A Single Amino Acid Substitution in the Phosphoprotein of Respiratory Syncytial Virus Confers Thermosensitivity in a Reconstituted RNA Polymerase System

Abstract

The single amino acid change Gly172 to Ser in the phosphoprotein (P) of respiratory syncytial virus (RSV) has previously been shown to be responsible for the thermosensitivity and protein-negative phenotype of tsN19, a mutant of the B subgroup RSN-2 strain. This single change was inserted into the P gene of the A subgroup virus RSS-2, and the resulting phenotype was observed in a plasmid-driven reconstituted RSV RNA polymerase system. Expression from a genome analogue containing two reporter genes was thermosensitive when directed by plasmids containing the N, L, M2, and mutant P genes cloned under the control of T7 promoters. Analysis of RNA synthesis showed that mutant P protein was unable to produce genome, antigenome, or mRNA at the restrictive temperature. At a semipermissive temperature, genome, antigenome, and mRNA synthesis were all reduced, 6- to 30-fold, relative to synthesis directed by a wild-type P plasmid. Binding of the mutant P protein to N protein in the absence of other viral proteins was unaffected by temperature, indicating that the lesion did not produce a large enough structural change to disrupt this binding. These data suggest that the plasmid rescue system is suitable for investigation of the role of thermosensitive mutations in RSV polymerase components in RNA synthesis.

The minimal RNA polymerase requirements of respiratory syncytial virus (RSV) have been defined from in vivo polymerase reconstitution experiments (i.e., a plasmid-driven rescue system) to be the ribonucleoprotein (RNA complexed with nucleoprotein [N]), the phosphoprotein (P), and the large protein (L) (11, 25). In combination with a genomic RNA analogue (minigenome) containing the terminal leader and trailer sequences, as well as gene start and gene end signals, these three proteins direct both replication and transcription of the synthetic minigenome. However, transcription is greatly enhanced by the presence of a further viral protein, M2 (also known as the 22K protein [2, 12]). The M2 gene (with two open reading frames [ORFs], M2-1 and M2-2) is unique to the genus Pneumovirus, but the functions of the N, P, and L proteins are believed to be equivalent to those of the other members of the Paramyxoviridae and of the members of the Rhabdoviridae. The L protein contains the sequence motifs for RNA-dependent RNA polymerase activity and is presumed to be responsible for capping and polyadenylation of mRNAs (22). P protein interacts with soluble N protein principally via the carboxyl-terminal 20 residues of P (9, 21) and is a component of purified nucleocapsids along with N and L proteins. An essential role for the P protein in RNA synthesis was suggested by the properties of the thermosensitive (ts) mutant tsN19. This mutant was isolated from the B subgroup strain RSN-2 following chemical mutagenesis (6) and was assigned to complementation group E (10). tsN19 is highly ts at 39°C (efficiency of plating, 7 × 10⁻⁸) and less so at 37°C (efficiency of plating, 2 × 10⁻⁶); it was observed to be protein negative (defined as having undetectable levels of antigen and intracellular viral polypeptides [1, 18]) at 39°C (18). P protein made by this mutant at the permissive temperature was not degraded at the restrictive temperature, suggesting that the ts defect did not involve thermolability of the P protein. Sequence analysis of the P genes of RSN-2, tsN19, and a revertant virus, ts⁺R3/6, identified the amino acid change Gly172 to Ser (G172S) as correlating with the ts phenotype, and the authors suggested that the mutation led to a defect in mRNA synthesis (1). It was noted that Gly at position 172 of the P protein is conserved in both A and B subgroups of human RSV and in bovine RSV (14).

In this report we describe the introduction of the G172S mutation into a subgroup A RSV background and show that the ts phenotype is expressed in a plasmid-driven minigenome RNA transcription-replication assay and hence that the single point mutation in the P protein alone is sufficient to abolish RNA synthesis.

Temperature dependence of the minigenome replication assay.

The assay used for transcription and replication was similar to that used previously for RSV and other members of the Paramyxoviridae (11; reviewed in reference 15). Recombinant vaccinia virus vTF7-3 (8) was used to supply T7 RNA polymerase (T7 pol) in the cytoplasm of transfected cells to transcribe both the minigenome analogue and the four protein components of the RSV replication complex. The N, P, L, and M2 ORFs of RSV strain RSS-2 (23) were cloned under the control of T7 promoters in vectors pTM1 (16) (T7-N, T7-L, and T7-M2) or pBluescribe (Stratagene) (T7-P). The dicistronic minigenome analogue, pCAT-Luc, contained, in the following order, the hepatitis delta virus ribozyme (17); the 3′-most 98 nucleotides (nt) of the RSS-2 genome, comprising the leader and NS1 gene start and 5′ untranslated region; the chloramphenicol acetyltransferase (CAT) ORF; the NS1 gene end, 6-nt intergenic region, and N gene start; the luciferase (Luc) ORF from vector pGL3-control (Promega); the 5′-most 166 nt of the RSS-2 genome, comprising the L gene end and trailer; and the T7 pol promoter. Transcription by T7 pol resulted in a 2,654-nt transcript of genomic (−) polarity, after autocleavage by the ribozyme to generate the correct 3′ end. Mixtures of plasmids were transfected into subconfluent monolayers of HEp-2 cells previously infected with vTF7-3 at a multiplicity of infection of 1 PFU per cell by using Lipofectace (Gibco-BRL Life Technologies). Typically, 0.2 μg of pCAT-Luc, 0.4 μg of T7-N, 0.8 μg of T7-P, 0.1 μg of T7-L, and 50 ng of T7-M2 were mixed with 4 μl of Lipofectace and used to transfect 5 × 10⁵ vaccinia virus-infected cells. These amounts of plasmids had previously been determined to be the optimum amounts as judged by maximal reporter protein production. After incubation at the appropriate temperature, the cells were harvested for lysis and reporter assay. CAT gene expression was determined as total protein by an antigen capture enzyme-linked immunosorbent assay (Boehringer-Mannheim) with purified CAT protein as the standard. Luc was determined with luciferase assay reagent (Promega) in a Labsystems Luminoskan luminometer and was expressed as relative light units (RLU). In both cases, the values were expressed as yield per 5 × 10⁵ cells.

The optimal temperature for vTF7-3 is 37°C; however, we wished to test the expression of reporter gene activity at the permissive and restrictive temperatures of tsN19 virus, namely, 33 and 39°C. Time courses of pCAT-Luc rescue at 33, 37, and 39°C showed that the optimum yields of CAT and Luc were obtained after 72 to 96, 48 to 72, and 48 h, respectively (data not shown). At the relatively low multiplicity of infection used, the cells were still alive, although exhibiting cytopathic effect, after 96 h at 33°C. For subsequent experiments, cells were harvested after 48 h (37 and 39°C) or 72 h (33°C). Levels of CAT and Luc expression were typically 100- to 200-fold above background. For wild-type (wt) P plasmid, the yields of CAT and Luc were essentially the same when rescue was performed at 33°C (72-h harvest) or 37°C (48-h harvest). At 39°C, CAT protein expression was reduced to 14% and Luc expression was reduced to 17% of the values at 33°C (these are not significantly different [P = 0.82]). Since wt RSV replicates well at 39°C (efficiency of plating at 39°C/37°C = 1.0 [6]), this suggests that the supply of T7 pol is reduced at the higher temperature. To avoid this complication, we compared ts P to wt P separately at each temperature.

Thermosensitivity of mutant P in reporter gene rescue.

The G172S mutation was introduced into the wt T7-P plasmid by PCR with the mutagenic primer RSS2PGS (5′ GATGCCATGGTTAGTTTAAGAGAAGAAATGAT 3′) and the reverse sequencing primer (5′ TTGTGAGCGGATAACAATTTC 3′). The NcoI site in RSS2PGS is underlined, and the mutagenic nucleotide is shown in bold type. After PCR with Taq DNA polymerase, the product was used to replace the equivalent HindIII-NcoI fragment of T7-P to produce T7-tsN19P. Expression of CAT and Luc gene activities by the T7-tsN19P plasmid was compared to those by T7-P, along with pCAT-Luc, T7-N, T7-L, and T7-M2 plasmids, at 33, 37, and 39°C (Fig. 1). At 33°C, T7-tsN19P is clearly functional, rescuing 57 to 68% of reporter gene activity compared to the levels seen with the wt protein expressed from T7-P. This reduction from the wt level may be due to insertion of the mutation into a heterologous background or to slight temperature sensitivity, since the permissive temperature was originally defined as 31°C (6). It is also possible that the mutation has a slight effect on the function of P protein even at the fully permissive temperature.

FIG. 1.

Reporter gene rescue by T7-tsN19P at permissive and restrictive temperatures. The values for T7-tsN19P were normalized to those for T7-P at each temperature, as were the values obtained when P plasmid was omitted. Error bars show the standard deviation of four replicate experiments.

At 37°C, the activity of T7-tsN19P is clearly reduced relative to that at 33°C; the CAT and Luc activities are 3 and 11% of the wt values, respectively. The more marked inhibition of CAT than of Luc was significant (P = 0.01), but the reason for this is not clear.

At 39°C, the ability of T7-tsN19P to direct the expression of the reporter genes was reduced even further, to 0.4 to 1% of the wt values. Compared to the controls with no P plasmid, T7-tsN19P was not significantly different, implying that no residual activity was detectable in our system at the restrictive temperature of the original mutant virus.

We did not attempt to test different concentrations of T7-tsN19P plasmid, since it has been established that nonoptimal amounts of the helper plasmids in minigenome rescues decrease the reporter activity (references 7 and 11 and unpublished data), thus making quantitative data uninterpretable.

Thermosensitivity of the mutant P gene in RNA synthesis.

RNA was extracted from transfected cells and analyzed by Northern blotting to determine if replication of the minigenome, as well as transcription, was affected. Total RNA was extracted with Trizol reagent (Gibco-BRL Life Technologies) and separated into poly(A)⁺ and poly(A)⁻ fractions on oligo(dT)-cellulose (Pharmacia type 7). RNAs were Northern blotted and probed with digoxigenin-labelled riboprobes as described previously (4). Transcription was analyzed by probing the poly(A)⁺ mRNAs with a negative-sense riboprobe (Fig. 2). At 33°C, no difference was seen between the wt and ts P plasmids, whereas omitting P produced only a background smear of RNA. The major bands seen are CAT mRNA (0.7 kb) and Luc mRNA (1.7 kb). The uppermost poly(A)⁺ mRNA band (2.5 kb) is presumably the CAT-Luc readthrough mRNA, since antigenome will not be selected by oligo(dT)-cellulose. At 37°C, transcription by T7-tsN19P was greatly reduced, with only very weak CAT and Luc mRNA bands being visible (lane 3). At 39°C, no mRNAs were visible from T7-tsN19P, just a smear resembling the sample with no P plasmid. These data agree well with the results of the reporter protein assays, demonstrating that reporter translation accurately reflects transcription levels.

FIG. 2.

Northern blot of mRNA selected by oligo(dT)-cellulose. The probe was a digoxigenin-labelled riboprobe corresponding to the minus strand of the pCAT-Luc minigenome. Lanes: 1, no P plasmid; 2, T7-P; 3, T7-tsN19P.

Replication was analyzed by digesting cell lysates with micrococcal nuclease S7 to remove unencapsidated RNAs (7) before extraction with Trizol reagent, leaving only replicative RNAs protected by N protein intact. Use of a positive-sense riboprobe to detect minigenome (Fig. 3A) or a negative-sense riboprobe to detect antiminigenome (Fig. 3B) produced very similar results; T7-tsN19P (lanes 3) was roughly equal in activity to T7-P (lanes 2) at 33°C, greatly reduced at 37°C (band intensities, 17 to 18% of wt levels), and indistinguishable from background at 39°C. The faint minigenome band seen in the no-T7-L tracks at 33 and 37°C (Fig. 3A, lanes 4) probably represent negative-sense primary transcripts from the pCAT-Luc plasmid which have been encapsidated by N (and P) but are not being replicated; the equivalent positive-sense band is not visible in Fig. 3B. These data suggest that mutant P is equally defective in the synthesis of plus- and minus-strand RNAs.

FIG. 3.

Northern blots of RNA which was protected from nuclease S7 digestion. The probes were the digoxigenin-labelled riboprobe corresponding to the minus strand of the pCAT-Luc minigenome (B) and the digoxigenin-labelled riboprobe corresponding to the plus strand of the CAT gene (A). Lanes: 1, no P plasmid; 2, T7-P; 3, T7-tsN19P, 4, T7-P but no T7-L. The 39°C tracks in panel B were exposed for 25 times as long as the other tracks, to show the fainter bands.

Thermosensitivity of mutant P protein in the N-binding assay.

Binding of P protein to the N protein of RSS-2 was assayed by using the Clontech Matchmaker mammalian two-hybrid system at 33 and 39°C, that is, the permissive and restrictive temperatures of the tsN19 mutation. The N gene was inserted into the binding-domain vector, pM, and the P gene was inserted into the activation domain vector, pVP16. This results in the production of N protein as a fusion at the C terminus of the yeast GAL4 binding domain and of P protein as a fusion at the C terminus of VP16 of herpes simplex virus type 1. The G172S mutation was introduced into pVP16-P on a NcoI-HindIII fragment from T7-tsN19P. The N and P plasmids were transfected into COS-7 cells along with the pG5CAT reporter plasmid as described by Slack and Easton (21). Plasmid pG5CAT contains CAT as the reporter gene, under the control of five tandem consensus GAL4 binding sites and the minimal promoter of the adenovirus E1b gene. Interaction of the N and P fusion proteins, concurrent with binding of the GAL4-N fusion protein to the GAL4 binding sites, brings the VP16 activating domain into proximity with the RNA polymerase II and results in transcription of the CAT gene. Cells were harvested for CAT enzyme-linked immunosorbent assay after 48 h (39°C) or 72 h (33°C) of incubation.

At each temperature, the level of CAT expression obtained with the tsN19 mutation in the P-activation domain vector was compared to the activity obtained with the wt P-activation domain vector (Fig. 4). At 33°C the relative activity (ts P/wt P) was 0.99 ± 0.07, and at 39°C the relative activity was 1.01 ± 0.09. These figures are not significantly different (P = 0.76), and since expression of CAT activity reflects the level of interaction of the N and P proteins, this demonstrates that the mutant P protein shows no thermosensitivity in its interaction with N protein. These data suggest that the ts P protein must be stable at 39°C to preserve the interaction with the N protein. This was confirmed directly by expressing the protein with vTF7-3 in the absence of other RSV proteins (Fig. 5). At 39°C, just as much protein is produced from T7-tsN19P as from wild-type T7-P.

FIG. 4.

Reporter gene activity produced by two-hybrid interactions. Values are expressed as picograms per 2 × 10⁵ cells. Error bars show the standard deviation of four or five replicate experiments. “no P” refers to empty pVP16 vector in place of the pVP16-P fusion.

FIG. 5.

Immunoblot analysis of wt and ts P proteins at the restrictive temperature. Following separation by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and blotting onto a nitrocellulose membrane, cell lysates were probed with a bovine polyclonal anti-RSV serum. Lanes 1 to 3 contain extracts of cells infected with vTF7-3 and transfected with T7-P (lane 1), T7-tsN19P (lane 2), no DNA (lane 3). Lane 4 contains RSS-2 virus-infected cell extract.

We have shown that the vaccinia virus-T7-based reconstituted RNA polymerase system can be used to identify functional conditional mutations in the RSV replication complex. The RSS-2-based minigenome rescue system showed all types of RSV RNA polymerase activity at temperatures between 33 and 39°C, namely, replication, transcription, polycistronic mRNA synthesis, and polyadenylation. Capping was also assumed to be occurring, since the CAT and Luc transcripts were readily translated.

A single point mutation at Gly172 has previously been implicated in the temperature sensitivity of the conditional lethal mutant tsN19: the tsN19 mutation was indicated to lie in the P protein, since the mutation caused the loss of reactivity of tsN19 virus with the P protein-specific monoclonal antibody 3-5 whereas revertant ts⁺R3/6 regained the ability to bind monoclonal antibody 3-5 (1). The evidence that this mutation was solely responsible for the phenotype, although compelling, was of necessity indirect. By inserting the mutation into a P gene used in a reverse genetics system, we have demonstrated directly that the G172S mutation alone confers thermosensitivity on the function of the expressed P protein in RNA synthesis. This occurs even though the mutation is inserted into a subgroup A genetic background instead of the subgroup B in which it was first identified. Furthermore, the increased sensitivity at 39 over 37°C observed for the original tsN19 virus was also reproduced in this system.

It has been suggested that this mutation may render the mutant P protein unable to interact functionally with one or more of the components of the ribonucleoprotein complex at the restrictive temperature (1). The region of the P protein around residue 172 is not required for binding to N protein; indeed, a mutant P protein in which residues 168 to 198 were deleted was still able to interact with N as strongly as the wt was (21). The two-hybrid data presented here show that binding of the mutant P protein to N protein is not thermosensitive, so that if the restrictive temperature induces an inactivating conformational change in the P protein, this change does not extend to the major N binding regions. However, other putative functions of the P protein such as oligomerization, binding to L protein, or direct interaction with the N-RNA template, have not been investigated, and the responsible domains of the P protein remain to be mapped.

Large panels of conditional lethal mutants have been isolated for a number of negative-strand viruses, in particular the paramyxoviruses RSV, Newcastle disease virus, Sendai virus, and measles virus (20); the rhabdoviruses vesicular stomatitis virus and Chandipura virus (19); and 11 members of the Bunyaviridae (5). Although many of the lesions remain unmapped, several complementation groups have been more or less tentatively assigned to polymerase component genes. In RSV, several ts lesions have been mapped to the L protein by genomic sequencing and reverse genetic methods (3, 13, 23, 24). For many of the above viruses, minigenome-RNA polymerase reconstitution systems analogous to the one described here are available or under development. These systems could be used to correlate complementation groups with specific genes involved in RNA synthesis and to identify residues critical for the function of the RNA polymerase complex.