- Email: [email protected]
The SV40 large T-p53 complex: Evidence for the presence of two immunologically distinct forms of p53
147,206-209
VIROLOGY
(1985)
The SV40 Large T-p53 Complex: Evidence for the Presence Immunologically Distinct Forms of ~53 Jo Divtiicm
of
MILNER’
AND
Virology, Department Tennis Court Road, Received
Am.1
JOHN
of Two
GAMBLE
<$ Pathology, University of Cambri&e, Cambridge, United Kingdom
2, 1985; accepted
Augwt
1, I.%%
The transforming protein of SV40 is the large T antigen. Large T binds a cellular protein, ~53, which is potentially oncogenic by virtue of its functional involvement in the control of cell proliferation. This raises the possibility that p53 may mediate, in part, the transforming function of SV40 large T. Two immunologically distinct forms of p53 have been identified in normal cells: the forms are cell-cycle dependent, one being restricted to nondividing cells (~53.Go) and the second to dividing cells (p53-G+). We have now dissociated and probed the multimeric complex of SV40 large T-p53 for the presence of immunologically distinct forms of ~53. Here we present evidence for the presence of p53-Go and p53-G+ complexed with SV40 large T. 0 1985 Academic press, h.
In normal cells the protein p53 is functionally implicated in the regulation of cell proliferation (1-h) and its expression is stringently controlled at the levels of gene transcription (I), mRNA abundance (4, 5), and protein turnover ($6). When this controlled expression of p53 breaks down, p53 may become oncogenic: indeed, the oncogenie potential of p53 has now been established by transfection of primary cell cultures with expression vectors for p53 (79). In cells transformed by simian virus 40 (SV40) the transforming viral protein, large T, binds (1~11) and stabilises p53 (6) in a multimeric complex (12): thus p53 may mediate, in part, the transforming effects of SV40 large T. In normal cells (primary cultures of lymphocytes) two cell-cycle dependent forms of p53 have been discovered (13). One form, p53-Go, is expressed in unstimulated lymphocytes in Go and is replaced by the second form, p53-Gt, following mitogenic stimulation. These two forms may represent different functional states of ~53, cells being held in Go in the presence of p53-Go and only able to enter the division cycle when p53-Go is replaced by p53G+. Certainly the switch from p53-Go to ’ Author addressed. 0042-6822185 Cop>riyht All rights
to whom
requests
for
$3.00
Cl 19% by Academic Press, Inc of rrproduction in any form reserved
reprints
should
p53-Gt in lymphocytes occurs within 4 hr of mitogenic stimulation and appears to be tightly linked with mitogenic commitment (13). This switch in response to mitogenic stimulation is readily effected since the half-life of both p53-Go and p53-Gt is very short, 5-10 min (13). Interestingly, when stabilised in the SV40 large T-p53 complex the p53 protein appears to bear epitopes that are mutually exclusive on p53-Go and p53-G+. These epitopes are recognised by the anti-p53 monoclonal antibodies RA3.2C2 (14) and PAb248 (15), specific for p53-Go (13); and PAb421 (16) and PAb122 (l7), specific for p53-G+ (13). All these monoclonals immunoprecipitate the p53large T complex, implying that their respective epitopes are available, simultaneously, for paratope binding on the complexed ~53 protein. Alternatively, since the large T-p53 complex is multimeric, both p53-Go and p53-Gt may be present: thus the complex would bear the combined epitopes of each p53 form. In order to determine whether one or more p53 forms are bound to SV40 large T we have now dissociated and probed the multimeric large T-p53 complex for different forms of ~53, using the monoclonal antibodies that discriminate between p53-Go and p53-Gt. Fresh lysates of cells labeled with
be
206
SHORT
COMMUNICATIONS
[?S]methionine were prepared as described in Fig. 1 legend. The first step was to determine optimal conditions for the dissociation of immune complexes and reimmunoprecipitation of ~53. Since we aimed probe for p53-Go and p53-Gt it was necessary, in these preliminary experiments, to establish that each form of p53 could be reimmunoprecipitated after dissociation: for this purpose a line of 3T6 cells known to express the two ~53 forms was included. A range of conditions for dissociation were 3T6 1
