Keywords

1 Introduction

Although much of the knowledge in microbiology is based on the study of free-floating planktonic cells, it is thought that the primary habitat for many microorganisms is biofilm. Bacterial biofilms are multicellular consortia in which the bacterial cells are embedded in an extracellular matrix or EPS that is often produced by themselves. Microbial mats in the natural environments or plaques that are developed on the surface of teeth are examples of surface-associated biofilms. Biofilms are not only necessarily attached to the surface but also are present as suspended cell aggregates, also known as flocs and granules (de Beer and Stoodley 2006). Natural biofilms can develop under most conditions were bacteria are present and their relevance in ecosystems has been recognized. Taking advantage of the multicellularity, biofilms are considered as a robust and sustainable system. Currently, biofilms are practically used for biological wastewater treatment. On the other hand, they are also associated with a wide range of problems in industry and also in clinical infections. For instance, biofilms have increased resistance against antimicrobial agents and is hard to eradicate them (Høiby et al. 2010). Given the potential benefits and drawbacks that biofilms can confer, there is high interest in their control.

Accumulating report demonstrates that biofilms are distinct from their free-floating counterparts (López et al. 2010). Cells in biofilms have different gene expression patterns compared to planktonic cells and numerous regulatory mechanisms take place during biofilm formation, suggesting that there is a unique program for biofilm development. Moreover, the gene expressions, phenotypes vary between the cells in a biofilm. This heterogenous nature of biofilm could be a consequence of differential gene expression that is initiated by a different gradient of nutrients, physicochemical factors effecting the cell activates. In addition, emergence of genotypically and phenotypically diverse mutants may lead to the heterogenous nature of biofilm. Remarkably, cells inside biofilms interact physically and chemically with each other to coordinate social behaviors. It is not surprising that the organization and differentiation of biofilm has led some scientists to compare biofilms with higher multicellular organisms.

In this chapter we will give a brief introduction to biofilms and discuss the feature of this multicellular bacterial community.

2 Sticking the Cells Together: The EPS Matrix

In order to form a multicellular community, the cells must have a way to connect to each other. EPS is one of the key components in biofilms and it could be simply said that “There is no biofilm without an EPS matrix” (Flemming and Wingender 2010). The EPS consists of different types of biopolymers which are polysaccharides, proteins, lipids, and nucleic acids. Also, polymers from other origin such as humic acid may be embedded into biofilms. It was long believed that polysaccharides are the main constituent of the EPS matrix; however, results from various studies indicate that the main EPS in terms of amounts and their impact on biofilm formation differs among species. The major role of EPS in biofilm is to connect the cells with the substratum or other cells and provide the mechanical stability of the biofilm. Therefore, the maintenance of the structured biofilm depends largely on the production of EPS. Several studies using pure culture have demonstrated that mutants lacking the component of EPS have altered biofilm morphology. For instance, in Pseudomonas aeruginosa, at least three types of exopolysaccharides that alter biofilm formation are identified. Psl is a mannose- and galactose-rich polysaccharide while Pel is a glucose-rich polysaccharide (Friedman and Kolter 2004a, b). Alginate is the major exopolysaccharide in mucoid strains that are the predominant morphotype of strains isolated from chronic infections (Govan and Deretic 1996); Psl and Pel exopolysaccharides are critical for biofilm formation on abiotic surfaces (Friedman and Kolter 2004a; Jackson et al. 2004; Matsukawa and Greenberg 2004). Further studies have demonstrated extracellular DNA (eDNA) to constitute an important component of the P. aeruginosa biofilm matrix (Whitchurch et al. 2002; Allesen-Holm et al. 2006). Although eDNA was initially thought to be a residual material from lysed cells, many groups indicate that it is in fact an integral part of the matrix. In Rhodovulum sp., flocculation was disrupted when treated with nucleic acid-degrading enzymes (Watanabe et al. 1998). Similar results are obtained with activated sludge that is composed of multispecies. DNase treatment of the activated sludge resulted in complete dissolution of the flocs in combination with shear force, whereas shear force alone did not (Dominiak et al. 2011). These results demonstrate that eDNA is related to structural integrity. In addition to the role in structural integrity, eDNA in bacteria may retain extracellular proteins as observed in the defensive web of neutrophil extracellular traps (NETs) (Brinkmann et al. 2004).

