Effect of material characteristics and/or surface topography on biofilm development
For figures, tables and references we refer the reader to the original paper.
Complexity of bacterial adhesion within the oro-pharynx
In the oro-pharyngeal areas, a dynamic equilibrium exists between the adhesion capacity of microorganisms and a variety of removal forces such as: swallowing, frictional removal by diet, tongue and oral hygiene, and the washing out by the salivary and crevicular flow. Most pathogenic organisms can only survive in the oro-pharynx when they firmly adhere to a non-shedding surface. The latter is clearly illustrated by the spontaneous disappearance of most pathogens after a full-mouth tooth extraction (Danser et al. 1994) and/or by the microbiological differences in peri-implant flora between partially and fully edentulous patients. In the latter, teeth and their periodontal pocket serve as a reservoir for pathogens (for a review, see Quirynen et al. 2002).
The initial bacterial adhesion to non-shedding surfaces, the first step in the formation of a biofilm after the formation of a conditioning film has been studied extensively over the past decades in many diverse areas, such as on solid surfaces in the oral cavity, on biomaterials implanted into the human body, on catheter surfaces, in water pipes, on ship hulls, and in the food industry. The adhesion process can be regarded either from a biochemical or from a physicochemical point of view.
The biochemical approach highlights the specific interaction between complementary surface components (the specific ligand–receptor interactions, Dalton & March 1998). Within the realm of cell–cell interaction, recent advances suggest that flagella, fimbriae, and other protein receptors are essential for bacterial attachment to surfaces. Gene expression changes and intra- and interspecies cell–cell communication further complicate the understanding of the process of adhesion and/or biofilm formation.
Other researchers emphasized the non-specific physicochemical mechanisms of bacterial adhesion (Bakker et al. 2004). They involve either a thermodynamic model (based on the interfacial free energies of liquids and interacting surfaces), or the Derjaguin, Landau, Verwey, Overbeek (DLVO) theory in which adhesion is regarded as the total sum of Lifshitz–Van der Waals, acid–base, and electrostatic interactions (the long-range forces). From a review by Bos et al. (1999), it became obvious that bacterial adhesion is unlikely ever to be captured in one generally valid mechanism.
The complexity of the oral flora with more than 500 species, together with the presence of a pellicle, a dynamic variety in availability of nutrients, temperature, and humidity, and the large variety of shear forces, renders a simple explanation of the adhesion processes nearly impossible. Advanced technologies recently provided novel insights into how dental plaque functions as a biofilm (Marsh 2004). Confocal microscopy confirmed that plaque has an open architecture with channels and voids. Within the biofilm, oral bacteria do not exist as independent entities but function as a coordinated, spatially organized, and fully metabolically integrated microbial community, whose properties are greater than the sum of the component species.
Owing to the tremendous complexity of the adhesion process and the fact that many aspects are still under investigation, this paper primarily aims to review the impact of surface characteristics on biofilm formation, especially from a clinical point of view.
Biofilm as a community
Bacteria within dental plaque do not exist as independent entities but rather function as a coordinated, metabolically integrated microbial community (Marsh & Bradshaw 1999; Marsh & Bowden 2000). This community life-style within dental plaque provides enormous potential benefits to the participating organisms (Caldwell et al. 1997; Shapiro 1998; Marsh & Bowden 2000; Marsh 2005) including: (i) a broader habitat range for growth (the metabolism of early colonizers alters the local environment, making conditions suitable for attachment and growth of later more fastidious species), (ii) an increased metabolic diversity and efficiency (molecules that are normally recalcitrant to catabolism by individual organisms can be broken down by the microbial consortia), and (iii) an enhanced resistance to environmental stress, antimicrobial agents, and the host defences.
Recent studies suggest that the environmental heterogeneity generated within biofilms promotes accelerated genotypic and phenotypic diversity that provides a form of ‘biological insurance’ that can safeguard the ‘microbial community’ (Boles et al. 2004). This diversity can affect several key properties of cells, including motility, nutritional requirements, secretion of products, detachment, and biofilm formation.
