Biomimetic cell-actuated artificial muscle with nanofibrous bundles

Skeletal muscle is composed of bundles of elongated, multinucleated myocytes whose myofilaments are organized and bound by connective tissue to regulate force and mechanical strength. Mimicking this bundled, fibrous architecture is key to reproducing muscular function. Prior biohybrid actuators have harnessed living muscle cells for motion via electrical or optical control, and artificial muscles based on functional materials (hydrogels, nylon, CNTs) have advanced actuation capabilities. However, recent biohybrid systems integrating skeletal muscle cells on CNT sheets have largely been limited to two-dimensional constructs. This study aims to create a three-dimensional, fiber-shaped, biscrolled biohybrid artificial muscle that integrates living skeletal muscle cells with nanofibrous HPU/CNT scaffolds, inspired by the bundled architecture of natural muscle, to enhance alignment, mechanical robustness, electrical conductivity, and contractile function under external stimulation.

Fiber-shaped constructs are advantageous for forming higher-order assemblies (bundling, weaving, folding) and are supported by anatomical evidence highlighting the role of fibrous bundles in force regulation and mechanical strength of muscle. Cellular constructs and functional materials have been used to form fiber-shaped scaffolds for tissue regeneration and artificial muscle. Biohybrid actuators driven by muscle cells (e.g., jellyfish- and ray-inspired systems) demonstrate controllable motion via electrical or optical stimulation. Artificial muscles built from hydrogels, nylon, and CNTs exhibit diverse actuation behaviors. A recent advance integrated skeletal muscle cells with CNT sheets, but remained two-dimensional. Nanofibrous scaffolds are well documented to align incorporated cells, promoting tissue function; thus, combining aligned HPU and CNT nanofibers is a promising strategy to achieve structural alignment, mechanical strength, and conductivity for biohybrid muscle.

- Scaffold fabrication: Hydrophilic polyurethane (HPU) nanofibers were produced via electrospinning to create highly aligned mats (observed on the collector). These HPU nanofibers were attached to a coverslide and overlapped with carbon nanotube (CNT) sheets drawn from a nanotube forest to reinforce mechanical strength. The resulting HPU/CNT matrix exhibited fully reversible strain up to ~4% under cyclic stretching/releasing and an electrical conductance of 3.55 (±0.41). - Surface functionalization: To support myoblast attachment and differentiation, the HPU/CNT matrix was uniformly coated with collagen at 4 mg mL^-1. - Cell seeding and initial culture: C2C12 myoblasts were seeded at 5 × 10^5 mL^-1 onto the collagen-coated HPU/CNT planar matrix and incubated for 5 h to allow attachment. Cells anchored, spread, and aligned along the CNT scaffold. - Biscrolling to form 3D fiber: Approximately 5 h post-seeding, the cell-laden HPU/CNT matrix was biscrolled parallel to the alignment direction of the CNT nanofibers to form a 3D fiber. Hydrophilic PU (containing water) provided viscosity/adhesion to maintain the biscrolled structure. Hoechst nuclear staining confirmed cells embedded throughout the scroll. SEM imaging showed a porous nanofibrous surface and a layer-by-layer cross-sectional structure conducive to nutrient/oxygen delivery. - Proliferation and differentiation protocol: After biscrolling, fibers were cultured in growth medium for 2 days (proliferation) and then switched to differentiation medium for 6 days to induce myotube formation. Live cell morphology and multinucleation were monitored using calcein-AM (cytosol) and Hoechst (nuclei) at days 5 and 8. - Viability assay: At day 8, live/dead staining used calcein-AM (live, green) and ethidium homodimer-1 (EthD-1; dead, red) to quantify viability across the fiber. - Immunostaining for myogenesis: Terminal differentiation was assessed by immunocytochemistry for myosin and α-actinin, co-stained with Hoechst to identify nuclei. - Contractility testing: For actuation assessment, fibers were labeled with a fluorescent tracer to track motion and subjected to electrical field stimulation at 0.5 Hz using 15-ms step pulses at 80 V. A custom device enabled imaging of relaxation and contraction states to measure displacement (dislocated distance) between marked points. - Data readouts: Mechanical characterization (strain reversibility), electrical conductance of scaffolds, cell alignment and morphology, multinucleation, viability (% dead), expression of myogenic markers, and contraction displacement under defined stimulation conditions.

- Successful 3D biohybrid fiber fabrication: Biscrolled HPU/CNT nanofibers formed a porous, layer-by-layer structure embedding aligned C2C12 myoblasts, resembling bundled skeletal muscle. - Scaffold properties: HPU/CNT matrices showed fully reversible strain up to ~4% and an electrical conductance of 3.55 (±0.41). - Cell viability and differentiation: After 8 days (2 days proliferation + 6 days differentiation), fibers exhibited robust viability with ~7.27 ± 0.82% dead cells. Fluorescence imaging showed arrays of elongated, multinucleated myotubes aligned along the fiber. Immunostaining confirmed terminal differentiation with positive myosin and α-actinin signals. - Reversible actuation under electrical stimulation: Electrical field stimulation (0.5 Hz, 15-ms pulses, 80 V) induced contraction and relaxation cycles. Measured contractile displacement was ~2.73 ± 0.27 µm, comparable to previously reported differentiated C2C12 contractile distances (~2–4 µm at 1–2 Hz). - Biomimicry and integration: The bundled nanofibrous architecture and aligned myotubes closely mimicked native skeletal muscle fiber organization, enabling functional contraction suitable for biohybrid actuator applications.

By integrating living skeletal muscle cells into aligned, conductive, and flexible HPU/CNT nanofibrous bundles, the study addresses the limitation of prior 2D biohybrid constructs, enabling a three-dimensional, fiber-shaped artificial muscle that more closely mimics the natural bundled architecture. The scaffold’s hydrophilicity supports cell attachment and differentiation, while CNTs provide mechanical reinforcement and electrical conduction, both critical for synchronized excitation-contraction coupling. The observed displacement (~2.73 µm) under defined electrical stimulation validates functional contractility in a 3D construct, aligning with known C2C12 performance. The high viability and expression of myogenic markers indicate successful terminal differentiation within the biscrolled architecture. These results are significant for the field of soft biohybrid robotics and therapeutic systems, suggesting applicability to implantable actuators, medibot locomotion, and controlled drug delivery. Potential enhancement of contractile outputs may be achievable by engineering macroscopic geometries (e.g., coils, helices) to amplify strain. The platform also lends itself to neuromuscular interfacing and advanced control modalities, including optogenetics and motor neuron innervation for human–machine interfaces.

This work demonstrates a biomimetic, biscrolled biohybrid artificial muscle composed of hydrophilic polyurethane and carbon nanotube nanofibers integrated with C2C12 skeletal muscle cells. The hydrophilic, stretchable HPU promotes cell attachment/differentiation, while CNTs add conductivity and mechanical strength. Following proliferation and differentiation, aligned, multinucleated myotubes formed within the nanofibrous bundles and exhibited reversible contractions under electrical stimulation with micrometer-scale displacements. The architecture emulates native muscle fiber bundles and offers promise for soft robotic actuators, implantable medibots, and drug delivery systems. Moreover, the system is compatible with remote cellular control strategies (e.g., optogenetic modulation) and could be advanced by neuron innervation for sophisticated interfaces. Future work should optimize structural designs to increase contractile output and evaluate long-term performance in vitro and in vivo.

The study does not establish long-term durability of contractile function; the maintenance of activity over extended periods in vitro and in vivo remains unknown and requires further investigation. Force generation was not reported, with contractility quantified primarily by displacement, leaving open questions about output force and efficiency. In vivo integration, immune response, and functional innervation by motor neurons were not assessed. Full author affiliation details for all contributors (e.g., for the third affiliation) are not provided in the excerpt.