Novel Three-Dimensional Preclinical Model for Investigating Cartilage Regeneration, Incorporating Physiological and Pathological Mechanical Loading

The development and implementation of novel treatments for temporomandibular joint disorders (TMDs) are limited by the lack of preclinical models that closely mimic <i>in vivo</i> conditions. Specifically, models that support three-dimensional (3D) cell culture and mechanical stimulation are needed, as both factors significantly influence cell behavior and, consequently, the success of TMDs treatments. We designed a novel 3D <i>in vitro</i> model using cells subjected to dynamic moderate loading (simulating chewing) or constant excessive loading (simulating clenching) applicable as an <i>in vivo</i>-mimetic system for TMDs research. In this study, cylindrical constructs (diameter: 6 mm, height: 3 mm), composed of 3% agarose containing mouse bone marrow mesenchymal stem cells (BMSCs), were cultured in chondrogenic medium. Cells in these constructs exhibited a 3.0-fold increase in <i>Sox9</i> RNA expression compared to cells inside agarose constructs cultured in proliferation medium and a 2.5-fold increase compared to micromass cultures (golden standard). After confirming the chondrogenic potential, we subjected the cell-agarose constructs to dynamic and constant mechanical loading using a custom-designed bioreactor. The constructs were divided into five groups: unloaded controls, dynamic loading (5% compression at 1 Hz), and constant loading at 10%, 20%, and 30% compression for 20 min per day. Finite element modeling analysis revealed that by increasing the strain compression level, the uniformity of displacement and von Mises stress distribution within the hydrogel will be reduced. Moreover, 30% constant loading compromised the structural integrity of the agarose constructs. Nitric oxide production by BMSCs was significantly elevated in response to 5% dynamic loading (2.8-fold), 20% constant loading (2.7-fold), and 30% constant loading (2.4-fold) after 20 min of stimulation compared to controls. Furthermore, 5% dynamic loading significantly upregulated <i>c-Fos</i> (2.0-fold) and <i>c-Jun</i> (1.5-fold) gene expression relative to controls, while 10% and 20% constant loading also significantly increased <i>c-Jun</i> expression (1.5-fold and 1.6-fold, respectively). Interestingly, 20% constant strain reduced the number of live cells compared with 10% constant loading. These findings demonstrate that a novel 3D preclinical model allowing to investigate cartilage regeneration under <i>in vivo</i>-like physiological and pathological mechanical stimuli has been established, providing a promising platform for future studies on chondroinductive agents and mechanical loading treatments for the TMDs.