Superhydrophilic Hierarchical Anatase Coating on Sandblasted, Acid-Etched Titanium: In Vitro Apatite Formation and Osteoblast Responses and the Role of Polar Surface Free Energy

Abstract

Physicochemical modification of titanium implants aims to enhance early osseointegration by improving bioactivity. This study deposited and evaluated an anatase TiO₂ film on clinically relevant sandblasted, acid-etched titanium (Ti-SLA) to enhance in vitro bioactivity and osteogenic responses. An ~8 µm TiO₂-anatase coating was deposited on Ti-SLA by reactive pulsed DC magnetron sputtering. Surface characterization included FE-SEM, helium ion microscopy, and XRD. Wettability and surface free energy (SFE) were evaluated by contact angle analysis. In vitro bioactivity was assessed by hydroxyapatite (HA) formation in twofold-concentrated simulated body fluid (2× SBF). Osteoblast responses were evaluated through cell adhesion, viability, alkaline phosphatase activity, gene expression, and mineralization. The coating produced hierarchical multi-globular microstructures decorated with faceted anatase nanocrystals. Ti-SLA’s initial hydrophobicity converted to a superhydrophilic, high-energy surface with increased polar SFE. Homogeneous HA crystallites deposited exclusively on SLA-anatase in 2× SBF. SAOS-2 cells showed enhanced metabolic activity, ALP activity, osteogenic gene upregulation, and improved mineralized matrix, while primary human osteoblasts exhibited increased metabolic activity and calcium deposition. The anatase coating produced a superhydrophilic, high-energy micro-nano surface that accelerates HA formation and enhances osteoblast function in vitro, warranting in vivo validation for early osseointegration.

1. Introduction

Titanium and its alloys remain the materials of choice for load-bearing orthopedic and dental implants because of their excellent mechanical strength, corrosion resistance, and biocompatibility [1]. While titanium and its alloys are highly biocompatible and often described as “bioinert”, they are not biologically inert and can elicit a regulated foreign-body response in vivo. Osseointegration on unmodified titanium therefore occurs mainly through bone apposition and mechanical interlocking at the bone–implant interface, rather than through direct bone bonding [2,3,4]. Successful osseointegration, however, requires materials with bioactive properties, defined as the ability to elicit specific cellular responses that promote a stable material–bone bond [5]. Accordingly, enhancing the bioactivity of titanium implant surfaces has become a key objective in contemporary surface engineering [6,7].

Implant surfaces, serving as the initial interface with peri-implant tissues, have therefore been the main target for surface engineering to improve bioactivity and biocompatibility [8,9]. Numerous surface modification techniques including anodic oxidation [10], sandblasting/acid etching [11,12], UV irradiation, and H₂O₂ treatment [13] have been developed to alter surface topography and/or chemistry and thereby modulate the biological response. Among these, sand-blasted, acid-etched (SLA) surfaces, characterized by a typical arithmetic mean roughness (Sa) of 1–2 µm, have become a well-established clinical standard with proven long-term survival rates [12,14]. However, SLA surfaces are initially hydrophobic, mainly because of air entrapment within the micro-pitted morphology and adsorption of airborne hydrocarbons during storage, with additional aging effects [9,15]. This intrinsic hydrophobicity can hinder protein adsorption and delay early stages of the osteoblast response [16,17,18]. Human histometric studies have shown that hydrophilic SLA variants such as SLActive achieve significantly higher bone-to-implant contact during early healing (e.g., 2–4 weeks) [16]. At the materials level, hydrophilic SLAactive modifications are achieved by post-etch processing under protective gas conditions and liquid storage, which reduces hydrocarbon contamination, increases surface free energy, and can yield near-zero water contact angles (≈0°) compared with hydrophobic SLA (≈140°) [15,19]. Taken together, these data support SLA’s clinical reliability but also reveal early-stage limitations, motivating modifications strategies that improve wettability and the early biological response while preserving the clinically validated moderate microroughness (Sa ≈ 1–2 µm) that underlies SLA’s high bone–implant contact and mechanical interlocking [9,12].

Nanostructuring of titanium implant surfaces has been proposed as a promising strategy to improve their physicochemical properties and osteogenic potential. Recent studies have investigated nanoscale topographies such as anodically formed nanotubes and nanocrystalline TiO₂ coatings and their effects on surface properties and osteoblast responses [20,21,22,23]. Titanium dioxide (TiO₂) coatings in anatase, rutile, or mixed phases represent versatile approaches for generating nanostructured surface topographies [24]. Anatase-phase coatings have been shown to increase protein adsorption and enhance osteoprogenitor cell differentiation, attributed to their nanoscale crystalline topography and changes in surface wettability [20]. Multiple studies show that anatase can promote osteoblastic activity, increase cell adhesion, and accelerate early bone formation [25,26,27,28]. Early biological responses to titanium surfaces depend not only on topography, but also on interfacial energetics (surface free energy) and wettability during initial protein conditioning and cell adhesion [15,18,29].

The integration of a bioactive implant with bone tissue is typically accompanied by the deposition of a bone-like hydroxyapatite (HA) layer on its surface [30,31]. In this context, the in vivo bioactivity of an implant material refers to its ability to form a stable chemical bond with bone. In contrast, in vitro bioactivity is used operationally for early screening and is commonly assessed by apatite-forming propensity in simulated body fluid (SBF; ISO 23317:2014) together with osteoblast functional responses [31,32,33]. Accordingly, the SBF assay is regarded as a useful initial screening tool, though it does not provide definitive proof of in vivo bioactivity [34,35]. Only a few studies have examined how TiO₂ coatings influence hydroxyapatite formation, reporting that bi-phasic anatase–rutile films can alter HA morphology [36] and that pure anatase coatings can induce apatite formation on bioinert surfaces [25].

The majority of studies on anatase coatings have primarily focused on smooth or laboratory-prepared, acid-etched titanium surfaces, leaving the influence of an anatase coating on clinically relevant microrough SLA substrates insufficiently understood. This gap in knowledge pertains particularly to its effects on surface topography, surface chemistry, and in vitro bioactivity. Additionally, the physicochemical mechanisms underlying the enhanced biological performance of such coatings remain incompletely understood. Given the clinical significance of the SLA surface and the potential of anatase coatings to improve wettability and early osteoblast activity, this study aimed to evaluate the effects of depositing an anatase film on Ti-SLA and characterize the resulting surface energy, wettability, and topography, as well as its downstream effects on in vitro apatite formation, osteoblast proliferation, and differentiation. The study also aimed to assess how the measured physicochemical surface characteristics relate to apatite formation in SBF and osteoblast responses in vitro, thereby supporting a physicochemical basis for interpreting the observed bioactivity. This work is an experimental in vitro study in which the surface condition was the independent variable and physicochemical and cellular readouts were predefined outcomes. It was hypothesized that the anatase coating would enhance in vitro bioactivity and osteoblast responses by introducing nanoscale structures and modulating surface free energy and wettability on a microrough substrate.

2. Materials and Methods

2.1. Sample Preparation

In this study, round titanium grade 4 discs (Medentis Medical GmbH, Bad Neuenahr-Ahrweiler, Germany) with three different surface modifications were examined. These included polished (Ti-machined) samples that were 12 mm in diameter and 2 mm thick. The second group was generated by sandblasting and acid-etching of Ti-machined surfaces (10 mm in diameter and 2 mm thick), as described previously. Briefly, the discs were first grit-blasted with medical-grade corundum and then subjected to thermochemical treatment in a strong acid bath, yielding a microrough surface. This surface resembles the sandblasted/acid-etched (SLA) surface used in commercial implants and is termed Ti-SLA in this study [27]. For the third group, the microrough Ti-SLA discs were further modified through a unipolar pulsed DC sputtering process to form an anatase layer (SLA-anatase; 12 mm in diameter and 2 mm thick), as previously described. Briefly, Ti-SLA discs were first pretreated with argon plasma (surface cleaning and plasma polishing) and then coated by unipolar pulsed-DC reactive magnetron sputtering in an oxygen-containing atmosphere to deposit TiO₂, followed by a thermal annealing step to obtain the anatase phase [27]. All sample discs were supplied by Medentis Medical GmbH, Bad Neuenahr-Ahrweiler, Germany. For consistency, all biological assay results were normalized to the surface area of the 10 mm discs. Prior to subsequent analyses, all samples underwent ultrasonic cleaning in 70% ethanol for 5 min, disinfection with 70% ethanol for 15 min, and drying for 1 h in laminar airflow on a clean bench. Cells grown on tissue culture coverslips (Sarstedt, Newton, NC, USA) were included as a standard in vitro reference (positive control) to verify expected osteoblast behavior and assay performance. Ti-machined discs served as a smooth titanium baseline to contextualize the effects of microroughness and to evaluate whether anatase modification improves cellular responses relative to the initially hydrophobic Ti-SLA surface.

2.2. Surface Roughness and Microtopography

The roughness of surface variants was quantified through optical 3D confocal microscopy (MarSurf CM Explorer, Mahr, Goettingen, Germany). Each test disc was evaluated across six measuring squares, each with an area of 0.64 mm² and an edge length of 0.8 mm, using a 20× objective (0.64 NA). First, the surface was leveled using the least square plane leveling method, followed by the application of an S-filter (λs): Gaussian 300:1 (0.8 µm) to eliminate short-scale components attributed to possible instrument and ambient noise. Additionally, an L-filter (λc): Gauss 0.25 mm was used to separate the roughness from the waviness. The Sa values were derived from measurements conducted on five sample discs per surface variant, utilizing MountainsMap software (version 9.1.9957, Digital Surf, Besançon, France). The microtopography of the sample discs was also visually evaluated using a scanning electron microscope (LEO 1430, Zeiss, Oberkochen, Germany).

2.3. Wettability and Surface Free Energy

The wettability of the samples was evaluated by measuring the static water contact angle using a high-resolution drop shape analysis system (OCA 200, Dataphysics, Filderstadt, Germany). Specifically, 1 µL droplets of ultrapure water (Milli-Q; Merck Millipore, Darmstadt, Germany) were placed on the sample surface, and the drop contour was analyzed after a 10 s interval using the measurement software (SCA202 V.6.1.11, Dataphysics, Filderstadt, Germany). For each surface, five discs were analyzed, with three droplets at various sites per disc. Static water contact angles ≤ 5° were treated as 0° (complete wetting) because sessile-drop fitting is not robust at very low angles. The surface free energy (SFE) and its dispersive and polar components were calculated using the Owens, Wendt, Rabel, and Kaelble (OWRK) and Lifshitz-van der Waals/acid-base (LW-AB) approaches, respectively [15]. The SFE was computed by applying the OWRK and LW-AB equations to contact angle data obtained using ultrapure water and ethylene glycol (Thermo Scientific Chemicals, 99%, Darmstadt, Germany) as polar components and diiodomethane (Thermo Scientific Chemicals, 99+%, Darmstadt, Germany) as a dispersive agent. The physical properties of the test liquids used to calculate the SFE are provided in Table S1.

2.4. Nanostructure, Crystallinity, and Film Thickness

Phase analysis of the TiO₂ anatase-coated samples was performed using an X-ray Diffractometer (XRD, X’Pert MRD Pro, Malvern Panalytical GmbH, Kassel, Germany) with Cu Kα₁ radiation operated with 45 kV at 40 mA, using a symmetric θ–2θ scan (ω = θ) in the 2θ range of 20–50°. Reference diffraction patterns for anatase TiO₂, rutile TiO₂, and α-Ti from the ICSD database (ICSD release 2025.2) [37], were included in the plots for comparison and phase identification. The nanostructure of the anatase layer on SLA-anatase discs was examined using field emission electron microscopy (Fe-SEM, JSM-6500F, Jeol, Tokyo, Japan) at an accelerating voltage of 5 kV. For higher-resolution imaging, helium ion microscopy (HIM, Orion Nanofab HIM, Carl Zeiss Microscopy GmbH, Oberkochen, Germany) analysis was conducted on SLA-anatase samples. Nanoscale surface feature dimensions were quantified from HIM micrographs using ImageJ (version 1.53a). Because the features were triangular/faceted, the edge length was selected as the primary size descriptor. Fifty features (n = 50) were measured, and the results are reported as mean ± SD. SLA-anatase discs were prepared for cross-sectional backscattered SEM (BSE) analysis to evaluate film–substrate integrity. Samples were first ground laterally using P1200 SiC abrasive paper (Buehler, Lake Bluff, IL, USA) until half of the circular diameter was removed. The grinding direction was consistently from the top surface toward the bottom surface. Subsequent fine polishing was performed with P2500 and P4000 papers (Buehler, Lake Bluff, IL, USA), followed by final polishing using a cloth and a 3 µm diamond suspension (Buehler, Lake Bluff, IL, USA). The discs were imaged using a backscattered electron detector at an acceleration voltage of 15 kV with a viewing angle of 20° relative to the top surface. The thickness of the anatase film was determined by fracturing a disc to expose its cross-section. This involved creating a 1.5 mm deep groove across the backside of an SLA-anatase disc using a diamond cutting wheel (Hager & meisinger, Neuss, Germany). One side of the cut plate was then clamped in a vice and cooled to −196 °C by submerging it in liquid nitrogen to increase its brittleness. Subsequently, the top half was struck with a hammer to cleanly fracture the plate into two halves. The resulting halves were placed on a sample holder and analyzed by SEM. The anatase layer thickness was quantified using SmartSEM software (Version 6.00, Carl Zeiss Microscopy GmbH, Oberkochen, Germany), with measurements taken at 12 distinct points along the layer’s height. Results are reported as mean ± SD.

2.5. In Vitro Apatite Formation in SBF

To evaluate the in vitro apatite-forming ability, simulated body fluid (1× SBF) and SBF with a doubled concentration of Ca²⁺ and HPO₄²⁻ (2× SBF) were used. The 1× SBF was prepared strictly in accordance with ISO 23317 [33], whereas for the 2× SBF, di-ammonium hydrogen phosphate and calcium nitrate tetrahydrate were added after the standard Ca²⁺ and HPO₄²⁻ sources, based on Nguyen et al. [38]. The rationale for the 2× SBF approach was to sensitize the test for titanium modifications whose apatite formation cannot be distinguished with the standard test. The pH of both SBF solutions was adjusted to 6.8 to prevent precipitation. All chemicals used were pro analysis grade and of the highest purity, sourced from Carl Roth (Karlsruhe, Germany). The samples were incubated in SBF solution at 37 °C for 7 days (1× SBF) and 5 days (2× SBF). Following incubation, the specimens were rinsed with ultrapure water and dried in a desiccator containing silica gel for at least 24 h. The samples incubated in 1× SBS were examined using SEM to assess any evidence of apatite formation. Following incubation in 2× SBF, the samples were examined using an SEM (LEO 1430, Zeiss, Oberkochen, Germany) equipped with an energy-dispersive X-ray (EDX) spectrometer (EDR288/SPU2, Röntec, Berlin, Germany) to analyze crystal formation and the chemical composition of the surfaces. To confirm apatite-like layer formation, XRD analysis was performed as described in Section 2.4.

2.6. Cell Culture

The human osteosarcoma-derived cell line SAOS-2, obtained from DSMZ (Deutsche Sammlung von Mikroorganismen und Zellkulturen), was cultured at a density of 1.8 to 2.0 × 10⁶ cells per 12 mL of McCoy’s-5A medium (Sigma Aldrich, Saint Louis, MO, USA) in a T75 cell culture flask. The medium was supplemented with 15% heat-inactivated fetal bovine serum (BIO&SELL, Feucht/Nuremberg, Germany), 1% GlutaMAX-1 (Gibco, Paisley, UK), and 1% PenStrep (Gibco, New York, NY, USA). SAOS-2 cells at 90–95% confluency were cultured at a density of 20,000 cells/cm² in 1 mL of complete medium on various discs and a tissue culture coverslip. A density of 10,000 cells/cm² was used to evaluate cell adhesion. For the experiments with HOB cells (PromoCell, Heidelberg, Germany), 490,000 HOB cells were cultured for 3 days prior to the experiment in 12 mL of growth medium (basal medium supplemented with supplement mix, PromoCell, Heidelberg, Germany). Cells with a confluency of over 95% were detached and subsequently seeded at a density of 12,600 cells/cm² onto disc samples in 1 mL of growth medium. The seeded discs were then incubated at 37 °C in a humidified atmosphere with 5% CO₂.

2.7. Cell Adhesion

To evaluate the early cell adhesion, the cells were incubated with the discs for 2 h at a density of 10,000 cells/cm². Subsequently, the cells were washed with phosphate-buffered saline (PBS; Gibco, Grand Island, NY, USA) and fixed in 3% paraformaldehyde (PFA, Sigma Aldrich, Steinheim, Germany) for 20 min. Following washing with PBS, the cells were stained with 1 μg/mL 4,6-diamidino-2-phenylindole (DAPI, Sigma Aldrich, Saint Louis, MO, USA) for 10 min and rinsed again with PBS. Imaging was performed using a Nikon Optiphot-2 fluorescence microscope (Nikon, Tokyo, Japan) equipped with a digital camera (Canon, Tokyo, Japan). Cell adhesion was assessed in five independent experiments. For each sample, six randomly selected fields of view were imaged using a 10× objective, and nuclei were counted using ImageJ (version 1.53a).

2.8. Live/Dead Viability Assay

The viability of SAOS-2 cells on different surfaces was evaluated through live/dead staining. After 72 h of cultivation on various surfaces, the cells were stained with a LIVE/DEAD BacLight Bacterial Viability Kit (Molecular Probes, Eugene, OR, USA) following the manufacturer’s instructions. The live/dead viability assay was performed using three independent biological replicates. Five images per sample were captured with a 10× objective using a fluorescence microscope. Quantitative viability was computed from image-integrated fluorescence intensities as: Cell viability (%) = Green/(Green + Red) × 100, where Green and Red denote the integrated intensities of the SYTO^® 9 (live) and propidium iodide (dead) channels, respectively (ImageJ v1.53a).

2.9. Relative Metabolic Activity

The metabolic activity of SAOS-2 and HOB cells cultured on different surface samples was assessed using a Cell Counting Kit-8 (CCK-8, Dojindo Laboratories, Japan). Following 24, 48, and 72 h of cultivation, the cells were rinsed with PBS and incubated with CCK-8 working solution (10% v/v in culture medium) for 1 h at 37 °C. The optical density (OD) of the medium at 450 nm, which correlates with the number of viable cells in culture, was measured using a microplate reader (Infinite f50, TECAN Austria, Salzburg, Austria). Experiments were performed using five independent replicates for SAOS-2 cells and six independent replicates for primary HOBs at each time point.

2.10. Cell Morphology (CLSM and SEM)

The morphology of osteoblasts on various surfaces was examined by confocal laser scanning microscopy (CLSM). After 24 h and 48 h of culturing, SAOS-2 cells were permeabilized with 0.2% Triton X-100 (Merck, Darmstadt, Germany) for 10 min and washed with PBS. Non-specific binding sites were then blocked with PBS containing 1% bovine serum albumin (VWR International, Leuven, Belgium) and 0.05% Tween-20 (Roth, Karlsruhe, Germany). The cells were subsequently stained with a 5 µg/mL FITC-Phalloidin conjugate solution (Sigma Aldrich, Saint Louis, MO, USA) for 1 h. Following three washes with PBS, the nuclei were counterstained with 10 µM DRAQ5 (Miltenyi Biotec, Bergisch Gladbach, Germany). Images were captured using a Leica TCS SP5 confocal microscope (Leica Microsystems, Mannheim, Germany) equipped with an argon laser (488 nm), a DPSS laser (561 nm), and a HeNe laser (633 nm). Imaging was performed in five independent experiments at 24 h and in three independent experiments at 48 h. Cell spreading was quantified by measuring the cell area across three independent experiments at each time point using ImageJ (v1.53a). Cellular morphology was further evaluated through SEM analysis. Following a 24 h incubation on the test surfaces, cells were fixed in a 2% (v/v) glutaraldehyde solution (Carl Roth, Karlsruhe, Germany) overnight. Subsequently, the samples underwent dehydration in ascending ethanol concentrations, followed by critical point drying. Thereafter, the samples were sputter-coated with Au-Pd and analyzed using SEM. SEM analysis was performed in three independent experiments. The images were randomly taken in three different areas of each sample disc and at 5 kV, with a working distance ranging from 9 to 10 mm. The SEM images were pseudo-colored using MountainsLab Premium 10 to improve the visual contrast of morphological features.

2.11. Osteoblasts Differentiation

At day 3 after seeding, when the SAOS-2 cells reached >95% confluency, they were treated with 1 mL of osteogenesis induction medium to facilitate differentiation of the cells. The induction medium was freshly prepared each time by supplementing McCoy’s complete medium with 100 µM of L-ascorbic acid 2-phosphate (Sigma Aldrich, Saint Louis, MO, USA), 10 mM of β-glycerophosphate (Sigma Aldrich, Saint Louis, MO, USA), and 4 µM of dexamethasone (Sigma Aldrich, Saint Louis, MO, USA). The medium was refreshed every 2 days, and the experiment concluded on day 21 post-differentiation. To induce differentiation of HOB cells, starting on day 3, at confluence, 1 mL of osteoblast mineralization medium (PromoCell, Heidelberg, Germany) was applied and refreshed every 2 days for 21 days.

2.12. Alkaline Phosphatase Activity

The ALP activity of SAOS-2 cells was assessed at days 3, 10, and 21 post-mineralization. Following a 15 min incubation of the cells in 400 µL of p-nitrophenyl phosphate (PNPP) solution (Thermo Fisher Scientific, Rockford, IL, USA), the reaction was stopped by the addition of 200 µL of 2 M NaOH (Sigma Aldrich, Saint Louis, MO, USA). The solution was diluted 1:1 with Ampuwa water (Fresenius Kabi, Bad Homburg, Germany), and the absorbance was measured photometrically at 405 nm. Five independent replicates were performed for the ALP assay at each time point.

2.13. Extracellular Matrix Calcium Deposition

Calcium deposition, serving as a late indicator of osseointegration, was evaluated at the 10th and 21st day of mineralization. After washing with PBS and drying for 30 min, the cells were incubated with 400 µL of 0.5 M HCl (Sigma Aldrich, Saint Louis, MO, USA) for 24 h at 4 °C. Quantification of the calcium content for samples and standards was conducted using a Calcium (Ca) Assay Kit (Novus Biologicals, Centennial, CO, USA), according to the manufacturer’s guidelines for tissue homogenate samples. The optical density (OD) of the reactions was measured photometrically at 620 nm, and the calcium concentration was calculated via interpolation using a standard curve. Calcium deposition by HOB cells was assessed on day 21. Calcium deposition was assessed using five independent replicates for SAOS-2 cells and three independent replicates for primary HOBs at each time point.

2.14. Gene Expression Analysis

At 3 and 10 days after mineralization, SAOS-2 cells were assessed for the expression of early and mid to late osteogenic differentiation markers. Total RNA was extracted using a RNeasy Micro Kit (Qiagen, Hilden, Germany) following the manufacturer’s protocol. Subsequently, 350 ng of total RNA was retrotranscribed using LunaScript RT SuperMix (New England Biolabs, Ipswich, MA, USA) to synthesize cDNA. qRT-PCR was performed using a Luna Universal qPCR Master Mix (New England Biolabs, Ipswich, MA, USA) on a QuantStudio 3 system (Applied Biosystems, Thermo Fisher Scientific, Waltham, MA, USA) using the following conditions: 95 °C for 3 min, followed by 40 cycles of 95 °C for 15 s and 55 °C for 30 s. Relative gene expression was determined by the 2^−ΔΔCt method, normalized to GAPDH. Gene expression analyses were performed using four independent biological replicates. The primer sequences are provided in Table 1. All primers were synthetized by standard solid-phase DNA synthesis and purified by HPLC (Ella Biotech GmbH, Amtsgericht Muenchen, Germany).

Table 1. Primer sequences used for gene expression analysis. BGN biglycan, DCN decorin, GAPDH glyceraldehyde 3-phosphate dehydrogenase, OSX osterix, RUNX2 runt-related transcription factor 2, SPARC osteonectin, ALP alkaline phosphatase, and OCN (BGLAP) osteocalcin.

	Forward Primer	Reverse Primer
GAPDH	GAGTCAACGGATTTGGTCGT	TTGATTTTGGAGGGATCTCG
ALP	GCTGTAAGGACATCGCCTACCA	CCTGGCTTTCTCGTCACTCTCA
SPARC	GTACATCGCCCTGGATGAGT	CGAAGGGGAGGGTTAAAGAG
BGN	CAGGAGTGAGTAGCTGCTTTCG	ATGGCGGATGGACCTGGAGA
COL1A1	AACCGTGGTAGGTTTGGTGA	CGAGCACCTTTACTACCACG
DCN	GGAGACTTTAAGAACCTGAAGAACC	CGTTCCAACTTCACCAAAGG
OSX	AAGTATGTATTGAGTGCCTAC-TCTG	TGGATGCACCACCAGATCA
RUNX2	TTACTTACACCCCGCCAGTC	CACTCTGGCTTTGGGAAGAG
OCN	GGCAGCGAGGTAGTGAAGA	AGCAGAGCGACACCCTAGA

2.15. Statistical Analysis

Statistical analysis was performed using GraphPad Prism v9.4.1. Statistical significance is indicated as follows: * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001. Non-significant comparisons are not annotated. All assays were repeated in independent experiments to confirm reproducibility (n = 3–6 per endpoint; the exact n for each assay is reported in the corresponding Methods subsection and in the figure legends). For each independent experiment, specimens from all surface groups were processed in parallel. The experimental unit was one disc per surface condition per experiment. For normally distributed data, differences were evaluated by one-way ANOVA followed by Tukey’s multiple comparison test, while nonparametric data were analyzed with the Friedman test followed by Dunn’s multiple comparison test.

3. Results

3.1. Characterization of the Modified Titanium Surfaces

3.1.1. Analysis of Surface Microroughness

Optical profilometry was used to quantify the arithmetical mean height (Sa) of the investigated surfaces. The average Sa values for the Ti-machined, Ti-SLA, and SLA-anatase surfaces were 0.078 ± 0.006 µm, 1.55 ± 0.06 µm, and 1.88 ± 0.43 µm, respectively (Figure 1A). Sa was significantly higher for Ti-SLA and SLA-anatase compared with Ti-machined, with SLA-anatase exceeded Ti-SLA.

Figure 1. Physico-chemical characterization of the different surfaces. (A) Surface roughness (Sa) of sample discs determined by optical profilometry. (B) Quantitative results of static contact angle measurements of various surfaces. The grid line indicates the 90° threshold separating hydrophilicity from hydrophobicity. Surface free energy analysis of evaluated surfaces using Owens–Wendt (C) and Lifshitz–van der Waals/acid base (D) methods. The SFE was calculated by applying contact angle measurements to the OWRK and LW-AB equations. (E) Dispersive and polar component calculation using the Owens–Wendt method. (F) Comparison of acid and base components calculated by Lifshitz–van der Waals/acid between different surfaces. Each value represents the mean ± SD (n = 5). Statistical significance is indicated as follows: ** p < 0.01; **** p < 0.0001.

3.1.2. Wettability and Surface Free Energy

Wettability was assessed by static water contact angle analysis (Figure 1B). The mean contact angles were 88.9 ± 5.9° for Ti-machined, 114.6 ± 0.8° for Ti-SLA, and 3.8 ± 5.4° for SLA-anatase. Thus, Ti-SLA exhibited higher hydrophobicity than Ti-machined, whereas anatase coating reduced the contact angle to near zero, confirming complete wettability. Surface free energy (SFE) values were calculated using the Owens–Wendt (Figure 1C) and Lifshitz–van der Waals/acid–base (LW/AB) methods (Figure 1D). The Ti-SLA surface demonstrated low surface energy, indicating hydrophobic behavior, whereas the SLA-anatase surface exhibited high surface energy, with a significantly higher SFE than the Ti-machined and Ti-SLA samples, explaining its superhydrophilic properties. Table S2 presents the measured contact angles for the test liquids. The SFE analysis further provides insight into the Lewis acid-base (γ^AB or polar) and Lifshitz–van der Waals (γ^LW or dispersive) components of SFE. Component analysis (Figure 1E,F) indicated a dispersive-dominated signature on Ti-SLA, whereas SLA-anatase was polar-enriched, with a predominance of the base (electron-donor) fraction. The SFE data, analyzed according to the Owens–Wendt and Lifshitz–van der Waals/acid-base methods, are summarized in Tables S3 and S4, respectively.

3.1.3. Micro-Nanotopography Analysis of Anatase-Coated Surface

The surface topography was analyzed using SEM, which provided detailed insights into the topographical characteristics of various surfaces (Figure 2A). The Ti-machined surface exhibited distinct grinding marks. Ti-SLA surfaces displayed rough microtopography with irregular ridges and pits characteristic of sandblasted and acid-etched treatments [9,27]. SLA-anatase displayed a uniform multi-globular microstructure (5.6 ± 0.8 µm diameter) over the microrough substrate. This structure likely resulted from anatase crystalline growth within the pits and rims of the micro-rough Ti-SLA surface. Higher-magnification Fe-SEM (Figure 2B) and HIM (Figure 2C) revealed that the micrometer-scale globular structures were decorated by evenly distributed, polygonal nanoscale surface features with distinct triangular facets. Quantitative ImageJ (v1.53a) analysis of the triangular features (edge length; n = 50) yielded an average size of 51 ± 17 nm, confirming that the anatase coating presents a nanoscale surface feature.

Figure 2. Analysis of surface topography and nanostructures. (A) Representative SEM micrographs of Ti-machined, Ti-SLA, and SLA-anatase surfaces (scale bar = 10 µm). (B) High-magnification FE-SEM images of Ti-SLA and SLA-anatase surfaces showing the anatase coating with nanoscale features on SLA-anatase surface (scale bars = 100 nm). (C) HIM analysis of the SLA-anatase surface at various magnifications reveals triangular nanoscale structures, discernible as distinct faceted faces on the surface (scale bars = 5 µm, 1 µm, 200 nm and 100 nm for different magnifications).

3.1.4. Characterization of TiO₂ Crystalline Phase, Integrity, and Thickness of Anatase Film

The crystallinity of the SLA-anatase surface was analyzed using X-ray diffraction and compared to Ti-machined and Ti-SLA samples. As shown in Figure 3A, the SLA-anatase pattern exhibited a dominant anatase diffraction peak at 2θ = 37.8° (A3) and a weaker peak at 25.2° (A1) (ICSD-142916). Although a faint reflection at 27.4° (R1), characteristic of rutile TiO₂ (ICSD-142923), was observed on the SLA-anatase surface, diffraction analysis confirmed that the coating was predominantly composed of anatase. In contrast, only titanium-related peaks (ICSD-43416) were detected for the Ti-machined and Ti-SLA samples. A single weak reflection at ≈ 41° (R4), assigned to rutile TiO₂, indicated a trace amount of rutile on the Ti-SLA surface. Cross-sectional BSE-SEM imaging of SLA-anatase surfaces demonstrated a uniform and continuous coating, with no evidence of delamination or discontinuities at the coating–substrate interface (Figure 3B). Finally, cross-sectional SEM imaging of fractured SLA-anatase discs showed an anatase layer thickness of 8.4 ± 0.4 µm (Figure 3C).

Figure 3. Characterization of anatase-TiO₂ film. (A) X-ray diffraction patterns were analyzed for SLA-anatase, Ti-machined, and Ti-SLA surfaces. The labeled peaks T, A, and R in the XRD plot correspond to titanium, anatase, and rutile, respectively. The primary anatase peak (A3) was identified at a 2θ angle of approximately 37.8°, exclusive to the SLA-anatase samples. Additional low-intensity anatase (A1) and faint rutile (R1) peaks were observed at 2θ angles of 25.2° and 27.4°, respectively, on the SLA-anatase surface. (B) Cross-sectional SEM image of the SLA-anatase surface acquired in backscattered mode (15 kV, 70° tilt). The anatase TiO₂ layer is continuous and uniformly adhered to the Ti-SLA substrate without detectable delamination or interfacial gaps. (C) The thickness of the anatase layer was measured at 12 distinct sites using SmartSEM (v.6.00) software, and the results are reported as mean ± SD.

3.2. Anatase Coating Impact on Hydroxyapatite Formation

As a preliminary assessment of bioactivity, an in vitro apatite formation assay was conducted by immersing titanium samples in 1× and 2× SBF. After 7 days in 1× SBF, no morphological changes were observed in SEM analysis across all groups (Figure S1). However, 2× SBF induced the formation of hydroxyapatite (HA)-like surface structures on the SLA-anatase samples, while the Ti-machined and Ti-SLA groups showed no evidence of apatite formation after 5 days (Figure 4A). Energy-dispersive X-ray (EDX) analysis confirmed the presence of HA through the detection of calcium (Ca) and phosphorus (P) in the SLA-anatase samples (Figure 4D). In contrast, no detectable levels of Ca and P were observed in the Ti-machined (Figure 4B) and Ti-SLA groups (Figure 4C) under identical conditions. X-ray diffraction (XRD) analysis provided phase-level evidence, revealing the emergence of distinct peaks at approximately 22.8°, 25.8°, and 31–33° (2θ) on SLA-anatase samples after 5 days of immersion in 2× SBF. These peaks are consistent with the formation of hydroxyapatite (HA; ICSD-180315) (Figure 4E). High-magnification SEM imaging of SLA-anatase after immersion in 2× SBF revealed a uniform layer of plate- or foil-like crystallites consistent with hydroxyapatite formation reported in conventional SBF tests [31,39] (Figure 4F). Together, the SEM, EDX, and XRD data indicate that SLA-anatase promotes apatite formation and growth in 2× SBF within 5 days, whereas the Ti-machined and Ti-SLA groups do not exhibit similar behavior within the same timeframe.

Figure 4. SBF-induced apatite formation. (A) Representative SEM images of Ti-machined, Ti-SLA, and SLA-anatase after immersion in 2× SBF at 37 °C for 5 days. (B–D) Corresponding EDX spectra for Ti-machined (B), Ti-SLA (C), and SLA-anatase (D); Ca and P peaks are detected only on SLA-anatase. Spectrum colors indicate the surface group (Ti-machined = gray, Ti-SLA = blue, SLA-anatase = red) and are used only for visual distinction. (E) XRD pattern of SLA-anatase after immersion in 2× SBF (37 °C, 5 days); new reflections consistent with hydroxyapatite (HA; ICSD-180315) appear near 2θ ≈ 22.8°, 25.8°, and 31–33° following immersion. Labels A, T, and HA indicate anatase, titanium, and hydroxyapatite, respectively. (F) Higher-magnification SEM reveals a uniform layer of plate/foil-like crystallites on SLA-anatase after immersion in 2× SBF.

3.3. Osteoblast Responses on Micro–Nano Structured SLA-Anatase Surfaces

3.3.1. SAOS-2 Cell Early Adhesion

Initial cell adhesion behavior was visualized through DAPI staining of nuclei. As illustrated in Figure 5A, after a 2 h incubation period, cells exhibited a uniform distribution across all titanium surfaces. The attached cell counts on Ti-machined, Ti-SLA, and SLA-anatase samples followed a consistent increasing trend. Quantification of cell numbers using ImageJ (v1.53a) further corroborated these observations. Figure 5B shows that the number of attached cells on all titanium surfaces was higher than that on the coverslip control, with statistically significant differences observed for both Ti-SLA and SLA-anatase surfaces. The highest level of cell attachment was detected on SLA-anatase, indicating that the anatase coating did not compromise the cell adhesion properties of Ti-SLA, but rather showed a tendency to further increase osteoblast attachment.

Figure 5. Early Adhesion and viability of SAOS-2 cells on different titanium surfaces. (A) Representative fluorescence images showing SAOS-2 cell attachment after 2 h of incubation. (B) Quantification of attached cells after 2 h. Nuclei were stained with DAPI, and six random images were captured per surface for counting using ImageJ software (v1.53a) (n = 5; mean ± SEM). Statistical significance is indicated as ** p < 0.01. (C) Live/dead fluorescence images of SAOS-2 cells cultured on the analyzed surfaces for 72 h. Live and dead cells are shown in green and red, respectively. (D) Quantification of cell viability expressed as the ratio of green/(green + red) fluorescence intensity (n = 3; mean ± SEM).

3.3.2. Viability of SAOS-2 Cells on Different Surfaces

SAOS-2 cell viability on different titanium surfaces was evaluated after 72 h using a live/dead staining assay. All groups exhibited predominantly green fluorescence, indicating high cell viability, with only minimal red fluorescence observed on the Ti-SLA surface (Figure 5C). The Ti-SLA group showed a slight decrease in the intensity and density of viable cells, with a more scattered distribution pattern. In contrast, cells on SLA-anatase appeared densely distributed, comparable to those on the control coverslip and Ti-machined surfaces. Quantitative analysis of cell viability supported these observations, revealing a minor reduction on Ti-SLA and near-complete recovery on SLA-anatase (Figure 5D). However, the differences among groups were not statistically significant.

3.3.3. Metabolic Activity of SAOS-2 Osteoblasts on the Investigated Implant Surfaces

Cell metabolic activity was quantified by CCK-8 assay, which measures cellular dehydrogenase activity and serves as an indirect index of viable cell number. After 24 h, osteoblast metabolic activity on the SLA-anatase surface exhibited a significant increase compared to the coverslip control group and the Ti-SLA surface. This effect became more pronounced after 48 h of incubation, with the SLA-anatase group exhibiting significantly higher metabolic activity than all other groups. At 72 h, cells continued to exhibit the highest metabolic activity on the SLA-anatase surface. In contrast, the Ti-SLA surface demonstrated the lowest metabolic activity across all observed time points (Figure 6A). The metabolic activity of cells over time revealed distinct trends across the surfaces. While the Ti-SLA surface exhibited a consistent decline in metabolic activity over the cultivation period, the SLA-anatase samples displayed a steady increase, eventually exceeding the performance of the standard coverslip samples (Figure 6B). The metabolic activity of the cells was also evaluated at 3, 10, and 21 days post-mineralization. SLA-anatase samples demonstrated the highest levels of metabolic activity compared to other titanium surfaces, aligning closely with the control coverslip group over the extended cultivation period from day 3 to day 21 (Figure 6C). The assessment of growth progression further corroborated the fluctuations in cell viability at days 10 and 21, with SLA-anatase demonstrating the highest metabolic activity at all time points (Figure 6D).

Figure 6. Analysis of osteoblast metabolic activity following incubation on different sample surfaces. (A) Quantitative osteoblast metabolic activity at 24, 48, and 72 h, as measured by CCK-8 assay. The coverslip group at each time point served as the control (set to 100%), and results for all other surfaces are expressed relative to that value. (B) Metabolic activity of SAOS-2 cells over time, normalized to the coverslip group at 24 h (set to 100%). (C) SAOS-2 cell metabolic activity at 3, 10, and 21 days after induction of differentiation. (D) Post-mineralization growth progression evaluated relative to the coverslip surface at day 3. n = 5; mean ± SEM. Statistical significance is indicated as follows: * p < 0.05; ** p < 0.01; *** p < 0.001.

3.3.4. Impact of Surface Modification on SAOS-2 Cell Morphology

The morphological features of osteoblasts were evaluated using CLSM at 24 h and 48 h post-seeding by fluorescence staining of actin filaments. As evidenced by fluorescence imaging, cellular expansion exhibited a continuous increase over the 24 h and 48 h periods following cultivation on all assessed surfaces. After 48 h, most cells exhibited a well-spread morphology across the examined surfaces. However, osteoblasts cultured on Ti-SLA exhibited distinctive behavior, appearing smaller and less spread out compared with those on the other surfaces (Figure 7A). Cell spreading was also further quantified by measuring the cell area. The quantification results revealed that cell spreading on coverslip and Ti-machined was almost identical and higher than on the other modified variants. Cell area was significantly higher on both coverslip and Ti-machined than on Ti-SLA after 48 h. In addition, comparison of cell area between 24 h and 48 h for each surface unveiled the highest difference (with the lowest p-value) for SLA-anatase, demonstrating the highest spreading rate during the time (Figure 7B).

Figure 7. Morphological features of osteoblasts on the evaluated surfaces. (A) SAOS-2 cells were cultured on all test surfaces and observed at 24 h (n = 5) and 48 h (n = 3) post-seeding. The cytoskeleton (actin) was stained with FITC-phalloidin (green), and nuclei were stained with DRAQ5 (purple). Confocal images were acquired using a 25× objective on a CLSM system with a 2.5× digital scan zoom. (B) Cell spreading was quantified by measuring the cell area using ImageJ software (v1.53a). The coverslip group after 48 h of incubation, serving as the control, was set to 100%. The numerical differences indicate statistical differences between two time points for each surface calculated by paired t-test; n = 3. Mean ± SEM. Statistical significance is indicated as ** p < 0.01. (C) Morphological differences in osteoblasts, as determined by SEM analysis. A total of 10,000 SAOS-2 cells/cm² was plated onto the different surfaces and analyzed 24 h after seeding. Images with a white scale bar have a magnification of 1000× and a scale bar of 30 µm. Images with a yellow scale bar have a magnification of 5000× and a scale bar of 5 µm (n = 3). The colored areas in the SEM inserts are false-color overlays used only to highlight the cell outlines/locations on each surface.

The morphology of the cells was analyzed in detail by SEM 24 h post-cultivation (Figure 7C). SEM analysis indicated that cells on Ti-machined implants displayed a spreading pattern comparable to that observed on coverslip, representing the most extensive spread among the implant types. Conversely, cells on the Ti-SLA surface exhibited reduced extension and varied shapes compared to those on the Ti-machined surface. These cells appeared small and confined within the microrough surface’s cavities. SAOS-2 cells cultured on the SLA-anatase surface demonstrated enhanced spreading relative to those on the Ti-SLA surface. At higher magnifications (5000×), cells on SLA-anatase displayed prominent filopodia extensions, akin to those observed on control surfaces. In contrast, such structures were not distinctly visible on Ti-SLA surfaces.

3.3.5. Differentiation of SAOS-2 Cells: ALP Activity

ALP, a marker indicative of early to midterm osteoblast differentiation, was assessed at 3, 10, and 21 days following mineralization (Figure 8A). ALP activity levels exhibited a consistent pattern across all time intervals. In both the Ti-machined and Ti-SLA groups, ALP activity was observed to be lower compared to the SLA-anatase group. By day 10, ALP activity in the SLA-anatase group showed a significant increase relative to the other titanium surfaces. However, the lowest ALP activity was consistently found in the Ti-SLA variant at all time points. The SLA-anatase surface exhibited an ALP activity profile most closely resembling that of the coverslip control across all assessed time points.

Figure 8. In vitro evaluation of phenotypic differentiation of SAOS-2 osteoblasts. (A) ALP activity of SAOS-2 cells at day 3, day 10, and day 21 of differentiation (n = 5). (B) Extracellular calcium deposition was analyzed at days 10 and 21 of differentiation (n = 5). Relative gene expression of early markers at day 3 (C) and mid and late osseointegration markers at day 10 (D) was analyzed relative to the GAPDH housekeeping gene. The relative mRNA expression was quantified by the 2^−ΔΔCt method. The data were normalized to gene expression on the reference coverslip surface, n = 4. Mean ± S.E.M. Statistical significance is indicated as follows: * p < 0.05; ** p < 0.01; **** p < 0.0001.

3.3.6. Differentiation of SAOS-2 Cells: Extracellular Matrix Mineralization

Cell-associated mineral deposition, a late-stage marker of osteogenic differentiation, was assessed on days 10 and 21 (Figure 8B). A marginal increase in calcium deposition was observed across all surfaces as the cultivation period progressed to day 10. By day 21, mineralization had increased on all surfaces, with the SLA-anatase and Ti-SLA surfaces exhibiting the highest and lowest calcium deposition, respectively. However, calcium deposition results at the evaluated time points did not reveal significant differences among the materials assessed. Notably, calcium deposition on the Ti-SLA surface was lower than that on the smooth Ti-machined surface. In contrast, the anatase coating reversed this trend, enhancing mineralization on the SLA-anatase surface beyond that observed on the Ti-machined surface and coverslip control.

3.3.7. Differentiation of SAOS-2 Cells: Osseointegration-Related Gene Expression

The phenotypic maturation of SAOS-2 osteoblasts was evaluated by quantifying mRNA expression using qRT-PCR. Early osteogenic genes associated with osteogenic maturation, including RUNX2, OSX, and ALP were analyzed on day 3 of differentiation. All three genes were upregulated in the SLA-anatase group compared to all other groups, although these differences were not statistically significant (Figure 8C).

Figure 8D shows the expression of mid- and late-stage osteogenic markers. On day 10, expression levels of all evaluated genes except OSX and SPARC were higher in SLA-anatase than in Ti-SLA; however, this increase reached statistical significance only for ALP (versus Ti-SLA) and RUNX2 (versus coverslip). For OSX and SPARC, both Ti-SLA and SLA-anatase displayed comparable and significantly higher expression than Ti-machined and the coverslip control. OCN expression in SLA-anatase was comparable to the Ti-machined and coverslip groups and higher than Ti-SLA, although the difference was not significant. Though, when Ti-SLA and SLA-anatase samples were compared by paired t-test, a significant increase in OCN expression was observed on the SLA-anatase surface (n = 4; mean ± SEM; p = 0.04). Finally, DCN expression showed a decreasing trend in Ti-machined and Ti-SLA compared with the coverslip, but returned to control-comparable levels in the SLA-anatase group.

3.3.8. Effects of the Anatase Coating on HOB Cell Metabolic Activity and Differentiation

To further validate the biological outcomes, primary HOB cells were assessed for metabolic activity and differentiation on the tested surfaces. HOB cells cultured on SLA-anatase exhibited the most favorable biological response. CCK-8 assay showed that SLA modification of the Ti-machined surface reduced HOB metabolic activity, whereas the subsequent anatase coating reversed this effect (Figure 9A). SLA-anatase samples consistently demonstrated higher metabolic activity than Ti-SLA, with a statistically significant difference at 48 h. Although the increase in metabolic activity at 72 h was not significant in the overall analysis, paired t-test comparison between Ti-SLA and SLA-anatase revealed a significant increase in metabolic activity on SLA-anatase (n = 6; mean ± SEM; p = 0.04). On SLA-anatase, metabolic activity trended upward over 24–72 h but stayed slightly below Ti-machined. Metabolic activity on Ti-SLA decreased then plateaued and was lowest overall (Figure 9B). HOB differentiation was evaluated on day 21 using a calcium-deposition assay (Figure 9C). The results indicated enhanced extracellular matrix mineralization on the modified surfaces, with a significant increase on SLA-anatase compared with Ti-machined and Ti-SLA. Although calcium concentrations were higher on Ti-SLA than on Ti-machined, this difference was not statistically significant.

Figure 9. Analysis of HOB cell metabolic activity and differentiation. (A) HOB cell metabolic activity after 24, 48, and 72 h post-seeding on different titanium surfaces, as measured by CCK-8 assay. The Ti-machined group served as the control and was set to 100%. (B) Metabolic activity trends over time expressed relative to Ti-machined at 24 h (set to 100%). n = 6; mean ± SEM. (C) Extracellular calcium deposition measured using a calcium assay kit on day 21 of differentiation. n = 3; mean ± SEM. Statistical significance is indicated as follows: * p < 0.05; ** p < 0.01; *** p < 0.001.

4. Discussion

Nanoscale surface coatings such as nanocrystalline anatase TiO₂ are widely pursued to strengthen osteogenic responses; however, reproducible application of these coatings onto clinically relevant microrough substrates and the mechanisms linking surface properties to biological responses remain inadequately understood. Here, we coated a clinically relevant SLA implant surface with anatase film and characterized surface crystallinity, topography, and chemistry, together with in vitro hydroxyapatite formation and osteoblast responses to assess in vitro bioactivity. Across assays, the anatase-modified surface exhibited exclusive HA formation and enhanced osteoblast function, supporting our hypothesis and motivating the mechanistic interpretation below.

Reactive pulsed magnetron sputtering was used to deposit an anatase film on Ti-SLA. The resulting SLA-anatase surface exhibited a hierarchical micro–nano topography, characterized by multi-globular microstructures, overlaid with faceted triangular nanoscale features. We previously demonstrated that optimizing pulsed-DC parameters allows controlled deposition of phase-pure anatase TiO₂ on various substrates [27,40]. On smooth titanium and quartz, this deposition method yielded columnar, faceted anatase nanocrystals predominantly oriented perpendicular to the surface [40]. In the present SLA context, the faceted nanoscale surface features show a range of apparent orientations in FE-SEM/HIM, which is consistent with conformal coverage over pore walls and ridge regions of the SLA microtexture. Here, “apparent orientation” refers to the geometric alignment of faceted features in projection; crystallographic texture was not determined from SEM/HIM. Consistent with this morphology, roughness analysis showed that the anatase coating did not substantially alter the microroughness of the SLA surface, which remained within the moderate roughness range. Cross-sectional BSE-SEM confirmed a continuous ~8 µm anatase layer without visible interfacial gaps or delamination, supporting coating integrity at the imaging scale.

The anatase coating altered not only surface topography, but also wettability, transforming hydrophobic Ti-SLA into a superhydrophilic SLA-anatase surface. This wetting objective parallels hydrophilic SLA variants such as SLActive, where controlled post-processing and wet storage maintain a high-energy, superhydrophilic surface and improve early bone-to-implant contact relative to conventional SLA [16,19]. In this context, our anatase-coating strategy aims to achieve similar early wettability on an SLA microrough substrate while incorporating a nanostructured TiO₂ surface topography, obviating the need for specialized storage conditions unlike SLActive. Additionally, prior reports indicate that anatase-coated titanium is typically rendered superhydrophilic only after post-deposition treatments such as UV photofunctionalization [9,40,41], oxidative and alkali conditioning [42,43,44], or wet storage [9] which remove adventitious hydrocarbons and enrich surface Ti–OH groups [9,40,45], thereby increasing the polar component of SFE and driving hydrophilicity [9,15]. In contrast, the present anatase coating achieved near-zero contact angles on an SLA substrate without requiring post-deposition activation or protected wet storage, offering a complementary approach for achieving early hydrophilicity that may streamline manufacturing and chairside handling. Given the improved early bone-to-implant contact reported for hydrophilic SLA concepts in humans [16], an activation-free route to superhydrophilicity on SLA substrates could be clinically relevant. However, translation will require the demonstration of hydrophilicity stability following sterilization and storage, as well as coating integrity under insertion-relevant mechanical loading, succeeded by in vivo evaluation of early BIC and long-term stability.

Mechanistically, wettability in biomaterials reflects the interplay of topography and SFE [46]. In this regard, capillary effects associated with nanoscale features can accelerate wetting by facilitating rapid water infiltration into nanotextures, thereby decreasing the contact angle [47]. The micro-globular and nano-triangular configurations observed on the SLA-anatase surface are therefore likely to promote such capillary effects, contributing to the high hydrophilicity. In addition, materials with higher SFE exhibit superior wettability due to stronger liquid–solid interactions, reducing the interfacial air, further enhancing wetting [15,29,48]. Consistent with this, Zhang et al. reported that an anatase honeycomb nanotexture increased SFE on smooth titanium, attributing the increase primarily to the anatase layer [49]. The Owens–Wendt and LW/AB methods were used to estimate SFE in our study [15]. However, since these thermodynamic models assume smooth, chemically homogeneous surfaces, the values obtained here should be interpreted as apparent SFE estimates, a widely adopted convention for microrough implant surfaces. Extensive work in implant surface science has shown that contact angles on rough titanium reflect a metastable wetting state influenced by air entrapment and microtexture, yet comparative SFE trends remain meaningful for understanding differences in wettability and early biological responses. On Ti-SLA, SFE was low and predominantly dispersive. However, following anatase deposition, the total SFE increased significantly, with the polar fraction, dominated by the base (electron donor), becoming more prominent. This aligns with the near-zero contact angles observed on SLA-anatase. Our findings collectively support a mechanism in which nanostructured anatase enhances wettability through a combination of increased polar SFE and capillary effects.

In vitro apatite formation in simulated body fluid (ISO 23317) serves as a commonly employed screening assay to evaluate a material’s apatite-forming propensity [25,31,33,34,50,51]. Standard 1× SBF (ISO 23317) was used initially, but after 7 days it did not discriminate between the titanium variants. Therefore, a twofold-concentrated SBF (2× SBF), adapted from Nguyen et al. [38], was employed as an accelerated screening medium to sensitize the assay for titanium modifications whose apatite-forming ability cannot be distinguished with the standard test. The higher Ca²⁺ and HPO₄²⁻ concentrations here are not intended to mimic physiological serum conditions, but to lower the nucleation barrier and reveal relative differences in apatite-forming tendency, consistent with the use of SBF as an in vitro screening tool rather than a model of in vivo bone remodeling. Under these conditions, only SLA-anatase developed a continuous Ca/P-rich, plate-like layer deposition, with HA reflection at ≈22.8°, 25.8°, and 31–33° 2θ (Figure 4), whereas Ti-machined and Ti-SLA remained negative, indicating lower surface reactivity in this assay. Although microscale roughness has been identified as a factor influencing Ca/P deposition, its precise role remains uncertain [52]. With comparable microroughness, ISO 23317 SBF testing showed apatite formation on hydrophilic SLActive but not on hydrophobic SLA [32]. Moreover, on titanium, SBF apatite formation is preferentially observed on hydrophilic, high-SFE surfaces [15,32,53]. This pattern suggests that wettability and the polar SFE component, rather than microroughness alone, favor early interfacial ion capture and nucleation in SBF [53,54]. Consistent with this view, nanostructured coatings that elevate hydrophilicity and polar SFE have been linked to enhanced apatite formation [32,44]. Porous or anodic TiO₂ films have been demonstrated to sustain a surface water layer that facilitates heterogeneous hydroxyapatite (HA) nucleation and subsequent growth [55]. Additionally, the apatite-inducing capability of titanium oxide layers has been linked to their higher surface energy [53]. In the present study, Ti-SLA and SLA anatase exhibited only minor differences in microscale roughness, yet their 2× SBF responses diverged markedly. We interpret the selective HA formation on SLA anatase as being driven primarily by its higher polar surface free energy and resultant superhydrophilicity. Elevated SFE and rapid wetting likely facilitate hydration layer formation and Ca²⁺ accumulation, lower the barrier to heterogeneous nucleation, and support successive Ca/P deposition. Although relatively few studies have deconvolved how specific SFE components regulate Ca²⁺ uptake, convergent evidence indicates that reducing the polar (electron donor/base) fraction suppresses Ca²⁺ adsorption and early calcium phosphate clustering in SBF, whereas higher polar SFE favors ion capture and nucleation [53,54]. In our samples, the anatase coating increased the polar/base contribution to SFE, which aligns with the enhanced Ca²⁺ attraction and HA formation observed in 2× SBF. Taken together, the 2× SBF data reinforce a mechanistic role for polar SFE (particularly its base fraction) and the resultant superhydrophilicity in interfacial mineral nucleation and apatite formation on SLA anatase.

Nanostructured titanium surfaces have frequently been reported to enhance cell adhesion, proliferation, and osteogenic differentiation in vitro and to promote osseointegration in vivo [22,56,57]. At the cell–material interface, the SLA-anatase surface fostered more favorable early interactions than Ti-SLA. Initial cell attachment and cell viability showed a consistent upward tendency on SLA-anatase relative to Ti-SLA, and cells displayed greater spreading between 24 and 48 h on SLA-anatase. SEM and CLSM supported these observations, showing that cells on SLA-anatase exhibited better cell spreading and numerous filopodia that followed the micro–nano topography, while Ti-SLA supported smaller, less spread morphologies. Taken together, these early events align with the view that higher polar surface free energy and superhydrophilicity accelerate interfacial hydration and protein conditioning, which in turn facilitate integrin-mediated attachment on hierarchical micro–nano surfaces [20,29,58]. Improved early behavior on SLA-anatase coincided with higher metabolic activity in SAOS-2 and HOBs from 24 to 72 h, which is consistent with adhesion-supported proliferative activity on hydrophilic, high-SFE micro–nano interfaces [20,29,58]. More specifically for TiO₂/anatase, nano-modified titanium surfaces have been reported to increase protein adsorption and to promote osteogenic proliferation and differentiation, consistent with the direction of our in vitro cell responses on SLA-anatase [20,25]. During osteogenic differentiation, alkaline phosphatase activity was significantly higher on SLA-anatase at day 10, with non-significant upward tendencies at days 3 and 21. Gene expression analysis revealed day-3 trends toward elevated early markers on SLA-anatase, followed by significant increases at day 10 in ALP (compared to Ti-SLA) and RUNX2 (compared to coverslip). Meanwhile, OSX and SPARC levels were similar between Ti-SLA and SLA-anatase but higher than those observed in controls. Extracellular calcium deposition in HOB cells was significantly higher on SLA-anatase at day 21, whereas SAOS-2 showed a non-significant upward trend. These aligned osteogenic outcomes across two cell models support a surface-driven effect.

The relatively lower early osteoblastic response on Ti-SLA is plausibly attributed to its initial hydrophobicity and reduced polar component of surface free energy, factors that hinder early protein conditioning and integrin engagement at the cell-material interface [9,15]. By contrast, the anatase TiO₂ coating produced a high-energy, superhydrophilic interface with nanoscale surface features. Evidence from the literature indicates that nanoscale features and hydrophilicity act together to optimize the adsorbed protein layer and focal adhesion formation, improving cell contact and spreading [15,27,59]. Lateral nanoscale spacing can further enhance adhesion and proliferative activity through increased protein adsorption [25,56,58], and cell adhesion and osteogenic differentiation on titanium are mediated by integrin engagement, with extracellular matrix proteins adsorbed onto the surface [60]. Within this framework, the hierarchical micro–nano cues on SLA-anatase likely stabilize cell adhesion by providing micro-scale anchorage overlaid with nanoscale adhesion sites, while the increased polar (base) fraction of surface free energy and superhydrophilicity accelerate interfacial hydration and protein conditioning [25,59]. Together, these factors provide a coherent, literature-supported explanation for early adhesion and spreading and, in turn, for subsequent proliferation and differentiation on SLA-anatase, while acknowledging that the initial protein conditioning layer and integrin signaling were not measured here.

Taken together, these findings support a model in which (i) increased polar (base) SFE and superhydrophilicity accelerate interfacial hydration and protein conditioning, and (ii) hierarchical micro–nano cues stabilize focal adhesions and sustain osteogenic activity, offering a plausible explanation for the enhanced apatite formation in 2× SBF and the osteogenic cell responses. We emphasis that this is a supported model rather than a proven causal chain, as we did not directly quantify protein adsorption kinetics, integrin/focal-adhesion signaling, or HA nucleation dynamics. Future work using time-resolved protein adsorption, integrin/focal adhesion immunostaining, and transcriptomic or proteomic readouts will be required to test this model more directly.

5. Conclusions

Coating an anatase TiO₂ layer onto a clinically relevant, commercially established SLA implant surface produced a superhydrophilic, high-energy hierarchically micro-nano structured surface, while retaining the underlying microroughness. This surface selectively facilitated homogeneous apatite formation in modified SBF and enhanced early osteoblast functions in vitro, demonstrating consistent improvements in cell metabolic activity and osteogenic differentiation across SAOS-2 and primary HOB cells. These findings suggest that an increased polar (base) fraction of surface free energy and complete wetting, combined with hierarchical micro-nano cues at the cell-material interface, promote interfacial hydration, support hydroxyapatite formation in SBF, and favor early osteogenic responses.

In this study, the anatase-modified SLA surface exhibited a superhydrophilic state without the need for additional post-deposition activation steps in our workflow, suggesting a potentially practical route to introducing nanoscale functionality and increasing polar surface energy on a microrough titanium implant substrate. Within the scope of in vitro testing, these results support the surface concept as a candidate for further evaluation in the early healing context. Priority next steps include determining the stability of wettability over storage time and handling, assessing coating stability under mechanical and chemical challenge, characterizing early protein conditioning at the cell–material interface, and verifying early bone–implant contact and stability in vivo.