|Home | About | Journals | Submit | Contact Us | Français|
Geobacter sulfurreducens required expression of electrically conductive pili to form biofilms on Fe(III) oxide surfaces, but pili were also essential for biofilm development on plain glass when fumarate was the sole electron acceptor. Furthermore, pili were needed for cell aggregation in agglutination studies. These results suggest that the pili of G. sulfurreducens also have a structural role in biofilm formation.
One of the hallmarks of Geobacter species is their ability to conserve energy from the transfer of electrons to a variety of extracellular electron acceptors, such as metals [Fe(III), Mn(IV), and U(VI)], humic acids, and electrodes (5, 6). Establishing an electrical connection with an extracellular electron acceptor poses challenges not faced by microorganisms that reduce soluble electron acceptors within the cell. In contrast to other Fe(III) oxide-reducing bacteria (4, 9-11), such as Shewanella and Geothrix species, Geobacter species do not excrete electron shuttles (8) and require direct contact with the electron-accepting surface (1, 10). Previous studies (14) have demonstrated that the pili of Geobacter sulfurreducens are conductive and that expression of pili is required for growth on Fe(III) oxides. These “microbial nanowires” are not required for attachment to the insoluble electron acceptor; rather, they function as electronic conduits to transfer electrons to the Fe(III) oxides, extending the electron transfer capabilities of the cell well beyond the outer surface (14). Pilus “nanowires” also serve as electric conduits to mediate long-range electron transfer across multilayer biofilms formed on anode electrodes, which is required to maximize current production per unit of anode surface area (15).
When G. sulfurreducens (2) was grown under strictly anaerobic conditions at 30°C in freshwater medium (7) with acetate (15 mM) as an electron donor and with Fe(III) oxide coatings [prepared on borosilicate coverslips (17) and providing 4.3 ± 0.7 μmol of Fe(III) per coverslip (mean ± standard deviation; n = 3)] as the sole electron acceptor, a biofilm grew on the Fe(III) coating (measured with a crystal violet assay ), but planktonic growth was not supported (Fig. (Fig.1A).1A). Viability staining with a BacLight viability kit (Molecular Probes) and confocal scanning laser microscopy (CSLM) analyses (14) of 48-h biofilms revealed a structured biofilm composed of cell clusters approximately 18 ± 1 μm high (Fig. (Fig.1B).1B). Control coverslips without the Fe(III) oxide coatings did not support biofilm growth (Fig. (Fig.1C),1C), suggesting that biofilm growth was not supported by any nutrient carried over in the inoculum. Viability staining suggested that even cells at a substantial distance from the Fe(III) oxide surface remained metabolically active (Fig. (Fig.1B).1B). This may be attributed to long-range electron transfer via the electrically conductive pili, as previously proposed for long-range transfer to the anode surface of microbial fuel cells (15). In contrast, a previously described (14) mutant in which the gene coding for PilA, the pilin structural subunit, was deleted grew poorly on the Fe(III) oxide coatings (Fig. (Fig.1D)1D) and produced 10-fold less biomass than the wild type produced after 72 h (Fig. (Fig.2),2), and complementation of the mutation in trans (14) restored the biofilm phenotype (data not shown). These findings are consistent with the previous finding that pili are required for growth on Fe(III) oxide (14).
Even though pili are not required for growth with fumarate as an electron acceptor (14), addition of fumarate (40 mM) to cultures with Fe(III) oxide-coated coverslips, while having little impact on the biofilm biomass of the wild type, increased the mutant biofilm biomass to approximately one-half of the wild-type biomass (Fig. (Fig.2).2). More wild-type biomass accumulated on glass coverslips when fumarate was provided as the electron acceptor, but the biomass of the pilin-deficient mutant biofilms remained approximately one-half that of the wild-type biofilms (Fig. (Fig.2).2). These results demonstrated that pili are required for optimal biofilm development even when the surface is not the electron acceptor.
This conclusion was consistent with CLSM images of the fumarate-grown biofilms (Fig. (Fig.3).3). In the presence of fumarate, wild-type cells formed pillars that were 19 ± 1.5 and 22 ± 0.5 μm high on Fe(III) oxide-coated surfaces and glass surfaces, respectively. In contrast, the maximum biofilm heights for the pilin-deficient mutant were 7.6 ± 1.5 and 8.5 ± 1.4 μm on the same surfaces. This difference was not apparent in the first 24 h, when both the mutant and wild-type biofilms formed short microcolonies. However, the wild-type microcolonies continued to grow, and both the height and width of the colonies increased to form mature biofilms (Fig. (Fig.3).3). Viability staining indicated that cells in all the layers of the wild-type and mutant biofilms were alive, suggesting that fumarate diffusion across the biofilms was not a limiting factor, as previously reported for other bacterial biofilms (16). These results indicate that the pili of G. sulfurreducens play a role in the development of the highly structured biofilms of G. sulfurreducens that is unlikely to be related to the electrical conductivity of the pili.
Pili in various bacteria mediate twitching motility during biofilm formation (12). Other pili, such as the toxin-coregulated pili (TCP) of Vibrio cholerae (13), and also the pili of G. sulfurreducens (14) do not appear to be involved in motility. Rather, TCP are a structural biofilm component that mediate cell interactions leading to microcolony development during colonization of the human intestine (3) or during biofilm formation on chitin surfaces (13). The ability of TCP to promote bacterial interactions also enables TCP-expressing cells to autoagglutinate in vitro (3). Similar agglutination studies were carried out with G. sulfurreducens by growing cells with fumarate as the electron acceptor at 25°C to induce pilus formation (14). The degree of agglutination was assayed by measuring the optical density at 600 nm of the cells that remained in suspension and subtracting the value obtained from the optical density of the culture after disruption of the aggregates with agitation. After 72 h of growth the wild-type strain formed large aggregates that settled at the bottom of the culture vessel (Fig. (Fig.4).4). There was no autoagglutination at 30°C, a temperature at which planktonic cells do not express pili (14). The mutant in which pilA was deleted did not agglutinate at 25°C (Fig. (Fig.4).4). Complementation of the mutation with a wild-type copy of the pilA gene expressed in trans produced a strain that agglutinated at levels that were much higher than the levels observed for the wild-type strain (Fig. (Fig.4),4), consistent with the fact that genetic complementation leads to overproduction of pili (14). These results suggest that the pili of G. sulfurreducens participate in cell-cell aggregation necessary for the development of microcolonies during biofilm differentiation.
The results presented here demonstrate that in addition to serving as electric conduits for electron transfer to Fe(III) oxides (14) and long-range electron transfer across anode biofilms in G. sulfurreducens fuel cells (15), the G. sulfurreducens pili also are required for maximum biofilm growth even when electron transfer to an electron-accepting surface is not required. This is an important consideration because the overall rate of electron transfer to an electron-accepting surface is dependent upon the number of metabolically active cells that can stack on the surface. Thus, high rates of electron transfer to an electron-accepting surface require not only the electronic capabilities of the pili but also their structural attributes that permit cells to stack at high densities on a given surface. These considerations make it clear that further evaluation of the contributions of pili and other outer cell components to the biofilm structure is essential in order to better understand, and perhaps optimize, electron transfer to electron-accepting surfaces.
This research was supported by grants DE-FG02-02ER63423 and DE-FC02-02ER63446 from the Office of Science (BER), U.S. Department of Energy, and by award N00014-03-1-0405 from the Office of Naval Research. G.R. acknowledges support provided by a postdoctoral fellowship from the Ministerio de Educación y Ciencia of Spain and by the European Social Fund.
Published ahead of print on 8 December 2006.