Inhibition of Sendai virus genome replication due to promoter-increased selectivity: a possible role for the accessory C proteins

(1)

Article

Reference

Inhibition of Sendai virus genome replication due to

promoter-increased selectivity: a possible role for the accessory C proteins

TAPPAREL, Caroline, et al .

Abstract

The role of the negative-stranded virus accessory C proteins is difficult to assess because they appear sometimes as nonessential and thereby of no function. On the other hand, when a function is found, as in the case of Sendai virus, it represents an enigma, in that the C proteins inhibit replication under conditions where the infection follows an exponential course.

Furthermore, this inhibitory function is exerted differentially: in contrast to the replication of internal deletion defective interfering (DI) RNAs, that of copy-back DI RNAs appears to escape inhibition, under certain experimental conditions (in vivo assay). In a reexamination of the C effect by the reverse genetics approach, it was found that copy-back RNA replication is inhibited by C in vivo as well, under conditions where the ratio of C to copy-back template is increased. This effect can be reversed by an increase in P but not L protein. The "rule of six"

was differentially observed in the presence or absence of C. Finally, a difference in the ability of the replicating complex to tolerate promoter modifications in RNA synthesis initiation was shown to [...]

TAPPAREL, Caroline, et al . Inhibition of Sendai virus genome replication due to

promoter-increased selectivity: a possible role for the accessory C proteins. Journal of Virology , 1997, vol. 71, no. 12, p. 9588-9599

DOI : 10.1128/JVI.71.12.9588-9599.1997 PMID : 9371623

Available at:

http://archive-ouverte.unige.ch/unige:139737

Disclaimer: layout of this document may differ from the published version.

1 / 1

(2)

Inhibition of Sendai Virus Genome Replication due to Promoter-Increased Selectivity: a Possible Role

for the Accessory C Proteins

CAROLINE TAPPAREL, STE´PHANE HAUSMANN, THIERRY PELET, JOSEPH CURRAN, DANIEL KOLAKOFSKY,^ANDLAURENT ROUX*

Department of Genetics and Microbiology, University of Geneva Medical School, CMU, 1211 Geneva 4, Switzerland Received 31 March 1997/Accepted 9 September 1997

The role of the negative-stranded virus accessory C proteins is difficult to assess because they appear sometimes as nonessential and thereby of no function. On the other hand, when a function is found, as in the case of Sendai virus, it represents an enigma, in that the C proteins inhibit replication under conditions where the infection follows an exponential course. Furthermore, this inhibitory function is exerted differentially: in contrast to the replication of internal deletion defective interfering (DI) RNAs, that of copy-back DI RNAs appears to escape inhibition, under certain experimental conditions (in vivo assay). In a reexamination of the C effect by the reverse genetics approach, it was found that copy-back RNA replication is inhibited by C in vivo as well, under conditions where the ratio of C to copy-back template is increased. This effect can be reversed by an increase in P but not L protein. The “rule of six” was differentially observed in the presence or absence of C. Finally, a difference in the ability of the replicating complex to tolerate promoter modifications in RNA synthesis initiation was shown to occur in the presence or the absence of C as well. We propose that C acts by increasing the selectivity of the replicating complex for the promoter cis-acting elements governing its activity.

The inhibitory effect of C becomes the price to pay for this increased selectivity.

Unless otherwise determined by specific genes (e.g., induc- ing latency), by specific viral constituents (e.g., defective interfering [DI] particles), or by a particular cellular environment (e.g., nonpermissive cells), viral infections spread exponentially, i.e., viral macromolecules tend to accumulate to provide all the components for efficient viral assembly and virion pro- duction. Unlike those of structural proteins and minimal essential genes, nonstructural functions always represent an enigma in this context. The set of Sendai virus (SeV) C proteins is a pertinent example of these. Not only are the C proteins marginally represented in virus particles, but also they appear to exert a negative effect on the extent of genome replication, this in the context of an otherwise exponentially spreading infection.

SeV is a member of the Paramyxovirus genus in the Para- myxovirinae subfamily. All members of this subfamily contain at least six transcription units (genes), encoding the nucleocapsid (N) protein, phosphoprotein (P protein), matrix (M) protein, two surface glycoproteins (H or HN and F proteins), and the large (L) RNA polymerase protein, which represent the minimal essential genes of this virus subfamily. Many of these viruses also encode other proteins from extra transcription units or from alternative expression mechanisms in the P gene (25). The SeV P gene is remarkable in this respect, in that it can encode as many as eight different polypeptides via overlapping open reading frames (ORF), mRNA editing, leaky ribosomal scanning, and a ribosomal shunt (6, 7, 22, 23, 26a).

The C proteins are a nested set of four polypeptides (called C9, C, Y1, and Y2), initiated at different translation start sites (in the 11 frame relative to that of P) but sharing a common C-terminal end. C9 translation starts on an unusual ACG

codon at position 81 of the P mRNA, upstream of the P protein initiation codon (AUG¹⁰⁴). C, Y1, and Y2 then initiate on ATGs at positions 114, 183, and 201, respectively. The four polypeptides end on TAA⁷²⁶and are thus, respectively, 215, 204, 181, and 175 amino acids in length. C proteins are ex- pressed by all viruses of the Paramyxovirus and the Morbillivirus genera but not necessarily as the whole array of polypeptides.

For instance, measles virus appears to express only a single C protein (33). In addition, an overlapping C ORF is absent in members of the Rubulavirus genus, which is the remaining genus of the Paramyxovirinae. In the more distantly related Rhabdoviridae family, vesiculoviruses such as vesicular stoma- titis virus (VSV) contain a C-like ORF in the P gene which expresses two C proteins (36), but this ORF is absent in lys- saviruses such as rabies virus.

The C proteins are basic polypeptides, exhibiting little spec- ificity in their intracellular localization (32). Due to their minimal representation in virus particles, they were first described as nonstructural proteins (26), but they have since been detected in virions in association with nucleocapsids by use of specific antibodies (38). Recombinant VSV and measles virus unable to express C protein have been prepared, and both of these viruses exhibit normal growth in cell cultures, suggesting that C has no essential role in virus multiplication under these conditions (24, 33). For SeV, there is evidence that C acts as an inhibitor of RNA synthesis. When RNA synthesis is studied in vitro by the addition of core nucleocapsids to transfected cell extracts containing P and L (and N), the abrogation of C protein expression leads to a strong increase in mRNA synthesis from nondefective templates and in the replication of either DI-E307 or DI-H4 (9) (see Fig. 1 for a description of these RNAs). When genome replication initiated from cDNA copies of the DI genomes is studied in transfected cells, C coexpression inhibits DI-E307 amplification but not that of the copy-back DI-H4. As copy-back DI RNA (H4 RNA) and an internal deletion DI RNA (E307 RNA) differ essentially by

* Corresponding author. Mailing address: Department of Genetics and Microbiology, University of Geneva Medical School, CMU, 9 Ave.

de Champel, 1211 Geneva 4, Switzerland. E-mail: Laurent.Roux

@medecine.unige.ch.

