F18 - Quantifying the microenvironmental forces present during early musculoskeletal tissue formation

Introduction: : Mechanical stimuli are critical for tissue formation and development, but there is a lack of measurement tools that can directly measure forces in cell-dense structures without significantly influencing tissue organization. Our group previously developed cell-sized microparticle force sensors that could be used to measure the local forces exerted by surrounding cells and matrix in live spheroids by 3D imaging their deformation over time. This work focused on individual spheroids but was more limited in its application to larger tissue structures. The objective of the current study was to adapt the measurement approach for 2D imaging to allow quantification of the microenvironmental forces present during tissue condensation of dog bone-shaped microtissues as a model for early musculoskeletal tissue development. We hypothesized that forces in the neck region between the two posts would be primarily tensile and in the longitudinal direction, while forces in the ring regions around the posts would be more varied and circumferential in direction. By showing feasibility in this model system, microenvironmental force measurements can be extended to a variety of other natural and engineered tissue structures.

Materials and

Methods: : Hyper-compliant microparticle force sensors were fabricated via droplet generation. Sensor size and stiffness was characterized using ImageJ and atomic force microscopy, respectively, for use in computational analysis. Normal human fibroblasts (NHFs) and Prrx1-expressing limb-bud progenitor-like cells were expanded and frozen between passage 6 to 9 in preparation for seeding. Microtissue wells were prepared by stamping molten agarose with a 3D printed mold of a 3 x 8 dog bone array. Microtissue wells were acclimated with serum-free media for at least 96 hours and then seeded with NHFs or Prrx1 cells, along with sensors such that there were ~20-100 sensors/dog bone. At predetermined time points (4h, 8h, 24h), 2D images of the dog bones were taken using an epifluorescent microscope with brightfield and Cy5 filters. To calculate microenvironmental measurements, we first extracted binary images of the sensors in MATLAB, followed by computational analysis in Wolfram Mathematica to compute mechanical parameters (e.g., average pressure, elastic energy, and elastic energy density). Additional MATLAB scripts were then used to visualize the computed parameters and compare mechanical and geometric measurements over time and in different regions of the dog bones. Statistical analyses were performed using GraphPad Prism 10. Outliers in the data were identified using the ROUT method with Q=1% and removed before statistical testing. One- or two-factor analysis of variances with Dunn’s post hoc test for nonparametric pairwise multiple comparisons were used to analyze differences between regions and timepoints. Statistical significance was achieved at p< 0.05.

Results, Conclusions, and Discussions:: Imaging data were collected for 18 NHF-seeded dog bones and 15 Prrx1-seeded dog bones at 4h, 8h, and 24h timepoints. The same sensor lot (30 μm-diameter particles with an E=240 Pa) was used for all experiments. Qualitative force map results indicated increased sensor elongation over time, shown by greater arrow lengths. In the neck region, a more longitudinal orientation of sensors was observed by the 24h timepoint, particularly in the Prrx1 microtissues (Figure 1). Analysis of the ratio of sensor major-to-minor axes to quantify elongation showed increased values over time, and degrees from horizontal measurements to quantify orientation in the neck region showed decreased values over time in the Prrx1 dog bones (Figure 2). These results supported qualitative observations and previous computational models of dog bone structures. In addition, forces were primarily tensile in the microtissues, with significantly greater mechanical parameter measurements in the rings compared to the neck regions (Figure 3). However, average pressure magnitudes generally remained within a similar range, and similar trends were observed for both force and elastic energy density measurements. These unexpected results were opposite our initial hypotheses but aligned well with a published computational model that described earlier timepoints for stretched dog bone structures, before significant necking occurred, which suggests that our 24h timepoint may have been too early to fully capture changing forces that exist just before failure. Greater mechanical parameter measurements in the rings also aligned with previous computational models of other microtissue shapes with similar posts. This work supports the efficacy of our force sensor and analysis system for use in larger microtissues. Findings for our direct mechanical measurements matched well with previous computational models, providing validation of those virtual results. Potential future applications of this technology include measurement of other microtissue shapes, investigation of cell-ECM mechanical interactions, and real-time monitoring of the microenvironmental forces present in tissue-engineered constructs.

Acknowledgements and/or References (Optional):: The authors would like to thank Graylen Chickering for her work in fabricating the force sensors used in this study. Funding for this work was provided by the National Science Foundation (CBET 1547819).