Download limited potential for mosquito transmission of genetically engineered

Am. J. Trop. Med. Hyg., 60(6), 1999, pp. 1041–1044
Copyright q 1999 by The American Society of Tropical Medicine and Hygiene
LIMITED POTENTIAL FOR MOSQUITO TRANSMISSION OF GENETICALLY
ENGINEERED, LIVE-ATTENUATED VENEZUELAN EQUINE ENCEPHALITIS VIRUS
VACCINE CANDIDATES
MICHAEL J. TURELL, GEORGE V. LUDWIG, JOHN KONDIG, AND JONATHAN F. SMITH
Virology Division, U.S. Army Medical Research Institute of Infectious Diseases, Fort Detrick, Frederick, Maryland
Abstract. In an attempt to improve the current live-attenuated vaccine (TC-83) for Venezuelan equine encephalitis
(VEE), specific mutations associated with attenuation of VEE virus in rodent models were identified. These mutations
were inserted into full-length cDNA clones of the Trinidad donkey strain of VEE virus by site-directed mutagenesis,
and isogenic virus strains with these mutations were recovered after transfection of baby hamster kidney cells with
infectious RNA. We evaluated 10 of these strains for their ability to replicate in and be transmitted by Aedes taeniorhynchus, a natural vector of epizootic VEE virus. Two vaccine candidates, one containing a deletion of the PE2
furin cleavage site, the other a combination of three separate point mutations in the E2 glycoprotein, replicated in
mosquitoes and were transmitted to hamsters significantly less efficiently than was either parental (wild type) VEE
virus or TC-83 virus. Although the attenuated strains were transmitted to hamsters by mosquitoes, after intrathoracic
inoculation, there was no evidence of reversion to a virulent phenotype. The mutations that resulted in less efficient
replication in, or transmission by, mosquitoes should enhance vaccine safety and reduce the possibility of environmental spread to unintentional hosts.
The current investigational new drug (IND) vaccine (TC83) for Venezuelan equine encephalitis (VEE) strain IAB
virus is a live-attenuated virus that causes reactogenicity in
20% of recipients and fails to elicit a positive seroresponse
in 20% of recipients.1 Current efforts to develop an improved
live-attenuated vaccine for VEE identified specific mutations
associated with attenuation of VEE virus in rodent models.2–
4
These attenuating mutations have been inserted into a fulllength cDNA clone of wild-type VEE virus (IAB) to produce
selected isogenic strains containing one or more attenuating
mutations. These attenuated strains are currently being evaluated for their potential as a live-attenuated VEE virus vaccine.
Live-attenuated vaccines typically offer many advantages
over inactivated immunogens (e.g., single immunization,
more efficient induction of mucosal immunity, longer duration of immunity). However, live arbovirus vaccines have
the potential to be transmitted to secondary hosts, and may
revert to a more virulent virus. This reversion to virulence
could occur in either the vertebrate or the arthropod host.
Thus, we examined VEE virus strains that contained attenuating mutations that might be included in a live-attenuated
vaccine for their ability to infect and be transmitted by mosquitoes. In addition, we evaluated these vaccine candidates
for reversion to a more virulent phenotype after intrathoracic
inoculation of Aedes taeniorhynchus mosquitoes.
Virus and virus assay. The strains of VEE virus evaluated are shown in Table 1. Potential vaccine candidate strains
were provided by N. Davis and R. Johnston (University of
North Carolina, Chapel Hill, NC).2,3,8 Plaque titers for specimens were determined on Vero cell monolayers grown in
6- or 12-well plastic cell culture plates. Serial 10-fold dilutions of each specimen were added to wells (0.1 ml/well).
After a 1-hr absorption period, a nutrient overlay (Eagle’s
basal medium with Earle’s salts, 7% heat-inactivated fetal
bovine serum, 0.75% agarose, and antibiotics [100 units/ml
of penicillin, 100 mg/ml of streptomycin, and 100 units/ml
of nystatin]) was added to each well and the plates incubated
at 358C for two days. Cells were then stained with 1 ml of
the above medium, except that 5% neutral red was used in
place of the fetal bovine serum and antibiotics. Plaques were
enumerated the following day.
Inoculation studies. Two- to 6-day-old female Ae. taeniorhynchus were inoculated intrathoracically9 with 0.3 ml
of a suspension containing about 105 plaque-forming units
(PFU)/ml (101.5 PFU/mosquito) of one of the strains of VEE
virus. Mosquitoes were placed in 0.5-liter cardboard containers with netting over the open end and held in an incubator maintained at 268C with a 16:8 hr light:dark photoperiod. To determine the potential for replication of each of
the VEE virus strains in mosquitoes, five mosquitoes were
removed from each cage at selected times, triturated individually in 1 ml of diluent (10% heat-inactivated fetal bovine
serum in medium 199 with Earle’s salts, antibiotics, and sodium bicarbonate), and frozen at 2708C until assayed for
virus.
To determine the ability of mosquitoes to transmit virus
by bite, mosquitoes inoculated $ 7 days previously were
allowed to feed individually on adult female Syrian hamsters. The hamsters were observed daily for 21 days. A sample of brain tissue was obtained from hamsters that died and
were tested by plaque assay to confirm the presence of virus.
Hamsters surviving $ 21 days were challenged with 104
50% lethal dose units (LD50) of the V3000 strain of VEE
virus. Infection with this strain is nearly always fatal for
MATERIALS AND METHODS
Mosquitoes. Two laboratory strains of Ae. taeniorhynchus, Vero Beach and Medical and Veterinary Entomology
Research Laboratory (MAVERL), were used during these
studies. Both of these strains have been in colonies for more
than 30 years and are derived from mosquitoes collected in
the late 1950s in Florida. Mosquitoes were held at 268C with
a 16:8 hr light:dark photoperiod and reared as described by
Gargan and others.5 Aedes taeniorhynchus is considered to
be a natural vector of VEE virus in the Americas,6 and both
strains are competent vectors of the epizootic IAB strain of
VEE virus7 (Turell MJ, unpublished data).
1041
1042
TURELL AND OTHERS
TABLE 1
Strains of Venezuelan equine encephalitis virus used in this study
Strain of
virus*
V3000
TC-83
V1000
V3014
V3038
V3040
V3042
V3043
V3520
V3526
Description
Parent clone†
Current live-attenuated IND vaccine
Deletion 5493-5595 (nucleotides)
E2-209 (Glu→Lys), E1-272 (Ala→Thr)
E3-59 (Arg→Asp)
E1-253 (Phe→Ser)
E1-81 (Phe→Ile)
Nt. #3 G→A
E2-76 (Glu→Lys), E2-209 (Glu→Lys), E1-81
(Phe→Ile)
PE-2 cleavage site deletion 1 E1-253 (Phe→Ser)
* Potential vaccine candidate strains were provided by N. Davis and R. Johnston.2,3,8
† Derived from the Trinidad donkey strain.
Syrian hamsters.3 Hamsters that died after mosquito feeding
were considered to have been infected with a virulent strain
of VEE virus. In contrast, hamsters that survived their initial
mosquito exposure, but died upon challenge with the V3000
strain were considered not to have been infected by mosquito
bite, while those that survived challenge were considered to
have been infected with an avirulent strain of VEE virus by
mosquito bite, and immunized by that infection.
Oral exposure studies. We used a membrane feeder to
feed female Ae. taeniorhynchus on a heparinized goose
blood-virus suspension containing serial dilutions of V3000,
V3526, or TC-83 viruses. After a 30-min feeding period,
unengorged mosquitoes were removed and discarded, while
engorged mosquitoes were held as described above for inoculated mosquitoes. After 14 days, the mosquitoes were
cold anesthetized and their legs and bodies were individually
triturated in 1 ml of diluent and tested for the presence of
virus as described above. Recovery of virus from the body,
but not the legs, indicated that viral infection was limited to
the midgut and had not disseminated to the hemocoel, while
recovery of virus from both legs and body indicated that the
mosquito had a disseminated infection.10
In conducting research using animals, the investigators adhered to the Guide for the Care and Use of Laboratory Animals, as prepared by the Committee on Care and Use of
Laboratory Animals of the Institute of Laboratory Animal
Resources, National Research Council (NIH Publication No.
86–23, Revised 1996). The facilities are fully accredited by
the Association for Assessment and Accreditation of Laboratory Animal Care, International.
TABLE 2
Replication and transmission of selected strains of Venezuelan
equine encephalitis virus in Aedes taeniorhynchus seven days after
inoculation of approximately 101.5 plaque-forming units of virus
Strain
of
virus
V3000
TC-83
V1000
V3014
V3038
V3040
V3042
V3043
V3520
V3526
Mean 6 SE
(no. tested)*
7.1
7.3
6.6
7.2
6.2
6.5
6.7
6.5
5.9
5.5
6
6
6
6
6
6
6
6
6
6
0.1
0.2
0.3
0.1
0.5
0.2
0.2
0.1
0.1
0.1
(21)
(6)
(9)
(13)
(5)
(5)
(10)
(4)
(15)
(21)
Transmission
rate
(no. tested)†
81%
64%
89%
93%
40%
100%
80%
75%
21%
18%
(43)a
(36)a
(9)a
(14)a
(5)abc
(5)a
(10)a
(4)ab
(14)bc
(33)c
Case fatality
rate
(no. infected)‡
94%
17%
100%
0%
100%
100%
100%
100%
0%
0%
(35)a
(23)c
(8)a
(13)c
(2)ab
(5)a
(8)a
(3)ab
(3)bc
(6)c
* Mean logarithm10 plaque-forming units per mosquito.
† Percentage of feeding Ae. taeniorhynchus that transmitted virus by bite as shown by
death from mosquito-transmitted virus or by hamster survival after challenge with the
V3000 parent strain. Numbers followed by the same letter are not significantly different at
a 5 0.05 by either a chi-square test or a Fisher’s exact test.
‡ Percentage of infected hamsters that died. Numbers followed by the same letter are not
significantly different at a 5 0.05 by either a chi-square test or a Fisher’s exact test.
tion as shown by hamster survival after challenge with the
V3000 parent strain (Table 2). Again, V3520 and V3526
viruses were transmitted significantly (x2 . 5.7, df 5 1, P
# 0.017) less efficiently than were either the parent virus or
the current vaccine virus (TC-83). Transmission rates for the
parent and TC-83 viruses were not significantly different (x2
5 2.25, df 5 1, P 5 0.13).
Although the strain containing the mutation at E2–209
(V3014) replicated to high titer and was efficiently transmitted by bite, none of the 13 hamsters infected with this
RESULTS
Mosquito inoculation studies. All of the strains of VEE
virus tested replicated in Ae. taeniorhynchus (Table 2). However, two viruses (V3520 and V3526) grew to significantly
lower titers (T $ 10.9, degrees of freedom [df] 5 $ 19, P
, 0.001) than did either the parent or TC-83 strains (Table
2 and Figure 1). In general, viral titers increased rapidly and
reached their highest levels about four days after inoculation,
and then gradually decreased with increased incubation (Figure 1).
Similarly, all of the strains were transmitted by bite after
the mosquitoes had been infected by intrathoracic inocula-
FIGURE 1. Replication of V3000 virus, the current live-attenuated Venezuelan equine encephalitis (VEE) virus vaccine strain (TC83), and three live-attenuated vaccine candidate strains of VEE in
Aedes taeniorhynchus after intrathoracic inoculation of approximately 102 plaque-forming units (PFU) of virus. Five mosquitoes
were sampled at each time point.
POTENTIAL MOSQUITO TRANSMISSION OF VEE VIRUS VACCINE STRAINS
TABLE 3
Replication of selected strains of Venezuelan equine encephalitis
virus in Aedes taeniorhynchus 14 days after ingestion of a viremic
blood meal from a membrane feeder*
Strain of
virus
V3000
TC-83
V3526
PFU/ml
ingested
No.
tested
Infection
rate†
Dissemination
rate‡
108
107
106
108
107
106
108
107
106
46
47
11
47
51
29
54
44
21
22
19
0
2
4
3
80
32
5
2
0
0
0
0
0
19
0
0
* PFU 5 plaque-forming units.
† Percentage of mosquitoes containing virus.
‡ Percentage of mosquitoes containing virus in their legs.
strain by mosquito bite died (Table 2). Similarly, none of
the nine hamsters infected with V3520 or V3526 by mosquito bite or 20 hamsters inoculated intraperitoneally with
105 PFU of V3526 died (Turell MJ, unpublished data). In
contrast, four (17%) of the 23 hamsters infected with TC-83
virus died of their initial infection. This was essentially identical to the mortality rate (16%) of the 25 hamsters that received an intraperitoneal inoculation of 105 PFU of TC-83
vaccine (Turell MJ, unpublished data).
Oral exposure studies. To test the oral susceptibility of
a potential vector mosquito, groups of Ae. taeniorhynchus
were fed blood-virus suspensions containing 104–108 PFU/
ml of V3000, V3526, or TC-83 viruses from a membrane
feeder. There were no significant differences in mosquito
susceptibility or dissemination rates between viral preparations unless mosquitoes received artificial blood meals containing $ 107 PFU/ml (Table 3). Mosquitoes were significantly more susceptible to V3526 virus than to TC-83 virus
when artificial blood meal titers contained 107 PFU/ml (x2
5 11.2, df 5 1, P # 0.001), and to both TC-83 and V3000
viruses when the artificial blood meal contained 108 PFU/ml
(x2 $ 31.1, df 5 1, P # 0.001). Differences in dissemination
rates were only apparent at the greatest viral concentration
tested. Again, the greatest rates occurred in mosquitoes ingesting V3526 virus (x2 . 5.1, df 5 1, P # 0.022).
DISCUSSION
A major environmental concern with the use of live-attenuated virus vaccines is the potential for spread of either
the vaccine virus or a pathogenic revertant to susceptible
hosts. This was documented by the isolation of TC-83, the
current IND live-attenuated VEE virus vaccine, from fieldcollected mosquitoes after TC-83 was used to vaccinate
equines during an outbreak of VEE in the early 1970s.11
Aedes taeniorhynchus was selected for these studies because it was incriminated as a vector in several outbreaks of
VEE.6,12,13 In addition to being a competent laboratory vector,7,14 it is a common mosquito that readily feeds on humans
throughout its range. Although female Ae. taeniorhynchus
became infected when they ingested the V3526 strain, a viremia $ 108 PFU/ml was needed for the development of a
disseminated infection in this potential mosquito vector. Be-
1043
cause no viremia was detected in any of six monkeys inoculated with V3526 (Pratt W, unpublished data), the viremia
needed for mosquito transmission is at least 100,000-fold
greater than the potential viremia that would have been observed in nonhuman primates, making dissemination of this
strain unlikely in the natural setting.
Other point mutations in VEE virus, such as the one at
E2–209 (V3014), had almost no effect on replication in, or
transmission by, Ae. taeniorhynchus, but these viruses remained avirulent after mosquito passage and protected hamsters challenged with 104 LD50 of virulent IAB virus.
The V3526 construct, which contains a four-amino acid
deletion in the PE2 furin cleavage site as well as a suppressor mutation in E1,8 proved to be more immunogenic in mice
than V3520 that contains three independently attenuating
point mutations. As expected, the deletion mutant was genetically stable, and showed no apparent phenotypic reversion on sequential passage in cell culture or in serial passage
in mice inoculated intracerebrally (Ludwig G, unpublished
data). For these reasons, V3526 was selected over V3520
and other potential candidates, and is currently being assessed in further preclinical studies.
Both V3526 and V3520 replicated in Ae. taeniorhynchus
after intrathoracic inoculation and both could be transmitted
by bite to a hamster. However, the two vaccine candidates,
V3526 and V3520, not only replicated less efficiently in
mosquitoes than did either TC-83 or the parent virus, but
also were transmitted less efficiently by mosquito bite. More
importantly, both of these vaccine candidates remained avirulent after mosquito passage. Based on the low viremias in
nonhuman primates, reduced ability to replicate in and be
transmitted by a known vector of epizootic VEE virus, and
evidence that V3526 does not revert to virulence after mosquito passage, it is unlikely that mosquito passage of this
virus vaccine candidate poses a significant environmental
danger.
Acknowledgments: We thank N. Davis and R. Johnston (University
of North Carolina, Chapel Hill, NC) for providing the various strains
of VEE virus used. We also thank D. Kline (Medical and Veterinary
Entomology Research Laboratory, Gainesville, FL) for providing the
MAVERL strain of Ae. taeniorhynchus; J. Johnson, J. Robles, and
B. Reynolds for assistance in rearing the mosquitoes; D. Dohm for
his assistance in processing specimens; and M. Hevey, G. Korch,
and K. Kenyon for their critical reading of the manuscript.
Disclaimer: The views of the authors do not purport to reflect the
positions of the Department of the Army or the Department of Defense.
Authors’ addresses: Michael J. Turell, John Kondig, and Jonathan F.
Smith, Virology Division, U.S. Army Medical Research Institute of
Infectious Diseases, 1425 Porter Street, Fort Detrick, MD 21702–
5011. George V. Ludwig, Diagnostic Systems Division, U.S. Army
Medical Research Institute of Infectious Diseases, 1425 Porter
Street, Fort Detrick, MD 21702–5011.
Reprint requests: Michael J. Turell, Vector Assessment Branch, Virology Division, U.S. Army Medical Research Institute of Infectious
Diseases, 1425 Porter Street, Fort Detrick, MD 21702–5011.
REFERENCES
1. Pittman PR, Makuch RS, Mangiafico JA, Cannon TL, Gibbs
PH, Peters CJ, 1996. Long-term duration of detectable neutralizing antibodies after administration of live-attenuated
1044
2.
3.
4.
5.
6.
7.
TURELL AND OTHERS
VEE vaccine and following booster vaccination with inactivated VEE vaccine. Vaccine 14: 337–343.
Davis NL, Willis LV, Smith JF, Johnston RE, 1989. In vitro
synthesis of infectious Venezuelan equine encephalitis virus
RNA from a cDNA clone: analysis of a viable deletion mutant. Virology 171: 189–204.
Davis NL, Powell N, Greenwald GF, Willis LV, Johnson BJB,
Smith JF, Johnston RE, 1991. Attenuating mutations in the E2
glycoprotein gene of Venezuelan equine encephalitis virus:
construction of single and multiple mutants in a full-length
cDNA clone. Virology 183: 20–31.
Grieder FB, Nguyen HT, 1996. Virulent and attenuated mutant
Venezuelan equine encephalitis virus show marked differences in replication in infection in murine macrophages. Microb
Pathog 21: 85–95.
Gargan II TP, Bailey CL, Higbee GA, Gad A, El Said S, 1983.
The effect of laboratory colonization on the vector-pathogen
interactions of Egyptian Culex pipiens and Rift Valley fever
virus. Am J Trop Med Hyg 32: 1154–1163.
Karabatsos N, ed, 1995. International Catalogue of Arboviruses
Including Certain Other Viruses of Vertebrates. Third edition.
San Antonio, TX: American Society of Tropical Medicine and
Hygiene.
Turell MJ, Ludwig GV, Beaman JR, 1992. Transmission of Venezuelan equine encephalomyelitis by Aedes sollicitans and
Aedes taeniorhynchus (Diptera: Culicidae). J Med Entomol
29: 62–65.
8. Davis NL, Brown KW, Greenwald GF, Zajac AJ, Zacny VL,
Smith JF, Johnston RE, 1991. Attenuated mutants of Venezuelan equine encephalitis virus containing lethal mutations
in the PE2 cleavage signal combined with a second-site suppressor mutation in El. Virology 212: 102–110.
9. Rosen L, Gubler D, 1974. The use of mosquitoes to detect and
propagate dengue viruses. Am J Trop Med Hyg 23: 1153–
1160.
10. Turell MJ, Gargan II TP, Bailey CL, 1984. Replication and dissemination of Rift Valley fever virus in Culex pipiens. Am J
Trop Med Hyg 33: 176—181.
11. Pedersen Jr CE, Robinson DM, Cole Jr FE, 1972. Isolation of
the vaccine strain of Venezuelan equine encephalomyelitis virus from mosquitoes in Louisiana. Am J Epidemiol 95: 490–
496.
12. Sellers RF, Bergold GH, Suarez OM, Morales A, 1965. Investigations during Venezuelan equine encephalitis outbreaks in
Venezuela— 1962–1964. Am J Trop Med Hyg 14: 460–469.
13. Rivas F, Diaz LA, Cardenas VM, Daza E, Bruzon L, Alcala A,
De la Hoz O, Caceres FM, Aristizabal G, Martinez JW, Revelo D, De la Hoz F, Boshell J, Camacho T, Calderon L,
Olano VA, Villarreal LI, Roselli D, Alvarez G, Ludwig G,
Tsai T, 1997. Epidemic Venezuelan equine encephalitis in La
Guajira, 1995. J Infect Dis 175: 828–832.
14. Kramer LD, Scherer WF, 1976. Vector competence of mosquitoes as a marker to distinguish Central American and Mexican
epizootic from enzootic strains of Venezuelan encephalitis virus. Am J Trop Med Hyg 25: 336–346.