Vaccination of sheep against M. paratuberculosis: immune parameters and protective efficacy

Vaccine 23 (2005) 4999–5008

Vaccination of sheep against M. paratuberculosis: immune parameters and protective efficacy D.J. Begg, J.F.T. Griffin ∗ Disease Research Laboratory, Department of Microbiology and Immunology, University of Otago, P.O. Box 56, Dunedin, New Zealand Received 20 January 2005; received in revised form 22 April 2005; accepted 26 May 2005 Available online 17 June 2005

Abstract Johne’s disease in ruminants is caused by the pathogenic bacterium Mycobacterium avium subspecies paratuberculosis (Map). Currently available Map commercial vaccines protect against clinical disease but not infection. In this study, the proprietary Johne’s vaccine NeoparasecTM and an aqueous formulation of Map 316F (AquaVax) were tested in sheep. Detailed immunological examination of blood and gut-associated lymphoid tissues was carried out on animals after vaccination and challenge with virulent Map to identify markers of protective immunity. NeoparasecTM vaccination provided significant protection against disease while AquaVax did not. Immune animals had stronger cell-mediated responses and altered proportions of CD4+ , CD8+ , CD25+ and B cells in blood, spleen and the gut lymphatics, than diseased animals. © 2005 Elsevier Ltd. All rights reserved. Keywords: Johne’s disease; Protective immunity; Vaccination

1. Introduction Vaccination has been used since 1926 [1] to control Johne’s disease. The types of vaccines used have included live attenuated strains of Mycobacterium avium subspecies paratuberculosis (Map) and heat killed or sonicated preparations. Mineral oil adjuvants are routinely included in the vaccine formulation. Vaccines are normally administered subcutaneously to sheep from 2 weeks to 4 months of age. A limitation for the widespread use of Johne’s vaccination is the development of abscesses at the injection site [2]. Protective efficacy studies show that while clinical disease may develop in some vaccinated animals, there is a significant reduction in the prevalence of diseased animals [3,4]. While vaccination reduces the total number of animals excreting organisms, it does not result in a decrease in the overall prevalence of infection [2,5]. Another complication following vaccination is that sensitisation with vaccines causes interference with immunolog∗

Corresponding author. Tel.: +64 3 4797718; fax: +64 3 4772160. E-mail address: [email protected] (J.F.T. Griffin).

0264-410X/$ – see front matter © 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.vaccine.2005.05.031

ical tests used for diagnosis of natural Map infection and tuberculosis (TB) due to M. bovis infection. This is a real concern for the use of Map vaccines [6] in animals such as cattle and deer that are naturally susceptible to tuberculosis. National Tuberculosis Eradication programmes used worldwide for cattle and deer are based on skin testing protocols that rely on immunodiagnostic testing. Map vaccination can interfere with TB surveillance schemes involving skin testing as a herd screening test. This is due to the high degree of antigenic cross reactivity between antigens in the vaccine strain of Map and mycobacterial pathogens such as M. bovis [6]. Protective immune responses to Map have not been looked at in detail previously in ruminants; however, as with other mycobacterial diseases, it is hypothesised that a vigorous CMI (Type 1) response is important for protection against Map infections [7–9]. Empirical approaches used historically for vaccines to control Johne’s disease use mineral oil adjuvants to evoke an aggressive immune response, on the premise that strong immune reactions are likely to be optimal. Recent advances in our understanding of the fundamental mechanisms of immunity require that prophylactic

D.J. Begg, J.F.T. Griffin / Vaccine 23 (2005) 4999–5008

5000

protocols be designed to evoke the most appropriate, rather than the most vigorous immune responses [10,11]. Vaccination of deer with an aqueous formulation of BCG, in a prime-boost protocol, has been shown to provide significant protection against M. bovis infection [12] involving predominantly a Type 1 pathway of immune reactivity. By contrast, vaccines using mineral oil adjuvants were not protective [12]. The first objective of this study was to monitor the immune responses in animals vaccinated with a commercial vaccine NeoparasecTM , compared with an aqueous suspension of Map strain 316F (AquaVax). The second, was to compare retrospectively, immune responses in unvaccinated animals that were uninfected following experimental challenge with Map, with those that developed disease. Finally, the immune responses seen in vaccinated animals and unvaccinated sheep that survived infection were compared to determine if there were definitive immune profiles typical for protection. Cellmediated immunity was monitored using T cell lymphocyte transformation (LT) and Interferon-␥ assays. Antibody assays (ELISA) were used to determine if there were qualitative differences between animals that were resistant to infection from those that developed disease. Comparative changes in the immune responsiveness and lymphocyte subpopulations of gut and peripheral tissues were examined to determine whether markers of systemic or localised immunity could be used as a signature for a protective response.

