JOURNAL OF VIROLOGY, Aug. 1992, p. 49574965

Vol. 66, No. 8

0022-538X/92/084957-09$02.00/0 Copyright © 1992, American Society for Microbiology

Identification and Genetic Definition of a Bovine Papillomavirus Type 1 E7 Protein and Absence of a Low-Copy-Number Phenotype Exhibited by E5, E6, or E7 Viral Mutants NICLAS JAREBORG, ANDERS ALDERBORN, AND STANLEY BURNETT* Department of Medical Genetics, Biomedical Center, Boax 589, S-751 23 Uppsala, Sweden Received 26 March 1992/Accepted 18 May 1992

The bovine papillomavirus type 1 (BPV-1) genome replicates as a multiple-copy plasmid in murine C127 cells transformed to neoplasia by virus infection or by transfection with BPV-1 DNA. It was reported previously that BPV-1 genomes harboring frameshift mutations in the E6 or E7 open reading frame (ORF) replicated in C127 cells transformed by these mutants at a low copy number. Furthermore, the characterization of a BPV-1 mRNA in which the E6 and E7 ORFs were spliced together in frame has led to the assumption that an E6/7 fusion protein is expressed in virus-transformed C127 cells. To define the number and nature of the E6 and E7 gene products expressed in BPV-1-transformed cells, we performed immunoprecipitation experiments with antisera raised to bacterially expressed BPV-1 E6 and E7 fusion proteins. By employing cell culture conditions which induce BPV-1 E2 transactivator expression and viral early region transcription in virus-transformed C127 cell lines, we detected a single immunoprecipitated E6 protein species with an apparent molecular mass of 17 kDa and a single E7 protein species with an apparent molecular mass of 15 kDa. To characterize further these E6 and E7 proteins, C127 cells were transformed by transfection with BPV-1 genomes containing mutations predicted to prevent expression of specific E6 or E7 gene products, and the transformed cells were subjected to immunoprecipitation analysis with the E6 or E7 antiserum. The results of these experiments confirmed that the E6 and E7 ORFs encode distinct proteins and failed to establish the existence of an E6/7 fusion protein. We did not find a significant difference in the viral genome copy number between clonal C127 cell lines transformed by wild-type BPV-1 or by mutant viral genomes unable to express the E6 or the E7 protein. Furthermore, in contrast to two previous reports suggesting that expression of the BPV-1 E5 gene was required for the establishment or maintenance of a high viral plasmid copy number, we observed a two- to fourfold increase over wild-type BPV-1 plasmid copy number in C127 cells transfected with a BPV-1 E5-minus mutant and subsequently selected by neoplastic focus formation.

Regulation of papillomavirus gene expression and viral DNA replication has been more extensively studied for bovine papillomavirus type 1 (BPV-1) than for any other wart virus. Consequently, much information has been obtained about the transcriptional organization of BPV-1, the function of specific open reading frames (ORFs) identified from DNA sequence analysis, and the molecular mechanism of viral DNA replication initiation (14, 32, 34). Such studies have relied heavily on the development of a cell culture system susceptible to the neoplastic transforming potential of molecularly cloned BPV-1 DNA, and the murine C127 cell line has become widely accepted as an appropriate target cell type (9). The 7,945-bp covalently closed circular DNA genome of BPV-1 replicates as a multiple-copy extrachromosomal plasmid in C127 cells transformed by it (15). Several years ago it was reported that frameshift mutations in the BPV-1 E6 and E7 ORFs resulted in the establishment of E6 and E7 mutant viral genomes at a very low plasmid copy number in C127 cells transformed by these mutants (2, 3, 17). However, some investigators have failed to confirm that similar or identical E6 or E7 BPV-1 mutants are affected in this property (19, 21). Frameshift mutations in the BPV-1 E5 ORF were also reported to result in a replication-defective phenotype in a long-term replication assay in which cell clones derived from single transfected cells were analyzed *

(11, 21). Results obtained recently by using a transient replication assay have indicated that expression of E5, E6, or E7 is not required in the primary stages of BPV-1 DNA plasmid replication (32). The proposed role of these viral genes in long-term or stable replication assays therefore remains obscure. The E5 and E6 ORFs encode independently functioning transforming proteins which are each capable of inducing morphological and neoplastic transformation of C127 cells (1, 4, 24, 25). Although an intact E7 ORF may be required for full transformation of C127 cells by BPV-1 (19), no independent transforming activity has been detected for the BPV-1 E7 ORF expressed from a strong heterologous promoter (24). This is in contrast to the situation for certain human papillomavirus types (types 16 and 18) in which E7 has been shown to be capable of inducing transformation in a number of in vitro cell transformation assays (reviewed by Vousden et al. [33]). No attempt has been made to identify a BPV-1 E7 protein. A variety of alternatively spliced BPV-1 transcripts initiated from the promoters P89 and P7940, just upstream of the E6 ORF, have been isolated (29, 35) from BPV-1-transformed C127 cells (see Fig. 1). Some of these transcripts (types I to III) had the potential to encode an E6/7 fusion protein comprising the N-terminal region of E6 and the C-terminal region of E7. The type VI transcript was colinear with the E6 and E7 ORFs and could potentially encode both a full-length E6 protein and a full-length E7 protein. Yet another spliced transcript (V) was predicted to encode an E6/4 fusion protein.

Corresponding author. 4957

4958

1I

JAREBORG ET AL.

D304

I

I

~,1. . I

I

I

I

predicted

A2557 D2505

D1234 A527 D864

I

J. VIROL.

I

II

-

7

E

~~~~~~~~~~~~~~~~~I I

I

I

*

I

VI

I

I

I

E6/E7 I

II

I

I

~~~~~~~~~~~E61E7,El1(N)

I

IIIII IV * I 0 II I I III I V I

E6/E7, E1

I I

11 * I I

protein product

A3224

I

E2

I

I

E6/E4

II

I

E6, E7

III I

I

..

I EI I

I

7945/1 * I PL

a

.1000.

.

P89

..

...............

|:::::-:-:::-::::::::-::-:: - - ::::

.

I

2000 l

LWP890

4000

3000.

II

I40..

I

I

I

l4.

P2443 P3080 [LP8 LP8 pA 4179 P7185 P7940 FIG. 1. Structure of spliced BPV-1 early transcripts initiated from P89/P7940. Six different species of spliced BPV-1 early region transcript (I to VI) which initiated close to P89 and terminated at the early mRNA polyadenylation site have been detected in previous investigations (29, 30, 35). The type I transcript was reported to be present mainly in nuclear RNA fractions (29). The structures of transcripts from other viral promoters are not shown. In the lower part of the figure are shown transcriptional regulatory sequences and translational ORFs (shaded) in the early (E) and late (L) regions. Initiator ATG codons are represented by vertical lines within the boxes. Arrows indicate major mRNA initiation sites. Vertical dashed lines show the positions of known early region splice donor and acceptor sites. pA, functional hexanucleotide polyadenylation signal.

To attempt to define how many E6- and E7-related proteins may be expressed in BPV-1-transformed C127 cells, we have performed immunoprecipitation experiments with antisera specific for the BPV-1 E6 and E7 proteins. We report here the detection of E6 and E7 proteins which appear to be encoded by their respective full-length ORFs. We are unable to confirm the existence of an E6/7 fusion protein or any other E6- or E7-related protein. In keeping with the hypothesis that the E5, E6, and E7 genes are positively regulated by the BPV-1 E2 transcriptional activator protein (12, 20, 28), expression of the E5, E6, and E7 proteins was induced under cell culture conditions which promoted E2 transactivator gene expression. In our study, we found that mutant BPV-1 genomes unable to express E5, E6, and E7 proteins did not exhibit a low-copy-number phenotype in C127 cells transformed by these mutants. MATERUILS AND METHODS Construction of pGEX/E7 recombinant plasmid and purification of fusion protein. A 220-bp XmaIII-NnmI fragment encoding amino acid residues 48 to 120 of the BPV-1 E7 ORF (total length, 127 amino acid residues) was cloned into the polylinker of the prokaryotic expression vector pGEX-3X (26), thus creating a fusion gene between Schistoma japonicum glutathione S-transferase (Sj26) and E7. The recombinant plasmid was analyzed by partial sequence analysis by the chain termination method (23) to confirm that the Sj26 and E7 coding regions were in frame. The fusion protein was expressed in Escherichia coli (strain DH5oa) by induction with isopropyl-,-D-thiogalactopyranoside and was purified

from cell lysates by absorption to glutathione-coupled agarose beads as described elsewhere (26). Protein preparations were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on 10% (wt/vol) polyacrylamide gels with a 5% (wt/vol) polyacrylamide stacking gel. Sera. Three rabbits were each immunized three times at 4-week intervals with 100 ,ug of E7 fusion protein dispersed in Freund's complete adjuvant. Sera were analyzed for the presence of antibodies to the fusion protein by Western blotting (immunoblotting) (31). The serum which showed the greatest reactivity to the Sj26/E7 fusion protein in Western blots was selected for use in immunoprecipitation experiments. The Sj26/E7 serum was blocked by overnight incubation of an aliquot of serum with purified Sj26 protein at a final concentration of 1 mg/ml. Antiserum to a full-length BPV-1 E6 protein expressed in E. coli (1) was kindly provided for these studies by Douglas Lowy (Laboratory of Cellular Oncology, National Cancer Institute, Bethesda, Md.) and Elliot Androphy (Department of Dermatology, New England Medical College, Boston, Mass.). Antiserum to BPV-1 E5 C-terminal synthetic peptide was obtained from Daniel DiMaio (Department of Genetics, Yale University School of Medicine, New Haven, Conn.). Cell lines and immunoprecipitation procedures. C127 cell lines wh.2 and cl.2 infected with replication-competent wildtype (wt) or mutant BPV-1 genomes, respectively, have been described elsewhere (6-8). The E6 and E7 mutant BPV-1 genomes used in this study were obtained from Daniel DiMaio and have been described previously (19). They were renamed in our study as follows: E6ocl (E61), E6fs2 (E62),

VOL. 66, 1992

IDENTIFICATION AND DEFINITION OF BPV-1 E7 PROTEIN

E67SA (E67), E7ocl (E71), and E7oc2 (E72), with the new code in parentheses. To induce cell growth arrest, mutant virus-infected cl.2 cells were maintained in confluent monolayer culture in medium containing 2.5% fetal calf serum. The culture medium was renewed every fourth day. To induce growth arrest of wt BPV-1-transformed wh.2 cells, the cells were grown to near confluence (75%) in Dulbecco modified Eagle's medium containing 10% (vol/vol) fetal calf serum and were then switched to serum-free Dulbecco modified Eagle's medium, which was thereafter renewed every second day. Confluent stationary-phase cl.2 or wh.2 cultures were used for immunoprecipitation experiments 12 days or longer after reaching confluence. Exponential-phase or stationary-phase cell cultures were labelled with 200 pCi each of [35S]methionine and [35S]cysteine for 1 to 3 h at 37°C in serum-free Dulbecco modified Eagle's medium lacking cold methionine and cysteine, in a final volume of 1.5 ml. Immunoprecipitations were then performed with preimmune serum or with E6- or E7-immune sera by using the method of Hubbert et al. (13). Immunoprecipitation with E5-immune sera was performed as described previously (5). Transfection of cells and Southern blot analysis of cell lines. C127 cells were transfected with molecularly cloned BPV-1 DNA by the calcium-phosphate coprecipitation procedure (10). Dense foci of transformed C127 cells induced by transfection with wt or mutant BPV-1 DNA were picked at 3 weeks posttransfection (or at 5 weeks for foci induced by E5-minus BPV-1 DNA) and were established as independent cell lines. Total cellular DNA was isolated from cells grown in exponential-phase culture and was analyzed by Southern blotting (27) and hybridization with a 32P-labelled BPV-1 virion DNA probe prepared by random priming. Prior to electrophoresis on 0.8% (wt/vol) agarose gels, the cellular DNA was treated with restriction endonucleases as indicated in Results. Hybridization and washing of filters were carried out under stringent conditions (6). RESULTS Production of antibodies to bacterially expressed BPV-1 E7 protein. We chose to express a portion of the E7 ORF in the form of a fusion protein in E. coli and to use the purified fusion protein to immunize rabbits. Since previous investigations have led us to believe that an important viral function is encoded by the spliced E6/7 RNA transcript, it was necessary that antibodies were raised to the C-terminal portion of E7 which is contained within the hypothetical E6/7 fusion protein. We therefore cloned the C-terminal half of the E7 ORF into the prokaryotic expression vector pGEX-3X (26), thus creating a fusion gene between Sj26 and E7. Fusion protein was purified from cell lysates as described in Materials and Methods and was then analyzed by SDS-PAGE. As shown in Fig. 2, visualization of the fusion protein with Coomassie brilliant blue indicated that the preparation was essentially pure, containing only the Sj26/E7 fusion protein and a putative degradation product which comigrated with the Sj26 protein. The apparent molecular mass of the largest species was 31 kDa, which is similar to that predicted from its amino acid composition (35

kDa). Sera from rabbits immunized with the purified Sj26/E7

protein preparation were analyzed for the presence of antibodies to the fusion protein by Western blotting. As shown in Fig. 2, Sj26/E7-immune serum reacted with both the Sj26 and the Sj26/E7 proteins, whereas preimmune serum did not react with these proteins. To investigate whether the antise-

CBB M

Sj26 Sj26 E7

o.-Sj26 E7

Sj26 Sj26 E7

Sj26 Blocked

4959 Pi

Sj26 Sj26E7 Sj26 E7

664536-

I

X6

29-5 24_

20--

FIG. 2. Electrophoretic and Western blot analyses of Sj26/E7 fusion protein. Purified Sj26 and Sj26/E7 proteins were fractionated by SDS-PAGE on a 10% (wt/vol) polyacrylamide gel and were electrotransferred to an Immobilon polyvinylidene difluoride membrane prior to Coomassie brilliant blue (CBB) staining or Western analysis with immune serum (a-Sj26/E7), Sj26 protein-blocked immune serum (Sj26-Blocked), or preimmune serum (PI). Molecular size markers (M) are as follows: bovine serum albumin (66 kDa), ovalbumin (45 kDa), glyceraldehyde-3-phosphate dehydrogenase (36 kDa), carbonic anhydrase (29 kDa), trypsinogen (24 kDa), and trypsin inhibitor (20 kDa).

rum contained E7-specific antibodies, an aliquot of the immune serum was incubated overnight with Sj26 protein to

block any reactivity specific for Sj26. This inhibited binding of antibodies to Sj26 but not to the Sj26/E7 protein, indicating that the antiserum contained antibodies directed to the E7 portion of the protein. The species that comigrated with Sj26 was also detected after blocking with purified Sj26 protein and was therefore assumed to be a degradation product of the fusion protein retaining some E7 peptide sequences. Detection of BPV-1 E6 and E7 proteins in C127 cell lines. We anticipated that it might be difficult to detect viral E6 and E7 proteins under normal cell culture conditions. For example, Androphy et al. have reported very low levels of E6 in BPV-1-transformed cells (1). Hence, we decided to begin our immunoprecipitation experiments by using virus-transformed cell lines exposed to culture conditions known to induce expression of the BPV-1 E2 transcriptional activator protein. As shown previously, E2 protein expression and viral early region transcription are induced by prolonged cell growth arrest in confluent culture (8). This is achieved for wt BPV-1-transformed C127 cells by serum deprivation or by contact inhibition for the minimally transformed C127 cell line, cl.2, containing a mutant BPV-1 genome (6, 7). Preliminary immunoprecipitation experiments were done by using cl.2 cells containing the previously characterized replication-competent mutant BPV-1 genome with intact E6 and E7 ORFs (7). Cultures of cl.2 cells maintained at confluence for 10 days were incubated with [35S]methionine plus [35S]cysteine, and cell lysates were immunoprecipitated with E7 antiserum or with preimmune serum. Uninfected C127 cells were labelled in parallel as a control. As shown in Fig. 3, the E7 antiserum precipitated a protein with an apparent molecular mass of 15 kDa from c1.2 cells but not

4960

JAREBORG ET AL.

P1

C127 E6

cI.2 conf. E6 Pi

C127 E7 Pi

cI.2 conf. P1 E7

46-

30

21.5-

14.5

if

J. VIROL.

from uninfected C127 cells. This molecular mass estimation was close to that calculated from the amino acid composition for a full-length E7 protein (13.6 kDa) but was much less than expected for an E6/7 fusion protein (20 kDa). This did not, however, rule out the possibility that this E7-related protein consisted of an E6/7 protein with an anomalous electrophoretic mobility. Precipitation of the 15-kDa protein could be specifically blocked by incubating the antiserum overnight with Sj26/E7 fusion protein but not by blocking with Sj26 protein alone (data not shown). If the putative E7 protein detected in cl.2 cells was an E6/7 fusion protein, then it should also be recognized by E6 antibodies. The production of polyclonal antibodies to a full-length E6 protein expressed in E. coli has been described previously (1). These E6 antibodies have been used to detect a BPV-1 E6-related protein with an apparent molecular mass of 15.5 kDa in virus-transformed murine NIH3T3 and C127 cell lines. By immunoprecipitation of cl.2 cell lysates with E6 antiserum, we detected an E6-related protein with an apparent molecular mass of 17 kDa, which is close to the value reported earlier. There was a slight, but reproducible, difference in the electrophoretic mobilities of the putative E6- and E7-related proteins expressed in growth-arrested cl.2 cell cultures. This was a further indication that the E6 and E7 ORFs encoded separate translation products in this cell line. To confirm and extend these findings, we sought expression of E6 and E7 proteins in the wt BPV-1-transformed C127 cell line, wh.2. We examined expression of E6 and E7 proteins in wh.2 cells maintained in exponential-phase culture or in stationary-phase cultures which had been deprived of serum for 10 days to induce viral early region transcription. We detected the same E6 and E7 proteins in stationaryphase wh.2 cultures as we detected in cl.2 cells (Fig. 4A and

E6

E7

FIG. 3. Detection of E6 and E7 proteins in stationary-phase cl.2 cells. 35S-labelled proteins were immunoprecipitated from cell lysates prepared from uninfected C127 cells or from 10-day-confluent cl.2 cell cultures (cl.2 conf.) by using control preimmune serum (PI), E6 serum, or E7 serum and were separated by SDS-PAGE on 15% (wt/vol) polyacrylamide gels. Positions of the E6 and E7 proteins are indicated by the arrows at the right side. The positions of radiolabelled molecular size markers (ovalbumin [46 kDa], carbonic anhydrase [30 kDa], trypsin inhibitor [21.5 kDa], and lysozyme [14 kDa]; Amersham (CFA 755) are indicated at the left side of the figure. C127 Pi E7

wt exp Pi E7

wt-S Pi

C127 Pi E6

E7

wt.exp

Pi

E6

wt -S Pi E6

C C127

wt exp.

wt-S

wt ind.

Ni Es NI E5 NI E5 NlE

*..'.i. i

*

466-

46 *.

3-W.-

3o-

4146 4 -30

30

1-21.5

21..5

21.5

-

4-14.3

;:.

2155IL t. er:

..

.. ..

_w

&t,w 14.5

4U-6.5 -

4

-

14.5

-

FIG. 4. Analysis of E5, E6, and E7 protein expression in exponential-phase and stationary-phase wh.2 cell cultures. (A and B) wt BPV-1-transformed wh.2 cells in exponential-phase (wt exp) or stationary-phase (wt-S) culture were labelled with [35S]methionine plus [35S]cysteine prior to immunoprecipitation of cell lysates with E7 (A) or E6 (B) antiserum or preimmune serum (PI). Lysates from uninfected C127 cells were immunoprecipitated as controls. Immunoprecipitated proteins were fractionated by SDS-PAGE on 15% (wt/vol) polyacrylamide gels. Positions of the E6 and E7 proteins are indicated by the arrowheads at the right side of each panel. The positions of molecular size markers are also shown (see legend to Fig. 3 for details). (C) Immunoprecipitation analysis of E5 protein expression under different cell culture conditions. wt BPV-1-transformed wh.2 cells were grown in exponential-phase culture (wt exp.), were deprived of serum for 12 days (wt-S), or were deprived of serum and then restimulated with serum (wt ind.), and cell lysates were immunoprecipitated with E5 peptide antiserum.

IDENTIFICATION AND DEFINITION OF BPV-1 E7 PROTEIN

VOL. 66, 1992

A

LI 7945.1

L7

P7940

B

E7 El

200

L-IP89

4961

400

600

D 304

A 527

461

E71 E67

E61

Boo00 8i

E14 4

B

-46

46-

-30

30-

-21.5

21.5E6 -0 _- E7-

--

14.5-

-14.5

FIG. 5. Genetic analysis of E6 and E7 ORF translation products. Transformed C127 cell lines harboring BPV-1 plasmid DNA genomes with specific mutations within the E6 and E7 ORFs were tested for expression of the BPV-1 E6 and E7 proteins by immunoprecipitation analysis. At the top are shown the locations of the E6 and E7 mutations examined in this experiment (see Materials and Methods for details). Cell lines were subjected to deprivation of serum for 10 days prior to labeling with [35S]methionine plus [35S]cysteine and immunoprecipitation with E6 or E7 antiserum as described in Materials and Methods. Control stationary-phase cl.2 cells (cl.2 confl) or uninfected C127 cells were tested as positive and negative control cell lines. Immunoprecipitation products were analyzed by SDS-PAGE on 15% (wt/vol) polyacrylamide gels, followed by autofluorography. Arrows indicate the positions of the immunoprecipitated E6 and E7 proteins. The positions migrated by molecular mass markers are indicated at the sides of the autoradiograms.

B). Exponential-phase wh.2 cells expressed less E6 protein than stationary-phase wh.2 cell cultures when cell lysates from similar numbers of cells were compared (Fig. 4B). We were unable to detect an E7 protein in exponential-phase wh.2 cells. We also failed to detect the E7-related protein when growth-arrested cl.2 or wh.2 cell cultures were labelled with [32P]phosphate, suggesting that this protein was not phosphorylated under these assay conditions. Biochemical fractionation analyses indicated that the E7 protein was present in the cytoplasm of wt BPV-1-transformed wh.2 cells (data not shown). To extend further the analysis of BPV-1 early gene regulation, we also examined expression of the E5 protein in wh.2 cells under different culture conditions. As shown in Fig. 4C, evidence was found for a significant induction of E5 protein synthesis in growth-arrested cell culture (wt-S). After stimulation with serum (wt ind.) to induce maximal amplification of the viral genome copy number in divisionarrested cells (6), there was a further increase in the level of immunoprecipitated E5 protein detected. Thus, we have observed that three different BPV-1 early proteins are coinduced with the E2 transactivator protein in stationary-phase C127 cell culture. Genetic definition of E6 and E7 proteins. The above analysis of E6 and E7 protein expression indicated that the E6 and E7 ORFs encode distinct protein products and that

neither the E6- nor the E7-related protein corresponded to an E6/7 fusion protein. To confirm the genetic identities of the E6 and E7 proteins, we examined their expression in C127 cell lines transformed with a series of BPV-1 genomes harboring mutations predicted to abolish expression of specific E6, E7, and E6/7 proteins (19). Cell lines were established from morphologically transformed foci of cells induced in separate transfections with each of the molecularly cloned mutant BPV-1 genomes. For analysis of E6 and E7 protein expression, each of the cell lines was subjected to serum deprivation to induce viral early gene expression. As shown in Fig. 5, immunoprecipitation analysis revealed that cells transformed by the E61 mutant, with a frameshift mutation in the E6 ORF downstream of the E6 splice donor site, expressed low levels of an E7 protein but did not express detectable E6 protein. The E71 mutant, with a nonsense mutation in the E7 ORF upstream of the splice acceptor site, expressed an E6 protein but did not express detectable E7 protein. Finally, the E67 mutant, with a point mutation at the splice acceptor site which abolished the splice (as previously demonstrated by Neary and DiMaio [19]) but did not disrupt the E7 ORF, expressed both the E6 and E7 proteins. These data were consistent with the proposal that the E6 and E7 proteins detected in our immunoprecipitation analyses were expressed from separate ORFs.

4962

J. VIROL.

JAREBORG ET AL. I--

c'

0 1

6.45.74.8-

2

E62 E61 3 4 1 2 3

__

M

___

1

E67 2 3 4

1

E71 2 3 4

_

_

_

1

E72 2 3 4

wt 2 3 4

1

____

4

*-

Eco/Bam A

4.33.7-

2.3-

_

w

e - Eco/Bam B

1.9-

FIG. 6. Copy number of E6 and E7 mutant viral genomes. Total cell DNA was purified from C127 cell lines in exponential growth phase derived from transformed foci induced by transfection with cloned wt or mutant viral genomes. Four independent cell lines (lanes 1 to 4) transformed by each mutant genome (with the exception of mutant E62, for which three lines were tested) were compared in this experiment. After digestion with a combination of EcoRI and BamHI (one site each in BPV-1), identical amounts of DNA from each cell line were electrophoresed on 0.8% (wt/vol) agarose gels, blotted onto nylon membranes, and then hybridized with a radioactively labelled BPV-1 virion DNA probe. C127, control nontransformed C127 cell line; E61 and E62, cell lines transformed by viral genomes with nonsense (E61) and frameshift (E62) mutations in the E6 ORF downstream and upstream of the E6 splice donor signal, respectively; E67, cell lines transformed by a viral genome lacking a functional E7 splice acceptor signal; E71 and E72, cell lines transformed by viral genomes with nonsense mutations in the E7 ORF upstream and downstream, respectively, of the E7 splice acceptor signal; wt, control wt BPV-1-transformed cell lines. The positions of the two BPV-1 EcoRI-BamHI fragments (Eco/Bam A and Eco/Bam B) are indicated at the right side of the figure. The positions of molecular size markers (in kilobase pairs) are indicated at the left side of the figure.

Furthermore, there was no evidence for expression of an E6/7 fusion protein in these cells. Plasmid copy number of E6 or E7 mutants. Conflicting data on the role of the E6 and E7 ORFs in BPV-1 plasmid copy number regulation have been published. On the basis of results presented in a series of papers it was proposed that the E6 protein and the predicted fusion protein product of the previously characterized E6/7 transcript acted in trans to establish a high viral plasmid copy number (2, 3, 17). Later studies performed in other laboratories have, however, failed to confirm that BPV-1 E6 or E7 mutants replicate at a low copy number in C127 cells (19, 21). We have analyzed the BPV-1 plasmid copy number in C127 cells transformed by each of the E6 or E7 mutants described in the previous section in addition to two other previously described mutants with frameshift mutations in the E6 and E7 ORFs (19). All these E6 and E7 mutants have earlier been reported to replicate at a normal copy number (19). For each mutant genome, four independent transformed cell lines (three lines for mutant E62) were isolated, and exponential-phase cultures were analyzed by Southern blot hybridization analysis with a 32P-labelled BPV-1 DNA probe. Purified total cellular DNA from each cell line was treated with SacI, which does not cut BPV-1 DNA, or with a combination of endonucleases EcoRI and BamHI, which each cut BPV-1 at a single site. Southern analysis of the SacI-digested DNA showed that the mutants replicated as monomeric and oligomeric circular plasmids (data not shown). As shown in Fig. 6 for DNA samples digested with EcoRI plus BamHI, the only mutant which consistently exhibited a plasmid copy number lower than that of wt BPV-1 DNA was mutant E67, which contained intact E6 and E7 ORFs but lacked a functional E6/7 splice acceptor signal.

This observation was consistent with the hypothesis that an mRNA which utilizes this splice acceptor signal encodes a factor required for high-copy-number BPV-1 plasmid replication. However, our results are nevertheless inconsistent with the notion that the E67 mutant genome exhibited a low copy number because of the absence of expression of an E6/7 fusion protein. Mutants E62 and E72 were also incapable of expressing a putative E6/7 fusion protein, yet they did not replicate at a low copy number. Isolation of C127 cells transformed by a BPV-1 E5-minus mutant. Investigation of the transforming properties of BPV-1 in several laboratories has indicated that the potent transforming activity of the viral genome in C127 cells is largely dependent on expression of the E5 transforming gene (4, 11, 19, 21). In keeping with these previous results, we found that a frameshift mutation (produced by insertion of an XhoI linker at the BstXI site [25]) in the E5 gene drastically reduced, but did not completely eliminate, the transforming activity of BPV-1 DNA in C127 cells (Fig. 7). The residual transforming activity of the E5 mutant BPV-1 genome is assumed to be due to expression of the E6 transforming protein. Transformed C127 foci induced by the E5-minus mutant were of reduced size compared with wt BPV-1induced foci, and they had an altered macroscopic appearance. It was of interest to know whether BPV-1 replicated normally in the absence of a functional E5 gene in these transformed cells. Previous investigators have reported that BPV-1 E5 mutant genomes either became integrated into the cellular chromosomes or replicated in a rearranged state in C127 cells (11, 21). We established cell lines from five independently arising foci obtained in two separate transfection experiments with E5 mutant viral DNA. By using Southern blot hybridization analysis with a 32P-labelled

IDENTIFICATION AND DEFINITION OF BPV-1 E7 PROTEIN

VOL. 66, 1992

4963

I

mock

E5-minus BPV-1

wt BPV-1

FIG. 7. Transformation of C127 cells by E5-minus BPV-1 genomic DNA. Subconfluent exponential-phase cultures of C127 cells in 5-cm-diameter culture dishes were transfected with 1 jig of BamHI-cleaved p142-6 (wt BPV-1) DNA or E5-minus BPV-1 [25]) or were mock transfected (no DNA) (mock). Transformed foci were visualized 3 weeks after transfection by staining with 1% (wt/vol) methylene blue after fixation of cells with 70% (vol/vol) 2-propanol. Representative plates are shown from one transfection experiment. The transforming activity of E5-minus BPV-1 DNA was calculated to be 1 to 2% compared with that of wt DNA BPV-1 in four separate transfection experiments.

BPV-1 DNA probe, we found that each of these five cell lines in exponential growth phase retained the mutant viral DNA and that it replicated as a multiple-copy plasmid (Fig. 8A). Each of these mutant BPV-1-transformed cell lines had an increased plasmid copy number (two- to fourfold) over the wt BPV-1-transformed cell line, wh.2, which we routinely use as a standard cell line containing a normal viral plasmid copy number (approximately 50 per cell) (Fig. 8B). We conclude that, under our assay conditions, expression of the E5 gene is not required for the establishment of the BPV-1 genome at a high plasmid copy number in C127 cells.

DISCUSSION In this study we have identified a protein encoded by the E7 ORF of BPV-1. Our accumulated biochemical and genetic data are consistent with the conclusion that this E7 protein corresponds to a translation product of the fulllength E7 ORF. Expression of the BPV-1 E7 protein, as well as of the E5 and E6 transforming proteins, was induced in virus-transformed C127 cells under culture conditions which inhibited cell proliferation. These observations suggest that the E5, E6, and E7 genes may be coordinately regulated by cell growth arrest in C127 cell lines. This induction process appears to resemble the natural induction of vegetative BPV-1 DNA synthesis, which takes place in terminally differentiated keratinocytes in virus-producing bovine skin warts (6). Our present data are also consistent with the hypothesis that expression of the E5, E6, and E7 genes is positively regulated by the E2 transcriptional activator protein. We have failed to confirm that an E6/7 fusion protein is expressed in BPV-1-transformed C127 cells. Our data therefore prompt us to challenge the conclusion that the previously identified E6/7 transcript encodes a functional E6/7 fusion protein (29, 35). Although it could be argued that our inability to detect an E6/7 protein could be explained by its

failure to be recognized by either the E6 or the E7 antiserum, by the instability of such a protein, or by a limitation on its expression to a defined stage in the cell cycle, these arguments either are weak or are difficult to test experimentally. The E6/7 and E6/4 mRNAs may represent redundant products of the complex alternative splicing pathways imposed on BPV-1 early region precursor transcripts. An alternative function for the splice acceptor signal located within the E7 ORF is suggested by our observation that a BPV-1 mutant genome lacking this splice acceptor reproducibly exhibited a low plasmid copy number, whereas other mutant genomes also lacking the potential to encode an intact E6/7 protein replicated at a normal copy number. It is therefore possible that an mRNA which utilizes this splice acceptor encodes a viral replication factor (other than E6/7) required for high-copy-number latent viral plasmid replication in proliferating C127 cells. As shown in Fig. 1, two different species of early region transcript which utilized this splice acceptor had the potential to encode truncated or full-length El proteins (types I and II), although fractionation studies indicated that the longer transcript was largely retained in the nuclei of virus-transformed C127 cells (29). We show here that, when tested under controlled experimental conditions, BPV-1 mutant genomes with single mutations which prevented expression of the E5, E6, and E7 proteins did not exhibit a low-copy-number phenotype in C127 cells transformed by these individual mutants. Our data are consistent with results obtained from a study of the in vivo replication properties of BPV-1 in a transient replication assay, in which it was shown that coexpression of full-length El and E2 proteins was necessary and sufficient to initiate viral DNA replication in C127 cells (32). Why have previous investigators occasionally observed low-copy-number or replication-defective phenotypes for E5, E6, and E7 BPV-1 mutants? A possible explanation for the low copy number of certain E6 and E7 mutants could be an indirect inhibitory effect of these mutations on transcrip-

JAREBORG ET AL.

4964

J. VIROL.

were selected for expression of the cotransfected neomycin resistance gene (11, 21), whereas in our study we selected for expression of a BPV-1-transforming gene and observed a viral plasmid copy number somewhat higher than normal. Although significant advances have recently been made in

E5-minus transformants wt

1

E

S

S

2

S

E

3

E

S

4

E

S

AL*

5

S

E

E

~ _*

understanding the mechanism of initiation of BPV-1 DNA synthesis in vitro and in vivo, we nevertheless still lack a coherent picture of the regulation of stable BPV-1 plasmid copy number. In particular it is important to clarify the functions of a number of previously described viral cis replication control elements, including the plasmid maintenance sequences (16) and negative control of replication elements (22), and to understand the role in replication control of a truncated El protein encoded by the N-terminal region of the El ORF (2, 18, 30). Although our present data do not support a role for ES, E6, or E7 in high-copy-number latent BPV-1 plasmid DNA replication in proliferating cells, it remains possible that these proteins are involved directly or indirectly in the induction of vegetative BPV-1 DNA amplification in postmitotic cells.

i; *- #t~~~~~

B Sl

5

-

ACKNOWLEDGMENTS We thank Douglas Lowy, Elliot Androphy, John Schiller, and Daniel DiMaio for providing sera and mutant BPV-1 plasmid DNA. This study was supported by the University of Uppsala and by the Swedish Cancer Society.

4-

E

0

3

-

a.

0

I_

2-

co

1

0-

1

wt

2

3

4

5

E5-minus transformants FIG. 8. Plasmid copy number of the E5-minus mutant BPV-1 (A) Southern blot of E5-minus mutant-transformed cell lines. Total cell DNA from exponentially growing cells was isolated from five independently cloned transformed cell lines picked from separate E5-minus mutant-induced foci. After digestion with restriction endonuclease Sacl (S; no sites in BPV-1) or EcoRI (E; one site in BPV-1), identical amounts of cell DNA from each line were electrophoresed in an 0.8% (wt/vol) agarose gel, blotted onto a nylon membrane, and then hybridized to a radioactively labelled BPV-1 virion DNA probe. Cellular DNA from the wt BPV-1-transformed C127 cell line, wh.2 (6), was used as a control for normal plasmid copy number (wt). (B) Bar chart showing relative copy numbers of wt and E5-minus mutant plasmids in transformed cell lines. The intensities of the radioactive hybridization signals in the Southern blot shown in panel A were measured with a phosphoimager gel scanner (Molecular Dynamics) and were normalized relative to the wt control, which was assigned a value of 1. genome.

tional processing

or

translation of

an

mRNA for

a

down-

stream replication factor, as suggested above for the E7 splice acceptor mutant. Nevertheless, this may not ade-

quately explain why investigators in different laboratories have obtained conflicting results with identical E6 or E7 mutants. Differences in selection procedures or in the properties of C127 cells have most often been invoked to account for such inconsistent results. It may therefore be relevant to note that in those studies in which BPV-1 ES-minus mutants were reported to be replication defective, transfected cells

REFERENCES 1. Androphy, E. J., J. T. Schiller, and D. R. Lowy. 1985. Identification of the protein encoded by the E6 transforming gene of bovine papillomavirus. Science 230:442-445. 2. Berg, L., M. Lusky, A. Stenlund, and M. R. Botchan. 1986. Repression of bovine papillomavirus replication is mediated by a virally encoded trans-acting factor. Cell 46:753-762. 3. Berg, L. J., K. Singh, and M. Botchan. 1986. Complementation of a bovine papilloma virus low-copy-number mutant: evidence for a temporal requirement of the complementing gene. Mol. Cell. Biol. 6:859-869. 4. Burkhardt, A., D. DiMaio, and R. Schlegel. 1987. Genetic and biochemical definition of the bovine papillomavirus E5 transforming protein. EMBO J. 6:2381-2385. 5. Burnett, S., N. Jareborg, and D. DiMaio. 1992. Localization of bovine papillomavirus type 1 E5 protein to transformed basal keratinocytes and permissive differentiated cells in fibropapilloma tissue. Proc. Natl. Acad. Sci. USA 89:5665-5669. 6. Burnett, S., U. Kiessling, and U. Pettersson. 1989. Loss of bovine papillomavirus DNA replication control in growth-arrested transformed cells. J. Virol. 63:2215-2225. 7. Burnett, S., J. Moreno-Lopez, and U. Pettersson. 1988. A novel spontaneous mutation of the bovine papillomavirus-1 genome. Plasmid 20:61-74. 8. Burnett, S., A.-C. Strom, N. Jareborg, A. Alderborn, J. Dillner, J. Moreno-Lopez, U. Pettersson, and U. Kiessling. 1990. Induction of bovine papillomavirus E2 gene expression and early region transcription by cell growth arrest: correlation with viral DNA amplification and evidence for differential promoter induction. J. Virol. 64:5529-5541. 9. Dvoretzky, I., R. Schober, S. K. Chattopadaya, and D. R. Lowy. 1980. A quantitative in vitro focus assay for bovine papilloma virus. Virology 103:369-375. 10. Graham, F. L., and A. J. van der Eb. 1973. A new technique for the assay of infectivity of human adenovirus 5 DNA. Virology 52:456-467. 11. Groff, D. E., and W. D. Lancaster. 1986. Genetic analysis of the 3' early region transformation and replication functions of bovine papillomavirus type 1. Virology 150:221-230. 12. Haugen, T. H., T. P. Cripe, G. D. Ginder, M. Karin, and L. P. Turelk 1987. Trans-activation of an upstream early gene promoter of bovine papillomavirus-1 by a product of the viral E2 gene. EMBO J. 6:145-152.

VOL. 66, 1992

IDENTIFICATION AND DEFINITION OF BPV-1 E7 PROTEIN

13. Hubbert, N. L., J. T. Schiller, D. R. Lowy, and E. J. Androphy. 1988. Bovine papilloma virus-transformed cells contain multiple E2 proteins. Proc. Natl. Acad. Sci. USA 85:5864-5868. 14. Lambert, P. F., C. C. Baker, and P. M. Howley. 1988. The genetics of bovine papillomavirus type 1. Annu. Rev. Genet. 22:235-258. 15. Law, M., D. R. Lowy, I. Dvoretzky, and P. M. Howley. 1981. Mouse cells transformed by bovine papillomavirus contain only extrachromosomal viral DNA sequences. Proc. Natl. Acad. Sci. USA 78:2727-2731. 16. Lusky, M., and M. R. Botchan. 1984. Characterization of the bovine papillomavirus plasmid maintenance sequences. Cell 36:391 401. 17. Lusky, M., and M. R. Botchan. 1985. Genetic analysis of bovine papillomavirus type 1 trans-acting replication factors. J. Virol. 53:955-965. 18. Lusky, M., and M. R. Botchan. 1986. A bovine papillomavirus type 1-encoded modulator function is dispensable for transient viral replication but is required for establishment of the stable plasmid state. J. Virol. 60:729-742. 19. Neary, K., and D. DiMaio. 1989. Open reading frames E6 and E7 of bovine papillomavirus type 1 are both required for full transformation of mouse C127 cells. J. Virol. 63:259-266. 20. Prakash, S. S., B. H. Horwitz, T. Zibello, J. Settleman, and D. DiMaio. 1988. Bovine papillomavirus E2 gene regulates expression of the viral E5 transforming gene. J. Virol. 62:3608-3613. 21. Rabson, M. S., C. Yee, Y. C. Yang, and P. M. Howley. 1986. Bovine papillomavirus type 1 3' early region transformation and plasmid maintenance functions. J. Virol. 60:626-634. 22. Roberts, J. M., and H. Wientraub. 1986. Negative control of DNA replication in composite SV40-bovine papilloma virus plasmids. Cell 46:741-752. 23. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463-5467. 24. Schiller, J. T., W. C. Vass, and D. R. Lowy. 1984. Identification of a second transforming region in bovine papillomavirus DNA.

4965

Proc. Natl. Acad. Sci. USA 81:7880-7884. 25. Schiller, J. T., W. C. Vass, K. H. Vousden, and D. R. Lowy. 1986. E5 open reading frame of bovine papillomavirus type 1 encodes a transforming gene. J. Virol. 57:1-6. 26. Smith, D. B., and K. S. Johnson. 1988. Single-step purification of polypeptides expressed in Escherichia coli as fusions with glutathione S-transferase. Gene 67:31-40. 27. Southern, E. M. 1975. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98:503-517. 28. Spalholz, B. A., S. B. Vande Pol, and P. M. Howley. 1991. Characterization of the cis elements involved in basal and E2-transactivated expression of the bovine papillomavirus P243 promoter. J. Virol. 65:743-753. 29. Stenlund, A., J. Zabielski, H. Ahola, J. Moreno-Lopez, and U. Pettersson. 1985. Messenger RNAs from the transforming region of bovine papilloma virus type I. J. Mol. Biol. 182:541-554. 30. Thorner, L., N. Bucay, J. Choe, and M. Botchan. 1988. The product of the bovine papillomavirus type 1 modulator gene (M) is a phosphoprotein. J. Virol. 62:2474-2482. 31. Towbin, H., T. Staehelin, and J. Gordon. 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. USA 76:4350-4354. 32. Ustav, M., and A. Stenlund. 1991. Transient replication of BPV-1 requires two viral polypeptides encoded by the El and E2 open reading frames. EMBO J. 10:449-457. 33. Vousden, K. H., E. J. Androphy, J. T. Schiller, and D. R. Lowy. 1989. Human papillomaviruses and cervical carcinoma. Cancer Cells 1:43-50. 34. Yang, L., R. Li, I. J. Mohr, R. Clark, and M. R. Botchan. 1991. Activation of BPV-1 replication in vitro by the transcription factor E2. Nature (London) 353:628-632. 35. Yang, Y.-C., H. Okayama, and P. Howley. 1985. Bovine papillomavirus contains multiple transforming genes. Proc. Natl. Acad. Sci. USA 82:1030-1034.