Jamestown Canyon virus comes into view: understanding the threat from an underrecognized arbovirus

Abstract

This review examines the epidemiology, ecology, and evolution of Jamestown Canyon virus (JCV) and highlights new findings from the literature to better understand the virus, the vectors driving its transmission, and its emergence as an agent of arboviral disease. We also reanalyze data from the Connecticut Arbovirus Surveillance Program which represents the largest dataset on JCV infection in mosquitoes. JCV is a member of the California serogroup of the genus Orthobunyavirus, family Peribunyaviridae, and is found throughout much of temperate North America. This segmented, negative-sense RNA virus evolves predominately by genetic drift punctuated by infrequent episodes of genetic reassortment among novel strains. It frequently infects humans within affected communities and occasionally causes febrile illness and neuroinvasive disease in people. Reported human cases are relatively rare but are on the rise during the last 20 yr, particularly within the northcentral and northeastern United States. JCV appears to overwinter and reemerge each season by transovarial or vertical transmission involving univoltine Aedes (Diptera: Culicidae) species, specifically members of the Aedes communis (de Geer) and Ae. stimulans (Walker) Groups. The virus is further amplified in a mosquito-deer transmission cycle involving a diversity of mammalophilic mosquito species. Despite progress in our understanding of this virus, many aspects of the vector biology, virology, and human disease remain poorly understood. Remaining questions and future directions of research are discussed.

Introduction

Jamestown Canyon virus (JCV) remains largely unknown to the general public and medical community, despite it being one of the most prevalent and widespread arboviruses in North America. The virus is transmitted by mosquitoes (Diptera: Culicidae) in temperate regions of the United States and Canada where it frequently infects humans (Grimstad 2001). Most human infections are asymptomatic or mild, but the virus can occasionally cause life-threatening neuroinvasive disease (Deibel et al. 1983). Although reported human cases are relatively rare in the U.S. (average 12 cases/year 2000–2022), they are on the rise with human cases concentrated in the midwestern and northeastern regions of the United States (Fig. 1). The factors responsible for this trend are not entirely clear but may relate to conditions of increased virus transmission and environmental risk, increased physician awareness and diagnostic testing, or some combination of these and other factors. JCV is maintained in a deer–mosquito transmission cycle, and the risk of infection may be increasing with the population expansion of deer in many parts of the country (Grimstad 1988). The proliferation of white-tailed deer (Odocoileus virginianus), in particular, has played a pivotal role in the emergence of tick-borne diseases such as Lyme disease in the eastern United States and may have a similar impact on JCV transmission (Barbour and Fish 1993). In addition, the development and adoption of new serological tests by state agencies and the CDC may have uncovered more human cases due to increased testing (Pastula et al. 2015, Matkovic et al. 2019).

In this review, we examine the epidemiology, evolution, and ecology of JCV to better understand this virus, the vectors driving its transmission, and its emergence as an agent of arboviral disease. We also analyze 26 yr of mosquito data from Connecticut (CT) as a case study to understand spatial-temporal patterns of virus activity and to identify the major vectors responsible for JCV perpetuation and spread. This analysis serves as a framework for identifying vectors and transmission patterns in other geographic regions where entomological data is sparse and the transmission cycle of JCV is poorly understood. In addition, remaining questions and future directions of research will be discussed.

Classification and Viral Evolution

Jamestown Canyon virus was first isolated from Culiseta inornata (Williston) mosquitoes collected in Colorado in 1961 and identified as a member of the California serogroup of the genus Orthobunyavirus, family Peribunyaviridae based on serological criteria and later by genetic identity (Sather and Hammon 1967, Bowen et al. 1995, Campbell and Huang 1999). The Orthobunyavirus genus includes more than 100 described species that are transmitted by either mosquitoes or biting midges and includes a number of other important human and livestock pathogens such as La Crosse virus (LACV), Oropouche virus, Cache Valley virus, and Schmallenberg virus (Karabatsos 1985). The taxonomy of bunyaviruses was recently revised by the by the International Committee on Taxonomy of Viruses by applying specific genetic criteria for designating species (Hughes et al. 2020). This has resulted in a consolidation of recognized species and a number of closely related viruses are considered strains of JCV. South River virus from New Jersey, Jerry Slough virus (JSV) from California, and Inkoo virus (INKV) from Finland are all considered strains of JCV rather than distinct species. For the purposes of this review, we will focus our discussion on JCV strains that circulate in North America where it is recognized as an emerging human pathogen.

