Previous Article | Next Article ![]()
Journal of Virology, June 2005, p. 7327-7337, Vol. 79, No. 12
0022-538X/05/$08.00+0 doi:10.1128/JVI.79.12.7327-7337.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.
D. Margaret Hunt,2 and
Sue A. Moyer1*
Department of Molecular Genetics and Microbiology, University of Florida College of Medicine, Gainesville, Florida 32610,1 Department of Microbiology and Immunology, University of South Carolina School of Medicine, Columbia, South Carolina 292082
Received 9 November 2004/ Accepted 8 February 2005
|
|
|---|
|
|
|---|
The VSV RdRp produces mRNA transcripts modified at their 5' end by capping and cap methylation (71). The mechanism of mRNA 5' capping in VSV and other NNS RNA viruses is unusual, where, in contrast to cellular cap structures, both the
and ß phosphates in the GpppA triphosphate bridge are derived from a GDP donor (2, 5, 23). The cytoplasmic localization of virus transcription and the unusual mechanism of capping suggest that the guanylyltransferase activity is virus encoded, although recent studies showed an association of cellular guanylyltransferase with the VSV RdRp complex, leading those authors to propose a catalytic role of this cellular enzyme in VSV mRNA capping (22, 47).
VSV mRNA 5'-cap structures are methylated at the guanine-N7 and 2'-O-adenosine positions (7mGpppAm) (40, 49, 50). Previously, in vitro results by Testa and Banerjee indicated the following order for the cap methylation in VSV: GpppA
GpppAm
7mGpppAm (69). While this may be a preferred order, an alternative order of the cap methylation (GpppA
7mGpppA
7mGpppAm) was shown to exist during VSV transcription in vivo (39) and in vitro (24). Unlike mRNA capping, which is tightly coupled with mRNA transcription in VSV, viral mRNA synthesis proceeds with a similar efficiency in the absence or presence of the methyl group donor, S-adenosyl-L-methionine (AdoMet) (1). However, mRNA cap methylation (particularly at the guanine-N7 position) is a prerequisite for the successful translation of viral mRNAs and is, therefore, essential for a productive VSV replication cycle (21, 29, 30). Simpson and coworkers isolated two host range (hr) mutants, hr1 and hr8, which were severely restricted for growth in many human cell lines, including HEp-2 cells, as well as in permissive cell lines, including BHK cells, at the nonpermissive temperature (44, 58, 59). Our laboratory further characterized these mutants and showed that the hr phenotype of both hr1 and hr8 mutants was dependent on a viral deficiency in mRNA guanine-N7 cap methylation, resulting in defective protein synthesis during infection of nonpermissive cells (29, 30). Later, we demonstrated that purified wild-type (wt) L protein can complement the cap methylation defect of these mutants during transcription in vitro (25), indicating that the VSV L protein possesses the viral methyltransferase (MTase) activities.
More than 15 years later, there has been no identification of residues in the L protein which catalyze mRNA cap MTase activities in VSV or any other NNS RNA virus. Previous alignments of L protein sequences from different NNS RNA viruses identified six conserved sequence regions, designated domains I to VI, which are believed to constitute various enzymatic activities of the L protein (45, 57, 65, 68). Domain II has a highly charged putative RNA-binding motif, and domain III contains a potential GDNQ phosphodiester bond-forming motif (33, 43, 45). Consistent with the predicted roles of domains II and III in viral RNA synthesis, most (domain II) or all (domain III) mutations within these motifs abolished RNA synthesis in several viruses (32, 53, 60, 62, 63). Although domains I, IV, and V are conserved among NNS viruses, no sequence elements were found which would indicate specific functions of these domains. A single mutation in domain IV specifically affected transcriptional termination of VSV L (9), and mutations in domains I and V have been shown to uncouple Sendai virus transcription and replication (12, 15). The N-terminal region of the L protein (overlapping with or including domain I) is also involved in the interaction of the L protein with the P protein and in the L-L oligomerization (10, 11, 27, 61, 64). Although the L protein domain VI was originally predicted as a general NTP-binding site (45) and thought to be involved in P protein phosphorylation occurring at least in some NNS RNA viruses (51), recent independently conducted computational analyses (8, 19) proposed it may, in fact, be associated with cap 2'-O MTase function in NNS RNA viruses.
In this study, we conducted sequencing and a functional analysis of the hr1 mutant of VSV and identified a single amino acid change, D1671V, in domain VI of the L protein, which specifically abolished viral mRNA cap methylation and was responsible for both the hr and temperature-sensitive (ts) phenotypes of mutant hr1.
|
|
|---|
Plasmids, mutagenesis, and recovery of recombinant VSV. The VSV minigenome plasmid pVSV-CAT2 has the chloramphenicol acetyltransferase (CAT) gene flanked by VSV trailer and leader regions under control of the T7 promoter. The CAT gene was amplified by PCR using primers with SphI and NcoI sites and p107MVCAT plasmid (56) as a template, with an internal NcoI site in CAT silently removed by overlap PCR mutagenesis. The PCR product was cut with SphI and NcoI/blunt and ligated into the SphI-SpeI and SpeI-BspHI/blunt-digested pBS-GMF (67). The resulting pVSV-CAT1 directs T7 transcripts comprised of the following regions: VSV trailer complement (1-70)/SphI/CAT open reading frame complement/NcoI-BspHI fusion/VSV leader complement (1-89)/ribozyme. Because this construct produced high backgrounds in the minigenome assay due to the presence of the T3 promoter in pBluescript, the minigenome expression cassette was subcloned into pGEM-3Zf by inserting the XmnI-SacI fragment from pVSV-CAT1 into XmnI-SacI-digested pGEM-3Zf, creating pVSV-CAT2 with no T3 promoter.
pBS-L, pBS-P, pBS-N, and pVSVFL(+), the plasmids for the expression of wt VSV (Indiana serotype) L, P, and N genes (66) and the full-length VSV antigenomic RNA (35), respectively, were kindly provided by John K. Rose. To construct pBS-L(HR1-1) with a single D1671V mutation, RNA was isolated from the hr1 mutant virus and used as the template for reverse transcription-PCR using VSV primers MH49 and MH59 (sequences available upon request). The PCR product was digested with FseI and SalI and cloned into pBS-L at those sites. pBS-L(HR1-0) with a single N505D mutation was constructed using the QuikChange XL site-directed mutagenesis kit (Stratagene). The PCR product with the mutation N505D and a BsrI silent restriction site was generated using primers SM580 and SM581 (sequences available upon request) and pBS-L as the template. Plasmid containing the L mutation was identified by the presence of the silent restriction site and digested with XbaI and BstBI, and the fragment was inserted into pBS-L at those sites. To construct the double mutant pBS-L(HR1-0,1), plasmids pBS-L(HR1-1) and pBS-L(HR1-0) were digested with XbaI and BstB1, and the fragment containing the "1" mutation (D1671V) from the pBS-L(HR1-1) was inserted into the digested pBS-L(HR1-0) containing the "0" mutation (N505D). All plasmids were sequenced to verify the correct mutations.
