Elsevier

Journal of Biomechanics

Volume 67, 23 January 2018, Pages 184-189
Journal of Biomechanics

Short communication
Novel image analysis methods for quantification of in situ 3-D tendon cell and matrix strain

https://doi.org/10.1016/j.jbiomech.2017.11.030Get rights and content

Abstract

Macroscopic tendon loads modulate the cellular microenvironment leading to biological outcomes such as degeneration or repair. Previous studies have shown that damage accumulation and the phases of tendon healing are marked by significant changes in the extracellular matrix, but it remains unknown how mechanical forces of the extracellular matrix are translated to mechanotransduction pathways that ultimately drive the biological response. Our overarching hypothesis is that the unique relationship between extracellular matrix strain and cell deformation will dictate biological outcomes, prompting the need for quantitative methods to characterize the local strain environment. While 2-D methods have successfully calculated matrix strain and cell deformation, 3-D methods are necessary to capture the increased complexity that can arise due to high levels of anisotropy and out-of-plane motion, particularly in the disorganized, highly cellular, injured state. In this study, we validated the use of digital volume correlation methods to quantify 3-D matrix strain using images of naïve tendon cells, the collagen fiber matrix, and injured tendon cells. Additionally, naïve tendon cell images were used to develop novel methods for 3-D cell deformation and 3-D cell-matrix strain, which is defined as a quantitative measure of the relationship between matrix strain and cell deformation. The results support that these methods can be used to detect strains with high accuracy and can be further extended to an in vivo setting for observing temporal changes in cell and matrix mechanics during degeneration and healing.

Introduction

Tendons are load-bearing tissues that are essential for joint motion and stability. Macroscopic loads modulate cellular responses prompting an array of biological outcomes such as maintained homeostasis, degeneration, or repair. However, a major gap in our understanding of mechanotransduction pathways stems from the fact that the influence of macroscopic deformations on the microstructural response remains unknown. Our overarching hypothesis is that the unique relationship between extracellular matrix strain and cell deformation dictates biological outcomes. A first step towards addressing this hypothesis is the ability to characterize the local strain environment. Two-dimensional methods have been used to determine matrix strain by applying speckle patterns and photobleached lines to the surface (Andarawis-Puri et al., 2009, Cheng and Screen, 2007, Luyckx et al., 2014), or by tracking the planar movement of nuclei centroids (Arnoczky et al., 2002, Screen et al., 2004). Additionally, nuclear deformation methods continually identify the 2-D focal plane of maximum cross-sectional area to observe changes in aspect ratio (Arnoczky et al., 2002). Despite the value gained from established 2-D methods, 3-D methods are integral for capturing the complexity of the local strain environment that can arise from out-of-plane motion, particularly in the highly cellular, injured state. Therefore, the objectives of this study were (1) to assess the accuracy of digital volume correlation (DVC) to quantify 3-D matrix strain using images of the innate tendon microstructure, (2) to develop and validate a method for 3-D cell deformation and (3) to develop and validate a method for cell-matrix strain, a measure of the relationship between matrix strain and cell deformation.

Section snippets

Image acquisition and processing

Patellar tendons from C57BL/6 mice (n = 3) were stained with Hoechst 33258 (Thermo-Fisher) nucleic acid stain and nominally loaded to 100 g. Z-stack images (20×) through a depth of 50 µm were acquired using multiphoton microscopy for stained nuclei imaging (Fig. 1A) and second harmonic generation (SHG) imaging of the collagen fiber matrix (Fig. 1B). For the purpose of cell deformation and cell-matrix strain method development, cell nuclei images were assumed to represent those of cell

3-D matrix strain

For naïve tendon cell images (Fig. 1A), level of applied strain positively correlated with error in the X-direction (r = 1, p = .0004) and standard deviation within each sample (r = 1, p = .0004) (Fig. 6). In the Z-direction (r = 0.4286, p = .3536), however, there was no correlation between level of applied strain and error (Fig. 7A). Results for injured tendon cell images (Fig. 1C) in the X- (r = 0.9286, p = .0067) and Z-directions (r = 0, p > .999) were similar (Fig. 7C). In contrast, for SHG

Discussion

We developed image analysis methods for calculating 3-D matrix strain, 3-D cell deformation, and 3-D cell-matrix strain in tendon. This 3-D approach is unique because it considers out-of-plane motion, which can occur from misalignment with the uniaxial loading direction or anisotropy. 2-D methods have been used extensively to quantify strain transfer from the macro to microscales (Han et al., 2013, Upton et al., 2008), but many studies acknowledge that results are confounded by artifacts from

Conflict of interest

The authors have no conflict of interest.

Acknowledgements

This study was supported by grants from the NIH/NIAMS (AR052743 & AR068301). Imaging data was acquired through the Cornell University Biotechnology Resource Center with NIH S10OD018516 funding.

References (13)

There are more references available in the full text version of this article.
View full text