A cocktail hydrogel promoting the functional interneurons regeneration of human neural progenitor cells for brain injury therapy

Graphical abstract

Highlights

• •A cocktail-loaded (BGA@GelMA) hydrogel are fabricated and synergistically improve the efficacy of hNPCs-based therapy for TBI.
• •The BGA@GelMA hydrogels mimic brain ECM in micro- and nano-architectures and reconstruct NVU by recruiting host endothelial cells.
• •The BGA@GelMA hydrogels enhance the secretion of glutamate and the activation of the MAPK signaling pathway, thereby promoting the maturation of interneurons and synaptic communication.

Abstract

Introduction

Traumatic Brain Injury (TBI) usually leads to substantial mortality and disability among adult populations. Neural progenitor cells (NPCs) transplantation exhibits great potential in TBI treatment. However, the differentiation rate of interneurons is relatively low, largely impeding the therapeutic effects of brain tissue repair.

Objectives

A cocktail hydrogel (BGA@GelMA) was developed to provide a sustained release of neurotrophic factors and neural signaling molecules, thereby promoting the maturation of cortical interneurons, which can be utilized to improve the therapeutic outcomes of hNPCs transplantation for cerebral injury treatment.

Methods

The cocktail strategy (BGA@GelMA) was developed by integrating a hydrogel with neural inducers, and was characterized using cryogenic scanning electron microscopy (cryo-SEM), rheological tests and release kinetics. The potential of the BGA@GelMA hydrogel to differentiate hNPCs into cortical interneurons and to facilitate neural networks formation was analyzed with transcriptomic analysis, transsynaptic rabies virus tracing, flow cytometric analysis, immunofluorescence staining, and quantitative reverse transcription polymerase chain reaction (RT-qPCR). To enhance the efficiency of cell-based therapies, human pluripotent stem cells (hPSCs)-derived hNPCs encapsulated within the BGA@GelMA hydrogel were transplanted into a rat TBI model. The brain tissue repair was assessed using hematoxylin and eosin (H&E) staining, immunofluorescence staining, magnetic resonance imaging (MRI), motor evoked potential (MEP) recordings, and behavior tests.

Results

The synergistic role of micro- and nano-characterization, mechanical properties and multiple inducers-rich environment facilitated the maturation of cortical interneurons and the formation of synapses. Furthermore, after transplantation into the motor cortex of a rat TBI model, hNPCs embedded within BGA@GelMA hydrogel successfully enhanced brain recovery by differentiating into cortical interneurons, reducing the inflammatory response, and regenerating Neurovascular-like Unit (NVU).

Conclusion

This cocktail hydrogel provided a novel strategy for improving the effect of hNPCs transplantation for cerebral injury treatment.

Introduction

Traumatic Brain Injury (TBI) leads to the irreversible loss of mature neurons and poses a significant challenge for treatment [[1], [2], [3]]. Currently, there is no effective clinical method to recover this loss. The suboptimal recovery of brain function following TBI is attributed to the limited regenerative capacity of endogenous neural progenitor cells (NPCs) within the central nervous system (CNS). Cell replacement therapies could provide an approach for injuries of the motor cortex, as cortical interneurons are crucial for the functionality of neural circuits [[4], [5], [6], [7], [8], [9]]. Previous studies demonstrated the migration of grafted NPCs through blood–brain barrier (BBB) to the peri-injured area, and the differentiation of NPCs directly increased the neuronal numbers [[10], [11], [12]]. In our previous study, implantation of human NPCs (hNPCs) derived from human pluripotent stem cells (hPSCs) resulted in the regeneration of multiple cortical neurons in vivo [13,14]. Thus, the transplanted hNPCs possess the ability to differentiate into functional neurons in situ, with the potential to guide the reconstruction of the injured area.

The survival of hNPCs after transplantation is recognized as a critical factor for behavioral recovery. However, the ischemic and inflammatory microenvironment at the injury site significantly compromises the survival of transplanted cells in the ischemic core. [15,16]. Moreover, the differentiation of transplanted cells into functional neurons, which is essential for effective integration with the host neural circuitry, was impeded by the injury-induced microenvironment within the ischemic core, leading to decreased differentiation efficiency and ultimately limiting therapeutic efficacy [[17], [18], [19], [20], [21]]. Firstly, the shear forces during injection without a delivery vehicle tend to cause mechanical damage to hNPCs [22]. Secondly, Extracellular Matrix (ECM) is important to provide physical support for graft retention and tissue restoration, but TBI causes the formation of irregularly shaped cavities devoid of ECM [23]. Finally, an ischemic environment results in severe deficiency of neurotrophic factors, dysfunction of Neurovascular-like Units (NVUs) and accumulation of inflammatory cytokines. These physiological events hinder the survival, differentiation and neuronal connectivity of transplanted cells [24]. Thus, the utilization of bioactive scaffolds that mimic the mechanical properties, micro/nanostructure and ECM of the brain tissue is expected to produce better therapeutic outcomes for TBI.

Bioactive hydrogels have been shown to serve as effective delivery vehicles for in situ cell-based transplantation [[24], [25], [26], [27], [28]]. For instance, methacrylated gelatin (GelMA) hydrogels retain the native properties of ECM, providing a conducive microenvironment for cellular growth and differentiation [29]. The arginine-glycine-aspartic acid (RGD) sequences of GelMA hydrogel promote integrin-mediated cell adhesion, synergistically enhancing the migratory behavior and interaction of the graft-host [11]. In addition, the matrix metalloproteinase (MMP)-responsive degradation property in GelMA hydrogel endows it with controlled biodegradability, accommodating the dynamics of the cytoskeleton and the morphogenic events pivotal to neural differentiation [30]. Furthermore, the mechanical properties and micro/nanoarchitectures can be tuned over the range of stiffness, making GelMA applicable in the elastic modulus of brain tissues [31]. However, pure GelMA hydrogels have limited potency of neuroregeneration [32]. Therefore, it is necessary to combine GelMA hydrogels with bioactive factors to improve neural maturation and synaptogenesis for brain repair [[33], [34], [35]].

Neural inducers, including the neurotrophic factors and neural signaling molecules, can differentiate NPCs into specialized neural subtypes by inductive signals pattern [36]. Recent studies demonstrated that neurotrophic factor-loaded hydrogels can promote the differentiation of the transplanted NPCs [37]. For example, hydrogels delivering brain-derived neurotrophic factors (BDNF) enhanced neurogenesis in peri-infarct area, improving motor function recovery of mice and monkey models [38,39]. In addition, hydrogel-laden glial cell line-derived neurotrophic factor (GDNF) promoted the differentiation toward neuron-like cells [40]. Nevertheless, relying on single neural inducers has limited efficacy for regeneration of complex neural networks, thereby the exploration of hydrogels assembled by multiple neural inducers is a logical step [14,41]. Importantly, our pre-study has shown that both the BDNF and GDNF play crucial roles in promoting the repair of the ischemic brain following the transplantation of hNPCs [14]. Moreover, recent studies have shown that the combination of cyclic adenosine monophosphate (cAMP) with BDNF facilitated the neuronal maturation in a two-dimensional culture system [36,42]. The combination of cAMP with GDNF enhanced the specification of neuronal subtypes by pomoting angiogenesis [10,43,44]. Consequently, the coordinated application of a cocktail comprising BDNF, GDNF, and cAMP is expected to enhance the maturation of interneurons, potentially improving the therapeutic efficacy of cellular replacement for cortical regeneration.

