Detection of Brucella abortus and Brucellla melitensis in cattle and sheep from southern Cameroon

Background: Brucellosis is an infectious disease caused by bacterial of the genus Brucella. Although investigations have identied Brucella antibodies in many animal species, little attention has been paid towards specic identication of different Brucella species in animals in Sub-Saharan African countries. This study was designed to identify Brucella abortus and Brucellla melitensis in cattle and sheep of several villages of southern Cameroon in order to improve our epidemiological knowledge on brucellosis in central Africa Methods: Blood samples were collected from 597 cattle and 315 sheep from 15 villages of two regions of Cameroon. DNA was extracted from blood samples and primers amplifying the bcsp31 gene locus were used for the identication of Brucella spp infections in these animals. Amongst animals found with Brucella spp infections, specic primes were used to identify B. abortus and B. melitensis respectively. Results: Of the 912 animals analyzed in this study, 159 (17.4%) were infected with Brucella spp. This gives a Brucella infection rate of 20.9% in cattle and 10.8% in sheep. Of the 159 animals harboring Brucella spp infections, 119 (74.8%) were infected by B. abortus and 35 (22.0%) by B. melitensis. The overall infection rates were 18.4% (110/597) for B. abortus and 1.7% for B. melitensis in cattle; 2.9% for B. abortus and 7.9% for B. melitensis in sheep. Co-infections of B. abortus and B. melitensis were found in 9 (1%) animals; 6 (1%) cattle and 3 (1%) sheep. The Brucella infection rate was signicantly higher in animals from the Noun division (20.6%) compared to those of Yoko (12.4%). Between animal species, signicant differences (P = 0.0001) were observed in infection rates of both abortus and B. melitensis. Within and between sampling sites, signicant differences (P = 0.02) were observed in the B. abortus infection rates. Conclusion: This study revealed B. melitensis and B. abortus infections in cattle and sheep from the Noun and Yoko divisions in southern Cameroon. The identication of B. melitensis and B. abortus in animals suggests further investigations on human brucellosis. Results of this study highlight the need of developing and implementing control measures against brucellosis in Cameroon. This study revealed the presence of B. melitensis and B. abortus in cattle and sheep from the villages of the Yoko and Noun division of southern Cameroon. Results of this study have allowed for the identication of villages with high risk of Brucella infections and where control operations could be implemented. They highlight the need for developing control measures to ght Brucella infections. The identication of B. melitensis and B. abortus, which are zoonotic bacteria, suggests investigations on human brucellosis, especially in villages showing high risk for Brucella infections. Such investigations could help to design control measures for both human and animal brucellosis.


