Spinal cord injury (SCI) is considered to be one of the most challenging central nervous system injuries, the serious complications and high rates of paraplegia caused by SCIs have brought great burden to individuals, families, and society (Venugopal et al., 2017). Once damaged or degenerated, nerve cells cannot be self-repaired, which results in the permanent loss of function. From a pathological point of view, SCI is caused by a primary injury and a series of secondary injuries. Primary injuries are mainly acute injuries caused by mechanical forces such as extrusion and dislocation, including damage to neurons and glial cells at corresponding segments, leading to the rupture of blood vessels (Schwab and Bartholdi, 1996; Sobrido-Camean and Barreiro-Iglesias, 2018). Secondary injuries are relatively complicated including ion homeostasis disorders, local edema, ischemia, free radical excess, and intense inflammatory responses (Rowland et al., 2008; Orr and Gensel, 2017). At the same time, both primary injuries and secondary injuries to the microenvironment are not conducive to the differentiation of neurons (Sabelstrom et al., 2014). This series of changes can lead to the formation of cystic cavities, while astrocyte proliferation leads to the formation of scar tissue (Liu and Chan, 2016). In the presence of cavities, the formation of glial scars further impedes the regeneration of neurons, eventually leading to the loss of original sensation and voluntary activity.

At present, the clinical treatment of spinal cord injury is mainly divided into non-surgical treatment and surgical treatment. One of the non-surgical treatments for SCIs is post-injury shock therapy with excessive doses of methylprednisolone (MP) (Breslin and Agrawal, 2012), which is a corticosteroid that inhibits lipid peroxidation. While acting as a scavenging agent for free radicals, it also restricts the inflammatory response, protecting the spinal cord–blood barrier, and increasing blood flow to the damaged spinal cord. However, since its therapeutic benefits are also controversial—including increasing risks of urinary tract, respiratory, and wound infections—these complications limit its use (Hurlbert, 2014). For surgical treatment, decompression and fixation have always been the most important methods to treat SCIs. The primary goal of SCIs treatment is to remove the compression factors of and restore the stability of spinal structure to the greatest extent. Multisystem medical management, hypothermia, and rehabilitation medical care are also the methods of treating SCIs (Silva et al., 2014). The aforementioned methods have made some progress in the repair of SCIs; however, at present, all clinical treatment methods can only remove the injury factors to the greatest extent but cannot make the injured nerve regenerate functionally. Therefore, it can be said that its repair of the injury is incomplete.

Nervous system regeneration refers to the re-establishment and repair of functional neural connections, nervous tissue, and cells. Methods to accomplish this through neural tissue engineering involve providing direct replacement of neural cells and/or the repair of circuitry by utilizing cell transplantation, bio/chemical molecular signaling, and directed guidance “bridge” scaffolding (Joung et al., 2020). The most fundamental part is the regeneration of cells. There are only three major cell types in the spinal cord, and all cell lineages can be derived from neural stem cells (NSCs). These three kinds of cells are neurons, astrocytes, and oligodendrocytes. The axon extends out of the neuron and forms a network with other neurons. After being electrically stimulated, the axon terminals release neurotransmitters to form chemical signals and transmits the chemical signals to the dendrites of other neurons to which they are functionally connected. These functional connections are called synapses, which form the basis of the neuronal network (Joung et al., 2020). Glial cells-astrocytes or oligodendrocytes play a supporting role in the structure of the spinal cord. Oligodendrocytes mainly form myelin around neurons, which plays a vital role in the electrochemical activities of neurons. Among the many phenotypes of astrocytes, scar-forming astrocytes mainly form glial scar which produce axonal growth inhibitors and prevents axonal regeneration. This sequential phenotypic change has long been considered to be unidirectional and irreversible; thus glial scarring is one of the main causes of the limited regenerative capability of the SCIs (Okada et al., 2018). Consequently, many researchers have focused on the directed regeneration of neurons in the repair of SCIs, hoping to reduce the formation of glial scar as much as possible while regenerating neurons, and expect the regenerated neurons to form synapses with the host neurons, so as to finally restore the function of the damaged site.

