| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 278, Issue 35, 32673-32682, August 29, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
¶
From the
Laboratory of Infectious Diseases, NIAID,
National Institutes of Health, Bethesda, Maryland 20892 and the
Department of Microbiology and Molecular
Genetics, Harvard Medical School, Boston, Massachusetts 02115
Received for publication, May 22, 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
The group A rotaviruses are the primary cause of acute dehydrating gastroenteritis in infants and young children throughout the world (8). Their genome consists of 11 segments of dsRNA, each positioned at one of the vertexes of the T = 1 icosahedral core contained within the triple-layered rotavirus virion (9). In addition to a segment of dsRNA, each vertex contains a copy of the RdRP, VP1 (10), and the mRNA-capping enzyme, VP3 (11). The core lattice protein, VP2, forms the shell of the core and serves to anchor the core components VP1, VP3, and the genome (12). Rotavirus entry is accompanied by the loss of the outer protein layer of the virion, yielding doublelayered particles, which catalyze the synthesis of the eleven viral mRNAs (13). The mRNAs direct protein synthesis and serve as templates for synthesis of (–) strand RNAs to form the dsRNA genome segments. The synthesis of the (–) strand RNAs occurs concurrently with the packaging of template mRNAs into nascent pre-virion cores and is coordinated such that packaging and replication lead to the formation of cores that contain one copy each of the eleven genome segments of dsRNA (4). Cores serve as intermediates in the morphogenesis of double- and triple-layered virus particles.
The eleven rotavirus mRNAs possess 5'-methylated caps but lack 3'-poly(A) tails and share no sequence identity except at their termini (14). For the group A rotaviruses, the mRNAs begin with the sequence 5'-GGC(U/A)7-3' and, in most cases, end with the 3'-consensus sequence (3'CS), 5'-UGUGACC-3' (14, 15). Cell-free replication assays performed with disrupted virion-derived cores (open cores) have shown that viral mRNAs must contain the 3'CS or related sequence to serve as templates for (–) strand synthesis (16–17). Additional cis-acting signals are contained with viral mRNAs that enhance but are not essential for (–) strand synthesis (16). These enhancing signals are located at the 5'-end and upstream of the 3'CS at the 3'-end of the viral mRNAs. The 5' and 3' enhancing signals are believed to base pair in cis to form panhandle structures from which the 3'CS extends as a single-stranded tail (18). The importance of the panhandle structure may be to ensure that the 3'CS is sterically accessible to the RdRP. Indeed, modifications made to the mRNA that alter the panhandle structure or that convert the 3'CS from a single-stranded to double-stranded form inhibit dsRNA synthesis (19–20). Thus, both sequence and structure are critical factors influencing the ability of rotavirus mRNAs to function as templates in replication. Although electrophoretic mobility shift assays (EMSA) have demonstrated that the RdRP specifically recognizes signals located at the 3'-end of rotavirus mRNAs, such studies have not shown whether the signals are located within the 3'CS (6, 21). Besides playing a critical role in RNA replication, the 3'CS also contains a recognition signal for NSP3, a viral protein that increases the translation efficiency of the mRNAs (22–24).
Studies with open cores have shown that the formation of the initiation complex for (–) strand synthesis is a salt-sensitive process that requires one or more core proteins, template mRNA, and GTP (25). Once formed, the initiation complex is resistant to salt-mediated disruption. Similarly, elongation of the (–) strand is a salt-resistant event. Replacement of open cores with purified recombinant (r) proteins in replication assays has shown that VP1 alone lacks polymerase activity, but when combined with the core lattice protein, the polymerase gains activity that directs dsRNA synthesis (7). Thus, the lattice protein provides an undefined, but essential, function in the synthesis of (–) strand RNA by the viral RdRP.
