Deletion of M2 Gene Open Reading Frames 1 and 2 of
http://www.100md.com
病菌学杂志 2005年第11期
Laboratory of Infectious Diseases, National Institute of Allergy and Infectious Diseases, Bethesda, Maryland 20892-8007
ABSTRACT
The M2 gene of human metapneumovirus (HMPV) contains two overlapping open reading frames (ORFs), M2-1 and M2-2. The expression of separate M2-1 and M2-2 proteins from these ORFs was confirmed, and recombinant HMPVs were recovered in which expression of M2-1 and M2-2 was ablated individually or together [rM2-1, rM2-2, and rM2(1+2)]. Each M2 mutant virus directed efficient multicycle growth in Vero cells. The ability to recover HMPV lacking M2-1 contrasts with human respiratory syncytial virus, for which M2-1 is an essential transcription factor. Expression of the downstream HMPV M2-2 ORF was not reduced when translation of the upstream M2-1 ORF was silenced, indicating that it is initiated separately. The rM2-2 mutants exhibited a two- to fivefold increase in the accumulation of mRNA, normalized to the genome template, suggesting that M2-2 has a role in regulating RNA synthesis. Replication and immunogenicity were tested in hamsters. Animals infected intranasally with rM2-1 or rM2(1+2) did not have recoverable virus in the lungs or nasal turbinates on days 3 or 5 postinfection and did not develop HMPV-neutralizing serum antibodies or resistance to HMPV challenge. Thus, M2-1 appears to be essential for significant virus replication in vivo. In animals infected with rM2-2, virus was recovered from only 1 of 12 animals and only in the nasal turbinates on a single day. However, all of the animals developed a high titer of HMPV-neutralizing serum antibodies and were highly protected against challenge with wild-type HMPV. The HMPV rM2-2 virus is a promising and highly attenuated HMPV vaccine candidate.
INTRODUCTION
Human metapneumovirus (HMPV) was first identified in 2001 in The Netherlands from infants and children with acute respiratory tract disease (38) and is now recognized to be worldwide in prevalence (22, 41). HMPV resembles human respiratory syncytial virus (HRSV) with regard to disease signs and the ability to infect and cause disease in infants as well as in individuals of all ages (7, 18, 20, 29, 32, 39, 41). The contribution of HMPV to respiratory tract disease remains to be fully defined but appears to be sufficient to warrant the development of a vaccine, especially for the pediatric population. Reverse genetic systems were recently developed for HMPV, allowing the generation of infectious virus from cDNA and providing an important tool for characterizing HMPV biology and for designing live-attenuated HMPV vaccines (5, 25).
HMPV has a negative-strand RNA genome of approximately 13 kb (4, 37). It has been classified, together with avian metapneumovirus, in the Metapneumovirus genus, Pneumovirus subfamily, Paramyxovirus family, of the order Mononegavirales. The Pneumovirus subfamily also contains the genus Pneumovirus, represented by HRSV. The Metapneumovirus gene order is N-P-M-F-M2-SH-G-L. By analogy to HRSV, the predicted HMPV proteins are the following: the nucleocapsid protein N, which encapsidates the RNA genome and, together with the phosphoprotein P and the RNA polymerase protein L, forms the ribonucleoprotein complex; the fusion glycoprotein F, the small hydrophobic protein SH, and the major attachment glycoprotein G that are the transmembrane surface glycoproteins; the matrix M protein; and the M2-1 and M2-2 proteins encoded by two overlapping open reading frames (ORFs) in the M2 mRNA. Among the HMPV proteins, only F, G, and SH have been identified and characterized by direct biochemical means (6). For the other HMPV proteins, there is no direct information available beyond assumptions based on deduced sequence relatedness with other pneumoviruses.
The M2 gene with its two overlapping ORFs is present in all known members of the Pneumovirus subfamily and is unique to this subfamily (4, 11, 37). The HMPV M2-1 ORF (strain CAN97-83) initiates with the first AUG at nucleotide position 14 in the predicted mRNA and would encode a protein of 187 amino acids (4, 37). The HMPV M2-2 ORF has the potential to initiate at two closely spaced AUGs at positions 525 and 537 in the mRNA, overlaps the M2-1 ORF by 53 or 41 nucleotides, respectively, and would encode a protein of up to 71 amino acids. In comparison, the M2-1 and M2-2 proteins of HRSV are 194 and 90 amino acids in length, respectively (11, 37). All of the M2-1 proteins of the Pneumovirus subfamily including HMPV contain a conserved Cys/His zinc finger-like motif (4, 11, 37).
Functions of the pneumovirus M2-1 and M2-2 proteins have been identified only in the case of HRSV (3, 12-14, 19, 24, 27). Studies with minigenome systems showed that the HRSV M2-1 protein is a transcription elongation factor that is necessary for full processivity of the viral transcriptase; in the absence of M2-1, the transcriptase terminates prematurely within several hundred nucleotides to yield a heterodisperse smear of early quitters (13). HRSV M2-1 also has an antitermination function that enhances the synthesis of readthrough mRNAs, an activity that might be important in determining the amount of polymerase delivered to downstream genes (19, 24). These activities of M2-1 are sensitive to mutations in the Cys/His motif (23). The HRSV M2-1 protein also has been shown to be an RNA-binding protein and binds to the N protein (9, 15). Expression of the HRSV M2-1 protein, either from an added support plasmid or by promiscuous expression from the antigenome cDNA, appears to be essential for the recovery of recombinant HRSV from transfected cDNA, and it has not been possible to recover infectious HRSV in which the integrity of the M2-1 protein has been drastically disturbed by introduced mutations or by extensive deletion (10, 12, 28, 34).
Deletion of the HRSV M2-2 protein resulted in a virus that replicated more slowly than wild-type HRSV in cell culture and directed the delayed and reduced production of genome and antigenome RNAs and an exaggerated production of mRNA and viral proteins (3, 27). This suggests that HRSV M2-2 functions as a regulatory factor that helps switch the viral RNA synthetic program from transcription to RNA replication, a switch that seems to operate beginning midway in the infection cycle and might be controlled by an increasing accumulation of M2-2 protein (3). In an HRSV minireplicon system, coexpression of the M2-2 protein had a strong inhibitory effect on RNA synthesis (13, 24), which could be consistent with a regulatory role but which has not yet provided insight into a mechanism of action. HRSV lacking M2-2 was highly attenuated, immunogenic, and protective in chimpanzees, and its phenotype of up-regulated antigen synthesis would be highly desirable for inclusion in a live-attenuated vaccine (3, 27, 36).
There are several other examples of mononegavirus proteins other than N, P, and L that have RNA regulatory functions. Perhaps the one most similar to the HRSV M2-1 protein is VP30 of the Marburg and Ebola viruses (31, 40). VP30 is a nucleocapsid-associated protein that interacts with a cis-acting element at the beginning of the first gene to strongly activate transcription. Thus, its mechanism of action appears to be distinct from that of HRSV M2-1 or M2-2, and its amino acid sequence is unrelated except for the common feature of a functionally important Cys/His zinc finger motif that is similar to the motif present in pneumovirus M2-1 proteins (30). Other examples are the C and V proteins of Sendai virus, which bear no apparent structural similarity to M2-1 or M2-2 apart from an unrelated Cys/His motif in the V protein. The C protein has the effect of specifically down-regulating the activity of the genome promoter and also inhibits transcription by binding the L protein (16, 21, 35). The V protein also down-regulates replication, perhaps by interacting with soluble NP complexes (26). Thus, there is diversity in the structures and functions of mononegavirus RNA regulatory factors.
As a first step to characterize the HMPV M2-1 and M2-2 proteins, their functions, and their potential as targets to make attenuated vaccines, we have generated and characterized a set of recombinant HMPVs in which the two ORFs were silenced individually or in combination. Unexpectedly, the M2-1 protein was not essential for virus recovery or growth in vitro, in sharp contrast to the situation for HRSV. However, expression of M2-1 was required for detectable replication in vivo. M2-2 also was dispensable for growth in vitro and had the effect of up-regulating transcription. HMPV lacking M2-2 was highly attenuated in vivo but was very immunogenic and protective.
MATERIALS AND METHODS
Cells and viruses. Vero cells (ATCC CCL-81) were grown in OptiProSFM (Invitrogen) supplemented with 4 mM L-glutamine. BSR T7/5 cells are baby hamster kidney 21 (BHK-21) cells that constitutively express T7 RNA polymerase (8). They were maintained in Glasgow minimal essential medium supplemented with glutamine and amino acids (Invitrogen) and 10% fetal bovine serum. The Canadian HMPV isolate CAN97-83 (33) and the recombinant HMPVs based on this strain were propagated in Vero or BSR T7/5 cells at 32°C in the absence of serum and the presence of 5 μg/ml of trypsin. For HMPV propagation on Vero cells, trypsin was replenished every second day as previously described (5).
Construction of pHMPVM2-1, pHMPVM2-2, and pHMPVM2(1+2) plasmids. The development of a reverse genetic system for HMPV based on strain HMPV CAN97-83 (GenBank accession number AY297749) was described earlier (5). The parental recombinant virus is identical to HMPV CAN97-83 except that it contains an NheI restriction site in the M-F intergenic region as a marker (Fig. 1A). Three modified forms of the M2 gene were prepared by PCR using the antigenome cDNA as template: (i) the fragment M2-1, in which the M2-1 ORF was silenced by introducing translational stop codons, was prepared by PCR with the forward primer M2-1 (5'-TTAGTTAATTAAAAATAAAATAAAATTTGGGACAAATCATAtagTCTCGCAAGGCTtagTGCAAtaaaGAAGTGCGGGGCAAtaaCAACAGAGGAAGTG-3'; the PacI restriction site is in boldface, the M2 gene start signal is underlined, and mutations to ablate expression of M2-1 ORF are shown in lowercase letters) and the reverse primer BsiWIr (5'-AAGGTACCgtAcGTGTTTTTACTAACTTAAGTAAGCCTTGAC-3'; the KpnI adapter is underlined, the BsiWI site is in boldface, and mutations to generate the BsiWI site are in lowercase letters); (ii) the fragment M2-2, in which the M2-2 ORF was silenced and partially deleted, was prepared by PCR using forward primer F4656 (HMPV nucleotides [nt] 4656 to 4676, upstream of a naturally occurring PacI site, HMPV nt 4691) and the reverse primer M2-2 (5'-AAGGTACCgtAcGTGTTTTTACTAACTTAAGCTCACTGCACTTGATTtATGCTTTCACTGTCTTGCAGGGCgTATGAAGAGTCgTTGTCAGCGCCTGC-3'; the KpnI adapter is underlined, the BsiWI site is in boldface, mutations to generate the BsiWI site are in boldface lowercase letters, and mutations to ablate translation of the M2-2 protein are in italic lowercase letters); and (iii) the fragment M2(1+2), in which the M2 gene was deleted altogether, was prepared by PCR using the positive sense primer SHM2 (5'-AAGGTACCTTAATTAAAAACACgTacGAGTGGGATAAGTG-3'; the KpnI adapter is in boldface and underlined, the PacI site is in italic boldface, the BsiWI site is in roman boldface, mutations to generate the BsiWI site are in boldface lowercase letters, and the partial SH gene start signal is underlined) and the reverse primer Kpn-rev (5'-TGTTGGTACCTACATGTTTTACTTTAGAGC-3'; HMPV nt 6968 to 6939 with the KpnI site underlined). These three PCR fragments, M2-1, M2-2, and M2(1+2), were digested with PacI and KpnI and cloned into a pBluescript vector with a multiple cloning site modified to contain a PacI and a KpnI site, and their sequences were verified. Separately, the HMPV antigenome cDNA was modified to insert a BsiWI restriction site at nt 5459 in the M2/SH intergenic region. The mutagenesis was done using the Stratagene MutaGene kit, with two complementary HMPV primers (5'-CTTAAGTTAGTAAAAACACGTACGAGTGGGATAAGTGACAATG-3' and 5'-CATTGTCACTTATCCCACTCGTACGTGTTTTTACTAACTTAAG-3', with the BsiWI restriction sites underlined and mutated nucleotides in boldface) and a template plasmid that contained the NheI/KpnI fragment of the HMPV antigenome cDNA containing the F, M2, SH, and G genes (HMPV nt 3038 to 6963) (Fig. 1A). After sequence confirmation, a restriction fragment was generated from the mutated plasmid containing the HMPV SH and G genes (nt 5460 to 6963), framed by the upstream BsiWI site and by the naturally occurring KpnI restriction site located in the HMPV G/L intergenic region (nt 6963). To assemble antigenome plasmids with the M2-1 and M2-2 mutations, three inserts were ligated simultaneously into the NheI-KpnI window of the HMPV antigenome plasmid: (i) the NheI/PacI restriction fragment containing the HMPV F gene (HMPV nt 3039 to 4691), (ii) the PacI/BsiWI fragment containing the M2-1 or M2-2 fragment described above, and (iii) the mutagenized BsiWI/KpnI fragment described above containing SH and G (HMPV nt 5460 to 6963) (Fig. 1A). To generate the antigenome plasmid containing the M2(1+2) mutant, two inserts were ligated simultaneously into the NheI-KpnI window of the HMPV antigenome plasmid: (i) the NheI/PacI fragment containing HMPV F and (ii) fragment M2, which contains SH and G, framed by PacI and KpnI. This placed HMPV F directly upstream of HMPV SH (Fig. 1A). Additionally, a second set of M2-1, M2-2, and M2(1+2) mutants was constructed by the same strategy except that they were based on an HMPV antigenome plasmid carrying a synthetic green fluorescent protein (GFP) gene in the first leader-proximal position (5) (Fig. 1A).
