Competitive Interactions in Mixed-Species Biofilms Containing the Marine Bacterium Pseudoalteromonas tunicata

Isolation of marine strains. P. tunicata and other isolates were originally isolated from the surface of the common marine alga U. lactuca, which was collected from the rocky intertidal zone near Sydney, Australia. The algal thallus was suspended in sterile nine-salt solution, and surface bacteria were removed by vortexing. Aliquots of the samples were then spread on the complex marine medium Vaatanen nine-salt solution (VNSS) (40) containing 1.5% agar and incubated at 23°C for 48 h. Morphologically distinct bacterial colonies were selected. Bacteria were stored at −70°C in 30% glycerol. While over 50 different colony phenotypes were observed, only 20 strains survived repeated subculturing on VNSS. These strains were routinely grown and maintained on VNSS agar at 25°C. Sequencing of these strains was part of another project and was carried out by Tujula et al. (N. Tujula, J. S. Webb, I. Dahllöf, C. Hölmström, and S. Kjelleberg, unpublished data). Four of the 20 culturable strains were chosen in addition to P. tunicata, as these strains are commonly found growing on algae. The strains were Pseudoalteromonas gracilis, Alteromonas sp., Cellulophaga fucicola, and Roseobacter gallaeciensis strains.

Fluorescent labeling of marine isolates.Marine strains were labeled with a green fluorescent protein (GFP) color tag by transconjugation by using the constitutive GFP expression plasmid pCJS10. This plasmid contains the gfpmut3 gene (11) on an RSF1010 backbone from the broad-host-range vector pHRP304 (3). In addition, strains were labeled with a red fluorescent protein (RFP) color tag by using the pCJS10-derived plasmid pCJS10R. This plasmid contains the RFP gene dsred (Clontech) in place of gfpmut3 on pCJS10. Plasmids pCJS10 and pCJS10R were gifts from Charles Svenson, University of New South Wales, Sydney, Australia. GFP triparental conjugations were carried out as described previously (19), and labeled transconjugants were grown on VNSS agar plates containing 15 μg of chloramphenicol per ml and 100 μg of streptomycin per ml. GFP- and RFP-labeled strains showed bright fluorescence after overnight culture, and we observed no differences in the growth rates or surface attachment properties of the labeled strains compared with the unlabeled parent strain (data not shown).

Detection of inhibitory compounds produced by marine bacterial isolates.In order to detect inhibitory compounds produced by marine bacterial isolates, concentrated supernatants from P. tunicata and other marine isolates were prepared by the method described previously (33), with some modifications. The strains were grown in 3M medium with 0.2% (wt/vol) trehalose and 0.2% (wt/vol) glucose as the carbon sources for 48 h, harvested by centrifugation (12,000 × g for 20 min), and resuspended in 3M medium (47) containing trehalose at a density of 0.7 g ml⁻¹. Each concentrated cell suspension was incubated without shaking at 28°C for 24 h. Cells were removed by centrifugation (14,000 × g for 1.5 h), and the concentrated supernatant was sterilized by using a 0.2-μm-pore-size sterile filter (Millipore).

Supernatant samples were assayed for inhibitory activity by the drop test assay (33). Briefly, 100 μl of an overnight broth culture of the target strain was spread on a VNSS agar plate, and the plate was dried at 30°C for 30 min. Drops containing 20 μl of the concentrated supernatant, as well as a control (nine-salt solution), were placed on the agar surface and incubated overnight at room temperature to allow formation of inhibition zones.

Biofilm experiments.Biofilms were grown in continuous-culture flow cells (channel dimensions, 1 by 4 by 40 mm) at room temperature as previously described (44). Channels were inoculated with overnight cultures and incubated with no flow for 1 h at room temperature. Cultures were adjusted so that biofilms were established with a flow rate of 150 μl min⁻¹. R. gallaeciensis, which could not be labeled with a fluorescent color tag, was stained with Syto 59 diluted to a concentration of 3 μl ml⁻¹ in biofilm media. (Biofilms were grown in 3M medium containing 0.01% trehalose, 0.01% glucose, and 0.01% fructose.) Biofilms were visualized with a confocal laser scanning microscope (Olympus) by using fluorescein isothiocyanate and tetramethyl rhodamine isocyanate optical filters.

