This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNB.2015.2455537, IEEE Transactions on NanoBioscience
Towards an endpoint cell motility assay by a microfluidic platform Chuan-Feng Yeh a , Wei Tai, b , Ching-Hui Lin a,c, Duane S. Juang a,d, Chia-Che Wu b, Ya-Wen Chen e and Chia-Hsien Hsu* a,c,f a
b c
d e f
Institute of Biomedical Engineering and Nanomedicine, National Health Research Institutes, Miaoli, Taiwan Department of Mechanical Engineering, National Chung Hsing University, Taichung, Taiwan Ph.D. Program in Tissue Engineering and Regenerative Medicine, National Chung Hsing University, Taichung, Taiwan Department of Life Science, National Tsing Hua University, Hsinchu, Taiwan National Institute of Cancer Research, National Health Research Institutes, Miaoli, Taiwan Institute of NanoEngineering and MicroSystems, National Tsing Hua University, Hsinchu, Taiwan
* To whom correspondence should be addressed: Chia-Hsien Hsu National Health Research Institutes 35, Keyan Road, Zhunan Town Miaoli County,Tawian 35053 Phone: +886-37-246-166 ext. 37105 Fax: +886-37-586-440 E-mail: [email protected] † Electronic Supplementary Information (ESI) available: Supplementary videos 1 and 2. See http://dx.doi.org/10.1039/b000000x/
Abstract In-vitro cell motility assays are frequently used in the study of cell migration in response to anti-cancer drug treatment. Microfluidic systems represent a unique tool for the in-vitro analysis of cell motility. However they usually rely on using time-lapse microscopy to record the spatial temporal locations of the individual cells being tested. This has created a bottleneck for microfluidic systems to perform high-throughput experiments due to requirement of a costly time-lapse microscopy system. Here, we describe the development of a portable microfluidic device for endpoint analysis of cell motility. The reported device incorporates a cell alignment feature to position the seeded cells on the same initial location, so that the cells’ motilities can be analyzed based on their locations at the end of the experiment after the cells have migrated. We show that the device was able to assess cancer cell motility after treatment with a migration inhibitory drug Indole-3-carbinol on MDA-MB-231 breast cancer cells, demonstrating the applicability of our device in screening anti-cancer drug compounds on cancer cells.
1536-1241 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNB.2015.2455537, IEEE Transactions on NanoBioscience
Introduction
Cell movement plays a critical role in many biological events such as embryogenesis[1], wound healing[2], immune responses[3], and cancer progressions[4]. In cancer disease, the motility of tumor cells is responsible for cancer dissemination to form metastases[5, 6] which is the main cause of cancer related deaths.[7] In-vitro cell motility assays are an important tool in studying the regulatory mechanisms of cancer cell migration, and is also routinely used to test the efficacy of potential anticancer drugs at the cell level. Traditionally, cell motility assays are carried out using the trans-well assay[8], or the wound-healing assay[9]. These assays provide an overall estimate of the average motility of the whole cell population, but are not suitable for studying the cellular heterogeneity that exists in the complex tissues of tumors. Cellular heterogeneity is believed to play an important role in cancer disease; for example it has been shown that metastatic cells arise from a specific but smaller population of the tumor cell mass[10]. These cells are significantly different from the rest of the population in terms of mutations[11], gene expression levels[12], drug resistance[13] and metastatic potential[14]. To assess the heterogeneous motility of individual cancer cells, most studies have relied on using time-lapse microscopy to continuously observe individual cells cultured on either conventional petri-dish and well-plates[15] or novel cell culture platforms such as microfluidic devices[16-18] which are advantageous due to their small dimensions, thus offering increased experimental throughput and reduced reagent consumption. However time-lapse microscopy has highequipment cost, is complicated to setup, and requires a lengthy image acquisition time and subsequent image analysis, making it less accessible to most biological laboratories. In addition, because a timelapse microscope system can only perform one experiment at a time, the throughput of the associated experiment is significantly limited by the number of available machines, making it unsuitable for the high throughput requirement of cell-based drug screening assays. Thus, we propose to develop a simple assay to measure cell motility based entirely on the endpoint results of the experiment. To
1536-1241 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNB.2015.2455537, IEEE Transactions on NanoBioscience
achieve this, a cell alignment strategy would be needed to lineup the cells on the same initial starting point, instead of by random seeding, before the migratory assays. Recently, many methods for positioning or patterning cells in a straight line in microfluidic devices have been reported, including centrifugal force[19], gravitational force[20], aspiration[21], sheath flow and hydrodynamic force[22, 23]. Among the above mentioned methods, hydrodynamic force is a comparatively easy, controllable and low shear stress method for cell positioning[24], thus should be suitable for aligning cells in our microfluidic device. In this report, we describe a simple and portable microfluidic device for screening the motilities of single cells in response to drug treatment. Instead of using time-lapse microscopy to record the temporal locations of individual cells to assess cell motility, our method evaluates the motility of a cell by analyzing the cell’s locations in the microfluidic channel from the image taken at the end of the cell migration experiment.
