Peripheral nerve injury is a global clinical issue, significantly impacting the quality of life for patients and implying a substantial socioeconomic impact (Raza et al., 2020; Felder and Ducic, 2022). When peripheral nerves undergo a traumatic injury, a sequence of pathophysiological events occurs at the site of the nerve injury, where the axons undergo Wallerian degeneration and the remaining Schwann cells (SCs) create a favorable environment for nerve regrowth towards the target organ, by forming bands of Bungner and releasing neurotrophic factors and extracellular matrix molecules (Faroni et al., 2015). In recent years, increasing attention has been devoted to the possibility to repair and regenerate nerve tissues by adopting targeted biomedical nanotechnology and tissue engineering approaches (Silva, 2006; Li et al., 2011; Ding et al., 2015; dos Santos et al., 2021). In fact, although peripheral nerves can spontaneously heal after traumatic injuries, poor regeneration outcomes are observed when surgical end-to-end nerve sutures are needed (i.e., in the presence of nerve segment loss) due to tension at the nerve repair site. To overcome this problem, a surgical approach employing grafts (i.e., small portions of a nerve tissue used to fill the gap of the nerve stumps) and tissue engineering nerve conduits (artificial structures used to bridge the nerve defects) is preferred (Faroni et al., 2015). The nerve conduits connect the nerve stumps and provide physical guidance for the axons, guaranteeing the correct connections of sensory and motor fibers of the distal and proximal stumps and the fast reinnervation of the motor end plates of distal target organs, to minimize the muscle fiber atrophy (Zhang et al., 2019).

Traditional direct suturing involves fascicular or epineural repairs. However, if the distance between nerve stumps exceeds 5 mm, grafting or conduits are utilized. The autologous grafts (i.e., autogenous donor nerves harvested from other parts of the patient’s body) are the gold standard due to their nontoxicity, non-immunogenic effects and good biocompatibility, but together with the allografts (i.e., nerve grafts from a donor) show limitations in terms of functional recovery (Faroni et al., 2015). Donor site morbidity, size mismatch between the injured nerve and the available donor nerves and significant healing times are some of the disadvantages of autografts, while tissue rejection and disease transmission are often related to allografts (Rebowe et al., 2018).

To overcome these problems and satisfy the demand for high-performance nerve conduits, new promising alternatives have been proposed to heal the damaged nerve, including the development of biocompatible tissue engineered nerve conduits that mimic the structure of an autograft and provide enough support and mechanical strength while being flexible (Sarker et al., 2018) (Figure 1). The use of these nerve guidance conduits improves nerve regeneration by guiding an ordered axon outgrowth and reducing scar formation, allowing for nutrient and waste exchange through a porous structure (Faroni et al., 2015; Sarker et al., 2018). Ideally, the conduits should guide axonal growth towards the severed distal nerve by reducing axonal dispersion and off-target reinnervation and improving the neural biomechanical microenvironment following nerve injury. This could be achieved by including topographic and biophysical cues sensed by the cells via the cellular mechanotransduction (Kong et al., 2022). The incorporation of neurotrophic factors and support cells, such as SCs, stem cells and macrophages, is commonly used as neuroprotective therapy (Li et al., 2022b; Kim et al., 2022), while guiding cues, including intraluminal multi-channels, grooves in the inner wall and extracellular matrix-like structures are known to promote cell polarization and neurite/axon outgrowth (Chighizola et al., 2019; Scaccini et al., 2021). In addition, electrical stimulation (ES) is another method for accelerating the regeneration of injured peripheral nerves and enhancing their functional recovery, even in the presence of large nerve gaps (Huang et al., 2012).

Current clinically approved nerve guides are primarily made from synthetic or natural materials, including poly-glycolic acid (PGA), polycaprolactone (PCL), polyvinyl alcohol (PVA), type I collagen, chitosan and porcine small intestine submucosa (Rebowe et al., 2018; Fornasari et al., 2020). Among the materials that have been proposed for conduit production, silicon lacks long-term stability and biodegradability, while stimulating excessive scar tissue formation and requiring a secondary procedure for removal (Ciardelli and Chiono, 2006). An ideal conduit should be easily available and implantable and it should satisfy some of these requirements: be biocompatible; without inducing immunogenic reactions in the host tissue: be porous for nutrient and oxygen diffusion; be biodegradable to avoid a second surgery; be flexible and mechanically stable to support nerve regeneration without long-term compression; and be transparent to help the surgeons in prosthesis positioning by evaluating the alignment of the nerve stumps and to detect regenerated tissue in preliminary assessments without the need to cut open the conduit (Rebowe et al., 2018; Stocco et al., 2023). However, no materials can satisfy all these characteristics together: usually natural polymers, such as collagen and chitosan, show excellent biocompatibility, low immunogenicity and high bio-absorbability and sustainability, but lack adequate mechanical and electrical properties, while synthetic polymers are better in terms of versatility, mechanical properties and structural stability, but they have poor biodegradability and there are still concerns regarding toxic residual monomers from incomplete polymerization (Pinho et al., 2016; Boni et al., 2018). The synthetic polymers family includes also conductive polymers, such as polypyrrole (PPy), polyaniline (PANI) and polyethylenedioxythiophene (PEDOT), widely used to fabricate nerve conduits capable of influencing cell proliferation and axonal extension by appropriate ES (Guo and Ma, 2018).

Knowing that neurons are electrically excitable cells that transmit electrical signals, an electrically conductive material represents an ideal substrate for them. Indeed, it has been demonstrated that conductive materials can enhance the electric field produced by the cell membrane, influencing the bioelectric properties of the cells (Guo et al., 2016). ES can also improve and direct neurite outgrowth (Meng, 2014) and can promote axonal elongation (Fraczek-Szczypta, 2014). In recent years, electrically conductive materials have been proposed as alternative candidates for tissue engineering applications. Traditional conductive polymers, like Ppy, PEDOT and PANI exhibit favorable conductivity. However, they show often poor mechanical properties, low solubility in available solvents, and poor processability and biocompatibility (Kaur et al., 2015). Consequently, there has been a growing interest on conductive polymeric composites, realized by incorporating conductive fillers that form a conductive network though the composite while still maintaining the polymeric characteristics (Marsden et al., 2018). Among the fillers that have been utilized to increase the conductivity range of the biocompatible polymer networks, carbon-based materials such as carbon nanotubes (CNTs) and graphene, have attracted great attention, due to their good electrical conductivity, chemical stability and easy functionalization (Ding et al., 2015; Shin et al., 2016; Geetha Bai et al., 2018; Kumar et al., 2021; Qian et al., 2021; Hui et al., 2022; Bao et al., 2023).

Graphene is a monolayer of sp²-hybridized carbon atoms arranged in a two-dimensional honeycomb lattice, first isolated from graphite in 2004 (Novoselov et al., 2004). It is a single layer of graphite that, due to the dimensional confinement, shows record electrical and thermal conductivity (Balandin, 2011), broadband light absorption (Bao and Loh, 2012), and exceptional mechanical properties (i.e., hosts extreme mechanical resistance and high flexibility) (Papageorgiou et al., 2017). The ensemble of these appealing properties opens up to the potential adoption of graphene for a large number of applications, which in the biomedical field span from biosensing to drug delivery and bio-imaging (Zhang et al., 2012; Reina et al., 2017). Furthermore, thanks to its carbon-based chemistry, graphene has been often presented as an excellent candidate material for neural interfacing devices (Kostarelos et al., 2017; Bramini et al., 2018). To date, significant focus has been paid to the perspective adoption of graphene-based materials (GBMs) in the central nervous system (CNS). The main applications include cell labeling and real-time live-cell monitoring, biomolecules delivering through the blood-brain barrier and highly sensitive electrodes that combine stimulation and recording and enable optoelectronic stimulation (Kuzum et al., 2014; Park et al., 2014; Bramini et al., 2018; Rauti et al., 2019). Since the integration of graphene with central neurons was observed to promote controlled elongation of neuronal processes, thereby facilitating neuronal regeneration (Li et al., 2011), GBMs have also been integrated in tissue engineering scaffolds (Ding et al., 2015; Bramini et al., 2018; Bai et al., 2019). These studies have paved the way to new research avenues, and while the CNS keeps being a central research topic, ever-increasing attention is being devoted to the adoption of graphene in the peripheral nervous system (PNS).

This review focuses on graphene and graphene-based neural interfaces and devotes particular attention to the potential use of graphene as a material for PNS regeneration, discussing how the complex system of cells involved in nerve regeneration could be affected by graphene. To date the overview of the adoption of different graphene and GBMs for this application is still incomplete. In fact, other reviews existing on the topic focus on graphene covalent-functionalized forms such as graphene oxide and reduced graphene oxide, or graphene flakes used in form of composites to design 2D or 3D scaffolds without including in the discussion other graphene forms (Grijalvo and Díaz, 2021; Aleemardani et al., 2022; Hui et al., 2022). As a matter of fact, the potentiality of planar highly crystalline pristine graphene in nerve regeneration is still little considered, although it has been reported in several works how the use of alternative graphene forms could improve the regeneration outcome (Convertino et al., 2018; Convertino et al., 2022). Herein, after a general introduction to nerve guide conduits for peripheral nerve regeneration, we elucidate graphene and GBMs synthetic methods and discuss the most common biomedical applications of these materials, their impact in central and peripheral neural interfaces, with a critical outlook on their potential cytotoxicity. We then discuss the existing studies on graphene and GBMs in nerve conduits, with a special focus on the potentiality of highly crystalline pristine planar graphene (i.e., epitaxial graphene grown on silicon carbide and graphene grown via chemical vapor deposition on copper or insulators). Finally, we address the interaction of graphene with different non-neuronal cell types involved in nerve tissue regeneration, a crucial point to understand and predict graphene performance in vivo upon implementation in nerve conduits.

Graphene and GBMs can be produced via different methods, resulting in different materials in terms of size, shape, number of layers, purity, lateral dimensions and chemical modification that can influence their application (Table 1) (Kostarelos and Novoselov, 2014; Backes et al., 2020). Recently, within the European Union’s GRAPHENE Flagship project, it has been proposed a classification system based on three physical-chemical properties of GBMs: the number of graphene layers, the average lateral dimensions, and the carbon-to-oxygen atomic ratio (Wick et al., 2014). Single-layer graphene is only one atom thick, in contrast, multilayer consists of a small number (2–10) of stacked graphene layers, named few-layer graphene if the layer number is between 2 and 5. Multiple stacking with a thickness up to 100 nm are referred to as nano-graphite, to distinguish them from the conventional thicker flakes of graphite powders (Bianco et al., 2013). Following the isolation of a single layer of graphene in 2004 via mechanical exfoliation, many efforts have been made to achieve large-scale synthesis of graphene (Li et al., 2021). Among the most common top-down production methods, liquid phase exfoliation (LPE) of graphite and reduction of exfoliated graphite oxide are typically used to produce dispersed graphene flakes. LPE of graphite, performed by ultrasonicating (or high-shear mixing) powdered graphite in solvents, yields graphene flakes with controlled thickness (from 1 to 100 layers), controlled dimension (lateral sizes from tens of nanometers to tens of microns), and good crystalline quality within a dispersing liquid (Backes et al., 2020; Tyurnina et al., 2021). While the resulting so-called graphene inks hold great potential for the realization of advanced composite materials and in energy conversion and storage, in the biomedical field the main applications include biosensing and intracellular delivery of small therapeutic molecules and drugs (Bonaccorso et al., 2016; Chen et al., 2023). For ultrahigh-sensitive biosensors, high-quality 2D crystals are usually required. However, for less demanding applications, such as electrochemical detection of analytes, LPE graphene has demonstrated good performances in detecting ascorbic acid, dopamine and uric acid (Qi et al., 2015; Banavath et al., 2022). Recently, LPE graphene was shown to enhance enzyme delivery to fibroblasts derived from patients with lysosomal storage disorders, demonstrating its potential as carriers for enzyme replacement therapy (Chen et al., 2023). Graphene oxide (GO), obtained by exfoliating oxidated graphite, is usually preferred for biomedical applications thanks to the easy functionalization of the oxygenated species introduced during the oxidation process (Smith et al., 2019). In addition, the resulting flakes are strongly hydrophilic and can be easily dispersed in water, with a broad potential in the biomedical sector. It can be easily integrated in a polymeric matrix, increasing surface hydrophilicity, therefore improving cell adhesion at the biomaterial surface (Pinto et al., 2013). However, in GO nanosheets, the large fraction of sp³ carbon covalently bonded to oxygen to form epoxy and hydroxyl groups degrades the electronic performances of the material that is an electrical insulator (Du et al., 2013). The GO properties can be improved by removing the oxygen functional groups via different reduction methods to produce reduced graphene oxide (rGO), however its properties are still poor if compared with pristine graphene due to the remaining oxygen-containing functional groups (Smith et al., 2019). But despite that, the presence of the functional groups in GO and rGO represents a great opportunity to tune their chemical and electrical properties (Li et al., 2021). Typical biomedical applications for GO and rGO include: 1) drug and therapeutic gene and cellular delivery, that take advantage of graphene chemical stability, the large surface area to increase the load rate and the easily functionalised surface chemistry; 2) phototherapy, that exploits GO photothermal effect to kill undesired cells and tumors; 3) biosensing, for which rGO is usually preferred with respect to GO for its higher conductivity; 4) bioimaging, that uses graphene as contrast agents in fluorescence, photoacoustic and magnetic resonance imaging; 5) sensing for RNA/DNA, glucose or disease biomarkers; and 6) antibacterial activity via oxidation and membrane stress (Bramini et al., 2018; Geetha Bai et al., 2018; Yan et al., 2019; Karki et al., 2020; Li et al., 2021; Cellot et al., 2022). Furthermore, GBMs such as LPE graphene, GO and rGO have been used in form of composites, foams, fibers and hydrogels to design tissue engineering 3D scaffolds with enhanced electrical and mechanical properties, to better mimic the in vivo environment (Bramini et al., 2018; Vlăsceanu et al., 2019; Savchenko et al., 2021; Lyu et al., 2022).

Bottom-up synthesis of graphene, on the other end, allows one to obtain highly crystalline graphene on large (i.e., up to wafer) scale. To this end, thermal decomposition of silicon carbide (SiC) (Riedl et al., 2007), and chemical vapor deposition (CVD) of graphene on transition metals (Yu et al., 2008; Miseikis et al., 2015) as well as on insulators such as sapphire (Mishra et al., 2019; Li et al., 2022a) can be adopted. In the former case, graphene is formed after the sublimation of silicon atoms from the SiC crystal surface, whereas in the latter case, graphene is grown by depositing the precursor’s carbon atoms into a substrate.

Epitaxial graphene on SiC combines high crystalline quality, thickness homogeneity and an extreme cleanliness (Convertino et al., 2018). It can be used directly on the growth substrate that is known to be highly biocompatible, with no need to be transferred thus avoiding possible contaminants (Coletti et al., 2007; Oliveros et al., 2011; Convertino et al., 2022). Despite it has been demonstrated to be a valid interface for peripheral neuron integration, to date, epitaxial graphene on SiC remains an ideal platform for proof-of-concept investigations due to the high production costs and the non-transferability that impedes its integration with other substrates (Convertino et al., 2018).

Concerning transition metal substrates, nickel is usually preferred to grow multilayer graphene due to the high solubility of carbon in nickel (Yu et al., 2008), while the low solubility of carbon in copper allows the formation of a single layer graphene film (Li et al., 2009). To date, CVD monolayer graphene grown on copper has a quality comparable to the crystals obtained via mechanical exfoliation (Pezzini et al., 2020), and is largely adopted for applications that require high transparency and electrical conductivity including biosensing (Keisham et al., 2016; Singh et al., 2018), stem cell differentiation (Park et al., 2011) and neural electrodes (Kuzum et al., 2014; Park et al., 2014).

With respect to epitaxial graphene on SiC, the use of CVD graphene on copper holds potential for several applications, being the substrate cheaper and graphene transferreable to substrates of choice. Indeed, its flexibility and mechanical resistance as well as its excellent electrical conductivity have shown to yield positive effects on cell viability and axon elongation (Li et al., 2011; Lee et al., 2015; Convertino et al., 2020). To date, most of the studies that use planar graphene as a neural interface investigate its effect on CNS neurons. The use of CVD graphene, with or without a polymeric coating is actually preferred to realize 2D devices as planar transparent electrodes to stimulate and record neural activity. Indeed, recently graphene and CNTs have been successfully used to improve recording and ES of neurons (Keefer et al., 2008; Kuzum et al., 2014) and, surprisingly, neural microelectrode arrays (MEAs) fabricated using graphene obtained via CVD performed better than gold and indium tin oxide (ITO), in terms of signal-to-noise ratio (Rastegar et al., 2017). Park et al. developed a CVD graphene-based, carbon-layered electrode array device that was implanted in rodent brain for high-resolution neurophysiological recording. Thanks to graphene’s biocompatibility, transparency and flexibility, the device showed long-term in vivo stability and viability for optogenetic activation of focal cortical areas, electrophysiology and cortical imaging (Park et al., 2014). Kuzum and coauthors developed a transparent flexible neural electrode based on CVD graphene for simultaneous electrophysiological recording and optical imaging, with an improved signal-to-noise ratio and substantial reduction in electrical interference noise (Kuzum et al., 2014). The technique has been recently improved to integrate 2-photon microscopy, optogenetic, stimulation and cortical recordings in the same in vivo experiment (Thunemann et al., 2018), and it was used for multimodal monitoring of transplanted organoids for a comprehensive evaluation of the development, maturation and integration between the organoid and the host brain. Recently, CVD graphene has also been used to realize a solution-gated field-effect transistor that showed a performance similar to platinum black electrodes in recording visual and auditory responses in rats (Hébert et al., 2018), with long-term stability and biocompatibility in epicortical chronic implant (Garcia-Cortadella et al., 2021).

Park et al., 2011 exploited graphene as a transparent electrode, observing a good electrical coupling between graphene and neural stem cells (NSCs) for ES and confirming the neural activity of the differentiated cells. The resulting neural network functionality and the graphene effect on the maturation of the NSC’s electrophysiological state was also investigated by Guo et al., 2016. Graphene influenced both passive and active bioelectric membrane properties, hyperpolarizing the resting membrane potential and increasing the firing of action potentials during development, resulting in an accelerated maturation and enhanced neural performance (Guo et al., 2016).

The primary consideration when employing GBMs in biomedical applications is their biocompatibility. Indeed, the family of GBMs includes materials with widely variable properties that greatly influence the cell response (Wick et al., 2014). The layer number affects the elasticity and thus the adsorptive capacity of biomolecules that affect the cell interactions. The lateral size influences the cellular uptake, the blood-brain barrier crossing and the clearance from the body. The carbon-to-oxygen atomic ratio highly influence the structural properties and the surface chemistry, going from the highly hydrophobic pristine graphene to the GO that presents oxygen functionalities and hydrophilic regions (Wick et al., 2014). The following section investigates the effect of the production methods on the potential toxicity of GBMs to cell functions in neural interfaces.

Following peripheral nerve injury several cell types besides neurons get involved. As soon as neurons get injured, SCs de-differentiate owing to their lost connection with axons, starting a vigorous proliferation and acquiring a precursor-like “repair” state, which accounts for the secretion of several pro-inflammatory factors such as interleukin (IL)-1, IL-2, IL-6, and tumor necrosis factor α (TNF-α) in the extracellular space (Balakrishnan et al., 2021). This promotes the recruitment of macrophages (Li et al., 2022b) to the injury site, which together with repair SCs are responsible for degrading and removing degenerated axons and myelin debris, a necessary step for axonal re-growth (Balakrishnan et al., 2021; Endo et al., 2022). The injury site is also rapidly invaded by fibroblasts (Parrinello et al., 2010) and other immune cells, including neutrophils (Kennedy and Deleo, 2009). Indeed, the innate arm of the immune system plays a crucial role in peripheral nerve regeneration (Kalinski et al., 2020).

Of note, other more cell types may be involved in the interaction with the injury site, when a cell therapy (typically, a stem cell therapy) approach is used either alone or in combination with the nerve conduit approach (Sullivan et al., 2016). The most thoroughly investigated cells for these applications are bone marrow mesenchymal stem cells, neural stem cells, adipose stem cells, skin derived precursor stem cells and, more recently, induced pluripotent stem cells (Yi et al., 2020). In this respect, understanding the interaction of these cell types with graphene is crucially important to unravel and predict its performance in vivo upon implementation in nerve conduits. For example, it was recently reported that GO and rGO-coated scaffolds, eventually coupled to ES, could positively affect SCs behavior by promoting the repair phenotype and by enhancing migration, proliferation, and myelination ability (Wang et al., 2019b; Fabbri et al., 2021; Aleemardani et al., 2022) as well as adipose mesenchymal stromal cells differentiation to SCs (Llewellyn et al., 2021).

GBMs were also shown to influence immune cells. Indeed, neutrophils were shown to recognize some carbon-based materials as pathogens, capturing and digesting GO and CNTs (Keshavan et al., 2019). Furthermore, GO and rGO were analyzed for their ability to influence macrophage response, and it was found that upon phagocytosis by the immune cells, they could decrease both oxidative stress and proinflammatory cytokine secretion, thus potentially boosting an appropriate immune response for tissue regeneration (Cicuéndez et al., 2021). Similar properties were ascribed to highly crystalline graphene particles (30–160 nm) derived by fragmenting CVD graphene grown on a nickel foam and phagocyted by human macrophages (Povo-Retana et al., 2021). The macrophage phagocytosis was confirmed also for pristine graphene flakes with an average size of 500 nm (Mcintyre et al., 2016). Concerning fibroblasts, monolayer CVD graphene was found to be non-toxic and to allow migration of murine subcutaneous connective tissue cells similar to controls (Lasocka et al., 2018). However, nano and micro sized GO were found to negatively impact embryonic fibroblast cell cycle and viability, underlying again the diverging performances that different graphene forms, with varying flake size, concentration and surface chemistry, can have (Liao et al., 2011; Hashemi et al., 2020). Finally, graphene was also studied for its ability to promote neural differentiation and affect the behavior of stem cells in the context of peripheral regeneration, as extensively reviewed elsewhere (Kenry et al., 2018; Yao et al., 2021; Hui et al., 2022). Conductive electroactive biomimetic scaffolds realized by combining GO with electrospun nanofibers were found to promote SC proliferation, migration and myelination, and to induce in vivo sciatic nerve repair with fiber regeneration similarly to the gold standard autograft (Wang et al., 2019b). Interestingly, graphene-based nerve conduits loaded with stem cell derived extracellular vesicles, i.e., biological nanoparticles displaying parental-cells-like pro-regenerative properties, has yielded promising results in an in vivo evaluation of peripheral nerve injury, while avoiding the issues of a cell therapy approach (Zhang et al., 2023). Overall, the interaction of GBMs with all the numerous cell types involved in peripheral nerve injury and regeneration remains largely unexplored, while it constitutes a crucial aspect to consider to avoid undesirable effects in designing graphene-based conduits. Further research is required in this direction to fully explore the implications of using graphene and GBMs-to achieve an optimal pro-regenerative performance.

In the field of neural regeneration, the evaluation of the cell’s response remains a relevant aspect. 3D scaffold should mimic the outer microenvironment, having physical, biological and biochemical features that favor cells proliferation, migration and differentiation. The majority of the FDA-approved conduits are made of bioresorbable polymeric hollow tube, that limits their repair efficiency to lesions less than 3 cm long (Meena et al., 2021). Despite the limitations of using hollow tubes, they still have a wide clinical acceptance when compared to most complex geometries. However, the use of innovative lumen wall designs, the addition of inner fillers and topological modifications, and the surface functionalization through conductive coatings have greatly improved the regenerative outcome. Among the conductive coatings that can be used to stimulate, direct and accelerate peripheral nerve regeneration, graphene and GBMs are bringing new perspectives thanks to their excellent conductivity and mechanical strength (Parker et al., 2021). To date, graphene oxide and graphene covalent-functionalized forms are preferred to realize three-dimensional scaffolds for in vivo nerve regeneration. They have been applied in polymeric composites, following the conduit topographical cues and allowing the scaffold loading with molecules, growth factors and stem cell, thus demonstrating a good integration with the current technologies (Jakus et al., 2015; Qian et al., 2018b; Qian et al., 2018a; Wang et al., 2019a; Wang et al., 2019b; Lu et al., 2023; Zhang et al., 2023). Yet, due to the possible cytotoxic effect of graphene nanosheets on nerve cells, further long-term investigations are needed (Fadeel et al., 2018). Planar graphene–such as graphene obtained via chemical vapor deposition and by thermal decomposition of SiC–has shown to have reduced cytotoxic effects (Fabbro et al., 2016; Veliev et al., 2016; Convertino et al., 2022). Indeed, CVD graphene has been successfully adopted for the realization of in vivo planar electrodes, which have been used for neural imaging and optogenetic applications (Kuzum et al., 2014; Park et al., 2014; Thunemann et al., 2018). Although conductive electrodes are the main field of application for planar graphene, the possibility to use CVD graphene to design implants should not be excluded. Indeed, CVD graphene has been reported to have a positive effect on axonal outgrowth of peripheral neurons, a result which suggests its possible use as an active conductive substrate for nerve regeneration (Convertino et al., 2020). The integration of CVD graphene with commercially available polymeric conduits adopted in neural regeneration might ultimately improve their electrical conductivity, while preserving the optimized biocompatibility and mechanical properties of the polymer. However, also for CVD graphene possible foreign-body reactions, long term degradation and bioresorption need to be carefully analyzed when considering in vivo applications. As discussed in this review, different graphene forms differently affect the neuronal and non-neuronal cells involved in the regeneration process (Kostarelos and Novoselov, 2014). For this reason, additional systematic research is needed in the coming years to truly unveil the effective potential of graphene and GBMs for the realization of novel nerve conduits meeting the diverse and complex requirements emerging at the nerve injured site.