Individual-based modeling for biofilm structure control
Date of Issue2017-04-10
Interdisciplinary Graduate School (IGS)
Singapore Centre for Environmental Life Sciences Engineering
Biofilm growth is affected by many factors such as substratum surface properties, fluidic conditions, and medium supply. In this thesis, we investigated biofilm structure development by considering the influence of three key factors: biofilm detachment, oxygen availability, and calcium concentration. This thesis mainly focused on the mathematical modeling of biofilms. Biofilm experiments were also conducted to validate the results of mathematical models. We first simulated the structure formation of Pseudomonas aeruginosa biofilms with three detachment mechanisms: shear detachment (SD), nutrient-limited detachment (NLD), and erosion detachment (ED). The simulation results agreed well with the reported data in terms of the effect of SD on producing smooth biofilms, NLD on hollowing the biofilms, and ED on isolating bacterial clusters. We also discovered that biofilm growth could achieve equilibrium when only SD was enabled. ED was important for biofilm structure formation throughout the simulation period, whereas SD and NLD demonstrated their importance only in late stages. In addition, the effect of SD on biofilm structure was less dependent on the detachment coefficients compared with that of NLD and ED. We conducted simulations and experiments to evaluate the effect of oxygen availability on the initial biofilm growth of Pseudomonas putida. Biofilms were cultured in a square chamber with controllable oxygen gradients. The oxygen distribution in the chamber was obtained from fluid simulation. Confocal images of six-hour biofilms were recorded at different oxygen concentrations, and biofilm surface coverage was higher with a higher oxygen concentration. A regression model was developed to optimize simulation parameters. Individual-based modeling of biofilms generated similar biofilm structures as experimental observations. In the simulated biofilms, oscillations of biofilm growth rates were characterized by surface coverage and cell number, and the periods of the oscillations were smaller with a higher oxygen concentration. We also investigated the effect of calcium on the biofilm structure formation of Shewanella oneidensis. A biofilm growth chamber, called the microfluidic biofilm reactor in this thesis, was developed for this purpose. Biofilm growth was monitored in the reactor under stable calcium gradients. The biovolume and surface coverage of biofilms were considerably greater in locations where calcium concentrations were higher. Based on the literature, calcium was hypothesized to affect biofilm detachment, EPS production, and bacterial surface motility. The hypotheses on calcium influence were tested by simulations, in which the results were comparable to those of experiments. When larger detachment coefficients and low calcium concentrations were applied, the simulated biofilms showed a slight decrease and large variations in biovolume and surface coverage in late stages, which were suspected to be caused by calcium deficiency as indicated by calcium profiles. This thesis contributes to the state-of-the-art in biofilm mathematical modeling. Unlike most of the existing biofilm mathematical modeling studies, simulations and experiments were compared in detail. Mathematical models were improved to simulate experimental observations and to test hypothesized mechanisms leading to those observations. The approaches proposed in this thesis demonstrates the potential of using a mathematical modeling approach to prioritizing the physicochemical factors according to their effect on biofilm structures.