Prevalence of Mycoplasma spp. in the Respiratory Tract of Healthy North American Bison (Bison bison) and Comparison with Serum Antibody Status

Mycoplasma bovis is a common cause of pneumonia, arthritis, and mastitis in cattle and is occasionally implicated in reproductive disorders (Maunsell et al. 2011; Ridley and Hately 2018). It imposes a considerable economic burden on cattle production worldwide (Dudek et al. 2020). Around the turn of this century M. bovis was recognized for the first time as a pathogen in North American bison (Bison bison; US Department of Agriculture 2013). Investigation of disease outbreaks revealed it as a primary cause of pneumonia, polyarthritis, necrotic pharyngitis, pleuritis, dystocia, and abortion, with high morbidity and significant mortality (Dyer et al. 2008; Janardhan et al. 2010; Dyer et al. 2013; Register et al. 2013b; Bras et al. 2016; Bras et al. 2017a). Mycoplasma bovirhinis, regarded as a commensal, is also frequently recovered from the respiratory tract of both healthy and diseased cattle (Maunsell and Donovan 2009), but whether it inhabits bison is unknown.

Efforts to control mycoplasmosis in cattle have been only marginally successful. Antibiotic treatment is frequently ineffective (Maunsell and Donovan 2009), and similarly poor outcomes are the norm in affected bison (US Department of Agriculture 2013). Cattle vaccines stimulate measurable immune responses but offer limited efficacy under field conditions (Maunsell and Donovan 2009) and are not recommended for use in bison. An autogenous vaccine containing bison isolates from outbreaks of mycoplasmosis is available in the US, but no data exist regarding its efficacy. Strategies to control the frequency and impact of mycoplasmosis in bison are urgently needed; their development is hampered by gaps in our understanding of M. bovis epidemiology. One of several key unanswered questions is whether healthy bison harbor M. bovis in the absence of clinical signs, thereby serving as hidden sources of transmission. Studies in which apparently healthy bison with no history of mycoplasmosis tested positively for antibodies reactive with M. bovis are consistent with inapparent infection (Register et al. 2021; Bras et al. 2017b), but there are no definitive data based on the results of samples cultured from healthy bison. We aimed to determine whether M. bovis can be recovered from the respiratory tract of healthy bison. Sera from the same bison were tested with an M. bovis enzyme-linked immunosorbent assay (ELISA), and the results were compared with those from culture.

During 2012–19, we collected deep nasal swabs from 391 bison, each sampled only once, at 11 locations in the US (n=304) and two locations in Canada (n=87; see Table 1). Bison at 10 locations were wide-ranging and minimally managed, handled yearly or biennially (n=333); those at the other three sites were restrained by fencing and handled more frequently (n=58). Most bison (304/391; 77.7%) were adults when sampled; 21.5% (84/391) and 0.8% (3/391) were juveniles or calves, respectively. All bison appeared healthy at the time of sample collection; none had been previously vaccinated against M. bovis. Twelve herds had no prior history of mycoplasmosis. Herd USA-8 had suffered a clinical outbreak of M. bovis approximately 6 mo prior to the date of sample collection but not more recently. Tonsil swabs from healthy bison processed at abattoirs in the US (n=55) and Canada (n=53) were also evaluated.

Swabs were used to inoculate selective PPLO broth (PPLO broth supplemented with 0.05% thallium acetate and 500 IU/ml penicillin G). Cultures were incubated at 37 C in 5% CO₂ for at least 24 h, or until growth was visually apparent, up to 72 h in the event of no detectable growth. At the end of the incubation period 5 µL of every culture was tested in an M. bovis–specific PCR (Clothier et al. 2010). To identify other mycoplasmas potentially present in M. bovis–negative cultures, an additional 5 µL was tested with a PCR amplifying ∼780-base pair of the 16S rRNA gene from Mycoplasma spp. known to infect cattle (Miles et al. 2004). Amplicons from positive reactions were treated with ExoSAPIT (Thermofisher, Waltham, Massachusetts, USA) and sequenced at the Iowa State University DNA Facility (Ames, Iowa, USA). Consensus sequences were deduced from a minimum of two high-quality reads, with at least one from each strand. An aliquot of each first-passage culture was serially diluted and plated on selective PPLO agar. Plates were incubated until Mycoplasma-like colonies were evident, or up to 5 d in the case of negative cultures. Well-isolated, suspect colonies from PPLO plates were used to inoculate fresh PPLO broth, which was incubated and tested using the M. bovis–specific and 16S rRNA gene PCRs, as described earlier. Samples were considered positive on the basis of PCR results obtained with the first-passage broth culture, whether or not colonies were recovered from the subsequent passage on PPLO agar.

We collected lung tissue postmortem from 25 bison being culled from herd USA-2 at the time of sampling. All lungs appeared normal. Tissues were weighed, minced with sterile scissors, diluted 1:10 (weight/volume) in PPLO broth, and homogenized in C tubes using a gentleMACS Octo Dissociator (Miltenyi Biotec, Auburn, California, USA). For each homogenate, 300 µL was cultured in selective PPLO broth and tested for M. bovis and other mycoplasmas as described for nasal swabs.

