Arch Virol (2007) 152: 827–832 DOI 10.1007/s00705-006-0881-1 Printed in The Netherlands
Brief Report Effects of single amino acid substitutions at the E residue in the conserved GDNE motif of the Nipah virus polymerase (L) protein D. E. Magoffin1;2 , K. Halpin1;3 , P. A. Rota3 , and L.-F. Wang1 1
CSIRO Livestock Industries, Australian Animal Health Laboratory, Geelong, Vic, Australia Curtin University of Technology, Perth, WA, Australia 3 Centers for Disease Control and Prevention, Atlanta, GA, U.S.A. 2
Received August 2, 2006; accepted October 20, 2006; published online December 4, 2006 # Springer-Verlag 2006
Summary
Nipah virus (NiV) is an emergent zoonotic paramyxovirus. The L proteins of most paramyxoviruses contain a GDNQ motif, thought to be part of the catalytic site for polymerase activity. Conversely, NiV L has GDNE in this position. We substituted the E residue with eight different amino acid residues and examined the effect on L function in an in vitro replication assay. Our results demonstrated that NiV L functioned with similar efficiency with either GDNE or GDNQ, but polymerase activity was severely reduced or abolished when a structurally destabilising residue (such as K, P or G) was introduced at this site. Nipah virus (NiV) was first isolated from cerebrospinal fluid of patients during an outbreak of febrile encephalitis in 1998–1999 that claimed the lives of more than 100 people in Malaysia [5] and Singapore [19]. The outbreaks were controlled by Author’s address: Dr. Lin-Fa Wang, CSIRO Australian Animal Health Laboratory, PO Bag 24, Geelong, Victoria 3220, Australia. e-mail:
[email protected]
culling of more than one million pigs. Serological and molecular characterisation of NiV revealed that the virus was closely related to Hendra virus (HeV), a zoonotic virus isolated during a disease outbreak in Australia in 1994 that resulted in the death of one human and 13 horses [16, 18, 10]. Recently, NiV was found to be the causative agent responsible for outbreaks of fatal febrile encephalitis in Bangladesh in 2001, 2003–2005 [11, 7, 9] and in India in 2001 [3]. Fruit bats in the genus Pteropus have been identified as the natural reservoir host for both NiV and HeV [29, 30]. Serological, virological and molecular studies have revealed the presence of NiV or NiV-like viruses in fruit bat populations in Malaysia, Cambodia, Thailand and Indonesia [29, 17, 23, 27, 25]. NiV and HeV have been classified in a new genus, Henipavirus, due to their unique virological and genomic features [28, 22]. For non-segmented negative-stranded (NNS) RNA viruses (rhabdoviruses, paramyxoviruses, filoviruses and bornaviruses), the L proteins (RNA polymerase) have a high level of sequence conservation. Sequence comparison of paramyxovirus L proteins revealed six conserved domains (I–VI) separated by variable regions, suggesting that the structure
828
consisted of concatenated functional domains [20]. Insertion of the EGFP reporter gene between domains II and III of the L protein of Rinderpest virus (RPV) [2] and measles virus (MeV) [6] did not disrupt polymerase function, indicating that the polymerase was a multi-domain protein. Domain II is essential for RNA recognition and domain III contains the GDNQ motif, which has been hypothesised to be an active site for phosphodiester bond formation. The GDNQ sequence is located within the C motif of domain III, with a b-turn-b structure critical for polymerase activity [21]. Specific locations of amino acid residues in these tight turns may be critical for correct orientation, cation binding, template specificity and catalytic processes. Prior to the discovery of HeV, all known NNS RNA viruses contained the highly conserved GDNQ motif. For henipaviruses, the Q residue has been replaced by an E, giving rise to a GDNE motif. This alternate motif has also been found in two recently characterised paramyxoviruses, Tupaia paramyxovirus (TuPV) and Mossman virus (MoV) [15]. The amino acid change introduces a second negatively charged residue into a predicted functionally important area of the protein. Initially, we postulated that the GDNE motif might be associated with paramyxoviruses with a relatively large genome size. However, it was recently found that J-virus and Beilong virus, two paramyxoviruses with the largest genomes of known NNS RNA viruses, contained the GDNQ motif in their L proteins [12, 14]. In this study, we examined the role of the E residue in the GDNE motif in the function of the NiV L protein. We replaced E with Q or other amino acid residues and assessed the activity of the L protein by using a NiV-GFP minigenome replication assay. Results demonstrated that the E residue within the GDNE motif was interchangeable with a Q residue. Furthermore, although the E residue could be re-
D. E. Magoffin et al.
placed with D, A, N, I and G residues, albeit with considerable reduction in activity, replacement with K or P residues resulted in abrogation of polymerase activity in the NiV minireplicon system. The NiV minigenome pNiV-CAT [8] was used as a template for the construction of the pNiV-GFP minigenome. A DNA fragment containing the NiV L gene untranslated region (UTR), NiV trailer region and T7 promotor sequence was amplified from the pNiV-CAT minigenome and was cloned into the BamHI and HindIII sites of a modified pUC19 vector. A second DNA fragment containing the NiV leader region, NiV N UTR, hepatitis delta virus ribozyme sequence and two T7 terminator sequences was amplified from the pNiV-CAT minigenome and cloned into the EcoRI and SmaI sites of the vector. The EGFP reporter gene was removed from the pCI-EGFP plasmid [13] and inserted into the NcoI and NotI sites of the NiV construct, generating the pNiV-GFP minigenome. The ‘rule of six’ was maintained in pNiV-GFP. A schematic representation of the pNiV-GFP plasmid is shown in Fig. 1. The NiV support plasmids consisted of the NiV N, P and L genes cloned individually into the pTM1 vector as previously described [8]. The recombinant fowlpox virus, fpEFLT7pol (FWPV-T7), which stably expresses T7 RNA polymerase in mammalian cells was kindly provided by Dr. M. Skinner [1].The function of the minireplicon system was assessed in Vero cells. Cells were inoculated with FWPV-T7 at an m.o.i. of 10. After infection, cells were transfected with a predetermined optimal concentration of minigenome and support plasmids (0.5 mg minigenome, 0.3 mg N, 0.2 mg P and 0.1 mg L) using Lipofectamine 2000 (Invitrogen). Transfected cells were examined after 72 hrs with an IX71 microscope (Olympus), and the images were captured by DP70 camera (Olympus) using the Analysis+ program (Soft Imaging System). Cells expres-
Fig. 1. Schematic representation of the NiV-GFP minigenome construct. Positions of major restriction sites used in this construct are shown above. HDV hepatitis delta virus ribozyme; UTR untranslated region
Mutational analysis of Nipah virus L protein GDNE motif
829
sing GFP were quantified as number of fluorescent cells per field of view using the Universal TEM Imaging Platform iTEM (Soft Imaging System). Each minireplicon assay was performed in replicates of eight, and experiments were carried out three times. Expression of GFP only occurred when all four plasmids were co-transfected, indicating the system was functional (data not shown). It should be noted that although this method is semi-quantitative, the direct measurement of fluorescence without additional processing steps makes this procedure much more convenient, allowing more replicate experiments to be conducted at once. This system is therefore as effective as other enzyme-based reporter systems such as CAT or luciferase in the quantitative analysis of minigenome replication activities. Mutagenesis of the GDNE motif was performed using a PCR-based approach. The GDNE coding region contains a StyI site that is unique between the BstBI and BsiWI sites of the NiV L gene. Two overlapping PCR products containing point mutations for an E to Q change were digested with StyI, ligated and used as a template for a third PCR. After digestion with BstBI and BsiWI, the mutated PCR fragment was cloned into pTM1-NiV-L, replacing the wild-type BstBI–BsiWI fragment. The L gene region between BstBI and BsiWI was sequenced to ensure that no other point mutation was unintentionally introduced. When the activity of NiV L-GDNE (wild-type) was compared to that of NiV L-GDNQ, the sequence
motif common to most other NNS RNA viruses, there was no significant difference between the two (Fig. 2A and B). GFP was not expressed when the L support plasmid was omitted from the minireplicon system (Fig. 2C). The same results were obtained when either the N or P support plasmid was omitted (data not shown). Considering that the Q residue is conserved among all NNS RNA viruses with the exception of HeV, NiV, TuPV and MoV, it was surprising to find that the exchange of E and Q at this site had minimal effect on GFP expression. Given that identical conditions were used for each rescue experiment with the same amount of L plasmid added, the level of GFP expression should directly correlate with NiV L protein activity and the replication efficiency of the NiV minigenome. To determine whether activity of the polymerase could be affected by introducing a bulky, positively charged residue or a residue known to disrupt protein secondary structure, two additional L gene mutants, GDNK and GDNP, were constructed using the same strategy as for GDNQ. These amino acid substitutions abolished GFP expression, demonstrating that the K and P residues are not tolerated at this position (Fig. 3). To determine the tolerance for amino acid substitutions at this position, five additional L gene mutants were constructed. These were GDND, GDNA, GDNN, GDNI and GDNG motifs. Although GFP expression was evident for most of these mutants, the level of GFP produced was significantly lower (P
