Strain-rate Dependence of Elastic Modulus Reveals Silver Nanoparticle Induced Cytotoxicity

Abstract

Force-displacement measurements are taken at different rates with an atomic force microscope to assess the correlation between cell health and cell viscoelasticity in THP-1 cells that have been treated with a novel drug carrier. A variable indentation-rate viscoelastic analysis, VIVA, is employed to identify the relaxation time of the cells that are known to exhibit a frequency dependent stiffness. The VIVA agrees with a fluorescent viability assay. This indicates that dextran-lysozyme drug carriers are biocompatible and deliver concentrated toxic material (rhodamine or silver nanoparticles) to the cytoplasm of THP-1 cells. By modelling the frequency dependence of the elastic modulus, the VIVA provides three metrics of cytoplasmic viscoelasticity: a low frequency modulus, a high frequency modulus and viscosity. The signature of cytotoxicity by rhodamine or silver exposure is a frequency independent twofold increase in the elastic modulus and cytoplasmic viscosity, while the cytoskeletal relaxation time remains unchanged. This is consistent with the known toxic mechanism of silver nanoparticles, where metabolic stress causes an increase in the rigidity of the cytoplasm. A variable indentation-rate viscoelastic analysis is presented as a straightforward method to promote the self-consistent comparison between cells. This is paramount to the development of early diagnosis and treatment of disease.

Keywords: VIVA; cell viscoelasticity; dextran; elastic modulus; nano-indentation; nanogel; silver cytotoxicity; silver nanoparticle; standard linear solid model; strain-rate dependent elasticity.

Figure 1.

Modulus of Polyacrylamide Gels. a. The bulk shear modulus is determined as a function of frequency by parallel-plate rheology and converted to an elastic modulus using equation 1. b. The elastic modulus of polyacrylamide gels is determined at different frequencies by fitting force-distance curves, which are taken at different rates with an AFM to the Hertz equation (see supplement).

Figure 2.

Rate dependence of the cell modulus. The strain-rate dependence of the elastic modulus for three cell types determined by force-distance curves at different rates fit to the Hertz equation. The modulus was measured in the cytosolic region of all three cell types and the solid lines represent fits to the SLSM.

Figure 3.

Optical micrographs of the THP-1 cells, which exhibit treatment dependent morphology. a. Untreated THP-1 Cells (control) b. Unloaded lysozyme-dextran conjugate nano-gel drug carriers (Dex-Gel). c. Rhodamine labelled dextran nanogels (Rd Dex-Gel). d. Dex-Gel loaded with 5 nm silver nanoparticles within the core (Ag Dex-Gel). The THP-1 cells are exposed to Ag Dex-Gels 24 h prior to imaging. A schematic of each drug/ carrier complex is drawn inset. Scale bars are 20 μm.

Figure 4.

Localization of the Rd Dex-Gel inside the THP-1 cell 24 h after exposure a. Simultaneous white light and fluorescent imaging that show distinguishable components of the cell. The cell periphery, P, cytosol, cyto, and nucleus, N, are labelled. The outline of the retracted AFM cantilever is visible and a red dot is drawn to the approximate size of the microsphere indenter. b. Red fluorescence image of the same cell shows the Rd Dex-Gels (white dots) and diffuse rhodamine (grey cloud) in the cell. The Dex-Gels are localized in the cytoplasmic region (Cyto) c. VIVA to identify the rate dependence of the elastic modulus of P, Cyto, and N of a THP-1 cell modulus results are N=28 measurements from different areas of a single representative HUVEC cell. Fits to SLSM have R² = 0.89, 0.86 and 0.80 for the periphery, nuclear and cytosol respectively. Scale bars 20 μm.

Figure 5.

Viability staining of the THP-1 cells 24 h after exposure to the Dex-Gels. a. The untreated THP-1 cells (control) appear predominantly healthy (blue staining dominates histogram) b. The Dex-Gel exposed THP-1 cells are also healthy at 24 h, indicating biocompatibility of the carrier. c. The Rd Dex-Gel loaded THP-1 cells require a different stain due to the fluorescence of the rhodamine dye, thus no histogram indicating stressed or unstressed population can be generated. d. The Ag Dex-Gel loaded THP-1 cells appear to be stressed, as cells expressing both dye (purple) dominate the histogram. e. Viability histogram for all populations. If the intensity of the blue is greater than the intensity of the red dye, a cell is counted as alive. Inset histograms (a, b, and d) show the proportion of each population, which expresses blue, red or both (purple) dyes.

Figure 6.

Elastic modulus distribution of THP-1 cell populations. a. Dex-Gel exposed THP-1 cell modulus distributions at four different indentation frequencies. The histograms are well fitted by a log-normal distribution at all frequencies. The circles represent the mean value of distribution. The P-values reflect Mann-Whitney test of distributions, compared to 1 rad/s data. b. The modulus distribution at 3 rad/s indentation frequency of the four cell populations. The histograms are fit by a single log-normal and the dots show the mean value. The P-values reflect Mann-Whitney test of distribution, compared to the THP-1 cell (control). N = 54; Dex-Gel, N=74; Rd Dex-Gel, N=32; Ag Dex-Gel, N = 67. Each modulus measurement (N) at each frequency is the result of between two and five independent indentations.

Figure 7.

Frequency dependence of the elastic modulus of treated and untreated THP-1 cells. The average elastic modulus determined by nano-indentation of THP-1 cells treated with different Dex-Gels is shown over two orders of magnitude of indentation frequency. The most simple viscoelastic model, which captures the data, is the standard linear solid model (illustrated top). The Ag Dex-Gel and Rd Dex-Gel exposed cells show higher average stiffness (black circles and red squares). The Dex-Gel and untreated cells are routinely softer (green diamonds and blue triangles). Solving the equation of motion for the model shown in the frequency space (top) yields equation 6.4, to which the data are well fitted (solid lines) THP-1 Cell, N = 54; Dex-Gel, N=74; Rd Dex-Gel, N=32; Ag Dex-Gel, N = 67. Each modulus measurement (N) at each frequency is the result of between two and five independent indentations.

Figure A1.

a. Model of SLSM numbering the spring and dashpot elements. The height of the cell, H, nucleus, N, and the stress applied to the cell, σ, are labelled. The elements of the SLSM are numbered so that their corresponding strain response, σ_1–3 can be catalogued below. b. The typical force-distance curve of a cell fit to the Hertz contact model for a spherical indenter. The maximum indentation depth into the cell, δ_max, is drawn on the plot. The strain-rate used was 2 μm/s and the trigger point was 10 nm.

Figure A2.

The frequency dependence of the elastic modulus is modelled for a standard linear solid. E₁ = 350, E₂ = 640, η = 60.