Identification, Genotyping and Antimicrobial Susceptibility Testing of Brucella spp. Isolated from Livestock in Egypt

Brucellosis is a highly contagious zoonosis worldwide with economic and public health impacts. The aim of the present study was to identify Brucella (B.) spp. isolated from animal populations located in different districts of Egypt and to determine their antimicrobial resistance. In total, 34-suspected Brucella isolates were recovered from lymph nodes, milk, and fetal abomasal contents of infected cattle, buffaloes, sheep, and goats from nine districts in Egypt. The isolates were identified by microbiological methods and matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS). Differentiation and genotyping were confirmed using multiplex PCR for B. abortus, Brucella melitensis, Brucella ovis, and Brucella suis (AMOS) and Bruce-ladder PCR. Antimicrobial susceptibility testing against clinically used antimicrobial agents (chloramphenicol, ciprofloxacin, erythromycin, gentamicin, imipenem, rifampicin, streptomycin, and tetracycline) was performed using E-Test. The antimicrobial resistance-associated genes and mutations in Brucella isolates were confirmed using molecular tools. In total, 29 Brucella isolates (eight B. abortus biovar 1 and 21 B. melitensis biovar 3) were identified and typed. The resistance of B. melitensis to ciprofloxacin, erythromycin, imipenem, rifampicin, and streptomycin were 76.2%, 19.0%, 76.2%, 66.7%, and 4.8%, respectively. Whereas, 25.0%, 87.5%, 25.0%, and 37.5% of B. abortus were resistant to ciprofloxacin, erythromycin, imipenem, and rifampicin, respectively. Mutations in the rpoB gene associated with rifampicin resistance were identified in all phenotypically resistant isolates. Mutations in gyrA and gyrB genes associated with ciprofloxacin resistance were identified in four phenotypically resistant isolates of B. melitensis. This is the first study highlighting the antimicrobial resistance in Brucella isolated from different animal species in Egypt. Mutations detected in genes associated with antimicrobial resistance unravel the molecular mechanisms of resistance in Brucella isolates from Egypt. The mutations in the rpoB gene in phenotypically resistant B. abortus isolates in this study were reported for the first time in Egypt.

in farm animals to promote growth or as prophylaxis also contributes to the development of resistant bacteria and plays a key role in their spread along the food chain [42]. Antimicrobial resistance in zoonotic pathogens is an additional risk because it will limit disease treatment options in public health and veterinary settings [43]. None of the available studies highlights detailed antimicrobial susceptibility patterns of Brucella isolates from livestock in Egypt.
The use of antimicrobial susceptibility testing is the solution for appropriate control and treatment of brucellosis [44,45]. Micro-dilution and/or gradient strip (E-test) methods are used to establish minimum inhibitory concentration (MIC) for antimicrobials [45,46]. PCR assays and the subsequent sequencing of genes associated with resistance are used to identify the genetic bases of resistance [47][48][49].
This study aimed to isolate, identify and biotype Brucella strains from livestock in various regions of Egypt. Antimicrobial resistance and its genetic basis are to be investigated in the gained Brucella isolates.

Isolation and Identification
A total of 34 suspected Brucella isolates were recovered from clinical specimens of lymph nodes, milk and fetal stomach contents from infected cattle, buffaloes, sheep and goats located in Giza, Beheria, Asyut, Qalyubia, Beni-Suef, Ismailia, Dakahlia, and Monufia governorates/districts in Egypt (Table 1).
Bacterial isolation and identification were performed in Biological Safety Level-3 (BSL-3) laboratory. Isolates were inoculated on calf blood agar, Brucella medium and Brucella selective medium plates (Oxoid GmbH, Wesel, Germany) at 37 • C in the absence and presence of 5-10% CO2 for up to 2 weeks. Typically, round, glistening, pinpoint and honey drop-like cultures were picked and stained with Gram and modified Ziehl-Neelsen staining (MZN) methods. Subsequent biochemical tests, motility test, hemolysis on blood agar and agglutination with monospecific sera were performed [24,53]. Isolates were stored at −20 • C for further processing.