The presence of proteins in the biofilm matrix has been reported in many samples and they are sometimes even more abundant than polysaccharides in natural cell aggregates (McSwain et al. 2005; Frølund et al. 1996; Urbain et al. 1993). While the functions of these proteins are not fully understood, it is reported that these proteins at least include extracellular enzymes as well as structural proteins. Various extracellular enzymes that are involved in degradation of biopolymers have been detected in the matrix suggesting its role as an external digestion system in addition to the immobilization of the cells (Baty et al. 2001; Flemming and Wingender 2010).

Structural proteins include CdrA which is an extracellular protein in P. aeruginosa that is induced upon biofilm formation. This protein cross-links polysaccharides and/or tethers the cells with the polysaccharides providing biofilm stability (Borlee et al. 2010). Carbohydrate-binding proteins, called lectins, are also involved in biofilm formation and stabilization in a wide range of species. Lectins have a very specific recognition of sugars, and therefore, they may be involved in recognizing particular polysaccharides and glycoconjugates on the cell surface of other cells allowing them to coaggregate (Rickard et al. 2003).

Another type of proteins that seems to be ubiquitous among bacterial biofilms are amyloid fibers. Amyloid fibers are defined by their characteristic cross-β-strand structure, where the β-sheets are oriented perpendicular to the fiber axis (Sunde et al. 1997). Amyloids of bacterial origins have been found in various natural biofilms suggesting their importance in the natural habitat (Otzen and Nielsen 2008). In certain species, amyloid proteins such as curli fimbriae in Escherichia coli and Salmonella typhimurium, TasA in Bacillus subtilis have critical roles in biofilm formation together with the polysaccharides (Chapman et al. 2002; Romero et al. 2010; Zogaj et al. 2001).

A group of surface proteins that are named biofilm-associated proteins (Bap) also plays a role in the cell adhesion to biotic and/or abiotic surfaces in Gram-positive and Gram-negative bacteria. These proteins contain core regions of amino acid repeats involved in cellular adhesion (Latasa et al. 2006).

Recent proteomic studies reveal that in addition to these extracellular enzymes, intracellular proteins that are normally localized to the cells are present in the matrix at a certain amount. Examples include membrane proteins, histone-like proteins, ribosomal proteins, and superoxide dismutase (Jiao et al. 2011; Eboigbodin and Biggs 2008). While the functions of the intracellular proteins in the matrix remain to be fully investigated, some reports suggest a role of intracellular proteins outside of the cell. Such proteins that display two distinct functions are known as moonlighting proteins (Jeffery 2003). For example, the elongation factor Tu as well as the chaperonin protein complex GroEL have been identified on the cell surface of lactobacilli and were shown to be involved in adhesion to mucin and human epithelial cells (Bergonzelli et al. 2006; Granato et al. 2004). A recent study suggested that a large amount of these intracellular proteins are delivered to the matrix by membrane vesicles (Toyofuku et al. 2012a).

3 The Life Cycle of a Biofilm

Based on the knowledge of a limited number of model organisms, similarity of the biofilm life cycle has been recognized so far. From these observations mainly from mono-species biofilms, it is proposed that the biofilm life cycle consists of initial attachment, production of EPS matrix, maturation to complex colonies, and finally dispersal of cells from the biofilm. Accumulated researches indicate that these steps involve complex regulatory mechanisms that are influenced by environmental conditions and cell interactions. As a consequence, the structure of biofilms becomes highly variable ranging from flat-layered mats to complex mushroomlike structures depending on the nutritional conditions and cell interactions (Davies et al. 1998). Thus, cells undergo flexible and dynamic lifestyles in biofilms.

Biofilm formation begins with attachment to a surface that can be a substratum such as metals, glasses, or a surface of another cell. This step is important for the bacteria so that they settle down in the right environment. Initial attachment to abiotic surfaces is well characterized in the attachment of Pseudomonas sp. It was indicated that during the cell attachment, the cells first attach to the surface loosely via adhesins, such as flagellar and pili (O'Toole and Kolter 1998a, b), that may readily detach. Once the cells attach to the surface along their axis, this step is called irreversible attachment, the bacteria commit to their life as a biofilm (Petrova and Sauer 2012). Even this single step in biofilm formation is highly regulated by several factors indicating that biofilm formation is an active biological process, and as discussed later, an intracellular second messenger whose production is controlled by several environmental modulates this step.

While the initial attachment is the beginning of the life as a biofilm, the community comes to an end when cells disperse from the biofilm and seek new environments. The mechanism by which bacteria regulate dispersal is not fully understood and strategies to manipulate biofilm dispersal are of great interest for broad application in controlling biofilm formation. Passive processes for cell dispersal such as the erosion of biofilms by shear stress are observed; however, many bacteria undergo active dispersal processes where the cells convert from sessile to planktonic free-swimming bacteria (Sauer et al. 2002). Several environmental cues such as carbon sources, oxygen depletion, and nitric oxide induce cell dispersal, which allows the cells to respond to changes in environments. As observed in P. aeruginosa, not all cells but a subpopulation of cells localized at the center of mature biofilms disperse while the majority of the cells remain in the biofilm. These cells that remain seem to be dead cells which is also an indicative of the heterogenous nature of the cells in the biofilm (Barraud et al. 2006). In the dispersal cells, genes that are related to biofilm formation such as exopolysaccharides are downregulated, whereas genes that are related to motility are induced. In addition to active motility, the cells must disassemble from the EPS matrix. EPS-degrading enzymes such as chitinase and nuclease are also reported that could potentially be involved in the cell dispersion (McDougald et al. 2012).

Several signals that are produced by the cells are also involved in this dispersal event indicating that the cells themselves can initiate cell dispersal in a coordinated fashion. For example, a fatty acid molecule, cis-11-mehyl-2-dodecenoic acid which is known as a diffusible signal factor (DSF), was found to be involved in dispersal in Xanthomonas campestris (Dow et al. 2003). DSF in X. campestris activate the production of endo-β1, 4-mannanase that disrupts cell aggregates. Further work indicated that DSF reduce the biosynthesis of a putative exopolysaccharide, providing more the detailed mechanism of how DSF regulate dispersal (Tao et al. 2010b). DSF-related molecules are also found in other species implicating its role in interspecies signaling (Zhang and Dong 2004; Ryan and Dow 2011). For instance, DSF produced by Stenotrophomonas maltophilia induced stress-related genes in P. aeruginosa through a putative sensor kinase whose homologous protein is widespread among pseudomonads (Ryan et al. 2008). P. aeruginosa itself also produces a DSF-family signal, cis-2-decenoic acid, that induces biofilm dispersion of which the mechanism is not clear (Davies and Marques 2009). Interestingly, DSF-family signals play an interkingdom role where it is recognized by Candida albicans (Davies and Marques 2009; Boon et al. 2008).

Signals that induce dispersal are also found from B. subtilis biofilms called pellicles that are formed at the air–liquid surface. The cells are held together in the matrix that consists of exopolysaccharide and amyloid fiber which is largely composed of TasA. d-amino acids that are produced during the maturation of biofilm are incorporated into the cell wall peptidoglycan and trigger the release of amyloid fibers from the cell wall. A second biofilm-disassembly factor was found recently from B. subtilis that is a polyamine norspermidine. Norspermidine, which is produced in the late life cycle of the pellicle directly, interact directly with the exopolysaccharide component of the matrix (Kolodkin-Gal et al. 2012). Though the mechanism is not fully understood, both d-amino acids and norspermidine can induce cell dispersal in other bacteria, including Gram-negative cells.

