Abstract
Vesicular stomatitis viruses are the causative agents of vesicular stomatitis, an economically important contagious disease of livestock that occurs in North, Central, and South America. Little is known regarding the early stages of infection in natural hosts. Twelve adult Holstein steers were inoculated with Vesicular stomatitis New Jersey virus (VSNJV) on the coronary bands (CB) of the feet via scarification (SC) or by VSNJV-infected black fly (Simulium vittatum) bite (FB). Three additional animals were inoculated on the neck skin using FB. Clinical disease and lesion development were assessed daily, and animals were euthanatized from 12 hours post inoculation (HPI) through 120 HPI. The animals inoculated in the neck failed to develop any clinical signs or gross lesions, and VSNJV was detected neither by in situ hybridization (ISH) nor by immunohistochemistry (IHC). Lesions on the CB were more severe in the animals infected by FB than by SC. In both groups, peak VSNJV replication occurred between 24 and 48 HPI in keratinocytes of the CB, as evidenced by ISH and IHC. There was evidence of viral replication limited to the first 24 HPI in the local draining lymph nodes, as seen through ISH. Successful infection via FB required logarithmically less virus than with the SC technique, suggesting that components in black fly saliva may facilitate VSNJV transmission and infection in cattle. The lack of lesion development in the neck with the same method of inoculation used in the CB suggests that specific characteristics of the CB epithelium may facilitate VSNJV infection.
Vesicular stomatitis New Jersey virus (VSNJV) is an arthropod-borne virus (arbovirus) in the family Rhabdoviridae, genus Vesiculovirus. Geographic distribution of VSNJV is restricted to the Americas, occurring from South America to the western United States, 11 and is one of the causative agents of vesicular stomatitis (VS), a contagious and vector-borne disease of livestock. Although VSNJV infects a wide range of species including wildlife 19 and laboratory and domestic animals, 11 the natural disease is only reported in cattle, pigs, and horses. The disease is clinically indistinguishable from other viral vesicular diseases, especially foot-and-mouth disease, when cattle and pigs are affected.
Vesicular stomatitis is characterized by fever from 48 to 72 hours post inoculation (HPI) with noticeable vesicle formation between 48 and 72 HPI. Vesicles can occur in the oral cavity (dorsal tongue epithelium, hard palate, and gums), muzzle, snout, lip, coronary bands (CB) of the feet, teats, and prepuce. 11,13,15 Vesicles can rupture, with consequent viral shedding and potential animal-to-animal contact transmission. 17 Lesions tend to heal around the second week after vesicle formation. Transmission also occurs by actively infected biting insects, such as black flies (Simulium spp.), 6 sand flies (Lutzomyia spp.), 18 and biting midges (Culicoides spp.). 12 VSNJV has been isolated from nonbiting insects during disease outbreaks, 3 and it is believed they might play a role in mechanical transmission of the virus during epizootics.
The detailed pathogenesis of VSNJV infection is unclear. Clinical disease can be consistently reproduced experimentally by application of at least 107 TCID50 per animal to scarified regions where lesions naturally occur. 14 When VSNJV is inoculated in haired skin of the neck, 14 flank, 14 or ear, 5 lesions do not occur but animals may seroconvert and/or shed virus through the tonsils (subclinical infection). Even though no lesions develop in those haired skin sites, it has been demonstrated that there can be limited viral replication. 8 Also, noninfected black flies physically separated from VSNJV-infected black flies became infected while cofeeding at these sites on a nonviremic host. 10 However, none of the previous studies used in situ hybridization (ISH) or immunohistochemistry (IHC) to detect virus in those areas.
Compared with inoculation via scarification (SC), feeding of VSNJV-infected black flies at sites where vesicle formation normally occurs will reproduce the disease, 6 although the amount of virus contained within the feeding flies is far less than what is required to reproduce disease via SC. 14 Previous studies investigating the pathogenesis of VSNJV in experimentally infected livestock have focused on the clinical course and late stages of the disease 4,9 following infection via SC or intradermal injection. However, the early events of experimental or natural VSNJV infection in livestock, including virus replication kinetics and distribution in tissues, have not been documented.
The present study was designed to sequentially assess and describe lesion development, virus distribution, and replication kinetics in cattle tissues during the early phases of VSNJV infection. Additional objectives were to compare insect bite transmission (VSNJV-infected black fly (Simulium vittatum) bite [FB]) to direct inoculation (SC) and to assess the presence of virus inoculated in susceptible versus nonsusceptible regions of skin via FB.
