Journal of Antimicrobial Chemotherapy (1992) 30, 587-596
Biochemical basis of mnpirodn resistance in strains of Staphylococcus awreus T. H. Fanner, J. GObart* and S. W. Ebon
Twenty one strains of Staphylococcus aureus, of varying resistance to mupirocin, were examined in order to determine the mechanism of resistance to this antibiotic; six of these strains were mupirocin sensitive (MIC 0-12-1-0 mg/L) nine moderately resistant strains (MIC 8-256 mg/L) and six highly resistant strains (MIC > 2048 mg/L). Mupirocin showed a time-dependent inhibition of the target enzyme, isoleucyl-tRNA synthetase (IRS); incubation of the antibiotic with this enzyme before adding the substrates markedly increased inhibition in sensitive strains. The IRS I n values (the antibiotic concentrations which cause a 50% decrease in enzyme activity) correlated well with the MIC values for each strain {P < 001). The mean lx value for sensitive strains was 3-3 x 10~2 mg/L, in moderately resistant strains it was 1-3 x 10~' mg/L and in highly resistant strains it was 7-5 mg/L. No degradation of mupirocin could be detected during extended incubation of the antibiotic with cell free extracts from four resistant S. aureus strains. We conclude that the production of a modified IRS enzyme is the major cause of mupirocin resistance in the strains studied.
Introduction Mupirocin (pseudomonic acid A), an antibiotic produced by Pseudomonas fluorescens, has activity against many Gram-positive and some Gram-negative bacteria (Sutherland et al., 1985). Early work with Escherichia coli (Hughes & Mellows, 1980) demonstrated that the target of mupirocin is bacterial isoleucyl-tRNA synthetase (IRS), the enyzme which charges the appropriate tRNA with isoleucine, an essential component of protein synthesis. This process is accomplished in two distinct steps as illustrated below (E = IRS enzyme): 1. Isoleucine + ATP+E ?± E-Isoleucine-AMP + PPi 2. E-Isoleucine-AMP + tRNA -»Isoleucine-tRNA -I- AMP + E Mupirocin has been shown to inhibit only the first of these reactions and it was postulated that part of the antibiotic molecule resembles isoleucine; the antibiotic competing for the active site of the enzyme (Hughes & Mellows, 1980; Mellows, 1987). This possible mechanism of action however, remains to be more fully investigated as other parts of the mupirocin molecule have been implicated with its interaction with the enzyme (Rechsteiner & Leisinger, 1989). 'Corresponding author. 587 0305-7453/92/110587 +10 S08.00/0
© 1992 The Briuih Society for AntimicrobUl Chemotherapy
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
SmithKline Beecham Pharmaceuticals, Brockham Park, Betchworth, Surrey, RH3 7AJ, UK
588
T. H. Fanner et aL
Methods Bacterial strains A total of 21 5. aureus strains were used in this study comprising 18 isolates from patients, supplied by clinical laboratories and three reference isolates from the National Collection of Type Cultures (NCTQ. The sources of these strains, their mupirocin MIC values and resistance phenotypes are given in Table I. Mupirocin sensitive strains are defined as having MIC values in the range, 0-12-4-0 mg/L. Moderately resistant, strains have MIC values in the range, 8-256 mg/L. Highly resistant strains have MIC values of 512 mg/L or greater. The MIC values were determined by an agar dilution method (Washington, 1985). Preparation of bacterial enzyme extracts Cell free extracts of S. aureus were used for both degradation studies and isoleucyl tRNA synthetase (IRS) assays and were prepared using a procedure similar to that described by Hughes, Mellows & Sough ton (1980) in their studies on E. coli and P. fluorescens. Starter cultures of each S. aureus strain were prepared by inoculating 50 mL volumes of nutrient broth No. 2 (Oxoid and incubated at 37°C with shaking (240 rpm) for 18 h. A 200 mL volume of nutrient broth was inoculated with 2 mL of culture from the starter culture and grown to late log phase (6 h) under the same conditions. The cells were harvested by centrifugation at 5000 g at 4°C and the pellet resuspended in 5 mL cold 0-1 M Tris/HCl buffer pH 7-5 containing 10 mM magnesium
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
Used as a topical agent, mupirocin has proved to be an effective treatment for staphylococcal infections (Casewell & Hill, 1987) and is active against methicillin resistant strains (Sutherland et al., 1985). More recently however, some clinical mupirocin resistant Staphylococcus aureus strains have been reported from hospitals in the United Kingdom (Rahman, Noble & Cookson, 1987) and the USA (M. Carter, personal communication). Mupirocin resistant strains are rarely isolated!; in a recent survey in the UK, only 0-3% were resistant to 2 mg/L mupirocin (Slocombe & Perry, 1991). Clinical strains of S. aureus have been isolated which were moderately resistant to mupirocin, (MIC 2-32 mg/L) compared with < 0-2 mg/L for sensitive strains (Baird & Coia, 1987; Kavi, Andrews & Wise, 1987). In addition, highly resistant strains (MIC > 700 mg/L) have been isolated from patients in a London hospital (Rahman, et al., 1987) and from a Bristol hospital (MIC > 1025 mg/L) (Smith & Kennedy, 1988). In terms of mupirocin resistance therefore, S. aureus strains appear to fit into three groups; the majority are sensitive (99-7%), a few are moderately resistant (0-25%) and still fewer are highly resistant (0-05%) (Slocombe & Perry 1991). In some strains of the latter group the resistance has been shown to be transferable (Rahman 1989). The number of resistant strains isolated remains low but a more detailed knowledge of this resistance would be desirable. The purpose of this study was therefore to examine a representative set of S. aureus strains which included isolates, sensitive, moderately and highly resistant to mupirocin, in an attempt to elucidate the resistance mechanism. Two possible mechanisms were considered, firstly, enzyme degradation (or modification) of mupirocin and secondly, production of a modified bacterial isoleucyltRNA synthetase.
ftfupirocin resistance
589
Spectrophotometric and HPLC assays of mupirocin stability Enzymic degradation or modification was sought using both direct spectrophotometric and HPLC assays. For the spectrophotometric assay mupirocin concentrations of 5 or 10 mg/L were used and changes in absorbence at 223 nm were monitored. The antibiotic was dissolved in 0-05 M sodium phosphate buffer, pH 7-3, warmed to 37°C, then staphylococcal extracts (equivalent to 1 mg/mL dry weight of bacterial cells) from Table L Sources and antibiotic resistance phenotypes of S. aureus strains used in this study
Strain NCTC 6571 NCTC 11561 MCI NCTC 10442 MC2 OC1 OC2 JN1 DB1 DAI MC3 MC4 CR1 PR1 SP1 DB2 MC5 WN1 WN2 WN3 WN4
Source reference strain reference strain skin isolate, New York, USA reference strain skin isolate, New York, USA skin isolate, Romford, UK skin isolate, Romford, UK skin isolate, Glasgow, UK skin isolate, Glasgow, UK skin isolate, Romford, UK skin isolate, New York, USA skin isolate, New York, USA skin isolate, Crawley, UK (kin isolate, Warrington, UK skin isolate, Bristol, UK skin isolate, Glasgow, UK skin isolate, New York, USA skin isolate, London SE., UK plasmid cured variant of WN1 skin isolate, London SE., UK plasmid cured variant of WN3
Mupirocin MIC (mg/L) 0-12 0-25 0-5 1 8 16 16 16 32 32 64 128 256 >2048 >2048 >2048 >2048 >2048 0-25 >2048 0-25
Resistance phenotype PenG11 PenG11 PenG11, TetR, Mec* PenG11 PenG*, Mec* PenG11, Mec* TetR, Ery*, Fuc* PenG\ TetR, Ery*, Fuc* PenGR, TetR, Ery*, Fuc* PenG11, TetR PenGR, TetR, Ery* PenGR, TetR, Ery*, Mec* PenGR, TetR, Ery*, Mcc* PenG* PenGR, TetR, Ery* PenG\ TetR, Ery* PenGR, TetR, Ery" PenGR, TetR, Ery"
R, Rettstant; PenG, penicillin O; Tet, tetracydine; Ery, erythromycin; Mec, methicilliii; Fuc, fundk acid.
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
chloride and 2 mM dithiothrcitol (omission of dithiothreitol resulted in substantial loss of enzyme activity). The cells were disrupted by sonication with a Soniprep ISO sonicator (MSE Loughborough, UK) for 8 x 30 s periods with IS s intervals for cooling. The sample vessel was cooled in ice/water during disruption. In some later experiments the activity of the IRS extract was found to be improved by adding lysostaphin (Sigma Poolc, UK), to a final concentration of 10 mg/L to the resuspended cells and shaking at 240 rpm at 37°C for 30 min prior to sonication. Comparative experiments indicated that lysostaphin treatment did not alter the inhibition of IRS activity by mupirocin. The suspensions were then recentrifuged at 15,000 g at 4°C for 20 min and the pellets discarded. The supernatants were treated with 10 mg/L ribonuclease free deoxyribonuclease (Sigma, Poole, UK) and incubated at 4°C for 20 min. These samples were then centrifuged at 100,000 g at 4°C for 3 h and the supernatants dialysed against the above buffer for 16 h at 4°C. Glycerol was then added to each sample to give a final concentration of 25% and the samples were mixed gently before storage at - 7 0 ° C .
590
T. H. Fanner et aL
Radiometric assay of isoleucyl-tRNA synthetase (IRS) inhibition The activity of IRS and the degree of its inhibition by mupirocin was determined using a radiometric assay modified from a previously described method (Hughes & Mellows, 1980). This procedure is based on the incorporation of [14C]isoleucine into tRNA which, at the end of the reaction, is precipitated and the radio-activity counted by scintillation as described below. In the presence of mupirocin however, this incorporation of radio-active label is diminished which results in a lower count. A 40 /xL volume of each S. aureus IRS enzyme extract was added to 10 uL of inhibitor (or water for the control) and 50 uL of a reagent mixture containing 12 uM L-[U-l4C]isoleucine, specific activity: 12-0 GBq/mmol (Amersham), 10 g/L bulk tRNA (Sigma Poole, UK) and 10 mM ATP in 100 mM Tris buffer pH 8-3, supplemented with 10 mM MgCl2 and 2 mM dithiothreitol. This reaction mixture was then incubated at 37°C for exactly 10 min. The antibiotic was either mixed with the other reactants prior to adding the isoleucyltRNA synthetase (to measure the effect without prc-incubation), or incubated with the enzyme for 5 min then the other reagents added (to determine the effect of preincubation). To stop the reaction and precipitate the isoleucyl-tRNA complex, 2-5 mL 7% trichloroacetic (TCA) acid was added and the vials left on ice for 20 min. The suspensions were then filtered through 25 mm GF/C filter papers (Whatman Maidstone, UK) which were washed with 10 mL TCA and 10 mL ethanol, then dried and placed in scintillation vials with 5 mL OptiPhase 'Safe' scintillant fluid (Pharmacia Milton Keynes, UK) and counted in a Tri-carb 460 liquid scintillation counter. The IRS inhibition for each antibiotic concentration was calculated by expressing the reduction in activity as a percentage of that in the control sample (where water was substituted for mupirocin). The mupirocin concentration required to give a 50% reduction in activity (Iy, value) was then derived from a plot of mupirocin concentration versus percent inhibition, as shown in Figure 1. Initially, approximate mupirocin Ijo values were determined using a ten-fold dilution range (between final concentrations of 10 mg/L and 1 pg/L) of antibiotic in the above reactions. For more precise assays, doubling dilution series of mupirocin were used in which the antibiotic concentrations were in appropriate ranges above and below the approximate IM value for each strain.
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
strains SP1, WN1, WN3 and OC1 were each added to separate mupirocin solutions and monitored at intervals over a period of several hours. Controls containing the same concentrations of mupirocin in buffer only were incubated under the same conditions and aliquots were withdrawn for HPLC analysis at the same time intervals. Peak areas in the resulting chromatograms were compared with those of samples where bacterial extract was included. Assay of mupirocin by HPLC was carried out using a C,, /j-Bondapak reverse phase column (Waters Watford, UK), eluted with 60% methanol, 40% 0-05 M ammonium acetate, pH 4-5. At a flow rate of 1 mL/min, mupirocin eluted after 15 min. Initially the same incubation conditions were used as for the direct assay, but subsequently, the effect of adding ATP to a concentration of 2-5 mM was examined for some preparations. The purpose of ATP addition was to investigate the possibility that mupirocin could be inactivated by phosphorylation of its hydroxyl groups as has been observed with some aminoglycosides (Naber et al., 1990). In addition, mupirocin (10 mg/L) was added to cultures of S. aureus strains, NCTC 11561, SP1 WN1 and WN3 and the concentration in the culture broth was measured at intervals by HPLC during a 5 h period.
Mupirocin resistance
591
Figure 1. Mupirocin inhibition curves for IRS enzyme extracts from three S. aureus strains: A, NCTC 6571 (mupirocin-semitive); B, JN1 (moderately resistant to mupirocin); C, WN3 (highly resistant to mupirocin).
All the values that are shown in Tables II and m were derived from these more accurate assays. Results Spectrophotometric and HPLC assays of mupirocin stability Neither spectrophotometric nor HPLC assays indicated detectable losses or modification of mupirocin after treatment with S. aureus cell extracts, whether these were from moderately resistant or highly resistant strains. Adding ATP did not stimulate loss or modification of the antibiotic. Similarly, when bacteria were grown in the presence of mupirocin, analysis of culture supernatants showed little if any ( < 10%) loss of this compound. Radiometric assays of IRS inhibition In an initial series of assays the action of mupirocin on IRS preparations from a limited set of six 5. aureus strains was examined. In Table II the MIC values and corresponding IJO values are shown for this set of strains. A general increase in lx value with MIC can be seen. Figure 1 shows some typical mupirocin inhibition curves for a sensitive, moderately resistant and a highly resistant strain. The lx values for all strains were derived from similar plots. At low mupirocin concentrations the enzyme activity for highly resistant strains in some cases appeared to increase before it decreases at higher levels. This could be the result of stabilization or activation of the enzyme by ligand binding, a function which has been previously reported for this and other enzymes (Baldwin & Berg, 1966). In order to determine the nature of the inhibition, the IRS activity assays were performed with and without pre-incubation of the enzyme and inhibitor. In Table II it
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
1-0 10-0 I00O 1000-0 10000-0 Mupirocin concentration (^xg/L)
592
T. H. Fanner et *L Table II. Inhibition by mupirocin of isoleucyl-tRNA synthetase activity in extracts from various S. aureus strains. Extracts were assayed with and without pre-incubab'on with mupirocin (mg/L) Strain NCTC 11561 WN2 OC1 WN1 WN3 SP1
MIC (mg/L)
No pre-incubation
Pre-incubation
0-25 0-12 16 >204« >2048 >2048
3-6x10-' 9-7x10"' llxlO"' 6-7 71 71
2-5 xlO" 4 1-8 x l 0 " J 3-0x10"' 7-3 1-3 1-3
Table m. Inhibition by mupirocin of isoleucyltRNA tynthetase activity from 21 S. aureus strains without pre-incubation with mupirocin • Strain NCTC 6571 NCTC 11561 WN2 WN4 MCI NCTC 10442 MC2 OC1 OC2 JN1 DB1 DAI MC3 MC4 CR1 PR1 SP1 DB2 MC5 WN1 WN3
MIC (mg/L)
I , (mg/L)
0-12 0-25 0-25 0-25 0-5 1 8 16 16 16 32 32 64 128 256 >2O48 >2048 >2048 >2048 >2048 >2048
8-7 x 10"' 7-0x10"' 2-9 xlO" 3 21 x 10"2 5-3 xlO" 2 3-6 xlO" 2 1-2x10-'2 1-9 xlO" 6-5 xlO" 2 7-5 x 10"J 6-9 x 10"2 9-2 xlO" 1 2-2x10-' 1-5x10-' 5-6x10-' 6-0 6-7 90 10O 100 1CK)
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
is apparent that pre-incubation increases the inhibitory action of mupirocin against IRS from the sensitive strain (NCTC 11561). This indicates that binding to the mupirocin sensitive enzyme is time dependent and suggests that the antibiotic is not a classical competitive inhibitor of the staphylococcal enzyme. In a competitive system pre-incubation with the inhibitor would have had little or no effect since irreversible or strong binding would not take place. In a further set of assays a larger number of strains were included, where, to simplify the procedure, pre-incubation was not performed. Table III shows the results of mupirocin IM determinations on this set of 21 S. aureus strains; the correlation of IRS inhibition and MIC is still clearly apparent. The overall correlation coefficient (r) for the data listed in Table III is 0-90 which indicates good linearity and is statistically
Mnptrodn resistance
593
lOOOO
l 000
'
3 O-IOO
0-O10
0-1
IO
10-0
I0O0 MC (mg/L)
1000-0 I0000O
Hgmc 2. Plot of mupirocin MIC value* vcnut. IRS enzyme I n values for the 21 5. aureus itraint included in this study, the plot shows the apparent correlation between these two parameters. Since MIC values of greater than 2048 mg/L could not be accurately determined, the points representing very resistant strains could not be placed on the graph with any certainty and are thus separated in the boxed area. These points were not included in the computed hne of best fit
significant (P < 0-01). A graphical illustration of this correlation is given in Figure 2. It can be seen that IRS extracts from the highly resistant strains were much less susceptible to inhibition by mupirocin than those from sensitive strains. Furthermore, the IJO values for mupirocin against the IRS extracts from the moderately resistant strains occupy an intermediate position between the values for the enzymes from the sensitive and highly resistant strains. For most strains we observed no correlation between level of IRS activity (i.e. the amount of isoleucyl-tRNA formed after 10 min) and resistance. The activity of each IRS enzyme preparation was variable despite the use of uniform quantities of bacterial cells for enzyme preparation and ensuring constant assay conditions. The protein • concentrations in the bacterial lysates after sonication were found to be fairly constant suggesting that this variability of IRS activity was not the result of inconsistent bacterial cell breakage. In two formerly resistant strains (WN2 and WN4), which had been made sensitive by culture at 42°C resulting in curing of plasmids, the IRS activities were reduced 20% and 58% in relation to their parental strains (WN1 and WN3, respectively). Strains WN2 and WN4 were isogenic with their uncured counterparts, WN1 and WN3 and thus were more suitable for comparison than the other unrelated isolates. Resistance was transferable in these strains. Another highly mupirocin resistant strain (SP1), was reproducibly shown to have 50% or lower IRS activity than the other highly resistant strains. No transferable plasmids encoding mupirocin resistance could be detected in SP1 and the resistance gene was believed to have been incorporated into the chromesome (C. Perry, personal communication). Discussion
Based on the HPLC and spectrophotometric assays, it appears unlikely that the strains included in this study owe their mupirocin resistance to overt modification or degradation of the antibiotic. These findings are in agreement with a recent report in which,
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
0-001
594
T. H. Fanner et aL
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
using growth inhibition studies and HPLC assays, no reduction in antibacterial activity or mupirocin concentration could be detected after extensive incubation of the antibiotic with the IRS enzyme extract (Cookson, 1989). The good correlation between mupirocin MIC and IRS IM reported above, suggests that the major difference between sensitive and resistant strains lies within the IRS enzyme itself. The effect of preincubating the IRS enzyme with mupirocin, as shown in Table I, indicates that this antibiotic is not a classical competitive inhibitor, the activity of the IRS was decreased if the antibiotic was mixed with it and incubated for a short period before adding the substrates. This effect was more marked with enzymes from more sensitive strains. A competitive inhibitor should not show this effect. In a previous report however (Hughes & Mellows, 1980), mupirocin was described as competitively inhibiting the E. coli IRS enzyme although a stable enzyme-inhibitor complex was observed. These findings indicate that the inhibitor is tightly bound but the exact nature of the binding is not fully understood. It is not currently known whether mupirocin resistant S. aureus strains produce a totally modified IRS enzyme concurrently with the sensitive enzyme or whether the resistant enzyme only is present. The slightly higher IRS activities of the two mupirocin resistant strains, WN1 and WN3 compared to the heat treated, sensitive revertants, WN2 and WN4, suggest that there could be two enzymes present whose activities may be additive. The low level of activity seen in the non-plasmid containing resistant strain, SP1 is further evidence for this additive effect Our data indicate however, that hyperproduction of the IRS enzyme is an unlikely resistance mechanism; much greater production of sensitive IRS enzyme than was observed would be required to overcome the effect of mupirocin. The genetic location of mupirocin resistance has been previously investigated. In a study using the archaebacterial species Methanobacterium thermoautotrophicum (Kiener, Rechsteiner & Leisinger, 1986), the inhibition of growth and IRS activity by mupirocin were similar to those for S. aureus reported above. The inhibition in M. thermoautotrophicum could be overcome by the addition of exogenous isoleucine. These authors used nitrosoguanidine to induce a mutation conferring mupirocin resistance and found that the IRS enzyme in the mutant had become less sensitive to the antibiotic. They suggested that this resistance was a result of a change in the IRS structural gene on the chromosome. In highly resistant S. aureus however, available evidence indicates that the resistance resides on a plasmid (Noble et al., 1988). High level mupirocin resistance in 5. aureus has been shown to be transferable at low frequency in conjugation experiments where highly resistant and sensitive clinical strains were mated on niters (Rahman, et al., 1989). Initial findings suggested that the resistance was located on a very large plasmid which was difficult to resolve from chromosomal DNA. Subsequently, mupirocin resistance in S. aureus has been located on plasmids ranging in size from 23 to 87 kb (Rahman et al., 1990). In more recent work the mupirocin resistance gene from an S. aureus strain was cloned from a small (25 kb) plasmid (Dyke et al., 1991). Mupirocin resistance can be eliminated from S. aureus by growth at 42°C, as was demonstrated in strains WN1 and WN3 in this study. This observation is consistent with the resistance being plasmid encoded since higher growth temperatures can cause curing. The plasmids involved in mupirocin resistance are of a novel type, their structure and epidemiology require further investigation (Rahman et al., 1989). Isolates of S. aureus can also be rendered resistant by culture in media containing successively higher mupirocin concentrations (Casewell
Mupfructn restasnce
995
Acknowledgements We wish to thank Dr R. Sutherland and Mr B. Slocombe for their assistance and helpful discussions. We also thank Mr. I. Macpherson for statistical assistance and the staff of the Anti-infective Product Support Unit, Brockham Park, for supplying the strains used in this study, and for determining the MICs. We also thank Dr. M. Carter of the Rockefeller Institute, New York, NY, USA, for supplying further clinical strains of S. aureus. References Baird, D. & Coia, J. (1987). Mupu-orin-resistant Staphyloccus aureus. Lancet li, 387-8. Baldwin, A. N. & Berg, P. (1966). Purification and properties of isoleucyl ribonucleic acid synthetase from Escherichia coli. Journal of Biological Chemistry 241, 831-8. Capobianco, J. O., Doran, C. C. & Goldman, R. C. (1989). Mechanism of mupirocin transport into sensitive and resistant bacteria. Antimicrobial Agents and Chemotherapy 33, 156-63. Casewell, M. W. &. Hill, R. L. R. (1985). In-vitro activity of mupirocin ('pseudomonic acid*) against clinical isolates of Staphylococcus aureus. Journal of Antimicrobial Chemotherapy 15, 523-31. CaseweH, M. W. & HUT, R. L. R. (1987). Mupirocin ("pseudomonic acid")—fl promising new topical antimicrobial agent Journal of Antimicrobial Chemotherapy 19, 1-5. Cookson, B. (1989). Failure of mupirocin-resistant staphylococci to inactivate mupirocin. European Journal of Clinical Microbiology and Infectious Diseases 8, 1038-40.
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
& Hill, 1985) The mechanism of this type of resistance is unknown. Although this phenomenon produces stable resistance, it is acquired slowly and the antibiotic levels achievable by topical therapy are substantially greater than the MIC seen. Resistance to mupirocin could alternatively result from alterations in membrane permeability. In a recent study by Capobianco, Doran & Goldman (1989) the mode of entry of mupirocin into bacterial cells including resistant and sensitive S. aureus was investigated. These authors concluded that mupirocin uptake into the cytoplasm was energy independent and recorded no differences in membrane penneability to the antibiotic between a resistant and a sensitive strain. From thesefindingsthey made the unusual conclusion that binding of the antibiotic to the IRS enzyme was the force responsible for development of high cytoplasmic concentrations of mupirocin in sensitive cells. In a resistant strain obtained by culture in media containing increasing concentrations of mupirocin, they speculated that the lower affinity of the IRS enzyme for the antibiotic reduced this attractive force to some extent and thus the intracellular concentration remained low. They also postulated that some other unknown mechanism may be operating. Although these findings suggest the IRS enzyme has a role in mupirocin resistance, further data on a greater number of S. aureus strains is required to support this novel hypothesis and to rule out the possible involvement of membrane permeability in the process. Mupirocin resistant clinical isolates of S. aureus remain a comparative rarity (Slocombe & Perry, 1991). Resistant isolates can however spread locally within hospital wards causing problems with treatment Understanding the mechanism resistance could aid epidemiological studies and establish a basis for the development of novel antibiotics of a similar type. The structure and function of the IRS enzyme in S. aureus requires further study as this report shows that alteration of the IRS enzyme is the cause of mupirocin resistance in S. aureus.
596
T. H. Fanner et *L
(Received 21 October 1991; revised version accepted 6 July 1992)
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
Dyke, K. G. H., Curnock, S. P., Golding, M. & Noble, W. C. (1991). Cloning of the gene conferring resistance to mupirocin in Staphylococcus aureus. FEMS Microbiology Letters 77, 195-8. Hughes, J. & Mellows, G. (1980). Interaction of pscudomonic acid A with Escheriehia coli B isoleucyl-tRNA synthetase. Biochemical Journal 191, 209-19. Hughes, J., Mellows, G. & Soughton, S. (1980). How does Pseudomonasfluorescent,the producing organism of the antibiotic pseudomonic acid A, avoid suicide? FEES Letters 122, 322-4. Kavi, J., Andrews, J. M. & Wise, R. (1987). Mupirocin-resistant Staphylococcus aureus. Lancet ii, 1472. Kiener, A., Rechsteiner, T. & Leisinger, T. (1986). Mutation to pseudomonic acid resistance of Methanobacterium thermoautotrophicum leads to an altered isoleucyl-tRNA synthetase. FEMS Microbiology Utters 33, 15-8. Mellows, G. (1987). The mode of action of mupirocin: inhibition of bacterial isoleucyl transferribonuckic acid synthetase. Fortschritte der Antimikrobiellen und Antineoplastischen Chemotherapie.6, 169-73. Naber, K. G., Grimm, H., Rosenthal, E. J. K., Shah, P. M. & Wiedemann, B. (1990). Resistance to aminoglycosides: the situation in the Federal Republic of Germany. Journal of International Medical Research 18, Suppl. 4, 6D-26D. Noble, W. C , Rahman, M., Cookson, B. & Phillips, I. (1988). Transferable mupirocinresistance. Journal of Antimicrobial Chemotherapy 22, 771-2. Rahman, M., Connolly, S., Noble, W. C , Cookson, B. & Phillips, I. (1990). Diversity of staphylococci exhibiting high-level resistance to mupirocin. Journal of Medical Microbiology 33, 97-100. Rahman, M., Noble, W. C. & Cookson, B. (1987). Mupirocin-resistant Staphylococcus aureus. Lancet ii. 387. Rahman, M., Noble, W. C. & Cookson, B. (1989). Transmissible mupirocin resistance in Staphylococcus aureus. Epidemiology and Infection 102, 261-70. Rechsteiner, T. &. Leisinger, T. (1989). Purification of isoleucyl-tRNA synthetase from Methanobacterium thermoautotrophicum by pseudomonic acid affinity chromatography. European Journal of Biochemistry 181, 41-6. Slocombe, B. & Perry, C. (1991). The antimicrobial activity of mupirocin: an update on resistance. Journal of Hospital Infection 19, Suppl. B. 19-25. Smith, G. E. & Kennedy, C. T. C. (1988). Staphylococcus aureus resistant to mupirocin. Journal of Antimicrobial Chemotherapy 21, 141-2. Sutherland, R., Boon, R. J., Griffin, K. E., Masters, P. J., Slocombe, B. &. White, A. R. (1985). Antibacterial activity of mupirocin (pseudomonic acid), a new antibiotic for topical use. Antimicrobial Agents and Chemotherapy 27, 495-8. Washington, J. A. (1985). Susceptibility tests: agar dilution. In Manual of Clinical Microbiology, 4th edn (Lennette, E. M., Balows, A. Hauskr, W. J. & Shadomy, H. J., Eds), pp. 967-71. American Society for Microbiology, Washington, DC.
Biochemical basis of mnpirodn resistance in strains of Staphylococcus awreus T. H. Fanner, J. GObart* and S. W. Ebon
Twenty one strains of Staphylococcus aureus, of varying resistance to mupirocin, were examined in order to determine the mechanism of resistance to this antibiotic; six of these strains were mupirocin sensitive (MIC 0-12-1-0 mg/L) nine moderately resistant strains (MIC 8-256 mg/L) and six highly resistant strains (MIC > 2048 mg/L). Mupirocin showed a time-dependent inhibition of the target enzyme, isoleucyl-tRNA synthetase (IRS); incubation of the antibiotic with this enzyme before adding the substrates markedly increased inhibition in sensitive strains. The IRS I n values (the antibiotic concentrations which cause a 50% decrease in enzyme activity) correlated well with the MIC values for each strain {P < 001). The mean lx value for sensitive strains was 3-3 x 10~2 mg/L, in moderately resistant strains it was 1-3 x 10~' mg/L and in highly resistant strains it was 7-5 mg/L. No degradation of mupirocin could be detected during extended incubation of the antibiotic with cell free extracts from four resistant S. aureus strains. We conclude that the production of a modified IRS enzyme is the major cause of mupirocin resistance in the strains studied.
Introduction Mupirocin (pseudomonic acid A), an antibiotic produced by Pseudomonas fluorescens, has activity against many Gram-positive and some Gram-negative bacteria (Sutherland et al., 1985). Early work with Escherichia coli (Hughes & Mellows, 1980) demonstrated that the target of mupirocin is bacterial isoleucyl-tRNA synthetase (IRS), the enyzme which charges the appropriate tRNA with isoleucine, an essential component of protein synthesis. This process is accomplished in two distinct steps as illustrated below (E = IRS enzyme): 1. Isoleucine + ATP+E ?± E-Isoleucine-AMP + PPi 2. E-Isoleucine-AMP + tRNA -»Isoleucine-tRNA -I- AMP + E Mupirocin has been shown to inhibit only the first of these reactions and it was postulated that part of the antibiotic molecule resembles isoleucine; the antibiotic competing for the active site of the enzyme (Hughes & Mellows, 1980; Mellows, 1987). This possible mechanism of action however, remains to be more fully investigated as other parts of the mupirocin molecule have been implicated with its interaction with the enzyme (Rechsteiner & Leisinger, 1989). 'Corresponding author. 587 0305-7453/92/110587 +10 S08.00/0
© 1992 The Briuih Society for AntimicrobUl Chemotherapy
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
SmithKline Beecham Pharmaceuticals, Brockham Park, Betchworth, Surrey, RH3 7AJ, UK
588
T. H. Fanner et aL
Methods Bacterial strains A total of 21 5. aureus strains were used in this study comprising 18 isolates from patients, supplied by clinical laboratories and three reference isolates from the National Collection of Type Cultures (NCTQ. The sources of these strains, their mupirocin MIC values and resistance phenotypes are given in Table I. Mupirocin sensitive strains are defined as having MIC values in the range, 0-12-4-0 mg/L. Moderately resistant, strains have MIC values in the range, 8-256 mg/L. Highly resistant strains have MIC values of 512 mg/L or greater. The MIC values were determined by an agar dilution method (Washington, 1985). Preparation of bacterial enzyme extracts Cell free extracts of S. aureus were used for both degradation studies and isoleucyl tRNA synthetase (IRS) assays and were prepared using a procedure similar to that described by Hughes, Mellows & Sough ton (1980) in their studies on E. coli and P. fluorescens. Starter cultures of each S. aureus strain were prepared by inoculating 50 mL volumes of nutrient broth No. 2 (Oxoid and incubated at 37°C with shaking (240 rpm) for 18 h. A 200 mL volume of nutrient broth was inoculated with 2 mL of culture from the starter culture and grown to late log phase (6 h) under the same conditions. The cells were harvested by centrifugation at 5000 g at 4°C and the pellet resuspended in 5 mL cold 0-1 M Tris/HCl buffer pH 7-5 containing 10 mM magnesium
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
Used as a topical agent, mupirocin has proved to be an effective treatment for staphylococcal infections (Casewell & Hill, 1987) and is active against methicillin resistant strains (Sutherland et al., 1985). More recently however, some clinical mupirocin resistant Staphylococcus aureus strains have been reported from hospitals in the United Kingdom (Rahman, Noble & Cookson, 1987) and the USA (M. Carter, personal communication). Mupirocin resistant strains are rarely isolated!; in a recent survey in the UK, only 0-3% were resistant to 2 mg/L mupirocin (Slocombe & Perry, 1991). Clinical strains of S. aureus have been isolated which were moderately resistant to mupirocin, (MIC 2-32 mg/L) compared with < 0-2 mg/L for sensitive strains (Baird & Coia, 1987; Kavi, Andrews & Wise, 1987). In addition, highly resistant strains (MIC > 700 mg/L) have been isolated from patients in a London hospital (Rahman, et al., 1987) and from a Bristol hospital (MIC > 1025 mg/L) (Smith & Kennedy, 1988). In terms of mupirocin resistance therefore, S. aureus strains appear to fit into three groups; the majority are sensitive (99-7%), a few are moderately resistant (0-25%) and still fewer are highly resistant (0-05%) (Slocombe & Perry 1991). In some strains of the latter group the resistance has been shown to be transferable (Rahman 1989). The number of resistant strains isolated remains low but a more detailed knowledge of this resistance would be desirable. The purpose of this study was therefore to examine a representative set of S. aureus strains which included isolates, sensitive, moderately and highly resistant to mupirocin, in an attempt to elucidate the resistance mechanism. Two possible mechanisms were considered, firstly, enzyme degradation (or modification) of mupirocin and secondly, production of a modified bacterial isoleucyltRNA synthetase.
ftfupirocin resistance
589
Spectrophotometric and HPLC assays of mupirocin stability Enzymic degradation or modification was sought using both direct spectrophotometric and HPLC assays. For the spectrophotometric assay mupirocin concentrations of 5 or 10 mg/L were used and changes in absorbence at 223 nm were monitored. The antibiotic was dissolved in 0-05 M sodium phosphate buffer, pH 7-3, warmed to 37°C, then staphylococcal extracts (equivalent to 1 mg/mL dry weight of bacterial cells) from Table L Sources and antibiotic resistance phenotypes of S. aureus strains used in this study
Strain NCTC 6571 NCTC 11561 MCI NCTC 10442 MC2 OC1 OC2 JN1 DB1 DAI MC3 MC4 CR1 PR1 SP1 DB2 MC5 WN1 WN2 WN3 WN4
Source reference strain reference strain skin isolate, New York, USA reference strain skin isolate, New York, USA skin isolate, Romford, UK skin isolate, Romford, UK skin isolate, Glasgow, UK skin isolate, Glasgow, UK skin isolate, Romford, UK skin isolate, New York, USA skin isolate, New York, USA skin isolate, Crawley, UK (kin isolate, Warrington, UK skin isolate, Bristol, UK skin isolate, Glasgow, UK skin isolate, New York, USA skin isolate, London SE., UK plasmid cured variant of WN1 skin isolate, London SE., UK plasmid cured variant of WN3
Mupirocin MIC (mg/L) 0-12 0-25 0-5 1 8 16 16 16 32 32 64 128 256 >2048 >2048 >2048 >2048 >2048 0-25 >2048 0-25
Resistance phenotype PenG11 PenG11 PenG11, TetR, Mec* PenG11 PenG*, Mec* PenG11, Mec* TetR, Ery*, Fuc* PenG\ TetR, Ery*, Fuc* PenGR, TetR, Ery*, Fuc* PenG11, TetR PenGR, TetR, Ery* PenGR, TetR, Ery*, Mec* PenGR, TetR, Ery*, Mcc* PenG* PenGR, TetR, Ery* PenG\ TetR, Ery* PenGR, TetR, Ery" PenGR, TetR, Ery"
R, Rettstant; PenG, penicillin O; Tet, tetracydine; Ery, erythromycin; Mec, methicilliii; Fuc, fundk acid.
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
chloride and 2 mM dithiothrcitol (omission of dithiothreitol resulted in substantial loss of enzyme activity). The cells were disrupted by sonication with a Soniprep ISO sonicator (MSE Loughborough, UK) for 8 x 30 s periods with IS s intervals for cooling. The sample vessel was cooled in ice/water during disruption. In some later experiments the activity of the IRS extract was found to be improved by adding lysostaphin (Sigma Poolc, UK), to a final concentration of 10 mg/L to the resuspended cells and shaking at 240 rpm at 37°C for 30 min prior to sonication. Comparative experiments indicated that lysostaphin treatment did not alter the inhibition of IRS activity by mupirocin. The suspensions were then recentrifuged at 15,000 g at 4°C for 20 min and the pellets discarded. The supernatants were treated with 10 mg/L ribonuclease free deoxyribonuclease (Sigma, Poole, UK) and incubated at 4°C for 20 min. These samples were then centrifuged at 100,000 g at 4°C for 3 h and the supernatants dialysed against the above buffer for 16 h at 4°C. Glycerol was then added to each sample to give a final concentration of 25% and the samples were mixed gently before storage at - 7 0 ° C .
590
T. H. Fanner et aL
Radiometric assay of isoleucyl-tRNA synthetase (IRS) inhibition The activity of IRS and the degree of its inhibition by mupirocin was determined using a radiometric assay modified from a previously described method (Hughes & Mellows, 1980). This procedure is based on the incorporation of [14C]isoleucine into tRNA which, at the end of the reaction, is precipitated and the radio-activity counted by scintillation as described below. In the presence of mupirocin however, this incorporation of radio-active label is diminished which results in a lower count. A 40 /xL volume of each S. aureus IRS enzyme extract was added to 10 uL of inhibitor (or water for the control) and 50 uL of a reagent mixture containing 12 uM L-[U-l4C]isoleucine, specific activity: 12-0 GBq/mmol (Amersham), 10 g/L bulk tRNA (Sigma Poole, UK) and 10 mM ATP in 100 mM Tris buffer pH 8-3, supplemented with 10 mM MgCl2 and 2 mM dithiothreitol. This reaction mixture was then incubated at 37°C for exactly 10 min. The antibiotic was either mixed with the other reactants prior to adding the isoleucyltRNA synthetase (to measure the effect without prc-incubation), or incubated with the enzyme for 5 min then the other reagents added (to determine the effect of preincubation). To stop the reaction and precipitate the isoleucyl-tRNA complex, 2-5 mL 7% trichloroacetic (TCA) acid was added and the vials left on ice for 20 min. The suspensions were then filtered through 25 mm GF/C filter papers (Whatman Maidstone, UK) which were washed with 10 mL TCA and 10 mL ethanol, then dried and placed in scintillation vials with 5 mL OptiPhase 'Safe' scintillant fluid (Pharmacia Milton Keynes, UK) and counted in a Tri-carb 460 liquid scintillation counter. The IRS inhibition for each antibiotic concentration was calculated by expressing the reduction in activity as a percentage of that in the control sample (where water was substituted for mupirocin). The mupirocin concentration required to give a 50% reduction in activity (Iy, value) was then derived from a plot of mupirocin concentration versus percent inhibition, as shown in Figure 1. Initially, approximate mupirocin Ijo values were determined using a ten-fold dilution range (between final concentrations of 10 mg/L and 1 pg/L) of antibiotic in the above reactions. For more precise assays, doubling dilution series of mupirocin were used in which the antibiotic concentrations were in appropriate ranges above and below the approximate IM value for each strain.
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
strains SP1, WN1, WN3 and OC1 were each added to separate mupirocin solutions and monitored at intervals over a period of several hours. Controls containing the same concentrations of mupirocin in buffer only were incubated under the same conditions and aliquots were withdrawn for HPLC analysis at the same time intervals. Peak areas in the resulting chromatograms were compared with those of samples where bacterial extract was included. Assay of mupirocin by HPLC was carried out using a C,, /j-Bondapak reverse phase column (Waters Watford, UK), eluted with 60% methanol, 40% 0-05 M ammonium acetate, pH 4-5. At a flow rate of 1 mL/min, mupirocin eluted after 15 min. Initially the same incubation conditions were used as for the direct assay, but subsequently, the effect of adding ATP to a concentration of 2-5 mM was examined for some preparations. The purpose of ATP addition was to investigate the possibility that mupirocin could be inactivated by phosphorylation of its hydroxyl groups as has been observed with some aminoglycosides (Naber et al., 1990). In addition, mupirocin (10 mg/L) was added to cultures of S. aureus strains, NCTC 11561, SP1 WN1 and WN3 and the concentration in the culture broth was measured at intervals by HPLC during a 5 h period.
Mupirocin resistance
591
Figure 1. Mupirocin inhibition curves for IRS enzyme extracts from three S. aureus strains: A, NCTC 6571 (mupirocin-semitive); B, JN1 (moderately resistant to mupirocin); C, WN3 (highly resistant to mupirocin).
