Herd and experimental schedule
A TB positive herd of Murciana-Grandina goats, located in Catalonia, with confirmed Mycobacterium caprae (spoligotype profile SB0416) infection and a reactor rate of 79% to the single intradermal comparative cervical tuberculin (SICCT) test, was selected for this study. A lot of 45 new born, approximately 1 month old, goat kids (16 males and 29 females) from the herd reposition were included in the study. Animals were randomly divided in two groups of 23 (8 males and 15 females) and 22 (8 males and 14 females), respectively. This sample size was deemed appropriate to estimate differences of proportions between treatment groups, i.e. 25% and 66% estimated proportion of infection in vaccinated and unvaccinated groups respectively (95% confidence interval, 90% statistical power, one-tailed test; source: www.winepi.net, University of Zaragoza).
At day 0, all experimental animals were negative to the SICCT test and the IFN-γ release assay (Bovigam™, Thermo Fisher Scientific, Schlieren, Switzerland). Twenty three animals were vaccinated subcutaneously with approximately 105 colony forming units (CFU) M. bovis BCG Danish strain (ATCC, Ref. 35,733™), prepared as previously described [16], and the other 22 remained as unvaccinated controls. Two months after vaccination, the kids were mixed with the rest of the herd and remained exposed to natural infection for 16 months, after this period, all the animals included in the study were humanely sacrificed.
All animals were fed following the regular farm conditions throughout the study (i.e. colostrum/milk initially, followed by forage supplemented with feed). Animals were followed-up for clinical signs every two months approximately.
Necropsy, tissue sampling and histopathology
The 45 experimental animals were euthanized by intravenous sodium pentobarbital overdose and carefully post-mortem examined in order to detect tuberculous lesions. Euthanasia was performed at 18 months after BCG vaccination. A complete necropsy procedure was conducted as previously described [17]. Briefly, tracheobronchial and mediastinal lymph nodes (LN) were aseptically removed making sure that the pleural lung surface was not sectioned, were sliced and examined for the presence of TB-like gross lesions, and then were stored at −20 °C as a pulmonary LN pool. Whole lungs were fixed with 10%-buffered formalin by pouring the fixative into the trachea. One month later, lungs were sliced at 1 cm thick intervals for TB-like gross lesion examination. All remaining viscera were also examined and all TB-like gross lesions detected in other tissues were collected at necropsy and subsequently fixed with 10%-buffered formalin, embedded in paraffin and examined by histopathology (haematoxylin and eosin and Ziehl Neelsen staining). For each animal at least one section of any tissues with visible lesions was included in the paraffin block, samples of the mesenteric lymph node and ileum were always included regardless whether lesions were observed or not macroscopically. The whole slide was evaluated under the microscope using the different magnification available in the microscope (4×, 10×, 20 x and 40× and, in the case of the Ziehl Neelsen Staining, the 100× magnification with immersion oil was used). The slides were analysed blindly by a pathologist. TB indicative lesions were set when the following histopathological features were observed: Presence of granulomas, with or without central necrosis and mineralization, surrounded by macrophages, multinucleated giant cells, sometimes partially or completely encapsulated. Enumeration (when present) of acid fast bacilli (AFB) was also performed. Four different categories were defined: None detected (no AFB detected after evaluation of the whole lesion under 100× magnification); 1 (observation of one AFB in at least one 100× field); 2–5 (observation of 2 to 5 AFB in at least one 100× field), >5 (observation of more than 5 AFB in at least one 100×).
MTBC culture
For microbiological culture, the pulmonary LN pools were thawed and a quantitative mycobacterial culture was subsequently carried out following the procedure described in previous studies [17]. Briefly, LN pools were homogenized, decontaminated, suspended and 100 μl of this suspension were twice serial diluted 1:10. Afterwards, 100 μl of each suspension dilution (1:100, 1:1000 and 1:10,000) were inoculated onto 7H11 agar plates (BD Diagnostics, MD, USA). The inoculated media were incubated at 37 °C. Bacterial counts were performed after 28 days and total CFUs in each pool of pulmonary LN were calculated. Isolates were confirmed as MTBC by multiplex PCR [18]. Then, DNA samples from MTBC isolates were analysed by DVR-spoligotyping [19].
Direct DNA detection
Once TB lesions were confirmed in non-respiratory tissues by histopathology, DNA extraction was carried out as following: Excess paraffin was trimmed off the sample block using a scalpel, and then 10 μm thick sections were obtained and dewaxed with xylene and absolute ethanol. Once the ethanol supernatant was removed and the remaining ethanol had evaporated the tissue was weighed to obtain an amount of 25 mg. Afterwards, DNA from resulting tissue powder was extracted using a DNA purification kit (Promega Biotech Iberica, Madrid, Spain). A semi-nested PCR, targeted to the MTBC-specific IS6110 sequence [17] was run under standard conditions and a M. avium subsp. paratuberculosis (MAP)-specific Real Time PCR (Vacunek, Derio, Spain) was run following the manufacturer instructions.
Serology
Blood samples were collected from all experimental animals at the end point of the study. Sera from all animals were analysed by ELISA to detect antibodies against MTBC as previously described [17] and against MAP (ID Screen® Paratuberculosis indirect, ID.vet, Grabels, France).
Data analysis
Differences in presence of TB lesions and histopathological parameters between groups were assessed by chi-square test. Significance was set at P value <0.05. Statistical analysis was performed with R package v2.15.0 (R Foundation for Statistical Computing, Vienna, Austria).
