Fluid Spatial Dynamics of West Nile Virus in the United States: Rapid Spread in a Permissive Host Environment.

UNLABELLED: The introduction of West Nile virus (WNV) into North America in 1999 is a classic example of viral emergence in a new environment, with its subsequent dispersion across the continent having a major impact on local bird populations. Despite the importance of this epizootic, the pattern, dynamics, and determinants of WNV spread in its natural hosts remain uncertain. In particular, it is unclear whether the virus encountered major barriers to transmission, or spread in an unconstrained manner, and if specific viral lineages were favored over others indicative of intrinsic differences in fitness. To address these key questions in WNV evolution and ecology, we sequenced the complete genomes of approximately 300 avian isolates sampled across the United States between 2001 and 2012. Phylogenetic analysis revealed a relatively star-like tree structure, indicative of explosive viral spread in the United States, although with some replacement of viral genotypes through time. These data are striking in that viral sequences exhibit relatively limited clustering according to geographic region, particularly for those viruses sampled from birds, and no strong phylogenetic association with well-sampled avian species. The genome sequence data analyzed here also contain relatively little evidence for adaptive evolution, particularly of structural proteins, suggesting that most viral lineages are of similar fitness and that WNV is well adapted to the ecology of mosquito vectors and diverse avian hosts in the United States. In sum, the molecular evolution of WNV in North America depicts a largely unfettered expansion within a permissive host and geographic population with little evidence of major adaptive barriers. IMPORTANCE: How viruses spread in new host and geographic environments is central to understanding the emergence and evolution of novel infectious diseases and for predicting their likely impact. The emergence of the vector-borne West Nile virus (WNV) in North America in 1999 represents a classic example of this process. Using approximately 300 new viral genomes sampled from wild birds, we show that WNV experienced an explosive spread with little geographical or host constraints within birds and relatively low levels of adaptive evolution. From its introduction into the state of New York, WNV spread across the United States, reaching California and Florida within 4 years, a migration that is clearly reflected in our genomic sequence data, and with a general absence of distinct geographical clusters of bird viruses. However, some geographically distinct viral lineages were found to circulate in mosquitoes, likely reflecting their limited long-distance movement compared to avian species.

Avian botulism and avian chlamydiosis in wild water birds, Benton Lake National Wildlife Refuge, Montana, USA.

In 1999, the U.S. Geological Survey (USGS) National Wildlife Health Center, Madison, Wisconsin, conducted a diagnostic investigation into a water bird mortality event involving intoxication with avian botulism type C and infection with avian chlamydiosis at the Benton Lake National Wildlife Refuge in Montana, USA. Of 24 carcasses necropsied, 11 had lesions consistent with avian chlamydiosis, including two that tested positive for infectious Chlamydophila psittaci, and 12 were positive for avian botulism type C. One bird tested positive for both avian botulism type C and C. psittaci. Of 61 apparently healthy water birds sampled and released, 13 had serologic evidence of C. psittaci infection and 7 were, at the time of capture, shedding infectious C. psittaci via the cloacal or oropharyngeal route. Since more routinely diagnosed disease conditions may mask avian chlamydiosis, these findings support the need for a comprehensive diagnostic investigation when determining the cause of a wildlife mortality event.

Pathogenicity of West Nile virus and response to vaccination in sandhill cranes (Grus canadensis) using a killed vaccine.

West Nile virus was introduced into the United States in the vicinity of New York, New York, USA in 1999. The virus has since killed large numbers of birds nationwide, especially, but not limited to, crows (Corvus brachyrhinchos). One sandhill crane (Grus canadensis) at the Bridgeport Zoo (Bridgeport, Connecticut, USA) reportedly died from West Nile virus, so sandhill cranes and endangered whooping cranes (Grus americana), both in the wild and in captive breeding colonies at United States Geological Service (USGS) Patuxent Wildlife Research Center (Laurel, Maryland, USA) were considered at risk. A killed vaccine in sandhill cranes was evaluated by vaccinating and then challenging these cranes with live West Nile virus. No sandhill cranes inoculated with the killed vaccine developed significant titers when compared with unvaccinated controls. No sandhill cranes inoculated with the vaccine and challenged with the virus died from West Nile virus infection. In addition, no unvaccinated challenged sandhill cranes died. However, 2 days postchallenge, vaccinated cranes had significantly less viremia (P < 0.05) than unvaccinated cranes. Seven days postchallenge vaccinated cranes had significantly less cloacal shedding of the virus (P < 0.05) than unvaccinated cranes and significantly less weight loss (P < 0.05) as compared with unvaccinated cranes. Vaccinated sandhill cranes developed significantly higher titers 14 days postchallenge and were viremic for shorter periods of time after challenge than unvaccinated individuals. Unvaccinated challenged cranes had glial cell aggregates in both the brain and brain stem areas, and this was not observed in vaccinated challenged cranes or in vaccinated unchallenged cranes.

Raptor mortality due to West Nile virus in the United States, 2002.

West Nile virus (WNV) has affected many thousands of birds since it was first detected in North America in 1999, but the overall impact on wild bird populations is unknown. In mid-August 2002, wildlife rehabilitators and local wildlife officials from multiple states began reporting increasing numbers of sick and dying raptors, mostly red-tailed hawks (Buteo jamaicensis) and great horned owls (Bubo virginianus). Commonly reported clinical signs were nonspecific and included emaciation, lethargy, weakness, inability to perch, fly or stand, and nonresponse to danger. Raptor carcasses from 12 states were received, and diagnostic evaluation of 56 raptors implicated WNV infection in 40 (71%) of these cases. Histologically, nonsuppurative encephalitis and myocarditis were the salient lesions (79% and 61%, respectively). Other causes of death included lead poisoning, trauma, aspergillosis, and Salmonella spp. and Clostridium spp. infections. The reason(s) for the reported increase in raptor mortality due to WNV in 2002 compared with the previous WNV seasons is unclear, and a better understanding of the epizootiology and pathogenesis of the virus in raptor populations is needed.

Isolation and characterization of a reovirus from common eiders (Somateria mollissima) from Finland.

Samples of brain, intestine, liver, lung, spleen, and bursa of Fabricius were collected from five common eider (Somateria mollissima) duckling carcasses during a die-off in the western Gulf of Finland (59 degrees 50'N, 23 degrees 15'E) in June 1996. No viral activity was observed in specific-pathogen-free chicken embryos inoculated with tissue suspensions, but samples of bursa of Fabricius from three birds were positive when inoculated into Muscovy duck (Cairina moschata) embryo fibroblasts. The isolates were characterized as nonenveloped RNA viruses and possessed several characteristics of the genus Orthoreovirus. Virus particles were icosahedral with a mean diameter of 72 nm and were stable at pH 3.0; their genome was separated into 10 segments by polyacrylamide gel electrophoresis. Mallard (Anas platyrhynchos) ducklings experimentally infected with the eider reovirus showed elevated serum activities of aspartate aminotransferase, creatine kinase, and lactate dehydrogenase enzymes and focal hemorrhages in the liver, spleen, and bursa of Fabricius. During 1997-99, the prevalence of neutralizing antibodies to the isolated virus ranged from 0 to 86% in 302 serum samples collected from incubating eider hens at three nesting areas along coastal Finland. The highest seroprevalence was found in Hanko in 1999, just weeks before reports of an uninvestigated mortality event resulting in the death of an estimated 98% of ducklings at that location. These findings raise the question of potential involvement of the virus in poor duckling survival and eider population declines observed in several breeding areas along coastal Finland since the mid-1980s.

An adenovirus linked to mortality and disease in long-tailed ducks (Clangula hyemalis) in Alaska.

An adenovirus was isolated from intestinal samples of two long-tailed ducks (Clangula hyemalis) collected during a die-off in the Beaufort Sea off the north coast of Alaska in 2000. The virus was not neutralized by reference antiserum against known group I, II, or III avian adenoviruses and may represent a new serotype. The prevalence of the virus was determined in live-trapped long-tailed ducks at the mortality site and at a reference site 100 km away where no mortality was observed. Prevalence of adenovirus antibodies in serum samples at the mortality site was 86% compared to 10% at the reference site. Furthermore, 50% of cloacal swabs collected at the mortality site and only 7% of swabs from the reference site were positive for adenoviruses. In 2001, no mortality was observed at either of the study areas, and virus prevalence in both serum and cloacal samples was low, providing further evidence that the adenovirus was linked to the mortality event in 2000. The virus was used to infect long-tailed ducks under experimental conditions and resulted in lesions previously described for avian adenovirus infections and similar to those observed in long-tailed duck carcasses from the Beaufort Sea. The status of long-tailed ducks has recently become a concern in Alaska due to precipitous declines in breeding populations there since the mid-1970s. Our findings suggest that the newly isolated adenovirus is a disease agent and source of mortality in long-tailed ducks, and thus could be a contributing factor in population declines.

An adenovirus associated with intestinal impaction and mortality of male common eiders (Somateria mollissima) in the Baltic Sea.

We examined 10 common eider (Somateria mollissima) males found dead in 1998 during a die-off in the northern Baltic Sea off the southwestern coast of Finland. We diagnosed impaction of the posterior small intestine with mucosal necrosis as the cause of death in all 10 and isolated adenoviruses from cloacal samples of six birds. The adenovirus isolates were not neutralized by reference antisera to group I, II, or III avian adenoviruses. Cloacal swabs from 22 apparently healthy eider females nesting at the mortality area were negative for viruses. An adenovirus isolated from one of the eiders caused clinical signs of illness and gastrointestinal pathology in experimentally infected mallard (Anas platyrhynchos) ducklings. These findings suggest that the adenovirus contributed to the mortality of common eider males in the Finnish archipelago.

Diagnostic and molecular evaluation of three iridovirus-associated salamander mortality events.

In 1998 viruses were isolated from tiger salamander larvae (Ambystoma tigrinum diaboli and A. tigrinum melanostictum) involved in North Dakota and Utah (USA) mortality events and spotted salamander (A. maculatum) larvae in a third event in Maine (USA). Although sympatric caudates and anurans were present at all three sites only ambystomid larvae appeared to be affected. Mortality at the North Dakota site was in the thousands while at the Utah and Maine sites mortality was in the hundreds. Sick larvae were lethargic and slow moving. They swam in circles with obvious buoyancy problems and were unable to remain upright. On the ventral surface, near the gills and hind limbs, red spots or swollen areas were noted. Necropsy findings included: hemorrhages and ulceration of the skin, subcutaneous and intramuscular edema, swollen and pale livers with multifocal hemorrhage, and distended fluid-filled intestines with areas of hemorrhage. Light microscopy revealed intracytoplasmic inclusions, suggestive of a viral infection, in a variety of organs. Electron microscopy of ultra thin sections of the same tissues revealed iridovirus-like particles within the inclusions. These viruses were isolated from a variety of organs, indicating a systemic infection. Representative viral isolates from the three mortality events were characterized using molecular assays. Characterization confirmed that the viral isolates were iridoviruses and that the two tiger salamander isolates were similar and could be distinguished from the spotted salamander isolate. The spotted salamander isolate was similar to frog virus 3, the type species of the genus Ranavirus, while the tiger salamander isolates were not. These data indicate that different species of salamanders can become infected and die in association with different iridoviruses. Challenge assays are required to determine the fish and amphibian host range of these isolates and to assess the susceptibility of tiger and spotted salamanders to heterologous virus isolates.

Changes in West Nile virus seroprevalence and antibody titers among Wisconsin mesopredators 2003-2006.

After the 2001 occurrence of West Nile virus (WNV) in Wisconsin (WI), we collected sera, during 2003-2006, from south-central WI mesopredators. We tested these sera to determine WNV antibody prevalence and geometric mean antibody titer (GMAT). Four-fold higher antibody prevalence and 2-fold higher GMAT in 2003-2004 indicated greater exposure of mesopredators to WNV during the apparent epizootic phase. The period 2005-2006 was likely the enzootic phase because WNV antibody prevalence fell to a level similar to other flaviviruses. Our results suggest that, in mesopredators, vector-borne transmission is the primary route of infection and WNV antibodies persist for < 1 year. Mesopredators may be sensitive indicators of West Nile virus spill-over into humans and horses. Mesopredator sero-surveys may complement dead crow surveillance by providing additional data for the timing of public health interventions. Research is needed to clarify the dynamics of WNV infection in these mammals and their role as potential WNV amplifiers.

Corvidae feather pulp and West Nile virus detection.

We evaluated cloacal swab, vascular pulp of flight feather, and kidney and spleen pool samples from carcasses of members of the family Corvidae as sources of West Nile virus (WNV). The cloacal swab, kidney and spleen pool, and feather pulp were the source of WNV in 38%, 43%, and 77%, respectively, of the carcasses.