2
s v 40 3
4
5
6
T a9
P 53
FIG. 1. Reimmunoprecipitation of p53 after dissociation of primary immune complexes from 3T6 and 3T3-SVA31E7 cells: the latter is a line of Balb/c-3T3 cells transformed with SV40 (Y. Ito). The cells were cultured in ‘75-cm2 flasks and labelled at 70% confluence with $Sjmethionine (1000 Ci mmol-‘, ‘70 PCi ml-‘) in methionine-free MEM for 2 hr. Immediately after labelling the cells were lysed in 2 ml lysis buffer (19) and 500 pl was taken for immunoprecipitation. The initial immunoprecipitations were with PAb421, except for track 2, for which PAb248 was used. The immune complexes were collected on Stczph$ococcus aureus and dissociated, as described in the text. After dilution of the dissociation buffer (see text) 500 ~1 was taken for reimmunoprecipitation with PAb421 (tracks 1 and 4), with PAb248 (track 2), or with PAb421 plus PAb419 (track 6). The supernatant from the immunoprecipitation for track 4 was reimmunoprecipitated with PAb419 for SV40 large T (track 5). Track 3 represents the S. aureus pellet after dissociation of primary immune complexes. Tracks 1 and 2 are for 3T6 cells and are from a different gel to tracks 3-6, which are for SV40-transformed cells. [36S&abelled p53 and large T in Figs. 1 to 3 were visualised by fluorography after electrophoresis on a 15% polyacrylamide gel.
207
tested, varying time, temperature, and buffer systems. The conditions giving optimal results were as follows. Immune complexes collected on Staphylococcus aureus Cowan I (18) were resuspended in dissociation buffer (40 ~1, pH 8.0) containing 0.25 M LiCl; 0.05 M Tris; 0.5% mercaptoethanol; 0.5% NP-40; and 1% SDS. To the resuspended pellet 2 ~1 dithiothreitol(1 M) was added and the dissociation mixture was incubated at 37” for 15 min. The dissociation mixture was then diluted with 1 ml ice-cold buffer containing 0.14 M NaCl; 10 mM Tris, and 0.5% NP-40, at pH 8.0. After microfuging, to pellet the S. aureus from the original immunoprecipitation step, the supernatant was reprobed for ~53 as described in the figure legends. The results presented in Fig. 1 demonstrate that the monoclonal antibodies PAb248 and PAb421 reimmunoprecipitate ~53 from 3T6 cells (tracks 1 and 2); similarly PAb421 and PAb419 (an anti-SV40 large T monoclonal antibody, Ref. 16) reimmunoprecipitated ~53 and SV40 large T from SV40-transformed cells (track 6). The next step was to demonstrate, by cascade immunoprecipitation, that the conditions of dissociation of the immune complex also dissociated ~53 from SV40 large T. Here ~53 was first immunoprecipitated with PAb421 (Fig. 1, track 4) and the remaining supernatant was immunoprecipitated with PAb419 to bring down SV40 large T (Fig. 1, track 5). The results indicate complete dissociation of ~53 from SV40 large T antigen. Using the above procedures we next examined the dissociated ~53 component for the presence of p53-Go and p53-G+-, using the monoclonal antibodies PAb248 and PAb421, respectively. Since p53-Go and p53-G+ comigrate during polyacrylamide gel electrophoresis (13) we aimed to deplete, from solution, ~53 bearing the PAb421 epitope. The depleted supernatant was then probed with PAb248 to immunoprecipitate any ~53 not bound by PAb421. The results presented in Fig. 2 are for the cascade sequence: PAb421-PAb421PAb421-PAb248 (tracks l-4). The reciprocal cascade was also adopted, i.e., preclearance with PAb248 followed by
208
SHORT 1
2
3
4
COMMUNICATIONS
5
P 53
FIG. 2. Cascade reimmunoprecipitation of p.53 and SV40 large T after dissociation of primary immune complexes precipitated by PAb421. SV40-transformed cells were labelled, immunoprecipitated with PAb421, and the immune complex was dissociated as detailed in Fig. 1 legend. After dissociation p53 was repeatedly immunoprecipitated from solution using PAb421 (tracks 1 to 3). Subsequently the supernatant from the third immunoprecipitation step was reprobed with PAb248 to detect any p53 not bound by PAb421 (track 4). The final supernatant was immunoprecipitated for SV40 large T using PAb419 (track 5). At each change of monoclonal antibody the supernatant was precleared of any residual antibody remaining from the preceding step, using S uureus.
