OUP user menu

Biofilm production by clinical strains of Acinetobacter baumannii isolated frompatients hospitalized in two tertiary care hospitals

Marta M. Wroblewska, Anna Sawicka-Grzelak, Halina Marchel, Miroslaw Luczak, Alex Sivan
DOI: http://dx.doi.org/10.1111/j.1574-695X.2008.00403.x 140-144 First published online: 1 June 2008


Microbial biofilms are considered as virulence factors. During the present study, 34 clinical strains of Acinetobacter baumannii, isolated from patients hospitalized in two tertiary care hospitals, were examined for biofilm formation. These strains showed high variability in biofilm formation. Furthermore, no relation could be found between the ability of biofilm production and molecular type, carbapenem resistance, site of isolation of the clinical strains of A. baumannii and disease severity. Interestingly, in two cases an increase in biofilm formation could be detected in A. baumannii isolates cultured from the same patient upon prolonged hospitalization.

  • biofilm
  • Acinetobacter baumannii
  • tertiary care

Strains of Acinetobacter spp., mainly Acinetobacter baumannii, are very important nosocomial pathogens, contributing significantly to morbidity and mortality of patients, particularly hospitalized in intensive care units (Bergogne-Bérézin & Towner, 1996). Moreover, recent emergence of carbapenem resistance among these isolates further stresses their importance in etiology of hospital-acquired infections. Infections of hospitalized patients with Acinetobacter spp., often preceded by colonization, are frequently associated with invasive procedures and implantable medical devices (Bergogne-Bérézin & Towner, 1996). The ability of a strain to form a biofilm may be a significant factor facilitating this process. However, there are only scarse reports on biofilm formation by clinical strains of A. baumannii isolated from hospitalized patients and the numbers of tested isolates were <20 (Inglis et al., 1995; Vidal et al., 1996; Can et al., 2005; de Breij et al., 2006; Lubeck & Gerischer, 2006). The study was undertaken to determine whether the capacity of clinical strains of A. baumannii, to produce a biofilm coincide with genotype of the bacterial strain, the site of isolation, resistance to carbapenems and duration of hospitalization.

Clinical strains of A. baumannii, isolated from patients hospitalized in two tertiary care hospitals in Warsaw were examined for biofilm formation. Twenty-four strains were isolated from patients hospitalized in a tertiary care hospital (Hospital A, 700 beds), specializing in care of surgical and injured patients. Ten strains were cultured from patients in a university-affiliated hospital (Hospital B, 1200 beds), comprising highly specialized internal medicine and surgical wards. Available medical charts of the patients were reviewed.

Bacteria were cultured according to standard bacteriological techniques and with an automated BacT/Alert system (bioMérieux, Marcy l'Etoile, France). Bacterial isolates were initially identified by API and VITEK tests (bioMérieux) and further confirmed as A. baumannii by tRNA spacer fingerprinting (Ehrenstein et al., 1996; Wroblewska et al., 2007). Susceptibility of the isolates to antibacterial agents was initially tested by a disk-diffusion method according to CLSI (formerly NCCLS) recommendations and with the VITEK system (bioMérieux). For carbapenems, the diameters of the inhibition zones were interpreted as: ≤13 mm, resistant; 14–15 mm, intermediate; and ≥16 mm, susceptible. Resistance of A. baumannii isolates to imipenem and/or meropenem was verified by determination of minimal inhibitory concentrations (MICs) with Etests (AB Biodisk, Solna, Sweden), using interpretation breakpoints of ≤4 mg L−1 as susceptible, 8 mg L−1 as intermediately-susceptible, and ≥16 mg L−1 as resistant to carbapenems.

The isolates were typed retrospectively by molecular techniques, initially using DAF4 randomly amplified polymorphic DNA (RAPD) analysis, as described by Grundmann (1997), and by pulsed-field gel electrophoresis (PFGE) of ApaI-digested total chromosomal DNA, prepared as described previously (Seifert et al., 2005), in a CHEF DRII apparatus (Bio-Rad, Hemel Hempstead, UK) at 200 V for 20 h at 14 °C, with a ramped pulse of 5–13 s. Type strains of European clones I, II and III were kindly supplied by Dr Lenie Dijkshoorn (LUMC, Leiden, the Netherlands).

Quantitative estimation of bacterial biomass in biofilms, formed by A. baumannii strains, was assessed in 96-well polystyrene microtiter plates, using the crystal-violet staining method (O'Toole et al., 2000). A reference biofilm-producing strain of Rhodococcus ruber was used in the assays (data not shown).

Two groups of 24 and 10 randomly selected strains of A. baumannii from hospitals A and B were tested. They originated from various clinical specimens, as shown in Table 1. Analysis of the available medical charts of 12 patients from hospital A showed that the isolates were cultured on average on day 14 of hospitalization (range: 2–46). In the case of isolates from hospital B this value was 16 days (range: 8–30 days).

View this table:
Table 1

Molecular type, carbapenem susceptibility and site of isolation of Acinetobacter baumannii strains isolated from patients hospitalized in Hospital A and Hospital B

Carbapenem susceptibility
HospitalBiofilm formationStrain no.Molecular typeImipenemMeropenemSite of isolation
Hospital AHigh843SSUrine
1082SSSurgical wound
1242SSVaginal swab
100SporadicSSWound swab
1103-RelatedSSIntubation tube
1163-RelatedSSChronic wound
1172SIGranulation tissues
1251SSBronchial aspirate
93SporadicSSEar swab
1073-RelatedSSChronic wound
118SporadicSSSurgical wound
1202SSBronchial aspirate
121SporadicSSBronchial aspirate
1222SSTracheal aspirate
1231SSWound swab
129SporadicSSWound swab
Hospital BHigh81RRWound swab
Medium61RRCatheter tip
92RRCatheter tip
102RRBiliary tract pus
112IIWound swab
Low71SRBronchial aspirate
122-LikeRRCatheter tip
  • S, susceptibility; I, intermediate susceptibility; R, resistance.

Among the isolates of A. baumannii cultured from patients in Hospital A all strains were fully susceptible to imipenem and all but one to meropenem (strain A-117 was characterized as intermediately susceptible) (Table 1). In contrast to this, strains isolated from patients in hospital B showed variable patterns of susceptibility to carbapenems. Molecular typing revealed cocirculation of at least three types of A. baumannii strains in hospital A and hospital B (referred to as types 1, 2 and 3).

Analysis of biofilm formation by the studied clinical strains of A. baumannii revealed three groups of strains, regarding their ability to produce a biofilm (Fig. 1). The ‘high producer group’ was determined empirically as OD ≥2.0 after 6 days of incubation and comprised the following isolates: A–84, –108, –124 and B–8. The ‘medium group’ showed the OD ranging from 1.0 to 1.9 at day 6 of the experiment: A–89, –100, –102, –103, –110, –116, –117, –125 and B–6, –9, –10, –11, –13, –15. The ‘low group’ was characterized by the OD ≤0.9 at day 6 of the experiment and contained the strains: A–82, –93, –107, –118, –119, –120, –121, –122, –123, –126, –127, –129, –130 and B–7, –12, –14. The data in Fig. 1 represent the mean value of 4 separate biofilm cultures (replicates). Each of these replicates comprised eight subreplicates (i.e. eight wells in the microtiter plate).

Figure 1

Biofilm formation by selected, representative clinical strains of Acinetobacter baumannii isolated from hospitalized patients, expressed as OD±SD: (a), the ‘high-producers’ group; (b), the ‘medium-producers’ group; (c), the ‘low-producers’ group.

There was no correlation between the ability of biofilm formation and molecular type, carbapenem resistance or site of isolation of the clinical strains of A. baumannii, cultured from patients hospitalized in two tertiary care hospitals (Table 1). Similarly, there was no relation of biofilm production to disease severity. However, strains A–107 and A–116 were cultured (as mentioned earlier) from the same patient, but 1 month apart, from a chronic, nonhealing wound. An increase in the ability of biofilm formation was observed with time. Similarly, strains B–12 and B–11, which were isolated from the same patient 5 days apart, showed an increase in biofilm production with time.