Materials and Methods
All experimental infections and all methods using nonfixed samples were conducted at the USDA-ARS Plum Island Animal Disease Center under biosafety level (BSL)-3Ag conditions. Tissues fixed in formalin were transferred to the Department of Pathology at the University of Georgia, for histological processing, immunohistochemistry, and in situ hybridization studies.
The virus used was a VSNJV field strain (95COB) isolated from a cow during an outbreak in Colorado in 1995. This virus was responsible for a large VS outbreak in 1995 and has been used previously to successfully reproduce clinical disease in cattle. 14
Fifteen 150- to 250-kg Holstein steers were used in this experiment. Scarification inoculation was done on 7 animals, 6 scarified with virus and 1 with cell culture media only. Eight animals were inoculated by FB: 7 with infected flies and 1 with noninfected flies. Three steers had flies feed on the neck skin, whereas for the other 5 FB animals, inoculation was on the CB. All animals had hair from the area of inoculation removed with clippers followed by shaving 24 hours prior to the inoculation time. They were sedated with an intramuscular injection of 0.3 mg/kg xylazine prior to the inoculation, and sedation was reversed after inoculation by intravenous injection of 2 mg/kg tolazoline.
Inoculation by Scarification
A plastic bifurcated needle (Duotip Tests, Lincoln Diagnostics, Decatur, IL) served as the inoculating tool and was used as previously described. 14 The line of scarification was approximately 8 cm long and located on the lateral aspect of the limbs, on the haired skin of the CB immediately proximal to the thick nonhaired skin (Fig. 1 ). The inoculum contained 107 TCID50 of VSVNJ in 50 μl of cell culture media with 1% antibiotic/antimycotic as previously described; 14 virus-free medium was used as the negative control. The scarification site was created with approximately 25 skin pricks, and then the inoculum or virus-free medium was applied with a pipette. The procedure was repeated, with 25 additional skin pricks and application of additional medium. In experimental animals, the viral inoculum was applied to the right front and rear CB, whereas the left front and rear CB were scarified and treated with virus-free medium. In the mock-inoculated animal, the right front and rear CB were scarified and treated with virus-free medium, whereas the left front and rear CB were left intact. The mock-inoculated animal was euthanatized at 24 HPI, and the virus-infected animals were euthanatized at 12, 24, 48, 72, 96, and 120 HPI.
Inoculation by Fly Bite
Inoculation by fly bite was done as in previous studies. 8 Briefly, black flies were intrathoracically injected, as previously described, 7 with approximately 1 μl containing 107 TCID50 of VSNJV. Infected black flies were allowed to feed on the designated CB of 4 animals and on sites on the necks of 3 animals. For the CB animals, 30 caged VSNJV-infected black flies were allowed to feed for approximately 20 minutes at the lateral aspect of the right front and rear CB (Fig. 2 ). Cages were placed in a position to allow flies to access either haired or glabrous CB skin. In general, flies probed or bit in the haired skin, approximately 1 cm proximal to the thick glabrous CB skin. The same number of uninfected black flies was allowed to feed for approximately 20 minutes at the lateral aspect of the left front and rear CB. These 4 FB CB experimental animals were euthanatized at 24, 48, 72, and 96 HPI. One mock-inoculated animal was treated with uninfected black flies on the right CB, with no treatment at the left CB, and was euthanatized at 24 HPI. The same method was used in the group of the 3 neck-inoculated animals. Infected black flies were allowed to feed on the right side of the neck, and uninfected black flies were allowed to feed on the left side of the neck. Of these 3 steers inoculated in the neck, 1 animal was euthanatized at 24 HPI and 2 animals at 72 HPI.
Analysis of Black Flies
As a quality control for the fly bite inoculation, all flies were tested for the presence or absence of mammalian blood by SYBR-Green real-time polymerase chain reaction (RT-PCR) for mammalian DNA (Power SYBR Green PCR Master Mix ABI 4367659) after feeding on cattle. Briefly, nucleic acid was extracted from each individual fly as described below for tissues. RT-PCR was conducted using primers designed specifically to target the 16s mammalian recombinant DNA gene. Forward primer sequence was 5′ CCTGTTTACCAAAAACATCAC 3′ and reverse primer sequence 5′ AYTGTCGATAKGRACTCTWRARTAG 3′ (Invitrogen, Carlsbad, CA). Primers were used at a final concentration of 300 nM per 25 µl reaction. The reaction was performed using 8 µl (per 25 µl reaction) of Power SYBR Green PCR Master Mix (Applied Biosystems 4367659). The amount of nucleic acid template was 2.5 µl per reaction, which was performed on an ABI 7000 sequence detection system using the following cycling conditions: 94°C for 2.5 minutes, and 40 cycles of 94°C for 45 seconds, 50°C for 45 seconds, 72°C for 60 seconds, and a final 72°C hold for 5 minutes. After the RT-PCR, a melt curve was performed to detect the Tm of SYBR Green-bound DNA products. Products with a Tm of approximately 80°C were considered positive for mammalian DNA.
