Comparison of various antimicrobial agents as catheter lock solutions:preference for ethanol in eradication of coagula-negative staphylococcal biofilms
Yue Qu,1Taghrid S.Istivan,1Andrew J.Daley,2Duncan A.Rouch1 and Margaret A.Deighton1
Correspondence Margaret A.Deighton
m_deighton@rmit.edu.au 1School of Applied Sciences,RMIT University,Plenty Road,Bundoora,Victoria3083,Australia
2Department of Microbiology,The Royal Children’s Hospital,Flemington Road,Parkville,Victoria 3052,Australia
Received4September2008 Accepted5January2009Coagula-negative staphylococci(CoNS)are the main causative agents of bacteraemia in infants managed in neonatal intensive care units(NICUs).Intraluminal colonization of long-term central venous catheters by the bacteria and subquent biofilm formation are the prerequisites of the bloodstream infections acquired in NICUs.The catheter lock technique has been ud to treat catheter colonization;however,the optimum choice of antimicrobial agents and their corresponding concentrations and exposure times have not be
en determined.The effectiveness of catheter lock solutions(CLSs)was assd by determining the minimal biofilm eradication concentration of antimicrobial agents against CoNS biofilms.Five conventional antibiotics (oxacillin,gentamicin,vancomycin,ciprofloxacin and rifampicin)alone or in combination,as well as ethanol,were evaluated.Ethanol was found to be superior to all of the conventional antibiotics when ud as a CLS.A time–kill study and confocal lar scanning microscopy revealed that exposure to40%ethanol for1h was sufficient to kill CoNS biofilm cells.To our knowledge,this is the first in vitro study to provide solid evidence to support the rationale of using ethanol at low concentrations for a short time as a CLS,instead of using conventional antibiotics at high concentrations for a long period to treat catheter-related bloodstream infections.
INTRODUCTION
Coagula-negative staphylococci(CoNS),predominantly Staphylococcus epidermidis,are the most common causative agents of bloodstream infections in neonatal intensive care units(NICUs)(Stoll et al.,2002;Villari et al.,2000). The bacteria are relatively avirulent compared with Staphylococcus aureus and other organisms that also cau bloodstream infections.The pathogenesis of CoNS infec-tion depends mainly on their ability to form biofilms on the surfaces of various polymers(Klingenberg et al.,2005; von Eiff et al.,2002).Long-term catheters are ud in both ambula
nt and hospitalized patients in areas such as intensive care and oncology to provide central venous access for various therapies(Ackoundou-N’guessan et al., 2006;Onland et al.,2006;Opilla et al.,2007;van de Wetering&van Woenl,2007).Major adver effects of long-term catheterization are catheter-related bloodstream infections(CRBSIs),due to intraluminal colonization of catheters by CoNS,and subquent biofilm formation on the surface of the catheter lumen.
Systemic antibiotics have been widely ud to treat CRBSIs, but failures have been reported frequently(Allon,2004; Benjamin et al.,2001;Berrington&Gould,2001;Gagnon et al.,1993)due to the inability of most conventional antibiotic therapies to eradicate biofilm-grown bacteria (Donlan,2000;Klingenberg et al.,2005)or to poor access of antibiotics to the surface of the catheter lumen(Bastani et al.,2000).The deficits can be overcome by the catheter lock technique(CLT),which involves filling the lumen of the catheter with an antimicrobial agent at high concen-tration(100–1000times higher than is ud systemically) and allowing the compounds to dwell for a period of time while the catheter is not in u(Messing et al.,1988). The effectiveness of various antimicrobial agents ud as catheter lock solutions(CLSs)to prevent or treat CRBSIs has been compared in numerous studies.Previous studies have shown that the effectiveness of antibiotics is
Abbreviations:CLS,catheter lock solution;CLSM,confocal lar-
scanning microscopy;CLT,catheter lock technique;CoNS,coagula-
negative staphylococci;CRBSI,catheter-related bloodstream infection;
MBEC,minimal biofilm eradication concentration;NICU,neonatal
intensive care unit;PI,propidium iodide.
