ica and biofilm

ORIGINAL ARTICLE 10.1111/j.1469-0691.2006.01382.x

Assessment of ica operon carriage and biofilm production inStaphylococcus epidermidis isolates causing bacteraemia in bone marrowtransplant recipientsE. Ninin1, N. Caroff1, E. Espaze1, J. Maraillac1, D. Lepelletier1, N. Milpied2 and H. Richet1

1Laboratoire de Bacteriologie-Virologie-Hygiene Hospitaliere and 2Service d’Hematologie, CHUNantes, Nantes, France

ABSTRACT

The clinical significance of coagulase-negative staphylococci isolated from blood culture is typicallyassessed on the basis of a combination of clinical and microbiological criteria. However, these criteria aredifficult to apply to haematology patients who are highly immunosuppressed and from whom bloodcultures are obtained most frequently through a central venous catheter. This study analysed 112episodes of Staphylococcus epidermidis bacteraemia that occurred in 79 bone marrow transplant recipients.In 73 (65%) episodes, only one blood culture set was positive for S. epidermidis, while 39 (35%) episodesgrew S. epidermidis from multiple blood cultures. Nine patients had two or more episodes of bacteraemiawith the same strain, as determined by pulsed-field gel electrophoresis (PFGE). The PFGE method alsoshowed that 34 (31%) isolates belonged to seven clusters, indicating the persistence of certain clones inthe environment. Of the 109 isolates analysed, 59 (54%) produced biofilm and 91 (83.5%) carried the icaoperon. Isolates that produced biofilm were observed to colonise central venous catheters faster thannon-biofilm-producing isolates (18 vs. 37 days; p 0.03). No clinical features were associated with carriageof the ica operon, but the ica operon was carried more frequently by the isolates that formed clusters.

Keywords Bacteraemia, biofilm, bone marrow transplant patients, coagulase-negative staphylococci, ica operon,

Staphylococcus epidermidis

Original Submission: 25 October 2004; Revised Submission: 29 August 2005; Accepted: 30 September 2005

Clin Microbiol Infect 2006; 12: 446–452

INTRODUCTION

Coagulase-negative staphylococci (CNS) haveemerged as a frequent cause of device-relatedinfection [1]. Criteria proposed for decisions onclinical significance and treatment remain difficultto apply in clinical practice, especially in haema-tology and neonatal intensive care units, whereblood cultures are obtained most frequentlythrough central venous catheters [2]. A singleblood culture growing CNS is usually consideredto result from contamination, whereas multiplepositive blood cultures are thought to represent

true infection. However, Peacock et al. [3] haveshown that a single positive blood culture bottlecan also reflect the presence of a true infection.

An alternative way to differentiate contami-nants from pathogens would be to investigate thepresence of virulence factors identified in CNSstrains, such as biofilm production and the inter-cellular adhesin (ica) ADBC operon [4,5]. It hasbeen shown that the ica operon, encoding theproduction of polysaccharide intercellular adhe-sin (PIA), is detected significantly more fre-quently in infecting than in contaminatingstrains [6–8].

The aims of the present study were toinvestigate S. epidermidis bacteraemia in bonemarrow transplant (BMT) recipients, to comparethe clinical features of bacteraemia when one orseveral blood cultures were positive, and toassess whether detection of biofilm production

Corresponding author and reprint requests: E. Ninin, Labora-toire de Bacteriologie-Virologie-Hygiene Hospitaliere, Institutde Biologie des Hopitaux de Nantes, CHU de Nantes, 9 QuaiMoncousu, 44093 Nantes Cedex 01, FranceE-mail: [email protected]

� 2006 Copyright by the European Society of Clinical Microbiology and Infectious Diseases

and ⁄ or the presence of the ica operon couldhelp physicians to decide whether to prescribeglycopeptides.