2. Methods 2.1. Animal ethics The animal experiments carried out in this study were carried out under ethics approval licences numbered: P453, P499, P518 and P594, approved by the Invermay AgResearch Animal Ethics Committee. 2.2. Experimental animals The experimental sheep comprised 130 Merinos, all of which were castrated males. The lambs were selected from flocks in which no Johne’s disease had been previously observed and were held with ewes until weaning at 2.5–3 months of age. After weaning, lambs were randomly allo-

cated to experimental treatment groups as required. The lambs were kept on pasture under standard New Zealand sheep farming conditions, with supplementary feeding during winter. 2.3. Immune profiles in lambs vaccinated with oil adjuvanted versus aqueous live formulations A group of 30 merino lambs were assigned randomly into three groups. One group was vaccinated with a commercial oil adjuvanted live (316F) vaccine (NeoparasecTM -Merial NZ Ltd.) and another group with a live aqueous (316F) vaccine (AquaVax). The third group was left as unvaccinated controls. The lambs were moved to the experimental facility after weaning (11 weeks old) and were vaccinated at 3 months of age. The NeoparasecTM vaccine was administered according to proprietary recommendations; 1 ml dose given subcutaneously. The AquaVax formulation consisted of live Map (316F) in a buffered saline solution at a concentration of 1 × 108 cfu/mL given as a 1 mL dose subcutaneously into the neck. The animals vaccinated with AquaVax were given a booster inoculum of the vaccine 1 month later. 2.4. Protective efficacy of vaccines Ninety merino lambs were assigned randomly into three groups each containing 30 animals. At docking (2 weeks of age), the lambs were vaccinated using the protocol outlined in Table 1. The NeoparasecTM vaccine was used according to the manufacturers specification in the first group of 30 lambs. AquaVax formulation (Map 316F at 1 × 108 cfu/mL in buffered saline) was given in 1 mL doses to the second group. One month after the primary vaccination a booster dose of AquaVax was given. The final group of 30 lambs were left as unvaccinated controls. The lambs were weaned at 2.5 months of age. At 3 months of age the lambs were challenged orally with virulent Map, given three times at weekly intervals as 1 mL doses (5 × 108 cfu/mL) of a gut tissue homogenate of Map, isolated from a sheep with clinical Johne’s disease (JD3). This experimental infection regime produces gut histopathology within 9 months and the onset of clinical disease within 11 months post-challenge [13]. A group of 10 sentinel unchallenged animals were included in the experiment to provide background immune parameters. Animals were slaughtered between 10 and 22 months post-

Table 1 Vaccination and infection protocols used in the second sheep study Vaccination protocols Treatment groupa Control NeoparasecTM Aqueous vaccine Unvaccinated a b

Vaccine formulation N/A 316F in oil 316F in saline N/A

Vaccine dose (cfu) N/A 6 × 108 1 × 108 N/A

The lambs were vaccinated at 2–4 weeks of age and infected at 3 months of age. Animals were challenged three times at weekly intervals.

Booster vaccination N/A None 1 month later N/A

Infectious challenge doseb N/A 5 × 108 5 × 108 5 × 108

D.J. Begg, J.F.T. Griffin / Vaccine 23 (2005) 4999–5008

challenge. The criteria for culling animals was that from 10 months post-challenge, the animals were weighed weekly and any animal that lost 10% of its weight before 1 year of age was culled. After 1 year, animals that lost 15% of their weight were culled, except during winter when the limit was raised to 18%. A small group of animals selected randomly at 13 months post-challenge were culled to establish representative histopathology. The other sheep were electively removed as indicated by weight loss, except for a small group of unaffected animals that were culled at 22 months of age.

2.7. Comparative intradermal skin test

2.5. Necropsy

2.8. Lymphocyte transformation assay using peripheral blood leucocytes

All animals were euthanased humanely using either a captive bolt stun gun or injected with barbiturate into the jugular vein. After exsanguination, the intestines were removed from below the abomasum through to the rectum. The small intestines were laid out to expose the jejunum, ileum, caecum and mesenteric lymph nodes (jejunal lymph nodes). Samples were taken from serial sections of the mesenteric lymph nodes, the ileocaecal lymph node (ICLN), jejunum, ileum and ileocaecal valve. After the animals were necropsied, the tissues were examined histologically to determine the grade of pathology at a microscopic level. Histological lesions were graded on a scale of 0–3 using the Perez classification [14], but without identifying the subtypes within each pathology grade. Based on the gross lesions and histopathological grading, animals were given a numerical disease score based on a 0–6 grading scale (0 = no pathology; 6 = severe Johne’s disease) to categorise the pathology found in individual animals. 2.6. Segregation of diseased and immune animals following challenge Six different groups of animals were identified for ex vivo examination of their immune response based on serial blood sample profiling. Six to ten animals that developed severe Johne’s disease were selected from each of the treatment groups (NeoparasecTM , AquaVax and unvaccinated controls) for more detailed study. At least 6 and up to 10 from each of the treatment groups that appeared to be immune, with no disease developing by 22 months post-challenge were also examined in detail. Tissues from selected animals were examined in detail to profile peripheral and localised gut immune responses specific to Map infection, after necropsy. The groups consisted of nine experimentally challenged sheep that developed severe clinical Johne’s disease (diseased; D), three non-challenged controls with no Johne’s disease as negative controls (unchallenged; Uc), four unvaccinated animals that survived the challenge (unvaccinated immune; Uv-I), and a final group of five NeoparasecTM vaccinated animals that did not develop disease after experimental infection (NeoparasecTM immune; Nv-I).

5001

Delayed type hypersensitivity response in the sheep was tested by the inoculation of 0.1 mL of purified protein derivative (PPD) from Map (PPDj, CSL 0.5 mg/mL), 0.1 mL of M. avium PPD (PPDa, 0.5 mg/mL) and 0.1 mL of M. bovis PPD (PPDb, 0.5 mg/mL) intradermally into the skin of the inner aspect of the hind legs of the sheep. The skin thickness was measured pre inoculation and 72 h later with digital callipers (Mitutoyo Digimatic CD-8, Japan). The skin test was carried out 6 months post-challenge.

