TY - JOUR
AU - Funamoto,, Kenichi
AB - Abstract The hypoxic microenvironment existing in vivo is known to significantly affect cell morphology and dynamics, and cell group behaviour. Collective migration of vascular endothelial cells is essential for vasculogenesis and angiogenesis, and for maintenance of monolayer integrity. Although hypoxic stress increases vascular endothelial permeability, the changes in collective migration and intracellular junction morphology of vascular endothelial cells remain poorly understood. This study reveals the migration of confluent vascular endothelial cells and changes in their adherens junction, as reflected by changes in the vascular endothelial (VE)-cadherin distribution, under hypoxic exposure. Vascular endothelial monolayers of human umbilical vein endothelial cells (HUVECs) were formed in microfluidic devices with controllability of oxygen tension. The oxygen tension was set to either normoxia (21% O2) or hypoxia (<3% O2) by supplying gas mixtures into separate gas channels. The migration velocity of HUVECs was measured using particle image velocimetry with a time series of phase-contrast microscopic images of the vascular endothelial monolayers. Hypoxia inducible factor-1α (HIF-1α) and VE-cadherin in HUVECs were observed after exposure to normoxic or hypoxic conditions using immunofluorescence staining and quantitative confocal image analysis. Changes in the migration speed of HUVECs were observed in as little as one hour after exposure to hypoxic condition, showing that the migration speed was increased 1.4-fold under hypoxia compared to that under normoxia. Nuclear translocation of HIF-1α peaked after the hypoxic gas mixture was supplied for 2 h. VE-cadherin expression was also found to be reduced. When ethanol was added to the cell culture medium, cell migration increased. By contrast, by strengthening VE-cadherin junctions with forskolin, cell migration decreased gradually in spite the effect of ethanol to stimulate migration. These results indicate that the increase of cell migration by hypoxic exposure was attributable to loosening of intercellular junction resulting from the decrease of VE-cadherin expression. collective cell migration, forskolin, hypoxia, microfluidic device, vascular endothelial monolayer, VE-cadherin Insight, innovation, integration Collective migration of vascular endothelial cells is fundamentally important for vasculogenesis and angiogenesis, and for maintenance of the monolayer integrity. Changes of the collective cell migration and adherens junction VE-cadherin by hypoxic exposure are investigated with monolayers of human umbilical vein endothelial cells formed in microfluidic devices with controllability of oxygen tension. Particle image velocimetry analysis of the monolayers has revealed increased cell migration under hypoxia. Immunofluorescence staining shows nuclear translocation of hypoxia inducible factor-1α and a decreasing tendency in VE-cadherin after hypoxic gas mixture is supplied. By contrast, strengthening of VE-cadherin with forskolin decreases cell migration. Consequently, loosening of the intercellular junction by hypoxic exposure increases cell migration. INTRODUCTION Oxygen tensions in vivo are well known to be lower than of atmospheric air [1, 2]. Cells sense hypoxic microenvironments through the inhibition of degradation of hypoxia inducible factor-1α (HIF-1α) and the subsequent increase of its concentration, leading to changes in their morphology and dynamics [3]. Collective cell migration [4, 5], by which adjacent cells exert forces on each other to move as a cluster, also changes [6, 7] and has important implications for various biological phenomena such as development [8], tissue regeneration [9–11] and tumour progression [12, 13]. To elucidate wound healing and tumour metastasis, the migration of leader cells or border cells in a group of epithelial cells has been investigated extensively, showing that collective cell migration depends on substrate stiffness [14], geometrical confinement [15, 16] and cell density [15, 17]. Moreover, the maturity of cell–cell and cell–extracellular matrix (ECM) adhesions that are responsible for the transduction mechanical forces of the cells (or mechanobiology [18]) affects the collective cell migration [19]. Intercellular adherens junctions based on E-cadherin are particularly important as influencing factors [20–22]. Similarly to epithelial cells, vascular endothelial cells, which form a monolayer covering the blood vessel lumen, show collective cell migration by forming a cluster of migrating cells with coherent velocity and direction [23]. Their collective migration is fundamentally important for vasculogenesis and angiogenesis [24, 25], and for the maintenance of monolayer integrity [26, 27]. When the vascular endothelial monolayer is scratched, the cells at the leading edge migrate faster than those at the inner region of the monolayer to cover the wounded area [27]. In addition, the vascular endothelial monolayer exposed to hypoxic stress reacts by releasing inflammatory mediators and growth factors and by changes in gene expression in accordance with the exposure time [28]. Wound healing assays revealed that hypoxic stress at some specific oxygen levels (1% O2 [29, 30], and 10% O2 but not 8% O2 [31]) promoted the migration of leader cells at the edge of the vascular endothelial monolayer to cover the scratched area. Hypoxic microenvironments also activate vascular endothelial angiogenic responses [32], but transient hypoxia and reoxygenation induced by ischaemia-reperfusion damage vascular endothelial monolayers, creating a risk of bleeding [33]. Investigation of the behaviours of vascular endothelial monolayers exposed to hypoxic stress is therefore key to elucidating the mechanisms of homoeostasis and pathological events. Cellular experiments using microfluidic devices enable high-resolution, real-time observation of individual cells cultured in three-dimensional spaces. Precise control over mechanical stimuli such as normal and shear stresses, and chemical stimuli by biochemical substances and drugs can also be provided to the cells. Microfluidic devices have been developed that control the oxygen tension inside the device precisely in order to facilitate observation of cell behaviours in hypoxic microenvironments [1, 34]. The authors previously reported a microfluidic device capable of controlling the oxygen tension in a three-dimensional microenvironment [35]. By supplying gas mixtures at predefined oxygen concentrations to gas channels located adjacent to the other channels for hydrogel and cell culture medium, both uniform hypoxic conditions (as low as 3% O2) and oxygen gradients were demonstrated. The microfluidic device facilitated rapid changes in oxygen tension around the cells, as needed in certain types of experiment, compared with the conventional methods such as hypoxia chambers and multi-gas incubators. Also, since the volume of gas needed to control the oxygen tension was small, use of the microfluidic device had the added advantage that it greatly reduced the volume of gas mixtures required. The device was used to measure vascular endothelial permeability under controlled oxygen tensions, demonstrating that hypoxic exposure impaired size-selective vascular endothelial barrier function observed under normoxic conditions [36]. Our study confirmed a decrease in the adherence junction of VE-cadherin to intracellular structures such as the cytoskeleton, cell membrane and nucleus. Because the intercellular adherens junctions, as reflected in the dynamics of VE-cadherins, are known to contribute to the collective migration of vascular endothelial cells [25, 37, 38], we speculate that it might also affect the migration of cells in confluency. However, some aspects of the changes of VE-cadherin in hypoxic conditions remain unclear. Specifically, changes of the migration of confluent vascular endothelial cells and the intercellular VE-cadherin in the early stage of hypoxic exposure are poorly understood. This study investigates changes of cell migration and the intracellular junction of vascular endothelial cells when vascular endothelial monolayers are exposed to hypoxia. Vascular endothelial monolayers are formed in microfluidic devices with oxygen tension controllability [36]. The vascular endothelial cell behaviours are observed with time-lapse imaging while controlling oxygen tension. Time series of phase-contrast microscope images of the vascular endothelial monolayer are analysed using particle image velocimetry (PIV) to measure the cell migration velocity. Immunofluorescence imaging is performed to elucidate changes of HIF-1α and VE-cadherin in the vascular endothelial cells after hypoxic exposure. The obtained results reveal the effects of loose intercellular adhesion of VE-cadherin in hypoxia and the resultant increase of migration of vascular endothelial cells. METHODS Microfluidic device A microfluidic device with oxygen tension controllability was used for this study (Fig. 1) [35]. The device is fabricated from poly(dimethylsiloxane) (PDMS), a polycarbonate (PC) film and a glass coverslip. Inside the device, a gel channel (1300 μm wide) into which hydrogel is filled, mimicking an extracellular matrix, is flanked by a Y-shaped media channel (500 μm wide). On both sides of the media channels are located a pair of gas channels (500 μm wide) into which gas mixtures are supplied for controlling oxygen tension. The media and gas channels are separated by 150 μm wide PDMS partitions, through which gas diffuses. The heights of all channels are 150 μm. The device diameter and thickness are, respectively, 52 mm and 6 mm. Additionally, a PC film (0.5 mm thick and 35 mm diameter) is embedded in the device at a distance of 2 mm from the bottom glass coverslip to prevent oxygen diffusion into the device from the atmosphere. During device fabrication, the channel pattern is produced on the silicon wafer using photolithography. It is then transferred to PDMS (Silgard 184 Silicone Elastomer Kit; Dow Corning Corp., USA) of 2 mm thick using soft lithography [39]. The channel pattern-transferred PDMS layer is placed on a cell culture dish. Then a PC film with holes to access the inlets and outlets of channels is placed on top of the layer. After additional PDMS is poured over the PC film, the PDMS layer is cured in an oven overnight. The cured PDMS is taken out of the dish and punched to create inlets and outlets. Then it is bonded to a glass coverslip by plasma treatment. The gel and media channels are coated with poly d-lysine (PDL) for better adhesion of ECM. The PDL hydrobromide solution of 1 mg/ml (P7886; Sigma-Aldrich Corp., USA) is infused into the gel and media channels, and the device is placed in an incubator at 37°C and 5% CO2 for more than 4 h. The gel and media channels are then washed twice with distilled water, followed by drying of the channels overnight in an oven at 70°C. Collagen type I gel solution (354 236; Corning Inc., USA) prepared to be 2.5 mg/ml and pH 7.4–7.5 is injected into the gel channel and polymerized in an incubator for 40 min. Cell culture medium (EGM-2, CC-3156; Lonza Group Ltd., Switzerland) is injected into the media channel. Then the device is incubated overnight for mechanical stabilization of the gel. Before cell seeding, the media channel is coated with Matrigel at 2.0 mg/ml for better adhesion for the cells. Figure 1. Open in new tabDownload slide Schematic diagram showing the experiment using the microfluidic device with the controllability of oxygen tension: (upper left) an overhead view of the device, (upper right) an enlarged view of the channel pattern, (lower left) a cross-sectional view of the device, and (lower right) HUVEC migration in a monolayer. The central gel channel (green) is sandwiched between media channels (purple), which merge downstream. The gas channels (blue) are placed next to the media channels with PDMS partitions between them. A polycarbonate film is embedded above the channels in the device. A monolayer of HUVECs is formed on the right-hand side media channel. Figure 1. Open in new tabDownload slide Schematic diagram showing the experiment using the microfluidic device with the controllability of oxygen tension: (upper left) an overhead view of the device, (upper right) an enlarged view of the channel pattern, (lower left) a cross-sectional view of the device, and (lower right) HUVEC migration in a