ABSTRACT

Streptococcus suis is an emerging zoonotic pathogen. With the lack of an effective vaccine, antibiotics remain the main tool to fight infections caused by this pathogen. We have previously observed a reserpine-sensitive fluoroquinolone (FQ) efflux phenotype in this species. Here, SatAB and SmrA, two pumps belonging to the ATP binding cassette (ABC) and the major facilitator superfamily (MFS), respectively, have been analyzed in the fluoroquinolone-resistant clinical isolate BB1013. Genes encoding these pumps were overexpressed either constitutively or in the presence of ciprofloxacin in this strain. These genes could not be cloned in plasmids in Escherichia coli despite strong expression repression. Finally, site-directed insertion of smrA and satAB in the amy locus of the Bacillus subtilis chromosome using ligated PCR amplicons allowed for the functional expression and study of both pumps. Results showed that SatAB is a narrow-spectrum fluoroquinolone exporter (norfloxacin and ciprofloxacin), susceptible to reserpine, whereas SmrA was not involved in fluoroquinolone resistance. Chromosomal integration in Bacillus is a novel method for studying efflux pumps from Gram-positive bacteria, which enabled us to demonstrate the possible role of SatAB, and not SmrA, in fluoroquinolone efflux in S. suis.

INTRODUCTION

Streptococcus suis is a worldwide-distributed zoonotic pathogen affecting pigs and humans (20). In industrialized countries, it is mainly regarded as an important economic burden to the pig industry that causes infections only sporadically in humans. Infection with S. suis is therefore defined as an occupational zoonosis, with the main risk groups being people in close contact with pigs and pork. Nevertheless, cases of S. suis meningitis have also been reported in patients with no history of contact with live pigs or their products, and recent data suggest that S. suis may be a food-borne disease (20, 26, 30). In several developing countries in Southeast Asia, S. suis is one of the main causes of meningitis in humans, even though its prevalence is probably still underestimated (18, 21, 24, 40). Several outbreaks of S. suis have occurred in China (47). The largest outbreak took place during the summer of 2005 in Sichuan Province and affected 204 people, causing 38 fatalities. Mortality was high as a result of the onset of a streptococcal toxic shock syndrome (STSS) similar to that produced by group A streptococci (GAS) (42). Consequently, concern about this pathogen has risen in these areas, and measures to prevent or hinder its impact are already being taken (18).
Antimicrobial resistance is one of the most concerning phenomena in infectious diseases. In S. suis, resistance to neomycin, erythromycin, spiramycin, lincomycin, trimethoprim-sulfamethoxazole, and tetracycline antibiotics is a common feature (33, 36, 46). Interestingly, comparison of the genomes of STSS-producing strains has led to the discovery of ICESsuBM4072, a putative mobile pathogenicity island bearing the tetracycline resistance determinants TetL and TetO, erythromycin ribosome methylase ErmB, a chloramphenicol acetyltransferase, and a dihydrofolate reductase (17). The role played by these genetic structures must not be overlooked, as it has recently been proven that genomic islands are prevalent in the genome of S. suis (48). Treatment of infections caused by S. suis remain, in general, effective by administration of beta-lactam antibiotics. Nevertheless, strains showing resistance to this family of drugs have been reported (14, 36). Fluoroquinolones (FQs) are the main alternative to beta-lactam antibiotics for treatment of streptococcal infections (9). We have reported increasing FQ resistance among clinical S. suis isolates in Spain (8). Substitutions in the quinolone resistance-determining regions (QRDR) of the genes encoding the antibiotic targets, gyrA and parC, were identified (8). Interestingly, the FQ resistance phenotype could be partially reduced by the efflux pump inhibitor reserpine, indicating that efflux may be involved in FQ resistance in S. suis. Several efflux pumps conferring resistance to fluoroquinolones have been described in other Gram-positive bacteria. Mainly, members of two families of transporters are involved: (i) proteins belonging to the major facilitator superfamily (MFS), such as Bmr of Bacillus subtilis (29), NorA of Staphylococcus aureus (43), Lde from Listeria monocytogenes (13), and PmrA in Streptococcus pneumoniae (12); and (ii) exporters belonging to the vast ATP binding cassette (ABC) family, such as PatA and PatB from S. pneumoniae (28) or LmrA from Lactococcus lactis (44). Efflux-mediated drug resistance is normally a consequence of overexpression of the genes encoding the pumps, either through mutations in regulators or in promoter regions thereof (31). Amino acid substitutions can also lead to resistance through a change in the substrate recognition of the pump (32, 34).
To date, no FQ efflux pumps have been described in S. suis. Here we report the possible role of SatAB, an ABC transporter homologous to PatA and PatB, in resistance to ciprofloxacin and norfloxacin in S. suis. Furthermore, intensive study of SmrA, a major facilitator superfamily transporter analogous to the pneumococcal FQ efflux pump PmrA, allows one to neglect its role in FQ resistance. Study of these pumps was accomplished with a knock-in strategy, integrating the genes encoding the pumps in monocopy in the chromosome of Bacillus subtilis 168 under the control of an inducible promoter. This system has proven to be a promising tool for the study of efflux pumps in Gram-positive bacteria.

MATERIALS AND METHODS

Bacterial strains, culture conditions, and susceptibility testing.

The characteristics of the strains and plasmids used in this study are listed in Table 1 . B. subtilis 168 and pDR67 were kindly donated by Tarek Msadek from Institut Pasteur in Paris.
Table 1.
Table 1. Strains and plasmids used in this study
Strain or plasmidRelevant genotypeDescriptionSource or reference
Strains   
    S. suis   
        BB1001Wild typeFluoroquinolone susceptible8
        BB1002Wild typeFluoroquinolone susceptible8
        BB1013Wild typeFluoroquinolone-resistant efflux8
    B. subtilis   
        BS168Reference strain used as recipient strainParental strainCourtesy T. Msadek
        BB1066amyE::Pspac lacI ΔlacZ catControl strainThis work
        BB1067amyE::Pspac smrABB1013 lacI ΔlacZ cat This work
        BB1068amyE::Pspac satABBB1001 lacI ΔlacZ catInducible resistance to NOR and CIPThis work
        BB1069amyE::Pspac satABBB1002 lacI ΔlacZ catInducible resistance to NOR and CIPThis work
        BB1070amyE::Pspac satABBB1013 lacI ΔlacZ catInducible resistance to NOR and CIPThis work
        BB1071amyE::Pspac ΔsatABBB1002 lacI ΔlacZ cat This work
    S. pneumoniae   
        R6Reference strain Rockefeller Institute
        BB1072R6/pB1011Control strainThis work
        BB1073R6/pB1012Strain bearing smrABB1013This work
Plasmids   
    pJS3catReplicative in Streptococcus spp.3
    pJS3EBEcoRI and BamHI restriction sites This work
    pB1011pJS3EB::Pspac lacIControl plasmidThis work
    pB1012pJS3EB::Pspac smrABB1013 lacIPlasmid bearing smrABB1013This work
S. suis and S. pneumoniae strains were cultured on Columbia sheep blood agar plates, Todd-Hewitt broth, and a casein hydrolysate-based medium with 0.3% sucrose (AGCH) as an energy source. B. subtilis 168 was cultured on LB agar or in LB broth. Media were supplemented with chloramphenicol to final concentrations of 5 μg/ml and 2.5 μg/ml for B. subtilis and S. pneumoniae transformants, respectively. Culture media were obtained from Oxoid (Oxoid Ltd., Basingstoke, United Kingdom) and bioMérieux (France). Chemicals, additives, and antibiotics were supplied by Merck (Merck KGaA, Darmstadt, Germany) and Sigma-Aldrich (Sigma Chemical Co., St. Louis, Mo). IPTG (isopropyl-β-d-thiogalactopyranoside) was purchased from Invitrogen (Carlsbad, CA).
Susceptibility tests were carried out by following CLSI guidelines (6). Efflux phenotype visualization was performed by adding 40 μg of ciprofloxacin, enrofloxacin, and levofloxacin and 100 μg of norfloxacin in a 10-μl volume to nonimpregnated discs that were used for diffusion antibiograms on standard media with and without 10 μg/ml of reserpine dissolved in acetone. Nonimpregnated and antibiotic discs were obtained from bioMérieux (France) and Oxoid (Oxoid Ltd., Basingstoke, United Kingdom). For sugar assays, API 50CH was obtained from bioMérieux and used according to the manufacturer's instructions.