A 10-mL heparinised vacutainer of blood was centrifuged at 1260 × g for 15 min. The buffy coat of mononuclear peripheral blood cells were removed and washed by centrifugation in 45 mL of phosphate buffered saline (PBS). The leucocytes were resuspended in 5 mL of 10% sheep serum (Gibco BRL, Grand Island, NY, USA) made up in RPMI 1640 (Gibco BRL) with l-glutamine and gentamicin (Gibco BRL). One hundred microlitres of cells were plated in a U bottom 96 well plate (Nunc Denmark), and 50 ␮L of antigen or mitogen added to quadruplicate wells. The concentrations of mitogen and antigen used in the assay were: Concanavalin A (ConA; Sigma, St. Louis, USA), 50 ␮g/mL and purified protein derivative from PPDj (50 ␮g/mL). The plates were then incubated under humidified conditions for 3 days at 37 ◦ C, in a 5% CO2 atmosphere. Fifty microlitres of H3 -thymidine (Amersham Pharmica, Piscataway, NJ, USA), with a specific activity of 10 ␮Ci/mL was added to each well after which the cells were incubated for a further 18 h, harvested (Cambridge Harvester, Watertown, MA, USA) and the radioactivity counted in a Wallac 1205 Betaplate counter (Turku, Finland). Results were expressed as mean counts per minute (cpm). 2.9. Isolation of sheep intestinal mucosal leucocytes Five to ten centimetres of the intestinal sample was removed (ileum or jejunum) and placed into 30 mL of 2% foetal calf serum (FCS) in RPMI as soon as was feasible after death (10–20 min) and transported back to the laboratory on ice. Using aseptic techniques, the intestinal tissues were opened using scissors and laid out flat. Tissue samples were trimmed removing excess fat and blood vessels. The intestinal tissues were washed in PBS. The mucosa was scraped off from the muscularis using a scalpel and put into 30 mL RPMI supplemented with 2% FCS, and left for 1 h shaking at 37 ◦ C. The supernatant was removed and replaced with medium containing 2% FCS and incubated shaking at 37 ◦ C for 1 h. This was repeated twice with a final incubation of 20 min. The medium was again removed and replaced with collagenase type XI (Sigma; 15 mg/100 mL) in 10% FCS. The samples were then incubated shaking at 37 ◦ C for 90 min. Cell debris

5002

D.J. Begg, J.F.T. Griffin / Vaccine 23 (2005) 4999–5008

and aggregated cells were removed by filtering the cell suspension through a cell strainer (70 ␮m BD Falcon) into a new 50 mL conical tube (BD Falcon). The supernatant was then centrifuged at 340 × g for 15 min and the supernatant discarded and replaced with 30 mL of PBS. This washing step was repeated and after discarding the supernatant, 5 mL of RPMI containing 10% FCS was added to the mucosal leucocytes. The cell suspension was counted using trypan blue diluent in a haemocytometer before adjusting the live cell count to 2.5 × 106 cells/mL. 2.10. Isolation of mononuclear leucocytes from lymph nodes for lymphocyte transformation assays Lymph nodes were removed from the animal immediately (5–15 min) following slaughter, placed into a petri dish with 30 mL 2% Foetal calf serum (FCS, Gibco BRL) in RPMI and macerated using a scalpel and forceps. The media containing the cells was poured through a 70 ␮m cell strainer (Becton Dickinson) into a 50 mL conical tube and held on ice during transportation back to the laboratory. The cells from the lymph nodes or blood were layered onto 7.5 mL of histopaque (Sigma) (δ 1.083) and centrifuged at 450 × g for 60 min. The cells at the interface were removed and resuspended in 35 mL of PBS and centrifuged at 340 × g for 15 min. After discarding the supernatant, the pellet was resuspended in 10 mL of RPMI containing 10% FCS. The live mononuclear cells from all tissues were counted using a haemocytometer and trypan blue diluent and adjusted to 2.5 × 106 live cells/mL in 10% FCS. One hundred microliters of cells were plated in a U bottom 96 well plate (Nunc) and 50 ␮L of antigen or mitogen added to replicate wells. The concentrations used were: ConA, 25 ␮g/mL; Pokeweed mitogen (Sigma), 6.25 ␮g/mL; and PPDj, 25 ␮g/mL. The concentrations of antigens were different from those used in serial blood sample analysis as optimisation studies showed these concentrations to be the best compromise between lymph node and blood samples. The plates were then incubated for 4 days at 37 ◦ C, in 5% CO2 . Fifty microlitres of H3 -thymidine (10 ␮Ci/mL) was then added to each well, after which the cells were incubated for a further 18 h, harvested and counted in a Wallac 1205 Betaplate counter. Results are expressed as a mean cpm. 2.11. Enzyme linked immunosorbant assay (ELISA) The ELISA test for detection of antibody was adapted from a method previously described by Griffin et al. [15]. Briefly flat-bottomed microtiter plates (Nunc Maxisorp immuno plate) were coated with 50 ␮L of antigen PPDj (12.5 ␮g/mL) and incubated overnight at 4 ◦ C. The unbound antigen solution was removed and the plates washed six times in wash buffer (PBS containing 0.05% Tween 20). Blood plasma, diluted 1:100 with wash buffer was dispensed in 50 ␮L aliquot’s to duplicate wells. Plasma from a con-