proved correct, as shown in Fig. 3. In Figs. 2 and 3 the fifth track shows SV40 large T immunoprecipitated with PAb419 at the end of each cascade series. It should be noted that, in Fig. 2 this represents the large T component that was originally complexed with p53 and with which it was coprecipitated. In Fig. 3, on the other hand, the large T band represents the component not complexed with p53 and, therefore, was not coprecipitated via PAb421 or PAb248. Interestingly, SV40 large T invariably appeared as a doublet after dissociation from the large T-p53 complex (Fig. 1, tracks 5 and 6; Fig. 2, track 5). This is under further investigation. The protein p53 has been shown to exhibit multiple, immunologically distinct forms (13, and J. Milner and Cook, in preparation), raising the possibility that p53 is multifunctional. In normal cells two forms of ~53, p53-Go and p53-Gt are under stringent control of expression and are cell-
TAg
p 53
PAb421: in both cases a strong band of ~53, visualised by autoradiography, was immunoprecipitated by the second monoclonal antibody (see for example, track 4 compared with track 3, Fig. 2). Similar results were obtained when RA3.2C2 was substituted for PAb248 and PAb122 substituted for PAb421. These results indicate that two forms of ~53, distinguished by those epitopes specific for p53-Go and p53G+, had been dissociated from the large T-p53 complex. For the undissociated complex we predicted that p53-Go and p53Gt (a) would be coprecipitated by either PAb421 or PAb248; (b) would comigrate and appear as a single protein band following electrophoresis; and (c) no additional p53 would be immunoprecipitable by PAb248 after depletion with PAb421 (and vice versa). Each of these predictions
-
-
FIG. 3. Cascade immunoprecipitation of p53 and SV40 large T from SV40-transformed cell lysate. The SV40 large T-p53 complex, undissociated, was subjected to a cascade immunoprecipitation identical to that described in Fig. 2 legend. Tracks 1-3 = repeated immunoprecipitation with PAb421; track 4 = subsequent immunoprecipitation with PAb248; track 5 = final immunoprecipitation with PAb419 to detect that component of large T not present in the large Tp53 complex.
SHORT
COMMUNICATIONS
cycle dependent (13). Using monoclonal antibodies that discriminate between p53Go and p53-G+ we have now probed the dissociated SV40 large T-p53 complex for two forms of ~53. Our results indicate that the SV40 large T antigen complexes with two forms of p53 distinguished by the same epitopes that distinguish p53-Go from p53G+. The nature of this complex is now under detailed investigation in order to assess the contribution of the cellular protein ~53, in terms of its different functional forms, to the oncogenic role of SV40 large T anACKNOWLEDGMENTS We thank Dr. Ed Harlow for hybridoma cells producing the monoclonal antibody PAb421, and Dr. D. Lane for access to PAb248. J. Gamble is an SERC Case Award Student (in cooperation with G. D. Searle, High Wycombe, England: Dr. Paul Bosely). This work was supported by a project grant (to J.M.) from the Cancer Research Campaign. REFERENCES 1. MILNER, Jo, and MILNER, Su., Virology 112, 785788 (1981). 2. MERCER, W. E., NELSON, D., DELEO, A. B., OLD, L. J., and BASERGA, R., Proc. NatL Acad. Sci. USA 79,6309-6312 (1982).