As a member of the family Peribunyaviridae, JCV is an enveloped virus that possesses 3 distinct negative-sense RNA segments designated as S (small), M (medium), and L (large) (Elliott 2014). The S segment is highly conserved and encodes for the nucleocapsid and nonstructural S protein in overlapping reading frames. The M segment encodes 2 envelope glycoproteins (Gn and Gc) and a nonstructural M protein in a continuous open reading frame, while the L segment encodes a single gene-RNA polymerase (Elliott and Blakqori 2011). Given the segmented nature of the JCV genome, there are 2 main mechanisms for generating genetic diversity: genetic drift and genetic shift (Briese et al. 2013). Genetic drift describes the accumulation of point mutations as the virus replicates due to error-prone nature of their RNA polymerases. Genetic drift generates incremental changes at the molecular level and is occurring constantly as the virus replicates. Genetic shift or segment reassortment, in contrast, occurs when 2 related viruses infect the same cell which allows them to exchange whole genetic segments to give rise to hybrid viruses. This mechanism produces rapid changes in virus genotype possibly with novel phenotypes and is expected to occur rarely when 2 different viruses coinfect the same cell.

Segment reassortment has been documented for many bunyaviruses both experimentally in the lab and by phylogenetic analysis of field isolates. In a study by Cheng et al. (1999), they demonstrated that mosquitoes were receptive hosts for segment reassortment between JCV and LACV. Both viruses successfully coinfected Aedes albopictus (Skuse) and formed all 6 possible reassortant genotypes. These hybrid viruses orally infected Ae. albopictus and were subsequently transmitted to suckling mice by mosquito bite. Reassortant viruses carrying the M segment from LACV against a JCV genetic background were more neurovirulent to mice than JCV or any of the other reassortment genotypes. In addition, these isolates could infect gerbils and Aedes triseriatus (Say), which are characteristics of LACV but not JCV, demonstrating the potential for these viruses to acquire new properties through reassortment.

Evidence for reassortment among field isolates may be inferred by comparing the evolutionary histories of each genome segment. To characterize the molecular evolution of JCV and determine the extent of reassortment among natural isolates, Armstrong and Andreadis (2007) compared the phylogenetic relationships of 56 JCV strains isolated from mosquitoes collected in CT over a 40-yr period. Their analysis revealed at least 2 major genetic lineages circulating in different regions of the state but with some overlap in their distribution. One of these lineages was detected in CT for over 40 yr with minimal nucleotide changes accumulating over time. They found no evidence for segment reassortment among the 2 major lineages circulating in CT; however, the 1961 prototype strain from Colorado carried the genetic signature of hybridization between JCV and the related strain JSV. Another study by Hughes et al. (2017) also suggested occasional episodes of reassortment among California serogroup viruses including among JCV and INKV based on incongruent branching patterns among different segment trees. Together, these studies suggest that JCV evolves predominately by genetic drift punctuated by infrequent episodes of reassortment among novel strains. Subsequent phylogenetic analyses of JCV strains from North Dakota and Maine found that isolates from these regions clustered with lineage A from CT (Anderson et al. 2015, Schneider et al. 2022). However, these studies analyzed only S segment sequences and therefore, could not identify reassortant genotypes.

Disease Patterns

In North America, reported cases of JCV infection occur annually and are most commonly reported from the midwestern and northeastern United States. Since 2000, human cases were reported in 25 states with more than half in 2 states: Minnesota and Wisconsin (Fig. 1). This observation may be driven, at least in part, by increased surveillance and testing within these states (Matkovic et al. 2019). A second focal area includes the northeastern United States with the most cases reported from Massachusetts, New Hampshire, and New York. During the last 10 yr, human cases are also appearing in the southeastern and western regions of the United States for the first time. The precipitous rise in cases since 2013 likely reflects changes in serological testing procedures adopted by the CDC. Starting in 2013, the CDC added JCV to their panel for antibody testing against domestic arboviruses (Pastula et al. 2015). Clinical cases are usually identified by detecting JCV IgM antibodies by ELISA and confirmed by more specific plaque reduction neutralization tests (Piantadosi and Kanjilal 2020). In addition, direct detection of viral RNA by PCR may be used in certain instances (Huang et al. 1999). Although increased testing is revealing a higher case burden than previously recognized, only a few public health laboratories are currently capable of performing these tests. Underdiagnosis of JCV cases is also possible due to lack of physician awareness, deficient public health outreach and education, nonspecific symptoms, and absence of commercial-available diagnostic testing.