firmed Johne’s disease positive sheep was used as a positive reference sample. Negative control plasma was obtained from a non-vaccinated, non-infected animal. Plates were incubated at 37 ◦ C for 1 h, and then washed six times in wash buffer. Rabbit anti-sheep horseradish peroxide labelled (Dako) antibody used to detect bound primary antibody. The secondary antibody was diluted 1:2000 and 50 ␮L was added to each well and the plates incubated at 37 ◦ C for 30 min. The plates were again washed. One hundred microlitres of substrate solution, equal volumes of 2.1% citric acid and 2.84% Na2 HPO4 in deionized water, 0.1% H2 O2 plus 0.4 mg/mL of orthophenylenediamine dihydrochloride (OPD) was then added to plates which were incubated in the dark at room temperature for 20 min. The reaction was stopped by addition of 50 ␮L 2 M H2 SO4 . The absorbance was read at 490 nm using an automated microplate reader (Bio-Rad model 3550, Japan). The results are expressed as ELISA units = (sample OD−negative control OD) × 100. 2.12. Interferon-γ assay The assay used involved the proprietary BOVIGAMTM kit (Pfizer Animal Health Ltd.) to detect IFN-␥. Briefly 1.5 mL of blood or cells (isolated from lymph nodes at 2.5 × 106 cells/mL) were placed into four wells of a 24 well plate (BD Falcon). Each well was stimulated with 100 ␮L of one of the following antigens or mitogen: saline control, PPDa (300 ␮g/mL) PPDj (300 ␮g/mL) and PWM (100 ␮g/mL). After 24 h, the plasma supernatant was removed and frozen at −20 ◦ C until tested. Analysis of IFN␥ was performed using the standard protocol recommended for the Bovigam ELISA kit. 2.13. Flow cytometric analysis of isolated cells Isolated leucocytes were resuspended to a concentration of 1 × 106 cells/mL in PBS containing 2% FCS. The cells were centrifuged at 340 × g for 7 min after which the media was removed and the antibodies added. The antibodies used were CD4 (44.97), CD8 (38.65), ␥␦-TCR (86D) and CD25 (9.14) (University of Melbourne) and a B antibody BAQ155A (VMRD). The cells and the antibodies were mixed well and placed on ice to incubate for 30 min. Unbound antibodies were removed by washing with 2% FCS in PBS. The fluorescein labelled secondary antibody (Donkey anti mouse IgG FITC Jackson Immunochemicals) was diluted 1/40 with 2% FCS in PBS. Ten microlitres of the diluted secondary antibody was added to the cells. The cells and antibody were mixed and incubated on ice in the dark for 30 min. The cells were then washed once in 2% FCS in PBS and again in FACS Buffer (5% FCS and 1% sodium azide in PBS). After the final wash, the cells were resuspended in 500 ␮L of FACS buffer and 500 ␮L 1% paraformaldehyde in PBS. The stained cells were analysed in a FACScalibur (Becton Dickinson).

D.J. Begg, J.F.T. Griffin / Vaccine 23 (2005) 4999–5008

2.14. Statistical analysis In order to examine statistical differences between the groups, taking into account outbred animal and treatment differences, the data were analysed using GenStat for Windows release 6.1.2002. Data were modelled using residual maximum likelihood [16]. The error for the repeated measurements on the same animal was modelled by a first order autoregressive process with heterogeneity of variance at the time points. Tests of significance were performed using the Wald statistic [17].

3. Results 3.1. Comparison of immune profiles in lambs vaccinated with oil adjuvanted versus aqueous live formulations The profile seen from the lymphocyte transformation response showed that the NeoparasecTM vaccines created a significantly higher (P < 0.001) response than was seen in

5003

either the AquaVax or unvaccinated control animals (Fig. 1a). Animals vaccinated with NeoparasecTM vaccine produced significant levels of IFN-␥ within 2 weeks postvaccination. This peaked at 6 weeks post-vaccination whereas the AquaVax group was largely unresponsive (Fig. 1b). The specific antibody responses in animals vaccinated with NeoparasecTM were significantly higher (P < 0.001) than the response to AquaVax or the unvaccinated group (Fig. 1c). The AquaVax appeared to evoke very little antibody production in sheep for the first 8 months. Increased antibody levels were seen consistently in all groups at 10 months post-vaccination. The LT response from the vaccinated animals tested at 10 months post-vaccination showed that blood mononuclear cells gave the strongest response, followed by splenic cells, while the intestinal tissues gave the weakest response (Fig. 1d). The unvaccinated showed minimal proliferation of peripheral blood lymphocytes. The lymph node draining the vaccine site (prescapular) and the spleen gave greater responses to the NeoparasecTM than to the AquaVax. The mesenteric and ileocaecal lymph nodes leucocytes gave

Fig. 1. Immune response profiles from animals vaccinated with NeoparasecTM or AquaVax. (a) Lymphocyte transformation responses; (b) IFN-␥ responses measured by the commercial Bovigam assay; (c) specific antibody responses (IgG) as measured by ELISA; (d) LT Responses from different tissues of animals obtained at necropsy. Group mean responses (n = 10) ± S.E.M.

