The success of tissue engineering constructs in restoring healthy tissue function is driven by the interplay of cells with their microenvironmental cues. Therefore, the design of tissue engineering materials is typically guided by ensuring adequate mimicry and regulation of the dynamic biochemical, mechanical, and electrical interactions that occur in the cellular and extracellular milieu. In this work, we introduce the current approaches and limitations to static and stimuli-responsive tissue engineering, with a focus on electroactive materials. We consider the mechanisms of material interactions and the development of electroactive polymers for soft robotics to address how these developments can pave the way for ‘smart’ tissue engineering devices that recapitulate key elements of tissue bioelectromechanics. By highlighting the successes and current challenges in the materials development to support such dynamic devices, we summarize our findings with design guidelines to direct the future development of clinically translatable and efficacious tissue engineering constructs with the dynamic functionality of soft robots.
I. INTRODUCTION
Tissue engineering (TE) approaches apply principles in physical and life sciences to formulate constructs that can replace, support, and restore native tissue functionality. When tissue damage is excessive or continuous, the healing process is both prolonged and limited.1,2 The fabrication of scaffolds that support tissue repair and growth has allowed the initial challenges in tissue damage to be overcome, such as tissue integration in orthopedic implants,3 large-area engineering alternatives for autologous skin grafts,4 and tissue mimicking resorbable sutures to minimize facial scar formation.5 Across these biomedical applications, a crucial design element involves the creation of a tissue mimetic framework for cells that can also support the rebuilding of the healthy extracellular matrix (ECM) found in native tissue.
As a result, the majority of the research in biomaterials for tissue engineering has focused on achieving three key criteria to mimic healthy ECM: (1) ensuring adequate and well-matched stability/degradation rates, (2) aligning mechanical properties with the tissue of interest, and (3) optimizing topography and architecture to direct cell growth, motility, proliferation, and infiltration through structural6–11 and/or chemical modifications.12–16 The hybridization of materials to combine the three aforementioned requirements has resulted in a range of successful static tissue engineering devices, such as Integra® and StrataGraft® for skin grafts and Osteomesh™ for craniofacial repair.17–19
However, such static tissue engineering approaches have seen limited clinical success in other areas such as cardiac and neuronal scaffolds.20,21 Static tissue engineering devices rely heavily on pre-clinical laboratory data for the optimization of critical parameters associated with degradation and mechanics, as part of a long and arduous journey from initial mechanistic research to device design, with biological efficacy validated through in vitro and in vivo studies.22 In practice, this translates to the need for continuous monitoring of ‘scaffold take,’ or its successful incorporation into the implanted site from early to late stages of healing.23 Such static devices cannot respond dynamically to changes in the physiology of the patient, thus offering little scope for personalization or mid-treatment intervention.22,24
Consequently, in recent years, there has been an increased interest in the development of stimuli-responsive materials in tissue engineering constructs. These materials can support the integration of electrical, biological, and biochemical properties, which can interact and respond dynamically to the surrounding environment,25 an advantage over passive constructs.26 Dynamic structures may offer both new research avenues and translational opportunities for tunable devices that are capable of stimulus modulation via control systems and biological feedback loops, overcoming current challenges in static tissue engineered constructs.27,28 Materials that elicit a response from exposure to single or multiple stimuli are often referred to as “smart” or “functional” materials. The functionalization of materials to adapt to internal biochemical, mechanical, and electrical stresses has shown promising results in a variety of therapeutic applications, particularly for improved induced pluripotent stem cell (iPSC) maturation, antimicrobial coatings, and controlled drug delivery through in vitro and in vivo models.29–33
Of these materials, electroactive polymers (EAPs), a broad group of polymers and composites that produce a mechanical deformation in response to an applied electrical field, have garnered interest for their ease of manipulation through exogenous electrical signals. EAPs are widely used in soft robotics since they allow for sensing, actuating, and structural control within a single device. Soft robots function by continuous deformation and have an infinite number of joints and degrees of freedom as compared with their rigid-body counterparts.34 Soft robots therefore rely on three main variables: (1) exogenous excitatory or inhibitory stimulus, (2) signal transduction, and (3) a stimuli sensitive output response. This is similar to smart materials, with the main difference being the ‘robotic’ control factor, which allows the operator to control the stimuli, only providing a response under a direct, regulated command.34,35 EAPs are therefore candidate materials to provide controlled actuation and bioelectric signal transduction in future biomedical applications by facilitating self-powered biomedical soft robotics, triboelectric energy harvesting, actuating scaffolds, and post-implant control.25,36,37
As the use of electromechanical stimulation moves to drive both the physical changes in devices as well as to induce biological effects in the tissues of interest, EAP-based tissue engineering devices will require a consideration of two key aspects: they must operate at both therapeutically safe and clinically relevant stimulation regimes in addition to possessing the force generation and high response rates that are typically observed in dynamic soft robots. This perspective aims to unify the two disparate strands of research by considering the current approaches to tissue engineering in Sec. II, the role of bioelectromechanics in tissues and tissue engineering in Sec. III, and the recent advances in materials development within electroactive polymers in Sec. IV. Finally, we highlight the key successes and limitations of strategies to introduce electroactivity into biopolymers and conclude with design guidelines for an ideal clinically translatable and efficacious dynamic tissue engineering construct in Sec. V.
II. STATIC AND DYNAMIC TISSUE ENGINEERING SCAFFOLDS: CURRENT APPROACHES AND LIMITATIONS
A. The Extracellular Matrix and Engineered Substitutes
The ECM is a dynamic 3D microenvironment that is in a constant state of degradation and remodeling. This network of proteins and polysaccharides is responsible for the architectural organization, mechanical properties, and stability of tissue, as well as for facilitating biochemical signaling and mediating each stage of the cell cycle via a combination of growth factors and bioactive cues.38,39
During tissue damage, the ECM loses integrity, limiting its ability to regulate healthy cell and tissue function by changing the biochemical and electromechanical cues to drive rapid, and often haphazard, repair.40 Promoting successful tissue repair, therefore, requires an understanding of the interplay between ECM compositions and the cellular species involved during both homeostatic function as well as in their diseased states.
Collagen, elastin, and glycosaminoglycans (e.g., hyaluronic acid) are the major ECM constituents in skin and neuronal tissue whose electromechanical properties are further explored in Table I. Collagen and elastin are the main structural units, responsible for the biomechanics of different tissues and subsequently their cellular behavior.39,41 GAGs bind to membrane proteins to form proteoglycans, which provide structural support, regulate migration, and regulate proliferation and differentiation, as well as contribute to cell–cell signaling.42,43 In particular, hyaluronic acid is involved in the dynamic remodeling of the ECM, forming porous networks for cell infiltration and tissue growth.
Summary of ECM derived material properties in dry and hydrated states, highlighting the mechanical properties (Young’s modulus, tensile strength, and strain to failure) alongside the electrical (dielectric and conductivity) and biological (sites for cell binding) attributes.
