1186

New Antimicrobial Agent

Activity of Meropenem against Imipenem-Resistant Bacteria and Selection in Vitro of Carbapenem-Resistant Enterobacteriaceae L.J.V. Piddock*, H.L. T u r n e r

The activity of meropenem against 106 imipenem-resistant (MIC > 8 mg/l) clinical isolates, and the frequency of resistance to meropenem and imipenem among 24 Enterobacteriaceae was determined. Both agents selected colonies on agar but 20-80 % were susceptible after one subculture and 72 % of the mutants reverted to susceptibility 1 to 6 months after selection. All isolates and stable mutants were inhibited by > 1 rag/! meropenem, although the MIC of hnipenem was 4-16 mg/l. Three of six Xanthomonas maltophilta isolates were susceptible to meropenem (MICs 2-4 mgB). Pseudomonas aeruginosa lacking outer membrane protein D2 were resistant to meropenem, although isolates with substantially reduced expression of this protein were susceptible. None of the imipenem-resistant gram-positive bacteria were susceptible to meropenem. There was no clear correlation between altered outer membrane protein expression and decreased susceptibility to carbapenems, and there was no apparent involvement of plasmid or chromosomal 13-1actamase.

Meropenem (SM-7338) is a new carbapenem with a broad spectrum of antibacterial activity (15). It is slightly less active than imipenem against gram-positive bacteria, but more active against gram-negative bacteria. Meropenem binds initially to penicillin binding protein (PBP) 2 in Escherichia coli and to PBPs 2 and 3 in Pseudomonas aeruginosa at similar concentrations. In addition, like imipenem, meropenem binds to PBP 1 and 2 at concentrations below the MIC (6). Escherichia coli lacking outer membrane protein (Omp) F are as susceptible to meropenem and Antimicrobial Research Group, Department of Medical Microbiology, Medical School, University of Birmingham, Birmingham BI5 2"IT, UK.

Vol. 11. No. 12

imipenem as wildtype bacteria, and Salmonella typhimurium rfaC (rough lipopolysaccharide) are hyper-susceptible (6). Like imipenem the MIC of meropenem is unaffected by increased expression of chromosomal class 1 13-1actamases (1-9). The excellent in vitro gram-negative activity of carbapenems is attributed to the zwitterionic character of the agents resulting in rapid penetration of the outer membrane, little hydrolysis by periplasmic 13-1actamases and high affinity for essential PBPs. Resistance to imipenem has occurred mainly in pseudomonads. Xanthomonas maltophilia expresses an inducible zinc-requiring fi-lactamase and imipenem-resistant Pseudomonas aeruginosa usually have decreased expression of Omp D2 resulting in decreased uptake (10). Recently, a plasmid-mediated metallo-13-1actamase that hydrolyses imipenem and meropenem was described in a n isolate of Pseudomonas aeruginosa isolated in Japan (11). Xanthomonas maltophilia have been shown in one study to be resistant to meropenem (3), and in another study 50 % of the isolates were susceptible to 2 mg/l meropenem, whereas all isolates in both studies were imipenem-resistant (MICs > 8 mg/l). A 4- to 32-fold increase in the MIC of imipenem and meropenem (MICs of imipenem 16-32 mg/l and of meropenem 4-8 rag/l) was shown for Pseudomonas aeruginosa lacking Omp D2 which was selected with imipenem or meropenem (12,13). It is thought that meropenem has better activity due to higher affinity for the PBPs of Pseudornonas aeruginosa as both agents have similar permeability coefficients (10). Imipenem-resistant (MICs > 8 mg/l) Enterobacteriaceae are rare and have only been described for Serratia marcescens and Enterobacter cloacae (A.A. Medeiros, R.S. Hare; 26th Interscience Conference on Antimicrobial Agents and Chemotherapy, 1986, Abstract 116; 14). All isolates contained 15-1actamases that hydrolysed imipenem; however only two of the Serratia marcescens isolates were examined for susceptibility to meropenem and both were inhibited by 0.12 mg/l (14). It has been assumed that due to the low numbers of imipenem-resistant Enterobacteriaceae described in the literature that carbapenem therapy would be most likely to fail in patients infected with Pseudomonas aeruginosa. However, in recent imipenem disc susceptibility testing in the UK 36 of 1395 (2.58 %) clinical isolates of Proteus spp. were not susceptible to 8 mg/l im-