3. MERCER,
209
W. E., AVIGNOLO, C., and BASERGA, R., Mol. Cell. BioL 4,276-281 (1984). 4. REICH, N. C., and LEVINE, A. J., Nature (London) 308,199-201 (1984). REICH, N. C., OREN, M., and LEVINE, A. J., Mol. ‘. Cell. BioL 3,2143-2150 (1983). 6, OREN, M., MALTZMAN, W., and LEVINE, A. J., MoL Cell. BioL 1, 101-110 (1981). 7. JENKINS, J., RUDGE, K., and CURRIE, G. A., Nature (London) 321,651-654 (1984). 8. ELIYAHU, D., RAZ, A., GRUSS, P., GIVOL, D., and OREN, M., Nature &mdm) 321,646-649 (1984). 9. PARADA, L. F., LAND, H., WEINBERG, R. A., WOLF, D., and ROTTER, V., Nature (km&m) 312,649651 (1984). LANE, D. P., and CRAWFORD, L. V., Nature (London) 278,261-263 (1979). LINZER, D. I. H., and LEVINE, A. J., Cell 17,43-52 (1979). 12. MCCORMICK, F., and HARLOW, E., J. ViroL 34,213224 (1980). 13. MILNER, J., Nature &w&n) 310,143-145 (1984). R., and BAL14. ROTTER, V., WITTE, 0. N., COFFMAN, TIMORE, D., J. ViroL 36,547-555 (1980). 15. YEWDELL, J., GANNON, J., and LANE, D. P., EMBO J. (submitted). 16. HARLOW, E., CRAWFORD, L. V., PIM, D., and WILLIAMSON, N. M., J. ViroL 34. 752-763 (1981). 17. GURNEY, E. G., HARRISON, R. O., and FENNO, J., J. ViroL 34,752-753 (1980). 18. KESSLER, S. W., J. ImmunoL 115,1617-1624 (1975). 19. MILNER, J., and MCCORMICK, F., Cell BioL Int. Rep. 4,663-667 (1980).
VIROLOGY
(1985)
The SV40 Large T-p53 Complex: Evidence for the Presence Immunologically Distinct Forms of ~53 Jo Divtiicm
of
MILNER’
AND
Virology, Department Tennis Court Road, Received
Am.1
JOHN
of Two
GAMBLE
<$ Pathology, University of Cambri&e, Cambridge, United Kingdom
2, 1985; accepted
Augwt
1, I.%%
The transforming protein of SV40 is the large T antigen. Large T binds a cellular protein, ~53, which is potentially oncogenic by virtue of its functional involvement in the control of cell proliferation. This raises the possibility that p53 may mediate, in part, the transforming function of SV40 large T. Two immunologically distinct forms of p53 have been identified in normal cells: the forms are cell-cycle dependent, one being restricted to nondividing cells (~53.Go) and the second to dividing cells (p53-G+). We have now dissociated and probed the multimeric complex of SV40 large T-p53 for the presence of immunologically distinct forms of ~53. Here we present evidence for the presence of p53-Go and p53-G+ complexed with SV40 large T. 0 1985 Academic press, h.
