Molecular detection and characterization of Brucella species in raw informally marketed milk from Uganda

Abstract

This study identified and characterized Brucella species in the informal milk chain in Uganda. A total of 324 cattle bulk milk samples were screened for the genus Brucella by real-time PCR with primers targeting the bcsp31 gene and further characterized by the omp25 gene. Of the samples tested, 6.5% were positive for Brucella species. In the omp25 phylogeny, the study sequences were found to form a separate clade within the branch containing B. abortus sequences. The study shows that informally marketed cattle milk in Uganda is a likely risk factor for human brucellosis and confirms that B. abortus is present in the cattle population. This information is important for potential future control measures, such as vaccination of cattle.

• Africa
• brucellosis
• bulk milk
• milk delivery chain
• PCR
• bcsp31
• omp25

Brucellosis is a widespread but neglected bacterial zoonosis of global importance (1). Besides being a public health threat, brucellosis also has a serious global economic impact as it causes severe production losses due to infertility, abortions, and reduced milk production in goats, sheep, cattle, and swine (2).

The bacterial genus Brucella consists of 11 species, of which some have been further subtyped into biotypes or biovars (3). The genomes of Brucella species (spp.) have a similarity at the nucleotide level exceeding 90% (4). Brucella melitensis, B. abortus, some B. suis biovars and B. canis are most commonly reported in domestic animals and are also known to be zoonotic (5). Even if Brucella exhibits host species preference, cross-infections to other animal species may occur (6). The major modes of transmission of Brucella infection to humans are through contact with infected animals, fetal membranes, aborted fetuses, and consumption of unpasteurized dairy products (7).

Cattle brucellosis is endemic in Uganda and human brucellosis is an important disease in the country (8). It has been shown that the Brucella seroprevalence at dairy herd level ranged between 28 and 44% (9–11) and that antibodies against Brucella spp. were present in 11 and 40% of samples of bulk milk (12). It has also been shown that B. abortus is infecting dairy cattle (9), but there is a lack of knowledge about presence of Brucella in the informal milk delivery chain in Uganda, here defined as small-scale dairy farmers selling fresh milk without uniform processing (12).

The aim of this study was to identify and characterize Brucella spp. in the final step of the informal milk delivery chain in Uganda, through molecular techniques.

Materials and methods

Study design and sample collection

The Gulu and Soroti Districts were included as they are two rapidly growing urban areas located in Northern and Eastern Uganda. Small-scale livestock holding is an important source of livelihood in these areas. In 2011 and 2012, 324 bovine bulk milk samples were collected from the two districts, see Rock et al. (12). In brief, the samples were collected directly from the containers of informal milk sellers and milk deliverers at the roadside, at milk-collecting centers, and at boiling points. Ethical clearance was obtained as described in Rock et al. (12).

Bacterial reference strains

DNA from the vaccine strains B. melitensis Rev. 1, B. abortus RB51, and B. suis in the commercial INgene Bruce-ladder V kit (Ingenasa, Madrid, Spain) was used as positive controls in all Brucella PCR-assays. In the 16S rRNA real-time PCR assay, the positive controls consisted of DNA from the bacterial strains Pseudomonas aeruginosa B683 and Treponema T2378.

Genomic DNA extraction and real-time PCR detection

Genomic DNA was extracted using a phase separation technique with phenol:chloroform:isoamyl alcohol (Sigma-Aldrich, St. Louis, MO, USA), recommended by the Animal Health and Veterinary Laboratories Agency Brucella research division in Weybridge, UK. The quantities and purities of the extracted DNA from all samples were determined by optical density measurement using a NanoDrop 2000c Spectrophotometer (Thermo Scientific, Wilmington, DE, USA). DNA extracts were stored at −20°C and were analyzed in 2014 for the genus Brucella as described by Probert et al. (13) with the modification that the assay was run as a singleplex with primers and probe targeting the bcsp31 gene. Extracts were randomly selected for analysis in order to prevent cross-contamination during handling. Two negative controls consisting of DEPC water (Invitrogen, Thermo Fisher Scientific, Stockholm, Sweden) were included in each run to detect PCR contamination. The limit of detection was determined to a DNA concentration of 3.6 ng µL⁻¹ extract. Samples with a cycle threshold (Ct) value of ≤40 were interpreted as positive.

Molecular typing of bulk milk samples by omp25

Brucella spp. were characterized in five strong positive extracts, four from Gulu and one from Soroti, using the omp25 gene (13). The limited number of samples characterized was due to limited amount of DNA extracts. The expected size of the omp25 amplicons was 523 base pairs (bp). Weak positive samples in the bcsp31 assay gave weak band in conventional PCR and were not enough for sequencing.

DNA sequencing and sequence analyses

Purified PCR products were sent to Macrogen Europe (Amsterdam, the Netherlands) for Sanger sequencing. Purification was performed with ExoSAP-IT (Affymetrix, USB, Santa Clara, CA, USA), according to the manufacturer's instructions. Sequencing primers for the omp25 amplicons were the same as for PCR.

Sequences were edited and assembled with the CLC Main Workbench 7 (CLC Bio-Qiagen, Aarhus, Denmark). Contigs and individual sequences were blasted (www.ncbi.nlm.nih.gov/BLAST/) and aligned in the MEGA6 software, using the MUSCLE algorithm. The total length of the alignment, excluding non-overlapping sequences, was 455 nucleotides. Phylogenetic trees were generated in MEGA6 and Ochrobactrum anthropi was used as outgroup to root the trees (14). The corresponding sequence of the O. anthropi omp25 gene was identified from the whole genome entry of O. anthropi (Accession number CP000758). Phylogenetic relationships were inferred using the neighbor-joining (NJ) and maximum-likelihood (ML) algorithms. Bootstrapping of the NJ method data based on 1,000 replicates assessed the resulting tree topology. All sequences in this study have been deposited in GenBank under the accession numbers KY038989-KY038992.

Results

Detection of Brucella spp. DNA in cattle bulk milk by real-time PCR

To investigate the presence of inhibitors and the effectiveness of the extraction, all samples were analyzed by a 16S real-time PCR. Bacterial DNA was present in the majority of the DNA extracts with similar fluorescence signal, indicating extraction success.

It was shown that Brucella spp. DNA was present in 6.5% (21/324; 95% confidence interval (CI) 3.8–9.2) of informally marketed raw cattle milk collected from street sellers and milk deliverers in two districts in Uganda. Of those were 5.3% (10/188; 95% CI 2.1–8.5) from Gulu and 8.1% (11/136; 95% CI 3.5–12.7) from Soroti.

Characterization of Brucella spp. in cattle bulk milk

Four (three from Gulu and one from Soroti) out of five extracts were successfully sequenced to ascertain Brucella spp. identification. Comparison of omp25 sequences revealed only single nucleotide substitutions and indicated a similarity of more than 98% between study and reference sequences of closely related Brucella spp. Study sequences were 100% identical to each other and exhibited the highest level of sequence similarity to B. abortus (99.6% nucleotide similarity, 453/455 bp).

Phylogenetic analysis of omp25 sequences indicated that B. abortus was present in the marketed raw cattle milk from Gulu and Soroti Districts. The sequences from the cattle bulk milk were found to form a separate clade within the branch containing B. abortus sequences. The NJ and ML trees exhibited similar topology and were in overall agreement with current Brucella taxonomy – separating three out of the four closely related classical Brucella spp. into separate sub-branches; B. melitensis, B. abortus, and B. canis. A B. abortus–B. melitensis clade appeared in both analyses.

Discussion

This study showed that B. abortus is the probable species found in bulk milk aimed for human consumption and complements previous findings by Mugizi et al. (9) and Rock et al. (12). Identifying infection species is important if control measures, such as vaccination, are to be implemented in the future. This study also indicates that informally marketed raw cattle milk is a probable risk factor for human brucellosis if consumed raw, even if presence of Brucella DNA does not give information of the presence of viable Brucella bacteria. The study also showed that proportionally more samples from Soroti contained Brucella DNA compared with samples from Gulu. The same relation with respect to antibodies against Brucella was shown by Rock et al. (12).

The real-time PCR results indicated low Brucella bacterial DNA load in the bulk milk extracts, which was expected since few Brucella bacteria in general are excreted in cattle milk (4) and pooling of milk from different sources might have a dilution effect. A greater analytical sensitivity has been observed when using the insertion element IS711 as target, in real-time PCR assays for detection of Brucella at the genus level (15). In this study, IS711 was not targeted due to limited amount of extracts.

All Brucella spp. have a genome similarity of more than 90%. Effective genetic markers for distinguishing closely related Brucella spp. and their biovars from each other are therefore difficult to identify. In this study, the Brucella genus–specific omp25 gene was used as a genetic marker. The similarity of the omp25 sequences proved to be more than 98%, clearly indicating that the sequences belong to the genus Brucella, and further confirming the presence of Brucella spp. In the omp25 phylogeny, study sequences were found to group with B. abortus sequences. Additionally, a close relationship was found between B. abortus and B. melitensis. This agrees with other analyses across multiple gene markers (16, 17).

This study confirms previous findings that Brucella abortus is the species infecting cattle in Uganda. This information is important if control options at the production level in the milk chain, such as vaccination of cattle against B. abortus, would be discussed.

Conflict of interest and funding

The authors have no conflict of interest to declare but acknowledge the SLU Global Food Security Research and Education Program 2010–2013 (UD40), the Ministry for Foreign Affairs, Sweden, for the financial support.

Acknowledgements

The authors wish to acknowledge Dr. Krishna Gopaul at the Animal Health and Veterinary Laboratories Agency (AHVLA) Brucella research division, Weybridge, UK, for assistance with extraction of bacterial DNA; Dr. Michelle Wille at the Zoonosis Science Center, Uppsala University, Sweden, for assisting during the phylogenetic analyses; and Dr. Anna Rosander at the Swedish University of Agricultural Sciences (SLU), Uppsala, Sweden, for kindly providing the bacterial strains Pseudomonas aeruginosa B683 and Treponema T2378. Additionally, the authors wish to express their gratitude to the milk sellers and milk deliverers for their generous cooperation.