Conductive MXene/adECM hydrogel promotes skeletal muscle regeneration and innervation through Ca²⁺ influx modulation and neuromuscular junction formation

Abstract

Background

Volumetric muscle loss (VML) leads to severe skeletal muscle dysfunction. While muscle tissue engineering offers a promising strategy, challenges persist due to insufficient neuromuscular innervation and poor reconstruction of neuromuscular junctions (NMJs). Conductive hydrogels can mimic the electrophysiological microenvironment and thus promote structural and functional regeneration, yet commonly used conductive materials still suffer from poor hydrophilicity, non-degradability, and potential cytotoxicity, while their underlying mechanisms remain unclear. Ti₃C₂T_x MXene, a class of two-dimensional nanomaterials with high conductivity and biocompatibility, shows potential for repairing electroactive tissues. In this study, we developed a novel biomimetic electroactive hydrogel by incorporating Ti₃C₂T_x MXene nanosheets into adipose-derived decellularized extracellular matrix (adECM). This study aimed to investigate the effects and mechanisms of MXene/adECM hydrogel on muscle regeneration and innervation.

Results

MXene/adECM hydrogel demonstrated excellent biocompatibility, biodegradability, and conductivity. Compared to the adECM hydrogel, the incorporation of MXene promoted myogenesis, along with increased expression of Desmin, MyoD1, and Myf5. Furthermore, the MXene/adECM hydrogel at the optimal concentration increased the average neurite length by 47.29 μm (p < 0.05) relative to the adECM group. Transcriptomic analysis combined with a neuromuscular co-culture system indicated that the MXene/adECM hydrogel promoted the formation of neuromuscular junctions (NMJs). The incorporation of MXene upregulated the expression of specific voltage-gated calcium channels at the motor endplate, with transcript levels of Cacna1a and Cacna1s increased to 2.1-fold and 3.1-fold, respectively. It was further observed that calcium signaling was enhanced in the MXene/adECM group, with the peak calcium signal intensity being 2.40 times that of the adECM group. In vivo rat VML model confirmed that, compared to the adECM hydrogel, the MXene/adECM hydrogel promoted an increase in regenerated muscle fiber area, reduced collagen deposition, and elevated the fluorescence intensity of CD31 and Tuj. The co-localization percentage of presynaptic and postsynaptic NMJ markers increased from 27.85 ± 8.69% to 42.21 ± 15.52%. Gait analysis showed significant improvements in print area, swing/stance ratio, and movement velocity. In the MXene/adECM group, the isometric tetanic force (ITF) upon sciatic nerve stimulation was significantly higher than that of the adECM group (0.082 ± 0.012 N vs. 0.057 ± 0.014 N, p < 0.05), approaching the level of the uninjured group.

Conclusion

Together, these findings demonstrate that the incorporation of MXenes into adECM provides a promising strategy that integrates microenvironmental support with endogenous electrical cues to modulate calcium influx and promote NMJ formation, offering a new paradigm for the treatment of VML.

Graphical Abstract

Study on the therapeutic strategy and underlying mechanism of conductive MXene/adECM hydrogel for volumetric muscle loss (VML) repair. Ti3C2Tx MXene nanosheets were integrated into an adipose-derived decellularized extracellular matrix (adECM) to fabricate a conductive composite hydrogel. Upon implantation, the MXene/adECM hydrogel promoted structural regeneration and reinnervation of the injured skeletal muscle. The bioactive microenvironment and endogenous electrical conductivity facilitated neuromuscular junction (NMJ) formation, accompanied by enhanced calcium influx

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04077-y.

Introduction

Volumetric muscle loss (VML) refers to extensive muscle damage (> 20% of total muscle mass) caused by traumatic injuries, combat wounds, tumor resection, or fractures. This condition exceeds the intrinsic regenerative capacity of skeletal muscle and is often accompanied by damage to motor nerve terminals, resulting in severe dysfunction [1, 2]. While the precise global incidence of VML is specifically limited due to clinical classification challenges, some evidence suggests that the societal burden associated with severe muscle trauma is progressively increasing. Global data shows that the overall burden of musculoskeletal disorders (under which VML is included) increased by 30.7% from 1990 to 2019 [3, 4]. It is estimated that the annual direct medical costs associated with VML exceed 176 billion dollars, with an additional 326 billion dollars annually in productivity losses, posing a significant economic burden on society [5]. Concurrently, the world’s rapidly aging population faces complex challenges. In geriatric population, weakened muscle strength and high rates of comorbidities like diabetes lead to more severe musculoskeletal injuries [6]. Trauma statistics indicate that road traffic accidents have become the second leading cause of trauma in the elderly, resulting in severe limb trauma and leading to a high proportion of permanent disability (73.4%) [6]. Collectively, these factors underscore the urgent and growing need for effective VML management strategies. Currently, composite tissue flap transplantation is the standard clinical approach for VML repair. However, due to limited donor muscle volume and a high risk of postoperative complications, functional recovery in the affected limbs remains suboptimal [7]. Hence, there is an urgent need to develop alternative therapies for VML, which remains a major challenge in the field of regenerative medicine.

Muscle tissue engineering (MTE) has recently demonstrated significant potential for VML treatment [5, 8]. Studies have demonstrated that engineered constructs can promote muscle regeneration, as evidenced by restored muscle volume and increased muscle fiber density [9]. However, functional recovery remains a major challenge [10]. Insufficient reinnervation of the regenerated muscle is a core bottleneck limiting functional recovery. Skeletal muscle is a dynamic organ composed of contractile fibers, vasculature, connective tissue, and nerve terminals [11]. Regeneration of damaged muscle requires a sequential process involving inflammation, myoblast proliferation and differentiation, muscle fiber formation and maturation, and restoration of neuromuscular junction (NMJ) [12]. However, during VML repair, the loss of regenerative cues and disruption of surrounding connective tissue trigger excessive collagen deposition, fibrotic scar formation, and compromised innervation [13]. Furthermore, inadequate physicochemical properties of scaffold materials can hinder host cell and nerve infiltration, further limiting functional recovery [14]. Therefore, achieving structural and functional muscle regeneration should be the primary goal in VML therapy, with the restoration of innervation to the regenerated muscle being the critical factor.

The physicochemical properties of the extracellular microenvironment play a crucial regulatory role in cell growth and function [15]. As an intrinsically electroactive tissue, nerve regeneration requires an appropriate electrophysiological microenvironment to promote axonal growth and functional recovery [16]. Researchers have incorporated conductive materials such as graphene, black phosphorus, and carbon nanotubes into biological scaffolds to provide intrinsic electrical fields, thereby promoting peripheral nerve regeneration [17]. Nevertheless, limitations such as poor hydrophilicity, non-degradability, and potential cytotoxicity associated with these conductive materials have prompted the need for further optimization of conductive biomaterials to achieve more efficient and safe neural functional restoration [18]. For MTE, carbon-based nanomaterials (e.g., graphene, carbon nanotubes) offer superior conductivity and mechanical reinforcement beneficial for muscle regeneration [19, 20]. Patel et al. demonstrated that carbon nanotube-based hierarchical scaffolds can promote the formation of mature myotubes [20]. However, poor hydrophilicity and propensity for aggregation hinder their practical application [18, 21]. Recently, MXenes have emerged as a novel class of two-dimensional (2D) nanosheet materials. Their rich surface functional groups, high conductivity, good hydrophilicity, biocompatibility, and antimicrobial properties have garnered increasing attention in tissue engineering and regenerative medicine [22–24]. Among them, Ti₃C₂T_x MXene possesses a trilayer Ti atomic structure and abundant -OH/-O surface groups, endowing it with superior conductivity and hydrophilicity [25]. These characteristics make it highly promising for the repair and regeneration of bioelectrically active tissues. Recent studies have shown that Ti₃C₂T_x MXene-based nanocomposites can promote nerve growth through their inherent electroactivity or in combination with electrical stimulation, supporting their application in neural tissue engineering [26–28]. However, the biological effects of Ti₃C₂T_x MXene-based materials on muscle reinnervation need further exploration.

On the other hand, although electroactive biomaterials have shown great potential in skeletal muscle regeneration, most current research still focuses on structural recovery [18, 29, 30]. The mechanism of muscle reinnervation still lacks in-depth study. Calcium ions (Ca²⁺), as an important second messenger in cells, play a critical role in the development, function, and regeneration of the neuromuscular system [31]. Previous studies have demonstrated that Ca²⁺ influx contributes to triggering acetylcholine release, activating postsynaptic signaling pathways, and regulating gene expression during the formation and signal transmission of NMJ [32, 33]. Rotzler et al. [34] revealed that the activity-dependent Ca²⁺ influx in the postsynaptic membrane of skeletal muscle regulates the metabolic stability of nAChR. Furthermore, Kaplan et al. [33] found that the nAChR cluster morphology was abnormal in calcium channel CaV1.1 gene knockout mice, severely affecting the establishment of NMJ. Meanwhile, conductive biomaterials can enhance intercellular electrical signal transmission and regulate ion channel activity by simulating endogenous electric fields [35]. A previous study has demonstrated that MXenes elevated intracellular Ca²⁺ levels in myoblast cells. The increase in Ca²⁺ signaling upregulated the iNOS and mTOR-AKT pathways to promote cell proliferation, survival, and myogenic differentiation [36]. However, its effect and mechanism on muscle reinnervation are still unclear. The activation and functional expression of voltage-gated calcium channels (VGCCs) are closely related to the extracellular electrical environment and may represent key targets for the biological effects of MXene-based conductive materials [37].

Currently employed scaffolding approaches in muscle tissue engineering, such as electrospinning and 3D printing, can provide tunable mechanical stiffness and topographical cues to guide the regeneration of highly aligned skeletal muscle [38]. However, electrospun scaffolds are known to have poor cell infiltration and suboptimal cellular adhesion [39, 40]. Limitations, including low printing speed and resolution, associated with 3D printing must be addressed to realize the full potential of large, 3D-printed bioconstructs [41]. In contrast, hydrogels offer the unique advantage of providing a highly hydrated microenvironment that closely mimics the natural cellular niche [42]. Injectable hydrogels can serve as fillers for irregular defects and also function as bioinks for 3D bioprinting [43]. Hydrogels are ideal carriers for conductive materials due to their ease of modification and integration [44]. Compared to synthetic polymers, natural materials such as adipose-derived decellularized extracellular matrix (adECM) offer better biocompatibility and bioactivity [45]. Our previous work showed that adECM supports muscle regeneration by providing a natural 3D microenvironment [46]. Therefore, in this study, we incorporated Ti₃C₂T_x MXene nanosheets into adECM to develop a novel biomimetic electroactive hydrogel (MXene/adECM). This hydrogel exhibited favorable physicochemical properties, including anti-swelling capacity, degradability, conductivity matching the physiological environment, and excellent biocompatibility. In vitro experiments demonstrated that the MXene/adECM hydrogel promoted myogenic differentiation and neurite outgrowth. In a neuromuscular co-culture system, the MXene/adECM hydrogel facilitated the directional alignment of nerve axons along myotubes and enhanced neuron–muscle connectivity. Transcriptomic results suggested that the MXene/adECM hydrogel upregulated the expression of Ca²⁺ ion channels and promoted NMJ formation. The experiment verified these results and observed the significantly enhanced Ca²⁺ influx signals. In vivo, MXene/adECM hydrogel significantly enhanced motor nerve conduction and muscle innervation, while concurrently promoting the structural and volumetric recovery of skeletal muscle. Collectively, we propose that the MXene/adECM hydrogel promotes structural and functional muscle regeneration following VML through modulating Ca²⁺ influx and enhancing NMJ formation.

Materials and methods

Materials

Sodium dodecyl sulfate (SDS, CH₃(CH₂)₁₁OSO₃Na), Triton X-100, and phosphate-buffered saline (PBS) were obtained from Solarbio (Beijing, China). Pepsin (P7125) was purchased from Sigma-Aldrich (USA). Sodium hydroxide (NaOH) and hydrochloric acid (HCl) were obtained from Macklin Biochemical Co., Ltd. (Shanghai, China). Isopropanol was purchased from Acmec Biochemical Co., Ltd. (Shanghai, China), and hydrogen peroxide (H₂O₂) was obtained from LABSELECT (China). Ti₃AlC₂ was acquired from 11 Technology Co., Ltd. (China), and lithium fluoride (LiF) was purchased from Aladdin Chemical Co., Ltd. (Shanghai, China). Unless otherwise specified, all reagents were used as received without further purification. All solutions were prepared using double-distilled water (ddH₂O).

Preparation of adECM

adECM was prepared from inguinal fat pads harvested from adult Sprague-Dawley rats. Following thorough rinsing with PBS, the adipose tissue was minced and lyophilized for 48 h. Decellularization was performed by incubating the dried tissue in a solution containing 0.5 wt% SDS and 1 wt% Triton X-100 (1:1, v/v) at 37 °C on a shaker (250 rpm) for 12 h. The tissue was then washed with ddH2O and subjected to delipidation using isopropanol under the same shaking conditions for 24 h. The tissue was then washed by ddH2O and 75% ethanol, followed by a second lyophilization for 48 h to obtain dry adECM. Subsequently, 15 mg of lyophilized adECM was enzymatically digested with 15 mg of pepsin in 1 mL of 0.1 M HCl, and stirred on a magnetic stirrer for 48 h until a homogenous solution without visible particulates was obtained. The pH of the solution was then adjusted to 7.3–7.4 using NaOH, and 10% (v/v) of 10× PBS was added to restore ionic balance. The final adECM solution was stored at 4 °C until further use.

Characterization of adECM

To evaluate the efficacy of decellularization and delipidation, native adipose tissue and processed adECM samples were embedded, sectioned, and subjected to histological staining, including hematoxylin and eosin (H&E), Masson, and Oil Red O staining, following standard protocols. For immunohistochemical (IHC) staining, tissue sections underwent antigen retrieval followed by incubation with 3% hydrogen peroxide (H₂O₂) for 30 min at room temperature to block endogenous peroxidase activity. Nonspecific binding was blocked using 5% goat serum. Primary antibodies against collagen I (#ab270993, Abcam, USA) and collagen IV (#ab236640, Abcam, USA) were diluted 1:1000 and incubated overnight at 4 °C. The next day, sections were treated with secondary antibodies for 30 min, followed by streptavidin–HRP complex incubation. Hematoxylin (SA00013-4, Proteintech, USA) was used for nuclear counterstaining. Residual DNA content in the tissue was quantified using the E.Z.N.A.^® Tissue DNA Kit (Omega, USA) according to the manufacturer’s instructions (n = 3).

Synthesis of Ti₃C₂T_x MXene nanosheets

Ti₃C₂T_x MXene nanosheets were synthesized following a previously established protocol [47]. Briefly, 1 g of LiF was added to 20 mL of 9 M HCl and stirred for 5 min. Subsequently, 1 g of Ti₃AlC₂ was added to the etching solution and stirred at 35 °C for 24 h to selectively remove the Al layer. The resulting acidic suspension was washed with deionized (DI) water and centrifuged at 3500 rpm for 5 min repeatedly until the supernatant reached a pH of 6. The precipitate was then redispersed in deionized water and subjected to 20 min of ultrasonication. Finally, the dispersion was centrifuged at 3500 rpm for 1 h to obtain the supernatant containing Ti₃C₂T_x nanosheets. The morphology of the MXene nanosheets was characterized by scanning electron microscopy (SEM, S-4700, Hitachi, Japan). Elemental composition was confirmed using energy-dispersive X-ray spectroscopy (EDS, Hitachi S-4800, Japan). Atomic force microscopy (AFM, Bruker, USA) was employed to characterize the morphology and thickness of monolayer MXene nanosheets. AFM measurements were performed using a Bruker Dimension FastScan system in ScanAsyst mode with a SCANASYST-AIR probe.

Formation and characterization of MXene/adECM hydrogels

The obtained Ti₃C₂T_x MXene nanosheets were dispersed in adECM solutions at final concentrations of 100, 300, and 500 µg mL⁻¹. The mixtures were ultrasonicated at 4 °C for 10 min to ensure uniform dispersion, resulting in MXene/adECM pre-gels. These were then incubated at 37 °C for 30 min to form physically crosslinked MXene/adECM hydrogels.

SEM analysis:

The internal microstructures of the hydrogels were visualized using SEM. Hydrogels were rapidly frozen in liquid nitrogen and subsequently freeze-dried for 24 h. The dried samples were then mounted on stubs, sputter-coated with gold, and observed under a scanning electron microscope (S-4800, Hitachi, Japan). The pore sizes of the hydrogels were measured from SEM images using ImageJ software.

Rheological test:

Rheological properties were measured using a rotational rheometer (Physica MCR 302e, Anton Paar, Austria) with a parallel plate (50 mm diameter, 1 mm gap). The storage modulus (G′) and loss modulus (G″) were measured at 37 °C and 1 Hz.

Compressive modulus:

The biomechanical properties were tested using a dynamic mechanical analyzer (E3000, UK). The samples were compressed with 0.1 N load. The stress-strain curve was drawn, and the compression modulus of each sample was calculated.

Swelling ratio:

The swelling ratio was determined by measuring the weight change of the hydrogel before and after equilibrium swelling in PBS. W₀ and W_s are the weights of the dry and swollen hydrogel, respectively. The swelling ratio (%) was calculated using the following equation.

Biodegradation evaluation:

Hydrogels were incubated in PBS at 37 °C. At intervals of every 3 days over a 15-day period, samples were collected and weighed (mt). The initial dry weight (m0) was recorded prior to incubation. The degradation rate (%) was calculated using the following equation.

A degradation profile was plotted to compare the degradation behaviors of adECM and MXene/adECM hydrogels over time.

Conductivity evaluation:

The electrical conductivity of hydrogels was measured using a four-point probe method (HPS2661, Helpass electronic technologies, China). Three independent samples were analyzed per group. The average conductivity was expressed in mS m⁻¹.

Cell culture and differentiation induction

L6 rat skeletal myoblasts were obtained from the Center of Basic Medical Research, Peking University Third Hospital (Beijing, China). Cells were cultured in high-glucose Dulbecco’s Modified Eagle’s Medium (DMEM; HyClone Laboratories, Logan, UT, USA) supplemented with 10% fetal bovine serum (FBS; TransSerum^® FQ, China) and 1% penicillin-streptomycin (HyClone, USA) at 37 °C in a humidified incubator with 5% CO2. For myogenic differentiation, L6 cells were switched to myogenic induction medium composed of DMEM supplemented with 2% horse serum (Invitrogen, USA) and 1% penicillin-streptomycin.

PC12 cells (CL-0480) and the growth medium (CM-0480) were purchased from Wuhan Pricella Biotechnology Co., Ltd (Wuhan, China). Cells were cultured at 37 °C in a humidified incubator with 5% CO2. To induce neuronal differentiation, PC12 cells were treated with DMEM supplemented with 1% horse serum and 100 ng mL⁻¹ recombinant human beta-nerve growth factor (β-NGF; 450-01, PeproTech, USA).

Cytocompatibility assay

To evaluate the cytocompatibility of the hydrogels, extract-based assays were conducted. Preformed hydrogels were immersed in growth medium at 37 °C for 72 h to obtain conditioned medium (CM). Cells were seeded into 96-well plates at a density of 5000 per well and incubated overnight at 37 °C with 5% CO2 in a humidified incubator. After attachment, the culture medium was replaced with CM, and cell viability was assessed at 24 and 48 h using the Cell Counting Kit-8 (CCK-8; KeyGEN, China). Briefly, 10% (v/v) CCK-8 reagent was added to each well and incubated at 37 °C for 1 h. Absorbance was measured at 450 nm using a microplate reader (Thermo Scientific, USA).

Distribution and proliferation of L6 cells in hydrogels

L6 cells were suspended in pre-gel solutions of adECM and MXene/adECM at a density of 8 × 10⁵ cells mL^− 1. A total volume of 300 µL of the cell-laden pre-gel solution was added to confocal dishes (20 mm in diameter), followed by incubation at 37 °C for 30 min to allow complete gelation. Subsequently, 1 mL L6 cell growth medium was added to the hydrogels. On days 1, 3, and 7 of culture, cell viability and spatial distribution were assessed by live/dead staining using a Calcein-AM/PI kit (KeyGEN, China) according to the manufacturer’s instructions. Samples were scanned layer-by-layer using a confocal laser scanning microscope (TCS-SP8 STED, Leica, Germany). All images were acquired and processed in LAS X software (Leica, Germany) to reconstruct 3D view graphs and visualize the internal distribution of L6 cells within the hydrogels. ImageJ software was used to quantify the proliferation from confocal images.

Evaluation of myogenic differentiation in L6 cells

L6 cells were suspended in pre-gel solutions of adECM and MXene/adECM hydrogels at a density of 8 × 10⁶ cells mL⁻¹. A total of 300 µL of the cell-laden pre-gel solution was added into confocal dishes (14 mm in diameter) and incubated at 37 °C for 30 min to achieve complete gelation. Subsequently, myogenic induction medium was added, and the constructs were cultured for 7 days to induce myogenic differentiation.

For immunofluorescence (IF) staining, the samples were fixed and incubated with a mouse monoclonal antibody against myosin heavy chain (MHC, 1:1000; ab37484, Abcam, UK), followed by incubation with a goat anti-mouse secondary antibody conjugated to Alexa Fluor 594 (1:200; ZSGB-BIO, China) for 2 h. Cell nuclei were counterstained with DAPI (BestBio, China). The expression of MHC was visualized using a confocal laser scanning microscope (TCS-SP8 STED, Leica, Germany). Imaging was performed using a 25x water-immersion objective, with the imaging depth confined to within 100 μm of the hydrogel surface layer.

For Western blotting analysis, 1 mL of the cell-laden hydrogels was added into each well of a six-well plate and cultured under myogenic induction medium for 7 days. The samples were lysed using RIPA buffer supplemented with protease inhibitor cocktail (P8340, Sigma, USA). Equal volumes (20 µL) of each lysate were loaded onto SDS-PAGE gels, separated electrophoretically, and transferred onto polyvinylidene fluoride (PVDF) membranes. After blocking with 5% non-fat milk for 1 h, the membranes were incubated overnight at 4 °C with primary antibodies against Desmin (1:100000; ab32362, Abcam, UK), Myod1 (1:1000; 18943-1-AP, Proteintech, China), and Myf5 (1:10000; ab125078, Abcam, UK). The membranes were then incubated with HRP-conjugated goat anti-rabbit secondary antibodies (ab99697, Abcam, UK) for 1 h at room temperature. Protein bands were visualized using an ECL chemiluminescence detection kit (GE Healthcare, Chicago, USA) and imaged with the TANON 5200 Multi system (Tanon, Beijing, China).

Evaluation of growth and neuronal differentiation of PC12 cells

The hydrogels were incubated in neuronal differentiation medium at 37 °C for 72 h to obtain the conditioned medium. PC12 cells were seeded at a density of 3,000 cells per well in 96-well plates and allowed to adhere overnight in growth medium. The next day, the medium was replaced with the conditioned neuronal differentiation medium. Cells were cultured under differentiation conditions for 72 h, after which the medium was refreshed. Time-lapse imaging was then performed continuously for an additional 72 h using the Incucyte S3 live-cell analysis system (Sartorius, Germany). Neurite length and neurite branch points were analyzed using the Incucyte^® NeuroTrack Analysis Software Module (Sartorius, Germany).

To evaluate the neuronal differentiation of PC12 cells on the hydrogels, IF staining for the neuronal marker β-III tubulin (Tuj1) was performed. PC12 cells were seeded onto the hydrogel at a density of 1.0 × 10³ cells cm^− 2. After 5 days of induction in neuronal differentiation medium, the samples were fixed and incubated with a primary antibody against Tuj1 (1:1000; ab18207, Abcam, UK), followed by a goat anti-rabbit secondary antibody conjugated to Alexa Fluor 488 (1:200; Jackson ImmunoResearch, USA). Cell nuclei were counterstained with DAPI (BestBio, China). Fluorescence images were acquired using a confocal laser scanning microscope (TCS-SP8 STED, Leica, Germany).

Neuromuscular co-culture system

The neuromuscular co-culture system was established based on previously reported protocols with slight modifications [48]. Briefly, L6 cells were first seeded onto the hydrogel at a density of 1.0 × 10⁴ cells cm^{− 2} and cultured in myogenic induction medium for 7 days. To inhibit further proliferation of undifferentiated myoblasts, cytosine β-D-arabinofuranoside (Ara-C, 0.5 µg mL^{− 1}; C1768, Sigma-Aldrich, USA) was added to the culture medium for 3 days. Subsequently, PC12 cells were seeded onto the same hydrogel at a density of 1.0 × 10³ cells cm^{− 2} and cultured in neuronal differentiation medium for an additional 7 days. After the completion of co-culture, samples were subjected to double immunofluorescence staining for MHC and Tuj1. The primary and secondary antibodies, DAPI nuclear staining, and image acquisition were performed following the same procedures as described above.

RNA-sequencing analysis

Total RNA was extracted using the TRIzol reagent (Invitrogen, CA, USA) according to the manufacturer’s protocol. RNA purity and quantification were evaluated using the NanoDrop 2000 spectrophotometer (Thermo Scientific, USA). RNA integrity was assessed using the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). Then the libraries were constructed using VAHTS Universal V10 RNA-seq Library Prep Kit (Premixed Version) according to the manufacturer’s instructions. The transcriptome sequencing and analysis were conducted by OE Biotech Co., Ltd. (Shanghai, China). The libraries were sequenced on an Illumina Novaseq 6000 platform and 150 bp paired-end reads were generated. Raw reads of fastq format were firstly processed using fastp and the low quality reads were removed to obtain the clean reads. The clean reads were mapped to the reference genome using HISAT2. FPKM of each gene was calculated and the read counts of each gene were obtained by HTSeq-count. Principal component analysis (PCA) was performed using R (v 3.2.0) to evaluate the biological duplication of samples. Differential expression analysis was performed using the DESeq2. Q value < 0.05 and fold change > 2 or fold change < 0.5 were set as the threshold for significantly differentially expressed genes (DEGs). Hierarchical cluster analysis of DEGs was performed using R (v 3.2.0) to demonstrate the expression pattern of genes in different groups and samples. Based on the hypergeometric distribution, GO and KEGG pathway enrichment analysis of DEGs were performed to screen the significantly enriched terms using R (v 3.2.0), respectively. R (v 3.2.0) was used to draw the chord diagram.

Real-time quantitative PCR (RT-qPCR)

L6 and PC12 cells in neuromuscular co-culture system were used to extract total RNA using TRIzol reagent (ThermoFisher Scientific, USA). RNA was reverse-transcribed to cDNA and 2 X qPCR Master Mix (Tiangen, China) was used for RT-qPCR (for primers, see Table S1). ΔΔCt method was used for calculating expression.

Viral transduction of PC12 cells

Green fluorescent protein (GFP)-expressing lentiviral vectors were obtained from a commercial provider (Genechem, Shanghai, China). PC12 cells were seeded in 6-well plates and infected with the lentivirus at a multiplicity of infection (MOI) of 10. After 72 h of incubation, GFP expression was assessed using confocal laser scanning microscopy. Cells exhibiting stable GFP fluorescence were subsequently selected using puromycin to obtain a purified GFP⁺ PC12 population.

Calcium imaging

Calcium influx was assessed using the red fluorescent calcium indicator Cal-630^™ AM (AAT Bioquest, USA). The indicator was diluted to a final concentration of 10 µM in HHBS buffer (CB1048, Yuabio, China) supplemented with 0.04% Pluronic^® F-127 (20053, AAT Bioquest, USA) to prepare the working solution. After washing the cells with HHBS, the Cal-630^™ AM working solution was added and incubated for 60 min at room temperature. Following incubation, cells were washed three times with fresh HHBS buffer. To stimulate calcium influx, 30 mM KCl dissolved in HHBS was applied directly to the culture dish. Time-lapse calcium imaging was performed using a confocal laser scanning microscope (TCS-SP8, Leica, Germany), with images acquired every 5 s for a total duration of 300 s. Fluorescence intensity within regions of interest (ROIs) and calcium signal variations were analyzed using LAS X software (Leica, Germany). The experiments were independently replicated 3 times per group with 3 ROIs. The baseline (F₀) was automatically calculated from background fluorescence by LAS X.

VML injury model

Healthy Sprague Dawley male rats (Vital Rive, Beijing, China) aged 8 weeks were used to establish the tibialis anterior (TA) muscle VML model. All rats were bred in a pathogen-free environment with a 12-h light/dark cycle. All experiments involving animals were conducted according to the ethical policies and procedures approved by the ethics committee of Peking University Third Hospital (No. A2022068) and were performed according to the National Institutes of Health regulations for the care and use of animals. The VML injury model was established according to the previous research [49]. Animals were anesthetized with isoflurane, and the skin and fascia of the right hindlimb were incised to expose the TA muscle. A standardized defect (approximately 7 × 10 × 3 mm³) was created in the middle portion of the TA muscle using a surgical scalpel. The excised muscle was weighed and recorded. Immediately after injury, the defect sites were treated according to group allocations.

A total of 24 rats were randomly assigned to four groups according to the random number table (n = 6 per group): (1) Uninjured, no TA defect created; (2) Untreated, TA defect without hydrogel implantation; (3) adECM, defect filled with 300 µL adECM hydrogel loaded with L6 cells (8 × 10⁶ cells mL^− 1); (4) MXene/adECM, defect filled with 300 µL MXene/adECM hydrogel (MXene: 300 µg mL^− 1) loaded with L6 cells (8 × 10⁶ cells mL^− 1). The sample size was estimated based on the degrees of freedom. No animals were excluded. All hydrogels were freshly prepared and added into the defect area under sterile conditions. Body weight was recorded on the day of surgery and subsequently on postoperative days 7, 14, 21, and 30 to monitor general health and recovery. To evaluate the time-dependent tissue response and regeneration, TA muscles were harvested at three time points. On day 7 and day 14, one rat from each group was euthanized for gross morphological observation and H&E staining. At the final endpoint (day 30), the remaining four rats in each group were subjected to functional assessments, including gait analysis and isometric tetanic force evaluation. Subsequently, the animals were sacrificed for imaging evaluation and comprehensive histological analysis. All histological evaluations and functional tests were conducted in a blinded manner. After data acquisition, further image processing and quantitative analysis were performed by another researcher who was not involved in the initial assessments.

The theoretical weight of the TA muscle was estimated based on body weight using the following formula [50]:

On the day of surgery, the initial TA weight (TA₀) was calculated, and the weight of the resected muscle fragment was measured as W₁. The proportion of muscle removed was defined as W₁/TA₀. The residual TA weight after defect creation was defined as TA₁. At day 30 post-surgery, the regenerated TA muscle was harvested and weighed as TA₂. Meanwhile, the theoretical increase in TA mass due to natural growth over 30 days (TA₃) was estimated based on body weight gain between day 0 and day 30 using the same formula. The improved regenerative muscle weight was calculated as:

Histologic evaluation

Rats were euthanized via carbon dioxide inhalation. The injured region of the TA muscle was harvested and processed into paraffin-embedded sections using standard histological procedures. Sections were subjected to H&E staining, Masson staining, and IF staining. For IF staining, tissue sections were incubated overnight at 4 °C with the following primary polyclonal antibodies: MHC (ab37484, Abcam, UK), laminin (PA1-16730, Thermo Fisher Scientific, USA), CD31 (ab182981, Abcam, UK), synaptic vesicle glycoprotein 2 (SV2; AB_2315387, DSHB, USA), α-BTX (90–1021, BTprobes, China), and AChE (RM2953, Biodragon, China). The sections were then incubated with species-appropriate secondary antibodies conjugated to Alexa Fluor 488 (ab150077, Abcam, UK) or Alexa Fluor 647 (ab150115, Abcam, UK), followed by nuclear counterstaining with DAPI. Quantitative analysis of stained sections was performed using ImageJ software. Parameters measured included the number, diameter, and cross-sectional area of muscle fibers within the injured site, collagen volume fraction, and fluorescence area and intensity for relevant markers.

Inductively coupled plasma mass spectrometry (ICP-MS)

Titanium (Ti) content in tissues was measured by Inductively Coupled Plasma Mass Spectrometry (ICP-MS, 7800, Agilent Technologies Inc., USA). The detection was surpported byBeijing Zhongkebaice Technology Co., Ltd.

Imaging evaluation

Imaging protocols were conducted in accordance with our previously established procedures [46]. At 30 days post-surgery, magnetic resonance imaging (MRI) and ultrasonographic assessments were performed to evaluate structural and volume changes in the TA muscles. For ultrasound imaging, both longitudinal and transverse planes of the TA region were acquired using a high-resolution photoacoustic imaging system (Vevo F2 LAZR-X, FUJIFILM VisualSonics, USA) equipped with a 30 MHz linear transducer. Rats were anesthetized, and the hair overlying the hindlimb was removed using depilatory cream. Ultrasonic gel was applied to the exposed region, and B-mode images were captured. MRI was conducted using a 3.0 T whole-body scanner (Discovery MR750, GE Healthcare, USA). Anesthesia was maintained with 2.5% isoflurane delivered in 100% medical-grade oxygen. To ensure consistent positioning, a preliminary scout scan of ten 2 mm slices was used to define the hindlimb orientation. Subsequently, 15 axial slices covering the region from the knee to the ankle were collected, with the field of view encompassing both hindlimbs for comparative analysis.

Gait analysis

A pressure-sensing walkway (Tekscan, USA) measuring 5 inches wide by 14 inches long was used to monitor paw pressure and impulse with a paw strike resolution of 15.5 sensels cm^− 2. The walkway was first calibrated with known weights which occupied a similar number of sensors as a rat paw, and the weight of each animal in Newtons was calculated and entered into the Tekscan operating system software (version 7.0, Tekscan, USA) prior to data collection. Data files consisted of stress diagrams, print area, swing duration, stance duration, stride length and speed of each rat.

Isometric tetanic force (ITF) evaluation

Under isoflurane anesthesia, the sciatic nerve was carefully exposed via a posterior gluteal incision in the hindlimb. A second incision was made anterior to the ankle and foot to expose the distal TA muscle-tendon unit. The femur and tibia were stabilized on a custom platform to ensure consistent limb positioning during testing. The distal TA tendon was isolated and securely tied to a force transducer using surgical sutures. A bipolar stimulating electrode (CJ-3, Pclab, China) was gently positioned beneath the sciatic nerve, with continuous application of saline to maintain tissue hydration and prevent thermal or mechanical injury. Isometric tetanic contractions were evoked by applying electrical stimulation to the sciatic nerve at parameters of 0.2 V amplitude, 100 Hz frequency, and 0.1 ms pulse duration. A biomedical signal acquisition system (Pclab-530 C, Pclab, China) was employed to deliver stimulation pulses and record force responses in real time.

Statistical analysis

GraphPad Prism 9.4.1 was used for statistical analysis. All statistical data were expressed as mean (± SD). Comparisons of two groups were done with t-test and comparisons of multiple groups were done with one way ANOVA. Each group contains at least three independent biological replicates. P-values less than 0.05 were considered statistically significant.

Results and discussion

Preparation and property characterization of MXene/adECM hydrogels

The MXene/adECM hydrogel was fabricated by incorporating Ti₃C₂T_x MXene nanosheets into an adECM pre-gel solution, followed by incubation at 37 °C to induce gelation. The adECM was derived from the decellularized and defatted inguinal fat pads of rats (Fig. 1a). Histological staining, including H&E and DAPI (Fig. 1b, c), as well as quantitative DNA analysis (Fig. 1i), confirmed the effective removal of cellular components. Oil Red O and Masson staining demonstrated thorough lipid removal and preservation of the native extracellular matrix architecture (Fig. 1d, e). Furthermore, immunohistochemistry staining and Sirius red staining were used to characterize the collagen types and distribution characteristics of the extracellular matrix. As shown in Fig. 1f, g type I and type IV collagen are well preserved in adECM, presenting a loose mesh structure. Polarized light images of Sirius red staining showed the presence of type III collagen fibers in both native adipose tissue and adECM, presenting soft green and green-yellow colors. Bright cord structures were observed, representing type I collagen (Fig. 1h). Structurally, adECM is less regular compared to native adipose tissue, with collagen fibers partially densely stacked, possibly due to the removal of cellular and lipid components. The resulting adECM retained a complex composition of structural proteins, glycosaminoglycans, and tissue-specific growth factors including VEGF, FGF, and IGF-1, which are known to support cell adhesion, proliferation, and myogenic differentiation [51, 52]. Compared to synthetic polymer-based hydrogels, adECM provides a more biomimetic microenvironment, improved bioactivity, and lower immunogenicity, making it a superior platform for skeletal muscle tissue engineering.

Fig. 1.

Fabrication and characterization of adECM. a) Schematic illustration of adECM preparation process from rat inguinal fat pads. b-e) H&E staining, DAPI staining, Oil Red O staining, Masson staining of native adipose tissue and adECM. Scale bar: 200 μm. f, g) Immunohistochemical staining of type I and type IV collagen. Scale bar: 200 μm. h) Sirius red staining under polarized light images of native adipose tissue and adECM. Scale bar: 200 μm. i) DNA content in native adipose tissue and adECM (n = 3). *** p < 0.001

Skeletal muscle-derived dECM (smdECM) is considered an ideal solution for skeletal muscle tissue engineering and the most widely used scaffold for VML applications [53]. However, a major challenge for the clinical use of smdECM is source availability [54]. In contrast, adipose tissue is an abundant and easily accessible autologous source that provides rich ECM [55]. This makes it a more viable candidate for clinical translation, particularly for repairing large defects where muscle-derived scaffolds would be insufficient in volume [56].

MXene nanosheets were synthesized by selectively etching the aluminum (Al) layer from the Ti₃AlC₂ MAX phase using in situ–generated hydrofluoric acid from a LiF/HCl mixture. Scanning electron microscopy (SEM) revealed the Ti₃AlC₂ MAX phase (Fig. S1a) and the characteristic accordion-like multilayered structure of the unexfoliated Ti₃C₂T_x (Fig. S1b). Subsequent ultrasonication yielded exfoliated ultrathin MXene nanosheets. SEM confirmed these nanosheets as monolayer flakes (Fig. S1c). The successful synthesis of MXene was further 1confirmed by energy-dispersive X-ray spectroscopy (EDS), which revealed elemental peaks corresponding to Ti, C, O, and F, consistent with the expected composition of Ti₃C₂T_x (Fig. S1d). To quantify the exfoliation outcome, atomic force microscopy (AFM) was performed on the nanosheets deposited on a silicon substrate (Fig. S2a). Statistical analysis indicated that the lateral size of the exfoliated Ti₃C₂T_x nanosheets was 181.90 ± 38.00 nm, while the thickness was 2.998 ± 0.088 nm (Fig. S2b, c). Although the theoretical thickness of Ti₃C₂T_x is ≈ 1 nm [57], the measured thickness often overestimated due to anomalies related to the measuring conditions of the AFM, instrument offset, and other artifacts [58, 59]. The measured thickness in this study corresponds to the monolayer thickness of Ti₃C₂T_x MXene reported in the literature, confirming successful exfoliation into ultrathin sheets [60].

After successful synthesis of Ti₃C₂T_x MXene nanosheets, we incorporated them into the adECM pre-gel solution at varying concentrations (100, 300, and 500 µg mL^{− 1}, denoted as AM100, AM300, and AM500, respectively). A 10-min ultrasonication facilitated uniform dispersion of the nanosheets, followed by incubation at 37 °C to induce gelation (Fig. 2a, b). As the MXene concentration increased, the color of the hydrogel deepened. SEM revealed that adECM hydrogel exhibited a uniformly crosslinked, three-dimensional (3D) porous structure (Fig. 2c). Notably, incorporation of MXene preserved the porous structure but led to reduction in pore size. Quantitative analysis showed that the average pore size decreased from 10.64 μm in the adECM group to 7.82–8.59 μm in MXene/adECM hydrogels (Fig. 2d, Fig. S3). The freezing temperature before lyophilization is a relevant factor for the pore size of hydrogel scaffolds, as governed by the ice-templating principle [61]. The high thermal conductivity of MXene may have contributed to a finer pore structure by influencing ice crystal nucleation during quenching [62]. This reduction may also result from enhanced interactions between the surface functional groups of MXene and components of the adECM, which promote the formation of a more compact network. As reported by Dey et al. [63], a lower MXene content compromises the integrity of the hydrogel network, leading to a porous structure with increased pore size, which is consistent with our observations.

Fig. 2.

Preparation and characterization of MXene/adECM hydrogels. (a) Schematic illustration of Ti₃C₂T_x synthesis and MXene/adECM hydrogels preparation. (b) Photographs of adECM pre-gel solution and the gel state of adECM and MXene/adECM. AM100, AM300, and AM500 respectively refer to 100 µg mL^− 1, 300 µg mL^− 1, and 500 µg mL^− 1 MXene/adECM hydrogel. (c) Scanning electron microscopy (SEM) images of adECM and MXene/adECM hydrogels. Scale bar: 10 μm. (d) Assessment of pore size distribution in adECM and MXene/adECM hydrogels. (e) Storage modulus (G′) and loss modulus (G″) of adECM and MXene/adECM hydrogels at 37 °C. (f) Compressive modulus of adECM and MXene/adECM hydrogels (n = 3). (g) Swelling ratio of adECM and MXene/adECM hydrogels in PBS solution (n = 3). (h) The degradation rate of adECM and MXene/adECM hydrogels (n = 3). (i) The conductivity of adECM and MXene/adECM hydrogels (n = 5). * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001, compared with adECM group, while ^###p < 0.001, ^####p < 0.0001, compared between MXene/adECM hydrogel groups

Rheological test further confirmed the thermosensitive sol–gel transition of the MXene/adECM hydrogels. Upon heating to 37 °C, a rapid gelation occurred within 10 min, marked by a sharp increase in storage modulus (G′), which consistently exceeded the loss modulus (G′′) (Fig. 2e). Interestingly, the AM300 group exhibited the highest G′. This nonlinear concentration-mechanical property relationship suggests a balance between MXene’s reinforcement effect and potential interference with ECM self-assembly at higher loadings. The mechanical properties of hydrogels are indeed critical, as they significantly influence cellular behavior [64]. An ideal scaffold for VML repair should ideally match the stiffness of native skeletal muscle, which has a Young’s modulus of approximately 12 kPa [65]. We performed compression modulus tests to demonstrate the mechanical properties of hydrogels (Fig. 2f). The compressive modulus of the adECM hydrogel was measured at approximately 157.3 ± 9.510 Pa. The incorporation of MXene notably enhanced the mechanical strength of the hydrogels, with the AM300 group reaching the highest value of about 375.6 ± 24.66 Pa.

Swelling behavior was assessed by incubating hydrogels in PBS. All hydrogels reached equilibrium after 24 h, exhibiting favorable water absorption with swelling ratios exceeding 60% (Fig. 2g). Good swelling capacity are important factors for materials to absorb nutrients and fluids from the extracellular microenvironment and promote cell survival. The swelling properties depend on the porosity and crosslinking ratio of the material [66, 67]. While the incorporation of MXene did lead to a reduction in the swelling ratio compared to the pure adECM hydrogel, the decrease was only 5.4% (66.55 ± 0.59% for adECM vs. 61.15 ± 0.67% for AM500). This slight reduction is likely attributable to the moderately smaller average pore size observed in the MXene-incorporated hydrogels, as discussed earlier.

The ideal muscle repair scaffold must maintain structural integrity during the critical early regeneration phase while progressively degrading to accommodate tissue ingrowth and remodeling [68]. To assess the degradability, all hydrogels were incubated in PBS at 37 °C, and their mass loss was monitored over time (Fig. 2h). The results showed no significant differences in degradation profiles among the adECM and MXene/adECM groups, with all hydrogels exhibiting over 80% mass loss by day 15. These findings indicated that the incorporation of MXene did not damage the collagen composition of adECM or impede the degradation of hydrogel.

The timeline and biological events of skeletal muscle regeneration have been extensively discussed and can be delineated into three interconnected phases: (1) degeneration and inflammation, (2) repair, and (3) remodeling [12, 69]. The primary events occurring in phase 1 (0–3 days) include the infiltration of inflammatory cells and the polarization of macrophages [69, 70]. In phase 2, the repair process around days 4–10 post-injury, involves the proliferation and myogenic differentiation of endogenous satellite cells or delivered seed cells [71, 72]. The proliferation of myoblasts peaks between 3 and 7 days post-injury, leading into a period of extracellular matrix remodeling approximately one week after injury [69, 73, 74].

Analyzing the hydrogel’s degradation behavior within this regenerative timeline, its degradation rate reached approximately 20% by day 3. This moderate degradation facilitates the infiltration of inflammatory factors and helps activate myogenesis, while the hydrogel continues to fill the defect and support the microenvironment. By days 7–9, the degradation rate reached about 50%. At this stage, myoblasts are highly active in proliferation and begin to differentiate. This level of hydrogel degradation facilitates the release and fusion of encapsulated seed cells, enabling the process to enter the matrix remodeling phase through a dynamic balance between scaffold degradation and new matrix formation, which aligns well with the physiological progression of skeletal muscle regeneration [75]. Nerve regeneration and NMJ formation are relatively later events in muscle repair [69]. A key factor for NMJ regeneration is the proper arrival of incoming regenerating axons to postsynaptic AChRs [76]. Therefore, even though the hydrogel is largely degraded by the later stages, the electrical cues it provides in the early phases can guide the directional growth and extension of axons [77]. Furthermore, the increased synthesis of AChRs promoted by muscle regeneration also favors the re-innervation and reconstruction of NMJs during the mid to late stages. Thus, the MXene/adECM hydrogel offers both early mechanical support and timely degradation, facilitating effective tissue regeneration and integration.

Notably, the incorporation of MXene was intended to enhance the electrical conductivity of the hydrogel, thereby creating a biomimetic electroactive microenvironment to support skeletal muscle and nerve regeneration. To evaluate this, we assessed the conductivity of the MXene/adECM hydrogels. As shown in Fig. 2i, the conductivity increased significantly with rising MXene concentrations, from 4.57 ± 0.14 mS m^{− 1} in adECM to 24.00 ± 0.40 mS m^{− 1} (AM100), 45.63 ± 7.36 mS m^{− 1} (AM300), and 59.05 ± 13.14 mS m^{− 1} (AM500), respectively. The conductivity range observed in the MXene/adECM hydrogels closely matches that of native neural tissues (30–600 mS m^{− 1}), highlighting their suitability as electroactive scaffolds for supporting peripheral nerve regeneration [78]. The uniformly dispersed, highly conductive MXene nanosheets within the adECM hydrogel network form an interconnected nanomaterial architecture that closely mimics the native electrophysiological microenvironment. This continuous conductive network enhances electrical coupling between native and regenerating tissues, thereby facilitating the transmission of bioelectrical signals essential for effective muscle regeneration [18].

In summary, we successfully engineered a conductive MXene/adECM hydrogel with favorable thermoresponsive gelation behavior. The hydrogel rapidly forms a stable, uniformly porous 3D structure at physiological temperature (37 °C), and exhibits excellent anti-swelling properties, biodegradability, and biomimetic electrical conductivity.

Evaluation of biocompatibility and myogenesis-promoting effects of MXene/adECM hydrogels in vitro

To evaluate the biocompatibility of MXene/adECM hydrogels and their effects on myogenic differentiation of L6 cells (rat skeletal myoblast), adECM and MXene/adECM hydrogels (AM100, AM300, AM500) were soaked in growth medium (DMEM supplemented with 10% FBS and 1% penicillin-streptomycin) for 72 h. The hydrogels were then removed, and the resulting supernatants were collected as conditioned medium (CM) for subsequent experiments. DMEM growth medium without hydrogel extract served as the Control group. L6 cells were cultured in these different CMs, and cell proliferation was assessed using the CCK-8 assay. As shown in Fig. 3a, absorbance values at both 24 h and 48 h were significantly higher in the MXene/adECM groups compared to the Control and adECM groups, indicating enhanced cell viability and proliferation. Among all groups, AM300 exhibited the most pronounced proliferative effect, suggesting an optimal MXene concentration for bioactivity. These results demonstrated the excellent biocompatibility of MXene/adECM hydrogels and supported their suitability for further biomedical applications. The mechanisms by which MXenes promote cell proliferation appear to be multifactorial. Firstly, the large surface area and excellent hydrophilicity of MXenes have been proven to promote the adhesion and proliferation of various cells [79]. The anti-inflammatory and antioxidant properties of MXenes further improve the microenvironment by reducing oxidative stress, providing favorable conditions for the proliferation of myoblasts [80]. Crucially, due to the inherent electrical conductivity, MXenes-based materials can significantly promote the proliferation and differentiation of myoblasts [81]. This enhancement is attributed to the improved electrical signal transmission provided by the conductive network [82]. The underlying molecular mechanism may involve the downregulation of nuclear factor-kappa B (NF-κB) signal, modulation of cytokine expression, and activation of calcium signaling pathways [83, 84]. Furthermore, Kang et al. [36] proposed that MXenes enhance the proliferation of myoblasts by elevating intracellular Ca²⁺ levels, thereby activating downstream iNOS and mTOR-AKT signaling pathways. While multiple factors appear to contribute to the pro-proliferative effects of MXenes, the dominant mechanism remains to be fully elucidated.

Fig. 3.

Biocompatibility and myogenesis-promoting effects of MXene/adECM hydrogels in vitro. (a) CCK-8 assay of L6 cells cultured in the conditioned mediums (CMs) at 24 h and 48 h (n = 3). AM100, AM300, and AM500 respectively refer to 100 µg mL^− 1, 300 µg mL^− 1, and 500 µg mL^− 1 MXene/adECM hydrogel. ns: not significant, * p < 0.05, ** p < 0.01, *** p < 0.001,**** p < 0.0001, compared with Control group, while ^#p < 0.05, ^##p < 0.01, ^###p < 0.001,^####p < 0.0001, compared between hydrogel groups. (b) Live/dead staining of L6 cells cultured in adECM and MXene/adECM hydrogels on day 1, 3, and 7. Scale bar: 100 μm. (c) Immunofluorescence staining of L6 cells cultured on glass substrate and in hydrogels after 10 days of myogenic differentiation induction. MHC (red) indicates the myogenic differentiation. DAPI (blue) stains nuclei. Scale bar: 100 μm. (d) Western blotting evaluation of myogenic regulatory proteins (Desmin, Myod1, Myf5) expressed by L6 cells in 2D culture or 3D culture within adECM and MXene/adECM hydrogels. (e) Quantitative analysis of western blotting (n ≥ 3). ns: not significant, * p < 0.05, ** p < 0.01, *** p < 0.001,**** p < 0.0001, compared with 2D group, while ^#p < 0.05, ^##p < 0.01, compared between hydrogel groups

Subsequently, L6 cells were seeded into the MXene/adECM hydrogels to examine their spatial distribution and proliferation within the 3D scaffold. Live/dead staining and confocal imaging revealed uniform cell distribution in both adECM and MXene/adECM hydrogels (Fig. 3b). On day 1, cell densities across all groups were comparable (Fig. S4). By day 3, a significant increase in cell number was observed in the AM300 group. After 7 days of culture, all MXene/adECM groups supported higher cell densities than the adECM hydrogel, with AM300 again showing the highest proliferation rate. These results confirm that the MXene/adECM hydrogels are non-cytotoxic and provide a favorable microenvironment for L6 cell growth in both 2D and 3D settings, with 300 µg mL^− 1 identified as the optimal concentration.

As a natural material, adECM has well-documented biocompatibility in previous studies [56, 85]. While MXene is also generally considered biocompatible, its cytotoxicity appears to be dose- and cell-type dependent [82, 86, 87]. MXenes tend to exhibit greater cytotoxicity toward cancerous cells, with evident toxicity observed at 62.5 mg mL^{− 1} [86]. Qi et al. [82] further demonstrated that coating poly(L-lactic acid) (PLLA) particles with 15 wt% MXene reduced cell proliferation compared to 10 wt%, underscoring the importance of dosage control. However, it is important to note that these studies primarily investigated MXene in solution or at the cell–substrate interface, where direct exposure may intensify cytotoxic effects. In contrast, the incorporation of MXene nanosheets within crosslinked hydrogel matrices offers a more controlled and biocompatible presentation. To mitigate potential risks, we selected MXene concentrations based on the study by Wei et al. [88], in which Ti₃C₂ MXene–Matrigel hydrogels below 500 µg mL^{− 1} demonstrated favorable cell compatibility. Our results were consistent with these findings and further supported the biosafety of MXene within ECM-based hydrogels at optimized doses.

Myosin heavy chain (MHC) is a widely recognized marker of myogenic differentiation [89]. To evaluate the myogenic effects of MXene/adECM hydrogels, we performed immunofluorescence staining of MHC after 10 days of differentiation induction. As shown in Fig. 3c, cells in the MXene/adECM hydrogels exhibited significantly higher MHC expression (red fluorescence) compared to the adECM group, with the most robust myotube formation observed in the AM300 group. Quantitative analysis of the MHC-positive area relative to DAPI showed that the degree of myogenic differentiation in the hydrogel 3D culture groups was significantly higher than in the 2D glass group, with the AM300 group showing the best myogenic differentiation effect (Fig. S5). To further validate the myogenic potential of MXene/adECM hydrogels, we assessed the expression of key myogenic regulatory proteins under different conditions (2D culture, adECM 3D hydrogel, and MXene/adECM 3D hydrogel) using Western blotting. Desmin, the major muscle-specific intermediate filament protein, indicates early myogenic commitment; MyoD1 is a master transcription factor initiating myogenic differentiation; and Myf5 plays a critical role in early myogenic lineage determination [90, 91]. As shown in Fig. 3d, expression levels of these markers were significantly upregulated in cells cultured within MXene/adECM hydrogels compared to the 2D condition after 7 days of differentiation, and the AM300 group exhibited markedly higher levels (Fig. 3e). Collectively, these findings demonstrated that MXene/adECM hydrogels effectively promoted myogenic differentiation of L6 cells, with 300 µg mL^{− 1} identified as the optimal concentration.

During the myogenic differentiation process, the cell-laden hydrogel did not completely disintegrate or disappear within the culture period due to the deposition of new ECM by the cells during 3D culture. Cell-derived ECM replaces a portion of the degrading scaffold, thereby helping to maintain the hydrogel’s structure [92]. In addition, cell-mediated gel compaction was observed in myogenic differentiation experiment. At the endpoint of myogenic differentiation, the hydrogels largely maintained their disc-like shape (Fig. S6) but exhibited varying degrees of areal contraction. Quantitative analysis revealed that the pure adECM hydrogel underwent the most significant contraction, approximately 80% of its initial area. In contrast, the MXene/adECM hydrogels showed less contraction, with the AM300 group retaining about 50% of its initial area.

Cell-mediated contraction results from a combination of factors, including cell activity, the mechanical properties of the hydrogel, and the culture system (whether it is a free-floating culture). Studies have shown that in free-floating culture systems using decellularized skeletal muscle ECM, good cell viability can lead to hydrogel contraction of up to 85% [93]. The higher contraction rate observed in our adECM group is consistent with this phenomenon. In comparison, the weaker contraction observed in the MXene/adECM hydrogels may be attributed to the enhanced mechanical strength imparted by MXene, or potentially to stronger adhesion to the bottom of the confocal dish, which could partially restrain overall contraction. Notably, within the MXene/adECM hydrogels, the AM300 group showed relatively more significant contraction, which may be related to its superior cell proliferation and myogenic induction activity.

Concerning strategies to prevent contraction, methods such as chemical crosslinking or physical confinement can be employed. This study did not specifically implement such strategies, primarily for two reasons: firstly, the hydrogels maintained their basic morphology and integrity throughout the culture period; secondly, literature suggests that moderate, cell-driven hydrogel remodeling and contraction can help guide cell alignment and promote the formation of engineered muscle tissue [94].

MXene/adECM hydrogels promote neurite outgrowth and neuronal differentiation of PC12 cells

To evaluate the neural regenerative potential of MXene/adECM hydrogels, PC12 cells were selected as model cells due to their well-established capacity for neuronal differentiation and widely used in in vitro neural regeneration research [95]. CCK-8 assay was used to assess the biocompatibility of MXene/adECM hydrogels with PC12 cells. Cell viability was assessed at 24 and 48 h. As shown in Fig. S7, all groups showed comparable levels of cell viability with no significant differences. This demonstrated that MXene/adECM hydrogels did not impair PC12 proliferation, even at the highest concentration (500 µg mL^{− 1}).

To further investigate the effects of MXene/adECM hydrogels on neuronal growth and differentiation, PC12 cells were cultured in conditioned neuronal differentiation medium supplemented with 100 ng mL^− 1 nerve growth factor (NGF). After 36 h of pre-culture, live-cell imaging was performed continuously for 36 h using the Incucyte^® imaging system. As shown in Fig. 4a, neurite extension was observed across all groups during neuronal induction. Notably, MXene/adECM groups exhibited more prominent neurite elongation, with axon-like projections extending beyond the diameter of the cell body. Among them, the AM300 group demonstrated the most pronounced neurite outgrowth, characterized by increased branch formation and the development of a complex neurite network. Quantitative analysis confirmed that AM300 significantly promoted neurite length compared to both adECM and other MXene concentrations (Fig. 4b). However, no statistically significant differences in neurite branching were observed between the MXene/adECM and adECM groups (Fig. 4c).

Fig. 4.

MXene/adECM hydrogels promote neurite growth and neuronal differentiation of PC12 cells. a) Continuous live-cell imaging of PC12 cells cultured in conditioned neuronal differentiation medium at 36, 48, 60, and 72 h. AM100, AM300, and AM500 respectively refer to 100 µg mL^− 1, 300 µg mL^− 1, and 500 µg mL^− 1 MXene/adECM hydrogel. Scale bar: 50 μm. b, c) Quantitative analysis of neurite length and neurite branch points (n = 3). ns: not significant, * p < 0.05, ** p < 0.01, **** p < 0.0001, compared with Control group; while ^#p < 0.05, ^##p < 0.01, ^###p < 0.001, compared between hydrogel groups. d) Immunofluorescence staining of PC12 cells cultured on glass substrate and hydrogels after 5 days of neuronal differentiation induction. Tuj1 (green) indicates the neuronal differentiation. DAPI (blue) stains nuclei. Scale bar: 20 μm. e) Quantitative analysis of neurite length and neurite branch points per cell (3 biological replicates). ns: not significant, ** p < 0.01, **** p < 0.0001, compared with Glass group; while ^###p < 0.001, ^####p < 0.0001, compared between hydrogel groups. f) Double immunofluorescence staining of MHC (red) and Tuj1 (green) in the neuromuscular co-culture system. DAPI (blue) stains nuclei. Scale bar: 50 μm

To more directly visualize the interactions between hydrogels and cells, PC12 cells were subsequently seeded onto the hydrogels for neuronal differentiation. The direct hydrogel contact offers a conductive interface that enhances cell–material coupling and facilitates the transmission of intrinsic electrical cues—an important factor in promoting neuronal growth and maturation [96, 97]. After 5 days of differentiation, β-III tubulin (Tuj1), a neuron-specific cytoskeletal marker, was assessed by immunofluorescence staining to evaluate neuronal differentiation [98]. As shown in Fig. 4d, cells cultured on MXene/adECM hydrogels exhibited significantly enhanced neurite outgrowth, with expanded cell bodies and fan-shaped growth cones at the tips of axons, indicative of dynamic axonal elongation [99]. Particularly in the AM300 group, dense neurite networks with numerous spine-like protrusions were observed, suggesting a highly active neuronal phenotype. Such protrusions are believed to represent early presynaptic terminals [100]. Quantitative analysis demonstrated that AM300 significantly promoted neurite outgrowth, with axonal lengths exceeding 100 μm, accompanied by a marked increase in neurite number (Fig. 4e). It is noticed that all groups exhibited relatively thick neurites after short-term (3–5 days) differentiation. This may be related to the specific PC12 cell line used in the experiments, the batch of differentiation reagents, or the relatively short observation time point for differentiation. As shown in Fig. 4d, all groups expressed the neuron-specific marker Tuj1, confirming the successful initiation of neuronal differentiation, despite the differences in morphological presentation. The quantitative metrics used to assess neuronal differentiation in this study are the objective measures of neurite length and neurite number. The data show that within the same morphological context, the AM300 group performed best in these quantitative metrics, with its neurite length and number significantly superior to those of other hydrogel groups and the Glass control group. Furthermore, in the neuromuscular co-culture system shown in Fig. 4f, PC12 cells in the AM300 group exhibited more typical axon-like projections. This indicates that the electroconductive microenvironment provided by AM300 can promote neuronal differentiation.

The neurogenic effects of MXene-based materials are largely attributed by researchers to their high electrical conductivity, which promotes the differentiation of neural stem cells (NSCs) [101]. Enhanced electrical conduction regulates the expression of calcium channels, realigns cell-matrix adhesion sites, and modulates the distribution of growth factor receptors [37, 102]. These changes collectively promote the differentiation of NSCs into neurons and enhance neurogenesis [103]. Furthermore, studies by Yang et al. [104] suggested that several signaling pathways associated with neural differentiation, such as the Hippo signaling pathway, PI3K-Akt signaling pathway, and Wnt signaling pathway, were involved in the pro-differentiation effects of MXenes.

The results have confirmed that MXene/adECM hydrogels promote myogenic differentiation as well as neuronal growth and maturation. However, the dynamic interplay between muscle regeneration and neural innervation remains insufficiently understood. To address this, we established a neuromuscular co-culture system to explore the role of MXene/adECM hydrogels in coordinating muscle regeneration and neural integration. L6 cells were first seeded on adECM or MXene/adECM hydrogels and induced to myogenic differentiation for 7 days. Cytosine β-D-arabinofuranoside (Ara-C), a cytosine nucleoside analog that inhibits DNA replication and synthesis, is commonly used in vitro to inhibit cell proliferation [105]. To suppress further myoblast proliferation, Ara-C was added to the culture medium at 0.5 µg mL^{− 1} for 3 days. Subsequently, PC12 cells were seeded onto the same hydrogel surface and cultured in neuronal differentiation medium for an additional 7 days. This strategy enabled the construction of an in vitro neuromuscular co-culture system. Immunofluorescence staining validated the successful establishment of the co-culture system (Fig. 4f). MHC (red) and Tuj1 (green) were co-stained to identify myogenic and neuronal components, respectively. In all groups, we observed MHC expression and neuronal differentiation, with neurite outgrowth aligned in parallel with the orientation of myotubes. Notably, in the MXene/adECM hydrogels—particularly the AM300 group—we observed enhanced MHC expression, increased myotube diameter, and longer neurite extension compared to the adECM group. Moreover, the extended neurites exhibited close contact with adjacent myotubes, indicating the preliminary formation of NMJ and suggesting the initiation of innervation.

Overall, MXene/adECM hydrogels effectively promoted myogenic differentiation, neurite outgrowth, and neuro-muscular integration in vitro, with the 300 µg mL^− 1 MXene concentration consistently demonstrating the most pronounced effects. Therefore, AM300 was selected for subsequent studies.

Mechanism of MXene/adECM hydrogel promoting muscle innervation

Previous findings suggested that MXene/adECM hydrogel contributed to myogenesis and neural differentiation. However, the underlying biological processes remain incompletely understood. Utilizing our established neuromuscular co-culture system, we performed RNA sequencing on cells seeded on the adECM hydrogel and 300 µg mL^{− 1} MXene/adECM hydrogel to identify the key gene networks modulated by the electrically active microenvironment conferred by MXene. Three biological replicates per group were used for sequencing. Principal component analysis (PCA) revealed a marked divergence in global gene expression profiles between the two groups (Fig. 5a), indicating that the electroactive environment significantly reshaped cellular transcriptional programs. The clustering analysis of transcriptomic data revealed a high degree of similarity among samples within the same group, while a clear distinction was observed between the two groups (Fig. S8). A total of 1,160 differentially expressed genes (DEGs) were identified in the MXene/adECM group relative to the adECM group, including 569 upregulated and 591 downregulated genes (Fig. 5b). Heat map further revealed distinct DEG patterns (Fig. 5c). Among the upregulated genes, Myod1, a myogenic regulatory factor, promotes the differentiation of muscle progenitor cells into myofibers [91]. IGF2BP3, a critical component of the insulin-like growth factor (IGF) signaling network, facilitates muscle repair and myofiber maturation [106].

Fig. 5.

RNA-sequencing analysis reveals the differential transcriptome profiling of neuromuscular co-culture cells in adECM and MXene/adECM hydrogels. (a) Principal component analysis (PCA) plot demonstrating the variance observed in RNA-seq data. (b) Volcano plot of differentially expressed genes (|log₂FC| > 1, q < 0.05; red: up-regulated, blue: down-regulated). (c) Hierarchical clustering heatmap of the top 20 differentially expressed genes (DEGs). (d) Gene ontology (GO) terms of the top 30 up-regulated biological processes between MXene/adECM and adECM groups. Blue boxes: myogenic differentiation-related terms. Red boxes: neural function-related terms. (e) Chord diagram visualizing enriched GO terms and associated genes (outer ring: pathways; inner arcs: DEGs). (f) KEGG pathway analysis according to DEGs between MXene/adECM and adECM groups

Gene Ontology (GO) enrichment analysis demonstrated significant enrichment of neuromuscular-associated biological processes in the MXene/adECM hydrogel group, including “skeletal muscle contraction,” “acetylcholine receptor activity,” and “regulation of membrane potential” (Fig. 5d), suggesting a potential role for MXene in supporting the structural and functional maturation of NMJ. Chord diagram visualization further elucidated the connections between these enriched pathways and key DEG (Fig. 5e). Notably, downregulated GO terms in the MXene/adECM group encompassed “collagen-containing extracellular matrix” and “extracellular matrix organization” with collagen-related genes such as Col1a1 and Col11a1 prominently suppressed. This reflects MXene’s capacity to inhibit excessive collagen deposition compared to adECM, aligning with our histological observations of reduced tissue fibrosis. Significantly upregulated GO processes included “ion transport,” “skeletal muscle contraction,” and “neuronal cell body,” all critically associated with neuromuscular conduction. Key upregulated genes included the calcium-binding protein S100G, which regulates intracellular Ca²⁺ homeostasis; the sarcomeric elastic protein TTN, essential for maintaining muscle structural integrity and mechanical stability; Neuroligin 3 (NLGN3), a pivotal regulator of neurite outgrowth and neuritogenesis [107–109]. Complementary KEGG pathway analysis corroborated these findings, revealing significant activation of key neuromuscular functional pathways in the MXene/adECM group, including “cholinergic synapse” “neuroactive ligand-receptor interaction” and “calcium signaling pathway” (Fig. 5f). These results indicated that under the influence of the electroactive microenvironment, the functional transmission between nerve axons and muscle fibers, along with NMJ formation, may be mediated through calcium signaling mechanisms.

These transcriptomic analyses comprehensively delineated the gene expression profiles distinguishing MXene/adECM from adECM hydrogels, highlighting two primary aspects: skeletal muscle regeneration and the establishment of neuromuscular innervation. To gain deeper insights into myogenic differentiation, we examined the expression of several key myogenic genes (Fig. 6a). Interestingly, Myf5 and Des were found to be downregulated in the MXene/adECM group, contrasting with their upregulation in L6 myogenic differentiation assays (Fig. 3d). Notably, both Myf5 and Des are early-stage markers of myogenesis [90, 91]. Their decreased expression, coupled with the upregulation of late-stage markers (Myd1, MyoG, and Myh1), suggests active myotube formation and maturation [52, 91]. This shift likely reflects the transcriptomic profile under the neuromuscular co-culture system, where the extended culture period (17 days vs. 7 days) and the presence of neuronal input contribute to the observed differences. Further analysis of muscle function regulators (Fig. 6b) demonstrated downregulation of ERBB receptor feedback inhibitor 1 (Errfi1) alongside upregulation of Erbb2 and Erbb3. As a negative regulator of ERBB signaling, suppressed Errfi1 releases inhibition on Erbb2/3, thereby potentiating neuromuscular synapse formation and muscle spindle development [110]. In addition, members of the troponin complex (Tnni, Tnnt, and Tnnc) were broadly upregulated, indicating enhanced excitation–contraction coupling under neural regulation.

Fig. 6.

DEGs analysis reveals the mechanisms by which MXene/adECM hydrogel promotes skeletal muscle regeneration and innervation. (a) Heatmap of key myogenic differentiation genes differentially expressed in MXene/adECM vs. adECM groups. (b) Heatmap of muscle function regulator genes differentially expressed in MXene/adECM vs. adECM groups. (c) Voltage-gated calcium channels (VGCCs) gene expression profiles. (d) Proposed mechanism: Ca²⁺ influx coordinates NMJ formation and neuromuscular transmission. (e) RT-qPCR analysis of acetylcholine receptor subunit genes (n = 3). * p < 0.05, ** p < 0.01. (f) Immunofluorescence staining of acetylcholine receptors (α-BTX, magenta) in neuromuscular co-culture system (green: GFP⁺ PC12 cells, blue: nuclei). Scale bar: 50 μm. (g) RT-qPCR analysis of calcium channel genes (n = 3). * p < 0.05. (h) Live calcium imaging in co-cultured myotubes (red: Cal-630, green: GFP⁺ PC12 cells). White dashed lines: regions of interest (ROI). Scale bar: 20 μm. (i) Relative change in fluorescence intensity of the ROI (300 s recording). (j) Quantification of Relative fluorescence intensity peak of the ROI (n = 3). * p < 0.05

Taken together, these results further elucidate the biological processes underlying muscle structural maturation and functional stabilization promoted by the MXene/adECM hydrogel. Notably, Ca²⁺ signaling appeared to be broadly involved in this process. We thus hypothesize that voltage-gated calcium channels (VGCCs)-mediated Ca²⁺ influx may serve as a central regulatory node for neuromuscular innervation. Transcriptomic visualization of calcium channels (CaVs) related genes (Fig. 6c) revealed that MXene/adECM hydrogel specifically upregulated L-type CaVs (Cacna1s, Cacna1c, Cacna1d) and the P/Q-type channel Cacna1a (localized to presynaptic membranes of NMJ), while downregulating T-type channels (Cacna1g, Cacna1i, Cacna1h). Critically, the skeletal muscle-specific L-type CaV1.1 governs postsynaptic Ca²⁺ influx, activating activity-dependent calcium signaling pathways that regulate both contraction and postsynaptic nAChR clustering [33, 34]. CaV2.1, the primary subtype of presynaptic P/Q-type VGCCs, is localized at the active zones of motor neuron terminals, where it mediates rapid, millisecond-scale calcium transients [111]. These pulsed Ca²⁺ signals bind to calcium sensors on synaptic vesicles (SVs), triggering their fusion with the presynaptic membrane and subsequent release of acetylcholine (Fig. 6d) [32]. Based on our transcriptomic findings, we propose that the MXene/adECM hydrogel enhances the expression of both CaV1.1 and CaV2.1 channels, facilitating Ca²⁺ influx as a key signaling hub for functional muscle regeneration. This coordinated mechanism is likely critical for NMJ formation, synaptic transmission, and ultimately, restoration of muscle contractile function.

Subsequently, we conducted experimental validation to verify the findings from the transcriptomic analysis. RT-qPCR analysis demonstrated significant upregulation of nAChR subunit genes Chrna1 and Chrnb1 in MXene/adECM hydrogel (Fig. 6e). To visualize receptor clustering, we established a co-culture system integrating GFP-labeled PC12 cells with L6 cells. We employed AF647-conjugated α-bungarotoxin (α-BTX), a fluorescently labeled polypeptide neurotoxin that binds with high specificity to nAChRs. As shown in Fig. 6f, the MXene/adECM group exhibited clustered fluorescence signals of α-BTX (magenta), which was co-localized with PC12 cells. These findings indicated the formation of nAChR clusters, suggesting that the conductive environment promoted the reconstruction of NMJ.

RT-qPCR analysis confirmed significantly elevated expression of Ca²⁺ channels of neuromuscular co-culture system in MXene/adECM hydrogel, with Cacna1a (CaV2.1) and Cacna1s (CaV1.1) levels increasing versus adECM group (p < 0.05, Fig. 6g). To further assess the functional consequence of increased channel expression, calcium influx assays were conducted in the neuromuscular co-culture system using red fluorescent calcium indicator (Cal-630^™ AM). High-concentration K⁺ solution was used as a depolarizing stimulus, triggering neuromuscular electrical activity. Upon stimulation with 30 mM KCl, time-lapse imaging was performed for 300 s, with frames captured every 5 s. Distinct Ca²⁺ transients were observed in myotubes adjacent to GFP-labeled PC12 neurons (dashed boxes, Fig. 6h). MXene/adECM group exhibited changed Ca²⁺ signals, while adECM group showed minimal responses (Supplementary Material 2, 3). Quantitative analysis of relative fluorescence intensity (ΔF/F₀) revealed significantly higher peak amplitudes (F_peak) and dynamic fluctuations in MXene/adECM groups (Fig. 6i, j), confirming enhanced Ca²⁺ influx.

It is important to compare the neuromuscular co-culture model with other models simulating neuromuscular junctions [112, 113]. The neuromuscular co-culture models established or reviewed in these works utilize human iPSC-derived cells within engineered microfluidic or 3D hydrogel matrices, excelling in recapitulating the spatial complexity and advanced functionality of native NMJs. In contrast, the co-culture model used in our work, which employs a sequential bilayer seeding of rat cell lines (L6 and PC12) on hydrogel surfaces, is conceptually and technically much simpler than the models described in the cited literature. However, this model serves to preliminarily yet effectively evaluate whether the conductive MXene/adECM hydrogel can support key nerve-muscle interactions, such as axonal extension, physical contact between neurons and myotubes, and the formation of NMJ structures. This simplified model provides a cost-effective tool suitable for assessing the bioactivity of novel biomaterials within a neuromuscular context.

These findings demonstrate that, upon upregulating calcium channel expression, MXene/adECM hydrogel promotes calcium signaling in response to neuromuscular stimulation. This enhancement of calcium dynamics at the neuromuscular interface is critical for functional maturation of regenerated muscle [33, 34]. In summary, these results verify that the conductive MXene/adECM hydrogel effectively facilitates calcium channel expression and activity, thereby promoting NMJ reconstruction and contributing to functional muscle regeneration, which is essential for VML therapy.

MXene/adECM hydrogel promotes skeletal muscle regeneration following volumetric muscle loss

To evaluate the in vivo muscle regeneration potential of the MXene/adECM hydrogel, we established a rat VML model and assessed the therapeutic effects of L6 cell-laden adECM hydrogel and MXene/adECM hydrogel on skeletal muscle repair. Structural regeneration and functional recovery were systematically and comprehensively analyzed through muscle mass measurement, histological staining, imaging, functional assessment, and biosafety evaluation (Fig. 7a). The VML model was created in the tibialis anterior (TA) muscle based on previously reported protocols, with a defect size of 7 × 10 × 3 mm [46]. To ensure that the removed muscle volume exceeded 20% of the total TA mass, we estimated the TA weight for each rat using a formula [50]. The proportion of the defect was calculated based on the excised tissue weight (Fig. S9). The excised TA proportion was approximately 30% in all groups, including the untreated group (defect without treatment), the adECM group (L6 cell-laden adECM hydrogel), and the MXene/adECM group (L6 cell-laden MXene/adECM hydrogel), with consistent defect ratios across all animals.

Fig. 7.

MXene/adECM hydrogel promotes skeletal muscle regeneration in VML rat model. (a) Schematic diagram of animal experiments. VML: volumetric muscle loss. TA: tibialis anterior. IF: immunofluorescence. MRI: magnetic resonance imaging. (b) Weekly body weight growth of animals in each group (n = 4). (c) Quantitative analysis of improved TA muscle weight (n = 4). (d,f) H&E staining and Masson staining of regenerated muscle at the VML site on day 30. Scale bar: 300 μm. Black dashed lines indicate the approximate interface along the remaining muscle mass. (e) Quantitative analysis of the area of myofibers per field (n = 4). (g) Quantitative analysis of collagen volume fraction per field (n = 4). (h) Immunofluorescence staining of regenerated muscle on day 30 (red: MHC, green: Laminin, blue: DAPI). Scale bar: 200 µm. White dashed lines indicate the approximate interface along the remaining muscle mass. (i) Quantitative analysis of MHC fluorescence intensity per field (n = 4). (j) Immunofluorescence staining of CD31 (green) on day 30 (blue: DAPI). Scale bar: 100 µm. (k) Quantitative analysis of CD31 fluorescence intensity per field (n = 4). (l) Immunofluorescence staining of Tuj1 (red) on day 30 (blue: DAPI). Scale bars: 100 µm (overview), 10 µm (in frame). (m) Quantitative analysis of Tuj1 fluorescence intensity per field (n = 4). (n) Immunofluorescence staining of NMJ on day 30 (green: SV2, red: α-BTX, blue: DAPI). Scale bars: 25 µm (overview), 5 µm (in frame). (o) Quantitative analysis of SV2/AChR overlap per NMJ (18–20 NMJs from 4 animals each group). ns: not significant, * p < 0.05, ** p < 0.01, *** p < 0.001,**** p < 0.0001, compared with Untreated group, while # p < 0.05, ## p < 0.01, ### p < 0.001, compared with adECM group

Biocompatibility is a fundamental requirement for the clinical application of biomaterials [114]. To evaluate the in vivo biosafety of the materials, we monitored animal body weights at 7, 14, 21, and 30 days post-implantation. As shown in Fig. 7b, all experimental groups exhibited a continuous increase in body weight throughout the treatment period, with no signs of distress or abnormal behavior observed by the end of the study. Further histopathological examination of major organs (heart, liver, spleen, lungs, and kidneys) harvested after 30 days revealed no evidence of morphological abnormalities, structural lesions, or inflammatory infiltrates upon H&E staining (Fig. S10). Then we performed quantitative ICP-MS analysis of major organs, muscle, and fascial tissue (Fig. S11). The results show that on day 30, the Ti content in major organs and the tibialis anterior muscle of the MXene/adECM group showed no significant difference from the uninjured group. However, a significant increase in Ti content was detected specifically in the fascial tissue of the MXene/adECM group. These data collectively indicate that MXene residues were primarily confined to the fascial layer at this time point, exhibiting delayed clearance.

The metabolic pathways and biocompatibility of MXene have been extensively studied, generally confirming its good biosafety profile [115]. For instance, Pan et al. [116] reported no toxic effects upon evaluating MXene-based implants in bone tissue over 24 weeks. It is understood that MXene can gradually degrade into smaller fragments under physiological conditions and be effectively cleared via pathways involving urine, feces, and the reticuloendothelial system (e.g., liver and spleen) [117]. To investigate the delayed clearance in fascia, we performed histological examination (Fig. S11c) and observed the presence of foreign body giant cells (FBGCs) and encapsulated MXene residues, but no severe tissue necrosis or suppurative inflammation. We hypothesize that the dense connective tissue structure of the fascia itself may have limited sufficient infiltration of inflammatory cells and clearance efficiency, leading to the encapsulation and retention of residues. In future work, we will focus on optimizing material degradation kinetics or adjusting implantation strategies to promote more complete clearance.

The morphological appearance of the regenerated TA muscle is shown in Fig. S12a. At both 14 and 30 days post-surgery, the untreated and adECM groups exhibited apparent volume loss compared to the uninjured group, while the MXene/adECM group displayed a more complete and fuller muscle contour. At the 30-day postoperative observation, the hydrogel had integrated well with the host tissue, and no obvious material boundaries or residual fragments were observed at the implantation site. This is consistent with the reported in vivo degradation kinetics of ECM hydrogels, which are typically fully absorbed within 3–4 weeks [118, 119]. Quantitative assessment of TA mass at 30 days post-operation showed no statistically significant differences among groups (Fig. S12b), likely attributable to the substantial baseline weight of TA muscles masking localized regenerative changes. To precisely evaluate neo-tissue formation specifically within the defect area, we applied a previously reported method for calculating regenerated muscle weight in VML models. Based on this analysis (Fig. 7c), both the adECM (72.58 ± 8.80 mg) and MXene/adECM (101.90 ± 15.25 mg) groups showed significantly greater regenerated mass compared to the untreated group (25.10 ± 3.63 mg, p < 0.001). Notably, the MXene/adECM group outperformed adECM group by 29.37 ± 7.34 mg (p < 0.01), suggesting enhanced myogenic restoration mediated by the conductive hydrogel.

These macroscopic findings were further substantiated through histological analyses to delineate the muscle regeneration process within the hydrogel scaffolds. As shown in Fig. S13a, H&E staining revealed pronounced fibrotic tissue in the untreated group at day 7, with no evidence of muscle regeneration. In the adECM group, a clear boundary remained between the implanted hydrogel and native muscle, with minimal neotissue formation. In contrast, the MXene/adECM group showed the emergence of nascent muscle fiber-like structures (arrow), accompanied by integration between the hydrogel and host tissue. By day 14, fibrotic overgrowth persisted in the untreated group, while limited new muscle fibers appeared in the adECM group. Notably, the MXene/adECM group exhibited abundant, fused muscle fibers within the defect region, suggesting accelerated myogenesis (Fig. S13a). These observations align with the known temporal dynamics of skeletal muscle regeneration, in which early fibrotic deposition competes with muscle fiber formation [12]. At day 30, serial sections of the defect site were stained with both H&E and Masson staining to provide comprehensive insights into the muscle regeneration and fibrosis (Fig. 7d, f). The untreated group displayed extensive blue collagen deposition, indicative of scar tissue formation and impaired functional restoration. Although the adECM group presented a greater number of regenerated fibers than the untreated group, excessive collagen accumulation remained evident. The MXene/adECM group showed a markedly improved muscle structure, with abundant muscle fibers and minimal collagen deposition confined primarily to the periphery of the scaffold. Quantitative analysis supported these findings. Although the adECM group had the highest number of muscle fibers per field (Fig. S13b), their diameters were significantly smaller than those in the MXene/adECM group (Fig. S13c), consistent with immature muscle fiber morphology observed in H&E sections. Given that muscle fiber cross-sectional area (FCSA) is a widely recognized metric of muscle regeneration quality, we measured the area of myofibers across all groups [120]. As shown in Fig. 7e, the uninjured control had an FCSA of 5146 ± 339.8 µm², while the untreated and adECM groups showed significantly reduced values (270.1 ± 46.39 µm² and 872.2 ± 308.7 µm², respectively). Importantly, the MXene/adECM group exhibited an FCSA of 2945 ± 855.7 µm², which was significantly higher than that of adECM group (p < 0.01), indicating enhanced fiber maturation. Quantitative analysis of Masson staining revealed that collagen deposition in the MXene/adECM group was significantly lower than that in the untreated and adECM groups (Fig. 7g). These results collectively suggested that the MXene/adECM hydrogel not only supported skeletal muscle regeneration but also inhibited pathological fibrosis, thereby providing a favorable microenvironment for functional muscle reconstruction. The observed collagen deposition in the adECM group is unlikely to be derived from the hydrogel itself. Firstly, the anti-inflammatory and immunomodulatory effects of adECM have been widely validated [56, 121]. The Masson staining and quantitative analysis in this study indicate that, compared to the untreated defect group, the adECM hydrogel actually reduced the overall degree of fibrosis. Therefore, the increased collagen level in this group is more likely to reflect the fibrotic repair process in VML that was not fully suppressed, rather than being solely caused by the implanted material. Combined with the in vitro degradation data, the adECM hydrogel is designed to provide temporary support in the early stages and to be sufficiently degraded and replaced by host cells in the mid to late stages (approximately 3–4 weeks).

To further assess the maturity and structural integrity of newly formed muscle tissue, we performed dual immunofluorescence staining of MHC (red) and Laminin (green) to specifically visualize newly formed myofibers and their surrounding basement membranes. As shown in Fig. 7h, the untreated and adECM groups displayed limited MHC⁺ fiber regeneration and immature basement membrane formation with fragmented Laminin expression. Strikingly, the MXene/adECM group showed abundant MHC⁺ fibers arranged in a highly aligned manner, with continuous Laminin surrounding individual fibers—features closely resembling native skeletal muscle tissue. Quantitative analysis confirmed a significant increase in MHC fluorescence intensity in the MXene/adECM group compared to the adECM group (Fig. 7i), indicating improved muscle regeneration. Notably, the MXene/adECM hydrogel did not significantly promote the directional alignment of myotubes in vitro (Fig. 4c). The transition from disordered myotubes in vitro to aligned myofibers in vivo can be explained by the guiding role of host tissue. Previous research has indicated that the residual muscle fibers after injury can guide the alignment and fusion of newly formed myofibers [122]. In future investigations, the use of fluorescently labeled transplanted cells could further elucidate whether regenerating myofibers integrate and align under the guidance of the native host muscle architecture. Furthermore, whether the conductivity of MXene contributes to guiding muscle alignment requires additional exploration through in vitro and in vivo experiments combined with directional electrical stimulation.

Vascularization plays a critical role in VML repair. Adequate angiogenesis can ensure the oxygen and nutrient supply required for muscle regeneration, as well as regulate immune responses [123]. As shown in Fig. 7j, the untreated and adECM groups exhibited disordered microvascular with small vessel diameters, suggesting immature and insufficient angiogenesis. In the MXene/adECM group, CD31⁺ vascular structures with enlarged lumens were observed, indicating enhanced vascularization (Fig. 7k). These findings suggested that the conductive hydrogel not only supports muscle fiber regeneration but also facilitates vascular network reestablishment, thereby providing a more favorable microenvironment for functional muscle recovery.

Nerve regeneration and muscle reinnervation were key focuses of this study. Tuj1 staining was employed to assess nerve infiltration within the regenerated muscle. As shown in Fig. 7l, the MXene/adECM group exhibited nerve bundle-like structures similar to those in the uninjured group, along with a significantly higher expression of Tuj1 compared to other groups (Fig. 7m). This indicates that the MXene/adECM hydrogel effectively promotes nerve regeneration in vivo. To further evaluate NMJ reformation, double immunofluorescence staining was performed using antibodies against the presynaptic marker SV2 and α-bungarotoxin (α--BTX) to label postsynaptic AChR. Normal muscle tissue displays characteristic pretzel-like structures with extensive overlap between pre- and postsynaptic markers, signifying innervated NMJs (Fig. 7n). In the MXene/adECM group, both AChR expression and the degree of NMJ marker co-localization were significantly increased compared to the adECM group (Fig. 7o, Fig. S14). Acetylcholinesterase (AChE) is an essential enzymatic component of NMJ [124]. To assess the formation of NMJ, AChE immunofluorescence staining was performed on tissue sections. The results showed that normal muscle tissue exhibited strong AChE positivity surrounding muscle fibers (Fig. S15a). The expression of AChE in Untreated and adECM groups markedly decreased. In contrast, the MXene/adECM group exhibited significantly enhanced AChE expression, which was similar to the clustered structure in native muscle. The quantitative analysis of the fluorescence intensity demonstrated that MXene/adECM hydrogels achieved the highest values across all parameters, suggesting superior expression of AChE and the regeneration of NMJ (Fig. S15b-d).

We attribute the enhanced innervation potential of MXene/adECM hydrogel primarily to the electroactive microenvironment. As mentioned in previous study, 200 µg mL^{− 1} MXene/GelMA hydrogel facilitated the connection between the newborn nerve and the injured axis, indicating the prospective application of MXene-based conductive scaffolds in neural tissue repair [125]. Wang et al. [17] utilized the high conductivity of MXene-based nerve conduits to facilitate the guided and directional migration of nerve cells, rapidly accelerating the healing of the peripheral nerve injury. As our in vitro observations, MXene/adECM hydrogel facilitated axonal outgrowth and guided their alignment along newly formed myotubes. Additionally, MXene/adECM hydrogel promoted myofiber regeneration and Laminin deposition while suppressing fibrotic scar formation, creating a supportive niche for axonal growth and functional integration. These effects are conducive to NMJ formation and functional reinnervation of regenerated muscle tissue.

The question of whether the regenerated NMJs originate from the transplanted L6 cells or the host cells is indeed highly relevant to understanding the underlying regenerative mechanism. We conducted additional experiments using acellular scaffolds (A-adECM and A-MXene/adECM) and performed comprehensive histopathological and immunofluorescence evaluations (Fig. S16). Histological (H&E) analysis confirmed that cell-free scaffolds exhibit limited muscle regeneration capacity, consistent with previous literature [126]. A-adECM group induced only a small number of regenerating muscle fibers. In contrast, the A-MXene/adECM group showed more regenerated myofibers (Fig. S16a), suggesting that the incorporation of MXene more effectively recruits host satellite cells and promotes their myogenic differentiation, likely due to the electroactive microenvironment it provides.

We further investigated whether MXene could promote host-derived NMJ regeneration in the absence of seed cells. We performed Tuj1 and SV2/AChR staining on both acellular groups and compared them with L6-loaded groups. The results showed that the A-MXene/adECM group significantly enhanced Tuj1-positive nerve growth compared to the untreated and A-adECM groups, indicating that MXene itself can independently promote host nerve infiltration (Fig. S16b, d). Regarding NMJ formation, the A-MXene/adECM group outperformed the A-adECM group, with significantly higher AChR synthesis (Fig. S16c, e). However, compared to the L6-loaded MXene/adECM group, the NMJ regeneration in the A-MXene/adECM group was markedly inferior, characterized by fragmented AChR structures and insufficient co-localization of pre- and post-synaptic markers. These results indicate that while MXene has the intrinsic capacity to facilitate NMJ reconstruction, such a process requires a foundation of robust muscle regeneration to fully support mature AChR clustering, axonal growth, and NMJ stabilization. Therefore, the delivery of seed cells with potent myogenic differentiation potential remains essential for optimal repair.

MXene/adECM hydrogel promotes skeletal muscle volume restoration and functional recovery

Following assessments of muscle mass and histological regeneration, we further performed a comprehensive evaluation to determine the volumetric and functional recovery of regenerated tissues. T2-weighted magnetic resonance imaging (MRI) revealed that the defect region of untreated group displayed diffuse hyperintense signals with indistinct margins, indicating structural disorganization (Fig. 8a). The adECM group showed partial filling of the defect area, but the signal remained heterogeneous. Notably, the MXene/adECM hydrogel group exhibited T2-weighted signals with intensity and distribution patterns closely resembling native muscle, suggesting superior structural restoration. These imaging observations were confirmed by ultrasound analysis. Both side-sectional and cross-sectional views showed that muscles in the MXene/adECM group were denser with well-defined contours (Fig. 8b). Quantitative analysis showed that the sectional muscle areas of the MXene/adECM group are significantly larger than those of the adECM group. Furthermore, 3D reconstruction confirmed the most complete volumetric restoration in the MXene/adECM group (Fig. 8c-e).

Fig. 8.

MXene/adECM hydrogel promotes skeletal muscle volume restoration and functional recovery. (a) Magnetic resonance imaging (MRI) of TA muscle on day 30. Yellow circles indicate the cross-sectional view of TA muscle. (b) Ultrasound images of TA muscle on day 30, including side-sectional view and cross-sectional view. c, d) Quantitative analysis of side-sectional area and cross-sectional area of TA muscle (n = 4). ns: not significant, ** p < 0.01, *** p < 0.001,**** p < 0.0001, compared with Untreated group, while ^#p < 0.05, ^##p < 0.01, compared with adECM group. e) Quantitative analysis of TA muscle based 3D reconstruction (n = 4). ns: not significant, *** p < 0.001,**** p < 0.0001, compared with Untreated group, while ^##p < 0.01, compared with adECM group. f) Schematic diagram of gait analysis experiment. g, h) Representative 2D and 3D stress diagrams of paw pressure. i, j,k) Quantitative analysis of print area, swing/stance, and speed (n = 4). ns: not significant, * p < 0.05, ** p < 0.01, compared with Untreated group, while ^#p < 0.05, ^##p < 0.01, compared with adECM group. l) Isometric tetanic force (ITF) of the TA muscle on day 30. m) Quantitative analysis of ITF (n = 4). ns: not significant, ** p < 0.01, *** p < 0.001, compared with Untreated group, while ^#p < 0.05 compared with adECM group

Gait analysis further revealed functional differences between groups. Paw pressure and impulse were monitored using a pressure-sensitive walkway (Fig. 8f). Representative 2D and 3D stress diagrams (Fig. 8g, h, based on gait traces in Fig. S17a) revealed that the untreated group exhibited impaired weight-bearing on the affected limb, with a markedly reduced contact area of injured hind foot. While the adECM group showed partial recovery of contact area, irregular stress distributions in the paw and toes persisted. In the MXene/adECM group, a near-normal stress distribution and a significantly restored print contact area were observed (Fig. 8i). Regarding core parameters of gait patterns, there were no significant differences in stride, swing duration, and standing duration between groups (Fig. S17b–d). This may be because these parameters reflect the rhythm and temporal characteristics of the gait cycle, changes in which typically require more severe neuromuscular damage [127, 128]. In comparison, parameters such as swing/stance ratio and walking speed are more sensitive indicators of gait coordination and voluntary motor function. The MXene/adECM group exhibited significantly higher swing/stance ratios and walking speeds than the adECM group (Fig. 8j, k), suggesting better functional recovery of the injured limb.

To further verify whether functional recovery of muscle was accompanied by effective neural regeneration, we assessed the isometric tetanic force (ITF) of the TA muscle at 30 days post-surgery, following the method described by Kim et al. [129]. As illustrated in Fig. 8l, the sciatic nerve was stimulated with regular electrical stimulation to induce contraction of the TA muscle, and the ITF was recorded in real time. In the untreated group, VML markedly impaired muscle contractility, and electrical stimulation triggered unstable responses with pronounced fluctuations and irregular baseline. The adECM group showed a slight improvement in ITF, but it remained significantly lower than that of the uninjured group (0.057 ± 0.014 N vs. 0.096 ± 0.001 N), indicating limited restoration of neuromuscular function (Fig. 8m). In contrast, the MXene/adECM hydrogel group achieved ITF levels comparable to those of the uninjured group and significantly higher than those of the adECM group (0.082 ± 0.012 N vs. 0.057 ± 0.014 N, p < 0.05). Additionally, stable ITF baseline and synchronized response patterns were observed, indicating not only enhanced muscle regeneration but also improved neuromuscular transmission. A limitation of the in vivo study is the absence of definitive electrophysiological evidence. Future investigations will incorporate electrophysiological assessments, such as compound muscle action potentials (CMAPs) and nerve conduction velocity (NCV) measurements, to provide a more comprehensive evaluation of neuromuscular signaling recovery.

Taken together, the structural-functional evaluation consistently demonstrated that the MXene/adECM hydrogel not only promoted muscle tissue regeneration and volume restoration, but also outperformed in improving gait behavior and neuromuscular transmission. These results highlight its strong therapeutic potential in the repair of VML.

Limitations and future perspectives

This study has several limitations. First, the mechanical strength of the constructed MXene/adECM hydrogel is relatively low and does not match the mechanical properties of native skeletal muscle tissue, which may affect the myogenic differentiation of cells and the mechanical support provided after implantation. In the future, it will be necessary to enhance the material strength by developing composite hydrogel systems or introducing chemical crosslinking strategies. Second, the model used to validate the conductive hydrogel’s promotion of NMJ regeneration is relatively simplified and may not fully reflect the complex structure and function of the neuromuscular system. Future research should transition to more physiologically relevant models, such as integrating primary or iPSC-derived cells and utilizing advanced culture systems like organ-on-a-chip platforms, to investigate the formation and maturation of functional NMJs on optimized hydrogels. Furthermore, although the degradability and biosafety of MXene have been widely discussed, this study observed delayed clearance of Ti₃C₂T_x MXene in the muscle fascial connective tissue, suggesting potential risks associated with long-term retention. Future studies should conduct more detailed and long-term evaluations of its biodistribution, degradation kinetics, and systemic toxicity. In the treatment of VML, optimizing the hydrogel’s encapsulation and in situ retention capacity to prevent leakage of MXene components into the fascial layer could also serve as a feasible improvement strategy.

Conclusion

In conclusion, our study reports the development and application of a novel conductive hydrogel for muscle tissue engineering. By incorporating Ti₃C₂T_x MXenes into adECM, we successfully constructed a biomimetic microenvironment that closely resembles the native electroactive neuromuscular tissue. The MXene/adECM hydrogel exhibits excellent biocompatibility and favorable physicochemical properties, including a uniformly crosslinked 3D porous structure, high conductivity, and biodegradability—features well aligned with the requirements for functional muscle regeneration. Both the in vitro experiments and the in vivo rat tibialis anterior VML model confirm that the MXene/adECM hydrogel effectively supports structural and functional muscle regeneration. Importantly, the incorporation of MXene markedly enhances muscle innervation compared to adECM hydrogel. Using a neuromuscular co-culture system, we further elucidated the underlying mechanism: MXene’s electrical coupling effect promotes the expression of specific voltage-gated calcium channels at the motor endplate, thereby facilitating calcium-mediated signaling and NMJ formation. These results support the promise of MXene/adECM hydrogels in VML therapy. It is noteworthy that three-dimensional (3D) bioprinting offers a promising approach for creating precise structures in tissue engineering. This technique could further facilitate vascularization and innervation in regenerated muscle tissue. The thermosensitive property of the MXene/adECM hydrogel provides a feasible foundation for its application in bioprinting, and future studies will explore its potential for constructing complex architectures through this approach. Nonetheless, future efforts should focus on optimizing material performance, improving cost-efficiency, and conducting comprehensive biosafety assessments to accelerate clinical translation.

Author contributions

**Mengying Jin** : Writing – original draft, Writing – review & editing, Data curation, Investigation, Methodology, Visualization, Conceptualization, Funding acquisition. **Yahui Zhang** : Writing – original draft, Data curation, Formal analysis, Methodology, Validation, Visualization. **Wei Liang** : Writing – original draft, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Supervision, Visualization. **Rigele Ao** : Data curation, Formal analysis, Investigation, Methodology, Resources. **Yuchen Zhou** : Formal analysis, Investigation, Methodology, Visualization. **Wanweng Dang** : Investigation, Methodology, Visualization. **Hongxu Wu** : Investigation. **Meng Han** : Investigation. Yonghuan Zhen: Resources. **Yang An** : Writing – review & editing, Funding acquisition, Resources, Supervision, Conceptualization.

Funding

Funding sources are inserted in acknowledgment part.

Data availability

The datasets and the materials used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

All experiments involving animals were conducted according to the ethical policies and procedures approved by the ethics committee of Peking University Third Hospital (No. A2022068) and were performed according to the National Institutes of Health regulations for the care and use of animals.

Consent for publication

All authors of this study agreed to publish.

Competing interests

The authors declare no competing interests.