Vector incompetency: its implication in the disappearance of epizootic Venezuelan equine encephalomyelitis virus from Middle America.

Mesenteronal infection and transmission thresholds for several Middle American epizootic and enzootic Venezuelan equine encephalomyelitis (VEE) viruses were evaluated in field-collected (Guatemala) and colonized individuals of the enzootic vector mosquito Culex (Melanoconion) taeniopus. Approximate thresholds for 3 epizootic, hemagglutination-inhibition (HI) subtype I-AB strains were 10 chick embryo cell (CEC) plaque-forming units (pfu) per blood meal. Infection rates never exceeded 33%, despite oral doses as high as 10 CEC pfu. Mosquitoes were incapable of transmitting epizootic viruses to hamsters even after oral doses of 10 4 CEC pfu and 21 days of extrinsic incubation. In contrast, infection thresholds for 3 enzootic, HI subtype I-E viruses were less than 1O CEC pfu, as were transmission thresholds. Mesenteronal 50% infectious doses were also less than 10 for all enzootic strains evaluated. The possibility of transovarial transmission of VEE viruses was evaluated by rearing progeny from Aedes taeniorhynchus mosquitoes infected with a Middle American epizootic VEE virus strain. No virus was recovered from 1,037 fourth-instar larvae reared at 20 or 27 °C. Culex taeniopus mosquitoes were also infected with enzootic viruses and progeny from 2nd and subsequent ovarian cycles tested for infectious virus. Vertical transmission was not detected despite assays of 5,077 fourth-instar larvae reared at 20 or 27 °C. Epidemiological implications of these findings related to postepizootic survival of HI subtype I-AB viruses in Middle America are discussed. Historically, only 2 epidemics and equine epizootics caused by the mosquito-borne Venezuelan equine encephalomyelitis (VEE) virus have been recognized in Middle America. The 1st occurred in 1966 near Tampico, Mexico, and was identified as VEE virus by antibody tests; no virus strains were isolated (Morilla-Gonzales & Mucha-Macias 1969; Dickerman et al. 1971). The 2nd, a large epidemic and equine epizootic, occurred during 1969-1972 (Sudia et al. 1971; Scherer et al. 1972; Sudia & Newhouse 1975; Scherer et al. 1976b). It began near the Guatemala-El Salvador border and spread 1. This research received support from the U.S. Army Medical Research and Development Command Contract No. DADA 17-72-C-2140 and U.S. Hatch Grant No. NYC-139417. Address reprint requests to Dr E.W. Cupp, Department of Entomology, Cornell University, Ithaca, New York 14853, USA. 2. Department of Microbiology, Cornell University Medical College, 1300 York Avenue, New York, New York 10021, USA. 3. Deceased. 4. Department of Entomology, Cornell University, Ithaca, New York 14853, USA. to Honduras, Nicaragua, Costa Rica, Mexico, and Texas, USA. Although VEE virus had been found in focal rain-forest and marsh habitats in Middle America before these outbreaks (Scherer et al. 1970; Scherer & Dickerman 1972), the strains isolated from endemic-enzootic habitats were different from strains recovered during the epizootics. Various differences in marker characteristics have been previously reported (Walton et al. 1973; Johnson 8c Martin 1974; Jahrling & Eddy 1977; Scherer & Chin 1977; Wiebe & Scherer 1980). There are 2 distinct cycles of VEE virus: epidemic-equine epizootic and endemic-enzootic. During epidemic-equine epizootic (henceforth referred to as epizootic) cycles, virus is transmitted by 1 or more mosquito species; Aedes taeniorhynchus (Christophers) and Psorophora confinnis (Lynch Arribalzaga) are proven vectors, while Aedes sollicitans (Walker), Mansonia titillans (Walker), Mansonia indubitans (Dyar 8c Shannon), and Culex nigripalpus (Theobald) are suspected, based on numerous virus isolations (Sudia & Newhouse 1975). The principal vertebrate hosts of epizootic cycles of VEE virus are equines and humans. However, additional vertebrates such as dogs, pigs, and cows are inapparently infected (Dickerman et al. 1973a,b). Viremic equines, humans, and possibly dogs serve as amplifying hosts. Equines are of primary importance because they are bitten by large numbers of mosquitoes. In contrast, the Central American enzootic cycle relies on a single vector, Culex (Melanoconion) taeniopus (Dyar & Knab); wild rodents and possibly marsupials act as vertebrate amplifying hosts (Scherer et al. 1985). The source of the epizootic strains of VEE virus that appeared in Middle America is unknown. The last recognized equine cases caused by a Middle American epizootic strain occurred in Nicaragua during 1972 (Scherer et al. 1976c; Dickerman 8c Scherer 1983). Scherer et al. (1976a,c) evaluated the possibility that epizootic strains of VEE virus are continuously cycling in occult natural habitats in Middle America. No persistent activity of epi24 J. Med. Entomol. Vol. 23, no. 1 zootic VEE virus strains was found in Guatemala, El Salvador, or Nicaragua by use of sentinel horses, hamsters, and guinea pigs, by serologic tests of humans and wild rodents, or by attempting to isolate virus from mosquitoes (Scherer et al. 1976c). However, 73 strains of VEE virus were recovered during 1970 and 1971 at marsh habitats in Guatemala from sentinel hamsters or mosquitoes. All of these produced predominantly large plaques in Vero cell cultures, suggestive of enzootic rather than epizootic VEE viruses (Scherer et al. 1976a). Enzootic VEE virus isolates made before and after the 1966 and 1969-1972 outbreaks were analyzed by hydroxylapatite chromatography and cloning in Vero cells to detect possible minority populations of epizootic virions. However, all 94 candidate epizootic clones were hemagglutinationinhibition (HI) subtype I-E, were avirulent for English shorthair guinea pigs inoculated subcutaneously, and had optimal pHs of hemagglutination at 6.2 (Wiebe et al. 1983). Survival of sentinel English shorthair guinea pigs in a Guatemalan enzootic habitat during 1977-1980 also suggested the absence of epizootic subpopulations, since these animals die from mixed enzootic/epizootic infections (Scherer et al. 1985). Thus, as of 1980, the use of multiple field and laboratory techniques has yielded no evidence of persistence of epizootic strains of VEE virus in Middle America following the outbreaks of 1966 and 1969-1972. These observations strongly suggest that the epizootic strains of Middle American outbreaks disappeared from this region. Recent findings have suggested that this disappearance may be related to the vector incompetence of the Middle American enzootic vector mosquito, Cx, taeniopus, for certain epizootic VEE virus strains (Scherer et al. 1982). To further investigate this phenomenon, we examined the ability of this mosquito to become infected with and transmit additional epizootic and enzootic viruses. We also examined the possibility that transovarial transmission might be involved in interepizootic VEE virus survival by evaluating this phenomenon in a vector mosquito associated with past outbreaks, Ae. taeniorhynchus (Sudia & Newhouse 1975). Vertical transmission was also examined in Cx. taeniopus to determine whether this phenomenon might aid in the survival of enzootic viruses during periods of inadequate horizontal transmission. MATERIALS AND METHODS Mosquitoes. Adult female Cx. taeniopus mosquitoes were either collected while host seeking at La Avellana, Guatemala (Scherer et al. 1976a; Cupp et al. 1979) or were obtained from progeny reared in laboratories at Cornell University in Ithaca, New York, or New York City. Mosquitoes were held in 45-cm cubical cages lined with plastic and containing clay pots inverted in water as resting and oviposition sites (Dziem & Cupp 1983). Larvae were fed a diet consisting of dry, active yeast and finely ground rabbit chow. Aedes taeniorhynchus mosquitoes were derived from a colony originating in La Avellana, Guatemala, in 1979. Larvae were fed a diet similar to that described above, and adult mosquitoes were held under similar conditions. Eggs were collected on moistened paper towels and incubated at 20 or 27 °C for 3 days before flooding with nutrient broth for hatching. VEE viruses and tests of mosquitoes for infection and ability to transmit. Data on the strains of VEE virus used are presented in Table 1. Virus strains selected were of the lowest vertebrate passage levels available, yet were generally well characterized. Mosquito suspensions were prepared and assayed routinely in chick embryo cells (CEC) as described previously (Cupp et al. 1979; Scherer et al. 1982). Some suspensions were tested in 1-4-dayold suckling albino mice (SM), inoculated intracranially (ic), to determine 50% lethal doses (SMicLD50). Transmissions of virus to hamsters were confirmed by testing suspensions of heart tissues from dead hamsters and showing cytopathic effects (plaques appearing in CEC) within 2-3 days at 37 °C. Infection and maintenance of mosquitoes. Culex taeniopus mosquitoes were infected by allowing them to engorge on viremic Syrian hamsters as described previously (Scherer et al. 1982; Dziem & Cupp 1983). During extrinsic incubation, mosquitoes were held at 26-28 °C and about 80% RH, with a 14:10 light:dark cycle including simulated dusk and dawn conditions. Mosquitoes used for experiments in Guatemala were held under local field conditions (Dziem & Cupp 1983). Transovarial infection attempts. Aedes taeniorhynchus mosquitoes were infected by intrathoracic injection with ca. 100 CEC pfu of virus strain 69U332. Oral infection was not attempted because infection rates rarely exceed 50% with high-titer hamster blood meals (Scherer, unpubl. data). Inoculated mosquitoes were transferred into 32-cm cubical cages and held at either 20 or 27 °C. Mosquitoes were offered blood meals from clean hamsters at 1986 Scherer et al.: Vector incompetency for VEE viruses 25 TABLE 1. Pedigrees of Middle American Venezuelan equine encephalomyelitis viruses used for infection experiments.