Genetic detection of quinolone resistance in Haemophilus parainfluenzae : Mutations in the quinolone resistance-determining regions of gyr A and par C

1Vaccine Preventable Bacterial Diseases, National Microbiology Laboratory, Public Health Agency of Canada, Winnipeg, Manitoba; 2Laboratoire de santé publique du Québec, Institut national de santé publique du Québec, Ste-Anne-de-Bellevue, Québec; 3Division of Medical Microbiology and Infection Control, Vancouver General Hospital, Vancouver, British Columbia Correspondence: Dr Raymond SW Tsang, Vaccine Preventable Bacterial Diseases, National Microbiology Laboratory, 1015 Arlington Street, Winnipeg, Manitoba R3E 3R2. Telephone 204-789-6020, fax 204-789-2018, e-mail raymond.tsang@phac-aspc.gc.ca Although generally not regarded as a pathogen, Haemophilus parainfluenzae does occasionally cause infections in humans, including pneumonia (1), peritonitis (2,3), biliary tract infection (4), endocarditis (5) and urinary tract infection (6). More recently, its role in chronic obstructive pulmonary disease has been discussed (7,8). Due to its potential role in the infectious process, susceptibility testing of H parainfluenzae is important to guide antibiotic treatment when required. This is especially relevant because beta-lactamase-mediated resistance to ampicillin has been observed more frequently in H parainfluenzae than in Haemophilus influenzae (9). Furthermore, fluoroquinolone-resistant H parainfluenzae has recently been reported in Japan (10). Of the 920 H parainfluenzae isolates recovered from respiratory specimens of patients admitted to outpatient clinics in Japan between 1998 and 2000, 31 isolates or 3.4% were resistant to fluoroquinolones based on the standard microbroth dilution method, and had a minimum inhibitory concentration (MIC) breakpoint of greater than 2 μg/mL for levofloxacin, according to the Clinical and Laboratory Standards Institute (USA) guidelines (11). The MIC required to inhibit the growth of 90% of organisms for five fluoroquinolones (ciprofloxacin, levofloxacin, moxifloxacin, sitafloxacin and sparfloxacin) for the fluoroquinolone-sensitive H parainfluenzae isolates ranged from 0.015 μg/mL or less to 0.06 μg/mL, and was 100-fold higher for the fluoroquinolone-resistant H parainfluenzae isolates. However, the molecular basis of this resistance was not studied. Mutations in both gyrA and parC have been documented to be associated with quinolone resistance in H influenzae and a number of other Gram-negative bacteria (12-15). Methods In the present study, the quinolone resistance-determining regions (QRDRs) of gyrA and parC for both quinolone-sensitive and quinolone-resistant H parainfluenzae strains were amplified by polymerase chain reaction (PCR) and then sequenced. PCR primers described for amplifying gyrA and parC of H influenzae (14) did not amplify the corresponding genes of H parainfluenzae. Based on the complete genome sequence of H parainfluenzae strain T3T1 (www.sanger.ac.uk/Projects/H_parainfluenzae/), primers were designed to target the QRDRs of both gyrA and parC of this organism. For gyrA, the forward and reverse primers GYRAF2 (5′-TCG ACG TGT GCT TTT CTC CAT G-3′) and GYRAR1 (5′-GTT GCC ATC CCC ACC GCA ATA CCG-3′), respectively, amplified a 405 base pair gene fragment. For parC, the forward and reverse primers PARCF1 (5'-TCT GAA CTG GGT TTA AAT GCC-3') and PARCR1 (5'-ACC ACG ACC GGT TTC ATA CAT-3'), respectively, amplified a 564 base pair gene fragment. The PCR mixture consisted of 2.5 mM magnesium chloride, 0.2 mM deoxyribonucleotide triphosphates, 10 mM primers and 1.25 units of HotStarTaq DNA polymerase (Qiagen, Canada). The PCR cycle consisted of an initial 15 min at 95°C, followed by 35 cycles at 95°C for 1 min, 54°C for 1 min and 72°C for 1 min and a final phase of 72°C for 10 min. The PCR amplicons were purified with the QIAquick PCR purification kit (Qiagen, Canada) before being sequenced by the DNA analyzer 3730x1 (Applied Biosystems, USA). The sequences of both strands were edited, assembled and aligned for comparison with published sequences obtained from oriGinAl ArtiCle

