Localized application of mechanical and biochemical stimuli in 3-D culture

Authors


Abstract

The goal of this study was to investigate the responses of isolated cells in 3-D culture to localized application of mechanical and biochemical signals. Corneal fibroblasts were plated inside collagen matrices for 24 hours, then imaged using time-lapse DIC. For mechanical perturbation, a microinjection needle (Femtotip) was inserted axially into the ECM, then displaced laterally to alter local ECM stress. For biochemical stimulation, PDGF or vehicle control solution was microinjected into the matrix. Compressing the ECM perpendicular to the cell axis had no appreciable effect on cell behavior. However, pushing the ECM parallel to the cell axis induced rapid cellular contraction, followed by secondary cell spreading and tractional force generation. Injection of PDGF induced a similar cell spreading response. Cells in 3-D matrices showed remarkable plasticity, and extension of pseudopodia could be induced at both the leading and trailing edges of migrating cells. Developmental Dynamics 237:2726–2736, 2008. © 2008 Wiley-Liss, Inc.

INTRODUCTION

Cell-matrix mechanical interactions play a defining role in a gamut of biological processes such as developmental morphogenesis and wound healing. During embryonic development, physical forces exerted by mesenchymal cells organize extracellular matrix (ECM) into a wide variety of spatial patterns whose mechanical properties lend structural support and give form and organization to vertebrate tissue (Bard and Hay,1975; Bard and Higginson,1977; Stopak and Harris,1982; Stopak et al.,1985). During corneal development, fibroblasts migrate into the acellular primary stroma and deposit and organize the extracellular matrix into orthogonal lamellae, which provide a unique combination of mechanical strength and optical transparency (Trelstad and Coulombre,1971; Bard and Hay,1975; Cintron et al.,1983). The process of corneal wound repair following lacerating injury or refractive surgery also depends upon mechanical processes such as the migration of activated keratocytes (corneal fibroblasts) into the wound from the surrounding stroma, apposition of the wound edges (wound contraction), and extracellular matrix reorganization (remodeling). As in development, these biomechanical mechanisms ultimately control corneal clarity and refractive (visual) outcome (Petroll et al.,1992; Moller-Pedersen et al.,1997,1998a,b; Jester et al.,1999; Dupps and Wilson,2006).

Historically, our ability to investigate cell mechanical behavior has been limited, in part, by the technical challenges associated with simultaneous imaging of cell activity and measurement of cellular forces. Most previous studies of cell mechanical behavior have used planar elastic substrates (Harris et al.,1980; Lee et al.,1994; Pelham and Wang,1997; Balaban et al.,2001; Beningo et al.,2001). Although these elegant studies have provided important insights into cell mechanical behavior, cells reside within 3-D extracellular matrices in vivo, and ECM geometry has been shown to effect cell morphology, mechanical activity, and adhesion organization and composition (Bard and Hay,1975; Tomasek et al.,1982; Doane and Birk,1991; Friedl and Brocker,2000; Cukierman et al.,2001,2002; Abbott,2003; Grinnell et al.,2003; Rhee et al.,2007).

An alternative to planar elastic substrates is the fibroblast populated collagen matrix model, in which cells are plated inside a 3-D fibrillar collagen matrix (Elsdale and Bard,1972; Bell et al.,1979; Stopak and Harris,1982; Grinnell and Lamke,1984). As originally demonstrated by Hay and coworkers, time-lapse DIC imaging can be used to assess the dynamic behavior of both cells and the surrounding collagen fibril organization in this model (Bard and Hay,1975; Tomasek et al.,1982; Tomasek and Hay,1984). This approach has been adapted and applied to determine the role of specific cytoskeletal and adhesive proteins in mediating cell-matrix mechanical interactions, and to investigate the underlying signaling pathways that regulate various aspects of cell motility (Petroll and Ma,2003; Petroll et al.,2003,2004a; Vishwanath et al.,2003). We have also used this model to investigate the response of corneal fibroblasts to local changes in extracellular matrix (ECM) tension, by using micropipettes to deform the collagen matrix parallel to the long axis of isolated corneal fibroblasts, and measuring the cellular response over time (Petroll et al.,2004b). In this study, we use smaller microneedles (Femtotips) to allow more precise control over ECM deformation, and compare the cellular response to changes in ECM tension along different directional axes. We also investigate whether these microinjection needles can be used to study the effects of localized application of a cytokine on dynamic cell behavior in 3-D collagen matrices. Our results demonstrate for the first time that the dynamic response of corneal fibroblasts to localized changes in ECM stress is highly dependent upon the spatial orientation of those stresses with respect to the cell. Fibroblasts in 3-D matrices showed remarkable plasticity, and both mechanical and biochemical stimulation could induce pseudopodial extension at the leading and trailing edges of migrating cells.

RESULTS

ECM Deformation Using Small Microneedles

Femtotips were easily inserted axially into the ECM without causing significant deformation (see Suppl. Movie 1, which can be viewed online), and subsequent downward displacement of the microneedle appeared to induce localized ECM compression in front of the needle. To demonstrate this quantitatively, we tracked the ECM deformation induced by pushing with microneedles (Fig. 1A and B, red tracks, cross marks starting position). These tracks were then used to develop FEM models. Principal strain vectors reveal that although some tension is produced perpendicular to the vertical axis in front of the needle (Fig. 1C, small white vectors), compression is clearly the dominant effect (larger blue vectors). FEM strain maps parallel to the direction of movement (y-axis) confirm that there was significant compression in front of femtotips (Fig. 1D, blue areas), with the maximum effect just in front of the needle. These responses are similar to what we have observed previously using large (35-μm diameter) glass pipettes for ECM manipulation; however, both the area affected and the magnitude of the compression were smaller when using Femtotips.

Figure 1.

ECM deformation using small microneedle. A: Femtotips were easily inserted axially into the ECM (from right to left on image) without causing significant deformation. B: Subsequent downward displacement of the microneedle by 50 μm induced localized ECM compression (ECM displacements are indicated by red tracks; cross marks starting position prior to pushing the needle). C: Principal strain vectors generated from the ECM displacements using finite element modeling: blue = compression, white = tension. D: Map of strain along the y-axis generated from the ECM displacements using finite element modeling. Pushing on the ECM with a needle caused matrix compression (blue) in front of and tension (red and orange) behind the region of contact of the needle. Scale bars at bottom are in dimensionless units ΔL/L (change in length/initial length). Negative values indicate compression and positive values indicate tension.

Fibroblast Response to ECM Compression Parallel to Long Axis

By one day after plating inside 3-D collagen matrices, both NRK and HTK cells generally had a bipolar morphology with thin pseudopodial processes; consistent with previous observations (Petroll and Ma,2003; Petroll et al.,2003). Cells were always aligned nearly parallel to the dish on which the collagen matrix was plated. Some cells develop stress fibers within the cell body (Kim et al.,2006), although these are not as prominent as those observed when cells were plated on rigid substrates. In summary, the cells have the features of activated corneal fibroblasts or protomyofibroblasts, as opposed to quiescent corneal keratocytes, which have a dendritic morphology and a cortical f-actin organization (Jester and Chang,2003).

DIC imaging allowed detailed visualization of the cells and the individual collagen fibrils surrounding them. Fibroblasts repeatedly extended and retracted pseudopodia at both 1 and 2 days after plating within 3-D matrices (Petroll and Ma,2003). Initial axial insertion of the microneedle into the ECM did not alter the normal pattern of cell behavior (compare Fig. 2A and B). Pushing the ECM toward a cell with a microneedle induced rapid cellular shortening (contraction) with corresponding ECM compression along the cell body (Fig. 2C, arrows; Suppl. Movie 2). Following this initial contraction, cell spreading was observed, and tractional force was generated as indicated by pulling in of the ECM (Fig. 2D, red tracks). This secondary spreading response is best appreciated in time-lapse movies (Suppl. Movie 3). Note that during the secondary spreading response, large amounts of ECM displacement were often produced for relatively small amounts of pseudopodial extension, particularly for large needle displacements. Compaction of collagen at the base of the pseudopodia was also often observed as new pseudopodial extensions pulled the collagen fibrils inward. As demonstrated by GFP-tubulin labeling (Fig. 3), re-spreading involved extension of existing processes as well as formation of new pseudopodia. In addition, there was apparent compression of microtubules following needle push, and decompression following secondary spreading. Overall, these responses are similar to what we have observed previously using large (35-μm diameter) glass pipettes for ECM manipulation; however, the magnitude of the responses was reduced when using Femtotips.

Figure 2.

Fibroblast response to ECM compression using small microneedle. A–D: Human corneal fibroblast 2 days after plating inside collagen matrix. Needle was inserted axially into the matrix above the cell (A) without inducing changes in cell behavior (B). Pushing on the ECM towards the leading edge of a cell induced rapid cellular contraction and ECM compression along the cell body (C, arrows). This initial contraction was followed by re-spreading and tractional force generation (D, red tracks; crosses mark position immediately after needle push). E–H: Rabbit corneal fibroblast moving downward. Following needle push, spreading is initiated at the trailing edge of the cell.

Figure 3.

A: Human corneal fibroblast transfected to express GFP-tubulin, 1 day after plating inside collagen matrix. Pushing the microneedle toward the cell (white arrow) induced an immediate rapid cellular contraction (B), followed by re-spreading and tractional force generation (C, red tracks, crosses mark position immediately following needle push). Re-spreading involved extension of existing processes as well as formation of new pseudopodia.

We also studied, for the first time, the cellular response to ECM compression at the trailing edge of a migrating cell (Fig. 2E–H). As the cell migrated slowly downward (Suppl. Movie 4), extension of pseudopodia at the leading edge was associated with tractional force generation (inward movement of collagen fibrils). At the rear of the cell, apparent rupture of cell-matrix adhesions led to elastic recoil of a cell process and release of ECM tension (as indicated by collagen movement away from the cell). Surprisingly, small extensions and retractions of pseudopodia and associated collagen fibril displacements were observed at the trailing edge of the cell throughout this process. Furthermore, when the needle was pushed toward the trailing edge, both an initial contraction and secondary spreading response (Suppl. Movie 5) were observed, similar to that identified following needle push at the leading edge of cells. These data suggest that similar cytoskeletal machinery and/or signaling networks may exist at both ends of slowly migrating cells, facilitating rapid responses to mechanical stimuli at either end.

To quantify the amount of initial cellular contraction in response to ECM micromanipulation, we measured the distance between landmarks at the ends of the cell and calculated the distance between them. Analysis was performed on experiments in which the needle was pushed 40 μm, and both ends of the cells and their surrounding matrix were clearly visible. All cells analyzed showed some amount of contraction immediately after needle push, with an average shortening of 9.9 ± 3.4% (n = 4); secondary spreading was also observed (Suppl. Movie 6). Note that this is substantially less than the amount observed previously using large microneedles (26.9 ± 6.9%) (Petroll et al.,2004b). Since some amount of passive cellular shortening would be expected due to the effect of the needle push alone, we assessed shortening following treatment with cytochalasin D. Incubation with cytochalasin D resulted in cell elongation and relaxation of cell-induced ECM stress (not shown). Cellular shortening after pushing with the needle was significantly reduced (4.5 ± 1.2%, P < 0.05). In addition, no secondary cell activity or ECM displacement was observed after the needle push following treatment with cytochalasin D (Suppl. Movie 7), confirming that these were active cell responses that do not occur in the absence of cellular tension. Similarly, pre-incubation with Y-27632 resulted in cell elongation and relaxation of cell-induced ECM tension, and blocked both the initial contraction and subsequent traction normally induced by pushing the needle toward the cell (not shown). It should be noted that for all experimental manipulations, similar responses were observed for NRK and HTK cells.

Fibroblast Response to ECM Compression Perpendicular to Long Axis

Pushing small microneedles toward the side of cells had no significant effect on cell morphology or tractional force generation (Fig. 4, Suppl. Movie 8). Some cells were transfected to express GFP-α-actinin, which incorporates into focal adhesions, stress fibers (when present), and the ruffling edge of lamellipodia (Lazarides and Burridge,1975; Jester et al.,1994; Katoh et al.,2001). Other cells were transfected to express GFP-zyxin, which more specifically labels cell-matrix adhesions. ECM compression perpendicular to the cell axis did not stimulate large changes in ruffling at the pseudopodial tips (Fig. 5A–D), or induce significant changes in the organization of cell-matrix adhesions (Fig. 5E–G, arrows). It should be noted that zyxin is a component of both focal complexes (point contacts) and focal adhesions (focal contacts) (Rottner et al.,2001). While we did not attempt to distinguish between these adhesive structures, in general, focal complexes have been associated with lamellipodia in 2-D culture and are thought to play a key role in protrusion and traction at the leading edge of migrating cells (Rottner et al.,1999; Beningo et al.,2001; Kaverina et al.,2002).

Figure 4.

DIC imaging of the fibroblast response to ECM compression adjacent to cell body. Human corneal fibroblast 1 day after plating inside collagen matrix. Pushing small microneedles toward the side of the cell had no significant effect on cell morphology or tractional force generation.

Figure 5.

A–D: Human corneal fibroblast transfected to express GFP-α-actinin, 2 days after plating inside collagen matrix. Pushing the microneedle toward the side of the cell (white arrow) had no appreciable effect on cell morphology or tractional force generation (C, red tracks, crosses mark position immediately following needle push). Significant cell ruffling was not observed. E–G: Rabbit corneal fibroblast transfected to express GFP-zyxin, 1 day after plating inside collagen matrix. Pushing the microneedle toward the side of the cell (white arrow) did not induce significant changes in the organization of cell-matrix adhesions (arrows).

Localized Cytokine Injection

Cells randomly extended and retracted cell processes both before needle insertion, and following vehicle control injections. Significant changes in cell morphology were not generally observed (Fig. 6). In contrast, injection of PDGF induced rapid cell spreading toward the microneedle tip, with both extension of existing pseudopodia and formation of new processes (Fig. 7, Suppl. Movie 9). Tractional force was generated during PDGF-induced spreading, as indicated by pulling in of individual collagen fibrils in front of extending pseudopodia (Fig. 7B, red tracks). While spreading was generally larger at the end of the cell closest to the needle, spreading at both ends was often observed. Overall, a significant increase in both cell length and cell area was observed at both 60 and 90 min following injection of PDGF (n = 6 cells) as compared to vehicle control (n = 4 cells) (Fig. 8, two-way repeated measures ANOVA, Holm-Sidak multiple comparison procedure). Rapid formation and turnover of focal adhesions and ruffling of extending pseudopodial tips was also revealed by GFP-α-actinin labeling (Fig. 9). The same response was observed when the needle was retracted following the injection of PDGF.

Figure 6.

Fibroblast response to injection of control solution in front of cell. Significant changes in cell morphology were not generally observed.

Figure 7.

Fibroblast response to injection of PDGF in front of cell. Both extension of existing pseudopodia toward the injection site and formation of new processes was observed. Tractional force was generated during PDGF-induced spreading, as indicated by pulling in of individual collagen fibrils in front of extending pseudopodia (red tracks, crosses mark initial position).

Figure 8.

Changes in cell length (A) and projected cell area (B) induced by localized microinjection (error bars represent standard deviations). A significant increase in both cell length and cell area was observed following injection of PDGF (n = 6 cells) as compared to vehicle control (n = 4 cells) at both 60 and 90 min. *Two-way repeated measures ANOVA, Holm-Sidak multiple comparison procedure.

Figure 9.

Maximum intensity projections of a human corneal fibroblast transfected to express GFP-α-actinin, 1 day after plating. A: Ten minutes prior to injection of PDGF. B: Thirty minutes after injecting PDGF in front of cell. Rapid formation and turnover of focal adhesions and ruffling of extending pseudopodial tips was revealed by GFP-α-actinin labeling.

Two cells underwent significant translocation toward the femtotip following microinjection of PDGF. An example of rapid cell migration is shown in Supplemental Movie 10. As the cell migrated to the right, extension of pseudopodia and tractional force generation at the leading edge was observed. At the rear of the cell, apparent rupture of cell-matrix adhesions led to elastic recoil of cell processes and release of ECM tension (as indicated by collagen movement away from the cell). Note that extension and protrusion of a new process is observed at the front of the cell simultaneous with the large retraction at the rear, consistent with the concept that reducing cellular tension stimulates cell spreading. Interestingly, small extensions and retractions of pseudopodia were often still observed at the trailing edge of the cell during cell migration (circles).

DISCUSSION

It is well established that mechanical stimuli play a key role in regulating growth and function in a variety of cell types (Sadoshima and Izumo,1997; Tummina et al.,1998; Liu et al.,1999; Brown,2000; Shyy and Chien,2002; Guo et al.,2006). During embryonic development, physical forces exerted by mesenchymal cells organize extracellular matrix (ECM) into a wide variety of spatial patterns, and feedback between cell and matrix mechanics has long been thought to be a key factor regulating this process (Bard and Hay,1975; Bard and Higginson,1977; Stopak and Harris,1982; Stopak et al.,1985; Krieg et al.,2008). For example, the stiffness of planar substrates in 2-D cell culture models plays a central role in directing human mesenchymal stem cells along neuronal, muscle, or bone lineages (Engler et al.,2006), and these effects are mediated by cellular force generation via nonmuscle myosin II. As shown by Bard and Hay, migrating mesechymal cells are in close proximity during corneal stromal development, thus localized alterations in effective matrix stiffness due to the activity of nearby cells are likely to occur (Bard and Hay,1975). In the current study, we use small microneedles (Femtotips) to allow precise local control over 3-D ECM deformation, and investigate the cellular responses to changes in localized ECM tension along different directional axes.

Changing local ECM stress using femtotips induced rapid and reproducible response patterns in both rabbit and human corneal fibroblasts. Reducing effective ECM stiffness by pushing the needle parallel to the long axis of a cell resulted in rapid cellular shortening with corresponding ECM compression along the cell body. This initial shortening is likely due to the release of pre-existing cellular contractile forces, since the response was blocked by treatment with cytochalasin D, or Rho kinase inhibition. These data demonstrate that there is dynamic feedback between cytoskeletal forces and local ECM stress that regulates corneal fibroblast mechanical behavior within 3-D matrices. Other researchers have used a culture force monitor to measure how dermal fibroblasts within 3-D collagen matrices respond to global changes in tensional loading (Brown et al.,1998; Eastwood et al.,1994). Their data suggests that cells within 3-D matrices alter their contractility in response to changes in mechanical loading in a way that maintains “tensional homeostasis” (constant tension) in their surrounding matrix. Our data on local feedback between cells and ECM is also consistent with the tensional homeostasis model, since following the reduction of tension induced by pushing the needle toward the cell, fibroblasts attempt to re-establish baseline tension by actively pulling in the ECM.

In contrast to ECM compression parallel to the long axis of cells, compressing the ECM perpendicular to the long axis had little effect on cell morphology or mechanical activity. This is also consistent with the tensional homeostasis model, since the cytoskeleton, focal adhesions, and contractile forces are all aligned parallel to the long axis of bipolar cells, and therefore reducing the effective stiffness of the ECM alongside of the cell should have little impact on cellular tension.

We observed similar response patterns from corneal fibroblasts derived from elderly human tissue as well as those from young rabbit eyes. Many aspects of the mechanical behavior of these ocular cells have been shown previously to be similar to that of dermal fibroblasts (Roy et al.,1999a; Jester and Chang,2003; Kim et al.,2006; Karamichos et al.,2007; Kim and Petroll,2007). Furthermore, the morphological changes these cells undergo during initial spreading and migration in 3-D culture are nearly identical to that of chick embryo fibroblasts. Thus, we hypothesize that tensional homeostasis may be a conserved mechanism that regulates mesenchymal cell behavior both during developmental morphogenesis and adult wound healing. Additional experiments using embryonic-derived cells are clearly needed to test this hypothesis.

The Rho-family of small GTPases such as Rho, Rac, and Cdc42 play a central role in regulating the cytoskeletal changes associated with cell spreading, migration, and contraction (Hall,2005; Jaffe and Hall,2005; Bustelo et al.,2007). These GTP binding proteins function as molecular switches; alternating between the active GTP-bound state and the inactive GDP-bound state. Previous studies suggest that Rho and Rac may be involved in the cellular response to a variety of mechanical signals. For example, in vascular smooth muscle cells, non-cyclic uniaxial mechanical stretching was shown to downregulate Rac and suppress cell spreading, whereas decreasing mechanical tension (by inhibiting Rho kinase or myosin light chain kinase) increased cell spreading through upregulation of Rac (Katsumi et al.,2002). A reduction in mechanical tension and increased spreading by corneal fibroblasts has also been demonstrated following Rho kinase inhibition in 3-D matrices (Vishwanath et al.,2003). In the current study, cell spreading and tractional force generation by corneal fibroblasts was observed after reducing ECM tension by pushing with microneedles. A similar spreading response was induced by local ECM microinjection of PDGF, which activates Rac (Sander et al.,1999; Grinnell,2000). Taken together, the data suggest that the interplay between Rho and Rac activation may play a central role in the fibroblast response to local mechanical stimulation. Additional studies more specifically targeting Rho and Rac signaling pathways are needed to clarify the molecular mechanisms underlying these important processes.

The cellular response to transient mechanical stimulation using microneedles has been investigated previously using planar elastic polyacrylamide substrates (Lo et al.,2000; Wang et al.,2001). The advantage of this model is that the mechanical properties of the substrate can be fully characterized, and precise mapping of cellular forces in response to mechanical or biochemical stimulation can been achieved. In this model, pushing the substrate toward the leading edge of a cell caused the cell to retract its leading edge and migrate away from the needle, a phenomenon termed “durotaxis.” We did not observe this type of migratory behavior in response to mechanical stimulation in the current study. Instead, cell spreading toward the microneedle was observed, even when the ECM was compressed at the rear of a migrating cell. This disparity is likely due, in part, to differences in the geometry and mechanical properties of planar elastic substrates and 3-D fibrillar matrices, as well as the fact that corneal fibroblasts are generally less migratory in our 3-D model. While durotaxis likely plays a role in modulating cell migration within collagen matrices, tensional homeostasis appears to dominate cell behavior under the conditions used in the current study.

Localized application of cytokines or peptides can be a useful tool for assessing mechanical behavior at the subcellular level. For instance, Wang and coworkers compared the effects of local application of the GRGDTP peptide to selectively disrupt substrate adhesions at the front and rear of migrating 3T3 fibroblasts on planar substrates (Munevar et al.,2001), and mechanical interactions were found to be distinctly different at the leading and trailing edge of the cells. In the current study, we investigated whether microinjection needles could be used to study the effects of localized application of a cytokine on dynamic cell behavior in 3-D collagen matrices. Femtotips could be inserted axially into the matrix without breaking the tip or significantly deforming the collagen organization surrounding the cells, and injection of control solution did not alter cell mechanical activity. In contrast, injection of PDGF BB induced rapid cell spreading and elongation, similar to that observed following matrix compression along the cell axis. Overall, local microinjection into 3-D collagen matrices may be a promising new approach for investigating mechano-regulation of corneal fibroblasts at the subcellular level.

Elizabeth Hay was a pioneer in live-cell imaging, and was the first to use DIC imaging to visualize corneal embryonic fibroblast migration in situ and within 3-D collagen matrices. While important insights into the mechanisms of cell migration were gained in these studies, recent advances in microscope optics and digital imaging technology allow these processes to be studied with much higher temoral and spatial resolution using current microscope systems. In this study, some cells underwent significant translocation either spontaneously in serum containing media, or following local microinjection of PDGF into serum-free media. In both cases, extension of pseudopodia and tractional force generation at the leading edge of migrating cells was observed. At the rear of the cells, apparent rupture of cell-matrix adhesions led to elastic recoil of cell processes and release of ECM tension (as indicated by collagen movement away from the cell). Extension of existing processes or protrusion of a new process was often observed at the front of the cell simultaneously with a large retraction at the rear, consistent with the concept that reducing cellular tension stimulates cell spreading. Interestingly, small extensions and retractions of pseudopodia and associated collagen fibril displacements were detected at the trailing edge of cells during cell migration. Furthermore, when the needle was pushed toward the trailing edge of migrating cells, both the initial contraction and secondary spreading response were still observed. These novel findings suggest that similar cytoskeletal machinery and/or signaling networks may be present to some extent at both the front and rear of migrating cells, facilitating remarkable plasticity and rapid responses to mechanical stimuli at either end.

EXPERIMENTAL PROCEDURES

Cells

Studies were performed using both primary rabbit corneal fibroblasts (NRK) and a previously characterized telomerase-infected, extended life-span human corneal fibroblast cell line, HTK (Jester et al.,2003). NRK cells were harvested from New Zealand White Albino rabbit eyes (Pel-Freez, Rogers, AR) as previously described (Petroll and Ma,2003). Both cell types were cultured in 25 cm2 tissue culture flasks (Costar, Cambridge, MA) using “complete media” consisting of Dulbecco's modified Eagle's medium (DMEM, GIBCO Invitrogen Cell Culture, Carlsbad, CA) supplemented with 1% Penicillin, 1% Streptomycin, and 1% Fungisome (Biowhitaker, Inc., Wakersville, MD) and 10% fetal bovine serum (FBS, Sigma Chemical Corp., St. Louis, MO). For some experiments, cells were transfected to express GFP-tubulin, GFP-zyxin, or GFP-α-actinin as previously described (Vishwanath et al.,2003; Kim and Petroll,2007).

Collagen Matrices

Hydrated collagen matrices were prepared by mixing neutralized bovine dermal collagen (Vitrogen 100; Collagen Corporation, Palo Alto, CA) with 10× DMEM to achieve a final collagen concentration of 2.48 mg/ml (Petroll and Ma,2003). For plating cells inside the matrix, a 50-μl suspension of NRK or HTK cells was mixed with 500 μL of collagen solution. The cell/collagen mixture was pre-incubated at 37°C for 5 min, and 30-μl aliquots (containing approximately 1,000 cells) were then poured onto Delta T culture dishes (Bioptechs, Inc., Butler, PA). Each aliquot was spread over a central 12-mm diameter circular region on the dish and was approximately 100 μm thick. The dish was then placed in a humidified incubator (37°C, 5% CO2) for 60 min for polymerization, and overlaid with either complete media or serum-free media. consisting of DMEM supplemented with 1% RPMI 1640 Vitamin and Glutathione mix, 1% ascorbic acid, and 1% MEM non-essential amino acids (Jester et al.,1994).

Time-Lapse Digital Imaging

Microscopy was performed as previously described (Petroll and Ma,2003; Petroll et al.,2003). Briefly, we used a Nikon TE300 inverted microscope with fluorescence and DIC imaging modules, two high-speed filter wheels for rapid selection of excitation and emission filters and shuttering of epifluorescent illumination, and a high-resolution cooled CCD camera (CoolSnap HQ, Roper Scientific, Tuscan, AZ). The hardware was controlled using a PC running MetaVue (Universal Imaging Corp., Downingtown, PA). To maintain cell viability during imaging, a Bioptechs microincubation system and objective heater was used (Bioptechs, Inc., Butler, PA). A Bioptechs microperfusion pump was used to continuously perfuse the cells while on the microscope stage with complete media containing HEPES buffer at the rate of 6 ml/hr.

Dishes were moved to the microscope stage 1–2 days after seeding on the gel; this allowed them to develop a more consistent bipolar spindle-shaped morphology as is observed during in vivo wound healing (Petroll et al.,1993; Moller-Pedersen et al.,1998a,c; Cukierman et al.,2001). In each experiment, cells were allowed to acclimate to the microincubation system for 30–60 min prior to time-lapse imaging. The cell density was sparse enough to focus on the mechanical activity of a single cell, minimizing the potential interference caused by neighboring cells. The activity of a single cell was imaged using either a 40× or 60× oil immersion or a 20× dry objective. Nomarski differential interference contrast (DIC) images and/or enhanced green fluorescent protein (EGFP) fluorescent images were automatically acquired at 1–3-min intervals using MetaVue. In most experiments, 3-D datasets were obtained at each time point by repeating the acquisition at 4–5 sequential focal planes in z steps of 2–3 μm. To minimize phototoxicity, neutral density filters and 2×2 on-chip camera binning was used for EGFP imaging.

Micromanipulation of ECM

In order to alter matrix tension, a glass microneedle (Femtotip) attached to a Narishige micromanipulator was used. Femtotip microneedles have an inner diameter of approximately 0.5 μm and an outer diameter of approximately 1 μm. Needles were positioned at a 45° angle to the microscope stage. Following 30 min of time-lapse imaging of a cell of interest, the needle was inserted axially into the collagen lattice 50–80 μm from the end (18 cells) or to the side (8 cells) of a cell of interest. After inserting the needle, time-lapse imaging was performed for an additional 20–30 min. The needle was then pushed toward the cell (30–50 μm) to compress the collagen ECM, thereby decreasing the effective matrix stiffness. All manipulations were visualized using DIC imaging. We generally selected cells that were less than 25 μm from the top of the collagen matrix, to minimize free body motion of the gel. The micropipette was inserted only far enough so that the tip was in the same focal plane as the cell. After pushing on the ECM, time-lapse imaging was continued for an additional 2–3 hr.

In 3 experiments, cells were treated with the Rho-kinase inhibitor Y-27632 40 minutes prior to pushing the needle toward the end of the cell. Four microliters of a 5-mM stock solution of Y-27632 was added to the Bioptechs culture dish to achieve a final concentration of 10 μM, and the perfusion media was simultaneously switched to complete media containing 10 μM Y-27632 (Vishwanath et al.,2003). In 4 experiments, cytochalasin D (25 μM) was added to the media 40 min prior to micromanipulation using a similar approach. Control experiments were also performed on collagen lattices without cells to determine the effect of needle pushing alone (i.e., without cellular force generation) on the pattern of ECM deformation.

Localized Cytokine Injection

Following 1 hr of time-lapse DIC imaging, a glass microneedle (Femtotip) loaded with either PDGF BB (100 mg/ml, 12 cells) or vehicle control solution (4 mM HCl in 0.1%BSA, 4 cells) was attached to a FemtoJet microinjection system. The microneedle was then inserted axially into the ECM approximately 25–50 μm from the leading edge of an isolated cell of interest, as described above. The loaded solution was then injected into the matrix, and time-lapse imaging was continued for another 2–4 hr. Typical settings for microinjection were as follows: Compensation pressure = 25hPa, Injection pressure = 100hPa, Injection time = 0.3 sec.

Image Processing and Analysis

Image processing was performed using MetaMorph. ECM deformation was quantified by measuring the x,y coordinates of landmarks in DIC images using the “measure pixel” feature in MetaMorph. In order to display the ECM displacements, a custom-written Visual Basic (Microsoft, Redmond, WA) program was used. The program generated cross-marks and tracks corresponding with the start-points and displacements, respectively, of ECM landmarks from the measured x,y coordinates. Cell area was measured by manually outlining the cell border from the DIC 3-D datasets of the time point from the time point of interest, using the “region overlay” tool in MetaMorph. Since cells often had processes extending above or below the primary focal plane, the z-plane within the stack was changed as necessary during the procedure to ensure that the entire projected cell area was outlined.

Finite Element Modeling (FEM)

FEM was used to visualize and quantify the pattern of matrix deformation due to needle pushing. Finite elements models were created using ANSYS engineering analysis software (Release 10.0, ANSYS Inc., Canonsburg, PA), as previously described by us (Roy et al.,1999b; Vishwanath et al.,2003). Briefly, nodes were defined at coordinates coinciding with ECM landmarks from the DIC images prior to ECM micromanipulation. Boundary nodes were placed at the periphery of a 600-μm diameter circular field around this central set of nodes. A two-dimensional plane stress model was created from the nodes using linear elastic triangular elements. For simplicity, the matrix was assumed to be isotropic, with a Young's modulus of 3.89 × 10-10 N/μm2, an effective thickness of 15 μm, and a Poisson's ratio of 0.3 (Roeder et al.,2002; Vishwanath et al.,2003). To generate maps of ECM deformation, the displacements measured from time-lapse recordings were applied to the corresponding nodes in the model. The resulting strains induced on the matrix were calculated and displayed.

For FEM analysis of matrix distortion, linear elastic, isotropic material properties were used. Our previous matrix calibration experiments suggest that this is a reasonable assumption when studying the normal cellular pattern of force generation (Roy et al.,1997), but it should be noted that large pushes or pulls with the microneedle may induce stresses that are outside the linear response range for the collagen matrix. For this reason, FEM was used to generate strain maps (to visualize the pattern of ECM deformation), but was not used to estimate forces.

Acknowledgements

This study was supported in part by a Senior Scientific Investigator award (W.M.P.) and unrestricted grant from Research to Prevent Blindness, Inc.

Ancillary