monolayer. The central gel channel (green) is sandwiched between media channels (purple), which merge downstream. The gas channels (blue) are placed next to the media channels with PDMS partitions between them. A polycarbonate film is embedded above the channels in the device. A monolayer of HUVECs is formed on the right-hand side media channel. Cell experiments Human umbilical vein endothelial cells (HUVECs, C2519A; Lonza Group Ltd., Switzerland) from the fourth and ninth passages were used. They were seeded only in the right-hand side media channel for this study [36]. By changing the normal cell culture medium of EGM-2 every day for 4 days, an endothelial monolayer was formed covering the media channel three dimensionally. The cell culture medium was changed to that without/with 10 μM forskolin (F6886; Sigma-Aldrich Corp., USA) prepared with ethanol (EtOH). Forskolin enhances the expression of VE-cadherin through increased production of cAMP [40, 41]. Additionally, cell culture medium containing 0.08% (14 mM) EtOH (solvent of FSK) was used for the other experiments since EtOH is known to promote the angiogenic activity of vascular endothelial cells [42, 43]. The device was then placed on a stage inside an incubator (INUBSF-ZILCS; Tokai HIT Co., Ltd., Japan) at 37°C and 5% CO2, which was mounted on a microscope (EVOS FL Cell Imaging System; Life Technologies Inc., USA). Humidified gas mixtures containing either 21% O2, 5% CO2, 74% N2, or 5% CO2, 95% N2, were then supplied into both gas channels to create a uniform normoxic (21% O2) or hypoxic (<3% O2) condition in the media channel (Fig. S1) [35]. The gas flow rate for each gas channel was set at 18 ml/min, found previously to be sufficient to control oxygen tensions. Phase-contrast microscope images of the vascular endothelial monolayers on the bottom glass coverslip were captured at 10 min intervals for 5 h. Particle image velocimetry analysis The migration velocity of the HUVECs was found from PIV analysis of a time-series of phase-contrast microscope images obtained using time-lapse imaging [18, 27]. PIV analysis was originally invented to measure the flow velocity from sequential images of particles distributed in flow. When PIV is applied to time-lapse microscope images of cells, the cell itself as well as its organelle (nucleus, mitochondria, endoplasmic reticulum and so on) are considered as the ‘particle.’ The PIV analysis was conducted with JPIV open-source software [44]. In the PIV analysis, a rectangular region of 512 × 960 pixels (440 × 825 μm) on the right-hand side media channel was divided into small regions of interest (ROIs) of 8 × 8 pixels (6.88 × 6.88 μm). Then, the displacement of each ROI between the images at two time points was measured by calculating the cross-correlation function. Here, images taken in the first hour were excluded from measurements because of the device movement and deviation of the focus due to a combination of thermal equilibration and the transient state of oxygen tension. Immunofluorescence imaging After supplying gas mixtures either for 0, 1, 1.5, 2, 3 or 5 h, the HUVECs were fixed with 4% paraformaldehyde phosphate buffer solution (163-20 145; Wako Pure Chemical Industries Ltd., Japan) for 10 min. The cells were then permeabilized with phosphate buffered saline (PBS) (P5119; Sigma-Aldrich Corp., USA) solution containing 0.1% Triton X-100 (Pharmacia Biotech, Sweden) for 5 min, and were blocked with PBS containing 1% Block Ace (DS Pharma Biomedical Co. Ltd., Japan) (BA-PBS) for 30 min to prevent non-specific absorption of the antibodies described below. HIF-1α or VE-cadherin was labelled with a mouse monoclonal antibody (ab1; Abcam plc., USA, or sc-9989; Santa Cruz Biotechnology Inc., USA) at a dilution of 1:100 in BA-PBS for 1 h, with subsequent staining with Alexa Fluor 488-conjugated goat anti-mouse IgG antibody (A11001; Invitrogen Corp., USA) or Alexa Fluor 594-conjugated goat anti-mouse IgG antibody (A11032; Invitrogen Corp., USA) at a dilution of 1:100 in PBS, respectively, for 1 h. Cell nuclei were stained with DAPI (D21490; Thermo Fisher Scientific Inc., USA) at 1 μg/ml. The immunofluorescence staining was conducted at room temperature. Between the individual processes, cells were washed twice with PBS. Imaging of the cells on the bottom glass coverslip of the media channel was performed using a confocal microscope (LSM800; Carl Zeiss Microscopy GmbH, Germany). Twenty confocal microscope images of HUVECs on the horizontal plane (xy-plane) were taken at 0.63 μm intervals in the vertical (z-axis) direction. Then, maximum intensity projections of the images to the xy-plane were created and analysed using open-source software (ImageJ, National Institutes of Health, USA). Data analysis and statistical analysis For each oxygen condition in normal cell culture medium, cell migration was measured using nine microfluidic devices, whereas four devices were measured using the same cell culture medium with addition of FSK or EtOH. Microscopic observation of the cells by immunofluorescence staining was performed with four devices for each target protein and each level of oxygen tension. Images of five arbitrarily determined locations on the bottom glass coverslip were obtained from one device: 20 images were obtained for each condition. Nuclear translocation of HIF-1α was assessed using image analysis. Fluorescence of HIF-1α in the nucleus was extracted, referring to the maximum intensity projection of confocal microscope images of DAPI. The average intensities of the fluorescence, Īnucleus and Īwhole, in the nuclei and in the whole image, respectively, were calculated. The relative average intensity Īnucleus/Īwhole was then evaluated as the ratio of the nuclear translocation of HIF-1α. Moreover, VE-cadherin of the cells was quantified by selecting three cells randomly from each microscope image of VE-cadherin, i.e., a total of 60 cells for each oxygen condition. The cell area Ain surrounded by VE-cadherin and the overlapping area Acad of the cell–cell junction were measured to calculate the fractional VE-cadherin area A*cad (= Acad/(Ain + Acad)) [45]. Significant differences between results under normoxia and hypoxia were assessed using two-tailed Welch’s t