Biofilm-control strategies based on enzymic disruption of the extracellular polymeric substance matrix – a modelling study

Joao B. Xavier1,2, Cristian Picioreanu2, Suriani Abdul Rani1, Mark C. M. van Loosdrecht2 and Philip S. Stewart1

1 Center for Biofilm Engineering and Department of Chemical and Biological Engineering, Montana State University-Bozeman, Bozeman, MT 59717-3980, USA
2 Department of Biotechnology, Delft University of Technology, Julianalaan 67, 2628 BC Delft, The Netherlands

Correspondence
Joao B. Xavier
J.Xavier{at}tnw.tudelft.nl


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
MODEL DESCRIPTION
RESULTS
DISCUSSION
REFERENCES
 
A kinetic model is proposed to assess the feasibility of strategies for the removal of biofilms by using substances that induce detachment by affecting the cohesiveness of the matrix of extracellular polymeric substances (EPSs). The model uses a two-state description of the EPS (natural EPS and compromised EPS) to provide a unified representation of diverse mechanisms of action of detachment-promoting agents (DPAs), which include enzymes that degrade the EPS and other agents described in the literature. A biofilm-cohesiveness factor describes local increases in detachment rates resultant from losses in cohesive strength. The kinetic model was implemented in an individual-based biofilm-modelling framework, including detachment rates dependent on local cohesiveness. The efficacy of treatments with DPAs was assessed by three-dimensional model simulations. Changes in treatment efficacy were evaluated quantitatively by using a Thiele modulus, which quantifies the relationship between diffusion of the DPA through the biofilm matrix and DPA decay rate, and a Damköhler number relating the rate of EPS reaction with a DPA and the rate of EPS production by the micro-organisms in the biofilm. This study demonstrates the feasibility and limits of implementing biofilm-control strategies based on attacking the EPS.


Abbreviations: DPA, detachment-promoting agent; EPS, extracellular polymeric substance; IbM, individual-based modelling


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
MODEL DESCRIPTION
RESULTS
DISCUSSION
REFERENCES
 
Biofilms forming on the surface of indwelling medical devices by organisms such as Staphylococcus epidermidis and Staphylococcus aureus constitute a leading cause of infections (Huebner & Goldmann, 1999). In industrial settings, unwanted biofilms are responsible for the fouling of cooling-water towers, water pipelines, membrane units or food-processing plants (Jass & Walker, 2000). Several strategies to remove unwanted biofilms exist that may be applied to a particular system, depending on its characteristics (Stewart et al., 2000). These include (i) mechanical cleaning, (ii) the use of antimicrobial agents, (iii) stopping biofilm growth by removing essential nutrients, (iv) inhibiting microbial attachment to a surface and (v) promoting biomass detachment. Mechanical cleaning and antimicrobial agents are the most-used methods. Mechanical cleaning can be costly, as it typically involves equipment down time or a significant labour expenditure. It may also not be applicable due to inaccessibility of the fouled surface. Still, toothbrushing remains the main method of removing dental plaque. Using biocides and disinfectants, in turn, may be ineffective due to the reduced susceptibility of micro-organisms in biofilms to antimicrobial agents [see Stewart et al. (2000) for a list of references]. Also, this strategy is insufficient when a clean surface, rather than one covered by an inactive biofilm, is required. Recent studies concerning induced detachment by depriving the biofilm of essential nutrients reveal the potential of this attractive strategy (Thormann et al., 2005), which, however, will not be applicable to cases where controlling the nutrients in the medium is not possible.

Promoting detachment is the least investigated of the possible strategies to remove unwanted biofilms (Stewart et al., 2000). The use of substances to induce biofilm removal directly by destroying the physical integrity of the biofilm matrix would be an attractive alternative for both medical and industrial applications where complete biofilm removal is essential. In industrial applications, this approach would also have the advantage of reducing reliance on inherently toxic antimicrobial agents, whose continued use is fundamentally at odds with the trend towards increasingly restrictive environmental regulations (Chen & Stewart, 2000).

Biofilms are primarily composed of bacteria, extracellular polymeric substances (EPSs) of microbial origin and other particulate substances. EPSs constitute a matrix embedding bacterial cells and almost certainly have essential roles in defining the cohesiveness and other physical properties of these attached microbial communities (Wingender et al., 1999). Biofilm EPSs are typically composed of diverse substances, including polysaccharides, proteins, nucleic acids, lipids and humic substances (e.g. Nielsen et al., 1996; Tsuneda et al., 2003). Substantial evidence exists supporting the role of polysaccharides in the cohesiveness of the EPS matrix (e.g. Boyd & Chakrabarty, 1995; Hughes et al., 1998). Multivalent cations such as calcium and magnesium also probably play a role in the cohesiveness of microbial aggregates, as evaluated from the study of anaerobic-sludge granules (Grotenhuis et al., 1991), activated-sludge flocs (Higgins & Novak, 1997) and biofilms (Chen & Stewart, 2002), by bridging negatively charged sites on extracellular polymers to create stable intermolecular and cell–EPS connections (Mayer et al., 1999).

Methods for promoting biofilm detachment by using chemical agents aimed at attacking the EPS have been investigated, so far with mixed results (Chen & Stewart, 2000). Table 1 lists several substances reported in the literature to affect the cohesiveness of EPS of bacterial origin. These substances, being enzymes, chelating agents or other agents, may reduce EPS cohesiveness through a variety of mechanisms. In the present article, all such substances that affect EPS cohesiveness and may potentially be used to promote biofilm detachment will generally be termed detachment-promoting agents (DPAs). Several enzymes have been observed to induce detachment by acting on the EPS matrix, by depolymerizing either polysaccharides (Allison et al., 1998; Boyd & Chakrabarty, 1994; Chen & Stewart, 2000; Itoh et al., 2005; Kaplan et al., 2004; Mayer et al., 1999) or extracellular DNA (Whitchurch et al., 2002). Chemicals that alter the ionic strength and composition of the liquid medium and affect electrostatic interactions involved in the cohesiveness of the EPS matrix, such as salts or chelating agents (Chen & Stewart, 2000; Mayer et al., 1999), may also be used as DPAs.


View this table:
[in this window]
[in a new window]
 
Table 1. Candidate substances found in the literature to be used as biofilm DPAs

 
Fig. 1 shows results from an experiment illustrating the use of chemical substances that affect biofilm cohesiveness to cause biofilm detachment. In this experiment, an S. epidermidis biofilm was cultivated in a capillary biofilm reactor for 24 h as described previously (Abdul Rani et al., 2005). The growth medium was then changed to a solution of 0·1 g NaOH l–1. The experiment was carried out without changing the flow velocity. Biofilm structures were static prior to the introduction of NaOH. Within 1 min of the introduction of dilute NaOH, biofilm cell clusters begin to deform in the direction of flow. This suggests that the chemical treatment weakened the mechanical properties of the biofilm. The reduction in biofilm cohesiveness resulted in the detachment of biofilm cell clusters in less than 4 min after the NaOH was added.