Accordingly, it is hypothesized that a synthetic neural inducers hydrogel, which is composed of BDNF, GDNF, and cAMP, could (i) provide sustained delivery of neurotrophic and neural signaling factors, (ii) enhance the viability and lineage-specific differentiation of hNPCs, and (iii) synergistically support the maturation of cortical interneurons through a brain-mimicking mechanical environment and multi-inducer cues. Such a strategy may further recapitulate aspects of cortical structure and function through host–graft interactions.

To validate this hypothesis, a biomimetic hydrogel delivery system, termed as BGA@GelMA, was prepared to facilitate the differentiation and maturation, as well as to promote the establishment of functional synaptic networks in vitro. Additionally, the hydrogel system promoted the regeneration of cortical interneurons and the formation of neurovascular-like structures through host-graft interactions in vivo after transplantation. These findings hold great promise for cell-based treatment of cerebral traumatic injury.

Material and methods

Preparation of BGA@GelMA hydrogel

GelMA was synthesized according to the previous description [45]. Briefly, 100 μL methacrylic anhydride was added dropwise to a 10 % gelatin solution (Sigma Aldrich, USA) at 60 °C for 3 h. The resulting 10 % GelMA solution was then dialyzed against Milli-Q water using 12–14 kDa cut-off dialysis membrane. The Milli-Q water was changed twice daily for 7 days at 60 °C to remove unreacted reagent. GelMA solution was freeze-dried for 2 days at −40 °C, then stored at −20 °C for later use. To prepare the BGA@GelMA hydrogel, lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP) solution was prepared with a concentration of 0.25 % (w/v). Then, the freeze-dried powder of GelMA was added to the LAP solution to achieve the concentration of 2.5 %, 5 %, and 10 % (w/v). Finally, cAMP at 1 μM, BDNF at 20 ng mL⁻¹, and GDNF at 20 ng mL⁻¹ were mixed together to ensure uniform distribution of the compounds.

The prepared solution was then transferred into an 8 Chambered Cover glass System (Cellvis) and exposed to 405 nm UV light (25 mW cm⁻2, Engineering For Life) for 20 s to form the BGA@GelMA hydrogel.

In vitro release analysis of neural inducers in BGA@GelMA hydrogel

The release kinetics of neural inducers in BGA@GelMA hydrogel were assessed using the human BDNF valukine enzyme-linked immunosorbent assay (ELISA) kit (VAL136, R&D Systems), GDNF ELISA kit (CSB-E04565h, Cusabio) and cAMP ELISA kit (CSB-E04488h, Cusabio), respectively. After solidifying the hydrogel solution at 37 °C, 5 mL of phosphate-buffered saline (PBS) was added to it for incubation at specified intervals (0 h, 6 h, 12 h, 48 h, 3 days, 7 days, and 14 days). Then 100 μL aliquots of the PBS were collected to quantify the released neural inducers. The concentration of the released neural inducers was quantified with available ELISA kit [25].

Morphological characterization of BGA@GelMA hydrogel

To evaluate the morphological characterization of the BGA@GelMA hydrogels assembly, cryogenic scanning electron microscopy (cryo-SEM) was used. The hydrogel samples were affixed to a holder and were rapidly frozen using liquid nitrogen. Subsequently, the suspensions were positioned into stub holes and plunge-frozen in liquid nitrogen. Then, the samples were introduced into a chamber etched by sublimation for 15 min at −90 °C followed by platinum sputtering (PP3010T, Quorum). Ultimately, the samples were examined using a scanning electron microscopy (SEM) to obtain the images (S-4800, Hitachi).

Mechanical properties of BGA@GelMA hydrogel

Rheological properties of the hydrogels were assessed utilizing a compact rheometer (MCR302, Anton Paar). The hydrogel samples were molded into cylindrical form with a diameter of 20 mm and a height of 2 mm and stabilized to ensure complete gelation at 25 °C. Amplitude Sweep (AS) experiments were conducted at a steady frequency of 1 Hz, with the strain varied from 0.01 % to 100 %. The storage modulus (G') and the loss modulus (G'') of all samples were recorded.

To evaluate the mechanical properties of the hydrogels, compression testing was performed. Gel discs (10 mm diameter, 2 mm thickness) were compressed to 100 % strain at a deformation rate of 10 mm/min (n = 3). The compressive stress–strain data were recorded, and the apparent compressive modulus was calculated as the slope of the stress–strain curve within the 60–80 % strain range, where the linear response was most prominently observed.

Directed differentiation of hNPCs derived from hPSCs

All cultures were maintained in a humidified incubator containing 5 % CO₂ at a temperature of 37 °C. hPSCs (H1 line) were cultured in mTeSR1 medium (85850, STEMCELL Technologies), and differentiation into hNPCs was performed following previously established protocols [69,70]. Briefly, after treatment with 0.5 mM ethylenediaminetetraacetic acid (EDTA) to dissociate the cells, hPSCs were passaged at a ratio of 2:1 onto plates (CCP01006-B, Vazyme) pre-coated with Matrigel. On the subsequent day, the culture medium was replaced with a 1:1 mixture of neurobasal medium (21103–049, Gibco) and Dulbecco’s modified Eagle’s medium/F12 (SH30023-018, Hyclone), enriched with 0.5 × N2 supplement (17502048, Gibco), 0.5 × B27 supplement (17504044, Gibco), 2 μM SB431542, and 2 μM dorsomorphin. After 8 days, the cells were passaged at a ratio of 1:2 onto plates pre-coated with Matrigel and further cultured in N2B27 medium for 4 days. To induce differentiation into hNPCs, 20 ng mL⁻¹ basic fibroblast growth factors (bFGFs) (10014-HNAE-1, Sino Biological) were added for an additional 4 days.

For the identification of hPSCs, protein markers including octamer-binding transcription factor 4 (OCT4) and stage-specific embryonic antigen 4 (SSEA4) were utilized [46]. For the characterization of hNPCs, Paired Box 6 (PAX6), Forkhead Box G1 (FOXG1), and Sex Determining Region Y-Box 2 (SOX2) were employed. The analysis of marker expression was conducted through flow cytometry and immunofluorescence (IF) staining. A detailed list of the antibodies in vitro used can be found in Tables S1, S2, Supporting Information.

