International Journal of Antimicrobial Agents 24 (2004) 369–373

In vitro activity of RBx 7644 (ranbezolid) onbiofilm producing bacteria

Tarun Mathura, Pragya Bhatejaa, Manisha Pandyaa,Tasneem Fatmab, Ashok Rattana,∗

a Department of Microbiology, Ranbaxy Research Laboratories, R&D II, Plot no. 20, Sector-18,Udyog Vihar Industrial Area, Gurgaon 12200, Haryana, India

b Jamia Millia Islamia University, New Delhi, India

Received 24 October 2003; accepted 14 April 2004

Abstract

Biofilm associated infections are becoming more common. Treatment outcome of device related infections cannot be predicted by theresults of a standard susceptibility test such as the MIC. Microorganisms involved in device related infections have a slow growth rate andadhere to surfaces. Activity against glass adherent bacteria has been shown to be correlated with treatment outcome in animal models ofcatheter related infections. Drug efficacy can be predicted if glass adherent staphylococci are killed by low drug concentration. The eradicationof bacteria adhering to glass beads and inhibition of biofilm formation by ranbezolid and three other antibiotics (quinupristin/dalfopristin,vancomycin and linezolid) was studied. The results indicated that ranbezolid required only (2–4× MIC) for total clearance of glass-adherentMRSA 562 and MRSE 879, compared with vancomycin (8×), quinupristin/dalfopristin (1–4×) and linezolid (4–16× MIC). In additionranbezolid inhibited biofilm formation to a greater extent at sub MIC and MIC level. In conclusion, this study indicated that ranbezolid hadpotent activity against adherent staphylococci isolates and may prove useful in the prevention and treatment of device related infections causedby staphylococci.© 2004 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.

Keywords: Biofilm; Device related infection; Glass adherent bacteria; Slime production; Ranbezolid

1. Introduction

Catheters are a frequent cause of nosocomial infec-tions and primary septicaemia[1]. Catheter related infec-tions are most commonly caused by coagulase-positiveor coagulase-negative staphylococci (CoNS), particularlyslime producing strains[2]. Routine antimicrobial suscepti-bility tests are not good predictors of treatment outcome ofcatheter related infections[3,4] leading to the dogma thatinfected implants have to be removed in order to achievecure. It has been shown that killing of non-growing andadherentStaphylococcus epidermidis defines drug efficacyin device related infections[3,4] and that the cure rate inexperimental device related infections can be predicted bythe in vitro bactericidal effect of antibiotics on non-growingand adherent bacteria[5]. In this study we have investi-

∗ Corresponding author. Tel.:+91 124 2397232;fax: +91 124 2343545.

E-mail address: ashok.rattan@ranbaxy.com (A. Rattan).

gated the in vitro activity of ranbezolid (RBx 7644)[6,7]on biofilm production and its bactericidal activity againstslime producing bacteria.

2. Materials and methods

2.1. Bacteria

The strains chosen for study were methicillin-resistantS. epidermidis (MRSE) 879, methicillin-resistantStaphylo-coccus aureus (MRSA) 562, MRSA 1026/99 and a clinicalisolate and moderate slime producerS. epidermidis ATCC35983.

2.2. Estimation of biofilm production

Slime production was determined by the microplate ad-herence assay[8]. Briefly, the cultures were grown overnightin brain heart infusion (BHI) broth. Cultures were diluted

0924-8579/$ – see front matter © 2004 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.doi:10.1016/j.ijantimicag.2004.04.012

370 T. Mathur et al. / International Journal of Antimicrobial Agents 24 (2004) 369–373

1:50 in BHI broth supplemented with 0.25% glucose. Twohundred microliters of the diluted cultures were added toa 96-well flat-bottom polystyrene microtitre plate in dupli-cates in each run. After incubation overnight at 37◦C un-der static conditions the amount of growth in each wellwas measured at an optical density at 544 nm (OD544 nm).Wells were emptied and washed three times with phosphatebuffered saline (PBS) (pH 7.3) and air-dried. Adherent bac-teria were stained with 0.1% safranin for 1 h, excess stainwas removed by rinsing the wells with distilled water, wellswere tap dried and adherent material was solubilized by in-cubation with 200�l of 0.2 M sodium hydroxide for 1 h at85

◦C. The OD544 nm was measured.

