The first discovered and reported fluorescent CDs were prepared using the top-down method. The method is employed in the presence of water vapor, using argon as a protective gas, and the laser ablation of graphite is performed to synthesize carbon dots after surface passivation and acid oxidation. Since then, researchers have developed top-down methods by reducing the structure of graphene, graphene or Go sheets, and CNTs to achieve a perfect sp² hybrid structure. For the top-down approach, scientists have developed a number of methods to fracture carbon structures into CDs, such as arc discharge, laser etching, nano-photolithography, electrochemical oxidation, hydrothermal or solvothermal, microwave-assisted, ultrasound-assisted, chemical stripping and nitric/sulfuric acid oxidation. These synthetic methods are always complex and uncontrollable, with relatively low yields and quantum yields, and some are even environmentally harmful, making them unsuitable for the mass production of CDs with high fluorescent quantum yields.

In contrast, bottom-up methods typically use designed precursors and preparation methods to control CDs with good molecular weight and properties, shape and size. At the same time, the bottom-up method is low-cost and can mass-produce fluorescent CDs, which is the prerequisite to ensure the practical application of CDs. Hydrothermal treatment of citric acid or carbohydrates, laser irradiation of toluene, stepwise solution chemistry of benzene derivatives and precursor thermolysis have been used to successfully prepare carbon points. In addition to the above precursors, glycerol, amino acids, ascorbic acid and other molecules rich in carboxyl, hydroxyl and amino groups are suitable carbon precursors. They are dehydrated at high temperatures and further carbonized to form carbon spots.Yang and co-workers synthesized nitrogen-doped carbon sites with about 80% quantum yield via hydrothermal treatment of citric acid and ethylenediamine (Zhu et al., 2013). In addition, the luminescence color and quantum yield of CDs can be regulated by adjusting the amount of auxiliary inorganic substrates such as KH₂PO₄, NaOH and H₃PO₄, or the ratio of reagents. For example, Bhunia et al. (2013) used different carbohydrates and dehydrating agents as carbon precursors to synthesize visible-light-tunable high-fluorescence CDs at different temperatures. In addition, nucleation and growth kinetics were controlled by changing the dehydrating agent and reaction temperature. This synthesis method can control particle sizes of less than 10 nm and can synthesize CDs with adjustable emission and relatively high fluorescence quantum yield.

Surface modification is a prerequisite for the further biomedical applications of CDs. Using different surfactants, the solubility of CDs in non-polar solvents can be enhanced, while their fluorescence properties can be adjusted. CDs with different groups on their surfaces can be synthesized to fulfill different functional requirements. Therefore, suitable biodegradable polymer precursors with different group components, such as amines, carboxylic acids and hydroxyl, can be selected to modify CDs. For example, the carboxyl groups on the CD surface make it hydrophilic during synthesis and can be mixed with polyethylene-glycol-containing amino groups to generate functionalized CDs. Thus, proper surface modification is beneficial to the further application and development of CDs in the biomedical field before biological application.

Various covalent-bond modification methods have been used in the functionalization of CNTs, among which oxidation reaction is the most important modification method. The oxidation of CNTs mainly uses nitric acid as oxidant. In this oxidation process, carboxyl groups appear at the ends and defect positions of CNTs, which makes CNTs water-soluble to a certain extent. However, the oxidized CNTs accumulate and precipitate in salt solution, which cannot be directly used in living systems (because most living systems contain high-salt solution). Further surface modification is necessary for biomedical applications of CNTs. Polyethylene glycol (PEG), as a widely used surfactant, has been used to covalently modify CNTs in vivo and in vitro. Another widely used method of modifying CNTs is the cycloaddition reaction, which occurs on the side of the aromatic ring rather than at the end of the nanotubes or at the site of the defects formed by oxidation. The cycloaddition reaction connects the azide to the CNTs through a photochemical reaction. Prato et al. developed a method to modify CNTs by 1, 3-dipole cycloaddition reaction (Tagmatarchis and Prato, 2004). This method is also one of the most common modification methods.

Because the covalent modification method destroys the structure of carbon nanotubes, the inherent physical properties of carbon nanotubes, such as photoluminescence and Raman scattering, are destroyed after the chemical reaction. For example, the Raman scattering and photoluminescence intensity of single-walled carbon nanotubes (SWCNTs) are greatly weakened after covalent modification, which reduces the potential optical applications of this class of materials.

Compared with covalent modification, the non-covalent modification of carbon nanotubes involves the modification of polymers or amphiphilic surfactant molecules onto the surface of CNTs. In the process of non-covalent modification, the chemical structure of CNTs is not damaged, but the length of CNTs is shortened by ultrasound in the process of functionalization, and other physical properties are retained. Therefore, the aqueous solution of CNTs, especially SWCNTs, can be used in various fields of biomedicine through non-covalent modification.

The aromatic carbon graphite surface of CNTs can be loaded with aromatic molecules by π–π stacking. Using the π–π interaction between CNTs and pyrene, Dai’s group used pyrene and its derivatives to non-covalently modify CNTs and found that proteins can be fixed to the CNTs modified by pyrene derivatives (Chen et al., 2001; Wu et al., 2008). In addition, Bertozzi et al. also modified CNTs with pyrene-linked sugar tree molecules (Moon et al., 2008). In addition to pyrene derivatives, single-stranded DNA is also widely used to modify carbon nanotubes, mainly through π–π stacking between the nanotube surfaces and the aromatic rings in DNA (Kam et al., 2005). However, Moon et al. found that the SWCNTs modified by DNA molecules were unstable in vivo because they contained nuclease to degrade DNA (Moon et al., 2008). Dai et al. found that fluorescein-labeled PEG chains could modify SWCNTs, and fluorescein labeled CNTs were obtained by π–π stacking between aromatic rings on fluorescein and CNTs for biological imaging (Nakayama-Ratchford et al., 2007). In addition, other aromatic molecules have also been used to non-covalently modify CNTs.

A variety of amphiphilic molecules are used to suspend CNTs by van der Waals forces and hydrophobicity to connect the hydrophobic part to the surface of the nanotubes, while the hydrophilic part is exposed to improve the water solubility of the nanotubes. In the process of biological detection based on SWCNTs, CNTs have been non-covalently modified with Tween-20 and block copolymer to reduce the non-specific adsorption between CNTs and proteins. Cherukuri et al. used block copolymer to modify CNTs for in vivo experiments (Cherukuri et al., 2006). However, using block copolymers to modify CNTs is not stable enough in vivo because proteins in plasma replace them. In addition, common surfactants such as sodium dodecyl sulfate and Tween-100 can be used to suspend CNTs, making them water-soluble. Amphiphilic-polymer-modified CNTs have a relatively high critical micelle concentration and are unstable in solution without the presence of other surfactants. This modification method is not suitable for use in biological systems, because large amounts of surfactant may dissolve cell membranes and denature proteins. An ideal non-covalent modification of CNTs for modifying biological systems should have the following characteristics: First, the modified molecule should be either non-toxic or biocompatible. Second, the polymer modified on the surface of CNTs should be stable enough to prevent dissociation in biological systems, especially in plasma solutions containing salts and proteins. Amphiphilic-polymer-modified CNTs should have a low critical micelle concentration value and maintain the stability of the CNTs after removing the excess modified molecules. Third, the modified molecule should have functional groups that can be used for biological coupling (such as linking antibodies or other molecules), allowing CNTs to be used for different aspects of biological applications.

