Neural repair and regeneration interfaces: a comprehensive review

Since the pioneering research by Shirakawa, Heeger, and MacDiarmid on bromine-doped polyacetylene in 2000, conductive polymers (CPs) such as polyacetylene, polypyrrole (PPy), poly(3,4-ethylenedioxythiophene) (PEDOT), polythiophene (PTH), and polyaniline (PANI), have been a focal point [13–15]. These materials, initially semiconductors or insulators without charge carriers, can be transformed into highly conductive materials with mobile charge carriers through a doping process involving oxidation or reduction of the polymer's conjugated structure [16]. CPs, while organic, conduct electricity due to the presence of conjugated single and double bonds along the polymer skeleton, enabling electron flow. CPs achieve high conductivities via doping, a process that involves redox reactions, resulting in charge transfer and subsequent formation of charge carriers such as solitons, polarons, and bipolarons. Doping can tune the density and mobility of these charge carriers, allowing the conversion from insulating to metallic conductivities, differing fundamentally from inorganic counterparts' doping mechanisms [17]. The rate of electron transfer at the electrolyte-polymer electrode interface is governed by the density of percolation paths within the material, indicating that the internal electronic transport in these polymers strongly correlates with the electron transfer rate to the molecules in the solution, thus significantly enhancing the efficiency of electrochemical technologies [18, 19]. The ion–electron transport capabilities of doped CPs allow them to respond directly to changes in ion concentration in biological media by integrating directly with ion channels and ion pump systems within organisms [20].

After early validations, conjugated polymers have shown excellent biocompatibility, promoting communication, growth, and regeneration of nerve cells, the proposed effect has been tested on PC12, sciatic nerve explants, and neuro2a neuroblastoma cells etc [21–23]. In 2004 and 2005, conjugated polymer electrodes were used for electrical stimulation of peripheral nervous system [24], and the central nervous system [25], leading to nerve repair. Subsequently, specialized CP composites were developed for various applications, including nerve regeneration using biodegradable composite materials, improving cochlear implant interfaces, and exploring biostability and biocompatibility [26–28]. Carbon nanomaterials are also introduced as dopants to improve the conductivity of various conducting polymer electrodes for neural interfaces [29–32]. Intrinsically biodegradability was also achieved for sciatic nerve regeneration, showing no inflammation after eight weeks of implantation [33]. These CPs often face limitations in conductivity. Therefore, degradable materials with attached CPs may offer greater potential for balancing degradability and conductivity [34].

Further research has focused on synthesizing electrodes in vivo to mimic the structure and morphology of target neuronal systems for seamless electro-neural interfaces. In vivo polymerization of PEDOT using iron chloride resulted in electrodes integrated with acellular muscle tissue structures [35]. Similar approaches have created PEDOT cloud electrodes that penetrate brain tissue without significantly affecting neural activity [36]. Enzymatic reactions have facilitated gene-targeted polymerization and hold the potential to form insulating structures in the future [37, 38]. A cocktail formulation containing hydrogen peroxide, surfactants, and trimmers has achieved enzyme-independent, multi-tissue adaptive in vivo polymerization with reduced primer toxicity [39].

ES promotes neural repair by orchestrating a complex and multifaceted suite of molecular mechanisms. At the cellular level, ES emulates the intracellular calcium fluctuations that transpire post-axonal rupture [40, 41]. These calcium waves are propagated in a retrograde manner to the soma of the neuron, thereby instigating inherent regenerative processes [42]. Subsequently, a calcium-dependent upregulation of brain-derived neurotrophic factor (BDNF) and its cognate receptor, tyrosine receptor kinase B (trkB), is elicited [43]. Simultaneously, the cyclic adenosine monophosphate (cAMP) pathway augments the expression of critical regeneration associated genes, including Tα1 tubulin and growth associated protein-43 [44, 45].