In vitro viability studies of BGA@GelMA hydrogel

For three-dimensional (3D) cell culture, dissociated hNPCs were mixed with BGA@GelMA hydrogel and subsequently photopolymerized using a 405 nm UV light. At days 0 and 14, the apoptotic cells were identified with an Annexin V-APC kit (KeyGEN Biotech) and the samples were analyzed using a flow cytometer (Beckman CytoFLEX S). The viability of cells encapsulated within the hydrogel was determined through a live/dead staining kit (KeyGEN Biotech) with calcein acetoxymethyl ester (Calcein AM) for live cells (green) and propidium iodide (PI) for dead cells (red).

hNPCs differentiation in BGA@GelMA hydrogel

For the differentiation of hNPCs within the 3D microenvironment of BGA@GelMA hydrogel, the hNPCs were first dissociated into single cells using Accutase (07922, Stemcell), and encapsulated into BGA@GelMA hydrogels at a concentration of 6 × 10⁶ cells mL⁻¹. After photo-crosslinking, the hydrogel was incubated in N2B27 medium.

For the 3D differentiation within Matrigel, hNPCs were dissociated into single cell and resuspended in Matrigel solution. The Matrigel solution was subsequently gelated at 37 °C, and cultured using the same medium as for the BGA@GelMA hydrogel. Reverse transcription quantitative polymerase chain reaction (RT-qPCR) was employed to quantify the expression levels of genes associated with neural differentiation. A detailed list of primers is provided in Table S3, Supporting Information. In addition, IF staining was employed to evaluate the cells for the markers of neurons, astrocytes, oligodendrocytes, and synapses.

Neural network formation in BGA@GelMA hydrogel

We employed a genetically edited hPSCs cell line (H1-CAG-GTRqp) for retrograde monosynaptic tracing. This cell line constitutively expresses the avian TVA receptor, the rabies virus glycoprotein, and GFP all regulated by the human CAG promoter [47]. The rabies virus of envelope protein A(EnvA)-pseudotyped glycoprotein-deleted (GRV) was added to the medium to assess retrograde monosynaptic tracing.

In vivo transplantation of hNPCs with BGA@GelMA hydrogel

All procedures related to housing, breeding, and conducting experiments on animals were performed in compliance with the ethical standards set by the Guangzhou Institutes of Biomedicine and Health (GIBH) Ethical Committee (IACUC, 2022124) and followed the guidelines of the GIBH Institutional Animal Care and Use Committee. Adult male Wistar rats, aged 8 weeks and sourced from Beijing Vital River Laboratory Animal Technology Co., Ltd, were maintained under the laboratory environment conditions and were allowed to acclimate to the settings for at least 1 week [48].

To visualize the implanted cells directly, hNPCs derived from the reporter human cell lines (H1-CAG-GFP) were used at a concentration of 6 × 10⁶ cells mL⁻¹. Rats were anaesthetized with 3 % isoflurane and secured to a brain stereotaxic device at 37 °C. Target areas were shaved, and skins were cut. A 2-mm skull defect was created using a drill within the right motor cortex, positioned at 1 mm to 3.0 mm anterior-posterior and 1.5 mm to 3.5 mm lateral to the bregma. Then the rats received punch treatment under 2 mm height for the TBI model. For the hNPCs + BGA@GelMA group, the hNPCs-loaded hydrogels were implanted into the lesion site. After that, the incision was closed using a 19 mm polyglycolic acid (PGA) suture, and the rats were accommodated in clean and temperature-regulated cages at 37 °C, with the monitor sustained until the subjects exhibited the capability for locomotion. To prevent potential infection, the penicillin G was administered to the rats at a dosage of 115 mU kg⁻¹ for 3 days. In addition, the immunosuppressor cyclosporine A (CsA) was treated subcutaneously at a daily dosage of 10 mg kg⁻¹ from 2 days before transplantation until the rats were sacrificed [49].

Staining of tissue slice

A 4 % paraformaldehyde (PFA) solution was applied to brain samples for 48 h for fixation. Subsequently, brain tissues were dehydrated, embedded in optimal cutting temperature compound (OCT), and then sectioned into 25 µm-thick slices for further analysis.

Hematoxylin and eosin (H&E) Staining Kit (G1120, Solarbio) was used to examine the micro/nanostructures of brain tissue after initial washing with PBS. The tissue sections were stained with 200 μL of hematoxylin solution for 10 min. The sections were then exposed to 1 % hydrochloric acid in ethanol for 10 s, followed by a 5-minute rinse under running water, which resulted in blue nuclei. The sections were rinsed with distilled water, stained with eosin for 1 min, dehydrated using a graded ethanol series, and cleared with xylene I and II. After that, the neutral resin was used to mount the sections.

IF staining was performed by fixing tissue sections for 30 min in 4 % PFA, washing the sections in PBS to remove non-specific binding after primary antibody incubation. The sections were incubated with secondary antibodies for 1 h and counterstained with 4′,6-diamidino-2-phenylindole (DAPI) to highlight nuclear morphology. A Zeiss LSM 800 microscope was used to collect fluorescent images. A detailed list of the antibodies in vivo used also can be found in Tables S1, S2, Supporting Information.

Electrophysiological assessment by MEP recording

To evaluate functional recovery of the corticospinal pathway after transplantation, motor evoked potentials (MEPs) were recorded using transcranial electrical stimulation [50]. Under anesthesia, rats were placed in a stereotaxic frame, and a single-pulse electrical stimulus (5 V, 0.1 ms duration) was applied through subdermal needle electrodes positioned over the motor cortex. Electromyographic (EMG) responses were recorded from the contralateral gastrocnemius muscle using bipolar electrodes.

The amplitude of the MEP signals was used as the primary index to reflect motor pathway integrity. Signals were acquired and processed using the BL-420N biological signal acquisition system (Chengdu Taimeng Technology Co., Ltd., China), and analyzed with the associated data processing software.

Neurovascular functional restoration assessed by MRI

To evaluate the structural and functional regeneration of neurovascular units (NVUs) at the lesion site post-transplantation, multimodal magnetic resonance imaging (MRI) was performed using a 9.4T Bruker BioSpec small animal scanner, incorporating arterial spin labeling (ASL), T2-weighted bBOLD imaging, and proton magnetic resonance spectroscopy (^1H-MRS). Specifically, cerebral blood flow (CBF) maps derived from ASL sequences were used to assess microvascular perfusion. T2 relaxation times obtained from blood oxygenation level-dependent (BOLD) imaging served as indicators of tissue oxygenation. The 3.23 ppm/creatine (Cr) ratio from ^1H-MRS spectra was employed to evaluate the relative glucose metabolism in the graft site.

Statistical analysis

The data was statistically processed using GraphPad Prism software and presented as the mean ± standard deviation. One-way analysis of variance (ANOVA) was used to evaluate in vitro cell viability and neurite extension, followed by Dunnett’s post hoc test for multiple comparisons. A two-tailed, unpaired t-test with Welch's correction was used to analyze in vivo graft survival. Two-way ANOVA was used to assess functional behavior in vivo, along with Tukey's post hoc test for multiple comparisons. Exact p-values are reported for all statistical comparisons in the Results section and figure legends.