In normal cells the protein p53 is functionally implicated in the regulation of cell proliferation (1-h) and its expression is stringently controlled at the levels of gene transcription (I), mRNA abundance (4, 5), and protein turnover ($6). When this controlled expression of p53 breaks down, p53 may become oncogenic: indeed, the oncogenie potential of p53 has now been established by transfection of primary cell cultures with expression vectors for p53 (79). In cells transformed by simian virus 40 (SV40) the transforming viral protein, large T, binds (1~11) and stabilises p53 (6) in a multimeric complex (12): thus p53 may mediate, in part, the transforming effects of SV40 large T. In normal cells (primary cultures of lymphocytes) two cell-cycle dependent forms of p53 have been discovered (13). One form, p53-Go, is expressed in unstimulated lymphocytes in Go and is replaced by the second form, p53-Gt, following mitogenic stimulation. These two forms may represent different functional states of ~53, cells being held in Go in the presence of p53-Go and only able to enter the division cycle when p53-Go is replaced by p53G+. Certainly the switch from p53-Go to ’ Author addressed. 0042-6822185 Cop>riyht All rights
to whom
requests
for
$3.00
Cl 19% by Academic Press, Inc of rrproduction in any form reserved
reprints
should
p53-Gt in lymphocytes occurs within 4 hr of mitogenic stimulation and appears to be tightly linked with mitogenic commitment (13). This switch in response to mitogenic stimulation is readily effected since the half-life of both p53-Go and p53-Gt is very short, 5-10 min (13). Interestingly, when stabilised in the SV40 large T-p53 complex the p53 protein appears to bear epitopes that are mutually exclusive on p53-Go and p53-G+. These epitopes are recognised by the anti-p53 monoclonal antibodies RA3.2C2 (14) and PAb248 (15), specific for p53-Go (13); and PAb421 (16) and PAb122 (l7), specific for p53-G+ (13). All these monoclonals immunoprecipitate the p53large T complex, implying that their respective epitopes are available, simultaneously, for paratope binding on the complexed ~53 protein. Alternatively, since the large T-p53 complex is multimeric, both p53-Go and p53-Gt may be present: thus the complex would bear the combined epitopes of each p53 form. In order to determine whether one or more p53 forms are bound to SV40 large T we have now dissociated and probed the multimeric large T-p53 complex for different forms of ~53, using the monoclonal antibodies that discriminate between p53-Go and p53-Gt. Fresh lysates of cells labeled with
be
206
SHORT
COMMUNICATIONS
[?S]methionine were prepared as described in Fig. 1 legend. The first step was to determine optimal conditions for the dissociation of immune complexes and reimmunoprecipitation of ~53. Since we aimed probe for p53-Go and p53-Gt it was necessary, in these preliminary experiments, to establish that each form of p53 could be reimmunoprecipitated after dissociation: for this purpose a line of 3T6 cells known to express the two ~53 forms was included. A range of conditions for dissociation were 3T6 1
2
s v 40 3
4
5
6
T a9
P 53
FIG. 1. Reimmunoprecipitation of p53 after dissociation of primary immune complexes from 3T6 and 3T3-SVA31E7 cells: the latter is a line of Balb/c-3T3 cells transformed with SV40 (Y. Ito). The cells were cultured in ‘75-cm2 flasks and labelled at 70% confluence with $Sjmethionine (1000 Ci mmol-‘, ‘70 PCi ml-‘) in methionine-free MEM for 2 hr. Immediately after labelling the cells were lysed in 2 ml lysis buffer (19) and 500 pl was taken for immunoprecipitation. The initial immunoprecipitations were with PAb421, except for track 2, for which PAb248 was used. The immune complexes were collected on Stczph$ococcus aureus and dissociated, as described in the text. After dilution of the dissociation buffer (see text) 500 ~1 was taken for reimmunoprecipitation with PAb421 (tracks 1 and 4), with PAb248 (track 2), or with PAb421 plus PAb419 (track 6). The supernatant from the immunoprecipitation for track 4 was reimmunoprecipitated with PAb419 for SV40 large T (track 5). Track 3 represents the S. aureus pellet after dissociation of primary immune complexes. Tracks 1 and 2 are for 3T6 cells and are from a different gel to tracks 3-6, which are for SV40-transformed cells. [36S&abelled p53 and large T in Figs. 1 to 3 were visualised by fluorography after electrophoresis on a 15% polyacrylamide gel.
