Protective immunity in cattle vaccinated with a commercial scale, inactivated, bivalent vesicular stomatitis vaccine

Vaccine 21 (2003) 1932–1937

Protective immunity in cattle vaccinated with a commercial scale, inactivated, bivalent vesicular stomatitis vaccine James A. House a,∗ , Carol House a , Philippe Dubourget b , Michel Lombard b a

United States Department of Agriculture, Animal and Plant Health Inspection Service, Veterinary Services, National Veterinary Services Laboratories, Foreign Animal Disease Diagnostic Laboratory, P.O. Box 848, Greenport, NY 11944, USA b Merial SAS, 69007 Lyon, France Received 23 July 2002; received in revised form 10 December 2002; accepted 11 December 2002

Abstract A commercially prepared oil-adjuvanted, inactivated vaccine containing antigens of vesicular stomatitis virus (VSV) serotypes New Jersey (NJ) and Indiana 1 (IND1) was administered to calves to determine its ability to induce protective immunity. Weekly serological studies were conducted. The 12 calves in Group I were vaccinated once and challenge inoculated with VSV New Jersey 28 days later. Two calves were fully protected and two were partially protected. The five calves in Group II were vaccinated twice 40 days apart and challenge inoculated on 14 days post-second vaccination (dp2v) with VSV Indiana 1. All animals were fully protected. The 14 calves in Group III were vaccinated twice 91 days apart and challenge inoculated on 91 dp2v with VSV Indiana 1. All animals were fully protected. All control calves in each group became clinically ill. Two calves inoculated with VSV Indiana 1 challenge virus on day 0 and 11 weeks later showed clinical disease after each inoculation. No virus was isolated from the blood of four acutely ill calves 48 h after challenge inoculation. © 2003 Elsevier Science Ltd. All rights reserved. Keywords: Vesicular stomatitis; Protective immunity; Inactivated vaccine

1. Introduction The clinical signs of vesicular disease in cattle are lesions in the mucous membranes of the mouth and nose and the epithelium of the feet and teats. Laboratory tests are necessary to determine the etiological agent because vesicular disease in cattle may be caused by vesicular stomatitis viruses (VSVs) and foot-and-mouth disease viruses (FMDVs). Members of the genus Vesiculovirus in the family Rhabdoviridae cause vesicular stomatitis. The economic impact of vesicular disease from VSVs and FMDVs results from loss of production, lameness, and most importantly, from trade embargos. In 1964, it was estimated that VS caused an annual loss of US$ 300,000 in 5000 dairy cattle in the area of Guatemala City, Guatemala [1]. Using a 4% annual rate of inflation, this would equate to a current annual loss of about US$ 1.3 million in this small affected population. Vesicular stomatitis is endemic in central America and in North and South America. Outbreaks of vesicular stomatitis ∗ Corresponding author. Present address: 29745 Main Road, Cutchogue, NY 11935, USA. Tel.: +1-631-734-6288; fax: +1-631-734-2591. E-mail address: [email protected] (J.A. House).

occur in southwestern and western North America sporadically. Serotype New Jersey (VSVNJ) is endemic on Ossabau Island, off the coast of Georgia. In 1995, outbreaks of VSVNJ occurred in southwestern USA [2]. In 1997, both VSVNJ and VSV Indiana 1 (VSVIND1) were isolated from infected animals in the same locale [3]. Vaccination for VS is not widely practiced. Experimental live VS vaccines have been evaluated in dairy cattle under field conditions with equivocal results [4–6]. Results indicating induced immunity in experimental cattle have been reported for a subunit vaccine [7] and for a vaccinia recombinant vaccine [8]. Experimental VS vaccines previously reported have not become commercially available. There are obvious safety problems administering live VS vaccines and vaccinia-vectored genes, since the former is a zoonotic agent and vaccinia has not yet been accepted as a safe vector for widespread usage. Subunit vaccines possess the utility of potentially being able to induce an immune response that could be differentiated from natural infection by assaying for responses to non-structural proteins [7]. The commercially prepared inactivated vaccine described here may have the same advantage since the virus does not replicate in the host. We studied the serological responses and protective immunity induced by a bivalent, inactivated vesicular

0264-410X/03/$ – see front matter © 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0264-410X(03)00008-2

J.A. House et al. / Vaccine 21 (2003) 1932–1937

stomatitis vaccine containing VSVNJ and VSVIND1 antigens as well as antigens of two serotypes of FMDV.

