In Situ Mapping of the Mechanical Properties of Biofilms by Particle-tracking Microrheology

Summary

Particle-tracking microrheology investigates the viscoelasticity of materials. Here, the technique is used to determine the viscoelasticity, creep compliance and effective crosslinking roles of different matrix components of a bacterial biofilm. The matrix consists of polymeric substances secreted by the bacteria and its components determine biofilm structure and mechanical properties.

Abstract

Bacterial cells are able to form surface-attached biofilm communities known as biofilms by encasing themselves in extracellular polymeric substances (EPS). The EPS serves as a physical and protective scaffold that houses the bacterial cells and consists of a variety of materials that includes proteins, exopolysaccharides and DNA. The composition of the EPS may change, which remodels the mechanic properties of the biofilm to further develop or support alternative biofilm structures, such as streamers, as a response to environmental cues. Despite this, there are little quantitative descriptions on how EPS components contribute to the mechanical properties and function of biofilms. Rheology, the study of the flow of matter, is of particular relevance to biofilms as many biofilms grow in flow conditions and are constantly exposed to shear stress. It also provides measurement and insight on the spreading of the biofilm on a surface. Here, particle-tracking microrheology is used to examine the viscoelasticity and effective crosslinking roles of different matrix components in various parts of the biofilm during development. This approach allows researchers to measure mechanic properties of biofilms at the micro-scale, which might provide useful information for controlling and engineering biofilms.

Introduction

Most bacterial cells are able to employ both planktonic (free-living) and surface-attached (sessile) modes of growth ¹. In the surface-attached mode of growth, bacterial cells secrete and encase themselves in large amounts of extracellular polymeric substances (EPS) to form biofilms. The EPS mainly consists of proteins, exopolysaccharide, extracellular DNA and is essential to biofilm formation ². It serves as a physical scaffold by which bacteria can use to differentiate spatially and protects the bacteria from harmful environmental conditions and host responses. Different components of EPS have distinct roles in biofilm formation ³ and changes in the expression of EPS components can dramatically remodel biofilm structures ⁴. EPS components can also function as signaling molecules ⁵, and recent studies has shown certain EPS components interacting with microbial cells to guide their migration and biofilm differentiation ^6-8.

Research on the EPS has greatly advanced based upon the morphological analyses of biofilms produced by mutants defective in a specific component of the EPS ^9,10. In addition, the EPS is usually characterized at the macro-scale (bulk characterization) ¹¹. Morphological analyses however can lack quantitative detail and bulk characterization, which returns average values, loses the detail that exists within the heterogeneity of the biofilm. There is now an increasing trend to progress to real-time characterization of the mechanic properties of EPS at the micro-scale. This protocol demonstrates how particle-tracking microrheology is able to determine the spatiotemporal effects of matrix components Pel and Psl exopolysaccharides on the viscoelasticity and effective crosslinking of Pseudomonas aeruginosa biofilms ⁴.

Passive microrheology is a simple and inexpensive rheology method that provides the highest throughput of spatial microrheological sampling of a material to date ^12,13. In passive microrheology, probe spheres are placed in the sample and their Brownian motion, driven by thermal energies (k_BT) is followed by video microscopy. Several particles can be tracked simultaneously, and the time-dependent coordinates of the particles follow a conventional random walk. Therefore, on average, the particles remain at the same position. However, the standard deviation of the displacements or the mean squared displacement (MSD) of the particles, is not zero. Since viscous fluids flow, the particle MSD in a viscous fluid grows linearly as time progresses. In contrast, the polymeric crosslinking found in viscoelastic or elastic substances help them to resist flow, and particles become limited in their displacement, leading to plateaus in the MSD curve (Figure 1A). This observation follows the relation MSD∝t^α , where α is the diffusive exponent that is related ratio of elastic and viscous contributions of the substance. For particles moving in viscous fluids α = 1, in viscoelastic substances 0 < α < 1, and in elastic substances α = 0. The MSD may also be used to calculate the creep compliance, which is the tendency of the material to deform permanently over time and estimates how easily a material spreads.