Vol. 11, 1992

ipenem (L.J.V. Piddock, R. Wise; 4th European Congress of Clinical Microbiology, 1989, Abstract 1166).

Materials and Methods. In this study 106 bacteria deemed imipenem-resistant (MIC _>8 mg/1) from the UK susceptibility survey were examined for susceptibility to meropenem (ICI Pharmaceuticals, UK); imipenem (Merck Sharpe and Dohme, UK); mecillinam (Leo Laboratories, Denmark); FCE 22101 (Farmitalia Carlo Erba, Italy); chloramphenicol (Warner Lambert-Parke Davis, UK); tetracycline (Lederle, UK); nalidixic acid (Sterling Research Group, UK). The MIC of all the agents was determined for each strain using a routine agar plate doubling dilution method, inocula of 104, 106 and 108 cfu, and aerobic incubation at 37 °C overnight. The study of frequency of mutation to resistance to meropenem and imipenem was performed with three clinical isolates (plus reference strain) of Proteus mirabilis (plus NCTC 10975), Proteus vulgaris (plus NCTC 4175), Morganella morganii (plus NCTC 235), Providencia stuartii (plus NCTC 10318), Citrobacter freundii (plus NCTC 9750) and Serratia marcescens (plus NCTC 10211) and antibiotic-containing agar at 3, 5, 8 and 10 times the MIC and 8 mg/l of each agent for each strain. Five agar plates for each antibiotic concentration were inoculated with bacteria that had been concentrated by centrifugation or diluted to give an inoculum range of 105 cfu-109 cfu. Ten randomly chosen colonies with the typical size and morphology of the wildtype parental strain were examined from each selecting concentration for susceptibility. One mutant of each observed resistance phenotype was retained. All mutants and parental strains were grown to late logarithmic phase and outer membrane proteins and IMactamase prepared essentially as described by Piddock et al. (15). All outer membrane protein samples were electrophoresed on two systems, 10 % and 14 %, of vertical SDSPAGE, using 20 ~tg protein per well. The crude [~-lactamase preparations were tested with 100 laM nitrocefin for the presence of 13-1actamase. For any strain with a iMactamase preparation that reacted faster with nitrocefin than the preparation from the parent wildtype, the Km (~tM) and Vmax (nmoles hydrolysed/ min/mg protein) for nitrocefin was obtained. All assays were performed at 370C using a spectrophotometer (Ultrospec, LKB, UK) with a heated cell and the data analysed as described elsewhere (15). Inducibility of chromosomal [5-

1187

lactamase was determined with the disc approximation test (16) and by growth of the strains in half the MIC of cefoxitin prior to determination of the MIC and the [$-lactamase assay.

Results and Discussion. "All nine Xanthomonas maltophilia were resistant to imipenem. Six isolates were also resistant to meropenem (> 8 mg/1), however the MIC of meropenem for two isolates was 2 rag/l, and for one isolate 4 mg/l (Table 1), confirming the finding of Edwards et al. (1) that not all Xanthomonas maltophilia are resistant to meropenem. The susceptibility of the gram-negative bacteria to imipenem was reduced with an increase in inoculum, such that at 108 cfu/spot the MIC was two- to four-fold higher. For nine isolates of Pseudomonas aeruginosa the MIC of imipenem increased to above the recommended breakpoint concentration (8 rag/l). The MIC of meropenem for the nine isolates lacking Omp D2 (identified by immunoblotting; J. Trias, personal communication) was 8-32 mg/l and for one isolate 4 mg/l. Four isolates with substantially reduced expression of Omp D2 were susceptible to 1-4 mg/l meropenem. Four isolates with a wildtype outer membrane protein profile but decreased susceptibility to imipenem due to an unknown mechanism (MIC of imipenem 2-64 mg/1, depending on the inoculum), had a corresponding MIC of 0.5-1 mg/1 meropenem. Imipenem-resistant isolates of Pseudomonas aeruginosa lacking Omp D2 have been shown to be cross-resistant to ciprofloxacin. Recently, Michea-Hamzepour et al. (17) showed an association between reduced expression of Omp D2 in two of three imipenem-resistant strains and quinolone resistance correlating with decreased uptake of [14C] imipenem or [14C] sparfloxacin. However, in our study only 4 of the 20 imipenemresistant Pseudornonas aeruginosa (one lacking Omp D2, two with decreased expression of Omp D2, and one with wild type expression) had MICs of ciprofloxacin greater than the recommended breakpoint concentration >_.1 mg/1. All 22 imipenem-resistant Enterobacteriaceae (7 Proteus mirabilis, 7 Morganella rnorganii, 4 Proteus vulgaris, 1 Proteus penneri and 3 Providencia stuartii, MICs of imipenem 8-32 mg/1 at 106 cfu) were susceptible to meropenem (MICs of meropenem < 0.12 mg/1). The 21 imipenemresistant Enterococcus faecium were resistant to meropenem, with meropenem MICs of > 64 mg/l for 19 isolates. The three imipenem-resistant Enterococcus faecalis isolates were also resistant to meropenem (MICs 64-128 mg/l). All methicil-

1188

Eur. J. Clin. Microbiol. Infect. Dis.

Table 1: Susceptibility of imipenem-resistant clinical isolates to imipenem and meropenem at an inoculum of 106 cfu/spot.

Species

M I C (mg/l)

(n) MICSO

MIC90

Range

Xanthomonas maltophilia

(9) imipenem meropenern

32 16

32 64

8 2

16 16

1 6 - 64 b 2 - > 128

Pseudomonas aeruginosa

(20) imipenem meropenem

1 - 16b 0 . 2 5 - 32

Enterobacteriaceae a (22) imipenem meropenem

4 0.06

4 0.12

1- 4b 0 . 0 3 - 0.12

Enterococcus faecium

(27) imipenem meropenem

32 128

64 256

0 . 5 - 128 1 - 256

1 4

1 8

0 . 5 - 64 2 - 128

Enterococcus faecalis (33) imipenem meropenem

a Includes 4 Proteus vulgaris, 1 Proteus penneri, 3 Providencia stuartii, 7 Morganella morganii, 7

Proteus mirabilis. blf the inoculum was increased to 108 cfu/spot, the MIC increased 2- to 4-fold.

lin-resistant Staphylococcus aureus isolates were resistant to imipenem and meropenem, It was difficult to select carbapenem-resistant

Enterobacteriaceae, the frequency of mutation to resistance being in the range 1.02 x 10-6 to 9.7 x 10-12 for imipenem and 1.2 x 10-6 to 5.38 x 10-12 for meropenem. More colonies of Proteus vulgaris grew on carbapenem-containing agar than of Providencia stuartii and Morganella morganii. The mutants that were selected were often unstable, 20 %-80 % of the selected colonies being susceptible to carbapenems after one subculture on antibiotic free agar (compared to 0-20 % selected with quinolones [18]). This suggests that the mutation was not favourable. Most of the 175 mutants (83 selected with meropenem and 92 selected with imipenem) had a four-fold decrease in susceptibility and were borderline for susceptibility or resistant to imipenem (MICs 4-16 mg/l) (Table 2). None were resistant to meropenem despite a two- to four-fold increase in the MIC (0.12-0.5 mg/l). Three mutants of a clinical isolate of Proteus mirabilis were inhibited by 16 mg/1 im-

ipenem. There was essentially no difference between the phenotypes of the mutants selected with either agent. Cross-resistance to FCE 22101 and mecillinam was seen in some mutants of two strains of Proteus mirabilis, one Morganella morganii and two Providencia stuartii; cross-resistance to mecillinam only was seen in mutants of three Citrobacter freundii. Three Proteus vulgaris, three Proteus mirabilis, two Morganella morganii, two Serratia marcescens and three Citrobacter freundii yielded mutants with a fourfold decrease in susceptibility to nalidixic acid, norfloxacin and ciprofloxacin. Only a few mutants were cross-resistant to chloramphenicol and tetracycline. During further studies it was noticed that many of the 175 mutants reverted to carbapenem-susceptibility, such that after six months storage on slopes at room temperature only 47 (28 %) mutants showed stable resistance (24 meropenem-selected, 23 imipenem-selected); 72 (42 %) mutants were inhibited at a concentration deviating plus/minus one dilution from that for

Vol, 11, 1992

1189

Table 2: Susceptibility to meropenem and imipenem of 175 carbapenem-resistant Enterobacteriaceae

selected in vitro compared to wildtype strains after one subculture on antibiotic-free agar. Strains

MIC (rag/l)

(n)

MIC50

MIC90

Range

meropenem imipenem

0.06 1

0.12 2

0.03- 0.12 0.5- 4

Meropenem-resistant (83) meropenem imipenem

0.25 8

0.5 8

0.06- 0.5 0.5- 16

Imipenem-resistant (92) meropenem imipenem

0.12 4

0.5 8

0,03- 1 2 - 16

Wildtype

(24)

45kD 30kD

17.3kD 12.4kD J31

J31/1

J31/2

J31/3

Figure 1: Outer membrane protein profile of Proteus mirabilis NCTC 10975 and mutants. Low molecular weight proteins decreased in expression or absent are marked with an arrow. The molecular weight standard is shown on the left. the wildtype strain and 51 (30 %) were fully susceptible. Five mutants died. After re-checking antimicrobial susceptibility, outer m e m b r a n e proteins and [3-1actamase preparations were made from the 47 stable mutants (14 Proteus vulgaris derived from 3

strains, 14 Proteus mirabilis from 1 strain, 8 Serratia marcescens from 4 strains, 6 Citrobacter freundii from 3 strains, 2 Providencia stuartii from 2 strains, 2 Morganella morganii from 1 strain). Four Proteus vulgaris mutants, four Proteus mirabilis mutants and one Providencia stuartii mutant lacked outer m e m b r a n e proteins with molecular weights similar to those of the porin proteins described for each species (40 kDa, 35 kDa and 36 k D a respectively). In addition, two high molecular weight proteins were absent in one Proteus vulgaris (83 kDa and 85 kDa) and one Providencia stuartii (100 k D a and 104 kDa). Other mutants lacked an outer m e m b r a n e protein of around 18 kDa (Figure 1; one Proteus mirabilis, 17 kDa; one Serratia marcescens 19 kDa; one Citrobacter freundii 18 kDa). T h r e e Providencia stuartii mutants (from two strains) had a new outer m e m b r a n e protein of 35 kDa. Despite a stable resistance phenotype, no consistent modification in expression of outer m e m b r a n e proteins correlated with carbapenemresistance. In addition, none of the mutants had mutationally derepressed class 1 ff-lactamase (data not shown). Some strains contained no detectable ff-lactamase despite exposure to half the MIC of cefoxitin. The mechanism of decreased susceptibility to carbapenems in the stable mutants is unclear and requires further investigation. However, the data from this study suggests that expression of proteins analogous to Pseudomonas aeruginosa O m p D2 was not altered, The role of the altered expression of low

1190

Eur. J. Clin. Microbiol. Infect. Dis.

molecular weight o u t e r m e m b r a n e proteins (possibly O m p X [19]) and high molecular weight outer m e m b r a n e proteins (possibly iron-regulated outer m e m b r a n e proteins) in c a r b a p e n e m resistance also requires further study. In addition, increased expression of plasmid or c h r o m o s o m a l 13-1actamases was not detected in any strain, and there was no suggestion of point mutations changing the s p e c t r u m of the I~-lactamases to include c a r b a p e n e m s as no hydrolysis was detected (data not shown).

3. Jones RN, Aldridge KE, Allen SD, Barry AL, Fuchs PC, Gerlack EH, Pfaller MA: Multicenter in vitro

In Enterobacter cloacae it has b e e n shown that i m i p e n e m resistance is due to a combination of reduced outer m e m b r a n e permeability due to the lack of the two m a j o r outer m e m b r a n e proteins ( p r e s u m a b l y porins) and high level class 1 13-1actamase expression (20, 21). H o w e v e r , of two mutants of Proteus rettgeri, one m u t a n t selected with m e r o p e n e m lacked the m a j o r o u t e r m e m b r a n e protein and had low outer m e m b r a n e permeability but showed no difference in 13-1act a m a s e expression (21). T h e other mutant, selected with imipenem, expressed wildtype o u t e r m e m b r a n e proteins and had no change in I~-lactamase expression, however outer m e m b r a n e permeability was decreased comp a r e d to the wildtype strain by an u n k n o w n mechanism. In addition, c a r b a p e n e m resistance was unstable (21). Likewise, in the present study, the m e c h a n i s m of decreased susceptibility to c a r b a p e n e m s in the mutants was unstable and apparently not attributable to outer m e m b r a n e protein changes or 13-1actamase activity. It may be that the m e c h a n i s m is associated with decreased accumulation of c a r b a p e n e m s , p e r h a p s due to altered lipopolysaccharide, or it m a y be due to altered target site (PBP).

6.

In conclusion, m e r o p e n e m can be expected to be active against imipenem-resistant Enterobacteriaceae, s o m e X a n t h o m o n a s maltophilia and some P s e u d o m o n a s aeruginosa but not against imipenem-resistant gram-positive bacteria.

References

1. Edwards JR, Turner PJ, Wannop C, Withnell ES, Grlndley A J, Nairn K: In vitro antibacterial activity

of SM-7338, a earbapenem antibiotic with stability to dehydropeptidase I. Antimicrobial Agents and Chemotherapy 1989, 33: 215-222. 2. Sumita Y, Inoue M, Mitsuhashi S: In vitro antibacterial activity and beta-lactamase stability of the new carbapenem SM-7338. New Antimicrobial Agents 1989, 8: 908-916.

4.

5.

7.

evaluation of SM-7338, a new carbapenem. Antimicrobial Agents and Chemotherapy 1989, 33: 562565. King A, Boothman C, Phillips I: Comparative in-vitro activity of meropenem on clinical isolates from the United Kingdom. Journal of Antimierobial Chemotherapy 1989, 24, Supplement A: 31--46. Bauernfeind A, Jungwirth R, Schweihart S: In vitro activity of meropenem, imipenem, and penem HRE 664 and eeftazidime against clinical isolates from West Germany. Journal of Antimicrobial Chemotherapy 1989, 24, Supplement A: 73--84. Kitzis MD, Acar JF, Gutmann L: Antibacterial activity of meropenem against gram-negative bacteria with a permeability defect against staphylococci. Journal of Arttimicrobial Chemotherapy 1989, 24, Supplement A: 125-132. Chanal C, Sirot D, Chanal M, Ouzel M, Sirot J, Ouzel R: Comparative in-vitro activity of meropenem against clinical isolates including Enterobacteriaceae with expanded spectrum 15-1actamases. Journal of Antimicrobial Chemotherapy 1989, 24, Supplement A: 133142.

8. Sanders CC, Sanders WE, Thomson KS, laconis JP:

Meropenem: activity against resistant gram-negative bacteria and interactions with I~-Iactamases. Journal of Antimicrobial Chemotherapy 1989, 24, Supplement A: 187-196. 9. Yang Y, Livermore DM: Interactions of meropenem with class I chromosomal I~-lactamases. Journal of Antimicrobial Chemotherapy 1989, 24, Supplement A: 207-218. 10. Trias J, Nikaido H: Outer membrane protein D2 catalyzes facilitated diffusion of carbapenems and penems through the outer membrane of Pseudornonas aeruginosa. Antimicrobial Agents and Chemotherapy 1990, 34: 52-57. 11. Watanabe M, Iyobe S, Inoue M, Mitsuhashi S: Transferrable imipenem resistance in Pseudomonas aeruginosa. Antimicrobial Agents and Chemotherapy 1991, 35: 147-151. 12. Livermore DM, Yang Y: Comparative activity of meropenem against Pseudomonas aeruginosa strains with well-eharacterised resistance mechanisms. Journal of Antimicrobial Chemotherapy 1989, 24, Supplement A: 149-160. 13. Margaret BS, Drusano GL, Standiford HC: Emergence of resistance to carbapenem antibiotics in Pseudomonas aeruginosa. Journal of Antimicrobial Chemotherapy 1989, 24, Supplement A: 161-168. 14. Yang Y, Wu P, Livermore DM: Biochemical characterisation of a 15-1actamasethat hydrolyses penems and carbapenems from two Serratia marcescens isolates. Antimicrobial Agents and Chemotherapy 1990, 34: 755-758. 15. Piddock LJV, Traynor EA, Wise R: A comparison of the mechanism of decreased susceptibility of aztreonam-resistant and ceftazidime-resistant Enterobacteriaceae. Journal of Antimicrobial Chemotherapy 1990, 26: 749-762. 16. Sanders CC, Sanders WE: Emergence of resistance to cefamandole: possible role of cefoxitin-inducible 15lactamase. Antimicrobial Agents and Chemotherapy 1979, 15: 792-797.

Vol. 11, 1992

17. Michea-Hamzepour M, Furet XY, Pechere JC: Role of protein D2 and lipopolysaccharide in diffusion of quinolones through the outer membrane of Pseudomonas aeruginosa. Antimicrobial Agents and Chemotherapy 1991, 35: 2209-2217. 18. Piddock LJV, Hall MC, Waiters RN: Phenotypic characterisation of quinolone-resistant mutants of Enterobacteriaceae selected from wild-type gyrA and multiply resistant (marA) type strains. AntimicrobiaI Agents and Chemotherapy 1991, 28: 185-198.

1191

19. Stoorvogel J, van Bussel MJ, Tommassen J, van de Klundert JAM: Molecular characterisation of an Enterobacter cloacae outer membrane protein (OmpX). Journal of Bacteriology 1991, 173: 156-160. 20. Lee EH, Nicolas MH, Kitzls MD, Pialoux G, Collatz E, Gutmann L: Association of two resistance mechanisms in a clinical isolate of Enterobacter cloacae with high-level resistance to imipenem. Antimicrobial Agents and Chemotherapy 1991, 35: 1093-1098, 21. Raimondi A, Traverso A, Nikaido H: Imipenem- and meropenem-resistant mutants of Enterobacter cloacae and Proteus rettgeri lack porins. Antimicrobial Agents and Chemotherapy 1991, 35: 1174-1180.

Activity of meropenem against imipenem-resistant bacteria and selection in vitro of carbapenem-resistant Enterobacteriaceae.

The activity of meropenem against 106 imipenem-resistant (MIC > or = 8 mg/l) clinical isolates, and the frequency of resistance to meropenem and imipe...
572KB Sizes 0 Downloads 0 Views