4 c-di-GMP, An Intracellular Second Messenger, Regulates Biofilm Formation

The observations of the stage-specific gene regulation in biofilms indicate that biofilm development is a biologically programmed process. For instance, once the cells find a suitable surface, they should alter their gene expression so that they can settle down to switch their lifestyle from planktonic cells to biofilm cells. Furthermore, cells respond to environmental cues in dispersal. These phenomena observed universally among bacterial suggest that there is a common mechanism that is involved in these processes. Recent studies have found out that a small molecule, bis-(3’-5’)-cyclic dimeric guanosine monophosphate (c-di-GMP), is an (almost) ubiquitous signal that plays a central role in biofilm development of diverse bacteria (Hengge 2009; Römling et al. 2005). C-di-GMP is a second messenger that passes through the information outside the cells into the cytoplasm. In general, high levels of c-di-GMP concentration correlate with the biofilm mode of lifestyle and low concentrations of c-di-GMP correlate with planktonic mode of growth. In other words, a high level of c-di-GMP stimulates the biosynthesis of adhesins and extracellular matrix and inhibits motility, thus allowing the cells to form and maintain biofilms. In dispersal cells, c-di-GMP levels decrease that allow the cells to become motile. The level of c-di-GMP is controlled by the enzymes that produce (diguanylate cyclases) and degrade (phosphodiesterases). Diguanylate cyclases activity is associated with the GGDEF domain, which is named after the highly conserved amino acid sequences they contain, and c-di-GMP-specific phosphodiesterases activity is associated with the EAL or HD-GYP domains. These domains are often found with N-terminal sensory domains, compromising sensor domains as well as phosphoacceptor. The great variety of these sensory domains indicates that c-di-GMP level responds to numerous signals (Hengge 2009). The extracellular signal DSF is also one of the signals that alter c-di-GMP levels by degrading c-di-GMP (Tao et al. 2010a; Deng et al. 2012). Hence, a diverse range of signals are integrated to the c-di-GMP-signaling system enabling the cells to respond to changing environments.

Although the target of c-di-GMP is not fully understood, it has been indicated that c-di-GMP binds to the target component and alter the structure and functions. Several examples demonstrate that the binding components of c-di-GMP are highly diverse that can be proteins as well as riboswitches. For instance, c-di-GMP binding to PilZ proteins is an essential step in c-di-GMP-mediated regulation (Ryjenkov et al. 2006). In riboswitches, c-di-GMP binds to mRNA motifs which alter the expression of the gene (Sudarsan et al. 2008).

5 Bacterial Communications that Influence Biofilm Formation

Multicellularity in bacteria can be defined as a group of cells that function as a greater entity. In order to accomplish this, intercellular interactions must play a crucial role to enable cells coordinate their activities (Shapiro 1998). Although it is a still-opening question whether bacteria generally function for the whole in biofilms, scientists have discovered that bacteria can communicate with each other through chemical molecules in order to coordinate the cell behavior. Moreover, cell-to-cell communication has been demonstrated to affect biofilm formation in many species. A typical signaling molecule produced by Gram-negative bacteria is the N-acyl homoserine lactones (AHLs). Over 100 species of Proteobacteria use AHLs as signal molecules. AHLs usually consist of a fatty chain acid coupled to a homoserine lactone. There are variations in the acyl-group length, substitution at the C3 position (hydrogen, hydroxyl, or oxo group), and the saturation of the fatty acid chain. These variations confer signal specificity that is recognized by their cognate LuxR family transcriptional regulators. AHLs are synthesized by LuxI family proteins. The luxI gene is induced in the presence of AHLs, consisting an autoregulatory system.

While the fatty acid side chains were typically found in these homoserine lactone signals, recently, a new class of non-fatty-acid-based AHL signals has been discovered. The phototrophic soil bacterium Rhodopseudomonas palustris possesses a pair of LuxIR homologues, RpaIR. RpaI synthesize p-coumaroyl-homoserine lactone, from exogenous p-coumaric acid, which is thought to be obtained from the plant host. These aryl-homoserine lactone homoserine productions have also been observed in Bradyrhizobium sp. BTAi1 and Silicibacter pomeroyi DSS-3(Schaefer et al. 2008; Ahlgren et al. 2011).

P. aeruginosa is one of the most studied bacteria in cell-to-cell communication and biofilms. This opportunistic human pathogen produces at least two AHLs that are N-butyryl-HSL (C4-HSL) and N-(3-oxododecanoyl)-HSL (3-oxo-C12-HSL). Each signal is produced by RhlI and LasI, respectively. The organized structure of biofilms evoked researchers to think that there must be some interactions between cells when forming biofilms. The involvement of AHL on biofilm formation was first reported in 1998, when Davies et al. reported that a lasI mutant formed biofilms that were flat compared to the wild type that formed structured microcolonies (Davies et al. 1998). Since then, many groups have confirmed that the effect of cell-to-cell communication on biofilm formation in P. aeruginosa depends on strain and experimental conditions (de Kievit 2009).

The mechanism is not clear how cell-to-cell communication influences biofilm formation because AHLs control a wide range of genes. However, several factors that influence biofilm formation are known to be controlled by cell-to-cell communication (de Kievit 2009). These are biofilm matrix, rhamnolipids, and denitrification. In P. aeruginosa, at least three types of exopolysaccharide (alginate, Pel, and Psl) that are related to biofilm formation have been confirmed. Thus far, only the pel-gene cluster has been identified as being regulated by cell-to-cell communication. Pel is a glucose-rich exopolysaccharide which is essential for biofilm formation (Friedman and Kolter 2004a). It was discovered that the Las system is mainly involved in the transcription of pel biosynthetic genes (Sakuragi and Kolter 2007). Another important component of the extracellular matrix, extracellular DNA (eDNA) is also suggested to be influenced by QS. The mechanism by which the DNA is exported outside of the cell is not fully understood. Interestingly, spatial analysis of P. aeruginosa biofilms revealed distinct patterns of eDNA localization. In the early stages of biofilm formation, eDNA was shown to accumulate on the surface of microcolonies, while in older biofilms eDNA was shown to accumulate at the junction of the stalk and cap of the typical mushroomlike structure of microcolonies (Allesen-Holm et al. 2006). It was observed that QS-deficient mutants contain less eDNA than the wild type, suggesting that QS induce eDNA (Allesen-Holm et al. 2006).

Rhamnolipids are amphipathic glycolipids that act as biosurfactants. The expression of rhamnolipid synthesis operon rhlAB is regulated by the rhl QS system (Pearson et al. 1997). A mutant deficient in rhamnolipid production forms biofilms that are flat and relatively homogenous compared to the structured wild-type biofilms (Davey et al. 2003). From these observations, it is proposed that rhamnolipids are important in forming distinct architecture of the biofilm. In addition to maintaining the biofilm structure, rhamnolipids are necessary for detachment of the cells from the biofilm (Boles et al. 2005). The influence of rhamnolipids on multiple facets of biofilm formation suggests that the production is regulated at particular location and time points during biofilm formation. Supporting this idea, microscopic analysis using gfp-fused rhlAB genes indicated that rhamnolipid production is maximal in the stalks of a developing microcolony (Lequette and Greenberg 2005).

P. aeruginosa is capable of growing anaerobically by using N-oxides as terminal electron acceptors instead of oxygen (Toyofuku et al. 2012b). Under denitrifying conditions, P. aeruginosa forms robust biofilms (Yoon et al. 2002). Inactivation of Rhl QS leads to nitric oxide (NO) accumulation that is toxic to the cells causing cell death. Further study demonstrated that QS repress denitrification activity (Toyofuku et al. 2007, 2008). One of the key factors influencing biofilm formation is NO. While high concentration leads to cell toxicity, sublethal concentrations trigger disperse of the cells from the biofilm (Barraud et al. 2006). Interestingly, the process of biofilm formation under anaerobic condition seems somewhat different from aerobic condition, where cells become filamentous under anaerobic condition (Yawata et al. 2008). The cell elongation is one of the important factors in biofilm formation under anaerobic conditions and it is influenced by the production of NO (Yoon et al. 2011). Hence, fine tune of NO accumulation by QS as well as other factors appears to be important in biofilm formation under anaerobic conditions.

In Gram-positive bacteria amino acid peptides are widely used as signal molecules. Once the precursor signal protein is synthesized within the cell, active signals are generated by cleaving the proteins to short form of peptide chains and these peptides are subject to posttranslational modification in some bacteria resulting in a great diversity of the signal (Dunny and Leonard 1997). The peptide signals are linear or cyclic usually ranging in the size of 5–26 amino acid residues. Interestingly, some signal peptides also function as bacteriocins. These peptide signals that do not diffuse across the cell membrane are secreted by ABC transporters. Once secreted, the cells then sense the peptide signals by a two-component regulatory system. Most typically, the peptide signal is sensed at the surface of the bacteria by a histidine kinase sensor. This results in the autophosphorylation of the sensor kinase and the phosphoryl group gets transferred to a response regulator that regulates the expression of genes.

In the case of Staphylococcus aureus, the autoinducing peptide (AIP) is encoded by agrD. The AgrB protein processes and exports AIP. The secreted AIP is detected by a histidine kinase AgrC, which activates the response regulator AgrA by phosphorylation. Activated AgrA then modulates gene expression including the positive control of the agrBDCA operon. As expected, the agr QS system is involved in biofilm formation. In S. aureus, the agr QS system mutant forms thicker biofilm than the wild type (Vuong et al. 2000). Furthermore, extracellular proteins such as proteases and nucleases that are involved in dispersal of biofilm are upregulated by the QS system. Hence, the agr QS system inhibits biofilm formation.

In addition to AHLs and signal peptides, AI-2 is discovered broadly among Gram-negative and Gram-positive bacteria. AI-2 was originally described in Vibrio harveyi, where it was involved in the regulation of bioluminescence (Chen et al. 2002). AI-2 is produced by a spontaneous reaction of 4,5-dihydroxy-2,3-pentanedione which is synthesized by LuxS. The AI-2 signal which is recognized by homologues of the luxS gene has been identified in over 350 sequenced bacterial genomes (Shrout and Nerenberg 2012). Thus, AI-2 appears to be an interspecies signal. Whether LuxS is involved in biofilm formation or not is not clear. For instance, in Staphylococcus epidermidis, LuxS has negative effects on biofilm formation (Li et al. 2008), while this is not the case in S. aureus (Doherty et al. 2006). Taking into account that AI-2 is an interspecies signal, AI-2 may be involved in multispecies biofilm formation.

6 Multispecies Biofilm

In the natural environment, bacteria form complex multispecies biofilm. Compared to the mono-species biofilm, our understanding of multispecies biofilm is limited. The common feature of these bacterial communities is that normally the cells are dense and physically close to each other compared to the free-living planktonic cells. As a result of active metabolism of the cells in biofilms, nutrient in the liquid surrounding may be consumed by cells in the outer layers and metabolites may accumulate in the biofilms. These gradients create microenvironments that become niches for other bacteria. One of the techniques that are commonly used for observation of complex natural biofilm is fluorescence in situ hybridization (FISH). With this method, localization of bacterial populations can be visualized by hybridizing a fluorescently labeled oligonucleotide probe to the target 16S rRNA (Amann et al. 1990; DeLong et al. 1989). Compartmentalized or mixed structures of species within natural biofilms have been observed based on this method. The stratification of species within the biofilm can mostly be explained by the difference of metabolism among the species. For instance, in a nitrifying activated sludge, nitrifying bacteria are split into layers according to their use of different electron acceptors. During nitrification, ammonia is oxidized by ammonia-oxidizing bacteria and the produced nitrite is further oxidized to nitrate by nitrite-oxidizing bacteria. Consequently, ammonia-oxidizing bacteria locate at the outer layer and the nitrite-oxidizing bacteria colonize immediately below the ammonia oxidizers in nitrifying biofilms (Okabe et al. 1999). Compartmentalized is also explained by computational simulations revealing that these structures result from simple diffusion of growth-limiting nutrient (Nadell et al. 2009). Some model experiments using two bacterial species has also revealed that physicochemical conditions determine the localization of bacterial colonies within the biofilm. For example, Burkholderia xenovorans LB400 and Pseudomonas sp. B13 are able to degrade a pollutant, 3-chlorobiphenyl, to water and carbon dioxide. In this case, B. xenovorans LB400 degrades 3-chlorobiphenyl only to chlorinated benzoate and no further, while Pseudomonas sp. B13 further degrades chlorobenzoate to water and carbon dioxide. Since Pseudomonas sp. B13 cannot utilize 3-chlorobiphenyl, it is able to grow only if it is together with B. xenovorans LB400 when 3-chlorobiphenyl is the sole carbon/energy source. When a mixed-species biofilm was developed by these two strains using 3-chlorobiphenyl as the only available carbon source, Pseudomonas sp. B13 was always mixed with B. xenovorans LB400, whereas separate colonies were observed when another carbon source that can be used by both strains was given (Nielsen et al. 2000). In another case using Acinetobacter sp. C6 and Pseudomonas putida KT2400, distinct colonies but not mixed-species colonies were formed even under experimental settings that support host-commensal relationship (Hansen et al. 2007). When benzyl alcohol is given as a sole carbon source, the presence of P. putida KT2400 is dependent on Acinetobacter sp. C6. Acinetobacter sp. C6 metabolize benzyl alcohol to benzoate that is partly excreted and can be utilized by P. putida KT2400. Under this condition, Acinetobacter sp. C6 and P. putida KT2400 form distinct colonies where Acinetobacter sp. C6 is loosely surrounded by P. putida KT2400. In this case, competition of oxygen might counteract the host-commensal partnership of carbon source. Hence, the examples show us how the final structure could be affected by different interactions that occur among the consortia. Interestingly, extended incubation of the Acinetobacter sp. C6 and P. putida KT2400 mixed-biofilm under the host-commensal condition led to the emergence of a variant distinct cell from the ancestral type of P. putida KT2400 (Hansen et al. 2007). The variant was not detected in chemostat culture with Acinetobacter sp. C6 or in single-species biofilms, indicating that the variant can evolve in the microenvironment within biofilms in the presence of Acinetobacter. The evolved community was more stable and more productive than the ancestral community. Altogether, these examples demonstrate how spatial organization is important for the function of the community and further leads to the understanding of how a community is optimized in their natural habitat.

In addition to these chemical interactions, direct contact of the cells also plays an important part in mixed-species biofilms. Interspecies electron transfer is known to be important during anaerobic degradation of organic matter to methane. During this process, close physical contact between syntrophs and methanogens is required. Direct observation of granular sludge indicated that these microbes locate close to each other (Harmsen et al. 1996). In Pelotomaculum thermopropionicum and Methanothermobacter thermautotrophicus, such symbiosis is mediated by a flagellum (Shimoyama et al. 2009). P. thermopropionicum can ferment propionate to acetate, bicarbonate, and H2; however, this conversion is endergonic and the partial pressure of H2 should be kept low for this process to proceed. In the presence of M. thermautotrophicus, H2 is consumed to CH4 as an energy source and enable P. thermopropionicum to oxidize propionate. When grown together, these two strains form aggregates that are maintained by flagellum of P. thermopropionicum. The flagellum not only maintains the symbiosis but also synchronizes their metabolism by inducing gene expression involved in methanogenesis of M. thermautotrophicus. More recently, another type of bacterial communication that is based on physical contact was discovered. This communication is mediated by nanotubes that connect neighboring cells in Bacillus subtilis (Dubey and Ben-Yehuda 2011). These tubes were not synthesized during planktonic growth but were produced in colonies. These tubes allow bacteria to exchange small molecules including proteins and plasmids even between other species. Although the relevance of this nanotube in the natural environment is yet to be understood, the findings open up a new era of bacterial interaction between species.

Another bacterial-derived structure that potentially has a role in interspecies communication inside biofilms is the outer membrane vesicles (OMVs). OMVs are particles ranging from approximately 20 to 200 nm that are mainly produced in Gram-negative bacteria (Kulp and Kuehn 2010; Tashiro et al. 2012). These vesicles are mainly composed of outer membrane, periplasmic components and also contain DNA. In P. aeruginosa, a hydrophobic QS molecule, Pseudomonas quinolone signal (PQS) is incorporated in OMVs allowing the PQS to freely diffuse from the cell and shuttle them to the neighboring cells (Mashburn and Whiteley 2005). Consequently, PQS carried by OMVs coordinate social behaviors in P. aeruginosa. In addition, PQS are able to induce OMV production in P. aeruginosa and other species (Tashiro et al. 2010; Mashburn and Whiteley 2005). Importantly, OMVs have been observed in the EPS fraction of mono-species biofilms such as P. aeruginosa and Myxococcus xanthus and also mixed-species biofilm of environmental samples (Schooling and Beveridge 2006; Toyofuku et al. 2012a). M. xanthus is a bacterium that shows complex social behavior and biofilm (also known as fruiting body) formation. In this bacterium, protein can be exchanged through direct cell contact in which the precise mechanism is not completely understood. As a result of protein exchange, individual cells can endow their neighbors with new characteristics without altering the genome. Direct observation of M. xanthus biofilm demonstrated that the intercellular region is packed with OMVs that are tethering on the cell surface (Palsdottir et al. 2009). From these observations it is hypothesized that these OMVs may play a role in the protein exchange of M. xanthus (Remis et al. 2010; Konovalova et al. 2010). OMVs are also found extensively in the EPS of P. aeruginosa biofilm. It is well studied that OMVs in P. aeruginosa are associated with virulence against the host (Tashiro et al. 2012). Moreover, these OMVs derived from P. aeruginosa have the potential to fuse and/or attach to other Gram-negative and Gram-positive cells that may confer new properties to the recipient by transporting proteins or DNA. The role of OMVs in biofilms is yet to be investigated while some report indicates that they are involved in the development of mixed-species biofilm. OMVs produced by Porphyromonas gingivalis induce co-aggregation of S. aureus and Prevotella intermedia via an adhesin localized to the OMVs (Kamaguchi et al. 2003), indicating the role of OMVs in interspecies biofilm formation. Given that there are numerous proteins in the EPS, OMVs could also be contributing to place the proteins in the EPS.

7 Recently Developed Methods to Study Biofilms

Since biofilm mode of growth differs from planktonic cells, several methods to cultivate and analyze biofilms have been developed. One of the most popular method used in the assay of biofilm development is the 96-well biofilm assay (O'Toole et al. 1999). In this method, biofilm are developed in a static batch culture and the cells attached to the 96-well plate are stained after washing. This method is very powerful for screening for factors that affect biofilm formation and thus has been used in various species by many scientists. Biofilm structures have been revealed by microscopy techniques such as electron microscopy and the bacterial community structure with their localizations is often analyzed by FISH. FISH can be combined with other useful techniques, for instance, with combination of microautoradiography, uptake of different organic substrate can be monitored in order to demonstrate active cells (Lee et al. 1999; Kindaichi et al. 2004). The limitation of these techniques is that they only give a snapshot of the biofilm images while the biofilm development is highly dynamic. The use of time-lapse microscopy with combination of flow-cell system where the intact biofilm development can be monitored for several days has overcome this limitation for some bacteria. By using the flow-cell system, the lifestyle of biofilms became clear that it consists of several stages. Moreover, the use of confocal laser scanning microscopy (CLSM) with the flow-cell system allows the observation of three-dimensional structure of the biofilm during the time course for fluorescent-labeled bacteria. The fluorescent protein can be expressed constitutively and can be used for mono-species and mixed-species biofilms. In addition, fluorescent protein can be expressed under control of a particular promoter which gene is of interest in order to monitor gene expression in biofilms.

Recently, a technique adapting the confocal reflection microscopy (CRM) was used in biofilm. The advantage of this technique is that intact biofilms can be observed directly by confocal laser scanning microscopy without staining, fixation, or labeling with fluorescent proteins. Because CRM relies on the reflected light of the sample, reflection artifact obtained from bright spots such as cover glasses often becomes a problem. To solve this problem, the signal intensity of the cells can be maintained to a constant level along the Z-axis continuously by adjusting the detector gain manually while scanning. This modified mode of CRM called continuous-optimizing-CRM (COCRM) enables direct imaging of mixed-species biofilms and also mono-species biofilms alive during time courses (Yawata et al. 2010) (Fig. 17.1). While microscopy observation enables detailed analysis of the biofilm structure and development, yet there are few nondestructive methods that allow both microscopy analysis and measurement of metabolites. Measurement of metabolites along with the biofilm structure is important in understanding the activity of the biofilm and further understanding how biofilm reacts to the environment as a whole. To this end, flow-cell system was improved that gaseous metabolites can be collected and measured simultaneously with gas chromatograph (Yawata et al. 2008). Another method commonly used to monitor microbial activity is the use of microsensors. Microsensors are used for profiling spatial distributions of specific substrates and products (de Beer and Stoodley 2006). These microsensors can be mounted to chambers that cultivate biofilms and the concentration of certain metabolites can be monitored real-time. Since microsensors are small with a tip size of 1–20 μm it is feasible for small chambers and was adapted to a micro-fluidic device to monitor NH4 + consumption of activated sludge with its structure (Toda et al. 2011). Raman microspectrometer is also a powerful tool in chemically characterizing the biofilms noninvasively without any pretreatment of the sample. Biomolecules along with second metabolites can be identified with its space-resolved information, which will give us a more detailed understanding of the activity and heterogeneity of the cells (Venkata et al. 2011; Patzold et al. 2006). These new techniques with others will provide a new insight into the structure functional correlation of biofilms.

Fig. 17.1
figure 1

Development of a mixed-species biofilm monitored by COCRM

Full size image