Journal of Applied Physiology Fuel your research with LabChart
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

J Appl Physiol 98: 1909-1921, 2005. First published December 23, 2004; doi:10.1152/japplphysiol.01137.2004
8750-7587/05 $8.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow All Versions of this Article:
98/5/1909    most recent
01137.2004v1
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (12)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Pizzo, A. M.
Right arrow Articles by Voytik-Harbin, S. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Pizzo, A. M.
Right arrow Articles by Voytik-Harbin, S. L.

HIGHLIGHTED TOPICS
Biomechanics and Mechanotransduction in Cells and Tissues

Extracellular matrix (ECM) microstructural composition regulates local cell-ECM biomechanics and fundamental fibroblast behavior: a multidimensional perspective

A. M. Pizzo,1 K. Kokini,1,2 L. C. Vaughn,4 B. Z. Waisner,3 and S. L. Voytik-Harbin2,3

1School of Mechanical Engineering, 2Weldon School of Biomedical Engineering, and 3Department of Basic Medical Sciences, School of Veterinary Medicine, Purdue University, West Lafayette, Indiana; and 4Department of Mechanical Engineering, University of Alabama, Tuscaloosa, Alabama

Submitted 11 October 2004 ; accepted in final form 21 December 2004


   ABSTRACT
TOP
 ABSTRACT
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
The extracellular matrix (ECM) provides the principal means by which mechanical information is communicated between tissue and cellular levels of function. These mechanical signals play a central role in controlling cell fate and establishing tissue structure and function. However, little is known regarding the mechanisms by which specific structural and mechanical properties of the ECM influence its interaction with cells, especially within a tissuelike context. This lack of knowledge precludes formulation of biomimetic microenvironments for effective tissue repair and replacement. The present study determined the role of collagen fibril density in regulating local cell-ECM biomechanics and fundamental fibroblast behavior. The model system consisted of fibroblasts seeded within collagen ECMs with controlled microstructure. Confocal microscopy was used to collect multidimensional images of both ECM microstructure and specific cellular characteristics. From these images temporal changes in three-dimensional cell morphology, time- and space-dependent changes in the three-dimensional local strain state of a cell and its ECM, and spatial distribution of {beta}1-integrin were quantified. Results showed that fibroblasts grown within high-fibril-density ECMs had decreased length-to-height ratios, increased surface areas, and a greater number of projections. Furthermore, fibroblasts within low-fibril-density ECMs reorganized their ECM to a greater extent, and it appeared that {beta}1-integrin localization was related to local strain and ECM remodeling events. Finally, fibroblast proliferation was enhanced in low-fibril-density ECMs. Collectively, these results are significant because they provide new insight into how specific physical properties of a cell’s ECM microenvironment contribute to tissue remodeling events in vivo and to the design and engineering of functional tissue replacements.

three-dimensional local strain; fibroblast three-dimensional morphology; confocal reflection microscopy; incremental digital volume correlation; collagen fibril density

MECHANICAL LOADS HAVE BEEN identified as critical determinants of cell behavior not only in vivo during physiological and pathological processes but also in vitro for engineering functional tissue constructs (46). These mechanical signals may be evoked via the contractile machinery of the resident cells or by a variety of environmental factors (5). The process by which these physical forces are converted into biochemical signals and integrated into cellular responses is referred to as mechanotransduction (31). A critical component of the mechanotransduction process is the extracellular matrix (ECM) and its interface with resident cells. The ECM interacts with cells to provide relevant microenvironmental information, biochemically through soluble and insoluble mediators and physically through imposition of structural and mechanical constraints. To date, significant advances have been made in our understanding of how specific molecules of the ECM affect fundamental cellular responses. However, less is known regarding the mechanisms by which specific structural and mechanical properties of the ECM microenvironment influence cell behavior, especially within a three-dimensional (3D) tissuelike context. Such fundamental information is critical to the field of tissue engineering, where one strategy to improve and/or accelerate tissue repair and replacement involves formulation of engineered extracellular microenvironments that would deliver specific growth-inductive signals.

Because biophysical cues such as those originating from ECM microstructure and mechanical properties are considered among the major signaling sources regulating growth and differentiation of cells (6), significant efforts have been made to identify, define, and prioritize specific signaling pathways and mechanisms involved in the mechanotransduction process. Unfortunately, this has been a difficult task, especially because events associated with mechanotransduction span size scales extending from the tissue to the molecular level. Furthermore, the ECM is extremely complex in its 3D molecular composition and organization and there exists a tremendous interdependence of ECM compositional, structural, and mechanical properties. Traditionally, two-dimensional (2D) model systems in which cells are seeded on the surface of a silicone membrane (28), polyacrylamide gel (48), or patterned elastomer (2) coated with individual ECM molecules have been used to explore how compositional and physical aspects of the extracellular microenvironment affect cell behavior. Such studies have shown modulation of fibroblast contractility, adhesion formation, migration, and other fundamental cell activities through the variation of ligand density (24, 40), substrate stiffness (39, 48), and application of external mechanical loads (1, 10, 11). However, the physiological relevance of such cell-ECM signaling as it occurs within a context of 3D structural-mechanical complexity is just being realized. Specifically, several investigators have utilized in vitro systems consisting of connective tissue cells (e.g., fibroblasts) seeded within 3D collagen matrices (4, 22) to study cell-ECM biomechanics and the events associated with mechanotransduction. Such a system has been instrumental in demonstrating that the mechanical environment of fibroblasts affects a number of cellular attributes, including morphology (21), proliferation (45), apoptosis (58), cyclic AMP signaling (30), and the expression of gene products (15, 33, 38, 44, 50, 58). However, the majority of such studies address the cell response from a population rather than an individual cell perspective. Furthermore, studies performed to date lack either quantitative information regarding specific structural and mechanical properties of the ECM scaffold at the microscopic level or a description of how loads are transferred from the global to the local level in 3D. Such information is imperative because the magnitude of the load at the tissue construct (global) level is different from that experienced by the cell and component collagen fibrils (local level) (9) and the differential load distribution is highly dependent on the material properties of the scaffold (43).