9588

(3)

which promoter sequences are present at P^L(le¹and le²for, respectively, E307 and H4), C appears to act as a promoter- specific repressor of RNA synthesis from P^L in transfected cells and as a general repressor in transfected cell extracts (1).

Further, in a recombinant SeV recovery system, C expression prevents the rescue of a normal infectious virus (FL3, containing the same promoter arrangement as E307) but allows the rescue of FL3-GP42, a nondefective virus whose promoter features resemble, at least partially, those of H4 (15).

From the data presented here, it is proposed that the C proteins act by increasing the selectivity of the replicating com- plex for the cis sequence elements governing its activity and, by doing so, limit the extent of replication. This interpretation ascribes a role to the C proteins and accounts for their apparent paradoxical effect.

MATERIALS AND METHODS

Virus and cells.Cells were grown in regular minimal essential medium (MEM) supplemented with 5% fetal calf serum (FCS) under a 5% CO2atmosphere.

Recombinant vaccinia virus expressing the T7 RNA polymerase, vTF7-3, a gift from Bernard Moss (National Institutes of Health, Bethesda, Md.), has been described by Fuerst and colleagues (13) and was used accordingly. vTF7-3 stocks were grown in HeLa cells to titers ranging from 5310⁸to 5310⁹PFU/ml.

Sequence and plasmids.The complete SeV RNA primary sequence (15,384 nucleotides) is taken from Shioda et al. (34, 35) and Neubert et al. (29). Plasmids expressing the SeV N (previously named NP), P plus C, P without C, C, and L proteins, i.e., pGem-N, pGem-P/C (here PC^wt), pGem-PC^stop, pGem-C9/C/

Y1/Y2 (here pGem-C), and pGem-L, respectively, under the control of the T7 RNA polymerase have been described previously (6, 9, 20, 22).

The cloning of the natural SeV DI RNAs H4 and E307 in pSP65 (pSV-DIH4 and pSV-E307, respectively) under the control of the T7 RNA polymerase promoter has been described previously (2, 4, 12). Their characteristics relative to those of the viral RNA are shown in Fig. 1. The DI-H4 and DI-E307 RNA derivatives generated by fusion PCR (17) and shown in Fig. 2 have been described before (1, 37). These contain reciprocal exchanges between each other’s P^L. They are identified as having an H4 or an E307 backbone, with subsequent information about the P^Lexchanges. For convenience, the 39end of the viral genome is called the genomic promoter (GP), and the 39end of the antigenome is called the antigenomic promoter (AGP). In the case of H4 RNA, which contains le²(AGP) on both the minus- and the plus-strand RNAs, the distinction between the two promoters is made by noting AGP(2) or AGP(1). Because in the data presented here only the minus-strand 39end has been substituted, the (2) has been omitted. The promoter accepting the substitution is indicated, as is the range of nucleotides from the donor promoter. For example, H4-AGP36 indicates an H4 RNA derivative where the first 36 nucleotides of the minus- strand 39end (AGP) have been substituted with the corresponding nucleotides from the E307 RNA minus-strand 39end. Conversely, E307-GP48/98 indicates an E307 RNA whose minus-strand 39end (GP) nucleotides 48 through 98 have been substituted with the corresponding nucleotides from the H4 AGP (see also the diagrams in Fig. 2).

Replication system.The replication system has been described previously (4).

In brief, CV1 cells (about 107) were infected (multiplicity of infection of 3) with vTF7-3. One hour later, in the basic reaction, the cells were transfected with 5mg of pGem-N, 5mg of pGem-PC^wt(or pGem-PC^stop), 1mg of pGem-L, and 5mg of the plasmid carrying the DI RNA sequence (transfection mixture) by incubation (3 h at 33°C) with 3 ml of MEM containing the plasmids mixed with 20ml of Transfectase (Bethesda Research Laboratories). Seven milliliters of MEM containing 2% FCS was then added. Cytoplasmic extracts were prepared 40 h later, and CsCl gradient-purified nucleocapsids were analyzed by Northern blotting with methods and³²P-labelled riboprobes of positive or negative polarity described before (28, 37). The signals were routinely quantitated by Phosphor- Imager scanning (Molecular Dynamics). When transfection of the template RNA-carrying plasmid was delayed (preloading experiments), the medium was removed at various times and replaced with 3 ml of transfection mixture containing 5mg of the template plasmid. After overnight incubation at 33°C, 7 ml of MEM containing 2% FCS was added for the rest of the incubation period (a total of 43 h). The cells were then processed as described above.

Limited primer extension.The sequence analysis of the editing site was done as described before (16), except that no PCR amplification was done in the present study. Limited primer extension was done with reverse transcriptase directly on purified nucleocapsid RNA by use of the³²P-labelled primer SeV edit (19), complementary to the sequence immediately downstream from the editing site.

Western blot analysis.Aliquots (about 1/100 and/or 1/600) of the cytoplasmic extracts prepared for the nucleocapsid RNA analysis (see above) were set aside for protein analysis by Western blotting. This analysis was performed as described previously (6) with mouse-derived antibodies against L, P, N, and C proteins (9, 11), at 1:8,000, 1:10,000, 1:50,000, and 1:1,000 dilutions, respectively.

The blots were developed with a goat anti-mouse antibody conjugated to per- oxidase by use of an enhanced chemoluminescence substrate (Amersham) at a 1:4,000 dilution. The exposed film was scanned, and the signals were quantitated by the program ONE-Dscan (Scanalytics).

RESULTS

A selective cis sequence target for the C inhibitory effect?

We monitored the effect of SeV C on the replication efficiency of DI RNAs in a cell transfection assay in which viral replicons (DI RNAs) and the viral functions required for their replication (N, P, C or no C, and L proteins) were expressed from pGem plasmids and the T7 RNA polymerase was generated by concomitant infection with a recombinant vaccinia virus. The replicon transcripts, made as positive strands with exact 59(T7 polymerase start) and 39(hepatitis delta virus ribozyme cleav- age) viral ends, were first encapsidated by the viral N protein.