Recombinant HMPV (rHMPV) recovery. Confluent BSR T7/5 cells in six-well dishes were transfected with 5 μg of an individual antigenome plasmid, 2 μg each of the support plasmids pT7-N and pT7-P, and 1 μg of pT7-L per well. Transfections were done with Lipofectamine 2000 (Invitrogen) in OptiMEM without trypsin or serum and maintained overnight at 32°C. The transfection medium was removed 1 day later and replaced with Glasgow minimal essential medium without trypsin or serum. Trypsin was added on day 3 to a final concentration of 5 μg/ml, and cell-medium mixtures were passaged onto fresh Vero cells on day 6. Parallel transfections were performed with versions of each mutant with and without GFP, and recovery and passage were monitored by GFP expression.
Replication and protective efficacy of M2-1, M2-2, and M(1+2) viruses in hamsters. Groups of 6-week-old Golden Syrian hamsters (18 animals per group) were infected intranasally, under light isoflurane anesthesia, with 0.1 ml of L15 medium containing 5.7 log10 50% tissue culture infective dose (TCID50) of rHMPV, rM2-1, rM2-2, or rM(1+2). On days 3 and 5 postinfection, the lungs and nasal turbinates were harvested from six animals from each group, and the virus titers of the individual specimens were quantified by titration of tissue homogenates on Vero cells. The level of immunogenicity and the protective efficacy of the recombinants were determined in the remaining six hamsters per group. Sera were collected 2 days prior to infection and 27 days postinfection. The titers of HMPV-neutralizing antibodies were determined by an endpoint dilution neutralization assay as described previously (6). On day 28 postinfection, hamsters were challenged by intranasal administration of 5.7 log10 TCID50 of HMPV CAN97-83 in a 0.1-ml inoculum. Nasal turbinates and lungs were harvested 3 days later, and the virus titer in each tissue homogenate was determined as described above.
RESULTS
Recovery of recombinant HMPV lacking the M2-1 and/or M2-2 proteins. Reverse genetics was used to design HMPV mutants in which the M2-1 and M2-2 ORFs were silenced individually or in combination. Two versions of each mutant were made, one based on rHMPV and one based on a version (rgHMPV) that expresses the enhanced GFP protein from an additional promoter-proximal gene but otherwise was identical (Fig. 1A). The GFP versions were constructed to provide a means of directly monitoring virus recovery and passage, which was helpful because of the slow growth and trypsin dependence of HMPV. The non-GFP versions were made because the evaluation of vaccine candidates should be done with versions that can be advanced directly to clinical trials, and the expression of GFP would not be acceptable for human vaccine purposes.
To silence the M2-1 ORF, its translation initiation codon was changed to a stop codon, a second in-frame stop codon was introduced five codons downstream (whereas the next in-frame ATG is 129 codons further downstream), and additional stop codons were introduced in the alternate reading frames to preclude possible frame-shifting into the M2-1 ORF (Fig. 1B). These changes did not alter the length of the M2 gene and did not affect the downstream M2-2 ORF. To silence the M2-2 ORF, the two putative translational start codons of M2-2, which are located at HMPV positions 5235 and 5247 and overlap with the M2-1 ORF, were mutagenized from ATG to ACG, and a translational stop codon was introduced 12 codons downstream. These changes did not affect amino acid coding by the M2-1 ORF. The ATG-to-ACG changes were not ideal because ACG has the potential to serve as a translational start site, as exemplified by the C ORF of Sendai virus (17), but the efficiency would be reduced compared to ATG, and these were the only changes possible at those loci that did not alter amino acid coding in the M2-1 ORF. In addition, the portion of the M2-2 ORF that is downstream of the M2-1 ORF was deleted, removing 152 nucleotides (HMPV nt 5289 to 5440). To delete the complete M2 gene, a PacI restriction site was introduced into the gene end signal preceding the M2/SH intergenic region at nucleotide 5450, and the complete M2 gene including its gene start and gene end signal was excised by digestion with PacI, which cleaves at the above-mentioned site bordering the M2/SH intergenic region as well as at a naturally occurring PacI site located at the F gene end signal (Fig. 1A). Thus, the genome of rM2(1+2) would be 759 nucleotides shorter than that of wild-type HMPV and would completely lack the M2 gene and both of its ORFs.
To recover the recombinant viruses, the individual mutant antigenome plasmids were transfected, together with a set of three support plasmids encoding the HMPV N, P, and L proteins, into BHK-21 cells that constitutively expresses T7 RNA polymerase (8). Each of the recombinant viruses, namely rM2-1, rgM2-1, rM2-2, rgM2-2, rM2(1+2), and rgM2(1+2) (Fig. 1A), was successfully produced, with the GFP-expressing viruses serving to guide recovery. In a typical transfection experiment of three replicate wells containing 106 cells each, the efficiency of recovery of rgHMPV and rgM2-2 was exactly the same, with an average of 31 positive cells (standard error, ± 0.3) for each recombinant. Means of 7 ± 1.0 and 18 ± 1.8 cells were observed for rgM2-1 and rgM2(1+2), respectively, suggesting that expression of M2-1 provided an advantage. Recovery of for rgM2 was usually more efficient than recovery of rgM2-1, as shown in this example, possibly reflecting an advantage due to the shorter genome. The recovered viruses were amplified on Vero cells to yield stocks that ranged between 2 x 106 and 9 x 106 PFU per ml for the GFP-expressing recombinants and between 7 x 106 and 3 x 107 for the non-GFP-expressing recombinants, illustrating efficient growth with a slight growth restriction associated with the addition of the GFP gene as noted previously (5).
The presence of the designed mutations in the recovered viruses was evaluated by reverse transcription-PCR (RT-PCR) performed with an F-specific forward primer (nt 4656 to 4676) and an SH-specific reverse primer (nt 5508 to 5529). Synthesis of a detectable RT-PCR product was dependent on the presence of reverse transcriptase in the initial RT reaction (data not shown), indicating that the template was RNA rather than contaminating DNA. In the case of the M2-2 and M2(1+2) mutants, electrophoresis of the RT-PCR products on a 2% agarose gel confirmed an increase in electrophoretic mobility consistent with the designed deletions (Fig. 1A). Direct sequence analysis of the RT-PCR product for each of the recovered viruses confirmed that the sequence between the F and SH genes was exactly as designed (data not shown).
Protein synthesis by the M2 mutant viruses. To characterize the expression of proteins and RNAs by the mutant viruses, Vero cells were infected at a multiplicity of infection (MOI) of 1 PFU per cell with the non-GFP-expressing versions of the M2 mutant viruses and harvested 72 h later. Each sample was divided into two aliquots for analysis of proteins and RNAs. For protein analysis, lysates were prepared and subjected to Western blotting using a polyclonal serum raised against gradient-purified HMPV (5). Since most of the HMPV proteins had not been previously identified, positive controls were prepared by transfecting BHK-21 cells that express the T7 RNA polymerase with the support plasmids encoding the N, P, and M2-1 proteins. The HMPV-specific antiserum detected bands of approximately 42 (N protein), 40 (P protein), and 22 (M2-1 protein) kDa in the plasmid transfection samples (Fig. 2A, lanes 1 to 3), consistent with the calculated molecular sizes. These species also corresponded to major bands in the samples from cells infected with rHMPV (Fig. 2A, lane 4) and wild-type HMPV CAN97-83 (lane 8). The HMPV-infected cells also contained an additional major band that might correspond to the M protein. Note that this particular serum did not react with the HMPV F, G, and SH proteins in this highly denaturing Western blot analysis; as noted previously, this serum did react efficiently in Western blotting with F protein that was not reduced and which migrated as homo-oligomers (6). Samples from cells infected with rM2-1 and rM2(1+2) lacked the M2-1 protein band (Fig. 2A, lanes 5 and 7), confirming that its ORF had been silenced. Conversely, silencing the M2-2 ORF did not affect expression of the upstream M2-1 ORF (Fig. 2A, lane 6), as would be expected. The Western blot analysis also showed that each of the mutant viruses directed the accumulation of these major HMPV proteins at a level that was comparable to that of CAN97-83 or rHMPV, in the case of rM2-1 (Fig. 2A, lane 5), or somewhat in excess of CAN97-83 or rHMPV, in the case of rM2-2 and rM2(1+2) (lanes 6 and 7).
We were unable to detect expression of the M2-2 protein using antiserum raised against gradient-purified HMPV or antisera from rabbits immunized with synthetic peptides representing two different regions of the M2-2 protein (data not shown). As an alternative strategy, we created two additional recombinants, based on rgHMPV and rgM2-1, in which the encoded M2-2 protein was modified to have a C-terminal extension consisting of a 9-amino-acid epitope tag (YPYDVPDYA) derived from the human influenza virus hemagglutinin (HA) protein and preceded by a spacer of three alanine residues. Western blotting of lysates of cells infected with either recombinant, rgHMPV/M2-2-HA or rgM2-1/M2-2-HA, yielded an HA-specific band of approximately 8 kDa, consistent with the expected size of the M2-2-HA fusion protein (Fig. 2B, lanes 4 and 5). Thus, expression of the M2-2 ORF was not dependent on expression of the upstream M2-1 ORF. Indeed, silencing the M2-1 ORF appeared to have no effect on the efficiency of expression of M2-2.
RNA synthesis by the M2 mutant viruses. The second set of cell pellets from the experiment shown in Fig. 2, in which Vero cells had been infected with the panel of M2 mutants and harvested 72 h later, was used to isolate total intracellular RNA. These samples were subjected to Northern blot analysis using double-stranded DNA probes against the M2 mRNA (Fig. 3). Hybridization with the M2 probe confirmed the lack of expression of M2 mRNA by the rM2(1+2) mutant (Fig. 3, lane 4) and confirmed that the rM2-2 mutant produced a shorter M2 mRNA (lane 3), consistent with the deletion of 152 nt. Otherwise, the pattern of monocistronic and polycistronic readthrough mRNAs was very similar among the different viruses. In particular, the RNA pattern of the rM2-1 mutant was essentially indistinguishable from that of rHMPV, with no evidence of premature termination or changes in the frequency of transcriptional readthrough. These observations were noteworthy, given that the M2-1 protein of HRSV is necessary for processive transcription and increases the production of polycistronic readthrough mRNAs. This did not appear to be the case for HMPV.
