Interfacial mild alkalinity shapes Neutrophil immunometabolism through sphingolipid remodeling to enhance antibacterial host defense

Abstract

Implant-associated infection is commonly addressed by surfaces designed for direct bacterial killing, yet early host immune events at the implant interface are critical for infection outcome. Here, we designed Ti-A, a mildly alkaline titanium interface, to examine whether interfacial pH could act as an immune-regulatory cue for host-directed infection control. Ti-A showed acceptable cytocompatibility and promoted neutrophil extracellular traps (NETs) formation, thereby enhancing neutrophil-mediated killing of S. aureus. The attenuation of antibacterial activity after DNase I treatment supported the functional contribution of NETs. Metabolomic and molecular analyses further suggested that sphingolipid remodeling may converge on enhanced SPHK1-S1PR2/ROS signaling during Ti-A-induced NET formation. In a bacteria-loaded implant-induced bone defect model, Ti-A reduced local bacterial burden with evidence of NET-associated host defense at the infected interface. Importantly, this immune-antibacterial response did not compromise bone repair; Ti-A preserved osteogenic activity, promoted matrix mineralization, and improved infected bone defect repair without evident systemic toxicity. These findings identify mild interfacial alkalinity as a material-encoded cue that tunes neutrophil immunometabolism toward NET-mediated infection control while supporting subsequent bone regeneration.

Graphical abstract

1. Introduction

Orthopedic implants have substantially improved the treatment of bone defects and fractures, but implant-associated infection remains a major barrier to long-term clinical success [1,2]. Early bacterial adhesion and colonization on implant surfaces can lead to biofilm formation [3], which sustains local inflammation, disrupts the peri-implant microenvironment, impairs osseointegration, and may ultimately cause implant failure [4,5]. In clinical practice, these infections often require revision surgery and prolonged systemic antimicrobial therapy, adding to healthcare costs and reducing patients’ quality of life [6].

Current efforts to prevent implant-associated infection have mainly focused on antibacterial surfaces, including systems that release antibacterial ions or antibiotics, present antimicrobial peptides, or respond to external stimuli to kill bacteria [[7], [8], [9]]. Although these systems differ in design, most are built on the expectation that the material itself can directly suppress bacterial growth [5,10,11]. Such strategies may reduce the initial bacterial burden, but their efficacy is often limited by unstable release, difficulty in maintaining effective local concentrations, and limited activity against established biofilms [12,13]. However, this view still places the implant and bacteria at the center of the problem, while the host response occurring at the early interface is often overlooked [14]. Importantly, increasing evidence suggests that the outcome of implant-associated infection is not determined solely by bacterial colonization itself, but also by the quality of the early host immune response at the implant interface. Therefore, strategies capable of actively engaging host antibacterial defense may provide a complementary route beyond direct material-mediated bactericidal activity.

Implant-associated infection is increasingly understood not as a simple material–pathogen interaction, but as a process shaped by a dysregulated interfacial microenvironment involving materials, bacteria, and host immunity [15]. At this interface, the functional state of immune cells can strongly influence the course of infection [[16], [17], [18]]. Neutrophils are among the first innate immune cells recruited to infected sites and constitute an early line of host defense [19,20]. Beyond phagocytosis and degranulation, neutrophils can release neutrophil extracellular traps (NETs), extracellular chromatin-based networks that capture pathogens and limit their local spread [21,22]. Given their role in bacterial clearance and early infection control, locally guiding neutrophil behavior at the implant interface, especially controlled NET formation, may offer an anti-infective strategy based on host intrinsic defense rather than direct material-mediated killing.

Achieving such regulation requires local signals that can be stably modulated at the implant interface and effectively sensed by immune cells. Among possible interfacial cues, pH is particularly relevant because it links the physicochemical properties of materials with cellular responses, yet its role in immune regulation has long been underestimated [[23], [24], [25]]. Leppkes et al. reported that bicarbonate-rich pancreatic fluid promoted neutrophil chromatin extrusion and PADI4-dependent aggregated NET formation [26]. Maueröder et al. further showed that NETs release was favored under a high bicarbonate-to-CO₂ ratio and moderately alkaline pH [27], whereas Behnen et al. demonstrated that extracellular acidosis suppressed ROS-dependent NETosis [28]. Together, these studies support pH as an extracellular regulator of NET formation, although how this signal is converted into neutrophil functional changes remains insufficiently understood. One possible link is immunometabolic regulation, as immune cell activity is tightly coupled to cellular metabolic state and can be reshaped by local microenvironmental signals [29,30]. Accordingly, local pH variations should not be regarded merely as passive physicochemical conditions, but as regulatory factors capable of influencing neutrophil immunometabolic states and functional outputs, including NET formation. In this context, a mildly alkaline interface may serve as a local immunometabolic cue that contributes to shaping early host antibacterial responses.

These considerations raise an important question for implant design. Can a local pH signal be deliberately built into the implant interface to guide neutrophil immunometabolic activity and strengthen host antibacterial defense without adding exogenous bactericidal agents? For bone implants, such a strategy must also remain compatible with tissue repair. An ideal bone implant interface should restrict early bacterial colonization while preserving conditions that support later bone regeneration. Whether a mildly alkaline interface can provide this balance among immune regulation, antibacterial defense, bone repair, and biosafety remains to be determined.

Here, we constructed a mildly alkaline implant interface to test whether local pH modulation could guide neutrophil-mediated antibacterial defense (Scheme 1). This interface was designed to regulate neutrophil immunometabolic activity and promote localized NET formation, thereby enhancing host-driven bacterial clearance without introducing exogenous bactericidal agents. We further examined the associated molecular mechanisms and evaluated the therapeutic performance of this strategy in an infected bone defect model. This work positions mild interfacial alkalinity as a host-responsive design cue for anti-infective bone implants, with the aim of coordinating early infection control and subsequent bone repair.

Scheme 1.

Mild interfacial alkalinity promotes NET formation for infected bone defect repair.

2. Materials and methods

2.1. Synthesis and characterization of the alkaline titanium interface

Ti plates (10 mm × 10 mm × 1 mm, Xi'an Saite Smar Titanium Co., Ltd.) were sequentially ultrasonically cleaned in ultrapure water, ethanol, and deionized water, and then dried. A mixed solution of sodium hydroxide (5 M, Sinopharm Chemical Reagent Co., Ltd.) and hydrogen peroxide (30 wt%, Sinopharm Chemical Reagent Co., Ltd.) was prepared at a volume ratio of 3:1. The titanium samples were placed in a hydrothermal reactor (Shanghai Yiyi Instrument Co., Ltd.), immersed in the prepared solution, and reacted at 80 °C for 24 h.

After treatment, the samples were rinsed thoroughly with ultrapure water and immersed in 0.1 M hydrochloric acid (Sinopharm Chemical Reagent Co., Ltd.) for 2 h to induce ion exchange. The samples were then washed again with ultrapure water, dried, and subsequently subjected to heat treatment in a muffle furnace (Shanghai Guangyi Instrument Co., Ltd.) at a heating rate of 5 °C/min up to 450 °C, followed by holding for 1 h and natural cooling to room temperature. The obtained samples were denoted as Ti-A.

