The Genome Length of Human Parainfluenza Virus Type 2 Follows the Rule of Six, and Recombinant Viruses Recovered from Non-Polyhexameric-Length Antigen
Respiratory Viruses Section, Laboratory of Infectious Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland 20892$2:^{8/, 百拇医药
Received 29 July 2002/ Accepted 1 October 2002$2:^{8/, 百拇医药
ABSTRACT$2:^{8/, 百拇医药
Members of the Paramyxovirinae subfamily of the Paramyxoviridae family of viruses have the unusual requirement that the nucleotide length of the viral genome must be an even multiple of six in order for efficient RNA replication, and hence virus replication, to occur. Human parainfluenza virus type 2 (HPIV2) is the only member of the genus that has been reported to have a genome length that is not an even multiple of six, and it has also been recovered from a full-length antigenomic-sense cDNA that did not conform to the "rule of six." To reexamine the issue of nucleotide length in natural isolates of HPIV2, a complete consensus genomic sequence was determined for three HPIV2 strains: Greer, Vanderbilt/1994 (V94), and Vanderbilt/1998. Each of these strains was found to have a genome length of 15,654 nucleotides (nt), thus conforming in each case to the rule of six. To directly examine the requirement that the genomic length of HPIV2 be an even multiple of six, we constructed six full-length antigenomic HPIV2/V94 cDNAs that deviated from a polyhexameric length by 0 to 5 nt. Recombinant HPIV2s were readily recovered from all of the cDNAs, including those that did not conform to the rule of six. One recombinant HPIV2 isolate was completely sequenced for each of the nonpolyhexameric antigenomic cDNAs. These were found to contain small nucleotide insertions or deletions that conferred polyhexameric length to the recovered genome. Interestingly, almost all of the length corrections occurred within the hemagglutinin-neuraminidase and large polymerase genes or the intervening intergenic region and thus were proximal to the insert that caused the deviation from the rule of six. These results demonstrate, in the context of complete infectious virus, that HPIV2 has a strong and seemingly absolute requirement for a polyhexameric genome.
INTRODUCTION/fu, 百拇医药
Human parainfluenza virus type 2 (HPIV2) is an enveloped single-stranded negative-sense RNA virus of the genus Rubulavirus in the family Paramyxoviridae. Other rubulaviruses include simian virus 41 (SV41), SV5, HPIV4 A and B, and mumps virus. Like HPIV3 and HPIV1, which are members of the genus Respiroviridae in the paramyxovirus family, HPIV2 is a major cause of acute respiratory tract disease in infants and young children (4). HPIV2 has a genomic organization that is similar to that of the other parainfluenza viruses (4, 24), encoding seven proteins from six mRNAs. The HPIV2 gene map, based on the encoded mRNAs, is 3'-N-P/V-M-F-HN-L. The nucleoprotein (N), phosphoprotein (P), and large polymerase (L) complexed with genomic RNA make up the nucleocapsid/polymerase complex. The P protein is encoded by a version of the P/V mRNA that is cotranscriptionally edited to contain an insertion of two guanosine residues at the P gene editing site (28). The accessory V protein is encoded from an unedited version of the P/V mRNA. HPIV2 does not have genes encoding the accessory C and D proteins that are found in HPIV1 or HPIV3 and does not have a gene for the SH protein that is found in SV5 and mumps virus. The matrix (M), fusion (F), and hemagglutinin-neuraminidase (HN) proteins are the envelope-associated polypeptides.
With the exception of the Toshiba strain of HPIV2 (19, 34) (see below), each of the previously reported complete genomic sequences for members of the subfamily Paramyxovirinae, including all other members of the Rubulavirus genus, have genome lengths that are an even multiple of six (4). The requirement for a polyhexameric genome length has been termed the "rule of six." A polyhexameric length has been shown to be required for efficient in vitro replication of paramyxovirus minigenomes and defective interfering (DI) genomes (7, 29, 30, 35) and is thought to be a consequence of each N protein subunit interacting with exactly 6 nucleotides (nt) in the viral RNA (vRNA)-nucleoprotein complex (2, 21, 44). For the most part, the rule of six in paramyxoviruses has not been evaluated by direct experimentation in the context of complete infectious virus.tx3/-$, http://www.100md.com
Previously, the genome length of the Toshiba strain of HPIV2 was reported to be 15,646 nt (GenBank accession number X57559) (16-18, 20), which exceeds an even multiple of six by 4 nt (6n + 4). Also, an HPIV2 Toshiba strain cDNA that did not conform to the rule of six was successfully used to recover a recombinant HPIV2 in cell culture (19). These observations suggested that HPIV2, unlike other members of the Paramyxovirinae, might not have a strict requirement for its genome length to conform to the rule of six (19). However, it was possible that the reported nucleotide length was in error and that the recovered recombinant virus had sustained one or more mutations that corrected the genome to make it polyhexameric. For example, we have recently discovered that mutant HPIV3s can be derived from cDNAs that do not conform to the rule of six (38), although HPIV3 has a requirement for a polyhexameric genome length (6, 7, 9, 41). Sequence analysis of HPIV3 recombinants derived from cDNAs that did not conform to the rule of six showed that they contained nucleotide insertions that corrected the length of the viral genome so that it conformed to the rule of six (38).
To investigate whether HPIV2 conforms to the rule of six, we (i) sequenced three natural isolates of HPIV2 to determine the naturally occurring genome length, (ii) constructed an antigenomic cDNA, recovered recombinant HPIV2 (rHPIV2) with a polyhexameric genome and wild-type (wt)-like growth characteristics in vitro and in vivo, and (iii) systematically modified the length of the antigenomic cDNA to determine whether recombinant virus could be recovered from nonpolyhexameric antigenomes. The genomes of recovered recombinant viruses were completely sequenced to determine their polyhexameric status.|ho, http://www.100md.com
MATERIALS AND METHODS|ho, http://www.100md.com
Cell lines and viruses. HEp-2 (ATCC CCL 23) and LLC-MK2 (ATCC CCL 7.1) cells were maintained in OptiMEM I (Life Technologies, Gaithersburg, Md.) supplemented with 5% fetal bovine serum and gentamicin sulfate (50 µg/ml). HPIV2 strain V9412-6 (HPIV2/V94) was isolated in Vero cells from a nasal wash specimen provided by Peter Wright of Vanderbilt University. HPIV2/V94 was biologically cloned and amplified on Vero cells for a total of nine passages (43), followed by one passage on LLC-MK2 cells. HPIV2 strain V9811-18 (HPIV2/V98), also provided by Peter Wright, was isolated in rhesus monkey kidney (RhMK) cells from a nasal wash specimen and was passaged three times on LLC-MK2 monkey kidney cells. The HPIV2 Greer strain was obtained from the American Type Culture Collection (VR-92; lot no. 1450727) and was passaged once on LLC-MK2 cells. Viruses were quantified by limiting dilution, with virus-infected cultures identified by hemadsorption with guinea pig erythrocytes, as described previously (12, 38).
Determination of complete consensus sequences for the Greer, V94, and V98 strains of HPIV2. vRNA was isolated from the supernatant of cells infected with HPIV2 strain Greer, V94, or V98, as described previously (27), and was subjected to reverse transcription (RT) using the Thermoscript RT-PCR system (Invitrogen, Inc., Carlsbad, Calif.) and random hexamer primers. PCR was carried out on the reverse-transcribed cDNA product using the Herculase enhanced polymerase blend (Stratagene, La Jolla, Calif.) or the Advantage cDNA PCR kit (Clontech, Palo Alto, Calif.) and primers designed from the published Toshiba strain sequence (GenBank accession number X57559) (16-18, 20) or sequence obtained during the course of the sequence analysis. Direct sequence analysis of the RT-PCR products was performed using a Perkin-Elmer ABI 3100 sequencer with the BigDye terminator cycle sequencing ready reaction kit (ABI Prism, Foster City, Calif.). The sequences were assembled using the Autoassembler program (Perkin-Elmer Applied Biosystems).
The 3' and 5' genomic termini of each HPIV2 strain were amplified using the 3' RACE System for Rapid Amplification of cDNA Ends (Invitrogen, Inc.) or the 5' RACE System for Rapid Amplification of cDNA Ends, version 2.0 (Invitrogen, Inc.), as specified by the manufacturer. For each strain, the sequence for the 3' and 5' ends was determined with multiple sequencing reactions of two independently derived PCR products, all of which were confirmatory.\ezt, http://www.100md.com
Sequence alignments. The nucleotide sequences of HPIV2 strains V94, V98, and Greer were subjected to pairwise comparisons to determine the percent nucleotide identity. The deduced amino acid sequences of the proteins of HPIV2 strains V94, V98, and Greer were compared to those of the following paramyxoviruses: HPIV2 Toshiba strain, GenBank accession no. X57559; SV5 W3A strain, GenBank accession no. AF052755; SV41 strain Toshiba/Chanock, GenBank accession no. X64275; and mumps virus isolate 88-1961, GenBank accession no. AF467767. Percent amino acid identities for the proteins in Table 1 were calculated with the Gap program of the Wisconsin Package Version 10.2 (Genetics Computer Group, Madison, Wis.).
fig.ommitteed4, http://www.100md.com
TABLE 1. Amino acid sequence identity between the proteins of HPIV2/V94 and the analogous proteins of HPIV2 strains Greer and V98, SV41, SV5, and mumps virus4, http://www.100md.com
Assembly of a full-length rHPIV2/V94 cDNA antigenomic clone. A full-length cDNA clone encoding the HPIV2/V94 antigenomic RNA was constructed from six cloned overlapping RT-PCR and RACE products using the following restriction sites in the HPIV2 genome: NheI (nt 892 position in the complete antigenomic sequence), XhoI (nt 3673), Asp718 (nt 8141), AatII (nt 10342), and BsiWI (nt 15280). Briefly, the first fragment was generated from vRNA using the 3' RACE System for Rapid Amplification of cDNA Ends and the Advantage-HF cDNA PCR kit with specific HPIV2 primers and a primer containing an MluI site followed by a T7 promoter, two nonviral G residues, and the antigenomic cDNA corresponding to the 3' end of the viral genome. The next four fragments were reverse transcribed using the Thermoscript RT-PCR System and PCR amplified with the Herculase enhanced polymerase blend. The final fragment containing the 5' end was reverse transcribed using the 5' RACE System for Rapid Amplification of cDNA Ends (Invitrogen, Inc.) and was amplified with the Advantage-HF cDNA PCR kit using primers that added a partial hepatitis delta virus ribozyme ending with an RsrII site that is part of the ribozyme sequence. The individual PCR products were cloned into a pUC19 plasmid that had been modified to contain MluI, BspEI, XhoI, Asp718, AatII, BsiWI, and RsrII restriction enzyme sites to accept the PCR fragments. The cDNA clones were sequenced using a Perkin-Elmer ABI 3100 sequencer with the BigDye sequencing kit and were compared to the consensus sequence of HPIV2/V94 prior to assembly. All six PCR products were assembled into the above-mentioned pUC19 vector. The MluI-to-RsrII cDNA, containing the complete HPIV2/V94 antigenome sequence, was transferred into a modified pBlueScript that contained the rest of the hepatitis delta virus ribozyme followed by a T7 polymerase terminator sequence (27, 33). The resulting antigenomic plasmid, pFLC-HPIV2/V94, contains the following elements: a T7 promoter followed by two nonviral guanosine residues, the complete antigenomic 15,654-nt sequence of HPIV2/V94, a hepatitis delta virus ribozyme, and a T7 polymerase transcriptional terminator as previously described for the generation of recombinant HPIV3, BPIV3, and HPIV1 (6, 27, 33). The sequence of the full-length cDNA was verified by DNA sequence analysis. Two point mutations were identified in the F and L open reading frames (ORFs) of the full-length cDNA that had not been present in the consensus sequence and presumably had occurred during the assembly of the complete clone. These occurred at positions 6265 (AGT to AGC; silent) and 15075 (GCT to GTT; Ala-2093 to Val). These mutations were used as markers to distinguish the recombinant HPIV2 from the biologically derived HPIV2/V94 parent.
HPIV2/94 N, P, and L support plasmids for the recovery of HPIV2/V94 from cDNA. A support plasmid encoding the N protein of HPIV2/V94 (pTM-N2) was derived from vRNA using the Thermoscript RT-PCR system and the Advantage-HF PCR kit using an antigenomic-sense oligonucleotide that contained an AflIII site spanning the N ORF ATG translation initiation codon site and an antisense oligonucleotide containing an EcoRI site. The PCR product was digested with AflIII and EcoRI and cloned into pTM1 (6-8) that had been digested with NcoI and EcoRI.*:q^%[, http://www.100md.com
The HPIV2/V94 P protein expression plasmid (pTM-P2) was generated from two overlapping PCR fragments (25) and was engineered to contain the two inserted guanosine residues at the HPIV2 P gene editing site (nt 2481 to 2487) that distinguish the complete P ORF from that of V. The P ORF was cloned into pTM1 as an NcoI-to-EcoRI fragment.*:q^%[, http://www.100md.com
An HPIV2 L polymerase expression plasmid (pTM-L2) was assembled from a PCR product and two subclones of L. The upstream end of L was amplified by PCR with a sense oligonucleotide containing an NcoI site spanning the L gene ATG translation initiation codon, and an antisense oligonucleotide downstream of a unique AatII site (nt 10342) in the L ORF. The remainder of the L ORF was derived from a subclone used to construct the HPIV2 full-length clone. The complete HPIV2/V94 L ORF was then cloned as an NcoI-to-RsrII fragment into a version of pTM1 that was modified to contain RsrII as a downstream site (27).
Recovery and sequence analysis of rHPIV2/V94. rHPIV2/V94 was recovered from the full-length antigenomic cDNA essentially as described previously (37). Briefly, HEp-2 cells in six-well plates were cotransfected with a full-length HPIV2/V94 cDNA plasmid and the three HPIV2 support plasmids (pTM-N2, pTM-P2, and pTM-L2), using Lipofectamine-2000 reagent (Invitrogen, Inc.). The HEp-2 cells were simultaneously infected with MVA-T7 as described previously (6, 33). The medium supernatant was harvested on day 3 or 4 posttransfection and passaged two times on LLC-MK2 cell monolayers. rHPIV2/V94 was then biologically cloned by plaque-to-plaque purification on LLC-MK2 monolayers and was further propagated on LLC-MK2 cells, as previously described (39). To confirm that the recovered viruses had been derived from cDNA, RT was performed and segments of the viral genome were amplified by PCR. Sequence analysis of the PCR products showed that the two point mutations in the F and L genes present in the antigenomic cDNA as described above were contained in the recombinant virus, designated rHPIV2/V94, and not in the wt parental virus. Two preparations of HPIV2/V94 (designated rAHPIV2/V94 and rBHPIV2/V94) were made, each derived from a separate transfection. vRNA was isolated from the biologically cloned rAHPIV2/V94 preparation as described previously (27) and was used as the template for RT-PCR and RACE using oligonucleotide primers. The amplified products were analyzed by DNA sequencing to determine the complete consensus genomic sequence of rAHPIV2/V94.