These plus-strand nucleocapsids then served as templates for the SeV RNA replicase (P, L, and N), leading to the pro- duction of minus-strand nucleocapsids. Multiple rounds of amplification resulted in the accumulation of genome and antigenome nucleocapsids. The extent of replication was monitored by Northern blotting for minus-strand nucleocapsid RNAs, which were made only by the SeV replicase; a reaction in which pGem-L was omitted controlled for this. This assay proved to be useful and reliable in our experiments. The requirement for the SeV template DI RNA to be a multiple of 6 nucleotides (the “rule of six”) was demonstrated in this way (3), as were other aspects of the SeV multiplication cycle, including the partial characterization of the cis signals involved in transcription, replication, and mRNA editing, the structure- function analysis of the P and the M proteins, the promoter- specific action of C and, last but not least, the recovery of infectious viruses from cDNA (1, 4, 15, 19, 27, 30, 37). An advantage of this assay is the ability to adjust the amount and the nature of the viral components. For example, either a wild-type P gene (PC^wt), in which the C proteins are coexpressed along with the P protein, or one containing a stop codon just downstream of the Y2 protein AUG²⁰¹(PC^stop) can be used (4, 37). The removal of the C proteins from the assay allowed the determination of the C inhibitory effect (see introduction and Cadd et al. [1]).

To more precisely characterize the cis-acting sequence tar- get for the inhibitory effect of C, we examined chimeric DI- E307 genomes, in which various amounts of the normal le¹ sequences at P^L were replaced with the equivalent le² se- quences, as well as DI-H4 genomes, in which the copy-back le² sequences at P^L were replaced with the equivalent le¹ sequences (for a schematic presentation of the SeV RNA features, see Fig. 1). Consistent with previous results (1), DI-H4 RNA replication was insensitive to the presence of the C proteins (Fig. 2A, H4, with or without C). On the other hand, there was a 7.5-fold-lower accumulation of DI-E307 RNA when C was coexpressed (Fig. 2A, E307). DI-E307 RNA also accumulated to lower levels than DI-H4 RNA in the absence of C coexpression, suggesting that the le²promoter is inher- ently more efficient than the le¹promoter at P^L. When the first 36 or 48 nucleotides (nt) of the le²sequence at P^Lof H4 were replaced by the equivalent le¹sequence, genome accumulation was inhibited about four- and ninefold, respectively, in the presence of C (Fig. 2A). When nt 24 to 55 were exchanged, genome accumulation was inhibited fivefold (Fig. 2B). Simi- larly, when the first 24, 33, or 48 nt of the le¹sequence at P^L of E307 were replaced by the equivalent le² sequences, genome accumulation was progressively less inhibited (from 7.5- to 2-fold) in the presence of C, and when nt 24 to 55 were exchanged, genome accumulation was reduced from 7.5-fold to

(4)

5.4-fold (Fig. 2). On other hand, when nt 48 to 98 of the le¹ region sequence at P^Lof E307 were replaced by the equivalent le² region sequence, genome accumulation continued to be inhibited in the presence of C (Fig. 2A). Thus, for both E307 and H4, the larger the le¹sequences at P^L, the stronger the inhibitory effect of C. Conversely, the larger the le²sequences at P^L, the weaker the inhibition. The target for the C inhibitory effect thus appears to be confined to and to extend throughout the le¹sequences at P^L.

H4 RNA is sensitive to the C inhibitory effect.Whereas H4 replication is insensitive to C inhibition in transfected cells, it is highly sensitive (up to 20-fold) to C coexpression in transfected cell extracts (see introduction and Cadd et al. [1]). The discrepancy may be explained by the difference between the

two assays. In the cell extract assay, the H4 template was added to a set of proteins which had accumulated for 24 h in the transfected cells. In the transfected cell assay, all the plasmids were transfected simultaneously, so that the levels of the viral proteins built up in concert with those of the templates. We therefore examined the effect of pretransfecting pGem-N, pGem-PC^wt, and pGem-L, followed by pSV-DIH4 transfection, on the intracellular accumulation of H4 (Fig. 3). When the pSV-DIH4 transfection was delayed by 1.5 h, there was a threefold difference in replication efficiency between the conditions in the presence or absence of C (Fig. 3B), and this difference increased to seven- to eightfold after a delay of 8 to 11 h. Preloading cells with C (along with N, P, and L) thus led to the inhibition of H4 replication in the transfected cells.

FIG. 1. SeV RNAs The SeV full-length RNAs are shown schematically in the conventional way, i.e., coding strand (1) 59to 39, from left to right. The minus-strand RNA (line 2) is depicted with its two promoter regions at left and at right, P^Land P^R, next to the le¹and le²template regions, respectively. The transcription products from these templates are shown below as the le¹RNA and the six mRNAs corresponding to the six transcription units. The features of the internal deletion DI E307 RNA are shown below, including its ability to encode only a single N/L mRNA due to the deletion of all the sequences in between. Note at P^Land P^Rthe same promoter regions as are present in the viral genome. In contrast, in the DI H4 RNA, the copy-back feature results in the presence of the same le²promoter region at P^Land P^R. Due to the lack of a transcription signal, no mRNA is produced from this type of DI RNA. nt, nucleotides.

(5)

FIG. 2. Target sequence for the C effect. CV1 cells (5310⁶) were infected with vTF7-3 and transfected with the plasmids pGem-N, pGem-PC^wtor pGem-PC^stop (1or2in C column, respectively), and pGem-L and the template plasmids listed on the left as described in Materials and Methods. At 43 h later, the cells were collected, and the encapsidated RNAs were analyzed by Northern blotting with a probe of positive polarity (see Materials and Methods and Results). The asterisks in the C column indicate samples from which the L plasmid was omitted. The C inhibitory effect was calculated by dividing the signals obtained in the absence of C (lanes2) by those obtained in the presence of C (lanes1). On the right, a graphic representation of the various RNA templates is shown as minus-strand RNAs, 39 to 59, from left to right. Except for the two parents, the H4 and E307 RNAs (uppermost and lowermost in panel A, respectively), for which the whole sequence is shown, only the minus-strand 39end (P^L) is shown to emphasize the substitutions made (dark gray box, H4 P^L; light gray box, E307 P^L). The replication efficiency was evaluated by PhosphorImager scanning of the Northern blots.

