OUP user menu

Pseudomonas aeruginosa rhamnolipids disperse Bordetella bronchiseptica biofilms

Yasuhiko Irie, George A. O'Toole, Ming H. Yuk
DOI: http://dx.doi.org/10.1016/j.femsle.2005.07.012 237-243 First published online: 1 September 2005


We have previously reported that the respiratory pathogen Bordetella bronchiseptica can form biofilms in vitro. In this report, we demonstrate the disruption of B. bronchiseptica biofilms by rhamnolipids secreted from Pseudomonas aeruginosa. This suggests that biosurfactants such as rhamnolipids may be utilized as antimicrobial agents for removing Bordetella biofilms.

  • Bordetella bronchiseptica
  • Biofilm
  • Rhamnolipids

1 Introduction

Bacterial biofilms are surface-adherent, multi-cellular communities that represent a fundamentally different physiological state compared to free-living planktonic bacteria [13]. Biofilms are thought to be representative of most bacterial growth on natural surfaces, and can be found in a wide array of natural and artificial environments ranging from rocks in a stream to shower curtains [4,5]. Biofilms are also important in pathogenesis of various bacterial infections. Nosocomial infections of pathogens such as Staphylococcus spp. occur via biofilm formation on catheters and other medical equipments. In addition, dental plaques, cystic fibrosis pneumonia, and infective endocarditis are thought to be caused by bacteria forming biofilms [5]. Various reports indicate that biofilms confer an enhanced resistance against antibiotics [6]. Hypotheses as to why biofilms show increased antibiotic resistance include decreased antibiotic penetration through biofilm structures [7], upregulation of expression of multi-drug efflux pumps [8], and expression of periplasmic glucans that directly bind and sequester antibiotics [9]. The phenotypes of antibiotic resistance and enhanced attachment to surfaces represent difficulties in combating bacterial biofilms in both medical and industrial settings.

Pseudomonas aeruginosa biofilms have been extensively studied due to its relative ease of biofilm formation under various conditions in vitro, its medical importance, and the genetic tractability of the organism [10]. During the later stages of P. aeruginosa biofilm development, the production of the biosurfactant rhamnolipid was shown to be important in modulating microcolony architecture by maintaining channels to allow fluids to flow through the biofilm [11]. Several surfactants produced by other bacteria have been shown to inhibit biofilm formation of heterospecific bacteria, and examples include surfactin produced by Bacillus subtilis [12] and surlactin produced by Lactobacillus spp. [1315]. P. aeruginosa rhamnolipids may therefore be able to affect biofilms of other bacterial species in a similar fashion [11].

Bordetella bronchiseptica is a Gram-negative coccobacilli species closely related to the human pathogens Bordetella pertussis and Bordetella parapertussis, which cause whooping cough. B. bronchiseptica infects the respiratory tract of a wide range of mammals, often leading to life-long chronic colonization [16]. Once the respiratory tract of animals have been infected, it is extremely difficult to eliminate B. bronchiseptica from the host. Infection by B. bronchiseptica has been closely associated with various respiratory diseases including atrophic rhinitis in pigs, kennel cough in dogs, snuffles in rabbits and bronchopneumonia in cats. The most severe disease symptoms usually occur in co-infections with secondary agents including Pasteurella multocida and various respiratory viruses. Bordetella species utilize the BvgAS (Bordetella virulence gene) two-component signal transduction system to sense environmental stimuli and regulate the expression of various genes [17]. Many of these genes are virulence factors, including adhesins such as filamentous hemagglutinin (FHA), fimbriae, and pertactin; and toxins such as the bifunctional adenylate cyclase/haemolysin (ACY), pertussis toxin, and type III secretion system [16]. The differential control of the BvgAS system results in at least three distinct phases of growth under various environmental conditions: the virulent Bvg+, intermediate Bvgi, and the non-virulent Bvg phases. We have previously reported that B. bronchiseptica forms a maximal biofilm phenotype in Bvgi phase [18]. This suggests a possible role of biofilm growth in infected hosts, as conditions in the nasal cavity of the upper respiratory tract is hypothesized to support a Bvgi phase growth. The difficulty in removing this organism from infected hosts is also indirect evidence that biofilms may be involved in the colonization process.