MATERIALS AND METHODS Microfluidic chip fabrication The microfluidic chip was made of polydimethylsiloxane (PDMS) using soft lithography techniques[25], and consists of a cell migration channel connected to a lower draining channel and an upper draining channel via 5 μm wide suction channels (Fig. 1 (a), (b)). The master was fabricated by a negative photoresist (SU-8, MicroChem, Newton, MA, USA) patterned on a silicon wafer with twolayer features. The heights of the two channel layers (3.45 µm and 75.38 µm, respectively) were measured with a surface profilometer (TE-200 Kosaka). PDMS pre-polymer (Sylgard 184, Dow corning) was poured on the master and allowed to cure in a conventional oven at 65 ℃ for 24 h. The cured PDMS replicas were then removed from the masters, and the inlet and outlet holes were punched out using sharpened needles (21 gauge). Chip assembly was performed by a brief oxygen plasma treatment on the PDMS replicas and glass substrates followed by bringing to contact to form bonded
1536-1241 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNB.2015.2455537, IEEE Transactions on NanoBioscience
devices. Type I collagen (BD Bioscience, USA) solution (2 μg ml-1) was used to coat the microchannels at 4℃ overnight prior to the cell experiments to improve cell adhesion to the substrate in the microfluidic device, as it is the most widely used extracellular matrix (ECM) protein for facilitating cell attachment. Cell culture and maintenance MCF7 and MDA-MB-231 breast cancer cells were cultured in RPMI1640 medium (Biowest, France) supplemented with 10% fetal bovine serum (FBS, Biowest, France) and 1% penicillin–streptomycin (Biowest, France) at 37 ℃ and 5% CO2 in a humidified incubator. For cell migration experiments, the cells were grown to 70 – 80% confluence in a 100 mm tissue culture dish (BD Bioscience, USA) and removed from the dish by using trypsin-EDTA (0.25% in PBS, Biowest, France), followed by centrifugation and re-suspension at a concentration of 106 cells per mL in RPMI1640 medium containing 1% FBS. Cell migratory experiment The device consists of two inlet holes at the two ends of the migration channel and one inlet hole for each of the draining channels. One inlet of the migration channel and the inlet of a draining channel were each connected to a medium (RPMI 1640 medium containing 1% FBS) containing reservoir tube via a silicone tubing. By raising both reservoirs the device was filled with the medium. Next, the inlet hole of the draining channel that was not connected to a reservoir was plugged followed by lowering both reservoir tubes to the same level as the device. A MCF7 or MDA-MB-231 cancer cell suspension-loaded pipette tip was then inserted into the inlet of the migration channel followed by lowering the reservoir tube connected to the drain channel. The resulting hydrostatic pressure created a flow which introduced the cells into the cell migration channel and pushes the cells against the channel's wall adjacent to the drain channel. The excess cells were then washed away by a backward flow generated by raising the reservoir container connected to the outlet of the migration channel. The
1536-1241 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNB.2015.2455537, IEEE Transactions on NanoBioscience
inlet of the migration channel was then plugged after the cell loading pipette tip was removed, followed by carefully leveling both reservoir tubes to the same level of the device, disconnecting the tubings and plugging the inlets. The device was then put in a humidified petri-dish and placed in a conventional tissue culture incubator. The operation is schematically depicted in Figure 2. Inhibition of cell motility by I3C drug treatment MDA-MB-231 cells cultured in tissue culture dishes were treated with or without 200 μM indole-3carbinol (I3C) for 24h prior to cell migration experiments. The I3C treated cells were stained with green fluorescent dye (calcein-AM, Invitrogen, USA) whereas the control (no I3C treatment) cells were stained with a red fluorescent dye (Dilc12(3), BD Bioscience, USA) to enable cell identification. Both cell staining were performed by incubating cells with the dye for 15 min at 