Surface morphology of the samples was characterized using scanning electron microscopy (SEM, FEI). Surface elemental composition and chemical states were analyzed by X-ray photoelectron spectroscopy (XPS, Physical Electronics Inc.). Crystal structure and phase composition were examined using X-ray diffraction (XRD, Rigaku).

Ti and Ti-A samples were individually placed in 12-well plates containing 3 mL of ultrapure water, PBS, or RPMI 1640 supplemented with 10% FBS. For the cell-containing condition, neutrophils were cultured in RPMI 1640 supplemented with 10% FBS in the presence of Ti or Ti-A. Samples were incubated statically at 37 °C for 24 h. The pH was then measured with the electrode positioned approximately 2 mm above the material–liquid interface. Corresponding liquid controls without materials were included, and three independent samples were analyzed for each condition.

Both untreated titanium samples and Ti-A samples were sterilized using ethylene oxide and stored under sterile conditions prior to subsequent experiments.

2.2. Cell viability assay

Rat mesenchymal stem cells (rBMSCs) were harvested, resuspended, and seeded onto Ti and Ti-A samples placed in 24-well plates at a density of 5 × 10⁴ cells per well. After incubation for 3–4 h to allow cell attachment, 500 μL of complete medium consisting of F-12 medium (Gibco, USA), fetal bovine serum (Biological Industries, Israel), and penicillin–streptomycin solution (Gibco, USA) was added, and the cells were cultured at 37 °C under 5% CO₂ in a cell incubator (Thermo Fisher Scientific, USA). Cell viability was evaluated at days 1, 3, and 5 using a CCK-8 assay kit (biosharp, China). Briefly, samples were rinsed twice with phosphate-buffered saline (PBS, Gibco, USA) and incubated with CCK-8 working solution prepared at a 1:10 dilution in serum-free medium for 30 min. Aliquots (100 μL) were transferred to a 96-well plate (Nest, China), and absorbance at 450 nm was measured using a microplate reader.

2.3. Cytoskeleton staining

Cells were cultured on samples for 24 h, fixed with 4% paraformaldehyde (biosharp, China) for 30 min, and permeabilized with 0.1% Triton X-100 for 10 min. After blocking with bovine serum albumin (Boster Biological Technology, China) for 30 min, cells were stained with ActinRed phalloidin probe (KeyGEN BioTECH, China) at a dilution of 1:40 for 20–25 min in the dark. Nuclei were counterstained with DAPI (10 μg/mL) for 7–10 min. Samples were rinsed with PBS and imaged using a fluorescence microscope (Nikon, Japan).

2.4. Flow cytometric analysis

rBMSCs were seeded onto Ti and Ti-A samples and cultured for 24 h. Both adherent and floating cells were collected, centrifuged using a low-speed centrifuge (Xiangyi, China), and washed with PBS (Gibco, USA). Cells were stained using a cell apoptosis detection kit (biosharp, China) containing FITC and PI according to the manufacturer's instructions and incubated for 20 min in the dark. Cell apoptosis was analyzed using a flow cytometer (Beckman Coulter, USA).

2.5. Isolation of mouse neutrophils

Bone marrow-derived neutrophils were isolated from BALB/c mice (4 weeks old) using a mouse bone marrow neutrophil isolation kit (Tianjin Haoyang, China). Briefly, femurs and tibias were harvested under sterile conditions, and bone marrow was flushed with PBS (Gibco, USA) using a syringe. The collected marrow was gently dispersed into a single-cell suspension and filtered through a 70 μm cell strainer (Corning, USA). After centrifugation at 450g for 10 min, red blood cells were removed using erythrocyte lysis buffer.

Neutrophils were further isolated by density gradient centrifugation according to the manufacturer's instructions. Briefly, a discontinuous gradient was prepared, and the cell suspension was layered onto the gradient and centrifuged at 700 g for 30 min. The neutrophil layer was carefully collected, washed, and subjected to additional red blood cell lysis. Cells were then washed with PBS (Gibco, USA) and resuspended in RPMI-1640 medium (Biological Industries, Israel) supplemented with heat-inactivated fetal bovine serum (Biological Industries, Israel) and penicillin–streptomycin (Gibco, USA).

2.6. SYTOX green staining

Neutrophils were seeded onto Ti and Ti-A samples placed in 24-well plates at a density of 1 × 10⁵ cells per well. A PMA-treated group (100 nM, Sigma-Aldrich, USA) was included as a positive control. Cells were incubated at 37 °C under 5% CO₂ for 6 h. After incubation, samples were washed with PBS (Gibco, USA), fixed with 4% paraformaldehyde (biosharp, China) for 30 min, and permeabilized with 0.1% Triton X-100 for 15 min. Cells were then stained with SYTOX Green (1 μM, Thermo Fisher Scientific, USA) for 25 min in the dark. Samples were rinsed with PBS and imaged using a fluorescence microscope (Nikon, Japan).

2.7. Immunofluorescence staining

Neutrophils were seeded and cultured as described above. Cells were fixed with 4% paraformaldehyde for 30 min, permeabilized with 0.1% Triton X-100 for 15 min, and blocked with bovine serum albumin (Boster Biological Technology, China) for 30 min. Samples were incubated with primary antibody against citrullinated histone H3 (H3Cit, 1:500, Abcam, UK) overnight at 4 °C, followed by incubation with Alexa Fluor 594-conjugated secondary antibody (1:500, Abcam, UK) for 1 h at room temperature in the dark. Nuclei were counterstained with DAPI (10 μg/mL) for 10 min. After washing with PBS, samples were observed using a fluorescence microscope (Nikon, Japan).

2.8. MPO–DNA ELISA

Neutrophils were cultured as described above, and the culture supernatants were collected and centrifuged at 400 g for 10 min to remove debris. The levels of MPO–DNA complexes were quantified using an ELISA kit (Enzyme-linked Immunoassay Co., China) according to the manufacturer's instructions. Briefly, standards were prepared at concentrations ranging from 20 to 320 ng/mL. Samples and standards were added to the wells and incubated at 37 °C for 30 min. After washing, enzyme conjugate was added and incubated, followed by substrate development. The reaction was terminated, and absorbance at 450 nm was measured using a microplate reader (BioTek, USA).

2.9. Antibacterial assay and bacterial viability staining

S. aureus suspensions were adjusted to 1 × 10⁸ CFU/mL based on OD₆₀₀. Ti and Ti-A samples were placed in 24-well plates and incubated with bacterial suspension at 37 °C for 6 h. For colony counting, 2 × 10⁵ CFU were added to each sample. After incubation, samples were rinsed with PBS, transferred into PBS, and vortexed to detach adherent bacteria. The resulting suspensions were spread onto agar plates and cultured at 37 °C for 18 h, after which colonies were counted and antibacterial rates were calculated. For bacterial live/dead staining, samples were incubated with 1 × 10⁶ CFU and stained with DMAO/EthD-III according to the manufacturer's instructions, followed by fluorescence imaging.

For neutrophil-mediated antibacterial assays, Ti and Ti-A samples were co-incubated with 2 × 10⁵ neutrophils and 1 × 10⁶ CFU of S. aureus for 6 h. DNase I (5 U/mL, Yeasen, China) or cytochalasin D (10 μg/mL, Takara, Japan) was added 30 min before the end of incubation where indicated. Antibacterial activity and bacterial live/dead staining were then evaluated as described above.

2.10. Metabolomic analysis of neutrophils under material stimulation

Neutrophils isolated from BALB/c mice were seeded onto Ti and Ti-A samples at a density of 1 × 10⁶ cells per sample and incubated for 6 h at 37 °C. After incubation, cells were collected by gentle pipetting, centrifuged at 500 g for 5 min, rapidly frozen in liquid nitrogen, and stored at −80 °C until analysis. For metabolite extraction, samples were thawed on ice and extracted with pre-cooled methanol/acetonitrile/water (2:2:1, v/v). After vortexing and sonication at low temperature, samples were incubated at −20 °C and centrifuged at 14,000 g for 20 min at 4 °C. The supernatants were dried under vacuum and reconstituted in acetonitrile/water (1:1, v/v) prior to analysis. Quality control (QC) samples were prepared by pooling equal volumes of all samples.

Metabolomic profiling was performed using an Agilent 1290 Infinity ultra-high-performance liquid chromatography system coupled to an AB Sciex TripleTOF 6600 mass spectrometer. Chromatographic separation was performed using an ACQUITY UPLC BEH Amide HILIC column (1.7 μm, 2.1 mm × 100 mm; Waters) maintained at 25 °C. The mobile phases consisted of water containing 25 mM ammonium acetate and 25 mM ammonium hydroxide (phase A) and acetonitrile (phase B). The flow rate was 0.5 mL min⁻¹, and the injection volume was 2 μL. Samples were maintained at 4 °C in the autosampler and analyzed in a randomized order. Mass-spectrometric data were acquired in both positive- and negative-ion electrospray ionization modes. The TOF-MS scan range was m/z 60–1000, and the product-ion scan range was m/z 25–1000. MS/MS spectra were acquired using information-dependent acquisition in high-sensitivity mode.

Raw data were converted to mzML format using ProteoWizard and processed using XCMS for peak detection, alignment, retention-time correction, and peak-area extraction. The principal XCMS parameters were set as follows: centWave m/z tolerance, 10 ppm; peak width, 10–60 s; prefilter, c(10, 100); bandwidth, 5; m/z width, 0.025; and minimum fraction, 0.5. Features with more than 50% missing values were excluded, and the remaining missing values were imputed using the K-nearest neighbor method. Extreme values were removed, and the data were normalized to the total peak area before statistical analysis. Metabolites were annotated by matching accurate m/z values, with a mass tolerance of <10 ppm, and MS/MS spectra against an in-house database established using available authentic standards. Differential metabolites were screened by combining multivariate orthogonal partial least-squares discriminant analysis (OPLS-DA) with univariate t-tests. Variable importance in projection (VIP) values were obtained from the OPLS-DA model, and fold change was calculated as the ratio of the mean normalized metabolite abundance in the Ti-A group to that in the Ti group. P values derived from the t-tests were adjusted for multiple comparisons using the Benjamini–Hochberg false discovery rate procedure. Metabolites simultaneously meeting VIP ≥ 1.0, fold change ≥ 2.0 or ≤ 0.5, and adjusted P < 0.05 were considered significantly differential metabolites.

2.11. Intracellular ROS detection in neutrophils

Neutrophils were seeded onto Ti and Ti-A samples and incubated for 6 h. Cells were then incubated with DCFH-DA (1:1000, Beyotime, China) in serum-free medium at 37 °C for 30 min in the dark. After washing with PBS (Gibco, USA), intracellular ROS levels were observed using a fluorescence microscope (Nikon, Japan). For flow cytometric analysis, cells were collected, stained with DCFH-DA under the same conditions, washed, and analyzed using a flow cytometer (Beckman Coulter, USA).

2.12. RT-qPCR analysis

Neutrophils cultured on Ti and Ti-A samples for 6 h were collected, and total RNA was extracted using TRIzol reagent (Thermo Fisher Scientific, USA). RNA purity and concentration were determined using a NanoPhotometer (IMPLEN, Germany). Equal amounts of RNA (1000 ng) were reverse-transcribed into cDNA using a reverse transcription kit (TransGen Biotech, China). Quantitative PCR was performed using SYBR Green Master Mix (TransGen Biotech, China) on a real-time PCR system (Bio-Rad, USA). Gene expression levels of SPHK1 and S1PR2 were normalized to GAPDH and calculated using the 2⁻ΔΔCt method. Primer sequences are listed in Table S1.

2.13. Western blot analysis

Neutrophils were cultured on Ti and Ti-A samples for 6 h and lysed using RIPA buffer. Protein concentrations were determined using a BCA assay kit. Equal amounts of protein were separated by SDS-PAGE and transferred onto PVDF membranes (Millipore, USA). Membranes were blocked with 5% BSA and incubated overnight at 4 °C with primary antibodies against SPHK1 and S1PR2 (1:500, Proteintech, China). After washing, membranes were incubated with HRP-conjugated secondary antibodies (1:500) for 1 h at room temperature. Protein bands were visualized using an enhanced chemiluminescence system.

2.14. Osteogenic differentiation assays

C3H10 cells were seeded onto samples and cultured under osteogenic induction conditions. After 24 h, the medium was replaced with osteogenic differentiation medium, which was refreshed every 3 days.

For alkaline phosphatase (ALP) activity, cells were cultured for 7 days, lysed using RIPA buffer, and centrifuged to collect supernatants. ALP activity was measured using a commercial assay kit according to the manufacturer's instructions, and normalized to total protein content determined by a BCA assay. Absorbance was recorded at 405 nm for ALP activity and 480 nm for protein concentration.

For ALP staining, cells were fixed with 4% paraformaldehyde after 7 days of induction and incubated with BCIP/NBT staining solution in the dark. Stained samples were rinsed with PBS and imaged.

For extracellular matrix mineralization, cells were cultured for 14 days, fixed, and stained with Alizarin Red S solution. After imaging, the bound dye was dissolved using cetylpyridinium chloride (CPC), and absorbance was measured at 562 nm for quantitative analysis.

2.15. In vivo infected bone defect model and evaluation

All animal procedures were approved by the Institutional Animal Care and Use Committee of Guangdong Provincial People's Hospital (KY2023-821-01). Ti and Ti-A implants were sterilized using ethylene oxide and incubated with a Staphylococcus aureus suspension (1 × 10⁷ CFU/mL) for 2 h before implantation. Female Sprague-Dawley rats (4–6 weeks old, approximately 200 g) were used, with an initial sample size of 12 animals in each group. After anesthesia, a cylindrical defect was created in the femoral condyle under sterile conditions. The bacteria-loaded Ti or Ti-A implants were inserted into the defects, followed by wound closure. Animals were sacrificed at 3 days and 2 months after surgery.