The size, density and surface chemistry of the particle are critical to the correct application of microrheological experiment and are chosen with respect to the system studied (in this case the polymers of the biofilm matrix, see Figure 1B). Firstly, the particle measures the rheology of the substance with structures that are much smaller than the particle itself. If the substance's structures are of similar scale to the particle, the motion of the particle is perturbed by the shape and orientation of the individual structures. However, if the structures surrounding the particle are much smaller, this effect is small and averaged out, presenting a homogeneous environment to the particle (Figure 1B). Secondly, the density of the particle should be similar to the medium (1.05 g ml^-1 for water based mediums) such that sedimentation is avoided and inertial forces are negligible. Most particles with polystyrene lattices meet the above criteria. Ideally, the particle does not interact with the polymers of the biofilm matrix as the rheological interpretation of particle MSD is only valid if motion is random, driven by thermal energy and collision with substance structures. This can be observed by checking whether the probe particle tends to bind or bounce off the surface of a pre-grown biofilm. However, despite the lack of attraction to the biofilm, the particles must be able to be incorporated into the matrix. In addition, the physiochemical heterogeneity of the biofilm may result in different particles being more suitable as probes in different regions of the biofilm. Thus, particles of different sizes and surface chemistry should be applied to the biofilm.

As such, the particle MSD is able to provide useful information on how different components contribute to the rheology and spreading of the biofilm. Furthermore, the use of different probes allows one to derive information on the spatial physiochemical heterogeneity of the biofilm. This method can be used to test the effect antimicrobial treatment on the mechanical properties of the biofilm, or applied to mixed species biofilms to investigate how the mechanical properties of the biofilm are changed from introduction of another species. Particle MSDs may also be useful for characterizing biofilm dispersal. Such studies would be helpful in our understanding of biofilms, potentially improving biofilm treatments and engineering of biofilms for useful activities.

Protocol

1. Biofilm Cultivation

1. Preparation of Bacterial Strains
   1. 1 day prior to biofilm cultivation, prepare planktonic bacterial cultures by inoculating 2 ml of appropriate growth medium from frozen bacterial culture. Use Luria-Broth medium (10 g L^-1 NaCl, 10 g L^-1 yeast extract, and 10 g L^-1 tryptone) for mucoid P. aeruginosa and its Δpel and Δpsl defective mutants. Incubate overnight at 37 °C and 200 rpm shaking conditions. Dilute overnight cultures to an OD₆₀₀ of 0.40 using a spectrophotometer.
   2. Assemble flow cell setup, which has been described previously ¹⁴, preferably on a mobile station or trolley that can be brought into the microscope room for imaging the flow cell without disassembly.
   3. Prepare sterile growth medium in 2 L or 5 L bottles for each flow cell. Use minimal medium M9 (48 mM Na₂HPO₄, 22 mM KH₂PO₄, 19 mM NH₄Cl, 9 mM NaCl, 2 mM MgSO₄, and 0.1 mM CaCl₂) supplemented with 0.04% glucose (wt/vol) and 0.2%(wt/vol) casamino acids) for P. aeruginosa flow cell biofilms.
   4. Aliquot fluorescent microspheres into microcentrifuge tubes and spin down in sterile H₂O for 5 min in centrifuge at 9,391.2 g. Remove supernatant and resuspend in 1 ml growth medium to remove sodium azide (disinfectant) prior to adding to growth medium in 2 L or 5 L medium bottles.
      NOTE: Different sizes of microspheres come in different concentrations and should be diluted accordingly to instructions. For example, disperse 120 µl of 1.0 µm diameter microspheres, 15 µl of 0.5 µm diameter microspheres and 0.96 µl of 0.2 µm diameter microspheres into 2 L of growth medium to give 2.18 × 10⁶ microspheres ml^-1 for each particle size.
   5. Attach medium bottle to upstream of flow cell and allow growth medium to flow through entire setup. Stop flow and inject diluted overnight cultures into flow cell chambers. Allow bacteria to attach to the coverslip substratum for 1 hr before continuing flow of growth medium at a flow rate of approximately 5.5 x 10^-3 m sec^-1,or 8 rpm with a peristaltic pump.
      NOTE: Biofilms are usually grown at room temperature (25 °C) for 3-7 days.
   6. Alternatively, for static culture setup, dilute overnight cultures 100 fold in microcentrifuge tubes by adding 10 µl of overnight culture to 1 ml growth medium with particles. Increase the particle concentration in the growth medium for static culture by approximately 500-fold compared to that used for flow biofilms (e.g. wash and resuspend 600 µl of 1.0 µm diameter spheres in 10 ml of growth medium) as flow cell biofilms are constantly replenished with particles but static cultures are not.
      1. Add 200 µl of diluted overnight culture with particles to the chambers of 8 well slides. Incubate at 37 °C under static conditions. Biofilms are usually grown for 1 day. Replace spent growth medium daily with fresh growth medium with particles if biofilm is grown for more than 1 day.

2. Microscopy

1. On day 3 and 5, switch off flow from peristaltic pump and bring flow cell setup into the microscopy room. Clamp tubing near the entrance and exit of the flow cell chambers to prevent drift (a systematic error cause by changes in the environment) from flow. Place the flow cell onto the microscope stage of an upright microscope.
   NOTE: Use an inverted microscope for imaging of microwells.
2. Use fluorescent microscopy with 40X oil objective to take videos of microsphere motion embedded in the biofilm at various locations (microcolonies and flat undifferentiated layers) and in different days (day 3 and 5) for temporal and spatial investigations. Take shorter videos with higher frame rate to investigate events occurring within short time scales (e.g. 1.5-3 min videos at frame rate of 25-50 frames per second for fast dynamics), and longer videos with lower frame rate to investigate events over longer time scales (e.g. 15-30 min videos at frame rate of 2.5-5 frames/sec for slower dynamics).
   NOTE: A video typically contains 5-10 particles, and is typically of 1.0-2.5 GB in file size. Larger microcolonies may hold more particles. Take 5-10 videos for each category (e.g. 5-10 videos of different microcolonies and of different locations in flat undifferentiated layers). The biofilm has a heterogeneous architecture that may consist of microcolonies, channels, voids and flat undifferentiated layers.
3. Save videos in appropriate format that can be read by Fiji/ImageJ (e.g. czi, tif).
4. Check videos for drift by scrolling through finished video and observing that the particles do not move in the same direction simultaneously. Drift can be caused by z-stage motion, currents in the flow cell, imbalance of anti-vibration table, air pressure or temperature changes. Correct minor drifting using post-processing techniques usually provided with microscope software. Discard any videos wherein drift cannot be corrected. Record the following parameters, which are required during Particle Tracking Analysis: Resolution (px/um), Number of Frames, Duration.
5. Remove flow cell from microscope stage and restart peristaltic pump to continue cultivating the biofilm.

3. Particle Tracking Analysis

1. Obtain particle trajectories by using the plug-in TrackMate in open source software (ImageJ). Download Fiji at  http://fiji.sc/Fiji. Open the video with Fiji and under the Plugins menu, select Tracking and TrackMate.
2. Check or adjust settings in the window suggesting initial calibration settings to match the video parameters as recorded in step 2.4.
3. Detect particles in TrackMate using ImageJ.
   1. Select 'LoG detector', input particle diameter and threshold value (e.g. 1,000). Scroll through the video whilst clicking on 'Preview' button to check that the purple circles follow the particles throughout the video. Adjust the diameter and threshold values as necessary: increase the threshold value if purple circles appear in the empty space. Reduce the threshold value if not enough particles are detected. Click the next button to complete detection the particles.
   2. Click the next button when presented with the option to add or remove particles detected in 'Initial thresholding', 'Select a view' and 'Select a filter' windows to continue without adjustment if initial particle detection was satisfactory.
   3. Select 'Simple LAP tracker' in the option bar of 'Select a tracker method' window as particles rarely move out of focus in the biofilm and are easily tracked. Use the suggested or low linking and gap-closing max distance values, and click next.
   4. Check that particle tracking is satisfactory by scrolling through the video to see that particles follow the tracks drawn by TrackMate, and that there are no unwanted or missing tracks. Skip the various filtering steps. Go to the final window 'Select an action'. Select 'Export tracks to xml file' from the drop down menu and click 'Execute'. Choose a folder to save the tracks.
      NOTE: There are various programs available in the public domain for the analysis of particle trajectories and use of microrheology. msdanalyzer and TrackArt are examples of particle tracking analysis programs that run in the Matlab environment.
4. Prepare Matlab to import the particle trajectories in the xml files by selecting in the menu of Matlab File>Set Path... and add the scripts folder available with the Fiji package.
5. To analyze the particle trajectories with msdanalyzer, download the link to the zip file or the tar.gz file at http://www.mathworks.com/matlabcentral/fileexchange/40692-mean-square-displacement-analysis-of-particles-trajectories.
   1. Extract the @msdanalyzer folder¹⁵ and drop it in a folder that belongs to the Matlab path (e.g. C:\Documents and Settings\\My Documents\Matlab). Start Matlab and initiate the analyzer by typing:
      ma = msdanalyzer(2, 'um', 'sec')
      where um and sec are the physical space and time units in the video.
      1. Import the particle trajectories by typing:
         [tracks, md] = importTrackMateTracks('FileName.xml', 'clipZ', 'scaleT')
         ma = ma.addAll(tracks);
         where 'clipZ' removes the z dimension for a 2D video, and 'scaleT'.
      2. Plot the trajectories onto graph using:
         ma.plotTracks;
         ma.labelPlotTracks;
      3. Compute the MSDs of the particles using:
         ma = ma.computeMSD;
         ma.msd
      4. Plot the MSDs of each particle:
         figure
         ma.plotMSD
      5. Combine particle trajectories from other videos of the same categery (e.g. particles embedded in wild-type microcolonies at day 3) by repeating 3.5.1.1 to 3.5.1.4).
      6. Plot the ensemble mean or the average over all curves:
         ma.plotMeanMSD
         Change the y- and x-axis from linear to logarithmic scale. At long lag times, the MSD curves become noisy due to insufficient statistics at longer timescales. The MSD curves may also rise steeply due to dynamic error. These regions of the curve can be removed by using the brush tool to select the end of the curve and selecting 'Brushing' > 'Remove Unbrushed' in the Tools menu.
   2. Download Ezyfit at http://www.fast.u-psud.fr/ezyfit/ and add Ezyfit to a folder belonging to the Matlab path (e.g. C:\Documents and Settings\\My Documents\Matlab). Install the Ezyfit menu by typing in Matlab workspace efmenu install. Restart MATLAB with Ezyfit menu in the figure window. Fit the MSD curves to the power law by selecting in the Ezyfit menu 'Show Fit' > 'Power' > 'a*x^n' were n is the estimated diffusive exponent α.
      NOTE: msdanalyzer requires the Curve Fitting Toolbox in Matlab for fitting of MSD curves. Ezyfit is an alternative and free software that can perform curve fitting.
   3. Clear current particle trajectories and MSDs to calculate new particle MSDs at other locations or time:
      clear
   4. After calculation of various mean MSD curves of particles in medium (control), and microcolonies and undifferentiated areas of various strains, copy and paste the curves into one graph for comparison. Sample stiffness and crosslinking increase with lower MSD values. Flat curves have lower α and sample is more elastic than steeper curves with higher α.
   5. Select the curve and go to "Tools" to look at the data statistics, such as the median and range. The MSD of the bead is proportional to the creep compliance, J(t) of the material in which the bead is embedded according to the relation,
      
      where J = creep compliance, d = particle diameter, k_B = Boltzman constant, T = temperature and t = time.

Results

The local viscoelastic properties of the biofilm in different regions of the biofilm, which included the voids (medium above the biofilm), plains (undifferentiated flat layer of cells) and microcolonies (see labels in Figure 2A) were investigated. The temporal changes in viscoelastic properties of the biofilm during maturation from days 3 to 5 were also determined. The MSD of the particles in the voids was used as a control and comparable to the MSD of particles in pure medium. In contrast, particles tra...

Discussion

Microrheology is a useful tool for local rheological measurements in heterogeneous systems, such as microbial biofilms. It is a non-destructive technique, enabling the real-time monitoring of rheological changes within the same biological sample over multiple time points. In this protocol, particle-tracking microrheology was applied to Pel and Psl exopolysaccharide mutants in order to investigate how they affect the elasticity and effective crosslinking of the biofilm matrix. Psl favors the development of elastic biofilm...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Fluorspheres	Invitrogen	F-8821	1.0 μm red fluorescent (580/605) microspheres with carboxylate modification
Zeiss Axio Imager M1	Carl Zeiss		Epifluorescent Microscope
Masterflex L/S Digital Drive 07523-80	Cole-Parmer	EW-07523-80	Peristaltic pump
Flow Cell Chambers	Technical University of Denmark
Bubble Trap	Technical University of Denmark
Silicone Tubing	Dow Corning		3 mm outer diameter, 1 mm inner diameter
Clear polypropylene plastic connectors	Cole Parmer	06365-83	1/16 in. (1.588 mm)
Binder Clips			To clamp tubing
Coverslips	Thermo Scientific™ Nunc™	50 x 24 mm
Syringe 3 ml	Terumo
27  G Needle	Terumo
2  L Storage/Media Bottles	VWR®
Trolley			To hold biofilm setup