High female mortality caused by an atypical Mycobacterium species closely related to the Mycobacterium ulcerans-marinum complex in a colony of bearded dragons (Pogona vitticeps)

Abstract

A commercial breeding colony of bearded dragons (Pogona vitticeps) experienced an increase in mortality that affected females only. Before death, the animals had lost appetite and weight, were dehydrated, and some had labored breathing. Necropsy revealed granulomas in many organs (ovaries, lungs, liver, kidneys, heart, bone marrow) in which numerous acid-fast bacteria were identified. Bacterial isolation confirmed Mycobacterium spp., which was identified by whole genome sequencing as closely related to the Mycobacterium ulcerans-marinum complex. Due to the zoonotic potential of this bacterium and the poor prognosis for the remaining sick animals, the entire colony was culled and 7 animals were evaluated. The possible routes for introduction of this bacterium, the female predisposition to the disease, as well as the zoonotic potential of this microorganism are discussed.

Key clinical message:

An atypical Mycobacterium species closely related to Mycobacterium ulcerans-marinum complex can cause high female morality in captive bearded dragons.

Bearded dragons (Pogona vitticeps), lizards native to Australia, are very popular in zoos and as pets as they are easy to maintain, and, owing to their placid and friendly nature, safe for children to handle. However, zoonotic agents such as Mycobacterium and Salmonella can be carried by these reptiles and could raise health concerns for those who handle them (1,2). Mycobacteriaceae are a family of Gram-positive, aerobic, acid-fast bacteria, occurring as slightly curved or straight rods (3). Infections caused by Mycobacterium pathogens are well-known to occur in humans, birds, and mammals, and have also been described in fish, amphibians, and reptiles (4). Very few mycobacteriosis cases have been reported in bearded dragons and always in single individuals: a Mycobacterium marinum septicemia, and an osteomyelitis presumably associated with Mycobacterium chelonae have been described (5,6). We report here an unusual Mycobacterium species closely related to Mycobacterium ulcerans-marinum complex that caused systemic infection in a colony of bearded dragons, with a unique sex susceptibility of females.

Case description

In 2018, a 2.5-year-old female (Female 1) bearded dragon (Pogona vitticeps), from a group of 10 animals raised for breeding (commercial facility, Québec, Canada) died after a few weeks of declining appetite and loss of body weight. Notably, 1 mo before this event 2 females from this group had died after developing similar clinical signs. Those bearded dragons were imported in 2016 at the age of 3 to 6 mo from a breeding farm in California; no quarantine was performed on arrival. The animals were housed in a dedicated room, separately from other reptile and amphibian species that were also raised in this facility. The dragons were kept individually in plastic bins lined with papers that were replaced daily, and the bins were disinfected weekly with bleach (1:10) or Oxyfresh (Coeur d’Alene, Idaho, USA). The ambient temperature was maintained at 27°C, and the humidity kept at approximately 40%. A temperature gradient was assured in each bin through a ceramic heating plate and an infra-red lamp; the hottest point in the environment reached 35°C. A fluorescent UV light was installed in each bin. The bearded dragons had permanent access to fresh water, were fed daily with a mix of fruits, vegetables, and canned food (Zoo Med Adult Bearded Dragon Food, San Luis Obispo, California, USA). Insects (crickets, mealworms, silkworms, goliath worms) were given 1 to 2 times per wk. The females began to breed at 1.5 y, and the males were rotated between females for breeding sessions of a maximum 24 h. Female reproduction cycles were generally spaced by 3 wk during the mating period, with at least 3 consecutive mo of rest/y.

Veterinary evaluation of the animal group was carried out after the third death (i.e., Female 1) in the colony. At that time, 2 more females had stopped eating and drinking. The breeder also noticed that 1 of the 2 females made occasional abnormal head movements, and both females breathed with their mouth open. Physical examination of both females revealed abnormal body score (2/5), dehydration (persistent skin fold, sunken eyes), and weak musculature. Neither of the 2 males serving the females in the group showed clinical signs.

