Necropsy of a wild chimpanzee found dead in Taï National Park, Côte d’Ivoire revealed a large, yellow-white granuloma of 5 × 6 × 3 cm in the liver and several smaller ones in the spleen parenchyma and the mesenterial lymph nodes. All other organs appeared unaffected; the lungs could not be evaluated in full because the leopard had consumed most of the tissue. 16S rDNA testing of the frozen tissue samples from the dead wild chimpanzee indicated the presence of a Mycobacterium sp. in various tissues. PCR–restriction fragment length polymorphism analysis confirmed MTBC in DNA preparations from spleen and mesenterial lymph nodes (28). MTBC was also confirmed by real-time PCR (29) in lung, spleen, liver, and colon abscesses.

Histopathologic examination of the liver, spleen, and lymph nodes revealed a chronic granulomatous inflammation within the altered tissues (Figure 2, panels A–C). Multiple unencapsulated granulomas of varying sizes were observed in the spleen (Figure 2, panel A) and liver (Figure 2, panel B). These lesions were composed of epitheloid macrophages, few granulocytes, and multinucleated Langhans giant cells. Larger tuberculoid lesions in the liver and the lymph nodes contained a prominent central necrotic core surrounded by epitheloid cells and a few scattered Langhans giant cells. The periphery of the granulomas was demarcated by variable amounts of fibrous connective tissue and infiltrates of lymphocytes interspersed with few Langhans giant cells (Figure 2, panel C). Both intra- and extracellular acid-fast bacilli were present in the lesions (Figure 2, panel D). Taken together, these lesions were characteristic of TB and indicative of hematogenous spread and generalization of the disease.

After 23 days’ incubation, the lymph node preparations yielded typical mycobacterial colonies. The isolated MTBC strain exhibited a slow growth on Middlebrook 7H11 agar and yielded colonies after 43 days, compared with 27 days for M. tuberculosis. The rough surface of the colonies and the irregular spreading margins were typical features of MTBC (Technical Appendix Figure 1).

Whole-genome sequencing was conducted by using the 454 and Illumina platforms (Technical Appendix Table 1). Phylogenetic reconstruction using previously published MTBC genomes representative of the MTBC’s global diversity (18) confirmed that the chimpanzee strain belonged to MTBC but not to any of the known phylogenetic lineages (Figure 1). Specifically, the chimpanzee strain grouped with strains from the human-associated lineage 6, sharing 32 SNPs with this lineage, but was separate from the lineage leading to most of the animal-adapted MTBC. To further test whether the chimpanzee strain represented a new lineage rather than a variant within lineage 6, we sequenced the genome of 9 lineage 6 clinical strains from TB patients originating from different West-African countries (Technical Appendix Table 1). Our phylogenomic analysis revealed that the chimpanzee strain harboured 881 exclusive SNPs, not found anywhere else in the global MTBC phylogeny, even when including these additional lineage 6 strains. Moreover, the maximum number of SNPs between the 2 most divergent lineage 6 strains was only about half (783 SNPs) of the minimum number of differences between the chimpanzee strain and the most closely related lineage 6 strain (1,405 SNPs). When all MTBC lineages were considered, pairwise SNP distances among any 2 strains belonging to a particular lineage were always markedly lower than the minimum number of differences between the chimpanzee strain and the most closely related lineage 6 strain (Technical Appendix Figure 3).

To investigate the distinctiveness of the chimpanzee strain we isolated, we used a principal component analysis as an additional clustering method. The first 3 principal components were used to generate a 3-dimensional scatter plot (Technical Appendix Figure 4). This analysis confirmed that the chimpanzee strain did not group with lineage 6. Taken together, these results strongly suggest that the chimpanzee strain belongs to a distinct MTBC population, separate from the human-associated lineage 6.

To further confirm the uniqueness of the chimpanzee strain, we compared the chimpanzee spoligotype (Figure 3) with 2 large international databases that encompass >8,702 spoligotyping profiles corresponding to >58,187 MTBC isolates from global sources [33]; www.mbovis.org). We found that none of the spoligotyping patterns included in these databases (including 64 spoligotyping profiles from Côte d’Ivoire) matched the pattern of the chimpanzee strain (30) (Figure 3).

Large-sequence polymorphisms have been used as phylogenetic markers for MTBC (4,7,8). We found that the region of difference (RD) 9 was absent in the chimpanzee strain (4) (Figure 1). In addition, this strain harbored deletions in RD7, RD8, and RD10, which supports its phylogenetic relationship with lineage 6 (Technical Appendix Table 2 and Figure 2). However, the chimpanzee strain harbored only 1 of 14 lineage 6–specific deletions and did not contain the lineage 6–specific region RD900 (Technical Appendix Table 3 (29,31). Hence, this deletion-based analysis also supports a related, yet separate, phylogenetic position of the chimpanzee strain relative to lineage 6.

To investigate the possible presence of MTBC in other chimpanzees, 115 tissue samples from 28 chimpanzees that died in the same area (many from the same community) within the previous 10 years were tested by real-time PCR (14). However, no test results were positive. Consistent with the molecular analyses, necropsies and pathological evaluation of these animals revealed no signs suggestive of TB.

Chimpanzees are known to be susceptible to TB, and MTBC strains have been reported previously in captive chimpanzees (32,33). These 2 studies concluded that the infecting strains belonged to M. africanum and M. tuberculosis, respectively. Direct comparison of those isolates was not possible because of the limitations of typing techniques at that time. However, close contact with humans (i.e., the persons caring for these animals) suggests those captive chimpanzees were infected with human strains, as reported for other captive nonhuman primates.

