1. Introduction
Organoids are
in vitro three-dimensional (3D) cell clusters derived from primitive tissues, embryonic stem cells (ESCs), or induced pluripotent stem cells (iPSCs). Organoids are capable of self-renewal and self-organization, and they exhibit some functions similar to those of primitive tissues [
1]. Organoid models mimic real organs in terms of cell morphology, tissue structure, and function, thereby serving as powerful tools for embryonic development, disease research, drug screening, regenerative medicine, and personalized medicine. In the field of regenerative medicine, many preclinical studies have evaluated organoid-based cell therapies, and the findings of these studies have demonstrated the potential value of organoids for treating diseases such as bone defects, diabetes mellitus, and ulcerative colitis [
2,
3].
Neurotrauma, including spinal cord injury (SCI) and peripheral nerve injury (PNI), is an important condition in orthopedic clinical practice and has been shown to lead to the loss of motor and sensory functions. Owing to the limited self-regenerative capacity of nerves and the scarcity of treatment options, the prognosis for neurotrauma is generally poor [
4,
5]. Tissue-engineered grafts combining stem cells and biomaterials have been considered to be potential therapeutic approaches [
6]. More recently, numerous advancements have been made in research on spinal cord and peripheral nerve organoids. Since Lancaster et al. [
7] first used human induced pluripotent stem cells (hiPSCs) to generate neural organoids, researchers have optimized cell culture protocols based on
in vitro simulation of molecular signaling regulation during embryonic development to generate spinal cord organoids with ventral-dorsal features of the spinal cord and dorsal root ganglion (DRG) organoids through targeted induction [
8], [
9], [
10], [
11], [
12]. When organoids are combined with co-culture, neuromuscular junctions are generated, partially reproducing the motor and sensory nerve-conduction pathways
in vitro [
13], [
14]. Spinal cord and peripheral nerve organoids represent promising new
in vitro disease models for hereditary peripheral neuropathies and motor neuron diseases [
15], [
16], [
17], [
18]. Additionally, they are expected to offer innovative platforms for disease research and drug screening related to SCI and PNI, while opening new avenues for tissue engineering. By integrating engineering principles, organoids with more complex 3D structures can be developed through manipulation of cell density and the deposition and the incorporation of physical and biochemical cues [
19]. The current challenges in SCI and peripheral PNI regeneration include slow axon regeneration, inhibitory microenvironmental factors, and neural scar formation. Cell therapy can address these challenges through differentiation into the required cell types to replenish lost cells and by modulation of the microenvironment through paracrine signaling to facilitate regeneration. In comparison with mature cells or stem cells, which are directly utilized in conventional tissue engineering, organoids exhibit intricate cellular phenotypes and intercellular interactions, which confer significant advantages for tissue regeneration.
In this review, we describe the genesis and structure of the spinal cord and the peripheral nerve system, and on the basis of this information, we comprehensively summarize the key factors for constructing spinal cord and peripheral nerve organoids as well as research advances in the formation of 3D organoid structures through engineering approaches. We also explore the potential applications of spinal cord and peripheral nerve organoids in neurotrauma. Finally, we discuss the challenges and prospects associated with the construction of spinal cord and peripheral nerve organoids as well as their clinical translation in the context of neurotrauma.
2. Physiology and embryology of the spinal cord and peripheral nerve system
To construct organoids that accurately replicate the cellular components and structures of the target organs, simulation of the biochemical cues present during embryonic development is essential. Research on the formation and development of the spinal cord and peripheral nerves during embryonic development offers a theoretical foundation for organoid induction. The spinal cord is composed of peripheral white matter and central gray matter. The white matter mainly consists of ascending and descending nerve fiber tracts, whereas the neuronal cell bodies are located in the vascular-rich gray matter of the spinal cord [
20]. The spinal cord tissue primarily includes neurons and glial cells that exhibit complex phenotypes. A recent study using spatial transcriptomics and antibody validation identified 29 glial clusters and 35 neuronal clusters in the human spinal cord, which mainly aggregated into spinal nuclei along anatomical locations [
21]. The dorsal horn of the spinal cord regulates sensory processing and consists of a complex network of projection neurons and excitatory and inhibitory interneurons. The ventral region forms different connectomes to regulate motor functions and contains various networks of interneurons and motor neurons, with motor neurons ultimately sending out motor fibers to innervate the peripheral effectors [
22]. The pseudo-unipolar neuronal cell bodies of the peripheral sensory nerves are located in the dorsal root of the spinal cord, forming the structure of the dorsal root ganglia (DRG). These neurons collect peripheral sensory signals and establish synaptic connections with different neurons in the dorsal horn to process and project corresponding types of sensory signals (
Fig. 1(a)). Oligodendrocytes and Schwann cells (SCs) surround the axons of the central and peripheral nerves, respectively. They form stable, multilayered, 3D myelin structures around certain fibers, contributing to the formation of the nodes of Ranvier. Fibers with these structures are known as myelinated fibers, whereas those without such structures are referred to as unmyelinated nerve fibers [
23].
