The observation of natural habitats has shown that bacteria generally aggregate in biofilm structures to persist [1]. Biofilm is an association of microbial cells which are surrounded by a matrix of polysaccharide material; this structure is an optimal environment for genetic material exchange between the different microorganisms [2]. Biofilm formation has been linked to the survival of pathogenic bacteria in the hospital environment and it has been connected to infections associated with indwelling medical devices. Indeed, biofilm bacterial communities confer a protection from environmental hazards [1, 3]. The surface colonization can take place either at the solid-liquid interface (SLI-biofilm) or at the air-liquid interface (ALI-biofilm) where it forms a pellicle on the top of the liquid media as observed in other microorganisms such as Pseudomonas aeruginosa or Salmonella spp. [4, 5].

The members of the genus Acinetobacter are ubiquitous Gram-negative cocobacilli that are frequently found in the environment but also in the hospital setting where they have been associated with outbreaks of nosocomial infections [6]. The so-called Acinetobacter calcoaceticus - Acinetobacter baumannii (ACB) complex contains phenotypically closely related species, i.e. the clinically most important species, A. baumannii, Acinetobacter Genospecies 3 and Acinetobacter Genospecies 13TU, also referred to as the A. baumanni group, and the environmental species A. calcoaceticus[7]. Although these species are very difficult to differentiate in the laboratory, their importance in the clinical environment is clearly different: A. baumannii and Acinetobacter Genospecies 13TU are responsible for most of the infections while Acinetobacter Genospecies 3 is less often associated with disease. On the other hand, A. calcoaceticus is mainly an environmental microorganism rarely involved in human infections [7]. Among them, A. baumannii has emerged as the most common pathogenic species involved in hospital-acquired infections [6, 8–10]; this multiresistant opportunistic pathogen can survive on nutrient-limited surfaces for several days, even in dry conditions and in the harsh hospital environment [11]. One reason for this emergence may be its persistence in the hospital wards, in particular in the intensive care unit; this persistence could be partially explained by the capacity of these microorganisms to form biofilm. Therefore, our main objective was to study the prevalence of the two main types of biofilm formed by the most relevant Acinetobacter species. Overall, this study has shown the capacity of the members of the ACB-complex to form two different types of biofilm (SLI and ALI), and it revealed that biofilm formation increased at 25°C, a condition that might contribute to their persistence in the hospital environment.

ALI-Biofilm formation

ALI-biofilm formation was performed in 5 ml polystyrene tubes with a diameter of 12.8 mm; biofilm formation was determined in Mueller Hinton broth (Oxoid, France) using an initial OD₆₀₀ of 0.01 and incubated at 25°C or 37°C for 72 h without shaking. Positive ALI-biofilm samples were identified visually (Figure 1); the isolates were considered positive when a pellicle was covering the whole liquid surface.

As shown in Table 1 SLI-biofilm was more frequently produced at both 25°C and 37°C in Acinetobacter Genospecies 3, Acinetobacter Genospecies 13TU, and A. baumannii than in A. johnsonii, A. lwoffii, and A. radioresistens (p < 0.05 for all comparisons). Thus, these results might explain, at least in part, the marked persistence of the three former species in hospitals and their involvement in nosocomial infections. Moreover, in a previous study (using the same collection of A. baumannii clinical isolates), Rodriguez-Baño et al.[16] showed that 63% of A. baumannii clinical isolates formed SLI-biofilm at 37°C. Based on these results, our study revealed that the rates of SLI-biofilm formation obtained for Acinetobacter Genospecies 13TU and A. baumannii were similar, as well as their variation related to the temperature (higher at 25°C than at 37°C). Likewise, ALI-biofilm was significantly more frequent at both 25°C and 37°C in Acinetobacter Genospecies 13TU and A. baumannii than in A. johnsonii, A. lwoffii, A. radioresistens, and Acinetobacter Genospecies 3 (p < 0.05). Overall, this similar behaviour in biofilm formation between A. baumannii and Acinetobacter G13TU is coherent with the fact that these two species are the most commonly found in the hospital.

Our results showed (see Table 1) that ALI-biofilm was mainly formed by the members of the ACB-complex. Although A. junii had some capacity to form this characteristic biofilm, the number of isolates studied was too small to draw conclusions. By contrast, as a member of the ACB-complex, A. calcoaceticus has shown a different pattern at 37°C with a complete absence of ALI-biofilm formation together with a reduced ability to form SLI-biofilm, both of which could be explained by the fact that this species' natural habitat is the environment where lower temperatures usually prevail.

Finally, as biofilm formation generally predominated at 25°C rather than 37°C, this mechanism could explain the observed persistence of the members of the A. baumannii group in the inanimate hospital environment. Previous studies have already reported that the members of the A. baumannii group and especially A. baumannii survive desiccation better than other Acinetobacter spp., comparing the survival rate of A. baumannii to those obtained for Staphylococcus aureus[7, 17]. In addition, in a recent study, Wisplinghoff et al showed that all disinfectants tested inhibited the growth of A. baumannii[18] which suggests that these microorganisms in their planktonic state are susceptible to most disinfectants. As already demonstrated for other bacterial species, extracellular polymeric substances from the biofilm matrix play an important role in the resistance and tolerance to dehydration [19], suggesting that biofilm formation contributes to the ability of A. baumannii, and possibly the other members of the A. baumannii group, to survive better in the hospital environment and to resist the action of disinfectants.

For A. baumannii, SLI-biofilm has already been described in several reports [20–22] and it has been linked to some device-associated infections [16]. Indeed, de Breij et al[21] have recently compared SLI-biofilm formation by the members of the ACB complex and concluded that there was no difference between clinically relevant and less-relevant Acinetobacter strains and species; in the same way, no temperature related differences were found for biofilm formation. Our results slightly differ from theirs possibly due to the higher number of strains and species analysed. They suggested a reduced biofilm formation for the Acinetobacter Genospecies 13TU although only 3 isolates had been studied; by contrast, after studying 60 isolates, our results clearly show that the behaviour for biofilm formation of this species is highly related to A. baumannii. On the other hand, although we have only studied 10 A. calcoaceticus isolates, the results also indicate an association between biofilm and temperature in this non-pathogenic species, a tendency that can be observed in most of the analysed species. The comparison with other non-pathogenic species outside the ACB-complex has shown a relevant difference for biofilm formation that was even increased when looking at the ALI-biofilm formation, an ability that to our knowledge has never been described in clinical isolates of the Acinetobacter spp.

In summary, the members of the A. baumannii group have a higher ability to form SLI-and ALI-biofilm than other less clinically related species. Nevertheless, A. calcoaceticus, the environmental representative of the ACB-complex, showed a reduced biofilm formation at 37°C when compared to the other members of this complex. This feature could be connected to the higher colonization rate of patients by pathogenic Acinetobacter species (mainly A. baumannii and Acinetobacter Genospecie 13TU), and probably contributing to the increased risk of clinical infection.

Due to the observational and retrospective design of the study, the Ethic Committee of the participating centres waived the need for obtaining written informed consent.