MATERIALS AND METHODS

BMT recipients

Seventy-nine patients (53 autologous and 26 allogeneic BMTrecipients) were investigated during the 4-year period of thestudy. The following data were collected for each patient:gender, age, disease, allogeneic or autologous graft, and typeof conditioning chemotherapy. For each blood culture set, bothaerobic and anaerobic bottles were inoculated with 10 mL ofblood obtained through the central venous catheter. Usually,two or three sets of blood cultures were obtained in the case offever (temperature of > 38.5�C or ‡ 38�C for > 12 h). Bacterae-mia was defined as a positive culture in at least one bloodculture bottle. Positive sequential blood cultures were consid-ered to belong to the same infectious episode if they occurredwithin 72 h of each other. The following data were collected foreach instance of S. epidermidis bacteraemia: date of catheterinsertion; date of onset of bacteraemia; administration ofantimicrobial prophylaxis with co-trimoxazole before the onsetof bacteraemia; date of catheter removal; inflammatory orpurulent aspect of the catheter insertion site; presence of fever,shivers, hypotension, septic shock or localised infection;increase in serum creatinin level (>110 lmol ⁄L); initiation ofantimicrobial therapy; outcome of therapy; and duration ofhospitalisation.

Bacterial isolates

S. epidermidis isolates (n = 112) from blood cultures obtainedfrom BMT recipients treated at Nantes hospital between 1993and 1998 were included in the study. Bacterial isolates werecharacterised by standard microbiological procedures. CNSwere identified to the species level using the coagulase testand the ID32 STAPH biotyping system (bioMerieux, Marcyl’Etoile, France). Antibiotic susceptibility was tested by theagar diffusion method according to CA-SFM guidelines(http://www.sfm.asso.fr). The biofilm-forming and PIA-pro-ducing strain S. epidermidis RP62A was used as a positivecontrol [9]. The biofilm-negative strain Staphylococcus hominisSP2 and the ica-negative strain Staphylococcus schleiferi schleiferiATCC 43808 were used as negative controls for biofilmproduction [9,10].

Pulsed-field gel electrophoresis (PFGE)

DNA isolation and PFGE were performed as describedpreviously with minor modifications [11]. S. epidermidis iso-lates were grown from a single colony overnight at 37�C. Next,

0.3 mL of this bacterial suspension was mixed with 2 lL oflysostaphin Ambicin L and 300 lL of low melting temperatureagarose 2% w ⁄v. The resulting agarose blocks were incubatedfor 3 h at 25�C in 120 lL of restriction buffer containing 40 U ofSmaI. Electrophoresis was performed in an agarose 1% w ⁄vgel in a CHEF DR III apparatus (Bio-Rad, Hercules, CA, USA),with a constant voltage of 200 V and a pulse time increasingfrom 5 to 40 s over 20 h. Strain NCTC 8523 was used as a DNAsize marker. Banding patterns were interpreted according toTenover et al. [12].

PCR and primers for icaABCD

In total, 109 isolates were available for analysis of ica operoncarriage and biofilm production. DNA was extracted byboiling five colonies for 5 min in 300 lL of distilled water.The resulting suspension was centrifuged at 12 000 g for5 min and the supernatant was used as a DNA template.Four primers were chosen to amplify fragments of the icaA,icaB, icaC and icaD genes of different lengths (Table 1).Universal primers targeting 16S rRNA were used as aninternal control [13]. Multiplex PCRs were performed in afinal volume of 50 lL containing 10 mM Tris-HCl pH 8.3,50 mM KCl, 1.5 mM MgCl2, 200 lM each nucleotide,0.25 lM each primer, 1.25 U Taq DNA polymerase (Promega,Charbonnieres, France) and 10 lL of the bacterial DNAextract.

After a 5-min initial denaturation at 94�C, PCRs comprised30 cycles of 30 s at 94�C, 30 s at 55�C and 1 min at 72�C,followed by a final extension step of 7 min at 72�C, in a Perkin-Elmer 480 thermal cycler (PE Applied Biosystems, Cergy-Pontoise, France). Each PCR run included a control reactionwith no added DNA.

Biofilm production

Biofilm production was determined by the method of Chris-tensen et al. [9] with some modifications. Isolates were grownovernight in 10 mL of trypticase soy broth (TSB) containingglucose 2.5% w ⁄v, then diluted 1:200 in fresh TSB. Portions(200 lL) of the resulting suspension were used to inoculatesterile flat-bottomed 96-well polystyrene microtitre plates(Greiner Labortechnik, Poitiers, France). After incubation for24 h at 37�C, the supernatant was decanted and the wells werewashed gently twice with phosphate-buffered saline. The cellsremaining in the well were fixed for 10 min with ethanol andstained with safranin for 30 s. Optical densities were read on aUniversal Microplate Reader (Bio-Tek Instruments, Winooski,VT, USA) at a wavelength of 492 nm [14,15]. Each isolate wastested in eight wells. The lower cut-off was chosen asdescribed by Christensen et al. [9], i.e., approximately threestandard deviations above the mean of non-inoculated (blank)wells.

Table 1. DNA sequences of amplification primers

Target

genes Primer DNA sequence (5¢-3¢)Product

size (bp) Reference

icaA ica A-F ica A-R GAATCCAAAATTAGGCGCAG AGACCCATGTAATCGATGCG 455 This workicaB ica B-F ica B-R TTAAAGCACACGACGCAAAG TTTCTTCATCGAATCCGTCC 603 This workicaC ica C-F ica C-R TTATCGCTGTTTCCGGTAGTG CTCTTAACATCATTCCGACGC 300 This workicaD ica D-F ica D-R ACAGAGGCAATATCCAACGG CGTGTTTTCAACATTTAATGCAA 184 This work16S rRNA GGAGGAAGGTGGGGATGACG ATGGTGTGACGGGCGGTGTG 241 [13]

Ninin et al. S. epidermidis bacteraemia in BMT recipients 447

� 2006 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 12, 446–452

Statistical analysis

Data were analysed using the Epi info computer database (v.6;CDC, Atlanta, GA, USA). Fisher’s two-tailed exact test and thechi-square test were used to test the significance of association.For comparison of duration, the F-test of ANOVAANOVA was used if thevariance of samples was homogeneous, and the test ofKruskal–Wallis was used if the variance was heterogeneous.

RESULTS

Characteristics of BMT recipients withS. epidermidis bacteraemia

In total, 112 S. epidermidis isolates were recoveredfrom the blood of 79 patients, of whom 26 wereallogeneic BMT recipients and 53 were autolo-gous BMT recipients. The male:female ratio was1.8, and the median age was 48 years (range 16–68years). The characteristics of the patients and theepisodes of bacteraemia are summarised in Tables2 and 3. Eleven (14%) patients died duringhospitalisation, but no deaths were attributed toS. epidermidis infection.

Characteristics of S. epidermidis blood cultures

In 73 (65%) episodes of bacteraemia, only one setof blood cultures was positive for S. epidermidis,while multiple blood cultures grew S. epidermidisin 39 (35%) episodes of bacteraemia (median:three positive blood cultures, range two to six).Most (80%) isolates were resistant to methicillin.Another microorganism was isolated concomit-

antly in 14 episodes of bacteraemia involvingsingle blood cultures, and in six episodes ofbacteraemia involving multiple blood cultures(Table 4).

When the characteristics of patients with asingle positive blood culture were compared withthose of patients with multiple positive bloodcultures, shivers was the only sign associatedsignificantly with the latter group (18 ⁄ 39 (46%)vs. 20 ⁄ 73 (27.4%); p 0.04, relative risk (RR) = 1.67,95% CI 1.02–2.73). Treatment with a glycopeptidewas started for 36 (92.3%) of 39 patients withmultiple positive blood cultures, and for 59(80.8%) of 73 patients with a single positive bloodculture (p 0.1). Apyrexia following the adminis-tration of a glycopeptide was observed more oftenin patients with multiple positive blood culturesthan in patients with a single positive bloodculture (22 ⁄ 36 (61.1%) vs. 24 ⁄ 59 (40.6%); p 0.05,

Table 2. Characteristics of 26 allogeneic and the 53autologous bone marrow transplant recipients

Characteristic

Allogeneic

bone marrowtransplantation

patients

Autologous

bone marrowtransplantation

patients

Male gender 14 (54%) 37 (70%)Age at transplantation, median years(range)

37 (16–56) 54 (27–68)