During early embryonic development, the neural plate bends dorsally to form the neural tube. Development of the neural tube proceeds along three axes: rostral–caudal (RC), dorsal–ventral (DV), and medial–lateral (ML). The RC axis determines the body axis from front to back, including the differentiation of regions such as the forebrain, midbrain, hindbrain, and spinal cord, which is determined by the spatiotemporal collinearity of homeobox (
Hox) genes [
24] and involves the gradient regulation of morphogens. The caudal end ultimately forms spinal cord segments under regulation of factors such as retinoic acid (RA), fibroblast growth factors (FGFs), and growth differentiation factor 11 (GDF11) (
Fig. 1(b)) [
25]. The DV axis is formed by the concentration gradients of factors such as Sonic hedgehog (SHH), bone morphogenetic proteins (BMPs), and wingless-type MMTV integration site family (WNTs). BMP and WNTs induce dorsal progenitor domains, whereas SHH induces ventral progenitor domains [
9]. Gradient regulation of these signals leads to the differentiation of neural progenitor cells into different domains, ultimately producing various subtypes of spinal interneurons and motor neurons from the patterned progenitor domains (
Fig. 1(c)). Recent studies have shown that the population of neuromesodermal progenitor (NMP) cells located at the edges of the somites of elongated embryos plays a crucial role in the formation of the spinal cord and paraxial mesoderm [
20]. NMPs are bipotent axial stem cells that serve as a common developmental origin for both the spinal cord and the musculoskeletal system it innervates [
24]. Neural crest cells (NCCs), located at the boundary of the neural plate, are the primary source of cells in the peripheral nervous system. After closure of the neural tube, NCCs lose their intercellular connections and undergo an epithelial-to-mesenchymal transition, allowing for extensive migration (
Fig. 1(d)). NCCs migrate from the neural tube to the DRG region, where they differentiate into ganglia containing various sensory neurons and satellite cells under the guidance of a range of signals in the microenvironment. The differentiation of stem cells into NCCs is regulated by the WNT and mitogen-activated protein kinase (MAPK) signaling pathways, while RA, neurotrophin-3 (NT-3), brain-derived neurotrophic factor (BDNF), and beta-nerve growth factor (βNGF) signaling pathways influence the final differentiation of NCCs into sensory neurons of varying diameters. Glial cells develop after the completion of neurogenesis. Neural stem cells (NSCs) in the spinal cord differentiate into oligodendrocytes and astrocytes, whereas peripheral NCCs give rise to SCs and satellite cells [
26].
By supplying biochemical cues at various developmental stages and simulating the morphogen gradients found in different tissues during embryonic development, fate determination can be replicated, and stem cell differentiation can be induced in vitro. This approach can facilitate the generation of organoids composed of specific cell types.
3. Design and fabrication of spinal cord and peripheral nerve organoids
Spinal cord and peripheral nerve organoids can perform physiological functions by recapitulating the synergistic interactions between various cell types [
27]. Therefore,
in vitro induction of multidirectional differentiation of stem cells into cells with distinct phenotypic characteristics, along with the establishment of a 3D culture environment, is crucial for organoid cultivation. The most common protocols involve formation of embryoid bodies, neural induction, and differentiation. The three key elements in this process are the cellular origin, signaling-modulating factors, and matrix materials for 3D culture (
Fig. 2).
3.1. Cellular origin
The stemness and multipotency status of the starting cells determine the activity of organoids and their response to morphogens, making them key factors in organoid development. In organoid cultures, the starting cells typically include pluripotent stem cells (PSCs) and tissue stem cells (TSCs).
The most common starting cells for neural organoids are ESCs and iPSCs, because they have the potential to differentiate into all three germ layers and can form complex cellular compositions by mimicking embryonic development. ESCs, with their robust stemness and pluripotency, have been widely used in organoid cultures [
28]. However, the limited availability of ESCs, coupled with ethical concerns regarding embryo use, has restricted their application in clinical settings. iPSCs are obtained through somatic cell reprogramming, which raises fewer ethical issues and is considered more promising for clinical applications. Additionally, reprogramming patient-derived cells facilitates disease modeling and personalized medicine [
29]. However, the reprogramming step increases the culture duration and may lead to organoid heterogeneity [
28]. In recent years, researchers have successfully developed various organoids resembling the neural tube [
30], [
31], spinal cord [
9], [
12], [
25], [
18], [
32], [
33], [
34], [
35], [
36], DRG [
11], [
26], [
37], and neuromuscular junctions [
14], [
38], [
39], primarily based on the guided induction of ESCs and iPSCs. Overall, organoids derived from PSCs are highly complex and can recapitulate the process of embryonic development
in vitro. However, they also have drawbacks such as highly demanding culture conditions and high heterogeneity.
Some researchers have explored the possibility of bypassing the iPSC induction step by directly reprogramming somatic cells to generate neural organoids. Generation of neural organoids directly from somatic cells through reprogramming can enable direct fate conversion between different cell types and bypass the intermediate states of embryonic development. This approach can significantly shorten the culture time, particularly for human cells [
40]. For example, fibroblasts can be reprogrammed to become pain receptor neurons [
41]. Xiao et al. [
10] successfully generated self-organizing sensory ganglion-like organoids and retinal ganglion cells from fibroblasts through the induction of the triple transcription factors (TFs) Ascl1, Brn3b/3a, and Isl1. Xu et al. [
42] found that human astrocytes can be directly reprogrammed by overexpressing octamer-binding transcription factor 4 (OCT4), inhibiting p53, and providing small molecules such as CHIR 99021, SB 431542, RepSox, and Y27632. This approach generated neurons that were reintroduced into a neurogenic state and subsequently cultured to produce brain and spinal cord neural organoids. The organoids generated in both studies exhibited similarities in cell type to those derived from PSCs, indicating that direct reprogramming of somatic cells is a promising strategy for organoid generation. Notably, the fibroblasts originated from the mesodermal mesenchyme rather than the ectoderm, which typically differentiates into the nervous system. Thus, cells reprogrammed across embryonic layers can also possess the ability to form organoids. Mesenchymal stem cells (MSCs) are of particular interest in regenerative medicine because of their high self-renewal capacity, low immunogenicity, and easy accessibility. Furthermore, MSCs can differentiate across embryonic layers into endodermal and ectodermal lineages [
43], potentially allowing the generation of neural organoids from MSCs. The use of organoids derived via direct reprogramming can notably shorten the culture duration and preserve aging-associated epigenetic signatures [
44]. However, these organoids are deficient in glial cell components.
