Tuberculosis is still a neglected disease in animals in the Sub- Saharan countries, where it is less studied [1]. M. bovis is the principal agent of animal tuberculosis and its transmission to humans constitutes a public health problem as this species is naturally resistant to pyrazinamid, one of the main anti-tuberculosis drugs [2]. In Africa, animal tuberculosis remains endemic and it causes significant losses in the animal industry, with implications for food safety and trade [3].

In Cameroon the breeding sector provides income of about 30% to rural population [4,5]. Sheep represents the second important livestock in Cameroon after cattle though it’s a great source of animal protein for the general population. Despite his usefulness, many reports of the Ministry of Livestock, Fisheries and Animal Industries and scarce research based only in cattle livestock, showed that animal tuberculosis is endemic and it is one of the main causes of seizures in abattoir [1,6-9]. Only few regions (two) in Cameroon have data on animal tuberculosis and these were essentially on bovine tuberculosis (tuberculosis in cattle). Moreover, all the ten regions are concerned by bovine tuberculosis.

It is now known that bovine tuberculosis like lesions are not only caused by a member Mycobacterium tuberculosis complex (MTBC) but also by Non Tuberculosis Mycobacteria (NTM) or atypical mycobacteria which in human need different treatment from that of patients infected by M. tuberculosis complex strains [10]. Accurate identification of Mycobacterium tuberculosis and NTP is thus, essential for TB control. We used in this study a new identification algorithm proposed by the International Union against Tuberculosis and Lung Disease (IUATLD), consisting by the use of p-nitrobenzoic acid (PNB) susceptibility testing permitting the differentiation between M. tuberculosis and NTM [11].

The aim of this study is to evaluate for the first time the prevalence of animal tuberculosis and the contribution of NTM and MTBC in sheep livestock in Cameroon. This will permit to gain more information which may significantly impact TB control in Cameroon.

Slaughterhouse sampling

The sheep’s slaughterhouse of central market in New-Bell district of Douala subdivision 2(4°01’26.3’’N; 9°42’53.4’’E) was chosen to screen slaughtered sheep. We made the choice due to its geographical situation (it is located in one of the populous markets of the town of Douala), the diverse origin of the animals coming from almost all the breeding regions in Cameroon, which are slaughtered there and by the existence of a system of notification of all the TB cases.

Sampling

Sampling for tuberculosis lesions occurred during routine inspection by veterinaries surgeon from the Ministry of Livestock, Fisheries and Animal Industries (MINEPIA) and us. This inspection was done according to the government’s legislation regulating veterinary health inspection and notification of contagious animal diseases and the recommendation of the World Organization for Animal Health (OIE) [12,13]. After obtaining consent from the relevant authorities, sampling was done from November 2013 to April 2014. About 2922 sheep were successively inspected for tuberculosis during this study. Lesion identification was done by postmortem examination of carcasses as earlier described [14]. This consisted of visual examination and palpation of lungs, liver, bowel, spleen, heart, and inside of the carcass of each slaughtered sheep for the research of caseous nodules, incision of thoracic, mesenteric lymph nodes and other nodes of the carcass for the research of tuberculous granulomas with caseous center.

Lesion collection

A sample was taken according to the World Organization for Animal Health recommendation [13], from any carcass presenting suspect lesions of tuberculosis or a fragment of lymph node presenting a caseous granuloma with or without purulence. Samples were transported dry in sterile plastic containers under controlled conditions and conveyed immediately to the laboratory of the Tuberculosis Research (LTR) of the Biotechnology Center (BTC) of the University of Yaounde I, where they were preserved at −20˚C. Information relating to each animal (type of lesion, sex, animal age and origin) was recorded postmortem.

Processing of samples

The processing of lesions samples was based on grinding and decontamination procedure using sodium lauryl sulfate as described in our previous study [7,15]. Briefly, the samples were first carved on sterile Petri dish using a sterile scalpel, then ground in a mortar by using a pestle before addition of sterile distilled water. Using a sterile 50 ml centrifuge tube with a screw cap, equal amounts of specimen and sodium lauryl sulfate were added. The centrifuge tube was capped and its content was vortexed until the specimen was liquefied. The mixture was allowed to stand at room temperature for 45 minutes with permanent gentle shaking by Khan Shaker. Prepared phosphate buffer was added to the mixture in the centrifuge tube and mixed, then centrifuged for 20 minutes at 3000 g. The supernatant was carefully decanted and 2 ml of sterile distilled water was added to re-suspend the sediment. The suspension was inoculated onto Lowenstein-Jensen (LJ) slopes with pyruvate and/or glycerol, para-nitro-benzoic acid and incubated at 37˚C for 8 to 12 weeks. All smears were stained by the conventional Ziehl-Neelsen method for the identification of acid-fast bacilli (AFB) and observed under a light microscope [12].

Molecular analysis

The GenoType® Mycobacterium CM/AS was used to identify mycobacteria from any positive culture media, according to the manufacturer’s instructions. The results were interpreted using a template sheet and interpretation charts supplied by the manufacturer. Standard spoligotyping was done as described by Kamerbeek and colleagues [16], using a commercially available kit (Isogen Bioscience BV, Maarssen, The Netherlands) for identification and typing of MTBC strains of positives culture on LJ and LJ+P media.

Data analysis

The spoligotypes patterns were introduced to the SITVIT_WEB database and assigned to a SIT number. Spoligotyping-based phylogeographic clades were assigned according to signatures provided by Brudey and collaborators [17]. The Pearson Chi-square or Fisher’s exact test was used to estimate the association between the sex, the infected organs, and the result of Ziehl-Neelsen, using statistical software R version 2.15.3 (www.r-project.org).

Postmortem inspection findings

Examination of the 2922 sheep carcasses, presented 810 sheep with the typical macroscopic lesions of tuberculosis, mainly characterized by caseous tubercles with a yellowish appearance (Figures 1a-1f). The organs or tissues from which tuberculosis lesion were taken were liver (55.55%), bowel (27.40%), lung (8.88%), lymph nodes (6.66%), muscle and spleen (0.74%). Among these, liver appeared to be the most frequently infected organ (χ²=67.02, p<0.0001).

Figure 1: Example of some tuberculosis lesions identified on sheep slaughtered in the abattoir of New-Bell (Cameroon). a. Caseous lesion on sheep lung; b. Hepatic lesions caseo-limestones generalized; c. Caseous nodular lesions on the sheep body in the milliary tuberculosis in abattoir of New-Bell (Cameroon); d. Spleen caseous lesion; e. Caseous Nodules in intestinal fabric; f. Greenish discharge mesenteric lymph nodes.

The apparent prevalence of tuberculosis-compatible lesions on the basis of macroscopic observation was 27.72% (810/2922). Female sheep presented more significant (χ²=52.06; p<0.0001) tuberculosis like lesion than male with 38.9% (768/1974) and 4.45% (42/948) respectively.

Ziehl-Neelsen stain findings

Acid Fast Bacilli (AFB), using Ziehl-Neelsen stained method, were detected in 630 (77.77%) of the 810 cases (Table 1). This corresponded to the apparent tuberculosis prevalence of 21.56%. There was no statistical association between the tissue lesions origin and Ziehl- Neelsen staining result in this study (p=0.6358).

Table 1: Macroscopic and microscopic identification of tuberculosis according to sheep organ.

Culture findings

From the 810 detected cases only 123 were positive in culture.

Non tuberculosis mycobacteria identification

Among the culture positive cases, 87 were positive in PNB indicating that 10.74% of cases were due to NTM. The used of the GenoType® Mycobacterium CM assay allows us to identify Mycobacterium interjectum , Mycobacterium fortuitum and Mycobacterium sp. (Figure 2).

Figure 2: GenoType^™ Mycobacterium CM profile of Non Tuberculosis Mycobacteria identified on slaughtered sheep in Cameroon.

Mycobacterium tuberculosis complex strain identification

Only 36 (4.44%) cases were positive in LJ and LJ supplemented with 0.4% of Pyruvate. Spoligotyping showed that all these cases were due to M. tuberculosis . No M. bovis nor M. caprea were identified. The use of SITVITWEB permitted to identify all these M. tuberculosis strains as belonging to the Ubiquitous T lineage called SIT53 (Table 2).

Table 2: Spoligotype profile of M. tuberculosis identified in Slaughtered sheep in Cameroon.

The detection of the lesions during a meat inspection of slaughtered animals is the only way for identifying animal tuberculosis in Cameroon. Many studies have speculated that sheep are resistant to tuberculosis [18-20], but our study shows a very high prevalence of TB-like lesions and very high acid fast bacilli prevalence in slaughtered sheep. This prevalence is very high compared to what is observed in other countries even by the use of comparative intradermal tuberculin skin test or ELISA test. Among these countries was Algeria (4.4%) [21], Ethiopia, 1.4% [22], Bangladesh, 9.15% [23]. However in Spain a study obtained a more high prevalence (59.4%) [18].

