Polymorphism in <i>gyrA</i> is associated to quinolones resistance in <scp>C</scp>hilean <i><scp>P</scp>iscirickettsia salmonis</i> field isolates

Abstract

Piscirickettsia salmonis is a gram-negative, pleomorphic coccoid-shaped and facultative intracellular fish pathogenic bacterium, that causes piscirickettsiosis, a disease also known as salmonid rickettsial septicaemia (SRS) (Fryer et al. 1990). In Chile, P. salmonis was isolated for the first time in 1989, but probably has been present since 1983 (Bravo & Campos 1989). The pathogen is highly virulent for salmonids and represents a major health concern for the local salmon farming industry due to economic losses (Fryer & Hedrick 2003; Rozas & Enriquez 2013). SRS has also been reported in important salmon producing countries such as Canada, Ireland, Scotland and Norway, but with a lower economic impact (Rozas & Enriquez 2013). P. salmonis cells are slow growing and hard to cultivate in free-cell media, which has hindered its isolation and further characterization. However, recently described supplemented media (Mikalsen et al. 2008) facilitate their growth and the antibiotic testing assays to evaluate effective antimicrobial therapy. In this regard, quinolones were the former most widely used antimicrobial compounds in Chilean aquaculture, especially as a first line drug against SRS. In Chile, a large proportion of antibiotics is used for prophylaxis rather than for chemotherapy. For instance, in comparison with Norway and Canada, 1500 and eight times more antimicrobials were needed to obtain similar amounts of salmon in Chile in 2007 (Cabello et al. 2013). The same year, 930 metric tons of antibiotics were mostly used in salmon production. Of those antibiotics, 24% were quinolones (Cabello et al. 2013). This practice has already altered the microbiota of sediments in close proximity to aquaculture sites, where a significant fraction of isolated bacteria is antibiotic resistant (Buschmann et al. 2012). Therefore, the emergence and potential spread of antibiotic-resistant P. salmonis isolates cannot be ruled out. Twenty P. salmonis isolates cultured from kidney samples of diseased fish collected from diverse geographic origin (Fig. 1) between 2010 and 2012 were phenotypically characterized for quinolones susceptibility. The isolates were grown on agar (Mikalsen et al. 2008), at 18 °C for 10 days and kept frozen in liquid nitrogen. For antibiotic sensitivity testing, the minimal inhibitory concentration (MIC) for enrofloxacin, flumequine and oxolinic acid were assessed according to the instructions given by the Clinical and Laboratory Standards Institute (CLSI), guide M49-A. 96-well microplates containing twofold serial dilutions (0.03–64 μg mL−1) of antibiotics were inoculated with 5 × 105 colony forming units/well and incubated at 18 °C for 10 days. The absorbance at 580 nm for each well was measured with a ELx800 Absorbance Microplate Reader (BioTek Instruments, Inc). The experiments were performed in triplicate. As a result, 10 of 20 isolates showed MIC values that ranged from 0.015 to 0.03, 0.06 to 0.125 and 0.03 to 0.125 μg mL−1 for enrofloxacin, flumequine and oxolinic acid, respectively, while the remaining 10 isolates proved resistant to those antibiotics and depicted values ranging from 2 to 8 μg mL−1 (Table 1). Well-known resistance mechanisms to quinolones comprise quinolone-binding alteration by amino acid replacement in topoisomerases enzymes or by the action of topoisomerases protecting proteins (Qnr proteins) and acetyltransferases, or the reduction of drug accumulation within bacteria by decreasing uptake or increasing efflux. The latter can arise from mutations in topoisomerases and porin genes or the presence of efflux pumps or plasmid-encoded resistance mechanisms. The above mechanisms have been found to occur in fish pathogens (Colquhoun, Aarflot & Melvold 2007; Shah et al. 2012a,b; recently reviewed in Miranda, Tello & Keen 2013). Regarding topoisomerases genes, specific mutations affecting genes encoding the essential enzyme DNA