Electrical stimulation induced structural 3D human engineered neural tissue with well-developed neuronal network and functional connectivity

Objective. Three-dimensional (3D) neural tissue engineering is expected to provide new stride in developing neural disease models and functional substitutes to aid in the treatment of central nervous system injury. We have previously detailed an electrical stimulation (ES) system to generate 3D mouse engineered neural tissue (mENT) in vitro. However, ES-induced human ENT (hENT) has not previously been either investigated or identified in structural and functional manner. Here, we applied ES as a stimulator to regulate human neural stem cells in 3D Matrigel, explored the components and functional properties of hENTs. Approach. By immunofluorescence chemical staining and electron microscope imaging, we evaluated the effects of ES on (1) neuronal differentiation and maturation, (2) neurites outgrowth and alignment in hENT, (3) formation of synapses and myelin sheaths in hENT. We further investigated the formation of synaptic connections between ex-vivo-fused mouse and human tissue. We used calcium imaging to detect activities of neurons in hENT culture. Results. ES could induce neuronal differentiation, the orderly growth of neurites and the maturation of neuron subtypes to construct a well-developed neuronal network with synapses and myelin sheaths. Most importantly, we discovered that raising extracellular K+ concentration resulted the increasing neuronal excitability in the hENT, indicating electrical activities in neuronal cells. Significance. We applied ES to generate the organised 3D hENTs and identified them in both structural and functional manner.


Introduction
Neural injuries and neurodegenerative diseases are great challenges in clinical practices. To satisfy these needs and deepening our understanding of neuroscience and stem cell research, in recent years, neural regeneration particularly engineered neural tissue (ENT) attracts lots of attention in the field (Harris et al 2020, Burrell et al 2022. ENT research is a multidisciplinary research field based on medicine, materials science, and engineering (Baklaushev et al 2019). It can be constructed in vitro using stem cells as seeds, partially reflecting the physiological structure and functional features of neural tissue in vivo. By providing a highly similar physiological system, ENT is not only a great model for neurodevelopmental research, but also a platform for high throughput screening of drug discovery, and a transplantable substitute for the treatment of neural injury diseases (George et al 2020).
Due to the extreme complexity of neural tissue, there are still some obstacles that need to be overcame on the way of constructing ENT in vitro, such as the induction of neural differentiation, the orderly growth of neural processes, the formation of synaptic connections and myelin sheaths, and the continuity of tissue vitality (Struzyna et al 2015). Therefore, identifying the most efficient way to induce neural differentiation of stem cells, to guide the orderly growth of neurites, and to get neuronal networks with functional synapses and myelin sheaths are all key challenges for neural tissue construction.
Neural stem cells (NSCs) are multipotent stem cells with the potentiation for neural repair. By optimizing conditions of culture system, NSCs could differentiate into neurons, astrocytes, and oligodendrocytes, which are the most important cell types in the central nervous system (Ostermann et al 2019, Nolbrant et al 2020. The physiological electric field (EF), a type of electrical stimulation (ES), with a range of 1-600 mV mm −1 can elevate cell proliferation at stimulated sites during neural development and regeneration (Erickson andNuccitelli 1984, Thompson et al 2014). As an intrinsic biophysical regulator, ES plays an important role in neural development. Previous studies have shown that the generation of ES can be closely related to the geometric changes of tissue morphology during the development (Dantzker andCallaway 1998, Moore andSheetz 2011). Moreover, ES could direct cell migration (Meng et al 2011a, Cao et al 2013, cause morphological changes of neural linage cells (Yamada et al 2007, Zhu et al 2019, Manousiouthakis et al 2022, regulate the process of neural development and neurogenesis (Medvedeva and Pierani 2020), and contribute to spatial patterning maturation in adult brain (Arias-Carrión et al 2004). For instance, endogenous EFs could promoted the axon initiation in the subplate and dendritogenesis in the marginal zone (Medvedeva and Pierani 2020). ES could also enhance the synaptic activities of the cortical neural network, particularly when it was applied by using a multielectrode array (Poli and Massobrio 2018). Therefore, ES has potential in resembling the neural tissue in vitro by regulating synaptogenesis and dendritogenesis, promoting the neurite outgrowth, increasing the number of branch points in the neurite and orienting axon growths (Yu et al 2022).
In our previous studies, we have developed a three-dimensional (3D) culture system by using the optimized ES protocol. The effects of ES have been evaluated in inducing the differentiation of mNSCs into neurons and the acquisition of long neurites. These optimized conditions may allow a well-developed neuronal network to be established within hydrogel droplets (Meng et al 2020).
In the current research, we tend to apply ES for the culture of human NSCs (hNSCs) to construct 3D human ENT (hENT) in vitro. Using ES to provide a platform for controlling and directing the neuronal differentiation, we fabricated hENT in spheroid pattern. We identified that ES sufficiently provided guidance of neuronal differentiation during its early development and determined the cell fate. Moreover, two main populations of neurons in cerebral cortex, glutamatergic excitatory neurons and GABAergic interneurons, were presented in hENT. Furthermore, the formation of highly branched neurites, synapses and myelin sheaths indicated that the complex neural networks were created in 3D hENT. Finally, we proved that the cells in hENT could form synaptic connections with cells derived from cocultured tissues, indicating a great potentiation to develop active neural networks. In addition to forming structural connections, raising extracellular K + concentration resulted the increasing neuronal excitability in the hENT, indicating electrical activities in neuronal cells, presenting an opportunity to accurately study the biology of human neural tissue.