Identification by MALDI-TOF MS
Bacterial identification was additionally carried out using MALDI-TOF MS as described previously [27,54]. Briefly, pure cultures of suspected Brucella were obtained by incubating inoculated chocolate PolyViteX (PVX) agar plates (bioMérieux, Marcy-l'Étoile, France) for 48 h at 37 • C in the presence of 5% CO 2 . Samples were reliably inactivated in Biological Safety Level-3 laboratory. Approximately 10 colonies from culture medium were suspended in 50 µL of sterile HPLC water and mixed carefully. Formic acid (v/v 70%) was added for the inactivation of brucellae and for extraction of proteins. Then, 1 µL of tested sample and Brucella reference strains were added onto spots of a steel target plate. After inactivation, the plate was dried at room temperature followed by the addition of 0.5 µL of 100% ethanol to each well. Finally, spots were overlaid with 1 µL of reconstituted alpha-cyano-4-hydroxycinnamic acid (Bruker Daltonics, Billerica, MA, USA).
Spectra were acquired with an Ultraflex instrument (Bruker Daltonics GmbH, Bremen, Germany). Analysis was done with the Biotyper 3.1 software (Bruker Daltonics GmbH, Germany) as per the manufacturer's instructions to exclude spectra with outlier peaks or anomalies.
Logarithmic score values (0-3.0) were determined by automatically calculating the proportion of matching peaks and peak intensities between the test spectrum and the reference spectra in the database. The identification was considered reliable when the score between 2.3 and 3.0. A logarithmic score of 1.7-2.299 was reported as 'probable genus identification', indicating that identification was reliable only at the genus level. When the logarithmic score was <1.7, the spectrum was reported as 'not reliable identification', indicating that sample could not be identified.

Genomic DNA Extraction and Purification
DNA was extracted from heat inactivated pure Brucella culture (biomass) using the HighPure PCR Template Preparation Kit (Roche Diagnostics, Mannheim, Germany) according to the manufacturer's instructions. DNA quantity and purity were determined using a NanoDrop™ 1000 spectrophotometer (Thermo Fisher Scientific, Wilmington, USA).
PCR condition was initiated by initial denaturation at 93 • C for 5 min, followed by 35 cycles of denaturation at 90 • C for 60 s, annealing at 60 • C for 60 s and elongation at 72 • C for 60 s and final elongation step at 72 • C for 5 min. PCR products (223 bp) were analyzed on 1.5% agarose gel, stained with ethidium bromide, and visualized under UV light.
The AMOS-PCR was performed to differentiate Brucella species [29,32] followed by a multiplex Bruce-ladder PCR assay for strain and biovar typing [30,56]. The list of primers and primer sequences for AMOS-PCR and Bruce-ladder PCR were geared from previously published [29] and [30], respectively. Briefly, for AMOS-PCR, PCR was performed using 25 µL reaction mixture containing 9.5 µL HPLC water, 12.5 µL of 2x Qiagen Master mix (Qiagen, Germany), 1 µL of 10 pmol primer mix and 2 µL DNA template. Initial denaturation at 95 • C for 5 min, was followed by 30 cycles of denaturation at 95 • C for 60 s, annealing at 58 • C for 2 min and elongation at 72 • C for 2 min and a final elongation step at 72 • C for 5 min. The Bruce-ladder PCR was performed using 12.5 µL reaction mixture containing 4.25 µL HPLC water, 6.25 µl of 2x Qiagen Master mix (Qiagen, Germany), 1 µL of 2 pmol/µL primer mix and 1 µL DNA template. Initial denaturation at 95 • C for 15 min, was followed by 25 cycles of denaturation at 94 • C for 30 s, annealing at 58 • C for 90 s, elongation at 72 • C for 3 min and a final elongation step at 72 • C for 10 min.
The PCR products from each PCR were separated by electrophoresis using 1.5% agarose gels (120 V for 60 min for conventional and AMOS-PCR and 130 V for 60 min for Bruce-ladder PCR). Gels were stained with ethidium bromide and photographed using a gene snap camera (Syngene Pvt Ltd., Cambridge, UK).

Antimicrobial Susceptibility Testing
The antimicrobial susceptibility of B. melitensis and B. abortus isolates was performed against eight clinically relevant antimicrobial agents (chloramphenicol, ciprofloxacin, erythromycin, gentamicin, imipenem, rifampicin, streptomycin and tetracycline) using gradient strip method (E-test, bioMerieux, Marcy L'Etoile, France) as described previously [48]. Briefly, a suspension of bacteria adjusted to 0.5 McFarland standard units was inoculated on Mueller-Hinton plates (Oxoid GmbH, Wesel, Germany) supplemented with 5% sheep blood and the gradient strips were applied. The plates were incubated at 37 • C with 5% CO 2 for 48 h before reading. As MIC breakpoints for clinically used antimicrobials are not yet established for brucellae, the guidelines for slow-growing bacteria (Haemophilus influenzae) were used as an alternative [57]. Quality control assays were performed using E. coli (161008BR3642, DSM 1103, ATCC 25922). The susceptibility profiles of Brucella isolates are presented as resistant and susceptible using minimum inhibitory concentrations (MIC), MIC 50 and MIC 90 . The interpretations were performed using CLSI (The Clinical and Laboratory Standards Institute) [57] and EUCAST (The European Committee on Antimicrobial Susceptibility Testing) [58] using the criteria for slow growing bacteria. For rifampin, the strains were also classified as intermediate (Table 2).

Molecular Detection of Antimicrobial Resistance-Associated Genes
The PCR assays were performed as described previously [47,49,52,59] to detect the antimicrobial resistance-associated genes, i.e., catB, gyrA and gyrB, rpoB, Aac genes and tet genes for chloramphenicol, ciprofloxacin, rifampicin, streptomycin, gentamicin and tetracycline, respectively (Supplementary Table S1). The primers used for amplification of the rpoB gene were designed by using submitted sequences for the rpoB gene of B. abortus (accession number AY562181) [47]. PCR was performed using 25 µL reaction mixture containing 2x Qiagen Mastermix, 10 pmol each forward and reverse primer ( Table 1) and 5 µl DNA template. PCR was carried out by initial denaturation at 95 • C for 10 min, followed by 35 cycles of denaturation at 95 • C for 45 s, annealing (temperatures for each primer are given in Table 1) for 60 s, elongation at 72 • C for 60 s and a final elongation step at 72 • C for 10 min. Twenty microliters of each reaction mixture were analyzed by gel electrophoresis (1% agarose gel with ethidium bromide).

PCR Amplicon Sequencing and Data Analysis
Amplified PCR products for gyrA, gyrB and rpoB genes were purified using Qiagen QIAquick Gel extraction kit (Qiagen, Germany) and sent for sequencing (Eurofins Genomics Germany GmbH, Ebersberg, Germany). All consensus sequences were aligned and compared to the reference Brucella genes obtained from NCBI for detection and evaluation of nucleotide diversity and mutations using the software Geneious ® R11.1.5 (https://www.geneious.com). The sequences of gyrA (CP034103 and AE017223), gyrB (CP007760 and SDWB01000001) and rpoB (AY562181 and AY540346) genes of B. melitensis and B. abortus were geared from Gene bank and used as reference. Amino acid sequences were determined along with nucleotide sequences to identify missense mutations using BLAST.

Microbiological Identification
Based on microbiological and biochemical characteristics, 21 strains were typed as B. melitensis biovar 3, eight strains were B. abortus biovar 1 and five samples were identified as Achromobacter species (Table 1). The results of MALDI-TOF MS confirmed five isolates as Achromobacter species while the remaining 29 isolates were identified as Brucella species (Table 1).

Molecular Identification and Differentiation
Brucella DNA of 24 isolates from cattle, three from buffaloes, one from a sheep and one from a goat were amplified with the genus specific assay. AMOS-PCR and Bruce-ladder PCR differentiated these 21 isolates as B. melitensis (17 from cattle, two from buffaloes, 1 from a sheep and 1 from a goat) and 8 isolates as B. abortus (seven from cattle and one from a buffalo). All isolates were confirmed as field strains (Table 1).

Antimicrobial Susceptibility Profiling
The in vitro MIC values against eight antimicrobial agents of all 29 Brucella isolates were determined by the gradient strip method (E-test). The MIC values along with MIC 50 and MIC 90 are summarized in Table 2.

Detection of Antimicrobial Resistance-Associated Genes and Mutations
Genes associated with antimicrobial resistance (catB, Aac and tet (tetA, tetB, tetM and tetO) conferring resistance to chloramphenicol, streptomycin/gentamicin and tetracycline, respectively) were not identified either in resistant or sensitive isolates. The gyrA, gyrB and rpoB genes were amplified in all isolates.
Mutations in rpoB gene associated with a rifampicin-resistant B. melitensis and B. abortus phenotypes were detected at different positions (Table 3).
Three-point mutations were also detected in gyrB gene at position 1141 (AAG to GAG/Lysine to Glutamine), 1144 (ATC to CTC/Isoleucine to leucine) and 1421 (TCA to TTA/Serine to Leucine) in phenotypically resistant B. melitensis isolates (Table 4).
Repeated mutations were detected at positions 676, 677 (TAC to CTC/tyrosine to leucine) and 1435 (AAG to CAG/lysine to glutamine) in the rpoB gene of phenotypic resistant B. melitensis isolates while the same was recorded at position 2890 (CGT to GGT/arginine to glycine) in the rpoB gene of B. abortus isolates. No mutation was detected in gyrA and gyrB gene of B. abortus strains.

Discussion
Brucellosis is a zoonotic disease of public health importance and is still endemic in many countries including Egypt [17,20]. In this study, the phenotypic and molecular characterization of Brucella isolates from cattle, buffaloes, sheep and goats obtained from different geographical locations of Egypt was performed. Additionally, the molecular basis of antimicrobial resistance in Brucella isolates from Egypt is reported for the first time. These results contribute to a better understanding of geographic transmission and spread of brucellae in livestock in Egypt and pave a way for specific treatment and control of the disease in animals and as well as in humans.
For the accurate diagnosis of brucellosis, isolation of bacteria or molecular proof along with suggestive clinical signs is needed. Brucellae were isolated in this study from milk, lymph nodes and fetal stomach contents as recommended in previous reports [24,60].
Twenty-one B. melitensis bv3 and 8 B. abortus bv1 were isolated from cattle, buffaloes, sheep and goats from Giza, Beheria, Asyut, Qalyubia, Beni-Suef, Ismailia, Dakahlia and Monufia governorates. Previous reports were described previously that Brucella was prevailing in the country [12]. The isolation of B. melitensis from cattle and buffaloes in this study may be attributed to mixed farming of large and small ruminants as mentioned previously [13].
Still brucellosis is a challenge to treat in humans, particularly after delayed diagnosis of the infection. The WHO (World Health Organization) recommended treatment include high oral doses of rifampicin, doxycycline or tetracycline and trimethoprim-sulfamethoxazole. Although streptomycin and tetracycline are considered as powerful therapeutic agents against brucellosis, their higher toxicity limits their use [52,61]. Quinolones are promising alternatives to treat human brucellosis as they have good bioavailability and affinity for bone and soft tissues [51].
Only one study from Brazil reported reduced antimicrobial sensitivity in brucellae isolated from cattle [62]. However, the emergence of brucellae isolated from humans phenotypically resistant to ciprofloxacin, gentamycin, streptomycin, rifampicin and trimethoprim-sulfamethoxazole was reported in Egypt, Iran, Qatar, China, Norway and Malaysia [46,48,[63][64][65]. Phenotypically rifampicin resistant B. melitensis isolates were also reported from Norway in imported cases from the Middle East, Asia or Africa [45]. Probable rifampicin resistance was noted in 19% of a large collection of B. melitensis isolates from humans in Egypt between 1999 to 2007 [65]. However, none of those isolates were investigated further to confirm the basis of resistance or reduced susceptibility.
In this study, a notable phenotypic resistance against ciprofloxacin (76.19%) was detected in B. melitensis strains isolated from animals. In contrast, none of the mentioned studies reported ciprofloxacin resistance in clinical isolates of humans and animals before. However, antimicrobial resistance against quinolones has been reported in in vitro studies of B. melitensis from Greece and France [49,52].
The most striking finding of the present study was the emergence of phenotypic antimicrobial resistance against erythromycin (19.04%), imipenem (76.19%) and streptomycin (4.76%) in B. melitensis isolates. However, the increased use of these antimicrobials in Egypt in veterinary and human practices may be the cause of the emerging of this resistance [37].
The phenotypic antimicrobial resistance against ciprofloxacin (25%), erythromycin (87.5%), imipenem (25%) and rifampicin (37.5%) of B. abortus isolated in this study was not proved previously. Multidrug resistant strains of B. abortus isolated from cattle in this study were reported previously in Brazil [62]. Four isolates of B. melitensis and one isolate of B. abortus showed multidrug resistance against ciprofloxacin, erythromycin, imipenem and rifampicin. These findings are in agreement with the results of Barbosa Pauletti et al. who find corresponding resistance among B. abortus isolates from cattle in brazil [62]. All B. melitensis and B. abortus isolates in this study were sensitive to chloramphenicol, gentamicin and tetracycline. These findings are comparable to previously published reports in Egypt, China, Qatar and Kazakhstan [46,48,65,66].
The target for rifampicin action in Brucella as well as in other bacteria is the beta-subunit of the DNA dependent RNA polymerase (RNAP) encoded by rpoB gene [47,51]. In this study, mutations were identified in rpoB gene associated with phenotypic rifampicin resistant Brucella strains isolated from clinical specimens of animals in Egypt. Mutations were detected in all phenotypically resistant brucellae. Multiple and variable mutations were noted in each isolate along with few commonly shared mutations among many isolates. Frequent mutations at positions 676, 677-TAC to CTC (tyrosine to leucine, 38%) and 1435-AAG to CAG (lysine to glutamine, 23.8%) in the rpoB gene of phenotypically resistant B. melitensis were detected. These mutations are different from previously reported mutations (in vitro mutations) associated with rifampicin resistance in Brucella [47].