DNA analysis and manipulation.

Primers used in this study were supplied by Roche (Germany) and Sigma-Genosys (Sigma Chemical Co., St. Louis, MO). A list of the primers used can be found in Table 2. PCRs were performed using Taq polymerase from Biotools (B&M Labs, Spain). Phusion polymerase (Finnzymes, Woburn, MA) was used for the obtention of long amplicons. pCR2.1 and pBAD plasmids and the Escherichia coli IncF′ (Top10) strain were purchased from Invitrogen (Carlsbad, CA). Plasmid DNA extraction and purification of PCR fragments were performed as previously described (37). For plasmid extraction from S. pneumoniae and S. suis, a modified first step with sodium deoxycholate was used as previously described (39). Automated sequencing was carried out at Secugen S. L. (Madrid, Spain). Sequence analysis was performed using DNA Strider 1.4f13 (CEA, France), 4Peaks 1.6 (Mekentosj, Netherlands), CLC DNA Workbench software (Denmark), ClustalW, NEBcutter (45), and NIH online analysis tools (http://www.ncbi.nlm.nih.gov). Promoter sequence analysis was performed with Bprom (Softberry, Inc., Mount Kisco, NY). Protein modeling was done using Phyre server (23) and illustrated using the PyMOL molecular graphics system (version 1.3; Schrödinger, LLC).
Table 2.
Table 2. List of primers used in this study
PrimerSequence 5′ → 3′aReference
Amplification and sequencing  
    satAFGTTGAGAACTTGTCCTAGGGThis work
    satBRTGACCAGTTCGAATCCACGGThis work
    smrAFATGGCTGCTCAGCTTTCTTTThis work
    smrARAAACTAAAAGACTGTATTTTGThis work
Operon determination  
    satAIntFCTGGTTTTGACAGAGAAGGGThis work
    satBIntRCCGCAATGGCATTTCCAAGGThis work
Cloning in pDR67 and insert checking  
    satAXbaIRBSGCGCTCTAGAGGTTAAACAGGTGGGCAATCThis work
    satBBglIIStopGCGCAGATCTACTTTATTCAAACACAAACTThis work
    smrAXbaIRBSGCTCTAGATTTTGGAGGAATTAAAGGATThis work
    smrABglIIRGCAGATCTTCACTACACATCCCTTACTTThis work
    pDR67FACATCCAGAACAACCTCTGCThis work
    pDR67RCTCGTTTCCACCGAATTAGCThis work
Mobilization to pJS3 and insert checking  
    pJS3EcoRIGAACCGAATTCTCCTTTTTCGCTTCThis work
    pJS3BamHIATATGGATCCGGAGCTGTAATATAAAAACThis work
    pDR67EcoRIATGAGAATTCTACACAGCCCThis work
    pDR67BamHICAGTGCAGGGATCCTAACTCThis work
    pJS3InsFAATGTCACTAACCTGCCCCGThis work
    pJS3InsRTGCCAAAAAGCTTCTGTAGGThis work
qRT-PCR  
    smrARTFAAGCAGAATTTGAAGGTGThis work
    smrARTRAAGGGCATTAACAGATACCGThis work
    satABRTFAATCCAGAACCTTGTCATThis work
    satABRTRAATAATCATCCACCAGAGTThis work
    rpoBsuisFAACTGGCGAGATCAAGACThis work
    rpoBsuisRAACGATGATACGCTCTGCThis work
    rpoBBSFGGATGGCTACAACTATGAGThis work
    rpoBBSRGCTTCTGATTCGTATTCTTCThis work
    rpoB428CGGTTGGTGAATTGCTTGCCAACC4
    rpoB468RACTGCAGCTGTTACAGGACGG4
a
Restriction sites are underlined.

RNA extraction and qRT-PCR.

RNA from all strains was obtained from liquid cultures in early exponential phase (optical density at 600 nm [OD600], 0.3). When ciprofloxacin was added to the culture medium, the concentrations used were one-fourth of the MIC of each strain. RNA extractions were performed twice for each strain and condition using the RNeasy kit (Qiagen, Inc., Chatsworth, CA). Digestions with DNase were performed by following the manufacturer's instructions. The absence of DNA was checked through PCR. Several digestions were performed until the PCR results obtained were negative. cDNA was obtained using SuperScript III reverse transcriptase (Invitrogen, Carlsbad, CA) and 6-bp random primers. Presence of cDNA was confirmed with conventional PCR. Quantitative reverse transcriptase PCR (qRT-PCR) primers (Table 2) were designed using AlleleID 7.7 software (Premier Biosoft, Palo Alto, CA). qRT-PCR was carried out in triplicate using a MyiQ thermal cycler (Bio-Rad Laboratories, Inc., Spain). Each condition and strain were tested three times. No primer dimer was obtained in any of the PCRs. The efficiency of all assays was 100% ± 5%. Results for the expression of smrA and satAB were normalized using rpoB. Primers rpoB428 and rpoB468R, targeting the rpoB gene of S. pneumoniae, have been previously described (4).

Cloning of smrA and satAB in the B. subtilis chromosome.

In order to clone smrA and satAB in pDR67, XbaI and BglII restriction sites were used. Amplification of the open reading frames (ORFs), including their putative ribosome binding sites (RBSs), was carried out using primers designed to bear both restriction sites (Table 2). Plasmids and inserts were digested using XbaI and BglII (Takara Bio, Inc., Japan) and ligated using T4 DNA ligase (New England BioLabs, Beverley, MA). The ligation product was directly transformed in B. subtilis using a two-step transformation procedure as previously described (15). Colonies were phenotypically tested for pDR67 integration in the chromosome. Briefly, amyE disruption leads to the loss of alpha-amylase activity. This was demonstrated by growth on 1% potato starch and staining with iodine pearls. Clones not showing a light halo around the colonies were selected, for their amylase activity was lost, presumably because of pDR67 integration. These observations were confirmed by PCR and sequencing.

Mobilization of the expression system bearing smrA to pJS3.

Streptococcal plasmid pJS3 was amplified by PCR with primers bearing EcoRI and BamHI restriction sites located close together and facing outwards (Table 2). The product was phosphorylated using polynucleotide kinase, ligated using T4 DNA ligase, and transformed in S. pneumoniae R6 as previously described (25). The resulting plasmid was called pJS3EB and bore both restriction sites plus a new unique NdeI site between the restriction sites. Pspac-lacI and Pspac-smrA-lacI from strains BB1066 and BB1067, respectively, were amplified using primers with EcoRI and BamHI restriction sites (Table 2). Both inserts and pJS3EB were digested using EcoRI and BamHI and ligated with T4 DNA ligase. All enzymes were purchased from New England BioLabs (Beverley, MA).