Assessment of the rule of six in complete recombinant virus. The full-length HPIV2/V94 antigenomic cDNA was modified by the insertion of oligonucleotide duplexes, which were 42, 43, 44, 45, 46, or 47 bp long, into the EcoRV restriction site present near the downstream end of the L ORF (nt 15554 to 15559) This resulted in six antigenomic cDNAs that are 6n, 6n + 1, +2, +3, +4, or +5 with regard to polyhexamer length. The correct orientation and sequence of each insert was confirmed. These antigenomic cDNAs were transfected into HEp-2 cells along with HPIV2 support plasmids and coinfected with MVA-T7 as described above. Recovered virus was passaged and biologically cloned on LLC-MK2 monolayers as described above, with one or two independent virus preparations made from each antigenomic plasmid. vRNA was isolated and used to generate RT-PCR products spanning the entire genome as described above. These were then directly sequenced to generate a complete consensus sequence for each virus.s;, http://www.100md.com
Replication of HPIV2 in vitro. Recombinant or biologically derived HPIV2 was inoculated in triplicate onto LLC-MK2 cell monolayers in six-well plates at a multiplicity of infection of 0.01, and cultures were incubated at 32°C with and without 5 µg of porcine trypsin/ml added to the culture medium, as described previously (38). Medium (0.5 ml) from each well was harvested and replaced with 0.5 ml of fresh medium at 0 h and at 1 to 6 days postinfection. Virus present in the samples was quantified by titration on LLC-MK2 monolayers in 96-well plates that were incubated for 6 days at 32°C. Virus grown in the presence of trypsin was titered with trypsin in the medium. Virus was detected by hemadsorption, and the titer is reported as log10 50% tissue culture infectious dose (TCID50)/ml.
Replication of recombinant and wt viruses in hamsters. Four-week-old Golden Syrian hamsters (Charles River Laboratories, Stoneridge, N.Y.) in groups of six were inoculated intranasally with 0.1 ml of L15 medium containing 106.0 TCID50 of virus. On day 3, 4, or 5 postinfection, the lungs and nasal turbinates were harvested and the virus was quantified by serial dilution of tissue homogenates on LLC-MK2 monolayers, as previously described (36). The mean titer of virus was calculated for each day for each group of hamsters and is expressed as log10 TCID50/gram of tissue.8yx5(, http://www.100md.com
Nucleotide sequence accession numbers. The nucleotide sequences of HPIV2 strains V94, V98, and Greer were submitted to GenBank under accession numbers AF533010, AF533011, and AF533012, respectively.8yx5(, http://www.100md.com
RESULTS8yx5(, http://www.100md.com
Determination of the complete nucleotide sequence of three strains of HPIV2. The biologically derived HPIV2 strain Vanderbilt 9412-6 (HPIV2/V94) was obtained in 1994 from a 13-month-old infant presenting with upper and lower respiratory tract infection, high fever, and otitis media and therefore represents a recent, low-passage-number isolate of HPIV2. The complete genomic sequence of HPIV2/V94 was determined from RT-PCR products amplified from purified vRNA. The sequence analysis was performed directly on RT-PCR products without a cloning step and thus yields a consensus sequence. The HPIV2/V94 genome was found to be 15,654 nt in length and thus conforms to the rule of six (2, 21). Two additional HPIV2 strains were also sequenced. The prototype HPIV2 Greer strain was originally isolated in 1955 from an 11-month-old infant with respiratory tract disease (3), and the HPIV2 Vanderbilt 9811-18 (V98) strain was isolated in 1998 from a 20-month-old infant with respiratory tract disease. The complete consensus genome sequence was determined for the HPIV2/Greer and HPIV2/V98 strains and, in each case, was 15,654 nt in length, the same as HPIV2/V94. Thus, three strains of HPIV2, isolated over a period of 40 years, have the identical polyhexameric genome length.
Sequence comparison of the HPIV2 strains. The percent identity of the nucleotide sequence of the Greer, V94, and V98 strains was 95.0% for the V94 and V98 strains, 97.4% for the V98 and Greer strains, and 96.2% for the Greer and V94 strains. Thus, the recent V94 and V98 isolates of HPIV2 are highly related to the 1955 Greer isolate, with each being more closely related to the Greer strain than to each other. The cis-acting transcription gene-start (GS), intergenic (IG), and gene-end (GE) signal sequences were highly conserved. In addition, the phasing of the GS signals, i.e., their position relative to the hexamer spacing of the genome, was the same for all three strains (not shown), which may be important for efficient expression of paramyxovirus mRNAs (21). As is typical for nonsegmented negative-strand viruses, the nucleotide sequences of the ORFs were more highly conserved than were the nontranslated gene sequences (not shown).m;5[kg&, http://www.100md.com
The predicted amino acid sequences of the HPIV2/V94 N, P, V, M, F, HN, and L proteins were compared to their counterparts in Greer, V98, and several other members of the rubulaviruses, and the percent identities are shown in Table 1. The predicted amino acid sequences are very highly conserved between the three HPIV2 strains, with the highest degree of divergence occurring in the HN glycoprotein. Comparison with SV41, SV5, and mumps virus showed a decreasing level of sequence identity between each and HPIV2, in the order of SV41 first with the highest level of identity, and then SV5, and then mumps virus, with L being the most highly conserved protein.
Construction of a full-length HPIV2/V94 antigenomic cDNA and recovery of recombinant wt HPIV2/V94. A complete HPIV2/V94 antigenomic cDNA, designated pFLC-HPIV2/V94, was constructed from six overlapping RT-PCR and RACE products. Nucleotide sequencing of the complete antigenomic cDNA showed that it contained two nucleotide substitutions in the F and L genes, respectively, compared to the consensus sequence of biologically derived HPIV2/V94. The substitution at position 6265 (T to C) in the F gene is translationally silent, whereas that at nt 15075 (C to T) encodes an alanine-2093-to-valine substitution in the L protein. The amino acid at this position is not conserved among other parainfluenza viruses and lies outside of the six major domains that are conserved among paramyxoviruses (31, 40). These nucleotide differences served as markers for a recombinant HPIV2 derived from cDNA.n!'d, 百拇医药
The antigenomic cDNA was transfected into HEp-2 cells, and virus was recovered by cotransfection with the HPIV2/V94 N, P, and L support plasmids and coinfection with MVA-T7. Virus was readily recovered after only a single passage of the HEp-2 transfection supernatant onto LLC-MK2 cell monolayers, and the presence of the nucleotide markers in the HPIV2 F and L genes was confirmed by RT-PCR of vRNA and sequence analysis. Amplification of RT-PCR products was dependent on the addition of RT, indicating that the template was indeed viral RNA and not contaminating DNA (data not shown). Two separate recombinant HPIV2 isolates were biologically cloned from independent transfections and were designated rAHPIV2/V94 and rBHPIV2/V94. The complete sequence of rAHPIV2/V94 was determined from RT-PCR products, including determination of the two ends by RACE. The size of the recovered recombinant HPIV2 genome was identical to that of its biologically derived parent and, thus, conformed to the rule of six. The genomic sequence of the recovered virus was identical to the cDNA from which it was derived except for a silent nucleotide substitution at nt 13786 (T to C) that presumably arose during propagation of the recovered virus in cell culture. rBHPIV2/V94 was not sequenced in its entirety, but it did not have this nucleotide change.
Characterization of the in vitro and in vivo growth properties of rHPIV2/V94. The growth kinetics and final yield of the two preparations of recombinant HPIV2, rAHPIV2/V94 and rBHPIV2/V94, were compared to those of their biologically derived HPIV2 parent at 32°C in LLC-MK2 cells and were found to be essentially indistinguishable (Fig. 1A). Growth at 32, 39, and 40°C was examined, and two rHPIV2/V94 isolates and their biologically derived parent were not temperature sensitive at the higher temperatures (data not shown). Thus, the incidental nucleotide substitutions in rHPIV2/V94 did not alter growth properties in vitro. To determine if HPIV2/V94 requires the addition of trypsin for efficient growth in vitro, we compared the effect of added porcine trypsin on the in vitro growth kinetics and final yield of the recombinant and biologically derived HPIV2s. The addition of trypsin did not increase the growth efficiency of the biologically derived or recombinant viruses (data not shown).
fig.ommitteed|!dnn, 百拇医药
FIG. 1. Replication of biologically derived and recombinant HPIV2/V94 in vitro and in vivo. (A) Multistep growth kinetics in vitro. LLC-MK2 cell monolayers were infected in triplicate at a multiplicity of infection of 0.01 with biologically derived HPIV2/V94 or with two preparations of rHPIV2/94 derived from separate transfections (rAHPIV2/V94 and rBHPIV2/V94). The cells were incubated at 32°C. Aliquots of the medium supernatants were harvested at 24-h intervals, and titer of virus was determined at 32°C. Titers of virus are expressed as mean log10 TCID50/ml ± standard error. (B) Replication in hamsters. Hamsters were inoculated intranasally with 106 TCID50 of the same viruses as in panel A. Nasal turbinates and lung tissues from six animals from each group were harvested on days 3, 4, and 5 postinfection. Virus present in tissues was quantified by serial dilution on LLC-MK2 monolayers at 32°C.|!dnn, 百拇医药
To determine whether the recovered rHPIV2/V94 retained the replicative properties of its biologically derived parent in vivo, the replication of the two preparations of rHPIV2/V94 in the respiratory tracts of hamsters was compared to that of the biologically derived HPIV2/V94 (Fig. 1B). The level of replication of the two preparations of rHPIV2/V94 was essentially indistinguishable from that of the biologically derived HPIV2/V94 parent on all of the days tested, demonstrating that the recombinant version of HPIV2/V94 has replicative properties similar to those of its biologically derived counterpart and therefore is an authentic wt virus. Thus, the incidental amino acid substitution at position 2093 of the L polymerase also did not affect the level of replication of rHPIV2/V94 in vivo. However, a full-length antigenomic cDNA encoding an alanine at position 2093 of the L protein, as occurs in biologically derived HPIV2/V94, was also generated for use in future studies, and this modified HPIV2/V94 cDNA was also used to recover recombinant HPIV2/V94 in vitro (data not shown).
Recombinant HPIV2s generated from cDNAs that do not conform to the rule of six contain nucleotide insertions or deletions that serve to yield polyhexameric genome lengths. The requirement for a polyhexameric genome length has not been directly examined for HPIV2. Also, it was of interest to systematically examine the requirements for the rule of six, and the consequences of deviations from the rule, in the context of complete infectious virus. We generated six full-length antigenomic cDNAs of HPIV2/V94 that were modified to contain oligonucleotide insertions (Fig. 2) at an EcoRV restriction site near the end of the L polymerase gene. This site was chosen since it did not lie within a known cis-acting sequence and would not interfere with the hexamer phasing of upstream cis-acting sequences. The inserts were designed to restore the L polymerase ORF sequence but increased the length of the downstream noncoding region of L, and hence the complete antigenomic cDNA, by 42 nt (6n with respect to hexamer spacing, a control cDNA conforming to the rule of six), 43 nt (6n + 1), 44 nt (6n + 2), 45 nt (6n + 3), 46 nt (6n + 4), or 47 nt (6n + 5) (Fig. 2).
fig.ommitteed1?zeh%, http://www.100md.com
FIG. 2. Structure of the nucleotide inserts used to make HPIV2 antigenomic cDNAs that do not conform to the rule of six. (A) The wt nucleotide sequence around the EcoRV restriction site near the end of the L ORF is shown. (B) The sequences of the six antigenomic-sense oligonucleotides inserted at the EcoRV restriction site spanning HPIV2 nt 15554 to 15559 is shown. Oligonucleotide duplexes were inserted between nt 15556 and 15557. Each oligonucleotide duplex contains a silent ATC-to-ATT mutation that destroys the EcoRV site and recreates the last 11 codons of the L ORF, and 12 HPIV2 nt including the TGA stop codon (bold) followed by zero to five additional nucleotides. The designation of the recombinant viruses generated from the cDNAs is indicated to the left: the virus designated rHPIV2/V94(+6) conforms to the rule of six, while the others contain the indicated number of additional nucleotides (+1 to +5). The length of each oligonucleotide insert is shown on the right, with the rule-of-six length and the total length of the antigenomic cDNA indicated in parentheses.1?zeh%, http://www.100md.com
The antigenomic cDNAs were transfected into HEp-2 cells along with HPIV2 support plasmids, as described above. The supernatants were harvested after 3 or 4 days and were passaged twice on LLC-MK2 monolayer cultures. Surprisingly, virus was readily recovered from every cDNA tested. This indicated that although naturally occurring HPIV2 genomes are polyhexameric, recombinant HPIV2 can be generated from a cDNA that does not conform to the rule of six. Recombinant HPIV2s recovered from cDNAs that did not conform to the rule of six [rHPIV2/V94(+1), rHPIV2/V94(+2), rHPIV2/V94(+3), rHPIV2/V94(+4), and rHPIV2/V94(+5)] were biologically cloned by plaque purification on LLC-MK2 monolayer cultures and were further amplified by passage on LLC-MK2 cells, as described above. vRNA was isolated from each of these recombinants and was used to perform RT-PCR and RACE. The PCR products were sequenced directly, yielding a complete consensus sequence for at least one recovered viral isolate for each of the nonpolyhexameric antigenomes (see below).
For each virus that was recovered from a nonpolyhexameric antigenomic cDNA, nucleotide sequence analysis revealed small nucleotide insertions or deletions that restored the recovered genome length to conform to the rule of six (Fig. 3). For example, the complete consensus sequence of the genome of the rHPIV2/V94(+3) isolate revealed that this virus had acquired a 2-nt (UU, genome sense) insertion in the HN gene 5' noncoding region (NCR) and a 1-nt (A, genome sense) insertion within a poly(A) tract in the HN gene 3' NCR (Fig. 3A). The genome size of the recovered rHPIV2/V94(+3) isolate was 15,702 nt, i.e., 3 nt longer than the HPIV2 cDNA from which it was derived. Thus, the 3-nt insertion served to modify the viral genome length so that it conforms to the rule of six. Complete sequencing of the rHPIV2/V94(+4) genome showed that this virus had sustained a 2-nt insertion (UA, genome sense) at the end of the intergenic region between the HN and L genes and immediately upstream of the L gene GS signal sequence (Fig. 3A). The genome size of rHPIV2/V94(+4) was 15,702 nt, i.e., 2 nt longer than the 6n + 4 HPIV2 cDNA from which it was derived. Thus, the 2-nt insertion served to generate a polyhexameric genome length. Two independently derived cloned isolates of rHPIV2/V94(+5) were prepared, rAHPIV2/V94(+5) and rBHPIV2/V94(+5). Complete consensus sequence analysis of rBHPIV2/V94(+5) showed that it has a single nucleotide insertion in the HN 3' untranslated region that restored polyhexameric length. The other [rAHPIV2/V94(+5)] was partially sequenced and was found to contain a single nucleotide insertion in the HN GE signal sequence that could serve to correct the genome length so that it conforms to the rule of six (Fig. 3A). rHPIV2/V94(+1) and rHPIV2/V94(+2) were completely sequenced and were found to have 1- and 2-nt deletions, respectively, that served to generate a polyhexameric genome (Fig. 3B). The 2-nt deletion that was found in rHPIV2/V94(+2) occurred within the intergenic sequence between the HN and L genes. Interestingly, the single nucleotide deletion in rHPIV2/V94(+1) occurred near the end of the coding region for the L polymerase. This resulted in a shift in the reading frame, such that the last 13 amino acids of the L protein were deleted and replaced by 21 amino acids encoded by the alternate reading frame. Thus, the majority of genome length corrections occurred at sites in noncoding regions. Interestingly, almost all of the changes were located within or between the HN and L genes, the two genes most proximal to the insert responsible for the deviation from the rule of six.
fig.ommitteedo+n\c^c, 百拇医药
FIG. 3. Nucleotide insertions and deletions detected in the genomic RNA of recombinant HPIV2s derived from cDNAs that do not conform to the rule of six (sequences are in antigenomic sense). (A) Four recombinant viruses that were produced from cDNAs that did not conform to the rule of six, rHPIV2/V94(+3), rHPIV2/V94(+4), rAHPIV2/V94(+5), and rBHPIV2/V94(+5), contained nucleotide insertions that resulted in a polyhexameric genome length. Two insertions containing a total of 3 nt were identified in the HN 5' and 3' noncoding regions (5' NCR and 3' NCR) of rHPIV2/V94(+3). A 2-nt insertion within the IG region between the HN and L genes was identified in rHPIV2/V94(+4). A 1-nt insertion was found in each clone of rAHPIV2/V94(+5) and rBHPIV2/V94(+5), in one case at the end of the HN GE signal and in the other in the HN 3' NCR. (B) Two recombinant viruses produced from cDNAs that did not conform to the rule of six, rHPIV2/V94(+2) and rHPIV2/V94(+1), contained nucleotide deletions that resulted in a polyhexameric genome length. rHPIV2/V94(+2) was found to have a 2-nt deletion (underlined) within the IG region between the HN and L genes. rHPIV2/V94(+1) was found to have a 1-nt deletion (underlined) near the end of the L polymerase ORF. This resulted in a frameshift in the L coding sequence that deleted the last 13 amino acids of L and replaced them with an unrelated sequence of 21 amino acids, as shown. The nucleotide sequence (antigenomic sense) in the region of the insertions is shown. The inserted nucleotides are shown, and the site of each insertion is indicated by an arrow pointing downward. Deleted nucleotides are shown, and the location of each deletion is indicated by an arrow pointing upward. The HN 5' or 3' NCR, transcription GE and GS (in bold type), and IG regions between the HPIV2 HN and L ORFs are indicated. The L polymerase translation initiation codon (ATG) and translation termination codon (TGA) are underlined. The single-letter amino acid designation is shown below the nucleotide sequence for the wt and a mutant version of the HPIV2 L polymerase.
DISCUSSIONj)r$-, 百拇医药
We investigated the question of whether the natural genome length of HPIV2 conforms to the rule of six by determining the complete genomic sequence for three strains of HPIV2, namely the prototype Greer strain and two recent HPIV2 isolates, V94 and V98. Each of these strains has a polyhexameric 15,654-nt genome length, demonstrating that the HPIV2 genome length is conserved and conforms to the rule of six. A polyhexameric antigenomic HPIV2/V94 cDNA was used to recover a recombinant HPIV2/V94 with wt growth characteristics, and the length of the recovered virus was confirmed to be 15,654 nt. Full-length antigenomic HPIV2 cDNAs that were engineered to deviate from the rule of six by 1 to 5 nt yielded viruses that contained small nucleotide insertions or deletions that served to generate polyhexameric genomes. These findings demonstrate that an HPIV2 polyhexameric genome can be generated from a cDNA that does not conform to the rule of six but that virus recovery appears to depend on the occurrence of spontaneous mutations that modify the length of the genome to conform to the rule of six. These findings suggest that HPIV2 has a strong preference for a polyhexameric genome. Thus, the requirement for a polyhexameric genome has now been directly confirmed in complete recombinant virus for HPIV2 and, in earlier work, HPIV3 (38), representing the Rubulavirus and Respirovirus genera of the Paramyxovirinae subfamily. This suggests that adherence to the rule of six and the ability to correct deviations from the rule may be universal for all members of this paramyxovirus subfamily.
Comparison of the genome sequence of the Greer strain with that reported previously for the Toshiba strain revealed that these genomes share 99.7% nucleotide sequence identity. Thus, the Toshiba and Greer strains are nearly identical, with a total of only 34 nucleotide differences. Surprisingly, these differences included a large number of apparent insertions and deletions in one strain versus the other. For example, 13 nt present in the Greer strain were missing from the reported Toshiba sequence, representing either deletions in the Toshiba strain or insertions in the Greer, and 5 nt reported for the Toshiba sequence were not present in Greer, representing either insertions in Toshiba or deletions in Greer. It seems unlikely that so many apparent insertions or deletions would exist in two such highly related strains, particularly since such spacing discrepancies were not evident between the Greer, V94, and V98 strains. Correction for these 18 insertions and deletions in the Toshiba strain would yield a genome length of 15,654, the same size as reported here for the V94, V98, and Greer strains of HPIV2. This suggests that these apparent insertions and deletions might represent errors in cDNA synthesis, cloning, and sequence analysis rather than true differences in nature. Regarding the previous demonstration that recombinant Toshiba strain virus can be recovered from a nonpolyhexameric antigenomic cDNA (19), the findings in the present paper indicate that there is a stringent mechanism for the very rapid emergence of polyhexameric genomes in recombinant virus derived from nonpolyhexameric antigenomes. Thus, it seems likely that the recovered recombinant Toshiba virus contains a corrected, polyhexameric genome.
In the present study, the source of the insertions and deletions that served to correct rHPIV2/V94 recovered from nonpolyhexameric antigenomes is not known. However, two likely sources are mutations introduced by T7 RNA polymerase during the synthesis of the initial antigenome from the transfected plasmid, mutations introduced by the HPIV2 polymerase during subsequent RNA replication, or both. In particular, the RNA polymerases of bacteriophage T7 and nonsegmented negative-strand viruses share the propensity to insert or delete nucleotides within homopolymeric tracts, which would be consistent with the finding in the present study that four of the five insertions, and one of the two deletions, occurred in a homopolymeric tract of three or more nucleotides. T7 RNA polymerase has been shown to insert or delete nucleotides within homopolymeric sequences consisting of three or more nucleotides at an estimated rate of up to 1 in 4.5 x 107 nt (22, 23). This misincorporation activity is likely a result of "stuttering," which involves slippage by the polymerase and nascent strand during the copying of repetitive sequences (5, 22, 42). Stuttering also has been described for eukaryotic and other viral polymerases, including the paramyxovirus L polymerases.
Indeed, stuttering is a normal feature of RNA synthesis by the L polymerase of nonsegmented negative-strand RNA viruses: stuttering at the poly(U) tract (genome-sense) of the GE signals is responsible for the synthesis of the poly(A) tails of mRNA, and stuttering at the editing site of the P gene is responsible for the insertion of one or more G residues (mRNA-sense) that give rise to frame-shifted versions of the P mRNA (15). There are numerous examples of changes in genome length during RNA replication by the negative-strand RNA polymerase. It was previously shown that Sendai virus minigenomes whose lengths did not conform to the rule of six and which contained the P gene editing site [a poly(G) stretch] underwent in vitro nucleotide insertions or deletions within the editing site that served to generate polyhexameric genome lengths (13). DI genomes of SV5 were also shown to undergo in vitro genome length modification by nucleotide insertions, and SV5 DI genomes whose length conformed to the rule of six replicated more efficiently than uncorrected genomes (26). Other negative-stranded RNA viruses have also been shown to generate modified genomes that contain nucleotide insertions. Vesicular stomatitis virus, which is a member of the Rhabdoviridae family and has a genomic organization similar to that of paramyxoviruses but which does not follow the rule of six, has been shown to accumulate many nucleotide insertions in the 3' noncoding region of the glycoprotein gene during the natural evolution of the virus (1). The nucleotide insertions identified included homopolymeric stretches of U and/or A, and it was suggested that these occurred by polymerase stuttering. Respiratory syncytial virus, a paramyxovirus that does not follow the rule of six, is able to generate neutralization resistant mutants by inserting or deleting nucleotides (A) within a homopolymeric A stretch in the G protein gene (10, 11). Another negative-stranded RNA virus, influenza virus type A, has also been shown to incorporate U's at a poly(U) stretch in the neuraminidase gene by stuttering during genome replication (45).
The length-correcting mutations identified in the present study had four characteristics: (i) they occurred predominantly in noncoding regions rather than in coding regions; (ii) they were exclusively A or U; (iii) they usually occurred in homopolymer tracts, as already noted; and (iv) they all occurred within one or two genes on the upstream side (relative to the genome map) of the insert that was responsible for the deviation from the rule of six. It is noteworthy that these same four characteristics also hold for the length corrections observed in a previous study in which rHPIV3s were recovered from antigenomes that deviated from the rule of six due to the insertion of a nonhexameric foreign gene (38).|ji@, 百拇医药
Interestingly, in this previous HPIV3 study, the insert that caused the deviation from the rule of six had been placed at two different genome locations, namely between the N and P or the P and M genes. Two of the six examples of correcting mutations that were found in this previous study occurred within the foreign gene insert and in four cases within the homopolymer tract of the GE signal of the first or second gene on the upstream side of the insert (reference 38 and unpublished data). Recently it was reported that recombinant polyploid measles viruses containing foreign gene insertions that were apparently deleterious to virus replication accumulated nucleotide insertions that inactivated the foreign gene and a compensating nucleotide deletion mutation that restored polyhexameric genome length downstream of the site of insertion (32). Thus, these genome length correction mutations differed from those described in the present study and previously (38) in that the location of the correcting mutation was downstream of the nucleotide insertion site that resulted in deviation of the genome length from the rule of six.
The low frequency of genome length corrections in ORFs is not surprising, since the insertion or deletion of 1 or 2 nt within an ORF would shift the reading frame and, if this affected an essential protein, would be lethal. A 3-nt insertion or deletion would add or delete an amino acid, which also would frequently be deleterious. The only genome length modification that occurred within an ORF was the nucleotide deletion in rHPIV2/V94 (+1). This occurred near the end of the L ORF, shifting the reading frame such that the C-terminal 13 amino acids of L were deleted and replaced by 21 heterologous amino acids encoded by the second reading frame. Since the recombinant bearing this mutation was readily recovered and grew to a high titer, it suggests that this region of the L protein is dispensable for viral replication in vitro. In the previous HPIV3 study, two of the six identified genome length corrections occurred within the ORF of the foreign gene insert (38), but this would not be deleterious to the virus since the foreign gene would not be required for virus replication.
It was unexpected to observe that the nucleotide insertions and deletions involved exclusively A or U. This also was observed in the previous HPIV3 study noted above. This might be a consequence of the inability of the replicase to modify genome length by adding or deleting G or C residues, or it might simply reflect a lower content of homopolymeric G and C in the regions most amenable to accepting a length correction, namely the nontranslated regions and the IG regions. The finding that most of the length-correcting mutations were insertions, and that most of them occurred in homopolymeric tracts, is consistent with a stuttering mechanism that might have some similarity to that involved in polyadenylation. Indeed, 5 of the 10 insertions identified in the two studies involved a GE signal. The present study also contained two examples involving the insertion or deletion of an AU from an AUAU sequence, each of which might represent a 2-nt slippage.*wm, 百拇医药
Perhaps the most unexpected finding of the present study was the distribution of the length-correcting mutations. It might have been expected that length corrections in recovered virus would be found throughout the genome, subject only to a preference for A or U homopolymers and a bias against disturbing an ORF or cis-acting signal. Instead, these mutations occurred with a biased distribution: specifically, they were always upstream of the insert causing the deviation from the rule of six, and they always occurred within the first or second upstream genes (in the present study, within the HN and L genes). This also had been observed for the previous HPIV3 study, in which the deviating insert was placed between N and P, or P and M, and the corrections always occurred on the upstream side and always within one or two genes. This suggest that the length correction by the two HPIVs operates by a common mechanism that might involve either (i) random length correction, followed by the stringent selection for virus in which the correction was close to the point of deviation, or (ii) nonrandom length correction, perhaps involving a replication complex that "senses" the deviation from the rule of six and acts to insert a correcting mutation at a second, downstream site in the nascent molecule.
One factor that could affect the location of the insertions and deletions is the phasing of the GS signals within the genome. It has been noted that the position of the first nucleotide of the GS signals of members within each of the three genera of the Paramyxovirinae is partially or completely conserved for each gene, suggesting that the position of the GS signals relative to the hexamer phasing of the genome might be important for transcription (21). The introduction of a length-correcting mutation into a genome would alter the hexameric phasing of all downstream GS signals. Thus, if the phasing of the GS signals indeed is important, there would be a selective advantage to viruses in which the length correction occurs as close as possible to the insert that caused the deviation from the rule of six, thereby disturbing the phasing of the fewest possible GS signals. For example, in the present study, the deviation from the rule of six was engineered in the downstream end of the L gene, and each of the correcting mutations occurred in the HN and L genes and altered the phasing of no more than one GS signal. In the previous HPIV3 study, the insert causing deviation from the rule of six was between N and P or P and M, and the correcting mutations occurred such that in three cases a single vector GS signal was out of phase, and in the other cases all of the GS signals were in phase (38).
Thus, the phasing of the GS signals might provide a selective pressure that accounts for the biased distribution of corrections in recovered virus. As a caveat, it should be noted that while the hexamer phasing of the GS signals for HPIV2 is conserved between the V94, V98, and Greer strains, the rubulaviruses have the lowest degree of conservation of GS phasing among the three genera of the Paramyxovirinae. Also, while the effect of GS phasing on paramyxovirus transcription and replication has not been carefully analyzed, one study in which the phasing of the L GS signal of SV5 was altered showed that the effect was modest. However, that study involved a single misphased GS signal, and it might be that a single misphased GS signal is tolerated, while the aggregate effect of two or more misphased signals is sufficient to provide a negative selective pressure (14). This model of random correction followed by biased selection has the advantage of simplicity, but it does not explain why the length corrections were always on the upstream side of the deviation from the rule of six.
Alternatively, nonrandom correction might be involved. By this model, the viral replication machinery somehow senses a deviation in the genome length or phasing in the template during synthesis and makes a length adjustment near the area where the genome has been altered. It is not clear how the replication machinery might sense a deviation, but one possibility is that it involves the hexamer phasing of some repeated genome sequence, such as the GS or GE signals or some other structure that has not been identified. Thus, a deviation in the phasing of a signal in the template would be detected by the polymerase in the process of RNA synthesis, and a length correction would follow. This model would account for the universal asymmetry of the length corrections with regard to the original deviation from the rule of six. The finding that all of the corrections occurred upstream in the gene map relative to the point of deviation would be consistent with length correction during the synthesis of the genome-sense strand.
It is not clear from the present study whether the ability to correct deviations in genome length in such a seemingly rapid and efficient manner is an artifact of the reverse genetics system or an intrinsic property of the viral replication machinery that is used during the natural evolution of the virus to maintain proper genome length. The conservation of the length of the HPIV2 genome in strains isolated more than 40 years apart, the numerous examples of negative-stranded RNA viruses that have the ability to modify their genome length, and the ease with which recombinant parainfluenza viruses can be derived from cDNAs that do not conform to the rule of six suggest that the latter may be the case. The infectious HPIV2 cDNA derived from a consensus wt HPIV2 sequence described here will be useful for further exploring this and other aspects of the biology of this important human pathogen.z]1c, 百拇医药
ACKNOWLEDGMENTSz]1c, 百拇医药
We thank Peter Wright and Sharron Tollefson for providing the V94 and V98 strains of HPIV2.
REFERENCES9+.hqe, 百拇医药
Bilsel, P. A., and S. T. Nichol. 1990. Polymerase errors accumulating during natural evolution of the glycoprotein gene of vesicular stomatitis virus Indiana serotype isolates. J. Virol. 64:4873-4883.9+.hqe, 百拇医药
Calain, P., and L. Roux. 1993. The rule of six, a basic feature for efficient replication of Sendai virus defective interfering RNA. J. Virol. 67:4822-4830.9+.hqe, 百拇医药
Chanock, R. M. 1956. Association of a new type of cytopathogenic myxovirus with infantile croup. J. Exp. Med. 104:555-575.9+.hqe, 百拇医药
Chanock, R. M., B. R. Murphy, and P. L. Collins. 2001. Parainfluenza viruses, p. 1341-1379. In D. M. Knipe, P. M. Howley, D. E. Griffin, R. A. Lamb, M. A. Martin, B. Roizman, and S. E. Strauss (ed.), Fields virology, 4th ed., vol. 1. Lippincott Williams & Wilkins, Philadelphia, Pa.9+.hqe, 百拇医药
Cunningham, P. R., C. J. Weitzmann, and J. Ofengand. 1991. SP6 RNA polymerase stutters when initiating from an AAA... sequence. Nucleic Acids Res. 19:4669-4673.
Durbin, A. P., S. L. Hall, J. W. Siew, S. S. Whitehead, P. L. Collins, and B. R. Murphy. 1997. Recovery of infectious human parainfluenza virus type 3 from cDNA. Virology 235:323-332.*/wow6, 百拇医药
Durbin, A. P., J. W. Siew, B. R. Murphy, and P. L. Collins. 1997. Minimum protein requirements for transcription and RNA replication of a minigenome of human parainfluenza virus type 3 and evaluation of the rule of six. Virology 234:74-83.*/wow6, 百拇医药
Elroy-Stein, O., T. R. Fuerst, and B. Moss. 1989. Cap-independent translation of mRNA conferred by encephalomyocarditis virus 5' sequence improves the performance of the vaccinia virus/bacteriophage T7 hybrid expression system. Proc. Natl. Acad. Sci. USA 86:6126-6130.*/wow6, 百拇医药
Galinski, M. S. 1991. Annotated nucleotide and protein sequences for selected Paramyxoviridae, p. 537-568. In D. W. Kingsbury (ed.), The paramyxoviruses. Plenum Press, New York, N.Y.*/wow6, 百拇医药
Garcia-Barreno, B., T. Delgado, and J. A. Melero. 1994. Oligo(A) sequences of human respiratory syncytial virus G protein gene: assessment of their genetic stability in frameshift mutants. J. Virol. 68:5460-5468.
Garcia-Barreno, B., A. Portela, T. Delgado, J. A. Lopez, and J. A. Melero. 1990. Frame shift mutations as a novel mechanism for the generation of neutralization resistant mutants of human respiratory syncytial virus. EMBO J. 9:4181-4187.(+@6%, http://www.100md.com
Hall, S. L., A. Stokes, E. L. Tierney, W. T. London, R. B. Belshe, F. C. Newman, and B. R. Murphy. 1992. Cold-passaged human parainfluenza type 3 viruses contain ts and non-ts mutations leading to attenuation in rhesus monkeys. Virus Res. 22:173-184.(+@6%, http://www.100md.com
Hausmann, S., J. P. Jacques, and D. Kolakofsky. 1996. Paramyxovirus RNA editing and the requirement for hexamer genome length. RNA 2:1033-1045.(+@6%, http://www.100md.com
He, B., and R. A. Lamb. 1999. Effect of inserting paramyxovirus simian virus 5 gene junctions at the HN/L gene junction: analysis of accumulation of mRNAs transcribed from rescued viable viruses. J. Virol. 73:6228-6234.(+@6%, http://www.100md.com
Jacques, J. P., and D. Kolakofsky. 1991. Pseudo-templated transcription in prokaryotic and eukaryotic organisms. Genes Dev. 5:707-713.
Kawano, M., H. Bando, S. Ohgimoto, K. Kondo, M. Tsurudome, M. Nishio, and Y. Ito. 1990. Sequence of the fusion protein gene of human parainfluenza type 2 virus and its 3' intergenic region: lack of small hydrophobic (SH) gene. Virology 178:289-292.$2:^{8/, 百拇医药
Kawano, M., H. Bando, S. Ohgimoto, K. Okamoto, K. Kondo, M. Tsurudome, M. Nishio, and Y. Ito. 1990. Complete nucleotide sequence of the matrix gene of human parainfluenza type 2 virus and expression of the M protein in bacteria. Virology 179:857-861.$2:^{8/, 百拇医药
Kawano, M., H. Bando, T. Yuasa, K. Kondo, M. Tsurudome, H. Komada, M. Nishio, and Y. Ito. 1990. Sequence determination of the hemagglutinin-neuraminidase (HN) gene of human parainfluenza type 2 virus and the construction of a phylogenetic tree for HN proteins of all the paramyxoviruses that are infectious to humans. Virology 174:308-313.$2:^{8/, 百拇医药
Kawano, M., M. Kaito, Y. Kozuka, H. Komada, N. Noda, K. Nanba, M. Tsurudome, M. Ito, M. Nishio, and Y. Ito. 2001. Recovery of infectious human parainfluenza type 2 virus from cDNA clones and properties of the defective virus without V-specific cysteine-rich domain. Virology 284:99-112.
Kawano, M., K. Okamoto, H. Bando, K. Kondo, M. Tsurudome, H. Komada, M. Nishio, and Y. Ito. 1991. Characterizations of the human parainfluenza type 2 virus gene encoding the L protein and the intergenic sequences. Nucleic Acids Res. 19:2739-2746./fu, 百拇医药
Kolakofsky, D., T. Pelet, D. Garcin, S. Hausmann, J. Curran, and L. Roux. 1998. Paramyxovirus RNA synthesis and the requirement for hexamer genome length: the rule of six revisited. J. Virol. 72:891-899./fu, 百拇医药
Kroutil, L. C., K. Register, K. Bebenek, and T. A. Kunkel. 1996. Exonucleolytic proofreading during replication of repetitive DNA. Biochemistry. 35:1046-1053./fu, 百拇医药
Kunkel, T. A., S. S. Patel, and K. A. Johnson. 1994. Error-prone replication of repeated DNA sequences by T7 DNA polymerase in the absence of its processivity subunit. Proc. Natl. Acad. Sci. USA 91:6830-6834./fu, 百拇医药
Lamb, R. A., and D. Kolakofsky. 2001. Paramyxoviridae: the viruses and their replication., p. 1305-1340. In D. M. Knipe, P. M. Howley, D. E. Griffin, R. A. Lamb, M. A. Martin, B. Roizman, and S. E. Strauss (ed.), Fields virology, vol. 1. Lippincott Williams & Wilkins, Philadelphia, Pa.
Moeller, K., I. Duffy, P. Duprex, B. Rima, R. Beschorner, S. Fauser, R. Meyermann, S. Niewiesk, V. ter Meulen, and J. Schneider-Schaulies. 2001. Recombinant measles viruses expressing altered hemagglutinin (H) genes: functional separation of mutations determining H antibody escape from neurovirulence. J. Virol. 75:7612-7620.tx3/-$, http://www.100md.com
Murphy, S. K., and G. D. Parks. 1997. Genome nucleotide lengths that are divisible by six are not essential but enhance replication of defective interfering RNAs of the paramyxovirus simian virus 5. Virology 232:145-157.tx3/-$, http://www.100md.com
Newman, J. T., S. R. Surman, J. M. Riggs, C. T. Hansen, P. L. Collins, B. R. Murphy, and M. H. Skiadopoulos. 2002. Sequence analysis of the Washington/1964 strain of human parainfluenza virus type 1 (HPIV1) and recovery and characterization of wild type recombinant HPIV1 produced by reverse genetics. Virus Genes 24:77-92.tx3/-$, http://www.100md.com
Ohgimoto, S., H. Bando, M. Kawano, K. Okamoto, K. Kondo, M. Tsurudome, M. Nishio, and Y. Ito. 1990. Sequence analysis of P gene of human parainfluenza type 2 virus: P and cysteine-rich proteins are translated by two mRNAs that differ by two nontemplated G residues. Virology 177:116-123.
Peeters, B. P., O. S. de Leeuw, I. Verstegen, G. Koch, and A. L. Gielkens. 2001. Generation of a recombinant chimeric Newcastle disease virus vaccine that allows serological differentiation between vaccinated and infected animals. Vaccine 19:1616-1627.|ho, http://www.100md.com
Pelet, T., C. Delenda, O. Gubbay, D. Garcin, and D. Kolakofsky. 1996. Partial characterization of a Sendai virus replication promoter and the rule of six. Virology 224:405-414.|ho, http://www.100md.com
Poch, O., B. M. Blumberg, L. Bougueleret, and N. Tordo. 1990. Sequence comparison of five polymerases (L proteins) of unsegmented negative-strand RNA viruses: theoretical assignment of functional domains. J. Gen. Virol. 71:1153-1162.|ho, http://www.100md.com
Rager, M., S. Vongpunsawad, W. P. Duprex, and R. Cattaneo. 2002. Polyploid measles virus with hexameric genome length. EMBO J. 21:2364-2372.|ho, http://www.100md.com
Schmidt, A. C., J. M. McAuliffe, A. Huang, S. R. Surman, J. E. Bailly, W. R. Elkins, P. L. Collins, B. R. Murphy, and M. H. Skiadopoulos. 2000. Bovine parainfluenza virus type 3 (BPIV3) fusion and hemagglutinin-neuraminidase glycoproteins make an important contribution to the restricted replication of BPIV3 in primates. J. Virol. 74:8922-8929.
Shimokata, K., Y. Ito, Y. Nishiyama, and Y. Kimura. 1980. Influence of trypsin on the infectivity and biological properties or parainfluenza type 2 (croup-associated) virus in Vero cells. J. Gen. Virol. 48:407-410.?jkv&, 百拇医药
Sidhu, M. S., J. Chan, K. Kaelin, P. Spielhofer, F. Radecke, H. Schneider, M. Masurekar, P. C. Dowling, M. A. Billeter, and S. A. Udem. 1995. Rescue of synthetic measles virus minireplicons: measles genomic termini direct efficient expression and propagation of a reporter gene. Virology 208:800-807.?jkv&, 百拇医药
Skiadopoulos, M. H., A. P. Durbin, J. M. Tatem, S. L. Wu, M. Paschalis, T. Tao, P. L. Collins, and B. R. Murphy. 1998. Three amino acid substitutions in the L protein of the human parainfluenza virus type 3 cp45 live attenuated vaccine candidate contribute to its temperature-sensitive and attenuation phenotypes. J. Virol. 72:1762-1768.?jkv&, 百拇医药
Skiadopoulos, M. H., S. R. Surman, A. P. Durbin, P. L. Collins, and B. R. Murphy. 2000. Long nucleotide insertions between the HN and L protein coding regions of human parainfluenza virus type 3 yield viruses with temperature-sensitive and attenuation phenotypes. Virology 272:225-234.
Skiadopoulos, M. H., S. R. Surman, J. M. Riggs, C. Orvell, P. L. Collins, and B. R. Murphy. 2002. Evaluation of the replication and immunogenicity of recombinant human parainfluenza virus type 3 vectors expressing up to three foreign glycoproteins. Virology 297:136-152.\ezt, http://www.100md.com
Skiadopoulos, M. H., S. R. Surman, M. St Claire, W. R. Elkins, P. L. Collins, and B. R. Murphy. 1999. Attenuation of the recombinant human parainfluenza virus type 3 cp45 candidate vaccine virus is augmented by importation of the respiratory syncytial virus cpts530 L polymerase mutation. Virology 260:125-135.\ezt, http://www.100md.com
Stec, D. S., M. G. Hill, and P. L. Collins. 1991. Sequence analysis of the polymerase L gene of human respiratory syncytial virus and predicted phylogeny of nonsegmented negative-strand viruses. Virology 183:273-287.\ezt, http://www.100md.com
Stokes, A., E. L. Tierney, B. R. Murphy, and S. L. Hall. 1992. The complete nucleotide sequence of the JS strain of human parainfluenza virus type 3: comparison with the Wash/47885/57 prototype strain. Virus Res. 25:91-103. (Erratum, 27:96, 1993).