TSC-derived organoids (such as intestinal organoids) are generated from primary tissue cells. Generally, TSC-derived organoids provide more robust, long-term expanding lines that allow direct studies of cell behavior and characteristics within specified tissues [
28]. These organoids have shorter culture durations and more closely resemble adult tissues, making them better suited for regenerative medicine. However, NSCs have historically been able to form relatively small and unorganized 3D cell aggregates known as neurospheres. In 2024, Hendriks et al. [
45] reported the generation of brain organoids from fetal brain tissue under specific culture conditions. A recent study also reported the generation of brain organoids from immortalized fetal NSCs [
46]. Similar reports on the spinal cord and peripheral nerve organoids are yet to be published. NSCs reside in the ependyma of the fourth ventricle of the mammalian brain and the central canal of the spinal cord; however, adult NSCs have limited potential for differentiation and proliferation. In the future, the generation of spinal cord and peripheral nerve organoids from fetal or adult NSCs may complement existing organoid research.
3.2. Signaling-modulating factors
The key factor affecting the formation and development of organoids
in vitro is the control of intrinsic signaling pathways that facilitate differentiation into various cell lineages. Studies on embryonic development have shown that gradient ecological signals
in vivo influence cell fate determination and patterning through a series of endogenous pathways. Therefore, based on the existing understanding of the spatiotemporal distribution of morphogenetic signals during embryonic development, organoids can be induced to produce specific cell clusters by adding ligands or compounds related to these key patterning signaling pathways at different time points to activate or inhibit these pathways. Commonly targeted pathways in organoid induction include the transforming growth factor (TGF)-β, BMP, WNT, FGF, and SHH pathways [
47], [
48]. Organoids with characteristics of the spinal cord and DRG can be generated by regulating these pathways with inducers and cytokines (
Table 1 [
9], [
11], [
12], [
18], [
22], [
26], [
32], [
33], [
34], [
35], [
36], [
10], [
42], [
49], [
50], [
51]).
For neural organoids, the first step is to guide the fate of the neuroectoderm. TGF-β is an important signal for determining germ layer fate. Unlike the endoderm and mesoderm, neuroectoderm differentiation is triggered by repressed external signals. By early inhibition of key downstream proteins of the TGF-β pathway, such as SMADs, the fate of neuroectodermal cells can be determined. SB-431542 and LDN-193189 are two commonly used compounds that, when added to the culture medium during the early stages of organoid generation, can enhance the conversion of PSCs into the neuroectoderm. After inducing the embryoid body to differentiate into the neuroectoderm, non-guided brain organoids can generate cell components representing the forebrain, midbrain, and hindbrain through embedding in Matrigel and cultivation in a rotating bioreactor, relying on intrinsic signaling for differentiation and maturation [
7], [
45].
Based on the existing understanding of embryonic development processes and the establishment of an
in vitro two-dimensional culture platform for small-molecule screening, researchers have generated specific lineages of cells through the regulation of biochemical signals [
52]. Specifically, for spinal cord organoids that require patterned tail region features, tailing can be induced by providing CHIR (an activator of WNT) and basic fibroblast growth factor (bFGF) [
9], [
12], [
32]. In addition, some induction protocols include RA and the Rho-associated coiled-coil kinase (ROCK) pathway inhibitor Y-27632. RA plays a crucial role in promoting neural differentiation, whereas Y-27632 enhances cell survival and promotes the formation of 3D structures during differentiation process [
9], [
12], [
34].
Mazzara et al. [
11] successfully induced the formation of DRG organoids using SU5402 (an FGF signaling pathway inhibitor), CHIR 99021, and DAPT (a Notch signaling pathway inhibitor). Lu et al. [
26] generated DRG organoids by adding CHIR 99021 and FGF2 to induce NCC formation. They also found that the addition of RA signaling could support the lineage of large-diameter subtype neurons.
Furthermore, in spinal cord organoids, the regulation of BMP and SHH signaling can induce dorsal and ventral spinal cord cell types, respectively [
9], [
18]. However, directly adding signaling regulatory molecules to the culture medium cannot create concentration gradients for these signaling molecules, precluding the formation of a complete spinal cord organoid model with DV spatial structures.
In contrast, less stringent guidance can lead to the formation of a broader range of cell lineages that can be used to study the co-occurrence and interactions between organs. In comparison with the process of co-culturing the resulting organoids into assembloids [
13], this strategy can better recapitulate
in vivo developmental processes. For instance, using FGF and WNT agonists to induce the production of neuromesodermal progenitors (NMPs), organoids can be differentiated to produce astrocytes, sensory motor neurons, and mesoderm-derived cells, successfully reflecting the formation of neuromuscular junctions and neuron-dependent skeletal muscle contractions [
14], [
38]. By performing dual SMAD inhibition, activating WNT, supplementing with FGF2 and RA, and activating SHH at different time points, organoids containing three lineages of neuromuscular and skeletal cells can be generated, which spatially self-organize into three parts and establish connections [
39].
In addition to simulating multi-component organoids that control movement, this approach has been used to describe the co-developmental processes of embryos. By activating WNT and SHH signaling and supplementing with RA, the co-morphogenesis of the spinal cord and vertebral column was simulated
in vitro [
53]. Recently, precise regulation of embryonic FGF and WNT agonists, along with the timely addition of BMP and nodal growth differentiation factor (NODAL) inhibitors at specific time points, was shown to yield organoids with structures such as the notochord and neural tube [
54].
In summary, different signaling pathways produce cells of various lineages through precise spatiotemporal regulation, and the induction of specific organoids requires the in vitro recreation of this regulation by the timely addition of exogenous signaling molecules. Precise temporal regulation requires a deeper understanding of the spatiotemporal specificity of signaling regulation and cell formation during embryonic development.
3.3. Matrix materials for 3D culture
Unlike the uniform mixing of different types of cells in co-culture systems, organoids form 3D structures of self-organized cell clusters with unified functions supported by the extracellular matrix (ECM) [
55]. These structures simulate interactions among different cell populations
in vivo and create specific local microenvironments [
56]. ECM is critical for constructing 3D structures and regulating cell patterning because it provides essential nutrients, adhesion sites, and biophysical cues that guide lineage differentiation [
57]. In most common organoid culture protocols, early suspension culture conditions allow cells to aggregate regionally based on their phenotypes. Subsequently, Matrigel serves as a mechanically supportive ECM that plays a crucial role in the maturation of organoid cells and the formation of spatial structures [
7]. Previous studies have shown that embedding aggregates formed from PSCs in Matrigel can regulate cell lineages, promoting the development of polarized trunk morphologies [
58] and the formation of tubular structures resembling neural tubes
in vitro [
53].
The mechanical parameters of 3D culture matrix materials, including material stiffness, viscoelasticity, stress relaxation, degradation rate, and geometric shape, all influence cellular behavior. Therefore, these parameters should be considered in organoid cultures [
59]. The key factors are the stiffness and viscoelasticity of the ECM [
60]. Viscoelastic properties of the ECM can regulate the patterning of spinal cord organoids derived from human PSCs. Specifically, stiffer hydrogels are more conducive to spinal cord organoid differentiation than softer hydrogels. Moreover, viscoelastic hydrogels promote the regional patterning of human spinal cord organoids more effectively than elastic hydrogels [
61]. In a study of
in vitro neural tube morphogenesis, ECM-mediated substrate adhesion, along with apical contraction of the neuroectoderm, facilitated the folding of the neural tube [
62]. Additionally, the stiffness of the ECM plays a role in the migration of NCCs during the formation of the peripheral nervous system. The stiffness gradient generated by the adjacent placodal tissue allows NCCs to follow this gradient through durotaxis, working in concert with chemotaxis to promote effective directional collective cell migration
in vivo [
63].
Matrigel is the most commonly used natural ECM material for 3D culture, and it offers excellent biocompatibility and a rich composition of ECM components (such as collagen, laminin, and fibronectin) that promote cell attachment, proliferation, and migration as well as neuronal differentiation and axon extension [
64]. However, Matrigel has several drawbacks, including batch-to-batch inconsistencies and limited control over mechanical and biochemical properties [
65]. Additionally, due to the uncertainty of its components and potential immunogenic factors, mouse-derived Matrigel is unsuitable for clinical applications [
64]. Consequently, many studies have focused on synthesizing engineered ECM materials to replace Matrigel.
The utilization of decellularized ECM (dECM) derived from specific tissues has been shown to yield tissue and organ specificity and represents a promising approach for matrix design. The dECM can retain key biomolecules that support the growth of the corresponding tissues and can be enhanced by artificially adding relevant proteins to facilitate the development and construction of organoids that closely mimic natural conditions [
66]. In the cultivation of spinal cord organoids, dECM derived from the brain [
67], spinal cord [
36], and placenta [
68] provide a suitable environment for neuronal growth, promoting the formation of organoids. Research has also shown differences in the physical properties of spinal cord ECM at various developmental stages, with decellularized spinal cord ECM from neonatal rabbits better supporting the proliferation, migration, and differentiation of spinal cord neural progenitor cells into neurons and promoting axon growth and synapse formation, indicating superior performance in comparison with Matrigel in organoid models [
36].
However, dECM still faces issues of inconsistency owing to the extraction methods and sources used, and the sources of brain or spinal cord dECM of human origin are limited. The use of animal-derived dECM requires a balance between the elimination of immunogenicity and the retention of active substances. Therefore, many studies have focused on using natural hydrogel materials such as alginate, hyaluronic acid, and gelatin or fully synthetic materials to synthesize engineered hydrogels with defined compositions [
33], [
61], [
65], [
69], [
70], [
71]. These materials can achieve control over parameters such as viscoelasticity through adjustment of component ratios and crosslinking designs [
61], [
69]. Moreover, these materials can be peptide-functionalized to enhance biochemical cues [
65]. These highly controllable ECM materials can also help determine how different parameters affect organoid cultures [
71]. Nevertheless, significant challenges remain in creating synthetic matrices that can closely replicate the structural complexity, biochemical diversity, and dynamic functions of natural tissues.
4. Formation of refined 3D structures through bioengineering approaches
The inability of self-organization to form refined spatial distributions of cells can lead to structural differences between organoids and normal tissues, thereby hindering the realization of more physiological functions in organoids [
39]. For example, spinal cord organoids cannot form models with a dorsoventral spatial structure or replicate the distribution of nuclei in the spinal cord cross-section. In addition, the axons do not develop into oriented nerve bundles. Combining these organoids with engineering approaches may provide a solution to this challenge. Engineering approaches can be used to create predetermined spatial structures through selective spatial interventions of biomechanical and biochemical signals.
4.1. Establishment of morphogen concentration gradient
Spatial delivery of biochemical cues to guide organoid development through engineering methods may be a promising strategy to address these limitations [
72]. The spatiotemporal gradients of morphogens can induce the heterologous growth and differentiation of organoids, and most studies have employed microfluidic devices to achieve this goal. By forming artificial signaling centers for BMP4, stem cells can generate unique, axially arranged differentiation domains [
73]. Through WNT activation gradients, Rifes et al. [
74] generated neural tissues that exhibited gradual caudalization from the forebrain to the midbrain and then to the hindbrain. Both studies established axial gradients of a single morphogen, thereby achieving axial cell alignment within the organoids. Xue et al. [
31] utilized microfluidic devices to achieve spatiotemporal activation of various morphogens, successfully replicating the dorsoventral differences in the brain and spinal cord regions, and creating a 3D, lumen-like tissue structure that resembles the
in vivo neural tube (
Fig. 3(a)). Microfluidic devices offer a controllable and automated approach to form morphogen gradients and facilitate the maintenance of nutrient perfusion. However, the existing tissue patterning and cellular organization still differ from those of
in vivo tissues, since the complexity of spatiotemporal patterning signals
in vivo has not yet been fully replicated.