37℃. After washing off the excess dye and refreshing the culture medium, the cells were removed from the culture dish using standard trypsinization procedure, washed with migratory assay medium, centrifuged down, and re-suspended in migratory assay medium. Cells were loaded and aligned according to our cell alignment procedure as described above. Cell motility analysis Cells in the device were imaged using a Leica AF 6000 LX microscope equipped with a Chargecoupled Device camera (SPOT MAC6000, France). We define the time that cells were aligned and seeded as 0 h. For each experiment sequential images of cells in the device were taken by the camera every 15 min or 1 h for 24 h. Post experiment, the images were analyzed using a cell tracking module of MetaMorph software (MetaMorph, Molecular Devices, USA) to generate results of cell migration trajectory, velocity and velocity of rate (i.e., speed). Statistical analyses Paired Student's t-tests were used for comparison of each group in this study. An asterisk was used to denote statistical significance at *p < 0.05 on figures. All experiments were performed in triplicate.
1536-1241 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNB.2015.2455537, IEEE Transactions on NanoBioscience
RESULTS AND DISCUSSION Device design and operation Our cell motility assay device consists of a cell migration channel connected to a lower draining channel and an upper draining channel via 5 μm wide suction channels (Figure 1 (a), (b)). This design allows for cell alignment to one side of the cell migration channel using hydrostatic pressure-driven flow generated by lowering the reservoir connected to the drain channel. (Figure 2). We chose this cell loading strategy because it does not require an external power source nor special equipment (e.g., syringe pump) and thus the device could be readily operated in a conventional biology laboratory setup. Note that during cell loading the cells in the device could be observed under a microscope, and the amount of hydrostatic force applied to cells needs to be controlled to avoid cell deformation, which can lead to cells entering the suction channels. We found that by using 491 Pa hydrostatic pressure (which translates to a 5 cm height difference between the medium levels in the cell loading tip inserted in the migration channel and the reservoir connected to the drain channel), cell alignment of both MDA-MB-231 and MCF7 cells could be achieved without squeezing the cells into the suction channels. The device design and operation method also results in a free-standing device after cell seeding, allowing the device to be conveniently placed in a conventional tissue culture incubator. Cell migratory assay in the microfluidic chip In our device, the cells are aligned prior to cell migration so each cell's endpoint location may represent its motility. To confirm this hypothesis, we tested our device with two cancer cell linesMDA-MB-231 (Video S1) and MCF7 (Video S2) - and used the MetaMorph software to analyze the migration paths of individual cells from images taken with time-lapse microscopy. As shown in Figure 3 (d) and 4 (d), our results demonstrated that the endpoint transversal positions (y) of the cells in the cell migration chamber were highly correlated with their migratory speeds. Our results also shows that MDA-MB-231 breast cancer cells (Figure 3) exhibits higher cell motility than that of MCF7 cells (Figure 4) confirming the higher aggressiveness of MDA-MB-231 as 1536-1241 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNB.2015.2455537, IEEE Transactions on NanoBioscience
previously reported[26]. From the endpoint image analysis results, the longest migration distance of MDA-MB-231 cells could reach up to 640 – 800 μm with an average y direction velocity of 26.93 μm h-1, whereas MCF7 cells displayed a displacement of less than 320 μm and a y direction velocity of 8.08 μm h-1 (Figure 3 (c) and 4 (c)). Interestingly, we observed that both cell types displayed heterogeneous cell migratory behavior within the populations (Figure 3 (b) and (c) and 4 (b) and (c)). This highlights the applicability of microfluidic device in studying heterogeneous cell motilities at single-cell resolution. Additionally, we observed from the cell migration trajectories that all the cells spontaneously migrated toward the opposite end of the chamber, without the presence of a chemokine attractant (Figure 3 (e) and 4 (e)). This is likely because the cells have an endogenous tendency to migrate away from areas of high cell density (the aligned row of seeded cells) to areas of lower density [23]. It has been reported that cells secrete diffusible autocrine factors for sensing the local density of cells. In areas of high cell density, the secreted factors can serve to inhibit cell proliferation and also act as a chemorepellent, causing cells to migrate away from the region of high cell density[27, 28]. Our results showed that as the cells migrated away from the starting point, a large portion of cells remained behind, resulting in the cell density being higher in the negative y-axis (Figure 5) during the entire cell migration experiment. So even after initial migration, when the cells had become dispersed, the cells continued to migrate primarily in the positive y-axis direction toward the opposite end of the migration channel (Figure 5). Additionally, as the opposite end of the migration channel in our device is connected via suction channels to a large draining channel filled with fresh medium, we believe that another driving force attracting the cells to migrate in the positive y-axis direction is due to the more abundant growth factors and lesser waste molecules present at the opposite end of the channel. Taken together, the migrating cells were simultaneously being chemically attracted to the opposite channel by the fresher medium, and repelled from the higher density of cells from behind. Although the fresh medium in the draining channel adjacent to the aligned cells may also attract cells to migrate in the negative y-axis direction at the beginning, we observed very few cells (in average,
Towards an endpoint cell motility assay by a microfluidic platform Chuan-Feng Yeh a , Wei Tai, b , Ching-Hui Lin a,c, Duane S. Juang a,d, Chia-Che Wu b, Ya-Wen Chen e and Chia-Hsien Hsu* a,c,f a
b c
d e f
Institute of Biomedical Engineering and Nanomedicine, National Health Research Institutes, Miaoli, Taiwan Department of Mechanical Engineering, National Chung Hsing University, Taichung, Taiwan Ph.D. Program in Tissue Engineering and Regenerative Medicine, National Chung Hsing University, Taichung, Taiwan Department of Life Science, National Tsing Hua University, Hsinchu, Taiwan National Institute of Cancer Research, National Health Research Institutes, Miaoli, Taiwan Institute of NanoEngineering and MicroSystems, National Tsing Hua University, Hsinchu, Taiwan
* To whom correspondence should be addressed: Chia-Hsien Hsu National Health Research Institutes 35, Keyan Road, Zhunan Town Miaoli County,Tawian 35053 Phone: +886-37-246-166 ext. 37105 Fax: +886-37-586-440 E-mail: [email protected] † Electronic Supplementary Information (ESI) available: Supplementary videos 1 and 2. See http://dx.doi.org/10.1039/b000000x/
Abstract In-vitro cell motility assays are frequently used in the study of cell migration in response to anti-cancer drug treatment. Microfluidic systems represent a unique tool for the in-vitro analysis of cell motility. However they usually rely on using time-lapse microscopy to record the spatial temporal locations of the individual cells being tested. This has created a bottleneck for microfluidic systems to perform high-throughput experiments due to requirement of a costly time-lapse microscopy system. Here, we describe the development of a portable microfluidic device for endpoint analysis of cell motility. The reported device incorporates a cell alignment feature to position the seeded cells on the same initial location, so that the cells’ motilities can be analyzed based on their locations at the end of the experiment after the cells have migrated. We show that the device was able to assess cancer cell motility after treatment with a migration inhibitory drug Indole-3-carbinol on MDA-MB-231 breast cancer cells, demonstrating the applicability of our device in screening anti-cancer drug compounds on cancer cells.
