Antemortem Diagnosis of Mycobacterium bovis Infection in Free-ranging African Lions (Panthera leo) and Implications for Transmission

Screening for mycobacterial diseases in free-ranging wildlife is complicated by the difficulties involved in obtaining good quality samples. Most investigators have relied on indirect diagnostic methods, such as serologic tests, the tuberculin skin test (TST), or in vitro cytokine assays. These tests may be limited by lack of species-specific reagents and validation, the need for multiple captures (i.e., TST), or other sample-handling requirements (Maas et al. 2013). For example, the TST in lions (Panthera leo) is performed by intradermal injection of 0.2 mg avian PPD and 0.2 mg bovine PPD in the cranial cervical area, right and left sides, respectively (Keet et al. 2010). Lions are immobilized again 3 d postinjection so that changes in skin thickness can be measured using tuberculosis calipers. A positive reaction is considered ≥2 mm increase in skin thickness at the bovine PPD site compared with the avian PPD site. However, this test provides only indirect evidence of infection. Acquiring biologic material for culture and molecular identification provides a direct method for detecting viable Mycobacterium bovis in the respiratory tract of lions. We describe a technique for tracheobronchial lavage of free-ranging lions under field conditions for this purpose.

Between February 2010 and August 2013, we collected tracheobronchial lavage samples from 134 lions (Panthera leo) using the technique described below. Research protocols were approved by the South African National Park Animal Care and Use Committee. Animals were captured in Kruger National Park (23°49′60″S, 31°30′0″E), South Africa, for other research or management purposes according to the South African National Parks standard operating protocols. Lions were immobilized with either tiletamine-zolazepam (Zoletil; Wildpharm, Queenswood, South Africa; total dose, 250–750 mg intramuscular [IM]) alone, or tiletamine-zolazepam (total dose, 80–120 mg IM) in combination with medetomidine (Kyron Laboratories Pty. Ltd., Benrose, South Africa; total dose, 4–6 mg IM) using a 3-mL CO₂-propelled dart (DAN-INJECT, International S.A., Skukuza, South Africa). Doses were based on estimated weight (approximately 2–3 mg/kg tiletamine-zolazepam or 0.55 mg/kg tiletamine-zolazepam with 0.027 mg/kg medetomidine). Lions that received medetomidine were partially reversed with atipamezole (Antisedan; Wildpharm) at a ratio of 5∶1 (mg atipamezole∶mg medetomidine) administered IM.

After immobilization, lions were blindfolded and placed in sternal recumbency. Using a three-piece portable metal frame (Fig. 1) to provide support for the head, the upper jaw was held open using a braided rope passed behind the canine teeth and hooked onto the metal frame. The operator could then use gravity to assist in holding the lower jaw, extend the tongue, and visualize the glottis with a 30-mm, straight laryngoscope blade (Fig. 1). A disinfected precut equine stomach tube with a wire stylet was passed along the laryngoscope blade into the glottis to approximately the level of the carina (premeasured by marking the length of the tube from the nose to the point of the shoulder). The carina is a sensitive area for eliciting cough reflex and collecting material from the lower airways. After removing the stylet, the tube was held in place and the animal placed in lateral recumbency. Approximately 2 mL/kg sterile saline was instilled (i.e., 240 mL/adult lion) while gently rocking the lion to distribute fluid. Thoracic coupage (a technique to loosen mucus by tapping the chest with cupped hands) was performed while aspirating fluid. Samples were collected by either manual aspiration using a sterile 60-mL catheter-tipped syringe or portable suction pump into a sterile canister. Samples were immediately transferred to sterile 50-mL conical tubes and placed on ice bricks until transported to the laboratory.

Lavage fluid was centrifuged at 1,500 × G for 10 min. The pellet was resuspended in approximately 5 mL of supernatant, and aliquots were frozen at −80 C until cultured. Samples were then transferred to a 50-mL conical tube and centrifuged at 1,500 × G for 15 min. The supernatant was decanted and the remaining pellet was suspended in 1 mL BD MycoPrep™ (Becton Dickinson, Franklin Lakes, New Jersey, USA) and incubated for 15 min at 37 C. Samples were neutralized with 18 mL phosphate-buffered saline (PBS), centrifuged for 15 min at 1,500 × G, and the supernatant was decanted. The remaining pellets were suspended in 1 mL PBS, and 500 µL of the suspension was inoculated into a Mycobacteria Growth Indicator Tube (MGIT) and incubated in a BACTEC™ MGIT 960 Mycobacterial Detection System (both Becton Dickinson). Cultures that were Ziehl-Neelsen stain-positive were identified to species by sequencing fragments of the 16S ribosomal DNA (Harmsen et al. 2003) and gyrB genes (Huard et al. 2006).

Using this technique, M. bovis was isolated and identified from eight living lions. M. bovis infection was first recognized in the Kruger National Park lion population in 1995 (Keet et al. 1996). It has been hypothesized that lions, as an apex predator, acquired M. bovis from infected African buffalo (Syncerus caffer). Lions have been considered potential maintenance hosts for M. bovis in this ecosystem, based on speculated transmission of organisms in respiratory secretions through aerosols, bite wounds, or grooming (Keet et al. 1996; Michel et al. 2006). However, there are no studies directly investigating spread among lions.

Similar to previous reports of tissue culture-positive animals (necropsy samples), adult lions were the animals from which most positive cultures were obtained (Table 1) (Keet et al. 2000). The median age of infected animals was 8.75 yr. Shedding of mycobacteria in respiratory secretions is associated with active pulmonary disease and is observed with M. tuberculosis and M. bovis infection in a variety of species (Mikota 2008; McNerney et al. 2010; Miller and Sweeney 2013). The median age of culture-positive lions is consistent with the chronic nature of mycobacterial infection and suggests progression to pulmonary disease. Clinical findings consistent with tuberculosis (i.e., emaciation, lymphadenopathy, elbow hygroma) were observed in all M. bovis culture–positive lions >2 yr old, except one (lion 6; Table 1).

Lions with active disease that are shedding viable mycobacteria present potential transmission risk to other lions. The viability of airborne M. bovis decreases with a half-life of 1.5 h (Gannon et al. 2007). However, M. bovis introduced into materials such as soil, water, and feedstuffs may persist in the environment for 43–112 d and may create the opportunity for infection through ingestion of contaminated material or contamination of wounds (Palmer and Whipple 2006; Fine et al. 2011). Investigations of the distribution of pulmonary lesions in lions suggest that 40% of cases were most likely the result of inhalation of infectious aerosols (Maas 2013). In our study, lions 2 and 3 (Table 1) were pride mates, suggesting either intraspecies transmission or a common infectious source.

Bronchoalveolar lavage (BAL) samples are reported to be more sensitive than sputum samples for detection of mycobacteria in human patients with known culture-positive tuberculosis (TB) (Liam et al. 1998; Tueller et al. 2005). In one study, only 39% of known TB cases had positive sputum smears compared with 83% detection by either positive smear or PCR using BAL fluid (Tueller et al. 2005). Bronchoalveolar lavage can also increase detection of M. tuberculosis in sputum-negative human patients (Liam et al. 1998). Previous studies in lions have inferred infection status based on TST results, postmortem pathologic changes, or confirmed infection by culture of tissues (Keet et al. 2010). In this study, we documented active shedding of pathogenic Mycobacterium in free-ranging lions using a minimally invasive technique for sample collection. The detection of M. bovis by bacterial shedding in respiratory secretions in 6.0% of lions tested was an important finding. Based on expectations of intermittent shedding, decreased recovery of organisms because of sampling technique (potential that only one lung was field sampled), and decreased viability associated with processing (freeze-thaw) and transport of samples, results from the tracheobronchial lavage cultures suggest that the prevalence of shedding of M. bovis may be >6.0% in this lion population. Because one of our aims was to determine whether lions could potentially be infectious, we did not perform PCR directly on lavage pellets. Although this technique could potentially increase the number of lions that would be classified as M. bovis-positive, it would not discern whether mycobacteria were viable and capable of transmitting infection. Future researchers should investigate techniques to improve recovery of mycobacteria in clinical samples including PCR of lavage samples, frequency of shedding, association of mycobacterial load with clinical disease presentation, and the risk factors associated with transmission of M. bovis in lions. These studies will improve our understanding of the role of lions as potential maintenance hosts for TB.

We acknowledge Marius Kruger, Nomkhosi Mathebula, Guy Hausler, and the staff of Veterinary Wildlife Services, South African National Parks for assistance with this study. Financial support was provided by Morris Animal Foundation grant D10ZO-039 and South African National Parks, Veterinary Wildlife Services. We acknowledge the National Research Foundation South African Research Chair Initiative for personal funding and support.