of duplicate determinations of pooled sera from five individually immunized mice. SEM values are shown as brackets. Anti-peptide antibodies from both strains failed to recognize the control antigen, HBsAg (data not shown).

R. K. BRIGHT et al.

1082

anti-peptide responses that bound SV40 T-ag failed to recognize a control antigen, HBsAg by ELISA, and in each instance, preincubation of the anti-peptide sera with SV40 T-ag inhibited binding to SV40 T-ag on the solid phase (data not shown). Antibodies generated against peptides 36-56 and 121-141 in both BALB/c and C57BL/6 mice failed to bind SV40 T-ag. These data indicate that murine antibodies generated to four out of six synthetic peptides defining predicted B-cell epitopes on SV40 T-ag are capable of recognizing epitopes expressed on intact recombinant SV40 T-ag by ELISA. Western peptide

blot analysis of SV40

T-ag reactioity

by anti-

responses

Since some of the antibodies produced by immunization with SV40 T-ag synthetic peptides were capable of binding recombinant SV40 T-ag by ELISA, it was of interest to ascertain if these murine anti-peptide responses were capable of recognizing denatured SV40 T-ag from SV40 transformed cells. Such reactivity would confirm the ELISA data and further indicate the recognition of sequential or linear epitopes on SV40 T-ag. Pooled immune sera from both inbred mouse strains were utilized at 1: 100 dilution to examine SV40 T-ag reactivity by Western blot. BALB/c antibodies from mice immunized with either recombinant SV40 T-ag or SV40 T-ag synthetic peptides 472490,632-652,660-679 and 69G-708 demonstrated reactivity, by Western blot, with a protein of an approximate molecular weight of 97 kD (Fig. 2A, lanes 2, 5, 6, 7 and 8, respectively). Similarly, antibodies from C57BL/6 mice induced by either recombinant SV40 T-ag or SV40 T-ag synthetic peptides 472490, 632652 and 690-708 recognized a 97 kD protein (Fig. 2B, lanes 2, 5, 6 and 8, respectively). Antibodies from both mouse strains, generated to synthetic peptides 36-56 and 121-141, as well as preimmune sera, failed to recognize any of the proteins from the SV40 transformed cell lysate by Western blot (Fig. 2, lanes 3, 4 and 1, respectively). Although C57BL/6 antibodies to the synthetic peptide corresponding to the amino acid sequence 660-679 bound SV40 T-ag by ELISA, they failed to recognize SV40 T-ag by Western blot (Fig. 2B, lane 7). This lack of reactivity may be the result of a conformational requirement for the recognition of this epitope that was subsequently destroyed by denaturation. The doublet observed at 97 kD is most likely due to variation in post-translational modification of the protein (glycosylation, acylation and phosphorylation) (Lanford, 1988). The additional lower molecular weight proteins that recognized antibodies to SV40 T-ag peptides 632-652 and 690-708 from both mouse strains, as well as 660-679 anti-peptide antibodies from BALB/c mice, are most likely the result of cross-reactivity with distinct cellular proteins that also have been shown to interact with SV40 T-ag (reviewed by Levine, 1990). The observed lack of reactivity with SV40 T-ag by antibodies generated to peptides 36-56 and 121-141 from both inbred strains does not appear to be titer related since antibodies to peptide 121-141 in BALB/c mice demonstrated a reciprocal peptide endpoint titer of 22,400, and

this compares to a peptide titer of 8000 for antibodies to peptide 632-652 in the same strain that recognize SV40 T-ag by Western blot. Reactivity of antibodies induced by recombinant SV40 T-ag from both inbred mouse strains with the 97 kD protein (Figs 2A, B, lane 2) further indicated that antibodies generated to several of the synthetic peptides defining epitopes on SV40 T-ag were capable of recognizing linear epitopes on denatured native SV40 T-ag by Western blot. Antibody surface