Preparation of hNSCs and mNSCs
hNSCs were isolated from brain of 10-15 weeks old human aborted fetuses and cultured using the neurosphere method as described previously with minor modifications (Meng et al 2007). Briefly, fetal brain tissue was micro-dissected and mechanically triturated to a single cell suspension and resuspended in serum-free-DMEM-F12 containing the N2/B27 supplements, human recombinant bFGF (20 ng ml −1 ) and human recombinant EGF (20 ng ml −1 ). Collected single cells were cultivated in Nunc T25 culture flasks (5000 cells cm −2 ), grown in culture medium to allow the formation of neurospheres in 5% CO 2 at 37 • C. The neurospheres were digested to single cells using Accutase before replating the cells under the same conditions. The medium was refreshed every 2-3 d.
After at least five passages, neurospheres were characterized before using them for experiments. The hNSCs were characterized by assessing the expression of NSC markers-Nestin and Musashi1 and lineage-specific markers-Tuj1, GFAP and MBP by immunocytochemistry.

Construction of hENT
hENT came from the culture of hNSCs mixing with naturally derived Matrigel droplet, which is widely used for neural tissue engineering (Murphy et al 2017). Electrotactic chambers and hENT-constructed protocol were established as described previously with minor modifications (Meng et al 2011a(Meng et al , 2020. Briefly, after at least eight passages, hNSCs formed neurospheres were mixed well in 50 µl Matrigel at a ratio of 1:5 on ice, and then the mixture was incubated at 37 • C to allow gelatinization. EFs were applied to the 3D culture in an electrotactic chamber via 5% FBS agar bridges, and the cultures were treated with a physiological strength 150 mV mm −1 DC in the BNb medium (DMEM/F12 basic medium with N2 (100×), B27 supplement (50×) and 20 ng ml −1 bFGF added). The neurospheres in 3D Matrigel were exposed to EFs for 30 min per day for three consecutive days, then hENTs were transferred into suspension culture on a shaker with a speed of 85 rpm until the end of the experiment. Culture medium was changed every 2-3 d. HENTs were analyzed at different timepoints by immunocytochemical staining following by confocal imaging to generate surface rendered 3D images for further analysis.

Immunofluorescence analysis
For hNSCs immunostaining, petri dishes were coated with poly-lysine at 4 • C overnight and dried in the hood. On the next day, cell suspension was centrifuged to collect neurospheres. Those cell pellets were plated onto previously prepared polylysine-coated petri dishes for immunostaining. About 4% paraformaldehyde was added to each dish and then incubated at room temperature for 20 min. Following that, 4% paraformaldehyde was aspirated, and samples were rinsed thrice with PBS. About 1 ml of permeabilize solution (0.03% Triton X-100 and 0.1% BSA in PBS) was added to each sample following by a 10 min incubation. Samples were rinsed by PBS for three times to remove any excess permeabilize solution. 1 ml blocking buffer (5% BSA in PBS) per well was used before applying primary antibodies (anti Nestin or anti Musashi-1). After overnight incubation, samples were washed thrice following by another 60 min of incubation for the secondary antibody (1:100; Life Technology, OR, USA). All samples were counter stained with Hoechst33342 for 10 min before imaging.
The protocol of 3D immunofluorescence analysis was following the method as described previously (Meng et al 2020). Samples were briefly rinsed in PBS and then fixed in 4% paraformaldehyde for 12 h at 4 • C. Then 3D tissue was incubated in blocking solution (1× PBS containing 0.1% Triton X-100, 2% BSA and 2% gelatin) for 30 min, following by permeabilization solution (0.5% Triton X-100 in PBS) treatment for another 30 min and then incubating with primary antibodies (table 1) at 4 • C overnight. After extensive washing with blocking solution, cells were incubated with secondary antibodies at 4 • C overnight, then washed with PBS overnight on the shaker. Samples were counterstained with Hoechst33342 for 10 min before imaging. All antibodies were diluted in the blocking solution (described above). Secondary antibodies included Alexa Fluor 488-conjugated goat anti-mouse, anti-rabbit IgG or anti-chicken IgY and Alexa Fluor 594/555-conjugated goat anti-mouse or anti-rabbit IgG (1:200; Life Technology, OR, USA). 0.01 M PBS was used to instead of primary antibody as negative control in immunofluorescence staining.
All antibodies are listed in table 1.