Introduction
Brucellosis is one of the most widespread zoonotic infectious diseases in the world [1]. It is amongst the top thirteen neglected zoonotic and contagious diseases [2]. Brucella infections are also considered as one of the major threat for livestock production in developing countries [3,4,5]. The economic losses resulting from Brucella infections in livestock can be estimated to about 427 million US$ per year in sub-Saharan African countries [6]. Livestock brucellosis has therefore detrimental socioeconomic effects in vulnerable low-income communities and the health of rural population [7].
Brucellosis is caused by bacteria of the genus Brucella which contains twelve species [8,9,10]. Amongst the classical Brucella spp, B. abortus and B. melitensis are of paramount zoonotic importance worldwide, with small ruminants and cattle serving as preferential hosts [11]. Human and several animal species including cattle, small ruminants, pigs, rodents and carnivores have been reported to be infected with Brucella [8,12]. In animals, Brucella infections induce abortion, infertility and decreases milk production, while in humans, they cause long debilitating illnesses [6]. Although human and various domestic animals can be infected by different Brucella species, little attention has been paid towards Brucella infections in developing countries due its neglected aspect.
In most sub-Saharan African countries, investigations undertaken on brucellosis have been mainly based on Brucella antibodies detection using various serological tests like the Rose Bengal Plate Test (RBPT), Enzyme Linked Immunosorbent assay (ELISA) and complement xation test [12,[13][14][15][16][17]. Although these serological tests have generated epidemiological data on the sero-prevalence of Brucella antibodies in humans and domestic animals from several sub-Saharan countries, these tests cannot differentiate different Brucella species [9,[13][14][15][16]. Results generated by these serological tests are of limited epidemiological values. For instance, the vertebrate host species of different Brucella species remain poorly characterized. Moreover, the distribution, transmission dynamics and the prevalence of different Brucella species remain under-investigated in most sub-Saharan countries.
Accurate identi cation of Brucella species in different hosts is of great importance for updating our epidemiological knowledge of brucellosis. Although culture and isolation of Brucella have been considered as the gold standard methods for brucellosis diagnosis [18], these methods are tedious, time consuming and di cult to implement in developing countries [19]. To overcome these challenges, molecular tools have been developed to identify different Brucella species in naturally infected hosts [20,21,22]. These molecular tools appear reliable, simple, easy to perform and able to identify natural infections of different Brucella species [12,[20][21][22][23][24][25]. Their use in West and East Africa has allowed for the identi cation and characterization of different Brucella species. In Zimbabwe, Uganda and Tanzania for instance, molecular tools have been used for the identi cation and the characterization of Brucella abortus and Brucella melitensis from cattle, sheep and pig [12,[24][25][26][27]. In central Africa, most studies on brucellosis have been focused on serological diagnosis of Brucella [14,15,[28][29][30]. Consequently, the Brucella species that infects humans and different animal species remain unknown. Species-speci c identi cation of Brucella species would undoubtedly shed light on the circulation and spread of different Brucella species. Furthermore, such investigations would allow for the establishment and the implementation of control strategies against brucellosis in each setting.
The present study was designed to identify Brucella abortus and Brucellla melitensis in cattle and sheep of several villages from the southern Cameroon in order to improve our epidemiological knowledge on brucellosis in central Africa.

Study zone
This cross-sectional study was conducted in nine villages (Foumban, Foumbot, Njimom, Massangam, Magba, Malentuen, Koutaba, Bangourain and Kouptamo) of the Noun division and six villages (Lena, Ngoun, Yoko Wankou, Megan, Kong and Kounde) of Yoko. The rst survey was performed in November 2017 at Yoko and the second from April to June 2018 in the Noun Division.
Yoko (5°31'60"N; 12° 18′ 57″E) is located in the "Mbam et Kim" Division of the center region of Cameroon ( Figure 1). It is situated between the southern green part and the northern Sahelian part of the country and covers about 15 000 km 2 . Yoko is known for large scale cattle and small ruminant's production within the center region of Cameroon. The presence of pastures and water favors transhumance of herds not only from the West, East and Adamawa regions of Cameroon, but also from neighboring countries like the Central Africa Republic [31,32].
The Noun division (4°95' 6°30'N; 10°30'12°E) is located in the western highlands of Cameroon. It covers about 7687 km 2 with a vegetation characterized by the savannah and degraded forest ( Figure 1). The Noun division is considered as the main cattle production area of the west region of Cameroon [30].
Animals in this locality move to other regions (Northwest, Center and Adamawa) of Cameroon for trade or transhumance.
In these two localities, the feeding system is essentially free grazing which is sometime associated to stall-feeding. Various animal species including goats, sheep and cattle share the same environment.

Ethics statement
The protocol of this study was approved by the Ministry of livestock, sheries and animals Industries of Cameroon with the reference number N°015/16/L/DDEPIA.NN. The review board of the molecular parasitology and entomology subunit of the Department of Biochemistry of the Faculty of Science of the University of Dschang gave also its approval. The local administration and traditional authorities of each sampling site were informed and their approvals were obtained. Verbal consent was obtained from each owner, after detailed explanation of the objective of the study.

Sample size estimation
For this study, a strati ed sampling strategy was applied to select herds and individual cattle per herd. The sample size was estimated for cattle using a standard formula for cross-sectional studies as described by Thrus eld [33].
See formula 1 in the supplementary les.
where i) N is the minimum sample size required; ii) Z (critical value for a given con dence interval) = 1.96; iii) P (expected prevalence) = 0.052; and iv) L (margin of error) = 0.05.
For this estimation, the bovine brucellosis prevalence of 5.2% previously reported in the North West Region of Cameroon was used [30]. Additionally, a 95% con dence level and a precision of 5% were also used to determine the sample size. In each herd and depending on its size, at least 20% of cattle were sampled. More than 20% of cattle of some herds were sampled for speci c reasons like the interests and cooperation of some farmers and advices from veterinarians. Selection of each cattle to be sampled in each chosen herd was done on the basis of systematic random sampling technique as described by Asgedom et al. [34]. A total of 37 farms were selected in 15 villages for a sample size of 597 cattle.
For the sheep, where there is no published data on the brucellosis prevalence in Cameroon, a prevalence of 50% was used for the sampling size estimation. Due to the small number of sheep in each village, all of them were sampled irrespective of the number of animals presented by each household. In consequence 315 sheep were sampled from the two agro-ecological zone of southern Cameroon.

Blood collection and DNA preparation
After approval from each herd owner, the farm characteristics and information regarding each animal including the name of the village (where each sample was collected), the geographical coordinates of each sampling site, the animal species found in farm (cattle, goat, sheep), the origin, sex, age, breed and the feeding system were recorded. Thereafter, about 5 ml of blood were collected, from the jugular vein of sheep and cattle, into EDTA coated tubes by a veterinarian. The tubes were labelled and carefully packed to avoid crossed contamination. In the eld, the blood samples were stored at 4°C in an electric cooler before being transported to the laboratory where they were kept at -20°c.
Genomic DNA was extracted from whole blood using the cetyl trimethylammonium bromide (CTAB) method adopted from Navajas et al. [35]. Brie y, frozen samples were thawed and 500µL of whole blood was pipetted into a micro-tube containing 1 mL of sterile water. The micro-tube was vigorously vortexed and then, centrifuged at 10.000 rpm for 5 min. To the resulting pellet, 500 µL of CTAB buffer (CTAB 2%; 1 M Tris, pH 8; 0.5 M EDTA pH 8; 1.4 M NaCl) was added. The pellet was re-suspended and incubated in a water bath at 60°C for 30 min. To the content of each micro-tube, 600 µL of chloroform/isoamyl alcohol (24/1) mixture was added. Each micro-tube was slowly homogenized for 15 min and the upper aqueous phase was removed and transferred to a new 1.5 ml micro-tube. DNA was precipitated with 600 µL of isopropanol. The mixture was gently homogenized for 5 min and then incubated overnight at -20° C. Thereafter, each micro-tube was centrifuged at 13,000 rpm for 15 min. DNA pellet was then washed twice with cold 70% ethanol and dried overnight at room temperature. The resulting DNA pellet was re-suspended in 50 µL of sterile nuclease free water and then stored at -20°C until use.

Detection of Brucella spp
The identi cation of bacteria of the genus Brucella was performed as described by Mitka etal. [23]. This was done using B4 (5'-TGGCTCGGTTGCCAATATCAA-3') and B5 (5'-CGCGCTTGCCTTTCAGGTCTG-3') primers that amplify a 223 bp DNA fragment corresponding to the bcsp31 gene locus of Brucella genus. PCR reactions were performed in a total volume of 25 µL containing 10 pmol of each primer, 2.5 µL of 10X PCR buffer, 2 mM MgCl 2 , 200 mM of each dNTP, 5 µL of DNA template and 0.5 Unit of Q5 high delity Taq polymerase (New England Biolab 5 U/µL). The ampli cation program comprised an initial denaturation step at 95°C for 5 min followed by 40 ampli cation cycles made up of a denaturation step at 95°C for 30 s, an annealing step at 60°C for 30 s and an extension step at 72°C for 45s. A nal extension step was performed at 72°C for 5 min.
The ampli ed products were resolved by electrophoresis on 1.5% agarose gel. The separation was done at 100 volts for 30 min. The gel was stained with ethidium bromide and DNA bands were visualized under ultraviolet light and photographed. Each PCR positive sample showing an amplicon of approximately 223 bp was considered as positive or having bacteria of the genus Brucella ( Figure 2). These samples were selected and subsequently subjected to the identi cation of different Brucella species.