D.J. Begg, J.F.T. Griffin / Vaccine 23 (2005) 4999–5008

5004

stronger responses than cells from the intestinal mucosal tissues, but both were considerably weaker than the response in the peripheral lymphatic tissues (prescapular lymph node, spleen or blood). The leucoctyes from the mucosal tissues—anterior jejunum (Ant JJ), posterior jejunum (Post JJ) and terminal ileum (TI) were reactive with mitogen (ConA or PWM) showing that viable cells with polyclonal reactivity were present. 3.2. Protective efficacy of vaccines The data shown in Table 2 identified that there was significant protection against Map pathology in animals vaccinated with NeoparasecTM vaccine. Only four of the NeoparasecTM vaccinated animals developed severe disease, while most had no pathology at all. The AquaVax did not appear to provide any protection, showing a similar range and severity of pathology as was seen in the unvaccinated group. Unvaccinated animals (Uv) that developed clinical disease following experimental challenge were first seen at 10 months after challenge. Within 15 months, up to 50% of the unvaccinated animals had been euthanased and by 18 months post-challenge 80% of the unvaccinated animals were culled. A small group of unaffected animals were euthanased at 22 months post-challenge. The AquaVax vaccinated (Av), challenged sheep had a similar onset of clinical symptoms as the unvaccinated ones. 3.3. Peripheral blood responses in immune and diseased animals seen following infectious challenge NeoparasecTM vaccinated animals that were immune (Nv-I) to experimental infectious challenge and those that developed disease (Nv-D) showed a significant difference (P < 0.01) in their specific lymphocyte transformation response (Fig. 2a) to Map antigens. Immune NeoparasecTM vaccinated (Nv-I) and unvaccinated immune animals (Uv-I) tended to have higher levels of lymphocyte reactivity than animals that developed disease. Unvaccinated diseased animals (Uv-D) had only baseline responses while the unvaccinated

Table 2 Protection against clinical disease by different vaccines Pathology score 0 normal 1 2 3 4 5 6 severe JD Total no. of animals examined

Unvaccinated

NeoparasecTM

Aqueous vaccine

7 5 2 1 2 5 8

21 2 1 1

4

6 1 2 3 6 1 9

30

29a

28a

a Three animals died accidentally and were excluded from the data set, as they could not be examined by detailed necropsy.

Fig. 2. Immune response profiles from vaccinated and non-vaccinated sheep that either develop disease or were immune, where NeoparasecTM vaccinated immune animals; Nv-I: NeoparasecTM vaccinated diseased animals; Nv-D: AquaVax animals immune to challenge; Av-I: AquaVax animals that develop disease; Av-D: unvaccinated animals immune to challenge; Uv-I: unvaccinated animals that develop disease; Uv-D. (a) Lymphocyte transformation responses to PPDj from NeoparasecTM and unvaccinated animals; (b) Specific antibody (IgG) to PPDj from NeoparasecTM and unvaccinated animals; (c) Lymphocyte transformation response to PPDj from AquaVax and unvaccinated animals. Group mean responses (n = 6) ± S.E.M.

immune sheep (Uv-I) began to develop LT responses between 6 and 9 months. The NeoparasecTM vaccinated animals had a consistently stronger response at challenge as a result of vaccination 2 months earlier. A further spike of lymphocyte reactivity was seen in the Nv-I groups 6 months post-challenge, while no change was seen in the response of the Nv-D group. The AquaVax (Av) treated sheep did not have the same initial high immune readouts at challenge as was seen in

D.J. Begg, J.F.T. Griffin / Vaccine 23 (2005) 4999–5008

the NeoparasecTM vaccinated sheep. The LT result showed the AquaVax animals produced an earlier response than the unvaccinated sheep and later developed a stronger response (Fig. 2c). The AquaVax immune (Av-I) animals gave a stronger LT response than was seen from the aqueous vaccinated diseased (Av-D) sheep. There were no differences in the levels of antibody from immune or diseased (NeoparasecTM or AquaVax) vaccinated animals at different time points post-challenge. The NeoparasecTM vaccinated animals produced high levels of PPDj specific antibody within 1 month post-vaccination (Fig. 1c) and had higher levels present at challenge, 2 months post-vaccination. The unvaccinated diseased sheep (Uv-D) showed a spike in antibody production at 9 months postchallenge (Fig. 2b), which was absent from the immune group of animals (Uv-I). The group mean comparative intradermal skin test results (Table 3) show that NeoparasecTM vaccinated groups of animals had the strongest skin test reactions among the sheep tested. NeoparasecTM vaccinated immune (Nv-I) animals tended to have larger group mean responses (5.5) than the diseased group (Nv-D) (4.8). A similar trend was seen in the ‘immune’ animals vaccinated with the AquaVax (2.5 versus 1.0) or the unvaccinated control animals (2.2 versus 0.7) where the diseased subgroup had overall lower levels of skin test reactivity. 3.4. Immunological responses of different tissues examined at necropsy The group mean LT response specific to PPDj varied between samples and treatments (Fig. 3a). The unchallenged animals (Uc) provided a reference background reactivity against which other responses could be measured. The peripheral LT response (blood, spleen and prescapular lymph node) was greater than the response from the gut tissues in all the challenged groups of animals. Both groups of immune animals (Uv-I and Nv-I) showed a strong response both from blood and spleen leucocytes, but the Uv-I produced lower responses from the prescapular lymph node cells. The IFN-␥ response (Fig. 3b) shows a significant difference between the Uc group and the other treatment groups (P < 0.01). Diseased animals had strong responses in the spleen and post JJLN and moderate responses in the other gut lymphatics. Unvaccinated immune animals had strong

5005

responses in the peripheral blood and relatively low responses from the gut lymphatics. The NeoparasecTM vaccinated animals had high responses in the peripheral and gut lymphatics. This group, not surprisingly, also had the highest response seen in the prescapular lymph node. This was logical based on the fact that this was the lymph node draining the injection site in the neck. Results given in Fig. 3c show that the percentage of CD4+ cells found in the peripheral and gut-associated lymphatic tissues of diseased, immune and unchallenged animals. In this experiment, the proportion of CD4+ cells found in the tissues of unchallenged control (Uc) animals provided background reference values. It appears that immune animals (Uv-I, Nv-I) have equivalent proportions of CD4+ cells to the unchallenged controls. In contrast, the tissues from diseased (D) animals had a significantly lower (P < 0.001) percentage of CD4+ cells in the peripheral and gut-associated lymphatic tissues. The spleen had fewer CD4+ cells than the other tissues examined. The diseased animals showed a significant reduction in percentage of CD8+ cells compared to the other treatment groups (P < 0.001), though this was not seen uniformly between the tissues (Fig. 3d). The blood and spleen from the immune (Uv-I and Nv-I) animals had a higher percentage of CD8+ cells than diseased (D) or unchallenged control (Uc) sheep. When the percentage of B cells was measured by FACS analysis, there was a large difference seen between the treatment groups (Fig. 3e). The Nv-I and the Uv-I animals showed a significantly higher percentage of B cells (P < 0.001) than the diseased animals and the unchallenged controls. Overall the percentage of B cells found in the tissues of unchallenged animals was similar to that found in diseased animals. There was no significant difference between the tissues. Diseased animals had a significantly lower (P < 0.01) percentage of CD25+ (activated) cells isolated from the different lymphatic tissues (Fig. 3f). The Nv-I animals had a higher percentage of CD25+ cells than the Uc sheep. The Nv-I sheep had a similar percentage of CD25+ cells in all the gut lymph nodes. The immune animals from both the NeoparasecTM vaccinated and unvaccinated group generally had a higher level of CD25+ cells than diseased (D) animals or unchallenged control (Uc) group.

4. Discussion Table 3 Mean intradermal skin test results in experimentally challenged sheep Group

Bovine PPD

Uv-I Uv-D Av-I Av-D Nv-I Nv-D

0.3 0.1 0.5 0.2 1.4 1.3

± ± ± ± ± ±

0.5 0.2 0.4 0.4 0.8 1.0

Avian PPD 1.2 0.3 1.5 0.3 2.7 2.0

± ± ± ± ± ±

1.4 0.5 0.8 0.7 0.7 2.0

Results expressed as the mean increase in skin thickness.

Johnin PPD 2.2 0.7 2.5 1.0 5.5 4.8

± ± ± ± ± ±

1.2 0.7 1.4 0.8 1.6 1.5

An ideal vaccine would either generate sterile immunity or help the animal contain the infection so there is no horizontal spread. Most of the current Map vaccines use mineral oil adjuvants to evoke more active immune responses. Adjuvants used range from mineral oil [18] to a mixture of liquid paraffin, olive oil and pumice powder [19]. Vaccination with these strong adjuvants often leads to lesions at the site of vaccination [2]. While they invariably result in a strong cellular immune response, it is overlayed with an equally

5006

D.J. Begg, J.F.T. Griffin / Vaccine 23 (2005) 4999–5008

Fig. 3. Immune profiles from different tissues at necropsy of unchallenged animals; Uc (n = 3), diseased animals; D (n = 9), unvaccinated animals that were immune to challenge (Uv-I; n = 4) and NeoparasecTM vaccinated animals that were immune to challenge; Nv-I (n = 5). (a) Lymphocyte transformation responses to PPDj; (b) IFN-␥ responses to PPDj stimulation as measured by the Bovigam assay; (c) percentage of CD4+ cells; (d) percentage of CD8+ cells; (e) percentage of B cells; (f) percentage of CD25+ cells measured from the different tissues. Results expressed as group mean values ± S.E.M.

strong humoral response [20]. The most likely protective response to mycobacterial infections would be an exclusive CMI response, with no humoral reactivity [8]. Hypothetically, the immune response resulting from optimal vaccination should comprise of a strong IFN-␥ and LT reactivity with no antibody, both hallmarks of a Type 1 response. The Types 1 and 2 dichotomy seen so clearly in inbred mice [10]

is not clear in outbred animals where one study of progressing bovine Johne’s disease showed decreased IgG1 antibody levels which is associated with the Type 2 immune response pathway [21]. The immune response to a commercial live vaccine (NeoparasecTM ) was examined to chart qualitative aspects of immunity in vaccinated sheep. The results show the

D.J. Begg, J.F.T. Griffin / Vaccine 23 (2005) 4999–5008

NeoparasecTM vaccine evoked strong peripheral immune responses, which comprised of both CMI and humoral reactivity. The Johne’s specific IFN-␥ production from NeoparasecTM vaccinated sheep in these trials demonstrated differences from previously published results [20]. In this series of experiments, the IFN-␥ levels peaked 1–2 months post-vaccination while Garcia Marin et al. [20] observed a gradual increase peaking at 1 year post-vaccination. This may be due to variables such as the breed of sheep, batch differences of vaccine, or natural exposure to environmental mycobacteria such as Map or M. avium subsp avium. Nonetheless, variations seen in the immune readout from different trials suggest caution in interpreting any one experimental finding as definitive. Specific antibody responses in sheep vaccinated with NeoparasecTM in Study I were first seen 1 month after vaccination of the sheep, with the levels rising again at 10 months post-vaccination. In the study of Garcia Marin et al. [20], maximum antibody levels were recorded from 60 to 120 days post-vaccination, using NeoparasecTM vaccine in sheep. The AquaVax, formulated to contain 108 cfu of live Map (316F) per inoculum, produced low level transient immune responses in sheep. In contrast, an aqueous low dose formulation of live BCG vaccine (2.5 × 106 cfu) has been used previously in deer and has produced consistent CMI responses [12] that were protective against tuberculosis. The reason for the low levels of immune reactivity seen in these trials is unknown. The second trial in which the animals were challenged after vaccination, confirmed that AquaVax provided very little protection against disease. Experimental challenge of the AquaVax and unvaccinated groups of animals with virulent Map produced lesions in 80% and 75% of the sheep, respectively. By contrast, only 28% of NeoparasecTM vaccinated animals had lesions following challenge. The data from the present study showed that sheep protected from disease had strong CMI responses: good IFN-␥ and LT reactivity, although they also produced elevated antibody levels. All the groups of animals that were immune to infection (unvaccinated, NeoparasecTM or aqueous vaccinated) showed a similar trend of higher LT and DTH responses, earlier following challenge, than case matched diseased sheep (Fig. 2 and Table 3). The unvaccinated diseased (Uv-D) sheep produced a higher antibody response at 12 months post-infection than the immune (Uv-I) animals. From this data, it is likely that an appropriate early CMI response needs to be activated to produce protective immunity in sheep, as the late CMI response seen in the NeoparasecTM vaccinated diseased animals was not protective. Animals that develop disease may have a late onset mixed CMI and antibody responses which are not protective. This is compatible with results from previous vaccination trials [4,18], which show a significant reduction in the prevalence of clinical disease without a corresponding decrease in infection, following vaccination of sheep. Map vaccination resulted in the development of a significant peripheral immune response with a low number of

5007

immune cells seeding to the gut, as seen by the low LT and IFN-␥ responses. In animals that cleared the infection, naturally or following vaccination, there are low levels of immune reactivity within the gut lymphoid tissues. Diseased animals appeared to produce non-protective, possibly immunopathologic cell activation in the gut-associated lymphoid tissues, causing increased production of IFN-␥ and LT responses and the proportion of lymphocyte subpopulations in the gutassociated lymphoid tissues (Fig. 3). The most obvious difference seen in the proportions of lymphoid cell subpopulations in the tissues of immune (NeoparasecTM immune or unvaccinated immune) animals and the other treatment groups (diseased or unchallenged) was the increase in proportion of B cells from the immune sheep. Changes in the proportion of B cells have been observed previously. Begara-McGorum et al. [22] found in the early stages of Map infection (8 weeks post-challenge) that sheep have a lower number of B cells in the gutassociated lymphatics than control animals. In cattle with clinical Johne’s disease, the number of B cells increase in blood, while the reactivity of the peripheral blood B cells decrease [23]. The increase in proportion of B cells seen in the present study may be associated with, rather than responsible for, the effector mediated protective immunity to Map. Further examination of this B cell population from the protected sheep is needed to determine if the cells are contributing to protection against disease. As with all mycobacterial diseases, the protective immune response to Johne’s disease should rely heavily on immunity that is strongly biased towards Type 1 pathways [9,24,25]. After the initial infection with Map, the mycobacteria seem to activate CD4+ T cells of the Type 1 pathway [26] resulting in the early induction of the CMI response observed. While NeoparasecTM vaccines appear to provide mixed Type 1/Type 2 activation, qualitative differences may occur where some animals develop protective immunity while others remain unprotected. Ideally, the vaccine should be refined to ensure that the response would be more heavily biased towards a Type 1 pathway of immunity. Mineral oil adjuvants may be required to evoke cellular responses to weak immunogens [27] though they may not be necessary to evoke adequate responses to live or killed mycobacterial vaccines [20] that contain strong bacterial immunogens. Less aggressive natural adjuvants have been observed to evoke adequate responses in deer vaccinated against M. bovis [12]. Studies are currently under way using lipid adjuvants as an alternative to mineral oil to improve the immunogenicity of Map vaccines. Immunological monitoring suggests that they have the potential to produce strong Type 1 patterns of immunity, compatible with protection.

Acknowledgments We thank the following individuals for their contributions to this work: Phil Farquhar (AgResearch Invermay) for ani-

5008

D.J. Begg, J.F.T. Griffin / Vaccine 23 (2005) 4999–5008

mal husbandry, Gary Clark and Dylan Turner histology, Pam Crosbie (University of Otago) and Gary Yates (AgResearch Wallaceville) for culture of Map and Peter Johnstone (AgResearch Invermay) for statistical analysis of the data. This research was supported by grants from the Foundation of Research, Science & Technology (FoRST) and the University of Otago Research Committee.

