OUP user menu

Ability of Pseudoalteromonas tunicata to colonize natural biofilms and its effect on microbial community structure

Dhana Rao , Torben Skovhus , Niina Tujula , Carola Holmström , Ingela Dahllöf , Jeremy S. Webb , Staffan Kjelleberg
DOI: http://dx.doi.org/10.1111/j.1574-6941.2010.00917.x 450-457 First published online: 1 September 2010


We investigated the effectiveness of surface colonization by the epiphytic marine bacterium Pseudoalteromonas tunicata firstly on a complex biofilm community on glass slides, and secondly, on the epiphytic community of Ulva australis. The effectiveness of P. tunicata was compared with the performance of Phaeobacter sp. 2.10, also a marine epiphytic isolate in the U. australis colonization experiments. Pseudoalteromonas tunicata cells were able to colonize the glass slide community at densities found naturally in the water column (9.7 × 104 cells mL−1). However, P. tunicata was a poor invader of the epiphytic community on U. australis at densities of 106 cells mL−1. At densities of 108 cells mL−1, P. tunicata again exerted little impact on the epiphytic community. Phaeobacter sp. 2.10 was also a poor invader at lower densities, but was able to invade and become dominant at densities of 108 cells mL−1. Differences in the ability of P. tunicata and Phaeobacter sp. 2.10 to invade natural communities may be due to differences in the antibacterial compounds produced by the two species. These experiments suggest that epiphytic communities may have protective effects compared with inanimate surfaces.

  • biofilm
  • invasion
  • Pseudoalteromonas tunicata
  • Phaeobacter
  • epiphytic community


In the marine environment, all surfaces are prone to colonization by microorganisms, leading to the formation of a potentially complex biofouling community. As microbial communities establish and mature, they are subjected to the challenge of colonization by planktonic cells. Whether or not an invading species can establish and grow within the existing community will depend on its ability to displace, compete or cooperate effectively with the community residents (Marsh & Bowden, 2000). Over time, the potential for invaders to colonize a biofilm changes as a function of the physiological and nutritional complexity of the community (McBain et al., 2000). The diversity of species within naturally occurring biofilm communities increases during establishment of the biofilms, until competitive interactions lead to the selection of a few successful partnerships, or a consortium of phenotypes (Jackson et al., 2001). Eventually, a mature biofilm community achieves a pseudosteady state, where sequestration of planktonic cells and the growth of attached cells are counterbalanced by losses through dispersal and cell death (Willcock et al., 1997). Once established, biofilm communities serve as a substrate for the attachment and growth of a range of other fouling organisms such as diatoms, invertebrates and algae. Attachment of macrofoulers adds further complexity to the community that continues to develop both through population growth within the biofilm and further recruitment from the surrounding seawater.

