Vesicular stomatitis New Jersey virus (VSNJV) infects

HAL Id: hal-00902859https://hal.archives-ouvertes.fr/hal-00902859

Submitted on 1 Jan 2007

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

Vesicular stomatitis New Jersey virus (VSNJV) infectskeratinocytes and is restricted to lesion sites and local

lymph nodes in the bovine, a natural hostCharles Scherer, Vivian O’Donnell, William Golde, Douglas Gregg, D. Mark

Estes, Luis Rodriguez

To cite this version:Charles Scherer, Vivian O’Donnell, William Golde, Douglas Gregg, D. Mark Estes, et al.. Vesicularstomatitis New Jersey virus (VSNJV) infects keratinocytes and is restricted to lesion sites and locallymph nodes in the bovine, a natural host. Veterinary Research, BioMed Central, 2007, 38 (3),pp.375-390. �10.1051/vetres:2007001�. �hal-00902859�

https://hal.archives-ouvertes.fr/hal-00902859

https://hal.archives-ouvertes.fr

Vet. Res. 38 (2007) 375–390 375c© INRA, EDP Sciences, 2007DOI: 10.1051/vetres:2007001

Original article

Vesicular stomatitis New Jersey virus (VSNJV)infects keratinocytes and is restricted to lesion sitesand local lymph nodes in the bovine, a natural host

Charles F.C. Sa,b,c, Vivian O’Da,d, William T. Ga,Douglas Ga, D. Mark Eb,c, Luis L. Ra*

a Plum Island Animal Disease Center, Agricultural Research Service, United States Department ofAgriculture, PO Box 848, Greenport, NY 11944, USA

b University of Missouri, College of Veterinary Medicine, Department of Veterinary Pathobiology,Columbia, MO 65251, USA

c Department of Microbiology and Immunology, University of Texas Medical Branch,Galveston, TX 77555, USA

d Department of Pathobiology, University of Connecticut, Storrs CT 06269, USA

(Received 16 August 2006; accepted 19 October 2006)

Abstract – Inoculation of vesicular stomatitis New Jersey virus (VSNJV) by skin scarification ofthe coronary-band in cattle, a natural host of VSNJV, resulted in vesicular lesions and 6−8 log10TCID50 increase in skin virus titers over a 72 h period. Virus infection was restricted to the lesionsites and lymph nodes draining those areas but no virus or viral RNA was found in the blood orin 20 other organs and tissues sampled at necropsy. Scarification of flank skin did not result inlesions or a significant increase in viral titer indicating that viral clinical infection is restricted toskin inoculation at sites where lesions naturally occur. Viral antigens co-localized primarily withkeratinocytes in the coronary band, suggesting these cells are the primary site of viral replication.Viral antigen also co-localized with few MHC-II positive cells, but no co-localization was observedin cells positive for macrophage markers. Although granulocyte infiltration was observed in lesions,little viral antigen co-localized with these cells. This is the first detailed description of VSNJV tissuedistribution and infected cell characterization in a natural host. The pathogenesis model shownherein could be useful for in-vivo tracking of virus infection and local immune responses.

vesicular stomatitis / bovine / pathogenesis / confocal microscopy / keratinocytes

1. INTRODUCTION

Vesicular stomatitis virus (VSV) iswidely used as a laboratory research toolfor RNA virus evolution and to evaluateimmune function. More recently it has had

* Corresponding author:[email protected]

application as a vector for experimentalvaccine delivery and for anti-tumor ther-apy [1,18]. In the nature VSV is an impor-tant livestock pathogen causing vesicularstomatitis (VS) a disease characterized bythe appearance of vesicular lesions in themouth, feet and udders, of cattle, pigsand horses. In cattle and swine theselesions are clinically undistinguishable

Article available at http://www.edpsciences.org/vetres or http://dx.doi.org/10.1051/vetres:2007001

http://www.edpsciences.org/vetres

http://dx.doi.org/10.1051/vetres:2007001

376 C.F.C. Scherer et al.

from foot-and-mouth disease, a devastat-ing disease of livestock [20]. VSV hasbeen shown to be transmitted by insectbites [6, 14, 27] but transmission by di-rect contact between animals has also beendemonstrated [15, 23]. The mechanisms ofdisease remain unclear, although field ob-servations indicate that many infectionsdo not result in overt clinical disease andmany susceptible species living in endemicareas possess neutralizing antibody titers toVSV [16, 19].

Most knowledge pertaining to VSVpathogenesis is derived from studies in-volving laboratory rodents where clini-cal signs are not vesicular in nature, butrather the infection manifests as encephali-tis and death depending on host factorssuch as age, route of inoculation and vi-ral strain [7, 9]. Neurological symptomshave not been reported in natural VSVhosts (cattle, swine, horses) [20]. The basicmechanisms of VSV infection remain un-clear, but experiments in swine suggest thatthe inoculation site determines clinical out-come. Vesicular lesions are observed onlywhen the virus is intradermally inoculatedat specific sites where lesions are observedduring natural infections (snout, lip, feet)and instead subclinical infection occurswhen virus is inoculated intradermally atother sites (i.e. ear, abdomen) or by theintranasal or intravenous routes [11, 15].Recently, it was further shown that vesic-ular lesions only developed when exper-imentally infected black flies (Simuliumvittatum) were allowed to feed on the snoutbut not when flies fed on the abdomen ofswine [14].