Johansen et al. reported mutations in phenotypic rifampicin resistant or intermediately resistant B. melitensis isolates [45], which in agreement with the findings of this study with additional mutations were detected as well as in intermediate rifampicin resistant B. melitensis.
To the best of our knowledge, this study is the first report that proved mutations in the rpoB gene of rifampicin resistant B. abortus strains. Frequent mutations were detected at position 2890-CGT to GGT (arginine to glycine, 37.5%).
Fluoroquinolone/quinolone resistance in Brucella is multifactorial by nature in addition to obvious mutations of the gyrA, gyrB, parC and parE genes [51,52]. In this study, the mutations in gyrA and gyrB genes in phenotypically resistant B. melitensis and B. abortus to ciprofloxacin were investigated. The mutations in gyrA did not correspond with fluoroquinolone resistance mutations described by Turkmani et al. [49], although they investigated mutations in vitro selected fluoroquinolone resistant Brucella mutants. The mutations in the gyrB gene detected at positions 1141-AAG to GAG (lysine to glutamine), 1144-ATC to CTC (isoleucine to leucine) and 1421-TCA to TTA (serine to leucine) of B. melitensis considered as novel findings of this study. None of these mutations was detected in B. abortus strains in gyrA or gyrB genes. However, the role of parC, parE and efflux systems cannot be ruled out for fluoroquinolone resistance [51] as we did not investigate the changes in parC and parE genes.
Genes responsible for resistance against chloramphenicol (catB), gentamicin (Aac) and tetracycline (tetA, tetB, tetM and tetO) were not detected in all investigated Brucella isolated in this study, which in accordance with the phenotypic antimicrobial susceptibility results of isolated Brucella isolates. It is also worth mentioning that all resistant Brucella strains were isolated from animals and they showed resistance to antimicrobials clinically used in humans practice, suggesting that the source of these Brucella strains may be of human origin. These findings point to the fact that inter-species and intra-host species Brucella transmission is common, but spillback may occur also when chronic human brucellosis is mistreated and resistant strains are shedded [67]. A likely scenario would be the animal keeper interface.
The emergence of antimicrobial resistance (AMR) in bacteria is a public health issue globally and already compromises the treatment options regarding effectiveness of antimicrobials and control of several bacterial infections especially caused by gram-negative bacteria [68]. Wide spreading AMR in these bacteria is likely to persist and even worsen in future due to the uncontrolled use of antimicrobials. Rifampicin and ciprofloxacin are effective against intracellular bacteria like Brucella [33]. Higher phenotypic resistance in Brucella against these antimicrobials is likely to limits the treatment effectiveness, owing to the increased number of infections. Emergence of multidrug resistance Brucella in livestock species in this study may pose serious threat to humans as these bacteria often transferred from animals to humans through food chain [69]. Being a zoonotic pathogen and given the emergence of increased antimicrobial resistance in Brucella species, the situation with respect to hospital care may worsen and limits the treatment options in public health settings.

Conclusions
Brucellosis is a contagious and often communicable worldwide zoonosis with high morbidity and low mortality. There has been a tremendous increase in inter host-species infection in the recent decades, especially in developing countries when farm animal species are kept on the same premises without biosecurity precautions. The disease is endemic in Egypt and B. melitensis and B. abortus have been reported as the main causative agents of brucellosis in humans and animals. High phenotypic resistance against ciprofloxacin, erythromycin, and imipenem were detected in Brucella spp. isolated from different districts and animals species reflecting a broad geographical distribution. The molecular identification of mutations in antimicrobial resistance associated genes highlight the mechanism of resistance in Brucella spp. There is a need for further insights into the epidemiology and spread of antimicrobial resistant Brucella in Egypt. The WHO regimes have to be reevaluated and awareness among physicians about AMR needs to be raised.