The carcass of Female 1 was submitted to the Laboratoire de Santé Animale de Saint-Hyacinthe, Québec, for post-mortem evaluation. The animal was in poor body condition (thin tail, poor internal fat reserves) and severely dehydrated (severe enophtalmia, sticky subcutaneous tissue). The extremity of its tongue was yellowish and necrotic and exhibited multifocal ulcers with underlying connective tissue mineralization. Numerous soft, yellowish granulomas, measuring 0.3 to 1 cm in diameter, were discovered in the coelomic cavity and within many organs (heart, liver, both lungs, kidneys, and ovaries) (Figure 1 A). The 2 ovaries were severely enlarged (5 cm × 3 cm × 3 cm) and their parenchyma completely effaced by numerous confluent granulomas (Figure 1 B). Microscopic examination revealed many round histiocytic granulomas in most organs examined, including the bone marrow, lungs, and liver (Figure 2 A, B). These granulomas were characterized by an accumulation of foamy and epithelioid macrophages, sometimes surrounding a caseous necrotic center, and surrounded by a small number of lymphocytes, plasma cells, fibrous tissue, and rare heterophils. Numerous acid-fast bacilli were revealed in the macrophage’s cytoplasm by a Ziehl-Neelsen stain (Figure 2 B, inset). Vitellus material was present on the serosa of some organs, eliciting a local granulomatous inflammation. In the liver, apart from the previously described granulomas, there was further inflammation, centered on bile ducts which were distended with bile, necrotic material, lipid vacuoles, cholesterol clefts, and random Gram-negative coccobacilli and bacilli. A routine bacterial culture on a Columbia blood agar medium was performed on the liver, ovary, and coelomic cavity, which yielded the growth of Salmonella spp. and contaminants. Based on the cumulative evidence, diagnoses of systemic mycobacteriosis and Salmonella septicemia were made.

Figure 1.

Necropsy of Female 1. A — The right lung (held by forceps) is covered with multiple yellowish, soft granulomas. B — Multiple granulomas are seen on serosae (arrows). The ovaries are severely enlarged, and their normal parenchyma is replaced by confluent, soft and yellow granulomas (asterisks). Note that the liver and heart have been removed from the coelomic cavity.

Figure 2.

Microscopic examination of granulomatous lesions in organs of Female 1. Hematoxylin-eosin-saffron staining, bar = 3 mm. A — Numerous round granulomas (some of them are indicated by asterisks) are randomly distributed in the lung parenchyma. B — Multiple confluent granulomas composed of foamy to granular macrophages and some multinucleated giant cells surround cores of caseous necrosis. The adjacent parenchyma is replaced by fibrous tissue and infiltrated by some lymphocytes and plasma cells. Acid-fast bacilli are present in the macrophages cytoplasm; Ziehl-Neelsen stain (inset).

Due to the zoonotic potential of Mycobacterium and Salmonella, in consideration of the absence of recommended treatment for this disease in reptiles plus the poor prognosis, the owner decided to euthanize the remaining bearded dragons in this group. Given concern of human exposure during the course of the animals’ disease, mycobacteriosis in particular, all animals (2 females and 5 males) were submitted for necropsy and Mycobacterium species identification. Only the females (designated further as Females 2 and 3) had lesions. In these females, many organs and serosae contained or were covered by granulomas similar to those seen in Female 1. Microscopic examination revealed many acid-fast bacilli in the lesions, suggesting mycobacteriosis. Salmonella spp. was identified in the liver of Female 3; no bacterial growth was observed for Female 2.

Tissue specimens from mycobacteriosis-positive animals were submitted to the Ottawa Animal Health Laboratory, Canadian Food Inspection Agency (OAHL, CFIA), where Mycobacterium isolation and identification was performed according to standard procedures (7,8). Visible growth of a subcultured isolate designated 2018/0516 was observed on Lowenstein-Jensen medium after 1 to 2 wk at temperatures of 25°C and 30°C, and after 3 to 4 wk at 37°C. Morphologically, the colonies were smooth, fairly flat, and opaque when grown on modified Middlebrook 7H11 agar medium. The isolate was non-chromogenic. Results of phenotypical and biochemical tests are presented in Table 1. Molecular identification procedures were based on amplification and sequencing of 16S rRNA (forward primer 5′-AGAGTTTGATCCTGGCTCAG-3′, reverse primer 5′-TGCACACAGGCCACAAGGGA-3′) and rpoB genes (forward primer 5′-GGCAAGGTCACCCCGAAGGG-3′, reverse primer 5′-AGCGGCTGCTGGGTGATCATC-3′) (9). According to the BLASTN analysis, the isolate 2018/0516 exhibited the highest 16S rRNA gene sequence similarities (99.7% across 960 bp) with Mycobacterium marinum and Mycobacterium ulcerans, and the highest rpoB sequence similarities (99.2% across 715 bp) with Mycobacterium marinum, Mycobacterium ulcerans, Mycobacterium ulcerans subsp. shinshuense, Mycobacterium liflandii, and Mycobacterium pseudoshottsii.

Table 1.

Phenotypic characteristics of 1) strain 2018/0516, and related Mycobacteria in the M. ulcerans-marinum complex; 2) M. marinum; 3) M. ulcerans; 4) M. ulcerans subsp. shinshuense; 5) M. liflandii; 6) M. pseudoshottsii.

Characteristics	1)	2)	3)	4)	5)	6)
Growth rate	slow	slow	slow	slow	slow	slow
Optimal temperature (°C)	25, 30	30	30	25, 32	28	23
Colony morphology	S	S	R	R	R	R
Pigment production	N	P	N	P	N	P
Growth, TCH	+	+	−	−	ND	+
Growth, 5% NaCl	−	−	−	−	ND	−
Arylsulfatase	+	+	−	−	ND	−
Nitrate reduction	−	−	−	−	ND	−
SQ Catalase	−	−	−	−	ND	−
Catalase 68°C	−	−	+	+	ND	−
Tween 80 hydrolisis	−	+	−	−	ND	−
Pyrazinamidase	+	+	−	−	ND	−
Urease	+	+	−	+	ND	+
Reference		(29,30)	(29,31)	(32)	(33)	(34,35)

S — Smooth; R — Rough; N — Non-chromogenic; P — Photochromogenic; TCH — Thiophene-2-carboxylic acid hydrazide; SQ — Semiquantitative; ND — No data.

Cumulatively, the results of the molecular, and phenotypic and biochemical characterization did not strongly correlate with either M. marinum or M. ulcerans species (Table 1). Biochemically, the isolate was most similar to M. marinum, but, unlike M. marinum, was nonchromogenic and negative in a Tween hydrolysis test. The genome of the isolate 2018/0516 was sequenced using Illumina sequencing technology (Nextera XT DNA library preparation, MiSeq reagent kit v2 2× 250, generated 1.1 Gbase) and de novo assembled (10). Evaluation of the strain 2018/0516 genome sequencing data in silico revealed the absence of genetic elements associated with mycolactone synthesis (pMUM plasmids, mls genes) and advocated that the strain 2018/0516 was not a mycolactone producing Mycobacterium spp. To determine the approximate phylogenetic relationship of the isolate 2018/0516, the isolate genome sequence was compared to NCBI GenBank RefSeq sequences of Mycobacterium spp. using an in-house computational script (https://github.com/duceppemo/genome_comparator) with the embedded MASH software (11). According to the resulting dendrogram (Figure 3), the isolate 2018/0516 was distinct but most closely related to the M. ulcerans-marinum complex (comprised of species M. marinum, M. ulcerans, M. pseudoshottsii, M. liflandii, and M. shinshuense) (12).

Figure 3.

Phylogenetic tree illustrating the relation of the isolate 2018/0516 to species of the Mycobacterium ulcerans-marinum complex. The tree, based on the comparison of whole genome sequences, is constructed using the MinHash algorithm and neighbour-joining method. The original analysis included the isolate 2018/0516 and Mycobacterium spp. RefSeq sequences (n = 426) available in the NCBI GenBank. Only the isolate 2018/0516 and the most closely related Mycobacterium spp. cluster (i.e., the M. ulcerans-marinum complex) are presented.

Discussion

We report here a case of a mycobacteriosis outbreak in a colony of captive bearded dragons, caused by an unusual Mycobacterium species closely related to Mycobacterium ulcerans-marinum complex. Nearly 200 Mycobacterium species are recognized, and divided into 3 groups based on their clinicopathological outcome: the Mycobacterium tuberculosis complex (pathogens causing tuberculosis), Mycobacterium leprae (causative pathogen of leprosy) which forms a group itself, and non-tuberculous mycobacteria (NTM) composed of all other species including Mycobacterium ulcerans-marinum complex (13). NTM are saprophytic organisms widely distributed in the environment, living in soil and aquatic habitats and sometimes in artificial environments such as aquariums (4). NTM infections of lizards are uncommon and sporadic, have been described only in captive animals, and are supposedly contracted through ingestion or wounds to skin, or respiratory and urogenital tracts (4). Fish tanks are the most common source of human infection by M. marinum and, interestingly, bearded dragons reportedly can contract this bacterium through ingesting contaminated aquarium fish (guppies) (6,14). In our case, the source of infection remains unknown, as lizards did not receive any fish or marine product, nor had any direct contact with an aquarium (although there were amphibians raised in partly submerged habitats in other sections of the building). We could not exclude the possibility that the fresh fruits, vegetables, or insects that were given as food were contaminated by this atypical Mycobacterium species. Aquatic insects could be carriers of M. ulcerans, and other NTM have been isolated from cockroaches in hospitals (15,16). Mealworms are often used as food for reptiles and birds, as was the case here, and in an experimental infection of mealworms with Mycobacterium avium avium and Mycobacterium avium paratuberculosis, viable mycobacteria were recovered from their bodies (16,17). Another report substantiates the ability of different NTM species to survive and grow in free-living amoebae in water networks (18). However, in this report, insects from the same supplier and water from the same source were given to other animals in the facility, with no signs of infection. Finally, we could not exclude the possibility that one or more of these dragons were already asymptomatic carriers of this mycobacteria when purchased. Latency is a well-known phenomenon in human tuberculosis, in which an infection with the microorganism in one individual can elicit a transient and asymptomatic infection in some other individuals, and then becomes quiescent, sometimes for years, until the immune system of the host is weakened, and enables the bacteria to replicate and spread (19). Notably, zebrafish latently infected by M. marinum are a well-known animal model developed to study latent tuberculosis in humans (20). However, the ability of reptiles to establish such a persistent mycobacteria infection is unknown.

Although we could not identify the primary source of introduction of the mycobacteria in the colony, we hypothesize that male bearded dragons could have acted as transmitters of the organism through frequent relocation for breeding purposes. Indeed, even if NTM are not part of the normal bacterial flora of animals, they may be isolated from the skin, the upper respiratory tract, and intestinal and genital tracts of asymptomatic animals (14). In our case, only females developed lesions and the males did not exhibit any signs of infection. The cause of this sex predilection is unknown.

Immunosuppression is known to favor the development of mycobacteriosis in humans (as in those with HIV infection) and in animals (3,4). In lizards, seasonality and reproduction, among other factors, have an effect on the efficiency of the immune system (21,22). Reproduction is generally prioritized by the body over the immune function during the mating season because these systems compete for energy and protein, and many steroid hormones involved in reproduction also have an immunomodulatory function. For example, during vitellogenesis, wound healing or microbiocidal activity are decreased in side-blotched and/or wild tree lizards (23,24). Interestingly, it was demonstrated that environment and nutrition had an effect on immunity, as wild free-ranging female dragons had a decrease in skin healing rate compared to well-nourished laboratory-kept female dragons at the same stage of the reproductive cycle (23). As our female dragons were in their breeding period, we could hypothesize that their immune response was weakened due to reproduction, and that other unknown husbandry factors could have enhanced immunosuppression. In another study, a greater phagocytic activity of macrophages was reported in female wall lizards compared with males (25). The ability of Mycobacterium species to survive in macrophages allow the bacteria to avoid detection and escape the immune system of the host (26). We could then speculate that female lizards, due to their increased capacity to phagocytize mycobacteria, could harbor more dormant organisms than males and be more prone to a systemic infection under stressful conditions that suppress immunity.

Salmonella spp. was cultured from tissues of the first female dragon and of one of the 2 females submitted later. Reptiles, both captive and free-ranging, are well-known carriers of Salmonella, which is a component of their normal bacterial intestinal flora (27). Most individuals remain asymptomatic, but debilitated animals could develop a systemic disease, generally concurrently to another infection (27). Interestingly, a cholangiohepatitis was present in Female 1, in which Gram-negative bacilli were visualized. Even if their exact taxonomic speciation was not determined, we could hypothesize that at least some of those were Salmonella coming from the intestine, eliciting a local inflammation, and terminally spreading to distant organs.

Infections caused by NTM are a growing human health problem, partially due to increasing interactions between immunologically compromised people (elderly, chemotherapy recipients, etc.) and companion animals. Diagnostic testing and therapy of NTM infections depend on the mycobacterial species involved. Also, an awareness of potential exposure to specific identified NTM could support risk mitigation strategies. Among NTM, 2 species are recognized as true pathogens for humans, namely, M. marinum and M. ulcerans (28). There have been no reports of mycobacteriosis in personnel handling the infected animals described in this study, and we acknowledge that the potential clinical implication caused by this uncharacteristic NTM species needs further consideration. Nevertheless, our case presents the importance of taxonomic speciation and characterization of infection causative agents, such as Mycobacterium, in companion animals.

In conclusion, mycobacteriosis could cause numerous mortalities in captive bearded dragons and specifically target females possibly due to sex-affiliated physiological specificities. This is the first time that an unusual Mycobacterium pathogen closely related to the M. marinum-ulcerans complex is reported and associated with lesions in animals.