207
tested, varying time, temperature, and buffer systems. The conditions giving optimal results were as follows. Immune complexes collected on Staphylococcus aureus Cowan I (18) were resuspended in dissociation buffer (40 ~1, pH 8.0) containing 0.25 M LiCl; 0.05 M Tris; 0.5% mercaptoethanol; 0.5% NP-40; and 1% SDS. To the resuspended pellet 2 ~1 dithiothreitol(1 M) was added and the dissociation mixture was incubated at 37” for 15 min. The dissociation mixture was then diluted with 1 ml ice-cold buffer containing 0.14 M NaCl; 10 mM Tris, and 0.5% NP-40, at pH 8.0. After microfuging, to pellet the S. aureus from the original immunoprecipitation step, the supernatant was reprobed for ~53 as described in the figure legends. The results presented in Fig. 1 demonstrate that the monoclonal antibodies PAb248 and PAb421 reimmunoprecipitate ~53 from 3T6 cells (tracks 1 and 2); similarly PAb421 and PAb419 (an anti-SV40 large T monoclonal antibody, Ref. 16) reimmunoprecipitated ~53 and SV40 large T from SV40-transformed cells (track 6). The next step was to demonstrate, by cascade immunoprecipitation, that the conditions of dissociation of the immune complex also dissociated ~53 from SV40 large T. Here ~53 was first immunoprecipitated with PAb421 (Fig. 1, track 4) and the remaining supernatant was immunoprecipitated with PAb419 to bring down SV40 large T (Fig. 1, track 5). The results indicate complete dissociation of ~53 from SV40 large T antigen. Using the above procedures we next examined the dissociated ~53 component for the presence of p53-Go and p53-G+-, using the monoclonal antibodies PAb248 and PAb421, respectively. Since p53-Go and p53-G+ comigrate during polyacrylamide gel electrophoresis (13) we aimed to deplete, from solution, ~53 bearing the PAb421 epitope. The depleted supernatant was then probed with PAb248 to immunoprecipitate any ~53 not bound by PAb421. The results presented in Fig. 2 are for the cascade sequence: PAb421-PAb421PAb421-PAb248 (tracks l-4). The reciprocal cascade was also adopted, i.e., preclearance with PAb248 followed by
208
SHORT 1
2
3
4
COMMUNICATIONS
5
P 53
FIG. 2. Cascade reimmunoprecipitation of p.53 and SV40 large T after dissociation of primary immune complexes precipitated by PAb421. SV40-transformed cells were labelled, immunoprecipitated with PAb421, and the immune complex was dissociated as detailed in Fig. 1 legend. After dissociation p53 was repeatedly immunoprecipitated from solution using PAb421 (tracks 1 to 3). Subsequently the supernatant from the third immunoprecipitation step was reprobed with PAb248 to detect any p53 not bound by PAb421 (track 4). The final supernatant was immunoprecipitated for SV40 large T using PAb419 (track 5). At each change of monoclonal antibody the supernatant was precleared of any residual antibody remaining from the preceding step, using S uureus.
proved correct, as shown in Fig. 3. In Figs. 2 and 3 the fifth track shows SV40 large T immunoprecipitated with PAb419 at the end of each cascade series. It should be noted that, in Fig. 2 this represents the large T component that was originally complexed with p53 and with which it was coprecipitated. In Fig. 3, on the other hand, the large T band represents the component not complexed with p53 and, therefore, was not coprecipitated via PAb421 or PAb248. Interestingly, SV40 large T invariably appeared as a doublet after dissociation from the large T-p53 complex (Fig. 1, tracks 5 and 6; Fig. 2, track 5). This is under further investigation. The protein p53 has been shown to exhibit multiple, immunologically distinct forms (13, and J. Milner and Cook, in preparation), raising the possibility that p53 is multifunctional. In normal cells two forms of ~53, p53-Go and p53-Gt are under stringent control of expression and are cell-
TAg
p 53
PAb421: in both cases a strong band of ~53, visualised by autoradiography, was immunoprecipitated by the second monoclonal antibody (see for example, track 4 compared with track 3, Fig. 2). Similar results were obtained when RA3.2C2 was substituted for PAb248 and PAb122 substituted for PAb421. These results indicate that two forms of ~53, distinguished by those epitopes specific for p53-Go and p53G+, had been dissociated from the large T-p53 complex. For the undissociated complex we predicted that p53-Go and p53Gt (a) would be coprecipitated by either PAb421 or PAb248; (b) would comigrate and appear as a single protein band following electrophoresis; and (c) no additional p53 would be immunoprecipitable by PAb248 after depletion with PAb421 (and vice versa). Each of these predictions
-
-
FIG. 3. Cascade immunoprecipitation of p53 and SV40 large T from SV40-transformed cell lysate. The SV40 large T-p53 complex, undissociated, was subjected to a cascade immunoprecipitation identical to that described in Fig. 2 legend. Tracks 1-3 = repeated immunoprecipitation with PAb421; track 4 = subsequent immunoprecipitation with PAb248; track 5 = final immunoprecipitation with PAb419 to detect that component of large T not present in the large Tp53 complex.