Influence of pH in the activity of SmrA.

LB was used as the growth medium, and pH was adjusted to 5, 6, 7, and 8 adding HCl or NaOH after autoclaving. Inocula were adjusted to a 0.5 McFarland standard and diluted 1:200 in the growth medium. IPTG was added to a final concentration of 1 mM. Ciprofloxacin and norfloxacin covering a wide range of concentrations, from subinhibitory to inhibitory (0.02 to 0.3 μg/ml for ciprofloxacin and 0.06 to 0.8 μg/ml for norfloxacin), were tested. Growth curves were performed in a Tecan Infinite 200 apparatus (Tecan Group Ltd., Männedorf, Switzerland). Twenty-four-well plates were incubated at 37°C, and OD620 measurements were taken every 10 min after 15 s of shaking with a 2.5-mm amplitude.

Phenotype MicroArrays.

Phenotype MicroArrays (PMs; Biolog, Hayward, CA) were performed at the Animal Health and Veterinary Laboratories Agency, Weybridge, United Kingdom. Strains were cultured on agar plates under selective pressure as described above. Cell suspensions were prepared as previously described (2) to directly obtain an 81% transmittance (T). All inoculating fluids were purchased from Biolog, Inc. Inocula were prepared by following the manufacturer's instructions, and IPTG was added to a final concentration of 1 mM for B. subtilis and 0.5 mM for S. pneumoniae. Microarray plates were incubated at 37°C, and colorimetric readings were performed at 15-min intervals over a 48-h period using the OmniLog reader. Kinetic data were analyzed using OmniLog PM software (Biolog). Phenotype MicroArrays used for both B. subtilis and S. pneumoniae strains included PM1 and PM2A for carbon pathways and PM11A to PM20B for sensitivity to 240 chemicals at 4 different concentrations. Furthermore, S. pneumoniae strains were also tested against panels PM9 and PM10 for ion/osmotic effects and pH, respectively. All assays were performed on at least two separate occasions.

Nucleotide sequence accession numbers.

The nucleotide sequences of this study have been deposited in GenBank under the following accession numbers: smrA, including the promoter region, in BB1001, JF416696; in BB1002, JF416694; and in BB1013, JF416698; and satRAB in BB1001, JF416697; in BB1002, JF416695; and in BB1013, JF416699.

RESULTS

Efflux is involved in FQ resistance in S. suis.

In our previous work, S. suis isolates of animal origin showing high-level FQ resistance were studied (8). These isolates carried resistance mutations in the quinolone resistance-determining regions (QRDR) of gyrA and parC, genes encoding the antibiotic targets. In order to assess the involvement of drug efflux in the FQ resistance phenotype, ciprofloxacin MICs were determined with and without the efflux pump inhibitor reserpine. In all but one resistant isolates, resistance levels were 2- to 4-fold lower in the presence of reserpine, indicating that efflux was involved in FQ resistance in these strains. No decrease in the MIC was observed in any of the FQ-susceptible strains in the presence of reserpine. To unequivocally prove these results, a diffusion antibiogram method that enabled a clear visualization of the efflux phenotype was developed. Custom-made antibiogram discs, bearing high concentrations of various fluoroquinolones, were used on standard medium with and without reserpine against BB1013, an isolate with ciprofloxacin MICs of 64 and 16 μg/ml in the absence and presence of reserpine, respectively, known to also carry ParC Ser79Tyr and GyrA Ser81Lys changes involved in FQ resistance (8). Clear growth inhibition zones were obtained by ciprofloxacin and norfloxacin only when reserpine was added, showing the existence of reserpine-sensitive FQ efflux. However, resistance to enrofloxacin and levofloxacin was unaffected by reserpine (Fig. 1).
Fig. 1.
Fig. 1. Diffusion antibiograms of BB1013 in medium with (A) and without (B) 10 μg/ml of reserpine. CIP, ciprofloxacin; NOR, norfloxacin; LEV, levofloxacin; ENR, enrofloxacin. Numbers indicate the micrograms of antibiotic in the disc.

Identification of smrA and satAB.

Three FQ efflux pumps have been widely studied in S. pneumoniae, a close member within the genus. PmrA, first described as a norfloxacin extruder (12), is an MFS transporter, whereas PatA and PatB are ABC transporters involved in ciprofloxacin and norfloxacin efflux among other compounds (28). We searched for homologues of these pumps in the available S. suis genomes in order to study their possible implication in FQ resistance in this species.

(i) smrA.

An ORF encoding a predicted protein showing 58% identity with PmrA was found in the genome of S. suis and named smrA. The genetic environment of smrA was analogous to that of pmrA (Fig. 2 ). Protein modeling of the 401-amino-acid-long product, SmrA, revealed a 12-transmembrane-segment structure found in most MFS transporters (Fig. 3). The complete gene sequence with the corresponding promoter region was obtained from resistant strain BB1013 and from the two susceptible, genetically unrelated, clinical isolates BB1001 and BB1002 (8). SmrA of BB1013 presented two residue substitutions (Thr107Ala and Ile126Val) in comparison to the susceptible alleles, which were identical to each other (Fig. 3). The promoter region of smrA presented a canonical −35 box (TTGACAA) and ribosome binding site (RBS) (GGAGG) at positions −66 and −14 (Fig. 2). No −10 box was identified. Interestingly, between the −35 box and the region where the −10 box should be located, a 7-bp inverted repeat, separated by 5 bp, was found. When any 2 of the 5 bp are erased in silico, a sequence resembling a −10 box is found immediately after the 3′ repeat, suggesting that smrA may be regulated the way in which bmr is regulated (16). Briefly, expression of bmr is controlled by BmrR, which binds to a dyad symmetry region in its promoter sequence and brings the −10 box 2 bp closer to the −35 box, to an optimal 17-bp distance. Interestingly, a mutation in the 3′ inverted repeat was found in the sequence of the resistant strain, suggesting a possible deregulation of smrA in BB1013. Further experiments were focused on elucidating whether overexpression of smrA and/or the amino acid substitutions found in SmrABB1013 may be involved in fluoroquinolone resistance in BB1013.
Fig. 2.
Fig. 2. (A) Genetic organization of smrA (left) and satAB (right) in S. suis strain ST3 (upper lane) compared to their homologs in S. pneumoniae R6 (lower lane, hatched arrows). Open reading frames are shown by arrows, with the direction of transcription indicated by the arrowhead. 1, Formamidopyrimidine-DNA glycosylase; 2, dephospho-coenzyme A kinase; 3, 50S ribosomal protein L33; 4, preprotein translocase subunit SecG; 5, exoribonuclease R; 6, UTP-glucose-1-phosphate uridylyltransferase; 7, glycerol-3-phosphate dehydrogenase; 8, satR; 9, dUTP nucleotidohydrolase; 10, phosphoglycerate mutase; 11, DNA repair protein RadA; 12, arginyl tRNA synthase (argS) repressor; 13, DNA mismatch repair protein MutS; 14, degenerate transposase; 15, glutamine amidotransferase; *, hypothetical protein; and 16, glucose-6-phosphate isomerase. Numbers between the ORFs represent the percentages of identity between their products. (B) Alignments of the promoter region of smrA and SatA and SatB from BB1001, BB1002, and BB1013. smrA, canonical −35 box and RBS are highlighted with black boxes. Two 7-bp inverted repeats (arrows) were found downstream the −35 box, separated by a 5-bp spacing region. In silico studies using Softberry Bprom software and the 133 bp immediately upstream from smrA showed that the loss of 2 base pairs in the spacer region led to the recognition of a previously undetected −10 box immediately downstream from the 3′ repeat (gray box). BB1013 bore three mutations compared to FQ-susceptible strains. G-43A is located in the 3′ inverted repeat and could interfere with regulation of smrA expression. SatA and SatB, alignment of the amino acid sequences of SatA and SatB from the three strains. Conserved residues are represented in black over a white background. Light gray and dark gray backgrounds represent one and two differences among sequences, respectively.
Fig. 3.
Fig. 3. Protein modeling of SmrA (A and B), SatA (C), and SatB (D). (A) Lateral view of SmrA showing a 12-transmembrane-segment structure typical of the major facilitator superfamily of transporters. Arrows show the amino acid substitutions present in BB1013 compared to BB1001 and BB1002. Red- and yellow-stained residues are T107A and I126V substitutions, respectively. (B) View down the channel of SmrA. (C and D) Modeling of SatA and SatB. A six-transmembrane-segment structure followed by a nucleotide binding domain is observed in each protein.

(ii) satA and satB.

Two homologues of the pneumococcal FQ efflux pumps PatA and PatB, showing 66% and 67% identity, were identified in S. suis genomes and named SatA and SatB. Protein modeling revealed an N-terminal transmembrane domain and a C-terminal nucleotide binding domain structure characteristic of ABC transporters (Fig. 3). Interestingly, in contrast to the pneumococcal organization of the ORFs, in which these genes seem to be independently cotranscribed, satA and satB showed in S. suis a classical operon conformation, with both genes immediately contiguous on the same strand and reading frame (Fig. 2). Furthermore, upstream from satA and partly overlapping it, an ORF coding for a regulator of the MarR family was found and named satR. To determine if satA and satB were cotranscribed on the same mRNA, a PCR from one gene to the other, using primers satAIntF and satBIntR (Table 2), was performed with cDNA from BB1001 and BB1013. A 2-kb fragment was amplified from the cDNA and the control DNA of the two strains, proving the suspected operon conformation of both ORFs (data not shown). This organization points to the function of SatA and SatB being intimately related and is in accordance with the suggestion by others of a heterodimeric SatAB transporter (11). Primers satAF and satBR, amplifying the whole operon and its regulator, were designed using the available S. suis genomes in GenBank. The sequences of satR and satAB of BB1001, BB1002, and BB1013 were obtained by primer walking and analyzed. Several amino acid differences, located mainly in the transmembrane segments, were found among the three alleles of both genes (Fig. 2). Interestingly, sequence analysis of the putative regulator SatR revealed in BB1013 a C238T mutation leading to a premature stop codon (CAA80TAA) in comparison to BB1001, BB1002, and GenBank sequences.

Expression analysis.

Efflux pumps can lead to antimicrobial resistance through overexpression. In order to assess if smrA and satAB were overexpressed in BB1013, the levels of mRNA of these pumps were measured by means of qRT-PCR using BB1001 as a control (Fig. 4).
Fig. 4.
Fig. 4. Levels of expression of smrA and satAB in BB1013 compared to BB1001. Hatched bars represent growth on subinhibitory concentrations of ciprofloxacin (one-fourth of the MIC of each strain). Pump expression was normalized against rpoB expression levels.
Expression levels of smrA were not significantly different between BB1013 and the control strain in the absence of fluoroquinolones. When subinhibitory concentrations of ciprofloxacin were added to the culture medium to a final concentration of one-fourth of the MIC (0.12 μg/ml for BB1001 and 16 μg/ml for BB1013), levels of expression of smrA increased in BB1013, leading to a significant 4-fold difference between the two strains.
Expression of satAB was found to be about 15 times higher in BB1013 than in the control strain in the absence of ciprofloxacin. Interestingly, expression was not enhanced but mildly repressed by the presence of ciprofloxacin in the culture medium.
Expression analysis showed that both pumps were overexpressed in BB1013, either under induction by ciprofloxacin, as is the case of SmrA, or constitutively, as for SatAB. Moreover, these data suggest that SatR acts as a repressor of satAB in S. suis.

Cloning and expression of smrA and satAB in B. subtilis.

Cloning of satAB and smrA in E. coli, as performed successfully with other pumps such as NorA or Bmr (29, 43), was investigated. Despite repeated attempts, the genes encoding the putative pumps could not be cloned in the pCR2.1 plasmid and transformed into an E. coli IncF′ strain. Thus, cloning under the control of the pBAD promoter was tested. Although repression was enhanced by the addition of glucose-6-phosphate (G6P) and fucose, cloning of S. suis efflux pumps was unsuccessful in all cases, suggesting a high toxicity of the insert in a Gram-negative environment. Thus, cloning in B. subtilis was explored as an alternative strategy.
pDR67 (22) is a plasmid capable of replicating in E. coli and integrating in the Bacillus subtilis chromosome. It is designed for cloning inserts under the control of the IPTG-inducible promoter Pspac. Expression is tightly repressed by the presence, downstream from the insert, of the promoter's operator, lacI. Furthermore, repression can be strengthened using lacIq E. coli strains (such as XL1-Blue). pDR67 bears two homology regions with the nonessential amyE chromosomal gene of B. subtilis, as well as a chloramphenicol marker suitable for selection in this species. It can therefore be used for mobilizing constructions obtained in E. coli to a Gram-positive environment as a single copy in the chromosome. IPTG induction is also feasible in this bacterium.
smrA and satAB were cloned with their corresponding RBS in pDR67 and transformed in E. coli XL1-Blue. Despite the high levels of repression of the insert, transformation efficiency was low (data not shown). Furthermore, growth of transformants was evidently impaired and, when these colonies were streaked onto new plates, various morphologies were recognizable. These clones did not seem to be suitable and were not used for further study. Interestingly, for each pump, one colony was found to be positive to the insert by PCR but did not display toxicity signs. Sequencing revealed premature stop codons in both inserts, suggesting the nonfunctionality of the pumps as the reason for the absence of toxicity (satA CAG→TAG in the case of satAB operon and a deletion of A1008 in smrA leading to a frameshift altering the protein sequence of the following 15 amino acids and a CTA→TAG premature stop codon). These strains were used as controls when needed. Altogether, our data suggest that a cloning procedure, including a passage of smrA and satAB through E. coli, is unfeasible. Thus, the ligation products of satAB from the three strains (BB1001, BB1002, and BB1013) and smrA from BB1013 with pDR67 were directly transformed into B. subtilis 168 (Fig. 5). Strains BB1068, B1069, and BB1070, bearing the three different alleles of satAB, and BB1067, bearing smrA from BB1013, were obtained (Table 1). Sequencing confirmed that the inserts harbored no mutations affecting the protein sequence. A transformant with the complete cloning system but no insert, namely BB1066, was obtained and conserved as a control. Last, plasmid pDR67 bearing a nonfunctional copy of satABBB1002 obtained in E. coli was also transformed in B. subtilis, giving rise to BB1071, a strain bearing the whole operon but with an interrupted satA gene.
Fig. 5.
Fig. 5. (A) Schematic representation of smrA and satAB integration in B. subtilis. Open reading frames are shown as arrows, with the direction of transcription indicated by the arrowhead. pDR67 bears regions of homology to the nonessential amyE locus of B. subtilis. These gene fragments are shown as broken arrows. Homologous recombination thereof with the chromosome allows for pDR67 integration. smrA and satAB amplicons were ligated to pDR67 and transformed directly in B. subtilis. (B) Phenotypic screening of pDR67 integration in the chromosome. Bacteria were grown on LB containing 1% potato starch. Upper lane, B. subtilis 168 with a functional copy of amyE. Degradation of starch appears as a light unstained halo. Lower lane, BB1066, with pDR67 disrupting amyE, has no amylase activity. Starch is stained, and no light zone is visible.

satAB confers resistance to fluoroquinolones, while smrA does not. (i) smrA.

Strain BB1067, bearing smrABB1013, was subjected to antimicrobial susceptibility testing using a wide variety of compounds. Strain BB1066 was used as a control. No differences were observed for any of the antibiotics and toxics tested. Primer walking confirmed the correct sequence of the insert. RT-PCR was also performed to confirm the induction of the system in this strain (data not shown). As MFS transporters obtain their energy from the proton gradient between the periplasm and the cytoplasm, we investigated whether pH of the medium could play a role in the activity of SmrA against fluoroquinolones. Growth curve analyses in LB medium at pH 5, 6, 7, and 8 were performed. Several norfloxacin and ciprofloxacin concentrations, ranging from subinhibitory to inhibitory, were tested in all pHs in the presence of 1 mM IPTG. No significant difference could be observed when comparing BB1066 and BB1067 under any condition (data not shown). Therefore, our data suggest that smrA is not involved in fluoroquinolone resistance in S. suis. MFS transporters have also been described as sugar transporters (27). To test whether smrA could be involved in carbohydrate import or export, API 50CH strips were inoculated with BB1067 with and without induction and compared to BB1066. No differences were observed among strains or conditions, suggesting that SmrA is not a carbohydrate pump either. In order to determine the natural substrate of smrA, our strains were tested using Biolog Phenotype MicroArrays (PMs). This high-throughput technology facilitated the rapid testing of a wide variety of compounds. A panel of 240 antibiotics and toxic molecules at four different concentrations were tested on BB1067 under induction conditions, using BB1066 as the control strain. Strikingly, of the entire set of molecules tested against it, BB1067 only showed a slight improvement of respiration with one of the dilutions of minocycline, a drug belonging to the family of tetracyclines (data not shown). Interestingly, minocycline is generally not considered to be a substrate of MFS pumps, except in the case of TetB (7). A Biolog panel of carbon pathways was also tested, confirming previous results on sugar usage and adding other compounds to SmrA's list of nonsubstrate molecules.
In order to rule out the possibility that SmrA is not active in B. subtilis, the construction from Pspac to lacI (see Fig. 5) was mobilized to Streptococcus pneumoniae R6, a close member of the genus. pJS3, a plasmid derived from streptococcal native plasmid pMV158 (3), was modified to bear EcoRI and BamHI restriction sites and named pJS3EB. The constructions in the chromosome of B. subtilis were amplified by PCR and cloned in pJS3EB using the new restriction sites, giving rise to plasmids pB1011, containing the empty system (Pspac-lacI) as in BB1066, and pB1012, with the system and smrABB1013 (Pspac-smrA-lacI) as in BB1067. Strains BB1072 and BB1073 (Table 1) were obtained by transformation of pB1011 and pB1012, respectively, into S. pneumoniae R6 (see Materials and Methods). The sequence of the insert was checked by PCR and sequencing. RT-PCR was performed in order to confirm that Pspac was inducible by IPTG in S. pneumoniae. Strains BB1072 and BB1073 were subjected to a Biolog PM panel of antibiotics, carbohydrates, and ion/osmotic effects, and the results confirmed previous observations on minocycline, proving B. subtilis to be a suitable model for the study of streptococcal MFS pumps. However, the natural substrate of this pump remains elusive.

(ii) satAB.

Antimicrobial susceptibility testing to a wide variety of antimicrobials of strains bearing the three alleles of satAB was performed (data not shown). When IPTG was added to the culture medium to a final concentration of 1 mM, strains bearing a functional copy of satAB showed a decrease in susceptibility to norfloxacin and ciprofloxacin but not to levofloxacin, enrofloxacin, moxifloxacin, or nalidixic acid (Table 3). In order to quantitate the effect of SatAB, MICs to ciprofloxacin and norfloxacin were determined with and without IPTG induction. All alleles conferred at least a 4-fold decrease in susceptibility to both antibiotics. Interestingly, the allele originating from BB1002 conferred an extra doubling dilution increase in resistance (up to 8-fold) against norfloxacin and the allele from BB1013 did so to both norfloxacin and ciprofloxacin. These observations point to the diversity in amino acid composition of the three alleles of SatAB as a possible factor in the levels of resistance observed through modifications in the affinity for the drug. The addition of reserpine to the medium reversed the resistance phenotype. No differences were observed between strain BB1066, bearing the expression system but no insert, and BB1071, bearing ΔsatABBB1002, confirming the nonfunctionality of the mutated insert. Fluoroquinolone efflux was detected only in the presence of IPTG, proving the system's tight regulation (Table 2). No other fluoroquinolone tested, or any of the antibiotics belonging to a variety of other families that were tested (data not shown), was a substrate of SatAB. SatAB is therefore a narrow-spectrum pump of norfloxacin and ciprofloxacin.
Table 3.
Table 3. MIC of strains from this study against several fluoroquinolones
StrainRelevant genotypeMIC (μg/ml) of antibiotic on indicated mediuma
CiprofloxacinNorfloxacinEnrofloxacinLevofloxacin
IIRIIRIIRIIR
BB1066No insert0.060.060.060.250.250.120.060.060.060.120.120.12
BB1067smrABB10130.060.060.060.250.250.120.060.060.060.120.120.12
BB1068satABBB10010.060.250.120.2510.50.060.060.060.120.120.12
BB1069satABBB10020.060.250.120.520.50.060.060.060.120.120.12
BB1070satABBB10130.060.50.120.2520.50.060.060.060.120.120.12
BB1071ΔsatABBB10020.060.060.060.250.250.120.060.060.060.120.120.12
a
—, standard medium; I, 1 mM IPTG added; IR, 1 mM IPTG and 10 μg/ml of reserpine added.

