Fig 1.
(a) Schematic of a membrane-integrated microfluidic device. (b) Experimental procedure used for coculturing BECs (cells in red) and LECs (cells in green) on opposite sides of the membrane in the microfluidic device.
Fig 2.
Characterization of BECs and LECs cocultured in the microfluidic device.
(a) Schematic of the microfluidic device. The cells in red are BECs and the cells in green are LECs. (b) Live/dead analysis. (c) Detection of the lymphatic endothelium by immunostaining for a lymphatic marker, podoplanin. (d) Detection of an endothelial-specific protein, claudin-5, by immunostaining. (e) Detection of an endothelial-specific adhesion molecule, VE-cadherin, by immunostaining.
Fig 3.
Permeability measured using the microfluidic system lacking or containing BECs and LECs.
No cells (x), BEC monolayer (yellow triangles), BEC-LEC coculture (blue diamonds), and BEC-BEC coculture (pink squares). The values are presented as means ± SD. The fluorescent tracers used were (a) TRITC-dextran and (b) Lucifer Yellow. The statistical significance of the differences between the values obtained at 60 min was evaluated using t tests: (a) P < 0.005, between No cells and BEC monolayer; P < 0.001, between BEC monolayer and BEC-LEC coculture; and P < 0.32, between BEC-LEC coculture and BEC-BEC coculture. (b) P < 0.001, between No cells and BEC monolayer; P < 0.001, between BEC monolayer and BEC-LEC coculture; and P < 0.45, between BEC-LEC coculture and BEC-BEC coculture. Each data point was obtained from three devices (n = 3).
Fig 4.
Permeability measured using the microfluidic system containing BECs and LECs cultured under 3 conditions.
Permeability tests were performed after cells were cultured for 24 h (BEC-LEC coculture) under a 1-μL/h pulsating-flow condition generated using a miniaturized infusion pump (pink triangles), a 1-μL/h continuous-flow condition generated using a syringe pump (yellow squares), or the static condition (blue diamonds). The values are presented as means ± SD. The fluorescent tracers used were (a) TRITC-dextran and (b) Lucifer Yellow. The significance of differences between the 60-min values was assessed by performing t tests: (a) P < 0.05, between static condition and continuous-flow condition; and (b) P < 0.05, between static condition and both flow conditions. Each data point was obtained from four devices (n = 4).
Fig 5.
Effects of histamine stimulation on BECs and LECs cultured in the microfluidic system.
A single-channel device without the permeable membrane was used instead of the coculture device. (a) Fluorescence imaging of BODIPY-FL-histamine bound to its receptor on BECs and LECs. (b) The time course of Ca2+ responses to histamine stimulation observed in BECs and LECs. (c) Immunostaining for claudin-5 and VE-cadherin after histamine stimulation for 30 min.
Fig 6.
Permeability of BECs and LECs measured using the microfluidic system, with or without histamine stimulation.
Permeability tests were performed after a 24-h cell culture (BEC-LEC coculture). Histamine (100 mM, red triangles) and control (blue diamonds). The values are presented as means ± SD. The fluorescent tracers used were (a) TRITC-dextran and (b) Lucifer Yellow. The results of t tests showed that the values obtained at 60 min with and without histamine stimulation were significantly different: (a) P < 0.001 and (b) P < 0.01. Each data point was obtained from seven devices (n = 7).
Fig 7.
Permeability of BECs and LECs stimulated with or without habu snake venom.
Permeability tests were performed after a 24-h cell culture (BEC-LEC coculture). Samples: (yellow squares) and control (blue diamonds). The values are presented as means ± SD. The fluorescent tracers used were (a) TRITC-dextran and (b) Lucifer Yellow. The t-test results showed that the 60-min values obtained without (control) and with stimulation with habu snake venom were significantly different: P < 0.0001, for both (a) and (b). Each data point was obtained from three devices (n = 3).