When the Party Is Over: a Signal for Dispersal of Pseudomonas aeruginosa Biofilms

The mechanisms that permit bacteria to make the transition from an unattached or planktonic cell to a surface-associated multicellular biofilm and back have become a subject of intense interest to microbiologists over the past several years. Studies by a research team led by Staffan Kjelleberg, reported in this issue (2), have identified nitric oxide (NO), an endogenous product of anaerobic metabolism or possibly a product thereof, as an elicitor of biofilm dispersal or detachment by Pseudomonas aeruginosa. Although NO was already known to have toxic effects on this bacterium, the levels that induced biofilm dispersal and inhibited biofilm formation were far below the toxic levels. Furthermore, low levels of NO were sufficient to sensitize the cells to killing by at least some bactericidal agents. How these effects of NO are mediated is yet to be determined. Nitric oxide is an important signaling molecule in eukaryotes, and this study provides tantalizing evidence that this also may be true for bacteria. In any case, these findings have advanced our understanding of one way in which biofilm dispersal can be triggered in response to the metabolic state of the culture and no doubt will stimulate new studies on biofilm dispersal mechanisms. They may also lead to practical applications in the search for improved therapies for infections caused by P. aeruginosa and perhaps other species.

The biofilm lifestyle.

Many or perhaps most bacteria have a strong propensity to form multicellular, matrix-enclosed assemblies, or biofilms, which are found on surfaces throughout the biological world. Cells within a biofilm have a number of advantages over their planktonic counterparts, including protection against the immune system and predation by protozoa, enhanced ability to transfer genetic information, and enhanced resistance to antimicrobial agents and other stresses (reviewed in reference 9). Thus, biofilm formation complicates a variety of chronic infections, including the devastating pulmonary infections that are caused by P. aeruginosa in cystic fibrosis patients and other opportunistic infections by this organism (31). On the other hand, the biofilm environment is restrictive for bacterial growth; the transcriptome of mature biofilm, on average, resembles that of stationary-phase cells (26). Entrapment in the biofilm matrix also precludes the ability of a cell to flee hostile conditions or migrate to a more propitious site when opportunities arise. Not surprisingly, microbes actively form, as well as exit or disperse, from biofilm. The resulting developmental cycle is characterized by a succession of changes in gene expression and phenotype (reference 21 and references therein).

Formation and dispersal of biofilm.

The universal task in building a biofilm is to alter the cell surface and immediate surroundings of the cell in order to promote intercellular and cell surface interactions. Molecular genetic investigations have revealed diverse factors that are necessary for biofilm formation by various bacteria and even by a single species. Motility (16), proteaceous adhesins (25), and polysaccharides, such as Psl and Pel of P. aeruginosa (7, 13), promote surface attachment and cell-cell adhesion, while acidic exopolysaccharides, such as alginate, tend to affect maturation of biofilm architecture (18). Small molecules and even extracellular DNA also contribute to biofilm structural stability (1, 22).

Biofilm dispersal processes have been slower to be defined than biofilm formation but are also yielding to molecular genetic investigation. In principle, regulation of adhesin synthesis could govern biofilm detachment (discussed in reference 24). Likewise, the destruction of an adhesin can lead to dispersal of certain species. For example, the oral pathogen Actinobacillus actinomycetemcomitans secretes a hydrolytic enzyme that cleaves poly-β-1,6-N-acetyl-d-glucosamine, a polysaccharide adhesin of diverse species (11, 14). Surfactant production can also lead to removal of attached cells and is necessary for modeling the three-dimensional architecture of P. aeruginosa biofilm (5). Biofilm dispersal can be coordinated in response to environmental shifts, such as increased or decreased nutrient availability or a change in temperature (10, 20, 24). Aging biofilms of P. aeruginosa exhibit a distinct and particularly dramatic form of dispersal, sometimes referred to as seeding dispersal. In this process, the surface-attached microcolonies of aging biofilms undergo internal disintegration, leaving behind “hollow” shell-like structures (17). This involves the induction of an endogenous prophage under conditions of anaerobic metabolism (28). While seeding dispersal is associated with cell death, viable cells that can colonize elsewhere are simultaneously released. Nutrient addition can induce dispersal of P. aeruginosa biofilm and is associated with sweeping changes in gene expression (20). This form of dispersal has been found recently to depend on BdlA, a transducer protein that apparently is located in the cytoplasm (15).

Anaerobic metabolism by an oxidative organism.

The realization that P. aeruginosa grows anaerobically in a biofilm-like form within the lungs of cystic fibrosis patients was an important advance in understanding this disease (30). In part, it led to the finding that NO is a toxic antimicrobial agent produced not only by host immune cells but also by the bacterium itself during anaerobic respiration in lung secretions. While the biofilms studied by Barraud et al. (2) were grown aerobically, oxygen availability was sufficiently restricted during the late stages of biofilm growth to permit anaerobic respiration, NO production, and seeding dispersal to occur simultaneously. Anaerobic conditions also induce dispersal of other biofilms (reference 24 and references therein).

Regulation of biofilm dispersal versus formation: two sides of a coin?

The studies by Barraud et al. implicated NO in seeding dispersal of aging biofilms as well as dispersal of younger-batch biofilms (2). Only the former process involved cell killing. Furthermore, when NO was present continuously at low levels in cultures, it inhibited biofilm formation without affecting cell growth and it stimulated motility. High levels of NO are known to induce stress responses in P. aeruginosa (6), but this is probably much different than the response of this bacterium to low NO concentrations. So, how might the effects of NO be mediated?

There is some support for the idea that biofilm formation versus dispersal and cell motility often represent opposite sides of a regulatory coin. A potential target for NO that exhibits such effects is the intracellular nucleotide cyclic di-GMP (4, 19, 24). This widespread bacterial second messenger is known to activate exopolysaccharide (e.g., cellulose) biosynthesis allosterically but also affects transcript levels by undefined mechanisms (3). The Csr/Rsm system is another possible target of NO signaling. In the best-studied example, the protein CsrA of Escherichia coli represses biofilm formation and induces motility and biofilm dispersal (12). The mechanisms for the first two processes involve CsrA binding to the mRNA leaders of the pgaABCD and the flhDC operons, respectively (27, 29). The way in which CsrA facilitates dispersal is still unclear. CsrA is sequestered by noncoding RNAs CsrB and CsrC, which are transcribed via the BarA-UvrY (GacS-GacA) two-component signal transduction system (23). Homologous regulatory circuitry of P. aeruginosa may control transitions between a biofilm-like persistent state of infection and a phenotype favoring acute, virulent infections (8). Although its regulatory circuitry is not yet established, BldA (15) is also a plausible candidate for mediating NO effects.

Questions for the future.

It is now evident that low concentrations of NO directly or indirectly induce biofilm dispersal and inhibit biofilm formation in P. aeruginosa, but many questions remain. How are these effects mediated? Does NO also affect dispersal of biofilms of other species, such as anaerobic dispersal of Shewanella oneidensis (24)? Can these findings be extrapolated to the behavior of P. aeruginosa in lung or other infections? Perhaps most importantly, will further delineation of the pathways and mechanisms responsible for NO effects suggest novel clinical applications? Concerning the broader search for understanding of biofilm dispersal, the party has just begun.