Shear Assay Protocol for the Determination of Single-Cell Material Properties

Summary

This protocol outlines the quantification of the mechanical properties of cancerous and non-cancerous cell lines in vitro. Conserved differences in the mechanics of cancerous and normal cells can act as a biomarker that may have implications in prognosis and diagnosis.

Abstract

Irregular biomechanics are a hallmark of cancer biology subject to extensive study. The mechanical properties of a cell are similar to those of a material. A cell's resistance to stress and strain, its relaxation time, and its elasticity are all properties that can be derived and compared to other types of cells. Quantifying the mechanical properties of cancerous (malignant) versus normal (non-malignant) cells allows researchers to further uncover the biophysical fundamentals of this disease. While the mechanical properties of cancer cells are known to consistently differ from the mechanical properties of normal cells, a standard experimental procedure to deduce these properties from cells in culture is lacking.

This paper outlines a procedure to quantify the mechanical properties of single cells in vitro using a fluid shear assay. The principle behind this assay involves applying fluid shear stress onto a single cell and optically monitoring the resulting cellular deformation over time. Cell mechanical properties are subsequently characterized using digital image correlation (DIC) analysis and fitting an appropriate viscoelastic model to the experimental data generated from the DIC analysis. Overall, the protocol outlined here aims to provide a more effective and targeted method for the diagnosis of difficult-to-treat cancers.

Introduction

Studying the biophysical differences between cancerous and non-cancerous cells allows for novel diagnostic and therapeutic opportunities¹. Understanding how differences in biomechanics/mechanobiology contribute to tumor progression and treatment resistance will reveal new avenues for targeted therapy and early diagnosis².

While it is known that cancer cell mechanical properties differ from normal cells (e.g., viscoelasticity of the plasma membrane and nuclear envelope)^3,4,5, robust and reproducible methods for measuring these properties in live cells are lacking⁶. The shear assay method is used to quantify the mechanical properties of cells by subjecting single cells to fluid shear stress and analyzing their individual responses and resistance to the applied stress^3,4,5,7,8,9. Although several methods and techniques have been used to characterize the mechanical properties of single cells, these tend to affect cell material properties by i) perforating/damaging the cell membrane due to the indentation depth, complex tip geometries, or substrate stiffening associated with atomic force microscopy (AFM)^10,11, ii) inducing cellular photodamage during optical trapping^12,13, or iii) inducing complex stress states associated with micropipette aspiration^14,15. These external effects are associated with significant uncertainties in the accuracy of cell viscoelasticity measurements^6,16,17.

To address these limitations, the shear assay method described here provides a highly controllable and simple approach to simulate physiological flow in the body without affecting cellular material properties in the process. Fluid shear stresses in this assay represent mechanical stresses experienced by cells in the body either by fluids within the tumor interstitium or in the blood during circulation^18,19,20. Further, these fluid stresses promote various malignant behaviors in cancer cells, including progression, migration, metastasis, and cell death^19,21,22,23 which vary between tumorigenic and non-tumorigenic cells. Moreover, the altered mechanical features of cancer cells (i.e., they are often "softer" than normal cells found within the same organ) allow them to persist in hostile tumor microenvironments, invade surrounding normal tissues, and metastasize to distant sites^24,25,26. By creating a pseudo-biological environment where cells experience physiological levels of fluid shear stress, a process that is physiologically relevant and not destructive to the cell is achieved. The cellular responses to these applied fluid shear stresses allow us to characterize cell mechanical properties.

