Domestic Cattle As a Nonconventional Amplifying Host of Vesicular Stomatitis New Jersey Virus

Medical and Veterinary Entomology (2010), doi: 10.1111/j.1365-2915.2010.00932.x

Domestic cattle as a non-conventional amplifying host of vesicular stomatitis New Jersey virus

P. F. S M I T H1, E.W.HOWERTH2, D.CARTER2,E.W.GRAY1, R.NOBLET1,G.SMOLIGA3, L.L.RODRIGUEZ3 and D. G. M E A D4 1Department of Entomology, University of Georgia, Athens, GA, U.S.A., 2Department of Pathology, College of Veterinary Medicine, University of Georgia, Athens, GA, U.S.A., 3Foreign Animal Disease Research Unit, Agricultural Research Service, U.S. Department of Agriculture, Plum Island Animal Disease Center, Orient Point, NY, U.S.A. and 4Southeastern Cooperative Wildlife Disease Study, College of Veterinary Medicine, University of Georgia, Athens, GA, U.S.A.

Abstract. The role of vertebrates as amplifying and maintenance hosts for vesicular stomatitis New Jersey virus (VSNJV) remains unclear. Livestock have been considered dead-end hosts because detectable viraemia is absent in VSNJV-infected animals. This study demonstrated two situations in which cattle can represent a source of VSNJV to Simulium vittatum Zetterstedt (Diptera: Simuliidae) by serving: (a) as a substrate for horizontal transmission among co-feeding black flies, and (b) as a source of infection to uninfected black flies feeding on sites where VSNJV-infected black flies have previously fed. Observed co-feeding transmission rates ranged from 0% to 67%. Uninfected flies physically separated from infected flies by a distance of up to 11 cm were able to acquire virus during feeding although the rate of transmission decreased as the distance between infected and uninfected flies increased. Acquisition of VSNJV by uninfected flies feeding on initial inoculation sites at 24 h, 48 h and 72 h post-infection, in both the presence and absence of vesicular lesions, was detected. Key words. Simulium, black flies, co-feeding transmission, vesicular stomatitis.

Introduction Vertebrate amplifying host(s) for VSNJV in endemic regions and during epidemic outbreaks have not been identiﬁed. Vesicular stomatitis (VS) epidemics associated with vesicular Livestock have been considered as dead-end hosts because stomatitis New Jersey virus (VSNJV) (a member of the family detectable viraemia is absent from VSNJV-infected animals Rhabdoviridae, genus Vesiculovirus) occur sporadically in (Shope & Tesh, 1987). This has led to the theory that a wildlife livestock, primarily cattle and horses, in the western U.S.A. reservoir(s) serves as a VSNJV maintenance and/or amplifying The virus also circulates subclinically and causes disease in host(s) (Webb et al., 1987). livestock in endemic areas throughout Mexico and Central In livestock, viraemia was not detected in VSNJV-infected America and in parts of South America. In the U.S.A., swine (Patterson et al., 1955; Redelman et al., 1989; Comer two patterns of VSNJV activity occur. In addition to the et al., 1995; Stallknecht et al., 1999, 2001, 2004; Mead et al., sporadic epidemic occurrence in livestock in the western states 2004a, 2004b), cattle (Cotton, 1926; Scherer et al., 2007; Mead (Rodriguez et al., 2002), VSNJV is endemic in the feral et al., 2009) or horses (Green, 1993; Howerth et al., 2006). swine population on Ossabaw Island, a barrier island near In studies involving wildlife, experimental VSNJV infection Savannah, Georgia (Stallknecht et al., 1985). Limited numbers of white-tailed deer (Karstad & Hanson, 1957; Comer et al., of domestic livestock were present on Ossabaw Island for many 1995) and pronghorn antelope (Thorne et al., 1983) did not years but were removed during the 1990s (W. O. Fletcher, result in viraemia, which suggests that these wild animals, which are naturally susceptible to VSNJV, are not amplifying personal communication, 2007).

Correspondence: Daniel G. Mead, Southeastern Cooperative Wildlife Disease Study, University of Georgia, 589 D. W. Brooks Drive, Wildlife Health Building, College of Veterinary Medicine, Athens, GA 30602, U.S.A. Tel.: +1 706 542 1741; Fax: +1 706 542 5865; E-mail: [email protected] © 2010 The Authors Medical and Veterinary Entomology © 2010 The Royal Entomological Society 1 2 P. F. Smith et al. hosts. Further, 24 species of wild vertebrates commonly found from tongue epithelium of naturally infected animals during within endemic areas in Panama were experimentally infected the 1982, 1995 and 2006 epidemics. Two virus samples with VSNJV and no viraemia was detected (Tesh et al., 1970). collected during the 1982 epidemic were used to produce virus More recently, viraemia was detected in juvenile and suckling stock. One was collected from tongue epithelium of a cow in deer mice that were infected with VSNJV by intranasal and Colorado (NJ82COB) and the other from tongue epithelium of intradermal inoculation, and in adult deer mice following a cow in Arizona (NJ82AZB). The 1995 virus was isolated intranasal inoculation (Cornish et al., 2001); however, whether from tongue epithelium of a cow in Colorado (NJ95COB) the viraemia was sufficient to infect blood-feeding insects was and the 2006 virus was isolated from tongue epithelium of not determined. a horse in Wyoming (NJ06WYE). In total, 23 5–9-month-old, Recent investigations demonstrated that black flies can mixed breed or Holstein steers or bulls weighing 175–400 kg acquire VSNJV unconventionally while feeding on uninfected were obtained from experimental livestock providers (Cabaniss vertebrate hosts. In experiments using mice and swine, Cattle Co., Stephens, GA; Alan Cagle, Madison, GA; Thomas uninfected Simulium vittatum acquired VSNJV when feeding D. Morris, Inc., Reisterstown, MD) and utilized in this study. in conjunction with VSNJV-infected black flies on naïve hosts (Mead et al., 2000, 2004b). This phenomenon of co-feeding transmission has often been observed with tick-borne viruses Donor fly infection (Jones et al., 1987, 1997; Labuda et al., 1993, 1996, 1997), but has only recently been reported for VSNJV and West Prior to the initiation of the experiments, donor black flies Nile virus, which are vectored by biting flies (Higgs et al., were infected with approximately 103.5 plaque-forming units 2005; McGee et al., 2007). In addition, uninfected black flies (PFU) of NJ82COB, NJ82AZB, NJ95COB or NJ06WYE via acquired VSNJV when they fed on virus-rich lesions on intra-thoracic inoculation and held for 3 days in extrinsic VSNJV-infected swine (Mead et al., 2004b). The efficiency incubation as previously described (Mead et al., 2004a). of VSNJV transmission to uninfected black flies via these routes in an experimental setting suggests that these may be viable transmission routes in natural settings and provide an alternative to viraemia in infected animals for subsequent Experiment 1 insect infection. If confirmed in livestock hosts that typically Co-feeding transmission. An initial experiment examining are found to be infected with VSNJV during epidemics or co-feeding transmission of VSNJV from infected donor flies to in endemic areas, these transmission routes will offer an uninfected recipient flies feeding on naïve cattle was conducted explanation for how VSNJV is amplified during epidemics at the PIADC. Donor flies were inoculated with NJ95COB. and potentially maintained in endemic areas. In the current Recipients were marked with a drop of yellow water-soluble study, our goal was to determine the extent to which domestic paint on the dorsum of the thorax 12 h prior to feeding to cattle, one of the most affected species during epidemics and in distinguish them from donors. endemic areas, serve as a source of VSNJV for biting insects. Five steers were randomly assigned to one of two feeding Specifically, we investigated whether VSNJV was transferred from infected to uninfected black flies co-feeding on the same site groups according to whether flies were to feed on the bovine host and whether uninfected black flies could acquire neck (steer IDs 01, 02 and 03) or coronary band (steer IDs VSNJV by feeding on sites where VSNJV-infected black flies 04 and 05). Feeding cages utilized on neck animals were had previously fed. designed to evaluate the effect of the spatial separation of donors and recipients on co-feeding transmission. A Plexiglas rectangular tube (2.0 × 4.0 cm) was cut into sections 1.5 cm Materials and methods deep. The edges of four sections were glued together with superglue to create a single multi-chambered feeding cage Animal studies were conducted in BSL-3Ag (Biosafety Level (Fig. 1). The open sides were enclosed with polyester mesh 3–agricultural) facilities at the University of Georgia, College (12 squares/cm) to allow flies to feed when the cage was of Veterinary Medicine’s Animal Health Research Center placed against the skin of the animal. Thirty donors and 20 and at the U.S. Department of Agriculture’s Plum Island recipients were placed in chamber A of the feeding cage. Animal Disease Center (PIADC) following standard BSL- Approximately 20 recipients were placed in each of chambers 3Ag biosafety and biosecurity protocols and procedures. The B and C (Fig. 1). Recipients in chamber B were separated use of animals was approved by the University of Georgia’s from donors by 0.3–6.4 cm. Recipients in chamber C were Institutional Animal Care and Use Committee (approval no. separated from donors by 3.3–11.0 cm. Cages designed for use 2008-08-22) and by the PIADC Institutional Animal Care and on the coronary band were made from one section of Plexiglas Use Committee (approval no. 199-05-R). rectangular tubing cut into a 1.5-cm thick section with one side removed. The two broad, open areas were sealed with plastic cut from a large Petri dish and the remaining narrow opening Black flies, experimental animals and VSNJV strains was sealed with polyester mesh. The feeding surface was cut as a curve to allow for better contact with the coronary band Laboratory-reared black flies, S. vittatum (IS-7 cytotype), surface. Approximately 30 infected donors and 20 marked, were utilized. The VSNJVs used in this study were isolated uninfected recipients were placed in each cage.

Chamber C Chamber A Chamber B Experiment 2

Infected flies Uninfected flies Uninfected flies Co-feeding transmission. As with Experiment 1, co-feeding 4.5 cm + transmission from infected donor flies to uninfected recipient flies feeding on the same naïve cattle host was examined. Uninfected flies (marked) Two animals were randomly assigned to one of three 2.0 cm feeding site groups for each of the three viruses tested (NJ82COB, NJ82AZB, NJ06WYE): the coronary band; the muco-cutaneous junction of the lower lip, and the neck. Cages Fig. 1. Diagram of feeding cages utilized on the necks of animals in designed for feeding on the lip were built in the same fashion Experiment 1, as viewed from above with the cage placed against the as coronary band cages from Experiment 1, except that they skin of the animal. Black lines indicate the Plexiglas frame. Chambers featured a straight feeding surface instead of a curved surface. were sealed on the top and bottom with mesh. All chambers were the same size. These cages were also utilized for feeding on neck animals in place of multi-chambered feeding cages. Each feeding cage contained approximately 30 infected donor flies and 20 The designated feeding sites were shaved using commer- marked, uninfected recipient flies. cially available double-bladed disposable razors 24 h prior As in Experiment 1, designated feeding sites were shaved to black fly feeding. During black fly feeding, animals were 24 h prior to black fly feeding. Animals were sedated, sedated (0.1–0.3 mg/kg xylazine, administered i.m.), feeding feeding cages were held against designated feeding sites for cages were manually held for 20–30 min on the designated 20–30 min, and flies were observed for feeding behaviour. areas, and flies were observed for feeding behaviour. At the Upon completion of feeding, donors and recipients were sorted same time, cages containing uninfected flies only were placed as in Experiment 1. Engorged recipients were individually ◦ on the opposite side of the neck of neck site animals and the sorted into cryovials and frozen at −80 C until processed for opposite coronary band of coronary band site animals (e.g. virus isolation. Recipients that were not visibly engorged were ◦ front left coronary band when infected flies were feeding on pooled into groups of three, frozen at −80 C and processed the front right coronary band) and flies were similarly allowed for virus isolation. The number of donors that fed on each to feed. Upon completion of feeding, donors and recipients animal was determined as in Experiment 1. were sedated using carbon dioxide, sorted, and estimates of feeding were made based on visual distension of the abdomen. Insect infection on non-viraemic cattle previously exposed Recipients with visibly distended abdomens (engorged) were ◦ to VSNJV. Uninfected black flies were allowed to feed on individually sorted into cryovials and frozen at −80 C until or near sites at which infected black flies had fed during processed for virus detection. Recipients that were not visi- the co-feeding experiment. Uninfected flies were allowed to bly engorged were pooled into groups of three and frozen at ◦ feed on cattle with clinical or subclinical disease at various −80 C and also processed for virus detection. The number of time-points following feeding by infected flies. Animals were donors that fed on each animal was determined by dissection observed daily for vesicle formation, and nasal, oral and bite and visual observation of blood in the crop or midgut. site swab samples were taken for virus isolation. Swab samples were placed in individual cryotubes containing 1 mL of viral transport medium (MEM supplemented with 3% foetal bovine Virus detection. Samples were processed and assayed using serum, 1000 U penicillin G, 1 mg streptomycin, 0.25 mg gen- two methodologies: virus isolation using African green mon- tamicin sulphate, 0.5 mg kanamycin monosulphate, 2.5 μg/mL key kidney (Vero) Middle America Research Unit [MARU amphotericin B). Blood was collected daily by jugular puncture (Vero M)] cells, and RNA extraction followed by real-time for virus isolation and serology. Additionally, punch biop- reverse transcription polymerase chain reaction (rRT-PCR) as sies were obtained from animals in the neck group infected previously described (Mead et al., 2009). For virus isolation, with NJ06WYE using disposable 6-mm skin biopsy punches ® flies were macerated in 1 mL of Sigma-Aldrich Minimum (Miltex, Inc., Bethpage, NY, U.S.A.) at 24 h and 48 h post- Essential Medium (MEM) (Sigma-Aldrich Corp, St Louis, inoculation. Biopsies were taken from secondary inoculation MO, U.S.A.) and clarified by centrifugation (12 000 × g for sites on the neck where feeding by infected flies had been 10 min). Sample supernatant was inoculated into 24-well tis- observed during the initial infection, according to the obser- sue culture plates containing confluent Vero cell monolayers, vation of swelling and inflammation of these sites. Biopsies ◦ incubated for 1 h at 37 C, and the medium replaced with were cut in half. One half was used for virus isolation and the fresh MEM. Cultures were observed for cytopathic effect at other was fixed in 10% buffered formalin for histopathology 24 h, 48 h and 72 h post-inoculation. Wells with cytopathic and immunohistochemistry as described previously (Howerth effect were confirmed as positive for VSNJV using rRT-PCR. et al., 2006). Approximately 30 uninfected flies were placed Total RNA was extracted using the Qiagen RNeasy kit (Qia- in a clean feeding cage and allowed to feed (as above) on gen, Inc., Valencia, CA, U.S.A.) following the manufacturer’s the initial inoculation site on the two neck animals from each ◦ protocol, and stored at −20 C. Viral RNA was detected by group and on the coronary band and lip animals that exhibited semi-quantitative rRT-PCR specific for the nucleocapsid gene evidence of lesion development at the site of inoculation at of VSNJV, as described previously (Wilson et al., 2009). 24 h, 48 h and/or 72 h post-inoculation. On the neck animals,

Table 1. Evaluation of co-feeding transmission of NJ95COB to uninfected Simulium vittatum (Experiment 1).

Uninfected fly feeding VSNJV+ flies, n/visibly distended flies, n(%) Feeding chamber Inoculation site (infected Opposite side of neck/opposite Animal ID Virus flies feeding, n)ABCcoronary band

01 NJ95COB Neck (3) 3/9 (33.0%) 1/8 (12.5%) 1/7 (14.3%) 0/13 02 NJ95COB Neck (5) 4/6 (66.7%) 0/10 0/12 0/31 03 NJ95COB Neck (3) 2/10 (20.0%) 0/10 0/10 0/18 04 NJ95COB Coronary band (1) 2/6 (33.3%) N/A N/A 0/18 05 NJ95COB Coronary band (1) 1/5 (20.0%) N/A N/A 0/17

Positive for VSNJV (VSNJV+) means positive by virus isolation and real-time reverse transcription polymerase chain reaction. VSNJV, vesicular stomatitis New Jersey virus; N/A, not applicable. uninfected flies were only allowed to feed on primary inocu- observed on two of six experimental animals with transmission lation sites and not where biopsies were collected. Uninfected rates of 11% and 7% (Table 2). In both cases, transmission flies were sorted and processed by virus isolation as described from donors to recipients was observed when fly feeding took below. place on the lip. Co-feeding transmission was not observed when NJ82COB or NJ82AZB were tested. Virus isolation and titration. African green monkey kidney (Vero) MARU (Vero M) cell culture monolayers were Insect infection on non-viraemic cattle previously exposed to used for all virus isolations and titrations. Briefly, black flies VSNJV. Transmission of VSNJV to female black flies feed- were homogenized using a Qiagen7 Mixer Mill 300 (Qiagen, ing on or near developing vesicular lesions was observed for Inc.) (22 cycles/s for 2 min). Swab samples were vortexed for all viruses (Table 3). An uninfected fly was able to acquire approximately 30 s. All samples were clarified by centrifu- virus when feeding on or near vesicular lesions on the lip gation (12 000 × g for 10 min) and 100 μL of the resulting of one steer infected with NJ82COB at 48 h post-inoculation, supernatant were transferred to individual Vero M cell mono- and on the lip of the second steer at 72 h post-inoculation. layers in 12-well plates and observed for cytopathic effect. Titrations of positive swab samples collected immediately Titration of virus isolation-positive samples was performed via endpoint titration on the original samples (Reed & Muench, 1938), and all isolates were confirmed as VSNJV utilizing Table 2. Evaluation of co-feeding transmission of vesicular stomatitis rRT-PCR or RT-PCR with VSNJV specific primers (Rodriguez New Jersey virus strains to uninfected Simulium vittatum (Experi- et al., 1993). ment 2). Inoculation site Uninfected fly feeding Results Animal (infected flies VSNJV+ flies, n/visibly ID Virus feeding, n) distended flies, n

Experiment 1 3 NJ82COB Lip (5) 0/7 26 NJ82COB Lip (1) 0/9 Co-feeding transmission. Transmission of NJ95COB viral 28 NJ82COB Neck (10) 0/5 RNA from infected to uninfected black flies co-feeding on the 29 NJ82COB Neck (2) 0/4 neck or coronary band of naïve cattle was documented on all 27 NJ82COB Coronary band (3) 0/5 five animals. Co-feeding transmission rates for infected and 35 NJ82COB Coronary band (6) 0/3 uninfected flies feeding in the same chamber were variable, 31 NJ82AZB Lip (6) 0/2 ranging from 67% to 20% (Table 1). Further, transmission of 235 NJ82AZB Lip (8) 0/11 NJ95COB to recipients that were physically separated from 32 NJ82AZB Neck (12) 0/9 33 NJ82AZB Neck (8) 0/5 donors by as far as 11 cm was detected. Transmission rates 236 NJ82AZB Coronary band (3) 0/6 from donors to recipients were higher when recipients were 238 NJ82AZB Coronary band (3) 0/7 located in the same chamber as donors rather than being 137 NJ06WYE Lip (10) 1/14 (7.2%) separated from them (Table 1). Uninfected flies feeding on the 278 NJ06WYE Lip (7) 2/19 (10.5%) opposite side of the animal from infected donor flies did not 450 NJ06WYE Neck (6) 0/5 acquire virus on any animal (Table 1). 508 NJ06WYE Neck (4) 0/6 443 NJ06WYE Coronary band (6) 0/4 566 NJ06WYE Coronary band (0) 0/5 Experiment 2 Positive for VSNJV (VSNJV+) means positive by virus isolation and Co-feeding transmission. Transmission of NJ06WYE from verified by reverse transcription polymerase chain reaction. infected to uninfected black flies co-feeding on naïve cattle was VSNJV, vesicular stomatitis New Jersey virus.

Table 3. Evaluation of transmission of vesicular stomatitis New Jersey virus strains to ﬂies feeding on initial inoculation sites at 24 h, 48 h and 72 h post-inoculation (Experiment 2).

Uninfected fly feeding VSNJV+ flies, n/visibly distended flies, n Time post-infection Inoculation site (infected flies Animal ID Virus feeding, n) 24h 48h 72h

3 NJ82COB Lip (5) 0/12 0/5 1/8 (12.5%) 26 NJ82COB Lip (1) 0/6 1/6 (16.7%) 0/4 28 NJ82COB Neck (10) 0/4 0/7 0/5 29 NJ82COB Neck (2) 0/6 0/5 0/8 27 NJ82COB Coronary band (3) 0/2 0/3 1/6 (16.7%) 35 NJ82COB Coronary band (6) 0/6 0/2 0/3 31 NJ82AZB Lip (6) 0/8 2/11 (18.2%) 0/7 235 NJ82AZB Lip (8) 0/12 2/10 (20.0%) 0/8 32 NJ82AZB Neck (12) 0/9 0/11 0/4 33 NJ82AZB Neck (8) 0/4 0/7 0/9 236 NJ82AZB Coronary band (3) 0/4 2/9 (22.2%) 0/2 238 NJ82AZB Coronary band (3) 0/5 0/8 3/10 (30.0%) 137 NJ06WYE Lip (10) 0/11 0/5 Not done 278 NJ06WYE Lip (7) 0/8 0/7 Not done 525 NJ06WYE Lip (4) 0/6 1/8 (12.5%) Not done 450 NJ06WYE Neck (6) 1/15 (6.7%) 0/9 Not done 508 NJ06WYE Neck (4) 3/23 (13.0%) 0/11 Not done 443 NJ06WYE Coronary band (6) 0/5 Not done Not done

Positive for VSNJV (VSNJV+) means positive by virus isolation and verified by reverse transcription polymerase chain reaction. VSNJV, vesicular stomatitis New Jersey virus. after the acquisition of virus by flies feeding on the lip of 0.57 1.45 these animals were 10 TCID50/mL and 10 TCID50/mL, respectively. Transmission of NJ82COB to an uninfected fly feeding on a coronary band animal occurred at 72 h post- inoculation (Table 3). Titration of a positive swab sample from 1.5 this site was 10 TCID50/mL. Transmission of NJ82AZB to 18% and 20% of uninfected black flies feeding on developing lesions on the two lip-feeding animals was observed at 48 h post-inoculation. Titrations of positive swab samples collected from these two animals at 48 h post-inoculation were 3.66 4.45 10 TCID50/mL and 10 CID50/mL, respectively. Trans- mission of NJ82AZB to uninfected black flies feeding on coronary band animals was observed at 48 h post-inoculation on one animal and at 72 h post-inoculation on the second animal (Table 3). Titrations of positive swab samples from these Fig. 2. Punch biopsy of haired skin from the neck of a steer (450) two animals following virus acquisition by uninfected black where VSNJV infected black flies had fed two days previously. 1.5 0.66 ∗ flies were 10 TCID50/mL and 10 TCID50/mL, respec- A minute epidermal vesicopustule (white ) filled with degenerate tively. Transmission of NJ06WYE to uninfected black flies eosinophils and neutrophils expands the epidermis. Small amounts feeding on the lip of a steer previously infected by the bite of of VSNJV antigen was detected in rare degenerate acantholytic cells NJ06WYE-infected black flies was also documented (Table 3). in the pustule and in epidermal cells at the edge of the pustule by immunohistochemistry. Hematoxylin and eosin stain. Bar = 100 mm. Titrations of positive swab samples from the lip of this ani- = × 4.17 Original magnification 200 . mal at 24 h and 48 h post-inoculation were 10 TCID50/mL 4.39 and 10 TCID50/mL, respectively. Additionally, black flies 3.10 feeding on the previously exposed neck region on two sepa- and 10 TCID50/mL, respectively. Skin biopsies were taken rate steers 24 h post-inoculation acquired NJ06WYE. Swabs from fly bite sites on both animals at 24 h and 48 h and of the initial inoculation sites on these two animals at 24 h showed small panepidermal vesicopustules filled with degen- 2.17 post-inoculation yielded VSNJV titres of 10 TCID50/mL erating eosinophils and neutrophils and a small number of 2.80 and 10 TCID50/mL, respectively. Additionally, skin biop- degenerating acantholytic cells surrounded by mild epider- sies taken from these locations on these two animals at 24 h mal spongiosis (Fig. 2). The superficial dermis surrounding 2.80 post-inoculation yielded VSNJV titres of 10 TCID50/mL these bites had congested capillaries and mild perivascular

© 2010 The Authors Medical and Veterinary Entomology © 2010 The Royal Entomological Society, Medical and Veterinary Entomology, doi: 10.1111/j.1365-2915.2010.00932.x 6 P. F. Smith et al. infiltrations of eosinophils and mononuclear cells. By immuno- (Scherer et al., 2007). Development of vesicular lesions was histochemistry, pustular acantholytic cells and a rare cell in the not observed. After needle inoculation of the coronary band, adjacent epidermis were positive for VSNJV in one animal at keratinocytes became infected and served as the primary repli- 24 h and the other at 48 h after feeding by infected flies. cation centre for VSNJV during an active infection, resulting in the formation of vesicles. Findings in the current study show that, in haired regions, microscopic vesiculpustules with sim- Discussion ilarities to larger vesicles caused by VSNJV in non-haired regions develop within 24–48 h of feeding by infected flies For decades, the lack of a known viraemic vertebrate host, and contain low levels of virus. This helps to explain how thought to be necessary for the infection of insects with an uninfected flies were able to acquire VSNJV when feeding on arbovirus such as VSNJV, has hindered our understanding typically haired skin at sites at which infected flies had fed, of VSNJV ecology. Although biological transmission of but grossly detectable vesicles do not form. VSNJV by black flies, sandflies and biting midges has Co-feeding transmission rates ranged from 0% to 67%. It been demonstrated (Comer et al., 1990; Mead et al., 2004a, is interesting to note that efficient transmission of the virus 2004b; Perez de Leon & Tabachnick, 2006), the role of from infected to uninfected black flies was detected in recipi- haematophagous insects in VSNJV transmission has been ent flies that were physically separated from donor flies. These questioned because the process by which they become infected results demonstrate that virus transmission occurs through the with the virus is unknown. vertebrate host, and not directly from infected to uninfected The results presented here corroborate and extend to cat- flies via contaminated faecal droplets or cleptohaematophagy. tle those of previous studies which demonstrate that black Although the specific mechanism for virus transfer during co- flies acquire VSNJV from a domestic animal species by routes feeding is not known, two possibilities are presented here. One which oppose convention. In addition, this work demonstrates hypothesis is that virus transfer during co-feeding is a passive for the first time that co-feeding transmission of VSNJV does process. Black flies are pool feeders and use their mouthparts not require direct contact between donor and recipient black to lacerate the epidermis to gain access to the dermal capillary flies and that, even in the absence of lesions, feeding sites bed, which they also lacerate. They then feed from the pools of of infected flies can remain infectious to recipient flies for blood that gather in the wound. Black fly salivary components at least 24 h. In the current study, the processing of recipient introduced into the site inhibit clotting and promote vasodi- flies can only indicate acquisition of VSNJV because recipients lation (Cupp & Cupp, 1997). Additionally, it is possible that were not held for an extrinsic incubation period. However, in uninfected flies feeding in the same capillary bed could inci- previous experiments, black flies became infectious following dentally ingest VSNJV introduced into the capillary bed by an oral inoculation and incubation with as little as 101.3 PFU of infected black fly. Alternatively, the transfer of VSNJV during VSNJV (Mead et al., 1997). These findings provide potential co-feeding may be an active process in which VSNJV migrates explanations for how VSNJV might be amplified during epi- from one feeding site to another via keratinocytes or migrating demics which involve livestock and, potentially, for how the cells, such as intra-epidermal or dermal dendritic cells. virus might be maintained in endemic regions. Specifically, When we consider these findings together with those of pre- it was demonstrated that black flies can acquire VSNJV via vious studies demonstrating that VSNJV can be transmitted three separate routes: (a) co-feeding in direct contact with, or mechanically to livestock by black flies (Smith et al., 2009), separated by up to 11 cm from, VSNJV-infected black flies; that the virus can be transmitted from animal to animal via (b) feeding on or near fly bite sites where vesicular lesions are contact transmission (Mead et al., 2004b; Stallknecht et al., not known to develop (i.e. haired regions of the neck) up to 2004) and that VSNJV is transovarially transmitted in sand- 24 h after VSNJV-infected black flies have previously fed, and flies (Comer et al., 1990), a new theory, which does not rely on (c) feeding on active vesicular lesions. a viraemic vertebrate host for VSNJV amplification and main- The acquisition of VSNJV by black flies feeding on vesicu- tenance, emerges. It is proposed that during epidemics and lar lesions was previously demonstrated using VSNJV-infected in endemic cycles that involve sandflies and domestic live- swine (Mead et al., 2004a, 2004b) and was validated in cattle stock, VSNJV is transmitted from VSNJV-infected livestock in the current study. Although viraemia is absent in livestock to blood-feeding insects via unconventional means. Although 2.3 4.6 hosts, virus titres ranging from <10 TCID50 to 10 TCID50 the current study focuses on cattle as a source of virus for per swab have been demonstrated in samples collected from an epidemic VSNJV vector, the proposed theory could also be the surface of vesicular lesions (Stallknecht et al., 2001, 2004) applied to cattle and endemic vectors. In regions where VSNJV and 109.15 PFU/mL in fluid aspirated from vesicular lesions is endemic, sandflies are thought to be the primary vector. (Marcus et al., 1998). Virus concentrations of this magnitude The relationship between sandflies and livestock in these areas can easily explain the acquisition of VSNJV by naïve flies has not been thoroughly investigated; however, based on the feeding on these regions. current findings with black flies, the following scenario is pro- The mechanism by which black flies acquire VSNJV when posed. Like black flies, sandflies are aggregate feeders and feed feeding on the neck at sites at which infected flies have on livestock (Schlein et al., 1984; Dias et al., 2003), so there fed previously is not clear. After needle inoculation of flank is a strong likelihood that uninfected sandflies co-feeding with skin, VSNJV was found to be associated with keratinocytes VSNJV-infected sandflies may ingest virus. Additionally, the at 24 h post-inoculation but not at 48 h post-inoculation probability that uninfected sandflies will ingest virus when they

© 2010 The Authors Medical and Veterinary Entomology © 2010 The Royal Entomological Society, Medical and Veterinary Entomology, doi: 10.1111/j.1365-2915.2010.00932.x Vesicular stomatitis New Jersey virus in domestic cattle hosts 7 feed on or near vesicular lesions of VSNJV-infected livestock Howerth, E.W., Mead, D.G., Mueller, P.O., Duncan, L., Murphy, M.D. is also strong. & Stallknecht, D.E. (2006) Experimental vesicular stomatitis virus Clearly, the possible scenarios presented here expand con- infection in horses: effect of route of inoculation and virus serotype. ventional theories of insect-vectored virus transmission and Veterinary Pathology, 43, 943–955. maintenance and provide an explanation that fits with current Jones, L.D., Davies, C.R., Steele, G.M. & Nuttall, P.A. (1987) A understanding of VSNJV ecology. It is possible that VSNJV novel mode of arbovirus transmission involving a non-viraemic host. may be maintained without a viraemic reservoir host and Science, 237, 775–777. that such a reservoir may not even exist. Additional work is Jones, L.D., Gaunt, M., Hails, R.S. et al. (1997) Transmission of required to investigate the relationships among VSNJV, sand- louping ill virus between infected and uninfected ticks co-feeding on mountain hares. Medical and Veterinary Entomology, 11, 172–176. flies and livestock hosts in endemic regions to test the validity Karstad, L. & Hanson, R.P. (1957) Vesicular stomatitis in deer. of the proposed theory. If verified, these alternative mecha- American Journal of Veterinary Research, 18, 162–166. nisms of VSNJV insect transmission may greatly influence Labuda, M., Jones, L.D., Williams, T., Danielova, V. & Nuttall, P.A. our current knowledge of VSNJV ecology and better guide (1993) Efficient transmission of tick-borne encephalitis virus strategies for the management of VSNJV outbreaks in livestock between co-feeding ticks. Journal of Medical Entomology, 30, populations. 295–299. Labuda, M., Austyn, J.M., Zuffova, E., Kosuch, O., Fuchsberger, N., Lysy, J. & Nuttall, P.A. (1996) Importance of localized skin Acknowledgements infection in tick-borne encephalitis virus transmission. Virology, 219, 357–366. Funding for this research was primarily provided by the Labuda, M., Kozuch, O., Zuffova, E., Eleckova, E., Hails, R.S. & National Research Initiative of the U.S. Department of Agri- Nuttall, P.A. (1997) Tick-borne encephalitis virus transmission culture (USDA) Cooperative State Research, Education and between ticks co-feeding on specific immune natural rodent hosts. Extension Service (grant no. 2005-35204-16102). Additional Virology, 235, 138–143. support was provided by the Georgia Research Alliance, Marcus, P.I., Rodriguez, L.L. & Sekellic, M.J. (1998) Interferon induction as a quasispecies marker of vesicular stomatitis virus USDA Animal and Plant Health Inspection Service (APHIS), populations. Journal of Virology, 72, 542–549. Veterinary Services Cooperative Agreement (08-9613-0032- McGee, C.E., Schneider, B.S., Girard, Y.A., Vanlandingham, D.L. & CA) and the USDA (CRIS-1940-32000-04000D). Higgs, S. (2007) Non-viraemic transmission of West Nile virus: evaluation of the effects of space, time, and mosquito species. American Journal of Tropical Medicine and Hygiene, 76, 424–430. References Mead, D.G., Mare, C.J. & Cupp, E.W. (1997) Vector competence of select blackfly species for vesicular stomatitis virus (New Jersey Comer, J.A., Tesh, R.B., Modi, G.B., Corn, J.L. & Nettles, V.F. serotype). American Journal of Tropical Medicine and Hygeine, 57, (1990) Vesicular stomatitis virus, New Jersey serotype: replication 42–48. in and transmission by Lutzomyia shannoni (Diptera: Psychodidae). Mead, D.G., Ramberg, F.B., Besselsen, D.G. & Mare, C.J. (2000) American Journal of Tropical Medicine and Hygiene, 42, 483–490. Efficient transmission of vesicular stomatitis virus between infected Comer, J.A., Stallknecht, D.E. & Nettles, V.F. (1995) Incompetence and non-infected blackflies co-feeding on non-viraemic deer mice. of domestic pigs as amplifying hosts of vesicular stomatitis virus Science, 287, 485–487. for Lutzomyia shannoni (Diptera: Psychodidae). Journal of Medical Mead, D.G., Gray, E.W., Noblet, R., Murphy, M.D., Howerth, E.W. Entomology, 32, 741–744. & Stallknecht, D.E. (2004a) Biological transmission of vesicular Cornish, T.E., Stallknecht, D.E., Brown, C.C., Seal, B.S. & How- stomatitis virus (New Jersey serotype) by Simulium vittatum erth, E.W. (2001) Pathogenesis of experimental vesicular stomatitis (Diptera: Simuliidae) to domestic swine (Sus scrofa). Journal of virus (New Jersey serotype) infection in the deer mouse (Peromyscus Medical Entomology, 41, 78–82. maniculatus). Veterinary Pathology, 38, 396–406. Mead, D.G., Howerth, E.W., Murphy, M.D., Gray, E.W., Noblet, R. Cotton, W.E. (1926) Vesicular stomatitis in its relation to the diagnosis & Stallknecht, D.E. (2004b) Blackfly involvement in the epi- of foot-and-mouth disease. Journal of the American Veterinary demic transmission of vesicular stomatitis New Jersey virus (Rhab- Medical Association, 22, 313–332. doviridae: Vesiculovirus). Vector-Borne and Zoonotic Diseases, 4, Cupp, E.W. & Cupp, M.S. (1997) Blackfly (Diptera: Simuliidae) 351–359. salivary secretions: importance in vector competence and disease. Mead, D.G., Rainwater-Lovett, K., Murphy, M.D. et al. (2009) Exper- Journal of Medical Entomology, 34, 87–94. imental transmission of vesicular stomatitis New Jersey virus from Dias, F.O., Lorosa, E.S. & Rebelo, J.M. (2003) Blood-feeding sources Simulium vittatum to cattle: clinical outcome is influenced by site and peridomiciliation of Lutzomyia longipalpis (Lutz & Neiva, of insect feeding. Journal of Medical Entomology, 46, 866–872. 1912) (Psychodidae, Phlebotominae). Cadernos de Saude Publica, Patterson, W.C., Jenney, E.W. & Holbrook, A.A. (1955) Experimental 19, 1373–1380. infections with vesicular stomatitis in swine. I. Transmission by Green, S.L. (1993) Vesicular stomatitis in the horse. Veterinary Clinics direct contact and feeding infected meat scraps. Proceedings of the of North America Equine Practice, 9, 349–353. United States Livestock Sanitation Association, 59, 368–374. Higgs, S., Schneider, B.S., Vanlandingham, D.L., Klingler, K.A. Perez de Leon, A.A. & Tabachnick, W.J. (2006) Transmission of & Gould, E.A. (2005) Non-viraemic transmission of West Nile vesicular stomatitis New Jersey virus to cattle by the biting virus. Proceedings of the National Academy of Sciences of the United midge Culicoides sonorensis (Diptera: Ceratopogonidae). Journal States of America, 102, 8871–8874. of Medical Entomology, 43, 323–329.

Redelman, D., Nichol, S., Klieforth, R., Van der Maaten, M. & Whet- Stallknecht, D.E., Howerth, E.W., Reeves, C.L. & Seal, B.S. (1999) stone, C. (1989) Experimental vesicular stomatitis virus infection of Potential for contact and mechanical vector transmission of vesicular swine: extent of infection and immunological response. Veterinary stomatitis virus New Jersey in pigs. American Journal of Veterinary Immunology and Immunopathology, 20, 345–361. Research, 60, 43–48. Reed, L.J. & Muench, H. (1938) A simple method of estimating fifty Stallknecht, D.E., Perzak, D.E., Bauer, L.D., Murphy, M.D. & percent endpoints. American Journal of Hygiene, 27, 493–497. Howerth, E.W. (2001) Contact transmission of vesicular stomatitis Rodriguez, L.L. (2002) Emergence and re-emergence of vesicular virus New Jersey in pigs. American Journal of Veterinary Research, stomatitis in the United States. Journal of General Virology, 83, 60, 43–48. 2475–2483. Stallknecht, D.E., Greer, J.B., Murphy, M.D., Mead, D.G. & How- Rodriguez, L.L., Letchworth, G.J., Spiropoulou, C.F. & Nichol, S.T. erth, E.W. (2004) Effect of strain and serotype of vesicular stomati- (1993) Rapid detection of vesicular stomatitis virus New Jersey tis virus on viral shedding, vesicular lesion development, and contact serotype in clinical samples by using polymerase chain reaction. transmission in pigs. American Journal of Veterinary Research, 65, Journal of Clinical Microbiology, 31, 2016–2020. 1233–1239. Scherer, C.F.C., O’Donnell, V., Golde, W.T., Gregg, D., Estes, D.M. Tesh, R.B., Peralta, P.H. & Johnson, K.M. (1970) Ecologic studies of & Rodriguez, L.L. (2007) Vesicular stomatitis New Jersey virus vesicular stomatitis virus. II. Results of experimental infection in (VSNJV) infects keratinocytes and is restricted to lesion sites and Panamanian wild animals. American Journal of Epidemiology, 91, local lymph nodes in the bovine, a natural host. Veterinary Research, 216–224. 38, 375–390. Thorne, E.T., Williams, E.S., Adrian, W.J. & Gillin, C.M. (1983) Schlein, Y., Warburd, A., Schnur, L.F., Le Blancq, S.M. & Gun- Vesicular stomatitis in pronghorn antelope: serologic survey and ders, A.E. (1984) Leishmaniasis in Israel: reservoir hosts, sandfly artificial infection. Proceedings of the United States Animal Health vectors and leishmanial strains in the Negev, Central Arava and Association, 87, 638–653. along the Dead Sea. Transactions of the Royal Society of Tropical Webb, P.A., McLean, R.G., Smith, G.C., Ellenberger, J.H., Francy, Medicine and Hygiene, 78, 480–484. D.B., Walton, T.E. & Monath, T.P. (1987) Epizootic vesicular Shope, R.E. & Tesh, R.B. (1987) The ecology of rhabdoviruses stomatitis in Colorado, 1982: some observations on the possible that infect vertebrates. The Rhabdoviruses (ed. by R.R. Wagner), role of wildlife in an enzootic maintenance cycle. Journal of Wildlife pp. 509–534. Plenum, New York, NY. Diseases, 23, 192–198. Smith, P.F., Howerth, E.W., Carter, D., Gray, E.W., Noblet, R. & Wilson, W.C., Letchworth, G.J., Jimenez, C. et al. (2009) Field evalu- Mead, D.G. (2009) Mechanical transmission of vesicular stomatitis ation of a multiplex real-time reverse transcription polymerase chain New Jersey virus by Simulium vittatum (Diptera: Simuliidae) to reaction assay for detection of vesicular stomatitis virus. Journal of domestic swine (Sus scrofa). Journal of Medical Entomology, 46, Veterinary Diagnostic Investigation, 21, 179–86. 1537–1540. Stallknecht, D.E., Nettles, V.F., Fletcher, W.O. & Erickson, G.A. (1985) Enzootic vesicular stomatitis New Jersey type in an insular Accepted 27 August 2010 feral swine population. American Journal of Epidemiology, 122, 876–883.