Identification of Internal Sequences in the 3' Lea
http://www.100md.com
病菌学杂志 2005年第4期
Division of Pathology and Neuroscience, University of Dundee, Dundee, United Kingdom
Laboratory of Infectious Diseases, National Institute of Allergy and Infectious Diseases, Bethesda, Maryland
ABSTRACT
Previous studies of respiratory syncytial virus have shown that the 44-nucleotide (nt) leader (Le) region is sufficient to initiate RNA replication, producing antigenome RNA, and that the Le and adjoining gene start (GS) signal of the first gene are sufficient to initiate transcription, producing mRNA. A cis-acting element necessary for both transcription and replication was mapped within the first 11 nt at the 3' end of Le. In the present study the remainder of the Le region was mapped to identify sequences important for transcription and replication. A series of minigenomes with mutant Le sequences was generated, and their ability to direct transcription and replication was determined by Northern blot analysis, which examined full-length antigenome and mRNA, and by primer extension analysis, which examined antigenome and mRNA initiation. With regard to transcription, nt 36 to 43, located immediately upstream of the GS signal, were found to be necessary for optimal levels of mRNA synthesis, although the GS signal in conjunction with the 3'-terminal region of Le was sufficient to direct accurate mRNA synthesis initiation. With regard to replication, the first 15 nt of Le were found to be sufficient to direct initiation of antigenome synthesis, but nt 16 to 34 were required in addition for efficient encapsidation and production of full-length antigenome. Analysis of transcripts produced from di- and tricistronic minigenomes indicated that a significant proportion of abortive replicases continue RNA synthesis to the end of the first gene and then continue in a transcription mode along the remainder of the genome.
INTRODUCTION
Human respiratory syncytial virus (RSV) is the major cause of pediatric respiratory disease worldwide and is increasingly recognized as an important pathogen in the elderly population (10). RSV is a member of the subfamily Pneumovirinae, of the family Paramyxoviridae within the order Mononegavirales. Vesicular stomatitis virus (VSV) (family Rhabdoviridae) and Sendai virus (SeV) (subfamily Paramyxovirinae, family Paramyxoviridae) are the most extensively studied viruses in this order and are models for mononegavirus RNA synthesis (for reviews, see references 25 and 32).
The mononegavirus genome is a single strand of negative-sense RNA that exists as a ribonucleoprotein complex in which the RNA is tightly and completely encapsidated with virus nucleoprotein N. The viral genome is transcribed to produce mRNA and replicated to produce progeny genome by a virus-encoded RNA-dependent RNA polymerase. Initiation of replication and transcription occurs at or near the 3' end of the genome (11) and is directed by cis-acting sequences located within a 3' extragenic region known as the leader (Le) (4, 14, 15, 20, 27, 39). For members of the heterologous Paramyxovirinae subfamily of Paramyxoviridae, the genome promoter is more complex and extends into the first gene (20, 21, 29, 37). During transcription, the polymerase responds to sequences at the boundaries of each gene to generate monocistronic mRNAs (1, 23, 35). A gene start (GS) signal directs the polymerase to initiate mRNA synthesis, and a gene end (GE) signal directs the polymerase to release the mRNA. These signals are also thought to be involved in modifying the 5' and 3' ends of the mRNA by capping and polyadenylation, respectively (1, 36). During replication, the polymerase disregards the gene junction sequences to synthesize a full-length complement of the genome, the antigenome. The antigenome contains a promoter at its 3' terminus that allows it to act as a template for genome RNA synthesis. Antigenome and genome RNAs are encapsidated with N protein as they are synthesized, and it is thought that concurrent encapsidation increases polymerase processivity, allowing RNA synthesis to continue through the gene junctions (17, 38). RSV follows the same strategy of RNA synthesis as do the prototype viruses, described above, although at a more detailed level there are differences between RSV and various other mononegaviruses in the structure of the gene junction regions (7), adherence to the "rule of six" (33), and organization of cis-acting elements at the genome termini (8, 14). In addition, RSV requires a transcription elongation factor, M2-1, for processive transcription (9) and encodes a second factor, M2-2, which appears to increase the efficiency of replication at the expense of transcription (2).
The initial steps in mononegavirus RNA synthesis are poorly understood, and it is currently unclear whether transcription and replication are initiated in the same way or by two distinct mechanisms. Replication is believed to initiate directly opposite the first nucleotide of the genome, whereas mRNA from the first gene is initiated at the 3'-proximal GS signal, approximately 50 nucleotides (nt) downstream of the 3' end of the genome. A longstanding model for mononegavirus RNA synthesis postulates that transcription is initiated at the 3' terminus of the Le in the same way as is replication. According to this model, the polymerase releases the nascent transcript at or near the end of the Le region, scans the template for the GS signal, and reinitiates mRNA synthesis at the GS site (3). Another model postulates that transcription is initiated directly at the GS signal and therefore is a process that is completely distinct from replication initiation (5). Although recent studies have suggested that VSV transcription is initiated directly at the GS signal (5, 31, 40), there are data that support the 3'-terminal entry model (reviewed in reference 22), and the mechanisms of transcription and replication initiation remain unresolved.
One means of characterizing the initial steps in mononegavirus RNA synthesis is to identify the necessary cis-acting sequences and determine their functions in transcription and replication. Although some cis-acting sequences have been mapped for several mononegaviruses, the functions of these sequences have not been fully defined. Based on the models described above, the roles of cis-acting sequences could include a promoter or promoters that direct polymerase binding and RNA synthesis initiation, an encapsidation nucleation signal present in nascent replication products, and sequences that direct internal initiation at the GS signal or, alternatively, release of the Le transcript and reinitiation at the GS signal.
In the case of RSV, previous studies have shown that all of the essential cis-acting signals for transcription and replication are contained within the 3'-terminal 54 nt: the 44-nt Le region contains all of the sequences necessary to direct antigenome RNA synthesis (8, 19), and the Le in conjunction with the 10-nt promoter-proximal GS signal is able to direct mRNA synthesis (19, 23, 24). Fine mapping of the first 26 nt of Le showed that positions 1 to 11 of the Le region are required for replication and that nt 3, 5, 8, 9, 10, and 11 are important for transcription (15), suggesting that these nucleotides comprise a common promoter element. In the present study, the RSV Le region downstream of nt 15 was broadly mapped by mutagenesis, cis-acting sequences important for transcription and replication were identified, and their functions were examined.
MATERIALS AND METHODS
Mutagenesis and plasmid construction. RSV Le regions containing substitutions or deletions were constructed by PCR with mutagenic oligonucleotides and inserted into plasmid C75, which was described previously (15). This plasmid carries a dicistronic minigenome with a single-nucleotide substitution at the penultimate nucleotide at the 5' end of the extragenic trailer (Tr) region, which inactivates the antigenomic promoter and limits replication to the antigenome synthesis step. The PCR products were inserted between the BtgI and XbaI sites of C75; the BtgI site is at the junction of the ribozyme and the Le region, and the XbaI site is at the junction between the NS1 nontranslated region and the open reading frame of the chloramphenicol acetyltransferase (CAT) reporter sequence. To insert spacer sequence into the Le region, an SphI site was introduced between nt 35 and 36 by performing PCR with a mutagenic primer and inserting the PCR product into the BstXI sites of C75. C75 has two BstXI sites, one naturally occurring at nt 35 to 46 of the RSV genomic sequence and one within the ribozyme. Thus, this modification introduced 6 nt between nt 35 and 36. Heterologous inserts of 66 or 106 nt were generated by PCR, amplifying nt 676 to 729 or 636 to 729 of the RSV N gene (numbering is with respect to the N GS signal) and inserting the products into the SphI site. The two inserts were ligated in opposite orientations from each other, and therefore the +66 and +106 minigenomes contain different spacer sequences. The C75 plasmid background was used for the experiments shown in Fig. 2 and 3A. For the experiments shown in Fig. 3B and 4 to 7, the mutant Le regions were inserted into a plasmid encoding a minigenome with a wild-type (wt) Tr region. This was done either by inserting the BtgI-XbaI fragment into plasmid MP-28, which has been described previously (15), or by inserting a PCR product containing the CAT 2 gene, the L nontranslated region, a wt Tr region, and the T7 promoter into the BglII and HindIII sites of the mutant C75 plasmids. The BglII site lies immediately downstream of the internal gene junction in the minigenome, and the HindIII site lies immediately downstream of the T7 promoter. In each of these mutant plasmids, all RSV-, T7 promoter-, and ribozyme-specific sequences that were generated by PCR were sequenced. To generate the tricistronic minigenome (see Fig. 6), the MP-28 plasmids containing the wt, 16-43S, and 16-34S Le regions (see Fig. 1) were modified by inserting a PCR product encoding the G-F sequence from nt 225 of the RSV G gene to nt 870 of the RSV F gene into the BglII site. The G-F gene junction in each of these mutants was sequenced.
Reconstituted minigenome transcription and replication. Minigenome plasmids were transfected into HEp-2 cells together with N, P, L and M2-1 plasmids as described previously (15), except that T7 polymerase was expressed from vaccinia virus MVA-T7 (41). In the experiments shown in Fig. 2 to 5, each transfection reaction was set up in duplicate: RNA was directly extracted from cells from one of the wells, and the cells in the other well were lysed with nonionic detergent and incubated with micrococcal nuclease (MCN) prior to RNA purification to digest unencapsidated RNA, as described previously (30). In the experiment shown in Fig. 6, parallel transfections in which the M2-1 protein was either included or omitted were carried out. Empty pTM1 vector was added to transfection mixes in which either the L polymerase or M2-1 protein expression plasmid was omitted to maintain a constant input plasmid concentration.
RNA isolation and Northern blot hybridization. RNA was isolated, electrophoresed, and transferred to nitrocellulose as described previously (15). Negative- or positive-sense 32P-labeled CAT-specific riboprobes were synthesized with T7 RNA polymerase. The template for the negative-sense probe was the minigenome-carrying plasmid C75 linearized with XbaI (15). The template for the positive-sense probe was PstI-linearized DM28 plasmid, which contains the CAT sequences from C75 inserted between the XbaI and PstI sites of pGEM3Zf(–) (Promega). Riboprobes were extracted with phenol and chloroform and hybridized to the Northern blot in a mixture of 6x SSC (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate), 2x Denhardt's solution, 0.5% sodium dodecyl sulfate (SDS), and 100 μg of sheared DNA per ml at 65°C for at least 12 h. The blots were washed in 2x SSC-0.1% SDS at 65°C for 2 h and then in 0.1x SSC-0.1% SDS at 65°C for 15 min.
The blots shown in Fig. 3B (upper panels), 6B, and 7A were hybridized with negative-sense 32P-labeled oligonucleotides. Oligonucleotides Le oligo (5' TTTATGCAAGTTTGTTGTACGCATTTTT), containing nt 7 to 34 of the wt Le, and LeS oligo (5' AAATACGTTCAAACAACATCGCATTTTT), containing nt 7 to 34 of the 16-34S mutant, were used to detect RNA encoded by wt Le or by Le with positions 16 to 34 substituted, respectively. Oligonucleotides were labeled by using [-32P]ATP and T4 polynucleotide kinase. The oligonucleotides were separated from unincorporated nucleotides by chromatography on a Sephadex G25 column. Probes were hybridized to Northern blots in a mixture of 6x SSC, 5x Denhardt's solution, 0.1% SDS, 0.05% sodium pyrophosphate, and 100 μg of tRNA per ml at 50°C for at least 12 h. Blots were washed in 100 ml of 6x SSC twice at room temperature for 15 min and twice at 50°C for 15 min. Phosphorimager analysis was carried out with a Molecular Imager FX and Quantity One quantitation software (Bio-Rad).
Primer extension analysis. A negative-sense CAT specific primer (5' GGGATATATCAACGGTGGTATATCCAGTG) was purified by polyacrylamide gel electrophoresis, and 100 ng of primer was end labeled with 32P as described above. The labeled primer was separated from unincorporated nucleotides by using a Sephadex G25 spin column (Amersham Biosciences). Total or MCN-treated RNA representing one-fifth of a well of cells was annealed to one-third of the 32P-labeled primer in 1x avian myeloblastosis virus cDNA synthesis buffer by heating the mixture to 65°C for 10 min and then cooling on ice for 5 min. The RNA-DNA hybrid was used as a template for reverse transcription with cloned avian myeloblastosis virus reverse transcriptase in 1x cDNA synthesis buffer, 1 mM dithiothreitol, deoxynucleoside triphosphates at 0.5 mM each, and 40 U of RNase inhibitor. Reverse transcription was carried out at 45°C for 1 h. The cDNA products were extracted with phenol and chloroform, precipitated with ethanol, and resuspended in 10 μl of loading buffer (95% formamide, 20 mM EDTA, 0.05% xylene cyanol). Five microliters of the cDNA products was electrophoresed on a 6% polyacrylamide gel containing 7 M urea and analyzed by autoradiography and phosphorimaging. The cDNAs corresponding to RNA initiated at the 3' terminus and the GS signal were initially identified by comparison with a sequencing ladder run in an adjacent well (14). Phosphorimager analysis was carried out with a Molecular Imager FX and Quantity One quantitation software (Bio-Rad).
RESULTS
Identification of transcription- and replication-specific sequences in the Le region. As described in the introduction, there is a cis-acting element contained within the first 11 nt at the 3' terminus of the Le that is important for both transcription and replication (15). To determine whether the remainder of the Le region is necessary for RNA synthesis, a series of mutant minigenomes in which various regions of the Le were replaced with complementary sequence was constructed (Fig. 1B, rows i to v). In each mutant, the first 15 nt of Le and nt 44 at the –1 position relative to the mRNA start signal were maintained as wt sequence to avoid an impact of the mutations on the functions of the cis-acting element at the 3' terminus of the genome and the GS signal, respectively. The Le mutations were introduced into plasmid C75, which carries a dicistronic minigenome containing a single-nucleotide substitution at the penultimate nucleotide in the 5' Tr region, as described previously (15). This mutation inactivates the antigenomic promoter, thus limiting minigenome replication to the antigenome synthesis step (30). The lack of minigenome amplification allows transcription and replication to be quantified as independent events. RSV RNA synthesis was reconstituted in HEp-2 cells in which minigenome RNA and the RSV N, P, L, and M2-1 proteins were expressed from transfected T7 plasmids (9, 16). The levels of mRNA and antigenome produced from the mutant minigenomes were measured by Northern blot analysis of total intracellular RNA with a negative-sense CAT-specific probe (Fig. 2A). The mRNA transcripts from the CAT 1 and CAT 2 genes are indicated as mRNA 1 and mRNA 2, respectively. To allow accurate quantitation of antigenome RNA, duplicate transfection cell lysates were treated with MCN prior to RNA purification. Encapsidated antigenome RNA is resistant to MCN treatment and is the highest-molecular-weight band in this panel (Fig. 2B). Some mRNA appears to be resistant to MCN treatment in this experiment. Apparent incomplete digestion of mRNA was typical and might have been due to incomplete solubilization of the lysate in nonionic detergent. Also, in this experiment, in which minigenome replication is blocked at the genome synthesis step, the reduced production of encapsidated RNAs might have resulted in an excess of free N protein, which could then have engaged in nonspecific encapsidation of mRNA. Duplicate blots were probed with a positive-sense CAT-specific probe as a control to confirm that similar levels of encapsidated minigenome template were produced by each of the minigenome cDNAs (data not shown).
The wt minigenome generated a large amount of mRNA and a small amount of antigenome (Fig. 2, lanes 1), as expected. It should be noted that on this particular Northern blot, the products from the wt minigenome in the total RNA fraction are somewhat overrepresented; in several other experiments, wt mRNA levels were more comparable to the levels shown in lanes 5 and 9, as indicated in the quantitative analysis shown below the blots. In contrast to the wt minigenome, mutant 16-43S, in which the entire region between the 3'-terminal cis-acting element and the GS signal was replaced, produced barely detectable levels of mRNA and antigenome (Fig. 2, lanes 3), demonstrating that the Le region downstream of the 3'-terminal promoter element contains sequences that are important for both transcription and replication. To map the locations of the cis-acting elements within this region of Le, minigenomes containing substitutions in nt 16 to 34 or 36 to 43 were generated. These regions were chosen based on the sequence alignments shown in Fig. 1A, which show that that the central region of the Le is well conserved with the antigenomic promoter, suggesting that it might be important for replication, and that the U-rich region at the end of the Le is well conserved in other pneumovirus Le regions but is absent in the antigenome promoter, suggesting that it might be important for transcription. Consistent with this supposition, minigenome 16-34S produced levels of mRNA similar to those produced by the wt minigenome (Fig. 2A, lane 5), indicating that the first 15 nt of Le, together with a U-rich region at the end of the Le and the GS signal, is sufficient to direct mRNA synthesis. However, this mutant replicated poorly (Fig. 2B, lane 5). A minigenome containing the converse mutation, in which the first 35 nt of Le were wt sequence and nt 36 to 43 were replaced, produced mRNA at low levels but antigenome at levels equivalent to, or greater than, wt (Fig. 2, lanes 7). Thus, these data show that in addition to the cis-acting sequence within nt 1 to 11, the RSV Le contains nonoverlapping regions that are important for transcription and replication: the central region of Le (nt 16 to 34) is necessary for efficient antigenome production, and the U-rich region at the end of Le (nt 36 to 43) is important for efficient mRNA synthesis. In an attempt to more precisely map the replication-specific sequence in the central region of Le, two additional substitution mutants were constructed. Lanes 8 of Fig. 2 show RNA from a minigenome containing bovine RSV (BRSV) Le sequence for positions 1 to 43. As shown in Fig. 1A, the human RSV and BRSV Le regions are highly conserved, but they diverge from nt 26 to 34; thus, this mutant allowed analysis of these nucleotides. Lanes 9 show RNA from a minigenome containing a substitution of complementary sequence in nt 16 to 26. Both of these mutants directed high levels of transcription, as expected. Minigenome 16-26S generated a low level of antigenome similar to that generated by the 16-34S mutant, indicating that this region is important for replication. Although the BRSV mutant replicated slightly more efficiently than the 16-26S and 16-34S mutants, its level of antigenome production was considerably lower than that of the wt. Therefore, these data show that there are replication-specific sequences extending from nt 16 to 34, with nt 16 to 26 being particularly significant.
The relative spacing of cis-acting sequences is important for transcription and replication. Having identified replication- and transcription-specific sequences in the Le region, it was of interest to determine whether the spacing of these elements is important. Deletions correlating with the regions of substituted sequence were introduced, as shown in Fig. 1B (rows vi to viii). Each of the deletion mutants directed lower levels of mRNA and antigenome synthesis than their corresponding substitution mutants (Fig. 2, compare lanes 2, 4, and 6 with lanes 3, 5, and 7). This demonstrated that the relative spacing between the 3' terminus and the GS signal is important for efficient mRNA synthesis and suggested that positioning the GS signal too close to the Le 3' end can interfere with replication.
To determine whether increasing the distance between transcription signal sequences affects RNA synthesis, 6, 66, or 106 nt of heterologous sequence was inserted between nt 35 and 36 of the wt Le, thus displacing the U-rich region and the GS signal from the 3' end of the minigenome, as illustrated in Fig. 1B (row ix). Analysis of total and MCN-treated RNAs with a negative-sense CAT-specific riboprobe showed that inserting 6 nt at position 34 had no significant effect on either transcription or replication (Fig. 3A, lanes 2 and 7), but as the length of the inserted sequence was increased to 66 or 106 nt, there was a significant decrease in mRNA and increase in antigenome (Fig. 3A, lanes 3, 4, 8, and 9).
The mRNA that was produced from the 106-nt insertion mutant migrated slightly more slowly than mRNA produced from the wt minigenome (compare lanes 1 and 4 of Fig. 3A), suggesting that it might be initiated at the 3' end of Le. Polyadenylated transcripts initiated at the 3' terminus of Le and extended to the first GE signal have previously been observed in RSV-infected cells and in the minigenome system (6, 23). To determine whether the mRNA detected in lane 4 contained Le-encoded sequence, Northern blot analysis was performed with an Le-specific oligonucleotide probe. Because of the relative insensitivity of an oligonucleotide probe, it was not possible to detect RNA produced from the nonreplicating minigenomes shown in Fig. 3A (data not shown). Therefore, the experiment was repeated with minigenomes with a wt Tr region, which can be amplified through multiple cycles of replication, resulting in an enhanced signal. Analysis of total RNA with a negative-sense, Le-specific probe showed that each minigenome produced two Le-containing RNAs (Fig. 3B, upper panels). The upper band is antigenome, and the lower band is mRNA 1 containing the complement of the Le, referred to as LeC-mRNA 1. Insertion of 6 nt had no significant effect on synthesis of either of these transcripts, but insertion of 66 or 106 nt of spacer sequence resulted in an increase in antigenome and LeC-mRNA 1 synthesis. Analysis of RNA purified from MCN-treated cell extracts showed that most of the LeC-mRNA 1 was digested (Fig. 3B, lanes 7 to 10), indicating that it was unencapsidated, as has been shown previously (14). The blot was stripped and reprobed with a negative-sense CAT specific riboprobe (Fig. 3B, lower panels) to allow comparison of the levels of mRNA detected by CAT- or Le-specific riboprobes. This analysis indicated that all detectable mRNA produced from the +106 mutant was attached to Le, suggesting that displacement of the transcription signals by 106 nt abolished transcription initiation at the GS signal. Thus, these data (i) confirm that the spacing of transcription signals is important for mRNA synthesis initiation from the GS signal and (ii) suggest that initiation at the 3' terminus of Le is subject to competition from the GS signal, which is aggravated if the GS is moved closer and alleviated if it is moved away.
Analysis of mRNA and antigenome initiation. The Northern blot analyses shown in Fig. 2 identified sequences necessary for synthesis of full-length antigenome and mRNA. However, Northern blotting does not provide information regarding the level or accuracy of initiation of these RNAs. To examine mRNA and antigenome initiation, primer extension analysis was carried out on total intracellular RNA with an excess of labeled, negative-sense primer that hybridized 130 nt from the 5' end of the antigenome and 86 nt from the GS sequence. Thus, this primer could be used to detect 3'-terminal and GS initiations within the same reaction (Fig. 4A and D). Primer extension products were barely detectable when nonreplicating minigenomes were used (data not shown), so this analysis was carried out on RNA produced from minigenomes that possessed a wt Tr region and thus were capable of amplification. To allow direct comparison with the levels of full-length antigenome and mRNA, aliquots of the same total RNA samples were analyzed by Northern blotting with a negative-sense CAT-specific riboprobe (Fig. 4B and E). MCN-treated samples prepared from duplicate wells were also analyzed to examine encapsidated antigenome levels (Fig. 4C and F).
Primer extension on RNA generated from the wt minigenome yielded two bands, corresponding to GS and 3'-terminal initiations (Fig. 4A and D, lane 2). Examination of GS initiations from the substitution mutants showed that the 16-34S, 16-26S, and BRSV minigenomes directed levels of GS initiations similar to those for the wt Le (Fig. 4A, lanes 4, 6, and 7), whereas minigenomes 16-43S and 36-43S directed considerably less (Fig. 4A, lanes 3 and 5). Thus, the relative levels of these initiations correlate with the levels of mRNA detected by Northern blot analysis (Fig. 4B). These data demonstrate that the GS signal and the 3'-terminal core promoter element are sufficient to direct accurate initiation at the GS sequence and confirm that nt 36 to 43 significantly augment levels of initiation at this site.
Analysis of bands representing RNA initiated at the 3' terminus of Le demonstrated that each of the substitution mutants directed efficient initiation compared to the wt minigenome (Fig. 4A, lanes 2 to 7). This was a surprising result given that minigenomes containing mutations involving nt 16 to 34 are significantly inhibited in their ability to produce full-length antigenome (Fig. 4C, lanes 3, 4, 6, and 7). These results were highly reproducible; for example, in three separate experiments the 16-34S mutant directed 3' initiation at 130% or more than wt but generated full-length antigenome at 26% or less. It is possible that the mutations in nt 16 to 34 rendered the antigenome RNA unstable, such that fragments that remained could be detected by primer extension but not by Northern blotting. However, this seems unlikely, and if such fragments existed, primer extension analysis should have yielded a more complex pattern. Therefore, these data suggest that the first 15 nt of Le contain all of the sequences required to direct antigenome initiation and that nt 16 to 34 are required specifically for elongation to produce full-length antigenome. To confirm that sequence within the first 15 nt is important for RNA synthesis initiation, RNA from a minigenome containing a C-to-A substitution at Le position 3 was examined. This substitution completely inhibits transcription and replication (15) (Fig. 4B and C, lanes 8) and eliminated the 3' and GS initiation bands detected by primer extension (Fig. 4A, lane 8), confirming that the 3' terminus of Le has the properties of a promoter.
The results obtained with the substitution mutants were confirmed with their deletion mutant counterparts. Analysis of the deletion mutants 16-43D and 36-43D by primer extension showed that the levels of GS initiations for these mutants were consistent with the levels of full-length mRNA detected by Northern blotting (Fig. 4D and E, lanes 3 and 4). Similar to the substitution mutants, both of these minigenomes directed initiation from the 3' terminus (Fig. 4D, lanes 3 and 4). Note that the sizes of the bands representing 3' initiations are reduced for the deletion mutants compared to the wt, as would be expected. Despite its ability to direct 3' initiation, minigenome 16-43D did not generate detectable full-length antigenome (Fig. 4F, lane 3), whereas in contrast, there was a good correlation between 3' initiation and full-length antigenome production for the 36-43D mutant (Fig. 4D and F, lanes 4). These data are consistent with the hypothesis that nt 16 to 34 contain a sequence that is essential for elongation to produce full-length antigenome.
nt 16 to 34 are important for antigenome encapsidation. Studies with SeV have suggested that antigenome elongation is dependent on concurrent encapsidation (17, 38). Thus, a possible function of nt 16 to 34 is to mediate encapsidation of the nascent antigenome RNA, either by recruiting encapsidation factors to the genome template or by templating for an encapsidation nucleation signal in the encoded antigenome. To test whether nt 16 to 34 are involved in encapsidation, total and MCN-treated RNA samples from duplicate transfections were analyzed by primer extension (Fig. 5). Analysis of MCN-treated RNAs produced by wt and mutant minigenomes showed that in each case at least some of the 3' initiation transcript could be detected (Fig. 5C and D). A proportion of transcripts initiated at the GS signal were also detected, indicating that MCN treatment likely did not completely digest unencapsidated RNA. This was not unexpected given that Northern blot analysis of MCN-treated RNA often showed a smear of partially digested mRNA, as described previously. Despite the detection of residual unencapsidated RNA, it was possible to determine the relative levels of encapsidation for each of the mutants by comparing their levels of 3' initiation transcripts to those of the wt in the total RNA and MCN-treated RNA samples. Analysis of minigenome 16-34S by this approach showed that in the total RNA fraction, the level of 3' initiation transcript was 141% of the level produced from the wt minigenome (Fig. 5A, lane 3). In contrast, in the MCN fraction, the level of 3' initiation transcript from the 16-34S mutant was only 34% of wt (Fig. 5C, lane 3). This demonstrates that the transcripts generated from the 3' terminus of the 16-34S mutant were encapsidated less efficiently than transcripts from the wt Le minigenome. Similar analysis of other minigenomes with substitutions covering nt 16 to 34 indicated that each of these mutations was associated with reduced encapsidation of the 3' initiation products (Fig. 5A and C, lanes 4 to 6). The difference was less significant in the case of BRSV Le (Fig. 5A and C, lanes 5), consistent with its slightly higher level of antigenome production compared to minigenomes 16-34S and 16-26S (as shown in Fig. 2 and 4). These data suggest that nt 16 to 34 are important for encapsidation of the transcripts generated from the 3' end of Le. As a control to confirm that the sequences required for encapsidation are located within the first 34 nt of Le, RNA from mutant 36-43D was analyzed in parallel with that from mutant 16-34S (Fig. 5B and D). The results for 16-34S (Fig. 5B and D, lanes 2) are similar to those shown in Fig. 5A and C. In contrast, mutant 36-43D produced 3' initiation transcripts at 36% of wt levels in the total RNA fraction and at 35% of wt levels in the MCN-treated fraction (Fig. 5B and D, lanes 3). Thus, this result confirms that while deletion of nt 36 to 43 inhibited initiation from the 3' terminus, it had no effect on encapsidation of the transcripts produced.
Abortive replicases can complete synthesis of the first gene and default to a transcription mode. The results shown in Fig. 4 suggest that mutation of nt 16 to 34 results in nonprocessive replication from the 3' terminus, as evidenced by the presence of 3' initiation product and absence of full-length antigenome. However, the abortive transcripts detected by primer extension analysis were at least 130 nt in length. Therefore, it was of interest to determine how far the polymerase was able to extend these transcripts. For this analysis, wt Le and mutant 16-34S and 16-43S Le sequences were introduced into a tricistronic minigenome backbone, in which the first gene contained CAT sequence, the second contained part of the RSV G gene, and the third contained part of the F gene fused with additional CAT sequence (Fig. 6A). The first and second gene junctions were derived from the N-P and G-F gene junction sequences, respectively. The minigenomes were used as templates in the intracellular transcription-replication assay, as described above, except that reactions were carried out both with and without the M2-1 expression plasmid. To examine RNA initiated from the 3' end of Le, total intracellular RNA was analyzed by Northern blotting with negative-sense, Le-specific oligonucleotide probes, one of which (Le oligo) was specific to RNA encoded by the wt Le and the other of which (LeS oligo) was specific to RNA encoded by the 16-34S and 16-43S Le sequences. Examination of RNA produced in the presence of M2-1 showed that the wt minigenome and the 16-43S and 16-34S mutants generated two LeC-containing RNAs, namely, antigenome and the LeC-mRNA 1 readthrough mRNA (Fig. 6B, lanes 1, 3, and 5). These same two species were detected when the blot was stripped and reprobed with a CAT-specific riboprobe, and in addition, the LeC-mRNA 1 band was augmented by the monocistronic mRNA 1, which was not resolved from the LeC-mRNA 1 species (Fig. 6C, lanes 1, 3, 5, and 9). The CAT-specific probe also detected mRNA 3, which migrated between the antigenome and mRNA 1 and was confirmed as mRNA 3 with an F-specific probe (data not shown). Examination of RNAs detected with the Le-specific probes showed no evidence of a smear of LeC-containing transcripts above or below the LeC-mRNA band, as would be expected if LeC-containing transcripts were terminated at various distances from the 3' terminus to a considerable extent (Fig. 6B, lanes 1, 3, and 5). These results suggest that in the presence of M2-1, a significant proportion of the abortive replication transcripts produced by 16-34S and 16-43S Le sequences were extended to the end of the first gene to produce LeC-mRNA 1.
To determine whether the polymerase that generated LeC-mRNA 1 was able to continue stop-start transcription along the length of the minigenome, the ratio of the third to the first gene transcripts, detected with the CAT-specific riboprobe, was analyzed. Because the 16-43S mutant generated very little mRNA 1 from the first GS signal (see, e.g., Fig. 4A) but produced amounts of LeC-mRNA 1 similar to those produced by the 16-34S minigenome, its gene 1 transcripts contained a much greater proportion of LeC-mRNA than the gene 1 transcripts produced from the 16-34S minigenome (compare Fig. 6B and C, lanes 3 and 5). If polymerase that synthesized LeC-mRNA 1 were unable to continue transcription beyond the first GE signal, the 16-43S mutant would be expected to produce very low levels of mRNA 3 relative to mRNA 1 compared to the 16-34S mutant. Examination of mRNAs produced in the presence of M2-1 with the CAT-specific riboprobe allowed comparison of the total population of mRNA 1 transcripts (including LeC-mRNA 1) and mRNA 3 transcripts on the same blot. Although the 16-43S mutant produced lower levels of mRNAs 1 and 3 than the 16-34S and wt minigenomes, as expected, the ratios of mRNA 3 to mRNA 1 were similar in each case (Fig. 6D). Therefore, this result indicates that polymerase that generates LeC-mRNA 1 can continue sequential stop-start transcription along the length of the genome.
RNA produced in the absence of M2-1 was also analyzed to determine the effect of this protein on RNA synthesis from the mutant minigenomes (Fig. 6B and C, lanes 2, 4, and 6). This analysis indicated that for each minigenome, a small amount of full-length LeC-mRNA 1 was generated, but a significant proportion was present as a smear of prematurely terminated product. Note that the premature termination products observed in Fig. 6B were less evident in Fig. 6C because in the latter case the more stringent hybridization and washing conditions necessary with a riboprobe, versus an oligonucleotide probe, were used. Thus, these data indicate that the synthesis of full-length LeC-mRNA 1 is dependent on M2-1 activity. Analysis of mRNAs 1 and 3 produced in the absence of M2-1 showed that a small amount of full-length mRNA was produced from the CAT 1 gene of each of the minigenomes (Fig. 6C, lanes 2, 4, 6, and 10). This was expected because the first gene is relatively short and a proportion of polymerase is able to complete its synthesis, as noted previously (13). However, for each minigenome, the third gene transcript was barely detectable in the absence of M2-1. Taken together, these data suggest that M2-1 protein activity is necessary to allow abortive replicases to continue RNA synthesis to the end of the first gene to produce LeC-mRNA 1 and then continue mRNA synthesis along the remainder of the genome. It should be noted that in this particular experiment there was variation in the levels of antigenome such that accumulation appeared to be reduced in the presence of M2-1. However, this likely reflects uneven transfer of the antigenome RNA in this experiment: for example, the RNA samples shown in Fig. 6C, lanes 1 and 2, are the same RNA samples as in lanes 9 and 10, which were electrophoresed and transferred from the same gel, and yet there was variation in the apparent abundance of antigenome.
As an additional step to confirm that polymerase that generated LeC-mRNA 1 could continue sequential transcription, we reexamined RNA synthesis by the +106-nt insertion mutant (Fig. 1B). This mutant had the desirable feature of producing a relatively large amount of LeC-mRNA 1 and little or no monocistronic mRNA 1, as shown in Fig. 3B. Here we examined whether this mutant directed the synthesis of mRNA 2, which would be evidence that sequential transcription followed the synthesis of LeC-mRNA 1. RNA produced in the presence of M2-1 from the wt minigenome and the +106-nt insertion mutant was analyzed with Le oligo to identify LeC-containing RNAs (Fig. 7A). Analysis of the same blot with a CAT-specific riboprobe showed that all detectable mRNA 1 from the insertion mutant migrated similarly to RNA detected with Le oligo and was of increased molecular weight compared to that produced from the wt minigenome, indicating that it was initiated at the 3' end of the Le region (Fig. 7B, lane 3). Importantly, analysis with the CAT-specific probe showed that the 106-nt insertion mutant generated mRNA 2 (Fig. 7B, lane 3). Quantitation of the total mRNA 1 transcripts (including LeC-mRNA 1) and mRNA 2 transcripts, shown in lanes 2 and 3 of Fig. 7B, demonstrated that the ratio between them was similar for the wt minigenome and the +106-nt insertion mutant (Fig. 7C), although it should be noted that the small size of mRNA 2 and its diffuse nature on the Northern blot precluded accurate quantitation of this mRNA. Thus, these data are consistent with a model in which polymerase that initiates at the 3' end of the Le region and terminates at the first GE signal is able to reinitiate mRNA synthesis at a downstream GS signal.
DISCUSSION
Elucidating the functions of cis-acting sequences in the RSV Le is an important step towards understanding the initial events involved in mononegavirus RNA synthesis. In a previous study it was shown that the 3' end of Le contains residues that are important for both transcription and replication. The present study identified nonoverlapping transcription- and replication-specific sequences in the remainder of Le and completes a broad mapping of the RSV Le region. The Le is split broadly into three regions: the 3'-terminal 15 nt contain sequence that is important for transcription and replication, the central region (nt 16 to 34) contains sequences important only for replication, and the downstream end of Le (nt 36 to 43) is important only for transcription. The functions of these sequences and the GS signal were further dissected, identifying replication initiation and elongation signals and demonstrating that the GS signal, in conjunction with Le nt 1 to 15, is sufficient for accurate initiation of mRNA synthesis.
The 3'-terminal 15 nt of Le act as an efficient promoter for RNA replication. This was highlighted in particular by analysis of mutants 16-43S and 16-43D, which were able to direct levels of 3' initiation comparable to that of the wt Le region (Fig. 4). The minimal region required for promoter activity was not determined in this study, but previously it was shown that most single-nucleotide substitutions in positions 1 to 11 significantly inhibited replication (15). A number of these substitutions also inhibited transcription, suggesting that these nucleotides comprise a common promoter element. In the case of VSV, it has been shown that a synthetic template containing 17 nt of wt sequence is a functional promoter in vitro (34), and the first 15 nt of wt sequence can act as a promoter in an intracellular minigenome assay (39). The first 12 nt of the genome and antigenome are conserved in all of the Paramyxovirinae, and these nucleotides have been shown to be critical for RNA replication in parainfluenza virus type 3 (20), suggesting that this is the typical size of the 3'-terminal element of the mononegavirus replication promoter. In the case of members of the Paramyxovirinae, the replication promoter contains a second essential cis-acting element located within the first gene (20, 21, 29, 37), but this additional element is not present in RSV or VSV.
The central portion of the RSV Le (nt 16 to 34) was required for efficient elongation of antigenome initiated at the 3' terminus of Le to produce full-length molecules and also was important for efficient encapsidation of the antigenome. It is tempting to speculate that Le nt 16 to 34 encode a sequence in the complementary nascent antigenome that binds N and acts as an encapsidation nucleation site. However, an attempt to identify an N-binding site in the RSV Le-encoded transcript failed to show evidence for a specific interaction (28). In addition, a previous study showed that LeC-mRNA transcripts were not encapsidated even under conditions in which N protein was in excess (14), suggesting that in RSV, encapsidation might not be initiated simply by interaction of N with a cis-acting sequence at the 5' end of the RNA. One possibility is that nt 16 to 34 encode an encapsidation nucleation site but that encapsidation is mediated by the polymerase during RNA synthesis. Alternatively, it is possible that these nucleotides constitute a signal in the genome template, rather than nascent antigenome, that is important for recruiting a polymerase that is able to mediate encapsidation, for example, a polymerase that is preloaded with N protein, which has been suggested to be the active form of the VSV replicase (18, 31).
The region at the downstream end of Le (nt 36 to 43) was necessary for efficient transcription. The specific nucleotides involved were not identified, but alignment of pneumovirus Le sequences shows that a U stretch immediately upstream of the first GS signal is highly conserved (Fig. 1A). Seven of the eight residues contained in nt 36 to 43 of the RSV Le are U, suggesting that this U-rich nature is an important feature. The finding that nucleotides at the end of Le are important for transcription is consistent with findings for VSV, SeV, and parainfluenza virus type 3 (4, 20, 27, 39), suggesting that the sequences immediately upstream of the promoter-proximal GS signal are important for many mononegaviruses. However, it should be emphasized that although nt 36 to 43 increased the efficiency of initiation from the GS signal, this region was not required for correct initiation, synthesis of complete mRNA, or the initiation of stop-start sequential transcription. It is interesting that in contrast to the results presented here, the sequences required for VSV transcription are located throughout the Le region (27, 39), suggesting that there are differences in the organization of the transcription promoters for RSV and VSV.
Both GS and 3'-terminal initiations were found to be inhibited if the GS signal was moved closer to the 3' terminus (Fig. 2 and 4), suggesting that the close proximity of the promoter and GS signal either inhibits polymerase binding to the template or causes the polymerase to stall shortly after initiation. If the distance between the 3' end of the Le and nt 36 to 43 and the GS signal was increased, mRNA synthesis was inhibited but antigenome and LeC-mRNA syntheses were augmented (Fig. 3). The decrease in transcription could be either because the U-rich region and the GS signal must be within a certain distance from the promoter for the polymerase to be able to terminate and reinitiate RNA synthesis (according to the model in which transcription is mediated by a polymerase that initiates at the 3' terminus of Le) or because these elements must be spaced appropriately for the polymerase to recognize them simultaneously (according to the model in which transcription initiates directly at the GS signal). The increase in antigenome and LeC-mRNA synthesis suggests that GS and 3' initiations are competitive processes. This is also suggested by the fact that a mutant in which nt 36 to 43 were replaced generated a higher level of antigenome than the wt Le (Fig. 2B, lane 7). Competition between synthesis from the 3' terminus and the GS signal has also been observed in SeV (26). Competition could exist because the same polymerase is used for both processes or because there is steric hindrance between replicases and transcriptases as they bind to their respective initiation sites. An experiment in a previous study, in which a spacer sequence was inserted into Le, showed no increase in LeC-mRNA and antigenome with increasing spacer, although mRNA synthesis was completely inhibited (14). The difference between the experiment described here and the previous study is that in the present study, nt 36 to 43 and the GS signal were displaced relative to the 3' terminus, whereas in the previous study, nt 36 to 43 were not displaced. Therefore, it appears that although moving the GS signal away from the 3' terminus causes a decrease in transcription, the U-rich region immediately upstream of the GS signal also must be displaced to alleviate competition on replication, suggesting that the U-rich region interacts directly with the polymerase.
The major product containing Le-encoded sequence from the minigenomes with substitution in nt 16 to 34 was LeC-mRNA 1. Indeed, this RNA was produced in considerable excess over the full-length antigenome. This suggests that LeC-mRNA transcripts are abortive replication products that are produced if the nascent RNA is not encapsidated (Fig. 6). LeC-mRNA has previously been detected in RSV-infected cells, indicating that it is a bona fide product of RSV RNA synthesis (6, 23). Results presented here (Fig. 6 and 7) and in a previous study (23) indicate that polymerase that synthesizes an LeC-mRNA transcript is able to continue stop-start RNA synthesis along the remainder of the genome. In the present study it was not determined whether the transcripts from downstream genes are capped and methylated at their 5' termini, but the study by Kuo and coworkers (23) indicated that polymerase that synthesizes LeC-mRNA goes on to produce translatable mRNA from a downstream gene. Therefore, taken together these data suggest that a replicase can default to a transcription mode in the absence of concurrent encapsidation and that this conversion is an efficient event. In contrast to the situation presented here, SeV abortive replicases were unable to extend the RNA beyond 358 nt (38). The difference in polymerase processivity between SeV and RSV might be due to the M2-1 transcription elongation protein, which is found in RSV but not in SeV.
As described in the introduction, there are two models for transcription initiation, one in which transcription is initiated at the 3' terminus of Le and one in which it is initiated directly at the GS signal. The data presented here could be interpreted to fit either of these initiation models and thus do not resolve this question. For example, nt 36 to 43 and the GS signal either could be involved in reinitiation of a polymerase that initiated at the 3' end of Le or could be a component of a distinct transcription promoter at which the polymerase initiates internally. With regard to competition between transcription and replication, as demonstrated by the effect of increasing the distance between the 3' end and the GS signal, this could also be interpreted to fit either model. For example, if polymerase initiates transcription at the 3' terminus of Le and subsequently terminates (possibly at a certain distance rather than at a specific cis-acting signal) and scans to locate the GS signal, then increasing the distance between the promoter and GS signal could decrease the likelihood that the polymerase will initiate at the GS signal and increase the likelihood that polymerases will reinitiate at the 3' terminus of the genome by scanning backwards, which is a known property of the RSV polymerase (12). On the other hand, if transcription is initiated directly at the GS signal by a distinct transcriptase complex, displacing this sequence could prevent transcriptases from forming a stable interaction with the template, thus allowing more replicases to bind to the Le region and initiate RNA synthesis at the 3' terminus. While these data do not clearly distinguish between these models for normal transcription initiation at the GS signal, the evidence suggests that the RSV polymerase can apparently shift from replication to transcription after the initiation of RNA synthesis, in a manner that seems to be controlled by encapsidation.
ACKNOWLEDGMENTS
We thank Bernard Moss for providing MVA-T7, Karen Jackson for assistance in subcloning some of the substitution mutants, and Vanessa Cowton and Frances Fuller-Pace for helpful discussions.
This work was funded by a Wellcome Trust grant (reference number 065568) to R.F.
REFERENCES
Barr, J. N., S. P. J. Whelan, and G. W. Wertz. 1997. cis-acting signals involved in termination of vesicular stomatitis virus mRNA synthesis include the conserved AUAC and the U7 for polyadenylation. J. Virol. 71:8718-8725.
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.
Blumberg, B. M., C. Giorgi, and D. Kolakofsky. 1983. N protein of vesicular stomatitis virus selectively encapsidates leader RNA in vitro. Cell 32:559-567.
Calain, P., and L. Roux. 1995. Functional characterisation of the genomic and antigenomic promoters of Sendai virus. Virology 211:163-173.
Chuang, J. L., and J. Perrault. 1997. Initiation of vesicular stomatitis virus mutant polR1 transcription internally at the N gene in vitro. J. Virol. 71:1466-1475.
Collins, P. L., and G. W. Wertz. 1985. Nucleotide sequences of the 1B and 1C non-structural protein mRNAs of human respiratory syncytial virus. Virology 143:442-451.
Collins, P. L., L. E. Dickens, A. Buckler-White, R. A. Olmsted, M. A. Spriggs, E. Camargo, and K. V. W. Coelingh. 1986. Nucleotide sequences for the gene junctions of human respiratory syncytial virus reveal distinctive features of intergenic structure and gene order. Proc. Natl. Acad. Sci. USA 83:4594-4598.
Collins, P. L., M. A. Mink, and D. S. Stec. 1991. Rescue of synthetic analogs of respiratory syncytial virus genomic RNA and effect of truncations and mutations on the expression of a foreign reporter gene. Proc. Natl. Acad. Sci. USA 88:9663-9667.
Collins, P. L., M. G. Hill, J. Cristina, and H. Grosfeld. 1996. Transcription elongation factor of respiratory syncytial virus, a nonsegemented negative strand RNA virus. Proc. Natl. Acad. Sci. USA 93:81-85.
Collins, P. L., R. M. Chanock, and B. R. Murphy. 2001. Respiratory syncytial virus, 1433-1485. In D. M. Knipe and P. M. Howley (ed.), Fields virology, 4th ed. Lippincott Williams & Wilkins, Philadelphia, Pa.
Emerson, S. U. 1982. Reconstitution studies detect a single RNA polymerase entry site on the vesicular stomatitis virus genome. Cell 31:635-642.
Fearns, R., and P. L. Collins. 1999. Model for polymerase access to the overlapped L gene of respiratory syncytial virus. J. Virol. 73:388-397.
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.
Fearns, R., P. L. Collins, and M. E. Peeples. 2000. Functional analysis of the genomic and antigenomic promoters of human respiratory syncytial virus. J. Virol. 74:6006-6014.
Fearns, R., M. E. Peeples, and P. L. Collins. 2002. Mapping the transcription and replication promoters of respiratory syncytial virus. J. Virol. 76:1663-1672.
Grosfeld, H., M. G. Hill, and P. L. Collins. 1995. RNA replication by respiratory syncytial virus (RSV) is directed by the N, P, and L proteins: transcription also occurs under these conditions but requires RSV superinfection for efficient synthesis of full-length mRNA. J. Virol. 69:5677-5686.
Gubbay, O., J. Curran, and D. Kolakofsky. 2001. Sendai virus genome synthesis and assembly are coupled: a possible mechanism to promote viral RNA polymerase processivity. J. Gen. Virol. 82:2895-2903.
Gupta, A. K., D. Shaji, and A. K. Banerjee. 2003. Identification of a novel tripartite complex involved in replication of vesicular stomatits virus genome RNA. J. Virol. 77:732-738.
Hallak, L. K., P. L. Collins, W. Knudson, and M. E. Peeples. 2000. Iduronic acid-containing glycosaminoglycans on target cells are required for efficient respiratory syncytial virus infection. Virology 271:264-275.
Hoffman, M. A., and A. K. Banerjee. 2000. Precise mapping of the replication and transcription promoters of human parainfluenza virus type 3. Virology 269:201-211.
Keller, M. A., and G. D. Parks. 2003. Positive- and negative-acting signals combine to determine differential RNA replication from the paramyxovirus simian virus 5 genomic and antigenomic promoters. Virology 306:347-358.
Kolakofsky, D., P. Le Mercier, F. Iseni, and D. Garcin. 2004. Viral RNA polymerase scanning and the gymnastics of Sendai virus RNA synthesis. Virology 318:463-473.
Kuo, L., H. Grosfeld, J. Cristina, M. G. Hill, and P. L. Collins. 1996. Effect of mutations in the gene-start and gene-end sequence motifs on transcription of monocistronic and dicistronic minigenomes of respiratory syncytial virus. J. Virol. 70:6892-6901.
Kuo, L., R. Fearns, and P. L. Collins. 1997. Analysis of the gene start and gene end signals of human respiratory syncytial virus: quasi-templated initiation at position 1 of the encoded mRNA. J. Virol. 71:4944-4953.
Lamb, R. A., and D. Kolakofsky. 2001. Paramyxoviridae: the viruses and their replication, p. 1305-1340. In D. M. Knipe, and P. M. Howley (ed.), Fields virology, 4th ed. Lippincott Williams & Wilkins, Philadelphia, Pa.
Le Mercier, P., D. Garcin, E. Garcia, and D. Kolakofsky. 2003. Competition between the Sendai virus N mRNA start site and the genome 3'-end promoter for viral polymerase. J. Virol. 77:9147-9155.
Li, T., and A. K. Pattnaik. 1999. Overlapping signals for transcription and replication at the 3' terminus of the vesicular stomatitis virus genome. 73:444-452.
Murphy, L. B., C. Loney, J. Murray, D. Bhella, P. Ashton, and R. P. Yeo. 2003. Investigations into the amino-terminal domain of the respiratory syncytial virus nucleocapsid protein reveal elements important for nucleocapsid formation and interaction with the phosphoprotein. Virology 307:143-153.
Murphy, S. K., and G. D. Parks. 1999. RNA replication for the paramyxovirus simian virus 5 requires an internal repeated CGNNNN sequence motif. J. Virol. 73:805-809.
Peeples, M. E., and P. L. Collins. 2000. Mutations in the 5' trailer region of a respiratory syncytial virus minigenome which limit RNA replication to one step. J. Virol. 74:146-155.
Qanungo, K. R., D. Shaji, M. Mathur, and A. K. Banerjee. 2004. Two RNA polymerase complexes from vesicular stomatitis virus-infected cells that carry out transcription and replication of genome RNA. Proc. Natl. Acad. Sci. USA 101:5952-5957.
Rose, J. K., and M. A. Whitt. 2001. Rhabdoviridae: the viruses and their replication, p. 1221-1244. In D. M. Knipe and P. M. Howley (ed.), Fields virology, 4th ed. Lippincott Williams & Wilkins, Philadelphia, Pa.
Samal, S. K., and P. L. Collins. 1996. RNA replication by a respiratory syncytial virus RNA analog does not obey the rule of six and retains a nonviral trinucleotide extension at the leader end. J. Virol. 70:5075-5082.
Smallwood, S., and S. A. Moyer. 1993. Promoter analysis of the vesicular stomatitis virus RNA polymerase. Virology 192:254-263.
Stillman, E. A., and M. A. Whitt. 1997. Mutational analyses of the intergenic dinucleotide and the transcriptional start sequence of vesicular stomatitis virus (VSV) define sequences required for efficient termination and initiation of VSV transcripts. J. Virol. 71:2127-2137.
Stillman, E. A., and M. A. Whitt. 1999. Transcript initiation and 5' end modification are separable events during vesicular stomatitis virus transcription. J. Virol. 73:7199-7209.
Tapparel, C., D. Maurice, and L. Roux. 1998. The activity of Sendai virus genomic and antigenomic promoters requires a second element past the leader template regions: a motif (GNNNNN)3 is essential for replication. J. Virol. 72:3117-3128.
Vidal, S., and D. Kolakofsky. 1989. Modified model for the switch from Sendai virus transcription to replication. J. Virol. 63:1951-1958.
Whelan, S. P. J., and G. W. Wertz. 1999. Regulation of RNA synthesis by the genomic termini of vesicular stomatitis virus: identification of distinct sequences essential for transcription, but not replication. J. Virol. 73:297-306.
Whelan, S. P. J., and G. W. Wertz. 2002. Transcription and replication initiate at separate sites on the vesicular stomatitis virus genome. Proc. Natl. Acad. Sci. USA 99:9178-9183.
Wyatt, L. S., B. Moss, and S. Rozenblatt. 1995. Replication-deficient vaccinia virus encoding bacteriophage T7 RNA polymerase for transient gene expression in mammalian cells. Virology 210:202-205.(David R. McGivern, Peter )
Laboratory of Infectious Diseases, National Institute of Allergy and Infectious Diseases, Bethesda, Maryland
ABSTRACT
Previous studies of respiratory syncytial virus have shown that the 44-nucleotide (nt) leader (Le) region is sufficient to initiate RNA replication, producing antigenome RNA, and that the Le and adjoining gene start (GS) signal of the first gene are sufficient to initiate transcription, producing mRNA. A cis-acting element necessary for both transcription and replication was mapped within the first 11 nt at the 3' end of Le. In the present study the remainder of the Le region was mapped to identify sequences important for transcription and replication. A series of minigenomes with mutant Le sequences was generated, and their ability to direct transcription and replication was determined by Northern blot analysis, which examined full-length antigenome and mRNA, and by primer extension analysis, which examined antigenome and mRNA initiation. With regard to transcription, nt 36 to 43, located immediately upstream of the GS signal, were found to be necessary for optimal levels of mRNA synthesis, although the GS signal in conjunction with the 3'-terminal region of Le was sufficient to direct accurate mRNA synthesis initiation. With regard to replication, the first 15 nt of Le were found to be sufficient to direct initiation of antigenome synthesis, but nt 16 to 34 were required in addition for efficient encapsidation and production of full-length antigenome. Analysis of transcripts produced from di- and tricistronic minigenomes indicated that a significant proportion of abortive replicases continue RNA synthesis to the end of the first gene and then continue in a transcription mode along the remainder of the genome.
INTRODUCTION
Human respiratory syncytial virus (RSV) is the major cause of pediatric respiratory disease worldwide and is increasingly recognized as an important pathogen in the elderly population (10). RSV is a member of the subfamily Pneumovirinae, of the family Paramyxoviridae within the order Mononegavirales. Vesicular stomatitis virus (VSV) (family Rhabdoviridae) and Sendai virus (SeV) (subfamily Paramyxovirinae, family Paramyxoviridae) are the most extensively studied viruses in this order and are models for mononegavirus RNA synthesis (for reviews, see references 25 and 32).
The mononegavirus genome is a single strand of negative-sense RNA that exists as a ribonucleoprotein complex in which the RNA is tightly and completely encapsidated with virus nucleoprotein N. The viral genome is transcribed to produce mRNA and replicated to produce progeny genome by a virus-encoded RNA-dependent RNA polymerase. Initiation of replication and transcription occurs at or near the 3' end of the genome (11) and is directed by cis-acting sequences located within a 3' extragenic region known as the leader (Le) (4, 14, 15, 20, 27, 39). For members of the heterologous Paramyxovirinae subfamily of Paramyxoviridae, the genome promoter is more complex and extends into the first gene (20, 21, 29, 37). During transcription, the polymerase responds to sequences at the boundaries of each gene to generate monocistronic mRNAs (1, 23, 35). A gene start (GS) signal directs the polymerase to initiate mRNA synthesis, and a gene end (GE) signal directs the polymerase to release the mRNA. These signals are also thought to be involved in modifying the 5' and 3' ends of the mRNA by capping and polyadenylation, respectively (1, 36). During replication, the polymerase disregards the gene junction sequences to synthesize a full-length complement of the genome, the antigenome. The antigenome contains a promoter at its 3' terminus that allows it to act as a template for genome RNA synthesis. Antigenome and genome RNAs are encapsidated with N protein as they are synthesized, and it is thought that concurrent encapsidation increases polymerase processivity, allowing RNA synthesis to continue through the gene junctions (17, 38). RSV follows the same strategy of RNA synthesis as do the prototype viruses, described above, although at a more detailed level there are differences between RSV and various other mononegaviruses in the structure of the gene junction regions (7), adherence to the "rule of six" (33), and organization of cis-acting elements at the genome termini (8, 14). In addition, RSV requires a transcription elongation factor, M2-1, for processive transcription (9) and encodes a second factor, M2-2, which appears to increase the efficiency of replication at the expense of transcription (2).
The initial steps in mononegavirus RNA synthesis are poorly understood, and it is currently unclear whether transcription and replication are initiated in the same way or by two distinct mechanisms. Replication is believed to initiate directly opposite the first nucleotide of the genome, whereas mRNA from the first gene is initiated at the 3'-proximal GS signal, approximately 50 nucleotides (nt) downstream of the 3' end of the genome. A longstanding model for mononegavirus RNA synthesis postulates that transcription is initiated at the 3' terminus of the Le in the same way as is replication. According to this model, the polymerase releases the nascent transcript at or near the end of the Le region, scans the template for the GS signal, and reinitiates mRNA synthesis at the GS site (3). Another model postulates that transcription is initiated directly at the GS signal and therefore is a process that is completely distinct from replication initiation (5). Although recent studies have suggested that VSV transcription is initiated directly at the GS signal (5, 31, 40), there are data that support the 3'-terminal entry model (reviewed in reference 22), and the mechanisms of transcription and replication initiation remain unresolved.
One means of characterizing the initial steps in mononegavirus RNA synthesis is to identify the necessary cis-acting sequences and determine their functions in transcription and replication. Although some cis-acting sequences have been mapped for several mononegaviruses, the functions of these sequences have not been fully defined. Based on the models described above, the roles of cis-acting sequences could include a promoter or promoters that direct polymerase binding and RNA synthesis initiation, an encapsidation nucleation signal present in nascent replication products, and sequences that direct internal initiation at the GS signal or, alternatively, release of the Le transcript and reinitiation at the GS signal.
In the case of RSV, previous studies have shown that all of the essential cis-acting signals for transcription and replication are contained within the 3'-terminal 54 nt: the 44-nt Le region contains all of the sequences necessary to direct antigenome RNA synthesis (8, 19), and the Le in conjunction with the 10-nt promoter-proximal GS signal is able to direct mRNA synthesis (19, 23, 24). Fine mapping of the first 26 nt of Le showed that positions 1 to 11 of the Le region are required for replication and that nt 3, 5, 8, 9, 10, and 11 are important for transcription (15), suggesting that these nucleotides comprise a common promoter element. In the present study, the RSV Le region downstream of nt 15 was broadly mapped by mutagenesis, cis-acting sequences important for transcription and replication were identified, and their functions were examined.
MATERIALS AND METHODS
Mutagenesis and plasmid construction. RSV Le regions containing substitutions or deletions were constructed by PCR with mutagenic oligonucleotides and inserted into plasmid C75, which was described previously (15). This plasmid carries a dicistronic minigenome with a single-nucleotide substitution at the penultimate nucleotide at the 5' end of the extragenic trailer (Tr) region, which inactivates the antigenomic promoter and limits replication to the antigenome synthesis step. The PCR products were inserted between the BtgI and XbaI sites of C75; the BtgI site is at the junction of the ribozyme and the Le region, and the XbaI site is at the junction between the NS1 nontranslated region and the open reading frame of the chloramphenicol acetyltransferase (CAT) reporter sequence. To insert spacer sequence into the Le region, an SphI site was introduced between nt 35 and 36 by performing PCR with a mutagenic primer and inserting the PCR product into the BstXI sites of C75. C75 has two BstXI sites, one naturally occurring at nt 35 to 46 of the RSV genomic sequence and one within the ribozyme. Thus, this modification introduced 6 nt between nt 35 and 36. Heterologous inserts of 66 or 106 nt were generated by PCR, amplifying nt 676 to 729 or 636 to 729 of the RSV N gene (numbering is with respect to the N GS signal) and inserting the products into the SphI site. The two inserts were ligated in opposite orientations from each other, and therefore the +66 and +106 minigenomes contain different spacer sequences. The C75 plasmid background was used for the experiments shown in Fig. 2 and 3A. For the experiments shown in Fig. 3B and 4 to 7, the mutant Le regions were inserted into a plasmid encoding a minigenome with a wild-type (wt) Tr region. This was done either by inserting the BtgI-XbaI fragment into plasmid MP-28, which has been described previously (15), or by inserting a PCR product containing the CAT 2 gene, the L nontranslated region, a wt Tr region, and the T7 promoter into the BglII and HindIII sites of the mutant C75 plasmids. The BglII site lies immediately downstream of the internal gene junction in the minigenome, and the HindIII site lies immediately downstream of the T7 promoter. In each of these mutant plasmids, all RSV-, T7 promoter-, and ribozyme-specific sequences that were generated by PCR were sequenced. To generate the tricistronic minigenome (see Fig. 6), the MP-28 plasmids containing the wt, 16-43S, and 16-34S Le regions (see Fig. 1) were modified by inserting a PCR product encoding the G-F sequence from nt 225 of the RSV G gene to nt 870 of the RSV F gene into the BglII site. The G-F gene junction in each of these mutants was sequenced.
Reconstituted minigenome transcription and replication. Minigenome plasmids were transfected into HEp-2 cells together with N, P, L and M2-1 plasmids as described previously (15), except that T7 polymerase was expressed from vaccinia virus MVA-T7 (41). In the experiments shown in Fig. 2 to 5, each transfection reaction was set up in duplicate: RNA was directly extracted from cells from one of the wells, and the cells in the other well were lysed with nonionic detergent and incubated with micrococcal nuclease (MCN) prior to RNA purification to digest unencapsidated RNA, as described previously (30). In the experiment shown in Fig. 6, parallel transfections in which the M2-1 protein was either included or omitted were carried out. Empty pTM1 vector was added to transfection mixes in which either the L polymerase or M2-1 protein expression plasmid was omitted to maintain a constant input plasmid concentration.
RNA isolation and Northern blot hybridization. RNA was isolated, electrophoresed, and transferred to nitrocellulose as described previously (15). Negative- or positive-sense 32P-labeled CAT-specific riboprobes were synthesized with T7 RNA polymerase. The template for the negative-sense probe was the minigenome-carrying plasmid C75 linearized with XbaI (15). The template for the positive-sense probe was PstI-linearized DM28 plasmid, which contains the CAT sequences from C75 inserted between the XbaI and PstI sites of pGEM3Zf(–) (Promega). Riboprobes were extracted with phenol and chloroform and hybridized to the Northern blot in a mixture of 6x SSC (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate), 2x Denhardt's solution, 0.5% sodium dodecyl sulfate (SDS), and 100 μg of sheared DNA per ml at 65°C for at least 12 h. The blots were washed in 2x SSC-0.1% SDS at 65°C for 2 h and then in 0.1x SSC-0.1% SDS at 65°C for 15 min.
The blots shown in Fig. 3B (upper panels), 6B, and 7A were hybridized with negative-sense 32P-labeled oligonucleotides. Oligonucleotides Le oligo (5' TTTATGCAAGTTTGTTGTACGCATTTTT), containing nt 7 to 34 of the wt Le, and LeS oligo (5' AAATACGTTCAAACAACATCGCATTTTT), containing nt 7 to 34 of the 16-34S mutant, were used to detect RNA encoded by wt Le or by Le with positions 16 to 34 substituted, respectively. Oligonucleotides were labeled by using [-32P]ATP and T4 polynucleotide kinase. The oligonucleotides were separated from unincorporated nucleotides by chromatography on a Sephadex G25 column. Probes were hybridized to Northern blots in a mixture of 6x SSC, 5x Denhardt's solution, 0.1% SDS, 0.05% sodium pyrophosphate, and 100 μg of tRNA per ml at 50°C for at least 12 h. Blots were washed in 100 ml of 6x SSC twice at room temperature for 15 min and twice at 50°C for 15 min. Phosphorimager analysis was carried out with a Molecular Imager FX and Quantity One quantitation software (Bio-Rad).
Primer extension analysis. A negative-sense CAT specific primer (5' GGGATATATCAACGGTGGTATATCCAGTG) was purified by polyacrylamide gel electrophoresis, and 100 ng of primer was end labeled with 32P as described above. The labeled primer was separated from unincorporated nucleotides by using a Sephadex G25 spin column (Amersham Biosciences). Total or MCN-treated RNA representing one-fifth of a well of cells was annealed to one-third of the 32P-labeled primer in 1x avian myeloblastosis virus cDNA synthesis buffer by heating the mixture to 65°C for 10 min and then cooling on ice for 5 min. The RNA-DNA hybrid was used as a template for reverse transcription with cloned avian myeloblastosis virus reverse transcriptase in 1x cDNA synthesis buffer, 1 mM dithiothreitol, deoxynucleoside triphosphates at 0.5 mM each, and 40 U of RNase inhibitor. Reverse transcription was carried out at 45°C for 1 h. The cDNA products were extracted with phenol and chloroform, precipitated with ethanol, and resuspended in 10 μl of loading buffer (95% formamide, 20 mM EDTA, 0.05% xylene cyanol). Five microliters of the cDNA products was electrophoresed on a 6% polyacrylamide gel containing 7 M urea and analyzed by autoradiography and phosphorimaging. The cDNAs corresponding to RNA initiated at the 3' terminus and the GS signal were initially identified by comparison with a sequencing ladder run in an adjacent well (14). Phosphorimager analysis was carried out with a Molecular Imager FX and Quantity One quantitation software (Bio-Rad).
RESULTS
Identification of transcription- and replication-specific sequences in the Le region. As described in the introduction, there is a cis-acting element contained within the first 11 nt at the 3' terminus of the Le that is important for both transcription and replication (15). To determine whether the remainder of the Le region is necessary for RNA synthesis, a series of mutant minigenomes in which various regions of the Le were replaced with complementary sequence was constructed (Fig. 1B, rows i to v). In each mutant, the first 15 nt of Le and nt 44 at the –1 position relative to the mRNA start signal were maintained as wt sequence to avoid an impact of the mutations on the functions of the cis-acting element at the 3' terminus of the genome and the GS signal, respectively. The Le mutations were introduced into plasmid C75, which carries a dicistronic minigenome containing a single-nucleotide substitution at the penultimate nucleotide in the 5' Tr region, as described previously (15). This mutation inactivates the antigenomic promoter, thus limiting minigenome replication to the antigenome synthesis step (30). The lack of minigenome amplification allows transcription and replication to be quantified as independent events. RSV RNA synthesis was reconstituted in HEp-2 cells in which minigenome RNA and the RSV N, P, L, and M2-1 proteins were expressed from transfected T7 plasmids (9, 16). The levels of mRNA and antigenome produced from the mutant minigenomes were measured by Northern blot analysis of total intracellular RNA with a negative-sense CAT-specific probe (Fig. 2A). The mRNA transcripts from the CAT 1 and CAT 2 genes are indicated as mRNA 1 and mRNA 2, respectively. To allow accurate quantitation of antigenome RNA, duplicate transfection cell lysates were treated with MCN prior to RNA purification. Encapsidated antigenome RNA is resistant to MCN treatment and is the highest-molecular-weight band in this panel (Fig. 2B). Some mRNA appears to be resistant to MCN treatment in this experiment. Apparent incomplete digestion of mRNA was typical and might have been due to incomplete solubilization of the lysate in nonionic detergent. Also, in this experiment, in which minigenome replication is blocked at the genome synthesis step, the reduced production of encapsidated RNAs might have resulted in an excess of free N protein, which could then have engaged in nonspecific encapsidation of mRNA. Duplicate blots were probed with a positive-sense CAT-specific probe as a control to confirm that similar levels of encapsidated minigenome template were produced by each of the minigenome cDNAs (data not shown).
The wt minigenome generated a large amount of mRNA and a small amount of antigenome (Fig. 2, lanes 1), as expected. It should be noted that on this particular Northern blot, the products from the wt minigenome in the total RNA fraction are somewhat overrepresented; in several other experiments, wt mRNA levels were more comparable to the levels shown in lanes 5 and 9, as indicated in the quantitative analysis shown below the blots. In contrast to the wt minigenome, mutant 16-43S, in which the entire region between the 3'-terminal cis-acting element and the GS signal was replaced, produced barely detectable levels of mRNA and antigenome (Fig. 2, lanes 3), demonstrating that the Le region downstream of the 3'-terminal promoter element contains sequences that are important for both transcription and replication. To map the locations of the cis-acting elements within this region of Le, minigenomes containing substitutions in nt 16 to 34 or 36 to 43 were generated. These regions were chosen based on the sequence alignments shown in Fig. 1A, which show that that the central region of the Le is well conserved with the antigenomic promoter, suggesting that it might be important for replication, and that the U-rich region at the end of the Le is well conserved in other pneumovirus Le regions but is absent in the antigenome promoter, suggesting that it might be important for transcription. Consistent with this supposition, minigenome 16-34S produced levels of mRNA similar to those produced by the wt minigenome (Fig. 2A, lane 5), indicating that the first 15 nt of Le, together with a U-rich region at the end of the Le and the GS signal, is sufficient to direct mRNA synthesis. However, this mutant replicated poorly (Fig. 2B, lane 5). A minigenome containing the converse mutation, in which the first 35 nt of Le were wt sequence and nt 36 to 43 were replaced, produced mRNA at low levels but antigenome at levels equivalent to, or greater than, wt (Fig. 2, lanes 7). Thus, these data show that in addition to the cis-acting sequence within nt 1 to 11, the RSV Le contains nonoverlapping regions that are important for transcription and replication: the central region of Le (nt 16 to 34) is necessary for efficient antigenome production, and the U-rich region at the end of Le (nt 36 to 43) is important for efficient mRNA synthesis. In an attempt to more precisely map the replication-specific sequence in the central region of Le, two additional substitution mutants were constructed. Lanes 8 of Fig. 2 show RNA from a minigenome containing bovine RSV (BRSV) Le sequence for positions 1 to 43. As shown in Fig. 1A, the human RSV and BRSV Le regions are highly conserved, but they diverge from nt 26 to 34; thus, this mutant allowed analysis of these nucleotides. Lanes 9 show RNA from a minigenome containing a substitution of complementary sequence in nt 16 to 26. Both of these mutants directed high levels of transcription, as expected. Minigenome 16-26S generated a low level of antigenome similar to that generated by the 16-34S mutant, indicating that this region is important for replication. Although the BRSV mutant replicated slightly more efficiently than the 16-26S and 16-34S mutants, its level of antigenome production was considerably lower than that of the wt. Therefore, these data show that there are replication-specific sequences extending from nt 16 to 34, with nt 16 to 26 being particularly significant.
The relative spacing of cis-acting sequences is important for transcription and replication. Having identified replication- and transcription-specific sequences in the Le region, it was of interest to determine whether the spacing of these elements is important. Deletions correlating with the regions of substituted sequence were introduced, as shown in Fig. 1B (rows vi to viii). Each of the deletion mutants directed lower levels of mRNA and antigenome synthesis than their corresponding substitution mutants (Fig. 2, compare lanes 2, 4, and 6 with lanes 3, 5, and 7). This demonstrated that the relative spacing between the 3' terminus and the GS signal is important for efficient mRNA synthesis and suggested that positioning the GS signal too close to the Le 3' end can interfere with replication.
To determine whether increasing the distance between transcription signal sequences affects RNA synthesis, 6, 66, or 106 nt of heterologous sequence was inserted between nt 35 and 36 of the wt Le, thus displacing the U-rich region and the GS signal from the 3' end of the minigenome, as illustrated in Fig. 1B (row ix). Analysis of total and MCN-treated RNAs with a negative-sense CAT-specific riboprobe showed that inserting 6 nt at position 34 had no significant effect on either transcription or replication (Fig. 3A, lanes 2 and 7), but as the length of the inserted sequence was increased to 66 or 106 nt, there was a significant decrease in mRNA and increase in antigenome (Fig. 3A, lanes 3, 4, 8, and 9).
The mRNA that was produced from the 106-nt insertion mutant migrated slightly more slowly than mRNA produced from the wt minigenome (compare lanes 1 and 4 of Fig. 3A), suggesting that it might be initiated at the 3' end of Le. Polyadenylated transcripts initiated at the 3' terminus of Le and extended to the first GE signal have previously been observed in RSV-infected cells and in the minigenome system (6, 23). To determine whether the mRNA detected in lane 4 contained Le-encoded sequence, Northern blot analysis was performed with an Le-specific oligonucleotide probe. Because of the relative insensitivity of an oligonucleotide probe, it was not possible to detect RNA produced from the nonreplicating minigenomes shown in Fig. 3A (data not shown). Therefore, the experiment was repeated with minigenomes with a wt Tr region, which can be amplified through multiple cycles of replication, resulting in an enhanced signal. Analysis of total RNA with a negative-sense, Le-specific probe showed that each minigenome produced two Le-containing RNAs (Fig. 3B, upper panels). The upper band is antigenome, and the lower band is mRNA 1 containing the complement of the Le, referred to as LeC-mRNA 1. Insertion of 6 nt had no significant effect on synthesis of either of these transcripts, but insertion of 66 or 106 nt of spacer sequence resulted in an increase in antigenome and LeC-mRNA 1 synthesis. Analysis of RNA purified from MCN-treated cell extracts showed that most of the LeC-mRNA 1 was digested (Fig. 3B, lanes 7 to 10), indicating that it was unencapsidated, as has been shown previously (14). The blot was stripped and reprobed with a negative-sense CAT specific riboprobe (Fig. 3B, lower panels) to allow comparison of the levels of mRNA detected by CAT- or Le-specific riboprobes. This analysis indicated that all detectable mRNA produced from the +106 mutant was attached to Le, suggesting that displacement of the transcription signals by 106 nt abolished transcription initiation at the GS signal. Thus, these data (i) confirm that the spacing of transcription signals is important for mRNA synthesis initiation from the GS signal and (ii) suggest that initiation at the 3' terminus of Le is subject to competition from the GS signal, which is aggravated if the GS is moved closer and alleviated if it is moved away.
Analysis of mRNA and antigenome initiation. The Northern blot analyses shown in Fig. 2 identified sequences necessary for synthesis of full-length antigenome and mRNA. However, Northern blotting does not provide information regarding the level or accuracy of initiation of these RNAs. To examine mRNA and antigenome initiation, primer extension analysis was carried out on total intracellular RNA with an excess of labeled, negative-sense primer that hybridized 130 nt from the 5' end of the antigenome and 86 nt from the GS sequence. Thus, this primer could be used to detect 3'-terminal and GS initiations within the same reaction (Fig. 4A and D). Primer extension products were barely detectable when nonreplicating minigenomes were used (data not shown), so this analysis was carried out on RNA produced from minigenomes that possessed a wt Tr region and thus were capable of amplification. To allow direct comparison with the levels of full-length antigenome and mRNA, aliquots of the same total RNA samples were analyzed by Northern blotting with a negative-sense CAT-specific riboprobe (Fig. 4B and E). MCN-treated samples prepared from duplicate wells were also analyzed to examine encapsidated antigenome levels (Fig. 4C and F).
Primer extension on RNA generated from the wt minigenome yielded two bands, corresponding to GS and 3'-terminal initiations (Fig. 4A and D, lane 2). Examination of GS initiations from the substitution mutants showed that the 16-34S, 16-26S, and BRSV minigenomes directed levels of GS initiations similar to those for the wt Le (Fig. 4A, lanes 4, 6, and 7), whereas minigenomes 16-43S and 36-43S directed considerably less (Fig. 4A, lanes 3 and 5). Thus, the relative levels of these initiations correlate with the levels of mRNA detected by Northern blot analysis (Fig. 4B). These data demonstrate that the GS signal and the 3'-terminal core promoter element are sufficient to direct accurate initiation at the GS sequence and confirm that nt 36 to 43 significantly augment levels of initiation at this site.
Analysis of bands representing RNA initiated at the 3' terminus of Le demonstrated that each of the substitution mutants directed efficient initiation compared to the wt minigenome (Fig. 4A, lanes 2 to 7). This was a surprising result given that minigenomes containing mutations involving nt 16 to 34 are significantly inhibited in their ability to produce full-length antigenome (Fig. 4C, lanes 3, 4, 6, and 7). These results were highly reproducible; for example, in three separate experiments the 16-34S mutant directed 3' initiation at 130% or more than wt but generated full-length antigenome at 26% or less. It is possible that the mutations in nt 16 to 34 rendered the antigenome RNA unstable, such that fragments that remained could be detected by primer extension but not by Northern blotting. However, this seems unlikely, and if such fragments existed, primer extension analysis should have yielded a more complex pattern. Therefore, these data suggest that the first 15 nt of Le contain all of the sequences required to direct antigenome initiation and that nt 16 to 34 are required specifically for elongation to produce full-length antigenome. To confirm that sequence within the first 15 nt is important for RNA synthesis initiation, RNA from a minigenome containing a C-to-A substitution at Le position 3 was examined. This substitution completely inhibits transcription and replication (15) (Fig. 4B and C, lanes 8) and eliminated the 3' and GS initiation bands detected by primer extension (Fig. 4A, lane 8), confirming that the 3' terminus of Le has the properties of a promoter.
The results obtained with the substitution mutants were confirmed with their deletion mutant counterparts. Analysis of the deletion mutants 16-43D and 36-43D by primer extension showed that the levels of GS initiations for these mutants were consistent with the levels of full-length mRNA detected by Northern blotting (Fig. 4D and E, lanes 3 and 4). Similar to the substitution mutants, both of these minigenomes directed initiation from the 3' terminus (Fig. 4D, lanes 3 and 4). Note that the sizes of the bands representing 3' initiations are reduced for the deletion mutants compared to the wt, as would be expected. Despite its ability to direct 3' initiation, minigenome 16-43D did not generate detectable full-length antigenome (Fig. 4F, lane 3), whereas in contrast, there was a good correlation between 3' initiation and full-length antigenome production for the 36-43D mutant (Fig. 4D and F, lanes 4). These data are consistent with the hypothesis that nt 16 to 34 contain a sequence that is essential for elongation to produce full-length antigenome.
nt 16 to 34 are important for antigenome encapsidation. Studies with SeV have suggested that antigenome elongation is dependent on concurrent encapsidation (17, 38). Thus, a possible function of nt 16 to 34 is to mediate encapsidation of the nascent antigenome RNA, either by recruiting encapsidation factors to the genome template or by templating for an encapsidation nucleation signal in the encoded antigenome. To test whether nt 16 to 34 are involved in encapsidation, total and MCN-treated RNA samples from duplicate transfections were analyzed by primer extension (Fig. 5). Analysis of MCN-treated RNAs produced by wt and mutant minigenomes showed that in each case at least some of the 3' initiation transcript could be detected (Fig. 5C and D). A proportion of transcripts initiated at the GS signal were also detected, indicating that MCN treatment likely did not completely digest unencapsidated RNA. This was not unexpected given that Northern blot analysis of MCN-treated RNA often showed a smear of partially digested mRNA, as described previously. Despite the detection of residual unencapsidated RNA, it was possible to determine the relative levels of encapsidation for each of the mutants by comparing their levels of 3' initiation transcripts to those of the wt in the total RNA and MCN-treated RNA samples. Analysis of minigenome 16-34S by this approach showed that in the total RNA fraction, the level of 3' initiation transcript was 141% of the level produced from the wt minigenome (Fig. 5A, lane 3). In contrast, in the MCN fraction, the level of 3' initiation transcript from the 16-34S mutant was only 34% of wt (Fig. 5C, lane 3). This demonstrates that the transcripts generated from the 3' terminus of the 16-34S mutant were encapsidated less efficiently than transcripts from the wt Le minigenome. Similar analysis of other minigenomes with substitutions covering nt 16 to 34 indicated that each of these mutations was associated with reduced encapsidation of the 3' initiation products (Fig. 5A and C, lanes 4 to 6). The difference was less significant in the case of BRSV Le (Fig. 5A and C, lanes 5), consistent with its slightly higher level of antigenome production compared to minigenomes 16-34S and 16-26S (as shown in Fig. 2 and 4). These data suggest that nt 16 to 34 are important for encapsidation of the transcripts generated from the 3' end of Le. As a control to confirm that the sequences required for encapsidation are located within the first 34 nt of Le, RNA from mutant 36-43D was analyzed in parallel with that from mutant 16-34S (Fig. 5B and D). The results for 16-34S (Fig. 5B and D, lanes 2) are similar to those shown in Fig. 5A and C. In contrast, mutant 36-43D produced 3' initiation transcripts at 36% of wt levels in the total RNA fraction and at 35% of wt levels in the MCN-treated fraction (Fig. 5B and D, lanes 3). Thus, this result confirms that while deletion of nt 36 to 43 inhibited initiation from the 3' terminus, it had no effect on encapsidation of the transcripts produced.
Abortive replicases can complete synthesis of the first gene and default to a transcription mode. The results shown in Fig. 4 suggest that mutation of nt 16 to 34 results in nonprocessive replication from the 3' terminus, as evidenced by the presence of 3' initiation product and absence of full-length antigenome. However, the abortive transcripts detected by primer extension analysis were at least 130 nt in length. Therefore, it was of interest to determine how far the polymerase was able to extend these transcripts. For this analysis, wt Le and mutant 16-34S and 16-43S Le sequences were introduced into a tricistronic minigenome backbone, in which the first gene contained CAT sequence, the second contained part of the RSV G gene, and the third contained part of the F gene fused with additional CAT sequence (Fig. 6A). The first and second gene junctions were derived from the N-P and G-F gene junction sequences, respectively. The minigenomes were used as templates in the intracellular transcription-replication assay, as described above, except that reactions were carried out both with and without the M2-1 expression plasmid. To examine RNA initiated from the 3' end of Le, total intracellular RNA was analyzed by Northern blotting with negative-sense, Le-specific oligonucleotide probes, one of which (Le oligo) was specific to RNA encoded by the wt Le and the other of which (LeS oligo) was specific to RNA encoded by the 16-34S and 16-43S Le sequences. Examination of RNA produced in the presence of M2-1 showed that the wt minigenome and the 16-43S and 16-34S mutants generated two LeC-containing RNAs, namely, antigenome and the LeC-mRNA 1 readthrough mRNA (Fig. 6B, lanes 1, 3, and 5). These same two species were detected when the blot was stripped and reprobed with a CAT-specific riboprobe, and in addition, the LeC-mRNA 1 band was augmented by the monocistronic mRNA 1, which was not resolved from the LeC-mRNA 1 species (Fig. 6C, lanes 1, 3, 5, and 9). The CAT-specific probe also detected mRNA 3, which migrated between the antigenome and mRNA 1 and was confirmed as mRNA 3 with an F-specific probe (data not shown). Examination of RNAs detected with the Le-specific probes showed no evidence of a smear of LeC-containing transcripts above or below the LeC-mRNA band, as would be expected if LeC-containing transcripts were terminated at various distances from the 3' terminus to a considerable extent (Fig. 6B, lanes 1, 3, and 5). These results suggest that in the presence of M2-1, a significant proportion of the abortive replication transcripts produced by 16-34S and 16-43S Le sequences were extended to the end of the first gene to produce LeC-mRNA 1.
To determine whether the polymerase that generated LeC-mRNA 1 was able to continue stop-start transcription along the length of the minigenome, the ratio of the third to the first gene transcripts, detected with the CAT-specific riboprobe, was analyzed. Because the 16-43S mutant generated very little mRNA 1 from the first GS signal (see, e.g., Fig. 4A) but produced amounts of LeC-mRNA 1 similar to those produced by the 16-34S minigenome, its gene 1 transcripts contained a much greater proportion of LeC-mRNA than the gene 1 transcripts produced from the 16-34S minigenome (compare Fig. 6B and C, lanes 3 and 5). If polymerase that synthesized LeC-mRNA 1 were unable to continue transcription beyond the first GE signal, the 16-43S mutant would be expected to produce very low levels of mRNA 3 relative to mRNA 1 compared to the 16-34S mutant. Examination of mRNAs produced in the presence of M2-1 with the CAT-specific riboprobe allowed comparison of the total population of mRNA 1 transcripts (including LeC-mRNA 1) and mRNA 3 transcripts on the same blot. Although the 16-43S mutant produced lower levels of mRNAs 1 and 3 than the 16-34S and wt minigenomes, as expected, the ratios of mRNA 3 to mRNA 1 were similar in each case (Fig. 6D). Therefore, this result indicates that polymerase that generates LeC-mRNA 1 can continue sequential stop-start transcription along the length of the genome.
RNA produced in the absence of M2-1 was also analyzed to determine the effect of this protein on RNA synthesis from the mutant minigenomes (Fig. 6B and C, lanes 2, 4, and 6). This analysis indicated that for each minigenome, a small amount of full-length LeC-mRNA 1 was generated, but a significant proportion was present as a smear of prematurely terminated product. Note that the premature termination products observed in Fig. 6B were less evident in Fig. 6C because in the latter case the more stringent hybridization and washing conditions necessary with a riboprobe, versus an oligonucleotide probe, were used. Thus, these data indicate that the synthesis of full-length LeC-mRNA 1 is dependent on M2-1 activity. Analysis of mRNAs 1 and 3 produced in the absence of M2-1 showed that a small amount of full-length mRNA was produced from the CAT 1 gene of each of the minigenomes (Fig. 6C, lanes 2, 4, 6, and 10). This was expected because the first gene is relatively short and a proportion of polymerase is able to complete its synthesis, as noted previously (13). However, for each minigenome, the third gene transcript was barely detectable in the absence of M2-1. Taken together, these data suggest that M2-1 protein activity is necessary to allow abortive replicases to continue RNA synthesis to the end of the first gene to produce LeC-mRNA 1 and then continue mRNA synthesis along the remainder of the genome. It should be noted that in this particular experiment there was variation in the levels of antigenome such that accumulation appeared to be reduced in the presence of M2-1. However, this likely reflects uneven transfer of the antigenome RNA in this experiment: for example, the RNA samples shown in Fig. 6C, lanes 1 and 2, are the same RNA samples as in lanes 9 and 10, which were electrophoresed and transferred from the same gel, and yet there was variation in the apparent abundance of antigenome.
As an additional step to confirm that polymerase that generated LeC-mRNA 1 could continue sequential transcription, we reexamined RNA synthesis by the +106-nt insertion mutant (Fig. 1B). This mutant had the desirable feature of producing a relatively large amount of LeC-mRNA 1 and little or no monocistronic mRNA 1, as shown in Fig. 3B. Here we examined whether this mutant directed the synthesis of mRNA 2, which would be evidence that sequential transcription followed the synthesis of LeC-mRNA 1. RNA produced in the presence of M2-1 from the wt minigenome and the +106-nt insertion mutant was analyzed with Le oligo to identify LeC-containing RNAs (Fig. 7A). Analysis of the same blot with a CAT-specific riboprobe showed that all detectable mRNA 1 from the insertion mutant migrated similarly to RNA detected with Le oligo and was of increased molecular weight compared to that produced from the wt minigenome, indicating that it was initiated at the 3' end of the Le region (Fig. 7B, lane 3). Importantly, analysis with the CAT-specific probe showed that the 106-nt insertion mutant generated mRNA 2 (Fig. 7B, lane 3). Quantitation of the total mRNA 1 transcripts (including LeC-mRNA 1) and mRNA 2 transcripts, shown in lanes 2 and 3 of Fig. 7B, demonstrated that the ratio between them was similar for the wt minigenome and the +106-nt insertion mutant (Fig. 7C), although it should be noted that the small size of mRNA 2 and its diffuse nature on the Northern blot precluded accurate quantitation of this mRNA. Thus, these data are consistent with a model in which polymerase that initiates at the 3' end of the Le region and terminates at the first GE signal is able to reinitiate mRNA synthesis at a downstream GS signal.
DISCUSSION
Elucidating the functions of cis-acting sequences in the RSV Le is an important step towards understanding the initial events involved in mononegavirus RNA synthesis. In a previous study it was shown that the 3' end of Le contains residues that are important for both transcription and replication. The present study identified nonoverlapping transcription- and replication-specific sequences in the remainder of Le and completes a broad mapping of the RSV Le region. The Le is split broadly into three regions: the 3'-terminal 15 nt contain sequence that is important for transcription and replication, the central region (nt 16 to 34) contains sequences important only for replication, and the downstream end of Le (nt 36 to 43) is important only for transcription. The functions of these sequences and the GS signal were further dissected, identifying replication initiation and elongation signals and demonstrating that the GS signal, in conjunction with Le nt 1 to 15, is sufficient for accurate initiation of mRNA synthesis.
The 3'-terminal 15 nt of Le act as an efficient promoter for RNA replication. This was highlighted in particular by analysis of mutants 16-43S and 16-43D, which were able to direct levels of 3' initiation comparable to that of the wt Le region (Fig. 4). The minimal region required for promoter activity was not determined in this study, but previously it was shown that most single-nucleotide substitutions in positions 1 to 11 significantly inhibited replication (15). A number of these substitutions also inhibited transcription, suggesting that these nucleotides comprise a common promoter element. In the case of VSV, it has been shown that a synthetic template containing 17 nt of wt sequence is a functional promoter in vitro (34), and the first 15 nt of wt sequence can act as a promoter in an intracellular minigenome assay (39). The first 12 nt of the genome and antigenome are conserved in all of the Paramyxovirinae, and these nucleotides have been shown to be critical for RNA replication in parainfluenza virus type 3 (20), suggesting that this is the typical size of the 3'-terminal element of the mononegavirus replication promoter. In the case of members of the Paramyxovirinae, the replication promoter contains a second essential cis-acting element located within the first gene (20, 21, 29, 37), but this additional element is not present in RSV or VSV.
The central portion of the RSV Le (nt 16 to 34) was required for efficient elongation of antigenome initiated at the 3' terminus of Le to produce full-length molecules and also was important for efficient encapsidation of the antigenome. It is tempting to speculate that Le nt 16 to 34 encode a sequence in the complementary nascent antigenome that binds N and acts as an encapsidation nucleation site. However, an attempt to identify an N-binding site in the RSV Le-encoded transcript failed to show evidence for a specific interaction (28). In addition, a previous study showed that LeC-mRNA transcripts were not encapsidated even under conditions in which N protein was in excess (14), suggesting that in RSV, encapsidation might not be initiated simply by interaction of N with a cis-acting sequence at the 5' end of the RNA. One possibility is that nt 16 to 34 encode an encapsidation nucleation site but that encapsidation is mediated by the polymerase during RNA synthesis. Alternatively, it is possible that these nucleotides constitute a signal in the genome template, rather than nascent antigenome, that is important for recruiting a polymerase that is able to mediate encapsidation, for example, a polymerase that is preloaded with N protein, which has been suggested to be the active form of the VSV replicase (18, 31).
The region at the downstream end of Le (nt 36 to 43) was necessary for efficient transcription. The specific nucleotides involved were not identified, but alignment of pneumovirus Le sequences shows that a U stretch immediately upstream of the first GS signal is highly conserved (Fig. 1A). Seven of the eight residues contained in nt 36 to 43 of the RSV Le are U, suggesting that this U-rich nature is an important feature. The finding that nucleotides at the end of Le are important for transcription is consistent with findings for VSV, SeV, and parainfluenza virus type 3 (4, 20, 27, 39), suggesting that the sequences immediately upstream of the promoter-proximal GS signal are important for many mononegaviruses. However, it should be emphasized that although nt 36 to 43 increased the efficiency of initiation from the GS signal, this region was not required for correct initiation, synthesis of complete mRNA, or the initiation of stop-start sequential transcription. It is interesting that in contrast to the results presented here, the sequences required for VSV transcription are located throughout the Le region (27, 39), suggesting that there are differences in the organization of the transcription promoters for RSV and VSV.
Both GS and 3'-terminal initiations were found to be inhibited if the GS signal was moved closer to the 3' terminus (Fig. 2 and 4), suggesting that the close proximity of the promoter and GS signal either inhibits polymerase binding to the template or causes the polymerase to stall shortly after initiation. If the distance between the 3' end of the Le and nt 36 to 43 and the GS signal was increased, mRNA synthesis was inhibited but antigenome and LeC-mRNA syntheses were augmented (Fig. 3). The decrease in transcription could be either because the U-rich region and the GS signal must be within a certain distance from the promoter for the polymerase to be able to terminate and reinitiate RNA synthesis (according to the model in which transcription is mediated by a polymerase that initiates at the 3' terminus of Le) or because these elements must be spaced appropriately for the polymerase to recognize them simultaneously (according to the model in which transcription initiates directly at the GS signal). The increase in antigenome and LeC-mRNA synthesis suggests that GS and 3' initiations are competitive processes. This is also suggested by the fact that a mutant in which nt 36 to 43 were replaced generated a higher level of antigenome than the wt Le (Fig. 2B, lane 7). Competition between synthesis from the 3' terminus and the GS signal has also been observed in SeV (26). Competition could exist because the same polymerase is used for both processes or because there is steric hindrance between replicases and transcriptases as they bind to their respective initiation sites. An experiment in a previous study, in which a spacer sequence was inserted into Le, showed no increase in LeC-mRNA and antigenome with increasing spacer, although mRNA synthesis was completely inhibited (14). The difference between the experiment described here and the previous study is that in the present study, nt 36 to 43 and the GS signal were displaced relative to the 3' terminus, whereas in the previous study, nt 36 to 43 were not displaced. Therefore, it appears that although moving the GS signal away from the 3' terminus causes a decrease in transcription, the U-rich region immediately upstream of the GS signal also must be displaced to alleviate competition on replication, suggesting that the U-rich region interacts directly with the polymerase.
The major product containing Le-encoded sequence from the minigenomes with substitution in nt 16 to 34 was LeC-mRNA 1. Indeed, this RNA was produced in considerable excess over the full-length antigenome. This suggests that LeC-mRNA transcripts are abortive replication products that are produced if the nascent RNA is not encapsidated (Fig. 6). LeC-mRNA has previously been detected in RSV-infected cells, indicating that it is a bona fide product of RSV RNA synthesis (6, 23). Results presented here (Fig. 6 and 7) and in a previous study (23) indicate that polymerase that synthesizes an LeC-mRNA transcript is able to continue stop-start RNA synthesis along the remainder of the genome. In the present study it was not determined whether the transcripts from downstream genes are capped and methylated at their 5' termini, but the study by Kuo and coworkers (23) indicated that polymerase that synthesizes LeC-mRNA goes on to produce translatable mRNA from a downstream gene. Therefore, taken together these data suggest that a replicase can default to a transcription mode in the absence of concurrent encapsidation and that this conversion is an efficient event. In contrast to the situation presented here, SeV abortive replicases were unable to extend the RNA beyond 358 nt (38). The difference in polymerase processivity between SeV and RSV might be due to the M2-1 transcription elongation protein, which is found in RSV but not in SeV.
As described in the introduction, there are two models for transcription initiation, one in which transcription is initiated at the 3' terminus of Le and one in which it is initiated directly at the GS signal. The data presented here could be interpreted to fit either of these initiation models and thus do not resolve this question. For example, nt 36 to 43 and the GS signal either could be involved in reinitiation of a polymerase that initiated at the 3' end of Le or could be a component of a distinct transcription promoter at which the polymerase initiates internally. With regard to competition between transcription and replication, as demonstrated by the effect of increasing the distance between the 3' end and the GS signal, this could also be interpreted to fit either model. For example, if polymerase initiates transcription at the 3' terminus of Le and subsequently terminates (possibly at a certain distance rather than at a specific cis-acting signal) and scans to locate the GS signal, then increasing the distance between the promoter and GS signal could decrease the likelihood that the polymerase will initiate at the GS signal and increase the likelihood that polymerases will reinitiate at the 3' terminus of the genome by scanning backwards, which is a known property of the RSV polymerase (12). On the other hand, if transcription is initiated directly at the GS signal by a distinct transcriptase complex, displacing this sequence could prevent transcriptases from forming a stable interaction with the template, thus allowing more replicases to bind to the Le region and initiate RNA synthesis at the 3' terminus. While these data do not clearly distinguish between these models for normal transcription initiation at the GS signal, the evidence suggests that the RSV polymerase can apparently shift from replication to transcription after the initiation of RNA synthesis, in a manner that seems to be controlled by encapsidation.
ACKNOWLEDGMENTS
We thank Bernard Moss for providing MVA-T7, Karen Jackson for assistance in subcloning some of the substitution mutants, and Vanessa Cowton and Frances Fuller-Pace for helpful discussions.
This work was funded by a Wellcome Trust grant (reference number 065568) to R.F.
REFERENCES
Barr, J. N., S. P. J. Whelan, and G. W. Wertz. 1997. cis-acting signals involved in termination of vesicular stomatitis virus mRNA synthesis include the conserved AUAC and the U7 for polyadenylation. J. Virol. 71:8718-8725.
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.
Blumberg, B. M., C. Giorgi, and D. Kolakofsky. 1983. N protein of vesicular stomatitis virus selectively encapsidates leader RNA in vitro. Cell 32:559-567.
Calain, P., and L. Roux. 1995. Functional characterisation of the genomic and antigenomic promoters of Sendai virus. Virology 211:163-173.
Chuang, J. L., and J. Perrault. 1997. Initiation of vesicular stomatitis virus mutant polR1 transcription internally at the N gene in vitro. J. Virol. 71:1466-1475.
Collins, P. L., and G. W. Wertz. 1985. Nucleotide sequences of the 1B and 1C non-structural protein mRNAs of human respiratory syncytial virus. Virology 143:442-451.
Collins, P. L., L. E. Dickens, A. Buckler-White, R. A. Olmsted, M. A. Spriggs, E. Camargo, and K. V. W. Coelingh. 1986. Nucleotide sequences for the gene junctions of human respiratory syncytial virus reveal distinctive features of intergenic structure and gene order. Proc. Natl. Acad. Sci. USA 83:4594-4598.
Collins, P. L., M. A. Mink, and D. S. Stec. 1991. Rescue of synthetic analogs of respiratory syncytial virus genomic RNA and effect of truncations and mutations on the expression of a foreign reporter gene. Proc. Natl. Acad. Sci. USA 88:9663-9667.
Collins, P. L., M. G. Hill, J. Cristina, and H. Grosfeld. 1996. Transcription elongation factor of respiratory syncytial virus, a nonsegemented negative strand RNA virus. Proc. Natl. Acad. Sci. USA 93:81-85.
Collins, P. L., R. M. Chanock, and B. R. Murphy. 2001. Respiratory syncytial virus, 1433-1485. In D. M. Knipe and P. M. Howley (ed.), Fields virology, 4th ed. Lippincott Williams & Wilkins, Philadelphia, Pa.
Emerson, S. U. 1982. Reconstitution studies detect a single RNA polymerase entry site on the vesicular stomatitis virus genome. Cell 31:635-642.
Fearns, R., and P. L. Collins. 1999. Model for polymerase access to the overlapped L gene of respiratory syncytial virus. J. Virol. 73:388-397.
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.
Fearns, R., P. L. Collins, and M. E. Peeples. 2000. Functional analysis of the genomic and antigenomic promoters of human respiratory syncytial virus. J. Virol. 74:6006-6014.
Fearns, R., M. E. Peeples, and P. L. Collins. 2002. Mapping the transcription and replication promoters of respiratory syncytial virus. J. Virol. 76:1663-1672.
Grosfeld, H., M. G. Hill, and P. L. Collins. 1995. RNA replication by respiratory syncytial virus (RSV) is directed by the N, P, and L proteins: transcription also occurs under these conditions but requires RSV superinfection for efficient synthesis of full-length mRNA. J. Virol. 69:5677-5686.
Gubbay, O., J. Curran, and D. Kolakofsky. 2001. Sendai virus genome synthesis and assembly are coupled: a possible mechanism to promote viral RNA polymerase processivity. J. Gen. Virol. 82:2895-2903.
Gupta, A. K., D. Shaji, and A. K. Banerjee. 2003. Identification of a novel tripartite complex involved in replication of vesicular stomatits virus genome RNA. J. Virol. 77:732-738.
Hallak, L. K., P. L. Collins, W. Knudson, and M. E. Peeples. 2000. Iduronic acid-containing glycosaminoglycans on target cells are required for efficient respiratory syncytial virus infection. Virology 271:264-275.
Hoffman, M. A., and A. K. Banerjee. 2000. Precise mapping of the replication and transcription promoters of human parainfluenza virus type 3. Virology 269:201-211.
Keller, M. A., and G. D. Parks. 2003. Positive- and negative-acting signals combine to determine differential RNA replication from the paramyxovirus simian virus 5 genomic and antigenomic promoters. Virology 306:347-358.
Kolakofsky, D., P. Le Mercier, F. Iseni, and D. Garcin. 2004. Viral RNA polymerase scanning and the gymnastics of Sendai virus RNA synthesis. Virology 318:463-473.
Kuo, L., H. Grosfeld, J. Cristina, M. G. Hill, and P. L. Collins. 1996. Effect of mutations in the gene-start and gene-end sequence motifs on transcription of monocistronic and dicistronic minigenomes of respiratory syncytial virus. J. Virol. 70:6892-6901.
Kuo, L., R. Fearns, and P. L. Collins. 1997. Analysis of the gene start and gene end signals of human respiratory syncytial virus: quasi-templated initiation at position 1 of the encoded mRNA. J. Virol. 71:4944-4953.
Lamb, R. A., and D. Kolakofsky. 2001. Paramyxoviridae: the viruses and their replication, p. 1305-1340. In D. M. Knipe, and P. M. Howley (ed.), Fields virology, 4th ed. Lippincott Williams & Wilkins, Philadelphia, Pa.
Le Mercier, P., D. Garcin, E. Garcia, and D. Kolakofsky. 2003. Competition between the Sendai virus N mRNA start site and the genome 3'-end promoter for viral polymerase. J. Virol. 77:9147-9155.
Li, T., and A. K. Pattnaik. 1999. Overlapping signals for transcription and replication at the 3' terminus of the vesicular stomatitis virus genome. 73:444-452.
Murphy, L. B., C. Loney, J. Murray, D. Bhella, P. Ashton, and R. P. Yeo. 2003. Investigations into the amino-terminal domain of the respiratory syncytial virus nucleocapsid protein reveal elements important for nucleocapsid formation and interaction with the phosphoprotein. Virology 307:143-153.
Murphy, S. K., and G. D. Parks. 1999. RNA replication for the paramyxovirus simian virus 5 requires an internal repeated CGNNNN sequence motif. J. Virol. 73:805-809.
Peeples, M. E., and P. L. Collins. 2000. Mutations in the 5' trailer region of a respiratory syncytial virus minigenome which limit RNA replication to one step. J. Virol. 74:146-155.
Qanungo, K. R., D. Shaji, M. Mathur, and A. K. Banerjee. 2004. Two RNA polymerase complexes from vesicular stomatitis virus-infected cells that carry out transcription and replication of genome RNA. Proc. Natl. Acad. Sci. USA 101:5952-5957.
Rose, J. K., and M. A. Whitt. 2001. Rhabdoviridae: the viruses and their replication, p. 1221-1244. In D. M. Knipe and P. M. Howley (ed.), Fields virology, 4th ed. Lippincott Williams & Wilkins, Philadelphia, Pa.
Samal, S. K., and P. L. Collins. 1996. RNA replication by a respiratory syncytial virus RNA analog does not obey the rule of six and retains a nonviral trinucleotide extension at the leader end. J. Virol. 70:5075-5082.
Smallwood, S., and S. A. Moyer. 1993. Promoter analysis of the vesicular stomatitis virus RNA polymerase. Virology 192:254-263.
Stillman, E. A., and M. A. Whitt. 1997. Mutational analyses of the intergenic dinucleotide and the transcriptional start sequence of vesicular stomatitis virus (VSV) define sequences required for efficient termination and initiation of VSV transcripts. J. Virol. 71:2127-2137.
Stillman, E. A., and M. A. Whitt. 1999. Transcript initiation and 5' end modification are separable events during vesicular stomatitis virus transcription. J. Virol. 73:7199-7209.
Tapparel, C., D. Maurice, and L. Roux. 1998. The activity of Sendai virus genomic and antigenomic promoters requires a second element past the leader template regions: a motif (GNNNNN)3 is essential for replication. J. Virol. 72:3117-3128.
Vidal, S., and D. Kolakofsky. 1989. Modified model for the switch from Sendai virus transcription to replication. J. Virol. 63:1951-1958.
Whelan, S. P. J., and G. W. Wertz. 1999. Regulation of RNA synthesis by the genomic termini of vesicular stomatitis virus: identification of distinct sequences essential for transcription, but not replication. J. Virol. 73:297-306.
Whelan, S. P. J., and G. W. Wertz. 2002. Transcription and replication initiate at separate sites on the vesicular stomatitis virus genome. Proc. Natl. Acad. Sci. USA 99:9178-9183.
Wyatt, L. S., B. Moss, and S. Rozenblatt. 1995. Replication-deficient vaccinia virus encoding bacteriophage T7 RNA polymerase for transient gene expression in mammalian cells. Virology 210:202-205.(David R. McGivern, Peter )