This paper provides a shear assay protocol for the extensive study of the mechanical properties and behavior of cancerous and non-cancerous cells under applied shear stress. Cells respond to external forces in an elastic and viscous manner and can therefore be idealized as a viscoelastic material³. This technique is categorized into: (i) cell culture of dispersed single cells, (ii) controlled application of fluid shear stress, (iii) in situ imaging and observation of cellular behavior (including resistance to stress and deformation), (iv) strain analysis of cells to determine the extent of deformation, and (v) characterization of the viscoelastic properties of single cells. By interrogating these mechanical properties and behaviors, complex cellular mechanobiology can be distilled to quantifiable data. A protocol outlining this method allows for the cataloging of and comparison between various malignant and non-malignant cell types. Quantifying these differences has the potential to establish diagnostic and therapeutic biomarkers.

Protocol

1. Preparation for the single-cell shear assay

1. Cell culture
   1. Seed approximately 50,000 suspended single cells in a 35 mm x 10 mm Petri dish containing 2 mL of culture media.
      NOTE: Vortex the suspended cells prior to seeding to break apart cell aggregates.
   2. Incubate the cells at 37 °C and allow between 10 to 48 h for cell attachment and complete cytoskeletal protein formation.
      ​NOTE: Consider the duration of cellular attachment, as well as proliferation and growth rates, to ensure adequate cellular growth and attachment while avoiding cell aggregation. These parameters vary with cell type.

2. Shear assay experiment

1. Preparation of shear assay viscous flow media
   1. To ensure a slightly viscous flow media (0.015-0.02 Pa·s), measure and add 0.05 wt% of non-toxic and non-allergenic methylcellulose (4  Pa·s) to the culture media.
   2. To ensure a homogeneous mix, preheat the base culture media for ~10-20 min at a temperature of ~60-70 °C with a magnetic stirrer/hot plate. While continually stirring the media, gently add the methylcellulose such that it quickly disperses, to avoid coagulation of the methylcellulose particles. Allow this process to continue for ~15-24 h to ensure a clear solution of media + cellulose.
      NOTE: Avoid excessive heating of the solution.
   3. To measure the viscosity of the flow media, test ~0.5-1 mL of the representative flow media using a rheometer. From the readout, determine the fluid viscosity and utilize this value to represent the viscosity of the shear fluid medium (μ) to compute shear stress using equation (2).
2. Shear apparatus setup
   1. Set up the shear assay system of dual syringes (60 mL or 100 mL) connected to a programmable syringe pump for infusion and withdrawal of the viscous culture media (Figure 1).
   2. Attach both syringes to the flow chamber via 1/16 inch tubing and tubing connectors.
   3. Fasten a rubber gasket to provide a controlled, uniform flow on single cells along the flow path (Figure 1). The rubber gasket comes in different sizes depending on the flow profile to be attained (laminar or turbulent) and the desired area of observation (e.g., length of 22.5 mm, width of 2.5 mm, and height of 0.254 mm) (Figure 1).
   4. Program the pump to infuse and withdraw a certain volume of fluid (e.g., 60 mL) at a designated rate (e.g., 1 mL/min) and select the corresponding syringes (e.g., 60 mL).
      NOTE: Account for the maximum infusion and withdrawal volume presets to avoid jamming or malfunction. Use equation (2) for the calculation of the required pump shear rate (assuming the required stress and viscosity are known).
3. Pre-shear setup
   1. Fill the syringe with the prepared viscous flow media.
   2. Attach the syringe, filled with 60 or 100 mL (or as needed) of viscous flow media, and an empty 60 mL syringe to their respective locations on the programmable syringe pump. Via tubing and tubing connectors, connect both syringes to the flow chamber.
   3. To ensure easy identification of single cells and a fastened connection between the flow chamber and the Petri dish, attach the rubber gasket to the flow chamber.
   4. Aspirate the cell culture media from the Petri dish containing the cells of interest.
   5. Wash off dead cells and loosely attached cells using phosphate-buffered saline (PBS).
   6. Aspirate the PBS.
   7. Insert and fix the flow chamber and rubber gasket (~34 mm x 9 mm) onto the Petri dish (35 mm x 10 mm) containing the attached cells.
   8. Place the fitted microfluidic flow chamber + cells on a cultured dish onto an inverted microscope, with a microscope objective high enough to obtain high-quality images with high pixel values (usually between 40x and 63x magnification) and a display monitor.
   9. Select the live image (time-lapse on some software) option from the microscope software on the display monitor. Ensure the microscope software on the PC has either a t (time-lapse) functionality or can take video recordings.
   10. Focus the microscope objective, ensuring adequate contrast and distinct cell edges. This is necessary for the image analysis post-shear. Move the microscope stage to ensure the cells are clearly visible on the display monitor and are live images.
   11. Select a cell or multiple distinct cells within the imaging/flow path of the fitted flow chamber + Petri dish (the area/path created by the fitting of the gasket to the flow chamber).
4. Shear and imaging
   1. To maintain a continuous uniform flow, select similar infusion and withdrawal rates and ensure laminar flow of the fluid, usually between 1 mL/min and 5 mL/min. For low laminar flow regimes, ensure a Reynold's number of Re < 100.
   2. Click on Run on the shear pump to inject and withdraw the shear fluid (prepared viscous media) at a continuous rate. Ensure that there are no bubbles during fluid infusion, as this might introduce external unaccounted stress on the cells.
   3. Begin recording a video by clicking record on the microscope software before the infused shear fluid makes contact with the cell(s) of interest under the microscope.
   4. Continue to record for 7 min, or for the desired duration of stress exposure, or until the cell(s) of interest shears off the bottom of the dish. Click stop recording on the microscope software when the run is completed as desired.
   5. Save the recording and extract as .tiff files. Preferably, extract images at 1 frame per second for facile analysis.