Preparing samples for the electron microscope imaging
The preparation protocol was carried out as described previously (Meng et al 2020). 3D hENTs were fixed by immersion in a mixture of 2% paraformaldehyde and 3% glutaraldehyde in PBS at 4 • C overnight. After three-times rinsing in PBS, samples were then incubated in 1% osmium tetroxide for 1 h post-fixation at room temperature and briefly rinsed in distilled water. Cultures were then dehydrated in ascending series of ethanol following by 100% acetone and consecutively immersed in two separate baths of 100% epoxy resin at 60 • C for 12 h, separately. Ultrathin sections (50 nm) were then prepared, stained by uranyl acetate solution followed by lead citrate, and imaged using a transmission electron microscope (EP5018/40/Tecnai Spirit Biotwin 120KV, FEI Czech Republic s.r.o, The Netherlands). 2.6. Establishing the hENT/EGFP-mouse engineered neural tissue (mENT) and the hENT/EGFP mouse cerebral slice 3D co-culture system Before generating co-culture system, petri dishes were coated with 0.05 mg ml −1 collagen I in basic medium at 4 • C overnight, then rinsed with PBS twice and allow them to dry in the hood. About 1 ml of complete BNb medium was added into each dish for 15 min balancing in the incubator. Preparation of EGFP-mENT and hENT coculture: two 50 µl-Matrigel-droplets mixed with hNSCs and EGFP-mNSCs respectively were plated in pre-established electrotactic chamber and left a certain gap between them. The cultures were stimulated with 150 mV mm −1 EF to construct hENT and EGFP-mENT in 3D Matrigel for 3 d. After removing the EF, hENT and EGFP-mENT continued to be co-cultured for another 14 d or 21 d.
Preparation of organotypic cerebral brain slices and hENT co-culture: in pre-established electrotactic chamber, hNSCs which mixed with 50 µl Matrigel droplets were treated with 3 d of 150 mV mm −1 ESs. Cerebral slices were prepared by using the protocol we published in 2011 (Meng et al 2011b). In brief, mice brains were dissected from 8 d-old EGFP mouse pups. Subsequently, brains were sliced into 400 mmsections with McIlwain tissue chopper (The Mickle Laboratory Engineering Co. Ltd, UK) and kept in cold basic medium. Under a dissecting microscope, each selected 400 µM cerebral slice was placed next to the hENT with a certain gap between two tissues. Then the co-culture was covered by a drop of the Matrigel (50 µl). Petri dishes with the cerebral slices were cultured in an incubator at 37 • C with 5% CO 2 . The medium consists of HEPES buffer (25 mM), fetal calf serum (15%) and DMEM/F-12 and was changed twice weekly.

Image analysis
To detect the cell composition and structural features of hENT, at least three parallel experiments were performed in each set of experiment. Neurons (Tuj1 or MAP2-positive cells), astrocytes (GFAP-positive cells) and subtype neurons (Tbr1, VGLUTI, γaminobutyric acid (GABA)-positive cells) were counted in 15 random selected visual fields derived from 3 replicates 3D-culture samples for each experiment, then quantified as a percentage of total Hoechstpositive nuclei.
Fluorescence microscopy (Olympus IX71, Japan) and confocal microscopy (FV3000, Olympus, Japan) were used for imaging. Z-stack confocal imaging was performed on FV3000, Olympus microscope. ImageJ software (National Institutes of Health, https:// imagej.nih.gov/ij/) or its plugin (3D viewer) was used to analyze the fluorescence intensity and generate surface rendered 3D images.

Measurement of the length of neurites and relative angles between neurites and the axis of EF
The length of neurites was measured by using ImageJ software with the plug-in, Simple Neurite Tracer (Plugins-Segmentation-Simple Neurite Tracer). The calibration ruler and the scale were setup in the unit of 1 µm. Neurites were traced, measured and labeled individually. Statistical analysis was carried out by using Origin Parametric Software (Origin Lab), and results were presented as mean ± standard deviation.
To measure relative angles between neurites and the axis of EF, target neurons were marked by Line Tool in ImageJ software, and the analytic-measure function was used to obtain an angle between the straight line of a target neuron and the positive half X axis of the EF. The Origin Parametric Software was used for statistical quantification according to four sections: 40-60, 60-80, 80-100 and 100-120 degrees of angles. The data were presented in terms of the proportion of different cell types in the four sections respectively.

Quantification of synapsin1 immunofluorescent intensity
To quantify the synapsin1 immunofluorescent intensity in hENT, 15 randomly selected fields in each group were imaged on a confocal microscope at 40× magnification at uniform exposure levels in both the control group and ES group (at least three replicate parallel experiments). Before measuring fluorescence intensity, background is determined based on the average intensity in several areas without any detectable immunoreactivity. The expression of syn-apsin1 in each image was measured by ImageJ software to determine the immunofluorescent intensity per unit area. To control the background noise, images were converted to 8 bit, adjusted the threshold and selected the best algorithm for the statistics according to Auto Threshold. Data were extracted by ImageJ software and analyzed by GraphPad Prism software.

Calcium imaging
To label calcium ions, hENTs were washed twice with PBS, then adding Fluo-3/AM-PBS solution with a concentration of 5 µM. All samples were incubated at 37 • C for 30 min and preventing light exposure. After washing all samples with PBS twice, the samples were incubated for 20 min at 37 • C to ensure that Fluo-3/AM was completely hydrolyzed into Fluo-3.
Optical recording was performed by a computercontrolled fluorescence imaging system with a highly sensitive CCD camera attached to an inverted phase contrast microscope (IX71, Olympus). 20× water immersion objective (Olympus) was used in this study. Fluo-3 was excited at 480 nm and fluorescence was recorded under emission wavelength of 525 nm. The baseline fluorescence values were first recorded for 60 s, and the intracellular Ca 2+ fluorescence at the distal end of hENT were recorded immediately after adding the high K + solution into the opposite end of hENT, with a total fluorescence recording time of 180 s. Images were taken at a frame rate of 2 Hz with an exposure time of 50 ms. ImageJ software was used to analyze regions of interest (ROI) of calcium ion imaging fluorescence signal and to collect data. OriginPro 8.5 software was used for further analyzing and data visualization. A total 16 neurons were analyzed across 3 fields of view from hENTs (n = 3 independent experiments). ROIs were drawn on the cell soma to determine fluorescence intensity changes over time. Data was extracted by ImageJ software, importing into the Excel format to match each ROI with time point (seconds from the beginning, calculated by frame number). Then processed data sheet was imported into OriginPro, selected all the data, right clicked plot and selected line to pop up graph.

Statistical analysis
Statistical analysis was carried out by using Origin Parametric software (Origin Lab). All data were presented as the mean ± SEM. Statistical analyses were performed by Student's t-test. A probability value (p) less than 0.05 was considered statistically significant.

The identification of hNSCs
Previous studies demonstrated that Nestin-positive cells are NSCs with self-renewal capacity and multipotential differentiation properties (Meng et al 2007). Here, we showed that the hNSC populations (figures 1(A) and (B)) used in this study were mostly Nestin-positive (figure 1(C)) and Musashi1positive cells (figure 1(D)), which are NSC specific markers. Almost 97.72 ± 0.62% of hNSCs showed immunoreactivity towards Nestin. When the mitogenic factors were removed, hNSCs differentiated into Tuj1-positive neurons, GFAP-positive astrocytes and MBP-positive oligodendrocytes respectively (figures 1(E) and (F)). This result confirmed that hNSCs used in the current study have multiple differentiation potentials.

Stimulation of 150 mV mm −1 EF for three consecutive days effectively induced neuronal differentiation in 3D Matrigel culture
When hNSCs were exposed to 150 mV mm −1 EFs for five consecutive days, cells showed poor condition together with increasing amount of cell debris. We therefore optimized the duration of ES to induce the highest neuronal differentiation rate while maintaining healthy cell condition as well as cell growing actively.
Our previous research suggested that the ES could induce the sustained expression of the transcription factor Ascl1 which is critical for neuronal cellfate-decision therefore inducing neuronal differentiation. We thus investigated the expression of Ascl1 in hNSCs after 3 d of ES. We found that the expression of Ascl1 were significantly enhanced among these cells compared to un-stimulated cells (figures 2(B) and (C)).
On the 7th day after ES, 18.67 ± 2.69% of stimulated hNSCs in 3D Matrigel were differentiated into Tuj1-positive immature neurons. Among them, 7.21 ± 5.41% cells were GFAP positive astrocytes (figures 3(A), (B) and (E)). The number of neuronal differentiated cells kept on increasing during the following week: at the 14th day, the number of Tuj1positive cells reached 68.53 ± 4.28% (figures 3(C) and (F)). To the contrary, this ratio was only 2.95 ± 1.14% among untreated cells. Further analysis identified that without ES, hNSCs in the 3D culture would primarily differentiate into astrocytes (86.9 ± 2.95%, figures 3(D) and (G)). Z-stacks of hENTs are shown in additional files Movies S1-S3.
These results demonstrated that the ES initiated neuronal differentiation process from an early stage   of the differentiation. Additionally, the progress of neuronal differentiation and maturation would continue afterward, and the trend of neuronal differentiation seems not rely on the following consecutive ES. It indicated that 3 d of 150 mV mm −1 EF would be strong enough to affect neural fate determination among hNSC 3D cultures, therefore effectively inducing neuronal differentiation for their later development stages even when the ES was terminated.
As a result, in this study, the neurospheres in 3D Matrigel were exposed to 150 mV mm −1 EFs for 30 min per day for three consecutive days, then hENTs were transferred into fluid suspension culture on a shaker with a speed of 85 rpm until the end of the experiment ( figure 4(A)).

ES promotes neurites outgrowth and alignment in 3D hENT
Following the process described above, we induced hNSCs to differentiate into a neuronal tissue rich in mature neurons at day 21, the percentage of MAP2-positive neurons reached 72.48 ± 7.54%. MAP2-positive neurons extend long dendrites with branches, presenting mature neuronal feature. Most of MAP2-positive neurites were grow at an angle to the field lines, and some are even perpendicular to the field lines ( figure 4(B)). The outgrowth of neurites is critical for both neural development and regeneration which is related to nerve fiber projection, branching, synapse formation and neuron maturation (Zhu et al 2019).
In this study, we measured the lengths of neurites and the angles of EF-guided neurites relative to the EF axis in hENT. We first analyzed the distribution of EF axis in hENT. The hENT had the form of a disk (figure 4(C-ii)), it was small enough (with a ∼0.6 cm diameter) relative to the electrode (with a ∼1 cm diameter) and was located between the two electrodes (figure 4(C-i)). As a result, we considered the field lines inside the center of tissue (in the yellow frame) as a set of parallel lines (figure 4(C-iii)). Outside the frame, on the edge of the circular Matrigel droplet, the cells arranged along the circumference where they had a radial orientation (figure 4(C-iii)), which might be related to the edge effect of the circular droplet. Therefore, we mainly analyzed the neurite growth within the center of tissue.
After the ES was applied for 3 d, it was observed that in 3D hENT, the processes of positive cells were arranged at a certain angle with the EF line. But interestingly, after the ES was removed, not only did the cells continue to differentiate into neurons, but their neurites also remained somewhat perpendicular to the EF axis. For all tested over the different time points, ES induced neurite alignment at 40 • -120 • angle to the applied current (figures 4(E)-(H)). It also significantly increased mean neurite length compared to un-stimulated cells, which including Nestinpositive, Tuj1-positive, and MAP2-positive neurites ( figure 4(D)). Z-stack of hENTs is shown in additional file Movie S4.
These results indicated that by applying ES, we created a 3D hENT with high percentage of neurons and orderly growing neurites.

Inducing matured neuronal subtypes within the 3D hENT
Matured neuronal subtypes such as excitatory glutamatergic neurons and GABAergic interneurons were generated in this 3D neural tissue culture model. The Tbr1 (T-box brain transcription factor 1, which is a marker of layer VI cortical projection neurons) positive cortical neurons appeared within 24 d (figures 5(A) and (D)), then VGLUT1 positive excitatory glutamatergic neurons (figures 4(B) and (E)) and GABAergic interneurons (figures 5(C) and (F)) emerged in the culture at approximately 32-45 d.
By day 45, 17.69% of these cells were VGLUT1positive glutamatergic neurons. The expression of GABA, a neuronal marker of GABAergic interneurons, was also detected in hENT at day 45, and about 26.29% of cells in this culture were GABA-positive. The MAP2 + /VGLUT1 + and MAP2 + /GABA + double positive cells reached 38.02% and 57.48% respectively (figures 5(E) and (F)). These results indicated that the early neurons continued to be differentiated into diverse kinds of mature neuronal subtypes in 3D hENT, including excitatory glutamatergic neurons and GABAergic interneurons.

Formation of neural networks in hENT
Our previous work found that combining EF stimulations and the application of bFGF could create neurospheroid with long projections and potential neural networks (Meng et al 2020). We thus evaluated the neural network developed within the 3D hENT.
Here, we show that by day 28, MAP2-positive neurons were distributed over the whole hENT, and synaptic expression could be observed in the whole tissue (figures 6(A) and (B)). In order to present neuronal morphology in more details, we exhibited a single neuron in high magnification (figure 6(C)), it showed numerous dotted synapses distributed along MAP2positive neurons and their neurites.
The ultrastructure of synapse (figures 6(D) and (E)) and myelin sheaths (figure 6(F)) were observed under electron microscopy, indicating formation of complex neural networks in the 3D culture of hENT (figure 6(G)). By day 45, MAP2-positive neurons displayed mature morphologies with complex dendrites and dendritic spines. Neuronal morphology displayed highly complex dendritic architectures (figure 6(H)). Movie S5 illustrated instances in which neurons developed a 3D neuronal network through their interconnected neurites across the Matrigel.
We also quantified the immunofluorescent intensity of synapsin1 in hENT, and the results indicated that the expression of synapsin1 were significantly enhanced among these cells compared to unstimulated hENTs (figures 6(J)-(L)).

Detection of neural integration between hENT and EGFP-mENT
To test whether the constructed hENT was capable to form neural connections with surrounding tissues, a co-culture system with both hENT and EGFP-mENT was set up ( figure 7(A)).
After the ES, hENT and EGFP-mENT were cocultured for either 14 d or 21 d, respectively. At the day 14, EGFP + /Tuj1 + neuron was not observed in the hNET, although a few of EGFP − /Tuj1 + neurons were found in the EGFP-mENT ( figure 7(B)).
At the 21 day of co-culture, we detected extensive cell migration between hENT and mENT (figure 7(C)-(i)). MAP2 immunostaining results demonstrated large amount of EGFP + /MAP2 + mouse neurons in hENT. Additionally, EGFP − /MAP2 + human neurons were also detected in EGFP-mENT as well (figure 7(Cii)), indicating robust neural migration across both of co-cultured neural tissues by fusing hENTs and EGFP-mENTs.
These results suggested that cells derived from both hENT and EGFP-mENT were able to form synaptic connections as well as axonal projections between each other, indicating a great potentiation to develop active neural networks.

Detection of neural integration between hENT and cerebral slices
Organotypic slice is the bridge between primary cell culture and living animal. The advantage of an organotypic slice is that it preserves the 3D structure of the tissue and includes all the cell types of central nervous system. (Ucar et al 2021). We thus assessed whether neurons in hENT could integrate into the mouse organotypic slice in our 3D co-culture and fusion experiment ( figure 8(A)).
HENTs were co-cultured with cerebral slices from embryonic EGFP-mice (EGFP-mCerebral slices) in which the architecture of cortical tissues was still  (ii) Showed that EGFP + /Synapsin1 + expression was detected in hENT. (iii) Showed a high magnification of (ii), red arrow indicated an EGFP + /Synapsin1 + cell and yellow arrow indicated an EGFP − /Synapsin1 + cell. Cell nuclei were counterstained with Hoechst33342 (blue). The data was collected from three independent experiments. intact ( figure 8(B)). At the day 7 post fusion, some of mouse original EGFP + cells started to egress, while extensive cell-migration continued during the following 2 weeks (figure 8(C)), which were different from the co-culture result between hENT and EGFP-mENT ( figure 7(B)). Immunostaining results demonstrated that EGFP + neurons originally from the mouse slice migrated into the hENT (figure 8(C)), while synaptic connections formed between human projecting axons and EGFP + neurons from the EGFP-mCerebral slice ( figure 8(D)). These findings provided evidence for the occurrence of synaptic connections between human neurons in hENT and mouse neurons in the organotypic slice cultures. The time procedure of hENT co-culture with EGFP-mCerebral slices. (B) hENT co-cultured with EGFP-mCerebral slice. At day 14, EGFP + cells migrated from mCerebral slice into hENT: the left side of the red dotted line was EGFP-mCerebral slice; the right side of the yellow dotted line was hENT. B-iii: A higher magnification image presented EGFP + positive cells within hENT. (C) By day 21 of co-culture, both hENT and EGFP-mCerebral slice healthily survived and projected outgrowth toward each other. EGFP-mCerebral slice on the left side and hENT on the right. C-ii: the yellow dashed box showed that the neurites growing from the EGFP-mCerebral slice have extended and entered the hENT. C-iii is a high magnification image: yellow arrow indicated that synapsin1 + /EGFP − human cells migrated into EGFP-mCerebral slice; the red arrow pointed out that synapsin1 + /EGFP + mouse cell also migrated into the growth area of hENT. (D) Immunofluorescence staining showing the expression of synapsin1 + and EGFP + neurons in the central location of hENT, the red arrow pointed out the synapsin1 + /EGFP + mouse cell, yellow arrow indicated the synapsin1 + /EGFP − human cell. Cell nuclei were counterstained with Hoechst33342 (blue). The data was collected from three independent experiments. Human cells had the capacity to enter the organotypic slice tissue, exemplified by a number of synapsin1 + /EGFP − cells found in the outer region of the slice (figure 8(C-iii)). Notably, in some specimens at the border between the hENT and slice, numbers of synapsin1 + /EGFP + cells were detected (figure 8(C-iii)). Additionally, these mouse synapsin1 + /EGFP + cells were detected in the center of hENTs, indicating that cells in brain slices migrated over long distances and formed synapses with neurons in the center of hENTs ( figure 8(D)). This indicated that neurons in hENT and EGFP-mCerebral slice migrated to each other, and synaptic connections were formed between neurons in different tissues.
Generally, by using the combination of hENT and coronal brain slices, we studied not only the survival of neurons but also the nerve fiber innervation towards its natural target. Co-cultured hENT and EGFP-mCerebral slices showed excellent survival and extensive projecting toward each other to form synaptic connections, suggesting that cells in hENT could respond to promoting cues exerted from the mouse tissue.

Calcium imaging detects activities among neurons in hENT culture
The Ca 2+ inward flow is an important sign for neural activities: the increased intracellular Ca 2+ level would trigger synaptic vesicle exocytosis and the release of neurotransmitters. The calcium imaging uses Ca 2+sensitive dyes to represent the Ca 2+ concentration in neurons for detecting neuronal activities. It tracks neuronal action potentials and visualizing activities of neuronal populations. In our current setup, we choose Fluo-3AM (Rao andSikdar 2004, Funke et al 2007) as the Ca 2+ indicator to represent neuronal action potentials.
After KCl stimulation, activities of six neurons (figure 9(A), arrowheads identified) were recorded and selected for ROI analysis (figure 9(B)) to identify the rise of their intracellular Ca 2+ concentration. In (figure 9(C)), curves demonstrated peaks of the mean fluorescence intensity for each of them. They represented that the maximum intracellular Ca 2+ concentration occurs soon after these cells being stimulated by the high extracellular K + concentration. Interestingly, we also observed diverse types of calcium events as calcium waves, spikes or a combination of both calcium waves and spikes (figures 9(C) and (D)) from recordings among these cells, determining that neurons in hENT had calcium oscillatory behaviors including fast-rising calcium spikes in neurons. Therefore, raising extracellular K + concentration resulted the increasing neuronal excitability in the hENT, indicating electrical activities in neuronal cells.
Together, these results demonstrated that functional neurons were efficiently established in hENT. Movie S6 showed changes of the peak of calcium ion concentration under confocal microscopy.

Discussion
Mimicking key features of early neural development could be a productive strategy for constructing hENTs in vitro. Pivotal events in reproducing neural tissue may include: (1) to direct neural differentiation in hENTs; (2) stimulating neurons in the hENTs to form branched neurites with highly ordered projection/reception fields; (3) the formation of complete myelin sheath; (4) to form a well-developed neuronal network (Yu et al 2022).
In the present study, we developed a customizable and reproducible hENT, which possessed certain tissue structures with multiple kinds of mature neuron subtypes, complex synaptic connections among cells and neuronal networks with physiological activations. This biofabrication could survive in vitro for more than 45 d.
In our previous studies, we have optimized ENT construction parameters which including EF strength, synergistic factors and the procedure (Meng et al 2020). These results indicated that EFs had following functions: (1) inducing neuronal differentiation of NSCs (Dong et al 2017); (2) promoting the parallel and ordered growth of neural processes in 3D cultured neural tissue (Meng et al 2020) (3) stimulating the formation of neurite branches; (4) triggering the formation of synaptic connections and myelin sheaths in the 3D culture environment, enhancing the formation and maturation of neural networks (Yu et al 2022).
By seeding hNSCs formed neurospheres evenly in grftMG, we customized the size and shape of hENT. By using the optimized 3 d of ES, our protocol sufficiently provided guidance of neuronal differentiation during early development and assisted the cell fate determination. After being stimulated by ES, hENTs were transferred to a shaker for extension culture and only medium changes are required throughout the rest of the culture period. This protocol was simple and efficient to construct the hENT with a high percentage of neuronal cells, a short developing time while minimizing interference other than artificial modification and ESs. By analyzing the cellular composition in hENT at different time points, we observed concomitant decrease of the expression of early neurogenic marker (Nestin) as well as an increase of the immature neuron marker (Tuj1). These data indicated a preference towards the neuronal lineage, and this was exactly what we expected. Because compared to the glia, the regenerative capacity of neurons is extremely limited after central nervous system injury.
However, the initial stage of neuronal differentiation was further prolonged in the construction of hENT compared to that of mENT (5-7 d in mouse). About 57.34% of the cells in hENT were Nestin + NSCs at day 7 of construction, and only 18.67% of the cells were Tuj1 + neurons. The Tuj1 + neurons reached to 68.53% in the 14 d of construction, while it reached a similar proportion at 7 d of mENT construction. Different from our results, Amanda et al examined the growth and differentiation of different hNSC cell lines in the 3D culture system they constructed, the maximum differentiation ratio of neurons was only 33% (Marchini et al 2020).
Further characterization of hENTs by immunofluorescence staining indicated that neurons were matured during a longer incubation time. About 100% of hENTs expressed the pan-neuronal marker, MAP2, at day 21 and expressed the early cortical marker, Tbr1, at day 28, demonstrating that there were precursors of cortical excitatory neuron presenting in the hENT. By 45 d of construction, a number of neurons in the hENTs were excitatory glutamatergic pyramidal neurons (17.69 %) and inhibitory GABA producing interneurons (26.29 %), which are two main populations of neurons in cerebral cortex (Harris and Shepherd 2015). This is similar to the culture system of Amanda M et al constructed culture system, where they also obtained more GABAergic neurons at the 6th week of construction (Marchini et al 2020).
Furthermore, both Tuj1-positive and MAP2positive neurons were distributed rather uniformly in the hENT, while young neurons had a more central localization in other tissue engineered neural constructs (Baklaushev et al 2019). We attributed this result to ES promoting more extensive migration of hNSCs and differentiated neurons in the 3D scaffold.
More importantly, the synaptic marker Synapsin1 and MAP2 double positive neurons presenting in hENTs suggested neuronal maturation and the presence of synapses at this time point. The outgrowth of neurites is the fundamental procedure during neural development and regeneration, which is related to neuron projections, neurites branching, synaptic formation and neural network maturation (Zhu et al 2019, Yu et al 2022. Highly branched neurites were also observed in hENTs indicating a well-developed framework of the neural tissue in this culture condition. In 3D Matrigel matrix, 150 mV mm −1 EFs stimulation not only induced the directed neuronal differentiation, but also regulated the orderly growth of neuronal processes perpendicular to the EF-axis. The guiding effect of ES on neurite growth and arrangement has been well demonstrated (Hinkle et al 1981, Patel and Poo 1982, Rajnicek et al 1992, Pan and Borgens 2012. In this study, the highly branched neurites of mature neurons were well developed and connected to each other into a network. In our viewpoint, the formation of a mature neural network is mainly based on the orderly neural maps, cell arrangement and well-organized axonal trajectories together with well-developed dendrites to form numerous synapses (Yu et al 2022). This process involves two events: dendritic branching and synaptic formation. Dendritic branching and synaptogenesis occur simultaneously, forming numerous functional synapses and helping to develop morphological recombination of neuronal network (Baldwin and Eroglu 2017, Ohtaka-Maruyama et al 2018, Lenk et al 2019. In this study, neurons were distributed over the whole hENT culture and formed a large number of synapses. This corroborates the view that ES has the potential to enhance 3D tissue structures toward functionally mature neural networks and more consistent structures (Warren et al 2021).
The platform of organotypic slices culture have been well established in our laboratory, including brain slices and spinal cord slices made with a tissue chopper (Meng et al 2011b). Slice model has the anatomical integrity and structures. Here, neurons survived well in brain slices of 300-400 µm thickness after 2-3 weeks in culture, which is in agreement with other studies (Ucar et al 2021). In this study, by using hENT and coronal cerebral slices co-culture, we tended to develop the in vitro 3D culture platform. It was not only a representing method to study human neuronal cell migration, but also an assay to study nerve fiber innervation/integration. Labeling cells of cerebral slices with a fluorescent reporter allowed the visualization of long-distance neuronal migration. Neurons derived from different tissues/slices migrated to the opposite side of each other, and then established synaptic connections. Axons projected toward each other and generated neural network activities, suggesting that the cells in hENT could respond to survival and migration promoting cues exerted from the mouse tissue. Especially, mCerebral slices might have some unique chemokines which could promote cell migration toward cocultured tissue while mutually stimulating neural projections.
However, this study has several limitations. The results of the present study only provide the presence of the synaptic connections formed by ENT and tissue slices, and lacks quantitative research on the synaptic function, such as acquiring and interpreting measures for synaptic strength and function, exact determination of the presynaptic origins of the recorded synaptic activity, and quantitative analysis of the function of neural circuits. Therefore, the objective, operational and generalized conclusions on ENT function are need to be further identified. In the future, we plan to use optogenetics and patch-clamp electrophysiology techniques to measure synaptic connectivity and strength.
We also encountered a common problem in fabrication researchers: limited supplements of oxygen and nutrients lead to a necrotic core in the 3D tissue center accompanied by tissue maturation during long-term culture (Bagley et al 2017, Xiang et al 2017, Rothenbücher et al 2021. Therefore, in the process of continuous practice, the construction of ENT needs further research to explore the intensity, time point and operation process of EF in 3D culture, and then to the direction of organoid culture systems that are closer to the real situation in vivo. The source and acquisition of NSCs is another important consideration. Several issues need to be addressed, such as ethical implications of the origin of human embryos, optimal cell source, the development of strictly defined culture procedures, strict quality control, long-term safety of stem cell therapy, and immune rejection. It is worth noting that NSCs can be obtained from human induced pluripotent stem cells (iPSCs). iPSCs are a class of dedifferentiated cells derived from human adult cells that have been induced to express specific genes, thus avoiding the controversial use of embryos. Since iPSCs can be obtained from the patient's own cells, it is believed that iPSC therapy can completely evade the immune response, and the establishment of iPSC biobanks would provide large-scale use in experimental and/or clinical studies. Therefore, considering the prospect of further clinical application, the next step will be applying ES to induce the neural differentiation of iPSCs to construct 3D hENT in vitro.
Regardless of these limitations, this study shows an alternative and feasible transplant substitute which should be incorporated into more animal experiments and clinical practice.5.

Conclusion
We concluded that the ES could be used to generate organized 3D human neural tissue with high percentage mature neurons and well-developed neuronal network. Moreover, we investigated and identified EF-induced hENT in both structural and functional manner. The key points for this method to succeed includes: (1) hNSC neurospheres could quickly fuse to form a3D tissue; (2) by using our protocol, hNSCs in the 3D tissue could be differentiated into the desired mature neuronal cells; (3) ES promoted the parallel growth of neural processes in hENT; (4) in a 3D-culture environment, ES promoted the formation of synapses and myelin sheaths, thus forming a mature 3D neural network. Overall, hENT could be a new model for neural development research and therapeutic drug screening, as well as a potential neural tissue substitute for the treatment of central nervous system (CNS) injury.

Data availability statement
All data that support the findings of this study are included within the article (and any supplementary files).