For antibacterial evaluation, peri-implant bone tissues were harvested, homogenized in PBS, and subjected to ultrasonic treatment to detach bacteria. The resulting suspensions were plated onto agar plates and incubated at 37 °C for 18 h for colony counting.

For bone regeneration assessment, femurs were collected, fixed in 4% paraformaldehyde, and analyzed by micro-CT (Siemens, Germany). Bone volume fraction (BV/TV) and bone mineral density (BMD) were quantified using VG Studio MAX software. Samples were subsequently decalcified, dehydrated, and embedded for histological analysis. Sections were prepared using a hard tissue grinding system and subjected to Giemsa staining and immunofluorescence staining according to standard protocols.

2.16. Statistical analysis

All experiments were performed with at least three independent samples. Data are presented as mean ± standard deviation (SD). Statistical analysis was conducted using GraphPad Prism 8.0 (GraphPad Software, USA) and SPSS 19.0 (IBM, USA). Differences between groups were evaluated by one-way or two-way analysis of variance (ANOVA) as appropriate. A value of P < 0.05 was considered statistically significant.

3. Results and discussion

3.1. Interfacial features and mildly alkaline microenvironment of Ti-A

To construct an alkaline surface on Ti, an alkali-heat treatment strategy was employed. As shown in Fig. 1A, Ti-A was fabricated through a NaOH/H₂O₂-based hydrothermal reaction on the Ti surface, followed by HCl-mediated ion exchange to remove excess Na⁺-containing alkaline species and stabilize the modified interface, and finally by thermal treatment. SEM (Fig. 1B) showed that pristine Ti exhibited a relatively smooth surface, whereas Ti-A developed a uniform nanoscale porous architecture, indicating that alkali-heat treatment induced pronounced surface reconstruction and generated a more favorable topographical basis for subsequent cell–material interactions. XRD analysis further confirmed these changes (Fig. 1C). Characteristic diffraction peaks corresponding to anatase and rutile phases were detected on Ti-A, indicating the formation of a mixed-phase TiO₂ structure at the interface. This finding suggests that the surface modification extended beyond morphological alteration and was accompanied by stable crystalline phase reconstruction. XPS (Fig. 1D) showed that, in addition to Ti, O, and C, a distinct Na signal was present on Ti-A, reflecting the incorporation of sodium-related species during alkali treatment and the associated change in surface chemical states. To characterize the interfacial pH of Ti-A, measurements were first performed after immersion in ultrapure water (Fig. 1E). Ti-A produced a local pH of approximately 7.88, which was significantly higher than that of pristine Ti and the aqueous control. Additional measurements were subsequently conducted under more biologically relevant conditions. After 24 h of incubation, the interfacial pH of Ti-A remained approximately 7.65 in PBS and approximately 7.60 in RPMI 1640 containing 10% FBS, both with and without neutrophils, and was significantly higher than that of the corresponding medium and Ti controls (Fig. S1–S3). Time-course measurements in RPMI 1640 containing 10% FBS further showed that the pH adjacent to Ti-A gradually decreased from approximately 7.78 at the initial measurement to approximately 7.49 at 48 h but remained within a mildly alkaline range throughout the observation period (Fig. S4). These results demonstrate that physiological buffering attenuated the alkalinizing effect observed in ultrapure water but did not abolish the ability of Ti-A to maintain a sustained, mildly alkaline interfacial environment.

Fig. 1.

Material characterization of Ti-A and its pH value. (A) Schematic illustration of the fabrication process of the Ti-A alkaline interface; (B) scanning electron microscopy images of Ti-A (1000×, scale bar = 20 μm; 5000×, scale bar = 4 μm); (C) X-ray diffraction analysis; (D) X-ray photoelectron spectroscopy analysis; (E) pH measured at 2 mm above the Ti-A surface; n = 3, ****P < 0.0001.

Taken together, these results show that alkali–heat treatment reconstructed the Ti surface at both the topographical and chemical levels. The uniform nanoporous architecture may increase the available contact area and influence initial protein adsorption, cell adhesion, and cell–material interactions. Recent studies have further demonstrated that the roughness, hydrophilicity, and nanoscale morphology of titanium surfaces can regulate neutrophil adhesion, activation, ROS production, and NET formation [31,32]. The mixed anatase/rutile phases and sodium-containing surface species may also contribute to the bioactivity of Ti-A, as alkali–heat-treated titanium surfaces containing nanostructured sodium titanate/TiO₂ layers have been reported to promote apatite formation and osseointegration [33,34]. Among these interfacial characteristics, the mildly alkaline environment is particularly relevant to the immunomodulatory purpose of the present study. Increasing evidence suggests that a moderately alkaline environment can support neutrophil activation and several NET-related processes, whereas acidic conditions tend to restrict ROS generation and NET formation [35]. Notably, high extracellular Na⁺ has been reported to suppress phagocyte oxidase-dependent neutrophil activity [36], indicating that alkalinity and sodium-associated effects should not be regarded as equivalent. These physicochemical characteristics provide the material basis for further investigating how Ti-A regulates neutrophil-mediated antibacterial defense.

3.2. Cytocompatible Ti-A interfaces elicit neutrophil NET-forming responses

Before investigating Ti-A-induced NET formation, the cytocompatibility of the mildly alkaline Ti-A interface was evaluated in both rBMSCs and neutrophils. Live/dead staining showed that rBMSCs on Ti and Ti-A were predominantly calcein-AM-positive, with only a few PI-positive dead cells. Although the number of adherent cells appeared lower on Ti-A than on pristine Ti, neither widespread PI staining nor evident membrane damage was observed (Fig. S5). Consistently, rBMSCs on Ti-A maintained well-organized F-actin cytoskeletons and intact nuclear morphology, without obvious cytoskeletal collapse or abnormal cell shape (Fig. 2A). Flow cytometry further showed that Ti-A did not markedly increase the proportion of apoptotic or dead rBMSCs (Fig. 2B). CCK-8 analysis revealed a modest but significant reduction in the viability signal of rBMSCs cultured on Ti-A at different time points; nevertheless, their relative metabolic activity remained approximately 92% of that observed on Ti, indicating that the decrease was limited in magnitude (Fig. 2C). Because the subsequent NET-related experiments were performed using neutrophils, their compatibility with Ti-A was also examined under the same 6-h exposure conditions. Neutrophils exposed to Ti-A retained approximately 89% of the CCK-8 signal measured in the Ti group, demonstrating that Ti-A did not cause extensive loss of neutrophil metabolic activity during the NET-induction period (Fig. S6).

Fig. 2.

Cytocompatibility assessment of Ti-A and neutrophil NETs-related analysis. (A) Phalloidin/DAPI staining images of rBMSCs cultured on Ti and Ti-A surfaces, showing F-actin cytoskeleton, nuclei, and merged images. (B) Flow cytometry analysis of rBMSCs cultured on Ti and Ti-A surfaces. (C) Cell activity of rBMSCs cultured on Ti and Ti-A surfaces for 1, 3, and 5 days. (D) SYTOX Green/DAPI staining images of neutrophils in the Ti, Ti-A, and PMA groups with or without DNase I treatment. (E) Quantification of NETs area under different treatment conditions. (F) Quantification of NETs-positive proportion under different treatment conditions. (G) H3Cit/DAPI immunofluorescence staining images of neutrophils in the Ti, Ti-A, and PMA groups with or without DNase I treatment. (H) ELISA analysis of MPO-DNA complex levels in each group without DNase I treatment. (I) ELISA analysis of MPO-DNA complex levels in each group with DNase I treatment. PMA was used as a positive NET-inducing control. Data are presented as mean ± SD; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Scale bar = 100 μm.

We next assessed NET formation from neutrophils seeded on the Ti-A interface. SYTOX Green staining revealed more extracellular DNA-positive structures on Ti-A than on pristine Ti, indicating enhanced extracellular chromatin release from neutrophils exposed to the Ti-A interface. PMA, used as a positive NET-inducing control, produced pronounced extracellular DNA networks (Fig. 2D). Because SYTOX Green labels extracellular DNA but does not by itself distinguish active NET formation from nonspecific DNA leakage, DNase I was introduced to digest extracellular DNA and test whether these structures were dependent on a DNA scaffold [37]. After DNase I treatment, the visible DNA networks in all groups were markedly reduced, accompanied by decreases in both NETs area and NETs-positive proportion (%) (Fig. 2D–F). This DNase sensitivity indicates that the Ti-A-induced extracellular structures were DNA scaffold-dependent, rather than nonspecific fluorescent debris or background signals. We further examined citrullinated H3Cit, a chromatin modification closely associated with NET formation [38]. Ti-A markedly increased H3Cit signals that colocalized with DAPI-positive extracellular chromatin regions, whereas DNase I weakened the extracellular network-like staining pattern (Fig. 2G). In parallel, NETs are typically composed of an extracellular DNA scaffold decorated with neutrophil-derived granule proteins, including MPO [39]. Therefore, MPO-DNA complexes were further measured as a biochemical indicator of NET formation. MPO-DNA ELISA showed higher complex levels in the Ti-A group than in the Ti group, further supporting that Ti-A promoted NET-associated co-release of extracellular chromatin and granule proteins (Fig. 2H and I). Together, these results indicate that Ti-A promotes bona fide NET formation rather than nonspecific extracellular DNA release, supporting mild interfacial alkalinity as an immune-active cue for neutrophil-mediated antibacterial defense.

3.3. NETs-dependent neutrophil antibacterial defense on Ti-A

To evaluate Ti-A-enhanced neutrophil-mediated antibacterial activity, neutrophils and S. aureus were co-incubated on Ti or Ti-A surfaces for 6 h, followed by colony counting and bacterial live/dead staining to assess bacterial survival at the interface (Fig. 3A). Plate counting showed that Ti-A had markedly fewer colonies than pristine Ti across all tested conditions (Fig. 3B and C). Even in the absence of neutrophils, Ti-A reduced bacterial survival on the material surface, demonstrating the intrinsic antibacterial activity of the modified interface. Following the addition of neutrophils, colony formation decreased on both Ti and Ti-A; notably, Ti-A achieved a substantially higher antibacterial rate than pristine Ti, reaching approximately 86.67% compared with 47.50%. This difference suggests that Ti-A enhanced neutrophil-associated antibacterial activity while retaining its intrinsic surface antibacterial effect. Consistent with the colony-counting results, bacterial live/dead staining showed predominantly green signals on Ti, whereas red dead-bacterial signals increased on Ti-A, particularly in the presence of neutrophils (Fig. 3D). These results indicate that the antibacterial advantage of Ti-A arises not only from direct suppression of bacterial survival at the material interface, but also from an enhanced neutrophil-mediated antibacterial response.

Fig. 3.

Neutrophil-assisted antibacterial assessment on Ti-A surfaces. (A) Schematic illustration of the antibacterial assay in which neutrophils and S. aureus were co-incubated on Ti or Ti-A surfaces. (B) Representative agar plate images of bacteria recovered from Ti and Ti-A surfaces under conditions without neutrophils, with neutrophils, with DNase I treatment, or with Cytochalasin D treatment. (C) Quantification of bacterial colony counts on Ti and Ti-A surfaces under different treatment conditions. (D) Bacterial live/dead staining images on Ti and Ti-A surfaces under different treatment conditions. (E) Quantification of the antibacterial rate on Ti-A surfaces under different treatment conditions. NE indicates neutrophils; S.a indicates S. aureus; DNase I was used to degrade extracellular DNA; Cytochalasin D was used to inhibit actin-dependent phagocytosis. Data are presented as mean ± SD; ***P < 0.001, ****P < 0.0001.

To distinguish the contribution of NETs from actin-dependent phagocytosis, DNase I and Cytochalasin D treatments were applied on both Ti and Ti-A surfaces. DNase I was used to degrade the extracellular DNA backbone of NETs and disrupt NETs structures, whereas Cytochalasin D was used to inhibit actin-dependent phagocytosis [40,41]. After DNase I treatment, colony formation on Ti-A markedly increased, and the antibacterial rate decreased from the high level observed with neutrophils to a level close to that of Ti-A alone. This indicates that NET-associated extracellular DNA structures were a major contributor to the Ti-A-enhanced neutrophil antibacterial effect. By contrast, Cytochalasin D also reduced the antibacterial rate on Ti-A, but the rate remained higher than that of Ti-A alone, suggesting that phagocytosis was involved but did not fully account for the neutrophil-mediated antibacterial advantage of Ti-A (Fig. 3B–E). Therefore, Beyond directly restricting bacterial survival, Ti-A further enhanced neutrophil-mediated antibacterial activity against S. aureus, with NET-associated extracellular killing making a major contribution.

3.4. SPHK1-S1PR2/ROS signaling in Ti-A-induced NET formation

After showing that Ti-A promoted neutrophil extracellular trap formation and antibacterial activity, we examined how this interfacial cue was translated into neutrophil functional responses at the metabolic level. Non-targeted metabolomic profiling was performed on neutrophils stimulated on Ti or Ti-A surfaces. The global metabolite distributions showed no obvious outliers, suggesting no obvious abnormality in the overall sample distribution (Fig. 4A). Volcano plot analysis revealed clear metabolic differences between the Ti and Ti-A groups (Fig. 4B). Among the differential metabolites, 77 were upregulated and 7 were downregulated in the Ti-A group (Fig. S7), indicating that Ti-A stimulation induced a marked shift in the neutrophil metabolic profile.

Fig. 4.

Non-targeted metabolomic analysis of neutrophils after Ti-A stimulation. (A) Violin plots showing the global metabolite distribution of individual samples in the Ti and Ti-A groups. (B) Volcano plot of differential metabolites between the Ti and Ti-A groups. (C) Clustered heatmap of representative differential metabolites in the Ti and Ti-A groups. (D) VIP analysis of differential metabolites. (E) KEGG top 20 enrichment analysis of differential metabolites. (F) KEGG pathway classification and enrichment distribution of differential metabolites.

Analysis of representative differential metabolites showed a clear separation between the Ti and Ti-A groups in several lipid-related metabolites. Heatmap analysis showed increased levels of sphingosine and N-palmitoyl-D-sphingosine in the Ti-A group (Fig. 4C). VIP analysis also identified sphingosine and N-palmitoyl-D-sphingosine as important metabolites distinguishing Ti-A-stimulated neutrophils from Ti controls (Fig. 4D), and the z-score distribution further confirmed their overall elevation in the Ti-A group (Fig. S5). Since sphingosine is a key intermediate in sphingolipid metabolism and can be phosphorylated by SPHK1 to generate S1P, which signals through S1P receptors, these changes suggested a connection between the Ti-A-induced neutrophil response and sphingolipid remodeling [42]. Notably, 6-phosphogluconic acid was also increased in the Ti-A group (Fig. 4D–S5). As a metabolite associated with the oxidative pentose phosphate pathway, its elevation may be related to NADPH supply and subsequent ROS generation, providing a metabolic clue to ROS changes during Ti-A-induced NET formation [43,44].

KEGG enrichment analysis further supported this metabolic direction. Sphingolipid metabolism and the sphingolipid signaling pathway both ranked among the top 20 enriched pathways, together with other lipid-related pathways such as glycerophospholipid metabolism (Fig. 4E). A supplementary KEGG bar plot showed a similar enrichment pattern, with sphingolipid metabolism ranking prominently alongside sphingolipid signaling (Fig. S8). Pathway annotation showed that the differential metabolites were mainly assigned to the metabolism category, with lipid metabolism accounting for a substantial proportion; energy metabolism-related entries were also observed (Fig. S9). Together with the changes in sphingosine, N-palmitoyl-D-sphingosine, and 6-phosphogluconic acid, these data indicate that Ti-A-stimulated neutrophils underwent metabolic changes centered on lipid metabolism, particularly sphingolipid-related processes, accompanied by metabolic features potentially linked to ROS generation.

These metabolomic changes pointed to a potential link between sphingolipid remodeling and ROS production. The increased levels of sphingosine and N-palmitoyl-D-sphingosine, together with the enrichment of sphingolipid metabolism and sphingolipid signaling pathways, indicated that Ti-A altered the sphingolipid metabolic state of neutrophils. Meanwhile, the elevation of 6-phosphogluconic acid was consistent with the possible involvement of the oxidative pentose phosphate pathway, which supplies NADPH for cellular ROS production. Mechanistically, sphingosine can be phosphorylated by SPHK1 to generate S1P, which subsequently signals through S1P receptors, including S1PR2, and may engage Ca²⁺- and MAPK-related pathways associated with PAD4 activation, histone H3 citrullination, chromatin decondensation, and NET formation [[45], [46], [47]]. Together, these metabolic features provided the basis for examining the involvement of the SPHK1–S1PR2/ROS axis in the neutrophil response to Ti-A.

Based on these metabolic clues, the proposed SPHK1–S1PR2/ROS axis was examined at the levels of intracellular oxidative activity and pathway-related gene and protein expression. DCFH-DA fluorescence imaging showed that neutrophils exposed to Ti-A exhibited a markedly stronger intracellular oxidative signal than those exposed to pristine Ti, while PMA also induced an evident oxidative response as a positive NET-inducing stimulus (Fig. 5A). Flow cytometry revealed a corresponding rightward shift in DCF fluorescence, and quantitative analysis confirmed that the oxidative signal was significantly increased in the Ti-A group (Fig. 5B and C). ROS has been reported to facilitate NET formation by promoting oxidative DNA damage, activating DNA-repair responses, and driving the subsequent chromatin remodeling required for extracellular DNA release [48,49]. Analysis of the sphingolipid-related components of the proposed pathway showed that the mRNA levels of SPHK1 and S1PR2 were both significantly elevated in Ti-A-stimulated neutrophils compared with the Ti and PMA groups (Fig. 5D). Western blotting confirmed the same expression pattern at the protein level (Fig. 5E). Notably, PMA enhanced intracellular oxidative activity without inducing a comparable increase in SPHK1 or S1PR2 expression. The different responses to Ti-A and PMA indicate that SPHK1/S1PR2 upregulation is not simply a general consequence of NET induction but is more closely associated with the sphingolipid remodeling induced by Ti-A.

Fig. 5.

Ti-A promotes NET formation through SPHK1/S1PR2-associated ROS signaling in neutrophils. (A) Representative DCFH-DA fluorescence images showing intracellular oxidative activity in neutrophils from the Ti, Ti-A, and PMA groups. (B) Representative flow-cytometry histograms of DCFH-DA fluorescence in the Ti, Ti-A, and PMA groups. (C) Quantitative analysis of DCFH-DA fluorescence under the indicated conditions. (D) RT-qPCR analysis of SPHK1 and S1PR2 mRNA expression in neutrophils from the Ti, Ti-A, and PMA groups. (E) Western blot analysis of SPHK1 and S1PR2 protein expression, with GAPDH used as the loading control. (F) Representative SYTOX Green staining of extracellular DNA/NET structures (upper panels) and DCFH-DA fluorescence images of intracellular oxidative activity (lower panels) in the Ti, Ti-A, Ti-A + PF-543, and Ti-A + DPI groups. (G) Quantification of NET formation under the indicated conditions. (H) Representative flow-cytometry histograms of intracellular DCFH-DA fluorescence in the untreated control, Ti, Ti-A, Ti-A + PF-543, and Ti-A + DPI groups. (I) Quantitative analysis of ROS-associated mean fluorescence intensity (MFI). (J) Schematic illustration of the proposed mechanism by which Ti-A-induced sphingolipid remodeling and SPHK1/S1PR2-associated ROS signaling promote NET formation. PMA was used as a positive NET-inducing control; PF-543 was used as a selective SPHK1 inhibitor, and DPI was used to suppress NADPH oxidase-dependent ROS production. Scale bars: 100 μm. Data are presented as mean ± SD; *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

The functional roles of SPHK1 and ROS were further examined by pharmacological blockade. PF-543, a selective inhibitor of SPHK1 that limits the phosphorylation of sphingosine to S1P, was used to interfere with the SPHK1-dependent branch of the pathway. DPI, an inhibitor of flavoprotein-dependent oxidases commonly used to suppress NADPH oxidase-derived ROS production, was applied to examine the contribution of the oxidative response. SYTOX Green staining showed abundant extracellular, web-like DNA structures in the Ti-A group, whereas treatment with either PF-543 or DPI markedly reduced the appearance of these structures (Fig. 5F). Quantitative analysis confirmed that Ti-A significantly increased NET formation relative to pristine Ti and that this increase was significantly attenuated by both inhibitors (Fig. 5G). DCFH-DA fluorescence imaging further showed that PF-543 and DPI each reduced the enhanced intracellular oxidative signal induced by Ti-A (Fig. 5F). Flow cytometry and MFI analysis yielded consistent results, with both inhibitor-treated groups displaying significantly lower ROS-associated fluorescence than the Ti-A group (Fig. 5H and I). The reduction in ROS after PF-543 treatment indicates that SPHK1 activity contributes to the oxidative response upstream of NET release, whereas the simultaneous suppression of ROS and NET formation by DPI confirms the functional importance of ROS generation. Together, these pharmacological results support the involvement of an SPHK1-associated, ROS-dependent mechanism in Ti-A-induced NET formation.

When considered together with the metabolomic results, these findings support a model in which Ti-A-induced sphingolipid remodeling engages SPHK1/S1PR2-associated signaling and converges on ROS-dependent NET formation (Fig. 5J). Recent experimental evidence provides independent support for this interpretation. Yang et al. demonstrated that S1P activated S1PR2-dependent Raf/MEK/ERK signaling in neutrophils and that PF-543 reduced S1P production, S1PR2 expression, and NET formation [50]. Pan et al. likewise reported an association among SPHK1, S1P, S1PR2, and H3Cit, and found that pharmacological inhibition of S1PR2 reduced NET accumulation in vivo [51]. In primary human neutrophils, MEK and p38 MAPK have been shown to regulate early PAD4-dependent histone citrullination during NET formation [52], whereas recent chromatin-accessibility analysis has linked PMA-induced NET formation to PKC/NOX-dependent ROS production and subsequent chromatin remodeling [53]. A separate study further demonstrated a DPI-sensitive NOX–ROS–PAD4 pathway during stimulus-induced NET formation [54]. In the present study, the selective induction of SPHK1 and S1PR2 by Ti-A, together with the reduction of both ROS and NET formation by PF-543 and DPI, connects the observed metabolic remodeling with a functional oxidative response. Because neither inhibitor completely abolished NET formation, this pathway is likely to make a substantial contribution while operating alongside other signaling events initiated at the Ti-A–neutrophil interface.

3.5. Ti-A sustains NET-associated host defense in vivo

To evaluate the in vivo antibacterial performance of Ti-A in an implant-associated infection setting, Ti and Ti-A implants were preloaded with S. aureus and inserted into femoral condyle defects in rats (Fig. 6A). At 3 days after implantation, fewer colonies were recovered from peri-implant bone tissues and implant-associated samples in the Ti-A group than in the Ti group. Giemsa staining also showed reduced bacteria-related signals around the implant interface in the Ti-A group (Fig. S11 and S12). At 2 months, colony formation and CFU counts from bone tissues remained significantly lower in the Ti-A group, and Giemsa-stained sections showed markedly fewer bacterial residues within the defect region and around the implant interface (Fig. 6B–D). These results indicate that Ti-A persistently restricted local bacterial colonization in the bacteria-loaded implant-induced bone defect infection model.

Fig. 6.

In vivo antibacterial activity and NET-associated host defense response of Ti-A in implant-associated bone defect infection. (A) Schematic illustration of the bacteria-loaded implant-induced femoral condyle defect infection model in rats. (B) Representative bacterial culture images of peri-implant bone tissues (left) and implant-associated samples (right) from the Ti and Ti-A groups at 2 months post-surgery. (C) Giemsa staining images of the defect region and implant interface in the Ti and Ti-A groups at 2 months post-surgery. (D) Quantitative analysis of bacterial CFU recovered from peri-implant bone tissues in the Ti and Ti-A groups at 2 months post-surgery. (E–L) Immunofluorescence staining images and corresponding mean fluorescence intensity quantification of Ly6G, H3Cit, SPHK1, and S1PR2 in the defect region of the Ti and Ti-A groups at 3 days post-surgery. Scale bar:100 μm. Data are presented as mean ± SD; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

The sustained reduction in bacterial burden prompted further examination of the early immune events at the implant interface; peri-implant tissues collected on postoperative day 3 were therefore analyzed for neutrophil recruitment and NET-associated markers. Ly6G-positive staining was stronger around the Ti-A interface than around the Ti interface, and quantitative analysis confirmed a significant increase in mean fluorescence intensity (Fig. 6E and F). Because this response was accompanied by a lower bacterial burden at the same early time point, the increased Ly6G signal is consistent with the involvement of recruited neutrophils in antibacterial host defense at the infected interface. H3Cit staining was also enhanced in the Ti-A group, indicating increased histone citrullination associated with NET formation (Fig. 6G and H). In parallel, SPHK1 and S1PR2 signals were higher in the Ti-A group than in the Ti group (Fig. 6I–L), in agreement with the in vitro finding that Ti-A increased SPHK1 and S1PR2 expression in neutrophils. Although immunofluorescence alone cannot establish a direct causal role of SPHK1/S1PR2 in NET formation, the concurrent increases in Ly6G, H3Cit, SPHK1, and S1PR2 signals, together with the reduced bacterial burden, support an association between the in vivo antibacterial effect of Ti-A and neutrophil-mediated, NET-associated host defense involving SPHK1/S1PR2 signaling. Collectively, these day-3 findings indicate that the inhibition of bacterial colonization by Ti-A was accompanied by an enhanced early neutrophil and NET-associated immune response at the infected implant interface.

These early immune findings provide a possible biological link between the initial antibacterial response and the sustained reduction in bacterial burden observed at 2 months. Neutrophil recruitment and NET formation occur predominantly during the initial inflammatory phase after biomaterial implantation, and peri-implant NET-associated responses have been shown to change substantially within the first several days following titanium implantation [32]. During bacterial infection, NETs can restrict pathogen dissemination and enhance cooperative bacterial clearance by macrophages [55]. The increased Ly6G and H3Cit signals observed on postoperative day 3 therefore suggest that Ti-A facilitates early bacterial containment at the implant interface, which may reduce persistent infectious stimulation and provide a more favorable environment for subsequent inflammation resolution and bone repair. Further time-resolved studies will be required to define the transition from early NET-associated antibacterial defense to later tissue regeneration.

3.6. Ti-A links infection control to bone repair

For anti-infective bone implants, bacterial clearance is only meaningful if the interface remains compatible with subsequent bone repair. We therefore examined whether Ti-A, after eliciting NET-associated antibacterial responses, could still support osteogenic activity. During in vitro osteogenic induction, ALP staining was stronger on Ti-A than on pristine Ti, accompanied by higher ALP quantification and enzymatic activity (Fig. 7A and B). ARS staining showed more extensive mineral deposition on Ti-A, indicating enhanced matrix mineralization (Fig. 7A). At the gene level, Ti-A increased Runx2 and Col1a1 expression, while ALP and OPN showed time-dependent changes with clearer upregulation at the later stage of induction (Fig. 7C). These findings show that the mildly alkaline Ti-A interface did not trade immune-antibacterial activity for impaired osteogenic function, but instead preserved the osteogenic differentiation and mineralization capacity of C3H10.

Fig. 7.

Evaluation of Ti-A-mediated infection control and bone defect repair. (A) ALP and ARS staining images of C3H10 after osteogenic induction in the Ti and Ti-A groups. (B) ALP quantification and ALP activity analysis of C3H10 after osteogenic induction in the Ti and Ti-A groups. (C) mRNA expression levels of OPN, Runx2, ALP, and Col1a1 in C3H10 after osteogenic induction in the Ti and Ti-A groups. (D) Micro-CT three-dimensional reconstruction images of the femoral condyle defect region in the Ti and Ti-A groups at 2 months post-surgery. (E,F) Quantitative analysis of BV/TV and BMD in the Ti and Ti-A groups at 2 months post-surgery. H&E (G), VG (H), and OPN immunohistochemical (I) staining images of the defect region in the Ti and Ti-A groups at 2 months post-surgery. Scale bar:100 μm. Data are presented as mean ± SD; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

We further evaluated bone repair outcomes under Ti-A-mediated infection control in vivo. Micro-CT reconstruction showed more newly formed bone within the defect region of the Ti-A group, with a more complete filling pattern than that observed in the Ti group (Fig. 7D). BV/TV and BMD were both significantly higher in the Ti-A group, supporting improved bone formation and mineralization under infected conditions (Fig. 7E and F). Histological staining showed the same trend. H&E sections from the Ti group displayed more evident tissue disruption and discontinuous repair, whereas the Ti-A group showed a more integrated local tissue structure (Fig. 7G). Early sections also showed heavier inflammatory cell infiltration around Ti, while Ti-A was associated with a milder local tissue reaction (Fig. S13). VG staining revealed richer bone matrix and collagen deposition in the Ti-A group (Fig. 7H), and OPN immunohistochemistry showed stronger osteogenesis-related positive staining within the defect region (Fig. 7I). These observations suggest that Ti-A created a local environment in which bone repair could proceed despite the initial bacterial challenge.

H&E staining of major organs showed no obvious pathological damage or structural abnormalities in the heart, liver, spleen, lung, or kidney in either group, indicating no evident systemic toxicity after Ti-A implantation (Fig. S14). Taken together, these findings connect the early antibacterial response at the Ti-A interface with the eventual regenerative outcome. Ti-A enhanced NET-associated host defense and reduced persistent bacterial colonization while maintaining cellular compatibility and supporting osteogenic differentiation and matrix mineralization. The improved bone repair observed in vivo should therefore be understood as the combined consequence of early infection control, modulation of the local inflammatory environment, and preservation of the osteogenic capacity required for defect reconstruction, rather than as the result of a single antibacterial or osteogenic mechanism.

Bone regeneration in an infected defect is closely coupled to the progression and resolution of the local immune response. Persistent bacterial colonization diverts the host response from tissue repair toward sustained antibacterial inflammation, disrupts osteoblast–osteoclast homeostasis, and ultimately compromises bone healing [56]. Consequently, successful repair requires not only the removal of bacteria but also the restoration of an immune microenvironment permissive to osteogenesis [57]. In the present study, the lower bacterial burden and milder inflammatory infiltration around Ti-A were accompanied by increased BV/TV, BMD, bone matrix deposition, and OPN expression. At the same time, the enhanced ALP activity, osteogenic gene expression, and mineral deposition observed in vitro indicate that the modified interface itself also supported osteogenic differentiation. Thus, the regenerative advantage of Ti-A most likely arose from the convergence of its direct compatibility with osteogenic cells and the indirect benefit of reducing infection-driven inflammatory interference, rather than from either effect operating independently.

The biological consequences of NET formation depend strongly on its timing, magnitude, and clearance. During the early stage of infection, NETs can immobilize bacteria, concentrate antimicrobial proteins, and assist macrophage-mediated bacterial killing [58]. If NETs persist or accumulate excessively, however, their DNA, histones, and granular enzymes may sustain inflammation and damage surrounding tissue [59]. This distinction is important in the setting of titanium implantation, where reducing NET formation under sterile conditions has been shown to accelerate inflammatory resolution and increase peri-implant bone formation [32]. In the present infected model, increased Ly6G and H3Cit signals were observed locally on postoperative day 3, together with an early decrease in bacterial burden. The later findings of lower bacterial colonization, reduced inflammatory infiltration, and improved bone repair are compatible with a beneficial early antibacterial response. The current experiments did not follow NET formation and clearance continuously, and therefore do not establish whether or when this response subsided. Serial measurements at early and intermediate time points will be needed to determine how the NET response is terminated as the interface shifts from infection control toward tissue repair.

4. Conclusion

In this work, we developed Ti-A as a mildly alkaline titanium interface that redirects implant-associated infection control from direct material-mediated killing toward neutrophil-centered host defense. Ti-A promoted NET formation, enhanced neutrophil-mediated antibacterial activity against S. aureus, and reduced bacterial colonization in a bacteria-loaded implant model. Mechanistically, metabolomic and molecular analyses suggested that sphingolipid remodeling may converge on enhanced SPHK1-S1PR2/ROS signaling during Ti-A-induced NET formation. Importantly, this immune-mediated antibacterial response remained compatible with osteogenesis and supported infected bone defect repair without evident systemic toxicity. These properties suggest that Ti-A may have particular translational value for patients with established implant-associated infections or infected bone defects, where infection control and bone reconstruction need to be achieved concurrently. Collectively, this work highlights interfacial mild alkalinity as a bioactive physicochemical signal capable of regulating neutrophil sphingolipid immunometabolism and provides a host-directed strategy for coordinating infection control and bone regeneration in orthopedic implants.

CRediT authorship contribution statement

Gaoquan Zheng: Writing – original draft, Validation, Investigation, Formal analysis, Data curation, Conceptualization. Dianqing Li: Writing – original draft, Visualization, Validation, Methodology, Investigation, Formal analysis, Data curation. Dengwen Zheng: Validation, Methodology, Investigation, Data curation. Yupeng Liu: Validation, Methodology, Investigation. Yaxi He: Methodology, Investigation, Data curation. Jialin Jiang: Visualization, Validation, Investigation. Dongdong Zhang: Visualization, Funding acquisition, Formal analysis. Xiurui Zhang: Resources, Investigation, Funding acquisition. Liping Tong: Software, Funding acquisition, Formal analysis. Feng Peng: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition, Conceptualization. Xiang Yu: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition, Conceptualization. Mei Li: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition, Conceptualization.

Declaration of generative AI and AI-assisted technologies in the manuscript preparation process

During the preparation of this work the authors used ChatGPT 5.5 in order to draw Fig. 1A. After using this tool, the authors reviewed and edited the content as needed and takes full responsibility for the content of the published article.

Declaration of competing interests

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

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Data availability

Data will be made available on request.