Another strategy involves loading morphogens with synthetic materials. Based on this strategy, Wylie et al. [
75] designed 3D protein-patterned scaffolds that spatially controlled the immobilization of different growth factors within hydrogels, thereby directing the differentiation of stem or progenitor cells. By forming composite microspheres with functional sustained-release smoothened agonist (SAG), Xue et al. [
50] generated human spinal cord organoids with dorsoventral characteristics within functionalized composite scaffolds (
Fig. 3(b)). This method is easier to implement and combines the regulation of scaffold materials using physical cues. However, control of complex temporal differentiation could not be achieved with this method.
Co-culture with engineered cells that secrete morphogens during organoid culture is also an effective strategy. Yamada et al. [
76] developed synthetic organizer cells that can self-assemble around ESCs through cell adhesion and express morphogenetic signals in specific structures, thereby generating different morphogen gradients. Similarly, Bosone et al. [
77] achieved patterned generation of assembloids with frontotemporal signatures by co-culturing brain organoids with organizer-like structures that produce FGF8. This method closely resembles the process of morphogen gradient formation by tissue organizer cells
in vivo. It does not require complex equipment or expensive purified morphogens. Moreover, these engineered organizer cells can be further modified to adhere to specific cell surfaces, thereby creating
in vivo morphogen gradients. However, this method can only form a single morphogen gradient and cannot simulate temporal patterning signals.
These studies created axial gradients from the outside in, whereas Afting et al. [
78] utilized microinjection of DNA microbeads into organoids to release morphogens in a light-triggered manner, providing spatially discrete and temporally controllable morphogen gradients within the organoids (
Fig. 3(c)). However, this method requires delivery of microbeads to a target location through microinjection. Achieving high-throughput and precise injection is a problem that must be resolved. Moreover, this procedure may exert mechanical stimulation.
4.2. Geometric confinements and biomechanical cues
Another strategy involves the active regulation of biomechanical cues. Xue et al. [
79] found that during the formation of neuroectodermal patterns, geometric confinement signals mediated by cell shape and cytoskeletal contractile forces guide the patterning of neuroepithelial and neural plate border cells through the BMP–SMAD signaling pathway. Ectopic mechanical activation can disrupt this patterning. Seo et al. [
22] found that pre-patterned geometric constraints can also generate ordered spinal cord organoids along the dorsoventral axis through symmetry breaking (
Fig. 4(a)). The geometric confinement method can also help control the size and morphology of organoids, ensuring consistency. However, the asymmetry of organoids formed through pre-patterned geometric constraints and the cell alignment in the spinal cord differ significantly. In comparison with regulation of biochemical signaling, this method falls short in terms of controllability over tissue patterning and cellular organization.
Artificially synthesized, dynamically variable ECM can provide dynamic physical cues to meet the needs of differentiation and maturation at different stages [
80]. Urciuolo et al. [
81] created a 3D hydrogel-in-hydrogel structure through photopolymerization, allowing soluble photoresponsive polymers to diffuse into the external solid hydrogel upon light activation, and patterned it through 3D printing (
Fig. 4(b)). This structure dynamically applied geometric constraints during the cultivation of spinal cord organoids, ultimately controlling the direction of neuronal axon projection [
81]. This ECM can dynamically simulate the
in vivo environment in terms of physical parameters, such as stiffness and viscoelasticity, and is also capable of guiding the axon direction. However, its effects on the orderly arrangement of cells within the organoids are limited.
Manipulation of mechanical forces may guide the controlled patterning of neuroectodermal cells. Magnetic nanoparticles can provide mechanical stimulation within tissues. Using magnetic nanoparticles, Abdel Fattah et al. [
82] applied precise and localized directional mechanical stimulation to human neuroectodermal stem cells, controlling their position and movement to guide tissue growth and patterning (
Fig. 4(c)). Acoustic radiation force is another potential method for mechanical regulation. Bouyer C achieved a bio-acoustic levitational assembly method using acoustic radiation force, forming multilayered 3D tissue structures from neural progenitor cells and establishing hierarchical connections [
83]. The acoustic virtual 3D scaffold has also enabled scaffold-free culture of tumor organoids, allowing co-cultured immune cells to interact directly with organoids [
84]. Although evidence has demonstrated the beneficial effects of mechanical force stimulation on stem cell function [
85], the precise application of mechanical forces through design to form organoids with specific tissue masses and cellular organizations requires further research.
4.3. 3D bioprinting
3D printing is a method for constructing 3D objects from computer-aided design models, typically using a layer-by-layer manufacturing approach. By utilizing 3D printing technology, micro-patterns can be created to provide biophysical cues. Moreover, in 3D bioprinting, biocompatible cell-laden materials can be used to create cell-laden materials with patterned tissue-like structures. Therefore, bioprinting has been widely applied to scaffolds for spinal cord and peripheral nerve tissue engineering [
86], [
87].
