VIROLOGY
188,
369-372 (1992)
Glucocorticoid-Dependent Transformation by Human Papillomavirus 16 E7 Coding and 3’ Noncoding Sequences ALAN
PATER, NARASIMHASWAMYS. BELAGULI,
Basic Medical Sciences,
Faculty of Medicine, Received
Memorial
A. R. GARDNER, ALKA MITHAL,
HUMPHREY University
of Newfoundland,
Type
AND MARY
St. John’s, Newfoundland,
M. PATER’
Canada A 1B 3V6
October 4, 199 1; accepted January 29, 1992
The establishment of transformation of primary rodent cells by human papillomavirus (HPV) type 16 DNA requires glucocorticoid hormones (Pater et a/., Nature 335, 832-835, 1988). Here we provide evidence by mutational analysis that, in the context of the hormone-regulated HPV 16 promoter/enhancer, the only protein coding sequences of HPV 16 required are those of the E7 gene. Moreover, additional sequences adjacent to the 3’ end of E7 coding sequences are also essential for the establishment of the transformed phenotype. Splice donor sites, especially an E7 ORF 3 proximal one, are implicated for this k-acting function, since specific deletion mutations of these splice sites greatly or completely reduced the frequency of transformation and the level of E7 RNA. o 1992 Academic PESS, IK-.
in Fig. 1. Plasmids pT64, pA7, pA41, pH4, and pCK1 were constructed as previously described (15, 16). Plasmid pA32 was generated by digestion, with 1.8 kU/ml Sl nuclease for 30 min at 27”, of partially IVarldigested HPV 16 DNA. Bal 31 was used to generate deletions at the Hindll site for pH 17 and at the Kpnl site for pHKB7, pHKB5, and pHKB1; the flush ending of the Narl site and /Varl to /Vcol fragment was used for pA5 and pAM16, respectively. For pHGl6A, the large nt 4466 Srul to 863 Ncol fragment and the small nt 4138 Accl to 4466 Srul fragment were flush ended and ligated. All deletions were confirmed by sequencing. Plasmids pT64 and pCK1, which have deletions removing the splice donor site for E6* (8) and important E6 ORF sequences, were capable of transformation (Fig. 1). These results indicate that neither E6 nor E6* are required. Plasmid pH4, containing a translation termination linker at position 683 which is in the E7 ORF, had no transforming ability, confirming the essential role of this gene in hormone-dependent establishment of transformation. Mutations within the 1947 nt of the El ORF demonstrated that the El gene product is not required, as follows. Plasmids pA7 and pA5 introduced small deletion mutations at 2 and 448 nt from the El ORF 5’ terminus, respectively, which disrupt the El ORF; plasmids pA32 and pH17 were deletion mutations of El ORF nt 413 to 460 and 1730 to 1947, respectively, from the 5’ terminus. All four of these constructs were fully transformation competent (Fig. 1). However, plasmid pHGl6A containing a 3275-bp deletion encompassing all of the El ORF as well as the rest of the early region E2, E4, and E5 ORFs failed to transform. We then generated smaller deletions to localize the E7 ORF 3’ region required for transformation. Plasmid
Squamous cell carcinoma of the uterine cervix is the second most common cancer of females in the world (I). Studies using the techniques of DNA hybridization and molecular cloning have shown the presence of human papillomavirus (HPV) types 16 and 18 in 70 to 80% of cervical tumors and tumor cell lines (2-4). Also, HPV types 16 and 18 RNA transcripts are detected in both cervical tumors and tumor cell lines (5-8). These RNA transcripts and expressed viral protein products are mostly from the E6-E7 region of viral DNA (7, 9), suggesting a role of these genes in the malignant conversion of cervical cells. In vitro transformation studies of rodent cells using the early region of HPV 16 or 18 expressed from a heterologous promoter have identified the transforming region as the E7 gene (10-12). Deletion mutagenesis studies have provided evidence for the essential role of this gene in the establishment of transformation (I 1, 12). Moreover, studies using HPV 16 early region sequences under the control of the hormone-inducible promoter of mouse mammary tumor virus (MMTV) have revealed the essential role of this gene in the maintenance of the transformed phenotype (13). In the present study, mutations in the context of the intact viral genome including its hormone-responsive promoter/enhancer are used. Results show that, while the E7 gene of this virus is required for the establishment of the transformed phenotype, sequences adjacent to the 3’ end of the E7 open reading frame (ORF) are also required. The various mutations in HPV 16 plasmids used to characterize essential sequences and results of transformation assays using 5 pg of HPV 16 and ras DNA with previously described procedures (14) are shown ’ To whom reprint requests should be addressed.
