Measuring the biomechanical properties of cell-derived fibronectin fibrils

Abstract

Embryonic development, wound healing, and organogenesis all require assembly of the extracellular matrix protein fibronectin (FN) into insoluble, viscoelastic fibrils. FN fibrils mediate cell migration, force generation, angiogenic sprouting, and collagen deposition. While the critical role of FN fibrils has long been appreciated, we still have an extremely poor understanding of their mechanical properties and how these mechanical properties facilitate cellular responses. Here, we demonstrate the development of a system to probe the mechanics of cell-derived FN fibrils and present quantified mechanical properties of these fibrils. We demonstrate that: fibril elasticity can be classified into three phenotypes: linearly elastic, strain-hardening, or nonlinear with a "toe" region; fibrils exhibit pre-conditioning, with nonlinear "toe" fibrils becoming more linear with repeated stretch and strain-hardened fibrils becoming less linear with repeated stretch; fibrils exhibit an average elastic modulus of roughly 8 MPa; and fibrils exhibit a time-dependent viscoelastic behavior, exhibiting a transition from a stress relaxation response to an inverse stress relaxation response. These findings have a potentially significant impact on our understanding of cellular mechanical responses in fibrotic diseases and embryonic development, where FN fibrils play a major role.

Keywords: Biomechanics; Elastic modulus; Extracellular matrix; Fibrils; Fibronectin; Mechanobiology; Nonlinear elasticity; Optical tweezers; Viscoelasticity.

Fig. 1

Optical tweezing of cell-derived fibronectin fibrils. A A schematic of the experimental system. Cells assemble FN fibrils onto the top surface of microfabricated pillars and are subsequently removed. Remaining FN fibrils are stretched with optical tweezers to measure stress–strain relationships. B–E Immunofluorescence images of B the microfabricated pillars, C actin, D fibronectin, and E the composite image. Cell-derived fibrils are highlighted in the red box. Scale bar is 20 microns

Fig. 2

Characterizing the composition of cell-derived FN fibrils. A To ensure that FN is not cleaved by the cell extraction buffer, FN was incubated with increasing time exposed to the buffer, and no cleavage was detected (lanes 4–6). As a control, FN was incubated with trypsin exposed for the same times (lanes 8–10), where all FN is cleaved. B Dual labeling of FN and common ECM proteins to determine if ECM proteins are colocalizing with FN fibrils. Scale bar = 20 microns

Fig. 3

Quantifying FN fibril dimensions. (A–B) Scanning electron micrographs were acquired and used to calculate FN fibril diameter (scale bar = 10 µm). C Fibril diameter was measured at five (5) locations along each fibril (N = 15 fibrils), and the distribution of fibril diameters was compared to D the fibril diameter calculated from immunofluorescence images of fibrils (N = 20 fibrils). E Fibril diameter distributions were fit to a Gaussian equation. F Fibril diameter (as measured by SEM) as a function of normalized length along the fibril. G Fibril length versus fibril diameter indicates that fibril diameter is not a function of fibril length. H Gaussian mapping was used to calculate the fibril diameter from immunofluorescence images

Fig. 4

Optical tweezing of suspended FN fibrils. A–C Timelapse phase contrast images of a fibril (yellow arrow) being stretched by the bead (blue arrow) trapped in the optical tweezers. Red-dashed line is included as a reference position

Fig. 5

Elastic moduli of FN fibrils. Representative raw A force–time data and B force–displacement data for an FN fibril exposed to a repeated load-hold-unload strain pattern C. Raw data were transformed to D stress–strain data, as detailed in Supplemental Materials. Note that displacement in 5B refers to the bead displacement perpendicular to the fibril, whereas strain in C and D refers to the stretching of the fibril parallel to its axis. The transformations are described in Results and in more detail in SI Methods

Fig. 6

FN fibrils exhibit three distinct elastic responses. A The stress–strain curves for each fibril were fit to a piece-wise linear regression model, with a crossover point at $ϵ$ = 0.01. Best-fit values were determined for slopes E_o and E. B Fibrils were classified into phenotypes shown based on relative values of E_o and E.C Plot for initial and final Eo and E values for each fibril phenotype. D-F Representative stress–strain curves for D a “toe” fibril, E a linear fibril, and F a strain-hardening fibril. G-I Changes in E_o and E as a function of successive stretches for G “toe” fibrils, H linear fibrils, and I strain-hardening fibrils. Arrows indicate direction from first to last stretch

Fig. 7

FN fibrils exhibit two distinct stress relaxation phases. A A schematic description of the stress relaxation and inverse stress relaxation phases. B The Maxwell standard linear solid viscoelastic model, in which a linear spring (elastic arm) is in parallel with a spring-dashpot combination (viscous arm). C–E Representative stress–time curves for C a fibril exhibiting stress relaxation, D a fibril exhibiting inverse stress relaxation, and E a fibril exhibiting both stress relaxation and subsequent inverse stress relaxation

Fig. 8

Frequency and correlation of elastic phenotypes and stress relaxation responses exhibited in FN fibrils. The number of fibrils exhibiting strain-hardening (SH), linear or toe elastic responses versus fibrils exhibiting stress relaxation (SR), and inverse stress relaxation (IR) or both (SR-IR)