(6)

Preloading cells with C could have two effects: (i) the level of C relative to that of templates at the early stages of amplification will be increased, and (ii) C will have been present intracellularly for a longer time before replication begins and therefore will be more subject to possible modification. To distinguish between these two possibilities, we transfected all the plasmids simultaneously, but with decreasing amounts of pSV-DIH4 (Fig. 4). Under these conditions, only the ratio of C to genome templates will vary. As before, when 5mg of pSV- DIH4 was transfected, DI genome accumulation was unaf- fected by C (Fig. 4, lanes 1 and 2). However, as progressively less pSV-DIH4 was transfected, the inhibition of H4 replication in the presence of C became progressively stronger (compare lanes 3 and 4, lanes 5 and 6, and lanes 9 and 10 with lanes 7 and 8 and lanes 11 and 12). A mean C inhibitory effect of ca.

4-6-fold was apparent at the lowest pSV-DIH4 amounts transfected. The experiments in Fig. 3 and 4 are thus consistent with the notion that an abnormally high ratio of C to templates results in the inhibition of H4 RNA replication.

Inhibition depends on the concentration of C proteins.The above-described results suggest that the inhibition of DI-H4

replication is dependent on the concentration of C. To further examine this idea, we preloaded cells with pGem-N, pGem-L, and pGem-PC^stop, as well as with increasing amounts of a plasmid expressing only the C proteins (pGem-C) or increasing amounts of pGem (with no insertions) as a control. Figure 5A shows that the extent of H4 replication was indeed decreased in proportion to increasing amounts of pGem-C (compare lane 2 with lanes 3, 5, and 7). Figure 5B shows that increasing amounts of pGem-C led to increasing levels of C intracellularly. The level of L protein, however, was unexpectedly reduced by eightfold at the higher plasmid concentrations (lanes 7 and 8), and this result appeared to be due to a nonspecific effect of excess plasmid transfection. When DI-H4 replication in the presence of increasing amounts of pGem-C (lanes 3, 5, and 7) was examined relative to that in the presence of increasing amounts of pGem (lanes 4, 6, and 8), the additional transfection of 1, 2.5, and 5mg of pGem-C resulted in 2-, 15-, and 20-fold inhibition, respectively. The inhibition of H4 replication clearly responded to C protein concentrations (Fig. 5C).

C inhibition can be reversed by increasing P but not L.The ability of the C proteins to inhibit transcription in cell extracts was strongest when C was coexpressed with P and L. C could interact with P to exert its effect, since P overexpression partially reversed C inhibition (9). Figure 6 shows that this coun- tering effect of excess P could be reproduced in transfected cells. In Fig. 6, lanes 4 and 5, P was increased intracellularly by the addition of pGem-PC^stop with pGem-PC^wt, resulting in about ninefold better replication (compare with lane 6). In contrast, an increase in L (lanes 2 and 3) could not relieve C inhibition. The inhibitory effect of C was relieved with an increase in P of less than threefold, while a fivefold increase in L was ineffective (Fig. 6B). An expression of the relative protein ratios, with the amounts of C as the common denominator, showed the correlation between a high P/C ratio (but not a high L/C ratio) and the reversal of the C inhibitory effect (Fig.

6B, bottom, numbers in parentheses).

C and the rule of six. C inhibition appears strongest on templates which replicate less well. H4 replicons which are not of hexamer length replicate very poorly. Although these RNAs can be shown to be amplified marginally above the background (no L protein present) in the presence of C, they generally are amplified ca. 50- to 100-fold less efficiently than H4 RNAs of exactly hexamer length (3, 27). The effect of C on the relative replication of H4 RNA (1,410, or 23536, nt long) and H411 RNA, i.e., H4 RNA 6n11 nt long (1,411 nt), is shown in Fig.

7. As before, H411 replicated 50-fold less efficiently than H4 (1,410 nt) in the presence of C. This difference was reduced to 10-fold in the absence of C (compare lanes 5 and 6 with lanes 9 and 10). This result was reproducibly observed not only for H411 but for the whole series of derivatives which violate the rule (H412, H413, and so forth; data not shown). This result could mean that hexamer length is less important in the absence of C or that the genomes are corrected to hexamer length during replication. Genome length correction has been shown to occur at the editing site in the P gene (a site prone to RNA polymerase stuttering) during antigenome synthesis (16).

H4 RNA does not contain this site but retains the L gene polyadenylation site, at which RNA polymerase stuttering also occurs. Limited primer extension through this sequence, however, found no evidence of genome length correction at this site (data not shown).

Rather than sequencing the entire H411 RNA in search of a possible length correction, we examined the effect of C on genome length correction by introducing the polypurine run (A6G3) of the P gene editing site into DI-H4, creating H4ed and H4ed11 RNAs. The replication of H4ed11 RNA in the

FIG. 3. Effect on H4 replication of preloading the cells with C. CV1 cells were infected and transfected as described in the legend to Fig. 2 (6C conditions), except that the addition of plasmid pSV-DIH4 was delayed as indicated.

(A) Northern blot analysis of the replicated nucleocapsid H4 RNA following 43 h of total incubation (as in Fig. 2). (B) Graphic representation of the C inhibitory effect as a function of the delay in plasmid pSV-DIH4 addition. Replication efficiency was evaluated as described in the legend to Fig. 2.

(7)

absence of C was previously shown to result in a 1-base deletion at the editing site (16). In the presence of C, H4ed was found to replicate ca. threefold less efficiently than H4, presumably as a result of this insertion (Fig. 7A); consequently, the difference between the H4ed and H4ed11 amplification levels was only ca. eightfold (compare lanes 7 and 8). In the absence of C, this difference was reduced to ca. twofold (compare lanes 11 and 12), due in large part to a higher level of replication of H4ed11. Limited primer extension across the editing site (Fig. 7B) showed that genome length correction took place in the presence of C (16), since ca. 30% of the accumulated H4ed11 RNAs had a 1-nt deletion (Fig. 7B, Corrected). In the absence of C,.90% of the accumulated genomes exhibited a corrected length. Thus, either the prob- ability of genome length correction increased in the absence of

C or there was a preferential amplification of the corrected genome due to the increased rounds of replication occurring in the absence of C (Fig. 6A) or both. In either case, genome length correction was more efficient in the absence of C. Note here that the intensity of the signals in Fig. 7B (under H4ed11) did not reflect the relative level of RNA amplification, since ninefold more cells were used in the presence of C to compensate for the lower level of replication. The relative level of replication could however, be noted from Fig. 7A.

C changes the flexibility of the replicating complex.The first interpretation of the rule of six, based on the suggestion that each N subunit contacts exactly 6 nt (10), postulates that the cis-acting sequence for replication at the 39end of the template must be exactly covered by the N protein subunit (3). This postulate indicates that the nucleotides are seen in the context

FIG. 4. Effect of decreasing the amount of H4 template plasmid. (A) CV1 cells were infected and transfected as described in the legend to Fig. 2 (simultaneous transfection of all the plasmids), but with gradually decreasing amounts of the replicon plasmid pSV-DIH4 as indicated. For 0.1 and 0.05mg of H4 plasmid, duplicate samples are shown (lanes 5 through 12). The replication efficiency, estimated by PhosphorImager scanning, is expressed relative to that obtained with 5mg of plasmid pSV-DIH4, arbitrarily taken as 100. (B) Extent of replication plotted as a function of the amounts of plasmid pSV-DIH4 added. Error bars are shown when applicable.

(8)

of the N subunit. The notion was then introduced that the six contacts between an N subunit and the 6 nt that it covers are not equivalent. For instance, a nucleotide normally in contact site 2 of the subunit is seen by the polymerase differently when positioned in contact site 3 or 4; i.e., the subunit imposes a phase on the sequence. This notion came from a series of constructs derived from E307 (E307A RNAs; see below) in which various insertions or deletions in P^L were made. The

FIG. 5. Effect of increasing the amounts of C proteins. CV1 cells were infected and transfected as described in the legend to Fig. 2 (plasmid pSV-DIH4 transfection delayed by 8.5 h), except that the C proteins were expressed (by transfection at time zero) from increasing amounts of plasmid pGem-C as indicated. (A) Northern blot analysis of the replicated nucleocapsid H4 RNA following 43 h of total incubation. In lanes 4, 6, and 8, pGem was added as a control for excess plasmid addition (see Results). (B) Western blot analysis of aliquots (1/200) of the cytoplasmic extracts prepared from the material tested in panel A.

The amounts of each protein, estimated by densitometer scanning (ONE- Dscan), are not absolute; they can only be compared within the same blot, since these proteins are detected by different antibodies (see Methods) and visualized after different exposure times. (C) Replication efficiency, determined as described in the legend to Fig. 4, plotted as a function of the amounts of plasmids pGem-C and pGem added.

FIG. 6. Reversal of C inhibitory effect by P but not by L. CV1 cells were infected and transfected as described in the legend to Fig. 3 (plasmid pSV-DIH4 transfection delayed by 8.5 h), except for the amounts of supporting plasmids, which varied from the standard as indicated. (A) Northern blot analysis of the replicated H4 RNA (lanes 2 and 3 and lanes 4 and 5 are duplicate samples).

Replication efficiency was determined as described in the legend to Fig. 4. (B) Western blot analysis. For each of the analyzed samples, two different amounts (13is 1/200 and 63is 6/200 of the total extract) were loaded on the gel. Relative ratios shown by numbers in parentheses, where C is the common denominator, illustrate the relevant pairs of proteins responsible for the reversal of the C inhibitory effect.

(9)

replication ability of these derivatives suggested that (i) the polymerase made multiple base contacts over at least 77 nt in the promoter region for initiation, (ii) that these bases were likely to be viewed in a certain context with regard to the N subunit, and (iii) that these contact sites may have been dis- continuous (30). These conclusions were drawn from results obtained without C coexpression, conditions under which the replicating complex was shown to tolerate variations either in the N subunit context or in the postulated distance between the cis-acting sequences (30).

The same derivatives were analyzed with C coexpression.

Figure 8A shows the derivatives and their sequences presented in hexamer units to highlight the phase context in which the nucleotides were found. In these E307A RNA derivatives, the introduction of two restriction sites (BglII and NsiI) allowed between one and five base insertions downstream of the BglII site to alter the base context. For instance, A-56, which represents nt 1 of the N transcript, progressively moved from its

original position in the hexamer. After a 6-base insertion at this site (16/10), the nucleotide context returned to normal, but all the nucleotides were shifted one hexamer downstream.

Note that, when required, compensatory modifications were made downstream, at the NsiI site, to maintain the hexamer length. From there downstream, then, the context was con- served, although the nucleotides were shifted by one hexamer.

Also noteworthy was the construct10/16, for which the context upstream from nt 66 was not changed in the interval but was shifted by one hexamer from the NsiI site downstream.

The left side of Fig. 8A shows the replication ability of the derivatives, measured in the presence of C and relative to that of E307A, the wild-type control (“wt”) for this experiment. The values are given in Fig. 8B (1C), side by side with those obtained previously in the absence of C coexpression (2C) (30). Figure 8C shows graphically the comparison. It is clear that the presence of C represented nonpermissive conditions for all the derivatives with insertions or deletions, in contrast to

FIG. 7. Effect of C on genome length correction. (A) Identical CV1 cell samples were infected and transfected as described in the legend to Fig. 2 (simultaneous transfection) with the indicated supporting and template plasmids. H411, H4ed, and H4ed11 indicate plasmids described in the text (see Results). Replication efficiency was determined as described in the legend to Fig. 2. (B) Limited primer extension of the purified nucleocapsid RNA was done as described in Materials and Methods. In lane 2 under H4ed11, ninefold more cells were used to start with to compensate for the lower level of replication (see panel A). Corrected and Not corrected indicate, respectively, the 1-nt deletion being present or absent at the editing site.

(10)

FIG. 8. Effect of C on replicons with phase context modifications. (A) CV1 cells were infected and transfected as described in the legend to Fig. 2 (simultaneous transfection) in the presence of C. The replicon RNAs were derived from a modified version of E307 RNA (E307A; see Results) depicted schematically in the first line;

modifications made in the sequence (presented as hexamers) of E307A are shown schematically in the second line (P^L) with the added BglII and NsiI sites. Letter code: lightface lowercase characters, base modifications used to create the two restriction sites in E307A; lowercase bold characters, insertions used to create the derivatives described by the additions at the two sites (10/16,11/15, and so forth); uppercase bold characters, context changes brought about by the insertions. On the left is shown a Northern blot analysis of the replicated RNAs. (B) Replication efficiency expressed as described in the legend to Fig. 2, taken from the Northern blot analysis in panel A and others (data not shown) for the1C conditions and from Pelet et al. (30) for the2C conditions. Note that 100% refers to different absolute replication efficiencies under the two conditions, as shown in Fig. 2A for E307. (C) Graphic comparison of derivative replication efficiencies achieved in the2C and1C conditions.

(11)

the results obtained in the absence of C, where some derivatives were efficiently amplified. For instance, the shift of one hexamer which did not change the context (10/16 and16/10;

see A56) obliterated replication in the presence of C, whereas that in the absence of C was as efficient as wild-type replication. Similar results were obtained when the context was changed (11/15 and15/11;12/22). In conclusion, whatever accommodation is possible in the absence of C appears not to be tolerated with C coexpression.

DISCUSSION

We previously found that the normal coexpression of C along with P and L in transfected cells inhibits the amplification of internal deletion DI-E307 but not that of copy-back DI-H4. As these DI genomes differ essentially by which le region is present at P^L and as the sensitivity of E307 to C inhibition could be relieved by substituting the le¹sequences normally present at P^Lwith the equivalent le²sequences, it appeared that the le²region promoter was insensitive to inhibition by C protein (1). The present study, however, shows that C can act on the le²region promoter as well, provided that (i) the transfected cells are preloaded with C (along with the other viral proteins) before replication starts, (ii) progressively less pSV-DIH4 is transfected into cells, or (iii) concomitant additional expression of C from a separate plasmid is present. In all of these situations, the level of C relative to that of genome templates is increased, especially during the early stages of the amplification, and this increased ratio presumably accounts for the inhibition of DI-H4, which contains the le² region pro- moter at both P^Land P^R. However, upon transfection of large amounts of pSV-DIH4 (5mg), the DI-H4 template concentra- tion presumably increases rapidly, due to cis-acting signals conferring high replication per se and/or because the sequences in pSV-DIH4 are relatively insensitive to the inhibitory effects of C. The amplification of DI-H4 in this situation is thus little affected by the presence of C. For DI-E307, repli- cating poorly due to cis-acting signals on the template, C has time to accumulate so that its inhibitory effect exaggerates the weakness of the promoter, leading to strong inhibition of amplification.

Although the C proteins are the most abundant products of the SeV P gene intracellularly, viral particles contain very few C proteins (32, 38). The work presented here suggests why the underrepresentation of C in virions might be important. In natural infections and in virus recoveries from cDNA (which require the suppression of C coexpression from the pGem-P supporting plasmid [15]), the C proteins slowly accumulate during the course of infection or recovery (8). This accumulation is also coordinated with that of the viral genomes and is thus apparently prevented from reaching ratios at which genome replication is strongly inhibited. On the other hand, when C proteins are allowed to accumulate in advance due to their coexpression from pGem-P, they presumably reach ratios relative to viral genomes which are too inhibitory for wild-type virus to be recovered. The replacement of the normal C-sen- sitive le¹ sequence at P^L with the relatively insensitive le² sequence, however, permits the recovery of this recombinant SeV made into a copy-back nondefective construct (rSeV- GP42) (15), presumably because the le²region promoter is less inhibited by the higher ratios of C proteins to genomes.

The relative exclusion of C protein from SeV virions thus appears to be an important feature of wild-type infectious virus assembly.

The C protein was first characterized as an inhibitor of mRNA synthesis from nondefective genomes and then as a

promoter-specific inhibitor of transcription and replication from P^L. The normal (le¹) replication promoter at P^L has recently been examined by insertion analysis of DI-E307 in the absence of C expression (30). Insertions of 6 nt on either side of the le¹-N junction were well tolerated, but insertions of 12 or 18 nt uniformly prevented amplification of the E307 replicon. In the presence of C coexpression, however, the insertion of even 6 nt at either site was very poorly tolerated (Fig. 8).

Antigenome synthesis from an inserted le¹region promoter is thus particularly sensitive to the presence of C coexpression. In a similar vein, the ability of copy-back DI-H4 genomes, which are not precisely of hexamer length, to be amplified in transfected cells is improved (fivefold) when C coexpression is suppressed (Fig. 7). The stringency of the rule of six also thus appears to depend on C coexpression. In both of the above- mentioned cases, C appears to act selectively to prevent replication from “substandard” promoters. C thus exerts one effect that is modulated by the type of promoter. It appears promoter specific in that it depends on the extent of le¹or le² sequences present at P^L. This effect is also complex, as it is possible to create chimeric promoters of high efficiency which are very sensitive to C inhibition and those of weak efficiency which are insensitive to C inhibition. Inherent promoter strength and promoter sensitivity to C are thus separate properties, difficult to distinguish, however, in an assay which only monitors the results of the two properties. It is only in the absence of C that inherent promoter strength can be estimated.

Precisely how C acts on the replicating complex remains unclear. One reason we know so little is that we are unable to reproduce a key aspect of this inhibition in vitro. When RNA synthesis is examined in transfected cell extracts to which core nucleocapsids (minus P and L) are added, the addition of C protein from a separate extract to a P3-L complex made in the absence of C has no effect on either transcription or replication (9). This is so even though the C added in trans appears to bind to the reconstituted holo-nucleocapsids (i.e., nucleocapsids active for RNA synthesis) normally (in either case, only a small fraction of the C protein is found on holo-nucleocapsids; 1a).

We thus only know that C must be coexpressed with P and L

FIG. 9. How C can add selectivity to the polymerase complex. (a) E307 P^Lis shown as gray vertical bars representing hexamer units exposing three sites (small horizontal rectangles) that the polymerase (lower open box) has to contact for productive initiation. (b) The three contact points are reinforced by C coexpression (1C) to indicate the higher stringency with which the cis elements are recognized. (c) After a hexamer insertion (black vertical bar), site three is displaced to the right, a move that the polymerase can still accommodate (via a possible conformational change shown as a wavy line) in the absence of C. (d) The presence of C somehow prevents this accommodation, and the polymerase can no longer contact the third site, resulting in aborted, initiation.

(12)

in transfected cells for its effect to be seen in cell extracts.

There can be several explanations for this fact, including that C may act indirectly during the formation of the P3-L complex.

Indeed, we have so far been unable to demonstrate the presence of C protein in a complex with either P3 or P3-L. Fur- thermore, when P3-L complexes are isolated from transfected cells on immunobeads and then used to drive viral RNA synthesis in vitro, P3-L assembled in the presence of C protein is just as active as that assembled in the absence of C protein (9a). The inhibitory effect of C on the activity of the polymerase complex is thus reversible and is lost upon purification of the P3-L complex. The ability of C to modulate polymerase activity might occur via low-affinity complexes (if it were di- rect), and this ability might explain the requirement for high intracellular levels of C. We also know that the levels of P protein are important for inhibition, since the inhibitory effect of C can be reversed by the overexpression of P (but not L) (Fig. 6), consistent with what has been observed with cell extracts (9). Finally, since C is as inhibitory for mRNA transcription as it is for genome replication (in vitro), we assume that C acts on the P3-L polymerase rather than on the P3-N assembly complex during genome replication.

Even though more specific information is lacking (and Fig. 9 reflects this fact), it is possible to visualize one way in which C might exert its effect. The template in Fig. 9 is shown as a horizontal stack of vertical bars representing N subunits, each associated with 6 nt which are not shown. Three cis-acting sequences which interact with the polymerase for initiating genome replication are shown as horizontal dark gray rectangles on the lower surface of the template. The polymerase is shown simply as an open box below; its P and L subunits are not indicated. The E307 replicon can tolerate a 6-nt insertion near the le¹-N junction (black bars in Fig. 9c and d) when C coexpression is suppressed but not the insertion of 12 or 18 nt.

This finding suggests that a cis-acting element important for the initiation of replication was found beyond the le¹ sequences, in the 59 untranslated region of the N gene. Since more than one cis-acting element within the le¹ sequences appears to be important (and two are drawn in Fig. 9) (30), it is possible that these elements are separated by 6 nt because the polymerase can alter its conformation sufficiently only when 6 (and not 12 or 18) nt are inserted between the last two elements (Fig. 9c). In Fig. 9d, the 6-nt insertion can no longer be accepted when C is coexpressed, because the polymerase cannot alter its conformation sufficiently to accommodate the last cis-acting element. This constraint on polymerase confor- mation causes it to become more selective with regard to the various promoters. C-insensitive promoters would thus be those requiring less conformational change.

By generating more selectivity as C proteins accumulate during infection, C might regulate the relative activities of the replication (and transcription?) promoters and would be expected to drive genome synthesis from antigenome templates.

In accord with this idea, SeV^MVC, the only natural mutant with mutations mapped to the C gene and whose C protein has mostly lost its ability to inhibit RNA synthesis in vitro (1), has an overtranscription phenotype in cell cultures (18) but does not accumulate more antigenomes, as might be expected (15a).

We can foresee another way that increased selectivity can be beneficial to the virus. Although very little nongenomic RNA is assembled into nucleocapsids in most cell culture infections, some infections lead to the encapsidation of significant amounts of viral mRNAs (15a). A more selective replicating complex would be less likely to engage these bogus templates.

Finally, the SeV C protein might have other functions of which

we are unaware, which might explain the presence of the C gene in so many paramyxoviruses.

In contrast to recoveries of infectious SeV from DNA, for which C gene expression must be suppressed, recoveries of respiratory syncytial virus (RSV) from DNA require the expression of the 22-kDa protein gene in addition to those of N, P, and L. In further contrast, the RSV 22-kDa protein exerts a positive effect on RNA synthesis as a processivity factor for mRNA synthesis (5). Given our limited understanding of the SeV C proteins, it is possible that they exert their negative effect on the polymerase by modulating its processivity, but we have no evidence for this suggestion. In both SeV and RSV, a polymerase cofactor which modifies polymerase activity is apparently required for virus viability; in contrast to the situation for measles virus and VSV, we have so far been unable to prepare recombinant SeV without the ability to express all of its C proteins. Recombinant SeV which cannot express either C9or C can, however, be prepared (26b), and its phenotypes are under investigation. Furthermore, a point mutation in the SeV C proteins which relieves their inhibitory effects on viral RNA synthesis has recently been found also to attenuate virulence for mice by preventing multiple rounds of virus infections. The SeV C proteins may thus also play a role in modulating the antiviral effects of the innate immune system of the mouse (14, 21).

The need to add selectivity to the RNA synthesis complex, however, could be an essential function that accounts for the failure to produce virus lacking all the C proteins. A nonselec- tive replicating complex could be trapped on nonfunctional templates to the point at which the infection aborts. What appears then an inhibitory property becomes the price to pay for selectivity. This selectivity may have evolved with the pres- sure to control a process such as mRNA editing. This, in turn, requires the rule of six to prevent the consequence of nucleotide addition during replication.

Finally, this reverse genetic system with minireplicons, as complex as it appears for the coinfection of recombinant vaccinia virus, has nevertheless provided in vivo data that corre- late with in vitro results produced before (9). The interpretation of the role of C given here also fits with the inability to rescue recombinant SeV expressing none of the C proteins, suggesting that the data obtained with this system are also relevant for SeV infection. This fact is noteworthy, since for VSV the effect of C on in vitro transcription (31) was not verified when the virus lacking C was produced (24).

ACKNOWLEDGMENTS

We are indebted to Dominique Garcin for advice to C. Tapparel as well as for the vivid but worthwhile criticisms that he raised about the interpretation of the data.

This work was supported by grants from the Swiss National Foun- dation for Scientific Research to D.K. and L.R. and from the de Reuter Foundation and the Socie´te´ Acade´mique de Gene`ve to L.R.

REFERENCES

1. Cadd, T., D. Garcin, C. Tapparel, M. Itoh, M. Homma, L. Roux, J. Curran, and D. Kolakofsky.1996. The Sendai paramyxovirus accessory C proteins inhibit viral genome amplification in a promoter-specific fashion. J. Virol.

70:5067–5074.

1a.Cadd, T. Unpublished data.

2. Calain, P., J. Curran, D. Kolakofsky, and L. Roux. 1992. Molecular cloning of natural paramyxovirus copy-back defective interfering RNAs and their expression from DNA. Virology 191:62–71.

3. Calain, P., and L. Roux. 1993. The rule of six, a basic feature for efficient replication of Sendai virus defective interfering RNA. J. Virol. 67:4822–4830.

4. Calain, P., and L. Roux. 1995. Functional characterisation of the genomic and antigenomic promoter of the Sendai virus. Virology 211:163–173.

5. Collins, P. L., M. G. Hill, J. Cristina, and H. Grosfeld. 1996. Transcription elongation factor of respiratory syncytial virus, a nonsegmented negative-

(13)

strand RNA virus. Proc. Natl. Acad. Sci. USA 93:81–85.

6. 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.

7. Curran, J., and D. Kolakofsky. 1989. Scanning independent ribosomal ini- tiation of the Sendai virus Y proteins in vitro and in vivo. EMBO J. 8:521–

8. Curran, J., and D. Kolakofsky. 1990. Sendai virus P gene produces multiple526.

proteins from overlapping open reading frames. Enzyme 44:244–249.

9. Curran, J., J.-B. Marq, and D. Kolakofsky. 1992. The Sendai virus nonstruc- tural C proteins specifically inhibit viral mRNA synthesis. Virology 189:647–

9a.Curran, J. Unpublished data.656.

10. Egelman, E. H., S.-S. Wu, M. Amrein, A. Portner, and G. Murti. 1989. The Sendai virus nucleocapsid exists in at least four different helical states.

J. Virol. 63:2233–2243.

11. Einberger, H., R. Mertz, P. H. Hofschnieder, and W. J. Neubert. 1990.

Purification, renaturation, and reconstituted protein kinase activity of the Sendai virus Large (L) protein. J. Virol. 64:4274–4280.

12. Engelhorn, M., R. Stricker, and L. Roux. 1993. Molecular cloning and characterization of a Sendai virus internal deletion defective RNA. J. Gen.

Virol. 74:137–141.

13. Fuerst, T. R., E. G. Niles, F. W. Studier, and B. Moss. 1986. Eukaryotic transient-expression system based on recombinant vaccinia virus that syn- thesizes bacteriophage T7 RNA polymerase. Proc. Natl. Acad. Sci. USA 83:8122–8126.

14. Garcin, D., M. Itoh, and D. Kolakofsky. A point mutation in the Sendai virus accessory C proteins attenuates virulence for mice, but not virus growth in cell culture. Virology, in press.

15. Garcin, D., T. Pelet, P. Calain, L. Roux, J. Curran, and D. Kolakofsky. 1995.

A highly recombinogenic system for the recovery of infectious Sendai paramyxovirus from cDNA: generation of a novel copy-back nondefective interfering virus. EMBO J. 14:6087–6094.

15a.Garcin, D. Unpublished data.

16. Hausmann, S., J. P. Jacques, and D. Kolakofsky. 1996. Paramyxovirus RNA editing and the requirement for hexamer genome length. RNA 2:1033–1045.

17. Ho, S. N., H. D. Hunt, R. M. Horton, J. K. Pullen, and L. R. Pease. 1989.

Site-directed mutagenesis by overlap extension using the polymerase chain reaction. Gene 77:51–59.

18. Itoh, M., and M. Homma. 1993. A mechanism of adaptation of Sendai virus to LLC-MK2 cells; single point mutation of C protein elevated the level of viral mRNA, abstr. 74. In Abstracts of the IXth International Congress of Virology.

19. Jacques, J. P., S. Hausmann, and D. Kolakofsky. 1994. Paramyxovirus mRNA editing leads to G deletions as well as insertions. EMBO J. 13:5496–

5503.

20. Kast, W. M., L. Roux, J. Curran, H. J. J. Blom, A. C. Voordouw, R. H.

Meloen, D. Kolakofsky, and C. J. M. Melief.1991. Protection against lethal Sendai virus infection by in vivo priming of virus-specific cytotoxic T lym- phocytes with a free synthetic peptide. Proc. Natl. Acad. Sci. USA 88:2283–

2287.

21. Kato, A., K. Kiyotani, K. Sakai, T. Yoshida, and Y. Nagai. 1997. Importance of the V protein for determining in vitro phenotypes and in vivo pathoge- nicity of Sendai virus. EMBO J. 16:678–687.

22. Keene, J. D., I. M. Chien, and R. A. Lazzarini. 1981. Vesicular stomatitis defective interfering particle containing a mutated, internal leader RNA gene. Proc. Natl. Acad. Sci. USA 78:2090–2094.

23. Kolakofsky, D., S. Vidal, and J. Curran. 1991. Paramyxovirus RNA synthesis and P gene expression, p. 215–233. In D. W. Kingsbury (ed.), The paramyxo- viruses. Plenum Press, New York, N.Y.

24. Kretzschmar, E., R. W. Peluso, M. J. Schnell, M. A. Whitt, and J. K. Rose.

1996. Normal replication of vesicular stomatitis virus without C proteins.

Virology 216:309–316.

25. Lamb, R. A., and D. Kolakofsky. 1996. Paramyxoviridae: the viruses and their Replication, p. 1177–1204. In B. N. Fields, D. M. Knipe, and P. M. Howley (ed.), Field’s virology.

26. Lamb, R. A., and R. G. Paterson. 1991. The nonstructural proteins of paramyxoviruses, p. 181–214. In D. W. Kingsbury (ed.), The paramyxovi- ruses. Plenum Press, New York, N.Y.

26a.Latorre, P. Unpublished data.

26b.Latorre, P., and J. Curran. Unpublished data.

27. Mottet, G., A. Mu¨hlemann, C. Tapparel, F. Hoffmann, and L. Roux. 1996. A Sendai virus vector leading to the efficient expression of mutant M proteins interfering with virus particle budding. Virology 221:159–171.

28. Mottet, G., and L. Roux. 1989. Budding efficiency of Sendai virus nucleo- capsids: influence of size and ends of the RNA. Virus Res. 14:175–188.

29. Neubert, W. J., C. Eckerskorn, and H. E. Homann. 1991. Sendai virus NP gene codes for a 524 amino acid NP protein. Virus Genes 5:25–32.

30. Pelet, T., C. Delenda, O. Gubbay, D. Garcin, and D. Kolakofsky. 1996.

Partial characterization of a Sendai virus replication promoter and the rule of six. Virology 224:405–414.

31. Peluso, R. W., J. C. Richardson, J. Talon, and M. Lock. 1996. Identification of a set of proteins (C9and C) encoded by the bicistronic P gene of the Indiana serotype of vesicular stomatitis virus and analysis of their effect on transcription by the viral RNA polymerase. Virology 218:335–342.

32. Portner, A., K. C. Gupta, J. M. Seyer, E. H. Beachey, and D. W. Kingsbury.

1986. Localization and characterization of Sendai virus monostructural C and C9protein antibodies against synthetic peptides. Virus Res. 6:109–121.

33. Radecke, F., and M. A. Billeter. 1996. The nonstructural C protein is not essential for multiplication of Edmonston B strain measles virus in cultured cells. Virology 217:418–421.

34. Shioda, T., Y. Hidaka, T. Kanda, H. Shibuta, A. Nomoto, and K. Iwasaki.

1983. Sequence of 3,687 nucleotides from the 39end of Sendai virus genome RNA and the predicted amino acid sequences of viral NP, P and C proteins.

Nucleic Acids Res. 11:7317–7330.

35. Shioda, T., K. Iwasaki, and H. Shibuta. 1986. Determination of the complete nucleotide sequence of the Sendai virus genome RNA and the predicted amino acid sequences of the F, HN, and L proteins. Nucleic Acids Res.

14:1545–1563.

36. Spiropoulou, C. F., and S. T. Nichol. 1993. A small highly basic protein is encoded in overlapping frame within the P gene of vesicular stomatitis virus.

J. Virol. 67:3103–3110.

37. Tapparel, C., and L. Roux. 1996. The efficiency of Sendai virus genome replication: the importance of the RNA primary sequence independent of terminal complementarity. Virology 225:163–171.

38. Yamada, H., S. Hayata, T. Omata-Yamada, H. Taira, K. Mizumoto, and K.

Iwazaki.1990. Association of the Sendai virus C protein with nucleocapsids.

Arch. Virol. 113:245–253.