Hyperbranched Biorefinery Molecule‐Regulated Switchable Adhesion and Noninvasive Healing

ABSTRACT

Switchable adhesives hold great potential in fields like flexible electronics, soft robotics, and tissue repair, etc. However, creating bio‐based switchable adhesives for biomedical use remains challenging due to the difficulty in balancing high strength and a broad adhesion span. Here, we report a dynamic chemical modification strategy based on biorefinery‐derived hyperbranched nanoconfinement to fabricate a switchable adhesive for biomedical applications. Microbial fermentation produces hyperbranched biorefinery molecules with nanostructures that address common biomass‐related issues like long production cycles, unstable batches, and by‐product generation, while also providing abundant binding sites for high‐density dynamic bonds. This hyperbranched nanoconfinement enables substantial energy dissipation, facilitating efficient stress redistribution and interfacial reconfiguration at fracture sites, thereby achieving simultaneously excellent adhesive strength and switchable adhesion. The resulting switchable wound bioadhesive exhibits a wide switching span (296 to 17 N/m) and high stability, enabling non‐invasive repair with a 94.80% wound closure rate after 10 days, significantly higher than the control (76.07%). This work explores the application of biorefinery molecules in the field of bio‐based adhesives and develops a promising candidate for innovating adhesion solutions in the biomedical field.

This work develops a reversible adhesive dressing from a hyperbranched polysaccharide produced by microbial fermentation. Its nanoconfinement provides abundant sites for high‐density dynamic disulfide bonds, enabling energy dissipation, stress redistribution, and interfacial reconfiguration, thus achieving strong yet switchable adhesion with a wide switching span and high stability for noninvasive repair.

1. Introduction

Switchable adhesives exhibit great potential in biomedical applications, flexible electronics, soft robotics, and recyclable materials [1, 2, 3, 4]. The adhesives used for wounds in biomedical fields require firm adhesion to prevent detachment and support wound healing, but extreme adhesion may cause secondary tissue damage and delay the healing process [5]. The dynamic balance of switchable adhesives between damage sealing and non‐invasive removal can effectively improve therapeutic effects and reduce complications, which reveals pivotal significance for achieving non‐damage healing promotion of wounds [6]. However, constrained by structural design and raw material sources, existing switchable adhesive materials observed weak adhesion, cytotoxicity, or uncontrollable biological activity [7, 8]. Therefore, the development of integrated multi‐performance, switchable adhesive materials suitable for the complex tissue microenvironment is highly desired but still challenging.

Although switchable adhesives hold significant potential, achieving switchable adhesion on target surfaces while maintaining stability remains challenging. Among the current strategies for realizing switchable adhesion are the modulation of surface energy at the adhesive interface, physical structural deformation, micro/nano‐structured contact, and the introduction of reversible bonds into polymer backbones [9, 10, 11]. For instance, Li et al. achieve high‐strength adhesion and easy detachment at the cardiac interface, utilizing dynamic lithium bonds, and bionic cascade structures. Liu et al. utilized N, N‐dimethylacrylamide to fabricate an adhesive that leverages the solid–liquid phase transition behavior of ionic crystals to achieve reversibility [12]. Danielson et al. introduced dynamic covalent vinylogous urethane bonds into poly(ethylene terephthalate)‐based adhesives, enabling thermal self‐healing and multiple cycles of on‐demand debonding and rebonding [13]. However, most switchable adhesives are limited by their low density of the dynamic bonds at the interface, resulting in weak interfacial adhesion. The dynamic network dominated by linear polymers causes irreversible chain slippage or breakage during mechanical stretching, reducing the reliability and reusability of the adhesion [14]. Maintaining sufficient mechanical robustness to provide physical barriers and protection, absorption capacity for wound exudate, and bioactivity to minimize the frequency of dressing changes are crucial yet challenging for switchable adhesives intended for wound healing [15, 16]. Therefore, the preparation method of on‐demand dressings with high mechanical strength, excellent wound management ability, and outstanding switchable adhesion performance is still worth exploring.

Biobased materials with inherent biocompatibility have emerged as potential candidates for developing switchable adhesives for biomedical applications [17, 18, 19, 20] However, their stability is often limited by challenges such as batch‐to‐batch variability of raw materials, side reactions during synthesis, and poor structural controllability. Recent advances in biorefining technologies, like microbial fermentation, offer promising pathways to produce biobased materials with consistent structures, and reproducible performance [21, 22]. The functional performance of switchable adhesion systems relies on the synergistic coupling of multi‐scale molecular topology, dynamic chemical bonds, and surface microstructures [23]. Hyperbranched polymers have abundant branching points that permit integration of dense dynamic bonds, diverse intermolecular interactions, and customized functional segments into the switchable adhesion system and facilitate effective energy dissipation during detachment [24]. However, it remains challenging to obtain bio‐based materials with hyperbranched structures through biorefining and apply them to the preparation of switchable adhesive materials.

