|
|
||||||||
| ABSTRACT |
|
|
|---|
| INTRODUCTION |
|
|
|---|
The ChimeriVaxTM platform developed by Acambis, Inc. uses the live attenuated 17D vaccine strain of YFV, which has a long history of safe and efficacious use. Vaccines for JEV, West Nile virus (WNV), and DENV have been developed based on this system and have been characterized both in vitro and in vivo.316 A YFV/St. Louis encephalitis chimeric virus has been produced as a diagnostic tool.17 Previous vector-competence studies of ChimeriVaxTM viruses have used both established colonies of mosquitoes and up to the fifth generation after collection of wild mosquito populations.18,19 Bhatt and others20 reported that ChimeriVaxTM-JE (YF/JE SA14-14-2) virus was not orally infectious for any of the species tested (Culex tritaeniorhynchus, Aedes aegypti, Ae. albopictus) despite the high titer of virus to which the mosquitoes were exposed (8.9 log10 pfu/mL). Although the virus replicated in Ae. aegypti and Ae. albopictus after intra-thoracic inoculation, it did not replicate in the natural vector Cx. tritaeniorhynchus. Replication in the vector did not result in genetic change. The ChimeriVaxTM-WNV is similarly attenuated in mosquitoes, with restricted replication and no dissemination in Cx. tritaeniorhynchus, Cx. nigripalpus, Cx. quinquefasciatus, Ae. aegypti, and Ae. albopictus.18 Vector competence studies with ChimeriVaxTM-DEN2 virus focused on Ae. aegypti and Ae. albopictus.21 When 6.37.7 log10 pfu/mL of virus was fed to mosquitoes, 17% of Ae. aegypti and 24% of Ae. albopictus were infected after 14 days; however, no dissemination occurred. Field-collected Ae. aegypti from Puerto Rico, DeQuincy, and the Republic of Vanautu were 100% susceptible to infection with DENV-2. However, ChimeriVaxTM-DEN2 virus replication was highly restricted in these populations, with infection rates ranging from 6% to 32% and no dissemination of virus in any population. A subsequent study19 used a highly sensitive real-time reverse transcriptase-polymerase chain reaction (RT-PCR) to determine the replication kinetics of the ChimeriVaxTM-DEN 1, 2, 3, 4 tetravalent virus mixture in Ae. aegypti. Growth rates of each chimeric virus in intrathoracically inoculated mosquitoes were similar regardless of whether it was used in the tetravalent mix or as a single serotype infection. However, the viruses were highly attenuated when presented to mosquitoes in a blood meal, and the authors concluded that there is very little transmission potential for these viruses.
Relatively few species of mosquito have been implicated in the DENV transmission cycle. In Asia, Ae. aegypti, Ae. albopictus, Ae. malaysensis, Ae. scutellaris group, and the Ochlerotatus niveus-group mosquitoes have been identified as potential vectors, but of these, most are probably of minor importance. For human epidemics, only Ae. aegypti and Ae. albopictus have played a significant role. In the South Pacific, Ae. polynesiensis may potentially be an important vector.
Recent concerns expressed about the potential for vaccine viruses to revert to a virulent phenotype22 may be unwarranted.23,24 However, because of, for example, the potential risks of inadvertently administering live vaccines to immunocompromized individuals (who may develop higher viremias), including those with HIV/AIDS living in areas at high risk for dengue, it is important to determine the transmission potential of live attenuated chimeric viral vaccines. The purpose of this research was therefore to determine the susceptibility of mosquitoes from natural populations in Asia, where DENV vaccines are most needed, to infection with ChimeriVax viruses. Mosquitoes were obtained from Thailand and reared at University of Texas Medical Branch (UTMB), Galveston, TX. The field-collected mosquitoes were expanded in the laboratory, and second-generation adults were orally challenged with the viruses to determine infection and dissemination rates. Genetically these mosquitoes should be representative of the natural populations with respect to their susceptibility to infection.
| MATERIALS AND METHODS |
|
|
|---|
Mosquitoes. Aedes aegypti eggs were collected in the spring of 2004 from Mae Sot Province, Thailand (16°45' N, 98°33' E), and Ae. albopictus eggs were collected in the summer 2004 from Surat Thani (ST) Province, Thailand (9°34' N, 90°07' E). Both species were reared for several generations, and F2 eggs were collected and stored for use in infectious feeds.
