Jonas Stenløkke Madsen:
Sociomicrobiology. Bacterial interactions and biofilm formation

Date: 01-03-2014    Supervisor: Søren Johannes Sørensen

Bacteria respond to and interact with their local environment by adjusting their behavioral output in order to optimize their own fitness. Behavioral plasticity is therefore key in bacterial adaptation and evolution. An important part of the microenvironment of a bacterial cell is other cells, and apt behavioral response via inter-bacterial interactions is therefore essential. Bacteria are single celled organisms but can function in multicellular communities where both coordinated and uncoordinated group behaviors can transpire, the latter being particularly important because many behaviors are futile if executed by individual bacteria. A bacterium therefore “make” a defining call in whether to be part of a multicellular community or act individually which has a huge implication for how bacteria adapt and evolve. Social interactions that trigger such as transition have been the main focus of this thesis. More specifically we have been studying social inter-bacteria interactions that induced biofilm behaviors. Bacteria can typically transition between a planktonic and a biofilms stage where the latter are multicellular communities that are typically sessile where members produce matrix compounds that encase and unify the community. 7 manuscripts have been produced during the duration of my PhD work. At the time of consigning this thesis, 3 manuscripts have been published, 2 are undergoing review, and 2 have been finalized for submission to peer reviewed scientific journals.

The research compiled in this PhD thesis focuses on bacterial interactions - a topic that relies on insights from a broad range of biological disciplines such as social evolutionary, population ecology, environmental microbiology and molecular biology. The introduction chapter is therefore organized to give a short background to highlight key concepts from the aforementioned topics and highlight their relevance to the work presented in the different manuscripts.

An initial goal was to develop and optimize a high throughput assay that would enable us to study biofilm formation in co-cultures in a reproducible manner while allowing us to estimate the number of each member of the biofilm individually. Manuscript 3 is the conclusion of this work. The optimized protocol was used in the majority of work presented in the other manuscripts.

Results compiled in manuscript 6 demonstrate that interactions between different bacteria (species/genotypes) can induce increased biofilm-formation in co-cultures (synergy) compared to when each strain is grown on its own. Biofilm synergy was found in 63% of four-species co-cultures while only 6% resulted in biofilm reduction. These results were based on 35 unique 4-member co-cultures mixed from seven different soil isolates. Especially strong biofilm synergy was found in co-cultures with members: Stenotrophomonas rhizophila, Xanthomonas retroflexus, Microbacterium oxydans, and Paenibacillus amylolyticus. Further analysis revealed that all members were essential for the strong biofilm synergy to transpire and that all 4 members were present in the actual biofilm. Cell numbers of each member in the planktonic and biofilm fraction were found using real-time qPCR and indicated that this biofilm synergy might be a cooperative interaction.

Findings reported in manuscript 6 raised the question of whether interspecific biofilm synergy is a common or rare event. Manuscript 5 was initiated mainly to approach this question. An important aspect of this query is whether interspecific biofilm induction is a response triggered between specific interacting partners or if it is a universal response among bacteria in general. Data presented in manuscript 5 support that biofilm synergy was common among the 1863 unique co-cultures, which were based on combinations of up to 8 members mixed from 45 different culturable bacteria, originating from various environments (freshwater, marine, two different soils, indoor and industry). Co-cultures with bacteria from the freshwater, marine and soil consortia were all dominated by biofilm synergy while indoor and industry consortia were dominated by biofilm reduction. Mixing bacteria originating from different environments resulted in a reduction in biofilm synergy suggesting that interspecific biofilm induction is an actor/recipient-specific event. Data also indicated that biofilm reduction was typically correlated with an induction in planktonic cells – a shift between a biofilm and planktonic state.

Exposing that biofilm synergy occurs, and does not seem to be an uncommon outcome of bacterial interactions, we wanted to study mechanisms that can enable intercellular biofilm induction. We decided to focus on horizontal gene transfer (HGT) as a trigger for interspecific biofilm induction because little is known about the connection between the two.

Manuscript 1 is a literature review that argues for and discusses the interconnection between biofilm formation and horizontal gene transfer. Here we hypothesize that fundamental differences among biofilm communities and planktonic populations may be important for how and at what frequency HGT occurs in addition to how mobile genetic elements evolve. We argue that theoretical considerations based on the cross field between biofilm biology and social evolution theory may be an important step to answer why certain genes are more frequently encoded on extra-chromosomal elements than chromosomes.

Manuscript 2 is a commentary paper based on a publication by Ma et al. (2013) and, like manuscript 1, discusses differences between HGT in biofilms and planktonic cultures. Based on data provided by Ma et al., we argue that their results may not be based on a difference in the frequency of actual plasmid loss events but based on competition (at the level of the cell), which is stronger in the biofilm due to spatial structure and steepened substrate gradients. More plasmid loss was found to occur in biofilms compared to planktonic cultures. This, we argue, is in line with our assumptions over a limited time period (laboratory relevant). Over a longer period of time (ecologically relevant time scale) it is, however, more likely that a plasmid will be sustained in a biofilm community than in a planktonic culture. This is because plasmid‐harboring cells at inner regions of the biofilm are less active and better protected from outside stress than highly active cells in suspension. This can potentially help explain why many plasmids persist in the environment without direct selection.

Manuscript 7 present experimental efforts to study some of the considerations described in manuscripts 1 and 2. In manuscript 7, we examine differences between type 3 fimbriae encoded on plasmids compared to chromosomes. Type 3 fimbriae enable biofilm formation on abiotic surfaces and are encoded by the mrkABCDF operon. We show that expression of type 3 fimbriae encoded on plasmids is not regulated in a c-di-GMP associated manner as is normally the case with type 3 fimbriae of chromosomes. The mrkABCDF promoter of plasmids is expressed constitutively and thus enforces an untraditional biofilm phenotype onto its host that does not influence its swimming motility. We illustrate that both the expression of type 3 fimbriae and swimming motility are factors that enhance the conjugative transfer frequency of mrkABCDF encoding plasmids, indicating social enforcement that promotes conditions for successful dispersal of the plasmid by horizontal transfer – a selfish trait of plasmids.

A pressing question when studying both social interactions and biofilm biology is whether or not the biofilm matrix is itself a shared resource. This we studied in the preparations for manuscript 4. The matrix is a defining feature of biofilms and is actively secreted by its members. Matrix substance is therefore a potential public good that could be exploited by non-producers. Using various P. aeruginosa PA14 mutants we illustrate that the biofilm matrix can not be exploited on by non-producers when electron acceptor availability is limited as we find that matrix producing cells gain a larger fitness advantage than cheaters, this we argue occurs because matrix producers will be positioned in the more saturated parts of the electron acceptor gradient. Most importantly, we demonstrate that clonal diversification is stimulated in biofilms because of spatial structure and heterogeneous electron acceptor availability.