Highly printable and multifunctional cell-laden collagen-based bioinks for precise DLP bioprinting and rapid diabetic wound regeneration

Abstract

Digital light processing (DLP) bioprinting has revolutionized tissue engineering by offering unprecedented speed and precision. However, its full biomedical potential is hindered by the scarcity of cell-laden bioinks that combine excellent printability with superior bioactivity. In this study, we introduce a novel cell-laden collagen-based bioink optimized for precise DLP bioprinting and diabetic wound regeneration. This bioink integrates methacrylated collagen (CMA) with dihydromyricetin (DHM) and selected additives, achieving a combination of low concentration, high printability, and superior cell bioactivity, along with antioxidant and anti-inflammatory effects. By employing a multi-crosslinking strategy that integrates free radical polymerization, Michael addition, Schiff base formation, and hydrogen bonding, the bioink achieves an ultra-fast gelation speed (375 % increase), a 161 % increase in stiffness, a 231 % improvement in mechanical resilience, and a 208 % enhancement in anti-biodegradation. These properties allow for the fabrication of intricate, cell-laden constructs with micron-scale precision, high cell viability, minimal swelling, and enhanced structural stability. The CMA-DHM system synergistically enhances 3D cell proliferation, mitigates oxidative stress, and modulates macrophage polarization, significantly outperforming conventional CMA hydrogels. Leveraging these properties, we developed biomimetic skin substitutes encapsulating human dermal fibroblasts (HDFs), which effectively facilitate diabetic wound progression through critical healing phases. These skin substitutes provide potent antioxidant and anti-inflammatory effects, accelerate re-epithelialization and collagen deposition, and enhance angiogenesis, thereby preventing chronic wound formation and facilitating efficient tissue regeneration. This study establishes a versatile, scalable DLP bioprinting platform, offering a rapid and effective solution for chronic wound treatment, representing a significant advancement in regenerative medicine.

Graphical abstract

A novel multifunctional cell-laden collagen-based bioink has been developed for precise DLP bioprinting and rapid diabetic wound regeneration.

Highlights

• •Fully soluble, low-concentration CMA-DHM bioink achieved precise DLP bioprinting.
• •Multi-crosslinking improved gelation, strength, and biodegradation resistance of hydrogels.
• •Printed hydrogels exhibited high resolution and cell viability, stability and low swelling.
• •CMA-DHM hydrogels exhibited superior in vitro and in vivo anti-oxidation and anti-inflammation.
• •Bioprinted skin substitutes enable rapid healing and tissue regeneration in diabetic wounds.

1. Introduction

3D bioprinting is revolutionizing tissue engineering and regenerative medicine by offering unparalleled precision and adaptability in the creation of intricate 3D structures [1]. This cutting-edge technology allows for the precise spatial deposition of biomaterials and cells, enabling the faithful recreation of the complex composition, structure, and functionality of tissues such as skin, cartilage, and heart [[2], [3], [4]]. Among the various bioprinting techniques, such as inkjet, extrusion-based, and laser-assisted approaches, digital light processing (DLP) bioprinting has become particularly prominent [5]. By utilizing a digital light projector, DLP enables rapid, layer-by-layer photopolymerization with micrometer-scale resolution, minimizing cellular damage while preserving cell viability [6]. These features make DLP an ideal platform for fabricating complex, cell-laden structures with high fidelity and enhanced cell viability for biomedical applications.

A major challenge in advancing DLP bioprinting is the development of high-performance bioinks that balance printability, biocompatibility, and structural integrity [7]. Ideal bioinks must exhibit optimal flowability, rapid photopolymerization, high cell viability, and functionality, fine resolution for bioprinting, and sufficient mechanical stability for prolonged culture [8]. Photocurable biomaterials, which form the backbone of bioinks, are pivotal in dictating essential properties such as rheological behavior, mechanical strength, and cellular responses [9]. Several biomaterials, including gelatin, hyaluronic acid (HA), silk fibroin, and polyethylene glycol (PEG), functionalized with photosensitive groups, have shown promise for DLP applications [10,11]. However, each material is constrained by specific limitations, such as the requirement for high concentrations or elevated printing temperatures, susceptibility to deformation due to excessive swelling, or reduced bioactivity owing to their non-native extracellular matrix (ECM) composition [12,13]. These deficiencies hinder their ability to replicate the mechanical properties and biological functions of the native ECM, which are essential for the bioprinting of functional human organs or tissues in vitro [14].

Collagen, as a structural component of the ECM, has garnered significant attention as a bioink material due to its remarkable biocompatibility and bioactivity [15]. Its ability to create a supportive microenvironment that promotes cell proliferation and wound regeneration makes collagen especially advantageous for soft tissue engineering [16]. For DLP bioprinting, methacrylated collagen (CMA) is preferable to gelatin methacrylate (GelMA) as it preserves the native triple-helical structure of collagen, maintaining its superior bioactivity [17]. However, CMA is hindered by limited solubility and inadequate mechanical properties, which restrict its application in high-precision DLP bioprinting. Strategies such as blending CMA with polyethylene glycol diacrylate (PEGDA) have been explored to improve DLP printability; however, concerns regarding the biocompatibility of these composite bioinks in cell-laden environments remain unresolved [18]. Therefore, developing a highly printable, bioactive, and cell-compatible collagen-based bioink for DLP bioprinting continues to pose a substantial challenge in tissue engineering.

Beyond advancing tissue engineering, DLP 3D printing utilizing collagen-based bioinks holds exceptional potential for addressing complex regenerative clinical challenges, particularly in the treatment of diabetic wounds [19]. These wounds present a formidable therapeutic obstacle, as the normal healing process, comprising inflammation, proliferation, and remodeling, is often disrupted by hyperglycemia and the accumulation of advanced glycation end-products (AGEs) [20,21]. These factors intensify chronic inflammation and oxidative stress, impeding wound closure and compromising ECM remodeling, thereby fostering a non-healing wound environment [22]. Current treatment strategies, including debridement, growth factor therapies, and wound dressings, often fail to address the multifactorial pathophysiology of diabetic wounds, resulting in limited therapeutic success [23]. Innovative approaches that combine collagen-based bioinks with cellular components such as fibroblasts, alongside bioactive compounds with antioxidant and anti-inflammatory properties, show great promise in overcoming these challenges. Consequently, the multifunctional applications of these bioinks in regenerative medicine impose increasingly stringent demands on their design and performance.

In this study, we have for the first time developed a novel, highly printable, multifunctional, cell-laden collagen-based bioink designed to enhance DLP bioprinting and accelerate diabetic wound regeneration (Fig. 1). This innovative bioink overcomes the limitations of conventional formulations by integrating high-substitution methacrylated collagen (CMA), dihydromyricetin (DHM), a photoinitiator (LAP), and a UV absorber to optimize printability, bioactivity, and mechanical properties. The CMA-DHM system leverages DHM's multi-crosslinking potential to address the mechanical challenges typically associated with collagen-based bioinks. Through a “one-step incorporation, multifunctional enhancement” strategy, the bioink demonstrates exceptional performance in both acellular and cell-laden DLP-based 3D printing, supporting enhanced cell proliferation, sustained antioxidant activity, rapid anti-inflammatory effects, and accelerated wound healing. When enriched with human dermal fibroblasts (HDFs), the CMA-DHM bioink was utilized to fabricate biomimetic dermal constructs that significantly improved wound closure, reduced oxidative stress, mitigated inflammation, enhanced epithelialization, and stimulated angiogenesis in diabetic wound models. These findings establish the CMA-DHM bioink as a robust and versatile platform for advanced 3D bioprinting, soft tissue engineering, and chronic wound management.

Fig. 1.

Schematic representation of the highly printable and multifunctional cell-laden collagen-based bioinks, designed for precise DLP bioprinting and accelerated diabetic wound regeneration. The figure highlights effective multi-crosslinking mechanisms, an advanced DLP 3D bioprinting strategy, superior cell-laden printing performance, enhanced bioactivity, immunomodulatory properties, and their efficient application in diabetic wound healing.

2. Materials and methods

2.1. Synthesis and characterization of CMA

CMA was synthesized by reacting type I collagen with methacrylic anhydride (MAA, Aladdin) at pH 8–9 for 24 h, using a 20:1 M ratio of MAA to collagen to ensure a high degree of amino substitution. The reaction mixture was dialyzed for 7 days and subsequently freeze-dried. CMA (10 mg/mL) and collagen (10 mg/mL) were dissolved in 10 mM HCl and analyzed using a Bruker AVANCE III 600 spectrometer to obtain ¹H NMR spectra.

The degree of substitution (DS) of CMA was evaluated using the 2,4,6-trinitrobenzene sulfonic acid (TNBS) assay. Three batches of CMA were selected for analysis. The procedure was performed as follows. Briefly, 11 mg of either collagen type I (Col I) or CMA was added into a 10 mL centrifuge tube. Then, 1 mL of 4 % NaHCO3 solution and 1 mL of 0.5 % TNBS solution were added. These samples served as the experimental group (CMA) and the control group (Col I), respectively. For the blank group, 11 mg of Col I was mixed with 3 mL of 6 M HCl, along with 1 mL of 4 % NaHCO3 solution and 1 mL of 0.5 % TNBS solution. All three groups were incubated in a shaker at 40 °C and 220 rpm for 4 h. After this initial reaction, 3 mL of 6 M HCl was added to both the experimental and control groups. Subsequently, all samples were incubated at 60 °C and 220 rpm for 1.5 h. After the reaction, the samples were diluted and extracted with diethyl ether. Excess ether was removed by evaporation in a 60 °C water bath. The absorbance of the samples was measured using a UV–Vis spectrophotometer. The absorbance values were recorded as follows: ODb for the blank group, ODs for the experimental group, and ODc for the control group.

The degree of substitution (DS) of CMA was calculated using the following formula:

$$DS(\%)=100\%-\frac{\left(ODs-ODb\right)}{\left(ODc-ODb\right)}\times 100\%$$ (1)

2.2. Preparation of bioinks

CMA was dissolved in acetic acid following sterilization with UV and 75 % ethanol, and the pH was adjusted to neutral. For cell and animal experiments, the samples were dissolved in DMEM containing acetic acid. DHM (0.01 %–0.04 %), the photoinitiator LAP (0.125 %–0.5 %), and the UV absorber tartrazine (0.025 %–0.1 %) were then added to the CMA solution, which was subsequently centrifuged to produce acellular CMA-DHM bioinks. For cell-laden bioinks, cells were cultured to 90 % confluency, digested with 0.25 % trypsin, centrifuged to obtain cell pellets, and resuspended in the bioink. Both cell and animal experiments utilized 0.01 % DHM.

2.3. Rheological characterization

In situ gelation behaviors of CMA and CMA-DHM bioinks were assessed on a rheometer (Anton Paar, MCR 302) using a 15 mm diameter parallel plate and a 405 nm light for a 70-s time sweep oscillation test (γ = 1 %, ω = 1 rad/s, gap 0.7 mm) to monitor the modulus changes. Oscillatory strain-sweep from 0.2 % to 20 % was tested (ω = 1 rad/s). Additionally, the dynamic viscosity was measured at shear rates ranging from 0.1 to 100 s⁻¹ (γ = 1 %, ω = 1 rad/s).

2.4. Mechanical test

Compression test was conducted on cylindrical hydrogel samples (10 mm × 10 mm) using a universal testing machine (Shimadzu, AGX-V500N, Japan) with a 20 mm/min displacement rate. At least three specimens were tested.

2.5. Circular dichroism (CD)

CD spectra (JASCO, J1500) of 1 mg/mL CMA and CMA-DHM solutions were obtained after equilibrating samples at 4 °C for 24 h. Measurements were performed using a 1 mm path length cuvette, with a scan range of 180–260 nm, a step size of 1.0 nm, and an averaging time of 1.0 s.

2.6. Swelling and degradation

The swelling ratio was calculated by measuring the weight of freeze-dried CMA and CMA-DHM gels after water absorption at different time points. For the degradation test, hydrogels were incubated in a 0.5 U/mL type I collagenase solution (0.05 M Tris-HCl buffer, pH 7.4, 37 °C) for 10 days. Samples were taken out at different time points, freeze-dried, and weighed to calculate the mass retention rate.

2.7. FTIR and FESEM characterization

FTIR (Nicolet, NEXUS 670, USA) was performed to analyze the chemical structure of CMA and CMA-DHM hydrogels. FESEM (FESEM Hitachi Limited, Japan) was utilized to observe the microstructure of freeze-dried CMA-DHM hydrogels.

2.8. DLP-based 3D printing

CMA-DHM bioink (DHM 0.01 %–0.04 %) with optimal printing parameters (15–25 mW/cm² of light intensity, 25–100 μm of slice thickness, 20–30 s of exposure time) was employed to print a comb model (1–5 mm tooth spacing), a flower model, a hollow scaffold, and a 100-pore mesh style model using a DLP 3D printer (EFL, BP8600, China) with a 405 nm light source. All the printed hydrogels were subjected to secondary curing under 405 nm light to ensure complete gelatin post-printing. The hydrogels were dyed with colors to facilitate the observation and then pictured.

2.9. Cell encapsulation biocompatibility

Cell-laden bioinks containing L929 fibroblasts were cured in a confocal dish (15 mm diameter) under 405 nm light for 30 s and then incubated in growth medium (DMEM, 5 % FBS, 1 % antibiotics) with the medium changed daily. On days 1, 4, and 7, samples were washed with 1 × AB buffer and subjected to Live/Dead staining (Calcein-AM/PI kit, Solarbio, CA1630, China). Images were captured using confocal laser scanning microscopy (CLSM) (Olympus, FV3000, Japan), and cells were counted using Image J.

2.10. Cell-laden bioprinting

A 10 × 10 mm mesh scaffold was printed using L929-laden CMA and CMA-DHM bioinks. Post-printing, scaffolds were cured for 20 s, washed, and incubated in the growth medium at 37 °C. CLSM images of CMA and CMA-DHM hydrogels were captured after 24 h of incubation and analyzed using Image J software. On days 1, 4, and 7, CMA-DHM hydrogels underwent Live/Dead staining. The CLSM image of CMA-DHM on day 1 was merged with its CAD model to assess printing accuracy. Live cell counts on days 1, 4, and 7 were conducted using Image J software. Swelling behaviors were assessed by measuring the width of the mesh scaffold. Additionally, six HUVEC-loaded models were also printed.

2.11. Antioxidant capacity of hydrogels

100 μL of human dermal fibroblasts (HDFs) at a density of 1 × 10⁶ cells/mL were seeded on cylindrical CMA and CMA-DHM hydrogels and incubated overnight. After cell adhesion, a fresh culture medium was added and incubated for 3 days. To observe cell proliferation and spreading, hydrogels were stained with rhodamine-phalloidin and Hoechst 33342. For antioxidant assessments, CMA and CMA-DHM hydrogels were treated with 88 μM (3.0 μg/mL) H₂O₂ solution for 40 min. This concentration was chosen to simulate the elevated oxidative stress commonly observed in diabetic wounds [36]. Untreated CMA was set as the blank control. 10 μM DCFH-DA (Solarbio, China) was added and incubated in the dark for 60 min. Fluorescent images were captured by CLSM and semi-quantitative analysis was performed using Image J. Additionally, the same procedures were conducted in a 96-well plate and the OD value at 529 nm was examined using a microplate reader (Tecan, Männedorf, Switzerland) to quantify the ROS level. Moreover, Live/Dead staining was performed to determine the ability of CMA-DHM hydrogels to improve cell viability under high oxidative stress.

2.12. Macrophages polarization by hydrogels

RAW 264.7 cells (1 × 10^5 cells/mL) were seeded in CLSM dishes (15 mm) and incubated for 24 h. Cells were treated with 0.1 mg/mL LPS for 12 h to induce M1 polarization. Subsequently, culture medium (Blank), 0.2 mg/mL CMA, and CMA-DHM hydrogels were added and incubated for 48 h. M1 sample (0 h) and the treated samples were stained for CD86 (M1 marker) and CD206 (M2 marker). Samples were washed with PBS, fixed, permeabilized, and blocked. Primary antibodies were incubated overnight at 4 °C, followed by secondary antibodies for 1 h at 37 °C. Fluorescent images were captured using CLSM under identical settings.

2.13. Design and fabrication of biomimetic skin

A bilayer porous artificial skin model with a diameter-to-height ratio of 9:2.5 was designed using AutoCAD software. For animal experiments, CMA, CMA-DHM and CMA-DHM@HDFs (HDFs-laden skin constructs) measuring 9 mm × 2.5 mm were fabricated via DLP 3D printing (layer thickness 25 μm, exposure time 16 s). After 14 days of in vitro culture, Live/Dead staining of CMA-DHM@HDFs hydrogels was performed and CLSM images were captured.

2.14. Chronic full-thickness wound model using diabetic rats

The animal protocol was approved by the Ethics Committee of the College of Chemistry and Chemical Engineering at Lanzhou University (Approval No. G09, 20220711). Male Sprague–Dawley (SD) rats (150–200 g) were used to establish a streptozotocin (STZ)-induced diabetic model via a single intraperitoneal injection of STZ (65 mg/kg; Macklin, Shanghai).

A total of 60 diabetic rats were randomly assigned to four groups (n = 15 per group): Control, CMA, CMA-DHM, and CMA-DHM@HDFs. An additional 5 non-diabetic rats were used to establish an acute wound model as healthy controls.

To ensure a stable diabetic state, blood glucose levels were monitored regularly after STZ injection. Rats exhibiting sustained hyperglycemia (blood glucose >16.7 mmol/L) for three consecutive weeks were considered eligible for subsequent wound modeling. This 3-week stabilization period was selected based on previous studies, which have shown that sustained hyperglycemia contributes to the reliability and consistency of diabetic wound models. While Cai et al. [34] and Chao et al. [35] used 15-day and 2-week stabilization periods, respectively, we extended this duration to 3 weeks to ensure a more robust and stable hyperglycemic state before wound creation, as Lee Y H et al. [36] reported.

Following confirmation of stable hyperglycemia, two full-thickness circular skin wounds (9 mm in diameter) were symmetrically created on the dorsal surface of each rat. Printed hydrogels including CMA, CMA-DHM, and CMA-DHM@HDFs hydrogels, were implanted into the wound sites of the respective treatment groups, while untreated wounds served as internal controls. All wounds were secured with Tegaderm^TM transparent dressings on day 0. Notably, wounds showed noticeable infection on days 7 and 10 in the diabetic control group. To prevent mortality from severe infection, all rats were disinfected with povidone-iodine, especially around the wound area. Wounds were photographed on days 0, 4, 7, 10, 15, and 18, and wound trace and closure rates were analyzed using Image J.

2.15. Histological analysis

On days 7 and 18, wounded tissues were collected and paraffin-embedded, with sections cut to a thickness of 5 μm. The sections were then subjected to H&E and Masson's trichrome staining and imaged using a digital panoramic scanning system (3DHISTECH/Pannoramic DESK, Hungary). Epidermal thickness was measured using CaseViewer software, and collagen deposition was calculated using ImageJ software.

2.16. ROS level and immunofluorescence staining of wound healing progress

Wound tissues from days 4, 7, and 18 were cryosectioned at a thickness of 10 μm. Tissues from days 4 and 7 were stained with 10 μM DCFH-DA for 60 min and washed with PBS to determine the ROS level in diabetic wounds during the inflammatory phase. Immunofluorescence staining was performed to evaluate the inflammatory response on day 7 and angiogenesis on day 18. The sections were incubated with primary antibodies against CD163 and iNOS (inflammatory markers) and CD31 and α-SMA (angiogenesis markers) overnight at 4 °C, and then treated with fluorescently labeled secondary antibodies, counterstained with DAPI to visualize cell nuclei, and mounted with an anti-fade medium. Fluorescent images were captured using CLSM, and fluorescence intensity was semi-quantitatively analyzed using Image J software.

2.17. Statistical analysis

For data analysis and statistical evaluation, Origin 2019 software (OriginLabs) was utilized. All data were presented as mean ± SD from at least 3 independent measurements. Statistical analyses were conducted using a two-tailed Student's t-test or one-way ANOVA with Tukey's post-hoc test. A p-value less than 0.05 was considered statistically significant. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ns = no statistical difference.

3. Results and discussions

3.1. Printability of novel CMA-DHM bioinks

Methacrylated collagen (CMA) is an emerging photo-curable bioink with substantial potential in biomedical applications. However, its mechanical limitations present challenges for precise DLP printing [24]. To overcome these challenges, we incorporated dihydromyricetin (DHM) as a novel crosslinker to enhance the mechanical robustness and printability of CMA (Fig. 2). CMA was synthesized under alkaline conditions and confirmed by ¹H NMR (Fig. S1) and TNBS results (Fig. S6), demonstrating a high degree of amino substitution (>90 %). DHM (0.01 %–0.04 %) was integrated into the CMA solution (0.75 %–1.0 %) along with a photoinitiator (LAP) and a UV absorber (UA) to formulate photo-curable CMA-DHM bioinks.

Fig. 2.

Printability of CMA-DHM bioinks. (a) Time-dependent photo-curing rheological characterization of CMA and CMA-DHM bioinks. (b) Circular dichroism spectrum of CMA and CMA-DHM bioinks. (c) Curing time of inks with different LAP concentrations. (d) Curing time of inks with different UA concentrations. (e) CAD model of a comb structure. (f) Comb structure after DLP printing using CMA-DHM bioinks. (g) The Rose model and multi-angle images after printing. (h) Stereoscopic hollow scaffold model and multi-angle images after printing. (i) Photographs of the stereoscopic hollow scaffold before and after compression. (j) Photos of the printed mesh scaffold with 100 holes. (k) SEM characterization of the surface and cross-section of the printed gel.

Time-dependent rheological behaviors of CMA and CMA-DHM bioinks were monitored to evaluate the kinetics of the photo-curing process (Fig. 2a). Both bioinks exhibited a progressive increase in modulus upon exposure to 405 nm light. The mechanical properties of CMA and CMA-DHM hydrogels are summarized in Table S1. Notably, CMA-DHM bioinks achieved a storage modulus of 1000 Pa within 8 ± 1 s, significantly outpacing the prolonged 30 ± 2 s required for CMA bioinks, underscoring the faster gelation speed (375 %-fold increase) of CMA-DHM bioinks. The final stable storage modulus of CMA-DHM hydrogels reached 1745.5 ± 4.51 Pa, remarkably higher than the 1084 ± 5.3 Pa observed for CMA hydrogels. This 161 % enhancement in viscoelasticity is likely attributable to the additional DHM crosslinking, making multi-crosslinked CMA-DHM bioinks potentially more effective for 3D printing by enhancing mechanical integrity.

Circular dichroism (CD) spectroscopy was employed to assess the secondary structure of CMA and CMA-DHM bioinks (Fig. 2b). CMA-DHM bioinks exhibited characteristic peaks at 222 nm and 198 nm, similar to those of CMA bioinks, affirming the preservation of the collagen triple-helical structure [25]. These findings suggest that the incorporation of DHM did not disrupt the native collagen structure of CMA.

The optimal formulation of CMA-DHM bioinks was investigated by examining gelation time via rheological characterization (Fig. 2c and d). LAP served as a photoinitiator to produce free radicals that initiate polymerization upon UV light exposure. Increasing concentrations of LAP (0.125 %–0.5 %) substantially reduced gelation time from 46 s to 15.8 s, thereby enhancing printing efficiency (Fig. 2c). However, due to cytotoxic concerns associated with excessive LAP content, a concentration of 0.25 % LAP was deemed optimal, balancing efficacy and biocompatibility. To achieve high-resolution printing of 3D constructs, a UV absorber (UA) was integrated into the bioinks. UA mitigated light scattering and prevented the over-curing of layers beyond the intended focal plane [26]. Incremental increases in UA concentration (0.025 %–0.1 %) extended gelation time from 11.3 s to 93.5 s (Fig. 2d). To achieve high printing resolution and balance gelation efficiency, a concentration of 0.05 % UA was identified as desirable.

The printability of dual-crosslinked CMA-DHM bioinks was evaluated by fabricating diverse 3D models using a DLP printer. A comb model designed with varying interdental spacing from 1 mm to 5 mm (Fig. 2e) exhibited precise printing outcomes, with the smallest tooth spacing discernible (Fig. 2f). A 1.5 cm tall flower model showcased intricate details such as petal morphology and detailed hollow interiors (Fig. 2g). Moreover, the ease of handling the printed rose with both fingertips and nozzles highlights its robust structural integrity. Additional fabrication of hollow scaffold models revealed distinct three-dimensional mesh patterns and the easy insertion of a transparent tube confirmed no over-curing mesh printing (Fig. 2h). Compression tests demonstrated the gel's remarkable resilience to regain its original shape within 10 s of compression (Fig. 2i). High-resolution printing was further demonstrated by the clearly defined 100 mesh squares in a mesh-style scaffold (Fig. 2j). In summary, these findings underscore the superior formality, printability, structural integrity, and high-fidelity capabilities of CMA-DHM bioinks for DLP 3D printing.

Furthermore, scanning electron microscopy (SEM) analysis was conducted to investigate the micromorphology of the printed CMA-DHM hydrogels (Fig. 2k). Both the surface and vertical cross-section of the printed gel exhibited an interconnected porous architecture, highlighting the effective crosslinking of CMA-DHM bioinks. The porous microstructure is crucial in tissue engineering as it facilitates cell proliferation and nutrient exchange. These results highlight the novel CMA-DHM bioink as a promising candidate for DLP 3D bioprinting.

3.2. Crosslinking mechanisms and characterization of CMA-DHM hydrogels

FTIR analysis was conducted to elucidate the crosslinking mechanisms of DHM, which was employed as an innovative crosslinking agent for CMA due to its identity as a polyphenol derivative with three active phenolic hydroxyl groups (Fig. 3a). The FTIR spectra exhibited the distinctive peaks of collagen within CMA, notably the amide A band at 3342 cm⁻¹(associated with N-H stretching vibrations), the amide B band at 3080.2 cm⁻¹ (related to C-H and -NH₃⁺ stretching vibrations), the amide I band at 1660.4 cm⁻¹(involving C=O stretching and N-H bending vibrations), the amide II band at 1561.5 cm⁻¹ (including N-H bending and C-N stretching vibrations), and the amide III band at 1237.6 cm⁻¹(comprising C-N and N-H vibrations). Notably, in the CMA-DHM hydrogel, the amide A, B, and I bands shifted to 3344.4 cm⁻¹, 3076 cm⁻¹, and 1663.7 cm⁻¹, respectively, indicating that the amino and carboxyl groups in CMA underwent reactions with DHM, probably involving hydrogen bond interactions as proposed (Fig. 3g).

Fig. 3.

Crosslinking mechanisms and characterization of CMA-DHM hydrogels. (a) FTIR spectra of CMA and CMA-DHM hydrogels. (b) Strain-dependent storage modulus (G′) and loss modulus (G″) of CMA and CMA-DHM hydrogels. (c) Viscosity characterization of CMA and CMA-DHM hydrogels across shear rates ranging from 0.1 to 100 s¹. (d) Compression stress-strain curves of CMA and CMA-DHM hydrogels. (e) Swelling ratios of CMA and CMA-DHM hydrogels. (f) Weight percentages of CMA and CMA-DHM hydrogels in collagenase over 10 days. (g) Schematic representation of the proposed non-covalent crosslinking mechanism between CMA and DHM. (h) Schematic representation of the proposed covalent crosslinking mechanism between CMA and DHM.

Furthermore, the amide I band exhibited broadening, accompanied by a new peak at 1637 cm⁻¹, indicative of C=N stretching vibrations possibly from Schiff's base formation. The peak at 1437.6 cm⁻¹ shifted and intensified, likely due to enhanced C-H bending or C-N stretching vibrations resulting from covalent bonding facilitated by DHM crosslinking. Additionally, two prominent absorption peaks at 1025.5 cm⁻¹ and 955.5 cm⁻¹ were observed, probably resulting from DHM's C-O-C stretching vibrations and new covalent bonds between DHM and CMA as illustrated (Fig. 3h). FTIR results underscored that DHM effectively crosslinked CMA through a combination of non-covalent and covalent crosslinking, as reported polyphenols that engaged with proteins via diverse interactions [27,28].

In the multi-crosslinking strategy of the CMA-DHM system, free radical polymerization initiates the network formation by generating covalent bonds between the collagen polymer chains, specifically through the reaction of methacrylate groups (-C=C) in CMA with radicals generated by the photoinitiator (LAP). This leads to the formation of covalent crosslinks, providing fast gelation and robust mechanical properties. Additionally, Michael addition occurs when the amine groups (-NH₂) in DHM react with the maleimide groups (-C=C) in CMA, forming covalent linkages that significantly enhance the mechanical strength of the network. This reaction contributes to the structural stability of the hydrogel. Furthermore, Schiff base formation takes place when the phenolic hydroxyl groups (-OH) in DHM are oxidized and react with the amine groups (-NH₂) in CMA, forming imine linkages (-C=N-), which further improve the network's mechanical integrity. The combination of covalent and non-covalent crosslinking mechanisms ensures that the CMA-DHM bioink exhibits superior mechanical stability and maintains excellent cell compatibility, which is crucial for its application in DLP-based 3D printing.

Efficient multi-crosslinking between CMA and DHM significantly enhanced the physicochemical properties of CMA hydrogels, as evidenced by comprehensive analyses encompassing viscoelasticity, dynamic viscosity, compressive strength, swelling behavior, and biodegradation (Fig. 3b–f). The storage modulus (G′) and loss modulus (G″) of CMA-DHM hydrogels exhibited marked elevation compared to single-crosslinked CMA hydrogels under various strains, indicative of improved viscoelastic robustness (Fig. 3b). Similarly, dynamic viscosity results demonstrated substantially higher viscosity in CMA-DHM hydrogels relative to CMA hydrogels. Both hydrogels displayed decreased viscosity with increasing shear rate from 0.1 to 100 s⁻¹, illustrating pronounced shear-thinning behavior (Fig. 3c). Compression stress-strain curves further validated superior compressive strength at break in CMA-DHM hydrogels (39.5 ± 5.8 kPa), higher than that of the CMA hydrogels (17.1 ± 3.69 kPa), underscoring significant mechanical enhancements of 231 ± 60 % facilitated by DHM crosslinking (Fig. 3d). Swelling tests indicated a reduced swelling ratio in CMA-DHM hydrogels (53 %) compared to CMA hydrogels (58 %), suggesting that DHM crosslinking likely augmented the density of crosslinked networks (Fig. 3e). Moreover, CMA-DHM hydrogels demonstrated a higher mass residue rate in collagenase solution across various time intervals. After 10 days of incubation, CMA-DHM hydrogels retained 45.86 ± 1.31 % of their weight, compared to only 22.04 ± 1.63 % retention for CMA hydrogels, highlighting significantly enhanced resistance (208 ± 17 %) to biodegradation (Fig. 3f). These findings collectively highlight the remarkable mechanical enhancement of CMA hydrogels achieved through the dual-crosslinking strategy, ensuring stable mechanical integrity post-printing and positioning CMA-DHM hydrogels as advanced gel systems for biomedical applications.

3.3. Enhanced cell-laden biocompatibility and printability of CMA-DHM bioinks than CMA bioinks

To evaluate cell encapsulation biocompatibility, Live/Dead staining was conducted on hydrogels incorporating L929 fibroblasts (Fig. 4a and b). Cells at a density of ∼1 × 10⁷ cells/mL were integrated into growth media dissolved in CMA and CMA-DHM bioinks. These mixtures were exposed to 405 nm light for 30 s to form hydrogels, followed by a 7-day incubation period. Despite the high degree of amino substitution in CMA, neither CMA nor CMA-DHM hydrogels exhibited a significant presence of dead cells, underscoring the superior cytocompatibility of these bioink systems (Fig. 4a). Notably, cells in CMA-DHM hydrogels demonstrated significantly higher proliferation both on days 4 and 7 compared to the CMA group (Fig. 4b). These results demonstrate the higher cell-laden biocompatibility of CMA-DHM bioinks compared to CMA bioinks.

Fig. 4.

Cell-laden biocompatibility and printability of CMA-DHM bioinks are better than CMA bioinks. (a) Live/Dead staining images of cell-laden CMA and CMA-DHM hydrogels after 1, 4, and 7 days of culture. (b) Quantification of live cells encapsulated in CMA and CMA-DHM hydrogels after 1, 4, and 7 days of culture. (c) TD image of cell-laden printed mesh scaffold using CMA bioinks after 1 day of incubation and treated with Image J to enhance contrast. Scale bar: 500 μm. (d) TD image of cell-laden printed mesh scaffold using CMA-DHM bioinks after 1 day of incubation and treated with Image J to enhance contrast. Scale bar: 500 μm. (e) Gray values of the green line in the image (c) analyzed by Image J. (f) Gray values of the green line in the image (d) analyzed by Image J. (g) CAD model, CLSM image, and their merged image of the printed cell-laden CMA-DHM mesh scaffold on day 1. (h) CLSM image of the printed cell-laden CMA-DHM mesh scaffold on day 4. (i) CLSM image of the printed cell-laden CMA-DHM mesh scaffold on day 7. (j) Average grid width of printed cell-laden CMA-DHM mesh scaffold after 1, 4, and 7 days of culture. (k) Fluorescence cell counts of the printed cell-laden CMA-DHM mesh scaffold on days 1, 4, and 7. ∗∗p < 0.01; ∗∗∗p < 0.001; ns = no significant differences. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

To investigate the printability of cell-laden bioinks, fibroblast-containing CMA and CMA-DHM bioinks were processed using the DLP bioprinter to fabricate mesh scaffolds (Fig. 4c–f). After a 1-day incubation, the printed constructs were imaged using confocal laser scanning microscopy (CLSM). The brightfield images revealed clear boundaries of CMA-DHM hydrogels, demonstrating high printing resolution (Figs. S2b and 4d), whereas significant over-curing was observed in CMA scaffolds (Figs. S2a and 4c). To quantify the printing quality, gray values of the green lines across eight squares of the scaffolds in Fig. 4c and d were examined. Uneven gray values between squares indicated unstable printing resolution for CMA, despite using optimal printing settings (Fig. 4e). In contrast, the gray values of eight squares in CMA-DHM scaffolds showed no significant differences, indicating uniform printing quality and high printing resolution of CMA-DHM bioinks.

Moreover, Live/Dead-stained CMA-DHM constructs were analyzed (Fig. 4g–k). The day 1 CLSM image was merged with the CAD model, revealing that the printed cell-laden hydrogels closely matched the model in terms of both size and pattern (Fig. 4g). Additionally, no dead cells were observed, and live cells (stained green) were uniformly dispersed throughout the printed scaffolds, demonstrating superior cell viability during the printing process. Notably, after incubation in the cell growth medium, smaller pores were observed, and the grid width increased on both day 4 (Fig. 4h) and day 7 (Fig. 4i), primarily due to gel swelling and potentially influenced by cell growth. However, there was no significant difference in the measured grid width (Figure S2 and Fig. 4j), indicating a low swelling ratio as determined in Fig. 3e. The low swelling behavior of the CMA-DHM system benefits the maintenance of shape fidelity during long-term in vitro incubation [29]. Furthermore, fibroblasts in hydrogels proliferated on day 4 and day 7 (Fig. 4k), indicating that the bioinks protect the cells from UV exposure during DLP bioprinting and promote cell proliferation. These results demonstrate the high printing resolution, cell viability, shape fidelity, and mechanical stability of CMA-DHM bioinks for efficient cell-laden bioprinting.

In addition to L929 cells, HUVECs (10⁶ cells/mL) were also employed for cell-laden bioprinting, resulting in the precise printing of six diverse models (Fig. S3). These findings collectively demonstrated that CMA-DHM bioinks possess enhanced cell-laden biocompatibility and printability compared to CMA bioinks, making them exceptionally desirable for tissue engineering.

3.4. Superior antioxidant efficiency of CMA-DHM hydrogels

The antioxidant capacity of CMA-DHM hydrogels was assessed through cell-based experiments (Fig. 5). Human dermal fibroblasts (HDFs) were cultured on CMA and CMA-DHM hydrogels for three days to prepare samples for antioxidant testing. To visualize cell growth, immunofluorescence staining was performed. HDFs exhibited significantly higher confluency and more elongated cytoskeletal spreading on CMA-DHM hydrogels compared to CMA hydrogels, providing evidence of the superior efficacy of the CMA-DHM system for 3D cell culture (Fig. 5c). Moreover, quantification analysis demonstrated that HDFs' proliferation on CMA-DHM hydrogels was double that observed on CMA hydrogels (Fig. 5b). These results indicate the enhanced bioactivity of CMA-DHM hydrogels, likely attributable to the potential reactive oxygen species (ROS) scavenging properties of DHM.

Fig. 5.

Antioxidant and anti-inflammatory performance of CMA-DHM hydrogels. a) Schematic representation of the antioxidant assays. CMA-DHM hydrogels exhibited sustained antioxidant properties, creating a ROS-free 3D cell culture microenvironment that promotes cell survival and proliferation. b) Quantification of HDFs after 3 days of incubation on CMA and CMA-DHM hydrogels. ∗∗∗p < 0.001. n = 4. c) Cytoskeleton staining of HDFs after 3 days of growth on CMA and CMA-DHM hydrogels. Scale bar: 50 μm. d) ROS levels in H₂O₂-treated HDFs cultured on Control, CMA, and CMA-DHM hydrogels, stained with DCFH-DA. Scale bar: 100 μm. e) Live/Dead staining of H₂O₂-treated HDFs on Control, CMA, and CMA-DHM hydrogels. Scale bar: 100 μm. f) Fluorescence intensity of DCFH-DA-stained cells on Control, CMA, and CMA-DHM hydrogels. ∗∗∗p < 0.001. n = 6. g) Relative content of dead cells in H₂O₂-treated Control, CMA, and CMA-DHM hydrogels. ∗∗∗p < 0.001. n = 3. h) Quantification of CD86 positive fluorescence intensity per cell. ∗p < 0.05, ∗∗p < 0.01, ns > 0.05, n = 3. i) Quantification of CD206 positive fluorescence intensity per cell. ∗p < 0.05, ∗∗p < 0.01, n = 3. j) Representative CLSM images of CD86 (M1 marker) immunofluorescence staining in LPS-activated M1 macrophages at 0 h and DMEM (blank), CMA, and CMA-DHM hydrogel-treated M1 cells at 48 h. Scale bars: 50 μm. k) Representative CLSM images of CD206 (M2 marker) immunofluorescence staining in LPS-activated M1 macrophages at 0 h and DMEM (blank), CMA, and CMA-DHM hydrogel-treated M1 cells at 48 h. Scale bars: 50 μm. l) Schematic illustration of CMA-DHM hydrogels potentially facilitating the transition of diabetic wounds from a pro-inflammatory to an anti-inflammatory state via efficient immunoregulation and ROS management, thereby accelerating diabetic wound healing.

To assess the antioxidant performance of CMA-DHM hydrogels, DCFH-DA assays and Live/Dead staining were performed. Hydrogels were treated with H₂O₂ to induce ROS production, thereby simulating a high oxidative stress microenvironment comparable to that in diabetic wounds [37]. Post-treatment, HDFs on CMA hydrogels exhibited intense DCFH-DA fluorescence, indicating elevated intracellular ROS levels. In contrast, cells on CMA-DHM hydrogels showed significantly weaker fluorescence intensity compared to both CMA and untreated control groups (Fig. 5d). This observation was corroborated by semi-quantitative fluorescence cell counts (Fig. S4) and the measured DCFH-DA fluorescence intensity at 529 nm (Fig. 5f). These results highlight the exceptional ROS scavenging ability of CMA-DHM hydrogels, which substantially release antioxidant DHM to capture ROS (Fig. 5a).

Live/Dead cell staining was further performed to assess cell viability on different hydrogels exposed to high ROS levels (Fig. 5e). A significant number of dead cells were observed in the CMA group, while minimal cell death occurred in the untreated control, indicating that high oxidative stress significantly induced cell death. However, negligible red cells were observed on H₂O₂-treated CMA-DHM hydrogels, attributed to their sustained ROS scavenging ability, which protected cells from free radical damage. Further semi-quantitative analysis indicated a markedly lower proportion of dead cells in the CMA-DHM group compared to other groups (Fig. 5g). These findings emphasize the crucial role of ROS scavenging in enhancing cell survival under high oxidative stress.

In summary, the superior antioxidant properties of CMA-DHM hydrogels make them particularly suitable for cell-laden 3D bioprinting, which relies on free radical polymerization. The sustained release of DHM from CMA-DHM hydrogels effectively mitigates excess free radicals generated during the bioprinting process, thereby creating a ROS-free 3D cell culture microenvironment that is highly conducive to cell survival, spreading, and proliferation (Fig. 5a). Furthermore, the exceptional antioxidant capacity of CMA-DHM hydrogels also renders them highly desirable for treating hard-to-heal diabetic wounds, which are often characterized by high oxidative stress and a prolonged inflammatory phase [30].

3.5. Efficient M1-to-M2 macrophage polarization by CMA-DHM hydrogels

The anti-inflammatory property was assessed by determining the efficacy of M1-to-M2 macrophage polarization using Raw 264.7 macrophages. Macrophages, fundamental to wound healing, exhibit the capacity to polarize into either pro-inflammatory (M1) or anti-inflammatory (M2) phenotypes, with M2 macrophages playing a crucial role in modulating inflammation resolution [31]. To evaluate the induction of the M1-to-M2 phenotype transition by CMA-DHM hydrogels, immunofluorescence staining was employed (Fig. 5j and k). CLSM images demonstrated that macrophages stimulated with lipopolysaccharide (LPS) displayed pronounced CD86 expression, characteristic of the M1 phenotype (0 h) (Fig. 5j). Following a 48-h incubation, there was a marked reduction in CD86 expression in cells treated with both hydrogels compared to the DMEM-treated group. Conversely, the expression of CD206, a definitive marker of the M2 phenotype, was significantly elevated. Notably, macrophages treated with CMA-DHM hydrogels exhibited a pronounced transition to an elongated CD206-positive M2 phenotype, in stark contrast to the limited morphological changes observed in CMA-treated cells (Fig. 5k). These results underscored the potent facilitation of M1-to-M2 macrophage polarization efficiency of CMA-DHM hydrogels.

Persistent hyperglycemia and excessive advanced glycation end products (AGEs) in diabetic wounds significantly impair the transition of M1 macrophages to the M2 phenotype. This impairment leads to the sustained activation and infiltration of pro-inflammatory M1 cells, which express inflammatory mediators at the wound site, causing prolonged inflammation [32]. CMA-DHM treatment effectively promotes the polarization of M1 to M2 macrophages, thereby facilitating the shift from the pro-inflammatory stage to the anti-inflammatory phase. Combined with its superior bioactivity and antioxidant properties, CMA-DHM hydrogels demonstrate significant potential as a novel therapeutic approach for accelerating diabetic wound healing (Fig. 5l).

3.6. Design and fabrication of biomimetic skin constructs for diabetic wound implantation

Inspired by the native skin structure, a meticulous bilayer multi-channel scaffold model with a diameter-to-height ratio of 9:2.5 was designed using AutoCAD software (Fig. 6a). Bio-ink was prepared and incorporated with HDFs, followed by DLP 3D printing to fabricate cell-laden CMA-DHM@HDFs hydrogels as biomimetic skin constructs (Fig. 6b). After 14 days of in vitro culture, the printed scaffold maintained exceptional structural integrity and high resolution (Fig. S5a and b). Live/Dead cell staining revealed extensive cellular proliferation within the hydrogel, with minimal cell death observed after the 14-day culture period (Fig. S5c and d), underscoring its potential wound-healing application as dermal equivalents. A diabetic wound model was established using STZ injection to evaluate the healing efficacy of the biomimetic skin constructs. A full-thickness skin defect with a diameter of 9 mm was created. The control group received no scaffold treatment, while the cultured CMA, CMA-DHM, and CMA-DHM@HDFs hydrogels were implanted into the wounds. To serve as acute wound controls, wound models were also established in healthy non-diabetic rats. Tegaderm™ transparent dressings were utilized to secure the wounds in all groups on day 0, while hydrogels were supplemented if detached.

Fig. 6.

Bioprinted cell-laden CMA-DHM biomimetic skin constructs for diabetic wound healing. a) Design of the multi-channel biomimetic skin scaffold model. b) Schematic of DLP 3D printing, in vitro culture, and wound treatments of cell-laden multi-channel scaffolds. CMA, CMA-DHM, and CMA-DHM@HDFs hydrogels were all printed in the designed scaffold pattern, cultured for 24 h, and implanted in the diabetic wounds. c) Representative images of wounds at different time points for Control, CMA, CMA-DHM, CMA-DHM@HDFs, and acute wound groups. d) Overlaid wound trace of Control, CMA, CMA-DHM, CMA-DHM@HDFs, and acute wound groups processed by Image J. e) H&E staining images of Control, CMA, CMA-DHM, and CMA-DHM@HDFs groups on day 18. Regenerated skin appendages were noted by green arrows. Neoepidermis were noted by blue arrows. f) Quantitative analysis of wound closure rates of Control, CMA, CMA-DHM, CMA-DHM@HDFs, and acute wound group analyzed by Image J. g) Counted skin appendages of Control, CMA, CMA-DHM, and CMA-DHM@HDFs groups on day 18. ∗∗p < 0.01, n = 4. h) Measured epidermal thickness of neoepidermis and normal epidermis in Control, CMA, CMA-DHM, and CMA-DHM@HDFs groups on day 18. ∗∗p < 0.01; ∗∗∗p < 0.001, ns > 0.05, n = 10. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Photographs of the wounds were taken on days 0, 4, 7, 10, 15, and 18 to monitor the healing process (Fig. 6c). Compared to the acute wound group, the diabetic control group exhibited significant delays in wound healing at every time point. This delay was attributed to increased oxidative stress, prolonged inflammation, and higher susceptibility to infection in diabetic wounds [33]. By day 10, while acute wounds showed further contraction (green notes), those in the diabetic control group deteriorated (red notes), likely due to infection, resulting in larger wound areas (Fig. 6d) and decreased healing rates (Fig. 6f). In contrast, the treated groups displayed significantly smaller wound areas and faster wound closure rates at various time points, demonstrating the effectiveness of the scaffold treatments. Among these, the CMA-DHM group showed a markedly accelerated wound closure rate compared to CMA treatments (Fig. 6f). Notably, the cell-laden CMA-DHM@HDFs group achieved the highest wound closure rate, with closure exceeding 85 % by day 10 and nearly complete healing by day 18, significantly faster than the other groups, indicating the efficacy of CMA-DHM@HDFs hydrogel in accelerating diabetic wound closure.

Histological analysis utilizing H&E staining was performed on skin sections harvested after 18 days of healing (Fig. 6e). The untreated control group exhibited classical signs of the inflammatory phase, including incomplete epidermal regeneration and substantial infiltration of inflammatory cells, indicative of prolonged inflammation. The CMA group showed a pronounced enhancement in re-epithelialization; however, the dermal layer displayed delayed regeneration, evidenced by extensive granulation tissue. In contrast, wounds treated with CMA-DHM demonstrated a regenerated epidermis and the presence of skin appendages within a significantly reduced granulation tissue area, indicating notably accelerated healing compared to the CMA group. Furthermore, the HDF-laden CMA-DHM skin constructs presented fully regenerated skin tissue with negligible granulation tissue, suggesting the efficacy of fibroblast-laden multifunctional therapy in repairing full-thickness diabetic wounds. Quantitative analysis of epidermal thickness revealed no significant differences between the neoepidermis and adjacent normal epidermis in the CMA-DHM and CMA-DHM@HDFs groups, while significant disparities were observed in the Control and CMA groups (Fig. 6h). This further substantiates the efficacy of CMA-DHM and CMA-DHM@HDFs in achieving normal epidermis. Additionally, the marked presence of skin appendages in the wound areas underscored the robust tissue remodeling enhanced by CMA-DHM@HDFs (Fig. 6g). These satisfactory wound-healing outcomes underscore the success of our multifunctional combined therapy using functional cell-laden skin substitutes as efficient wound dressings for accelerating diabetic wound healing.

3.7. Regenerative diabetic wound healing via efficient ROS management, M1-to-M2 transition, Re-epithelialization, collagen deposition, blood vessel formation, and tissue remodeling

To elucidate the regenerative processes facilitated by biomimetic skin constructs in diabetic wound healing, we systematically investigated wounded tissues across three main phases of wound healing. During the inflammatory phase, ROS levels and inflammatory cell phenotypes were assessed via DCFH-DA fluorescence imaging and immunofluorescence staining. On the infection-prone day 4, the control group displayed intense ROS fluorescence, indicating elevated ROS levels. The CMA group showed similar fluorescence levels to the control, which were significantly higher than those of the other treatment groups. In contrast, the CMA-DHM and CMA-DHM@HDFs groups exhibited significantly reduced DCFH fluorescence (Fig. 7a and b). By day 7, ROS levels in the CMA-DHM and CMA-DHM@HDFs groups diminished to nearly negligible levels, signifying a transition to the resolution of inflammation and the proliferative phase (Fig. 7c and d), while the control and CMA groups continued to exhibit strong fluorescence. Immunofluorescence staining corroborated these healing trends. On day 7, the number of M1 macrophages (iNOS⁺) was significantly lower in the CMA-DHM and CMA-DHM@HDFs groups compared to other groups, with the CMA-DHM@HDFs group showing the fewest iNOS-positive cells (Fig. 7e and f). Simultaneously, there was a significant increase in the infiltration and distribution of M2 macrophages (CD163⁺), especially in the CMA-DHM@HDFs group (Fig. 7g and h), indicating that the implantation of HDFs further enhanced the anti-inflammatory effect. These results demonstrate the excellent anti-inflammatory and antioxidant activity of CMA-DHM@HDFs in vivo, efficiently scavenging reactive oxygen species in diabetic wounds, alleviating high oxidative stress during the inflammatory phase, and facilitating the transition of M1 macrophages to the M2 phenotype, thereby advancing to the anti-inflammatory and proliferative phases. Histological analysis with H&E and Masson staining on day 7 further confirmed that the treatment groups entered the proliferative phase, as evidenced by effective re-epithelialization, granulation tissue development, and increasing trends in collagen deposition across different groups (Fig. 7i and j). The CMA-DHM@HDFs group displayed the highest collagen deposition fraction at 60.6 % after 7 days of treatment, significantly higher than the other groups (Fig. 7k), signifying its superior bioactivity in enhancing ECM construction.

Fig. 7.

Evaluation of the healing progress across three wound healing phases. a) DCFH-DA staining of the ROS level of wounds on day 4. b) Quantification analysis of ROS level on day 4 (n = 6). Scale bar: 50 μm. c) DCFH-DA staining of the ROS level of wounds on day 7. d) Quantification analysis of ROS level on day 7 (n = 6). Scale bar: 50 μm. e) Immunofluorescence staining of M1 macrophage marker (iNOS) on day 7. Scale bar: 100 μm. f) Quantification analysis of iNOS positive area (n = 4). g) Immunofluorescence staining of M2 macrophage marker (CD163) on day 7. Scale bar: 100 μm. h) Quantification analysis of CD163 positive area (n = 4). i) H&E staining on day 7. The newly formed epidermis was noted by green curves and green “E”. Granulation tissue was noted by green “GT”. Scale bar: 100 μm. j) Masson's trichrome staining on day 7. Newly synthesized collagen was indicated by green arrows. Scale bar: 200 μm. k) Collagen deposition on day 7 (n = 6). l) CD31 staining on day 18, Scale bar: 100 μm. s) Quantification of blood vessel (BV) diameters (n = 4). t) α-SMA staining on day 18, Scale bar: 100 μm. o) Quantification of BV density (n = 6). p) Masson's trichrome staining on day 18. Newly synthesized collagen was indicated by green arrows. Scale bar: 200 μm. q) Collagen deposition on day 18 (n = 6). ∗p < 0.05, ∗∗p < 0.01; ∗∗∗p < 0.001. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

By day 18, fluorescence staining of CD31⁺ cells in mature vessels and α-SMA⁺ cells in newly formed vessels was more pronounced in the CMA-DHM and CMA-DHM@HDFs treatment groups than in the CMA and control groups, indicating more effective stimulation of angiogenesis during the proliferative and remodeling phases (Fig. 7l–t). The CMA-DHM@HDFs group exhibited larger blood vessel diameters (Fig. 7s) and higher blood vessel density (Fig. 7o), likely due to the direct participation of HDFs or their secretion of angiogenic factors that further stimulate angiogenesis. Masson trichrome staining further demonstrated the efficacy of CMA-DHM@HDFs in promoting tissue regeneration during the remodeling phase (Fig. 7p), as evidenced by the regeneration of skin appendages and the presence of robust, well-organized, mature collagen fibers, with a collagen deposition rate reaching 69.8 % (Fig. 7q). The overall healed structure closely resembled normal skin, indicating successful regenerative diabetic wound repair. In contrast, due to infection, the wounds in the control group failed to re-epithelialize and exhibited delayed granulation tissue development, resulting in a prolonged inflammatory stage and progression to chronic wounds. Overall, the multifunctional bioengineered skin (CMA-DHM@HDFs) effectively guided diabetic wounds through the inflammatory phase, promoting angiogenesis and extracellular matrix remodeling during the proliferative and remodeling phases, achieving efficient and high-quality tissue regeneration.

4. Conclusion

In summary, we have developed a streamlined yet highly effective CMA-DHM bioink for DLP 3D bioprinting, enabling the creation of biomimetic skin constructs (CMA-DHM@HDFs) as an advanced, multifunctional therapeutic solution for diabetic wound care. Composed of methacrylated collagen (CMA) and dihydromyricetin (DHM), this bioink overcomes key challenges in DLP cell-laden printing by improving printability, mechanical integrity, and bioactivity at low concentrations. Its potent antioxidant properties create a ROS-free microenvironment that promotes cell proliferation while mitigating oxidative stress-induced damage. Furthermore, the CMA-DHM hydrogels induce macrophage polarization from the pro-inflammatory M1 phenotype to the pro-regenerative M2 phenotype, effectively modulating inflammation in diabetic wounds. The hydrogels also accelerate re-epithelialization and collagen deposition during the proliferative phase. The encapsulation of human dermal fibroblasts (HDFs) further enhances angiogenesis and tissue regeneration. This work presents a scalable, versatile DLP bioprinting platform that offers a transformative therapeutic strategy for chronic wound management. Its simplicity and efficacy position it for promising clinical translation, marking a significant advancement in tissue engineering and regenerative medicine.

CRediT authorship contribution statement

Caihong Fu: Writing – review & editing, Writing – original draft, Visualization, Validation, Software, Methodology, Investigation, Formal analysis, Data curation. Guangyu Liu: Software, Methodology, Investigation, Formal analysis, Data curation. Yirui Fan: Methodology, Investigation, Data curation. Lang Xiao: Methodology, Data curation, Conceptualization. Wenhua Li: Methodology. Xinyu Tian: Methodology. Jianxi Xiao: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition.

Declaration of competing interest

We declare that no known competing financial interests or personal relationships could have appeared to influence the work reported in our paper “Highly Printable and Multifunctional Cell-Laden Collagen-based Bioinks for Precise DLP Bioprinting and Rapid Diabetic Wound Regeneration”

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Data availability

Data will be made available on request.