Establishing correlations in the en-mass migration of dermal fibroblasts on oriented fibrillar scaffolds

Acta Biomaterialia 25 (2015) 230–239

Contents lists available at ScienceDirect

Acta Biomaterialia journal homepage: www.elsevier.com/locate/actabiomat

Establishing correlations in the en-mass migration of dermal fibroblasts on oriented fibrillar scaffolds Sisi Qin a, Richard A.F. Clark b, Miriam H. Rafailovich a,⇑ a b

Materials Sciences and Engineering Department, Stony Brook University, Stony Brook, NY, USA Department of Materials Science and Engineering, Stony Brook University, Stony Brook, NY, USA

a r t i c l e

i n f o

Article history: Received 25 January 2015 Received in revised form 14 June 2015 Accepted 23 June 2015 Available online 24 June 2015 Keywords: Three dimensional scaffolds Dermal fibroblasts Cell migration Oriented fibers

a b s t r a c t Wound healing proceeds via fibroblast migration along three dimensional fibrillar substrates with multiple angles between fibers. We have developed a technique for preparation of three dimensional fibrillar scaffolds with where the fiber diameters and the angles between adjacent fiber layers could be precisely controlled. Using the agarose droplet method we were able to make accurate determinations of the dependence of the migration speed, focal adhesion distribution, and nuclear deformation on the fiber diameter, fiber spacing, and angle between adjacent fiber layers. We found that on oriented single fiber layers, whose diameters exceeded 1 lm, large focal adhesion complexes formed in a linear arrangement along the fiber axis and cell motion was highly correlated. On multi layered scaffolds most of the focal adhesion sites reformed at the junction points and the migration speed was determined by the angle between adjacent fiber layers, which followed a parabolic function with a minimum at 30°. On these surfaces we observed a 25% increase in the number of focal adhesion points and a similar decrease in the degree of nuclear deformation, both phenomena associated with decreased mobility. These results underscore the importance of substrate morphology on the en-mass migration dynamics. Statement of Significance En-mass fibroblast migration is an essential component of the wound healing process which can determine rate and scar formation. Yet, most publications on this topic have focused on single cell functions. Here we describe a new apparatus where we designed three dimensional fibrillar scaffolds with well controlled angles between junction points and highly oriented fiber geometries. We show that the motion of fibroblasts undergoing en-mass migration on these scaffolds can be controlled by the substrate topography. Significant differences in cell morphology and focal adhesions was found to exist between cells migrating on flat versus fibrillar scaffolds where the migration speed was found to be a function of the angle between fibers, the fiber diameter, and the distance between fibers. Ó 2015 Published by Elsevier Ltd. on behalf of Acta Materialia Inc.

1. Introduction Collective migration where cells migrate together while maintaining cell–cell contacts has been studied extensively especially for epithelial and endothelial cells, which migrate as a sheet [1– 5]. Other studies have focused on migration of multiple cell types together in order to determine the role of soluble factors in the process [6,7]. In vivo wound healing is a combination of multiple factors, where the rate of the migration process can be a determinant factor in controlling the probable outcome-i.e. low mobility ⇑ Corresponding author at: Department of Materials Science and Engineering, Stony Brook University, Stony Brook, USA. E-mail address: [email protected] (M.H. Rafailovich). http://dx.doi.org/10.1016/j.actbio.2015.06.030 1742-7061/Ó 2015 Published by Elsevier Ltd. on behalf of Acta Materialia Inc.

can prevent healing as has been reported for diabetics while overly rapid rates are associated with hypertrophic scar formation [8–10]. The rapid developments in the stem cell field has amply demonstrated that the mechanical, chemical, and morphological properties of the substrates on which the stem cells reside, both in vivo and in vitro, can be determining factors [11–14]. Since wound healing also involves tissue regeneration and renewal, these factors need to be considered as well. Fibroblasts, which play a major role in collagen contraction, are particularly sensitive to environmental cues such as the diameter of the collagen fibers [15] or their rigidity, where both factors were shown in vitro and in vivo to influence cell migration and healing [16,17]. Even though fibroblasts do not migrate as a sheet, it was shown that single cell migration can differ fundamentally from the

231

S. Qin et al. / Acta Biomaterialia 25 (2015) 230–239

migration which occurs when multiple cells are involved. This type, referred to as ‘‘en-mass’’ migration [18,19], relies on cell–cell communication and interactions via the substrate properties, where studies that have shown that fibroblasts prefer fibrillar structure over than flat surfaces [20–23]. Yet despite its importance, most studies have focused on the dynamics of isolated cells, [24–27] and only few studies have been reported regarding en-mass migration [19,28]. For example, Liu et al. [28] compared en-mass migration of fibroblasts out of an agarose droplet to single cell migration onto flat and fibrillar surfaces. On flat surfaces, fibroblasts migrated outward into a ‘‘sunburst’’ type pattern in a trajectory aimed increasing the separation between adjacent cells while the speed of the cells decreased with increasing distance from the droplet, reaching the single cell migration value after 24 h. In contrast, cells migrating outward from droplets placed on fibers experienced a ‘‘bottleneck’’ effect, as they emerged en-mass with the single cell velocity which remained unchanged for the first 24 h. It has also been shown that three dimensional migration of human dermal fibroblasts on a fibrous scaffolds, having diameters 1 lm or larger, did not proceed via cell migration within the pores of the scaffold [29]. Rather, the migration occurred only along the fibers, where motion from one layer to the next involved a sudden change in trajectory at the junction of fibers in different layers. Hence it appeared that the major factors determining the magnitude and direction of the migration speed were the nature of the fiber junctions, rather than the porous structure of the scaffold. It has been shown in several studies that cell migration was affected by the topography of the underlying substrate [20,30– 35], which in turn determined other parameters such as density and distribution of the focal adhesion points [36–40] the degree of nuclear deformation [41–44]. In the case of en-mass migration of cells on scaffolds, the angle between fibers, or the angle at the junction points, can also be element determining the substrate topography. Hence the question arises whether cells could discriminate between junction points having different angles and exhibit a preference for migration in a specific direction. In order to address these questions, we have devised a technique for producing a multi-layered scaffold where the angle between fibers can be precisely controlled and the influence of the three dimensional substrate morphology on the migration behavior determined. We chose to use the well-established droplet technique [45–47] for studying en-mass migration behavior rather than scratch assays [48–50] since it allows for better quantification of the cell migration trajectories and the behavior of individual cells within the migrating ensemble. The answers to these questions can further our understanding of the elements within the substrate that control outward cell migration and can enhance our ability to engineer structures essential for wound healing applications.

2. Materials and methods 2.1. Fabrication of electrospun PMMA scaffolds of variable angle The application of the multilayered scaffold is preceded by preparation of the substrate with a flat spun cast film. The film allows for adhesion of the scaffold, as well as for a direct comparison between cells on flat vs fibrillar surfaces. Poly (methyl methacrylate) (PMMA) (Mw = 120,000 Da, Mw/Mn = 3; Sigma– Aldrich) was dissolved in toluene (Fisher Scientific, Pittsburgh, PA) at 30 mg/mL, and then spin-casted on 1.5 cm diameter glass cover slips at 2500 RPM for 30 s. Electrospun fibers of different diameters were produced by dissolving PMMA in different solvents, as listed in Table 1, where the

Table 1 Solvent and concentration of PMMA for producing different sizes of electrospun fibers. Concentration of PMMA (%)

Solvent

Results

30 30

Tetrahydrofuran + Dimethylformamide (1:1) Dimethylformamide + Chloroform (1:1)

20

Chloroform

1 lm fibers 4 lm fibers 8 lm fibers

solvent and concentrations were optimized to minimize beading and produce fibers of the desired diameters. The set up for obtaining oriented multi layered scaffolds is illustrated in Fig. 1a. Fiber alignment was accomplished by placing the PMMA coated cover slips on a drum whose diameter, 10 cm, was much larger than that of the cover slips, and which was rotating at a speed of 6750 RPM. Analysis of multiple images and more than 500 fibers indicated that the degree of alignment, as determined from the ratio of aligned vs. curved, non-aligned fibers was greater than 98%. The distance between adjacent fibers was controlled by the electrospinning time, where mean distances of 30, 50, and 100 lm were achieved with spinning times of 1, 2.5, and 5 min. The angular orientation of the cover slips was marked by fiduciary on the drum as shown. The angle between different layers could then be precisely controlled via rotation of the cover slip after deposition of each layer. The samples were then annealed in a vacuum of 107 Torr at 120 °C overnight to remove the remaining solvent and sterilize the samples. Fig. 1b–f shows the phase contrast images of samples with angles of 15°, 30°, 45°, 75°, and 90° between adjacent layers of fibers. From the figure we can see that even though the spacing between fibers has a large variation, the angle between fibers in adjacent layers is precise to within 5°. We have previously shown that for an orthogonal scaffold, the spacing between fibers need only be large enough to allow cells attached to fibers to penetrate, and hence did not affect the migration when they were in a range from 10 to 100 lm. The spacing between the fibers shown in Fig. 1b–f was easily controlled within this range and hence no further processing [51,52] was applied to try to reduce the variation. 2.2. Cell culture Adult Human Fibroblast Cell (CF29, passage 10-passage 12 cells were used in this study) was purchased from ATCC (Manassas, VA), and cultured in Dulbecco’s Modified Eagle Medium (DMEM), together with 10% fetal bovine serum (Hyclone, Logan, UT), 1% antibiotic mix of penicillin, streptomycin, and L-glutamine (GIBCO BRL/Life Technologies, Grand Island, NY), (referred to as full-DMEM), in a humidified incubator at 37 °C with 5% CO2. Before seeding the cells, the samples were further sterilized by exposure to the antimicrobial UVC light within the BSL-2 enclosure for 20 min, and then submerged in 0.5 ml of 30 mg/mL of intact human plasma fibronectin (Calbiochem, San Diego, CA) in serum-free DMEM for 2 h in the incubator. Sterilization with alcohol was avoided since it pitted the PMMA surfaces. 2.3. Cell membrane staining Fibroblasts cell membranes were live stained with 1,10 -dioctade cyl-3,3,30 ,30 -tetramethylindodicarbocyanine perchlorate (DiD, Invitrogen, Carlsbad, CA). Cells were washed with phosphate buffer saline (PBS), and suspended in the 3.5 lg/ml DiD-serum free

232

S. Qin et al. / Acta Biomaterialia 25 (2015) 230–239

Fig. 1. Electrospun PMMA scaffolds. (a) Illustration of electrospinning setup for fabricating designed angle scaffolds. (b) 15°, (c) 30°, (d) 45°, (e) 75°, and (f) 90° (scale bar = 100 lm).

DMEM solution at a density of 1  106 cells/mL, then incubated at 37 °C for 30 min. The suspension tubes were then centrifuged at 1500 RPM for 5 min, and the supernatant was removed. 2.4. Droplet migration method After staining the membrane, the fibroblasts were re-suspended in a volume of 0.2% (w/v) agarose solution to obtain a final cell density of 1.5  107 cells/ml. Several droplets of the agarose/cell suspension, each 1.25 lL in volume, were seeded onto the substrates, and solidified by placing at 4 °C for 10 min. Full-DMEM, warmed to 37 °C was then slowly added to the sample in order to avoid displacement. 2.5. Measurement of cell migration After cell seeding, samples were incubated in the humidified incubator, with 5% CO2, at 37 °C for 24 h. The migration velocity was measured via time lapse photography by placing the samples on a 37 °C incubator stage attached to a Nikon Diaphot-TMD inverted microscope and fitted with a MetaMorphÒ-operated CoolSNAP™HQ camera (Universal Imaging Corporation, Downingtown, PA). Images were automatically taken every 15 min over the period of one hour, and the migration velocity of the cells was calculated using MetaMorph. Only the cells that were migrating at the leading edge were chosen for the measurement in order to avoid cell–cell interference. In order to determine the location of the moving front relative to the initial droplet position, fluorescent microscope pictures were overlapped by the images of

cells incubated for 4 h, stained with DiD (red) and the cells incubated for 24 h, fixed, and stained with Alexa-fluor 488 (green).

2.6. Immunofluorescent staining Location of vinculin was visualized by a Leica TCS SP2 laser scanning confocal microscope (Leica Microsystems, Bannockburn, IL). After incubation, cells were fixed with 3.7% formaldehyde for 20 min, permeabilized with 0.4% Triton for 7 min, and blocked with 2% BSA in PBS for 30 min at room temperature. Immunofluorescent staining for vinculin (Sigma, Saint Louis, MO) was performed at a 1:600 dilution for 1 h then incubated with the Oregan Green 488 goat anti-mouse secondary antibody (Invitrogen, Carlsbad, CA) with the same procedure. To observe the nuclei, cells were stained with 40 ,6-diamiadino-2-phenylindole (DAPI, Sigma–Aldrich, Inc., St. Louis, USA) for 10 min at room temperature, and the results were quantified using Image J software.

2.7. Statistical analysis Twenty cells were measured in each sample, and 3 replicate samples were prepared for each condition. Each data point then signified an average over at least 60 cells. Statistical analysis was performed using GraphPad5 software to calculate the statistical significance level using Tukey’s HSD test. Reduced Chi squared analysis was also performed by OriginPro 9.1 and the statistical significance is determined where p < 0.05 was considered significant.

S. Qin et al. / Acta Biomaterialia 25 (2015) 230–239

3. Results 3.1. Migration on oriented fibers 3.1.1. Dependence on fiber diameter The en-mass migration profile of cells emerging from droplets positioned on aligned parallel fibers of different diameters is shown in Fig. 2a–c. The figures were obtained by superimposing the image of the droplet taken after 4 h of incubation on top of the image obtained after incubation for 24 h. The edges of the droplet at the two time points are delineated by the solid lines. From the figure we can see that as the fiber diameter increases from 1, 4 to 8 lm the aspect ratio of the migration front increases (Fig. 2d). Closer examination of the area along the shorter axis shows that most of the cells comprising the front are migrating along the direction of the fiber orientation, hence the non-spherical distribution. As the fiber diameter increases we see that the amount of cells migrating in a direction perpendicular to the fiber orientation decreases (inserts to Fig. 2a–c), indicating that cell alignment along the fiber direction becomes easier as the fiber diameter increases. In Fig. 3a–c we show confocal microscopy of the cells migrating on fibers of different diameters where the focal adhesion points are imaged after immunofluorescent staining for vinculin. The focal adhesions seem to be distributed randomly on the cytoplasm, without regarding to the underlying fibrillar structure. On the substrates with larger fiber diameters the cells are clearly oriented along the fiber direction, with nearly all the cells firmly attached to one fiber (Fig. 3d). A well-defined line of focal adhesion points is observed on most cells, which follows the contour of the fiber to which the cells are attached. The degree of orientation of the cells and the number of aligned focal adhesions increases with increasing fiber diameter. In Fig. 3e we plot the migration velocity after 24 h of incubation on fibers of different diameters. From the figure we can see the migration speed decreases with increasing alignment along the fibers. Hence increased orientation also hinders the initial speed with which the cells emerge en mass from the droplet. This is consistent with the previous observation [28] where they reported that cells emerged en mass on oriented fibers at a lower speed than those on flat substrates as the cells waited in an orderly manner to find a place on the fiber.

3.1.2. Distance between fibers The ability of all cells to find a position on a fiber when they emerge from the droplet can be thwarted if the distance between fibers is increased. In this case the pressure from other cells forces them out of the droplet in the space between fibers. In Fig. 4a we show a typical image of cells emerging on 8 lm fibers, approximately 100 lm apart. From the figure we can see that a large number of cells are now in between fibers. In Fig. 4b we plot the number of cells between fibers at a fixed distance of 500–900 lm from the droplet, as a function of the mean distance between

233

fibers. From the figure we see that this number decreases with decreasing spacing and increasing distance from the edge of the droplet. In Fig. 4b we plot the migration speed of cells in the flat areas between fibers, as a function of inter fiber distance. From the figure we find that as the number of cells increases the migration speed decreases. This is in complete contrast with the behavior reported for en-mass migration of fibroblasts from a droplet, where the chosen trajectory is such as to maximize the distance between adjacent cells and to where the migration speed is highest at the largest cell densities. A closer observation of the cell migration dynamics in the space between fibers offers a possible explanation for this effect. Fibroblasts extend long processes emanating from the cells on the fibers, where the process senses the local environment. As a result, cells on fibers will move in tandem, where Table 2 lists the amount of neighboring cells whose motion may be correlated along one fiber or along adjacent fibers. Observation of the motion of the cells was shown in Fig. 5 and [supplementary material] that, in contrast to a flat open space where cells can move in random directions, the cells in the space between fibers also undergo correlated motion with a correlation number of approximately 3 cells. It is interesting to note that despite having emerged on the flat region of the sample, the cells somehow are aware of the presence of the fibers, and migrate toward the fibers. At that point, they slow down and wait for a space to present itself where they can attach to the fibers. As a result, the cell density on the flat areas decreases quickly with increasing incubation time and distance from the edge of the droplet.

3.2. Migration into a fibrillar scaffold 3.2.1. Pattern formation In order to migrate in three dimensions, cells must also be able to change course from fiber to fiber. We therefore also placed the agarose droplet on scaffolds where the fiber layers were arranged at a fixed angle relative to each other and observed the fraction of cells moving along the initial fiber direction relative to those which switched fibers and moved into another layer after 24 h incubation (Fig. 6a). The geometry of the scaffolds was shown in Fig. 6b, where the distance between the aligned fibers was chosen to be approximately 30 lm. From the figure we can see that the majority of the cells preferred continuing straight on the oriented fibers. Approximately 30% switched layers, when the scaffold angles were 30° and 45° and slightly more, 35%, switched on scaffolds with a 90° angle. This can be seen even more clearly in Fig. 7 where we superimpose the initial droplet over the migration pattern that evolved after 24 h. From the figure we can see that the patterns traced out reflect the ratio of cells emerging on layers placed at different angles. This is quantified in Fig. 7g where we plot the aspect ratio of the patterns as a function of the angle between the fibers, and the cell motion between layers is determined by the angle between fibers.

Fig. 2. (a–c) Fluorescent microscope overlap images of cells on different diameters of fibers (scale bar = 400 lm). (d) Aspect ratio of migration area after 24 h of incubation.

234

S. Qin et al. / Acta Biomaterialia 25 (2015) 230–239

Fig. 3. Cell migration on different diameter fibers. Confocal images of vinclulin on (a) 1 lm, (b) 4 lm, and (c) 8 lm fibers. (d) Nuclei location on 8 lm fibers (scale bar = 100 lm). (e) Cell migration velocity of on different diameter fibers.

Fig. 4. Cell migration on different spacing of fibers. (a) Phase contrast image of cells plated on 100 lm spacing fibers (scale bar = 100 lm). (b) Cell migration velocity and number of cells as a function of spacing between fibers.

3.2.2. Migration velocity In Fig. 8a we plot the en-mass migration speed as a function of the angle between fibers in adjacent layers in the different scaffolds. From the figure we see that the migration speed has a parabolic form with respect to angle, with the maximum reduction in

speed occurring at 30° (Reduced Chi square = 0.94). This result was quite unexpected since one would predict that the degree with which the cells would slow down should scale with the deformation they would encounter in switching between fibers at the junction points. We therefore explored this phenomenon in more detail

S. Qin et al. / Acta Biomaterialia 25 (2015) 230–239 Table 2 Correlation distance of cells migrating on the same fiber and adjacent fibers. Spacing between fibers (lm)

10 30 50

On the same fiber

On adjacent fibers

Number of cells in correlation

Number of cells in correlation

6.25 ± 1.56 4.05 ± 0.93 4.20 ± 1.61

4.05 ± 1.18 2.93 ± 1.53 2.47 ± 0.72

by observing the migration speed of individual cells as they approached a fiber/fiber junction and switched fibers, for the 30° and 90° scaffolds. From Fig. 8b we can see that on the 30° scaffold when the cells come within approximately 40 lm, they begin to slow down till they reach a value nearly half of their original speed at the junction itself. In contrast the cells on the 90° scaffolds migrate from one fiber to the next, undisturbed, at the same speed. In Fig. 8c–h we show confocal images of the cells 24 h after emerging from an agarose droplet onto scaffolds with fibers oriented at 30°, 45° and 90°. In the inset we also plot SEM images of cells at the 30° and 90° junction points which show that they are well extended along the two fibers forming the junction with long processes extended along the fibers. Comparing these images with those shown above in Fig. 3c of the cells incubated for 24 h on the oriented single layer substrate one can immediately see that the cells on the single layer substrate have more focal adhesion sites per cell. In the case of the cells migrating on the multi-layers substrates there are fewer focal adhesions along the fiber perimeter and most seem localized along the junction region. This is consistent with the breaking of the focal adhesions on one fiber and reforming at the junction point in preparation of the cells changing direction. The spacing of junction points, approximately 40 lm, is roughly the same as the length of the cells, and hence the cells are constantly choosing their orientation direction, which may prevent the formation of the focal adhesion sites along the fiber perimeter and concentrating them at the junction. The number of focal adhesion sites per cell was plotted in Fig. 9a for 0°, 30°, and 90° scaffolds. From the figure we can also see that the number of focal adhesions is about 25% larger on the 30° scaffold than on the 90° scaffold (p < 0.01) which is consistent with the cells moving faster at the 90° junctions, having fewer focal adhesion complex to break as they change direction. Nuclear deformation has also been reported as a driver for cell migration [41–44]. On flat surfaces migration was shown to occur as a consequence of redistribution of the traction forces as the cells

Fig. 5. Number of cells migrate in correlation on the flat film between adjacent fibers.

235

attempted to reduce nuclear deformation. On fibers a permanent deformation is imposed via the orientation of the cells along the fiber axis. This is shown in Fig. 9b where the nuclei of the cells on the single layer substrate have an aspect ratio of 2.0. Measurement of the nuclear deformation of the cells at the junction points on the multilayer substrates shows that the nuclear deformation is maintained on the 90° scaffold, as the cells are changing direction, while it is reduced by 25% on the 30° scaffold while the cells are resident on the junction point indicating that the distribution of traction forces exerted by cells to change direction may also be a function of angle.

4. Discussion In vivo fibroblasts are residents on extra cellular matrix whose structure is fibrillar. The size and shape of the fibers, which are mostly collagen in nature, is organ dependent. Individual collagen fibrils which are typically cylindrical in shape, with diameters of ranging from 10 to 500 nm can form bundles whose diameters can range from 1 to 20 lm [15]. Subcutaneous collagen for example, has been shown to consist of poorly oriented cord shaped fibers, approximately 8 lm in diameter [15]. In an elegant paper, Bowes et al. used small angle light scattering to show that hypertrophic scar formation was correlated to a significant increase in the degree of orientation of the collagen fibers [53]. The enhanced degree of orientation was found immediately post-wound formation and was larger in wounds treated with TGF-B than in the untreated controls, which were in turn more oriented than the normal unwounded skin [54]. Numerous groups have shown that the migration of fibroblasts into a wound, which occurs in the early stages of wound healing, is directed by matrix orientation, and moderated by chemotactic factors. While excessive migration is known to be associated with hypertrophic scarring, poor migration is associated with unfavorable prognosis in the healing of the wound [17]. Wound healing is therefore a complex process moderated by multiple factors. In order to better understand the process, and enable the design of a wound healing scaffold, we have attempted to study the influence on migration only of the structure

Fig. 6. Cell migration on two layers of fibers. (a) Residence of cells on 30°, 45°, and 90° scaffolds. (b) Illustration of cell migrating on 30°, 45°, and 90° scaffolds.

236

S. Qin et al. / Acta Biomaterialia 25 (2015) 230–239

Fig. 7. The overlapped image of en-mass cell migration of live cells stained with DiD (red), and incubated for 4 h, onto the image of cells incubated for 24 h, fixed and stained for F-actin with Alexa Fluor (green) on (a) 0°, (b) 15°, (c) 30°, (d) 45°, (e) 75° and (f) 90° scaffolds (scale bar: 400 lm). (g) Aspect ratio of cell migration pattern after 24 h incubation on different angle scaffolds.

of the substrate, in the absence of other external chemotactic factors. We therefore designed the apparatus shown in Fig. 1 where we could control multiple aspects of the morphology separately and correlate them directly to the cell migration speed. The process of electrospinning for tissue engineering has been the subject of multiple reviews [55–57], where it has been shown to be a simple and economical technique to produce scaffolds with high porosity and large surface to volume ratios for delivery of cytokines or chemotactic factors. The focus in most of these papers has been in the potential of these fibers as drug delivery vehicles or anti-bacterial inserts. These scaffolds are also shown to have huge potential for the delivery of cells into non-healing wounds. In this case, in addition to adhesion or porosity, the influence of the fibers on cell migration also becomes important. Hence we focused on the design of the fiber structures which would enable control of the migration and possibly allow integration of the cells with minimal disruption of the natural tissue. We have shown that dermal fibroblast cells strongly prefer migrating along fibers even when executing en-mass migration from an agarose droplet. This preference is established as soon as the cells emerged onto the fibers. Observation of the cells after 24 h shows that the cells formed strong focal adhesion sites and are oriented along the fibers. The most pronounced degree of orientation occurred on fibers of diameter greater than 1 lm. Even though some degree of orientation was also observed on fibers 1 lm in diameter, rather than forming along the fiber, focal adhesions were observed throughout the cytoplasm and the degree of orientation was lower than on the larger diameter fibers. Measurements of the migration speed indicated that the cells migrated slower as the degree of orientation along the fiber increased. SEM images of the cells on the fibers showed large process also formed which extended large distances from the cytoplasm. These processes were very thin and were difficult to image using florescent stains. Nevertheless, these processes allowed the cells to sense the presence of other cells and resulted in highly correlated en-mass motion pattern. For cells emerging on flat films even though we observed en mass migration, no correlations were observed between individual cells, which migrated independently of each other. For cells migrating on the fibrillar surfaces a correlation length extending over approximately 4–6 cells

along the fiber and approximately 2–4 cells in adjacent fibers was observed where the cells moved in a lock-step fashion. These multiple correlations were also thought to be responsible for the initial lower migration speed of the cells on the fibers relative to those on the flat films. Motion in three dimensions gave rise to distinct differences in the cell behavior. The frequent presence of the fiber junctions forced the cells along multiple migration trajectories, which interfered with the correlated behavior observed on the single layer scaffold. Furthermore, the frequent changes in orientation imposed at the junction points, also interfered with the formation of the long lines of focal adhesions which formed along the oriented fibers. Instead, the focal adhesions were now concentrated along the junction points and long processes were observed to attach to the junctions allowing the cells to alter their migration speed or orientation in preparation of the approach. Therefore, orientation of the fibrous scaffolds results in a more ordered cell migration pattern, which in turn minimizes the conditions leading to hypertrophy, and may be one of the natural mechanisms to promote healing while controlling scarring. A very surprising detail was the observation that the migration speed on the multi-layer scaffold was a parabolic function of the angle between fibers in adjacent layers, with a distinct minimum in speed observed at an angle of 30°. This minimum was consistent with the measurement with a 25% increase in both numbers of focal adhesion points as well as a 25% decrease in nuclear aspect ratio for cells migrating on the 30° scaffolds relative to those migrating on the 90° scaffolds. Cell migration has been shown to proceed via polymerization and de-polymerization of filaments followed by placement of focal adhesions [58–64]. Preferential angles involved in actin filament orientation have been reported by several groups where it has been shown that dendrites polymerizing on actin filaments form at a distinct angle of 70° from the main filament backbone [65]. Recently it was also shown that cells subjected to cyclic stress deformation react by forming actin stress fibers oriented at 60° to the stress direction [66]. Hence, even though we do not have a model as yet, it is not surprising to find that cell re-orientation during migration should also have a pronounced angular dependence since it too is a process which also involves directional actin polymerization.

237

S. Qin et al. / Acta Biomaterialia 25 (2015) 230–239

b

a

90° 30°

40 cell migration velocity/(µm/h)

Cell migration velocity/(µm/h)

32 30 28 26 24 22

35 30 25 20 15 10 5 0 45min

30min

15min

0min

time before reaching junction

20 0

20 40 60 degree between fibers

80

100

c

e

g

d

f

h

Fig. 8. Cell migration on different angle scaffolds. (a) Cell migration velocity on electrospun fibers with different angles. (b) Migration velocity of cells approaching the junction on 30° and 90° surfaces. (c and d) Confocal images of vinculin on 30°, (e and f) on 45° surfaces, (g and h) 90° surfaces (inserts: SEM images of cells at fiber junction).

Fig. 9. (a) Number of focal adhesion sites per cell on 0°, 30°, 45° and 90° surfaces. (b) Aspect ratio of nucleus on the surfaces. (c) Fluorescent image of nucleus on 30° surface. (d) Fluorescent image of nucleus on 45° surface. (e) Fluorescent image of nucleus on 90° surface (scale bar = 100 lm).

5. Conclusions En-mass migration on fibrillar surfaces is an essential component of the wound healing process. We have therefore

concentrated on studying the en-mass behavior of cells emerging from an agarose droplet onto single and multi-layered scaffolds of oriented fibers. On the single layered substrates we found that, for fibers with diameters greater than 1 lm, the cells formed focal

238

S. Qin et al. / Acta Biomaterialia 25 (2015) 230–239

adhesions oriented on the fiber axis and migrated along the fibers in a highly correlated manner. On the multi-layered scaffolds the en-mass migration speed was found to be a parabolic function of the angle between adjacent layers, with a well-defined minimum occurring at an angle of 30° between adjacent layers. In this case the focal adhesions were concentrated at the fiber/fiber junctions with the number of adhesions being larger and the nuclear deformation being smaller than the corresponding values for those migrating on the 30° scaffolds relative those migrating on the 90° scaffolds, underscoring the role of fibrillar substrate alignment in orienting the en-mass migration of cells in three dimensions.

[18]

[19]

[20]

[21]

[22]

Acknowledgment The authors would like to acknowledge the National Science Foundation: Inspire Program Award number: 1344267.

[23]

Appendix A. Figures with essential color discrimination

[24]

Certain figures in this article, particularly Figs. 1–4, 6–8 and 10, are difficult to interpret in black and white. The full color images can be found in the on-line version, at http://dx.doi.org/10.1016/ j.actbio.2015.06.030.

[25]

[26]

[27]

Appendix B. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.actbio.2015.06. 030. References [1] T. Gupta, A. Giangrande, Collective cell migration: ‘‘all for one and one for all’’, J. Neurogenet. 28 (2014) 190–198. [2] M.L. Woods, C. Carmona-Fontaine, C.P. Barnes, I.D. Couzin, R. Mayor, K.M. Page, Directional collective cell migration emerges as a property of cell interactions, PLoS One (2014) 0104969. [3] C.J. Weijer, Collective cell migration in development, J. Cell Sci. 122 (2007) 3215–3223. [4] Olga Ilina, Peter Friedl, Mechanisms of collective cell migration at a glance, J. Cell Sci. 122 (2009) 3203–3208. [5] P. Friedl, D. Gilmour, Collective cell migration in morphogenesis, regeneration and cancer, Nat. Rev. Mol. Cell Biol. 10 (2009) 445–457. [6] L. Rodriguez-Menocal, M. Salgado, D. Ford, E. Van Badiavas, Stimulation of skin and wound fibroblast migration by mesenchymal stem cells derived from normal donors and chronic wound patients, Stem Cells Transl. Med. 1 (2012) 221–229. [7] S. Werner, T. Krieg, H. Smola, Keratinocyte-fibroblast interactions in wound healing, J. Invest. Dermatol. 127 (2007) 998–1008. [8] E.E. Tredget, B. Nedelec, P.G. Scott, A. Ghahary, Hypertrophic scars, keloids, and contractures. The cellular and molecular basis for therapy, Surg. Clin. North Am. 77 (1997) 701–730. [9] W.M. van der Veer, M.C. Bloemen, M.M. Ulrich, G. Molema, P.P. van Zuijlen, E. Middelkoop, F.B. Niessen, Potential cellular and molecular causes of hypertrophic scar formation, Burns 35 (2009) 15–29. [10] W. Cui, L. Cheng, C. Hu, H. Li, Y. Zhang, J. Chang, Electrospun poly(L-lactide) fiber with ginsenoside rg3 for inhibiting scar hyperplasia of skin, PLoS One 8 (2013) e68771. [11] N.D. Evans, E. Gentleman, The role of material structure and mechanical properties in cell–matrix interactions, J. Mater. Chem. B 2 (2014) 2345–2356. [12] C. Zhao, A. Tan, G. Pastorin, H.K. Ho, Nanomaterial scaffolds for stem cell proliferation and differentiation in tissue engineering, Biotechnol. Adv. 31 (2013) 654–668. [13] L.A. Turner, M.J. Dalby, Nanotopography – potential relevance in the stem cell niche, Biomater. Sci. 2 (2014) 1574–1594. [14] X. Yao, R. Peng, J. Ding, Cell-material interactions revealed via material techniques of surface patterning, Adv. Mater. 25 (2013) 5257–5286. [15] T. Ushiki, Collagen fibers, reticular fibers and elastic fibers. A comprehensive understanding from a morphological viewpoint, Arch. Histol. Cytol. 65 (2002) 109–126. [16] H.N. Kim, Y. Hong, M.S. Kim, S.M. Kim, K.Y. Suh, Effect of orientation and density of nanotopography in dermal wound healing, Biomaterials 33 (2012) 8782–8792. [17] S. McDougall, J. Dallon, J. Sherratt, P. Maini, Fibroblast migration and collagen deposition during dermal wound healing: mathematical modelling and

[28]

[29]

[30]

[31]

[32]

[33]

[34] [35]

[36]

[37] [38]

[39]

[40]

[41]

[42] [43]

[44]

clinical implications, Philos. Trans. A Math. Phys. Eng. Sci. 364 (2006) 1385– 1405. S.A. McClain, M. Simon, E. Jones, A. Nandi, J.O. Gailit, M.G. Tonnesen, D. Newman, R.A. Clark, Mesenchymal cell activation is the rate-limiting step of granulation tissue induction, Am. J. Pathol. 149 (1996) 1257–1270. Z. Pan, K. Ghosh, V. Hung, L.K. Macri, J. Einhorn, D. Bhatnagar, M. Simon, R.A. Clark, M.H. Rafailovich, Deformation gradients imprint the direction and speed of en masse fibroblast migration for fast healing, J. Invest. Dermatol. 133 (2013) 2471–2479. H. Jeon, H. Hidai, D.J. Hwang, K.E. Healy, C.P. Grigoropoulos, The effect of micronscale anisotropic cross patterns on fibroblast migration, Biomaterials 31 (2010) 4286–4295. Q. Cheng, B.L.P. Lee, K. Komvopoulos, S. Li, Engineering the microstructure of electrospun fibrous scaffolds by microtopography, Biomacromolecules 14 (2013) 1349–1360. S.H. Chen, Y. Chang, K.R. Lee, Y.J. Lai, A three-dimensional dual-layer nano/ microfibrous structure of electrospun chitosan/poly(D, L-lactide) membrane for the improvement of cytocompatibility, J. Membr. Sci. 450 (2014) 224– 234. S. Heydarkhan-Hagvall, K. Schenke-Layland, A.P. Dhanasopon, F. Rofail, H. Smith, B.M. Wu, R. Shemin, R.E. Beygui, W.R. MacLellan, Three-dimensional electrospun ECM-based hybrid scaffolds for cardiovascular tissue engineering, Biomaterials 29 (2008) 2907–2914. R.G. Flemming, C.J. Murphy, G.A. Abrams, S.L. Goodman, P.F. Nealey, Effects of synthetic micro- and nano-structured surfaces on cell behavior, Biomaterials 20 (1999) 573–588. J.Y. Lim, H.J. Donahue, Cell sensing and response to micro- and nanostructured surfaces produced by chemical and topographic patterning, Tissue Eng. 13 (2007) 1879–1891. C.P. Connolly, A.S.G. Curtis, J.T.A. Dow, Wilkinson CDW. Topographical control of cell behavior. 2. Multiple grooved substrata, Development 108 (1990) 635– 644. D.H. Kim, C.H. Seo, K. Han, K.W. Kwon, A. Levchenko, K.Y. Suh, Guided cell migration on microtextured substrates with variable local density and anisotropy, Adv. Funct. Mater. 19 (2009) 1579–1586. Y. Liu, A. Franco, L. Huang, D. Gersappe, R.A.F. Clark, M.H. Rafailovich, Control of cell migration in two and three dimensions using substrate morphology, Exp. Cell Res. 315 (2009) 2544–2557. Y. Liu, Y. Ji, K. Ghosh, R.A.F. Clark, L. Huang, M.H. Rafailovich, Effects of fiber orientation and diameter on the behavior of human dermal fibroblasts on electrospun PMMA scaffolds, J. Biomed. Mater. Res. 90A (2009) 1092–1106. M. Nikkhah, F. Edalat, S. Manoucheri, A. Khademhosseini, Engineering microscale topographies to control the cell-substrate interface, Biomaterials 33 (2012) 5230–5246. D.M. Brunette, B. Chehroudi, The effects of the surface topography of micromachined titanium substrata on cell behavior in vitro and in vivo, J. Biomech. Eng. 121 (1999) 49–57. C. Oakley, N.A. Jaeger, D.M. Brunette, Sensitivity of fibroblasts and their cytoskeletons to substratum topographies: topographic guidance and topographic compensation by micromachined grooves of different dimensions, Exp. Cell Res. 234 (1997) 413–424. K. Kolind, A. Dolatshahi-Pirouz, J. Lovmand, F.S. Pedersen, M. Foss, F. Besenbacher, A combinatorial screening of human fibroblast responses on micro-structured surfaces, Biomaterials 31 (2010) 9182–9191. J.P. Kaiser, A. Reinmann, A. Bruinink, The effect of topographic characteristics on cell migration velocity, Biomaterials 27 (2006) 5230–5241. D.H. Kim, K. Han, K. Gupta, K.W. Kwon, K.Y. Suh, A. Levchenko, Mechanosensitivity of fibroblast cell shape and movement to anisotropic substratum topography gradients, Biomaterials 30 (2009) 5433–5444. A. Diener, B. Nebe, F. Lüthen, P. Becker, U. Beck, H.G. Neumann, J. Rychly, Control of focal adhesion dynamics by material surface characteristics, Biomaterials 26 (2005) 383–392. D.H. Kim, D. Wirtz, Focal adhesion size uniquely predicts cell migration, FASEB J. 27 (2013) 1351–1361. R.M. Saunders, M.R. Holt, L. Jennings, D.H. Sutton, I.L. Barsukov, A. Bobkov, R.C. Liddington, E.A. Adamson, G.A. Dunn, D.R. Critchley, Role of vinculin in regulating focal adhesion turnover, Eur. J. Cell Biol. 85 (2006) 487–500. D.W. Dumbauld, H. Shin, N.D. Gallant, K.E. Michael, H. Radhakrishna, A.J. García, Contractility modulates cell adhesion strengthening through focal adhesion kinase and assembly of vinculin-containing focal adhesions, J. Cell. Physiol. 223 (2010) 746–756. D.W. Dumbauld, K.E. Michael, S.K. Hanks, A.J. García, Focal adhesion kinasedependent regulation of adhesive forces involves vinculin recruitment to focal adhesions, Biol. Cell 102 (2010) 203–223. N. Wang, J.D. Tytell, D.E. Ingber, Mechanotransduction at a distance: mechanically coupling the extracellular matrix with the nucleus, Nat. Rev. Mol. Cell Biol. 10 (2009) 75–82. P. Friedl, K. Wolf, J. Lammerding, Nuclear mechanics during cell migration, Curr. Opin. Cell Biol. 23 (2011) 55–64. K. Wolf, M. Te Lindert, M. Krause, S. Alexander, J. Te Riet, A.L. Willis, R.M. Hoffman, C.G. Figdor, S.J. Weiss, P. Friedl, Physical limits of cell migration: control by ECM space and nuclear deformation and tuning by proteolysis and traction force, J. Cell Biol. 201 (2013) 1069–1084. Z. Pan, K. Ghosh, Y. Liu, R.A. Clark, M.H. Rafailovich, Traction stresses and translational distortion of the nucleus during fibroblast migration on a physiologically relevant ECM mimic, Biophys. J. 96 (2009) 4286–4298.

S. Qin et al. / Acta Biomaterialia 25 (2015) 230–239 [45] T.B. Sorensent, M. Kjaer, Agarose microdroplet leucocyte migration technique for detection of cell-mediated hypersensitivity to PPD and renal carcinoma antigen, Clin. Exp. Immunol. 30 (1977) 168–172. [46] C. Morita, T. Okada, H. Izawa, M. Soekawa, Agarose droplet method of macrophage migration-inhibition test for newcastle disease virus in chickens, Avian Dis. 20 (1975) 230–235. [47] H.L. Wiggins, J.Z. Rappoport, An agarose spot assay for chemotactic invasion, Biotechniques 48 (2010) 121–124. [48] C.C. Liang, A.Y. Park, J.L. Guan, In vitro scratch assay: a convenient and inexpensive method for analysis of cell migration in vitro, Nat. Protoc. 2 (2007) 329–333. [49] W.J. Ashbya, A. Zijlstra, Established and novel methods of interrogating twodimensional cell migration, Integr. Biol. 4 (2012) 1338–1350. [50] K.I. Hulkower, R.L. Herber, Cell migration and invasion assays as tools for drug discovery, Pharmaceutics 3 (2011) 107–124. [51] A.S. Nain, M. Sitti, A. Jacobson, T. Kowalewski, C. Amon, Dry spinning based spinneret based tunable engineered parameters (step) technique for controlled and aligned deposition of polymeric nanofibers, Macromol. Rapid Commun. 30 (2009) 1406–1412. [52] A.S. Nain, J.A. Phillippi, M. Sitti, J. Mackrell, P.G. Campbell, C. Amon, Control of cell behavior by aligned micro/nanofibrous biomaterial scaffolds fabricated by spinneret-based tunable engineered parameters (STEP) technique, Small 4 (2008) 1153–1159. [53] A. Livne, E. Bouchbinder, B. Geiger, Cell reorientation under cyclic stretching, Nat. Commun. 5 (2014) 3938. [54] L.E. Bowes, M.C. Jimenez, E.D. Hiester, M.S. Sacks, J. Brahmatewari, P. Mertz, W.H. Eaglstein, Collagen fiber orientation as quantified by small angle light scattering in wounds treated with transforming growth factor-beta2 and its neutalizing antibody, Wound Repair Regen. 7 (1999) 179–186. [55] M. Abrigo, S.L. McArthur, P. Kingshott, Electrospun nanofibers as dressings for chronic wound care: advances, challenges, and future prospects, Macromol. Biosci. 14 (2014) 772–792.

239

[56] D.B. Khadka, D.T. Haynie, Protein- and peptide-based electrospun nanofibers in medical biomaterials, Nanomedicine 8 (2012) 1242–1262. [57] J. Venugopal, S. Ramakrishna, Applications of polymer nanofibers in biomedicine and biotechnology, Appl. Biochem. Biotechnol. 125 (2005) 147– 158. [58] J.C. Kuo, Mechanotransduction at focal adhesions: integrating cytoskeletal mechanics in migrating cells, J. Cell Mol. Med. 17 (2013) 704–712. [59] H. Hirata, H. Tatsumi, M. Sokabe, Mechanical forces facilitate actin polymerization at focal adhesions in a zyxin-dependent manner, J. Cell Sci. 121 (17) (2008) 2795–2804. [60] C. Ciobanasu, B. Faivre, C. Le Clainche, Actin dynamics associated with focal adhesions, Int. J. Cell Biol. 2012 (2012) 941292. [61] U.S. Schwarz, S.A. Safran, Physics of adherent cells, Rev. Mod. Phys. 85 (2013) 1327. [62] E. Puklin-Faucher, M.P. Sheetz, The mechanical integrin cycle, J. Cell Sci. 122 (2009) 179–186. [63] I. Thievessen, P.M. Thompson, S. Berlemont, K.M. Plevock, S.V. Plotnikov, A. Zemljic-Harpf, R.S. Ross, M.W. Davidson, G. Danuser, S.L. Campbell, C.M. Waterman, Vinculin-actin interaction couples actin retrograde flow to focal adhesions, but is dispensable for focal adhesion growth, J. Cell Biol. 202 (2013) 163–177. [64] T. Shemesh, B. Geiger, A.D. Bershadsky, M.M. Kozlov, Focal adhesions as mechanosensors: a physical mechanism, Proc. Natl. Acad. Sci. USA 102 (2005) 12383–12388. [65] R.D. Mullins, J.A. Heuser, T.D. Pollard, The interaction of Arp2/3 complex with actin: nucleation, high affinity pointed end capping, and formation of branching networks of filaments, Proc. Natl. Acad. Sci. USA 95 (1998) 6181– 6186. [66] J. Qian, H. Liu, Y. Lin, W. Chen, H. Gao, A mechanochemical model of cell reorientation on substrates under cyclic stretch, PLoS One (2013) 0065864.