By contrast, several lines of evidence support the view that the chimpanzee strain we report was not acquired from humans. First, its position on the MTBC phylogeny strongly suggests it belongs to a novel lineage. This notion is sustained by the fact that the minimum genetic distance between the chimpanzee strain and any of the nearest human strains (i.e., lineage 6) was larger than the corresponding distance between any 2 strains from the same lineage. Second, our PCA analysis showed that the chimpanzee strains did not group with lineage 6. Third, spoligotyping revealed a novel pattern among 58,187 clinical isolates from 102 countries, including Côte d’Ivoire. Fourth, genome deletion analyses corroborated the distinct phylogenetic position of the chimpanzee strain compared with known MTBC lineages. Fifth, researchers and their assistants who are in proximity of the chimpanzees at Taï National Forest are regularly screened for TB, but none has ever had a positive test result.

Two other animal-associated members of MTBC are known to cluster with lineage 6 rather than with the classical animal-adapted lineages: M. mungi and the Dassie Bacillus, which infect African mongooses and hyraxes, respectively (11,12). Whole genome analyses are not available for these organisms, but genomic deletion data have been reported (34). The chimpanzee strain we isolated did not harbor any of the specific deletions found in M. mungi or Dassie Bacillus (Technical Appendix Table 2) (34). Moreover, spoligotyping confirmed that the chimpanzee strain was distinct from M. mungii and Dassie Bacillus and from any other MTBC strain genotyped to date (30). However, 1 of the 32 SNPs shared between the chimpanzee genome and the lineage 6 strains also occurred in M. mungii or Dassie Bacillus. On the basis of this 1 SNP in Rv1510, which has been reported before (5), one could hypothesize that the chimpanzee strain and M. mungii and Dassie Bacillus might be related. However, genome-wide data will be necessary to define the exact phylogenetic position of M. mungii and Dassie Bacillus, and their relationship with the chimpanzee strain, in the global MTBC tree.

Even though chimpanzees maintain close social contacts with other members of their group, extensive necropsies and molecular screening of 28 chimpanzees from the same region yielded no additional case of TB infection, which suggests TB is rare in this chimpanzee population. This low prevalence could have several explanations. While we can likely disregard a human origin of the chimpanzee strain described here, we cannot exclude the possibility that this strain was acquired from another unidentified animal host, including other primates; chimpanzees are known to hunt other animals, including monkeys and small antelopes. The chimpanzee strain we isolated shared >1 SNP with M. mungi and the Dassie Bacillus, which are pathogens of 2 other small African mammals. On a more speculative note, and if it is assumed that the MTBC strain described is indeed chimpanzee-specific, this MTBC variant might be relatively attenuated and only marginally affect a chimpanzee’s health and longevity. This would enable sustained transmission and persistence of the pathogen in small host populations (35).

Although more work is needed to establish the prevalence, diversity, and clinical outcome of MTBC infection in wild chimpanzees and other great apes, from a conservation point of view, MTBC may join Ebola virus, Bacillus cereus biovar anthracis, and simian immunodeficiency viruses as a microorganism capable of threatening great apes in the wild (36). Our results suggest the effect of this MTBC strain on chimpanzee populations might be limited, but small outbreaks or single deaths can have a strong influence on the viability of isolated populations, particularly in great apes, which exhibit a slow reproductive rate and a high juvenile mortality rate (37).

Our study also sheds new light on the overall diversity of MTBC with implications for understanding the evolution of this pathogen. Together with other recent reports (11,12), our work suggests wider MTBC diversity, particularly among African mammals. Moreover, these data indicate that the theory that MTBC originally evolved as a human pathogen and jumped into animals is overly simplistic and may apply mainly to domestic animals. These data indicate 1 possible model could be that the common ancestor of MTBC was a generalist capable of infecting many mammals, including humans. From here, only few descendants spread around the world through human and animal migrations, creating the human-dominated phylogenetic picture we see today. We may well expect to find a much higher diversity of this MTBC, extending well outside the human-associated MTBC strains that infect various species, represented mainly through wildlife, including our closest relatives, the great apes. M. canettii and the other smooth TB bacilli show a high genetic diversity and are largely limited to the Horn of Africa, although whether these bacilli should be formally considered part of MTBC is controversial (3). Hence, they have been proposed to be part of the mycobacterial population that gave rise to the classical members of MTBC. Together with the apparent lack of human-to-human transmission of M. canettii (38), this suggestion would be consistent with a wider host range and/or environmental reservoir for the original ancestor of MTBC.

In conclusion, we report a microbiologically confirmed case of TB in a wild chimpanzee. Our molecular data show that the chimpanzee strain described here belongs to a novel lineage, more closely related to human-associated lineage 6 than to the other classical animal MTBC and possibly related to M. mungii and the Dassie Bacillus. This strain could represent a chimpanzee-specific pathogen or an MTBC variant acquired from another source. Because of our limited understanding of the ecology of this microbe, we propose at this stage to name it “Chimpanzee Bacillus” rather than to develop a dedicated species or subspecies name. Further studies are warranted, not only to better understand the natural history of TB in great apes and the biology of the Chimpanzee Bacillus, but also to estimate a possible risk for transmission of new types of MTBC to humans (e.g., through hunting and consumption of bushmeat). Moreover, further characterization of MTBC diversity will be crucial for understanding the origins of TB and the potential for the emergence of new strains through proximity between humans and wildlife.

Sequences of Mycobacterium tuberculosis complex isolates from previous studies compared with isolates from this study; genomic deletions respective to the H37Rv genome in the Chimpanzee Bacillus genome; coverage of the Chimpanzee Bacillus sequence reads in M. africanum–specific deletions; broth culture of M. tuberculosis complex isolate; maximum-likelihood phylogeny of M. tuberculosis complex isolate; observed pairwise distance between strains within lineages of M. tuberculosis; and PCA scatter plot of M. tuberculosis complex and related strains.