SHORT
COMMUNICATIONS
cycle dependent (13). Using monoclonal antibodies that discriminate between p53Go and p53-G+ we have now probed the dissociated SV40 large T-p53 complex for two forms of ~53. Our results indicate that the SV40 large T antigen complexes with two forms of p53 distinguished by the same epitopes that distinguish p53-Go from p53G+. The nature of this complex is now under detailed investigation in order to assess the contribution of the cellular protein ~53, in terms of its different functional forms, to the oncogenic role of SV40 large T anACKNOWLEDGMENTS We thank Dr. Ed Harlow for hybridoma cells producing the monoclonal antibody PAb421, and Dr. D. Lane for access to PAb248. J. Gamble is an SERC Case Award Student (in cooperation with G. D. Searle, High Wycombe, England: Dr. Paul Bosely). This work was supported by a project grant (to J.M.) from the Cancer Research Campaign. REFERENCES 1. MILNER, Jo, and MILNER, Su., Virology 112, 785788 (1981). 2. MERCER, W. E., NELSON, D., DELEO, A. B., OLD, L. J., and BASERGA, R., Proc. NatL Acad. Sci. USA 79,6309-6312 (1982).
3. MERCER,
209
W. E., AVIGNOLO, C., and BASERGA, R., Mol. Cell. BioL 4,276-281 (1984). 4. REICH, N. C., and LEVINE, A. J., Nature (London) 308,199-201 (1984). REICH, N. C., OREN, M., and LEVINE, A. J., Mol. ‘. Cell. BioL 3,2143-2150 (1983). 6, OREN, M., MALTZMAN, W., and LEVINE, A. J., MoL Cell. BioL 1, 101-110 (1981). 7. JENKINS, J., RUDGE, K., and CURRIE, G. A., Nature (London) 321,651-654 (1984). 8. ELIYAHU, D., RAZ, A., GRUSS, P., GIVOL, D., and OREN, M., Nature &mdm) 321,646-649 (1984). 9. PARADA, L. F., LAND, H., WEINBERG, R. A., WOLF, D., and ROTTER, V., Nature (km&m) 312,649651 (1984). LANE, D. P., and CRAWFORD, L. V., Nature (London) 278,261-263 (1979). LINZER, D. I. H., and LEVINE, A. J., Cell 17,43-52 (1979). 12. MCCORMICK, F., and HARLOW, E., J. ViroL 34,213224 (1980). 13. MILNER, J., Nature &w&n) 310,143-145 (1984). R., and BAL14. ROTTER, V., WITTE, 0. N., COFFMAN, TIMORE, D., J. ViroL 36,547-555 (1980). 15. YEWDELL, J., GANNON, J., and LANE, D. P., EMBO J. (submitted). 16. HARLOW, E., CRAWFORD, L. V., PIM, D., and WILLIAMSON, N. M., J. ViroL 34. 752-763 (1981). 17. GURNEY, E. G., HARRISON, R. O., and FENNO, J., J. ViroL 34,752-753 (1980). 18. KESSLER, S. W., J. ImmunoL 115,1617-1624 (1975). 19. MILNER, J., and MCCORMICK, F., Cell BioL Int. Rep. 4,663-667 (1980).