Fatal Outbreak of Mycoplasma capricolum Pneumonia in Endangered Markhors

A pneumonia outbreak reduced the numbers of a wild population of endangered markhors (Capra falconeri) in Tajikistan in 2010. The infection was diagnosed by histologic examination and bacteriologic testing. Mycoplasma capricolum subsp. capricolum was the sole infectious agent detected. Cross-species transmission from domestic goats may have occurred.

A pneumonia outbreak reduced the numbers of a wild population of endangered markhors (Capra falconeri) in Tajikistan in 2010. The infection was diagnosed by histologic examination and bacteriologic testing. Mycoplasma capricolum subsp. capricolum was the sole infectious agent detected. Cross-species transmission from domestic goats may have occurred.
M ycoplasma capricolum subsp. capricolum and M. capricolum subsp. capripneumoniae are closely related subspecies of the M. mycoides cluster (1). Whereas M. capricolum subsp. capripneumoniae is the etiologic agent of contagious caprine pleuropneumonia (CCPP), a severe and typically lethal respiratory disease, M. capricolum subsp. capricolum infection is usually not fatal and instead results in chronic infl ammation in a variety of organs, including joints, udder, eyes, and lungs (2). M. capricolum subsp. capricolum infection occurs worldwide and appears widespread but has rarely been found in species of small ruminants other than domestic goats and, more occasionally, sheep (2,3). This lack of evidence may be partially because few studies have applied sensitive molecular techniques for its detection in nondomestic ruminants (2,3). Domestic goats can carry M. capricolum subsp. capricolum asymptomatically, notably in the ear canal (4), and pose an insidious risk for cross-species transmission with sympatric wild caprines (2,3).

The Study
The markhor (Capra falconeri) is an endangered wild goat in a continuous decline; the global population is <2,500 mature animals (5). In Tajikistan, <350 animals may survive in fragmented subpopulations in the remote Hazratishoh and Darvaz mountain ranges along the Afghanistan border (6). They live sedentarily over relatively small home ranges, moving <5 km per day (7). Throughout its range, the markhor has to forage in close proximity to domestic goats (8) and is therefore prone to infections of contagious agents transmitted by these animals.
During September 17-October 18, 2010, eleven markhors that displayed labored breathing and 64 markhors that had recently died were found in 5 localities ( Table) in the district of Shuroabad, Khatlon Province, usually in close proximity to water sources ( Figure 1, panel A). All but 4 carcasses were too scavenged for thorough examination in the fi eld, and 1 dying adult female was sent on September 20 to the Republican Veterinary Laboratory in Dushanbe for necropsy. The most relevant necropsy fi ndings noted in the fi eld were the following: an abundant serous to mucopurulent nasal discharge; and internally, severe pneumonia associated with a variable level of pleural fl uid. The female markhor examined in Dushanbe showed gray areas of consolidations in the apical and cardiac lobes of the right lung and yellow pleural fl uid ( Figure 1, panel B). The cut surface of the affected lobes revealed a fi ne granular texture, and mucopurulent exudate could be expressed from the bronchi. The joints, eyes, and udder were not affected. Regarding indications of CCPP, although gross lesions were limited to the thoracic cavity, fi brinous pleurisy (9) was not observed. The histopathologic fi ndings were consistent with proliferative interstitial pneumonia associated with multifocal suppurations ( Figure 1, panel C). Findings included diffuse thickening of the interlobular septa, alveolar epithelialization, and polymorphonuclear leukocytes in alveolar and bronchiolar spaces ( Figure  1, panel D). Also, disseminated and abundant neutrophil infi ltration of alveolar spaces was found that, when  ruptured, created multifocal nodules. The left lung was congested.
No bacteria were isolated from pleural fl uid that was obtained aseptically and inoculated onto 5% sheep blood agar. However, when a healthy 5-month-old domestic goat (C. hircus) was inoculated intratracheally with 5 mL of this pleural fl uid, pneumonia developed within 8 days. To evaluate the possibility that the outbreak might have been caused by M. capricolum subsp. capripneumoniae, which was fi rst identifi ed in Tajikistan in 2009 (10), samples were sent to Centre de Coopération International en Recherche Agronomique pour le Developpement, Montpellier, France, a reference laboratory for CCPP for the World Organisation for Animal Health. Pleural fl uid from the markhor was centrifuged at low speed (500 g for 10 min) to eliminate infl ammatory cells, and DNA was extracted from the supernatant with the DNeasy blood and tissue kit (QIAGEN, Courtaboeuf, France), according to the manufacturer's instructions. Real-time PCR (11), specifi c for M. capricolum subsp. capripneumoniae, provided negative results, whereas a partial 16S rRNA gene sequence could be amplifi ed and sequenced (12). BLAST analysis (http://blast.ncbi.nlm.nih.gov/Blast.cgi) showed that the sequence corresponded to that of a mycoplasma from the M. mycoides cluster. Markhor pleural fl uid and lung tissue from the domestic goat were then injected into modifi ed Hayfl ick broth (1) and onto agar plates and incubated at 37°C in 5% CO 2 . Typical Mycoplasma colonies appeared on solid medium after 1-2 days. A clone from the markhor culture, named 10074 (1.4), was used for PCR amplifi cation and sequencing of fusA, glpQ, gyrB, lepA, and rpoB partial gene sequences as described (1). The fusA sequence was identical to a previously submitted sequence (GenBank accession no. EF071735). All other sequences were novel and have been assigned new GenBank accession nos.: HQ882179 (glpQ), HQ882180 (gyrB), HQ882181 (lepA), HQ882178 (rpoB). The 5 protein-coding sequences were then concatenated and incorporated in the M. mycoides cluster phylogenetic tree (Figure 2). The strain clustered with M. capricolum subsp. capricolum and was therefore identifi ed as belonging to this subspecies. The isolate from the inoculated domestic goat was undistinguishable from 10074 (1.4) by sequencing of the 5 housekeeping genes.
The pattern of infl ammation in the right lung in the necropsied markhor was consistent with M. capricolum subsp. capricolum respiratory infection in domestic goats (2), and the presence of M. capricolum subsp. capricolum in pleural fl uid, the lack of fi ndings indicative of alternative etiologic agents, and the reproduction of the disease in the domestic goat with isolation of M. capricolum subsp. capricolum from the lungs support a causal relationship between M. capricolum subsp. capricolum and disease in markhors.

Conclusions
Although because of diffi cult fi eld conditions, only 1 markhor could be thoroughly investigated, these fi ndings support the hypothesis that M. capricolum subsp.
capricolum, a newly recorded pathogen in free-ranging wild ruminants, may be responsible for the pneumonia epizootic observed in endangered markhors. This outbreak claimed ≈20% of the population remaining in Tajikistan, and more markhor deaths might have remained undetected.
The source of Mycoplasma infection in markhors is unknown, but domestic goats, which have contact with markhors, particularly in the summer, might have been responsible for the emergence of M. capricolum subsp. capricolum in this wild species. In November 2008, a disease resembling CCPP affected domestic goats and, to a lesser extent, sheep in Shuroabad District (<40 km from the area where the dead markhors were found) with a case-fatality rate of 20%-30% (13). However, the origin of this outbreak could not be investigated. Although the clinicopathologic features of the disease in the markhors resembled CCPP, M. capricolum subsp. capricolum was identifi ed as the most probable causative agent. In fact, several mycoplasmas have been associated with respiratory diseases in ruminants, including M. capricolum subsp. capricolum infection in young domestic goats (2,14). Further testing should evaluate the presence of sick animals or asymptomatic carriers of M. capricolum subsp. capricolum in livestock and wildlife in Shuroabad District.
Environmental and nutritional stressors may exacerbate the susceptibility of ruminants to Mycoplasma infections (2,15). The disease appeared at the end of summer, when markhors are forced by livestock and guard dogs to retreat to suboptimal habitats with poor forage (8). It was also the end of the dry season, when markhors may have contact with livestock at the few remaining water sources.
Findings from this study support the conclusion that the markhor is vulnerable to M. capricolum subsp. capricolum. In the generalized context of increasing encroachment of livestock into wild habitats, markhors and other wild caprines might be at risk for future mycoplasmosis outbreaks. This case emphasizes the need for continuous disease surveillance in domestic animals that have contact with valuable wildlife resources.

Acknowledgments
We thank L. Boulouha for performing the histopathologic examination; I. Ikromov and his staff who reported the outbreak, centralized reports of dead markhors, and captured the sick female markhor; and P. Zahler and 2 anonymous reviewers for their helpful comments.  capricolum strains, as well as type strains corresponding to other species or subspecies from this cluster and M. putrefaciens, used as outgroup. The tree, derived from distance analysis of 5 concatenated protein-coding sequences (fusA, glpQ, gyrB, lepA, rpoB), was constructed by using the neighbor-joining algorithm. Bootstrap percentage values were calculated from 500 resamplings, and values >90% are indicated. Scale bar indicates distance equivalent to 1 substitution per 100 nt positions. Note that the branch corresponding to the outgroup has been shortened, as indicated by 2 parallel bars.