The blood meal analysis was done in all flies used in this study and determined that approximately 20% of the flies (which corresponds to about 6 flies) fed during the period of mock and viral inoculation (data not shown).
Samples
Samples collected immediately after euthanasia included lateral aspect of all 4 CB and potential lesions extending from the original site of the inoculation, right and left neck skin, and local draining lymph nodes (LN) including prescapular, axillary, popliteal, and prefemoral. Spleen, tongue, and tonsils were also collected. Approximately 30 mg of collected tissues were quick-frozen in liquid nitrogen for later transfer to –70°C for virus isolation (VI) and reverse transcriptase RT-PCR (rRT-PCR) for detection of live virus and viral nucleic acid respectively. Additionally, serum was collected just prior to euthanasia for VI and rRT-PCR. Sections <1 cm thick of the remaining tissues were fixed in a 10% buffered formalin solution, for histopathology, immunohistochemistry, and in situ hybridization. The whole neck and CB skin at and around fly cage location and scarified areas were collected and trimmed for histopathologic evaluation. For each of these sites, 4 tissue cassettes were collected, each with 3 samples of epidermis. All 4 hematoxylin and eosin (HE)–stained slides from these blocks (equaling at least 12 sections from each site) were examined; a minimum of 2 slides (6 sections) were selected, and corresponding sections were analyzed by ISH and IHC. A semiquantitative score was attributed to the ISH and IHC slides: negative (N) was assigned for no signal, + for minimal to mild or faint signal, ++ for a moderately discrete but localized signal, and +++ for a prominent and diffuse signal.
Real Time Reverse Transcriptase PCR (rRT-PCR)
Tissues were thawed and immediately macerated by adding two 5-mm stainless steel beads (Qiagen, Catalog No. 69989) and 0.9 ml of MEM-25mM HEPES and then were shaken in a TissueLyser bead beater (Qiagen) for 2 minutes at a frequency of 22 per second. After maceration, 50 μl of sample was transferred to 96-well plates (King Fisher No. 97002540) containing 150 μl of lysis/binding solution. RNA was then extracted using Ambion’s MagMax-96 Viral RNA Isolation Kit (Ambion, catalog No. 1836) on a King Fisher-96 Magnetic Particle Processor (Thermo Electron Corp.). Briefly, after the initial 5-minute lysis/binding step, the RNA sample underwent a series of 4 washing steps, a drying step, and a final elution. RNA was eluted in a final volume of 25 μl. At each of the above steps, RNA was magnetically bound to the beads contained in the lysis/binding solution and was transferred to the different extraction solutions. The extracted RNA was analyzed by rRT-PCR using 2.5 μl of RNA on the ABI 7000 as described below. The remaining macerated tissue was clarified (1000 g for 2 minutes at 4°C), and the supernatant was cleared of possible bacterial contamination using centrifuge tube filters (Spin-X, Costar). Samples were then stored at –70°C for VI. The rRT-PCR was performed using primers and probe designed specifically for the nucleocapsid region of isolate used for the inoculations (95COB). Forward primer sequence was 5′ GCACTTCCTGATGGGAAATCA 3′ and reverse primer sequence 5′ GGGAAGCCATTTATCATCCTCA 3′ (Invitrogen, Carlsbad, CA), and a 6-carboxyfluorescein (6-FAM)–labeled probe 5′ ACCCTGACCGTTCTG 3′ (Applied Biosystems, Foster City, CA) was used. Primers were used at a final concentration of 300 nM, and probes were used at 100 nM per 25 μl reaction. The reaction was performed using the Taqman EZ RT-PCR Core Reagents (Applied Biosystems N808-0236) at the following volumes/concentrations per 25 µL reaction: 5 μl of 5× buffer, 3 μl of 25 mM Mn (OAc)2, 3 μl of dNTPs (combined 1:1:1:1 by volume), 1 μl of rTth, 0.25 μl of Amperase, and 9.7 μl of H20. The amount of RNA template was 2.5 µl per reaction, which was performed on an ABI 7000 sequence detection system using the following cycling conditions: 60°C for 25 minutes, 95°C for 2 minutes, and 40 cycles of 95°C for 10 seconds and 60°C for 1 minute. Results were expressed as cycle threshold (CT) values. A CT value <40 was considered positive for the presence of virus. Negative control consisted of all rRT-PCR reagents and water in place of template.