369
0042-6822/92
$3.00
CopyrIght 0 1992 by Academic Press, Inc All rights of reproduction in any form reserved
SHORT COMMUNICATIONS
370
Plrsmid
Transformation
HPV 16
23
pT64
20
pCK1
18
pH4
0
PA7
22
PAS
20
~A32
24
pHG16A
0
225-294
.E6
132-39p ; E6 663;
E7
667;. El 1313; 1278-1325;
El El
2965-4004; PA41
26
pH17
25
pAnI
0
pHKB7
1
pHKB5
2
pHKB1
2
E2, E4. ES
2595-3907; 863-1309;
El
a77,-893;
El
8794387; I
El
881-884;
El
FIG. 1. Transformation assays of BRK ceils with HPV 16 mutant constructs. El to E7 are early ORFs; Ll and L2 are late ORFs. The arrows and P indicate initiation sites for transcription; A’s indicate polyadenylation signals. Large deletions are indicated as black boxes; small deletions and insertions are indicated by vertical lines. Affected nucleotides and ORFs are indicated. The results are the average of three independent experiments using four plates per experiment. The results for the last three plasmids are from five experiments with four plates each, and the average number of colonies transformed by wild-type HPV 16 plasmid for these experiments was 19.
pA41, containing a 1019-bp deletion encompassing E2, E4, and E5 ORFs, was transformation competent. Also, plasmid pH 17, with a 1312-bp deletion encompassing the 3’end of the El ORF and the entire E2, E4, and E5 ORFs, was transformation competent. However, plasmid PAM 16 with a 546-bp deletion in the 5’ end of the El ORF failed to transform. Since the possibility of a trans-acting function of the El ORF had been eliminated, this region must have a c&-acting function. Thus, the data of Fig. 1 provide evidence that E7 is the only ORF required for HPV 16 enhancer/promoterand glucocorticoid hormone-dependent transformation of primary BRK cells. This is consistent with the work of others in which HPV coding sequences were expressed from a heterologous promoter and, accordingly, in the absence of hormones (12, 17). However, Fig. 1 further shows that 3’ noncoding sequences are also required. To further analyze the nature of the sequences at the 5’ end of the El sequences required for E7-mediated transformation, we generated smaller deletion mutations. These were at the nt 880 splice donor site, 3’ to the E7 ORF and present in E7 pre-mRNA, which is utilized in both cell lines derived from cervical lesions as well as HPV 16-immortalized keratinocytes (18, 19, our unpublished observations). In addition, all three cDNA
clones isolated from a tumorigenic rodent cell line transformed by HPV 16 DNA were the products of utilization of this splice donor site (20). The results of Fig. 1 for pHKB7, pHKB5, and pHKB1 show that, in the context of the homologous HPV enhancer/promoter-containing constructs, three nt 880 splice site-specific deletions as small as 4, 9, and 16 bp greatly reduced, although they did not completely eliminate, transformation. This is in contrast to the larger deletion mutations which repeatedly failed to transform (Fig. 1; pHGl6A and pAM16). Within E7 pre-mRNA there are two other sites at nt 226 and 1302, and both sequences are more similar to the consensus sequence than is the 880 site. The 1302 sequence had been shown by cDNA analysis to be functional in a cell line from a precancerous cervical lesion (18). In addition, the potential of 1302 site splicing events may be obscured by its position within the intron generated from the nt 880 site. Since utilization appears to be low for this site, the influence of splicing at this site may be less than that of the 880 site. Indeed, deletion of the 1302 site from HPV 16 DNA in pA32 (Fig. 1) failed to affect transformation. Similarly, deletion of the 226 site in pT64 and pCK1 (Fig. 1) had no effect on transformation. Moreover, the splice donor site at nt 226 does not substitute for the 880 site since deletions of the 880
371
SHORT COMMUNICATIONS
Experiment
SVP/E
Plasmid
1
2
pD7
102
79
pD7KB
0
0
---c-a---------
---------------e
881-884 '
------a--
of BRK cells with the intact and mutant early region of HPV 16 expressed from the SV40 enhancer/promoter. The FIG. 2. Transformation results are the number of colonies from a total of seven plates for experiment 1 and six plates for experiment 2. The expression vector sequences are indicated by the dashed lines and HPV sequences are as in Fig. 1. The SV40 promoter/enhancer is indicated @VP/E) by the arrow.