ECM derived material . | Biopolymer type . | Young’s modulus . | Tensile strength . | Strain to failure (%) . | Dielectric permittivity (ɛ0) . | Conductivity (S/cm) . | Cell binding motifs . |
---|---|---|---|---|---|---|---|
Collagen | Protein | 20 MPa–2.3 GPa46,47 | 10–100 MPa13,46 | 10–3013,46 | 1.29–2.3048,49 | 1 × 10−13, 2.98 × 10−1050 | GFOGER12 |
Gelatin | Denatured protein | 31 MPa–1.14 GPa51,52 | 56–65 MPa51,52 | 2.2–6.151,52 | 0.78–1353 | 3 × 10−1054 | DGEA, RGD39 |
Elastin | Protein | 0.60–1.1 MPa46,55 | 1–2 MPa46,55 | 100–15046,55 | 2.7–4.056 | ⋯ | VGVAPG, EBP57,58 |
Keratin | Protein | Dry: 1.05 × 109 wet: 1.00 × 10659 | Dry: 50 × 106 wet: 0.5 × 10659 | 12 | ⋯ | 0.13 × 10−3 to 75 × 10−1260 | ⋯ |
Bone apatite/hydroxyapatite | Bone mineral | 80–110 GPa61 | 40–300 MPa61 | ⋯ | 4–962 | ⋯ | ⋯ |
Hyaluronic acid | Non-sulfated glycosaminoglycan | ⋯ | 0.35–0.62 Pa63 | 6.39–10.1263 | 5–4064 | 1 × 10−1564 | CD44 |
ECM derived material . | Biopolymer type . | Young’s modulus . | Tensile strength . | Strain to failure (%) . | Dielectric permittivity (ɛ0) . | Conductivity (S/cm) . | Cell binding motifs . |
---|---|---|---|---|---|---|---|
Collagen | Protein | 20 MPa–2.3 GPa46,47 | 10–100 MPa13,46 | 10–3013,46 | 1.29–2.3048,49 | 1 × 10−13, 2.98 × 10−1050 | GFOGER12 |
Gelatin | Denatured protein | 31 MPa–1.14 GPa51,52 | 56–65 MPa51,52 | 2.2–6.151,52 | 0.78–1353 | 3 × 10−1054 | DGEA, RGD39 |
Elastin | Protein | 0.60–1.1 MPa46,55 | 1–2 MPa46,55 | 100–15046,55 | 2.7–4.056 | ⋯ | VGVAPG, EBP57,58 |
Keratin | Protein | Dry: 1.05 × 109 wet: 1.00 × 10659 | Dry: 50 × 106 wet: 0.5 × 10659 | 12 | ⋯ | 0.13 × 10−3 to 75 × 10−1260 | ⋯ |
Bone apatite/hydroxyapatite | Bone mineral | 80–110 GPa61 | 40–300 MPa61 | ⋯ | 4–962 | ⋯ | ⋯ |
Hyaluronic acid | Non-sulfated glycosaminoglycan | ⋯ | 0.35–0.62 Pa63 | 6.39–10.1263 | 5–4064 | 1 × 10−1564 | CD44 |
While collagen provides a stiff matrix (MPa–GPa), elastin introduces tissue elasticity (100%–150% strain to failure), allowing the tissue to return to its original shape after deformation.39,41 Although hyaluronic acid does not possess comparable values for stiffness and elongation, its ability to retain moisture aids with shock absorption and lubrication of tissues.44 Tissue organization can result in isotropic or anisotropic strains under an applied load. This results in a dominant tensile, compressive, torsional, or shear strength depending on tissue location. It is important to consider that the mechanical properties of tissues are not exclusively a result of the base material itself but also its conformation, which may change from interactions with the cellular components.45
Biological interactions between ECM materials and cells can be characterized by the peptide adhesion sequences available on the material substrate. Collagen mediates integrin binding via GFOGER sites.12,39 Elastin is less active in cell binding, targeting non-integrin-based glycoside protein through the VAPG cell binding motif.57 Degradation of these materials is also controlled by the presence of substrate sequences for enzymatic action (collagenase for collagen and elastase for elastin).46
Ligand binding capabilities and changes in physical, bioelectric, and biochemical properties of the ECM are critical considerations in the design of tissue engineering constructs.21,65,66 The attributes of both the substrate and the specific cells of interest contribute to optimal tissue function, allowing for an uninterrupted cell cycle for tissue renewal and restoration of homeostatic conditions in the event of tissue damage.
Archetypal biopolymers, such as those naturally derived from the ECM, are advantageous over synthetic polymers in achieving the promotion of cell–ECM binding, the mimicry of tissue stiffness and composition, and the ability to be degraded by local enzymes for remodeling into new, healthy ECM.39,67 However, batch-to-batch variability, the heterogeneity of natural materials, and associated costs in obtaining standardized precursors have dissuaded their extensive use within commercial TE devices.67 Furthermore, the extraction, processing, and purification of biopolymers can alter their material properties, giving rise to wider ranges of mechanical, physicochemical, and electrical properties in practice.
Synthetic polymers such as polycaprolactone (PCL), poly(glycolic acid) (PGA), and polyvinyl alcohol (PVA), on the other hand, are often developed under precise manufacturing processes to create homogeneous scaffolds with fine-tuned control over structural design and mechanical properties (elastic moduli, tensile and compressive strength).68 As such, PCL and PLGA electrospun scaffolds have been comprehensively studied for use in applications such as biodegradable artificial blood vessels, with a view to improve the long term mechanical stability (elastic moduli, tensile and compressive strength until uniaxial and dynamic loading) over biologically derived alternatives.69
The lack of intrinsic bioactive sites available for cell interactions in synthetic polymers has also been attributed to implant rejection or aggravated foreign body responses, as non-degradable structures are encapsulated in a fibrous capsule rather than integrated into the host tissue.67,70 Hybrid constructs incorporating biopolymer fillers and coatings have been designed to present the necessary biochemical cues for host tissue integration. Specifically, multiple studies report the addition of collagen, fibrin, gelatin, and other ECM-derived polymers to improve cell adhesion and viability while reducing the overall immunological response to the construct.71–74
B. Stimuli-responsive scaffolds
Stimuli-responsive, or “smart,” materials display large changes in their properties in response to small perturbations in their environment. The impact of these dynamic smart materials often manifests as the manipulation of electrical and mechanical cues in tissues, which can determine cell fate and be used therapeutically to support repair.20,40,75 Exogenous stimuli such as UV,76 temperature,32,77 and pressure78 can drive shape changes with the aim of generating mechanical stimulation, whereas pH79 and magnetism80,81 can manipulate internal fields and ionic gradients in conjunction with the mechanical action of a base substrate.
The advantages and limitations of the use of various stimuli-responsive modalities are outlined in Table II, yet in brief, three recurring themes can be identified surrounding the limitations in smart polymeric devices for tissue engineering: (1) tissue damage during activation, (2) control in physiological environments, and (3) activation in situ.
Smart materials for tissue engineering: A summary of the most common stimuli-responsive materials used in tissue engineering constructs with a list of their advantages and limitations.
Stimulus . | Advantages . | Limitations . | Refs. . |
---|---|---|---|
Photo-responsive | • Activation using specific wavelengths, typically ±80 nm, can allow for controlled responses.82 | • Dependent on light penetration depths of tissues in vivo (maximum of 5 mm through the skin at 750 nm83). | 76 |
• No physical stimulation required. | • Intense light exposure (3.6 J/cm2 for white light) can damage cells, particularly at low wavelengths (445 nm).84 | ||
Thermo-responsive | • Can respond to internal temperature changes. | • Narrow operating temperatures (37 ± 5 °C) in vivo.85 | 32, 77, and 88 |
• Exogenously induced temperature changes can trigger homeostatic mechanisms.86,87 | |||
pH-sensitive | • Sensitive to small changes (pH ± 2) in the local environment.89,90 | • Deviation from homeostatic pH can damage proteins.91 | 79 and 92 |
• Useful for controlled degradation, drug delivery, or sensing.79,89,90 | • pH change triggers homeostatic mechanisms.91 | ||
• Limited control in vivo. | |||
Magneto-responsive | • Control direction and magnitude by varying the intensity and polarity. | • High magnetic field strengths (10 s of mT) often required in situ.93,94 | 80, and 81 |
• Non-invasive. | • Secondary heating can cause tissue damage (±1500 mT can cause a ±150 °C heating/cooling effect).94 | ||
Electroactive | • Can mimic biomechanoelectric cues, e.g., wound contraction or cardiomyocyte contractile function95,96 | • Tissue impedance affects the potential with conductivities varying from 0.93 S/m for muscle to 1–10 S/m in the spinal cord.97 | 99–101 |
• Controlled response for drug delivery and actuation. | • Invasive electrodes can be painful. | ||
• Control of response by varying the stimulation intensity and duration. | • Non-degradable electrodes require surgical removal. | ||
• Can be non-invasive. | • Tissue damage more likely at high voltages (300–500V) and through secondary heating.98 | ||
Mechano-responsive | • Can respond to internal stress/strains in real time. | • Hysteresis: Reduced mechanical output over repeated cycles. | 88 and 103 |
• Self-healing polymers can improve product lifetime. | • Tissue constraints in vivo can alter actuation behavior. | ||
• Can provide mimetic mechanical cues, similar to the native environment, to encourage natural behavior (typically stresses between 3 and 5 kPa for fibroblasts).102 | • Activation methods are difficult to design for in vivo translation. |
Stimulus . | Advantages . | Limitations . | Refs. . |
---|---|---|---|
Photo-responsive | • Activation using specific wavelengths, typically ±80 nm, can allow for controlled responses.82 | • Dependent on light penetration depths of tissues in vivo (maximum of 5 mm through the skin at 750 nm83). | 76 |
• No physical stimulation required. | • Intense light exposure (3.6 J/cm2 for white light) can damage cells, particularly at low wavelengths (445 nm).84 | ||
Thermo-responsive | • Can respond to internal temperature changes. | • Narrow operating temperatures (37 ± 5 °C) in vivo.85 | 32, 77, and 88 |
• Exogenously induced temperature changes can trigger homeostatic mechanisms.86,87 | |||
pH-sensitive | • Sensitive to small changes (pH ± 2) in the local environment.89,90 | • Deviation from homeostatic pH can damage proteins.91 | 79 and 92 |
• Useful for controlled degradation, drug delivery, or sensing.79,89,90 | • pH change triggers homeostatic mechanisms.91 | ||
• Limited control in vivo. | |||
Magneto-responsive | • Control direction and magnitude by varying the intensity and polarity. | • High magnetic field strengths (10 s of mT) often required in situ.93,94 | 80, and 81 |
• Non-invasive. | • Secondary heating can cause tissue damage (±1500 mT can cause a ±150 °C heating/cooling effect).94 | ||
Electroactive | • Can mimic biomechanoelectric cues, e.g., wound contraction or cardiomyocyte contractile function95,96 | • Tissue impedance affects the potential with conductivities varying from 0.93 S/m for muscle to 1–10 S/m in the spinal cord.97 | 99–101 |
• Controlled response for drug delivery and actuation. | • Invasive electrodes can be painful. | ||
• Control of response by varying the stimulation intensity and duration. | • Non-degradable electrodes require surgical removal. | ||
• Can be non-invasive. | • Tissue damage more likely at high voltages (300–500V) and through secondary heating.98 | ||
Mechano-responsive | • Can respond to internal stress/strains in real time. | • Hysteresis: Reduced mechanical output over repeated cycles. | 88 and 103 |
• Self-healing polymers can improve product lifetime. | • Tissue constraints in vivo can alter actuation behavior. | ||
• Can provide mimetic mechanical cues, similar to the native environment, to encourage natural behavior (typically stresses between 3 and 5 kPa for fibroblasts).102 | • Activation methods are difficult to design for in vivo translation. |
Parameters such as the time and intensity of exposure to exogenous stimuli should be carefully accounted for in the design of activation regimes to ensure that the desired scaffold responsiveness is achieved while mitigating the risks of short or long-term damage. Narrow operating conditions and changes in the internal microenvironment (e.g., temperature and pH) can also reduce the ability to control the materials with exogenous stimuli post-implantation. Light penetration depth, magnetic field application strengths, and the electromechanical impedance of tissue also limit the activation via electromagnetic stimuli in translational applications.
