We investigated the use of infectious cDNA for the production of poliovirus type 1-type 3 recombinants. ... molecular basis of attenuation in the poliovirus vaccine.
Vol. 57, No. 3
JOURNAL OF VIROLOGY, Mar. 1986, p. 1187-1190
0022-538X/86/031187-04$02.00/0 Copyright C 1986, American Society for Microbiology
Construction of Poliovirus Intertypic Recombinants by Use of cDNA GLYN STANWAY,l* PAMELA J. HUGHES,' GARETH D. WESTROP,' DAVID M. A. EVANS,2 GLYNIS DUNN,2 PHILIP D. MINOR,2 GEOFFREY C. SCHILD,2 AND JEFFREY W. ALMOND'
Department of Microbiology, University of Leicester, Leicester, LE] 7RH,1 and National Institute for Biological Standards and Control, London, NW3 6RB,2 United Kingdom Received 5 July 1985/Accepted 16 October 1985
We investigated the use of infectious cDNA for the production of poliovirus type 1-type 3 recombinants. One such recombinant virus was produced, but a second construct involving the transfer of part of the capsid protein region was not infectious. Our results suggest that the approach may prove valuable but that not all cDNA constructs will give rise to viable viruses.
Attenuated derivatives of each of the three poliovirus serotypes were developed by A. Sabin in the 1950s, and they are used as the seed for most of the trivalent oral polio vaccines (20). However, it is known that the type 2 and 3 strains are genetically unstable, and there is strong evidence that they can revert to a neurovirulent phenotype and produce paralytic disease in vaccine recipients or their contacts (4, 11, 27). The type 1 vaccine strain seems to be much more stable, and, therefore, it may be possible to produce safer poliovirus type 2 and 3 vaccines by making recombinants which possess the appropriate antigenic properties but which contain regions of the type 1 genome conferring the stably attenuated phenotype. The complete nucleotide sequence of the vaccine strain of each of the poliovirus serotypes has been determined, and 70% of the nucleotides are common to all three serotypes (22, 26). cDNAs of the three types therefore contain a number of restriction enzyme sites in common that can be used for recombination. In this communication, we describe the use of full-length cDNAs for the production of recombinants from the poliovirus type 1 and 3 Sabin vaccine strains, P1/LSc, 2ab and P3/Leon 12alb. We first explored whether the two types of poliovirus have sufficient functional homology for specific recombinants to be made between them. The 5' nontranslated region was chosen for replacement because this region of about 740 nucleotides, although of unknown function, is highly conserved between the serotypes and, indeed, between members of different picornavirus genera (22, 23, 26). There is, however, a region of 120 nucleotides, just before the start of the large open reading frame, which has greatly diverged between the types and which may therefore be involved in some type-specific function which renders certain recombinant viruses nonviable. In addition, our recent work has indicated that a single mutation in the 5' nontranslated region of the genome of the poliovirus type 3 vaccine strain is rapidly selected for upon passage in the gut of vaccine recipients and that virus apparently containing only this mutation possesses significantly increased neurovirulence (7). This indicates that this region plays some role in the attenuation and reversion to neurovirulence of the type 3 vaccine strain. It is possible that such reversion would be overcome by the production of specific intertypic recombinants. When these experiments commenced, we did not have a full-length, infectious cDNA from P3/Leon 12alb, and to test the applicability of the method we constructed the
The three poliovirus serotypes are the causative agents of poliomyelitis, a disease which is now controlled effectively in developed countries through the use of live-attenuated or killed vaccines (16). The polioviruses are typical members of the Picornaviridae family, possessing a single-stranded RNA genome of positive polarity, approximately 7,500 nucleotides in length (17). The viruses have been studied extensively, and biochemical work has shed a great deal of light on the details of their replicative cycle (19). In contrast, genetic studies have proved less productive, largely because the nature of the poliovirus genome makes it difficult to produce and analyze mutant and recombinant viruses. Although it has been shown that picornaviruses are capable of recombination in vivo and coinfection of cells with viruses carrying selectable markers, such as antigenicity and guanidine resistance, has been used to produce recombinants of footand-mouth disease virus (9) and poliovirus (2, 6, 25), these recombinants have been of limited practical use. Variation in recombination frequency means that the location of markers is inaccurate, and, because the crossover point cannot be accurately controlled, it is impossible to produce specific recombinants (5). In 1981, Racaniello and Baltimore (18) reported that a complete cDNA copy of the Mahoney strain of poliovirus type 1 is infectious when transfected into susceptible cells. This observation had profound implications for the study of picornavirus genetics, because it opened up the possibility of circumventing the above difficulties to produce specifically designed mutant or recombinant viruses by transfecting cells with cDNA which has been modified by DNA manipulation techniques. One of the simplest applications is the use of conserved restriction enzyme sites to produce recombinants between closely related virus strains. We and others have constructed full-length cDNA copies of several strains of poliovirus types 1 and 3 and shown them to be infectious (14, 18, 21, 24; unpublished data). These have already been used to construct interstrain recombinants for the analysis of the molecular basis of attenuation in the poliovirus vaccine strains (10; unpublished data) and to confirm the presence of a second neutralizing antigenic site in poliovirus type 3 (12). We are also interested in producing intertypic recombinants between the three poliovirus serotypes, because they may be of use for the further study of antigenicity. Such recombinants may also have potential as novel poliovirus vaccines. *