and 1302 sites in pHG16A and pAMl6
(Fig. 1) completely eliminated transformation. These results indicate that the 880 site has a strong dominant role in transformation and that, in the absence of this site, the splice sites at 226 and 1302 may function, resulting in a low level of transformation. Since the effect of 880 splice donor site mutations was so great, we also examined the effect of the splice acceptor sites for the 880 donor site. These are at nt 2708 and 3357. The removal of either the 3357 site in pA41 or both sites in pH 17 had no effect on transformation (Fig. 1). Thus, additional splice acceptor sites would appear to function to allow full transformation. However, the nt 880 splice donor site appears to be critically important since loss of this site severely reduced transformation, even in the presence of the two alternative donor sites at nucleotides 226 and 1302 (Fig. 1, pHKB5, pHKB7, and pHKBl), and completely abolished transformation in the presence of only the 226 donor site (Fig. 1, pAMl6 and pHGl6A). To examine the mechanism by which the splice donor site at nt 880 affects transformation, we probed with RNase protection assays, using Cos-1 cells which amplify SV40 origin region-containing sequences. Plasmid pD7KB, containing the same 4-nt 880 deletion as pHKB1, was generated by T4 DNA polymerase I incubation of Kpnl-linearized pD7 (2 I). First, transformation assays were done as before (14) except in the absence of dexamethasone. The pD7 plasmid gave transformed colonies while pD7KB gave no colonies (Fig. 2). Since pD7 and pD7KB lack the 226/409 splice donor/acceptor sites, the effect on transformation of deletion of the 880 plus 226/409 splice sites is the same as the effects in pA32 and pHGl6A of the deletion of the nt 880 plus 1302 splice donor sites. This reinforces the implication that both the 226 and the 1302 sites have some effect on E7 expression but that this effect on transformation can only be detected in the absence of the predominating effect of the 880 splice donor site. Next, the effect of 880 splice site mutation on E7 expression was analyzed by RNase protection (22) of 5 pg total cellular RNA, prepared as described by Chirgwin et a/. (23). Cos-1 cells were
transfected with 10 pg of plasmid 48 hr prior to harvesting. Probe for E7 was from Bluescript KS(+) containing the SV40 nt 5 190 Stul to HPV nt 720 Sspl fragment of pD7. The SV40 T antigen RNA level was an internal control; probe was from a similar construct for the SV40 3072-3225 Rsal fragment. Hybridization was with 1O5 cpm each of both probes for 14 hr at 56” and digestion was with 700 U/ml RNase Tl. E7 products were resolved on 5% polyacrylamide denaturing gels. The splice site mutation in pD7KB reduced the RNA level dramatically (Fig. 3). Thus, the strong &-acting effect of the functional 880 site, in transformation assays (Fig. l), is also observed for E7 oncogene expression. Since splicing is a post-transcriptional event, the 123456
E7-
FIG. 3. RNase protection assays. On the left, the larger type letters indicate E7 (E7) and T antrgen (T) probes. Smaller type letters rndicate the products of RNase protection, which are smaller for CaSki E7 RNA since it lacks SV40 sequences. Molecular weight marker sizes in bases are indicated on the right. Lane 1, E7 and T antigen probes; lane 2. CaSki; lane 3. pD7; lane 4, pD7KB; lane 5, Cos-1; lane 6. Escherichia co/i tRNA.
