Background

With latest advances in biological sciences, transplantation of neural stem cells (NSCs) shows great potential for treating neurodegenerative diseases. In order for this strategy to be successful, transplanted NSCs must replace ailing and dead neurons and form neural connections with remaining healthy neurons, to a significant extent so that at least part of original neural functions can be restored [1, 2]. Therefore, it is critically important to promote NSC development and maturation in targeted brain region [3]. However, low survival rate, unwanted cell differentiation and lack of functional integration are among the biggest obstacles for clinical application of NSCs in therapy [4, 5].

The brain consists of a complex network of neurons located in the extracellular matrix (ECM) with heterogenous components and three-dimensional structures [6]. Under physiological conditions, ECM is composed of proteins and polysaccharides, secreted and assembled by cells, providing channels for them to send and receive chemical, electrical and mechanical signals [7, 8]. In cell transplantation, extracellular matrix proteins are shown to possess biological properties in guiding cell behavior and/ or tissue regeneration but often lack sufficient mechanical properties. However, these properties are required for biomaterials and physically guide cell fate and function [9] through elastic modulus, electrical conductivity, topology, spatial dimension, and dynamic degradability, such as cell adhesion [10], migration [11], proliferation [12], differentiation [13] and stem cell maintenance [14]. It has been shown that the composition, physical properties, surface and structural features of biomaterials can modulate cell behavior and guide the process of tissue regeneration, mainly through the direct interaction between cells and biomaterials [15, 16].

Ion channels are essential for regulating the physiological function of cells and maintaining the stability of the intracellular environment. Because of their unique molecular structure and contact sites, ion channels have become potential targets for extracellular stimuli, including synthetic drugs and biomaterials [17]. In recent years, many studies have reported that biomaterials can interact with ion channels directly and / or indirectly, change ion homeostasis, ion channel currents, channel dynamics, ion channel-related RNA and protein expression levels, and trigger a variety of intracellular signal transduction pathways [17,18,19].

Among the biomaterials tested so far, graphene [20,21,22,23], carbon nanotubes [24, 25] and their derivatives have been widely used to create microenvironment for neural stem cells, and have shown the potential to promote nerve regeneration and the ability to regulate ion currents [26,27,28,29,30]. In addition, their excellent thermal, electrical and mechanical properties make them show great prospects in terms of high-resolution and stable neural interfaces [31]. However, the hydrophobicity of these carbon-based materials limits their surface functionalization and further applications [32].

In contrast, Ti3C2Tx MXene, a new kind of two-dimensional (2D) crystal nanomaterial in a general formula of Mn+1XnTx, represents a new and more promising direction. Due to the presence of transition metals (M in the formula, such as Ti and Zr) and surface groups (T in the formula, such as O2− and OH−), Ti3C2Tx MXene is not only conductive but also hydrophilic [33, 34], affording rich anchoring sites and modifiability and making it attractive for many applications, including nanomedicine [35,36,37,38,39,40], biosensor [41,42,43,44,45], antimicrobial therapy [46, 47], self-cleaning [48, 49], biological imaging [50, 51] and therapeutic diagnostics [52, 53].

With the increasing research on MXenes, these materials have been successfully applied in the field of biomedicine [54]. 2D structures of MXenes with abundant surface anchoring sites have potential to be used as drug or protein carriers [55]. MXenes is considered as a high potential antimicrobial agent for inhibiting the growth of bacteria and fungi because of its large specific surface, operable surface functionalization and the potential to load different antibacterial functional groups [47, 56]. MXenes absorbs non-toxic electromagnetic radiation in the near infrared spectrum and is used in photothermal therapy and photoacoustic imaging of cancer [57, 58]. Due to the interaction of surface hydrophilicity, metal conductivity and two-dimensional layered atomic structure [59], MXenes can be used to sense methanol and other gases, glucose, dopamine and nitrite in the field of biosensors [60].

Recently, several different compositions of MXenes have been proved to be biocompatible, with no cytotoxicity detected in living nerve tissue [61, 62]. Two-dimensional Ti3C2 MXene is also used in nerve recording in vivo, providing a high-resolution neural interface for neuroelectronic devices. Dissociated cortical neurons cultured on Ti3C2 could adhere normally, grow neurites and form neural networks [31]. In addition, NSCs cultured on laminin-coated Ti3C2Tx MXene nanosheets formed stable adhesion, retained the ability of proliferation and showed neurite growth, increased differentiation rate and synaptic formation [63]. Therefore, from the existing research results, it can be clearly seen that the outstanding electrical and surface functional properties of MXenes make them have potentially important applications in nerve tissue engineering.

