Substitution or deletion mutations between nt 54 and 70 in the 59 non-coding region of dengue type 2 virus produce variable effects on virus viability

Transcription

1 Journal of General Virology (2007), 88, DOI /vir Short Communication Correspondence Vijittra Leardkamolkarn Substitution or deletion mutations between nt 54 and 70 in the 59 non-coding region of dengue type 2 virus produce variable effects on virus viability Wipawan Sirigulpanit, 1 Richard M. Kinney 2 and Vijittra Leardkamolkarn 1 1 Department of Anatomy and Center for Vectors and Vector-Borne Diseases, Faculty of Science, Mahidol University, Bangkok 10400, Thailand 2 Division of Vector-Borne Infectious Diseases, Centers for Disease Control and Prevention, US Department of Health and Human Services, Public Health Service, Fort Collins, CO 80522, USA Received 9 August 2006 Accepted 17 February 2007 A C57U nucleotide mutation in a predicted RNA stem structure (nt 11 16/56 61) of the 59 non-coding region (59NCR) of dengue 2 (DEN-2) virus is partially attenuating, but unstable during serial passage of certain candidate DEN-2 PDK-53-based vaccine viruses containing this mutation. Here, 11 different mutations (one or more point substitution and/or deletion) between nt 54 and 70 in the 59NCR of the pd2/ic-30p-a (16681) infectious clone are described. Four mutants were infectious. Three mutants with single point substitutions replicated well in cell culture and exhibited variable neurovirulence in mice. Constructs containing multiple substitutions or any deletions failed to produce infectious viruses. Unexpectedly, a double C57U+G58C mutant replicated as efficiently as D2/IC-30P-A virus, and was more neurovirulent for newborn ICR mice. Thus, despite its predicted additional disruption of the RNA stem structure, the engineered contiguous secondary G58C mutation caused reversion of the partially attenuated phenotype caused by the 59NCR-C57U mutation. Dengue viruses are flaviviruses (genus Flavivirus, family Flaviviridae) that are transmitted to humans principally by Aedes aegypti or Aedes albopictus mosquitoes in tropical regions of the world. Based on immunological studies, dengue viruses are divided into four distinct antigenic serotypes: DEN-1, -2, -3 and -4. The disease caused by DEN virus infection ranges from a non-specific viral syndrome to severe and often fatal dengue haemorrhagic fever/dengue shock syndrome. DEN-2 is the most prevalent serotype that has been identified in outbreaks. DEN viruses cause more than 100 million cases of infection and an estimated deaths worldwide each year (Gubler, 1998). An effective vaccine against dengue is not yet available. The DEN virus genome is a single-stranded, positivesense RNA that is approximately 11 kb in length. It contains a type I cap at the 59 end, but lacks a 39-end poly(a) tail. The genome organization is 59 non-coding region (59NCR) capsid protein premembrane (prm)/ membrane protein envelope (E) protein non-structural protein 1 (NS1) NS2A NS2B NS3 NS4A NS4B NS5 39NCR (Lindenbach & Rice, 2003). The 59NCRs of flaviviruses and several other positive-strand RNA viruses are predicted to form functional secondary RNA structures (Brinton & Dispoto, 1988). These structures can interact with cellular or viral proteins to regulate viral RNA translation (Ali & Siddiqui, 1995), RNA transcription and packaging (Guesdon et al., 2001; Kuhn et al., 2002). 59NCR sequences may also interact with downstream sequences, particularly 39NCR sequences, in the viral genome during virus replication (Alvarez et al., 2005; You et al., 2001). The functional significance of the 59NCR of DEN-4 virus has been demonstrated by the attenuating nature of mutations in this region, including changes that disrupt predicted base pairing (Cahour et al., 1995). A C57U mutation in the 96 nt long 59NCR of the candidate DEN-2 PDK-53 vaccine virus has been associated with viral attenuation (Butrapet et al., 2000; Kinney et al., 1997). The phenotypic markers of attenuation of the PDK-53 virus, which was derived from wild-type DEN virus, are encoded by the three mutations NS1-Gly Asp PDK-53, 59NCR-C57U and NS3-Glu250Val, in order of dominance