gyrase (topoisomerase II) seem to be the most frequent (Yoshida et al. 1990; Nakamura 1997; Gibello et al. 2004; Izumi & Aranishi 2004). The protein is composed of two subunits, A and B, encoded by gyrA and gyrB genes, respectively, and acquired-resistance mutations are frequently accumulated in gyrA, especially in a region within the N-terminal domain spanning 40 amino acids between positions 67 and 106. This region, known as the Quinolone Resistance-Determining Region (QRDR), is a mutation-prone locus associated with resistance to quinolones in gram-negative bacteria (Hawkey 2003). Therefore, we hypothesized that Chilean quinolone-resistant P. salmonis isolates may carry mutations on their QRDR as well. To prove this hypothesis, we investigated the sequence of the QRDR of P. salmonis field isolates. DNA samples were obtained using E.Z.N.A™ Bacterial DNA Kit (Omega-Biotek) following the manufacturer's instructions. Previously, we massively sequenced the genome of all P. salmonis quinolone-resistant isolates (our unpublished results). Based on sequencing data, a specific primer pair targeting the QRDR of P. salmonis was designed (GYRA1PS-F; 5′ AATATCGAAGAAGAACTGAA 3′ and GYRA1PS-R; 5′ GCATGAGCAAATTTATCC 3′). PCR rounds were carried out in a reaction volume of 25 μL. The reaction mixture contained 15 pmol of each primer, 0.1 nmol of each deoxynucleotide triphosphate, 2 mm MgCl2, 1.0 U of Taq DNA polymerase (Invitrogen) and 2.5 μL of template. The thermal profile for PCR was: preheating at 95 °C for 2 min; 35 cycles of denaturation at 95 °C for 40 s, annealing at 55 °C for 30 s, and extension at 72 °C for 40 s; and a final extension step at 72 °C for 10 min, using a thermal cycler 2720 (Applied Biosystems). PCR amplification yielded a fragment of 365 bp for each isolate. Purified fragments were sequenced using BigDye™ Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems) in an ABI PRISM™ 310 Genetic Analyzer (Applied Biosystems) following the manufacturer's instructions. Interestingly, the sequence analysis revealed a unique nucleotidic change (G–T) in position 259 of gyrA in all of 10 quinolone-resistant P. salmonis isolates. This nucleotide change resulted in an amino acid sequence where aspartic acid in position 87 was replaced by tyrosine (Fig. 2, Table 1). This finding suggested that this polymorphism is linked to the reduced bactericidal action of quinolones in P. salmonis, which is clearly manifested in the considerably augmented MIC values (in some resistant isolates more than 200 times higher when compared to susceptible samples). Other well-recognized mutations associated to quinolones resistance are those in gyrB and the genes encoding subunits of topoisomerase IV (parC and parE) (Hawkey 2003). We looked for the sequences of the above genes in the genome of resistant isolates, and the alignment with the sequence of GenBank reference strain (Eppinger et al. 2013) did not show polymorphisms (not shown). Hence, the absence of mutations on gyrB, parC and parE genes and the fact that plasmids were not detected support the hypothesis that the single point mutation found in gyrA of quinolone-resistant P. salmonis isolates could be responsible for the phenotype. To our knowledge, this is the first report describing such an antibiotic resistance mechanism in this relevant fish pathogen. Piscirickettsiosis will continue being a threat for the Chilean salmon industry in the upcoming years. Therefore, the rationalization and rotation of the use of antimicrobials may decrease the selective pressure imposed on bacteria and thus may avoid the selection of multidrug resistant pathogens. We are currently developing a molecular assay for the detection of QRDR mutation in P. salmonis in order to facilitate a faster identification of potential resistant field isolates and to guide the decision for the appropriate therapy to be applied during SRS outbreaks. We thank Dr. Melanie Kaiser for her valuable suggestions and comments about the manuscript. This research was funded by CORFO INNOVA project 12BPC2-13471.