Volumetric Deformation of Live Cells Induced by Pressure-Activated Cross-Membrane Ion Transport

In this work, we developed a method that allows precise control over changes in the size of a cell via hydrostatic pressure changes in the medium. Specifically, we show that a sudden increase, or reduction, in the surrounding pressure, in the physiologically relevant range, triggers cross-membrane fluxes of sodium and potassium ions in leukemia cell lines K562 and HL60, resulting in reversible volumetric deformation with a characteristic time of around 30 min. Interestingly, healthy leukocytes do not respond to pressure shocks, suggesting that the cancer cells may have evolved the ability to adapt to pressure changes in their microenvironment. A model is also proposed to explain the observed cell deformation, which highlights how the apparent viscoelastic response of cells is governed by the microscopic cross-membrane transport.

Figure 1

(a) Evolution of the mean radius of K562 cells, normalized by its initial value $r_0=10.5\mu \mathrm{m}$, under a suddenly applied pressure of 70 mm Hg which was then removed after 2 h. Experimental data are represented by markers while theoretical prediction by Eq. (3) is shown by the solid curve. The error bar indicates the standard error of the mean (SEM). Representative cell morphologies at different stages are given in the insets. (b) Fluorescent intensities of intracellular sodium and potassium as functions of time during the loading-unloading process. Both intensities are normalized by their initial values before the pressure treatment. (c) Fluorescent images of both ions at different time points as indicated in (b). Scale bars in (a) and (c) correspond to 7.5 and $10\mu \mathrm{m}$, respectively. Results shown in (a) and (b) were based on measurement on 50 live cells with $p<0.03$ by the $t$ test.

Figure 2

Schematic diagram illustrating the pressure-triggered cross-membrane ion transport and volumetric deformation of cells. Here the term “membrane” refers to the lipid-protein bilayer and the associated actin cortex underneath. (a) Initially, cells maintain a steady-state size in the culture fluid, and the pressure difference across the membrane is balanced by the tension generated inside. (b) The cell shrinks in response to increasing hydrostatic pressure in the medium by activating ion channels that “pump” out intracellular potassium and sodium ions. (c) Pressure-induced cell shrinkage will be stopped if all channels have been blocked.

Figure 3

(a) The size of K562 cells, with initial radius of $r_0=10.5\mu \mathrm{m}$, as a function of time after the introduction of extra hydrostatic pressure in the medium. Predictions shown here are from Eq. (3) with parameter values given in Table S1 in the Supplemental Material [12]. (b) and (c) Evolutions of the number of intracellular potassium (b) and sodium (c) under different applied pressure. Theoretical fittings by Eq. (1) are also shown. (d) Fluorescent intensity drops (2 h after the pressure was applied) of intracellular sodium and potassium as functions of the magnitude of the pressure introduced. No detectable drop was observed when the applied pressure is below $\sim 7.5\mathrm{mm}$ Hg. (e) The radii of pressurized cells, treated with quinidine (potassium channel inhibitor), quinidine and sotalol (sodium channel inhibitor), or without drug treatment, as functions of time. Micrographs show the representative morphologies of these cells after 2 h of pressure treatment. (f) Evolution of the mean radius of mixed leukocytes (normalized by the initial value $r_0=5\mu \mathrm{m}$) under a suddenly applied hydrostatic pressure of 30 or 70 mm Hg. In comparison, the average sizes of K562 and HL60 cells subjected to the same treatment are also shown. Micrographs here show the representative morphologies of normal leukocytes during the experiment. Scale bars in (e) and (f) represent 10 and $5\mu \mathrm{m}$, respectively. Cell radii shown in (a),(e), and (f) were based on measurements on 50 live cells ($p<0.03$), while 2000 cells were examined by flow cytometry to render the results given in (b),(c), and (d) with $p<0.01$.