Corresponding author. 1187
1188
NOTES
J. VIROL.
Translation M. G A V S S O K V 6 A M E ype i CAAUUGUAUCAUAA5GGUGCUCAGGUUUCAUCACA6AAAGU66iCCSACM,GAA Ty,p 3 UUUCAGUGUCACAA#J6GAtCUCAA6UAUCAUCCCAAAAAGUA6iCSCUCACGA6 CFo I FIG. 1. Nucleotide and amino acid sequence of the type 1 (P1/LSc, 2ab) and type 3 (P3/Leon 12a,b) Sabin vaccine strains of poliovirus in the region of the translation start (22, 26). The open reading frame is preceded by a long untranslated region, nucleotide 1 of the AUG codon of P3/Leon 12alb being located at position 743 in the genomic sequence.
reciprocal recombinant. P1/LSc, 2ab RNA was cloned by the cDNA:RNA hybrid method (3, 24), and the resultant subgenomic clones were used to construct a full-length infectious cDNA, pOLIO 1 (unpublished data). The cDNA corresponding to the 5' noncoding region was then replaced with P3/Leon 12a1b cDNA. The nucleotide sequence of P1/LSc, 2ab and P3/Leon 12a1b cDNA around position 743, the start of the long open reading frame, is shown in Fig. 1. By making use of the conserved CfoI site, it is possible to produce a poliovirus type 1-type 3 cDNA chimera in the plasmid pAT 153 without perturbing the amino acid sequence of the type 1 polyprotein. The complete construction of the full-length recombinant cDNA, pOT 9, is shown in Fig. 2. HEp-2c cells were grown to 60% confluence in 9-cm plastic dishes and transfected with 20 ,ug of pOT 9 per plate by the calcium phosphate technique using a 20% glycerol shock (15). Cytopathic effect was observed after 5 to 7 days of incubation at 34°C, and viral RNA was prepared for sequencing after one plaque passage. The identity of the recombinant virus (OT 9) was confirmed by sequencing using two synthetic oligonucleotides to prime reverse transcription of the RNA in the presence of dideoxynucleotides. One of these was complementary to the sequence 510 to 522 (in the 5' noncoding region) of type 1 and 3 viral RNA, and one was complementary to the sequence 3006 to 3020 (in the capsid protein region) of type 1 and 3 viral RNA. The sequence obtained indicated that the virus was type 3-like in the 5' noncoding region and type 1-like in the capsid protein region and thus that the expected type 1-type 3 intertypic recombinant had been produced (data not shown). The successful production of OT 9 led us to assess the suitability of the cDNA method for the construction of other intertypic recombinants. A major antigenic site for the neutralization of poliovirus type 3 has been located between amino acid positions 89 and 100 of capsid protein VP1 (8, 13). One attractive possibility is to substitute this region of P3/Leon 12a1b for the corresponding region of P1/LSc, 2ab to produce a virus which is predominantly type 1 and thus presumably stable with respect to attenuation, but which has the major antigenic determinant of type 3. Between the two types, conserved Hinfl and SphI sites 228 nucleotides apart flank the cDNA corresponding to this region. In addition to the seven amino acid changes in the antigenic site, substitution of this region introduces a further seven amino acid changes. Initially, a PstI fragment of type 1 cDNA (position 2243 to 3417) was subcloned into pAT 153. The Hinfl-SphI fragment was then replaced by type 3 cDNA, as shown in Fig. 3, to give the intermediate pOT 1. The desired fulllength product, pOT 5, was obtained by ligating the PstI fragment of pOT 1 into pOLIO 1 partially digested with PstI. pOT 5 was transfected into HEp-2c cells as described above. Although several transfections were performed, each in parallel with the transfection of a control full-length cDNA which yielded virus, pOT 5 always failed to give a cytopathic effect, indicating that no virus was produced. It is possible that the noninfectious nature of pOT 5 is due to the intro-
duction of errors into the cDNA during consiruction, particularly because the manipulations were relatively complex, and it was therefore analyzed extensively. Regtriction enzyme maps were exactly as expected, indicating that there were no gross errors and that restriction enzyme sites used for the construction were restored upon ligation. Furthermore, the whole of the chimeric 1,174-base-pair PstI fragment upon which all the manipulations were performed (Fig. 3) was found to have the correct nucleotide sequence. This
SABIN 3 Pst
I
c
c
SABIN 1 Ava I
PstI
PstI
c
c
pAT l53xPst I
I
...