In this report, we demonstrate that B. bronchiseptica biofilms can be effectively disrupted by treatment with the rhamnolipids secreted by P. aeruginosa PAO1 in vitro. Therefore, P. aeruginosa rhamnolipids may function to disrupt bacterial biofilms in nature, and this biosurfactant may be a potential candidate for use in treatment strategies to eliminate B. bronchiseptica infections.

2 Materials and methods

2.1 Bacterial strains and growth conditions

B. bronchiseptica strains RB50 (wild type) and RB53i (Bvgi phase-locked mutant) were previously reported and characterized [19,20]. RB50 and RB53i were cultured in Stainer–Scholte (SS) liquid medium [21] or on BG agar (Becton Dickinson) supplemented with defibrinated sheep blood at 37 °C. For Bvg phase modulation, bacteria were grown in SS media with nicotinic acid (Sigma) added to appropriate final concentrations. P. aeruginosa PAO1 (wild type), 12985 (rhlA′::ISlacZ/hah Tn5 mutant strain), and 45577 (rhlB′::ISphoA/hah Tn5 mutant strain) were obtained from the University of Washington Genome Center P. aeruginosa PAO1 mutant collection. Unless otherwise stated, P. aeruginosa were propagated on either LB agar or in LB broth.

2.2 Preparation of conditioned medium

Pseudomonas spent medium was prepared from P. aeruginosa cultures grown in 100 ml SS for 36 h at 37 °C. Cultures were centrifuged at 10,000g at 4 °C, and the supernatant was filtered through 0.22 μm cellulose acetate filter unit (Corning). Conditioned medium was prepared by adding fresh 2× concentrated SS to the spent medium.

2.3 Purification of rhamnolipids

P. aeruginosa rhamnolipids were isolated using a modified protocol from previous reports [22,23]. Polystyrene resin Amberlite XAD-2 (Supelco) was incubated with methanol for 30 min prior to preparation. Resin was then washed with 0.1 M phosphate buffer, pH 6.1, and centrifuged to remove the supernatant. Twenty millilitres of P. aeruginosa spent medium was incubated with the equal volume of resin at 37 °C with agitation for 12 h, and washed with phosphate buffer three times. Rhamnolipids were extracted with methanol, evaporated to dryness under vacuum, and then re-suspended with 500 μl de-ionized H2O.

2.4 Biofilm growth conditions

Quantitative assay of biofilm formation was performed in a 96-well plate format as previously described [18]. Overnight B. bronchiseptica cultures were inoculated 1:20 into 100 μl SS medium and incubated statically for 24 h at 37 °C. Biofilms were then washed vigorously with running water, and stained with 150 μl crystal violet for 30 min. The plates were dried, crystal violet stains were solubilised with 200 μl 33% acetic acid, and the biofilm formation was quantified by reading the absorbance at 595 nm using a plate reader. In experiments where biofilms were incubated with Pseudomonas conditioned media or rhamnolipids, media were first removed and replaced with 125 μl fresh medium containing antibiotics, conditioned medium, or rhamnolipid, and incubated at 37 °C. Biofilms for microscopy were grown in non-tissue culture-coated glass chamber slides (BD Falcon). B. bronchiseptica was inoculated 1:20 into 200 μl SS/chamber, supplemented with 0.1 M CaCl2 as previously reported to stabilize biofilm during the wash steps of staining procedures [24]. Addition of CaCl2 was shown not to affect B. bronchiseptica biofilm formation both by crystal violet staining assay and microscopically compared to paraformaldehyde-fixed biofilm (data not shown). Chamber slides were incubated statically at 37 °C for 24 h, washed with PBS three times, and proceeded to staining procedures.

