Molecular analysis of fusidic acid resistance in Staphylococcus aureus

Molecular Microbiology (2003)

47

(2), 463–469

© 2003 Blackwell Publishing Ltd

Blackwell Science, LtdOxford, UKMMIMolecular Microbiology0950-382XBlackwell Publishing, 200347Original Article

Molecular analysis of fusidic acid resistance in S. aureusS. Besier, A. Ludwig, V. Brade and T. A. Wichelhaus

Accepted 9 October, 2002. *For correspondence. [email protected]; Tel. (

+

49) 69 6301 6438; Fax (

+

49)69 6301 5767.

†

These authors contributed equally to this work.

Molecular analysis of fusidic acid resistance in

Staphylococcus aureus

Silke Besier,

†

Albrecht Ludwig,

†

Volker Brade and Thomas A. Wichelhaus*

Institut für Medizinische Mikrobiologie, Klinikum der J. W. Goethe-Universität, Paul-Ehrlich-Str. 40, 60596 Frankfurt am Main, Germany.

Summary

Fusidic acid is a potent antibiotic against severeGram-positive infections that interferes with the func-tion of elongation factor G (EF-G), thereby leading tothe inhibition of bacterial protein synthesis. In thisstudy, we demonstrate that fusidic acid resistance in


results from point mutationswithin the chromosomal

fusA

gene encoding EF-G.Sequence analysis of

fusA

revealed mutationalchanges that cause amino acid substitutions in 10fusidic acid-resistant clinical

S. aureus

strains as wellas in 10 fusidic acid-resistant

S. aureus

mutants iso-lated under fusidic acid selective pressure

in vitro

.Fourteen different amino acid exchanges were identi-fied that were restricted to 13 amino acid residueswithin EF-G. To confirm the importance of observedamino acid exchanges in EF-G for the generation offusidic acid resistance in

S. aureus

, three mutant

fusA

alleles encoding EF-G derivatives with the exchangesP406L, H457Y and L461K were constructed by site-directed mutagenesis. In each case, introduction ofthe mutant

fusA

alleles on plasmids into the fusidicacid-susceptible

S. aureus

strain RN4220 caused afusidic acid-resistant phenotype. The elevated mini-mal inhibitory concentrations of fusidic acid deter-mined for the recombinant bacteria were analogousto those observed for the fusidic acid-resistant clini-cal

S. aureus

isolates and the

in vitro

mutants con-taining the same chromosomal mutations. Thus, thedata presented provide evidence for the crucial impor-tance of individual amino acid exchanges within EF-G for the generation of fusidic acid resistance in

S.aureus

.

Introduction

Fusidic acid is a steroid antibiotic derived from the fungus

Fusidium coccineum

(Godtfredsen

et al

., 1962). Thesodium salt of fusidic acid was introduced into clinical usein 1962, and its activity against staphylococci has contin-ued to draw interest as resistance towards antistaphylo-coccal substances such as

b

-lactams, quinolones andmacrolides continues to advance in this genus (Livermore,2000).

Fusidic acid inhibits protein synthesis by acting directlyon the elongation factor EF-G (Laurberg

et al

., 2000). EF-G in turn hydrolyses GTP to GDP in order to provideenergy for the translocation of the peptidyl-tRNA from theA site to the P site within the ribosome. In the presenceof fusidic acid, EF-G remains bound to the ribosome afterGTP hydrolysis, sterically blocking the next stage in pro-tein synthesis (Bodley

et al

., 1969; Martemyanov

et al

.,2001).

With respect to


, resistance tofusidic acid has been demonstrated to occur rapidly

invitro

(Price and Gustafson, 2001). Nonetheless, resis-tance in clinical isolates of

S. aureus

, although observedwith increasing frequency, still remains at a low level(Brown and Thomas, 2002). This feature may be attribut-able to several facts: (i) administration of fusidic acid isusually limited to very few patients suffering from severe

S. aureus

infections, thereby exerting only a relatively lowselection pressure; (ii) fusidic acid is generally adminis-tered in combination with other antibiotics, which almostcertainly reduces the mutation frequency towards resis-tance; (iii) fusidic acid resistance is often associated withfitness costs during growth

in vivo

, thus enabling the sus-ceptible counterpart to keep pace with the competition(Shanson, 1990; Nagaev

et al

., 2001; Brown and Thomas,2002).