DISCUSSION

Antimicrobial resistance is a major threat in the treatment of infectious diseases. Efflux is a worrying mechanism of resistance, as efflux pumps are proteins involved in many basic life functions, such as osmoregulation or detoxification, and are hence ubiquitous in bacteria. Furthermore, efflux pumps often confer resistance to a wide array of unrelated compounds, allowing for a one-step selection of multiresistant clones as well as for an undesirable coselection phenomenon.
Here we present SatAB, an efflux pump involved in fluoroquinolone resistance in S. suis. We also assessed the involvement of SmrA, an MFS transporter homologous to FQ efflux pump PmrA, in this phenotype. Our data show that SatAB but not SmrA is involved in fluoroquinolone efflux in S. suis.
Expression analysis of both pumps has been performed in clinical isolate BB1013. In the case of satAB, overexpression was constitutive, probably as a consequence of the loss of function of SatR due to a premature stop codon in its coding sequence. The expression of smrA was not significantly different from that of susceptible strain BB1001 in the absence of antibiotic. Nevertheless, when subinhibitory concentrations of ciprofloxacin were added to the medium, smrA was significantly overexpressed in BB1013 compared to BB1001, while satAB levels of expression remained stably high. Overexpression of efflux pumps in the presence of ciprofloxacin has already been observed for transporters not related to the efflux of this drug (28). Therefore, the link between expression and resistance is inconclusive. Genetic approaches, such as knockout or knock-in mutations, are needed to confirm the implication of a pump in a given resistance phenotype. Cloning in E. coli is an easy and convenient approach for studying gene function. Interestingly, some MFS pumps from Gram-positive bacteria, such as NorA and Bmr, have been cloned and are functional in E. coli (29, 43), while others, such as PmrA, have, to our knowledge, never been cloned in a Gram-negative species, despite the interest this pump raises. Together our data suggest that smrA and satAB are extremely toxic for E. coli. We hypothesize that cell wall destabilization due to structural differences between E. coli and S. suis is at the basis of this phenomenon. It would be interesting to elucidate the features that make some Gram-positive MFS pumps, and not others, toxic in Gram-negative bacteria even though they all share the same basic 12-transmembrane-segment structure. When cloning toxicity is an impediment, knockout mutants are the main strategy to study and characterize pumps. Nevertheless, tools to perform these assays have not been developed for all species. Even when available, wild-type strains are often difficult to manipulate, and modifiable laboratory strains might not show the desired efflux phenotype. In S. suis, although some genetic tools are available (38), DNA transformation in BB1013 was found to be a bottleneck for further experiments. SmrA and SatAB were found to be very toxic in E. coli, rendering cloning and further study unviable for this species. Thus, a system enabling the introduction of SmrA and SatAB pumps, while avoiding toxicity, was required. pDR67 and B. subtilis 168 were used to effectively obtain knock-in mutants that were suitable for studying efflux pumps. This system has proven to be a fast, easy, and reliable tool in the study of MFS and ABC transporters of S. suis and will surely be helpful for the study of other pumps from different Gram-positive species.
PmrA is an MFS efflux pump from S. pneumoniae that has been, at least for some time, considered to be the main norfloxacin exporter in this bacterium. Since its first description (12), PmrA has generally failed to be further related to FQ resistance in either wild-type strains or laboratory mutants (1, 5, 28, 35). Its role in fluoroquinolone resistance therefore remains unclear and is now questioned. The function of SmrA, the homolog of pneumococcal PmrA in S. suis, was intensively studied. Our results point to SmrA as not being involved in fluoroquinolone resistance in BB1013. Among more than 450 compounds tested, only minocycline, an antibiotic from the family of tetracyclines, seemed to be affected by SmrA. These differences were subtle, and further investigations are needed to define the possible clinical impact of SmrA in minocycline resistance. Nevertheless, they were consistent in genetic constructions in both B. subtilis and S. pneumoniae bearing an inducible copy of the pump, allowing the inference that the whole expression system works not only for ABC but for MFS transporters as well. Also, S. pneumoniae seems a suitable species for the use of the Pspac-lacI expression system.
Expression levels of smrA were measured in BB1013. Interestingly, expression of smrA increased in this strain in the presence of ciprofloxacin. As mentioned above, similar results have also been observed in S. pneumoniae for other transporters that are not involved in FQ extrusion (28). S. pneumoniae's chromosome has been shown to be organized in topology-reacting gene clusters. DNA relaxation by a gyrase inhibitor modified global gene expression, which affected both pmrA (downregulation) and patA-patB (upregulation) in opposite ways (10). It would be relevant, from both a clinical and a basic research perspective, to elucidate the reasons why a subset of unrelated transporters are overexpressed when in the presence of antibiotics that are not a substrate thereof. Unraveling whether this is a general response to FQ-induced stress, such as an increased superoxide production, or an indirect consequence of the effect of these antibiotics on the coiling and topology of the chromosome (10) may be important for antimicrobial therapy.
Our results prove that SatAB can extrude the fluoroquinolones norfloxacin and ciprofloxacin. Interestingly, the alleles from both susceptible and resistant strains increased the levels of resistance to fluoroquinolones when cloned in B. subtilis. The activity of SatAB was partially reverted by reserpine, as was the case for BB1013. Interestingly, the MIC of susceptible S. suis strains was unaffected by reserpine (8), suggesting that the effect of this pump in the intrinsic levels of FQ resistance in S. suis is probably low. It is unlikely that the natural function of SatAB is FQ efflux, as these antibiotics are fully synthetic. Therefore, other functions of this transporter are yet to be unveiled. The genetic environment of satAB (Fig. 2) (19) and the presence of SatA and SatB homologues in other streptococci (Fig. 6 ) point to a possible role in basic metabolic functions. Also, the phylogenetic relationship of the two genes with their homologues shows a clear species-dependent clustering, suggesting that satAB coevolves with the whole bacterial genome and has not yet been the subject of horizontal gene transfer events.
Fig. 6.
Fig. 6. Dendrograms from the amino acid sequence of SatA and SatB and closely related homologues of streptococci and enterococci. Trees were constructed with NCBI BLAST by using Fast minimum evolution and Grishin parameters and edited using MEGA software. The bar denotes genetic distance. A clear species-dependent clustering was observed among the hits obtained. Note, however, that (i) some S. mitis, S. oralis, and S. pneumoniae strains inter-mixed and that (ii) S. sanguinis strain ATCC 49296 clusters together with viridans group streptococci, away from the 21 other members of this species represented by S. sanguinis SK49. Although this suggests a horizontal gene transfer event, it is rather a misidentification of S. sanguinis strain ATCC 49296, as its 16S rRNA gene also clusters with S. oralis and S. mitis (data not shown), pointing to SatAB as a good evolutionary marker. To clearly illustrate this species-specific grouping, only one representative of each species is shown (with the number of strains included in each branch indicated by gray numbers), except for S. mitis and S. sanguinis, for which two strains have been represented, and S. suis, for which all nonredundant hits were included. It is noteworthy that SatB yields, in S. suis, one hit less than SatA. This is due to the fact that strain 05ZYH33 has an A1056 deletion in satB that leads to a premature stop codon, rendering the identity with the query sequence too low to be represented in the tree. The similarity of both trees and the species-specific clustering suggests the coevolution of both ORFs together and with the genome and points to the nonimplication of satAB in horizontal gene transfer events.
Interestingly, levels of resistance conferred by SatAB differed depending on the allele. A possible influence of the amino acid differences found among alleles cannot be ruled out. Still, SatAB overexpression seems to be the main requisite to obtain fluoroquinolone efflux. As is the case for other pumps, of all fluoroquinolones tested, only ciprofloxacin and norfloxacin were affected by SatAB. This is interesting from both a drug development and a structural point of view, as the substrate ciprofloxacin and the nonsubstrate enrofloxacin differ only in a methyl group. An increased bulkiness of the methylated substituent at position C-7, which would impede enrofloxacin from traversing the pump's channel, is likely to be at the basis of these differences (41). It is also relevant from a therapeutic perspective, as selection of efflux or gyrA or parC mutants may depend on the molecule used for treatment.
Together our data on SmrA allow one to neglect the possibility of its being a fluoroquinolone efflux pump of clinical relevance. Therefore, in order to further investigate S. suis mechanisms of FQ resistance, efforts should be directed toward performance of an in-depth characterization of SatAB and the discovery of other possible transporters. Experiments to elucidate the implication of SatAB in other resistant clinical isolates showing efflux are under way.

ACKNOWLEDGMENTS

We thank T. Msadek for the kind donation of B. subtilis 168 strain and pDR67 plasmid. Natalia Montero is acknowledged for excellent technical assistance. L. Gutmann and S. Coyne are acknowledged for helpful discussion.
We thank the Universidad Complutense de Madrid for the Ph.D. scholarship of J.A.E. and the Spanish Ministry of Science and Innovation (MICINN) for supporting the Ph.D. scholarship of A.S.M. B.G.-Z. acknowledges WP29 of the Med-Vet-Net Network of Excellence (FOOD-CT-2004-506122), the MICINN (BIO 2010-20204 and GEN2006-27767-E/PAT), and the Programa de Vigilancia Sanitaria 2009 AGR/4189 of the Comunidad de Madrid (Spain). A. G. de la Campa acknowledges MICINN grants BIO2008-02154 and BIO2011-2543 and Ciber Enfermedades Respiratorias, an initiative from Instituto de Salud Carlos III.

REFERENCES

1.
Avrain L. et al. 2007. Selection of quinolone resistance in Streptococcus pneumoniae exposed in vitro to subinhibitory drug concentrations. J. Antimicrob. Chemother. 60:965–972.
2.
Bailey A., Paulsen I., and Piddock L. 2008. RamA confers multidrug resistance in Salmonella enterica via increased expression of acrB, which is inhibited by chlorpromazine. Antimicrob. Agents Chemother. 52:3604–3611.
3.
Ballester S., Lopez P., Alonso J. C., Espinosa M., and Lacks S. A. 1986. Selective advantage of deletions enhancing chloramphenicol acetyltransferase gene expression in Streptococcus pneumoniae plasmids. Gene 41:153–163.
4.
Balsalobre L. and de la Campa A. 2008. Fitness of Streptococcus pneumoniae fluoroquinolone-resistant strains with topoisomerase IV recombinant genes. Antimicrob. Agents Chemother. 52:822–830.
5.
Brenwald N., Appelbaum P., Davies T., and Gill M. 2003. Evidence for efflux pumps, other than PmrA, associated with fluoroquinolone resistance in Streptococcus pneumoniae. Clin. Microbiol. Infect. 9:140–143.
6.
CLSI. 2009. Performance standards for antimicrobial susceptibility testing; nineteenth informational supplement. Clinical and Laboratory Standards Institute, Wayne, PA.
7.
Courvalin P., Leclercq R., and Rice L. 2010. Antibiogram. ESKA Pub., Portland, OR.
8.
Escudero J. A. et al. 2007. First characterization of fluoroquinolone resistance in Streptococcus suis. Antimicrob. Agents Chemother. 51:777–782.
9.
Feldman C. and Anderson R. 2011. Bacteraemic pneumococcal pneumonia: current therapeutic options. Drugs 71:131–153.
10.
Ferrándiz M. J., Martín-Galiano A. J., Schvartzman J. B., and de la Campa A. G. 2010. The genome of Streptococcus pneumoniae is organized in topology-reacting gene clusters. Nucleic Acids Res. 38:3570–3581.
11.
Garvey M. and Piddock L. 2008. The efflux pump inhibitor reserpine selects multidrug-resistant Streptococcus pneumoniae strains that overexpress the ABC transporters PatA and PatB. Antimicrob. Agents Chemother. 52:1677–1685.
12.
Gill M. J., Brenwald N. P., and Wise R. 1999. Identification of an efflux pump gene, pmrA, associated with fluoroquinolone resistance in Streptococcus pneumoniae. Antimicrob. Agents Chemother. 43:187–189.
13.
Godreuil S., Galimand M., Gerbaud G., Jacquet C., and Courvalin P. 2003. Efflux pump Lde is associated with fluoroquinolone resistance in Listeria monocytogenes. Antimicrob. Agents Chemother. 47:704–708.
14.
Gottschalk M., Turgeon P., Higgins R., Beaudoin M., and Bourgault A. M. 1991. Susceptibility of Streptococcus suis to penicillin. J. Vet. Diagn. Invest. 3:170–172.
15.
Harwood C. and Cutting S. 1990. Molecular biological methods for Bacillus. John Wiley & Sons, Inc., Hoboken, NJ.
16.
Heldwein E. E. and Brennan R. G. 2001. Crystal structure of the transcription activator BmrR bound to DNA and a drug. Nature 409:378–382.
17.
Holden M. T. et al. 2009. Rapid evolution of virulence and drug resistance in the emerging zoonotic pathogen Streptococcus suis. PLoS One 4:e6072.
18.
Horby P. et al. 2010. Stimulating the development of national Streptococcus suis guidelines in Viet Nam through a strategic research partnership. Bull. World Health Organ. 88:458–461.
19.
Hu P. et al. 2011. Complete genome sequence of Streptococcus suis serotype 3 strain ST3. J. Bacteriol. 193:3428–3429.
20.
Huang Y. T., Teng L. J., Ho S. W., and Hsueh P. R. 2005. Streptococcus suis infection. J. Microbiol. Immunol. Infect. 38:306–313.
21.
Hui A. C. et al. 2005. Bacterial meningitis in Hong Kong: 10-years' experience. Clin. Neurol. Neurosurg. 107:366–370.
22.
Ireton K., Rudner D. Z., Siranosian K. J., and Grossman A. D. 1993. Integration of multiple developmental signals in Bacillus subtilis through the Spo0A transcription factor. Genes Dev. 7:283–294.
23.
Kelley L. A. and Sternberg M. J. 2009. Protein structure prediction on the Web: a case study using the Phyre server. Nat. Protoc. 4:363–371.
24.
Kerdsin A. et al. 2011. Genotypic profile of Streptococcus suis serotype 2 and clinical features of infection in humans, Thailand. Emerg. Infect. Dis. 17:835–842.
25.
Lacks S. A., Lopez P., Greenberg B., and Espinosa M. 1986. Identification and analysis of genes for tetracycline resistance and replication functions in the broad-host-range plasmid pLS1. J. Mol. Biol. 192:753–765.
26.
Manzin A. et al. 2008. Streptococcus suis meningitis without history of animal contact, Italy. Emerg. Infect. Dis. 14:1946–1948.
27.
Marger M. D. and Saier M. H. 1993. A major superfamily of transmembrane facilitators that catalyse uniport, symport and antiport. Trends Biochem. Sci. 18:13–20.
28.
Marrer E. et al. 2006. Involvement of the putative ATP-dependent efflux proteins PatA and PatB in fluoroquinolone resistance of a multidrug-resistant mutant of Streptococcus pneumoniae. Antimicrob. Agents Chemother. 50:685–693.
29.
Neyfakh A. A., Bidnenko V. E., and Chen L. B. 1991. Efflux-mediated multidrug resistance in Bacillus subtilis: similarities and dissimilarities with the mammalian system. Proc. Natl. Acad. Sci. U. S. A. 88:4781–4785.
30.
Nghia H. D. et al. 2011. Risk factors of Streptococcus suis infection in Vietnam. A case-control study. PLoS One 6:e17604.
31.
Nikaido H. 1996. Multidrug efflux pumps of Gram-negative bacteria. J. Bacteriol. 178:5853–5859.
32.
Ohshita Y., Hiramatsu K., and Yokota T. 1990. A point mutation in norA gene is responsible for quinolone resistance in Staphylococcus aureus. Biochem. Biophys. Res. Commun. 172:1028–1034.
33.
Palmieri C., Princivalli M. S., Brenciani A., Varaldo P. E., and Facinelli B. 2011. Different genetic elements carrying the tet(W) gene in two human clinical isolates of Streptococcus suis. Antimicrob. Agents Chemother. 55:631–636.
34.
Paulsen I. T., Brown M. H., Littlejohn T. G., Mitchell B. A., and Skurray R. A. 1996. Multidrug resistance proteins QacA and QacB from Staphylococcus aureus: membrane topology and identification of residues involved in substrate specificity. Proc. Natl. Acad. Sci. U. S. A. 93:3630–3635.
35.
Piddock L. J., Johnson M. M., Simjee S., and Pumbwe L. 2002. Expression of efflux pump gene pmrA in fluoroquinolone-resistant and -susceptible clinical isolates of Streptococcus pneumoniae. Antimicrob. Agents Chemother. 46:808–812.
36.
Prieto C., García F. J., Suárez P., Imáz M., and Castro J. M. 1994. Biochemical traits and antimicrobial susceptibility of Streptococcus suis isolated from slaughtered pigs. Zentralbl. Veterinarmed. B 41:608–617.
37.
San Millan A. et al. 2009. Multiresistance in Pasteurella multocida is mediated by coexistence of small plasmids. Antimicrob. Agents Chemother. 53:3399–3404.
38.
Smith H., Wisselink H., Vecht U., Gielkens A., and Smits M. 1995. High-efficiency transformation and gene inactivation in Streptococcus suis type 2. Microbiology 141:181–188.
39.
Stassi D. L., Lopez P., Espinosa M., and Lacks S. A. 1981. Cloning of chromosomal genes in Streptococcus pneumoniae. Proc. Natl. Acad. Sci. U. S. A. 78:7028–7032.
40.
Suankratay C., Intalapaporn P., Nunthapisud P., Arunyingmongkol K., and Wilde H. 2004. Streptococcus suis meningitis in Thailand. Southeast Asian J. Trop. Med. Public Health 35:868–876.
41.
Takenouchi T. et al. 1996. Hydrophilicity of quinolones is not an exclusive factor for decreased activity in efflux-mediated resistant mutants of Staphylococcus aureus. Antimicrob. Agents Chemother. 40:1835–1842.
42.
Tang J. et al. 2006. Streptococcal toxic shock syndrome caused by Streptococcus suis serotype 2. PLoS Med. 3:e151.
43.
Ubukata K., Itoh-Yamashita N., and Konno M. 1989. Cloning and expression of the norA gene for fluoroquinolone resistance in Staphylococcus aureus. Antimicrob. Agents Chemother. 33:1535–1539.
44.
van Veen H. W. et al. 1996. Multidrug resistance mediated by a bacterial homolog of the human multidrug transporter MDR1. Proc. Natl. Acad. Sci. U. S. A. 93:10668–10672.
45.
Vincze T., Posfai J., and Roberts R. J. 2003. NEBcutter: a program to cleave DNA with restriction enzymes. Nucleic Acids Res. 31:3688–3691.
46.
Wasteson Y., Høie S., and Roberts M. C. 1994. Characterization of antibiotic resistance in Streptococcus suis. Vet. Microbiol. 41:41–49.
47.
Yu H. et al. 2006. Human Streptococcus suis outbreak, Sichuan, China. Emerg. Infect. Dis. 12:914–920.
48.
Zheng X. et al. 2011. Identification of genes and genomic islands correlated with high pathogenicity in Streptococcus suis using whole genome tilling microarrays. PLoS One 6:e17987.

Information & Contributors

Information

Published In

cover image Antimicrobial Agents and Chemotherapy
Antimicrobial Agents and Chemotherapy
Volume 55Number 12December 2011
Pages: 5850 - 5860
PubMed: 21930876

History

Received: 13 April 2011
Returned for modification: 8 August 2011
Accepted: 13 September 2011
Published online: 11 November 2011

Permissions

Request permissions for this article.

Contributors

Authors

Jose Antonio Escudero
Departamento de Sanidad Animal, Facultad de Veterinaria, Universidad Complutense de Madrid, Madrid, Spain
Centro de Vigilancia Sanitaria Veterinaria (VISAVET), Universidad Complutense de Madrid, Madrid, Spain
Alvaro San Millan
Departamento de Sanidad Animal, Facultad de Veterinaria, Universidad Complutense de Madrid, Madrid, Spain
Centro de Vigilancia Sanitaria Veterinaria (VISAVET), Universidad Complutense de Madrid, Madrid, Spain
Belen Gutierrez
Departamento de Sanidad Animal, Facultad de Veterinaria, Universidad Complutense de Madrid, Madrid, Spain
Centro de Vigilancia Sanitaria Veterinaria (VISAVET), Universidad Complutense de Madrid, Madrid, Spain
Laura Hidalgo
Departamento de Sanidad Animal, Facultad de Veterinaria, Universidad Complutense de Madrid, Madrid, Spain
Centro de Vigilancia Sanitaria Veterinaria (VISAVET), Universidad Complutense de Madrid, Madrid, Spain
Roberto M. La Ragione
Department of Bacteriology, Animal Health and Veterinary Laboratories Agency (AHVLA), Weybridge, New Haw, Addlestone, Surrey KT15 3NB, United Kingdom
Faculty of Health and Medical Sciences, University of Surrey, Guildford, Surrey GU2 7XH, United Kingdom
Manal AbuOun
Department of Bacteriology, Animal Health and Veterinary Laboratories Agency (AHVLA), Weybridge, New Haw, Addlestone, Surrey KT15 3NB, United Kingdom
Marc Galimand
Institut Pasteur, Unité des Agents Antibactériens, Paris, France
María José Ferrándiz
Unidad de Genética Bacteriana, Centro Nacional de Microbiología, and CIBER Enfermedades Respiratorias, Instituto de Salud Carlos III, 28220 Majadahonda, Madrid, Spain
Lucas Domínguez
Centro de Vigilancia Sanitaria Veterinaria (VISAVET), Universidad Complutense de Madrid, Madrid, Spain
Adela G. de la Campa
Unidad de Genética Bacteriana, Centro Nacional de Microbiología, and CIBER Enfermedades Respiratorias, Instituto de Salud Carlos III, 28220 Majadahonda, Madrid, Spain
Consejo Superior de Investigaciones Científicas (CSIC), Madrid, Spain
Bruno Gonzalez-Zorn [email protected]
Departamento de Sanidad Animal, Facultad de Veterinaria, Universidad Complutense de Madrid, Madrid, Spain
Centro de Vigilancia Sanitaria Veterinaria (VISAVET), Universidad Complutense de Madrid, Madrid, Spain

Metrics & Citations

Metrics

Note:

  • For recently published articles, the TOTAL download count will appear as zero until a new month starts.
  • There is a 3- to 4-day delay in article usage, so article usage will not appear immediately after publication.
  • Citation counts come from the Crossref Cited by service.

Citations

If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. For an editable text file, please select Medlars format which will download as a .txt file. Simply select your manager software from the list below and click Download.

View Options

Figures and Media

Figures

Media

Tables

Share

Share

Share the article link

Share with email

Email a colleague

Share on social media

American Society for Microbiology ("ASM") is committed to maintaining your confidence and trust with respect to the information we collect from you on websites owned and operated by ASM ("ASM Web Sites") and other sources. This Privacy Policy sets forth the information we collect about you, how we use this information and the choices you have about how we use such information.
FIND OUT MORE about the privacy policy