3. Data processing

1. Digital image correlation procedure (image analysis)
   1. If the recording from the microscope was extracted as a video file, convert it to image frames (preferably .tiff file format).
   2. Import the images derived from the shear assay recording to Davis 10.1.2 software (DIC software) to track the movement of the naturally patterned structures of the cell by locating each block of pixel (subset) of the reference image in the respective new images (deformed images) (Figure 2).
   3. For an optimized correlation, utilize a subset size of 31 x 31 pixels, a step size (deformation distance of each subset) of 20 pixels, and the sum of the differential track option, which tracks the deformation of a new image with respect to the last image. The result of this correlation is a strain-time plot (Figure 3) that can be exported as a .csv file for further analysis in MATLAB.
   4. Map out the region of interest for a chosen single cell. Select arbitrary points within the mapped cell to track deformation. For an irregular shape, such as the cell, use a polygon mask to map out the cell's geometry.
   5. After mapping, choose specific points on the cell (nucleus or cytoplasm) to be analyzed by clicking add strain gauge and drawing out individual strain gauges at points within the defined cellular boundary.
   6. Click on Run to begin the strain processing and obtain strain versus time data.
   7. Double-click (or right-click) on the generated strain-time plot and select export data as a spreadsheet.

4. Mechanical property characterization

1. Viscoelastic property characterization
   1. Save the .csv file containing the strain-time data from the DIC software in a separate folder for easy readability by MATLAB.
   2. Run MATLAB and click on the editor tab to open an editor page to write a code that reads the spreadsheet, cell by cell.
   3. Change the MATLAB path (the folder path that accesses the file of interest) to access the folder containing the data to be analyzed, for example, Users/Username/Desktop/data.
   4. On the MATLAB editor page, access the spreadsheet data using the customized code. For example: a1= xlsread('data','run1','A4:A183'), where a1 represents the identifier, xlsread is the MATLAB function that reads the .csv file (in this case, as a spreadsheet), data is the file name, run1 is the sheet name, and A4:A183 is the range of data of interest in cell A of the spreadsheet data to be analyzed. For a complete fit, analyze x and y (time and strain, respectively). For example:
      a1=xlsread('data','run1','A4:A183');
      b1=xlsread('data','run1','B4:B183');
      a1 = x (time), and b1 = y (strain).
   5. In MATLAB, click Apps | Curve Fitter | Custom Equation. Clear the representative custom equation and input the viscoelastic model equation [equation (1)], where ε represents the x variable and t represents the y variable.
           (1)
      Here, ε represents the strain, σ represents the shear stress, η₁ represents the viscosity, E represents the elasticity, t represents the time, and τ represents the relaxation time, which characterizes the maximum time required for the cell to return to its original shape after primary deformation. It is expressed as τ = , where η₂ is the secondary viscosity term for the second dashpot (Figure 4).