1536-1241 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNB.2015.2455537, IEEE Transactions on NanoBioscience
Introduction
Cell movement plays a critical role in many biological events such as embryogenesis[1], wound healing[2], immune responses[3], and cancer progressions[4]. In cancer disease, the motility of tumor cells is responsible for cancer dissemination to form metastases[5, 6] which is the main cause of cancer related deaths.[7] In-vitro cell motility assays are an important tool in studying the regulatory mechanisms of cancer cell migration, and is also routinely used to test the efficacy of potential anticancer drugs at the cell level. Traditionally, cell motility assays are carried out using the trans-well assay[8], or the wound-healing assay[9]. These assays provide an overall estimate of the average motility of the whole cell population, but are not suitable for studying the cellular heterogeneity that exists in the complex tissues of tumors. Cellular heterogeneity is believed to play an important role in cancer disease; for example it has been shown that metastatic cells arise from a specific but smaller population of the tumor cell mass[10]. These cells are significantly different from the rest of the population in terms of mutations[11], gene expression levels[12], drug resistance[13] and metastatic potential[14]. To assess the heterogeneous motility of individual cancer cells, most studies have relied on using time-lapse microscopy to continuously observe individual cells cultured on either conventional petri-dish and well-plates[15] or novel cell culture platforms such as microfluidic devices[16-18] which are advantageous due to their small dimensions, thus offering increased experimental throughput and reduced reagent consumption. However time-lapse microscopy has highequipment cost, is complicated to setup, and requires a lengthy image acquisition time and subsequent image analysis, making it less accessible to most biological laboratories. In addition, because a timelapse microscope system can only perform one experiment at a time, the throughput of the associated experiment is significantly limited by the number of available machines, making it unsuitable for the high throughput requirement of cell-based drug screening assays. Thus, we propose to develop a simple assay to measure cell motility based entirely on the endpoint results of the experiment. To
1536-1241 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNB.2015.2455537, IEEE Transactions on NanoBioscience
achieve this, a cell alignment strategy would be needed to lineup the cells on the same initial starting point, instead of by random seeding, before the migratory assays. Recently, many methods for positioning or patterning cells in a straight line in microfluidic devices have been reported, including centrifugal force[19], gravitational force[20], aspiration[21], sheath flow and hydrodynamic force[22, 23]. Among the above mentioned methods, hydrodynamic force is a comparatively easy, controllable and low shear stress method for cell positioning[24], thus should be suitable for aligning cells in our microfluidic device. In this report, we describe a simple and portable microfluidic device for screening the motilities of single cells in response to drug treatment. Instead of using time-lapse microscopy to record the temporal locations of individual cells to assess cell motility, our method evaluates the motility of a cell by analyzing the cell’s locations in the microfluidic channel from the image taken at the end of the cell migration experiment.
MATERIALS AND METHODS Microfluidic chip fabrication The microfluidic chip was made of polydimethylsiloxane (PDMS) using soft lithography techniques[25], and consists of a cell migration channel connected to a lower draining channel and an upper draining channel via 5 μm wide suction channels (Fig. 1 (a), (b)). The master was fabricated by a negative photoresist (SU-8, MicroChem, Newton, MA, USA) patterned on a silicon wafer with twolayer features. The heights of the two channel layers (3.45 µm and 75.38 µm, respectively) were measured with a surface profilometer (TE-200 Kosaka). PDMS pre-polymer (Sylgard 184, Dow corning) was poured on the master and allowed to cure in a conventional oven at 65 ℃ for 24 h. The cured PDMS replicas were then removed from the masters, and the inlet and outlet holes were punched out using sharpened needles (21 gauge). Chip assembly was performed by a brief oxygen plasma treatment on the PDMS replicas and glass substrates followed by bringing to contact to form bonded
1536-1241 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNB.2015.2455537, IEEE Transactions on NanoBioscience
devices. Type I collagen (BD Bioscience, USA) solution (2 μg ml-1) was used to coat the microchannels at 4℃ overnight prior to the cell experiments to improve cell adhesion to the substrate in the microfluidic device, as it is the most widely used extracellular matrix (ECM) protein for facilitating cell attachment. Cell culture and maintenance MCF7 and MDA-MB-231 breast cancer cells were cultured in RPMI1640 medium (Biowest, France) supplemented with 10% fetal bovine serum (FBS, Biowest, France) and 1% penicillin–streptomycin (Biowest, France) at 37 ℃ and 5% CO2 in a humidified incubator. For cell migration experiments, the cells were grown to 70 – 80% confluence in a 100 mm tissue culture dish (BD Bioscience, USA) and removed from the dish by using trypsin-EDTA (0.25% in PBS, Biowest, France), followed by centrifugation and re-suspension at a concentration of 106 cells per mL in RPMI1640 medium containing 1% FBS. Cell migratory experiment The device consists of two inlet holes at the two ends of the migration channel and one inlet hole for each of the draining channels. One inlet of the migration channel and the inlet of a draining channel were each connected to a medium (RPMI 1640 medium containing 1% FBS) containing reservoir tube via a silicone tubing. By raising both reservoirs the device was filled with the medium. Next, the inlet hole of the draining channel that was not connected to a reservoir was plugged followed by lowering both reservoir tubes to the same level as the device. A MCF7 or MDA-MB-231 cancer cell suspension-loaded pipette tip was then inserted into the inlet of the migration channel followed by lowering the reservoir tube connected to the drain channel. The resulting hydrostatic pressure created a flow which introduced the cells into the cell migration channel and pushes the cells against the channel's wall adjacent to the drain channel. The excess cells were then washed away by a backward flow generated by raising the reservoir container connected to the outlet of the migration channel. The
1536-1241 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNB.2015.2455537, IEEE Transactions on NanoBioscience
inlet of the migration channel was then plugged after the cell loading pipette tip was removed, followed by carefully leveling both reservoir tubes to the same level of the device, disconnecting the tubings and plugging the inlets. The device was then put in a humidified petri-dish and placed in a conventional tissue culture incubator. The operation is schematically depicted in Figure 2. Inhibition of cell motility by I3C drug treatment MDA-MB-231 cells cultured in tissue culture dishes were treated with or without 200 μM indole-3carbinol (I3C) for 24h prior to cell migration experiments. The I3C treated cells were stained with green fluorescent dye (calcein-AM, Invitrogen, USA) whereas the control (no I3C treatment) cells were stained with a red fluorescent dye (Dilc12(3), BD Bioscience, USA) to enable cell identification. Both cell staining were performed by incubating cells with the dye for 15 min at 37℃. After washing off the excess dye and refreshing the culture medium, the cells were removed from the culture dish using standard trypsinization procedure, washed with migratory assay medium, centrifuged down, and re-suspended in migratory assay medium. Cells were loaded and aligned according to our cell alignment procedure as described above. Cell motility analysis Cells in the device were imaged using a Leica AF 6000 LX microscope equipped with a Chargecoupled Device camera (SPOT MAC6000, France). We define the time that cells were aligned and seeded as 0 h. For each experiment sequential images of cells in the device were taken by the camera every 15 min or 1 h for 24 h. Post experiment, the images were analyzed using a cell tracking module of MetaMorph software (MetaMorph, Molecular Devices, USA) to generate results of cell migration trajectory, velocity and velocity of rate (i.e., speed). Statistical analyses Paired Student's t-tests were used for comparison of each group in this study. An asterisk was used to denote statistical significance at *p < 0.05 on figures. All experiments were performed in triplicate.
