- Email: [email protected]
Genetic analysis of CRPV pathogenesis: The L1 open reading frame is dispensable for cellular transformation but is required for papilloma formation
VIROLOGY
170,321-325
(1989)
Genetic Analysis of CRPV Pathogenesis: The Ll Open Reading Frame Is Dispensable Transformation but Is Required for Papilloma Formation
for Cellular
M. NASSERI,*+’ C. MEYERS,* AND F. 0. WE-rrSTEIN*+2 *Department
of Microbiology
and Immunology, School of Medicine, tMo/ecu/ar Biology Institute and Jonsson Comprehensive University of California, Los Angeles, California 90024- 1747 Received October 7, 1988; acceptedJanuary
Cancer Center,
12, 1989
Genetic studies to elucidate the role of papillomaviruses in the development and progression of tumors have been severely hampered because the viruses cannot be grown in tissue culture and therefore mutants are not available. We have employed recombinant DNA for papilloma induction to identify essential sequences involved in papillomavirus pathogenesis. Here, we demonstrated that deleting most of the open reading frame (ORF) L2 did not affect the potential of viral cottontail rabbit papillomavirus (CRPV) DNA to induce papillomas. The extrachromosomally maintained DNA in the papillomas was not rearranged and the major transcripts of 1.3 and 2.0 kb encoded E7 and E6, respectively. A recombinant DNA containing a larger deletion lacking the 3’terminal half of ORF L2 and all of ORF Ll (pdlBc/l) did not induce papillomas. The results indicate that sequences in the late region not required for transformation of NIH 3T3 o 1999Academic PUSS, IN. cells by bovine papillomavirus type-l essential in CRPV for induction of papillomas.
and/or maintenance of carcinomas. This notion is supported by two facts: (i) both transcripts are present in the transplantable cancer line VX-2 (9, 11, 12), established about 50 years ago, and (ii) they are the two major transcripts which have been expressed in transformed cell lines and the nude mouse tumors induced by these lines (13). Furthermore, the analogy between oncogenic human papillomaviruses and CRPV is reflected in the expression of the equivalent proteins in the cancers associated with both groups of viruses (14-16). The viral genetic elements involved in the development of papillomas and progression to cancers in their natural hosts have not been characterized for any of the papillomaviruses since, in the absence of a tissue culture system, mutant virus cannot be isolated. Although a genetic analysis based on transfection of cells in culture has provided information about viral genes involved in cell transformation (17~23), there are important differences between transformation of cells in culture and development of carcinomas in a natural host. This is illustrated by the findings with BPV-1, a nononcogenic papillomavirus (a papillomavirus not associated with malignancies in its natural .host) which does, however, transform cells in culture which are tumorigenic for nude mice (24). We are now using CRPV as a model to define the role of different genetic elements in the development of papillomas and their progression to carcinomas. We took advantage of the precise transcript mapping data and DNA sequence information (25) and constructed specific deleted cecom-
Papillomaviruses are a group of closely related small DNA tumor viruses which have recently received special attention because of their strong association with human genital cancer (1). Despite their role in the induction of papillomas and carcinomas in both humans and in some animals, very little is known about the viral genetics of neoplasia formation. This is for two reasons: (i) the lack of permissive cell cultures, and (ii) their strict host and tissue specificity. Therefore, in order to study the pathogenesis of these medically important viruses, new approaches have to be developed. Cottontail rabbit papillomavirus (CRPV) occupies a unique place among papillomaviruses since papillomas induced by this virus progress to cancers and the progression is particularly high (759/o) with papillomas induced in domestic rabbits (2) and therefore the virus serves as an animal model for oncogenic human papillomaviruses. Papillomas and carcinomas from both domestic and cottontail rabbits have been examined and the structure and transcription of viral DNA has been characterized (3-7). Several viral transcripts in virusproducing and non-virus-producing papillomas and in carcinomas have been precisely mapped within both early and late regions (8-10). Two of the early transcripts of 1.3 and 2.0 kb encoding E7 and E6, respectively, are present in both papillomas and carcinomas and are believed to play a critical role in establishment ’ Current address: Department of Microbiology and Immunology, Sherman Fairchild Science Building, Stanford University School of Medicine, Stanford, CA 94305. ’ To whom requests for reprints should be addressed. 321
0042-6822189
$3.00
Copyright Q 1989 by Academic Press. Inc. All rights of reproduction I” any form reserved.
322
SHORT COMMUNICATIONS
-Xbal
\
\
’ \
\
\
I’ :
\ ‘\
PVUll 4464
\
: \
\
5-h ‘,4671
5775
, , I \ , , I \ Xbal 665
AT6
..*. .. -
1.3
kb RNA 3’-end.