The HR1-0, HR1-1, and HR1-0,1 mutations were also introduced into the full-length genomic VSV plasmid pVSVFL(+) g.1 (35) for recovery of recombinant viruses. Plasmids pBS-L(HR1-0), pBS-L(HR1-1), and pBS-L(HR1-0,1) were cut with SalI and HpaI, and the fragment containing the L mutation was inserted into pVSVFL(+) digested at those sites. The recombinant viruses were rescued as described previously (35). Five 35-mm wells of BHK cells in a six-well plate were infected with VVT7 at an MOI of 5.0 PFU/cell for 45 min at 37°C. The cells were then transfected with the plasmids pBS-P (5 µg), pBS-N (3 µg), pBS-L (1 µg), and the full-length plasmid (5 µg) containing the wt or mutant L gene and incubated at 34°C. The transfection was conducted using Opti-MEM medium (GIBCO) and Lipofectamine (Invitrogen) according to the manufacturer's protocol. A sixth well of BHK cells was infected with VVT7 as above and received Lipofectamine but no plasmids. Four hours posttransfection (p.t.), the medium containing the transfection reagents was aspirated, 2 ml of BHK growth medium containing 5% fetal bovine serum (FBS) was added, and cells were incubated at 34°C. To remove VVT7 and amplify recombinant viruses, 3 days p.t. the medium was filtered through a Millex GV 0.22-µm syringe-driven filter unit (Millipore) directly onto a 35-mm well of confluent BHK cells in a six-well plate. An additional 1.5 ml of BHK growth medium containing 5% FBS was added, and the cells were incubated at 34°C. On average, two of five wells showed visible cytopathic effect. The medium from such wells was passed onto a 100-mm dish of confluent BHK cells as described above. When cells had visible cytopathic effect, the medium was harvested by pelleting cellular debris for 15 min at 1,500 rpm. The titer of each virus was determined on BHK cells. Individual plaques were picked and grown on BHK cells, and the mutations in the L gene of each virus were confirmed by sequencing or restriction analysis using appropriate silent restriction sites.
In vitro transcription.
For VSV in vitro transcription with expressed L and P proteins, 60-mm dishes of BHK or HEp-2 cells were infected with VVT7 at an MOI of 2.5 PFU/cell for 1 h at 37°C, washed with Opti-MEM (Gibco), transfected with 1.5 µg of pBS-P and 1.5 µg of pBS-L (wt L or one of the mutant L genes) plasmids with Lipofectamine (Invitrogen), and incubated at 34°C or 40°C in Opti-MEM. At 18 h p.t., cytoplasmic extracts were prepared exactly as described previously for Sendai virus (12) by lysolecithin permeabilization of the cells in the incomplete reaction mix, digesting cell extracts (100 µl) with micrococcal nuclease, and then supplementing extracts with magnesium acetate, RNasin (Promega), actinomycin D, creatine phosphate, and creatine phosphokinase. To assay for VSV mRNA synthesis, 1 µg of wt VSV RNA-N template and 20 µCi of [
-32P]CTP were added to each extract and reaction mixtures were incubated for 2 h at 30°C.
VSV in vitro transcription by detergent-activated purified viruses was conducted as described earlier (41). For CTP-labeled RNA, 15 µg of purified virus was used in a 100-µl reaction mixture containing 0.1 M NaCl, 5 mM MgCl, 0.05 M Tris-HCl (pH 8.0), 0.05% Triton N101, 1 mM each of ATP, GTP, and UTP, 100 µM CTP, 50 U of RNasin (Promega), and 20 µCi of [
-32P]CTP. For GTP-labeled RNA, a similar reaction mixture was used, except with 1 mM CTP, 5 µM GTP, and 50 µCi of [
-32P]GTP. All reaction mixtures were incubated at 30°C for 4 h and analyzed by 1.5% agarose-8 M urea gel electrophoresis. The transcription products were visualized by autoradiography and quantitated using a PhosphorImager (Molecular Dynamics).
To generate GTP-labeled uncapped mRNA, PstI-digested pBS-N was transcribed in vitro by T7 polymerase with [
-32P]GTP and in the absence of cap analogs.
VSV CAT minigenome replication. HEp-2 or BHK cells in 35-mm dishes were infected with MVA-T7 at an MOI of 2.5 PFU/cell and then transfected with 1 µg N, 0.3 µg L (wt or mutant), 0.5 µg P, and 0.5 µg VSV-CAT2 plasmids in Opti-MEM medium using Lipofectamine (Invitrogen). The cells were incubated for 48 h at 34°C or 40°C, and then the transfection medium was aspirated, fresh BHK growth medium containing 5% FBS was added, and cells were incubated again for 24 h at the appropriate temperature. The following day the medium was aspirated, and the cells were washed with cold phosphate-buffered saline, scraped into 200 µl of 0.25 M Tris-HCl (pH 7.8) and 0.5% Triton X-100, and centrifuged for 10 min at 13,000 rpm at 4°C. A CAT enzyme-linked immunosorbent assay (Roche) was used to test supernatant samples (20 µl) for CAT enzyme expression according to the manufacturer's protocol as a measure of viral transcription and replication by mutant L proteins compared to wt VSV L. All mutants were tested in triplicate in a minimum of two separate experiments.