Results

Preparation and characterization of the BGA@GelMA hydrogel

Modulus matching brain tissue is considered as an essential requirement for applications of hydrogels [51]. Rheological tests were conducted to determine the viscoelastic properties of three concentrations of BGA@GelMA hydrogels at 2.5%, 5%, and 10% (w/v). The AS test showed that the G' value of the 5 % group and the 10 % group remained almost exactly at the initial 'gel' state with the increase in amplitude, indicating that both the 5 % and 10 % hydrogels showed adequate stability with amplitude elevation (Fig. 1A).

Fig. 1.

Preparation and characterization of the BGA@GelMA hydrogel. (A) The storage modulus (G') and the viscous modulus (G") of the 2.5 %, 5 % and 10 % (w/v) BGA@GelMA hydrogels by AS test. (B) Quantification of the compressive modulus of the samples, n = 3, *p = 0.0472 (5 % vs 2.5 %). **p = 0.0013 (5 % vs 10 %). (C) Representative stress strain curves of three hydrogel samples. (D) Cryo-SEM images of the 2.5 %, 5 % and 10 % BGA@GelMA hydrogel. (E and F) Quantification of the average pore sizes, and the ratio of the longitudinal dimension to the lateral dimension of the BGA@GelMA hydrogels with three different mechanical stiffnesses, n = 30, **p = 0.0056 (aspect ratio, 5 % vs 10 %) and **p = 0.0021 (pore size, 5 % vs 10 %), respectively. (G and H) The morphology of GFP labeled hNPCs under microscope after 3 days. Images below are the enlarged image of the area shown in the white box, scale bar = 50 μm. (I) Quantification of distance from cell body center of samples, n = 9, ****p = 0.00001 (5 % vs 2.5 %) and ****p = 0.00002 (5 % vs 10 %). (J) Quantification of hNPCs spheres under light microscope, n = 3, ***p = 0.0008 (5 % vs 2.5 %).

The hydrogel scaffolds with a low modulus similar to native CNS tissue can promote NPCs differentiation, and significantly enhance functional recovery [52]. To compare the mechanical performance of the hydrogels, the stress–strain curves obtained from compression testing were analyzed. The apparent compressive modulus was calculated as the slope of the stress–strain curve within the 60–80 % strain range, where the response was relatively linear and differences between groups became evident. Based on this analysis, both the 5 % and 10 % groups exhibited higher apparent compressive modulus compared to the 2.5 % group (Fig. 1B), indicating enhanced resistance to deformation. Additionally, the 5 % group maintained structural integrity and showed a modulus value (4.1 ± 0.1 kPa) close to native brain tissue stiffness (∼3.5 kPa), suggesting more favorable elasticity. Notably, the 10 % group showed signs of material failure at approximately 90 % strain (Fig. 1C).

The cryo-SEM was employed for examining the micro- and nano-architectures of three concentrations of BGA@GelMA hydrogels at 2.5 %, 5 %, and 10 % (w/v). With increasing GelMA concentration, the pore size in the hydrogels decreased, indicating a denser structure (Fig. 1D). The aspect ratio of the 5 % group is 1.1 ± 0.4, which is more uniform compared to the 2.5 % group (2.0 ± 0.4) (Fig. 1E). The pore sizes of the 2.5 % and 10 % groups were 78.5 μm ± 46.3 μm and 0.70 μm ± 6.6 μm, respectively. The pore size of the 5 % group was 35.0 μm ± 5.7 μm, which is suitable for hNPCs attachment (Fig. 1F) [53].

The length of neurofilaments extending from hNPCs serves as a morphological indicator to validate the efficiency of neural differentiation [36]. To study the effects of micro- and nano-structure on the cellular behavior of the encapsulated hNPCs, BGA@GelMA hydrogel of selected groups at 2.5 %, 5 %, and 10 % (w/v) was fabricated and cultured for 3 days. hNPCs in the 5 % group extended to filopodia-rich shape, while hNPCs in the 2.5 % and 10 % groups retained neurospheres and rounded morphology (Fig. 1G and H). The neurites emerging from the cell soma in 5 % hydrogel were 79.12 μm ± 35.59 μm, which was significantly longer than the 1.24 μm ± 0.78 μm in 2.5 % hydrogel and 6.77 ± 6.94 μm in the 10 % group (Fig. 1I). The spheres of hNPCs in the 2.5 % group were found to increase nearly 20-fold than the 5 % and 10 % groups, as the hydrogel degraded and led to insufficient support neurite extension (Fig. 1J). These results revealed that neurite elongation was superior in the 5 % group, suggesting that micro- and nano-structures of BGA@GelMA hydrogel have an important effect on neurite and outgrowth.

The in vitro release behavior of neural inducers and biocompatibility of BGA@GelMA hydrogel

The 5 % BGA@GelMA hydrogels had the most suitable physical properties and micro- and nano-structure, hence were subjected to further cytocompatibility testing. A double-antibody sandwich ELISA was employed for quantifying the release of BDNF, GDNF, and cAMP at specific time intervals (0 h, 6 h, 12 h, 48 h, 3 days, 7 days, and 14 days). The cumulative release data showed that 96.61 % of the BDNF was released within 72 h (Fig. 2A). Additionally, 96.47 % of the GDNF and 93.91 % of the cAMP were released within 96 h, respectively (Fig. 2B and C). The results revealed that BDNF, GDNF, and cAMP are capable of a sustained release within the BGA@GelMA hydrogel, thus allowing new axons formation within 3 days. Moreover, the release behavior of neural inducers within the BGA@GelMA hydrogel is also appropriate for the acute phase of brain injury (3 days) [54]^.

Fig. 2.

The in vitro release behavior of neural inducers and biocompatibility of BGA@GelMA hydrogel. (A to C) The relative release profile of BDNF, GDNF, and cAMP of the BGA@GelMA hydrogel measured with ELISA kit, n = 3. (D) Schematic representation of 3D culture system via the BGA@GelMA and the Matrigel hydrogel. (E) Flow cytometry was used to detect APC of cell apoptosis. (F) Statistical diagram of apoptotic cell distribution, n = 3, ***p = 0.0005. (G) Quantitative analysis of live cells and dead cells per field in the live/dead assay, n = 3. (H) Representative live/dead images of hNPCs co-cultured with the BGA@GelMA hydrogels and the Matrigel hydrogels after 14 days.

According to the flow cytometry and IF analysis, the hPSCs of the H1 cell line showed that over 98.5 % of cells expressed the pluripotent markers OCT4 and SSEA4. hNPCs derived from hPSCs exhibited the markers of neural progenitors with PAX6, FOXG1, and SOX2 (Fig. S1A to D, Supporting Information). Subsequently, hNPCs were embedded in 3D systems established by the BGA@GelMA hydrogels, with Matrigel hydrogels used for comparison (Fig. 2D). Annexin V-APC/PI staining showed that hNPCs in BGA@GelMA group exhibited almost no apoptosis cells (1.48 % ± 1.07 %) compared to Matrigel hydrogel (14.93 ± 0.67 %) within 14 days (Fig. 2E and F). Additionally, live/dead staining revealed that most of the hNPCs in the BGA@GelMA group remained alive (93.62 ± 1.83 %) compared to the Matrigel group (73.67 % ± 3.48 %) (Fig. 2G and H). These results demonstrated that BGA@GelMA hydrogels had excellent biocompatibility to promote the survival of hNPCs in vitro.

Effect and mechanism of BGA@ GelMA hydrogel on hNPCs differentiation

To determine the effects of BGA@GelMA hydrogels on neural differentiation of encapsulated hNPCs, neurons relevant markers were examined by RT-qPCR and IF staining. After 2 weeks, the gene expression of Tuj1 (β-tubulin III, a protein associated with early neural generation) in the BGA@GelMA group was 2.26-fold higher than those in the Matrigel group (Fig. 3A). The gene expression of MAP2 (microtubule-associated protein 2, a dendritic cytoskeletal protein) was 26.67-fold higher than those in the Matrigel group (Fig. 3B). In addition, IF staining showed that the expression levels of Tuj1 and MAP2 in the BGA@GelMA group were 71.00 % and 62.67 % respectively, whether both more than 5 % those in the Matrigel group (Fig. 3D, E, G, and H). The results showed that the BGA@GelMA hydrogel promoted hNPCs to differentiate into mature neurons.

Fig. 3.

hNPCs in the BGA@GelMA hydrogel had more highly expressed genes related to neuronal differentiation than the Matrigel. (A to C) expression of TUJ1, MAP2 and GFAP of the BGA@GelMA and the Matrigel group as determined by RT-qPCR, n = 3, ***p = 0.0002, ***p = 0.0004, **p = 0.0090. Quantification of percentages of (D) TUJ1⁺, (E) MAP2⁺, (F) stem123⁺ cells, n = 3, ***p = 0.0001, ***p = 0.0008, ***p = 0.00092. The macroscopic image and quantification of immature neuron marker (G) TUJ1, mature neuron marker (H) MAP2, the astrocyte marker (I) stem123, Scale bar = 50 μm. (J) Schematic diagram of the workflow for RNA sequencing of the BGA@GelMA and the Matrigel hydrogel. (K) The diagram of PCA shows overall gene expression profiles at day-0 and day-14, respectively. (L) Volcano map with a threshold of log2 (fold change) > 1 differentially expressed genes between the BGA@GelMA and the Matrigel group. (M) Expression of genes related to the MAPK pathway between the BGA@GelMA and the Matrigel group. (N) The biological processes of the top-100 highly expressed genes by GO terms enrichment analysis.

During the early in vitro differentiation phase, the spontaneous differentiation of hNPCs into the glial lineage (including astrocytes and oligodendrocytes) reduces the efficiency of neuronal differentiation [25,55,56]. Thus, glial fibrillary acidic protein (GFAP), a marker of astrocytes, was evaluated to assess astrogliosis of hNPCs within the hydrogel. The transcription level of GFAP in BGA@GelMA group accounted for 21.48 % of the Matrigel group (Fig. 3C). IF staining showed that approximately 0.67 % of cells expressed stem123 (human GFAP specific mouse monoclonal antibody) compared with Matrigel group (70.67 %) (Fig. 3F and I). Additionally, oligodendrocyte transcription factor 2 (Olig2) and adenomatous polyposis coli (APC) were evaluated as markers of oligodendrocyte lineage differentiation. Relative to the Matrigel group, immunofluorescence staining demonstrated that the expression levels of Olig2 and APC in the BGA@GelMA group were 4.02- and 4.93-fold lower, respectively (Fig. S2A, B, and E, Supporting Information). These results suggested that the BGA@GelMA hydrogel suppressed spontaneous differentiation of hNPCs into the astrocyte and oligodendrocyte lineage, potentially contributing to enhanced neural differentiation potential.

The uncontrolled cellular expansion is a potential risk of tumorigenicity after transplantation [57]. Quantification of cellular numbers and Ki-67 antigen (Ki67) positive staining demonstrated that the BGA@GelMA hydrogels significantly decreased the uncontrolled proliferation (0.67 %) compared to the Matrigel group (52.33 %) (Fig. S2C, D, F, and G, Supporting Information). These results showed that the BGA@GelMA system decreased astrogliosis and uncontrolled proliferation, exhibiting excellent long-term safety for hNPCs-loaded therapy.

Additionally, RNA-sequencing was employed to assess the global transcriptional landscape of 3D-cultured hNPCs in the BGA@GelMA and Matrigel groups (Fig. 3J). Principal component analysis (PCA) indicated no significant differences between the two groups at an initial stage, but significant variations were observed after 14 days (Fig. 3K). Compared with the Matrigel group, the BGA@GelMA group upregulated 2803 genes and downregulated 2158 genes (Fig. 3L). Gene Ontology (GO) enrichment analysis identified GO terms highly related to glutamatergic transmission, L-glutamate transport among the top 50 enriched terms. In contrast, most GO terms captured in the Matrigel group were stem cell proliferation and glia cell differentiation (Fig. 3M, N, Fig. S3A, Supporting Information). These results further confirmed that the BGA@GelMA hydrogel promoted the maturation of interneurons through glutamatergic transmission.

Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis was applied to elucidate the molecular pathways associated with the differentially expressed genes of the BGA@GelMA hydrogel (Fig. S3B, Supporting Information). The KEGG-enriched bubble maps revealed that the mitogen-activated protein kinase (MAPK) pathway was linked to differential genes expression. MAPKBP1 (mitogen-activated protein kinase binding protein 1, a member of the MAPK phosphatase family) and GRIA3 (glutamate AMPA receptor, ionotropic type 3, the primary receptors for excitatory neurotransmission) in the MAPK signaling pathway were screened in the BGA@GelMA group. Moreover, the BGA@GelMA hydrogel enhanced the expression of MAPKBP1 and GRIA3 to 1.51-fold and 1.67-fold that of the Matrigel group, respectively. These results indicated that the neural differentiation produced by BGA@GelMA hydrogel mainly by activating the MAPK signaling pathway.

The BGA@GelMA hydrogel promoted interneurons differentiation and synaptic plasticity in vitro

The balance between excitatory (glutamatergic) and inhibitory (GABAergic) transmission is crucial for maintaining normal cerebral function [8,58,59]. However, TBI disrupts the regulation and balance of cortical interneurons [60]. Therefore, it is essential to support the differentiation of functional interneurons to improve the brain repair [61].

The RT-qPCR results showed that transcript levels of Glutamate (a marker of glutamatergic neurons) in the BGA@GelMA group were 2.05-fold as those in the Matrigel group, while the transcript levels of GABA (Gamma-Aminobutyric Acid, a marker of GABAergic neurons) were found to be 51.20 % lower in the BGA@GelMA group (Fig. 4A). Immunostaining revealed that the BGA@GelMA group increased the proportion of L-glutamic acid (L-Glu) positive glutamatergic neurons (27.67 % ± 2.08 %), compared to that in the Matrigel group (0.27 % ± 0.21 %) (Fig. 4B, C), as revealed by RNA-seq analysis (Fig. 4E). The BGA@GelMA group decreased the proportion of GABA positive neurons (6.03 % ± 1.95 %), compared to that in the Matrigel group (15.33 % ± 3.21 %) (Fig. 4D and F). The transcript levels of TH (tyrosine hydroxylase, a marker of cortical interneurons) were found to be 28.71-fold in the BGA@GelMA group. IF staining showed TH⁺ cells were observed within the BGA@GelMA group (Fig. 4H). In addition, GO enrichment analysis also identified genes related to synaptic neurotransmission were upregulated in the BGA@GelMA group (Fig. 4G). These results suggested that BGA@GelMA hydrogel has the potential to induce the differentiation of the cortical interneurons.

Fig. 4.

Maturation of interneurons and formation of neural circuits in the BGA@GelMA hydrogel in vitro. (A) Expression of Glutamate, GABA and TH of the BGA@GelMA and Matrigel group as determined by RT-qPCR, n = 3, **p = 0.001, **p = 0.0040, **p = 0.0011. (B) Quantification of percentages of L-Glu⁺, GABA⁺, and TH⁺ cells, n = 3, **p = 0.0023, *p = 0.0128, **p = 0.0049. IF staining with (C) L-Glu⁺, (D) GABA⁺, and (H) TH⁺ cells. Expression of genes related to (E) glutamate secretion, (F) GABAergic secretion and (G) neurotransmission in the BGA@GelMA and Matrigel group as determined by RNA sequencing. The macroscopic image of (I) mCherry⁺ rabies virus, (J) MAP2⁺ and PSD95⁺. (K) Quantification of PSD95⁺, n = 3, ***p = 0.00068. (L) The percentage of mCherry⁺/GFP^- from mCherry⁺/GFP⁺ infected neurons, n = 10, ***p = 0.00037. (M) Expression of genes related to synapse plasticity.

Synaptic integration is essential to functional restoration in the adult brain [8,58,62]. Immunostaining demonstrated the BGA@GelMA group significantly enhanced postsynaptic scaffolding protein 95 (PSD95)-positive synapses colocalized with MAP2-positive neurons (18.67 ± 7.77), compared to the Matrigel group (Fig. 4J, K). The results were consistent with the RNA-sequential analysis (Fig. 4M). To demonstrate the BGA@GelMA hydrogel's ability to establish neural circuits, hNPCs derived from the H1-CAG-GTRgp cell line were utilized (Fig. S1C and D, Supporting Information). After treatment of rabies virus labeled by mCherry, the newly tracer neurons (mCherry⁺/GFP^-) in the BGA@GelMA group were 11.0-fold as those in the Matrigel group (Fig. 4I, L). In addition, the patch clamp experiment showed that the neurons derived from hNPCs had a prominent spike after depolarization in the BGA@GelMA group, improving the nerve conduction (Fig. S3C, Supporting Information). Collectively, these results showed that the BGA@GelMA hydrogels facilitated the establishment the of neural network and induced the synaptic integration.

The BGA@GelMA hydrogel supported structural restoration of brain tissue in the TBI model

The efficacy of the hNPCs-loaded BGA@GelMA hydrogel (hNPCs + BGA@GelMA) in repairing TBI was further evaluated using a rat model. The experimental workflow is illustrated in Fig. 5A. At 4 weeks post-treatment, the hNPCs + BGA@GelMA group exhibited a significantly higher left forelimb touch rate in the cylinder test, as 1.49-fold and 1.15-fold compared to the hNPCs group and the hNPCs + Matrigel group, respectively (Fig. S4A, Supporting Information). In the mNSS assessment, neurological deficit scores in the hNPCs + BGA@GelMA group were significantly reduced by 2.21- and 2.42-fold compared to the hNPCs group and the hNPCs + Matrigel group, respectively (Fig. S4B, Supporting Information). In the rotarod test, the latency to fall was markedly increased by 5.93- and 2.30-fold relative to the hNPCs group and the hNPCs + Matrigel group, respectively (Fig. S4C, Supporting Information). These results indicated that the neural inducers-loaded hydrogel significantly improved motor coordination, suggesting that the cocktail strategy may facilitate structural and functional repair of the injured brain.

Fig. 5.

Survival and differentiation of grafted hNPCs in the collapsed core. (A) Schematic diagram and surgical procedure of the in vivo experimental processes. (B) The gross appearance of the injury site within the whole brain 4 weeks after transplantation. (C) H&E staining of brain injured cavities. (D) Representative images of survival grafts in the injured cerebral cortex. (E) Statistical analyses of injured cavity, n = 3, **p = 0.0016 (hNPCs + BGA@GelMA vs TBI). The percentage and mean intensity of (F) TUJ1⁺ and (G) MAP2⁺ neurons at the injury site, n = 3, ****p = 0.00007 (TUJ1⁺/hNA⁺ ratio, hNPCs + BGA@GelMA vs hNPCs), *p = 0.0211 (TUJ1⁺ intensity, hNPCs + BGA@GelMA vs hNPCs), **p = 0.0013 (MAP2⁺/hNA⁺ ratio, hNPCs + BGA@GelMA vs hNPCs), ****p = 0.00004 (MAP2⁺ intensity, hNPCs + BGA@GelMA vs hNPCs). Representative images of (H) TUJ1⁺ and (I) MAP2⁺ at the injury site.

The photograph of isolated brain showed that the cavities were filled by newborn tissue in the hNPCs + BGA@GelMA group at 4 weeks post-transplantation (Fig. 5B). Furthermore, H&E staining showed that the hNPCs + BGA@GelMA group was filled with an ECM-like microstructure, decreasing the injured cavities (17.33 % ± 25.70 %), compared to the sparse and disordered tissue in the TBI group (85.33 % ± 4.73 %), hNPCs (83.10 % ± 13.12 %), and hNPCs + Matrigel groups (71.20 % ± 5.95 %) (Fig. 5C and E). These findings indicated that the micro- and nano- structure of the BGA@GelMA hydrogel facilitated effective tissue reconstruction following injury.

BGA@GelMA hydrogel facilitates cortical interneuron regeneration and synaptic connectivity in situ

To track the fate of the transplanted hNPCs contributing to tissue regeneration in vivo, GFP fluorescence by CRISPER/Cas9 gene editing was infected before transplantation. The brains from hNPCs + BGA@GelMA group displayed the GFP-positive cells in the collapsed cortex through interval sections (Fig. 5D). Compared with the hNPCs group and the Matrigel group, the BGA@GelMA group exhibited significantly enhanced differentiation into Tuj1⁺ and MAP2⁺ neurons (Fig. 5F to I), Additionally, compared to the hNPCs and hNPCs + Matrigel groups, the hNPCs + BGA@GelMA group exhibited reduced expression of stem123⁺ and APC⁺ glial lineage markers (Fig. S4D to H, Supporting Information), as well as Ki67⁺, Nestin⁺, and OCT4⁺ cells associated with undifferentiated hNPCs (Fig. S4I to M, Supporting Information).

Consistent with in vitro results, the proportion of L-Glu⁺ glutamatergic neurons were significantly elevated in the BGA@GelMA group (47.79 % ± 6.62 %) compared to the Matrigel group (4.04 % ± 2.22 %) (Fig. 6A and G). A low proportion of GABAergic neurons (6.97 % ± 4.18 %) in the BGA@GelMA group was also detected, indicating a relatively balanced excitatory/inhibitory neuronal composition (Fig. S5A and E, Supporting Information). In addition, the TH-positive cortical interneurons were observed at the lesion in the hNPCs + BGA@GelMA group (Fig. S5B and F, Supporting Information). Together, the results showed that the hNPCs + BGA@GelMA group, with its unique micro- and nano-structure and neural inducers-rich microenvironment, enabled the restoration of brain tissue and regeneration of cortical interneurons in situ.

Fig. 6.

Functional interneurons generated from the graft and injured microenvironment recovery. The immunofluorescence staining of (A) L-Glu⁺ and (B) TBR1⁺ at the injury site. The percentage of (G) L-Glu⁺ and (H) TBR1⁺ co-labeled with hNA⁺ grafts at the injury site, n = 3, ****p = 0.00009 (hNPCs + BGA@GelMA vs hNPCs). Representative images of (C) NF200⁺ and (D) synapsin1⁺/PSD95⁺ at the injury site. Quantification of the (I) NF200 positive area and co-immunostaining of (J) synapsin1⁺/PSD95⁺ at the injury site, n = 3, **p = 0.0038 (hNPCs + BGA@GelMA vs hNPCs). Representative images of (E) Glut1⁺/PDGFR-β⁺ and (F) Glut1⁺/GFAP⁺ at the injury site. Quantification of the (K) Glut1 positive and (L) PDGFR-β positive area at the injury site, n = 3, *p = 0.0499 (hNPCs + BGA@GelMA vs hNPCs). (M) Schematic diagram of brain recovery in the hNPCs + BGA@GelMA group.

Accurate positional identity is essential for appropriate synaptic integration and function. To determine whether regenerated neurons possessed cortical layer-specific characteristics, co-staining for T-box brain protein 1 (TBR1, layer VI markers) and COUP-TF-interacting protein 2 (Ctip2, layer V markers) with human nuclear antigen (hNA) was performed. The percentage of TBR1⁺/hNA⁺ neurons was significantly increased in the BGA@GelMA group (48.50 % ± 4.83 %), compared to the hNPCs group (2.75 % ± 0.76 %) and the Matrigel group (4.10 % ± 5.63 %) (Fig. 6B and H). Similarly, the proportion of Ctip2⁺/hNA⁺ neurons was relatively higher in the BGA@GelMA group (Fig. S5C and G, Supporting Information). This indicated that graft-derived neurons acquired regionally appropriate cortical identity.

To determine whether regenerated interneurons supported functional repair, axonal and synaptic integration were evaluated [63]. NF200⁺ and synapsin1⁺/PSD95⁺ staining showed significantly enhanced axon growth and synaptogenesis in the BGA@GelMA group (Fig. 6C, D, I, and J). Despite the limited number of traced neurons observed at the transplantation site, rabies virus tracing demonstrated a relatively higher detection in the hNPCs + BGA@GelMA group compared to the hNPCs and hNPCs + Matrigel groups (Fig. S5D and H, Supporting Information). Additionally, the BGA@GelMA group exhibited a 2.95- and 3.43-fold increase in signal amplitude by MEP recording compared to the hNPCs and hNPCs + Matrigel groups, respectively (Fig. S5I to L, Supporting Information). These results indicated that BGA@GelMA treatment facilitated the reconstruction of cortical neural connectivity and the restoration of electrophysiological function in situ.

Collectively, the micro- and nano-structured BGA@GelMA hydrogel enriched with neural inducers provided a regenerative microenvironment that directed transplanted hNPCs toward cortical interneuron fates and promoted synaptic connectivity with host tissue.

The BGA@GelMA hydrogel promoted regeneration of NVU and reduced inflammation after transplantation

Regeneration of NVU and optimization of neuroinflammation are pivotal in augmenting the restoration of motor cortex [39,64]. The function of the NVU is controlled by interactions of endothelial cells, pericytes, astrocytes, and neurons [65,66]. Therefore, recruiting host cells by hydrogel can facilitate the reconstruction of NVU in defect area [67]. To examine the potential role of the hNPCs + BGA@GelMA group in recruiting host cells, IF staining was performed at 4 weeks post-transplantation. The host GLUT1⁺ (glucose transporter type 1⁺) vascular cells and the PDGFR-β⁺ (platelet-derived growth factor receptor beta)-positive pericytes in situ were examined in the hNPCs + BGA@GelMA group (Fig. 6E, F, K, and L). The results indicated that the micro- and nano-structure of the BGA@GelMA hydrogel promoted recruitment of host vascular cells. IF staining showed that and GFAP-positive astrocytes colocalized with GLUT1⁺ were abundant and uniformly distributed around GFP⁺ grafts (Fig. S6A, Supporting Information). These results indicated that the hNPCs + BGA@GelMA group promoted reconstruction the structure of NVU through host–graft integration.

To further validate functional reconstruction of NVU, multi-modal MRI techniques were performed. ASL imaging indicated increased cerebral blood flow at the lesion site in the hNPCs + BGA@GelMA group compared to the hNPCs and hNPCs + Matrigel groups (Fig. S6B and C, Supporting Information). BOLD imaging revealed prolonged T2* signals, suggesting improved tissue oxygenation (Fig. S6D, Supporting Information). Additionally, ^1H-MRS demonstrated an elevated signal at glucose normalized to Cr, suggesting relatively enhanced glucose metabolism in the BGA@GelMA group compared to the control groups (Fig. S6E, Supporting Information). These results suggested that the micro- and nano-structured and neural inducers-enriched hydrogel facilitated structural and functional reconstruction of the NUV at the injury site.

Additionally, studies have shown that M1 polarization of microglia can exacerbate brain damage by releasing pro-inflammatory cytokines, while promoting a shift towards M2 polarization is beneficial for brain repair [68,69]. Therefore, the inflammatory responses were further investigated in the core and peri-area of the implanted site. Compared with the hNPCs and hNPCs + Matrigel groups, the expression of pro-inflammatory M1 microglial markers, including Iba1⁺/IL-1β⁺ (ionized calcium-binding adapter molecule 1/interleukin-1 beta) and CD68 (cluster of differentiation 68), was significantly decreased in the hNPCs + BGA@GelMA group, whereas the expression of the anti-inflammatory M2 marker CD206 (macrophage mannose receptor 1) was notably increased (Fig. S6F to I). The results showed that the BGA@GelMA hydrogel did not significantly exacerbate the inflammatory response at the injury site, compared with hNPCs and hNPCs + Matrigel groups. Collectively, these data suggested that the micro- and nano-structure of BGA@GelMA hydrogel, sustained release of neural inducers and grafted hNPCs synergistically promoted brain recovery through restoring NVU and reducing inflammation.