The objective of the present study was to determine whether specific microstructural features of the ECM could regulate fundamental cellular behavior. This topic was approached from a multidimensional perspective and at a cellular (local) level of function. Specifically, the working hypothesis tested was that morphometric, ECM-remodeling, and proliferative properties of fibroblasts are dependent on the fibril density of a 3D collagen ECM. Fibroblasts were provided a tissuelike context by seeding them within 3D collagen ECMs in which the microstructural organization of collagen fibrils was systematically varied and quantified. Previous studies by our laboratory have shown that collagen ECMs polymerized at increasing collagen concentrations (e.g., 0.5 to 3 mg/ml) possess increased fibril densities but no significant difference in fibril width (51). The model system also involved a relatively low cell density such that the mechanical properties of the tissue construct were primarily determined by the ECM component, with the contribution from cells being negligible (62). Confocal microscopy was used in a reflection or a combined reflection-epifluorescence mode for visualization (up to 4 dimensions: x, y, z, and time) of the dynamic events associated with cell-ECM interactions over the first 12 h after seeding. These images provided the basis for a rigorous 3D morphometric analysis. In addition, ECM remodeling was quantified on an individual cell basis by using an incremental digital volume correlation (IDVC) algorithm (52) to determine the time- and space-dependent changes in the 3D strain state of the cell and its associated ECM components. To show that the ECM is instrumental in regulating not only early events of cell-ECM interaction but also longer term cellular responses that characterize phenotype and function, the effect of ECM microstructure on fibroblast proliferation at 24- and 48-h time points was also determined. Results from these studies provide new fundamental information regarding how the ECM participates in the transfer of mechanical signals to and from resident cells and how specific biophysical features can be used to control cell fate. Such insight contributes to the further elucidation of the mechanotransduction process as well as to the design and development of biomimetic environments for tissue engineering applications.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
Cell culture.   Primary neonatal human dermal fibroblasts (NHDFs) were obtained from Clonetics (San Diego, CA) and propagated in fibroblast basal medium supplemented with 1 µg/ml human recombinant fibroblast growth factor, 5 mg/ml insulin, 50 mg/ml gentamicin, 50 mg/ml amphotericin B, and 2% FBS. Cells were grown and maintained in a humidified atmosphere of 5% CO2 at 37°C. For most studies, parallel experiments were conducted on NHDFs and on Swiss mouse 3T3 fibroblasts obtained from American Type Culture Collection (Manassas, VA). 3T3 fibroblasts were propagated in DMEM with 1.5 g/l NaHCO3, 10% calf serum, 100 U/ml penicillin, 100 µg/ml streptomycin, and 2 mM L-glutamine. For both types of fibroblasts, subcultures were established every 2–3 days to prevent cells from exceeding 75% confluence. Cells from 20 or fewer passages were used for all experiments.

Preparation of engineered 3D tissue constructs.   3D collagen ECMs were prepared by dissolving native, acid-solubilized type I collagen from calf skin (Sigma Chemical, St. Louis, MO) in 0.01 M hydrochloric acid to achieve desired collagen concentrations. For sterile preparations of collagen, the collagen solution was layered onto a volume of chloroform. After incubation for 18 h at 4°C, the collagen solution layer was carefully removed so as not to include the collagen-chloroform interface layer.

To produce 3D collagen ECMs that varied in fibril density while maintaining a constant fibril diameter, native type I collagen was polymerized under different conditions as previously described (51). Specifically, collagen solutions were polymerized at final collagen concentrations of 0.5 to 3.0 mg/ml. The polymerization buffer consisted of 10x PBS with an ionic strength of 0.14 M and a pH of 7.4. Fibroblasts were harvested in complete medium, collected by centrifugation, and added as the last component before polymerization. Tissue constructs were prepared at a relatively low cell density of 5 x 104 cells/ml. Previous studies by our laboratory have shown that this cell density is suitable for maintaining cell viability, minimizing cell-cell interaction, and allowing the study of the dynamic relationship between an individual cell and its surrounding ECM. Polymerization of tissue constructs was conducted in four-well Lab-Tek coverglass chambers (Nalge Nunc International, Rochester, NY) maintained in a humidified environment at 37°C. Immediately after polymerization (20 min or less), complete medium was added and the tissue constructs were cultured at 37°C in a humidified environment consisting of 5% CO2 in air.

Multidimensional imaging of tissue constructs.   All multidimensional imaging was performed on a Bio-Rad Radiance 2100 MP Rainbow (Bio-Rad, Hemel Hempstead, UK) multiphoton/confocal system adapted to a TE2000 (Nikon, Tokyo, Japan) inverted microscope. The system has a Mai-Tai (Spectra-Physics, Mountain View, CA) tunable red/infrared pulse laser as well as four-line argon, green helium-neon, and red diode lasers. The system is also equipped with a motorized XY Scan IM stage (Märzhäuser, Wetzlar, Germany). For live-cell experiments, a heating/cooling microincubator stage (ALA Scientific Instruments, Westbury, NY) with a Peltier temperature control system (npi electronic, Tamm, Germany) was used to maintain temperature at 37°C. A custom-designed environmental chamber was adapted to the microscope to provide tissue constructs with a sterile environment of 5% CO2 in humidified air. CO2 levels were monitored and adjusted through a feedback controller (Digital Control Systems, Portland, OR) and regulated with a solenoid valve (KIP, Farmington, CT).

Time-lapse imaging of cell-ECM interactions.   Tissue constructs representing NHDFs seeded at 5 x 104 cells/ml within collagen ECMs with defined microstructural compositions were evaluated by use of time-lapse confocal microscopy. Beginning 1 h after polymerization, two to three cells were repeatedly monitored using the confocal microscope in a reflection (backscattered light) mode to obtain image stacks of the individual cell and its surrounding matrix as described previously (7, 60, 61). Images were collected at 30-min intervals and a z-step of 0.5 µm to minimize exposure of the tissue constructs to radiation from the argon laser. This time-lapse imaging procedure was performed on a total of six cells for each of the ECM microstructural compositions studied.

Determination of volumetric strain.   Consecutive confocal reflection images representing temporal deformation induced by a resident cell on its surrounding ECM microstructure provided the basis for the quantification of local displacements and strains in 3D. Within each image, subvolumes of 32 x 32 x 20 pixels in the x, y, and z directions, respectively, were established. Each subvolume represented a group of voxels centered around a given point at which displacement values were sought. Each image subvolume provided a unique 3D voxel intensity pattern that allowed correlation pattern matching between consecutive images using an IDVC algorithm developed previously by our laboratory (52). The IDVC algorithm provided strain-state data, including principal strains and their associated directions, for all grid point locations. Grid points were established in 512 x 512-pixel images that were 32 pixels apart in both x- and y-directions, with 24-pixel spacing in the z-direction. Principal strains determined for the length (EL), width (EW), and height (EH) directions were used to calculate volumetric strain (EV) on the basis of the following formula (41):

Qualitative and quantitative determination of 3D cell morphology.   Before imaging at either 6 or 12 h after construct polymerization, tissue constructs were stained with the vital dye Cell Tracker Green (Molecular Probes, Eugene, OR) to facilitate discrimination of the cell from the surrounding collagen ECM. Confocal image stacks were then collected in a combined reflection-epifluorescence mode for determination of cell morphology and ECM microstructural organization. Morphological evaluation was conducted on a total of 12–20 cells for each of the ECM microstructural compositions studied. Each of 512 x 512-pixel confocal images was first cropped in the x-, y-, and z-dimensions to define a region of interest containing an individual cell. Then a threshold pixel-intensity value was chosen objectively on the basis of the number of z-sections detected over a range of threshold values. The threshold value was then used by the Sobel image-gradient edge-detection algorithm in MATLAB (The MathWorks, Natick, MA) to identify the cell boundary and convert the pixel intensity map to a binary image. Cell volume (V; in µm3) was calculated on the basis of the Cavalieri estimator

where T is the distance between confocal z-sections (z-step), Ai is the cross-sectional area of the cell within the ith z-section, and n is the total number of z-sections (36). 3D cell surface area, As, was found by using the formula

where A1 and An are the areas of the top and bottom sections of the cell, respectively (55). The perimeter of the cell within the jth z-section (Pj) is summed over the entire z-range. As in the equation for cell volume, T is the distance between confocal z-sections and n is the total number of z-sections.