Researchers began by injecting NSCs and/or cell growth factors directly into the site of the injury, but the study did not consider introducing scaffolds into the injury site. Although these studies achieved good results in animal experiments, the lack of scaffolding structure and difficulties in the selection of optimal performance cells have been the great challenges to translate this method into practical clinical application (Joung et al., 2020). With the development of tissue engineering, researchers consider that the repair of SCIs needs to provide a certain support structure to the damaged site, to restore the mechanical structure, and provide an effective protective barrier for implanted cells (Struzyna et al., 2014; Struzyna et al., 2015).

The spinal cord is linear in structure, which provides structural support for the body’s reflex response, the axons extending from neurons form synapses with other neurons, which is the basis of body’s reflex response. For example, the corticospinal tract is the most important motor system of human beings, but the corticospinal projection remains largely refractory to regeneration. The fundamental cause of this failed regeneration is that synapses cannot be formed between neurons (Kadoya et al., 2016).

For the repair of SCIs, the ideal biomimetic scaffolds should satisfy the following criteria:

1) The designed biological scaffold should conform to the mechanical properties of the spinal cord as far as possible, which is ideally approximately 10 kPa (Joung et al., 2020).

In order to achieve such ideal conditions, the most important things are scaffold fabrication materials and the scaffold fabrication process. First, soft materials are needed to make the scaffolds to avoid mechanical mismatch. Hydrogels are widely used in the application of scaffolds because most of them have good biocompatibility and predictable biodegradability (Huang et al., 2017), and most natural hydrogels have specific cell binding sites, which are required for cell attachment, proliferation, growth, and differentiation (Gungor-Ozkerim et al., 2018). Hydrogels can also facilitate the exchange of gases and soluble nutrients (Yeh et al., 2006). At the same time, hydrogel is more suitable for the strength of spinal scaffolds because of its low mechanical properties, and because of their high water content and non-toxicity, excellent mimics of the extracellular matrix (ECM) is rendered (Holzl et al., 2016).

They can be broadly divided into two categories—that is, natural and synthetic hydrogels. Natural hydrogels—such as collagen, chitosan, agarose, and alginate—have been widely used because of their good biocompatibility. They can be further classified based on their source. Hydrogels such as collagen, fibrin, and gelatin typically come from vertebrates, and therefore have inherent cell-adhesion signaling molecules, while hydrogels such as alginate and agarose come from other organisms, such as algae, which lack these signaling molecules (Hospodiuk et al., 2017). For example, collagen contains some adhesion motifs as RGD (Arg–Gly–Asp) -an important tripeptide for the interaction between a variety of cells and the ECM (Gasperini et al., 2014). Agarose does not have adhesion motifs to cells, so when it acts as a hydrogel matrix, there is no interaction with cells (Tang et al., 2007). However, it can also be supplemented cell adhesion motifs to the matrix, such as fibronectin (Karoubi et al., 2009) or RGD soluble peptide (Guaccio et al., 2008). Sharp et al. (2012) treated a rat model of SCI with salmon fibrin, the results demonstrated that the sensory, motor, and bladder function of the rats recovered to a certain extent after treatment. Ansorena et al. (2013) made progress in an experiment using alginate to conduct nerve growth factor repair in the spinal cord hemi-section model. In their experiment, the hydrogel scaffold-carrying growth factor had a certain effect on the functional recovery of animals after SCIs.

Although natural hydrogels have good biocompatibility, their very soft mechanical properties cannot provide the strength of biological tissues, which can easily cause structural collapse after implantation, leading to repair failure. Conversely, synthetic hydrogels can be chemically modified with domains that enhance the mechanical properties of biological printing structures. For example, adding methacryloyl groups into gelatin to form a synthetic hydrogel–gelatin methacrylate (GELMA), the hydrogel can exhibit different mechanical strengths under different ultraviolet light times and intensities. At the same time, the mechanical strength can also be adjusted according to the concentration of methacryloyl groups (Gungor-Ozkerim et al., 2018). The degradability of the synthetic hydrogels can also be modulated. For example, McKinnon et al. (2013) synthesized a polypeptide cross-linked poly (ethylene glycol) (PEG) hydrogel and investigated the proliferation and differentiation of embryonic stem cell-derived motor neurons (ESMNs) in it. Their result demonstrated that the hydrogels were able to promote neuronal survival and axon outgrowth using cell-extracellular matrix interactions and allowed neurons to remodel their extracellular environment using matrix metalloproteinase (MMP)-mediated hydrogel degradation. In addition, improved synthetic hydrogels can carry a variety of bioactive factors that can affect the proliferation and differentiation of NSCs. For example, research works printed fibroblast growth factor-2 (FGF2), ciliary neurotrophic factor (CNTF), and fetal bovine serum (FBS) on a polyacrylamide-based hydrogel. After NSCs were seeded on the scaffold, NSCs were proven to respond properly to printed macromolecules, and suggested that printed scaffold could successfully achieve effective control of stem cell differentiation (Ilkhanizadeh et al., 2007).

Another factor that has an impact on scaffolds fabrication is the manufacturing process. The research on tissue engineering scaffolds is not uncommon. The traditional manufacturing methods of tissue engineering scaffolds, such as gas foaming, melt molding, electrospinning, and phase separation, have been used in the production of scaffolds composed of synthetic and natural polymers (Liao et al., 2004; Cheng et al., 2018).

Bioprinting technology has attracted extensive attention in recent years. Bioprinting is defined as the use of 3D printing technology to deposit bioink or biomaterials ink on a receiving solid or gel substrate or liquid reservoir (Demirci and Montesano, 2007; Mironov et al., 2008; Moon et al., 2010).3D printing technology is an emerging technology, which can accurately “reproduce” tissue using the computer-aided design (CAD). 3D printing provides four important features (Joung et al., 2020) as follows:

1) It can be combined with 3D imaging technology to achieve anatomical accuracy.

It is different from traditional “subtractive” manufacturing, which processes the raw material, removing some of it to shape the tissue as required. “Additive” manufacturing, on the other hand, uses layers of materials that are combined to create a target tissue sample. In recent years, it has rapidly developed in the field of biomedical science and has been effectively applied in areas such as the osteochondral interface, peripheral nerves, kidneys, skin, cardiovascular, livers, and artificial ears (Mannoor et al., 2013; Johnson et al., 2015; Kesti et al., 2015; Homan et al., 2016; Gungor-Ozkerim et al., 2018; Rastogi and Kandasubramanian, 2019).

With the development of 3D printing, many researchers have focused on the theoretically feasible combination of SCI repairs using this technique. First, the combination of 3D printing and 3D images can customize the shape of the scaffold based on the distinctive SCI of different individuals, forming a “personalized” manufacturing scheme, so that the printed tissue structure can fit the microenvironment of the injury and achieve maximum anatomical fidelity. At the same time, 3D printing can provide direction for the regeneration of axons. This directional regeneration is conducive to the formation of synapses between axons, so as to achieve the regeneration of neural pathways in the spinal cord, and then promote the formation of neural networks (Kadoya et al., 2016). Second, 3D printing technology can be combined with multi-nozzle technology to achieve the function of printing multiple materials on a single printer at the same time. The combination of these different materials allows the structural simulation at the injury site to more closely resemble the natural spinal cord structure (Cao et al., 2003; Mironov et al., 2003; Hollister, 2005; Fedorovich et al., 2007; Ladd et al., 2013; Tabatabai et al., 2013). Furthermore, bioprinting has advantages in the establishment of in vitro models, as compared with traditional tissue engineering, bioprinting samples can be formed faster (Johnson et al., 2016). The establishment of in vitro models brings “preoperative simulation” to the reconstruction of nervous system tissue more effectively, which can achieve the goals of injury repair more accurately and effectively. Consequently, compared with traditional tissue engineering, 3D printing technology has created the potential for personalized tissue manufacturing, offering a new therapeutic trend for SCI regeneration.

The concepts of bioink and biomaterials ink are similar. The main difference is that the former, where cells belong to the mandatory component of the printing formulation, they can also contain bioactive factors. The latter, biomaterials are used for printing, and the cell contact occurs after manufacturing (Groll et al., 2018). Therefore, bioprinting can produce bioactive scaffolds with cell seeding by combining 3D printing with biomaterials ink; “living scaffold” or “cell-laden” can also be produced by combining 3D printing with bioink (Joung et al., 2020). Due to the limitations of the traditional scaffold fabrication process, which are not mild, it is impossible to make scaffolds containing bioactive factors, let alone the living scaffold. Even if the cells are cultured after scaffold fabrication, the growth direction of cells cannot be effectively controlled. At the same time, the scaffolds made by the traditional “subtraction” process cannot have a specific shape, such as 3D grid, which is composed of multi-dimensional linear structure. Most of the scaffolds manufactured by them only have porous structure on the surface of the scaffolds, while the interior is mostly solid structure. As such, traditional tissue engineering can be challenging in its quest to mimic the complexity of the spinal cord (Faroni et al., 2015).

In this article, we summarize the application of bioprinting in SCI repairs and show how this technique has led to further advances in neural tissue engineering. First, we introduce several main methods for spinal cord bioprinting. Second, we present landmark research on spinal cord bioprinting in recent years. Finally, we discuss future application prospects and the related disadvantages of this technique.

Bioprinting of the nervous system can be broadly divided into the following categories—that is, inkjet/droplet bioprinting, extrusion bioprinting, and light-assisted bioprinting (Khalil and Sun, 2009; Wust et al., 2011; Dababneh and Ozbolat, 2014; Wang et al., 2015). These methods are illustrated schematically in Figure 1, and the comparison of different bioprinters is listed in Table 1.

Due to the limitations of the traditional tissue engineering scaffold fabrication process, it cannot meet the requirements of complex structure manufacturing. Bioprinting can be divided into cell-laden and acellular ink printing. Compared with traditional tissue engineering scaffolds, acellular scaffolds made by bioprinting can have more meticulous spatial structure, and its “additive” manufacturing method can produce multi-dimensional linear structure, which conform to the spinal cord. Then, for cell-laden scaffolds, cells can be placed in a specific spatial position, and multi-cell scaffolds can also be made. This can meet the needs of the spinal cord for the direction of cell regeneration. With the increase of bioprinting, effectively manufacturing or recreating patient-specific constructs of clinically relevant size, shape, and structural integrity has been advanced using the combination of neural stem and progenitor cells with 3D printing biocompatible scaffolds to test new therapeutic options for SCIs (Joung et al., 2020). Table 2 summarizes the landmark research works of spinal cord bioprinting in recent years.

Chen et al. (2017) developed collagen/heparin sulfate scaffolds, which were used for extrusion bioprinting. In this study, a multi-dimensional grid scaffold with channels separated by approximately 400 μm was preapared, and the grid structure also had porous structure on the surface of the scaffold, which provided the structural basis of multi-dimensional distribution for subsequent cell inoculation. This structure could simulate the linear structure of the spinal cord, and the cells could grow relatively orderly along the channel of the scaffold after implantation. Considering that the mechanical properties of collagen are very soft relative to the spinal cord, the mechanical properties of collagen were improved by applying heparin sulfate modification to the scaffold; the results showed that the mechanical strength of the modified collagen was enhanced, and the strength of the modified scaffold after bioprinting was significantly enhanced. NSCs were then loaded onto biological scaffolds. Improvements in motor functions were observed after SCI. Jiang et al. (2020) also used collagen as a base and added silk fibroin protein to strengthen it to construct biological scaffolds. They also used extruded bioprinting technology and NSCs for inoculation. Different from Chen’s study, the scaffold model in this study attempted to imitate the butterfly-like structure of the gray matter of the spinal cord, which is generally an oval structure, with four linear pore structures in the middle gray of the matter and the surrounding white matter, and the rest is a solid structure. This structure is more consistent with the real structure of the spinal cord than that of the grid structure. The experimental results showed that NSCs adhered well and extended the scaffold, indicating that the scaffold had good biocompatibility. Subsequent kinematic function tests also showed that the motor function of rats improved after the implantation of the scaffold.

Koffler et al. (2019) developed a 3D-printed hydrogel spinal cord scaffold to evaluate spinal cord regeneration by planting cells on it (Figure 2). This study creatively developed a new microscale continuous projection printing technique (µCPP) using PEGDA–GELMA as an acellular hydrogel ink. This technology was improved based on DLP to achieve continuous printing. While inkjet or extrusion printing can compromise mechanical integrity through artificial interfaces between the printed drops or lines, µCPP’s layerless printed structures do not exhibit planar artifacts (interfaces). Similarly, this technology could print a biological scaffold suitable for the size of a patient’s spinal cord in 1.6 s and could also be used to customize the scaffold based on different shapes of human SCIs. They loaded neural progenitor cells (NPCs) onto bio-printed scaffolds in the hope that they would regenerate axons and thus restore functions after SCI. The experimental results showed that NPCs could be differentiated into neurons and grow along the channels formed by the scaffold. Cells were not only seen to enter the scaffold from the host end at the injured cranial side but regenerated axons were also seen to extend from the inside of the scaffold to the host end at the injured caudal side. They demonstrated that such “axon elongation” was capable of synaptic conduction and biological function. The results of their study are exciting and prove that 3D biomimetic scaffolds can be an effective option for promoting the regeneration of the spinal cord (Koffler et al., 2019).

As mentioned in the earlier studies, which pay more attention on the reconstruction of spinal cord structure, Zhu et al. (2017) focused on the study of cell differentiation. They applied GELMA and PEGDA as scaffold substrates and planted NSCs on biological scaffolds, and the channels of the scaffold were about 250 μm. Low-level light therapy (LLLT) has been shown to have positive effects on the rehabilitation of degenerative nerves and neural disorders. Consequently, low-dose light was used to explore the effects of light on the proliferation and differentiation of NSCs. The results showed that low-dose light promoted the differentiation of NSCs into neurons and inhibited the differentiation of glial cells. These findings suggested that integration of 3D printing and LLLT might provide a powerful methodology for neural tissue engineering (Zhu et al., 2017). Li et al. (2020) also focused on the differentiation of ectomesenchymal stem cells (EMSCs) but applied the cell-laden scaffold for research. They applied sodium alginate-Matrigel (SA-MA) hydrogel and extrusion bioprinting to reconstruct structures. At the same time, EMSCs were blended with hydrogel to form bioink as cell-laden. Traditional 2D cell culture was used for the control group, in which the cells were cultured directly in the culture medium, focusing on evaluating the differentiation of EMSCs between cell-laden and 2D culture. Experimental results showed that compared with traditional 2D cell culture, the microenvironment provided by 3D-printed scaffolds could promote the growth and proliferation of cells, and EMSCs could differentiate into neurons more effectively. The results of their study provided a new strategy for the differentiation of cells and the application of biological scaffolds in the treatment of SCIs.

Joung et al. (2018) also developed a cell-laden scaffold based on extrusion technology, using alginate (AG) and methylcellulose (MC) for printing. In contrast to other studies, this study enabled multiple neural cell types to be co-printed in a specific channel—clusters of NPCs and oligodendrocyte progenitor cells (OPCs) were placed with a spatial distribution of 200 µm within 150 µm wide channels. They expected the cells could differentiate into neurons and oligodendrocytes, then formed myelinated nerve fibers, and finally, while exploring the directional differentiation of cells, they could accurately locate cells in space. The experimental results showed that both types of progenitor cells could grow along the scaffold channel and extend and differentiate into corresponding cell lines, these results provided an important experimental basis for cell regeneration in spinal cord tissue engineering. It was also confirmed that AG combined with MC was the only material capable of carrying a variety of neural precursor cell types in the process of 3D bioprinting. In addition to differentiating NSCs into neurons, they could also differentiate into other neuron-related tissues. To repair the biological function of a damaged nerve, it is necessary to ensure the NSCs differentiate in the direction of the neuron as only in this way synapses structures can be formed and biological functions can be generated. This study proposed a new biological scaffold, which provided new ideas for simulating the structure of the nervous system in vitro and provided an option for the future development of new clinical treatment methods for SCIs.

Liu et al. (2021) developed a composite hydrogel scaffold composed of hydroxypropyl chitosan (HBC), thiolated hyaluronic acid (HA-SH), vinyl sulfonated hyaluronic acid (HA-VS), and Matrigel (MA). This study is the first in vivo experimental validation study using the cell-laden scaffold; the scaffold was also a multi-dimensional grid structure with 400 μm channels. When printed, the cells were evenly distributed in the grid hydrogel. They used extrusion printing, the cell survival rate being approximately 95%. The results of in vitro experiments showed that the composite scaffold had a degradation rate, porosity, and mechanical strength suitable for the growth of NSCs, providing an ideal microenvironment for the proliferation and differentiation of NSCs. The results of in vivo experiments and biomechanical measurements showed that the scaffold simulated the parallel linear structure of the spinal cord, realized the regeneration and connection of neurons, and promoted the recovery of motor functions in SCI models. The results of their study have promoted the development of the cell-laden scaffold in the regeneration of spinal cord injury (Figure 3).

Bioprinting is emerging as a promising tool in tissue engineering, providing bioengineering researchers with the ability to design complex 3D biological structures. It can print specific nerve cell subtypes and growth factors in a region-specific manner, promoting the differentiation of NSCs into neurons using the ratio of different components of bioink, providing the precisely orchestrated re-establishment of neural networks and connections. This provides a new direction for the repair of SCIs. Bioprinting technology can also be used to create individualized SCI models to test the feasibility of a treatment plan, including the biocompatibility of the cells and scaffolds, cell–cell and cell–matrix interactions with cell–matrix interactions, and even improve our understanding of the mechanisms of neuromodulation—in short, 3D printing can provide personalized treatment for SCIs. It can create unique models for different individuals with different injuries and is more in line with the “from the bottom up” regeneration of the body compared to traditional scaffolds.

However, although bioprinting has many advantages, it still faces many challenges. To date, only a few specific cell types and scaffold models have been studied in spinal cord bioprinting. As mentioned earlier, researchers pay attention to the differentiation of NSCs and the production of scaffold three-dimensional structure, but the structure of scaffold only focuses on the production of multi-dimensional grid structure. Although this simulates the structure of spinal cord linear conduction bundle, it does not completely restore the butterfly-like gray matter and surrounding white matter structure of the spinal cord. Current bioprinting technology has been able to simulate a variety of organs in the body, but the spinal cord has a complex tissue structure and many corresponding neural pathways, making it extremely challenging to complete bionic reconstruction.

At present, bioprinting for the spinal cord has not been combined with relevant studies such as vascularization, immunosuppression, and inflammation; this gives it a lot of room for development. To this end, it is necessary to encourage interdisciplinary research in multiple disciplines to promote the techniques which may have the greatest impact on SCI repair and provide guidance for the transformation of basic experiments into clinical efficacy. The cell-laden scaffold has been successfully used in the in vivo experiment of spinal cord injury, it is reasonable to believe that with the development of bioprinting technology, it will not be difficult to make a scaffold that fits the spinal cord structure more closely.