S. epidermidis ATCC 12228, a poor slime producer andS.epidermidis ATCC 35983, a moderate slime producer weretaken as control strains for slime production. Each exper-iment was performed three times and the mean OD withstandard deviation was calculated.

2.3. MIC determination

MICs of linezolid, vancomycin, quinupristin/dalfopristin(commercial sources) and ranbezolid (Ranbaxy Labs. Ltd.,India) for the bacterial strains selected for biofilm studieswere determined using cation adjusted Mueller Hinton brothaccording to the NCCLS micro broth dilution method[9].

2.4. In vitro colonization of glass adherent bacteria

MRSE 879 and MRSA 562 were used for the study andthe method followed was that described by Zimmerli et al.[5]. The adherent inocula were prepared using sinteredglass beads containing pores of 60–300�m. These beadswere placed in a flask-containing medium (1% glucose and4.2% trypticase soya broth (TSB, Difco) supplemented with50 mg/l of Ca2+ and 25 mg/l Mg2+ in PBS).

The media was inoculated with overnight cultures and theflasks were incubated at 37◦C on a shaker at 70 rpm. After24 h the medium was aspirated and the beads with biofilmwere washed using phosphate buffered saline to remove non-adherent bacteria. The viable count of adherent bacteria wasdetermined by placing the washed glass beads in 2 ml ofsaline containing ethylene diamine tetra acetic acid (EDTA)(0.15%) and Triton-X 100 (0.1%). Vigorous mixing and son-ication for 3 min removed the adherent bacteria and cfu/beadwas determined by a serial dilution method. The mean ofviable count recovered from six different beads is definedas the adherent inoculum.

2.5. Killing of glass adherent bacteria

Concentrations 0.125–32 mg/l of all four antibiotics wereprepared in 5 ml cation adjusted Mueller Hinton broth intriplicates. Each tube was inoculated with one sintered glassbead coated with adherent bacteria. The tubes were incu-bated at 37◦C for 24 h. Adherent bacteria were washed and

harvested after incubation[5] and the mean cfu/bead of eachconcentration was determined. The mean log killing of ad-herent bacteria for all four drugs was calculated as follows:log cfu (geometric mean) of untreated beads (inoculum)−log cfu (geometric mean) of antibiotic treated beads.

The concentration of each drug required to kill adherentbacteria was decided after performing three independent tri-als. The mean log killing and standard deviation were de-termined for each concentration.

2.6. Inhibition of biofilm formation

2.6.1. Bacteria and antibioticsModerate slime producing isolates MRSE 879, MRSA

1026/99 andS. epidermidis ATCC 35983 were used for thestudy along with the four antibiotics used before.

2.6.2. ProcedureThe cultures were inoculated in TSB, supplemented with

2% glucose and incubated at 37◦C for 24 h. Inhibition ofbiofilm formation was confirmed by an adherence biofilmassay using 96-well flat bottom polystyrene microtitre plates[10].

The overnight cultures were diluted 1:10 in TSB (2%glucose), two-fold dilutions of antibiotics were also pre-pared using the same diluent to obtain a concentration rangeof 0.03–16 mg/l in triplicate. Wells with sterile TSB aloneserved as a negative control and the mean OD544 nmfor thesewells were subtracted from the OD544 nm value of the testwells.

The plates were incubated in an incubator shaker at 37◦C,

110 rpm for 24 h. Following incubation the liquid was aspi-rated and wells were washed with PBS (pH 7.3). Staining ofadherent bacteria was performed as described in estimationof slime. The OD544 nm was measured. The relative inhibi-tion of biofilm (expressed as mean percentage) was deter-mined as follows:

Percent inhibition

= 100−[(

OD544 nmof antibiotic well

OD544 nmof positive control well

)× 100

]

Each experiment was repeated three times and the meanpercent inhibition and standard deviation was consideredwhen comparing the results of all tested drugs.

2.7. Statistical analysis

The statistical significance of killing of adherent bacteriaand relative percentage of inhibition were determined. Com-parison of results between the groups used one-way anal-ysis of variance and multiple comparisons used Dunnet’smultiple comparison posttest. AP value of<0.01 was con-sidered statistically significant. The statistical analysis wasperformed using Graph Pad prism 2.01(2) version for Win-dows 2001 (Graph Pad software, San Diego.)

T. Mathur et al. / International Journal of Antimicrobial Agents 24 (2004) 369–373 371

Table 1Biofilm production byS. aureus and S. epidermidis isolates

Strain Optical density (OD544)(mean+ S.D.)

1 MRSE 879 0.224+ 0.0142 MRSA1026/99 0.215+ 0.0193 MRSA 562 0.389+ 0.0284 S. epidermidis ATCC 35983 0.210+ 0.0255 S. epidermidis ATCC 12228 0.104+ 0.010

3. Results

3.1. Estimation of biofilm production

Bacteria showed different levels of slime production,S.epidemidis ATCC 35983 showed a mean OD544nm of 0.21± 0.025, whereas the OD544nm for S. epidermidis ATCC12228 was 0.104± 0.010. The clinical isolates MRSE 879and MRSA1026/99 classified as moderate slime producersshowed mean OD544nmof 0.224± 0.014 and 0.215± 0.019,respectively and that of MRSA 562 was slightly higher at0.389 ± 0.028 (Table 1). These slime-producing isolateswere selected for further studies.

3.2. MIC of isolates

The four selected slime-producing bacteria were sen-sitive to all tested drugs and MICs (mg/l) were linezolid(2 mg/l), vancomycin (1–2 mg/l), quinupristin/dalfopristin(0.25–0.5 mg/l) and ranbezolid (0.25–2 mg/l) (Table 2).

3.3. Killing of adherent staphylococci

The adherent inoculum was between 5.41× 106 and1.9 × 107 cfu/bead for MRSA 562 and 4.9× 106 and 5.3× 107 cfu/bead for MRSE 879. The mean of adherent in-oculum was used to determine the reduction of adherentstaphylococci. Ranbezolid significantly (P < 0.01) reducedadherent MRSA 562 and MRSE 879 and total clearance wasachieved at 2 mg/l (2× and 4× MIC), when compared withvancomycin 16 and 8 mg/l (8×), linezolid 8 and 32 mg/l(4× and 16× MIC), respectively.

Quinupristin/dalfopristin produced bacterial clearance at0.5 and 2 mg/l (1× and 4×) for these strains (Table 3).

Table 2MICs (mg/l) of isolates selected in the study

Strain Vancomycin Linezolid Quinupristin/dalfopristin

Ranbezolid

MRSE 879 1 2 0.5 0.5MRSA 562 2 2 0.25 1MRSA 1026/99 2 2 0.5 2S. epidermidis ATCC 35983 2 2 0.5 0.25

Table 3Killing of glass adherent bacteria

Strains Antibiotics MIC(mg/l)

Concentration forclearance of glassadherent bacteria(mg/l)

MRSE 879 Vancomycin 1 8Quinupristin/dalfopristin

0.5 2

Linezolid 2 32Ranbezolid 0.5 2

MRSA 562 Vancomycin 2 16Quinupristin/dalfopristin

0.5 0.5

Linezolid 2 8Ranbezolid 1 2

3.4. Inhibition of biofilm formation

In the presence of sub inhibitory concentrations (1/2MIC), ranbezolid significantly inhibited biofilm forma-tion. Mean percent inhibition against MRSE 879, MRSA1026/99 andS. epidermidis ATCC 35983 were 76, 62 and68, respectively, significantly higher (P < 0.01) than theother three antibiotics.

At 1× MIC ranbezolid produced 93 and 91% biofilminhibition against MRSE 879 and MRSA 1026/99, comparedwith linezolid (76 and 63%) and quinupristin/dalfopristin (69and 50%) (P < 0.01). No significant inhibition was foundwith vancomycin and the mean percent inhibition for MRSE879 and MRSA 1026/99 were 69 and 29%, respectively.

Ranbezolid and linezolid significantly inhibited biofilmformation at MIC level (85 and 86.7%) againstS. epider-midis ATCC 35983, whereas quinupristin/dalfopristin andvancomycin showed 77 and 21% inhibition (Fig. 1).

4. Discussion

Intravascular catheters are essential in complex medicaland surgical interventions. More than 150 million intravas-cular catheters are used annually by clinics and hospitals inthe USA alone[11].

Maki et al. [12] found that the risk of venous catheter-related blood stream infections (CRBSI) was directlyproportional to the number of organisms counted by roll

372 T. Mathur et al. / International Journal of Antimicrobial Agents 24 (2004) 369–373

Fig. 1. Mean percent inhibition (±S.D.) of biofilm formation of MRSE 879, MRSA 1026/99 andS. epidermidis ATCC 35983.

T. Mathur et al. / International Journal of Antimicrobial Agents 24 (2004) 369–373 373

plate technique. Several studies have examined the effectof various types of antimicrobial treatment in controllingbiofilm formation on catheters[13–16]. Another area ofgreat importance is the role of biofilms in antimicrobialdrug resistance. Bacteria within biofilms are intrinsicallymore resistant to antimicrobial agents than planktonic cellsbecause of the altered rate of mass transport of antimicro-bial molecules to the biofilm-associated cells or becausebiofilm cells differ physiologically from planktonic cells.Antibacterial concentrations sufficient to inactivate plank-tonic organisms are generally not adequate to inactivatebiofilm organisms, especially those deep within the biofilm;this may select for a resistant subpopulation[17–19]. Ran-bezolid has potent activity against bothS. aureus and CoNS[6] and inhibits biofilm formation at sub MIC and MIC con-centrations, thereby decreasing the likelihood of catheterrelated infections.

Vancomycin exerted a limited effect on inhibition ofbiofilm observed in the study and this agrees with previousstudies[10,20–22].

In this study we showed that ranbezolid and quin-upristin/dalfopristin have significant activity against slimeproducing staphylococci and that the MIC of the two celltypes changes marginally from 0.5 to 2 mg/l. There was,however, a significant shift in MIC under the same condi-tions for vancomycin and linezolid.

Due to its potent activity against slime producing staphy-lococci and no significant loss of activity in the presence ofa foreign body, ranbezolid may emerge as an important an-tibiotic for the prevention and treatment of these infections.

Acknowledgements

We appreciate the help of Biswajit Das and Ajay Yadavfor providing ranbezolid, Sachin Malhotra and Vikesh Shri-vastava for statistical analysis and Smita Singhal for criticalreview.

References

[1] Maki DG, Infections due to infusion therapy. In: Bennett JV, Brach-man PS, editors. Hospital infections. Boston: Little Brown & Co.;1992.

[2] Raad I, Darouiche R, Hachem R. Antibiotics and prevention ofmicrobial colonization of catheters. Antimicrob Agents Chemother1995;39:741–6.

[3] Widmer AF, Frei R, Rajacic Z. Correlation between in vivo and invitro efficacy of antimicrobial agents against foreign body infections.J Infect Dis 1990;162:96–102.

[4] Widmer AF, Wiestener A, Frei R. Killing of non-growing and ad-herent Escherichia coli determines drug efficacy in device relatedinfections. Antimicrob Agents Chemother 1991;35:741–6.

[5] Zimmerli W, Frei R, Widmer AF, Rajacic Z. Microbiological test topredict treatment outcome in experimental device related infectiondue toStaphylococcus aureus. J Antimicrob Chemother 1994;33:959–67.

[6] Hoellman DB, Lin G, Ednie LM, Rattan A, Jacobs MR, AppelbaumPC. Antipneumococcal and antistaphylococcal activities of ranbezolid(RBx 7644), a new oxazolidinone, compared to those of other agents.Antimicrob Agents Chemother 2003;47:1148–50.

[7] Ednie LM, Rattan A, Jacobs MR, Appelbaum PC. Antianaerobeactivity of RBx 7644 (ranbezolid), a new oxazolidinone, comparedwith those of eight other agents. Antimicrob Agents Chemother2003;47:1143–7.

[8] Blake JE, Metcalfe MA. A shared noncapsular antigen is respon-sible for false positive reactions byStaphylococcus epidermidis incommercial agglutination test forStaphylococcus aureus. J Clin Mi-crobiol 2001;39:544–50.

[9] National Committee for Clinical Laboratory Standards. Methods fordilution antimicrobial susceptibility tests for bacteria that grow aero-bically. 5th ed. Approved standard. NCCLS Publication no. M7-A5.Pennsylvania, Wayne: National committee for Clinical LaboratoryStandards; 2000.

[10] Polonio RE, Mermel LA, Paquette GE, Sperry JF. Eradication ofbiofilm–forming Staphylococcus epidermidis (RP 62A) by a com-bination of sodium salicylate and vancomycin. Antimicrob AgentsChemother 2001;45:3262–6.

[11] Raad I. Intravascular-catheter-related infections. Lancet1998;351:893–8.

[12] Maki DG, Weise CE, Sarafin HW. A semiquantitative culture methodfor identifying intravenous-catheter-related infections. New Engl JMed 1977;296:1305–9.

[13] Mandell GL. The antimicrobial activity of rifampin: emphasis on therelation to phagocytes. Rev Infect Dis 1983;5(Suppl 5):S463–7.

[14] Minuth JN, Holmes TM, Musher DM. Activity of tetracycline, doxy-cycline and minocycline against methicillin susceptible and resistantstaphylococci. Antimicrob Agents Chemother 1974;6:411–4.

[15] Norwood S, Ruby A, Civetta J, Cortes V. Catheter related infectionsand associated septicemia. Chest 1991;99:968–75.

[16] Raad I, Darouiche, R. Central venous catheters (CVC) coated withminocycline and rifampin (M/R) for the prevention of catheter- re-lated bacteremia (CRB). In: Program and abstracts of the 35th In-terscience Conference on Antimicrobial Agents and Chemotherapy.Abstract J7. USA, Washington, DC: American Society for Microbi-ology; 1995. p. 258.

[17] Evans DJ, Allison DG, Brown MRW, Gilbert P. Susceptibility ofPseudomonas aeruginosa and Escherichia coli biofilms towardsciprofloxacin: effect of specific growth rate. J Antimicrob Chemother1991;27:177–84.

[18] Roberts AP, Pratten J, Wilson M, Mullany P. Transfer of a conjugativetransposon, Tn 5397, in a model of oral biofilm. FEMS MicrobiolLett 1999;177:63–6.

[19] Donlan RM. Biofilms and device associated infections. Emerg InfectDis 2001;7:277–81.

[20] Drago L, Vecchi EDe, Valli M, Nicola L, Gismondo MR. Effect oflinezolid in comparison with that of vancomycin on glycocalix pro-duction: in vitro study. Antimicrob Agents Chemother 2002;46:598–9.

[21] Tsidorias S, Gold HS, Sakoulas G, Eliopoulos GM, WennerstenC, Venkataraman L. Linezolid resistance in a clinical isolate ofStaphylococcus aureus. Lancet 2001;358:207–8.

[22] Carsenti-Etesse H, Durant J, Entenza J, et al. Effects of sub in-hibitory concentrations of vancomycin and teicoplanin on adherenceof staphylococci tissue culture plates. Antimicrob Agents Chemother1993;37:921–3.