Aminoglycoside Resistance in Pseudomonas aeruginosa Isolated from Cystic Fibrosis Patients

Aminoglycoside Resistance in Pseudomonas aeruginosa Isolated from Cystic Fibrosis Patients WILLIAM F. MCNEILL, M.D., JOSEPH F. JOHN, JR., M.D., A JAMES A. TWITTY, B.S. The authors studied 30 gentamicin-resistant and 17 gentamicinsensitive strains of Pseudomonas aeruginosa isolated from respiratory cultures of patients with cystic fibrosis from five United States cities for the presence of plasmids, cross-resistance to other aminoglycosides, and the production of aminoglycosidemodifying enzymes. Four of 30 resistant strains and 3 of 17 sensitive strains contained one or more plasmids. Aminoglycoside cross-resistance to tobramycin, amikacin, and netilmicin was seen in 21 of 30 gentamicin-resistant strains. Seven strains that had low-level gentamicin resistance (minimum inhibitory concentrations [MIC] = 8-32 /ig/ml) were sensitive to one or more of the other three aminoglycosides. Two strains with high-level gentamicin resistance (MIC ^ /ug/ml) were sensitive to amikacin. These two strains, each containing three plasmids, were the only isolates of nine tested that produced an aminoglycoside-modifying enzyme with activity against gentamicin. None of the plasmids was transferable by conjugation. Four strains, three of which contained one or more plasmids, produced an aminoglycoside 3'-0-phosphotransferase II. The authors propose that the mechanism of gentamicin resistance in P. aeruginosa from patients with cystic fibrosis is not commonly plasmidmediated and likely is due to membrane impermeability to aminoglycosides. (Key words: Pseudomonas aeruginosa; Gentamicin resistance; Cystic fibrosis) Am J Clin Pathol 198; 81:72-77 MOST PATIENTS with cystic fibrosis die of progressive pulmonary disease often complicated by recurrent infection with Pseudomonas aeruginosa.*' 9 In the last decade the aminoglycosides (gentamicin, tobramycin, and amikacin) have become the primary antibiotic therapy of severe infections due to P. aeruginosa. Resistance to gentamicin and tobramycin in P. aeruginosa has been a growing clinical problem in patients with cystic fibrosis. 212 Mechanisms of aminoglycoside resistance in P. aeruginosa involve either resistance (R-) plasmid-mediated enzymes that inactivate aminoglycosides or mutations of chromosomal genes affecting the energy requirements for transport of aminoglycosides into the cell. 3 Aminoglycoside resistance due to inactivating enzymes encoded by R-plasmids occurs in isolates of P. aeruginosa from hos- Received August 2, 1983; received revised manuscript and accepted for publication October 2, 1983. Supported by Cystic Fibrosis Foundation Grant #H008 2-01 and by the Veterans Administration. Dr. McNeill's present address is Department of Pathology, Emory University School of Medicine, Atlanta, Georgia. Address reprint requests to Dr. John: 9 Clinical Science Building, Medical University of South Carolina, Charleston, South Carolina 2925. Department of Medicine, Infectious Diseases and Immunology Division, Medical University of South Carolina, VA Medical Center, Charleston, South Carolina pitalized patients 5101 '! 7 ' 25 but has not been reported in P. aeruginosa isolated from cystic fibrosis patients. In hospitalized patients gentamicin resistance in P. aeruginosa has been associated with increased exposure to aminoglycosides and other antibiotics. 12,20 Cystic fibrosis patients constitute another group who undergo recurrent exposure to aminoglycosides as well as other antibiotics for the control of respiratory infections. In order to elucidate the mechanism of aminoglycoside resistance in P. aeruginosa from patients with cystic fibrosis from different United States cities, we studied the pattern of aminoglycoside cross-resistance, the plasmid content, and the presence of aminoglycoside modifying enzymes in strains known to be gentamicin resistant. Methods Bacterial Strains Forty-seven strains of P. aeruginosa (% resistant to gentamicin) isolated from respiratory cultures from cystic fibrosis patients were collected in Charleston, South Carolina, Pittsburgh, Pennsylvania, Philadelphia, Pennsylvania, Atlanta, Georgia, and Chicago, Illinois. Gentamicin-resistant strains were maintained on Mueller-Hinton agar (Difco, Detroit, MI) containing 10 Mg/mL gentamicin; sensitive strains were maintained on trypticase soy agar (Difco). Determination of Antibiotic Resistance Minimal inhibitory concentrations (MIC) were determined at 37 C on Mueller-Hinton agar (Difco, control lot #682923, containing 20.7 mg/l Mg ++ and 50. mg/ L Ca ++ ) by the agar dilution technic 23 using a Steers' replicating device and an inoculum of 10 to 10 5 colony forming units. MICs were determined for amikacin, gentamicin, kanamycin, netilmicin, streptomycin, and tobramycin. P. aeruginosa ATCC 27853, having standardized MIC values, 18 was used as a control in the MIC determinations. 72

Vol. 81 -No. 6 RESISTANT PSEVDOMONAS AERUGINOSA 73 Percent Resistant 20 30 0 50 60 70 80 90 100 Gentamicin Tobramycin V//////////77777A V//////////////////////////////////////////////////A FIG. 1. Percentage of resistance by disc sus- Amikacin \//////////////////////////////////////////////;/////////A ceptibility testing to other antibiotics for 30 gentamicin-treated and 11 gentamicin-sensitive Kanamycin strains of Pseudomonas aeruginosa from patients with cystic fibrosis. Gentamicin-resistant '///////, w»»»»»»»»». V//////////777J»»»»»»»»»* strains are denned as those that grew on Corbenicillin r //////A Mueller-Hinton agar containing 10 Mg/mL gentamicin. Tetracycline K7/>///////>////////////////////^^^^ Cefotaxime rssssssssssssssssssssssssssssssssss. Moxalactam ^ B - W///A- Gentamicin-sensitive Gentamicin-resistant Antibiotic susceptibility testing by the disc diffusion method of Bauer and associates' was likewise done on Mueller-Hinton agar for the following antibiotics: amikacin, carbenicillin, cefotaxime, gentamicin, kanamycin, moxalactam, streptomycin, tetracycline, and tobramycin. Determination ofplasmid Content Three methods were employed to prepare cleared lysates containing plasmid deoxyribonucleic acid (DNA). In the standard method of Clewell and Helinski, 6 lysates were prepared from 10 ml tryptic soy broth (Difco) overnight culture. In the method of Birnboim and Doly 2 lysates were prepared from 0.5 ml of broth culture employing an alkaline (ph 12.0-12.5) denaturation of chromosomal DNA. Finally, multiple strains were tested for plasmid content by the method of Kado and Liu, which through a heating step minimizes chromosomal DNA contamination. 13 Aliquots of plasmid DNA were subjected to electrophoresis in 0.7% agarose at 50 amps for three hours by the method of Meyers and associates. Control plasmid DNAs of known molecular weights were electrophoresed in gels with cleared lysates for calculation of the molecular weights of unknown plasmids. Assay for Aminoglycoside-modifying Enzymes Cell-free preparations of P. aeruginosa were obtained by ultrasonic disruption for 2 minutes, frozen at -70 C, and assayed as a group for the presence of aminoglycosidemodifying enzymes by methods previously described." Conjugation Procedure Matings were performed on nitrocellulose filters (0.5 jim; Millipore Corp., Bedford, MA) by the method of McHugh and associates 15 using P. aeruginosa PA038 tmp + (gift of Dr. George A. Jacoby) as a recipient. P. aeruginosa PA038 was grown overnight at both 37 C and 3 C prior to use in matings to insure induction of a restriction-deficient phenotype. Selection medium was Mueller-Hinton agar containing 2 mg/ml trimethoprim and 10 Mg/mL gentamicin. Susceptibilities Results Figure 1 shows the per cent of 30 gentamicin-resistant strains and 11 susceptible strains of P. aeruginosa that were resistant by disc susceptibility testing to five aminoglycosides and to four nonaminoglycoside antibiotics. Among gentamicin-resistant strains, the most active drugs were tobramycin, carbenicillin, and moxalactam. Among gentamicin-sensitive strains was resistant to tobramycin or amikacin and only one strain was resistant to carbenicillin. Two gentamicin-sensitive strains were found to be gentamicin-resistant by disc testing, each with zone sizes of 12 mm, but did not grow on Mueller-Hinton agar containing 10 Mg/mL gentamicin. 19 Minimum Inhibitory Concentrations Table 1 shows median MIC values, MIC ranges, and percentage of strains resistant to six aminoglycosides for 30 gentamicin-resistant strains. Although only 27% of our gentamicin-resistant strains were tobramycin resistant by disc testing, 77% were resistant by the agar dilution technic. 9 The most striking finding was cross-resistance to all aminoglycosides tested in all but nine strains. Seven

7 MCNEILL, JOHN, A TWITTY AJ.C.P. -June 198 Table 1. Minimum Inhibitory Concentrations of Six Aminoglycosides for 30 Gentamicin-Resistant Strains of Pseudomonas aeruginosa from Patients with Cystic Fibrosis Antibiotic Gentamicin Tobramycin Amikacin Netilmicin Kanamycin Streptomycin * Resistance is2=8 ^g/ml. t Resistance is ^32 Mg/mL. Median MIC (Mg/mL) Range to > 2 to > 2 to to to 32 to % Resistant 100* 77* 87t 97* loot 100t of these nine strains had low level gentamicin resistance (MICs = 8-32 jug/ml) and were sensitive to one or more of the group tobramycin/amikacin/netilmicin. Only two strains, strains 3-96 and 111, had high-level gentamicin resistance (MICs ^ jug/ml) and corresponding amikacin MICs in the sensitive range. Plasmid Content Of 30 gentamicin-resistant strains, only four contained one or more plasmids by any of three methods. Figure 2 and Figure 3 show examples of agarose gel electrophoreses of cleared lysates prepared by the method of Birnboim and Doly compared with that of Kado and Liu. The former method showed smaller (<10 Md) plasmids better and tended to give brighter bands for all plasmids. The latter method resulted in almost no chromosomal DNA but tended not to show smaller plasmids. The method of Clewell and Helinski gave variable results and plasmid bands were often dull and diffuse. Aminoglycoside Modifying Enzymes Seven gentamicin-resistant and two gentamicin-sensitive strains of P. aeruginosa were assayed for production of aminoglycoside-modifying enzymes (Table 2). Only two resistant strains, 3-96 and 111, elaborated enzymes with activity against gentamicin: 3-96, which made an aminoglycoside-nucleotidyltransferase (ANT), was crossresistant to tobramycin, and 111, which made an aminoglycoside-acetyltransferase (AAC), was cross-resistant to tobramycin and netilmicin. Two gentamicin-resistant and two sensitive strains, three of which contain plasmids, produced an aminoglycoside 3'-0-phosphotransferase II [APH (30-H], which has activity against kanamycin but not against gentamicin, tobramycin or amikacin. 7 Three strains with no demonstrable plasmids lacked aminoglycoside-modifying enzymes. Transfer of Resistance P. aeruginosa strains 111 and 3-96 were tested further for transfer of gentamicin resistance. Because these strains elaborated enzymes with activity against gentamicin and the resistance pattern observed in the two strains was suggestive of aminoglycoside modifying enzymes, the likelihood of their resistance being transferable was greater than for the other 28 resistant strains. However, transfer experiments failed to yield trimethoprim-gentamicin-resistant organisms having the phenotypic properties of the recipient organism. Discussion Among the strains of P. aeruginosa from the respiratory tract of patients with cystic fibrosis in our study, gentamicin resistance is not likely conferred by R-plasmids. Our conclusion is supported by the following findings: (1) Plasmids could not be visualized in the majority of resistant strains. We employed three different methods of extracting plasmid DNA because of the reported unreliability in visualizing pseudomonas plasmids. In fact, transferrable resistance to gentamicin has been observed in strains in which plasmid DNA had not been visualized. 10 Very few reports compared methods of Pseudomonas plasmid preparation. We found variability among the three methods we employed. Based on our results, the rapid method of Bimboim and Doly 2 should be the most applicable of currently available methods to a clinical laboratory. (2) Only two of seven selected gentamicinresistant strains produced aminoglycoside-modifying enzymes capable of inactivating gentamicin, an important finding since P. aeruginosa plasmids confer aminoglycoside resistance by encoding aminoglycoside-modifying enzymes. 3 In these two strains, we were unable to demonstrate transfer of gentamicin resistance. (3) Seventyseven per cent of the gentamicin-resistant strains were cross-resistant to all other aminoglycosides; in addition, 87% were cross-resistant to amikacin. A high degree of FIGS. 2 (upper) and 3 (lower). Agarose gel electrophoresis of cleared lysates prepared by two different methods: Figure 2, the method of Bimboim and Doly 17; Figure 3 the method of Kado and Liu. 13 C and H are control plasmid DNAs of known molecular megadalton (Md) mass, top to bottom, of 63 Md, 5 Md, 3 Md, 26 Md, 5.8 Md, and 2.6 Md. A, B, F, and G have no plasmid visible in either method. D, strain 111, contains two large plasmids, 81 Md and 8 Md, by both methods and third small faint 3 Md plasmid in Figure 2. E, strain 3-96, contains a 65 Md and 53 Md plasmid in both methods and again a small plasmid by the method of Bimboim and Doly. 2 I, strain CF, has a 50 Md plasmid in both preparations and a small.2 Md plasmid by the method of Bimboim and Doly. 2 Finally, J, strain W5, has a single 62 Md plasmid by both methods.

RESISTANT PSEUDOMONAS AERUGINOSA Vol. 81 - N o. 6 W*T$lfc-K G T? fclmnmmp l^^ Fir 75

76 MCNEILL, JOHN, A TWITTY A.J.C.P. June 198 Strain Table 2. Geographic Origin, Level of Aminoglycoside Resistance, Plasmid Content, and Aminoglycoside Modifying Enzyme Production of Pseudomonas aeruginosa from Patients with Cystic Fibrosis Source Gentamicin-resistant Strains 99 Charleston 111 Charleston 3-96 Philadelphia W27 Pittsburgh W5 Pittsburgh W09 Pittsburgh JG Chicago 3-10 Philadelphia Gentamicin -sensitive Strains P3 Pll CF Charleston Charleston Atlanta Gm* >t > 32 > 2 8 Aminoglycoside Susceptibility Tm > > 32 1 *Gm = gentamicin:tm = tobramycin: An = amikacin: Nt = netilmicin. t APH phosphotransferase: AAC acetyltransferase; ANT nucleotidyltransferase. An 2 32 8 Nt 8 Plasmid Size X10 6 Daltons None 81,8,3 65, 53,.2 71 62 60 57 50,.2 X Minimum inhibitory concentration (ng/ml). Aminoglycoside Modifying Enzymesf APH(3')-II AAC ANT APH(3')-II APH(3')-H APH(3')-II aminoglycoside cross-resistance suggests membrane impermeability to this class of antibiotics. 17 A possible mechanism for aminoglycoside resistance in the P. aeruginosa from our study is chromosomally mediated reduced uptake of the aminoglycoside. Bryan and associates showed aminoglycoside cross-resistance in representative clinical isolates of P. aeruginosa that contained no evidence of gentamicin-modifying enzymes. Plasmid DNA was detected in two of three strains studied by Bryan and associates, but aminoglycoside resistance could not be transferred. A non-plasmid-mediated mechanism for reduced energy-dependent accumulation of gentamicin was proposed. Other workers have found non-plasmid-mediated, nzymatic resistance to gentamicin in P. aeruginosa. John and co-workers 12 found plasmid DNA in 1 of 38 gentamicin-resistant P. aeruginosa from patients with nosocomial infections. However, gentamicin resistance could not be transferred and aminoglycoside-modifying enzymes could not be detected in selected strains. Phillips and associates 17 found 3 of hospital isolates of gentamicin-resistant P. aeruginosa did not produce gentamicin-inactivating enzymes but did show wide aminoglycoside cross-resistance. Our data indirectly support reduced uptake of aminoglycosides as a common mechanism of gentamicin resistance in isolates of P. aeruginosa colonizing cystic fibrosis patients. To support this condition, uptake of intracellular aminoglycoside could be measured in such strains. Strain 111 was highly resistant to gentamicin but sensitive to amikacin; strain 3-96 was also highly gentamicin resistant but sensitive to amikacin and netilmicin. These two strains each contained three plasmids of similar molecular weights (Table 2). Strain 111 was from Charleston, strain 3-96 from Philadelphia. Strain 111 produced an AAC and 3-96 produced an ANT, though the substrate profiles obtained were inadequate to identify accurately the specific enzymes (data not shown). In these two strains, the mechanism of resistance may be mediated by nonconjugative R-plasmids as we were unable to transfer gentamicin resistance in these two strains by the membrane filter technic. 15 If transfer of these plasmids can be accomplished by transformation or transduction, it would be interesting to assess molecular relatedness of the R-plasmids present and to compare their relationship with other R-plasmids of nosocomial P. aeruginosa known to produce enzymes capable of acetylating or adenylating gentamicin and other aminoglycosides. 7 We propose that the most common mechanism of aminoglycoside resistance in P. aeruginosa from patients with cystic fibrosis is chromosomal mutation affecting aminoglycoside transport systems. Such mutation is a likely outgrowth of repeated exposure in cystic fibrosis patients to these agents. Our data discourage the use of alternative aminoglycosides for the treatment of pulmonary infections with gentamicin-resistant P. aeruginosa and suggest that alternative agents may be more effective. Two recently marketed /3-lactam agents, cefotaxime and moxalactam, occasionally were active against gentamicinresistant and gentamicin-sensitive strains 22 ; however, other workers have found increased resistance to these two agents among gentamicin-resistant strains of P. aeruginosa. 26 Acknowledgments. The authors thank the following people for furnishing them with strains: Dr. Ellen Wald, Pittsburgh, Pennsylvania; Dr. Peter H. Gilligan, Philadelphia, Pennsylvania; Dr. John R. Boring, Atlanta, Georgia; and Dr. Ram Yogev, Chicago, Illinois. In addition, they thank Dr. Kenneth Price and Mr. David G. Bobey for performing the enzyme determinations, Dr. W. Edmund Farrar, Jr. for manuscript suggestions, and Ms. June Cox for word processing.

Vol.81 No. 6 RESISTANT PSEUDOMONAS AERUGINOSA 77 References 1. Bauer AW, Kirby WMM, Sherris JC, Turck M: Antibiotic susceptibility testing by standardized single disc method. Am J Clin Pathol 1966; 5:93-96 2. Birnboim HC, Doly J: A rapid alkaline extraction procedure for screening recombinant plasm id DNA. Nucleic Acid Research 1979;7:1513-1523 3. Bryan LE: Resistance to antimicrobial agents: the general nature of the problem and the basis of resistance, Pseudomonas aeruginosa. Clinical Manifestations of Infection and Current Therapy. Edited by RG Doggett. New York, Academic Press, 1979; 219-270. Bryan LE, Haraphongse R, Van Den Elzen HM: Gentamicin resistance in clinical-isplates of Pseudomonas aeruginosa associated with diminished gentamicin accumulation and no detectable enzymatic modification. J Antibiot 1976; 29:73-753 5. Bryan LE, Shahrabadi MS, Van Den Elzen HM: Gentamicin resistance in Pseudomonas aeruginosa: R-factor-mediated resistance. Antimicrob Agents Chemother 197; 6:191-199 6. Clewell DB, Helinski DR: Properties of a super coiled deoxyribonucleic acid-protein relaxation complex and strand specificity of the relaxation event. Biochemistry 1970; 9:28-0 7. Davies J, Smith DI: Plasmid-determined resistance to antimicrobial agents. Ann Rev Microbiol 1978; 32:69-518 8. di Sant' Agnese PA, Davis PB: Research in cysticfibrosis.n Engl J Med 1976; 295:81-85, 53-51, 597-602 9. Hoiby N: Pseudomonas aeruginosa infection in cysticfibrosis.acta Pathol Microbiol Scand Sect Suppl 1977; 262:1-96 10. Jacoby GA: Plasmid determined resistance to carbenicillin and gentamicin in Pseudomonas aeruginosa, Plasmids and transposons environmental effects and maintenance mechanisms. Edited by C Stutard, KR Rozee. New York, Academic Press, 1980; 83-96 11. John JF, McNeill WF, Price KE, Kresel PA: Evidence for a chromosomal site specifying amikacin resistance in multiresistant Serratia marcescens. Antimicrob Agents Chemother 1982; 21:587-591 12. John JF, Rubens CE, Farrar WF: Characteristics of gentamicinresistance in nosocomial infections. Am J Med Sci 1980; 279:25-30 13. Kado CI, Liu ST: Rapid procedure for detection and isolation of large and small plasmids. J Bacteriol 1981; 15:1365-1373 1. Kawakami Y, Mikoshiba F, Nagaski S, Matsumoto H, Taxaki T: Prevalence of Pseudomonas aeruginosa strains passing R-factor in a hospital. J Antibiot 1972; 25:607-609 15. McHugh GL, Moellering RC, Hopkins CC, Swartz MN: Salmonella typhimurium resistant to silver nitrate, chloramphenicol, and ampicillin. Lancet 1975; 1:235-20. Meyers JA, Sanchez D, Elwell LP, Falkow S: Simple agarose gel electrophoresis method for the identification and characterization of plasmid deoxyribonucleic acid. J Bacteriol'1976; 127:1529-1537 17. Philips I, King BA, Shannon KP: The mechanisms of resistance to aminoglycosides in the genus Pseudomonas. J Antimicrob Chemother 1978;:121-129 18. Reimer LG, Stratton CW, Reller LB: Minimum inhibitory and bactericidal concentrations of antimicrobial agents against three standard control strains in broth with and without human serum. Antimicrob Agents Chemother 1981; 19:1050-1055 19. Reller LB, Schoenknecht FD, Kenny MA, Sherris JC: Antibiotic susceptibility testing of Pseudomonas aeruginosa: Selection of a control strain and criteria for magnesium and calcium content in media. J Infect Dis 197; 130:5-r63 20. Ruben FL, Norden CW, Hruska E: Factors associated with acquisition of Pseudomonas aeruginosa resistant to gentamicin. Am J Med Sci 1978;275:173-179 21. Seale TW, Thirkill H, Tarpay M, Flux M, Rennert OM: Serotypes and antibiotic susceptibilities of Pseudomonas aeruginosa isolates from single sputa of cysticfibrosispatients. J Clin Microbiol 1979; 9:72-78 22. Slack MPE: Antipseudomonal /3-lactams. J Antimicrob Chemother 1981;8:5-170 23. Steers E, Foltz EL, Graves BS: An inocula replicating apparatus for routine testing of bacterial susceptibility to antibiotics. Antibiot Chemother 1959; 9:307-311 2. Thomassen MJ, Demko CA, Boxerbaum B, Stem RC, Kunchenbrod PJ: Multiple isolates of Pseudomonas aeruginosa with differing antimicrobial susceptibility patterns frqm patients with cystic fibrosis. J Infect Dis 1979; 10:873-880 25. Weinstein RA, Nathan C, Gruensfelder R, Kabins SA: Endemic aminoglycoside resistance in gram-negative bacilli: Epidemiology and mechanisms. J Infect Dis 1980; 11:338-35 26. Woolfrey BF, Fox JMK, Quail CO: Susceptibility of Pseudomonas aeruginosa to cefoperazone, cefotaxime and moxalactam with special reference to isolates resistant to aminoglycosides, carbenicillin and ticarcillin. J Antimicrob Chemother 1981; 8:205-211