Pre-priming cell sheet therapy enabled by dynamic wrinkled electroactive substrate for muscle reconstruction

Abstract

Current therapeutic approaches for muscle reconstruction face considerable challenges, particularly in generating sufficiently dense cell aggregates and in establishing effective methods for reactivating the function of exogenous cells. Herein, we developed a pre-priming cell sheet therapy for volumetric muscle loss (VML) that leverages highly dense, electro-mechanically bioactive constructs. To achieve this goal, we fabricated a multifunctional cell culture platform based on a near-infrared (NIR)-responsive, wrinkle-patterned, conductive substrate. This system enables scalable preparation (>6 mm in diameter), non-invasive harvesting, and bioactive pre-priming of cell sheets for transplantation. Non-invasive harvesting of the sheets is achieved via a NIR-triggered release mechanism, in which dynamic changes in wrinkle morphology induce a sufficient shift in mechanical stress at the cell-substrate interface, thereby disrupting focal adhesions. Compared with conventional cell-suspension therapy, the microstructured electroactive surface demonstrated superior efficacy for VML repair, as evidenced by integrated in vitro electrophysiology, RNA sequencing, and in vivo analysis. This enhancement is attributed to the substrate's provision of combined electrical and mechanical priming cues, which collectively promote myogenic differentiation, growth, and pro-regenerative calcium signaling in C2C12 myoblasts. In conclusion, this work establishes that engineering interfacial dynamics—rather than relying solely on static material properties—is pivotal for the development of advanced cell therapies. The dynamic electroactive substrate offers a versatile strategy for fabricating pre-functionalized tissue constructs, with immediate promise for regenerating electroexcitable tissues and broad application prospects in regenerative medicine.

Graphical abstract

Highlights

• •A dynamic wrinkled electroactive substrate non-invasive harvests scaffold-free cell sheets with intact ECM and junctions.
• •NIR-triggeredreversible surface switching enables spatiotemporal cell sheet detachment without enzymes or chemicals.
• •The resulting cell sheets are pre-primed, dense assemblies with electroactivity and mechanical cues for muscle reconstruction.
• •This pre-functionalized cell sheet therapy significantly enhances functional muscle regeneration in a rat VML model.

1. Introduction

Volumetric muscle loss (VML) refers to skeletal muscle injuries with extensive, large-area defects resulting from combat injuries, traumatic accidents, surgical resections, or degenerative diseases, where the intrinsic regenerative capacity of muscle is overwhelmed by the magnitude of tissue loss [[1], [2], [3]]. Pathologically, VML is characterized by substantial loss of muscle fibers and their supporting extracellular matrix (ECM), triggering a self-perpetuating cycle of secondary pathologies, including chronic inflammation, fibrotic scarring, and aberrant tissue remodeling, which collectively preclude functional recovery [4]. Although functioning free muscle transplantation (FFMT) remains the clinical gold standard for VML [5], its widespread application is constrained by donor-site morbidity, prolonged surgery, and unreliable graft re-innervation, often leading to suboptimal functional outcomes [6]. Consequently, extensive effort has been devoted to developing regenerative strategies ranging from cell suspension injections to ECM-mimetic scaffolds, aimed at addressing the dual challenges of achieving sufficient cell retention and promoting functional tissue integration at the injury site. However, these approaches have intrinsic limitations, specifically that injected cells rapidly disperse and exhibit poor retention. At the same time, conventional scaffolds fail to recapitulate the dynamic bioactivity and topographical guidance cues of the native muscle niche [7]. As a result, both established and emerging strategies yield, at best, modest functional improvements and fail to address the underlying pathophysiology initiated by VML, leaving a critical gap in effective rehabilitation [8].

To overcome the limitations of cell injection and conventional grafting, biomaterial-based strategies have been developed [9]. Acellular ECM scaffolds, in particular, have attracted intensifying interest because they simultaneously provide structural support and bioactive cues for host cell recruitment and tissue remodeling, thereby exhibiting considerable promise for soft-tissue regeneration [10]. Building on this concept, subsequent studies have seeded muscle progenitor or stem cells into such ECM scaffolds to couple cellular therapeutics with matrix-driven support [11]. Nevertheless, the regenerative efficacy of these ECM-based constructs remains inconsistent [7,12], as the engrafted cells still exhibit poor retention, limited long-term survival after implantation, and insufficient functional maturation. These limitations underscore that mere cellular seeding onto a static scaffold is insufficient to reconstitute a fully functional tissue unit.

To address these interconnected limitations, we posit that next-generation VML therapies must achieve both high-density cell delivery and preemptive construct functionality, enabling rapid functional initiation in vivo. Herein, we proposed a pre-priming cell sheet therapy driven by a dynamic wrinkled electroactive substrate [13] that establishes a biomimetic microenvironment to promote cellular integration, electroactivity, and mechanical stability, thereby achieving elevated cell density and enhanced bioactivities for muscle reconstruction. In contrast to conventional methods (e.g., cell injection or scaffold-based approaches), engineered cell sheets inherently increase cell density while preserving native cell–cell connections [14]; however, conventional harvesting techniques, such as enzymatic digestion or thermoresponsive surfaces, inevitably disrupt ECM integrity and reduce cell viability [12,13]. Our approach addresses this critical bottleneck: the dynamic substrate enables non-destructive, scaffold-free harvest, preserving the endogenous ECM and intercellular networks [14]. This resulting pre-primed, high-density cellular assembly, endowed with inherent electroactivity and mechanical cues, directly counteracts low cell retention and poor functional maturation, offering a more integrated and effective strategy for VML repair.

Hence, we developed a method for the non-invasive fabrication and pre-functional programming of cell sheets (Scheme 1). By leveraging the superior biocompatibility and spatiotemporal controllability of near-infrared (NIR) light, we engineered a novel bilayer construct comprising a carbon nanotube (CNT)-embedded polydimethylsiloxane (PDMS) elastomer that acts as an NIR-responsive actuator and an electroactive polypyrrole (Ppy) surface that interfaces directly with cells. This architecture enables reversible, NIR-driven modulation of the cell-substrate interface via thermomechanical coupling [[15], [16], [17]], enabling non-invasive sheet harvesting. Owing to the high photothermal conversion efficiency of the embedded CNTs, NIR irradiation rapidly converts photonic energy into localized heat, inducing a thermomechanical stress field that drives localized deformation of the bilayer [18]. The resulting dynamic shape change enables on-demand switching between cell-adhesive topographies that promote cell sheet formation and non-adhesive states that facilitate gentle detachment, while preserving cell-cell junctions and ECM integrity through purely physical stimuli [19]. This functional switching is driven by a NIR-triggered evolution in wrinkle morphology that rapidly modulates interfacial tension, significantly downregulating focal adhesion protein expression yet maintaining critical cell-cell connections, thereby enabling the harvest of intact cell sheets with preserved ECM architecture—a notable advance over conventional enzymatic or mechanical methods [12,15]. The resulting electroactive dynamic biointerface further exhibits enhanced biofunctionality, promoting electrophysiological maturation in primary cardiomyocytes and underscoring its utility for engineering electroexcitable tissues [16]. To validate the translational potential of this approach, we implanted the primed cell sheets into a rat VML model and observed significant muscle recovery through in situ tissue remodeling [2]. We believe that this robust strategy establishes a foundational framework for electroresponsive tissue repair, with direct relevance to VML rehabilitation and potential applicability to other regenerative contexts.

Scheme 1.

Schematic illustration of the NIR-responsive dynamic wrinkle platform for the non-invasive harvesting of pre-primed cell sheets and their application in volumetric muscle loss (VML) repair. (A) The process of obtaining and applying pre-conditioned cell sheets for VML repair. (B) NIR-triggered dynamic reconfiguration of the wrinkle topography remotely switches the interfacial adhesion state. (C) This reconfiguration alters cellular mechanotransduction and focal adhesion density, leading to cell sheet detachment when the interfacial mechanical force (Fm) surpasses the cell-substrate adhesion force (Fc). (D) Immunofluorescence staining of focal adhesion-related markers (integrin β1, talin, pFAK(Y397), paxillin, and YAP/TAZ) and cytoskeleton in cells under control and mechanical stimulation conditions.

2. Results and discussion

2.1. Development and photothermal responsive properties of NIR-light responsive dynamic wrinkled electroactive surfaces

To engineer a versatile platform for precise control of cell sheet adhesion, we developed a smart bilayer system capable of photothermally reconfiguring both surface and interfacial properties. As illustrated in Fig. S1A, the system was fabricated via in situ polymerization of a stiff Ppy layer onto a compliant PDMS elastomer substrate embedded with photothermal CNTs [18]. Ppy was selected as the electroactive top layer due to its well-established electrical conductivity, biocompatibility, and positive effects on cell adhesion and proliferation [17]. The polymerization process spontaneously generated hierarchical wrinkled structures across the PDMS/CNTs-Ppy surface, driven by the mismatch in mechanical modulus and thermal expansion coefficient between the soft underlying substrate and the stiff top Ppy film (Fig. 1A, S1B) [20]. Cross-sectional SEM characterization confirmed that the CNTs were uniformly distributed within the PDMS matrix and were not present as isolated structures (Fig. S1C), ensuring efficient photothermal conversion for dynamic wrinkle actuation. This integrated design combines the electroactive function of the continuous Ppy film with the photothermal efficiency of the uniformly dispersed CNTs, enabling both cellular electrical stimulation and on-demand topological switching.