reactiuity qf SV40

with SV40

trawformed

T-ag

expressed

on the

cells

In the next set of experiments, we used flow cytometry to examine the ability of antibodies induced by immunization with the six SV40 T-ag synthetic peptides to recognize the surface of live SV40 transformed (mKSA) cells. Recombinant SV40 T-ag induced antibodies from BALB/c mice were capable of recognizing mKSA cells with fluorescence intensity values of approximately 200, nearly ten-fold greater than that of the preimmune sera (Fig. 3A). C57BL/6 antibodies induced by recombinant SV40 T-ag demonstrated a fluorescence intensity of 320 (Fig. 3B). The observed fluorescence intensity value of 50 for antibodies with specificity for the control HIV-l gp120 synthetic peptide is most likely the result of nonspecific reactivity with cell surface expressed glycoproteins by antibodies elicited to the carrier protein, KLH. Since all synthetic peptides used for immunization in this study were coupled to KLH to enhance immunogenicity, the SV40 T-ag synthetic peptide responses were compared to the control peptide response to establish specificity as assessed by fluorescence intensity. Antibodies from BALB/c mice with specificities for the synthetic peptides analogous to SV40 T-ag amino acid residues 36-56, 121-141, 4722490 and 66&679 failed to demonstrate reactivity with mKSA cell surfaces, in that the fluorescence intensities were similar to that induced by the control peptide. However, BALB/c antibodies to SV40 T-ag synthetic peptides 632-652 and 690-708 demonstrated fluorescence intensity values greater than that of the control peptide (fluorescence intensities of 90 and 80, respectively). In addition, SV40 T-ag at a concentration of 50 pg/ml was capable of inhibiting the mKSA cell surface reactivity of antibodies induced by both 632-652 and 690-708 (data not shown), such that the respective histograms resembled the histograms of the negative controls by flow cytometry, thus indicating specificity for the surface of the SV40 transformed cells (Fig. 3A). The greater fluorescence intensity values observed for antibodies with specificity for SV40 T-ag compared to antibodies with specificity for the synthetic peptides was not surprising since the SV40 T-ag induced antibodies are likely to have multiple SV40 T-ag epitope specificities, With the exception of antibodies to the SV40 T-ag peptide corresponding to amino acid residues 36-56 that failed to demonstrate cell surface recognition, all the anti-peptide responses for C57BL/6 mice appeared to recognize the surface of mKSA cells with fluorescence intensity values greater than that of the control peptide. Fluorescence intensity values for

Synthetic peptide induced tumor immunity

12345678 200 116-

976745-

312-I14-

12345678

Fig. 2. Western blot analysis of SV40 transformed mouse kidney fibroblast (mKSA) whole cell lysates. (A) Sera from BALB/c mice. (B) Sera from C57BL/6 mice. Immunization: none (lane l), SV40 T-ag (lane 2), peptide 36-56 (lane 3) peptide 121-141 (lane 4), peptide 472490 (lane 5) peptide 632-652 (lane 6), peptide 660-679 (lane 7), and peptide 690-708 (lane 8). Arrows indicate SV40 T-ag at a molecular weight of approximately 97,000 dahons.

1083

R. K. BRIGHT et al.

1084 300

(A) r

400

1 :lOO Dilution

of serum

1: 100 Dilution

of serum

Fluorescence Intensity

@I r

Fluorescence Intensity

Fig. 3. Flow cytometric analysis of live mKSA cells stained with antibodies from mice immunized with SV40 T-ag synthetic peptides or recombinant SV40 T-ag. (A) Sera from BALB/c mice. (B) Sera from C57BL/6 mice. Immunization: none (pre), peptide 36-56 (36) peptide 121-141 (121) peptide 472490 (472), peptide 632-652 (632), peptide 660-679 (660), peptide 690-708 (690), and recombinant SV40 T-ag (T-ag). Bars indicating staining with sera from mice immunized with the control peptide (5077) are also shown.

Fig. 4. Representative histograms for flow cytometric analysis of viable mKSA cells stained with: (A) Sera collected from mice prior to immunization (preimmune). (B) Sera from mice following immunization with SV40 T-ag. Cells were incubated with sera and FITC conjugates as described in Materials and Methods. The fluorescence of 5000 cells was quantitated by flow cytometry with logarithmic amplification.

antibodies induced by SV40 T-ag peptides 121l141, 472-490, 632-652, 660-679 and 690-708 were 200, 150, 150, 110 and 100, respectively (Fig. 3B). Additionally, the CS7BL/6 antibody response to synthetic peptide 121-141 demonstrated cell surface reactivity equal to that of the BALB/c anti-SV40 T-ag antibodies (fluorescence intensity of 200). The observed increase in reactivity with mKSA (a BALB/c transformed cell line) by peptide induced antibody responses in C57BL/6 mice compared to anti-peptide antibodies in BALB/c mice, is probably not due to pre-existing antibodies that exhibit anti-cellular activity since sera collected from C57BL/6 mice prior to immunization demonstrated levels of reactivity comparable to that of the BALB/c preimmune sera (Fig. 3). The reason for this increased reactivity based on flow cytometry is unknown. These data suggest that antibodies generated by immunization with synthetic peptides corresponding to predicted Bcell epitopes on SV40 T-ag appear to be capable of recognizing native, conformationally intact SV40 T-ag expressed on the surface of tumorigenic, SV40 transformed cells.

Since murine antibodies induced by immunization with SV40 T-ag associated synthetic peptides were capable of recognizing SV40 T-ag by a variety of immunologic assays, it was of interest to ascertain whether those SV40 T-ag peptide induced responses were capable of protecting mice from a lethal inoculation with syngeneic SV40 transformed cells. The SV40 transformed C57BL/6 kidney fibroblast cell lines available do not possess the ability to form tumors in vivo. Therefore, this in vivo tumor protection study was performed utilizing BALB/c mice and the SV40 transformed cell line of BALB/c origin, designated mKSA. Individual groups of BALB/c mice were immunized with either SV40 T-ag or one of the six SV40 T-ag synthetic peptides, as described in Materials and Methods. The HIV-l associated synthetic peptide and alum each served as control immunizations. mKSA cells injected into control immunized mice were 100% lethal, with mean survival times ranging from 18-21 days and 15-24 days for alum and control peptide immunized mice, respectively. Likewise, mice immunized with the SV40 T-ag synthetic peptides corresponding to amino acid residues 3656, 121-141

EfSect of synthetic peptide immunization tumors in vivo

of SV40 induced

Synthetic peptide induced tumor immunity

and 66Q-679 did not survive more than 26 days (mean survival time ranged from 17-20.4 days) (Table 3). However, mice immunized with synthetic peptide 472490 demonstrated a higher mean survival time (27.2 days) compared to either the control groups or the three other SV40 T-ag peptide immunized groups of mice, but this was not statistically significant (p > 0.05). Moreover, the percent survival rates (greater than 60 days) for mice immunized with SV40 T-ag peptides 632-652 and 690-708 were approximately 50% of those of the SV40 T-ag immunized mice (Table 3). Compared to control, peptides 632-652 and 690-708 demonstrated statistically significant levels of protection (p < 0.05). The observed 100% survival rate of mice immunized with recombinant SV40 T-ag was not surprising since in earlier studies, immunizations with inactivated SV40 transformed cells appeared to protect mice from subsequent challenge with live tumor cells. These data demonstrate that immunization of BALB/c mice with specific synthetic peptides corresponding to predicted B-cell epitopes on SV40 T-ag induces partial tumor immunity capable of protecting up to 50% of the animals from a lethal tumorigenic dose of syngeneic SV40 transformed cells. DISCUSSION The purpose of this study was to examine and compare (using a variety of immunologic assays) the ability of antibodies induced by immunization of two genetically distinct inbred strains of mice with synthetic peptides corresponding to predicted B-cell epitopes on SV40 T-ag to recognize native epitopes on SV40 T-ag. Moreover, it was determined that some of the SV40 T-ag peptide induced immune responses in BALB/c mice were capable of in uivo tumor protection. Table 3. In uivo tumor immunity induced by immunization with SV40 T-ag peptides” Immunization Alum SV40 T-ag 50776 36-56 121-141 472-490 632452 660-679 690-708

Percentage of survivors 0 100 0 0 0 0 50 0 50

Mean survival time (days) 19.4 20.0 20.4 17.5 27.2 17.0

(18-21) (> 60.0) (15-24) (I 8-26) (7-23) (18-56) (8.0-> 60.0)d (7-22) (19.0->60.0)d

“Each group received six injections, then was rested for appoximately 30 days. A final injection was given 14 days prior to challenge with 5 LD~ of mKSA cells. bControl peptide associated with gp120 of HIV-l. ‘Values in parentheses represent the range of survival time in days. dDifferences in mean survival times for groups immunized with peptides 632-652 and 690-708 compared to controls were found to be significant (p > 0.05) based on statistical analysis utilizing a two tailed Student’s t-test for samples with unequal variance.

108.5

Both the C57BL/6 and BALB/c strains of mice generated antibodies to all six of the SV40 T-ag associated synthetic peptides. Interestingly, antibodies induced by SV40 T-ag synthetic peptide 472-490 in BALB/c mice demonstrated cross-reactivity with the peptide defining a carboxyl-terminal epitope on SV40 T-ag (690-708). The cross-reactivity was observed only in one direction and may have resulted from overall chemical similarities between the amino acids KGTGG of peptide 472-490 and KPPTPP of peptide 690-708. For example, both synthetic peptides share a positively charged lysine residue followed by uncharged, structure influencing residues (glycine and proline), and then a polar threonine residue that is capable of forming hydrogen bonds via the hydroxyl R-group. The reason for the lack of reciprocal cross-reactivity is not known; however, this suggests that a five amino acid continuous sequence within peptide 472-490 may define the cross-reactive epitope. Anti-peptide reagents used to examine cross-reactive variable region epitopes expressed on murine monoclonal antibodies with specificity for CD4 demonstrated that five amino acids in succession may be sufficient to define a minimal continuous epitope capable of being recognized by an anti-peptide reagent (Attanasio et al., 1993). Examination of the ability of peptide specific antibodies to bind SV40 T-ag by various immunologic assays revealed differences in SV40 T-ag reactivity between the two strains. Antibodies induced by four of the six synthetic peptides (472490, 632-652, 660-679 and 690-708) from both inbred mouse strains were capable of recognizing recombinant SV40 T-ag by ELISA. However, reactivity with SV40 T-ag was not as comparable when assessed by Western blot analysis of SV40 transformed cell lysates. BALB/c antibodies to peptides 472-490, 632-652, 66&679 and 690-708 were capable of recognizing epitopes on denatured SV40 T-ag by Western blot (based on recognition of a protein with an approximate molecular weight of 97 kD), whereas C57BL/6 antibodies to only three of the six peptides (472-490, 632-652 and 690-708) recognized SV40 T-ag by Western blot. The inability of the C57BL/6 antibodies induced by peptide 660-679 to recognize SV40 T-ag by Western blot suggested that this may represent a conformationally dependent epitope not expressed on denatured SV40 T-ag. To further characterize the antipeptide epitope reactivity, we examined the ability of these reagents to recognize the surface of SV40 transformed cells. Investigators from other laboratories have shown that SV40 T-ag is expressed on the surface of SV40 transformed cells as a whole protein with the amino and carboxy-terminal regions accessible to solution and the middle portion of the protein embedded in the plasma membrane (Butel and Jarvis, 1986). Native SV40 T-ag appeared to be recognized by peptide induced antibodies from both inbred strains of mice by flow cytometric analysis. It is not clear why C57BL/6 antibodies induced by five of the six SV40 T-ag peptides recognized the surface of SV40 transformed cells with greater fluorescence intensity than peptide induced

1086

R. K. BRIGHT

antibodies from BALB/c mice. The reactivity was most likely not the result of pre-existing anti-cellular responses in the C57BL/6 mice, since preimmune sera were not reactive, and C57BL/6 antibodies to the SV40 T-ag peptide 3656, as well as an irrelevant peptide, demonstrated only minor reactivity. Interestingly, the C57BL/6 anti-peptide antibodies induced by peptide 121-141 recognized SV40 T-ag by flow cytometry, but not by ELISA or Western blot. The lack of reactivity by Western blot may be due to a conformational dependence that is lost upon denaturation. The accessibility of epitopes on an antigen that is bound to a solid phase is sometimes hindered, this may reflect the lack of SV40 T-ag reactivity of the C57BL/6 anti-peptide (121-141) antibodies by ELISA. In addition, the peptide 121-141 corresponds to an epitope in the extreme amino-terminus of SV40 T-ag. It has been previously reported that the amino-terminus and the carboxy-terminus of SV40 T-ag are expressed and accessible as part of the intact native protein on the surface of SV40 transformed cells (Butel and Jarvis, 1986). Therefore, the C57BL/6 antibody response to peptide 121-141 appears to be restricted in reactivity with SV40 T-ag to the native conformational form expressed on the surface of SV40 transformed cells. BALB/c antibodies to two of the six SV40 T-ag peptides (632-652 and 690-708) that define distinct carboxylterminal epitopes appeared to recognize native SV40 T-ag expressed on the surface of mKSA cells. This reactivity indicated that antibodies to peptides 632-652 and 69g-708 not only recognized denatured continuous epitopes on SV40 T-ag, but were also capable of recognizing SV40 T-ag as it is expressed on tumor cell surfaces. A number of studies have examined peptide epitopes for their ability to induce protective immune responses to a variety of pathogenic organisms (reviewed in Steward and Howard, 1987; Rothbard, 1992). Alternatively, synthetic peptides have been used to define and map neutralizing epitopes and/or epitopes associated with protective immunity. For example, synthetic peptides defining epitopes on influenza virus hemagglutinin induced antibodies not normally elicited during the immune response to the native antigen. These antibodies were capable of neutralizing viral infection regardless of the serotype of the native hemagglutinin molecule (Muller et al., 1982). Additional studies demonstrated that immunization with synthetic peptides representing distinct regions of VP1 of polio virus induced virus neutralizing antibodies that were capable of priming animals for a subimmunogenic dose of virus (Emini et al., 1983). Our data demonstrate that synthetic peptides can be used to induce antibodies capable of recognizing native protein based tumor antigens. Additionally, immune responses elicited by synthetic peptides defining predicted B-cell epitopes on the native tumor antigen are capable of inducing protective tumor immunity in vivo. Epitopes defined by SV40 T-ag associated synthetic peptides corresponding to distinct carboxyl-terminal antigenic determinants induced SV40 T-ag specific tumor immune responses in BALB/c mice that were

et al.

capable of partially protecting animals from a lethal challenge with syngeneic SV40 transformed cells. This protection was not complete as only 40 to 60% of the mice exhibited survival times greater than 60 days. However, this was significantly greater than the control groups of mice and mice immunized with four of the six SV40 T-ag synthetic peptides. Previous studies demonstrated that BALB/c antibodies induced by immunization with baculovirus derived SV40 T-ag were capable of recognizing the synthetic peptide 690-708 that corresponds to a carboxyl-terminal epitope on SV40 T-ag. In addition, the epitope specificity of a monoclonal antibody generated to SV40 T-ag, which expresses an idiotype that may be involved in protective tumor immunity (Shearer et al., 1990; Bright et al., 1993) also recognized the carboxylterminal epitope defined by peptide 690-708 (Bright et al., 1994). Thus, it was concluded that this carboxylterminal epitope of SV40 T-ag may likely represent an immunodominant epitope in BALB/c mice that could serve as a focal point for manipulating protective humoral tumor immunity. Preliminary studies from our laboratory present evidence that antibodies with specificity for SV40 T-ag are capable of antibody dependent cell-mediated cytotoxicity of radiolabeled SV40 transformed cells in vitro (unpublished observation). Together, these data demonstrate that the C57BL/6 and BALB/c humoral immune responses to synthetic peptides defining selected epitopes on SV40 T-ag are capable of recognizing the native denatured protein, as well as cell surface expressed SV40 T-ag. The ability of selected SV40 T-ag peptides to induce partial tumor immunity in vivo further supports the role of the carboxyl-terminus of SV40 T-ag in the protective tumor immune response in BALB/c mice. Acknowledgements-The

authors would like to thank Dr Robert Lanford and MS Lena Notvall for the generous gift of the recombinant SV40 T-ag extract and MS Yolanda Arias for technical assistance. We would also like to thank Mr Angel Delgado for technical assistance with the flow cytometry experiments, and MS Tina Salazar and MS Jo Fletcher for assistance in preparing this manuscript. This study was supported in part by a grant from the American Cancer Society. REFERENCES Attanasio R., Kennedy R. C., Allan J. S., Maino V. C., Buck D. and Kanda P. (1990) Anti-idiotypic antibodies of a predefined specificity generated against CDR3Vn synthetic peptides define a private anti-CD4 idiotype. Molec. Immun. 27, 5 133522. Attanasio R., Kanda P., Stunz G. W., Buck D. W. and Kennedy R. C. (1993) Anti-peptide reagent identifies a primary-structure-dependent, cross-reactive idiotype expressed on heavy and light chains from a murine monoclonal anti-CD4. Molec. Immun. 30, 9-17. Bright R. K., Shearer M. H. and Kennedy R. C. (1993) Comparison of the murine humoral immune response to recombinant simian virus 40 large tumor antigen: epitope specificity and idiotype expression. Cancer Immunol. Im-

munother. 37, 31-39.