A lthough generally not regarded as a pathogen, Haemophilus parainfluenzae does occasionally cause infections in humans, including pneumonia (1), peritonitis (2,3), biliary tract infection (4), endocarditis (5) and urinary tract infection (6).More recently, its role in chronic obstructive pulmonary disease has been discussed (7,8).Due to its potential role in the infectious process, susceptibility testing of H parainfluenzae is important to guide antibiotic treatment when required.This is especially relevant because beta-lactamase-mediated resistance to ampicillin has been observed more frequently in H parainfluenzae than in Haemophilus influenzae (9).Furthermore, fluoroquinolone-resistant H parainfluenzae has recently been reported in Japan (10).Of the 920 H parainfluenzae isolates recovered from respiratory specimens of patients admitted to outpatient clinics in Japan between 1998 and 2000, 31 isolates or 3.4% were resistant to fluoroquinolones based on the standard microbroth dilution method, and had a minimum inhibitory concentration (MIC) breakpoint of greater than 2 μg/mL for levofloxacin, according to the Clinical and Laboratory Standards Institute (USA) guidelines (11).The MIC required to inhibit the growth of 90% of organisms for five fluoroquinolones (ciprofloxacin, levofloxacin, moxifloxacin, sitafloxacin and sparfloxacin) for the fluoroquinolone-sensitive H parainfluenzae isolates ranged from 0.015 μg/mL or less to 0.06 μg/mL, and was 100-fold higher for the fluoroquinolone-resistant H parainfluenzae isolates.However, the molecular basis of this resistance was not studied.
Mutations in both gyrA and parC have been documented to be associated with quinolone resistance in H influenzae and a number of other Gram-negative bacteria (12)(13)(14)(15).

Methods
In the present study, the quinolone resistance-determining regions (QRDRs) of gyrA and parC for both quinolone-sensitive and quinolone-resistant H parainfluenzae strains were amplified by polymerase chain reaction (PCR) and then sequenced.PCR primers described for amplifying gyrA and parC of H influenzae (14) did not amplify the corresponding genes of H parainfluenzae.Based on the complete genome sequence of H parainfluenzae strain T3T1 (www.sanger.ac.uk/Projects/H_parainfluenzae/), primers were designed to target the QRDRs of both gyrA and parC of this organism.For gyrA, the forward and reverse primers GYRAF2 (5′-TCG ACG TGT GCT TTT CTC CAT G-3′) and GYRAR1 (5′-GTT GCC ATC CCC ACC GCA ATA CCG-3′), respectively, amplified a 405 base pair gene fragment.For parC, the forward and reverse primers PARCF1 (5'-TCT GAA CTG GGT TTA AAT GCC-3') and PARCR1 (5'-ACC ACG ACC GGT TTC ATA CAT-3'), respectively, amplified a 564 base pair gene fragment.
The PCR mixture consisted of 2.5 mM magnesium chloride, 0.2 mM deoxyribonucleotide triphosphates, 10 mM primers and 1.25 units of HotStarTaq DNA polymerase (Qiagen, Canada).The PCR cycle consisted of an initial 15 min at 95°C, followed by 35 cycles at 95°C for 1 min, 54°C for 1 min and 72°C for 1 min and a final phase of 72°C for 10 min.The PCR amplicons were purified with the QIAquick PCR purification kit (Qiagen, Canada) before being sequenced by the DNA analyzer 3730x1 (Applied Biosystems, USA).The sequences of both strands were edited, assembled and aligned for comparison with published sequences obtained from The quinolone resistance-determining regions of gyrA and parC of both quinolone-sensitive and quinolone-resistant Haemophilus parainfluenzae strains were amplified and sequenced.Similar to Haemophilus influenzae, resistance to quinolones in H parainfluenzae is associated with mutations in the quinolone resistance-determining regions of both gyrA and parC.The present study discusses the importance of this finding.

ResuLts And dIsCussIon
Two H parainfluenzae isolates recovered from respiratory specimens were referred to the National Microbiology Laboratory (NML; Winnipeg, Manitoba) for further characterization: isolate 1, NML# 07-028, from the pleural fluid of a 52-year-old woman from British Columbia; and isolate 2, NML# 07-020, from the sputum of an 83-year-old man from Quebec.Both isolates were confirmed as H parainfluenzae by biochemical tests and 16S ribosomal RNA sequencing.Isolate 1 belonged to biotype I, and isolate 2 belonged to biotype II.Detection of betalactamase was performed using the BBL DrySlide nitrocefin (Becton Dickinson, Canada).Antibiotic susceptibility testing by the disk-diffusion method was performed according to the Clinical and Laboratory Standards Institute guidelines (16).Etest (bioMérieux, Canada) was used to determine the MIC of fluoroquinolones (ciprofloxacin, levofloxacin and moxifloxacin) to H parainfluenzae.Microbroth dilution method using the Sensititre plates (Trek Diagnostic Systems Inc, USA) was also performed to test the MIC of levofloxacin.Both isolates did not produce beta-lactamase and were found by disk-diffusion susceptibility testing to be resistant to ciprofloxacin, moxifloxacin, clarithromycin and azithromycin, but sensitive to ampicillin, amoxicillin-clavulanic acid, cefaclor, ceftriaxone, trimethoprimsulfamethoxazole and chloramphenicol.In addition, 11 quinolone-sensitive H parainfluenzae isolates (sensitive to ciprofloxacin, moxifloxacin and levofloxacin) were included for comparison.These 11 isolates included ATCC 33392, three from respiratory sources, three from stool/appendix, and one each from cerebral spinal fluid, eye, gallbladder and one of unknown source.
Table 1 compares the MICs of fluoroquinolones and the amino acid compositions of the QRDRs of gyrA and parC in the H parainfluenzae strains examined in the present study.Similar to H influenzae (14), resistance to quinolones in H parainfluenzae is associated with mutations in both gyrA and parC.In the two resistant strains examined, the mutations involved substitutions of Ser84Phe and Asp88Tyr of gyrA, and Ser84Phe of parC.The difference in the MICs (2 μg/mL versus 12.0 μg/mL) of levofloxacin in these two resistant strains suggest that strain 07-028 might possess additional resistance mechanisms, such as additional mutations in the drug targets GyrB and ParE, decreased permeation of the drug, or presence and/or overexpression of an endogenous efflux pump system that causes reduced accumulation of the drug in the target site.These possibilities have not been examined in the strains.
Besides being capable of causing a variety of infections, H parainfluenzae has been reported to be associated with acute exacerbations of chronic bronchitis (AECB) (17).Empirical antibiotic treatment of patients with AECB has been proven to be beneficial in those with severe airflow obstruction and presenting with at least two or all three cardinal symptoms (increased dyspnea, increased sputum volume and increased sputum purulence) (18).With increasing reports of resistance of many bacteria to multiple antibiotics including the beta-lactams, fluoroquinolones have become the antibiotic of choice to treat AECB (19).

suMMARy
We report new primers for PCR amplification of the QRDRs of gyrA and parC of H parainfluenzae for molecular detection of quinolone resistance.We confirm that quinolone resistance in H parainfluenzae, similar to that reported for H influenzae and other Gram-negative bacteria, is associated with genetic mutations that lead to amino acid substitutions in the QRDRs of both gyrA and parC.Because these two random isolates were identified at the local hospital or provincial public health laboratories, the prevalence of quinolone-resistant H parainfluenzae strains in the general population is not available from the present study.Further studies are required to understand the extent of fluoroquinolone resistance in H parainfluenzae, and the potential involvement of other DNA gyrase and topoisomerase genes, such as gyrB and parE, in the resistance of these organisms to quinolones.

ACKnoWLedGeMent:
The authors thank the staff of the National Microbiology Laboratory's DNA Core Facility in Winnipeg, Manitoba, for the DNA sequencing work.