The pellicle coating
When microorganisms and substratum surfaces are in an aqueous environment, in which organic material is present (e.g. sea water, milk, tear fluid, urine, blood or saliva), they immediately (within seconds) become covered with a layer of adsorbed, organic molecules. This is commonly called ‘conditioning film’, simply because transport and adsorption of molecules to a substratum proceed relatively fast compared with that of microorganisms. This conditioning film in the oral cavity, called pellicle, consists of numerous components including glycoproteins (mucins), proline-rich proteins, phosphoproteins (e.g. statherin), histidine-rich proteins, enzymes (e.g. α-amylase), and other molecules that can function as adhesion sites for bacteria (receptors). The bacterial adhesion thus occurs between a with pellicle-coated bacterium and a with pellicle-coated surface.
The mechanisms involved in pellicle formation include electrostatic, van der Waals, and hydrophobic forces. Studies of early (2 h) pellicle on tooth enamel revealed that for example its amino acid composition differs from that of saliva. This clearly indicates that the pellicle is formed by a selective adsorption of environmental macromolecules. The physicochemical surface properties of a pellicle, including its composition, packing, density, and/or configuration, are largely dependent on the physical and chemical nature of the underlying hard surface (Lee et al. 1974; BaierGlantz 1978; de Jong et al. 1984; Fine et al. 1984; Ruan et al. 1986; Pratt-Terpstra et al. 1989, 1991; RykkeSonju 1991; Sipahi et al. 2001). Thus, the characteristics of the underlying hard surface are transferred through the pellicle layers, and as such will still have its influence on the initial bacterial adhesion. Absolom et al. (1987) even observed a clear relationship between the type of proteins adsorbed in the pellicle and the substratum surface free energy (SFE) of the underlying surface. For example, on polyethylene hydrophobicity gradients exposed to blood serum, less proteins were absorbed at the hydrophobic end with relatively more fibrinogen, while on the hydrophilic end more albumin was present (Rakhorst et al. 1999). Busscher et al (1995) also observed that the detachment of adhering bacteria could occur through a cohesive failure within the pellicle.
Specific biochemical mechanisms of bacterial adhesion
The bonding between bacteria and pellicle is mostly mediated by specific extracellular proteinaceous components (adhesins) of the organism and complementary receptors (i.e. proteins, glycoproteins, or polysaccharides) on the surface (e.g. the pellicle), and is species specific. These specific interactions are in fact non-specific forces acting on highly localized regions of the interacting surfaces over distances smaller than 5 nm. Most bacteria possess several specific mechanisms for adherence (Whittaker et al. 1996). Porphyromonasgingivalis strains possess the armament to coaggregate, to bind to saliva-coated hydroxyapatite, to haemagglutinate, and to adhere to and/or to invade epithelial cells. This bacterium also binds to several matrix molecules, like fibronectin, fibrinogen, and collagen, and produces proteases that may promote adherence. Streptococcus gordonii PK488 adheres to saliva-coated hydroxyapatite and coaggregates with Actinomycesnaeslundii PK606, other streptococci, and fusobacteria by different mechanisms. Mutants of strain PK488 that fail to coaggregate with PK606 retain the lactose-inhibitablecoaggregations with streptococci and the lactose-noninhibitablecoaggregations with fusobacteria, and bind to saliva-coated hydroxyapatite. Multiple adhesins on a given cell are likely to mediate distinct interactions with different surfaces, which can be animate or inanimate.
Streptococcal and actinomyces strain, the early colonizers, bind specific salivary molecules (Fachon-Kalweit et al. 1985; Fives-Taylor & Thompson 1985; Mergenhagen et al. 1987). Streptococci (especially Streptococcus sanguis), the principal early colonizers, bind to acidic proline-rich proteins and to other receptors in the pellicle like α-amylase and sialic acid (Hsu et al. 1994; Scannapieco et al. 1995). Actinomycesviscosus possesses fimbriae that containsadhesins that specifically bind to proline-rich proteins of the dental pellicle (Mergenhagen et al. 1987; Gibbons et al. 1988). Some molecules from the pellicle (e.g. proline-rich proteins) evidently undergo a conformational change when they adsorb to the tooth surface, so that new receptors become exposed. Indeed, for example A. viscosus recognizes cryptic segments of the proline-rich proteins that are only available in adsorbed molecules (Gibbons & Hay 1988; Gibbons et al. 1988). This provides a microorganism with a mechanism for efficiently attaching to teeth and also offers a molecular explanation for their sharp tropisms.
After the formation of a monolayer on the surface, biofilm formation can start, either by multiplication of adhering species and/or the adhesion of new species. From this stage on, new mechanisms are involved, because each newly accreted cell itself becomes a nascent surface and therefore may act as a coaggregation bridge to the next potentially accreting cell type that passes by. At least 18 genera from the oral cavity have shown some form of coaggregation (Kolenbrander et al. 1993). Essentially all oral bacteria possess surface molecules that foster some sort of cell-to-cell interaction (Kolenbrander & London 1993). This process occurs primarily through the highly specific stereo-chemical interaction of protein and carbohydrate molecules located on the bacterial cell surfaces in addition to the less specific interactions resulting from hydrophobic, electrostatic, and Van der Waals forces (Kolenbrander 1989; Kolenbrander & London 1993). Fusobacteriacoaggregate with all other human oral bacteria, while veillonellae, capnocytophagae, and prevotellae species preferably bind to streptococci and/or actinomyces (Kolenbrander & London 1993; Kolenbrander et al. 1995; Whittaker et al. 1996). Most coaggregations among strains of different genera are mediated by lectin-like adhesins and can be inhibited by lactose and other galactosides. The significance on biofilm formation of coaggregation in oral colonization has been documented in in vitro studies as well as in animal studies (Bradshaw et al. 1998). Secondary colonizers (e.g. Prevotellaintermedia, Prevotellaloescheii, Capnocytophaga spp., Fusobacteriumnucleatum, and P. gingivalis) do not initially colonize clean tooth surfaces but adhere to bacteria already in the plaque mass (Kolenbrander & London 1993).
In the saliva, each strain of early colonizers becomes quickly coated with distinct molecules. Identical cells therefore can agglutinate, which will lead to a microconcentration and juxta-positioning of a particular strain. Alternatively, growth of a particular accreted strain can lead to a microcolony, coated with specific salivary molecules.
Non-specific physicochemical mechanisms of bacterial adhesion
So far, no completely satisfactory picture has been proposed for the physicochemical mechanisms involved in bacterial adhesion to non-shedding surfaces (MorraCassinelli 1997; Bos et al. 1999). The following concept can help to understand most aspects of this adhesion process. In the formation of a biofilm to a non-shedding surface in an aqueous environment, such as the oral cavity, four well-defined stages (Fig. 1) have been described (BusscherWeerkamp 1987; Busscher et al. 1990; Van LoosdrechtZehnder 1990; Van Loosdrecht et al. 1990a, 1990b; Scheie 1994; Bos et al. 1999). The same steps apply to the marine environment, pipe lines, cardiovascular prostheses, air planes wings, etc.
Figure 1.
(a) Schematic representation of the dynamic plaque formation process as a four-stage sequence: (I) random transport of bacterium to the surface, (II) initial adhesion at secondary minimum (which often does not reach large negative values so that the adhesion is reversible), or directly at the primary minimum (with an irreversible binding) depending on the resultant of the van der Waals attractive force (GA) and the electrostatic repulsive force (GE), (III) attachment of bacterium to the surface by specific interactions after bridging the separation gap or after passing the energy barrier, (IV) colonization of the surface and biofilm formation (primarily by cell dividing and by bacterial intra- and/or intergenericcoaggregation). (b) Long-range interaction between a negatively charged bacterium and a negatively charged surface according to the Derjaguin, Landau, Verwey, Overbeek (DLVO) theory (Rutter & Vincent 1984). The Gibbs energy of interaction (GTOT) is calculated, in relation to the separation gap (D), as the summation of the van der Waals force (GA) and the electrostatic interaction (GE). Electrostatic interactions start when the electrical double layers overlap each other (see the upper part of the figure with T: solid surface (e.g. tooth) and C: bacterial cell). Adapted from Busscher et al. (1990), Quirynen & Bollen (1995), Van Loosdrecht et al. (1990b).
Phase 1: transport to the surface
The initial transport of a bacterium to the surface may occur through Brownian motion (average displacement of 40 μm/h), through sedimentation of the bacterium in the solution, through liquid flow (several orders of magnitude faster than diffusion), or through active bacterial movement (chemotactic activity). Alternatively, microorganisms in suspension may also be transported towards each other from microbial (co)aggregates.
Phase 2: initial adhesion
This stage results in a weak and reversible adhesion of the bacterium via its interaction with the surface at a certain distance (50 nm) through long- and short-range forces. The organisms will be attracted or repelled by the surface, depending on the resultant of the different non-specific interaction forces. Two physicochemical approaches, initially considered distinctly different, are available to describe microbial adhesive interactions: the thermodynamic and the DLVO approach.
- a.Thethermodynamic approach is based on the SFE of the interacting surfaces and does not include an explicit role for electrostatic interactions. Before a bacterium can come in direct contact with a surface, the water film between the interacting surfaces has to be removed. The interaction energy for this process can be calculated from the assumption that the interfaces between bacterium/liquid (bl) and surface/liquid (sl) are replaced by a surface/bacterium (sb) interface. The change in the interfacial excess Gibbs energy upon adhesion is described by the formula (Absolom et al. 1983; Bellon-Fontaine et al. 1990): ΔGadh=γsb−γsl−γbl in which the interfacial free energy of adhesion for bacteria (ΔGadh) is correlated with the surface–bacterium interfacial free energy (γsb), the surface–liquid interfacial free energy (γsl), and the bacterium–liquid interfacial free energy (γbl). If ΔGadh is negative (nature tends to minimize free energy), adhesion is thermodynamically favoured and will proceed spontaneously.
- b.Theclassical DLVO approach describes the interaction energies between surface and bacterium. When a bacterium approaches a surface, it will interact with that surface by means of two forces: the Lifshitz–van der Waals attractive forces (GA: the first force becoming active at distances even above 50 nm), and the electrostatic repulsive forces (GF: available at a closer distance). The latter force occurs due to the formation, in water, of a counter-charged layer, diffusely distributed around the particle, to neutralize the negative charge of the bacterium and of the surface (the electrical double layer or Stern layer, Figs 1a and b). When this double layer overlaps the double layer of the surface (the pellicle coating confers a negative charge to all surfaces), an electrostatic interaction will take place. As both surfaces have the same charge, this electrostatic interaction is repulsive in nature. The distance at which this interaction appears depends on the thickness of the double layers, which themselves depend on the ionic charge of the surface and the ionic concentration of the suspension medium.
DLVO have postulated that, above a separation distance of 1 nm, the summation of the above-mentioned two forces describes the total long-range interaction between bacterium and surface (Rutter & Vincent 1984). Figure 1b shows the total interaction energy (also called the total Gibbs energy (GTOT), as the result of this summation of the above-mentioned forces (GTOT=GA+GE), and in function of the distance between bacterium and surface. For most bacteria, GTOT consists of a secondary minimum (where a reversible binding takes place: 5–20 nm from the surface), a positive maximum (an energy barrier B to adhesion), and a steep primary minimum (located <2 nm away from the surface) where an irreversible adhesion is established. For bacteria in the mouth, the secondary minimum does not frequently reach large negative values (Van LoosdrechtZehnder 1990), which means a ‘weak’ reversible adhesion (defined as a deposition to a surface in which the bacterium continues to exhibit Brownian motion and can readily be removed from the surface by mild shear or by the bacterium's own mobility).
Both approaches have proven merits for microbial adhesion, when certain collections of strains and species are considered. They have, however, failed so far to yield a generalized description of all aspects of microbial adhesion valid for each and every strain (Van LoosdrechtZehnder 1990; Van Loosdrecht et al. 1990b). Van Oss et al. (1986) therefore introduced a so-called extended DLVO theory. This theory considers four fundamental, non-covalent interactions: Lifshitz–van der Waals, electrostatic, Lewis acid–base and Brownian motion forces. The acid–base interactions are based on electron-donating and electron-accepting interactions between polar moieties in aqueous solutions. The polar or acid–base interfacial free energy balance ΔGadhAB is incorporated into the extended DLVO approach by attributing a decay function to this balance. The influence of the acid–base interactions is enormous when compared with electrostatic and Lifshitz–van der Waals interactions. However, the acid–base interactions are also relatively short ranged, and a close approach between the interacting surfaces (less than 5 nm) is required before these forces can become operative. This new concept has been very useful for the prediction of bacterial adhesion in several in vitro experiments (Liu & Zhao 2005).
Phase 3: attachment
After contact is established between bacterium and surface, either directly or via bridging of the gap (fimbriae), a firm anchorage between bacteria and surface can be established by specific interactions (covalent, ionic, or hydrogen bonding). After adhesion, most organisms also start to secrete slime and embed themselves in a slime layer, the glycocalix, which forms an important virulence factor as it provides protection against humoral and cellular immune components.