Models created by neurons derived from immortalized cell lines (e.g., SHSY-5Y; human-derived neuroblastoma cell line) do not provide reliable results, as they show a quite different amyloidogenesis phenotype from degenerated neurons in AD [34]. 1–5 μM Aβ concentrations are required to cause cell death in vitro, while it is known that this concentration is 50 nM in natural brain tissue [35]. In studies with PC-12 cells, Aβ1–42 treatment increased the phosphorylation of tau at 10 µM, but not at lower concentrations [36,37]. In another study, 25 μM Aβ1–42 administration on PC-12 cells increased cell apoptosis [38].

3D in vitro models can mimic the physiology more realistically, as the 3D configuration of neurons affects cytoskeletal dynamics [5]. Similar to AD pathophysiology, there was glutamatergic rather than GABAergic toxicity in human embryonic stem cell (hESC)-derived neurons exposed to oligomeric pre-fibrillar Aβ forms. In this study, time-dependent Aβ accumulations and subsequent cell death were reported in Matrigel-coated ultra-low attachment plates [34]. Induced pluripotent stem cell (iPSC) based models enabled the discovery of some basic differences between healthy control groups and individuals with AD. Researchers have characterized several iPSC lines from donor cells from familial AD and sporadic AD patients. Studies on these cells have revealed differences such as increased Aβ, specifically amyloid β42 (Aβ42) and tau hyperphosphorylation compared to those in iPSCs derived from healthy controls. iPSC lines also reflected additional AD-related pathologies such as increased glycogen synthase kinase-3 beta (GSK3β) activation, increased large endosome number, and accumulation of intraneuronal Aβ oligomers [39].

PSEN1 and PSEN2 mutations are genetic factors that cause early-onset familial AD. In a study conducted by differentiating iPSCs produced from AD patient-derived fibroblasts with these mutations into neurons on gelatin-coated plates, the ratio of Aβ40 and Aβ42 synthesized by these neurons was higher than healthy controls in embryoid bodies [40]. This shows that the model reflects the familial AD pathophysiology.

Since it is impossible to obtain primary human neuronal cells from healthy (or diseased) donors, attention has been drawn to stem-cell-induced neurons for more realistic models. In a model with iPSCs obtained from blood cells of AD patients, 3D neuronal spheroids were created in 100 mm ultra-low-attachment plastic plates by inducing iPSC colonies with ROCK inhibitor and growth factors (epidermal growth factor (EGF), fibroblast growth factor-2 (FGF2), brain-derived neurotrophic factor (BDNF), neurotrophin-3 (NT3)). Researchers compared 3D spheroids with 2D culture and found that while there were differences in Aβ clearance between individual iPSCs in 2D, these differences were less in 3D [5]. Another research group using iPSCs derived from AD patients also generated brain organoids showing AD-like pathophysiologies such as hyperphosphorylated tau protein, Aβ aggregation, and endosome aberration. These pathologies were observed in a time-dependent manner harboring APP duplication or PSEN1 mutation [41]. This was a result revealing the similarity of the model to in vivo. In a similar study with the differentiation of iPSCs into neurons, it was reported that p21-activated kinase-mediated sensing of Aβ oligomers was only in the 3D model created in Puramatrix and F-actin-related proteins were clearly associated with disease processes in 3D [42].

Another in vitro AD model is produced by differentiating iPSCs derived from AD patients with the PSEN1 ΔE9 mutation into astrocytes. They demonstrated the pathological features of AD, such as increased Aβ production, altered cytokine release, and impaired Ca²⁺ homeostasis. These pathologies caused oxidative stress in astrocytes and suppressed neuronal support by reducing lactate production. This study also shows that PSEN1 mutant astrocytes display a serious disease phenotype and may contribute to AD progression [43].

Scaffolds that allow for greater physiological interactions in three dimensions between neurons and glia have also been used in AD models. Microenvironments that mimic the ECM, the presence of biochemical factors, and functional and structural molecules play a direct role in in vitro disease models.

In their study, Ranjan et al. formed a porous poly (lactic-co-glycolic acid) scaffold using a wet electrospinning technique and cultured familial AD-derived iPSCs on these scaffolds. Increased neuronal differentiation and spontaneous enhanced pathogenic Aβ42 and p-tau levels were observed in neurons carrying familial AD mutations, compared to age-matched healthy controls on scaffolds within seven days [44].