   6. Reassign new variables to the viscoelastic parameters within the custom equation interface. (ε, η₁, E, σ, t, and τ = ) will represent (x, a, b, K, y, and c), respectively. Here, x and y are the independent and dependent variables, respectively. Shear stress (σ) can be determined using equation (2):
           (2)
      ​Here, μ represents the viscosity of the shear fluid medium, Q is the pump flow rate, and w and h are the width and height of the flow channel shown in Figure 1C, respectively.
   7. Click on Select Data to select the Time (a1) and Strain (b1) for each set of data.
   8. Ensure the Auto Fit box option is checked. This runs the fit automatically when data (x and y) are selected.
   9. Select the fitting methods to tighten the boundary conditions. Click on Advanced Options under the Methods category and select Nonlinear Least Squares. Under Robust, select Off, and under Algorithm, select Trust-Region. Leave every other parameter as it is.
   10. The new variables post-fitting (a, b, and c) represent the viscoelastic properties, viscosity, elasticity, and relaxation time (η₁, E, EQUAT), of the cell, respectively (Figure 5).
   11. Look for a high R-square value of the fit (R² > 80%) to ensure that the data output can be considered a true fit of the viscoelastic model (Figure 3)

Representative Results

The shear assay protocol coupled with deformation analysis using DIC and a viscoelastic model is successful in quantifying the mechanical properties of a single cell in vitro. This method has been tested on human and murine cell lines, including normal human breast cells (MCF-10A)^3,4,9, less metastatic triple-negative breast cancer cells (MDA-MB-468)³, triple-negative breast cancer cells (MDA-MB-231)³, human osteosarcoma cells^7,8, and most recently glioblastoma cells (Figure 3). DIC of image frames extracted from shear assay recordings is successful in producing strain-time data compatible with the creep stress response (Figure 3), which is fit to a viscoelastic model (Figure 4). With the use of MATLAB, the viscoelastic model can be used to fit the strain-time data to obtain the viscoelastic properties of the different cells.

Figure 5 describes the range of properties that have been previously studied using the shear assay technique. Additional mechanical properties can be characterized using this method, depending on the parameter of interest. The results of these studies using the shear assay technique showed that cancer cells were generally more mechanically compliant and less viscous than normal cells (Figure 6)^3,4,7,8. Figure 6A-C describes the mechanical properties of normal breast cells (MCF-10A) and triple-negative breast cancer cells (less metastatic MDA-MB-468 and highly metastatic MDA-MB-231). In Figure 6A,B, the stiffness and viscosities of the cells can be observed to decrease with increased cancer progression, from a normal cell state, to a slightly metastatic state, and to a highly metastatic state. Statistically significant variations were found between the normal MCF-10A cells and highly metastatic MDA-MB-231 cells, and between the less metastatic MDA-MB-468 cells and the highly metastatic MDA-MB-231 cells. There were no significant variations between the normal MCF-10A and the less metastatic MDA-MB-468 cells. Figure 6C shows the relaxation time for these cell types, which showed no significant variations; this may vary with other types of cells.

