The development of tendon tissue engineering and regenerative medicine depends heavily on whether in vitro mechanical stimulation can faithfully reproduce the physiological loading environment experienced in vivo. For load-bearing tissues such as rotator cuff tendons, appropriate mechanical cues are essential not only for cell proliferation, differentiation, and extracellular matrix remodeling, but also for the maturation and repair potential of engineered grafts. "However, conventional bioreactor platforms mostly rely on uniaxial stretching and therefore fail to capture the complex biomechanics of human motion, including multiaxial loading, interlayer sliding, and shear. As a result, the biological effects of multiaxial stimulation on cellular mechanotransduction and tendon-related phenotypic regulation remain insufficiently understood." said the author Zekun Liu, a researcher at University of Oxford, "Against this backdrop, developing an in vitro stimulation platform that more closely replicates human biomechanics has become a key challenge in advancing physiologically relevant tendon tissue engineering."
In this study, the authors developed a humanoid robotic bioreactor capable of reproducing human-like shoulder abduction–adduction motions and integrated a flexible strain sensor into the system for real-time monitoring of strain in engineered tendon constructs during multiaxial dynamic loading. Human mesenchymal stem cells were seeded onto decellularized tendon scaffolds and cultured under dynamic stimulation in either the humanoid robotic platform or a conventional uniaxial loading platform, with matched peak strain conditions used to directly compare the biological effects of different loading modes. Throughout the culture period, the transparent soft bioreactor design enabled noninvasive and continuous observation of cell morphology and alignment, while cell viability assays, morphological analysis, transcriptomic profiling, and protein expression studies were further performed to systematically evaluate how biomimetic multiaxial mechanical stimulation regulates cellular mechanotransduction and tendon-related phenotypic responses.
The results showed that the humanoid robotic bioreactor could deliver stable and measurable multiaxial mechanical stimulation to engineered tendon constructs, with robotic shoulder abduction–adduction generating peak strains of approximately 3.5% and 9.5% under external forces of 25 and 50 N, respectively. Compared with static culture and conventional uniaxial stretching, robotic loading markedly enhanced cell alignment along the loading direction and induced more pronounced mechanotransduction-related responses. Although dynamic stimulation led to a moderate reduction in cell viability, particularly under the 3.5% strain condition, transcriptomic and protein analyses suggested that this change was more consistent with a mechanically driven shift from proliferation toward phenotypic adaptation rather than dominant acute cytotoxic damage. Further analyses revealed that the multiaxial robotic platform triggered broader gene-expression changes than the uniaxial platform and showed stronger enrichment and activation of the PI3K/Akt signaling pathway. At the protein level, PI3K/Akt activation was most evident under the 3.5% strain condition, and the magnitude of p-PI3K/PI3K and p-Akt/Akt changes was overall greater in the robotic platform than in the uniaxial control, indicating that a more human-like multiaxial mechanical environment can more effectively reshape cellular mechanosensing and promote early tendon-related biological responses.
Overall, this study developed a humanoid robotic bioreactor capable of delivering human-like multiaxial shoulder motions and integrated a real-time strain-monitoring system, allowing engineered tendon constructs to be stimulated under a more physiologically relevant mechanical environment. The significance of this work lies in showing that biomimetic multiaxial loading can more effectively reshape cellular mechanosensing, alignment behavior, and activation of key pathways such as PI3K/Akt than conventional uniaxial stretching, thereby providing a mechanobiological basis for the development of more physiologically relevant tendon grafts. "At the same time, we observed that although cell viability decreased to some extent following mechanical stimulation, this change was not predominantly associated with acute cytotoxic damage, but was more likely indicative of a mechanically driven adaptation toward tendon-related phenotypes under a multiaxial loading environment. This further suggests that future tendon tissue-engineering platforms should not focus solely on cell survival, but should also place greater emphasis on how mechanical cues regulate early phenotypic programming and tissue maturation." said Zekun Liu.
Authors of the paper include Zekun Liu, Jinrong Lin, Tania Choreno Machain, Muhammad Hanif Nadhif, Yuyang Wei, Nicole Dvorak, Dylan Yeo, Yu Kiu Victor Chan, Alona Kharchenko, Rafael Hostettler, Antoine Jerusalem, Sarah Waters, Sarah Snelling, and Pierre-Alexis Mouthuy.
Authors would like to acknowledge financial support from the United Kingdom's Engineering and Physical Sciences Research Council (project number: EP/S003509/1) and the National Institute for Health Research (NIHR) Oxford Biomedical Research Centre (BRC).
The paper, "Humanoid Robotic Loading Enhances Mechanotransduction in Tendon Tissue Engineering" was published in the journal Cyborg and Bionic Systems on Mar 24, 2026, at https://doi.org/10.34133/cbsystems.0542.