Cui et al. investigated the effect of self-assembled peptide hydrogel containing the YIGSR (Tyr-Ile-Gly-Ser-Arg) functional domain on the survival and neuronal differentiation of neural stem cells (NSCs) in in vitro and in vivo AD models. To examine the neuroprotective effect, 40 µM Aβ1–40 was applied to the cells simultaneously with the peptides. The researchers found that self-assembled peptides not only increased cell survival, but also reduced the number of cells undergoing apoptosis. In addition, neuronal differentiation was increased on these peptides when compared to self-assembled peptides without YIGSR. In rat AD models, NSCs with self-assembled peptide containing YIGSR increased cognitive rescue and decreased the number of cells proceeding to apoptosis and the Aβ production [45].

The development of cells in an organoid can also be achieved by using scaffolds. In the study of Choie et al., a 3D organoid was developed with human neural stem cells (hNSC) in Matrigel scaffolds, and extracellular Aβ aggregation and accumulation of tau in the dystrophic neurites and cell soma was observed in vitro for the first-time [46]. AD virus models were also used to understand disease pathophysiology. A 3D AD model was created in a study with Herpes simplex virus type I (HSV-1), which has recently been shown to play a role in AD development pathology. Silk sponge scaffolds with 500 to 600 µm pores were prepared as 6 mm discs, and 2 mm holes were drilled in the center to form annular scaffolds. After the scaffolds were coated with laminin (0.5 mg/mL), iPSC-derived hNSCs were seeded onto the scaffolds, and the next day, type I rat tail collagen gel was added. Organoids cultured for 4 weeks were infected by HSV-1, and Aβ plaque-like formations, gliosis, and neuroinflammation were reported [47].

Basic pathologies of AD were provided in a model tri-culture of neurons, astrocytes, and microglia on a 3D co-centric two-chamber microfluidic platform. High levels of Aβ-producing human neural progenitor cells (hNPCs) with FAD mutations and astrocyte-differentiated cells were cultivated in the central chamber, while microglia were cultivated in the outer angular chamber. The central and angular chambers were connected by migration channels. Using this platform, migration of microglial cells has been observed. Aβ aggregation, p-tau deposition, axonal cleavage, and neuroinflammatory activity such as NO, CCL2, IL8, TNF-α, and IFN-γ release were demonstrated [48]. This model, which is useful in elucidating neuro-glial interactions, can be used not only for AD but also for understanding other disease pathologies such as PD and MS.

Due to the experimental challenges of neurospheroid cultures and the fact that they are affected by variability, including size differences, variable cell culture procedures, including cell density and media composition, the researchers further developed uniformly sized neurospheroids. These neurospheroids were derived from NSCs with FAD mutations and human-derived iPSCs (hiPSCs) with a PDMS microarray platform produced by photolithography for high-throughput drug screening [49]. In the developed platform, dendrite formations were observed in neurospheroids containing mature neurons. After 8 weeks of culturing, these neurospheroids were reported to possess Aβ plaques and p-tau deposits [49].

The disruption of blood-brain barrier (BBB) in AD has been shown to precede disease pathology. In this regard, to investigate the role of BBB in AD, a three-channel microfluidic system was developed. Researchers used NSCs with FAD mutations in AD modeling and hCEMC/D3 cells (brain endothelial cell line) in BBB modeling. To modulate the differentiation periods of these two cells separately, sequential seeding was performed on the developed chip, and the intermediate channel was left filled with air until co-culture was initiated. Human NSCs were cultivated in Matrigel, and hCEMC/D3 cells were cultivated in collagen-filled channels. After the endothelial barrier was formed, the intermediate channel was filled with collagen type 1, and the interaction between differentiated NSCs, and the endothelial layer was ensured by diffusion of soluble factors. It was stated that this model showed pathologies such as increased BBB permeability, decreased endothelial junction markers (claudin-1, claudin-5, zonula occludens-1, occludin, VE-cadherin), increased expression of matrix-metalloproteinase-2 and ROS, and accumulation of Aβ peptides in the vascular endothelium [50].

The researchers investigated the effects of Aβ on accumulations of microglial cells in a microfluidic chemotaxis platform by subjecting the cells to different Aβ42 gradients and surface bound Aβ concentrations. Microglia chemotaxis for Aβ monomers and oligomers was determined to be most effective at two different Aβ42 concentrations (23 pg/mL and 23 ng/mL), and at lower concentrations of Aβ42, microglial migration was further enhanced by monocyte chemoattractant protein-1 (MCP-1) secreted by the microglia [51]. This study created a model to elucidate the microglial pathology in AD.

Various in vitro culture procedures have been developed to model the pathogenesis of PH. The most commonly used cell lines in PD studies are the PC12 cell line derived from rat pheochromocytoma and human-derived SH-SY5Y neuroblastoma cells with ion channels and neurotransmitter receptors that secrete catecholamines such as dopamine and can exhibit DA neuron-like properties [56,57]. Since these cells can be obtained more easily than primary cells and can maintain their structural stability, ensuing genetic manipulations, they provide great advantages in the production of toxicity-induced PD models and the use of these models in pharmaceutical trials.

Neurodegeneration, oxidative stress, loss of function of mitochondria, and formation of protein aggregates are among the known causes of PD. Models created by inducing PD by triggering the conversion of ROS by neurotoxins such as 6-hydroxydopamine (6-OHDA), 1-methyl 4-phenyl 1,2,3,6-tetrahydro-pyridine (MPTP), paraquat and rotenone have led to important discoveries in understanding the molecular pathways that occur during the degeneration and death of DA neurons [58]. One of the catecholaminergic neurotoxins, 6-OHDA, and its noradrenergic analog have been used for more than 30 years in the creation of in vitro and in vivo experimental models for PD studies. Since 6-OHDA has a chemical structure similar to dopamine and norepinephrine, it is recognized with high affinity by the dopamine transporters (DATs) and norepinephrine transporters (NETs) in the plasma membrane. Thus, it causes oxidative stress on the catecholaminergic pathways of the peripheral nervous system (PNS) and central nervous system (CNS) through ROS and hydroxyl radicals and causes degeneration of DA and noradrenergic neurons [59].

A significant change in PC-12 morphology and DNA and chromosome breaks in cells was observed with the apoptosis-triggering effect of 6-OHDA [60]. Similarly, 6-OHDA has been tested with neuroblastoma, cerebellar granule, and primary mesencephalic DA cell lines, which are frequently used as model cell lines in the literature, and the data obtained support that 6-OHDA activates oxidative stress and apoptotic pathways in 2D culture [61,62].

MPTP, unlike 6-OHDA, can quickly pass through the BBB and be metabolized to its active form. 1-methyl-4-phenylpyridinium (MPP+), which is stored in neurons via the DAT, and as a result, it plays a role in the pathophysiology of mitochondrial dysfunction by affecting the mitochondrial electron transport chain [63].

Many researchers have generated iPSCs derived from PD patients to obtain cells with susceptibility alleles in PD-related genes such as leucine-rich repeat kinase 2 (LRRK2), glucocerebrosidase (GBA), PARKIN, α-Syn, synaptophysin, and phosphatase and tensin homolog-(PTEN)-induced gene (PINK1) [48].

Developing an idiopathic PD model for the first time with organoids generated from iPSCs, the researchers were able to demonstrate a complex network of DA neurons. They observed statistical differences in expression levels of the LIM homeobox transcription factor alpha and tyrosine hydroxylase genes between organoids developed from healthy and PD cells [64]. In a study with iPSCs developed from a patient with a mutation in the LRRK2 gene, 3D neuro-ectodermal and intestinal organoids were created to examine the brain-gut relationship dependent on LRRK2. As a result of the differences seen in the gene expression profiles of these organoids, the researchers stated that PD may affect intestinal functions [65].

In a study aiming to standardize reproducible organoid production, human neuronal progenitor cells (NPCs) were cultured for 8 days in 96-well ultra-low adhesion plates until they formed colonies. On day 8, the colonies were embedded in geltrex (a gel containing ECM proteins such as laminin, fibronectin, collagen, entactin, and heparin sulfate proteoglycans) droplets, and maturation was initiated by differentiation media in 24-well ultra-low adhesion plates under shaking media for 30 days. This protocol developed standardized organoids that retained the basic aspects of midbrain development, such as DA neuron and astrocyte differentiation and compartmentalization, and can be utilized in PD models [66].

In a PD model developed using 3D bioprinting, DA neurons were cultured in ultra-short self-assembled tetrapeptide Ac-Ile-Val-Phe-Lys-NH2 (IVFK) and Ac-Ile-Val-Cha-Lys-NH2 (IVZK) scaffolds. DA neurons encapsulated in tetrapeptide gels responded to 6-OHDA. They recorded spontaneous electrophysiological activity of encapsulated DA neurons by multi-electrode arrays. The model supported vascularization (by human umbilical vein endothelial cells (HUVECs)), and neuronal culture was continued for more than one month [67].

In another study, PD was modeled by culturing SH-SY5Y and U-87MG glioblastoma cell lines on polyacrylonitrile (PAN) and Jefamine® doped polyacrylonitrile (PJ) electrospun scaffolds. PAN and PJ nanofibrous scaffolds promoted proliferation, differentiation, and 3D interaction of DA neuron and astrocyte-like cells and increased neuronal survival against PD mimetic toxin treatments, including rotenone, MPTP, MG132 proteasome inhibitor and buthionine sulphoximine (BSO) [68].

Researchers developed a different approach using iPSC, genetic, and scaffold technologies together in a model. Human iPSC-induced DA neurons derived from healthy people and PD patients harboring LRRK2/GBA mutations were cultured on 3D silk scaffolds produced by the melt-molding technique [69]. Sodium chloride particles with a diameter of 415–500 μm were added to the silk solution and removed by washing, creating pores on the silk scaffolds. The 3D platform supported DA neuronal network formation, whereas α-Syn and purine metabolite profiles showed changes in PD-derived cells, offering a realistic approach to PD pathophysiology [69].

There are several microfluidic models for PD. One of them is the model in which primary cortical neurons were cultivated on a microfluidic platform in the presence of recombinant α-Syn fibrils. They reported that neurons were able to internalize the α-Syn fibrils and transported them along axons resembling the characteristic spreading pattern of LBs [70].

Differentiating neuroepithelial cells derived from iPSCs into DA neurons within phase-guided microtiter plate format bioreactors brought a distinct perspective to PD modeling. Cells differentiated to functional DA neurons in a Matrigel-loaded microfluidic bioreactor with differentiation media in 30 days, and 19% showed high tyrosine hydroxylase levels [71].

In another study using the same type of microfluidic reactor, patient-specific PD-derived iPSCs carrying the LRRK2-G2019S mutation were differentiated into DA neurons. The DA differentiation, neurite length, and branching complexity in LRRK2-G2019S neurons decreased, and cell death increased due to changing mitochondrial morphology. DA phenotypes were found to be improved with the administration of an LRRK2 inhibitor [72].

The differentiation of PD-derived human neuroepithelial stem cells into DA neurons was performed on an automated cell culture platform integrating the microtiter plates with a robotic system for long-term monitoring and maintenance of the culture system. This system, called Pelican, has the potential to automate patient-specific disease modeling, enabling a versatile set of in vitro analyses [73].

Different toxins or immune-mediated inducers are frequently used in demyelination studies. One of the toxin-based demyelination models is the use of lysolecithin (LPC), which disrupts cell membranes, including myelin, by acting on lipid bilayers to form micelles. It has been shown to damage directly the myelin layer in vitro [85–88].

Lipopolysaccharide (LPS), which induces innate immune cells, is also used in modeling MS pathology [89–92]. Researchers investigating the effect of dimethyl fumarate (DMF) used in the treatment of relapsing-remitting MS in vitro used the oli-neu (primary mouse oligodendrocyte) cell line. By applying LPS-mediated inflammation to cells, researchers demonstrated that DMF induces differentiation of oli-neu cells into mature and potentially myelinating cells, and it has anti-inflammatory and antioxidant effects [93].

The cuprizone demyelination model is of particular interest among toxic models [94]. Cuprizone [bis (cyclohexanone) oxaldihydrazone] is a copper chelator. After cuprizone administration, cytochrome oxidase and monoamine oxidase are inhibited in mature oligodendrocytes, resulting in apoptosis due to mitochondrial damage. Cerebral microglia activation is observed in vivo as a result of the demyelination of axons and the production of tumor necrosis factor-alpha (TNFα) and interferon-gamma (IFNγ) by oligodendrocytes is increased [95–97].

In vitro systems that culture OPCs with CNS or PNS neurons are relatively simple, inexpensive, high-throughput models [98]. Oligodendrocytes express genes encoding cell-specific myelin structural proteins [99–101]. Isolated rat and mouse primary oligodendrocytes were shown to express most of the proteins and lipids associated with myelin membranes, but primary cultures of human oligodendrocytes are difficult to study due to the difficulty of reaching live human brain tissue. For this reason, immortalized human oligodendrocyte cell lines are preferred [82]. The human oligodendrocyte cell line (HOG) is a cell line created from a surgically removed oligodendroglioma and has been shown to express myelin basic protein (MBP) and some oligodendrocyte-specific proteins such as A2B5 [102]. It is frequently used in myelination and remyelination studies [103–105].

Myelinated brain organoids have recently turned into a new and important model system for myelination. Human cortical organoids contain GABAergic interneurons, a small number of oligodendrocytes, and OPCs located in the ventral forebrain, mimicking the dorsal forebrain composition [106]. A single-cell sequencing study reported that oligodendrocytes in ventral organoids go through developmental stages similar to those observed in the developing human brain [107]. During ventral organoid differentiation, with the addition of growth factors such as platelet-derived growth factor (PDGF-AA), Insulin-like growth factor 1 (IGF1), and NT3, the number of OPCs increases for about 2 months [108].

Neurons derived from iPSCs can be co-cultivated with iPSC-derived or primary OPCs and oligodendrocytes. Mixed culture approaches like this could enable the investigation of interactions between oligodendroglia and neurons. A group working with hiPSCs studied oligodendrocyte developmental stages by developing human oligodendrocyte spheroids in microwells. The researchers reported that cell morphologies change as they mature and begin to myelinate neurons by 100 days. They also showed that LPC causes oligodendrocyte loss, a process that can be monitored with live imaging [107].

Another study on hiPSCs reproducibly obtained oligodendrocytes and myelin in ‘oligocortical spheroids’ using MSC-derived extracellular vesicles (EVs). Myelin-related protein expressions were reached at week 20, and myelin structures were observed at week 30 [109].

Chesnut et al. developed another iPSC-derived brain organoid that can be used to study myelin damage. They examined the damage by applying multiple toxins to examine the myelination disruption on the brain organoid containing highly myelinated axons they developed in eight weeks in non-coated plates under constant shaking at 88 rpm. Researchers who tried cuprizone, bisphenol A, tris(1,3-dichloro-2-propyl) phosphate, and ibuprofen determined that myelination was affected only in cuprizone application [110]. All spheroid studies revealed that co-culture conditions and interactions of oligodendrocytes with neurons and/or astrocytes are important for mature myelination in in vitro scaffold-free culture conditions.

Studies to investigate the need for axonal signals to initiate oligodendrocyte myelination have shown that myelination does not depend solely on co-culture with neurons, but oligodendrocytes can also form myelin on surfaces such as paraformaldehyde-fixed axons [111].

The development of scaffolds that mimic axons has prompted the investigation of their effects on oligodendrocytes. With this approach, OPCs or oligodendrocytes are studied by culturing on nanofibers or microcapillary arrays. Such synthetic approaches are important not only to screen drugs for multiple sclerosis therapeutics but also to screen for oligodendrocyte association of all NDs.

Appropriate biomaterials that mimic the structure of axons and the three-dimensional microenvironment in the brain with their fibrous microarchitecture can also support cell migration, attachment, and survival, and increase the success of ND models as well as therapeutic studies. Electrospinning is an advantageous method for the reproducible and modifiable production of the physical properties of the extracellular matrix (e.g., pore and fiber size), chemical properties (e.g., cell adhesion proteins), and mechanical properties (e.g., Elasticity). In the literature, beneficial effects of electrospun fiber scaffolds on neural cell proliferation, viability, differentiation, and neurite formation have been demonstrated [112–114]. Furthermore, the functionalization of electrospun scaffolds with ECM components can control cell attachment and neurite outgrowth.

Lee et al. reported that the use of nanofibers offers occasions to study myelination mechanisms and is a complementary approach to induce myelination without requiring the presence of co-cultured neurons. Nanofiber scaffolds facilitate the investigation of oligodendrocyte dynamics and enable the testing of therapeutic agents without confounding neuronal effects. As demonstrated in the study, OPCs and oligodendrocytes are sensitive to the minimum fiber diameter threshold of 0.4 µm to initiate myelination-related processes (Fig. 2) [83].

A research group has also developed synthetic mimics of biological axons that represent key geometric and mechanical characteristics by 3D printing of polydimethylsiloxane (PDMS), poly (2-hydroxyethyl methacrylate) (poly-HEMA) and polyhexanediol diacrylate-co-star-polyethylene glycol (poly (HDDA-co-star PEG)) by projection micro-stereolithography. The fiber surfaces were coated with laminin and poly-D-lysine. These artificial axons reflected the mechanical cues presented by neurons in vivo with similar stiffness to axons, enabling the study of myelination. Researchers reported that laminin-coated pillars were more ensheathed by oligodendrocytes compared to poly-D-lysine ones. For ~16 μm diameter laminin-coated pillars, the artificial axon with a stiffness of 140 kPa showed approximately three times more oligodendrocyte interaction compared to those with a stiffness of 0.4 kPa. Also, more oligodendrocyte interactions were observed in 10 µm axons compared to 20 µm diameter axons [115].

In addition to fibrous microstructures, myelination, and demyelination modeling studies are also carried out in hydrogels that allow cells to move in 3D. It was reported that myelination increased in Schwann cells co-cultivated with dorsal root ganglion (DRG) cells in gelatin meth acryloyl (GelMA) hydrogels with Young’s modulus of 32.6 ± 1.9 kPa and mean pore size of 10.3 μm. Schwann cells had a higher rate of myelination on GelMA compared to poly-D-lysine and collagen-coated surfaces [116].

Co-culture systems using rat DRG/Schwann cells are widely used to fundamentally study myelination in vitro. Researchers developing co-culture using polyvinylidene fluoride-trifluoro ethylene (PVDF-TrFE) scaffolds showed that aligned fibers-oriented Schwann cells and neurites and supported longer neurite extension of the DRG compared to cultures without scaffolds and monoculture. In addition, they reported that myelin structures formed around neurites in scaffolds and emphasized the importance of oriented fibers in myelination studies [117].

In another study, OPCs cultured on collagen microspheres differentiated into oligodendrocytes and formed myelin sheath in co-culture with DRG neurons [118]. This study shows that microspheres produced from biocompatible materials can also be used in myelination and neurite modeling studies and provide a supportive environment for cell growth.

Scaffold-based myelination models provide substantial evidence of the importance of mechanical as well as biochemical regulation of myelination.

Microfluidic devices are advantageous platforms for co-culturing neurons and oligodendrocytes in in vitro myelination models. Spontaneous formation and random orientation of myelin sheaths have made the analysis of myelination difficult in systems reported to be used in scaffold-based and scaffold-free myelination models. In organ-on-chip models, myelin formations can be examined by channeling the chip. In a study performed on the 3D-printed microchip, oligodendrocytes co-cultivated with DRG neurons interacted with neurites extending into an independent channel from the somas. Myelination capacity was demonstrated by overlapping MBP- and βIII-tubulin-positive staining and MBP segments were reported to be located along neurites. The myelin segment distribution followed the neurite network distribution, and the differences in myelin appearance were evident when compared with the classical culture [119].

Sharma et al. developed a microphysiological biomimetic chip that demonstrated neurite outgrowth (0.4–0.55 µm), myelination of Schwann cells (4.7%), and assessment of nerve conduction velocity (0.13–0.28 m/s) [120]. They created hydrogel scaffolds using micro-photolithography. Cell adhesion restrictive surfaces were created using 30% w/v polyethylene glycol dimethacrylate 1000 and 1.1 mM lithium phenyl-2,4,6-trimethylbenzoylphosphinate solution. After the formed channels were filled with Matrigel, motor neurons and Schwann cell co-cultured spheroids were transferred into the chip. For 4 weeks, the elongation of neurites along the canal and myelin formation of Schwann cells were monitored and evaluated [120].

Researchers who developed another 3D neuron culture platform on a microfluidic chip (Neuro-IMPACT) modeled neuronal network, BBB, and myelination processes in this high-throughput scanning platform [121]. The developed chip enabled 3D neuron culture for a long time and enabled the examination of neuro-endothelial and neuro-glial interactions on the same platform. The functionality of the model was demonstrated by BBB permeability testing, imaging of myelin sheaths, and observing neuronal functionality with intracellular calcium changes.

Table 1 gives a summary of the selected studied in vitro models of ND.