The processed image stack was also used to determine fundamental morphological parameters, including length, width, and height. Because each cell had a relatively unique orientation within the 3D matrix, these morphological parameters were defined on the basis of the cellular coordinate system as shown in Fig. 1. This method accounts for the orientation of the cell in 3D space as opposed to only considering the cell bounding box in the xyz-coordinate (confocal image) system. First the longest axis of the cell was found in 3D by determining the two points on the cell surface that were the farthest distance apart. The distance between these two points defined the cell length. Then, vector projections using the equation for the dot product were used to determine the perpendicular width dimension relative to the length vector of the cell. Once two perpendicular dimensions were determined (length and width), the cross product of these two vectors was used to find the orthogonal basis for the height measurement.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 1. Schematic demonstrating fundamental morphometric parameters (length, width, and height) that were defined on the basis of an assigned cellular coordinate system and used to describe and compare 3-dimensional (3D) fibroblast morphology.

 
The number of cell projections was determined visually by using 3D rotations in the Confocal Assistant (Bio-Rad) software application. A projection was counted if it showed any protrusion from the main cell body that was at least 1 µm wide and 5 µm long. The dimensions of the cell projections were determined by using the measurement features in Imaris 4.0 (Bitplane, Saint Paul, MN).

Immunolocalization of {beta}1-integrin and cell-matrix adhesions.   A monoclonal antibody specific for the {beta}1-integrin subunit (MAB1977; Chemicon, Temecula, CA) was first labeled directly with Alexa Fluor 488 (Molecular Probes, Eugene, OR). Immediately after polymerization, tissue constructs were incubated in complete medium at 37°C in a humidified environment of 5% CO2 in air for 2, 4, 12, or 24 h. Samples then were fixed in a solution containing 4% formaldehyde and 5% sucrose in PBS, pH 7.4. After blocking for nonspecific binding with 1% BSA in PBS, specimens were incubated with the Alexa Fluor 488-conjugated {beta}1-integrin antibody for 2 h at 37°C. 3D images representing {beta}1-integrin and the ECM microstructure were obtained by using confocal microscopy in a combination reflection-epifluorescence mode.

Qualitative and quantitative determination of cell proliferation.   Cell proliferation experiments involved seeding fibroblasts in tissue constructs prepared in a 24-well tissue-culture plate. For comparison purposes, an equivalent number of fibroblasts were seeded directly on the plastic surface of the plate. After 24 or 48 h, each well and tissue construct was examined microscopically to observe the viability, number, and morphology of the cells. The medium from each well then was replaced with fresh medium containing 10% (vol/vol) alamarBlue, a metabolic indicator dye. Dye reduction was monitored spectrofluorometrically 24 h later using a FluoroCount Microplate Fluorometer (Packard Instruments, Meriden, CT) with excitation and emission wavelengths of 560 and 590 nm, respectively. Background fluorescence measurements were determined from wells containing only dye reagent in culture medium. Maximum levels of relative fluorescence were determined from alamarBlue solutions that were autoclaved to induce complete dye reduction. The mean and the standard deviation for all fluorescence measurements were calculated and subsequently normalized with respect to the background and maximum fluorescence readings. All experiments were performed in triplicate and repeated at least three times.

Statistical analysis.   When relevant, statistical analyses were performed using MATLAB and included an ANOVA. The Tukey-Kramer method for multiple comparisons (P < 0.05) was then applied.


   RESULTS
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
ECM microstructural composition affects the 3D morphology of fibroblasts.   To determine how fibril density of a 3D collagen ECM affected fibroblast morphology, time-lapse confocal reflection microscopy was used to follow the progressive changes in individual cells as they interacted with their surrounding matrix for a time period representing 1–5 h after polymerization. In addition, a rigorous 3D morphometric analysis was conducted on cells at time points representing 6 and 12 h postpolymerization. Because each cell had a relatively unique orientation within its 3D ECM, morphometric parameters were defined on the basis of a cellular coordinate system as shown in Fig. 1.

Upon seeding within 3D collagen ECMs, fibroblasts generally adapted from an initial rounded shape to a more stellate or bipolar spindle shape. Interestingly, the observed time course of events associated with these morphological changes depended on the collagen fibril density of the surrounding ECM (Fig. 2). Fibroblasts grown within 3D ECMs characterized by a low fibril density probed their extracellular microenvironment by rapidly developing a number of cytoplasmic projections. In some cases, these projections were evident within the first hour after seeding. Fibroblasts grown within high-fibril-density matrices took longer to develop projections, but these cells consistently demonstrated a significantly greater number of projections at 6- and 12-h time points (Fig. 4C). In addition, cell projections within high-fibril-density matrices were typically longer and thinner than those within low-fibril-density matrices. Regardless of fibril density, the projections decreased in number but increased in size over time as some retracted and others stabilized.



View larger version (98K):
[in this window]
[in a new window]
  		Fig. 2. Primary human dermal fibroblasts develop distinct morphologies when grown in extracellular matrices (ECMs) that differ in collagen fibril density. 2D projections of confocal image stacks of neonatal human dermal fibroblasts (NHDFs) grown within collagen ECMs at 1 mg/ml collagen concentration (low fibril density) 6 h (A) and 12 h (B) postpolymerization and within ECMs at 3 mg/ml collagen concentration (high fibril density) 6 h (C) and 12 h (D) postpolymerization. Cells were stained with the vital fluorescent dye Cell Tracker Green (Molecular Probes). Scale bar represents 20 µm.

 


View larger version (18K):
[in this window]
[in a new window]
  		Fig. 4. Primary human dermal fibroblasts develop a greater surface area and more cytoplasmic projections when grown in collagen ECMs of high fibril density. Cell volume (A), surface area (B), and number of cytoplasmic projections (C) were quantified for NHDFs cultured for 6 and 12 h within collagen ECMs at 1 mg/ml and 3 mg/ml collagen concentrations. Results represent the means and standard deviations for 11 ≤ n ≤ 20 cells analyzed for each ECM microstructural composition at a given time point. All relationships showing statistically significant differences (P < 0.05) are indicated with an asterisk (*).

 
That this transformation of fibroblast shape was dependent on ECM microstructure was also evident on comparison of other morphometric parameters determined for 6- and 12-h time points (Fig. 3). Although fibroblast length did not seem to be significantly affected by collagen fibril density, fibroblast length increased significantly between 6 and 12 h for both ECM microstructures (Fig. 3A). In contrast, fibroblast width and height were affected by the fibril density of the collagen ECM. Specifically, fibroblasts grown in low-fibril-density matrices were significantly decreased in their width and height dimensions as shown in Fig. 3, B and C. In addition, comparison of cellular aspect ratios (Fig. 3, D–F) showed that fibroblasts grown in low-fibril-density matrices were more spindle or bipolar in shape than those grown in high-fibril-density matrices. This was evident from the fact that fibroblasts grown in low- vs. high-density matrices were characterized by increased length/width (Fig. 3D) and length/height (Fig. 3F) parameters. The morphological features, together with the decreased number of cytoplasmic projections, exhibited by fibroblasts within low-fibril-density ECMs provided these cells with a significantly decreased 3D surface area (Fig. 4B). Interestingly, as indicated in Fig. 4A, no significant effect of ECM microstructure was noted on cell volume. In summary, results showed that fibroblast morphology is dependent on the microstructural composition of the surrounding 3D ECM.



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 3. Primary human dermal fibroblasts show distinct 3D morphometric features when grown in ECMs that differ in collagen fibril density. Morphometric parameters including length (A), width (B), height (C), and aspect ratios [length/width (D), width/height (E), and length/height (F)] were determined and compared for NHDFs cultured for 6 and 12 h within collagen ECMs at 1 mg/ml and 3 mg/ml collagen concentrations. Results represent the means and standard deviations for 12 ≤ n ≤ 20 cells analyzed for each ECM microstructural composition at a given time point. All relationships showing statistically significant differences (P < 0.05) are indicated with an asterisk (*).

 
ECM remodeling and cell-matrix adhesions are dependent on growth format and ECM microstructure.   Repeated monitoring of interactions between an individual cell and neighboring collagen fibrils within a single living construct by confocal reflection microscopy also provided a means of quantifying and comparing the ECM remodeling properties of cells grown within collagen ECMs of different fibril density. An IDVC algorithm developed by our laboratory (52) was applied to consecutive 3D image stacks to determine 3D displacements and strains as they occurred locally to a given cell and component collagen fibrils of its ECM over the first 5 h after polymerization. The bottom panels of Fig. 5, A and B, represent 2D projections of 3D time-lapse confocal image stacks for representative fibroblasts grown within high- and low-fibril-density matrices, respectively. Time-dependent changes in the associated calculated volumetric strain are provided in the top panels of Fig. 5, A and B. Movies demonstrating these cell-ECM interactions qualitatively and quantitatively over time can be found at http://www.cyto.purdue.edu/jappphys/videos. From these images, it is evident that although fibroblasts in both low- and high-fibril-density ECMs attach to and reorganize nearby collagen fibrils, the extent of remodeling is dependent on the original ECM microstructure. Within 2–3 h after seeding within a low-fibril-density ECM, a condensation or increase in the collagen fibril density immediately surrounding the fibroblast was noted. In many cases, condensation events appeared to correlate with sites of cytoplasmic projection development. As tensional forces exerted by the fibroblast increased with time, regions within the matrix more distant from the cell experienced significant strain and showed preferential alignment of collagen fibrils. At 5 h after seeding, resident fibroblasts within these low-fibril-density matrices were able to generate volumetric strain levels of –0.2 at distances up to 60 µm from the center of the cell. On the other hand, high-fibril-density ECMs were better able to resist the contractile forces and matrix remodeling of fibroblasts. Although fibroblasts within high-fibril-density matrices developed, in general, a significantly greater number of long, thin projections, they were not able to induce a high level of reorganization in the relatively stiff surrounding matrix (Fig. 5B). In fact, 5 h after seeding, maximum volumetric strain levels of only –0.1 were observed, and this low-level deformation occurred in relative proximity to the cell (<30 µm from the center of the cell). On the basis of these results, it was apparent that ECM microstructure plays an important role in regulating matrix remodeling events of fibroblasts.



View larger version (118K):
[in this window]
[in a new window]
  		Fig. 5. Primary human dermal fibroblasts grown within low-fibril-density ECMs remodel and reorganize surrounding collagen fibrils to a greater extent. Matrix remodeling by individual NHDFs resident within collagen ECMs at 1 mg/ml (A) and 3 mg/ml collagen concentrations (B) was determined both qualitatively and quantitatively. Bottom panels represent 2-dimensional (2D) projections of confocal reflection image stacks showing changes to NHDF morphology and matrix microstructure observed at designated time points. Top panels show corresponding quantified levels of local volumetric strain. Refer to http://www.cyto.purdue.edu/jappphys/videos for time-lapse movies (Video 1, low fibril density and Video 2, high fibril density) demonstrating these dynamic interactions.

 
One way that cells interact with their ECM is through assembly of cell-matrix adhesions. The assembly process involves the initial binding of integrins (e.g., {alpha}2{beta}1-integrin) to matrix protein ligands (e.g., collagen) followed by colocalization of a complex array of cytoplasmic proteins, catalytic signaling molecules, and membrane-associated molecules that together ultimately provide extracellular linkage to major cytoskeletal proteins such as actin. On the basis of the observed temporal and spatial differences in matrix modeling by fibroblasts and the fact that such events were dependent on ECM microstructure, we speculated that {beta}1-integrin subunit distribution would be distinct for these constructs. Therefore, {beta}1-integrin was localized in 3D for fibroblasts grown within collagen ECMs with defined microstructural compositions at time points up to 24 h. For comparison purposes, {beta}1-integrin localization was also performed on fibroblasts grown in a traditional 2D format, on the surface of collagen ECMs. Confocal microscopy in a combined reflection-epifluorescence mode provided a means to relate {beta}1-integrin localization with ECM microstructure and remodeling events. Initial comparison of fibroblasts grown on the surface of vs. within a collagen ECM (2D vs. 3D) showed a dramatically different 3D distribution of {beta}1-integrin. The dependence of {beta}1-integrin localization on the cell growth format was best realized by calculating and plotting the surface area occupied by {beta}1-integrin as a function of z-section (Fig. 6). For cells grown in 2D, {beta}1-integrin was polarized to the lower surface of the fibroblast that was in intimate contact with matrix. In this case, the {beta}1-integrin pattern was characterized by a high density of punctate staining regions that were relatively uniform in length, as shown previously by others (17). In contrast, fibroblasts grown within a 3D collagen ECM demonstrated a {beta}1-integrin pattern that was decreased in frequency and distributed across the 3D surface area of the cell. Evaluation of {beta}1-integrin distribution for fibroblasts seeded within collagen ECMs over time showed that at early time points (2 h) {beta}1-integrin was diffusely distributed across the cell surface, for cells seeded within both high- and low-fibril-density matrices (Fig. 7). At 4 h and time points thereafter, {beta}1-integrin seemed to be preferentially distributed along cytoplasmic projections rather than over the body of the cell. Interestingly, the location and size of regions occupied by {beta}1-integrin appeared to be related to matrix-remodeling events. More specifically, larger areas of {beta}1-integrin were repeatedly observed along the length of long, thin projections immediately adjacent to regions of local matrix condensation. In contrast, multiple small punctate staining regions were noted at the ends of cytoplasmic projections that were associated with significant matrix deformation.



View larger version (27K):
[in this window]
[in a new window]
  		Fig. 6. 3D spatial distribution of cell-matrix adhesions within 3T3 fibroblasts as indicated by localization of {beta}1-integrin is different for 2D and 3D growth formats. Histograms represent {beta}1-integrin-stained area as a function of z-section for a representative 3T3 fibroblast grown on the surface of (A) or within (B) a collagen ECM for 24 h.

 


View larger version (81K):
[in this window]
[in a new window]
  		Fig. 7. Two-dimensional projections of confocal image stacks demonstrate temporal changes in {beta}1-integrin distribution for 3T3 fibroblasts grown within collagen ECMs at 1 mg/ml (A) and 3 mg/ml collagen concentrations (B) at specified time points. Top panels show spatial distribution of {beta}1-integrin and represent entire confocal image stack (26–76 optical sections or 16–45 µm thickness). Bottom panels show corresponding 3T3 fibroblast morphology and ECM microstructure and represent a 6-µm-thick portion (10 optical sections) of the confocal image stack. 3D blind deconvolution was performed on the {beta}1-integrin images by use of AutoDeblur and AutoVisualize (AutoQuant, Watervliet, NY).

 
Proliferative properties of cells are regulated by ECM microstructure.   To determine the effect of collagen fibril density on cell proliferation, NHDFs were seeded within ECMs prepared at collagen concentrations ranging from 1 to 3 mg/ml. For comparison purposes, fibroblasts were also seeded and cultured in parallel on tissue culture plastic. Proliferation of the fibroblasts within the 3D ECM constructs was determined at 24- and 48-h time points by using the metabolic (redox) dye alamarBlue. This specific dye is reduced by the biochemical reactions of normal cellular metabolism and provides an indirect measure of live fibroblast number (59). In addition, qualitative evaluations of the cells cultured within ECMs were conducted to verify that the amount of dye reduced correlated with the relative number of live cells.

Results from this study demonstrated that, in general, cells grown in a 3D format within collagen ECM constructs proliferated at rates that were significantly decreased compared with those grown in a 2D format on tissue culture plastic. Interestingly, it was also noted that the proliferation of fibroblasts within collagen ECMs prepared at lower collagen concentrations (1 mg/ml) was significantly greater compared with those grown within ECMs prepared at higher collagen concentrations (3 mg/ml; Fig. 8). That is, systematically decreasing the collagen fibril density resulted in enhanced cell proliferative properties. Although the total number of cells within both ECM microstructures increased between 24 and 48 h, the total number of fibroblasts was greatest in the low-fibril-density matrices at both time points.



View larger version (15K):
[in this window]
[in a new window]
  		Fig. 8. Primary human dermal fibroblasts proliferate to a greater extent when seeded within low-fibril-density ECMs. NHDFs were seeded on tissue-culture plastic and within collagen ECMs characterized by a collagen concentration of 1 and 3 mg/ml. Relative proliferation was quantified indirectly by measuring the reduction of a metabolic indicator dye 24 and 48 h after polymerization. All relationships showing statistically significant differences (P < 0.05) are indicated with an asterisk (*).

 

   DISCUSSION
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
The ECM has long been appreciated for its role in providing structural support and organization for resident cells as well as endowing tissues with specific mechanical properties. However, the results presented herein collectively show that the ECM component of tissues has a much more sophisticated role in determining cell fate and therefore the structure and function of tissues. In the present study, the major structural-mechanical element of the ECM, type I collagen fibrils, was systematically varied to determine how the 3D microstructural composition, specifically fibril density, affected cell-ECM biomechanics and fundamental fibroblast behavior. A unique aspect of this study was that an individual cell and the component collagen fibrils of its ECM microenvironment were simultaneously visualized in three (x, y, and z) or four dimensions (x, y, z, and time) as they underwent dynamic interaction. This qualitative information, in turn, provided the basis for rigorous quantification of 1) 3D cell morphometric parameters; 2) time- and space-dependent changes in the 3D strain state of a cell and its associated matrix components; and 3) 3D spatial distribution of molecules (e.g., {beta}1-integrin) involved in the formation of cell-matrix adhesions.

It is well documented that, compared with fibroblasts grown in a 2D format, those grown within 3D collagen matrices develop a morphology that is more reminiscent of that observed in vivo (4, 22, 56). Furthermore, on the basis of qualitative and 2D microscopic analyses, fibroblast morphology and orientation within collagen matrices have been shown to be dependent on the mechanical loading state (21), the cell density (54), and the milieu of soluble factors (26). Herein, we show for the first time that fibroblast morphology is also responsive to systematic variation of the 3D collagen fibril density (0.5–3.0 mg/ml collagen concentration). Recently, Loftis et al. (40) reported no qualitative morphological differences for fibroblasts grown within free-floating collagen matrices varying in collagen concentration (0.75–1.25 mg/ml). However, tissue constructs for the present studies were restrained (mechanically loaded) rather than free-floating (mechanically unloaded).

Quantification of a number of 3D morphometric parameters for cells grown within 3D ECMs with different microstructures allowed meaningful statistical comparison of results. From a cell biomechanics standpoint, it was interesting to note that projections developed by fibroblasts within high-fibril-density matrices were greater in number and longer and thinner than those observed in low-fibril-density matrices. The number of cell projections decreased with time regardless of the fibril density. However, even 12 h after seeding the number of cytoplasmic projections for fibroblasts grown in high-fibril-density ECMs was double that for those grown in low-fibril-density ECMs. The observed retraction of some projections along with the stabilization of others over time was consistent with previous reports by others (26, 56). This network of projections, also referred to as "dendritic network," was less pronounced or even absent in fibroblasts grown on 2D collagen-coated surfaces (26) and thus appeared to provide cells with a mechanism to explore and establish a mechanical equilibrium ("tensional homeostasis") with its local 3D ECM microenvironment (8).

Determination and comparison of the morphometric parameters (including length, width, height, volume, 3D surface area, and aspect ratios) also showed that fibroblasts grown within low-fibril-density ECMs took on a bipolar spindle shape. In a high-fibril-density matrix, fibroblasts maintained a more stellate morphology characterized by increased width, height, and 3D surface area. Even after 12 h of culture, fibroblasts within high-fibril-density matrices exhibited length-height and length-width ratios that were significantly less than those for fibroblasts grown in low-fibril-density matrices. The observed differences in shape owing to ECM microstructure are important because cellular phenotype and function have traditionally been linked to cell form or shape. As expected, because variation of the 3D ECM microstructure did not affect the osmotic properties of the microenvironment, cell volume was similar for fibroblasts grown within ECMs of different collagen fibril density.

A critical determinant of morphology and ultimately of the overall phenotype and function expressed by a fibroblast is the cell’s ability to attach to and deform the surrounding ECM components. This cell-induced remodeling process has been said to resemble matrix morphogenesis during development and wound repair (4, 25, 29, 57). The fundamental ability of fibroblasts to generate tensional forces has been documented at a global or tissue level by monitoring progressive contraction of free-floating fibroblast-populated collagen matrices (4) or by measuring the contractile forces of tissue constructs using direct force measurement instruments (19, 20, 35). More recently, matrix deformation and contractile forces by fibroblasts within 3D collagen matrices have been quantified at a cellular or local level. Such studies involved monitoring the displacement of collagen embedded beads (54) or collagen fibrils imaged using high-magnification differential interference contrast microscopy (49). These approaches have provided insightful information regarding the effect of cell density and soluble factors on matrix remodeling by fibroblasts (54) as well as relating cell-matrix adhesion movement with local force generation (49). However, local deformation and strain data have been limited to a 2D (x-y) perspective. In the present study, the contractile properties and matrix deformation exhibited locally by a fibroblast on the surrounding ECM were monitored and quantified from a multidimensional perspective and the effect of ECM microstructural composition on this process was determined. It was evident, both visually and quantitatively, that the extent of matrix remodeling and reorganization induced by an individual fibroblast was dependent on the microstructural composition of the ECM, more specifically the collagen fibril density. For fibroblasts seeded within a collagen ECM of low fibril density, deformation of the surrounding matrix first occurred in regions adjacent to the cell. As early as 2.5 h after polymerization, volumetric strain levels of –0.1 were observed adjacent to the cell membrane as the cell condensed or increased the local density of collagen fibrils. As time progressed, these fibroblasts induced significant strain within the collagen fibrils resulting in fibril alignment at sites more distal to the cell. The amount of local ECM deformation and remodeling induced by fibroblasts was significantly less for fibroblasts seeded within high-fibril-density ECMs. Furthermore, matrix remodeling events were limited to fibril condensation immediately surrounding the cell. The fact that matrix remodeling by fibroblasts is dependent on the ECM microstructure can be attributed in part to differences in mechanical properties of these ECM constructs (e.g., stiffness or linear modulus). Traditional analyses of "engineering" stress-strain properties have shown that collagen ECMs prepared at collagen concentration of 1 mg/ml (low fibril density) and 3 mg/ml (high fibril density) have a linear modulus of 10.7 kPA (SD 1.93) and 24.3 kPa (SD 4.16), respectively (51). Recently, more sophisticated analyses of "true" stress-strain behavior have shown that at low strain levels the linear modulus 3 mg/ml collagen ECMs is about four- to sixfold greater than that for 1 mg/ml collagen ECMs (unpublished data). Interestingly, these dynamic interactions between a cell and the ECM microstructure determined at the local level translate well to previous global-level results where free-floating fibroblast-populated collagen ECMs prepared at increasing collagen concentrations contracted progressively less and at decreasing rates (4, 40). The present results also show that ECM remodeling by cells is not a displacement-limited phenomenon. This is consistent with studies reporting that fibroblasts contract the substrate to attain a particular value of force per cell rather than a particular value of displacement per cell (23). Taken together, these results emphasize that ECM microstructure is a critical determinant of how contractile forces are exerted by fibroblasts and how the resulting mechanical loads are distributed locally.

An important aspect of cell-ECM interaction and the associated exchange of mechanical loads is the number and quality of cell-matrix adhesions (37). Previous studies involving cells grown on 2D substrates consisting of silicone rubber or polyacrylamide coated with individual ECM molecules (e.g., collagen) have shown that integrin binding and cell-matrix adhesion formation are responsive to physical and mechanical cues (14, 16, 32, 63). However, the recent differences noted in the morphology and distribution of cell-matrix adhesion for cells grown in 2D vs. 3D culture formats emphasize that the true physiological significance of these adhesion structures cannot be realized without pursuing the mechanisms of cell-ECM interaction in a 3D tissuelike context (17, 18, 65). Here, using a 3D type I collagen matrix as a simplified model of the ECM, the morphology and 3D spatial distribution of {beta}1-integrin within fibroblasts grown on the surface and within 3D collagen ECMs were compared. The {beta}1-integrin receptor subunit was chosen because the binding of cells to type I collagen involves mainly {alpha}1{beta}1, {alpha}2{beta}1, and {alpha}3{beta}1 integrins (3, 12, 13, 34, 53, 64). {beta}1-Integrin was found to polarize to the lower surface of the cell, which was in immediate contact with collagen fibrils of the substrate. On the other hand, {beta}1-integrin was widely distributed across the 3D surface area of a fibroblast grown within a 3D collagen ECM. In addition, the frequency and size distribution of the {beta}1-integrin pattern was different for cells cultured under these two conditions. We extended these findings by determining the effect of ECM microstructure on the time- and space-dependent distribution of {beta}1-integrin within fibroblasts. Over the first 12 h, fibroblasts adapted from displaying a diffuse pattern of {beta}1-integrin to one in which {beta}1-integrin was preferentially localized along the length and/or at the ends of cytoplasmic projections. Interestingly, the 3D morphology and 3D spatial distribution of {beta}1-integrin correlated well with the local ECM microstructure and matrix remodeling events. Specifically, multiple small punctate regions of {beta}1-integrin were consistently noted at the ends of stable, mature cytoplasmic projections where significant contraction and matrix remodeling had occurred. In contrast, {beta}1-integrin-labeled regions that were considerably longer in length were repeatedly identified along the length of long, thin cytoplasmic projections. Associated with these regions was a highly localized condensation of collagen fibrils. Recently, Petroll and Ma (49) compared the distribution of green fluorescent protein-zyxin (an established marker for cell-matrix adhesions) and the organization of collagen fibrils as determined by confocal fluorescence microscopy and differential interference contrast imaging, respectively. Interpretation of images obtained in that study provided results that were consistent with the findings of the present study, suggesting that the size and morphology of cell-matrix adhesions may be linked to local strain field and matrix remodeling events. More extensive studies correlating 3D cell-matrix adhesion patterns with 3D ECM microstructure and organization are necessary to determine the significance of these findings. Taken together, the findings of this study continue to support the notion that distribution, morphology, and composition of cell-matrix adhesions provide the cell with the ability to sense and respond to a particular type and magnitude of mechanical load.

Because ECM microstructural composition influenced the initial events of cell-ECM interaction, in particular cell adhesion and morphology, it was hypothesized that the program of fundamental cellular responses, including proliferation, migration, and differentiation, would also be affected. As a first step to test this hypothesis, the proliferative properties of fibroblasts grown on plastic and within collagen ECMs with defined microstructural composition were determined and compared. Consistent with previous reports, the proliferative properties of fibroblasts grown on plastic were markedly greater than those of fibroblasts grown within a 3D collagen matrix (42, 47). However, significant differences were also noted between proliferative properties of fibroblasts within collagen ECMs of low and high fibril density. Fibroblasts seeded within collagen ECMs prepared at increasing collagen concentrations showed a decrease in their proliferative properties. Recall that significant contractile and matrix remodeling properties exhibited by fibroblasts within collagen matrices of low collagen fibril density result in a relatively high level of local strain or mechanical loading. Thus the proliferative properties of fibroblasts appeared to be enhanced by a local increase in mechanical load. Such observations translate well to previous studies showing that fibroblasts within restrained matrices proliferated to a greater extent than those within matrices that were free floating (45). In this case, fibroblasts within restrained constructs experienced an increase in mechanical load as the applied strain was transferred from the global to the local level. The fact that we observed a decrease in fibroblasts proliferation with increasing collagen fibril density is also analogous to the sequence of events associated with wound healing in vivo. During the remodeling phase of wound healing, the provisional ECM, which is characterized by low collagen-fibril density and stiffness, is associated with increased proliferation and matrix deposition by fibroblasts. With time, as collagen concentration and stiffness of the ECM increase, there is an associated decrease in fibroblast number (57).

In summary, these studies provide new insight into how ECM microstructure regulates fundamental fibroblast behavior and tissue remodeling. Such findings are important because they provide new perspective regarding the tissue repair process as it occurs in vivo as well as important design information for engineering tissue constructs in vitro. This critical role played by ECM microstructure is largely owing to that fact that it controls how mechanical loads are transferred between a cell and its microenvironment as part of the mechanotransduction process. Furthermore, the microstructure-dependent differences in {beta}1-integrin distribution and morphology suggest that cell-matrix adhesions provide at least one mechanism by which mechanical loads are differentially transferred between structural elements of the ECM and the cell interior. It should be noted that within tissues in vivo, the ECM represents a complex composite material consisting of an insoluble 3D network of collagen fibrils surrounded by soluble proteins, glycoproteins, proteoglycans, and carbohydrates. In the present work, a simplified model of the ECM was necessary to systematically determine how 3D microstructural organization of component type I collagen fibrils regulated the overall cell response. It is likely that cell-ECM biomechanics and fundamental fibroblast behavior are affected by microenvironmental signals other than ECM microstructural composition, including seeding cell density (cell-cell interactions), concentration of serum or individual growth factors (soluble factors), and composition of the ECM. Studies to determine the interplay between microstructural-mechanical cues provided by the ECM and these other major cell signaling modalities are currently under way.


   GRANTS
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
ACKNOWLEDGMENTS
 REFERENCES
 
This research was funded by the National Institute for Biomedical Imaging and Bioengineering Grant 1R01EB000165. This work was also supported by the National Science Foundation (NSF) Integrative Graduate Education and Research Training (IGERT) Program in Therapeutic and Diagnostic Devices Grant DGE-99-72770. A. M. Pizzo is an NSF/IGERT fellow. Finally, the authors would like to thank the National Science Foundation (0353901-EEC; principal investigators Karen Haberstroh and Tom Webster) for providing support to L. C. Vaughn as part of a Research Experiences for Undergraduates (REU) program at Purdue University.


   ACKNOWLEDGMENTS
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
REFERENCES
 
The authors acknowledge Dr. J. Paul Robinson and Jennifer E. Sturgis for support and assistance with the confocal imaging aspects of this work. All confocal microscopy was conducted within the Purdue University Cytometry Laboratories. We also thank Dr. Blayne A. Roeder for designing and building the custom environmental chamber as well as for providing consultation and suggestions regarding stress-strain analyses and the incremental digital volume correlation algorithm. Finally, the authors are grateful to Gretchen Lawler for helpful editorial comments.


   FOOTNOTES
 

Address for reprint requests and other correspondence: S. L. Voytik-Harbin, Weldon School of Biomedical Engineering/Dept. of Basic Medical Sciences, School of Veterinary Medicine, Lynn Hall, Purdue Univ., 625 Harrison St., West Lafayette, IN 47907-2026 (E-mail:harbins{at}purdue.edu)

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


   REFERENCES
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 

  1. Atance J, Yost MJ, and Carver W. Influence of the extracellular matrix on the regulation of cardiac fibroblast behavior by mechanical stretch. J Cell Physiol 200: 377–386, 2004.[CrossRef][ISI][Medline]
  2. Balaban NQ, Schwarz US, Riveline D, Goichberg P, Tzur G, Sabanay I, Mahalu D, Safran S, Bershadsky A, Addadi L, and Geiger B. Force and focal adhesion assembly: a close relationship studied using elastic micropatterned substrates. Nat Cell Biol 3: 466–472, 2001.[CrossRef][ISI][Medline]
  3. Belkin VM, Belkin AM, and Koteliansky VE. Human smooth muscle VLA-1 integrin: purification, substrate specificity, localization in aorta, and expression during development. J Cell Biol 111: 2159–2170, 1990.[Abstract/Free Full Text]
  4. Bell E, Ivarsson B, and Merrill C. Production of a tissue-like structure by contraction of collagen lattices by human fibroblasts of different proliferative potential in vitro. Proc Natl Acad Sci USA 76: 1274–1278, 1979.[Abstract/Free Full Text]
  5. Bershadsky AD, Balaban NQ, and Geiger B. Adhesion-dependent cell mechanosensitivity. Annu Rev Cell Dev Biol 19: 677–695, 2003.[CrossRef][ISI][Medline]
  6. Bottaro DP, Liebmann-Vinson A, and Heidaran MA. Molecular signaling in bioengineered tissue microenvironments. Ann NY Acad Sci 961: 143–153, 2002.[Abstract/Free Full Text]
  7. Brightman AO, Rajwa BP, Sturgis JE, McCallister ME, Robinson JP, and Voytik-Harbin SL. Time-lapse confocal reflection microscopy of collagen fibrillogenesis and extracellular matrix assembly in vitro. Biopolymers 54: 222–234, 2000.[CrossRef][ISI][Medline]
  8. Brown RA, Pajapati R, McGrouther DA, Yannas IV, and Eastwood M. Tensional homeostasis in dermal fibroblasts: mechanical responses to mechanical loading in three-dimensional substrates. J Cell Physiol 175: 323–332, 1998.[CrossRef][ISI][Medline]
  9. Brown TD. Techniques for mechanical stimulation of cells in vitro: a review. J Biomech 33: 3–14, 2000.[CrossRef][ISI][Medline]
  10. Butt RO, Laurent GJ, and Bishop JE. Mechanical load and polypeptide growth factors stimulate cardiac fibroblasts activity. Ann NY Acad Sci 752: 387–393, 1995.[ISI][Medline]
  11. Carver W, Nagpal ML, Nachtigal M, Borg TK, and Terracio L. Collagen expression in mechanically stimulated cardiac fibroblasts. Circ Res 69: 116–122, 1991.[Abstract/Free Full Text]
  12. Carver W, Molano I, Reaves TA, Borg TK, and Terracio L. Role of the alpha 1 beta 1 integrin complex in collagen gel contraction in vitro by fibroblasts. J Cell Physiol 165: 425–437, 1995.[CrossRef][ISI][Medline]
  13. Cherish DA and Mecham RP. Integrins: Molecular and Biological Response to the Extracellular Matrix. San Diego, CA: Academic, 1994.
  14. Chicurel ME, Chen CS, and Ingber DE. Cellular control lies in the balance of forces. Curr Opin Cell Biol 10: 232–239, 1998.[CrossRef][ISI][Medline]
  15. Chiquet-Ehrismann R, Tannheimer M, Koch M, Brunner A, Spring J, and Martin D. Tenascin C expression by fibroblasts is elevated in stressed collagen gels. J Cell Biol 127: 2093–2101, 1994.[Abstract/Free Full Text]
  16. Choquet D, Felsenfeld DP, and Sheetz MP. Extracellular matrix rigidity causes strengthening of integrin-cytoskeleton linkages. Cell 88: 39–48, 1997.[CrossRef][ISI][Medline]
  17. Cukierman E, Pankov R, Stevens DR, and Yamada KM. Taking cell-matrix adhesions to the third dimension. Science 294: 1708–1712, 2001.[Abstract/Free Full Text]
  18. Cukierman E, Pankov R, and Yamada KM. Cell interactions with three-dimensional matrices. Curr Opin Cell Biol 14: 1–7, 2002.
  19. Delvoye P, Wiliquet P, Leveque JL, Nusgens BV, and Lapiere CM. Measurement of mechanical forces generated by skin fibroblasts embedded in a three-dimensional collagen gel. J Invest Dermatol 97: 898–902, 1991.[CrossRef][ISI][Medline]
  20. Eastwood M, McGrouther DA, and Brown RA. A culture force monitor for measurement of contraction forces generated in human dermal fibroblast cultures: evidence for cell-matrix mechanical signaling. Biochim Biophys Acta 1201: 186–192, 1994.[Medline]
  21. Eastwood M, Mudera VC, McGrouther DA, and Brown RA. Effect of precise mechanical loading on fibroblast populated collagen lattices: morphological changes. Cell Motil Cytoskeleton 40: 13–21, 1998.[CrossRef][ISI][Medline]
  22. Elsdale T and Bard J. Collagen substrata for studies on cell behavior. J Cell Biol 54: 626–637, 1972.[Abstract/Free Full Text]
  23. Freyman TM, Yannas IV, Yokoo R, and Gibson LJ. Fibroblast contractile force is independent of the stiffness which resists the contraction. Exp Cell Res 272: 153–162, 2002.[CrossRef][ISI][Medline]
  24. Gaudet C, Marganski WA, Kim S, Brown CT, Gunderia V, Dembo M, and Wong JY. Influence of type I collagen surface density on fibroblast spreading, motility, and contractility. Biophys J 85: 3329–3335, 2003.[Medline]
  25. Grinnell F. Fibroblasts, myofibroblasts, and wound contraction. J Cell Biol 124: 401–404, 1994.[Free Full Text]
  26. Grinnell F. Fibroblast biology in three-dimensional collagen matrices. Trends Cell Biol 13: 264–269, 2003.[CrossRef][ISI][Medline]
  27. Grinnell F, Ho CH, Tamariz E, Lee DJ, and Skuta G. Dendritic fibroblasts in three-dimensional collagen matrices. Mol Biol Cell 14: 384–395, 2003.[Abstract/Free Full Text]
  28. Harris AK, Wild P, and Stopak D. Silicone rubber substrata: a new wrinkle in the study of cell locomotion. Science 208: 177–179, 1980.[Abstract/Free Full Text]
  29. Harris AK, Stopak D, and Wild P. Fibroblast traction as a mechanism for collagen morphogenesis. Nature 290: 249–251, 1981.[CrossRef][Medline]
  30. He Y and Grinnell F. Stress relaxation of fibroblasts activates a cyclic AMP signaling pathway. J Cell Biol 126: 457–464, 1994.[Abstract/Free Full Text]
  31. Huang H, Kamm RD, and Lee RT. Mechanics and mechanotransduction: pathways, probes, and physiology. Am J Physiol Cell Physiol 287: C1–C11, 2004.[Abstract/Free Full Text]
  32. Katz BZ, Zamir E, Bershadsky A, Kam Z, Yamada KM, and Geiger B. Physical state of the extracellular matrix regulates the structure and composition of cell-matrix adhesions. Mol Biol Cell 11: 1047–1060, 2000.[Abstract/Free Full Text]
  33. Kessler D, Dethlefsen S, Haase I, Plomann M, Hirche F, Krieg T, and Eckes B. Fibroblasts in mechanically stressed collagen lattices assume a "synthetic" phenotype. J Biol Chem 276: 36575–36585, 2001.[Abstract/Free Full Text]
  34. Kirchhofer D, Languino LR, Ruoslahti E, and Pierschbacher MD. {alpha}2{beta}1-Integrins from different cell types show different binding specificities. J Biol Chem 265: 615–618, 1990.[Abstract/Free Full Text]
  35. Kolodney MS and Wysolmerski RB. Isometric contraction by fibroblasts and endothelial cells in tissue culture: a quantitative study. J Cell Biol 117: 73–82, 1992.[Abstract/Free Full Text]
  36. Kubínová L. Advantages of stereological methods in biomedicine: efficiently obtaining unbiased estimates of geometrical characteristics for 3-D structures. IEEE Eng Med Biol Mag 17: 110–115, 1998.[Medline]
  37. Lafrenie RM and Yamada KM. Integrin-dependent signal transduction. J Cell Biochem 61: 543–553, 1996.[CrossRef][ISI][Medline]
  38. Lambert CA, Soudant EP, Nusgens BV, and Lapiere CM. Pretranslational regulation of extracellular matrix molecules and collagenase expression in fibroblasts by mechanical forces. Lab Invest 66: 444–451, 1992.[ISI][Medline]
  39. Lo CM, Wang HB, Dembo M, and Wang YL. Cell movement is guided by the rigidity of the substrate. Biophys J 79: 144–152, 2000.[ISI][Medline]
  40. Loftis MJ, Sexton D, and Carver W. Effects of collagen density on cardiac fibroblast behavior and gene expression. J Cell Physiol 196: 504–511, 2003.[CrossRef][ISI][Medline]
  41. Malvern LE. Introduction to the Mechanics of a Continuous Medium. Upper Saddle River, NJ: Prentice-Hall, 1969.
  42. Mio T, Adachi Y, Romberger DJ, Ertl RF, and Rennard SI. Regulation of fibroblast proliferation in three-dimensional collagen gel matrix. In Vitro Cell Dev Biol Anim 32: 427–33, 1996.[ISI][Medline]
  43. Mow VC, Bachrach NM, Setton LA, and Guilak F. Stress, strain, pressure and flow fields in articular cartilage and chondrocytes. In: Cell Mechanics and Cellular Engineering, edited by Mow VC, Guilak F, Tran-Son-Tay R, and Hochmuth RM. New York: Springer-Verlag, 1994.
  44. Mudera VC, Pleass R, Eastwood M, Tarnuzzer R, Schultz G, Khaw P, McGrouther DA, and Brown RA. Molecular responses of human dermal fibroblasts to dual cues: contact guidance and mechanical load. Cell Motil Cytoskeleton 45: 1–9, 2000.[CrossRef][ISI][Medline]
  45. Nakagawa S, Pawelek P, and Grinnell F. Extracellular matrix organization modulates fibroblast growth and growth factor responsiveness. Exp Cell Res 182: 572–582, 1989.[CrossRef][ISI][Medlin