Materials and Methods

Animals—The study was approved by the University of Pennsylvania Institutional Animal Care and Use Committee. Twelve healthy newborn male calves were obtained from a dairy farm that was free of paratuberculosis. Calves were separated from their dams within 1 hour after birth and fed 4 L of colostrum that was collected from their own dam. Calves were housed in individual pens and fed milk replacer and calf-starter grain throughout the study. Calves were randomly assigned (in blocks) to the control or vaccination group (n = 6/group).

Experimental design—At 14 days after birth (day 0), calves were injected SC in a side of the neck, midway between the shoulder region and ear and midway between the dorsal and ventral midlines, with 1.0 mL of heat-inactivated, strain 316F MAP vaccine^a (vaccination group) or 1.0 mL of sterile saline (0.9% NaCl) solution (control group). The vaccine contained 2.5 mg of dried microorganisms/mL and was compounded with mineral oil in a multiple emulsion as an adjuvant. The person who administered the injections had no other involvement in the project; therefore, all project personnel involved in data generation were blinded to calf treatment. On 2 occasions (days 21 and 22), 1.2 × 10⁹ CFUs of a live field strain MAP (American Type Culture Collection 700535; prepared as described previously⁹) was administered orally to each calf. Briefly, first-passage MAP seed stock that was originally cultured from a bovine fecal sample was frozen at −70°C. Prior to the start of this experiment, an aliquot of seed stock was thawed and propagated in mycobacterial growth medium^b and stored at 4°C until used in the mycobacterial challenge. The MAP was suspended in milk replacer, and the calves voluntarily suckled the suspension from a syringe. On day 83 or 84 (9 weeks after challenge), calves were euthanized by IV administration of pentobarbital and tissues were harvested for analysis.

Clinical observations and sample collection schedule—For each calf, rectal temperature was measured before injections were administered on day 0 and once daily for the following 3 days; measurements were also obtained immediately before live MAP were administered and once daily for the 3 days following challenge. Injection-site reaction was measured in 3 dimensions with a ruler on days 3, 7, 14, 21, 28, 35, 42, 56, 70, and 84. Body weight was measured at the start and conclusion of the study (prior to euthanasia).

A blood sample (8 mL) was collected from each calf into heparinized collection tubes on days 0 (prior to vaccination or injection of saline solution; baseline), 7, 14 (before challenge), 21, 28, 42, and 84 for an IFN-γ–release assay specific for MAP antigens. At the same collection times on days 0, 21, and 84, an additional blood sample (6 mL) was collected from each calf to obtain serum for detection of antibodies against MAP by use of an ELISA. At necropsy, prescapular lymph nodes (ipsilateral to the injection site) were harvested and processed to collect cells for an IFN-γ–release assay specific for MAP antigens.

Various samples from each calf were collected for mycobacterial culture: ileocecal lymph node (n = 2), hepatic lymph node, duodenum, jejunum (10 sites distributed at regular intervals along the length of the jejunum; samples included the corresponding mesenteric lymph node [10]), ileum (3 sites), ileocecal valve, cecum, spiral colon, prescapular lymph node ipsilateral to the injection site, and feces. A separate set of sterile instruments was used to harvest each specimen. Samples were stored at −70°C and processed for mycobacterial culture within 6 weeks of collection.

Samples of ileum, ileocecal lymph nodes, prescapular lymph nodes, and, where present, injection-site reactions were preserved in neutral-buffered 10% formalin. Histologic examination was performed by use of staining with H&E and Ziehl-Nielsen stains.

Sample processing—Serum samples were tested for antibodies against MAP by use of a commercial ELISA kit^c according to manufacturer's recommendations. For the IFN-γ–release assay, 4 aliquots (1.5 mL each) of whole blood were placed into tissue culture plates. The stimulants for each of the 4 wells were PBS solution (negative control sample), 2 johnin PPD preparations (designated as PPDa^d and PPDb^e; 10 μg/mL each), or pokeweed mitogen^f (1 μg/mL; positive control sample). The samples were incubated for 18 hours at 37°C in 5% CO₂, and supernatants were then harvested via centrifugation. Cells from prescapular lymph nodes were suspended in RPMI tissue culture medium and incubated with PPDa MAP antigen (10 μg/mL), a suspension of heat-killed MAP (American Type Culture Collection 700535; 10⁶ CFU/mL), or medium alone. Cells were incubated for 18 hours at 37°C in 5% CO₂, and supernatants were then harvested via centrifugation. Whole blood and lymph node cell supernatants were assayed for IFN-γ concentration by use of a commercial bovine IFN-γ ELISA kit.^g The optical density values were converted to arbitrary IFN-γ units by use of a standard curve that was based on data obtained from serial dilutions of a conditioned-medium standard created in our laboratory. A standard curve was created for each ELISA plate.

Tissue and fecal samples were processed for MAP culture by use of hexadecylpyridinium chloride decontamination.^10,11 The inoculum prepared from these samples was plated on HEYM^h (4 tubes/sample [100 μL/tube]) or into liquid culture medium (1 tube/sample [100 μL/tube]). The samples in the liquid culture medium were incubated in a unit with an automated growth detector.ⁱ The number of colonies on each tube of HEYM after 16 weeks of incubation, as well as the duration of incubation until bacterial growth exceeded the detection threshold in the liquid culture system (TTPR), was recorded. In the liquid culture system, samples that did not trigger growth detection by 56 days of incubation were deemed negative for MAP growth; for purposes of analysis, a TTPR of 56 days was assigned to cultures that yielded negative results.

Statistical analysis—All statistical tests were 2-sided at the 5% level of significance (ie, values of P ≤ 0.05 were considered significant). Rectal temperature, injection site size, results of serologic analyses, and IFN-γ concentrations were analyzed by use of general linear mixed models with repeated measures.^j Treatment group, time point, and the interaction of these effects were fitted as fixed effects, and block and animal were fitted as random effects. Least squares means and associated 95% confidence intervals for each treatment group at each time point were calculated. Linear combinations of the variable estimates (obtained from the model) were used to compare treatment groups at each time point when an overall significant treatment effect or an overall significant interaction between treatment and time point was initially detected.

For HEYM MAP culture, the primary outcome variable was the total number of MAP CFUs per gram of tissue among all samples for each calf. For automated liquid culture, the primary outcome variable was the mean TTPR among all tissues for each calf. The MAP CFUs or TTPR was analyzed by use of a general linear mixed model^j that included the fixed effect of treatment and the random effect of block. Data were logarithmically transformed prior to analysis, and the results were back-transformed for evaluation of least squares means and associated 95% confidence intervals for each treatment group. Linear combinations of the variable estimates (obtained from the model) were used to compare treatment groups. The mean liquid culture TTPR and HEYM culture results (total number of MAP CFUs/g of tissue) for each calf were compared via correlation analysis.^k

The total number of tissue samples from which MAP was isolated, as determined via HEYM culture, was analyzed by use of a generalized mixed model.^j The model included the fixed effect of treatment and the random effect of block. Linear combinations of the variable estimates (obtained from the model) were used to compare treatment groups.

Results

Clinical observations—On day 1, rectal temperature in the calves in the vaccination group was increased (least square mean difference, 0.72 ± 0.23°C; P = 0.013), compared with rectal temperature in calves in the control group. The increased rectal temperature persisted on day 2 (least square mean difference, 0.67 ± 0.22°C; P = 0.012). By day 3, there was no difference in rectal temperature between the groups. Vaccinated calves also developed swelling at the injection site; on day 1, the injection sites were significantly (P < 0.001) larger than those in the control calves. The swelling persisted throughout the study, as indicated by significant differences in injection site size between groups at most time points (Figure 1). Gains in weight from the start to the end of the study did not differ significantly (P = 0.33) between the groups. Histologic examination of samples of the ileum, ileocecal lymph nodes, and prescapular lymph nodes did not reveal any lesions in any calf from either group. Acid-fast bacteria were seen rarely in samples of ileum from 2 vaccinated calves. The injection sites from vaccinated calves had coalescing, necrotizing granulomata with core mineralization and clusters of acid-fast bacteria.

View Full Size

Least square mean injection site size versus time following SC injection of MAP bacterin (1.0 mL [squares]) or saline (0.9% NaCl) solution (1.0 mL [diamonds]) in 12 calves (6 calves/group). Day 0 measurements were obtained prior to injection. Injections were administered in a side of the neck, midway between the shoulder region and ear and midway between the dorsal and ventral midlines on day 0. Calves received 1.2 × 10⁹ CFUs of a live field strain MAP orally on days 21 and 22. Bars indicate the upper 95% confidence limit. *At this time point, the value in the vaccinated group is significantly (P b 0.05) different from that in the control group.

Citation: American Journal of Veterinary Research 70, 4; 10.2460/ajvr.70.4.493

• Download Figure
• Download figure as PowerPoint slide

Immune response—Calves were seronegative for antibodies against MAP at all time points throughout the study. However, at 7 days following injection, calves in the vaccinated group had increased IFN-γ production by whole blood cells following in vitro incubation with johnin PPDa or PPDb. This increased IFN-γ production was significant, compared with the value in blood samples collected on day 0 (P = 0.01) and compared with the value in control calves at day 7 (P = 0.006), and persisted throughout the study (Figure 2). The MAP antigen-stimulated IFN-γ production in whole blood cells from control calves did not vary from the baseline value up to day 42. However, at the end of the study period, there was a small increase in IFN-γ production that was significantly (P = 0.004) different from baseline. The IFN-γ production in response to stimulation with PBS solution alone (negative control) did not vary from baseline values throughout the study for either group. Similarly, IFN-γ production by cells from the prescapular lymph node that drained the injection site was greater (20-fold difference; P = 0.028) in the vaccination group, compared with the control group, when whole-cell MAP antigen was used as the stimulus, but the difference was not significant (P = 0.15) when johnin PPDa was used as the stimulus.

View Full Size

Least square mean concentration of IFN-γ in supernatants derived from whole blood samples that were collected and incubated with johnin PPDa at intervals during an 84-day period from 12 calves that were injected SC with MAP bacterin (1.0 mL [squares]; n = 6) or saline solution (1.0 mL [diamonds]; 6) on day 0. Injections were administered in a side of the neck, midway between the shoulder region and ear and midway between the dorsal and ventral midlines. Calves received 1.2 × 10⁹ CFUs of a live field strain MAP orally on days 21 and 22 (challenge [double-headed arrow]). Similar results were obtained following incubation of whole blood samples with johnin PPDb (data not shown). Bars indicate the upper 95% confidence limit. *At this time point the value in the vaccinated group is significantly (P ≤ 0.01) different from that in the control group. †Within a treatment group, the value at this time point is significantly (P ≤ 0.05) different from the value at day 0.

Citation: American Journal of Veterinary Research 70, 4; 10.2460/ajvr.70.4.493

• Download Figure
• Download figure as PowerPoint slide

Tissue colonization—When evaluated by use of culture with HEYM, tissues from calves in the vaccinated group yielded 2,417 CFUs/g of tissue (expressed as geometric least squares mean; 95% confidence interval, 489 to 11,926 CFUs/g of tissue) and was significantly (P = 0.045) lower than the value in the control group (15,709 CFUs/g of tissue; 95% confidence interval, 3,185 to 77,486 CFUs/g of tissue). Also, the geometric least squares mean of number of tissues from which MAP was isolated for the vaccination group (21.8 tissues/calf; 95% confidence interval, 17.7 to 26.9 tissues/calf) was significantly (P = 0.041) lower than the number of tissues in the control group (27.6 tissues/calf; 95% confidence interval, 25.6 to 29.8 tissues/calf). The results of liquid culture (TTPR) were highly and inversely correlated (r = −0.96) with the results of culture with HEYM (CFUs/g of tissue), regardless of treatment. The geometric least squares mean TTPR for vaccinated calves (33.1 days; 95% confidence interval, 26.1 to 41.9 days) was significantly (P = 0.034) longer than that for control calves (22.7 days; 95% confidence interval, 17.8 to 28.8 days), which also indicated less tissue colonization with MAP in vaccinated calves. For selected tissues, the difference in the culture results (median CFUs/g of tissue or median TTPR) between the treatment groups was significant (P ≤ 0.05; Table 1).

Discussion

In the study reported here, SC injection with the strain 316F MAP vaccine resulted in less extensive intestinal and lymph node tissue colonization in vaccinated calves, compared with findings in control calves, at 9 weeks following oral challenge with a heterologous, live field strain of MAP. The reduction in tissue colonization was accompanied by induction of antigen-specific stimulation of IFN-γ production by circulating lymphocytes and cells from lymph nodes ipsilateral to the injection site. It is possible that this early reduction in tissue colonization by MAP would eventually result in less severe or delayed onset of clinical disease, fecal shedding, or both, but studies have not been conducted to investigate such effects. However, in a similar study,⁶ vaccine-associated reduction in MAP infection at 120 days after challenge was also present at 270 days. The findings of the present study suggested that the strain 316F MAP vaccine would be a useful aid in the control of MAP infection in cattle. However, even though vaccinated calves had reduced tissue colonization, compared with control calves, MAP infection was not prevented in any calf. This underscores the importance of management practices (eg, calving stall hygiene and colostrum source and handling) that reduce the possibility of transmission of MAP, rather than relying on vaccination alone, in an MAP control program.

In a previous investigation⁷ in which the experimental MAP challenge method applied in the present study was used, vaccination with a commercially available paratuberculosis vaccine derived from M avium strain 18 did not reduce tissue colonization; however, vaccination with a heat-killed, whole-cell preparation of an MAP field strain did reduce tissue colonization. These findings, combined with those of the present study, suggest that successful vaccination against paratuberculosis requires the use of an appropriate MAP strain rather than the closely related strain 18 of M avium.

Vaccination resulted in the rapid development of an antigen-specific, cell-mediated immune response that was detected 7 days following injection and persisted through the duration of the study. Following the oral challenge with live MAP, calves in both groups had a secondary, steady increase in the IFN-γ response from whole blood cells. However, the response in unvaccinated calves was smaller than the response in vaccinated calves. The present study did not include an unchallenged group of vaccinated calves to determine whether the immune response from vaccination would have persisted without the stimulus of oral challenge with live MAP. The lack of detectable serum antibodies in the vaccinated calves in the present study is consistent with findings of a previous study,¹² in which the earliest time points at which antibodies were detected were 60 to 100 days following vaccination and in some calves as late as 180 days. Similarly, in a previous investigation⁹ that used the same experimental challenge method as that applied in the present study, pathologic lesions and acid-fast bacteria were rarely observed despite the recovery of MAP from intestinal and lymphoid tissues by use of culture techniques.

Strong correlation between results for the MAP liquid culture and HEYM culture was detected in the present study. The liquid culture system has the advantage of a shorter duration of incubation and automated MAP detection. Quantification of MAP, which can be used for comparisons of interventions such as vaccination or antimicrobial treatment, is also possible with the liquid culture system. Additionally, both forms of culture appeared to have greater sensitivity for detection of MAP organisms than histologic examination of tissue samples. This apparent difference in sensitivity was likely attributable, in part, to the amount of tissue examined with each technique—the amount of tissue that underwent HEYM or liquid culture was much larger than the amount of tissue that underwent histologic examination.

Vaccinated calves developed a granulomatous swelling at the injection site and had a mild increase in rectal temperature for 48 hours following injection. This was not accompanied by a change in attitude or appetite, and body weight gain during the study period among the vaccinated calves was not different from that among the control calves. The swellings did not develop drainage and did not enlarge during the final 30 days of the study.

The results of the study reported here suggested that use of the strain 316F MAP vaccine stimulated an appropriate immune response to MAP in vaccinated calves. This resulted in reduced tissue colonization with the organism following oral MAP challenge. The strain 316F MAP vaccine could be useful as an aid to control MAP infections in cattle.