Speci c identi cation of different Brucella species
All samples that were positive for the bcsp31 gene locus of Brucella genus (ampli cation of 223 bp DNA fragment) were further analyzed to determine the speci c infecting Brucella species. This identi cation was done using AMOS PCR as described by Bricker and Halling [36].  Results of Brucella antibodies detection and molecular identi cation of Brucella species were compared in the present study to nd the strength of agreement between serological and molecular tests (

Discussion
In most sub-Saharan African countries, investigations undertaken on brucellosis have generated epidemiological data that helped to understand the seroprevalence of Brucella antibodies in humans and animals. However, it is well-established that the dominance and overlapping nature of the C epitope of smooth Brucellae that is widely used in serological tests does not enable to ascertain the infecting Brucella species, [37]. To ll the gap linked to speci c identi cation of Brucella species, molecular tools have been used to determine the infection rates of B. abortus and B. melitensis in cattle and sheep of southern Cameroon.
Although the Brucella infection rate of 17.4% revealed by PCR-based tests is higher than its sero-prevalence (13.37%), the value of Kappa Cohen coe cient of 0.70 indicates a good strength of agreement between serological and molecular tests. Our results are in line with previous results; highlighting perfect concordance between these tests [22,38,39]. The animals found with Brucella DNA, but in which Brucella antibodies could not be detected may be carriers of early infections or with antibodies title bellow the detection threshold of the serological tests used in this study. However, we cannot rule out the fact some of these animals could be immune compromised and unable to seroconvert following exposure to Brucella infections. It is also important to point out the fact that serological tests could be negative if Brucella, which is an intracellular organism, is hidden in some tissues [38,39,40].
The high Brucella infection rates revealed by PCR-based test could be explained by the higher sensitivity of the molecular marker used in this study. Previous studies reported that the bcsp 31 gene ampli ed here for the detection of bacteria of the genus Brucella has higher sensibility and speci city compared to other markers like omp 2, Omp28, 16sRNA [41,42]. Indeed, PCR targeting bcsp 31 gene can detect very small amount of Brucella DNA that was extracted from blood [39,23,43]. It appears therefore as an appropriate molecular tool for the detection of Brucella infections in blood samples.
Our results showing high Brucella infection rate in cattle (20.93%) compared to sheep (10.79%) is consistent with results reported in South Africa [26]. The difference observed could be explained by the larger herd sizes of cattle and the management system [4,44]. The extensive cattle management system and the larger size of herds have been reported as risk factors for brucellosis [45]. Moreover, the relative short life expectancy of sheep compared to cattle is additional factor explaining the low Brucella infection rate in sheep. In fact, cattle remain in the herd for many years and therefore, are more exposed to Brucella infections than sheep. This hypothesis is strengthened by observations reporting that as long as infected animals remain in contact with the rest of the herd, the number of infected animals will increase [13,46,47,48].
The identi cation of B. abortus and B. melitensis in both cattle and sheep is consistent with results obtained in the South and East Africa [12,24,26,38].  [12,20,25,50]. In fact, B. abortus has been reported as the most commonly Brucella species that has been isolated and characterized in cattle from sub-Saharan countries [12,16,50]. The high infection rate of B. melitensis in sheep is consistent with results of most sub-Saharan countries [16]. These results indicate that the epidemiological patterns of some Brucella species could be similar in sub-Saharan countries. Compared to B. abortus, the low B. melitensis infection rate reported in domestic animals suggests its low transmission in sub Saharan countries [51]. The small size of sheep' herd is another factor playing for the low transmission of B. melitensis and consequently, its low infection rate in domestic animals [14,30].
The identi cation of B. abortus in sheep and B. melitensis in cattle is not common because these animals are not their preferential hosts. Nevertheless, it is important to point out that several B. abortus isolates have been obtained from milk and abortion products of sheep and goat [52,53]; highlighting the probability of sheep to be infected by B. abortus. In addition, other animal species like sheep has been reported to be susceptible to B. abortus infections [54,55]. The detection of B. melitensis in cattle is in agreement with previous observations reporting the transmission of B. melitensis to cattle [56,57]. The cross-transmission of B. abortus to sheep and B. melitensis to cattle could be explained by the fact that these animals share the same environment. In such context, the transmission of B. melitensis to cattle and B. abortus to sheep can occur. This hypothesis is strengthened by previous observations reporting the transmission of B. melitensis to cattle, especially where both cattle and small ruminants share the same breeding sites [56,57]. Indeed, inter and intra-speci c transmissions of different Brucella species can occur in larger herds and livestock production system where cattle, goat and sheep are put together [4,45,29]. Animals found with Brucella infections, but for which B. abortus and B. melitensis could not be identi ed may be carriers of bacteria load below the sensitivity threshold of AMOS-PCR. Some of these animals could be infected by other Brucella species like B. ovis which is a nonzoonotic bacteria species that is restricted to sheep. Although not investigated in the present study, this Brucella species has been already reported in sub-Saharan countries [58,59,60].
The signi cant difference in the Brucella infection rates reported between villages of the Noun division; but not at Yoko highlights some variations occurring within and between sampling sites. This difference results probably from environmental factors and/or livestock management system that vary between villages and that have impacts on the transmission pattern of Brucella species. Within and between sampling sites, our results showing no signi cant difference (P =0.6; X 2 =0.26) in the infection rates of B. melitensis could be explained by the low size of herds and the limited movement of sheep. In such context, the transmission of B. melitensis from infected to uninfected sheep is limited because animals of different herds cannot mix together and, most of time, sheep are regularly sale and killed for different ceremonies. For B. abortus where the infection rates vary signi cantly (P =0.02; X 2 =5.96) between villages of the two localities, it is likely that high transmission of B. abortus occur in some villages like Kong in Yoko, Foumban, Magba, Bangourain and Koutaba in the Noun division. These villages can be considered as hot spots where particular attention must be paid during the implementation of control measures. They are also of great epidemiological interest for further investigations on human brucellosis. The reasons explaining the high infection rate, and probably the high transmission of B. abortus in these villages are not well understood. However, we can speculate about: i) the large size of herds in these villages, ii) the environmental factors that could be more favorable for brucellosis transmission, iii) the poor cattle management systems, iv) the regular mixing of animals from different herds. Most of these factors have been recognized as important contributors for the transmission and the spread of brucellosis in livestock [29,31,61]. Viewed the high prevalence of B. abortus recorded in these villages and the zoonotic nature of this bacterium, measures should be taken to prevent the transmission of B. abortus to humans.
Results of this study highlight the need to raise awareness for brucellosis control and to design control measures that could be implemented to ght this neglected zoonotic disease. While waiting for investigations that will enable to identify risk factors linked to brucellosis transmission in villages showing high Brucella infection rates, the feeding and management systems must be improved; the sharing of common grazing areas as well as the mixing of animals of different herds must be limited because these factors have been recognized elsewhere as risk factors for brucellosis transmission [62,63,64].
The main limitations of this study rely on the fact that no Brucella species was isolated and characterized and speci c identi cation of Brucella species was based on band size. Consequently, the genetic structure of Brucella circulating within and between the two localities remains unknown. Moreover, all Brucella species for which cattle and sheep are susceptible have not been investigated and consequently, other Brucella species that could be found in animals of Yoko and the Noun division also remain unknown. With these limitations, the transmission dynamics within and between villages, and also between different animal species remains not well understood for e cient planning of control operations against brucellosis.

Consent for publication
Not applicable Availability of data and materials All data generated and/or analyzed are included in this article.

Competing interests
The authors declare that they have no competing interests.

Funding
This project received no speci c grant from any funding agency in the public, commercial, or not-for-pro t sectors. The Molecular Parasitology and Entomology subunit of the Faculty of Science University of Dschang of Cameroon provided reagents required for lab experiments that scheduled in this project.
Authors' contributions KNRM helped in the sample collection, identi cation of Brucella species and the drafting of the manuscript. SAB participated in the collection of samples, the study design and the drafting of the manuscript. FO helps in the collection of samples. GS participated in the conception, data collection and the drafting of the manuscript. All authors read and approved the nal version of the manuscript.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.