Underlying diseaseMyeloma 1 (4%) 20 (38%)Lymphoma 1 (4%) 17 (32%)Hodgkin’s disease 1 (2%)Acute leukaemia 17 (65%) 5 (9%)Chronic leukaemia 5 (19%) 7 (13%)Other 2 (8%) 3 (6%)

Conditioning regimenMelphalan, 200 mg ⁄m2 16 (30%)Cyclophosphamide and TBI 15 (58%) 10 (19%)VP16, cyclophosphamide and TBI 3 (11.5%)Melphalan, 140 mg ⁄m2, and TBI 8 (15%)BEAM 8 (15%)Melphalan ⁄ busulfan 2 (4%)CBV and TBI 4 (15%) 2 (4%)Cyclophosphamide and busulfan 1 (4%) 2 (4%)Other 3 (11.5%) 5 (9%)

BEAM, carmustine, etoposide, aracytine and melphalan; CBV, cyclophosphamide,carmustine and etoposide; TBI, total body irradiation.

Table 3. Data collected for 112 episodes of Staphylococcusepidermidis bacteraemia

Characteristic n %

Prophylaxis 61 54.5Fever: median (range) 39 �C (37–40.5 �C)Shivers 38 34Hypotension 4 3.5Septic shock 2 2Increased serum creatinin level 18 16Localised infection 36 32

Catheter 11Lung 11Skin 8Other 6

Inflammatory aspect of the catheter 40 36Purulent aspect of the catheter 7 6Length of hospitalisation: median (range) 21 days (0–98 days)Duration between catheter insertion and

bacteraemia: median (range)22 days (1–99 days)

Administration of glycopeptide 94 85Apyrexia following glycopeptide administration 32a 34Administration of amphotericin B 59 53

aTotal n = 94.

Table 4. Other microorganisms isolated concomitantlywith Staphylococcus epidermidis

Microorganism

Single positive blood

culture (n = 14)

Multiple positive

blood cultures (n = 6)

Staphylococcus haemolyticus 3 1Pseudomonas aeruginosa 2Viridans streptococci 2 3Enterobacteriaceae 2Enterococcus faecalis 1Candida albicans 1Candida parapsilosis 1Corynebacterium spp. 1Corynebacterium spp.

plus S. haemolyticus1

Viridans streptococciplus Enterobacteriaceae

1

Viridans streptococciplus Staphylococcus hominis

1

448 Clinical Microbiology and Infection, Volume 12 Number 5, May 2006


RR = 1.67, 95% CI 0.98–2.86). The results obtainedwere similar when bacteraemias in which anothermicroorganism was associated with S. epidermidiswere excluded.

PFGE

In total, 23 (29%) of 79 patients had more than oneepisode of S. epidermidis bacteraemia. Of these,PFGE demonstrated that episodes in 11 patientswere caused by unrelated strains (two episodes ineight patients; three episodes in three patients),and in nine patients by related (‘persistent’)strains (two episodes in six patients; three epi-sodes in two patients; four episodes in onepatient), and in three patients by both relatedand unrelated strains.

Thirty-four (30.3%) isolates belonged to sevenclusters: cluster 1 included 11 related (> 85%similarity) isolates from 1993 to 1996; cluster 2included 11 related isolates from 1993 to 1995;clusters 3 and 4 each included three relatedisolates; and clusters 5, 6 and 7 each includedtwo related isolates.

PCR for ica operon

All 109 isolates analysed were positive for 16SrRNA, with 91 (83.5%) also carrying the entire icaoperon (icaA, icaB, icaC and icaD genes). Detectionof the ica operon was not associated with anyspecific clinical or biological features, and did notdiffer significantly (p 0.8) among isolates fromepisodes involving single (60 ⁄ 71; 84.5%) or mul-tiple (31 ⁄ 38; 81.6%) blood cultures. The medianinterval between catheter insertion and occur-rence of bacteraemia was 25 days (range 1–99days) for the isolates carrying the ica operon and15 days (range 7–47 days) for the ica-negativeisolates (p 0.07). The results did not differ signi-ficantly when the analysis was restricted topatients who had another microorganism isolatedconcomitantly from blood.