The use of PNB and GenoType® Mycobacterium CM/AS permitted to identify NTM in 10.74% of sheep presenting gross TB-like lesion. This result can justify the high prevalence of AFB observed but remain difficult to understand since cattle and local sheep live in closely proximity but many TB studies in cattle in Cameroon did not found any atypical mycobacteria [7,8]. An explanation could be the fact that many sheep sold and slaughtered in Douala is coming from Chad passing through Far North. In Chad, some study has showed that 38.3% cattle TB-like lesion, are due to Non tuberculosis mycobacteria (atypical mycobacteria) [24], and knowing that in the countries of Central Africa, cattle and sheep live in close contact, we think therefore that Chadian sheep and cattle might share the common source of TB contamination and when this sheep arrive in Douala it is directly sold for human consumption and is very rarely mixed with Cameroonian cattle. This assertion is supported by the fact that M. fortitum which have been identified mainly in sheep slaughtered in Cameroonian had been described as dominant in Chadian slaughtered cattle sheep. This assertion needs however to be verified with a large transversal study in Chadian and local Cameroonian since one study done in many regions of Cameroon, have identified M. furtitum in Cattle slaughtered [25].

This result can be of great interest in public health and can be disturbing since M. fortitum may cause infection in patients with or without HIV infection [26]. Moreover Non Tuberculous Mycobacteria (NTM) can be associated to increasing virulence outbreaks and emergence of antibiotic resistance [27], it may also lead to ill-treatment of the patients to whom a non-adverse treatment would be administered.

The female sheep was significantly more associated to tuberculosis like lesions than male in our study. This result is in line with that obtained in Algeria [21]. This could be explained by the fact that the females are maintained in the herd longer in order to maintain the reproductive performance of the herd and therefore increasing their chance to be exposed to the TB infection [28]. Our result is nevertheless contrary to that obtained by Tschopp et al., in Ethiopia, who showed that the males like the females are more at the risk to develop tuberculosis [29].

An interesting result of this study was the identification of M. tuberculosis in Sheep TB like lesion, only represented by it ubiquitous T family clad SIT 53. Identification of M. tuberculosis in animal has been also described [8,29-31]. The M. tuberculosis lineage SIT 53 has been more descript in human TB in Cameroon but never in cattle, goat, pig or sheep [8,32-35]. Knowing that human remain the true host for M. tuberculosis, the finding of M. tuberculosis in sheep may traduce an indirect zoonosis case as it has been suggested for cattle [8]. This result shows the urge to make a large epidemiological study to evaluate the degree of implication of M. tuberculosis in animal tuberculosis in Cameroon. This may significantly impact TB control in Cameroon.

The absence of M. bovis and M. caprea among slaughtered sheep in Cameroon can be intriguing but it have longer known that tuberculosis caused by M. bovis is rare in sheep, mainly because of the nature of sheep husbandry and the fact that sheep are rarely exposed to infectious material [36]. Moreover others studies also did not identified M. bovis and M. caprea in sheeps [37]. Nevertheless, the fact that 630 specimens were Ziehl-Neelsen stain positive but that only 123 were positive in culture may also explain the absence of M. bovis or M. caprea because of the difficulty of recovery of these MTBC species by culture. In fact, difficulties to recover M. bovis due to its cultural habit have been longer highlight [38].

This study is the first study on molecular epidemiology of tuberculosis in sheep in Cameroon. It shows a high prevalence of TB in sheep due principally by M. tuberculosis and the implication of atypical mycobacteria represented mostly by M. fortitum and M. interjectum in the sheep TB like lesions. This highlight the urge to make a large study in Cameroonian livestock in order to identify the risk factors associated to M. tuberculosis transmission in animal in Cameroon.

R Ateugieu Guemechieu and AT Onana are Ph.D. students of the University of Douala. This project did not receive external grant funding. The cost for all aspect of this research was covered by the funds from the research grant granted by the Cameroonian government to researchers of the state universities.

FX Etoa, RA Ngono Ngane, and F Koro Koro conceived and designed the experiments. R Ateugieu Guemechieu and AT Onana, EM Tchamba Kombou and Y Kamdem Simo performed the experiments. F Koro Koro, R Ateugieu Guemechieu and AT Onana analysed the data. F Koro Koro wrote the first draft of the paper. R Ateugieu Guemechieu designed figures. All authors provided critical input.

I would like to undertake the responsibility for this submitted manuscript, I did not receive reimbursements, fees, funding, or salary from an organization that may in any way gain or lose financially from the publication of this manuscript, either now or in the future. I did not hold any stocks or shares in an organization that may in any way gain or lose financially from the publication of this manuscript, either now or in the future. Moreover, I did not hold or was currently applying for any patents relating to the content of the manuscript even received reimbursements, fees, funding, or salary from an organization that holds or had applied for patents relating to the content of the manuscript. In the best of my knowledge I did not have any other financial and non-financial competing interests.

We could not complete this work without any time to thank the head for the zootechnical and veterinary center as well as the veterinary agents of the slaughter-house of the sheep of the central market of Douala (Cameroon) for their collaboration. We thank the head of the research laboratory in tuberculosis of the center of biotechnology of the University of Yaoundé I as well as the Mycobacteria National Reference Laboratory of Centre Pasteur of Cameroon for their logistical support.