View larger version (95K):
[in this window]
[in a new window]
 
Fig. 1. Transmission-mode scanning laser microscopy of S. epidermidis biofilms grown for 24 h in a glass capillary tube. The image sequence, from (a) to (f), shows the deformation of a biofilm cluster and its consequent detachment after changing the influent medium to a solution with 0·1 g NaOH l–1. The time shown in each panel is the time elapsed from the change in medium to the NaOH solution. The flow velocity was kept constant during the experiment [flow direction is from top to bottom, as represented by the arrow in (a)]. Bar, 10 µm. A video of the experiment is also available at http://www.biofilms.bt.tudelft.nl/biofilmControl/index.html.

 
The feasibility of strategies for removing unwanted biofilms by using DPAs is evaluated here by using mathematical modelling. A generic methodology for accessing the effectiveness of a DPA treatment based on computer simulations is proposed. Parameters concerning biofilm cohesiveness and DPA action may be customized. The mathematical model provides a three-dimensional (3D) description of biofilm dynamics based on mass conservation of the reactions involved and uses individual-based modelling (IbM; Kreft & Wimpenny, 2001), implemented in a generic software framework introduced recently (Xavier et al., 2005a).

IbM was previously applied to study biofilm-structure dynamics in the presence of erosion forces in monospecies biofilms (Xavier et al., 2004, 2005b), EPS production (Kreft & Wimpenny, 2001; Xavier et al., 2005a, b) and multispecies biofilms (Picioreanu et al., 2004; Xavier et al., 2005a). Previous studies using other 2D- or 3D-modelling approaches addressed the effect of detachment caused by mechanical shear (Hermanowicz, 2001; Picioreanu et al., 2001), biomass decay (Pizarro et al., 2001), detachment induced by a chemical produced by the biofilm (Hunt et al., 2003) and starvation (Hunt et al., 2004). The present study is the first application of a multidimensional biofilm model for the evaluation of strategies for the removal of unwanted biofilms by treatment with chemical substances.


   MODEL DESCRIPTION
TOP
ABSTRACT
INTRODUCTION
MODEL DESCRIPTION
RESULTS
DISCUSSION
REFERENCES
 
The biofilm model used here is an extension of the framework reported previously (Xavier et al., 2005a, b). Only a brief introduction of the essential model elements and description of the extensions implemented will be presented here.

Biofilm system.
As in other biofilm models (e.g. Wanner & Gujer, 1986), the dynamics of two types of components are followed here. First, the soluble matter comprises any dissolved substance, such as substrates, products or DPA. Second, the particulate matter consists of solid matter, such as microbial cells and EPS. The computational volume used for the simulations represents a typical biofilm system composed of a biofilm (solid) phase and a liquid phase. Fig. 2 is a schematic representation of the system phases and geometry.



View larger version (56K):
[in this window]
[in a new window]
 
Fig. 2. Schematic representation of the system used in simulations, describing a computational volume of 1000x1000x1000 µm3. The biofilm phase, comprising active biomass and EPS, develops on a planar solid substratum. The liquid phase on top of the biofilm is divided into a bulk region of constant concentration of soluble species (an infinite reservoir) and a boundary layer concentrating the whole external resistance to mass transfer of soluble species.

 
Dynamics of soluble components.
Soluble matter exists both in the liquid and biofilm phases. The liquid phase is located above the biofilm and is, in turn, composed of a bulk liquid compartment and a concentration-boundary layer. The bulk liquid is assumed to be completely mixed, where solute concentrations are homogeneous and constant in time. The solute species diffuse to or from the biofilm through the concentration-boundary layer. In the biofilm phase, dynamics of solute species are governed by diffusion and reaction, as described by

{3817equ1}
here written for a growth-limiting substrate S. In equation (1), CS is the substrate concentration, DS is the diffusivity and rS the net reaction rate. Two solute species – a growth-limiting substrate (S) and a soluble DPA – are considered in the present model. The two boundary conditions for each diffusion/reaction equation of type (1) are (i) a specified concentration at the interface between boundary layer and bulk liquid and (ii) no-flux condition ({3817equ21}) at the substratum on which the biofilm grows.

Dynamics of particulate components
Representation.
Biofilm matter (biomass or particulate components) is represented in terms of its composition in active biomass (i.e. living micro-organisms) and EPS. For simplicity, it is assumed here that particulate matter exists only in the biofilm phase and not in the liquid phase. The IbM approach is used for the spreading of the biofilm matter as described previously (Xavier et al., 2005a). In the original framework, only one EPS type was allowed for each active biomass species. In this study, the framework is extended to allow EPS composition to include any number of particulate species. This extension permits implementation of two EPS states, ‘cohesive’ (natural EPS) and ‘compromised’ (EPS altered by DPA, denoted here as EPS*), with consequences that will be explained further. In IbM, biomass is represented by using spherical particles that act independently. These particles are entities with an internal state, defined in the present work in terms of their biomass composition (in active mass, EPS and EPS*), size and location in space. The size (volume, Vp) of a particle is related to its composition. Each of the biomass components, such as the active mass (MX), the EPS mass (MEPS) and the EPS* mass ({3817equ22}), accounts for a volume related to its specific mass ({rho}):

{3817equ2}
The radius of a spherical particle, Rp, is determined directly from that volume by using

{3817equ3}

Processes.
Biomass particles follow behaviour rules that mimic the behaviour of a microbial cell. They can: (i) grow by intake of nutrients; (ii) divide, creating an offspring agent; (iii) move (in continuous spatial coordinates) when pushed by neighbouring particles; and (iv) produce and excrete EPS. In addition, biomass particles can be removed from the biofilm body by any detachment mechanism. Growth of the spherical biomass particles occurs when active biomass or EPS is produced, the production of which is governed by equations (4) and (6), respectively. Moving follows the iterative procedure of the IbM, where the individuals (biomass particles) shove each other every time step to undo any overlap between neighbours generated by divisions and growth.

