Polyphasic Characterization of Brucella spp. in Livestock Slaughtered from Abattoirs in Eastern Cape, South Africa

In livestock, brucellosis is mainly an asymptomatic disease except when abortion occurs; therefore, two serological tests are used for diagnosis as no single test is suitable. Abattoir samples enable a combination of culture, molecular, and serological tests to detect brucellosis. This study assessed Brucella-specific PCR (ITS-PCR) to detect brucellosis and to conduct a molecular characterization of Brucella spp. isolated from PCR-positive livestock (n = 565) slaughtered at abattoirs and the appropriate sample tissue(s). ITS-PCR detected Brucella DNA in 33.6% of cattle, 14.5% of sheep, and 4.7% of pig tissues. Impure Brucella cultures from PCR-positive tissues were 43.6% (44/94) of cattle, 51.7% (15/29) of sheep, and 50% (2/4) of pigs with predominantly B. abortus identification with AMOS-PCR and low isolation of mixed B. abortus and B. melitensis in all species. In cattle, 33% of isolates were from lymph nodes, while in sheep 38.0% were from the liver and kidney and only from tonsils in pigs (2/4). Brucella infections identified with AMOS-PCR were present in seropositive and mainly seronegative (75.6–100%) livestock with the potential to cause brucellosis during pregnancy or breeding. This study demonstrated the value of the polyphasic approach, especially with chronic infections and the potential risk of these asymptomatic animals.


Introduction
Brucellosis is a highly contagious zoonotic infection of humans, domestic, and marine animals [1].The Brucella species are facultative intracellular Gram-negative, non-spore forming, cocco-bacilli bacteria [2][3][4] causing a disease called brucellosis.In animals, Brucella invades the host without any clinical symptoms, resulting and progressing to acute infection only when the bacteria replicate actively within the macrophages and other replication sites [4].The infection in animals remains asymptomatic in most cases, or until the first pathological signs/symptoms appear [4].Chronic infections occur when the bacterial load decreases after plateauing, with sporadic clinical symptoms when the infection localizes in the reproductive system of sexually mature animals, resulting in sterility in males and placentitis as well as abortion in females during pregnancies.It continues to spread amongst animals in the herd [5][6][7][8].Infection is described as mostly self-limiting [8] due to low activation of phagocytosis and other host defences related to innate immunity [9,10].After the initial phase of the illness has passed, the majority of brucellosis symptoms are not pathognomonic, and the organism can persist over time in the mammary glands and supramammary lymphatic nodes of 80% of infected animals [11].Brucella replicates extensively in the endoplasmic reticulum (ER) compartment within the host cells [12].The host cells' specialized compartment where intracellular pathogens reside prevents antibiotics from reaching them, thus affecting the efficacy of current therapies [12].The association of Brucella within the host cell ER provides optimal bacterial growth conditions and replications in organs such as the spleen, lymph nodes, liver, bone marrow, epididymis, and placenta, which is rich in reticuloendothelial cells [8].In the chicken embryo model, the replication of Brucella abortus spreads to all tissues, with the liver and spleen being the most severely infected [13].
Gonzalez-Espinozo et al. [8] reviewed investigations to improve cultures other than blood, such as bone marrow aspirates, liver, and lymph nodes, based on the rationale to obtain specimens from macrophage-rich tissues where Brucella organisms multiply and concentrate as this may increase bacterial recovery.Culture from these tissues remains far from resolved due to its non-specific signs and symptoms that are comparable to other febrile diseases, its slow growth rate on culture, and the complexity of its sero-detection, and thus brucellosis remains difficult to diagnose [14,15].The sensitivity of culture depends on the disease stage, Brucella spp.culture medium and the technique being used, the quality of circulating bacteria, and the number of contaminants present in the sample [16].The skin, hair, limbs, blood, stomach, gut contents, bile, and other excretions of the animal, as well as the facilities, can all contaminate the sample taken from the carcasses of the animals throughout the slaughter process in the abattoirs [17].It is crucial to minimize any surface contamination occurring in the abattoirs during the slaughtering process, using a hazard analysis critical control point (HACCP) plan, in order to effectively handle and regulate the microbiological hazards connected with meat products [18].
Several molecular and high-resolution phenotypic assays that allow the differentiation of Brucella spp., the biovars, and the traceability of the source have been published [19,20].However, only the gold standard (culture) is capable of reliably diagnosing brucellosis [21,22].Brucella abortus and B. melitensis isolates take up to 4-7 days for growth in the laboratory; however, an increased number of contaminants including fungi and bacteria are usually observed, resulting in the regular contamination of culture plates and the decreased sensitivity of bacteriological diagnosis [23].The most common sample for brucellosis diagnosis is serum for serological tests, which is not an appropriate sample for culture.Investigation at abattoirs where various sample types can be collected offers a polyphasic approach.Rose Bengal test (RBT) is used as a serological screening test with high sensitivity and low specificity resulting in false positives.RBT positive samplescan be confirmed through a compliment fixation est (CFT) and/or indirect enzyme-linked immunosorbent assay (iELISA), which both have high specificity, but the CFT is less sensitive than iELISA, resulting in false-negative results [24].Modern molecular approaches are currently not widely implemented in low-income nations where brucellosis is endemic in livestock [14,25].This may be due to resource constraints.However, Brucella genus-specific PCR assays (conventional and real-time) such as 16 The Brucella genus currently consists of twelve species of which four species are pathogenic to humans [27].Brucella melitensis and B. abortus commonly infect small ruminants and cattle, respectively, leading to abortions and infertility thus resulting in significant economic losses [28].Five Brucella spp.have been discovered in wildlife and marine mammals, while four additional distinct strains have been discovered in rodents, frogs, baboons, and humans [29].The most pathogenic species for human brucellosis is B. melitensis, followed by B. suis, and then B. abortus [29].Brucella ceti, B. inopinata, and B. canis (rarely) are also known to cause human brucellosis [29].
Brucellosis-infected animals are the primary cause of human brucellosis, a persistent illness with serious side effects if neglected [30].Despite brucellosis being a notifiable disease in many countries, official statistics do not accurately reflect the number of cases that are reported each year [31].Most developing countries in Africa have listed brucellosis as an under-reported endemic infection; due to the limited number of studies and the lack of epidemiological data [25].In South Africa (SA), B. abortus and B. melitensis have been reported in humans, cattle, sheep, and goats at the turn of the century [32,33].The cattle population contributes to the majority of the income in SA and thus bovine brucellosis has a significant negative economic impact on the country's dairy and beef industries [34].A nationwide bovine brucellosis scheme existed in SA since 1979 [35] and includes the vaccination of heifers and test and slaughter of high-risk bovines such as dairy and export.This scheme is voluntary for other animal owners and depends on the resources and willingness of the owners [36].Most owners are aware that a positive test results in quarantine, which limits participation in this scheme focusing mainly on high-risk bovines as well as knowledge about brucellosis seroprevalence amongst livestock in SA.Despite the scheme, bovine brucellosis seroprevalence has increased from 3.74% to 9.18% based on retrospective results reported in SA from 2007 to 2015 [37].The aim of this study was to use a polyphasic approach to (I) screen tissue samples using Brucella ITS-PCR to detect Brucella DNA; (II) isolate Brucella from ITS-PCR-positive tissues using a selective medium; (III) assess the most appropriate sample type (lymph nodes, spleen, kidney, liver, and tonsils (the latter only from cattle and pigs)) to isolate Brucella; and (IV) characterize Brucella spp.isolates using AMOS-PCR and Bruce-ladder multiplex PCR assays from seropositive and seronegative livestock (cattle, sheep, and pigs) slaughtered at abattoirs in the Eastern Cape Province, SA.

Description of the Study Area
This study was based on voluntary participation from abattoirs in the Eastern Cape Province, SA.The Brucella isolates characterized in this study were recovered from cattle, pigs, and sheep (lymph nodes, liver, spleen, kidney, and tonsils (the latter from cattle and pigs)) collected from the abattoirs.The Eastern Cape (at 168,966 km 2 ) has the largest percentage of livestock in the country [38] and stretches along the Indian Ocean between Western Cape and KwaZulu-Natal provinces.The collection of samples was from five abattoirs in the Eastern Cape Province, but the livestock slaughtered were not only from the Eastern Cape but included livestock transported from bordering provinces such as KwaZulu-Natal and Free State.The latter two provinces do not have any movement control for livestock.In SA, Western Cape is the only province that enforces movement control of foot and mouth disease-susceptible animals into and out of the province [39].

Study Design and Sample Size
The abattoirs recruited for this study included both high-throughput and low-throughput abattoirs.The target animal population was sheep, cattle, and pigs and included apparently healthy animals with unknown Brucella status.During the abattoir visits, blood (serum) and tissue (kidney, spleen, liver, tonsils, and lymph nodes) samples were collected from corresponding animals.For this study, tissue samples were collected from 565 animals, comprising 280 cattle, 200 sheep, and 85 pigs.This number may not represent the population ratio of 12.7 million cattle, 22.3 million sheep, and 1.4 million pigs in SA (https://www.agriseta.co.za/wpcontent/uploads/2021/02/Agriseta_Red_Meat_SSSP_DIGITAL.pdf,accessed on 13 November 2023) because the number of animals sampled was dependent on the number slaughtered at the abattoir on the day of visit.The sample size was calculated using the following formula n = z 2 P exp Q/L2, where n is the sample size, P exp is the expected prevalence, and L is the precision of the estimate (also called "the allowable error" or margin of error), which is taken to be 0.05 for this study.Q = 1 − P exp , and Z is the (1 − α2) percentile of a standard normal distribution [40]; for α = 0.05, Z = 1.96.In order to estimate the sample size, 11% prevalence was applied based on recent data from other South African provinces [41] due to a lack of recent data on Brucella seropositivity in Eastern Cape province.This resulted in a sample size of 151 for each species, which adheres to the sample size per the calculation, except the porcine samples, since the abattoirs received fewer pigs during the sampling period.Animals were sequentially sampled using a randomly selected subset of a single species.Samples were collected in a sterile plastic bag and stored at −20 • C at the University of Pretoria, Department of Veterinary Tropical Diseases Biosafety Level 2+ laboratory prior to processing.

Sample Collection Procedure
An opportunistic sampling procedure was followed for the collection of the samples.Multiple animal species are slaughtered in these selected abattoirs on any given day.Animals were sampled consecutively from within a randomly selected subset of a single species.That is, for every species, the daily quota of animals was sampled one after the other to ensure the accurate sampling and assignation of samples per species.Our approach was carried out in three steps: (I).Planning: The relevant information was requested from the abattoir managers regarding the animals and herd information, and this included the age, sex, vaccination status, location, and owners or seller of the animals.However, the animals and herd vaccination information were not available, and only the abattoir and regional veterinary services are allowed access to the location and owner information, which must be requested through official procedures by the regional state veterinary office which were obtained.(II).Sample collection: The selected abattoirs in the Eastern Cape Province are located more than 100 km apart, except for two which are located within the same region.We aimed to collect samples from herds as Brucella infection is a known herd disease [42].Upon the slaughtering of animals by butchers, the animals were immediately eviscerated, and all the organs were removed from the carcasses (Figure 1A,B).To avoid animal-to-animal contamination, the knifes were cleaned with boiling hot water between uses.Approximately 100 g of each tissue was excised (Figure 1C).The sample collection procedure was lengthy, since it also included a cursory meat inspection by our team and the abattoir meat inspector (Figure 1D-F).Therefore, only one abattoir could be sampled per day.The tissues were stored in a clearly labelled sterile plastic bag followed by ~4 • C cold chain in the abattoir.(III).Packing and transportation: The samples were stored in a −20 freezer before transportation to the University of Pretoria, Department of Veterinary Tropical Diseases, SA, in triple-layer packaging for processing in accordance with the National Road Traffic Act, 1996 (Act No. 93 of 1996).

Sample Processing
The excised tissues were processed according to set laboratory protocols in a bio-safety level (BSL) 2 plus laboratory.The kidney, spleen, liver, tonsils, and lymph nodes were examined for lesions and calcification.A cubic centimetre of healthy-looking tissue was dissected with a sterile surgical blade and aliquoted into two separate cryovials.These matching tubes were submitted for direct DNA extraction, PCR, and microbial isolation, respectively.The serological test results were determined by [43] using serum samples; these were subjected to the RBT from Onderstepoort Biological Products, Pretoria, South Africa, CFT (conducted at Onderstepoort Veterinary Institute laboratory where the test is SANAS accredited for bovine, but not sheep and pigs), and the iELISA (IDVet, Grabels, France) as per the manufacturer's recommendations.

Genomic DNA Extraction
DNA was extracted directly from all the tissue samples for Brucella spp.screening.This was performed using the Pure-Link Genomic DNA Kit (tissue protocol) according to the instructions of the manufacturer (Thermo Fisher Scientific, Waltham, MA, USA).

Brucella Genus PCR Screening Using ITS
DNA amplification for the detection of the target Brucella gene using genus-specific 16S-23S rRNA interspacer region (ITS) primers (ITS66: ACATAGATCGCAGGCCAGTCA and ITS279: AGATACCGACGCAAACGCTAC) was used for the detection of Brucella DNA in the tissues [44].During the culturing process, suspected Brucella isolates were screened with ITS-PCR to detect Brucella spp.DNA.Briefly, a PCR master mix of 12 µL was prepared as follows: 6.5 µL Dream Taq polymerase, 0.3 µL (0.2 µM) Forward primer, 0.3 µL reverse primer (0.2 µM) and 4.9 µL of nuclease-free water (Thermo Fisher Scientific, Johannesburg, South Africa).From each sample, 3 µL of DNA was used in a 15 µL PCR reaction.The mix was amplified on a thermal cycler (Veriti, Thermo Fisher Scientific, Waltham, MA, USA) with a heated lid, preheated to 105

Sample Processing
The excised tissues were processed according to set laboratory protocols in a biosafety level (BSL) 2 plus laboratory.The kidney, spleen, liver, tonsils, and lymph nodes were examined for lesions and calcification.A cubic centimetre of healthy-looking tissue was dissected with a sterile surgical blade and aliquoted into two separate cryovials.These matching tubes were submitted for direct DNA extraction, PCR, and microbial isolation, respectively.The serological test results were determined by [43] using serum samples; these were subjected to the RBT from Onderstepoort Biological Products, Pretoria, South Africa, CFT (conducted at Onderstepoort Veterinary Institute laboratory where the test is SANAS accredited for bovine, but not sheep and pigs), and the iELISA (IDVet, France) as per the manufacturer's recommendations.

Genomic DNA Extraction
DNA was extracted directly from all the tissue samples for Brucella spp.screening.This was performed using the Pure-Link Genomic DNA Kit (tissue protocol) according to the instructions of the manufacturer (Thermo Fisher Scientific, Waltham, MA, USA).

Sample Preparations, Brucella Culture, and Bacteriological Examination
Each tissue was homogenized with 1 mL of ddH 2 O in a Precellys 24 lysis and tissue homogenizer (Bertin technologies, Paris, France).About 200 µL of the tissue homogenate from pre-screened Brucella ITS-PCR-positive tissues was inoculated onto the modified CITA medium [45] and incubated at 37 • C with 5.0% CO 2 for 5-14 days, followed by subculturing for purification where necessary.Culture plates were considered negative and discarded following 14 days of incubation with no growth observed.Brucella suspected isolates were selected based on their morphology on the CITA medium, Gram staining, and modified Ziehl-Neelsen staining [40].DNA was extracted from all the Brucella suspected (Gramnegative and modified Ziehl-Neelsen) isolates for molecular characterization using the Pure-Link Genomic DNA Kit (Gram-negative protocol).During purification, suspect Brucella single colonies were transferred to CITA medium and screened using staining and/or ITS-PCR.Fast-growing bacteria kept on overgrowing slow-growing Brucella colonies observed with Gram staining and ITS-PCR resulted in the impure Brucella isolates.Gram-negative fast-growing isolates were selected on the culture plates and submitted for genomic DNA extraction (Pure-Link Genomic DNA Kit; Section 2.5) and 16S sequencing (see Section 2.8).These fast-growing contaminants grew on modified CITA medium in the presence of antibiotics (natamycin, nitrofurantoin, amphotericin B, colistin, nystatin, and vancomycin).AMOS-PCR assay was used to identify Brucella spp.from DNA extracted (Pure-Link Genomic DNA Kit; Section 2.5) from impure Brucella culture isolates from livestock tissues (see Section 2.10).

AMOS-PCR and Bruce-Ladder PCR Assays
The multiplex AMOS-PCR includes species-specific primers, B. abortus (F-GAC GAA CGG AAT TTT TCC AAT CCC), B. melitensis (F-AAA TCG CGT CCT TGC TGG TCT GA), B. ovis (F-CGG GTT CTG GCA CCA TCG TCG GG), B. suis (F-GCG CGG TTT TCT GAA GGT GGT TCA), and reverse primer IS711 (R-TGC CGA TCA CTT AAG GGC CTT CAT) as described [22].Four species-specific forward primers were used at a final concentration of 0.1 µM with 0.2 µM reverse primer IS711.PCR cycling conditions consisted of an initial denaturation at 95 • C for 5 min followed by 35 cycles of 95 • C for 1 min, 55.5 • C for 2 min, 72 • C for 2 min, and a final extension step at 72 • C for 10 min.Specific amplicon sizes were determined using agarose electrophoresis.
As described by [19,46], a multiplex Bruce-ladder PCR was performed to identify and distinguish between vaccine strains and field isolates of Brucella spp.The positive controls used were B. abortus bv 1 strain (REF 544, BCCN R4), B. abortus S19 (Design Biologix, Pretoria, South Africa), and B. melitensis Rev 1 (Onderstepoort Biological Products, Pretoria, South Africa).The amplified products were examined via electrophoresis in a 2% agarose gel and stained with ethidium bromide (0.03 µL/mL).The gel ran at 120 volts for 1 h.The gel was documented under UV light by a molecular imager (Bio-rad, ChemiDoc TM XRS).

Statistical Analysis
Descriptive analysis was used to determine the frequency (percentage) of Brucella PCR positivity among the different variables (abattoir, throughput, animal species, and sex of the animal).The chi-squared or Fisher's exact test was used to determine the association between PCR positivity, on the one hand, and each of the four variables, on the other hand, in univariate analyses.The four variables, regardless of their p-value, from univariate analyses were included in subsequent multivariable models.The multivariable analysis was conducted with generalized linear models with a stepwise backward elimination procedure and Akaike Information Criteria to determine the risk factors for Brucella infection.We determined the level of agreement between PCR and culture results for Brucella spp.using the Cohen's kappa (k) test [47].The kappa result was interpreted as follows: kappa ≤ 0 = no agreement; 0.01-0.20 = none to slight; 0.41-0.60= moderate; 0.61-0.80= substantial; and 0.81-1.00= almost perfect agreement.Data analyses were performed using R statistical software version 4.21 [48], employing the packages "MASS" and "epiR", and a 0.05 level of significance.Of the tissue samples from the 280 slaughtered cattle tested using the Brucella ITS-PCR, the frequency of detection was 33.57% (94/280) (Supplementary Figure S1).Of the 200 slaughtered sheep tested using ITS-PCR, the frequency of detection was 14.5% (29/200).Of the tissue samples from the 85 slaughtered pigs tested using the ITS-PCR, the frequency of detection was 4.71% (4/85).

Identification of Gram-Negative Isolates Using Gram Staining
Tissues from Brucella ITS-PCR positive animals (127/565) were included in culturing, after which round, smooth margin, translucent, and yellowish-white coloured colonies on modified CITA medium were examined using microscopy and staining.Of the 94 cattle tissues that tested positive with ITS-PCR, 41 Brucella isolates were identified based on Gram-negative cocco-bacilli using Gram staining and were positive with modified Ziehl-Neelsen staining.Fifteen (15) Brucella suspect cultures from 29 ITS-PCR-positive sheep tissues were identified using microscopy.Additionally, two Brucella cultures were observed from four ITS-PCR pig tissues based on microscopy.Brucella colonies were further subjected to several rounds of streaking and dilution to purify the colonies.Additional fast-growing Gram-negative bacteria were also observed on culture.Spingomonas was identified among other bacteria.The organism has an identical antibiogram, thus making it impossible to select and purify Brucella from this faster growing contaminant using antibiotics.None of the isolations could be purified and thus remained impure isolates, which we identified using AMOS-PCR.

Characterization of Brucella spp. Using AMOS-PCR Assay and Seropositivity
Of the 41/94 (43.6%)Brucella suspect isolates observed on microscopy from ITS-PCR positive cattle tissues, AMOS-PCR characterized 38 as B. abortus and a mixed infection of both B. abortus and B. melitensis was observed in three cattle (Table 1).From the 15/29 (51.7%)Brucella suspect isolates from ITS-PCR positive sheep tissues, AMOS-PCR characterized 11 as B. abortus and a mixed infection of both B. abortus and B. melitensis was observed in four sheep (Figure 2A).Of the 2/4 (50%) Brucella suspect isolates from ITS-PCR positive pig tissues, AMOS-PCR characterized 1 as B. abortus and 1 as a mixed infection of B. abortus and B. melitensis (Table S1, supplementary data).The single-plex AMOS-PCR was used to separate and confirm the mixed infection of B. abortus and B. melitensis (Figure 2B,C).Using the AMOS-PCR and Bruce-ladder PCR assays, the isolates were distinguished from the 16S vaccine strain (Supplementary Figure S1).

Association between Brucella ITS-PCR Positivity and Predictor Variables
Three variables (abattoir, throughput, and animal species) out of four analysed in univariate analyses showed statistical significance (p ≤ 0.05) (Table 2).The four variables regardless of p value were included in a multivariable logistic regression model.After multivariable analysis, which followed a backward stepwise elimination procedure, only three variables (sex, species, and abattoir) out of the four comprised the final regression model (Table 3).The abattoir factor was a significant determinant for positivity amongst

Association between Brucella ITS-PCR Positivity and Predictor Variables
Three variables (abattoir, throughput, and animal species) out of four analysed in univariate analyses showed statistical significance (p ≤ 0.05) (Table 2).The four variables regardless of p value were included in a multivariable logistic regression model.After multivariable analysis, which followed a backward stepwise elimination procedure, only three variables (sex, species, and abattoir) out of the four comprised the final regression model (Table 3).The abattoir factor was a significant determinant for positivity amongst the specimens from different animal species.With abattoir B as the reference level, animals in abattoir D (39.1%; OR = 7.0, p = 0.00014), abattoir E (41.7%; OR = 5.13, p < 0.0001), and abattoir A (38.0; OR = 4.9, p < 0.0001) were more likely to be PCR positive for Brucella spp., while abattoir C (15.6%, OR = 0.91, p = 0.85) had a similar positivity rate (Tables 2 and 3).There was an almost similar likelihood of Brucella positivity between male (21.1%) and female animals (23.9%) with an odds ratio between the two levels of 0.5 (Tables 2 and 3).There was moderate agreement between the PCR and culture results (kappa = 0.57; 95% CI 0.47, 0.66; p < 0.0001).Of the 565 samples tested, 58 were positive with both PCR and culture methods, while 69 samples were positive on PCR but negative on culture.A total of 438 samples were negative with both methods.

Sequence Identification of Additional Gram-Negative and Gram-Positive Isolates from Culture
Faster-growing contaminants were a recurring hindrance to obtain pure Brucella isolates.To identify the contaminants, and, in doing so, attempting to improve the selective medium, isolates were submitted for sequencing.The following isolates were identified by nucleotide identity using QIIME2.Proteus vulgaris (21%), Cuktibacterium acnes (3%), Brevundiminas terrae, Brevundimonas naejangsanensis (20%), Serratia nematodiphila (3%), and Serratia marcescens (24%) were identified on culture from the livestock tissue samples (Figure 3).

Sequence Identification of Additional Gram-Negative and Gram-Positive Isolates from Culture
Faster-growing contaminants were a recurring hindrance to obtain pure Brucella isolates.To identify the contaminants, and, in doing so, attempting to improve the selective medium, isolates were submitted for sequencing.The following isolates were identified by nucleotide identity using QIIME2.Proteus vulgaris (21%), Cuktibacterium acnes (3%), Brevundiminas terrae, Brevundimonas naejangsanensis (20%), Serratia nematodiphila (3%), and Serratia marcescens (24%) were identified on culture from the livestock tissue samples (Figure 3).

Discussion
This study used samples available from abattoirs to investigate brucellosis, which allowed a polyphasic approach and thus serology, molecular, and bacteriology detection.Most studies only use serology, and few continue to obtain Brucella culture isolates.Brucella-specific PCR on tissues from livestock followed by culture and AMOS-PCR identification detected mainly B. abortus with a few mixed infections (B.abortus and B. melitensis) in 14.6% (41/280) of cattle, 7.5% (15/200) of sheep and 2.4% (2/85) of pig tissues collected from Eastern Cape Province abattoirs.This study demonstrated the value of the polyphasic approach, especially to identify the potential risk of brucellosis in asymptomatic animals with possible chronic infections.
This study isolated Brucella spp.from the liver, spleen, kidney, lymph nodes (mesenteric and mandibular), and tonsils of apparently healthy livestock from the abattoirs in the Eastern Cape province.Tissue samples from livestock including the liver, spleen, kidneys, lungs, and lymph nodes have previously been processed for the isolation of Brucella spp.[8,49].In this study, of the 58 AMOS-PCR identified Brucella isolates, 19.0% (11/58) were seropositive using either RBT, CFT, or iELISA [43] with the majority being seronegative.The isolation of Brucella spp.from seronegative animals (see supplementary data) may be an indication of chronic infection in the animals [50], with these asymptomatic animals posing a risk of spreading the pathogen once they become pregnant or during

Discussion
This study used samples available from abattoirs to investigate brucellosis, which allowed a polyphasic approach and thus serology, molecular, and bacteriology detection.Most studies only use serology, and few continue to obtain Brucella culture isolates.Brucellaspecific PCR on tissues from livestock followed by culture and AMOS-PCR identification detected mainly B. abortus with a few mixed infections (B.abortus and B. melitensis) in 14.6% (41/280) of cattle, 7.5% (15/200) of sheep and 2.4% (2/85) of pig tissues collected from Eastern Cape Province abattoirs.This study demonstrated the value of the polyphasic approach, especially to identify the potential risk of brucellosis in asymptomatic animals with possible chronic infections.
This study isolated Brucella spp.from the liver, spleen, kidney, lymph nodes (mesenteric and mandibular), and tonsils of apparently healthy livestock from the abattoirs in the Eastern Cape province.Tissue samples from livestock including the liver, spleen, kidneys, lungs, and lymph nodes have previously been processed for the isolation of Brucella spp.[8,49].In this study, of the 58 AMOS-PCR identified Brucella isolates, 19.0% (11/58) were seropositive using either RBT, CFT, or iELISA [43] with the majority being seronegative.The isolation of Brucella spp.from seronegative animals (see supplementary data) may be an indication of chronic infection in the animals [50], with these asymptomatic animals posing a risk of spreading the pathogen once they become pregnant or during breeding as Brucella will then start to replicate.Disease surveillance from live animals using serological tests is limiting and cannot detect latent or chronically infected animals and thus shows the value of the sample availability combined with molecular methods at abattoirs to determine the risk of contribution to disease spread and spillover [51].
Although the culture technique is not a sensitive procedure, bacterial isolation is considered as the gold standard for diagnosing Brucella spp. in humans and animals [52].Aborted tissues from a B. abortus abortion episodic yield more than 10 14 microbial organisms, which constitutes 10 5 times the presumed infectious dosage of heifers vaccinated with S19 [53].Hence, increased isolation of Brucella spp. on culture has been reported when sampling from aborted materials and vaginal swaps [54], as compared to tissues from asymptomatic and apparently healthy animals.Thus, Brucella-specific PCR was used for tissue screening before attempting low-sensitivity isolation, especially as the samples were collected from asymptomatic livestock.However, tissues were collected from organs with ER cells such as the spleen, liver, kidney, and lymph nodes as these are macrophagerich tissues where Brucella organisms multiply and concentrate and thus increase culture sensitivity [8] and can only be collected from dead animals [23].The Brucella spp.isolation frequency of cattle was higher in lymph nodes (31/94), followed by the liver (25/94) and spleen (20/94).The frequency of isolation of Brucella spp. in sheep was higher in the liver (11/29) and kidney (11/29), followed by the spleen (10/29).Brucella isolates were only recovered from the tonsils in 2/4 pigs.Brucella spp. was isolated and detected with AMOS-PCR from the lymph nodes, liver, spleen, and kidney samples from animals showing no clinical signs of brucellosis infection.Thus, this suggests that the above-mentioned tissues may be utilized for brucellosis screening purposes and diagnostics in slaughtered abattoir animals.This study also highlights improved assessment standards and procedures that may result from routine sampling, such as obtaining tonsils from monogastric animals and the liver, kidneys, and spleen from ruminants.
This study further shows the presence of fast-growing contaminants which makes the isolation of low-concentration Brucella from asymptomatic animals impossible despite various attempts.It has been reported that Brucella isolation from vaginal secretions, placenta, foetal tissues, milk, and semen from animals are normally impaired by contaminants that overgrow the slow-growing brucellae, even on selected media [8].The presence of these other fast-growing Gram-negative bacteria on culture affects the growth of Brucella spp.through competitive inhibition, thus resulting in impure/contaminated isolates.The present study reports the isolation of other pathogenic organisms such as Proteus Vulgaris, Serratia marcescens, and Brevundimonas naejangsanensis.Proteus vulgaris has been reported as a zoonotic infection, which is mainly known for causing wound and urinary infections in humans [55].Previous researchers have reported S. marcescens as a common cause of mastitis and early abortions in cows [56,57].Brevundimonas naejangsanensis is an environmental Gram-negative bacterium, which has been isolated from the soil [58].The risk of zoonotic diseases is increased by the isolation of potentially harmful foodborne pathogens such as Brucella spp.and P. vulgaris [59], from apparently healthy abattoir livestock.Microbial contamination of the abattoir meat may occur during the exsanguination process, particularly if a sterile environment is not maintained [60].Based on our study, it was observed that the butcher only washes their knife to remove the excessive amount of blood and not to avoid contamination from one animal to the other.It was also observed that the operators clean/spray the floors frequently for blood removal.However, this process allows contaminated water/blood to splash onto the meat.In [60], the authors reported the blood removal procedure on the floor as being unhygienic.To reduce these contaminants, it is advisable to surface-sterilize the tissues before culturing, which can reduce these contaminants.
In this study, on gross pathological examination, yellowish-white lesions, discoloration/bruises, abscesses, and cysts were observed on some cattle tissues.This included the mesenteric lymph nodes, skin, liver, and the spleen.The presence of lesions in the mesenteric lymph nodes can indicate Mycobacterium tuberculosis complex infection.As reported by a similar study conducted in the Eastern Cape abattoirs, the presence of nodular lesions was observed in 162 cattle lymph node samples with visible inflammation [61].Their study reported the isolation of Mycobacterium bovis and M. tuberculosis [61].Feed-lot cattle may develop liver abscesses as a result of vigorous grain-feeding programs, which are also influenced by a number of nutritional and management factors [62].Our findings are in agreement with other studies which identified major causes of offal and carcass condemnation in the Eastern Cape abattoirs including tongue and spleen abscesses, bruises, actinobacillosis, heart and kidney cysts, inflammatory conditions, and improper evisceration [63].However, the underlying causes of the conditions remain unknown.Due to the tissue condemnation and decreased meat yield, the presence of pathological evidence on the tissue has a major economic impact on the animal industry [62] and increases the risks of zoonotic infections to humans.
Multivariable analyses showed that sheep (14.5%;OR = 5.6, p = 0.0043) and cattle (33.6%;OR = 17.1, p < 0.0001) were significantly more likely to be AMOS-PCR-positive for Brucella species compared to pigs (4.7%).The current study reports the isolation of Brucella spp. in 43.6% (41/94) cattle, 51.7% (15/29) sheep, and 50% (2/4) pig samples using AMOS-PCR, which only detects B. abortus bv 1, 2, and 4, B. melitensis bv 1-3, B. ovis, and B. suis bv 1.A similar study conducted in the Eastern Cape Province reported an increased isolation of Brucella spp.from cattle (62.3%) as compared to goats (25.4%) and sheep (12.3%), also using AMOS-PCR.The current bovine brucellosis scheme includes the mandatory vaccination of heifers aged 4-8 months using 16S vaccine, serological testing, and surveillance of high-risk farms, particularly dairy and breeding cattle with suspected or proven brucellosis infections [64,65].However, the participation of the farmers is voluntary and self-funded, thus negatively affecting the role and importance of early vaccination.None of the B. abortus isolates from the livestock tissue were the S19 vaccine strain.This study indicates an almost similar likelihood of Brucella positivity between male (21.1%) and female animals (23.9%), with an odds ratio between the two levels of 0.5.The abattoirs (except abattoir C), species, and sex were significant determinants for positivity in our study with a p ≤ 0.05.The Brucella positivity in male animals may be due to high exposure of the bacteria or through the consumption of milk from infected females.An increased positivity was observed from low-throughout abattoirs (39.1%) as compared to high-throughput abattoirs (21.8%).As reported by [41], an increased sero-positivity and isolation of Brucella spp. was also observed from low-throughput abattoirs as compared to high-throughput abattoirs in Gauteng Province.This may be because low-throughput abattoirs receive animals from the local community alongside animals from the same herd or animals grazing together, thus increasing the possibility of transmission between each other.
Brucellosis is a controlled zoonotic infection in animals and a notifiable disease in humans in SA [66].The infection is a major public health challenge and still predominant as a neglected endemic zoonosis requiring proactive considerations in numerous communities worldwide [67].Serological tests have been used to detect brucellosis throughout SA in bovine.However, brucellosis outbreaks have been reported mainly in the central and highveld regions [68].The brucellosis scheme in SA is focused on bovine, and, from this study, B. abortus was the dominant species detected with AMOS-PCR in Brucellainfected animals.Brucella abortus was not only detected in bovine but also sheep and pigs, which indicates spill over to these species in SA.A previous study conducted in the Eastern Cape, reported the isolation of B. abortus in cattle, sheep, and goats, whereas the isolation of B. melitensis was observed in sheep and goats [69].As reported by [41], the first case in SA of B. melitensis in cattle was isolated from abattoirs in Gauteng Province.The current study reports the isolation of B. melitensis from abattoirs cattle in the Eastern Cape Province.Serological tests cannot differentiate between Brucella species, therefore, brucellosis seropositive bovines are presumed to be B. abortus, while seropositive sheep and goats are presumed to be infected with B. melitensis.Mixed B. abortus and B. melitensis infections were also detected in all livestock in this study and need further investigations.Despite the tremendous efforts of the SA government in the eradication of infection, an increased number of reports continue to indicate the presence of brucellosis in livestock in SA [41,69].Surveillance schemes in countries where brucellosis has been eradicated mainly focus on the vaccination of livestock as well as the test and slaughtering schemes of all relevant species [70], unlike SA, which focuses only on high-risk bovines [71].Brucellosis eradications takes decades, and it is a costly exercise [72].In endemic countries such as SA, serological tests will have their limitations due to chronically infected animals since the antibody level is below detection in these animals.However, serological tests will identify some infected animals, but the results in this study indicate that testing should be expanded to all bovine and to other livestock species as well, especially sheep and goats, to increase the detection of brucellosis.

Conclusions
This study has demonstrated the importance of multiple tests in the diagnosis and surveillance of brucellosis, as is evidenced by the isolation and identification of B. abortus and B. melitensis from seropositive but mainly seronegative asymptomatic livestock.The use of only serological tests in chronic infected animals results in false negative results.This study demonstrates the value of the polyphasic approach using the molecular method in combination with samples from abattoirs, especially to identify the potential risk of brucellosis in asymptomatic animals with possible chronic infections.This study also emphasizes refined evaluation criteria, and processes could come from routine sampling, i.e., collecting the liver, kidneys, and spleen from ruminants and tonsils from slaughtered animals.Abattoirs prove to be a valuable surveillance resource as their tissues are easily accessible post-slaughter.More data included from such sites would allow for a much clearer epidemiological picture of brucellosis in provinces across SA.This could, in turn, provide better data with which to plan targeted surveillance for both B. abortus and B. melitensis infections in livestock and to make effective management decisions against this devastating herd disease.

Limitations of the Study
The isolation of Brucella spp. was recovered from the livestock samples.However, due to the increased growth of other fast-growing Gram-negative bacteria, impure cultures were observed.The Bruce-ladder PCR assay requires highly concentrated Brucella DNA to amplify the multiple targets of this assay.Mixed infections of B. abortus and B. melitensis were observed in all the species (cattle, sheep, and pigs).However, due to the confluent growth of contaminants, the mixed Brucella spp.could not be isolated separately.Further investigation, which will be possible in a larger study, is recommended.This could include the surface sterilization of tissues to reduce the growth of other organisms, thus allowing the Brucella spp. to grow confluently.

Supplementary Materials:
The following supporting information can be downloaded at https: //www.mdpi.com/article/10.3390/microorganisms12010223/s1.All supplementary materials used in this study are attached as a separate file; Table S1: Molecular and serological identification of Brucella spp. in livestock.; .Risk assessment and appropriate health and safety measures were taken during the sample collection and processing.The identity of the abattoir is kept private in accordance with the terms of the research ethics approval, participant permission, and agreement regarding inclusion in this study.

18 Figure 1 .
Figure 1.Collection of samples during slaughtering process workflow and inspection in this study.(A,B): Livestock in holding pens at the abattoirs.(C) Processing of approximately 100 g of each tissue (D) Atypical cattle lymph node with visible lesions.(E) Cyst/abscess on a liver.(F) Bruised carcass post slaughter.

Figure 1 .
Figure 1.Collection of samples during slaughtering process workflow and inspection in this study.(A,B): Livestock in holding pens at the abattoirs.(C) Processing of approximately 100 g of each tissue (D) Atypical cattle lymph node with visible lesions.(E) Cyst/abscess on a liver.(F) Bruised carcass post slaughter.

2. 11 .
Ethical Considerations Approvals from the Research and Animal Ethics Committees of the University of Pretoria (Ref: REC 028-22), section 20 of the Animal Diseases Act, (Act No. 35 of 1984) from the Department of Agriculture, Land Reform and Rural Development (DALRRD) were obtained.Appropriate health and safety precautions with risk assessments were followed throughout the collection and processing of the samples.

1 .
Identification of Brucella spp.Directly from the Tissues Using 16S-23S Ribosomal DNA Interspacer (ITS) Region PCR Assay

Figure 2 .
Figure 2. Amplification from Brucella isolates from sheep tissues using AMOS-PCR with 498 bp amplifying B. abortus target region and 731 bp amplifying the B. melitensis target region.(A) Multi-plex AMOS-PCR with a mixed infection of both B. abortus and B. melitensis isolated from the kidneys in lanes 1-3; B. abortus isolated from the liver in lane 4; negative control in lane 5 and 9 (-ve); B. melitensis Rev 1 and B. abortus positive controls in lane 6 and 7. (B) Single-plex B. abortus-specific primer of AMOS-PCR with negative water control and 731 bp B. melitensis PCR product using B. melitensis Rev 1 positive control; lanes 1-7 included tissues that were AMOS-negative in lanes 1 and 2; and lane 3-7 included mixed B. abortus and B. melitensis sheep isolates from 4 sheep (with the same animals repeated in lanes 3 and 7).(C) Single-plex B. abortus-specific primer of AMOS-PCR with negative water control and 498 bp B. abortus PCR product using B. abortus S19 positive control; lanes 1-7 included tissues that were AMOS-negative in lanes 1 and 2; and lane 3-7 included mixed B. abortus and B. melitensis sheep isolates from 4 sheep (with the same animals repeated in lanes 3 and 7).

Figure 2 .
Figure 2. Amplification from Brucella isolates from sheep tissues using AMOS-PCR with 498 bp amplifying B. abortus target region and 731 bp amplifying the B. melitensis target region.(A) Multi-plex AMOS-PCR with a mixed infection of both B. abortus and B. melitensis isolated from the kidneys in lanes 1-3; B. abortus isolated from the liver in lane 4; negative control in lane 5 and 9 (-ve); B. melitensis Rev 1 and B. abortus positive controls in lane 6 and 7. (B) Single-plex B. abortus-specific primer of AMOS-PCR with negative water control and 731 bp B. melitensis PCR product using B. melitensis Rev 1 positive control; lanes 1-7 included tissues that were AMOS-negative in lanes 1 and 2; and lane 3-7 included mixed B. abortus and B. melitensis sheep isolates from 4 sheep (with the same animals repeated in lanes 3 and 7).(C) Single-plex B. abortus-specific primer of AMOS-PCR with negative water control and 498 bp B. abortus PCR product using B. abortus S19 positive control; lanes 1-7 included tissues that were AMOS-negative in lanes 1 and 2; and lane 3-7 included mixed B. abortus and B. melitensis sheep isolates from 4 sheep (with the same animals repeated in lanes 3 and 7).

Figure 3 .
Figure 3. Sequencing identification of other bacterial organisms present in cultures from abattoir livestock tissues on modified CITA medium.

Figure 3 .
Figure 3. Sequencing identification of other bacterial organisms present in cultures from abattoir livestock tissues on modified CITA medium.
Figure S1: Gel electrophoresis of Bruce-Ladder PCR amplification to differentiate the field strains from vaccine strains.Lanes 1-5 and 10 show amplification of B. abortus, lanes 6-8 show amplification of B. melitensis, lane 9 show Rev 1 positive and lane 11 and 12 show negative controls.Author Contributions: Conceptualization: K.D.M., F.B.K. and H.v.H.; Formal analysis: K.D.M., C.B. and H.v.H.; Funding acquisition, H.v.H.; Investigation: K.D.M.; Methodology: K.D.M., A.H. and H.v.H.; Project administration, H.v.H.; Resources: H.v.H.; Supervision: F.B.K., I.F.J. and H.v.H.; Writing-original draft: K.D.M.; Writing-review & editing: F.B.K., C.B., A.H. and H.v.H.All authors have read and agreed to the published version of the manuscript.Funding: The authors would like to thank the Department of Veterinary Tropical Diseases (AgriSETA), South Africa, the Institute of Tropical Medicine Antwerp, Belgium (FA4 & FA5), and UNICEF for the research funding.The corresponding author would also like to acknowledge the NRF-DAAD scholarship fund.Institutional Review Board Statement: The approvals from the Research, and Animal Ethics Committees of the University of Pretoria (Ref: REC 028-22), and the Eastern Cape Department of Agriculture and Rural Development were obtained prior to the commencement of the study.Section 20 of the Animal Diseases Act, (Act No. 35 of 1984) approval was obtained from Department of Agriculture, Land Reform and Rural Development (DALRRD)

Table 2 .
Descriptive and univariate analyses to determine the association between various factors and occurrence of Brucella spp. in the tissue was determined using ITS-PCR.
3.6.Level of Agreement between PCR and Culture Results for the Detection of Brucella spp.