Biofouling communities also develop on living surfaces, with subsequent detrimental effects on the host organism (Wahl, 2008). However, the extent of biofouling on marine organisms is markedly less than that on inanimate surfaces and it is evident that host algae can regulate the bacterial colonization of their surfaces (Kjelleberg et al., 1997; Skovhus et al., 2007). The ecological significance of most epiphytic microbial communities is not clear, but it has been suggested that for the green alga Ulva australis, they may play a role in host defence against biofouling (Holmström et al., 1992; Egan et al., 2002). A number of inhibitory bacteria have been isolated from U. australis and the best studied of these are Pseudoalteromonas tunicata (Holmström et al., 1998) and Phaeobacter sp. 2.10 (formerly Roseobacter gallaeciensis) (Martens et al., 2006). Pseudoalteromonas tunicata produces a diverse range of biologically active compounds that specifically target marine fouling organisms (Holmström & Kjelleberg, 1999), and Phaeobacter sp. 2.10 displays strong antibacterial activity (Ruiz-Ponte et al., 1998; Brinkhoff et al., 2004; Hjelm et al., 2004). Both P. tunicata and Phaeobacter sp. 2.10 display antibacterial activity against monospecies biofilms under laboratory conditions (Rao et al., 2005, 2006), and monospecies biofilms of P. tunicata and Phaeobacter sp. 2.10 inhibit the settlement and attachment of fouling organisms (Rao et al., 2007). Colonization studies using high cell densities of P. tunicata and Phaeobacter sp. 2.10 on nonaxenic Ulva have been performed (Rao et al., 2006), but the effect of these two species on complex natural biofilm communities at low, ecologically relevant cell densities has not been explored previously. In fact, the integration potential of inhibitory marine bacteria into a complex biofilm on inanimate surfaces or on algae remains largely unstudied. The aim of this study was primarily to examine whether ecologically relevant densities of P. tunicata can change the composition of a natural seawater community on an inanimate surface. It has been shown previously that the numbers of P. tunicata cells on marine eukaryotic hosts are in the range of 0.06–1.5% (<1 × 103 cells cm−2) (Skovhus et al., 2004), with a relative abundance of 1% in seawater (Skovhus et al., 2007). Secondly, we aimed to determine whether P. tunicata is able to invade and impact an established epiphytic community on U. australis. For comparison, we also included the epiphytic bacterium Phaobacter sp. 2.10, which also exerts pronounced antibacterial activity (Ruiz-Ponte et al., 1998). For the first time, we studied the integration of inhibitory marine bacteria into the pre-existing complex community on U. australis.

Materials and methods

Establishment of a natural seawater community on glass slides

The entire experiments with glass slides were conducted in Aarhus, Denmark. Glass slides were arranged in open microscope slide storage boxes filled with seawater collected from Aarhus Bay. The storage boxes were then incubated at 14 °C with agitation and light intensities of 20 μE m−2 s−1, supplied over a 12-h photoperiod. After a 2-week incubation period, 12 slides were transferred to each of three seawater tanks containing 1200 mL of seawater. Green fluorescent protein (GFP)-labelled P. tunicata cells were prepared as described previously (Rao et al., 2005). Labelled cells were added at in situ densities (9.7 × 104 cells mL−1) to Tank 2 and at an approximately 1000-fold higher density (7.8 × 107 cells mL−1) to Tank 3. Tank 1 was a control with no labelled P. tunicata added. Slides were incubated for an additional week using the same conditions as described above. Of the slides in each tank, nine were used for DNA extraction (three slides were pooled for each DNA extraction).

Biofilm sampling and DNA extraction of the natural seawater community on glass surfaces

After incubation, the biofilm was removed from the slides using sterile razor blades and transferred directly into DNA extraction tubes. Slides were selected for DNA extractions using a table of random numbers. Total nucleic acid was extracted using the FastDNA Spin Kit for Soil (Qbiogene, Carlsbad, CA). The concentration of the extracted DNA was measured using a fluorometer (TD-700 fluorometer; Turner Design, Sunnyvale, CA) and PicoGreen (Molecular Probes). DNA extracts were diluted 1 : 10 in molecular biology-grade water (Sigma, St. Louis, MO) before DNA amplification.

Diversity of the natural seawater community on glass slides

PCR-denaturant gradient gel electrophoresis (DGGE) analysis was used to assess the taxonomic diversity of the biofilm community. PCR reaction was carried out using DGGE primers designed to be Pseudoalteromonas specific. The primers Eub341F-GC and Psalt815R and the amplification mixtures are as described by Skovhus et al. (2004). A second PCR reaction amplified eubacteria and the primers were Eub341F-GC and Univ907RC. PCR amplifications were thermocycled in a PT-200 Peltier thermal cycler (MJ Research, Reno, NV). The thermocycling conditions consisted of an initial denaturation at 93 °C for 60 s, followed by 25 cycles of 30-s denaturing at 92 °C, 60-s annealing at 57 °C and 45-s (increasing by 1 s per cycle) extension at 72 °C. The reaction was completed by a final 5-min extension step at 72 °C.

