Abstract
A broth microdilution method was used to determine the antimicrobial susceptibility of 20 Brachyspira intermedia isolates obtained from different layer flocks in Belgium and The Netherlands between 2008 and 2010. The antimicrobial agents used were tylosin, tilmicosin, tiamulin, valnemulin, doxycycline, and lincomycin. The minimal inhibitory concentration (MIC) distribution patterns of tylosin, tilmicosin, lincomycin, and doxycycline were bimodal, demonstrating acquired resistance against doxycycline in three strains, against the macrolides in two strains, and against lincomycin in one strain. The MICs of tiamulin and valnemulin showed a monomodal distribution, but with tailing toward the higher MIC values, possibly suggesting low-level acquired resistance in six isolates. Sequencing revealed a G1058C mutation in the 16S rRNA gene in all doxycycline-resistant strains. The strain resistant to tylosin, tilmicosin, and lincomycin had an A2058T mutation in the 23S rRNA gene.
Introduction
To date, only a few publications deal with antimicrobial susceptibility testing of Brachyspira spp. from chickens. Trampel et al. 32 tested two B. pilosicoli and two B. alvinipulli isolates from chickens in the United States. All four isolates were highly susceptible to lincomycin, carbadox, and tiamulin. Variable minimal inhibitory concentrations (MICs) were seen for chlortetracycline, oxytetracycline, tylosin, bacitracin, erythromycin, neomycin, and penicillin, suggesting that these antimicrobials may be effective against some isolates but not against others. The MIC of streptomycin was consistently high. Hampson et al. 8 tested 25 B. intermedia isolates from chickens (24 Australian isolates and 1 isolate from The Netherlands) and 17 B. pilosicoli isolates from chickens (13 Australian isolates, 3 isolates from The Netherlands, and 1 US isolate). Based on available breakpoint values for Brachyspira hyodysenteriae or other Gram-negative enteric veterinary pathogens, isolates of both species were considered to be generally susceptible to tiamulin, lincomycin, metronidazole, and tetracycline, although one or more strains had elevated MIC ranges for tiamulin, lincomycin, tetracycline, and ampicillin. About half of the B. intermedia and B. pilosicoli isolates had increased MIC ranges for tylosin.
Only a few molecular mechanisms for antimicrobial resistance in Brachyspira spp. have been described. A point mutation at position 2058 in the 23S rRNA gene in B. hyodysenteriae from pigs and at position 2058 or 2059 in the 23S rRNA gene in B. pilosicoli was described as the genetic basis of macrolide and lincosamide resistance.15,20 A point mutation G1058C in the 16S rRNA gene of B. hyodysenteriae was described as the genetic background for decreased susceptibility to doxycycline. 28
The aim of this study was to determine the antimicrobial resistance pattern of recent layer field isolates of B. intermedia from Belgium and The Netherlands. The molecular mechanism of acquired resistance was also studied.
Materials and Methods
Bacterial isolates
Twenty isolates of B. intermedia were obtained from the ceca of chickens of 20 different laying flocks in Belgium (n = 10) and The Netherlands (n = 10) during the period 2008–2010. For primary isolation, samples were cultured on trypticase soy agar (TSA) supplemented with 5% defibrinated sheep blood, 0.1% yeast extract, spectinomycin (400 μg/ml), vancomycin (25 μg/ml), and colistin (25 μg/ml) 12 and incubated for 4 days at 39°C in an anaerobic atmosphere (84% N2, 8% H2, and 8% CO2). Subculturing was done on TSA plates to obtain pure cultures, which were harvested in a peptone–glycerol medium and stored at −80°C. Identification and typing of the isolates were done using Brachyspira genus-specific and species-specific polymerase chain reactions.11,25 The type strains B. intermedia PWS/AT (ATCC 51140) and B. hyodysenteriae B78T (ATCC 27164) were included as control strains for standardization of antimicrobial susceptibility testing. 27
Antimicrobial susceptibility testing
Six antimicrobial agents were used (range of final concentrations in mg/L of culture medium): tylosin (0.016–128), tilmicosin (0.016–128), tiamulin (0.031–4), valnemulin (0.016–2), doxycycline (0.063–8), and lincomycin (0.016–128) (Sigma-Aldrich, Seelze, Germany). The compounds were dissolved and diluted according to the Clinical and Laboratory Standards Institute (CLSI) recommendations to make stock solutions. 1
The broth dilution method as described by Karlsson and Franklin 14 was used, with some modifications. In the present study, fresh antimicrobial solutions were used instead of precoated multiwell plates. Briefly, 100 μl of twofold serial dilutions of the antimicrobial agents in brain heart infusion (BHI) broth (Merck, Darmstadt, Germany) supplemented with 10% fetal calf serum was transferred to wells of 48-well culture plates (Greiner Bio-One, Frickenhausen, Germany). Wells containing 100 μl BHI broth supplemented with 10% fetal calf serum but no antimicrobials were used as growth controls. The panels were prepared and stored at 4°C for less than 24 hr before use.