Biofilm production

In total, 109 isolates were analysed at least threetimes. Biofilm was produced by 59 (54.1%) isolates,was not produced by 39 (35.8%) isolates, and majordiscrepancies were found with 11 (10.1%) isolates(strong positive vs. negative), which were excludedsubsequently from the study.

With the exception of serum creatinine level,biofilm production was not associated signifi-cantly with any of the clinical features analysed orwith the number of positive blood cultures (63%of isolates from a single positive blood cultureproduced biofilm, compared with 54.5% of iso-lates from multiple positive blood cultures; p 0.4).The serum creatinine level increased significantlymore frequently during bacteraemias caused bybiofilm-producing isolates than during bacterae-mias caused by non-biofilm-producing isolates(14 ⁄ 59 vs. 3 ⁄ 39; p 0.03, RR = 1.50, 95% CI 1.11–2.01). The median interval between catheter inser-tion and onset of bacteraemia was significantlyshorter (p 0.03) for bacteraemias caused bybiofilm-producing isolates (18 days) than forbacteraemias caused by non-biofilm-producingisolates (37 days).

Relationship between ica operon carriage andbiofilm production

After excluding the isolates with discrepantmeasures of biofilm production, no significantassociation was observed between ica operoncarriage and production of biofilm: 49 (60.5%) ofthe 81 ica-positive isolates produced biofilm and32 (39.5%) did not, compared with ten (58.8%) ofthe 17 ica-negative strains that produced biofilmand seven (41.2%) that did not (p 0.8). Nodifference was found in clinical features, numberof blood cultures or outcome following glycopep-tide administration between biofilm-producingica-positive (n = 49) and non-producing ica-posit-ive (n = 32) isolates. A significant increase inserum creatinine level was confirmed in associ-ation with ica-positive biofilm-producing isolates(13 ⁄ 49 (26.5%) vs. 2 ⁄ 32 (6%); p 0.02, RR = 1.59,95% CI 1.18–2.14). The median interval betweencatheter insertion and onset of bacteraemia wassignificantly shorter (p 0.02) for bacteraemiascaused by biofilm-producing ica-positive isolates(22 days) than for bacteraemias caused bynon-producing ica-positive isolates (40 days).

Relationship between PFGE profiles, biofilmproduction and ica operon carriage

Of the 28 ‘persistent’ isolates from patients pre-senting with more than one episode of bacterae-mia, 26 ⁄ 26 carried the ica operon, while theremaining two isolates were not available for



analysis. In contrast, of the 83 unrelated isolates,65 (78.3%) carried the ica operon (p 0.004, RR =1.28, 95% CI 1.14–1.23). No difference in biofilmproduction was observed between these groupsof isolates. The ica operon was found morefrequently in clustered isolates than in unclus-tered isolates (31 ⁄ 33 (93.9%) vs. 60 ⁄ 76 (78.9%); p0.05, RR = 1.28, 95% CI 1.14–1.43). Again, nodifference in biofilm production was observedbetween these groups of isolates.

DISCUSSION

Despite their supposed low virulence, CNS, andespecially S. epidermidis, are now the most com-mon organisms isolated from infections ofimplanted medical devices [1,16]. The immunestatus of the patients and the capacity of thebacteria to adhere to foreign bodies are the mainfactors involved in the pathogenesis of S. epider-midis. Despite attempts to determine a referencestandard for assessing the pathogenicity of thesestrains by a combination of clinical and microbio-logical findings, these criteria remain difficult toapply in clinical practice, especially in haemato-logical and neonatal intensive care units [2].Therefore, the present study aimed to determinewhether biofilm production and ⁄ or detection ofthe ica operon could be used to assess thevirulence of S. epidermidis strains isolated fromblood cultures in BMT recipients, and thus helpphysicians in deciding whether to prescribeglycopeptides. Unfortunately, the present resultssuggest that these indicators cannot be used forthat purpose. Nevertheless, there are some inter-esting findings.

