Curcumin-loaded alkali-heat-treated titanium modulates macrophage inflammatory responses Authors: Andrew Cedillo, Niloufar Abedi, Sahar Vahabzadeh, Miya Kang Published: 2026 ABSTRACT Chronic inflammation at the implant–tissue interface is a major contributor to implant failure. Therefore, enhancing the biological response of titanium (Ti) surfaces is critical for promoting early osseointegration and improving long-term implant stability. In this study, alkali-heat-treated (AHT) Ti was evaluated as a platform for the controlled delivery of curcumin (Cur), a natural anti-inflammatory compound. The release kinetics of curcumin and the interactions between curcumin-loaded AHT Ti surfaces and RAW264.7 macrophages were investigated in vitro. Various polymer coating strategies were applied to modulate curcumin release. The presence of poly(lactic-co-glycolic acid) (PLGA) reduced the curcumin release under both physiological (pH 7.4) and acidic (pH 5.0) conditions, with uncoated AHT samples exhibiting the greatest cumulative release (∼21.3 μg) in Study 1. Increasing the loaded curcumin amount from 100 to 300 μg resulted in a decrease in the release for both pH 7.4 and pH 5.0, whereas the incorporation of polyethylene glycol/polycaprolactone (PEG/PCL) coating enhanced the gradual curcumin release (Study 2). Increasing the curcumin loading from 50 to 100 and 300 μg increased the total amount of curcumin released, with the 300 μg group exhibiting a more gradual release profile. However, the percentage of curcumin released decreased as the loading amount increased (Study 3). When macrophages were cultured on Ti surfaces, AHT samples exhibited higher cell viability than polished (P) control samples, and the presence of curcumin reduced cell viability in a dose-dependent manner. While macrophages cultured on AHT surfaces exhibited a transient increase in pro-inflammatory marker expression at 24 h compared with polished surfaces, the expression of these markers was reduced by 48 h. Notably, curcumin significantly decreased pro-inflammatory cytokine expression at both the gene and protein levels and suppressed NF-κB activation in macrophages exposed to inflammatory stimuli. Overall, the sustained and controlled release of curcumin from AHT Ti surface is highly dependent on the curcumin loading amount as well as the presence and composition of the polymer coating. Moreover, both AHT and curcumin significantly reduced inflammatory responses in macrophages, demonstrating the potential of the curcumin-loaded AHT Ti platform as an immunomodulatory surface for improving the biological performance and long-term success of Ti implants. Introduction Titanium and titanium-based alloys are widely used in orthopedic and dental implants because of their biocompatibility, corrosion resistance, and mechanical stability,1,2 and the number of primary total hip and knee arthroplasties performed in the United States is projected to more than double by 2040.3 Although titanium implants have a strong record of clinical use, the material itself does not strongly stimulate osteogenesis. As a result, inadequate biological integration at the implant surface may limit direct bonding with the surrounding bone.1,4 Poor early integration may lead to excessive micromotion, fibrous tissue formation, implant instability, and eventual failure.5 Chronic inflammation around prosthetic materials can also contribute to periprosthetic osteolysis and aseptic loosening, which remain major causes of long-term implant failure and revision surgery.6 Improving the biological response at the titanium surface is therefore an important goal for both early implant fixation and long-term stability. Implant healing is influenced by several early events at the bone–implant interface, including protein adsorption, immune and inflammatory signaling, cell attachment, and the eventual transition toward osteogenesis and bone remodeling;2,7 thus, the biological response to titanium depends heavily on its surface properties rather than its bulk composition alone. Surface modification can alter roughness, topography, wettability, surface energy, and chemistry, which in turn affect protein adsorption, cell attachment, proliferation, and differentiation at the bone–implant interface.8,9 Alkali-heat treatment (AHT) is a chemical modification method that has received considerable attention because it is relatively simple and produces a reactive titanate-derived surface layer.10 During treatment, titanium is immersed in concentrated sodium hydroxide (NaOH) or potassium hydroxide (KOH), forming a sodium or potassium titanate layer that assists with the accumulation of calcium and phosphate ions and supports hydroxyapatite nucleation. Subsequent heat treatment stabilizes and crystallizes the modified layer while improving its mechanical integrity.10,11 Increased hydroxyapatite formation and surface hydrophilicity have both been associated with improved bone bonding and osseointegration.10,12 We have shown that AHT induces vascular endothelial growth factor and osteocalcin expression and thus accelerates osteoblast cell maturation and bone remodeling.13 In addition to enhancing bioactivity, the porous and reactive structure produced by may provide a useful foundation for functional coatings and local drug delivery. Modified titanium surfaces can also be designed to deliver therapeutic agents directly at the bone–implant interface. Local delivery may improve antibacterial, anti-inflammatory, or bone-regenerative effects by concentrating a therapeutic compound at the surgical site. This approach may also reduce some of the limitations of systemic administration, including poor targeting, short therapeutic duration, and off-target toxicity.14 Micro- and nanostructured titanium surfaces, formed by processes such as anodization and, are particularly useful for local delivery because their increased surface area can improve drug loading and influence release behavior.14,15 For example, alkali-treated orthopedic screws with hydroxyapatite coatings have been used to release tobramycin locally, producing antibacterial activity while maintaining in vitro cell compatibility.16 Polymer coatings applied to nanostructured titanium have also been used to reduce burst release and extend drug delivery, with release behavior depending on the thickness and composition of the coating.15,17,18 However, additional coating strategies may be needed to retain the therapeutic agent and provide more controlled release over time.14,18 Curcumin is a potential candidate for local implant-based delivery because of its reported antioxidant, anti-inflammatory, antimicrobial, and bone-regenerative properties.19,20 These biological effects may be valuable during early implant healing, when successful osseointegration depends on both osteogenic activity and appropriate regulation of inflammation.21,22 However, curcumin is hydrophobic, poorly soluble in water, and has limited bioavailability. When deposited directly onto a surface without a carrier, it may also be difficult to retain and release in a controlled manner.19,21 Polymer coatings can address some of these limitations by affecting water penetration, drug diffusion, polymer degradation, and overall release kinetics.23,24 Biocompatible polymers such as PLGA, PCL, and PEG are commonly used in drug-delivery systems because their hydrophilicity and degradation behavior can be adjusted to influence release.23–26 Combining curcumin with polymer-coated AHT titanium may therefore allow local therapeutic delivery while preserving the bioactive properties of the modified implant surface. Macrophages are among the first immune cells to respond following biomaterial implantation and play an important role in determining whether the surrounding tissue progresses toward chronic inflammation or healing.27,28 Their response can be affected by adsorbed proteins, surface chemistry, topography, and therapeutic compounds released from the implant.28,29 This is particularly relevant for titanium implants because macrophages contribute to the regulation of inflammation, tissue repair, and bone remodeling throughout osseointegration.27,30 A prolonged pro-inflammatory macrophage response may promote chronic inflammation, osteoclast activation, fibrous encapsulation, and aseptic loosening. In contrast, a more pro-healing response can support inflammatory resolution, angiogenesis, extracellular matrix remodeling, and osteogenic activity.6,27,30 Macrophage-based cell culture models can therefore provide useful information about both the cytocompatibility and inflammatory effects of curcumin-releasing AHT titanium surfaces.28,29 Despite previous reports on AHT titanium surfaces and curcumin-releasing biomaterial systems, the influence of coating methodology, polymer incorporation, and polymer composition on curcumin release remains poorly understood. Given the hydrophobic nature of curcumin and its typically limited release from biomaterial platforms, the objective of this study was to develop curcumin-loaded AHT titanium surfaces and determine whether variations in polymer composition and coating strategy could modulate curcumin release profiles. To address this objective, three delivery systems employing different polymers (PLGA, PEG, and PEG/PCL) and coating approaches were evaluated. In addition, the biological effects of released curcumin on macrophage inflammatory responses were assessed by examining M1 macrophage polarization, pro-inflammatory cytokine expression and secretion, and NF-κB activation. These studies were conducted to identify an effective curcumin delivery strategy for AHT titanium that provides sustained drug release while mitigating inflammatory activation at the biomaterial interface. Materials and methods Commercially pure titanium (CP-Ti grade #2, President Titanium, MA, USA), ethanol (200 proof anhydrous, Decon Laboratories, Inc.), acetone (certified ACS, Fisher BioReagents), sodium hydroxide (NaOH, Thermo Fisher Scientific), curcumin (Sigma Aldrich, St Louis, MO, USA), PLGA (LA : GA – 65 : 35, Sigma Aldrich, St Louis, MO, USA), polyethylene glycol (PEG, Sigma Aldrich, St Louis, MO, USA), and polycaprolactone (PCL, Sigma Aldrich, St Louis, MO, USA) were used in this study. Alkali-heat treatment (AHT) Commercially pure titanium (CP-Ti) discs were used as the substrate material for all subsequent surface modification, drug loading, and release studies. AHT was performed using our previously optimized procedure.13 Briefly, the CP-Ti discs with a diameter of ∼11 mm and a thickness of ∼2 mm were ground and polished, and then rinsed with DI water and ethanol for 10 min. AHT was conducted using 10 mL of 7.5 M NaOH solution at 60 °C and 60 rpm for 24 h, followed by heating at 600 °C for 1 h, and then cooling back to room temperature. Curcumin and polymer loading Study 1: PLGA overcoat on 100 µg Cur-loaded AHT CP-Ti For all non-control groups, 100 µg of curcumin (Cur) was loaded onto AHT CP-Ti by applying 50 µl ethanol/curcumin solution (0.2% w/v) using a drop-by-drop (DbD) method. For PLGA-coated samples, 20 µl of PLGA/acetone (5% w/v) solution was subsequently deposited onto the curcumin-loaded surfaces. After evaporation of acetone at room temperature, approximately 1.0 mg of PLGA formed a uniform coating across the surface. The loading parameters were selected based on preliminary optimization studies. The comparison groups can be seen in Table 1. Table 1Surface treatment and composition of samples for Study 1GroupCurcumin loaded (µg)Polymer loaded (µg)pHGroup 1 (Ti 7.4)007.4Group 2 (Ti 5.0)005Group 3 (Ti + Cur 7.4)10007.4Group 4 (Ti + Cur 5.0)10005Group 5 (Ti + Cur + PLGA 7.4)10010007.4Group 6 (Ti + Cur + PLGA 5.0)10010005 Surface treatment and composition of samples for Study 1 Study 2: dual PEG/PCL overcoat on 300 µg Cur-loaded AHT CP-Ti 300 µg of curcumin was deposited onto AHT CP-Ti from 30 µl solution (1% w/v) via the DbD method. Selected groups then received a PEG/PCL coating where PEG and PCL were dissolved in chloroform at a respective ratio of 35/65 and a total polymer weight of 5% w/v. The polymer coating was applied via the DbD addition of 30 µl of chloroform/PEG/PCL solution. Chloroform was evaporated, leaving a 1.5 mg polymer coating (0.525 mg PEG/0.975 mg PCL). Comparison groups are summarized in Table 2. Table 2Curcumin and PEG/PCL loading conditions and testing pH for each trial groupGroupCurcumin loaded (µg)PEG loaded (mg)PCL loaded (mg)pHGroup 1 (Cur 7.4)300007.4Group 2 (PEG/PCL 7.4)3000.5250.9757.4Group 3 (Cur 5.0)300005Group 4 (PEG/PCL 5.0)3000.5250.9755 Curcumin and PEG/PCL loading conditions and testing pH for each trial group Study 3: polymer loading method and Cur-loading quantity on release from AHT CP-Ti In Study 3, three different concentrations of curcumin (0.167% w/v, 0.333% w/v, and 1.00% w/v) were prepared in 200-proof ethanol and used to load separate sample groups. Each sample received 30 µl of curcumin solution to create loading groups with 50 µg, 100 µg, and 300 µg curcumin after ethanol evaporation. A PEG solution of 8.33% w/v was incorporated using two different approaches: mixed with the curcumin/ethanol solution before deposition (/P) or deposited on top of curcumin after loading via DbD addition of 30 µl DI-water/PEG solution (+P). Both PEG loading methods resulted in a total of 2.5 mg PEG on the CP-Ti samples. Release was evaluated at pH 7.4. Comparison groups are summarized in Table 3. Table 3Curcumin and PEG loading conditions for each trial groupGroupCurcumin loaded (µg)Polymer loaded (mg)Polymer loading methodGroup 1 (50)500N/AGroup 2 (50 + P)502.5CoatedGroup 3 (50/P)502.5MixedGroup 4 (100)1000N/AGroup 5 (300)3000N/AGroup 6 (300 + P)3002.5CoatedGroup 7 (300/P)3002.5Mixed Curcumin and PEG loading conditions for each trial group Drug release study For each test group, 3 samples were prepared under the same loading and pH conditions. Each sample was placed in an individual glass vial and 4 mL of phosphate buffer solutions (pH 7.4 or 5.0, depending on the study) were added. The vials were covered using aluminum foil to prevent light degradation of the curcumin, and they were then placed in an incubator shaker set at 37 °C and 60 rpm. The buffer solutions were removed at 1, 3, 6 and 24 hours during the first 24 hours and then every 1 to 3 days and replaced with fresh buffer solution. 600 µl of solution was pipetted into a 96-well plate in triplicate, and the curcumin concentration was measured using a UV-Vis spectrometer at a wavelength of 430 nm. Morphology of the samples before and after the drug release was evaluated using a Hitachi TM-1000 scanning electron microscope. In vitro interaction with macrophages Cell culture The NF-κB luciferase reporter RAW264.7 stable cell line (T3015; Applied Biological Materials) was cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco) supplemented with 10% fetal bovine serum (FBS; Gibco) and 1% antibiotic–antimycotic solution (Gibco). For M1 polarization, RAW264.7 macrophages (M0) were stimulated with 100 ng mL−1 lipopolysaccharide (LPS; Sigma) and 25 ng mL−1 recombinant mouse interferon-γ (IFN-γ; 315-05, PeproTech). For NF-κB pathway activation studies, cells were stimulated with a pro-inflammatory cytokine cocktail (TII), consisting of 10 ng mL−1 tumor necrosis factor-α (TNF-α; 315-01A, PeproTech), 10 ng mL−1 interleukin-1β (IL-1β; 211-11B, PeproTech), and 5 ng mL−1 IFN-γ to induce an inflammatory response. Cytotoxicity assay (cell viability assay) The cytotoxicity of curcumin was evaluated using a luminescent cell viability assay. Polished (P) and alkali-heat-treated (AHT) titanium disks were placed in 24-well plates and sterilized with ethanol prior to cell seeding. AHT disks were treated with curcumin at concentrations of 150, 75, and 37.5 µg mL−1 for 1 hour. The concentration range was selected based on preliminary observations indicating that 300 µg mL−1 curcumin exhibited reduced cytocompatibility. M0 RAW264.7 macrophages were seeded onto each disk at a density of 150 000 cells per disk and incubated for 48 hours (n = 3 per group). Cell viability was assessed using the CellTiter-Glo® 2.0 Luminescent Cell Viability Assay (G9241, Promega), which measures intracellular ATP levels as an indicator of metabolically active cells via a luciferase-based reaction. Immunocytochemical evaluation of cell attachment on titanium disk surfaces Polished, AHT, and AHT loaded with 37.5 µg mL−1 curcumin titanium disks were placed in 24-well plates. RAW264.7 macrophages were seeded onto the disk surfaces at a density of 150 000 cells per disk and incubated for 48 hours. After incubation, the culture medium was removed, and the disks were gently washed with phosphate-buffered saline (PBS). Cells were fixed with 10% formalin for 30 minutes, followed by permeabilization with 0.5% Triton X-100 in PBS for 30 minutes at room temperature. For immunofluorescence staining, samples were blocked with 5% bovine serum albumin (BSA) for 1 hour at room temperature. Cells were then incubated overnight at 4 °C with phospho-FAK (Tyr397) rabbit monoclonal antibody (1 : 200; ABT135, Sigma). After washing, the samples were incubated with goat anti-rabbit Alexa Fluor 647 secondary antibodies (1 : 200; A-21245, Invitrogen) for 1 hour at room temperature in the dark. F-actin was then stained with Alexa Fluor 568 phalloidin (1 : 200; A-11011, Invitrogen), and nuclei were counterstained with DAPI. Samples were imaged using a Zeiss LSM 980 laser scanning confocal microscope. Proinflammatory responses of macrophages on titanium disk surfaces RAW264.7 macrophages were seeded onto the polished, AHT and AHT disks loaded with two doses of curcumin (75 and 37.5 µg mL−1) at a density of 150 000 cells per disk and incubated for 4 hours to allow cell attachment. Cells were then polarized to the M1 phenotype (n = 3 per group). Total RNA was isolated at 24 and 48 hours using the Direct-zol RNA Miniprep Kit (R2051, Zymo Research) according to the manufacturer's instructions. Complementary DNA (cDNA) was synthesized from 0.5 µg of total RNA using the RevertAid RT Reverse Transcription Kit (K1691, Thermo Scientific). Quantitative real-time PCR (qRT-PCR) was performed to evaluate pro-inflammatory gene expression, including IL 1β, IL 6, and Nos2. Gene expression levels were normalized and presented as mean fold change relative to the polished titanium disk group. To evaluate pro-inflammatory protein expression, conditioned media were collected after 48 hours of incubation. The levels of IL-6 and TNF-α were then quantified using enzyme-linked immunosorbent assay (ELISA) kits (R&D Systems), following the manufacturer's protocols. NF-κB reporter assay Macrophages were seeded onto the disks at a density of 150 000 cells per disk and incubated for 4 hours to allow cell attachment, followed by treatment with the TII cytokine cocktail for 48 hours. After stimulation, the culture medium was removed, and the disks were washed with PBS. The cells were then lysed using cell lysis buffer for 15 minutes at 4 °C. The lysates were centrifuged at 12 000g for 10 minutes to remove cellular debris, and the supernatants were collected for luciferase activity measurement using a luciferase assay kit (Promega). NF-κB p65 nuclear translocation analysis The translocation of phosphorylated NF-κB p65 was evaluated by immunocytochemistry. RAW264.7 macrophages were seeded onto titanium disks and stimulated with the TII cytokine cocktail for 1 hour. Following stimulation, the cells were fixed, permeabilized, and processed for immunofluorescence staining. Samples were first blocked with 5% BSA for 1 hour, then incubated overnight at 4 °C with NF-κB p65 (L8F6) mouse monoclonal antibodies (1 : 200; 6956S, Cell Signaling Technology). After washing, the samples were incubated with donkey anti-mouse Alexa Fluor 647 secondary antibodies (1 : 200; A-31571, Invitrogen) for 1 hour at room temperature in the dark. Nuclei were counterstained with DAPI. Samples were imaged using a Zeiss LSM 980 laser scanning confocal microscope. Statistical analysis For curcumin loading and release studies, three samples from each group were prepared. The curcumin release was measured from three independent samples, with triplicate readings from each sample. The release data were plotted as mean ± standard deviation for each time point. For comparisons between two groups, statistical analysis was performed using an unpaired Student's t-test. For experiments involving comparisons among more than two groups, one-way analysis of variance (ANOVA) was performed with a 95% confidence interval, followed by Tukey's post hoc multiple comparisons test. Differences were considered statistically significant at P < 0.05. Results Curcumin release Study 1: PLGA overcoat on 100 µg Cur-loaded AHT CP-Ti In Study 1, all curcumin-loaded groups showed a rapid initial release followed by a slower, more gradual phase over the remainder of the study. This can be seen in Fig. 1A and B where the largest burst release occurred in the uncoated pH 7.4 group, which reached ∼17.7 μg release by 24 h and then increased more slowly to ∼21.3 μg by 96 h; the curve began to level off after about 24 to 48 h. The uncoated pH 5.0 group showed a smaller initial burst, reaching ∼11.1 μg release by 24 h, then gradually increasing to ∼16.5 μg by 96 h, with a less pronounced plateau than the pH 7.4 group. Both PLGA-coated groups released less curcumin throughout the study and showed flatter profiles overall. The PLGA pH 7.4 group showed the lowest burst release, reaching only ∼3.3 μg release by 24 h and ∼6.0 μg by 96 h, with the curve appearing relatively flat after the first 24 h. The PLGA pH 5.0 group showed an intermediate profile, reaching ∼7.9 μg release by 24 h and ∼12.5 μg by 96 h, with a more gradual increase afterward. Overall, the release profiles indicate that most release occurred during the early time points, especially in the uncoated groups, while the PLGA-coated groups showed reduced burst release and earlier flattening of the curves. Study 1 curcumin release profile and post-release surface morphology of AHT CP-Ti samples. (A) Curcumin burst release during the first 24 h and (B) cumulative curcumin release over 96 h. (C) Representative SEM images of Ti surfaces after the release study at 500× (bar = 100 μm) and 1000× magnifications (insets, bar = 50 μm). The uncoated Cur group showed minimal residual curcumin at pH 7.4 but crystal-like residues at pH 5.0. PLGA-coated groups retained rounded deposits after release, with smaller and finer deposits at pH 5.0 than at pH 7.4, indicating pH- and coating-dependent effects on post-release surface morphology. Fig. 1C shows clear differences in the surface appearance of the Study 1 samples after release as a function of both pH and PLGA coating. In the uncoated curcumin groups, the pH 7.4 surface appears to have scattered small deposits and less distinct large-scale structure, whereas the pH 5.0 surface shows more prominent localized dark, needle-like or star-shaped features, suggesting a more concentrated or crystalline residue. In contrast, both PLGA-coated groups display a much more continuous film-like morphology with numerous rounded, droplet-like or circular domains across the surface. These rounded features are especially apparent in the PLGA samples and suggest that the polymer formed a coating distributed across the surface rather than isolated crystalline deposits. Study 2: dual PEG/PCL overcoat on 300 µg Cur-loaded AHT CP-Ti As shown in Fig. 2, all groups showed a relatively small initial burst release followed by a slow, gradual increase in release over the rest of the study. By 24 h, the percent release remained low under all conditions, at ∼1.56% for Cur pH 7.4, ∼1.78% for Cur pH 5.0, ∼1.44% for PEG/PCL pH 7.4, and ∼1.55% for PEG/PCL pH 5.0. The largest early release occurred in the uncoated pH 5.0 group with ∼5.3 μg, followed by the uncoated pH 7.4 group (∼4.7 μg), while the polymer-coated groups showed slightly lower burst release with ∼4.3 μg and ∼4.6 μg for pH 7.4 and pH 5.0, respectively. After the first day, all curves continued to rise gradually without a strong plateau, indicating continued low-level release through the later time points. By the end of the study at 360 h, the total percent release was ∼3.63% for Cur pH 7.4, ∼4.26% for Cur pH 5.0, ∼5.63% for PEG/PCL pH 7.4, and ∼4.50% for PEG/PCL pH 5.0. The PEG/PCL pH 7.4 group showed the greatest final release of ∼16.9 μg, while the Cur pH 7.4 group showed the lowest at ∼10.9 μg. Standard deviations in the percent-remaining data were generally small at early time points and remained modest overall, although variability increased somewhat at selected later time points, particularly around 168 h for Cur pH 7.4 and 336 h for PEG/PCL pH 5.0. Study 2 curcumin release profile. (A) Curcumin burst release during the first 24 h and (B) cumulative curcumin release over 360 h. Study 3: polymer loading method and Cur-loading quantity on release from AHT CP-Ti In Study 3, both curcumin loading amount and PEG loading method affected release behavior at pH 7.4. As shown in Fig. 3A and B, all groups exhibited an initial burst release followed by a slower increase in cumulative release over time. In terms of cumulative release, the 50 µg and 50 + P groups reached ∼5.8 and ∼5.6 µg by 24 h, while the 100 µg group reached ∼12.7 µg and the 300 µg group reached ∼9.0 µg. The 50/P group showed the largest early release among all groups, reaching ∼8.9 µg by 24 h, while the 300 + P and 300/P groups reached ∼7.9 and ∼7.7 µg, respectively. By the final time point, cumulative release was highest for 300 + P at ∼25.5 µg, followed by 300/P at ∼23.2 µg, 300 µg at ∼20.0 µg, 100 µg at ∼17.8 µg, 50/P at ∼11.1 µg, 50 + P at ∼8.1 µg, and 50 µg at ∼8.0 µg. When expressed as percent release, the ranking differed because of the varied starting loads. By 24 h, the 50 µg, 50 + P, and 100 µg groups released 11.6%, 11.1%, and 12.7%, respectively, whereas the 300 µg, 300 + P, and 300/P groups released only 3.0%, 4.2%, and 4.1%. The 50/P group showed the greatest early percent release at 17.7%. By the final time point, the total percent release was highest for 50/P at about 22.1%, followed by 100 µg at about 17.8%, 50 + P at about 16.2%, 50 µg at about 16%, 300 + P at about 8.5%, 300/P at about 7.8%, and 300 µg at about 6.7%. Overall, the lower-loading groups showed greater fractional release, while the higher-loading groups released more total curcumin in µg but a smaller percentage of the loaded amount. Study 3 curcumin release profile and post-release surface morphology of AHT samples (300 µg groups). (A) Curcumin burst release during the first 24 h. (B) Cumulative curcumin release over 480 h. (C) Representative SEM images of AHT surfaces after release at 500× (bar = 100 μm) and 2500× magnifications (insets, bar = 20 μm), showing differences in residual surface debris. The 300/P group exhibited the least residual debris, the uncoated 300 µg group showed an intermediate amount, and the 300 + P group retained the greatest amount of post-release debris. Fig. 3C shows that the 300 µg groups had clearly different post-release surface morphologies depending on the PEG loading method. The uncoated 300 µg sample showed several relatively large, irregular residual deposits scattered across the titanium surface. The 300/P sample appeared much more uniform at low magnification, with fewer large, isolated features and a finer distribution of small residual particles across the surface. The 300 µg + P sample showed a greater number of visible surface deposits than the mixed PEG condition, including several larger bright features and more scattered residue over the surface. At 2500× magnification, differences can be seen in local deposit structure among the three groups, indicating that PEG addition changed not only the amount of visible residue but also how the residual curcumin was distributed on the AHT titanium after release. Overall, the SEM images suggest that the mixed PEG condition produced the most uniform post-release surface among the 300 µg groups, while the uncoated and post-coated PEG samples retained more distinct localized deposits. In vitro interaction with macrophages Effects of surface treatment on macrophage morphology Confocal fluorescence microscopy imaging of F-actin and nuclei revealed distinct differences in macrophage morphology between the two surfaces (Fig. 4A). Cells cultured on the polished surface exhibited extensive spreading and pronounced actin-rich protrusions, along with a greater vertical profile, as confirmed by orthogonal (XZ) projections. In contrast, cells on the AHT surface remained adherent but displayed a more rounded and flattened morphology with reduced spreading. Notably, curcumin treatment did not significantly alter cell attachment or morphological characteristics on the AHT surface. Macrophage morphology and viability on titanium surfaces. (A) Representative confocal fluorescence images of RAW264.7 macrophages cultured on polished (P), alkali-heat-treated (AHT), and curcumin-loaded AHT (Cur@AHT) Ti surfaces. F-actin is stained red and nuclei are stained blue. (B) Cell viability of macrophages cultured on Ti disks, measured using a luciferase-based metabolic activity assay. Data are presented as mean ± SD and normalized to the polished surface (n = 3 per group). *p < 0.05, determined by one-way ANOVA followed by Tukey's post hoc test. Cell viability on titanium surfaces and curcumin treatment To evaluate the effects of titanium surface properties and curcumin loading on macrophage viability, RAW264.7 cells were cultured on titanium disks. Cells on AHT surfaces exhibited significantly higher metabolic activity compared to those on polished surfaces (Fig. 4B). In the presence of curcumin, all tested concentrations significantly reduced cell viability compared to untreated AHT surfaces. However, the highest dose (150 µg mL−1) resulted in a greater decrease in cell viability, comparable to levels observed on polished surfaces. In contrast, the 75 and 37.5 µg mL−1 curcumin treatments maintained higher cell viability, with no significant difference between these two groups. Based on these findings, the 75 and 37.5 µg mL−1 curcumin concentrations were selected for subsequent cytokine expression studies to evaluate their effects on M1 macrophage polarization. Effects of surface properties and curcumin on classically activated (M1) macrophages M1 macrophages are among the first inflammatory cells to colonize biomaterial surfaces following implantation and play a critical role in initiating the early foreign body and inflammatory response. Therefore, evaluating the effects of released curcumin on M1 polarization is particularly relevant for assessing the potential of this delivery system to modulate the early inflammatory environment at the biomaterial interface. Accordingly, to evaluate the effects of titanium surface properties and curcumin treatment on pro-inflammatory macrophage responses, RAW264.7 cells were cultured on different titanium disk surfaces and polarized toward the M1 phenotype. Two concentrations of curcumin-loaded AHT (75 and 37.5 µg mL−1) were selected for the anti-inflammatory assays based on the cytocompatibility results to minimize potential confounding effects arising from reduced cell viability. At 24 hours, macrophages cultured on AHT surfaces exhibited elevated expression of pro-inflammatory markers compared to polished surfaces (Fig. 5A). However, by 48 hours, the expression of these markers on AHT surfaces was reduced (Fig. 5B), indicating a time-dependent modulation of the inflammatory response. Curcumin loading on AHT surfaces resulted in a marked reduction in M1-associated gene expression at both time points. At 24 hours, both curcumin concentrations (75 and 37.5 µg mL−1) demonstrated comparable inhibitory effects, with no significant differences between the two groups. At 48 hours, the lower dose (37.5 µg mL−1) further reduced Il6 expression compared to the 75 µg mL−1 treatment group. At the protein level, these findings were consistent with the gene expression results. The secretion of pro-inflammatory cytokines IL-6 and TNF-α was significantly decreased in macrophages cultured on AHT surfaces compared to polished surfaces. Curcumin treatment further reduced TNF-α secretion relative to the AHT group (Fig. 5C). Overall, AHT surfaces exhibited a time-dependent reduction in pro-inflammatory gene expression, and curcumin loading further modulated macrophage phenotype, suggesting an additive synergistic effect on inflammatory regulation. Pro-inflammatory gene and cytokine expression in M1-polarized macrophages cultured on titanium surfaces. (A and B) Relative mRNA expression of the pro-inflammatory genes Il1b, Il6, and Nos2 in M1-polarized RAW264.7 macrophages cultured on polished (P), alkali-heat-treated (AHT), and curcumin-loaded AHT (Cur@AHT) Ti surfaces for 24 h (A) and 48 h (B). Gene expression levels are presented as mean fold change ± SD and normalized to the polished surface control (n = 3 per group). (C) Relative protein levels of the pro-inflammatory cytokines IL-6 and TNF-α in conditioned media collected from M1-polarized macrophages cultured on each surface for 48 h. Data are presented as mean ± SD and normalized to the polished surface control (n = 3 per group). *p < 0.05, determined by one-way ANOVA followed by Tukey's post hoc test. Effects of surface properties and curcumin on macrophages’ NF-κB activity NF-κB is a key transcription factor involved in the regulation of inflammatory responses and cell survival. Under basal conditions, macrophages cultured on polished and AHT surfaces exhibited differences in NF-κB activity. Confocal imaging showed a slightly higher NF-κB signal in cells on polished surfaces compared to AHT surfaces (Fig. 6A). Consistently, the NF-κB luciferase reporter assay demonstrated significantly lower NF-κB activity in macrophages cultured on AHT surfaces, indicating reduced basal inflammatory signaling (Fig. 6B). Upon stimulation with the TII cytokine cocktail, NF-κB activity markedly increased on both polished and AHT surfaces, as evidenced by enhanced nuclear localization of phosphorylated p65 (Fig. 6C). Notably, curcumin treatment (37.5 µg mL−1) on AHT surfaces significantly reduced the nuclear translocation of phosphorylated p65. This effect was further confirmed by the NF-κB reporter assay, which showed a significant decrease in NF-κB activity in the curcumin-treated group compared to both polished and untreated AHT groups (Fig. 6D). Overall, AHT surfaces reduced basal NF-κB activation, and curcumin loading further suppressed TII-induced NF-κB signaling, demonstrating a combined inhibitory effect on inflammatory activation in macrophages. NF-κB activation in macrophages cultured on titanium surfaces. (A) Representative confocal microscopy images of RAW264.7 macrophages cultured on polished (P) and alkali-heat-treated (AHT) Ti surfaces under basal conditions. NF-κB p65 is shown in green and nuclei are stained blue. (B) NF-κB reporter activity of macrophages cultured on Ti disks for 48 h under basal conditions. Data are presented as mean ± SD and normalized to the polished surface control (n = 3 per group). *p < 0.05, Student's t-test. (C) Representative confocal microscopy images of macrophages stimulated with the TII cytokine cocktail and cultured on P, AHT, and curcumin-loaded AHT (Cur@AHT; 37.5 µg mL−1) surfaces. NF-κB p65 is shown in green and nuclei are stained blue. (D) NF-κB reporter activity of TII-stimulated macrophages cultured on Ti disks for 48 h. Data are presented as mean ± SD and normalized to the polished surface control (n = 3 per group). *p < 0.05, one-way ANOVA followed by Tukey's post hoc test. Discussion Drug release studies The release profiles in Study 1 show that both pH and the presence of a PLGA overcoat influenced curcumin transport from the AHT titanium surface. In the uncoated groups, greater release was observed at pH 7.4, particularly during the first 24 h, suggesting that curcumin partitioned more readily into the neutral medium than into the acidic medium. At pH 5.0, curcumin exists in its protonated form, which is highly hydrophobic and has low solubility. As a result, even when curcumin detaches from AHT Ti, it may remain adsorbed to or closely associated with the surface rather than diffusing into the surrounding medium.31 On the other hand, with an increase in pH to 7.4, a large portion of curcumin becomes deprotonated, increasing its solubility and release into the solution.32 Additionally, the presence of a hydroxyl-rich titanate (Ti–OH) surface has a secondary effect in AHT Ti. Curcumin can interact with the surface through hydrogen bonding, and changes in pH can alter the protonation of the surface hydroxyl group, as well as curcumin, resulting in even greater release at pH 7.4, even though the molecule itself is less stable.33–35 The presence of a PLGA coating reduced curcumin release under both pH conditions, suggesting that the polymer layer acted as a diffusion-limiting barrier, with a more pronounced effect at pH 7.4. The decrease in curcumin release in the presence of PLGA is consistent with PLGA drug-delivery mechanisms, in which early release is governed by water ingress, drug dissolution, and diffusion through the intact polymer matrix before slower degradation-controlled processes become more important.36,37 In addition, the overall release was higher in coated samples at pH 5.0 than at pH 7.4 (Fig. 1A and B). PLGA contains hydrolyzable ester bonds, which allow hydrolysis and water uptake to accelerate and facilitate erosion and degradation in an acidic environment.38 PLGA degradation occurs primarily through hydrolytic cleavage after water penetrates the polymer matrix, and its degradation rate depends on factors such as lactic acid : glycolic acid ratio, molecular weight, coating thickness, and release medium conditions.36,37,39 In addition, PLGA hydrolysis produces lactic acid and glycolic acid close to the surface of AHT Ti, which can further lower the local pH and promote curcumin release. Overall, the findings from Study 1 suggest that the PLGA overcoat has a pH-dependent effect on curcumin release from AHT Ti and fundamentally changes the release mechanism. These results indicate that even a simple drop-cast PLGA overcoat can alter curcumin transport, although further optimization of polymer composition, coating uniformity, and drug–polymer interactions will be needed to achieve more predictable and tunable release behavior. In Study 2, our goal was to evaluate whether the release behavior depends on the amount of loaded curcumin and the polymer coating composition. The curcumin content was increased to 300 µg (compared with 100 µg in Study 1), and PLGA was substituted with 35/65 PEG/PCL. Curcumin release was higher at pH 5.0 than at pH 7.4 when AHT samples were loaded with a high concentration of curcumin. Moreover, at earlier time points, the presence of a polymer coating suppressed curcumin release; however, higher release was observed in PEG/PCL coated samples at later time points. Alkali-heat treatment creates a porous, web-like structure, and we have previously shown that the microstructure is highly dependent on the processing conditions.17 At high curcumin concentrations, the porous structure that serves as the adsorption site for curcumin can become saturated. Although curcumin is more soluble and less stable at pH 7.4, the lower release might be attributed to possible dynamic desorption–readsorption of curcumin, a phenomenon reported for other stable coatings on Ti at neutral pH.40 Alternatively, partial curcumin degradation may have contributed to the reduced measured release; however, no clear evidence of degradation was observed in the present study. Unlike Study 1, where the PLGA overcoat mainly suppressed release, the 35/65 PEG/PCL coating in Study 2 only slightly reduced the early burst while ultimately allowing greater cumulative release, especially at pH 7.4. This suggests that the PEG/PCL layer did not act as a purely diffusion-limiting barrier. Instead, it likely reduced the earliest surface loss while also improving longer-term wetting and drug accessibility. This interpretation is consistent with the different transport roles and degradation behavior of the polymers. PEG is hydrophilic and water-soluble, which can promote water uptake, improve hydration of the deposited curcumin, and increase the accessibility of the drug reservoir to the release medium. In contrast, PCL is hydrophobic, semicrystalline, and slow degrading, with degradation typically occurring over a much longer timescale than the present release study; therefore, it tends to slow diffusion and prolong transport through the coating.41–43 As a result, the PEG/PCL blend likely behaved as a mixed system with counteracting effects, in which PEG-supported hydration and PCL-supported transport resistance acted simultaneously, shifting release away from a single sharp burst and toward a slower and more accessible release phase over time. This also helps explain why the PEG/PCL coating in Study 2 differed from the PLGA overcoat in Study 1. PLGA release behavior depends on water penetration, drug dissolution, polymer degradation, and the interaction among the drug, polymer, titanium surface, and release medium.39 Therefore, the PLGA overcoat in Study 1 likely behaved more like an intact hydrophobic barrier during the release period, while the PEG/PCL coating in Study 2 provided both water-accessible PEG domains and diffusion-limiting PCL domains. This distinction is important because the 300 µg curcumin loading in Study 2 created a larger drug reservoir that may have required improved hydration before additional drug could dissolve and diffuse into the medium. The PEG-containing coating may therefore have increased the accessibility of this larger reservoir over time, while the PCL phase still moderated the early burst. This explains why the PLGA-coated samples in Study 1 showed lower release, while the PEG/PCL-coated samples in Study 2 could show moderated early release but greater cumulative release. The pH dependence in Study 2 was also more complex than in Study 1. During the first 24 h, the uncoated pH 5.0 group released slightly more curcumin than the uncoated pH 7.4 group, whereas over the full study duration the PEG/PCL pH 7.4 group gave the highest cumulative release. This mixed behavior may reflect competing effects of solubility, stability, and polymer wetting. As mentioned above, curcumin generally shows very low aqueous solubility under acidic conditions but is also known to degrade more rapidly near neutral and alkaline pH, especially in phosphate-containing media.33,35,44 As a result, pH 7.4 may favor partitioning of curcumin out of the coating while also increasing the likelihood of degradation of the released drug, whereas pH 5.0 may preserve curcumin better once released but still limit its dissolution from the surface. At the higher 300 µg loading used here, another likely factor is that a greater fraction of the drug remained as poorly wet or partially aggregated material on the surface in the uncoated groups, which would restrict the percentage released. Incorporating the PEG/PCL layer may have improved dispersion and hydration of that larger curcumin reservoir, allowing more of the loaded drug to eventually enter the medium, even though early release remained moderated by the PCL phase. Overall, these results suggest that at a higher loading concentration of 300 µg, the PEG/PCL coating was more effective at prolonging access to curcumin than at strongly suppressing total release, and that the balance between PEG-driven wetting and PCL-driven transport resistance became more important than in the lower-loading PLGA system. This is relevant to the present study because it shows that polymer coatings on AHT titanium can influence not only burst release, but also how much of a larger curcumin reservoir remains available for longer-term release. After evaluating the effects of pH and polymer coating, the goal of Study 3 was to evaluate whether the curcumin loading amount and the method of polymer incorporation influence curcumin release behavior. Among the uncoated groups, increasing the loading amount increased the absolute amount of curcumin released, but decreased the percentage of the loaded drug that was released. This is evident from the much lower percent release of the 300 µg group compared with the 50 µg and 100 µg groups, both during the burst phase and over the full study. This suggests that at lower loadings of 50 µg and 100 µg, the majority of the curcumin is successfully adsorbed within the AHT porous structure, and once the releasable fraction leaves the surface, the remaining curcumin is strongly retained, resulting in a plateau in the release profile. However, at a higher loading of 300 µg, a greater portion of the curcumin remained in less accessible deposits on the AHT surface, likely because thicker or more aggregated hydrophobic domains formed during drying. This is consistent with the general challenge of curcumin delivery, as its low aqueous solubility, chemical instability, and tendency to remain in poorly wetted hydrophobic domains can limit release even when the total loaded mass is high.33–35 Notably, the release from the 300 µg samples continued to increase, whereas the release from the 50 µg and 100 µg samples reached a plateau. In addition, another interesting finding was that the release from the 300 µg sample at early time points was lower than that from the 100 µg samples, which supports our interpretation of the release behavior; at early time points, the 300 µg layer may be thicker or more aggregated, such that the primary exposed curcumin fraction does not necessarily diffuse out faster. The diffusion path can be longer, and the curcumin–curcumin interactions may slow down the initial release. The effects of PEG incorporation and configuration further indicate that not only polymer presence but also polymer distribution relative to the drug influenced release behavior. In the 300 µg specimens, when PEG was coated on the curcumin surface (300 + P), hydration of hydrophilic PEG likely increased curcumin wetting in the underlying layers and facilitated the transport of curcumin into the release medium. When PEG was mixed with curcumin (300/P), PEG not only facilitated wetting but also improved the distribution and stability of curcumin.45 However, the effect of PEG is not as strong as that observed with 300 µg, as at a lower loading of 50 µg, most curcumin was associated with the AHT surface rather than PEG as discussed earlier. At the 50 µg loading, mixing PEG directly with curcumin (50/P) produced both the greatest burst and the greatest overall release, whereas applying PEG afterward as a surface coating (50 + P) had little effect relative to the uncoated 50 µg group. This suggests that when PEG was intimately mixed with curcumin, it more effectively increased local wettability and water access within the deposited layer, thereby accelerating dissolution and outward transport. This behavior is consistent with the general drug/polymer/media interaction concept, where release depends on water penetration into the coating or deposited layer, partial drug dissolution, diffusion, and the relative interactions among the drug, polymer, titanium surface, and release medium.39 Unlike PLGA or PCL, PEG is not expected to control release primarily through slow degradation during the time frame of this study. Instead, because PEG is hydrophilic and water-soluble, its main effect was likely rapid hydration and partial dissolution or leaching from the deposited layer, which could create water-accessible pathways and improve curcumin transport.41–43 PEG-containing polymer systems can increase hydrophilicity and water uptake, while PCL-containing systems provide more hydrophobic transport resistance and slower degradation behavior.41–43 The SEM image of the 300 µg groups supports this interpretation qualitatively. The uncoated 300 µg surface showed larger and more distinct residual deposits, while the 300/P surface appeared more uniform with fewer large, isolated features. The 300 + P surface showed more visible residue than the mixed PEG condition, including several larger bright deposits, suggesting that PEG mixed directly with curcumin may have improved dispersion of the hydrophobic drug more effectively than a PEG overcoat applied after deposition. Together, the findings from Study 3 suggest that, at high curcumin loadings, PEG improved hydration and mobilization of the curcumin reservoir; however, the extent of this effect depended on how the polymer was incorporated. Overall, these results indicate that PEG can influence release in two different ways: when mixed with curcumin, it can strongly enhance early release by increasing drug accessibility, whereas when applied after loading, it may have a more modest or delayed effect that depends on the size and structure of the underlying drug deposit. This is important for the present study because it shows that both loading amount and polymer placement are meaningful design variables for controlling burst release and longer-term curcumin availability from AHT titanium surfaces. In vitro interaction with macrophages Upon implantation of titanium-based prostheses, such as dental implants and orthopedic joint replacements, macrophages represent the primary cellular interface with the material, initiating the early host response. The success of these treatments largely depends on favorable interactions between the implant surface and host immune cells following implantation. In the present study, we first evaluated macrophage attachment and morphology on polished and AHT titanium surfaces. Although all surfaces supported macrophage attachment, AHT modification provided a more favorable microenvironment for cell attachment, survival, and function. This was evidenced by enhanced cell morphology, well-organized F-actin structures observed through immunofluorescence staining, and increased metabolic activity measured by cell viability assays. The improved cell attachment on AHT surfaces may be attributed, at least in part, to their more uniform surface characteristics, including reduced and homogeneous surface roughness. These features are known to enhance cell–material interactions, promote cytoskeletal organization, and improve adhesion stability.46 Curcumin, a bioactive compound derived from turmeric, is widely recognized for its anti-inflammatory properties, primarily through inhibition of NF-κB activation.47 In this study, we aimed to evaluate the immunomodulatory activity of curcumin released from AHT surfaces. Therefore, inflammatory responses were assessed using curcumin-loaded AHT surfaces without polymer coatings to isolate the biological effects of released curcumin and avoid potential confounding effects from the polymers. While the influence of polymer composition and coating strategy on macrophage responses is an important topic for future investigation, the primary objective of this study was to establish the bioactivity of released curcumin. When macrophages were cultured on the curcumin-loaded AHT surfaces, a modest reduction in viability was observed relative to AHT alone. However, all curcumin-treated groups exhibited equal or higher viability than polished surfaces, indicating that curcumin incorporation remained cytocompatible within the tested concentration range. To minimize potential confounding effects, subsequent inflammatory analyses were performed using concentrations selected based on these viability results. While macrophage metabolic activity can influence inflammatory activation,48 the observed reductions in inflammatory gene expression and cytokine production are unlikely to be explained solely by decreased metabolic activity. Notably, macrophages cultured on polished titanium also exhibited reduced metabolic activity but did not show a corresponding decrease in inflammatory responses, suggesting that the anti-inflammatory effects were primarily attributable to curcumin-mediated immunomodulation rather than reduced cell metabolism. Macrophages are highly plastic immune cells that dynamically alter their phenotype in response to functional demands.49 In response to tissue injury, they are activated by cytokines and local microenvironmental cues, which drive their polarization into distinct functional states. In this study, RAW264.7 macrophages were cultured on different titanium surfaces and polarized toward the classically activated (M1) phenotype. This phenotype is characterized by the upregulation of pro-inflammatory cytokines, reflecting its role in initiating and propagating the early inflammatory response.50 Quantitative analysis of macrophage inflammatory responses using qRT-PCR revealed a time-dependent effect of titanium surface properties. Although a modest increase in M1-associated gene expression was observed at 24 h on AHT surfaces compared to polished surfaces, this response was attenuated by 48 h. The transient elevation in pro-inflammatory markers may reflect early macrophage activation in response to the unique topographical and chemical characteristics of the AHT surface. However, with prolonged exposure, macrophages appeared to adapt to the bioactive surface environment, resulting in reduced inflammatory signaling and lower expression of pro-inflammatory markers. Importantly, the presence of curcumin significantly suppressed the expression of pro-inflammatory cytokines at both the mRNA and protein levels at both time points. These findings highlight the ability of curcumin to attenuate macrophage inflammatory responses, likely through modulation of NF-κB signaling, a central regulator of inflammatory gene expression and an attractive target for therapeutic intervention. To further validate this mechanism, immunofluorescence staining was performed to assess the nuclear translocation of NF-κB p65 in macrophages. Consistent with the gene and protein expression findings, AHT surfaces exhibited lower basal levels of nuclear p65 compared to polished surfaces. Upon stimulation with the TII cytokine cocktail, which mimics the inflammatory cytokine surge during early-stage inflammation, macrophages cultured on polished surfaces displayed a pronounced increase in p65 expression and nuclear translocation relative to AHT surfaces. In contrast, curcumin-treated groups showed markedly reduced nuclear localization of p65, confirming the inhibitory effect of curcumin on NF-κB activation. These results further support the role of curcumin in modulating inflammatory signaling pathways and attenuating macrophage activation when loaded onto AHT surfaces. Conclusion Collectively, these findings demonstrate the potential of AHT and curcumin-loaded AHT titanium surfaces as an effective strategy for regulating macrophage behavior and mitigating inflammation at the biomaterial interface. The release profile of curcumin was highly dependent on the loading amount and the presence and composition of polymer coatings. Acidic conditions reduced curcumin release, while higher curcumin loading increased the total amount released and promoted a more sustained release profile; however, the percentage of curcumin released decreased with increasing curcumin loading. Biologically, AHT surface modification alone enhanced macrophage compatibility and reduced inflammatory signaling over time, while curcumin incorporation provided an additional suppressive effect on pro-inflammatory responses. Together, AHT and curcumin effectively modulated macrophage activity, combining favorable cytocompatibility with potent anti-inflammatory effects. These findings suggest that the curcumin-loaded AHT Ti platform may promote a more favorable early inflammatory microenvironment by attenuating M1 macrophage activation. Future studies investigating macrophage-mediated effects on osteogenic cells and in vivo implant integration are needed to further determine the translational significance of these findings. Author contributions Conceptualization: S. V. and M. K.; methodology, experiments, investigation, data interpretation and data curation, writing and revision: A. C., N. A., S. V., and M. K.; supervision, project administration, and funding acquisition: S. V. and M. K.; all authors have read and agreed to the published version of the manuscript. Conflicts of interest The authors declare no conflicts of interest.