t 2.0 Poly A signal 4348
FIG. 1. Map and construction of recombinant plasmids; the 7%kb CRPVgenome inserted into the SalI site of pLAll resulting in the 1 1.3kb CRPV-pLAII. El-7, Ll, and L2 are the early and late ORFs as deduced from DNA sequence studies (25). The expanded region (nucleotides 4265 to 4665) is depicted to show the position of Sal1 and Xbal with respect to E5 and L2 ORFs in CRPV-pLAII, and pdt%bal, respectively. The ATG initiation codon for the E5 and L2 are indicated. The position of AATAAA used for termination and polyadenylation of the early transcripts is shown. Linear DNA was prepared by SalI digestion of &/l-cloned CRPV-pLAII recombinant DNA and circular DNA was prepared by in vitro ligation of CRPV DNA separated on glycerol gradients from the vector DNA. DNA (1 O-20 pg) was inoculated per site as previously described (13). The plasmid pdlXbal was constructed by Xbal digestion removing nts 4665 to 5775 from CRPV-pLAII. The lO.l-kb DNA band was prepared by gel purification followed by religation, cloning and selection for the deleted plasmid. The structure of the pdlXbal was further confirmed by enzyme digestion. The plasmid pdlBc/l was constructed by&/l digestion of CRPVpLAII, prepared from a dam-Escherichia co/i strain EM 119, removing the &/I fragment (nts 4899-7213) followed by its religation and cloning. The structure of the selected pdlBc/l mutants was further confirmed by restriction enzyme analysis.
binant molecules and identified their potential to induce papillomas. The experiments relied on a previoustyestablished technique of papilloma induction in rabbits by direct administration of DNA (13, 26). Here we show that a recombinant DNA with a 1 .l -kb deletion in the late region retained the potential to induce papillomas while deletion of larger segment of the late region abolished this potential.
To induqe papillomas with cloned viral DNA, both linear and circular CRPV DNA were tested. Both DNA preparations induced papillomas in rabbits. The state of viral DNA in the papillomas was supercoiled and the DNA could be linearized by Sal1 digestion producing a 7.8-kb band corresponding to unit-sized CRPV (Fig. 2A). These results demonstrated that circular and linear CRPV DNA induced papillomas in animals and that the viral DNA was maintained extrachromosomally as in virus induced papillomas. Furthermore, several transcripts including the major 1.3- and 2.0-kb species were present in these papillomas (data not shown). Previously, we and others have shown that all early transcripts utilized the polyadenylation signal located at nucleotide (nt) 4348 (Fig. 1). Since the Sal1 site is located at nt 4571 which is 223 nts downstream of the signal, it appeared likely that viral DNA interrupted at the Sal1 site would induce papillomas. When CRPV cloned into the “poisonless” vector pLAll (27) was inoculated into rabbits, papillomas developed (Table 1). Analysis of papilloma DNA by Southern blot hybridization showed that the recombinant DNA was maintained extrachromosomally (Fig. 2B). Digestion of the papilloma DNA as well as of the input DNA with Sal1 produced two bands of 3.5 and 7.8 kb corresponding to the pLAll vector and to the CRPV DNA, respectively (Fig. 2B). The digest of the papilloma DNA was not complete, as indicated by bands of nicked and linear DNA. The incomplete digestion is probably the consequence of a low level of methylation at the SalI sites (3). The results demonstrated that the interruption at the SalI site did not affect the ability of the viral DNA to induce papillomas. It should be mentioned that insertion of the vector DNA at the Sal1 site removed the stop codon for the E5 ORF (Fig. 1); however, a new stop codon was introduced with the vector DNA 48 bp TABLE 1 PAPILLOMA~NDUCTIONWITHCRPV RECOMEINANTDNA
Experiment 1 2 3 4 5 All experiments
wt
pdlXbal
pdlBc/l
131178 314 z/13 4/19 7117 29170
8114 N.D.b 5110 o/3 9114 22141
Of9 o/5 o/9 N.D. o/7 o/34
Note. Wild-type CRPV DNA (wt). or CRPV DNA with the following deletions: nt 4665 to nt 5775 (pdlXbal) and nt 4899 to nt 7213 (pdlBc/l) cloned in the “poisonless” pBR322 derivation pLAll and inoculated into domestic rabbits as described previously (13). a Number of tumors per sites inoculated with DNA. b Not done.
323
SHORT COMMUNICATIONS
I
TUMOR ONA
B
C
B
A
3
TUMOR DNA -
p*
r
TUMOR DNA
pdl Xbal
3
2.3 -
FIG. 2. Southern blot analysis of DNA from papillomas induced by wild-type or mutant recombinant plasmids in rabbits; papillomas were induced in rabbits as described (13). (A) DNA (5 pg) isolated from a tumor induced by Sall-linearized CRPV-DNA undigested (U) or Sall-digested were separated on a 0.7% agarose gel, transferred to filters, and hybridized into CRPV DNA 32P, labeled by nick-translation. (6) DNA (5 rg) isolated from a papilloma induced by CRPV-pLAII DNA or plasmid DNA (100 pg) (pCRPV) were digested with SalI or left undigested and separated on a 0.7% agarose gel, transferred to filters, and hybridized as under (A). (C) DNA (5 fig) isolated from a tumor induced by pdlXbal DNA or plasmid DNA (100 pg) (pdV(bal) undigested or digested with &/I (tumor only) or BamHl were electrophoresed on a 0.7% agarose gel, transferred to filters, and hybridized as under (A). The positions and sizes in kilobases of molecular markers, phage X DNA HindIll-digested, are shown. The position of forms I, II, and Ill plasmid DNA is shown in (A).
downstream of the Ml site, thus increasing the size of a putative E5 protein by 16 amino acids. In this recombinant DNA L2 was also affected; it was interrupted 193 bp downstream of the first ATG (nt 4378), thus, an intact L2 could not be synthesized. These results suggested that L2 was not essential for papilloma induction in domestic rabbits. It would, however, be possible that a truncated L2 initiated at an ATG further downstream could still be made. To further investigate a potential role of the late region of the viral DNA in inducing papillomas, a CRPVpLAll plasmid (pdlMal), deleted for the Xbal fragment (nt 4665 to nt 5775) was tested. This plasmid, lacking 80% of the carboxy terminal portion of L2, still induced papillomas in rabbits (Table 1). This result clearly showed that L2 was not required, and furthermore it indicated that this region did not contain essential regulatory sequences for papilloma formation. Southern blot analysis of viral DNA in one such papilloma is shown in Fig. 2C. Papilloma DNA undigested or di-
gested with &/I or BamHl were analyzed by Southern blotting parallel to plasmid DNA (pdlXbal). The results showed that the recombinant DNA remained extrachromosomal in the papillomas (Fig. 2C). In both undigested DNA lanes, bands representing multimeric circular forms were also present (3). Digestion with Bell produced a single band corresponding to the unit-sized linear DNA and BarnHI digestion resulted in three bands with the expected sizes of 4037, 3877, and 2322 bp. This demonstrated that the structure of the plasmid in papillomas had remained unchanged. Finally, RNA blot analysis of CRPV transcripts from papillomas induced by pdlXbal revealed the major 1.3- and 2.0-kb transcripts common to papillomas and carcinomas induced by the virus (Fig. 3). We had previously shown that a CRPV DNA variant lacking most of ORF Ll and the 3’ half of ORF L2, but not wild-type DNA, was present extrachromosomally in malignantly transformed 3T3 mouse fibroblasts (13). In contrast, in rabbit papillomas, the variant DNA was always maintained in addition to the wild-type CRPV DNA, suggesting that the genetic requirements for transformation of fibroblasts are different from those for tumor formation (13). To test directly if deletion of a large segment of the late region affected the ability of the CRPV DNA to induce papillomas, a deletion similar
TUMOR RNA dlXbal WW
285
-
18s
-
4.8
2.6 2.0 1.3 0.9
FIG. 3. Analysis of CRPV poly(A)+ RNA from papillomas induced by pdlXbal DNA. Poly(A)+ RNA (5 pg) was isolated from a pdKbal-induced tumor (dlXbal) and from a virus-induced cottontail rabbit papilloma (1 or 0.5 pg) (WW). electrophoresed on a 1.1% agarose gel containing 2.2 Mformaldehyde, transferred to GeneScreen, and hybridized to CRPV DNA 32P-labeled by nick-translation (sp act 10’ cpm/ rg). The positions of 18 and 28 S ribosomal RNA are indicated on the left and the positions and sizes in kilobases of cottontail rabbit papilloma RNAs are on the right.
324
SHORT COMMUNICATIONS
to that of the variant was constnlpted by Bell digestion of CRPV-pLAII. The plasmid, pdlbcll, had a deletion from nt 4899 to nt 7213. Inoculation of rabbits with this recombinant failed to induce papillomas (Table 1). The fact that rabbits inoculated with pdlBc/l developed papillomas at other sites inoculated with pCRPV-pLAII or pdKbal indicated that the negative results were not due to a general unresponsiveness. Furthermore, it is unlikely that other alterations of the viral DNA in pdlf3cll were responsible for the inability to induce tumors since independently cloned pdlBc/l deletions were tested. Finally, a Pvull deletion construct (nt 4454-nt 6290) lacking both the Ll and L2 region (Fig. 1) did not induce papillomas (no tumors developed at 29 sites inoculated). In the absence of mutant virus, we have employed DNA to induce papillomas to initiate experiments aimed at genetic analysis of viral pathogenesis. In this first phase, we were particularly interested in determining if, and to what extent, the late region could be deleted without impairing the ability of the DNA to induce papillomas in domestic rabbits. Our results indicated that L2 is not required. Deletion of a larger segment in pdlBc/l (nt 4899 to nt 7213) lacking the whole Ll ORF and leaving the long control region (LCR) intact abolished this ability. An equivalent deletion in BPV-1 does not affect its ability to transform NIH 3T3 cells (24). This could indicate that some sequences in the late region of papillomaviruses are not needed for transformation of fibroblasts but are required for papilloma induction. This notion is further supported by the findings with the spontaneous CRPV variant which also had deleted most of the late region but still transformed NIH 3T3 cells (13). There are several ways ORF Ll sequences could affect the potential to induce papillomas. One possibility would be that the Ll protein itself is required. Although late proteins have not been detected in domestic rabbit papillomas and we have not been able to detect late transcripts, a low level of infectivity has been reported in some domestic rabbit papillomas (28). In addition, domestic rabbits inoculated with the transplantable VX7 carcinoma develop an antibody response to the Ll protein (29). Regulatory effects of late viral proteins have been demonstrated for the polyomavirus capsid protein VP1 which induces the cellular oncogenes c-myc and c-fos (30). Similarly, the HSVI tegument protein Vmw 65 initiates the gene expression by direct binding to upstream activator sequences of immediate early genes (31). Another possibility is that a regulatory element(s) located within ORF Ll plays a definite role in papilloma induction but not in transformation of cells in culture. No such regulatory element(s) has been mapped yet in CRPV, but in BPV-1 an
enhancer-like element is located in Ll which plays a role in DNA replication (32, 33). The requirement for this element, however, appears to be conditional since some BPV-1 plasmid constructs lacking this element can be maintained extrachromosomally. Thus it is possible that such an enhancer, if indeed present in CRPV, is essential for induction of papillomas but may not be required for transformation of NIH 3T3 cells in culture. In conclusion, our results presented here have shown that induction of papillomas in a natural host requires the function of genetic elements which may not be needed for transformation of cells in culture.
ACKNOWLEDGMENTS We thank Margerita Tayag for excellent technical help and Jeannie Martinez and Lori Kenton for help with the manuscript. M.N. was the recipient of a postdoctoral fellowship from the Cancer Research Coordinating Committee of the University of California and CM. was a recipient of a predoctoral fellowship from U.S. Public Health Service Grants GM 07104 and Al 07323. The research was supported by Grant CA 1815 1 awarded by the National Cancer Institute, National Institutes for Health, Department of Health and Human Services, and a grant from the W.M. Keck Foundation.
REFERENCES 1. PETO, R., and ZUR HAUSEN, H., Eds. “Banbury Report,” Vol. 21 Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 2. SYVERTON,J. T., Ann. NYAcad. Sci. 54, 1126-l 140 (1952). 3. WETTSTEIN,F. O., and STEVENS,J. G., Proc. Nat/. Acad. Sci. USA 79,790-794 (1982). 4. NASSERI, M., WETTSTEIN, F. O., and STEVENS,J. G., J. Viral. 44, 263-268 (1982). 5. MC VAY, P., FRETZ, M., WETSTEIN, F. O., STEVENS,J. G., and ITO, Y., 1. Gen. Viol. 60,27 l-278 (1982). 6. PHELPS,W. C., LEARY,S. L., and FARAS,A. J., L/irology 146, 120129 (1985). 7. WETTSTEIN, F. O., In “The Papovaviridae” (N. Salzman and P. Howley, Eds.), Vol. 2, Chap. 6. Plenum, New York. 8. WE~TSTEIN,F. O., BARBOSA,M. S., and NASSERI,M., Virology 159, 321-328 (1987). 9. NASSERI,M., and WETTSTEIN,F. O., /. viral. 51, 706-712 (1984). 10. NASSERI, M., and WETTSTEIN, F. O., !&o/ogy 138, 362-367 (1984). 11. GEORGES,E., CROISSANT,O., BONNEAUD,N., and ORTH, G..f. l/ire/. 51,530~538 (1984). 12. DANOS, O., GEORGE% E., ORTH. G., and YANIF. M., J. Viral. 53, 735-741 (1985). 13. NASSERI, M., and WETTSTEIN, F. O., virology 161, 541-548 (1987). 14. SMOTKIN, D., and WE-STEIN, F. O., Proc. Narl. Acad. Sci USA 83,4680-4684 (1986). 15. SCHNEIDER-GADICKE,A., and SCHWARZ, E., EMBO J. 5, 22852292 (1986). 16. INAGAKI,Y., TSUNOKAWA,Y., TAKEBE, N., NAWA, H., NAKANISHI,S., TERADA, M., and SUGIMURA,T., 1. viral. 62, 1640-l 646 (1988). 17. ANDROPHY, E. J., SCHILLER,J. T., and Lowv, D. R., Science 230, 442-445 (1985). 18. SCHLEGEL,R., WADE-GLASS, M., RAESON, M. S., and YANG, Y.-C., Science 233,464-467 (1986).
SHORT COMMUNICATIONS 19. 20. 21. 22. 23. 24. 25.
M., JONES, K., and LAIMINS, L., J. Viral. 61, 3635-3640 (1987). KANDA, T., FURUNO, A., and YOSHIIKE, Y., 1. Viral. 62, 61 O-61 3 (1988). MATLASHEWSKI,G., SCHNEIDER,J., BANKS, L., JONES, N., MURRAY, A., and CRAWFORD,L., EMBOI. 6,1741-1746 (1987). PHELPS, W. C., LEE, C. L., MUNGEA. K.. and HOWLEY, P. M., Cell 53,539-547 (1988). STOREY, A., PIM, D., MURRAY, A., OSEIORN, K., BANKS, L.. and CRAWFORD,L., fMBO/. 7,1815-1820 (1988). Lowv, D. R., DVORETZKY,I., SHOBER, R.. !-Aw, M.-F., ENGEL. L., and HOWLEY, P. M., Nature (London) 287,72-74 (1980). GIRI, I., DANOS, O., and YANIF, M., Proc. Nat/. Acad. Sci. USA 82, 1580-l 584 (1985). BEDELL,
325
26. ITO,Y.,~~~EVANS,C.A.,J.&J. Med. 114,485-500(1961). 27. MELLON, P., PARKER,V., GLUZMAN, Y., and MANIATIS, T., Ce//27, 279-288 (1981). 28. SHOPE, R. E., Proc. Sot. Exp. Med. 32, 830 (1935). 29. ORTH, G., BREITBARD,F., and FAVRE, M., Virology 91, 243-255 (1978). 30. ZULLO, J., STILES, C. D., and GARCEA, R. L., Proc. Nat/. Acad. Sci, USA84, 1210-1214(1987). 31. O’HARE, P., and GODING, C. R., Ce//52,435-445 (1988). 32. LUSKY, M.. and BOTCHAN, M. R.. Proc. Nat/. Acad. Sci. USA 83, 3609-3613 (1986). 33. STENLUND, A., BREAM, G. L., and BOTCHAN, M. R., Science 236, 1666-1671 (1987).
170,321-325
(1989)
Genetic Analysis of CRPV Pathogenesis: The Ll Open Reading Frame Is Dispensable Transformation but Is Required for Papilloma Formation
for Cellular
M. NASSERI,*+’ C. MEYERS,* AND F. 0. WE-rrSTEIN*+2 *Department
of Microbiology
and Immunology, School of Medicine, tMo/ecu/ar Biology Institute and Jonsson Comprehensive University of California, Los Angeles, California 90024- 1747 Received October 7, 1988; acceptedJanuary
Cancer Center,
12, 1989
Genetic studies to elucidate the role of papillomaviruses in the development and progression of tumors have been severely hampered because the viruses cannot be grown in tissue culture and therefore mutants are not available. We have employed recombinant DNA for papilloma induction to identify essential sequences involved in papillomavirus pathogenesis. Here, we demonstrated that deleting most of the open reading frame (ORF) L2 did not affect the potential of viral cottontail rabbit papillomavirus (CRPV) DNA to induce papillomas. The extrachromosomally maintained DNA in the papillomas was not rearranged and the major transcripts of 1.3 and 2.0 kb encoded E7 and E6, respectively. A recombinant DNA containing a larger deletion lacking the 3’terminal half of ORF L2 and all of ORF Ll (pdlBc/l) did not induce papillomas. The results indicate that sequences in the late region not required for transformation of NIH 3T3 o 1999Academic PUSS, IN. cells by bovine papillomavirus type-l essential in CRPV for induction of papillomas.
and/or maintenance of carcinomas. This notion is supported by two facts: (i) both transcripts are present in the transplantable cancer line VX-2 (9, 11, 12), established about 50 years ago, and (ii) they are the two major transcripts which have been expressed in transformed cell lines and the nude mouse tumors induced by these lines (13). Furthermore, the analogy between oncogenic human papillomaviruses and CRPV is reflected in the expression of the equivalent proteins in the cancers associated with both groups of viruses (14-16). The viral genetic elements involved in the development of papillomas and progression to cancers in their natural hosts have not been characterized for any of the papillomaviruses since, in the absence of a tissue culture system, mutant virus cannot be isolated. Although a genetic analysis based on transfection of cells in culture has provided information about viral genes involved in cell transformation (17~23), there are important differences between transformation of cells in culture and development of carcinomas in a natural host. This is illustrated by the findings with BPV-1, a nononcogenic papillomavirus (a papillomavirus not associated with malignancies in its natural .host) which does, however, transform cells in culture which are tumorigenic for nude mice (24). We are now using CRPV as a model to define the role of different genetic elements in the development of papillomas and their progression to carcinomas. We took advantage of the precise transcript mapping data and DNA sequence information (25) and constructed specific deleted cecom-
Papillomaviruses are a group of closely related small DNA tumor viruses which have recently received special attention because of their strong association with human genital cancer (1). Despite their role in the induction of papillomas and carcinomas in both humans and in some animals, very little is known about the viral genetics of neoplasia formation. This is for two reasons: (i) the lack of permissive cell cultures, and (ii) their strict host and tissue specificity. Therefore, in order to study the pathogenesis of these medically important viruses, new approaches have to be developed. Cottontail rabbit papillomavirus (CRPV) occupies a unique place among papillomaviruses since papillomas induced by this virus progress to cancers and the progression is particularly high (759/o) with papillomas induced in domestic rabbits (2) and therefore the virus serves as an animal model for oncogenic human papillomaviruses. Papillomas and carcinomas from both domestic and cottontail rabbits have been examined and the structure and transcription of viral DNA has been characterized (3-7). Several viral transcripts in virusproducing and non-virus-producing papillomas and in carcinomas have been precisely mapped within both early and late regions (8-10). Two of the early transcripts of 1.3 and 2.0 kb encoding E7 and E6, respectively, are present in both papillomas and carcinomas and are believed to play a critical role in establishment ’ Current address: Department of Microbiology and Immunology, Sherman Fairchild Science Building, Stanford University School of Medicine, Stanford, CA 94305. ’ To whom requests for reprints should be addressed. 321
0042-6822189
$3.00
Copyright Q 1989 by Academic Press. Inc. All rights of reproduction I” any form reserved.
322
SHORT COMMUNICATIONS
-Xbal
\
\
’ \
\
\
I’ :
\ ‘\
PVUll 4464
\
: \
\
5-h ‘,4671
5775
, , I \ , , I \ Xbal 665
AT6
..*. .. -
1.3
kb RNA 3’-end.
t 2.0 Poly A signal 4348
FIG. 1. Map and construction of recombinant plasmids; the 7%kb CRPVgenome inserted into the SalI site of pLAll resulting in the 1 1.3kb CRPV-pLAII. El-7, Ll, and L2 are the early and late ORFs as deduced from DNA sequence studies (25). The expanded region (nucleotides 4265 to 4665) is depicted to show the position of Sal1 and Xbal with respect to E5 and L2 ORFs in CRPV-pLAII, and pdt%bal, respectively. The ATG initiation codon for the E5 and L2 are indicated. The position of AATAAA used for termination and polyadenylation of the early transcripts is shown. Linear DNA was prepared by SalI digestion of &/l-cloned CRPV-pLAII recombinant DNA and circular DNA was prepared by in vitro ligation of CRPV DNA separated on glycerol gradients from the vector DNA. DNA (1 O-20 pg) was inoculated per site as previously described (13). The plasmid pdlXbal was constructed by Xbal digestion removing nts 4665 to 5775 from CRPV-pLAII. The lO.l-kb DNA band was prepared by gel purification followed by religation, cloning and selection for the deleted plasmid. The structure of the pdlXbal was further confirmed by enzyme digestion. The plasmid pdlBc/l was constructed by&/l digestion of CRPVpLAII, prepared from a dam-Escherichia co/i strain EM 119, removing the &/I fragment (nts 4899-7213) followed by its religation and cloning. The structure of the selected pdlBc/l mutants was further confirmed by restriction enzyme analysis.
binant molecules and identified their potential to induce papillomas. The experiments relied on a previoustyestablished technique of papilloma induction in rabbits by direct administration of DNA (13, 26). Here we show that a recombinant DNA with a 1 .l -kb deletion in the late region retained the potential to induce papillomas while deletion of larger segment of the late region abolished this potential.
To induqe papillomas with cloned viral DNA, both linear and circular CRPV DNA were tested. Both DNA preparations induced papillomas in rabbits. The state of viral DNA in the papillomas was supercoiled and the DNA could be linearized by Sal1 digestion producing a 7.8-kb band corresponding to unit-sized CRPV (Fig. 2A). These results demonstrated that circular and linear CRPV DNA induced papillomas in animals and that the viral DNA was maintained extrachromosomally as in virus induced papillomas. Furthermore, several transcripts including the major 1.3- and 2.0-kb species were present in these papillomas (data not shown). Previously, we and others have shown that all early transcripts utilized the polyadenylation signal located at nucleotide (nt) 4348 (Fig. 1). Since the Sal1 site is located at nt 4571 which is 223 nts downstream of the signal, it appeared likely that viral DNA interrupted at the Sal1 site would induce papillomas. When CRPV cloned into the “poisonless” vector pLAll (27) was inoculated into rabbits, papillomas developed (Table 1). Analysis of papilloma DNA by Southern blot hybridization showed that the recombinant DNA was maintained extrachromosomally (Fig. 2B). Digestion of the papilloma DNA as well as of the input DNA with Sal1 produced two bands of 3.5 and 7.8 kb corresponding to the pLAll vector and to the CRPV DNA, respectively (Fig. 2B). The digest of the papilloma DNA was not complete, as indicated by bands of nicked and linear DNA. The incomplete digestion is probably the consequence of a low level of methylation at the SalI sites (3). The results demonstrated that the interruption at the SalI site did not affect the ability of the viral DNA to induce papillomas. It should be mentioned that insertion of the vector DNA at the Sal1 site removed the stop codon for the E5 ORF (Fig. 1); however, a new stop codon was introduced with the vector DNA 48 bp TABLE 1 PAPILLOMA~NDUCTIONWITHCRPV RECOMEINANTDNA
Experiment 1 2 3 4 5 All experiments
wt
pdlXbal
pdlBc/l
131178 314 z/13 4/19 7117 29170
8114 N.D.b 5110 o/3 9114 22141
Of9 o/5 o/9 N.D. o/7 o/34
Note. Wild-type CRPV DNA (wt). or CRPV DNA with the following deletions: nt 4665 to nt 5775 (pdlXbal) and nt 4899 to nt 7213 (pdlBc/l) cloned in the “poisonless” pBR322 derivation pLAll and inoculated into domestic rabbits as described previously (13). a Number of tumors per sites inoculated with DNA. b Not done.
323
SHORT COMMUNICATIONS
I
TUMOR ONA
B
C
B
A
3
TUMOR DNA -
p*
r
TUMOR DNA
pdl Xbal
3
2.3 -
FIG. 2. Southern blot analysis of DNA from papillomas induced by wild-type or mutant recombinant plasmids in rabbits; papillomas were induced in rabbits as described (13). (A) DNA (5 pg) isolated from a tumor induced by Sall-linearized CRPV-DNA undigested (U) or Sall-digested were separated on a 0.7% agarose gel, transferred to filters, and hybridized into CRPV DNA 32P, labeled by nick-translation. (6) DNA (5 rg) isolated from a papilloma induced by CRPV-pLAII DNA or plasmid DNA (100 pg) (pCRPV) were digested with SalI or left undigested and separated on a 0.7% agarose gel, transferred to filters, and hybridized as under (A). (C) DNA (5 fig) isolated from a tumor induced by pdlXbal DNA or plasmid DNA (100 pg) (pdV(bal) undigested or digested with &/I (tumor only) or BamHl were electrophoresed on a 0.7% agarose gel, transferred to filters, and hybridized as under (A). The positions and sizes in kilobases of molecular markers, phage X DNA HindIll-digested, are shown. The position of forms I, II, and Ill plasmid DNA is shown in (A).
downstream of the Ml site, thus increasing the size of a putative E5 protein by 16 amino acids. In this recombinant DNA L2 was also affected; it was interrupted 193 bp downstream of the first ATG (nt 4378), thus, an intact L2 could not be synthesized. These results suggested that L2 was not essential for papilloma induction in domestic rabbits. It would, however, be possible that a truncated L2 initiated at an ATG further downstream could still be made. To further investigate a potential role of the late region of the viral DNA in inducing papillomas, a CRPVpLAll plasmid (pdlMal), deleted for the Xbal fragment (nt 4665 to nt 5775) was tested. This plasmid, lacking 80% of the carboxy terminal portion of L2, still induced papillomas in rabbits (Table 1). This result clearly showed that L2 was not required, and furthermore it indicated that this region did not contain essential regulatory sequences for papilloma formation. Southern blot analysis of viral DNA in one such papilloma is shown in Fig. 2C. Papilloma DNA undigested or di-
gested with &/I or BamHl were analyzed by Southern blotting parallel to plasmid DNA (pdlXbal). The results showed that the recombinant DNA remained extrachromosomal in the papillomas (Fig. 2C). In both undigested DNA lanes, bands representing multimeric circular forms were also present (3). Digestion with Bell produced a single band corresponding to the unit-sized linear DNA and BarnHI digestion resulted in three bands with the expected sizes of 4037, 3877, and 2322 bp. This demonstrated that the structure of the plasmid in papillomas had remained unchanged. Finally, RNA blot analysis of CRPV transcripts from papillomas induced by pdlXbal revealed the major 1.3- and 2.0-kb transcripts common to papillomas and carcinomas induced by the virus (Fig. 3). We had previously shown that a CRPV DNA variant lacking most of ORF Ll and the 3’ half of ORF L2, but not wild-type DNA, was present extrachromosomally in malignantly transformed 3T3 mouse fibroblasts (13). In contrast, in rabbit papillomas, the variant DNA was always maintained in addition to the wild-type CRPV DNA, suggesting that the genetic requirements for transformation of fibroblasts are different from those for tumor formation (13). To test directly if deletion of a large segment of the late region affected the ability of the CRPV DNA to induce papillomas, a deletion similar
TUMOR RNA dlXbal WW
285
-
18s
-
4.8
2.6 2.0 1.3 0.9
FIG. 3. Analysis of CRPV poly(A)+ RNA from papillomas induced by pdlXbal DNA. Poly(A)+ RNA (5 pg) was isolated from a pdKbal-induced tumor (dlXbal) and from a virus-induced cottontail rabbit papilloma (1 or 0.5 pg) (WW). electrophoresed on a 1.1% agarose gel containing 2.2 Mformaldehyde, transferred to GeneScreen, and hybridized to CRPV DNA 32P-labeled by nick-translation (sp act 10’ cpm/ rg). The positions of 18 and 28 S ribosomal RNA are indicated on the left and the positions and sizes in kilobases of cottontail rabbit papilloma RNAs are on the right.
324
SHORT COMMUNICATIONS
to that of the variant was constnlpted by Bell digestion of CRPV-pLAII. The plasmid, pdlbcll, had a deletion from nt 4899 to nt 7213. Inoculation of rabbits with this recombinant failed to induce papillomas (Table 1). The fact that rabbits inoculated with pdlBc/l developed papillomas at other sites inoculated with pCRPV-pLAII or pdKbal indicated that the negative results were not due to a general unresponsiveness. Furthermore, it is unlikely that other alterations of the viral DNA in pdlf3cll were responsible for the inability to induce tumors since independently cloned pdlBc/l deletions were tested. Finally, a Pvull deletion construct (nt 4454-nt 6290) lacking both the Ll and L2 region (Fig. 1) did not induce papillomas (no tumors developed at 29 sites inoculated). In the absence of mutant virus, we have employed DNA to induce papillomas to initiate experiments aimed at genetic analysis of viral pathogenesis. In this first phase, we were particularly interested in determining if, and to what extent, the late region could be deleted without impairing the ability of the DNA to induce papillomas in domestic rabbits. Our results indicated that L2 is not required. Deletion of a larger segment in pdlBc/l (nt 4899 to nt 7213) lacking the whole Ll ORF and leaving the long control region (LCR) intact abolished this ability. An equivalent deletion in BPV-1 does not affect its ability to transform NIH 3T3 cells (24). This could indicate that some sequences in the late region of papillomaviruses are not needed for transformation of fibroblasts but are required for papilloma induction. This notion is further supported by the findings with the spontaneous CRPV variant which also had deleted most of the late region but still transformed NIH 3T3 cells (13). There are several ways ORF Ll sequences could affect the potential to induce papillomas. One possibility would be that the Ll protein itself is required. Although late proteins have not been detected in domestic rabbit papillomas and we have not been able to detect late transcripts, a low level of infectivity has been reported in some domestic rabbit papillomas (28). In addition, domestic rabbits inoculated with the transplantable VX7 carcinoma develop an antibody response to the Ll protein (29). Regulatory effects of late viral proteins have been demonstrated for the polyomavirus capsid protein VP1 which induces the cellular oncogenes c-myc and c-fos (30). Similarly, the HSVI tegument protein Vmw 65 initiates the gene expression by direct binding to upstream activator sequences of immediate early genes (31). Another possibility is that a regulatory element(s) located within ORF Ll plays a definite role in papilloma induction but not in transformation of cells in culture. No such regulatory element(s) has been mapped yet in CRPV, but in BPV-1 an
enhancer-like element is located in Ll which plays a role in DNA replication (32, 33). The requirement for this element, however, appears to be conditional since some BPV-1 plasmid constructs lacking this element can be maintained extrachromosomally. Thus it is possible that such an enhancer, if indeed present in CRPV, is essential for induction of papillomas but may not be required for transformation of NIH 3T3 cells in culture. In conclusion, our results presented here have shown that induction of papillomas in a natural host requires the function of genetic elements which may not be needed for transformation of cells in culture.
ACKNOWLEDGMENTS We thank Margerita Tayag for excellent technical help and Jeannie Martinez and Lori Kenton for help with the manuscript. M.N. was the recipient of a postdoctoral fellowship from the Cancer Research Coordinating Committee of the University of California and CM. was a recipient of a predoctoral fellowship from U.S. Public Health Service Grants GM 07104 and Al 07323. The research was supported by Grant CA 1815 1 awarded by the National Cancer Institute, National Institutes for Health, Department of Health and Human Services, and a grant from the W.M. Keck Foundation.
REFERENCES 1. PETO, R., and ZUR HAUSEN, H., Eds. “Banbury Report,” Vol. 21 Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 2. SYVERTON,J. T., Ann. NYAcad. Sci. 54, 1126-l 140 (1952). 3. WETTSTEIN,F. O., and STEVENS,J. G., Proc. Nat/. Acad. Sci. USA 79,790-794 (1982). 4. NASSERI, M., WETTSTEIN, F. O., and STEVENS,J. G., J. Viral. 44, 263-268 (1982). 5. MC VAY, P., FRETZ, M., WETSTEIN, F. O., STEVENS,J. G., and ITO, Y., 1. Gen. Viol. 60,27 l-278 (1982). 6. PHELPS,W. C., LEARY,S. L., and FARAS,A. J., L/irology 146, 120129 (1985). 7. WETTSTEIN, F. O., In “The Papovaviridae” (N. Salzman and P. Howley, Eds.), Vol. 2, Chap. 6. Plenum, New York. 8. WE~TSTEIN,F. O., BARBOSA,M. S., and NASSERI,M., Virology 159, 321-328 (1987). 9. NASSERI,M., and WETTSTEIN,F. O., /. viral. 51, 706-712 (1984). 10. NASSERI, M., and WETTSTEIN, F. O., !&o/ogy 138, 362-367 (1984). 11. GEORGES,E., CROISSANT,O., BONNEAUD,N., and ORTH, G..f. l/ire/. 51,530~538 (1984). 12. DANOS, O., GEORGE% E., ORTH. G., and YANIF. M., J. Viral. 53, 735-741 (1985). 13. NASSERI, M., and WETTSTEIN, F. O., virology 161, 541-548 (1987). 14. SMOTKIN, D., and WE-STEIN, F. O., Proc. Narl. Acad. Sci USA 83,4680-4684 (1986). 15. SCHNEIDER-GADICKE,A., and SCHWARZ, E., EMBO J. 5, 22852292 (1986). 16. INAGAKI,Y., TSUNOKAWA,Y., TAKEBE, N., NAWA, H., NAKANISHI,S., TERADA, M., and SUGIMURA,T., 1. viral. 62, 1640-l 646 (1988). 17. ANDROPHY, E. J., SCHILLER,J. T., and Lowv, D. R., Science 230, 442-445 (1985). 18. SCHLEGEL,R., WADE-GLASS, M., RAESON, M. S., and YANG, Y.-C., Science 233,464-467 (1986).
SHORT COMMUNICATIONS 19. 20. 21. 22. 23. 24. 25.
M., JONES, K., and LAIMINS, L., J. Viral. 61, 3635-3640 (1987). KANDA, T., FURUNO, A., and YOSHIIKE, Y., 1. Viral. 62, 61 O-61 3 (1988). MATLASHEWSKI,G., SCHNEIDER,J., BANKS, L., JONES, N., MURRAY, A., and CRAWFORD,L., EMBOI. 6,1741-1746 (1987). PHELPS, W. C., LEE, C. L., MUNGEA. K.. and HOWLEY, P. M., Cell 53,539-547 (1988). STOREY, A., PIM, D., MURRAY, A., OSEIORN, K., BANKS, L.. and CRAWFORD,L., fMBO/. 7,1815-1820 (1988). Lowv, D. R., DVORETZKY,I., SHOBER, R.. !-Aw, M.-F., ENGEL. L., and HOWLEY, P. M., Nature (London) 287,72-74 (1980). GIRI, I., DANOS, O., and YANIF, M., Proc. Nat/. Acad. Sci. USA 82, 1580-l 584 (1985). BEDELL,
325
26. ITO,Y.,~~~EVANS,C.A.,J.&J. Med. 114,485-500(1961). 27. MELLON, P., PARKER,V., GLUZMAN, Y., and MANIATIS, T., Ce//27, 279-288 (1981). 28. SHOPE, R. E., Proc. Sot. Exp. Med. 32, 830 (1935). 29. ORTH, G., BREITBARD,F., and FAVRE, M., Virology 91, 243-255 (1978). 30. ZULLO, J., STILES, C. D., and GARCEA, R. L., Proc. Nat/. Acad. Sci, USA84, 1210-1214(1987). 31. O’HARE, P., and GODING, C. R., Ce//52,435-445 (1988). 32. LUSKY, M.. and BOTCHAN, M. R.. Proc. Nat/. Acad. Sci. USA 83, 3609-3613 (1986). 33. STENLUND, A., BREAM, G. L., and BOTCHAN, M. R., Science 236, 1666-1671 (1987).