Clinically, human infections may present as a nonspecific febrile illness to severe neuroinvasive disease including meningitis and encephalitis. Some case reports also note respiratory involvement which is not typical for arboviruses (Grimstad 1988). Fatalities are rare with several cases documented in recent years, of which many were immunocompromised or had other underlying health conditions (Curren et al. 2018, Savard et al. 2018, VanderVeen et al. 2020, Solomon et al. 2021). Analysis of national cases from 2003 to 2013 revealed that 48% of cases were hospitalized and the most common symptoms were encephalitis and/or meningitis (54%) (Pastula et al. 2015). Most human cases occurred between May and September which corresponds to the mosquito season. The median age was 48 yr (range 10–69) and 68% were male. Another analysis of 30 cases of JCV infection from Wisconsin showed similar patterns with 50% of cases showing neuroinvasive disease, 56% were male, and a median age of 54 yr (Matkovic et al. 2019). JCV appears to affect all age groups unlike many other arboviruses that disproportionately impact children (LACV) or older age groups (West Nile virus [WNV]).

Reported human cases are highly skewed towards the most severe JCV infections that are referred for diagnostic testing. Underlying symptomatic cases is a much larger population of exposed individuals as revealed by serosurveys of human populations. Seropositivity to JCV among humans was reported at 28% in Michigan (Grimstad et al. 1986a), 4–10% in CT (Mayo et al. 2001), 18% in Alaska (Walters et al. 1999), 9–24% in Quebec (Sampasa-Kanyinga et al. 2013), 21% in Nova Scotia (Patriquin et al. 2018), and 27% in New Brunswick (Mincer et al. 2021). Antibodies increased with age in studies that analyzed this variable, suggesting long-term endemic transmission of JCV within these communities (Walters et al. 1999, Patriquin et al. 2018, Mincer et al. 2021). Seropositivity was statistically higher in males in 3 surveys (Grimstad et al. 1986a, Patriquin et al. 2018, Mincer et al. 2021) but there was no difference among sexes in native populations from Alaska (Walters et al. 1999). Although reported clinical cases are rare, serological evidence suggests that residents are frequently exposed to JCV and that most infections are mild or asymptomatic.

Vectors and Transmission Cycles

Jamestown Canyon virus is widely distributed throughout North America and throughout its range, the principal vectors and amplification hosts will vary by geographic region. Nevertheless, some common trends can be discerned. (i) Single generation (univoltine) Aedes mosquitoes serve as important vectors and overwintering hosts in many endemic regions (Grimstad 2001). These species can maintain the virus in overwintering eggs via transovarial transmission and are involved in early season amplification of the virus (Boromisa and Grimstad 1986, Boromisa and Grayson 1990, Campbell et al. 1991, Hardy et al. 1993, Farquhar et al. 2022). (ii) JCV infects a diverse array of mosquito species and does not rely on any 1 mosquito species to maintain transmission throughout the summer (Grayson and Calisher 1983, Grimstad 2001, Andreadis et al. 2008, McMillan et al. 2020). Local transmission of JCV appears to be driven by the involvement of multiple mammalian-biting mosquito species that have overlapping seasonal activity to extend the length of the transmission season. This contrasts with arboviruses such as WNV or eastern equine encephalitis virus (EEEV) which involve 1 or 2 key species that are responsible for local enzootic transmission (Rochlin et al. 2019, Armstrong and Andreadis 2022). (iii) Deer appear to serve as the main amplification hosts in regions where human cases are prevalent (Grimstad 2001). The virus infects a variety of free ranging ungulates including bison, moose, and elk (Zarnke et al. 1983, Grimstad et al. 1986b); however, white-tailed deer are recognized as the principal amplification host based on experimental infections and prevalence of neutralizing antibodies (Watts et al. 1982, Boromisa and Grimstad 1986, 1987, Grimstad et al. 1987, Murphy 1989, Neitzel and Grimstad 1991, Zamparo et al. 1997, Patriquin et al. 2018, Hollis-Etter et al. 2019, Dupuis et al. 2020). This species thrives in suburban and periurban residential communities in the eastern half of the United States and parts of Canada (Urbanek 2013). Other species such as mule deer and black-tailed deer are also frequently exposed to the virus and may play a similar role in western North America (Eldridge et al. 1987, Campbell et al. 1989).

