Replication of the Primary Dog Kidney-53 Dengue 2 Virus Vaccine Candidate in Aedes aegypti Is Modulated by a Mutation in the 5′ Untranslated Region and Amino Acid Substitutions in Nonstructural Proteins 1 and 3

Abstract

Previous studies have demonstrated reduced replication of the cell culture-adapted Dengue-2 virus (DENV-2) vaccine candidate, primary dog kidney (PDK)-53, compared with the parental DENV-2 strain, 16681, in C6/36 cells. Various DENV-2 mutants incorporating PDK-53 substitutions singly and in combination into the 16681 genetic backbone were used to identify the genetic basis for impaired replication of the vaccine candidate in vitro in Aedes aegypti cell culture (Aag2 cells) as well as the reduced in vivo infectivity and transmissibility within Ae. aegypti infected by intrathoracic inoculation. 5′ untranslated region (UTR-c57t) and nonstructural protein 1 (NS1-G53D) mutations were required to completely attenuate in vitro replication. In contrast, incorporation of the PDK-53-specific NS3-250V mutation into the 16681 virus resulted in reduced replication in mosquitoes but had no effect on in vitro replication. Further, reversion of the PDK-53 NS3-250 site to that of the wild-type 16681 virus (NS3-V250E) failed to increase either in vitro or in vivo replication. Intrathoracic inoculation of Ae. aegypti with mutants containing the PDK-53 NS1 substitution exhibited in vivo replication indistinguishable from the parental PDK-53 virus, implicating this mutation as the dominant determinant for impaired mosquito replication of the PDK-53 candidate; however, further attenuation of in vivo replication was magnified in mutants including the additional 5′UTR-c57t mutation.

Introduction

D engue viruses consist of four distinctive antigenic serotypes (DENV-1, DENV-2, DENV-3, and DENV-4) within the family Flaviviridae, all four of which are transmitted among humans by Aedes aegypti and Aedes albopictus mosquitoes. These viruses cause large-scale human epidemics of febrile disease that are often associated with cases of dengue hemorrhagic fever (DHF) and dengue shock syndrome (Gubler et al. 1979, Gubler 1988, 1997, 2002). Epidemiologically, DHF/dengue shock syndrome has been associated with sequential infection with heterologous serotypes of DENV (Gubler 1988, Gubler and Trent 1994). Vaccination of humans has been one method proposed for intervention because there are no alternate hosts in the urban epidemic transmission cycle of this virus. As infection with a single DENV serotype could increase the risk of DHF, it is important to assess the potential for live-attenuated candidate DENV vaccines to be transmitted by mosquitoes.

The DENV-2 primary dog kidney (PDK)-53 vaccine candidate was generated by 53 serial passages of the parental Thai DENV-2 strain 16681 in PDK cells (Yoksan et al. 1986). The PDK-53 vaccine candidate virus has been shown to be well tolerated and elicit long-term humoral and cellular immunity in early clinical trials (Bhamarapravati et al. 1987, Dharakul et al. 1994, Vaughn et al. 1996). The phenotypic markers associated with DENV-2 PDK-53 attenuation include reduced plaque size and temperature sensitivity in mammalian cells, reduced replication capacity in C6/36 mosquito (Ae. albopictus) cells, and attenuation of neurovirulence for newborn ICR mice, relative to the wild-type, parental DENV-2 16681 strain (Butrapet et al. 2000). Recently, live-attenuated, chimeric candidate dengue and West Nile virus (WNV) vaccine viruses have been engineered that express the prM/E gene regions of the heterologous wild-type viruses (DENV-1, DENV-3, DENV-4, or WNV) in the attenuated genetic background of the DENV-2 PDK-53 virus (Huang et al. 2000, 2005).