Two different mechanisms of resistance to fusidic acidhave been suggested previously: first, alteration of thetarget protein EF-G and, secondly, decreased perme-ability of the bacterial envelope for the antibiotic (Chopra,1976; Turnidge and Collignon, 1999). With regard to

S.aureus

, however, neither of these two mechanisms hasyet been proven on the molecular level. In

Salmonellaenterica

serovar Typhimurium at least, it has been con-cluded from a phage transduction assay that mutations inthree specific regions of the EF-G structural gene,

fusA

,

464

S. Besier, A. Ludwig, V. Brade and T. A. Wichelhaus

© 2003 Blackwell Publishing Ltd,

Molecular Microbiology

,

47

, 463–469

result in an altered target and cause the minimal inhibitoryconcentration (MIC) of fusidic acid to increase (Johansonand Hughes, 1994). Recently, several mutations in

fusA

that cause amino acid substitutions in EF-G have beenidentified in a number of

in vitro

-generated and clinicallyisolated fusidic acid-resistant strains of

S. aureus

(Nagaev

et al

., 2001), but the causal relationships between thesemutations and the fusidic acid-resistant phenotypes of thecorresponding strains have not yet been demonstrated.

This study is the first to provide evidence at the molec-ular level that distinct point mutations within the

fusA

geneare responsible for fusidic acid resistance in

S. aureus

.

Results

Identification of mutations within the

fusA

gene in fusidic acid-resistant

S. aureus

isolates

The complete

fusA

gene was sequenced from 10 fusidicacid-resistant clinical

S. aureus

isolates as well as from10 fusidic acid-resistant derivatives of the

S. aureus

strainRN4220, which were selected

in vitro

. Among the

in vitro

mutants, eight different point mutations causing aminoacid exchanges in EF-G were identified as shown inTable 1. No silent mutations were observed. Single aminoacid substitutions were detected in all but one mutant that,in turn, exhibited a double mutational change within

fusA

.

With respect to the clinical isolates, all showed aminoacid substitutions within EF-G, and multiple mutationalchanges were found in seven of the 10 strains. Further-more, all clinical isolates exhibited a diverse number ofsilent mutations within

fusA

, thereby demonstrating theheterogeneity of these strains. In addition, the clinicalisolates were proven by pulsed-field gel electrophoresis(PFGE) to be of polyclonal origin (Fig. 1). Different levelsof resistance to fusidic acid were observed for thetested

S. aureus

strains, i.e. low-level resistance(MIC

=

8

m

g ml

-

1

), intermediate-level resistance (MIC

=

64

m

g ml

-

1

) and high-level resistance (MIC

>

256

m

g ml

-

1

)(Table 1). Interestingly, high-level resistance was onlyfound in clinical isolates, which might reflect adaptationto the strong antibiotic selective pressure in the clinicalsetting.

Cloning and site-directed mutagenesis of the

fusA

gene from

S. aureus

The

S. aureus fusA

gene was amplified by polymerasechain reaction (PCR) from the chromosome of strainRN4220 and cloned into the

Sph

I–

Sal

I site of pUC19,resulting in plasmid pSB1. To study the importance ofsingle amino acid exchanges in EF-G for the generation

Table 1.

Mutational changes in EF-G and minimal inhibitory concen-tration (MIC) of fusidic acid in

S. aureus

isolates.