372
SHORT COMMUNICATIONS
influence of the 880 splice site for transformation is most likely due to post-transcriptional factors such as RNA stability and/or processing which influence the levels of RNA available for E7 product synthesis. Therefore, the integrity of the 880 splice donor site is apparently specifically important to maintain sufficient levels of E7 RNA to allow transformation. ACKNOWLEDGMENTS We thank G. Hughes and D. Hyslop for technical assistance, S. Atkins for typing the manuscript, and H. zur Hausen and L. Gissmann for HPV 16 plasmid. This work was supported in part by grants from the MRC and NCI of Canada.
REFERENCES PARKIN, D. M., LAURA, E., and MUIR, C. S., Int. J. Cancer 41, 184-197, 1988. DURST, M., GISSMANN, L., IKENBERG,H., and ZUR HAUSEN, H., Proc. Nat/. Acad. Sci. USA 80, 3812-3815. 1983. PATER, M. M., and PATER, A., L/iro/ogy 145, 313-318, 1985. SCHWARZ, E., FREESE,U. K., GISSMANN, L., MAYER, W.. ROGGENBUCK,B., STREMLAU,A., and ZUR HAUSEN, H., Nature 314,111114, 1985. BAKER,C. C., PHELPS,W. C., LINDGREN,V., BRAUN, M. J., GONDA, M. A., and HOWLEY, P. M., J. Vii-o/. 81, 962-971, 1987. SCHNEIDER-GADICKE,A., and SCHWARZ, E., EMBO J. 5, 22852292, 1986. SEEDORF,K., OLTERSDORF,T., KRAMMER, G., and ROWEKAMP,W., EMBO J. 6, 139-144, 1987. 8. SMOTKIN, D., and WE-STEIN, F. 0. Proc. Nat/. Acad. Sci. USA 83,4680-4684, 1986.
9. SMOTKIN, D.. and WETSTEIN, F. O., J. Vifol. 61, 1686-1689, 1987. 10. BEDELL, M. A., JONES, K. H., GROSSMAN,S. R.. and LAIMINS, L. A., J. Viral. 63, 1247-l 255, 1989. 11. MATLASHEWSKI,G., SCHNEIDER,.I.. BANKS, L., JONES, N., MURRAY, A., and CRAWFORD, L., EMBO J. 6, 1741-1746, 1987. 12. PHELPS,W. C., YEE, C. L., MUNGER, K., and HOWLEY, P. M., Cell 53, 539-547, 1988. 13. CROOK, T., MORGENSTERN,J. P., CRAWFORD, L., and BANKS, L., EMBOJ. 8, 513-519, 1989. 14. PATER, M. M., HUGHES, G. A., HYSLOP, D. E., NAKSHATRI,H., and PATER, A., Nature 335, 832-835, 1988. 15. PATER. M. M., MARSHALL, T., NAKSHATRI,H.. and PATER,A. UCLA Sym. Mol. Cell Biol. New Series 126, 313-32 1, 1990. 16. PATER, M. M., NAKSHATRI,H., KISAKA,C., and PATER,A., Virology 186, 348-351, 1992. 17. STOREY, A., PIM, D., MURRAY, A., OSBORN, K., BANKS, L., and CRAWFORD, L., EMBO J. 7, 1815-1820, 1988. 18. DOOREIAR,J., PARTON, A., HARTLEY, K., BANKS, L., CROOK, T., STANLEY, M., and CRAWFORD, L., Virology 178, 254-262, 1990. 19. ROHLFS, M., WINKENBACH,S., MYER, S., RUPP,T., and DURST, M., virology 183, 331-342, 1991. 20. TANAKA, A., NODA, T.. YAJINA, H., HATANAKA, M., and ITO, Y., J. Viral. 63, 1465-l 469, 1989. 27. MARSHALL, T., PATER, A., and PATER. M. M., J. Med. Viral. 29, 115-126, 1989. 22. KREIG, P. A., and MELTON, D. A., In “Methods in Enzymology” (R. Wu, Ed.), Vol. 155, pp. 397-415. Academic Press, San Diego. 23. CHIRGWIN, 1. M., PRYZYBYL%A. E., MACDONALD, R. C., and RUTTER, W. J., Biochemistry 18, 5294-5299, 1979.
188,
369-372 (1992)
Glucocorticoid-Dependent Transformation by Human Papillomavirus 16 E7 Coding and 3’ Noncoding Sequences ALAN
PATER, NARASIMHASWAMYS. BELAGULI,
Basic Medical Sciences,
Faculty of Medicine, Received
Memorial
A. R. GARDNER, ALKA MITHAL,
HUMPHREY University
of Newfoundland,
Type
AND MARY
St. John’s, Newfoundland,
M. PATER’
Canada A 1B 3V6
October 4, 199 1; accepted January 29, 1992
The establishment of transformation of primary rodent cells by human papillomavirus (HPV) type 16 DNA requires glucocorticoid hormones (Pater et a/., Nature 335, 832-835, 1988). Here we provide evidence by mutational analysis that, in the context of the hormone-regulated HPV 16 promoter/enhancer, the only protein coding sequences of HPV 16 required are those of the E7 gene. Moreover, additional sequences adjacent to the 3’ end of E7 coding sequences are also essential for the establishment of the transformed phenotype. Splice donor sites, especially an E7 ORF 3 proximal one, are implicated for this k-acting function, since specific deletion mutations of these splice sites greatly or completely reduced the frequency of transformation and the level of E7 RNA. o 1992 Academic PESS, IK-.
in Fig. 1. Plasmids pT64, pA7, pA41, pH4, and pCK1 were constructed as previously described (15, 16). Plasmid pA32 was generated by digestion, with 1.8 kU/ml Sl nuclease for 30 min at 27”, of partially IVarldigested HPV 16 DNA. Bal 31 was used to generate deletions at the Hindll site for pH 17 and at the Kpnl site for pHKB7, pHKB5, and pHKB1; the flush ending of the Narl site and /Varl to /Vcol fragment was used for pA5 and pAM16, respectively. For pHGl6A, the large nt 4466 Srul to 863 Ncol fragment and the small nt 4138 Accl to 4466 Srul fragment were flush ended and ligated. All deletions were confirmed by sequencing. Plasmids pT64 and pCK1, which have deletions removing the splice donor site for E6* (8) and important E6 ORF sequences, were capable of transformation (Fig. 1). These results indicate that neither E6 nor E6* are required. Plasmid pH4, containing a translation termination linker at position 683 which is in the E7 ORF, had no transforming ability, confirming the essential role of this gene in hormone-dependent establishment of transformation. Mutations within the 1947 nt of the El ORF demonstrated that the El gene product is not required, as follows. Plasmids pA7 and pA5 introduced small deletion mutations at 2 and 448 nt from the El ORF 5’ terminus, respectively, which disrupt the El ORF; plasmids pA32 and pH17 were deletion mutations of El ORF nt 413 to 460 and 1730 to 1947, respectively, from the 5’ terminus. All four of these constructs were fully transformation competent (Fig. 1). However, plasmid pHGl6A containing a 3275-bp deletion encompassing all of the El ORF as well as the rest of the early region E2, E4, and E5 ORFs failed to transform. We then generated smaller deletions to localize the E7 ORF 3’ region required for transformation. Plasmid
Squamous cell carcinoma of the uterine cervix is the second most common cancer of females in the world (I). Studies using the techniques of DNA hybridization and molecular cloning have shown the presence of human papillomavirus (HPV) types 16 and 18 in 70 to 80% of cervical tumors and tumor cell lines (2-4). Also, HPV types 16 and 18 RNA transcripts are detected in both cervical tumors and tumor cell lines (5-8). These RNA transcripts and expressed viral protein products are mostly from the E6-E7 region of viral DNA (7, 9), suggesting a role of these genes in the malignant conversion of cervical cells. In vitro transformation studies of rodent cells using the early region of HPV 16 or 18 expressed from a heterologous promoter have identified the transforming region as the E7 gene (10-12). Deletion mutagenesis studies have provided evidence for the essential role of this gene in the establishment of transformation (I 1, 12). Moreover, studies using HPV 16 early region sequences under the control of the hormone-inducible promoter of mouse mammary tumor virus (MMTV) have revealed the essential role of this gene in the maintenance of the transformed phenotype (13). In the present study, mutations in the context of the intact viral genome including its hormone-responsive promoter/enhancer are used. Results show that, while the E7 gene of this virus is required for the establishment of the transformed phenotype, sequences adjacent to the 3’ end of the E7 open reading frame (ORF) are also required. The various mutations in HPV 16 plasmids used to characterize essential sequences and results of transformation assays using 5 pg of HPV 16 and ras DNA with previously described procedures (14) are shown ’ To whom reprint requests should be addressed.
369
0042-6822/92
$3.00
CopyrIght 0 1992 by Academic Press, Inc All rights of reproduction in any form reserved
SHORT COMMUNICATIONS
370
Plrsmid
Transformation
HPV 16
23
pT64
20
pCK1
18
pH4
0
PA7
22
PAS
20
~A32
24
pHG16A
0
225-294
.E6
132-39p ; E6 663;
E7
667;. El 1313; 1278-1325;
El El
2965-4004; PA41
26
pH17
25
pAnI
0
pHKB7
1
pHKB5
2
pHKB1
2
E2, E4. ES
2595-3907; 863-1309;
El
a77,-893;
El
8794387; I
El
881-884;
El
FIG. 1. Transformation assays of BRK ceils with HPV 16 mutant constructs. El to E7 are early ORFs; Ll and L2 are late ORFs. The arrows and P indicate initiation sites for transcription; A’s indicate polyadenylation signals. Large deletions are indicated as black boxes; small deletions and insertions are indicated by vertical lines. Affected nucleotides and ORFs are indicated. The results are the average of three independent experiments using four plates per experiment. The results for the last three plasmids are from five experiments with four plates each, and the average number of colonies transformed by wild-type HPV 16 plasmid for these experiments was 19.
pA41, containing a 1019-bp deletion encompassing E2, E4, and E5 ORFs, was transformation competent. Also, plasmid pH 17, with a 1312-bp deletion encompassing the 3’end of the El ORF and the entire E2, E4, and E5 ORFs, was transformation competent. However, plasmid PAM 16 with a 546-bp deletion in the 5’ end of the El ORF failed to transform. Since the possibility of a trans-acting function of the El ORF had been eliminated, this region must have a c&-acting function. Thus, the data of Fig. 1 provide evidence that E7 is the only ORF required for HPV 16 enhancer/promoterand glucocorticoid hormone-dependent transformation of primary BRK cells. This is consistent with the work of others in which HPV coding sequences were expressed from a heterologous promoter and, accordingly, in the absence of hormones (12, 17). However, Fig. 1 further shows that 3’ noncoding sequences are also required. To further analyze the nature of the sequences at the 5’ end of the El sequences required for E7-mediated transformation, we generated smaller deletion mutations. These were at the nt 880 splice donor site, 3’ to the E7 ORF and present in E7 pre-mRNA, which is utilized in both cell lines derived from cervical lesions as well as HPV 16-immortalized keratinocytes (18, 19, our unpublished observations). In addition, all three cDNA
clones isolated from a tumorigenic rodent cell line transformed by HPV 16 DNA were the products of utilization of this splice donor site (20). The results of Fig. 1 for pHKB7, pHKB5, and pHKB1 show that, in the context of the homologous HPV enhancer/promoter-containing constructs, three nt 880 splice site-specific deletions as small as 4, 9, and 16 bp greatly reduced, although they did not completely eliminate, transformation. This is in contrast to the larger deletion mutations which repeatedly failed to transform (Fig. 1; pHGl6A and pAM16). Within E7 pre-mRNA there are two other sites at nt 226 and 1302, and both sequences are more similar to the consensus sequence than is the 880 site. The 1302 sequence had been shown by cDNA analysis to be functional in a cell line from a precancerous cervical lesion (18). In addition, the potential of 1302 site splicing events may be obscured by its position within the intron generated from the nt 880 site. Since utilization appears to be low for this site, the influence of splicing at this site may be less than that of the 880 site. Indeed, deletion of the 1302 site from HPV 16 DNA in pA32 (Fig. 1) failed to affect transformation. Similarly, deletion of the 226 site in pT64 and pCK1 (Fig. 1) had no effect on transformation. Moreover, the splice donor site at nt 226 does not substitute for the 880 site since deletions of the 880
371
SHORT COMMUNICATIONS
Experiment
SVP/E
Plasmid
1
2
pD7
102
79
pD7KB
0
0
---c-a---------
---------------e
881-884 '
------a--
of BRK cells with the intact and mutant early region of HPV 16 expressed from the SV40 enhancer/promoter. The FIG. 2. Transformation results are the number of colonies from a total of seven plates for experiment 1 and six plates for experiment 2. The expression vector sequences are indicated by the dashed lines and HPV sequences are as in Fig. 1. The SV40 promoter/enhancer is indicated @VP/E) by the arrow.
and 1302 sites in pHG16A and pAMl6
(Fig. 1) completely eliminated transformation. These results indicate that the 880 site has a strong dominant role in transformation and that, in the absence of this site, the splice sites at 226 and 1302 may function, resulting in a low level of transformation. Since the effect of 880 splice donor site mutations was so great, we also examined the effect of the splice acceptor sites for the 880 donor site. These are at nt 2708 and 3357. The removal of either the 3357 site in pA41 or both sites in pH 17 had no effect on transformation (Fig. 1). Thus, additional splice acceptor sites would appear to function to allow full transformation. However, the nt 880 splice donor site appears to be critically important since loss of this site severely reduced transformation, even in the presence of the two alternative donor sites at nucleotides 226 and 1302 (Fig. 1, pHKB5, pHKB7, and pHKBl), and completely abolished transformation in the presence of only the 226 donor site (Fig. 1, pAMl6 and pHGl6A). To examine the mechanism by which the splice donor site at nt 880 affects transformation, we probed with RNase protection assays, using Cos-1 cells which amplify SV40 origin region-containing sequences. Plasmid pD7KB, containing the same 4-nt 880 deletion as pHKB1, was generated by T4 DNA polymerase I incubation of Kpnl-linearized pD7 (2 I). First, transformation assays were done as before (14) except in the absence of dexamethasone. The pD7 plasmid gave transformed colonies while pD7KB gave no colonies (Fig. 2). Since pD7 and pD7KB lack the 226/409 splice donor/acceptor sites, the effect on transformation of deletion of the 880 plus 226/409 splice sites is the same as the effects in pA32 and pHGl6A of the deletion of the nt 880 plus 1302 splice donor sites. This reinforces the implication that both the 226 and the 1302 sites have some effect on E7 expression but that this effect on transformation can only be detected in the absence of the predominating effect of the 880 splice donor site. Next, the effect of 880 splice site mutation on E7 expression was analyzed by RNase protection (22) of 5 pg total cellular RNA, prepared as described by Chirgwin et a/. (23). Cos-1 cells were
transfected with 10 pg of plasmid 48 hr prior to harvesting. Probe for E7 was from Bluescript KS(+) containing the SV40 nt 5 190 Stul to HPV nt 720 Sspl fragment of pD7. The SV40 T antigen RNA level was an internal control; probe was from a similar construct for the SV40 3072-3225 Rsal fragment. Hybridization was with 1O5 cpm each of both probes for 14 hr at 56” and digestion was with 700 U/ml RNase Tl. E7 products were resolved on 5% polyacrylamide denaturing gels. The splice site mutation in pD7KB reduced the RNA level dramatically (Fig. 3). Thus, the strong &-acting effect of the functional 880 site, in transformation assays (Fig. l), is also observed for E7 oncogene expression. Since splicing is a post-transcriptional event, the 123456
E7-
FIG. 3. RNase protection assays. On the left, the larger type letters indicate E7 (E7) and T antrgen (T) probes. Smaller type letters rndicate the products of RNase protection, which are smaller for CaSki E7 RNA since it lacks SV40 sequences. Molecular weight marker sizes in bases are indicated on the right. Lane 1, E7 and T antigen probes; lane 2. CaSki; lane 3. pD7; lane 4, pD7KB; lane 5, Cos-1; lane 6. Escherichia co/i tRNA.
