Bursting out: linking changes in nanotopography and biomechanical properties of biofilm-forming Escherichia coli to the T4 lytic cycle

Abstract

The bacteriophage infection cycle has been extensively studied, yet little is known about the nanostructure and mechanical changes that lead to bacterial lysis. Here, atomic force microscopy was used to study in real time and in situ the impact of the canonical phage T4 on the nanotopography and biomechanics of irreversibly attached, biofilm-forming E. coli cells. The results show that in contrast to the lytic cycle in planktonic cells, which ends explosively, anchored cells that are in the process of forming a biofilm undergo a more gradual lysis, developing distinct nanoscale lesions (~300 nm in diameter) within the cell envelope. Furthermore, it is shown that the envelope rigidity and cell elasticity decrease (>50% and >40%, respectively) following T4 infection, a process likely linked to changes in the nanostructure of infected cells. These insights show that the well-established lytic pathway of planktonic cells may be significantly different from that of biofilm-forming cells. Elucidating the lysis paradigm of these cells may advance biofilm removal and phage therapeutics.

Fig. 1. The procedure used for E. coli attachment to AFM coupons.

Cells were attached to an AFM glass coupon via a positively charged lipid bilayer (LBL), which served as a rigid and adhesive support. a Fluorescence micrograph of a sparse E. coli cell coverage on a glass coupon without LBL coating. b Schematic illustration of the molecular interactions between E. coli cells and the glass coupon, pointing on repulsive electrostatic interactions together with weak van der Waals adhesion between the extracellular polysaccharide substances (EPS) and the surface. c LBL was prepared on an AFM glass coupon using vesicle fusion techniques, where lipid vesicles adhere to the surface, rupture, and form an LBL coating. d, e AFM images (in physiological solution) of positively charged LBL on an AFM glass coupon with a corresponding illustration of the LBL coating. Holes in the LBL allowed us to measure the layer thickness. f Fluorescence images of E. coli cells attached to a glass coupon that was precoated by a positively charged LBL. g Schematic illustrations that exhibit the electrostatic adhesion that attaches E. coli cells to the positively charged coated surface.

Fig. 2. Changes in the physiological properties of E. coli cells following phage infection.

a–c In situ fluorescence and d–h concomitant AFM images following the infection of biofilm-forming E. coli cells by T4 phages. i Cross-sectional analysis of the corresponding AFM nanotopography image (h). Epifluorescence images of “live” (green) and “dead” (red) cells, as well as the corresponding AFM scans, which were captured immediately after T4 infection (t < 10 min) as well as 60 min and 120 min following T4 addition. “Live” and “dead” bacteria were identified by staining the cells with SYTO 9 and PI, respectively (see also Supplementary Fig. 1). Complementary microtome slices of E. coli cells on a polycarbonate surface covered with LBL were captured by TEM before T4 addition as well as 10, 20, 40, and 90 min after T4 were added. Black circles within the TEM inserts (t = 40, 90 min) indicate the presence of T4 virions inside the cells. TEM scale bar was 1 µm (dashed line) for the low magnification and 0.4 µm (circles) for the larger magnifications. Similar images were captured from 12 other cells at 5 individual experiments.

Fig. 3. Three-dimensional, time-series images of an E. coli cell during T4 phage infection captured in situ by AFM.

a Images of uninfected E. coli cells, namely with no addition of T4. b The first image was captured before the addition of T4 phages, defined as t = 0. c–i Time series of AFM images that show the changes in nanotopography of the same cell during the infection cycle under hydrated conditions. Representative sections were enlarged to identify distinct nanostructures at different time points (b, d, e, f, h). Similar images were captured from 12 other cells at 5 individual experiments.

Fig. 4. Stiffness measurements of the E. coli cell envelope during T4 infection.

a, b Schematic illustration of AFM stiffness measurement using a pyramid AFM tip (not to scale). The force of the nanometer end-tip (F) that was applied caused an indentation (δ) of the cell envelope, including the outer/inner membrane (OM/IM) and the peptidoglycan layer (PG). c A representative AFM force vs. indentation that demonstrates how the stiffness was calculated. Approach velocity was 190 ± 20 µm/s. d Stiffness (rigidity) changes with time for different bacteria with (box plot with red circles indicating the average) and without (empty circles) the addition of T4 phages. Average stiffness and the corresponding standard deviation (s.d.) were based on 45,000 force measurements taken from 4 to 5 areas on the surface of each bacterium. Asterisks indicate a statistically significant difference between two time points (p < 0.01, n = 12 at 3–5 independent experiments).

Fig. 5. AFM measurements of applied force, F, and cell deformation, h, vs. time of applied force for infected and uninfected (carried out in separate experiments with no addition of T4 phages) E. coli cells.

Two hours after the addition of T4 phages, infected E. coli were identified and analyzed according to the red fluorescence (i.e., “dead”) of cells with defined structure. a–d Schematic illustration of the tipless cantilever as it deforms the uninfected and e–h infected E. coli cells at different applied forces. Black full lines a, e indicate the initial position of the tipless AFM cantilever, while the gray lines b–d, f–h mark the position of that cantilever following the application of a specific force. i A representative curve where the deformation (indentation), h, was set to zero when the AFM cantilever applied F = 10 nN. The deformations of uninfected or infected E. coli cells (Δh₁, Δh₂, and Δh₃) were measured by applying 25 nN, 50 nN, and 10 nN, respectively. Scatterplot indicates the average and s.d. from 12 different bacteria measured at 3–5 independent experiments. Asterisks indicate a statistically significant difference between uninfected and infected E. coli cells (p < 0.01, n = 12).