1536-1241 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNB.2015.2455537, IEEE Transactions on NanoBioscience
RESULTS AND DISCUSSION Device design and operation Our cell motility assay device consists of a cell migration channel connected to a lower draining channel and an upper draining channel via 5 μm wide suction channels (Figure 1 (a), (b)). This design allows for cell alignment to one side of the cell migration channel using hydrostatic pressure-driven flow generated by lowering the reservoir connected to the drain channel. (Figure 2). We chose this cell loading strategy because it does not require an external power source nor special equipment (e.g., syringe pump) and thus the device could be readily operated in a conventional biology laboratory setup. Note that during cell loading the cells in the device could be observed under a microscope, and the amount of hydrostatic force applied to cells needs to be controlled to avoid cell deformation, which can lead to cells entering the suction channels. We found that by using 491 Pa hydrostatic pressure (which translates to a 5 cm height difference between the medium levels in the cell loading tip inserted in the migration channel and the reservoir connected to the drain channel), cell alignment of both MDA-MB-231 and MCF7 cells could be achieved without squeezing the cells into the suction channels. The device design and operation method also results in a free-standing device after cell seeding, allowing the device to be conveniently placed in a conventional tissue culture incubator. Cell migratory assay in the microfluidic chip In our device, the cells are aligned prior to cell migration so each cell's endpoint location may represent its motility. To confirm this hypothesis, we tested our device with two cancer cell linesMDA-MB-231 (Video S1) and MCF7 (Video S2) - and used the MetaMorph software to analyze the migration paths of individual cells from images taken with time-lapse microscopy. As shown in Figure 3 (d) and 4 (d), our results demonstrated that the endpoint transversal positions (y) of the cells in the cell migration chamber were highly correlated with their migratory speeds. Our results also shows that MDA-MB-231 breast cancer cells (Figure 3) exhibits higher cell motility than that of MCF7 cells (Figure 4) confirming the higher aggressiveness of MDA-MB-231 as 1536-1241 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNB.2015.2455537, IEEE Transactions on NanoBioscience
previously reported[26]. From the endpoint image analysis results, the longest migration distance of MDA-MB-231 cells could reach up to 640 – 800 μm with an average y direction velocity of 26.93 μm h-1, whereas MCF7 cells displayed a displacement of less than 320 μm and a y direction velocity of 8.08 μm h-1 (Figure 3 (c) and 4 (c)). Interestingly, we observed that both cell types displayed heterogeneous cell migratory behavior within the populations (Figure 3 (b) and (c) and 4 (b) and (c)). This highlights the applicability of microfluidic device in studying heterogeneous cell motilities at single-cell resolution. Additionally, we observed from the cell migration trajectories that all the cells spontaneously migrated toward the opposite end of the chamber, without the presence of a chemokine attractant (Figure 3 (e) and 4 (e)). This is likely because the cells have an endogenous tendency to migrate away from areas of high cell density (the aligned row of seeded cells) to areas of lower density [23]. It has been reported that cells secrete diffusible autocrine factors for sensing the local density of cells. In areas of high cell density, the secreted factors can serve to inhibit cell proliferation and also act as a chemorepellent, causing cells to migrate away from the region of high cell density[27, 28]. Our results showed that as the cells migrated away from the starting point, a large portion of cells remained behind, resulting in the cell density being higher in the negative y-axis (Figure 5) during the entire cell migration experiment. So even after initial migration, when the cells had become dispersed, the cells continued to migrate primarily in the positive y-axis direction toward the opposite end of the migration channel (Figure 5). Additionally, as the opposite end of the migration channel in our device is connected via suction channels to a large draining channel filled with fresh medium, we believe that another driving force attracting the cells to migrate in the positive y-axis direction is due to the more abundant growth factors and lesser waste molecules present at the opposite end of the channel. Taken together, the migrating cells were simultaneously being chemically attracted to the opposite channel by the fresher medium, and repelled from the higher density of cells from behind. Although the fresh medium in the draining channel adjacent to the aligned cells may also attract cells to migrate in the negative y-axis direction at the beginning, we observed very few cells (in average,