of effect on virus replication in vitro and attenuation of neurovirulence for newborn mice (Butrapet et al., 2000). The 59NCR-C57U mutation contributes to the PDK-53 viral phenotypes of small plaque size, decreased replication in C6/36 cells and attenuation of neurovirulence for mice (Butrapet et al., 2000). The 59NCR-57 locus lies within a predicted 6 bp stem structure consisting of nt and (Butrapet et al., 2000). This stem structure is one of several stems in a larger predicted 59 structure, designated SLA and consisting of nt 1 70, which has recently been shown to be recognized by the viral RNA polymerase (NS5) and to participate in cyclization of the genome through Printed in Great Britain

2 Effect of mutations in the 59NCR of DEN-2 virus interaction with the 39-end sequence, thereby promoting minus-strand RNA synthesis in D2/IC-30P-A clonederived DEN virus (Filomatori et al., 2006). Presumably, the C57U mutation impairs this function in PDK-53 virus. Serial in vitro passage of chimeric DEN viruses expressing the prm/e gene region of DEN-1, -2, -3 or -4 virus in the genetic background of PDK-53 virus has demonstrated high genetic stability of the NS1-53-Asp and NS3-250-Val loci. However, these chimeric viruses, as well as the PDK- 53 virus itself, showed varying propensities for 59NCR- U57C reverse mutation during serial passage. The extent of reversion in these viruses was recently examined quantitatively (Butrapet et al., 2006). The present study was undertaken to determine whether other mutations near nt 57 in the 59NCR could attenuate wild-type DEN virus. Such mutations could prove useful in designing live-attenuated DEN vaccine viruses. We engineered 11 different mutations, each containing one or more point substitution and/or deletion (Table 1), into the 59NCR of the DEN virus-specific infectious cdna clone pd2/ic-30p-a by using standard in vitro mutagenesis procedures. Deletion or multiple point-substitution mutations in the 59NCR of viable, attenuated mutant viruses should resist reversion to wild-type sequence. We considered the planned double C57U+G58C mutation, with predicted greater destabilization of the stem structure, as a potentially more attenuating and reversion-resistant locus in a viable, clone-derived virus. Several other single point mutations were engineered to begin to examine the uniqueness of the 59NCR-C57U attenuating effect. DEN variants were derived by transfection of LLC-MK 2 cells with viral genomic RNA transcribed from plasmid DNA as described previously (Kinney et al., 1997). RNA transcripts of the single point-mutated plasmids pd2/ic- 55, -57, -60 and -69, as well as the double mutant pd2/ic- 5758, yielded fully infectious viruses, as indicated by the expression of immunofluorescence assay (IFA)-detectable DEN-2 viral antigen in 100 % of cells in the transfected cell culture following 4 7 days incubation (Table 1). In our non-optimized transfection experiments with pd2/ic- 30P-A clone-transcribed RNA, IFA-detectable DEN-2 viral antigen was detected routinely in a minor fraction of cells in the transfected culture during the first 2 days after electroporation (unpublished data), and the extent of IFA positivity has been a reliable indicator of clone-derived viral viability in our laboratory. The pd2/ic-v5 clone and D2/IC-V5 virus reported previously (Butrapet et al., 2000) were utilized in the current study and, for clarity, their designations are changed here to clone pd2/ic-57 and D2/ IC-57 virus, respectively. Rescued viruses were amplified once more in LLC-MK 2 cells, and the complete genomes (except for the 59- and 39-terminal primer target sequences) of the resulting virus seeds were sequenced following cdna amplification by RT-PCR. The single and double deletionmutant plasmids pd2/ic-d57 and pd2/ic-d5758 produced crippled, replication-deficient viruses that resulted in IFAdetectable DEN-2 antigen in approximately 5 and 10 % of transfected cells in the cultures, respectively, in four separate experiments. The remaining five mutations engineered into plasmids pd2/ic-d62/63 (deletion of nt 62, substitution at nt 63), -5760, -5661, -d5759 and -d5661 were Table 1. Mutations