$$\rightleftharpoonup{xx}$$
$$\longleftharp{xx}$$,
$$\longrightharp{xx}$$,
Fabricated PDMS-PC hybrid microfluidic cell culture device. Fig. 1 shows a photo and an illustration of the microfluidic device. The bottom layer contains four levels of serpentine-shaped channels to generate solutions from reagents introduced from two separated inlets with six different mixing ratios. Theoretically, the six different mixing ratios are 1:0, 4:1, 3:2, 2:3, 1:4, and 0:1 (left:right) between the two solutions introduced from the inlets. The chemical gradients constructed by the six different mixing-ratio solutions can be generated in the cell culture chamber, located downstream. The top and bottom layers are separated by a PDMS membrane. In the top layer, the reagents for the oxygen scavenging chemical reaction are introduced into the microfluidic channel from two separate inlets. The reagents are mixed with each other for the reaction immediately before flowing on top of the cell culture chamber to scavenge the oxygen from the bottom channel, without direct chemical contact. The embedded PC film, with a smaller gas diffusion coefficient compared to PDMS, acts as a diffusion barrier that makes oxygen scavenging more efficient. Oxygen gradually diffuses back to the cell culture chamber through the PDMS in the downstream area to form an oxygen gradient along the flow direction. Since the oxygen scavenging chemical reaction is spatially confined, only local oxygen tensions are affected. As a result, the device can be utilized in a conventional cell incubator without altering its global oxygen tension. In the migration experiments, cells are seeded inside the cell culture chamber for observation. The growth medium and chemical reagents are introduced to the device using syringe pumps with precisely controlled flow rates.
Characterization of chemical and oxygen gradients generated inside the device. Due to the laminar flow nature of microfluidics, flow behaviors can be predicted using computational fluidic dynamics (CFD) simulation. In this paper, we constructed a 3D model and performed the simulation using a commercially available multiphysics modeling software. Fig. 2(a) shows a comparison between experimentally characterized fluorescein concentration profiles across the width of the cell culture chamber based on fluorescence intensity measurements and the numerical simulation results. The agreement between the experimental and simulation results suggests that the CFD model can well estimate the chemical gradients generated inside the device. Fig. 2(b) plots the simulated SDF-1α gradient generated in the cell culture chamber. Fig. 3 shows the oxygen gradient characterization results by flowing the oxygen-sensitive fluorescence dye inside the cell culture chamber before the cell experiments. The result indicates that an oxygen gradient, ranging from about 1 to 16%, can be established using the aforementioned protocol.
Cell migration results. As a demonstration, we performed A549 cell migration studies under 4 combinations of chemokine (SDF-1α) and oxygen gradients: (1) no chemokine and no oxygen gradients as a control, (2) with a chemokine gradient and without an oxygen gradient, (3) with an oxygen gradient and without a chemokine gradient, and (4) with both chemokine and oxygen gradients. Fig. 4 shows the photo of the entire experimental setup. The experiments were all performed in a conventional cell culture incubator with the entire setup (including the microfluidic devices, syringe pumps, and live cell imaging microscopes) placed inside it. The cell migration results are shown in Fig. 5. Fig. 5(a) shows the images collected during the experiments using the live cell imaging analyzer, and Fig. 5(b) and (c) plots the cell migration trajectories and average movements under the four combinations analyzed by the ImageJ software with the plugins. The results show that the average cell migration distance in the control approaches zero, which suggests random movement of the cells in the experiment. In contrast, with only the chemokine gradient, the average movement of the cells is towards the left, where the SDF-1α concentration is higher. The results suggest the SDF-1α chemotaxis behavior of A549 cells, which has been previously reported. In the experiment with only oxygen gradients, the average movement of the cells is upwards, where the oxygen tension is lower. More interestingly, in the experiment with perpendicular chemokine and oxygen gradients, the average movement of the cells is upwards and without any obvious movement in the horizontal direction (chemokine gradient direction).

Figure 1: Fabricated PDMS-PC microfluidic cell culture device. (a) The experimental photo of the fabricated device capable of reliably generating perpendicular chemical and oxygen gradients for cell migration studies. The chemical gradient channel is filled with blue and yellow food dyes to demonstrate the gradient generation inside the cell culture chamber. The oxygen gradient channel is filled with red food dye. The scale bar is 1 cm. (b) The schematic of the microfluidic device. The top layer is fabricated using PDMS with an embedded PC layer as a gas diffusion barrier for efficient oxygen gradient control inside the cell culture chamber. (c) The master molds for the fabrication of the top and bottom layers. Please click here to view a larger version of this figure.

Figure 2: Chemical gradient inside the microfluidic cell culture device. (a) Numerically simulated and experimentally characterized fluorescein concentration gradient inside the cell culture chamber across the width of the cell culture chamber (Y direction). The similarity between the simulated and experimentally measured gradients indicates that the simulation can well predict the chemical gradient. The figure inset shows the three-dimensional (3D) model constructed for the simulation. (b) Numerical simulation result of the SDF-1α chemokine gradient across the width of the cell culture chamber for cell migration studies. Please click here to view a larger version of this figure.

Figure 3: Oxygen gradient inside the microfluidic cell culture device. Experimentally measured oxygen gradients within the cell culture chamber along the flow direction. The gradients were estimated using the oxygen-sensitive fluorescence dye and image analysis. The gradients, from left to right of the chamber, are characterized, and the results show consistent gradient profiles across the width of the chamber.

Figure 4: Photos of the experimental setup. The entire setup, including microfluidic devices, syringe pumps, and a live cell imaging microscope, is placed inside a conventional cell culture incubator for optimized cell culture conditions during the experiments. Please click here to view a larger version of this figure.

Figure 5: Cell migration study results under perpendicular SDF-1α and oxygen gradients. (a) Images captured before and after the 12-h cell migration study. The cell migration paths can be analyzed from the captured time-lapsed images using the live cell imaging microscope. (b) The cell migration paths and the analyzed average migration movement from the captured images under 4 different gradient combinations: no gradient, only chemokine gradient, only oxygen gradient, and both chemokine and oxygen gradients. The images were captured every 15 min. The scale bar is 250 µm. (c) Plots of the average cell migration distances in the perpendicular (oxygen gradient) and horizontal (chemokine gradient) directions under four different gradient combinations. The data are expressed as the mean ± SD, obtained from three independent experimental sets, and 10 cells were analyzed in each experiment. The statistical significantly different (unpaired Student's t-test, p < 0.01) results are designated by different letters (a and b). Please click here to view a larger version of this figure.