Vol. 66, No. 5
JOURNAL OF VIROLOGY, May 1992, p. 2724-2730
0022-538X/92/052724-07$02.00/0 Copyright © 1992, American Society for Microbiology
Nucleotide 880 Splice Donor Site Required for Efficient Transformation and RNA Accumulation by Human Papillomavirus Type 16 E7 Gene NARASIMHASWAMY S. BELAGULI, MARY M. PATER,
AND
ALAN PATER*
Division of Basic Medical Sciences, Faculty of Medicine, Memorial University of Newfoundland, St. John's, Newfoundland, Canada AIB 3V6 Received 1 October 1991/Accepted 31 January 1992
Mutations within coding sequences of the various human papillomavirus type 16 (HPV-16) genes have been used to demonstrate that the HPV-16 E7 gene is necessary and sufficient for transformation of rodent cells. We now provide evidence that, in addition to E7 coding sequences, a small cis-acting region immediately flanking the 3' end of E7 coding sequences is also required for transformation. This was shown by translation termination linker insertion, progressive deletion analysis, and site-directed mutagenesis. Disruption of the nucleotide (nt) 880 splice donor site within the 3'-flanking region by deletion of as few as 4 nt or substitution of 3 nt totally abolished transformation. Regeneration of the wild-type sequence in a previously transformationincompetent splice site mutant restored transformation. Mutating the wild-type splice donor site to the consensus splice site resulted in a stronger transformation phenotype, while mutating the +2 position of the consensus sequence significantly reduced the frequency of transformation. It was shown with RNase protection assays that the amount of E7 mRNA in transformation-deficient splice site mutants was much lower. Nuclear runoff experiments revealed that there was no change in the rate of synthesis of E7 message in the nt 880 splice site mutant. Furthermore, mutations of HPV-16 sequences indicated that the two other early region splice donor sites have no more than minor roles in transformation and efficient RNA accumulation. These results indicate that the specific integrity of the nt 880 splice donor site is essential for both accumulation of E7 RNA and efficient E7-mediated transformation.
presence of both E7 coding sequences and a short region of immediately 3-flanking sequences. Various mutations within this short region implicate the role of a functional splice donor site. The results also indicate that efficient accumulation of E7 message by posttranscriptional events is determined by the integrity of this splice site.
Human papillomaviruses (HPVs) constitute a group of highly epitheliotropic viruses that are closely associated with various benign and malignant epithelial lesions at distinct anatomical sites. Of the more than 60 different genotypes (6), types 16 and 18 (HPV-16 and -18) are found frequently in anal, penile, and uterine cervical mucosal lesions (36) and are found in premalignant and malignant cervical cancers (41). Recent studies have shown that approaching 100% of cervical cancers and 60% of abnormal cervical smears and biopsies harbor HPV DNA (6, 12, 27, 40). HPV-16 DNA is integrated into the cellular genome in cervical cancers and cancer cell lines (10, 31) and has been shown to express the early region E6 and E7 genes (37). In experimental studies, efficient immortalization of primary human keratinocytes, which requires the presence of both E6 and E7 open reading frames (ORFs), has been achieved (9, 23). HPV-16 DNA can also immortalize primary rodent cells (16), and the HPV-16 E7 ORF can fully transform primary rodent cells in cooperation with activated EJ ras oncogene (20, 33). Previously, the E7-mediated transformation of rodent cells has been investigated with expression of E7 under the control of strong heterologous transcriptional regulatory elements. The E7-expressing plasmids used in some of these investigations consisted of the E7 coding region cloned upstream of heterologous message-processing signals (11, 39). Other studies have shown that of the various HPV-16 ORFs, only E7 ORF is required (20, 33), but the other studies did not test for the requirement of cis-acting HPV-16 sequences. The present report demonstrates that the transformation of baby rat kidney (BRK) cells is dependent on the
*
MATERUILS AND METHODS Plasmids. Plasmid pAl was constructed by ligating the blunt-ended nucleotide (nt) 700 HpaII to nt 4335 Saul fragment of HPV-16 to the Hindlll to nt 2635 EcoRI fragment of pSV2cat. pAl was then generated by a nt 2898 BstXI to nt 3762 NdeI deletion. Plasmid pD7, containing most of the HPV-16 early region under the control of the simian virus 40 (SV40) regulatory region, was constructed as described previously (19). Plasmids pY71, pY72, pY73, pY74, and pY75 (Fig. 1) were generated by deleting sequences from the common nt 3762 NdeI site to the nt 1310 Narl, nt 867 NcoI, nt 884 KpnI, nt 879 PstI, and nt 1178 SspI sites, respectively, from pD7 (see Fig. 3). Plasmids pY7PX, pY7SX, and pY7RX were constructed by inserting translation termination linkers at the nt 685 PvuII, nt 720 SspI, and nt 767 RsaI sites, respectively, in pY71 (Fig. 1). Small deletions initiated from the nt 884 KpnI site of pY71 (see Fig. 3) and large deletion plasmids from the Narl site of pD7 were generated by Bal31 digestion (Fig. 2). All the E7 plus SV40 fragments containing deletions were purified after complete PvuI digestion and ligated to the NdeI-PvuI fragment of pD7. Site-directed mutagenesis was performed by the uracil incorporation method of Kunkel (18). The SspI site at nt 1178 of pY71 was modified to an EcoRI site, and the 313-bp NcoI-to-EcoRI fragment was cloned into Bluescript KS(+) (Stratagene).
Corresponding author. 2724
HPV-16 SPLICE DONOR SITE REQUIRED FOR TRANSFORMATION
VOL. 66, 1992
Transformation Expt.1 Expt.2 Expt.3 ATG
TA 855
E7
pY71 502 562 TTL pY7PX
(El)
4338
112
128
1179
pY73
502 562
1185
pY713'61S
1144
pY713'M17
1083
pY713'Al
1054
pY713'&29
1004
pY713'&16
pY713'&1.2 pY713h2
963
9
954
0
0
0
0
0
112
116
105
0
0
0
116
118
pD7
0
0
0
pD7KB
0
0
0
FIG. 3. Transformation by small deletion mutants. Diagram and
0
o
0
conditions are as described in the legend to Fig. 1, except that small
EP 100 433
107
112
107
109
11S
111
103
114
I11
'
0
103
Transformation Expt.2 Expt3
1312 3763
0
pY7KB7
E3:xp.l
855
96
0
pY7KBS
Mutations were introduced by using either 19- or 29-mer oligonucleotide primers. The mutated Ncc?I-EcoRI fragments were inserted into pY71 or pY7KB1. P'lasmid pD7KB (Fig. 3) was constructed by cleavage of the r)D7 plasmid at its unique KpnI site and was blunt ended by T4 DNA polymerase and then ligated. For the pHKBlL, pHKB5, and pHKB7 plasmids (Fig. 4), HPV-16 DNA cloneed at its unique BamHI site in PBR322 was partially cleav ed with KpnI treated with Bal31, blunt ended by reverse trainscriptase, and ligated. All deletion and site-directed mutatiions were confirmed by sequencing. Cell culture and transformation assays. Sub(confluent BRK cells prepared from 5-day-old rats were transfiected with 5 ,ug of each of the indicated plasmids and activate d EJ ras by the method of Chen and Okayama (4). Two days after transfection the medium was changed to include 2% ftztal calf serum The total number of colonies in each plate was counted at the end of 4 weeks. All cells were maintained in Dulbecco's modified Eagle medium. RNA preparation and analysis. Cos-1 ce lls in 100-mm plates were transfected with 10 ,ug of the indi vcated plasmids and harvested 48 h after transfection. Total cellular RNA was prepared by the method of Chirgwin e t al. (5). Five micrograms of RNA was analyzed by RN ase protection
pY713'A7
113
0
0
liner;mnumr
pY7l
104
0
conEPn; El
(El)
EP
Nm--4=mm6m 43: 138 1312 3763
0
867
TAA
(El)
0
FIG. 1. Transformation of BRK cells by HPV -16 deletion mutants. Thick lines represent HPV-16 sequences, and thin lines represent deletions. Labels are as follows: ATG, E7 start E7 ORF; (El), truncated El ORF; TAA, E7 stop ecodon, , ear y polyadenylation signal; TTL, translation terminatiion bers below lines, HPV nucleotide sequence numb er. Diagrams are not to scale. The total number of colonies in eight 60-mm plates is given.