To cultivate mixed biofilms, flow chambers were inoculated with 500 μl of a mixture of stationary-phase cultures of P. tunicata and a competitive marine strain. In order to ensure that the initial ratio of attached cells of the two competing strains was 1:1, we first monitored the initial attachment of the mixed culture after a 1-h adhesion period. In most cases, the initial levels of attachment to the glass surface were approximately equal, as determined by counting the numbers of red- and green-labeled cells by epifluorescence microscopy by using an eyepiece grid. However, for C. fucicola and Alteromonas it was necessary to increase the ratios of cells to 4:1 and 2:1, respectively, in order to achieve equal levels of attachment for the two competing strains.

To investigate in more detail whether microcolonies improved the competitiveness of strains, marine strains were allowed to preestablish and form microcolonies for 24 h (P. gracilis) and 48 h (C. fucicola and Alteromonas sp.). Each preformed biofilm was inoculated with ∼10 ⁷ cells of wild-type P. tunicata, and the flow was stopped for 1 h. After resumption of the flow, the biofilm was examined for red and green fluorescence. Experiments were repeated in three separate rounds with three independent flow cells running in parallel.

Characterization of marine strains. (i) 16S rRNA gene sequencing.We first carried out partial 16S rRNA gene sequencing of bacteria isolated from U. lactuca (Table 1). We found that the four isolates studied all exhibited 16S rRNA gene homology with organisms that are commonly isolated from marine eukaryotic surfaces. C. fucicola belongs to the Flavobacterium group of the Bacteroidetes, a diverse group with members commonly found in coastal marine regions (49). C. fucicola is frequently found on the surface of marine benthic macroalgae (6) and has been found to decompose highly polymeric material from the brown alga Fucus serratus (34). The sequenced strains in our collection also contained Pseudoalteromonas and Alteromonas spp. The recently described genus Pseudoalteromonas resulted from division of the genus Alteromonas into two genera, Alteromonas and Pseudoalteromonas, based on the phylogenetic comparison of Gauthier et al. (21). Numerous bacteria belonging to these genera are frequently isolated from marine waters and are found in association with marine invertebrates, algae, plants, and animals (31). They are readily culturable and are also capable of surviving in nutrient-poor environments. P. gracilis is frequently found on algae such as Gracilaria and produces disease symptoms due to its agarolytic activity (52). Members of the genus Roseobacter are often found on the surfaces of algae (9, 43).

(ii) Production of extracellular antibacterial compounds.We examined the ability of the bacterial isolates from U. lactuca to inhibit the growth of each of the other isolates from the plant. Studies of the antibacterial activity in concentrated supernatants indicated that P. tunicata and R. gallaeciensis were the most inhibitory (Table 2). Each of these strains could inhibit the growth of several of the bacterial strains used in this study. P. tunicata exhibited strong inhibitory activity against itself and C. fucicola and weak inhibitory activity against P. gracilis. The 190-kDa protein (AlpP) responsible for the antibacterial activity of P. tunicata has now been well characterized (33). This protein contains at least two subunits (60 and 80 kDa), which are joined together by noncovalent bonds, and it was shown to be released during the stationary phase (33). P. tunicata possesses a ToxR-like regulon which appears to control determinants for the expression of iron uptake and also regulates expression of AlpP (S. Stelzer, S. Egan, and S. Kjelleberg unpublished data).

R. gallaeciensis exhibited strong activity against both P. tunicata and P. gracilis. Production of secondary metabolites by members of this group has been reported previously (5, 25, 51). One of the inhibitory compounds (proposed to be a peptide) has been reported to be produced only in the presence of other bacteria (51), although in this study we observed strong inhibitory activity with pure-culture supernatants of this strain. More recent work indicated that R. gallaeciensis also produces a new antibiotic called tropodithietic acid (8). This compound showed strong inhibitory activity against marine bacteria belonging to various taxa and marine algae.

C. fucicola did not exhibit antibacterial activity against the strains used in this study. However, a Cytophaga strain (RB1057) has previously been shown to produce an extracellular inhibitor of expansion of colonies of closely related Cytophaga strains (10). The inhibitor (a glycoprotein) inhibited the competing strain's ability to adhere to and glide on a substrate. The inhibitor had no measurable impact on several other related strains of gliding bacteria. Thus, inhibitors produced by this genus may only be effective against closely related bacteria, which may explain why no activity was detected for the Cellulophaga strain tested in this study. In their studies on antagonistic interactions among marine pelagic bacteria, Long and Azam found that the members of the Bacteroidetes group were the most sensitive to inhibition by other marine bacteria and were also the least inhibitory (37). Similarly, Grossart and coworkers found that the Flavobacterium-Sphingobacterium group had the lowest percentage of inhibitory strains among a diverse group of 51 marine bacterial isolates (27).