Mosquito infection.
Experimental mosquitoes, 710 days after emergence, were deprived of sugar for 24 hours and of water for 12 hours before being presented with the artificial infectious blood meal. The artificial blood meal contained equal parts of defibrinated sheep blood and freshly harvested virus, presented through a Hemotek feeder. Mosquitoes were allowed to feed for 30 minutes. To obtain freshly harvested virus, C6/36 (Ae. albopictus) cells were infected with wild-type dengue viruses at a multiplicity of infection (MOI) of 0.01. Supernatant medium was harvested at 6 (DENV-1) or 7 days (DENV-2, -3, -4) post-infection (p.i.) for immediate incorporation into the infectious blood meal. Vero (green monkey kidney) cells were infected with YF-VAX and ChimeriVaxTM-DEN1, 2, 3, and 4 viruses, also at an multiplicity of infection (MOI) of 0.01, and the supernatant was harvested after 6 days. For the tetravalent mix, 1.5 mL each of ChimeriVaxTM-DEN1, 2, 3, and 4 was combined, and 5 mL of this mixture was added to 5 mL of defibrinated sheep blood. Each blood meal was heated to 37°C just before feeding. Mosquitoes engorged to
70% capacity were kept and incubated at 28°C. Three mosquitoes were collected on days 0 and 1. All remaining mosquitoes were collected on day 21. Mosquitoes were stored at 80°C until assayed for infection. Oral infection was divided into two sets. In the first set, Ae. albopictus and Ae. aegypti were exposed to wild-type DENV-1, 2, 3, and 4 and the tetravalent mix. In the second set, Ae. albopictus and Ae. aegypti were exposed to ChimeriVaxTM-DEN1, 2, 3, and 4 and YF-VAX.
As negative controls, Ae. albopictus F2, Thailand were fed a non-infectious meal containing equal parts of culture supernatant collected from uninfected cells. One negative control mosquito was titrated with all sets of infected mosquitoes. To produce positive control mosquitoes, aliquots of wild-type DENV-1, 2, 3, and 4; ChimeriVaxTM-DEN1, 2, 3, and 4; and YF-VAX viruses were thawed and immediately used for in-trathoracic inoculation of 7- to 10-day-old mosquitoes25 (0.5 µL/mosquito). Inoculated mosquitoes were maintained for 21 days at 28°C and collected and stored at -80°C pending analysis. Positive control mosquitoes were titrated with all sets of orally infected mosquitoes.
Determination of virus titer. Mosquito bodies and heads were assayed for infectious virus separately to determine overall infection rate and disseminated infection rate, respectively. Mosquito bodies were triturated in 1 mL of L-15 media (10% fetal bovine serum (FBS) + 10% tryptose phosphate broth (TPB) + 100 units/mL penicillin + 100 µg/mL streptomycin + 1 µg/mL fungizone). One hundred microliters of each sample was loaded in duplicate and titrated in serial 10-fold dilutions in Vero cell culture in the first 8 rows of a 96-well plate. Mosquito heads were triturated in 150 µL of L-15 medium (same as above) and titrated in serial 10-fold dilutions in Vero cells in the last 4 wells of the same rows as corresponding bodies. Titration plates were incubated at 37°C for 10 days and fixed with 3:1 acetone-phosphate-buffered saline (PBS) for 10 minutes, dried, and stored at 20°C until analyzed by immunofluorescence assay (IFA). Mosquitoes that had ingested the tetravalent mix of chimeric viruses were titrated in quadruplicate, so that samples could be assayed for each of the four ChimeriVaxTM vaccine serotypes. Therefore, only 10 µL of both bodies and heads were loaded into plates to ensure there would be enough sample for each assay.
Immunofluorescence assay. Viruses differed in their capacity to produce cytopathic effect (cpe) on Vero cells. While all of the ChimeriVaxTM viruses consistently caused cpe, wild-type viruses were variable. In preliminary tests using titrations of stock ChimeriVaxTM viruses, it was determined that wells without cpe were always negative when tested for IFA. All titration plates containing wild-type DENV were therefore tested by IFA26 to determine the endpoint (i.e., the highest dilution at which antigen was detected). Because the ChimeriVaxTM-DEN-1, 2, 3, and 4 and YF-VAX viruses caused cpe in Vero cells, only plates in which cpe were observed were tested by IFA.
A variety of primary antibodies were used. Selection of the antibody and optimal dilution to give consistently bright and readable IFA data was based on visual evaluation using Vero cells infected with each of the viruses, grown on glass cover-slips,27 and using positive and negative control mosquitoes titrated on Vero cells in 96-well plates. ChimeriVaxTM-DEN1, 2, 3, and 4 and the tetravalent mix plates were stained using serotype-specific antibodies6 provided by Acambis. Acambis DENV-1, -2, or -4 antibodies were used at a 1:200 dilution, and the Acambis DENV-3 antibody was used at a 1:100 dilution. Plates assayed for YF-VAX were stained using a hyperimmune mouse serum (designated MA93, produced by S. Higgs in 1993) at a 1:500 dilution. All of the plates assayed for the viruses mentioned above were incubated overnight at 4°C, before the application of a secondary antibody, anti-mouse Ig, biotinylated, species-specific whole antibody from sheep (Amersham Biosciences, Piscataway, NJ) at 1:200 for 40 minutes at 37°C, followed by streptavidin fluorescein (1: 200) for 10 minutes at 37°C. DENV-1, 2, 3, and 4 plates were stained with a rabbit-anti-DENV hyperimmune serum at 1:100 (produced by S. Higgs in 1994) for 40 minutes at 37°C followed by a secondary antibody, anti-rabbit, Ig, biotinylated, species-species whole antibody from donkey (Amersham Biosciences) at 1:100 for 40 minutes at 37°C and streptavidin fluorescein (1:200) for 10 minutes at 37°C. 1.4 Diazobicyclo(2,2,2)Octane (DABCO)-glycerol mounting solution was added to all plates, which were scored using an Olympus IX70 inverted epifluorescence microscope. Infection and dissemination rates were compared with Fisher exact test using SPSS version 11.5 (SPSS, Chicago, IL).
| RESULTS |
|
|
|---|
|
|
|
Comparison of total dissemination rates, defined as the number of positive heads per total number mosquitoes tested, revealed differences between wild-type DENV-2 (77%) and ChimeriVaxTM-DEN2 in the tetravalent mix (0), between wild-type DENV-4 (25%) and ChimeriVaxTM-DEN4 in the tetravalent mix (0), and between YF-VAX (5%) and DENV-1 (100%) and DENV-2 (77%) (Table 2
).
Infection and dissemination rates of the ChimeriVaxTM vaccines, both individually and in the tetravalent mix, were generally lower than those of wild-type DEN viruses and close to that of YF-VAX (Figure 1A
; Table 2
). The infection rates of ChimeriVaxTM-DEN3 (5/20 mosquitoes) and DENV-3 (6/20 mosquitoes) were similar, but the dissemination rate of ChimeriVaxTM-DEN3 was considerably lower than that of its wild-type counterpart (5% versus 30%, respectively; Figure 1A
).
Only ChimeriVaxTM-DEN4 had an infection rate higher than that of its wild-type counterpart (8/20 versus 6/20, respectively; Figure 1A
; Table 2
); however, the difference was not significant (P > 0.05). The infection rate of the individually fed ChimeriVaxTM-DEN4 (8/20) was higher than that of ChimeriVaxTM-DEN4 in the mix (2/20) (Table 2
). Importantly, the dissemination rate of ChimeriVaxTM-DEN4 was lower than that of DENV-4 (10% versus 25%; Figure 1A
).
One difference between the ChimeriVaxTM-DEN4 alone and the ChimeriVaxTM-DEN4 in the mix that might have contributed to the observed difference in the infection rate was that the titer of the ChimeriVaxTM-DEN4 blood meal was ~1.5 logs higher than that of ChimeriVaxTM-DEN4 in the tetravalent mix blood meal (Table 1
). However, this was also true of the blood meals containing ChimeriVaxTM-DEN1 and ChimeriVaxTM-DEN1 as part of the mix where the infection rates in Ae. albopictus did not differ as markedly (Table 1
). Overall, 19/160 (11.8%) of Ae. albopictus that were fed the ChimeriVaxTM vaccines became infected, but only 4/160 (2.5%) developed a disseminated infection.
Ae. aegypti, Thailand.
The infection rates of the wild-type DEN viruses were all significantly higher than their vaccine counterparts, either individually or as part of a tetravalent mix (Table 2
). All of the wild-type DENV infection rates were also significantly higher than that of YF-VAX (Table 1
; Figure 1B
). No significant difference (P > 0.05) was found between the infection rates of the ChimeriVaxTM-DEN-1, 2, 3 vaccine viruses, when used singly (all 0) or in the tetravalent mix (all 0), compared with ChimeriVaxTM-DEN4 (1/17) (Figure 1B
). There was a significant difference (P < 0.05) between ChimeriVaxTM-DEN-4 (1/17) and ChimeriVaxTM-DEN-4 in the tetravalent mix (0/14) and between ChimeriVaxTM-DEN4 (1/17) and YF VAX (0/14) (Table 2
). No other significant differences were determined for infection rates. Among the four ChimeriVaxTM vaccines, ChimeriVaxTM-DEN4 had the highest rate of infection in Ae. aegypti and Ae. albopictus (1/17 and 8/20, respectively) and also had the highest dissemination rate in Ae. albopictus (10%) (Table 2
).
Overall, the infection rates of the ChimeriVaxTM vaccines in Ae. aegypti (Thailand) were extremely low (06%), compared with those of the wild-type viruses (4777%) and similar to that of the vaccine strain YF-VAX (0%) (Table 2
), which is known to have a safe track record and not to be transmitted by mosquitoes.29 Only 1 of 124 Ae .aegypti mosquitoes, which ingested the ChimeriVaxTM vaccines either individually or as part of the mix, became infected (0.8%; Table 2
), and none developed a disseminated infection (Figure 1B
; Table 2
).
| DISCUSSION |
|
|
|---|
Vector competence studies have been an important component of previous studies to characterize candidate live attenuated vaccine viruses. Typically, these viruses are poorly infectious to mosquitoes, with low transmission rates. For the DEN-2 S-1 vaccine virus, an average of 16% of mosquitoes became infected, and none transmitted (versus 56% and 14%, respectively, for the parent).53 It was estimated that the vaccine strain was ~63 times less efficient than the parent strain in orally infecting Ae. aegypti. A few mosquitoes that were allowed to feed on humans infected with the S-1 strain became infected, but dissemination and transmission did not occur.54 By intra-thoracic inoculation, both viruses produced similar infection and dissemination rates (98100%). In similar studies, a DENV-4 vaccine candidate (PDK35-TD3 FRhL p3) was attenuated for Ae. aegypti, but DENV-1 (45AZ5) and DENV-3 candidates (CH53489) were not.55,56 In comparison with its wild-type DENV-4 strain 814669 parent, a live attenuated DENV-4 vaccine (2A
30) produced by reverse genetics57 is slightly restricted in its ability to infect Ae. aegypti midguts and restricted in its ability to disseminate to the salivary glands.58 Importantly, the 2A
30 vaccine virus was not transmitted to Ae. albopictus mosquitoes fed on human vaccines. JEV vaccine candidate strain SA14-2-8 was attenuated in mosquitoes compared with the parent (SA14) and remained avirulent after mosquito passage.59
Of particular relevance to this study, it has long been known that, whereas wild-type YFV can infect, disseminate, and be transmitted by Ae. aegypti, the 17D vaccine derivative can only infect midgut epithelial cells but cannot disseminate and be transmitted when mosquitoes are infected orally.29,6063 The French neurotropic vaccine (FNV) is similarly attenuated in mosquitoes.6466 In one study, when Ae. aegypti were presented with virus in a blood meal (viral titers 7.28.3 log10 PFU/mL), dissemination rates ranged from 90% to 100% for wild-type YFV but were only 3% for YFV 17D.62 Jenning and others63 reported that midguts became infected with 17D but also observed a lack of dissemination.
In this study, test viruses were propagated using the optimal conditions recommended by Acambis to maximize the titer presented in blood meals to mosquitoes. The highest titer was produced by YF-VAX virus, which as our gold standard control, is regarded as a safe and efficacious attenuated vaccine virus that is not transmissible by mosquitoes. Experimental mosquitoes imbibed relatively high viral titers compared with those titers typically observed in primates vaccinated with the ChimeriVaxTM viruses58 and those that they would likely encounter if feeding on vaccinated humans.9,12,13 Despite this deliberate bias to maximize the potential for vector infection, the candidate vaccines were highly attenuated, although the development of disseminated infections with all of the wild-type DENV serotypes showed susceptibility. Lack of infection with ChimeriVaxTM viruses is therefore indicative of viral attenuation and not of mosquito refractoriness. In the principle DENV vector, Ae. aegypti, no infections were observed with ChimeriVaxTM-DEN1, 2, and 3 viruses. A 6% infection rate of Ae. aegypti with ChimeriVaxTM-DEN4 failed to result in dissemination. In Ae. albopictus, ChimeriVaxTM-DEN4 infected 10% more mosquitoes than its wild-type counterpart, but disseminated in 15% fewer mosquitoes than the wild-type. These apparent differences between the DENV-4 serotype viruses and other serotypes seem to reflect the natural phenotype of the wild-type DENV-4 (strain 1288) as reported by others.19 The ChimeriVaxTM-DEN4 virus does, however, replicate to lower titer than YF-VAX,19 and so at titers anticipated in vaccinees is highly unlikely to infect vectors. The transient relatively low viremias observed in humans after immunization with, for example, the ChimeriVaxTM-DEN2 vaccine9 provides an additional safety characteristic likely to prevent vector infection.
In an overall comparison of the two mosquito species, Ae. aegypti was generally less susceptible to infection than Ae. albopictus, with lower dissemination rates and typically with lower viral titers attained after 21 days of incubation. Importantly, Ae. aegypti, the predominant vector of DENV, had the lowest infection and dissemination rates for the ChimeriVaxTM vaccines when fed individually and the lowest infection rate for all ChimeriVaxTM vaccines when presented as a tetravalent mix.
ChimeriVaxTM-DEN1, ChimeriVaxTM-DEN2, ChimeriVaxTM-DEN3, and the ChimeriVaxTM-DENV tetravalent mix viruses were never detected in Ae. aegypti at 21 days p.i., despite the presence of up to 4 log10TCID50 of virus immediately after feeding. When fed separately, ChimeriVaxTM-DEN4 virus was detected by whole body titration at 21 days p.i., but the titer was only 1.52 log10TCID50, with the virus presumably restricted to the midgut because no dissemination was observed. YF-VAX failed to infect Ae. aegypti. In Ae. albopictus, disseminated infections were observed for ChimeriVaxTM-DEN1, ChimeriVaxTM-DEN3, and ChimeriVaxTM-DEN4 viruses; however, the rates were considerably lower than observed for the wild-type parent viruses. We hypothesize that if these mosquitoes were exposed to the low viral titers developed in vaccinated humans,9 these infection and dissemination rates would be lower, or zero, because the titers will likely be below the infection threshold. Although ChimeriVaxTM-DEN2 and ChimeriVaxTM-DEN4 viruses were detected in Ae. albopictus after feeding with the tetravalent mix, infection rates were only 5% and 10%, respectively, with no dissemination. This is important because it is likely that these viruses will be used in the context of a mixture to induce broad spectrum protection and minimize the risks of adverse cross-serotypic responses. The overall infection and dissemination rates for Ae. albopictus fed with the ChimeriVaxTM-DEN vaccines were 19/160 (11.8%) and 4/160 (2.5%), respectively, whereas those for YF-VAX were 10% (2/20) and 5% (1/20), suggesting that the Acambis strategy for chimerization to produce vaccine viruses may contribute further to restrict dissemination in the mosquito vector.
Although some mosquitoes were infected with the vaccine viruses, dissemination rates were very low, < 10% for the individual vaccine viruses and always 0% in the tetravalent mix. These low rates were despite experimental mosquitoes imbibing relatively high viral titers compared with the titers that they would encounter if feeding on vaccinated humans. Furthermore, the whole body viral titers never exceeded 2.88 ± 0.56 log10TCID50/mosquito. Such a low titer is unlikely to facilitate transmission, although this will be evaluated in future studies. Based on these data, it seems highly unlikely that mosquitoes from these Asian populations would be infected with the ChimeriVaxTM-DEN viruses if feeding on vaccinated humans and so do not represent an opportunity for viral reversion and transmission to naïve persons. Although a low rate of dissemination did occur in Ae. albopictus, there was no dissemination in Ae. aegypti, which is considered to be the more important vector of DENV for humans.1 Moreover, the experiments were designed to exceed the likely conditions that would occur in the wild. This is therefore very reassuring.
Received May 1, 2006. Accepted for publication June 21, 2006.
Acknowledgments: The authors thank J. Huang for assistance in rearing mosquitoes.
Financial support: This work was supported by Sanofi-Pasteur, Marcy-LEtoile, France. Kate McElroy and Charles McGee are recipients of CDC Fellowships for training in vector-borne infectious diseases (TO1/CCT622892).
Disclosure: Some of the authors wish to disclose that they have financial interest in Acambis, the company that sponsored the study. Some authors are current or former employees of Acambis. These statements are being made in the interest of full disclosure and not because the authors consider this to be a conflict of interest.
* Address correspondence to Stephen Higgs, Department of Pathology, Keiller 2.104 301 University Boulevard, Galveston, TX 77555-0609. E-mail: sthiggs{at}utmb.edu ![]()
Authors addresses: Stephen Higgs, Dana Vanlandingham, Kimberly Klingler, Kate McElroy, and Charles E. McGee, Department of Pathology, University of Texas Medical Branch, Galveston, TX 77555-0609. Laura Harrington, Department of Entomology, 3138 Comstock Hall, Cornell University, Ithaca, NY 14853. Jean Lang, Sanofi Pasteur, Campus Mérieux, 1541 Avenue Marcel Mérieux, Marcy-Létoile F-69280, France. Thomas Monath and Farshad Guirakhoo, Acambis, Inc., 38 Sidney Street, Cambridge, MA 02139, Telephone: 617-761-4323, Fax: 617-494-1741. Current address for Thomas Monath, Kleiner Perkins Caufield & Byers, 21 Finn Road, Harvard MA 01451. Farshad Guirakhoo, Acambis, Inc., 38 Sidney Street, Cambridge, MA 02139, Telephone
| REFERENCES |
|
|
|---|
This article has been cited by other articles:
![]() |
N. C. Arrigo, D. M. Watts, I. Frolov, and S. C. Weaver Experimental Infection of Aedes sollicitans and Aedes taeniorhynchus with Two Chimeric Sindbis/Eastern Equine Encephalitis Virus Vaccine Candidates Am J Trop Med Hyg, January 1, 2008; 78(1): 93 - 97. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |