Radiation cross-linked collagen scaffolds facilitate root coverage and keratinized gingival regeneration

Abstract

The study aimed to develop a radiation cross-linked collagen scaffold (RCS) and assess its potential for root coverage and keratinized gingival regeneration, addressing the prevalent issue of gingival recession and limitations of traditional treatments. RCS was prepared through electron beam irradiation and cross-linking followed by freeze-drying. Its properties were evaluated, including Fourier transform infrared analysis, swelling behavior, microscopic observation, porosity measurement, compression modulus and structural stability. In a rat gingival recession model with 96 rats divided into four groups, the root coverage index and gingival health indices were measured, and histological analyses were conducted. The cross-linked network structure of RCS provided excellent mechanical properties and stability. In the rat model, RCS effectively promoted gingival regeneration, with the RCS group achieving a root coverage index of 87.7 ± 2.7 %, which was 54.13 %, 42.83 % and 8.41 % higher than that of the sham operation group, non-crosslinked group and chemical crosslinked group respectively. Histological analysis showed that RCS promoted anti-inflammatory macrophage polarization, enhanced collagen deposition and gingival lamina propria fiber density and increased angiogenesis. Additionally, RCS exhibited good biosafety, as blood indices and organ coefficients remained normal. In conclusion, RCS effectively promotes gingival regeneration and is a promising keratinized gingiva substitute for gingival recession, offering a new option for oral tissue repair.

Graphical abstract

Root coverage and gingival regeneration enabled by radiation cross-linked collagen scaffolds.

Highlight

• •RCS shows great potential for keratinized gingiva in promoting root coverage and gingival regeneration.

• •RCS improves the root coverage index in gingival regeneration.

• •RCS, with its mechanical stability, eliminates soft tissue volume deficiencies in gingival tissue.

1. Introduction

Gingival recession (GR) is characterized by the apical migration of the gingival margin relative to the cementoenamel junction, leading to exposure of the root surface to the oral cavity [1,2]. It's associated with issues like root surface caries, tooth sensitivity and chewing function decline. Meta-analysis indicates a high prevalence, affecting 78 % of the global population [3,4].

Over the past 50 years, autogenous keratinized gingival grafting has been used for GR treatment but presents drawbacks, including donor site complications, limited tissue availability and about 30 % shrinkage during healing, impacting treatment success [[5], [6], [7]]. Therefore, novel graft substitutes utilizing natural or synthetic polymers have been developed to improve root coverage and gingival regeneration [8,9]. These substitutes aim to reduce periodontal surgery morbidity and aid tissue repair, offering a predictable approach for gingival regeneration and root coverage [10,11]. Natural polymer scaffolds can promote regeneration but often suffer from low strength, unstable structure, immunogenicity and inflammation [12]. Synthetic polymer scaffolds offer controllable properties, but their degradation products may cause rejection and suboptimal repair effects [13].

However, existing graft materials face challenges in application. Graft materials may be exposed or contaminated, and the removal of non-degradable materials or dealing with the degradation products with biotoxicity can damage newly formed tissue [14]. When using polytetrafluoroethylene or bioabsorbable membranes for treatment, the root coverage range from 54 % to 87 %, with an average of 74 % [4]. Additionally, poor color matching, misalignment of the mucogingival junction, and keloid-like texture often lead to aesthetically unacceptable outcomes, rendering such treatments unsuccessful [15]. Therefore, there is an urgent need for new substitute materials.

Previous work reported irradiation parameters and preparation of collagen hydrogel for tissue regeneration [[16], [17], [18], [19]]. Herein, radiation cross-linked collagen scaffold (RCS) with excellent compliance and long-term stability was designed and constructed through programmed freeze-drying following electron beam (EB) irradiation. RCS exhibited enhanced mechanical resistance, structure adaptability, stability and superior biocompatibility. Using a rat model, biomechanical tests, histological evaluations and immunohistochemical analyses were conducted to explore RCS's potential in root coverage, particularly for keratinized gingival defect repair. These findings demonstrated that RCS promotes gingival regeneration while improving stability and aesthetics, providing a new option for GR treatment. RCS also holds promise for addressing gingival recession and serving as a dental implant material, advancing stomatology and bioactive materials.

2. Results

2.1. Preparation and characterization

2.1.1. Preparation

RCS was fabricated through irradiation-induced crosslinking and lyophilization. Crosslinking of collagen was induced by EB irradiation. The programmed freeze-drying process precisely controlled the freezing and drying stages (Fig. 1a). During the freezing stage, the temperature was maintained at approximately −4 °C with a stable temperature gradient, enabling water molecules to crystallize orderly at an appropriate rate, forming fine and uniformly distributed ice crystals. This laid the foundation for a favorable three-dimensional structure during the subsequent drying process. By programming parameters such as the vacuum degree and temperature in the drying stage, efficient sublimation of ice crystals into water vapor was ensured, minimizing damage to the material structure. Eventually, RCS with three-dimensional sponge-like structure was obtained [20].

Fig. 1.

Preparation and characterization of collagen scaffolds for root coverage. (a) Freeze-drying; (b) Fourier transform infrared spectroscopy; (c) liquid absorption capacity; (d) Pore size, surface and pore microstructure; (e) Swelling and stability; (f) Compressive properties.

2.1.2. Fourier transform infrared spectroscopy

Fourier Transform Infrared Spectroscopy (FTIR) was used to investigate the chemical structure of collagen material and RCS (Fig. 1b). Typical characteristic peaks of collagen included the N-H stretching vibration of amide A at around 3310 cm⁻¹, the C-H stretching vibration of amide B at approximately 3080 cm⁻¹, the C=O stretching vibration of amide I in the range of 1600–1700 cm⁻¹, amide II (1500 - 1550 cm⁻¹) generated by the C-N stretching and N-H bending vibrations, as well as the amide III band (1200 - 1300 cm⁻¹) arising from the C-N stretching, N-H bending and C-C stretching vibrations. After irradiation-induced crosslinking and freeze-drying, the shifts in these major amide bands were negligible, indicating that collagen retained conformation.

2.1.3. Liquid absorption and swelling properties

The hydrophilicity of RCS was evaluated. Collagen molecule, with abundant hydroxyl, carboxyl, amino and amide groups, enhanced the hydrophilicity of RCS through hydrogen bonding with water. RCS absorbed 16–39 times its dry weight of solutions such as PBS and water within 1 h (Fig. 1c, Fig. S1), and reached an equilibrium swelling state after 3 h (Table S1).

2.1.4. Morphology

Focused Ion Beam Scanning Electron Microscopy revealed the pore structure and surface characteristics of RCS. At 250 × magnification, a biomimetic thin-layer was observed on the surface, facilitating adhesion to the gingival injury surface, with fence-like interconnected pores crucial for cell migration and vessel growth (Table S2). RCS exhibited an average pore size of approximately 114.00 ± 49.52 μm. At high magnification, pore surfaces displayed fine convex or concave patterns conducive to cell adhesion (Fig. 1d). Micro-CT analysis indicated a porosity of 95.7 ± 2.5 %. After 7-day co-incubation with various liquids followed by re-freeze-drying, RCS maintained structural integrity in pure water. In PBS and DMEM, crystal particles formed and covered the surface, indicating liquid entry and storage. When immersed in rabbit anticoagulant venous blood, its surface exhibited a honeycomb-like structure due to blood cell agglutination (Fig. 1e).

2.1.5. Mechanical properties

RCS demonstrated excellent resilience, fully recovering its original shape after compression and release (Fig. 1f). The mechanical properties of RCS were assessed in various states, namely after freeze-drying, rehydration, secondary freeze-drying and rehydration swelling, with stress-strain curves exhibiting consistent trends. The mechanical compression moduli for these states were 43.70 ± 3.87 kPa, 39.98 ± 17.46 kPa and 25.51 ± 1.27 kPa, respectively (Fig. S2). Compression to 75 % of the original height was selected based on previous studies in biomaterial mechanical property testing [16,21]. This compression level effectively simulate the physiological pressure environment that a wound dressing might encounter, such as pressure from surrounding tissues during woundhealing providing an accurate and comparable assessment of RCS's mechanical performance.

2.1.6. Hemolysis and coagulation tests

The hemolysis rate of RCS was 0.61 ± 0.06 %, meeting the requirements of GB/T 16886.4 (ISO 10993-4). The blood clotting index (BCI) of RCS was 38.12 ± 1.91 %, lower than that of the collagen raw material (47.67 ± 4.79 %).

2.2. General observation and HE staining

2.2.1. General observation

Four groups were established: the RCS group (experimental group), the chemical cross-linked collagen scaffold group (CCS, positive control group), the non-crosslinked collagen scaffold group (CS, negative control group) and the sham operation group (SOG). Post-operative early appearance and healing varied among groups (Fig. 2a). On day 28, the RCS group exhibited newly formed tissues that integrated well with surrounding gingiva, resembling normal gingiva in morphology, color and texture. The CCS group showed similarly favorable healing conditions. The CS group displayed a slower healing process with thinner newly formed gingiva and poor attachment. The SOG group's healing remained insufficient with unresolved recession.

Fig. 2.

Appearance and tissue healing assessment of root coverage and keratinized gingival regeneration. (a) Representative gross appearance of the gingiva in SD Rats; (b) Representative results of Hematoxylin and Eosin staining on day 28; (c) Gingival index; (d) Root coverage index (mean ± standard deviation, ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001; ∗∗∗∗P < 0.0001).

2.2.2. Tissue regeneration features by Hematoxylin and Eosin staining

H&E staining revealed effective tissue healing in the RCS and CCS groups, with no material residue or significant inflammatory cell infiltration on day 28. In the RCS group, on the buccal side, the dento-gingival unit spontaneously regenerated, showing masticatory mucosa characteristics and increased thickness. The connective tissue below the keratinized epithelium exhibited more complete fiber structures (Fig. 2b). Further analysis of RCS at earlier time points demonstrated progressive degradation and integration. On day 3, H&E staining showed visible scaffold residue with moderate inflammatory cell infiltration, indicating an early healing response. On day 7, scaffold residue was significantly reduced, with increased tissue infiltration, fewer inflammatory cells and early collagen fiber formation. On day 14, no scaffold residue remained and the tissue exhibited well-organized collagen fibers, suggesting advanced integration (Fig. S3). After autogenous gingival tissue was excised and re-implanted, H&E staining on day 28 showed well-regenerated tissue with dense collagen fibers, similar to the RCS group (Fig. S4d). These findings confirm that RCS degrades rapidly while facilitating tissue regeneration without chronic inflammation.

2.2.3. Gingival index

The gingival index (GI) changes were assessed (Fig. 2c). On day 3, the SOG group exhibited a relatively high GI, indicating inflammation, while the RCS group showed the lowest GI. On day 7, the GI of RCS group further decreased. On day 14 and 21, the RCS and CCS groups exhibited no bleeding or inflammation, unlike the SOG and CS groups. The autogenous graft (AG) using harvested gingiva with donor-site suturing, also showed a low GI, comparable to the RCS and CCS groups (Fig. S4b).

2.2.4. Attachment loss incidence

On day 3, the SOG group exhibited a high frequency of attachment loss (AL), while the RCS group performed better. On day 7, AL in the RCS group dropped to 0. From day 14, all groups except the SOG group performed well in AL (Table S3).

2.2.5. Analysis of root coverage index

The root coverage index (RCI) showed significant differences existed among treatment groups (Fig. 2d–Table S4). The RCS group outperformed the CS and SOG groups at multiple time points. On day 3, the RCS group achieved an RCI of 34.2 ± 11.3 %, higher than the SOG (15.7 ± 5.4 %) and CS (23.8 ± 8.4 %) groups. On day 28, the RCS group (87.7 ± 2.7 %) outperformed the other groups, being 54.13 % higher than the SOG (56.9 ± 2.4 %) group, 8.41 % higher than the CCS (80.9 ± 6.3 %) group, and 42.83 % higher than the CS (61.4 ± 3.7 %) group. To further evaluate RCS against the clinical gold standard, the AG, in which harvested gingiva with donor-site suturing was performed, was assessed on day 21 (82.6 ± 3.0 %) and 28 (93.1 ± 7.2 %), with results comparable to RCS (Fig. S4a and S4d).

2.3. Immunohistochemistry and macrophage polarization

Macrophages play a crucial role in tissue repair and regeneration [22]. Immunohistochemical (IHC) staining showed that in the early healing stage, the expression of M1 macrophage marker (iNOS) in SOG group was higher than in the RCS and CCS groups (Fig. 3a). Quantitative analysis revealed that the iNOS-positive area in the SOG group was 2.7–4.6 times that of the RCS group (Fig. 3c). For the M2 marker (CD163), expression in the RCS and CCS groups was 1.8–4.2 times higher than in the SOG group (Fig. 3b and d). Cytokine detection found that the serum levels of C-reactive protein and TNF-α in the RCS group were significantly lower than those in the SOG group, while the anti-inflammatory factor IL-10 increased (Fig. S5).

Fig. 3.

Inflammatory response and macrophage analysis. (a) Expression of iNOS in gingival tissue. (b) Expression of CD163 in gingival tissue. (c) Quantitative analysis of the positive expression area of iNOS. (d) Quantitative analysis of the positive expression area of CD163 (bar = 20 μm; mean ± standard deviation, ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001; ∗∗∗∗P < 0.0001).

2.4. Gingival collagen fibers and angiogenesis

2.4.1. Collagen fiber characteristics

Masson staining revealed that the RCS, CCS and CS groups had greater collagen fiber density than the SOG group. The RCS group exhibited more orderly arranged gingival fibers compared to the CS group (Fig. 4a). Sirius red staining showed more newly formed collagen fiber bundles in the RCS group were orderly arranged compared to others (Fig. 4b). IHC staining of type I collagen fiber confirmed it preponderance (Fig. 4c). The collagen volume fraction (CVF) data demonstrated the RCS group had higher collagen fiber density at various time points compared to other groups (Fig. 4d). On day 3, the CVF of RCS was 67.88 ± 2.26 %, higher than the SOG group (55.91 ± 7.03 %, P < 0.05), and also superior to the CCS and CS groups. On day 7, with a CVF of 72.88 ± 6.31 %, its leading position persisted. On day 14, the CVF of RCS was 75.90 ± 3.05 %, showing stable superiority. Even on day 28, the CVF of RCS was 80.29 ± 5.20 %, higher than that of other groups.

Fig. 4.

Distribution of collagen fibers and vascularization in gingival tissue and their characteristics in different groups.(a) Representative Masson staining showing the distribution and arrangement of collagen fiber. (b) Sirius red staining depicting the newly formed collagen fiber bundles and their organization. (c) Immunohistochemistry of collagen- I, confirming the preponderance of collagen fibers. (d) Collagen volume fraction of gingival tissue, demonstrating the variation in fiber density over time and across different groups. (e) Immunohistochemistry of CD31 for vascularization characterization, revealing the differences in angiogenesis and the relationship between collagen fibers and blood vessels. (f) Vascularization characteristics, showing distinct blood vessel distribution patterns (bar = 50 μm, mean ± standard deviation, ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001; ∗∗∗∗P < 0.0001).

2.4.2. Effect of RCS for promoting angiogenesis

IHC staining of CD31 for blood vessels on days 14 and 28 showed that the RCS group had a greater number of both large and small blood vessels compared to other groups. The SOG group exhibited sparsely distributed collagen fibers and fewer blood vessels. The CCS group displayed a certain complexity in the collagen fiber structure with interspersed blood vessels. The CS group had loosely arranged fibers and relatively fewer small blood vessels (Fig. 4e and f).

2.5. Assessment of postoperative health status

2.5.1. Blood biochemical indexes

Hematological analysis indicated good biocompatibility and non-toxic in all groups. Serum biochemical indicators, including alanine aminotransferase (ALT), aspartate aminotransferase (AST), creatinine (CRE) and blood urea nitrogen (BUN), showed no significant differences among the groups (Fig. 5a).

Fig. 5.

Blood routine examination, blood biochemistry, organ coefficients and H&E staining results on day 28. (a) Biochemical parameters such as ALT, AST, BUN and CRE levels among all groups showed no significant change (n = 6). (b) Blood routine parameters such as RBC, WBC, HGB and PLT levels among all groups showed no significant change (n = 6). (c) Organ coefficients parameters including heart, liver, spleen and kidney levels among all the groups showed no significant change (n = 6). (d) The histochemical analysis of the heart, liver, spleen, and kidney revealed no apparent differences compared to the SHAM control(bar = 50 μm; mean ± standard deviation).

2.5.2. Blood routine indexes

Analysis of blood routine parameters, such as white blood cell count (WBC), hemoglobin (HGB), red blood cell count (RBC) and platelet count (PLT), revealed no significant differences among the four groups. These parameters remained within the normal range, indicating stable hematological status across all groups (Fig. 5b).

2.5.3. Major organs indexes analysis

The organ coefficient levels of the heart, liver, spleen and kidney showed no significant differences among the groups and remained within the normal range (Fig. 5c). Additionally, histological examination using H&E staining of the sections from these major organs revealed no significant changes in tissue microstructure, suggesting that the experimental treatments had no adverse effects on the structure and function of these vital organs (Fig. 5d).

3. Discussion

This study comprehensively characterized the RCS and evaluated its performance in root coverage and keratinized gingiva regeneration. Type I collagen, the most abundant collagen in the gingival tissue and the primary target for this application, features a relatively thick and strong fibril structure, thus offering superior mechanical support compared to type II collagen. The preparation of RCS involves irradiation-induced crosslinking and freeze-drying techniques. During the crosslinking process, EB irradiation acts on collagen dispersion, triggering a series of reactions where reactive species, like hydroxyl radicals (·OH), are generated from the radiolysis of water. These hydroxyl radicals induce dehydrogenation of collagen polypeptide chains, forming radicals that recombine to create new covalent bonds for crosslinking (Fig. S6) [[16], [17], [18], [19],23]. This precise crosslinking mechanism endows RCS with several advantages over traditional preparation methods. Irradiation-induced crosslinking achieves a more uniform crosslinking distribution compared to chemical methods, as high-energy rays generate uniformly distributed free radicals in the polymer system, rapidly initiating crosslinking reactions throughout the material [24]. In contrast, chemical crosslinking relies on the diffusion of cross-linkers, which is often affected by steric hindrance and local concentration differences, leading to uneven crosslinking. Additionally, the strong penetration ability of high-energy rays ensures consistent crosslinking both on the surface and within the material, whereas chemical cross-linkers may have limited diffusion, resulting in inconsistent crosslinking between the surface and the interior [25]. Studies on hydrogels indicated chemical crosslinking led to uneven crosslinking, causing variable swelling/shrinking properties and limited gel functionality compared to irradiation-crosslinked hydrogels, which may affect the mechanical properties and biocompatibility of the final product [26,27]. In contrast, the irradiation-induced crosslinking in RCS preparation ensures a stable and consistent structure, as evidenced by the FTIR analysis. The FTIR results indicated that the irradiation-induced crosslinking and freeze-drying processes did not significantly alter the original conformation of collagen, which is essential for RCS to mimic the normal oral gingiva and utilize the inherent properties of collagen to promote cell migration and vascularization [28].

The hydrophilicity and swelling properties of RCS suggest its potential to maintain tissue moisture and absorb blood during initial implantation in the defective oral mucosa, which is beneficial for early healing. This high hydrophilicity also contributes to its hemostasis function. Absorbed blood can form a natural clot-like structure, which, combined with the mechanical properties of RCS, plays an important role in preventing bleeding [29]. The mechanical properties of RCS enable it to withstand pressure, which is crucial for stabilizing early wounds. Consistent trends in stress-strain curves across different states of RCS, along with the specific values of its mechanical compression moduli, indicate stable mechanical behavior. Compression to 75 % of the original height, aligned with established practices in previous research, reflects RCS's mechanical adaptability [16,18,21]. Moreover, the favorable biocompatibility of RCS is evidenced by its low hemolysis rate and enhanced procoagulant effect following irradiation-induced crosslinking modification. These characteristics highlight RCS's suitability as a wound dressing material and strongly indicate its potential for clinical application. The selection of the 75 % compression level also ensures accurate evaluation of its mechanical properties, which is essential for predicting performance in wound care applications [30,31].

Although RCS and CCS exhibited no statistically significant differences in most parameters, such as macrophage polarization and angiogenesis markers, several advantages justify the preference for RCS in gingival regeneration. RCS achieved a root coverage index of 87.7 ± 2.7 %, slightly higher than CCS, which may reflect its enhanced biocompatibility, facilitating better tissue integration and a stable tissue-material interface. The EB irradiation-induced crosslinking process, characterized by high efficiency and elimination of elution steps compared to chemical crosslinking, streamlines manufacturing and reduces production costs [32]. Unlike CCS, which often relies on chemical agents like glutaraldehyde that carry risks of residual toxicity, RCS's irradiation-based approach minimizes potential adverse reactions, as evidenced by its low hemolysis rate and absence of genotoxicity. Future research on long-term safety and cost-effectiveness could further confirm RCS's superiority over CCS, supporting its potential as an ideal scaffold for root coverage and keratinized gingiva regeneration.

Based on the results of this study and recent research findings, mechanisms driving the observed effects of RCS are hypothesized [33]. The RCS group exhibited enhanced M2 macrophage polarization, marked by increased CD163 expression and elevated IL-10 levels, alongside reduced TNF-α and C-reactive protein, suggesting an anti-inflammatory microenvironment conducive to gingival regeneration [34,35]. Although signaling cascades were not directly elucidated, RCS may activate the PI3K-Akt-mTOR pathway, which is critical for M2 polarization by suppressing pro-inflammatory signals and promoting tissue repair [36]. Recent evidence indicates that biomaterial scaffolds can trigger this pathway via integrin-mediated mechanotransduction, consistent with RCS's porous structure facilitating cell adhesion [37,38]. Additionally, the JAK/STAT6 pathway may contribute, as it is known to drive IL-10 expression and M2 differentiation in response to biomaterial cues [39]. For angiogenesis, the RCS group's increased CD31 expression indicates enhanced vascularization, likely mediated by VEGF-A signaling [[40], [41], [42]]. Hypoxia-inducible factor-1α (HIF-1α) may further amplify this response by promoting endothelial cell migration within RCS's interconnected pores [43]. Integrin signaling, particularly through αvβ3, could also facilitate angiogenesis by enhancing endothelial cell interactions with the scaffold's extracellular matrix [44,45]. These hypothesized mechanisms are consistent with RCS's ability to promote root coverage (87.7 ± 2.7 %) and collagen deposition, underscoring its potential as a gingival regeneration scaffold.

RCS exhibits a pore size of 114.00 ± 49.52 μm and porosity of 95.7 ± 2.5 %, achieved through controlled EB irradiation and programmed freeze-drying, which forms fine, interconnected ice crystals for a biomimetic 3D structure [16,21]. This pore size aligns with optimal ranges for gingival fibroblast infiltration (<160 μm) and ensures cells remain within 200 μm of blood supply for angiogenesis, as evidenced by increased CD31 expression and orderly collagen deposition [46,47]. Prior work demonstrates that adjusting the absorbed dose (from 5 kGy to 25 kGy) and cooling rates tailors pore size, with lower doses and moderate cooling yielding pores of 100–150 μm, ideal for vascularization and fibroblast migration [16,21]. The interconnected, fence-like pores observed via SEM further support endothelial cell migration and microchannel formation. Reducing pore size below 50 μm may impede fibroblast migration, while larger pores (>250 μm) compromise mechanical stability. These design choices underpin RCS's efficacy, achieving 87.7 ± 2.7 % root coverage and robust tissue integration.

Blood-related and organ-related indices analyses provided evidence of RCS's biocompatibility. No significant disparities were observed in blood biochemical indexes, blood routine parameters or organ coefficients among the groups. Additionally, histological examination of major organs revealed no substantial changes in tissue microstructure, indicating that RCS has minimal impact on the overall physiological status of the experimental subjects. In contrast, glutaraldehyde cross-linkers have potential genotoxicity and can cause adverse reactions, which highlights the biocompatibility advantage of RCS [48]. During EB irradiation, transient hydroxyl radicals facilitate collagen crosslinking, with most recombining to form stable covalent bonds. Post-processing steps, including freeze-drying and sterilization, further minimize residual radicals, reducing the likelihood of adverse biological effects. Results showing low hemolysis rates, no genotoxicity and enhanced M2 macrophage polarization (increased CD163, elevated IL-10), suggest that any residual byproducts do not impair biocompatibility or tissue integration. Recent studies confirm that irradiation-crosslinked biomaterials exhibit negligible cytotoxicity and promote tissue healing [49]. These findings support RCS's safety profile, reinforcing its potential for gingival regeneration.

RCS exhibited progressive degradation, with visible scaffold residue on day 3, significant reduction on day 7, and complete degradation by day 14. This rapid degradation aligns with RCS's high porosity and the metabolically active gingival environment. Tissue integration progressed concurrently, with early collagen fiber formation on day 7, well-organized fibers on day 14, and complete regeneration of the dento-gingival unit by day 28, displaying masticatory mucosa characteristics and increased thickness. Inflammatory cell infiltration decreased over time, with minimal presence by day 14, indicating no chronic inflammation, further supported by an anti-inflammatory microenvironment (increased CD163, elevated IL-10, reduced TNF-α). The rapid degradation and integration of RCS minimize the risk of prolonged foreign body responses, enhancing its suitability for gingival regeneration.

A comparison between RCS and AG, the gold standard for GR treatment, is included to enhance clinical relevance. RCS achieved the RCI of 87.7 ± 2.7 % on day 28, slightly below the AG group's RCI of 93.1 ± 7.2 %, but both showed greater RCI values than CCS (80.9 ± 6.3 %), CS (61.4 ± 3.7 %) and SOG (56.9 ± 2.4 %).These results are comparable to reported outcomes for autografts, where mean root coverage ranges from 80 % to 90 % using techniques like coronally advanced flaps with connective tissue grafts [30,50]. Both RCS and AG exhibited low GI on day 21, indicating minimal inflammation. Histologically, autogenous graft specimens on 28 days post-operation revealed well-regenerated tissue with dense collagen fibers and minimal inflammation, like those of RCS. Unlike autografts, which require donor site surgery (e.g., palate), RCS eliminates donor site morbidity, reducing postoperative pain and complications such as bleeding or infection. Cost-wise, RCS may offer advantages by eliminating surgical harvesting, and its EB irradiation-induced crosslinking process, characterized by high efficiency and the elimination of elution steps compared to chemical crosslinking, streamlines manufacturing and reduces production costs. Although scalability for large-scale production requires further evaluation, these process characteristics suggest promising cost-effectiveness for clinical translation. Future clinical studies directly comparing RCS with autografts in terms of efficacy, patient-reported outcomes, cost-effectiveness and long-term population benefit evaluation are essential to validate its role as a viable alternative.

4. Conclusion

RCS was comprehensively investigated as a promising substitute for gingival regeneration, evaluated through physical, chemical, and preclinical performance in root coverage procedures, including aesthetic outcomes. The newly formed tissues exhibited a smooth surface, integrating well with surrounding gingiva, thereby enhanced aesthetics. Compared to other groups, the RCS group demonstrated lower GI, reduced AL incidence and a higher RCI, confirming its efficacy in promoting root coverage and tissue regeneration. RCS promotes macrophage polarization toward the M2 phenotype, fostering an anti-inflammatory environment and supporting tissue regeneration, while generating high-density collagen fibers beneficial for tissue repair and effectively promoting blood vessel growth. Analysis of blood and organ-related indexes verified that RCS exhibits no sub-chronic toxicity to liver and kidney functions. Future research is needed to further explore its long-term effects, underlying mechanisms, and optimize the application of RCS for better clinical outcomes in periodontal tissue repair.

5. Materials and methods

5.1. Preparation

Reagents used in this study are listed in Table S5. Unless otherwise specified, all reagents were of analytical grade, and the collagen raw material was of medical grade. RCS was prepared as follows. Collagen type I and sterile water for injection were mixed at a ratio of 0.8 % (w/v). The mixture was then stirred using an ARE-310 hybrid mixer for 3 h to form collagen dispersion. The dispersion was filled into test molds and subjected to EB irradiation using the IS1020 system (0.2 MeV–10 MeV) at a dose rate of 15 Gy/min at room temperature to form a collagen hydrogel. Subsequently, the hydrogel was frozen for 4 h and freeze-dried at an atmospheric pressure of less than 50 Pa for 72 h to obtain RCS.

5.2. FTIR analysis

RCS and collagen raw material were prepared. FTIR spectroscopy was recorded using an FTIR spectrometer (Nicolet iS10, Thermo Scientific, USA). Scanning was performed with a resolution of 1 cm⁻¹ and a wavenumber range from 650 to 4000 cm⁻¹ to analyze and compare structural changes between RCS and collagen samples.

5.3. Swelling behavior

Liquid absorption performance in various aqueous media was tested. The initial weight of RCS (M₀) was weighed. Samples were soaked and incubated at 37 ± 1 °C, and the swelling weight (M_w) was weighed at different time points (n = 6/group). The swelling state was recorded with a camera. The calculation formula for liquid absorbency was as follows.

$$\mathrm{L}\text{iquid}\text{absorbency}(\%)=\frac{M_w-M_0}{M_0}\times 100$$ (1)

5.4. Microscopic morphology observation

RCS samples were visualized using a focused ion beam scanning electron microscope (ThermoFisher Helios 5 UC, Thermo Fisher Scientific, USA). The specific parameters during the scanning process were as follows. The specific parameters during scanning were as follows: high voltage at 5.00 kV, current at 0.10 nA, detector as TLD, mode as SE, and magnification at 250 × and 5000 × .

5.5. Porosity

RCS samples were scanned using a high-resolution micro-CT imaging system (SKYSCAN 1276, Bruker, Germany) operated at 50 kV and 200 μA without a filter for optimal scanning. Cross-sectional images of RCS samples (n = 6) were captured and transformed into interpretable data via NRecon v.1.6.3 software (Bruker-micro-CT). Porosity was determined by analyzing bone density of the transformed images using CTAn v.1.12 software (Bruker-micro-CT).

5.6. Compression modulus

The samples (n = 6/group) were compressed to 75 % of their original height using a texture analyzer (TA.XT PlusC, Stable Micro System, UK). Force-displacement data obtained during compression were used to calculate the compression modulus.

5.7. Structural stability

RCS samples after swelling in different aqueous media for 7 days were freeze-dried following the same procedure as during RCS preparation (n = 6/group). The re-lyophilized RCS samples were visualized using a focused ion beam scanning electron microscope (ThermoFisher Helios 5 UC, Thermo Fisher Scientific, USA).

5.8. Hemolysis experiment and coagulation index

Assessment methods were employed in accordance with the Chinese Pharmacopoeia 2020. In the hemolysis experiment, RCS samples were exposed to blood, and the degree of hemolysis was measured and evaluated. For the coagulation index determination, specific tests were conducted to analyze the effect of RCS on blood coagulation (n = 10).

5.9. Animal ethics and experimentation

96 Sprague-Dawley rats were used and housed in individual ventilated cages at 20–25 °C, 40 %–70 % humidity, with a 12-h light/dark cycle. The animal experimentation adhered to relevant regulations, following standards in GB/T 35892-2018, NIH guidelines. The ARRIVE guidelines were used for result description.

After one week of adaptive feeding, the preparation of the gingival recession model in SD rats was carried out. Anesthesia was induced, and the rats' four limbs were fixed on a rat board. A 4 × 4 tooth spreader separated the upper and lower jaws, exposing the rats’ oral cavities. The skin around the oral cavity was disinfected with 75 % alcohol, and the oral surface, including the oral mucosa, was disinfected with 0.12 % chlorhexidine and povidone-iodine solution. The buccal gingival tissue of the first premolar of the right mandible was resected using a disposable surgical blade. After placement of the corresponding graft materials, all group except SOG were intermittently sutured with absorbable sutures to ensure that the implanted or covered materials were closely attached to the surfaces of the teeth and alveolar bones. A periodontal dressing (Shaanxi Junxing Biomedical Technology Co., Ltd., China) was used for secondary fixation, consistent with clinical operation.

5.10. Determination of the sample size of experimental animals

This study adopted a randomized controlled experimental design, designating one experimental site per animal (Fig. S7). The experiment was divided into four groups, including positive control group (CCS), experimental group (RCS), negative control group (CS) and SOG, with 24 rats in each group. After creating gingival defects, four monitoring time points were set: days 3, 7, 14, and 28 post-operation. The root coverage rate was selected as the primary statistical indicator. Sample size was determined by referencing multiple control studies [3,27,51]. Adhering to the 3R principles (replacement, reduction, refinement) in animal experiments and considering risks such as animal mortality, the number of rats per group was set at 24, ensuring 6 rats per group at each monitoring time point, totaling 96 rats. One piece of sample material was used for each animal.

5.11. Measurement of root coverage index

The RCI was measured by observing and measuring the distance from the gingival margin to the cementoenamel junction before and after treatment, on days 0, 3, 7, 14, 21 and 28. Based on previous studies, it was calculated as the ratio of the reduction in gingival recession width after treatment to the initial width before treatment, multiplied by 100 % [2,4].

5.12. Measurement of gingival index

Based on the WHO guidelines for periodontal examination, the GI was evaluated through periodontal probe exploration and visual inspection. Gingiva's overall condition was observed for signs of inflammation, and scores from 0 to 3 were assigned according to its characteristics (0: healthy; 1: mild; 2: moderate; 3: severe). Data were analyzed to track gingival changes over time or due to treatment.

5.13. Measurement of attachment loss incidence

Attachment loss was measured using a periodontal probe. A distance greater than 1 mm between the gingival margin and the cementoenamel junction during examination indicated attachment loss.

5.14. Morphological observation by HE staining

Tissue samples were prepared through fixation, processing and sectioning, followed by H&E staining. The stained sections were observed and analyzed under a Zeiss Axioscan 7 microscope slide scanning system to understand tissue morphology and pathology.

5.15. Analysis of new type I collagen and by immunohistochemistry

For IHC analysis of newly formed type I collagen in decalcified tissues, paraffin sections were dewaxed and rehydrated, followed by antigen retrieval and PBS washing. After blocking endogenous peroxidase, diluted primary antibody (Wuhan Servicebio Technology Co., Ltd.) was added, and sections were incubated overnight at 4 °C. Slides were washed, HRP-labeled secondary antibody was added, and sections were incubated at room temperature. DAB color development was carried out, followed by rinsing with tap water. Stained sections were examined under a white light microscope.

5.16. Analysis of collagen deposition by masson staining and sirius red staining

For collagen deposition analysis via Masson and Sirius red staining, paraffin sections of decalcified tissues were dewaxed and hydrated. Masson and Sirius red staining were performed following reagent instructions and previous experience. Stained sections were examined under bright-field microscopes for visual analysis. For CFV% quantification in Masson-stained slides, images were captured and imported into ImageJ. The color threshold was adjusted to select collagen fiber regions, and the software's functions were used to calculate the area within the ROI. CFV% was calculated by dividing this area by the total ROI area and multiplying by 100.

5.17. Analysis of new blood vessels by immunohistochemistry

For CD31 IHC analysis of new blood vessels, paraffin sections were dewaxed and antigen retrieved first, followed by blocking of endogenous peroxidase and serum. Sections were incubated overnight at 4 °C with Anti-CD31 Rabbit pAb (Servicebio, China, 1:500 dilution). Secondary antibody addition and DAB color development were performed, followed by nuclei staining. Stained sections were inspected under a microscope for analysis.

5.18. Analysis of inflammation regulation through macrophage polarization

For analysis of inflammation regulation via macrophage polarization, 5 μm decalcified paraffin sections underwent microwave-based antigen retrieval. Using an immunohistochemical kit, primary antibodies (anti-iNOS, 1:500 dilution; anti-CD163, 1:200 dilution, both from Servicebio, Wuhan, China) were incubated overnight at 4 °C. Secondary antibody incubation and DAB development followed. Sections were counterstained with hematoxylin. Tissue structure and target protein expressions were observed under the Zeiss microscope. ImageJ was used to analyze images and count positively expressed cells. Serum samples were collected on day 3. The levels of TNF-α, IL-10 and CRP were detected using enzyme-linked immunosorbent assay kits according to the manufacturer's instructions.

5.19. Blood routine examination, biochemical indicators and organ tissue analysis

Following GB/T 16886-11 (ISO 10993-11), blood samples from each group were collected over 28 days for complete blood count and serum biochemical indicator detection to assess the RCS's toxicological implications. Using a hemocytometer and Mindray's automated hematology analyzer (China), parameters like WBC, RBC, HGB, PLT, CRE, BUN, AST and ALT were analyzed. Instruments were calibrated with calibrators and control sera beforehand. After sacrificing SD rats, their hearts, livers, spleens and kidneys were removed for weighing and measurement. Organ coefficients were calculated, and HE staining was performed for tissue observation to check for toxic reactions.

5.20. Statistical analyses

Data are presented as the means ± standard deviations and were analyzed using SPSS. Statistical analysis was performed using One-way ANOVA and Tukey's post hoc test. P-value less than 0.05 was considered to indicate a statistically significant difference. The symbols ∗, ∗∗, ∗∗∗ and ∗∗∗∗ represent P < 0.05, P < 0.01, P < 0.001 and P < 0.0001, respectively.

CRediT authorship contribution statement

Hongwei Li: Writing – review & editing, Writing – original draft, Project administration, Formal analysis. Xin Chen: Validation, Formal analysis. Bozhao Wang: Validation, Formal analysis. Caiming Wu: Validation, Formal analysis. Jian Li: Writing – review & editing, Methodology, Formal analysis, Conceptualization. Ling Xu: Writing – review & editing, Supervision, Funding acquisition.

Ethics approval and consent to participate

This study was approved by the Xiamen University Laboratory Animal Center (study approval code: XMULAC20230243).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article.