References [1] Vallee H, Rinjard P. Etudes sur l’ enterite paratuberculeuse des bovides (note preliminaire). Rev Gen Med Vet 1926;35:1–9. [2] Windsor PA, Eppleston J, Whittington RJ, Jones SL, Britton A. Efficacy of a killed Mycobacterium avium subsp. paratuberculosis vaccine for the control of OJD in Australian sheep flocks. In: Juste RA, Geijo MV, Garrido J, editors. Proceedings of the Seventh International Colloquium on Paratuberculosis. Bilbao, Spain: International Association for Paratuberculosis; 2002. p. 420–3. [3] Wentink GH, Bongers JH, Zeeuwen AA, Jaartsveld FH. Incidence of paratuberculosis after vaccination against M. paratuberculosis in two infected dairy herds. Zentralbl Veterinarmed B 1994;41:517– 22. [4] Cranwell MP. Control of Johne’s disease in a flock of sheep by vaccination. Vet Rec 1993;133:219–20. [5] Kormendy B. The effect of vaccination on the prevalence of paratuberculosis in large dairy herds. Vet Microbiol 1994;41:117–25. [6] Mackintosh CG, Labes R, Rodgers C, Griffin F. Paratuberculosis vaccination and Tb testing trial in red deer. Proc N Z V A Deer Br 2001;18:36–43. [7] Hamilton HL, Follett DM, Siegfried LM, Czuprynski CJ. Intestinal multiplication of Mycobacterium paratuberculosis in athymic nude gnotobiotic mice. Infect Immun 1989;57:225–30. [8] Orme IM, Lee BY, Appelberg R, Miller ES, Chi DL, Griffin JP, et al. T cell response in acquired protective immunity to Mycobacterium tuberculosis infection. Bull Int Union Tuberc Lung Dis 1991;66:7–13. [9] Chiodini RJ. Immunology: resistance to paratuberculosis. Vet Clin North Am Food Anim Pract 1996;12:313–43. [10] Estes DM, Brown WC. Type 1 and type 2 responses in regulation of Ig isotype expression in cattle. Vet Immunol Immunopathol 2002;90:1–10. [11] Brown WC, Rice-Ficht AC, Estes DM. Bovine type 1 and type 2 responses. Vet Immunol Immunopathol 1998;63:45–55. [12] Griffin JF, Mackintosh CG, Slobbe L, Thomson AJ, Buchan GS. Vaccine protocols to optimise the protective efficacy of BCG. Tuber Lung Dis 1999;79:135–43.

[13] Begg DJ, O’Brien RL, Mackintosh CG, Griffin JFT. Parameters of experimental infection with M. avium subsp. paratuberculosis to develop a sheep model for Johne’s disease. Infect Immun, in press. [14] Perez V, Garcia Marin JF, Badiola JJ. Description and classification of different types of lesion associated with natural paratuberculosis infection in sheep. J Comp Pathol 1996;114:107–22. [15] Griffin JFT, Cross JP, Chinn DN, Rodgers CR, Buchan GS. Diagnosis of tuberculosis due to Mycobacterium-bovis in New Zealand red deer (Cervus-elaphus) using a composite blood-test and antibodyassays. N Z Vet J 1994;42:173–9. [16] Patterson HD, Thompson R. Recovery of inter-block information when block sizes are unequal. Biometrika 1971;58:545–54. [17] Wald A. Tests of statistical hypotheses concerning several parameters when the number of observations is large. Transact Am Math Soc 1943;54:426. [18] Sigurdsson B. A killed vaccine against paratuberculosis (Johne’s disease) in sheep. Am J Vet Res 1960;21:54–67. [19] Saxegaard F, Fodstad FH. Control of paratuberculosis (Johne’s disease) in goats by vaccination. Vet Rec 1985;116:439–41. [20] Garcia Marin JF, Tellechea J, Gutierrez M, Corpa JM, Perez V. Evaluation of two vaccines (killed and attenuated) against small ruminant paratuberculosis. In: Manning EJ, Collins MT, editors. Sixth International Colloquium on Paratuberculosis, Melbourne, Australia; 1999. p. 234–41. [21] Koets AP, Rutten VP, de Boer M, Bakker D, Valentin-Weigand P, van Eden W. Differential changes in heat shock protein-, lipoarabinomannan-, and purified protein derivative-specific immunoglobulin G1 and G2 isotype responses during bovine Mycobacterium avium subsp. paratuberculosis infection. Infect Immun 2001;69:1492–8. [22] Begara-McGorum I, Wildblood LA, Clarke CJ, Connor KM, Stevenson K, McInnes CJ, et al. Early immunopathological events in experimental ovine paratuberculosis. Vet Immunol Immunopathol 1998;63:265–87. [23] Waters WR, Stabel JR, Sacco RE, Harp JA, Pesch BA, Wannemuehler MJ. Antigen-specific B-cell unresponsiveness induced by chronic Mycobacterium avium subsp. paratuberculosis infection of cattle. Infect Immun 1999;67:1593–8. [24] Clarke CJ. The pathology and pathogenesis of paratuberculosis in ruminants and other species. J Comp Path 1997;116:217–61. [25] Little D, Alzuherri HM, Clarke CJ. Phenotypic characterisation of intestinal lymphocytes in ovine paratuberculosis by immunohistochemistry. Vet Immunol Immunopathol 1996;55:175–87. [26] Burrell C, Clarke CJ, Colston A, Kay JM, Porter J, Little D, et al. Interferon-gamma and interleukin-2 release by lymphocytes derived from the blood, mesenteric lymph nodes and intestines of normal sheep and those affected with paratuberculosis (Johne’s disease). Vet Immunol Immunopathol 1999;68:139–48. [27] Aucouturier J, Dupuis L, Ganne V. Adjuvants designed for veterinary and human vaccines. Vaccine 2001;19:2666–72.