Bluetongue virus non-structural protein 1 is a positive regulator of viral protein synthesis
© Boyce et al.; licensee BioMed Central Ltd. 2012
Received: 31 October 2011
Accepted: 24 August 2012
Published: 29 August 2012
Skip to main content
© Boyce et al.; licensee BioMed Central Ltd. 2012
Received: 31 October 2011
Accepted: 24 August 2012
Published: 29 August 2012
Bluetongue virus (BTV) is a double-stranded RNA (dsRNA) virus of the Reoviridae family, which encodes its genes in ten linear dsRNA segments. BTV mRNAs are synthesised by the viral RNA-dependent RNA polymerase (RdRp) as exact plus sense copies of the genome segments. Infection of mammalian cells with BTV rapidly replaces cellular protein synthesis with viral protein synthesis, but the regulation of viral gene expression in the Orbivirus genus has not been investigated.
Using an mRNA reporter system based on genome segment 10 of BTV fused with GFP we identify the protein characteristic of this genus, non-structural protein 1 (NS1) as sufficient to upregulate translation. The wider applicability of this phenomenon among the viral genes is demonstrated using the untranslated regions (UTRs) of BTV genome segments flanking the quantifiable Renilla luciferase ORF in chimeric mRNAs. The UTRs of viral mRNAs are shown to be determinants of the amount of protein synthesised, with the pre-expression of NS1 increasing the quantity in each case. The increased expression induced by pre-expression of NS1 is confirmed in virus infected cells by generating a replicating virus which expresses the reporter fused with genome segment 10, using reverse genetics. Moreover, NS1-mediated upregulation of expression is restricted to mRNAs which lack the cellular 3′ poly(A) sequence identifying the 3′ end as a necessary determinant in specifically increasing the translation of viral mRNA in the presence of cellular mRNA.
NS1 is identified as a positive regulator of viral protein synthesis. We propose a model of translational regulation where NS1 upregulates the synthesis of viral proteins, including itself, and creates a positive feedback loop of NS1 expression, which rapidly increases the expression of all the viral proteins. The efficient translation of viral reporter mRNAs among cellular mRNAs can account for the observed replacement of cellular protein synthesis with viral protein synthesis during infection.
Bluetongue virus (BTV) is the prototype member of the Orbivirus genus of the Reoviridae family. BTV is an arbovirus vectored between ruminant hosts by biting midges of the Culicoides genus [1, 2], causing disease in wild and domesticated ruminants, including sheep, goats and cattle. The virus has a global distribution which has extended into mainland Europe since 1998 and has caused a severe economic impact on European livestock farming [3, 4]. BTV has a segmented genome consisting of ten linear double-stranded RNA (dsRNA) molecules [5, 6]. Like other members of the Reoviridae the viral messenger RNA (mRNA) is synthesised by transcription of the double-stranded RNA (dsRNA) genome segments which remain packaged within the uncoated core particle [7–9]. The ten viral mRNAs are synthesised as complete end-to-end copies of the genome segments by the RNA-dependent RNA polymerase (RdRp) component of the transcription complexes, VP1, present at the vertices of the core particle [9–11]. Before extrusion into the cytosol each transcript is capped and methylated by VP4 at the 5′ end, to produce a cap1 structure (7mGpppGm) identical to that found in cellular messenger RNAs [12–14]. The 5′ cap1 structure is almost ubiquitous among cellular mRNAs and is important for mRNA stability and efficient translation . The 3′ end of viral transcripts lack a poly(A) tail, which normally functions in eukaryotic mRNAs to increase the rate of translation initiation and RNA stability [15–18]. Following the infection of mammalian cells BTV proteins rapidly accumulate to be the most abundant cytoplasmic proteins, with the viral proteins being the major products of active translation by 8 hours post-infection . This rapid replacement of cellular protein synthesis with viral protein synthesis indicates that BTV has effective mechanism(s) for ensuring the translation of its mRNA, but how the viral mRNAs of orbiviruses effectively compete with cellular mRNAs for translation has not been addressed to date.
A characteristic of Orbivirus infections is the accumulation of highly abundant cytoplasmic tubular structures composed of the 64 kDa non-structural protein 1 (NS1), for which no homologue has been found in other genera of the Reoviridae[20–22]. NS1 is the most highly expressed BTV protein, becoming the most abundant cytoplasmic protein before completion of the replication cycle. Understanding of the functions of NS1 remains limited to a proposed role in regulating the mechanism of virus release from the mammalian and insect host cells . In this model the high ratio of NS1 to NS3 in mammalian cells is correlated with cell lysis, and the higher level of NS3 expression observed in Culicoides cells is correlated with non-lytic release and persistent infection in culture.
Here we use a synthetic mRNA reporter assay system to screen the BTV gene products for upregulation of the expression of a reporter gene consisting of a fusion between the termini of segment 10 and GFP. The mRNA reporter contains GFP flanked by the 5′ end (152nt) and 3′ end (149nt) of segment 10, creating a reporter mRNA which has both termini of segment 10. Using this approach NS1 is identified as sufficient to upregulate the reporter protein abundance. The specificity of NS1 for viral mRNAs is investigated using reporter mRNAs consisting of BTV untranlsated regions (UTRs) flanking the reporter ORF, and using mRNAs with 3′ poly(A) extensions. Using the reverse genetics system we demonstrate that NS1-dependent upregulation of gene expression also occurs in the context of viral replication in infected BSR cells. To our knowledge this is the first report addressing how orbiviruses upregulate their gene expression.
The BTV proteins have been shown to be the most highly expressed genes in BHK cells by 8 hours post infection . The approach used here identifies the tubule forming protein characteristic of orbiviruses, NS1, as a positive regulator of the translation of viral mRNAs. The screen of the BTV genes for proteins which regulate the expression from the segment 10-GFP mRNA identified NS1 as sufficient, and no effect on GFP expression was observed with other BTV proteins. This approach demonstrates that NS1 is a positive regulator of BTV gene expression, but the possibility that complexes containing more than one viral protein also regulate BTV gene expression is not excluded.
The increased expression was also found when the reporter mutants retained the 5′ and 3′ regions of segment 10, but did not encode the amino and carboxy termini of NS3. The NS1-responsiveness of these variants demonstrates that the short NS3 protein sequences fused to GFP cannot be responsible for the increased accumulation of the reporter protein. This was confirmed by the NS1 responsiveness of transcripts which consist of only the UTRs of BTV mRNAs fused to the Renilla luciferase ORF. Together these results suggest that NS1 acts at the level of the mRNA to either increase its stability and therefore abundance, or increase the rate at which it is used as a template in translation.
The fusion of the Renilla luciferase gene with segment 10 in the S10-Rluc mRNA produced a quantitative and more sensitive reporter system for measuring NS1 responsiveness, with NS1 enhancing the expression from the S10-Rluc mRNA in transfected cells by 15–20 fold. The upregulation of expression from reporter mRNAs containing the UTRs of five BTV genes demonstrates that NS1-dependent gene expression is not restricted to the expression of the NS3 protein from segment 10, but is a wider phenomenon affecting the expression of other BTV genes.
Importantly, the type of transcripts upregulated by NS1 does not extend to the Rluc(A)n mRNA which lacks viral sequences and has a typical cellular 3′ poly(A) tail, demonstrating that NS1 is not a general upregulator of expression from all mRNAs. The 3′ extension of the NS1-responsive S10-Rluc mRNA with a poly(A) sequence was sufficient to prevent increased expression, showing that the nature of the 3′ end of the mRNA is crucial to the upregulation function of NS1. Moreover the absence of upregulation of the polyadenylated mRNAs confirms that NS1 does not increase viral gene expression by globally increasing protein stability. Together these data support a model where NS1 is a positive regulator of viral mRNA expression, but does not increase the expression of cellular mRNAs. The competitive advantage this would give to the translation of viral transcripts can account for the replacement of cellular protein synthesis with viral protein synthesis in BTV infected BHK cells .
The BTV UTRs themselves were found be one level of regulation of the expression of the viral proteins, with the absolute levels of expression from each of the UTR-containing reporters varying markedly between the segments. This variation occurred both in the presence and absence of NS1 pre-expression, with NS1 increasing the quantity of luciferase produced from each mRNA by 10–30 fold. The BTV UTRs were found to be determinants of the protein expression level as is established for the UTRs of cellular mRNAs , and the upregulatory effect of NS1 is an additional level of regulation. The contribution of the UTRs to the expression level is particularly marked for the segment 6 UTRs which flank the NS1 gene itself, the most highly expressed BTV gene during infection. In this case the level of translation seen in the absence of NS1 is already high, and the absolute quantity of luciferase produced when NS1 is upregulating expression is higher than the other UTRs tested.
When rescued into the genome of infectious virus using the reverse genetics system, the expression of the S10-Rluc reporter genome segment was also responsive to NS1. The expression of NS1 using pCAG NS1 increased the expression of the reporter at early time points, demonstrating that the NS1-dependent upregulation occurs in the context of virus replication where the viral mRNA is synthesised by transcriptionally active cores. The diminishing effect observed with time is an expected consequence of the abundant and rapid synthesis of NS1 by the infecting virus, which would be expected to rapidly exceed the initial difference in NS1 abundance between cells transfected with pCAG NS1 and those transfected with the pCAGGS vector. The higher level of reporter expression seen at early times after infection with the BTV-1 S10Rluc virus demonstrates the competitive advantage given to viral gene expression by the presence of NS1 in the infected cells. Indeed the recovery of viruses by reverse genetics with no NS1 initiation codon, or coding changes which abrogate the regulatory function was not possible (data not shown). Knockdown by siRNA of the moderate level of NS1 expression from pCAG NS1, compared to BTV infection, have been insufficient to reduce reporter activity by more than 2 fold (data not shown). This observation is consistent with the amount of NS1 produced by transfection with small quantities of pCAG NS1 plasmid being sufficient to highly upregulate expression from reporter mRNAs (Figure 2B).
NS1 performing the role of a positive regulator of BTV gene expression is consistent with it being the most highly expressed BTV protein encoded by the most abundant transcript produced by transcribing BTV cores [9, 19]. We propose a model of translational regulation in which NS1 increases the expression of the BTV genes, including itself, creating a positive feedback loop of NS1 expression which increases the rate of synthesis of all the viral proteins, as has been found for the rotavirus NSP3 protein . The increased translation of viral transcripts in the presence of NS1 suggests that a competitive advantage is given to the translation of viral transcripts over cellular polyadenylated mRNAs, and is consistent with the observed shutdown of cellular gene expression and its simultaneous replacement with viral gene expression in BTV infected cells by 8 hours p.i. reported by Huismans . Altogether the data presented here demonstrate that the effective competition of viral mRNAs with cellular mRNAs, mediated by NS1, is one way that this is achieved. BTV NS1 and rotavirus NSP3 do not share amino acid sequence homology and are not similar size proteins (64 kDa versus 34 kDa respectively), but NS1 clearly performs the translational upregulation function also reported for rotavirus NSP3 [29–32], although the mechanism leading to increased expression in the presence of NS1 remains to be elucidated. The existence of cis-acting sequences within the viral UTRs, particularly the 3′ UTR, which allow NS1 to discriminate between viral mRNAs and cellular mRNAs, is one prediction of the model which is being investigated.
This is the first report addressing how orbiviruses replace cellular protein synthesis with viral protein synthesis. NS1 is identified as sufficient to increase protein synthesis from chimeric virus-reporter gene mRNAs in both transfected and infected cells. The UTRs of segments 1, 4, 6, 8, and 10 are sufficient to confer NS1-responsiveness to protein synthesis. The abrogation of NS1 responsiveness by 3′ poly(A) extension of the mRNA demonstrates that the 3′ end is a critical determinant of NS1 responsiveness. Knowledge of the structure of NS1 and its interaction with viral mRNAs will be important in understanding how NS1 specifically increases viral protein synthesis in orbiviruses.
BSR cells (a BHK-21 subclone) were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 5% (v/v) foetal bovine serum (FBS) at 35°C in 5% CO2. The NS3 expressing cell line BSR NS3 was cultured in the presence of puromycin (7.5 μg ml-1) .
BTV-1 stocks were generated by infecting BSR cells at a multiplicity of infection (MOI) of 0.1 and harvesting the medium at 3 days for wildtype BTV-1 or 7 days for BTV-1 S10-Rluc. Viral stocks were stored at 4°C.
Transfected BSR cells were infected at an MOI of 1 at room temperature for 45 minutes. Wells were washed four times in PBS, and incubated in DMEM supplemented with 5% (v/v) FBS at 35°C in 5% CO2.
The GFP-containing clone pS10-GFP consists of the 5′ 152nt and 3′ 149nt of BTV-10 segment 10 fused in frame with the GFP ORF. Mutants of pS10-GFP which did not fuse the remaining BTV protein coding sequences to GFP were generated by site directed mutagenesis using the method of Weiner et al. , with the primer pairs S10GFP_mut1F and S10GFP_mut1R (mutants 1 and 5), S10GFP_mut2F and S10GFP_mut2R (mutants 2, 4, and 5), S10GFP_mut3F and S10GFP_mut3R (mutants 4 and 5), S10GFP_mut4F and S10GFP_mut4R (mutants 4 and 5).
The Renilla luciferase-containing clone pS10-Rluc consists of the 5′ 152nt and 3′ 149nt of BTV-10 segment 10 fused in frame with the Rluc ORF.
The Renilla luciferase clones containing the 5′ and 3′ UTRs of segments 1, 4, 6, 8 or 10 of BTV-1 consist of the Renilla luciferase ORF immediately flanked by the UTRs.
The pRL-null (Promega) vector contains the Renilla luciferase ORF downstream of a T7 promoter.
BTV protein expression plasmids were generated by cloning the ten BTV-10 ORFs downstream of the chicken β actin promoter of the pCAGGS expression vector reported previously , to generate pCAG VP1, pCAG VP2, pCAG VP3, pCAG VP4, pCAG VP5, pCAG VP6, pCAG VP7, pCAG NS1, pCAG NS2 and pCAG NS3.