Journal of Medical Microbiology(2009),58,442–450DOI10.1099/jmm.0.006387-0 442006387G2009SGM Printed in Great Britain
suboptimal(Allon,2003;Curtin et al.,2003;Fernandez-Hidalgo et al.,2006;Kite et al.,2004).In addition,the application of antibiotics may be associated with undesir-able effects,such as the development of resistance,allergic reactions and toxicity.In three recent cohort studies, ethanol was found to be extremely effective in the treatment of CRBSIs when applied as a CLS(Metcalf et al.,2004;Onland et al.,2006;Opilla et al.,2007). However,there is still a lack of scientific data on the effectiveness of ethanol in eradication of biofilm-grown bacteria.
Besides the paucity of data on the most effective anti-microbial agents ud as CLSs,their optimum concentrations and exposure times also need to be clarified further(Bailey et al.,2002;Berrington&Gould,2001).Most investigators have chon concentrations and exposure time arbitrarily, and hence there is much variation in the effectiveness reported in different studies(Berrington&Gould,2001;Lee et al.,2006;Onland et al.,2006;Opilla et al.,2007). METHODS
Bacterial isolates and growth conditions.Eight CoNS isolates were ud in this study,including two reference strains(RP62a,a biofilm-positive S.epidermidis strain,and SP2,a biofilm-negative Staphylococcus hominis strain)and six invasive clinical CoNS isolates from newborns with confirmed CRBSI.The isolates were obtained from blood cultures from infants at the Royal Women’s Hospital NICU,Victoria,Australia.Biofilm production of the isolates was examined quantitatively in a previous study by staining the biofilms with Hucker’s crystal violet and measuring the A600(Bradford et al., 2006).The isolates were:a biofilm-positive S.epidermidis(isolate no.
3),a biofilm-weak S.epidermidis(no.4),another biofilm-positive S. epidermidis(no.5)and three Staphylococcus capitis isolates,which only produce biofilms under stress induced by sodium chloride or ethanol(nos6,8and9).The icaA gene was found in all of the biofilm-positive S.epidermidis and S.capitis isolates,except for the biofilm-weak isolate S.epidermidis no.4(Bradford et al.,2006). Biofilm
s were developed in tryptone soy broth(TSB;Oxoid)in96-well microplates as described by Deighton et al.(2001).As S.capitis only produces biofilms when stresd with salt or ethanol,TSB with 4%(w/v)sodium chloride was ud for biofilm production of S. capitis nos6,8and9.
Antimicrobial agents.Reprentatives of five different class of antibiotic and one disinfectant were chon for the catheter lock experiments:oxacillin,gentamicin,vancomycin,ciprofloxacin,rifam-picin(Sigma-Aldrich)and ethanol(Merck).Antibiotic stock solu-tions were prepared according to the manufacturer’s instructions.To compare the effectiveness of single antimicrobial agents in catheter lock experiments,dilutions were prepared with sterile saline,starting from very high concentrations ud in the CLT as reported previously (Lee et al.,2006;Mermel et al.,2001;Onland et al.,2006;Opilla et al., 2007;Sherertz et al.,2006).The lected high concentrations were also consistent with published guidelines for the management of CRBSIs (Mermel et al.,2001).The starting concentrations were10000m g ml21for gentamicin and rifampicin,5000m g ml21for oxacillin, vancomycin and ciprofloxacin and80%(v/v)for ethanol.The concentrations are defined as pharmacological concentrations at which no apparent precipitation of antibiotic powder or evaporation of solutions occurs.The double or triple combinations of conventional antibiotics were chon bad on their reported effectiveness against CoNS biofilms (Saginur et al.,2006).Therapeuti
c concentrations were ud for individual antibiotics in combinations that have been defined as the Clinical and Laboratory Standards Institute breakpoints for deter-mining resistance(Table1).