A
855
o
875
pY72
TAA
o
880
pY74
87
-
502 562
0
767
pY75
ATG
pY71-
o
720Tn
pY7RX
115
o
685 Tn.
pY7SX
Transformation Expt.1 Expt.2 Expt.3
EP
1312 3763
2725
pY7KB1
881-884 I
879-887
7-8 877-893 0
881-884
deletions are indicated as boxes.
assays as described by Kreig and Melton (17). The fragment of from the SV40 nt 5190 StuI to HPV nt 720 SspI was cloned into Bluescript KS(+) at the SmaI site, and the BamHI-cleaved plasmid was then used to synthesize uniformly labelled antisense E7 RNA probe. Similarly, the nt 3072 to 3225 RsaI fragment of SV40 was used for large T-antigen probe. Then, 100,000 cpm each of E7 and T-antigen probes was hybridized with 5 ,ug of total RNA at 56°C for 14 h and digested with 700 U of RNase Ti per ml. RNaseresistant hybrids were resolved on 5% denaturing polyacrylamide-urea sequencing gels. Comparisons given in the text are based on an average of at least three independent experiments in which the E7 signals were normalized with respect to T-antigen signals with laser densitometry. Nuclear runoff assays. Transfections were done as described above. Nuclear runoff assays were performed as described by Ausubel et al. (1). Approximately 5 x 107 nuclei were isolated by Dounce homogenization and incubated with 100 ,uCi of [a-32P]UTP. Approximately 106 cpm of labelled RNA was hybridized for 36 h at 65°C with 5 ,ug (each) of BamHI-cleaved genomic HPV-16 and SV40 DNA immobilized on nitrocellulose filters. Filters were washed thrice in 2x SSC (lx SSC is 0.15 M NaCl plus 0.015 M sodium citrate) at 65°C and treated with 80 ,ug of RNase A.
pY71
RESULTS
Requirement for HPV-16 E7 ORF and immediately 3'flanking sequences for transformation. To examine essential sequences for E7 expression in the context of homologous 3'-flanking sequences, we constructed pY71. This plasmid contains the HPV-16 E7 ORF under the transcriptional control of the SV40 early promoter-enhancer, 455 bp of the E7 3'-flanking region, and the homologous early polyadenylation signal (Fig. 1). Although deleted of the majority of the
E7, 81 E _LI
n i miA IU lUl iz1
99
112
115
93
99
121
pIlKB1
96
113
106
pIIKB5
103
11S
120
0
0
0
876
FIG. 2. Transformation by large deletion mutants. Diagram and conditions are as described in the legend to Fig. 1.
pHKB7
El
E4 E5
E2
AA^AZAA^ I
2
LI
Transformation Expt.2
Expt.1
^Aw
28
23
4
5
3
4
3
3
881-884 879-887
~~~~~~87893
FIG. 4. Transformation by small deletion mutants of pHPV16 plasmid. Labels are as shown in Fig. 3. ORFs are labelled and indicated by lines at the top. Media contained dexamethasone as described previously (28).
2726
J. VIROL.
BELAGULI ET AL.
early region coding sequences, pY71 could transform BRK same frequency as parental pD7 (Fig. 3). Further experiments examining possible effects due to E5 gene products, using a nt 1176 to 3976 deletion mutation, showed no effect on transformation or RNA levels (data not shown). Transfection of pY71 or activated EJ ras alone was not sufficient for transformation. Truncated forms of E7 proteins generated from translation termination linker mutants pY7PX, pY7SX, and pY7RX failed to transform BRK cells (Fig. 1). These results are in agreement with earlier studies
TABLE 1. Transformation by nt 880 splice site-specific mutations
cells at the
and show the essential role of the HPV-16 E7 ORF in transformation of BRK cells without the requirement of other HPV ORFs. In addition to the E7 ORF, 450 bp of the 5' end of the El ORF are also present in the construct pY71. Unexpectedly, deletion of the 3-flanking sequences including the adjacent truncated El ORF in constructs pY72, pY73, and pY74 (Fig. 1) resulted in loss of transformation. It was possible that the severely truncated forms of El may have participated in E7-mediated transformation. This, however, is unlikely, since cotransfection of a separate El expression construct, pAl, with the nontransforming E7expressing construct pY72 failed to give rise to colonies in triplicate experiments in assays which were part of the Fig. 1 experiments. As a control, pAl transfection alone also failed to transform BRK cells. These results indicate that cis-acting 3-flanking sequences are required for E7-mediated transformation. To determine the minimum size of the 3'-flanking region required for transformation, deletion mutation analysis of this region was performed. Progressive unidirectional deletion toward the E7 ORF was undertaken. Deletion of up to 358 bp in pY713'Al.2 did not affect transforming activity, while deletion of an additional 78 bp, which are immediately 3' to the E7 ORF, in pY713'A2 totally abolished transformation (Fig. 2). No change in the frequency of transformation attributable to progressive deletions was observed, implicating the requirement of a single and narrowly defined region. Smaller deletions extending bidirectionally from nt 880 were introduced into pY71. Deletions of 4 bp in pY7KB1, 9 bp in pY7KB5, or 18 bp in pY7KB7 totally abolished transformation (Fig. 3). Deletions in these constructs overlap a splice donor site at nt 880 which shares 6 bp of homology with the CAG GTAAGT consensus splice donor site sequence (for HPV-16 splice sites, see diagrams in references 8 and 34). Although there is an additional potential splice donor site at nt 1301, apparently this site could not functionally substitute for the nt 880 splice site. Effects of nt 880 splice donor site mutations on transformation. To investigate in more detail the role of the splice site at nt 880 in E7-mediated transformation, site-directed mutants were constructed. The plasmid substituting taa in the wild-type sequence was transformation incompetent (Table 1). Converting the wild-type splice site into the consensus splice site, CAG GTAagt, resulted in the appearance of 1.5to 2-fold more colonies at 2 to 3 weeks after transfection. A single-base substitution at the +2 position of the consensus splice donor site in the gAagt construct markedly reduced the frequency of transformation. In addition, regeneration of the wild-type splice site by inserting the tacc sequence into the transformation-incompetent splice site deletion mutant, pY7KB1 (Fig. 3), restored transforming activity to wild-type levels (Table 1). These results clearly indicate that the integrity of the splice donor site located at nt 880 is important for efficient transformation. Effects of splice acceptor sites. Occasionally, cellular splice sites appear to function to introduce cellular cleavage and
Splice site
sequence' GAG Cta CAG CAG CAG
GTACCA aTACCA GTAagt GgAagt GtaccA
Expt 1
144 0 156 28 133
Transformation (no. of colonies)" Expt 2 Expt 3
104 0 121 13 115
122 0 135 16 113
ant sequences are 878 to 886, with wild-type and mutated nucleotides indicated by capital and lowercase letters, respectively. The tacc mutation was introduced into pY7KB1 plasmid, while others were in pY71. b Conditions are as described in the legend to Fig. 1.
polyadenylation sites. However, utilization of the splice donor site located at nt 880 has been observed for most HPV-16 E7 cDNAs analyzed to date (8, 34, 38; our unpublished data), and alternate splice acceptor sites located at positions 2708 or 3357 were utilized for these E7 cDNAs. Neither of the authentic splice acceptor sites are present in the pY7 series constructs. We examined the effect of acceptor sites and found that disruption or deletion of the authentic HPV-16 splice acceptor site of pY71 resulted in apparent use of a cryptic HPV splice. This site was located upstream of the early polyadenylation signal at nt 4213 and approximately at the splice acceptor site-like sequence at nt 4019 (data not shown). Also, it was noted that deletion of the two HPV splice acceptor sites from pD7 in pY71 did not influence transformation (Fig. 3). In addition, a 1,313-bp deletion from pHPV16, which removed both the nt 2708 and 3357 splice acceptor sites, did not alter transformation (data not shown). For an analysis of the converse, i.e., whether the disruption of the splice donor site has the same effect in the presence of authentic splice acceptor sites, splice site mutations were introduced into the plasmids pD7 and pHPV16. The pD7 splice site mutant, pD7KB, was transformation incompetent (Fig. 3). However, in the context of the complete HPV-16 genome, the disruption of the splice donor site significantly reduced but did not completely eliminate the transforming ability (Fig. 4). Effects of splice donor site mutations on mRNA levels. To investigate the molecular mechanism by which integrity of the HPV-16 splice site affects the transformation function of E7, the levels of E7 message generated by various transforming and nontransforming constructs were compared by RNase protection assays. Since BRK cells were not transformed by splice site mutants, assays used RNA from transiently transfected Cos-1 cells. The transforming constructs pY71 and pY75 generated more than eightfold-higher levels of E7 message than the nontransforming constructs pY72 and pY73 (Fig. SA, compare lanes 4 and 7 with lanes S and 6). The pY75 construct, with a deletion of 136 bp of El coding sequences, transformed at wild-type frequency and had only a slight difference in the level of E7 message (Fig. 5A, lane 7). This difference was only 1.4-fold, on the basis of the average value of densitometric scans of autoradiograms for three experiments. The nontransforming construct pY72, which has a large deletion in the 3'-flanking region, produced levels of E7 message comparable to that of a nontransforming construct pY7KB1, which has only a 4-bp deletion at the splice donor site (Fig. SB, compare lanes 5 and 6). This demonstrates that the size of E7 mRNA is not critical. The splice donor site taa mutant produced E7 message at 10-foldlower levels than parental pY71 (Fig. SC, compare lanes 5
HPV-16 SPLICE DONOR SITE REQUIRED FOR TRANSFORMATION
VOL. 66, 1992
1 2 ~3 4 5 6 7 8 9 10
1
B
2727
2 3 4 5 6 7 8 9 404
..K;
;,''.' .g.'T'SJ'' 0;
:
0X'404
309 309
.s ''
;$
^
E7
242
,
;238
3
E7 , E7
242 238 u's
217 199
199
190 190
:
180
1 8C 1 1 60
T
160
147
14 7
C
1
2 3 4 5 6 7 8 9 10 !
