Sequential hydrogen-release implant for promoting the soft tissue integration through immunomodulatory and pro-remodeling coupling

Abstract

The weaker soft tissue integration around implants compared to natural teeth poses a substantial challenge to the long-term success of implants. To enhance soft tissue integration, we develop an on-demand and long-lasting H₂-releasing implant to achieve precise sequential regulation of the soft tissue integration through immunomodulation and pro-remodeling coupling. In the inflammatory phase, the system on-demand releases H₂ responded to the local mild acidic microenvironment, which eliminates 73.6 % reactive oxygen species to induce M2 macrophage polarization, thereby establishing a pro-remodeling microenvironment. During the subsequent remodeling phase, the implant sustains release H₂ based on the hierarchical nanostructure, effectively promoting collagen fiber formation and angiogenesis. Surprisingly, we propose that H₂ can coordinately activate MAPK signaling in both gingival fibroblasts and vascular endothelial cells, coupled with stimulating pro-angiogenic paracrine of gingival fibroblasts. This implant achieves the on-demand transition of H₂ release kinetics that matches the temporal progression of soft tissue integration, implying great potential of enhancing soft tissue integration.

Graphical abstract

Highlights

• •The established soft tissue integration (STI) has less fiber and vasculature, making it prone to disruption and infection.
• •The STI process can be broadly categorized into two progressive phases: the inflammatory phase and the remodeling phase.
• •An innovative H₂-releasing implant was designed to sequentially regulate STI through immunomodulation and pro-remodeling.
• •H₂ can activate MAPK signaling for proliferation, coupled with stimulating pro-angiogenic paracrine of gingival fibroblasts.

1. Introduction

The demand for dental implants grows annually, with the global dental implants market projected to reach USD 8.17 billion by 2032, due to their proven effectiveness in replacing missing teeth [1]. The prevalence of peri-implant inflammation, including peri-implant mucositis and peri-implantitis, is reported to affect 63 % of patients, negatively impacting patients’ quality of life, and placing a considerable financial burden on global healthcare systems [2]. As the first barrier against pathogen invasion and infection, the soft tissue integration (STI) plays a crucial role in preventing inflammation and maintaining implant stability [3,4]. Nevertheless, compared to that around natural tooth, STI around the implant is weaker and more susceptible to pathogen invasion and damage from external stimuli [5,6]. Therefore, enhancing STI is crucial to reduce the incidence of inflammation and ensure long-term success, underscoring the urgent need for the development of effective strategies to enhance STI.

The STI around the implant is a multifaceted process, and comprises two key phases: inflammation phase and remodeling phase [4,7]. In the inflammatory phase, immune cells recruited to the implant site produce reactive oxygen species (ROS) and pro-inflammatory cytokines to eliminate cellular debris and protect tissues from pathogenic infections. Then, the remodeling phase is primarily characterized by soft tissue cell proliferation and migration, collagen deposition, and angiogenesis [8,9]. Although inflammation plays a critical role in the formation process, excessive oxidative stress-induced inflammatory imbalance can delay the transition from inflammation phase to pro-remodeling phase, impeding STI [10]. Recently, Chen et al. developed an IL-4/PDA (polydopamine)-coated titanium alloy implants to modulate macrophages differentiation for inducing pro-remodeling microenvironment, thereby improving STI [11]. Meanwhile, considering that the weak STI primarily results from the insufficient collagen fibers and blood vessels, promoting the formation of collagen fibers and blood vessels can enhance STI [[12], [13], [14]]. It was reported that a CCN2 delivery system was constructed on Ti implant surfaces to enhance STI by promoting the cell behavior of fibroblasts [13]. Notably, these previous studies focused on a single phase, employing limited strategies that exclusively target either immunomodulation or pro-remodeling, lacking the coordinate regulation of both phases [12,13].

As a bioactive gas molecule, molecular hydrogen (H₂) selectively reduces highly cytotoxic ROS such as the hydroxyl radical (•OH) and possesses inherent biocompatibility, making it a promising therapeutic agent for various inflammatory diseases [[15], [16], [17]]. Strikingly, emerging evidences suggested H₂ can accelerate mucosal and skin wound repair by promoting collagen deposition and angiogenesis, demonstrating favorable properties for promoting tissue remodeling [[18], [19], [20]]. Inspired by this, we infer that H₂ holds promise for coordinately modulating both phases of STI through immunomodulation and pro-remodeling coupling. The general clinic H₂ therapy includes inhalation of H₂, oral intake of hydrogen-rich water, and injection of hydrogen-rich saline [21]. However, due to the low solubility and high dispersibility of H₂, the frequent administration and short effect duration of these methods pose significant challenges to the application around implants [22,23]. To enhance STI, an exquisite design that enabling localized and sustained H₂ delivery at the implantation site are required.

In this study, we developed an on-demand and long-lasting H₂-releasing system on titanium surfaces for sequentially regulating the STI (Fig. 1). Remarkably, this system achieved the dynamic release rate transition that matches the need of STI by employing ammonia borane (AB) as a pH-dependent hydrogen donor [17,24]. The efficacy of this approach was validated using both cellular models and an implantation model in SD rats. In the inflammatory phase, the resultant implant (HP-MTi) on-demand released H₂ in response to the locally mild acidic microenvironment. As expected, H₂ eliminated ROS notably to induce macrophages polarization, accelerating the transition from the inflammatory to the pro-remodeling phase. Then in the remodeling phase, the hierarchical nanostructure of the implant limited the hydrolysis of AB, thereby ensuring a sustained and stable H₂ release, which effectively facilitated the formation of collagen fibers and blood vessels. Particularly, we proposed that H₂ promotes tissue remodeling through coordinated activation of MAPK signaling in both gingival fibroblasts and vascular endothelial cells, coupled with stimulation of pro-angiogenic paracrine crosstalk of gingival fibroblasts. In summary, the proposed local H₂-releasing strategy sequentially modulates the formation of STI around the implant by adapting to local microenvironmental pH changes, thereby enhancing the STI to ensure long-term success of the implant.

Fig. 1.

Schematic illustration for the preparation and proposed mechanism of the sequential H₂-release system for enhancing STI. (A) The preparation of HP-MTi. (B) The sequential regulation of STI process. (C) The proposed immunomodulation and pro-remodeling mechanism of H₂ for enhancing STI.

2. Results

2.1. The physiological structure and formation process of STI

To elucidate the physiological mechanisms underlying the susceptibility of STI around implants to disruption, an implantation model in Sprague-Dawley rats was constructed to compare the differences of STI between implants and contralateral natural molars (Fig. 2A and Fig. S1). Histological analysis (Hematoxylin-Eosin staining, Van Gieson staining and Masson staining) demonstrated significant morphological differences between peri-implant connective tissue and periodontal connective tissue (Fig. 2B). In stark contrast to natural teeth, the connective tissue fibers surrounding implants were sparser, with a distinct difference in fiber orientation (Fig. 2C and Fig. S2). In natural teeth, gingival fibers extending from the cementum at the cervical region toward the crown were observed. These fibers function to pull the gingiva, ensuring gingival tissue close attachment to the teeth and stabilizing the gingival tissue [25]. Instead, soft tissues around implants exhibited a markedly distinct fiber architecture characterized by fibers running parallel to the implant surface, extending from the alveolar crest to the free gingival margin. No fibers perpendicular to the implant were present owing to the lack of cementum and periodontal ligament structures on the implant surface. Furthermore, quantitative histomorphometric analysis revealed significantly reduced collagen fiber density in peri-implant connective tissues compared to natural periodontal tissues (Fig. 2D). This structural deficiency coincided with impaired vascularization, as evidenced by diminished CD31⁺ expression in the soft tissue around implants (Fig. 2E). With absence of periodontal ligament, the blood supply to the soft tissue around the implant is only derived from supra-periosteal blood vessels [26]. The reduced fiber and vascular around implants weaken the STI compared to natural teeth, making it more susceptible to disruption by bacterial infection and mechanical external forces, ultimately leading to implant failure [25,27].

Fig. 2.

The physiological structure and formation process of the STI around the implant. (A) Flow diagram of the establishment of the implantation model in rats. (B) Schematic illustration of comparison of the STI between natural tooth and implant. (C) Representative images of soft tissue around natural teeth and implants stained with VG, Masson's trichrome, and CD31 immunohistochemistry. Scale bars, 200 μm. Magnify bar, 100 μm. (D) Quantitative analysis of collagen density, n = 6. (E) Quantitative analysis of CD31 expression level, n = 6. (F) Schematic illustration of the formation of the STI around the implant. (G) Representative images of STI around implants in different time points, stained with Hematoxylin-Eosin (H&E). Scale bars, 200 μm. Magnify bar, 50 μm. The value is expressed as the means ± SD. (∗∗∗P < 0.001).

A comprehensive investigation of STI process will provide critical mechanistic insights into the structural disparities. To systematically characterize the temporal progression of STI, we performed longitudinal histological analyses at critical healing timepoints (the 3^rd, 7^th, 14^th, and 28^th days post-implantation) (Fig. 2G). On the 3^rd day post-implantation, the soft tissue around the implant exhibited significant inflammatory infiltration, with collagen fiber degeneration and epithelial dysplasia, which reflecting a typical phenomenon of acute inflammation. By the 7^th day, persistent inflammatory infiltration and a disorganized, loosely arranged collagen network were observed. By the 14^th day, the inflammatory infiltration had significantly subsided, and fiber arrangement became more uniform, suggesting that the state of tissue remodeling. At day 28, histological analysis demonstrated more compact collagen fiber arrangement with epithelial regression, indicating progressive STI maturation. On the basis of the above results, it is suggested that the formation of the STI around the implant proceeds through a coordinated series of stages, which are divided into two main sequential phases: the inflammatory phase and the remodeling phase (Fig. 2F) [4,7,28]. During the inflammatory phase, immune cells rapidly infiltrate the implantation site, initiating an acute inflammatory response to clear cellular debris and eliminate pathogens. This process is supported by a metabolic shift to anaerobic glycolysis, leading to substantial production of ROS and lactate, which create a markedly acidic local microenvironment [29,30]. Then the immune microenvironment progressively transitions from a pro-inflammatory state to a pro-remodeling state. Due to their impressive plasticity, macrophages play an important role in coordinating the STI process, undergoing marked phenotypic and functional changes depending on different phases [31]. In the remodeling stage, the process primarily involves the generation of collagen fibers and blood vessels, which is mainly mediated by gingival fibroblasts and umbilical vein endothelial cells [28]. Consequently, coordinated immunomodulation during the inflammatory phase to accelerate pro-remodeling microenvironment establishment, coupled with enhanced collagen fiber regeneration and vascularization during the remodeling phase, represents a promising strategy to optimize STI.

2.2. Preparation and characterization of H₂-releasing titanium surface

To sequentially regulate the STI process around the implant, an on-demand and long-lasting H₂-releasing titanium surface was designed. As illustrated in Fig. 1A, AB@mPDA NPs were synthesized and subsequently loaded into the titanium surface treated with micro-arc oxidation (MAO).

Firstly, we synthesized mesoporous polydopamine nanoparticles (mPDA NPs) via the soft-template method based on previous studies [32,33], and subsequently utilized these mPDA NPs for AB loading. The mesoporous structure of mPDA NPs provides a significantly larger contact area which enables the efficient loading and limited hydrolysis of AB [34]. Transmission electron microscope (TEM) images showed that the prepared nanoparticles exhibit a well-defined and highly ordered mesoporous structure, with an average particle size of approximately 144.3 nm (Fig. 3A, and Fig. S3). In elemental mapping, it clearly illustrated the element distribution of C and B, indicating the successful encapsulation of AB. The results from the dynamic light scattering (DLS) revealed that the average hydration sizes of mPDA NPs and AB@mPDA NPs were about 300 nm (Fig. S4). Furthermore, the zeta potential of mPDA NPs was -29.8 mV, while the zeta potential of AB@mPDA NPs increased to -9.09 mV (Fig. 3B). Fourier transform infrared spectroscopy (FTIR) confirmed the successful loading of AB. The AB@mPDA NPs exhibited characteristic peaks of AB at 2385 cm⁻¹ and mPDA at 1506 cm⁻¹ [33], with an AB loading efficiency in mPDA NPs of approximately 24.7 wt% (Fig. S5).

Fig. 3.

Characterization of HP-MTi. (A) Representative TEM mapping images of AB@mPDA NPs. Scale bar, 100 nm. (B) Zeta potentials of mPDA NPs and AB@mPDA NPs. (C) Representative SEM images of HP-MTi. Scale bar, 5 μm. (D) High-resolution XPS spectrum of mPDA NPs and AB@mPDA NPs and the corresponding resolution spectrum for (E) N 1s. (F) Quantitative analysis of water contact-angle measurements and the representative images inserted on the top of each column. (G) Quantitative analysis of roughness, and (H) representative AFM images of HP-MTi. (I) Cumulative release curve of H₂ from AB and HP-MTi in pH6.8, pH7.0, and pH7.4 for 72 h. (J) Quantitative analysis of the •OH scavenging ratio of MTi, P-MTi, and HP-MTi in pH6.8, pH7.0, and pH7.4. (∗P < 0.05, ∗∗P < 0.01, and ∗∗∗∗P < 0.0001.)

Subsequently, we loaded AB@mPDA NPs onto the titanium surface treated with MAO, thereby fabricating the H₂-releasing system on the titanium surface (HP-MTi). The titanium surface was observed using SEM, showing that the MAO-Ti surface (MTi) exhibits a three-dimensional porous structure (Fig. S6), which facilitates enhanced AB@mPDA NPs loading [35,36]. As shown in Fig. 3C, a substantial number of AB@mPDA NPs were observed to be loaded into the pores. Based on the nanoconfinement effect, this hierarchical nanostructure limits the hydrolysis of AB to prolong H₂ generation [37]. The elemental composition and surface bonding of HP-MTi were further investigated using X-ray photoelectron spectroscopy (XPS) (Fig. 3D, E and Fig. S7). The high-resolution XPS spectra of HP-MTi showed a new peak at 390.00 eV, which was correspond to the structure of B-N [17]. In addition, the results of water contact-angle measurements showed that the surfaces of MTi, mPDA NPs-loaded MAO-Ti (P-MTi) and HP-MTi exhibit outstanding hydrophilic properties, and the loading of mPDA NPs and AB@mPDA NPs caused a decrease in the water contact angle (Fig. 3F), which is beneficial to improve biocompatibility [38,39]. The AFM measurement indicated that the loading of mPDA NPs and AB@mPDA NPs had no influence on the surface roughness (Fig. 3G and H).

More importantly, H₂ release from HP-MTi was analyzed under mild acidic (pH 6.8, simulating an inflammation environment), neutral (pH 7.0), and normal physiological (pH 7.4) conditions using gas chromatography [40]. As displayed in Fig. 3I, HP-MTi sustainably released H₂ for at least 72 h under these three pH conditions. Especially, the release rate of H₂ markedly accelerates under acidic conditions, whereas under neutral conditions, it exhibits a relatively sluggish release behavior, thus demonstrating distinct pH-dependent characteristics. The in vivo release data demonstrated that HP-MTi achieved a cumulative release rate of 89.23 ± 1.92 % within the first 7 days, whereas mathematical modeling projected a total release duration of up to 21 days (Fig. S8). Furthermore, we investigated the ROS scavenging capabilities of different surfaces against •OH radicals under different pH conditions, together with bare Ti as a negative control. As illustrated in Fig. 3J, HP-MTi exhibited remarkable •OH scavenging ability, highlighting the antioxidant properties of H₂ for biomedical applications. Notably, this scavenging efficacy was distinctly strengthened in an acidic microenvironment, aligning with the pH-responsive nature of AB.

2.3. Evaluation of biocompatibility

Following successful fabrication of the H₂-releasing implant on titanium substrates, a comprehensive evaluation of its biocompatibility properties was conducted through systematic in vitro and in vivo experiments. To evaluate the cytocompatibility of HP-MTi, RAW 264.7, human gingival fibroblasts (HGFs), and human umbilical vein endothelial cells (HUVECs) were separately inoculated on three titanium surfaces and cultured for 24 h. Live/dead staining demonstrated excellent cellular compatibility across all modified surfaces, with all three cell types maintaining high viability rates and exhibiting negligible cytotoxicity (Fig. S9A). CCK-8 assays further substantiated these findings, revealing consistently high cellular proliferation rates across all experimental groups. Notably, for HGFs, the cell viability rate of the HP-MTi group was significantly higher compared to the other two groups (Fig. S9B), suggesting a potential role of H₂ in promoting cell proliferation. Given that the implant comes into direct contact with blood upon implantation, hemolysis experiments were further conducted to assess the biocompatibility of each surface with blood cells. As demonstrated in Fig. S9C, all groups exhibited the hemolysis rate below 5 %, confirming excellent blood compatibility. Furthermore, major organs were examined to visualize histological changes after implantation. H&E staining confirmed the biosafety of HP-MTi as no significant pathological alterations were observed in the heart, liver, spleen, lungs, or kidneys (Fig. S9D). These findings demonstrate the outstanding biocompatibility of this implant, fulfilling a critical prerequisite for in vivo implantation.

2.4. Evaluation of immunomodulatory effects

After implantation, blood clots form rapidly, which triggers the recruitment of immune cells to the implantation site. Then, the soft tissue around the implant experiences an acute inflammatory phase. Although inflammation serves as a critical mechanism in the regeneration process, immunomodulation to achieve the timely and effective resolution of inflammation and forming a pro-remodeling microenvironment is essential for the formation of STI [4,41]. To examine the immunomodulatory effects of the H₂-releasing system, an in vitro inflammatory model was established by culturing macrophages on various modified titanium substrates under lipopolysaccharide (LPS) stimulation (Fig. 4A). The intracellular ROS levels were quantified using DCFH staining and flow cytometry. Following LPS stimulation, the ROS content in cells significantly increased, reflecting oxidative stress alterations within the cells. Notably, macrophages cultured on HP-MTi surfaces exhibited substantially reduced ROS accumulation, highlighting the potent antioxidant capacity of this H₂-releasing implant. (Fig. 4B–D). Given mitochondria's pivotal role as the major intracellular ROS source, mitochondrial membrane potential was assessed via JC-1 staining. HP-MTi surface most effectively maintained mitochondrial membrane potential changes, further substantiating that H₂ preserves mitochondrial function by mitigating oxidative damage (Fig. 4C, E).

Fig. 4.

Evaluation of immunomodulatory effects in the inflammation phase. (A) Schematic illustration of HP-MTi scavenging ROS to reduce macrophages polarization to pro-remodeling phenotype. (B) Flow cytometry analysis of intracellular ROS of LPS-stimulated RAW 264.7 after treatment of MTi, P-MTi, and HP-MTi. (C)Representative laser confocal microscopy images of intracellular ROS and JC-1 of LPS-stimulated RAW 264.7 after different treatment. Scale bar, 20 μm. Quantitative analysis of (D) DCFH and (E) JC-1. (F) Representative laser confocal microscopy images and 3D surface plot of iNOS and Arg-1 labled LPS-stimulated RAW 264.7 after different treatment. Scale bar, 20 μm. Quantitative analysis of fluorescence intensity of iNOS (G) and Arg-1 (H) in LPS-stimulated RAW 264.7 after different treatment. (I) Levels of IL-6, TNF-α, IL-10 and TGF-β in LPS-stimulated RAW 264.7 after different treatment. (J) Relative mRNA level of IL-6, TNF-α, CD 86, iNOS and TGF-β in LPS-stimulated RAW 264.7 after different treatment. The value is expressed as the means ± SD, with a minimum sample size of 3. (∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, and ∗∗∗∗P < 0.0001.)

Critically, emerging evidences indicated that intracellular ROS can dynamically regulate macrophage polarization [42,43]. To evaluate the immunomodulatory effects of HP-MTi on macrophage polarization, immunofluorescence staining for the pro-inflammatory M1 marker inducible nitric oxide synthase (iNOS) and the anti-inflammatory M2 marker arginase-1 (Arg-1) were performed. LPS stimulation robustly induced classical M1 polarization, as evidenced by significantly elevated iNOS expression concomitant with suppressed Arg-1 expression relative to unstimulated controls. No statistically significant differences were observed between the P-MTi and MTi groups compared to the LPS-treated group. In contrast, macrophages cultured on HP-MTi exhibited a significant reduction in iNOS expression and a notable increase in Arg-1 expression (Fig. 4F–H). These findings collectively demonstrate that HP-MTi actively modulates macrophage, inducing anti-inflammatory M2 polarization while suppressing pro-inflammatory M1 polarization. It is well established that macrophages are key regulators of tissue regeneration, undergoing marked phenotypic and functional changes to ensure proper STI [31]. M1 macrophages predominantly secrete pro-inflammatory mediators (e.g. IL-6, TNF-α) that amplify inflammatory cascades, whereas M2 macrophages produce anti-inflammatory cytokines (e.g. IL-10, TGF-β) that resolve inflammation and promote tissue repair [44]. The regulatory effects of HP-MTi on cytokines factors were validated using the enzyme-linked immunosorbent assay (ELISA) and qPCR. From Fig. 4I and J, HP-MTi significantly reduced pro-inflammatory cytokine expression (TNF-α, IL-6) while enhancing anti-inflammatory factors (IL-10, TGF-β) compared to both P-MTi and MTi groups.

Based on these findings, we speculated that H₂ mediated ROS scavenging to reprogram macrophages polarization from pro-inflammatory M1 to pro-remodeling M2 phenotypes, accelerating the formation of a pro-remodeling immune microenvironment. M2 macrophages can product numerous growth factors, including transforming growth factor β (TGF-β) and vascular endothelial growth factor A (VEGFA), which promote gingival fibroblasts and vascular endothelial cells proliferation, as well as angiogenesis, thereby enhancing STI around the implant (Fig. 4A) [31].

2.5. Evaluation of pro-remodeling effects

After the inflammatory phase, the soft tissue around the implant progressively transitions into the remodeling phase [7]. As main cell types in soft tissue around the implant, the biological activities of HGFs and HUVECs, including proliferation, migration and secretion, are crucial to the STI.

The biological response of HGFs to H₂ was assessed through systematic evaluation of cellular behavior across differentially modified titanium substrates. Firstly, the proliferation behavior of gingival fibroblasts was assessed using Edu experiment and CCK-8 assays. The Edu experiment labeled the proliferation of HGFs cells within approximately 10 % of the cell cycle (Fig. 5A and B), showing that the proliferation rate of HGFs cultured on HP-MTi increased significantly. CCK-8 assays quantitatively demonstrated the superior proliferative capacity of HGFs on HP-MTi surfaces throughout the 5-day culture period (Fig. 5C). Subsequently, the cytoskeleton of HGFs was labeled with F-actin after inoculation for 12 h. HGFs cultured on HP-MTi exhibited the greatest number of pseudopodia, suggesting an enhancement in their cell migration capability (Fig. 5D). The scratch assay and Transwell assay further demonstrated that HP-MTi significantly enhanced the migratory capability of HGFs, which is conducive to the formation of STI (Fig. 5E–G, and Fig. S10). Furthermore, quantitative assessment through immunofluorescence staining and ELISA revealed that HP-MTi significantly enhanced the expression of key extracellular matrix components, including fibronectin (FN), Collagen types I (COL1), and Collagen types III (COL3), in cultured HGFs compared to control surfaces (Fig. 5H–K). It has been extensively documented that FN serves as a critical component of the integration between the soft tissue and the implant surface [45,46]. COL1 and COL3 constitute the fundamental structural and functional components of soft tissue around implants, playing essential roles in barrier defense, inflammation regulation, tissue repair and regeneration [13]. These findings collectively demonstrate that H₂ can pronouncedly enhance the functional behavior of HGFs, including proliferative, migratory, and secretory activities, which are crucial for promoting STI around implants.

Fig. 5.

Evaluation the influence on the biological behaviors of HGFs in the remodeling phase. (A) Representative images and (B) quantitative analysis of Edu staining of HGFs cultured in MTi, P-MTi, and HP-MTi. Scale bar, 100 μm. (C) Quantitative analysis of the relative proliferation ratio of HGFs cultured in different surfaces. (D) Representative images of the actin cytoskeleton of HGFs cultured on various substrates with phalloidin-FITC (green) staining. Scale bar, 20 μm. (E) Illustration of the Transwell assay scheme. (F) Quantitative analysis and (G) representative images of transwelled HGFs cultured in different surfaces. Scale bar, 200 μm. (H) Representative IF images of FN-1 and COL1A, and (I) representative laser confocal microscopy images of COL3A1 in HGFs. Scale bar, 50 μm. (J) Quantitative analysis of fluorescence intensity of FN-1, COL1A and COL3A1. (K) Levels of FN-1, COL1A and COL3A1 in HGFs. The value is expressed as the means ± SD, with a minimum sample size of 3. (ns, not significant; ∗P < 0.05; ∗∗P < 0.01; ∗∗∗∗P < 0.0001.)

Similarly, we investigated the regulatory effect of H₂ on HUVECs. The Edu staining and the CCK-8 experiment demonstrated that the proliferation rate of HUVECs cultured on HP-MTi was significantly higher than that on other surfaces (Fig. 6A–C). It indicated that H₂ also exerts the promoting effect on proliferation of HUVECs. Comprehensive evaluation of HUVECs migratory capacity through cytoskeletal organization analysis, scratch assay, and Transwell assay consistently demonstrated significantly superior migration capacity in the HP-MTi group compared to control surfaces (Fig. 6D–G and Fig. S11). Particularly, angiogenesis is considered a critical factor in the process of tissue remodeling [47]. To investigate the regulatory effect of H₂ on the angiogenesis, the expression levels of the angiogenesis-related VEGF-A and bFGF were measured using immunofluorescence, ELISA assays and qPCR. Both transcriptional and translational expression levels of VEGF-A and bFGF in the HP-MTi group were significantly higher than those in the P-MTi and MTi groups (Fig. 6H, K, L and Figure S12). This pro-angiogenic effect was functionally validated through in vitro tube formation assays. As presented in Fig. 6I and J, HP-MTi presented significantly increase in branch points and tubular junctions relative to P-MTi and MTi groups, reflecting a superior capacity for angiogenesis.

Fig. 6.

Evaluation the influence on the biological behaviors of HUVECs in the remodeling phase. (A) Representative images and (B) quantitative analysis of Edu staining of HUVECs cultured in MTi, P-MTi, and HP-MTi. Scale bar, 100 μm. (C) Quantitative analysis of the relative proliferation ratio of HUVECs cultured in different surfaces. (D) Representative images of the actin cytoskeleton of HUVECs cultured on various substrates with phalloidin-FITC (green) staining. Scale bar, 20 μm. (E) Illustration of the Transwell assay scheme. (F) Quantitative analysis and (G) representative images of transwelled HUVECs cultured in different surfaces. Scale bar, 200 μm. (H) Representative laser confocal microscopy images of VEGFA in HUVECs. Scale bar, 50 μm. (I) Representative images and (J)quantitative analysis of tube formation assays. Scale bar, 500 μm.Magnify scale bar, 200 μm. (K) Relative mRNA levels of VEGFA and bFGF in HUVECs. (L) Levels of VEGFA and bFGF in HUVECs. The value is expressed as the means ± SD, with a minimum sample size of 3. (ns, not significant; ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001; and ∗∗∗∗P < 0.0001.)

2.6. Evaluation of in vivo therapeutic efficacy

To further demonstrate the superiority of our strategy, the therapeutic effect of this H₂-releasing system was tested in the rat implantation model (Fig. 7A). On the basis of the STI process discussed in Fig. 2F and G, we harvested specimens on 7th and 28th day post implantation, corresponding to the inflammatory phase and remodeling phase, respectively.

Fig. 7.

HP-MTi notably enhanced the STI around implants in both inflammation phase and remodeling phase. (A) Schematic illustration of in vivo evaluation plan. (B) Representative IF images of iNOS and F4/80 and in the soft tissue around implants from different groups, and (C) quantitative analysis. Scale bar, 20 μm. (D) Representative IF images of Arg-1 and F4/80 and in the soft tissue around implants from different groups, and (E) quantitative analysis. Scale bar, 20 μm. (F) Levels of IL-6, TNF-α, IL-10 and TGF-β in the soft tissue around implants from different groups. (G) Relative mRNA level of IL-6, TNF-α, IL-10 and TGF-β in the soft tissue around implants from different groups. (H) Representative VG and Masson staining image of soft tissue around implants from different groups. Scale bar, 200 μm. Magnify scale bar,100 μm. (I) Representative laser confocal microscopy images and (J)quantitative analysis of CD31 in the soft tissue around implants form different groups. Scale bar, 50 μm. (K)Quantitative analysis of collagen density. (L) Level of COL1A and VEGFA in the soft tissue around implants form different groups. (M) Relative mRNA level of COL1A and VEGFA in the soft tissue around implants form different groups. The value is expressed as the means ± SD, with a minimum sample size of 3. (ns, not significant; ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001; and ∗∗∗∗P < 0.0001.)

The in vivo immunomodulatory capacity of H₂-releasing implant was validated on 7th day post implantation. Analysis of oxidative stress markers revealed that the HP-MTi group displayed significantly lower levels of malondialdehyde (MDA) and superoxide dismutase (SOD), as well as a notably higher reduced glutathione/oxidized glutathione (GSH/GSSG) ratio relative to the control groups (Fig. S13). Immunofluorescence analysis revealed that HP-MTi significantly reduced iNOS expression while increasing Arg-1 expression in F4/80⁺ macrophages compared to controls (Fig. 7B–E). Complementary ELISA and qPCR analyses confirmed these polarization shifts, with HP-MTi exhibiting downregulation of pro-inflammatory cytokines (TNF-α, IL-6) and upregulation of anti-inflammatory cytokines (IL-10, TGF-β) at both protein and transcriptional levels (Fig. 7F and G). H&E staining revealed significant inflammatory cell infiltration in the soft tissue of both the MTi and P-MTi groups, suggesting a severe inflammatory response. In contrast, the HP-MTi group exhibited a marked reduction in inflammatory cell infiltration, indicating basically resolution of inflammation (Fig. S14).

Subsequently, at day 28 post-implantation, in vivo validation studies revealed substantial improvements in STI. Histological evaluation using Masson and Van Gieson staining showed that the HP-MTi group exhibited denser collagen fiber arrangement and significantly increased collagen deposition compared to control groups, indicating markedly enhanced STI (Fig. 7H–K). In addition, comparative analysis revealed that the HP-MTi group exhibited substantially increased COL-1 expression in soft tissue around implants compared to other treatment groups (Fig. 7L and M). Meanwhile, the expression of CD31 and VEGF in the soft tissue around the implants in the HP-MTi group was significantly upregulated, with a notable increase in the number of blood vessels (Fig. 7I–M). In summary, comprehensive evaluation confirmed dual capacity of H₂ to promote collagen fiber formation and functional neovascularization, ultimately yielding structurally robust STI marked by mature collagen architecture and stable vascular networks.

The osseointegration cannot be neglectable for the long-term success of implants [48]. Consequently, we further discussed the potential influence of this H₂-releasing implant on osseointegration. As illustrated in Fig. 3G, the maximal and stable concentration achieved by this H₂ release system is lower than the saturated solubility of hydrogen gas in water (800 × 10⁻⁶ M) [18], thereby preventing bubble formation to hinder osseointegration. Micro-CT analysis further confirm that the system has no adverse effects on osseointegration when compared to MTi (Fig. S15).

2.7. Exploring the mechanism of the pro-remodeling effect

While existing researches has extensively characterized the antioxidant properties of H₂ [[49], [50], [51], [52]], its potential role in modulating tissue remodeling remains poorly understood. Building upon our confirmation that H₂ promotes soft tissue remodeling, we explored the underlying molecular mechanisms of this phenomenon (Fig. 8A).

Fig. 8.

RNA-seq analysis revealing the potential mechanize of H₂ promoting STI. (A) Schematic illustration of H₂ activating the MAPK signaling pathway in HGFs and HUVECs to promote cell proliferation, and enhancing the secretion of paracrine angiogenic cytokines by HGFs to induce angiogenesis. (B) The volcano plot depicting different expressed genes between HP-MTi and P-MTi. (C) GO enrichment analysis revealing the up-regulated terms between HP-MTi and P-MTi. GSEA analysis of (D) the positive regulation of cell proliferation and (E) angiogenesis. (F) KEGG enrichment analysis revealing the up-regulated pathways between HP-MTi and P-MTi. (G) Visualizing the top ten significantly enriched pathways and their corresponding differentially expressed genes through KEGG enrichment analysis with a chord diagram. (H)GSEA analysis of MAPK signaling pathway. (I) Heat map of differential expressed genes associated with the MAPK signaling pathway.

RNA sequencing analysis was performed on HGFs cultured on HP-MTi and P-MTi substrates. Differential expression analysis identified 228 significantly regulated genes, comprising 104 upregulated and 124 downregulated transcripts (Fig. 8B and Fig. S16). The GO enrichment analysis and GSEA analysis indicated that H₂ has the potential to upregulate critical biological processes, including cell proliferation, cell migration, and angiogenesis (Fig. 8C–E). The KEGG pathway enrichment analysis suggested that H₂ could upregulate several critical pathways, including cytokine-cytokine receptor interaction, PI3K-Akt signaling, MAPK signaling, JAK-STAT signaling, and others (Fig. 8F–H). Critically, the MAPK signaling pathway plays a pivotal role in regulation of cell proliferation and angiogenesis [53]. Meanwhile, MAPK signaling pathway is a highly conserved signal transduction pathway widely distributed across various cell types, its core kinase cascade (Ras-Raf-MEK-ERK) exhibits significant functional conservation among different cell types [54]. Inspired by these findings, we hypothesized that H₂ may exert its effects by upregulating the MAPK signaling pathway in both HGFs and HUVECs. Furthermore, the analysis of differential genes in the MAPK signaling pathway revealed that the gene expressions of angiogenesis-related genes CXCL8 and VEGF were significantly upregulated in the HP-MTi group (Fig. 8G, I). There is a complex interdependent relationship between angiogenesis and fibroplasia, which is driven by the dynamic reciprocity among cellular components, matrix proteins, and bioactive molecules [55,56]. The essential role of fibroblasts in angiogenesis under both physiological and pathological conditions has been widely acknowledged [57]. The secretion of angiogenesis-related proteins CXCL8 and VEGF by HGFs can stimulate angiogenesis, indirectly further facilitating STI. In summary, we speculated that H₂ orchestrates STI through complementary mechanisms involving both direct cellular activation and paracrine-mediated effects. By stimulating MAPK signaling pathways in HGFs and HUVECs, H₂ promotes cellular proliferation while simultaneously enhancing the angiogenic paracrine activity of HGFs through upregulated CXCL8 and VEGF secretion. This dual action creates a synergistic microenvironment conducive to collagen fiber production and neovascularization, ultimately leading to improved structural and functional STI (Fig. 8A).

To investigate the regulatory effect of H₂ on the MAPK signaling pathway (RAS/RAF/MEK/ERK pathway) in HGFs and HUVECs (Fig. 9A), Western blot analysis targeting the key molecules p-Erk and Erk in MAPK signaling pathway was performed. Following H₂ treatment, the phosphorylation level of Erk protein was markedly upregulated in both HGFs and HUVECs (Fig. 9B, C and Fig. S17), a finding corroborated by immunofluorescence analysis showing elevated p-Erk expression and nuclear accumulation (Fig. 9D and Fig. S18). Complementary immunofluorescence staining and Western blot analyses revealed a coordinated upregulation of Ras, Raf, and pMEK/MEK, which are established upstream regulators of the MAPK signaling pathway (Fig. 9E–G and Fig. S19). Functional assays confirmed that MAPK pathway activation translated to increased proliferative capacity in both cell types (Fig. S20). Consistent with in vitro results, in vivo analysis demonstrated significantly higher p-Erk and Ki67 levels in peri-implant soft tissue from the HP-MTi group (Fig. S21, S22). Collectively, the aforementioned findings established that H₂ can activate the MAPK signaling pathway in HGFs and HUVECs, driving cellular proliferation (Fig. 9A).

Fig. 9.

The regulatory mechanism of H₂ on tissue remodeling. (A) Schematic illustration of H₂ activating the MAPK signaling pathway in HGFs and HUVECs to promote cell proliferation. (B) Western blot image and (C) quantitative analysis of Erk and p-Erk expression in HGFs and HUVECs. (D) Representative IF images and co-location analysis of p-Erk and Dapi in HGFs and HUVECs. Scale bar, 10 μm. (E) Representative laser confocal microscopy images of Ras in HGFs and HUVECs. Scale bar, 20 μm. Quantitative analysis of fluorescence intensity of Ras in (F) HGFs and (G) HUVECs. (H) Schematic illustration of the collection of the supernatants from HGFs for culturing HUVECs. Levels of (I) CXCL8 and VEGFA of the supernatants from HGFs. (J) Representative immunofluorescence images and (K) quantitative analysis of VEGFs in HUVECs cultured by the supernatants from HGFs. Scale bar, 20 μm. (L) Representative images and (M)quantitative analysis of tube formation assays. Scale bar, 200 μm. The value is expressed as the means ± SD, with a minimum sample size of 3. (ns, not significant; ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001; and ∗∗∗∗P < 0.0001.)

Subsequently, to assess the regulatory effect on the paracrine function of HGFs, the supernatants from HGFs cultured on three distinct types of titanium surfaces were collected and utilized for culturing HUVECs (Fig. 9H). Firstly, the expression levels of CXCL8 and VEGF in the supernatants of HGFs cultured on different titanium surfaces were analyzed by ELISA. As expected, the HP-MTi group presented significantly higher expression levels of CXCL8 and VEGF, compared to those in the P-MTi group and the MTi group (Fig. 9I). Immunofluorescence images also demonstrated that VEGF expression level in HUVECs was up-regulated in the HP-MTi group (Fig. 9J and K). As before, angiogenic potential was assessed through tube formation assays. As depicted in Fig. 9L and M, treatment with the supernatants from HP-MTi resulted in a significant increase in the number of branch nodes, junctions, and mesh structures, compared to that in P-MTi group and the MTi group, implying an enhanced angiogenic capacity. These data demonstrate that H₂ treatment upregulates pro-angiogenic cytokine secretion (VEGF and CXCL8) in HGFs through paracrine mechanisms, functionally promoting angiogenesis. Therefore, the above results showed H₂ coordinately enhances STI through MAPK pathway-mediated cellular activation and amplified paracrine signaling. This synergistic effect drives collagen fiber formation and functional neovascularization, ultimately achieving stable STI around the implant.

3. Discussion and conclusion

In this study, the established STI around the implant was demonstrated to differ from that around the natural tooth in fiber and vasculature. This inherent physiological structure predisposes the STI to easy disruption and infection, which is a critical factor contributing to implant failure [58]. A sound investigation of the STI process will enhance our understanding of this process and offer actionable insights for developing effective strategies to promote STI. We found that the STI process can be broadly categorized into two progressive phases: the inflammatory phase and the remodeling phase. After implantation, the surrounding soft tissues initiate an acute inflammatory response to eliminate tissue debris and pathogens as the normal defensive physiological mechanism. Through the regulation of immune cells, the local microenvironment progressively transitions from the inflammation phase to the remodeling phase. During the remodeling phase, activated soft tissue cells undergo coordinated proliferation and differentiation, orchestrating the collagen fiber production and angiogenesis, ultimately establishing a stable STI [4,27,28]. Currently, emerging surface modification strategies for implants exhibit limitations in effectively regulating both two stages. More importantly, there is a conspicuous absence of sequential dynamic regulation linking these two stages.

Given the potential promise of H₂ for coordinately immunomodulation and pro-remodeling coupling, we developed an on-demand and long-lasting H₂ release system in the titanium surface for enhancing STI. With the pH-responsive properties of AB and the pH changes during the STI, the prepared implant can on-demand release H₂ for immunoregulatory in the inflammatory phase. In the remodeling phase, this system sustainably releases H₂ with the hierarchical nanostructure, facilitating tissue regeneration. This implant achieved the on-demand transition of H₂ release kinetics that matches the temporal progression of soft tissue integration, yielding promising outcomes in both in vitro and in vivo experiments. In the inflammatory phase, H₂ treatment effectively scavenged excessive ROS, thereby promoting the polarization of macrophages from the pro-inflammatory M1 phenotype to the pro-remodeling M2 phenotype, which is conductive to accelerate the transition to the remodeling phase. Then in the remodeling phase, H₂ treatment effectively promoted the formation of fibers and blood vessels in soft tissues around the implant, thereby enhancing resistance to external stimuli and preventing the invasion of pathogens.

With its potent antioxidant activity, high biosafety, and excellent tissue permeability, H₂ possesses significant potential as a bioactive gas molecule in biomedicine [59,60]. Current studies have mostly focused on the antioxidant mechanism of H₂, whereas few investigations have explored its impacts on soft tissue cells under normal physiological conditions [51,61]. Few studies indicated that H₂ may function in promoting tissue remodeling, while the underlying mechanism of H₂ on tissue remodeling still require more comprehensive and deep exploration [20,62]. Herein, on the basis of the results of transcriptomic analysis and relevant literature reports, we propose that H₂ may promote the proliferation of gingival fibroblasts and umbilical vein endothelial cells by activating the MAPK signaling pathway, directly benefiting fiber and blood vessel formation. Additionally, H₂ may enhance the paracrine function of gingival fibroblasts, leading to the release of pro-angiogenic factors and thereby indirectly promoting angiogenesis. Emerging evidence indicates that Fe-porphyrin acts as a molecular target for hydrogen molecules, significantly accelerating the reaction between H₂ and •OH via catalytic hydrogenation [[63], [64], [65], [66]]. This, in turn, suggests that the initiation of MAPK signaling by H₂ is a multifaceted process. This process may be orchestrated through an indirect cellular response to the modulation of redox homeostasis or the direct modulation of Fe-porphyrin-containing proteins, ultimately converging on the MAPK cascade.

Despite the substantial therapeutic effects observed in rats implanted with H₂ releasing implants, several limitations in our study must be acknowledged. Firstly, STI is a multifactorial process. While our findings suggest H₂ may enhance this process, the underlying molecular mechanisms remain to be fully elucidated. Secondly, as a proof-of-concept, this study reveals the potential role of H₂ in enhancing STI, it should be noted that our experimental conditions are unable to fully replicating the physiological microenvironment, including dynamic fluid flow and physiological pH variations. Third, the lack of perpendicularly inserted collagen fibers at the implant interface is one of key factors undermining STI. While our current strategy improved STI, it did not achieve guided fiber orientation. Developing engineering implant surfaces with specific biomechanical cues or incorporating bioactive agents to induce collagen fiber insertion may further enhance STI.

In conclusion, this local H₂ releasing strategy effectively enhance STI to ensure long-term success of the implant, with substantial potential for possible clinical translation for both trans-mucosal implant collars and healing abutments. Notably, the investigation of the underlying mechanisms paves the way for further research of therapeutic strategies to improve STI and the extended application of H₂ therapy in diseases related to tissue remodeling.

4. Experimental section

Establishment of rat implantation model: Screw-shaped Ti implants (Φ1.1 × 4 mm) were manufactured by Shandong Weigao Orthopedic device Co., Ltd. (Shandong, China) and Ti implants was treated in the same way as Ti substrates in vitro. Rat implantation model was established as our pervious study [14]. Briefly, 4-week-old male Sprague Dawley rats (∼250 g) were used. One month following the extraction of the maxillary first molar in rats, different Ti implants were implanted into the extraction socket. After 3,7,14, and 28 days, the maxillary and the soft tissue around implants of the rats were collected for histological evaluation, Immunofluorescence, micro-CT scans, ELISA, q-PCR and RNA-sequencing. The animal experiments were approved by The Laboratory Animals Welfare and Ethical Center of School of Stomatology, Air Force Military Medical University (No. 2024kq-039).

Histological evaluation: The maxillary bone tissues of rats were decalcified with a 10 % EDTA solution for 4 weeks after fixation with 4 % paraformaldehyde. Conventional paraffin embedding after decalcification was performed. Paraffin-embedded tissues were cut into 4-μm-thick sections. H&E staining, VG and Masson staining were performed according to the manufacturer's protocols.

Preparation of HP-MTi: First, MAO-Ti plates were prepared according to our previous study. After grinding with 400–1200 grit silicon carbide sandpaper, smooth Ti plates (2 mm thick, 10 mm diameter) were treated by micro arc oxidation with an electrolyte consisting of 0.04 M β-glycerophosphate and 0.2 M calcium acetate using a Ti sheet as an anode, and a stainless-steel plate as a cathode. The applied voltage, oxidizing time, frequency, and duty ratio were 300 V, 5 min, 400 Hz, and 8 %, respectively.

Next, 1.0 g Triblock copolymer Pluronic F127 was dissolved in an ethanol/water mixture (50 ml:50 ml). After stirring for 15 min, 1.0 mL TMB was added to the solution for the formation of F127/TMB single micelles. 750 mg Dopamine hydrochloride was added to the aforementioned solution, stirring for 10 min to form the F127/TMB/DA composite unimicelles. 1.5 mL Ammonia water was added to the monomer solution to initiate the polymerization of DA, finally assemble F127/TMB/DA single micelles into mPDA. Subsequently, the sample was purified using a sequential treatment with ethanol and deionized water, and then carried out freeze-drying. 100 mg of AB was fully dissolved in 1 mL of a 5 mg/mL mPDA aqueous solution. Following vortex mixing for 1 min, the mixture was centrifuged at 10,000 rpm for 5 min. The supernatant was discarded, and the pellet was washed with water three times to obtain AB@mPDA.

Finally, HP-MTi was prepared according to our previous study. In detail, 20 mg of AB@mPDA were added to 5 mL of phosphate-buffered saline (PBS) and sonicated for 20 min. Next, 0.2 mL of the suspension was dropped onto the MTi surface and then oscillated for 1 h. As a control, the same concentration of mPDA solution was added dropwise to the MTi surface, which was denoted as P-MTi.

In vitro H₂ Release Measurement: The quantitative analysis of the gaseous H₂ accumulated during the reaction of AB and HP-MTi with water was obtained using gas chromatography. Specifically, the AB and HP-MTi was added to PBS with pH values of 6.8, 7.0 and 7.4) in a self-made airtight container filled with nitrogen atmosphere. Subsequently, the signal of H₂ was collected by a gas chromatography (GC, Dataphysics OCA20) at different points in time. The results are obtained with reference to the integrated area of the pure hydrogen gas as standard sample. The volume of H₂ generation per unit mole of AB and HP-MTi has been calculated and normalized according to the state equation of ideal gas.

Intracellular ROS analysis of RAW264.7: RAW264.7 cells were inoculated onto samples (MTi, P-MTi, and HP-MTi) and treated with LPS (100 ng/ml) for 12 h. The 2,7-dichlorodihydrofluorescein diacetate (DCFH-DA, Beyotime, S0033) assay was used to detect ROS. A working solution of DCFH-DA (20 μM, prepared in FBS-free cell culture medium) was added to each group at a volume of 1 mL and incubated for 30 min at 37 °C. After washing with PBS, intracellular DCF fluorescence were observed by CLSM. Fluorescence intensity was analyzed using ImageJ software. Furthermore, ROS was measured by flow cytometry and then analyzed by FlowJo software. cells were incubated with the JC-1 reagent for 20 min and subsequently imaged using confocal laser scanning microscopy (CLSM).

In vitro polarization of RAW264.7: RAW264.7 cells were seeded onto the surface of titanium sheets. After incubating for 12 h, RAW264.7 cells were treated with LPS (100 ng/ml) to induce macrophages polarization. After treating for 12 h, the cells were stained with anti-iNOS for the flow cytometer and CLSM analysis to analyze the M1 polarization. After continuing treatment for 36 h, the cells were stained with anti-Arg-1 for the flow cytometer and CLSM analysis to analyze the M2 polarization.

Cell proliferation assay: 5-Ethynyl-2′-deoxyuridine (EdU) staining was conducted using BeyoClick™ EdU Cell Proliferation Kit with Alexa Fluor 594 (Betotime, C0078S, China). Based on the incorporation of the thymidine analog EdU during the DNA synthesis phase, EdU is specifically labeled with Alexa Fluor 594 followed by a Click chemistry reaction. HGFs and HUVECs were seeded onto the surface of titanium sheets for 12 h. After adding Edu working solution, cells were incubated for 18 h. Following the completion of Edu labeling, cells were fixed with paraformaldehyde for 15 min, then permeabilized using 0.3 % Triton X-100 for 15 min. After adding Click reaction solution, cells were incubated in the dark for 30 min, and then the nuclei were stained with Dapi. Finally, observe the stained cells under CLSM. The cell proliferation ability was evaluated by CCK-8. HGFs, and HUVECs were cultured respectively into three groups of Ti plates, and incubated at 37 °C with 5 % CO₂ for 1, 3, 5 days. Then, the absorbance was measured by a microplate reader (TECAN) at 450 nm after incubation of cell counting kit-8 (CCK-8) solution for 1 h. The proliferation ratio was calculated using the following formula and plot the proliferation curve: Cell proliferation ratio (%) = (OD _sample - OD _control)/OD _control × 100 %

Migration and invasion assay: An in vitro wound healing assay was employed to evaluate cell migration capabilities. HGFs and HUVECs were seeded in different Ti plates. Following a 24-h incubation period, a uniform scratch was introduced using a 200-μl pipette tip, after which the cells were cultured in a serum-free medium to minimize proliferation effects. The cells were stained with ghost pen cyclic peptide at 0, 12 and 24 h after the scratch, and then observed under a fluorescence microscope.

Cell invasion was measured using the Transwell assay. HGFs and HUVECs were first seeded onto the surfaces of various titanium sheets for 24 h, then digested and inoculated into the upper chamber at a density of 3 × 10⁴ cells/ml. The medium with FBS were added to the lower chamber, while the medium without FBS were added to the upper chamber. A final fixation and staining procedure were performed after 24 h. The invading cells were then observed using optical microscopy.

Tubule formation assays: A tubule formation assay was conducted using Matrigel (HY-K6001; MCE) to evaluate the vascularization potential. After being seeded onto the surface of different titanium sheets for 24 h, HUVECs were trypsinized and subsequently added to each well of a 24-well plate pre-coated with Matrigel. Images were captured using an inverted microscope after 12 h.

Immunofluorescence: Immunohistochemistry was used to stain the cells and tissues. Cells were fixed with pre-cooled paraformaldehyde for 30 min, and then permeabilized with 0.1 % Triton X-100 for 30 min. Subsequently, they were blocked with 5 % bovine serum albumin (BSA; Gibco) for 1 h. Followed by being incubated with the specific primary antibody (anti-FN-1, anti-COL1, anti-COL3, anti-VEGF, anti-p-Erk, and anti-Ras) at 4 °C overnight, they were incubated with a fluorescent secondary antibody for 2 h at room temperature in the dark. The nuclei were labeled with DAPI. Finally, they were observed under CLSM.

After fixation, decalcification, paraffin embedding, and cutting, tissues were cut into 4-μm-thick sections. The sections were deparaffinized, rehydrated, and blocked with 5 %BSA, then incubated for 12 h with specific antibodies (anti-iNOS, anti-Arg-1, anti-F4/80, anti-p-Erk, anti-Ki67, and anti-CD31) at 4 °C. Subsequently, the samples were incubated with the corresponding secondary antibodies for 2 h at room temperature in the dark. The nuclei were stained with DAPI for visualization. Finally, the samples were observed under CLSM.

RNA isolation and quantitative real-time PCR (qPCR): Total RNA was extracted from the primary cell culture using Trizol reagent. First-strand cDNA was synthesized using a RevertAid First Strand cDNA Synthesis kit (Termo Fisher Scientifc, Waltham, USA). qPCR was performed on a Real-time quantitative PCR instrument (ABI, 7500) using the SYBR Green Real-time PCR Master Mix (TOYOBO, Shanghai, China). Gene expression levels were presented as mRNA levels, normalized to the reference housekeeping gene (Gapdh) using the △△CT method. The sequences of the primer pairs are provided in the supplementary information (Table S1).

Enzyme-linked immunosorbent assay (ELISA): The supernatants of the cell culture were collected and purified by centrifugation. The supernatants were analyzed using ELISA kits (Elabscience, China) to measure IL-1β, IL-10, TNF-α,TGF-β,FN-1,COL1A,COL3A1,VEGF, bFGF, and CXCL8 levels. All assays were performed strictly in accordance with the manufacturer's instructions.

Micro-CT analysis: After 28 days of implantation, rats were euthanized for Micro-CT scanning to assessed the osseointegration of implants. The maxillaries of rats were scanned by Micro-CT (AX-2000, Always Imaging, China) with 130 kV in voltage and 110 μA in current. Three dimensional reconstructions were processed by VG Studio MAX 3.5 (Volume Graphics, Heidelberg, Germany) to quantitatively analyze the ratio of bone volume to total volume (BV/TV), trabecular thickness (Tb.Th), trabecular number (Tb.N), and trabecular spacing (Tb.Sp).

Western blotting: RIPA buffer supplemented with phosphatase and protease inhibitors was used to extract proteins from HGFs and HUVECs. We separated 10 μg of protein per lane by 4–20 % SDS–polyacrylamide gel electrophoresis and subsequently transferred the samples onto polyvinylidene difluoride (PVDF) membranes. After blocking with 5 % nonfat milk, the membranes were incubated with primary antibodies overnight at 4 °C, then were incubated with anti-mouse or anti-rabbit secondary antibodies (Immunoway, China). Subsequently, enhanced chemiluminescence (ECL) reagents were utilized to detect the protein bands on the membranes with high sensitivity and specificity.

Statistical analysis: GraphPad Prism (version 10.0) was used for statistical analyses. Data from at least three independent experiments are presented as means ± SD. Two independent groups were compared using an unpaired two-tailed Student's t-test, while multiple groups were assessed using one-way analysis of variance (ANOVA) followed by Tukey's post hoc tests for multiple comparisons. A two-sided P value < 0.05 was considered statistically significant.

CRediT authorship contribution statement

Yue Yuan: Writing – review & editing, Writing – original draft. Zishuo Hou: Writing – original draft, Investigation. Miaomiao Chen: Software, Investigation. Jingwei Yu: Investigation, Formal analysis. Minghao Zhou: Visualization, Validation. Jiaxin Kang: Formal analysis. Tengjiao Wang: Writing – review & editing, Supervision, Project administration, Conceptualization. Peng Li: Project administration, Investigation, Funding acquisition. Hongbo Wei: Supervision, Project administration, Investigation, Funding acquisition.

Ethics approval and consent to participate

All the animal experimental procedures were conducted in accordance with institutional guidelines for the care and use of laboratory animals and protocols, which were approved by The Laboratory Animals Welfare and Ethical Center of School of Stomatology, Air Force Military Medical University (No. 2024kq-039).

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: