Michael J. Turell

As part of a field ecology study of arbovirus and malaria activity in the Amazon Basin, Loreto Department, Peru, we collected mosquitoes landing on humans at a forest site and inside and outside of residences and military barracks at periurban, rural, and village sites. We collected 11 Anopheles spp. from these four sites. An. darlingi, the principal malaria vector in the region, accounted for 98.7% of all Anopheles spp. collected at Puerto Almendra. Peaks in landing activity occurred during the December and April collection periods. However, the percent of sporozoite-positive Anopheles spp. was highest 1–2 months later, when landing activity decreased to approximately 10% of the peak activity periods. At all sites, peak landing activity occurred about 2 hours after sunset. These data provide a better understanding of the taxonomy, population density, and seasonal and habitat distribution of potential malaria vectors within the Amazon Basin region.

Rift Valley fever virus (RVFV) causes outbreaks of severe disease in domestic ungulates as well as humans in Africa. There is a logical concern that RVFV could be introduced into the Americas and cause significant health and economic damage based on the precedent of the introduction and spread of West Nile virus (WNV). Unfortunately, there are currently no licensed diagnostic assays available for RVFV in the Americas. In this work, we report on the ability of a novel dipstick assay, VectorTest RVFV antigen assay, modeled on the VecTest assay for WNV, to detect a RVFV-infected female within a pool of mosquitoes. The dipsticks provided results in <20 min, were easy to use, and did not require a laboratory with containment facilities. Although readily able to detect a mosquito with a disseminated RVFV infection, it only occasionally detected RVFV in a mosquito with a nondisseminated infection, and therefore may fail to detect some pools that actually contain one or more positive mosquitoes. The RVFV dipstick assay was highly specific and did not react with samples to which had been added yellow fever, West Nile, Venezuelan equine encephalitis, sandfly fever Naples, sandfly fever Sicilian, or sandfly fever Toscana viruses. The RVFV assay can provide a rapid, safe, easy-to-use assay to alert public health personnel to the presence of RVFV in mosquitoes. Results from this assay will allow a rapid threat assessment and the focusing of vector control measures in high-risk areas.

To determine which arthropods should be targeted for control should Rift Valley fever virus (RVFV) be detected in North America, we evaluated Culex erraticus (Dyar and Knab), Culex erythrothorax Dyar, Culex nigripalpus Theobald, Culex pipiens L., Culex quinquefasciatus Say, Culex tarsalis Coquillett, Aedes dorsalis (Wiedemann), Aedes vexans (Meigen), Anopheles quadrimaculatus Say, and Culicoides sonorensis Wirth and Jones from the western, midwestern, and southern United States for their ability to transmit RVFV. Female mosquitoes were allowed to feed on adult hamsters inoculated with RVFV, after which engorged mosquitoes were incubated for 7-21 d at 260C, then allowed to refeed on susceptible hamsters, and tested to determine infection, dissemination, and transmission rates. Other specimens were inoculated intrathoracically, held for 7 d, and then allowed to feed on a susceptible hamster to check for a salivary gland barrier. When exposed to hamsters with viremias > or =10(8.8) plaque-forming units/ml blood, Cx. tarsalis transmitted RVFV efficiently (infection rate = 93%, dissemination rate = 56%, and estimated transmission rate = 52%). In contrast, when exposed to the same virus dose, none of the other species tested transmitted RVFV efficiently. Estimated transmission rates for Cx. erythrothorax, Cx. pipiens, Cx. erraticus, and Ae. dorsalis were 10, 8, 4, and 2%, respectively, and for the remaining species were < or = 1%. With the exception of Cx. tarsalis and Cx. pipiens, all species tested had moderate to major salivary gland barriers. None of the C. sonorensis became infected and none of the An. quadrimaculatus tested transmitted RVFV by bite, even after intrathoracic inoculation, indicating that these species would not be competent vectors of RVFV. Although Ae. vexans from Florida and Louisiana were relatively efficient vectors of RVFV, specimens of this species captured in Colorado or California were virtually incompetent, illustrating the need to evaluate local population for their ability to transmit a pathogen. In addition to laboratory vector competence, factors such as seasonal density, host feeding preference, longevity, and foraging behavior should be considered when determining the potential role that these species could play in RVFV transmission.

This study was conducted as part of a field-ecology study of arboviral and malarial activity in the Amazon Basin, Loreto Department, Peru, to determine the relative abundance, species diversity, and seasonal and vertical distributions of potential mosquito vectors. Mosquitoes were captured either by volunteers using mouth aspirators while mosquitoes attempted to land on the collectors or in dry ice–baited ABC light traps. Anopheles darlingi, the principal malaria vector in the region, was the most commonly captured anopheline mosquito in Puerto Almendra village (99%) while landing on humans, with a mean of 37.1 mosquitoes captured per 24-h period, representing nearly one half of all mosquitoes collected. An. darlingi human landing activity began shortly after sunset, peaked at 2000–2100 hours, and declined gradually until sunrise. This species readily entered houses, because 51% of the An. darlingi captured by paired collectors, stationed inside and outside houses, were captured indoors. Human landing collections provided a more accurate estimate of human attraction of An. darlingi, capturing 30 times as many as co-located dry ice–baited ABC light traps. In contrast, eight times as many Culex (Melanoconion) species, including known arbovirus vectors, were captured in light traps as by co-located human collectors. Despite being located within 300 m of the village collection site, only a few Anopheles species were captured at the forest collection site, including only 0.1 An darlingi/24 h, thus indicating that An. darlingi activity was directly associated with the rural village. These data provide a better understanding of the taxonomy, population density, and seasonal distribution of potential mosquito vectors of disease within the Amazon Basin region and allow for the development of appropriate vector and disease prevention strategies that target vector populations.

Mosquitoes were collected in the Amazon Basin, near Iquitos, Peru, and used in experimental studies to evaluate their susceptibility to strains of ea

In an attempt to improve the current live, attenuated vaccine (TC-83) for Venezuelan equine encephalitis virus (VEEV), specific mutations associated with attenuation of VEEV in rodent models were inserted into a full-length cDNA clone of the Trinidad donkey strain of VEEV by site-directed mutagenesis. Because some viruses have been reported to be more pathogenic when introduced by mosquito bite than the same virus introduced by needle inoculation, there were concerns that the presence of mosquito saliva, or changes in the virus caused by replication in a mosquito, might allow the virus to overcome the protective effects of prior vaccination with V3526. Therefore, we determined if hamsters vaccinated with V3526 were protected from challenge with the virulent Trinidad donkey strain of VEEV. All non-vaccinated hamsters died after intraperitoneal challenge or after being fed on by VEEV-inoculated Aedes taeniorhynchus. In contrast, hamsters vaccinated with V3526 were resistant to intraperitoneal challenge and infection by VEEV-infected Ae. taeniorhynchus. Therefore, the V3526 candidate vaccine elicits protection against VEEV infection by mosquito bite.

Little is known about the replication of Karshi virus, a member of the mammalian tick-borne flavivirus group, in its rodent hosts. Therefore, we developed a quantitative real-time RT-PCR assay and measured the amount of viral RNA in selected tissues of infected Swiss Webster mice. Two-day-old mice were highly susceptible, with 100% fatality 9 to 12 days after infection, whereas infection in 9-day-old mice was less virulent, with death occurring only rarely. In nearly all cases, mice inoculated when 2 days old contained similar numbers of viral genome equivalents from blood and liver samples from any given mouse, with titers declining after day 7. In contrast, the amount of viral RNA in the brain began to rise rapidly 4 days after exposure, peaked at about 6 days after virus exposure (titer of > 10(13) genome equivalents/g), and remained at that level until euthanasia or death. Viral profiles were similar in needle-inoculated or tick-exposed mice.