We collected serum at the time that nasal swabs were obtained for 381 bison. Each was tested twice with the M. bovis BIO K 260 ELISA (Bio-X Diagnostics, Rochefort, Belgium) using independently made dilutions, as described previously (Register et al. 2013a). This ELISA has not been validated for use with bison but performed well in a prior study with bison with known exposure history to M. bovis (Register et al. 2013a). Mean test values were used to assign each sample a numeric score, from 0 (negative) to 5+, using the metric provided by the manufacturer. Sera were recorded as positive only when a result of 1+ or higher was obtained on both occasions tested. Results for 284 sera have been reported previously (Register et al. 2021).

Overall, 14/391 nasal swabs from bison (3.6%) were positive for M. bovis, with viable colonies recovered from each sample. Positive animals were identified in two herds in the US and one in Canada. Herd USA-1 accounted for 9/14 positives (64.3%), with positive bison found on two of the three occasions sampled. One positive bison was detected in herd USA-8, which had experienced an outbreak of mycoplasmosis 6 mo earlier. The remaining four positive bison were from herd Can-1. Herd-specific prevalence varied from 2.5% for USA-8 to 87.5% for USA-1 (in 2012).

First-passage nasal swab cultures from 133 M. bovis–negative bison tested positive with the Mycoplasma-specific 16S rRNA PCR (Table 1). The species in each positive culture was discerned based on the highest-scoring match obtained when the PCR amplicon sequence was used to query the National Center for Biotechnology Information nonredundant nucleotide database (2020). All top matches were ≥98% identical to ≥98% of the query sequence, with E values of 0.0. Mycoplasma bovirhinis was identified in 120 bison (30.7% of all bison), and Mycoplasmalike colonies identified as M. bovirhinis were recovered from all but 22 of the corresponding first-passage cultures. Mycoplasma bovirhinis was found in 10/13 herds or locations; the three remaining cannot confidently be considered free of M. bovirhinis since only a few individuals from each were tested (1–8 bison). Herd-level prevalence among those positive for M. bovirhinis ranged from 9.1% (USA-4) to 77.8% (USA-3, in 2019), mean 34.0%.

First-passage cultures from 11 bison representing three herds in the US were positive for Mycoplasma bovoculi, with colonies recovered from all but one. One bison from USA-5 was positive for Mycoplasma arginini; colonies of that organism were also obtained from that animal. One sample from USA-3 was positive for Mycoplasma dispar, but no colonies were recovered.

Only one abattoir-sampled bison tonsil, from the US, tested positively for M. bovis (0.9%); an isolate was obtained from that sample. Ten tonsils from the US and 12 from Canada were positive for M. bovirhinis (20.4%), with isolates recovered from 16.

Considering culture results from both nasal swabs (n=391) and tonsils (n=108), M. bovis was isolated from the upper respiratory tract of 15/499 bison (3.0%), while M. bovirhinis was detected in 142 (28.5%).

Mycoplasma bovirhinis was detected in first-passage cultures of four lung samples (16.0%), and M. bovirhinis colonies were recovered from two of those. Additionally, M. bovirhinis was detected in the nasal cavities of two bison with M. bovirhinis–positive lungs; no mycoplasmas were identified in the nasal cavities of the remaining two. No other Mycoplasma spp. were found in any other lung sample. Findings are summarized in Table 2, with results from nasal swabs and ELISA (described soon) included for comparison.

Antibodies reactive with M. bovis were detected in 25/381 bison sera tested (6.6%) but in only 46.2% of bison from which the mycoplasma was recovered (Table 3). Positive results were also obtained for 5.4% and 9.1% of bison culture-positive for M. bovirhinis or M. bovoculi, respectively, and 4.6% of bison in which no mycoplasmas were detected.

Our data reveal that M. bovis may reside in the upper respiratory tract of healthy bison with no apparent history or clinical signs of mycoplasmosis and that carriers may not produce detectable antibodies. Accordingly, the absence of neither clinical signs nor M. bovis–reactive antibody provides definitive evidence of freedom from infection with M. bovis. Whether healthy carriers are at higher risk for subsequent disease requires further investigation.

Our finding that 5.2% of M. bovis–negative bison were seropositive (19/368; Table 3), most with moderate to high levels of antibody (a score ≥2+), suggests possible cross-reactivity between antibodies elicited by other mycoplasmas and the ELISA capture antigen. Similarly, Bras et al. (2017b) found that 23/552 bison (4.2%) with no history of mycoplasmosis tested positively with a custom ELISA. We cannot dismiss the possibility that some bison M. bovis–negative at the time of sampling were seropositive owing to a prior, unrecognized infection. Also unknown is the limit of detection for M. bovis with our methods. Nonetheless, these explanations seem unlikely to fully account for our results, especially as we cultured other mycoplasmas more difficult to propagate than M. bovis. Studies directly addressing the question of cross-reactivity, including comparative immunoblotting with extracts from the mycoplasmas identified here as most prevalent among bison, are ongoing.

This study was conducted in accordance with protocols approved by the Institutional Animal Care and Use Committee of the National Animal Disease Center and other participating institutions. Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the US Department of Agriculture. USDA is an Equal Opportunity Employer.