Modulation of osteoclast differentiation by epigallocatechin-3-gallate (EGCG)/type I collagen–coated titanium implant surface in vitro

Abstract

Background/purpose

Implant osseointegration depends not only on osteogenesis but also on coordinated bone remodeling. Our previous study demonstrated that epigallocatechin-3-gallate (EGCG)/type I collagen–coated titanium (Ti) surfaces enhance angiogenesis and osteogenic differentiation in vitro. However, their influence on osteoclast differentiation remains unclear. This study evaluated the modulatory effects of this coating on osteoclast responses in vitro.

Materials and methods

Sandblasted and acid-etched (SLA) Ti surfaces followed by alkaline treatment (SLAA) were prepared. Type I collagen was immobilized onto SLAA surfaces using the natural crosslinker EGCG (10 or 50 μg/mL). Surface morphology, roughness, hydrophilicity, and functional groups were characterized. RAW 264.7 macrophages were used to assess cell adhesion, proliferation, viability, tartrate-resistant acid phosphatase (TRAP) expression, and F-actin ring formation. Osteoclast differentiation was induced using receptor activator of nuclear factor kappa-B ligand (RANKL). Statistical analysis was performed using one-way ANOVA with Dunnett’s test.

Results

Surface characterization confirmed successful coating without compromising surface stability. Compared with SLA surfaces, collagen-coated surfaces with or without EGCG enhanced initial cell spreading while showing less increase in metabolic activity without cytotoxicity. EGCG- and EGCG/type I collagen–coated surfaces significantly reduced TRAP expression and F-actin ring size relative to SLA surfaces. Although the 50 μg/mL EGCG groups showed lower TRAP expression and smaller F-actin rings than the 10 μg/mL groups, the differences were not statistically significant.

Conclusion

EGCG/type I collagen surface modification may modulate osteoclast-related cellular responses in RAW 264.7 cells without affecting cell viability, suggesting a potential strategy to modulate peri-implant bone resorption and support long-term implant stability.

Introduction

Long-term stability of titanium (Ti) dental implants depends on preservation of peri-implant bone, yet marginal bone loss (MBL) and peri-implantitis remain major biological complications.1 Peri-implant disease involves not only bacterial biofilms but also host-mediated inflammatory and bone remodeling responses,1,2 and early radiographic MBL has been linked to implant prognosis.3 Accordingly, implant surface strategies should promote osseointegration while limiting excessive bone resorption. Because bone remodeling is governed by osteoblasts and osteoclasts mainly through the RANKL–RANK–OPG pathway,4,5 and increased osteoclastic activity with altered RANKL/OPG balance has been observed in peri-implant lesions,6–8 modulation of osteoclast differentiation may help preserve peri-implant bone stability.9

Among commercially available Ti dental implant surfaces, sandblasted/acid-etched (SLA) Ti is a widely used micro-roughened surface with substantial clinical documentation.10 Accordingly, SLA is commonly adopted as a reference in Ti dental implant surface research. Our previous study further generated a superhydrophilic nanonetwork structure on SLA Ti via alkaline treatment to enhance cellular responses in vitro and in vivo.11 Although surface microtopography promotes early bone apposition, emerging osteoimmunology evidence suggests that surface-mediated regulation of inflammatory and osteoclast-related responses may also affect long-term peri-implant bone maintenance.12 Consistently, biomimetic surface modifications have been reported to attenuate macrophage activation and osteoclastogenesis, potentially supporting osseointegration under compromised systemic conditions.13

In addition to topographical modification, biofunctional coating incorporating extracellular matrix (ECM) components on Ti and its alloys has been proposed to improve biological performance.14,15 Type I collagen coating enhances early bone cell responses to Ti surfaces.16,17 However, Ti surfaces optimized for osteogenesis may not adequately modulate osteoclast-mediated resorption under inflammatory conditions, highlighting the need for strategies that promote bone formation while controlling osteoclast activity.

Epigallocatechin-3-gallate (EGCG), a major green tea polyphenol, exhibits anti-inflammatory and anti-osteoclastogenic properties.18–21 In our previous study, EGCG (50 μg/mL) served as a natural crosslinker for type I collagen immobilization on alkaline-treated SLA Ti surfaces and enhanced osteogenic and angiogenic responses in vitro.15 However, its effects on osteoclast differentiation remain unclear. Therefore, this study investigated whether EGCG and/or type I collagen coating modulates osteoclas-trelated cellular responses on Ti surfaces in vitro. We hypothesized that EGCG-containing and collagen-coated surfaces would attenuate RANKL-induced osteoclast differentiation without impairing cell viability. Two EGCG concentrations (10 and 50 μg/mL) were evaluated, and RAW 264.7 cell adhesion, proliferation, viability, and RANKL-induced osteoclast differentiation were analyzed.

Materials and methods

Materials preparation

Biomedical grade IV pure Ti discs (15 mm diameter, 1.5 mm thickness) were used as substrates. Sandblasting and acid etching produced SLA surfaces (control). Subsequent alkaline immersion treatment generated a superhydrophilic nano-network structure on SLA, termed SLAA. Detailed surface preparation procedures were described previously.15

Biomolecular functionalization of SLAA specimens was performed using collagen and/or EGCG. SLAA specimens immersed in 0.05 wt% type I collagen at 37 °C for 24 h were designated the collagen group (C). To assess the effect of EGCG on osteoclastogenesis, SLAA specimens were treated with 10 or 50 μg/mL EGCG under the same conditions and labeled E1 and E2, respectively. For EGCG-mediated type I collagen crosslinking, SLAA specimens were incubated in 0.05 wt% type I collagen containing either 10 or 50 μg/mL EGCG and designated E1C and E2C accordingly. The stated EGCG concentrations (10 and 50 μg/mL) refer to the preparation concentrations used in the coating solution during surface functionalization and do not represent directly measured surface-bound amounts or soluble exposure concentrations during cell culture. Additional details of the surface functionalization protocol have been described previously.15

Surface characterizations

Surface morphology was examined by scanning electron microscopy (SEM; JSM-6500F, JEOL, Tokyo, Japan) to observe high-resolution structural features. Surface roughness was quantitatively evaluated by measuring the arithmetic mean height (Sa) using a white-light interferometric 3D optical profilometer (Profilm3D, KLA-Filmetrics, San Diego, CA, USA). After being stored in a moisture-proof box for a long time (30 d), the surface hydrophilicity of test specimens was determined via static water contact angle measurements with a contact angle goniometer (100SB, Sindatek, New Taipei City, Taiwan) to assess whether the modified surfaces retained their hydrophilicity after prolonged dry storage before potential use. Type I collagen immobilization was verified by X-ray photoelectron spectroscopy (XPS; ESCA-003800, ULVAC-PHI, Kanagawa, Japan), focusing on the N1s signal. Functional groups were further characterized using Fourier transform infrared spectroscopy (FTIR; Vertex 80v, Bruker, Ettlingen, Germany).

It was noted that EGCG was incorporated into the immersion/coating solution as a natural crosslinking-functionalization component during surface preparation. In the present study, the mode of EGCG retention on the Ti surface was not directly distinguished as physical adsorption versus chemical binding. Therefore, coating reproducibility was inferred indirectly from the consistent SEM, wettability, XPS, and FTIR findings across the prepared groups.

Osteoclast responses: adhesion, proliferation, viability, differentiation, and activity

Before cell seeding, all specimens were sterilized by exposure to UV light (254 nm) at a distance of 80 cm for approximately 10 min, and the same sterilization protocol was applied to all groups to minimize potential effects on the surface coatings. RAW 264.7 murine macrophages were obtained from the Bioresource Collection and Research Center (BCRC, Hsinchu, Taiwan) and seeded onto prepared specimens at 2 × 10⁴ cells/disc. Cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum and sodium bicarbonate (Sigma–Aldrich, St. Louis, MO, USA). Experimental procedures are briefly summarized below.

For adhesion analysis, cells were incubated for 1 d, rinsed with phosphate-buffered saline (PBS), and fixed in 2% glutaraldehyde/4% paraformaldehyde at 37 °C for 30 min, then stored at 4 °C. After washing, specimens were dehydrated through graded ethanol, critical point dried, sputter-coated with platinum, and observed by SEM (JEOL) to examine cell morphology.

Cell proliferation was assessed using the alamarBlue® assay. Cells cultured on different Ti surfaces were evaluated on days 1, 3, and 5 by incubation with 10% alamarBlue® reagent (Thermo Fisher Scientific, Eugene, OR, USA) for 4 h. Absorbance at 570 and 595 nm was measured using a microplate reader, and the reduction ratio (%) was calculated according to the manufacturer’s instructions.

Cell viability on days 1, 3, and 5 was determined using a LIVE/DEAD® Viability Kit (calcein AM/ethidium homodimer-1) (Thermo Fisher Scientific). After staining at 37 °C for 20 min in the dark, fluorescence images were captured using an upright fluorescence microscope (AZ100, Nikon, Tokyo, Japan) and analyzed with Image-Pro Express 6.0 software (Media Cybernetics, Bethesda, MD, USA).

For osteoclast differentiation, cells (2 × 10⁴ cells/disc) were cultured for 24 h, then induced with osteoclastogenic medium containing 50 ng/mL RANKL, refreshed every 48 h. Cells were analyzed after 1, 4, and 7 d. Following fixation (4% paraformaldehyde), permeabilization (0.3% Triton X-100), and blocking (5% normal goat serum), specimens were incubated with anti-TRAP primary antibody (4 °C, 16–20 h) and Alexa Fluor® 488 secondary antibody (1.5 h) (Thermo Fisher Scientific). F-actin was stained with rhodamine phalloidin, and nuclei with DAPI (Thermo Fisher Scientific). Fluorescence images were obtained using an upright fluorescence microscope (Nikon). Image-Pro Express 6.0 software (Media Cybernetics) was used to quantify F-actin ring size per cell to evaluate osteoclast activity across surface treatments.

Statistical analysis

All cell-based experiments were conducted as three independent experiments (biological replicates, N = 3), with five parallel specimens analyzed per group at each time point within experiment (technical replicates, n = 5). Data are expressed as mean ± standard deviation. Differences among groups were analyzed separately at each time point by one-way analysis of variance (ANOVA), followed by Dunnett’s post hoc test using SLA as the reference control. Statistical significance was defined as P < 0.05. Significance levels are denoted as *P < 0.05, **P < 0.01, and ***P < 0.001 compared with the SLA control group.

Results

Fig. 1 presents representative SEM images of the test surfaces, along with corresponding surface roughness (Sa) values and 30-d water contact angles provided below the images. SLA showed the typical micron/submicron topography, whereas alkali treatment generated an additional nanoscale mesh-like network, resulting in a hierarchical structure. This multiscale morphology was preserved after type I collagen, EGCG, or collagen/EGCG coating. All groups exhibited comparable roughness (Sa: 0.7–1.0 μm), with only slight reductions after biomolecular coating. In contrast, wettability changed markedly: SLA was relatively hydrophobic (79 ± 9°), SLAA was hydrophilic (34 ± 1°), and all biomolecule-coated surfaces remained superhydrophilic (<10°). Because all coated groups showed contact angles below 10°, small differences among the biomolecule-coated groups were not overinterpreted.

Figure 1.

Surface morphology, roughness, and wettability of the test Ti surfaces.

Representative SEM images of SLA, SLAA, C, E1, E2, E1C, and E2C surfaces. SLA exhibited the characteristic micron/submicron topography, whereas alkaline treatment (SLAA) introduced an additional nanoscale mesh-like network, resulting in a hierarchical nano-/submicron-/micron-scale structure. This multiscale morphology was preserved after type I collagen, EGCG, or collagen/EGCG coating. The arithmetic mean height (Sa) and 30-d static water contact angle are shown below the corresponding SEM micrographs. All groups exhibited comparable roughness (Sa: 0.7–1.0 μm); however, surface wettability differed markedly, with all biomolecule-coated surfaces remained superhydrophilic (<10°). SEM, scanning electron microscopy; EGCG, epigallocatechin-3-gallate; SLA, sandblasted and acid-etched Ti; SLAA, SLA followed by alkaline treatment; C, type I collagen-coated SLAA; E1 and E2, SLAA treated with 10 and 50 μg/mL EGCG, respectively; E1C and E2C, type I collagen-coated SLAA crosslinked with 10 and 50 μg/mL EGCG, respectively.

Fig. 2 shows the XPS and FTIR analysis results used to confirm biomolecule coating on the test surfaces. These analyses were used as confirmatory evidence of surface functionalization rather than as full quantitative compositional analyses. A marked N1s signal was detected in the type I collagen-coated groups, irrespective of EGCG incorporation, indicating the presence of collagen (Fig. 2A). Since the N1s peak area, which indirectly represents the surface collagen content, did not show a clear trend among the type I collagen-coated groups, quantitative comparison was not conducted. Therefore, the XPS data were used mainly for qualitative interpretation of surface chemical characteristics. FTIR spectra showed the characteristic bands of type I collagen at approximately 1650 cm⁻¹ (amide I), 1550–1560 cm⁻¹ (amide II), and 1240 cm⁻¹ (amide III), together with EGCG-related bands at 3200–3500 cm⁻¹ (O–H stretching) and ~1620 cm⁻¹ (aromatic C＝C) (Fig. 2B), confirming the presence of collagen and/or EGCG on the coated surfaces.

Figure 2.

Surface chemical characterization of the test Ti surfaces.

(A) X-ray photoelectron spectroscopy (XPS) spectra showing the N1s signal used to verify type I collagen immobilization on the modified surfaces. (B) Fourier transform infrared spectroscopy (FTIR) spectra of the test surfaces. Characteristic collagen bands were observed at around 1650 cm⁻¹ (amide I), 1550–1560 cm⁻¹ (amide II), and 1240 cm⁻¹ (amide III), whereas EGCG-related bands were observed at 3200–3500 cm⁻¹ (O–H stretching) and around 1620 cm⁻¹ (aromatic C＝C), confirming the presence of collagen and/or EGCG on the coated surfaces. EGCG, epigallocatechin-3-gallate; SLA, sandblasted and acid-etched Ti; SLAA, SLA followed by alkaline treatment; C, type I collagen-coated SLAA; E1 and E2, SLAA treated with 10 and 50 μg/mL EGCG, respectively; E1C and E2C, type I collagen-coated SLAA crosslinked with 10 and 50 μg/mL EGCG, respectively.

Fig. 3 summarizes the early cellular responses of RAW 264.7 cells, including 1-h adhesion morphology and metabolic activity up to 5 d. Results showed that collagen-containing surfaces (C, E1C, and E2C) promoted greater initial cell spreading and exhibited less further increase in metabolic activity from day 3 to day 5. These findings suggest modulation of early cell behavior rather than simple enhancement of cell proliferation. Accordingly, a live/dead viability assay was performed to determine whether the lower metabolic increase was associated with controlled cell growth rather than cytotoxicity.

Figure 3.

Initial adhesion morphology and metabolic activity of RAW 264.7 cells on the test surfaces.

(A) Representative scanning electron microscope images of RAW 264.7 cells after 1 h of adhesion. Greater cell spreading was observed on type I collagen–containing surfaces (C, E1C, and E2C). (B) AlamarBlue assay showing metabolic activity on days 1, 3, and 5. Collagen-containing surfaces showed less further increase in metabolic activity from day 3 to day 5. Data are presented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001 versus SLA. EGCG, epigallocatechin-3-gallate; SLA, sandblasted and acid-etched Ti; SLAA, SLA followed by alkaline treatment; C, type I collagen-coated SLAA; E1 and E2, SLAA treated with 10 and 50 μg/mL EGCG, respectively; E1C and E2C, type I collagen-coated SLAA crosslinked with 10 and 50 μg/mL EGCG, respectively.

Fig. 4 further shows that RAW 264.7 cell viability remained generally preserved on all test surfaces over the same 5-day period. Live/dead staining confirmed that cells remained predominantly viable on all surfaces during the early culture period. Although live cell expansion appeared less marked on collagen-containing surfaces, dead-cell staining remained minimal, suggesting the interpretation that the lower metabolic increase on collagen-containing groups observed in Fig. 3 was associated with controlled cell growth rather than cytotoxicity.

Figure 4.

Live/dead viability of RAW 264.7 cells cultured on the test surfaces.

Representative fluorescence images of live cells (green) and dead cells (red) on days 1, 3, and 5. Live cells were observed on all surfaces, with minimal dead-cell staining at early time points. By day 5, dead cells became more evident on non-collagen-coated surfaces, whereas collagen-containing surfaces (C, E1C, and E2C) still showed minimal cell death. EGCG, epigallocatechin-3-gallate; SLA, sandblasted and acid-etched Ti; SLAA, SLA followed by alkaline treatment; C, type I collagen-coated SLAA; E1 and E2, SLAA treated with 10 and 50 μg/mL EGCG, respectively; E1C and E2C, type I collagen-coated SLAA crosslinked with 10 and 50 μg/mL EGCG, respectively.

In contrast, Figs. 5 and 6 show later osteoclast-related responses of RAW 264.7 cells after RANKL induction, including F-actin ring formation and qualitative TRAP expression. All biomolecule-modified surfaces reduced F-actin ring perimeter compared with SLA, and the C, E1, E2, E1C, and E2C groups showed significantly smaller actin rings, indicating attenuation of osteoclast maturation on these modified surfaces (Fig. 5). Although the 50 μg/mL EGCG groups showed lower mean values than the 10 μg/mL EGCG groups, no significant difference was detected between the two concentrations. Consistently, qualitative TRAP fluorescence images showed lower TRAP expression on collagen- and/or EGCG-modified surfaces than on SLA (Fig. 6), suggesting the overall interpretation that these surface modifications attenuated osteoclast-related differentiation and activity. TRAP fluorescence intensity was not quantitatively analyzed because reliable image-based quantification was limited by the hierarchical nano-/submicron-/micron-scale surface topography.

Figure 5.

F-actin ring formation of RANKL-induced RAW 264.7 cells on the test surfaces.

Representative fluorescence images and quantitative analysis of actin ring perimeter per cell after osteoclast induction. All biomolecule-modified surfaces reduced F-actin ring size relative to SLA, and the C, E1, E2, E1C, and E2C groups exhibited significantly smaller actin ring perimeters, indicating suppression of mature osteoclast formation. Data are presented as mean ± SD. *P < 0.05 and **P < 0.01 versus SLA. EGCG, epigallocatechin-3-gallate; SLA, sandblasted and acid-etched Ti; SLAA, SLA followed by alkaline treatment; C, type I collagen-coated SLAA; E1 and E2, SLAA treated with 10 and 50 μg/mL EGCG, respectively; E1C and E2C, type I collagen-coated SLAA crosslinked with 10 and 50 μg/mL EGCG, respectively.

Figure 6.

TRAP expression of RANKL-induced RAW 264.7 cells on the test surfaces.

Representative fluorescence images showing TRAP staining of RAW 264.7 cells after osteoclast induction on the indicated surfaces. Lower qualitative TRAP fluorescence was observed on the I collagen- and/or EGCG-modified surfaces than on SLA. TRAP fluorescence intensity was not quantitatively analyzed because reliable image-based quantification was limited by the hierarchical nano-/submicron-/micron-scale surface topography. EGCG, epigallocatechin-3-gallate; SLA, sandblasted and acid-etched Ti; SLAA, SLA followed by alkaline treatment; C, type I collagen-coated SLAA; E1 and E2, SLAA treated with 10 and 50 μg/mL EGCG, respectively; E1C and E2C, type I collagen-coated SLAA crosslinked with 10 and 50 μg/mL EGCG, respectively.

Discussion

In the present study, the effects of EGCG and/or type I collagen coating on osteoclast differentiation on Ti surfaces were evaluated in vitro. For clarity, the following discussion is organized to address surface characterization first, followed by cellular responses, potential biological mechanisms, and study limitations.

Surface characterization confirmed successful coating without marked compromise of surface integrity (Figs. 1 and 2). In the following comparisons, SLA was treated as the principal reference surface, whereas SLAA was regarded as the intermediate post-alkaline-treatment surface. Because surface roughness (Sa), hydrophilicity, and overall surface topography were comparable among the biomolecule-coated groups, the observed biological differences are more likely attributable to differences in surface chemistry. Accordingly, the biological findings are better interpreted as surface chemistry-dependent modulation of osteoclast-related behavior rather than as a nonspecific consequence of altered topography. The combined reduction in TRAP expression and F-actin ring size on EGCG-containing surfaces suggests attenuation of osteoclast maturation under RANKL induction, while the preserved viability profile argues against a cytotoxic explanation.

Maintenance of peri-implant bone levels depends on balanced bone remodeling. Although osteoclast activity is necessary for physiologic turnover, excessive osteoclastogenesis has been associated with MBL around implants.1,6 Recent clinical studies indicate that peri-implant crevicular fluid levels of RANKL relative to OPG are elevated in peri-implantitis cases, and that RANKL/OPG imbalance correlates with disease severity.1,8,22 These findings suggest that modulation of osteoclast differentiation may be relevant to limiting bone loss in peri-implant contexts.

With respect to cellular responses, the greater early spreading observed on type I collagen-containing surfaces suggests that collagen facilitated initial cell–surface interaction, whereas the smaller subsequent increase in alamarBlue activity may reflect altered cell behavior during differentiation rather than overt cytotoxicity. When interpreted together with the preserved live/dead viability and the smaller per-cell F-actin ring perimeter, these findings support the view that surface chemistry influenced osteoclast functional maturation instead of simply suppressing cell survival. This interpretation is consistent with previous reports that type I collagen can regulate RAW 264.7 proliferation through PI3K- and MAPK-dependent modulation of cyclin expression.23 In addition, EGCG-containing surfaces may further attenuate RANKL-induced maturation, potentially through mechanisms related to NF-κB suppression reported in previous studies.24 Because the abovementioned assays were conducted under different culture conditions, direct quantitative comparison may not be appropriate; however, the combined results suggest that EGCG/collagen surface modification may influence both baseline macrophage responses and RANKL-induced osteoclast maturation.

From a biological and mechanistic perspective, EGCG has been associated with anti-inflammatory and anti-osteoclastogenic properties in various experimental models.18,19 In addition to inhibition of NF-κB and downstream transcriptional regulators such as c-Fos and NFATc1,20,21 recent studies suggest that EGCG can attenuate osteoclastogenesis under inflammatory and pathological conditions.21 Although intracellular signaling pathways were not examined in the present study, the reduced TRAP expression and F-actin ring formation observed here are consistent with previously reported anti-osteoclastogenic effects of EGCG.

In our previous study, 50 μg/mL EGCG was used primarily to investigate osteogenic responses.15 In the present study, both 10 and 50 μg/mL EGCG inhibited osteoclast differentiation. Although the 50 μg/mL group showed lower mean TRAP expression and smaller F-actin rings than the 10 μg/mL group, the difference was not statistically significant. Thus, a definitive concentration-dependent effect could not be established under the present experimental conditions. Variations from previous reports may reflect differences in cell type, microenvironment, and mode of presentation.19,25 These factors may partly account for differences between the present findings and those reported previously. In this context, the present findings may offer preliminary information for the future development of EGCG-related biofunctional approaches in dental implant applications. Nevertheless, further in vitro and in vivo studies are needed to further clarify the dose–response relationship and clinical relevance of these findings.

Although alamarBlue showed less further increase on type I collagen–containing surfaces, this assay reflects metabolic reduction capacity rather than absolute cell number.26 Live/dead staining and DAPI labeling suggested that the reduced signal was not primarily attributable to overt cytotoxicity, but rather to altered cell behavior during differentiation.27,28 In addition, reduced F-actin ring perimeter on EGCG-containing surfaces may reflect altered osteoclast functional maturation rather than reduced cell survival.29

Importantly, RANKL-induced osteoclast maturation is characterized by cytoskeletal reorganization and actin ring formation, which reflect functional polarization rather than proliferation status.29 Thus, the reduction in per-cell F-actin ring perimeter observed on EGCG-containing surfaces (Fig. 6) may indicate altered functional maturation of osteoclasts. Future studies incorporating normalization of differentiation markers to nuclei count or additional mechanistic evaluation would further clarify the relationship between metabolic activity and osteoclast differentiation.

Finally, several limitations should be acknowledged. First, this was an in vitro study using RAW 264.7 cells; therefore, the findings may not fully represent the complexity of peri-implant bone remodeling in vivo. Second, osteoclastogenesis was assessed mainly by TRAP expression and F-actin ring formation, whereas additional functional markers, such as resorption pit formation and osteoclast-related gene expression, were not analyzed. Third, although EGCG-containing surfaces reduced osteoclast differentiation, the underlying molecular mechanisms were not directly examined. Further studies are needed to clarify the signaling pathways involved and to evaluate the in vivo relevance of this coating strategy.

In summary, EGCG-mediated type I collagen coating on Ti surfaces attenuated osteoclast differentiation without compromising RAW 264.7 cell viability. Both EGCG concentrations inhibited osteoclast differentiation and activity, although no statistically significant difference was observed between 10 and 50 μg/mL. Together with previously reported pro-osteogenic and pro-angiogenic effects, these findings provide preliminary in vitro evidence that EGCG/type I collagen coating may support coordinated regulation of peri-implant bone remodeling. Nevertheless, further mechanistic and in vivo studies are needed.

Funding Statement

This research was funded by National Science and Technology Council (No. NSTC 112-2314-B-A49-020-MY3) and MacKay Memorial Hospital (Nos. MMH-114-54, MMH-114-056, and MMH-114-88) in Taiwan.