Amplified 16S rRNA gene fragments were analysed by DGGE using the D-GENE DGGE system (Bio-Rad, Hercules, CA) with an 8% acrylamide gel and a denaturating gradient ranging from 30% to 70% (100% denaturant defined as 7 M urea and 40% v/v formamide). Both the Pseudoaltermonas and the eubacterial DGGE analysis were run at 100 V at 60 °C for 15 h to provide maximum band separation. The gels were analysed using the Plot Profile tool in scion image software package (scion beta version 4.0.2; http://www.scioncorp.com) and scored for the presence and absence of specific DGGE bands. PCR-amplified P. tunicata DNA was used in the control DGGE ladder to verify the presence of P. tunicata in biofilm samples.

anova and the Student–Newman–Keuls post hoc test was performed using gmav5 developed at the Department of Marine Ecology, Sydney University, Australia. Homogeneity of variance was checked using Cochran's C test included in the program.

Epiphytic communities on U. australis

Experiments with epiphytic communities were conducted in Sydney, Australia. The common marine alga U. australis was collected from the rocky intertidal zone at Shark Point, Australia. Algal discs were excised from U. australis thallus tissue as described previously (Rao et al., 2006) and incubated in Saarsted 24-well plates on a shaker at 60 r.p.m. and 25 °C with a 16-h photoperiod at light intensities of 20 μE m−2 s−1. A longer photoperiod was used for the discs (compared with the glass slide experiments) to prevent them from becoming bleached. Although untreated algal tissue has a pre-existing biofilm community on its surface, excised discs were incubated in natural seawater for 6 days to allow further growth and development of this epiphytic community. The treatment resulted in epiphytic biofilms with a thickness of 10–15 μm.

Bacterial strains were isolated from the surface of U. australis as described previously (Rao et al., 2005). Wild-type P. tunicata, Phaeobacter sp. 2.10 and the P. tunicata AlpP mutant (Mai-Prochnow et al., 2004) and P. tunicata WmpR mutant (W3) (Egan et al., 2000) were used in the invasion experiments. The alpP mutant, which does not produce the antibacterial protein alpP, and the wmpR mutant, which is defective in the master regulator wmpR and does not produce any of the antifouling compounds, served as controls. All the strains used for colonization experiments were GFP labelled (Rao et al., 2005). Inoculae were prepared by culturing for 24 h at 25 °C in VNSS broth (Marden et al., 1985). Ulva australis tissue discs were inoculated with either 106 or 108 cells mL−1 in filtered seawater and incubated for 3 h without shaking at room temperature. Higher densities were used to inoculate discs as compared with glass slides as our earlier work indicated that at densities of <106 cells mL−1, P. tunicata did not attach to algal surfaces and the persistence of Phaeobacter sp. 2.10 was much lower (Rao et al., 2006).

Inoculated discs were rinsed three times with filtered seawater and transferred to sterile seawater in Saarsted 24-well plates. The discs were rinsed a further two times in filtered seawater and transferred to fresh Saarsted 24-well plates containing 2 mL of filtered seawater. The plates were incubated on a shaker at 60 r.p.m. and 25 °C with a 16-h photoperiod at light intensities of 20 μE m−2 s−1. Five tissue discs were randomly sampled 1, 2, 4 and 8 days following inoculation and each disc was examined by 12 random fields of view. The epiphytic community was visualized by staining with acridine orange and colonization and invasion of GFP-labelled P. tunicata and Phaeobacter sp. 2.10 within the community was examined using a confocal laser scanning microscope (Olympus).


Effects of P. tunicata on the biofilm diversity on glass surfaces

The community diversity on the glass slides increased when P. tunicata was added to the seawater. The overall species richness, as assessed by the number of distinct bands scored on DGGE gels for the eubacterial primer set (Fig. 1), was equivalent for the control (8–13 bands) and the treatment to which low densities of P. tunicata had been added (11–16 bands). In contrast, the richness was significantly higher (anova, SNK P<0.1) for the treatment to which high densities of P. tunicata had been added (16–24 bands). No bands matching the P. tunicata control in the DGGE ladder were observed in any of the treatments.


DGGE profiles of eubacterial communities in the three biofilm systems. The first system was a control with no labelled Pseudoalteromonas tunicata added. The second system had labelled cells added at in situ densities (9.7 × 104 cells mL−1) and the third system had labelled cells added at approximately 1000-fold higher density (7.8 × 107 cells mL−1). Profiles were obtained with the primers Eub341F-GC and Univ907RC. Numbers indicate the species richness in each lane. Asterisks indicate bands that could be found at the same location in all nine lanes.

When using the Pseudoalteromonas-specific primer set, the diversity of Pseudoalteromonas within treatments was more similar than the diversity observed between different treatments inoculated with different densities of P. tunicata (Fig. 2). The number of DGGE bands was the highest in samples from the treatment to which low densities of P. tunicata had been added (five bands) and the lowest in samples from the control and the treatment to which high densities of P. tunicata had been added (both with three bands). The intensity of the bands corresponding to the P. tunicata control in the DGGE ladder corresponds to differences in the cell densities of the treatments used. It was observed that P. tunicata cells were able to colonize an existing mixed community on glass slides at in situ densities. Increasing the density of P. tunicata cells 1000-fold resulted in a change in community diversity.


DGGE profiles of Pseudoalteromonas communities in the three biofilm systems. The first system was a control with no labelled Pseudoalteromonas tunicata added. The second system had labelled cells added at in situ densities (9.7 × 104 cells mL−1) and the third system had labelled cells added at approximately 1000-fold higher density (7.8 × 107 cells mL−1). Profiles were obtained with the primers Eub341F-GC and Psalt815R. Numbers indicate the species richness in each lane. The asterisk indicates that the band belongs to P. tunicata.

Invasion of an epiphytic community on U. australis by P. tunicata and Phaeobacter sp. 2.10

When P. tunicata was inoculated at densities of 106 cells mL−1, very few cells attached and persisted on the surface of U. australis thallus discs (data not shown). Even at densities of 108 cells mL−1, P. tunicata did not invade effectively and had no discernable effect on biofilm structure. There were no substantive differences among biofilm communities following invasion by wild-type P. tunicata (Fig. 3a), the alpP mutant (Fig. 3b) or the wmpR mutant (data not shown) nor did such communities differ from the control (Fig. 3d) treatments. Similarly, at low cell densities of 106 cells mL−1, Phaeobacter sp. 2.10 was a poor invader, but at high cell densities (108 cells mL−1), invasion was much more successful. Indeed, at high densities, Phaeobacter sp. 2.10 formed a confluent mat that eventually blanketed the epiphytic community, becoming dominant within 8 days (Fig. 3c).


Confocal microscopy images of biofilms on the surface of Ulva australis. Nonaxenic U. australis was allowed to develop an epiphytic community for 6 days before being inoculated with Pseudoalteromonas tunicata (a), P. tunicata AlpP mutant (b), Phaeobacter sp. 2.10 (c), control natural seawater community (d). Scale bar=5 μm.


Microbial communities are ubiquitous in nature, and it has been suggested that the environmental heterogeneity generated within biofilm communities provides a form of ‘biological insurance’ that can safeguard the microbial community under adverse conditions (Boles et al., 2004; Marsh, 2005; Boles & Singh, 2008). In this study, we compared the ability of inhibitory bacteria to invade existing natural and artificially generated biofilms and show for the first time that P. tunicata was able to colonize seawater-derived biofilms formed on glass surfaces, but was not effective in incorporating into natural epiphytic biofilms. Conversely, Phaeobacter sp. 2.10 was able to colonize and also dominate an epiphytic community. The evidence presented here suggests that compared with the natural seawater community on glass slides, the epiphytic bacterial community has protective benefits not present on inanimate surfaces.

Natural seawater community on glass surfaces

We observed that the addition of high numbers of P. tunicata cells to the natural seawater community on glass slides resulted in an increased species richness. This may be due to the release of inhibitory compounds (James et al., 1996), which may result in the removal of some dominant members of the microbial community, paving the way for an invasion by opportunistic bacteria leading to an increase in diversity. Likewise, earlier studies suggest that the expression of antibacterial activity is important for maintaining microbial activity and diversity within microhabitats (Defreitas & Fredrickson, 1978).

No increase in Pseudoalteromonas species richness, however, was observed for the treatment with high cell numbers of P. tunicata as compared with the control (Fig. 2). It has been shown that at high cell densities, P. tunicata is not only antagonistic to other marine bacteria but also displays antibacterial activity against members of the Pseudoalteromonas genus (Holmström et al., 2002). Moreover, the antibacterial protein AlpP produced by P. tunicata has autotoxic effects (James et al., 1996) and causes cell death and detachment in dense and mature P. tunicata biofilms (Mai-Prochnow et al., 2004).

Unexpectedly, there was an increase in diversity and Pseudoalteromonas species richness in the community that had P. tunicata added at in situ densities. The inhibitory activity of P. tunicata would be expected to decrease bacterial diversity, and this is borne out by an extensive study that was conducted to assess the abundance and diversity of Pseudoalteromonas strains in 11 types of marine samples from Danish coastal waters. Pseudoalteromonas tunicata was only detected on three unfouled eukaryotic hosts (Ciona intestinalis, Ulva lactuca and Ulvaria fusca), and Pseudoaltermonas diversity on these hosts was low (Skovhus et al., 2007). This suggests that the epiphytic community, together with the host organism itself, plays a role in the defence against colonizers.

When the converse experiments were conducted, where P. tunicata was allowed to establish biofilms on or glass slides, there was a slower return of the microbial community (data not shown). However, this was not necessarily due to the effect of bacteria producing inhibitory compounds, as mutant strains that do not produce inhibitory compounds also had the same effect. However, initially established strains appear to impair the ability of new colonizers to gain a foothold and establish. Once a new community is established eventually, it is no different from that of the controls that had no pre-established biofilms of marine bacteria (Rao, 2005).

Epiphytic community on U. australis

Pseudoalteromonas tunicata

Major differences in the ability of P. tunicata to colonize and invade a complex community were observed when inoculation was conducted on an epiphytic community on U. australis, as compared with a natural seawater community on glass slides. Pseudoalteromonas tunicata did not colonize at in situ cell densities, and even when inoculated at high densities to yield a final density of 108 cells mL−1, P. tunicata cells did not persist on the surface, and therefore had minimal effects on the epiphytic community. The lack of success supports the suggestion that the interaction between the epiphytic community and the host is actively involved in resisting changes to the epiphytic community towards some colonizers. One of the biologically active compounds produced by P. tunicata is an antibacterial protein that is effective against both gram-negative and gram-positive bacteria from a range of environments (Mai-Prochnow et al., 2004). The antibacterial protein does confer a competitive advantage to P. tunicata during biofilm growth in both flow cells and on algal surfaces. However, the protein does not seem to be effective against a mixture of bacteria and multispecies biofilms have been shown to resist the effects of P. tunicata (Burmolle et al., 2006).

A stable epiphytic community on U. australis may be more resistant to invasion as climax communities are better able to withstand environmental perturbations (Alexander, 1971). Studies on epiphytic communities on macroalgae show that although microbial populations change with the season or the age of the host (Sieburth & Tootle, 1981), they appear to form spatially structured, temporally stable communities (Longford et al., 2007). Despite some variation in the epiphytic community over space and time, a stable subpopulation has been shown to be consistently associated with the algal surface in U. australis (Tujula et al., 2010). In stable diverse communities, there is greater resource utilization and extensive networks of interactions among species, making invasion of the community more difficult (Wilsey & Polley, 2002). Beneficial interactions in multispecies biofilms derived from the epiphytic community of U. australis can protect one or several species from eradication when the biofilm is exposed to external stress factors (Burmolle et al., 2006). Thus, epiphytic bacteria may exploit the benefits of a community lifestyle to enhance resistance to extracellular compounds produced by inhibitory bacteria.

In contrast to inanimate surfaces, which are colonized in a rapid and predictable manner by a diverse assemblage of marine microorganisms, biotic surfaces frequently harbour species-specific microbial communities (Wahl, 1989). These communities are variable and distinct from those found in the surrounding environment (Hentschel et al., 2002; Lee & Qian, 2004; Taylor et al., 2004). Microbial community diversity is higher on living surfaces compared with inanimate surfaces (Dobretsov et al., 2005), suggesting that the seawater community on glass slides is more homogenous than the epiphytic community on Ulva.

Phaeobacter sp. 2.10

The resistance offered by a stable epiphytic community on U. australis against invasion by P. tunicata appears not to be an impediment to Phaeobacter sp. 2.10. Phaeobacter, which was able to invade and dominate the community at high densities, has been shown to be an aggressive colonizer of monospecies biofilms in flow cells (Rao et al., 2005) and on algal surfaces (Rao et al., 2006). This study indicates that it is also an effective colonizer of complex natural communities. Cell surface interactions are a defining feature of the marine Roseobacter group (Slightom & Buchan, 2009). Phaeobacter sp. 2.10 produces potent antibacterial compounds (Ruiz-Ponte et al., 1998; Brinkhoff et al., 2004; Hjelm et al., 2004) that kill or impair other species that might already occupy niches on the seaweed (Rao et al., 2006). The antibiotic has been identified as tropodithietic acid, and recent studies have revealed that much of the activity is due to plasmid-borne genes (Geng et al., 2008). Given the potential for high rates of genetic exchange, this might be a mechanism for the spread of active metabolites within biofilms (Egan et al., 2008).

The high densities required for successful integration into the community suggest that colonization may be regulated by quorum sensing. Phaeobacter sp. 2.10 has been shown to produce signalling molecules (Rao et al., 2007), but the traits controlled by quorum sensing in this organism are not known. Signalling molecules have been shown to exert not only intraspecies control but also interspecies control on the growth and expression of specific phenotypes, such as the interference of inhibitor production in other bacterial strains (Egland et al., 2004). In niches where bacteria are competing for nutrients in a mixed community, the production of signalling molecules may confer an advantage, as acyl-homoserine lactones have also been shown to function as antimicrobials in Pseudomonas aeruginosa (Kaufmann et al., 2005). Although the role of signalling molecules in Phaeobacter sp. 2.10 remains unclear, bacterial intercellular signalling has been shown to mediate parameters of complex microbial communities (Valle et al., 2004), and the success of members of the Roseobacter clade has been linked to the production of secondary metabolites and signalling molecules (Brinkhoff et al., 2008). Quorum sensing may be a significant factor that adds complexity to interactions in multispecies biofilm communities.


The physiology and metabolism of multispecies biofilm communities are immensely complex (Rickard et al., 2003) and the mechanisms that control microbial interactions in these biofilms are poorly understood (Komlos et al., 2005). Studying microbial communities in the laboratory may be an approach for identifying key attributes that allow dominant microorganisms to effectively compete and coexist in natural environments. An understanding of microbial interactions may enable better predictions of community behaviour in natural complex systems. Ultimately, this could serve as a basis for modulating interactions between biofilm residents, resulting in novel approaches for controlling biofilms. Identifying the selection pressures that favour certain interactions is also the key to gain an evolutionary understanding of microbial communities (Little et al., 2008).


  • Editor: Patricia Sobecky


View Abstract