Development of a Biohybrid Tendon Interface for Muscle-Powered Robots

<b>Introduction: </b>Unlike abiotic materials like metal and plastic, biotic materials can communicate with their surroundings, adapt to stimuli, and self-repair damage. Incorporating these materials into engineered systems could foster smarter, more adaptable machines. We have shown that engineered skeletal muscle stretched around a flexible elastomer ‘skeleton’ can generate force and drive locomotion. Since these locomotive robots are powered by ‘living’ actuators, they can strengthen with exercise and heal from damage, capabilities not possible with abiotic systems [1]. However, the interface between biotic (muscle) and abiotic (skeleton) components in our robot is driven by friction, leading to slippage and inefficient force transmission. In the body, muscle is covalently tethered to the skeleton via tendons, which efficiently transmit force. Thus, we have developed a synthetic tendon to act as a biohybrid interface, enabling the design of more modular, efficient, and adaptive biohybrid machines.<br/><br/><b>Methods: </b>We developed an adhesive tendon in collaboration with Professor Xuanhe Zhao’s lab at MIT. It is composed of a poly(acrylic acid) hydrogel functionalized for tissue adhesion with N-Hydroxysuccinimide ester groups. To characterize the interface between synthetic tendon and muscle, we performed tensile tests on differentiated and undifferentiated muscle samples (n=3) bound to the synthetic tendon. Muscle samples were manufactured from C2C12 mouse myoblasts. Cell populations were seeded into a fibrin and matrigel matrix, then cast into polydimethylsiloxane molds. Tensile tests were performed by layering muscle between two pieces of tendon, and pulling the assembly apart until interface failure was observed.<br/>To understand the effect of the synthetic tendon on cell viability, a colorimetric MTS assay was performed, along with a pH exposure test. Absorbance readings were taken at 490nm from control and experimental C2C12 myoblast cultures that had been incubated with synthetic tendon overnight (n=3). pH readings were probed at 0 and 30 minutes of control and experimental samples of DMEM culture medium that had also been exposed to the synthetic tendon (n=3).<br/><br/><b>Results and Discussion: </b>Our experiments demonstrated that differentiated muscle had a greater mean force-at-break value (~680mN) than undifferentiated muscle (~520mN) at their interfaces, although this difference was not statistically significant. Both values are magnitudes greater than the mean contractile force generated by our muscle actuators (~300uN) [1], indicating that the interfacial strength far exceeds our target metric and that the muscle-tendon assembly should remain intact over multiple cycles of use. We also observed no statistically significant difference in cell viability from absorbance readings or media pH in response to tendon exposure, indicating biocompatibility.<br/>Furthermore, current preliminary tests evaluating the tendon-skeleton interface indicate a robust connection as well, asserting that the whole biohybrid assembly should be resilient to repeated use. We are also now quantifying the improvement in force transmission efficiency enabled by the interface, increasing the performance of our muscle bioactuators.<br/><br/><b>Conclusion:</b> We have developed a hydrogel tendon that serves as a robust biocompatible musculoskeletal interface. The unconstrained nature of this innovation will enable the design of novel robotic architectures capable of functional applications like walking, swimming, pumping, and gripping that were previously not possible to manufacture. We also anticipate our biohybrid interface will advance the fundamental understanding of how passive tension generated by tendons guides muscle fiber alignment during growth and in response to injury. Our work sets the stage for fabricating next generation biohybrid robots that can dynamically adapt to the unpredictability of the real-world.<br/><br/><b><u>Reference:</u></b> [1] R. Raman <i>et al.</i> (2017). <i>Nat. Protoc.</i>