However, it remains unclear how MXenes interact with ion channels in close contact, and how the electrophysiological properties of neurons are impacted by it. In the present study, we prepared biocompatible and conductive Ti3C2Tx MXene films and cultured neural stem cells on them. We focused on dissecting the effects of MXenes on the ion channels of NSC-derived neurons and synaptic connections between them, expanding our knowledge on the electrophysiological mechanisms of MXenes in promoting the maturation of neurons and neural circuits.

Materials and methods

Ti3C2Tx MXene solution preparation

Multilayer Ti3C2Tx were synthesized by selective etching of the Al atomic layer in Ti3AlC2 in a mixture of HCl and LiF. Ti3AlC2 (1 g, MAX phase) was slowly introduced into the mixture of HCl (8 M, 10 mL) and LiF (1 g) in ice bath and then incubated at 40 °C for 30 h with stirring. With the Al layer removed, Ti3C2Tx solution was washed with deionized water and centrifuged at 4000 rpm for 10 min. The washing and centrifugation steps were repeated about 10 times until the pH reached ~ 6. The Ti3C2Tx was next mixed with deoxidized and deionized water and sonicated for 1 h in ice bath. After centrifugation at 4000 rpm for 20 min, the Ti3C2Tx MXene colloidal solution in water was collected as the dark supernatant, and then spin-coated on the tissue culture polystyrene (TCPS). The area of each TCPS (10 mm) is about 0.785 cm2, and the amount of Ti3C2Tx MXene is 6 μg.

Ti3C2Tx MXene film characterization

Transmission electron microscopy images and corresponding SAED of Ti3C2Tx MXene flakes were obtained by transmission electron microscopy (TEM, JEOL JEM-2100F). The adhesion and growth of NSCs on Ti3C2Tx MXene were observed by scanning electron microscopy (SEM, Hitachi Regulus8100). Raman measurements of Ti3C2Tx MXene film were conducted on a confocal Raman microscope (Horiba LabRAM HR800) at room temperature with an excitation wavelength of 633 nm. X-ray photoelectron spectroscopy (XPS) was performed with a PHI QUANTERA II (760Zi-A 6 GHz) and the wide scanning of samples was selected in the range of 0–1200 eV. The contact angle measurements were taken on an OCA 15 plus (Dataphysics).

NSCs extraction and culture

NSCs were taken from the left and right hippocampus of 18 days old mouse embryos, at a time when these cells are highly proliferate, followed the guidelines and agreements approved by the Animal Care and Use Committee of Southeast University. Briefly, blood vessels and meninges on the hippocampus were removed in phosphate-buffered saline (PBS) at 4 °C. Then PBS was fully removed and Accutase solution (Thermo, USA) was added to digest the hippocampus in a 5% CO2 incubator at 37 °C for 15 min. After Accutase solution was fully washed, DMEM-F12 medium containing 2% B27, 1% Penicillin–Streptomycin, 10 ng/mL EGF (Thermo, USA) and 10 ng/mL FGF (Pepro tech, USA) was added. Then, gently grinding was applied to the dissociated tissues with a pipette tip till they were dispersed as much as possible, and cells were filtered and transferred to culture flask (Corning, USA) for proliferation in the incubator at 37 °C. After the purification of the neurospheres for three generations, individual NSCs were seeded on either TCPS or monolayer Ti3C2Tx MXene (MXenes, TCPS spin-coated with Ti3C2Tx MXene) coated with laminin (Sigma, USA) at a density of 2 \(\times\) 106/mL, then the differentiation medium (Stemcell, Canada) was changed after 24 h of proliferation. The differentiation process lasted for 14 days, and the medium was changed every 3 days.

Immunofluorescence

For viability test, cells were washed with PBS after three days in vitro in proliferation medium, transferred to DMEM/F12, containing 0.05% Calcein-AM and 0.025% EthD-1, in the incubator for 20 min, and then washed with DMEM/F12. After sealing with DAKO, cells were mounted for confocal imaging. For immunofluorescence staining, cells were washed with PBS at 7 DIV (differentiation in vitro), fixed in 4% paraformaldehyde for 50 min, blocked in PBS containing 2% BSA, and permeated with 0.1% TritonX-100 for 90 min. The cells were incubated with primary antibodies overnight at 4 °C, and then stained with DAPI (Solarbio, China) following incubation with secondary antibodies for 60 min. The antibodies used in these experiments include primary antibodies against β-tubulin (Beyotime, China), Nestin (Beyotime, China), MAP2 (Abcam, UK), Synapsin-1 (Cell Signaling Tech, USA) and PSD95 (Millipore, Germany). After sealing, cells were mounted for imaging under a Zeiss 700 laser scanning confocal microscope.

Electrophysiological recordings

Cells grown on TCPS or MXenes were immersed in an oxygenated extracellular solution containing (in mM) 135 NaCl, 4 KCl, 2 CaCl2, 1 MgCl2, 10 HEPES, 10 D-glucose (pH 7.40, osmolarity 300 mOsm). The cells were visualized through a 60X water-immersion objective in an up-right microscope (Olympus), and patch-clamp recordings were made through an EPC10/2 amplifier (HEKA Electronics, Lambrecht Pfalz, Germany), driven by a PC computer running Patchmaster (HEKA Electronics). Recording pipettes were pulled from borosilicate glass capillaries (Sutter) and filled with an internal solution containing (in mM) 135 K-methane sulfonate, 20 KCl, 2 EGTA, 3 Mg-ATP, 10 HEPES, and 0.5 Na-GTP (pH 7.20, osmolarity 300 mOsm). Cells were all held at − 90 mV and the uncompensated series resistance had a typical value of 6 ~ 10 MΩ. For experiments on voltage-gated Ca2+ or Na+ current, a Cs+-based internal solution containing (in mM) 135 Cs-methane sulfonate, 10 CsCl, 10 TEA-Cl, 10 HEPES, 2 EGTA, 3 Mg-ATP and 0.5 Na-GTP (pH 7.20, osmolarity 300 mOsm) was used. To examine the spiking behavior of these neurons, current steps of 500 ms with an amplitude of 10 ~ 100 pA were applied under the current-clamp mode. Between stimulations, a negative leaking current was injected to maintain a membrane potential of ~ − 90 mV, and spontaneous action potentials and resting membrane potentials were accessed by holding cells at zero current. Signals were filtered at 2 kHz and sampled at 100 kHz. All patch-clamp experiments were performed at room temperature and the liquid junction potential (~ 10 mV) was corrected offline.

Statistical analysis

Morphological analysis was carried out in Image J Pro Plus, and statistical analyses were performed in GraphPad Prism 7. A home-made macro written in Igor Pro 6.22 was used for detection of spontaneous synaptic events. Origin 2018 was used to draw activation and inactivation curves. Throughout the study, data are presented as mean ± SEM. Statistical significance was assessed by either two-way ANOVA test followed by Bonferroni multiple comparison test, or Student’s t-test, and p < 0.05 was considered significant.

Results and discussion

Ti3C2Tx MXene fabrication

We fabricated Ti3C2Tx MXene through selective HF etching of the Al layer in Ti3AlC2 and characterized its morphology with transmission electron microscopy (TEM). As shown in the TEM image in Fig. 1A, the prepared MXene nanosheets display monolayer structure and two-dimension features with lateral sizes in the range of 100 nm ~ 800 nm, and the inset on the TEM image depicts a representative example of selected area electron diffraction (SAED), demonstrating a hexagonal crystal structure of Ti3C2Tx MXene. The structure and chemical composition of Ti3C2Tx MXene was investigated by Raman spectroscopy and X-ray photoelectron spectroscopy (XPS). Figure 1B shows the contact angle of water droplets on the nanosheets (62.47° ± 2.77°) is steeper when compared to that on the surface of tissue culture polystyrene (TCPS, 41.47° ± 3.31°), which may be due to the large surface roughness of Ti3C2Tx MXene. After coating with laminin, there was almost no difference in the hydrophilicity of MXenes surface (69.90° ± 0.81°), and was as close as possible to the coated TCPS (60.56° ± 0.96°). As shown in Fig. 1C, Raman fingerprint of Ti3C2Tx MXene was located in the range of 100 ~ 800 cm−1. The peaks at 199 and 715 cm−1 were the A1g symmetry out-of-plane variation of Ti and C atoms, respectively. Furthermore, the peaks at 286, 369 and 623 cm−1 were the Eg group vibration of Ti, C and surface functional group atoms, respectively. In addition, the image of the nanosheets taken with X-ray photoelectron spectroscopy (XPS) indicates mainly diffraction peaks of such elements, including Ti 2 s, Ti 2p, O 1 s, C 1 s, and F 1 s, as expected from Ti3C2Tx MXene (Fig. 1D). Figure 1E–G displayed the C1s, Ti 2p, and O 1 s XPS spectra of Ti3C2Tx MXene as well. Finally, for hydrophilic test, the contact angles of water droplets on different surface states were measured after laminin coating. Taken together, these results demonstrated successful fabrication of Ti3C2Tx MXene in our hands. All of the following experimental TCPS and MXenes have been treated with laminin.

Fig. 1
figure 1

Characterization of Ti3C2Tx MXene. A Transmission electron microscopy (TEM) image of monolayer Ti3C2Tx MXene flakes. The inset shows a representative example of selected area electron diffraction (SAED). B Images of water contact angle on TCPS and Ti3C2Tx MXene before and after laminin treatment. C Raman spectroscopy of Ti3C2Tx MXene. D X-ray photoelectron spectroscopy (XPS) of Ti3C2Tx MXene. E–G XPS spectra of the C 1 s (E), Ti 2p (F), and O 1 s (G) regions

Full size image

Ti3C2Tx MXene promotes differentiation and morphological maturation of NSCs

It has been reported previously NSCs can adhere, grow and differentiate into neurons to form neural networks on MXenes and survive up to at least 7 days [63]. However, it is not clear if NSCs on MXenes can continue to develop to functionally mature stages. Therefore, we isolated NSCs from mouse hippocampus and stained for nestin, and the immunofluorescence images showed that almost all cells were nestin positive, indicating that the extracted cells were NSCs (Additional file 1: Fig. S1). The NSCs were then seeded on TCPS and Ti3C2Tx MXene substrate, respectively. The Live/Dead assay demonstrated that the Ti3C2Tx MXene displayed great cytocompatibility (Additional file 2: Fig. S2A and 2B). Afterwards, SEM was used to observe the adhesion, axon elongation and growth on TCPS and Ti3C2Tx MXene (Additional file 2: Fig. S2C). It was suggested that the NSCs grown on Ti3C2Tx MXene had similar growth morphology to those on TCPS. Subsequently, we conducted differentiation assay to examine the morphological and electrophysiological properties of neurons derived from NSCs.

After 7 days of differentiation, the cells were stained with β-tubulin, a specific marker for neurons (Fig. 2A). Compared to those seeded on TCPS, the proportion of NSC-differentiated neurons (i.e. β-tubulin positive cells) on Ti3C2Tx MXene was increased significantly (TCPS: 21.57 ± 0.73%, n = 16 cultures; MXene: 24.90 ± 0.76%, n = 18 cultures; p < 0.01, Fig. 2B), and the MAP2 positive cells on Ti3C2Tx MXene was increased as well (TCPS: 25.18 ± 0.92%, n = 10 cultures; MXene: 30.12 ± 0.92%, n = 12 cultures; p < 0.01, Additional file 3: Fig. S3). In addition, we found that neurons grown on MXenes exhibited typical morphological maturation with neurite extension (Fig. 2C). Remarkably, the average length of neurites on MXenes was significantly longer (TCPS: 29.84 ± 2.19 μm, n = 11 cells; MXenes: 41.75 ± 2.75 μm, n = 13 cells; p < 0.01, Fig. 2D), and the numbers of branch points (TCPS: 3.45 ± 0.43, n = 11 cells; MXenes: 4.82 ± 0.46, n = 11 cells; p < 0.05, Fig. 2E) and branch tips (TCPS: 7.64 ± 0.66, n = 11 cells; MXenes: 10.82 ± 0.82, n = 11 cells; p < 0.01, Fig. 2F) on MXenes increased significantly as well. Consistent with a previous study [63], our results demonstrated that MXenes could significantly promote the differentiation of NSCs into neurons and even increase the length of their neurites.

Fig. 2
figure 2

Ti3C2Tx MXene promoted NSCs development and growth. A and B Representative images of NSC-derived neurons stained with β-tubulin antibody and DAPI at 7 DIV. Scale bar = 50 μm. (A), and the percentage of β-tubulin positive cells on TCPS and MXenes (B). C–F Two representative new generated neurons cultured on TCPS and MXenes. Scale bar = 10 μm. (C), the average length of neurites (D), the number of branch points (E), and the number of branch tips (F) for new neurons cultured on TCPS and MXenes. * indicates p < 0.05, ** indicates p < 0.01

Full size image

Ti3C2Tx MXene increases voltage-gated Ca2+ current in NSC-derived neurons

During the development and maturation of neurons, they must acquire a variety of ion channels that are essential for spiking and synaptic transmission [17, 64,65,66]. Given the close contact of Ti3C2Tx MXene with the plasma membrane of neurons cultured on it, it is conceivable that ion channels imbedded in the plasma membrane could be altered by Ti3C2Tx MXene, directly or indirectly. To address this question, we decided to perform patch-clamp recording on NSC-derived neurons cultured on TCPS or Ti3C2Tx MXene and investigate the currents flew through voltage-gated Na+, K+ and Ca2+ ion channels.

First, we examined voltage-gated Na+ current (INa), which is essential for the depolarization phase of spikes [17]. To isolate INa from the whole-cell current, we used a Cs+-based internal solution and included 2 mM 4-AP and 10 mM TEA-Cl in the external solution to block voltage-gated K+ current (IK), and added 0.2 mM CdCl2 in the external solution to block voltage-gated Ca2+ current (ICa). After breaking into the whole-cell mode, we held the cell at − 90 mV under voltage-clamp, applied voltage steps of − 100 ~ + 10 mV, 5 mV a step, for 150 ms, and recorded the resulting currents (Fig. 3A). As shown in the current–voltage (IV) curve plotted in Fig. 3B, INa recorded from neurons on both TCPS and Ti3C2Tx MXene was activated rapidly at about − 60 mV and reached the peak at − 40 ~ − 30 mV, and there is no significant difference in the peak amplitude between the two groups (TCPS: 1.63 ± 0.14 nA, n = 15; MXenes: 1.63 ± 0.15 nA, n = 18; p > 0.05).

Fig. 3
figure 3

Ti3C2Tx MXene bore no effect on the activation of INa or IK. A–C Representative INa recorded in two NSC-derived neurons under voltage-clamp, in response to voltage steps (from − 100 mV to + 10 mV, 150 ms, 5 mV a step). Their current–voltage (IV) and activation curves, pooled from multiple cells, are shown in B and C, respectively. D–F IK,A recorded under voltage-clamp, induced through voltage steps (− 80 ~ + 70 mV, 80 ms, 10 mV a step). G–I IK,dr recorded under voltage-clamp, in response to the same voltage steps as in D. Neither the amplitude (B, E and H) nor the activation (C, F and I) of INa, IK,A or IK,dr was significantly altered by Ti3C2Tx MXene

Full size image

We then converted current into conductance and fitted the activation of INa with a Boltzmann equation:

$${\text{G}}/{\text{G}}_{\text{max}}\text{=1/}\left\{{1} \, \text{+} \, {\text{exp}}\left[\left({\text{V}}_{\text{m}}-{\text{V}}_{1/2}\right)\text{/k}\right]\right\}$$

where G is the conductance, Vm is the membrane potential, V1/2 is the half-activation potential, and k is the slope factor (shown in Fig. 3C). We found that there is no significant difference in V1/2 or k between TCPS and MXene groups (TCPS: V1/2 = − 47.85 ± 0.43 mV, k = 3.51 ± 0.37, n = 15; MXenes: V1/2 = − 45.67 ± 0.14 mV, k = 2.83 ± 0.13, n = 18; p > 0.05), suggesting the activation of voltage-gated Na+ channels were not significantly altered by Ti3C2Tx MXene. Next, we studied the inactivation of INa with double-pulse stimulation: 150 ms conditioning pre-pulse from − 100 to 0 mV in 5 mV increment, followed by a 50 ms depolarizing test pulse to 0 mV (as shown in Fig. 4A). The peak amplitudes of INa were normalized and the inactivation curve was fitted with the same equation above (Fig. 4B). We found that \({\text{V}}_{1/2}\) and k of INa inactivation in the TCPS group was comparable to that of the MXene group (TCPS: V1/2 = − 50.92 ± 1.46 mV, k = 7.12 ± 0.83, n = 23; MXenes: V1/2 = − 50.96 ± 1.92 mV, k = 7.82 ± 0.98, n = 25; p > 0.05), indicating the inactivation of voltage-gated Na+ channels was not significantly altered by Ti3C2Tx MXene either.

Fig. 4
figure 4

Ti3C2Tx MXene had no effect on the inactivation of INa or IK,A. A and B INa recorded in two NSC-derived neurons under voltage-clamp, in response to double-pulse stimulation (conditioning pre-pulse: from − 100 mV to 0 mV, 150 ms, 5 mV a step; depolarizing test pulse: 0 mV for 50 ms). C and D IA recorded under voltage-clamp, in response to double-pulse stimulation (pre-pulse: from − 120 mV to + 30 mV, 80 ms, 10 mV a step; test pulse: + 50 mV for 80 ms)

Full size image

Second, we turned to voltage-gated K+ current (IK), which dictates the repolarization phase of spikes, and to a large extent, repetitive spiking, mainly control the shape of the action potential, adjust the frequency of the action potential and the resting membrane potential [17]. To isolated the total IK, we included 1 μM tetrodotoxin (TTX) and 0.2 mM CdCl2 in the external solution to block INa and voltage-gated Ca2+ current, respectively. In addition, a subtype specific IK blocker, TEA-Cl or 4-AP was also included to isolated the transient K+ current (IK,A, sensitive to 4-AP at 2 mM) and delayed rectifier K+ current (IK,dr, sensitive to TEA-Cl at 10 mM), respectively. As shown in Fig. 3, neither IK,A nor IK,dr was significantly altered by Ti3C2Tx MXene, in terms of the peak amplitude or activation. Given that IK,dr exhibited only minimal inactivation, we examined the inactivation of IK,A only, and we once again found that it was not significantly altered (Fig. 4). Both INa and IK are critically important for spiking, a central functional requirement for neurons. The fact that no significant alteration by Ti3C2Tx MXene was observed in either current, by far is the ultimate testament of Ti3C2Tx MXene’s safety and biocompatibility.

Lastly, we studied voltage-gated Ca2+ current (ICa), which controls neurotransmitter release and regulates many other neuronal functions, including excitability and axon growth [17]. To isolated ICa, we used a Cs+-based internal solution and added TTX (1 μM), TEA-Cl (10 mM) and 4-AP (2 mM) in the external solution to block INa and IK. When depolarization steps of − 70 ~ + 40 mV (5 mV a step) were applied to cells under voltage-clamp, small resulting inward currents were recorded (Fig. 5A), and the IV curve was constructed. As shown in Fig. 5B, ICa in NSC-derived neurons on both TCPS and MXenes was activated at about − 60 mV. However, ICa in neurons cultured on MXenes has a significantly larger amplitude at multiple voltages (For a voltage step of -10 mV, TCPS: 168.5 ± 15.5 pA, n = 45; MXenes: 248.1 ± 21.0 pA, n = 41; p < 0.0001). In alignment with our analysis on INa and IK, we also examined the activation and inactivation of ICa, and we found neither of the two characteristics was significantly altered by Ti3C2Tx MXene (Fig. 5). Besides depolarizing the membrane potential, Ca2+ influx from ICa serves as triggers for a variety of intracellular signal transduction cascades, including synaptic vesicle release [67] and axon growth [68,69,70]. For establishing the relationship between Ca2+ influx and neurites growth, as previously reported [71,72,73], Ca2+ influx was increased by depolarizing with elevated extracellular K+ (10 mM KCl). After increasing intracellular Ca2+, the average length of neurites of NSC-derived neurons cultured on TCPS was significantly longer, (TCPS + 0 mM KCl: 55.15 ± 2.12 μm, n = 20 cells; TCPS + 10 mM KCl: 79.43 ± 4.15 μm, n = 23 cells; p < 0.0001, Additional file 4: Fig. S4) and comparable to the length of neurites cultured on MXenes without extra Ca+ influx (MXenes + 0 mM KCl: 88.43 ± 4.53 μm, n = 10 cells), suggesting MXenes lengthen neurites through increasing Ca2+ influx. Furthermore, there was no additional influence on neurites elongation on Ti3C2Tx MXene film after adding K+ to elevate intracellular Ca2+ (MXenes + 10 mM KCl: 91.98 ± 4.99 μm, n = 12 cells), indicating that the effects of MXenes on promoting neurite growth and maturation of neurons may be mediated merely through increasing Ca2+ influx. Our finding of an increased ICa on Ti3C2Tx MXene could contribute to the longer neurites of neurons described above, as well as the boosted spiking and enhanced synaptic transmission that are to be followed.

Fig. 5
figure 5

Ti3C2Tx MXene increased the amplitude of ICa without changing its voltage-dependent activation or inactivation. A–C Representative ICa recorded in two NSC-derived neurons under voltage-clamp, in response to voltage steps from -70 mV to + 40 mV, 200 ms, 5 mV a step, and their IV and activation curves, with data pooled from multiple cells. D and E ICa recorded under voltage-clamp, induced by double-pulse stimulation (pre-pulses: − 90 ~ + 20 mV, 200 ms, 5 mV a step; test pulse: + 20 mV for 200 ms). * indicates p < 0.05, ** indicates p < 0.01, *** indicates p < 0.001, **** indicates p < 0.0001

Full size image

Ti3C2Tx MXene boosts spiking of NSC-derived neurons

In addition to expression of voltage-gated currents, the therapeutic potential of NSCs relies on their continuation of development and maturation to acquire capabilities to fire action potential spikes and form synaptic connections. We therefore decided to examine the electrophysiological properties of NSC-derived neurons cultured on Ti3C2Tx MXene from 8 to 14 DIV. Each day, we picked one batch of NSC-derived neurons and performed whole-cell patch-clamp recording. We first focused on their passive membrane properties. To this end, we kept cells under current-clamp, injected step currents and recorded voltage responses (Fig. 6). Based on these voltage responses, we determined the input resistance, resting membrane potential and action potential threshold, which are widely considered to be indicators of cell maturation and health [74]. We found that there is no significant difference between neurons cultured on TCPS or Ti3C2Tx MXene in any of the three parameters: the input resistance (TCPS: 1267 ± 82.28 MΩ, n = 57 cells; MXenes: 1325 ± 66.39 MΩ, n = 110 cells), resting potential (TCPS: − 49.65 ± 1.37 mV, n = 57 cells; MXenes: − 49.33 ± 1.02 mV, n = 110 cells) and action potential threshold (TCPS: − 44.84 ± 0.85 mV, n = 57 cells; MXenes: − 45.70 ± 0.68 mV, n = 110 cells), consistent with previous studies [27, 75,76,77].

Fig. 6
figure 6

Passive membrane properties of NSC-derived neurons cultured on TCPS and MXenes after 7 DIV. A Representative voltage responses to step current injection, recorded from a single NSC-derived neuron. B Input resistance of new neurons cultured on TCPS and MXenes, determined by taking the largest voltage response with no spikes triggered and dividing the voltage change with the current injected (see the lower trace in A). C Resting membrane potential determined by holding cells at zero current. D Action potential threshold measured at the inflex point before an action potential (see the upper trace in A)

Full size image

Next, we investigated the spiking pattern of neurons cultured on TCPS and Ti3C2Tx MXene. We first recorded spontaneous action potentials by holding cells at zero current under the current-clamp mode (Fig. 7A). We found that neurons cultured on Ti3C2Tx MXene fired spontaneous spikes as frequently as those on TCPS (TCPS: 0.97 ± 0.151 Hz, n = 62 cells; MXenes: 1.29 ± 0.17 Hz, n = 112 cells; Fig. 7B). We then applied step currents from 10 to 100 pA with 10 pA increment and recorded the spiking responses (Fig. 7C). We found that for step current injection of 10 ~ 70 pA, there is no significant difference in the spiking frequency between neurons cultured on TCPS or Ti3C2Tx MXene. This is consistent with the previous findings that no significant alteration by Ti3C2Tx MXene was observed in either INa and IK (Figs. 3 and 4). For stronger current injection of 80 ~ 100 pA, however, neurons cultured on Ti3C2Tx MXene fired significantly more spikes when compared to those cultured on TCPS (For current injection of 80 pA, TCPS: 3.57 ± 1.01, n = 21 cells; MXenes: 7.08 ± 0.81, n = 38 cells; p < 0.01, Fig. 7D). While the exact underlying mechanisms are not clear, increased voltage-gated Ca2+ current in neurons cultured on Ti3C2Tx MXene, as described above, was likely to be involved in the boosted spiking observed in these neurons.

Fig. 7
figure 7

Ti3C2Tx MXene selectively enhanced NSC-derived neurons spiking under strong depolarization. A Representative spontaneous spikes recorded from two new generated neurons cultured on TCPS (black) and MXenes (red), by holding the cells at zero current. B Frequency of spontaneous spikes in new neurons cultured on TCPS and MXenes. C Representative voltage responses evoked by current injection, recorded from two new neurons cultured on TCPS (black) and MXenes (red). D Number of spikes evoked by step current of 10 ~ 100 pA. ** indicates p < 0.01

Full size image

Taken together, our results on the passive membrane properties and spiking behavior strongly argue that Ti3C2Tx MXene are safe and highly compatible with living cells, making them an attractive biomaterial for constructing neural interface and scaffold. The increased spiking frequency for NSC-derived neurons cultured on Ti3C2Tx MXene in response to strong depolarization suggests Ti3C2Tx MXene is capable to facilitate or promote NSC development and maturation.

Ti3C2Tx MXene enhances synaptic transmission between NSC-derived neurons

During the development of NSCs, synaptic connections are formed and strengthened [26]. We therefore decided to investigate if Ti3C2Tx MXene could bear any difference in these synaptic connections between NSC-derived neurons. First, we examined the number of synapses. Neurons derived from NSCs were co-stained with PSD95 and Synapsin-1, puncta of synapses can be found in these neurons grown on both TCPS and MXenes (Fig. 8A), and there is no significant difference in the number of synapses between the two groups (TCPS: 33.15 ± 1.43, n = 13 cells; MXenes: 33.57 ± 1.43, n = 14 cells; Fig. 8B). Next, we performed whole-cell patch-clamp recording and recorded spontaneous excitatory postsynaptic currents (EPSCs) from these neurons cultured on TCPS or MXenes. As shown in Fig. 8C, EPSCs were detected in both experimental groups, indicating that synaptic connections revealed in immunostaining are functional. Although there is no significant difference in the amplitude of ESPCs between the two groups (TCPS: 18.28 ± 1.28 pA, n = 48 cells; MXenes: 19.74 ± 1.18 pA, n = 41 cells; Fig. 8D and E), EPSCs recorded from neurons cultured on Ti3C2Tx MXene are more frequent (TCPS: 0.25 ± 0.05 Hz, n = 48 cells; MXenes: 0.50 ± 0.11 Hz, n = 41 cells; p < 0.05, Fig. 8F and G).

Fig. 8
figure 8

Ti3C2Tx MXene increased the frequency but not the amplitude of spontaneous synaptic currents in NSC-derived neurons. A and B Representative dual staining images of PSD95 and Synapsin-1 at 7 DIV for typical neuron cultured on TCPS and MXenes, Scale bar = 10 μm. (A) and the average number of synapses (B). C Typical spontaneous excitatory postsynaptic currents (EPSCs) recorded from two new neurons. D and E Average of EPSC amplitudes across different cells (D) and cumulative amplitude distribution of two new neurons (E). F and G Average of EPSC frequencies across different cells (F) and cumulative interevent interval distribution of two new neurons (G). * indicates p < 0.05

Full size image

Taken together, these results show that regenerated neurons on Ti3C2Tx MXene could make functional synapses with one another and form mature neural networks, further reinforce its safety and high biocompatibility. Furthermore, Ti3C2Tx MXene enhances synaptic transmission by selectively increasing the frequency but not the amplitude of synaptic events, affording additional advantageous for Ti3C2Tx MXene to be used in neural interface and scaffold.

Conclusion

In this study, we investigated the effects of two-dimensional Ti3C2Tx MXene film on the electrophysiological maturation of NSCs. Through the whole-cell patch-clamp recording technique, the direct effects of Ti3C2Tx MXene on NSC-derived neurons and its influence on the formation and performance of neural networks were examined. Ti3C2Tx MXene films with a thickness of several hundred nanometers were fabricated and spin-coated on TCPS, and they exhibited good biocompatibility as NSCs cultured on them are as viable as the control group. In addition, NSCs cultured on Ti3C2Tx MXene films differentiated into neurons with higher efficiency and longer neurites, highlighting their capability to promote NSCs maturation. Furthermore, close contact with Ti3C2Tx MXene bears no appreciative difference in voltage-gated Na+ or K+ current, but selectively increases the amplitude of voltage-gated Ca2+ current, which could contribute to the longer neurites observed. Finally, Ti3C2Tx MXene does not alter the passive membrane properties of matured neurons and promotes neuronal firing only for strong depolarization. Furthermore, Ti3C2Tx MXene does not change the number synapses or the amplitude of synaptic events, but it is capable to enhance synaptic transmission by increasing the frequency of synaptic releases. In summary, our results strongly suggest two-dimensional Ti3C2Tx MXene represents a new and promising direction for conductive neural interface or scaffold in stem cell therapy and nerve tissue engineering from morphology, physiology and functionality.

However, there are still limitations in our study. Firstly, we used laminin coating like other studies [7, 75, 78,79,80,81], which helps the adhesion of NSCs, but weakens the most direct influence of Ti3C2Tx MXene on NSCs. Secondly, in order to fully reveal the interaction between Ti3C2Tx MXene and neural tissue, it is necessary to accurately control its size, composition and surface functional groups and evaluate in a more systematic way its impact on functions of individual neurons and neural circuits among them. Finally, the physical interacting of MXenes with K+, Na+ and Ca2+ channels and the underlying mechanisms also need to be further studied [26].