Fig. 1.

The fabrication and photothermal responsive behavior of the NIR-responsive dynamic wrinkled PDMS/CNTs-PPy composites. (A) The process and principle of in-situ generation of Ppy wrinkle-patterned surface. (B) Conductive atomic force microscopy (C-AFM) current mapping of PDMS/CNTs (left) and PDMS/CNTs-Ppy (right) nanocomposites. (C) It presents AFM characterization of the PDMS/CNTs-PPy composites at different time points, showing the evolution of the surface morphology from smooth to a disordered labyrinthine wrinkle pattern through 2D topography, phase, and 3D morphology images. (D–F) The evolution of width, height, and surface roughness of Ppy film with polymerization time (t), respectively. (G) Hydrophobicity change before (left) and after (right) Ppy coating. (H) Schematic of surface morphology change under NIR ON/OFF cycling. (I) Temperature profiles under NIR irradiation in different media. (J) Temperature profiles in phosphate-buffered saline (PBS) under NIR irradiation, comparing PDMS/CNTs-Ppy, PDMS/CNTs, and pristine PDMS. (K) Temperature profiles for composites with different CNT content (0, 0.025, 0.05, 0.075, and 0.1%). (L) Temperature profiles at varying NIR power densities for the 0.05 wt% CNT composite. (M) Equilibrium temperatures at different power densities. (N) Temperature cycling stability test (10 min ON/OFF intervals) under ∼1.33 W cm⁻² NIR irradiation. Note: Unless specified (e.g., J-L), experiments used the 0.05 % CNT composite and ∼1.33 W cm⁻² NIR power density. Data are mean ± S.D. (n = 5).

The surface morphology of the films was characterized by atomic force microscopy (AFM). AFM characterization demonstrates a marked morphological transition from a smooth baseline to a complex wrinkled architecture. The pristine PDMS/CNTs substrate (0 h) exhibited minimal topographic variation, appearing uniformly smooth in 2D, with a flat height curve, and a near-planar 3D structure. Following Ppy polymerization, a disordered labyrinthine wrinkle pattern emerged. At 4 h, a dense network of wrinkles initiated, progressively increasing in amplitude and definition at 8 h and 24 h. This morphological evolution is quantitatively reflected in the expanding range of the height curves and visually confirmed by the 3D morphologies, which distinctly contrast the initial flat state with the final highly textured surface (Fig. 1C). To further investigate the formation and evolution of the surface wrinkling morphology during the polymerization of Ppy on PDMS/CNTs, we performed time-resolved AFM imaging of the as-prepared sample at various polymerization times (t) (Fig. S2) [17]. The surface wrinkles remained remarkably stable throughout the polymerization process. Their evolution revealed three distinct growth stages: an initial approximately linear growth phase (t < 1 h), a slow growth phase (1–8 h), and a saturated growth phase (t > 8 h). At t = 8 h, the film exhibited well-defined wrinkles with an average width of 0.88 (±0.06) μm (Fig. 1D). The evolution of wrinkle height with polymerization time followed a trend similar to that of wrinkle width (Fig. 1E). Notably, the Ppy polymerization reached a saturation stage after 8 h, a pattern closely resembling the wrinkle evolution observed in previously reported similar material systems [15]. Additionally, surface root-mean-square (RMS) roughness, calculated from AFM images [21], increased with t and plateaued at approximately 8 h (Fig. 1F). The parallel trends in RMS roughness and wrinkle dimensions (width and height) indicate that the Ppy polymerization reached saturation after 8 h, stabilizing both the film and the wrinkle pattern [15]. This supports the selecting of the 8 h time point for subsequent experiments. Wettability measurements revealed that both surfaces exhibited moderate hydrophobicity, with water contact angles in the range of 90–100° (Fig. 1G) [16]. Despite this hydrophobic character, the higher surface roughness of the wrinkled Ppy film, compared to the smooth PDMS/CNTs surface, is more conducive to promoting cell adhesion. This enhanced adhesion can be attributed to the greater number of contact points and microstructures on the rougher surface, which enlarges the effective cell-substrate contact area and facilitates cellular anchoring [21].

As efficient photosensitizers, CNTs can absorb NIR light and convert it into localized thermal energy. Upon 808 nm irradiation, the photothermal effect drives thermal expansion of the PDMS/CNTs substrate, which, in turn, imposes tensile strain on the overlying Ppy film, thereby dynamically deforming its wrinkled morphology (Movie S1). This topographic switch modulates interfacial adhesion, thereby initiating cell-sheet release. When NIR light is withdrawn, the strain is relieved, and the elastic mismatch between the two layers drives immediate recovery of the original wrinkle pattern. Consequently, simply switching the NIR light on or off alone switches the surface between two distinct topographic states (Fig. 1H) [18]. To quantify responsiveness and stability, we measured surface temperature changes over time during irradiation. As shown in Fig. 1I, in dry air, the surface temperature increased rapidly to approximately 91 °C within seconds and then stabilized. To further evaluate its performance under biologically relevant conditions, the as-prepared sample was immersed in phosphate-buffered saline (PBS) solution, and the temperature reached a stable plateau of ∼37 °C within 8 min. Crucially, cell sheet release is governed not by this final equilibrium temperature but by the transient thermal stress generated during the rapid, localized photothermal expansion of the CNT-embedded PDMS layer. This dynamic stress, which is absent during slow equilibrium heating in a 37 °C incubator, provides the mechanical impetus for interfacial restructuring, as visualized in Movie S1 and quantified by surface topography analysis. Control experiments confirmed that pure PDMS exhibited negligible photothermal response, while the temperature profile of PDMS/CNTs-Ppy is dominated almost exclusively by the CNTs; the Ppy film contributes negligibly to heat generation (Fig. 1J). While previous studies report notable photothermal effects from bulk or thick Ppy coatings [22,23], in our composite, Ppy was formed as an ultra-thin film (tens to hundreds of nanometers thick) to enable surface wrinkling and electroactivity. At this scale, its thermal contribution is masked by the efficient heat conduction of the CNTs/PDMS substrate. Thus, CNTs primarily drive photothermal wrinkling, while the thin Ppy layer provides a continuous conductive interface for cellular signaling. This design separates the photothermal and electrical functions, enabling combined thermal actuation and bio-electrical activity.

The aforementioned experiments demonstrate that PDMS/CNTs-Ppy immersed in PBS realizes stable, biocompatible photothermal heating (∼37 °C). To identify conditions that generate reversible light-responsive wrinkling without compromising cell sheet viability, we systematically investigated the effect of two key factors: the photothermal conversion efficiency (influenced by CNT concentration) and the NIR laser power intensity. We first examined the impact of different CNT concentrations (0%, 0.025%, 0.05%, 0.075%, and 0.1%) while keeping the laser intensity constant. As shown in Fig. 1K, samples across all CNT concentrations showed similar temperature-increase trends; only the 0.05% CNT formulation achieved a stable temperature closest to standard cell culture conditions (∼37 °C). Using this CNT concentration, we next explored the effect of different NIR laser power intensities (0.004, 1.328, 2.512, and 3.784 W cm⁻²) on surface temperature. As expected, higher laser power produced greater temperature increases; a power of 1.328 W⋅cm⁻² yielded a stable temperature of approximately 37 °C (Fig. 1M–N and S3). To evaluate the safety and stability of the photothermal platform under physiological conditions, samples were immersed in PBS maintained at 37 °C. Measurements from multiple samples showed that the temperature consistently stabilized within the safe range of 37–38 °C upon irradiation, demonstrating a mild, controllable, and self-stabilizing heating profile. This reliable, biocompatible thermal performance makes the platform particularly suitable for repeated use in harvesting intact cell sheets that require gentle thermal stimulation (Fig. S4). Consequently, 0.05% CNTs and 1.328 W⋅cm⁻² laser power were selected for all subsequent experiments. To achieve further practical application in the biomedical field, the thermal robustness under these optimized settings was verified by 20-min NIR on/off cycling. The material exhibited reproducible and reversible temperature amplitudes over five cycles, confirming reliable actuation for cell sheet harvesting (Fig. 1M). Besides, cells cultured on Ppy polymerized for 8 h showed higher detachment efficiency than at other polymerization times (Fig. S5), confirming the importance of a fully matured wrinkle architecture.