In this study, we have examined the requirements for template recognition by the RdRP and compared those for formation of the (–) strand initiation complex. Our analysis reveals that multiple functionally independent recognition signals are present at the 3'-end of viral mRNAs, some positioned in the non-conserved sequences upstream of the 3'CS. RdRP recognition signals are distinct from cis-acting signals involved in the formation of (–) strand initiation complexes, because deletion of portions of the 3'CS can cause the viral mRNAs to be poorly replicated while still being efficiently recognized by the RdRP. Although the RdRP can specifically bind to recognition signals in the absence of other proteins, the polymerase requires the core lattice protein to form the salt-resistant (–) strand initiation complex. Hence, template recognition and the formation of (–) strand initiation complexes by the RdRP differ in sequence and protein requirements.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
To prepare rVP1, rBVg1-infected cells were pelleted at 2 days post
infection, washed with phosphate-buffered saline, resuspended in 25
mM Tris-HCl, pH 7.8, 150 mM NaCl, 1 mM EDTA,
and 10 mM 
-mercaptoethanol, and homogenized. The membrane
fraction was removed by low speed centrifugation, and rVP1 was precipitated
from the supernatant between 20 and 25% (w/v) of ammonium sulfate. rVP1 was
dissolved in elution buffer (50 mM Hepes-NaOH, pH 7.8, 50
mM NaCl, 1 mM EDTA, 10% glycerol, and 2 mM
-mercaptoethanol), applied to a heparin-Sepharose column, and eluted
with a gradient of NaCl. The peak fraction containing rVP1 was concentrated
with a Centricon filtration unit and passed through a cationic exchange column
(Amersham Biosciences). Fractions containing rVP1 were chromatographed on a
Superdex-200 size exclusion column (Amersham Biosciences) using the elution
buffer. The eluted rVP1 was stored at 4 °C.
rVP2 was purified by CsCl centrifugation as described earlier (7). The concentration of purified proteins was determined by comparison with known amounts of bovine serum albumin electrophoresed on SDS-polyacrylamide gels and stained with Coomassie Brilliant Blue.
Preparation of Templates for Transcription—The T7 transcription vector SP65g8R (16), which contains a full-length cDNA of rotavirus gene 8 RNA was linearized with SacII and blunt-ended by treatment with T4 DNA polymerase prior to transcription with T7 RNA polymerase.
The vector SP72/T3–0/EcoRV (21) was linearized with EcoRV and used in T3 transcription reactions to produce a nonviral 60-nt RNA (NonV), which initiates with sequence gggatttgttgca, followed by a sequence corresponding to residues 29–78 of SP72. The vector T7B/T3–60(+) (21) was linearized with SacII, treated with T4 DNA polymerase, and used in T3 transcription reactions to synthesize the RNA probe g8-3'60. The sequence of the g8-3'60 begins with three G residues followed by the 3'-terminal 60 residues of the SA11 gene 8 RNA.
The vectors SP65g8-5'-3' (21) and SP65g8-gfp (26) were linearized with SacII and SpeI, respectively, treated with T4 DNA polymerase, and used in T7 transcription reactions to synthesize the RNA probes g8-5'-3' and g8-5'1–88. The g8-5'-3' probe links through the nonviral sequence UUAUU the first 88 nt and the last 50 nt of the SA11 gene 8 RNA, whereas g8-5'1–88 represents only the first 88 nt of the gene 8 RNA.
The vector SP72/T3–60minus was generated by annealing these synthetic oligonucleotides: 5'-agcttcgccggcgAATTAACCCTCACTAAAgggaggtcacataagc-3', 5'-gctttctattcttgctaagccatcatcatcctcaaattgatagcgaatcccgggg-3', 5'-aattccccgggattcgctatcaatttgaggatgatgatgatggcttagcaagaatagaaagcgcttatgtgacctccc-3', and 5'-TTTAGTGAGGGTTAATTcgccggcga-3' (sequences representing the T3 promoter are in uppercase). The annealed product was ligated into the plasmid SP72 previously linearized by digestion with HindIII and EcoRI. This vector was used to produce an RNA representing the inverse complement of the last 60 nucleotides of gene 8 mRNA.
Alternatively, PCR was used to prepare cDNAs linked to T7 or T3 promoters. Primers used in these reactions are listed in Table I. The PCR products were treated with T4 DNA polymerase and gel-purified with a Qiaex II Kit (Qiagen) prior to transcription. Nucleotide sequences were confirmed using an Applied Biosystems automated sequencer.
|
In Vitro Synthesis of RNAs—Full-length g8, g8
CC and
g83'CS, g8+AA, g8+CC, and g8–5'1–88
plus-sense RNAs were produced with the Ambion T7 MEGAscript system
(21). The RNA products of
reaction mixtures were purified by phenol-chloroform extraction and ethanol
precipitation and passed through Sephadex G-25 columns (Roche Applied Science)
to remove unincorporated NTPs. The quality of the mRNAs was assessed by
electrophoresis on 1.25% agarose gels containing formaldehyde (Reliant gel
system, Cambrex). RNA concentrations were calculated from optical densities at
260 nm.
RNA probes used in electrophoretic mobility shift assays (EMSAs) were made
with the Ambion T3 or T7 MEGAscript systems. Transcription was performed
according to the manufacturer's protocols except that the concentration of
cold UTP was reduced by one-fourth and 200 µCi of
[
-32P]UTP (800 Ci/mmol) was included per 80 µl of
reaction mixture. Cold probes were trace-labeled by including
[-33P]UTP in reaction mixtures. The probes were purified by
electrophoresis on 8% polyacrylamide gels containing 7 M urea
(21).
To generate g8-3'60 dsRNA, equal amounts of g8-3'60 (+) and (–) sense RNAs were heat-denatured and annealed by cooling in 50 mM Tris-HCl, pH 8.0, 5 mM MgCl2, and 55 mM KCl. The annealed product was resolved by electrophoresis on a 3% agarose gel and recovered with a Qiaex II Kit (Qiagen).
Electrophoretic Mobility Gel Shift Assay—Procedures for characterization of rVP1-RNA interactions by EMSA were similar to those described earlier (21). Typically, 32P-labeled probes (10 pmol) were incubated with rVP1 (2 pmol) in a final volume of 10 µl for 30 min at room temperature in the presence or absence of competitor RNA. The reaction mixtures were analyzed by electrophoresis for 3–4 h at 175 V on nondenaturing 8% polyacrylamide gels containing 50 mM Tris-HCl and 50 mM glycine, pH 9.1. Protein·probe complexes were detected by autoradiography, and the intensities of radiolabeled bands were quantified with an Amersham Biosciences PhosphorImager.
Replicase Assay—Based on results optimizing the replicase
activity of rVP1 in the presence of rVP2, reaction mixtures (20 µl)
contained 50 mM Tris-HCl, pH 7.1, 1.5% polyethylene glycol, 2
mM dithiothreitol, 1.5 unit of RNasin, 20 mM magnesium
acetate, 2 mM manganese acetate, 1.25 mM each of ATP,
CTP, and UTP, 5 mM GTP, 10 µCi of [-32P]UTP
(800 Ci/mmol), 5.6 pmol of each template mRNA, 2 pmol of rVP1, and 40 pmol of
rVP2. Reaction mixtures were incubated at 37 °C for 4 h. The
32P-labeled dsRNA products were analyzed by SDS-PAGE and
autoradiography and quantified with a PhosphorImager.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
= 1.30 g/cm3
(7)
(Fig. 1).
|
rVP1 Specifically Binds to the 3'-End of Gene 8 mRNA— Virion-derived VP1 has RNA-binding activity that specifically recognizes the 3'-end of rotavirus mRNAs (6). To test whether rVP1 had similar activity, the purified protein was incubated with 32P-labeled g8-3'60, an RNA probe with the same 3'-terminal 60-nt sequence as the viral gene 8 mRNA. Analysis of the reaction mixture by EMSA showed that the protein and probe interacted, forming rVP1·g8-3'60 complexes (Fig. 2A, lane 1). To address the specificity of the RNA-binding activity of rVP1, the protein was incubated with 32P-labeled g8-3'60 in the presence of increasing amounts of the unlabeled competitor RNAs g8-3'60, NonV, or NonV/A. The nonviral RNAs, NonV and NonV/A, are identical except the 3'-end of the latter contains a homopolymer A tail of 20-nt. The competition assays showed that the presence of NonV or NonV/A RNA had a minimal effect on the formation of rVP1·g8-3'60 complexes even when these competitors were in 20-fold excess over the probe (Fig. 2, A and B). In contrast to the nonviral RNAs, the viral-specific competitor RNA g8-3'60 successfully competed with the formation of rVP1·probe complexes to an extent approximately proportional to the ratio of probe to competitor in the reaction mixtures. These results show that rVP1 has specific affinity for the 3'-terminal 60-nt sequence of the gene 8 mRNA.
|
To determine whether rVP1 possessed a single or multiple RNA-binding sites, we took advantage of the fact that the migration of rVP1·probe complexes on nondenaturing gels is affected by the size of the probe associated with the protein. For example, rVP1·probe complexes containing the 63-nt g8-3'60 RNA migrate more rapidly than complexes containing the 143-nt g8-5'-3' RNA (Fig. 3, lanes 2 and 6). Electrophoretic analysis of reaction mixtures containing rVP1 and combinations of these two probes showed that the only complexes that were formed migrated like those that were formed in reaction mixtures containing rVP1 and either one of the probes (lanes 3–5 versus lanes 2 and 6). The absence of complexes in reaction mixtures containing both probes that migrated at a position intermediate of complexes formed with either one of the probes indicates that rVP1 was not able to simultaneously bind the g8-3'60 and g8-5'-3' probes. These data suggest that the rotavirus RdRP has only a single RNA-binding site.
|
Analysis of reaction mixtures containing 2 pmol of rVP1 and 10 pmol of
32P-labeled g8-3'60 by EMSA
(Fig. 2, lane 1)
showed that 
18% (n = 6, S.D. = 4.9), or 1.8 pmol, of the probe
interacted with the protein to form complexes. Given the assumption that each
molecule of RdRP binds one molecule of the probe, this suggests that almost
all rVP1 in the assay was functional with respect to RNA-binding activity.
Replicase Activity Is Dependent on rVP1 and rVP2—Replicase assays were used to test whether purified rVP1 was enzymatically active. The reaction mixtures included full-length gene 8 mRNA as template for (–) strand synthesis and [32P]UTP to label newly made dsRNAs. Analysis of the radiolabeled products showed that rVP1 alone lacked replicase activity (Fig. 4, lane 1). Similarly, rVP2 alone did not direct dsRNA synthesis (lane 2). In contrast, reaction mixtures containing rVP1 and any amount of rVP2 possessed replicase activity that catalyzed the synthesis of gene 8 dsRNA (lanes 3–7). These results are consistent with an earlier report, which indicated that the rotavirus RNA polymerase requires the core lattice protein for activity (7). The level of replicase activity was affected by the molar ratio of rVP1:rVP2 in reaction mixtures. When the ratio of rVP1:rVP2 was similar to the ratio of these proteins present at the vertexes of the core (1 molecule of VP1:5 dimers of VP2), the level of activity was maximal.
|
Taken together, the results presented in Figs. 2 and 4 indicate that, although the rotavirus RNA polymerase can interact specifically with the 3'-end of viral mRNA in the absence of any other protein, the polymerase requires VP2 for catalytic activity. Hence, the protein requirements for template recognition and (–) strand synthesis are distinct.
Factors Affecting Replicase Activity—Optimization of the replicase activity associated with rVP1 and rVP2 showed that maximal dsRNA synthesis occurred at 20 mM Mg2+ (Fig. 5A). The need for Mg2+ was specific, because the level of dsRNA synthesis was much lower when Mn2+ was used at any concentration in place of Mg2+ (Fig. 5B). However, the synthesis of dsRNA increased an additional 3- to 4-fold when low concentrations (1–4 mM) of Mn2+ were included in replicase assays that contained the optimal concentration (20 mM) of Mg2+ (Fig. 5B). Thus, the two divalent cations have distinct roles in dsRNA synthesis, and both are necessary for the rotavirus RNA polymerase to achieve maximal catalytic activity.
|
Time-course experiments showed that dsRNA synthesis continued for at least
24 h in replicase assays containing rVP1 and rVP2
(Fig. 6A). During the
first 8 h of incubation, synthesis occurred in a near linear manner. To
determine the elongation rate for (–) strand synthesis, a reaction
mixture containing rVP1, rVP2, gene 8 mRNA, and GTP was incubated
under conditions required for the formation of initiation complexes
(25). Afterward, the
elongation of (–) strand RNA was allowed to commence by the addition of
[32P]UTP and cold NTPs to reaction mixtures. Aliquots taken from
the mixture during the incubation period were analyzed for the synthesis of
full-length gene 8 dsRNA by SDS-PAGE
(Fig. 6B). The 1059-nt
gene 8 dsRNA product was initially detected in reaction mixtures
after 4 min of incubation indicating a maximum elongation rate of 270
nucleotides per min. This value is indistinguishable from the elongation rate
of open cores (data not shown). The optimal synthesis of dsRNA by rVP1 and
rVP2 occurred at 30 to 37 °C (data not shown), as was observed earlier
with open core replication assays.
|
Importance of 5'- and 3'-Sequences in
(–) Strand Synthesis— Previous experiments performed
with open cores have shown that the 3'CS has a critical role in
(–) strand synthesis
(16). In agreement with these
earlier results, we observed that replicase assays containing rVP1 and rVP2
did not produce dsRNA from gene 8 template mRNAs, which lacked the
3'CS (Fig. 7, A and
B). Replicase assays carried out with the recombinant
proteins also confirmed earlier results showing that the 3'-CC of viral
mRNA is important for (–) strand synthesis
(19), because deletion of
these residues reduced replication by two-thirds
(Fig. 7B). We found
that the addition of two extra nucleotides to the 3'-end (AA or CC) of
the gene 8 mRNA had little effect on its ability to serve as a
template for dsRNA synthesis (Fig.
7C). Notably, this is different from the results of a
previous study, which showed that, when such extended template RNAs were used
in open core replicase assays, a >50% reduction occurred in dsRNA synthesis
(19). The explanation for the
different findings is not known but might be correlated with VP3, a component
present only in the open core replicase assays. Consistent with results from
earlier studies (7,
16), we also observed that
deletion of the 5'-terminal 88 nucleotides from the gene 8 mRNA
led to a significant reduction (5-fold) in the synthesis of dsRNA by rVP1
and rVP2 (Fig.
7D).
|
Recognition Signals for the Viral RNA Polymerase—EMSA
performed with 32P-labeled g8-3'60 showed that the
3'-terminal 60-nt of the gene 8 mRNA contains cis-acting
signals, which allow it to be specifically recognized by the viral RNA
polymerase (Fig. 2). To
determine whether the 3'CS represented part of the recognition signal,
we analyzed the ability of the unlabeled competitor RNAs,
g8-3'60CS and NonV+CS, to interfere with the interaction between
rVP1 and 32P-labeled g8-3'60 by EMSA. The
g8-3'60CS RNA is identical to the g8-3'60 probe, except
that it lacks the 3'CS, and the NonV+CS is identical to the nonviral
RNA, NonV, except that the 3'CS has been added to its 3'-end. The
results showed that the g8-3'60CS and NonV+CS RNAs competed with
the 32P-labeled g8-3'60 probe for binding to rVP1 to an
extent intermediate of that of the unlabeled NonV and g8-3'60 RNAs
(Fig. 8, A and
B). These results indicate that g8-3'60CS,
despite the absence of the 3'CS, retained part of the recognition signal
for rVP1. Hence, signals specifically recognized by the viral RNA polymerase
exist within the nonconserved and highly variable region positioned upstream
of the 3'CS (residues –8 to –60).
|
In addition, EMSA showed that a competitor RNA, which lacked the
3'-terminal 30-nt of the gene 8 mRNA but contained residues
between –31 and –60 (g8–3'1–30), competed
to an intermediate extent with 32P-labeled g8-3'60 RNA in the
formation of rVP1·probe complexes
(Fig. 8C). Similarly,
EMSA showed that a competitor RNA, which contained only the viral specific
region from –8 to –29
(g8–3'CS/30–60), competed to an intermediate
extent with 32P-labeled g8-3'60 RNA for rVP1. These results
indicate that recognition signals for the RNA polymerase are present in
regions, 30–60 and 8–29 nucleotides upstream from the 3'-end
of the gene 8 mRNA. Because the NonV+3'CS RNA was able to
competitively interfere with the formation of the rVP1·probe complex,
albeit at an intermediate level, the 3'CS must also include a signal
recognized by the RNA polymerase (Fig.
8B). Collectively, these data suggest that the
3'-end of the gene 8 mRNA contains multiple functionally
independent polymerase recognition signals. Notably, the gene
8CS RNA was not able to serve as a template for dsRNA
synthesis (Fig. 7B),
despite containing the same sequence that allowed the g8-3'60 RNA to be
recognized by the viral RNA polymerase at an intermediate level
(Fig. 8B). Thus, there
is no direct connection between whether an mRNA can be recognized by the RNA
polymerase and whether the mRNA can function as a template for (–)
strand synthesis.
To address the importance of the 3'-terminal CC residues of rotavirus
mRNAs in polymerase recognition, we analyzed the ability of the unlabeled
g8-3'60CC and NonV+CC RNAs to compete with 32P-labeled
g8-3'60 in the formation of rVP1·probe complexes. The sequences
of the g8-3'60CC and g8-3'60 RNAs are identical, except the
former lacks the 3'-terminal CC. NonV+CC is like the NonV RNA but in
addition contains a 3'CC. Over the range of competitor RNAs used in the
EMSA, no significant difference was noted between the capacity of
g8-3'60CC and g8-3'60 to competitively interfere with the
formation of complexes between rVP1 and 32P-labeled g8-3'60
(Fig. 8D). This
indicates that the 3'-terminal conserved CC of rotavirus mRNAs does not
form part of the polymerase recognition signal contained in the
3'-terminal 60-nt of the gene 8 mRNA. In contrast, replicase
assays have shown that the 3'CC contributes significantly to the ability
of the gene 8 mRNA to function as a template for (–) strand
synthesis (Fig. 7B).
The inability of NonV+CC, even at high concentrations, to compete with
32P-labeled g8-3'60 for rVP1
(Fig. 8D) provides
additional evidence that the 3'CC is not involved in the recognition of
the template mRNA by the RdRP.
To examine the effect of extra nucleotides at the 3'-end on polymerase recognition, g8-3'60 RNAs with two added A (g8-3'60+AA) or C (g8-3'60+CC) residues were used as competitors in EMSA containing rVP1 and the 32P-labeled g8-3'60. The analysis showed that g8-3'60 RNAs with the extra 3'AA and 3'CC competitively interfered with the formation of the rVP1·probe complex in a manner similar to the g8-3'60 RNA (Fig. 8E). This finding suggests that to at least some extent the position of the polymerase recognition signals in the gene 8 mRNA relative to the 3'-end is not an important factor affecting polymerase binding.
The impact of strandedness of the g8-3'60 RNA on its recognition by the RNA polymerase was analyzed using the double-stranded form of g8-3'60 RNA as competitor in EMSA. As shown in Fig. 8F, the conversion of the g8-3'60 RNA from a single-stranded to double-stranded form resulted in a loss of the ability of the g8-3'60 RNA to be specifically recognized by the RNA polymerase. Thus, the function of the polymerase recognition signals is dependent on single-stranded regions or higher order structures formed by the 3'-terminal 60-nt of the gene 8 mRNA. This finding is consistent with the observation that annealing of complementary oligonucleotides to the 3'-end of the gene 8 mRNA prevents it from serving as a template for dsRNA synthesis (20).
Finally, we found that an RNA probe containing only sequences corresponding to the 5'-end of the gene 8 mRNA, i.e. g8-5'1–88, did not compete for binding of the 32P-labeled g8-3'60 probe to rVP1 (Fig. 8G). Hence, the 5'-end of the gene 8 mRNA apparently does not contain polymerase recognition signals. Remarkably, this contrasts with the results of replicase assays performed with rVP1 and rVP2 (Fig. 7D) or open cores (16), which showed that the 5'-end plays an important role in promoting synthesis of dsRNA from viral mRNAs.
Essential Components for the Formation of a Functional Initiation
Complex for (–) Strand RNA Synthesis—Replicase
assays performed with open cores have shown that monovalent salts are an
effective inhibitor of dsRNA synthesis
(25). Previous analysis of
this phenomenon has shown that it is the process of (–) strand
initiation that is salt sensitive and not elongation, which can proceed even
in concentrations of salt exceeding 1 M. To assemble an initiation
complex that remains functional when subsequently exposed to salt requires one
or more core proteins, GTP, and template mRNA
(25). In agreement with these
results, we found that the synthesis of gene 8 dsRNA by rVP1 and rVP2
was inhibited by the presence of NaCl, with an approximate 10-fold reduction
in RNA replication occurring at 100 mM
(Fig. 9). To assess the
possibility that salt inhibited dsRNA synthesis by reducing the ability of the
RNA polymerase to interact with the template mRNA, rVP1 and
32P-labeled g8-3'60 probe were co-incubated in the presence
of 0–250 mM NaCl. rVP1·probe complexes in the reaction
mixtures were analyzed by EMSA, and their levels compared with the levels of
dsRNA synthesis that occurred in replicase assays containing NaCl
(Fig. 9). The results showed
that a correlation did not exist between the extent of dsRNA synthesis and the
levels of rVP1·probe complexes formed. For example, at 100
mM NaCl, where dsRNA synthesis was reduced by 10-fold, only a
20% reduction was observed in the amount of probe bound to the polymerase. And
at 250 mM NaCl, where no dsRNA occurred, the extent or interaction
between the polymerase and probe was reduced by only 40%. These data provide
evidence that binding of the RNA polymerase to the 3'-end of the gene
8 mRNA does not in itself satisfy the requirements that must be met
for forming the (–) strand initiation complex.
|
To clarify the protein requirement for forming the initiation complex for
(–) strand synthesis, rVP1 and/or rVP2 were incubated with gene
8 mRNA, three NTPs (ATP, CTP, and GTP), and divalent cations
(Mg2+ and Mn2+) (initiation
phase). Afterward, NaCl was added the reaction mixtures to prevent further
de novo formation of initiation complexes. To promote elongation, the
reaction mixtures were then supplemented with UTP and any components not
present in the initiation phase (e.g. rVP1 or rVP2; ATP, CTP, and
GTP; and gene 8 mRNA). The analysis showed that dsRNA was synthesized
only in reaction mixtures in which both rVP1 and rVP2 were included during the
initiation phase (Fig.
10A, lane 2 versus lanes 4 and 5).
dsRNA synthesis also required the presence of ATP, CTP, and/or GTP (lane 1
versus 2), and viral mRNA (lane 2 versus 3) during the
initiation phase. Further assays showed that the need for NTPs during
initiation was satisfied by GTP alone
(Fig. 10B), a result
similar to that seen in the analysis of (–) strand synthesis by open
cores (25). The levels of
functional initiation complexes were 3-fold higher when both
Mg2+ and Mn2+ were present during
the initiation phase compared with the levels obtained when
Mg2+ was present alone, even when the latter assays were
supplemented with Mn2+ during elongation
(Fig. 10C, lanes
2 and 6). Taken together, the results indicate that, although
the viral RNA polymerase contains the intrinsic ability to recognize viral
mRNA, the assembly of the polymerase into a complex with catalytic activity
requires the core lattice protein, viral mRNA, GTP, and, for greatest
efficiency, both Mg2+ and
Mn2+.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
The 3'-terminal CC of the 3'CS (UGUGACC) plays a critical role in the assembly of stable initiation complexes for (–) strand synthesis (25). In contrast, EMSA indicates that the CC is not involved in recognition of the viral mRNA by the RdRP. Therefore, it is the remaining portion of the 3'CS, i.e. UGUGA, that seems likely to form a polymerase recognition signal. However, although the vast majority of rotavirus mRNAs end with the 3'CS and, thus, contain the UGUGA region, some do not. Rather, analysis of genomic dsRNAs of viable rotaviruses has revealed that viral mRNAs ending with the sequence 5'-URN0–5CC-3' (R = purine, N = any base) can function efficiently as templates for (–) strand synthesis.2 When this finding is considered in combination with our EMSA results, it can be predicted that the UG (or UR) portion of the 3'CS is a critical part of the polymerase recognition signal. Interestingly, the study of rotavirus mRNAs has shown that the distance between the 5'-UR and 3'-CC elements of the minimal 3' replication motif can range from 0 to 5 nucleotides.2 Thus, a fixed distance between the residues involved in RdRP recognition and in initiation complex formation is not required for productive (–) strand synthesis.
From our studies, we predict that, during genome replication, the RdRP
initially interacts with the upstream recognition signals of the viral mRNA to
form a stable RNA·protein complex. This interaction then facilitates
the recruitment of the 3'-CC to a site within the polymerase that
mediates initiation. Without the upstream recognition signals, the recruitment
of the CC to the initiation site is inefficient, because the addition of two C
residues to the 3'-end of nonviral RNAs does not provide sufficient
information for the RNAs to function as templates for dsRNA synthesis
(16). The stabilization of the
CC at the initiation site is likely dependent on one or more of the components
required for forming the (–) strand initiation complex, which includes
divalent cations, GTP, and VP2. Our studies indicate that
Mg2+ and Mn2+ have distinct roles
in promoting initiation and RNA synthesis. This suggests that the rotavirus
RdRP may have specific binding sites for each cation and thus have similarity
to the polymerase of the segmented dsRNA phage, 
6, which has been
previously shown to bind both ions specifically, even in the absence of NTPs
(27). Like the rotavirus RdRP,
the 6 polymerase also requires Mg2+ and
Mn2+ to achieve maximal levels of RNA synthesis in
vitro (27).
The need for VP2 in RNA replication has been established by studies showing that (i) dsRNA is not made in cells infected with a mutant rotavirus containing a temperature-sensitive lesion in the VP2 gene (28), (ii) replicase activity is associated only with those replication intermediates recovered from infected cells which contain VP2 (29), and (iii) VP2 must be present in replicase assays for VP1 to synthesize dsRNA (7). Here, we have shown that VP2 is required for formation of the (–) strand initiation complex. The question of whether the protein is required for elongation remains unresolved. Nevertheless, the requirement of the core lattice protein VP2 in initiation provides a potential mechanism by which replication of the rotavirus genome is made dependent upon the availability of sufficient levels of the core lattice protein to allow packaging of newly made dsRNAs.
Each vertex of the virion core is composed of a copy of VP1 and VP3 and 10 copies of VP2 and directs (+) strand synthesis from one of the genome segments. The fact that dsRNA synthesis in replicase assays was maximal at a ratio of VP1 and VP2 of 1:10 suggests that replicase activity is associated with structures much like those that form the vertexes of the core. Moreover, the 1:10 ratio implies that polymerase activation requires the assembly of VP2 into a decamer. Because VP2 lacks detectable helix unwinding activity (data not shown), it seems unlikely that the protein functions in RNA replication to remove or alter secondary structures in template mRNAs that could impede formation of initiation complexes or (–) strand elongation. Instead, based on physical studies showing that VP1 and VP2 interact within the core (30), it may be speculated that VP2 induces structural changes in the polymerase that relieve blocks to the formation of stable initiation complexes that can catalyze (–) strand synthesis. These structural changes may parallel those observed for other RNA polymerases, e.g. T7 RNA polymerase, which in some cases has been shown to undergo dramatic rearrangement as they transition from conformations that allow only abortive initiation to those that allow promoter escape (1, 2).
The 5'-end contributes importantly to the ability of rotavirus mRNAs to serve as templates for (–) strand synthesis in replicase assays containing open cores (16) or rVP1 and rVP2 (7). Our EMSA results suggest that the mechanism by which the 5'-end up-regulates dsRNA synthesis is not due to the presence of recognition signals for RdRP in this portion of template mRNA. Rather, our data are consistent with the earlier proposal that the 5'-end contributes to efficient replication by interacting with complementary regions at the 3'-end, to generate panhandle-like structures in which the RdRP recognition signals at the 3'-end are presented in forms that are accessible and comformationally stable (18).
An unexpected finding of our replicase assays with rVP1 and rVP2 was that the addition of two extra nucleotides (AA or CC) to the 3'-end did not effect the template activity of the gene 8 mRNA. This is in contrast to earlier studies with open cores, which showed that such additions reduced dsRNA synthesis by >50% (19). The increased specificity of the replicase activity of open cores for template mRNAs that terminate with authentic 3'-ends correlates with the presence of VP3, the multifunctional capping enzyme of the virus that has associated guanylyltransferase and methyltransferase activities. The protein also has sequence-independent affinity for RNA but exhibits preference for uncapped over capped RNA (21). A copy of VP3 resides at each of the vertexes of the core, positioned toward the inner side where it contacts VP2 and possibly VP1 (30). In this configuration, the interaction of VP3 with the 5'-end of viral mRNAs containing 5'-3' panhandles may assist in the positioning of the 3'-end into the initiation pocket of the RdRP. Addition of extra nucleotides to the 3'-end of the mRNA would obviously affect the juxtaposition of the ends with respect to each other in the panhandle structure. Such a change may decrease the capacity of VP3 to appropriately assist in placement of the 3'-end into the initiation pocket, causing a reduction in the efficiency of dsRNA synthesis by open cores.
|
FOOTNOTES |
|---|
¶ To whom correspondence should be addressed: Laboratory of Infectious Diseases, NIAID, National Institutes of Health, 50 South Dr., MSC 8026, Rm. 6314, Bethesda, MD 20892. Tel.: 301-594-1615; Fax: 301-496-8312; E-mail: jpatton{at}niaid.nih.gov.
1 The abbreviations used are: RdRP, RNA-dependent RNA polymerase; dsRNA,
double strand RNA; CS, consensus sequence; EMSA, electrophoretic mobility
shift assay; nt, nucleotide(s); NonV, nonviral RNA.
2 K. Kearney, D. Chen, Y. Hoshino, Z. Taraporewala, P. Vende, and J. Patton,
unpublished data.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  P. Roy Bluetongue virus: dissection of the polymerase complex J. Gen. Virol., August 1, 2008; 89(8): 1789 - 1804. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. McDonald and J. T. Patton Molecular Characterization of a Subgroup Specificity Associated with the Rotavirus Inner Capsid Protein VP2 J. Virol., March 15, 2008; 82(6): 2752 - 2764. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. TORTORICI, B. A. SHAPIRO, and J. T. PATTON A base-specific recognition signal in the 5' consensus sequence of rotavirus plus-strand RNAs promotes replication of the double-stranded RNA genome segments RNA, January 1, 2006; 12(1): 133 - 146. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. van den Born, D. A. Stein, P. L. Iversen, and E. J. Snijder Antiviral activity of morpholino oligomers designed to block various aspects of Equine arteritis virus amplification in cell culture J. Gen. Virol., November 1, 2005; 86(11): 3081 - 3090. [Abstract] [Full Text] [PDF]
|
|