Moreover, ES demonstrates the capacity to escalate intraneuronal cAMP levels, thereby fostering the growth of dorsal root ganglion neurons [46]. In the wake of this, ES provokes cAMP to stimulate the protein kinase A-mediated phosphorylation of the cAMP response element-binding protein [47]. These series of events subsequently activate downstream signaling pathways, leading to the inhibition of the Rho protein within the p75-Nogo receptor pathway, bolstering cytoskeletal assembly, enhancing BDNF expression, and promoting neuronal growth [48]. Furthermore, BDNF, by inhibiting phosphodiesterase activity, forestalls the degradation of cAMP, thereby preserving elevated levels of cAMP [49]. ES also modulates intracellular calcium dynamics within Schwann cells (SCs), thereby triggering the upregulation of neurotrophic factors, such as nerve growth factor, Neurotrophin-3, and BDNF, advancing neural regeneration and migration via the Mitogen-activated protein kinase pathway [50]. The collective influence of these underlying mechanisms encapsulates the profound and transformative impact of electrical stimulation on neural repair.

Electric field stimulation (EFS) also plays a crucial role in influencing the direction and velocity of neural cell migration [51]. Constant EFS has been demonstrated to promote the migration of various cell types toward the cathode [52–54], exhibiting field intensity-dependent properties [55]. Additionally, a constant EF can also perpendicularly align neural stem cells (NSC) to the EF vector [56], consistent with the results observed in primary rat brain cells, such as astrocytes [57] and neurons [58]. This may be due to the formation of a relatively stable ion distribution on the cell membrane and the influence of electrophoresis and electroosmosis on the movement of membrane receptors [59]. Moreover, EF can enhance NSC neural differentiation and maturation [60, 61]. Based on the aforementioned principles, EF is employed to guide NSC migration to injured brain regions to replace lost cells, serving as a practical therapeutic strategy for brain damage [62].

In addition to stimulation, drug delivery systems show promise in neural regulation and repair. Various drug delivery systems have been integrated into neural interfaces, primarily including micro-electro-mechanical systems (MEMS) [63], redox-induced release systems [64], and electrophoresis systems [65]. MEMS demonstrate the potential for achieving multi-component drug delivery. However, the delivery strategies pertaining to MEMS are not within the scope of this review [66].

In the doping process of CPS, dopant ions bearing opposite charges to the conjugated chains are absorbed. This process augments the π–π interactions between the conjugated chains, thereby increasing conductivity [67]. Stemming from this principle, if the dopant ions are drug ions, the redox doping process can be utilized as a drug delivery system, wherein the drugs are released into the solution during the dedoping process. This controlled release can achieve high temporal precision through the application of voltage pulses to regulate the on-demand release of bioactive compounds, hence realizing precise and localized drug delivery systems [64]. Given the high temporal resolution of electrophysiological signals, closed-loop drug delivery systems hold promise in addressing numerous challenging diseases characterized by abnormal neural discharges, such as epilepsy [68].

One drawback of conducting polymers, however, is their low drug load capacity. In an effort to enhance the drug-loading capacity of CPs, CP composites have been developed. Nanoparticles like graphene [69, 70], and carbon nanotubes [71, 72] are frequently employed in the drug delivery field due to their abundant surface functional groups, controllable surface charge, and large relative surface area (figures 2(a) and (b)). Conductive polymer composites combine the ionic response capability of CPs with the superior drug-loading capacity of nanoparticles.

Zoom In Zoom Out Reset image size

Leveraging the distinctive ionic conductivity in organic electronics, researchers have developed ion-conducting drug delivery systems known as organic electronic ion pumps (OEIPs). These pumps utilize ion exchange membranes and electrophoretic transport to deliver ions with high precision and minimal liquid transport, enabling accurate control of delivery rates, similar to synaptic signaling [75]. OEIPs were first used to induce Aβ protein aggregation in Alzheimer's disease and for neuromodulation [76, 77]. A system integrating ion pumps and electrical signals managed to regulate and record epileptic activity in mouse hippocampus [78]. Microfluidic system-integrated ion pumps (µFIP) facilitate localized drug delivery to the brain and rapid gamma-aminobutyric acid (GABA) transport at low voltages [79]. A cutting-edge brain implant device merges a µFIP with an electrocorticography apparatus, providing on-demand therapeutic drug delivery and concurrent local neural activity recording [4, 73] (figure 2(c)). A primary challenge, however, lies in restricting drug leakage during idle states. Researchers have incorporated proton capture electrodes into OEIPs, bolstering selective electrophoretic transport and minimizing leakage while negligibly affecting active drug transport rates [74, 80] (figure 2(d)).

Recent advancements in drug delivery methods for neural repair and regeneration have the potential to transform clinical applications significantly. These methods can be categorized into two representative classes: nanoparticle-related drug delivery and device-based drug delivery.

In the realm of nanoparticle-related drug delivery, Chen et al [81] developed a multilayered nerve guidance conduit (NGC) loaded with melatonin and Fe₃O₄ magnetic nanoparticles. This NGC allows sequential and sustained drug release, thus enhancing nerve regeneration and functional recovery in both in vitro and in vivo experiments. This advancement suggests potential as a promising treatment for long-term nerve defects. Another significant development in this class was made by Tao et al [82] who introduced a functional nanoparticle-enhanced nerve conduit. This conduit is composed of gelatin-methacryloyl hydrogels with drug-loaded poly(ethylene glycol)-poly(3-caprolactone) nanoparticles, fabricated using a continuous 3D printing process. It efficiently promotes peripheral nerve regeneration by providing a physical microenvironment for axonal elongation and sustained drug release of a Hippo pathway inhibitor. The process leads to improved proliferation and migration of SCs and up-regulation of neurotrophic factors genes, thus holding promising potential for clinical application in peripheral nerve repair.

In the class of device-based drug delivery, Manoukian et al [83] optimized a NGC with aligned microchannels. This device uses a biodegradable chitosan structure reinforced with drug-loaded halloysite nanotubes, for sustained release of 4-Aminopyridine as a growth factor alternative, showing promising potential for nerve repair and regeneration in preclinical studies. Similarly, Wang et al [3] devised a novel method for peripheral nerve regeneration, employing a fully biodegradable, self-electrified, miniaturized device composed of dissolvable galvanic cells on a biodegradable scaffold. This innovation offers both structural guidance and electrical cues for nerve regeneration, leading to successful motor functional recovery in rodent models. Finally, Fadia et al [84] developed biodegradable polymer scaffolds that release glial cell line-derived neurotrophic factor from embedded double-walled polymeric microspheres. As nerve conduits for large nerve gaps in nonhuman primates, these scaffolds have shown improved nerve conduction velocity, increased SC recruitment, and functional recovery comparable to autograft treatment, thus providing a potential alternative for peripheral nerve regeneration and repair. Together, these advancements illustrate the remarkable potential of innovative drug delivery methods for neural repair and regeneration.

Poly(3-hexylthiophene) (P3HT), an organic photoactive material, has been extensively investigated for neural stimulation interfaces due to its wide applications in photovoltaic polymers, facile solution processing, and high absorption coefficient [23, 87–89]. Injectable P3HT has successfully generated neuronal action potentials, cell depolarization, and light-stimulated retinal tissues [90]. In the subsequent strides of organic photoactive material-based devices, organic electrolytic photocapacitors (OEPC) have emerged prominently. OEPCs, fundamentally, are devices that exploit the photo-induced reactions occurring at the interface of a semiconductor material and an electrolyte. Upon exposure to light, the semiconductor material in the OEPC absorbs photons and generates electron-hole pairs. These charges then migrate to the interface, where reactions with the electrolyte produce or consume ions, leading to an ionic current in the solution. This mechanism allows OEPCs to convert light pulses directly into ionic currents, establishing them as promising tools for neural stimulation. OEPCs have demonstrated outstanding biocompatibility in vitro and in vivo stimulation tests with oocyte cells [91], and retinal tissues [92].

OEPCs can be applied for in vivo percutaneous stimulation of peripheral nerves [85] (figure 3(a)). Flexible OEPC stimulators have been integrated into zipper-locking mechanisms, enabling long-term neural stimulation [2] (figure 3(b)). Ultra-thin OEPCs have been directly applied to the cortical surface, successfully activating the vibrissal motor cortex of mice and eliciting behavioral and electrophysiological responses in the sensory cortex [93]. A multi-electrode array and two wireless organic photovoltaic electrodes were designed to activate the vagus nerve in the mouse's neck. This study achieved real-time heart rate monitoring and reduced heart rate by activating the vagus nerve, circumventing the limitations of traditional electrodes that require insertion into the body [94]. OEIPs were also integrated with OEPCs to achieve on-demand and local delivery of pharmacologically active species, including alkali ions and neurotransmitters [95].

Zoom In Zoom Out Reset image size

Optogenetic interfaces, compared to electrophysiology, possess higher spatial resolution, even capable of targeting and regulating the physiological activities of individual neurons [97, 98]. However, for specific neural circuit analysis and nerve regeneration, laser-based systems lack portability and motion compatibility. Therefore, the portability of optogenetic devices has become a major focus in design.

Fiber optics is an important branch in this field. Light transmitted through fiber optics to specific nerves has achieved activation of the vagus nerve in live animals and suppressed systemic inflammation [99]. Inspired by patch-clamp techniques, multifunctional optical fibers with electrolyte microfluidic channels have been developed, achieving simultaneous optical and electrical stimulation, which is of great significance for closed-loop control of neural function reconstruction [100]. However, because fiber optic devices always rely on external light sources, they are not suitable for free-moving applications. In addition, due to the mismatch of the light modulus with biological tissues, normal physiological movements can easily cause repeated damage near the implanted items, accelerating the formation of scar tissue and eventually leading to device failure.

Recently, micro-light-emitting diode (µLED) has been used to manufacture highly integrated optogenetic devices [101]. Because small devices can be directly implanted into target tissues, efficient light transmission and high-resolution control can be achieved through external power devices. By modifying the design parameters of µLED probes, such as size and substrate, they can be adapted to the needs of different application scenarios. Multisite controllers for different brain areas have been developed and have achieved control over limb movements [102]. For the motor function impairment and reconstruction difficulties caused by spinal cord injury, flexible optogenetic devices have been used to implant into the spinal cord and intervertebral space, assisting in deciphering various neuronal subtypes, sensory pathways, and supraspinal projections in spinal-cord injured mice [86] (figure 3(c)). For the stomach, optogenetic devices can be stably fixed at the target location, resist cyclic stress caused by peristalsis, and regulate malfunctioning neural circuits, achieving control over eating [96](figure 3(d)).

Both electrophysiological and optogenetic interfaces applied in neural modulation require flexibility and stretchability. This is because the tissues that these interfaces often interact with, such as the brain, nerves, skin, and muscles, typically have a low Young's modulus. Consequently, to optimize tissue integration, the development of flexible and stretchable interface techniques has become an essential pursuit in the field. Viscoelasticity and stretchability are particularly important for achieving long-term stable perineural or spinal interfaces [103, 104]. During the chronic implantation process of rigid electrode materials, discomfort in the post-operative state often incites animals to repeatedly collide with their skull, specifically at the site where the electrode is implanted. The relative displacement and microvibrations between the skull and the brain continually stimulate the cells surrounding the neural electrode. As a result, over prolonged implantation periods, rigid electrodes can inflict severe damage, irrespective of whether it manifests as scar formation or neural tissue damage. On the contrary, the flexible electrode's wiring can effectively prevent the transmission of mechanical motion from the skull fixation site to the electrode's forefront, thereby significantly mitigating these injuries [105, 106]. To achieve the flexibility and stretchability of neural interfaces, three strategies are commonly used: hydrogels, material thinning, and elastomers.

The conductive hydrogel was developed based on the swelling behavior of CPs. The earliest record found that adding 5% ethylene glycol to PEDOT:PSS (polystyrene sulfonate) can promote swelling and reduce the elastic modulus to 100 MPa [107]. Subsequently, various additives were developed to reduce Young's modulus of PEDOT:PSS film and improve its elasticity, such as polyethylene glycol [108]. Recently, a formula for electrically conductive hydrogels was used to achieve 90% water content and matched tissue elastic modulus [109, 110]. Through 3D printing, these materials can be patterned into high-density flexible electronic circuits and soft neural probes [111] (figure 4(a)). A stretchable neuromorphic implant, utilizing low power and consisting of hydrogel electrodes and a stretchable transistor, has demonstrated promising results in restoring coordinated leg motions in mice with neurological motor disorders, offering a potential advancement in neurorehabilitation [5].

Zoom In Zoom Out Reset image size

By diminishing the thickness of polymers such as Parylene, polyimide, and SU-8 to a few micrometers, a congruity with the bending stiffness of tissue can be achieved. Williamson et al developed a 4-micrometer-thick styrene probe with gold conductors, PEDOT: PSS electrodes, and electrochemical transistors. After removing the rigid shuttle post-implantation that helps with insertion, the implant stimulated local neuronal groups, and shows no obvious glial scar formation [115]. Recently, analogous designs have been employed to fabricate mesh neural electrodes. By joining the shuttle and mesh electrode with soluble Dextran, the fully deployed tissue-level flexible mesh electronics minimize damage during the implantation process. This innovative approach has facilitated long-term implantation throughout the entire lifespan of a mouse, with no discernable tissue damage [112] (figure 4(b)). Inkjet printers can directly pattern polymer substrates into neural regeneration cuff electrodes with PEDOT: PSS and heat-shrinkable polymers to regenerate motor axons and incite muscle responses three months post-implantation [116]. A miniaturized, wireless, and bioresorbable electrotherapy system has been demonstrated to effectively accelerate wound closure, particularly in chronic cases associated with diabetes, by guiding epithelial migration, modulating inflammation, and promoting vasculogenesis, offering a practical solution for electrical stimulation therapy [6].

Flexible and stretchable conductive elastomers have also enhanced tissue stability in neural interfaces [117, 118]. Stretchable CPs can be achieved by customizing material structures and combining them with water swelling or by forming CP-elastomer composites [119]. Plasma activation is commonly used to improve adhesion problems caused by the low surface energy of flexible elastomers [120]. Viscoelastic flexible elastomers have achieved expandable electrodes that can grow with the surrounding tissue and have been implanted around the sciatic nerve of growing rats for long-term stimulation [1] (figure 4(c)). Recently, a molecular engineering strategy utilizing topological supramolecular networks has been developed to achieve high mechanical strength and good conductivity in elastomers, to control organ-specific activities through the delicate brainstem [113] (figure 4(d)). Stretching during continuous motion can often affect the electrical performance of elastomers, which can potentially affect the effectiveness of stimulation. Therefore, a series of motion-insensitive flexible electrodes have been developed [121, 122]. The examples demonstrate the benefits of soft and stretchable CP electrodes for neural modulation; however, a limiting factor is the relatively low conductivity of these materials. A development trend in the field is to combine high-performance conductive materials that are insensitive to stretching, such as aligned nanowires and nanosheets, with stretchable polymers to achieve high-performance stretchable devices and high-performance devices for the next generation of neural modulation interfaces [114, 123, 124] (figure 4(e)).

Biohybrid neural interface devices have been developed due to a deeper understanding of tissue engineering. Biohybrid neural interfaces may also integrate electrophysiological and optical neural interfaces. Still, their development is focused on optimizing electrode materials through tissue engineering to primarily meet the growth needs of cells and tissues. Additionally, they can directly provide cells for tissue repair to restore damaged tissues and optimize the immune rejection response of the implant and surrounding tissues [125, 126]. Biohybrid neural interface devices are in a very early stage of development, with the main unresolved issues being quality control and ethical limitations on exogenous implanted cells in clinical applications.

Metal electrodes and conductive hydrogels compose a hybrid neural interface that supports the growth of neural progenitor cells and glial cells [127]. Through experimental evaluations, hybrid neural interface structures exhibit lower impedance, higher charge storage capacity, and higher charge injection limits. In other experiments, researchers studied the effects of hybrid neural interfaces on the extracellular matrix, finding that they promote the uniform distribution of collagen and adhesive proteins [128]. Hybridization of common Si probe electrodes with nucleus pulposus cells in the mouse cortex induced differentiation of NSC and significantly reduced foreign body reactions [129]. Hybrid interfaces carrying NSC exhibited good cell viability and could differentiate into neurons and glial cells in vivo. NSCs secrete neurotrophic factors that promote regeneration of damaged nerves around the implant site and reduce the formation of glial scars [130, 131]. Hybrid devices can achieve synaptic connections between loaded cells and implant site cells. Such hybrid synapses can enhance the modulatory capacity of neural interfaces through high-specificity and high-density information transmission [132]. Hydrogel-based hybrid interfaces have been shown in vitro to support the neurite growth of glutamatergic, dopaminergic, and GABAergic neuronal subtypes and form sub-millimeter and centimeter synaptic connections [133]. Hybrid interfaces have loaded GFP-modified cortical neurons and formed synaptic connections between the rat cortex and thalamus in vivo. Histological examination revealed the presence of presynaptic protein synaptophysin near the transplanted neurons, confirming synapse formation with host neurons [134].

Hybrid interfaces for peripheral nerve repair can be either cellular or acellular. The loaded cells primarily include macrophages, SCs, and NSC. Cell selection targets different physiological processes during the injury repair process. In the first five days after injury, macrophages and SCs clean up the axons and myelin debris, leaving empty endoneurium tubes [135]. Later, SCs proliferate and transdifferentiate into specialized repair cells, filling the empty endoneurium tubes and forming characteristic bungner bands or tubes. Simultaneously, macrophages are recruited nearby, together with SCs, secreting various neurotrophic factors, which support axonal sprouting and growth towards injured neurons, or induce the differentiation of stem cells into neuron-like cells [136–138]. Ideally, cellular hybrid interfaces can intervene in the injury process and promote nerve repair by modulating abnormal immune responses [3, 139–141].

Optogenetics has been utilized in the development of tissue-engineered neural interfaces for therapeutic purposes that involve ex vivo cell engineering, where cells are altered to express specific light-responsive proteins before being transplanted into host organisms [142, 143]. Notably, long-term cell replacement enabled by transgenic stem cells offers the potential to eliminate the requirement for ongoing drug administration [144]. Moreover, in the case of human embryonic stem cell (ESC) (hESC)-derived midbrain dopaminergic neurons expressing halorhodopsin (HR) that are transplanted into mice with motor impairments, green light activation allows for control of the graft's neural activity and dopamine release, subsequently affecting the connectivity between transplanted neurons and host cells [145]. Optogenetic devices typically entail related cellular modification techniques. Encapsulating cells in protective hydrogels can improve cell survival, as demonstrated by the co-loading of genetically engineered non-neuronal human kidney cells and LED light sources into implantable hydrogels [146]. Shao et al developed a smartphone-assisted, semi-automatic treatment for diabetes in mice by engineering cells with optogenetic circuits to release glucose-lowering hormones in response to far-red light. The implanted hydrogel capsules contained both the engineered cells and wirelessly powered far-red light LEDs, which allowed a smartphone to remotely control hormone production and maintain glucose homeostasis in diabetic mice over several weeks [147]. The clinical application of optogenetic devices combined with transplanted cells is progressing, especially in the eye and brain [148, 149].