For preparation of the inocula, bacteria harvested from 4 days incubated TSA plates were suspended in BHI broth supplemented with 10% fetal calf serum. The procedure described by Karlson and Franklin 14 was followed to obtain final inoculum concentrations of 1 × 106 to 5 × 106 CFU/ml. Each well in the panels was filled with 0.4 ml of this inoculum. The panels were incubated in an anaerobic atmosphere (84% N2, 8% H2, and 8% CO2; Hypoxic Workstation, Ruskinn Technology, South Wales, UK) for 4 days on a rotary shaker (60–80 rpm) at 37°C. The MIC was read as the lowest concentration of the antimicrobial agent that prevented visible growth after 4 days of incubation. All growth control wells were checked for pure growth by examining aliquots under a phase-contrast microscope.
Molecular mechanism of resistance
Amplification of a 388-bp fragment of the 23S rDNA (position 1858–2244 according to the Escherichia coli 23S rRNA gene sequence) was accomplished with primers, as described by Karlsson et al. 15 Amplification of a 644-bp fragment of the 16S rDNA (position 800–1443 according to the E. coli 16S rRNA gene sequence) was accomplished with the previously described primers 5′-GTAGTCCACGCCGTAAACG-3′ 13 and 5′-GCTAACGACTTCAGGTAAAAC-3′. 25 The PCR was performed to amplify the fragments by using the following conditions: 15 min at 95°C, 30 cycles of 30 sec denaturation at 94°C, 30 sec of annealing at 55°C, and 1 min of elongation at 72°C, followed by a final extension of 2 min at 72°C. The sequencing reaction was performed using the BigDye Terminator v3.1 Cycle Sequencing Kit (AB Applied Biosystems) and the aforementioned primers. The purified PCR products were sequenced using an ABI Prism™3100 Genetic Analyzer. Contigs were created using the ContigExpress program, included in the Vector NTI Advance™10 (Invitrogen). Sequence alignment was done with the ClustalW2-Multiple Sequence Alignment program (EMBL-EBI).
Results
Antimicrobial susceptibility testing
The MIC results are presented in Table 1.
Brachyspira intermedia isolates considered to have acquired resistance according to the microbiological criterion are represented in bold.
MIC for Brachyspira hyodysenteriae B78T (ATCC 27164).
MIC for B. intermedia PWS/AT (ATCC 51140).
MIC, minimal inhibitory concentration.
For the field isolates, the MIC distribution patterns of tylosin, tilmicosin, lincomycin, and doxycycline were bimodal, demonstrating acquired resistance against doxycycline in three strains (MIC 2 mg/L), against tylosin and tilmicosin in two strains (MIC ≥ 128 mg/L), and against lincomycin in one strain (MIC 8 mg/L). The lincomycin-resistant isolate was also resistant to the macrolides. The MICs of tiamulin and valnemulin showed a monomodal distribution, but with tailing toward the higher MIC values.
Molecular mechanism of resistance
The three doxycycline-resistant strains showed a G1058C mutation in the 16S rRNA gene, whereas all doxycycline-susceptible isolates had the wild-type sequence. The strain that was resistant to tylosin, tilmicosin, and lincomycin had an A2058T mutation in the 23S rRNA gene. No mutations were found at position 2059. The type strains B. intermedia PWS/AT (ATCC 51140) and B. hyodysenteriae B78T (ATCC 27164) showed the wild-type sequences for both fragments.
Discussion
There is no generally accepted or standardized method for antimicrobial susceptibility testing of Brachyspira species. Although agar dilution is a commonly used method for susceptibility testing of B. hyodysenteriae isolates from pigs, a broth dilution method has been described by Karlsson and others.14,17,19 The appearance of hemolysis is used to indicate growth of Brachyspira spp. in the agar dilution method. 21 As the weak β-hemolysis of B. intermedia is much less visible on agar than the strong hemolysis of B. hyodysenteriae, we found a broth dilution method more suitable for the present study. In the present study, the MIC results of tylosin, tiamulin, valnemulin, doxycycline, and lincomycin for the type strain B. hyodysenteriae B78T (ATCC 27164) fit into the quality-control ranges suggested by Pringle et al. 27 The MIC results of tylosin and tiamulin for type strain B. intermedia PWS/AT (ATCC 51140) were comparable to those previously reported. 19
Most of the investigated B. intermedia isolates had very low MICs for tylosin and tilmicosin, but two strains showed acquired resistance to tylosin and tilmicosin. One of these two strains additionally showed resistance to lincomycin, whereas all other strains had low MICs for lincomycin. Karlsson et al.15,20 described a single base mutation at position 2058 or 2059 in the 23S rRNA gene as a genetic basis of macrolide and lincosamide resistance in B. hyodysenteriae and B. pilosicoli strains of pigs. In the present study, the A2058T point mutation in the 23S rRNA gene was found only in the strain that was resistant to both macrolides and lincomycin. The other macrolide-resistant strain had the wild-type sequence at positions 2058–2059. The genetic background of acquired resistance to macrolides but not to lincosamide is not known in Brachyspira spp.
A G1058C mutation in the 16S rRNA gene in B. hyodysenteriae has been described as a genetic background for decreased susceptibility to doxycycline. 28 All three B. intermedia strains resistant to doxycycline had this G1058C mutation, whereas it was absent in all susceptible strains.
Several authors from different countries reported on pleuromutilin resistance of B. hyodysenteriae and B. pilosicoli from pigs with a parallel decreased susceptibility to tiamulin and valnemulin.6,10,16,18,22,23,26,29,34 In this study, the MICs of tiamulin and valnemulin showed a monomodal distribution, but with tailing toward the higher MIC values, possibly suggesting a low-level acquired resistance to both pleuromutilins in 6 B. intermedia isolates. Analysis of a larger collection of B. intermedia field strains is necessary to more efficiently determine pleuromutilin wild-type distributions of this microorganism, which may allow defining epidemiological cutoff values.
There are no accepted clinical breakpoints for Brachyspira species in chickens. 1 In the present study, the epidemiological cutoff value or microbiological criterion was used for interpretation of MIC results. 33 Using this criterion, the interpretation was obvious for tilmicosin, tylosin, lincomycin, and doxycycline, because MIC distributions of these antimicrobials were bimodal, indicating acquired resistance in isolates in the higher range of MIC values. For the pleuromutilins, MICs rather showed an extended frequency distribution range wherein the division between isolates with or without acquired resistance was more difficult to establish. The microbiological criterion refers to direct in vitro interactions between the antimicrobial agents and the B. intermedia isolates and is not necessarily linked to therapeutic success. However, for lincomycin and doxycycline, MIC values were at least 8 times higher for isolates with acquired resistance and were up to 100 times higher for the macrolides. The likelihood that chickens infected with these isolates will respond less well to treatment with these antimicrobials is high.
In conclusion, acquired resistance to tilmicosin, tylosin, lincomycin, and doxycycline was demonstrated in B. intermedia field strains from Belgian and Dutch layers. An A2058T mutation in the 23S rRNA gene was found as the genetic base for macrolide and lincosamide resistance and a G1058C mutation in the 16S rRNA gene as the genetic base for doxycycline resistance.
Footnotes
Disclosure Statement
No competing financial interests exist.