Discussion

The variations in micro- and nano-structure, mechanical properties and environmental factors determined stem cells destiny [54]. However, hydrogels loading a single growth factor struggle to modulate cells exhibiting diverse morphological characteristics and to accomplish neural maturation [55,71]. In this study, a cocktail-loaded hydrogel was constructed to enhance the formation of functional interneurons and synaptic circuits. Park et al. reported that BDNF-loaded hydrogel can contribute to the axon elongation following a 7-day culture [72]. Garbayo et al. reported that encapsulated cells treated with GDNF reached survival rates that over 90 %, with 11.5 ± 4.04 μm neurites visualized [40]. The cocktail hydrogel here enhanced the axon extension to 79.12 μm ± 35.59 μm within 3 days (Fig. 1G to I). The dynamic cytoskeleton of hNPCs was mainly due to the micro-and nano-characterizations of the BGA@GelMA hydrogel (Fig. 1D) and the sustained release of neural inducers (Fig. 2A to C). Moreover, the BGA@GelMA hydrogel maintained excellent biocompatibility, with 1.48 % ± 1.07 % apoptosis cells and 93.62 % ± 1.83 % live cells within 14 days (Fig. 2E to H). Hence, hydrogel delivering multiple factors provides the advantage of inhibiting apoptosis and promoting neurogenesis.

Neural networks dictate the cortical-region functions, including motor regulation, which are essential in brain recovery [73]. The development of advanced therapeutic strategies through co-transplantation with single factor-based hydrogels could promote brain repair [74]. Lippmann et al. reported that a hydrogel with an N-cadherin had propagated to 423 PSD95 puncta per 75 μm³ and synaptic connections after 21 days [75]. Liu et al. reported that the hydrogel treated with BDNF formed 22.05 ± 0.37 synaptic spots for repairing synaptopathy [76]. In this study, the BGA@GelMA hydrogel not only differentiated hNPCs into PSD95⁺ synapses co-localized MAP2⁺ neurons, but also formed functional monosynaptic connections within 14 days (Fig. 4I to M). Thus, the multi-component hydrogel is more effective for the maturation of interneurons and the establishment of neural circuits.

Additionally, NVU reconstruction is important for facilitating brain injury recovery through modulating neuron-glia-vascular interactions [77]. Pre-study suggested that hydrogel that delivers neurotrophic factors has been effective in improving NVU development in vitro [78]. However, the complex post-injury brain milieu significantly impedes the in vivo regeneration of NVUs [29]. In this work, the prepared BGA@GelMA hydrogel facilitated the NVU formation through endogenous recruitment and graft-host integration (Fig. 6E, F, K, L and Fig. S6A, Supporting Information). Additionally, this cocktail strategy facilitated NVU functional recovery by promoting cerebral blood flow, tissue oxygenation, and metabolic activity (Fig. S6B to E, Supporting Information). Hence, the micro- and nano-characterizations of the BGA@GelMA hydrogel provide a biomimetic scaffold for reconstruction of NVU, improving the efficiency of hNPCs-loaded therapy. Collectively, the therapeutic efficacy of the BGA@GelMA hydrogel appears to derive from a synergistic interplay between transplanted hNPCs and the hydrogel-mediated modulation of the post-injury microenvironment, beyond serving as a vehicle for cellular delivery.

Previous studies suggested that the activation of the MAPK pathway is crucial for neural differentiation by promoting the phosphorylation of extracellular signal-regulated kinase (ERK) 1/2 and the expression of runt-related transcription factor 2 (RUNX2) [79]. In this study, the BGA@GelMA hydrogel provided a neuronal-specific microenvironment for neural differentiation, after which the hNPCs were induced to secrete glutamate, and the MAPK signaling pathway was activated (Fig. 3N). Furthermore, our findings are validated by the research of Wang et al., which demonstrated that the neural networks promoted recovery of motor function through neurotransmitters [80]. Hence, the utilization of hydrogel to modulate glutamate reuptake and metabolic pathways holds the potential for enhancing the regeneration of cerebral tissue [81].

However, there are some limitations to this study. Firstly, a precise proportion of neural subtypes differentiated from hNPCs is necessary to form the specific neural networks. Moreover, while the results showed the survival and differentiation of transplanted hNPCs as well as graft vascularization at 4 weeks post-transplantation, the longer-term integration and functional contribution should be further evaluated. The four-week time point was selected based on both prior studies and preliminary data, which demonstrated meaningful cell survival, differentiation, and behavioral recovery at this stage [82]. Nonetheless, to better understand the durability and physiological relevance of the observed regeneration, future studies should extend the follow-up period and incorporate longitudinal analyses of graft-host integration and functional performance. Furthermore, given the essential role of oligodendrocytes in axonal myelination and the maintenance of long-term neuronal function, future studies should also aim to elucidate the dynamics and extent of oligodendrocytic maturation over extended post-transplantation periods.

In future investigations, the current therapeutic strategy is intended to be further validated by extension from rodent models to non-human primates to better evaluate its translational potential. Additionally, the hydrogel formulation can be scaled from millimeter to centimeter dimensions to enable sustained release of neural inducers appropriate for larger injury volumes. Moreover, by leveraging the photosensitive properties of the hydrogel combined with advanced 3D printing technologies, personalized and precise therapeutic scaffolds can be engineered to accommodate individualized treatment needs. Finally, long-term preclinical studies (three months or longer) are needed to confirm chronic biosafety and durable efficacy. Collectively, these efforts will support the development of the current strategy as a novel and clinically translatable approach for brain repair.

Conclusion

In summary, the neural inducers-loaded hydrogel was developed for differentiation and functional synaptogenesis of cortical interneurons. The BGA@GelMA hydrogel with excellent cytocompatibility and biocompatibility enhanced the maturity and synaptic networks of hNPCs by secreting glutamatergic transmitters in vitro. In vivo, the micro- and nano-structure of the BGA@GelMA hydrogel facilitated the generation of cortical interneurons by the restoration of NVU and the mitigation of the inflammatory response. Collectively, this finding opens an avenue to enable differentiation and maturation of the transplanted hNPCs under brain repair, which will provide definite evidence for further pre-clinical cell-laden therapy.

Data availability statement

All data needed to evaluate the conclusions are present in the paper and the supplementary materials. Additional data related to this paper may be requested from the authors.

Compliance with Ethics Requirements

All Institutional and National Guidelines for the care and use of animals (fisheries) were followed.

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.

Appendix A. Supplementary data

The following are the Supplementary data to this article: