Previous Article | Next Article ![]()
Journal of Virology, September 2008, p. 8927-8932, Vol. 82, No. 17
0022-538X/08/$08.00+0 doi:10.1128/JVI.00239-08
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Unité de Biologie des Virus Entériques,1 Unité de Pharmacoépidémiologie et Maladies Infectieuses, Institut Pasteur, Paris, France,2 Unité de Virologie Médicale, Institut Pasteur de Madagascar, Antananarivo, Madagascar3
Received 2 February 2008/ Accepted 9 June 2008
|
|
|---|
|
|
|---|
Recent poliomyelitis outbreaks due to circulating VDPVs (cVDPVs) have brought to the forefront the depth of enterovirus evolution and the influence it may have on the eradication program (14). cVDPV outbreaks occurred subsequently to a period of low vaccine coverage in nine different countries, including Madagascar (8, 12, 17, 31, 32, 37-39). All cVDPV outbreaks but one were due to recombinants originating from OPV strains; large parts of the genomic regions encoding the nonstructural proteins in these strains and the 3' untranslated region (UTR), and in some cases the 5' UTR, were derived from HEV species C (HEV-C) (8, 12, 17, 31, 32, 37-39), enteroviruses that are phylogenetically closely related to PVs (4). Genetic analyses of recombinant cVDPV strains and cocirculating enteroviruses have recently shown that HEV-C isolates (coxsackievirus strains A13 and A17) exhibit highest similarity with these cVDPVs in the genomic regions encoding the nonstructural proteins and in the 3' UTR (3, 26).
Recombination is a common evolution mechanism for enteroviruses that is thought to contribute at least to the repair of deleterious mutations in genomes (1). However, the role of interspecific genetic exchanges in the evolution and functions of OPV/HEV-C recombinants is still unclear. Two type 2 recombinant cVDPV lineages with sequences from the OPV type 2 strain in the 5' half of the genomes and with sequences derived from HEV-C in the 3' half have been isolated from children with poliomyelitis in Madagascar in 2002 (31). These cVDPV strains showed phenotypic characteristics similar to those of wild-type neurovirulent viruses, including neurovirulence in PV receptor-transgenic (PVR-Tg) mice (26). To investigate the impact of genetic exchanges between PV and HEV-C species on the characteristics of recombinant viruses, we generated and characterized chimeric viruses with various crossover regions between one type 2 Madagascan cVDPV isolate (MAD04) and the original Sabin OPV type 2 strain (Sabin2).
Nomenclature for viruses.
Natural viruses were Sabin2 (WHO laboratory stock of Sabin2) and MAD04 (accession number AM084223). The viral stock used in this experiment was obtained following the amplification of a plaque-purified virus derived from the original MAD04 isolate described in reference 26. Although this plaque-purified virus differs from the original isolate by a few nucleotides (nt) (nt 704 C
T, 949 C
T, 3576 T
C, 5334 A
G, 6897 C
T, 6910 T
C [Sabin2 sequence numbering]), it exhibited similar phenotypic characteristics.
Viruses derived from cDNA (MAD04 series) were strains cS2 and cMAD04 for Sabin2 and MAD04, respectively (parental), and strains cS2/MAD04, cMAD04/S2, and cMAD042A/S2 (chimeric).
Construction of cDNA-derived viruses. Taking into account the natural recombination site between the vaccine-derived part and the nonvaccine (HEV-C) part of the MAD04 genome in the protease 2A coding region (nearby nt 3801), infectious viral cDNAs were cloned and three recombinants were constructed (Fig. 1). Briefly, PCR amplifications of four half-viral genomes (from Sabin2 and MAD04) were performed using (i) MAD04 viral RNA extracted by the Trizol reagent (Invitrogen) from MAD04 viral particles purified by ultracentrifugation or plasmid pT7S2 (a plasmid harboring the whole Sabin2 cDNA; generous gift of A. Macadam) as a template, (ii) specific primers flanked with either an AscI restriction site and the T7 RNApol promoter sequence at the 5' end or a NotI restriction site at the 3' end of the genome, and (iii) oligonucleotide primers on either side of a BssSI restriction site (nt 3303; 3' end of the VP1 region, codon 274). Fragments were then TA cloned into the pGEM-T vector (Promega) and fully sequenced. No differences in the 5' UTR and 3' UTR sequences or in the viral polypeptides inferred from the coding regions between natural and cDNA-derived viruses were observed. The half-viral genomes were then sequentially subcloned into a modified pBR322 vector containing a custom-made cloning cassette (HindIII-AscI-BssSI-NotI-EcoRI). Using the BssSI restriction site, we generated the parental full-length cDNAs, cS2 and cMAD04, and the chimeric ones, cS2/MAD04 and cMAD04/S2 (Fig. 1). The chimeric cDNA cMAD042A/S2 (recombination site, nt 3827; protease 2A codon 148) was generated by replacing in cMAD04/S2 a BssSI-BseRI restriction fragment (nt positions 3,303 to 3,827) amplified from cMAD04. Recombination did not modify the whole VP1 and 2A polypeptide sequences found in MAD04 and Sabin2. Constructs and mutations were verified by DNA sequencing.