S. aureus

strain Mutation in EF-GFusidic acidMIC (

m

g ml

-

1

)

a

RN4220 – 0.032

In vitro

mutants

b

T 964 P406L 8T 965 P406L 8T 966 P406L 8T 967 P406L 8T 980 T436I 8T 963 G452S 8T 978 L456F 8T 962 H457Y 64T 982 R464C 8T 975 F652S, Y654N 8

Clinical isolatesT 52 A67T, P406L 8T 54 H457Y 64T 121 H457Y, S416F 64T 16 L461K

>

256T 55 L461K

>

256T 255 V90I, H457Q, L461K, A655V

>

256T 57 V90I, H457Q, L461K, A655V

>

256T 120 V90I, H457Q, L461K, A655V

>

256T 138 V90I, H457Q, L461K, A655V

>

256T 268 V90I, H457Q, L461K, A655V

>

256

a.

MIC values were measured by the agar dilution method accordingto NCCLS criteria.

b.

Mutants were selected on MH agar that contained 1

m

g ml

-

1

sodiumfusidate.

Fig. 1.

Pulsed-field gel electrophoresis patterns of

Sma

I digests of total DNA from fusidic acid-resistant clinical

S. aureus

isolates (indi-cated by T-numbers above lanes). Lanes m, size markers. Molecular sizes of the marker fragments (in kb) are indicated on the right.

Molecular analysis of fusidic acid resistance in

S. aureus 465

© 2003 Blackwell Publishing Ltd,

Molecular Microbiology

,

47

, 463–469

of fusidic acid resistance in

S. aureus

, the cloned

fusA

gene was subjected to site-directed mutagenesis. In par-ticular, three

fusA

mutants were constructed that encodedEF-G derivatives containing the amino acid exchangesP406L, H457Y and L461K (Fig. 2). As shown in Table 1,these amino acid substitutions were also detected in theamino acid sequences deduced from the

fusA

alleles offusidic acid-resistant

S. aureus

isolates. The wild-type

fusA

gene and the generated

fusA

mutants were sub-cloned into the shuttle expression vector pHPS9, result-ing in the plasmids pSB2 (containing the wild-type

fusA

gene), pSB2P406L, pSB2H457Y and pSB2L461K.Finally, these plasmids were introduced by electroporationinto the

S. aureus

strain RN4220 to determine the effectof altered EF-G on fusidic acid susceptibility in

S. aureus

.

Fusidic acid susceptibility of recombinant

S. aureus

RN4220 derivatives harbouring wild-type fusA or mutant

fusA

alleles


RN4220 containing the wild-type

fusA

gene on plasmid pHPS9 (pSB2) exhibited a MIC forfusidic acid of 0.032

m

g ml

-

1

, which corresponds to the

MIC obtained for the parental strain RN4220 (Fig. 3A). Incontrast, in the case of

S. aureus

RN4220 carrying plas-mid pSB2P406L, a fusidic acid MIC of 4

m

g ml

-

1

wasobserved (Fig. 3B). The

S. aureus

RN4220 clones carry-ing the plasmids pSB2H457Y and pSB2L461K exhibitedeven higher fusidic acid MICs of 64 and

>

256

m

g ml

-

1

respectively (Fig. 3C and D). Thus, the recombinant

S.aureus

strains that carried the mutant fusA alleles not onlyexhibited fusidic acid resistance but also showed similarMIC values to the clinical isolates or in vitro mutantscontaining the same mutational changes in the chromo-somal fusA gene.

Analysis of fusA expression by RT-PCR

To confirm that the wild-type and mutant fusA allelespresent in pSB2, pSB2P406L, pSB2H457Y andpSB2L461K are indeed expressed in the recombinant S.aureus RN4220 derivatives harbouring these plasmids,the fusA-specific mRNA levels in log-phase cultures of twoof these strains, i.e. S. aureus RN4220/pSB2 and S.aureus RN4220/pSB2L461K, were analysed by reversetranscription (RT)-PCR. When the oligonucleotides P9and P10 were used as RT-PCR primers, the expectedfusA-specific 410 bp fragment was amplified from bothstrains (Fig. 4). A unique EcoRV site in fusAL461K wasused to discriminate between RT-PCR products that wereamplified from transcripts of the chromosomal wild-typefusA gene of S. aureus RN4220 and those that wereamplified from transcripts of the plasmid-encoded fusAgene. This EcoRV site, which is located 248 bp down-stream from the 5¢ end of the 410 bp RT-PCR product,was introduced together with the mutation causing theamino acid substitution L461K to permit the identificationof mutant clones by restriction analysis (Fig. 2). Indeed,most of the 410 bp fragment amplified from total RNA ofS. aureus RN4220/pSB2L461K was cleaved by EcoRVinto two fragments corresponding in size to the anticipated248 bp and 162 bp fragments, showing that the plasmid-encoded mutant fusA gene was transcribed efficiently(Fig. 4). In contrast, the 410 bp RT-PCR product amplifiedfrom RNA of S. aureus RN4220/pSB2 was not cleaved byEcoRV, which is consistent with the fact that pSB2 con-tains the wild-type fusA gene.

Discussion

The data presented in this study clearly demonstrate thatdistinct mutations within the fusA gene causing aminoacid substitutions in EF-G result in fusidic acid resistancein S. aureus.

Three mutant fusA alleles, each containing a putativeresistance-mediating mutation, were generated by site-directed mutagenesis and introduced on a plasmid into

Fig. 2. Construction of EF-G derivatives containing the amino acid substitutions P406L, H457Y and L461K. The nucleotide sequences of three fusA regions from S. aureus and the deduced amino acid sequences of EF-G are shown. Nucleotide and amino acid substitu-tions generated by site-directed mutagenesis are depicted below the corresponding wild-type sequences. The codon numbering is that of the S. aureus sequence. Nucleotide and amino acid substitutions are indicated in bold. Silent mutations causing either destruction or gen-eration of a restriction site resulted in modified restriction patterns, thereby facilitating the identification of clones harbouring the muta-tional changes.

466 S. Besier, A. Ludwig, V. Brade and T. A. Wichelhaus

© 2003 Blackwell Publishing Ltd, Molecular Microbiology, 47, 463–469

the fusidic acid-susceptible S. aureus strain RN4220.Expression of these mutant fusA alleles in the recombi-nant S. aureus strains gave rise to different levels offusidic acid resistance, thereby proving the causal rela-tionship between individual amino acid exchanges inEF-G and fusidic acid resistance in S. aureus. In previousreports, mutations in fusA of S. aureus have beendescribed and proposed to be associated with fusidic acidresistance (Laurberg et al., 2000; Nagaev et al., 2001).

These mutations, however, could only be regarded asputative resistance mediating insofar as a correlationbetween mutations in EF-G and fusidic acid resistancewas deduced solely from phage transduction experimentsperformed in S. enterica serovar Typhimurium (Johansonand Hughes, 1994).

EF-G has been described as consisting of at least sixdomains, so that domain III is believed to function as thepossible fusidic acid binding site (Laurberg et al., 2000).

Fig. 3. Determination of the fusidic acid susceptibility of S. aureus RN4220 harbouring the following plasmids: (A) pSB2; (B) pSB2P406L; (C) pSB2H457Y; and (D) pSB2L461K. The MIC for fusidic acid was measured using E-test strips (AB Biodisk).

A B

C D

Molecular analysis of fusidic acid resistance in S. aureus 467


Interestingly, it could be shown in the present study thatall but one of the tested fusidic acid-resistant S. aureusisolates displayed mutational changes within this centraldomain. Seven of the 14 different amino acid substitutionsfound here within EF-G have also been described byLaurberg et al. (2000) and Nagaev et al. (2001) and sug-gested to be associated with fusidic acid resistance:V90I, P406L, T436I, G452S, L456F, H457Y and R464C.However, the other seven mutational changes were notdetected previously. Among them, the amino acid substi-tution L461K, which demonstrably resulted in high-levelfusidic acid resistance, was shown to exist in seven out of10 clinical isolates. In view of the polyclonal origin of theclinical isolates analysed, the high prevalence of both themutation L461K as well as the mutational changes atamino acid position 457 (H457Y, H457Q) might be indic-ative of the crucial importance of these substitutions forthe generation of fusidic acid resistance in vivo. Otheramino acid exchanges at position 461 (L461S, L461F)were found recently in fusidic acid-resistant clinical S.aureus isolates (Nagaev et al., 2001). Remarkably, thefusidic acid MIC value determined here for the strainsexhibiting the L461K substitution in EF-G (>256 mg ml-1)is significantly higher than the MIC values obtained byNagaev et al. (2001) for two clinical strains encodingan EF-G derivative with the mutation L461S (6 and8 mg ml-1). This is in line with the assumption that different

amino acid exchanges at the same site may confer mark-edly different levels of resistance to fusidic acid.

Of further interest was the finding that, in contrast to invitro mutants, fusidic acid-resistant clinical isolates fre-quently harbour multiple mutational changes within EF-G.It remains a matter of speculation whether multiple aminoacid substitutions are associated with an evolutionaryadvantage. On the one hand, they might cause higherlevels of resistance to fusidic acid owing to a synergisticeffect of individual mutations whereas, on the other hand,certain secondary mutations might compensate for biolo-gical fitness costs associated with some resistance-mediating mutational changes. Indeed, several investiga-tors have shown that antibiotic resistance is frequentlyassociated with a biological fitness cost for the resistantbacteria (Björkman et al., 1998; Andersson and Levin,1999; Macvanin et al., 2000; Nagaev et al., 2001;Wichelhaus et al., 2002). In addition, it has been con-vincingly argued that resistant strains easily acquirefitness-compensatory mutations, usually without loss ofresistance (Björkman et al., 2000; Levin et al., 2000). Inthis sense, multiple mutations within EF-G of fusidic acid-resistant clinical S. aureus isolates could be interpretedas a response to the selection pressure in vivo that callsfor both fusidic acid resistance and EF-G functionality (i.e.fitness). The same concept might also be true for thedouble mutational change H457Y/S416F in EF-G. Amongthe clinically isolated fusidic acid-resistant S. aureusstrains, one exhibited the amino acid substitution H457Yin EF-G, whereas in another strain, H457Y was found incombination with a further amino acid exchange, S416F,that has not been described previously. It is unlikely thatthe S416F substitution contributes to fusidic acid resis-tance because the same fusidic acid MICs were deter-mined for the two strains that displayed the singlemutation H457Y and the double mutation H457Y/S416F.It remains to be seen in additional studies whether theS416F exchange represents a fitness-compensatorymutation.

The present paper suggests that alteration of EF-Gby amino acid substitutions, a phenomenon that hasbeen observed in all the clinical S. aureus isolates analy-sed, is a major mechanism for achieving fusidic acidresistance. However, mutations affecting the amino acidsequence of EF-G have not been detected in severalother clinical fusidic acid-resistant isolates (Nagaev et al.,2001), thus indicating that other resistance mechanismsmay also play a role. In particular, the clinical relevanceof the plasmid-mediated resistance mechanism, thoughtto be attributable to a fusidic acid permeability barrier(Lacey and Rosdahl, 1974), still remains to beelucidated.

In conclusion, this is the first report to provide directevidence at the molecular level that individual amino acid

Fig. 4. Analysis of the expression of the plasmid-encoded fusA alleles in S. aureus RN4220/pSB2 and S. aureus RN4220/pSB2L461K by RT-PCR. RT-PCR was performed with the fusA-specific forward primer P9 and the reverse primer P10, using 10 pg of total RNA as template. The PCR products obtained from S. aureus RN4220/pSB2 (lanes 1 and 2) and S. aureus RN4220/pSB2L461K (lanes 3 and 4) were either directly separated by agarose gel electrophoresis (lanes 1 and 3) or incubated with EcoRV before separation on the agarose gel (lanes 2 and 4).

468 S. Besier, A. Ludwig, V. Brade and T. A. Wichelhaus


exchanges in EF-G are causative for elevated fusidic acidMICs in S. aureus.

Experimental procedures

Bacterial strains, plasmids and culture conditions

The clinical fusidic acid-resistant S. aureus isolates analysedin this study were obtained from patients with wound infec-tions or bacteraemia hospitalized in the university clinic ofFrankfurt am Main, Germany.