Although electrical stimulation (ES) allows for exceptional control through external circuitry, in many clinical studies the application has also required invasive or implantable non-degradable electrodes for deep tissue stimulation, which require secondary removal and introduce risk alongside surgical wounding.104–106
Electroactive polymers (EAPs) can be designed to overcome these issues, as they do not rely on internal charges to drive activation but can be locally controlled by the magnitude, polarity, frequency, and duration of an applied current, with soft and flexible electrodes providing exogenous routes for electrical current induction.25,99 However, the material compliance required and the parameter ranges useable for activation in situ are entangled by the electromechanical communication pathways embedded in tissue functions and cellular processes.
III. UNDERSTANDING ELECTROMECHANICAL INTERACTIONS IN TISSUES
A. Role of bioelectricity
Bioelectricity is an innate system driven by ion transport to create relative charge differences between cells that trigger location-specific responses to exogenous and endogenous cell stress. These serve as a form of cellular communication that allows for control of a variety of cellular and tissue processes.40,107
While cellular bioelectricity is characterized by changes in charge and ion accumulation within the intracellular and extracellular environments, bioelectricity at the tissue level is identified through quick and intense current amplitudes. For example, when an action potential depolarizes the sarcoplasmic reticulum of muscle fibers, voltage-gated calcium channels are opened.108 A mass binding of calcium ions to troponin (a protein preventing uncontrolled contraction) causes a conformational change, exposing the actin filament. Actin–myosin cross-bridges form, and myosin flexes to contract muscle fibers. Hydrolysis of ATP allows for the simultaneous cyclic formation and breaking of actin and myosin throughout muscle contraction.108 Without this bioelectric stimulus, muscle tissue would not contract (Fig. 1).
Representative bioelectric signals at the cellular level, where changes in ion concentration contribute to the intracellular (Vin) and extracellular (Vout) potentials, alter the membrane potential (Vmem). At the tissue level, bioelectric signals are apparent in muscle contraction, when an action potential triggers calcium production observed by a short pulsed depolarization. Using the skin as a model, a disruption at the tissue level, for instance through epithelial wounding as depicted in the schematic, results in an electric field generation across the injured site (i.e., the epithelium), causing cellular migration and proliferation in response to the change in bioelectric signaling. Similarly, during the regenerative process, the endogenous bioelectric pattern will return to its homeostatic norm.40
Representative bioelectric signals at the cellular level, where changes in ion concentration contribute to the intracellular (Vin) and extracellular (Vout) potentials, alter the membrane potential (Vmem). At the tissue level, bioelectric signals are apparent in muscle contraction, when an action potential triggers calcium production observed by a short pulsed depolarization. Using the skin as a model, a disruption at the tissue level, for instance through epithelial wounding as depicted in the schematic, results in an electric field generation across the injured site (i.e., the epithelium), causing cellular migration and proliferation in response to the change in bioelectric signaling. Similarly, during the regenerative process, the endogenous bioelectric pattern will return to its homeostatic norm.40
The bioelectricity of tissues is often modified during damage and disease. For instance, skin has an intrinsic trans-epithelial potential (10–60 mV between the dermis and stratum corneum) that is disrupted when the tissue is damaged.75,109 The changes in the electric and strain fields have been suggested to drive the migration of cells toward a wound site. For instance, electric field strengths between 50 and 400 mV/mm (similar to native wounds) have been shown to drive keratinocyte migration toward the cathode,110 with similar studies corroborating these effects in fibroblasts.111 Similarly, nervous, cardiac, and bone tissues all experience endogenous bioelectric signaling, and the alterations that arise during both successful and unsuccessful healing are, therefore, critical in guiding engineered tissue therapies.20,21,112,113 Therefore, wounding triggers a short circuit in tissue by creating a relative negative charge at the site of injury, generating an electric field due to the potential difference between the surrounding skin and the damaged tissue.
B. Mechanical Cues Mediate Cell Behavior
Exogenous mechanical forces exerted on cells and tissues are key inputs that allow activation of a range of responses at the cellular and tissue level. All stages of the cell cycle, from migration, proliferation, and differentiation to apoptosis, involve various forms of mechanotransduction114 that translate endogenous and exogenous mechanical cues into biochemical signals, allowing cells to sense and respond to the surrounding microenvironment.80 This transfer of exogenous signals follows the hierarchical structure of tissue, distributing forces among increasingly smaller units until individual membrane mechanotransducers (strain-sensitive components) are activated.114
Mechanotransducers include receptors, ion channels, and enzymes in the cell membrane and cytoskeleton. One type of mechanotransduction is dependent on the cell–ECM integrin binding, which transmits mechanical signals through the architectural tensegrity of the overarching network, i.e., the distribution of force throughout the hierarchical structure of tissue in states of pre-stress.114,115 Hence, the structure and composition of the ECM determine the maximum force withstood by cells. When stimulated, mechanotransducers exhibit changes that either activate or deactivate proteins. These receptors undergo a conformational change triggering a change in their binding specificity. Consequently, ion channels open, inducing electrochemical diffusion of ions in/out of cells.114,116 Enzyme activation subsequently triggers a variety of cellular responses that, in combination with other mechanotransducers, regulate signaling pathways.
Mechanically sensitive tissues such as muscle, cartilage, and bone are responsive to mechanical cues present in their environment or to those transferred by exogenous stimuli. In cartilage, mechanically gated ion receptors and mechanosensitive integrins initiate transcriptional cascades that result in chondrogenesis.117 Bone remodeling is mediated by the mineralization and demineralization of tissue via osteocyte mechanosensors for controlled activation/deactivation of osteoblasts and osteoclasts.117,118 Aside from tissues whose functionality relies on mechanotransduction, mechanical signaling cascades are also implicated in the initiation of ECM protein deposition in fibroblasts,78 increasing contraction of cardiomyocytes,119 driving action potentials in neurons,120 and activating or deactivating specialized cell pathways for induced pluripotent stem cell (iPSC) differentiation.20,121
Within the context of tissue engineering, it is important to consider how the routes of mechanotransduction lost through damaged ECM components can be replaced while simultaneously considering the impact of exogenous loads on other cellular processes, particularly apoptosis and necrosis. A change in cell dynamics is transferred through the extracellular components (into strain), which connect all cells in the tissue, resulting in time-dependent and elastic responses to deformation. As a result, optimization of the topography and overall morphology of engineered constructs that interact with damaged tissues can help recapitulate the tensegrity experienced in native tissue.20,114,122 However, the dynamic loading regimes manipulated by both endogenous and exogenous mechanical stimuli ultimately control the physiological responses. For example, activating actomyosin motors can generate local contraction/relaxation to stiffen/soften the surrounding ECM. Kumar et al. showed that cytoskeletal fibers are in a state of pre-stress, which induces changes in cellular mechanics depending on their anchorage to stiff and compliant substrates.122,123 This extended interaction was evaluated in tissue models by Van Oosten et al., which demonstrated the change in substrate elasticity under tensile and compressive axial strains in particle-loaded fibrin networks. Adherent and free particles were found to change the dynamic behavior from the tensile strain stiffening to both compression and tensile stiffening networks.45 Compared to the characteristic response of ECM proteins presented by Storm et al., the introduction of particles and cells to the matrix material resulted in an alteration of the mechanical properties from compression softening to stiffening.123 Consequently, over the course of repair, structural hierarchy and dynamic remodeling in situ will alter the mechanical forces experienced throughout the tissues.
Mechanical stimulation holds promise as a regenerative cue, particularly in cardiac applications, where an increase in oxygenated blood circulation and patient activity was observed following a surgical procedure that delivered shock waves (peak positive pressure of 120 MPa and peak negative pressure of 10 MPa) to cardiac tissue.105 Physical cues have also been shown to influence the differentiation lineage of human Mesenchymal Stem Cells (MSCs). Engler et al. demonstrated that the response of MSCs was influenced by matrix stiffness, with neurogenic commitment observed in low stiffness structures (0.1–1 kPa) and osteogenic on high stiffness substrates (25–40 kPa).124 Furthermore, Wahlsten et al. introduced dynamic mechanical loads to proliferating fibroblasts via pressure regulated bioreactors, demonstrating that cyclic loading at a frequency of 0.01 Hz between 0.1 and 1.5 kPa increased fibroblast population by 75%.78 Therefore, mechanical cues may play an underappreciated role in tissue regeneration, particularly in their potential as a proliferation rate limiting factor in static engineered constructs.
C. Interplay of Electrical and Mechanical Cues
Electromechanical coupling enables the functioning of multiple tissue types. In the peripheral nervous system, electromechanical coupling enables detection of changes in tactile low-threshold mechanoreceptors (e.g., Pacinian corpuscles and Merkel’s disks), which communicate physiological conditions via action potentials along myelinated axons to central nervous system effectors.116 In blood vessels, variations in blood pressure initiate changes in the frequency of action potentials, allowing stretch mediated mechanoreceptors known as baroreceptors to relay pressure and volume information through the peripheral nervous system in a negative feedback loop.125 Neurotransmitters (such as acetylcholine) are released at the medulla synaptic junctions, triggering a response to the effector organ (the heart). This response is then relayed to cardiomyocytes via the vagus nerve to initiate contraction to either increase or decrease the heart rate in an attempt to restore homeostatic conditions.125 This is an example of effectors recognizing and responding to the triggered neurotransmitter at synaptic junctions to cause a cascade of responses in an electrically and mechanically coupled system.
Although the mechanical and electrical pathways are activated by different stimuli, the indirect effect of stimulation often results in secondary stresses, be they chemical, mechanical, or bioelectric in nature.20,40 While changes in the electric field are important for the activation of ion channels to initiate action potentials, secondary results include activation of ionic pathways and pressure induced by osmosis; such uncontrolled electrical and mechanical signals have been suggested to trigger apoptosis.126,127 Two critical mechanisms for scaffold design consideration are discussed below, which describe the indirect electrical and mechanical cues that can facilitate electromechanical coupling.
1. Electroosmosis
Electroosmosis is an electrokinetic phenomenon where the presence of an electric field/electrochemical gradient causes diffusion of liquid through a porous medium (in this case, the membrane between cells or the ECM in tissues).128 Hydrated cations move toward their respective counter electrode, in turn causing fluid flow down the electrochemical gradient. Depending on the direction and intensity, cells can swell, becoming turgid or shriveled. Combined with shear forces from fluid flows, strains on the cell or surrounding microenvironment are established as homeostatic pressure is destabilized. This effect is observed experimentally in bone, where a piezoelectric potential was found to drive flow between bone canaliculi.129 Therefore, an electrical system can produce mechanical stimuli from secondary responses indirectly associated with the original mechanistic pathway. It is important to note that this process is reliant on a constant viscosity of Newtonian fluids across the applied strain range, an assumption that does not hold for many biological fluids that are thixotropic (e.g., blood, mucus, and semen).130–132
2. Flexoelectricity
A strain on the cell membrane can result in an associated change in curvature. The flexoelectric effect states that there is a coupling between the membrane curvature and the electric field of cells.133–136 Therefore, an alteration of the internal electric field causes redistribution of charges in the phospholipid bilayer, resulting in a strain on cells and subsequently the organization of charges within the cell. This behavior has been modeled by Harland et al. and previously observed experimentally by Li et al. with voltage clamped cochlear outer hair cells, where an applied voltage modified the tensile force when placed under tension via optical tweezers.133,137 Therefore, mechanical loads can cause a net flexoelectric effect, which is presented as an electric field in tissues when the polarity of cells differs from its resting or unloaded state. In simplified models, the direct flexoelectric effect has been described by Petrov as the change in curvature and resultant change in transmembrane electric field, given by , where c1 and c2 are the principal membrane curvatures, E is the transmembrane electric field, K is the curvature elastic modulus, and f is the flexo-coefficient, a parameter that accounts for the effects of proteins, ions, and molecular structure of membranes in summation.135
IV. ELECTROACTIVE POLYMERS FOR TISSUE ENGINEERING
Due to the inherent coupling of biochemical, bioelectrical, and biomechanical signaling, the selection of materials that are able to target multiple biological pathways through the application of a single exogenous stimulus is highly desirable. Electroactive polymers can produce a direct physical or biomechanical change in response to an applied electric field and, conversely, output an electrical signal in response to a range of external stimuli (e.g., temperature, mechanical agitation, pH, light).25,88 This controllable response caters to multiple applications with materials developed for cell stimulation in TE,138,139 electromechanical transduction,140 controlled drug delivery,119,141 and biocidal structures.142
Although unmodified polymers are typically insulators, EAP composites have seen electrical conductivities ranging from semiconducting (10−7 S cm−1) to metallic (10−3 S cm−1) ranges,99,143 which may be particularly useful for skin, nerve, and cardiac regeneration due to the innate electrical signaling implicated in their functionality.20,75,144 For typical soft robotic applications such as artificial muscles, pumps, valves, and sensors, the performance outputs such as maximum force and strain are significantly dependent on the electromechanical, thermal, and environmental properties of their constituent materials. These have been explored in greater detail by Bar-Cohen et al., who have developed a comprehensive review on the metrics pertaining to EAPs.145 Briefly, the magnitude and frequency of the applied electrical inputs used to obtain the desired electromechanical responses are directly linked to the mechanisms of electroactivity in EAPs. As a result, careful consideration should be given to the type of EAPs chosen in implantable devices to ensure that the desired functionality is achieved without causing further damage to the surrounding ECM in this tissue. EAPs are broadly divided into two main groups based on their mechanism of energy and charge transfer: (1) dielectric (dEAPs) and (2) ionic (iEAPs). dEAPs operate due to direct charge transfer, allowing them to function in both dry and wet environments but requiring larger voltages to achieve fast response deformation. On the other hand, iEAPs rely on the diffusion of ions through a matrix that requires an electrolyte for charge transfer. Therefore, iEAPs are typically limited to wet environments, as response time and deformation magnitude are restricted by the availability and the diffusion of counterions.146,147
A. Dielectric
1. Piezoelectric EAPs
As one of the quintessential classes of electromechanical energy transduction, piezoelectric polymers are capable of generating a potential difference under mechanical loads, achieving μV–V voltage outputs.148 Piezoelectric polymers rely on the same mechanism of action as ceramics, where the orientation of dipoles in the polymer structure induces an electrical potential under mechanical loading. Piezoelectric polymers are lightweight, flexible, and malleable, overcoming the low mechanical strains achievable in lead-based ceramics, despite their lower piezoelectric coefficients.
Synthetic piezoelectric polymers include polyvinylidene fluoride or polyvinylidene difluoride (PVDF), poly-L-lactide (PLLA), and polyacrylonitrile (PAN), which have been used in tissue engineered scaffolds as sensing (mechanoelectro) or stimulating (electromechano) platforms.148 The use of polylactic acid based nerve conduits by Chen et al. highlighted the successful application of piezoelectric polymers as a wireless ultrasound-activated in vivo electrical stimulation platform for enhanced nerve regeneration, producing a current density of 4.44 μA/cm2, with a peak acoustic pressure of 150 kPa when driven by a 1 MHz sinusoidal wave with a 5 μs pulse width and a 10 ms pulse interval.149 Similar applications in cardiac scaffolds by Doustvandi et al. employed a PVDF/graphene oxide (GO) electrospun cardiac patch to mimic the electromechanical coupling of native cardiac tissue, which is lost following an ischemic event. This composite was shown to be viable as both a sensor and stimulation matrix with a conductivity range between 0.014 and 0.017 S m−1 and voltage outputs of 1.02–9.44 V under loads of 0.235–0.627 N.149,150
The incorporation of ECM-derived constituents has been been investigated to improve the biocompatibility of the final device, both locally (tissue integration) and systemically (degradation and clearance). Some ECM-sourced biopolymers are inherently piezoelectric, with their piezoelectric nature purported to aid repair in native tissues such as bone and skin.151 While the piezoelectric coefficients of human tissue are generally in the pico/millicoulomb range, controlling both the architecture and chemistry of the ECM-derived components could improve their power outputs.151–153 Examples include recent demonstrations of a fourfold increase in piezoelectric response in aligned aerosol jet printed collagen fibers compared to drop-cast controls,154 as well as chemical modification of collagen side groups through routine crosslinking methods.152 Similarly, hydroxyapatite, a structural facsimile and compositional approximation to bone apatite, is intrinsically piezoelectric48,62 and may present additional routes for an improved electromechanical response in purely biologically derived constructs.49 However, control via exogenous signals in such ECM-derived piezoelectric constructs is still likely to require significantly higher input currents and voltages than in the synthetic alternatives without further functionalization. Aside from the inverse piezoelectric effect observed in piezoelectric polymers, activation voltages and current amplitudes exceed the limits for effective tissue engineering in dEAPs. A more detailed comparison of the different types of EAPs has been discussed by Shi and Yeatman.155 Therefore, in Sec. IV A, we have only discussed piezoelectric dEAPs alongside iEAPs in the context of tissue engineering.
B. Ionic EAPs
For biomedical applications, iEAPs are favorable due to their low activation voltages (1–7 V), their ability to function in aqueous environments, and their small actuation forces (reducing the risk of mechanical tissue damage).99 Actuating iEAPs change their shape in response to ion diffusion; typically, hydrated membranes with a charged polymer backbone facilitate counter-ion diffusion toward the electrode in a fixed system with three main flux gradients governing actuation: (1) counter-ion concentration gradient, (2) electric potential gradient, and (3) pressure gradient (Fig. 2).156
Ionic polymer composite actuation in thin films as a result of the electric potential, cation concentration gradient, and pressure gradient, modified from Pugal.157 In a polymeric matrix with bound ionic charges, the mobile counter-ions (often hydrated cations) are free to move under the influence of an applied electric field, resulting in a gradient in the counter-ions and, consequently, film deflection through changes in the osmotic pressure.
Ionic polymer composite actuation in thin films as a result of the electric potential, cation concentration gradient, and pressure gradient, modified from Pugal.157 In a polymeric matrix with bound ionic charges, the mobile counter-ions (often hydrated cations) are free to move under the influence of an applied electric field, resulting in a gradient in the counter-ions and, consequently, film deflection through changes in the osmotic pressure.
An immobilized polymeric backbone and the incompressibility of water induce an osmotic pressure, resulting in actuation from asymmetric volume distribution. Therefore, iEAPs can facilitate both actuation and sensing through changes in deformation under an applied voltage and changes in electrical resistance under applied strains. Table III summarizes common types of ionic EAPs, focusing on their mechanisms of action and accompanying advantages and disadvantages within tissue engineering. In particular, we will focus on two main types of EAPs in the next two subsections: conductive polymer composites and ionic polymer composites.
Comparison of EAPs in tissue engineering. Acronyms for the materials used in the table are as follows: Polyvinylidene fluoride or polyvinylidene difluoride (PVDF), Poly(3,4-ethylenedioxythiophene) (PEDOT), PSS (polystyrene sulfonate), Polyethylene glycol diacrylate (PEGDA), Calcium titanyl oxalate (CTO), 2-hydroxyethyl-trimethylammonium dihydrogen phosphate ([Ch][DHP]), 1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide (EMITFSI), Polypyrrole (PPy), and Polyurethane (PU).
Type of ionic EAP . | General structure . | Tissue engineering examples . | Advantages . | Disadvantages . | Refs. . |
---|---|---|---|---|---|
Piezoelectric polymers | Asymmetric dipole arrangements in the polymer network | • PVDF149 | • Facilitates electric potential generation under mechanical strain. | • Activation requires additional mechanical loading to sites of injury, which could progress damage. | 158 |
• Chitosan139 | • Under rest, a change in output voltage gives insight into resting stresses. | • Difficult to degrade at physiological temperatures | |||
• Collagen151 | • Mimics the mechanotransduction pathways in tissues for local electrical stimulation therapy. | ||||
Conductive polymers | Conjugated polymer network | • PEDOT:PSS72 | • Mediate bioelectric signaling and cell stimulation. | • Limited translational outcomes from in vitro to in vivo. | 158 |
• Poly(pyrrole)159 | • Allow sensing of internal bioelectric signals. | • Difficult to maintain functionality in complex environments. | |||
• polyaniline160 | • Potential for integration with tissue-specific control loops. | • Compliance decreases (by ∼47%) with an increase in the conducting polymer ratio (with 20% PPy in PU)161 | |||
Ionic polymer metal composites | Bound ionic functional groups along a polymer backbone | • Nafion162,163 | • Low actuation voltage ( V).164,165 | • Limited directional control | 169 |
• Fast response (10 s to maximum displacement164 | • Limited force generation ( N at 4 V.168 | ||||
• Large deformations (20%–100% of free film length).165–167 | |||||
Ionic polymer gels | Viscous ionic liquids suspended in a polymer matrix | • PEGDA170 | • No hydration needed for in vivo applications for aqueous-based systems. | • Slow response times (156 μ/sec for PEGDA170). | 99 |
• Collagen-actomyosin171 | • Low voltage activation (20 V170) | • Hysteresis limits performance capabilities over time. | |||
• Capable of large force generation and deformation (μN–mN170,171). | |||||
Electro-rheological fluids | Inorganic particle suspension | • Hybrid CTO composite172 | • Hydrostatic pressure like native mechanical loading. | • Sedimentation of particles reduces performance over time. | 99 |
• Fast response time (36 ms.172 | • Biocompatibility of suspension medium limits biomedical application. | ||||
Ionic liquids | Dissolved cation–anions in liquid medium. | • [Ch][DHP]173 | • Improves counterion migration in ionic polymers by maintaining actuator hydration with non-volatile solvents101 | • Biocompatibility | 99 and 174 |
• EMITFSI101 | • Cost of isolation/processing. |
Type of ionic EAP . | General structure . | Tissue engineering examples . | Advantages . | Disadvantages . | Refs. . |
---|---|---|---|---|---|
Piezoelectric polymers | Asymmetric dipole arrangements in the polymer network | • PVDF149 | • Facilitates electric potential generation under mechanical strain. | • Activation requires additional mechanical loading to sites of injury, which could progress damage. | 158 |
• Chitosan139 | • Under rest, a change in output voltage gives insight into resting stresses. | • Difficult to degrade at physiological temperatures | |||
• Collagen151 | • Mimics the mechanotransduction pathways in tissues for local electrical stimulation therapy. | ||||
Conductive polymers | Conjugated polymer network | • PEDOT:PSS72 | • Mediate bioelectric signaling and cell stimulation. | • Limited translational outcomes from in vitro to in vivo. | 158 |
• Poly(pyrrole)159 | • Allow sensing of internal bioelectric signals. | • Difficult to maintain functionality in complex environments. | |||
• polyaniline160 | • Potential for integration with tissue-specific control loops. | • Compliance decreases (by ∼47%) with an increase in the conducting polymer ratio (with 20% PPy in PU)161 | |||
Ionic polymer metal composites | Bound ionic functional groups along a polymer backbone | • Nafion162,163 | • Low actuation voltage ( V).164,165 | • Limited directional control | 169 |
• Fast response (10 s to maximum displacement164 | • Limited force generation ( N at 4 V.168 | ||||
• Large deformations (20%–100% of free film length).165–167 | |||||
Ionic polymer gels | Viscous ionic liquids suspended in a polymer matrix | • PEGDA170 | • No hydration needed for in vivo applications for aqueous-based systems. | • Slow response times (156 μ/sec for PEGDA170). | 99 |
• Collagen-actomyosin171 | • Low voltage activation (20 V170) | • Hysteresis limits performance capabilities over time. | |||
• Capable of large force generation and deformation (μN–mN170,171). | |||||
Electro-rheological fluids | Inorganic particle suspension | • Hybrid CTO composite172 | • Hydrostatic pressure like native mechanical loading. | • Sedimentation of particles reduces performance over time. | 99 |
• Fast response time (36 ms.172 | • Biocompatibility of suspension medium limits biomedical application. | ||||
Ionic liquids | Dissolved cation–anions in liquid medium. | • [Ch][DHP]173 | • Improves counterion migration in ionic polymers by maintaining actuator hydration with non-volatile solvents101 | • Biocompatibility | 99 and 174 |
• EMITFSI101 | • Cost of isolation/processing. |
1. Conductive polymers and composites
While most polymers are insulators ( S/m), conductive polymers, a class of polymers capable of facilitating free electron transfer, can achieve conductivities from 10−10–107 S/m, similar to ranges for metallic counterparts.143,175,176
Although it is uncommon, intrinsic conductivity is possible in conjugated polymers such as polyacetylene (PA) and polypyrrole (PPy).175–177 Contrary to the electron delocalization observed with typical π-conjugated systems (such as 2D graphene or 3D graphite), 1D polymeric systems such as PA are subjected to the Peierls’ distortion, giving rise to an electron mobility that is comparable to semiconductors.177,178 This is facilitated by a net current via a stepped system of partially localized electrons, achieving intrinsic conductivities of up to 107 S/m.143 Further conductivity enhancement can be achieved through doping, which can both improve electron mobility as well as improve thermal stability to prevent premature oxidation that can increase electrical resistance.143,177
Biological materials (such as collagen and elastin) have high electrical resistivity and require the incorporation of conductive materials such as metals or polymers.50,64,154,179 At high filler concentrations, or using high aspect ratio nanomaterials, electrons can flow through any conductive pathways formed by continuous contact of the particles through the polymer matrix. Considerations should be taken to retain the cytocompatibility of composite systems with high aspect ratios and concentrations of nanomaterials (such as graphite and gold) that may otherwise induce apoptosis.180–182 Bishal et al. reinforce this idea, where a platinum-coated collagen film improved the conductivity of the biomaterial by a factor of 1012 but reduced the cell population by 27% when compared to collagen control samples.183
Alternatively, conductive polymers have also been used as a more biocompatible composite constituent in studies with collagen-PPy and (poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate)) PEDOT:PSS scaffolds, when the device possessed an enhanced conductivity ( resistance) while maintaining or increasing in vitro viability.30,72 Many of these materials are stable at room temperature but are only electrically conductive within specific temperature ranges, which may limit performance in the body. For example, Chiang et al. report a five-fold increase in the electrical conductivity of undoped polyacetylene from 273 to 300 K.178,184
Conducting polymers are also widely used and sought after as electrodes for biomedical applications, serving predominantly as the link to deliver exogenous electrical stimulus29,138,185,186 or as electrochemical sensors to monitor biological signals (such as pH, temperature, oxygen saturation, and blood glucose).60,187,188 While distinct from the main focus of this perspective, the use of conducting polymer electrodes for such biomedical applications shares several similar (yet distinct) design requirements, as detailed in reviews by Zhou et al. and Hwangbo et al.189,190 For such implantable electrodes, future solutions should address the temporary nature of tissue engineered scaffolds, allowing adequate routes for removal, degradation, or resorption while maintaining adequate conductivity during its active lifetime.
As electroactive substrates, conducting polymers are capable of generating a mechanical strain in response to an applied current, as illustrated in Fig. 3.29,186 When a potential is applied to a conducting polymer, an oxidation/reduction reaction takes place in doped systems. Here, the dopants act as counterions to the charged polymer.25,29 Electrochemical oxidation/reduction causes the collection/ejection of ions in the polymer matrix. Ion migration alters the relative water potential of the system, causing macroscopic swelling/contraction of the polymer network from the osmotic-stress gradient that arises during the oxidation/reduction reactions.25,29,186 Electromechanical actuation in these systems, therefore, relies on free counterions in the solution to create an electroosmotic gradient, limiting application for implantable scaffolds. Such systems are generally non-degradable,31 can be cytotoxic,100,191 or are difficult to fabricate with biological carriers (e.g., collagen).50,192
Oxidation and reduction of conducting polymers to initiate electromechanical actuation via electro-osmotic swelling. The flow of the counter-ions in and out of the polymer produces a volumetric strain due to oxidation reduction reactions with the polymer. Figure adapted from Otero and Sansieña.193
Oxidation and reduction of conducting polymers to initiate electromechanical actuation via electro-osmotic swelling. The flow of the counter-ions in and out of the polymer produces a volumetric strain due to oxidation reduction reactions with the polymer. Figure adapted from Otero and Sansieña.193
2. Ionic polymer composites
Nafion® and Flemion™ are commercial examples of ionic polymer metal composites (IPMCs), originally popularized as membranes for proton exchange in fuel cells. Each possesses a similar backbone with highly charged functional groups that allow for directed cation electro-osmosis by immobilizing a hydrophilic anion group to their fluorocarbon backbone between two metallic electrode layers. In these cases, two groups are used, sulfonic acids (Nafion) and carboxylic acids (Flemion), respectively, due to their net negative charge.164,169 Osmosis is directed by relative water potentials across a membrane. Hence, increasing the net negative charge along the EAP backbone would enhance electroosmotic activity due to the larger gradient and electric field. The voltage-to-strain sensitivity of these EAPs is characterized by the back relaxation of the devices after removing the electric potential. For instance, during both a pulsed signal and a sustained potential, Nafion, the more strongly charged system, was found to hold the strained position whereas Flemion drifted back toward the anode over time.169
Commercial synthetic materials, used in soft robotic applications such as Nafion and Flemion, often lack the chemical ligands necessary for the desired cell–scaffold adhesion, representing a potential barrier to successful integration and repair. To improve the translation of EAPs in TE, biocompatible materials are needed. Huang et al. developed a cellulose derived actuator by functionalizing the backbone with 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO), where the oxidation of the primary hydroxyl groups to carboxylate groups facilitates ionic transport and osmotic swelling.167 The modified cellulose nano-fibrous films with sputter coated gold electrodes produced an electroactive displacement ranging between 12.8 and 32.1 mm (32%–80.35% of the free film length) under 10 V DC, comparable to Nafion responses at 2 V (120% of the free film length).167 Alginate and chitosan have also been explored as biocompatible and degradable solutions for iEAPs; however, these appear as copolymers in membranes that already contain iEAPs, instead of direct functionalization for ionic actuation.166,194 Currently, biopolymeric iEAPs rely on materials that are not native to the human or mammalian ECM which may induce pro-inflammatory responses in biomedical applications and foreign body responses, thus reducing post implantation functionality.195
In the interest of improving cytocompatibility and biocompatibility, ECM-derived biomolecular fillers have gained attention for promoting cell–material interactions.50,53,154,196 Some tissue-derived materials intrinsically express electromechanical properties, which could allow for robotic control if enhanced.151,152,154 A few approaches to enhance the electrical activity of ECM-derived materials, including the manipulation of material compositions, controlled fabrication, and various post-processing modifications (e.g., conductive coatings), have already been discussed in Sec. IV B 1 on conductive polymer composites.48,50,53,196 In the realm of ionic polymer composites, ECM-derived polymers such as collagen have not been subject to the same chemical modifications as cellulose, alginates, and chitosan. Several examples of co-mixed iEAP-ECM compounds have been observed with collagen and gelatin that include doping with ionic liquids or conjugation with organic molecules.72,184,197,198 Therefore, a gap appears in this area of research for directly functionalized ECM-sourced, biocompatible, actuating, and conductive EAPs. Identification of functionalization methods would widen the range of bio-safe EAPs, creating space for implantable devices that can directly sense, stimulate, and repair tissue. This would provide alternatives currently absent from the permanent synthetic materials (with varying degrees of cytotoxicity) for use in such devices.
V. CONCLUSIONS: CONSIDERATIONS AND KEY CHALLENGES FOR DYNAMIC TISSUE ENGINEERING SCAFFOLDS
EAPs are promising materials for combining soft robotics with tissue engineering, but challenges remain in actuation control, output forces, and degradation/resorption and response times. While dEAPs materials can be optimized to maximize strain outputs in artificial muscles (e.g., a fourfold improvement of strain (45%) at 11 kV199), they require prohibitively high voltages for biological applications. For sustained (non-pulsed) applications, such high voltages can exceed the dielectric breakdown voltage of most tissue, for instance, around 300–500 V for the skin.98 Hence, iEAPs are more desirable for artificial muscles that have a direct tissue interface, having demonstrated the ability to produce strains between 0.5% and 10% at much lower driving voltages (0.5–7 V).155,169,200 The examples in Table IV present several reoccurring limitations. For instance, conductive polymers such as PPy and polyaniline (PANi) are not degradable, and ideal structures should avoid surgical removal to prevent further damage at the implant site.138,142 Current iEAP scaffolds require highly ionic solutions for large displacements, which can be cytotoxic.197 Although encapsulation or surface modification techniques may mitigate cytotoxicity.139 While EAPs may deliver appropriate stimuli, their surface properties, degradation products, or other characteristics may hinder cellular adhesion, proliferation, or differentiation. Unlike EAPs, pH-sensitive and thermo-responsive polymers rely on internal mechanisms for actuation, limiting control.88 Activation conditions may not be physiologically relevant or may denature proteins. External control mechanisms may be necessary for precise responses in TE applications, a feasible route in EAP. These limitations highlight the need for further research in biomaterial design, surface modification, stimuli-response, biocompatibility, and control systems to fully harness the potential of EAPs in TE soft robotics. This perspective has also only considered the base materials that will facilitate the cell–material interactions and the primary drivers of electromechanical stimulation; the integration of sensors and feedback systems will also need further development for effective control of such devices.
Summary of the key findings in in vitro and in vivo studies of EAP scaffolds and the effect of electrical, mechanical, and electromechanical stimulation across different biomedical applications. Acronyms for the materials used in the table are as follows: PCL (Polycaprolactone), PEDOT (Poly(3,4-ethylenedioxythiophene)), PPy (Polypyrrole), TEMPO (2,2,6,6-tetramethylpiperidine-1-oxyl), CS (chitosan), QCS (quaternized chitosan), PANi (Polyaniline), Odex (oxidized dextran), Dex-AT (dextran-graft-aniline tetramer), CECS (N-carboxyethyl chitosan), PGS (Poly(glycerol sebacate)), MAETAC (2-(methacryloyloxy)ethyl-trimethylammonium chloride), PEGDA (poly(ethylene glycol)diacrylate), PSS (polystyrene sulfonate), P3HT (poly(3-hexylthiophene)), MEH-PPV (Poly(2-methoxy-5(2′-ethyl)hexyloxy phenylene vinylene)), PLA (Polylactic acid), Gel (Gelatin), PEO (polyethylene oxide), NBR (nitrile butadiene rubber), EMITFSI (ionic liquid 1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl imide)), KNN (potassium sodium niobate), AA (acrylic acid), CMA (collagen methacrylate), P(VDF-TrFE) (poly(vinylidene fluoride-trifluoroethylene)), polyHIPE (Polymerized High Internal Phase Emulsion).
Scaffold Format . | Material . | Application . | Main findings . | Limitations . | Refs. . |
---|---|---|---|---|---|
3D-printed macroporous | PCL with PEDOT | Bone regeneration | Conductive scaffold similar to PCL | No ES testing, no enhanced cell proliferation | 201 |
Conductive polymer films | Laminin coated PPy | Nerve TE | Enhanced neurite growth, good biocompatibility | Performance depends on dopant, no degradation data | 138 |
Biopolymer fibrous hydrogel | Collagen-actomyosin | Artificial muscles | Biomimetic actuator, repeated actuation | ATP activation may be limited, low mechanical strength | 171 |
Ionic nanofiber film | TEMPO-oxidized cellulose | Biopolymer actuator | Large displacements, fast response | High Young’s modulus, high activation voltage | 167 |
Shell/core nanofibrous film | PEDOT/CS | Nerve TE | Conductive, piezoelectric, increased neurite length | No degradation studies, potential toxicity | 139 |
Conductive hydrogel | QCS–PANi/Odex | Nerve TE | Increased cell proliferation, antibacterial properties | Potential toxicity, poor solubility, difficult degradation | 142 |
Electrode integrated fibrous patch | PCL:gelatin, gold electrodes | Cardiac patch | Functional cardiac patch, drug release | Long-term performance not assessed, risk of rejection | 119 |
Injectable hydrogel | Dex-AT/CECS | Skeletal muscle regeneration | Degradable, self-healing, low toxicity | Limited fibroblast viability, cytotoxic effects | 185 |
Conductive thin films | PGS, aniline pentamer | Nervous TE | Promotes Schwann cell myelin gene expression | Lacks bioactive cues, reduced in vivo conductivity | 202 |
Ionic EAP hydrogel | MAETAC modified PEGDA | Artificial muscles | Increased cell binding, capable of actuation | Binding of cations limits actuation speed | 140 |
Conductive thin films | PEDOT:PSS, P3HT, MEH-PPV | Neuronal TE | Increased fibroblast proliferation, stem cell differentiation | Not biodegradable, needs further cytotoxicity evaluation | 100 |
3D printed hydrogel | PEGDA | Soft robotics | Muscle force generation, controlled locomotion | Long-term viability not discussed | 170 |
Conductive nanofibrous film | PLA blending with PANi | Cardiac TE | Enhanced cell interaction, maturation, beating | PANi cytotoxic in high concentrations | 203 |
Hydrogel | PEDOT/Cs/Gel | Neural TE | Increased stem cell proliferation, enhanced neurite outgrowth | Non-degradable, limited adhesion specificity | 31 |
Ionic EAP film | PEO, NBR with EMITFSI | Artificial muscles | Linear actuators with 0.55% strain | Response time limits application | 101 |
Rolled non-woven mat | PLA with KNN nanowires | Neural TE | Output voltages from 0.52 to 17.9 V, promotes differentiation | High ultrasound pressure induces apoptosis | 149 |
Ionic hydrogel | PEGDA, AA with CMA | Muscle TE | Controllable actuation, improved compatibility | Acidic media from AA, reduced metabolic activity | 197 |
Non-woven mat | P(VDF-TrFE) | Neural TE | Enhanced differentiation, improved cell interactions | Difficult to scale, large output error | 204 and 205 |
4D porous | polyHIPE/PEDOT | Cell culture platform | Viable with fibroblasts, electromechanical stimulation | Non-degradable, required fibronectin coating | 29 |
Scaffold Format . | Material . | Application . | Main findings . | Limitations . | Refs. . |
---|---|---|---|---|---|
3D-printed macroporous | PCL with PEDOT | Bone regeneration | Conductive scaffold similar to PCL | No ES testing, no enhanced cell proliferation | 201 |
Conductive polymer films | Laminin coated PPy | Nerve TE | Enhanced neurite growth, good biocompatibility | Performance depends on dopant, no degradation data | 138 |
Biopolymer fibrous hydrogel | Collagen-actomyosin | Artificial muscles | Biomimetic actuator, repeated actuation | ATP activation may be limited, low mechanical strength | 171 |
Ionic nanofiber film | TEMPO-oxidized cellulose | Biopolymer actuator | Large displacements, fast response | High Young’s modulus, high activation voltage | 167 |
Shell/core nanofibrous film | PEDOT/CS | Nerve TE | Conductive, piezoelectric, increased neurite length | No degradation studies, potential toxicity | 139 |
Conductive hydrogel | QCS–PANi/Odex | Nerve TE | Increased cell proliferation, antibacterial properties | Potential toxicity, poor solubility, difficult degradation | 142 |
Electrode integrated fibrous patch | PCL:gelatin, gold electrodes | Cardiac patch | Functional cardiac patch, drug release | Long-term performance not assessed, risk of rejection | 119 |
Injectable hydrogel | Dex-AT/CECS | Skeletal muscle regeneration | Degradable, self-healing, low toxicity | Limited fibroblast viability, cytotoxic effects | 185 |
Conductive thin films | PGS, aniline pentamer | Nervous TE | Promotes Schwann cell myelin gene expression | Lacks bioactive cues, reduced in vivo conductivity | 202 |
Ionic EAP hydrogel | MAETAC modified PEGDA | Artificial muscles | Increased cell binding, capable of actuation | Binding of cations limits actuation speed | 140 |
Conductive thin films | PEDOT:PSS, P3HT, MEH-PPV | Neuronal TE | Increased fibroblast proliferation, stem cell differentiation | Not biodegradable, needs further cytotoxicity evaluation | 100 |
3D printed hydrogel | PEGDA | Soft robotics | Muscle force generation, controlled locomotion | Long-term viability not discussed | 170 |
Conductive nanofibrous film | PLA blending with PANi | Cardiac TE | Enhanced cell interaction, maturation, beating | PANi cytotoxic in high concentrations | 203 |
Hydrogel | PEDOT/Cs/Gel | Neural TE | Increased stem cell proliferation, enhanced neurite outgrowth | Non-degradable, limited adhesion specificity | 31 |
Ionic EAP film | PEO, NBR with EMITFSI | Artificial muscles | Linear actuators with 0.55% strain | Response time limits application | 101 |
Rolled non-woven mat | PLA with KNN nanowires | Neural TE | Output voltages from 0.52 to 17.9 V, promotes differentiation | High ultrasound pressure induces apoptosis | 149 |
Ionic hydrogel | PEGDA, AA with CMA | Muscle TE | Controllable actuation, improved compatibility | Acidic media from AA, reduced metabolic activity | 197 |
Non-woven mat | P(VDF-TrFE) | Neural TE | Enhanced differentiation, improved cell interactions | Difficult to scale, large output error | 204 and 205 |
4D porous | polyHIPE/PEDOT | Cell culture platform | Viable with fibroblasts, electromechanical stimulation | Non-degradable, required fibronectin coating | 29 |
The development of active TE constructs should build upon passive designs that have optimized structural, mechanical, and biological properties while avoiding the incorporation of EAPs that are inherently cytotoxic and non-degrading.25,206,207 The desired response must also be considered; while large electromechanical stimulation may be beneficial for cardiac engineering, it could damage neuronal tissue or hinder electrokinetic drug release profiles. Finally, the activation mechanisms of the chosen EAP (electronic, ionic) are critical in ensuring ideal robotic control but should also be operable in ranges feasible for clinical translation. These considerations have been laid out in a product design specification (PDS) as a guide for further research in Table V.
Product design specification for an electroactive tissue engineered implantable scaffold.
1. User . | |||
---|---|---|---|
ID . | Origin . | Requirements . | Advantage . |
1.1 | Must not harm or cause degenerative progression | Intended use of the device should not cause direct or indirect harm to patients or clinicians. | Improves product safety. |
1.2 | Ease of handling | Clinicians must be able to handle the scaffold without damaging the structure and function during implantation. | Ensure the device works as intended post implantation. |
2. Regulatory and statutory | |||
2.1 | Medical device regulations. Requirements for general safety and performance. | Test product through standard procedures for regulatory approval. EU: Medical device regulations, USA: Food and Drug Administration device regulation | Conforms with regulatory standards208 |
3. Technical/material | |||
3.1 | Degradable/resorbable | Degradation rate should match the rate of new ECM production. | Prevents the need for a secondary surgical procedure to remove the device. Allows seamless integration with native tissue. |
3.2 | Material derived from the native ECM | Materials used should already exist in the ECM. | Improves biochemical, mechanical, and structural similarities to healthy tissue. |
3.3 | Non-toxic | Does not directly harm cells and tissues or produce any compounds that may damage the patients throughout its entire lifetime. | Reduces risk of scaffold rejection with potential for integration into the native tissue. |
4. Performance | |||
4.1 | Mechanically stable under operating loads | Able to withstand tensile, compressive, and torsional loads such as those experienced in situ. | Prevents premature failure and further patient injury. |
4.2 | Biomimetic structural features | Fibers and pores (including their interconnectivity) should replicate that of native tissue. | Enhances cellular penetration depths |
4.3 | Biochemical sites for cell–scaffold interactions | Various sites can be suppressed or exposed in biomaterials for cell interactions. | Allows specific cells to anchor to the matrix. |
4.4 | Permeable | Highly porous scaffold with high volume of free space. | Allows nutrient exchange in the seeded cells. |
4.5 | High surface area to volume ratio | Exposed surface area should exceed total volume. | Supports cell proliferation and internal angiogenic processes. |
4.6 | Electromechanical transduction capabilities | Sensitive, tuned response to changes in applied endogenous and/or exogenous electrical signals, including the voltage and frequency values. | Allows directed stimulation for enhanced proliferation, differentiation, and maturation in vivo and/or in vitro. |
5. Manufacturing | |||
5.1 | Scalable manufacture | Scaffold should be able to be made in batches. | Improves potential for successful commercialization. |
5.2 | Low batch to batch variation | Scaffolds made under the same process should be able to function identically. | Reduces error between batches. |
5.3 | Ability to sterilize post processing. | Chemical sterilization methods should not damage the product. Otherwise, aseptic techniques must be used. | Removes or inactivates micro-organisms to a probability of micro-organism presence of , in line with the International Organization for Standardization requirements.209 |
1. User . | |||
---|---|---|---|
ID . | Origin . | Requirements . | Advantage . |
1.1 | Must not harm or cause degenerative progression | Intended use of the device should not cause direct or indirect harm to patients or clinicians. | Improves product safety. |
1.2 | Ease of handling | Clinicians must be able to handle the scaffold without damaging the structure and function during implantation. | Ensure the device works as intended post implantation. |
2. Regulatory and statutory | |||
2.1 | Medical device regulations. Requirements for general safety and performance. | Test product through standard procedures for regulatory approval. EU: Medical device regulations, USA: Food and Drug Administration device regulation | Conforms with regulatory standards208 |
3. Technical/material | |||
3.1 | Degradable/resorbable | Degradation rate should match the rate of new ECM production. | Prevents the need for a secondary surgical procedure to remove the device. Allows seamless integration with native tissue. |
3.2 | Material derived from the native ECM | Materials used should already exist in the ECM. | Improves biochemical, mechanical, and structural similarities to healthy tissue. |
3.3 | Non-toxic | Does not directly harm cells and tissues or produce any compounds that may damage the patients throughout its entire lifetime. | Reduces risk of scaffold rejection with potential for integration into the native tissue. |
4. Performance | |||
4.1 | Mechanically stable under operating loads | Able to withstand tensile, compressive, and torsional loads such as those experienced in situ. | Prevents premature failure and further patient injury. |
4.2 | Biomimetic structural features | Fibers and pores (including their interconnectivity) should replicate that of native tissue. | Enhances cellular penetration depths |
4.3 | Biochemical sites for cell–scaffold interactions | Various sites can be suppressed or exposed in biomaterials for cell interactions. | Allows specific cells to anchor to the matrix. |
4.4 | Permeable | Highly porous scaffold with high volume of free space. | Allows nutrient exchange in the seeded cells. |
4.5 | High surface area to volume ratio | Exposed surface area should exceed total volume. | Supports cell proliferation and internal angiogenic processes. |
4.6 | Electromechanical transduction capabilities | Sensitive, tuned response to changes in applied endogenous and/or exogenous electrical signals, including the voltage and frequency values. | Allows directed stimulation for enhanced proliferation, differentiation, and maturation in vivo and/or in vitro. |
5. Manufacturing | |||
5.1 | Scalable manufacture | Scaffold should be able to be made in batches. | Improves potential for successful commercialization. |
5.2 | Low batch to batch variation | Scaffolds made under the same process should be able to function identically. | Reduces error between batches. |
5.3 | Ability to sterilize post processing. | Chemical sterilization methods should not damage the product. Otherwise, aseptic techniques must be used. | Removes or inactivates micro-organisms to a probability of micro-organism presence of , in line with the International Organization for Standardization requirements.209 |
In summation, future research avenues should, therefore, address the following key challenges for the development of ECM derived electroactive scaffolds: (1) integration of electrical, mechanical, and biological signaling, (2) control of degradation, cell behavior, and material mechanics, (3) lack of post-implant tunability or control, and (4) cytotoxicity and biodegradability. By bridging the gap between ECM-derived materials and electroactive synthetic polymers, there is significant potential to combine ECM-mimetic properties (biological, chemical, electrical, and mechanical) with post-implant controlled responses to stimulation to create novel classes of clinically translatable dynamic tissue engineering scaffolds.
ACKNOWLEDGMENTS
MKB was funded by the EPSRC Doctoral Training Partnership Studentship 2886355 (Grant No. EP/W524311/1), and MN acknowledges support from the EPSRC BIONIC Hearts New Investigator (Award No. EP/Y004434/1). We would also like to acknowledge Victor Choi for his suggestions and review of the article.
AUTHOR DECLARATIONS
Conflict of Interest
The authors have no conflicts to disclose.
Author Contributions
Matthew K. Burgess: Conceptualization (supporting); Visualization (lead); Writing – original draft (lead); Writing – review & editing (supporting). Malavika Nair: Conceptualization (lead); Funding acquisition (lead); Supervision (lead); Visualization (supporting); Writing – original draft (supporting); Writing – review & editing (lead).
DATA AVAILABILITY
Data sharing is not applicable to this article as no new data were created or analyzed as part of this perspective. All synthesized values are directly cited from the original works they have been quoted from.