In this study Alteromonas sp. showed activity against P. tunicata, C. fucicola, and R. gallaeciensis. Alteromonas spp. have been shown to produce a wide range of inhibitory compounds. Some species, such as Alteromonas citrea and Alteromonas rubra, produce only a macromolecular polyanionic substance, whereas other species, such as Alteromonas luteoviolacea, produce both a diffusible macromolecule and two intracellular low-molecular-weight brominated compounds (22). Barja et al. (4) found that Alteromonas species isolated from seaweeds produced thermolabile low-molecular-weight inhibitors (molecular weights, less than 2,000), whereas strains isolated from seawater produced a high-molecular-weight glycoprotein (molecular weight, 90,000), which displayed a broad spectrum of inhibitory activity against medical and environmental isolates. An Alteromonas strain (SWAT5) derived from particulate organic matter was found to produce 2-alkyl-quinolinols. The antibiotics were produced only on polysaccharide matrices and were found to be hydrophobic (38).

Single-species biofilm development.Biofilm development for each of the marine strains in monoculture was monitored in glass flow cells (Fig. 1). For all strains, single cells were observed to attach to the substratum after inoculation. Each strain established a stable biofilm with a characteristic architecture after 72 h. All of the strains exhibited microcolony formation to various extents during the course of biofilm development. P. tunicata and P. gracilis each formed well-defined spherical microcolonies that were up to 50 μm in diameter and 100 μm high within 48 h after inoculation. Alteromonas sp. also formed large, distinct microcolony structures which were frequently over 100 μm in diameter and 50 μm high after 72 h of biofilm development. In contrast, the biofilms formed by R. gallaeciensis and C. fucicola were less structured. The C. fucicola biofilms consisted of both flat unstructured regions and many small microcolonies that were up to 10 μm in diameter and up to 10 μm high. R. gallaeciensis initially formed cell chains which later developed into small microcolonies (10 to 20 μm) during the early stages of biofilm development. These clusters then formed a relatively flat, mat-like biofilm whose thickness approached 10 μm. For each strain, the characteristic architecture formed after 72 h persisted for at least another 2 days within flow cells. However, after 8 days we noted that the majority of the biomass had detached from the flow cells for all of the strains studied, which left only single cells attached to the flow cells (data not shown).

• Open in new tab
• Download powerpoint

FIG. 1.

Single-species biofilm development for bacteria isolated from the marine alga U. lactuca expressing GFP. Magnification, ×600.

The marine strains studied here comprise both organisms that form microcolonies rapidly and organisms that form less-defined microcolonies. Previously, it has been proposed that microcolonies can result from carbon flux gradients during growth within biofilms (56). In addition, many genetically encoded regulatory and structural determinants of microcolony formation have been revealed over the past few years. Some recent examples include the roles of conjugative plasmids (23, 50) and antigen 43 (13, 50) in enhancing microcolony formation in Escherichia coli biofilms, suggesting that these cell surface structures might act as cellular adhesins in stabilizing microcolony structures. Generally, the formation of multicellular structures, such as microcolonies, is thought to be an adaptive response that mediates survival under unfavorable conditions.

Of the marine organisms studied here, only P. tunicata has previously been characterized for biofilm development under continuous-culture conditions. In agreement with the present study, P. tunicata was observed previously to form prominent microcolony structures during biofilm development (39). Biofilm development has been well characterized in the marine bacterium Vibrio cholerae, which also can form microcolonies (45, 58, 59). To our knowledge, however, there have been no reports that have considered the ecological role of biological structures in mediating interactions between bacteria in mixed-community marine biofilms. In several studies the workers investigated interactions in marine biofilm communities (14, 26, 36). However, these workers focused mainly on succession events or addressed the issue of attachment and detachment of defined pure and mixed cultures on agar beads. In this study, we also examined the role of microcolonies during mixed-species biofilm development.

Mixed-species biofilm development.Because of its strong inhibitory activity against a broad range of surface-fouling organisms, we hypothesized that P. tunicata would exhibit an aggressive biofilm-forming lifestyle and would dominate other marine strains during competitive biofilm development. We therefore prepared mixed 1:1 inocula containing P. tunicata (GFP tagged) and a second organism (dsRed tagged) and allowed them to attach in a flow cell. In this experiment, P. tunicata totally outcompeted C. fucicola and Alteromonas sp. (Fig. 2). No cells of the latter competing organisms were visible after 24 h. This is consistent with our observation that P. tunicata produces inhibitory compounds. The results of the drop test assay (Table 2) showed that both C. fucicola and Alteromonas sp. are susceptible to inhibitors from P. tunicata. P. gracilis, which has a lower susceptibility to AlpP, was able to coexist with P. tunicata for up to 72 h before it was removed. The outcome of the competition may be correlated with the results of the drop test assay, which showed that P. gracilis is relatively insensitive to the antibacterial protein.

• Open in new tab
• Download powerpoint

FIG. 2.

Competitive biofilm development in two-species biofilms containing P. tunicata (expressing GFP) and other marine bacterial isolates (expressing RFP). Magnification, approximately ×600.

In contrast to the dominance of P. tunicata over C. fucicola and Alteromonas, P. tunicata was outcompeted by R. gallaeciensis. While some cells of P. tunicata persisted on the surface throughout the experimental period, these cells did not grow, and R. gallaeciensis rapidly grew and formed a biofilm. This is again consistent with the results of the drop test assay, which showed that P. tunicata is highly susceptible to the inhibitory compounds produced by R. gallaeciensis.

Role of the P. tunicata antibacterial protein AlpP in competitive biofilm development.To evaluate whether the antibacterial protein AlpP can play a role in competitive biofilm development, a ΔalpP insertional knockout mutant (39) was used in competition studies. We first repeated the mixed-inoculum biofilm experiments described above by using the ΔalpP strain instead of the wild type. In these experiments we observed no difference in the outcome of the biofilm competition experiments when the wild-type and ΔalpP strains were used. When inoculated together with C. fucicola, Alteromonas sp., or P. gracilis, the ΔalpP mutant became dominant after 72 h, similar to the results obtained for the wild-type strain (data not shown). Also, R. gallaeciensis was dominant over the P. tunicata ΔalpP strain in a manner similar to the manner observed with wild-type P. tunicata.

However, we found that AlpP can play an important role in the colonization of preestablished biofilms of other marine strains. As shown in Fig. 3, the ΔalpP strain had a greatly reduced ability to colonize 24-h-old biofilms of P. gracilis. Wild-type P. tunicata could become established within such a biofilm, form microcolonies, and completely remove P. gracilis in a manner similar to the manner observed for the mixed-inoculum biofilm shown in Fig. 2. In contrast, introduction of the ΔalpP strain led to establishment of a mixed-species biofilm containing separate microcolonies of each strain, but after 108 h the ΔalpP mutant strain was completely overtaken by P. gracilis (Fig. 3.). When the P. tunicata ΔalpP strain was introduced into preestablished biofilms of C. fucicola and Alteromonas sp., mixed-species biofilms were formed that persisted for over 9 days. In contrast, the wild-type strain was able to remove these competitors after 120 h (data not shown).

• Open in new tab
• Download powerpoint

FIG. 3.

A P. tunicata mutant that did not produce the antibacterial protein AlpP had reduced ability to compete against P. gracilis. The images show two-species biofilm development for P. tunicata ΔalpP (green) and P. gracilis (red) strains. After 96 h, microcolonies of P. tunicata and P. gracilis coexisted within the biofilm (compare with Fig. 2, which shows that wild-type P. tunicata could outcompete P. gracilis after 72 h). After 108 h, P. gracilis outcompeted the P. tunicata ΔalpP mutant strain. Magnification, ×600.

These results suggest that AlpP can provide a competitive advantage in certain ecological situations, such as the colonization of established biofilms. The production of inhibitory compounds by marine organisms may be a response to various ecological pressures in the environment. Epiphytic bacteria live in a highly competitive environment in which space and access to nutrients are limited. It is possible that P. tunicata up-regulates production of the antibacterial protein in the presence of competitors; in future studies we will examine AlpP expression in mixed-species biofilms by using a Gfp reporter protein fused to the alpP promoter. Some bacteria exhibit an inducible chemical antagonism when they are grown in the presence of competing marine bacteria (42, 54). It is also possible that the antibacterial protein is a key factor when there is competition between microcolonies. Evans and coworkers (20) compared protease production in planktonic and biofilm cells and found that the production was higher in the latter cells, suggesting that antimicrobial agents were important in biofilms. Furthermore, it has been shown that strains that are normally considered organisms that do not produce inhibitors can express inhibitory activity when they are growing as a biofilm (61). Moreover, there is compelling evidence that mature microcolonies may be the site of inhibitor production in P. tunicata. Production of inhibitors has been linked to pigmentation in P. tunicata (18), and we have observed mature microcolonies become pigmented in biofilms (data not shown).

Role of microcolonies in competitive interactions within biofilms.To evaluate the role of microcolonies in competition, microcolonies of Alteromonas sp., P. gracilis, and C. fucicola were allowed to preform before they were colonized by P. tunicata. Our data suggest that microcolony formation may enhance the ability of an organism to compete against P. tunicata and persist within a flow cell reactor. It was found that P. gracilis biofilms containing microcolonies could coexist with P. tunicata for more than 5 days, in contrast to biofilms without microcolonies (Fig. 4). Moreover, preestablished biofilms of C. fucicola and Alteromonas sp. (which normally persist for no more than 24 h in the presence of P. tunicata) were able to persist for up to 72 h before they were dispersed. Thus, high densities of cells within microcolonies may allow enhanced persistence during coculture with the superior competitor P. tunicata. This conclusion is supported by the outcome of competition studies conducted by other researchers. In two-species biofilms, bacteriocin-producing enteric bacteria prevented the colonization of a potential competitor in a preestablished biofilm (55). Al-Bakri et al. (1) also found that preestablished biofilms of P. aeruginosa conferred colonization resistance to Burkholderia cepacia.

• Open in new tab
• Download powerpoint

FIG. 4.

Microcolonies enhanced the coexistence of strains in competition with P. tunicata. Competing marine strains (red) were allowed to form microcolonies before introduction of the superior competitor P. tunicata (green) into the flow cell. Microcolonies were able to persist for more than 72 h in competition with P. tunicata. In contrast, in mixed-inoculum experiments P. tunicata completely removed inferior competitors after 72 h (see Fig. 2).

Conclusions.By labeling marine bacteria isolated from the green alga U. lactuca with genetic color tags, we were able to examine the colonization, biofilm formation behavior, and competitive interactions of the strains during mixed-species biofilm development in real time. This study is the first study to demonstrate that inhibitory compounds produced by marine bacteria can provide a competitive advantage during competitive growth within biofilms. We further suggest that microcolony formation may represent one adaptive strategy that increases the ability of bacteria to persist in mixed-species biofilms under competitive biofilm formation conditions.

Our studies showed that the AlpP protein, a potent antibacterial protein produced by P. tunicata, can enhance the ability of P. tunicata to colonize biofilms formed by other bacteria. The wild-type P. tunicata strain aggressively colonized and dominated strains that were sensitive to the AlpP protein. In contrast, a P. tunicata alpP mutant was defective in the ability to colonize and dominate biofilms under the same conditions. Moreover, marine strains that were tolerant to the AlpP protein in drop plate assays were recalcitrant to invasion and displacement by P. tunicata during biofilm development.

Previous studies performed in our laboratory have suggested that bacteria present on the surface of some higher marine eukaryotes may play an important role in the chemical defense against biofouling in the marine environment (30). Many members of the genus Pseudoalteromonas, for example, produce multiple extracellular inhibitory compounds that target different classes of marine fouling organisms (28). Competitive interactions between bacteria, such as those demonstrated in this study, may play an important role in determining the composition of such antifouling communities, because they likely lead to predominance of inhibitory bacteria on the host surface.

We also suggest that specific interactions between inhibitor-producing bacteria and eukaryotic host surfaces may enhance the antifouling defense of the host organism. For example, recent studies in our laboratory have revealed that P. tunicata possesses a mannose-sensitive type IV pilus that promotes attachment to the cellulose-containing surfaces of U. lactuca and Ciona intestinalis, the principal hosts of P. tunicata in the marine environment (D. Saludes, A. Scheffel, S. James, J. Webb, C. Holmstrom, and S. Kjelleberg, unpublished data). Both U. lactuca and C. intestinalis are thought to rely on inhibitor-producing bacteria for defense against biofouling; thus, it may be an advantage to these organisms to become associated with antifouling bacteria such as P. tunicata. The present study provides a platform for further studies of interactions between marine bacteria in surface-associated communities and between biofilms and their eukaryotic host surfaces.