Poster W1 - Soft Folding Skeletal Muscle Bio-bot Fabricated with PEDOT:PSS

Introduction: • The use of differentiated muscle as actuators in soft robotics, termed biohybrid robotics, is a growing field combining biological and synthetic materials to create actuation in centimeter-scale soft robots that exhibit motions such as walking and gripping. While several prior studies have leveraged the rhythmic contraction of 2D monolayers of cardiac muscle to power swimming robots, culturing skeletal muscle in this format has proven highly challenging and has led to delamination of cells from their substrates due to the large passive tension forces generated by muscle fibers during development and even larger active forces generated during stimulated contraction. Given that skeletal muscle offers significant advantage over cardiac muscle in robotics, such as on/off control and adaptation to gain-of-function (exercise) and loss-of-function (damage) cues, there is a significant need to develop methods to control skeletal muscle contraction in 2D formats. We have developed a method for culturing C2C12 murine myoblast-derived muscle on a bi-continuous conducting polymer hydrogel, namely, poly(3,4-ethylenedioxythiophene): poly(styrene sulfonate or PEDOT:PSS) coated with extracellular matrix hydrogels known to promote muscle differentiation, Matrigel and fibrin. These conductive hydrogels exhibit robust mechanical and electrical properties under a range of strains and enable targeted stimulation of skeletal muscle. Unlike previous studies in which electrical stimulation triggers contraction of entire tissues, our platform enables us to control muscle activation precisely and spatiotemporally in specific regions of the gel substrate. Our method of differentiating skeletal muscle on conductive substrates can be used as a building block to create complex, controllable soft robots.

Materials and

Methods: To prepare the conductive substrate for cell culture with optogenetic C2C12 mouse myoblasts, first we prepared a 3.5% PEDOT:PSS to solvent (ethanol) solution from a 7% stock solution by mixing a 1:1 solution of 7% PEDOT:PSS and 70% ethanol. Then using a manual coater, set to a 2mm thickness, the PEDOT solution was spread onto a hydrophobic sheet for easy removal post drying. To dry the ethanol from the solution, the gel layer was left in ambient 1 day and then put in a 38C oven for 1 hour. Following this, the paper and gel layer was cut into 20mm by 7mm beams and left under UV for 1 hour for sterilization.
Since preliminary cell cultures with direct cell seeding on PEDOT gels resulted in minimal to no cell growth, we have developed a method to layer of fibrin gel on top of the PEDOT layer. PEDOT beams were sterilized and placed in growth medium for 1 day to drive swelling and protein adsorption that functionalized the gel surface for the binding of fibrin. Post swelling, a solution of 8mg/ml fibrinogen mixed with serum-free media, mixed with 4ul of thrombin was added to the PEDOT beams and left to polymerize with a weighted stamp to produce a thin gel layer (Fig. 1a-d). After the fibrin/PEDOT bilayer was created, 500,000 cells per well were seeded. After 2 days of growth media samples were switched to differentiation medium and after 4 more days they were optically and electrically stimulated.

Results, Conclusions, and Discussions: Comparing light stimulation of optogenetic cells to electrical stimulation shows that both methods generate muscle contraction, indicating our ability to grow mature contractile muscle on conductive PEDOT substrates. Leveraging our open-source computational frameworks to map muscle displacement in 2D monolayers, we noted no significant difference in contractile displacement for optical and electrical stimulation (Fig 2a-d).

While preliminary data doesn’t show significant difference in contractility between optical and electrical stimulation, we anticipate that future studies where we align muscle fibers unidirectionally will provide a more quantitative and controlled comparison between groups. Concurrent with experiments, we have generated computational models for rapid in silico iteration of biohybrid robot design. Experimentally determined values of beam geometry, PEDOT mechanical properties, fibrin viscoelasticity, and muscle force were input into our custom model (Fig 1d?). In our current configuration, we anticipate that muscle layer applying a unidirectional shear force of about 0.7 mN to the bilayer beam will generate a maximum deflection of about 700 microns. This shows that the conducting substrate has a stiffness that allows for significant beam deflection to be output via muscle contraction, showcasing the promise of our platform for generating biohybrid robots powered by 2D skeletal muscle monolayers.

Further work will include designing a bio-bot structure with separate PEDOT components and showing that we can control each PEDOT separately by sending different electrical signals to each one. This would show a step forward in the controllability of muscle-powered bio-bots and allow a platform for researchers to design and control more complex structures with conducting substrates.

Acknowledgements (Optional): Thank you to Hyunmin Moon and the Zhao Lab for facilitating PEDOT used in this work.