In vitro model to predict the efficacy of CLSs.The efficacy of single antibiotics,antibiotic combinations and ethanol against CoNS biofilms was examined.Preformed24h biofilms were expod to different concentrations and combinations of antimicrobial agents in saline for24h as follow.Two hundred microlitres of single agents at concentrations ranging from4m g ml21to5000or10000m g ml21, double or triple antibiotic combinations at therapeutic concentrations (Table1)or ethanol(10–80%)were added to the biofilms.After24h of exposure,the biofilms were washed three times with saline to remove any residual antimicrobial agents.Sterile resin beads(0.2g per well,Amberlite XAD-16;Rohm and Haas)were then added to minimize the chance of antibiotic carryover.TSB(200m l)was then added to each well and microplates were incubated at37u C for a further48h.A microplate shaker(speed2,Titretek;Flow Laboratories)was ud to facilitate the multiplication and relea of any living cells remaining in the biofilms.After48h,150m l of the contents in each well of the microplate was transferred to a U-bottomed microplate and examined visually for turbidity.The lowest concentration of antimicrobial agent corresponding to clear wells was defined as the minimal biofilm eradication concentration(MBEC)for successful u in the CLT.This concentration is generally higher 新概念英语官网
than the MIC for planktonic cultures and also higher than the minimum biofilm inhibitory concentration for biofilms,and targets complete killing of biofilm-embedded cells.This method was defined as the broth recovery method,in contrast to the plate recovery method, which employs agar plates to recover the remaining living cells after treatment.
Optimum concentration and duration of ethanol exposure for catheter lock applications.Two hundred microlitres of ethanol at concentrations ranging from10to80%was added to24h CoNS biofilms and allowed to remain in contact for1min,15min,1h,4h, 8h or24h.After the specified contact time,the MBEC of ethanol was determined as described above,but without adding resin beads to the wells.The ethanol-treated biofilms of RP62a and SP2in the original microplate were then collected with sterile cotton-tipped Table1.Antibiotic combinations and the test concentrations of individual agents
Combination*Therapeutic concentrations(m g ml”1)D G:R16:4
元宵节快乐的英语O:R0.5:4
V:R32:4
C:R4:4
G:O:R16:0.5:4
G:V:R16:32:4
G:C:R16:4:4
O:V:R0.5:32:4
O:C:R0.5:4:4
V:C:R32:4:4
*G,Gentamicin;O,oxacillin;V,vancomycin;C,ciprofloxacin;R, rifampicin.
D G:R ratio,for example,of16:4reprents16m g gentamicin ml21 plus4m g rifampicin ml21.
Preference of ethanol for CoNS biofilm eradication
jmm.sgmjournals443
swabs and cultured on nutrient agar plates to detect any viable cells embedded in the biofilms that could not be detected by the broth recovery method.
Confocal lar-scanning microscopy(CLSM).Sterile polystyrene pieces(~0.560.5cm)were aptically cut from the bottom of a24-well microplate with extreme caution to avoid any scratch to the tissue culture treated surface.The polystyrene pieces were then transferred to a microwell containing an S.epidermidis RP62a bacterial suspension of~107c.f.u.ml21and incubated overnight at 37u C.The pieces,with biofilm,were then rind three times with 0.9%saline to remove planktonic bacteria.One of the following treatments was applied to the S.epidermidis RP62a biofilms grown on the polystyrene pieces:(i)0.9%saline for24h,as an untreated biofilm control;(ii)5000m g oxacillin ml21for24h;(iii)5000m g vancomycin ml21for24h;(iv)256m g ciprofloxacin ml21for24h; (v)40%ethanol for1h;or(vi)40%ethanol for4h.The treated biofilms were then stained with a LIVE/DEAD BacLight viability kit, containing3.35m M SYTO-9and20m M propidium iodide(PI),at 22u C for15min in the dark.The fluorescently stained polystyrene pieces were then washed twice with0.9%saline.The viability of biofilm cells was examined immediately using a Zeiss LSM510Pa confocal lar-scanning microscope with Zeiss Axioplan upright microscopes(Leica Microsystems).To minimize artefacts associated with simultaneous dual wavelength excitation,all samples were quentially scanned,frame-by-frame,first at488nm and then at 561nm.A663oil objective was ud in all imaging experiments. Data analysis.All experiments were run in duplicate and repeated at least three times.A fourth duplicate was performed if the results were inconsistent.
RESULTS AND DISCUSSION
Comparison of the effectiveness of antibiotics and ethanol as a CLS
Ethanol was found to be the most efficient of the antimicrobial agents examined for eradication of the CoNS biofilms after24h exposure.A low concentration of20%ethanol killed all cells embedded in the biofilms after overnight exposure.Gentamicin,oxacillin and vancomycin,even at very high pharmacological concentra-tions(10000,5000and5000m g ml21,respectively),failed to eradicate the biofilm cells of most CoNS isolates.The only exceptions were the biofilms formed by S.epidermidis nos3and4,which responded to gentamicin.In addition, ciprofloxacin and rifampicin achieved biofilm killing of all isolates when exposure lasted for24h,but only at relatively high concentrations(32–128and256–512m g ml21, respectively)(Table2).
The inability of conventional antibiotics commonly ud in NICUs(gentamicin,oxacillin and vancomycin)at very high concentrations to kill all biofilm cells indicates their unsuitability as a CLS.This in vitro finding is supported by frequent reports of failure of CRSBIs to respond to oxacillin,gentamicin,vancomycin or teicoplanin in the abnce of catheter removal(Allon,2004;Castagnola et al., 2006;Fernandez-Hidalgo et al.,2006;Guedon et al.,2002). Alth
ough vancomycin and teicoplanin lock solutions have been found to be effective in the prevention of catheter-related infections(van de Wetering&van Woenl,2007), their effectiveness on established CoNS biofilms was found to be poor in this study.Ciprofloxacin and rifampicin were able to completely kill biofilm-embedded cells.However, the value of ciprofloxacin and rifampicin in biofilm killing is limited by the high dos and related expen and drug toxicity,and the ea with which rifampicin induces resistance in bacterial populations(Dunne et al.,1993). Our findings were consistent with a recent study by Lee et al.(2006),which found that ciprofloxacin(1000–5000m g ml21)and rifampicin(5000m g ml21)completely eradicated staphylococcal biofilms after a short-term exposure,in contrast to vancomycin(5000m g ml21)and gentamicin(10000m g ml21),which were much less effective,even after long-term exposure(Lee et al.,2006). Differences in the MBEC values of single antibiotics reported in our study and the study by Lee et al.(2006) could be due to the different diluents ud for preparing the CLS or the different ages of the biofilms.Sterile saline was ud in our study as the test medium to mimic the clinical situation,whilst TSB,a rich medium for bacterial growth,was ud in the study by Lee et al.(2006).We examined the effect of antimicrobial agents on24h
Table2.MBECs of single antimicrobial agents against24h CoNS biofilms
Values are geometric means of at least three duplicates.Variation between three duplicates was less than fourfold in all instances.
Strain identity Gentamicin
(m g ml”1)Oxacillin
(m g ml”1)
Vancomycin
(m g ml”1)
Ciprofloxacin暮光之城男女主角
(m g ml”1)
Rifampicin
(m g ml”1)
Ethanol(%)
S.epidermidis RP62a.10000.5000.50006451220
S.hominis SP2.100005000.50003251220
S.epidermidis no.3128.5000.50006451220
S.epidermidis no.48.5000.500012825620
S.epidermidis no.5.10000.5000.500012851220
S.capitis no.6.10000.500050006451220
S.capitis no.8.10000.5000500012851220
S.capitis no.9.10000.500050006451220
Y.Qu and others
444Journal of Medical Microbiology58
biofilms,whilst Lee et al.(2006)ud5-day biofilms,which could be expected to behave differently. Inte
restingly,there were no significant differences found in the prent study in the MBECs of single antimicrobial agents for strains with different biofilm-forming abilities. The concentrations of the agents required to eradicate the adherent cells of the biofilm-negative S.hominis SP2 were the same as tho required for S.epidermidis RP62a biofilms,although the latter formed much thicker biofilms than the former(Table2).Although classified as a biofilm-negative strain,SP2does form an adherent monolayer on various biomaterials(Christenn et al.,1985).It is therefore likely that the biofilm mode of growth,rather than the ability to form den biofilm,leads to its resistance to antimicrobial agents.A parallel study by us comparing the antimicrobial susceptibilities of biofilms,adherent monolayers and planktonic cells at similar densities reported the same conclusion when targeting99.99% biofilm bacterial killing(unpublished data).Similar results were also found in a recent study,in which a biofilm-negative mutant strain,M187sn3,had the same or an even higher resistance level to most antibiotic locks as its biofilm-positive ancestor(Edmiston et al.,2006).
At therapeutic concentrations,most combinations failed to kill the biofilm cells,indicating the ineffectiveness of biofilm killing of antibiotic combinations at achievable rum concentrations(Table3).Combinations containing gentamicin were effective against biofilms of S.epidermidis no.4,unless they contained oxacillin(Table3).Van-comycin combined with rifampicin ha朋友的英文单词
s been suggested for the treatment of S.epidermidis biofilm-associated infec-tions(Monzon et al.,2001;Shama et al.,2002),but our study found no advantage of this combination over rifampicin alone in biofilm eradication.Vancomycin may,however,minimize the risk of rifampicin resistance developing(Zimmerli et al.,1994).Besides the suboptimal effectiveness of conventional antibiotics,the risk of developing resistance,the high cost of antibiotic usage and the possibility of the accidental infusion of highly concentrated solutions into the circulation during the lock procedure and related toxicity cannot be ignored.Optimum concentration of ethanol and duration of treatment for catheter locks
Biofilm cells of all isolates were killed within1min of exposure to60–80%ethanol and in1h of exposure to 40%ethanol.Exposure for.1h was required only for 20%ethanol and no further advantage was achieved by increasing the dwell time beyond4h(Fig.1).
In agreement with Sherertz et al.(2006),we showed that ethanol eradicates biofilm bacteria much more effectively than commonly ud antibiotics if ud as a CLS. Moreover,the u of ethanol for the CLT is simpler and easier than previously recommended antibiotic regimens. We also extended the findings further by showing that exposure of CoNS biofilms to40%ethanol for1h was sufficient to eradicate the biofilm bacteria.In addition,low concentrations of ethanol(25–60%)do not cau cath
爱好的英文
eter occlusion(Laird et al.,2005)or degradation of the catheter biomaterial(Opilla et al.,2007),and ethanol is compatible with heparin and EDTA,the anticoagulant agents com-monly ud in the CLT(Ackoundou-N’guessan et al.,2006; Raad et al.,2007).A shorter dwell time also reduces the risk of spillage of ethanol into the circulation and its associated toxicity(Ackoundou-N’guessan et al.,2006).We therefore suggest that this concentration and exposure time is optimum for clinical u.
The superiority of ethanol over conventional antibiotics can be explained by its hydrophilic nature and small size,which enable it to obtain access to deep biofilm cells,as the biofilm matrix is highly hydrated.This is in contrast to other antimicrobial agents that may have restricted diffusion gradients through the biofilm matrix.Although it has been suggested that1–2mm thick biofilms may have a diffusion delay for ethanol to pass through the den biofilm structure, no solid evidence has been prented(Sissons et al.,1996).In addition,as ethanol has extremely high bactericidal activity against planktonic cells,it could also be effective in killing the ‘less active’cells in he cells prenting biofilm-specific genes or the more tolerant‘persister’cells.
广州话翻译Our CLSM study revealed that exposure to40%ethanol for1–4h,or256m g ciprofloxacin ml21for24h,worked
Table3.Regrowth of biofilm bacteria after exposure to double and triple combinations of antibiotics at therapeutic concentrations For abbreviations and therapeutic concentrations of individual antibiotics,e Table1.+,Regrowth;2,no regrowth.
Strain identity R:G R:V R:O R:C R:G:O R:G:V R:G:C R:O:V R:O:C R:V:C S.epidermidis RP62a++++++++++
S.hominis Sp2++++++++++
S.epidermidis no.3++++++++++
S.epidermidis no.42++++22+++
S.epidermidis no.5++++++++++
S.capitis no.6++++++++++
S.capitis no.8++++++++++
S.capitis no.9++++++++++
2013考研英语二真题
Preference of ethanol for CoNS biofilm eradication jmm.sgmjournals445
most efficiently in biofilm cell killing (Fig.2d–f).Exposure to first-line conventional antibiotics,such as oxacillin and vancomycin,at very high concentrations only partially killed the biofilm cells (Fig.2b,c).
CLSM in combination with a LIVE/DEAD BacLight viability kit has been ud successfully to examine the efficacy of antimicrobial agents on S.epidermidis biofilm eradication (Qin et al.,2007a,b).The LIVE/DEAD BacLight viability kit is a fluorescence-bad stain for determining bacterial cell viability and contains two components.SYTO-9is a membrane-permeant nucleic acid-labelling dye that labels all cells and fluoresces green.The cond dye,PI,is another nucleic acid-labelling dye that only enters cells with compromid or damaged cytoplasmic membranes.PI has been reported to have a stronger affinity for nucleic acids,competing with and quenching the SYTO-9and then fluorescing red (Stocks,2004).According to the manufacturer,once stained with the LIVE/DEAD BacLight viability kit,the dead cells in a bacterial population prent as a bright red colour,whilst live cells are a bright green colour,leaving the background non-fluorescent.This theory is consistent with our results when cell-wall-attacking antibiotics,such as oxacillin and vancomycin,were tested (Fig.2b,c).However,when antimicrobial agents with intracellular targets or more complex killing mechanisms were tested in our study,the dead cells prented as a reduced gr
een colour or displayed an orange colour,indicating that only partial quenching of SYTO-9by PI or simultaneous staining of cells by PI and SYTO-9had occurred.Similar results were obtained in a previous study by Berney et al.(2007).The authors found that Escherichia coli with an intact outer membrane showed strong green fluorescence intensity and tho cells with cytoplasmic membrane damage displayed a decread green fluorescence intensity.An incomplete replacement
and quenching of SYTO-9fluorescence by PI has been propod to explain this phenomenon (Berney et al.,2007).Bad on the report by Berney et al.(2007)as well as the current study,reduction of green fluorescence density,but not the absolute abnce of green signal,should be ud as indicators of bacterial killing when the BacLight kit is ud to examine viability.This theory is more significant when antimicrobial agents with intracellular activity rather than cell-wall-active agents are iprofloxacin and ethanol.The incomplete replacement could be explained by the poor effect of the antimicrobial agents on membrane integrity in Staphylococcus spp.(O’Neill et al.,2004),the much lower fluorescent intensity of PI compared with SYTO9(8:300)when excited (Stocks,2004)or the prence of viable but non-culturable cells in the biofilms after antimicrobial exposure (Alam et al.,2007).
Concerning the methodology,the definition of successful catheter lock application ud in the prent
paymentstudy was the complete abnce of recoverable living cells following in vitro exposure.Successful treatment using the catheter lock technique has been defined as negative blood cultures and resolution of fever a few days after completion of a 3-week treatment cour (Allon,2004).However,this definition has not been validated by in vitro studies.Unless complete eradication of biofilm cells is achieved,the remaining living cells will rebuild the biofilm once antimicrobial stress diminishes,as host immune mechanisms can neither reach the lumen of catheters nor eradicate any remaining living cells in biofilms (Lewis,2005;Spoering &Lewis,2001).Currently,there is no universally accepted in vitro model of catheter-related biofilm infection.The model of biofilm formation and antimicrobial treatment ud in our study was intended to replicate the clinical situation as cloly as possible.Biofilm development and treatment were carried出国留学申请流程
100
2030405060708090E t h a n o l c o n c e n t r a t i o n (%)
S. epidermidis RP62a S. hominis SP2S. epidermidis no. 3S. epidermidis no. 4S. epidermidis no. 5S. capitis no. 6S. capitis no. 8S. capitis no. 9
1 min
leyes
15 min
1 h 4 h 8 h
24 h
Exposure time
Fig.2.Biofilms of S.epidermidis RP62A treated with antimicrobial agents and then stained with SYTO-9(bright green for live cells)and PI (red or orange,or loss of bright green for dead cells).(a)Untreated biofilm control;(b)biofilm treated with 5000m g oxacillin ml ”1for 24h;(c)biofilm treated with 5000m g vancomycin ml ”1for 24h;(d)biofilm treated with 256m g ciprofloxacin ml ”1for 24h;(e)biofilm treated with 40%ethanol for 1h;and (f)biofilm treated with 40%ethanol for 4h.Y.Qu and others
446
Journal of Medical Microbiology 58
Preference of ethanol for CoNS biofilm eradication
(a)(b)
(c)(d)
(e)(f)
jmm.sgmjournals447