Bioprinting of organoids refers to the technique of using bioinks composed of self-organized organoids or cells capable of self-organizing into organoids after printing along with matrix materials to produce complex tissue structures through precise deposition of building units [
88]. Organoid technology generates cell clusters that more closely resemble the cell types and functions of real tissues, whereas 3D printing technology facilitates the realization of customized complex structures. Organoid bioprinting combines the advantages of both of these methods. Additionally, organoid bioprinting may significantly enhance the scalability and reproducibility of organoid-based models and can be used to manufacture larger-scale organ-like tissues [
89]. For instance, renal organoids generated from PSCs through extrusion-based bioprinting have increased manufacturing throughput ninefold while providing high-quality control in terms of cell number, diameter, and cell viability [
90]. Moreover, the bioprinted line conformation increased the number of nephrons, which may be attributed to a more uniform cell arrangement and thinner tissue, which facilitates nutrient access. Brassard et al. [
91] employed stem cells and organoids as spontaneously self-organizing building blocks and spatially arranged them to form interconnected cellular constructs, thereby creating large-scale cellular structures at the centimeter level.
Furthermore, dynamic and precise control can be achieved by integrating engineered ECM materials and designing biomechanical and biochemical cues [
92], [
93], [
94]. For example, the direction of neuronal axon projections can be controlled by designing bioinks to dynamically apply geometric constraints during the cultivation of spinal cord organoids [
81].
Despite these advantages, the existing organoid bioprinting technologies struggle to simultaneously ensure printability, high resolution, and adaptability to the cellular environment. This limitation restricts the generation of highly complex microscale structures.
5. Potential applications of spinal cord organoids and peripheral nerve organoids for regenerative medicine in neurotrauma
5.1. Cell therapy and tissue engineering
Cell transplantation therapies have rapidly progressed to the clinical trial stage, demonstrating remarkable potential in treating neural diseases such as Parkinson disease, multiple sclerosis, and amyotrophic lateral sclerosis [
95]. Stem cell therapy offers a major advantage since it directly targets the site of nerve injury. This approach facilitates nerve regeneration and repair through multiple mechanisms, including differentiation into functional neurons and glial cells to replace lost or damaged cells, secretion of neurotrophic factors, and modulation of immune responses. Consequently, stem cell therapy holds considerable therapeutic potential for a range of diseases characterized by permanent neurological damage. In the realm of spinal cord and peripheral nervous system disorders, cell transplantation is considered to enhance repair and regeneration in SCI and PNI.
After SCI, neuronal axons break, leading to neuronal death and dysfunction of neural circuits. The existing treatment modalities primarily focus on alleviating inflammation during the acute phase and relieving compression to reduce spinal cord pressure [
5], [
96]. However, the therapeutic outcomes of these modalities are unsatisfactory. Cell transplantation for repairing spinal cord injuries is regarded as one of the most promising therapeutic strategies, with various cell types used for transplantation [
97]. MSCs can participate in immune modulation and neuroprotection. Olfactory ensheathing cells (OECs) and oligodendrocyte precursor cells (OPCs) offer advantages in myelin formation and tissue repair. NSCs and neural progenitor cells may differentiate into neurons, replacing damaged cells, forming local networks at the injury site, and establishing connections with endogenous neurons [
96]. A recent clinical study examined the long-term safety and feasibility of NSC transplantation in the treatment of chronic SCI. Four patients with chronic spinal cord injuries were followed up for up to five years, and two patients showed lasting evidence of improvement in neurological function following treatment. This finding supports the potential of cell therapy for SCI [
98].
However, cell suspension injection therapy still faces challenges, including uncontrolled cell distribution, the potential for tumor formation, and limited capacity to fill spatial defects in the treatment area [
99]. Tissue-engineered grafts can integrate cell transplantation, biomaterials, and cytokines to increase cell survival and differentiation efficiency, thereby enhancing therapeutic outcomes [
100], [
101], [
102]. By integrating 3D printing technology, customized scaffolds that match the shape and size of spinal cord injuries can be created, effectively filling spatial defects [
103]. Additionally, several studies designed multi-modular grafts to target specific spinal cord regions for enhanced tissue repair [
6], [
86], [
104].
In PNI, despite the inherent regenerative capacity, the slow regeneration rate and misdirected axonal growth affect the outcomes of peripheral nerve repair [
4]. Autologous nerve grafting is the gold standard treatment; however, donor nerve sources are limited, and complications at the donor site and nerve mismatches may occur. Current research and treatment methods primarily focus on cell repair, pharmacological interventions, and multi-factorial scaffolds [
105], [
106], [
107]. Tissue-engineered nerve grafts are reliable alternatives to autologous nerves, with living biological cells being a crucial component of nerve tissue engineering [
108], [
109]. SCs promote nerve regeneration due to their ability to support axons and form myelin, particularly in long-segment defects. However, SCs need to be harvested from nerves and have limited proliferative capacity, which restricts their clinical application. NSCs can differentiate into neurons and SC-like cells (SCLCs), secrete various critical neurotrophic factors, and promote angiogenesis, nerve growth, and myelin formation [
110]. Additionally, bone marrow-derived MSCs, adipose-derived MSCs, and skin-derived precursor cells can modulate immune responses and secrete factors that promote nerve regeneration. These cells can also differentiate into various cell types, including SCLCs. These advantages make these cells suitable seed cells for applications in neural tissue engineering [
111].
In summary, the core features of stem cell therapy include its multidirectional differentiation potential and paracrine effects. However, several challenges remain. First, the microenvironment exerts detrimental effects on NSCs. The post-injury microenvironment induces the formation of glial scars, which impede the elongation of functional neuronal axons. In addition, host immune responses can trigger the rejection of transplanted cells. On the other hand, while MSCs possess robust paracrine effects that can modulate the microenvironment, they exhibit limited capacity for neurogenic differentiation. Second, barriers to achieving functional maturation remain unresolved. The efficiency of differentiation after transplantation is often suboptimal, and the integration of transplanted cells into functional neural networks remains challenging. Finally, precise spatiotemporal control is lacking. Current approaches cannot precisely target regions with functional deficits or dynamically adapt to the evolving microenvironment following injury.
5.2. Organoid-based regenerative medicine approaches
Researchers have also focused on organoid transplantation [
112], [
113]. In central nervous system organoid transplantation, transplanted cortical organoids can develop into mature cell types, integrate into the neural circuits of the host, and be activated by optogenetics [
114] (
Fig. 5(a)). After transplantation, human brain organoids respond to host niche factors, aligning the fate of transplanted cells with the implanted brain region [
115], resulting in progressive neuronal differentiation and maturation, microglial integration, and the formation of internal vasculature and functional neural networks [
116] (
Fig. 5(b)). These studies have demonstrated the ability of organoids to survive long after transplantation and integrate into host tissues.
However, research on organoid transplantation into the spinal cord and peripheral nerves is limited. Xu et al. [
42] transplanted spinal cord organoids generated through direct reprogramming of astrocytes into an SCI model (
Fig. 5(c)). These organoids survived, differentiated, migrated, and formed synaptic connections with host neurons post-transplantation, although spontaneous motor function improvement was not significant, possibly due to the lack of spatial architecture of the spinal cord. Sun et al. [
36] produced a decellularized neonatal rabbit spinal cord ECM hydrogel to culture spinal cord motor neuron organoids as grafts, which led to motor function recovery in a rat model of SCI (
Fig. 5(d)). This study highlights the support provided by neonatal rabbit decellularized spinal cord ECM in terms of physical parameters and proteins related to neural development and maturation. These studies have demonstrated the potential of spinal cord organoids in regenerative medicine therapy for SCI and highlighted the possibility of enhancing neural repair through the integration of biomaterials, pharmacological treatments, and other approaches. However, the currently available organoids for transplantation cannot achieve modular precision repair by generating DV differences. Additionally, the lack of
in vitro vascularization may slow the repair process. Future studies may be able to address these challenges by incorporating engineering methods. Moreover, investigating the mechanisms underlying the differentiation of functional neurons and their integration into neural networks after transplantation is essential.
In PNI research, Nishijima et al. [
117] attempted to isolate axons from motor and sensory nerve organoids derived from human iPSCs for use as structural components within conduits; however, this approach did not involve live-cell transplantation. To date, no reports have described the use of neural organoids to promote nerve regeneration after PNI. Only a few studies have reported the use of other organoids containing neuronal and glial cell components to repair epidermal soft tissues [
118]. The heterogeneity of induction protocols and the unclear mechanisms by which NSCs promote repair after PNI may explain the lack of such applications. However, the beneficial effects of SC transplantation on myelination and axonal regeneration after PNI are well-established [
119]. Studies have shown that scaffolds loaded with NSCs can promote nerve regeneration [
120]. Some studies have also generated myelin structures
in vitro under specific conditions in organoid culture [
37]. These findings suggest the potential of neural organoids in the treatment of PNI; however, further research is needed.
Thus, organoid transplantation is a promising regenerative medicine therapy for SCI and PNI because organoids contain complex cell types and can respond to post-transplantation niches for differentiation and maturation, integrate microglia, and show vascularization, all of which can help achieve complex post-transplantation functions. Unlike diseases such as Parkinson disease, which are associated with specific types of neurons, SCI and PNI involve tissue defects with multiple cellular components, making organoid-based grafts more suitable for these conditions. Combining tissue engineering with biomaterials and cytokines can improve survival rates after transplantation and further support repair and regeneration. Future strategies may include the integration of tissue engineering with biomaterials and cytokines to modulate the microenvironment and enhance survival after transplantation, thereby further supporting repair and regeneration. Additionally, engineering approaches that enable highly controlled morphology and tissue structure can facilitate integration of host and graft tissues. Dynamic repair strategies based on the timing of injury can help respond to the highly variable microenvironment following injury.
5.3. Organoid extracellular vesicles
Extracellular vesicles (EVs) secreted by organoids are also considered potential therapeutic agents. EVs derived from cellular sources are rich in bioactive substances that can modulate gene expression and function of target cells, regulate immune responses, and promote neural repair following injury [
121]. One study demonstrated that EVs from bone marrow-derived MSCs delivered via noninvasive intranasal administration could promote vascular regeneration after SCI, thereby improving functional recovery [
122]. Exosomes derived from hypoxia-preconditioned human umbilical vein endothelial cells (HUVECs) preconditioned under hypoxia can enhance the angiogenic activity of MSCs during cell therapy [
123]. Exosomes from SCLCs can promote axonal regeneration and SC proliferation
in vivo [
124].
Many studies have shown that EVs secreted by organoids closely resemble those isolated from human body fluids, with a greater abundance of bioactive substances and higher yields, conferring advantages in immune regulation [
125], [
126]. Moreover, in neural organoids,
in vitro simulations of embryonic development can generate exosomes matching the characteristics of different stages and brain regions [
127]. Exosomes from neural, vascular, and mesenchymal cell-derived organoids may better promote repair and regeneration following neural trauma.
In comparison with cell- and organoid-based transplantation therapies, treatments based on organoid EVs offer several advantages, including rapid regulatory effects, low immunogenicity, no tumorigenic risk, and high delivery efficiency. Additionally, EVs can be combined with transplantation to modulate the immune microenvironment and enhance neural repair by the transplanted cells. However, organoid-derived exosomes may show batch-to-batch variability owing to the heterogeneity of organoids and cellular differentiation processes during in vitro culture. Future research should explore high-throughput automated production of organoids as well as the underlying molecular mechanisms of efficacy.
6. Challenges and future perspectives
6.1. High heterogeneity
Organoid models have shown great potential in disease modeling, mechanistic exploration, drug screening, and regenerative medicine for the treatment of spinal cord and peripheral nerve disorders. However, similar to all organoids, the use of spinal cord and peripheral nerve organoids presents several challenges, including high heterogeneity among neural organoids [
128]. The cellular origin, including the type and activity of primary cells used to generate iPSCs, lack of precision in induction protocols, and absence of standardized culture systems are potential reasons for the observed heterogeneity. To address these issues, researchers have optimized organoid generation protocols, identified synthetic ECM materials, and employed engineered manufacturing platforms [
129]. Additionally, by establishing the cellular transcriptome profiles of neural organoids, new culture protocols can be identified and evaluated [
8].
6.2. Maturation and myelination
Another limitation is that current organoids only partially replicate organ functions. This limitation is partly due to insufficient cell maturity and the inability of existing methods to fully mimic the
in vivo environment. For neural organoids, establishing neural circuits and generating bioelectrical signal stimulation are crucial for neuronal maturation in addition to the biophysical and biochemical signals of the niche [
130]. Several studies have shown that hierarchical neural innervation established through
in vivo transplantation can enhance neuronal maturity and the development of complex branching structures [
131], [
132]. Furthermore, the formation of multi-organoid assemblies or self-organizing multi-tissue organs induced by multidirectional differentiation can produce synaptic connections and projections, potentially better simulating neural pathways
in vitro [
39], [
133]. Simulating the physiological microenvironment through synthetic matrix materials, applying biomechanical stimuli such as micropatterned cues and mechanical forces using electrical stimulation [
134], and optimizing culture conditions during the stages of neural differentiation [
135] may all be directions for enhancing maturation.
The myelin sheath is a multilayered cell membrane structure that wraps around the axons. In the peripheral nervous system, the myelin sheath is primarily produced by SCs, while in the central nervous system, it is primarily generated by oligodendrocytes. Myelination is crucial for neuronal function and plays an important role in nerve repair and regeneration.
In vitro formation of myelin structures is important for studying demyelinating diseases and the process of remyelination. Researchers have generated spinal cord and peripheral nerve organoids with myelin characteristics by combining intercellular communication and mechanical contact under culture conditions that promote the maturation of oligodendrocytes and SCs [
32], [
37], [
136]. These organoids have been used to study multiple sclerosis and Charcot-Marie-Tooth disease. Future studies should aim to address the fact that peripheral nerve organoids lack axonal guidance and cannot form structures similar to myelinated nerve fibers.
6.3. Vascularization and immune microenvironment construction
Vascular and immune cell components play crucial roles in spinal cord and peripheral nerve repair; however, the existing methods struggle to induce or co-culture intrinsic vascular networks. This lack of vascularization limits organoid size due to insufficient oxygen and nutrient supply [
137]. Although
in vivo transplantation generates vascular networks and integrates microglia, it does not fully resolve the issue, since these cells are mainly concentrated in the periphery. In terms of angiogenesis, cultivation of organoids containing both vascular and neural tissues is challenging. One approach involves using 3D-printed mesh channels to simulate the diffusion function of vascular networks, thereby improving the supply of oxygen and nutrients to organoids [
138]. However, this method is limited by the printing resolution and cannot yield microvascular structures. Another primary method is the co-cultivation or culturing of organoids within pre-vascularized matrices [
139]. However, existing vascular organoids cannot form perfusable networks
in vitro and rely on integration with
in vivo vascular networks. Co-culture systems can be used to study the interactions of microglia and macrophages with neural organoids.
6.4. Limitations of disease modeling
Although organoids can replicate certain aspects of central and peripheral nervous system diseases, many disease models still show limitations. For instance, although peripheral nerve organoids can simulate certain aspects of peripheral neuropathies, they often lack the complexity needed to fully capture disease mechanisms such as chronic pain or neuroplasticity changes following injury. These limitations are primarily due to insufficient cellular maturity and absence of certain cell types.
Especially in the field of neurotrauma, although recent research successfully created organoid models of traumatic brain injury using high-intensity focused ultrasound [
140], organoid models that can fully replicate the pathological processes associated with SCI and PNI are currently lacking. This gap is particularly evident in the simulation of changes in the immune microenvironment after acute injury and subsequent secondary damage. To address this issue, future engineering strategies combined with co-culture techniques involving immune cells or co-developing organoids may help develop more accurate models.
6.5. Efficacy and safety for regenerative medicine
Safety and efficacy are critical factors for applications in regenerative medicine. Many scientists believe that transplantation of ESCs and induced iPSCs carries the risk of tumor formation [
97]. Because organoids contain multiple cell types at various developmental stages, their safety requires further investigation. Determining the optimal timing for transplantation and improving the survival rate of organoids after transplantation may enhance their efficacy. The development of tissue engineering materials and advanced manufacturing methods tailored for neural organoids is crucial for more effective translation of organoids in spinal cord and peripheral nerve repair.
CRediT authorship contribution statement
Jiaqi Su: Writing – original draft, Visualization, Validation, Software, Methodology, Investigation, Formal analysis, Conceptualization. Zhiwen Yan: Visualization, Software, Methodology, Investigation, Formal analysis. Xiaoxuan Tang: Visualization, Software, Methodology, Investigation, Formal analysis. Tong Wu: Writing – review & editing, Supervision, Project administration, Investigation, Funding acquisition. Jue Ling: Writing – review & editing, Supervision, Project administration, Investigation, Funding acquisition. Yun Qian: Writing – review & editing, Supervision, Resources, Project administration, Investigation, Funding acquisition, Conceptualization.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This paper is supported by the National Key Research and Development Program of China (2021YFC2400801), the National Natural Science Foundation of China (82372409, 32471412, 32171322), the Sino-German Mobility Programme (M-0699), the Excellent Youth Cultivation Program of Shanghai Sixth People’s Hospital (ynyq202201), the Laboratory Open Fund of Key Technology and Materials in Minimally Invasive Spine Surgery (2024JZWC-ZDA05), the Special Funds for Taishan Scholars Project of Shandong Province (tsqn202211125), and the Natural Science Foundation of Shandong Province (ZR2024JQ026).
PDF