2. Materials and methods 2.1. Vaccine The vaccine was a mixture made in Lyon of two single emulsions. A single emulsion of FMD vaccine 01 and A24 strains made at Pirbright, UK with a 1 ml volume dose and a single emulsion of VS vaccine New Jersey (NJ) and Indiana (IND) made in Lyon at 2 ml volume dose. Total vaccine dose injected at Foreign Animal Disease Diagnostic Laboratory (FADDL) was 3 ml by intramuscular route to each bovine. Vaccine strains were O1 BFS and A 24 Cruzeiro for FMDV and VS New Jersey (ATCC-1952) and VS Indiana 1 (ATCC-1925). The VS antigen load was expressed in equivalent micrograms after ultracentrifugation on sucrose gradient and reading at 254 nm and was determined by previous dose–effect correlation measured by VNT. VS Antigens were inactivated by binary ethyleneimine 1.2 mM with transfer in a second vessel (GMP requirements). Control of inactivation was carried out on three consecutive passages in BHK monolayer cells. High purification of inactivated antigens was made according to a patented process. The vaccine was first injected at FADDL 26 months after the manufacturing date. The combined vaccine (two FMD strains and two VS strains) a Aftovesibov, had similar FMD antigenic properties to the bivalent FMD vaccine Aftobov. The batch number was Aftovesibov-4VSV 5E 060. 2.2. Challenge viruses Tongue tissues were obtained from a field case of VS caused by VSVNJ that occurred in Colorado in 1982–1983 (VSVNJ COL82/83) and from a field case of VS Indiana 1 that occurred in Guatemala in 1994. The tissues were prepared as a sterile 10% suspension in Eagles minimum essential medium with 10% fetal bovine serum. Two calves were inoculated intradermally (i.d.) in the tongue with each suspension, and the vesiculated tongue tissue was harvested 2 days later. A 2% suspension was prepared as above and filtered through a 0.2 ␮ filter, aliquoted and frozen in the gas phase of liquid nitrogen. The titer of each preparation was determined in cell culture and in two calves that were injected intradermally in the tongue with 10-fold dilutions (1:10 through 1:10,000), using four sites per dilution. The cattle titrations were read 2 days later. The VSVNJ challenge virus (VSVNJ-CV) had a titer of 106.1 TCID50 /0.1 ml in Vero cell culture and 105.0 lesion forming units (LFU50 )/0.1 ml in cattle. Therefore, 13 TCID50 was equal to 1 LFU50 for VSVNJ-CV. The VSVIND1 challenge virus (VSVIND1-CV) had a titer of 106.4 TCID50 /0.1 ml in

1933

Vero cell culture and 105.0 LFU50 /0.1 ml in cattle. Therefore, 25 TCID50 was equal to one LFU50 for VSVIND1-CV. 2.3. Serology Vero cells (CCL48, American Type Culture Collection) were used between passages 136 and 236. The viruses used in the virus neutralization (VN) test were Vero cell culture adapted VSVNJ strain COL 82/83 and VSVIND1 strain Fort Lupton, both from the repository of the FADDL. Virus neutralizing antibody was measured with a microtiter test using approximately 100 TCID50 /0.025 ml. The test was incubated for 3 days at 37 ◦ C in a humid 5% CO2 atmosphere. Titers were based on the final dilution of serum and were calculated by the Spearman–Karber method of estimating 50% endpoints [9]. Geometric mean titers (GMT) were calculated. 2.4. Experimental design Holstein or Hereford calves weighing 350–500 lbs were housed in biological safety level 3 containment facilities at the FADDL throughout the studies. Persons in contact with VS-infected animals wore personal protective gear (Air Mate HEPA 10, Racal Health and Safety Inc., Frederick, MD 21704, USA) and followed Plum Island Animal Disease Center approved guidelines for handling zoonotic agents. Animals were observed and rectal temperatures recorded daily. Blood samples for serums were collected weekly. All vaccinations were done by intramuscular injections in the neck. Challenge inoculations were i.d. injections of a total estimated 10,000 LFU50 divided between four sites in the tongue (0.1 ml per site). Vaccinated animals showing no lesion were considered fully protected. Vaccinated animals showing lesions at one, two, or three inoculation sites were considered partially protected, and animals showing lesions at four sites were not protected. A summary of the treatment groups is given in Table 1. The 12 animals in Group I were vaccinated once and challenge inoculated on 28 days post-vaccination (dpv) with VSVNJ-CV. The five animals in Group II were vaccinated on 0 and 40 dpv and challenge inoculated on 14 days post-second vaccination (dp2v) with VSVIND1-CV. Additionally, at the time of challenge inoculation of the Group II animals, the two animals that were used 11 weeks earlier to titrate the VSVIND1-CV inoculum were inoculated with VSVIND1-CV a second time. Groups I and II each included two control animals that received the challenge inoculation. Two and three days post-challenge (dpc), all animals were observed for lesions of VS daily under RompunTM (xylazine 100 mg/ml; Bayer Corporation, Agricultural Division Animal Health, Shawnee Mission, KS 66201, USA) anesthesia. The 14 animals in Group III were vaccinated on 0 and 91 dpv, and challenge inoculated on 91 dp2v with VSVIND1-CV. Group III had four control animals: two

1934

J.A. House et al. / Vaccine 21 (2003) 1932–1937

Table 1 Summary of experimental design and selected results Feature

Group I

Group II

Group III

Animals vaccinated Dates of vaccination GMTb of vaccinated animals to challenge virus on 28 dpv GMT of vaccinated animals to heterologous virus on 28 dpv Challenge virus Challenge time GMT of vaccinated animals to challenge virus on 0 dpcf GMT of vaccinated animals to heterologous virus on 0 dpc GMT of vaccinated animals to challenge virus on 14 dpc GMT of vaccinated animals to heterologous virus on 14 dpc Vaccinated animals fully protected Vaccinated animals partially protected Vaccinated animals not protected Control animals receiving full challenge dose Control animals receiving dilutions of challenge virus Control animals exposed to challenge virus 11 weeks previously GMT of control animals to challenge virus on 14 dpc GMT of control animals to heterologous virus on 14 dpc Control animals clinically ill

12 0 dpva 1:592 1:1235 VSVNJ-CVc 28 dpv 1:592 1:1235 1:30902 1:3324 2/12 2/12 8/12 2 0 0 1:6096 1:77 2/2

5 0 and 40 dpv 1:1406 1:1770 VSVIND1-CVd 14 dp2ve 1:16827 1:24889 1:18793 1:25590 5/5 0/5 0/5 2 0 2 1:861 <1:32 4/4

14 0 and 91 dpv 1:1940 1:2592 VSVIND1-CV 91dp2v 1:13634 1:19707 1:17073 1:29408 13/13 0/13 0/13 2 2 0 1:1145 <1:32 4/4

a

Days post-vaccination. Geometric mean titer. c Vesicular stomatitis virus New Jersey-challenge virus. d Vesicular stomatitis virus Indiana 1-challenge virus. e Days post-second vaccination. f Days post-challenge. b

received VSVIND1-CV and two animals received 10-fold dilutions of VSVIND1-CV to determine the titer of the challenge inoculum. 2.5. Virus isolation studies The four control animals in Group III exhibited high fevers (41.1–41.7 ◦ C) and fulminating clinical signs 2 days after challenge inoculation. A heparinized blood sample was collected aseptically and 1 ml was inoculated into a 75 cm2 -culture flask of Vero cells. The flasks were incubated at 37 ◦ C and observed daily for cytopathogenic effects. Two blind cell culture passages of the samples were made in Vero cells.

3. Results 3.1. Group I A summary of selected results is given in Table 1. No animals in Group I (one vaccination followed by VSVNJ-CV on 28 dpv) had any reaction following vaccination. On 28 dpv, the GMT to VSVNJ was 1:592 and to VSVIND1 was 1:1235 (see Fig. 1). Following challenge inoculation, two animals were completely protected from any clinical sign and showed no febrile response. Two animals were partially protected and had febrile responses, respectively on 2, 3, and 4 dpc. On 14 dpc, the GMT to VSVNJ had risen anamnes-

tically 52-fold, to 1:30,902 and the GMT to VSVIND1 rose slightly to 1:3,324. Both control animals and the remainder of the vaccinated animals exhibited clinical signs of VS, with lesions developing at all four sites of challenge inoculation. The GMT of the control animals to VSVNJ was 1:1135 on 7 dpc and 1:6096 on 14 dpc, and to VSVIND1, 1:77 on 14 dpc. 3.2. Group II No animal in Group II (vaccinated twice on 0 and 40 dpv and challenge inoculated with VSVIND1-CV on 14 dp2v) had any reaction following vaccination. On 28 dpv, the GMT to VSVNJ was 1:1770 and to VSVIND1 was 1:1406. On 40 dpv, the GMT to VSVNJ was 1:2,831 and to VSVIND1 was 1:1950 (see Fig. 2). On 14 dp2v, the mean GMTs had risen to 1:24,889 for VSVNJ and to 1:16,827 for VSVIND1. All five vaccinated animals were completely protected and showed no febrile response. By 7 dpc, the GMTs had risen, respectively to 1:35,481 for VSVNJ and 1:24,774 for VSVIND1. By 14 dpc, the GMTs to both viruses essentially returned to pre-challenge levels, to 1:25,590 for VSVNJ and to 1:18,793 for VSVIND1. Both control animals showed clinical signs of VS. The control animals’ GMTs to VSVIND1 were 1:85 on 7 dpc and 1:861 on 14 dpc; neither control had antibody to VSVNJ. The animals that were inoculated twice with VSVIND1CV 11 weeks apart showed typical acute signs of VS following both inoculations. They developed an anamnestic

J.A. House et al. / Vaccine 21 (2003) 1932–1937

1935

Fig. 1. Serological responses of 12 calves vaccinated once with a bivalent VS vaccine and challenge inoculated with VSVNJ-CV.

response. Their mean GMTs to VSVIND1 rapidly increased from 1:490 to 1:6310 at 7 days post-re-challenge (dprc) and 1:9120 at 14 dprc, an 18-fold increase (see Fig. 3). These animals developed a low level of antibody to VSVNJ (GMT of 1:81 at 14 dprc).

3.3. Group III No animal in Group III (vaccinated twice and challenge inoculated with VSVIND1-CV on 91 dp2v) had any reaction to the first vaccination. One of the vaccinated animals died

Fig. 2. Serological responses of five calves vaccinated twice with a bivalent VS vaccine and challenge inoculated with VSVIND1-CV.

1936

J.A. House et al. / Vaccine 21 (2003) 1932–1937

Fig. 3. Serological responses and clinical signs of two calves inoculated ID in the tongue twice with VSVIND1-CV 11 weeks apart.

of bloat 10 weeks after the first vaccination. After the second dose of vaccine, five animals developed a transient febrile response of 40.0–41.1 ◦ C. One developed a febrile response for 4 days, as well as a 6–8 cm diameter swelling that subsided by 10 dpv. This calf was quite intractable and thrashed about during vaccination; the lesion was attributed to trauma, likely a hematoma. One non-febrile animal also had a transient swelling of 2.5 cm at the site of injection. On 28 and 91 dpv, the mean GMTs to VSVNJ were 1:2592 and 1:9683 and to VSVIND1 were 1:1940 and 1:5785, increasing approximately three-fold over the 63-day period. Three weeks after the second vaccination, the mean GMTs increased to 1:24,765 to VSVNJ and of 1:20,309 to VSVIND1, again an approximately three-fold increase. At the time of challenge (91 dp2v) the GMTs were 1:19,707 to VSVNJ and 1:13,634 to VSVIND1. None of the 13 vaccinated animals showed a lesion after challenge inoculation. All but one of the vaccinated animals maintained normal temperatures (rectal temperature below 40 ◦ C) following challenge inoculation. One animal had a fever of 40 ◦ C on 3 dpc, but otherwise was clinically normal. At 14 dpc, the GMTs remained essentially unchanged at 1:29,408 to VSVNJ and of 1:17,073 to VSVIND1. Both control animals given VSVIND1-CV developed severe lesions at all four sites of inoculation. The quantity of virus detected in the two animals inoculated with 10-fold dilutions of VSVIND1-CV was 103.5 and 103.75 LFU50 /0.1 ml, giving an average of 4000 LFU50 /0.1 ml of challenge virus or a total of 12,000 LFU50 . One of the control animals receiving the dilutions lost approximately 100 lbs and had severe damage to the tongue. This animal was humanely

euthanized on 10 dpc. At 14 dpc, the GMTs of the three remaining control and titration animals to VSVIND1 was 1:1145. There were no titers to VSNJ. No virus was isolated from the blood of the four control animals taken at the time of highest fever in the acute disease.

4. Discussion One vaccination with the subject VSV vaccine afforded full protection in two animals and partial protection in two out of 12 animals against a virulent challenge with VSVNJ-CV on 28 dpv. Therefore, a single vaccination did not provide complete protection. Two vaccinations afforded full protection to 19 out of 19 animals against a virulent challenge inoculation with VSVIND1-CV on either 14 or 91 dp2v. The GMTs of the animals receiving two vaccinations remained stable following challenge inoculation, indicating that the challenge virus failed to replicate enough to induce an anamestic response. The criteria for protection against a VS challenge inoculation were quite rigorous. During the development of the challenge system, we tried to obtain isolates from current field cases. One Indiana 1 candidate was a strain designated as Morelas 88 (received as a gift from Dr. Farouk Hamdy, USDA International Services, Mexico). This isolate was passed in Vero cells only four times. This challenge virus pool required 400 TCID50 to form a lesion (unpublished data) in contrast to 25 and 13 TCID50 for the challenge viruses used in this study. When an abundance of non-virulent virus is present, interferon or other interference

J.A. House et al. / Vaccine 21 (2003) 1932–1937

mechanisms such as defective particles may mask the lack of immunity in a challenge test. As well, with a high ratio of non-virulent to virulent viral particles, virulent viral particles may be blocked from reaching cell receptors and expressing their virulence, again masking a lack of protective immunity. Mackett et al. [8] did not observe secondary lesions even in susceptible cattle following VSV challenge inoculation. Therefore, for VS, the site of challenge inoculation is where protection must be measured by the lack of lesion development. In this study, we have described a reproducible model challenge system for assessing the immunity induced by VS vaccines that could be used to monitor future VS vaccines in the livestock host. The immune mechanism that affords protection against a massive local VSV challenge has not been determined. It may be speculated that following infection of mucous membranes or epithelium, the protection against VSV challenge may be mediated by local immunity (immunoglobulin A; IgA). The IgA response has a short duration. The waning of local immunity could account for the susceptibility to reinfection and clinical disease in two cattle given the same virulent VSVIND1-CV 11 weeks after full recovery from their original severe clinical disease. Indeed, Brandly et al. [10] reported reinfection as recent as 3 weeks after recovery from clinical VS. Cell-mediated immunity was not measured in this study or in Brandly’s study, but it may play a role in the protection against VS. A detectable viremia that could stimulate an animal to produce immunoglobulin G (IgG) is not associated with VS [11]. The lack of detectable viremia was substantiated in this study, as virus could not be isolated from heparinized blood of four control calves taken at the peak of acute disease. Parenteral vaccination normally induces an IgG response that is associated with a persistent immunity. IgG class antibodies may cross mucous membranes and could afford protection against viral invasion and lesion extension. In this study, the calves in Group III demonstrated a peak GMT to VSVIND1 at 70 dpv following the first vaccination. The GMT to VSVNJ was still rising when the animals were revaccinated at 91 dpv. The animals in Group I were challenged with VSVNJ-CV on 28 dpv and only 2 of the 12 were fully protected. If IgG levels predict activity that affords protection at the challenge site, the one vaccination scheme could possibly have shown improved protection if the interval between vaccination and challenge inoculation was increased. It is not possible to estimate the threshold GMT above which cattle would be protected with a designated probability since the numbers of cattle in this study is small. Gearhart et al. [12] studied the average response of dairy cattle to two doses of an inactivated VS vaccine. Twenty-one days after two doses of vaccine the average VN titer was 1:530, finally decreasing to 1:65 by 175 days. Serological tests vary so direct comparison of the data with our results is not possible. Also, protective immunity data is not available to correlate the described titer with resistance to a VS challenge inoculation. The solid immunity against a virulent chal-

1937

lenge inoculation induced by the described inactivated, bivalent VS vaccine sharply contrasts the lack of a protective response induced by mucosal infection. Two doses of commercial vaccine 3 months apart provided protection against challenge inoculation of VSVIND1-CV 6 months after the first vaccination. As VS has a relatively predictable cycle in the southwest United States of America, it is feasible that this vaccine could be applied in the winter months and afford protection under field conditions in the late summer and fall.

Acknowledgements The authors thank Brenda Rodd and Geoffrey Ferman for their excellent technical assistance and Jeffrey Babcock, Jose Sierra, Jim Liszanckie and Thomas Garner for their excellent animal care. References [1] Correa WM. Prophylaxis of vesicular stomatitis: a field trial in Guatemalan dairy cattle. AJVR 1964;25(107):1300–2. [2] Bridges VE, McCluskey BJ, Salman MD, Hurd HS, Dick J. Review of the 1995 vesicular stomatitis outbreak in the western United States. JAVMA 1997;211(5):556–60. [3] McCluskey BJ, Hurd HS, Mumford EL. Review of the 1997 outbreak of vesicular stomatitis in the western United States. JAVMA 1999;215(9):1259–62. [4] Lauerman LH, Kuns ML, Hanson RP. Field trial of live virus vaccination procedure for prevention of vesicular stomatitis in dairy cattle. I. Preliminary immune response. In: Proceedings of the 66th Annual Meeting of US Livestock San. Association; 1962. p. 365–9. [5] Lauerman LH, Hanson RP. Field trial of live virus vaccination procedure for prevention of vesicular stomatitis in dairy cattle. II. Second year evaluation in Panama. In: Proceedings of the 67th Annual Meeting of US Livestock San. Association, vol. 67, 1963. p. 483–90. [6] Lauerman LH, Hanson RP. Field trial of live virus vaccination procedure for prevention of vesicular stomatitis in dairy cattle. III. Evaluation of emergency vaccination in Georgia. In: Proceedings of the 67th Annual Meeting of US Livestock San. Association, vol. 67, 1963. p. 473–82. [7] Yilma T, Breeze RG, Ristow S, Gorham JR, Leib SR. Immune responses of cattle and mice to the G glycoprotein of vesicular stomatitis virus. In: Atassi MZ, Bachrach HL, editors. Immunology of proteins and peptides, vol. III. New York: Plenum Press; 1985. p. 101–15. [8] Mackett M, Yilma T, Rose JK, Moss B. Vaccinia virus recombinants: expression of VSV genes and protective immunization of mice and cattle. Science 1985;227:433–5. [9] Cottral GE. In: Cottral GE, editor. Manuel of standardized methods for veterinary microbiology. Ithaca (NY): Cornell University Press; 1978. p. 81–2. [10] Brandly CA, Hanson RP, Chow TL. Vesicular stomatitis with particular reference to the 1949 Wisconsin epizootic. Proc Am Vet Med Assoc 1951;88:61–70. [11] Redelman D, Nicol S, Klieforth R, Van Der Maaten M, Whetstone C. Experimental vesicular stomatitis virus infection of swine: extent of infection and immunological response. Vet Immunol Immunopathol 1989;20(4):345–61. [12] Gearhart MA, Webb PA, Knight AP, Salman MD, Smith JA, Erickson GA. The serum neutralizing antibody titers in dairy cattle administered an inactivated vesicular stomatitis virus vaccine. JAVMA 1987;191(7):819–22.