Kinetics.
The growth rate of active biomass (rX,prod) is governed by a Monod-type expression:

{3817equ4}
where µmax is the maximum specific growth rate of the micro-organisms, CS is the concentration of the growth limiting substrate, KS is the Monod half-saturation constant and CX is the concentration of active biomass. The consumption rate of substrate, rS,cons, is related to growth of active biomass through a yield coefficient YS, according to

{3817equ5}

The rate of EPS production (rEPS,prod) is assumed to be coupled to biomass growth and related to rX,prod by a yield coefficient YEPS, as described previously (Stewart, 1993):

{3817equ6}

The model proposed here also addresses the possibility of inactivation of the DPA. An example of an inactivation process is the decay of the DPA, applicable for instance in the case where the DPA is an enzyme. For simplicity, first-order decay rate will be used to model this case:

{3817equ7}

Stoichiometry and rates of reactions used by the current biofilm model are presented in Table 2.


View this table:
[in this window]
[in a new window]
 
Table 2. Stoichiometry and rates of reactions defined in the model system

Columns represent the species in the system, rows represent the reactions and entries in the table represent the stoichiometric coefficients (empty entries are effectively 0). The last column shows the expressions for reaction rates of each process.

 
Biomass detachment.
Detachment occurring from both erosion and sloughing was included following a methodology introduced previously (Xavier et al., 2005b). Briefly, the detachment model uses a detachment-speed function, Fdet, to calculate the local biomass erosion. Fdet(x) is the retraction speed of the biofilm–liquid interface at a point x located on the biofilm–liquid interface. The speed is defined in the direction normal to the surface and expressed by

{3817equ8}
where n(x) is the vector normal to the surface at point x. Point x is defined as x=(x,y,z), with x being the direction perpendicular to the planar substratum and y and z being the directions parallel to the substratum. The expression for Fdet is arbitrary, as it is possible to use a function of any state variable. Sloughing occurs whenever biomass clusters become disconnected from the biofilm body.

Introducing the influence of biofilm cohesiveness in the detachment expression.
Detachment occurs when external forces acting on the biofilm exceed the local biofilm cohesiveness. A biofilm cohesiveness factor, {chi}, is proposed here, so that the local detachment speed is inversely proportional to {chi}:

{3817equ9}

In equation (9), {chi}(x) refers to the local value of the biofilm cohesiveness at location x on the biofilm/liquid interface.

A detachment-speed function used previously for modelling biofilm development on a planar substratum (Xavier et al., 2004, 2005b) included a dependence on the square of the distance to the solid substratum, x2, and the local biofilm density, {rho}(x), as follows:

{3817equ10}

The value for the local biofilm density, {rho}(x), used in this expression is the total biomass (all types of EPS and active mass) per unit volume. It is noteworthy that this modelling framework allows for other types of detachment functions to be used, allowing the model to be customized for each particular system. The dependence on x2 was chosen because it is used commonly in biofilm studies (Stewart, 1993; Wanner & Gujer, 1986; Xavier et al., 2004, 2005b) and ensures the existence of a steady state (Stewart, 1993). The detachment-speed coefficient in equation (10), kdet, can be used to alter the effect of the external forces acting on the biofilm. Equation (10) may be interpreted as a balance between external forces acting on the biofilm (the product kdetx2) and its cohesive forces, here proportional to the local biofilm density, {rho}(x):

{3817equ11}

In the present study, equation (10) will be extended to include the role of EPS In the cohesiveness of the biofilm matrix. In order to encompass the generality of mechanisms for the action of DPAs, two states for EPS, the ‘cohesive state’ and the ‘compromised state’, are proposed. The cohesive state (represented by subscript EPS) refers to EPS in its normal state, which provides the biofilm with its natural characteristics. This is a general representation of the EPS that groups all its possible components (polysaccharides, proteins, etc.). The compromised state (represented by subscript EPS*) refers to a denatured state of the EPS after reacting with a DPA. Consequently, two EPS concentrations are distinguished, CEPS and {3817equ23}, as mass of EPS per biofilm volume.

The rate of EPS decomposition, rEPS,dec, will depend on the type of the DPA used. For example, if the DPA is an enzyme, Michaelis–Menten kinetics may be used:

{3817equ12}

In equation (12), kEPS,dec is a decomposition-rate coefficient, CDPA is the DPA concentration and KEPS is the Michaelis–Menten (or saturation) coefficient.

Conversion of EPS into EPS* will depend on the particular composition of the biofilm EPS. The following generic expression is proposed here:

{3817equ13}
where {3817equ24} is the yield of mass of EPS* produced per mass of EPS decomposed. In practice, {3817equ25} will have a value of 1 for DPAs that affect matrix cohesiveness without degrading the EPS. This may be the case for chelating agents and other substances that alter electrostatic interactions (and hence biofilm cohesiveness), but that do not decrease the mass of EPS in the biofilm. As the biofilm EPS may comprise a variety of polymers, enzymes that degrade only certain polymers in the EPS will generate a value between 0 and 1 for {3817equ26}. In these cases, a quantity ({3817equ27}) of soluble products from the EPS degradation would be formed. These products are assumed to have no subsequent influence on the system (i.e. are not biodegradable and do not react with any of the system's components) and therefore will not be included in the model formulation.

If we define the local fraction of EPS in its cohesive state, relative to the total EPS (cohesive and compromised states), as

{3817equ14}
then equation (10) can be extended to include the effect of the fEPS on the biofilm cohesiveness, as a power–law dependency:

{3817equ15}

This states that biofilm cohesiveness decreases with a decreasing fraction of cohesive EPS, fEPS. Any order ({gamma}) of the dependence on fEPS may be used, depending on the particular case being modelled. The order {gamma} will reflect the importance of the EPS components affected by the DPA for the biofilm cohesiveness. An expression that extends the detachment function (equation 10) is then derived from equation (15):

{3817equ16}
and includes the dependence on fEPS of the biofilm cohesiveness.

Simulations carried out.
Parameters for biomass growth and EPS production obtained from the literature were used in the simulations and are listed in Table 3. The effectiveness of biofilm removal using a DPA was evaluated by the time course of the simulated reduction of active mass in the biofilm. Several scenarios were evaluated, considering various characteristics of the DPA. A set of six variables defines each particular scenario. These variables are the diffusivity of the DPA (DDPA), the decay-rate coefficient of the DPA (kDPA,dec), the concentration of DPA used in the treatment (CDPA,treat), the decay-rate coefficient of the EPS (kEPS,dec), the yield ({3817equ28}) and the order of the EPS cohesiveness dependence on fEPS ({gamma}). The first two variables may be grouped, together with the biofilm thickness, Lf, in a dimensionless number, the Thiele modulus {pi}, defined as

{3817equ17}


View this table:
[in this window]
[in a new window]
 
Table 3. Parameters used in simulations

 
The value for Lf used in equation (17) is the maximum biofilm thickness at the time of application of the DPA treatment. This {pi} number thus relates the timescales of DPA diffusion and decay. kEPS,dec and CDPA,treat may be grouped in the dimensionless Damköhler number, defined as

{3817equ18}

This Da number relates the timescale of cohesive EPS decay to the timescale of EPS production by the micro-organisms. The remaining two variables from the set, {3817equ29} and {gamma}, are addressed independently. {3817equ30} was varied to represent scenarios in which the EPS mass is degraded at varying degrees. The effect of parameter {gamma} on biofilm removal was also assessed by carrying out simulations with values of {gamma} ranging from 0·2 to 10.

Animations concerning all of the simulations analysed in the present study were also produced. These animations may be obtained in the form of digital video files from our website at http://www.biofilms.bt.tudelft.nl/biofilmControl/index.html.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
MODEL DESCRIPTION
RESULTS
DISCUSSION
REFERENCES
 
Characterization of growth of undisturbed biofilms
The dynamics of an undisturbed biofilm (i.e. growing in the absence of any DPA) was first studied by simulating a period of 100 days growth. Four replicate simulations (U1, U2, U3 and U4) were carried out with the aims of (i) assessing reproducibility of the steady state and (ii) allowing statistically significant comparison with the results from the treatment-scenario simulations later. All simulations started from an initial state consisting of 3200 biomass particles placed randomly on the solid substratum surface. The difference between replicates is given by the initial position of biomass particles and by the sequence of random events in the shoving algorithm [see Xavier et al. (2004) for details]. Results of the replicate simulations showed the presence of a reproducible initial development stage in the first 10 days growth. After the occurrence of the first sloughing events at approximately t=12 days, structure dynamics are no longer quantitatively reproducible, similarly to what has been reported from experiments (Lewandowski et al., 2004) and other simulations (Xavier et al., 2004). Fig. 3 shows results from biofilm simulation U1. Observation of the computed biofilm dynamics throughout the entire simulated period allows the conclusion that the biofilm reached a ‘noisy’ steady state after 20 days growth. From this time, long-term biofilm accumulation is negligible (Fig. 3a), with time average biomass-production rate practically balancing the biomass losses by detachment (Table 4). The steady state of the undisturbed biofilm was further characterized by the occurrence of frequent sloughing events. Fig. 3(b, c) shows that erosion accounted for about 80 % of the total biomass losses, whereas sloughing accounted for only 20 %. Detachment by sloughing originated a distribution of sizes of detached particles, shown in Fig. 4, which was fitted to a normal distribution (also shown).



View larger version (29K):
[in this window]
[in a new window]
 
Fig. 3. Results from simulation U1 of undisturbed biofilm growth for 100 days (2400 h). (a) Time course of maximum biofilm thickness (Lf) and equivalent thickness of total biomass ({lambda}total), of active biomass ({lambda}X) and of EPS ({lambda}EPS) in the biofilm. Equivalent thickness of a biomass component i in the biofilm is defined by {3817equ46}, where Mi,j is the mass of component i in biomass particle j. The whole biofilm contains Np particles. (b) Sloughing events occurring in the course of biofilm development, shown as the mass of biofilm particles detached by sloughing. (c) Time course of the rate of biomass erosion from the biofilm.

 

View this table:
[in this window]
[in a new window]
 
Table 4. Mean biofilm properties at ‘steady state’ for the simulations of undisturbed development

Averaging was done between days 20 and 100.

 


View larger version (27K):
[in this window]
[in a new window]
 
Fig. 4. Distribution of biomass particle sizes sloughed during simulation of undisturbed biofilm. Particle count corresponds to the total of the four replicate simulations carried out. Detachment events are discriminated as either big (particles>0·01 µg, ~330 particles in total) or small (particles<=0·01 µg, more than 22 000 particles in total, considered as erosion); see text for details. Also shown is the normal distribution of big particles sloughed, showing a mean±SD value of 0·060±0·032 µg.

 
Simulation of treatment with a DPA
Several treatment scenarios were simulated by using different sets of parameter values. These simulations were all carried out using the same initial state, i.e. the biofilm from simulation U1 at day 60. Fig. 5 shows the results from a sample simulation, following a typical pattern observed for all simulations in which the biofilm was fully removed. The simulated conditions correspond to treatment with an enzymic DPA with the relatively low values of {pi}=3·6 and Da=0·71 and at {3817equ31} (total EPS conversion to EPS*). In this simulation, complete biofilm removal was achieved after 23·3 h DPA treatment. This case corresponds to a practical case where the DPA is an enzyme that diffuses rapidly through the biofilm (quantified by a low value of {pi}), catalysing EPS conversion to EPS* very quickly relative to the rate at which EPS is produced (quantified by the low Da value). Most removal, corresponding to 99 % of the active biomass, occurred in the relatively short period ranging from t=0 to 2·7 h (as seen in Fig. 5e). During the same period, large changes in the biofilm porosity [represented in Fig. 5(f)] were also observed. Biofilm finger-like structures [observable in the 3D representation in Fig. 5(a)] were destroyed first by the action of the DPA. As a result, a more homogeneous structure was formed temporarily [as seen in Fig. 5(b)] and the porosity of the biofilm, {varepsilon}, decreased until t=1·4 h. After 1·4 h, the porosity increased again ({varepsilon}->1) as the biomass at the solid substratum started to detach and colonization at the substratum became sparse (Fig. 5c). Detachment of the last 1 % of the micro-organisms required a comparatively long time, corresponding to 94 % of the total treatment time. Most of the detachment occurred from enhanced erosion and not from sloughing. Erosion rates [red line in Fig. 5(g)] at early treatment time (until t=2·4 h) are well above the mean for undisturbed biofilm growth (Fig. 3c). This enhanced erosion rate is a direct consequence of the higher fraction of the weak and detachable compromised EPS [see red areas in Fig. 5(b, c)] produced by DPA action. Sloughing during the treatment [blue line in Fig. 5(g)] accounted for <3 % of the total detached biomass (black line). No sloughing of big biofilm sections occurred.



View larger version (48K):
[in this window]
[in a new window]
 
Fig. 5. Results for simulation of the action of a DPA on a 60-day-old biofilm. Simulations were carried out at {pi}=3·6, Da=0·71 and {3817equ47}. (a–d) Biofilm structure after 0·1, 1·4, 2·2 and 2·5 h treatment with DPA. For each 3D structure, a 2D cross section through the biofilm is shown together with isoconcentration curves of DPA (normalized values). (e) Decrease in active biomass in the biofilm. (f) Change in biofilm porosity during the treatment. (g) Dynamics of biomass detachment (by erosion and sloughing) in the course of the treatment.

 
Treatment simulations were carried out for a range of values of {pi} and Da, while keeping all other parameters constant. A high value of {pi} means fast DPA decay relative to its diffusion rate (equation 17). A high value of Da means fast EPS degradation by DPA relative to the EPS production rate (equation 18). Therefore, a high Da number corresponds to high DPA concentration, high EPS-decay rate or slow EPS-production rate. Fig. 6 shows results obtained from the simulations, where the efficacy of a treatment is evaluated by the treatment time necessary for a tenfold decrease in biofilm active mass (T10 %). As expected, efficacy of the treatment (characterized by a lower T10 %) increases with increasing values of Da and decreasing values of {pi}. For cases where Da is low (production of EPS is much faster than decay due to DPA) and {pi} is high (degradation of DPA in biofilm is relatively fast compared with diffusion of DPA), the full removal of biofilm may not be accomplished, even for treatments with the DPA up to 100 h. In these conditions, DPA either decays too quickly or penetrates the biofilm too slowly and EPS either decays slowly or is produced quickly by the micro-organisms. Interestingly, for these cases, the model simulations predict the existence of persistent biofilm even in the presence of the DPA, with a steady state in which detachment is balanced by biofilm growth.



View larger version (44K):
[in this window]
[in a new window]
 
Fig. 6. Efficiency of biofilm removal with a DPA. The bar plot shows the treatment time necessary for a tenfold decrease in biofilm active mass (T10 %) for simulations carried out at a range of values of {pi} and Da. Shading of the bars represents the range of the T10 % for each simulation, with lower T10 % values corresponding to a more efficient treatment. Asterisks shown next to a bar represent cases where a tenfold decrease in the biofilm active mass was not achieved up to the end of the simulation at 100 h treatment. Simulations were carried out with {3817equ48} and {gamma}=10.

 
Treatment with DPA – effect of {3817equ32}
Values of {3817equ33} of <1 represent cases where the EPS degradation to EPS* by DPA leads to a reduction of the total EPS mass in the matrix. Fig. 7(a) presents the time course of the reduction of active mass in the biofilm for a series of simulations with {3817equ34} ranging from 1 to 0 ({pi}=3·6, Da=0·71). The results show decreased treatment effectiveness with decreasing {3817equ35}. Due to the practically complete decay of the EPS when {3817equ36} and 0·01, the treatment with DPA produces a more dense and confluent biofilm, as EPS is assumed to be less dense than active mass. As detachment speed is inversely proportional to the local biofilm density in the expression used (equation 16), the formation of a biofilm with less EPS generates a structure less susceptible to erosion forces. Consequently, it is observed from the time course of active mass that, after an initial reduction in biomass, the biofilm starts to grow despite the presence of the DPA. This derives from the fact that using a low value for {3817equ37} effectively represents cases where EPS does not play a significant role in the biofilm cohesiveness, i.e. in binding the biofilm together. This issue is central for concluding that a biofilm with a low content of EPS may in fact develop, which may be the case in Pseudomonas aeruginosa biofilms. For such biofilms, experimental results suggest that alginate may not play a role in the cohesiveness of the biofilm matrix (Christensen et al., 2001). Modelling the cases where the EPS plays a role in biofilm cohesiveness is represented by using higher values of {3817equ38}.



View larger version (24K):
[in this window]
[in a new window]
 
Fig. 7. (a) Time course of active mass in the biofilm in the presence of a DPA when the EPS matrix is decomposed in different degrees, simulated by four different {3817equ49} values: 0, 0·01, 0·1, 0·5 and 1. Simulations were executed with {pi}=3·6 and Da=0·71. (b) Time course of active mass in the biofilm for the case where the DPA is involved in a stoichiometric reaction with the EPS. Simulations were carried out with values for {pi}' of 3·6, 6·2, 11·3, 13·9 and 16·0, Da=0·71 and {3817equ50}. (c) Time course of active mass in the biofilm for simulations carried out at different orders of the dependence of the biofilm cohesiveness on fEPS, {gamma} from equations (15) and (16) (with {pi}=3·6 and Da=0·71). Simulations were performed with {gamma}=1 (indicated in plot), five simulations with {gamma}<1 (indicated in plot, with arrow representing series carried out with increasing values of {gamma}) and four simulations with {gamma}>1 (indicated in plot, with arrow representing series carried out with increasing values of {gamma}).

 
DPA involved in a stoichiometric reaction with the EPS
The theoretical approach proposed here can be further adapted to represent the case in which the DPA is consumed as a result of its involvement in a stoichiometric reaction. This corresponds, for example, to the case in which the DPA is a chelating agent that sequesters multivalent cations from the EPS structure. In this case, there is no degradation of the EPS to soluble material. All of the EPS mass will remain in the biofilm and only its material properties will be affected (i.e. EPS is converted to EPS*). Also, the DPA is inactivated by reacting with the EPS. The decay of DPA is no longer governed by equation (7). Instead, the following equation is used:

{3817equ19}
where YDPA is the yield of DPA consumed per decaying EPS. Table 5 contains the stoichiometry and reaction rates for the system adapted to account for DPA consumption in EPS degradation (cf. equation 19). For this case, the Thiele modulus for the DPA must be redefined as follows:

{3817equ20}


View this table:
[in this window]
[in a new window]
 
Table 5. Stoichiometry and rates of reactions adapted for the case where DPA is involved in a stoichiometric reaction with EPS

 
This new number, {pi}', groups the yield YDPA with kEPS,dec in a dimensionless group. Fig. 7(b) shows the time course of the decrease in active biomass in the biofilm for five simulations carried out at different values of {pi}'. These results show that, also for this case, the effectiveness of the treatment decreases with the increase in {pi}', thus with a less-diffusible DPA or with a slowly degradable EPS.

Effect of {gamma} on biofilm removal
Parameter {gamma} defines the order of the dependence of biofilm cohesiveness on the fraction fEPS through equation (15). The effect of {gamma} on the effectiveness of the DPA treatment was assessed by carrying out simulations with values of {gamma} ranging from 0·2 to 10. Fig. 7(c) shows the time course of the decrease in active biomass in the biofilm obtained from those simulations (carried out with {pi}=3·6 and Da=0·71). The results show that the predicted effectiveness of the biofilm treatment is highly dependent on the value of {gamma} used. Simulations carried out with {gamma}>1 predicted full biofilm removal within 100 h simulated treatment with a DPA. At {gamma}<=1, meaning a weaker dependency of biofilm cohesiveness on the EPS fraction, the simulations did not predict full removal within 100 h. Furthermore, all simulations performed with values of {gamma} lower than 1 predicted a subsequent boost in biofilm development after an initial decrease in biofilm active mass. Presently, there are no experimental data available on the literature that could be used to determine {gamma}. Possible experimental methods to measure {gamma} are proposed in the Discussion.


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
MODEL DESCRIPTION
RESULTS
DISCUSSION
REFERENCES
 
Model-based simulations of biofilm removal by treatment with a DPA provide a quantitative framework for evaluating the importance of mass-transport characteristics of the DPA and its reaction with biofilm EPS. The Thiele moduli {pi} and {pi}' quantify the relationship between DPA decay rate and its diffusion into the biofilm structure. The decay of EPS by reacting with a DPA, quantified by the Damköhler number (Da), is also highly important in defining the efficacy of a treatment. Simulations carried out at diverse values of {pi} and Da suggest that efficacy of a treatment decreases with decreasing {pi} and increasing Da (Fig. 6). That is, as one would logically expect, if the DPA decays too quickly or it diffuses across the biofilm too slowly, if the EPS decays slowly or if it is produced too quickly, then diminished biofilm removal is achieved. In those conditions, a thick biofilm with a high EPS content may persist in spite of the presence of a DPA.

The dual-state representation for EPS proposed here and the parameter {3817equ39} allow the description of several DPA action scenarios. Such scenarios range from cases where EPS mass is converted entirely to a non-cohesive substance, the compromised state, to cases where the EPS does not play a role in the biofilm cohesiveness. The latter case, represented by setting {3817equ40}, assumes that no conversion to a compromised state exists. Intermediate scenarios, where the solid EPS matrix is partially converted to EPS* (‘compromised’ EPS) and partially dissolved, are simulated by using values of {3817equ41} between 0 and 1. The simulations carried out at a range of {3817equ42} values (Fig. 7a) indicate that DPAs whose action is directed at affecting the biofilm cohesiveness (represented by values of {3817equ43} closer to 1) may be the most effective in removing the biofilm. DPAs whose action is directed at degrading EPSs that do not influence the biofilm cohesiveness (cases represented by using low {3817equ44} values) may not be a suitable strategy for removing the biofilm. Results from simulations for these cases predict that biofilms that are thin and dense and that have a high content of active microbial biomass will result from the treatment, rather than full removal of the biofilm. This prediction is purely theoretical and only valid if the assumptions presented in the model formulation are themselves valid. Nevertheless, the mechanism observed here may be a possible explanation for the experimentally observed failure of alginate lyase to remove P. aeruginosa biofilms when the same enzyme is able to degrade alginate in other settings (Christensen et al., 2001).

The parameter {gamma}, used here to define the order of the dependence of the biofilm cohesiveness on the fraction fEPS, was also observed to be highly important for the efficacy of a treatment scenario. This was revealed from the results shown in Fig. 7(c). The value of {gamma} represents the importance that the EPS component affected by the DPA has on the biofilm cohesiveness. {gamma}=1 indicates that biofilm cohesiveness is affected linearly as EPS is converted to EPS*. Values of {gamma}>1 may represent, for example, cases where the component affected by DPA is involved in several structural bonds in the EPS matrix. In such cases, biofilm cohesiveness could be affected non-linearly and with an order of fEPS higher than 1. Values of {gamma}<1 represent, in turn, the opposite scenario, where the affected EPS component does not play an important role in defining the EPS cohesiveness. Experimental values for parameters such as {gamma} are presently unavailable, although experimental techniques exist that may potentially be applied to obtain such information. The method proposed by Ohashi & Harada is based on applying both tensile force and shear force and was used successfully to measure adhesion strength of denitrifying biofilms (Ohashi & Harada, 1996) and biofilms grown under different conditions (Ohashi et al., 1999). A method developed recently that uses micro-cantilevers (Poppele & Hozalski, 2003) to measure the tensile strength of biofilms and microbial flocs could be used to determine the alteration in biofilm strength as a consequence of the presence of a DPA substance. Results from such experiments would allow derivation of a function for the alteration of the biofilm cohesive strength.

The study presented here comprises four simulations of undisturbed biofilm growth and 43 simulations of biofilm-treatment scenarios. Animated visualizations of biofilm-structure evolution during the treatment with DPA may be obtained for all simulations from our website at http://www.biofilms.bt.tudelft.nl/biofilmControl/index.html.

Of all of the simulations of biofilm treatment, however, only a limited number predicted a successful, complete biofilm removal. A common pattern was observed in all of the simulations predicting complete biofilm removal: an initial rapid removal of the largest part of the biofilm occurs within the first moments after introduction of the DPA in the medium. Following this initial period, removal of the remaining portion of the biofilm requires a relatively long period, still in the presence of the DPA. This last portion of the biofilm often consists of a thin and sparse layer of biofilm. The simulation results shown in Fig. 5 exemplify this observation, as analysed previously in the Results. From an analysis of these simulation results, we can conclude that the persistence of a last sparse layer of biomass results (i) from a lower detachment rate acting on thinner biofilms (because the detachment-speed function decreases as the biofilm thickness decreases) and (ii) from higher specific-growth and EPS-production rates for the thin biofilm (because of lower mass-transfer limitations and lower amounts of biomass sharing the substrate). The removal of the last fraction of the biofilm may constitute the majority of the time required for full biofilm removal, as was the case for the simulation shown in Fig. 5. For this simulation, as stated earlier, detachment of the last 1 % of the micro-organisms required 94 % of the total treatment time. This fact alerts to the necessity of a more precise knowledge of the detachment dynamics of the bacteria attached directly to the solid substratum for an accurate prediction of the efficiency of treatment using modelling techniques.

Other parameters related to the action of a DPA, i.e. reaction-rate coefficients, reaction yields and diffusivity coefficients, are also presently unavailable in the literature. Due to the absence of extensive information on the mechanisms of action of DPAs, the model presented here was kept as simple and generic as possible, but was still based on widely accepted principles. The framework used to construct the model allows even more detailed representations of the biomass. Biofilm models defined in this framework may include inert formation and multiple microbial species, as shown in previous studies (Xavier et al., 2005a, b). Introducing more detail into the biomass description here would have been undesirable, as it would create the necessity for more assumptions and complicate derivation of the trends reported. However, if more information is made available in the future from experimental characterization of biofilm-removal experiments, the model can easily be extended to accommodate new processes or biomass components. For example, it is expected that, for older biofilms, the formation of inert material will play a role in biofilm cohesiveness, as inert material will not grow or produce more EPS and its presence may affect the biofilm cohesiveness.

Conclusions
The modelling approach proposed here, consisting of a theoretical framework and its numeric implementation by using IbM, allowed the prediction of the efficacy of biofilm-control scenarios by using DPAs. The results of simulations carried out by using several sets of parameters produced the following conclusions.

The efficacy of using a DPA for removing a biofilm will be highly dependent on: (i) the balance between the mass-transport properties of the DPA through the biofilm matrix and its decay rate (quantified by Thiele moduli); (ii) the balance between the kinetics of action of the DPA on the EPS and EPS production by the micro-organisms [which may be quantified by the Damköhler number proposed in equation (18)]; and (iii) the mechanisms of influence of the DPA on the EPS, included in the model through the parameters {3817equ45} and {gamma}.

For cases where treatment with a DPA does result in complete removal of the biofilm, it is expected that the removal will follow a pattern where the majority of the biofilm biomass is removed in the initial instants of the treatment, but the removal of the last fraction of the biofilm will take a considerably larger fraction of the total treatment time.

The results from simulations presented here alert for the necessity to investigate experimentally the effect of DPAs on biofilm cohesiveness. This is essential for the accurate prediction of the efficiency of control strategies using modelling approaches.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
MODEL DESCRIPTION
RESULTS
DISCUSSION
REFERENCES
 
Abdul Rani, S., Pitts, B. & Stewart, P. S. (2005). Rapid diffusion of fluorescent tracers into Staphylococcus epidermidis biofilms visualized by time lapse microscopy. Antimicrob Agents Chemother 49, 728–732.[Abstract/Free Full Text]

Allison, D. G., Ruiz, B., SanJose, C., Jaspe, A. & Gilbert, P. (1998). Extracellular products as mediators of the formation and detachment of Pseudomonas fluorescens biofilms. FEMS Microbiol Lett 167, 179–184.[CrossRef][Medline]

Boyd, A. & Chakrabarty, A. M. (1994). Role of alginate lyase in cell detachment of Pseudomonas aeruginosa. Appl Environ Microbiol 60, 2355–2359.[Abstract]

Boyd, A. & Chakrabarty, A. M. (1995). Pseudomonas aeruginosa biofilms: role of the alginate exopolysaccharide. J Ind Microbiol Biotechnol 15, 162–168.

Chen, X. & Stewart, P. S. (2000). Biofilm removal caused by chemical treatments. Water Res 34, 4229–4233.[CrossRef]

Chen, X. & Stewart, P. S. (2002). Role of electrostatic interactions in cohesion of bacterial biofilms. Appl Microbiol Biotechnol 59, 718–720.[CrossRef][Medline]

Christensen, B. E., Ertesvag, H., Beyenal, H. & Lewandowski, Z. (2001). Resistance of biofilms containing alginate-producing bacteria to disintegration by an alginate degrading enzyme (AlgL). Biofouling 17, 203–210.

Cui, W., Winter, W. T., Tanenbaum, S. W. & Nakas, J. P. (1999). Purification and characterization of an intracellular carboxylesterase from Arthrobacter viscosus NRRL B-1973: physical and spectroscopic characterization and evaluation as models for cellulose triacetate. Enzyme Microb Technol 24, 200–208.[CrossRef]

Gordon, C. A., Hodges, N. A. & Marriott, C. (1991). Use of slime dispersants to promote antibiotic penetration through the extracellular polysaccharide of mucoid Pseudomonas aeruginosa. Antimicrob Agents Chemother 35, 1258–1260.[Medline]

Grotenhuis, J. T. C., van Lier, J. B., Plugge, C. M., Stams, A. J. M. & Zehnder, A. J. B. (1991). Effect of ethylene glycol-bis({beta}-aminoethyl ether)-N,N-tetraacetic acid (EGTA) on stability and activity of methanogenic granular sludge. Appl Microbiol Biotechnol 36, 109–114.[CrossRef]

Hermanowicz, S. W. (2001). A simple 2D biofilm model yields a variety of morphological features. Math Biosci 169, 1–14.[CrossRef][Medline]

Higgins, M. J. & Novak, J. T. (1997). The effect of cations on the settling and dewatering of activated sludges: laboratory results. Water Environ Res 69, 215–224.

Horn, H., Neu, T. R. & Wulkow, M. (2001). Modelling the structure and function of extracellular polymeric substances in biofilms with new numerical techniques. Water Sci Technol 43, 121–127.

Huebner, J. & Goldmann, D. A. (1999). Coagulase-negative staphylococci: role as pathogens. Annu Rev Med 50, 223–236.[CrossRef][Medline]

Hughes, K. A., Sutherland, I. W. & Jones, M. V. (1998). Biofilm susceptibility to bacteriophage attack: the role of phage-borne polysaccharide depolymerase. Microbiology 144, 3039–3047.[Medline]

Hunt, S. M., Hamilton, M. A., Sears, J. T., Harkin, G. & Reno, J. (2003). A computer investigation of chemically mediated detachment in bacterial biofilms. Microbiology 149, 1155–1163.[CrossRef][Medline]

Hunt, S. M., Werner, E. M., Huang, B., Hamilton, M. A. & Stewart, P. S. (2004). Hypothesis for the role of nutrient starvation in biofilm detachment. Appl Environ Microbiol 70, 7418–7425.[Abstract/Free Full Text]

Itoh, Y., Wang, X., Hinnebusch, B. J., Preston, J. F., III & Romeo, T. (2005). Depolymerization of {beta}-1,6-N-acetyl-D-glucosamine disrupts the integrity of diverse bacterial biofilms. J Bacteriol 187, 382–387.[Abstract/Free Full Text]

Jass, J. & Walker, J. T. (2000). Biofilms and biofouling. In Industrial Biofouling: Detection, Prevention and Control, pp. 1–12. Edited by J. T. Walker, S. Surman & J. Jass. New York: Wiley.

Johansen, C., Falholt, P. & Gram, L. (1997). Enzymatic removal and disinfection of bacterial biofilms. Appl Environ Microbiol 63, 3724–3728.[Abstract]

Kaplan, J. B., Ragunath, C., Ramasubbu, N. & Fine, D. H. (2003). Detachment of Actinobacillus actinomycetemcomitans biofilm cells by an endogenous {beta}-hexosaminidase activity. J Bacteriol 185, 4693–4698.[Abstract/Free Full Text]

Kaplan, J. B., Ragunath, C., Velliyagounder, K., Fine, D. H. & Ramasubbu, N. (2004). Enzymatic detachment of Staphylococcus epidermidis biofilms. Antimicrob Agents Chemother 48, 2633–2636.[Abstract/Free Full Text]

Kreft, J.-U. & Wimpenny, J. W. T. (2001). Effect of EPS on biofilm structure and function as revealed by an individual-based model of biofilm growth. Water Sci Technol 43, 135–141.

Lewandowski, Z., Beyenal, H. & Stookey, D. (2004). Reproducibility of biofilm processes and the meaning of steady state in biofilm reactors. Water Sci Technol 49, 359–364.

Mayer, C., Moritz, R., Kirschner, C., Borchard, W., Maibaum, R., Wingender, J. & Flemming, H.-C. (1999). The role of intermolecular interactions: studies on model systems for bacterial biofilms. Int J Biol Macromol 26, 3–16.[CrossRef][Medline]

Nielsen, P. H., Frølund, B. & Keiding, K. (1996). Changes in the composition of extracellular polymeric substances in activated sludge during anaerobic storage. Appl Microbiol Biotechnol 44, 823–830.[Medline]

Ohashi, A. & Harada, H. (1996). A novel concept for evaluation of biofilm adhesion strength by applying tensile force and shear force. Water Sci Technol 34, 201–211.

Ohashi, A., Koyama, T., Syutsubo, K. & Harada, H. (1999). A novel method for evaluation of biofilm tensile strength resisting erosion. Water Sci Technol 39, 261–268.

Picioreanu, C., van Loosdrecht, M. C. M. & Heijnen, J. J. (2001). Two-dimensional model of biofilm detachment caused by internal stress from liquid flow. Biotechnol Bioeng 72, 205–218.[CrossRef][Medline]

Picioreanu, C., Kreft, J.-U. & van Loosdrecht, M. C. M. (2004). Particle-based multidimensional multispecies biofilm model. Appl Environ Microbiol 70, 3024–3040.[Abstract/Free Full Text]

Pizarro, G., Griffeath, D. & Noguera, D. R. (2001). Quantitative cellular automaton model for biofilms. J Environ Eng 127, 782–789.

Poppele, E. H. & Hozalski, R. M. (2003). Micro-cantilever method for measuring the tensile strength of biofilms and microbial flocs. J Microbiol Methods 55, 607–615.[CrossRef][Medline]

Rittmann, B. E., Schwarz, A. O., Eberl, H. J., Morgenroth, E., Perez, J., van Loosdrecht, M. & Wanner, O. (2004). Results from the multi-species Benchmark Problem (BM3) using one-dimensional models. Water Sci Technol 49, 163–168.

Skillman, L. C., Sutherland, I. W. & Jones, M. V. (1999). The role of exopolysaccharides in dual species biofilm development. J Appl Microbiol 85 (Suppl. 1), 13S–18S.[CrossRef]

Staudt, C., Horn, H., Hempel, D. C. & Neu, T. R. (2004). Volumetric measurements of bacterial cells and extracellular polymeric substance glycoconjugates in biofilms. Biotechnol Bioeng 88, 585–592.[CrossRef][Medline]

Stewart, P. S. (1993). A model of biofilm detachment. Biotechnol Bioeng 41, 111–117.[CrossRef]

Stewart, P. S., McFeters, G. A. & Huang, C. T. (2000). Biofilm control by antimicrobial agents. In Biofilms II: Process Analysis and Applications, pp. 373–405. Edited by J. D. Bryers. New York: Wiley.

Sutherland, I. W. (2001). Biofilm exopolysaccharides: a strong and sticky framework. Microbiology 147, 3–9.[Medline]

Thormann, K. M., Saville, R. M., Shukla, S. & Spormann, A. M. (2005). Induction of rapid detachment in Shewanella oneidensis MR-1 biofilms. J Bacteriol 187, 1014–1021.[Abstract/Free Full Text]

Tsuneda, S., Aikawa, H., Hayashi, H., Yuasa, A. & Hirata, A. (2003). Extracellular polymeric substances responsible for bacterial adhesion onto solid surface. FEMS Microbiol Lett 223, 287–292.[CrossRef][Medline]

Turakhia, M. H., Cooksey, K. E. & Characklis, W. G. (1983). Influence of a calcium-specific chelant on biofilm removal. Appl Environ Microbiol 46, 1236–1238.

van Casteren, W. H. M., Dijkema, C., Schols, H. A., Beldman, G. & Voragen, A. G. J. (1998). Characterisation and modification of the exopolysaccharide produced by Lactococcus lactis subsp. cremoris B40. Carbohydr Polymers 37, 123–130.[CrossRef]

Wanner, O. & Gujer, W. (1986). A multispecies biofilm model. Biotechnol Bioeng 28, 314–328.[CrossRef]

Whitchurch, C. B., Tolker-Nielsen, T., Ragas, P. C. & Mattick, J. S. (2002). Extracellular DNA required for bacterial biofilm formation. Science 295, 1487.[Free Full Text]

Wingender, J., Neu, T. R. & Flemming, H.-C. (1999). What are bacterial extracellular polymeric substances? In Microbial Extracellular Polymeric Substances: Characterization, Structure and Function, pp. 1–20. Edited by J. Wingender, T. R. Neu & H.-C. Flemming. New York: Springer.

Xavier, J. B., Picioreanu, C. & van Loosdrecht, M. C. M. (2004). A modelling study of the activity and structure of biofilms in biological reactors. Biofilms 1, 377–391.[CrossRef]

Xavier, J. B., Picioreanu, C. & van Loosdrecht, M. C. M. (2005a). A framework for multidimensional modelling of activity and structure of multispecies biofilms. Environ Microbiol 7, 1085–1103.[CrossRef][Medline]

Xavier, J. B., Picioreanu, C. & van Loosdrecht, M. C. M. (2005b). A general description of detachment for multidimensional modelling of biofilms. Biotechnol Bioeng 91, 651–669.[CrossRef][Medline]

Xun, L., Mah, R. A. & Boone, D. R. (1990). Isolation and characterization of disaggregatase from Methanosarcina mazei LYC. Appl Environ Microbiol 56, 3693–3698.[Medline]

Received 29 April 2005; revised 29 July 2005; accepted 21 September 2005.



This Article
Abstract
Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Xavier, J. B.
Articles by Stewart, P. S.
PubMed
PubMed Citation
Articles by Xavier, J. B.
Articles by Stewart, P. S.
Agricola
Articles by Xavier, J. B.
Articles by Stewart, P. S.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
INT J SYST EVOL MICROBIOL MICROBIOLOGY J GEN VIROL
J MED MICROBIOL ALL SGM JOURNALS
Copyright © 2005 Society for General Microbiology.