Virus Isolation
A 250-µl aliquot of the supernatant described above was placed in individual T-25 tissue-culture flasks containing a monolayer of VERO cells. The virus was allowed to adsorb for 1 hour at 37°C on a rocker plate, followed by addition of 5 ml of maintenance media with 3 subsequent days of incubation at the same temperature. Flasks were examined under an inverted microscope to observe cytopathic effect (CPE) at 24, 48, and 72 HPI. After 72 HPI, CPE-positive samples were stored at –70°C for rRT-PCR, and 250 µl of supernatant from each CPE-negative sample was transferred to a new culture flask (second passage). After 72 HPI, all flasks (CPE positive or negative) from the second passage and all CPE-positive materials from the first passage were processed for rRT-PCR.
Histopathology and Immunohistochemistry
All formalin fixed samples were routinely embedded in paraffin, cut at 3- to 4-μm thickness, and stained with HE for histological examination to assess inflammatory reaction, extent of microvesicle formation, and any other pathologic changes.
For IHC, deparaffinized sections were subjected to antigen retrieval with 80 μg/ml proteinase K for 15 minutes at 37°C, followed by blocking of nonspecific epitopes. Then samples were incubated overnight at 4°C with primary antibody, anti-VSNJV polyclonal, a mouse hyperimmune ascitic fluid (kindly provided by Dr. Robert Tesh, Department of Pathology, Medical Branch, University of Texas, Galveston), at a dilution of 1:1800. Primary antibody was followed by incubation with an alkaline phosphatase-linked polymer system (LabVision). The reaction was revealed using Vector Red chromogen. The slides were lightly counterstained with Mayer’s hematoxylin and then cover-slipped using Permount for permanent record. For each IHC protocol, anti-VSV antibody was applied to a left (mock-inoculated side) CB or neck skin and popliteal or prefemoral lymph nodes for negative control. Additionally, each IHC protocol had a slide containing a section of CB skin with a recognizable histological vesicular lesion or a lymph node draining a CB with lesion. For these sections, normal mouse serum was used instead of the anti-VSV antibody. Selected lymph node sections were additionally subjected to an IHC protocol to highlight dendritic cells, using the same procedure outlined above, except that the primary antibody was anti-CD1b (monoclonal ascites fluid antibody; VMRD catalog No. TH97A) at a dilution of 1:100.
In Situ Hybridization
Negative sense riboprobes were used to detect positive-stranded (replication) viral RNA. The riboprobes consisted of 410 bases corresponding to the nucleoprotein (N) gene using methods previously described. 2 The following primers were used to produce the N gene construct: 5′-GAA GAT GGT CTT GAC TTC TTT G-3′ (forward) and 5′-CGA GTT GAT CTT AGC AAG AGT G-3′ (reverse). The N gene segment was cloned and amplified in a pGEM-4Z vector (Promega, Madison, WI). The vector was digested with BglII restriction enzyme to create a template for the negative sense probe, and digoxigenin-labeled nucleotides were added during transcription of the probe. Incorporation of the digoxigenin was verified by dot blot. Sections from tissues that were positive either on VI or on rRT-PCR were processed for ISH as previously described. 1 Tissue sections were treated with 100 μg/ml proteinase K solution and then incubated with the negative sense VSNJV riboprobe (corresponding to the N gene) at 42°C for 12 hours, followed by incubation with antidigoxigenin conjugated to alkaline phosphatase (1:300). Negative controls included noninfected tissues. The reaction was visualized using tetrazolium-based (NBT-BCIP) chromogen. The slides were lightly counterstained with Mayer’s hematoxylin then cover-slipped with Permount for permanent record.
Results
Animals Inoculated by FB on the Neck
None of the animals inoculated in the neck skin developed fever or any other clinical sign of disease. They had only multiple small (approximately 1 mm in diameter) focal areas of hyperemia at the site of FB in both right and left (infected and uninfected FB) neck skin.
Histologically, the superficial dermis had minimal to mild foci of inflammatory infiltrates in the right and left neck skin of all 3 steers. Dermal infiltrates were slightly more severe in the animal euthanatized at 24 HPI and were composed of small numbers of macrophages with fewer neutrophils. There were no lesions in the epidermis, except for the animal euthanatized at 24 HPI that had scant small (approximately 0.5 mm in diameter) focal areas of epidermal necrosis within the FB region in both right and left sides.
There were very limited positive results by any of the techniques used to detect virus in these tissues. Inoculation site skin in the 24 HPI animals was positive by both VI and rRT-PCR on the right (infected) side. At 72 HPI, 1 of the 2 euthanatized animals at this time had inoculated neck skin positive on rRT-PCR only, with no viral detection with VI, ISH, or IHC. At no time were the tissues of any of these 3 animals positive by IHC or ISH, nor was any evidence of epidermal disruption seen on HE. All other tissues as well as serum collected from these animals were consistently negative for the presence of virus by all techniques (data not shown).
