Structural and mutational analyses of cis-acting sequences in the 5'-untranslated region of satellite RNA of bamboo mosaic potexvirus

The satellite RNA of Bamboo mosaic virus (satBaMV) contains on open reading frame for a 20-kDa protein that is flanked by a 5'-untranslated region (UTR) of 159 nucleotides (nt) and a 3'-UTR of 129 nt. A secondary structure was predicted for the 5'-UTR of satBaMV RNA, which folds into a large stem-loop (LSL) and a small stem-loop. Enzymatic probing confirmed the existence of LSL (nt 8-138) in the 5'-UTR. The essential cis-acting sequences in the 5'-UTR required for satBaMV RNA replication were determined by deletion and substitution mutagenesis. Their replication efficiencies were analyzed in Nicotiana benthamiana protoplasts and Chenopodium quinoa plants coinoculated with helper BaMV RNA. All deletion mutants abolished the replication of satBaMV RNA, whereas mutations introduced in most of the loop regions and stems showed either no replication or a decreased replication efficiency. Mutations that affected the positive-strand satBaMV RNA accumulation also affected the accumulation of negative-strand RNA; however, the accumulation of genomic and subgenomic RNAs of BaMV were not affected. Moreover, covariation analyses of natural satBaMV variants provide substantial evidence that the secondary structure in the 5'-UTR of satBaMV is necessary for efficient replication.     

Two cis-acting elements in negative RNA strand of Hepatitis C virus involved in synthesis of positive RNA strand in vitro

Sequences at the 3'-ends of both positive and negative strands of Hepatitis C virus (HCV) RNA harbor cis-acting elements required for RNA replication. However, little is known about the properties of the negative RNA strand as a template for the synthesis of positive RNA strand. In this study, a purified recombinant HCV RNA-dependent RNA polymerase (RdRp) was used to investigate the synthesis of positive RNA strand using the 3'-terminal region of negative RNA strand ((-)3'T RNA) as template. A mutagenesis analysis was performed to evaluate the role of the 3'-proximal stem-loop and the first 3'-cytidylate (3'C) of the negative RNA strand in the synthesis of the positive RNA strand. A negative RNA strand of wild type (wt) HCV as template was able to direct the synthesis of a full-length positive RNA strand. Deletion of the 3'-proximal stem-loop resulted in an approximately 90% decrease in RNA synthesis. Disruption of the 3'-proximal stem-loop structure by nucleotide substitutions led to a 70-80% decrease in RNA synthesis. However, the restoration of the stem-loop by compensatory mutations in the stem region restored also the RNA synthesis. Likewise, the deletion or substitution of the first 3'C by guanylate (G) led to a 90% decrease in the RNA synthesis; while the substitution by adenylate (A) or uridylate (U) resulted in a 60-80% decrease in the RNA synthesis only. These findings demonstrate that the 3'-proximal stem-loop and the first 3'C of the negative RNA strand of HCV are two cis-acting elements involved in the synthesis of the positive RNA strand.     

3'-Terminal RNA secondary structures are important for accumulation of tomato bushy stunt virus DI RNAs

The plus-strand RNA genome of tomato bushy stunt virus (TBSV) contains a 351-nucleotide (nt)-long 3'-untranslated region. We investigated the role of the 3'-proximal 130 nt of this sequence in viral RNA accumulation within the context of a TBSV defective interfering (DI) RNA. Sequence comparisons between different tombusviruses revealed that the 3' portion of the 130-nt sequence is highly conserved and deletion analysis confirmed that this segment is required for accumulation of DI RNAs in protoplasts. Computer-aided sequence analysis and in vitro solution structure probing indicated that the conserved sequence consists of three stem-loop (SL) structures (5'-SL3-SL2-SL1-3'). The existence of SLs 1 and 3 was also supported by comparative secondary structure analysis of sequenced tombusvirus genomes. Formation of the stem regions in all three SLs was found to be very important, and modification of the terminal loop sequences of SL1 and SL2, but not SL3, decreased DI RNA accumulation in vivo. For SL3, alterations to an internal loop resulted in significantly reduced DI RNA levels. Collectively, these data indicate that all three SLs are functionally relevant and contribute substantially to DI RNA accumulation. In addition, secondary structure analysis of other tombusvirus replicons and related virus genera revealed that a TBSV satellite RNA and members of the closely related genus Aureusvirus (family Tombusviridae) share fundamental elements of this general structural arrangement. Thus, this secondary structure model appears to extend beyond tombusvirus genomes. These conserved 3'-terminal RNA elements likely function in vivo by promoting and/or regulating minus-strand synthesis.     

Two classes of subgenomic RNA of grapevine virus A produced by internal controller elements

Grapevine virus A (GVA), a species of the recently established genus Vitivirus, consists of an approximately 7.3-kb single-stranded RNA genome of positive polarity, organized into five open reading frames (ORFs). The virus, which is closely associated with the grapevine rugose wood disease complex, has been poorly investigated genetically. We explored the production of viral RNAs in a GVA-infected Nicotiana benthamiana herbaceous host and characterized one nested set of three 5'-terminal sgRNAs of 5.1, 5.5, and 6.0 kb, and another, of three 3'-terminal sgRNAs of 2.2, 1.8, and 1.0 kb that could serve for expression of ORFs 2-3, respectively. Neither 3'- nor 5'-terminal sgRNAs, which would correspond to ORF5, was detected, suggesting that expression of this ORF occurs via a bi- or polycistronic mRNA. The 5'-terminal sgRNAs were abundant in dsRNA-enriched extracts. Cloning and sequence analysis of the 3' end of 5.5-kb 5'-terminal sgRNA and the 5' end of the 1.8-kb 3'-terminal sgRNA suggested that a mechanism other than specific cleavage was involved in production of these sgRNAs. Apparently, the production of the 5'- and 3'-terminal sgRNAs was controlled by sequences upstream of the 5'-terminus of each of ORFs 2-4. Detection of both plus and minus strands of the 5'- and 3'-terminal sgRNAs, though in different levels of accumulation, suggested that each of these cis-acting elements is involved in production of four RNAs: a 3'-terminal plus-strand sgRNA which could act as an mRNA, the corresponding 3'-terminal minus-strand RNA, a 5'-terminal plus-strand sgRNA, and the corresponding 5'-terminal minus-strand RNA.     

The RNAs of cucumoviruses: 3'-terminal sequence analysis of two strains of tomato aspermy virus

The sequences of 189 residues from the 3' terminus of three RNAs of one strain and two RNAs of another strain of tomato aspermy virus have been determined; there was almost complete sequence homology between the RNAs. A base-paired transfer RNA-like structure is proposed for tomato aspermy virus RNAs which is similar in many aspects to the structure proposed for the 3'-terminal 172 residues of RNA 4 of the Q-strain of cucumber mosaic virus (R. H. Symons, Nucleic Acids Res.7, 825-837, 1979). These 172 residues of cucumber mosaic virus RNA 4 can be aligned to show 73% sequence homology with the 3'-terminal 189 residues of the tomato aspermy virus RNAs.     

cis-acting elements at opposite ends of the Citrus tristeza virus genome differ in initiation and termination of subgenomic RNAs

Citrus tristeza virus (CTV), a member of the Closteroviridae with a plus-stranded genomic RNA of approximately 20 kb, produces 10 3'-coterminal subgenomic (sg) RNAs that serve as messenger (m)RNAs for its internal genes. In addition, a population of 5'-terminal sgRNAs of approximately 700 nts are highly abundant in infected cells. Previous analysis demonstrated that the controller elements (CE) are responsible for the 3'-terminal mRNAs and the small 5'-terminal sgRNAs differ in the number of additional sgRNAs produced. A feature of both types of CE is production of 5'- and 3'-terminal positive-stranded sgRNAs, but the 3' CEs additionally produce a negative-stranded complement of the 3'-terminal mRNAs. Here, we found that the termination (for 5'-terminal sgRNAs) and initiation (for 3'-terminal sgRNAs) sites of the 5' vs. the 3' CEs occur at opposite ends of the respective minimal active CEs. The initiation site for the 3' CE of the major coat protein gene, and probably those of the p20 and p23 genes, was outside (3' in terms of the genomic RNA) the minimal unit, whereas the termination sites were located within the minimal CE, 30-50 nts upstream of the initiation site (referring to the positive-strand sequence). In contrast, the initiation site for the 5' CE was in the 5' region of the minimal unit, with the termination sites 20-35 nts downstream (referring to the positive-strand sequence). Furthermore, the CEs differ in initiation nucleotide and response to mutagenesis of that nucleotide. The 3' CE initiates sgRNA synthesis from a uridylate, whereas the 5' CE initiates from a cytidylate. We previously found that the 3' CEs were unusually tolerant to mutagenesis of the initiation sites, with initiation proceeding from alternative sites. Mutagenesis of the initiation site of the 5' CE prevented synthesis of either the 5'- or 3'-terminal sgRNAs. Thus, the cis-acting elements at opposite ends of the genome are remarkably different, perhaps having arisen from different origins and or with different functions in the life cycle of this virus.     

In vitro translation of cucumoviral satellites. I. Purification and nucleotide sequence of cucumber mosaic virus-associated RNA 5 from cucumber mosaic virus strain S

The satellite cucumber mosaic virus (CMV)-associated RNA 5 (CARNA 5) of CMV strain S (CMV-S) which previously had been assigned the capability both to direct the synthesis of two small proteins in vitro (R. A. Owens and J. M. Kaper, 1977, Virology, 80, 196-203) and to induce the tomato necrosis disease in the presence of its helper virus (J. M. Kaper and H. E. Waterworth, 1977, Science, 196, 429-431), has been reinvestigated. Polyacrylamide gel electrophoretic analyses under partially denaturing conditions of CARNA 5 preparations from CMV-S grown in tobacco reveal a mixture of three distinct RNA species which have been isolated and partially characterized. In order of decreasing electrophoretic mobility they have been designated RNA 5, (n)CARNA 5, and (S)CARNA 5, respectively. RNA 5 has been partially sequenced and shown to represent 3'-terminal fragments of the CMV genomic RNAs. (n)CARNA 5 is responsible for the tomato necrosis-inducing properties of the mixture and coelectrophoreses with tomato necrosis-inducing CARNA 5 from CMV strain D. (S)CARNA 5 does not cause tomato necrosis; its complete nucleotide sequence was determined and is compared here to the published sequences of the CARNA 5s of four other CMV strains. A companion paper (M. J. Avila-Rincon et al., 1986, Virology, 152, 455-458) provides unequivocal evidence that the in vitro translation of nonnecrotic (S)CARNA 5 produces two small polypeptides resembling those described earlier.     

Terminal structures of West Nile virus genomic RNA and their interactions with viral NS5 protein

Genome cyclization is essential for flavivirus replication. We used RNases to probe the structures formed by the 5′-terminal 190 nucleotides and the 3′-terminal 111 nucleotides of the West Nile virus (WNV) genomic RNA. When analyzed individually, the two RNAs adopt stem-loop structures as predicted by the thermodynamic-folding program. However, when mixed together, the two RNAs form a duplex that is mediated through base-pairings of two sets of RNA elements (5′CS/3′CSI and 5′UAR/3′UAR). Formation of the RNA duplex facilitates a conformational change that leaves the 3′-terminal nucleotides of the genome (position − 8 to − 16) to be single-stranded. Viral NS5 binds specifically to the 5′-terminal stem-loop (SL1) of the genomic RNA. The 5′SL1 RNA structure is essential for WNV replication. The study has provided further evidence to suggest that flavivirus genome cyclization and NS5/5′SL1 RNA interaction facilitate NS5 binding to the 3′ end of the genome for the initiation of viral minus-strand RNA synthesis.     

Sequence and secondary structure analysis of the 5'-terminal region of flavivirus genome RNA

The 5'-terminal noncoding region sequences were determined for the genome RNAs of seven strains of St. Louis encephalitis virus (SLEV) and one strain of West Nile virus (WNV) using a single synthetic cDNA primer complementary to the 5'-terminus of the coding region of a strain of WNV RNA. The 5'-terminal sequences obtained for the SLEV and WNV RNAs were compared with published sequences for yellow fever virus (YFV), Murray Valley encephalitis virus (MVEV), and dengue virus. While only short regions within the 5'-noncoding sequence were conserved among different flavivirus RNAs, significant homology was observed in this region among members of the same flavivirus subgroup and almost complete conservation was observed between different strains of the same virus. For example, seven strains of SLE, isolated from different geographic locations over a 17-year period and differing in their neurovirulence phenotype, contained only two to four nucleotide changes in the 5'-noncoding region. Interestingly, each of three low-virulence strains shared the same unique base substitution at position 16. Secondary structures predicted to be formed by the 5'-termini of each of the different flavivirus genome RNAs were of similar size and shape, in each case consisting of a stem with a small top loop and a larger side loop. The prediction of a common structure among a number of different flaviviruses, despite the lack of extensive sequence homology, suggests that this secondary structure is functionally important. An additional stem and loop structure is predicted to be formed in the region spanning the translation initiation codon. This structure showed significantly less conservation of size and shape than the 5'-terminal secondary structure.     

Cross protection between strains of cucumber mosaic virus: effect of host and type of inoculum on accumulation of virions and double-stranded RNA of the challenge strain

Two strains of cucumber mosaic virus (CMV) differed in three characteristics of value for cross-protection experiments. Their virions and also two of their double-stranded RNAs could be separated and distinguished by electrophoresis on polyacrylamide gels, and the symptoms of one strain were milder than the other in tobacco, tomato, and squash. The mild strain, CMV-S, protected plants of these three hosts from the effects of the second strain, CMV-P, and also prevented the accumulation of virions and ds RNAs of the challenge strain. Protection was detected in leaves inoculated with the challenge strain and also in later formed leaves. The only exception to this result was the accumulation of ds RNAs and to a lesser extent virions of the challenge strain when infectious viral RNA was used as the challenge inoculum instead of virus particles. This breakdown of cross protection occurred only in those leaves inoculated with the challenge strain RNA. No accumulation of challenge ds RNAs or virions occurred in later formed leaves. Tomato and tobacco plants infected with CMV-S were not protected from infection by tobacco mosaic virus.     

Nucleotide sequence and translation of satellite tobacco mosaic virus RNA

Satellite tobacco mosaic virus (STMV) is a plant virus with a 17-nm icosahedral particle encapsidating a 0.3 X 10(6) Mr ssRNA genome that depends on tobamoviruses for its replication. The complete nucleotide sequence of STMV RNA deduced in the experiments described here was 1059 nucleotides in length. The efficiency of labeling viral RNA with [gamma-32P]ATP using T4 polynucleotide kinase was not affected by treatment with tobacco acid pyrophosphatase and/or bacterial alkaline phosphatase, indicating that the majority of the 5' termini of encapsidated STMV RNAs were not phosphorylated. The 240 3'-terminal nucleotides of STMV RNA and either tobacco mosaic virus (TMV) U1 RNA or TMV U2/U5 RNA had greater than 65% overall sequence similarity, with two nearly identical regions of 40 and 50 bases, respectively. There were no other regions of sequence relatedness to TMV RNA. The 19 5'-terminal nucleotides of STMV RNA had greater than 65% sequence similarity with the 16 5'-terminal nucleotides of brome mosaic virus (RNA 3 and 50% sequence similarity with the 12 5'-terminal nucleotides of the Q strain of cucumber mosaic virus RNA 3. The first open reading frame (ORF) beginning at base 53 encoded a 6800 Mr protein that corresponded in size to a major in vitro translation product directed by STMV RNA. A second ORF, beginning at nucleotide 163, had the capacity to code for a protein that corresponded in size (17,500 Mr) to the other major in vitro translation product. The first 12 codons of this ORF corresponded to the sequence of the N-terminal amino acids of the capsid protein. Western-blot analysis of the in vitro translation products revealed that the 17,500 Mr protein had the same electrophoretic mobility as the authentic capsid protein; it was also antigenically related to the capsid protein, but the 6800 Mr protein was not. Time course analysis of in vitro translation demonstrated that the 6800 Mr protein was synthesized at the same time as the capsid protein and did not arise by the proteolytic cleavage of a larger precursor polypeptide. These results suggest that the genome of STMV functioned as a polycistronic messenger RNA. It has not been determined if the 6800 Mr protein is synthesized in vivo. STMV RNA had untranslated regions of 52 and 418 nucleotides at its 5' and 3' termini, respectively.(ABSTRACT TRUNCATED AT 400 WORDS)     

Intermediates in the disassembly of tobacco mosaic virus at alkaline pH Infectivity, self-assembly, and translational activities

Disassembly of tobacco mosaic virus at alkaline pH produces a series of discrete, kinetically metastable rods of intermediate length, all containing the original 3'-terminus, together with a "stable fraction" of apparently undegraded virus. The physical properties of these various rods and their protected RNAs were already known and some biological properties are now described. The stable fraction of rods was found to have an infectivity less than 5% of that of the control virus, irrespective of whether or not it had been treated with micrococcal nuclease after exposure to the alkaline conditions. This residual infectivity was probably due to contamination with some remaining undegraded virus. In the wheat germ cell-free protein synthesis system, the RNA from the stable fraction that had not been treated with micrococcal nuclease, and which still contained the 5'-terminal "cap" structure m7G(5')ppp(5')G, was translated indistinguishably from control viral RNA. The RNAs from the nuclease-treated stable fraction and from the other partly stripped particles were translated only with greatly reduced efficiency, but the fragment containing the 2,000 nucleotides from the 3' end of the original RNA yielded disproportionate amounts of a 30,000-dalton polypeptide. Reassembly with the viral coat protein occurred readily only with those fragments of the RNA longer than about 1,400 nucleotides from the 3'-terminus, consistent with earlier evidence that the nucleation site for virus assembly is about 1,000 nucleotides from the 3' end. Paradoxically, however, the RNA from the nuclease-treated stable fraction, which lacks only a few nucleotides from the 5' end of the native viral RNA, would not reassemble. The reason for this is unclear, but change in the secondary structure of the RNA in the assembly initiation region is a possible explanation.     

Full-length tobacco mosaic virus RNAs and defective RNAs have different 3' replication signals

The viral replicase complex of positive-stranded RNA viruses interacts with cis-acting elements that are usually located at the termini of the viral RNAs. On comparison of the replication requirement of a tobacco mosaic virus (TMV)-based defective RNA (dRNA) and its helper virus, we found different requirements for replication of TMV RNAs in cis and in trans. The level of replication of full-length TMV RNA decreased substantially in the absence of pseudoknot (pk) 1 and/or 2, whereas identical deletions in dRNAs did not affect their replication. However, pk3 was required for replication of both full-length TMV RNAs and dRNAs. The requirements for homologous sequences were greater for dRNA replication than for replication of full-length TMV RNAs. Defective RNAs with heterologous 3' nontranslated regions (NTRs) failed to be replicated or replicated minimally, whereas replication of similarly mutated full-length RNAs was much less affected. Increasing amounts of contiguous heterologous sequences in the dRNAs compensated for the impaired interactions between the replicase and 3' NTR. The precision requirement appeared to involve the terminal 28 nucleotides, specifically the pseudoknot in the aminoacyl acceptor arm of the tRNA like structure, which was important in replication of both dRNAs and full-length TMV RNAs.     

Enhancer-like properties of an RNA element that modulates Tombusvirus RNA accumulation

Prototypical defective interfering (DI) RNAs of the plus-strand RNA virus tomato bushy stunt virus contain four noncontiguous segments (regions I-IV) derived from the viral genome. Region I corresponds to 5'-noncoding sequence, regions II and III are derived from internal positions, and region IV represents a 3'-terminal segment. We analyzed the internally located region III in a prototypical DI RNA to understand better its role in DI RNA accumulation. Our results indicate that (1) region III is not essential for DI RNA accumulation, but molecules that lack it accumulate at significantly reduced levels ( approximately 10-fold lower), (2) region III is able to function at different positions and in opposite orientations, (3) a single copy of region III is favored over multiple copies, (4) the stimulatory effect observed on DI RNA accumulation is not due to region III-mediated RNA stabilization, (5) DI RNAs lacking region III permit the efficient accumulation of head-to-tail dimers and are less effective at suppressing helper RNA accumulation, and (6) negative-strand accumulation is also significantly depressed for DI RNAs lacking region III. Collectively, these results support a role for region III as an enhancer-like element that facilitates DI RNA replication. A scanning-type mutagenesis strategy was used to define portions of region III important for its stimulatory effect on DI RNA accumulation. Interestingly, the results revealed several differences in the requirements for activity when region III was in the forward versus the reverse orientation. In the context of the viral genome, region III was found to be essential for biological activity. This latter finding defines a critical role for this element in the reproductive cycle of the virus.     

The 3'-nucleotides of flavivirus genomic RNA form a conserved secondary structure

The terminal noncoding regions of viral RNA genomes are presumed to contain signal sequences and sometimes also secondary structures involved in regulating viral RNA synthesis. Such signals would be expected to be highly conserved among related viruses. In order to identify replication signal features for flaviviruses we have compared the 3'-terminal nucleotide sequences of West Nile virus (WNV), Saint Louis encephalitis (SLE) virus, and yellow fever virus (YFV) genome RNAs. The existence of a stable 3'-terminal secondary structure was previously predicted by a cDNA sequence obtained from YFV genome RNA. We have confirmed the existence of this structure by direct RNA sequencing methods. Even though the size and shape of the 3'-terminal secondary structure is highly conserved, sequence conservation is restricted to the loop regions of the secondary structure and to 27 nucleotides immediately adjacent to the 5' side of the structure. The regions of conserved sequence represent likely signals for viral polymerase recognition and binding. However, the preservation of the configuration of the secondary structure by a means other than sequence conservation indicate that this structure is important for the survival of the virus. A WNV mutant, which replicates progeny genome RNA more efficiently than parental WNV, was found to have a 3'-genomic sequence identical to that of its parent virus. The sequence change conferring the phenotype of this mutant is therefore located in another region of the genome.     

cis-acting sequences required for in vivo amplification of genomic RNA3 are organized differently in related bromoviruses

Cowpea chlorotic mottle virus (CCMV) is a positive-strand RNA virus that infects dicotyledonous plants. The genome comprises three capped RNAs: RNA1 (3.2 kb), RNA2 (2.9 kb), and RNA3 (2.1 kb). cis-Acting sequences required for amplification in vivo were explored for RNA3, which does not contribute trans-acting factors to viral RNA replication. Using a CCMV cDNA expression system, deletions throughout RNA3 were constructed and tested for successful replication in barley protoplasts coinoculated with RNAs 1 and 2. As previously found for RNA3 of the related brome mosaic virus (BMV) (R. French and P. Ahlquist, 1987, J. Virol. 61, 1457-1465), either of the two coding regions can be individually deleted without blocking RNA3 amplification. However, in striking contrast to BMV, the entire intercistronic noncoding region separating these genes is also dispensable for CCMV RNA3 amplification. Moreover, although simultaneous deletions of the 3a and coat protein genes were deleterious for BMV RNA3 accumulation, CCMV RNA3 derivatives bearing larger deletions encompassing the 3a gene, intercistronic region, and coat protein gene amplify to high levels. Thus, unlike BMV RNA3, cis-acting sequences required for CCMV RNA3 amplification map solely in the 5' and 3' noncoding regions. Normal levels of CCMV RNA3 accumulation require over 125 but no more than 220 bases from the 3' noncoding region, and no more than the first 89 bases of the 238-base-long 5' noncoding region.