![]() View larger version (46K): [in a new window] |
FIG. 1. Genomic structure of in vitro-generated parental and chimeric viruses. (A) The genetic organization of the PV genome is shown (top row), including the 5' and 3' UTRs. Genomic regions encoding viral proteins (VP4 to 3Dpol) and their precursors P1 to P3 are indicated. (B) The cDNA-derived parental and chimeric viruses of the MAD04 series are shown. The cDNA-derived Sabin2 (cS2) and MAD04 viruses (cMAD04) are presented. The mutated vaccine-derived part (the 5' half) and the HEV-C part (the 3' half) of the cMAD04 genome are indicated together with the natural recombinant site (filled triangles). This natural site and another intentionally determined site located at the boundary between the capsid and noncapsid genomic regions (open triangles) were used for constructing chimeric cDNAs (cS2/MAD04, cMAD04/S2, and cMAD042A/S2). (C) Nucleotide sequence alignments showing the natural and engineered recombination sites. Nucleotide numbering is indicated according to the Sabin2 genome. Nucleotide positions of the natural recombination site in the MAD04 isolate (approximate position) (white dotted line) and of the recognition sequences of the restriction enzymes used for engineering chimeric cDNAs (horizontal white solid lines) and their respective cleavage sites (white vertical lines) are given. Nucleotides differentiating cS2 from cMAD04 are in white.
|
Phenotypic characteristics of chimeric viruses. Phenotypic characteristics of parental cDNA-derived viruses were compared to those of natural viruses. No significant differences were observed. For each construct used, two different cDNA-derived viruses obtained from separate cDNA clones were tested.
Replication kinetics of natural and MAD04 series viruses were compared in HEp-2c cells (6). For single-step growth experiments, cells were infected at a multiplicity of infection (MOI) of 25 50% tissue culture infectious doses (TCID50) per cell and incubated at 37°C, and viral titers (TCID50/ml) were determined at various time points postinfection (6) (Fig. 2A and B). Overall, virus replication kinetics between natural and cDNA-derived viruses (Fig. 2A) are very similar, and growth curves of chimeric viruses were rather similar to those of both parental viruses (Fig. 2B). Moreover, results using a low MOI (1 TCID50 per cell) do not indicate major differences between all viruses (Fig. 2C). Although the MAD04 virus appears to replicate slightly better than the other viruses (Fig. 2B) and the growth of cMAD04/S2 and cMAD042A/S2 appeared slightly delayed compared to cS2, cS2/MAD04, and cMAD04 (Fig. 2C), the differences do not appear to be significant according to longitudinal data analysis. These results indicated that recombination exerts only a mild effect or no effect at all on viral growth in HEp-2c cells.
![]() View larger version (24K): [in a new window] |
FIG. 2. Growth curves of parental and chimeric viruses. HEp-2c cells were infected with the indicated viruses, cells and supernatants were harvested at various time points, and each sample was titrated by a microtitration method. Standard errors from the means of different samples are shown as error bars. (A) Natural viruses (Sabin2 and MAD04) and parental cDNA-derived viruses (cS2 and cMAD04). (B and C) MAD04 series viruses. Whereas growth curves for panels A and B were obtained following infection of HEp-2C cells at a high MOI (25 TCID50 per cell for single-step growth curves), curves for panel C were obtained following infection at a low MOI (1 TCID50 per cell). (B) Mean values obtained from two different experiments using two batches of MAD04 series viruses derived from different clones of the same plasmidic chimeric genomic cDNA constructs.
|
|
View this table: [in a new window] |
TABLE 1. Phenotypic markers of the parental and chimeric PV strains
|
Phenotypic characterization of the cDNA-derived MAD04 series indicated that determinants located in both the vaccine part and the nonvaccine part of the MAD04 genome contribute to the temperature sensitivity and plaque size phenotypes and may slightly influence its replication capacity. However, we cannot exclude that these aspects result from cooperative functional interactions between both genomic parts and/or corresponding viral proteins.
