- Email: [email protected]
Enhancement of Epstein-Barr virus membrane protein (LMP) expression by serum, TPA, or n-butyrate in latently infected Raji cells
VIROLOGY
159, 161-165 (1987)
Enhancement of Epstein-Barr Virus Membrane Protein (LMP) Expression by Serum, TPA, or n-Butyrate in Latently Infected Raji Cells HARALD Boos, RUDOLFBERGER,CORNELIAKUKLIK-ROOS,THOMAS IFTNER,* AND NIKOLNJS MUELLER-LANTZSCH’ lnstitut fsir Virologie, Zentrum fur Hygiene der Universitat Virologie der Universit& Erlangen-Nijmberg,
Freiburg, Hermann-Herder Strasse 11, 7800 Freiburg, and *i~@%~t fsir Klinische Loschgestrasse 7, 8520 Eriangen, Federal Republic of Germany
Received September
24, 1986; accepted
February
12, 1987
The 5amHl Nhet region of the EBV DNA is known to code for two proteins. One is a membrane protein (LMP) with an apparent molecular weight of 60,000 on SDS-PAGE which is expressed in ‘tatentiy EBV infected cells. The second protein, so far unidentified, is presumably a late protein with a calculated molecular weight of 28,000 !I&, Antisera against both proteins were generated by immunizing rabbits with either a fusion protein containing 166 amino acids of the C-terminus of LMP and a 37,000 mol wt piece of the bacterial anthranilate synthase or with a C-term&at synthetic peptide of 7 amino acids. These sera reacted with a protein varying in size between 60,000 and 65,000,mol wt on SDS-PAGE, found in all cell lines harboring EBV. In addition, these sera identified a second protein with,en apparent molecular weight of 49,000 on SDS-PAGE in 895-8, P3HR-1, and M-ABA ceils, which is presumably identZcal’with the 28,QOO-Daprotein mentioned above. Furthermore, with these sera a positive cytoplasmic immutifluorescence in 1 to 10% of the cells was obtained, depending on the cell line examined. Ana’lyzing the nonproducer Raji cell line, the number of immunofluorescence-positive cells and the amount of the 60,000 protein, as judged by immunoblotting,~ was rapidly increased by addition of fresh medium with 10% fetal calf serum as well as by the tumor promoter TPA or to an even higher extend by n-butyrate. The kinetics of induction reached 9 maximum 24 hr after add&&n of medium plus 10% fresh serum or TPA or n-butyrate and decreased after 24 to 48 hr. Since the induction of’ the E&V early antigen (EA) associated proteins by TPA or n-butyrate exhibits a diverse kinetic with a maximum at 72 hr, the regulation of the 60,000 protein synthesis appears to be different from known EA-associated proteins. o 1987 Academic Press, I~C. Epstein-Barr virus (EBV) is infectious mononucleosis and
the causative agent of is also associated with two human malignant tumors, Burkitt’s lymphoma and nasopharyngeal carcinoma (1). Furthermore, EBV has the ability to immortalize human B lymphocytes by establishing a latent infection in which the majority of viral genes is not expressed (2). Immortalized B cells, as well as tumor cells harboring EBV genomes, express an EBV-associated nuclear antigen (EBNA) which can be detected by anticomplementary immunofluorescence (ACIF; 3). So far, four viral genes have been found to be expressed in latently EBV-infected cells (4-9). These genes code for at least three EBNA-associated nuclear proteins and one potential membrane protein (4, 6, 7, 9-l 1). The potential membrane protein is encoded by the BarnHI Nhet legion (6, 8, 12) and exhibits an apparent mo~lecularweight of 60,000 but of varying sizes in different EBV genome-harboring lymphoid cell lines. Since two promoters (ED-L1 and ED-LlA) map in the BamHl Nhet region, two proteins were predicted with calculated molecular weights of 43,000 and 28,000 sharing a common C-terminal region (8; Fig. 1). The 43,000 protein has been proven to be identical with
the protein LMP of apparent mol wt 60,000 (6, 12). The second, presumably late gene product has not been identified so far. Here we report that two antisera derived either against a bacterial fusion protein containing 155 amino acids from the C-terminus or against a synthetic peptide of 7 amino acids from the C-terminus of both predicted proteins recognize, beside the 60,000 protein, a second protein of apparent mol wt 49,000 in 895-8 and M-ABA EBV producer cell lines., This 49,000 protein was not detectable in several other producer and nonproducer cell lines. We provide evidence that in nonproducer Rajjicells the synthesis of the 60,000 protein is.rapidly!induced by serum or the tumor promotter 124O-tetrad,ecanoylphorbol-13-acetate (TPA) or n-butyrate. Since the kirietics of induction of this protein: differs from the induction of knowu early antigen (EA)-associ!ted proteins, the synthesis of this ‘60,000 protein r;rjight be regulated differetitly from the exprestiion of tt-$ EA-associated proteins. Antisera were~ raised in rabbits either against the TrpE-LMP fusion protein or, against the cautioxy-terminal’ seven amino acids as descrtbad in the i’ftgend to Fig. 2. These sera,,were a’ssayed by eleaZrophoretic immunobl!otting vsing EBV genome-negative and -pasitive’ lymphoid cell lines. The results (Figs. 2Asand 2B)
’ To whom requests for reprints should be addressed. 161
0042-6822187
$3.00
Copyright G3 1987 by Academic Press, Inc. All rights of reproduction in any form resewed.
162
SHORT COMMUNICATIONS BamHI
EmHI
N fragment 169
168 e : F I
I
of EBV genome 170
171
latent
transcript
late
11
expressed synthetic
BamHI ILMP)
in Ecoli peptide
FIG. 1. Diagram of promoter studies from a recent report (8). Open bars show the coding region of the latent membrane protein (LMP) and the late 28,000 protein. The closed bars indicate the parts of the proteins which were used for immunization.
demonstrate that both sera react with a viral protein having a variable mol wt of 60,000 to 65,000, although in some cell lines (i.e., P3HR-I) the protein was hardly visible on the original picture. This latter result is in agreement with recent reports where quantitative differences of LMP expression in different Burkitt lymphoma cell lines observed (6, 13). In addition, a smaller protein with apparent molecularweight of about 49,000 could be observed in B95-8 cells as well as in M-ABA cells (Fig. 2). We believe that this protein with the apparent molecular weight of 49,000 is identical to the 28,000 mol wt protein deduced by EBV-DNA sequence analysis and promotor studies (8). Differences in calculated and apparent molecular weights on SDS-PAGE are often observed. For example, the human p53 has a calculated molecular weight of 43,000. On SDSPAGE this protein exhibits an apparent molecular weight of 53,000 as estimated by relative molecular mass (I’&) markers (74). The exact reasons for the discrepancy between the calculated and apparent molecular weig,hts of the 49,000 protein on SDS-PAGE are uncertain. Nevertheless, this feature may result from a high number of prolin residues (and particularly ProPro pairs) in the 49,000 protein. Furthermore, these differences could be due to post-translational modifications. The existence of the 49,000 mol wt protein is confirmed by recent observations (12, 13). Interestingly, this 49,000 protein was not found in all EBV genomeharboring lymphoid cell lines, indicating virus strainspecific d’ifferences with regard to the structure or the expressioin of this particular viral DNA region. The reaction of the anti-peptide serum with LMP in Raji and BL74 cells as well as with LMP and 49,000 mol wt protein in ‘B95-8 cells could be specifically inhibited by the addition of synthetic peptides prior to the immunoreaction (Fig. 2C), while unspecific bands were not diminished. Therefore, we conclude that both types of antisera react specifically and that the proteins identified are the gene product of the BNLFla-c reading frame (15). Furthermore, EBV genome-negative BJAB cells and EBV-positive Jijoye, B95-8, and Raji cells were
assayed by indirect immunofluorescence using antisera against the fusion protein or synthetic peptide. Both types of sera exhibited titers greater than 1280 on Jijoye cells as judged by endpoint dilution. Even a 64-fold higher concentration (dilution 1:20) of both sera was able to stain only a part of the Jijoye cell population (510%) while a reasonable number of cells were negative in immunofluorescence. A representative result is shown in Fig. 3. No positive reaction could be observed using preimmune sera (Fig. 3). The positive reaction with the antipeptide serum could be inhibited completely by addition of synthetic peptide (data not shown) providing direct evidence for the specificity of the reaction. Furthermore, the amount of immunofluorescence-positive cells could not be increased by applying the anticomplementary immunofluorescence, although the nonspecific background was slightly elevated in EBV genome-negative and -positive lymphoid cell lines (data not shown). Therefore, this result indicates that the low number of positive cells (5-100~) was not due to an insensitive test. The small percentage of immunofluorescence-positive cells indicates that LMP is expressed in different amounts within a cell population. These data are in contrast to the present understanding
.I
-9L -67
r**
--*-“‘.-, I
-L3 -30
-+-+-+-+--C-+--f-+ A
B
C
(24) of EBV genomeFIG. 2. Protein analysis by immunoblotting negative (BJAB) and -positive lymphoid cell lines (695-8, Raji, Jijoye, Raji, BL74, QIMR-WIL, M-ABA, and PBHR-1). Identical quantities of proteins from each cell extract were applied to the SDS-PAGE (25, 26). (A) Cells were harvested 3 days after cell subculturing without TPA treatment (--) or with TPA treatment (+) (20 rig/ml) using an rabbit antiserum against a bacterial fusion protein. This fusion protein was expressed in the tryptophan-regulated E. co/i expression vector PATH 1 (Koerner, unpublished results; 27) containing 155 amino acids from the carboxy-terminal part of the BNLFla-c open reading frame (15) of cloned EBV M-ABA DNA (28) and the amino-terminal part of the anthranilate synthase of f. co/i (pAH5, H. Boos, unpublished results). (B) Indicated cell lines were analyzed by a rabbit antiserum against a synthetic peptide consisting of seven amino acids (H-Val-Gin-Leu-Ser-Tyr-Tyr-Asp-OH) purchased from Bachem AG, Switzerland. The procedure for generating the antisera was described recently (29). (C) Competition of the immunoreaction as described in (B) with free peptide. Serum (40 pl) was incubated with 10 pl PBS containing the free synthetic peptide (1 mg/ml) for 1 hr before immunoreaction. The apparent molecular weights are estimated by comparison with comigrating marker proteins.
SHORT COMMUNICATIONS
FIG. 3. Indirect immunofluorescence serum.
analysis of Jijoye cells using antisera against pAH5 fusion protein. (A) Preimmune
of the nature of a protein expressed in latently EBVinfected ceils. For example, EBNA proteins identified so far could be detected in 100% of the ceils from an EBV genome-positive cell line. Therefore, we have attempted to characterize the regulation of the membrane protein synthesis in latently EBV-infected cell lines. In a first series of experiments different lymphoid cell lines harboring EBV genomes were treated with the phorbol ester TPA which is a potent inducer of EBV
163
serum. (B) Immune
proteins associated with the EA and late virus capsid antigen (VCA) complex ( 16- 18). Cells were harvested at Day 3 after TPA treatment. At this time maximal EA and VCA protein synthesis can be observed (19). Cell extracts were furthermore analyzed by electrophoretic immunoblotting. The results (Fig. 2A) demonstrate that some cell lines (Raji, BL74, QUIMR-WIL, P3HR-1) exhibit increased synthesis of the LMP or the 49,000 mol wt protein (M-ABA), while in 895-8 and Jijoye cells this
SHORT COMMUNICATIONS
164
effect could not be observed. Since in Raji cells the induction of the 60,000-Da membrane protein by TPA was very clear, the synthesis of this protein in Raji cells was further characterized. Freshly subcultured cells were harvested after addition of TPA, n-butyrate, or a combination of TPA and n-butyrate at the hours indicated in Fig. 4 and were analyzed in parallel either by electrophoretic immunoblotting or by indirect immunofluorescence. Freshly subcultered cells without TPA or n-butyrate treatment served as a control. The results of one out of three independent experiments are shown (Fig. 4). The synthesis of this protein was induced to a reasonable extent at 7 hr after addition of fresh medium with 10% fetal calf serum alone. On the other hand, the results again clearly demonstrate the increased expression of the 60,000 membrane protein in the presence of TPA or n-butyrate, reaching a maximum at 24 hr. These results are confirmed by data obtained from immunofluorescence assays. The amount of LMP in immunofluorescence-positive cells as judged by a more brilliant staining as well as the number of LMPpositive cells is increased by the addition of TPA or nbutyrate (data not shown). Treatment of Raji cells with n-butyrate, for example, increased the number of LMPpositive cells from approximately 3% to 50-609~ (data not shown). When a combination of TPA and n-butyrate was applied to Raji cells no synergistic effect was observed and the expression of LMP was similar to the amount found with TPA alone (Fig. 4). These findings are confirmed by .data obtained from immunofluorescence assays. In addition to LMP, the induction of the
0
0
7 2G. 48 72
7 2L L8 72 0
d
7 2L L8 72
7 2L
L8 72
FIG. 4. Protein analysrs of Raji cells cultivated with medium and 10% fresh fetal calf serum (A) and additional TPA (6). n-butyrate (C). or a combination of TPA and n-butyrate (D) using the technique and rabbit antiserum as described in Fig. 2. Cells were harvested at in drcated hours after subculturing.
EA complex in these TPA, n-butyrate, or TPA and nbutyrate treated Raji cells was assayed in parallel using a human anti-EA positive serum. This protein complex was synthesized to a maximal amount at 72 hr after addition of both drugs as estimated by immunofluorescence and electrophoretic immunoblotting and exhibited a different kinetic to LMP (data not shown). Seventy-two hours after a combined treatment with TPA and n-butyrate, more than 60% EA immunofluorescence-positive cells were determined (data not shown). These data are in agreement with previous reports (19, 20). No enhanced synthesis of EBNA 1 and EBNA 2 proteins could be observed after TPA, n-butyrate, or TPA and n-butyrate application (data not shown). The most surprising observation in these experiments is that LMP is expressed only in a distinct percentage of an EBV latently infected cell population and that expression of the 60,000 mol wt LMP can be enhanced in Raji cells by addition of medium with 10% fresh serum, TPA, or n-butyrate. From these data we suggest that the 60,000 LMP is synthesized during a defined period of the cell cycle, or the presence of this protein is not necessarily a prerequisite for the maintenance of the immortalized state of a cell. The latter consideration would link this 60,000 LMP to a class of viral proteins associated with the EA complex of EBV. The induction of this membrane protein by TPA or nbutyrate would favor this idea. The. kinetics of LMP induction in Raji cells by inductors and the induction by serum ~itself, however, argue.against the possibility that the LMP belongs to the EA complex. Furthermore, LMP is expressed in latently infected Raji and lijoye cells to a much higher degree (3-15%) than the proteins of the EA-complex (approximately O.l%), as judged by immunofluorescence (data not shown). In addition, in contrast to the EA complex, LM,P expression is not enhanced in a synergistic manner by a combination of TPA and n-butyrate. These~data indicate that LMP is not a member of the EA complex. At the present time one cannot rule out the possibility that the 60,000 mol wt LMP is one of the first proteins of the EA protein cascade or a prerequisite for the induction of the lytic or abortive cycle of the EBV. Preliminary experiments showed no evidence for a cell cycle-dependent expression of the 60,000 mol wt protein (R. Berger and W. MOller, unpublished results). At the present time nothing is known about the mechanism of the 60,000Da membrane protein induction by TPA, n-butyrate, or serum alone. Our results on the enhanced expression of LMP by TPA or n-butyrate are in some contrast to very recent findings (73). These differences might be due to different cell lines, different methods, and the combined treatment of TPA and n-butyrate, since in this report
SHORT COMMUNICATIONS
B95-8 immortalized ceils treated with TPA and n-butyrate for 24 hr were labeled metabolically with [2”S]methionine. Furthermore, these cells were then analyzed by immunoprecipitation using an antipeptide serum against the N-terminal part of LMP. No 60,000 mol wt LMP could be identified under these conditions. The fun.ction of the 60,000 and 49,000 proteins are still unknown. From previous studies evidence suggests localization in the cell membrane (72, 27). But so far, there is no strong proof that this 60,000 LMP is the lymphocyte detected membrane antigen (LYDMA) and itself the target for cytotoxic T cells, although recently some evidence for this idea was provided (22). DNA sequence homology studies did not reveal any significant homology to any known protein sequence, although a minor homology to the 7-amino acid Glu-Arg-Gly-Pro-Pro-Gly-Pro stretch in the carboxy-terminal part of the protein to the sequence of collagen could be observed. Furthermore, data were presented for the transforming capacity of the 60,000 LMP gene (23). In addition, this protein could be identified in virus particles (72). It remains to be elucidated which functional role this membrane protein has during the viral life cycle and for the immortalization of a B lymphocyte. ACKNOWLEDGMENTS We are grateful to T. .I. Koerner for his supply of the expression vector pATH. These studies were supported by the Deutsche Forschungsgameinschaft (SFB 31 -Medizinische Virologie: Tumorentstehung und -Entwicklung)‘and by the Bundesministerium fiir Forschung and Technologie (ECT 3llA).
REFERENCES 7. EPSTEIN, M. A., and ACHONG, B. G. (Eds.), “The Epstein-Barr Virus.” Springer-Varlag, Berlin, 1979. 2. HENLE, W., DIEHL, V., KO~IN, G., ZUR HAUSEN, H., and HENLE, G., Science 157, 1064-1065 (1967). 3. REEDMAN, B., and KLEIN, G., Inr. J. Cancer 11, 599-620 (1973). 4. SUMMERS, W. P., GROGAN, W. A., SHEDD, D., ROBERT, M., LIU, D-R., and MILLER, G., Proc. Nat/. Acad. Sci. USA 79, 5688-
5692 (1982). 5. HWNESSY, K., and KIEFF,E., Proc. Natl. Acad. Sci. USA 80, 56655669 (1983).
‘I65
6. HENNESSY,K., FENNEWALD,S., HUMMEL, M.. COLE, T., and KIEFF, E., Proc. Natl. Acad. Sci. USA 81, 7207-7211 (1984). 7. HENNESSY,K., FENNNVALD, S., and KIEFF, E., Proc. Natl. Acad.
Sci. USA 82,5944-5948
(1985).
8. HUDSON, G. S., FARRELL, P. J., and BARR&L, 5. G., /. l&o/. 53,
528-535 (1985). 9. MUELLER-LANTZSCH,N., LENOIR, G. M., SAUTER, M., TAKAKI, K., B~CHET, J. -M., KUKLIK-ROES,C., WuNDsstIcr-r, D., and BORNKAMM, G. W., E/‘&30/. 4, 1805-1811 (1985). 10. HENNESSY,K., WANG, F., WOODLAND:BUSHMAN, E., and KIEFF, E.,
Proc. Natl. Acad. Sci. USA 83, 5693-5697 (1986). 11. KALLIN, B., DILLNER,J., EIMBERG,J., EHLIN-HENRIKSSON,B,, ROS&N, A., HENLE, W., HENLE, G., and KLEIN, G., fioc. Nati. Acad. Sci. USA 83, 1499-l 503 (1986). 12. MANN, K. P., STAUNTON, D., and THORLEY-LAWSON,D. A., J. Viral. 55,7 1O-720 (1985). 13. MODROW,S., and WOLF, H., Proc. Nat/. Acad. Sci. USA 83,57035707 (1986). 14. OREN, M., Biochim. Biophys. Acfa 623, 67-78 (1985). 15. BAEA, R., BANKIER,A. T., BIGGIN, M. D., DEINI~~GER,P. L., FARR~L, P. J., GIBSON, R. J., HATFULL, G.; HUOSON, G. S.. SATCHWELL, S. C., %GUIN, C., TUFFNELL, P. S., and BARRELL, B. G., Narure (London) 310,207-211 (1984). 16. ZUR HAUSEN, H., O’NEILL, F. J., FREESE,U. K., and HECKER, E.,
Nature (London) 272,373-375 X1978). 17. ZUR HAUSEN,‘H., BORNKAMM,G. W., SCH~MIDT,I?., and HECKER,E., Prac. NatI. Acad. Sci. USA 76, 782-765 (1879). 18. MUELLER-L&~~STZSCH, N., YAMAMOTO, N., and ZUR HAUSEN, H., I/iroloby 97, 378-387 (1979). 19. MUELLER-LAFITZSCH,N., GEORG,B., YAMAMOTO, and ZUR HAUSEFJ, H., Virology 102, 4Oi-411 (1980). 20. ITO, Y., QW~NISHI, M., HARAYMA,T., ~~~‘TA~BAYAsHI, S., Cancer Lett, 12,175-180
(1981).
21. LIEBOWITZ,D., WANG, D., and KIEFF, E., 1. l&o/. 58, 233-237 (1986). 22. SULITZEANU,D., SZIGETI, R., KLEIN, G., HENNESSY,K., and KIEFF, E., 1. Viral. 58, 230-232 (1986). (1984). 23. WANG, D., LIEBOWIR, D.; and KIEFF, E., @l/43,831-840 24. TO~BIN, H., STAEHELIN.T., and GORDON% J., Pruc. Nat/. Acad. Sci. USA 76, 43&O-4354,(1 979). 25. LOWRY, 0. H., RO~EBR~WGH!N. J., FARR, A. L., and RANDALL, R. J., 1. Biol. Chem, 193, 265-275 (1651). 26. LAEMMLI, U. K., Nature (London) 277, 680-686 (1970). 27. SPINDLER, K. R., ROUSER,D. S. E., and BERK, A. J., J. Viral. 49, 132-141 (1984). 28. POLACK,A., HARTL, G., ZIMCTER,U., FREE% U. K., LAux, G., TAKAKI, K., HOLM, B,. GISSMANN, L., and BORNKAMM, G. W., Gene 27,
279-288 (1984). 29. WALTER, G., SCWEIDTMANN,K, -H., CAR&ONE,A., LAUDANO, A. P.. and DOOLITTLE, R. F., Proc. Nat/. Acad. Sci. USA 77, 51975200 (1 980).
159, 161-165 (1987)
Enhancement of Epstein-Barr Virus Membrane Protein (LMP) Expression by Serum, TPA, or n-Butyrate in Latently Infected Raji Cells HARALD Boos, RUDOLFBERGER,CORNELIAKUKLIK-ROOS,THOMAS IFTNER,* AND NIKOLNJS MUELLER-LANTZSCH’ lnstitut fsir Virologie, Zentrum fur Hygiene der Universitat Virologie der Universit& Erlangen-Nijmberg,
Freiburg, Hermann-Herder Strasse 11, 7800 Freiburg, and *i~@%~t fsir Klinische Loschgestrasse 7, 8520 Eriangen, Federal Republic of Germany
Received September
24, 1986; accepted
February
12, 1987
The 5amHl Nhet region of the EBV DNA is known to code for two proteins. One is a membrane protein (LMP) with an apparent molecular weight of 60,000 on SDS-PAGE which is expressed in ‘tatentiy EBV infected cells. The second protein, so far unidentified, is presumably a late protein with a calculated molecular weight of 28,000 !I&, Antisera against both proteins were generated by immunizing rabbits with either a fusion protein containing 166 amino acids of the C-terminus of LMP and a 37,000 mol wt piece of the bacterial anthranilate synthase or with a C-term&at synthetic peptide of 7 amino acids. These sera reacted with a protein varying in size between 60,000 and 65,000,mol wt on SDS-PAGE, found in all cell lines harboring EBV. In addition, these sera identified a second protein with,en apparent molecular weight of 49,000 on SDS-PAGE in 895-8, P3HR-1, and M-ABA ceils, which is presumably identZcal’with the 28,QOO-Daprotein mentioned above. Furthermore, with these sera a positive cytoplasmic immutifluorescence in 1 to 10% of the cells was obtained, depending on the cell line examined. Ana’lyzing the nonproducer Raji cell line, the number of immunofluorescence-positive cells and the amount of the 60,000 protein, as judged by immunoblotting,~ was rapidly increased by addition of fresh medium with 10% fetal calf serum as well as by the tumor promoter TPA or to an even higher extend by n-butyrate. The kinetics of induction reached 9 maximum 24 hr after add&&n of medium plus 10% fresh serum or TPA or n-butyrate and decreased after 24 to 48 hr. Since the induction of’ the E&V early antigen (EA) associated proteins by TPA or n-butyrate exhibits a diverse kinetic with a maximum at 72 hr, the regulation of the 60,000 protein synthesis appears to be different from known EA-associated proteins. o 1987 Academic Press, I~C. Epstein-Barr virus (EBV) is infectious mononucleosis and
the causative agent of is also associated with two human malignant tumors, Burkitt’s lymphoma and nasopharyngeal carcinoma (1). Furthermore, EBV has the ability to immortalize human B lymphocytes by establishing a latent infection in which the majority of viral genes is not expressed (2). Immortalized B cells, as well as tumor cells harboring EBV genomes, express an EBV-associated nuclear antigen (EBNA) which can be detected by anticomplementary immunofluorescence (ACIF; 3). So far, four viral genes have been found to be expressed in latently EBV-infected cells (4-9). These genes code for at least three EBNA-associated nuclear proteins and one potential membrane protein (4, 6, 7, 9-l 1). The potential membrane protein is encoded by the BarnHI Nhet legion (6, 8, 12) and exhibits an apparent mo~lecularweight of 60,000 but of varying sizes in different EBV genome-harboring lymphoid cell lines. Since two promoters (ED-L1 and ED-LlA) map in the BamHl Nhet region, two proteins were predicted with calculated molecular weights of 43,000 and 28,000 sharing a common C-terminal region (8; Fig. 1). The 43,000 protein has been proven to be identical with
the protein LMP of apparent mol wt 60,000 (6, 12). The second, presumably late gene product has not been identified so far. Here we report that two antisera derived either against a bacterial fusion protein containing 155 amino acids from the C-terminus or against a synthetic peptide of 7 amino acids from the C-terminus of both predicted proteins recognize, beside the 60,000 protein, a second protein of apparent mol wt 49,000 in 895-8 and M-ABA EBV producer cell lines., This 49,000 protein was not detectable in several other producer and nonproducer cell lines. We provide evidence that in nonproducer Rajjicells the synthesis of the 60,000 protein is.rapidly!induced by serum or the tumor promotter 124O-tetrad,ecanoylphorbol-13-acetate (TPA) or n-butyrate. Since the kirietics of induction of this protein: differs from the induction of knowu early antigen (EA)-associ!ted proteins, the synthesis of this ‘60,000 protein r;rjight be regulated differetitly from the exprestiion of tt-$ EA-associated proteins. Antisera were~ raised in rabbits either against the TrpE-LMP fusion protein or, against the cautioxy-terminal’ seven amino acids as descrtbad in the i’ftgend to Fig. 2. These sera,,were a’ssayed by eleaZrophoretic immunobl!otting vsing EBV genome-negative and -pasitive’ lymphoid cell lines. The results (Figs. 2Asand 2B)
’ To whom requests for reprints should be addressed. 161
0042-6822187
$3.00
Copyright G3 1987 by Academic Press, Inc. All rights of reproduction in any form resewed.
162
SHORT COMMUNICATIONS BamHI
EmHI
N fragment 169
168 e : F I
I
of EBV genome 170
171
latent
transcript
late
11
expressed synthetic
BamHI ILMP)
in Ecoli peptide
FIG. 1. Diagram of promoter studies from a recent report (8). Open bars show the coding region of the latent membrane protein (LMP) and the late 28,000 protein. The closed bars indicate the parts of the proteins which were used for immunization.
demonstrate that both sera react with a viral protein having a variable mol wt of 60,000 to 65,000, although in some cell lines (i.e., P3HR-I) the protein was hardly visible on the original picture. This latter result is in agreement with recent reports where quantitative differences of LMP expression in different Burkitt lymphoma cell lines observed (6, 13). In addition, a smaller protein with apparent molecularweight of about 49,000 could be observed in B95-8 cells as well as in M-ABA cells (Fig. 2). We believe that this protein with the apparent molecular weight of 49,000 is identical to the 28,000 mol wt protein deduced by EBV-DNA sequence analysis and promotor studies (8). Differences in calculated and apparent molecular weights on SDS-PAGE are often observed. For example, the human p53 has a calculated molecular weight of 43,000. On SDSPAGE this protein exhibits an apparent molecular weight of 53,000 as estimated by relative molecular mass (I’&) markers (74). The exact reasons for the discrepancy between the calculated and apparent molecular weig,hts of the 49,000 protein on SDS-PAGE are uncertain. Nevertheless, this feature may result from a high number of prolin residues (and particularly ProPro pairs) in the 49,000 protein. Furthermore, these differences could be due to post-translational modifications. The existence of the 49,000 mol wt protein is confirmed by recent observations (12, 13). Interestingly, this 49,000 protein was not found in all EBV genomeharboring lymphoid cell lines, indicating virus strainspecific d’ifferences with regard to the structure or the expressioin of this particular viral DNA region. The reaction of the anti-peptide serum with LMP in Raji and BL74 cells as well as with LMP and 49,000 mol wt protein in ‘B95-8 cells could be specifically inhibited by the addition of synthetic peptides prior to the immunoreaction (Fig. 2C), while unspecific bands were not diminished. Therefore, we conclude that both types of antisera react specifically and that the proteins identified are the gene product of the BNLFla-c reading frame (15). Furthermore, EBV genome-negative BJAB cells and EBV-positive Jijoye, B95-8, and Raji cells were
assayed by indirect immunofluorescence using antisera against the fusion protein or synthetic peptide. Both types of sera exhibited titers greater than 1280 on Jijoye cells as judged by endpoint dilution. Even a 64-fold higher concentration (dilution 1:20) of both sera was able to stain only a part of the Jijoye cell population (510%) while a reasonable number of cells were negative in immunofluorescence. A representative result is shown in Fig. 3. No positive reaction could be observed using preimmune sera (Fig. 3). The positive reaction with the antipeptide serum could be inhibited completely by addition of synthetic peptide (data not shown) providing direct evidence for the specificity of the reaction. Furthermore, the amount of immunofluorescence-positive cells could not be increased by applying the anticomplementary immunofluorescence, although the nonspecific background was slightly elevated in EBV genome-negative and -positive lymphoid cell lines (data not shown). Therefore, this result indicates that the low number of positive cells (5-100~) was not due to an insensitive test. The small percentage of immunofluorescence-positive cells indicates that LMP is expressed in different amounts within a cell population. These data are in contrast to the present understanding
.I
-9L -67
r**
--*-“‘.-, I
-L3 -30
-+-+-+-+--C-+--f-+ A
B
C
(24) of EBV genomeFIG. 2. Protein analysis by immunoblotting negative (BJAB) and -positive lymphoid cell lines (695-8, Raji, Jijoye, Raji, BL74, QIMR-WIL, M-ABA, and PBHR-1). Identical quantities of proteins from each cell extract were applied to the SDS-PAGE (25, 26). (A) Cells were harvested 3 days after cell subculturing without TPA treatment (--) or with TPA treatment (+) (20 rig/ml) using an rabbit antiserum against a bacterial fusion protein. This fusion protein was expressed in the tryptophan-regulated E. co/i expression vector PATH 1 (Koerner, unpublished results; 27) containing 155 amino acids from the carboxy-terminal part of the BNLFla-c open reading frame (15) of cloned EBV M-ABA DNA (28) and the amino-terminal part of the anthranilate synthase of f. co/i (pAH5, H. Boos, unpublished results). (B) Indicated cell lines were analyzed by a rabbit antiserum against a synthetic peptide consisting of seven amino acids (H-Val-Gin-Leu-Ser-Tyr-Tyr-Asp-OH) purchased from Bachem AG, Switzerland. The procedure for generating the antisera was described recently (29). (C) Competition of the immunoreaction as described in (B) with free peptide. Serum (40 pl) was incubated with 10 pl PBS containing the free synthetic peptide (1 mg/ml) for 1 hr before immunoreaction. The apparent molecular weights are estimated by comparison with comigrating marker proteins.
SHORT COMMUNICATIONS
FIG. 3. Indirect immunofluorescence serum.
analysis of Jijoye cells using antisera against pAH5 fusion protein. (A) Preimmune
of the nature of a protein expressed in latently EBVinfected ceils. For example, EBNA proteins identified so far could be detected in 100% of the ceils from an EBV genome-positive cell line. Therefore, we have attempted to characterize the regulation of the membrane protein synthesis in latently EBV-infected cell lines. In a first series of experiments different lymphoid cell lines harboring EBV genomes were treated with the phorbol ester TPA which is a potent inducer of EBV
163
serum. (B) Immune
proteins associated with the EA and late virus capsid antigen (VCA) complex ( 16- 18). Cells were harvested at Day 3 after TPA treatment. At this time maximal EA and VCA protein synthesis can be observed (19). Cell extracts were furthermore analyzed by electrophoretic immunoblotting. The results (Fig. 2A) demonstrate that some cell lines (Raji, BL74, QUIMR-WIL, P3HR-1) exhibit increased synthesis of the LMP or the 49,000 mol wt protein (M-ABA), while in 895-8 and Jijoye cells this
SHORT COMMUNICATIONS
164
effect could not be observed. Since in Raji cells the induction of the 60,000-Da membrane protein by TPA was very clear, the synthesis of this protein in Raji cells was further characterized. Freshly subcultured cells were harvested after addition of TPA, n-butyrate, or a combination of TPA and n-butyrate at the hours indicated in Fig. 4 and were analyzed in parallel either by electrophoretic immunoblotting or by indirect immunofluorescence. Freshly subcultered cells without TPA or n-butyrate treatment served as a control. The results of one out of three independent experiments are shown (Fig. 4). The synthesis of this protein was induced to a reasonable extent at 7 hr after addition of fresh medium with 10% fetal calf serum alone. On the other hand, the results again clearly demonstrate the increased expression of the 60,000 membrane protein in the presence of TPA or n-butyrate, reaching a maximum at 24 hr. These results are confirmed by data obtained from immunofluorescence assays. The amount of LMP in immunofluorescence-positive cells as judged by a more brilliant staining as well as the number of LMPpositive cells is increased by the addition of TPA or nbutyrate (data not shown). Treatment of Raji cells with n-butyrate, for example, increased the number of LMPpositive cells from approximately 3% to 50-609~ (data not shown). When a combination of TPA and n-butyrate was applied to Raji cells no synergistic effect was observed and the expression of LMP was similar to the amount found with TPA alone (Fig. 4). These findings are confirmed by .data obtained from immunofluorescence assays. In addition to LMP, the induction of the
0
0
7 2G. 48 72
7 2L L8 72 0
d
7 2L L8 72
7 2L
L8 72
FIG. 4. Protein analysrs of Raji cells cultivated with medium and 10% fresh fetal calf serum (A) and additional TPA (6). n-butyrate (C). or a combination of TPA and n-butyrate (D) using the technique and rabbit antiserum as described in Fig. 2. Cells were harvested at in drcated hours after subculturing.
EA complex in these TPA, n-butyrate, or TPA and nbutyrate treated Raji cells was assayed in parallel using a human anti-EA positive serum. This protein complex was synthesized to a maximal amount at 72 hr after addition of both drugs as estimated by immunofluorescence and electrophoretic immunoblotting and exhibited a different kinetic to LMP (data not shown). Seventy-two hours after a combined treatment with TPA and n-butyrate, more than 60% EA immunofluorescence-positive cells were determined (data not shown). These data are in agreement with previous reports (19, 20). No enhanced synthesis of EBNA 1 and EBNA 2 proteins could be observed after TPA, n-butyrate, or TPA and n-butyrate application (data not shown). The most surprising observation in these experiments is that LMP is expressed only in a distinct percentage of an EBV latently infected cell population and that expression of the 60,000 mol wt LMP can be enhanced in Raji cells by addition of medium with 10% fresh serum, TPA, or n-butyrate. From these data we suggest that the 60,000 LMP is synthesized during a defined period of the cell cycle, or the presence of this protein is not necessarily a prerequisite for the maintenance of the immortalized state of a cell. The latter consideration would link this 60,000 LMP to a class of viral proteins associated with the EA complex of EBV. The induction of this membrane protein by TPA or nbutyrate would favor this idea. The. kinetics of LMP induction in Raji cells by inductors and the induction by serum ~itself, however, argue.against the possibility that the LMP belongs to the EA complex. Furthermore, LMP is expressed in latently infected Raji and lijoye cells to a much higher degree (3-15%) than the proteins of the EA-complex (approximately O.l%), as judged by immunofluorescence (data not shown). In addition, in contrast to the EA complex, LM,P expression is not enhanced in a synergistic manner by a combination of TPA and n-butyrate. These~data indicate that LMP is not a member of the EA complex. At the present time one cannot rule out the possibility that the 60,000 mol wt LMP is one of the first proteins of the EA protein cascade or a prerequisite for the induction of the lytic or abortive cycle of the EBV. Preliminary experiments showed no evidence for a cell cycle-dependent expression of the 60,000 mol wt protein (R. Berger and W. MOller, unpublished results). At the present time nothing is known about the mechanism of the 60,000Da membrane protein induction by TPA, n-butyrate, or serum alone. Our results on the enhanced expression of LMP by TPA or n-butyrate are in some contrast to very recent findings (73). These differences might be due to different cell lines, different methods, and the combined treatment of TPA and n-butyrate, since in this report
SHORT COMMUNICATIONS
B95-8 immortalized ceils treated with TPA and n-butyrate for 24 hr were labeled metabolically with [2”S]methionine. Furthermore, these cells were then analyzed by immunoprecipitation using an antipeptide serum against the N-terminal part of LMP. No 60,000 mol wt LMP could be identified under these conditions. The fun.ction of the 60,000 and 49,000 proteins are still unknown. From previous studies evidence suggests localization in the cell membrane (72, 27). But so far, there is no strong proof that this 60,000 LMP is the lymphocyte detected membrane antigen (LYDMA) and itself the target for cytotoxic T cells, although recently some evidence for this idea was provided (22). DNA sequence homology studies did not reveal any significant homology to any known protein sequence, although a minor homology to the 7-amino acid Glu-Arg-Gly-Pro-Pro-Gly-Pro stretch in the carboxy-terminal part of the protein to the sequence of collagen could be observed. Furthermore, data were presented for the transforming capacity of the 60,000 LMP gene (23). In addition, this protein could be identified in virus particles (72). It remains to be elucidated which functional role this membrane protein has during the viral life cycle and for the immortalization of a B lymphocyte. ACKNOWLEDGMENTS We are grateful to T. .I. Koerner for his supply of the expression vector pATH. These studies were supported by the Deutsche Forschungsgameinschaft (SFB 31 -Medizinische Virologie: Tumorentstehung und -Entwicklung)‘and by the Bundesministerium fiir Forschung and Technologie (ECT 3llA).
REFERENCES 7. EPSTEIN, M. A., and ACHONG, B. G. (Eds.), “The Epstein-Barr Virus.” Springer-Varlag, Berlin, 1979. 2. HENLE, W., DIEHL, V., KO~IN, G., ZUR HAUSEN, H., and HENLE, G., Science 157, 1064-1065 (1967). 3. REEDMAN, B., and KLEIN, G., Inr. J. Cancer 11, 599-620 (1973). 4. SUMMERS, W. P., GROGAN, W. A., SHEDD, D., ROBERT, M., LIU, D-R., and MILLER, G., Proc. Nat/. Acad. Sci. USA 79, 5688-
5692 (1982). 5. HWNESSY, K., and KIEFF,E., Proc. Natl. Acad. Sci. USA 80, 56655669 (1983).
‘I65
6. HENNESSY,K., FENNEWALD,S., HUMMEL, M.. COLE, T., and KIEFF, E., Proc. Natl. Acad. Sci. USA 81, 7207-7211 (1984). 7. HENNESSY,K., FENNNVALD, S., and KIEFF, E., Proc. Natl. Acad.
Sci. USA 82,5944-5948
(1985).
8. HUDSON, G. S., FARRELL, P. J., and BARR&L, 5. G., /. l&o/. 53,
528-535 (1985). 9. MUELLER-LANTZSCH,N., LENOIR, G. M., SAUTER, M., TAKAKI, K., B~CHET, J. -M., KUKLIK-ROES,C., WuNDsstIcr-r, D., and BORNKAMM, G. W., E/‘&30/. 4, 1805-1811 (1985). 10. HENNESSY,K., WANG, F., WOODLAND:BUSHMAN, E., and KIEFF, E.,
Proc. Natl. Acad. Sci. USA 83, 5693-5697 (1986). 11. KALLIN, B., DILLNER,J., EIMBERG,J., EHLIN-HENRIKSSON,B,, ROS&N, A., HENLE, W., HENLE, G., and KLEIN, G., fioc. Nati. Acad. Sci. USA 83, 1499-l 503 (1986). 12. MANN, K. P., STAUNTON, D., and THORLEY-LAWSON,D. A., J. Viral. 55,7 1O-720 (1985). 13. MODROW,S., and WOLF, H., Proc. Nat/. Acad. Sci. USA 83,57035707 (1986). 14. OREN, M., Biochim. Biophys. Acfa 623, 67-78 (1985). 15. BAEA, R., BANKIER,A. T., BIGGIN, M. D., DEINI~~GER,P. L., FARR~L, P. J., GIBSON, R. J., HATFULL, G.; HUOSON, G. S.. SATCHWELL, S. C., %GUIN, C., TUFFNELL, P. S., and BARRELL, B. G., Narure (London) 310,207-211 (1984). 16. ZUR HAUSEN, H., O’NEILL, F. J., FREESE,U. K., and HECKER, E.,
Nature (London) 272,373-375 X1978). 17. ZUR HAUSEN,‘H., BORNKAMM,G. W., SCH~MIDT,I?., and HECKER,E., Prac. NatI. Acad. Sci. USA 76, 782-765 (1879). 18. MUELLER-L&~~STZSCH, N., YAMAMOTO, N., and ZUR HAUSEN, H., I/iroloby 97, 378-387 (1979). 19. MUELLER-LAFITZSCH,N., GEORG,B., YAMAMOTO, and ZUR HAUSEFJ, H., Virology 102, 4Oi-411 (1980). 20. ITO, Y., QW~NISHI, M., HARAYMA,T., ~~~‘TA~BAYAsHI, S., Cancer Lett, 12,175-180
(1981).
21. LIEBOWITZ,D., WANG, D., and KIEFF, E., 1. l&o/. 58, 233-237 (1986). 22. SULITZEANU,D., SZIGETI, R., KLEIN, G., HENNESSY,K., and KIEFF, E., 1. Viral. 58, 230-232 (1986). (1984). 23. WANG, D., LIEBOWIR, D.; and KIEFF, E., @l/43,831-840 24. TO~BIN, H., STAEHELIN.T., and GORDON% J., Pruc. Nat/. Acad. Sci. USA 76, 43&O-4354,(1 979). 25. LOWRY, 0. H., RO~EBR~WGH!N. J., FARR, A. L., and RANDALL, R. J., 1. Biol. Chem, 193, 265-275 (1651). 26. LAEMMLI, U. K., Nature (London) 277, 680-686 (1970). 27. SPINDLER, K. R., ROUSER,D. S. E., and BERK, A. J., J. Viral. 49, 132-141 (1984). 28. POLACK,A., HARTL, G., ZIMCTER,U., FREE% U. K., LAux, G., TAKAKI, K., HOLM, B,. GISSMANN, L., and BORNKAMM, G. W., Gene 27,
279-288 (1984). 29. WALTER, G., SCWEIDTMANN,K, -H., CAR&ONE,A., LAUDANO, A. P.. and DOOLITTLE, R. F., Proc. Nat/. Acad. Sci. USA 77, 51975200 (1 980).