Synthetic

peptide

induced

Bright R. K., Shearer M. H. and Kennedy R. C. (1994) Fine specificity of the murine response to SV40 large tumour antigen utilizing synthetic peptides that define selected epitopes. Clin. exp. Immun. 96, 495499. Butel J. S. (1986) SV40 large T-antigen: dual oncogene. Cancer Surv. 5, 343-365. Butel J. S. and Jarvis D. L. (1986) The plasma-membraneassociated form of SV40 large tumor antigen: biochemical and biological properties. Biochim. biophys. Acta 865, 171-195. Butel J. S., Tevethia S. S. and Melnick J. L. (1972) Oncogenicity and cell transformation of papovavirus SV40: the role of the viral genome. Adv. Cancer Res. 15, l-55. Deckhut A. M., Tevethia M. J., Haggerty S., Frisque R. J. and Tevethia S. S. (1991) Localization of common cytotoxic T lymphocyte recognition epitopes on Simian papovavirus SV40 and human papovavirus JC virus T antigens. ViroIogy 183, 122-132. Dorries K., Loeber G. and Meixensberger J. (1987) Association of polyomaviruses JC, SV40, and BK with human brain tumors. Virology 160, 268-270. Emini E. A., Jameson B. A. and Wimmer E. (1983) Priming for and induction of anti-poliovirus neutralizing antibodies by synthetic peptides. Nature 304, 699-703. Fiori M. and di Mayorca G. (1976) Occurrence of BK virus DNA in DNA obtained from certain human tumors. Proc. natn. Acad. Sci. U.S.A. 73, 46624666. Frisque R. J., Bream G. L. and Canella M. T. (1984) Human polyomavirus JC virus genome. J. Viral. 51, 458469. Gurney E. G., Harrison R. 0. and Fenno F. (1980) Monoclonal antibodies against simian virus 40 T antigens: evidence for distinct subclasses of large T antigen and for similarities among nonviral T antigens. J. Virof. 34,752-763. Harlow E., Crawford L. V., Pim D. C. and Williamson N. M. (1981) Monoclonal antibodies specific for simian virus 40 tumor antigens. J. Virol. 39, 861-869. Kit S., Kurimura T. and Dubbs D. R. (1969) Transplantable mouse tumor line induced by injection of SV40-transformed mouse kidney cells. Int. J. Cancer 4, 384-392. Krieg P., Amtmann E., Jonas D., Fischer H., Zang K. and Sauer G. (198 1) Episomal simian virus 40 genomes in human brain tumors. Proc. natn. Acad. Sci. U.S.A. 78, 64466450. Lanford R. E. (1988) Expression of simian virus 40 T antigen in insect cells using a baculovirus expression vector. ViroIogy 167, 72-8 1.

tumor

immunity

1087

Levine A. J. (1990) The p53 protein and its interactions with the oncogene products of the small DNA tumor viruses. Virology 177, 419426. Mernaugh R. L., Shearer M. H., Bright R. K., Lanford R. E. and Kennedy R. C. (1992) Idiotype network components are involved in the murine immune response to simian virus 40 large tumor antigen. Cancer Immunol. Immunother. 35, 113-118. Muller G. M., Shapira M. and Arnon R. (1982) Anti-influenza response achieved by immunization with a synthetic conjugate. Proc. natn. Acad. Sci. U.S.A. 79, 569-573. Pauletti D., Simmonds R., Dreesman G. R. and Kennedy R. C. (1985) Application of a modified computer algorithm in determining potential antigenic determinants associated with the AIDS virus glycoprotein. Anal. Biochem. 151, 540-546. Rosner B. (1990) Fundamentals of Biostatistics, 3rd edn, p. 268. PWS-KENT Publ. Co., Boston, MA. Rothbard J. B. (1992) Synthetic peptides as vaccines. In Vaccines: New Approaches to Immunological Problems (Edited by Ellis R. W.), pp. 451465. Butterworth-Heinemann, Boston. Shearer M. H., Lanford R. E. and Kennedy R. C. (1990) Monoclonal antiidiotypic antibodies induce humoral immune responses specific for simian virus 40 large tumor antigen in mice. J. Immun. 145, 932-939. Shearer M. H., Bright R. K., Lanford R. E. and Kennedy R. C. (1993) Immunization of mice with baculovirus-derived recombinant SV40 large tumor antigen induces protective tumour immunity to a lethal challenge with SV40-transformed cells. Clin. exp. Immun. 91, 266271. Simanis V. and Lane D. P. (1985) An immunoaffinity purification procedure for SV40 large T antigen. Virology 144, 88-100. Soule H. R., Lanford R. E. and Butel J. S. (1980) Antigenic and immunogenic characteristics of nuclear and membraneassociated simian virus 40 tumor antigen. J. Viral. 33, 887-901. Steward M. W. and Howard C. R. (1987) Synthetic peptides: a next generation of vaccines? Immun. Today 8, 51-58. Tevethia S. S. (1980) Immunology of simian virus 40. In Viral Oncology (Edited by Klein G.), pp. 581601. Raven Press, New York. Tevethia S. S. (1990) Recognition of simian virus 40 T antigen by cytotoxic T lymphocytes. Molec. biol. Med. 7, 83-96.