Continuously perfused microbubble array for 3D tumor spheroid model

Abstract

Multi-cellular tumor spheroids (MCTSs) have been established as a 3D physiologically relevant tumor model for drug testing in cancer research. However, it is difficult to control the MCTS testing parameters and the entire process is time-consuming and expensive. To overcome these limitations, we developed a simple microfluidic system using polydimethylsiloxane (PDMS) microbubbles to culture tumor spheroids under physiological flow. The flow characteristics such as streamline directions, shear stress profile, and velocity profile inside the microfluidic system were first examined computationally using a COMSOL simulation. Colo205 tumor spheroids were created by a modified hanging drop method and maintained inside PDMS microbubble cavities in perfusion culture. Cell viability inside the microbubbles was examined by live cell staining and confocal imaging. E-selectin mediated cell sorting of Colo205 and MDA-MB-231 cell lines on functionalized microbubble and PDMS surfaces was achieved. Finally, to validate this microfluidic system for drug screening purposes, the toxicity of the anti-cancer drug, doxorubicin, on Colo205 cells in spheroids was tested and compared to cells in 2D culture. Colo205 spheroids cultured in flow showed a threefold increase in resistance to doxorubicin compared to Colo205 monolayer cells cultured under static conditions, consistent with the resistance observed previously in other MCTS models. The advantages presented by our microfluidic system, such as the ability to control the size uniformity of the spheroids and to perform real-time imaging on cells in the growth platform, show potential for high throughput drug screening development.

Figure 1

A COMSOL model was used to demonstrate the flow characteristics inside the PDMS microbubbles that were assembled within a Glyotech parallel plate flow chamber. Streamline side view (a) and top view (b) show the three-dimensional circular flow paths produced inside one microbubble. The COMSOL model slice plot (c) shows the flow velocity magnitude inside one microbubble. The velocity decreases sharply with distance from the bubble top opening. The shear stress plot (d) shows the shear stress magnitude exerted by the perfusion flow at the inner surface of the microbubble. COMSOL simulation conditions: Perfusion flow rate: 0.403 ml/min; fluid velocity: 2.12 cm/sec; shear stress at the PDMS surface: 10 dyn∕cm²; 600 μm diameter microbubble with circular 100 μm top opening.

Figure 2

(a) Sequential images of the microbeads trapped and circling inside the microbubble. Black arrows indicate the direction of the perfusion flow. Red and green arrows follow two different beads in their circular trajectories. These values are in agreement with the COMSOL simulation results (Fig lc). (b) The circular paths taken by the two beads were highlighted using ImageJ software, (c) Schematic of the parallel plate flow chamber set up. (d) Micrograph of the experimental set up used for the spheroid perfusion culture including syringe pump, parallel flow chamber, and cell solution, (d3) Microdroplets placed on the petri dish lid for Colo205 spheroid formation.

Figure 3

(a) Micrograph showing Colo205 cells rolling on the PDMS surface functionalized with E-selectin recombinant protein at a shear stress of 2 dyne∕cm². The edges of the microbubble and its triangular top opening are outlined in green. (b) Higher magnification micrograph showing rolling cells entering the microbubble openings. (cl, c2, and c3) Micrographs showing Colo205 cells captured inside the microbubbles and cultured for two days under flow condition at 37 °C and 5% CO₂ at humidified conditions. (dl‚ d2‚ and d3) Micrographs showing Colo205 tumor spheroids cultured inside the microbubbles in 4X, 10X, and 20X, respectively.

Figure 4

Top view (al) and side view (a2) of a spheroid cultured inside a microbubble. The spheroids are pseudo-colored to indicate the distance from the microbubble bottom surface, (b) 3D image of the Colo205 spheroid cultured inside a microbubble by perfusion, stained with Propidium Iodide (PI) and imaged with confocal microscopy for the cell viability assay. Live cells were stained using Invitrogen Celltracker probe. (c) Viability of Colo205 spheroids cultured inside the microbubbles by perfusion culture as a function of seeding duration.

Figure 5

(a) Toxicity of doxorubicin drug on Colo205 cells cultured in monolayers. The percentage of cells viable after 48 h was 16.67%±1.49 SD. (b) Doxorubicin toxicity on Colo205 spheroids cultured in agar-coated plates under static conditions. The percentage of cells viable after 48 h of treatment was 42.85%±3.97 SD. (c) Doxorubicin toxicity on Colo205 spheroids captured and cultured inside microbubbles by perfusion flow. The plotted values are the average number of viable cells in each microbubble. Three microbubbles were used for the perfusion experiments and two microbubbles for the untreated group. The percentage of cells viable after 48 h of perfusion culture was 51.36%±1.93 SD. Doxorubicin at 10 μM concentration was used in all three conditions.

Figure 6

Results of specific cell capture and enrichment study. Colo205 tumor cells stained with Invitrogen Cell Tracker Green were mixed with MDA-MB-231 tumor cells stained with Invitrogen Cell Tracker Blue at different ratios and used for cell rolling and capture inside the bubbles. MDA-MB-231 cells did not roll on E-selectin coated PDMS. The principal mechanism behind the enrichment is that the perfused cells rolling on the PDMS surface have an elevated probability of entering the bubbles compared to non-rolling cells.