Comparative analysis of G1 glycoprotein-coding sequences of Cache Valley virus (Bunyaviridae: Bunyavirus) isolates.

The complete 4463 nucleotide sequence for the medium segment viral RNA of Cache Valley virus has been cloned and sequenced in four isolates; in addition, the G1 glycoprotein extracellular coding domains are completed for nine additional isolates, including two subtypes, Ft. Sherman (86MSP18) and Tlacotalpan (61D240) viruses. The 13 represent isolations spanning over 45 years and a large geographic area, including the U.S., Mexico, Canada, and Panama. Glycosylation sites in G1 are generally conserved among all except the Ft. Davis, Panama (90P686) isolate, which loses a site otherwise conserved within the serogroup. Comparison of the G1 coding regions indicates a number of shared amino acid substitutions within a centrally located 70 amino acid hypervariable domain, which seems to fall outside the primary antigenic domains of G1, most of which are found within the amino half of the protein, while a less antigenic region is predicted for the carboxyl half of the protein encoded beyond the hypervariable domain. Numerous amino acid substitutions are found within various antigenic regions, which may be an indication of altered neutralization or hemagglutination sites. Putative phosphorylation sites are indicated, most of which are well conserved, with the exception of the absence of a specific protein kinase C site for the prototype (6V633) virus isolated in Utah. The overall nucleotide identity between isolates ranges from 91% (Ft. Sherman subtype, 86MSP18) to 99.4% (North Dakota, 1508-A52) as compared to the prototype virus (Utah, 6V633).

Sequence analysis of the medium (M) segment of Cache Valley virus, with comparison to other Bunyaviridae.

The complete sequence of the medium (M) segment of Cache Valley virus (CVV), a human neuropathogen, has been determined using a series of overlapping cDNA clones. The viral complementary-sense RNA is comprised of 4463 nucleotides which encodes a polyprotein precursor of 1435 amino acids, starting at AUG at bases 49-51 to a UGA stop codon at bases 4351-4353. This polyprotein-encoding sequence is arranged as G2-NSm-G1. The base composition of the segment is 34.9% A, 17.0% C, 19.4% G and 28.7% U. Comparison of the nucleotide sequence to the prototype Bunyamwera virus sequence shows an identity of 63%, indicating several differences exist within the individual coding regions, most notably within the NSm and G1 coding regions. Based on two presumed cleavage points within the precursor, the G2 glycoprotein, encoded from nt 94-951, is 286 amino acids long, and has two sites of potential glycosylation. NSm, encoded from nt 952-1476, is 175 amino acids, while the largest glycoprotein, G1, encoded from nt 1477-4350, consists of 958 amino acids, and has five potential glycosylation sites, two of which appear to be unique to CVV. The subsequent study of these glycosylation sites and potential differences between the sequence of this prototype CVV strain and other geographic isolates may suggest the means for improving detection of human infections as well as mapping differences in neurovirulence, neuroinvasiveness and other aspects of pathogenicity.

Cache Valley and Potosi viruses (Bunyaviridae) in white-tailed deer (Odocoileus virginianus): experimental infections and antibody prevalence in natural populations.

Cache Valley virus (CVV) and Potosi virus (POTV) are two closely related mosquito-borne viruses (Bunyaviridae: Bunyamwera group) that appear to circulate in several regions of the United States, especially the Midwest. We determined the prevalence of specific neutralizing antibodies to both viruses in Indiana white-tailed deer and conducted infection experiments to assess whether deer could serve as an vertebrate-amplifying host. Cross-infection experiments also were carried out to investigate the level of antibody cross-reactivity and cross-protection between the two viruses. The seroprevalence rate was high for both CVV (> 66%) and POTV (> 43%) in adult deer statewide. Antibodies neutralizing CVV were more common among deer harvested in the northern part of Indiana whereas the prevalence of POTV antibodies suggested a more southern distribution for this virus. Experimental infections of captive deer showed that they may serve as amplifying hosts for either virus. Deer infected with CVV or POTV developed a 1-3-day viremia with 3.0 and 4.1 log10 plaque-forming units/ml mean peak titers, respectively. However, significant levels of antibody cross-reactivity between the two viruses were observed. Viremia was lower and shorter when animals immune to either CVV or POTV were cross-infected with the alternate virus and antibody responses following cross-infections resembled original antigenic sin with higher titers of antibodies against the primary agent.

Introduction of Aedes albopictus into a La Crosse virus--enzootic site in Illinois.

In late summer and fall 1997, Aedes albopictus mosquitoes were found in Peoria, Illinois, a long recognized focus of La Crosse virus transmission. Larvae were found in tires and other artificial containers, biting adults were recovered, and eggs were collected in oviposition traps within a 25-ha area. One chipmunk trapped < 0.25 km from the infested area tested positive for neutralizing antibodies against La Crosse virus.

Role of Anopheles quadrimaculatus and Coquillettidia perturbans (Diptera: Culicidae) in the transmission cycle of Cache Valley virus (Bunyaviridae: Bunyavirus) in the midwest, USA.

Midwestern populations of Coquillettidia perturbans (Walker) and Anopheles quadrimaculatus (Say) were tested for their ability to transmit Cache Valley virus (CV), a recognized human and animal pathogen. Field-collected mosquitoes were fed artificial blood meals containing 5.2-6.2 log10 pfu/ml of CV. After 9-23 d at 28 degrees C, 75-93% of blood-fed Cq. perturbans had disseminated infections and 6-62% transmitted the virus to suckling mice. However, when infected with a lower virus titer (3.3 log10 pfu/ml), only 10-36% of the mosquitoes had disseminated infections and 0-10% transmitted the virus to suckling mice. A similar infection rate (21%) was observed in Cq. perturbans fed on viremic (3.2 log10 pfu/ml) hamsters. An. quadrimaculatus were infected (81-100%) by both doses used, with transmission rates ranging from 13-67% after 16-23 d of incubation. Transmission rates for the laboratory strain An. quadrimaculatus SAVANNAH ranged from 20 to 33% after 7-14 d of incubation. Our data show that although An. quadrimaculatus is more susceptible to CV infections than Cq. perturbans, both mosquito species could be involved in the midwestern transmission cycle of the virus.

Variation in the efficiency of vertical transmission of dengue-1 virus by strains of Aedes albopictus (Diptera: Culicidae).

Five geographical strains of Aedes albopictus (Skuse) were compared for their ability to transmit vertically a dengue-1 isolate from Jamaica. The OAHU strain of Ae. albopictus and a strain of Aedes aegypti (L.) from the United States were included as controls. The offspring of orally infected females were assayed individually for vertical infection. Vertical transmission rates among strains ranged from 11 to 41%, and filial infection rates of strains ranged from 0.5 to 2.9%. Filial infection rates of individual positive families within strains ranged from 1.4 to 17.4%. These rates were higher than those previously recorded for Ae. albopictus. The observed differences in rates of vertical transmission among the strains were not statistically significant, because 95% of the measured variation was attributed to families within strains. The most significant source of variation in vertical transmission of dengue-1 by Ae. albopictus was at the individual level.

Vector competence of Aedes hendersoni (Diptera: Culicidae) for La Crosse virus: lack of impaired function in virus-infected salivary glands and enhanced virus transmission by sporozoite-infected mosquitoes.

Previous research has shown Aedes hendersoni Cockerell to be an incompetent vector of La Crosse (LAC) virus because of a salivary gland escape (SGE) barrier; that is, the salivary glands are infected but the mosquito fails to transmit the virus orally. Intradermal probing behavior and ability to locate blood were studied in infected mosquitoes as indicators of salivary gland impairment to determine if the SGE barrier was due to virus-induced pathology of the salivary glands. No evidence of salivary gland impairment as a result of virus infection was detected in infected Ae. hendersoni. This was also true for Aedes triseriatus (Say), a competent vector of LAC virus, which was used as a control. However, coinfection of Ae. hendersoni with Plasmodium gallinaceum and LAC virus dramatically increased virus transmission (72 versus 8%), whereas transmission by coinfected Ae. triseriatus was not significantly affected. Possible causes for the SGE barrier in Ae. hendersoni are discussed.

Aedes triseriatus (Diptera: Culicidae) and La Crosse virus. IV. Nutritional deprivation of larvae affects the adult barriers to infection and transmission.

Groups of Aedes triseriatus (Say) were reared either as nutritionally deprived (two regimens) or well fed (one regimen) throughout larval development, and the vector competence of resulting small, normal, and large females was assessed for La Crosse virus. When fed a high dose of virus (4.6 log10/0.025 ml in Vero cell culture), 90% of small Ae. triseriatus females transmitted La Crosse virus to suckling mice compared with 70% of normal and 42% of large females. Among small females, 100% had disseminated infections as did 86% of normal females, whereas only 69% of large females had disseminated infections. All females had infected mesenterons (midguts). When fed a low dose of virus (2.2 log10/0.025 ml in Vero cell culture) in a second experiment, 15% of small females transmitted compared with 0% of large females; 50% of small females developed disseminated infections compared with 16% of large females. mesenteronal infection occurred in 70% of small but only 32% of large females. Electron microscopy of mesenteronal tissues from large and small females revealed physical differences in the basement membranes (basal laminae). The mesenterons of small females had 3-6 laminae (mean thickness of the basement membrane = 0.14 microns) compared with 9-16 laminae (mean thickness of the basement membrane = 0.24 microns) in large females. These morphological differences indicated that the mesenteronal escape barrier, which accounted for the difference in the percentage of small and large females with disseminated infections, may be, in part, a physical barrier that was modified by nutritional deprivation in the larval instars.

Serological evidence of California group and Cache Valley virus infection in Minnesota white-tailed deer.

Blood samples were obtained from 138 white-tailed deer (Odocoileus virginianus) harvested at three sites surrounding the greater Minneapolis-St. Paul, Minnesota, metropolitan area (USA) and tested for neutralizing antibody to Cache Valley virus and three California serogroup (Jamestown Canyon, La Crosse, trivittatus) viruses (Bunyaviridae). Deer at each site had neutralizing antibody to one or more California serogroup viruses and/or Cache Valley virus. The majority of adult deer (85%) had antibody to both a California serogroup virus and Cache Valley virus. Antibody prevalence varied significantly with age of the deer. Fawns had a significantly lower prevalence of antibody to either a California serogroup (17%) or Cache Valley virus (39%) than did older (greater than 1-yr-old) deer (89% for a California serogroup virus and 91% for Cache Valley virus). The geometric mean titers of antibody in fawns to California serogroup (1:6) and Cache Valley viruses (1:17) were also less than that seen in older animals (1:11 and 1:28 for California serogroup and Cache Valley viruses, respectively). Of 76 older deer with antibody to the California serogroup, 91% had antibody specific for Jamestown Canyon virus. Jamestown Canyon is the primary California serogroup virus circulating in the suburban/rural Minneapolis-St. Paul area. Transmission occurs in an enzootic pattern similar to that documented in Indiana and Michigan. Cache Valley virus also appears to be enzootically transmitted in this area. However, the impact on domestic or wild animal populations is unknown.

Laboratory transmission of Jamestown Canyon and snowshoe hare viruses (Bunyaviridae: California serogroup) by several species of mosquitoes.

The ability of 14 species of mosquitoes to biologically transmit Jamestown Canyon virus was tested. Four species not previously described as vectors of that virus transmitted to suckling mice. Among membrane-fed mosquitoes with disseminated infections, field-collected Aedes canadensis (1/3), Anopheles punctipennis (1/12), Coquillettidia perturbans (2/14) and a laboratory strain of Ae. epactius (19/67) transmitted virus. Two species were tested for their ability to transmit snowshoe hare virus: field-collected Ae. provocans (4/20) and Ae. abserratus-punctor (2/20) successfully transmitted to suckling mice. Evidence regarding the role of these species as field vectors is summarized.

Isolation of Jamestown Canyon virus (California serogroup) from Aedes mosquitoes in an enzootic focus in Michigan.

Twenty isolates of Jamestown Canyon virus were obtained from adult females of 5 Aedes species collected at the Houghton Lake Wildlife Research Area, Missaukee County, in north-central Michigan between 1985 and 1989. Fourteen were from Aedes provocans, and 6 were from 4 other snowmelt Aedes species. One isolate of trivittatus virus and one Cache Valley-like virus were also obtained. Seasonal succession patterns for numerous mosquito species were recorded over 4 years. The temporal association of adult mosquito emergence, virus isolations, and infection and seroconversion of sentinel deer suggest that Ae. provocans is a primary enzootic vector of Jamestown Canyon virus in that focus. We hypothesize that Ae. provocans provides an overwintering reservoir for Jamestown Canyon virus at the study site. A large dry ice-baited "tent trap" was the most productive method for collecting numerous aedine and other mosquito species.

Replication and dissemination of La Crosse virus in the competent vector Aedes triseriatus and the incompetent vector Aedes hendersoni and evidence for transovarial transmission by Aedes hendersoni (Diptera: Culicidae).

The time course and pattern of the replication and dissemination of La Crosse virus was studied in orally infected Aedes triseriatus (Say) and Ae. hendersoni Cockerell. Development of La Crosse virus was approximately the same in both species when plaque assay titers of intact mosquitos or dissected tissues were compared. The mosquitoes were equally susceptible to infection; all Ae. hendersoni and 99% of the Ae. triseriatus tested showed detectable midgut infections. Virus was first detected in hemolymph, salivary glands, and ovaries 10-13 d after infection in both species. The pattern of infection suggests virus dissemination beyond the midgut to be via the hemolymph. By 21 d after infection, 100% (10 of 10) of Ae. triseriatus and 70% (7 of 10) of Ae. hendersoni had infected salivary glands, and the geometric mean titer of Ae. hendersoni salivary glands was 10 times higher than the geometric mean titer of those of Ae. triseriatus, However, when tested for transmission 22 d after infection by refeeding on suckling mice, only 9% (2 of 22) of the Ae. hendersoni with disseminated infections transmitted virus versus 71% (12 of 17) of the Ae. triseriatus. A salivary gland escape barrier was shown to be primarily responsible for the failure of Ae. hendersoni to orally transmit La Crosse virus. However, eight parenterally infected Ae. hendersoni females transovarially transmitted the virus to 25% (5 of 20) of their progeny.

Recently introduced Aedes albopictus in the United States: potential vector of La Crosse virus (Bunyaviridae: California serogroup).

A population of Aedes albopictus collected in 1986 in Harris County, Texas, was evaluated for its vector competence with 4 California serogroup viruses (Jamestown Canyon, Keystone, La Crosse and trivittatus). Rates of midgut infection, dissemination of virus beyond the midgut and oral transmission to suckling mice were markedly different for the 4 viruses in a pattern representative of the antigenic relationships known for the California serogroup. Only La Crosse virus was shown to be efficiently transmitted by this recently introduced mosquito population. The results suggest that populations of Ae. albopictus originating from the Harris County population might well be as efficient in transmitting La Crosse virus as are populations of the natural mosquito vector. Aedes triseriatus, from the midwestern La Crosse virus enzootic region. The public health implications of these results are discussed in relation to the rapid spread of Ae. albopictus throughout the eastern half of the United States and into regions where La Crosse virus is known to be enzootic.

Midgut and salivary gland barriers to La Crosse virus dissemination in mosquitoes of the Aedes triseriatus group.

Vector competence for La Crosse virus (LACV) was compared for four species in the Aedes (Protomacleaya) triseriatus group: Ae. triseriatus (Say), Ae. hendersoni Cockerell, Ae. zoosophus Dyar and Knab and Ae. brelandi Zavortink (Diptera: Culicidae). Rates of replication and dissemination of virus in the mosquito hosts were compared and rates of oral transmission of virus to suckling mice were determined. Barriers to virus dissemination which limited the ability of each species to transmit virus were identified. Ae. zoosophus displayed the highest vector competence for LACV. Both infection and transmission rates were high: 99% and 85% respectively; no significant barriers to LACV were found. Disseminated infection of Ae. triseriatus with LACV was controlled primarily be a midgut escape barrier. When virus was introduced directly into the haemocoel, transmission rates were significantly increased (37% v. 79%). Ae. hendersoni showed high susceptibility to LACV infection but a very low rate of oral transmission (7%). Ae. brelandi was also highly susceptible to infection by LACV and transmitted virus at an intermediate rate (27%). Modulation of vector competence in both Ae. hendersoni and Ae. brelandi resulted from a salivary gland escape barrier. As these four species of mosquitoes comprise a closely related monophyletic series, their differences of vector competence for LACV provide an excellent model for studying the genetic basis of the barriers involved.

Variation in the vector competence of geographic strains of Aedes albopictus for dengue 1 virus.

Eight geographic strains of Aedes albopictus from Asia and North America and one North American strain of Aedes aegypti were tested for their vector competence with dengue 1 virus. Three groups of Ae. albopictus were established based on their vector competence: a) the OAHU laboratory strain, b) the three Malaysian strains, and c) the TOKYO and three North American strains. The three North American strains were similar to the strain of Ae. aegypti from Houston, Texas in their ability to transmit dengue 1 virus. A comparison of barriers to infection and transmission suggests that Ae. albopictus HOUSTON represents an introduced strain distinct from the more similar MEMPHIS and NEW ORLEANS strains. Based on these studies the North American strains were seen as more similar to a northern Asian strain (TOKYO) than to the three Malaysian (southern Asia) strains, supporting the current hypothesis that the indigenous strains of Ae. albopictus recently introduced into the United States had a northern Asian origin.

Production and use of a hemagglutinin for detecting antibody to Jamestown Canyon virus.

A procedure was developed for producing a hemagglutinin for the California serogroup (family Bunyaviridae, genus Bunyavirus) virus Jamestown Canyon, a human pathogen. Serum samples from humans putatively infected with this virus or with La Crosse virus were tested by hemagglutination inhibition. Each antigen detected antibody to the respective virus, with little cross-reactivity. These results suggest that both antigens should be used when the hemagglutination inhibition test is applied to the diagnosis of human infections with California serogroup viruses in North America.

Seroconversion rates to Jamestown Canyon virus among six populations of white-tailed deer (Odocoileus virginianus) in Indiana.

The annual seroconversion of fawns, yearlings, and adult white-tailed deer (Odocoileus virginianus) to Jamestown Canyon virus (California group) was followed at six Indiana sites from 1981 through 1984. In all, sera from 1,642 deer (515 fawns, 618 yearlings, and 509 adults) were tested for neutralizing antibody to three California serogroup viruses: Jamestown Canyon, La Crosse, and trivittatus. Virtually all deer with specific neutralizing antibody showed evidence of a prior infection with Jamestown Canyon virus; only three deer showed evidence of a prior infection with only La Crosse virus and none showed evidence of an infection with only trivittatus virus. While there were no significant differences in antibody prevalence to Jamestown Canyon virus between yearling and adult deer at any site, fawns had significantly lower antibody prevalences than either of the two older age groups. Significant differences in antibody prevalence were found between northern versus southern populations of white-tailed deer in Indiana, however, no significant differences were found among the four northern populations or between the two southern populations. The mean antibody prevalences in the two southern fawn, yearling, and adult populations were 15%, 38%, and 41% respectively, while the prevalences in the four northern fawn, yearling, and adult populations were 5%, 67%, and 67% respectively. These different prevalences (northern vs. southern) correlate with the higher Jamestown Canyon virus antibody prevalence in human residents of northern Indiana (2-15%) compared to residents of southern Indiana (less than 2%) found in other studies. The significantly lower prevalence of antibody to Jamestown Canyon virus in fawns is attributed to maternal antibody protecting them from a primary infection their first summer. Yearling deer showed high rates of seroconversion following their second summer of life. These results suggest that infection of white-tailed deer in Indiana with Jamestown Canyon virus is a common phenomenon.