Ter channel of the device, we flow fluoresceinlabeled dextran (0.1 mM, MW = 10 kDa, Invitrogen) and buffer through the two side channels, and then take the time-lapse fluorescence Title Loaded From File images of all three channels. The fluorescence intensity profile cross all three channels are used to represent the chemical concentration gradients (For more details please see ref. [8]).3D Cell CultureA malignant breast cancer cell line, MDA-MB-231, was obtained as a gift from the Cornell University Center on the Microenvironment and Metastasis. The basal medium for the cell line was DMEM (Invitrogen, Carlsbad, CA), supplemented with 10 FBS (Atlanta Biologicals, Lawrenceville, GA) and antibiotics (100 units penicillin and 100 mg streptomycin, Invitrogen). Cell cultures were maintained every 2? days at a T75 flask (Corning, Lowell, MA) with 5,10 of initial confluency (percentage of cell area coverage) in a humidified, CO2-controlled incubator at 37uC. SDF-1a (10 mg/ml in PBS with 0.1 BSA) and EGF (200 mg/ml in 20 mM acetic acid) were purchased from R D Systems (Minneapolis, MN) and stored at 220uC after reconstitution as instructed by the suppliers. Type I collagen was extracted from rat tails (Pel-Freez, Rogers, AR) using a Title Loaded From File modified protocol [34] and stored at 5 mg/ml in 0.1 acetic acid at 4uC. Cell pellets from 50,75 confluency from T75 cultures were re-suspended in DMEM with 10 FBS and then mixed at 16106 cells/ml with 1N NaOH (for pH,7), 10X M199 and 0.15 collagen on ice. Cell numbers were counted using a hemocytometer (Bright-Line Hemocytometer, Hausser Sci., Horsham, PA). For a typical composition for 500 ml mixture, 150 ml 5 mg/ml collagen, 50 ml 106M199, 3.3 ml 1N NaOH, and 296.7 ml cell culture at 1 million/mL cell concentration were mixed.Gel Filling and Device SetupA volume of 20 ml of cell embedded collagen was introduced into the middle channel of each of the 4 devices using a gel-loading tip. All the inlets and outlets are plugged for preventing slow flow in the center channels during polymerization process. To polymerize the collagen gel, the device was placed in a 37uC incubator for at least 20 minutes of which the device was placed upside down for the first 7 minutes for better distributing cells in the z-direction. Cell distribution in 3D was visually confirmed using a bright field microscope (Nikon Eclipse TS100, Nikon Instruments, Melville, NY) right after the gelation (See Figure 1). Cells were incubated for 24 hours in the device so that cells will have time to attach to the matrix. We start imaging the cells at the same time when the chemical/buffer were introduced in the two side channels where we define t = 0. For a typical experiment, one device was used as a control where media were pumped through both side channels. Flows of three different chemical concentrations and buffers were introduced to the other three source and sink channels respectively. The flows ran at a rate of 1 ml/min through a medical grade tubing (ID = 0.51 mm, PharMed BPT, Cole-Parmer, Vernon Hills, IL) using a syringe pump (KDS230,Materials and Methods Microfluidic Chemoinvasion Device Design and CharacterizationA microfluidic chemoinvasion device previously developed in our lab was modified for this experiment [8,32]. Chemoinvasion here is defined as tumor cell migration within 3D biomatrices under the influences of chemokines and growth factors. Briefly, four three ?parallel ?channel devices were patterned on a 1 mm thick agarose gel membrane using a si.Ter channel of the device, we flow fluoresceinlabeled dextran (0.1 mM, MW = 10 kDa, Invitrogen) and buffer through the two side channels, and then take the time-lapse fluorescence images of all three channels. The fluorescence intensity profile cross all three channels are used to represent the chemical concentration gradients (For more details please see ref. [8]).3D Cell CultureA malignant breast cancer cell line, MDA-MB-231, was obtained as a gift from the Cornell University Center on the Microenvironment and Metastasis. The basal medium for the cell line was DMEM (Invitrogen, Carlsbad, CA), supplemented with 10 FBS (Atlanta Biologicals, Lawrenceville, GA) and antibiotics (100 units penicillin and 100 mg streptomycin, Invitrogen). Cell cultures were maintained every 2? days at a T75 flask (Corning, Lowell, MA) with 5,10 of initial confluency (percentage of cell area coverage) in a humidified, CO2-controlled incubator at 37uC. SDF-1a (10 mg/ml in PBS with 0.1 BSA) and EGF (200 mg/ml in 20 mM acetic acid) were purchased from R D Systems (Minneapolis, MN) and stored at 220uC after reconstitution as instructed by the suppliers. Type I collagen was extracted from rat tails (Pel-Freez, Rogers, AR) using a modified protocol [34] and stored at 5 mg/ml in 0.1 acetic acid at 4uC. Cell pellets from 50,75 confluency from T75 cultures were re-suspended in DMEM with 10 FBS and then mixed at 16106 cells/ml with 1N NaOH (for pH,7), 10X M199 and 0.15 collagen on ice. Cell numbers were counted using a hemocytometer (Bright-Line Hemocytometer, Hausser Sci., Horsham, PA). For a typical composition for 500 ml mixture, 150 ml 5 mg/ml collagen, 50 ml 106M199, 3.3 ml 1N NaOH, and 296.7 ml cell culture at 1 million/mL cell concentration were mixed.Gel Filling and Device SetupA volume of 20 ml of cell embedded collagen was introduced into the middle channel of each of the 4 devices using a gel-loading tip. All the inlets and outlets are plugged for preventing slow flow in the center channels during polymerization process. To polymerize the collagen gel, the device was placed in a 37uC incubator for at least 20 minutes of which the device was placed upside down for the first 7 minutes for better distributing cells in the z-direction. Cell distribution in 3D was visually confirmed using a bright field microscope (Nikon Eclipse TS100, Nikon Instruments, Melville, NY) right after the gelation (See Figure 1). Cells were incubated for 24 hours in the device so that cells will have time to attach to the matrix. We start imaging the cells at the same time when the chemical/buffer were introduced in the two side channels where we define t = 0. For a typical experiment, one device was used as a control where media were pumped through both side channels. Flows of three different chemical concentrations and buffers were introduced to the other three source and sink channels respectively. The flows ran at a rate of 1 ml/min through a medical grade tubing (ID = 0.51 mm, PharMed BPT, Cole-Parmer, Vernon Hills, IL) using a syringe pump (KDS230,Materials and Methods Microfluidic Chemoinvasion Device Design and CharacterizationA microfluidic chemoinvasion device previously developed in our lab was modified for this experiment [8,32]. Chemoinvasion here is defined as tumor cell migration within 3D biomatrices under the influences of chemokines and growth factors. Briefly, four three ?parallel ?channel devices were patterned on a 1 mm thick agarose gel membrane using a si.