engineered into the 59NCR of DEN-2 virus 5 NCR plasmid mutant IFA-detectable DEN-2 antigen (%)* DEN-2 genomic nucleotide positiond pd2/ic-30p-ad 100 C A A C G T A G T T C T A A C A G pd2/ic C A A T G T A G T T C T A A C A G pd2/ic C G A C G T A G T T C T A A C A G pd2/ic C A A C G T G G T T C T A A C A G pd2/ic C A A T C T A G T T C T A A C A G pd2/ic C A A C G T A G T T C T A A C T G pd2/ic C A A T C C G G T T C T A A C A G pd2/ic C A G T C C G A T T C T A A C A G pd2/ic-d62/63 0 C A A C G T A G A C T A A C A G pd2/ic-d57 5 C A A G T A G T T C T A A C A G pd2/ic-d C A A T A G T T C T A A C A G pd2/ic-d C A A A G T T C T A A C A G pd2/ic-d C A T T C T A A C A G *100 %, All cells in the transfected LLC-MK 2 culture expressed DEN-2 antigen; 0 %, DEN-2 antigen was not detected in the transfected cell culture. DUnderlined letters indicate nucleotide substitutions; indicates deletion of the nucleotide at the indicated position. Nucleotides (shown in bold) form a predicted stem structure with nt Plasmid cdna (T), rather than viral RNA (U), nucleotides are shown. dinfectious cdna clone of wild-type DEN virus (Kinney et al., 1997). Described previously as plasmid D2/IC-V5 (Butrapet et al., 2000)

3 W. Sirigulpanit, R. M. Kinney and V. Leardkamolkarn considered lethal, because they failed to yield infectious viruses in five attempts. These results indicated that mutations involving certain 1 or 2 nt substitutions between nt 54 and 70 were well tolerated and resulted in viruses of high infectivity. However, all of the engineered deletion mutations, as well as those substitutions involving more than 2 nt, failed to yield infectious viruses. The results agree with a previous report showing that multiple contiguous deletions between nt 50 and 72 in the 59NCR of DEN-4 virus are fatal for the virus (Cahour et al., 1995). Viable clone-derived viruses were analysed for plaque size and replication in LLC-MK 2 cells and for replication in C6/36 cells, as described previously (Butrapet et al., 2000). At 9 days after infection, the mean plaque diameters (Table 2) of D2/IC-30P-A virus and D2/IC-69 virus were equivalent ( mm) (P.0.05, Student s t-test). D2/ IC-55 and -60 viruses produced significantly smaller plaques with a mean diameter of 2.5 mm (P,0.05), which was similar to the plaque size reported previously for D2/IC-57 virus (Butrapet et al., 2000). The D2/IC-5758 virus had a statistically significantly larger mean plaque size than the D2/ IC-30P-A virus (P,0.05). All of these viruses produced plaques that were much larger than the pinpoint (,1 mm) plaques of the candidate DEN-2 PDK-53 vaccine virus included in the same assays (data not shown). All three of the 59NCR-C57U, NS1-Gly53Asp and NS3-Glu250Val mutations have been shown to contribute independently, and additively, to the reduced plaque size of PDK-53 virus (Butrapet et al., 2000). The D2/IC-d57 and -d5758 viruses failed to produce plaques, even by 2 weeks post-infection (data not shown). Virus growth curves were determined in LLC-MK 2 or C6/ 36 cells infected at an m.o.i. of approximately p.f.u. per cell. Aliquots of culture medium were removed at 48 h intervals for 12 days. Plaque titrations were performed in confluent monolayers of Vero or LLC-MK 2 cells as described previously (Huang et al., 2000; Miller & Mitchell, 1986). Viruses D2/IC-30P-A, D2/IC-55, -57, -60, and -69 replicated well to high peak titres of p.f.u. ml 21 at days after infection of LLC-MK 2 cells (Fig. 1a). D2/IC-60 virus replicated more slowly than the other viruses between days 2 and 8 after infection, but attained a similarly high titre by days The clone-derived viruses showed variable replication in C6/36 cells (Fig. 1b). D2/IC-30P-A, -60, and -69 viruses replicated to peak titres of p.f.u. ml 21 at day 8 after infection, although replication of D2/IC-60 virus appeared to be slightly delayed in these cells as well. D2/IC-57 and -55 viruses exhibited restricted growth that was characterized by 100- and 10-fold decreases in peak titres, respectively, relative to the replication of D2/IC-30P-A virus in C6/36 cells. The peak titres of these two viruses also occurred at day 8 after infection. DEN-2 PDK-53 virus replicated to p.f.u. ml 21 after 8 days growth in C6/36 cells (data not shown). The severely crippled replicative phenotype of the DEN-2 PDK-53 virus in C6/36 cells is caused by the 59NCR-C57U and NS1-Gly53Asp mutations, both of which have been shown to inhibit virus replication independently, as well as additively when both mutations were engineered together into the DEN virus genome (Butrapet et al., 2000). Wild-type DEN virus routinely causes % mortality in newborn ICR mice challenged intracranially (i.c.) with 10 4 p.f.u. virus, whereas the candidate PDK-53 vaccine virus is reproducibly attenuated in this model (Butrapet et al., 2000; Huang et al., 2000, 2003; Kinney et al., 1997). Attenuation of neurovirulence for ICR mice is Table 2. Plaque size and neurovirulence of DEN-2 59NCR mutant viruses Abbreviations: MST, mean survival time; NA, not applicable. Virus Plaque size (mm) (mean±sd)* Mortality (%)D MST±SD (days) Body mass (g) (mean±sd)d Diluent NA 0.00 NA 16.2±1.9 D2/IC-30P-A 3.0± ± ±1.9 D2/IC ± ±4.2 D2/IC ± ± ±1.6 D2/IC ± ± ±1.0 D2/IC ±0.3# ± ±1.3 D2/IC ± ± ±1.9 *Diameter of 20 randomly selected plaques. Dn532 mice per group, except for D2/IC-57 virus (n516). dbody mass of surviving mice 3 weeks after infection. Plaque size and neurovirulence results are from a previous study (Butrapet et al., 2000). Mean body mass of surviving mice was significantly lower (P,0.001, Student s t-test) than that of diluentinoculated control mice 3 weeks after infection (Butrapet et al., 2000). Significantly smaller than mean plaque size of D2/IC-30P-A virus (P,0.05, Student s t-test). #Significantly greater than mean plaque size of D2/IC-30P-A virus (P,0.05, Student s t-test) Journal of General Virology 88

4 Effect of mutations in the 59NCR of DEN-2 virus Fig. 1. Growth curves of D2/IC-30P-A virus (clone-derived, wildtype DEN virus) and its viable clone-derived mutants in LLC-MK 2 (a) and C6/36 (b) cells. Cells were infected at an approximate m.o.i. of p.f.u. per cell. D2/IC-57 is a previously described virus (Butrapet et al., 2000). m, D2/IC-30P-A; $, D2/ IC-57; ¾, D2/IC-55;, D2/IC-60; #, D2/IC-5758; g, D2/IC-69. * caused by the NS1-Gly53Asp and 59NCR-C57U mutations (Butrapet et al., 2000). More recently, the NS3-Glu250Val mutation was also found to contribute to the attenuated phenotype of PDK-53 virus in the more DEN-2 virussensitive Swiss Webster mouse model (Huang et al., 2005). In the present study, the mutant 59NCR viruses were tested for neurovirulence in newborn ICR mice. In two separate experiments, groups of 16 mice were inoculated i.c. with 10 4 p.f.u. clone-derived D2/IC virus and observed for 35 days for morbidity, paralysis and/or death. All mice surviving for 21 days after infection were weighed individually on that day. The pooled results (32 mice per group) from the two experiments are shown in Table 2. Previously published data for D2/IC-57 virus (n516 mice) are shown for comparison (Butrapet et al., 2000). The D2/IC-30P-A and D2/IC-60 viruses resulted in % mortality, with a mean survival time (MST) of 13.6 and 15.8 days, respectively. Because we previously identified only % mortality (MST of 21.7 days) in newborn ICR mice challenged i.c. with D2/IC-57 virus, the 87.5 %, wild-type level of mortality in mice challenged with D2/IC-5758 virus was unexpected (Table 2). Mice succumbing to challenge with this virus had the lowest MST (9.9 days), and the survivors at 21 days exhibited the lowest mean body mass (9.8 g), of all mouse groups tested. Viruses D2/IC-55 and D2/IC-69 showed lower levels of mortality of 50 and % (MSTs of 16.8 and 19.2 days), respectively, compared with D2/IC- 30P-A virus. At 21 days after i.c. infection, the surviving mice in all of the virus-challenged groups exhibited a significantly lower mean body mass than the diluentinoculated control mice (P,0.001, Student s t-test). Even though the ICR mouse model of neurovirulence for DEN virus is fairly insensitive, the D2/IC-69 virus appeared to be partially attenuated for neurovirulence, although it exhibited the robust replication phenotype of the D2/IC-30P-A virus in terms of plaque size and replication in LLC-MK 2 and C6/36 cells. The A69U mutation is located at the leading end of the first stem (nt 4 9/64 69), situated at the base of the predicted large SLA structure (Filomatori et al., 2006). This A69U substitution has been identified in wild-type DEN-2 virus and was reported to constitute a significant genetic marker in differentiating between South-East Asian DEN-2 strains and DEN-2 strains of the American genotype (Leitmeyer et al., 1999). Given the apparent partial attenuation of mouse neurovirulence afforded by the A60U mutation and the robust replication of the D2/IC-69 virus in both mammalian (LLC-MK 2 ) and mosquito (C6/36) cell cultures, it might be fruitful to investigate a combination of this mutation with another attenuating 59NCR mutation, such as the C57U mutation, in a potentially attenuated double mutant that might resist complete genetic reversion in the 59NCR more readily during viral passage. D2/IC-57 virus exhibited 50 % morbidity in ICR mice. Although it is difficult to interpret such a result in the ICR model, it is possible that D2/IC-55 is slightly attenuated for neurovirulence. This virus did exhibit decreased replicative ability in C6/36 cells (although not to the extent of D2/IC-55 virus) and smaller plaque size, relative to D2/IC-30P-A virus. The A55G mutation might be another possible candidate for inclusion in an attenuated virus containing a double mutation in the 59NCR. The D2/IC-60 virus was reproducibly less robust during the first 6 8 days of replication in LLC-MK 2 cells in two independent experiments, but replicated well in C6/36 cells and was fully neurovirulent for mice. As we were unable to derive infectious viruses from the plasmid constructs containing one or more deletion mutation or more than two nucleotide substitutions, a defect in viral protein synthesis that might have prevented virus replication was suspected. The capacity of the RNA mutants to direct protein synthesis was examined in the rabbit reticulocyte lysate system in the presence of [ 35 S]Met. The translation products were analysed by SDS-PAGE (10 % gels) and translation efficiencies were determined by measuring incorporation of acid-precipitated [ 35 S]Met. All

5 W. Sirigulpanit, R. M. Kinney and V. Leardkamolkarn of the RNA mutants were translated efficiently to variable extents, with [ 35 S]Met incorporation ranging from to 2.13-fold relative to the pd2/ic-30p-a RNA transcript (data not shown). The results obtained failed to show an obvious impact of mutation on translation efficiency. Nevertheless, this finding is in agreement with a recent study demonstrating that the 59NCR SLA structure functions in minus-strand RNA synthesis, rather than RNA translation (Filomatori et al., 2006). Interestingly, the D2/IC-5758 mutant, which was expected to be less neurovirulent in mice than the D2/IC-V5 mutant, yielded unexpected results. RNA secondary-structure prediction of the 59NCR of each mutant DEN-2 virus was performed with the Mfold program (Zuker, 2003). The predicted stem structure formed by nt and was more perturbed in the D2/IC-5758 mutant than in the D2/IC-57 virus, as expected by the incorporation of the G58C mutation in addition to the partially attenuating 59NCR-C57U mutation in this stem structure (data not shown). Rather than exhibiting more pronounced attenuation markers, the D2/IC-5758 mutant replicated as well as the wild-type virus in both LLC-MK 2 and C6/36 cell cultures, exhibited a large plaque size and appeared to be more neurovirulent for newborn ICR mice than the wild-type (D2/IC-30P-A) virus. Thus, despite its predicted additional disruption of the RNA stem structure, the engineered contiguous secondary G58C mutation caused reversion of the partially attenuated phenotype caused by the 59NCR-C57U mutation. This observation is particularly relevant for candidate live-attenuated, chimeric DEN virus and other flavivirus vaccines based on the genetic background of the attenuated DEN-2 PDK-53 virus. Genetic analyses of such vaccine virus seeds would be able to verify the apparent absence of potentially compensatory secondary mutations in the 59NCR, as well as a low level or absence of reversion at the 59NCR-57U locus. Acknowledgements This study was supported by the Royal Golden Jubilee PhD Programme of Thailand Research Fund (RGJ/PHD/2541) and Thailand- Tropical Diseases Research Programme (T2), National Center for Genetic Engineering and Biotechnology. The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention. References Ali, N. & Siddiqui, A. (1995). Interaction of polypyrimidine tractbinding protein with the 59 noncoding region of the hepatitis C virus RNA genome and its functional requirement in internal initiation of translation. J Virol 69, Alvarez, D. E., Lodeiro, M. F., Luduena, S. J., Pietrasanta, L. I. & Gamarnik, A. V. (2005). Long-range RNA-RNA interactions circularize the dengue virus genome. J Virol 79, Brinton, M. A. & Dispoto, J. H. (1988). Sequence and secondary structure analysis of the 59-terminal region of flavivirus genome RNA. Virology 162, Butrapet, S., Huang, C. Y., Pierro, D. J., Bhamarapravati, N., Gubler, D. J. & Kinney, R. M. (2000). Attenuation markers of a candidate dengue type 2 vaccine virus, strain (PDK-53), are defined by mutations in the 59 noncoding region and nonstructural proteins 1 and 3. J Virol 74, Butrapet, S., Kinney, R. M. & Huang, C. Y. (2006). Determining genetic stabilities of chimeric dengue vaccine candidates based on dengue 2 PDK-53 virus by sequencing and quantitative TaqMAMA. J Virol Methods 131, 1 9. Cahour, A., Pletnev, A., Vazielle-Falcoz, M., Rosen, L. & Lai, C. J. (1995). Growth-restricted dengue virus mutants containing deletions in the 59 noncoding region of the RNA genome. Virology 207, Filomatori, C. V., Lodeiro, M. F., Alvarez, D. E., Samsa, M. M., Pietrasanta, L. & Gamarnik, A. V. (2006). A59RNA element promotes dengue virus RNA synthesis on a circular genome. Genes Dev 20, Gubler, D. J. (1998). Resurgent vector-borne diseases as a global health problem. Emerg Infect Dis 4, Guesdon, F. M., Greatorex, J., Rhee, S. R., Fisher, R., Hunter, E. & Lever, A. M. (2001). Sequences in the 59 leader of Mason-Pfizer monkey virus which affect viral particle production and genomic RNA packaging: development of MPMV packaging cell lines. Virology 288, Huang, C. Y., Butrapet, S., Pierro, D. J., Chang, G. J., Hunt, A. R., Bhamarapravati, N., Gubler, D. J. & Kinney, R. M. (2000). Chimeric dengue type 2 (vaccine strain PDK-53)/dengue type 1 virus as a potential candidate dengue type 1 virus vaccine. J Virol 74, Huang, C. Y., Butrapet, S., Tsuchiya, K. R., Bhamarapravati, N., Gubler, D. J. & Kinney, R. M. (2003). Dengue 2 PDK-53 virus as a chimeric carrier for tetravalent dengue vaccine development. J Virol 77, Huang, C. Y., Silengo, S. J., Whiteman, M. C. & Kinney, R. M. (2005). Chimeric dengue 2 PDK-53/West Nile NY99 viruses retain the phenotypic attenuation markers of the candidate PDK-53 vaccine virus and protect mice against lethal challenge with West Nile virus. J Virol 79, Kinney, R. M., Butrapet, S., Chang, G. J., Tsuchiya, K. R., Roehrig, J. T., Bhamarapravati, N. & Gubler, D. J. (1997). Construction of infectious cdna clones for dengue 2 virus: strain and its attenuated vaccine derivative, strain PDK-53. Virology 230, Kuhn, R. J., Zhang, W., Rossmann, M. G., Pletnev, S. V., Corver, J., Lenches, E., Jones, C. T., Mukhopadhyay, S., Chipman, P. R. & other authors (2002). Structure of dengue virus: implications for flavivirus organization, maturation, and fusion. Cell 108, Leitmeyer, K. C., Vaughn, D. W., Watts, D. M., Salas, R., Villalobos, I., de Chacon, Ramos, C. & Rico-Hesse, R. (1999). Dengue virus structural differences that correlate with pathogenesis. J Virol 73, Lindenbach, B. D. & Rice, C. M. (2003). Molecular biology of flaviviruses. Adv Virus Res 59, Miller, B. R. & Mitchell, C. J. (1986). Passage of yellow fever virus: its effect on infection and transmission rates in Aedes aegypti. Am J Trop Med Hyg 35, You, S., Falgout, B., Markoff, L. & Padmanabhan, R. (2001). In vitro RNA synthesis from exogenous dengue viral RNA templates requires long range interactions between 59- and 39-terminal regions that influence RNA structure. J Biol Chem 276, Zuker, M. (2003). Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res 31, Journal of General Virology 88