In this study, an integrated wound bioadhesive with excellent switchable adhesion and physiological activity was developed by in situ gelation of α‐lipoic acid (LA)‐grafted hyperbranched biorefinery molecule nanostructures on a cellulose nonwoven fabric (NWF). Biorefinery molecules obtained through fermentation further undergo self‐assembly facilitated by non‐covalent interactions to form hyperbranched nanostructures, which offer abundant binding sites for the incorporation of high‐density dynamic covalent bonds (disulfide bonds), helping form an adaptive nanoconfined network with efficient energy dissipation capacity. The bioadhesive based on nanostructure biorefinery molecules demonstrates promising performance in the complex microenvironment of postoperative wounds by combining excellent switchable adhesion with bioactive functionality. Thus, this work pioneers the application of biorefinery molecules in switchable adhesion while simultaneously developing a promising candidate for innovating adhesion solutions in the biomedical field.

2. Results and Discussion

2.1. Switchable Adhesives Constructed with the Hyperbranched Biorefinery Molecule

Traditional bio‐based macromolecules, such as cellulose and lignin, are widely employed in the development of switchable adhesive materials [25, 26, 27]. However, their application is often hindered by rigid linear or complex network structures, leading to challenges such as poor solubility and difficult modification. Moreover, heterogeneity in the raw materials and complexities in the processing result in batch‐to‐batch inconsistencies and the formation of numerous by‐products, which undermine the fabrication of stable switchable adhesives and increase the cost of material preparation. Therefore, it is imperative to develop novel bio‐based polymeric materials with a stable and controllable structure, high solubility, and ease of modification for constructing high‐performance switchable adhesive systems.

On the basis of these discussions, a biorefinery molecule (Levan) with hyperbranched nanostructures was obtained via microbial fermentation. Owing to its inherent hyperbranched structure and abundant functional groups, Levan exhibits high water solubility and ease of modification. This microbial fermentation approach overcomes the raw material heterogeneity inherent to traditional extraction methods, while offering advantages such as low feedstock cost, easy scalability, and stable product structure [28]. Then, an ester exchange reaction occurs between methacrylate anhydride (MA) and the active hydroxyl group on the Levan chain, thereby grafting the methacrylate with a ─C═C─ bond onto the backbone to yield Lev‐MA [29]. Subsequent further modification with LA was performed. During this process, LA underwent spontaneous ring‐opening polymerization to form polylipoic acid (PLA), and the terminal radicals of PLA reacted with the ─C═C─ bond on Lev‐MA, thereby preventing depolymerization of PLA and resulting in the final product PLA‐LM [30]. Subsequently, PLA‐LM was in situ gelated on the cellulose‐based NWF processed from plants to obtain PLA‐LM/NWF dressings (Figure 1a). A dynamic covalent network with adaptive properties and efficient energy dissipation capabilities is constructed by leveraging the hyperbranched nanoconfinement in synergy with the dynamic disulfide bond exchange within PLA chains. The abundant hydroxyl and carboxyl groups in PLA‐LM form high‐density intermolecular hydrogen bonds with cellulose molecules in NWF, enabling PLA‐LM to firmly adhere to the fabric surface and produce stable PLA‐LM/NWF dressings (Figure 1b). PLA‐LM/NWF dressings can toughly adhere to the skin and continuously release LA‐based active molecules, demonstrating potent antibacterial, antioxidant, and anti‐inflammatory effects, along with a marked pro‐angiogenic effect (Figure 1c). These attributes collectively contribute to effective infection eradication, promotion of angiogenesis, and comprehensive tissue regeneration in wounds [31]. Furthermore, the high‐density dynamic cross‐linking formed by plentiful dynamic covalent bonds (disulfide bonds) and non‐covalent bonds in PLA‐LM/NWF endows it with reversible adhesion properties. Such a significant reduction in the adhesion strength of the PLA‐LM/NWF dressing after glutathione treatment, owing to the disruption of the disulfide bond, permits its easy and painless removal from the skin [32]. demonstrating the excellent switchable adhesion performance (Figure 1d). This feature minimizes damage to newly formed tissues and alleviates patient discomfort during dressing changes.

FIGURE 1.

Design and formation of PLA‐LM/NWF dressing. (a) The preparation process of PLA‐LM/NWF dressing. (b) The hyperbranched nanoconfinement‐based dynamic network and internal multiple hydrogen bond interaction of PLA‐LM/NWF dressing. (c) Illustration of the proposed mechanism of PLA‐LM/NWF dressing for promoting wound healing. (d) Schematic diagram of the detachment of PLA‐LM/NWF dressing from the skin surface after glutathione treatment.

The fabrication of PLA‑LM/NWF is accomplished via a one‑pot synthesis of PLA‑LM followed by its in situ gelation on the NWF, requiring no additional steps and offering a simple, rapid process. Notably, as an entirely bio‑based reversible adhesive wound dressing, PLA‑LM/NWF provides non‑invasive debonding, promotes wound healing, and exhibits high biocompatibility. Therefore, from a life‑cycle and clinical‑benefit perspective, PLA‑LM/NWF represents a cost‑effective and translatable solution.

2.2. Hyperbranched Biorefinery Molecule

The levan investigated in this study was synthesized from sucrose by the halotolerant bacterium Bacillus sp. SCU‐E108 via a series of biochemical reactions during fermentation (Figure S1), followed by isolation and purification. Monosaccharide composition analysis (Table S1) revealed a fructose content exceeding 99% in the obtained levan. Notably, the prepared Levan exhibits a hyperbranched structure (Figure 2a). Building on our earlier work on levan and relevant literature [33]. we further elucidated the fine structural features of the polymer using one‐ and two‐dimensional NMR spectroscopy. The ¹H NMR spectrum (Figure 2b) displayed major signals between 3.4 and 4.8 ppm, with no detectable anomeric proton signals in the downfield region, consistent with a typical Levan framework. In the ^{1 3}C NMR spectrum (Figure 2c), three resonances were observed at 104.18, 104.08, and 103.59 ppm, indicating the presence of three distinct fructosyl residues, designated as residue A, residue B, and residue C. 2D NMR analyses‐including COSY (Figure 2d), NOESY (Figure 2e), HSQC (Figure 2f), and HMBC (Figure 2g)‐were employed to further characterize the levan structure. Based on these spectra, the ¹H and ^{1 3}C chemical shifts were assigned (Table S2). Comparison with previously reported Levan units revealed that the C‐2 of residue A appears downfield, suggesting a β‐anomeric configuration. Notably, the observed downfield shift of the C‐6 signal in residue A confirmed the presence of a 2,6‐linked β‐D‐fructofuranosyl unit, supporting the structure →6)‐β‐D‐Fruf‐(2→ for residue A [34]. Similarly, the HMBC correlation at 3.59 ppm (C H‐2)/104.27 ppm (B C‐2) indicated that the O‐2 of residue C is linked to C‐1 of residue B, corresponding to the segment β‐D‐Fruf‐(2→1,6)‐β‐D‐Fruf‐(2→). NOESY correlations at 3.70 ppm (A H‐1)/3.81 ppm (A H‐6), 3.70 ppm (A H‐1)/3.82 ppm (B H‐6), and 3.66 ppm (B H‐1)/3.48 ppm (A H‐6) further revealed linkages between O‐2 of residue A and C‐6 of residue A, O‐2 of residue A and C‐6 of residue B, and O‐2 of residue B and C‐6 of residue A, respectively. Additionally, the branching degree, calculated from the integral ratio (4.0:1.9:1.2) of the C‐2 signals for residues A, B, and C, was as high as 43% (Figure S2). Collectively, these NMR data support the possible structure of the obtained Levan (Figure S3), confirming its hyperbranched nanostructure. Meanwhile, the TEM image (Figure 2h) shows the spherical shape of Levan, which is a typical hyperbranched aggregate structure. In conclusion, we believe that the Levan chains obtained through microbial fermentation self‐assemble and aggregate to form a spherical hyperbranched nanostructure during separation and purification processes (Figure 2i).

FIGURE 2.

Synthesis and Characterization of PLA‐LM. (a) Levan with a hyperbranched structure (b). The ¹H, (c) ¹³C, (d) COSY, (e) NOESY, (f) HSQC, and (g) HMBC NMR spectra of the obtained Levan. (h) The TEM image of the obtained Levan. (i) The preparation process of Levan.

2.3. Switchable Adhesion Performance

The surface of spherical and hyperbranched Levan possesses numerous terminal functional groups (‐OH), providing abundant reactive sites for chemical modification and contributing to the construction of a high‐density dynamic network. The synthetic route for preparing PLA‐LM from levan is illustrated in Figure 3c. The ¹H NMR spectra of Lev‐MA and PLA‐LM after modification with MA and LA are shown in Figure 3a and Figure S4, respectively. Compared with the original levan, two new signals appear at 5.7 and 6.1 ppm in the spectrum of Lev‐MA, indicating successful introduction of the olefinic protons from MA [35]. In the spectrum of PLA‐LM, the emergence of a carboxylic acid proton peak at 12 ppm, along with the disappearance of the olefin proton signals, suggests that the terminal radical of LA underwent spontaneous ring‐opening polymerization to form PLA, which subsequently grafted onto the ─C═C─ bonds of Lev‐MA [36]. This transformation is further supported by the Raman spectrum (Figure 3b), which shows a shift in the C─S stretching vibration (from 681 to 674 cm^{− 1}) and splitting of the S─S vibration band [25]. The crystallinity of the samples before and after modification was analyzed by X‐ray diffraction. As depicted in Figure S5, the Levan maintains an amorphous structure both pre‐ and post‐modification, which may be related to the inherent structural irregularity of hyperbranched architectures hindering the regular arrangement of polymer chains, thereby suppressing crystallization.

FIGURE 3.

Characterization and adhesion performance of PLA‐LM. (a) The ¹H NMR spectra of Levan and Lev‐MA. (b) The Raman spectrum of LA and PLA‐LM. (c) The synthetic route of PLA‐LM. The temperature‐variable FT‐IR in (d) 3500–2800 cm⁻¹ and 1780–1665 cm⁻¹, 2DCS spectra of PLA‐LM in the range of (e) 3500–3070 cm⁻¹ and 1770–1670 cm⁻¹. (f) Images of PLA‐LM adhering to various materials. (g) The shear strength of PLA‐LM adhering to various materials. (h) The disruption and recovery of the hyperbranched nanoconfinement‐based dynamic network. (i) Images of the changes in adhesion strength of PLA‐LM after heating. (j) Reversible adhesion performance of PLA‐LM. (k) Representative force‐displacement curves of PLA‐LM/NWF before and after treatment with glutathione, with ≈1741% switching ratio in adhesion force.

The thermal response kinetics of molecular interactions in PLA‐LM were investigated in detail using temperature‐dependent FT‐IR spectroscopy combined with generalized two‐dimensional correlation spectroscopy (2DCS). As shown in Figure 3d, the broad O─H absorption band at 3188 cm^{− 1}, associated with strong hydrogen bonding, decreases markedly as temperature increases. Concurrently, the characteristic ─C═O stretching peak shifts from 1700 to 1710 cm^{− 1}, indicating a blue shift consistent with the weakening of hydrogen bonding. These changes reflect a typical hydrogen bond dissociation process, confirming the presence of strong dynamic interactions within the material [37]. Analysis of the 2DCS asynchronous maps (Figure 3e) and PCMW2D (Figure S6) further reveals that interfacial hydrogen bonds dissociate progressively over a broad temperature range. This dissociation facilitates the reconfiguration of the dynamic network with a low activation energy barrier [38], which is meaningful for understanding the reversibility of the PLA‐LM. PLA‐LM exhibits excellent adhesive properties, observing strong, and stable adhesion to diverse materials (including glass, paper, iron, and plastic) as well as biological tissues (such as skin, muscle, liver, and heart) (Figure 3f). Meanwhile, its shear strength on porcine skin exceeds 125 kPa (Figure 3g), which adequately meets the requirements for stable attachment of wound dressings to skin [39], The outstanding interfacial adhesion of PLA‐LM stems from synergistic interactions involving multiple mechanisms, including hydrogen bonds, coordination bonds, dipole‐dipole interactions, and physical adsorption (Figure S7). The cleavage and reformation of disulfide bonds underpin the reversible adhesion behavior of PLA‐LM. Raman spectroscopy performed at room temperature and upon heating revealed that the characteristic disulfide bond peak disappeared at elevated temperature and reappeared after cooling back to room temperature, confirming the reversible breakage, and reformation of disulfide bonds (Figure S8). Upon heating, numerous disulfide bonds in PLA‐LM undergo homolytic cleavage by directly absorbing energy, generating sulfur radicals. This process markedly disrupts the hyperbranched nanoconfinement‐based dynamic network structure of PLA‐LM, leading to a pronounced reduction in adhesion strength. After removal of the heating source, sulfur radicals recombine to regenerate the original or form new disulfide bonds, thereby restoring the dynamic network and recovering the initial adhesion strength (Figure 3h) [40, 41]. This reversible dissociation‐reconstruction process causes no permanent chemical damage. The macroscopic manifestation of this behavior was captured visually (Figure 3i): PLA‐LM adhered to glass firmly and supported a 200 g weight under ambient conditions. After heating, the weight dropped as the adhesive strength declined. Subsequent removal of the heat source restored the tough adhesive performance. PLA‐LM demonstrates robust and reversible adhesion performance that can maintain stable adhesion strength after multiple adhesive‐debonding cycles, achieving an optimal balance between high adhesion strength and damage‐free cyclability. To fabricate a practical wound dressing, PLA‐LM was combined with NWF to form PLA‐LM/NWF composites. PLA‐LM/NWF dressing exhibits strong adhesion to skin (≈296 N·m^{− 1}, Figure 3j), attributed to the high density of hydroxyl and carboxyl groups in PLA‐LM that form dynamic non‐covalent bonds and multiple interactions with the skin surface. Notably, the results of Tian et al. suggested that PLA‐LM/NWF may establish stable dynamic cross‐links via disulfide bonds (S─S) with sulfhydryl groups present in skin tissue, further enhancing adhesion strength [42]. Except for heating, the use of a moderate approach (such as glutathione processing) can also reduce the adhesion strength. As shown in Figure 3k, the adhesion strength of PLA‐LM/NWF decreases dramatically by approximately 1741% after being treated with glutathione, highlighting an excellent on‐off switching ratio and outstanding switchable adhesion performance [5, 43]. The reversible adhesive dressing developed in this work was compared with other previously reported wound dressings recognized for their excellent performance, in order to evaluate its adaptability in wound adhesion (Figure S9). The results demonstrate that our dressing exhibits superior overall performance, including higher adhesion strength than most existing dressings, as well as outstanding reversible capability, making it suitable for non‑invasive wound repair.

2.4. Mechanical Properties and Biological Activities

As depicted in Figure 4a, the NWF derived from plants exhibits a loose and porous network structure. After integrating with PLA‐LM to form the PLA‐LM/NWF dressing, the structure becomes denser while retaining a certain degree of porosity, which facilitates exudate management and maintains breathability. To comprehensively evaluate its mechanical compliance for application on dynamically deforming skin, we further characterized the full tensile stress–strain behavior (Figure S10). The integration of PLA‐LM has significantly enhanced the mechanical strength of the NWF, with the tensile strength of PLA‐LM/NWF exceeding 9 MPa. The PLA‐LM/NWF exhibits a Young's modulus of 18.3 MPa, and an elongation at break of 55.2%, which is higher than that of NWF (Figure S11). These values indicate that the PLA‐LM/NWF dressing possesses both enhanced stiffness and improved flexibility compared to the pristine NWF, enabling it to maintain robust structural integrity while accommodating skin deformation during daily movements. Notably, both native Levan and MA‐modified Lev‐MA exist as low‐viscosity liquids at room temperature due to their hyperbranched spherical architecture, and thus cannot adhere effectively to the NWF (Figure S12). In contrast, after grafting with PLA chains, the resulting PLA‐LM forms a viscous gel that can stably adhere to the NWF surface. Recognizing the critical role of PLA in the fabrication and stability of the dressing, we evaluated the interfacial toughness between PLA and NWF (PLA/NWF system), as well as between PLA‐LM and NWF (PLA‐LM/NWF system), via peel tests (Figure 4b). As shown in Figure 4c, the high average interfacial toughness for the PLA/NWF system (108 N·m^{− 1}) indicates excellent energy dissipation and resistance to steady‐state debonding. Remarkably, the interfacial toughness of the PLA‐LM/NWF system increases by 3.72 times and reaches 402 N·m^{− 1}. This enhancement is attributed to the hyperbranched spherical Levan grafted onto the PLA chains, which strengthens the interaction with the NWF and suggests the dressing's high stability and environmental resistance.

FIGURE 4.

Characterization of the PLA‐LM/NWF dressing. (a) The SEM images of NWF and PLA‐LM/NWF dressing. (b) Schematic of peeling test. (c) The interfacial toughness of PLA/NWF and PLA‐LM/NWF. (d) Binding energies and optimized models of the Levan/NWF system, the PLA/NWF system, and the PLA‐LM/NWF system were obtained through molecular dynamics simulations. (e) Images of PLA‐LM/NWF dressing on dry and wet skin. (f) The antioxidant activity of Levan, PLA, and PLA‐LM/NWF dressing. (g) The cytotoxicity of PLA‐LM/NWF dressing. (h) The antibacterial rate, (i) the agar plates images, and (j) the SEM images of E. coli and S. aureus treated with Levan, PLA, and PLA‐LM/NWF dressing.

Furthermore, we employed molecular dynamics simulations to determine the optimal conformations and binding energies of the Levan/NWF, PLA/NWF, and PLA‐LM/NWF systems. As summarized in Figure 4d, the Levan and PLA to the NWF are observed to have low binding energies of 101.92 and 156.44 kJ/mol, respectively. In contrast, the binding energy of PLA‐LM to the NWF increases significantly to 310.32 kJ/mol. This increase correlates with enhanced interfacial interactions: the abundant hydroxyl groups in Levan and the incorporation of PLA chains collectively enable PLA‐LM to form a dense [44], strong non‐covalent cross‐linked network with the NWF. These findings align well with the peel test results and confirm the improved stability of the composite structure. To further validate the unique contribution of the hyperbranched architecture of Levan to interfacial binding, we performed comparative molecular dynamics simulations using a linear polysaccharide (amylose) with the same degree of polymerization. Both the hyperbranched Levan and linear amylose were modified with MA and then simulated in interaction with the cellulose‐based NWF (Figure S13). The binding energy of the modified linear amylose (PLA‐AM) to NWF was calculated to be 726.34 kJ/mol, whereas that of the modified hyperbranched levan was higher, reaching 774.64 kJ/mol. These results demonstrate that the hyperbranched nanostructure provides a substantially larger interfacial contact area and more abundant binding sites, leading to enhanced energy dissipation and more efficient stress redistribution compared to its linear counterpart [45].

The wettability of NWF, PLA/NWF, and PLA‐LM/NWF was evaluated by water contact angle measurements (Figure S14). It should be noted that water droplets spread and wetted the NWF substrate almost instantaneously, making it impractical to obtain a reliable contact angle value. The NWF conjointed with PLA exhibited high hydrophilicity, whereas the PLA‐LM/NWF dressing showed the lowest hydrophilicity among the three. The optimal medium water contact angle close to 90° allows the PLA‐LM/NWF dressing to absorb excess wound exudate while preventing excessive drying of the wound and surrounding skin tissue, as well as obtaining a certain level of waterproof capability [46]. Furthermore, the practical adhesion stability of the dressing under wet conditions was tested by applying the PLA‐LM/NWF dressing to skin and immersing it in tap water for approximately 30 s (Figure 4e). Both the volume of water exposure and the duration in this test far exceed the typical water contact expected at the wound site. After that, the dressing still maintained strong adhesion performance, proving it can effectively avoid adhesion failure caused by contact with water in daily life.

PLA‐LM/NWF dressing can release natural active molecules LA with biological activities such as antioxidation, anti‐inflammation, and antibacterial properties (Figure S15), which are crucial for conducting its biological activity [47]. The PLA‐LM/NWF dressing exhibited remarkable antioxidant activity, with DPPH and ABTS radical scavenging rates as high as 77% and 91%, comparable to that of PLA (Figure 4f). The excellent antioxidant activity indicates its potential to effectively alleviate oxidative stress in wounds, a key factor in reducing subsequent scar formation [48]. Meanwhile, the PLA‐LM/NWF dressing presented low cytotoxicity with cell viability remaining above 85% after 12 and 24 h of incubation with 293T cells at different dosages, confirming its excellent biocompatibility (Figure 4g). The antibacterial performance of PLA‐LM/NWF dressing was evaluated using E. coli and S. aureus as representative Gram‐negative and Gram‐positive bacteria commonly associated with wound infections. Notably, while unmodified Levan showed no antibacterial activity, the PLA‐LM/NWF dressing inhibited both bacterial types by nearly 90%, a level comparable to that of PLA alone (Figure 4h,i). SEM imaging further revealed obvious shrinkage and rupture of bacteria treated with PLA and PLA‐LM/NWF (Figure 4j), underscoring the essential role of the PLA chains in the antimicrobial performance of PLA‐LM/NWF dressing.

2.5. Application in Non‐Damage Healing

Following confirmation of the favorable bioactivity and biocompatibility of PLA‐LM/NWF, its wound healing efficacy was evaluated using a mouse full‐thickness wound excision model. The mice, with a circular wound (6 mm in diameter) on the dorsal surface, were randomly divided into three groups: control (medical gauze), PLA/NWF, and PLA‐LM/NWF. The dressings were changed every 2 days, and wound images were recorded at each time point. The skin tissues were collected on day 10 for histopathological analysis to further evaluate the microscopic therapeutic effects of the different treatments. As depicted in Figure 5a, the PLA/NWF and PLA‐LM/NWF groups exhibited better healing outcomes compared with the control group. Specifically, the PLA/NWF and PLA‐LM/NWF groups achieved wound closure rates of 94.80% ± 2.93% and 86.72% ± 3.24% after 10‐day treatment, significantly surpassing the control group (76.09% ± 3.40%) (Figure 5b). The PLA/NWF group exhibited higher wound healing effectiveness than PLA‐LM/NWF, which may be attributed to the cross‐linking between Levan and PLA, facilitating sustained release of LA in the wound surface. H&E and Masson staining (Figure 5c) revealed that all groups displayed different levels of inflammatory cell infiltration, fibroblast migration and proliferation, and subsequent granulation tissue formation. Notably, interstitial voids accompanied by epidermal necrosis and detachment at the dermal‐epidermal junction were observed in the control group. Meanwhile, the PLA‐LM/NWF group demonstrated reduced inflammatory infiltration, thicker granulation tissue, and a more continuous epidermal layer compared with the control group. In the Masson staining images, a clear boundary was observed between the collagen fibers and the regenerated normal skin in the control group. In contrast, the PLA‑LM/NWF group showed a much more indistinct interface, suggesting a lower degree of fibrotic repair [49]. Furthermore, collagen volume fraction analysis (Figure 5d) showed that the PLA‐LM/NWF group achieved the highest level of collagen deposition (70.83% ± 3.65%) within the wound surface, representing a 31.02% increase over the control group. These results confirm that the PLA‐LM/NWF dressing can accelerate tissue regeneration by augmenting collagen encapsulation [50, 51].

FIGURE 5.

Effect of PLA‐LM/NWF dressings in the full‐thickness wound model of mice. (a) Photographs of mouse skin wounds. (b) Quantitative analysis of wound healing effect. (c) H&E and Masson staining images of wound tissues in different groups at day 10. (d) Quantitative result of the collagen volume fraction of skin wounds. (e) Immunofluorescence staining images of CD31, α‐SMA, CD68, and IL‐6. (f) Quantitative immunofluorescence results of CD31 and α‐SMA. (g) Quantitative immunofluorescence results of CD68 and IL‐6. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. All data are shown as mean ± S.D.

To further investigate the mechanism of PLA‐LM/NWF in wound healing, immunofluorescence analysis was performed on skin tissues collected on day 10 (Figure 5e). Staining for angiogenesis‐related markers, including platelet endothelial cell adhesion molecule (CD31) and α‐smooth muscle actin (α‐SMA), was conducted. CD31 served as an endothelial marker to identify newly formed capillaries, while α‐SMA was used to label pericytes and smooth muscle cells, indicating mature vasculature [52]. Staining activity results (Figure 5f) demonstrated that the PLA‐LM/NWF group presented a markedly higher level of angiogenesis at the wound site compared to the PLA/NWF and control groups, highlighting its excellent pro‐angiogenic capability. Immunofluorescence staining for the macrophage marker CD68 and interleukin‐6 (IL‐6) was employed to quantitatively assess the degree of inflammation in the injured area [53, 54] as shown in Figure 5g, the expression levels of both CD68 and IL‐6 in the PLA‐LM/NWF group were significantly lower than those in the control group, indicating a reduced inflammatory response. The immunofluorescence results indicated that the PLA‐LM/NWF group significantly modulated the wound vascular environment and local immune response. Overall, the PLA‐LM/NWF group demonstrated superior re‐epithelialization, collagen deposition, vascular maturation, and inflammation control during the healing process.

2.6. Transcriptomic Analysis

To investigate the mechanism by which PLA‐LM/NWF promotes wound healing, transcriptomic analysis was performed on tissues collected on day 10. Principal component analysis (PCA) revealed a clear separation between the control and PLA‐LM/NWF groups (Figure S16). Additionally, a high degree of reproducibility was observed among biological replicates (Pearson's Correlation Coefficient > 0.94), supporting the reliability of downstream analyses (Figure 6a) [55]. Comparative transcriptomics identified significant alterations in gene expression. Volcano graph showed 418 upregulated and 141 downregulated differentially expressed genes (DEGs) in the PLA‐LM/NWF group compared to the control (|log₂FoldChange| > 1, p < 0.05) (Figure 6b). Heat map of DEGs further confirmed distinct transcriptional profiles between groups (Figure 6c).

FIGURE 6.

(a) Heatmap visualization of the Pearson correlation coefficient matrix among all analyzed samples. (b) Volcano plots of the DEGs. (c) The heatmap of the DEG expression profiles between the control and PLA‐LM/NWF groups. (d,e) GO pathway enrichment analysis identified differentially expressed pathways between the Control and the PLA‐LM/NWF groups. (f,g) KEGG pathway enrichment analysis identified differentially expressed pathways between the Control and the PLA‐LM/NWF groups.

Gene Ontology (GO) enrichment analysis of DEGs revealed significant enrichments across cellular components (CC), molecular functions (MF), and biological processes (BP) (Figure S17). Over‐representation analysis of key DEGs highlighted several biological functions and processes related to skin tissue formation, including skin development, establishment of skin barrier, myofibril, and actin binding (Figure 6d,e) [52]. Upregulated GO terms were associated with skin development, keratinization, and epidermal development, while downregulated terms were enriched for muscle cell development, muscle cell differentiation, and myofibril (Figure S18). This expression pattern suggests that during epidermal wound repair, resources are prioritized for re‐epithelialization, driving keratinocyte proliferation, migration, and redifferentiation, thereby upregulating skin development‐related genes. Conversely, differentiation pathways unrelated to repair‐such as those involved in muscle development—are transiently suppressed, which may help prevent excessive fibrosis during healing [56]. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis indicated enrichments in organismal systems, environmental information processing, and metabolism (Figure S19). Further analysis showed that PLA‐LM/NWF modulates several key biological response networks during healing (Figure 6f,g), including regulation of the MAPK signaling pathway acting as a growth kinase, activation of the estrogen signaling pathway to accelerate wound closure, and modulation of leukocyte transendothelial migration to mitigate local inflammation (Figure S20).

Based on transcriptomic enrichment results, genes associated with skin development and barrier formation were further examined. Several key genes—including Alox12b, Cdsn, Hrnr, and Rptn‐were significantly upregulated following PLA‐LM/NWF treatment (Figure S21). These genes play distinct roles in terminal keratinocyte differentiation and cornified envelope formation: Alox12b encodes an enzyme involved in generating specific oxidized lipids essential for the lipid envelope; Cdsn and Hrnr encode structural proteins critical for cornified envelope assembly and corneocyte adhesion; and Rptn is associated with intermediate filament organization in differentiating keratinocytes [57, 58]. The coordinated upregulation of these genes suggests that PLA‐LM/NWF supports the re‐establishment of a functional, multi‐layered epidermis by promoting keratinocyte differentiation. This mechanism is closely associated with the restoration of the epidermal barrier during wound healing.

3. Conclusion

In conclusion, we have successfully developed a switchable wound bioadhesive for non‐damaging wound healing. The biorefinery molecules with hyperbranched nanostructures were prepared by microbial fermentation, and high‐density dynamic covalent bonds were introduced to construct a dynamic network. The hyperbranched nanoconfinement significantly enhances energy dissipation and promotes efficient stress redistribution and interface reconfiguration at fracture sites, thus achieving both excellent adhesion strength and a broad adhesion span (296 to 17 N/m). This design imparts robust mechanical strength, outstanding interfacial adhesion, and remarkable structural stability to the material. Furthermore, the bioadhesive exhibits low cytotoxicity as well as excellent antibacterial, antioxidant, and pro‐healing activities, significantly accelerating wound closure with a 94.80% wound closure rate after 10 days, significantly higher than the control (76.07%). Transcriptome analysis was employed to elucidate the underlying mechanisms through which the dressing promotes wound healing, thereby establishing a foundation for its further biomedical applications. This work addresses a key challenge in wound management by realizing effective switchable adhesion and is expected to inspire new avenues for the development of advanced switchable adhesives in the biomedical field.

Author Contributions

X.Z. supervised the project and provided Funding; G.J. performed data analysis, wrote the original draft, and provided Funding. K.H. performed all experiments and created the figures; C.C. and L.H. supervised the project and Conceptualization; W.S. contributed to the figure creation, revision of the paper, investigation, and formal analysis. The other authors revised the manuscript and provided some suggestions.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: smll73951‐sup‐0001‐SuppMat.docx.

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.