In the following discussion, we review the evidence supporting the role of various mosquito species in the transmission of JCV among wildlife and human hosts. Identifying the major vectors of JCV is perhaps more complicated given that this virus infects a wide diversity of mosquito species throughout its geographic range. Nevertheless, patterns emerge when interpreting the data in aggregate which leads to incriminating certain species as likely vectors. The most important criteria for incriminating vectors include (i) species that readily feed on the vertebrate host of the pathogen, (ii) spatial-temporal association between suspected vectors and vertebrate infections, (iii) species that are frequently infected with the pathogen in nature, and (iv) demonstration of pathogen transmission by the suspected vector under laboratory conditions. Satisfying all of these criteria represents the highest standard of proof for identifying vectors, but much of this information is incomplete or fragmentary for the potential vectors of JCV.

Connecticut

The most comprehensive dataset on JCV infection in mosquitoes arguably comes from the CT surveillance program. For this paper, we reanalyzed this dataset to update our earlier publications and now include 26 seasons of mosquito data from 1997 to 2022. The surveillance program was originally developed to monitor the risk of EEEV and later expanded to monitor WNV, but JCV was also detected every year under established sampling and testing procedures. Details of collection and testing methods are previously described but briefly, mosquitoes were collected at up to 108 fixed-trapping sites statewide by operating CO₂-baited CDC light traps and gravid traps overnight at each site on a weekly rotation from the beginning of June-late October (Andreadis et al. 2008). All collected mosquitoes were identified to the species level and grouped into pools for virus testing. Processed mosquito pools were inoculated into Vero cell cultures and any isolated viruses were identified by serological or PCR-based assays.

A total of 4,889,926 mosquitoes were collected, identified, and processed in 333,932 mosquito pools for virus testing from 1997 to 2022. This effort yielded 597 JCV isolations from 26 species representing 6 genera (Table 1, Fig. 2). The 22 species that tested negative for the virus are not shown and together comprise 11.2% of total collections. JCV infected mostly mammalian-feeding mosquito species with the highest number of isolates derived from Aedes canadensis (Theobald) (n = 123), Aedes cantator (Coquillett) (n = 83), Anopheles punctipennis (Say) (n = 55), Aedes aurifer (Coquillett) (n = 54), Aedes abserratus (Felt & Young) (n = 53), and Coquillettidia perturbans (Walker) (n = 44). All of these species except Ae. aurifer were previously incriminated as the most likely vectors based on a prior analysis of the CT data from 1997 to 2006 (Andreadis et al. 2008). Aedes canadensis remained the dominant source of JCV isolates with positive mosquitoes collected in more years and trapping locations than any of the other species examined.

Virus infection rates were estimated by the maximum likelihood estimate (MLE) method (Biggerstaff 2009) and ranged from 0.004 to 11.13 per 1,000 mosquitoes (Table 1). MLE’s were generally highest for mammalophilic univoltine species and lowest for ornithophilic multivoltine species. Aedes provocans (Walker) exhibited the highest infection rate but this species was infrequently collected and represented only 3.6% of the total JCV isolations. Together, these findings support a prior analysis showing that JCV was strongly associated univoltine Aedes species by odds ratio analysis of positive mosquito species (McMillan et al. 2020).

Jamestown Canyon virus is the earliest mosquito-borne virus to emerge in CT and typically proceeds WNV and EEEV by 1–2 months. This early season cycle is likely due to the importance of univoltine Aedes species serving as both overwintering hosts of JCV by vertical transmission and early season amplification vectors. The seasonal curve of weekly JCV infections in mosquitoes from 1997 to 2022 is given in Fig. 3. Virus infections peaked in late June and then rapidly declined in late July and August and closely followed the seasonal abundance of mosquito collections in the state. The virus was isolated from a diverse assemblage of mosquito species throughout the transmission season indicating that multiple species participate in local transmission. Mammalophilic univoltine species predominated in June and continued to have a strong presence throughout the month of July. Multivoltine species, which occur throughout the summer, then became dominant sources of infection during the latter part of the transmission season. In the following section, we discuss the role of individual species that were incriminated as important vectors of JCV in CT.