To further evaluate the apparent increased transcription associated with deletion of the M2-2 gene, replicate cultures of Vero cells were infected with the M2 mutants, and total intracellular RNA was isolated at 12-, 24-, 36-, 48-, and 72-h time points and analyzed by Northern blot hybridization with strand-specific riboprobes for the F gene. Calculations were done based on three independent time course experiments using the GFP-expressing mutants and two experiments using rHMPV mutants; the Northern blots for one of the experiments using rgHMPV are shown in Fig. 4. The reason for using the rgHMPV virus was to monitor the uniformity of the infection and to gauge the appropriate harvest times. To compare the various viruses, the amount of mRNA for each mutant for the 24-, 36-, 48-, and 72-h time points quantified by phosphor imager was divided by the amount of genome for that mutant from a replicate gel lane for the same time point; the 12-h time point was not used because of the low levels of RNA. Then, the value for each M2 mutant virus for each time point was normalized relative to the corresponding value for rgHMPV or rHMPV (Table 1). This showed that the amount of F mRNA relative to the genome in cells infected with the rgM2-2 mutant was increased approximately three- to fivefold compared to the parental virus and two- to threefold for the non-GFP mutants, results that we consider to be essentially the same. Examination of the original Northern blots (Fig. 4) confirmed the high level of mRNA expression. In cells infected with the M2(1+2) mutant, F mRNA/genome ratios were comparable to those of cells infected with the parental virus. The mRNA/genome ratio for the M2-1 mutant was essentially the same as for the parent (Table 1), and direct examination of the original Northern blots (Fig. 4) confirmed that the intracellular level was similar to, or somewhat lower than, that of the parent virus. The finding that transcription was not elevated for the M2(1+2) mutants despite the deletion of M2-2 suggests that the combination of the M2-1 and M2-2 mutations might be more complicated than expected. Further studies using recombinant virus and minireplicons will be needed to better understand the effects of the M2-1 and M2-2 deletions and, in particular, the effect of the combined mutations.
In vitro growth of the M2 mutant viruses. To evaluate the kinetics of multicycle growth, replicate cultures of Vero cells were infected with the non-GFP-expressing and GFP-expressing versions of the M2 mutants at an MOI of 0.01 PFU/cell. At 24-h intervals, medium supernatant samples were taken, frozen, and analyzed later in parallel for infectious virus. As shown in Fig. 5A for the GFP-expressing versions, each of the M2 mutants was capable of efficient multicycle growth, showing that M2-1 and M2-2 are dispensable either singly or in combination. The rgM2-1 mutant replicated somewhat less efficiently than its rgHMPV parent during the first 7 days of incubation, although the final titers were indistinguishable. The rgM2-2 and rgM2(1+2) viruses replicated similarly to the rgHMPV parent for the first 7 days but yielded slightly higher titers at later time points. The shorter genome lengths of these two deletion mutants might have conferred a slight growth advantage. Thus, silencing the M2-1 ORF seemed to confer a very small growth restriction in vitro, whereas silencing M2-2 did not reduce the efficiency of growth.
Indirect immunofluorescence staining of permeabilized Vero cells infected with the non-GFP-expressing recombinants (Fig. 5 B) showed that the absence of M2-1 and/or M2-2 did not have a visible effect on the expression of viral antigen or the formation of viral filaments.
Replication, immunogenicity, and protective efficacy of the M2 mutants in hamsters. To assay the ability of the M2 mutants to replicate in vivo, hamsters in groups of 18 were infected intranasally with the non-GFP-expressing versions of the M2 mutants and rHMPV. Six animals from each group were sacrificed on days 3 and 5 following inoculation, and the lungs and nasal turbinates were removed and analyzed for infectious virus. rHMPV replicated to a mean titer of 5.7 and 5.4 log10 TCID50 per g of tissue in the upper respiratory tract on days 3 and 5, respectively, and to 4.0 and 3.1 log10 TCID50 in the lower respiratory tract on days 3 and 5, respectively (Table 2). In contrast, none of the 12 sacrificed animals infected with the rM2-1 or rM2(1+2) viruses yielded detectable infectious virus from either the nasal turbinates or the lungs on either day. Among the 12 sacrificed animals infected with the rM2-2 virus, a trace amount of infectious virus was recovered from the nasal turbinates of a single animal on a single day (day 3) but not from the lungs of this animal and not from any other animal on either day (Table 2).
For the remaining six animals in each group, serum samples were taken on days –2 (i.e., 2 days before infection) and 27 and assayed for the ability to neutralize HMPV CAN97-83 in vitro (Table 3). Animals that had been infected with rM2-1 or rM2(1+2) did not have detectable serum HMPV-neutralizing antibodies (Table 3), providing further evidence that these viruses did not replicate in vivo. In contrast, the rM2-2 virus induced a strong serum HMPV-neutralizing antibody response in each of the inoculated animals, with a mean titer that was only twofold less than that induced by rHMPV. This was remarkable given that only 1 of 12 animals had recoverable infectious virus. It seems unlikely that these antibodies were induced by the inoculum alone, given the lack of a detectable response to inoculation with the rM2-1 and rM2(1+2) viruses noted above. Thus, the uniformly strong antibody response to rM2-2, coupled with the ability to recover infectious virus from a single animal, suggests that a low level of replication occurred in each animal.
The animals were challenged intranasally with HMPV CAN97-83 on day 28 postinoculation and were sacrificed 3 days later, and the lungs and nasal turbinates were assayed for infectious virus. The group previously immunized with rHMPV was completely protected from challenge virus replication in both the upper and lower respiratory tract, whereas the mock-infected control group had mean challenge virus titers of 6.9 log10 PFU per g of tissue in the nasal turbinates and 5.3 log10 PFU per g of tissue in the lungs (Table 3). The mean titers of challenge virus in the animals previously immunized with rM2-1 and rM2(1+2) were essentially the same as that of the mock-immunized group, confirming the lack of a significant protective response from these highly debilitated mutants. In contrast, of the animals that had been immunized with rM2-2, all six animals were completely protected in the lungs, and four of the six animals were completely protected in the nose. The two positive animals yielded titers of challenge virus that were barely above the limit of detection. Thus, HMPV recombinants that lacked the M2-1 protein provided no evidence of replication or immunogenicity in vivo, whereas HMPV lacking M2-2 was highly restricted in vivo but nonetheless was highly immunogenic and protective and is a promising vaccine candidate.
DISCUSSION
The M2 gene with its two overlapping ORFs is present in all known members of the Pneumovirus and Metapneumovirus genera (4, 11, 37), but functions for the encoded proteins had previously been identified only for HRSV. In the present study, the M2-1 and M2-2 ORFs of recombinant HMPV were silenced individually and in combination. The two ORFs were completely dispensable for efficient replication in vitro, but their deletion was highly attenuating in vivo.
The HMPV M2-1 and M2-2 ORFs were each confirmed to express a separate protein product during HMPV infection in cell culture. The M2-1 protein was readily detected in infected-cell lysates by Western blot analysis using antibodies raised against gradient-purified HMPV. Identification was confirmed by comparison with M2-1 protein expressed from a transfected cDNA. Although we did not specifically address the issue of incorporation into virus particles in the present study, an abundant band of the appropriate size was detected previously in gradient-purified virus (6). Thus, HMPV M2-1 appears to be a virion protein, like its HRSV counterpart.
The HMPV M2-2 protein was identified by adding an epitope tag to M2-2 encoded by recombinant HMPV. This confirmed expression of the M2-2 ORF, in contrast to results with avian metapneumovirus for which it was concluded that the M2-2 ORF was not expressed (1). Unexpectedly, the M2-2 protein was expressed even when the upstream M2-1 ORF was silenced; indeed, its level of expression appeared to be completely unaffected when M2-1 was silenced. This is in contrast to the situation with HRSV, where expression of the M2-2 ORF appeared to depend on translational stop-restart by ribosomes exiting the upstream M2-1 ORF (2). However, even in the case of HRSV, details such as identification of the translational start site of M2-2 remain to be described. The mechanism by which ribosomes gain access to the HMPV M2-2 ORF remains to be identified. Typically, eukaryotic ribosomes enter an mRNA at the 5' end and scan to find appropriate start sites. It is difficult to imagine that ribosome scanning could account for the expression of the M2-2 ORF, given its location deep within the mRNA. However, this cannot be ruled out. Alternatively, scanning-independent initiation of internal start sites has been reported, as exemplified by the mechanism of ribosomal shunt described for the Y1, Y2, and X proteins of Sendai virus (17).
The finding that the HMPV M2-1 ORF is dispensable for virus recovery and replication in vitro contrasts sharply with previous results with HRSV, where it has not been possible to recover virus in which the integrity of the M2 protein was disrupted by mutation or deletion. In the case of the rM2-1 virus, the possibility of a low level of expression of a fragment of the M2-1 ORF could not be absolutely excluded because most of the ORF remained intact, but in the case of the rM2(1+2) mutant, the entire gene was deleted, excluding any possibility of residual expression. However, silencing of HMPV M2-1 was not without any effect in vitro: there was a clear and consistent reduction in the efficiency of recovery as assayed by the number of green cells produced following transfection with the GFP-expressing mutants. Also, multicycle growth for the M2-1 mutant was modestly but clearly less efficient for approximately the first 7 days compared to its rgHMPV parent, although the final titers were not different. This modest difference might reflect effects at any stage of the infectious cycle, including entry, gene expression, RNA replication, and virion production, although effects on entry seem less likely, given the presumption that M2-1 is an internal protein. The M2-1 protein of HRSV, the only one studied in any detail to date, was shown to be a transcription processivity factor, without which the polymerase can transcribe for up to several hundred nucleotides before terminating prematurely (13). HRSV M2-1 also has an antitermination activity that acts at the gene end signals to promote transcriptional readthrough (19, 24). In the case of HMPV, the loss of expression of M2-1 did not appear to affect the production of discrete, full-length mRNAs, nor did it modify the frequency or occurrence of readthrough mRNAs. Thus, the function of the HMPV M2-1 protein might be quite different from that of HRSV, even though the two proteins share the Cys/His motif and are among the most highly conserved counterparts between HMPV and HRSV. Alternatively, our understanding of the functions of HRSV M2-1 may be incomplete, and functions common to M2-1 of both viruses might remain to be identified.
In vivo, in the hamster model, HMPV lacking the M2-1 protein [rM2-1 and rM2(1+2)] was highly debilitated for replication and may not have replicated at all. Infectious virus was not recovered on days 3 and 5 postinoculation from the lungs or nasal turbinates of inoculated animals, and inoculated animals did not develop detectable HMPV-neutralizing antibodies or protection against challenge. In any case, the lack of a detectable immune response clearly indicates that the inoculum alone was not significantly immunogenic and suggests that little antigen production occurred. The lack of a detectable immune response following the administration of an inoculum containing nearly 6.0 log10 PFU of HMPV—and probably containing a much larger number of non-plaque-forming particles, given the typical high particle-to-PFU ratio of paramyxoviruses—illustrates the need to have at least some virus replication in order to stimulate a significant immune response and provides an indication that an alternative strategy involving nonreplicating virus-like particles would require a very high dose.
The ability to delete HMPV M2-2 without compromising growth in vitro is reminiscent of results with HRSV (3, 13, 27, 36). For HRSV, deletion of M2-2 resulted in a decrease in RNA replication and an increase in transcription, whereas for HMPV only the second effect was evident. The rM2-2 virus also directed an increased level of accumulation of intracellular HMPV proteins as detected by Western blotting. An increase in the synthesis of viral antigens might increase the immunogenicity of the virus, which would be a desirable property for a vaccine. However, it is clear that we have more to learn about the role of M2-2 (and M2-1), since the up-regulation of transcription associated with M2-2 was not observed when this deletion was combined with the deletion of M2-1, at least not for the F gene.
The restriction on replication in vivo associated with silencing M2-2 appeared to be greater in the case of HMPV than for HRSV. Whether this greater attenuating effect is related to the slower growth of HMPV is unclear. However, even though replication of the HMPV rM2-2 virus could not be directly detected in most of the animals, each animal developed a high titer of virus-neutralizing antibodies and was highly protected against challenge with wild-type HMPV. This suggests that a low level of replication did occur. The high degree of immunogenicity observed in the near absence of detectable infectious virus might be an indication that rM2-2 is indeed more immunogenic due to its up-regulated transcription. Thus, the HMPV rM2-2 mutant is a highly attenuated, highly immunogenic, and protective vaccine candidate that warrants further evaluation in nonhuman primates and in the clinic.
ACKNOWLEDGMENTS
We thank Guy Boivin of the Centre Hospitalier Universitaire de Québec and Laval University and Dean Erdman, Teresa Peret, and Larry Anderson of the Centers for Disease Control and Prevention for providing initial stocks of CAN97-83 and Chris Hanson for sequencing.
REFERENCES
Ahmadian, G., P. Chambers, and A. J. Easton. 1999. Detection and characterization of proteins encoded by the second ORF of the M2 gene of pneumoviruses. J. Gen. Virol. 80:2011-2016.
Ahmadian, G., J. S. Randhawa, and A. J. Easton. 2000. Expression of the ORF-2 protein of the human respiratory syncytial virus M2 gene is initiated by a ribosomal termination-dependent reinitiation mechanism. EMBO J. 19:2681-2689.
Bermingham, A., and P. L. Collins. 1999. The M2-2 protein of human respiratory syncytial virus is a regulatory factor involved in the balance between RNA replication and transcription. Proc. Natl. Acad. Sci. USA 96:11259-11264.
Biacchesi, S., M. H. Skiadopoulos, G. Boivin, C. T. Hanson, B. R. Murphy, P. L. Collins, and U. J. Buchholz. 2003. Genetic diversity between human metapneumovirus subgroups. Virology 315:1-9.
Biacchesi, S., M. H. Skiadopoulos, K. C. Tran, B. R. Murphy, P. L. Collins, and U. J. Buchholz. 2004. Recovery of human metapneumovirus from cDNA: optimization of growth in vitro and expression of additional genes. Virology 321:247-259.
Biacchesi, S., M. H. Skiadopoulos, L. Yang, E. W. Lamirande, K. C. Tran, B. R. Murphy, P. L. Collins, and U. J. Buchholz. 2004. Recombinant human metapneumovirus lacking the small hydrophobic SH and/or attachment G glycoprotein: deletion of G yields a promising vaccine candidate. J. Virol. 78:12877-12887.
Boivin, G., Y. Abed, G. Pelletier, L. Ruel, D. Moisan, S. Cote, T. C. Peret, D. D. Erdman, and L. J. Anderson. 2002. Virological features and clinical manifestations associated with human metapneumovirus: a new paramyxovirus responsible for acute respiratory-tract infections in all age groups. J. Infect. Dis. 186:1330-1334.
Buchholz, U. J., S. Finke, and K. K. Conzelmann. 1999. Generation of bovine respiratory syncytial virus (BRSV) from cDNA: BRSV NS2 is not essential for virus replication in tissue culture, and the human RSV leader region acts as a functional BRSV genome promoter. J. Virol. 73:251-259.
Cartee, T. L., and G. W. Wertz. 2001. Respiratory syncytial virus M2-1 protein requires phosphorylation for efficient function and binds viral RNA during infection. J. Virol. 75:12188-12197.
Collins, P. L., E. Camargo, and M. G. Hill. 1999. Support plasmids and support proteins required for recovery of recombinant respiratory syncytial virus. Virology 259:251-255.
Collins, P. L., R. M. Chanock, and B. R. Murphy. 2001. Respiratory syncytial virus, p. 1443-1485. In D. M. Knipe, P. M. Howley, D. E. Griffin, R. A. Lamb, M. A. Martin, B. Roizman, and S. E. Straus (ed.), Fields virology, 4th ed. Lippincott Williams & Wilkins, Philadelphia, Pa.
Collins, P. L., M. G. Hill, E. Camargo, H. Grosfeld, R. M. Chanock, and B. R. Murphy. 1995. Production of infectious human respiratory syncytial virus from cloned cDNA confirms an essential role for the transcription elongation factor from the 5' proximal open reading frame of the M2 mRNA in gene expression and provides a capability for vaccine development. Proc. Natl. Acad. Sci. USA 92:11563-11567.
Collins, P. L., M. G. Hill, J. Cristina, and H. Grosfeld. 1996. Transcription elongation factor of respiratory syncytial virus, a nonsegmented negative-strand RNA virus. Proc. Natl. Acad. Sci. USA 93:81-85.
Collins, P. L., M. G. Hill, and P. R. Johnson. 1990. The two open reading frames of the 22K mRNA of human respiratory syncytial virus: sequence comparison of antigenic subgroups A and B and expression in vitro. J. Gen. Virol. 71:3015-3020.
Cuesta, I., X. Geng, A. Asenjo, and N. Villanueva. 2000. Structural phosphoprotein M2-1 of the human respiratory syncytial virus is an RNA binding protein. J. Virol. 74:9858-9867.
Curran, J., R. Boeck, and D. Kolakofsky. 1991. The Sendai virus P gene expresses both an essential protein and an inhibitor of RNA synthesis by shuffling modules via mRNA editing. EMBO J. 10:3079-3085.
Curran, J., P. Latorre, and D. Kolakofsky. 1998. Translational gymnastics on the Sendai virus P/C mRNA. Semin. Virol. 8:351-357.
Falsey, A. R., D. Erdman, L. J. Anderson, and E. E. Walsh. 2003. Human metapneumovirus infections in young and elderly adults. J Infect. Dis. 187:785-790.
Fearns, R., and P. L. Collins. 1999. Role of the M2-1 transcription antitermination protein of respiratory syncytial virus in sequential transcription. J. Virol. 73:5852-5864.
Greensill, J., P. S. McNamara, W. Dove, B. Flanagan, R. L. Smyth, and C. A. Hart. 2003. Human metapneumovirus in severe respiratory syncytial virus bronchiolitis. Emerg. Infect. Dis. 9:372-375.
Grogan, C. C., and S. A. Moyer. 2001. Sendai virus wild-type and mutant C proteins show a direct correlation between L polymerase binding and inhibition of viral RNA synthesis. Virology 288:96-108.
Hamelin, M. E., Y. Abed, and G. Boivin. 2004. Human metapneumovirus: a new player among respiratory viruses. Clin. Infect. Dis. 38:983-990.
Hardy, R. W., and G. W. Wertz. 2000. The Cys3-His1 motif of the respiratory syncytial virus M2-1 protein is essential for protein function. J. Virol. 74:5880-5885.
Hardy, R. W., and G. W. Wertz. 1998. The product of the respiratory syncytial virus M2 gene ORF1 enhances readthrough of intergenic junctions during viral transcription. J. Virol. 72:520-526.
Herfst, S., M. de Graaf, J. H. Schickli, R. S. Tang, J. Kaur, C. F. Yang, R. R. Spaete, A. A. Haller, B. G. van den Hoogen, A. D. Osterhaus, and R. A. Fouchier. 2004. Recovery of human metapneumovirus genetic lineages A and B from cloned cDNA. J. Virol. 78:8264-8270.
Horikami, S. M., S. Smallwood, and S. A. Moyer. 1996. The Sendai virus V protein interacts with the NP protein to regulate viral genome RNA replication. Virology 222:383-390.
Jin, H., X. Cheng, H. Z. Zhou, S. Li, and A. Seddiqui. 2000. Respiratory syncytial virus that lacks open reading frame 2 of the M2 gene (M2-2) has altered growth characteristics and is attenuated in rodents. J. Virol. 74:74-82.
Jin, H., D. Clarke, H. Z. Zhou, X. Cheng, K. Coelingh, M. Bryant, and S. Li. 1998. Recombinant human respiratory syncytial virus (RSV) from cDNA and construction of subgroup A and B chimeric RSV. Virology 251:206-214.
Maggi, F., M. Pifferi, M. Vatteroni, C. Fornai, E. Tempestini, S. Anzilotti, L. Lanini, E. Andreoli, V. Ragazzo, M. Pistello, S. Specter, and M. Bendinelli. 2003. Human metapneumovirus associated with respiratory tract infections in a 3-year study of nasal swabs from infants in Italy. J. Clin. Microbiol. 41:2987-2991.
Modrof, J., S. Becker, and E. Muhlberger. 2003. Ebola virus transcription activator VP30 is a zinc-binding protein. J. Virol. 77:3334-3338.
Muhlberger, E., M. Weik, V. E. Volchkov, H. D. Klenk, and S. Becker. 1999. Comparison of the transcription and replication strategies of Marburg virus and Ebola virus by using artificial replication systems. J. Virol. 73:2333-2342.
Osterhaus, A., and R. Fouchier. 2003. Human metapneumovirus in the community. Lancet 361:890-891.
Peret, T. C., G. Boivin, Y. Li, M. Couillard, C. Humphrey, A. D. Osterhaus, D. D. Erdman, and L. J. Anderson. 2002. Characterization of human metapneumoviruses isolated from patients in North America. J. Infect. Dis. 185:1660-1663.
Tang, R. S., N. Nguyen, X. Cheng, and H. Jin. 2001. Requirement of cysteines and length of the human respiratory syncytial virus M2-1 protein for protein function and virus viability. J. Virol. 75:11328-11335.
Tapparel, C., S. Hausmann, T. Pelet, J. Curran, D. Kolakofsky, and L. Roux. 1997. Inhibition of Sendai virus genome replication due to promoter-increased selectivity: a possible role for the accessory C proteins. J. Virol. 71:9588-9599.
Teng, M. N., S. S. Whitehead, A. Bermingham, M. St. Claire, W. R. Elkins, B. R. Murphy, and P. L. Collins. 2000. Recombinant respiratory syncytial virus that does not express the NS1 or M2-2 protein is highly attenuated and immunogenic in chimpanzees. J. Virol. 74:9317-9321.
van den Hoogen, B. G., T. M. Bestebroer, A. D. Osterhaus, and R. A. Fouchier. 2002. Analysis of the genomic sequence of a human metapneumovirus. Virology 295:119-132.
van den Hoogen, B. G., J. C. de Jong, J. Groen, T. Kuiken, R. de Groot, R. A. Fouchier, and A. D. Osterhaus. 2001. A newly discovered human pneumovirus isolated from young children with respiratory tract disease. Nat. Med. 7:719-724.
Viazov, S., F. Ratjen, R. Scheidhauer, M. Fiedler, and M. Roggendorf. 2003. High prevalence of human metapneumovirus infection in young children and genetic heterogeneity of the viral isolates. J. Clin. Microbiol. 41:3043-3045.
Weik, M., J. Modrof, H. D. Klenk, S. Becker, and E. Muhlberger. 2002. Ebola virus VP30-mediated transcription is regulated by RNA secondary structure formation. J. Virol. 76:8532-8539.
Williams, J. V., P. A. Harris, S. J. Tollefson, L. L. Halburnt-Rush, J. M. Pingsterhaus, K. M. Edwards, P. F. Wright, and J. E. Crowe, Jr. 2004. Human metapneumovirus and lower respiratory tract disease in otherwise healthy infants and children. N. Engl. J. Med. 350:443-450.(Ursula J. Buchholz, Stéph)
ABSTRACT
The M2 gene of human metapneumovirus (HMPV) contains two overlapping open reading frames (ORFs), M2-1 and M2-2. The expression of separate M2-1 and M2-2 proteins from these ORFs was confirmed, and recombinant HMPVs were recovered in which expression of M2-1 and M2-2 was ablated individually or together [rM2-1, rM2-2, and rM2(1+2)]. Each M2 mutant virus directed efficient multicycle growth in Vero cells. The ability to recover HMPV lacking M2-1 contrasts with human respiratory syncytial virus, for which M2-1 is an essential transcription factor. Expression of the downstream HMPV M2-2 ORF was not reduced when translation of the upstream M2-1 ORF was silenced, indicating that it is initiated separately. The rM2-2 mutants exhibited a two- to fivefold increase in the accumulation of mRNA, normalized to the genome template, suggesting that M2-2 has a role in regulating RNA synthesis. Replication and immunogenicity were tested in hamsters. Animals infected intranasally with rM2-1 or rM2(1+2) did not have recoverable virus in the lungs or nasal turbinates on days 3 or 5 postinfection and did not develop HMPV-neutralizing serum antibodies or resistance to HMPV challenge. Thus, M2-1 appears to be essential for significant virus replication in vivo. In animals infected with rM2-2, virus was recovered from only 1 of 12 animals and only in the nasal turbinates on a single day. However, all of the animals developed a high titer of HMPV-neutralizing serum antibodies and were highly protected against challenge with wild-type HMPV. The HMPV rM2-2 virus is a promising and highly attenuated HMPV vaccine candidate.
INTRODUCTION
Human metapneumovirus (HMPV) was first identified in 2001 in The Netherlands from infants and children with acute respiratory tract disease (38) and is now recognized to be worldwide in prevalence (22, 41). HMPV resembles human respiratory syncytial virus (HRSV) with regard to disease signs and the ability to infect and cause disease in infants as well as in individuals of all ages (7, 18, 20, 29, 32, 39, 41). The contribution of HMPV to respiratory tract disease remains to be fully defined but appears to be sufficient to warrant the development of a vaccine, especially for the pediatric population. Reverse genetic systems were recently developed for HMPV, allowing the generation of infectious virus from cDNA and providing an important tool for characterizing HMPV biology and for designing live-attenuated HMPV vaccines (5, 25).
HMPV has a negative-strand RNA genome of approximately 13 kb (4, 37). It has been classified, together with avian metapneumovirus, in the Metapneumovirus genus, Pneumovirus subfamily, Paramyxovirus family, of the order Mononegavirales. The Pneumovirus subfamily also contains the genus Pneumovirus, represented by HRSV. The Metapneumovirus gene order is N-P-M-F-M2-SH-G-L. By analogy to HRSV, the predicted HMPV proteins are the following: the nucleocapsid protein N, which encapsidates the RNA genome and, together with the phosphoprotein P and the RNA polymerase protein L, forms the ribonucleoprotein complex; the fusion glycoprotein F, the small hydrophobic protein SH, and the major attachment glycoprotein G that are the transmembrane surface glycoproteins; the matrix M protein; and the M2-1 and M2-2 proteins encoded by two overlapping open reading frames (ORFs) in the M2 mRNA. Among the HMPV proteins, only F, G, and SH have been identified and characterized by direct biochemical means (6). For the other HMPV proteins, there is no direct information available beyond assumptions based on deduced sequence relatedness with other pneumoviruses.
The M2 gene with its two overlapping ORFs is present in all known members of the Pneumovirus subfamily and is unique to this subfamily (4, 11, 37). The HMPV M2-1 ORF (strain CAN97-83) initiates with the first AUG at nucleotide position 14 in the predicted mRNA and would encode a protein of 187 amino acids (4, 37). The HMPV M2-2 ORF has the potential to initiate at two closely spaced AUGs at positions 525 and 537 in the mRNA, overlaps the M2-1 ORF by 53 or 41 nucleotides, respectively, and would encode a protein of up to 71 amino acids. In comparison, the M2-1 and M2-2 proteins of HRSV are 194 and 90 amino acids in length, respectively (11, 37). All of the M2-1 proteins of the Pneumovirus subfamily including HMPV contain a conserved Cys/His zinc finger-like motif (4, 11, 37).
Functions of the pneumovirus M2-1 and M2-2 proteins have been identified only in the case of HRSV (3, 12-14, 19, 24, 27). Studies with minigenome systems showed that the HRSV M2-1 protein is a transcription elongation factor that is necessary for full processivity of the viral transcriptase; in the absence of M2-1, the transcriptase terminates prematurely within several hundred nucleotides to yield a heterodisperse smear of early quitters (13). HRSV M2-1 also has an antitermination function that enhances the synthesis of readthrough mRNAs, an activity that might be important in determining the amount of polymerase delivered to downstream genes (19, 24). These activities of M2-1 are sensitive to mutations in the Cys/His motif (23). The HRSV M2-1 protein also has been shown to be an RNA-binding protein and binds to the N protein (9, 15). Expression of the HRSV M2-1 protein, either from an added support plasmid or by promiscuous expression from the antigenome cDNA, appears to be essential for the recovery of recombinant HRSV from transfected cDNA, and it has not been possible to recover infectious HRSV in which the integrity of the M2-1 protein has been drastically disturbed by introduced mutations or by extensive deletion (10, 12, 28, 34).
Deletion of the HRSV M2-2 protein resulted in a virus that replicated more slowly than wild-type HRSV in cell culture and directed the delayed and reduced production of genome and antigenome RNAs and an exaggerated production of mRNA and viral proteins (3, 27). This suggests that HRSV M2-2 functions as a regulatory factor that helps switch the viral RNA synthetic program from transcription to RNA replication, a switch that seems to operate beginning midway in the infection cycle and might be controlled by an increasing accumulation of M2-2 protein (3). In an HRSV minireplicon system, coexpression of the M2-2 protein had a strong inhibitory effect on RNA synthesis (13, 24), which could be consistent with a regulatory role but which has not yet provided insight into a mechanism of action. HRSV lacking M2-2 was highly attenuated, immunogenic, and protective in chimpanzees, and its phenotype of up-regulated antigen synthesis would be highly desirable for inclusion in a live-attenuated vaccine (3, 27, 36).
There are several other examples of mononegavirus proteins other than N, P, and L that have RNA regulatory functions. Perhaps the one most similar to the HRSV M2-1 protein is VP30 of the Marburg and Ebola viruses (31, 40). VP30 is a nucleocapsid-associated protein that interacts with a cis-acting element at the beginning of the first gene to strongly activate transcription. Thus, its mechanism of action appears to be distinct from that of HRSV M2-1 or M2-2, and its amino acid sequence is unrelated except for the common feature of a functionally important Cys/His zinc finger motif that is similar to the motif present in pneumovirus M2-1 proteins (30). Other examples are the C and V proteins of Sendai virus, which bear no apparent structural similarity to M2-1 or M2-2 apart from an unrelated Cys/His motif in the V protein. The C protein has the effect of specifically down-regulating the activity of the genome promoter and also inhibits transcription by binding the L protein (16, 21, 35). The V protein also down-regulates replication, perhaps by interacting with soluble NP complexes (26). Thus, there is diversity in the structures and functions of mononegavirus RNA regulatory factors.
As a first step to characterize the HMPV M2-1 and M2-2 proteins, their functions, and their potential as targets to make attenuated vaccines, we have generated and characterized a set of recombinant HMPVs in which the two ORFs were silenced individually or in combination. Unexpectedly, the M2-1 protein was not essential for virus recovery or growth in vitro, in sharp contrast to the situation for HRSV. However, expression of M2-1 was required for detectable replication in vivo. M2-2 also was dispensable for growth in vitro and had the effect of up-regulating transcription. HMPV lacking M2-2 was highly attenuated in vivo but was very immunogenic and protective.
MATERIALS AND METHODS
Cells and viruses. Vero cells (ATCC CCL-81) were grown in OptiProSFM (Invitrogen) supplemented with 4 mM L-glutamine. BSR T7/5 cells are baby hamster kidney 21 (BHK-21) cells that constitutively express T7 RNA polymerase (8). They were maintained in Glasgow minimal essential medium supplemented with glutamine and amino acids (Invitrogen) and 10% fetal bovine serum. The Canadian HMPV isolate CAN97-83 (33) and the recombinant HMPVs based on this strain were propagated in Vero or BSR T7/5 cells at 32°C in the absence of serum and the presence of 5 μg/ml of trypsin. For HMPV propagation on Vero cells, trypsin was replenished every second day as previously described (5).
Construction of pHMPVM2-1, pHMPVM2-2, and pHMPVM2(1+2) plasmids. The development of a reverse genetic system for HMPV based on strain HMPV CAN97-83 (GenBank accession number AY297749) was described earlier (5). The parental recombinant virus is identical to HMPV CAN97-83 except that it contains an NheI restriction site in the M-F intergenic region as a marker (Fig. 1A). Three modified forms of the M2 gene were prepared by PCR using the antigenome cDNA as template: (i) the fragment M2-1, in which the M2-1 ORF was silenced by introducing translational stop codons, was prepared by PCR with the forward primer M2-1 (5'-TTAGTTAATTAAAAATAAAATAAAATTTGGGACAAATCATAtagTCTCGCAAGGCTtagTGCAAtaaaGAAGTGCGGGGCAAtaaCAACAGAGGAAGTG-3'; the PacI restriction site is in boldface, the M2 gene start signal is underlined, and mutations to ablate expression of M2-1 ORF are shown in lowercase letters) and the reverse primer BsiWIr (5'-AAGGTACCgtAcGTGTTTTTACTAACTTAAGTAAGCCTTGAC-3'; the KpnI adapter is underlined, the BsiWI site is in boldface, and mutations to generate the BsiWI site are in lowercase letters); (ii) the fragment M2-2, in which the M2-2 ORF was silenced and partially deleted, was prepared by PCR using forward primer F4656 (HMPV nucleotides [nt] 4656 to 4676, upstream of a naturally occurring PacI site, HMPV nt 4691) and the reverse primer M2-2 (5'-AAGGTACCgtAcGTGTTTTTACTAACTTAAGCTCACTGCACTTGATTtATGCTTTCACTGTCTTGCAGGGCgTATGAAGAGTCgTTGTCAGCGCCTGC-3'; the KpnI adapter is underlined, the BsiWI site is in boldface, mutations to generate the BsiWI site are in boldface lowercase letters, and mutations to ablate translation of the M2-2 protein are in italic lowercase letters); and (iii) the fragment M2(1+2), in which the M2 gene was deleted altogether, was prepared by PCR using the positive sense primer SHM2 (5'-AAGGTACCTTAATTAAAAACACgTacGAGTGGGATAAGTG-3'; the KpnI adapter is in boldface and underlined, the PacI site is in italic boldface, the BsiWI site is in roman boldface, mutations to generate the BsiWI site are in boldface lowercase letters, and the partial SH gene start signal is underlined) and the reverse primer Kpn-rev (5'-TGTTGGTACCTACATGTTTTACTTTAGAGC-3'; HMPV nt 6968 to 6939 with the KpnI site underlined). These three PCR fragments, M2-1, M2-2, and M2(1+2), were digested with PacI and KpnI and cloned into a pBluescript vector with a multiple cloning site modified to contain a PacI and a KpnI site, and their sequences were verified. Separately, the HMPV antigenome cDNA was modified to insert a BsiWI restriction site at nt 5459 in the M2/SH intergenic region. The mutagenesis was done using the Stratagene MutaGene kit, with two complementary HMPV primers (5'-CTTAAGTTAGTAAAAACACGTACGAGTGGGATAAGTGACAATG-3' and 5'-CATTGTCACTTATCCCACTCGTACGTGTTTTTACTAACTTAAG-3', with the BsiWI restriction sites underlined and mutated nucleotides in boldface) and a template plasmid that contained the NheI/KpnI fragment of the HMPV antigenome cDNA containing the F, M2, SH, and G genes (HMPV nt 3038 to 6963) (Fig. 1A). After sequence confirmation, a restriction fragment was generated from the mutated plasmid containing the HMPV SH and G genes (nt 5460 to 6963), framed by the upstream BsiWI site and by the naturally occurring KpnI restriction site located in the HMPV G/L intergenic region (nt 6963). To assemble antigenome plasmids with the M2-1 and M2-2 mutations, three inserts were ligated simultaneously into the NheI-KpnI window of the HMPV antigenome plasmid: (i) the NheI/PacI restriction fragment containing the HMPV F gene (HMPV nt 3039 to 4691), (ii) the PacI/BsiWI fragment containing the M2-1 or M2-2 fragment described above, and (iii) the mutagenized BsiWI/KpnI fragment described above containing SH and G (HMPV nt 5460 to 6963) (Fig. 1A). To generate the antigenome plasmid containing the M2(1+2) mutant, two inserts were ligated simultaneously into the NheI-KpnI window of the HMPV antigenome plasmid: (i) the NheI/PacI fragment containing HMPV F and (ii) fragment M2, which contains SH and G, framed by PacI and KpnI. This placed HMPV F directly upstream of HMPV SH (Fig. 1A). Additionally, a second set of M2-1, M2-2, and M2(1+2) mutants was constructed by the same strategy except that they were based on an HMPV antigenome plasmid carrying a synthetic green fluorescent protein (GFP) gene in the first leader-proximal position (5) (Fig. 1A).
Recombinant HMPV (rHMPV) recovery. Confluent BSR T7/5 cells in six-well dishes were transfected with 5 μg of an individual antigenome plasmid, 2 μg each of the support plasmids pT7-N and pT7-P, and 1 μg of pT7-L per well. Transfections were done with Lipofectamine 2000 (Invitrogen) in OptiMEM without trypsin or serum and maintained overnight at 32°C. The transfection medium was removed 1 day later and replaced with Glasgow minimal essential medium without trypsin or serum. Trypsin was added on day 3 to a final concentration of 5 μg/ml, and cell-medium mixtures were passaged onto fresh Vero cells on day 6. Parallel transfections were performed with versions of each mutant with and without GFP, and recovery and passage were monitored by GFP expression.
Replication and protective efficacy of M2-1, M2-2, and M(1+2) viruses in hamsters. Groups of 6-week-old Golden Syrian hamsters (18 animals per group) were infected intranasally, under light isoflurane anesthesia, with 0.1 ml of L15 medium containing 5.7 log10 50% tissue culture infective dose (TCID50) of rHMPV, rM2-1, rM2-2, or rM(1+2). On days 3 and 5 postinfection, the lungs and nasal turbinates were harvested from six animals from each group, and the virus titers of the individual specimens were quantified by titration of tissue homogenates on Vero cells. The level of immunogenicity and the protective efficacy of the recombinants were determined in the remaining six hamsters per group. Sera were collected 2 days prior to infection and 27 days postinfection. The titers of HMPV-neutralizing antibodies were determined by an endpoint dilution neutralization assay as described previously (6). On day 28 postinfection, hamsters were challenged by intranasal administration of 5.7 log10 TCID50 of HMPV CAN97-83 in a 0.1-ml inoculum. Nasal turbinates and lungs were harvested 3 days later, and the virus titer in each tissue homogenate was determined as described above.
RESULTS
Recovery of recombinant HMPV lacking the M2-1 and/or M2-2 proteins. Reverse genetics was used to design HMPV mutants in which the M2-1 and M2-2 ORFs were silenced individually or in combination. Two versions of each mutant were made, one based on rHMPV and one based on a version (rgHMPV) that expresses the enhanced GFP protein from an additional promoter-proximal gene but otherwise was identical (Fig. 1A). The GFP versions were constructed to provide a means of directly monitoring virus recovery and passage, which was helpful because of the slow growth and trypsin dependence of HMPV. The non-GFP versions were made because the evaluation of vaccine candidates should be done with versions that can be advanced directly to clinical trials, and the expression of GFP would not be acceptable for human vaccine purposes.
To silence the M2-1 ORF, its translation initiation codon was changed to a stop codon, a second in-frame stop codon was introduced five codons downstream (whereas the next in-frame ATG is 129 codons further downstream), and additional stop codons were introduced in the alternate reading frames to preclude possible frame-shifting into the M2-1 ORF (Fig. 1B). These changes did not alter the length of the M2 gene and did not affect the downstream M2-2 ORF. To silence the M2-2 ORF, the two putative translational start codons of M2-2, which are located at HMPV positions 5235 and 5247 and overlap with the M2-1 ORF, were mutagenized from ATG to ACG, and a translational stop codon was introduced 12 codons downstream. These changes did not affect amino acid coding by the M2-1 ORF. The ATG-to-ACG changes were not ideal because ACG has the potential to serve as a translational start site, as exemplified by the C ORF of Sendai virus (17), but the efficiency would be reduced compared to ATG, and these were the only changes possible at those loci that did not alter amino acid coding in the M2-1 ORF. In addition, the portion of the M2-2 ORF that is downstream of the M2-1 ORF was deleted, removing 152 nucleotides (HMPV nt 5289 to 5440). To delete the complete M2 gene, a PacI restriction site was introduced into the gene end signal preceding the M2/SH intergenic region at nucleotide 5450, and the complete M2 gene including its gene start and gene end signal was excised by digestion with PacI, which cleaves at the above-mentioned site bordering the M2/SH intergenic region as well as at a naturally occurring PacI site located at the F gene end signal (Fig. 1A). Thus, the genome of rM2(1+2) would be 759 nucleotides shorter than that of wild-type HMPV and would completely lack the M2 gene and both of its ORFs.
To recover the recombinant viruses, the individual mutant antigenome plasmids were transfected, together with a set of three support plasmids encoding the HMPV N, P, and L proteins, into BHK-21 cells that constitutively expresses T7 RNA polymerase (8). Each of the recombinant viruses, namely rM2-1, rgM2-1, rM2-2, rgM2-2, rM2(1+2), and rgM2(1+2) (Fig. 1A), was successfully produced, with the GFP-expressing viruses serving to guide recovery. In a typical transfection experiment of three replicate wells containing 106 cells each, the efficiency of recovery of rgHMPV and rgM2-2 was exactly the same, with an average of 31 positive cells (standard error, ± 0.3) for each recombinant. Means of 7 ± 1.0 and 18 ± 1.8 cells were observed for rgM2-1 and rgM2(1+2), respectively, suggesting that expression of M2-1 provided an advantage. Recovery of for rgM2 was usually more efficient than recovery of rgM2-1, as shown in this example, possibly reflecting an advantage due to the shorter genome. The recovered viruses were amplified on Vero cells to yield stocks that ranged between 2 x 106 and 9 x 106 PFU per ml for the GFP-expressing recombinants and between 7 x 106 and 3 x 107 for the non-GFP-expressing recombinants, illustrating efficient growth with a slight growth restriction associated with the addition of the GFP gene as noted previously (5).
The presence of the designed mutations in the recovered viruses was evaluated by reverse transcription-PCR (RT-PCR) performed with an F-specific forward primer (nt 4656 to 4676) and an SH-specific reverse primer (nt 5508 to 5529). Synthesis of a detectable RT-PCR product was dependent on the presence of reverse transcriptase in the initial RT reaction (data not shown), indicating that the template was RNA rather than contaminating DNA. In the case of the M2-2 and M2(1+2) mutants, electrophoresis of the RT-PCR products on a 2% agarose gel confirmed an increase in electrophoretic mobility consistent with the designed deletions (Fig. 1A). Direct sequence analysis of the RT-PCR product for each of the recovered viruses confirmed that the sequence between the F and SH genes was exactly as designed (data not shown).
Protein synthesis by the M2 mutant viruses. To characterize the expression of proteins and RNAs by the mutant viruses, Vero cells were infected at a multiplicity of infection (MOI) of 1 PFU per cell with the non-GFP-expressing versions of the M2 mutant viruses and harvested 72 h later. Each sample was divided into two aliquots for analysis of proteins and RNAs. For protein analysis, lysates were prepared and subjected to Western blotting using a polyclonal serum raised against gradient-purified HMPV (5). Since most of the HMPV proteins had not been previously identified, positive controls were prepared by transfecting BHK-21 cells that express the T7 RNA polymerase with the support plasmids encoding the N, P, and M2-1 proteins. The HMPV-specific antiserum detected bands of approximately 42 (N protein), 40 (P protein), and 22 (M2-1 protein) kDa in the plasmid transfection samples (Fig. 2A, lanes 1 to 3), consistent with the calculated molecular sizes. These species also corresponded to major bands in the samples from cells infected with rHMPV (Fig. 2A, lane 4) and wild-type HMPV CAN97-83 (lane 8). The HMPV-infected cells also contained an additional major band that might correspond to the M protein. Note that this particular serum did not react with the HMPV F, G, and SH proteins in this highly denaturing Western blot analysis; as noted previously, this serum did react efficiently in Western blotting with F protein that was not reduced and which migrated as homo-oligomers (6). Samples from cells infected with rM2-1 and rM2(1+2) lacked the M2-1 protein band (Fig. 2A, lanes 5 and 7), confirming that its ORF had been silenced. Conversely, silencing the M2-2 ORF did not affect expression of the upstream M2-1 ORF (Fig. 2A, lane 6), as would be expected. The Western blot analysis also showed that each of the mutant viruses directed the accumulation of these major HMPV proteins at a level that was comparable to that of CAN97-83 or rHMPV, in the case of rM2-1 (Fig. 2A, lane 5), or somewhat in excess of CAN97-83 or rHMPV, in the case of rM2-2 and rM2(1+2) (lanes 6 and 7).
We were unable to detect expression of the M2-2 protein using antiserum raised against gradient-purified HMPV or antisera from rabbits immunized with synthetic peptides representing two different regions of the M2-2 protein (data not shown). As an alternative strategy, we created two additional recombinants, based on rgHMPV and rgM2-1, in which the encoded M2-2 protein was modified to have a C-terminal extension consisting of a 9-amino-acid epitope tag (YPYDVPDYA) derived from the human influenza virus hemagglutinin (HA) protein and preceded by a spacer of three alanine residues. Western blotting of lysates of cells infected with either recombinant, rgHMPV/M2-2-HA or rgM2-1/M2-2-HA, yielded an HA-specific band of approximately 8 kDa, consistent with the expected size of the M2-2-HA fusion protein (Fig. 2B, lanes 4 and 5). Thus, expression of the M2-2 ORF was not dependent on expression of the upstream M2-1 ORF. Indeed, silencing the M2-1 ORF appeared to have no effect on the efficiency of expression of M2-2.
RNA synthesis by the M2 mutant viruses. The second set of cell pellets from the experiment shown in Fig. 2, in which Vero cells had been infected with the panel of M2 mutants and harvested 72 h later, was used to isolate total intracellular RNA. These samples were subjected to Northern blot analysis using double-stranded DNA probes against the M2 mRNA (Fig. 3). Hybridization with the M2 probe confirmed the lack of expression of M2 mRNA by the rM2(1+2) mutant (Fig. 3, lane 4) and confirmed that the rM2-2 mutant produced a shorter M2 mRNA (lane 3), consistent with the deletion of 152 nt. Otherwise, the pattern of monocistronic and polycistronic readthrough mRNAs was very similar among the different viruses. In particular, the RNA pattern of the rM2-1 mutant was essentially indistinguishable from that of rHMPV, with no evidence of premature termination or changes in the frequency of transcriptional readthrough. These observations were noteworthy, given that the M2-1 protein of HRSV is necessary for processive transcription and increases the production of polycistronic readthrough mRNAs. This did not appear to be the case for HMPV.
To further evaluate the apparent increased transcription associated with deletion of the M2-2 gene, replicate cultures of Vero cells were infected with the M2 mutants, and total intracellular RNA was isolated at 12-, 24-, 36-, 48-, and 72-h time points and analyzed by Northern blot hybridization with strand-specific riboprobes for the F gene. Calculations were done based on three independent time course experiments using the GFP-expressing mutants and two experiments using rHMPV mutants; the Northern blots for one of the experiments using rgHMPV are shown in Fig. 4. The reason for using the rgHMPV virus was to monitor the uniformity of the infection and to gauge the appropriate harvest times. To compare the various viruses, the amount of mRNA for each mutant for the 24-, 36-, 48-, and 72-h time points quantified by phosphor imager was divided by the amount of genome for that mutant from a replicate gel lane for the same time point; the 12-h time point was not used because of the low levels of RNA. Then, the value for each M2 mutant virus for each time point was normalized relative to the corresponding value for rgHMPV or rHMPV (Table 1). This showed that the amount of F mRNA relative to the genome in cells infected with the rgM2-2 mutant was increased approximately three- to fivefold compared to the parental virus and two- to threefold for the non-GFP mutants, results that we consider to be essentially the same. Examination of the original Northern blots (Fig. 4) confirmed the high level of mRNA expression. In cells infected with the M2(1+2) mutant, F mRNA/genome ratios were comparable to those of cells infected with the parental virus. The mRNA/genome ratio for the M2-1 mutant was essentially the same as for the parent (Table 1), and direct examination of the original Northern blots (Fig. 4) confirmed that the intracellular level was similar to, or somewhat lower than, that of the parent virus. The finding that transcription was not elevated for the M2(1+2) mutants despite the deletion of M2-2 suggests that the combination of the M2-1 and M2-2 mutations might be more complicated than expected. Further studies using recombinant virus and minireplicons will be needed to better understand the effects of the M2-1 and M2-2 deletions and, in particular, the effect of the combined mutations.
In vitro growth of the M2 mutant viruses. To evaluate the kinetics of multicycle growth, replicate cultures of Vero cells were infected with the non-GFP-expressing and GFP-expressing versions of the M2 mutants at an MOI of 0.01 PFU/cell. At 24-h intervals, medium supernatant samples were taken, frozen, and analyzed later in parallel for infectious virus. As shown in Fig. 5A for the GFP-expressing versions, each of the M2 mutants was capable of efficient multicycle growth, showing that M2-1 and M2-2 are dispensable either singly or in combination. The rgM2-1 mutant replicated somewhat less efficiently than its rgHMPV parent during the first 7 days of incubation, although the final titers were indistinguishable. The rgM2-2 and rgM2(1+2) viruses replicated similarly to the rgHMPV parent for the first 7 days but yielded slightly higher titers at later time points. The shorter genome lengths of these two deletion mutants might have conferred a slight growth advantage. Thus, silencing the M2-1 ORF seemed to confer a very small growth restriction in vitro, whereas silencing M2-2 did not reduce the efficiency of growth.
Indirect immunofluorescence staining of permeabilized Vero cells infected with the non-GFP-expressing recombinants (Fig. 5 B) showed that the absence of M2-1 and/or M2-2 did not have a visible effect on the expression of viral antigen or the formation of viral filaments.
Replication, immunogenicity, and protective efficacy of the M2 mutants in hamsters. To assay the ability of the M2 mutants to replicate in vivo, hamsters in groups of 18 were infected intranasally with the non-GFP-expressing versions of the M2 mutants and rHMPV. Six animals from each group were sacrificed on days 3 and 5 following inoculation, and the lungs and nasal turbinates were removed and analyzed for infectious virus. rHMPV replicated to a mean titer of 5.7 and 5.4 log10 TCID50 per g of tissue in the upper respiratory tract on days 3 and 5, respectively, and to 4.0 and 3.1 log10 TCID50 in the lower respiratory tract on days 3 and 5, respectively (Table 2). In contrast, none of the 12 sacrificed animals infected with the rM2-1 or rM2(1+2) viruses yielded detectable infectious virus from either the nasal turbinates or the lungs on either day. Among the 12 sacrificed animals infected with the rM2-2 virus, a trace amount of infectious virus was recovered from the nasal turbinates of a single animal on a single day (day 3) but not from the lungs of this animal and not from any other animal on either day (Table 2).
For the remaining six animals in each group, serum samples were taken on days –2 (i.e., 2 days before infection) and 27 and assayed for the ability to neutralize HMPV CAN97-83 in vitro (Table 3). Animals that had been infected with rM2-1 or rM2(1+2) did not have detectable serum HMPV-neutralizing antibodies (Table 3), providing further evidence that these viruses did not replicate in vivo. In contrast, the rM2-2 virus induced a strong serum HMPV-neutralizing antibody response in each of the inoculated animals, with a mean titer that was only twofold less than that induced by rHMPV. This was remarkable given that only 1 of 12 animals had recoverable infectious virus. It seems unlikely that these antibodies were induced by the inoculum alone, given the lack of a detectable response to inoculation with the rM2-1 and rM2(1+2) viruses noted above. Thus, the uniformly strong antibody response to rM2-2, coupled with the ability to recover infectious virus from a single animal, suggests that a low level of replication occurred in each animal.
The animals were challenged intranasally with HMPV CAN97-83 on day 28 postinoculation and were sacrificed 3 days later, and the lungs and nasal turbinates were assayed for infectious virus. The group previously immunized with rHMPV was completely protected from challenge virus replication in both the upper and lower respiratory tract, whereas the mock-infected control group had mean challenge virus titers of 6.9 log10 PFU per g of tissue in the nasal turbinates and 5.3 log10 PFU per g of tissue in the lungs (Table 3). The mean titers of challenge virus in the animals previously immunized with rM2-1 and rM2(1+2) were essentially the same as that of the mock-immunized group, confirming the lack of a significant protective response from these highly debilitated mutants. In contrast, of the animals that had been immunized with rM2-2, all six animals were completely protected in the lungs, and four of the six animals were completely protected in the nose. The two positive animals yielded titers of challenge virus that were barely above the limit of detection. Thus, HMPV recombinants that lacked the M2-1 protein provided no evidence of replication or immunogenicity in vivo, whereas HMPV lacking M2-2 was highly restricted in vivo but nonetheless was highly immunogenic and protective and is a promising vaccine candidate.
DISCUSSION
The M2 gene with its two overlapping ORFs is present in all known members of the Pneumovirus and Metapneumovirus genera (4, 11, 37), but functions for the encoded proteins had previously been identified only for HRSV. In the present study, the M2-1 and M2-2 ORFs of recombinant HMPV were silenced individually and in combination. The two ORFs were completely dispensable for efficient replication in vitro, but their deletion was highly attenuating in vivo.
The HMPV M2-1 and M2-2 ORFs were each confirmed to express a separate protein product during HMPV infection in cell culture. The M2-1 protein was readily detected in infected-cell lysates by Western blot analysis using antibodies raised against gradient-purified HMPV. Identification was confirmed by comparison with M2-1 protein expressed from a transfected cDNA. Although we did not specifically address the issue of incorporation into virus particles in the present study, an abundant band of the appropriate size was detected previously in gradient-purified virus (6). Thus, HMPV M2-1 appears to be a virion protein, like its HRSV counterpart.
The HMPV M2-2 protein was identified by adding an epitope tag to M2-2 encoded by recombinant HMPV. This confirmed expression of the M2-2 ORF, in contrast to results with avian metapneumovirus for which it was concluded that the M2-2 ORF was not expressed (1). Unexpectedly, the M2-2 protein was expressed even when the upstream M2-1 ORF was silenced; indeed, its level of expression appeared to be completely unaffected when M2-1 was silenced. This is in contrast to the situation with HRSV, where expression of the M2-2 ORF appeared to depend on translational stop-restart by ribosomes exiting the upstream M2-1 ORF (2). However, even in the case of HRSV, details such as identification of the translational start site of M2-2 remain to be described. The mechanism by which ribosomes gain access to the HMPV M2-2 ORF remains to be identified. Typically, eukaryotic ribosomes enter an mRNA at the 5' end and scan to find appropriate start sites. It is difficult to imagine that ribosome scanning could account for the expression of the M2-2 ORF, given its location deep within the mRNA. However, this cannot be ruled out. Alternatively, scanning-independent initiation of internal start sites has been reported, as exemplified by the mechanism of ribosomal shunt described for the Y1, Y2, and X proteins of Sendai virus (17).
The finding that the HMPV M2-1 ORF is dispensable for virus recovery and replication in vitro contrasts sharply with previous results with HRSV, where it has not been possible to recover virus in which the integrity of the M2 protein was disrupted by mutation or deletion. In the case of the rM2-1 virus, the possibility of a low level of expression of a fragment of the M2-1 ORF could not be absolutely excluded because most of the ORF remained intact, but in the case of the rM2(1+2) mutant, the entire gene was deleted, excluding any possibility of residual expression. However, silencing of HMPV M2-1 was not without any effect in vitro: there was a clear and consistent reduction in the efficiency of recovery as assayed by the number of green cells produced following transfection with the GFP-expressing mutants. Also, multicycle growth for the M2-1 mutant was modestly but clearly less efficient for approximately the first 7 days compared to its rgHMPV parent, although the final titers were not different. This modest difference might reflect effects at any stage of the infectious cycle, including entry, gene expression, RNA replication, and virion production, although effects on entry seem less likely, given the presumption that M2-1 is an internal protein. The M2-1 protein of HRSV, the only one studied in any detail to date, was shown to be a transcription processivity factor, without which the polymerase can transcribe for up to several hundred nucleotides before terminating prematurely (13). HRSV M2-1 also has an antitermination activity that acts at the gene end signals to promote transcriptional readthrough (19, 24). In the case of HMPV, the loss of expression of M2-1 did not appear to affect the production of discrete, full-length mRNAs, nor did it modify the frequency or occurrence of readthrough mRNAs. Thus, the function of the HMPV M2-1 protein might be quite different from that of HRSV, even though the two proteins share the Cys/His motif and are among the most highly conserved counterparts between HMPV and HRSV. Alternatively, our understanding of the functions of HRSV M2-1 may be incomplete, and functions common to M2-1 of both viruses might remain to be identified.
In vivo, in the hamster model, HMPV lacking the M2-1 protein [rM2-1 and rM2(1+2)] was highly debilitated for replication and may not have replicated at all. Infectious virus was not recovered on days 3 and 5 postinoculation from the lungs or nasal turbinates of inoculated animals, and inoculated animals did not develop detectable HMPV-neutralizing antibodies or protection against challenge. In any case, the lack of a detectable immune response clearly indicates that the inoculum alone was not significantly immunogenic and suggests that little antigen production occurred. The lack of a detectable immune response following the administration of an inoculum containing nearly 6.0 log10 PFU of HMPV—and probably containing a much larger number of non-plaque-forming particles, given the typical high particle-to-PFU ratio of paramyxoviruses—illustrates the need to have at least some virus replication in order to stimulate a significant immune response and provides an indication that an alternative strategy involving nonreplicating virus-like particles would require a very high dose.
The ability to delete HMPV M2-2 without compromising growth in vitro is reminiscent of results with HRSV (3, 13, 27, 36). For HRSV, deletion of M2-2 resulted in a decrease in RNA replication and an increase in transcription, whereas for HMPV only the second effect was evident. The rM2-2 virus also directed an increased level of accumulation of intracellular HMPV proteins as detected by Western blotting. An increase in the synthesis of viral antigens might increase the immunogenicity of the virus, which would be a desirable property for a vaccine. However, it is clear that we have more to learn about the role of M2-2 (and M2-1), since the up-regulation of transcription associated with M2-2 was not observed when this deletion was combined with the deletion of M2-1, at least not for the F gene.
The restriction on replication in vivo associated with silencing M2-2 appeared to be greater in the case of HMPV than for HRSV. Whether this greater attenuating effect is related to the slower growth of HMPV is unclear. However, even though replication of the HMPV rM2-2 virus could not be directly detected in most of the animals, each animal developed a high titer of virus-neutralizing antibodies and was highly protected against challenge with wild-type HMPV. This suggests that a low level of replication did occur. The high degree of immunogenicity observed in the near absence of detectable infectious virus might be an indication that rM2-2 is indeed more immunogenic due to its up-regulated transcription. Thus, the HMPV rM2-2 mutant is a highly attenuated, highly immunogenic, and protective vaccine candidate that warrants further evaluation in nonhuman primates and in the clinic.
ACKNOWLEDGMENTS
We thank Guy Boivin of the Centre Hospitalier Universitaire de Québec and Laval University and Dean Erdman, Teresa Peret, and Larry Anderson of the Centers for Disease Control and Prevention for providing initial stocks of CAN97-83 and Chris Hanson for sequencing.
REFERENCES
Ahmadian, G., P. Chambers, and A. J. Easton. 1999. Detection and characterization of proteins encoded by the second ORF of the M2 gene of pneumoviruses. J. Gen. Virol. 80:2011-2016.
Ahmadian, G., J. S. Randhawa, and A. J. Easton. 2000. Expression of the ORF-2 protein of the human respiratory syncytial virus M2 gene is initiated by a ribosomal termination-dependent reinitiation mechanism. EMBO J. 19:2681-2689.
Bermingham, A., and P. L. Collins. 1999. The M2-2 protein of human respiratory syncytial virus is a regulatory factor involved in the balance between RNA replication and transcription. Proc. Natl. Acad. Sci. USA 96:11259-11264.
Biacchesi, S., M. H. Skiadopoulos, G. Boivin, C. T. Hanson, B. R. Murphy, P. L. Collins, and U. J. Buchholz. 2003. Genetic diversity between human metapneumovirus subgroups. Virology 315:1-9.
Biacchesi, S., M. H. Skiadopoulos, K. C. Tran, B. R. Murphy, P. L. Collins, and U. J. Buchholz. 2004. Recovery of human metapneumovirus from cDNA: optimization of growth in vitro and expression of additional genes. Virology 321:247-259.
Biacchesi, S., M. H. Skiadopoulos, L. Yang, E. W. Lamirande, K. C. Tran, B. R. Murphy, P. L. Collins, and U. J. Buchholz. 2004. Recombinant human metapneumovirus lacking the small hydrophobic SH and/or attachment G glycoprotein: deletion of G yields a promising vaccine candidate. J. Virol. 78:12877-12887.
Boivin, G., Y. Abed, G. Pelletier, L. Ruel, D. Moisan, S. Cote, T. C. Peret, D. D. Erdman, and L. J. Anderson. 2002. Virological features and clinical manifestations associated with human metapneumovirus: a new paramyxovirus responsible for acute respiratory-tract infections in all age groups. J. Infect. Dis. 186:1330-1334.
Buchholz, U. J., S. Finke, and K. K. Conzelmann. 1999. Generation of bovine respiratory syncytial virus (BRSV) from cDNA: BRSV NS2 is not essential for virus replication in tissue culture, and the human RSV leader region acts as a functional BRSV genome promoter. J. Virol. 73:251-259.
Cartee, T. L., and G. W. Wertz. 2001. Respiratory syncytial virus M2-1 protein requires phosphorylation for efficient function and binds viral RNA during infection. J. Virol. 75:12188-12197.
Collins, P. L., E. Camargo, and M. G. Hill. 1999. Support plasmids and support proteins required for recovery of recombinant respiratory syncytial virus. Virology 259:251-255.
Collins, P. L., R. M. Chanock, and B. R. Murphy. 2001. Respiratory syncytial virus, p. 1443-1485. In D. M. Knipe, P. M. Howley, D. E. Griffin, R. A. Lamb, M. A. Martin, B. Roizman, and S. E. Straus (ed.), Fields virology, 4th ed. Lippincott Williams & Wilkins, Philadelphia, Pa.
Collins, P. L., M. G. Hill, E. Camargo, H. Grosfeld, R. M. Chanock, and B. R. Murphy. 1995. Production of infectious human respiratory syncytial virus from cloned cDNA confirms an essential role for the transcription elongation factor from the 5' proximal open reading frame of the M2 mRNA in gene expression and provides a capability for vaccine development. Proc. Natl. Acad. Sci. USA 92:11563-11567.
Collins, P. L., M. G. Hill, J. Cristina, and H. Grosfeld. 1996. Transcription elongation factor of respiratory syncytial virus, a nonsegmented negative-strand RNA virus. Proc. Natl. Acad. Sci. USA 93:81-85.
Collins, P. L., M. G. Hill, and P. R. Johnson. 1990. The two open reading frames of the 22K mRNA of human respiratory syncytial virus: sequence comparison of antigenic subgroups A and B and expression in vitro. J. Gen. Virol. 71:3015-3020.
Cuesta, I., X. Geng, A. Asenjo, and N. Villanueva. 2000. Structural phosphoprotein M2-1 of the human respiratory syncytial virus is an RNA binding protein. J. Virol. 74:9858-9867.
Curran, J., R. Boeck, and D. Kolakofsky. 1991. The Sendai virus P gene expresses both an essential protein and an inhibitor of RNA synthesis by shuffling modules via mRNA editing. EMBO J. 10:3079-3085.
Curran, J., P. Latorre, and D. Kolakofsky. 1998. Translational gymnastics on the Sendai virus P/C mRNA. Semin. Virol. 8:351-357.
Falsey, A. R., D. Erdman, L. J. Anderson, and E. E. Walsh. 2003. Human metapneumovirus infections in young and elderly adults. J Infect. Dis. 187:785-790.
Fearns, R., and P. L. Collins. 1999. Role of the M2-1 transcription antitermination protein of respiratory syncytial virus in sequential transcription. J. Virol. 73:5852-5864.
Greensill, J., P. S. McNamara, W. Dove, B. Flanagan, R. L. Smyth, and C. A. Hart. 2003. Human metapneumovirus in severe respiratory syncytial virus bronchiolitis. Emerg. Infect. Dis. 9:372-375.
Grogan, C. C., and S. A. Moyer. 2001. Sendai virus wild-type and mutant C proteins show a direct correlation between L polymerase binding and inhibition of viral RNA synthesis. Virology 288:96-108.
Hamelin, M. E., Y. Abed, and G. Boivin. 2004. Human metapneumovirus: a new player among respiratory viruses. Clin. Infect. Dis. 38:983-990.
Hardy, R. W., and G. W. Wertz. 2000. The Cys3-His1 motif of the respiratory syncytial virus M2-1 protein is essential for protein function. J. Virol. 74:5880-5885.
Hardy, R. W., and G. W. Wertz. 1998. The product of the respiratory syncytial virus M2 gene ORF1 enhances readthrough of intergenic junctions during viral transcription. J. Virol. 72:520-526.
Herfst, S., M. de Graaf, J. H. Schickli, R. S. Tang, J. Kaur, C. F. Yang, R. R. Spaete, A. A. Haller, B. G. van den Hoogen, A. D. Osterhaus, and R. A. Fouchier. 2004. Recovery of human metapneumovirus genetic lineages A and B from cloned cDNA. J. Virol. 78:8264-8270.
Horikami, S. M., S. Smallwood, and S. A. Moyer. 1996. The Sendai virus V protein interacts with the NP protein to regulate viral genome RNA replication. Virology 222:383-390.
Jin, H., X. Cheng, H. Z. Zhou, S. Li, and A. Seddiqui. 2000. Respiratory syncytial virus that lacks open reading frame 2 of the M2 gene (M2-2) has altered growth characteristics and is attenuated in rodents. J. Virol. 74:74-82.
Jin, H., D. Clarke, H. Z. Zhou, X. Cheng, K. Coelingh, M. Bryant, and S. Li. 1998. Recombinant human respiratory syncytial virus (RSV) from cDNA and construction of subgroup A and B chimeric RSV. Virology 251:206-214.
Maggi, F., M. Pifferi, M. Vatteroni, C. Fornai, E. Tempestini, S. Anzilotti, L. Lanini, E. Andreoli, V. Ragazzo, M. Pistello, S. Specter, and M. Bendinelli. 2003. Human metapneumovirus associated with respiratory tract infections in a 3-year study of nasal swabs from infants in Italy. J. Clin. Microbiol. 41:2987-2991.
Modrof, J., S. Becker, and E. Muhlberger. 2003. Ebola virus transcription activator VP30 is a zinc-binding protein. J. Virol. 77:3334-3338.
Muhlberger, E., M. Weik, V. E. Volchkov, H. D. Klenk, and S. Becker. 1999. Comparison of the transcription and replication strategies of Marburg virus and Ebola virus by using artificial replication systems. J. Virol. 73:2333-2342.
Osterhaus, A., and R. Fouchier. 2003. Human metapneumovirus in the community. Lancet 361:890-891.
Peret, T. C., G. Boivin, Y. Li, M. Couillard, C. Humphrey, A. D. Osterhaus, D. D. Erdman, and L. J. Anderson. 2002. Characterization of human metapneumoviruses isolated from patients in North America. J. Infect. Dis. 185:1660-1663.
Tang, R. S., N. Nguyen, X. Cheng, and H. Jin. 2001. Requirement of cysteines and length of the human respiratory syncytial virus M2-1 protein for protein function and virus viability. J. Virol. 75:11328-11335.
Tapparel, C., S. Hausmann, T. Pelet, J. Curran, D. Kolakofsky, and L. Roux. 1997. Inhibition of Sendai virus genome replication due to promoter-increased selectivity: a possible role for the accessory C proteins. J. Virol. 71:9588-9599.
Teng, M. N., S. S. Whitehead, A. Bermingham, M. St. Claire, W. R. Elkins, B. R. Murphy, and P. L. Collins. 2000. Recombinant respiratory syncytial virus that does not express the NS1 or M2-2 protein is highly attenuated and immunogenic in chimpanzees. J. Virol. 74:9317-9321.
van den Hoogen, B. G., T. M. Bestebroer, A. D. Osterhaus, and R. A. Fouchier. 2002. Analysis of the genomic sequence of a human metapneumovirus. Virology 295:119-132.
van den Hoogen, B. G., J. C. de Jong, J. Groen, T. Kuiken, R. de Groot, R. A. Fouchier, and A. D. Osterhaus. 2001. A newly discovered human pneumovirus isolated from young children with respiratory tract disease. Nat. Med. 7:719-724.
Viazov, S., F. Ratjen, R. Scheidhauer, M. Fiedler, and M. Roggendorf. 2003. High prevalence of human metapneumovirus infection in young children and genetic heterogeneity of the viral isolates. J. Clin. Microbiol. 41:3043-3045.
Weik, M., J. Modrof, H. D. Klenk, S. Becker, and E. Muhlberger. 2002. Ebola virus VP30-mediated transcription is regulated by RNA secondary structure formation. J. Virol. 76:8532-8539.
Williams, J. V., P. A. Harris, S. J. Tollefson, L. L. Halburnt-Rush, J. M. Pingsterhaus, K. M. Edwards, P. F. Wright, and J. E. Crowe, Jr. 2004. Human metapneumovirus and lower respiratory tract disease in otherwise healthy infants and children. N. Engl. J. Med. 350:443-450.(Ursula J. Buchholz, Stéph)