RNA cap analysis. To test for mRNA cap methylation, in vitro transcription by detergent-activated purified viruses was conducted as described above, but RNA was synthesized in the presence of cold nucleoside triphosphates (NTPs; 1 mM each) and 2.5 µCi of [3H]AdoMet (81.5 Ci/mmol). Total RNA was purified using RNeasy columns (QIAGEN), diluted in 15 µl of H2O, and used for gel analysis (10 µl), measurement of [3H]Met incorporation (1 µl), and analysis of methylated cap structures by nuclease P1 digestion (3 µl). To visualize methylated products, the RNA products of the in vitro transcriptions were analyzed by 1.5% agarose-8 M urea gel electrophoresis, and the gels were fixed in 7% acetic acid, processed for fluorography (6), dried, and exposed to Kodak X-Omat film for 48 h at 80°C. RNA methylation was measured by the incorporation of 3H into mRNA as assayed by binding to DEAE-cellulose paper and scintillation counting (46). For the structure analysis of methylated caps, 3 µl of methylated RNA was digested with the nuclease P1 in a 6-µl reaction mixture containing 25 mM sodium acetate (pH 5.3), 2.5 mM MgCl2, and 3 µg of P1 nuclease (MP Biomedicals) for 1 h at 37°C. The reaction volume was then adjusted to 50 µl, the P1 was removed by phenol-chloroform treatment, and the products were dried, diluted in 10 µl of H2O, and spotted (1 µl) onto polyethyleneimine-cellulose (PEI) plates (EMD Chemicals), along with unlabeled nucleotides (Sigma) and cap standards GpppA, 7mGpppA, GpppAm, and 7mGpppAm (P-L Biochemicals). After the PEI plates were dry, they were developed in 1.2 M LiCl at room temperature (13). Markers were developed under UV light, and tritiated cap structures were visualized by treating the PEI plates with 5% 2,5-diphenyloxazole in acetone for fluorography and exposing them to Kodak X-Omat film for 72 h at 80°C.
To radioactively label VSV mRNA cap structure independently of cap methylation, in vitro transcription by detergent-activated purified viruses was conducted with [
-32P]GTP in the absence of AdoMet. Total RNA was purified using RNeasy columns (QIAGEN) and diluted in 20 µl of H2O. To release the 5'-terminal GMP from GpppA caps, 3 µl of [
-32P]GTP-labeled RNA was digested with 2.5 U of tobacco acid pyrophosphatase (TAP; Epicentre) in a 20-µl reaction mixture for 1 h at 37°C, according to the manufacturer's protocol. The reaction volume was then adjusted to 50 µl, the TAP was removed by phenol-chloroform treatment, and the products were dried and resuspended in 10 µl of H2O. Each sample (1 µl) was spotted onto PEI plates, along with unlabeled GMP standard (Sigma). After the PEI plates were dry, they were developed in the buffer (100 ml) containing 0.1 M sodium phosphate (pH 6.8; 100 ml), ammonium sulfate (30 g), and n-propanol (2 ml) (34) at room temperature, GMP marker was developed under UV light, and the released [
-32P]GMP was visualized by exposing the plates to Kodak X-Omat film at 80°C for 16 h.
Western blot analysis. To compare the amount of L protein, total protein samples from transfected cytoplasmic lysates (5 µl of a total 100 µl of lysate) or from purified viruses (200 ng) were separated by 7.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electroblotted onto a nitrocellulose membrane. The blots were incubated with polyclonal rabbit antibodies against a VSV L fusion protein (the N-terminal half of the L protein fused to TrpE) and developed with horseradish peroxidase-conjugated goat anti-rabbit secondary antibody using the Enhanced Chemiluminescence Plus protein detection system (Amersham).
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. L protein sequence differences of wt VSV, the original hr1 host range mutant, and recombinant VSV
|
Effect of hr1 L mutations on VSV mRNA synthesis in vitro.
Our previous studies showed that infection of nonpermissive HEp-2 cells in vivo by the hr1 mutant was characterized by severely reduced synthesis of viral mRNAs and suggested that this defect was a result of the demonstrated lack of viral mRNA cap methylation and the consequent defect in the translatability of primary VSV transcripts (29). However, we were unable to determine how hr1 mutations affected the ability of viral polymerase to synthesize mRNAs under nonpermissive conditions using assays available at the time. To address this question now, we employed an in vitro transcription system (12, 28), which is independent of the translatability of VSV mRNAs and is based on the exogenously provided polymerase-free genomic RNA-N template, and cell extracts containing T7 RNA polymerase-expressed VSV L and P proteins. The N505D and D1671V mutations (individually or together) (Table 1) were introduced into the wt L gene, and the mutant L proteins were expressed under conditions that are nonpermissive for hr1 and tested for their capacity to synthesize viral mRNA in vitro. VVT7-infected cells (BHK or HEp-2) were transfected with wt P and wt or mutant L plasmids and incubated overnight at 34°C or 40°C. Cytoplasmic cell extracts (prepared as described under Materials and Methods) were incubated with polymerase-free wt VSV RNA-N template in the presence of [
-32P]CTP at 30°C, and the total mRNA product was analyzed. The empty "mock" lanes (Fig. 1 and 2) corresponded to the VVT7-infected cells, but not transfected cells, demonstrating that the added RNA-N template had no VSV polymerase activity and that the mRNA accumulation detected in the other lanes was dependent on VSV proteins expressed from plasmids. To confirm that the wt and mutant L proteins accumulated in cells to similar levels, cell extracts were analyzed by Western blotting using an antibody to L (Fig. 1 and 2, lower panels). For these and other RNA synthesis experiments, we only considered changes of 50% or greater as a dividing line for the determination of phenotype. As shown in Fig. 1 and 2 and summarized in Table 2, while L protein synthesized at 34°C in HEp-2 and BHK cells with the N505D mutation (HR1-0) gave levels of transcription actually exceeding wt, D1671V (HR1-1) and the double mutant (HR1-0,1) decreased transcription by at most 30%. These data show that viral mRNA synthesis itself is not significantly host restrictive in hr1. However, at 40°C the presence of the D1671V L mutation in HR1-1 and HR1-0,1 resulted in a reduction of mRNA synthesis (50 to 66%) in both cell lines (Fig. 1B and 2B and Table 2), suggesting that this mutation is at least in part accountable for the ts phenotype of hr1 in cells (58, 59).
![]() View larger version (23K): [in a new window] |
FIG. 1. In vitro mRNA synthesis with VSV L mutants in HEp-2 cytoplasmic extracts. VVT7-infected HEp-2 cells were transfected with either no plasmids (mock) or wt VSV P plasmid plus wt or mutant VSV L plasmid and incubated at 34°C (A) or 40°C (B) for 18 h. (Upper panels) Cytoplasmic extracts were prepared and incubated at 30°C with polymerase-free VSV RNA-N template in the presence of [ -32P]CTP. Labeled RNA products were purified and separated by agarose-urea gel electrophoresis. The positions of the N and P mRNAs are indicated. (Lower panels) Immunoblots with L antibody of a portion of the extracts used for in vitro transcription. The position of the L protein is indicated.
|
![]() View larger version (25K): [in a new window] |
FIG. 2. In vitro mRNA synthesis with VSV L mutants in BHK cytoplasmic extracts. VVT7-infected BHK cells were transfected with either no plasmids (mock) or wt VSV P plasmid plus wt or mutant VSV L plasmid and incubated at 34°C (A) or 40°C (B) for 18 h. (Upper panels) Cytoplasmic extracts were prepared from transfected cells and incubated at 30°C with polymerase-free VSV RNA-N template in the presence of [ -32P]CTP. Labeled RNA products were purified and separated by agarose-urea gel electrophoresis. The positions of the N and P mRNAs are indicated. (Lower panels) Immunoblots with L antibody of a portion of the extracts used for in vitro transcription. The position of the L protein is indicated.
|
|
View this table: [in a new window] |
TABLE 2. In vitro transcription with VSV wt and mutant L proteins expressed under different conditions
|
|
View this table: [in a new window] |
TABLE 3. VSV CAT minigenome reporter assay with wt and mutant L proteins in HEp-2 cells
|
|
View this table: [in a new window] |
TABLE 4. Comparative titers of recombinant VSV mutants in BHK and HEp-2 cell lines at 34°C and 40°C
|
-32P]CTP or [
-32P]GTP as the labeled substrates. In agreement with the previous results (1, 30), the presence of AdoMet did not affect the efficiency of viral mRNA synthesis (data not shown). Consistent with the plasmid-based in vitro transcription results, the N505D mutation in rHR1-0 did not affect the L activity, whereas the presence of D1671V in the L protein of rHR1-1, rHR1-0,1, and the original hr1 virus somewhat reduced (30 to 60%) mRNA accumulation (Fig. 3A). Immunoblot analysis showed that similar amounts of L protein were present in each reaction mixture (Fig. 3B).
![]() View larger version (48K): [in a new window] |
FIG. 3. In vitro transcription with detergent-activated purified mutant VSV. BHK cells were infected with recombinant wt or mutant viruses or the original hr1 mutant at 34°C, and viruses were purified and used for in vitro mRNA synthesis in the presence of [ -32P]CTP as described under Materials and Methods. (A) Labeled RNA products were purified and separated by agarose-urea gel electrophoresis. The positions of the N and P mRNAs are indicated. The products were quantitated using a phosphorimager, and results are shown as the percent transcription (Txn) relative to the level of wt VSV transcription (set as 100%). The number represents the average of four to five experiments where the variation was less than 15%. (B) Immunoblot analysis of the viruses used for in vitro transcription using L antibody. The position of the L protein is indicated.
|
-32P]GTP (in the absence of AdoMet). The RNA products were untreated or TAP treated as indicated, and the products were separated by thin-layer chromatography (TLC) as described under Materials and Methods (Fig. 4). No release of [
-32P]GMP was observed when an uncapped [
-32P]GTP-labeled T7-directed VSV N transcript was used or when GpppA (5 mM) competitor was added to the TAP reaction mixture with wt VSV-produced mRNAs (Fig. 4), confirming the strict specificity of TAP for capped mRNAs (16, 55). However, TAP digested all the mRNA transcribed by each virus, as shown by the release of GMP (Fig. 4). Therefore, each mRNA was capped at the 5' end regardless of the L mutation, which is in agreement with our previous results indicating that hr1 mutations did not affect the mRNA capping ability of the viral polymerase (30).
![]() View larger version (20K): [in a new window] |
FIG. 4. Mutant L VSV synthesize 5'-capped mRNAs. [ -32P]GTP-labeled viral mRNA produced by the indicated detergent-disrupted purified viruses (lanes 3 to 9) or uncapped mRNA controls (lanes 1 and 2) were tested for the presence of the 5'-cap structure. In vitro transcription by purified viruses was conducted in the absence of AdoMet. Equivalent labeled RNA products normalized to wt RNA level were purified and digested with TAP to remove the 5'-terminal GMP specifically from the capped mRNAs (lanes 2 and 4 to 9) or incubated with the same reaction components but in the absence of TAP (lanes 1 and 3). Lane 4 shows wt VSV mRNA treated with TAP in the presence of competitor GpppA cap analog (5 mM). The reaction products were spotted onto PEI plates and analyzed by TLC, along with unlabeled GMP standard. The positions of the GMP (marker was developed under UV light) and undigested mRNA are indicated.
|
![]() View larger version (24K): [in a new window] |
FIG. 5. Analysis of mRNA cap methylation by mutant VSV. The indicated purified viruses were used for in vitro mRNA synthesis in the presence of 0.3 µM [3H]AdoMet and in the presence or absence of 100 µM competitor AdoMet or 100 µM inhibitor AdoHcy, and the RNA products were purified. (A) The RNA products were separated by agarose-urea gel electrophoresis and visualized by fluorography. The positions of the N and P mRNAs are indicated. (B) Relative methylation of the RNA products shown in panel A was measured by [3H]RNA binding to DEAE-cellulose paper and scintillation counting. (C) The RNA products shown in panel A were digested with nuclease P1, and the products were spotted onto PEI plates and analyzed by TLC, along with cap standards. The developed PEI plates were visualized by fluorography. The positions of the GpppAm and 7mGpppAm markers (developed under UV light) are indicated.
|
|
|
|---|
In addition to studies on the expressed L proteins, recombinant viruses with the individual L mutations as well as one with both mutations to reconstruct the original mutant were characterized. RNA synthesis with purified viruses, which was done at the permissive temperature, showed a reduction (30 to 40%) with rHR1-1 similar to that of the original hr1 mutant (Fig. 3), consistent with the analysis of L protein above. The reconstructed rHR1-0,1 gave less RNA synthesis than hr1, but this might be accounted for by differences in the genetic background of the L genes in the two viruses or perhaps differences in virus preparations. The virus infection data clearly show that the D1671V L mutation in the rHR1-1 virus is solely accountable for both the hr and ts phenotypes of the original hr1 mutant (Table 4) and, further, this mutation completely abolishes all cap methylation while not affecting capping of the mRNA (Fig. 4 and 5).
Together, these data are in agreement with our previous studies demonstrating that the coinfection of nonpermissive HEp-2 cells with hr1 and poxvirus yielded a permissive infection for hr1 and the accumulation of VSV mRNAs having fully methylated caps (7mGpppAm), suggesting that poxvirus rescued hr1 by converting the VSV mRNAs to a translationally active form due to the methylation by the cytoplasmic poxvirus mRNA MTase enzymes (29). Accordingly, we concluded that the hr phenotype of hr1 was independent of the activity of the viral MTase in different cell lines but was a result of differences between cytoplasmic levels (higher in permissive cells) of the host MTase enzymes, such that the unmethylated VSV mRNA synthesized by the viral polymerase becomes properly modified in permissive cells (29, 30).
Previously, in vitro results by Testa and Banerjee indicated that two different activities, RNA guanine-N7-MTase and RNA 2'-O-MTase, were associated with VSV RdRp, and they proposed the following order of MTase reactions: GpppA plus AdoMet (low concentration)
GpppAm plus AdoMet (high concentration)
7mGpppAm (69). In agreement with this scheme, wt VSV L protein supported synthesis of viral mRNA that had monomethylated GpppAm cap or the fully methylated 7mGpppAm cap in the presence of 0.3 µM AdoMet. However, the L D1671V mutation abolished mRNA methylation at both positions, guanine-N7 and 2'-O-adenosine, resulting in the synthesis of viral mRNAs having unmethylated GpppApApCp 5' ends. We propose two possibilities explaining how a single amino acid change abolishes two different activities of the L protein. The first hypothesis is based on the possibility that a single AdoMet-binding site serves both MTase activities in the L protein. Therefore, a loss of AdoMet-binding ability in the L D1671V mutant would consequently inactivate both RNA guanine-N7 MTase and RNA 2'-O MTase activities. The presence of a single AdoMet-binding site serving both RNA guanine-N7 and RNA 2'-O MTase domains was also proposed for the
2 protein of reoviruses (36). Alternatively, the D1671V mutation could specifically inactivate the 2'-O MTase activity of the L protein and, consequently, also guanosine methylation in the ordered process observed in vitro (69). However, our previous in vivo data on VSV mRNA synthesis in the presence of the methylation inhibitor cycloleucine (39) and in vitro transcription data on VSV New Jersey serotype (24) suggest that the reverse order of VSV mRNA methylation (GpppA
7mGpppA
7mGpppAm) can also occur. Therefore, a complete defect of the L mutant D1671V in mRNA methylation would not be explained by a specific block of a single MTase activity.
Crystal structures exist for various AdoMet-dependent MTases (38), including the cellular guanine-N7 cap MTase (18), the nucleoside-2'-O cap MTases of vaccinia virus (26) and dengue fever virus (17), and the guanine-N7 and nucleoside-2'-O cap MTase domains of the
2 core protein of reovirus (48). While AdoMet-dependent MTases show a very low degree of homology at the amino acid level, they share a high degree of structural similarity and the presence of the glycine-rich motif (named motif I), which is a part of the catalytic domain directly involved in AdoMet binding and where methylation actually occurs (31, 38).
The hr1 mutation D1671V resides within domain VI of the L protein (Fig. 6A). Based on the presence of the glycine-rich region, this domain was originally predicted as a general NTP-binding site (45) and thought to be involved in P protein phosphorylation occurring at least in some NNS RNA viruses (51). Contrary to that assignment, recent independently conducted computational analyses (8, 19) proposed that domain VI of the L protein may, in fact, be associated with the MTase function in NNS RNA viruses. These data predicted that domain VI of L proteins has distinct features of a prototypical MTase fold which incorporates seven ß-strands and six
-helices, forming a seven-stranded ß-sheet with three
-helices on each side (38). Moreover, these predictions (8, 19) argued that the glycine-rich region, while present in many protein kinases (72), is a canonical signature motif shared by all members of the AdoMet-dependent MTase superfamily. Remarkably, the D1671V mutation lies within this glycine-rich motif (G-D1671-G-S-G in VSV), a region which is highly conserved among representative members of the order Mononegavirales (Fig. 6A) (57) and shares homology with the glycine-rich motifs of various mRNA guanine-N7 cap MTases (Fig. 6B) and mRNA nucleoside-2'-O cap MTases (Fig. 6C). For NNS RNA viruses, charge is conserved at position 1671; however, in other enzymes neither amino acid nor charge is conserved. The importance of the glycine motif, especially the second glycine residue, for function was previously demonstrated by site-directed mutagenesis of several cap mRNA MTases, including those of vaccinia virus (37, 52), reovirus (36), and eukaryotic cells (70, 74). Mutational analysis was not done on these other enzymes at a position equivalent to 1671 in VSV L. Our findings experimentally support the computational predictions (8, 19) assigning the MTase function(s) to an AdoMet-binding site in domain VI in the L protein in NNS RNA viruses, but further studies will be required to understand the required regions of both the guanine-N7 and 2'-O MTase activities.
![]() View larger version (64K): [in a new window] |
FIG. 6. Sequence alignment of the putative AdoMet-binding motif of VSV L protein with various homologs. Multiple alignment of the putative AdoMet-binding motif within domain VI of the L proteins of Mononegavirales is shown (A), along with their comparison to the AdoMet-binding motifs of known mRNA guanine-N7 cap MTases (B) and mRNA nucleoside-2'-O cap MTases (C). Database accession numbers are shown for all the presented proteins, and starting amino acid numbers are shown for the L proteins. Gaps in the sequences are indicated by dashes. Positions shown in bold are conserved within panels A, B, and C, while the grey shaded boxes show amino acid positions conserved among all the proteins.
|
This work was funded by NIH grant AI14594 (to S.A.M.).
Present address: Vaccinex, Inc., Rochester, NY 14620. ![]()
|
|
|---|
2 protein. Evidence for activities in mRNA cap methylation. J. Biol. Chem. 273:23773-23780.
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»