Synthetic single-stranded RNAs were prepared by run-off in vitro transcription from BsmBI digested plasmid clones, using T7 RNA polymerase. Transcripts were prepared with anti-reverse cap analogue (ARCA) from linear plasmid DNA using the mMESSAGE mMACHINE T7 Ultra kit (Ambion) as previously described . Poly(A) tailed transcripts were generated using poly(A) polymerase according to the manufacturer’s instructions.
Transcripts were analysed under denaturing conditions by electrophoresis on 1% agarose in morpholinepropanesulfonic acid (MOPS) electrophoresis buffer in the presence of formaldehyde .
Samples were resolved in 10% SDS-PAGE gels and transferred to nitrocellulose membranes. Standard immunoblotting techniques were used with rabbit polyclonal antiserum R633 raised against NS1. Proteins were visualised using alkaline phosphatase-conjugated goat anti-rabbit IgG (Sigma) with BCIP/NBT substrate (Sigma).
BSR cultures were infected with BTV at an MOI of 0.1. Transcriptionally active BTV-1 core particles were purified as previously described and stored at 4°C .
BTV core particles were incubated at 40 μg ml-1 at 30°C for 5 to 6 h in core transcription buffer (100 mM Tris HCl pH8.0, 4 mM ATP, 2 mM GTP, 2 mM CTP, 2 mM UTP, 500 μM S-adenosylmethionine, 6 mM DTT, 9 mM MgCl2, 0.5U μl-1 RNAsin Plus [Promega]). BTV core-derived mRNAs were purified as previously described , followed by an additional precipitation in isopropanol and resuspension in DEPC-treated water.
70-90% confluent BSR monolayers in 96 well plates were transfected with plasmid DNA using Lipofectamine 2000™ reagent (Invitrogen), according to the manufacturer’s instructions. Incubation of transfected cells was continued at 35°C in 5% CO2.
Growth medium was replaced with 100 μl fresh growth medium 18 h after transfection with BTV protein expression plasmids. Reporter transcripts were diluted in OptiMEM-1, 0.1U μl-1 RNAsin Plus (Promega), before mixing with Lipofectamine™ 2000 reagent. Monolayers were transfected with 25 ng of the reporter transcript unless otherwise indicated, and incubated at 35°C in 5% CO2 for 20 h.
GFP expression in transfected BSR cells was visualised using a Nikon Eclipse TS100 microscope and recorded using a Nikon Coolpix 995 digital camera, with the number of fluorescing cells per visual field recorded. When transfection included a complete infectious set of BTV genome segments then fixation in 4% paraformaldehyde in PBS for 10 minutes was performed prior to recording fluorescence.
The expression of Renilla luciferase in transfected or infected BSR cells was quantified with a Turner Biosystems Glomax luminometer using the Dual Luciferase Reporter Assay System (Promega) or the Renilla Luciferase Assay System (Promega), according to the manufacturer’s instructions. The mean and standard deviation were calculated for triplicate monolayers transfected or infected in 96 well plates.
Mutagenic primers used to generate variants of pS10-GFP were the following:
Mutants 1 and 5 (Introduce a termination codon at the C terminus of the GFP ORF and a −1 frame shift of NS3 relative to GFP); S10GFP_mut1F (5′GGCATGGACGAGCTGTACAAGTAAGTGGAAGTTCAAGATCTATCCCG3′) and S10GFP_mut1R (5′CGGGATAGATCTTGAACTTCCACTTACTTGTACAGCTCGTCCATGCC3′).
Mutants 2, 4 and 5 (NS3 initiation codon changed to AAG); S10GFP_mut2F (5′CTATAGTTAAAAAGTGTCGCTGCCAAGCTATCCGGGCTGATCC3′) and S10GFP_mut2R (5′GGATCAGCCCGGATAGCTTGGCAGCGACACTTTTTAACTATAG3′).
Mutant 4 and 5 (NS3 2nd ATG codon changed to ACG); S10GFP_mut3F (5′CCAAAGGTTCGAAGAAGAAAAGACGAAACACAATCAAGATAGAGTTGAAG3′) and S10GFP_mut3R (5′CTTCAACTCTATCTTGATTGTGTTTCGTCTTTTCTTCTTCGAACCTTTGG3′).
Mutants 4 and 5 (Reading frame altered so that GFP is in the +1 frame compared to NS3 N terminus); S10GFP_mut4F (5′CGAGTGCACCGACGCGTATCCAGTGGTGGAGGTATGGTG3′) and S10GFP_mut4R (5′CACCATACCTCCACCACTGGATACGCGTCGGTGCACTCG3′).Mutagenic bases are underlined.
This work was jointly supported by the Biotechnology and Biological Sciences Research Council (UK) and the National Institute of Health (USA).
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.