Mechanical Comparison of Escherichia coli Biofilms with Altered Matrix Composition: A Study Combining Shear-Rheology and Microindentation

Abstract

The mechanical properties of bacterial biofilms depend on the composition and microstructure of their extracellular matrix (ECM), which constitutes a network of extracellular proteins and polysaccharide fibers. In particular, Escherichia coli macrocolony biofilms were suggested to present tissue-like elasticity due to a dense fiber network consisting of amyloid curli and phosphoethanolamine-modified cellulose (pEtN-cellulose). To understand the contribution of these two main ECM components to the emergent mechanical properties of E. coli biofilms, we performed shear-rheology and microindentation experiments on biofilms grown from E. coli strains that produce different ECM. We measured that biofilms containing curli fibers are stiffer in compression than curli-deficient biofilms. We further quantitatively demonstrate the crucial contribution of pEtN-cellulose, and especially of the pEtN modification, to the stiffness and structural stability of biofilms when associated with curli fibers. To compare the differences observed between the two methods, we also investigated how the structure and mechanical properties of biofilms with different ECM compositions are affected by the sample preparation method used for shear-rheology. We found that biofilm homogenization, used prior to shear-rheology, destroys the macroscale structure of the biofilm while the microscopic ECM architecture may remain intact. The resulting changes in biofilm mechanical properties highlight the respective advantages and limitations of the two complementary mechanical characterization techniques in the context of biofilm research. As such, our work does not only describe the role of the ECM on the mechanical properties of E. coli biofilms. It also informs the biofilm community on considering sample preparation when interpreting mechanical data of biofilm-based materials.

Keywords: E. coli; biofilm; extracellular matrix; mechanical properties.

1

ECM composition influences macrocolony biofilm spreading, texture, dry mass and water content. (A) Morphologies and textures of biofilms grown from strains with various ECM compositions (Table S1). The sketches at the bacteria scale are based on existing knowledge on the respective strains. , The images were acquired after 7 days of growth using a stereomicroscope in transmission. The darker lines result from the emergence of wrinkles in the third dimension. Textures are qualitative descriptions based on biofilm properties when scraped from the agar surface. (B–D) Biofilm diameter (n = 12 biofilms produced in 4 different plates), dry mass (N = 3–4 plates per strain) and density as a function of composition; median, quartiles and extreme values are represented by the line, the box limits and the whiskers respectively; Mann–Whitney U-tests were performed for statistical analyses. Statistical significance values are reported in Tables S2–4.

2

Oscillatory shear-rheology of biofilms with different ECM compositions. (A) Schematic image of sample preparation steps for bulk rheology measurements of homogenized biofilms. (B) Representative strain amplitude sweep (here for AR3110, Figure S2) used to extract the viscoelastic properties of homogenized biofilms (ω = 10 rad s^–1, γ = 0.01 to 100%). (C) Plateau of the storage modulus (G′₀). (D) Plateau of the loss modulus (G″₀). (E) Yield stress (τ_Y). (F) Flow stress (τ_F). (C–F) Bars represent median values and error bars represent standard deviation of bootstrapped data sets. A Mann–Whitney U test was used (*: p-value = 0.1, n = 3 measurements with biofilms grown on different plates but from the same bacteria suspension). More details are given in Methods section.

3

Mechanical characterization of macrocolony biofilms with microindentation. (A) Schematic image of the microindentation experiment performed on the native biofilms. (B) Representative force–displacement curves of biofilms with different ECM compositions, after the contact point is set at (0;0) for further analysis. (C) Reduced elastic moduli E _r. (D) Representative force-time curves of each strain during the holding time at a maximum penetration depth of 20 μm. (E) Portion of the force relaxation happening before the tip instability ΔF _fast/ΔF _total, derived from the relaxation curves. (F) Plasticity during the holding time, defined as the ratio of energy dissipated during relaxation and energy stored upon indentation. (G) Adhesion strength σ_Adh measured from the minimum force recorded during tip retraction divided by the contact area at maximum indentation. Data in E–G are derived from indentation curves with a maximum penetration depth of 20 μm. See Methods section and Figure S3 for technical details. Median, quartiles and extreme values are represented by the thick line, the box limits and the whiskers respectively; n = 10 curves from N = 4 different biofilms. Statistical analysis was performed with a Mann–Whitney U test and is reported in Tables S5–8.

4

Structure of macrocolony biofilms grown from the different strains. Confocal images of cross sections of native biofilms grown at 28 °C for 7 days on salt-free LB agar supplemented with Direct Red 23 to stain the ECM (red). For each strain, the images are representative of images acquired for 3 biofilms seeded on different agar plates from the same bacterial suspension. The images were acquired with 552 nm laser excitation and photon collection in the range from 600 to 700 nm. All images are displayed with the same brightness and contrast settings.

5

Influence of a homogenization step (mixing) on the structure and mechanics of selected macrocolony biofilms with different ECM composition. (A) Confocal images of native vs homogenized biofilms, formed by bacteria producing no ECM, only pEtN-cellulose, only curli or both ECM fibers. (B) Comparison of the reduced indentation elastic moduli E _r of native and homogenized biofilms with the shear modulus G obtained from rheology. Note that a quantitative comparison between E _r and G was not performed because of the unknown Poisson ratio. (C) Adhesion strength σ_Adh measured from the minimum load recorded during retraction. The values are normalized by the contact area at maximum indentation. Median, quartiles and extreme values are represented by the horizontal line, the box limits and the whiskers respectively; n = 7 curves from N = 3 different biofilms. Statistical analysis was performed using a Mann–Whitney U test and significance values are reported in Table S10.

6

Shear modulus vs reduced elastic modulus obtained on macrocolony biofilms with different ECM compositions. Each point represents one median value. The error bars show the 0.25 and 0.75 quantiles of the data distribution. (A) The data are compiled from the independent experiments reported in Figures and . (B) The data correspond to the comparative experiments performed with biofilms grown on the same plate (N = 3 plates, n = 3 biofilms per plate). The dashed lines represent theoretical relations between the shear modulus G and the reduced modulus E _r for ideal materials with different Poisson ratios.