Clinical and Gross Findings in Animals Inoculated in CB
All experimentally infected animals in this group developed fever (>39.2°C) and mild depression between 24 and 48 HPI. No other clinical abnormalities were noted. In the case of the SC-inoculated animals, at 12 HPI, the only lesion was a mild hyperemia along the line of inoculation. Subtle gross lesions, including mild edema and hyperemia, were present at the site of inoculation at 24 HPI in both SC- and FB-inoculated animals. At 48 HPI there was moderate edema and hyperemia restricted to the site of SC. However, the FB-inoculated steer euthanatized at the same time had much more prominent hyperemia and swelling of the right (infected) rear CB, with a few 1- to 2-cm diameter, dark and circumscribed areas of ulcerations that coincided with the sites where the flies were feeding. In addition, in a region beyond where the cage was positioned, in the right rear interdigital space, this animal had marked vesicular formation, characterized by a focally extensive white and elevated area that exuded serous fluid when cut. No gross lesions were present at 48 HPI in the right (infected) front FB-inoculated CB, and analysis of fly blood meals revealed that no flies fed on this foot. At 72 and 96 HPI, the right CB infected by SC had moderate hyperemia and edema, multifocal areas of hemorrhage, and multiple 2- to 3-mm ulcerations corresponding to zones of SC. A serous fluid frequently oozed from the small ulcers. However, at the same time points (72 and 96 HPI), the lesions in the FB-inoculated animals were more severe, with diffuse edema, hyperemia, and a few dark 2- to 3-cm diameter ulcerated areas corresponding to where fly cages were positioned at the time of inoculation. In these FB-inoculated animals, lesions extended far beyond the site of FB with vesicles reaching the interdigital space. At 120 HPI, the SC lesions were similar to or slightly more severe than those seen in the 96 HPI SC steer. In the mock-inoculated animals, CBs developed a mild hyperemia at the site of SC or FB at 24 HPI and occasionally at 48 HPI. The 96 HPI FB and the 120 HPI SC had moderately to markedly enlarged popliteal and prescapular LNs.
Histological Findings in Animals Inoculated in CB
Only 1 SC-inoculated steer was euthanatized at 12 HPI. In this animal in both right and left (virus inoculated and mock inoculated, respectively) CB, small areas of serocellular crusting with underlying epidermal disruption developed, characterized by poorly defined focal eosinophilic areas with loss of definition of cellular borders in the upper strata spinosum and granulosum. There was focal loss of the epidermis at the areas of scarification (Fig. 3a ). There was mild to moderate focal neutrophilic exocytosis in the epidermis. The superficial dermis had focal perivascular infiltration of numerous neutrophils and fewer macrophages.
At 24 HPI, both the SC- and FB-inoculated animals presented large coalescing areas of mild to moderate intercellular edema (spongiosis) with stretching of intercellular bridges and shrunken keratinocytes in the deep stratum spinosum (Fig. 4a). Cells from the strata basale and granulosum were consistently spared at this time. There was mild neutrophilic exocytosis throughout the affected epidermis, with moderate numbers of neutrophils and macrophages around perivascular areas of the superficial dermis.
At 48 HPI in the SC-inoculated animal, the narrow zone of inoculated haired CB skin had necrosis of epithelium of the epidermis and hair follicles with accompanying neutrophilic infiltration; the thicker nonhaired skin closer to the hoof had severe intercellular edema and hypereosinophilic, pyknotic, and shrunken keratinocytes limited to the stratum spinosum (Fig. 5a). Multifocal areas of this thick glabrous skin formed cavitations partially filled with numerous neutrophils, fewer macrophages, cell debris, and acantholytic cells. Similar to what was observed at 24 HPI, the strata basale and granulosum were consistently spared. The superficial dermis had moderate perivascular to interstitial infiltration of numerous macrophages and neutrophils, with fewer lymphocytes.
The FB-inoculated steer at 48 HPI had minimal lesions in the right front CB, whereas the right rear CB had severe epidermal disruption affecting a much larger zone of haired skin. These lesions were more severe than those observed in the SC-inoculated animal at this time. Hair follicles were frequently effaced by marked epidermal necrosis and accumulation of degenerated neutrophils. In the nonhaired thick skin (located immediately distal the haired skin but not in contact with the flies), there were extensive areas of full-thickness necrosis from the stratum basale to the stratum granulosum, with much greater damage than seen in the SC-inoculated animal at this time point. Within the stratum spinosum were multiple large coalescing cavitations filled with numerous neutrophils, few macrophages, cell debris, and occasional acantholytic cells. There was cleft formation at the stratum basale, with separation of this layer from the stratum spinosum. The dermis had moderate to severe perivascular to interstitial infiltration of numerous macrophages and neutrophils, with fewer lymphocytes and infrequent mast cells.
By 72 HPI, there was extensive necrosis of the thick nonhaired skin at the CB of the SC-inoculated animal, characterized by large pale eosinophilic areas, with loss of cellular detail (Fig. 6a). Areas of necrosis were centered in the stratum spinosum, spared the stratum basale, and extended to but did not efface the stratum granulosum. Numerous neutrophils infiltrated through the intercellular spaces around necrotic keratinocytes and filled areas where there was loss of keratinocytes. The dermis had moderate to severe perivascular to interstitial infiltration of numerous macrophages and neutrophils, with fewer lymphocytes. The animal inoculated by FB had lesions similar to those observed in the FB-inoculated steer at 48 HPI, with more extensive areas of necrosis and loss of epidermis.
At 96 HPI, epidermal lesions in the animals inoculated by either method were similar. There were full-thickness necrosis and cleft formations at the level under the stratum basale, with frequent loss of the entire epidermis of the CB (ulceration).
At 120 HPI, lesions were similar to those observed at 96 HPI but showed evidence of initial epidermal regeneration, in which the basal layer contained numerous and prominent digitiform projections composed of hyperplastic epithelial cells extending into the dermis.
The lymph nodes draining the right (inoculated) CBs presented with nonspecific reactive changes. No difference was observed between SC- and FB-inoculated animals. Mild paracortical hyperplasia started at 24 HPI, with gradual development of more paracortical and also medullary cord hyperplasia from 72 to 120 HPI.
Detection of Virus in Animals Inoculated in the Coronary Bands
All results from animals inoculated in the CBs are shown online in a supplementary table (to view online supplement, please go to http://www.vet.sagepub.com/supplement). Detection of VSNJV by VI and rRT-PCR was restricted to the site of inoculation and draining LNs. All samples from mock-inoculated animals and left mock-inoculated CB (on the left side), tongue, tonsil, spleen, and serum from experimental animals were consistently negative by these techniques in all animals using either method of inoculation. For the right CBs (inoculated side), positive VI and low rRT-PCR CT values were positive at all times post inoculation, from 12 to 120 HPI, in both FB- and SC-inoculated groups of animals. However, the right front CB from a steer sampled at 48 HPI, where inoculation with infected black flies was attempted, was negative on VI and only weakly positive (high CT value) on the rRT-PCR. Flies from this site were negative for mammalian DNA, indicating that flies fed very poorly, if at all. For the draining lymph nodes of inoculated animals, live virus was isolated as early as 12 HPI in a scarified animal and no later than 72 HPI in both animal groups. The rRT-PCR results from the draining lymph nodes tended to have much higher CT values (lesser amount of virus) than the inoculated CBs.
Detection of Vesicular stomatitis New Jersey virus (VSNJV) by Virus Isolation (VI), Reverse Transcriptase Real Time PCR (rRT-PCR), In Situ Hybridization (ISH), and Immunohistochemistry (IHC) in Steers Inoculated by Scarification or Fly Bite in the Coronary Bands (CB) and Euthanatized at Different Times Post Inoculation (HPI).
N, negative; NA, not applicable (there were no animals in those time points); NP, not processed; P, positive. rRT-PCR results expressed as cycle threshold value (CT value <40 was considered positive for the presence of virus).
In Situ Hybridization for Animals Inoculated in CB
Using the negative sense probe, which is complementary to the viral messenger RNA (mRNA) and antigenome strand (which are both sense/positive strands), for the N gene resulted in segmentally diffuse staining of the keratinocyte cytoplasm in the inoculated CBs. The SC-inoculated steer at 12 HPI had ISH signal from the stratum basale extending up into the midstratum spinosum (Fig. 3b). Signal distribution and intensity varied from darkly distinct small foci at 12 HPI to extensive coalescing large areas at 24 HPI (Fig. 4b) for both methods of inoculation. The signal had the same intensity at 48 HPI, but the positive keratinocytes were scattered within (Fig. 5b) or located around the periphery of necrotic areas (vesicle) for both SC and FB animals. At 72 HPI, both SC- and FB-inoculated animals had positive keratinocytes with fainter but diffuse signal throughout the cytoplasm (Fig. 6b). When we compared the 2 methods of inoculation at 48 HPI, the location of ISH labeling was distinct between SC- (Fig. 7 ) and FB-inoculated (Fig. 8 ) animals. For both, the ISH signal was scattered within the stratum spinosum and restricted to the CB epithelium. However, in the FB-inoculated steer, ISH signal was deeper, occurring in the stratum basale of the CB and expanding beyond the fly-feeding area into the interdigital region, with cleft formation under the level of the stratum basale of both CB and interdigital skin (Fig. 8). ISH-positive keratinocytes were present throughout the necrotic epidermis. At 96 HPI, SC- and FB-inoculated animals had faint diffuse signal in the cytoplasm of keratinocytes, as did the SC-inoculated animal at 120 HPI. Positivity was multifocal and located more frequently at the periphery of necrotic areas.
Use of a negative sense probe to detect replicating virus in the draining LNs highlighted scattered large cells with abundant cytoplasm, morphology suggestive of dendritic cells or macrophages, in subcapsular areas of cortex and paracortex regions at 24 HPI only in both SC- and FB-inoculated animals (Fig. 9 ).
Immunohistochemistry in Animals Inoculated in the Coronary Bands
There was signal representing viral protein in the cytoplasm and in the plasma membrane of keratinocytes in the inoculated CB. This signal was consistently restricted to the epidermis. Epidermal location of the signal followed in a similar fashion with the development of the histological lesion described above, except for the 12 HPI scarified steer, which had IHC signal from the stratum basale up through the midstratum spinosum (similar to the ISH staining). The IHC signal was present at later time points than ISH signal but was otherwise similar in the epidermis, yet it had many similarities to that seen with negative sense riboprobe ISH. That is, staining intensity varied from focal and intense, being restricted to small foci at 12 HPI (Fig. 3c) to more extensive, affecting large coalescing areas from 24 until 72 HPI (Figs. 4c, 5c, and 6c), after which it became fainter and more localized at 96 and 120 HPI. However, the patterns of IHC and ISH diverged in the lymph nodes. In contrast to what was observed with ISH, there was IHC positivity until 120 HPI in the draining lymph nodes, with positive cells in the cortical subcapsular regions at 24 HPI and in perifollicular and follicular regions at 48, 72 (Fig. 10 ), 96, and 120 HPI. On sections of 24 HPI lymph nodes, IHC specific for dendritic cells (anti-CD1b antibody) highlighted the same types of cells in the same areas as were positive by ISH at 24 HPI (data not shown).
Discussion
In this study, acute clinical disease and typical lesion development were successfully reproduced in cattle experimentally infected with VSNJV by application of virus to a scarified region on the CB and by feeding of VSNJV-infected black flies at the CB. The kinetics of viral replication and dissemination that occur during acute stages of VSNJV infection were analyzed using IHC for detection of viral protein, VI, rRT-PCR, and ISH for detection of viral replication (negative sense riboprobe ISH). Intense ISH signal was observed in CB keratinocytes from 24 to 48 HPI using the negative-sense riboprobe, which recognizes mRNA and replicative intermediate RNA, indicating that viral replication is taking place in these cells. In the same tissues, IHC labeling of viral proteins was observed beginning at 12 HPI and increasing through 72 HPI, with the signal becoming more localized and faint from 96 to 120 HPI. Therefore, these results demonstrate for the first time that the in vivo peak of viral replication occurs between 24 and 48 HPI and demonstrate clearly that the keratinocytes from the CB are permissive cells for early viral replication. The fact that the intensity of IHC signal persists until 72 HPI, with faint ISH labeling at that time, is likely associated with the natural cycle of viral replication, with virus replicating early and viral protein remaining in the cells longer. The presence of virus and therefore potential viral shedding from CBs persisted to the end of the experiment, 120 HPI, as shown by VI and rRT-PCR.
In the local draining lymph nodes of animals inoculated in the CBs, whether cattle were infected by SC or FB, replicating VSNJV was detected for the first time via ISH (negative sense riboprobe) and was restricted to the 24 HPI time point. ISH signal was scattered and distributed predominantly in the subcapsular areas within individual cells with morphology suggestive of macrophages or dendritic cells. On adjacent sections, anti-CD1b antibody highlighted cells with similar morphology and distribution of those containing replicating virus at 24 HPI. Morphologic features and anti-CD1b labeling in subcapsular sinuses strongly suggest that these cells are dendritic cells. With the negative sense riboprobe, any positive signal represents the presence of positive sense RNA, which would be either a replicative intermediate of the virus or viral mRNA. In either case, positive ISH signal is an indication of active viral replication in the lymph node. In the LNs, live virus or its nucleic acid was detectable by VI and/or rRT-PCR for up to 72 or 120 HPI, respectively. However, neither of these parameters provides definitive evidence that the virus is replicating there, as virus could be passively carried to the lymph node. So this study is the first to show VSV replication beyond the site of the vesicle in cattle.
Both SC and FB inoculation methods resulted in disease when introduced at the CB, and the character and time course of infection had some definite similarities. However, there were also notable differences, with more severe and extensive gross and histological lesions in the FB-inoculated animals.
Feeding of infected black flies at the CBs could reproduce the disease, although the amount of virus contained within the total FB inoculum is far less than with SC. Previous pilot studies demonstrated that inoculation of virus via SC with smaller amounts than 107 TCID50 per animal, as used in this study, did not consistently produce clinical disease in cattle (L. Rodriguez, personal communication) or pigs. 16 The amount of VSNJV shed in saliva when a black fly is probing and blood feeding is not known; however, in previous studies that used an in vitro method to collect fly saliva, 9 VSNJV titers between 101.5 and 102.34 plaque-forming units (pfu)/ml were found in the saliva of black flies infected in the same manner as in our experiment. Flies probably cannot inoculate much more than 10–30 µl of saliva each. In this experiment, blood meal analysis revealed that only 6 flies out of 30 per cage participated in feeding. Consequently the total amount of inoculated fly saliva is very low. Although it is technically not feasible to quantitatively compare pfu with TCID50, it would seem that a total dose of less than 200 pfu is considerably less than 107 TCID50. Even with this lower dose, the lesion severity increased earlier and more rapidly in the FB compared with the SC animals. Therefore, it appears that infection by FB can be successful with markedly decreased amounts of virus, compared with SC, which suggests that there may be a facilitating factor supplied by the insect inoculation.
In this study, we also inoculated virus into neck skin using FB and examined the tissues using VI, HE, IHC, and ISH. Previous VSNJV studies in cattle 14 with SC of the CBs and flank skin could produce lesions only in the inoculated CBs, with no lesion development in the flank or in any other tissue. 14 However, it has been shown that noninfected black flies physically separated from VSNJV-infected black flies became infected while cofeeding at these nonvesicular sites on the same nonviremic host. 10 Our findings confirmed the lack of gross lesion development and demonstrated no histological viral-associated lesions. In addition, none of the multiple examined sections were positive either by IHC or ISH, although there was limited positivity using VI and rRT-PCR. Therefore, presence of the virus in this type of tissue is probably short and transient, although enough to be transmitted to other flies, with no resultant vesicle development or detectable virus by ISH or IHC.
Lesions developed only when the inoculation was in the CB, and they were restricted to the site of inoculation (CB), with viral replication and dissemination to the draining lymph node. These data indicate that the characteristics of the epithelium may well play a pivotal role in susceptibility to the virus infection.
The epidermis of the CBs and the epidermis of other areas where vesicular lesions develop, such as oral cavity (tongue, hard palate and gums), muzzle, teats, and prepuce, share similar histological and physiological features. These areas are unhaired, have thicker epidermis with a prominent stratum spinosum, and are more subjected to mechanical stress. These areas require higher amounts and more specialized intercellular adhesion molecules for proper tissue cohesiveness. In contrast, the neck and the flank skin are haired and have a markedly thinner epidermis compared with those areas where VSNJV replicates and forms vesicles, with less mechanical stress and therefore decreased concentrations of intercellular bridges. Comparing the morphophysiological features of the anatomical locations where VSNJV lesions develop to those sites where vesicles do not form, it seems plausible that the virus uses cell membrane receptors (perhaps intercellular junction molecules) for viral entry and/or uses part of specialized cell pathways for viral replication.
In summary, our study demonstrates that VSNJV replicates successfully and extensively in the keratinocytes from specific anatomical locations with thick unhaired skin. Vesicular lesions were limited to the site of inoculation, with peak of viral replication between 24 and 48 HPI in the CB. Replicating virus was detected in draining lymph nodes but only at 24 HPI. Lesion development was markedly more severe and extensive when the virus was transmitted via fly bite, which suggests that components in the saliva facilitate VSNJV infection.
Footnotes
Acknowledgements
We acknowledge Jian Zhang from the Pathology Department, UGA, for development of ISH probe; Paul Smith from the Entomology Department, UGA, for infecting black flies and helping with animal inoculation; Ethan Hartwig from Plum Island, ARS-USDA, for viral isolation; and Steve Pauszek and Jonathan Artz from Plum Island, ARS-USDA, for animal inoculation and necropsy assistance respectively.
The authors declared that they had no conflicts of interest with respect to their authorship or the publication of this article.
The CAPES–Fulbright Doctoral Exchange program and the USDA-ARS Plum Island Animal Disease Center provided funding for this project.