Fusidic acid-resistant in vitro mutants were selected byplating S. aureus strain RN4220 (Kreiswirth et al., 1983) ontoMueller–Hinton (MH) agar containing 1 mg ml-1 sodium fusi-date. Escherichia coli strain DH5a [F¢ f80 dlacZDM15 D(lac-ZYA-argF) U169 deoR recA1 endA1 phoA hsdR17 (rK

– mK+)

supE44 l– thi-1 gyrA96 relA1] was used as host strain for theconstruction and propagation of plasmids as well as for site-directed mutagenesis. The plasmids pUC19 (Ampr) andpHPS9 (Cmr, Emr) (Haima et al., 1990) were used as cloningvectors. Bacteria were grown aerobically at 37∞C in 2¥ yeastextract–tryptone (2¥ YT) medium (Ludwig et al., 1999) or onYT medium solidified with 1.5% (w/v) agar. Antibiotics wereused at the following final concentrations: ampicillin (Amp),100 mg ml-1; chloramphenicol (Cm), 10 mg ml-1; erythromycin(Em), 10 mg ml-1.

Fusidic acid susceptibility testing

MIC values of fusidic acid-resistant clinical S. aureus isolatesas well as of fusidic acid-resistant mutants isolated underfusidic acid selective pressure in vitro were determined bythe agar dilution method following NCCLS criteria. MIC val-ues of recombinant S. aureus strains harbouring the wild-typefusA gene or a mutant fusA allele on a plasmid were deter-mined using E-test strips (AB Biodisk).

DNA standard techniques

Isolation of chromosomal and plasmid DNA, agarose gelelectrophoresis and all general DNA cloning procedures werecarried out as described by Sambrook and Russell (2001).Nucleotide sequences of DNA fragments were determinedby cycle sequencing using an ABI Prism DNA sequencer(Applied Biosystems). PCR amplification of DNA was con-ducted with Pfu DNA polymerase (Stratagene) according tothe recommendations of the manufacturer. Site-directedmutagenesis was performed using the QuikChangeTM site-directed mutagenesis kit (Stratagene).

Oligonucleotides

The following oligonucleotides were synthesized by LifeTechnologies:P1, 5¢-CATTAAGCGTGCATGCTTAGGGCATC-3¢ (SphI);P2, 5¢-GGCTAGCTAGTCGACCAAGTTATATTATTCACC-3¢(SalI); P3, 5¢-CAATGGAATTTCCAGAGCTAGTTATTCAC-3¢(D EcoRI, P406L); P4, sequence complementary to P3; P5,5¢-GGGTGAGCTTTACTTAGATATCTTAGTAG-3¢ (EcoRV,

H457Y); P6, sequence complementary to P5; P7, 5¢-GCTTCACTTAGATATCAAAGTAGACCGTATG-3¢ (EcoRV,L461K); P8, sequence complementary to P7; P9, 5¢-GCGGTAGGTCTTAAAGATACAGG-3¢; P10, 5¢-GAATTCAATGTGAACATCACCG-3¢.

Underlined sequences within the oligonucleotides indicateintroduced or destroyed (D) restriction sites for the enzymesgiven in parentheses. Nucleotide substitutions are shown inbold and altered fusA codons in mutagenic primers causingthe amino acid exchanges given in parentheses are shownin boxes.

Construction of plasmids and site-directed mutagenesis

The oligonucleotides P1 and P2 were used as forward andreverse primers for the PCR amplification of the fusA genefrom S. aureus RN4220. To construct plasmid pSB1, thefusA-carrying 2.185 kb PCR product was cleaved in its 5¢-and 3¢-terminal regions with SphI and SalI, respectively, andthe resulting 2.16 kb SphI–SalI fragment was subsequentlyinserted into pUC19, which had been linearized by cleavagewith SalI and SphI. The insert of the resulting plasmid, pSB1,carries, in addition to the complete fusA gene, the 66 bppreceding the fusA start codon as well as the first 12 bp afterthe fusA stop codon.

Mutant fusA alleles encoding EF-G derivatives with theamino acid substitutions P406L, H457Y and L461K wereconstructed by site-directed mutagenesis, using pSB1 astemplate and the primer pairs P3/P4, P5/P6 and P7/P8,respectively, for introduction of these mutations. pSB1 deriv-atives carrying these codon substitutions in fusA could beselected by restriction analysis because the mutagenic prim-ers were designed such that they also introduced silent muta-tions that generate or destroy a restriction site (an EcoRI sitein fusA was destroyed by primer pair P3/P4, whereas EcoRVsites were introduced by primer pairs P5/P6 and P7/P8)(Fig. 2). The presence of the mutations in the selectedplasmids was subsequently confirmed by DNA sequenceanalysis. Finally, the wild-type fusA gene and the mutant fusAalleles were excised from the corresponding plasmids bycleavage with SphI and SalI and inserted into the E. coli/S.aureus shuttle expression vector pHPS9, resulting in theplasmid pSB2 (containing the wild-type fusA gene) andthe mutant derivatives pSB2P406L, pSB2H457Y andpSB2L461K. In all these plasmids, fusA is under the controlof the promoter that is provided by the vector.

Isolation of RNA and RT-PCR

To isolate the total RNA of S. aureus RN4220/pSB2 and S.aureus RN4220/pSB2L461K, both strains were grown in 2¥YT medium supplemented with erythromycin (10 mg ml-1)and chloramphenicol (10 mg ml-1). The bacteria were har-vested in the mid-logarithmic growth phase (optical densityat 600 nm = 0.7–0.9). Before cell lysis, the bacteria weretreated with RNAprotectTM Bacteria Reagent (Qiagen)according to the manufacturer’s instructions, a procedure thatprovides immediate stabilization of the RNA. Lysis of the cellswas achieved by suspending the bacteria in TE buffer (10 mMTris-HCl, 1 mM Na-EDTA, pH 8.0) containing lysostaphin at

Molecular analysis of fusidic acid resistance in S. aureus 469


a concentration of 1 mg ml-1, followed by incubation for30 min at 37∞C. Subsequently, the total RNA of the bacteriawas isolated using the RNeasy Mini kit (Qiagen), accordingto the protocol provided by the manufacturer. Residualamounts of DNA were removed during RNA isolation by anon-column DNase I digestion using the RNase-free DNaseset from Qiagen.

RT-PCR analysis of the expression of plasmid-encodedwild-type and mutant fusA was performed with the QiagenOneStep RT-PCR kit, according to the manufacturer’s recom-mendations. The fusA-specific oligonucleotide P10 was usedas primer for reverse transcription of the fusA mRNA; PCRamplification of the generated cDNA was conducted with thefusA-specific forward primer P9 and the reverse primer P10to yield a PCR product of 410 bp. Best results were obtainedwhen ª10 pg of template RNA was used for the RT-PCR.

Acknowledgements

We would like to express our thanks to Wilma Ziebuhr andKnut Ohlsen for kindly supplying S. aureus strain RN4220 aswell as the plasmid pHPS9. We also thank Denia Frank,Christine von Rhein and Bernd Striebeck for their excellenttechnical assistance.

References

Andersson, D.I., and Levin, B.R. (1999) The biological costof antibiotic resistance. Curr Opin Microbiol 2: 489–493.

Björkman, J., Hughes, D., and Andersson, D.I. (1998) Viru-lence of antibiotic-resistant Salmonella typhimurium. ProcNatl Acad Sci USA 95: 3949–3953.

Björkman, J., Nagaev, I., Berg, O.G., Hughes, D., andAndersson, D.I. (2000) Effects of environment on compen-satory mutations to ameliorate costs of antibiotic resis-tance. Science 287: 1479–1482.

Bodley, J.W., Zieve, F.J., Lin, L., and Zieve, S.T. (1969)Formation of the ribosome-G factor-GDP complex in thepresence of fusidic acid. Biochem Biophys Res Commun37: 437–443.

Brown, E.M., and Thomas, P. (2002) Fusidic acid resistancein Staphylococcus aureus isolates. Lancet 359: 803.

Chopra, I. (1976) Mechanisms of resistance to fusidic acidin Staphylococcus aureus. J Gen Microbiol 96: 229–238.

Godtfredsen, W., Roholt, K., and Tybring, L. (1962) Fucidin.A new orally active antibiotic. Lancet i: 928–931.

Haima, P., van Sinderen, D., Schotting, H., Bron, S., andVenema, G. (1990) Development of a beta-galactosidasealpha-complementation system for molecular cloning inBacillus subtilis. Gene 86: 63–69.

Johanson, U., and Hughes, D. (1994) Fusidic acid-resistant

mutants define three regions in elongation factor G of Sal-monella typhimurium. Gene 143: 55–59.

Kreiswirth, B.N., Lofdahl, S., Betley, M.J., O’Reilly, M.,Schlievert, P.M., Bergdoll, M.S., and Novick, R.P. (1983)The toxic shock syndrome exotoxin structural gene is notdetectably transmitted by a prophage. Nature 305: 709–712.

Lacey, R.W., and Rosdahl, V.T. (1974) An unusual ‘penicilli-nase plasmid’ in Staphylococcus aureus; evidence for itstransfer under natural conditions. J Med Microbiol 7: 1–9.

Laurberg, M., Kristensen, O., Martemyanov, K., Gudkov,A.T., Nagaev, I., Hughes, D., and Liljas, A. (2000) Structureof a mutant EF-G reveals domain III and possibly thefusidic acid binding site. J Mol Biol 303: 593–603.

Levin, B.R., Perrot, V., and Walker, N. (2000) Compensatorymutations, antibiotic resistance and the population genet-ics of adaptive evolution in bacteria. Genetics 154: 985–997.

Livermore, D.M. (2000) Antibiotic resistance in staphylococci.Int J Antimicrob Agents 16: S3–S10.

Ludwig, A., Bauer, S., Benz, R., Bergmann, B., and Goebel,W. (1999) Analysis of the SlyA-controlled expression, sub-cellular localization and pore-forming activity of a 34 kDahaemolysin (ClyA) from Escherichia coli K-12. Mol Micro-biol 31: 557–567.

Macvanin, M., Johanson, U., Ehrenberg, M., and Hughes, D.(2000) Fusidic acid-resistant EF-G perturbs the accumula-tion of ppGpp. Mol Microbiol 37: 98–107.

Martemyanov, K.A., Liljas, A., Yarunin, A.S., and Gudkov,A.T. (2001) Mutations in the G-domain of elongation factorG from Thermus thermophilus affect both its interactionwith GTP and fusidic acid. J Biol Chem 276: 28774–28778.

Nagaev, I., Bjorkman, J., Andersson, D.I., and Hughes, D.(2001) Biological cost and compensatory evolution infusidic acid-resistant Staphylococcus aureus. Mol Micro-biol 40: 433–439.

Price, C.T., and Gustafson, J.E. (2001) Increases in themutation frequency at which fusidic acid-resistant Staphy-lococcus aureus arise with salicylate. J Med Microbiol 50:104–106.

Sambrook, J., and Russell, D.W. (2001) Molecular Cloning:a Laboratory Manual, 3rd edn. Cold Spring Harbor, NY:Cold Spring Harbor Laboratory Press.

Shanson, D.C. (1990) Clinical relevance of resistance tofusidic acid in Staphylococcus aureus. J AntimicrobChemother 25: S15–S21.

Turnidge, J., and Collignon, P. (1999) Resistance to fusidicacid. Int J Antimicrob Agents 12: S35–S44.

Wichelhaus, T.A., Böddinghaus, B., Besier, S., Schäfer, V.,Brade, V., and Ludwig, A. (2002) Biological cost of rifampinresistance from the perspective of Staphylococcus aureus.Antimicrob Agents Chemother 46: 3381–3385.

Documents

Molecular analysis of fusidic acid resistance in Staphylococcus aureus