2.5 Microscopy

Biofilms formed in chamber slides were stained using BacLight LIVE/DEAD staining kit (Molecular Probes) as directed by the manufacturer at room temperature. Samples were subsequently washed with PBS three times, and the top chamber was removed. Deconvolution micrographs of the biofilms were taken from the slides with a Leica DM R epifluorescence microscope with deconvolution software (Improvision Volocity).

3 Results

3.1 P. aeruginosa PAO1 secretes a substance that disrupts B. bronchiseptica biofilms

In the course of examining the effects of secreted molecules from various bacteria on B. bronchiseptica biofilm formation, we observed that the addition of P. aeruginosa PAO1 conditioned medium to B. bronchiseptica biofilms on polystyrene and glass surfaces rapidly disrupted the biofilms. Biofilms formed by B. bronchiseptica on various surfaces were incubated with conditioned media from various sources for up to 24 h. As shown by quantitative crystal violet staining assay of B. bronchiseptica biofilms in 96-well polystyrene plates in Fig. 1(a), addition of P. aeruginosa PAO1 conditioned medium to pre-formed B. bronchiseptica biofilms under all conditions were disrupted. We have shown previously that Bvgi-locked mutants form robust biofilms [18]– the addition of conditioned medium of PAO1 also disrupted the biofilm formed by this strain (Fig. 1(b)). Similar experiments performed with Bordetella biofilms grown on polystyrene tubes on continuous rolling conditions showed that the biofilms can be disrupted by PAO1 conditioned medium within 10 min under these conditions of higher mechanical shear (data not shown). The disruption of the Bordetella biofilm by PAO1 conditioned medium was also observed with biofilms grown on glass surfaces (Fig. 1(c) and (d)). Treatment of the biofilm with control medium for 4 h (Fig. 1(c)) had no apparent effect on the biofilm; however, treatment with PAO1 spent medium for 4 h eliminated a large portion of the attached cells. Interestingly, biofilm disruption by P. aeruginosa conditioned medium was apparently not caused by the reported bactericidal activity of rhamnolipids [25], because a majority of the detached cells (not shown) and remaining attached cells (Fig. 1(d)) were viable as judged by BacLight staining. Furthermore, B. bronchiseptica grown in non-biofilm forming conditions and incubated with PAO1 conditioned medium for periods up to 24 h did not show a decrease in viability (Fig. 2).

Figure 1

P. aeruginosa PAO1 conditioned medium disrupts B. bronchiseptica biofilm. (a) Wild type B. bronchiseptica was grown in 96-well polystyrene plates for 24 h in various Bvg conditions (0.4 mM nicotinic acid being intermediate phase with maximal biofilm formation), and growth media were then replaced with conditioned media from cultures of PAO1 (black bar) or control fresh media (open bar), each supplemented with respective nicotinic acid concentrations, for 4 h. (b) The Bvgi phase constitutive mutant RB53i (which forms high biofilm under all conditions) was grown on polystyrene for 24 h and growth medium was then replaced with conditioned medium from cultures of PAO1 (right) or control fresh medium (left) for 24 more hours. Biofilm formation was quantified by absorbance of solubilised crystal violet stains. Error bars in (a) and (b) represent standard deviations from the experiments done in at least triplicates. (c and d) RB53i biofilm was grown on glass chamber slides for 24 h, and growth medium was replaced by either fresh control medium (c), or PAO1 conditioned medium (d), and incubated for 4 h. Biofilms were stained with LIVE/DEAD staining kit, and the images were de-convolved. Green represents live cells and red represents dead cells.

Figure 2

P. aeruginosa conditioned medium does not display bactericidal effects on B. bronchiseptica. B. bronchiseptica RB50 was grown in 1 ml SS overnight in the Bvg+ phase (non-biofilm phase). One millilitre of either SS medium (control) or PAO1 spent medium were then added to the cultures and the colony forming units (cfu) were counted by plating at 0, 4, and 24 h after addition of the media. Data shown is representative of at least three independent experiments.

3.2 Conditioned media from the P. aeruginosa rhlA and rhlB mutants are deficient in the disruption of B. bronchiseptica biofilm

Several bacterial species have been characterized to secrete biosurfactants, and these substances have been shown to be able to inhibit biofilm formation of other species [1215]. P. aeruginosa secretes biosurfactants known as rhamnolipids [22,23]. These compounds have been implicated in such roles as biofilm architecture maintenance [11] and they also possess antimicrobial activities [26,27]. Rhamnolipid synthesis in P. aeruginosa is controlled by quorum sensing and mainly occurs during late stationary phase and in resting cells [26]. The rhlA gene in P. aeruginosa encodes for a rhamnosyltransferase that is critical for rhamnolipid synthesis, and an rhlA mutant is defective in the production of rhamnolipid [28,29]. The rhlB gene is thought to catalyze a reaction converting a rhamnolipid precursor 3-(3-hydroxyalkanoyloxy)alkanoic acid (HAA) to mono-rhamnolipid, which serves as the substrate for RhlC that catalyzes the formation of di-rhamnolipid. HAA was previously reported to be the sufficient form of the biosurfactant for swarming motility of P. aeruginosa [30]. When we compared the conditioned media from either rhlA or rhlB mutants with that from the wild type PAO1 in their abilities to disrupt B. bronchiseptica biofilms, we found that the conditioned media from both mutants are defective in the disruption of B. bronchiseptica biofilms (Fig. 3). This result strongly suggests that mono- and/or di-rhamnolipid, and not its precursor HAA, is primarily responsible for the disruption of the B. bronchiseptica biofilms.

Figure 3

Conditioned media from P. aeruginosa rhlA and rhlB mutants are defective in B. bronchiseptica biofilm disruption. The Bvgi phase-locked B. bronchiseptica mutant RB53i was grown on polystyrene for 24 h, followed by incubation with control medium, P. aeruginosa PAO1 conditioned medium, rhlA or rhlB mutant conditioned media for another 4 h. Biofilm was quantified by absorbance of solubilised crystal violet stains. Error bars represent standard deviations from the experiment done in hextuplicates.

3.3 Purified rhamnolipids from wild type P. aeruginosa disrupts B. bronchiseptica biofilm

Based on a modified protocol developed from previous reports [22,23], we purified rhamnolipids from the culture medium of wild type P. aeruginosa PAO1 and also made a control preparation from the rhlA mutant. The rhamnolipid preparations were examined with thin-layer chromatography based on a published protocol [31], and we observed the expected rhamnolipid staining on the TLC plate only from the preparation of the wild type bacterial culture (data not shown). The rhamnolipid preparation from the culture medium of wild type P. aeruginosa PAO1 was capable of disrupting a significant portion of the B. bronchiseptica biofilm in vitro, but the control preparation from the rhlA mutant did not (Fig. 4). The purified rhamnolipid preparation was apparently less effective than the total conditioned medium from PAO1 in the disruption of the Bordetella biofilm, and this may be due to the presence of other substances in the culture medium that can contribute to the disruption of the biofilms, or the presence of higher concentrations of rhamnolipids in the unpurified medium. Serial dilution of PAO1 conditioned media of up to 10% concentration showed biofilm dispersion, albeit lower effectivity than at 100% concentration (data not shown), suggesting the likelihood of the latter hypothesis.

Figure 4

Rhamnolipid preparation from wild type P. aeruginosa disrupts B. bronchiseptica biofilm. The Bvgi phase-locked B. bronchiseptica mutant RB53i was grown on polystyrene for 24 h, followed by incubation with control medium, the rhamnolipid preparation from wild type P. aeruginosa PAO1 culture medium (1:10 dilution in culture medium), or a similar preparation from rhlA mutant culture medium, for another 24 h. Biofilm was quantified by absorbance of solubilised crystal violet stains. Error bars represent standard deviations from the experiment done in hextuplicates.

4 Discussion

Understanding the biology of biofilms and formulating new strategies to combat biofilm formation or facilitate their removal are pressing issues in medical and industrial microbiology. Most bacterial biofilms display resistance against antimicrobials such as antibiotics and various host immune responses [32,33]. Bordetella bronchiseptica, like many other bacterial species, displays an increased resistance to antibiotics in the biofilm state compared to planktonic cells [34, our unpublished data]. As previously suggested, it is possible that B. bronchiseptica establishes a life-long niche within the host respiratory tract by forming biofilms to evade host immune surveillance and resist antibiotic therapies [18]. Although planktonic cells of B. bronchiseptica are sensitive to gentamicin with MIC < 10 μg/ml [35,36, our unpublished data], it is impossible to eliminate B. bronchiseptica from infected mice with gentamicin treatment via various routes (personal communications from Dr. Eric Harvill, Penn State University, and our unpublished data). Other antibiotics have also been unsuccessful. The in vitro data of increased antibiotics resistance of B. bronchiseptica biofilms is consistent with the formation of B. bronchiseptica biofilms in vivo.

In this report, we demonstrate that rhamnolipids secreted by P. aeruginosa can lead to a significant disruption of the B. bronchiseptica biofilm structure in vitro. Although the mechanism of rhamnolipids on biofilm disruption is not known, a more generalized activity of altering charge-charge properties is hypothesized [11], which likely decreases the chances for the bacteria acquiring resistance due to spontaneous mutations. Rhamnolipid has been implicated as an antimicrobial agent secreted by P. aeruginosa in order to exclusively acquire their ecological niche [26,27] and may represent one of the mechanisms of “microbial warfare.” Previous reports have shown that rhamnolipid is effective in preventing or slowing the formation of, or accelerating the dispersion phase of biofilm development of other bacterial species [1215]. In this report, rhamnolipid-induced dispersion was shown to be effective on a fully mature maximal biofilm formed by B. bronchiseptica.

Planktonic B. bronchiseptica has demonstrated a higher resistance against the bactericidal effects of P. aeruginosa rhamnolipids compared to other bacterial species in various reports [22,23]. We also showed that cells released from B. bronchiseptica biofilms after treatment with P. aeruginosa rhamnolipids were mostly alive. Therefore, the primary effect of the P. aeruginosa rhamnolipids on the B. bronchiseptica biofilm does not appear to be bactericidal. As our purified rhamnolipid preparations were less efficient in the disruption of the biofilms compared to conditioned medium, it is possible that other secreted substances from P. aeruginosa also contribute to the biofilm disruption effect. Other secreted molecules from P. aeruginosa that possess antimicrobial activities include pyocyanin [37]. The majority of cystic fibrosis patients in the later stages of infection are almost exclusively colonized by P. aeruginosa, and this success in niche acquisition to the exclusion of other pathogens may indicate the importance of production of antimicrobial substances that disrupt the biofilms of heterospecific organisms.

The use of rhamnolipids or other biosurfactants that have potent activities to disrupt biofilm structures, in combination with antibiotic treatment, may represent another productive antimicrobial strategy. Antibiotics have strong bactericidal or bacteriostatic effects against various growing planktonic bacterial cells but are generally less effective against biofilms. The biosurfactant/antibiotic combination therapeutic strategies are therefore likely to be more effective against various chronic bacterial infections that may be associated with biofilm formation, including B. bronchiseptica infections. The ability of the rhamnolipids to disrupt pre-formed biofilms, as demonstrated in this report, is critical for these treatment strategies, as it can then facilitate the access of antibiotics to the bacterial cells that have been released from biofilms. This report also sheds light on the fact that not all biosurfactants are necessarily effective against biofilms. Although HAA is also a biosurfactant, the precursor to rhamnolipid evidently has no effect on biofilm dispersion. It is likely that for each biofilm system formed by particular bacterial species, the effective biosurfactants may be different.


The authors would like to thank the University of Washington Genome Center for supplying the P. aeruginosa strains. This work was supported in part by NIH grant AI04936 to M.H.Y. and AI051360 to G.A.O.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
  35. [35].
  36. [36].
  37. [37].
View Abstract