tests. Statistical significance was inferred for P < 0.05. RESULTS Increased migration of vascular endothelial cells due to hypoxic exposure Velocity vectors of HUVECs were obtained from PIV analysis using a time-series of phase-contrast microscope images (Fig. 2(a)). Regions with high migration speed (red vectors) and those with low migration speed (blue vectors) were distributed randomly (Fig. 2(b) and (c)). Comparison of the images in the two oxygen conditions represented a large number of regions with high migration speed under hypoxic condition. For more detailed evaluation, the frequency distributions of cell migration speed measured by the displacements of the ROIs were assessed every 30 min (Fig. 2(d) and (e)). The ROIs with low migration speed were more prevalent in both oxygen conditions, but only in hypoxia, regions with high migration speed greater than 20 μm/h were detected. The distribution of migration speed under hypoxia tended over time to approach the result obtained under normoxia because of the increasing numbers of ROIs with low migration speed. Variations in the migration speeds of HUVECs under the two oxygen conditions are presented for comparison in Fig. 2(f). Linear least squares regression of migration speed over the time under hypoxia yielded a negative regression coefficient of −0.20 (coefficient of determination R2 = 0.50), whereas that under normoxia was nearly zero with the value of −0.01 (R2 = 0.002). Thus, the cell migration speeds decreased gradually as time progressed under hypoxia but were almost constant under normoxia. At every time point, the average migration speed of the HUVECs was ~1.4-fold faster for hypoxia than for normoxia. Spatial extent of cells migrating in the same direction was evaluated by calculating the spatial autocorrelation function Cvv of the velocity fluctuation vectors δv obtained by subtracting the spatial average velocity v̅ from the velocity vector v: Cvv=⟨∑iδv(ri)⋅δv(ri+r)∑iδv(ri)⋅δv(ri)⟩,(1) where ri is the positions where velocity vector of cell migration was measured, and the angle brackets denote an average over all directions and time [46]. The autocorrelation function exponentially decreased over more than 100 μm under both oxygen conditions (Fig. 2(g)). By hypoxic exposure, migration speed of HUVECs was increased, but the cluster size of cells migrating in the same direction was little changed. Figure 2. Open in new tabDownload slide Migration of HUVECs on the media channel in the microfluidic device. (a) Representative enlarged phase-contrast images starting one hour after the supply of normoxic gas mixture (21% O2) at 30 min intervals showing resultant velocity vectors of migration calculated by PIV with the sequential images. Velocity vectors superimposed on phase-contrast microscope images under (b) normoxic and (c) hypoxic conditions between 2 h and 2.5 h. Respective upper left panels are enlargements of the rectangular region positioned in the media channel. Stacked bar charts of cell migration speed measured by the displacements of ROIs within 30 min under (d) normoxic and (e) hypoxic conditions. (f) Variation of the migration speed and its linear least squares regression. Data shown are presented as mean ± standard deviation. *P < 0.05, **P < 0.01. (g) Autocorrelation function of velocity fluctuation vectors (mean ± standard deviation). The inset shows the decay of the velocity correlation function on a logarithmic scale. Figure 2. Open in new tabDownload slide Migration of HUVECs on the media channel in the microfluidic device. (a) Representative enlarged phase-contrast images starting one hour after the supply of normoxic gas mixture (21% O2) at 30 min intervals showing resultant velocity vectors of migration calculated by PIV with the sequential images. Velocity vectors superimposed on phase-contrast microscope images under (b) normoxic and (c) hypoxic conditions between 2 h and 2.5 h. Respective upper left panels are enlargements of the rectangular region positioned in the media channel. Stacked bar charts of cell migration speed measured by the displacements of ROIs within 30 min under (d) normoxic and (e) hypoxic conditions. (f) Variation of the migration speed and its linear least squares regression. Data shown are presented as mean ± standard deviation. *P < 0.05, **P < 0.01. (g) Autocorrelation function of velocity fluctuation vectors (mean ± standard deviation). The inset shows the decay of the velocity correlation function on a logarithmic scale. Nuclear translocation of hypoxia inducible factor-1α Confocal microscope images of HIF-1α and DAPI of HUVECs forming the monolayers were obtained after exposure to each oxygen condition (Fig. 3(a)). Here, images at 0 h were obtained after 4 days of incubation without gas mixtures supplied to the gas channels. Although HIF-1α was observed only slightly at either time point under normoxia, it was observed accumulating in the cytoplasm and translocating into the cell nuclei after an hour under hypoxia. Nuclear translocation of HIF-1α was evaluated by the relative value Īnucleus/Īwhole of the average fluorescence intensity Īnucleus in nuclei over those Īwhole in the whole image, as depicted in Fig. 3(b). Even under normoxia, the average intensity of HIF-1α in the nuclei was higher than that in the cytoplasm. However, the value of Īnucleus/Īwhole was almost constant around 1.5, which implied no nuclear translocation of HIF-1α. By contrast, the value of Īnucleus/Īwhole increased significantly under hypoxia, showing a peak value after 2 h of hypoxic gas supply, followed by a gradual decrease. These results show that the HUVECs sensed the hypoxic environment generated inside the device. Figure 3. Open in new tabDownload slide Nuclear translocation of HIF-1α in HUVECs by hypoxic exposure. (a) Maximum intensity projections of confocal microscope images of HUVECs to the xy-plane after normoxic or hypoxic exposure. The images at 0 h were obtained after 4-day incubation in a normal incubator (5% CO2, 37°C) without supplying gas mixtures to the gas channels. HIF-1α and nuclei were, respectively, stained with green and blue. Scale bar shows 50 μm. (b) Variation of the average intensity of HIF-1α in nucleus relative to that in the whole image, Īnucleus/Īwhole. Data shown are presented as mean ± standard deviation. *P < 0.05, **P < 0.01. Figure 3. Open in new tabDownload slide Nuclear translocation of HIF-1α in HUVECs by hypoxic exposure. (a) Maximum intensity projections of confocal microscope images of HUVECs to the xy-plane after normoxic or hypoxic exposure. The images at 0 h were obtained after 4-day incubation in a normal incubator (5% CO2, 37°C) without supplying gas mixtures to the gas channels. HIF-1α and nuclei were, respectively, stained with green and blue. Scale bar shows 50 μm. (b) Variation of the average intensity of HIF-1α in nucleus relative to that in the whole image, Īnucleus/Īwhole. Data shown are presented as mean ± standard deviation. *P < 0.05, **P < 0.01. Changes of hypoxic response by forskolin The migration speeds of HUVECs with the addition of FSK in the cell culture medium under controlled oxygen tensions were also measured using PIV analysis (solid symbols in Fig. 4(a)). Although the migration speed tended to be higher under hypoxia than under normoxia, no significant difference was found between the two oxygen conditions, each showing a gradual decrease in speed with time. In the experiment in which only EtOH was added, the migration speeds of HUVECs were almost constant at a high value (open symbols in Fig. 4(a)). The migration speeds of HUVECs between different cell culturing and oxygen tension conditions at three representative time periods were compared: the first and last time period (1–1.5 h and 4.5–5 h), and the time period (2–2.5 h) for which the largest difference was observed in the nuclear translocation of HIF-1α between different oxygen conditions (Fig. 4(b)). Regarding time periods and cell culture conditions, the migration speed of HUVECs was higher under hypoxia than under normoxia. The migration speed under normoxia was increased significantly immediately after the addition of FSK. However, it decreased gradually to a value similar to the speed in the experiment with normal cell culture medium. It showed a significant difference from the speed in the experiment with the addition of EtOH in the last time period. The migration speed under hypoxia with either cell culture medium showed similar values that were larger than those under normoxia. In the experiment with FSK under hypoxia, a significant decrease in the migrating speed was observed in the middle of the experiment period. Figure 4. Open in new tabDownload slide Migration speed V of HUVECs treated with/without forskolin (FSK) or ethanol (EtOH) under normoxic or hypoxic exposure. (a) Variation of migration speed and (b) comparison of migration speed between different conditions during three time periods of 1–1.5 h, 2–2.5 h and 4.5–5 h. Error bars show the standard deviation. *P < 0.05, **P < 0.01. Figure 4. Open in new tabDownload slide Migration speed V of HUVECs treated with/without forskolin (FSK) or ethanol (EtOH) under normoxic or hypoxic exposure. (a) Variation of migration speed and (b) comparison of migration speed between different conditions during three time periods of 1–1.5 h, 2–2.5 h and 4.5–5 h. Error bars show the standard deviation. *P < 0.05, **P < 0.01. After 2 h of the gas mixtures infusion, at which time the nuclear translocation of HIF-1α was observed to be greatest in the experiments with normal cell culture medium under hypoxia, HIF-1α in HUVECs in the experiments with FSK/EtOH were also observed using immunofluorescence imaging (Fig. 5(a)). All experiments conducted under hypoxia demonstrated nuclear translocation of HIF-1α. Relative values Īnucleus/Īwhole under hypoxia were significantly larger than those under normoxia when the same cell culture medium was used (Fig. 5(b)). Comparison of the relative values indicated that the added FSK significantly increased nuclear translocation of HIF-1α under normoxia, contrary to EtOH which significantly decreased the translocation under hypoxia. Figure 5. Open in new tabDownload slide HIF-1α in HUVECs, treated with/without forskolin (FSK) or ethanol (EtOH), after 2 h of normoxic (N) or hypoxic (H) exposure. (a) Maximum intensity projections of confocal microscope images of HUVECs to the xy-plane. Scale bar shows 50 μm. (b) Average intensity of HIF-1α in nucleus relative to that in the whole image, Īnucleus/Īwhole. Error bars show the standard deviation. *P < 0.05, **P < 0.01. Figure 5. Open in new tabDownload slide HIF-1α in HUVECs, treated with/without forskolin (FSK) or ethanol (EtOH), after 2 h of normoxic (N) or hypoxic (H) exposure. (a) Maximum intensity projections of confocal microscope images of HUVECs to the xy-plane. Scale bar shows 50 μm. (b) Average intensity of HIF-1α in nucleus relative to that in the whole image, Īnucleus/Īwhole. Error bars show the standard deviation. *P < 0.05, **P < 0.01. Changes of the adherens junction of VE-cadherin The immunofluorescence images of VE-cadherin in HUVECs after 5-h exposure to each oxygen condition are depicted in Fig. 6(a). Bands of VE-cadherin were observed surrounding each cell. They were apparently thicker because of the addition of FSK. In the experiment without FSK, narrow bands of VE-cadherin were visible locally. For quantitative evaluation of expression of VE-cadherin, the ratio A*cad of the VE-cadherin area to the total cell area was calculated as shown in Fig. 6(b). The addition of FSK increased the value of A*cad. Results show a significant difference compared to that without FSK under hypoxia. In experiments with normal cell culture medium, the value A*cad tended to be larger in normoxia than in hypoxia, although no significant difference was confirmed between the two oxygen conditions. Figure 6. Open in new tabDownload slide VE-cadherin of HUVECs, treated with/without forskolin (FSK) or ethanol (EtOH), after 5 h of normoxic (N) or hypoxic (H) exposure. (a) Maximum intensity projections of confocal microscope images of HUVECs to the xy-plane. Scale bar shows 50 μm. (b) Ratio A*cad of the VE-cadherin area to the total cell area. The metric was quantified for 60 cells from four device averages under each condition. Error bars show the standard deviation. **P < 0.01. Figure 6. Open in new tabDownload slide VE-cadherin of HUVECs, treated with/without forskolin (FSK) or ethanol (EtOH), after 5 h of normoxic (N) or hypoxic (H) exposure. (a) Maximum intensity projections of confocal microscope images of HUVECs to the xy-plane. Scale bar shows 50 μm. (b) Ratio A*cad of the VE-cadherin area to the total cell area. The metric was quantified for 60 cells from four device averages under each condition. Error bars show the standard deviation. **P < 0.01. DISCUSSION Hypoxic stress increases the speed of vascular endothelial cell migration, which is related to a decrease of adherens junction of VE-cadherin. In a hypoxic condition (<3% O2) generated by supplying a gas mixture of 0% O2 concentration to the gas channels [35], the migration speeds of HUVECs were higher than those in the normoxic condition though there was little difference in cluster size of cells migrating in the same direction by hypoxic exposure (Fig. 2). After an hour of the supply of 0% O2 concentration gas mixture, nuclear translocation of HIF-1α was observed in HUVECs, suggesting that the cells sensed the surrounding hypoxic condition (Fig. 3). As observed in an earlier study [36], VE-cadherin in HUVECs tended to decrease under hypoxic exposure (Fig. 6). By contrast, in the experiment with the addition of FSK which potentiates VE-cadherin to the cell culture medium, the migration speed of HUVECs did not show a significant difference between the two oxygen conditions, but it was slightly higher under hypoxic exposure (Fig. 4). Moreover, the migration speed of HUVECs after the addition of FSK gradually approached the speed with normal cell culture medium under normoxia. Consequently, the reinforcement of VE-cadherin suppressed the migration of the vascular endothelial cells. In this study, the migration speed under normoxia was almost constant, ~4 μm/h, during the experiments (Fig. 2). This migration speed was lower than the reported value of ~30 μm/h in a wound healing assay [27]. The difference in the migration speed probably derives from that the cells in a wound healing assay migrated to fill an open space while being coordinated by inflammatory factors. Also, there is a possibility that Matrigel coated onto the media channel yielded better cell adhesion, resulting in the slower migration speed of HUVECs [47]. An increase in the migration of HUVECs by hypoxic exposure was consistent with promoted migration observed in wound healing assays under hypoxia [29, 30]. Variation of the HUVECs migration speed under hypoxia showed a tendency to approach the migrating speed under normoxic conditions (Fig. 2). In addition, the ratio of the nuclear translocation of HIF-1α decreased with time (Fig. 3). These results imply that HUVECs were adjusting to the hypoxic environment. With the addition of FSK, the migration speed decreased gradually to that under normoxia with normal cell culture medium (Fig. 4). FSK simultaneously promotes gene expression of HIF-1α through increased production of cAMP [48]. Nuclear translocation of HIF-1α was observed by adding FSK to the cell culture medium under both oxygen conditions (Fig. 5). After 2 h of exposure to hypoxia with the addition of FSK, the nuclear translocation of HIF-1α was equivalent to that under hypoxia without FSK. However, with the addition of EtOH, the migration speed of HUVECs increased and remained high throughout the experiment period. These results are consistent with the fact that EtOH promotes the angiogenesis and vasculogenesis of vascular endothelial cells [42, 43]. Moreover, with the addition of EtOH, nuclear translocation of HIF-1α after 2 h of hypoxic exposure was not remarkably high compared with the other cell culture conditions. Consequently, HIF-1α translocated into the cell nucleus due to hypoxic exposure, but no distinct relation was found between nuclear translocation and cell migration speed. An earlier study demonstrated that the vascular endothelial permeability increased when the vascular endothelial monolayer was exposed to hypoxic stress [36]. In addition, a study using FSK has revealed that increased cAMP controls vascular permeability by strengthening VE-cadherin [40]. Corresponding with these studies, the present results suggest that the loosening of intercellular junctions because of the decrease of VE-cadherin engenders an increase of vascular endothelial permeability and the cell migration speed (Fig. 6). The present study with the microfluidic device provided new insights into the changes of migration of confluent vascular endothelial cells and intercellular VE-cadherin in the early stage of hypoxic exposure. However, there were several limitations on this approach that need to be further investigated. The microfluidic device used for this study still requires more than an hour to control the oxygen tension from a normoxic 21% O2 condition to a hypoxic <3% O2 condition [35]. In addition, a slight difference in oxygen tension existed across the media channel, of ~1% O2 (Fig. S1). A microfluidic device with a shorter transition time and more precise control of oxygen tension is necessary to investigate quick changes of cell migration. Further experiments conducted with control of oxygen tension and the exposure time must be undertaken to elucidate the dependence of cell migration on oxygen tension. Long-term variations of the migration of vascular endothelial monolayers and the expression of HIF-1α and cell–cell junctions at various oxygen levels are worthwhile topics to be investigated. Although three-dimensional migration of HUVECs in the collagen gel was not observed in the early stage of hypoxic exposure, the effect of long-term hypoxic exposure on HUVEC migration is expected to provide insight into hypoxia-induced angiogenesis [30, 49]. Moreover, the addition of EtOH used as a solvent of FSK increased the cell migration. Its effects could have influenced cell migration in the experiment with FSK. Use of a lower concentration of EtOH or a different solvent such as dimethyl sulfoxide might provide more persuasive results. For investigation under more physiologically relevant conditions, considerations of blood flow and hemodynamic environments are fundamentally important because they are known to influence angiogenesis in in vivo microenvironments. Furthermore, intracellular signal transduction to enhance cell mobility originating from the accumulation of HIF-1α inside the cell should be clarified. CONCLUSION Changes of the migration and intracellular junction morphology of vascular endothelial cells in a monolayer in the early stage of hypoxic exposure were investigated using microfluidic devices. Vascular endothelial monolayers were formed on the media channel inside the device, with the oxygen tension controlled by supplying gas mixtures flowing through the gas channels. Subsequent PIV analysis with a time-series of phase-contrast microscope images of the vascular endothelial monolayer revealed that hypoxic exposure increases migration speed of the cells with little changes of the cluster size of cells migrating in the same direction. Image analyses of immunofluorescence staining revealed that the vascular endothelial cells sensed hypoxia and reduced the adherens junction of VE-cadherin. The addition of FSK to the cell culture medium strengthened VE-cadherin, leading to a gradual decrease in the migration of the vascular endothelial cells. In conclusion, a loosening of the intercellular junction by the decrease of VE-cadherin caused by a hypoxic environment results in increased cell migration. ACKNOWLEDGEMENTS Part of the work was conducted under the Creative Interdisciplinary Research Project of FRIS, Tohoku University. KeF and DY acknowledge the Japan Society for the Promotion of Science for their funding under the Young Researcher Overseas Visits Program for Vitalizing Brain Circulation. FUNDING KeF acknowledges Japan Society for the Promotion of Science KAKENHI Grant Number 15K12500. RK acknowledges funding from the National Science Foundation Science and Technology Center for Emergent Behaviors of Integrated Cellular Systems (CBET-0939511). REFERENCES 1 Byrne MB , Leslie MT, Gaskins HR et al. Methods to study the tumor microenvironment under controlled oxygen conditions . Trends Biotechnol 2014 ; 32 : 556 – 63 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Carreau A , El Hafny-Rahbi B, Matejuk A et al. Why is the partial oxygen pressure of human tissues a crucial parameter? Small molecules and hypoxia . J Cell Mol Med 2011 ; 15 : 1239 – 53 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Semenza GL . Oxygen sensing, hypoxia-inducible factors, and disease pathophysiology . Annu Rev Pathol 2014 ; 9 : 47 – 71 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Haeger A , Wolf K, Zegers MM et al. Collective cell migration: guidance principles and hierarchies . Trends Cell Biol 2015 ; 25 : 556 – 66 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Hakim V , Silberzan P. Collective cell migration: a physics perspective . Rep Prog Phys 2017 ; 80 : 076601 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Djagaeva I , Doronkin S. Hypoxia response pathway in border cell migration . Cell Adhes Migr 2010 ; 4 : 391 – 5 . Google Scholar Crossref Search ADS WorldCat 7 Lehmann S , Te Boekhorst V, Odenthal J et al. Hypoxia induces a HIF-1-dependent transition from collective-to-amoeboid dissemination in epithelial cancer cells . Curr Biol 2017 ; 27 : 392 – 400 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Scarpa E , Mayor R. Collective cell migration in development . J Cell Biol 2016 ; 212 : 143 – 55 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Brugues A , Anon E, Conte V et al. Forces driving epithelial wound healing . Nat Phys 2014 ; 10 : 683 – 90 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Farooqui R , Fenteany G. Multiple rows of cells behind an epithelial wound edge extend cryptic lamellipodia to collectively drive cell-sheet movement . J Cell Sci 2005 ; 118 : 51 – 63 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Ravasio A , Cheddadi I, Chen T et al. Gap geometry dictates epithelial closure efficiency . Nat Commun 2015 ; 6 : 7683 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Clark AG , Vignjevic DM. Modes of cancer cell invasion and the role of the microenvironment . Curr Opin Cell Biol 2015 ; 36 : 13 – 22 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Friedl P , Alexander S. Cancer invasion and the microenvironment: plasticity and reciprocity . Cell 2011 ; 147 : 992 – 1009 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Sunyer R , Conte V, Escribano J et al. Collective cell durotaxis emerges from long-range intercellular force transmission . Science 2016 ; 353 : 1157 – 61 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Doxzen K , Vedula SR, Leong MC et al. Guidance of collective cell migration by substrate geometry . Integr Biol 2013 ; 5 : 1026 – 35 . Google Scholar Crossref Search ADS WorldCat 16 Vedula SR , Leong MC, Lai TL et al. Emerging modes of collective cell migration induced by geometrical constraints . Proc Natl Acad Sci USA 2012 ; 109 : 12974 – 79 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Lee RM , Kelley DH, Nordstrom KN et al. Quantifying stretching and rearrangement in epithelial sheet migration . New J Phys 2013 ; 15 : 025036 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Ladoux B , Mege RM. Mechanobiology of collective cell behaviours . Nat Rev Mol Cell Biol 2017 ; 18 : 743 – 57 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Garcia S , Hannezo E, Elgeti J et al. Physics of active jamming during collective cellular motion in a monolayer . Proc Natl Acad Sci USA 2015 ; 112 : 15314 – 19 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Bazellieres E , Conte V, Elosegui-Artola A et al. Control of cell–cell forces and collective cell dynamics by the intercellular adhesome . Nat Cell Biol 2015 ; 17 : 409 – 20 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Lee RM , Stuelten CH, Parent CA et al. Collective cell migration over long time scales reveals distinct phenotypes . Converg Sci Phys Oncol 2016 ; 2 : 025001 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Vedula SR , Hirata H, Nai MH et al. Epithelial bridges maintain tissue integrity during collective cell migration . Nat Mater 2014 ; 13 : 87 – 96 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Michaelis UR . Mechanisms of endothelial cell migration . Cell Mol Life Sci 2014 ; 71 : 4131 – 48 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Arima S , Nishiyama K, Ko T et al. Angiogenic morphogenesis driven by dynamic and heterogeneous collective endothelial cell movement . Development 2011 ; 138 : 4763 – 76 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Bentley K , Franco CA, Philippides A et al. The role of differential VE-cadherin dynamics in cell rearrangement during angiogenesis . Nat Cell Biol 2014 ; 16 : 309 – 21 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Vitorino P , Meyer T. Modular control of endothelial sheet migration . Genes Dev 2008 ; 22 : 3268 – 81 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Yang Y , Jamilpour N, Yao B et al. Probing leader cells in endothelial collective migration by plasma lithography geometric confinement . Sci Rep 2016 ; 6 : 22707 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Michiels C , Arnould T, Remacle J. Endothelial cell responses to hypoxia: initiation of a cascade of cellular interactions . Biochim Biophys Acta 2000 ; 1497 : 1 – 10 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Inoue T , Kohro T, Tanaka T et al. Cross-enhancement of ANGPTL4 transcription by HIF1 alpha and PPAR beta/delta is the result of the conformational proximity of two response elements . Genome Biol 2014 ; 15 : R63 . Google Scholar Crossref Search ADS PubMed WorldCat 30 Yang L , Zheng J, Xu R et al. Melatonin suppresses hypoxia-induced migration of HUVECs via inhibition of ERK/Rac1 activation . Int J Mol Sci 2014 ; 15 : 14102 – 21 . Google Scholar Crossref Search ADS PubMed WorldCat 31 Tao B , Wang R, Sun C et al. 3-mercaptopyruvate sulfurtransferase, not cystathionine β-synthase nor cystathionine γ-lyase, mediates hypoxia-induced migration of vascular endothelial cells . Front Pharmacol 2017 ; 8 : 657 . Google Scholar Crossref Search ADS PubMed WorldCat 32 Michaelis UR , Fisslthaler B, Barbosa-Sicard E et al. Cytochrome P450 epoxygenases 2C8 and 2C9 are implicated in hypoxia-induced endothelial cell migration and angiogenesis . J Cell Sci 2005 ; 118 : 5489 – 98 . Google Scholar Crossref Search ADS PubMed WorldCat 33 Funamoto K , Ito T, Funamoto K et al. Ultrasound imaging of mouse fetal intracranial hemorrhage due to ischemia/reperfusion . Front Physiol 2017 ; 8 : 340 . Google Scholar Crossref Search ADS PubMed WorldCat 34 Wu H , Lee T, Ko P et al. Review of microfluidic cell culture devices for the control of gaseous microenvironments in vitro . J Micromech Microeng 2018 ; 28 : 043001 . Google Scholar Crossref Search ADS WorldCat 35 Funamoto K , Zervantonakis IK, Liu Y et al. A novel microfluidic platform for high-resolution imaging of a three-dimensional cell culture under a controlled hypoxic environment. Lab Chip 2012 ; 12 : 4855 – 63 . Google Scholar Crossref Search ADS PubMed WorldCat 36 Funamoto K , Yoshino D, Matsubara K et al. Endothelial monolayer permeability under controlled oxygen tension . Integr Biol 2017 ; 9 : 529 – 38 . Google Scholar Crossref Search ADS WorldCat 37 Kametani Y , Takeichi M. Basal-to-apical cadherin flow at cell junctions . Nat Cell Biol 2007 ; 9 : 92 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat 38 Hayer A , Shao L, Chung M et al. Engulfed cadherin fingers are polarized junctional structures between collectively migrating endothelial cells . Nat Cell Biol 2016 ; 18 : 1311 – 23 . Google Scholar Crossref Search ADS PubMed WorldCat 39 Shin Y , Han S, Jeon JS et al. Microfluidic assay for simultaneous culture of multiple cell types on surfaces or within hydrogels . Nat Protoc 2012 ; 7 : 1247 – 59 . Google Scholar Crossref Search ADS PubMed WorldCat 40 Fukuhara S , Sakurai A, Sano H et al. Cyclic AMP potentiates vascular endothelial cadherin-mediated cell-cell contact to enhance endothelial barrier function through an Epac-Rap1 signaling pathway . Mol Cell Biol 2005 ; 25 : 136 – 46 . Google Scholar Crossref Search ADS PubMed WorldCat 41 Noda K , Zhang J, Fukuhara S et al. Vascular endothelial-cadherin stabilizes at cell-cell junctions by anchoring to circumferential actin bundles through α- and β-catenins in cyclic AMP-Epac-Rap1 signal-activated endothelial cells . Mol Biol Cell 2010 ; 21 : 584 – 96 . Google Scholar Crossref Search ADS PubMed WorldCat 42 Lamichhane TN , Leung CA, Douti LY et al. Ethanol induces enhanced vascularization bioactivity of endothelial cell-derived extracellular vesicles via regulation of microRNAs and long non-coding RNAs . Sci Rep 2017 ; 7 : 13794 . Google Scholar Crossref Search ADS PubMed WorldCat 43 Morrow D , Hatch E, Hamm K et al. Flk-1/KDR mediates ethanol-stimulated endothelial cell notch signaling and angiogenic activity . J Vasc Res 2014 ; 51 : 315 – 24 . Google Scholar Crossref Search ADS PubMed WorldCat 44 Vennemann P , JPIV, https://www.jpiv.vennemann-online.de/, (accessed August 2018 ). 45 Seebach J , Donnert G, Kronstein R et al. Regulation of endothelial barrier function during flow-induced conversion to an arterial phenotype . Cardiovasc Res 2007 ; 75 : 596 – 607 . Google Scholar Crossref Search ADS PubMed WorldCat 46 Angelini TE , Hannezo E, Trepat X et al. Cell migration driven by cooperative substrate deformation patterns . Phys Rev Lett 2010 ; 104 : 168104 . Google Scholar Crossref Search ADS PubMed WorldCat 47 Kick K , Nekolla K, Rehberg M et al. Hypoxia and estrogen are functionally equivalent in breast cancer-endothelial cell interdependence . Arterioscler Thromb Vasc Biol 2016 ; 36 : 2346 – 57 . Google Scholar Crossref Search ADS PubMed WorldCat 48 Busca R , Berra E, Gaggioli C et al. Hypoxia-inducible factor 1α is a new target of microphthalmia-associated transcription factor (MITF) in melanoma cells . J Cell Biol 2005 ; 170 : 49 – 59 . Google Scholar Crossref Search ADS PubMed WorldCat 49 George AL , Rajoria S, Suriano R et al. Hypoxia and estrogen are functionally equivalent in breast cancer-endothelial cell interdependence . Mol Cancer 2012 ; 11 : 80 . Google Scholar Crossref Search ADS PubMed WorldCat © The Author(s) 2019. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Migration of vascular endothelial cells in monolayers under hypoxic exposure
JF - Integrative Biology
DO - 10.1093/intbio/zyz002
DA - 2019-01-01
UR - https://www.deepdyve.com/lp/oxford-university-press/migration-of-vascular-endothelial-cells-in-monolayers-under-hypoxic-CeV1tYmSE6
SP - 26
VL - 11
IS - 1
DP - DeepDyve
ER -