372
SHORT COMMUNICATIONS
influence of the 880 splice site for transformation is most likely due to post-transcriptional factors such as RNA stability and/or processing which influence the levels of RNA available for E7 product synthesis. Therefore, the integrity of the 880 splice donor site is apparently specifically important to maintain sufficient levels of E7 RNA to allow transformation. ACKNOWLEDGMENTS We thank G. Hughes and D. Hyslop for technical assistance, S. Atkins for typing the manuscript, and H. zur Hausen and L. Gissmann for HPV 16 plasmid. This work was supported in part by grants from the MRC and NCI of Canada.
REFERENCES PARKIN, D. M., LAURA, E., and MUIR, C. S., Int. J. Cancer 41, 184-197, 1988. DURST, M., GISSMANN, L., IKENBERG,H., and ZUR HAUSEN, H., Proc. Nat/. Acad. Sci. USA 80, 3812-3815. 1983. PATER, M. M., and PATER, A., L/iro/ogy 145, 313-318, 1985. SCHWARZ, E., FREESE,U. K., GISSMANN, L., MAYER, W.. ROGGENBUCK,B., STREMLAU,A., and ZUR HAUSEN, H., Nature 314,111114, 1985. BAKER,C. C., PHELPS,W. C., LINDGREN,V., BRAUN, M. J., GONDA, M. A., and HOWLEY, P. M., J. Vii-o/. 81, 962-971, 1987. SCHNEIDER-GADICKE,A., and SCHWARZ, E., EMBO J. 5, 22852292, 1986. SEEDORF,K., OLTERSDORF,T., KRAMMER, G., and ROWEKAMP,W., EMBO J. 6, 139-144, 1987. 8. SMOTKIN, D., and WE-STEIN, F. 0. Proc. Nat/. Acad. Sci. USA 83,4680-4684, 1986.
9. SMOTKIN, D.. and WETSTEIN, F. O., J. Vifol. 61, 1686-1689, 1987. 10. BEDELL, M. A., JONES, K. H., GROSSMAN,S. R.. and LAIMINS, L. A., J. Viral. 63, 1247-l 255, 1989. 11. MATLASHEWSKI,G., SCHNEIDER,.I.. BANKS, L., JONES, N., MURRAY, A., and CRAWFORD, L., EMBO J. 6, 1741-1746, 1987. 12. PHELPS,W. C., YEE, C. L., MUNGER, K., and HOWLEY, P. M., Cell 53, 539-547, 1988. 13. CROOK, T., MORGENSTERN,J. P., CRAWFORD, L., and BANKS, L., EMBOJ. 8, 513-519, 1989. 14. PATER, M. M., HUGHES, G. A., HYSLOP, D. E., NAKSHATRI,H., and PATER, A., Nature 335, 832-835, 1988. 15. PATER. M. M., MARSHALL, T., NAKSHATRI,H.. and PATER,A. UCLA Sym. Mol. Cell Biol. New Series 126, 313-32 1, 1990. 16. PATER, M. M., NAKSHATRI,H., KISAKA,C., and PATER,A., Virology 186, 348-351, 1992. 17. STOREY, A., PIM, D., MURRAY, A., OSBORN, K., BANKS, L., and CRAWFORD, L., EMBO J. 7, 1815-1820, 1988. 18. DOOREIAR,J., PARTON, A., HARTLEY, K., BANKS, L., CROOK, T., STANLEY, M., and CRAWFORD, L., Virology 178, 254-262, 1990. 19. ROHLFS, M., WINKENBACH,S., MYER, S., RUPP,T., and DURST, M., virology 183, 331-342, 1991. 20. TANAKA, A., NODA, T.. YAJINA, H., HATANAKA, M., and ITO, Y., J. Viral. 63, 1465-l 469, 1989. 27. MARSHALL, T., PATER, A., and PATER. M. M., J. Med. Viral. 29, 115-126, 1989. 22. KREIG, P. A., and MELTON, D. A., In “Methods in Enzymology” (R. Wu, Ed.), Vol. 155, pp. 397-415. Academic Press, San Diego. 23. CHIRGWIN, 1. M., PRYZYBYL%A. E., MACDONALD, R. C., and RUTTER, W. J., Biochemistry 18, 5294-5299, 1979.
