Seventeen strains of S. Choleraesuis, obtained from different diseased pigs between 2001 and 2005 (Asai et al., 2010), were used in this study (Table 1). These strains comprised 6 nalidixic acid-resistant strains (low enrofloxacin-accumulation), and 11 nalidixic acid-susceptible strains (two strains: low enrofloxacin-accumulation; three strains: intermediate enrofloxacin-accumulation; six strains: high enrofloxacin-accumulation) as previously recorded (Usui et al., 2009). The intracellular concentration of enrofloxacin was significantly lower in nalidixic acid-resistant isolates, and the nalidixic acid-susceptible isolates ZSC-8 and 582 as compared to other susceptible isolates with the exception of isolates 143, ZSC-12, and 1002 (P < 0.1; Usui et al., 2009). The intracellular enrofloxacin concentrations of isolates 143, ZSC-12, and 1002 were at an intermediate level. Minimum inhibitory concentrations (MIC) of enrofloxacin, ampicillin, oxytetracycline, and kanamycin had been determined in our previous study (Asai et al., 2010). MIC of spectinomycin was determined by the broth microdilution method according to CLSI guidelines (Clinical and Laboratory Standard Institute, 2008).

Five strains (582, 143, 916, 1002, and 419), susceptible to all of the antimicrobial agents evaluated in this study, were selected as parent strains from which mutants were isolated. Cultures grown on agar containing various agents was employed, including three AcrAB substrates (ampicillin, oxytetracycline, and kanamycin) and one non-substrate (spectinomycin; Bailey et al., 2008; Ricci and Piddock, 2009a,b). For selection of mutants, 0.1 mL of each strain, adjusted to 10⁸–10⁹cfu/mL were exposed to 4 × MIC of the respective substance for each strain in the agar medium, as described previously (Ricci and Piddock, 2009b). A single strain was selected for each combination. In addition, as fluoroquinolone is an AcrAB substrates, enrofloxacin-selected mutants obtained in our previous report were used as control (Usui et al., 2010). MICs of antimicrobials for mutants were determined as described above.

To evaluate the genes conferring efflux pump activity and their multiple global regulators, quantitative PCR was applied to the detection and quantification of mRNAs (Usui et al., 2010, 2011). The preparation of cDNA was performed as described by Zheng et al. (2009), with slight modifications. In brief, extraction of total RNA from 5 mL of bacterial suspensions was carried out with an ISOGEN kit (Nippon Gene, Tokyo, Japan). cDNA was synthesized from these RNA samples using the PrimeScript RT reagent kit (TaKaRa, Shiga, Japan). Quantitative PCR was performed with SYBR premix EX TaqII (TaKaRa) on a One Step real-time system (Applied Biosystems, Foster City, CA, USA) according to the manufacturer’s instructions. The oligonucleotide primers used for the detection of cDNA are listed in Table 2. The yields of amplicons from acrA, acrB, ramA, soxS, marA, and rob were normalized to those originating from 16S rRNA. Within bacterial cells, 16S rRNA was assumed to be transcribed at a constant rate throughout the growth conditions used in this study.

After normalization to the levels of 16S rRNA, gene expression was compared with the average values for the six high enrofloxacin-accumulating susceptible strains.

Mutations in the local repressor gene, acrR, and the regulator-binding site were detected by direct DNA sequencing, using previously reported primer sets (Zheng et al., 2009). Amplification of the gene and purification of the resulting amplicon were performed as described before (Zheng et al., 2009). Nucleotide sequences were determined using a BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems) with a 3130 Genetic Analyzer (Applied Biosystems).

Student’s t-test was used to compare results between and within experiments. P-values < 0.05 were considered significant.

Expression of acrA and acrB in the six resistant strains and in the two low enrofloxacin-accumulation nalidixic acid-susceptible strains was significantly higher than that in the other susceptible strains (P < 0.05; Table 1). Expression levels of acrB correlated significantly with those of acrA in this study (correlation coefficient: 0.89).

Of the eight strains with a high acrB expression level, six strains showed higher levels of both ramA and soxS expression compared with the average values for the six high accumulating susceptible strains (P < 0.05; Figure 2). The remaining strains showed a higher level of ramA or soxS expression (P < 0.05): the 14-PLS-21 strain showed only a higher expression level of ramA (107.1-fold ± 42.0-fold), while the ZSC-8 strain showed only a higher expression level of soxS (17.4-fold ± 2.7-fold). Neither the expression of marA nor that of rob changed in any of the strains tested.

A stop codon mutation in AcrR (glutamine-78-to-terminal; Q78ter) was found in 13 strains (Table 1). No significant relationship was found between the Q78ter stop codon mutation in AcrR and the expression of acrB. None of the strains tested had a point mutation in the regulator-binding site, which affects the acrB expression, previously determined using the S. serovar Typhimurium strain BN18 (Olliver et al., 2004).

Expression levels of both acrB and ramA increased significantly in AcrAB substrate-selected mutants compared to these levels in parental strains (P < 0.05; Table 3). Among four AcrAB substrates, a higher level of ramA expression was observed in the enrofloxacin- and ampicillin-selected mutants than in the oxytetracycline- and kanamycin-selected mutants (P < 0.05). In contrast, the expression of soxS decreased significantly in AcrAB substrate-selected mutants compared to that in parental strains (P < 0.05; Table 3). MICs of AcrAB substrates were increased in AcrAB substrate-selected mutants as compared with those in the parental strains (Table 3).

Expression of acrB and multiple regulator genes remained stable in spectinomycin-selected mutants. In spectinomycin-selected mutants, MICs of AcrAB substrates were of the same level as compared with those in the parental strains (Table 3).