$Wit
;522At, ;2¢H2'4i
r2
404
309
0:.!
E7 E7
,; ,
'O.
and 4); this mutant precisely delineated the
sequence essen-
242
tial for both
238
ifying that potential second site mutations are not involved,
217
190
180
T
FIG. 5. RNase protection assays. For all panels the 4 left and 3 right lanes are: T-antigen probe, E7 probe, CaSki, pY71, Cos-1, tRNA, and molecular weight markers, respectively. (A) Lane 5, pY72; lane 6, pY73; lane 7, pY75; (B) lane 5, pY72; lane 6, pY7KB1; (C) lane 5, Cta aTACCA; lane 6, CAG GtaccA; lane 7, CAG GTAagt. Constructs in panel C are as in Table 1. E7- and T-antigenprotected fragments are indicated on the left, and molecular weight marker sizes (in bases) are indicated on the right. CaSki E7 RNA fragment is smaller since it lacks SV40 sequences present in the probe and pY7-type constructs.
160
14 7
transformation
and
mRNA
accumulation.
Ver-
the plasmid in which the wild-type tacc sequence was inserted into the pY7KB1 deletion mutant produced a nearwild-type level of E7 message (Fig. 5C, compare lanes 6 and 4). In addition, the mutant containing a consensus splice donor site sequence had a 1.5-fold-higher level of E7 message (Fig. SC, compare lanes 7 and 4). In experiments in which RSVcat was used as a cotransfection control, the results (data not shown) are in agreement with the five repetitions of the experiments shown in Fig. 5 and 6. The significance of the residual E7 RNA signal for splice site
mutants
in Fig. 5
was
examined. First, the results in Fig.
5 present data with the strongest signals in five repetitiobs of the experiments. Second, analysis for splicing events with a probe spanning exon-intron junctions failed to detect signals for the taa construct, while controls gave readily detectable signal. Third, examination of cytoplasmic RNA, by using the Fig. 5 probe, failed to detect any signal for the pY72 mutant (data not shown). We, therefore, conclude that no detectable
2728
J. VIROL.
BELAGULI ET AL. 1
E 7 .*..
2
3
4
F..*:..
FIG. 6. Nuclear runoff assays. Immobilized HPV-16 (E7) and SV40 (T) DNA were hybridized with RNA from SiHa (lane 1), Cos-1 transfected with pY71 (lane 2), Cos-1 transfected with pY72 (lane 3), and Cos-1 (lane 4).
E7 mRNA signal of donor site mutants reaches the cytoplasm since (i) it is not spliced and/or (ii) it is not transported. These results strongly suggest that both transformation and accumulation of E7 mRNA is dependent on the integrity of the splice donor site located at nt 880. Effects of splice site mutations on E7 mRNA synthesis. The lower levels of E7 message could be due to either lower levels of synthesis or higher rates of degradation of this message. Nuclear runoff assays were performed to compare the synthesis of E7 mRNA in cells transiently transfected with either the transforming pY71 or nontransforming pY72 plasmids. Relative to the internal, T-antigen control, synthesis of E7 was comparable in both pY71- and pY72-transfected nuclei (Fig. 6). Taken together, the results of RNase protection and nuclear runoff assays suggest that the lower levels of E7 message accumulated by splice site mutants are due to higher rates of degradation of E7 message. DISCUSSION Earlier studies on transformation have established that the E7 ORF in cooperation with the activated EJ ras is necessary and sufficient for transformation of rodent cells. The possible role of the E7 3'-flanking sequences in E7-mediated transformation of such cells was not clear, in part because this question was not addressed or the 3'-flanking sequences in E7 expression vectors were replaced by heterologous message-processing signals. In the present report we show that when the E7 ORF is expressed in the context of homologous 3'-flanking sequences, both the E7 coding region and a small region of 3'-flanking sequences are required for transformation. The latter was shown by the deletion and mutation of the E7 3'-flanking sequences contained within the 5' end of the El ORF (Fig. 1). These El sequences are cis-acting, since the El-expressing construct pAl failed to act in trans to restore transformation when cotransfected with the nontransforming construct pY72. In addition, the E7 translation termination linker mutant data (Fig. 1) demonstrated that the truncated El region does not have an independent transforming potential. The fact that deletion of this cis-acting 3'-flanking region resulted in lower levels of E7 mRNA and complete loss of transformation suggested the presence of either a transcriptional enhancer or an mRNA stabilizing element(s) within the 3'-flanking region. The possible presence of a transcriptional enhancer for HPV-16 E7 was clearly eliminated, since comparable levels of RNA synthesis were found in the transforming construct pY71 and the nontransforming construct pY72 by nuclear runoff experiments (Fig. 6). In support of a second possible mechanism, specific 3-flanking sequences of certain mRNAs have been shown to regulate the stability of mRNAs (3, 21). However, no homologies or
similarities with any of the known sequences or structural elements were apparent within the E7 immediately 3'-flanking region. This small region does, however, contain a splice donor site. Our results indicate that it is the integrity of this splice site that determines the normal level of E7 message and transformation competency. This is supported by the following observations. (i) Deletion of as few as 4 nt in pY7KB1 (Fig. 3 and 5B) or mutating 3 nt in the taa mutant of this splice site (Table 1; Fig. 5C) completely eliminated transformation and greatly reduced E7 mRNA levels. Both of these modified splice sites are nonconforming to the GT-AG rule observed for 100% of splice junctions (15, 35). (ii) Insertion of tacc into the splice donor site of transformation-incompetent splice site deletion mutant pY7KB1 restored wild-type sequence and both the transformation phenotype (Table 1) and E7 message (Fig. 5C) to levels comparable to that of the parental wild-type construct pY71. (iii) Mutating the wild-type splice donor site to a consensus splice site in the agt mutant resulted in a 1.5-fold-higher level of E7 message (Fig. 5C), slightly higher frequency of transformation (Table 1), and earlier appearance of transformed colonies. (iv) A single-base mutation at the +2 position within the consensus splice donor site in the gAagt mutant greatly reduced the efficiency of transformation (Table 1). (v) The E7 nt 501 to 863 fragment of the nontransforming plasmid pY72 can transform BRK cells when the messageprocessing signals of SV40 are provided (data not shown). Taken together, these results support the crucial role of the splice donor site at nt 880 in E7 mRNA stability or the failure of unspliced mRNA to be transported to the cytoplasm, resulting in its degradation in nuclei and, thus, the loss of transformation. Our results on the nt 880 splice site are in agreement with the finding that efficient expression of certain genes is dependent on the presence of an intron in the transcription unit (13) and that the presence of functional splice sites can promote accumulation of mRNA (2, 14). Furthermore, the higher frequency of transformation (Table 1), earlier colony formation, and 1.5-fold-higher E7 RNA level (Fig. 5C) observed with the agt construct could be due to a higher efficiency of processing, and thus, the slight increase in total E7 mRNA may be due to an even larger increase in the level of processed, functional cytoplasmic mRNA. The consensus splice donor site present in this construct is fully complementary to the first 9 nt of Ul small nuclear RNA, and the degree of complementarity has been shown to determine the efficiency of splicing (24). Another possible mechanism by which the splice site mutations at nt 880 result in loss of transformation may be through the activation of a cryptic splice site located within the E7 coding region resulting in the production of nonfunctional truncated E7 protein; however, RNase protection assays with a complete E7 coding region probe failed to identify products of such an event (data not shown). Other possible mechanisms whereby splicing affects functional E7 mRNA levels remain to be addressed. First, the unspliced E7 message may fail to undergo correct processing and be retained within the nucleus and, consequently, be rapidly degraded. In addition, splicing has been shown to increase polyadenylation (14, 25), which is known to stabilize mRNAs. The splice site mutations in the context of the whole HPV-16 genome greatly reduced but did not completely eliminate transformation. One factor to consider in comparing the HPV-16- and SV40-based constructs is that HPV-16 but not SV40 promoter-enhancer-driven constructs require the presence of the glucocorticoid hormone dexamethasone
HPV-16 SPLICE DONOR SITE REQUIRED FOR TRANSFORMATION
Voi.. 66, 1992
for transformation (28). The stabilizing effect of glucocorticoid hormones on growth hormone mRNA (26) and phosphoenolpyruvate carboxykinase mRNA (32) have been reported. The same may be occurring for HPV-16, and dexamethasone might have a dual role in the regulation of E7-mediated transformation, one at the transcriptional level and the other at the posttranscriptional level. Another possible explanation could that the other two splice donor sites at nt 226 and 1301 may function additively to substitute for the splice site at nt 880. It should be noted that effects of the nt 226 and 409 splice sites are very small, on the basis of our recent results with mutants in the context of the full-length HPV-16 genome (30). The present results demonstrate that the nt 880 splice donor site has a strong and dominant effect on E7-mediated BRK cell transformation in a mechanism that probably involves E7 mRNA stabilization due to splicing events. ACKNOWLEDGMENTS We thank L. Gissmann and H. zur Hausen for HPV-16 DNA, J. Lilly for technical assistance, A. Mithal for pAl, Y. Shi for assistance with site-directed mutagenesis, and S. Atkins for typing the manuscript. This work was supported in part by grants from the MRC and NCI of Canada.
REFERENCES 1. Ausubel, F. M., R. Brent, F. K. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl. 1990. Current protocols in molecular biology, vol. 1, p. 4.10.1-4.10.9. John Wiley & Sons., Inc., New York. 2. Buchman, A. R., and P. Berg. 1988. Comparison of introndependent and intron-independent gene expression. Mol. Cell. Biol. 8:4395-4405. 3. Casey, J. L., D. M. Koeller, V. C. Ramin, R. D. Klausner, and J. B. Harford. 1989. Iron regulation of transferrin receptor mRNA levels requires iron-responsive elements and a rapid turnover determinant in the 3' untranslated region of the mRNA. EMBO J. 8:3693-3699. 4. Chen, C., and H. Okayama. 1987. High-efficiency transformation of mammalian cells by plasmid DNA. Mol. Cell. Biol.
7:2745-2752. 5. Chirgwin, J. M., A. E. Przybyla, R. C. MacDonald, and W. J. Rutter. 1979. Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 18:5294-
5299. 6. de Villiers, E. M. 1989. Heterogeneity of the human papilloma-
virus group. J. Virol. 63:4898-4903. 7. de Villiers, E. M., A. Schneider, H. Miklaw, U. Papendick, D. Wagner, H. Wesch, J. Wahrendorf, and H. zur Hausen. 1987. Human papillomavirus infections in women with and without abnormal cervical cytology. Lancet ii:703-706. 8. Doorbar, J., A. Parton, K. Hartley, L. Banks, T. Crook, M. Stanley, and L. Crawford. 1990. Detection of novel splicing patterns in a HPV 16 containing keratinocyte cell line. Virology
178:254-262. 9. Durst, M., T. Dzarlieva-Petrusevska, P. Boukamp, N. E. Fusenig, and L. Gissman. 1987. Molecular and cytogenetic analysis of immortalized human primary keratinocytes obtained after transfection with human papillomavirus type 16 DNA. Onco-
gene 1:251-256.
10. Durst, M., L. Gissmann, H. Ikenberg, and H. zur Hausen. 1983. A papillomavirus DNA from a cervical carcinoma and its presence in cervical cancer biopsy samples from different geo-
graphic areas. Proc. Natl. Acad. Sci. USA 80:3812-3815.
11. Edmonds, C., and K. H. Vousden. 1989. A point mutational analysis of human papillomavirus type 16 E7 protein. J. Virol.
63:2650-2656. 12. Gergley, L., J. Czegledy, and Z. Hernady. 1987. Letter. Lancet
ii:513.
2729
13. Goff, S. P., and P. Berg. 1979. Construction, propagation and expression of simian virus 40 recombinant genomes containing the Escherichia coli gene for thymidine kinase and a Saccharomyces cerevisiae gene for tyrosine transfer RNA. J. Mol. Biol. 133:359-383. 14. Huang, M. T. F., and C. M. Gorman. 1990. Intervening sequences increase efficiency of RNA 3' processing and accumulation of cytoplasmic RNA. Nucleic Acids Res. 18:937-947. 15. Jackson, I. J. 1991. A reappraisal of non-consensus mRNA splice sites. Nucleic Acids Res. 19:3795-3798. 16. Kanda, T., S. Watanabe, and K. Yoshiike. 1988. Immortalization of primary rat cells by human papillomavirus type 16 subgenomic DNA fragments controlled by the SV40 promoter. Virology 165:321-325. 17. Kreig, P. A., and D. A. Melton. 1987. In vitro RNA synthesis with sp6 RNA polymerase. Methods Enzymol. 155:397-415. 18. Kunkel, T. A. 1985. Rapid and efficient site-specific mutagenesis without phenotypic selection. Proc. Natl. Acad. Sci. USA 82: 488-492. 19. Marshall, T., A. Pater, and M. M. Pater. 1989. Trans-regulation and differential cell specificity of human papillomavirus types 16, 18, and 11 cis-acting elements. J. Med. Virol. 29:115-126. 20. Matlashewski, G., J. Schneider, L. Banks, N. Jones, A. Murray, and L. Crawford. 1987. Human papillomavirus type 16 DNA co-operates with activated ras in transforming primary cells. EMBO J. 6:1741-1746. 21. Meijlink, F., T. Curran, A. D. Miller, and I. Verma. 1985. Removal of a 67-base-pair sequence in the noncoding region of protooncogene fos converts it to a transforming gene. Proc. Natl. Acad. Sci. USA 82:4987-4991. 22. Mullner, E. W., and L. C. Kuhn. 1988. A stem-loop in the 3' untranslated region mediates iron-dependent regulation of transferrin receptor mRNA stability in the cytoplasm. Cell 53:815825. 23. Munger, K., W. C. Phelps, V. Bubb, P. M. Howley, and R. Schlegel. 1989. The E6 and E7 genes of the human papillomavirus type 16 together are necessary and sufficient for transformation of primary human keratinocytes. J. Virol. 63:4417-4421. 24. Nelson, K. K., and M. R. Green. 1990. Mechanism for cryptic splice site activation during pre-mRNA splicing. Proc. Natl. Acad. Sci. USA 87:6253-6257. 25. Niwa, M., S. D. Rose, and S. M. Berget. 1990. In vitro polyadenylation is stimulated by the presence of an upstream intron. Genes Dev. 4:1552-1559. 26. Paek, I., and R. Axel. 1987. Glucocorticoids enhance stability of human growth hormone mRNA. Mol. Cell. Biol. 7:1496-1507. 27. Pater, M. M., J. Dunne, G. Hogan, P. Ghatage, and A. Pater. 1986. Human papillomavirus types 16 and 18 sequences in early cervical neoplasia. Virology 155:13-18. 28. Pater, M. M., G. A. Hughes, D. E. Hyslop, H. Nakshatri, and A. Pater. 1988. Glucocorticoid-dependent oncogenic transformation by type 16 but not type 11 human papilloma virus DNA. Nature (London) 335:832-835. 29. Pater, M. M., T. Marshall, H. Nakshatri, and A. Pater. 1990. Glucocorticoid-dependent transformation of primary cells by human papillomavirus type 16 DNA. UCLA Symp. Mol. Cell. Biol. 126:313-321. 30. Pater, M. M., H. Nakshatri, C. Kisaka, and A. Pater. 1992. The first 124 nucleotides of the E7 coding sequences of HPV16 can render the HPV11 genome transformation competent. Virology 186:348-351. 31. Pater, M. M., and A. Pater. 1985. Human papillomavirus type 16 and 18 sequences in carcinoma cell lines of the cervix. Virology 145:313-318. 32. Petersen, D. D., S. R. Koch, and D. K. Granner. 1989. 3' noncoding region of phosphoenolpyruvate carboxykinase mRNA contains a glucocorticoid-responsive mRNA-stabilizing element. Proc. Natl. Acad. Sci. USA 86:7800-7804. 33. Phelps, W. C., C. L. Yee, K. Munger, and P. M. Howley. 1988. The human papillomavirus type 16 E7 gene encodes transactivation and transformation functions similar to those of adenovirus EIA. Cell 53:539-547. 34. Rohlfs, M., S. Winkenbach, S. Meyer, T. Rupp, and M. Durst.
2730
35.
36. 37.
38.
BELAGULI ET AL.
1991. Viral transcription in human keratinocyte cell lines immortalized by human papillomavirus type-16. Virology 183:331342. Senapathy, P., M. B. Shapiro, and N. L. Harris. 1990. Splice junctions, branch site points, and exons: sequence statistics, identification, and applications to genomic project. Methods Enzymol. 183:252-278. Shah, K. V., and P. M. Howley. 1990. Papillomaviruses, p. 1651-1676. In B. N. Fields and D. M. Knipe (ed.), Virology. Raven Press, Ltd., New York. Smotkin, D., and F. 0. Wettstein. 1986. Transcription of human papillomavirus type 16 early genes in a cervical cancer and a cancer-derived cell line and identification of the E7 protein. Proc. Natl. Acad. Sci. USA 83:4680-4684. Tanaka, A., T. Noda, H. Yajima, M. Hatanaka, and Y. Ito. 1989.
J. VIROL.
Identification of a transforming gene of human papillomavirus type 16. J. Virol. 63:1465-1469. 39. Watanabe, S., T. Kanda, H. Sato, A. Furuno, and K. Yoshiike. 1990. Mutational analysis of human papillomavirus type 16 E7 functions. J. Virol. 64:207-214. 40. Young, L. S., I. S. Bevan, M. A. Johnson, P. I. Blomfield, T. Brombridge, N. J. Maitland, and C. B. J. Woodman. 1989. The polymerase chain reaction: a new epidemiological tool for investigating cervical human papillomavirus infection. Br. Med. J. 298:14-18. 41. zur Hausen, H., and A. Schneider. 1987. The role of papillomaviruses in anogenital cancer, p. 245-263. In N. P. Salzman and P. M. Howley (ed.), The Papovaviridae: the papillomaviruses. Plenum Publishing Corp., New York.
JOURNAL OF VIROLOGY, May 1992, p. 2724-2730
0022-538X/92/052724-07$02.00/0 Copyright © 1992, American Society for Microbiology
Nucleotide 880 Splice Donor Site Required for Efficient Transformation and RNA Accumulation by Human Papillomavirus Type 16 E7 Gene NARASIMHASWAMY S. BELAGULI, MARY M. PATER,
AND
ALAN PATER*
Division of Basic Medical Sciences, Faculty of Medicine, Memorial University of Newfoundland, St. John's, Newfoundland, Canada AIB 3V6 Received 1 October 1991/Accepted 31 January 1992
Mutations within coding sequences of the various human papillomavirus type 16 (HPV-16) genes have been used to demonstrate that the HPV-16 E7 gene is necessary and sufficient for transformation of rodent cells. We now provide evidence that, in addition to E7 coding sequences, a small cis-acting region immediately flanking the 3' end of E7 coding sequences is also required for transformation. This was shown by translation termination linker insertion, progressive deletion analysis, and site-directed mutagenesis. Disruption of the nucleotide (nt) 880 splice donor site within the 3'-flanking region by deletion of as few as 4 nt or substitution of 3 nt totally abolished transformation. Regeneration of the wild-type sequence in a previously transformationincompetent splice site mutant restored transformation. Mutating the wild-type splice donor site to the consensus splice site resulted in a stronger transformation phenotype, while mutating the +2 position of the consensus sequence significantly reduced the frequency of transformation. It was shown with RNase protection assays that the amount of E7 mRNA in transformation-deficient splice site mutants was much lower. Nuclear runoff experiments revealed that there was no change in the rate of synthesis of E7 message in the nt 880 splice site mutant. Furthermore, mutations of HPV-16 sequences indicated that the two other early region splice donor sites have no more than minor roles in transformation and efficient RNA accumulation. These results indicate that the specific integrity of the nt 880 splice donor site is essential for both accumulation of E7 RNA and efficient E7-mediated transformation.
presence of both E7 coding sequences and a short region of immediately 3-flanking sequences. Various mutations within this short region implicate the role of a functional splice donor site. The results also indicate that efficient accumulation of E7 message by posttranscriptional events is determined by the integrity of this splice site.
Human papillomaviruses (HPVs) constitute a group of highly epitheliotropic viruses that are closely associated with various benign and malignant epithelial lesions at distinct anatomical sites. Of the more than 60 different genotypes (6), types 16 and 18 (HPV-16 and -18) are found frequently in anal, penile, and uterine cervical mucosal lesions (36) and are found in premalignant and malignant cervical cancers (41). Recent studies have shown that approaching 100% of cervical cancers and 60% of abnormal cervical smears and biopsies harbor HPV DNA (6, 12, 27, 40). HPV-16 DNA is integrated into the cellular genome in cervical cancers and cancer cell lines (10, 31) and has been shown to express the early region E6 and E7 genes (37). In experimental studies, efficient immortalization of primary human keratinocytes, which requires the presence of both E6 and E7 open reading frames (ORFs), has been achieved (9, 23). HPV-16 DNA can also immortalize primary rodent cells (16), and the HPV-16 E7 ORF can fully transform primary rodent cells in cooperation with activated EJ ras oncogene (20, 33). Previously, the E7-mediated transformation of rodent cells has been investigated with expression of E7 under the control of strong heterologous transcriptional regulatory elements. The E7-expressing plasmids used in some of these investigations consisted of the E7 coding region cloned upstream of heterologous message-processing signals (11, 39). Other studies have shown that of the various HPV-16 ORFs, only E7 ORF is required (20, 33), but the other studies did not test for the requirement of cis-acting HPV-16 sequences. The present report demonstrates that the transformation of baby rat kidney (BRK) cells is dependent on the
*
MATERUILS AND METHODS Plasmids. Plasmid pAl was constructed by ligating the blunt-ended nucleotide (nt) 700 HpaII to nt 4335 Saul fragment of HPV-16 to the Hindlll to nt 2635 EcoRI fragment of pSV2cat. pAl was then generated by a nt 2898 BstXI to nt 3762 NdeI deletion. Plasmid pD7, containing most of the HPV-16 early region under the control of the simian virus 40 (SV40) regulatory region, was constructed as described previously (19). Plasmids pY71, pY72, pY73, pY74, and pY75 (Fig. 1) were generated by deleting sequences from the common nt 3762 NdeI site to the nt 1310 Narl, nt 867 NcoI, nt 884 KpnI, nt 879 PstI, and nt 1178 SspI sites, respectively, from pD7 (see Fig. 3). Plasmids pY7PX, pY7SX, and pY7RX were constructed by inserting translation termination linkers at the nt 685 PvuII, nt 720 SspI, and nt 767 RsaI sites, respectively, in pY71 (Fig. 1). Small deletions initiated from the nt 884 KpnI site of pY71 (see Fig. 3) and large deletion plasmids from the Narl site of pD7 were generated by Bal31 digestion (Fig. 2). All the E7 plus SV40 fragments containing deletions were purified after complete PvuI digestion and ligated to the NdeI-PvuI fragment of pD7. Site-directed mutagenesis was performed by the uracil incorporation method of Kunkel (18). The SspI site at nt 1178 of pY71 was modified to an EcoRI site, and the 313-bp NcoI-to-EcoRI fragment was cloned into Bluescript KS(+) (Stratagene).
Corresponding author. 2724
HPV-16 SPLICE DONOR SITE REQUIRED FOR TRANSFORMATION
VOL. 66, 1992
Transformation Expt.1 Expt.2 Expt.3 ATG
TA 855
E7
pY71 502 562 TTL pY7PX
(El)
4338
112
128
1179
pY73
502 562
1185
pY713'61S
1144
pY713'M17
1083
pY713'Al
1054
pY713'&29
1004
pY713'&16
pY713'&1.2 pY713h2
963
9
954
0
0
0
0
0
112
116
105
0
0
0
116
118
pD7
0
0
0
pD7KB
0
0
0
FIG. 3. Transformation by small deletion mutants. Diagram and
0
o
0
conditions are as described in the legend to Fig. 1, except that small
EP 100 433
107
112
107
109
11S
111
103
114
I11
'
0
103
Transformation Expt.2 Expt3
1312 3763
0
pY7KB7
E3:xp.l
855
96
0
pY7KBS
Mutations were introduced by using either 19- or 29-mer oligonucleotide primers. The mutated Ncc?I-EcoRI fragments were inserted into pY71 or pY7KB1. P'lasmid pD7KB (Fig. 3) was constructed by cleavage of the r)D7 plasmid at its unique KpnI site and was blunt ended by T4 DNA polymerase and then ligated. For the pHKBlL, pHKB5, and pHKB7 plasmids (Fig. 4), HPV-16 DNA cloneed at its unique BamHI site in PBR322 was partially cleav ed with KpnI treated with Bal31, blunt ended by reverse trainscriptase, and ligated. All deletion and site-directed mutatiions were confirmed by sequencing. Cell culture and transformation assays. Sub(confluent BRK cells prepared from 5-day-old rats were transfiected with 5 ,ug of each of the indicated plasmids and activate d EJ ras by the method of Chen and Okayama (4). Two days after transfection the medium was changed to include 2% ftztal calf serum The total number of colonies in each plate was counted at the end of 4 weeks. All cells were maintained in Dulbecco's modified Eagle medium. RNA preparation and analysis. Cos-1 ce lls in 100-mm plates were transfected with 10 ,ug of the indi vcated plasmids and harvested 48 h after transfection. Total cellular RNA was prepared by the method of Chirgwin e t al. (5). Five micrograms of RNA was analyzed by RN ase protection
pY713'A7
113
0
0
liner;mnumr
pY7l
104
0
conEPn; El
(El)
EP
Nm--4=mm6m 43: 138 1312 3763
0
867
TAA
(El)
0
FIG. 1. Transformation of BRK cells by HPV -16 deletion mutants. Thick lines represent HPV-16 sequences, and thin lines represent deletions. Labels are as follows: ATG, E7 start E7 ORF; (El), truncated El ORF; TAA, E7 stop ecodon, , ear y polyadenylation signal; TTL, translation terminatiion bers below lines, HPV nucleotide sequence numb er. Diagrams are not to scale. The total number of colonies in eight 60-mm plates is given.
A
855
o
875
pY72
TAA
o
880
pY74
87
-
502 562
0
767
pY75
ATG
pY71-
o
720Tn
pY7RX
115
o
685 Tn.
pY7SX
Transformation Expt.1 Expt.2 Expt.3
EP
1312 3763
2725
pY7KB1
881-884 I
879-887
7-8 877-893 0
881-884
deletions are indicated as boxes.
assays as described by Kreig and Melton (17). The fragment of from the SV40 nt 5190 StuI to HPV nt 720 SspI was cloned into Bluescript KS(+) at the SmaI site, and the BamHI-cleaved plasmid was then used to synthesize uniformly labelled antisense E7 RNA probe. Similarly, the nt 3072 to 3225 RsaI fragment of SV40 was used for large T-antigen probe. Then, 100,000 cpm each of E7 and T-antigen probes was hybridized with 5 ,ug of total RNA at 56°C for 14 h and digested with 700 U of RNase Ti per ml. RNaseresistant hybrids were resolved on 5% denaturing polyacrylamide-urea sequencing gels. Comparisons given in the text are based on an average of at least three independent experiments in which the E7 signals were normalized with respect to T-antigen signals with laser densitometry. Nuclear runoff assays. Transfections were done as described above. Nuclear runoff assays were performed as described by Ausubel et al. (1). Approximately 5 x 107 nuclei were isolated by Dounce homogenization and incubated with 100 ,uCi of [a-32P]UTP. Approximately 106 cpm of labelled RNA was hybridized for 36 h at 65°C with 5 ,ug (each) of BamHI-cleaved genomic HPV-16 and SV40 DNA immobilized on nitrocellulose filters. Filters were washed thrice in 2x SSC (lx SSC is 0.15 M NaCl plus 0.015 M sodium citrate) at 65°C and treated with 80 ,ug of RNase A.
pY71
RESULTS
Requirement for HPV-16 E7 ORF and immediately 3'flanking sequences for transformation. To examine essential sequences for E7 expression in the context of homologous 3'-flanking sequences, we constructed pY71. This plasmid contains the HPV-16 E7 ORF under the transcriptional control of the SV40 early promoter-enhancer, 455 bp of the E7 3'-flanking region, and the homologous early polyadenylation signal (Fig. 1). Although deleted of the majority of the
E7, 81 E _LI
n i miA IU lUl iz1
99
112
115
93
99
121
pIlKB1
96
113
106
pIIKB5
103
11S
120
0
0
0
876
FIG. 2. Transformation by large deletion mutants. Diagram and conditions are as described in the legend to Fig. 1.
pHKB7
El
E4 E5
E2
AA^AZAA^ I
2
LI
Transformation Expt.2
Expt.1
^Aw
28
23
4
5
3
4
3
3
881-884 879-887
~~~~~~87893
FIG. 4. Transformation by small deletion mutants of pHPV16 plasmid. Labels are as shown in Fig. 3. ORFs are labelled and indicated by lines at the top. Media contained dexamethasone as described previously (28).
2726
J. VIROL.
BELAGULI ET AL.
early region coding sequences, pY71 could transform BRK same frequency as parental pD7 (Fig. 3). Further experiments examining possible effects due to E5 gene products, using a nt 1176 to 3976 deletion mutation, showed no effect on transformation or RNA levels (data not shown). Transfection of pY71 or activated EJ ras alone was not sufficient for transformation. Truncated forms of E7 proteins generated from translation termination linker mutants pY7PX, pY7SX, and pY7RX failed to transform BRK cells (Fig. 1). These results are in agreement with earlier studies
TABLE 1. Transformation by nt 880 splice site-specific mutations
cells at the
and show the essential role of the HPV-16 E7 ORF in transformation of BRK cells without the requirement of other HPV ORFs. In addition to the E7 ORF, 450 bp of the 5' end of the El ORF are also present in the construct pY71. Unexpectedly, deletion of the 3-flanking sequences including the adjacent truncated El ORF in constructs pY72, pY73, and pY74 (Fig. 1) resulted in loss of transformation. It was possible that the severely truncated forms of El may have participated in E7-mediated transformation. This, however, is unlikely, since cotransfection of a separate El expression construct, pAl, with the nontransforming E7expressing construct pY72 failed to give rise to colonies in triplicate experiments in assays which were part of the Fig. 1 experiments. As a control, pAl transfection alone also failed to transform BRK cells. These results indicate that cis-acting 3-flanking sequences are required for E7-mediated transformation. To determine the minimum size of the 3'-flanking region required for transformation, deletion mutation analysis of this region was performed. Progressive unidirectional deletion toward the E7 ORF was undertaken. Deletion of up to 358 bp in pY713'Al.2 did not affect transforming activity, while deletion of an additional 78 bp, which are immediately 3' to the E7 ORF, in pY713'A2 totally abolished transformation (Fig. 2). No change in the frequency of transformation attributable to progressive deletions was observed, implicating the requirement of a single and narrowly defined region. Smaller deletions extending bidirectionally from nt 880 were introduced into pY71. Deletions of 4 bp in pY7KB1, 9 bp in pY7KB5, or 18 bp in pY7KB7 totally abolished transformation (Fig. 3). Deletions in these constructs overlap a splice donor site at nt 880 which shares 6 bp of homology with the CAG GTAAGT consensus splice donor site sequence (for HPV-16 splice sites, see diagrams in references 8 and 34). Although there is an additional potential splice donor site at nt 1301, apparently this site could not functionally substitute for the nt 880 splice site. Effects of nt 880 splice donor site mutations on transformation. To investigate in more detail the role of the splice site at nt 880 in E7-mediated transformation, site-directed mutants were constructed. The plasmid substituting taa in the wild-type sequence was transformation incompetent (Table 1). Converting the wild-type splice site into the consensus splice site, CAG GTAagt, resulted in the appearance of 1.5to 2-fold more colonies at 2 to 3 weeks after transfection. A single-base substitution at the +2 position of the consensus splice donor site in the gAagt construct markedly reduced the frequency of transformation. In addition, regeneration of the wild-type splice site by inserting the tacc sequence into the transformation-incompetent splice site deletion mutant, pY7KB1 (Fig. 3), restored transforming activity to wild-type levels (Table 1). These results clearly indicate that the integrity of the splice donor site located at nt 880 is important for efficient transformation. Effects of splice acceptor sites. Occasionally, cellular splice sites appear to function to introduce cellular cleavage and
Splice site
sequence' GAG Cta CAG CAG CAG
GTACCA aTACCA GTAagt GgAagt GtaccA
Expt 1
144 0 156 28 133
Transformation (no. of colonies)" Expt 2 Expt 3
104 0 121 13 115
122 0 135 16 113
ant sequences are 878 to 886, with wild-type and mutated nucleotides indicated by capital and lowercase letters, respectively. The tacc mutation was introduced into pY7KB1 plasmid, while others were in pY71. b Conditions are as described in the legend to Fig. 1.
polyadenylation sites. However, utilization of the splice donor site located at nt 880 has been observed for most HPV-16 E7 cDNAs analyzed to date (8, 34, 38; our unpublished data), and alternate splice acceptor sites located at positions 2708 or 3357 were utilized for these E7 cDNAs. Neither of the authentic splice acceptor sites are present in the pY7 series constructs. We examined the effect of acceptor sites and found that disruption or deletion of the authentic HPV-16 splice acceptor site of pY71 resulted in apparent use of a cryptic HPV splice. This site was located upstream of the early polyadenylation signal at nt 4213 and approximately at the splice acceptor site-like sequence at nt 4019 (data not shown). Also, it was noted that deletion of the two HPV splice acceptor sites from pD7 in pY71 did not influence transformation (Fig. 3). In addition, a 1,313-bp deletion from pHPV16, which removed both the nt 2708 and 3357 splice acceptor sites, did not alter transformation (data not shown). For an analysis of the converse, i.e., whether the disruption of the splice donor site has the same effect in the presence of authentic splice acceptor sites, splice site mutations were introduced into the plasmids pD7 and pHPV16. The pD7 splice site mutant, pD7KB, was transformation incompetent (Fig. 3). However, in the context of the complete HPV-16 genome, the disruption of the splice donor site significantly reduced but did not completely eliminate the transforming ability (Fig. 4). Effects of splice donor site mutations on mRNA levels. To investigate the molecular mechanism by which integrity of the HPV-16 splice site affects the transformation function of E7, the levels of E7 message generated by various transforming and nontransforming constructs were compared by RNase protection assays. Since BRK cells were not transformed by splice site mutants, assays used RNA from transiently transfected Cos-1 cells. The transforming constructs pY71 and pY75 generated more than eightfold-higher levels of E7 message than the nontransforming constructs pY72 and pY73 (Fig. SA, compare lanes 4 and 7 with lanes S and 6). The pY75 construct, with a deletion of 136 bp of El coding sequences, transformed at wild-type frequency and had only a slight difference in the level of E7 message (Fig. 5A, lane 7). This difference was only 1.4-fold, on the basis of the average value of densitometric scans of autoradiograms for three experiments. The nontransforming construct pY72, which has a large deletion in the 3'-flanking region, produced levels of E7 message comparable to that of a nontransforming construct pY7KB1, which has only a 4-bp deletion at the splice donor site (Fig. SB, compare lanes 5 and 6). This demonstrates that the size of E7 mRNA is not critical. The splice donor site taa mutant produced E7 message at 10-foldlower levels than parental pY71 (Fig. SC, compare lanes 5
HPV-16 SPLICE DONOR SITE REQUIRED FOR TRANSFORMATION
VOL. 66, 1992
1 2 ~3 4 5 6 7 8 9 10
1
B
2727
2 3 4 5 6 7 8 9 404
..K;
;,''.' .g.'T'SJ'' 0;
:
0X'404
309 309
.s ''
;$
^
E7
242
,
;238
3
E7 , E7
242 238 u's
217 199
199
190 190
:
180
1 8C 1 1 60
T
160
147
14 7
C
1
2 3 4 5 6 7 8 9 10 !
$Wit
;522At, ;2¢H2'4i
r2
404
309
0:.!
E7 E7
,; ,
'O.
and 4); this mutant precisely delineated the
sequence essen-
242
tial for both
238
ifying that potential second site mutations are not involved,
217
190
180
T
FIG. 5. RNase protection assays. For all panels the 4 left and 3 right lanes are: T-antigen probe, E7 probe, CaSki, pY71, Cos-1, tRNA, and molecular weight markers, respectively. (A) Lane 5, pY72; lane 6, pY73; lane 7, pY75; (B) lane 5, pY72; lane 6, pY7KB1; (C) lane 5, Cta aTACCA; lane 6, CAG GtaccA; lane 7, CAG GTAagt. Constructs in panel C are as in Table 1. E7- and T-antigenprotected fragments are indicated on the left, and molecular weight marker sizes (in bases) are indicated on the right. CaSki E7 RNA fragment is smaller since it lacks SV40 sequences present in the probe and pY7-type constructs.
160
14 7
transformation
and
mRNA
accumulation.
Ver-
the plasmid in which the wild-type tacc sequence was inserted into the pY7KB1 deletion mutant produced a nearwild-type level of E7 message (Fig. 5C, compare lanes 6 and 4). In addition, the mutant containing a consensus splice donor site sequence had a 1.5-fold-higher level of E7 message (Fig. SC, compare lanes 7 and 4). In experiments in which RSVcat was used as a cotransfection control, the results (data not shown) are in agreement with the five repetitions of the experiments shown in Fig. 5 and 6. The significance of the residual E7 RNA signal for splice site
mutants
in Fig. 5
was
examined. First, the results in Fig.
5 present data with the strongest signals in five repetitiobs of the experiments. Second, analysis for splicing events with a probe spanning exon-intron junctions failed to detect signals for the taa construct, while controls gave readily detectable signal. Third, examination of cytoplasmic RNA, by using the Fig. 5 probe, failed to detect any signal for the pY72 mutant (data not shown). We, therefore, conclude that no detectable
2728
J. VIROL.
BELAGULI ET AL. 1
E 7 .*..
2
3
4
F..*:..
FIG. 6. Nuclear runoff assays. Immobilized HPV-16 (E7) and SV40 (T) DNA were hybridized with RNA from SiHa (lane 1), Cos-1 transfected with pY71 (lane 2), Cos-1 transfected with pY72 (lane 3), and Cos-1 (lane 4).
E7 mRNA signal of donor site mutants reaches the cytoplasm since (i) it is not spliced and/or (ii) it is not transported. These results strongly suggest that both transformation and accumulation of E7 mRNA is dependent on the integrity of the splice donor site located at nt 880. Effects of splice site mutations on E7 mRNA synthesis. The lower levels of E7 message could be due to either lower levels of synthesis or higher rates of degradation of this message. Nuclear runoff assays were performed to compare the synthesis of E7 mRNA in cells transiently transfected with either the transforming pY71 or nontransforming pY72 plasmids. Relative to the internal, T-antigen control, synthesis of E7 was comparable in both pY71- and pY72-transfected nuclei (Fig. 6). Taken together, the results of RNase protection and nuclear runoff assays suggest that the lower levels of E7 message accumulated by splice site mutants are due to higher rates of degradation of E7 message. DISCUSSION Earlier studies on transformation have established that the E7 ORF in cooperation with the activated EJ ras is necessary and sufficient for transformation of rodent cells. The possible role of the E7 3'-flanking sequences in E7-mediated transformation of such cells was not clear, in part because this question was not addressed or the 3'-flanking sequences in E7 expression vectors were replaced by heterologous message-processing signals. In the present report we show that when the E7 ORF is expressed in the context of homologous 3'-flanking sequences, both the E7 coding region and a small region of 3'-flanking sequences are required for transformation. The latter was shown by the deletion and mutation of the E7 3'-flanking sequences contained within the 5' end of the El ORF (Fig. 1). These El sequences are cis-acting, since the El-expressing construct pAl failed to act in trans to restore transformation when cotransfected with the nontransforming construct pY72. In addition, the E7 translation termination linker mutant data (Fig. 1) demonstrated that the truncated El region does not have an independent transforming potential. The fact that deletion of this cis-acting 3'-flanking region resulted in lower levels of E7 mRNA and complete loss of transformation suggested the presence of either a transcriptional enhancer or an mRNA stabilizing element(s) within the 3'-flanking region. The possible presence of a transcriptional enhancer for HPV-16 E7 was clearly eliminated, since comparable levels of RNA synthesis were found in the transforming construct pY71 and the nontransforming construct pY72 by nuclear runoff experiments (Fig. 6). In support of a second possible mechanism, specific 3-flanking sequences of certain mRNAs have been shown to regulate the stability of mRNAs (3, 21). However, no homologies or
similarities with any of the known sequences or structural elements were apparent within the E7 immediately 3'-flanking region. This small region does, however, contain a splice donor site. Our results indicate that it is the integrity of this splice site that determines the normal level of E7 message and transformation competency. This is supported by the following observations. (i) Deletion of as few as 4 nt in pY7KB1 (Fig. 3 and 5B) or mutating 3 nt in the taa mutant of this splice site (Table 1; Fig. 5C) completely eliminated transformation and greatly reduced E7 mRNA levels. Both of these modified splice sites are nonconforming to the GT-AG rule observed for 100% of splice junctions (15, 35). (ii) Insertion of tacc into the splice donor site of transformation-incompetent splice site deletion mutant pY7KB1 restored wild-type sequence and both the transformation phenotype (Table 1) and E7 message (Fig. 5C) to levels comparable to that of the parental wild-type construct pY71. (iii) Mutating the wild-type splice donor site to a consensus splice site in the agt mutant resulted in a 1.5-fold-higher level of E7 message (Fig. 5C), slightly higher frequency of transformation (Table 1), and earlier appearance of transformed colonies. (iv) A single-base mutation at the +2 position within the consensus splice donor site in the gAagt mutant greatly reduced the efficiency of transformation (Table 1). (v) The E7 nt 501 to 863 fragment of the nontransforming plasmid pY72 can transform BRK cells when the messageprocessing signals of SV40 are provided (data not shown). Taken together, these results support the crucial role of the splice donor site at nt 880 in E7 mRNA stability or the failure of unspliced mRNA to be transported to the cytoplasm, resulting in its degradation in nuclei and, thus, the loss of transformation. Our results on the nt 880 splice site are in agreement with the finding that efficient expression of certain genes is dependent on the presence of an intron in the transcription unit (13) and that the presence of functional splice sites can promote accumulation of mRNA (2, 14). Furthermore, the higher frequency of transformation (Table 1), earlier colony formation, and 1.5-fold-higher E7 RNA level (Fig. 5C) observed with the agt construct could be due to a higher efficiency of processing, and thus, the slight increase in total E7 mRNA may be due to an even larger increase in the level of processed, functional cytoplasmic mRNA. The consensus splice donor site present in this construct is fully complementary to the first 9 nt of Ul small nuclear RNA, and the degree of complementarity has been shown to determine the efficiency of splicing (24). Another possible mechanism by which the splice site mutations at nt 880 result in loss of transformation may be through the activation of a cryptic splice site located within the E7 coding region resulting in the production of nonfunctional truncated E7 protein; however, RNase protection assays with a complete E7 coding region probe failed to identify products of such an event (data not shown). Other possible mechanisms whereby splicing affects functional E7 mRNA levels remain to be addressed. First, the unspliced E7 message may fail to undergo correct processing and be retained within the nucleus and, consequently, be rapidly degraded. In addition, splicing has been shown to increase polyadenylation (14, 25), which is known to stabilize mRNAs. The splice site mutations in the context of the whole HPV-16 genome greatly reduced but did not completely eliminate transformation. One factor to consider in comparing the HPV-16- and SV40-based constructs is that HPV-16 but not SV40 promoter-enhancer-driven constructs require the presence of the glucocorticoid hormone dexamethasone
HPV-16 SPLICE DONOR SITE REQUIRED FOR TRANSFORMATION
Voi.. 66, 1992
for transformation (28). The stabilizing effect of glucocorticoid hormones on growth hormone mRNA (26) and phosphoenolpyruvate carboxykinase mRNA (32) have been reported. The same may be occurring for HPV-16, and dexamethasone might have a dual role in the regulation of E7-mediated transformation, one at the transcriptional level and the other at the posttranscriptional level. Another possible explanation could that the other two splice donor sites at nt 226 and 1301 may function additively to substitute for the splice site at nt 880. It should be noted that effects of the nt 226 and 409 splice sites are very small, on the basis of our recent results with mutants in the context of the full-length HPV-16 genome (30). The present results demonstrate that the nt 880 splice donor site has a strong and dominant effect on E7-mediated BRK cell transformation in a mechanism that probably involves E7 mRNA stabilization due to splicing events. ACKNOWLEDGMENTS We thank L. Gissmann and H. zur Hausen for HPV-16 DNA, J. Lilly for technical assistance, A. Mithal for pAl, Y. Shi for assistance with site-directed mutagenesis, and S. Atkins for typing the manuscript. This work was supported in part by grants from the MRC and NCI of Canada.
REFERENCES 1. Ausubel, F. M., R. Brent, F. K. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl. 1990. Current protocols in molecular biology, vol. 1, p. 4.10.1-4.10.9. John Wiley & Sons., Inc., New York. 2. Buchman, A. R., and P. Berg. 1988. Comparison of introndependent and intron-independent gene expression. Mol. Cell. Biol. 8:4395-4405. 3. Casey, J. L., D. M. Koeller, V. C. Ramin, R. D. Klausner, and J. B. Harford. 1989. Iron regulation of transferrin receptor mRNA levels requires iron-responsive elements and a rapid turnover determinant in the 3' untranslated region of the mRNA. EMBO J. 8:3693-3699. 4. Chen, C., and H. Okayama. 1987. High-efficiency transformation of mammalian cells by plasmid DNA. Mol. Cell. Biol.
7:2745-2752. 5. Chirgwin, J. M., A. E. Przybyla, R. C. MacDonald, and W. J. Rutter. 1979. Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 18:5294-
5299. 6. de Villiers, E. M. 1989. Heterogeneity of the human papilloma-
virus group. J. Virol. 63:4898-4903. 7. de Villiers, E. M., A. Schneider, H. Miklaw, U. Papendick, D. Wagner, H. Wesch, J. Wahrendorf, and H. zur Hausen. 1987. Human papillomavirus infections in women with and without abnormal cervical cytology. Lancet ii:703-706. 8. Doorbar, J., A. Parton, K. Hartley, L. Banks, T. Crook, M. Stanley, and L. Crawford. 1990. Detection of novel splicing patterns in a HPV 16 containing keratinocyte cell line. Virology
178:254-262. 9. Durst, M., T. Dzarlieva-Petrusevska, P. Boukamp, N. E. Fusenig, and L. Gissman. 1987. Molecular and cytogenetic analysis of immortalized human primary keratinocytes obtained after transfection with human papillomavirus type 16 DNA. Onco-
gene 1:251-256.
10. Durst, M., L. Gissmann, H. Ikenberg, and H. zur Hausen. 1983. A papillomavirus DNA from a cervical carcinoma and its presence in cervical cancer biopsy samples from different geo-
graphic areas. Proc. Natl. Acad. Sci. USA 80:3812-3815.
11. Edmonds, C., and K. H. Vousden. 1989. A point mutational analysis of human papillomavirus type 16 E7 protein. J. Virol.
63:2650-2656. 12. Gergley, L., J. Czegledy, and Z. Hernady. 1987. Letter. Lancet
ii:513.
2729
13. Goff, S. P., and P. Berg. 1979. Construction, propagation and expression of simian virus 40 recombinant genomes containing the Escherichia coli gene for thymidine kinase and a Saccharomyces cerevisiae gene for tyrosine transfer RNA. J. Mol. Biol. 133:359-383. 14. Huang, M. T. F., and C. M. Gorman. 1990. Intervening sequences increase efficiency of RNA 3' processing and accumulation of cytoplasmic RNA. Nucleic Acids Res. 18:937-947. 15. Jackson, I. J. 1991. A reappraisal of non-consensus mRNA splice sites. Nucleic Acids Res. 19:3795-3798. 16. Kanda, T., S. Watanabe, and K. Yoshiike. 1988. Immortalization of primary rat cells by human papillomavirus type 16 subgenomic DNA fragments controlled by the SV40 promoter. Virology 165:321-325. 17. Kreig, P. A., and D. A. Melton. 1987. In vitro RNA synthesis with sp6 RNA polymerase. Methods Enzymol. 155:397-415. 18. Kunkel, T. A. 1985. Rapid and efficient site-specific mutagenesis without phenotypic selection. Proc. Natl. Acad. Sci. USA 82: 488-492. 19. Marshall, T., A. Pater, and M. M. Pater. 1989. Trans-regulation and differential cell specificity of human papillomavirus types 16, 18, and 11 cis-acting elements. J. Med. Virol. 29:115-126. 20. Matlashewski, G., J. Schneider, L. Banks, N. Jones, A. Murray, and L. Crawford. 1987. Human papillomavirus type 16 DNA co-operates with activated ras in transforming primary cells. EMBO J. 6:1741-1746. 21. Meijlink, F., T. Curran, A. D. Miller, and I. Verma. 1985. Removal of a 67-base-pair sequence in the noncoding region of protooncogene fos converts it to a transforming gene. Proc. Natl. Acad. Sci. USA 82:4987-4991. 22. Mullner, E. W., and L. C. Kuhn. 1988. A stem-loop in the 3' untranslated region mediates iron-dependent regulation of transferrin receptor mRNA stability in the cytoplasm. Cell 53:815825. 23. Munger, K., W. C. Phelps, V. Bubb, P. M. Howley, and R. Schlegel. 1989. The E6 and E7 genes of the human papillomavirus type 16 together are necessary and sufficient for transformation of primary human keratinocytes. J. Virol. 63:4417-4421. 24. Nelson, K. K., and M. R. Green. 1990. Mechanism for cryptic splice site activation during pre-mRNA splicing. Proc. Natl. Acad. Sci. USA 87:6253-6257. 25. Niwa, M., S. D. Rose, and S. M. Berget. 1990. In vitro polyadenylation is stimulated by the presence of an upstream intron. Genes Dev. 4:1552-1559. 26. Paek, I., and R. Axel. 1987. Glucocorticoids enhance stability of human growth hormone mRNA. Mol. Cell. Biol. 7:1496-1507. 27. Pater, M. M., J. Dunne, G. Hogan, P. Ghatage, and A. Pater. 1986. Human papillomavirus types 16 and 18 sequences in early cervical neoplasia. Virology 155:13-18. 28. Pater, M. M., G. A. Hughes, D. E. Hyslop, H. Nakshatri, and A. Pater. 1988. Glucocorticoid-dependent oncogenic transformation by type 16 but not type 11 human papilloma virus DNA. Nature (London) 335:832-835. 29. Pater, M. M., T. Marshall, H. Nakshatri, and A. Pater. 1990. Glucocorticoid-dependent transformation of primary cells by human papillomavirus type 16 DNA. UCLA Symp. Mol. Cell. Biol. 126:313-321. 30. Pater, M. M., H. Nakshatri, C. Kisaka, and A. Pater. 1992. The first 124 nucleotides of the E7 coding sequences of HPV16 can render the HPV11 genome transformation competent. Virology 186:348-351. 31. Pater, M. M., and A. Pater. 1985. Human papillomavirus type 16 and 18 sequences in carcinoma cell lines of the cervix. Virology 145:313-318. 32. Petersen, D. D., S. R. Koch, and D. K. Granner. 1989. 3' noncoding region of phosphoenolpyruvate carboxykinase mRNA contains a glucocorticoid-responsive mRNA-stabilizing element. Proc. Natl. Acad. Sci. USA 86:7800-7804. 33. Phelps, W. C., C. L. Yee, K. Munger, and P. M. Howley. 1988. The human papillomavirus type 16 E7 gene encodes transactivation and transformation functions similar to those of adenovirus EIA. Cell 53:539-547. 34. Rohlfs, M., S. Winkenbach, S. Meyer, T. Rupp, and M. Durst.
2730
35.
36. 37.
38.
BELAGULI ET AL.
1991. Viral transcription in human keratinocyte cell lines immortalized by human papillomavirus type-16. Virology 183:331342. Senapathy, P., M. B. Shapiro, and N. L. Harris. 1990. Splice junctions, branch site points, and exons: sequence statistics, identification, and applications to genomic project. Methods Enzymol. 183:252-278. Shah, K. V., and P. M. Howley. 1990. Papillomaviruses, p. 1651-1676. In B. N. Fields and D. M. Knipe (ed.), Virology. Raven Press, Ltd., New York. Smotkin, D., and F. 0. Wettstein. 1986. Transcription of human papillomavirus type 16 early genes in a cervical cancer and a cancer-derived cell line and identification of the E7 protein. Proc. Natl. Acad. Sci. USA 83:4680-4684. Tanaka, A., T. Noda, H. Yajima, M. Hatanaka, and Y. Ito. 1989.
J. VIROL.
Identification of a transforming gene of human papillomavirus type 16. J. Virol. 63:1465-1469. 39. Watanabe, S., T. Kanda, H. Sato, A. Furuno, and K. Yoshiike. 1990. Mutational analysis of human papillomavirus type 16 E7 functions. J. Virol. 64:207-214. 40. Young, L. S., I. S. Bevan, M. A. Johnson, P. I. Blomfield, T. Brombridge, N. J. Maitland, and C. B. J. Woodman. 1989. The polymerase chain reaction: a new epidemiological tool for investigating cervical human papillomavirus infection. Br. Med. J. 298:14-18. 41. zur Hausen, H., and A. Schneider. 1987. The role of papillomaviruses in anogenital cancer, p. 245-263. In N. P. Salzman and P. M. Howley (ed.), The Papovaviridae: the papillomaviruses. Plenum Publishing Corp., New York.