Little is known about the basic cellu-lar and molecular mechanisms mediatingVSV pathogenesis in its natural hosts. In-formation regarding the primary sites ofvirus replication and the cell types involvedin supporting viral growth and those in-volved in controlling the infection remainscarce. Previous studies on VSV inocu-lation in natural hosts have been limited

to the description of gross pathology andhistopathology of the lesions without spe-cific identification of cell types involved inearly viral infection.

This study describes early events ofvesicular stomatitis New Jersey virus (VS-NJV) infection in cattle utilizing a novelcoronary band scarification inoculationmodel combined with analyses by im-munohistochemistry, confocal microscopyand real-time RT-PCR. The tissue distribu-tion of virus and the identity of the celltypes infected during early phases of dis-ease are described.

2. MATERIALS AND METHODS

2.1. Animals and virus

Adult (18 to 24 months) Holsteinsteers weighing 500−700 lb were obtainedfrom an experimental-livestock provider(Thomas-Morris Inc., PA, USA) and keptin the biosafety level 3 facility at PlumIsland Animal Disease Center for at leastone week prior to initiation of the exper-iments. All animal inoculations were per-formed with a VSNJV field strain obtainedfrom tongue epithelium of a bovine nat-urally infected during the 1995 epidemicin Colorado (95COB). This virus wasidentified as VSNJV by virus neutraliza-tion and sequencing of the complete viralgenome [21]. The virus was propagated bypassing once in baby hamster kidney cells(BHK-21) infected at 0.01 multiplicity ofinfection. Viral stock was titrated in BHK-21 cells, and kept in aliquots at −70 ◦C.

2.2. Inoculation procedure

Animals were sedated with xylazine andthe coronary band areas were shaved, priorto the epidermis being pricked 20 times us-ing a dual tip skin test applicator (Duotip-Test, Lincoln Diagnostics, Decatur, IL,

Vesicular stomatitis: pathogenesis in cattle 377

USA). Virus inoculum was placed on thescarified area in 100 µL of Dulbecco Mod-ified Eagle Medium containing 2% fe-tal bovine serum (FBS) (DMEM2). Thearea under the inoculum was then scar-ified 20 additional times with the ani-mals restrained in a stationary position for1−2 min until the inoculum was adsorbed.A total of thirteen animals were used in thisstudy. Six animals, housed in three sep-arate rooms, received 107 TCID50/foot ineach of all four feet; one animal (#102) wasinoculated only on the right feet and mock-inoculated on the left feet. Animals 12and 148 were inoculated with 107 TCID50of VSNJV by intradermal injection or scar-ification on the skin of the flank and keptin separate rooms. One animal (#699) wasinoculated by injection of 107.0 TCID50distributed in four sites on the dorsal ep-ithelium of the tongue and house alone. Allinoculations were performed with the sameviral stock.

Three non-inoculated animals kept inseparate rooms were used as negative con-trols and sampled similarly. Clinical signs,temperature, appetite, and attitude wereevaluated daily for all animals. Clinicaldisease was scored by determining the sizeand number of lesions, a value of 1 wasgiven to small lesions and a value of 2 forlarge lesions on each foot or on the tongue(a maximum score of 10 would indicatelarge vesicles in all feet and on the tongue).

2.3. Sampling

Punch biopsies were obtained from se-dated animals using disposable 6 mm skinbiopsy punch (Miltex, Inc., Bethpage, NY,USA) before inoculation, and at 6, 12, 24,48 and 72 hours post-inoculation (hpi).Two biopsies were taken at each timepoint from different feet; one was fixed in10% buffered formalin for 24 h before pro-cessing for histological examinations andthe other one was snap-frozen in liquid ni-

trogen for RNA extraction, virus isolationand confocal microscopy. Oropharyngealfluid (OPF), plasma and blood sampleswere also obtained and kept at −70 ◦C un-til being processed. In four of the animalsinoculated in the coronary band, biopsieswere also taken 20 min after inoculation.

Eight animals were euthanized at 72 hpiand the following tissues were collected forRT-PCR and virus isolation: prescapular,popliteal, axillary, mediastinal, mesenteric,iliac, prefemoral, retropharyngeal, parotid,and submandibular lymph nodes; coronaryband, tongue, tonsils, lung, heart, liver,spleen, kidney, small intestine (duode-num), large intestine (cecum), snout skin,mandibular salivary glands, nasal epithe-lium and brain (olfactory area). Animals 12and 148, inoculated on the flank, were eu-thanized at 48 hpi and only skin samplesand major lymph nodes were collected.

2.4. Virus isolation

Tissues were macerated using a mor-tar and pestle in 2 mL of Minimal Es-sential Medium containing 400 U/mL ofpenicillin, 400 U/mL of streptomycin, and10 µg/mL of amphotericin B (MEM). Mac-erated samples were centrifuged for 5 minand dilutions of the clarified supernatantswere inoculated onto monolayers of BHK-21 cells as previously described [8]. Blood,plasma and OPF samples were tested sim-ilarly by direct inoculation of cell mono-layers with serial sample dilutions. Mono-layers were rinsed with MEM 2 hpi andreplaced with fresh medium. Cultures wereobserved for cytopathic effect (CPE) at 24,48 and 72 hpi. After 72 h supernatant fromeach CPE-positive well was saved and thepresence of VSNJV antigens was con-firmed by direct staining of the fixed cellsusing biotinylated VSNJV-specific anti-bodies as previously described [12]. Flasksnegative for CPE after 72 h were frozen,


thawed and supernatants were reconfirmednegative by real-time RT-PCR.

2.5. Immunohistochemistry

Immunohistochemistry (IHC) was per-formed as previously described by Suret al. [25]. Briefly, 3 µM paraffin sectionswere placed on ProbeOnTM Plus slides(Fisher Scientific, Pittsburgh, PA, USA),deparaffinized in xylene and dehydratedin graded alcohol. Antigen retrieval wasdone in two ways, either using citrate so-lution (DAKO Cytomation, Denmark) for5 min in an autoclave or 3 cycles of3 min at medium power in a microwaveusing 0.1 M Tris-HCL pH 8.0−8.2 as de-scribed by Tanimoto and Ohtsuki [26].Tissues were blocked with a 5% solutionof normal rabbit serum in PBS contain-ing 0.01% Tween 20 (PBST) for 10 minat room temperature, washed in PBST andincubated overnight at 4 ◦C with anti-VSNJV guinea pig antibody (1:5000 di-lution in PBST) in a humid chamber. Af-ter three washes with PBST, a secondaryanti-guinea pig IgG conjugated with per-oxidase (Vector Laboratories, Burlingame,CA, USA) was applied according to themanufacturer’s instructions. After a finalrinse with 0.1 M Tris pH 8.0, substratesolution was applied (Vector Red kit, Vec-tor Laboratories) following the manufac-turer’s protocol. After 20 min, the slideswere washed and counterstained withGill Haematoxylin. Sections of VSNJV-infected tissue were used as positive con-trols and sections of mock inoculated tis-sues were used as the negative control.

2.6. Confocal immunofluorescencemicroscopy

For confocal microscopy, 3−5 µmthick sections of cryopreserved tissueswere sectioned with a cryomicrotome

and fixed with acetone for 10 minat −20 ◦C. After fixation, the sectionswere blocked for 1.5 h in PBS, 20% fetalbovine serum, 2% BSA (blocking buffer)at 37 ◦C. Primary antibodies, anti-VSNJV(guinea pig polyclonal, 1/1000), anti-cytokeratin (a marker for keratinocytes,IgG2a, clone K8.13, Sigma, 1/100), anti-human HLA-DR (a marker for dendriticcells, B-cells, macrophages and mono-cytes, IgG1, clone 1B5, DAKO Cytoma-tion, 1/500), or MAC-387 (a markerfor granulocytes, monocytes and tissuemacrophages, DAKO Cytomation, 1/200)were diluted in blocking buffer and incu-bated with the slides overnight at 4 ◦Cin a humid chamber. When double label-ing was performed, the slides were incu-bated with both antibodies together. Afterbeing washed five times with PBS, theslides were incubated with the appropri-ate secondary antibodies; goat anti-guineapig (1/400, Alexa Fluor 594, MolecularProbes, Eugene OR, USA), or goat anti-mouse isotype specific (1/400, Alexa Fluor488, Molecular Probes); and diluted inblocking buffer for 1.5 h at 37 ◦C. Fol-lowing this incubation, the slides werewashed five times with PBS, counter-stained with the nuclear staining TOPRO-iodide 642/661 (Molecular Probes) for5 min at RT, mounted and examined us-ing a Leica scanning confocal microscope.Data were collected using appropriate con-trols lacking the primary antibodies, aswell as using uninfected sections to givethe negative background levels. The cap-tured images were adjusted for contrastand brightness using Adobe Photoshopsoftware.

2.7. RNA extraction and QuantitativeReal Time PCR (Q-RT-PCR)

Total RNA extraction from tissues,plasma, and OPF samples was done usingTRIZOL reagent (Invitrogen Corporation,


Carlibad, CA, USA) following the pro-tocol supplied by the manufacturer. TotalRNA from PBMC was extracted usingthe Qiagen RNEasy kit (Qiagen, Valen-cia, CA, USA), per manufacturer’s proto-col. RNA samples were kept at −20 ◦Cuntil analysis. Semi-quantitative real-timePCR specific for VSNJV was done using anucleocapsid-specific test with the follow-ing primers: forward 5’GCACTTCCTG-ATGGGAAATCA3’, reverse 5’GGGAA-GCCATTTATCCTCA3’ and FAM-labeledprobe 5’ACCCTGACCGTTCTG3’ (Ap-plied Biosystems, Foster City, CA, USA).The RT-PCR reaction was done using theTaqMan� EZ RT PCR core reagent kit(Applied Biosystems), containing 300 nMof each primer, 25 mM Manganese acetate,10 mM dNTP with 20 mM of dUTP, 2.5 Uof rTth polymerase, 0.25 U of AmpEraseUNG�, 2 µL of template RNA and rnase-free water in a 25 µL reaction volume.The cycling profile was as follows: 50 ◦Cfor 2 min, 58 ◦C for 30 min for reversetranscription, hold for 95 ◦C for 5 min toinactivate AmpErase and then 40 cyclesof 95 ◦C for 20 s and 60 ◦C for 1 min.The results were expressed as Ct values.Relative sensitivity of the rRT-PCR wasdetermined to be 8 TCID50 utilizing se-rial dilutions infected cell supernatant andcomparing level of detection by rRT-PCRand virus isolation utilizing BHK-21 cellsas described above (see Appendix A).

3. RESULTS

3.1. Clinical outcome and virusdistribution

Animals inoculated in the coronarybands exhibited transient fever (≥ 40 ◦C)and blanching on the entire extension ofthe coronary bands by 24 hpi followedby vesicles in the epithelium of the coro-nary band by 48 hpi. All coronary band-inoculated animals developed vesicular le-

sions in the inoculated feet by 48 hpi re-sulting in clinical scores of 8 accordingto the scale described in the materials andmethods above. Animal #102 inoculatedonly on the right feet developed lesionsonly at inoculation sites and had the max-imum score of 4. The animal inoculatedin the tongue showed blanching and fluid-filled vesicles in the tongue by 24 h thatwere ruptured by 48 h with epithelium losson the tongue, leaving a large eroded sur-face. No lesions were observed at othersites. Fever and/or lesions were absent inthe animals inoculated in the skin of theflank or neck; no clinical signs were ob-served in mock-inoculated animals.

Virus was not detected in whole blood,or plasma obtained at 0, 12, 24, 48 and72 hpi from nine inoculated animals re-gardless of the route of inoculation ortime after infection. OPF was negativeas most coronary band or flank skin-inoculated animals, except in animals 30and 31, where virus was detected in OPFat 72 hpi. Tongue-inoculated animal 699had vesicular lesions on the tongue; there-fore OPF samples were not obtained sincethey would become contaminated by virusin the mouth.

Cattle were euthanized and various tis-sues were collected and analyzed both byvirus isolation and real-time RT-PCR. In4 of 6 coronary band-inoculated animals,virus was only found at the inoculation siteand at the primary draining lymph nodes;i.e. prescapular or popliteal draining thefront or the rear feet respectively. Coro-nary band-inoculated animals 30 and 31,in addition to the inoculation site, hadvirus in OPF but no lesions in the mouthor tongue, yet at necropsy, virus wasfound in tonsil and lymph nodes drainingthe mouth (retropharyngeal, and parotid)(Tab. I). In the case of animal 699, in-oculated in the tongue, VSNJV virus wasfound in the tongue epithelium, retropha-ryngeal, parotid and submandibular lymphnodes and in the tonsil, but not in coronary


Table I. Distribution of VSNJV in inoculated cattle determined by virus isolation and real-timeRT-PCR.

Inoculation site

Four coronary bands Right Flank/ Tongue

coronary neck

bands skin

Animal ID** 10 30 31 725 754 102 12 / 148** 699

Tissuea Virus Isolation positive samples (left/right when applicable)

Coronary band 2/2 2/2 2/2 2/2 2/2 0/2 0/0 0/0

Prescapular ln* 1/0 1/1 1/0 1/1 1/1 0/1 0/0 0/0

Popliteal ln* 1/1 1/0 1/1 1/1 1/1 0/1 0/0 0/0

Axillary ln# 0/0b 0/0 1/0 0/0 0/0 0/1d 0/0 0/0

Prefemoral ln# 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0

Tongue 0 0 0 0 0 0 0 1

Snout skin 0 0 0 0 0 0 n.d.c 0

Nasal epithelium 0/0 0/0 0/0 0/0 0/0 0/0 n.d. 0/0

Tonsil 0/0 1/0 0/1 0/0 0/0 0/0 0/0 1/1

Retropharyngeal ln 0/0 1/0 0/0 0/0 0/0 0/0 0/0 1/1

Parotid ln 0/0 1/0 0/0 0/0 0/0 0/0 0/0 1/1

Submandibular ln 0/0 0/0 0/0 0/0 0/0 0/0 0/0 1/1

Mandibular salivary gland 0/0 0/0 0/0 0/0 0/0 0/0 n.d. 0/0

Brain (Olfactory Bulb) 0/0 0/0 0/0 0/0 0/0 0/0 n.d. 0/0

Mediastinal ln 0 0 0 n.d. n.d. 0 n.d. 0

Mesenteric ln 0 0 0 0 0 0 n.d. 0

Iliac ln 0 0 0 n.d. n.d. 0 n.d. 0

Heart 0 0 0 n.d. n.d. 0 n.d. n.d.

Lung 0/0 0/0 0/0 0/0 0/0 0/0 n.d. 0/0

Kidney 0/0 0/0 0/0 0/0 0/0 0/0 n.d. 0/0

Liver 0 0 0 0 0 0 n.d. 0

Spleen 0 0 0 0 0 0 0 0

Payer patches 0 0 0 n.d. n.d. 0 n.d. 0

Small intestine/duodenum 0 0 0 n.d. n.d. 0 n.d. 0

Large intestine/colon 0 0 0 n.d. n.d. 0 n.d. 0

Flank/neck skin n.d. n.d. n.d. n.d. n.d. n.d. 0/1 n.d.

a All animals euthanized at 72 hpi except #12 euthanized at 48 hpi.b Negative in virus isolation.c n.d., not done.d Positive by RT-PCR only.* Primary draining lymph node (ln).# Secondary draining ln.** Neck skin inoculated animal (#148) had the same results as animal 12 except that residual virus wasfound on the inoculation site skin at necropsy (48 h).


Figure 1. A. Growth curve of VSNJV in skinfrom animals inoculated on the right coronarybands (#102) or on skin of the flank (#12).B. Growth curve of four additional animals in-oculated in the coronary bands.

bands or in lymph nodes draining the feet(Tab. I).

To explore the effect of the inoculationsite on lesion development and virus dis-tribution we inoculated three cattle; oneanimal was inoculated on the right coro-nary bands, leaving the left coronary bandsas mock-inoculated controls (animal 102),and two other animals were inoculated in-tradermally on the skin of the flank orneck, sites with VSV lesions are not re-ported (animals 12 and 148). Typical vesic-ular lesions and an increase in viral titerof 8 log10TCID50 by 48 hpi were ob-served only on the right coronary bandsof animal 102 (Fig. 1A). Similar resultswere observed in four other coronary band-inoculated animals, as shown in Figure 1B.

In contrast, intradermal inoculation of theflank or neck skin in animals 12 and 148(not shown) respectively, resulted in no le-sions and a virus titer of less than 3 log10TCID50 was detected in skin biopsies(Fig. 1A). Virus was recovered from theright coronary bands and draining lymphnodes in animal 102, but not from the left-side samples nor from any other organor lymph nodes from this animal or ani-mals 12 and 148 (Tab. I).

3.2. VSV distribution in infected tissues

Distribution of VSV in infected tissueswas determined by immunohistochemistryutilizing VSNJV-specific antibodies. In thecoronary band, virus antigens were primar-ily associated with the upper layers of theskin; particularly the “stratum spinosum”and “stratum granulosum”. Early after in-fection (6−12 h), only cells in the upperlayers stained for viral antigens; these cellswere mostly associated with micro vesi-cles that coalesced into larger vesicles by72 hpi (Fig. 2A). As vesicles became largerand filled with fluid (48 to 72 hpi), amixed population of inflammatory cells in-filtrated the lesion site and some distinctVSV-antigen containing cells with den-dritic cell-like morphology were observedin the lower layers of the dermis (Fig. 2Ainset). In the skin of the flank sporadic anti-gen positive cells were observed, mostlyassociated with hair follicles and connec-tive tissues in the dermis (Fig. 2B).

In the draining lymph nodes, fewantigen-positive cells were observed by72 hpi, mostly in the paracortex area andto a lesser extent in the trabeculae, man-tle and germinal centers (Fig. 3). In anearlier experiment, virus was detected byvirus isolation and real-time RT-PCR asearly as 24 hpi in draining lymph nodes butantigen-positive cells were not observeduntil 48-72 hpi (data not shown).


Figure 2. Skin sections of animals inoculated in the coronary band (A) or flank skin (B) stainedwith anti-VSNJV antibodies and counterstained with Gill Haematoxilin. VSNJV antigen-positivecells are shown in pink. Panel A shows a vesicular lesion at low (×100) and high magnification(×400) detail of the dermis area (inset). Panel B shows low (×100) and high (×400) magnificationof inoculated flank skin. Arrows mark VSNJV antigen-positive cells.

3.3. Characterization of infected cellsby immunofluorescence confocalmicroscopy

In order to characterize VSNJV antigen-positive cells in coronary band and flankskin, we utilized confocal microscopy anddouble staining with specific antibodiesagainst VSNJV and antibodies to three

cellular markers: cytokeratin (a markerfor keratinocytes), MHC-II clone TH14B(a marker for dendritic cells, B-cells,macrophages and monocytes) and MAC-387 (a marker for granulocytes, monocytesand tissue macrophages). The majority ofcells staining with VSNJV-specific anti-bodies in coronary bands were also posi-tive for cytokeratin staining suggesting that


Figure 3. Sections of prescapular lymph nodes from an animal inoculated with VSNJV 48 h earlier,stained with guinea pig antibodies to VSNJV and counterstained with Gill Haematoxilin. VSNJVantigen-positive cells are shown in the paracortex area (A) and in the trabeculae and germinal cen-ters (B).

keratinocytes are primary targets for vi-ral replication early during infection andsupport virus growth (Fig. 4). VSNJVantigen-positive cells of the middle and up-per stratum spinosum and in the stratumgranulosum of the coronary band showedstrong staining with anti-cytokeratin an-tibodies particularly in those areas withmicro vesicles. In flank skin, VSV stainingwas only observed at 24 hpi co-localizingwith cytokeratin-positive staining in cellsassociated with hair follicles (Fig. 4). By48 hpi VSV positive cells were no longerobserved in the flank (not shown).

To further characterize VSV infectedcells, we stained coronary band and To:Command not found.

flank skin tissue sections with antibod-ies to human MHC-II antigens. Despite thefact that a number of MHC-II-positive cellswere found in and around the coronaryband lesions, particularly at 48−72 hpi,only a few of these cells stained for VSVantigens and most of them were located inthe deep dermis (Fig. 5). In flank skin sec-tions MHCII-positive cells were observedbut no co-localization with VSV antigenswas observed (Fig. 5). These results wereconfirmed utilizing an antibody specific forbovine MHC-II antigens (not shown).

Cells stained by monoclonal antibodyMAC-387 were frequently observed in the

coronary band particularly after 48−72 hpi,indicating inflammatory cell infiltration.However, these cells did not stain forVSNJV antigen (Fig. 6). Few MAC-387-positive cells were observed in the flankskin at any time post inoculation and nonewere stained for VSNJV-antigen (Fig. 6).

4. DISCUSSION

Little is known about the mechanismsof VSV transmission, tissue tropism, virusdistribution and factors determining thelocalization of vesicular lesions in itsmost commonly affected natural host; thebovine. Pathogenesis studies in laboratorymice are not relevant to clinical diseasein cattle since mice present neurologicalclinical signs and not vesicular lesionslike those observed in livestock species.Early studies in cattle utilized tongue in-oculation, an unlikely route of infectionresulting from insect bite [5, 17, 22]. Theinoculation model described here is moresimilar to natural skin infection and is thefirst in cattle that consistently results invesicular lesions and at the same time al-lows sequential sampling of the skin andtracking of viral growth using skin punchbiopsies.


Figure 4. Localization of VSNJV antigen in frozen sections of tissues from infected steers. Cryosec-tioned tissues from the coronary band (48 hpi, a–c inoculated, g−i non-inoculated) and skin flank(24 hpi, d−f inoculated, j−l non-inoculated) were processed for immunofluorescence staining andconfocal microscopy with anti-VSNJV and anti-cytokeratin antibodies. VSNJV was visualized withAlexa Fluor 594 (red), cytokeratin was visualized with Alexa Fluor 488 (green). Cells were coun-terstained with TOPRO-iodide 642/661 (blue) to reveal the nuclei.


Figure 5. Localization of VSNJV antigen in frozen sections of tissues from two infected cattle at48 hpi. Cryosectioned tissues from inoculated coronary band (a−c inoculated, g−i non-inoculated)and inoculated skin of the neck (d−f inoculated, j−l non-inoculated) were processed for immunoflu-orescence staining and confocal microscopy with anti-VSV and anti-HLADR Class II antibodies.VSV was visualized with Alexa Fluor 594 (red), HLADR class II was visualized with Alexa Fluor488 (green). Cells were counterstained with TOPRO-iodide 642/661 (blue) to reveal the nuclei.


Figure 6. Localization of VSNJV antigen in frozen sections of tissues from two cattle, one inocu-lated on the coronary band and one on the skin of the flank 48 earlier. Cryosectioned tissues fromthe inoculated coronary band (upper panel) and inoculated skin of the flank (lower panel) were pro-cessed for immunofluorescence staining and confocal microscopy with anti-VSNJV and MAC-387antibodies. VSV was visualized with Alexa Fluor 594 (red), HLADR class II was visualized withAlexa Fluor 488 (green).

Virus distribution in infected animalshas been a long standing question that hasimplications not only for the pathogenesisand transmission of this virus but also forregulatory issues related to the approval ofVSV as a potential vaccine and anti-tumorvector and for resumption of trade and an-imal movement after quarantines imposedduring VSV outbreaks. Postmortem exam-ination demonstrated that VSNJV causedlocalized infections with virus recoveredonly from the inoculation site and regionaldraining lymph nodes, but not from bloodor internal organs. Furthermore, no viruswas detected on the left feet of animals in-oculated on the right side coronary bandsand only residual virus was detected onetime on a left foot, likely the result of sur-

face contamination from virus shedding ofthe right feet. Only in two out of sevencoronary band-inoculated cattle, was virusfound in the pharyngeal fluid, tonsils orlymph nodes draining the oropharynx. Thisis an interesting finding since tonsils seemto be an important site for viral replicationin swine but had not previously been re-ported in cattle [2,11,14,23]. We could notdetermine the source of virus in these twocattle, but it is possible that their mouthcame in close contact with their inoculatedfeet during the experiment.

Virus was not detected in whole bloodor in plasma at any time post inocula-tion. This result confirms previous reportsin swine, horses and cattle [10, 11, 14]. Itcould be argued that the samples tested had


low titers or that infectious virus was notdetectable due to virus inhibitory factorssuch as interferon, in the blood. However,this is unlikely the case since we failedto detect viral RNA using a real-time RT-PCR capable of detecting 8 TCID50 of VS-NJV in “spiked” normal blood and plasmasamples (Appendix A). Therefore, we hy-pothesize that the mechanism by whichvirus traveled from the inoculation site tothe regional lymph nodes was cell associ-ated via the lymphatic system. It is alsopossible that the virus was cell-associatedin the blood in amounts undetectable byreal-time RT-PCR. The lack of a viremicphase in cattle for an arthropod-borne viruslike VSV is puzzling. However, it is pos-sible that livestock are dead-end hostsand that other mammalian hosts mightbe responsible for maintaining the naturalvirus-insect cycle [3, 4]. Alternatively, re-cent studies have shown horizontal insect-to-insect transmission of VSV while co-feeding in non-viremic mammalian hoststheoretically making viremia unnecessaryfor insect to insect transmission [13].

The localized nature of VSNJV infec-tion in cattle was also confirmed aftertongue inoculation, where in postmortemexamination virus was only found in thetongue and head associated tissues, but notin coronary bands or other organs. Theseresults were consistent with field clinicalobservations in cattle where lesions rarelyoccur at more than one site, but contrastthose in swine, where lesions in the mouthand feet are commonly observed both infield infections and laboratory infected an-imals [12, 23]. The basis of this differenceis not clear but swine are known to shedvirus in saliva (from tonsil infection) forextended periods and numerous skin abra-sions occur during fighting when housed ingroups together [24].

In natural VSV infections, vesicularlesions appear at specific sites includ-ing the feet, mouth, or teats in lactatinganimals [28]. Previous studies in swine,

showed that only intradermal inoculationof the snout or coronary band resultedin lesion formation [11]. Similar resultswere obtained by inoculation with blackflies, with skin lesions observed only whenflies were allowed to feed on the skin ofthe snout but not when they fed on theskin of the abdomen in swine [14]. Wedemonstrate that in cattle, the site of skininoculation not only determines the clinicaloutcome but also the ability of VSNJV toinduce local replication. Inoculation of theflank skin resulted in no lesions and littleviral replication at the inoculation site. Themechanism of this restriction remains un-clear, but we showed a marked contrast indistribution of virus antigen-positive cellsbetween the coronary band, where ex-tensive replication occurs and flank skinwhere replication is limited and infectiondoes not progress to clinical lesions. Incoronary bands there was a clear associ-ation of viral antigen with keratinocyteslocated in the upper layers of the epidermiswhere vesicular lesions later developed,whereas in flank skin, only few antigen-positive keratinocytes were transiently ob-served in association with hair follicles.There are important histological and struc-tural differences between the skin of thecoronary band and the flank skin that mayexplain this difference in supporting vi-ral growth. The coronary band skin hasmultiple layers of keratinocytes formingthick epidermal layers (stratum basale andstratum spinosum), while the flank skinhas thin epidermal layers with a smallernumber of keratinocytes that may not besufficient to support viral replication andvesicle formation.

This study is the first detailed descrip-tion of VSV pathogenesis in cattle, themost frequently affected species duringVSV outbreaks. The inoculation modelwill be useful in future pathogenesis stud-ies, such as determining the cellular andmolecular events after VSV infection by


insect bite, or the effect of individual viralgenes on VSV virulence in a natural host.

Utilizing the model presented here wehave shown that after penetration throughscarification of the coronary band skin VS-NJV infects and replicates primarily inkeratinocytes of the stratum granulosumand stratum spinosum resulting in microvesicles that coalesce to form larger vesi-cles at the coronary band. Virus antigenwas not observed in significant quantitiesassociated with MHC-II or MAC-387 cel-lular markers indicating that keratinocytesare the primary and most important celltype supporting VSNJV infection.

ACKNOWLEDGEMENTS

This work was supported by the UnitedStates Department of Agriculture (CRIS 1940-32000-040-00D). We thank Dr Jose Del C. Bar-rera for carrying out virus isolation and GeorgeSmoliga for real-time PCR tests. Dr BarbaraDrolet helped on the animal experiments andprovided antibody to VSNJV, Dr Corrie Brownfor valuable advice and Amy Kozer for histo-logical technical assistance and Melanie Praratfor reading the manuscript and providing valu-able suggestions.

REFERENCES

[1] Balachandran S., Barber G.N., Vesicularstomatitis virus (VSV) therapy of tumors,IUBMB Life (2000) 50:135−138.

[2] Clarke G.R., Stallknecht D.E., HowerthE.W., Experimental infection of swine witha sandfly (Lutzomyia shannoni) isolateof vesicular stomatitis virus, New Jerseyserotype, J. Vet. Diagn. Invest. (1996)8:105−108.

[3] Comer J.A., Stallknecht D.E., Nettles V.F.,Incompetence of domestic pigs as amplify-ing hosts of vesicular stomatitis virus forLutzomyia shannoni (Diptera: Psychodidae),J. Med. Entomol. (1995) 32:741−744.

[4] Comer J.A., Stallknecht D.E., Nettles V.F.,Incompetence of white-tailed deer as ampli-fying hosts of vesicular stomatitis virus for

Lutzomyia shannoni (Diptera: Psychodidae),J. Med. Entomol. (1995) 32:738−740.

[5] Cotton W.E., Vesicular stomatitis and its re-lation to the diagnosis of foot-and-mouthdisease, J. Am. Vet. Med. Assoc. (1926)69:313−332.

[6] Cupp E.W., Mare C.J., Cupp M.S., RambergF.B., Biological transmission of vesicularstomatitis virus (New Jersey) by Simuliumvittatum (Diptera: Simuliidae), J. Med.Entomol. (1992) 29:137−140.

[7] Falke D., Rowe W.P., Mouse disease due tovesicular stomatitis virus. II. Pathology oforgan lesions and the involvement of the cen-tral and peripheral nervous systems, Arch.Gesamte Virusforsch. (1965) 17:560−576(in German).

[8] Flanagan E.B., Zamparo J.M., Ball L.A.,Rodriguez L.L., Wertz G.W., Rearrangementof the genes of vesicular stomatitis viruseliminates clinical disease in the natural host:new strategy for vaccine development, J.Virol. (2001) 75:6107−6114.

[9] Fultz P.N., Holland J.J., Differing responsesof hamsters to infection by vesicular stomati-tis virus Indiana and New Jersey serotypes,Virus Res. (1985) 3:129−140.

[10] House J.A., House C., Dubourget P.,Lombard M., Protective immunity in cattlevaccinated with a commercial scale, inacti-vated, bivalent vesicular stomatitis vaccine,Vaccine (2003) 21:1932−1937.

[11] Howerth E.W., Stallknecht D.E., DorminyM., Pisell T., Clarke G.R., Experimentalvesicular stomatitis in swine: effects of routeof inoculation and steroid treatment, J. Vet.Diagn. Invest. (1997) 9:136−142.

[12] Martinez I., Rodriguez L.L., Jimenez C.,Pauszek S.J., Wertz G.W., Vesicular stom-atitis virus glycoprotein is a determinant ofpathogenesis in swine, a natural host, J.Virol. (2003) 77:8039−8047.

[13] Mead D.G., Ramberg F.B., Besselsen D.G.,Mare C.J., Transmission of vesicular stom-atitis virus from infected to noninfectedblack flies co-feeding on nonviremic deermice, Science (2000) 287:485−487.

[14] Mead D.G., Gray E.W., Noblet R., MurphyM.D., Howerth E.W., Stallknecht D.E.,Biological transmission of vesicular stomati-tis virus (New Jersey serotype) by Simuliumvittatum (Diptera: Simuliidae) to domesticswine (Sus scrofa), J. Med. Entomol. (2004)41:78−82.


[15] Patterson W.C., Jenney E.W., HolbrookA.A., Experimental Infections with vesicularstomatitis in swine I. Transmission by directcontact and feeding infected meat scraps,United States Livestock Sanitary AssociationProceedings, 1955, 368 p.

[16] Remmers L., Perez E., Jimenez A., VargasF., Frankena K., Romero J.J., Salman M.,Herrero M.V., Longitudinal studies in theepidemiology of vesicular stomatitis onCosta Rican dairy farms, Ann. NY Acad. Sci.(2000) 916:417−430.

[17] Ribelin W., The cytopathogenesis of vesicu-lar stomatitis virus infection in cattle, Am. J.Vet. Res. (1958) 19:66−73.

[18] Roberts A., Buonocore L., Price R., FormanJ., Rose J.K., Attenuated vesicular stomati-tis viruses as vaccine vectors, J. Virol. (1999)73:3723−3732.

[19] Rodriguez L.L., Vernon S., Morales A.I.,Letchworth G.J., Serological monitoring ofvesicular stomatitis New Jersey virus in en-zootic regions of Costa Rica, Am. J. Trop.Med. Hyg. (1990) 42:272−281.

[20] Rodriguez L.L., Nichol S.T., WebsterR.G., Granoff A., in: Vesicular StomatitisViruses, Encyclopedia of Virology, 2ndedition, Academic Press, London, 1999,pp. 1910−1919.

[21] Rodriguez L.L., Bunch T.A., Fraire M.,Llewellyn Z.N., Re-emergence of vesicularstomatitis in the western United States is as-sociated with distinct viral genetic lineages,(2000) 271:171−181.

[22] Seibold H., Sharp J., A revised concept ofthe pathological changes of the tongue in cat-tle with vesicular stomatitis, Am. J. Vet. Res.(1960) 21:35−51.

[23] Stallknecht D.E., Howerth E.W., ReevesC.L., Seal B.S., Potential for contact andmechanical vector transmission of vesicularstomatitis virus New Jersey in pigs, Am. J.Vet. Res. (1999) 60:43−48.

[24] Stallknecht D.E., Perzak D.E., Bauer L.D.,Murphy M.D., Howerth E.W., Contact trans-mission of vesicular stomatitis virus NewJersey in pigs, Am. J. Vet. Res. (2001)62:516−520.

[25] Sur J.H., Doster A.R., Christian J.S., GaleotaJ.A., Wills R.W., Zimmerman J.J., OsorioF.A., Porcine reproductive and respiratorysyndrome virus replicates in testicular germcells, alters spermatogenesis, and inducesgerm cell death by apoptosis, J. Virol. (1997)71:9170−9179.

[26] Tanimoto T., Ohtsuki Y., Evaluation ofantibodies reactive with porcine lympho-cytes and lymphoma cells in formalin-fixed, paraffin-embedded, antigen-retrievedtissue sections, Am. J. Vet. Res. (1996)57:853−859.

[27] Tesh R.B., Chaniotis B.N., Johnson K.M.,Vesicular stomatitis virus, Indiana serotype:multiplication in and transmission by ex-perimentally infected phlebotomine sand-flies (Lutzomyia trapidoi), Am. J. Epidemiol.(1971) 93:491−495.

[28] Vanleeuwen J.A., Rodriguez L.L., Waltner-Toews D., Cow, farm, and ecologic risk fac-tors of clinical vesicular stomatitis on CostaRican dairy farms, Am. J. Trop. Med. Hyg.(1995) 53:342−350.


Appendix A. Sensitivity of real-time RT-PCR (rRT-PCR). Supernatant from BHK-21 cells infectedwith VSNJ-95COB was titrated in serial dilutions by virus isolation in BHK-21 cells or by rRT-PCR. The results represent averages of two independent assays.

µL

To access this journal online:www.edpsciences.org/vetres

Documents

Vesicular stomatitis New Jersey virus (VSNJV) infects