Quantification of Mechanical Stresses Within 3D Matrices Using Quantitative polarization Microscopy

Introduction: : Mechanical forces existing in tissues are critical in the regulation of a number of physiological and pathological processes, such as embryogenesis, angiogenesis and tumor cell metastasis. Cells can both respond to forces exerted on them and apply force to their surroundings. It is now well established that 3D cell cultures better represent the in vivo cell condition for the majority of physiological and pathological processes. Our ability to map forces distribution within a complex 3D biological system is therefore critical to shed light on underlaying biological processes. However, the vast majority of readily available techniques are tailored for 2D culture conditions, and establishing suitable techniques to quantify existing mechanical stresses/forces within 3D cell culture samples remains an ongoing challenge. This requires the development of robust, yet simple methodologies to quantify these 3D mechanical stresses/forces with minimal intrusion into the cell microenvironment. Several techniques have been developed to address this challenge, including 3D traction force microscopy (TFM), microfabricated sensor-based approaches, e.g. micropillar devices and microparticle-based force sensors. Most of these techniques offer several advantages but also come with critical drawbacks. Thus, adopting a methodology that does not assume a material model or introduce exogenous materials that disrupt the cell environment would lead to more accurate measurements of 3D forces/stresses within the matrix.

Materials and

Methods: : Cantilevers and U-shaped holders were made of poly(dimethylsiloxane) with solution with a curing ratio of 10:1 monomer cured at 100 °C and then sculpted to the required dimensions. Then collagen hydrogel was poured in the cantilevers and was allowed to polymerize for 1h at 37 °C under humidified atmosphere and 5% CO2. The deflections of the cantilever arm in pre-release and post-release modes were measured using a digital caliper or based on microscopy images of the cantilevers. Polarized images were acquired on a Zeiss Axio Observe 7 with a motorized rotating linear polarizer located above the condenser and a circular analyzer in the filter holder below the objectives. The image sequences were analyzed with an in-house MatLab code to obtain pixel-by-pixel retardance images for the subsequent analyses using the ImageJ software. Spheroids were made in ultra-low adhesion 96 well plates using either 10 000 Mµ2F uveal melanoma cell line, Human hepatic stellate cells or Human umbilical vein endothelial cells. After 48h hours, spheroids were encapsulated in a 1.5 mg ml-1 collagen hydrogel, left to invade for an additional 48h before fixation. The Rho Kinase inhibitor Y27632 was used for contractility inhibition assays at a concentration of 10 µM. The stress optic coefficient in collagen hydrogel was calculated using a custom optical setup that allowed for the measurement of the retardance under a known tensile strength and material properties.

Results, Conclusions, and Discussions:: Using U-shaped cantilever systems to apply external forces, we first assessed the ability of the QPOL to map the strain inside collagen hydrogels (Fig. 1). The correlation between the retardance and applied external forces on collagen was then evaluated by comparing the pre- and post-release images. Our results show a strong positive correlation between the retardance of the loaded collagen and the applied external force on collagen hydrogels. As the external force increased, higher strains were induced within collagen, resulting in an increase in the retardance values. We then proceeded to measure and compare the signal from the 3 cell lines using spheroids. The QPOL revealed a heterogeneous retardance distribution in the vicinity of Mµ2F and HHSteC spheroids. In contrast, the HUVEC spheroids displayed a relatively homogeneous retardance distribution. Using the measured stress optic coefficient, the retardance values were converted into a mechanical stress map and compute the radial traction stress in proximity of the spheroids (HHSteC spheroids shown, Fig. 2A,B). A 3D traction stress distribution map was generated to extract the force distribution along the optical axis. To further confirm that our system was measuring the cellular force transmitted to the collagen gels, we performed a contractility inhibition assay using HHSteC spheroid treated with Y27632. Notably, we observed a steady decline in the retardance signal and overall traction force over the span of 6h in spheroids treated with Y2762, whereas the signal in untreated spheroids was stable or showed a slight increase (Fig. 2C,D). Overall, we observed a strong correlation between the optical retardance and force-related parameters, i.e. external forces and maximum shear stress, in loaded 3D collagen. Going forward, the retardance signal obtained can be paired with the orientation angle between the optical setup axis and the measured change in the birefringence slow axis to decompose the stress in its σx and σy components, thus providing information of the stress orientation (Fig. 3). Together, the presented results demonstrate that QPOL can quantify mechanical stresses within nonlinear heterogeneous 3D matrices at high resolution.

Acknowledgements and/or References (Optional)::