Aedes canadensis accounted for 123 (20.6%) of JCV isolates during 25 of 26 yr of sampling in CT, the highest for any species collected during that period (Table 1). Broadly described as univoltine, this mosquito species typically produces a large generation in the spring, with larvae developing in vernal, leaf-lined woodland pools (Crans 2004). Additional hatching of eggs has been documented following high levels of precipitation in late summer or early fall, but it is undetermined if this late-season population is due to a delayed hatch or progeny from early-season adults. Nonetheless, late season abundance is minor compared to the population present throughout June and early July (Means 1979, Crans 2004, Andreadis et al. 2005). Aedes canadensis was found to feed primarily on white-tailed deer in CT (95% of blood meals identified) but will also feed on humans, medium-sized mammals, and birds (Molaei et al. 2008). Isolates of JCV from Ae. canadensis peaked in mid to late June and decreased through the end July, coincident with longer lived adults. Since this is one of the most common mosquito species collected (14.4% of JCV+ species), the resulting bias-corrected MLE is relatively low (0.20 per 1000). Aedes canadensis may serve as a potential bridge vector of JCV to humans given its propensity to feed on both deer and humans and demonstrated its ability to become infected, disseminate and transmit this virus under laboratory conditions (Watts et al. 1986, Heard et al. 1991).

Aedes aurifer and Ae. abserratus respectively account for 54 (9.0%) and 53 (8.9%) of JCV isolates and 1.6% and 1.4% of all JCV positive species collected. Virus isolations were obtained from Ae. abserratus during 19 of 26 yr of sampling and more occasionally from Ae. aurifer (9 yr). These species are univoltine, developing in woodland pools, with peak adult collections occurring in the first 2 wk of June (Means 1979, Andreadis et al. 2005). Isolations of JCV in CT peak in the last 2 wk of June and are rarely made after mid-July. White-tailed deer were the primary common source of bloodmeals for both species (Molaei et al. 2008). Aedes abserratus was also identified as an important vector species in an earlier arbovirus survey in CT from 1969 to 1978 (Main et al. 1979). The vector competency of a mixed collection of Ae. abserratus and Aedes punctor (Kirby) was previously assessed for JCV. Members of this cohort developed disseminated infection, but virus transmission could not be demonstrated in a limited sample individuals tested (Heard et al. 1991).

Aedes provocans is collected in northern sections of CT which represents the southern edge of its traditional range (Andreadis et al. 2005). Larvae develop in snow melt pools and adults emerge during April and May (Means 1979, Smith and Gadawski 1994). Because Ae. provocans populations are already in decline once surveillance begins in CT in June and their distribution is geographically restricted, this species represents only 0.06% of JCV+ species collected in the state. Nevertheless, there are a relatively high number of isolates from this species (n = 22, 3.69% of total) and the resulting MLE is a robust 11.13 per 1,000 individuals. Most isolates from this species are obtained in the first 2 wk of June during 10 of 26 yr of sampling, with infrequent isolates in the last week of the month. Ae. provocans has been identified as an important vector of JCV based on a high odds ratio of testing positive (McMillan et al. 2020), as well as laboratory and field studies confirming JCV infection, horizontal, or vertical transmission (Boromisa and Grayson 1990, Heard et al. 1990, 1991).

Often grouped with the Aedes stimulans (Walker) Group (Edwards 1932, Eldridge 1990), due to the presence of light scaled bands on the legs, Aedes cantator (Coquillett) is a multivoltine species with larval development in shallow brackish or freshwater pools located in the upper portions of coastal salt marshes (Andreadis et al. 2005). In CT, 83 (13.9%) JCV isolates during 21 yr, have been isolated from Ae. cantator samples, which strongly implicate this species as an important vector species in JCV transmission, especially in shoreline communities. The quantity and frequency of JCV isolations likely reflect high density areas of white-tailed deer in the state (QDMA 2009), and bloodmeals have been obtained with high frequency from this host based on blood meal analysis (Molaei et al. 2008). Likewise, Aedes sollicitans (Walker) and Aedes taeniorhynchus (Wiedemann) are also active later in the season and infected with JCV in coastal salt marsh habitats, but not with the intensity observed with Ae. cantator. Peak JCV isolations from Ae. cantator are obtained from mid-June to mid-July, which is consistent with peak collections of this mosquito, however, because Ae. cantator is multivoltine, JCV has been isolated from this species into early October.

Multivoltine, mammalophillic Anopheles, notably An. punctipennis are frequently infected with JCV and may serve as important vectors. Larvae of An. punctipennis, develop in semi- and permanent freshwater habitats, in slow moving steams in areas of emergent vegetation, and from artificial containers at times (Andreadis et al. 2005). Since this species overwinters as an adult female, it may also serve as an overwintering host of JCV, reinitiating virus transmission when females exit diapause and host seek in early spring (Grimstad 2001). Mosquito trapping programs likely do not detect this early transmission, as the bulk of adult surveillance does not commence until early June, but isolations of JCV can be obtained in the first week of June on occasion in CT. Summer populations of adults build gradually over multiple generations, and JCV has been isolated over a broad period throughout the summer, peaking in July, extending through August and, at low levels, to the end of September. Overall An. punctipennis was frequently identified with JCV, accounting for 55 (9.2%) of CT isolates, and is the second most commonly associated species on a yearly basis with isolates obtained during 23 seasons. Given strong feeding patterns on white-tailed deer, An. punctipennis likely serves as an enzootic vector, and may have the potential to expose humans to the virus later in the season and the following spring if conditions are favorable (Grimstad 2001, Molaei et al. 2008, 2009, Poggi et al. 2023). Vector competency in An. punctipennis has been demonstrated, but at low levels (Heard et al. 1991). A limited number of JCV isolations have also been obtained from pools of Anopheles quadrimaculatus Say (n = 4) and Anopheles walkeri Theobald (n = 1). In a laboratory study, An. quadrimaculatus was able to become infected, show dissemination, but not able to transmit JCV infection (Dieme et al. 2022).

Coquillettidia perturbans is considered a host generalist, obtaining bloodmeals from a variety of avian and mammalian hosts. Larvae develop in permanent freshwater habitats and marshes where mud and emergent vegetation provide the microhabitats where larvae are commonly found. This univoltine species produces a large population of adults accounting for 20.9% of species tested for JCV in CT. A relatively low proportion of these mosquitoes were infected, with 41 JCV isolations (6.9%) obtained during 18 trapping seasons from 1997 to 2022. Isolations peak in late June through mid-July and can extend into August, as adult populations of Cq. perturbans decrease in CT. While white-tailed deer are a common mammalian source of blood-meals, Cq. perturbans has been associated with a broad range of hosts including birds, which may reduce the frequency that JCV may be acquired (Molaei et al. 2008, Shepard et al. 2016, Poggi et al. 2023). Vector competency has been demonstrated, but at low levels (Heard et al. 1991), which may limit the potential of Cq. perturbans to serve as a vector of JCV.

Eastern North America (Except Connecticut)

Field studies from other eastern US states, the Midwest, and southeastern Canada show broad consistency with the results in CT and indicate that JCV infects a diverse array of mammalian-biting mosquito species and strongly associates with univoltine Aedes species. Nevertheless, many gaps remain in the consistency of sampling and geographic coverage across this region. The most comprehensive surveys, outside of CT, come from New York and Michigan, and elsewhere published studies are more limited in scope. Entomological data is particularly sparse in the southeastern United States despite the involvement of human cases in this region (Fig. 1). In the following discussion, we review the available evidence incriminating different vector species and highlight where there are regional differences throughout the endemic range of JCV.

In the northeastern United States, JCV has been frequently isolated from Aedes communis (de Geer) Group mosquitoes (Edwards 1932, Eldridge 1990). This group includes Ae. abserratus, Ae. communis, Aedes intrudens Dyar, Ae. provocans, Ae. punctor, Aedes sticticus (Meigen), and Aedes thibaulti Dyar & Knab, which can be difficult to distinguish adult female specimens by morphology and are often grouped together for testing purposes. These species are univoltine snow pool Aedes that share similar phenologies, larval habitats, and host associations. In New York state, Ae. communis Group mosquitoes accounted for more JCV isolates (41.8%) and the highest infection rate (0.78 per 1,000 mosquitoes) than any other species tested during statewide surveillance efforts from 1972 to 1980 (Grayson and Calisher 1983). A subsequent New York study that differentiated and separated these species incriminated Ae. provocans as the dominant vector due to its high infection rate in field collected adults and evidence for vertical transmission (Boromisa and Grayson 1990). Recently, JCV was detected in Ae. provocans in Maine from a small number of specimens and pools (Schneider et al. 2022), reinforcing that this species is an important vector in regions where this species is abundant. Other members of the communis Group found infected with virus include Aedes abserratus/punctor and Ae. intrudens from western Massachusetts (Walker et al. 1993), Ae. aurifer, Ae. communis Ae. intrudens, and Ae. punctor from New York (Grayson and Calisher 1983, Boromisa and Grayson 1990), and in Ae. abserratus/punctor and Ae.sticticus collected in New Hampshire (Poggi et al. 2023). Given that adult populations peak in early summer, and their potential to maintain the virus by vertical transmission, these species are likely important enzootic vectors of JCV.

Other regionally important mosquito species include the Aedes stimulans (Walker) Group (as annulipes Group Edwards 1932, Eldridge 1990) which includes Ae. cantator, Aedes excrucians (Walker) Aedes euedes Howard, Dyar & Knab, Aedes fitchii (Felt & Young), and Ae. stimulans. These species were identified as the second most important source of JCV infection in New York State (Grayson and Calisher 1983), 2 viral isolations from southern Ontario (Artsob et al. 1983), 4 isolations from mixed pools of Ae. excrucians/stimulans in New Hampshire (Poggi et al. 2023), and a single virus isolate from Maine in Ae. cantator (Schneider et al. 2022). These species are readily distinguished as larvae but they share a similar appearance as adult females, with tarsal claw morphology as a distinguishing character that can lead to misidentification. Due to their morphological similarity, these species have often been grouped and tested together as part of the stimulans Group. Bloodmeal analysis showed these species frequently feed upon white-tailed deer (Molaei et al. 2008, Poggi et al. 2023).

Jamestown Canyon virus has been detected a wide diversity of other mammalian-biting mosquito species in the eastern United States but their relative importance is less clear based on limited sampling efforts. Species with multiple JCV infections reported include Aedes cinereus Meigen, Ae. triseriatus, and Aedes vexans (Meigen) from New York (Grayson and Calisher 1983), Ae. canadensis from New York, Massachusetts, and New Hampshire (Grayson and Calisher 1983, Kinsella et al. 2020, Poggi et al. 2023), Cq. perturbans from New York and New Hampshire (Grayson and Calisher 1983, Poggi et al. 2023), An. punctipennis from New Hampshire and New York (Grayson and Calisher 1983, Poggi et al. 2023), and Ae. sollicitans from New York and Maine (Grayson et al. 1983, Schneider et al. 2022). In the southern United States, JCV has been reported only from 1 pool of Anopheles crucians Wiedemann in South Carolina (Wozniak et al. 2001) and interestingly, 1 pool of male Ae. albopictus reared from field collected eggs in eastern Tennessee (Gottfried et al. 2002). The identity of potential JCV vectors from the southeastern and southcentral regions of the United States remains an important knowledge gap.

In the midwestern United States, early examination of JCV activity was associated with isolates from mosquitoes in the Ae. communis and Ae. stimulans groups similar to findings in the northeastern United States (DeFoliart et al. 1969). Building upon this research in Michigan, JCV was identified most frequently from Ae. provocans (n = 14), with additional isolations from Ae. abserratus/punctor (n = 2), Ae. aurifer (n = 1), Ae. canadenisis (n = 1), and Ae. intrudens (n = 2), reinforcing the roles of these univoltine species which have strong host association with white-tailed deer (Heard et al. 1990). Virus isolates from Ae. provocans were made frequently within the first week (8 of 14 isolates), and all within 2 wk after adult emergence (late May to mid-June), providing circumstantial evidence of transovarial transmission and an overwintering mechanism for JCV in Ae. provocans populations. Additionally, the isolations from Ae. abserratus/punctor and Ae. intrudens were made within 2 wk of adult emergence (late May to early June), which support their involvement in early season transmission. In another study from Wisconsin, more than 40,000 mosquitoes, representing 23 species, were collected in tested for JCV infection of which only Ae. provocans tested positive (Farquhar et al. 2022). This includes 1 pool of mosquitoes that were collected as larvae further implicating Ae. provocans as an important vector and overwintering host. JCV was also isolated from field collected females and transovarially infected male Ae. stimulans in Indiana (Boromisa and Grimstad 1986), strengthening support for Ae. stimulans (and stimulans Group species) as important regional vectors.

In North Dakota, JCV isolations were made from mosquitoes collected during late June through the end of August from 2003 to 2006 (Anderson et al. 2015). The largest number of JCV isolates came from Ae. vexans (n = 58) representing 66% of all virus isolations. This species was the dominant species collected (83% of all mosquito collections) and the JCV infection rate was less than 0.1 per thousand. Additional isolations of JCV from North Dakota were obtained from Aedes trivittatus (Coquillett) (n = 12), Aedes dorsalis (Meigen) (n = 8), Cs. inornata (n = 4), Aedes melanimon Dyar (n = 2), and Ae. sticticus (n = 1), species that had previously identified with JCV infection in North America. Virus isolations were also reported for the first time from Culex tarsalis Coquillett (n = 2) and Aedes flavescens (Müller) (n = 1), but little is known about the potential role these species may play in JCV transmission.

Western North America

Jamestown Canyon virus was originally discovered in the western U.S. when it was first isolated from Cs. inornata from Colorado in 1961; nevertheless, despite this long history with the virus, the vector biology/ecology remains poorly understood within this region. Additional JCV isolates were obtained from Cs. inornata from California and Utah, Ae. melanimon from California, and Ae. communis Group from Alberta, Canada during the 1960’s, as a part of collection and testing efforts by the CDC and other research teams (reviewed in Sudia et al. 1971). This indicates that the virus is broadly distributed throughout western North America which is also supported by the detection of seropositive deer in California and humans in Alaska as well as reported human cases from California, Idaho, Oregon, Montana, and Wyoming (Fig. 1) (Eldridge et al. 1987, Campbell et al. 1989, Walters et al. 1999).

To better understand the regional vector ecology of JCV, mosquito populations were intensively sampled and tested from the high-elevation habitats from California, Nevada, Oregon, and Washington during 1990–1992 (Hardy et al. 1993). JCV was isolated solely from California populations of the Ae. communis Group comprised of: Aedes tahoensis Dyar (n = 10), Aedes cataphylla Dyar (n = 2), and Aedes hexodontus Dyar (n = 1) (Hardy et al. 1993). This included isolations from immature stages of Ae. tahoensis, providing evidence of transovarial transmission. These results support findings from a prior survey in California which recovered JCV from pupae and larvae Ae. cataphylla, and adult Ae. communis senso lato and Ae. hexodontus (Campbell et al. 1991).

Vector competence of orally infected species including those from the communis group: Ae. cataphylla, Ae. hexodontus, Ae. tahoensis and those belonging to the stimulans group: Aedes clivis Lazaro and Eldridge, Aedes increpitus Dyar, and Ae. clivius/Aedes washinoi Lazaro and Eldridge, was demonstrated in laboratory studies for certain strains of JCV in northern California (Kramer et al. 1993). Additionally, transovarial transmission by intrathoracically infected females in the aforementioned species was also documented (Kramer et al. 1993). A container-associated species the Western Rock-Pool Mosquito, Aedes epacticus Dyar & Knab, was shown to be a competent vector of JCV in the laboratory, however, transmission rates varied among colony populations and viral doses evaluated. Additionally, the range Ae. epacticus is not known to overlap the range where JCV is likely to be encountered in the North America (Heard et al. 1991).

Interestingly, few specimens of Cs. inornata (n = 73) were collected and tested by Hardy et al. (1993) reiterating that while the original isolation of JCV was made from Cs. inornata, this species does not frequently test positive (Eldrige 1990) as compared to many boreal Aedes species. However, Cs. inornata, collected from the Central Valley region of California, was shown to be a competent vector of JCV, including demonstrated vertical transmission of the virus when acquired by oral ingestion and intrathoracic injection (Kramer et al. 1993).

Conclusions

Jamestown Canyon virus causes human disease throughout much of North America where it circulates in an enzootic cycle involving mainly univoltine Aedes mosquitoes and deer hosts. Despite progress in our basic understanding of this virus, many questions remain. Jamestown Canyon virus frequently infects human hosts and occasionally causes severe neurological illness, yet the risk factors for disease development are unknown within affected communities. Prior research has focused on the vector ecology of JCV but the identity of the major vectors is poorly understood or unknown in many regions of its endemic range. Vector competence trials indicate that a diversity of mosquito species are capable of transmitting the virus; nevertheless, many of these studies could be improved by increasing the sample sizes so that accurate conclusions can be made. Members of the Aedes communis and stimulans Groups are capable of vertical transmission to their progeny but transovarial transmission has not been assessed for many other woodland species that acquire the virus in nature. Finally, more research is needed to assess the evolutionary potential of JCV via reassortment and genetic drift, and whether circulating strains differ in transmissibility and virulence. Such studies will help us better understand the major determinants of virus transmission and disease that could lead to more effective treatments and public health interventions.

Acknowledgments

We would like to thank Michael Thomas, Tanya Petruff, Angela Bransfield, Michael Misencik, and numerous seasonal assistants for their assistance in coordinating trap collections, mosquito identification, and virus testing of mosquito pools. We give special thanks to Dr. Theodore Andreadis for overseeing past operations of the surveillance program and his helpful input on the development of this review paper. This publication was supported in part by Cooperative Agreement Numbers U01CK000509 and U50CK000524, funded by the Centers for Disease Control and Prevention. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the Centers for Disease Control and Prevention or the Department of Health and Human Services. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References