One mutation within the 5′ untranslated region (5′UTR-c57t), five coding amino acid differences (prM-D29V, NS1-G53D, NS2A-L181F, NS3-E250V, and NS4A-G75A) and three silent mutations are present in the PDK-53 vaccine candidate compared with the parental 16681 strain (Kinney et al. 1997). The dominant determinants of the PDK-53 phenotypic markers of attenuation previously were mapped to the NS1-53, 5′UTR-57, and NS3-250 loci, in order of predominance of effect (Butrapet et al. 2000). The reduced replication phenotype in C6/36 cells previously has been demonstrated to be encoded additively by the 5′UTR-57 and NS1-53 mutations (Butrapet et al. 2000). A previous study has also demonstrated the PDK-53 vaccine virus to poorly infect Ae. aegypti after oral exposure of high viral titers with no subsequent dissemination (Khin et al. 1994). To further characterize the potential replication of the PDK-53 vaccine candidate in the principal urban vector mosquito as well as to identify potential attenuation determinants for in vivo mosquito replication, we characterized the replication phenotypes of DENV-2 16681, PDK-53, and recombinant viruses containing single or combinations of point mutations in the genetic backbone of either 16681 or PDK-53 viruses, respectively, within Aag2 mosquito (Ae. aegypti) cells and in intrathoracically (IT) inoculated Ae. aegypti mosquitoes from Thailand. Transmission potential was also assessed 7 days after IT inoculation in Ae. aegypti using an in vitro technique (Aitken 1977). No evidence for the transmission of the vaccine strain was detected. Similarly, parental 16681 DENV-2 variants containing the NS1 mutation replicated poorly in vivo and resulted in undetectable virus levels in mosquito saliva. These results indicate that nonstructural genetic elements modulate the reduced mosquito competence phenotype. Since the vaccine-associated viremia generated in human vaccinees has been experimentally observed to be of low magnitude (Vaughn et al. 1996), it is unlikely that Ae. aegypti would become infected or transmit live-attenuated PDK-53-based vaccines containing dengue or heterologous flaviviral structural (prM/E) genes.

Materials and Methods

Mosquitoes

Ae. aegypti mosquitoes were kindly provided by Dr. Laura Harrington of Cornell University. Mosquitoes used for these studies were from a colony established in 2002 from a village near Mae Sot (16′ N, 33′ E), Thailand, and had been in colonization for ∼30 generations. Genetic diversity was maintained by annual addition of field-collected eggs to the colony. Eggs were hatched in 1 L of distilled water, and developing larvae were fed a finely ground mixture of rabbit chow and fish food. Pupae were allowed to emerge and adults to mate in 1.7 L carton cages, where they were offered 10% sucrose. Females were utilized for IT inoculation at 3–4 days postemergence.

Viruses/mosquito infection

Table 1 delineates the genetic differences between the parental DENV-2 16681 and recombinant DENV-2 containing single or combinations of mutations in the 16681 (V513, V5, V1, V51, and V3) or PDK-53 (P3) genetic backbones utilized for this study. The DENV-2 16681 virus and derivation of its pD2/IC-30P-A infectious clone have been previously reported (Kinney et al. 1997). All other viruses were derived from infectious cDNA clones of DENV-2 16681 and PDK-53 viruses, as previously described (Butrapet et al. 2000). All viruses were diluted to 10⁶ plaque-forming units (PFU)/mL, as determined by plaque formation on Vero cells, and mosquitoes were inoculated with ∼100 PFU of each virus in a 0.1-μL inoculum.

Table 1.

Nucleotide and Amino Acid Residues Present in the Wild-Type Dengue-2 16681 Virus and Recombinant Dengue Viruses Assessed for Replication in Aedes aegypti (Thailand)

	Genetic backbone	5′UTR-57	prM-29	NS1-53	NS2A-181	NS3-250	NS4A-75
16681^a	NA	c	D	G	L	E	G
VV45R^b	PDK-53-parental	t	V	D	F	V	A
30P-A^c	16681-parental	—^d	—	—	—	—	—
V5	16681	t	—	—	—	—	—
V1	16681	—	—	D	—	—	—
V3	16681	—	—	—	—	V	—
V51	16681	t	—	D	—	—	—
V513^e	16681	t	—	D	—	V	—
P3^f	PDK-53	t	V	D	F	—	A

Wild-type DENV-2 16681 (parent to the candidate PDK-53 vaccine virus), genetic residues shown in bold font.

D2/IC-VV45R virus derived from the infectious clone of one of the two variants present in the PDK-53 vaccine. The VV45R variant contains NS3-250V, whereas the other variant retains the NS3-250E of the wild-type 16681 virus (Huang et al. 2000).

Derived from the pD2/IC-30P-A infectious clone of DENV-2 16681 (Kinney et al. 1997).

Residue is that of the wild-type DENV-2 16681.

V513 contains the 5′ NC-57t, NS1-53D, and NS3-250V of the PDK-53 vaccine virus in the genetic background of the wild-type DENV-2 16681 (Kinney et al. 1997).

P3 contains the NS3-250E residue of wild-type 16681 virus in the PDK-53 virus-specific VV45R genetic background (Kinney et al. 1997).

A, alanine; c, cytosine; D, aspartic acid; DENV-2, dengue-2 virus; E, glutamic acid; F, phenylalanine; G, glycine; L, leucine; t, thymine; V, valine.

Aag2 growth curves

Aag2 cells were kindly provided by Drs. Carol Blair and Katie Poole of Colorado State University. Cells were grown to confluent monolayers in six-well plates (Costar, Bethesda, MD), and triplicate cultures were inoculated at a multiplicity of infection of 0.01 at 28°C for 1 h, at which point the cells were washed with phosphate-buffered saline twice and 3 mL of fresh Schneider's Drosophila media supplemented with 2% fetal bovine serum (FBS), 1% P/S, 1% L-glutamine, and 1% nonessential amino acids was added. Based on previous findings that demonstrated extremely slow replication of the PDK-53 virus in C6/36 cells (Huang et al. 2000), Aag2 culture media was sampled [50 μL of infectious culture medium placed in 450 μL minimum essential media (MEM) supplemented with 10% FBS] at 4, 8, and 12 days postinfection (dpi). Samples were stored at −80°C until titrated by plaque assay to estimate infectious units.

Transmission assessment

After 7 days of extrinsic incubation at 28°C, IT-inoculated mosquitoes were anesthetized with triethylamine and their proboscis was inserted into a capillary tube containing FBS (Aitken 1977, Kramer et al. 1990). Mosquitoes were allowed to expectorate for 30 min at which point the bodies were placed in an individual cryovial and frozen at −80°C until titrated as described below for infectious virus. Capillary tubes were placed in a 1.5 mL vial containing 0.25 mL of MEM supplemented with 20% FBS and centrifuged to expel all fluid from the capillary tube. Expectorants were frozen at −80°C until titrated for infectious virus. One hundred microliters of the expectorant was used for a plaque titration resulting in a detection limit of 2.5 PFU.

Viral quantification

0.5 mL of MEM supplemented with 5% FBS, 50 mg/mL penicillin/streptomycin, and 0.1% Amphotericin B was added to each tube containing a single IT-inoculated female mosquito. Mosquitoes were triturated in a 1.5 mL cryovial using a sterile pestle until no visible mosquitoes tissues could be identified. Vials containing triturated mosquito were centrifuged at 3000 g for 5 min to pellet homogenized mosquito tissues. Tenfold dilutions of the supernatants from mosquito homogenates, as well as Aag2 time point harvests, were titrated on Vero cells as previously described (Miller and Mitchell 1986, Butrapet et al. 2000, Huang et al. 2000). Titers were expressed as PFU/mosquito body or PFU/mL of culture medium, respectively.

Statistical analyses

Body titers from IT-inoculated mosquitoes among viruses were compared by one-way analysis of variance (ANOVA) (Hintze 1998). Titers among replicated cell cultures were compared using a repeated measures ANOVA. Because there was a highly significant interaction effect among strains that did or did not replicate well in cell culture over time (F = 70.4, df = 16, 35, p < 0.0001), final strain comparisons focused on titers measured at 12 dpi using a one-way ANOVA, with multiple pairwise differences among means performed using the Tukey-Kramer test.

Results

Aag2 replication

When compared by one-way ANOVA followed by a Tukey-Kramer multiple comparison test, the nine viral strains grouped into three significantly (p < 0.05) different growth patterns by day 12 (Fig. 1). The parental 16681 DENV-2 replicated well in Aag2 cells, with a mean peak titer of 7.3 log₁₀ PFU/mL by 12 dpi. The DENV-2 16681 infectious clone-derived 30P-A virus, as well as the clone-derived V3 virus containing the PDK-53-specific NS3-250V substitution in the 16681 background, demonstrated efficient replication profiles that were indistinguishable from that of the wild-type 16681 virus by day 12 (p > 0.05). In contrast, the PDK-53-derived VV45R strain demonstrated a steady decrease in titer in Aag2 cells between 4 and 12 dpi following its peak levels observed at 3.3 log₁₀ PFU/mL at 4 dpi (Fig. 1).

FIG. 1.

Replication of dengue-2 virus in Aedes aegypti (Aag2) cells. Cells were infected in triplicate at a multiplicity of infection of 0.01, and infectious units were titrated as plaque-forming units (PFU)/mL of culture medium by plaque assay from sampling on 4, 8, and 12 days postinfection (dpi) (no detectable virus [ < 1.7 log₁₀ PFU/mL] was identified at 0 dpi). Bars represent standard deviations from the mean. The detection limit of 1.7 log₁₀ PFU/mL resulted from the initial 1:10 dilution of culture medium. Data points have been slightly offset on the x-axis for observation purposes.

The V51 (PDK-53-specific 5′UTR-57t and NS1-53D loci in the 16681 background), P3 (16681-specific NS3-250E in the PDK-53 background), and V513 viruses exhibited a retarded replication phenotype in Aag2 cells that was not significantly different from VV45R virus on 12 dpi when all nine means were compared using a Tukey-Kramer multiple comparison test (p > 0.05). However, when viruses within this group were retested by ANOVA and the means again compared by the Tukey-Kramer test, the mean V513 virus titer on day 12 was significantly (p < 0.05) greater than the means observed for the VV45R, V51, and P3 viruses; these latter three viruses were not significantly (p > 0.05) different from one another (Fig. 1). The somewhat greater replication efficiency of the V513 virus indicated the possibility that, in combination with the 5′UTR-57t and NS1-53D substitutions in the 16681 background, the NS3-250V might have had a subtle resuscitating effect in Aag2 cells. Contradicting this was the finding that incorporation of the NS3-250V mutation into the 16681 background (V3 virus) did not adversely affect the efficient replication phenotype of the wild-type virus, nor did the reciprocal NS3-250E mutation in the PDK-53 background (P3 virus) impart increased replication efficiency to the vaccine-specific VV45R virus (Fig. 1). The V5 and V1 viruses, containing either the single 5′UTR-57t or single NS1-53D of PDK-53 virus in the 16681 background, both demonstrated intermediate growth phenotypes in Aag2 cells, replicating to peak titers of 5.0 and 4.5 log₁₀ PFU/mL, respectively, by 12 dpi (Fig. 1) that were not significantly different (p > 0.05).

Mosquito replication

All parental and recombinant DENV-2 viruses replicated in Ae. aegypti from Thailand after IT inoculation (Table 1). However, virus was only detected in 35% and 40% of the mosquitoes inoculated with the mutants containing the PDK-53 NS1-53D mutation and 5′UTR-57t/NS1-53D mutations, respectively (Table 2). The parental, wild-type 16681 virus and its clone-derived 30P-A virus replicated the most efficiently, reaching mean titers of 4.3 and 4.4 log₁₀ PFU/mosquito body, respectively (Table 2), significantly greater (p < 0.05) than the clone-derived VV45R virus containing the 5′UTR-57t and all five amino acid mutations present in the attenuated PDK-53 virus (Tables 1 and 2). The VV45R virus replicated to only 2.1 log₁₀ PFU/mosquito body (Table 2), about 100-fold lower than the wild-type virus, in agreement with previous findings by oral feeding (Khin et al. 1994).

Table 2.

Replication Titers/Phenotypes of Dengue-2 Virus Constructs in Aedes aegypti (Thailand) Measured by Plaque Assay on Vero Cells

Virus	VV45R mutations	Tested	Positive	Percent positive	Body titer mean	Standard deviation	Statistical comparison ^a
16681	NA	22	22	100	4.3	0.3	a
VV45R	5′UTR/prM/NS1/NS2A/NS3/NS4A	29	29	100	2.1	0.6	c,d
30P-A	NA	38	38	100	4.4	0.3	a
V5	5′UTR	21	21	100	3.6	0.4
V1	NS1	46	16	35	2.4	0.8	b,c
V3	NS3	14	14	100	2.8	0.4	b
V51	5′UTR/NS1	53	21	40	1.6	0.6	e
V513	NS5/NS1/NS3	18	18	100	1.7	0.5	d,e
P3	5′UTR/prM/NS1/NS2A/NS4A	30	30	100	1.9	0.4	d,e

Means for virus-positive mosquitoes followed by the same letter were not significantly different using a Tukey-Kramer multiple range test (p > 0.05).

The V1 virus, containing the single PDK-53-specific NS1-53D substitution in the 30P-A (16681) backbone, exhibited a mean Ae. aegypti titer of 2.4 log₁₀ PFU/mosquito body. This titer was indistinguishable from that of the VV45R virus. Further, the other two viruses that contained the PDK-53-specific NS1-53D substitution as well as the 5′UTR-57t substitution in the 16681 background, V51 and V513, also exhibited the crippled replication phenotype of the PDK-53 virus in Ae. aegypti. The single NS3-E250V substitution incorporated into the V3 virus also was associated with reduced replication in Ae. aegypti. This virus had a mean titer of 2.8 log₁₀ PFU/mosquito body that was intermediate between the titers produced by the 30P-A and VV45R viruses. Incorporation of the single parental 16681 NS3-V250E substitution into the VV45R backbone (P3) produced a virus that replicated to only 1.9 log₁₀ PFU/mosquito, failing to resuscitate the crippled mosquito replication phenotype of the vaccine strain. Incorporation of the single 5′UTR-57t mutation into the 16681 background generated the V5 virus that replicated to a mean titer of 3.6 log₁₀ PFU/mosquito body that was statistically unique from all viruses assayed (Table 2). There was evidence of a cumulative effect of PDK-53 mutations on the reduced mosquito replication phenotype. A statistically significant difference was identified between titers of viruses that contained the NS1-G53D only (V1) and viruses that additionally contained either the 5′UTR mutation (V51) or 5′UTR and the NS3 mutations (V513).

Transmission capacity

Saliva samples obtained from 15 IT-inoculated mosquitoes per group were tested for the presence of infectious virus by plaque titration in Vero cells with an associated detection limit of 2.5 PFU per expectoration. Virus was detected in the saliva of 60% of the 30P-A virus-inoculated group (mean 1.3 ± 0.7 log₁₀ PFU/expectoration), indicating the potential of susceptible mosquitoes infected with the wild-type virus to transmit the virus to a vertebrate host (Table 3). Previous studies have demonstrated a positive correlation between body titer and the capacity for transmission of dengue (Hanley et al. 2008), St. Louis encephalitis virus (Mahmood et al. 2004), and WNV (Reisen et al. 2006); however, there was no significant difference observed between the body titers of mosquitoes that successfully transmitted 30P-A and those for which virus was not detected in the expectorants (4.3 vs. 4.4 log₁₀ PFU/body, respectively; p > 0.05, t-test). Despite the fact that only mosquitoes with bodies positive for DENV were utilized for saliva sampling, none of the VV45R, V1, and V51 viruses containing the PDK-53-specific NS1-G53D substitution had detectable virus in their saliva (Table 3). This failure of the mutants to transmit precluded any comparisons to be made between body titers of mosquitoes infected with these viruses and transmission capacity.

Table 3.

Transmission Capacity of Intrathoracically Inoculated Aedes aegypti (Thailand)

Virus	Transmission ^a	Percent
VV45R	0/15	0
30P-A	9/15	60
V1	0/15	0
V51	0/15	0

Transmission was assessed by plaque assay on Vero cells (detection limit 2.5 plaque-forming units) of saliva obtained from female Aedes aegypti that had been intrathoracically inoculated 7 days before with the corresponding parental or recombinant DENV-2 and determined to have detectable virus from their bodies.

Discussion

Live-attenuated arboviral vaccines must meet criteria for safety assuring that no transmission of the vaccine virus occurs after arthropod blood feeding on vaccinees. This study provides evidence that a sympatric strain of the primary mosquito vector of DENV-2 in Thailand is incapable of transmitting a DENV-2 16681 variant containing a single amino acid substitution. The clone-derived PDK-53 virus-specific VV45R virus, as well as viruses derived from the wild-type 16681 virus containing the NS1-G53D substitution, failed to produce detectable virus titers in the saliva of IT-inoculated Ae. aegypti mosquitoes. This evident transmission defect was likely due to reduced replication of these viruses in the mosquito vector, as demonstrated by low-level viral replication in the body of IT-inoculated mosquitoes, relative to that of the wild-type 16681 virus and its clone-derived 30P-A virus. The finding that mutations within the 5′UTR, as well as in NS3, also had an inhibiting effect on replication in mosquitoes indicated the presence of additional, redundant viral genetic restrictions for mosquito vector competence of the candidate PDK-53 vaccine virus.

This study highlights the importance of UTR nucleotide regions and nonstructural proteins for the modulation of vector competence of arboviruses and supports targeting these genomic regions for the development of live-attenuated vaccines that are incapable of being transmitted by arthropod vectors. The preponderance of reports identifying vector infection, dissemination, and transmission determinants for arboviruses have focused on the structural genetic elements associated with receptor binding and viral entry of arthropod cells (Brault et al. 2002, 2004, McElroy et al. 2006b). Some data have been established to support the role of nonstructural elements for the modulation of vector competence. For example, chimeric yellow fever viruses generated between wild-type virus and the 17D vaccine strain demonstrated that genetic elements of the NS2A and NS4B, in addition to structural elements, were responsible for reduced dissemination of the vaccine strain in Ae. aegypti (McElroy et al. 2006a). An additional study comparably indicated that a single NS4B amino acid substitution could significantly impair the ability of DENV-4 virus to replicate in mosquito cells and infect adult mosquitoes while resulting in increased fitness in vertebrate cells (Hanley et al. 2003). A DENV-4 construct with a 30-nucleotide deletion within the 3′UTR had lower replicative capacity in C6/36 cells as well as lowered dissemination rates from the midguts of orally infected Ae. aegypti mosquitoes (Troyer et al. 2001). Similarly, a DENV-3 vaccine candidate containing a 30-nucleotide deletion in the 3′UTR has demonstrated incompetence for replicating in C6/36 mosquito cells and retarded replication in IT inoculated Toxorhynchites mosquitoes (Blaney et al. 2008). Chimeric viruses containing the structural proteins of WNV in a DENV-4 3′UTR deletion backbone did not demonstrate efficient infection, dissemination, or transmission by Culex tarsalis (Hanley et al. 2005), indicating that vector restrictions could be modulated by nonstructural proteins. However, in many of the studies cited above, the effects of chimerization with diverse flaviviruses could not be excluded as a contributing factor to restricted replication. Finally, incorporation of the prM/E structural elements from Modoc virus, a flavivirus with no known arthropod vector that is incompetent for replicating in mosquito cells, into the genetic backbone of dengue and yellow fever viruses failed to demonstrate a restricted replication phenotype in C6/36 cells, indicating the importance of genetic elements exclusive of the prM/E structural proteins for modulating arthropod replicative capacity (Charlier et al. 2010).

Testing of DENV-2 in Aag2 cells demonstrated the particular importance of both the PDK-53 virus-specific 5′UTR-c57t and the NS1-G53D mutations for the decreased replication phenotype in vitro. These results were very similar to previous studies indicating that the 5′UTR-c57t and the NS1-G53D mutations were the dominant determinants of crippled replication in Ae. albopictus (C6/36) cells (Butrapet et al. 2000). A similar effect of the V51 virus was observed in vivo in Ae. aegypti after IT inoculation. As such, the NS1 mutation has been identified as the dominant contributor to the PDK-53-retarded viral replication phenotype in both mosquitoes and mosquito cell cultures. NS1 is a secreted (in vertebrate cells), glycosylated protein of flaviviruses that forms homodimers and is involved in the RNA replication complex (Winkler et al. 1988). NS1 has been associated with inhibiting TLR3-mediated IFN activation after WNV infection in one report (Wilson et al. 2008); however, alternative studies have failed to demonstrate such a downregulation associated with WNV infection or with other mosquito-borne flaviviruses (Baronti et al. 2010). Previous experiments have not directly assessed the effects of amino acid substitutions within NS1 on in vivo mosquito infectivity, dissemination, or transmission; however, removal of one of the two glycosylation motifs within the NS1 protein of DENV-2 has been shown to restrict replication in C6/36 cells (Crabtree et al. 2005). The direct evidence provided in this study for the role of the NS1 in flaviviral replication kinetics in mosquitoes is intriguing. Additional studies have demonstrated the reduced stability of DENV-2 16681 viral mutants containing nucleotide substitutions or deletions between nucleotides 54–70 in the 5′UTR (Sirigulpanit et al. 2007). The 5′UTR-c57t modulating mutation encodes a potential upstream start codon (AUG) that could result in reduced translational efficiency (Chiu et al. 2005) of the DENV-2 polyprotein for both mammalian and mosquito cells.

Given the present data, it appears that all three major determinants of the phenotypic markers of attenuation (Butrapet et al. 2000) adversely affect replication in the intact mosquito. In this context, the finding that the PDK-53 virus-specific 5′UTR, NS1, and NS3 mutations all contributed significantly to the poor replication phenotype of this candidate vaccine virus in intact Ae. aegypti mosquitoes could make it unlikely that reverting mutations would render a DENV-2 PDK-53-based vaccine capable of efficiently infecting and being transmitted by Aedes mosquitoes. The three dominant 5′UTR, NS1, and NS3 loci affecting the attenuated phenotypic markers of the DENV-2 PDK-53 are located outside of the prM/E structural gene region. This has permitted the generation of live-attenuated chimeric viruses expressing the C/prM/E or prM/E structural gene region of the heterologous, wild-type DENV-1, DENV-3, DENV-4, and WNV in the genetic background of the PDK-53 virus (Huang et al. 2000, 2003, 2005). These chimeric viruses retained the PDK-53-characteristic phenotypic markers of small plaque size, temperature sensitivity, reduced replication in C6/36 cells, and attenuation of neurovirulence for mice. Further, by comparison with chimeric viruses engineered in the parental genetic background of the wild-type DENV-2 16681, it was shown that the PDK-53 genotype was the predominant factor controlling the attenuation of these viruses as opposed to impairment through chimerization (Huang et al. 2003, 2005). It is likely that these chimeric viruses will exhibit reduced capacity for replicating in intact mosquitoes. Further studies investigating the relative mosquito infectivity of chimeric DENV-2 PDK-53 viruses expressing heterologous DENV-1, DENV-3, and DENV-4 prM/E genes will be required to further assess the potential of sequences in the heterologous structural proteins to resuscitate a more competent replication phenotype in mosquitoes. The high degree of variability in vector competence of different mosquito populations for dengue (Bennett et al. 2002) as well as other flaviviruses, such as WNV and St. Louis encephalitis virus (Reisen et al. 2008), indicate the need for more extensive assessment of vector competency of the PDK vaccine candidates in Ae. aegypti populations from Asia as well as the Americas.

Footnotes

Acknowledgments

We thank Vincent Martinez and Brian Carroll, Center for Vector borne Diseases for their excellent technical service. The Ae. aegypti mosquitoes from Thailand used for these studies were kindly provided by Dr. Laura Harrington of Cornell University. We also thank Drs. Tom Scott and Rajeev Vaidyanathan for access to and assistance with insectary use, and Brian Carroll for mosquito propagation.

Disclosure Statement

This work was partially funded by a research contract to the University of California, Davis from the Centers for Disease Control and Prevention. Claire Huang and Richard Kinney are among the inventors, and recipients of an awarded patent, of candidate live-attenuated, chimeric dengue vaccine viruses that are based on the attenuated genetic background of the DENV-2 PDK-53 strain. CDC has licensed these candidate vaccine viruses to Inviragen Inc. for commercial manufacture and clinical trial.

References

Aitken

THG

. An in vitro feeding technique for artificially demonstrating virus transmission by mosquitoes. Mosq News, 1977; 37:130–133.

Baronti

, Sire

, de Lamballerie

, Querat

. Nonstructural NS1 proteins of several mosquito-borne Flavivirus do not inhibit TLR3 signaling. Virology, 2010; 404:319–330.

Bennett

, Olson

, Munoz Mde

, Fernandez-Salas

et al. Variation in vector competence for dengue 2 virus among 24 collections of Aedes aegypti from Mexico and the United States. Am J Trop Med Hyg, 2002; 67:85–92.

Bhamarapravati

, Yoksan

, Chayaniyayothin

, Angsubphakorn

et al. Immunization with a live attenuated dengue-2-virus candidate vaccine (16681-PDK 53): clinical, immunological and biological responses in adult volunteers. Bull World Health Organ, 1987; 65:189–195.

Blaney

Jr. , Sathe

, Goddard

, Hanson

et al. Dengue virus type 3 vaccine candidates generated by introduction of deletions in the 3′ untranslated region (3′-UTR) or by exchange of the DENV-3 3′-UTR with that of DENV-4. Vaccine, 2008; 26:817–828.

Brault

, Powers

, Ortiz

, Estrada-Franco

et al. Venezuelan equine encephalitis emergence: enhanced vector infection from a single amino acid substitution in the envelope glycoprotein. Proc Natl Acad Sci U S A, 2004; 101:11344–11349.

Brault

, Powers

, Weaver

. Vector infection determinants of Venezuelan equine encephalitis virus reside within the E2 envelope glycoprotein. J Virol, 2002; 76:6387–6392.

Butrapet

, Huang

, Pierro

, Bhamarapravati

et al. Attenuation markers of a candidate dengue type 2 vaccine virus, strain 16681 (PDK-53), are defined by mutations in the 5′ noncoding region and nonstructural proteins 1 and 3. J Virol, 2000; 74:3011–3019.

Charlier

, Davidson

, Dallmeier

, Molenkamp

et al. Replication of not-known-vector flaviviruses in mosquito cells is restricted by intracellular host factors rather than by the viral envelope proteins. J Gen Virol, 2010; 91:1693–1697.

10.

Chiu

, Kinney

, Dreher

. Control of translation by the 5′- and 3′-terminal regions of the dengue virus genome. J Virol, 2005; 79:8303–8315.

11.

Crabtree

, Kinney

, Miller

. Deglycosylation of the NS1 protein of dengue 2 virus, strain 16681: construction and characterization of mutant viruses. Arch Virol, 2005; 150:771–786.

12.

Dharakul

, Kurane

, Bhamarapravati

, Yoksan

et al. Dengue virus-specific memory T cell responses in human volunteers receiving a live attenuated dengue virus type 2 candidate vaccine. J Infect Dis, 1994; 170:27–33.

13.

Gubler

. Dengue. Monath

. The Arboviruses: Epidemiology and Ecology, II. Boca Raton, FL: CRC Press, 1988; 223–260.

14.

Gubler

. Dengue and dengue hemorrhagic fever: its history and resurgence as a global public health problem. Gubler

, Kuno

. Dengue and Dengue Hemorrhagic Fever. New York: CAB International, 1997; 1–22.

15.

Gubler

. The global emergence/resurgence of arboviral diseases as public health problems. Arch Med Res, 2002; 33:330–342.

16.

Gubler

, Nalim

, Tan

, Saipan

et al. Variation in susceptibility to oral infection with dengue viruses among geographic strains of Aedes aegypti. Am J Trop Med Hyg, 1979; 28:1045–1052.

17.

Gubler

, Trent

. Emergence of epidemic dengue/dengue hemorrhagic fever as a public health problem in the Americas. Infect Agents Dis, 1994; 2:383–393.

18.

Hanley

, Goddard

, Gilmore

, Scott

et al. Infectivity of West Nile/dengue chimeric viruses for West Nile and dengue mosquito vectors. Vector Borne Zoonot Dis, 2005; 5:1–10.

19.

Hanley

, Manlucu

, Gilmore

, Blaney

Jr.

et al. A trade-off in replication in mosquito versus mammalian systems conferred by a point mutation in the NS4B protein of dengue virus type 4. Virology, 2003; 312:222–232.

20.

Hanley

, Nelson

, Schirtzinger

, Whitehead

et al. Superior infectivity for mosquito vectors contributes to competitive displacement among strains of dengue virus. BMC Ecol, 2008; 8:1.

21.

Hintze

. NCSS Statistical Software. Kaysville, UT: NCSS, 1998.

22.

Huang

, Butrapet

, Pierro

, Chang

et al. Chimeric dengue type 2 (vaccine strain PDK-53)/dengue type 1 virus as a potential candidate dengue type 1 virus vaccine. J Virol, 2000; 74:3020–3028.

23.

Huang

, Butrapet

, Tsuchiya

, Bhamarapravati

et al. Dengue 2 PDK-53 virus as a chimeric carrier for tetravalent dengue vaccine development. J Virol, 2003; 77:11436–11447.

24.

Huang

, Silengo

, Whiteman

, Kinney

. 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, 2005; 79:7300–7310.

25.

Khin

, Jirakanjanakit

, Yoksan

, Bhamarapravati

. Infection, dissemination, transmission, and biological attributes of dengue-2 PDK53 candidate vaccine virus after oral infection in Aedes aegypti. Am J Trop Med Hyg, 1994; 51:864–869.

26.

Kinney

, Butrapet

, Chang

, Tsuchiya

et al. Construction of infectious cDNA clones for dengue 2 virus: strain 16681 and its attenuated vaccine derivative, strain PDK-53. Virology, 1997; 230:300–308.

27.

Kramer

, Presser

, Houk

, Hardy

. Effect of the anesthetizing agent triethylamine on western equine encephalomyelitis and St. Louis encephalitis viral titers in mosquitoes (Diptera: Culicidae) J Med Entomol, 1990; 27:1008–1010.

28.

Mahmood

, Chiles

, Fang

, Barker

et al. Role of nestling mourning doves and house finches as amplifying hosts of St. Louis encephalitis virus. J Med Entomol, 2004; 41:965–972.

29.

McElroy

, Tsetsarkin

, Vanlandingham

, Higgs

. Manipulation of the yellow fever virus non-structural genes 2a and 4b and the 3′non-coding region to evaluate genetic determinants of viral dissemination from the Aedes aegypti midgut. Am J Trop Med Hyg, 2006a. 75:1158–1164.

30.

McElroy

, Tsetsarkin

, Vanlandingham

, Higgs

. Role of the yellow fever virus structural protein genes in viral dissemination from the Aedes aegypti mosquito midgut. J Gen Virol, 2006b. 87:2993–3001.

31.

Miller

, Mitchell

. Passage of yellow fever virus: its effect on infection and transmission rates in Aedes aegypti. Am J Trop Med Hyg, 1986; 35:1302–1309.

32.

Reisen

, Barker

, Fang

, Martinez

. Does variation in Culex (Diptera: Culicidae) vector competence enable outbreaks of West Nile virus in California? J Med Entomol, 2008; 45:1126–1138.

33.

Reisen

, Fang

, Martinez

. Effects of temperature on the transmission of West Nile virus by Culex tarsalis (Diptera: Culicidae) J Med Entomol, 2006; 43:309–317.

34.

Sirigulpanit

, Kinney

, Leardkamolkarn

. Substitution or deletion mutations between nt 54 and 70 in the 5′ non-coding region of dengue type 2 virus produce variable effects on virus viability. J Gen Virol, 2007; 88:1748–1752.

35.

Troyer

, Hanley

, Whitehead

, Strickman

et al. A live attenuated recombinant dengue-4 virus vaccine candidate with restricted capacity for dissemination in mosquitoes and lack of transmission from vaccinees to mosquitoes. Am J Trop Med Hyg, 2001; 65:414–419.

36.

Vaughn

, Hoke

Jr. , Yoksan

, LaChance

et al. Testing of a dengue 2 live-attenuated vaccine (strain 16681 PDK 53) in ten American volunteers. Vaccine, 1996; 14:329–336.

37.

Wilson

, de Sessions

, Leon

, Scholle

. West Nile virus nonstructural protein 1 inhibits TLR3 signal transduction. J Virol, 2008; 82:8262–8271.

38.

Winkler

, Randolph

, Cleaves

, Ryan

et al. Evidence that the mature form of the flavivirus nonstructural protein NS1 is a dimer. Virology, 1988; 162:187–196.

39.

Yoksan

, Bhamarapravati

, Halstead

. Dengue virus vaccine development: study on the biological markers of uncloned dengue 1–4 viruses serially passaged in primary kidney cells. Kay

, Blok

. Proceedings of the Fourth Symposium of Arbovirus Research in Australia. Brisbane, Australia: T.D St. George, 1986; 35–38.