Probing the nanoscale viscoelasticity of intracellular fluids in living cells

Abstract

We have used fluorescence correlation spectroscopy to determine the anomalous diffusion properties of fluorescently tagged gold beads in the cytoplasm and the nucleus of living cells. From the extracted mean-square displacement v(tau) approximately tau(alpha), we have determined the complex shear modulus G(omega) approximately omega(alpha) for both compartments. Without treatment, all tested cell lines showed a strong viscoelastic behavior of the cytoplasm and the nucleoplasm, highlighting the crowdedness of these intracellular fluids. We also found a similar viscoelastic response in frog egg extract, which tended toward a solely viscous behavior upon dilution. When cells were osmotically stressed, the diffusion became less anomalous and the viscoelastic response changed. In particular, the anomality changed from alpha approximately 0.55 to alpha approximately 0.66, which indicates that the Zimm model for polymer solutions under varying solvent conditions is a good empirical description of the material properties of the cytoplasm and the nucleoplasm. Since osmotic stress may eventually trigger cell death, we propose, on the basis of our observations, that intracellular fluids are maintained in a state similar to crowded polymer solutions under good solvent conditions to keep the cell viable.

FIGURE 1

Representative FCS curves C(τ) of gold particles diffusing in the cytoplasm and the nucleus of untreated HeLa cells (solid circles). A clear change is seen in both cases when cells were osmotically stressed by adding 500 mM sucrose to the medium (open squares). (Inset) The autocorrelation decay of the fluorescent gold particles in water (crosses) is well described by normal diffusion. Solid lines are best fits with Eq. 1.

FIGURE 2

Average MSD v(τ) as obtained by fitting FCS curves with Eq. 1 (cf. Tables 1 and 2). Solid circles correspond to untreated HeLa cells, and open squares to osmotically stressed cells. The power-law increase v(τ) ∼ τ^α(α < 1) is highlighted by dashed lines.

FIGURE 3

Average real (elastic) and imaginary (viscous) parts of the complex shear modulus (G′(ω) (solid circles) and G″(ω) (open squares), respectively) in the cytoplasm and nucleoplasm of HeLa cells (cf. Tables 1 and 2). The power-law increase, G(ω) ∼ ω^α, is highlighted by dashed lines.

FIGURE 4

(a) The distribution of residence times p(τ_s) resembles the previously reported exponential distribution of masses in the cytoplasm (dashed line). (b) Upon introduction of osmotic stress, the distribution broadens and approaches a more uniform distribution. (Insets) The distribution of anomalities α is roughly Gaussian and a shift of the mean α is clearly visible. For better comparison, the axes have been given the same scaling.

FIGURE 5

(a) The degree of anomality α measured in Xenopus egg extracts (initial concentration c₀) tends to unity upon dilution. The full line is a guide to the eye. (b) The ratio between elastic and viscous modulus tends toward zero upon diluting the extract. c/c₀ = 100%, 10%, and 1% for squares, diamonds, and triangles, respectively.

FIGURE 6

(a) Representative FCS curves C(τ) of gold particles diffusing in a crowded dextran solution (100 g/l) based on different solvents (solid symbols, pure water; open symbols, 60% water, 40% ethanol). A clear change in the diffusional behavior is visible. (b) Corresponding average modulus of the complex shear modulus (|G(w)|) as derived from FCS measurements. A clear change in the scaling from α = 0.85 to α = 0.97 is observed when changing the solvent from water to a water/ethanol mixture.