2.2. Versatile bioactivity gain of NIR light-responsive dynamic wrinkled electroactive surface for electroexcitable cells

Electroactive biomaterials have garnered significant interest for biomedical applications, particularly for establishing functional interfaces with electroexcitable cells and tissues [24]. Among these materials, Ppy is a representative conductive polymer, notable for its inherent conductivity [[25], [26], [27]]. This property accelerates electrical signal propagation and thereby promotes synchronous contraction of engineered myocardial tissue [28]. To evaluate the versatility of our NIR-light-responsive, dynamically wrinkled, electroactive platform in supporting electroexcitable cells, we first assessed the bioactivity of the Ppy-based surface. We confirmed that the physicochemical properties of Ppy support robust adhesion and growth of various cell types (Fig. S6). Notably, we observed enhanced adhesion and proliferation for electroexcitable cells. Building on this observation, we selected cell models with inherent electrical signalling capability for subsequent functional studies. These were L6 myoblasts, C2C12 myoblasts, and primary cardiomyocytes (CMs). Live/dead staining on days 1 and 3 revealed that C2C12 cells adhered and proliferated effectively on all substrates (Fig. 2A). Significantly higher cell densities were seen on PDMS/CNTs-Ppy and tissue culture plastic (TCP) compared to PDMS/CNTs alone. Quantitative analysis via a CCK-8 assay at days 1, 3, and 5 supported these findings (Fig. 2B). Collectively, these results demonstrate that the wrinkled electroactive surface (PDMS/CNTs-Ppy) offers a favourable microenvironment for the cultivation of dense, viable cell sheets.

Fig. 2.

The biocompatibility and functional characteristics of different surfaces. Live/dead staining (A) and CCK8 OD value (B) of C2C12 cells cultured on TCP, PDMS/CNTs, and PDMS/CNTs-Ppy. (C) Calcium transient of CMs on different surfaces at day 5 of culture. (D)Expression of cardiac-specific proteins of α-actinin (green) and CX-43 (red) in the CMs on non-electroactive PDMS/CNTs-PLA and electroactive PDMS/CNTs-Ppy. (E) and (F) Immunofluorescence intensity statistics were calculated based on CX-43 and α-actinin immunostaining images. (G)Heatmap of DEGs between PDMS/CNTs-Ppy and PDMS/CNTs-PLA after hierarchical cluster analysis (n = 3 rats per group). (H) PCA plot of proteomics data in these two groups. (I) GO terms enriched from up-regulation DEGs associated with C2C12 differentiation (PDMS/CNTs-Ppy versus PDMS/CNTs-PLA. (J) Volcano plot of protein expression in PDMS/CNTs-Ppy and PDMS/CNTs-PLA. (K) GO terms enriched from up-regulation DEGs associated with C2C12 differentiation (PDMS/CNTs-Ppy versus PDMS/CNTs-PLA). (L) Network of the muscle differentiation-related GO enriched from all DEGs with a fold change >2.0. The data were expressed as mean ± S.D. (n = 3).

To further reveal the bioactivity advantage conferred by conductivity, we used CMs with electrophysiological behaviour as a bioindicator and cultured them on our wrinkled electroactive PDMS/CNTs-Ppy surface [18]. As a reference, a non-electroactive control surface was prepared by coating non-electroactive PLA on the same substrate (PDMS/CNT), yielding a PDMS/CNTs-PLA surface. To further investigate the effect of the conductive microenvironment of PDMS/CNTs-Ppy on CMs' systolic synchrony, we recorded Ca²⁺ transients at three random fields per sample. These transients indirectly reflect the propagation of action potentials. The results demonstrated that PDMS/CNTs-Ppy supported strong, high-frequency, and synchronous calcium signal changes across CM clusters.

In contrast, PDMS/CNTs-PLA produced only weak, asynchronous Ca²⁺ puffs (Fig. 2C and Movie S2). On day 4 of culture, spontaneous contractions of CMs-seeded PDMS/CNTs-Ppy were visible under a microscope. By day 5, the contraction amplitude had markedly increased. This high-frequency, monolithic beating behaviour is mainly attributed to the formation of tightly connected, mature CMs on PDMS/CNTs-Ppy. The presence of conductive biomaterials, Ppy, endows PDMS/CNTs-Ppy with favourable electrical conductivity.

We hypothesized that the conductive PDMS/CNTs-Ppy surface would facilitate intercellular electrical signalling within cell sheets, thereby promoting charge transfer between cells. As shown in Fig. 1B, the PDMS/CNTs-Ppy surface exhibited a maximum current of 65.4 pA, which is approximately 11.7 times higher than that of the PDMS/CNTs surface (5.6 pA). This order-of-magnitude improvement quantitatively demonstrates that the Ppy layer significantly enhances local charge transport capability at the cell-material interface. This property facilitates the transmission of electrical signals between CMs and the establishment of synchronous contraction. Upon contact, electrons can traverse the Ppy films [26]. This film electrically couples adjacent cells and synchronizes their activity. After identical culture periods on PDMS/CNTs-PLA (non-conductive) and PDMS/CNTs-Ppy (conductive), we assessed CMs maturation via immunofluorescence staining for sarcomeric α-actinin and connexin 43 (CX43). α-Actinin is a key microfilament protein located on the fine myofilament of the contractile protein. It plays an essential regulatory role in muscle contraction and relaxation [29]. CX43 is a well-known gap junction protein between CMs. This protein forms the gap-junctional channels that mediate the propagation of electrical signals between cells [30]. As shown in Fig. 2D, a large number of sarcomeric structures could be observed in CMs cultured on PDMS/CNTs-Ppy. This indicates that these CMs on the electroactive surface exhibited mature contraction function, likely attributed to the presence of conductive materials. Statistical analyses presented in Fig. 2E and F revealed significant differences in the expression levels of CX43 and α-actinin between CMs cultured on PDMS/CNTs-PLA and PDMS/CNTs-Ppy, suggesting that the non-conductive PDMS/CNTs-PLA has a weaker ability to promote the maturation of CMs. These findings further validate the beneficial role of electrical conductivity in enhancing CM maturation. Collectively, the results indicate that the wrinkled electroactive PDMS/CNTs-Ppy enhances the bioactivity of cultured cells and tissue-like cell sheets. On the one hand, micro- and nanoscale wrinkle structures formed on the Ppy surface increase the effective surface area, providing more adhesion sites and mechanical support for CMs spreading. On the other hand, the PDMS/CNTs-Ppy also provides a conductive microenvironment for CMs and cell sheets, modulating electroactive molecular interactions and thereby reinforcing both intercellular communication and functional maturation.

To clarify how an electroactive cell culture platform facilitates cellular functionalization, including the enhancement of intercellular communication and the promotion of stem cell differentiation into skeletal muscle cells, we performed RNA-seq analysis of C2C12 cells. Three independent replicates were sequenced for the conductive wrinkle group (PDMS/CNTs-Ppy) and its non-conductive wrinkle counterpart (PDMS/CNTs-PLA). The heatmap of differentially expressed genes (DEGs) following clustering analysis revealed significant differences in gene expression between the two groups (Fig. 2G). Principal component analysis (PCA) indicated apparent clustering of the three replicates within each group. A total of 14,384 genes were identified based on the volcano plot, with 94 DEGs up-regulated and 31 DEGs down-regulated when comparing PDMS/CNTs-Ppy to PDMS/CNTs-PLA (Fig. 2H). To further understand the roles of these differentially expressed genes (DEGs), we conducted Gene Ontology (GO) enrichment analysis. According to the results, several GO terms highly correlated with skeletal muscle regeneration were identified, including “muscle system process,” “muscle contraction, " “muscle cell development,” “regulation of muscle system process,” “muscle cell differentiation,” “muscle tissue development,” and “response to muscle stretch” (Fig. 2I). These terms primarily relate to the development and functional maintenance of skeletal muscle, suggesting that these two groups effectively promote skeletal muscle regeneration in the presence of VML defects, consistent with histologic observations (Fig. 6). Comparative analysis of PDMS/CNTs-Ppy to PDMS/CNTs-PLA further revealed significant alterations in various muscle-related biological processes, cellular components, and molecular functions. Specifically, the biological processes (BPs) included “muscle system process,” “muscle contraction,” and “muscle cell development.” The cellular components (CCs) identified were “M band,” “A band,” “sarcomere,” “sarcoplasm,” “myofibril,” and “contractile fiber.” For molecular functions (MFs), the relevant activities included “calcium-release channel activity” and “ligand-gated calcium channel activity”. These findings further confirm the prominent role of the electroactive Ppy films in promoting muscle regeneration (Fig. 2K). Fig. 2L illustrates the network associated with the electrically active cell culture platform and C2C12 cell differentiation, encompassing pathways such as “muscle cell development,” “muscle cell differentiation,” “regulation of muscle system process,” “muscle contraction,” “muscle tissue development,” “response to mechanical stimulus,” and “action potential.” Collectively, the transcriptomic data demonstrate that the conductive Ppy film primes C2C12 cell for robust myogenesis, underscoring its promise for VML tissue repair. Furthermore, considering the directional alignment of myofibers is beneficial for inducing muscle repair and maturation [31], we constructed the pre-patterned PDMS/CNTs-Ppy substrates with varying microgroove widths (20 μm, 50 μm, 100 μm and 200 μm) to explore the potential cellular alignment effect on the functionality of cell sheets. Our systematic evaluation using micro-grooved PDMS/CNTs-Ppy substrates confirmed that only grooves with 20 μm and 50 μm width exhibited effectively cell alignment for different adherent cells, e.g., C2C12, L6, HUVECs), in contrast to (Figs. S7 and S8) [32]. These results indicate that pre-patterned microstructural cues may serve as an important factor in enhancing the functional performance of engineered cell sheets.