......
... ..
..
...
\ ru
pOLIO
I
FIG. 2. The 5'-terminal 1,314 nucleotides of P3/Leon 12alb cDNA is contained within a PstI-AvaI fragment. Partial CfoI digestion produced the PstI-Cfol fragment indicated. The 5' 1,809 nucleotides of P1/LSc, 2ab cDNA is contained within a PstI fragment, and CfoI digestion gives the CfoI-PstI fragment indicated. The two fragments were ligated, together with PstI-digested, phosphatasetreated plasmid pAT 153, and the product was used to transform competent Escherichia coli JA221 cells. The chimera pH 1 was selected fom the other possible products (head-to-head dimers or the chimera in the wrong orientation) by restriction mapping of the Tetr clones produced. pOT 9, a full-length cDNA chimera, was then made from pH 1 and pOLIO 1 (full-length P1/LSc, 2ab cDNA) by digestion at the two NruI sites, followed by ligation of the appropriate fragments.
NOTES
VOL. 57. 1986 0
I1 VV4
VP?9
I
p
cDNA p
~
r-P2-3bP25 P-
VPI
pP
-P3-lb
1 .2PR O°
LAS E
PYERASE
p H HS
S
P
\~~~~~
p
FIG. 3. cDNA representing most of the region of P1/LSc, 2ab coding for VPI is contained within a 1,174-base-pair Pstl fragment (position 2243 to 3417). The region encoding the major neutralizing antigenic site of poliovirus type 3 (8, 13) is encoded within this fragment and flanked by Hinfl (position 2695) and SphI (position 2923) sites conserved between P1/LSc, 2ab and P3/Leon 12a,b. The PstI fragment was excised from the full-length P1/LSc, 2ab clone pOLIO 1 and subcloned into Pst 1-digested pAT 153 to give pO 1. In addition, the fragment was digested with SplI, and the appropriate PstI-Sphl fragment was isolated and subcloned into PstI-digested pAT 153 to give a head-to-head dimer, pS 1. This was linearized by partial digestion with Sphl and then digested with BarnHI, and the indicated fragment was isolated. Complete BamHI and Hinfl digestion of pO 1 gave the fragment indicated. The 228-base-pair type 3 component was obtained by complete Sphl and partial Hinfl digestion of the cDNA insert of subclone pSGA 28 (3). The three fragments were ligated together, and the ligation mixture was used to transform E. coli JA221. The correct constructs, pOT 1, which were found to make up about 75% of the 200 clones obtained, were identified by restriction enzyme analysis. pOLIO 1 was partially digested with Pstd, and the full-length plasmid from which the 1,174-base-pair PstI fragment had been removed was isolated. The full-length chimera, pOT 5, was then produced by ligation of the PstI-digested pOLIO 1 and the purified PstI insert from pOT 1. pOT 1 is therefore full-length P1/LSc, 2ab cDNA in which nucleotides 2695 to 2923 have been replaced by the equivalent region from P3/Leon 12a,b, part of which is known to specify the major neutralizing antigenic site (8, 13).
intertypic poliovirus recombinants. However, the noninfectious nature of pOT 5 suggests that not all recombinants will be viable, owing to the detailed interaction of the virus proteins or RNA or both, and, this may limit the applicability of the approach. Recombinants have been produced in the past by mixed infection of cells with viruses carrying selectable markers, and they have been used in attempts to locate the attenuating mutations in poliovirus (1, 2). However, recombination seems to occur usually around the middle of the genome, and the absence of a range of crossover points suggests that this method is of limited usefulness (2, 6). Furthermore, it is difficult to see how fine control over the crossover point could be effected so as to produce truly specific recombinants. The specificity of the infectious cDNA approach therefore makes this a potentially important technique for the study of polioviruses. To increase the power of the method, however, it would be useful to know more in detail about the protein-protein or protein-RNA interactions within the poliovirus capsid. As information on such interactions becomes available, for example from X-ray crystallographic studies, it may be possible to predict which recombinants would be likely to give rise to viable viruses. This would then circumvent the problem of performing several necessarily complex manipulations only to find the final construct noninfectious. Given such a predictive capability, the cDNA method should greatly facilitate the study of the biological properties of the polioviruses. Indeed, in view of their extensive nucleotide homology (23), it may even be possible to produce chimeric viruses between different members of the enterovirus and rhinovirus genera of Picornasiridae, and these may be of use in defining the molecular basis of biological characteristics, such as different temperature optima, acid lability, host range, and
tissue tropism. We are grateful for financial support from the Medical Research Council of Great Britain (grant G83242556CB). LITERATURE CITED 1. Agol, V. I., S. G. Drozdov, M. P. Frolova, V. P. Grachev, MI. S. Kolesnikova, V. G. Kozlov, N. M. Ralph, L. 1. Romanova, E. A. Tolskaya, and E. G. Victorova. 1985. Neurovirulence of the intertypic poliovirus recombinant v3/al-25: characterization of
strains isolated from the spinal cord of diseased monkeys and evaluation of the contribution of the 3' half of the genome. J. Gen. Virol. 65:309-316. 2. Agol, V. I., V. P. Grachev, S. G. Drozdov, M. S. Kolesnikova,
V. G. Kozlov, N. M. Ralph, L. I. Romanova, E. A. Tolskaya, A. V. Tyufanov, and E. G. Viktorova. 1984. Construction and properties of intertypic poliovirus recombinants: first approxi-
mation mapping of the major determinants of neurovirulence. Virology 136:41-55. 3. Cann, A. J., G. Stanway, R. Hauptmann, P. D. Minor, G. C. Schild, L. D. Clarke, R. C. Mountford, and J. W. Almond. 1983.
Poliovirus type 3: molecular cloning of the genome and
strongly suggests that the nonviability was not due to an artifact generated during manipulation but was inherent in the final construction. Presumably, the protein-protein interactions in the poliovirus capsid are so specific that even a small number of amino acid changes, such as were transferred in pOT 5, are sufficient to prevent correct assembly or functioning of the capsid. We are now attempting to transfer larger and smaller regions of the type 3 capsid into type 1 to circumvent this problem. In conclusion, the results presented here indicate that the transfection of recombinant cDNA can be used to produce
1189
nuLcle-
otide sequence of the region encoding the protease and polymerase proteins. Nucleic Acids. Res. 11:1267-1281.
4. Cann, A. J., G. Stanway, P. J. Hughes, P. D. Minor, D. M. A. Evans, G. C. Schild, and J. W. Almond. 1984. Reversion to
neurovirulence of the live-attenuated Sabin type 3 oral poliovirus vaccine. Nucleic Acids Res. 12:7787-7792. 5. Cooper, P. D. 1977. Genetics of picornaviruses, p. 133-207. In
H. Fraenkel-Conrat and R. Wagner (ed.). Comprehensive virology, vol. 9. Plenum Publishing Corp., New York. 6. Emini, E. A., J. Leibowitz, D. C. Diamond, J. Bonin, and E. Wimmer. 1984. Recombinants of Mahoney and Sabin strain poliovirus type 1: analysis of in vitro phenotypic markers and evidence that resistance to guanidine maps in the non-structural
1190
NOTES
proteins. Virology 137:74-85. 7. Evans, D. M. A., G. Dunn, P. D. Minor, G. C. Schild, A. J. Cann, G. Stanway, J. W. Almond, K. Currey, and J. V. Maizel, Jr. 1985. A single nucleotide change in the 5' non-coding region of the genome of the Sabin type 3 poliovaccine is associated with increased neurovirulence. Nature (London) 314:548-550. 8. Evans, D. M. A., P. D. Minor, G. C. Schild, and J. W. Almond. 1983. Critical role of an eight-amino acid sequence of VP1 in neutralization of poliovirus type 3. Nature (London) 304:459-462. 9. King, A. M. Q., D. McCahon, W. R. Slade, and J. W. I. Newman. 1982. Recombination in RNA. Cell 29:921-928. 10. Kohara, M., T. Omata, A. Kameda, B. L. Semler, H. Itoh, E. Wimmer, and A. Nomoto. 1985. In vitro phenotypic markers of a poliovirus recombinant constructed from infectious cDNA clones of a neurovirulent Mahoney strain and the attenuated Sabin 1 strain. J. Virol. 53:786-792. 11. Minor, P. D. 1982. Characterization of strains of type 3 poliovirus by oligonucleotide mapping. J. Gen. Virol. 59:307-317. 12. Minor, P. D., D. M. A. Evans, M. Ferguson, G. C. Schild, G. Westrop, and J. W. Almond. 1985. Principal and subsidiary antigenic sites of VP1 involved in the neutralization of poliovirus type 3. J. Gen. Virol. 66:1159-1165. 13. Minor, P. D., G. C. Schild, J. Bootman, D. M. A. Evans, M. Ferguson, P. Reeve, M. Spitz, G. Stanway, A. J. Cann, R. Hauptmann, L. D. Clarke, R. C. Mountford, and J. W. Almond. 1983. Location and primary structure of a major antigenic site for poliovirus neutralization. Nature (London) 301:674-679. 14. Omata, T., M. Kohara, Y. Sakai, A. Kameda, N. Imura, and A. Nomoto. 1984. Cloned infectious complementary DNA of the poliovirus Sabin 1 genome: biochemical and biological properties of the recovered virus. Gene 32:1-10. 15. Parker, B. A., and G. R. Stark. 1979. Regulation of simian virus 40 transcription: sensitive analysis of the RNA species present early in infections by virus or viral DNA. J. Virol. 31:360-369. 16. Paul, J. R. 1971. A history of poliomyelitis. Yale University Press, New Haven, Conn. 17. Putnak, J. R., and B. A. Phillips. 1981. Picornaviral structure
J. VIROL.
and assembly. Microbiol. Rev. 45:287-315. 18. Racaniello, V. R., and D. Baltimore. 1981. Cloned poliovirus complementary DNA is infectious in mammalian cells. Science 214:916-919. 19. Rueckert, R. R. 1976. On the structure and morphogenesis of picomaviruses, p. 131-213. In H. Fraenkel-Conrat and R. Wagner (ed.), Comprehensive virology, vol. 6. Plenum Publishing Corp., New York. 20. Sabin, A., and L. Boulger. 1973. History of Sabin attenuated poliovirus oral live vaccine strains. J. Biol. Stand. 1:115-118. 21. Semler, B. L., A. J. Dorner, and E. Wimmer. 1984. Production of infectious poliovirus from cloned cDNA is dramatically increased by SV40 transcription and replication signals. Nucleic Acids Res. 12:5123-5141. 22. Stanway, G., A. J. Cann, R. Hauptmann, P. Hughes, L. D. Clarke, R. C. Mountford, P. D. Minor, G. C. Schild, and J. W. Almond. 1983. The nucleotide sequence of poliovirus type 3 Leon 12alb: comparison with poliovirus type 1. Nucleic Acids Res. 11:5629-5643. 23. Stanway, G., P. J. Hughes, R. C. Mountford, P. D. Minor, and J. W. Almond. 1984. The complete nucleotide sequence of a common cold virus: human rhinovirus 14. Nucleic Acids Res. 12:7859-7875. 24. Stanway, G., R. C. Mountford, S. D. J. Cox, G. C. Schild, P. D. Minor, and J. W. Almond. 1984. Molecular cloning of the genomes of poliovirus type 3 strains by the cDNA:RNA hybrid method. Arch. Virol. 81:67-78. 25. Tolskaya, E. A., L. A. Romanova, M. S. Kolesnikova, and V. I. Agol. 1983. Intertypic recombination in poliovirus: genetic and biochemical studies. Virology 124:121-132. 26. Toyoda, H., M. Kohara, Y. Kataoka, T. Suganuma, T. Omata, N. Imura, and A. Nomoto. 1984. Complete nucleotide sequences of all three poliovirus serotype genomes: implication for genetic relationship, gene function and antigenic determinants. J. Mol. Biol. 174:561-585. 27. World Health Organization. 1981. Markers of poliovirus strains isolated from cases temporally associated with the use of live poliovirus vaccine: report of a WHO collaborative study. J. Biol. Stand. 9:163-184.