Neurovirulence of chimeric viruses. The pathogenicity of MAD04 series viruses was first evaluated by determining the paralytic dose affecting 50% of the inoculated mice (Table 1). This was achieved by intracerebral inoculation of serial dilutions of viral stocks into PVR-Tg21 mice genetically modified to express the human PV receptor gene (generous gift from A. Nomoto; experiments were approved by and conducted in accordance with the guidelines of the Office of Laboratory Animal Care at the Institut Pasteur) (10, 26). Virus cMAD04 showed potent neurovirulence that was comparable to that of the virulent S2/4568, a highly neurovirulent type 2 VDPV isolated from environmental waters (33). In contrast, the cMAD04/S2 and cMAD042A/S2 viruses exhibited only moderate neurovirulence in PVR-Tg21 mice. Finally, the cS2/MAD04 virus did not induce paralysis in mice inoculated with the highest doses (6.6 TCID50/mouse), thus showing an attenuated phenotype similar to that of cS2. Similar differences between neurovirulence levels were obtained when PVR-Tg21 mice were inoculated, via the intranasal (IN) and intraperitoneal routes (IP), with given amounts of parental and chimeric viruses (Fig. 3). Groups of mice inoculated with cMAD04/S2 and cMAD042A/S2 were significantly less affected than those that received S2/4568 and cMAD04 (P < 0.01 and P < 0.05 for mice inoculated via IN and IP routes, respectively [log rank test]).
![]() View larger version (22K): [in a new window] |
FIG. 3. Neurovirulence of MAD04 series viruses in Tg mice. A given dose of virus was inoculated via IN (A) or IP (B) routes in groups of PVR-Tg21 mice expressing the human PV receptor (6 to 16 animals per virus). Animals were checked daily for 21 days for the appearance of paralysis or death. Healthy-mouse ratios following inoculation of either parental or chimeric viruses or the neurovirulent positive-control virus S2/4568 (33) are represented. No additional paralyzed or dead mice were detected after day 17 (A) and day 9 (B) postinoculation.
|
G and aa 143 Ile
Thr, respectively) as well as in previously described pathogenic Sabin2 mutants (5, 7, 19, 26, 28). However, our results indicated that the HEV-C-derived 3' half of the MAD04 genome modulates pathogenicity maybe by slightly improving the capacity of the virus to replicate in certain tissues. Actually, it has been previously shown that the P3-3' UTR regions of PV, coxsackievirus B3, and enterovirus 71 strains contribute to pathogenicity in animal models and/or to temperature sensitivity (2, 9, 22, 35). Comparative analysis of cMAD04/S2 and cMAD042A/S2 properties suggested that protease 2A (P2 region) plays only a minor role or no role at all in the pathogenicity and in other characteristics of MAD04, despite a few aa residues differing from those of Sabin2 (26). Impact of genetic exchanges on recombinant PV/HEV-C. Recently, the characterization of recombinants between the highly pathogenic PV strain Mahoney and a prototype HEV-C (coxsackievirus A20) suggested that the P2 or P3 region of the PV genome can be exchanged with that of coxsackievirus A20 without any effect on viral multiplication or, at least in some cases, on pathogenicity for PVR-Tg mice (11). This suggests that recombination between PV and HEV-C can in some cases be neutral. Our experiments indicate that regardless of the evolution process following genetic recombination events, recombination between OPV strains and HEV-C brings together different genomic fragments that may contribute through their intrinsic properties and their interactions with the specific features of recombinant cVDPVs. The differences between our results and those mentioned above (11) can be explained by some parameters, including the nature of the PV genomes used in each study of either wild-type or vaccine origin and the HEV-C sequences used. Based on this study, recombination of PVs with other enteroviruses appeared to be not solely an indicator of intense circulation of the vaccine strains (15) but also an evolutionary process that may influence crucial viral properties, such as circulation and neurovirulence. Interspecific genetic recombination with HEV-C contributes to both genetic and phenotypic biodiversities of PVs and may be responsible, in part, for the emergence of cVDPVs.
This study was partly supported by grants from the Direction of Foreign Affairs, the Transverse Research Programs (PTR120 and PTR276) of the Institut Pasteur, and the French Ministry of Foreign Affairs (FSP 2001-168).
Published ahead of print on 25 June 2008. ![]()
|
|
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»