Biophysical studies of cancer cells' traverse-vessel behaviors under different pressures revealed cells' motion state transition

Abstract

Circulating tumor cells (CTCs) survive in the bloodstream and then seed and invade to foster tumor metastasis. The arrest of cancer cells is favored by permissive flow forces and geometrical constraints. Through the use of high-throughput microfluidic devices designed to mimic capillary-sized vessels, we applied pressure differences to cancer cells (MCF-7 cell line) and recorded the cell traverse-vessel behaviors. Our results showed that cancer cells transform from a Newtonian droplet state to an adhesion/migration state when cancer cells traverse artificial vessels. To explain these phenomena, a modified Newtonian droplet model was also proposed. These phenomena and the modified model may reveal how CTCs in the blood seed and invade vessels under suitable conditions.

Figure 1

The design principle of the microfluidic device and time-lapse microscopy images of two typical cells with different traverse-vessel behaviors. (A) Schematic diagram of the double-layer microfluidic chip with CO₂ layer (blue) and cell cultivation layer (red) to maintain cell culture environment and culture cells. (B) A photograph of the microfluidic chip with CO₂ layer (upper, blue) and cell cultivation layer (bottom, red). Scale bar: 1 cm. (C) Detailed schematic diagrams of the cell cultivation layer with Inlet, Filter, Trap Unit and Outlet. (D) Schematic illustration of experimental design in the research. (E,F) Time-lapse microscopy images and the protrusion length as a function of time for two typical cells under applied pressure difference $\mathrm{\Delta }\mathit{P}$, 200 mbar and 100 mbar respectively. Scale bar: 20 μm.

Figure 2

Left: Dynamic behaviors of cancer cells traversing microvessels with 7.5 μm × 6 μm × 40 μm over time under four applied pressure differences $\mathrm{\Delta }\mathit{P}$ (400 mbar, 200 mbar, 100 mbar and 50 mbar); Right: Scatter plot of R-square and fitting velocities for cells traversing microvessels with 7.5 μm × 6 μm × 40 μm under four applied pressure differences $\mathrm{\Delta }\mathit{P}$ (400 mbar, 200 mbar, 100 mbar and 50 mbar). Red lines and Blue lines indicated cells with R-squared of linear fitting larger or smaller than 0.85.

Figure 3

The model and the simulation of traverse-vessel behaviors. (A) Sketch map of cell elongate, and adhere/migration in the microvessel. (B) The model for the dynamic behavior of cells in the microvessel. (C) The simulation of dynamic behaviors for cells under different pressure differences (50 mbar and 200 mbar) using the microfluidic chips with the same size of microvessels (7.5 μm × 6 μm). (D) The simulation of dynamic behaviors for cells under the same pressure difference (100 mbar) using the microfluidic chips with different sizes of microvessels (6 μm × 5 μm and 7.5 μm × 6 μm).

Figure 4

Distribution analysis of cells’ behaviors illustrated the transition. (A) Probability density distribution of cell apparent viscosities for 50, 100, and 200 and 400 mbar at various periods (0–10 min, 20–40 min, 80–120 min, 120–240 min). Major peak in blue, minor peak in pink and total in purple. (B) Probability density distribution of the R-squared of linear fitting for 50, 100, and 200 and 400 mbar at various times. The left y-axis shows frequencies and the right y-axis gives counts for both (A) and (B).

Figure 5

(A,B) Heatmap depicting the number of cells with corresponding R-squared and apparent viscosities under the pressure difference 50 mbar. (A) Counts cells between 20 and 40 min in experiments. (B) Counts cells between 40 and 80 min in simulations. (C,D) Medians of apparent viscosities and R-squared of cells (up to 400 cells in the first time frame under the pressure difference 50 mbar, no less than 20 cells in the final time frame under the pressure difference 400 mbar) over time under different pressure differences in experiments. Insets are in simulations. (E) Ratio of residual cells (larger than 200 cells in any condition at the start of experiments) to total throughout the experiments under different pressure differences. Inset is in simulations. Dashed lines represent results of linear fitting for two distinct stages.

Figure 6

The summarization of the R-squared and the apparent cell viscosity of the cell traverse-vessel behaviors at different conditions. (A,B) The comparison of plots of R-squared and cell apparent viscosity as functions of pressure applied for experiments and simulations under different pressure differences (median of all available cells in one condition, and no less than 80 cells in any condition at any time). (C) Illustration of cell transition from Newtonian liquid mode to adhesion and migration mode.