Streisinger, G., Y. Okada, J. Emrich, J. Newton, A. Tsugita, E. Terzaghi, and M. Inouye. 1966. Frameshift mutations and the genetic code. Cold Spring Harb. Symp. Quant. Biol. 31:77-84.4, http://www.100md.com
Tao, T., F. Davoodi, C. J. Cho, M. H. Skiadopoulos, A. P. Durbin, P. L. Collins, and B. R. Murphy. 2000. A live attenuated recombinant chimeric parainfluenza virus (PIV) candidate vaccine containing the hemagglutinin-neuraminidase and fusion glycoproteins of PIV1 and the remaining proteins from PIV3 induces resistance to PIV1 even in animals immune to PIV3. Vaccine 18:1359-1366.4, http://www.100md.com
Vulliemoz, D., and L. Roux. 2001. "Rule of six": how does the Sendai virus RNA polymerase keep count? J. Virol. 75:4506-4518.4, http://www.100md.com
Zheng, H., H. A. Lee, P. Palese, and A. Garcia-Sastre. 1999. Influenza A virus RNA polymerase has the ability to stutter at the polyadenylation site of a viral RNA template during RNA replication. J. Virol. 73:5240-5243.(Mario H. Skiadopoulos Leatrice Vogel Jeffrey M. Riggs Sonja R. Surman Peter L. Collins and Brian R. )
Received 29 July 2002/ Accepted 1 October 2002$2:^{8/, 百拇医药
ABSTRACT$2:^{8/, 百拇医药
Members of the Paramyxovirinae subfamily of the Paramyxoviridae family of viruses have the unusual requirement that the nucleotide length of the viral genome must be an even multiple of six in order for efficient RNA replication, and hence virus replication, to occur. Human parainfluenza virus type 2 (HPIV2) is the only member of the genus that has been reported to have a genome length that is not an even multiple of six, and it has also been recovered from a full-length antigenomic-sense cDNA that did not conform to the "rule of six." To reexamine the issue of nucleotide length in natural isolates of HPIV2, a complete consensus genomic sequence was determined for three HPIV2 strains: Greer, Vanderbilt/1994 (V94), and Vanderbilt/1998. Each of these strains was found to have a genome length of 15,654 nucleotides (nt), thus conforming in each case to the rule of six. To directly examine the requirement that the genomic length of HPIV2 be an even multiple of six, we constructed six full-length antigenomic HPIV2/V94 cDNAs that deviated from a polyhexameric length by 0 to 5 nt. Recombinant HPIV2s were readily recovered from all of the cDNAs, including those that did not conform to the rule of six. One recombinant HPIV2 isolate was completely sequenced for each of the nonpolyhexameric antigenomic cDNAs. These were found to contain small nucleotide insertions or deletions that conferred polyhexameric length to the recovered genome. Interestingly, almost all of the length corrections occurred within the hemagglutinin-neuraminidase and large polymerase genes or the intervening intergenic region and thus were proximal to the insert that caused the deviation from the rule of six. These results demonstrate, in the context of complete infectious virus, that HPIV2 has a strong and seemingly absolute requirement for a polyhexameric genome.
INTRODUCTION/fu, 百拇医药
Human parainfluenza virus type 2 (HPIV2) is an enveloped single-stranded negative-sense RNA virus of the genus Rubulavirus in the family Paramyxoviridae. Other rubulaviruses include simian virus 41 (SV41), SV5, HPIV4 A and B, and mumps virus. Like HPIV3 and HPIV1, which are members of the genus Respiroviridae in the paramyxovirus family, HPIV2 is a major cause of acute respiratory tract disease in infants and young children (4). HPIV2 has a genomic organization that is similar to that of the other parainfluenza viruses (4, 24), encoding seven proteins from six mRNAs. The HPIV2 gene map, based on the encoded mRNAs, is 3'-N-P/V-M-F-HN-L. The nucleoprotein (N), phosphoprotein (P), and large polymerase (L) complexed with genomic RNA make up the nucleocapsid/polymerase complex. The P protein is encoded by a version of the P/V mRNA that is cotranscriptionally edited to contain an insertion of two guanosine residues at the P gene editing site (28). The accessory V protein is encoded from an unedited version of the P/V mRNA. HPIV2 does not have genes encoding the accessory C and D proteins that are found in HPIV1 or HPIV3 and does not have a gene for the SH protein that is found in SV5 and mumps virus. The matrix (M), fusion (F), and hemagglutinin-neuraminidase (HN) proteins are the envelope-associated polypeptides.
With the exception of the Toshiba strain of HPIV2 (19, 34) (see below), each of the previously reported complete genomic sequences for members of the subfamily Paramyxovirinae, including all other members of the Rubulavirus genus, have genome lengths that are an even multiple of six (4). The requirement for a polyhexameric genome length has been termed the "rule of six." A polyhexameric length has been shown to be required for efficient in vitro replication of paramyxovirus minigenomes and defective interfering (DI) genomes (7, 29, 30, 35) and is thought to be a consequence of each N protein subunit interacting with exactly 6 nucleotides (nt) in the viral RNA (vRNA)-nucleoprotein complex (2, 21, 44). For the most part, the rule of six in paramyxoviruses has not been evaluated by direct experimentation in the context of complete infectious virus.tx3/-$, http://www.100md.com
Previously, the genome length of the Toshiba strain of HPIV2 was reported to be 15,646 nt (GenBank accession number X57559) (16-18, 20), which exceeds an even multiple of six by 4 nt (6n + 4). Also, an HPIV2 Toshiba strain cDNA that did not conform to the rule of six was successfully used to recover a recombinant HPIV2 in cell culture (19). These observations suggested that HPIV2, unlike other members of the Paramyxovirinae, might not have a strict requirement for its genome length to conform to the rule of six (19). However, it was possible that the reported nucleotide length was in error and that the recovered recombinant virus had sustained one or more mutations that corrected the genome to make it polyhexameric. For example, we have recently discovered that mutant HPIV3s can be derived from cDNAs that do not conform to the rule of six (38), although HPIV3 has a requirement for a polyhexameric genome length (6, 7, 9, 41). Sequence analysis of HPIV3 recombinants derived from cDNAs that did not conform to the rule of six showed that they contained nucleotide insertions that corrected the length of the viral genome so that it conformed to the rule of six (38).
To investigate whether HPIV2 conforms to the rule of six, we (i) sequenced three natural isolates of HPIV2 to determine the naturally occurring genome length, (ii) constructed an antigenomic cDNA, recovered recombinant HPIV2 (rHPIV2) with a polyhexameric genome and wild-type (wt)-like growth characteristics in vitro and in vivo, and (iii) systematically modified the length of the antigenomic cDNA to determine whether recombinant virus could be recovered from nonpolyhexameric antigenomes. The genomes of recovered recombinant viruses were completely sequenced to determine their polyhexameric status.|ho, http://www.100md.com
MATERIALS AND METHODS|ho, http://www.100md.com
Cell lines and viruses. HEp-2 (ATCC CCL 23) and LLC-MK2 (ATCC CCL 7.1) cells were maintained in OptiMEM I (Life Technologies, Gaithersburg, Md.) supplemented with 5% fetal bovine serum and gentamicin sulfate (50 µg/ml). HPIV2 strain V9412-6 (HPIV2/V94) was isolated in Vero cells from a nasal wash specimen provided by Peter Wright of Vanderbilt University. HPIV2/V94 was biologically cloned and amplified on Vero cells for a total of nine passages (43), followed by one passage on LLC-MK2 cells. HPIV2 strain V9811-18 (HPIV2/V98), also provided by Peter Wright, was isolated in rhesus monkey kidney (RhMK) cells from a nasal wash specimen and was passaged three times on LLC-MK2 monkey kidney cells. The HPIV2 Greer strain was obtained from the American Type Culture Collection (VR-92; lot no. 1450727) and was passaged once on LLC-MK2 cells. Viruses were quantified by limiting dilution, with virus-infected cultures identified by hemadsorption with guinea pig erythrocytes, as described previously (12, 38).
Determination of complete consensus sequences for the Greer, V94, and V98 strains of HPIV2. vRNA was isolated from the supernatant of cells infected with HPIV2 strain Greer, V94, or V98, as described previously (27), and was subjected to reverse transcription (RT) using the Thermoscript RT-PCR system (Invitrogen, Inc., Carlsbad, Calif.) and random hexamer primers. PCR was carried out on the reverse-transcribed cDNA product using the Herculase enhanced polymerase blend (Stratagene, La Jolla, Calif.) or the Advantage cDNA PCR kit (Clontech, Palo Alto, Calif.) and primers designed from the published Toshiba strain sequence (GenBank accession number X57559) (16-18, 20) or sequence obtained during the course of the sequence analysis. Direct sequence analysis of the RT-PCR products was performed using a Perkin-Elmer ABI 3100 sequencer with the BigDye terminator cycle sequencing ready reaction kit (ABI Prism, Foster City, Calif.). The sequences were assembled using the Autoassembler program (Perkin-Elmer Applied Biosystems).
The 3' and 5' genomic termini of each HPIV2 strain were amplified using the 3' RACE System for Rapid Amplification of cDNA Ends (Invitrogen, Inc.) or the 5' RACE System for Rapid Amplification of cDNA Ends, version 2.0 (Invitrogen, Inc.), as specified by the manufacturer. For each strain, the sequence for the 3' and 5' ends was determined with multiple sequencing reactions of two independently derived PCR products, all of which were confirmatory.\ezt, http://www.100md.com
Sequence alignments. The nucleotide sequences of HPIV2 strains V94, V98, and Greer were subjected to pairwise comparisons to determine the percent nucleotide identity. The deduced amino acid sequences of the proteins of HPIV2 strains V94, V98, and Greer were compared to those of the following paramyxoviruses: HPIV2 Toshiba strain, GenBank accession no. X57559; SV5 W3A strain, GenBank accession no. AF052755; SV41 strain Toshiba/Chanock, GenBank accession no. X64275; and mumps virus isolate 88-1961, GenBank accession no. AF467767. Percent amino acid identities for the proteins in Table 1 were calculated with the Gap program of the Wisconsin Package Version 10.2 (Genetics Computer Group, Madison, Wis.).
fig.ommitteed4, http://www.100md.com
TABLE 1. Amino acid sequence identity between the proteins of HPIV2/V94 and the analogous proteins of HPIV2 strains Greer and V98, SV41, SV5, and mumps virus4, http://www.100md.com
Assembly of a full-length rHPIV2/V94 cDNA antigenomic clone. A full-length cDNA clone encoding the HPIV2/V94 antigenomic RNA was constructed from six cloned overlapping RT-PCR and RACE products using the following restriction sites in the HPIV2 genome: NheI (nt 892 position in the complete antigenomic sequence), XhoI (nt 3673), Asp718 (nt 8141), AatII (nt 10342), and BsiWI (nt 15280). Briefly, the first fragment was generated from vRNA using the 3' RACE System for Rapid Amplification of cDNA Ends and the Advantage-HF cDNA PCR kit with specific HPIV2 primers and a primer containing an MluI site followed by a T7 promoter, two nonviral G residues, and the antigenomic cDNA corresponding to the 3' end of the viral genome. The next four fragments were reverse transcribed using the Thermoscript RT-PCR System and PCR amplified with the Herculase enhanced polymerase blend. The final fragment containing the 5' end was reverse transcribed using the 5' RACE System for Rapid Amplification of cDNA Ends (Invitrogen, Inc.) and was amplified with the Advantage-HF cDNA PCR kit using primers that added a partial hepatitis delta virus ribozyme ending with an RsrII site that is part of the ribozyme sequence. The individual PCR products were cloned into a pUC19 plasmid that had been modified to contain MluI, BspEI, XhoI, Asp718, AatII, BsiWI, and RsrII restriction enzyme sites to accept the PCR fragments. The cDNA clones were sequenced using a Perkin-Elmer ABI 3100 sequencer with the BigDye sequencing kit and were compared to the consensus sequence of HPIV2/V94 prior to assembly. All six PCR products were assembled into the above-mentioned pUC19 vector. The MluI-to-RsrII cDNA, containing the complete HPIV2/V94 antigenome sequence, was transferred into a modified pBlueScript that contained the rest of the hepatitis delta virus ribozyme followed by a T7 polymerase terminator sequence (27, 33). The resulting antigenomic plasmid, pFLC-HPIV2/V94, contains the following elements: a T7 promoter followed by two nonviral guanosine residues, the complete antigenomic 15,654-nt sequence of HPIV2/V94, a hepatitis delta virus ribozyme, and a T7 polymerase transcriptional terminator as previously described for the generation of recombinant HPIV3, BPIV3, and HPIV1 (6, 27, 33). The sequence of the full-length cDNA was verified by DNA sequence analysis. Two point mutations were identified in the F and L open reading frames (ORFs) of the full-length cDNA that had not been present in the consensus sequence and presumably had occurred during the assembly of the complete clone. These occurred at positions 6265 (AGT to AGC; silent) and 15075 (GCT to GTT; Ala-2093 to Val). These mutations were used as markers to distinguish the recombinant HPIV2 from the biologically derived HPIV2/V94 parent.
HPIV2/94 N, P, and L support plasmids for the recovery of HPIV2/V94 from cDNA. A support plasmid encoding the N protein of HPIV2/V94 (pTM-N2) was derived from vRNA using the Thermoscript RT-PCR system and the Advantage-HF PCR kit using an antigenomic-sense oligonucleotide that contained an AflIII site spanning the N ORF ATG translation initiation codon site and an antisense oligonucleotide containing an EcoRI site. The PCR product was digested with AflIII and EcoRI and cloned into pTM1 (6-8) that had been digested with NcoI and EcoRI.*:q^%[, http://www.100md.com
The HPIV2/V94 P protein expression plasmid (pTM-P2) was generated from two overlapping PCR fragments (25) and was engineered to contain the two inserted guanosine residues at the HPIV2 P gene editing site (nt 2481 to 2487) that distinguish the complete P ORF from that of V. The P ORF was cloned into pTM1 as an NcoI-to-EcoRI fragment.*:q^%[, http://www.100md.com
An HPIV2 L polymerase expression plasmid (pTM-L2) was assembled from a PCR product and two subclones of L. The upstream end of L was amplified by PCR with a sense oligonucleotide containing an NcoI site spanning the L gene ATG translation initiation codon, and an antisense oligonucleotide downstream of a unique AatII site (nt 10342) in the L ORF. The remainder of the L ORF was derived from a subclone used to construct the HPIV2 full-length clone. The complete HPIV2/V94 L ORF was then cloned as an NcoI-to-RsrII fragment into a version of pTM1 that was modified to contain RsrII as a downstream site (27).
Recovery and sequence analysis of rHPIV2/V94. rHPIV2/V94 was recovered from the full-length antigenomic cDNA essentially as described previously (37). Briefly, HEp-2 cells in six-well plates were cotransfected with a full-length HPIV2/V94 cDNA plasmid and the three HPIV2 support plasmids (pTM-N2, pTM-P2, and pTM-L2), using Lipofectamine-2000 reagent (Invitrogen, Inc.). The HEp-2 cells were simultaneously infected with MVA-T7 as described previously (6, 33). The medium supernatant was harvested on day 3 or 4 posttransfection and passaged two times on LLC-MK2 cell monolayers. rHPIV2/V94 was then biologically cloned by plaque-to-plaque purification on LLC-MK2 monolayers and was further propagated on LLC-MK2 cells, as previously described (39). To confirm that the recovered viruses had been derived from cDNA, RT was performed and segments of the viral genome were amplified by PCR. Sequence analysis of the PCR products showed that the two point mutations in the F and L genes present in the antigenomic cDNA as described above were contained in the recombinant virus, designated rHPIV2/V94, and not in the wt parental virus. Two preparations of HPIV2/V94 (designated rAHPIV2/V94 and rBHPIV2/V94) were made, each derived from a separate transfection. vRNA was isolated from the biologically cloned rAHPIV2/V94 preparation as described previously (27) and was used as the template for RT-PCR and RACE using oligonucleotide primers. The amplified products were analyzed by DNA sequencing to determine the complete consensus genomic sequence of rAHPIV2/V94.
Assessment of the rule of six in complete recombinant virus. The full-length HPIV2/V94 antigenomic cDNA was modified by the insertion of oligonucleotide duplexes, which were 42, 43, 44, 45, 46, or 47 bp long, into the EcoRV restriction site present near the downstream end of the L ORF (nt 15554 to 15559) This resulted in six antigenomic cDNAs that are 6n, 6n + 1, +2, +3, +4, or +5 with regard to polyhexamer length. The correct orientation and sequence of each insert was confirmed. These antigenomic cDNAs were transfected into HEp-2 cells along with HPIV2 support plasmids and coinfected with MVA-T7 as described above. Recovered virus was passaged and biologically cloned on LLC-MK2 monolayers as described above, with one or two independent virus preparations made from each antigenomic plasmid. vRNA was isolated and used to generate RT-PCR products spanning the entire genome as described above. These were then directly sequenced to generate a complete consensus sequence for each virus.s;, http://www.100md.com
Replication of HPIV2 in vitro. Recombinant or biologically derived HPIV2 was inoculated in triplicate onto LLC-MK2 cell monolayers in six-well plates at a multiplicity of infection of 0.01, and cultures were incubated at 32°C with and without 5 µg of porcine trypsin/ml added to the culture medium, as described previously (38). Medium (0.5 ml) from each well was harvested and replaced with 0.5 ml of fresh medium at 0 h and at 1 to 6 days postinfection. Virus present in the samples was quantified by titration on LLC-MK2 monolayers in 96-well plates that were incubated for 6 days at 32°C. Virus grown in the presence of trypsin was titered with trypsin in the medium. Virus was detected by hemadsorption, and the titer is reported as log10 50% tissue culture infectious dose (TCID50)/ml.
Replication of recombinant and wt viruses in hamsters. Four-week-old Golden Syrian hamsters (Charles River Laboratories, Stoneridge, N.Y.) in groups of six were inoculated intranasally with 0.1 ml of L15 medium containing 106.0 TCID50 of virus. On day 3, 4, or 5 postinfection, the lungs and nasal turbinates were harvested and the virus was quantified by serial dilution of tissue homogenates on LLC-MK2 monolayers, as previously described (36). The mean titer of virus was calculated for each day for each group of hamsters and is expressed as log10 TCID50/gram of tissue.8yx5(, http://www.100md.com
Nucleotide sequence accession numbers. The nucleotide sequences of HPIV2 strains V94, V98, and Greer were submitted to GenBank under accession numbers AF533010, AF533011, and AF533012, respectively.8yx5(, http://www.100md.com
RESULTS8yx5(, http://www.100md.com
Determination of the complete nucleotide sequence of three strains of HPIV2. The biologically derived HPIV2 strain Vanderbilt 9412-6 (HPIV2/V94) was obtained in 1994 from a 13-month-old infant presenting with upper and lower respiratory tract infection, high fever, and otitis media and therefore represents a recent, low-passage-number isolate of HPIV2. The complete genomic sequence of HPIV2/V94 was determined from RT-PCR products amplified from purified vRNA. The sequence analysis was performed directly on RT-PCR products without a cloning step and thus yields a consensus sequence. The HPIV2/V94 genome was found to be 15,654 nt in length and thus conforms to the rule of six (2, 21). Two additional HPIV2 strains were also sequenced. The prototype HPIV2 Greer strain was originally isolated in 1955 from an 11-month-old infant with respiratory tract disease (3), and the HPIV2 Vanderbilt 9811-18 (V98) strain was isolated in 1998 from a 20-month-old infant with respiratory tract disease. The complete consensus genome sequence was determined for the HPIV2/Greer and HPIV2/V98 strains and, in each case, was 15,654 nt in length, the same as HPIV2/V94. Thus, three strains of HPIV2, isolated over a period of 40 years, have the identical polyhexameric genome length.
Sequence comparison of the HPIV2 strains. The percent identity of the nucleotide sequence of the Greer, V94, and V98 strains was 95.0% for the V94 and V98 strains, 97.4% for the V98 and Greer strains, and 96.2% for the Greer and V94 strains. Thus, the recent V94 and V98 isolates of HPIV2 are highly related to the 1955 Greer isolate, with each being more closely related to the Greer strain than to each other. The cis-acting transcription gene-start (GS), intergenic (IG), and gene-end (GE) signal sequences were highly conserved. In addition, the phasing of the GS signals, i.e., their position relative to the hexamer spacing of the genome, was the same for all three strains (not shown), which may be important for efficient expression of paramyxovirus mRNAs (21). As is typical for nonsegmented negative-strand viruses, the nucleotide sequences of the ORFs were more highly conserved than were the nontranslated gene sequences (not shown).m;5[kg&, http://www.100md.com
The predicted amino acid sequences of the HPIV2/V94 N, P, V, M, F, HN, and L proteins were compared to their counterparts in Greer, V98, and several other members of the rubulaviruses, and the percent identities are shown in Table 1. The predicted amino acid sequences are very highly conserved between the three HPIV2 strains, with the highest degree of divergence occurring in the HN glycoprotein. Comparison with SV41, SV5, and mumps virus showed a decreasing level of sequence identity between each and HPIV2, in the order of SV41 first with the highest level of identity, and then SV5, and then mumps virus, with L being the most highly conserved protein.
Construction of a full-length HPIV2/V94 antigenomic cDNA and recovery of recombinant wt HPIV2/V94. A complete HPIV2/V94 antigenomic cDNA, designated pFLC-HPIV2/V94, was constructed from six overlapping RT-PCR and RACE products. Nucleotide sequencing of the complete antigenomic cDNA showed that it contained two nucleotide substitutions in the F and L genes, respectively, compared to the consensus sequence of biologically derived HPIV2/V94. The substitution at position 6265 (T to C) in the F gene is translationally silent, whereas that at nt 15075 (C to T) encodes an alanine-2093-to-valine substitution in the L protein. The amino acid at this position is not conserved among other parainfluenza viruses and lies outside of the six major domains that are conserved among paramyxoviruses (31, 40). These nucleotide differences served as markers for a recombinant HPIV2 derived from cDNA.n!'d, 百拇医药
The antigenomic cDNA was transfected into HEp-2 cells, and virus was recovered by cotransfection with the HPIV2/V94 N, P, and L support plasmids and coinfection with MVA-T7. Virus was readily recovered after only a single passage of the HEp-2 transfection supernatant onto LLC-MK2 cell monolayers, and the presence of the nucleotide markers in the HPIV2 F and L genes was confirmed by RT-PCR of vRNA and sequence analysis. Amplification of RT-PCR products was dependent on the addition of RT, indicating that the template was indeed viral RNA and not contaminating DNA (data not shown). Two separate recombinant HPIV2 isolates were biologically cloned from independent transfections and were designated rAHPIV2/V94 and rBHPIV2/V94. The complete sequence of rAHPIV2/V94 was determined from RT-PCR products, including determination of the two ends by RACE. The size of the recovered recombinant HPIV2 genome was identical to that of its biologically derived parent and, thus, conformed to the rule of six. The genomic sequence of the recovered virus was identical to the cDNA from which it was derived except for a silent nucleotide substitution at nt 13786 (T to C) that presumably arose during propagation of the recovered virus in cell culture. rBHPIV2/V94 was not sequenced in its entirety, but it did not have this nucleotide change.
Characterization of the in vitro and in vivo growth properties of rHPIV2/V94. The growth kinetics and final yield of the two preparations of recombinant HPIV2, rAHPIV2/V94 and rBHPIV2/V94, were compared to those of their biologically derived HPIV2 parent at 32°C in LLC-MK2 cells and were found to be essentially indistinguishable (Fig. 1A). Growth at 32, 39, and 40°C was examined, and two rHPIV2/V94 isolates and their biologically derived parent were not temperature sensitive at the higher temperatures (data not shown). Thus, the incidental nucleotide substitutions in rHPIV2/V94 did not alter growth properties in vitro. To determine if HPIV2/V94 requires the addition of trypsin for efficient growth in vitro, we compared the effect of added porcine trypsin on the in vitro growth kinetics and final yield of the recombinant and biologically derived HPIV2s. The addition of trypsin did not increase the growth efficiency of the biologically derived or recombinant viruses (data not shown).
fig.ommitteed|!dnn, 百拇医药
FIG. 1. Replication of biologically derived and recombinant HPIV2/V94 in vitro and in vivo. (A) Multistep growth kinetics in vitro. LLC-MK2 cell monolayers were infected in triplicate at a multiplicity of infection of 0.01 with biologically derived HPIV2/V94 or with two preparations of rHPIV2/94 derived from separate transfections (rAHPIV2/V94 and rBHPIV2/V94). The cells were incubated at 32°C. Aliquots of the medium supernatants were harvested at 24-h intervals, and titer of virus was determined at 32°C. Titers of virus are expressed as mean log10 TCID50/ml ± standard error. (B) Replication in hamsters. Hamsters were inoculated intranasally with 106 TCID50 of the same viruses as in panel A. Nasal turbinates and lung tissues from six animals from each group were harvested on days 3, 4, and 5 postinfection. Virus present in tissues was quantified by serial dilution on LLC-MK2 monolayers at 32°C.|!dnn, 百拇医药
To determine whether the recovered rHPIV2/V94 retained the replicative properties of its biologically derived parent in vivo, the replication of the two preparations of rHPIV2/V94 in the respiratory tracts of hamsters was compared to that of the biologically derived HPIV2/V94 (Fig. 1B). The level of replication of the two preparations of rHPIV2/V94 was essentially indistinguishable from that of the biologically derived HPIV2/V94 parent on all of the days tested, demonstrating that the recombinant version of HPIV2/V94 has replicative properties similar to those of its biologically derived counterpart and therefore is an authentic wt virus. Thus, the incidental amino acid substitution at position 2093 of the L polymerase also did not affect the level of replication of rHPIV2/V94 in vivo. However, a full-length antigenomic cDNA encoding an alanine at position 2093 of the L protein, as occurs in biologically derived HPIV2/V94, was also generated for use in future studies, and this modified HPIV2/V94 cDNA was also used to recover recombinant HPIV2/V94 in vitro (data not shown).
Recombinant HPIV2s generated from cDNAs that do not conform to the rule of six contain nucleotide insertions or deletions that serve to yield polyhexameric genome lengths. The requirement for a polyhexameric genome length has not been directly examined for HPIV2. Also, it was of interest to systematically examine the requirements for the rule of six, and the consequences of deviations from the rule, in the context of complete infectious virus. We generated six full-length antigenomic cDNAs of HPIV2/V94 that were modified to contain oligonucleotide insertions (Fig. 2) at an EcoRV restriction site near the end of the L polymerase gene. This site was chosen since it did not lie within a known cis-acting sequence and would not interfere with the hexamer phasing of upstream cis-acting sequences. The inserts were designed to restore the L polymerase ORF sequence but increased the length of the downstream noncoding region of L, and hence the complete antigenomic cDNA, by 42 nt (6n with respect to hexamer spacing, a control cDNA conforming to the rule of six), 43 nt (6n + 1), 44 nt (6n + 2), 45 nt (6n + 3), 46 nt (6n + 4), or 47 nt (6n + 5) (Fig. 2).
fig.ommitteed1?zeh%, http://www.100md.com
FIG. 2. Structure of the nucleotide inserts used to make HPIV2 antigenomic cDNAs that do not conform to the rule of six. (A) The wt nucleotide sequence around the EcoRV restriction site near the end of the L ORF is shown. (B) The sequences of the six antigenomic-sense oligonucleotides inserted at the EcoRV restriction site spanning HPIV2 nt 15554 to 15559 is shown. Oligonucleotide duplexes were inserted between nt 15556 and 15557. Each oligonucleotide duplex contains a silent ATC-to-ATT mutation that destroys the EcoRV site and recreates the last 11 codons of the L ORF, and 12 HPIV2 nt including the TGA stop codon (bold) followed by zero to five additional nucleotides. The designation of the recombinant viruses generated from the cDNAs is indicated to the left: the virus designated rHPIV2/V94(+6) conforms to the rule of six, while the others contain the indicated number of additional nucleotides (+1 to +5). The length of each oligonucleotide insert is shown on the right, with the rule-of-six length and the total length of the antigenomic cDNA indicated in parentheses.1?zeh%, http://www.100md.com
The antigenomic cDNAs were transfected into HEp-2 cells along with HPIV2 support plasmids, as described above. The supernatants were harvested after 3 or 4 days and were passaged twice on LLC-MK2 monolayer cultures. Surprisingly, virus was readily recovered from every cDNA tested. This indicated that although naturally occurring HPIV2 genomes are polyhexameric, recombinant HPIV2 can be generated from a cDNA that does not conform to the rule of six. Recombinant HPIV2s recovered from cDNAs that did not conform to the rule of six [rHPIV2/V94(+1), rHPIV2/V94(+2), rHPIV2/V94(+3), rHPIV2/V94(+4), and rHPIV2/V94(+5)] were biologically cloned by plaque purification on LLC-MK2 monolayer cultures and were further amplified by passage on LLC-MK2 cells, as described above. vRNA was isolated from each of these recombinants and was used to perform RT-PCR and RACE. The PCR products were sequenced directly, yielding a complete consensus sequence for at least one recovered viral isolate for each of the nonpolyhexameric antigenomes (see below).
For each virus that was recovered from a nonpolyhexameric antigenomic cDNA, nucleotide sequence analysis revealed small nucleotide insertions or deletions that restored the recovered genome length to conform to the rule of six (Fig. 3). For example, the complete consensus sequence of the genome of the rHPIV2/V94(+3) isolate revealed that this virus had acquired a 2-nt (UU, genome sense) insertion in the HN gene 5' noncoding region (NCR) and a 1-nt (A, genome sense) insertion within a poly(A) tract in the HN gene 3' NCR (Fig. 3A). The genome size of the recovered rHPIV2/V94(+3) isolate was 15,702 nt, i.e., 3 nt longer than the HPIV2 cDNA from which it was derived. Thus, the 3-nt insertion served to modify the viral genome length so that it conforms to the rule of six. Complete sequencing of the rHPIV2/V94(+4) genome showed that this virus had sustained a 2-nt insertion (UA, genome sense) at the end of the intergenic region between the HN and L genes and immediately upstream of the L gene GS signal sequence (Fig. 3A). The genome size of rHPIV2/V94(+4) was 15,702 nt, i.e., 2 nt longer than the 6n + 4 HPIV2 cDNA from which it was derived. Thus, the 2-nt insertion served to generate a polyhexameric genome length. Two independently derived cloned isolates of rHPIV2/V94(+5) were prepared, rAHPIV2/V94(+5) and rBHPIV2/V94(+5). Complete consensus sequence analysis of rBHPIV2/V94(+5) showed that it has a single nucleotide insertion in the HN 3' untranslated region that restored polyhexameric length. The other [rAHPIV2/V94(+5)] was partially sequenced and was found to contain a single nucleotide insertion in the HN GE signal sequence that could serve to correct the genome length so that it conforms to the rule of six (Fig. 3A). rHPIV2/V94(+1) and rHPIV2/V94(+2) were completely sequenced and were found to have 1- and 2-nt deletions, respectively, that served to generate a polyhexameric genome (Fig. 3B). The 2-nt deletion that was found in rHPIV2/V94(+2) occurred within the intergenic sequence between the HN and L genes. Interestingly, the single nucleotide deletion in rHPIV2/V94(+1) occurred near the end of the coding region for the L polymerase. This resulted in a shift in the reading frame, such that the last 13 amino acids of the L protein were deleted and replaced by 21 amino acids encoded by the alternate reading frame. Thus, the majority of genome length corrections occurred at sites in noncoding regions. Interestingly, almost all of the changes were located within or between the HN and L genes, the two genes most proximal to the insert responsible for the deviation from the rule of six.
fig.ommitteedo+n\c^c, 百拇医药
FIG. 3. Nucleotide insertions and deletions detected in the genomic RNA of recombinant HPIV2s derived from cDNAs that do not conform to the rule of six (sequences are in antigenomic sense). (A) Four recombinant viruses that were produced from cDNAs that did not conform to the rule of six, rHPIV2/V94(+3), rHPIV2/V94(+4), rAHPIV2/V94(+5), and rBHPIV2/V94(+5), contained nucleotide insertions that resulted in a polyhexameric genome length. Two insertions containing a total of 3 nt were identified in the HN 5' and 3' noncoding regions (5' NCR and 3' NCR) of rHPIV2/V94(+3). A 2-nt insertion within the IG region between the HN and L genes was identified in rHPIV2/V94(+4). A 1-nt insertion was found in each clone of rAHPIV2/V94(+5) and rBHPIV2/V94(+5), in one case at the end of the HN GE signal and in the other in the HN 3' NCR. (B) Two recombinant viruses produced from cDNAs that did not conform to the rule of six, rHPIV2/V94(+2) and rHPIV2/V94(+1), contained nucleotide deletions that resulted in a polyhexameric genome length. rHPIV2/V94(+2) was found to have a 2-nt deletion (underlined) within the IG region between the HN and L genes. rHPIV2/V94(+1) was found to have a 1-nt deletion (underlined) near the end of the L polymerase ORF. This resulted in a frameshift in the L coding sequence that deleted the last 13 amino acids of L and replaced them with an unrelated sequence of 21 amino acids, as shown. The nucleotide sequence (antigenomic sense) in the region of the insertions is shown. The inserted nucleotides are shown, and the site of each insertion is indicated by an arrow pointing downward. Deleted nucleotides are shown, and the location of each deletion is indicated by an arrow pointing upward. The HN 5' or 3' NCR, transcription GE and GS (in bold type), and IG regions between the HPIV2 HN and L ORFs are indicated. The L polymerase translation initiation codon (ATG) and translation termination codon (TGA) are underlined. The single-letter amino acid designation is shown below the nucleotide sequence for the wt and a mutant version of the HPIV2 L polymerase.
DISCUSSIONj)r$-, 百拇医药
We investigated the question of whether the natural genome length of HPIV2 conforms to the rule of six by determining the complete genomic sequence for three strains of HPIV2, namely the prototype Greer strain and two recent HPIV2 isolates, V94 and V98. Each of these strains has a polyhexameric 15,654-nt genome length, demonstrating that the HPIV2 genome length is conserved and conforms to the rule of six. A polyhexameric antigenomic HPIV2/V94 cDNA was used to recover a recombinant HPIV2/V94 with wt growth characteristics, and the length of the recovered virus was confirmed to be 15,654 nt. Full-length antigenomic HPIV2 cDNAs that were engineered to deviate from the rule of six by 1 to 5 nt yielded viruses that contained small nucleotide insertions or deletions that served to generate polyhexameric genomes. These findings demonstrate that an HPIV2 polyhexameric genome can be generated from a cDNA that does not conform to the rule of six but that virus recovery appears to depend on the occurrence of spontaneous mutations that modify the length of the genome to conform to the rule of six. These findings suggest that HPIV2 has a strong preference for a polyhexameric genome. Thus, the requirement for a polyhexameric genome has now been directly confirmed in complete recombinant virus for HPIV2 and, in earlier work, HPIV3 (38), representing the Rubulavirus and Respirovirus genera of the Paramyxovirinae subfamily. This suggests that adherence to the rule of six and the ability to correct deviations from the rule may be universal for all members of this paramyxovirus subfamily.
Comparison of the genome sequence of the Greer strain with that reported previously for the Toshiba strain revealed that these genomes share 99.7% nucleotide sequence identity. Thus, the Toshiba and Greer strains are nearly identical, with a total of only 34 nucleotide differences. Surprisingly, these differences included a large number of apparent insertions and deletions in one strain versus the other. For example, 13 nt present in the Greer strain were missing from the reported Toshiba sequence, representing either deletions in the Toshiba strain or insertions in the Greer, and 5 nt reported for the Toshiba sequence were not present in Greer, representing either insertions in Toshiba or deletions in Greer. It seems unlikely that so many apparent insertions or deletions would exist in two such highly related strains, particularly since such spacing discrepancies were not evident between the Greer, V94, and V98 strains. Correction for these 18 insertions and deletions in the Toshiba strain would yield a genome length of 15,654, the same size as reported here for the V94, V98, and Greer strains of HPIV2. This suggests that these apparent insertions and deletions might represent errors in cDNA synthesis, cloning, and sequence analysis rather than true differences in nature. Regarding the previous demonstration that recombinant Toshiba strain virus can be recovered from a nonpolyhexameric antigenomic cDNA (19), the findings in the present paper indicate that there is a stringent mechanism for the very rapid emergence of polyhexameric genomes in recombinant virus derived from nonpolyhexameric antigenomes. Thus, it seems likely that the recovered recombinant Toshiba virus contains a corrected, polyhexameric genome.
In the present study, the source of the insertions and deletions that served to correct rHPIV2/V94 recovered from nonpolyhexameric antigenomes is not known. However, two likely sources are mutations introduced by T7 RNA polymerase during the synthesis of the initial antigenome from the transfected plasmid, mutations introduced by the HPIV2 polymerase during subsequent RNA replication, or both. In particular, the RNA polymerases of bacteriophage T7 and nonsegmented negative-strand viruses share the propensity to insert or delete nucleotides within homopolymeric tracts, which would be consistent with the finding in the present study that four of the five insertions, and one of the two deletions, occurred in a homopolymeric tract of three or more nucleotides. T7 RNA polymerase has been shown to insert or delete nucleotides within homopolymeric sequences consisting of three or more nucleotides at an estimated rate of up to 1 in 4.5 x 107 nt (22, 23). This misincorporation activity is likely a result of "stuttering," which involves slippage by the polymerase and nascent strand during the copying of repetitive sequences (5, 22, 42). Stuttering also has been described for eukaryotic and other viral polymerases, including the paramyxovirus L polymerases.
Indeed, stuttering is a normal feature of RNA synthesis by the L polymerase of nonsegmented negative-strand RNA viruses: stuttering at the poly(U) tract (genome-sense) of the GE signals is responsible for the synthesis of the poly(A) tails of mRNA, and stuttering at the editing site of the P gene is responsible for the insertion of one or more G residues (mRNA-sense) that give rise to frame-shifted versions of the P mRNA (15). There are numerous examples of changes in genome length during RNA replication by the negative-strand RNA polymerase. It was previously shown that Sendai virus minigenomes whose lengths did not conform to the rule of six and which contained the P gene editing site [a poly(G) stretch] underwent in vitro nucleotide insertions or deletions within the editing site that served to generate polyhexameric genome lengths (13). DI genomes of SV5 were also shown to undergo in vitro genome length modification by nucleotide insertions, and SV5 DI genomes whose length conformed to the rule of six replicated more efficiently than uncorrected genomes (26). Other negative-stranded RNA viruses have also been shown to generate modified genomes that contain nucleotide insertions. Vesicular stomatitis virus, which is a member of the Rhabdoviridae family and has a genomic organization similar to that of paramyxoviruses but which does not follow the rule of six, has been shown to accumulate many nucleotide insertions in the 3' noncoding region of the glycoprotein gene during the natural evolution of the virus (1). The nucleotide insertions identified included homopolymeric stretches of U and/or A, and it was suggested that these occurred by polymerase stuttering. Respiratory syncytial virus, a paramyxovirus that does not follow the rule of six, is able to generate neutralization resistant mutants by inserting or deleting nucleotides (A) within a homopolymeric A stretch in the G protein gene (10, 11). Another negative-stranded RNA virus, influenza virus type A, has also been shown to incorporate U's at a poly(U) stretch in the neuraminidase gene by stuttering during genome replication (45).
The length-correcting mutations identified in the present study had four characteristics: (i) they occurred predominantly in noncoding regions rather than in coding regions; (ii) they were exclusively A or U; (iii) they usually occurred in homopolymer tracts, as already noted; and (iv) they all occurred within one or two genes on the upstream side (relative to the genome map) of the insert that was responsible for the deviation from the rule of six. It is noteworthy that these same four characteristics also hold for the length corrections observed in a previous study in which rHPIV3s were recovered from antigenomes that deviated from the rule of six due to the insertion of a nonhexameric foreign gene (38).|ji@, 百拇医药
Interestingly, in this previous HPIV3 study, the insert that caused the deviation from the rule of six had been placed at two different genome locations, namely between the N and P or the P and M genes. Two of the six examples of correcting mutations that were found in this previous study occurred within the foreign gene insert and in four cases within the homopolymer tract of the GE signal of the first or second gene on the upstream side of the insert (reference 38 and unpublished data). Recently it was reported that recombinant polyploid measles viruses containing foreign gene insertions that were apparently deleterious to virus replication accumulated nucleotide insertions that inactivated the foreign gene and a compensating nucleotide deletion mutation that restored polyhexameric genome length downstream of the site of insertion (32). Thus, these genome length correction mutations differed from those described in the present study and previously (38) in that the location of the correcting mutation was downstream of the nucleotide insertion site that resulted in deviation of the genome length from the rule of six.
The low frequency of genome length corrections in ORFs is not surprising, since the insertion or deletion of 1 or 2 nt within an ORF would shift the reading frame and, if this affected an essential protein, would be lethal. A 3-nt insertion or deletion would add or delete an amino acid, which also would frequently be deleterious. The only genome length modification that occurred within an ORF was the nucleotide deletion in rHPIV2/V94 (+1). This occurred near the end of the L ORF, shifting the reading frame such that the C-terminal 13 amino acids of L were deleted and replaced by 21 heterologous amino acids encoded by the second reading frame. Since the recombinant bearing this mutation was readily recovered and grew to a high titer, it suggests that this region of the L protein is dispensable for viral replication in vitro. In the previous HPIV3 study, two of the six identified genome length corrections occurred within the ORF of the foreign gene insert (38), but this would not be deleterious to the virus since the foreign gene would not be required for virus replication.
It was unexpected to observe that the nucleotide insertions and deletions involved exclusively A or U. This also was observed in the previous HPIV3 study noted above. This might be a consequence of the inability of the replicase to modify genome length by adding or deleting G or C residues, or it might simply reflect a lower content of homopolymeric G and C in the regions most amenable to accepting a length correction, namely the nontranslated regions and the IG regions. The finding that most of the length-correcting mutations were insertions, and that most of them occurred in homopolymeric tracts, is consistent with a stuttering mechanism that might have some similarity to that involved in polyadenylation. Indeed, 5 of the 10 insertions identified in the two studies involved a GE signal. The present study also contained two examples involving the insertion or deletion of an AU from an AUAU sequence, each of which might represent a 2-nt slippage.*wm, 百拇医药
Perhaps the most unexpected finding of the present study was the distribution of the length-correcting mutations. It might have been expected that length corrections in recovered virus would be found throughout the genome, subject only to a preference for A or U homopolymers and a bias against disturbing an ORF or cis-acting signal. Instead, these mutations occurred with a biased distribution: specifically, they were always upstream of the insert causing the deviation from the rule of six, and they always occurred within the first or second upstream genes (in the present study, within the HN and L genes). This also had been observed for the previous HPIV3 study, in which the deviating insert was placed between N and P, or P and M, and the corrections always occurred on the upstream side and always within one or two genes. This suggest that the length correction by the two HPIVs operates by a common mechanism that might involve either (i) random length correction, followed by the stringent selection for virus in which the correction was close to the point of deviation, or (ii) nonrandom length correction, perhaps involving a replication complex that "senses" the deviation from the rule of six and acts to insert a correcting mutation at a second, downstream site in the nascent molecule.
One factor that could affect the location of the insertions and deletions is the phasing of the GS signals within the genome. It has been noted that the position of the first nucleotide of the GS signals of members within each of the three genera of the Paramyxovirinae is partially or completely conserved for each gene, suggesting that the position of the GS signals relative to the hexamer phasing of the genome might be important for transcription (21). The introduction of a length-correcting mutation into a genome would alter the hexameric phasing of all downstream GS signals. Thus, if the phasing of the GS signals indeed is important, there would be a selective advantage to viruses in which the length correction occurs as close as possible to the insert that caused the deviation from the rule of six, thereby disturbing the phasing of the fewest possible GS signals. For example, in the present study, the deviation from the rule of six was engineered in the downstream end of the L gene, and each of the correcting mutations occurred in the HN and L genes and altered the phasing of no more than one GS signal. In the previous HPIV3 study, the insert causing deviation from the rule of six was between N and P or P and M, and the correcting mutations occurred such that in three cases a single vector GS signal was out of phase, and in the other cases all of the GS signals were in phase (38).
Thus, the phasing of the GS signals might provide a selective pressure that accounts for the biased distribution of corrections in recovered virus. As a caveat, it should be noted that while the hexamer phasing of the GS signals for HPIV2 is conserved between the V94, V98, and Greer strains, the rubulaviruses have the lowest degree of conservation of GS phasing among the three genera of the Paramyxovirinae. Also, while the effect of GS phasing on paramyxovirus transcription and replication has not been carefully analyzed, one study in which the phasing of the L GS signal of SV5 was altered showed that the effect was modest. However, that study involved a single misphased GS signal, and it might be that a single misphased GS signal is tolerated, while the aggregate effect of two or more misphased signals is sufficient to provide a negative selective pressure (14). This model of random correction followed by biased selection has the advantage of simplicity, but it does not explain why the length corrections were always on the upstream side of the deviation from the rule of six.
Alternatively, nonrandom correction might be involved. By this model, the viral replication machinery somehow senses a deviation in the genome length or phasing in the template during synthesis and makes a length adjustment near the area where the genome has been altered. It is not clear how the replication machinery might sense a deviation, but one possibility is that it involves the hexamer phasing of some repeated genome sequence, such as the GS or GE signals or some other structure that has not been identified. Thus, a deviation in the phasing of a signal in the template would be detected by the polymerase in the process of RNA synthesis, and a length correction would follow. This model would account for the universal asymmetry of the length corrections with regard to the original deviation from the rule of six. The finding that all of the corrections occurred upstream in the gene map relative to the point of deviation would be consistent with length correction during the synthesis of the genome-sense strand.
It is not clear from the present study whether the ability to correct deviations in genome length in such a seemingly rapid and efficient manner is an artifact of the reverse genetics system or an intrinsic property of the viral replication machinery that is used during the natural evolution of the virus to maintain proper genome length. The conservation of the length of the HPIV2 genome in strains isolated more than 40 years apart, the numerous examples of negative-stranded RNA viruses that have the ability to modify their genome length, and the ease with which recombinant parainfluenza viruses can be derived from cDNAs that do not conform to the rule of six suggest that the latter may be the case. The infectious HPIV2 cDNA derived from a consensus wt HPIV2 sequence described here will be useful for further exploring this and other aspects of the biology of this important human pathogen.z]1c, 百拇医药
ACKNOWLEDGMENTSz]1c, 百拇医药
We thank Peter Wright and Sharron Tollefson for providing the V94 and V98 strains of HPIV2.
REFERENCES9+.hqe, 百拇医药
Bilsel, P. A., and S. T. Nichol. 1990. Polymerase errors accumulating during natural evolution of the glycoprotein gene of vesicular stomatitis virus Indiana serotype isolates. J. Virol. 64:4873-4883.9+.hqe, 百拇医药
Calain, P., and L. Roux. 1993. The rule of six, a basic feature for efficient replication of Sendai virus defective interfering RNA. J. Virol. 67:4822-4830.9+.hqe, 百拇医药
Chanock, R. M. 1956. Association of a new type of cytopathogenic myxovirus with infantile croup. J. Exp. Med. 104:555-575.9+.hqe, 百拇医药
Chanock, R. M., B. R. Murphy, and P. L. Collins. 2001. Parainfluenza viruses, p. 1341-1379. In D. M. Knipe, P. M. Howley, D. E. Griffin, R. A. Lamb, M. A. Martin, B. Roizman, and S. E. Strauss (ed.), Fields virology, 4th ed., vol. 1. Lippincott Williams & Wilkins, Philadelphia, Pa.9+.hqe, 百拇医药
Cunningham, P. R., C. J. Weitzmann, and J. Ofengand. 1991. SP6 RNA polymerase stutters when initiating from an AAA... sequence. Nucleic Acids Res. 19:4669-4673.
Durbin, A. P., S. L. Hall, J. W. Siew, S. S. Whitehead, P. L. Collins, and B. R. Murphy. 1997. Recovery of infectious human parainfluenza virus type 3 from cDNA. Virology 235:323-332.*/wow6, 百拇医药
Durbin, A. P., J. W. Siew, B. R. Murphy, and P. L. Collins. 1997. Minimum protein requirements for transcription and RNA replication of a minigenome of human parainfluenza virus type 3 and evaluation of the rule of six. Virology 234:74-83.*/wow6, 百拇医药
Elroy-Stein, O., T. R. Fuerst, and B. Moss. 1989. Cap-independent translation of mRNA conferred by encephalomyocarditis virus 5' sequence improves the performance of the vaccinia virus/bacteriophage T7 hybrid expression system. Proc. Natl. Acad. Sci. USA 86:6126-6130.*/wow6, 百拇医药
Galinski, M. S. 1991. Annotated nucleotide and protein sequences for selected Paramyxoviridae, p. 537-568. In D. W. Kingsbury (ed.), The paramyxoviruses. Plenum Press, New York, N.Y.*/wow6, 百拇医药
Garcia-Barreno, B., T. Delgado, and J. A. Melero. 1994. Oligo(A) sequences of human respiratory syncytial virus G protein gene: assessment of their genetic stability in frameshift mutants. J. Virol. 68:5460-5468.
Garcia-Barreno, B., A. Portela, T. Delgado, J. A. Lopez, and J. A. Melero. 1990. Frame shift mutations as a novel mechanism for the generation of neutralization resistant mutants of human respiratory syncytial virus. EMBO J. 9:4181-4187.(+@6%, http://www.100md.com
Hall, S. L., A. Stokes, E. L. Tierney, W. T. London, R. B. Belshe, F. C. Newman, and B. R. Murphy. 1992. Cold-passaged human parainfluenza type 3 viruses contain ts and non-ts mutations leading to attenuation in rhesus monkeys. Virus Res. 22:173-184.(+@6%, http://www.100md.com
Hausmann, S., J. P. Jacques, and D. Kolakofsky. 1996. Paramyxovirus RNA editing and the requirement for hexamer genome length. RNA 2:1033-1045.(+@6%, http://www.100md.com
He, B., and R. A. Lamb. 1999. Effect of inserting paramyxovirus simian virus 5 gene junctions at the HN/L gene junction: analysis of accumulation of mRNAs transcribed from rescued viable viruses. J. Virol. 73:6228-6234.(+@6%, http://www.100md.com
Jacques, J. P., and D. Kolakofsky. 1991. Pseudo-templated transcription in prokaryotic and eukaryotic organisms. Genes Dev. 5:707-713.
Kawano, M., H. Bando, S. Ohgimoto, K. Kondo, M. Tsurudome, M. Nishio, and Y. Ito. 1990. Sequence of the fusion protein gene of human parainfluenza type 2 virus and its 3' intergenic region: lack of small hydrophobic (SH) gene. Virology 178:289-292.$2:^{8/, 百拇医药
Kawano, M., H. Bando, S. Ohgimoto, K. Okamoto, K. Kondo, M. Tsurudome, M. Nishio, and Y. Ito. 1990. Complete nucleotide sequence of the matrix gene of human parainfluenza type 2 virus and expression of the M protein in bacteria. Virology 179:857-861.$2:^{8/, 百拇医药
Kawano, M., H. Bando, T. Yuasa, K. Kondo, M. Tsurudome, H. Komada, M. Nishio, and Y. Ito. 1990. Sequence determination of the hemagglutinin-neuraminidase (HN) gene of human parainfluenza type 2 virus and the construction of a phylogenetic tree for HN proteins of all the paramyxoviruses that are infectious to humans. Virology 174:308-313.$2:^{8/, 百拇医药
Kawano, M., M. Kaito, Y. Kozuka, H. Komada, N. Noda, K. Nanba, M. Tsurudome, M. Ito, M. Nishio, and Y. Ito. 2001. Recovery of infectious human parainfluenza type 2 virus from cDNA clones and properties of the defective virus without V-specific cysteine-rich domain. Virology 284:99-112.
Kawano, M., K. Okamoto, H. Bando, K. Kondo, M. Tsurudome, H. Komada, M. Nishio, and Y. Ito. 1991. Characterizations of the human parainfluenza type 2 virus gene encoding the L protein and the intergenic sequences. Nucleic Acids Res. 19:2739-2746./fu, 百拇医药
Kolakofsky, D., T. Pelet, D. Garcin, S. Hausmann, J. Curran, and L. Roux. 1998. Paramyxovirus RNA synthesis and the requirement for hexamer genome length: the rule of six revisited. J. Virol. 72:891-899./fu, 百拇医药
Kroutil, L. C., K. Register, K. Bebenek, and T. A. Kunkel. 1996. Exonucleolytic proofreading during replication of repetitive DNA. Biochemistry. 35:1046-1053./fu, 百拇医药
Kunkel, T. A., S. S. Patel, and K. A. Johnson. 1994. Error-prone replication of repeated DNA sequences by T7 DNA polymerase in the absence of its processivity subunit. Proc. Natl. Acad. Sci. USA 91:6830-6834./fu, 百拇医药
Lamb, R. A., and D. Kolakofsky. 2001. Paramyxoviridae: the viruses and their replication., p. 1305-1340. In D. M. Knipe, P. M. Howley, D. E. Griffin, R. A. Lamb, M. A. Martin, B. Roizman, and S. E. Strauss (ed.), Fields virology, vol. 1. Lippincott Williams & Wilkins, Philadelphia, Pa.
Moeller, K., I. Duffy, P. Duprex, B. Rima, R. Beschorner, S. Fauser, R. Meyermann, S. Niewiesk, V. ter Meulen, and J. Schneider-Schaulies. 2001. Recombinant measles viruses expressing altered hemagglutinin (H) genes: functional separation of mutations determining H antibody escape from neurovirulence. J. Virol. 75:7612-7620.tx3/-$, http://www.100md.com
Murphy, S. K., and G. D. Parks. 1997. Genome nucleotide lengths that are divisible by six are not essential but enhance replication of defective interfering RNAs of the paramyxovirus simian virus 5. Virology 232:145-157.tx3/-$, http://www.100md.com
Newman, J. T., S. R. Surman, J. M. Riggs, C. T. Hansen, P. L. Collins, B. R. Murphy, and M. H. Skiadopoulos. 2002. Sequence analysis of the Washington/1964 strain of human parainfluenza virus type 1 (HPIV1) and recovery and characterization of wild type recombinant HPIV1 produced by reverse genetics. Virus Genes 24:77-92.tx3/-$, http://www.100md.com
Ohgimoto, S., H. Bando, M. Kawano, K. Okamoto, K. Kondo, M. Tsurudome, M. Nishio, and Y. Ito. 1990. Sequence analysis of P gene of human parainfluenza type 2 virus: P and cysteine-rich proteins are translated by two mRNAs that differ by two nontemplated G residues. Virology 177:116-123.
Peeters, B. P., O. S. de Leeuw, I. Verstegen, G. Koch, and A. L. Gielkens. 2001. Generation of a recombinant chimeric Newcastle disease virus vaccine that allows serological differentiation between vaccinated and infected animals. Vaccine 19:1616-1627.|ho, http://www.100md.com
Pelet, T., C. Delenda, O. Gubbay, D. Garcin, and D. Kolakofsky. 1996. Partial characterization of a Sendai virus replication promoter and the rule of six. Virology 224:405-414.|ho, http://www.100md.com
Poch, O., B. M. Blumberg, L. Bougueleret, and N. Tordo. 1990. Sequence comparison of five polymerases (L proteins) of unsegmented negative-strand RNA viruses: theoretical assignment of functional domains. J. Gen. Virol. 71:1153-1162.|ho, http://www.100md.com
Rager, M., S. Vongpunsawad, W. P. Duprex, and R. Cattaneo. 2002. Polyploid measles virus with hexameric genome length. EMBO J. 21:2364-2372.|ho, http://www.100md.com
Schmidt, A. C., J. M. McAuliffe, A. Huang, S. R. Surman, J. E. Bailly, W. R. Elkins, P. L. Collins, B. R. Murphy, and M. H. Skiadopoulos. 2000. Bovine parainfluenza virus type 3 (BPIV3) fusion and hemagglutinin-neuraminidase glycoproteins make an important contribution to the restricted replication of BPIV3 in primates. J. Virol. 74:8922-8929.
Shimokata, K., Y. Ito, Y. Nishiyama, and Y. Kimura. 1980. Influence of trypsin on the infectivity and biological properties or parainfluenza type 2 (croup-associated) virus in Vero cells. J. Gen. Virol. 48:407-410.?jkv&, 百拇医药
Sidhu, M. S., J. Chan, K. Kaelin, P. Spielhofer, F. Radecke, H. Schneider, M. Masurekar, P. C. Dowling, M. A. Billeter, and S. A. Udem. 1995. Rescue of synthetic measles virus minireplicons: measles genomic termini direct efficient expression and propagation of a reporter gene. Virology 208:800-807.?jkv&, 百拇医药
Skiadopoulos, M. H., A. P. Durbin, J. M. Tatem, S. L. Wu, M. Paschalis, T. Tao, P. L. Collins, and B. R. Murphy. 1998. Three amino acid substitutions in the L protein of the human parainfluenza virus type 3 cp45 live attenuated vaccine candidate contribute to its temperature-sensitive and attenuation phenotypes. J. Virol. 72:1762-1768.?jkv&, 百拇医药
Skiadopoulos, M. H., S. R. Surman, A. P. Durbin, P. L. Collins, and B. R. Murphy. 2000. Long nucleotide insertions between the HN and L protein coding regions of human parainfluenza virus type 3 yield viruses with temperature-sensitive and attenuation phenotypes. Virology 272:225-234.
Skiadopoulos, M. H., S. R. Surman, J. M. Riggs, C. Orvell, P. L. Collins, and B. R. Murphy. 2002. Evaluation of the replication and immunogenicity of recombinant human parainfluenza virus type 3 vectors expressing up to three foreign glycoproteins. Virology 297:136-152.\ezt, http://www.100md.com
Skiadopoulos, M. H., S. R. Surman, M. St Claire, W. R. Elkins, P. L. Collins, and B. R. Murphy. 1999. Attenuation of the recombinant human parainfluenza virus type 3 cp45 candidate vaccine virus is augmented by importation of the respiratory syncytial virus cpts530 L polymerase mutation. Virology 260:125-135.\ezt, http://www.100md.com
Stec, D. S., M. G. Hill, and P. L. Collins. 1991. Sequence analysis of the polymerase L gene of human respiratory syncytial virus and predicted phylogeny of nonsegmented negative-strand viruses. Virology 183:273-287.\ezt, http://www.100md.com
Stokes, A., E. L. Tierney, B. R. Murphy, and S. L. Hall. 1992. The complete nucleotide sequence of the JS strain of human parainfluenza virus type 3: comparison with the Wash/47885/57 prototype strain. Virus Res. 25:91-103. (Erratum, 27:96, 1993).
Streisinger, G., Y. Okada, J. Emrich, J. Newton, A. Tsugita, E. Terzaghi, and M. Inouye. 1966. Frameshift mutations and the genetic code. Cold Spring Harb. Symp. Quant. Biol. 31:77-84.4, http://www.100md.com
Tao, T., F. Davoodi, C. J. Cho, M. H. Skiadopoulos, A. P. Durbin, P. L. Collins, and B. R. Murphy. 2000. A live attenuated recombinant chimeric parainfluenza virus (PIV) candidate vaccine containing the hemagglutinin-neuraminidase and fusion glycoproteins of PIV1 and the remaining proteins from PIV3 induces resistance to PIV1 even in animals immune to PIV3. Vaccine 18:1359-1366.4, http://www.100md.com
Vulliemoz, D., and L. Roux. 2001. "Rule of six": how does the Sendai virus RNA polymerase keep count? J. Virol. 75:4506-4518.4, http://www.100md.com
Zheng, H., H. A. Lee, P. Palese, and A. Garcia-Sastre. 1999. Influenza A virus RNA polymerase has the ability to stutter at the polyadenylation site of a viral RNA template during RNA replication. J. Virol. 73:5240-5243.(Mario H. Skiadopoulos Leatrice Vogel Jeffrey M. Riggs Sonja R. Surman Peter L. Collins and Brian R. )