Mechanical interactions between bacteria and hydrogels

Abstract

Mechanical interactions between bacterial cells and extracellular polymeric substance are essential in determining biofilm assembly and disassembly as well the mechanical characteristics of biofilms. However, the physics of these mechanical interactions in different cell culture conditions are poorly understood. We created typical artificial biofilm consisting of planktonic bacteria and hydrogel, in the absence of metabolic or regulatory effect. We have demonstrated that the cell culture medium can significantly affect the mechanical interactions between bacterial cells and hydrogels. The stiffness of the bacteria-hydrogel artificial biofilm cannot be simply attributed by the summation of the contribution from the bacteria and hydrogel based on the mathematical models and computational models. We have revealed that the tryptone component of Luria-Bertani broth medium plays an important role in stiffening effect of bacteria-hydrogel construct. Such significant stiffening effect can be explained by the following mechanism: the presence of tryptone in cell culture medium may enable the bacteria itself to crosslink the hydrogel polymer chains. Our findings have also demonstrated the synergy of modelling and innovative experiments which would potentially impact the biofilm control strategies.

Figure 1

Instantaneous elastic moduli of 1% agarose hydrogels without encapsulation obtained from Hooke’s law. The symbols on the plot * and ** indicate $p\le 0.05$ and $p\le 0.01$ respectively between two different gels (at 0.5% and 2% strains only PBS – LB gels showed significant differences and at 5% strain both PBS – LB gels and NB – LB gels showed significant differences). Data represented as mean ± standard deviation, $n\ge 5$.

Figure 2

Normalised stiffness values of different LB-based hydrogels when (a) 0.5% strain, (b) 2% strain and (c) 5% strain were applied (the bars represent the gels with bacteria, the dashed line shows the normalised value for gels without encapsulation and error bars represent standard deviation). This normalisation represents the fold change in the stiffness in hydrogels containing encapsulated cells compared with those without cells. The symbols on plots * and ** indicate $p\le 0.05$ and $p\le 0.01$ respectively between the gels made with the same liquid media/buffer with and without encapsulation of bacterial cells, determined from 3 independent experiments.

Figure 3

Normalised stiffness values obtained from mathematical models (group I - Voigt model, Reuss model, and Hashin and Strikman model with upper and lower bounds), simulations (group II) and experiments (group III) respectively for 5% applied strain. All the data were normalised by either the adapted gel stiffness, i.e. matrix without particles and separation (for simulations), or the corresponding 1% agarose gel stiffness with the same media/buffer (for experimental data). The dashed line represents the case without particles/encapsulated cells and the volume fraction of particles is 1%.

Figure 4

(a) E. coli and (b) S. epidermidis cell growth in different liquid growth media. Stationary phase E. coli cell elongation in 1% agarose hydrogels made with (c) LB, (d) NB, (e) LB-no yeast extract and (f) LB-no tryptone. Data points and error bars represent mean values and standard deviation, respectively. Number of replicates for each experiment is indicated at their individual sections. The images embedded in (c) illustrates how the bacteria elongation was determined: an ellipse was fitted around the identified edges of the bacteria and the major axis of the ellipse indicated the length of the bacteria. The same approach was used for all of the images captured at different times and the differences between the axis lengths indicated the bacteria growth at a particular time interval.

Figure 5

Representation of the model for the compression tests. Applied boundary conditions: symmetric boundary conditions on the matrix-particle structure and the planes provided us the opportunity to only model half of structure while restraining the linear movement in x direction and the rotational movement in y and z directions, fixed support applied at the bottom plane for the restriction of movement and a 5% strain was applied through the top plane. Mesh distribution of the matrix-particle structure: applied mesh is finer to correctly represent the matrix-particle interface and at the symmetry plane with an element size of 0.05, elsewhere the element size is 0.1. The element type used is CAX8R (An 8-node biquadratic axisymmetric quadrilateral with reduced integration).