Fig. 6.

Skeletal muscle regeneration in a rat model of VML after implantation of the cell sheets for 1 week and 4 weeks. (A) Schematic illustration of cell sheets repairing the VML defects in vivo. Sham and injury were used as the control groups. Skeletal muscle regeneration in a rat model of VML after implantation for 1 week, representative histological evaluation by H&E (B) and Masson staining (D), and immunofluorescence of anti-MHC-Ⅱ (F). Skeletal muscle regeneration in a rat model of VML after implantation for 4 weeks, representative histological evaluation by H&E (C) and Masson staining (E), and immunofluorescence of anti-MHC-Ⅱ (G). (H) shows the statistical analysis of the TA wet weight rate, (I) the number of centronucleated myofibers (new myofibers), and (J) the perimeter of myofibers around the defective area in week 1. (K) presents the statistical analysis of the TA wet weight rate (L), the number of centronucleated myofibers (new myofibers), and (M) the perimeter of myofibers around the defective area in week 4. Data were expressed as mean ± S.D. (n = 6).

In summary, the dynamically wrinkled electroactive PDMS/CNTs-Ppy surface shows significant potential to enhance cellular functionalization and promote muscle regeneration. This work highlights the favourable physicochemical properties of Ppy that facilitate the adhesion, proliferation, and maturation of electroexcitable cells, including CMs and C2C12 myoblasts. RNA-seq analysis revealed significant differences in gene expression associated with muscle regeneration, underscoring the electroactive platform's ability to influence key biological processes. Notably, the Ppy-CNT network facilitated intercellular electrical signalling and synchronous contraction of CMs, as evidenced by enhanced calcium transient responses and immunofluorescence staining for sarcomeric structures. These data demonstrate that the PDMS/CNTs-Ppy platform not only supports muscle cell growth and maturation but also promotes coordinated cell-cell communication, offering a promising strategy for engineering functional muscle constructs and treating VML [33].

2.3. The damage-free cell sheets harvesting based on tunable adhesion and de-adhesion of NIR light-responsive dynamic wrinkled electroactive surfaces

Cells adhere to the matrix substrate via an integrin-mediated adhesion mechanism, which comprises both passive and active adhesion processes [34]. Active site activation induces cell recruitment and aggregation of more adhesion proteins, and the greater the traction force between adhesion proteins and substrate, the stronger the cell interface force. Cells sense chemical, mechanical, and topological cues of the surrounding environment through cell-substrate adhesion sites known as focal adhesions (FAs) [35] and thus decide whether to translocate or change their morphodynamics [36]. In FAs, integrins are the primary adhesion receptors to bind ECM proteins and connect to the actin cytoskeleton through a large number of FAs clumping proteins. The appearance of FAs is dependent on the tension mechanically exerted by contractile myosin [37]. Talin, which links integrins to the actin cytoskeleton, is also necessary for the re-recruitment of neuregulin to the FAs [38]. Vinculin is ideally positioned to coordinate force-induced signaling [39], thereby regulating the recruitment and release of core focal adhesion proteins in a force-dependent manner. Therefore, the interaction of cell adhesion proteins with microenvironmental ligands is pivotal for cytomechanical sensing [39]. By modulating the force between the cell and the substrate, affecting vinculin expression, reducing the number of focal adhesions, and altering the state of cell adhesion, it's possible to obtain intact cell sheets. To investigate the cellular behavior of on-demand cell adhesion and detachment and to reveal the underlying molecular mechanisms during manipulation of tissue-like constructs in our dynamic wrinkle cell culture system, we subjected cells adhering to PDMS/CNTs-Ppy surfaces to NIR light irradiation. Subsequently, we performed corresponding tests to assess cell viability, morphology, adhesion status, and intracellular DNA damage, accordingly.

As shown in Fig. 3A, cells displayed excellent viability on the PDMS/CNTs-Ppy surface both before and after NIR irradiation, confirming good biocompatibility of the material. Interestingly, cell detachment occurred exclusively within the NIR-irradiated spots, indicating that NIR light plays a dominant role in driving localized cell release. Dynamic micro/nanowrinkles provide an effective approach for on-demand tuning of surface adhesion properties, enabling the realization of a smart surface [40]. On this basis, cells experience a traction force generated by NIR-induced deformation of the Ppy film. Under this force, the cell-substrate interfacial stress increases, the integrin-ECM connection is disrupted, vinculin expression decreases, and focal adhesions disassemble, thereby causing cells to detach precisely from the irradiated region of the Ppy films' surface.

Fig. 3.

Biological activity and adhesion status of cells on PDMS/CNTs-Ppy before and after NIR irradiation. (A) Live/dead assay of C2C12 cells before (up) and after (down) cell detachment (green, live cells; red, dead cells). (B) Changes in cell morphology before (left) and after (right) NIR irradiation by SEM. (C) Immunofluorescence staining of F-actin (red) and vinculin (green). (D) Immunofluorescence staining of γ-H2AX (green). (E) Fluorescence intensity statistics were calculated based on Vinculin immunofluorescence images. (F) Fluorescence intensity statistics were calculated based on γ-H2AX immunofluorescence images. (G) Heatmap of DEGs between PDMS/CNTs-Ppy and TCP-Ppy after hierarchical cluster analysis. (H) PCA plot of proteomics data in these two groups. (I) Volcano plot of protein expression in PDMS/CNTs-Ppy and TCP-Ppy. (J) GO terms enriched from up-regulation DEGs associated with C2C12 differentiation (PDMS/CNTs-Ppy versus TCP-Ppy). (K) GO terms enriched from up-regulation DEGs associated with C2C12 adhesion and differentiation (PDMS/CNTs-Ppy versus TCP-Ppy). (L) GO terms enriched from down-regulation DEGs associated with C2C12 differentiation (PDMS/CNTs-Ppy versus TCP-Ppy). (M) Network of the cell adhesion-related GO enriched from all DEGs with a fold change>2.0. The data were expressed as mean ± S.D. (n = 3); ns means no significance.

Subsequently, to elucidate the mechanism underlying cell detachment, we further charactered the surface morphology of PDMS/CNTs-Ppy before and after NIR illumination. SEM images revealed that C2C12 cells on PDMS/CNTs-Ppy surface exhibited significantly greater adhesion and more extensive spreading than those on PDMS/CNTs. This enhancement is attributed to the crumpled structure of Ppy, which provides a larger surface area for initial cell attachment and thereby promotes cell adhesion, spreading, and proliferation. After NIR stimulation, the Ppy film remained structurally intact, but the cells completely detached. In contrast, although cells remained adhered to the PDMS/CNTs surface, their number decreased slightly (Fig. 3B).

At cell-substrate adhesion sites, talin is subject to tensile forces and exposes cryptic vinculin-binding sites. The stretching of talin induces conformational changes of vinculin, which in turn reinforces F-actin anchorage and enables the formation of additional linkages between integrins and the actin cytoskeleton. The ultimate consequence is increased integrin clustering and FA maturation [41]. Therefore, vinculin promotes traction force generation by stabilizing integrin connections to the actin cytoskeleton. To characterize morphological changes of cells during the adhesion-to-detachment transition, immunofluorescence staining for F-actin and vinculin was performed in C2C12 cells (Fig. 3C). After culturing the cells on the PDMS/CNTs and PDMS/CNTs-Ppy surfaces for 3 days, vinculin expression was significantly higher on the PDMS/CNTs-Ppy surface than on the PDMS/CNTs surface, consistent with more robust adhesion on the conductive substrate. Stable cell adhesion is a prerequisite for sustained cell growth.

To clarify the active cellular program behind NIR-triggered detachment, we systematically analyzed the dynamics of focal adhesion and mechanotransduction components. After NIR stimulation, immunofluorescence showed a coordinated molecular response. The mechanosensors integrin β1 and talin formed aggregates at the cell periphery, suggesting heightened mechanical sensing of the substrate's deformation (Fig. 4A and B). Total FAK levels increased, but the phosphorylation of its active form, pFAK, decreased markedly (Fig. 4C). This drop in pFAK suggests that signals that maintain focal adhesion stability are being actively shut down. The scaffold protein paxillin was also recruited to the cell periphery (Fig. 4D).

Fig. 4.

Immunofluorescence staining of focal adhesion-related markers and cytoskeleton in cells under control and mechanical stimulation conditions. Left (Control group): Immunofluorescence staining shows focal adhesion-associated proteins (green channels). These correspond to integrin β1(A), talin (B), pFAK/FAK (C), paxillin (D), YAP (E), and TAZ (F) in each row. F-actin and FAK are shown in the red channel. The rightmost column of each row displays merged images. These integrate focal adhesion marker signals (green), F-actin (red), and DAPI-stained nuclei (blue). Yellow indicates co-localization of focal adhesion markers and F-actin. Right (After mechanical stimulation group): Immunofluorescence staining displays the same set of focal adhesion-related proteins and F-actin in cells after mechanical stimulation. Merged images are presented in the same format as the control group.

Following these molecular changes, vinculin expression decreased after NIR stimulation, coinciding with a reduction in the number of FAs and facilitating cell detachment (Fig. 3C) [39]. The downregulation of vinculin weakened the integrin-mediated linkage to the extracellular matrix. Together, the subsequent downregulation of pFAK and the recruitment of other components (integrin β1, talin, paxillin) describe a tightly regulated process that unfolds after NIR stimulation (Fig. 4D). Upon mechanical signal detection, cells enhance mechanosensing and cytoskeletal mobilization, then actively disassemble adhesion-stabilizing signals, thereby advancing detachment.

We observed a distinct pattern in the core mechanosensitive transcriptional coactivators YAP and TAZ linked to dynamic destabilization signals. After NIR stimulation, TAZ moved into the nucleus, and YAP accumulated near the nucleus (Fig. 4E and F). This differs from the classic model in which sustained tension activates YAP/TAZ. We hypothesize this is the cell's reaction to a specific cue, anchorage destabilization, as focal adhesion stability is compromised by reduced pFAK signaling. Cells may control YAP/TAZ localization to initiate transcriptional or non-transcriptional processes linked to morphological remodeling and preparation for programmed detachment. In this way, they actively exit the stable adhesion state.

Together, these molecular events shift the force balance so that cell-driven contractile forces (Fm) overcome substrate adhesion (Fc). With continuous NIR stimulation, the crumpled Ppy layer sends stronger mechanical forces. The cell responds by weakening adhesion, disrupting force transmission, and changing shape. This response drives detachment from the PDMS/CNTs-Ppy surface [41]. This molecular process explains the efficient, intact harvesting of cell sheets.

γ-H2AX is a DNA damage marker that becomes phosphorylated upon DNA double-strand breaks [42]. To investigate whether the cell-sheet-harvesting protocol induces DNA damage, we used γ-H2AX staining to label double-stranded DNA molecules damaged by DNA double-strand breaks. We compared the results with those obtained using the conventional cell-digestion method with pancreatic enzymes. Notably, no significant difference was observed (Fig. 3D–F), confirming that the sheet-harvesting method does not inflict DNA damage.

To elucidate the role of a wrinkled cell culture platform in facilitating cell adhesion, this study employed RNA-seq analysis using C2C12 cells. Three independent sequencing replicates were performed for the conductive wrinkle group (PDMS/CNTs-Ppy) and the conductive smooth group (TCP-Ppy). The heatmap of DEGs following clustering analysis revealed significant differences in gene expression between the two groups (Fig. 3G). PCA indicated apparent clustering of the three replicates within each group (Fig. 3H). A total of 17590 genes were identified based on the volcano plot, with 1543 DEGs up-regulated and 517 DEGs down-regulated when comparing PDMS/CNTs-Ppy to PDMS/CNTs-PLA (Fig. 3I). To further understand the roles of these DEGs, GO enrichment analysis was performed. This analysis identified several GO terms highly correlated with C2C12 adhesion, including " cell-substrate adhesion,” " extracellular matrix organization,” " extracellular structure organization,” and " positive regulation of response to external stimulus " for BPs; “collagen-containing extracellular matrix,” “myofibril,” “sarcomere,” “sarcolemma,” “I band,” “collagen trimer,” “Z disc,” “receptor complex,” and “stress fiber” for CCs; and “extracellular matrix structural constituent,” “growth factor binding,” “cell adhesion molecule binding,” “integrin binding,” “actin binding,” and “extracellular matrix binding,” for MFs (Fig. 3K). These terms primarily relate to mechanisms underlying cell adhesion, extracellular matrix interactions, and the structural integrity of muscle tissue, suggesting they play a critical role in enhancing the regenerative potential and functional properties of C2C12 cells during muscle repair and regeneration. Comparative analysis of PDMS/CNTs-Ppy to TCP-Ppy revealed significant alterations in multiple muscle-related biological processes, cellular components, and molecular functions (Fig. 3J and L), further confirming the prominent role of the wrinkled cell culture platform in promoting cell adhesion and muscle regeneration. Fig. 3M illustrates the network associated with the wrinkled cell culture platform and C2C12 cell adhesion, encompassing pathways such as “focal adhesion,” “ECM-receptor interaction,” “P13K-Akt signaling pathway,” “Rap1 signaling pathway,” “Regulation of actin cytoskeleton,” “MAPK signaling pathway,” and “calcium signaling pathway.” In conclusion, the RNA-seq analysis demonstrated that the wrinkled cell culture platform (PDMS/CNTs-Ppy) significantly enhances cell adhesion and influences various muscle-related biological processes in C2C12 cells. The identified differentially expressed genes and associated signaling pathways underscore the platform's potential to improve regenerative outcomes in muscle repair by facilitating extracellular matrix interactions and cell signaling.

Based on the facile operation and demonstrated controllability of our NIR light-responsive, dynamically wrinkled, electroactive surface, cell sheets can be harvested remotely and non-invasively under NIR light stimulation (Fig. S9). As illustrated in Fig. 5A, cells cultured on PDMS/CNTs-Ppy were obtained through two routes: trypsin digestion to generate a cell suspension or NIR light stimulation to release an intact cell sheet. The cell suspension obtained after trypsin digestion lost cell-cell junctions, resulting in cells becoming round and displaying condensed nuclei (Fig. 5B and C).

Fig. 5.

High integrity of cell sheets acquired by the NIR response dynamic wrinkle system. (A) Schematic illustrating the process of obtaining cell sheets, controlled remotely with on-off NIR light. (B) Immunofluorescence images of F-actin (red) and vinculin (green) in C2C12 and L6 cells. (C) Immunofluorescence images of C2C12 and L6 cell suspensions. (D) Bulk cell sheets obtained from the PDMS/CNTs-Ppy surface. (E) Viability test of cells transferred onto PDMS/CNTs-Ppy after transfer from the PDMS/CNTs-Ppy surface: Calcein-AM (green, live), EthD-1 (red, dead). (F) Immunofluorescence images of F-actin (red) and vinculin (green) in C2C12 and L6 cells before cell sheet detachment. (G) Corresponding images after detachment. (H) Immunofluorescence images of F-actin (red) and fibronectin (green) in C2C12 cell sheets. (I) Immunofluorescence images of ZO-1 (green) in C2C12 cell sheets.

Benefiting from robust interfacial adhesion, we achieved gentle, intact harvesting of cell sheets. The PDMS/CNTs surface is hydrophobic and forms strong bonds with Ppy through in-situ polymerization (Fig. S10A). Micro-scratch tests and SEM morphology verified this after cyclic use. The Ppy layer remained continuous, intact, and tightly adherent throughout the dynamic wrinkling process (Fig. S10B). As a result, the cell sheets obtained by NIR stimulation were smooth and undamaged (Fig. S11–12). No visible cellular residue was left on the PDMS/CNTs-Ppy surface, and the harvested cells maintained high viability (Figs. 3B, 5E, and S13). These results demonstrate that robust interfacial integration between Ppy and the substrate provides the fundamental material basis for reversible, stable photothermal-driven topographic switching. Ultimately, this enables the programmed detachment of high-quality cell sheets (Fig. 5E).

The crucial feature of this technology is the retention of intact cell-cell junctions and deposited ECM on the cell sheet (Movie S3) [43]. Previous studies have identified vinculin as an essential biomarker for cell-cell adherens junctions, and specific deletion of vinculin from these junctions (while leaving its functions at cell-matrix adhesions unperturbed) reduces cell-cell adhesion and results in a loss of E-cadherin from the cell surface [41]. As presented in Fig. 5F and G, immunostaining of the C2C12 and L6 cell sheets for F-actin and vinculin confirmed that cytoskeletal structure and cell-to-cell, cell-to-ECM adhesion remained fully intact after detachment (Fig. 5H and I). We attribute the preservation of intact cell sheets to the exclusive application of mechanical forces at the cell-matrix interface, with a magnitude just sufficient to break the bonds between cells and the matrix without affecting cell-cell junctions. Because this purely physical interaction (mechanical force) has less interference with cells than chemical processes such as enzymatic digestion or ion chelation [44], the structural integrity of the cell sheet is maintained throughout the cell release process.

2.4. In-vivo therapeutic potential of cell sheets prepared by dynamic wrinkled electroactive surface

To further evaluate the in vivo therapeutic potential of the cell sheet constructs fabricated on our dynamic wrinkled electroactive surface platform, we established a VML defect model in the rat TA muscle [45,46]. Notably, the cell sheets, though extremely thin (∼20 μm) and flexible, demonstrated robust handling integrity—when gently manipulated, they elastically returned to their original shape without damage (Fig. S14). This confirmed their suitability for surgical transfer. Building on this, these intact L6 cell sheets were transplanted into the TA muscle defect. To assess outcomes, muscle regeneration was evaluated through histological staining and functional tests at 1 and 4 weeks post-transplantation (Fig. 6A).

One week after surgery, hematoxylin-eosin (H&E) staining revealed a large amount of immune cell infiltration (dark purple) in both the cell-suspension and cell-sheets groups. In contrast, in the VML-operated group, the immune cells were scattered in typical fibrotic tissues (Fig. 6B and S15). Masson's trichrome staining demonstrated pronounced fibrotic deposition in the injury control group, which was significantly reduced in the cell sheets group and moderately improved in the suspension group (Fig. 6D and F). High-magnification images identified centronucleated myofibers in all groups, and quantitative analysis demonstrated that the cell sheet groups significantly increased nascent myofiber density compared to the injury control (Fig. 6F and I), the perimeter of these nascent myofibers remained significantly lower than that in the uninjured control group (Fig. 6J). Analysis of the muscle wet-weight ratio further showed that the cell-sheet group's value was closer to that of the sham group than the cell-suspension group (Fig. 6H and Table S1), indicating its superior ability to preserve tissue mass and volume at the early stage.

By 4 weeks, the cell sheets group displayed robust tissue reconstruction composed of densely packed, mature myofibers that closely resembled sham muscle architecture (Fig. 6C and E). In contrast, the injury control and suspension groups showed sparse, disorganized myofibers with residual fibrosis (Fig. 6E). Quantification of centronucleated myofibers confirmed superior regeneration in the cell sheets group (Fig. 6L). MHC-II immunofluorescence revealed a significant increase in fast-twitch myofiber density in the cell sheets group compared with both suspension and injury control groups, indicating enhanced functional maturation (Fig. 6G). Although newly formed myofibers were also present in the defect areas of the cell suspension group and the VML group, their perimeter and number remained significantly lower than in the cell sheets group (Fig. 6M). The final wet-weight ratio results were consistent with the histological outcomes, with the cell-sheets group again showing values closest to the sham group (Fig. 6K and Table S1), which macroscopically confirmed its exceptional repair efficacy.

These results demonstrated that the pre-primed cell sheets significantly enhance skeletal muscle regeneration in vivo. Pre-priming L6 cell sheets outperforms cell suspensions in VML repair due to their improved structural and priming properties. These sheets maintain strong intercellular connections, which facilitate better cell communication and coordinated repair processes. The structural integrity of the sheets provides a stable scaffold, promoting efficient tissue reconstruction. Moreover, partially differentiated cells within the sheets integrate more rapidly into the damaged tissue, accelerating the healing process. Additionally, these cell sheets can continuously release bioactive factors, which are essential for stimulating cell proliferation and improving regenerative outcomes.

2.5. Cell sheets promote skeletal muscle injury repair by enhancing angiogenesis

Vascular regeneration is pivotal for repairing skeletal muscle defects, as the newly formed vascular system can restore blood flow, deliver oxygen, and mitigate damage following VML [47]. To assess vascular formation, we performed immunofluorescent staining for von Willebrand factor (vWF, indicative of microvasculature) and α-smooth muscle actin (α-SMA, indicative of arterioles) at 1 and 4 weeks post-operation.

In the early repair phase (week 1), the cell sheet group exhibited a significantly higher density of both vWF⁺ microvessels and α-SMA⁺ small arteries compared to the injury and cell suspension groups (Fig. 7A and F). This accelerated vascularization can be attributed to the unique pro-angiogenic paracrine activity of the cell sheets. To directly elucidate this mechanism, we performed in vitro assays using conditioned medium from cell sheets cultured on the electroactive PDMS/CNTs-Ppy platform. This medium robustly promoted human umbilical vein endothelial cell (HUVEC) migration (Fig. 7E) in scratch wound healing and Transwell assays (Fig. 7C), and significantly enhanced tube formation in a Matrigel-based angiogenesis assay (Fig. 7D). These results confirm that the electroactive microenvironment promotes the secretion of soluble factors from cell sheets, which actively stimulate endothelial cell recruitment and vascular morphogenesis.

Fig. 7.

(A, B) Immunofluorescence staining of anti-vWF (red) and anti-α-SMA (green) in skeletal muscle injury areas at 1 week (A) and 4 weeks (B) post-implantation of cell sheets. (C) Transwell migration assay of human umbilical vein endothelial cells (HUVECs). (D) Tube formation assay of HUVECs on Matrigel matrix. (E) Cell scratch healing assay for different material groups. (F, G) Statistical analysis of α-SMA and vWF counts at 1 week (F) and 4 weeks (G) post-implantation, based on fluorescence. Data are presented as mean ± S.D. (n = 5).

By week 4, as tissue repair progressed, the density of new vessels decreased to similar levels across all groups (Fig. 7B and G), indicating resolution of the initial angiogenic burst and a transition to a more stable tissue state. The temporal enhancement of early vascularization by the cell sheets is critically important: it ensures rapid re-establishment of nutrient and oxygen supply, facilitates the clearance of tissue debris, and supports the inflammatory-to-regenerative transition within the defect microenvironment. This robust and timely vascular support, driven by the cell sheet's paracrine function, creates a conducive foundation for effective muscle fiber regeneration and functional recovery, thereby explaining the superior therapeutic outcomes observed with cell sheet implantation.

2.6. Mechanisms of skeletal muscle repair and evaluation of inflammation

To further assess the inflammatory status and explore the mechanisms underlying skeletal muscle repair, we performed immunofluorescence staining for the macrophage markers CD86 (M1) and CD206 (M2) at 1 and 4 weeks post-operation. The results indicated a dynamic shift in macrophage polarization during the repair process. In the early phase (1 week), the injured group showed increased numbers of both M1 and M2 macrophages compared with the sham group, accompanied by tissue disorganization and diffuse inflammatory infiltration, reflecting a compensatory response to acute trauma (Fig. 8A and D). By the late phase (4 weeks), the injury group showed a marked decline in both macrophage subsets alongside persistent chronic inflammation. In contrast, cell-based treatments—especially the cell-sheet group—effectively counteracted this trend (Fig. 8B and E). Notably, the cell-sheet group demonstrated a significantly higher M2/M1 ratio than both the injury group and the cell-suspension group at both time points [48], suggesting that the preserved cell-cell and cell-ECM interactions within the sheets promoted a more pronounced and sustained M2-polarizing microenvironment [35].

Fig. 8.

Immunohistochemical staining of anti-CD86 and anti-CD206 in skeletal muscle tissue. (A) CD86 (red) and CD206 (green) expression in skeletal muscle 1 week after operation. (B) CD86 (red) and CD206 (green) expression in skeletal muscle 4 weeks after operation. (C) Immunofluorescence staining of in vitro macrophage polarization. (D) Rates of CD206-positive M2 and CD86-positive M1 macrophage cells 1 week after cell sheet implantation. (E) Rate of CD206-positive M2 macrophage cells 4 weeks after cell sheet implantation. Data are mean ± S.D. (n = 6).

To elucidate the molecular basis of this immunomodulatory effect, RNA sequencing was performed. To specifically decouple the effect of electrical conductivity, we compared the transcriptional profiles of tissues surrounding smooth conductive versus smooth non-conductive substrates. The data revealed that the electroactive substrate (smooth_con group) activated pathways involved in positive regulation of wound healing while suppressing pro-inflammatory signaling, thereby fostering a “low-inflammation and pro-repair” transcriptional profile (Fig. S16). This pattern aligns with the observed promotion of M2 macrophage polarization. In support of this mechanism, we performed an in vitro assay in which RAW264.7 macrophages were treated with conditioned medium from cell sheets cultured on the electroactive wrinkled PDMS/CNTs-Ppy substrate. Immunofluorescence quantification confirmed a significant increase in CD206 (M2) intensity and a decrease in CD86 (M1) intensity (Fig. 8C), providing direct evidence that paracrine signals derived from material-supported cells actively drive macrophages toward an anti-inflammatory phenotype.

Collectively, these findings establish a multi-level evidence chain demonstrating that the electroactive substrate facilitates skeletal muscle repair by orchestrating a favourable immune microenvironment. Transcriptomic profiling reveals its ability to induce a pro-repair, anti-inflammatory molecular program; in vitro functional validation directly confirms that the substrate, via cellular paracrine signaling, promotes M2 polarization; and in vivo phenotypic data ultimately verify that this modulation translates into improved tissue repair outcomes. The cell-sheet construct, by preserving native cellular interactions and ECM architecture, further amplifies this effect, leading to sustained suppression of chronic inflammation and enhanced tissue regeneration [49].

3. Conclusion

In this study, we developed a pre-priming cell sheet therapy for VML that exploits an NIR-responsive, dynamic-wrinkling electroactive substrate (PDMS/CNTs-Ppy). This biomimetic platform enables non-invasive harvesting of high-integrity cell sheets through programmed focal adhesion disassembly triggered by substrate deformation. Mechanistically, NIR-induced wrinkle reconfiguration initiates a coordinated cellular response: enhanced peripheral recruitment of mechanosensors (integrin β1 and talin) is followed by decreased active pFAK and redistribution of paxillin, indicating destabilization of mature focal adhesions. This precisely regulated force-sensing and adhesion-remodeling process allows directional traction forces to weaken cell-matrix adhesion while preserving intercellular junctions selectively. Furthermore, the conductive, wrinkled surface delivers combined electromechanical stimulation via calcium-mediated signaling, significantly enhancing myogenic differentiation and electrophysiological maturation of C2C12 myoblasts, thereby achieving cellular pre-activation.

In a rat VML model, transplantation of L6-derived sheets demonstrated three synergistic repair mechanisms: the retained intercellular connections and paracrine function of the sheet attenuated inflammation and promoted angiogenesis. Specifically, in vitro assays confirmed that conditioned medium from these sheets significantly enhanced endothelial cell migration and tube formation, and promoted macrophage polarization toward the M2 phenotype, which aligns with the observed in vivo outcomes of increased vascular density and elevated M2 macrophage infiltration. The pre-primed cells enhanced the activation of skeletal muscle regeneration pathways, and the electroactive microenvironment guided macrophage polarization and tissue remodeling (Fig. 9). Collectively, these effects accelerated recovery after muscle injury.

Fig. 9.

Schematic diagram showing the mechanism of VML tissue repair. L6 cell sheets obtained from the PDMS/CNTs-Ppy conductive wrinkled cell culture platform are transplanted into the VML site for skeletal muscle repair. The microstructured, electroactive substrate pre-primes the L6 cells, while preserving cell-cell connections, cell-ECM interactions, and a high-density, high-activity cell sheet. Consequently, the L6 cell sheets promote myogenic differentiation, maintain electromechanical coupling, reduce inflammation, and enhance neovascularisation, enabling rapid functional repair at the injury site.

4. Experimental section

Materials: Pyrrole was obtained from Shanghai Macklin Biochemical Co. 184 Silicone Elastomer Kit (PDMS) was purchased from Dow (USA). CNT (Tanfeng Tech Inc.; 20–40 nm diameter, 10–30 μm length) Ltd. Dulbecco's modified Eagle's medium (DMEM)/high-glucose medium, fetal bovine serum (FBS), and trypsin were purchased from Gibco (USA). The following reagents were obtained from Invitrogen: Alexa Fluor-568 and Alexa Fluor-488 donkey anti-rabbit IgG (H&L), Alexa Fluor-488 and Alexa Fluor-568 donkey anti-mouse IgG (H&L), and F-actin dye. The primary antibodies against α-actinin (ab9465), connexin 43 (CX-43, ab11370), Anti-Fast Myosin Skeletal Heavy chain (MHC-Ⅱ,ab91506), CD86 (ab220188), CD206 (ab64693), vinculin (ab129002), ZO-1 (ab221547), N-cardherin (ab18203), alpha-smooth muscle actin (α-SMA, ab5694), and von Willebrand factor (vWF, ab6994) were ordered from Abcam. FAK Antibody (sc-271126), pFAK(Y3978556s), talin (sc: 4021s), paxillin (sc: 365379), integrin β1 (sc: 374429), and YAP (cst: 14074s), TAZ (cst: 83669s) were ordered from Santa Cruz Biotechnology.

Fabrication of the PDMS/CNTs substrate: A polydimethylsiloxane (PDMS) substrate was prepared by combining the base and curing agent (Sylgard 184, Dow Corning) at a predetermined weight ratio of 10:1. In the first step, carbon nanotubes (CNTs) were dispersed in the curing agent and subjected to ultrasonic dispersion for 2 h. The mixture was then filtered using a 70 μm filter. Next, the base and curing agent were mixed at a 10:1 wt ratio, with CNTs accounting for 0.05% of the total PDMS mass. The PDMS mixture was degassed under vacuum to eliminate any bubbles. Finally, the mixed base and curing agent were poured into a culture dish and baked at 80 °C for 2 h to obtain an elastic PDMS/CNTs substrate.

Growth of Ppy films on the PDMS/CNTs substrate with in situ self-reinforcing surface wrinkling: Anhydrous ferric chloride and pyrrole monomer were separately dispersed in a 1 M hydrochloric acid solution to obtain a mixture of pyrrole (67 μm) and 1 M HCl (referred to as A). Similarly, a mixed solution consisting of FeCl₃ (0.3244 g) and 1 M HCl (10 mL) was prepared (referred to as B). Then, the PDMS/CNT substrates were immersed in a 1:1 mixture of A and B, resulting in final concentrations of 0.05 M for A and 0.1 M for B. After the designed reaction time (T) at 4 °C, the PDMS/CNTs-Ppy samples were rinsed with water and air-dried.

Surface Topography and Roughness Analysis by Atomic Force Microscopy (AFM): Surface topography of the PDMS/CNTs-Ppy composites was analyzed using an MFP-3D Origin™ Atomic Force Microscope. Measurements were performed in tapping mode under ambient conditions with standard silicon nitride cantilevers. Multiple areas of each sample were scanned at an appropriate rate to ensure representative surface imaging. For composites polymerized for different durations, comprehensive datasets—including 2D topographic images, height profiles, and 3D-reconstructed morphologies—were obtained. This allowed for quantitative evaluation of surface roughness parameters and systematic tracking of the structural evolution of Ppy-based wrinkled patterns.

Macroscopic Surface and Wrinkle Characterization Using 3D Super-Depth-of-Field Microscopy: Macroscopic surface examination was performed with an OLYMPUS DSX1000 optical microscope equipped with a high-magnification zoom lens. Samples were observed under optimized bright-field illumination. The integrated depth-of-field fusion function was employed to acquire clearly focused, extended-depth images over large viewing areas. This enabled detailed analysis of the distribution characteristics and microscale dimensional parameters of wrinkles.

High-Resolution Surface Imaging via Scanning Electron Microscopy (SEM): High-resolution surface imaging was conducted using a field-emission scanning electron microscope. Before imaging, samples were sputter-coated to ensure sufficient surface conductivity. Images were captured under optimized operating conditions at various magnifications to characterize fine structural details of the wrinkled surfaces, including morphological features and topological continuity.

Effect of Medium on Photothermal Performance: To investigate the influence of the surrounding medium on the photothermal response, the PDMS/CNTs-Ppy sample was placed separately in air and in phosphate-buffered saline (PBS) solution. Both setups were subjected to NIR irradiation at a power density of 1.328 W cm⁻². The temperature was recorded at 1-min intervals over a 10-min irradiation period.

Comparison of Photothermal Performance among Different Composites: The photothermal efficiency of different material groups—pure PDMS, PDMS/CNTs, and PDMS/CNTs-Ppy—was evaluated. All samples containing 0.05 wt% carbon nanotube (CNT) were immersed in PBS and exposed to NIR irradiation at a power density of 1.328 W cm⁻². Took temperature measurements every minute during the 20-min irradiation period.

Effect of CNT Concentration on Photothermal Performance: The role of CNT concentration in the photothermal response was studied using PDMS/CNTs-Ppy composites with five different CNT loadings: 0, 0.025, 0.05, 0.075, and 0.1 wt%. Each sample was immersed in PBS and irradiated with NIR light at a fixed power density of 1.328 W cm⁻², with the temperature monitored at 1-min intervals for 15 min.

Effect of NIR Power Density on Photothermal Heating: To examine the power-dependent photothermal behavior, PDMS/CNTs-Ppy samples with a fixed CNT concentration of 0.05 wt% were irradiated in PBS using four different NIR power densities: 0.004, 1.328, 2.512, and 3.748 W cm⁻². The temperature was recorded every minute over a 10-min irradiation period for each power setting.

Photothermal Stability and Cyclic Performance: The photothermal stability of the PDMS/CNTs-Ppy composite was assessed through a cyclic heating-cooling test. A sample with 0.05 wt% CNTs was irradiated in PBS at 1.328 W cm⁻² for 10 min, followed by a 10-min cooling period with the laser turned off. This cycle was repeated five times consecutively, with temperature recorded at 1-min intervals throughout.

Photothermal Performance under Cell Culture Conditions: To simulate a physiological environment, the PDMS/CNTs-Ppy composite was immersed in PBS maintained at 37 °C and exposed to continuous NIR irradiation. Temperature was measured at 1-min intervals over 10 min to evaluate the photothermal response under biologically relevant conditions.

Contact angle measurement: The advancing contact angle was measured on PDMS/CNTs and PDMS/CNTs-Ppy using deionized water (Theta Flex, Biolin Scientific, Finland). 3 μL droplets were dispensed onto the surface of each sample and imaged immediately using an optical contact angle meter.

Local Conductivity Mapping via Conductive Atomic Force Microscopy (C-AFM): The local electrical conductivity of PDMS/CNTs and PDMS/CNTs-Ppy nanocomposites was characterized using conductive atomic force microscopy (C-AFM). Measurements were performed on an MFP-3D Origin™ Atomic Force Microscope operated in contact mode. A Pt/Ir-coated conductive silicon tip was used, and a constant bias voltage of 1 V was applied between the tip and the sample during scanning. Current mapping images were acquired over a 400.0 nm × 400.0 nm scan area at 1 Hz, enabling direct visualization and comparison of nanoscale current distributions across the two composite surfaces.

Cell culture: Mouse myoblasts (C2C12) and Rat myoblasts (L6) were obtained from the China Academic Cell Bank. C2C12 and L6 were cultured in high-glucose Dulbecco's modified Eagle's medium (DMEM) (Gibco) supplemented with 10% fetal bovine serum (Gibco) and 1% penicillin-streptomycin. Rat Primary Cardiomyocytes were isolated from the hearts of 2-day-old Sprague-Dawley rats using a method reported by our group and cultured in high-glucose DMEM (Gibco) supplemented with 20% fetal bovine serum (Gibco) and 1% penicillin-streptomycin. The cells were incubated at 37 °C in a humid atmosphere containing 5% CO2, and the culture medium was refreshed every 2 days.

Transcriptome sequencing and data analysis: C2C12 cells (2 × 10⁶ cells/ml) were cultured for 3 days on PDMS/CNTs-Ppy, PDMS/CNTs-PLA, or TCP-Ppy as described above. Cells were then lysed using TRIzol reagent (Ambion, Carlsbad, CA) and stored at −80 °C before sequencing. RNA-seq was performed by Illumina HiSeq × 10 (Illumina, USA). Genefund Biotech (Shanghai, China) conducted quality control, comparison, and quantification of raw RNA-seq data to assess expression levels. RNA-seq expression data were analyzed using R (version 4.2.2). The limma package was used to normalize gene expression levels and detect DEGs, applying a minimum 1.5-fold change and a P-value cutoff of <0.05 as filtering criteria. The clusterProfiler package was used to perform functional enrichment analysis of the identified DEGs.

L6 and C2C12 cell Live-Dead staining: The collected cells were seeded in the 48-well (1 × 10⁴ cells per well) and cultured with medium soaked with the PDMS/CNTs and PDMS/CNTs-Ppy for 1, 3, or 5 days, respectively. After culture, samples were washed 3 times with PBS, incubated with a live-dead reagent for 10–15 min at room temperature in the dark, and then imaged on a laser-scanning confocal microscope (LSM 880, Zeiss).

Immunocytochemistry: For immunofluorescence staining, the samples were incubated in primary antibodies overnight at 4 °C, including mouse anti-α-actinin (1:250), rabbit anti-CX 43 (1:1000), and then treated with Alexa Fluor-488 Donkey anti-mouse IgG (H&L) (1:500) or Alexa Fluor-568 Donkey anti-rabbit IgG (H&L) (1:500) for 1 h. The stained samples were further incubated with 4′,6-diamidino-2-phenylindole (DAPI) for 30 min. L6 and C2C12 were seeded onto each substrate, including PDMS/CNTs, PDMS/CNTs-Ppy, or TCP, and fixed with 4% paraformaldehyde solution (Sigma) for 15 min at 3 days in vitro. The cells were then incubated with Triton X-100 (0.2% w/v, Sigma) and BSA (1% w/v) in PBS for 20 min at RT to permeabilize and block. The samples were incubated overnight at 4 °C with mouse anti-vinculin (Abcam) to assess cell adhesion before and after detachment, and then washed three times with PBS. Subsequently, the substrates were stained with an anti-mouse secondary antibody (1:200, Invitrogen) for 2 h and rewashed with PBS. To visualize actin and nuclei, the samples were additionally stained with rhodamine phalloidin (1:50 Invitrogen), fluorescein isothiocyanate (FITC)-phalloidin (1:100 Thermo), and 4′,6-diamidino-2-phenylindole (DAPI, Sigma), respectively, for 40 min at RT. All fluorescence images were captured using a confocal laser-scanning microscope (LSM 880, Zeiss).

Calcium transients imaging: After 7 days of culture, the CMs on hydrogels were stained with Fluo-4 AM to observe Ca²⁺ activity according to the manufacturer's instructions. Briefly, the samples were washed with PBS and then incubated in Fluo-4 AM work solution at 37 °C in 5% (v/v) CO₂ for 45 min. Finally, the samples were washed with buffer solution, and the calcium activity was recorded under a fluorescence microscope (Olympus BX53 software). The video data were analyzed by ImageJ software and Prism 8.

Cell sheet harvesting: The PDMS/CNTs-Ppys were UV-sterilized for 2 h, then immersed in a 10% penicillin-streptomycin mixture for 2 h, and washed three times with PBS before seeding the cells. After cells reached confluence, the media were replaced with PBS, and NIR light was used to stimulate cell sheet detachment for 15 min.

Volumetric muscle loss (VML) injury model: All the animal studies were approved by Southern Medical University Animal Ethics Committee (SYXK (yue) 2021–0167). Male Sprague-Dawley (S.D.) rats (10 weeks old, weight 200 ± 20 g) were obtained from the Laboratory Animal Center of the Academy of Southern Medical University (Guangzhou, China) and used as the animal model for all implantation studies, in accordance with the ethics committee guidelines. Isoflurane was used to anesthetize the animals. Under anesthesia, the right leg skin of the rat was incised, and the fascia and muscle were separated. According to the equation, approximately 20% ablation of the TA muscle mass was conducted. The pre-functionalized, intact cell sheet was carefully spread flat into the VML defect cavity in the rat tibialis anterior muscle using a sterile pipette. 7–0 absorbable sutures were used to suture together the preserved muscle fascia surrounding the defect area. By tightly suturing the surrounding fascia, a relatively enclosed “fascial compartment” was created. This naturally encapsulated the cell sheet, securing it firmly against the defect bed while preventing displacement in the early postoperative period due to limb movement or tissue fluid flow.

Histological analysis of skeletal sections: Four weeks after the transplanted, all the animals were euthanized. The skeletal muscle was harvested and cut into 6 μm-thick paraffin sections. The histologic features of VML were assessed by Masson tricolor staining according to the manufacturer's instructions. The area of skeletal muscle defects based on the ratio of collagenous area (blue) to skeletal muscle area (red). The data were analyzed using ImageJ and Prism 8. The immunofluorescence staining procedure was as follows: The skeletal muscle defect sections were washed 3 times in PBS, permeabilized with 0.2% Triton X-100 at room temperature for 20 min. After washing with PBS, the samples were blocked in 2% BSA at room temperature for 1 h. The primary antibodies of CD86 (1:200), CD206 (1:200), vWF (1:250), α-SMA (1:200), and MHC-Ⅱ (1:250) were added at 4 °C overnight. The samples were then incubated in the secondary antibody working solution for 2 h at room temperature. Finally, the nuclei were marked with DAPI. The pictures were obtained using a fluorescence microscope (LSM 880, Zeiss).

CRediT authorship contribution statement

Chunyi Pu: Writing – original draft, Methodology, Formal analysis, Data curation, Conceptualization. Siyu Liang: Validation, Methodology, Formal analysis. Yue Ma: Validation, Methodology, Formal analysis. Xingyun Fan: Validation, Methodology. Jin Li: Methodology. Jingyu Guan: Methodology, Investigation. Rurong Lin: Writing – review & editing, Methodology. Shuai Liu: Methodology, Formal analysis. Jie Zhang: Validation, Methodology. Dong Yang: Methodology. Xuesong Jiang: Writing – review & editing, Supervision, Funding acquisition. Xiaozhong Qiu: Writing – review & editing, Supervision, Project administration, Funding acquisition. Honghao Hou: Writing – review & editing, Supervision, Project administration, Funding acquisition, Conceptualization.

Ethics approval and consent to participate

Experimental and animal care procedures were approved by the Southern Medical University Ethics Committee (SYXK (Guangdong)2021–0167).

Declaration of competing interest

The authors do not have any possible conflicts of interest, and declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article: