Poster Q9 - Nonlinear elasticity and buckling of fibers regulate the compaction of extracellular matrix by contractile cells

Introduction: Cells in a fibrous matrix mechanically interact with each other by long-range force transmission and by deforming the matrix. This leads to the compaction of the matrix, when sufficient conditions are met (i.e. critical cell density). The mechanism by which the individual cell-level interactions initiate the macroscopic compaction remains unclear. Here, we provide a chronological description of the mechanical events at the single-cell protrusion level that lead to a collective network formation and subsequent compaction of fibrous matrices. We identify two mechanical properties of the matrix – nonlinear elasticity and buckling of constituent fibers as the key elements that regulate the compaction of ECM by cells. Based on our findings – we introduce a two-step mechanical model for compaction – formation of a cell network through strain-stiffening of the matrix and collective contraction by the buckling of the fibers within the matrix.

Materials and

Methods: We prepared free-floating disc-shaped 3D collagen gels using a high throughput PDMS scaffold. 3T3 fibroblast cells were mixed with liquid type I collagen, dispensed into the scaffolds and were polymerized in 37C temperature. The strain in the region between a pair of interacting cells was quantified as a function of time from the displacements of 2-micron tracer particles embedded into the matrix from bright field time-lapse images obtained for several hours.
The linear elastic 3D hydrogels with embedded cells were prepared using the same method described above. For this purpose, we used ColT gel - a commercially available product consisting of gelatin and crosslinked with transglutaminase. The stiffness of these gels can be tuned by varying the concentration of transglutaminase crosslinkers. We prepared gels with two different stiffness - 0.5kPa and 10kPa. The rate of compaction of these gels was quantified by measuring the area of the disc-shaped gels over time.
Confocal reflectance images of collagen fibers were obtained using a Zeiss LSM 800 confocal microscope. The buckling of the collagen fibers in cell-free native collagen gels and partially compacted collagen gels was quantified from the curviness of the fibers (i.e. average curvature of a fiber over its entire length). This was done from the confocal reflectance images using a custom MATLAB script. Tracking of a single collagen fiber in a compacting collagen disc was done by confocal time-lapse imaging of the gels with live cells.

Results, Conclusions, and Discussions: Mechanical interaction between cells within a 3D matrix is initiated when the protrusions of two cells persistently move toward each other by stretching the matrix fibers between them. This leads to the formation of a network of cells connected by the protrusion tips. To understand the mechanism of this network formation, we quantified the rate of strain in the region of the stretched fibers between two cells. We found that the strain increases up to 30% over time, which is above the critical strain for the nonlinear elastic response of collagen. By tracking the distance between the tips of two persistently approaching protrusions, we found that the persistent motion trails the rise in strain in the region between them, implying that strain-induced stiffening of collagen guides the persistent approach of the cells and the formation of the cell network (Fig. 1A-D).
To test if strain-stiffening or nonlinear elasticity of collagen is indeed responsible for initiating the mechanical crosstalk, we prepared free-floating ColT Gel discs with fibroblast cells and measured the rate of compaction over time for two different stiffnesses (0.5kPa and 10 kPa). Unlike collagen, the discs did not exhibit any significant compaction. We also did not observe any formation of cell network in these gels (Fig. 2A-D). The cells are randomly oriented throughout the matrix, suggesting that they do not establish mechanical communication with neighboring cells by force transmission and deformation of the matrix when the matrix is linear elastic.
From confocal reflectance images, the structural difference in the fibers of cell-free collagen gels and gels with cellular networks suggests that the fibers surrounding the interacting cells undergo buckling. We quantified this by calculating the average curvature of the fibers for the two cases (Fig. 3A-C). This finding is further supported by time-lapse imaging of a single fiber whose curvature increases as the collagen gel compacts (Fig. 4A-F).
Our study delves into the role of mechanical and structural properties of ECM in guiding the compaction of collagen gels by cells. It has potential implications for understanding diseases involving ECM remodeling (e.g. fibrosis) and for engineering functional and active biomaterials.

Acknowledgements (Optional): The author thanks the staff at the core facilities of Institute for Genomic Biology, University of Illinois Urbana Champaign for their assistance in doing the live-cell confocal imaging.