First, there were some clinical differencesbetween patients presenting with single or mul-tiple positive blood cultures. In the present study,a number of S. epidermidis bacteraemias wereassociated with only one positive blood culture,and > 80% of these were treated with glycopep-tides. Only one clinical sign, the presence ofshivers, was significantly more common inpatients with multiple positive blood cultures.

Second, although biofilm production per secannot be used to decide on the need to treat S.epidermidis bacteraemias, differences were foundbetween patients infected by isolates eitherproducing or failing to produce biofilm. Severalstudies have attempted to evaluate the role ofbiofilm production in virulence, but the results

have been somewhat contradictory [17–20]. How-ever, recent clinical studies have suggested thatPIA is important for biofilm accumulation, and thatthe presence of the icaADBC operon coding for PIAis related to the clinical significance of S. epidermidisisolates [4–8,21–23]. In the present study, biofilmproduction was associated with a shorter intervalbetween catheter insertion and the onset of bacter-aemia, suggesting that biofilm-producing strainscolonise catheters more rapidly than non-produc-ers. Biofilm production was also associated with anincrease in creatinine serum level.

Third, although detection of the ica operoncannot be used to assess the virulence of S.epidermidis isolates, the presence of this operonis associated with important epidemiologicalcharacteristics. As the present study involvedpatients with central venous catheters, the pres-ence of the ica operon in 83% of the isolatesanalysed was consistent with published datashowing that the ica operon is found morefrequently in catheter-colonising isolates [22–24].These data indicate that strains carrying the icaoperon have a competitive advantage for colo-nising indwelling medical devices, especiallywhen the duration of catheterisation is long. Inthe present study, the mean interval betweencatheter insertion and the onset of bacteraemiawas 22 days, and the duration of catheterisationwas longer in patients with bacteraemias invol-ving ica-positive isolates (25 days) than in thosewith isolates without the ica operon (15 days).This long duration of catheterisation could havefavoured the selection of ica-positive strainsfrom among ica-negative strains present on theskin.

The present study found that 32 of 91 isolatescarrying the ica operon did not produce biofilm.Regulation of the ica operon appears to be verycomplex. Production of PIA is certainly subjectedto on-off switching, and may be involved in S.epidermidis phase-variation that might improvebacterial survival and growth under changingenvironmental conditions in vivo [25]. Environ-mental regulation may play an important role inbiomaterial-related disease [8,22,23,26–31].

The present study also showed that, despitewide dispersion of the S. epidermidis isolates, somestrains could persist in the BMT unit. Persistenceof S. epidermidis clones for a long period has beendemonstrated previously in neonatal, haemato-logical and cardiac surgery units [32–35].



However, the pathogenic potential of clonal S.epidermidis strains remains unclear. In the presentstudy, clustered isolates were neither related tothe severity of the clinical features nor associatedwith catheter-related infection, but carried the icaoperon more frequently than sporadic isolates.The presence of the ica operon may be evidence forthe persistence of a particular S. epidermidis strainin a unit [22,27]. This hypothesis was reinforced bythe fact that the time between insertion of acatheter and the onset of bacteraemia was longerin bacteraemia occurring with clustered isolatesand with ica-positive isolates, and by the frequentpresence of the ica operon in persistent strainsisolated from the blood of patients presentingmultiple episodes of bacteraemia.

Despite the limitations associated with retro-spective studies, S. epidermidis bacteraemia asso-ciated with a single positive blood culture wasshown to be a very common event in BMTrecipients. The presence of the ica operon wasvery common among S. epidermidis isolates frompatients with multiple or single positive bloodcultures. The presence of the ica operon does notseem to be associated with specific clinical orbiological features in BMT recipients, but could bean indicator of strain persistence.

ACKNOWLEDGEMENTS

The authors acknowledge financial support from the FrenchMinistry of Health (PHRC No 329799), and also thank N. Melatand the team in the Medical Archives for their assistance incollecting the clinical data.

REFERENCES

1. Rupp ME, Archer GL. Coagulase negative staphylococci:pathogens associated with medical progress. Clin Infect Dis1994; 19: 231–245.

2. Herwaldt LA, Geiss M, Kao C, Pfaller MA. The positivepredictive value of isolating coagulase-negative staphylo-cocci from blood cultures. Clin Infect Dis 1996; 22: 14–20.

3. Peacock SJ, Bowler ICJW, Crook DWM. Positive predictivevalue of blood cultures growing coagulase-negative sta-phylococci. Lancet 1995; 346: 191–192.

4. Heilmann C, Schweitzer O, Gerke C, Vanittanakom N,Mack D, Gotz F. Molecular basis of intercellular adhesionin the biofilm-forming Staphylococcus epidermidis. MolMicrobiol 1996; 20: 1083–1091.

5. Mack D. Molecular mechanisms of Staphylococcus epider-midis biofilm formation. J Hosp Infect 1999; 43 (suppl): 113–125.

6. Arciola CR, Baldassarri L, Montanaro L. Presence of icaAand icaD genes and slime production in a collection of

staphylococcal strains from catheter-associated infections.J Clin Microbiol 2001; 39: 2151–2156.

7. Barbier Frebourg N, Lefebvre S, Baert S, Lemeland JF.PCR-based assay for discrimination between invasive andcontaminating Staphylococcus epidermidis strains. J ClinMicrobiol 2000; 38: 877–880.

8. Ziebuhr W, Heilmann C, Gotz F et al. Detection of theintercellular adhesion gene cluster (ica) and phase vari-ation in Staphylococcus epidermidis blood culture strains andmucosal isolates. Infect Immun 1997; 65: 890–896.

9. Christensen GD, Simpson WA, Younger JJ et al. Adherenceof coagulase negative staphylococci to plastic tissue cul-ture plates: a quantitative model for the adherence ofstaphylococci to medical devices. J Clin Microbiol 1985; 22:996–1006.

10. Cramton SE, Gerke C, Schnell NF, Nichols WW, Gotz F.The intercellular adhesion (ica) locus is present in Staphy-lococcus aureus and is required for biofilm formation. InfectImmun 1999; 67: 5427–5433.

11. Bannerman TL, Hancock GA, Tenover FC, Miller M.Pulsed-field gel electrophoresis as a replacement for bac-teriophage typing of Staphylococcus aureus. J Clin Microbiol1995; 33: 551–555.

12. Tenover FC, Arbeit RD, Goering RV et al. Interpretingchromosomal DNA restriction patterns produced bypulsed-field gel electrophoresis: criteria for bacterial straintyping. J Clin Microbiol 1995; 33: 2233–2239.

13. Martineau F, Picard FJ, Roy PH, Ouellette M, BergeronMG. Species-specific and ubiquitous DNA-based assaysfor rapid identification of Staphylococcus epidermidis. J ClinMicrobiol 1996; 34: 2888–2893.

14. Deighton MA, Balkau B. Adherence measured bymicrotiter assay as a virulence marker for Staphylococcusepidermidis infections. J Clin Microbiol 1990; 28: 2442–2447.

15. Pfaller M, Davenport D, Bale M, Barrett M, Koontz F,Massanari RM. Development of the quantitative micro-testfor slime production by coagulase-negative staphylococci.Eur J Clin Microbiol Infect Dis 1988; 7: 30–33.

16. Thylefors JD, Harbarth S, Pittet D. Increasing bacteremiadue to coagulase negative staphylococci: fiction or reality?Infect Control Hosp Epidemiol 1998; 19: 581–589.

17. Christensen GD, Baldassarri L, Simpson WA. Colonizationof medical devices by coagulase negative staphylococci. In:Bisno AL, Waldvogel FA, eds, Infections associated withindwelling medical devices, 2nd edn. Washington DC:American Society for Microbiology, 1994; 45–78.

18. Cree RGA, Phillips I, Noble WC. Adherence characteristicsof coagulase negative staphylococci isolated from patientswith infective endocarditis. J Med Microbiol 1995; 43: 161–168.

19. Deighton MA, Borland R, Capstick JA. Virulence of Sta-phylococcus epidermidis in a mouse model: significance ofextracellular slime. Epidemiol Infect 1996; 117: 267–280.

20. Souto MJ, Ferreiros CM, Criado MT. Failure of phenotypiccharacteristics to distinguish between carrier and invasiveisolates of Staphylococcus epidermidis. J Hosp Infect 1991; 17:107–115.

21. Arciola CR, Campoccia D, Gamberini S, Donati ME,Baldassarri L, Montanaro L. Occurrence of ica genes forslime synthesis in a collection of Staphylococcus epidermidisstrains from orthopedic prosthesis infections. Acta OrthopScand 2003; 74: 617–621.



22. Fitzpatrick F, Humphreys H, Smyth E, Kennedy CA,O’Gara JP. Environmental regulation of biofilm formationin intensive care unit isolates of Staphylococcus epidermidis.J Hosp Infect 2002; 42: 212–218.

23. de Silva GDI, Kantzanou M, Justice A et al. The ica operonand biofilm production in coagulase-negative sta-phylococcci associated with carriage and disease in aneoanatal intensive care unit. J Clin Microbiol 2002; 40: 382–388.

24. Vandecasteele SJ, Peetermans WER, Merckx Rijnders BJ,Van Eldere J. Reliability of the ica, aap and atlE genes in thediscrimination between invasive, colonizing and contam-inant Staphylococcus epidermidis isolates in the diagnosis ofcatheter-related infections. Clin Microbiol Infect 2003; 9:114–119.

25. Ziebuhr W, Krimmer V, Rachid S, LoBner I, Gotz F,Hacker J. A novel mechanism of phase variation of viru-lence in Staphylococcus epidermidis: evidence for control ofthe polysaccharide intercellular adhesin synthesis byalternating insertion and excision of the insertion sequenceelement IS256. Mol Microbiol 1999; 32: 345–356.

26. O’Gara JP, Humphreys H. Staphylococcus epidermidis bio-films: importance and implications. J Med Microbiol 2001;50: 582–587.

27. Cho SH, Naber K, Hacjer J, Ziebuhr W. Detection of theicaADBC gene cluster and biofilm formation in Staphylo-coccus epidermidis isolates from catheter-related urinarytract infections. Int J Antimicrob Agents 2002; 19: 570–575.

28. Dobinsky S, Bartscht K, Mack D. Influence of Tn917insertion on transcription of the icaADBC operon in sixbiofilm-negative transposon mutants of Staphylococcusepidermidis. Plasmid 2002; 47: 10–17.

29. Dobinsky S, Kiel K, Rohde H et al. Glucose-relateddissociation between icaADBC transcription and biofilmexpression by Staphylococcus epidermidis: evidence for anadditional factor required for polysaccharide intercellularadhesin synthesis. J Bacteriol 2003; 185: 2879–2886.

30. Handke LD, Conlon KM, Slater SR et al. Genetic andphenotypic analysis of biofilm phenotypic variation inmultiple Staphylococcus epidermidis isolates. J Med Microbiol2004; 53: 367–374.

31. Conlon KM, Humphreys H, O’Gara JP. icaR encodes atranscriptional repressor involved in environmentalregulation of ica operon expression and biofilm formationin Staphylococcus epidermidis. J Bacteriol 2002; 184: 4400–4408.

32. Huebner J, Pier GB, Maslow JN et al. Endemic nosocomialtransmission of Staphylococcus epidermidis bacteremia iso-lates in a neonatal intensive care unit over 10 years. J InfectDis 1994; 169: 526–531.

33. Nouwen JL, van Belkum A, de Marie S et al. Clonalexpansion of Staphylococcus epidermidis strains causingHickman catheter-related infections in a haemato-onco-logic department. J Clin Microbiol 1998; 36: 2696–2702.

34. van Belkum A, Kluijmans J, van Leeuwen W et al.Monitoring persistence of coagulase negative staphylo-cocci in a hematology department using phenotypic andgenotypic strategies. Infect Control Hosp Epidemiol 1996;7: 660–667.

35. van Belkum A, Kluijtmans J, van Leeuwen W, GoessensW, ter Averst E, Verbrugh H. Investigation into therepeated recovery of coagulase-negative staphylococcifrom blood taken at the end of cardiopulmonary by-pass.J Hosp Infect 1995; 31: 285–293.



ica and biofilm

Documents