The present study aimed at revealing whether biofilm formation by clinical strains of A. baumannii coincides with bacterial genotype, antibiotic resistance and with the isolation site. Although all bacterial strains represented a single species (A. baumannii), they exhibited great variability in biofilm formation, which could be classified into high-, medium- and low-producers. These findings were in agreement with other reports, however based on only several clinical strains (de Breij et al., 2006; Lubeck & Gerischer, 2006). A study by de Breij (2006) found a wide variation in biofilm formation and adherence of six analysed Acinetobacter strains, with no correlation of biofilm formation, adherence and presence of csuE or AcuA genes. In our study 12% of A. baumannii strains represented high-producers, 41%– medium-producers and 47%– low-producers of biofilm.

Similarly to reports on biofilms formed by other genera of bacteria, different morphotypes of Acinetobacter cells represented specialized adaptations to adherence and colonization in the nutrient-rich (bacillary forms — poorly adhesive) or nutrient-deficient system (coccoid forms — strongly adhesive) (James et al., 1995). In another report a correlation was found between gene PER-1 and cell adhesion of Acinetobacter spp. strains (Sechi et al., 2004). Furthermore, biofilm formation by Acinetobacter spp. may be influenced by its interaction within a biofilm with other bacterial species (Moller et al., 1998). A gene cup (chaperone-usher pathways) has been reported in Pseudomonas aeruginosa as an important factor in biofilm formation (Vallet et al., 2004). Interestingly, a similar gene has also been described in A. baumannii as being important for pili formation and adherence to plastic surfaces, with subsequent formation of a biofilm (Tomaras et al., 2003). In another study on biofilms formed by a tested strain of Acinetobacter spp., natural genetic transformation of bacteria remained effective (Hendrickx et al., 2003).

Genotype analysis of A. baumannii isolates from both hospitals revealed that molecular type 1 was identical to the European clone I, while type 3 — to the European clone II. Molecular type 2 was not related to any of the European clones known to date. In our study no correlation could be found between the ability of biofilm formation and molecular type, carbapenem resistance or site of isolation of the clinical strains of A. baumannii, cultured from hospitalized patients. However, no such data could be found in the literature for comparison.

Biofilm formation by Acinetobacter spp. strains has been observed on tracheal tubes upon prolonged mechanical ventilation (Inglis et al., 1995). Interestingly, in our study in two cases (one patient from hospital A and one patient from hospital B) an increase in biofilm production has been detected in A. baumannii isolates, cultured from the same patient within 1 month or 5 days, respectively. These findings suggest, that ability of a strain to form a biofilm may be time-dependent and require further studies. Indeed, in reports analysing biofilm dynamics it was shown, that up to 48 h predominated ‘sessile cells’. After 48 h in the biofilm appeared amorphous exopolysaccharide-like substance (Vidal et al., 1996). In a report by Lubeck & Gerischer (2006) biofilm formation by A. baumannii strains on polystyrene microtiter plates was tested over 24 h only, showing time-dependancy over a period of 8 h. At 24-h a level of biofilm formation was increased in the case of one strain and decreased in relation to two other tested strains (Lubeck & Gerischer, 2006). In our study, in the majority of analysed strains the highest level of biofilm formation was observed within 24–48 h of incubation (Fig. 1). However, so called ‘high-producers’ showed a time-dependent increase in biofilm formation upon prolonged incubation for up to 6 days (Fig. 1a).

More research is needed on possible links between biofilm formation and patient's infection, the role of biofilm in antimicrobial drug resistance as well as methods for control of biofilm-related infections. A better understanding of biofilm formation by Acinetobacter and genetic basis for control of this process are required to develop novel strategies for dealing with infections caused by these opportunistic and often multi-drug resistant nosocomial pathogens.


We thank Dr Kevin Towner (Department of Microbiology, University Hospital, Queen's Medical Centre, Nottingham, UK) for his help in molecular typing of the strains and invaluable suggestions regarding the manuscript.


  • Editor: Alex van Belkum


View Abstract