Dai and co-workers developed non-covalent PEG-phospholipid (PL-PEG)-modified CNTs to meet the various needs of biomedicine and obtained highly water-soluble CNTs and their various functionalities (Figure 2) (Liu et al., 2008). Because phospholipids are a major component of cell membranes, they are safer than biological systems. During the modification process, two hydrocarbon chains of liposomes were anchored on the surface of the CNTs, while the hydrophilic chains extended to the water phase, making the CNTs water-soluble and biocompatible. Different from surfactant-modified CNTs, PL-PEG-modified SWCNTs have good stability in solution, including serum. PL-PEG of different PEG lengths and structures can be used to modify SWCNTs. Binding of biomolecules may be affected by amino interactions with PEG terminals. Using this strategy, PEG-modified single-walled CNTs have been successfully applied in the biomedical field, mainly including biological detection, biological imaging and drug delivery.

Many active chemical groups (such as hydroxyl groups, epoxy and carboxyl) exposed at the GO edges and defect sites can be covalently linked to modify nanographene (Park and Ruoff, 2009). PEG is a strong hydrophilic polymer, which is widely used for modifying all kinds of nanometer materials, so as to improve their biocompatibility, reduce the nanomaterials non-specific adsorption of biological molecules and cells, and further improve the nanomaterials in vivo to better pharmacokinetically promote tumor targeting (Liu et al., 2007a). In 2008, Dai and co-workers used the amino-terminal-branched PEG to modify GO for the first time, mainly connecting the amino group of PEG to the carboxyl group on the surface of GO, so as to obtain PEG-modified GO (NGO-PEG) of ultra-small size (10–50 nm). NGO-PEG has shown excellent stability in a variety of physiological solutions (Yang et al., 2010). Since then, PEG-modified GO has been widely used in cell and in vivo biomedical research (Yang et al., 2010; Yang et al., 2011b)

In addition to PEG modification, many other hydrophilic macromolecules are used to covalently modify nanographene. Liu’s group covalently modified GO with aminated dextran, which could significantly improve the stability of GO in physiological solution (Zhang et al., 2011). Bao et al. used chitosan to covalently modify nanographene for drug and gene delivery (Bao et al., 2011). In addition, Gollavelli’s and Ling’s groups modified the GO functional complex with polyacrylic acid to improve its biocompatibility (Gollavelli and Ling, 2012). In addition to reacting with the carboxyl group on the GO surface, the epoxy group on the GO surface can also connect to the polymer to improve its stability. In a recent study, Niu et al. modified GO using an amino group on polylysine linked to an epoxide group on the surface of GO (Shan et al., 2009).

In addition, besides using hydrophilic polymers to modify GO, there are some other methods, such as using some small molecules to modify GO. Zhang and co-workers reported that sulfonic acid can covalently connect to the GO surface to improve the stability of GO in physiological solution (Zhang et al., 2010). Prato’s group modified GO with methylene nitride via a 1,3-dipole cycloaddition reaction (Quintana et al., 2010). However, the use of small molecule modifications of GO may require more testing in biological experiments to determine their biocompatibility.

In addition to covalently modifying nanographene, nanographene can also be non-covalently modified with polymers or biological macromolecules through hydrophobic interactions, π–π stacking or electrostatic adsorption. GO can interact with surfactants and hydrophilic polymers to improve their water solubility. However, for biomedical applications, biocompatible-polymer-modified GO is better than small-molecule-surfactant-modified GO. Park et al. (2010) used the biocompatible molecule triton to modify reduced nanographene. Hu et al. (2012) used Pluronic F127 (PF127) to modify nanographene. The hydrophobic part of PF127 was connected to the GO surface through hydrophobic action, and the hydrophilic part extended into the solution, so as to prepare a very stable GO PF127 complex. Recently, Professor Dai Hongjie’s group prepared an ultra-small RGO, which was non-covalently modified by amphiphilic PEGylated polymer and attached with Arg-Gly-Asp (RGD) to improve its biocompatibility (Robinson et al., 2011). In recent work, Liu’s group used branched PEG to modify RGO and obtained RGO–PEG with very stable properties. RGO–PEG has an extremely long blood circulation time and can be used for photothermal therapy of tumors (Yang et al., 2012b).

The non-covalent modification methods listed above are based on the hydrophobic forces between hydrophilic polymers on the GO surface. Since the surface of GO has a strong negative charge, a polymer with positive charge can bind to the surface of GO through the effect of charge, so as to achieve the effect of modification. Liu’s group modified GO with a positively charged polymer, polythylene-imine (PEI), commonly used in gene transfection to obtain a very stable GO–PEI complex in physiological solution, while reducing the toxicity of PEI alone and improving the efficiency of GO-based gene transfection (Feng et al., 2011). In addition, Dong et al. (2011) used the same method to modify graphene nanoribbons to obtain PEI–NGR complexes for cell gene transfection and in situ microRNA detection. Misra’s group used the same approach for drug delivery. DOX-loaded GO was modified with positively charged folic-acid-linked chitosan to obtain DOX–GO–Chitosan–Folate nanocarriers for pH-controlled drug release (Depan et al., 2011).

In addition to the use of synthetic polymers to modify GO, some natural biomolecules such as proteins and DNA can also modify GO. Recently, Huang et al. modified GO with fetal bovine serum to obtain a GO–protein complex. Compared with GO without surface modification, GO–protein significantly reduced cytotoxicity (Figure 3) (Hu et al., 2011). In addition, non-polarized amino acids in BSA can also bind to the GO surface by hydrophobic forces. In another work, Liu et al. (2011) modified GO with gelatin. Graphene and GO have very high electrons on the surface, so GO can be linked to many aromatic molecules through π–π stacking.

Ren et al. prepared p-phenylenediamine functionalized CD-mediated silk fibroin-PLA (SF-PLA) nanofibrous scaffolds. It has better mechanical properties and can be applied in cardiac-tissue engineering and nursing care (Figure 7) (Yan et al., 2020). In the absence of any external electrical supply, the incorporation of CDs into SF-PLA scaffolds significantly enhanced cell adhesion, proliferation and mRNA expression of cardiac genes (Tnnc1, Tnnt2, Cx43, and Atp2a2).

Shafiei et al. developed nanofibrous scaffolds for efficient bone-tissue engineering application by an electrospinning method (Shafiei et al., 2019). Non-invasive scaffolds based on CD-peptide-mixed tannin-polyurethane (CDP-f-PU) fabricated by Gogoi et al. (2017) exhibited higher biocompatibility, osteoconductivity and osteodifferentiation ability in bone-tissue regeneration. Ghorghi et al. (2022) fabricated Captopril/CQDs/polycaprolactone (CP-CQDs-PCL) nanocomposite scaffolds with reduced fiber diameter due to the conductive properties of CQDs in the scaffolds (Figure 8). The scaffold with 0.5% CQD significantly increased the adhesion, proliferation and ALP activity of MG-63 cells, which was sufficient for bone-tissue engineering applications.

It has been reported that CNTs can support neuronal attachment, promote the generation of neurites and be directly used as substrates for neuronal-tissue engineering (Koppes et al., 2016; Arslantunali et al., 2014; Lee et al., 2018; Shao et al., 2018; Shrestha et al., 2018; Wang et al., 2018; Wu et al., 2017; Xia et al., 2019). However, the potential toxicity of biological systems and the difficulty in generating canonical 2D/3D structures limit their utilization. Although polymer scaffolds have excellent biocompatibility, their poor electrical conductivity and the fact that they are not easily stretched limit their application in electrically propagating tissues. However, in combination with carbon nanotubes with strong electrical conductivity and good mechanical properties, they can be used in neuronal-tissue engineering, while the polymer scaffold minimizes the toxic effects of carbon nanotubes. Numerous studies have shown that carbon nanotubes and biocompatible polymers combined with each other can be used as two-dimensional conductive meshes or membranes for neural-tissue engineering (Table 1).

Shao et al. (2018) synthesized membranes composed of carbon nanotubes and poly (Dimethyl Diallyl Ammonium chloride). These membranes can provide potent regulatory signals for the adhesion, differentiation and growth of neural-stem-cell-derived neurons. Wu et al. (2017) studies on Schwann cells showed that the biocompatible chitin/carbon nanotube (Ch/CNT) composite hydrogel not only had no cytotoxicity nor neurotoxicity but also exhibited increased proliferation, adhesion and diffusive bodies, and cells exhibited greater extension and elongation. Arslantunali et al. (2014) designed the MWCNT/poly (2-hydroxyethyl methacrylate) composite catheters, which can promote peripheral nerve regeneration.

More recently, Xia et al. (2019) reported electrospun fibrous scaffolds constructed using polyvalent polyanion-dispersed carbon nanotubes, which can be used for neural-tissue regeneration. The results demonstrated that the scaffold acts as a biocompatible platform for pluripotent stem cell (IPS) adhesion and proliferation, while being able to induce higher differentiation of IPSs. Furthermore, aligned fibers can guide neurites to grow in specific directions. Lee et al. (2018) synthesized MWCNT and PEGDA composite scaffolds; when combined with electrical stimulation, they showed synergistic effects on promoting nerve growth in neural-tissue engineering.

Shrestha et al. (2018) developed a self-electrically stimulating bilayer nerve guide catheter (NGC) for neural-tissue engineering. The results of cell experiments showed that this NGC can enhance the proliferation, differentiation and adhesion of PC 12 cells and Schwann cells. Furthermore, cells can be observed to distribute along the aligned fibers. Wang et al. (2018) produced a poly (lactic-co-glycolic acid) (PLGA) composite scaffold that could guide the growth of PC12 cells and DRG neurons along the fiber direction. Furthermore, by applying electrical stimulation, the material PC12 cells and DRG neurons loaded with polylactic-co-glycolic acid (PLGA) composite scaffolds exhibited longer neurite lengths, and PC12 cells under 40 mV electrical stimulation also increased compared with the control differentiation. Similarly, Koppes and colleagues (Koppes et al., 2016) synthesized composite scaffolds by mixing carboxylated SWCNTs with hydrogels, which could be applied by direct current stimulation. With the increase in SWNT loading, the electrical conductivity of the composite scaffolds increased significantly, resulting in a marked improvement in neurite outgrowth. Furthermore, by applying electrical stimulation, the growth of the SWCNT-loaded material was increased by a factor of 7.0 compared with the control. However, the exact mechanism of neurite elongation is not well understood, so further research is required.

Biomaterials with sufficient mechanical strength and good electrical conductivity can be used in cardiac-tissue engineering. Despite their biocompatibility, polymeric materials are generally inert and require further improvement of their mechanical properties, which limit cell-to-cell interactions and signaling (Shin et al., 2013b; Kharaziha et al., 2014; Pok et al., 2014; Wickham et al., 2014; Sun et al., 2015; Liu et al., 2016; Yu et al., 2017; Mombini et al., 2019). Various polymer/CNTs composites with high mechanical and electrical properties of biopolymers have been investigated (Table 2).

Shin et al. (2013a) prepared CNT–GelMA and cultured it with neonatal rat cardiomyocytes. Compared with controls, cardiomyocytes exhibited a 3-fold increase in spontaneous beating rate, an 85% lower excitation threshold and significantly improved cell adhesion. The experimental results obtained by Yu et al. (2017) showed that CNT–collagen scaffolds increased the rhythmic contractile area of cardiomyocytes (CMs), indicating improved function of CMs.

Wickham et al. (2014) showed that PCL and PCL/T-CNTs membranes did not improve the attachment and proliferation of cardiac stem progenitor cells (CPCs). In the case of electrospun meshes, the incorporation of T-CNTs can be observed to increase mechanical strength and cell proliferation, but no effects on cell differentiation are observed. Similarly, Liu et al. (2016) showed that electrospun poly (lactic-co-glycolic acid)/MWCNTs induce elongation of neonatal rat cardiomyocytes while enhancing the production of sarcomeric α-actin and troponin I in cardiomyocytes. The application of carbon nanotubes helps to overcome the limitation of insufficient remodeling of intercalated discs (IDs) connecting cardiomyocytes. Kharaziha et al. (2014) produced grids that were able to support cardiomyocyte ingrowth. Compared with the grids without MWCNTs, the grids containing MWCNTs exhibited stronger spontaneous and synchronous beating behavior, with a 3.5-fold lower excitation threshold and a 2.8-fold higher maximum trapping rate.

Pok et al. (2014) studies have shown that composite hydrogels can achieve excitation conduction velocities (22 ± 9 cm/s) similar to native myocardial tissue to support the function of cardiomyocytes. Therefore, they can be used as electrical nanobridges for communication between cardiomyocytes. Sun et al. (2015) developed a CNT/collagen patch that enhanced the assembly of intercalated discs in cardiomyocytes. Mombini et al. (2019) produced electrospun cardiac conductive scaffolds composed of chitosan (CS), polyvinyl alcohol (PVA) and carbon nanotubes at different concentrations (1, 3, and 5 wt%). Biological results showed that the electrical stimulation of scaffolds containing 1 wt% CNTs promoted gene expression of cardiac markers.

Engineered tissue is used to produce synthetic 3D structural bone scaffolds that provide support for cell differentiation, attachment and proliferation (Gupta et al., 2014; Oyefusi et al., 2014; Duan et al., 2015; Khalid et al., 2015; Gonçalves et al., 2016; Tanaka et al., 2017a; Tanaka et al., 2017b; Jing et al., 2017; Wang et al., 2019; Świętek et al., 2019). Bone generates an electrical current when pressure is applied, so bone is a tissue with piezoelectric properties, whereas piezoelectric materials exhibit the reverse piezoelectric effect (The material compresses when an electrical current is applied) (Table 3); therefore, bone-tissue regeneration can be accelerated by using electroactive scaffolds and applying electrical stimulation.

Tanaka et al. (2017a) fabricated CNT-containing 3D bulk structures and investigated their efficiency as scaffolds for bone repair. Mechanical tests showed that the compressive strengths of the fabricated structures and rat femurs were 62.1 and 61.86 MPa, respectively, with no significant differences. In the presence of recombinant human BMP-2, CNT blocks showed higher ALP activity favoring osteogenic behavior. Tanaka et al. (2017b) concluded that carbon nanotubes have better protein uptake and release capacity than hydroxyapatite (HA). CNT porous structures combined with recombinant human BMP-2 showed higher cell proliferation, better osteoconductivity and more osteogenesis. Oyefusi et al. (2014) grafted HA onto carbon nanotubes and graphene nanosheets and then used human fetal osteoblast cell lines to study their effects on cell proliferation and differentiation. The experimental results showed that both CNTs–HA and graphene--HA materials can promote cell growth and differentiation. Khalid et al. (2015) synthesized MWCNT/HA scaffolds containing different loadings of CNTs (1 wt%, 3 wt%, and 5 wt%). The cytotoxicity of the composite scaffolds was dose-dependent with the CNT content. Experiments using a human osteosarcoma cell line showed that cell viability decreased with the increase in CNT content.

Duan et al. (2015) fabricated poly (L-lactic acid) (PLLA)/CNTs scaffolds using a freeze-drying method. In vitro studies have shown that CNTs can promote the cell proliferation and osteogenic differentiation of bone-marrow mesenchymal stem cells (BMSCs). In vivo experiments showed that CNTs significantly promoted the expression of osteogenesis-related proteins and the formation of type I collagen. Jing et al. (2017) reported that MWCNT scaffolds could enhance the proliferation rate of BMSCs and their protein expression of bone sialoprotein (BSP) and osteocalcin (OCN).

Goncalves et al. (2016) synthesized 3D-printed porous CNT scaffolds, which can be used for bone-tissue engineering. Cell proliferation experiments (6 days) were performed with MG63 osteoblast-like cells, and the results showed that the use of scaffolds containing 10 wt% MWCNTs had the most positive effect on cell proliferation.

Recently, Wang et al. (2019) introduced a polymethylmethacrylate (PMMA) bone cement loaded with different MWCNTs (0.1, 0.25, 0.5, and 1 wt%) that could be injection-molded for high-load joint replacement. In vitro studies have shown that this bone cement can promote the adhesion, proliferation and osteogenic differentiation of rat bone-marrow mesenchymal stem cells (rBMSCs). In vivo experiments have shown that PMMA bone cement containing 1 wt% MWCNTs could significantly enhance bone ingrowth, and the ingrowth rate was as high as 42% 12 weeks after surgery. Swietek et al. (2019) reported a solvent-casting method to combine functionalized CNTs (fCNTs) and iron oxide (ION) in two different mass ratios (1:1 and 1:4) to synthesize PCL-based 3D composite scaffolds for in-bone regeneration. The toxicity of composite scaffolds was assessed using the SAOS-2 human cell line. The results showed that fCNTs containing 1 wt% significantly improved cell attachment, and ion concentrations below 1 wt% increased cellular metabolic activity, indicating that the composite scaffold could be used for bone-tissue regeneration.

Gupta et al. (2014) fabricated SWCNT composite microspheres and polylactic-co-glycolic acid (PLAGA) for bone-tissue engineering. Compared with pure PLAGA scaffolds, MC3T3-E1 cells added SWCNT showed no abnormality in adhesion and growth, but the cell proliferation rate and gene expression were enhanced. The composites were further implanted in Sprague Dawley rats for 2, 4, 8, and 12 weeks. The results showed that the SWCNT/PLAGA composite has good in vivo biocompatibility. However, possibly due to the lack of a controlled porous structure, the material did not significantly improve bone-tissue repair.

Charge-conducting polymers and carbon-based materials make biological interfaces conductive (Huang et al., 2013; Sinclair et al., 2013; Tu et al., 2013; Serrano et al., 2014; Song et al., 2014; Walters, 2014; Munoz et al., 2015; Pascual et al., 2015; Zhang et al., 2016). Among these conductive materials, graphene composites with excellent chemical, electrical and thermal properties and extremely low cytotoxicity are ideal materials for tissue engineering and prosthetics. Graphene and its derivatives are often used as neural structures in combination with neurotransmitters, anticoagulants and growth factors (Table 4).

Serrano et al. (2014) synthesized 3D free-standing porous and flexible GO scaffolds for nerve regeneration. Cell cultures showed that highly interconnected and viable neural networks formed on these scaffolds for 14 days. However, there is a lack of in vivo studies confirming that these scaffolds can guide nerve regeneration. Song et al. (2014) observed that 3D graphene had good biocompatibility and was able to promote the growth of microglia.

After coating the aligned PLLA nanofiber scaffolds with GO in the presence of the nerve growth factor (NGF), their hydrophilicity and surface roughness were significantly improved. This fibrous scaffold promotes the proliferation and differentiation of Schwann cells and rat pheochromocytoma 12 (PC12) cells (Zhang et al., 2016). Tu et al. (2013) constructed intelligent biomimetic GO-based composites. The acetylcholine-like unit (dimethyl aminoethyl methacrylate) and the phosphorylcholine-like unit MPC on the surface of GO can promote the germination and growth of neurites after being covalently linked together.

Neural-tissue engineering has also been developing materials that can induce neuroinflammation (Sinclair et al., 2013). Astrocytes, peripheral macrophages and microglia in the brain mediate neuroinflammatory responses to hypoxia, ischemia and viral and bacterial infections (Huang et al., 2013; Walters, 2014; Munoz et al., 2015; Pascual et al., 2015). Observations suggested that the neuroinflammatory response of 3D graphene to microglia is milder than that of 2D graphene, suggesting that topographical features may influence inflammatory behavior. Furthermore, 3D graphene foam promoted the growth of neural stem cells and PC-12 cells (derived from the neural crest), demonstrating their use in neural repair and neurogenesis.

In cardiac-tissue engineering, graphene is often used to treat cardiovascular disease and prevent blood clotting in artificial heart valves (Ong et al., 2015; Sergi et al., 2015; Weijmans et al., 2015). Cardiovascular disease affects human health due to its high incidence and prevalence and is intensely studied in regenerative medicine (Shi et al., 2010; Şenel et al., 2014). The scarcity of donors and the high number of complications limit the use of heart transplantation for the treatment of cardiovascular disease (Shin et al., 2013b; Gálvez-Montón et al., 2013). The heart has dynamic functional characteristics and a complex tissue structure. Therefore, in cardiac-tissue engineering, an ideal artificial cardiac scaffold needs to be similar to the heart in structure, electrophysiology and mechanical properties to stimulate angiogenesis and maintain the viability of transplanted cells (Shin et al., 2013a; Radhakrishnan et al., 2014).

GO/gelatin methacrylate (GM) hydrogel is a synthetic cardiac scaffold that can be used to treat cardiovascular disease (Nguyen et al., 2015) GM facilitates the uniform distribution of GO in the hydrogel matrix and is a biocompatible surfactant. GO/GM hydrogels support cell spreading and alignment and enhance viability and proliferation in a 3D environment. Their tunable mechanical stiffness and electrical conductivity enable engineered cardiac patches to treat myocardial infarction (Kim et al., 2013; Kim and Song, 2014; Nguyen et al., 2015). Hydrogels can increase the local retention time of the carriers at the target site and increase their likelihood of being internalized by the tissue (Bae et al., 2012; Qi et al., 2014). Studies have reported that PLL can enhance the biological activity of GO in tissue engineering. GO/PLL composites have strong biocompatibility and high mechanical properties, which can accelerate cell growth and differentiation (Shin et al., 2014).

Graphene is a multifunctional carbon-based material that can serve as a framework for multicomponent nanostructured biomimetic scaffolds with bone-like morphology and chemical structure (Deepachitra et al., 2013; Lalwani et al., 2013; Depan et al., 2014; Oyefusi et al., 2014). The application of graphene nanocomposites containing polymers (chitosan, collagen, polypropylene fumarate) in bone-tissue engineering is discussed in detail in the following sections (Table 5).

GO is uniformly dispersed in the matrix, generates strong interfacial adhesion through polar and hydrogen bonding interactions and provides sufficient stiffness and strength in the biological state, providing effective support for the formation of bone tissue (Depan et al., 2014). Recently, a research group has synthesized a layered hybrid system composed of chitosan, GO and HPA (Oyefusi et al., 2014). The GO layer provides biocompatibility and mechanical strength; the HPA layer provides higher bioactivity; and chitosan in the middle layer further strengthens the hybrid material. Incubation of pre-osteoblasts (MC3T3-E1) with this hybrid system promotes uniform cell mineralization, proliferation and enhancement of proteins (actin, vinculin and fibronectin). These properties are important for cell adhesion and osteoblast morphogenesis (Oyefusi et al., 2014). Lalwani et al. (2013) synthesized PPF 2D nanocomposites containing GO nanosheets, and MoS2 nanosheets showed the highest mechanical strength and were ideal candidates for BTE. Deepachitra et al. (2103) studies on osteosarcoma bone cell lines (SaOS-2, CPC-2, MG-63, etc.) showed that fibrin is an important marker of osteoblast differentiation. Colorimetric assays (MTT, alkaline phosphatase (ALP) and in vitro calcium release) have shown that fibrin-modified GO has good biocompatibility and enhances osteoconductivity, making it an ideal scaffold for BTE.