All the values that are shown in Tables II and m were derived from these more accurate assays. Results Spectrophotometric and HPLC assays of mupirocin stability Neither spectrophotometric nor HPLC assays indicated detectable losses or modification of mupirocin after treatment with S. aureus cell extracts, whether these were from moderately resistant or highly resistant strains. Adding ATP did not stimulate loss or modification of the antibiotic. Similarly, when bacteria were grown in the presence of mupirocin, analysis of culture supernatants showed little if any ( < 10%) loss of this compound. Radiometric assays of IRS inhibition In an initial series of assays the action of mupirocin on IRS preparations from a limited set of six 5. aureus strains was examined. In Table II the MIC values and corresponding IJO values are shown for this set of strains. A general increase in lx value with MIC can be seen. Figure 1 shows some typical mupirocin inhibition curves for a sensitive, moderately resistant and a highly resistant strain. The lx values for all strains were derived from similar plots. At low mupirocin concentrations the enzyme activity for highly resistant strains in some cases appeared to increase before it decreases at higher levels. This could be the result of stabilization or activation of the enzyme by ligand binding, a function which has been previously reported for this and other enzymes (Baldwin & Berg, 1966). In order to determine the nature of the inhibition, the IRS activity assays were performed with and without pre-incubation of the enzyme and inhibitor. In Table II it
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
1-0 10-0 I00O 1000-0 10000-0 Mupirocin concentration (^xg/L)
592
T. H. Fanner et *L Table II. Inhibition by mupirocin of isoleucyl-tRNA synthetase activity in extracts from various S. aureus strains. Extracts were assayed with and without pre-incubab'on with mupirocin (mg/L) Strain NCTC 11561 WN2 OC1 WN1 WN3 SP1
MIC (mg/L)
No pre-incubation
Pre-incubation
0-25 0-12 16 >204« >2048 >2048
3-6x10-' 9-7x10"' llxlO"' 6-7 71 71
2-5 xlO" 4 1-8 x l 0 " J 3-0x10"' 7-3 1-3 1-3
Table m. Inhibition by mupirocin of isoleucyltRNA tynthetase activity from 21 S. aureus strains without pre-incubation with mupirocin • Strain NCTC 6571 NCTC 11561 WN2 WN4 MCI NCTC 10442 MC2 OC1 OC2 JN1 DB1 DAI MC3 MC4 CR1 PR1 SP1 DB2 MC5 WN1 WN3
MIC (mg/L)
I , (mg/L)
0-12 0-25 0-25 0-25 0-5 1 8 16 16 16 32 32 64 128 256 >2O48 >2048 >2048 >2048 >2048 >2048
8-7 x 10"' 7-0x10"' 2-9 xlO" 3 21 x 10"2 5-3 xlO" 2 3-6 xlO" 2 1-2x10-'2 1-9 xlO" 6-5 xlO" 2 7-5 x 10"J 6-9 x 10"2 9-2 xlO" 1 2-2x10-' 1-5x10-' 5-6x10-' 6-0 6-7 90 10O 100 1CK)
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
is apparent that pre-incubation increases the inhibitory action of mupirocin against IRS from the sensitive strain (NCTC 11561). This indicates that binding to the mupirocin sensitive enzyme is time dependent and suggests that the antibiotic is not a classical competitive inhibitor of the staphylococcal enzyme. In a competitive system pre-incubation with the inhibitor would have had little or no effect since irreversible or strong binding would not take place. In a further set of assays a larger number of strains were included, where, to simplify the procedure, pre-incubation was not performed. Table III shows the results of mupirocin IM determinations on this set of 21 S. aureus strains; the correlation of IRS inhibition and MIC is still clearly apparent. The overall correlation coefficient (r) for the data listed in Table III is 0-90 which indicates good linearity and is statistically
Mnptrodn resistance
593
lOOOO
l 000
'
3 O-IOO
0-O10
0-1
IO
10-0
I0O0 MC (mg/L)
1000-0 I0000O
Hgmc 2. Plot of mupirocin MIC value* vcnut. IRS enzyme I n values for the 21 5. aureus itraint included in this study, the plot shows the apparent correlation between these two parameters. Since MIC values of greater than 2048 mg/L could not be accurately determined, the points representing very resistant strains could not be placed on the graph with any certainty and are thus separated in the boxed area. These points were not included in the computed hne of best fit
significant (P < 0-01). A graphical illustration of this correlation is given in Figure 2. It can be seen that IRS extracts from the highly resistant strains were much less susceptible to inhibition by mupirocin than those from sensitive strains. Furthermore, the IJO values for mupirocin against the IRS extracts from the moderately resistant strains occupy an intermediate position between the values for the enzymes from the sensitive and highly resistant strains. For most strains we observed no correlation between level of IRS activity (i.e. the amount of isoleucyl-tRNA formed after 10 min) and resistance. The activity of each IRS enzyme preparation was variable despite the use of uniform quantities of bacterial cells for enzyme preparation and ensuring constant assay conditions. The protein • concentrations in the bacterial lysates after sonication were found to be fairly constant suggesting that this variability of IRS activity was not the result of inconsistent bacterial cell breakage. In two formerly resistant strains (WN2 and WN4), which had been made sensitive by culture at 42°C resulting in curing of plasmids, the IRS activities were reduced 20% and 58% in relation to their parental strains (WN1 and WN3, respectively). Strains WN2 and WN4 were isogenic with their uncured counterparts, WN1 and WN3 and thus were more suitable for comparison than the other unrelated isolates. Resistance was transferable in these strains. Another highly mupirocin resistant strain (SP1), was reproducibly shown to have 50% or lower IRS activity than the other highly resistant strains. No transferable plasmids encoding mupirocin resistance could be detected in SP1 and the resistance gene was believed to have been incorporated into the chromesome (C. Perry, personal communication). Discussion
Based on the HPLC and spectrophotometric assays, it appears unlikely that the strains included in this study owe their mupirocin resistance to overt modification or degradation of the antibiotic. These findings are in agreement with a recent report in which,
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
0-001
594
T. H. Fanner et aL
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
using growth inhibition studies and HPLC assays, no reduction in antibacterial activity or mupirocin concentration could be detected after extensive incubation of the antibiotic with the IRS enzyme extract (Cookson, 1989). The good correlation between mupirocin MIC and IRS IM reported above, suggests that the major difference between sensitive and resistant strains lies within the IRS enzyme itself. The effect of preincubating the IRS enzyme with mupirocin, as shown in Table I, indicates that this antibiotic is not a classical competitive inhibitor, the activity of the IRS was decreased if the antibiotic was mixed with it and incubated for a short period before adding the substrates. This effect was more marked with enzymes from more sensitive strains. A competitive inhibitor should not show this effect. In a previous report however (Hughes & Mellows, 1980), mupirocin was described as competitively inhibiting the E. coli IRS enzyme although a stable enzyme-inhibitor complex was observed. These findings indicate that the inhibitor is tightly bound but the exact nature of the binding is not fully understood. It is not currently known whether mupirocin resistant S. aureus strains produce a totally modified IRS enzyme concurrently with the sensitive enzyme or whether the resistant enzyme only is present. The slightly higher IRS activities of the two mupirocin resistant strains, WN1 and WN3 compared to the heat treated, sensitive revertants, WN2 and WN4, suggest that there could be two enzymes present whose activities may be additive. The low level of activity seen in the non-plasmid containing resistant strain, SP1 is further evidence for this additive effect Our data indicate however, that hyperproduction of the IRS enzyme is an unlikely resistance mechanism; much greater production of sensitive IRS enzyme than was observed would be required to overcome the effect of mupirocin. The genetic location of mupirocin resistance has been previously investigated. In a study using the archaebacterial species Methanobacterium thermoautotrophicum (Kiener, Rechsteiner & Leisinger, 1986), the inhibition of growth and IRS activity by mupirocin were similar to those for S. aureus reported above. The inhibition in M. thermoautotrophicum could be overcome by the addition of exogenous isoleucine. These authors used nitrosoguanidine to induce a mutation conferring mupirocin resistance and found that the IRS enzyme in the mutant had become less sensitive to the antibiotic. They suggested that this resistance was a result of a change in the IRS structural gene on the chromosome. In highly resistant S. aureus however, available evidence indicates that the resistance resides on a plasmid (Noble et al., 1988). High level mupirocin resistance in 5. aureus has been shown to be transferable at low frequency in conjugation experiments where highly resistant and sensitive clinical strains were mated on niters (Rahman, et al., 1989). Initial findings suggested that the resistance was located on a very large plasmid which was difficult to resolve from chromosomal DNA. Subsequently, mupirocin resistance in S. aureus has been located on plasmids ranging in size from 23 to 87 kb (Rahman et al., 1990). In more recent work the mupirocin resistance gene from an S. aureus strain was cloned from a small (25 kb) plasmid (Dyke et al., 1991). Mupirocin resistance can be eliminated from S. aureus by growth at 42°C, as was demonstrated in strains WN1 and WN3 in this study. This observation is consistent with the resistance being plasmid encoded since higher growth temperatures can cause curing. The plasmids involved in mupirocin resistance are of a novel type, their structure and epidemiology require further investigation (Rahman et al., 1989). Isolates of S. aureus can also be rendered resistant by culture in media containing successively higher mupirocin concentrations (Casewell
Mupfructn restasnce
995
Acknowledgements We wish to thank Dr R. Sutherland and Mr B. Slocombe for their assistance and helpful discussions. We also thank Mr. I. Macpherson for statistical assistance and the staff of the Anti-infective Product Support Unit, Brockham Park, for supplying the strains used in this study, and for determining the MICs. We also thank Dr. M. Carter of the Rockefeller Institute, New York, NY, USA, for supplying further clinical strains of S. aureus. References Baird, D. & Coia, J. (1987). Mupu-orin-resistant Staphyloccus aureus. Lancet li, 387-8. Baldwin, A. N. & Berg, P. (1966). Purification and properties of isoleucyl ribonucleic acid synthetase from Escherichia coli. Journal of Biological Chemistry 241, 831-8. Capobianco, J. O., Doran, C. C. & Goldman, R. C. (1989). Mechanism of mupirocin transport into sensitive and resistant bacteria. Antimicrobial Agents and Chemotherapy 33, 156-63. Casewell, M. W. &. Hill, R. L. R. (1985). In-vitro activity of mupirocin ('pseudomonic acid*) against clinical isolates of Staphylococcus aureus. Journal of Antimicrobial Chemotherapy 15, 523-31. CaseweH, M. W. & HUT, R. L. R. (1987). Mupirocin ("pseudomonic acid")—fl promising new topical antimicrobial agent Journal of Antimicrobial Chemotherapy 19, 1-5. Cookson, B. (1989). Failure of mupirocin-resistant staphylococci to inactivate mupirocin. European Journal of Clinical Microbiology and Infectious Diseases 8, 1038-40.
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
& Hill, 1985) The mechanism of this type of resistance is unknown. Although this phenomenon produces stable resistance, it is acquired slowly and the antibiotic levels achievable by topical therapy are substantially greater than the MIC seen. Resistance to mupirocin could alternatively result from alterations in membrane permeability. In a recent study by Capobianco, Doran & Goldman (1989) the mode of entry of mupirocin into bacterial cells including resistant and sensitive S. aureus was investigated. These authors concluded that mupirocin uptake into the cytoplasm was energy independent and recorded no differences in membrane penneability to the antibiotic between a resistant and a sensitive strain. From thesefindingsthey made the unusual conclusion that binding of the antibiotic to the IRS enzyme was the force responsible for development of high cytoplasmic concentrations of mupirocin in sensitive cells. In a resistant strain obtained by culture in media containing increasing concentrations of mupirocin, they speculated that the lower affinity of the IRS enzyme for the antibiotic reduced this attractive force to some extent and thus the intracellular concentration remained low. They also postulated that some other unknown mechanism may be operating. Although these findings suggest the IRS enzyme has a role in mupirocin resistance, further data on a greater number of S. aureus strains is required to support this novel hypothesis and to rule out the possible involvement of membrane permeability in the process. Mupirocin resistant clinical isolates of S. aureus remain a comparative rarity (Slocombe & Perry, 1991). Resistant isolates can however spread locally within hospital wards causing problems with treatment Understanding the mechanism resistance could aid epidemiological studies and establish a basis for the development of novel antibiotics of a similar type. The structure and function of the IRS enzyme in S. aureus requires further study as this report shows that alteration of the IRS enzyme is the cause of mupirocin resistance in S. aureus.
596
T. H. Fanner et *L
(Received 21 October 1991; revised version accepted 6 July 1992)
Downloaded from http://jac.oxfordjournals.org/ at The Univesity of Calgary on May 25, 2015
Dyke, K. G. H., Curnock, S. P., Golding, M. & Noble, W. C. (1991). Cloning of the gene conferring resistance to mupirocin in Staphylococcus aureus. FEMS Microbiology Letters 77, 195-8. Hughes, J. & Mellows, G. (1980). Interaction of pscudomonic acid A with Escheriehia coli B isoleucyl-tRNA synthetase. Biochemical Journal 191, 209-19. Hughes, J., Mellows, G. & Soughton, S. (1980). How does Pseudomonasfluorescent,the producing organism of the antibiotic pseudomonic acid A, avoid suicide? FEES Letters 122, 322-4. Kavi, J., Andrews, J. M. & Wise, R. (1987). Mupirocin-resistant Staphylococcus aureus. Lancet ii, 1472. Kiener, A., Rechsteiner, T. & Leisinger, T. (1986). Mutation to pseudomonic acid resistance of Methanobacterium thermoautotrophicum leads to an altered isoleucyl-tRNA synthetase. FEMS Microbiology Utters 33, 15-8. Mellows, G. (1987). The mode of action of mupirocin: inhibition of bacterial isoleucyl transferribonuckic acid synthetase. Fortschritte der Antimikrobiellen und Antineoplastischen Chemotherapie.6, 169-73. Naber, K. G., Grimm, H., Rosenthal, E. J. K., Shah, P. M. & Wiedemann, B. (1990). Resistance to aminoglycosides: the situation in the Federal Republic of Germany. Journal of International Medical Research 18, Suppl. 4, 6D-26D. Noble, W. C , Rahman, M., Cookson, B. & Phillips, I. (1988). Transferable mupirocinresistance. Journal of Antimicrobial Chemotherapy 22, 771-2. Rahman, M., Connolly, S., Noble, W. C , Cookson, B. & Phillips, I. (1990). Diversity of staphylococci exhibiting high-level resistance to mupirocin. Journal of Medical Microbiology 33, 97-100. Rahman, M., Noble, W. C. & Cookson, B. (1987). Mupirocin-resistant Staphylococcus aureus. Lancet ii. 387. Rahman, M., Noble, W. C. & Cookson, B. (1989). Transmissible mupirocin resistance in Staphylococcus aureus. Epidemiology and Infection 102, 261-70. Rechsteiner, T. &. Leisinger, T. (1989). Purification of isoleucyl-tRNA synthetase from Methanobacterium thermoautotrophicum by pseudomonic acid affinity chromatography. European Journal of Biochemistry 181, 41-6. Slocombe, B. & Perry, C. (1991). The antimicrobial activity of mupirocin: an update on resistance. Journal of Hospital Infection 19, Suppl. B. 19-25. Smith, G. E. & Kennedy, C. T. C. (1988). Staphylococcus aureus resistant to mupirocin. Journal of Antimicrobial Chemotherapy 21, 141-2. Sutherland, R., Boon, R. J., Griffin, K. E., Masters, P. J., Slocombe, B. &. White, A. R. (1985). Antibacterial activity of mupirocin (pseudomonic acid), a new antibiotic for topical use. Antimicrobial Agents and Chemotherapy 27, 495-8. Washington, J. A. (1985). Susceptibility tests: agar dilution. In Manual of Clinical Microbiology, 4th edn (Lennette, E. M., Balows, A. Hauskr, W. J. & Shadomy, H. J., Eds), pp. 967-71. American Society for Microbiology, Washington, DC.