Figure 1: Shear assay experimental setup. The method involves the use of (A) a programmable syringe pump to both infuse and withdraw viscous flow media into and out of (B) a flow chamber at a constant shear rate. (C) A rubber gasket containing a rectangular window (dimensions: 20.5 mm x 2.5 mm x 0.254 mm) and vacuum suction is fit in a 35 mm x 10 mm Petri dish containing cultured cells of interest to establish a flow field. (D) The cells are subjected to flow within the established field and are deformed by fluid shear stress, which is derived via (E) an equation for wall shear stress. (F) A cell of interest within the flow field is imaged using a microscope at 63x magnification, and real-time deformation is captured and recorded using the microscope software on the monitor. The induced fluid shear stress is utilized along with the imaged deformation to quantify the mechanical properties of the single cell. This figure is modified from Hu et al.³. Please click here to view a larger version of this figure.

Figure 2: Strain analysis using digital image correlation. Images extracted from a recording of the shear assay show the maximum shear strain experienced by individual cells throughout the experiment. DIC tracks changes in the naturally patterned structure of the cells by locating and tracking a block of pixels subset within each frame or image. (A) The image of the cell before the DIC analysis. (B) Masking of the cells, selecting the region of interest, and seeding the cells to aid deformation tracking. (C) The onset of deformation. (D) Deformation build-up. (E) Increasing deformation over time. (F) Application of strain gauges to quantify strain during deformation. Abbreviation: DIC = digital image correlation. This figure is modified from Hu et al.³. Please click here to view a larger version of this figure.

Figure 3: Representative sample of strain versus time data of glioblastoma cells and mathematical modeling of the viscoelastic model. (A) The strain versus time data obtained by DIC show different creep regimes (primary, secondary, and tertiary) of the cell. (B) The strain versus time data is exported to MATLAB for quantification of the mechanical properties of the individual cell. Abbreviation: DIC = digital image correlation. This figure is modified from Hu et al.³. Please click here to view a larger version of this figure.

Figure 4: Viscoelastic model for the determination of viscoelastic properties. A three-parameter generalized Maxwell model, which comprises Kelvin (dashpot) and Voigt (dashpot + spring) models connected in series, for characterization of the viscoelastic behavior of cells. The result of the viscoelastic modeling using the shear technique is the cell's elasticity, viscosity, and relaxation time. This figure is modified from Hu et al.³. Please click here to view a larger version of this figure.

Figure 5: Graphical depiction of the types of material properties of cells that can be extracted from the shear assay technique, including stiffness, viscosity, creep, and relaxation time. Please click here to view a larger version of this figure.

Figure 6: Material properties of cells obtained from the shear assay technique. Cell stiffness was significantly different between the nucleus and cytoplasm of normal breast cells (MCF-10A) and highly metastatic breast cells (MDA-MB-231), and between less metastatic cells (MDA-MB-468) and highly metastatic cells (MDA-MB-231) (A). However, there was no significant difference between the stiffness of the nucleus and cytoplasm of normal cells (MCF-10A) and less metastatic cells (MDA-MB-468). Similar trends were also found in nuclear and cytoplastic viscosities (B). The relaxation times of these cells did not exhibit significant mechanical differences (C), though other cell types may. This figure is modified from Hu et al.³. Please click here to view a larger version of this figure.

Discussion

The shear assay method, which includes setting up an pseudo-mechanobiological environment to simulate the interaction of cells with the surrounding mechanical microenvironment and their responses to mechanical stresses, has produced a catalog of cellular mechanical properties, whose patterns show conserved physical atypia among cancerous cell lines^3,4,5,7,8. This method combines an understanding of basic fluid mechanics and physics to characterize unique mechanical properties of cells and provides potential material or mechanical biomarkers for the detection of several difficult-to-treat cancers^3,4.

The shear assay technique has certain limitations. Ensuring a single-cell culture can be challenging, owing to the quick growth and proliferation of certain cell types, making it difficult to visualize and analyze single cells. Stress tolerance and the cellular response can be heterogeneous from one cell to another depending on the cell line, and as such requires many cells and troubleshooting, to determine the average critical stress required to structurally deform the cells while maintaining their viability to analyze their material properties as live cells.

Troubleshooting for this technique includes determining i) the growth and proliferation rates of cell lines of interest and ii) the time needed for cells to adhere to the substrate prior to shear stress. Researchers using this protocol may choose to assess the expression of key cytoskeletal and adherent proteins, such as actin, cadherins, and other focal adhesion proteins. It is also necessary to determine the critical stress magnitudes required to induce the optimal cell deformation for the strain analysis software (DIC).

The shear assay technique can serve as an effective technique to explore the unique mechanical behaviors and inherent mechanical properties of different types of cells, potentially aiding in specific and sensitive identification and diagnoses that are independent of cell surface markers typically used in current cancer cell diagnostic methods. It also provides significant improvements over existing mechanical characterization methods that rupture the cell membrane prior to testing.

The unique shear assay technique has been extensively explored for over two decades for characterization of the behavior of biological cells and the determination of their mechanical properties. This experimental setup closely mimics the effects of fluid shear stresses in the body by applying controlled fluid shear stress on cells in vitro. Normal and cancerous cells are observed to respond to these stresses differently, and thus this technique provides insights to their structural and mechanical properties and behaviors. This technique has been explored in several prior studies to determine the viscoelastic properties of cells, and interestingly, significant cellular mechanical differences were observed in healthy normal cells and cancerous cells^3,4,9. Further, significant intracellular mechanical differences were observed between the nucleus and cytoplasm of cells. These unique mechanical properties can therefore be explored as potential mechanical biomarkers for the detection of different cancers in clinical settings.

Materials

Name	Company	Catalog Number	Comments
CELL CULTURE
.25% Trypsin, 2.21 mM EDTA, 1x[-] sodium bicarbonate	Corning	25-053-ci	For cellular detachment from substrate in cell culture
15 mL Centrifuge tubes	Falcon by Corning	05-527-90
35 mm Petri dishes	Corning	430165
50 mL Centrifuge tubes	Falcon by Corning	14-432-22
Centrifuge	any		For sterile cell culture
Dulbecco's Modification of Eagle's Medium (DMEM) 1x	Corning	10-013-cv	Or any other media for culturing cells. DMEM was used for culturing U87 cells
Gloves	any		For sterile cell culture
Heracell Vios 160i CO2 Incubator	Thermo Scientific	51033770	For Incubation during cell culture
Hood	any		For sterile cell culture
Micropipette	any		For sterile cell culture
Micropipette tips	any		For sterile cell culture
Microscope	Leica/any		For sterile cell culture
Phosphate Buffered Saline without calcium and magnesium PBS, 1x	Corning	21-040-CM
Pipetman	any		For sterile cell culture
Pipette tips	any		For sterile cell culture
Precision GP 10 liquid incubator	Thermo Scientific	TSGP02
T25 Flask	Corning	430639
T75 Flask	Corning	430641U
SHEAR ASSAY
100 mL beaker	any		For creating DMEM + methyl cellulose viscous shear media
DMEM	Corning
Flow chamber + rubber gasket	Glycotech	31-001	Circular Flow chamber Kit ( for 35 mm tissue culture dishes)
Hybrid Rheometer	HR-2 Discovery Hybrid Rheometer		For determination of shear fluid viscosity
Magnetic stir bar	any		For creating DMEM + methyl cellulose viscous shear media
Magnetic stir plate	any		For creating DMEM + methyl cellulose viscous shear media
Methyl cellulose	any		To increase viscosity of DMEM in flow media
Syringe Pump	KD Scientific Geminin 88 plus	788088	For programming fluid infusion and withdrawal
Syringes, tubing, and connectors			For shear apparatus setup
SOFTWARE
ABAQUS software	Simulia
Digitial Image Correlation software	LaVision, Germany	DAVIS 10.1.2
Imaging software	Leica/any microscope software
MATLAB	MATLAB	MATLAB_R2020B