Guanidine-quaternary ammonium polymer-derived carbon dots differentially modulate bacterial respiration and osteoblast mitochondria to resolve infection and foster osteointegration Authors: Xinxin Bai, Anhui Qin, Zhenguo Wang, Guang Kong, Menghan Zhang, Hongyu Li, Rui Zhao, Yue Yu, Yongfeng Zhang, Qi Liu, Ziqing Zhu, Bing Xia, Shengyou Li, Yiming Hao, Mingze Qin, Dong Lei, Lingli Guo, Meijia Su, Zhuojing Luo, Shaohuang Weng, Teng Ma, Jinghui Huang Published: 2027 ABSTRACT Implant-associated infections (IAIs) represent one of the most challenging complications following orthopedic implantation. Current antibacterial strategies struggle to achieve complete pathogen eradication while maintaining sustained osteogenic activity at the implant-bone interface. Leveraging the spatial divergence in energy metabolism between prokaryotic and eukaryotic cells, we designed and synthesized guanidine-quaternary ammonium polymer-derived carbon dots (GQCDs) with confirmed structure. Under non-infected conditions, GQCDs enhance mitochondrial metabolism to promote bone regeneration. In infected microenvironments, GQCDs not only selectively target bacterial membranes and inhibit the respiratory chain to exert potent antibacterial effects, but also downregulate host inflammatory responses and restore mitochondrial function in osteoblasts, thereby creating a favorable microenvironment for bone repair. GQCDs can be stably integrated onto various clinical implant substrates via a coating process. GQCDs-coated implants possessed the dual antibacterial and osteogenic functionalities, which were validated using in vitro co-culture systems and in vivo implant infection models in rats and beagles, offering a promising strategy with encouraging translational prospects to address IAIs. Introduction Implant-associated infections (IAIs) represent a formidable complication following orthopedic implantation [1,2]. Although implants are designed to provide mechanical support and promote osteointegration, their surfaces can inadvertently serve as a nidus for pathogen colonization, culminating in biofilm formation and persistent infection even from minimal microbial contamination [[3], [4], [5]]. The establishment of this biofilm, compounded by a locally immunosuppressive microenvironment, renders IAIs particularly recalcitrant to treatment [6]. Therefore, surface modification of implants to confer durable antibacterial activity is a critical strategy. Epidemiological data identify six principal pathogens responsible for IAIs: Staphylococcus epidermidis (S. epidermidis), Staphylococcus aureus (S. aureus), methicillin-resistant Staphylococcus aureus (MRSA), Pseudomonas aeruginosa (P. aeruginosa), Escherichia coli (E. coli), and Enterobacter cloacae (E. cloacae) [7,8]. Among these, MRSA is particularly notorious. Its high virulence, multidrug resistance, and strong biofilm-forming capacity enable it to establish persistent, recurrent chronic infections [9]. Such infections often necessitate high-dose systemic antibiotic therapy. This carries a significant risk of severe adverse effects and highlights the urgent need for broad-spectrum antibacterial strategies [10]. Despite advances in antimicrobial technologies, a central challenge remains: achieving complete pathogen eradication while simultaneously promoting sustained osteogenic activity at the bone-implant interface. Existing approaches have yet to adequately resolve this “antibacterial-osteogenic paradox”. They often fail to address these dual objectives through mechanistically distinct yet compatible pathways [11,12]. A largely unexploited therapeutic window lies in the fundamental evolutionary divergence of energy metabolism between prokaryotes (bacteria) and eukaryotes (host cells). Prokaryotes generate ATP via the electron transport chain (ETC) located on their cytoplasmic membrane [13,14]. In contrast, eukaryotes compartmentalize this core bioenergetic process within mitochondria [15]. This spatial and functional dichotomy provides a theoretical foundation for developing dual-target agents capable of selectively disrupting bacterial bioenergetics while enhancing mitochondrial function in host cells to promote tissue repair. Currently, most antibacterial strategies for implants containing modifications utilize antibiotics, which do not promote bone regeneration directly and can induce host cell toxicity and antibiotic resistance when retained in the bone tissue [16,17]. The growing prevalence of antimicrobial resistance underscores the urgent need for alternative therapeutic strategies [18]. Emerging antibacterial approaches, such as polyphenolamine-based chemical sterilizers [19] and electroactive compounds [20], have partially overcome resistance but still lack osteogenic enhancement. Therefore, the design and synthesis of dual-targeted antibacterial agents that harness the energy metabolism differences between eukaryotes and prokaryotes are urgently needed to promote bone regeneration during infection [21]. Recently, as an emerging class of zero-dimensional carbon-based nanomaterials, functional carbon dots (CDs) with unique size effects and charge characteristics have demonstrated immense potential in the management of implant-associated infections. Inherently, they not only directly compromise bacterial membrane integrity via mechanical stress but also offer extensive opportunities for surface functionalization. Leveraging their highly tunable surface chemistry, we rationally designed functionalized CDs by covalently grafting guanidino and quaternary ammonium groups onto the carbon surface via a facile one-step hydrothermal process. This strategy enables the precise modulation of surface positive charge density, endowing the CDs with a dual-targeting mechanism: they can strongly anchor to negatively charged bacterial membranes via electrostatic interactions while concomitantly responding to the mitochondrial membrane potential gradient (ΔΨm), thereby exhibiting a highly flexible “modular” adaptability. Furthermore, while traditional antimicrobial agents are prone to deactivation within the complex implant infection microenvironment [22], the inherent rigid carbon core of CDs confers extraordinary physicochemical stability and chemical inertness, effectively resisting microenvironmental degradation to ensure long-lasting bioactivity. Consequently, the synergistic integration of size matching, surface programmability, and structural stability affords CDs a unique advantage in constructing a smart-responsive antibacterial-tissue repair integrated platform, making them ideal candidates to navigate the multifaceted challenges of the implant infection microenvironment. The bacterial cell membrane and the mitochondrial inner membrane, which serve as the primary loci for the respiratory chain in prokaryotes and eukaryotes, respectively, share substantial structural homology, both comprising phospholipid bilayers rich in negatively charged lipids. Leveraging this structural homology alongside the distinct spatial localization of respiratory chains in prokaryotes and eukaryotes, we developed spatially specific guanidine-quaternary ammonium carbon dots (GQCDs) via a hydrothermal approach to overcome the traditional antibacterial-osteogenic trade-off in implant materials. GQCDs operate through a spatiotemporally adaptive mechanism. Upon infection, they preferentially target and disrupt bacterial membranes, paralyzing respiratory chain-mediated energy metabolism. Following bacterial eradication, they subsequently localize to the host's mitochondrial network to dampen inflammation, suppress ferroptosis, and restore mitochondrial function, thereby fostering bone repair in the infected microenvironment. Conversely, under sterile conditions, they directly accumulate in mitochondria to promote osteogenesis via metabolic enhancement. This dual functionality was comprehensively validated in vitro as well as in vivo using both rat and beagle dog implant-associated infection models. Furthermore, we developed a versatile coating strategy to integrate GQCDs onto a wide spectrum of clinically relevant substrates (Ti, Ni, Mg, Ta, and PEEK), endowing these substrates with stable antibacterial and osteogenic properties. Ultimately, this work provides an innovative synthetic paradigm for smart antibacterial design and offers a highly promising translatable solution for the management of implant-associated infections. Results and discussion Synthesis and characterization of dual-functional GQCDs As illustrated in Fig. 1a, spatially specific GQCDs were synthesized via a hydrothermal method using anhydrous citric acid (CA), polyhexamethylene guanidine (PHMG), and diallyldimethylammonium chloride (DMDAAC). Transmission electron microscopy (TEM) revealed that GQCDs were quasi-spherical, well-dispersed, and uniform, with a lattice spacing of 0.23 ± 0.02 nm (Fig. 1b), characteristic of the (100) crystal plane of graphitic carbon [23,24]. The average particle size was determined to be 2.43 nm, a finding consistent with atomic force microscopy (AFM) measurements, which showed an average height of less than 2.5 nm (Fig. 1c). X-ray diffraction (XRD) revealed a discernible diffraction peak at approximately 23° (Fig. 1d), which can be assigned to the (002) plane of graphitic carbon, indicating the presence of short-range ordered graphitic domains. This broad diffraction feature suggests limited long-range crystallinity, which may be influenced by the presence of abundant surface functional groups.Fig. 1Characterization of GQCDs. a) Schematic diagram of GQCDs synthesis. b) TEM images of GQCDs (inset shows particle size distribution (right) and high-resolution TEM image (left)). c) AFM images of GQCDs. d) XRD patterns of GQCDs. e) FTIR of GQCDs. f) XPS spectra of GQCDs for the C, N, and O elements. g) GPC analysis of GQCDs. h) Proposed molecular and spatial structure of GQCDs. i) Zeta potential of GQCDs. j) Effect of GQCDs intervention on bacterial zeta potential across six species (Data represent the mean ± s.d., n = 3). Characterization of GQCDs. a) Schematic diagram of GQCDs synthesis. b) TEM images of GQCDs (inset shows particle size distribution (right) and high-resolution TEM image (left)). c) AFM images of GQCDs. d) XRD patterns of GQCDs. e) FTIR of GQCDs. f) XPS spectra of GQCDs for the C, N, and O elements. g) GPC analysis of GQCDs. h) Proposed molecular and spatial structure of GQCDs. i) Zeta potential of GQCDs. j) Effect of GQCDs intervention on bacterial zeta potential across six species (Data represent the mean ± s.d., n = 3). The potential functional groups of GQCDs were analyzed. As shown in the Fourier transform infrared spectrum (FTIR) (Fig. 1e), the absorption peak at 1476 cm−1 was attributed to the C-H scissoring vibration of quaternary ammonium groups [-N+(CH3)2-] [25], while peaks at 1649 cm−1 (N-H bending or C=N stretching) and 1420 cm−1 (C-N stretching) were indicative of successful modification of the guanidinium group. In addition, the absorption peaks around 1004 cm−1 and 963 cm−1 were related to C-N or C-O stretching vibrations, as well as = C-H out of plane bending vibrations in allyl structures, indicating that kept structures derived from DMDAAC in GQCDs [26,27]. X-ray photoelectron spectroscopy (XPS) revealed the predominant elemental composition of C (77.2%), N (10.24%), and O (12.56%), suggesting the high nitrogen content corroborating the successful surface functionalization (Fig. 1f). Gel permeation chromatography (GPC) revealed an average molecular weight of approximately 922 (Fig. 1g). Collectively, these results demonstrate that the GQCDs surface is enriched with alkyl chains, guanidinium moieties, and quaternary ammonium functional groups. Based on this comprehensive characterization, a putative molecular and spatial structure of GQCDs, retaining carbon core conjugated structure, key guanidinium and quaternary ammonium functional groups, was simulated (Fig. 1h). Bacterial cells typically exhibit a strongly negative surface charge due to the chemical composition of their walls and membranes, and disruption of this charge has been shown to induce bactericidal effects [28,29]. Thus, the zeta potentials of both GQCDs and the six bacterial strains were measured and compared. As shown in Fig. 1i, GQCDs exhibited a positive surface charge of +7.48 mV. In contrast, all tested bacteria showed negatively charged surfaces: S. epidermidis (−22.4 to −24.0 mV), S. aureus (−17.3 to −18.8 mV), MRSA (−13.6 to −15.5 mV), P. aeruginosa (−20.4 to −21.0 mV), E. coli (−25.5 to −25.8 mV), and E. cloacae (−14.5 to −16.3 mV). This electrostatic contrast is expected to facilitate the adsorption of GQCDs onto bacterial membranes. Consistent with this, treatment with GQCDs led to a significant reduction in the absolute zeta potential values across all six bacterial strains (Fig. 1j), indicating disruption of the native electronegativity of the bacterial surface. These results suggest strong electrostatic binding between GQCDs and bacterial cells. GQCDs exhibit potent and broad-spectrum antibacterial activity The introduction of guanidinium groups would endow the newly synthesized GQCDs with antibacterial efficacy through multiple mechanisms, including bacterial membrane disruption, inhibition of genetic material synthesis, and suppression of energy metabolism (Supplementary Fig. 1a). The antibacterial activity of GQCDs against six bacterial strains was evaluated. The minimum inhibitory concentration (MIC) values of GQCDs were determined as 2.5 μg/mL for S. epidermidis and E. cloacae, 5 μg/mL for S. aureus and MRSA, 7.5 μg/mL for E. coli, and 10 μg/mL for P. aeruginosa (Supplementary Fig. 1b), confirming their superior antibacterial efficacy. To assess the potential resistance risk induced by GQCDs, a 30-day serial passage experiment was conducted using vancomycin as a positive control. S. aureus and MRSA were selected as model bacteria, which were continuously exposed to sub-inhibitory concentrations (1/2 MIC) of either GQCDs or vancomycin for 30 consecutive passages, with MIC values monitored at each passage. The results demonstrated that the MIC of vancomycin increased more than 3-fold from the initial value by the 30th passage, whereas the MIC of GQCDs remained essentially stable throughout the entire passaging process, without any evident stepwise elevation (Supplementary Fig. 1c). These findings indicate that GQCDs pose a relatively low risk of inducing bacterial resistance or adaptive tolerance. To investigate the source of its antibacterial properties, as shown in Supplementary Fig. 1d, the MIC values of the raw materials were measured separately. It was ultimately found that CA had no antibacterial properties, while PHMG had antibacterial properties close to GQCDs and significantly better than DMDAAC. Therefore, it can be inferred that PHMG is the significant factor contributing to the antibacterial properties of GQCDs. However, to facilitate their broader application, further investigation into the specific mechanisms underlying their antibacterial effects is required. The antibacterial performance of GQCDs was further compared with that of PHMG. As shown in Supplementary Table 1, the minimum inhibitory concentrations (MICs) of GQCDs against all six tested bacterial strains were lower than those of PHMG. Additionally, at a concentration of 10 μg/mL, agar plate assays (Supplementary Fig. 1e) revealed that GQCDs completely inhibited bacterial growth in all six strains, whereas PHMG-treated groups still exhibited visible bacterial colonies. These results collectively confirm the superior antibacterial activity of GQCDs. To investigate the antibacterial performance of GQCDs, six common infectious bacteria were cultured and validated through agar plate culture, Live/Dead staining, and scanning electron microscopy (SEM) (Fig. 2a). In plate culture assays, GQCDs completely suppressed bacterial growth compared to the heavily colonized control (Fig. 2b). Live/dead staining further revealed predominantly red fluorescence (dead bacteria) in the GQCDs group, in contrast to the viable blue-stained control (Supplementary Fig. 2a–c). Furthermore, the inhibitory and clearance capabilities of GQCDs against bacterial biofilms were evaluated (Fig. 2c). The results demonstrated that GQCDs significantly inhibited the formation and development of biofilm (achieving 99% inhibition at concentrations >60 μg/mL, Fig. 2d) and exhibited concentration-dependent biofilm eradication efficacy, GQCDs at concentrations above 60 μg/mL resulted in greater than 50% biofilm clearance across all six bacterial strains (Fig. 2e).Fig. 2Antibacterial verification and mechanism exploration. a) Schematic illustration of the antimicrobial assay. b) Antibacterial smears of GQCDs. c) Schematic of the antibiofilm mechanism of GQCDs. d) Inhibition assay of GQCDs against biofilms (Data represent the mean ± s.d., n = 3. analyzed by one-way ANOVA). e) Eradication assay of GQCDs against biofilms (Data represent the mean ± s.d., n = 3. analyzed by one-way ANOVA). f) Antibacterial kinetic curves of GQCDs. g) Antibacterial stability of GQCDs (Data represent the mean ± s.d., n = 3. analyzed by one-way ANOVA). h) SEM images of six bacterial species following GQCDs intervention. i, j) ONPG assay results of six bacterial species treated with GQCDs at different concentrations (Data represent the mean ± s.d., n = 4. analyzed by one-way ANOVA). Antibacterial verification and mechanism exploration. a) Schematic illustration of the antimicrobial assay. b) Antibacterial smears of GQCDs. c) Schematic of the antibiofilm mechanism of GQCDs. d) Inhibition assay of GQCDs against biofilms (Data represent the mean ± s.d., n = 3. analyzed by one-way ANOVA). e) Eradication assay of GQCDs against biofilms (Data represent the mean ± s.d., n = 3. analyzed by one-way ANOVA). f) Antibacterial kinetic curves of GQCDs. g) Antibacterial stability of GQCDs (Data represent the mean ± s.d., n = 3. analyzed by one-way ANOVA). h) SEM images of six bacterial species following GQCDs intervention. i, j) ONPG assay results of six bacterial species treated with GQCDs at different concentrations (Data represent the mean ± s.d., n = 4. analyzed by one-way ANOVA). Antibacterial kinetics assays revealed a concentration-dependent inhibitory effect of GQCDs against all six bacterial strains (Fig. 2f). To further evaluate the storage stability and batch-to-batch reproducibility of their antimicrobial efficacy, agar plate-based tests were conducted. The MICs of GQCDs against these bacteria remained unchanged over a 20-week storage period (Fig. 2g). Coating experiments further confirmed that GQCDs retained stable antibacterial activity after 4, 8, and 12 weeks of storage (Supplementary Fig. 2f), demonstrating their long-term functional stability. Moreover, GQCDs synthesized in different months exhibited consistent antibacterial performance (Supplementary Fig. 2g), indicating high reproducibility of the preparation approach. The antibacterial mechanism of GQCDs was investigated. SEM images exhibited extensive morphological damage to the bacteria, including loss of membrane integrity, leakage of cytoplasmic material, and the formation of cellular fragments (Fig. 2h). o-nitrophenyl-β-D-galactopyranoside (ONPG) assays revealed significantly increased extracellular ONPG, further demonstrated the loss of bacterial membrane integrity, as evidenced by the leakage of intracellular contents (Fig. 2i and j). These findings indicated that GQCDs exert antibacterial efficacy by disrupting membrane integrity. Due to the fact that bacterial energy metabolism mainly relies on the respiratory chain on the cell membrane, disruption of bacterial cell membrane integrity is highly likely to affect its energy metabolism. Finally, consistent with this membrane disruption, a dose-dependent increase in extracellular ATP levels was also observed (Supplementary Fig. 2d and e). These results suggest that GQCDs may directly disrupt the integrity of cell membranes and strongly interfere with energy metabolism [30], leading to ATP leakage or stress-induced release, ultimately reducing ATP in bacterial cells to achieve antibacterial efficacy. Transcriptomic analysis reveals GQCDs disrupt core metabolic, virulence, and resistance pathways The antibacterial molecular mechanisms of GQCDs were investigated. The RNA sequencing (RNA-seq) was first performed (Fig. 3a). Volcano plots for each species of the six pathogenic strains revealed extensive transcriptional remodeling in response to GQCDs, indicating profound alterations in their metabolic and homeostatic states (Supplementary Fig. 3a). We then conducted detailed pathway enrichment analysis on the differentially expressed genes (DEGs) for each major bacterial group (Fig. 3c–h). Among Gram-positive bacteria, GQCDs exerted multifaceted disruptive effects. In S. epidermidis, Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis showed significant downregulation of genes associated with ABC transporters and nitrogen metabolism, suggesting that GQCDs impair transmembrane transport and energy production. Gene Ontology (GO) analysis corroborated this by revealing significant inhibition of pathways related to membrane protein complexes, peptidoglycan catabolism, and nucleotide biosynthesis (Fig. 3c).Fig. 3Antibacterial verification and mechanism exploration. a) Differential analysis of six bacterial species treated with GQCDs. b) Grouped volcano plots showing the effects of GQCDs on six bacterial species. c-h) Differential enrichment analysis of S. epidermidis (c), S. aureus (d), MRSA (e), P. aeruginosa (f), E. coli (g), and E. cloacae (h) in response to GQCDs. i, j) Molecular docking analysis of key proteins in S. aureus (i) and E. coli (j) with GQCDs. Antibacterial verification and mechanism exploration. a) Differential analysis of six bacterial species treated with GQCDs. b) Grouped volcano plots showing the effects of GQCDs on six bacterial species. c-h) Differential enrichment analysis of S. epidermidis (c), S. aureus (d), MRSA (e), P. aeruginosa (f), E. coli (g), and E. cloacae (h) in response to GQCDs. i, j) Molecular docking analysis of key proteins in S. aureus (i) and E. coli (j) with GQCDs. Similarly, in S. aureus, KEGG analysis again highlighted the inhibition of ABC transporters and key virulence pathways. Furthermore, metabolic pathways for arginine and histidine were significantly suppressed. The disruption of arginine metabolism, which generates ornithine crucial for DNA and membrane integrity, and histidine metabolism, which feeds into the energy-providing tricarboxylic acid (TCA) cycle, suggests a multi-pronged attack on S. aureus survival. GO analysis further indicated that GQCDs inhibit the formation of extracellular structures and interfere with amino acid and protein synthesis in S. aureus (Fig. 3d). Notably, in MRSA, transcriptomic analysis confirmed the suppression of core metabolic processes like ribosome function and ABC transporters, and also revealed a unique mechanism for overcoming drug resistance (Fig. 3e). KEGG analysis showed that GQCDs significantly downregulated pathways associated with beta-lactam antibiotic resistance. GO enrichment analysis provided a comprehensive picture of the damage, highlighting the disruption of MRSA's cell membrane and anatomical structure, the inhibition of transmembrane transport and protein synthesis, and profound interference with the respiratory chain and coupled ATP synthesis. GQCDs also exerted profound effects on Gram-negative bacteria by targeting their central energy metabolism. In P. aeruginosa, KEGG analysis showed direct inhibition of the carbon fixation pathway, the TCA cycle, and oxidative phosphorylation, as well as suppression of ribosome-mediated protein synthesis. GO analysis further detailed this disruption, showing downregulation of terms associated with structural molecule activity, ribosome subunits, and the catabolism of amino and organic acids (Fig. 3f). In E. coli, a similar potent inhibition of core energy-producing pathways was observed, including carbon fixation, oxidative phosphorylation, and the TCA cycle. KEGG analysis also revealed the suppression of cysteine, histidine, and arginine biosynthesis. GO analysis provided deeper mechanistic insight, revealing direct inhibition of specific respiratory components, including NADH dehydrogenase and proton-transporting ATP synthase complexes, which collectively disrupt cellular respiration and ATP synthesis (Fig. 3g). Secondly, for E. cloacae, KEGG enrichment showed that GQCDs can also directly inhibit its carbon fixation pathway, TCA cycle, and oxidative phosphorylation energy metabolism processes, while also inhibiting cysteine metabolism; GO enrichment analysis showed that GQCDs can directly inhibit the NADH dehydrogenase activity, electron transfer activity, and proton channel protein activity of E. cloacae, suppress cell respiration, electron transport chain, and ATP synthesis process coupled with electron transport chain (Fig. 3h). Therefore, according to the transcription results, GQCDs can disrupt the energy metabolism process of E. cloacae by inhibiting NADH dehydrogenase activity, electron transport chain, respiratory system, proton transport ATP complex enzymes, and other mechanisms, ultimately achieving stable antibacterial effects through multiple mechanisms. Correspondingly, as shown in Fig. 3b, a consolidated analysis across all six species identified several key genes consistently downregulated by GQCDs. This suppression targeted essential cellular processes, including respiratory chain (icd, sdhA, cydB) and the TCA cycle, cell wall synthesis (tagH), DNA replication (ssb), oxidative stress response (lon), and critically, antibiotic resistance (blaR1). To gain deeper mechanistic insight, molecular docking studies were performed using GQCDs—representing the molecular structure of GQCDs (representing the molecular structure of GQCDs) with key proteins from S. aureus (Gram-positive) and E. coli (Gram-negative). Six essential bacterial proteins were selected as potential targets: FtsA (involved in bacterial division), ABC transporter ATP-binding protein (metabolism and transport), intercellular adhesion protein R (biofilm formation), DNA primase (DNA replication), coenzyme Q, and cytochrome oxidase (both associated with the respiratory chain). As shown in Fig. 3i, Supplementary Fig. 3b, GQCDs exhibited high binding affinity to all six target proteins in S. aureus, with the strongest interactions observed for respiratory chain-related proteins: coenzyme Q (−12.381 kcal/mol) and cytochrome oxidase (−10.625 kcal/mol). Similarly, in E. coli (Fig. 3j, Supplementary Fig. 3c), GQCDs showed strong binding (all > −8 kcal/mol), with the highest affinity again for coenzyme Q (−12.560 kcal/mol), followed by cytochrome oxidase (−9.091 kcal/mol). These results indicate a preferential targeting of respiratory chain components by GQCDs in both Gram-positive and Gram-negative bacteria. It has been reported that disruption of the bacterial respiratory chain is closely associated with rapid accumulation of reactive oxygen species (ROS) and depletion of glutathione (GSH) [31,32]. To further investigate the antibacterial mechanism of GQCDs, we treated these strains with four different concentrations of GQCDs and quantified ROS and GSH levels using ELISA and flow cytometry, respectively (Supplementary Fig. 3d and e). The results demonstrated a concentration-dependent increase in ROS and a concurrent decrease in GSH across all strains, supporting the role of respiratory chain inhibition in the antibacterial activity of GQCDs. In summary, this integrated analysis confirms that GQCDs exert robust antibacterial effects through multiple pathways, including respiratory chain inhibition, disruption of energy metabolism, and impairment of membrane integrity. Among these, interference with energy metabolism, namely the inhibition of the bacterial respiratory chain, represents a central mechanism underlying the broad-spectrum antibacterial activity of GQCDs. GQCDs exhibit biocompatibility and promote osteogenesis in vitro To assess their potential for clinical application, we first evaluated the cytotoxicity and effects of GQCDs on osteoblasts (Fig. 4a). CCK-8 assays demonstrated excellent biocompatibility (Fig. 4b). The optical density (OD) values of osteoblasts treated by 30–90 μg/mL of GQCDs were comparable to or even surpassing the control group. Notably, GQCDs exhibited a significant pro-proliferative effect at concentrations between 30 and 60 μg/mL, with optimal enhancement observed at 60 μg/mL. RT-qPCR and immunofluorescence staining were further carried out to validate the pro-osteogenic potential of GQCDs. As shown in Fig. 4c and d, GQCDs significantly up-regulated the expression of key osteogenic genes of Runx2 and OCN with dose-dependence. Furthermore, immunofluorescence staining (Supplementary Fig. 4a–d) showed that GQCDs significantly enhanced the protein expression of Runx2 and OCN within the 30–60 μg/mL range, which was consistent with the RT-qPCR results. These results demonstrate that, beyond its bactericidal properties, GQCDs further stimulates osteoblast proliferation as well as differentiation. Additionally, GQCDs (30-60 μg/mL) significantly increased intracellular ATP level of osteoblasts (Fig. 4e). This suggests that GQCDs exert divergent effects on bacterial (prokaryotic) and eukaryotic cells, likely attributable to fundamental structural differences between these cell types.Fig. 4Osteogenic performance validation and mechanism exploration. a) Schematic diagram illustrating the action of GQCDs on cells. b) Cytotoxicity of GQCDs at different concentrations on osteoblasts assessed by CCK-8 assay (Data represent the mean ± s.d., n = 3. analyzed by one-way ANOVA). c, d) Effects of GQCDs on the mRNA expression of osteogenic genes Runx2 and OCN measured by RT-qPCR (Data represent the mean ± s.d., n = 3. analyzed by one-way ANOVA). e) Effect of GQCDs on ATP levels in osteoblasts (Data represent the mean ± s.d., n = 3. analyzed by one-way ANOVA). f) PCA plot and volcano plot of transcriptional profiles in osteoblasts treated with GQCDs. g) KEGG pathway analysis of GQCDs-dependent upregulated and downregulated genes. h) KEGG pathway analysis revealing significantly impacted metabolic pathways in osteoblasts treated with GQCDs compared to the control group. i) Integrated transcriptomic and metabolomic analysis of osteoblasts. j) Quantitative expression analysis of COX4 and GPX4 by RT-qPCR (Data represent the mean ± s.d., n = 3. Analyzed by one-way ANOVA). k) Mitochondrial membrane potential in osteoblasts after GQCDs intervention measured by flow cytometry (Data represent the mean ± s.d., n = 3. Analyzed by one-way ANOVA). l) Schematic representation of the transcriptional response in osteoblasts infected with S. aureus or E. coli. m, n) Transcriptomic enrichment analysis of osteoblasts infected with S. aureus (m) and E. coli(n). o) Fluorescence staining of mitochondria and lysosomes in osteoblasts after S. aureus and GQCDs intervention. p) Transmission electron microscopy images of mitochondria in cells following S. aureus and GQCDs intervention. q) Western blot analysis of System Xc− and GPX4 protein expression in cells following S. aureus and GQCDs intervention. r-u) Changes in ROS (r), MDA (s), cell viability (t), and GSH (u) levels after treatment with GQCDs, Lip-1, and S. aureus (Data represent the mean ± s.d., n = 3 for r, u; n = 4 for s, t; analyzed by one-way ANOVA). Osteogenic performance validation and mechanism exploration. a) Schematic diagram illustrating the action of GQCDs on cells. b) Cytotoxicity of GQCDs at different concentrations on osteoblasts assessed by CCK-8 assay (Data represent the mean ± s.d., n = 3. analyzed by one-way ANOVA). c, d) Effects of GQCDs on the mRNA expression of osteogenic genes Runx2 and OCN measured by RT-qPCR (Data represent the mean ± s.d., n = 3. analyzed by one-way ANOVA). e) Effect of GQCDs on ATP levels in osteoblasts (Data represent the mean ± s.d., n = 3. analyzed by one-way ANOVA). f) PCA plot and volcano plot of transcriptional profiles in osteoblasts treated with GQCDs. g) KEGG pathway analysis of GQCDs-dependent upregulated and downregulated genes. h) KEGG pathway analysis revealing significantly impacted metabolic pathways in osteoblasts treated with GQCDs compared to the control group. i) Integrated transcriptomic and metabolomic analysis of osteoblasts. j) Quantitative expression analysis of COX4 and GPX4 by RT-qPCR (Data represent the mean ± s.d., n = 3. Analyzed by one-way ANOVA). k) Mitochondrial membrane potential in osteoblasts after GQCDs intervention measured by flow cytometry (Data represent the mean ± s.d., n = 3. Analyzed by one-way ANOVA). l) Schematic representation of the transcriptional response in osteoblasts infected with S. aureus or E. coli. m, n) Transcriptomic enrichment analysis of osteoblasts infected with S. aureus (m) and E. coli(n). o) Fluorescence staining of mitochondria and lysosomes in osteoblasts after S. aureus and GQCDs intervention. p) Transmission electron microscopy images of mitochondria in cells following S. aureus and GQCDs intervention. q) Western blot analysis of System Xc− and GPX4 protein expression in cells following S. aureus and GQCDs intervention. r-u) Changes in ROS (r), MDA (s), cell viability (t), and GSH (u) levels after treatment with GQCDs, Lip-1, and S. aureus (Data represent the mean ± s.d., n = 3 for r, u; n = 4 for s, t; analyzed by one-way ANOVA). To further elucidate the underlying mechanism of this differential cellular behavior and investigate the subcellular localization of GQCDs within eukaryotic cells, we performed colocalization tracking by leveraging the intrinsic photoluminescence of GQCDs using Super-resolution confocal microscopy (ZEISS Elyra 7). Following the co-incubation of osteoblasts with GQCDs, the cells were counterstained with the mitochondria-specific probe MitoTracker Deep Red FM (Supplementary Fig. 4e). The imaging results visually revealed a high degree of spatial overlap between the intrinsic blue fluorescence of GQCDs and the red fluorescence of MitoTracker. Colocalization scatter plots (Supplementary Fig. 4g) and correlation coefficients (Supplementary Fig. 4f) were generated using the Colocalization Finder plugin in ImageJ. The quantitative analysis demonstrated a Pearson correlation coefficient (Rr) of 0.880 ± 0.040 and a Manders’ colocalization coefficient (M2) of 0.875 ± 0.057. Both metrics significantly exceed the strong correlation threshold of 0.7, unequivocally confirming the robust colocalization of GQCDs with the osteoblastic mitochondrial network. These findings strongly suggest that upon internalization by eukaryotic cells, the GQCDs predominantly accumulate within mitochondria to exert their subsequent biological functions. Meanwhile, transcriptomic and metabolomic analyses were performed on treated osteoblasts to elucidate the mechanisms underlying the differential effects of GQCDs on eukaryotic versus prokaryotic cells. Principal component analysis (PCA) of the transcriptomic data revealed a clear separation and distinct expression profiles between the GQCDs -treated and control groups (Fig. 4f). A total of 741 DEGs were identified, with 653 genes upregulated and 88 downregulated following GQCDs treatment (Fig. 4f). KEGG analysis of these DEGs showed that, in stark contrast to the metabolic suppression observed in bacteria, GQCDs promoted a hypermetabolic state in osteoblasts. This was characterized by the significant upregulation of pathways for ABC transporters and amino acid metabolism (phenylalanine, tyrosine). Critically, the analysis also revealed a dual signature of improved mitochondrial health: a significant enhancement of glutathione metabolism coupled with a robust suppression of the ferroptosis pathway (Fig. 4g). These transcriptional changes provide a strong molecular basis for the functional improvements, including increased mitochondrial membrane potential and ATP production. This conclusion was further corroborated by the metabolomic data and a combined multi-omics analysis (Fig. 4h and i). Collectively, the integrated results raise that the pro-osteogenic effect of GQCDs is primarily mediated through the enhancement of mitochondrial metabolism. RT-qPCR analysis of key respiratory chain components of COX4 and GPX4 revealed that GQCDs upregulate their expression in a concentration-dependent range of 30–60 μg/mL (Fig. 4j), demonstrating that GQCDs indeed promote energy metabolism in osteoblasts. Subsequent assessment of mitochondrial membrane potential using flow cytometry (Fig. 4k) and fluorescence staining (Supplementary Fig. 4h and i) showed that GQCDs dose-dependently enhance mitochondrial membrane potential. Additionally, fluorescence staining (Supplementary Fig. 4j and k) and ELISA (Supplementary Fig. 4l) were employed to measure cellular ROS and GSH levels. The results indicated that GQCDs (30–60 μg/mL) significantly increased intracellular GSH production while dose-dependently suppressing ROS generation (30–90 μg/mL), further confirming that GQCDs exert their osteogenic-promoting effects by scavenging ROS, and enhancing mitochondrial metabolism. Therefore, we propose that the structural dichotomy between prokaryotes and eukaryotes—specifically, the presence of mitochondria in eukaryotes—underpins the dual targeting mechanism of GQCDs. Specifically, positively charged GQCDs exhibit cationic characteristics. According to the Nernst equation, a transmembrane potential difference of 59.2 mV increases the concentration of cations on the negatively charged side of the membrane by approximately 10-fold. Consequently, compared with the plasma membrane (<−30 mV), the higher negative potential of the mitochondrial inner membrane (approximately −120 mV) drives the sequential translocation of GQCDs across the cell membrane and promotes their extensive accumulation within the mitochondrial membrane system [33]. This mechanism is supported by experimental data. Zeta potential measurements indicate that GQCDs treatment did not alter the surface charge of osteoblasts (Supplementary Fig. 4m), while previous studies have confirmed that GQCDs significantly enhance mitochondrial membrane potential (Fig. 4k). This suggests that GQCDs primarily function within the intermembrane space between the mitochondrial inner and outer membranes. Moreover, the nanoscale size of GQCDs (<2.5 nm) allows them to pass through the pores of the mitochondrial outer membrane (2–3 nm) and accumulate in the intermembrane space, thereby enhancing mitochondrial membrane potential without interfering with core metabolic processes. Further investigation in a bacteria-cell co-culture system revealed that under non-infected conditions, GQCDs did not affect the surface charge of osteoblasts (Supplementary Fig. 4m). While, under infected conditions, GQCDs treatment induced an increase in membrane potential, indicating that GQCDs preferentially interact with bacteria and disrupt bacterial membrane potential. To directly evaluate the binding selectivity of GQCDs in a complex infection microenvironment, we established an in vitro competitive co-culture model comprising osteoblasts and E. coli. Following 24 h of cell seeding, the dynamic interactive processes were tracked using confocal microscopy (ZEISS LSM900) (Supplementary Fig. 5a). In the untreated control, the osteoblast membranes (labeled with CellMask™) remained intact, indicating a healthy cellular state. Conversely, in the bacterial infection group, copious E. coli adhered to the osteoblast membranes, subsequently triggering severe morphological aberrations in the host mitochondria (labeled with MitoTracker Green), including shrinkage and structural fragmentation. Strikingly, in the GQCDs-treated competitive co-culture group, GQCDs exhibited remarkable spatiotemporal selectivity. After 3 h of co-incubation, the intrinsic blue fluorescence of GQCDs overwhelmingly co-localized with the bacterial clusters, with negligible signal detected inside osteoblasts. Concurrently, abundant PI-positive (dead) bacteria and debris were observed alongside minimal SYTO 9-positive (live) bacteria, indicating that GQCDs preferentially targeted and efficiently eradicated bacteria during the early contact phase. Extending the co-incubation to 6 h revealed a near-total clearance of viable bacteria, accompanied by a significant intracellular accumulation of GQCDs’ blue fluorescence within the osteoblasts. Fluorescence quantitative analysis (Supplementary Fig. 5b–d) further corroborated this dynamic shift. These observations demonstrate that in a mixed infection microenvironment, GQCDs possess a sequential targeting capability: they preferentially bind to bacteria to exert bactericidal efficacy before gradually internalizing into the host cells. To unravel the thermodynamic and molecular basis driving this selective binding, we performed in silico molecular docking simulations between GQCDs and representative membrane proteins of bacteria, host cells, and mitochondria (Supplementary Fig. 5e). The results revealed a distinct binding affinity gradient: GQCDs exhibited the most thermodynamically favorable interactions with the outer membrane proteins of Gram-negative E. coli (BamA and OmpC), yielding the lowest binding energies of −11.0 and −10.6 kcal/mol, respectively. This was followed by affinities for Gram-positive S. aureus proteins (Cna and FnBP) at −8.8 and −8.3 kcal/mol. In stark contrast, the predicted binding energy with Caveolin-1, a representative osteoblast plasma membrane protein, was notably higher at −7.9 kcal/mol (indicating weaker affinity). This computational hierarchy perfectly mirrors the in vitro observation of GQCDs’ preferential bacterial binding. Furthermore, GQCDs demonstrated strong affinities for mitochondrial membrane proteins (SLC25A3, TOM70, and TOM20) with binding energies of −10.2, −9.5, and −8.7 kcal/mol, respectively, which significantly outperformed their affinity for the host plasma membrane. Collectively, the competitive imaging and computational analyses elucidate that GQCDs preferentially bind and eliminate bacteria in an infection microenvironment. Once the bacterial threat is neutralized, or in sterile conditions, they effectively target the osteoblastic mitochondrial network, showcasing a smart-responsive targeting switch. To further investigate the regulatory role of GQCDs on osteoblasts under infectious conditions, transcriptomic analysis on cells infected with representative Gram-positive (S. aureus) and Gram-negative (E. coli) pathogens was comparatively performed with and without GQCDs treatment (Fig. 4l). KEGG revealed that GQCDs treatment significantly suppressed infection-induced ferroptosis, apoptosis, and cellular senescence in osteoblasts (Fig. 4m and n). Additionally, GQCDs inhibited the PPAR signaling pathway and attenuated chronic inflammatory responses, further supporting their dual functionality in possessing both antibacterial activity and osteogenic potential under infected conditions. The effect of GQCDs on mitochondrial function was validated using co-localization fluorescence staining of mitochondria and lysosomes (Fig. 4o, Supplementary Fig. 6a). The results showed that infected cells exhibited diminished mitochondrial function, cellular atrophy, and typical apoptotic features. In contrast, GQCDs intervention significantly improved mitochondrial function in infected cells. TEM of mitochondria provided further morphological evidence (Fig. 4p). Compared with the control group, mitochondria in the infection group displayed reduced volume and partially disrupted cristae. Following GQCDs treatment, mitochondrial volume increased, cristae remained largely intact, and mitochondrial function was markedly restored. Simultaneously, Western blot analysis further validated this underlying mechanism, as demonstrated in Fig. 4q and Supplementary Fig. 6b–c. GQCDs treatment significantly enhanced mitochondrial function and dose-dependently upregulated the expression of the key anti-ferroptotic protein of GPX4, as well as the cystine/glutamate antiporter (System Xc−). Meanwhile, as shown in Supplementary Fig. 6d, osteoblasts were cultured and analyzed using the Seahorse XF96 Pro Analyzer to evaluate the effects of GQCDs treatment under infection on cellular oxygen consumption rate (OCR), basal respiration, maximal respiration, and ATP production. The results indicated that GQCDs significantly enhanced cellular respiration, further confirming the ability of GQCDs to promote mitochondrial metabolism (Supplementary Fig. 6e–h). Furthermore, in a ferroptosis model induced by RSL3, GQCDs, similar to the ferroptosis inhibitor of Liproxstatin-1 (Lip-1), effectively rescued cell viability by suppressing RSL3-induced ROS and MDA elevation while restoring intracellular GSH levels (Fig. 4r–u). Although the regulation of the GPX4 and System Xc− axis constitutes a classical pathway for ferroptosis suppression, the onset of ferroptosis within the complex implant infection microenvironment is actually a multifaceted cascade orchestrated by pathogen invasion, host immune responses, and microenvironmental stress [34]. The core pathological mechanism lies in the fact that bacterial virulence factors not only trigger severe inflammation but also drastically disrupt host cell iron homeostasis, collectively driving the progression of ferroptosis Driven by this rationale, to further validate the intervention efficacy of GQCDs on infection-induced ferroptosis, we directly evaluated the lipid peroxidation levels in osteoblasts using the C11-BODIPY 581/591 probe. Flow cytometry analysis revealed that, compared to the uninfected control, S. aureus infection significantly intensified the oxidized C11-BODIPY signal (green fluorescence), confirming that the infection microenvironment strongly provoked lipid peroxidation (Supplementary Fig. 6i and j). Remarkably, upon GQCDs intervention, the green fluorescence intensity was substantially attenuated. This phenotype visually demonstrates that GQCDs effectively suppress infection-triggered lipid peroxidation and the pathogenic accumulation of lipid reactive oxygen species (lipid ROS). Subsequently, we quantified the expression of core targets within the ferroptosis-driving network: ACSL4 (acyl-CoA synthetase long-chain family member 4) and TFR1 (Transferrin receptor 1). RT-qPCR results indicated that bacterial infection dramatically upregulated the mRNA levels of ACSL4 and TFR1 in host osteoblasts, whereas GQCDs treatment significantly reversed this trend at the transcriptional level (Supplementary Fig. 6k and l). Combined with our previous RSL3 rescue experiments, these phenotypic and molecular findings collectively verify that GQCDs not only scavenge lipid peroxides but also block ferroptosis signaling by downregulating the pro-ferroptotic factors ACSL4 and TFR1, thereby effectively preserving the osteogenic differentiation potential of osteoblasts under severe infection stress. To explore the far-upstream mechanism by which GQCDs regulate ferroptosis—specifically, whether they directly interfere with iron metabolism or merely modulate downstream antioxidant signals—we quantitatively tracked intracellular free Fe2+ (the labile iron pool, LIP) using the specific fluorescent probe FerroOrange. The results (Supplementary Fig. 6m and n) demonstrated that infection stimuli induced an explosive surge in intracellular LIP levels. Conversely, GQCDs treatment significantly suppressed this intracellular Fe2+ overload in a strictly dose-dependent manner. This finding unequivocally indicates that GQCDs possess the capacity to directly deplete the excessive LIP required to drive the Fenton reaction. Taken together, within the complex infection microenvironment, GQCDs comprehensively antagonize infection-induced ferroptosis via a unique dual-pronged mechanism: (1) Upstream Blockade (Source quenching): Effectively depletes the LIP and intrinsically quenches the Fenton reaction-mediated lipid peroxidation cascade.(2)Downstream Restoration (Pathway remodeling): Upregulation of System Xc− (SLC7A11) and GPX4, which remodels the endogenous enzymatic antioxidant barrier and lipid metabolism homeostasis of the host. (2)Downstream Restoration (Pathway remodeling): Upregulation of System Xc− (SLC7A11) and GPX4, which remodels the endogenous enzymatic antioxidant barrier and lipid metabolism homeostasis of the host. Downstream Restoration (Pathway remodeling): Upregulation of System Xc− (SLC7A11) and GPX4, which remodels the endogenous enzymatic antioxidant barrier and lipid metabolism homeostasis of the host. Based on the above findings, GQCDs exhibit differential targeting between prokaryotic bacteria and eukaryotic cells. First, GQCDs can bind to negatively charged bacterial membranes through electrostatic interactions, thereby disrupting bacterial energy metabolism and achieving potent antibacterial effects. In eukaryotic cells, positively charged GQCDs (+7.48 mV) with lipophilic cationic properties, are able to specifically accumulate in mitochondria driven by the transmembrane mitochondrial potential and exert their pro-osteogenic effects by scavenging ROS and inhibiting ferroptosis. Surface modification of clinical implants with GQCDs confers potent antibacterial and pro-osteogenic properties To translate the therapeutic potential of GQCDs into clinical applications, we developed a polyvinyl butyral (PVB)-based coating technique for titanium (Ti), tantalum (Ta), nickel (Ni), magnesium (Mg), and polyetheretherketone (PEEK) implants (Fig. 5a). Biocompatibility assays using osteoblasts confirmed that coatings with final GQCDs concentrations of 50, 100, and 200 μg/mL exhibited excellent cytocompatibility (Fig. 5b–f). Notably, the 100 μg/mL GQCDs coating showed the strongest proliferative effect on osteoblasts and was thus selected for further implant modification. Using titanium as a representative substrate, SEM analyses revealed that Ti-GQCDs had a smoother and more homogenous surface compared to uncoated titanium (Fig. 5g). SEM imaging of osteoblasts cultured on different surfaces (Supplementary Fig. 7a) demonstrated limited cell attachment and spreading on uncoated implants, whereas Ti-GQCDs facilitated robust cell growth and extensive spreading.Fig. 5Performance evaluation of GQCDs-coated implants. a) Schematic illustration of preparing GQCDs-coated implants. b–f) Cytotoxicity of GQCDs-coated Ti, Ta, Ni, Mg, and PEEK (Data represent the mean ± s.d., n = 3. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, analyzed by one-way ANOVA). g) SEM images of pure Ti and Ti-GQCDs. h, i) Stress–displacement curves and elastic moduli of original and modified implants (Data represent the mean ± s.d., n = 3. ns, p > 0.05, analyzed by Unpaired t-test). j) SEM images of MRSA incubated with various implants and their modified counterparts. k) Schematic and operational diagrams illustrating the implantation model in rats. l–p) X-ray analysis of surgical rats at 0, 2, and 4-weeks post-implantation. q, r) Footprint analysis of rats implanted with Ti, Ta, and their modified counterparts. s) Quantitative analysis of footprint areas across different groups (Data represent the mean ± s.d., n = 3. *p < 0.05, analyzed by Unpaired t-test). Performance evaluation of GQCDs-coated implants. a) Schematic illustration of preparing GQCDs-coated implants. b–f) Cytotoxicity of GQCDs-coated Ti, Ta, Ni, Mg, and PEEK (Data represent the mean ± s.d., n = 3. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, analyzed by one-way ANOVA). g) SEM images of pure Ti and Ti-GQCDs. h, i) Stress–displacement curves and elastic moduli of original and modified implants (Data represent the mean ± s.d., n = 3. ns, p > 0.05, analyzed by Unpaired t-test). j) SEM images of MRSA incubated with various implants and their modified counterparts. k) Schematic and operational diagrams illustrating the implantation model in rats. l–p) X-ray analysis of surgical rats at 0, 2, and 4-weeks post-implantation. q, r) Footprint analysis of rats implanted with Ti, Ta, and their modified counterparts. s) Quantitative analysis of footprint areas across different groups (Data represent the mean ± s.d., n = 3. *p < 0.05, analyzed by Unpaired t-test). Mechanical testing demonstrated that GQCDs coating did not adversely affect the elastic modulus of the substrates, indicating that the modification preserved the mechanical integrity of the implants (Fig. 5h and i). Furthermore, leveraging the intrinsic fluorescence of GQCDs, the release kinetics of GQCDs from the modified implants were monitored. As shown in Supplementary Fig. 7b, the release of GQCDs from all modified implants (Ti-GQCDs, Ta-GQCDs, Ni-GQCDs, Mg-GQCDs, and PEEK-GQCDs) was largely complete within 8–10 days. These results indicate that GQCDs-coated implants enable sufficient early-stage release of the active agent post-implantation. To systematically evaluate the in vitro antibacterial performance of the modified implants, MRSA was selected as a representative strain and analyzed via a 7-day continuous antibacterial assay. SEM observation revealed robust adhesion and growth of MRSA on the surfaces of uncoated materials (Ti, Ta, Ni, Mg, PEEK). In contrast, bacteria on GQCDs-coated surfaces exhibited evident structural disruption and only sporadic bacterial debris, demonstrating the potent antibacterial effect of the GQCDs coating (Fig. 5j). Furthermore, the 7-day continuous antibacterial assay showed that the GQCDs-coated implants maintained high bactericidal activity (inhibition rate >99%) for at least 6 days, highlighting the sustained antibacterial capacity conferred by the coated GQCDs (Supplementary Fig. 7c). Furthermore, we precisely quantified the coating thickness of both unmodified and GQCDs–coated implants via morphological characterization (Supplementary Fig. 8a). And quantitatively analyzed the drug loading of different implants based on the in vitro drug release curve (Supplementary Fig. 8b). Meanwhile, considering that orthopedic implants are chronically exposed to physiological fluids and must endure severe mechanical shear forces during surgical insertion, we subjected the coatings to rigorous physicochemical stability evaluations. The GQCDs–coated samples were either immersed in simulated body fluid (SBF) for 24 h or subjected to high-power ultrasonication in pure water for 15 min. Morphological observations revealed that, despite experiencing these extreme chemical and mechanical stresses, the GQCDs coatings maintained excellent structural integrity, with no visible signs of delamination, microcracking, or severe degradation. Crucially, this robust interfacial adhesion directly translated into durable biological efficacy. Agar diffusion assays demonstrated that, compared to the control, the Ti-GQCDs substrates subjected to SBF immersion and ultrasonic perturbation still exhibited distinct and prominent zones of inhibition (Supplementary Fig. 8c). Further surface wash-and-plate assays confirmed this: while the unmodified control group (①) presented dense colonies of viable bacteria, all tested GQCDs–coated groups (including unperturbed and stressed groups ②, ③, and ④) showed zero bacterial growth on the agar plates, achieving a nearly 100% surface bacterial eradication rate (Supplementary Fig. 8d). Collectively, these results unequivocally demonstrate that the GQCDs coating not only withstands harsh mechanical and chemical disruptions through robust interfacial bonding but also retains exceptional and long-lasting antibacterial activity under extreme conditions, highlighting its immense potential for clinical translation. A rat implant infection model was established to systematically assess the in vivo performance (Fig. 5k). Specifically, the knee joint was exposed through a patellar ligament incision, and a standardized bone defect was created in the distal femur using a 2 mm drill. After injecting MRSA, either unmodified or GQCDs–coated implants (Ti, Ta, Ni, Mg, PEEK) were introduced (Fig. 5k). The progression of infection and osteogenesis at the implant site was dynamically monitored by X-ray imaging at 0, 2, and 4 weeks post-surgery (Fig. 5l–p). Radiographic analysis showed that immediately after implantation (0 weeks), all implants were well-fixed, with intact femoral morphology and uniform bone density, indicating standardized surgical procedures. At 2 weeks, the unmodified implant groups (Ti, Ta, Ni, Mg, PEEK) displayed typical signs of the destruction of the distal femoral cortex, low-density shadows in the medullary cavity, soft-tissue swelling, and osteolytic lesions. In sharp contrast, none of the GQCDs-coated groups (Ti-GQCDs, Ta-GQCDs, Ni-GQCDs, Mg-GQCDs, PEEK-GQCDs) exhibited visible infectious signs at this time point, suggesting that the modified implants exerted significant antibacterial effects early after implantation. By 4 weeks, infection in the unmodified implant groups had worsened further, manifested as severe bone destruction in the distal femur (Red triangle indicating part), loosening of Ti and Ta groups, and spread of infection to the knee joint, severely compromising limb function. In contrast, the GQCDs-coated groups remained free of local infection around the implant sites with stable fixing and no progressive bone destruction. These results demonstrate that all GQCDs-coated implants maintain excellent and durable antibacterial performance even in a complex in vivo environment, effectively controlling implant-associated infection and preserving bone structural stability. At 8 weeks, the function of the surgical limbs was further evaluated using footprint analysis. Compared to the Ti-GQCDs and Ta-GQCDs groups, rats implanted with unmodified Ti and Ta exhibited significantly reduced footprint areas in the right hind limb (Fig. 5q and r, Supplementary Fig. 9a and b). A similar trend was observed in the other implant groups (Ni, Mg, and PEEK) (Supplementary Fig. 9d–f). Quantitative analysis further confirmed larger footprint areas and faster movement speeds in all GQCDs-coated groups, indicating preserved limb function (Fig. 5s, Supplementary Fig. 9c). Collectively, these results demonstrate that GQCDs–coated implants maintain durable antibacterial efficacy and promote bone healing and functional recovery in vivo. Performance validation of GQCDs-Coated implants in rats To further compare the in vivo performance of different coated implants, we performed histological and immunological analyses at eight weeks post-implantation (Fig. 6a). Femurs from all five groups receiving unmodified implants showed severe infection, evidenced by pronounced bone expansion and cortical destruction at the distal femur. In contrast, all GQCDs–coated groups displayed intact bone cortex without signs of infection (Supplementary Fig. 9g), confirming the robust in vivo antibacterial and regeneration efficacy of the GQCDs coating. Moreover, pro-inflammatory cytokines such as IL-1β, IL-6, TNF-α, and IL-8 tested by ELISA were markedly lower in all GQCDs-coated groups compared to unmodified controls (Fig. 6b–e, Supplementary Fig. 9h), revealing the significantly reduced inflammation of the aspirated bone marrow from implant sites of all GQCDs-coated groups. Together, these results demonstrate that GQCDs efficiently suppress the inflammatory response in the bone marrow microenvironment under infectious conditions, thereby preserving bone integrity and facilitating a niche more conducive to osteointegration.Fig. 6In vivo performance of GQCDs-coated implants in rats. a) Schematic illustration of the in vivo experimental workflow after implant insertion. b-e) Detection of inflammatory cytokines in bone marrow following implantation of Ti, Ta, Ni, and their modified counterparts (Data represent the mean ± s.d., n = 3. **p < 0.01, ***p < 0.001, ****p < 0.0001, analyzed by Unpaired t-test). f-h) Micro-CT reconstruction and osteogenic analysis of rat femurs 8 weeks post-implantation (Data represent the mean ± s.d., n = 3. *p < 0.05, **p < 0.01, analyzed by Unpaired t-test). i) Schematic diagram of hard-tissue sectioning and staining. j) Hard-tissue sections of rat femurs with HE and Goldner's trichrome staining across different groups. k) TEM images of femoral hard tissues from Ti, Ta, Ni, and their modified groups. l, m) Statistical analysis of mitochondrial swelling rate (l) and maximum mitochondrial diameter (m) in femoral hard tissue (Data represent the mean ± s.d., n = 3. *p < 0.05, **p < 0.01, ***p < 0.001, analyzed by Unpaired t-test). In vivo performance of GQCDs-coated implants in rats. a) Schematic illustration of the in vivo experimental workflow after implant insertion. b-e) Detection of inflammatory cytokines in bone marrow following implantation of Ti, Ta, Ni, and their modified counterparts (Data represent the mean ± s.d., n = 3. **p < 0.01, ***p < 0.001, ****p < 0.0001, analyzed by Unpaired t-test). f-h) Micro-CT reconstruction and osteogenic analysis of rat femurs 8 weeks post-implantation (Data represent the mean ± s.d., n = 3. *p < 0.05, **p < 0.01, analyzed by Unpaired t-test). i) Schematic diagram of hard-tissue sectioning and staining. j) Hard-tissue sections of rat femurs with HE and Goldner's trichrome staining across different groups. k) TEM images of femoral hard tissues from Ti, Ta, Ni, and their modified groups. l, m) Statistical analysis of mitochondrial swelling rate (l) and maximum mitochondrial diameter (m) in femoral hard tissue (Data represent the mean ± s.d., n = 3. *p < 0.05, **p < 0.01, ***p < 0.001, analyzed by Unpaired t-test). Clinically, the early postoperative period (24–72 h) represents a critical “window of vulnerability” for bacterial colonization on implants. To verify whether the localized concentration of the modified implants during the initial stage is sufficient to exert potent bactericidal effects, we conducted a longitudinal in vivo antibacterial evaluation using a rat model with titanium implants challenged by S. aureus. The contents of the bone marrow cavity were aspirated on days 1, 3, 10, and 30 post-implantation for agar plating and colony-forming unit (CFU) counting (Supplementary Fig. 10a and b). The results demonstrated that within the first 24 h (the peak colonization phase), only negligible surviving colonies were observed in the GQCDs group, translating to a drastic reduction in bacterial burden of approximately 2 orders of magnitude (∼2 log reduction) compared to the unmodified control. Strikingly, at all subsequent time points (days 3, 10, and up to 30), the GQCDs group exhibited absolutely zero colony growth on the plates. This dynamic in vivo tracking confirms that the GQCDs coating not only swiftly truncates the initial bacterial colonization cascade but also maintains a sterile, infection-free microenvironment within the bone marrow cavity over the long term, demonstrating exceptional in vivo anti-infection proficiency. To elucidate how GQCDs-coated implants modulate the local immune microenvironment during infection, we performed immunohistochemical staining of peri-implant tissues and further quantified the expression of key inflammatory factors and osteogenic proteins within the bone marrow cavity (Supplementary Fig. 10c). GQCDs–coated groups stimulated the robust anti-inflammatory response with decreased expression of TNF-α and increased expression of the anti-inflammatory cytokine IL-10 (Supplementary Fig. 10d–h). Additionally, GQCDs-coated groups showed a significant reduction in INOS (M1 macrophage marker) and increased CD163 (M2 macrophage marker) expression, indicating an enhanced pro-regenerative M2 macrophage phenotype. Consistent with these improvements in the immune environment, expression of osteogenic markers—including OCN, OPG, and Runx2—was significantly upregulated in the GQCDs-coated groups (Supplementary Fig. 11a–d). A major challenge in addressing IAIs is the dysregulated inflammatory environment impeding bone repair. Here, immunohistochemistry demonstrated that GQCDs displayed suppression of TNF-α and IL-6 and promoted a shift from M1 to pro-regenerative M2 macrophage phenotype. This immune reprogramming underlies the observed upregulation of osteogenic markers like OCN and Runx2, promoting a favorable local inflammatory milieu for osteointegration. For detailed evaluation of infection control and osteointegration, micro-computed tomography (micro-CT) was performed at eight weeks post-implantation (Fig. 6f). Rats receiving unmodified implants had severe femoral infections and significant bone destruction. Micro-CT images and 3D reconstructions showed minimal new bone formation, limited to sparse coverage of the implant surfaces, indicating that uncontrolled infection severely compromised integration. Conversely, all GQCDs-coated implants were uniformly surrounded by newly formed bone. Femoral reconstructions confirmed intact morphology without infection. To further precisely quantify the quality of new bone formation and the extent of infection-induced bone destruction, we systematically evaluated the microarchitectural parameters within the rat femur defect model. Regarding osteogenic efficacy, quantitative analysis of the core parameters revealed that the GQCDs-treated group exhibited robust bone regeneration, evidenced by significantly enhanced bone mineral density (BMD) and bone volume fraction (BV/TV) (Fig. 6g and h). Concurrently, prominent increases were also observed in trabecular thickness (Tb.Th) and trabecular number (Tb.N) (Supplementary Fig. 11g and h). Crucially, to quantitatively gauge the severity of infection-induced osteolysis and localized bone defects, we measured the TbCav.CT Value. The results demonstrated a drastic reduction in TbCav.CT Value in the GQCDs group compared to the severe bone resorption seen in the infected control (Supplementary Fig. 11i). The comprehensive improvement across these microarchitectural indices powerfully demonstrates that GQCDs not only efficiently eradicate local infections but also successfully rescue infection-associated localized bone loss by resolving the inflammatory microenvironment, achieving a synergistic anti-infection and pro-healing dual function. To further validate these findings, histological analysis was performed on surgical sections (Fig. 6i). Hematoxylin-eosin staining of unmodified groups revealed substantial inflammatory infiltration around implants. GQCDs–coated groups had minimal inflammation and were well integrated with host bone. Goldner's trichrome staining reinforced this dichotomy: unmodified implants were surrounded by fibrous tissue with little new mineralized bone, whereas GQCDs–coated implants were in direct contact with abundant newly mineralized bone. Using Ti as a representative, unmodified Ti exhibited extensive inflammatory cell accumulation and clear gaps between implant and host bone. Goldner's staining showed scant immature bone. In contrast, Ti-GQCDs group was fully encircled by a continuous layer of mature and mineralized bone, highlighting the exceptional osteointegration afforded by GQCDs modification (Fig. 6j). In light of in vitro findings that GQCDs’ pro-osteogenic effect was mediated by enhanced mitochondrial metabolism, we sought to validate the in vivo mechanism. Mitochondrial ultrastructure of the femoral bone tissue adjacent to implants were examined using TEM (Fig. 6k–m, Supplementary Fig. 11e and f). In groups implanted with unmodified Ti, Ta, Ni, Mg, and PEEK, morphological alterations of mitochondria were observed, including a significant reduction in cell volume, plasma membrane invagination, and loss of intercellular connections. Chromatin within the nucleus exhibited condensation and marginalization, accompanied by dissolution of numerous cellular structures and a marked decrease or even complete disappearance of mitochondrial cristae. These phenomena collectively presented typical morphological features of both apoptosis and ferroptosis, suggesting that unmodified implants can induce significant mitochondrial dysfunction. In contrast, cells in the groups implanted with Ti-GQCDs, Ta-GQCDs, Ni-GQCDs, Mg-GQCDs, and PEEK-GQCDs showed a notable improvement in cellular status. Mitochondrial cristae structure was well-preserved, and mitochondrial volume was significantly increased. These results indicate that GQCDs modification effectively enhances mitochondrial function in the local bone tissue and suppresses infection-induced mitochondrial ferroptosis, which is consistent with the previously demonstrated ferroptosis-inhibitory effects observed in vitro. Furthermore, to comprehensively delineate the spatiotemporal biodistribution dynamics, organ accumulation profiles, and ultimate metabolic fate of GQCDs in vivo, implants modified with Cy5.5-labeled GQCDs were evaluated in the rat implant-associated infection model. In vivo and ex vivo fluorescence tracking was conducted on days 1, 3, 10, 15, and 30 post-implantation (Supplementary Fig. 12a). The imaging results visually revealed that during the acute infection phase (particularly days 1–3), intense fluorescence signals were highly concentrated around the implant and the surrounding infected microenvironment. This characteristic indicates that GQCDs possess an excellent local retention effect during the initial release phase, thereby ensuring a sufficient localized bactericidal concentration for early-stage therapy. Over time (beyond day 3), the focal signal gradually attenuated, indicating the progressive release of GQCDs into the systemic circulation. Ex vivo imaging of major organs (heart, liver, spleen, lungs, and kidneys) (Supplementary Fig. 12b) further unraveled their metabolic trajectory: by day 10, the fluorescence predominantly accumulated in the reticuloendothelial system (RES)-rich liver and spleen, as well as the excretory kidneys, with negligible distribution in the heart and lungs. By day 15, the fluorescence in the liver was virtually cleared, leaving only a faint residual signal in the kidneys. Remarkably, at the day 30 observation endpoint, the fluorescence intensities in all harvested organs had entirely reverted to baseline background levels. This time-dependent metabolic pattern—characterized by the rapid clearance from the liver/spleen and relatively prolonged presence in the kidneys—strongly suggests that GQCDs are safely eliminated from the body primarily via the renal clearance pathway. Crucially, this month-long dynamic tracking unequivocally confirms the absence of any persistent long-term tissue accumulation in major organs, demonstrating an exceptionally favorable in vivo biosafety profile. Finally, histological analysis of liver and kidney sections from all groups revealed no signs of toxicity, confirming the excellent in vivo biocompatibility of the GQCDs-coated implants (Supplementary Fig. 12c and d). These results collectively demonstrate that GQCDs enhance bone tissue regeneration by ameliorating the local bone tissue microenvironment with whole biosafety. In vivo validation of Ti-GQCDs in a beagle dog model An implant-associated infection model in 10-month-old beagle dogs was established to validate the significantly improved effectivities of GQCDs modified implants (Fig. 7a). Custom-sized Ti rods (4 mm × 22 mm) were fabricated to fit the beagle femur and coated with GQCDs. Mechanical property tests demonstrated that the compressive and tensile performance of Ti-GQCDs was comparable to that of unmodified Ti, as evidenced by the overlapping compression-displacement and tension-displacement curves (Fig. 7b). Moreover, no significant changes were observed in the tensile modulus (Supplementary Fig. 13a) or peak compressive stress (Supplementary Fig. 13b). These results confirm that the coated GQCDs do not compromise the structural integrity of large-scale implants. Furthermore, water contact angle measurements showed that the Ti-GQCDs surface exhibited significantly more hydrophilic, a property favorable for subsequent osteointegration (Fig. 7c).Fig. 7In vivo performance of Ti-GQCDs in beagle dogs. a) Schematic illustration of the beagle dog model for bone implantation with infection. b) Mechanical properties of Ti and Ti-GQCDs. c) Water contact angle of the modified implants (Data represent the mean ± s.d., n = 3. **p < 0.01, analyzed by Unpaired t-test). d) Surgical procedure for beagle dog modeling. e, f) Gait analysis of beagle dogs in different groups 3 months post-implantation. g, h) Micro-CT reconstruction (g) and osteogenic analysis (h) of beagle dog femurs at 12 weeks post-implantation (Data represent the mean ± s.d., n = 3. **p < 0.01, ***p < 0.001, ****p < 0.0001, analyzed by Unpaired t-test). i, j) Hard-tissue sections of femurs with HE and Goldner's trichrome staining across different groups. k) Harvesting and Masson staining of gastrocnemius muscle from beagle dogs. l) TEM images of femoral hard tissue. In vivo performance of Ti-GQCDs in beagle dogs. a) Schematic illustration of the beagle dog model for bone implantation with infection. b) Mechanical properties of Ti and Ti-GQCDs. c) Water contact angle of the modified implants (Data represent the mean ± s.d., n = 3. **p < 0.01, analyzed by Unpaired t-test). d) Surgical procedure for beagle dog modeling. e, f) Gait analysis of beagle dogs in different groups 3 months post-implantation. g, h) Micro-CT reconstruction (g) and osteogenic analysis (h) of beagle dog femurs at 12 weeks post-implantation (Data represent the mean ± s.d., n = 3. **p < 0.01, ***p < 0.001, ****p < 0.0001, analyzed by Unpaired t-test). i, j) Hard-tissue sections of femurs with HE and Goldner's trichrome staining across different groups. k) Harvesting and Masson staining of gastrocnemius muscle from beagle dogs. l) TEM images of femoral hard tissue. As shown in Fig. 7d, pure Ti and Ti-GQCDs were implanted into the 4 mm hole in the distal femur with MRSA infection. After 8 weeks of implantation, the locomotor performance of all animals was evaluated through systematic gait analysis. Sequential analysis of dynamic locomotion revealed significant disparities in motor function recovery between the two groups. Dogs in the pure Ti group exhibited distinct signs of paresis in the implanted hindlimb, characterized by impaired weight-bearing and an uncoordinated gait. These motor deficits indicate that persistent infection severely compromised functional recovery. Conversely, the Ti-GQCDs group demonstrated a robust restoration of motor function. These animals achieved normal weight-bearing on the operated hindlimb and displayed a coordinated, balanced gait with a regular pace, signifying a successful functional outcome (Fig. 7e). Furthermore, dynamic reconstruction of the implanted hindlimb positions throughout the gait cycle provided a clearer and more accurate visualization of locomotor recovery in the experimental animals (Fig. 7f), corroborating that Ti-GQCDs implantation led to a significant improvement in motor function. Gastrocnemius muscle samples were harvested at 12 weeks for Masson's trichrome staining. Compared with the healthy contralateral side, the Ti group displayed evident muscle atrophy, with Masson staining revealing substantial blank spaces due to myofiber atrophy. In contrast, the Ti-GQCDs group exhibited favorable muscle recovery, showing significant improvement over the Ti group. Analysis of the mean muscle fiber area further confirmed these findings (Supplementary Fig. 13c), indicating superior recovery of motor function in the Ti-GQCDs group (Fig. 7k). Micro-CT analysis of the femurs at 12 weeks provided striking evidence of the coating's efficacy (Fig. 7g). Unmodified Ti implants were associated with large, infection-induced osteolytic defects and were covered by only minimal new bone. In stark contrast, the femurs of dogs with Ti-GQCDs showed no signs of infectious defects, and the implant surfaces were extensively covered with new bone. Quantitative analysis confirmed that BMD, Tb.Th, BV/TV and Tb.N were significantly higher in the Ti-GQCDs group than that of the unmodified Ti group. In parallel, the parameters indicative of bone resorption (TbCav.CT) were also markedly decreased, collectively underscoring the GQCDs coating's dual protective and osteogenic effects. (Fig. 7h, Supplementary Fig. 13d and e). Hard tissue histology provided further evidence of enhanced osteointegration. HE and Goldner's trichrome staining showed significant inflammatory cell infiltration and poor bone apposition around unmodified Ti. In contrast, Ti-GQCDs were tightly surrounded by newly formed, well-integrated bone (Fig. 7i and j). Finally, TEM of the peri-implant bone tissue revealed that infection in the unmodified Ti group led to poor cellular integrity and dissolution of the peri-cellular structure. In contrast, the Ti-GQCDs group exhibited distinct mineralized crystals surrounding the cells, indicating that Ti-GQCDs significantly enhanced bone mineralization (Fig. 7l). This confirms that GQCDs-coated implants maintain stable osteogenic-promoting performance in the large-animal model. To visually assess the antibacterial efficacy of GQCDs in beagle dogs, peri-implant bone tissues were aseptically harvested at the experimental endpoint, and the bacterial load was quantitatively determined as colony-forming units per gram of tissue (CFU/g). The quantitative analysis revealed striking inter-group disparities: the untreated infected control group sustained a severe and persistent bacterial burden throughout the observation period, with bacterial loads in the surrounding bone reaching 2.47 × 106 CFU/g and 4.70 × 106 CFU/g on days 3 and 30, respectively, indicating stubborn colonization and proliferation within the local bone defect. In stark contrast, the bone tissue homogenates from the GQCDs-treated group exhibited absolutely no bacterial growth (zero CFU/g) on agar plates at both critical time points, achieving complete sterilization of the local tissue microenvironment (Supplementary Fig. 13f–h). This compelling outcome decisively demonstrates that even in the highly complex and mechanically demanding in vivo environment of a large animal model, the GQCDs coating can exert potent and durable bactericidal effects, underscoring its immense potential for clinical translation. Moreover, to further validate the anti-infective efficacy of Ti-GQCDs, hematological analysis was performed. The results revealed that the Ti group exhibited typical signs of chronic bacterial infection, characterized by elevated white blood cell counts (Supplementary Fig. 14a), increased neutrophil percentages (Supplementary Fig. 14b), and reduced hemoglobin levels indicative of anemia (Supplementary Fig. 14c). In contrast, implantation of Ti-GQCDs led to decreased white blood cell counts, lower neutrophil proportions, and amelioration of anemic conditions (Supplementary Fig. 14a–c). These findings demonstrate that Ti-GQCDs exerted stable in vivo antibacterial activity and suppressed the development of chronic bacterial infection. Serological analysis further corroborated these results. The bacterial infection related markers of C-reactive protein (CRP, Supplementary Fig. 14d), IgG-2 (Supplementary Fig. 14e), and IgM-2 (Supplementary Fig. 14f), and the indicator of chronic inflammation IgA-2 (Supplementary Fig. 14g) were reduced in the Ti-GQCDs group than those in the Ti group. Together, these data confirm the anti-inflammatory and anti-infective properties of GQCDs-coated implants in large animals. In addition, key indicators reflecting nutritional status (albumin, ALB) and hepatic and renal function-including alanine aminotransferase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (UREA), and creatinine (CREA)—were evaluated. The results showed no significant differences or abnormal elevations between the Ti and Ti-GQCDs groups (Supplementary Fig. 14h–l), indicating that GQCDs modification did not induce notable hepatic or renal impairment in the large-animal model. Furthermore, histopathological examination of vital organs (heart, liver, spleen, lungs, and kidneys) via H&E staining revealed no obvious pathological alterations; tissue architecture remained intact and function appeared normal (Supplementary Fig. 14m). Collectively, these findings demonstrate that Ti-GQCDs implants retain favorable systemic biosafety while delivering stable antibacterial and osteogenic-promoting performance. Limitations Admittedly, while this study comprehensively substantiates the “anti-infection and pro-healing” dual benefits of GQCDs, several hurdles remain before ultimate clinical translation can be realized. First, the long-term in vivo degradation kinetics and deep-tissue clearance mechanisms of GQCDs necessitate complete elucidation through extended longitudinal tracking. Second, although the GQCDs coatings demonstrated favorable interfacial stability herein, their tribological durability under long-term (e.g., >6 months) complex physiological mechanical loading—particularly in large animal models—requires further verification. Furthermore, the capacity of GQCDs to avert bacterial resistance evolution within the authentic, complex in vivo microenvironment demands more rigorous assessment beyond in vitro serial passaging. Finally, it is noteworthy that the five implant substrates used in this study—Ti [35], Ta [36], Ni [37], Mg [38], and PEEK [39]—possess well-established clinical safety profiles and are associated with a minimal risk of severe foreign body reactions. However, it should be acknowledged that the current experimental design lacked a sham-implant absolute infection control group. Consequently, while current data sufficiently prove the efficacy of the coating intervention, the extent to which local osteolysis is independently driven by bacterial infection versus synergistically exacerbated by residual mild FBR cannot be absolutely and precisely decoupled. Future iterations will incorporate refined, isolated control cohorts to accurately quantify the independent contributions of infection, FBR, and nano-intervention on bone metabolism. In addition, the relatively small sample size represents another limitation of this study. Although the current sample size (n = 3) was sufficient to demonstrate the profound antibacterial and osteogenic efficacy of GQCDs with statistical significance, we acknowledge the limitation imposed by this relatively small cohort. Given the individual variability inherent in large animal infection models, future studies with expanded sample sizes are warranted to further validate the long-term clinical safety and efficacy of the coatings. Ultimately, given the inherent species disparities in immune responses and bone remodeling rates between rodent/canine models and humans, the clinical safety and efficacy of GQCDs await definitive validation through rigorous prospective clinical trials. Conclusion In this study, we innovatively engineered guanidino/quaternary ammonium-derived carbon quantum dots (GQCDs) endowed with a “spatiotemporal sequential targeting” capability, systematically validating their immense potential to synergistically couple potent antibacterial efficacy with enhanced osteointegration in implant-associated infections. The core design rationale of GQCDs elegantly capitalizes on a fundamental evolutionary divergence regarding the spatial segregation of energy metabolism systems: the bacterial respiratory chain is localized to the cell membrane, whereas the eukaryotic respiratory chain resides entirely within the inner mitochondrial membrane. Driven by this profound anatomical disparity, GQCDs exhibit remarkable smart-responsive adaptability within the complex infection microenvironment. During the initial phase, they preferentially target and disrupt negatively charged bacterial membranes via strong electrostatic interactions, paralyzing respiratory chain functions to achieve robust sterilization. Subsequently, upon neutralizing the local infection threat, GQCDs are directed by the mitochondrial membrane potential (ΔΨm) to accumulate specifically within the host's osteoblastic mitochondrial network, thereby remodeling energy metabolism and promoting bone regeneration. Mechanistically, GQCDs orchestrate a multidimensional networked intervention on both pathogens and host cells. Transcriptomic and molecular docking analyses collectively revealed that on the bacterial front, GQCDs not only bind with exceptionally high affinity to crucial respiratory chain proteins (coenzyme Q and cytochrome c oxidase) but also simultaneously repress ABC transporters, disrupt peptidoglycan synthesis and DNA replication, and markedly downregulate resistance genes (blaR1). This pleiotropic, multi-target paradigm fundamentally circumvents the inherent vulnerability of conventional antibiotics, which easily provoke resistance mutations due to their single-target mechanisms. On the host front, multi-omics data indicate that GQCDs trigger beneficial metabolic reprogramming in osteoblasts by upregulating amino acid and glutathione metabolic pathways. Crucially, under severe S. aureus infection stress, GQCDs dose-dependently upregulate the GPX4 and System Xc− antioxidant axis, significantly enhancing cellular respiration and thoroughly blockading the infection-induced ferroptosis cascade. This mechanism robustly shields the osteogenic potential of host cells even in a highly hostile inflammatory microenvironment. In summary, GQCDs—designed based on eukaryotic-prokaryotic structural distinctions—exhibit synergistic antibacterial and osteogenic effects, highlighting their potential as an ideal surface-modifying agent for implants. To assess clinical applicability, GQCDs were successfully loaded onto five commonly used orthopedic implants of Ti, Ni, Mg, Ta, and PEEK. The modified implants consistently demonstrated stable antibacterial and pro-osteogenic functions in vitro. Further in vivo evaluation using rat and beagle dog models of IAIs confirmed that GQCDs modification not only preserved the inherent biosafety but also significantly enhanced their antibacterial and osteointegrative performance. These findings underscore the considerable research value and broad clinical potential of GQCDs and GQCDs modified implants in the prevention and treatment of orthopedic infections. Methods Materials and reagents Anhydrous citric acid (CA) and diallyldimethylammonium chloride (DMDAAC) were obtained from Aladdin Reagent Co., Ltd. (Shanghai, China). Polyhexamethylene guanidine (PHMG) was purchased from Shanghai Macklin Biochemical Technology Co., Ltd. Polyvinyl butyral (PVB) was supplied by China National Pharmaceutical Group Co., Ltd. Mueller–Hinton (MH) broth and nutrient agar were acquired from Qingdao Hope Bio-Technology Co., Ltd. Bacterial strains—Staphylococcus aureus (ATCC 25923), Staphylococcus epidermidis (ATCC 12228), methicillin-resistant S. aureus (MRSA, ATCC 43300), Pseudomonas aeruginosa (ATCC 27853), Escherichia coli (ATCC 25922), and Enterobacter cloacae (ATCC 700323)—were obtained from the American Type Culture Collection (ATCC). Polyetheretherketone (PEEK) discs of various specifications were sourced from GEHR (Germany). Titanium (Ti), tantalum (Ta), nickel (Ni), and magnesium (Mg) implants were supplied by Taiyuxin Metal Factory. Male Sprague-Dawley rats (8 weeks old) and male beagles (10 months old) were provided by the Animal Center of the Fourth Military Medical University. All animal procedures were approved by the Institutional Animal Care and Use Committee of the Fourth Military Medical University. Synthesis of GQCDs Guanidine–coated carbon quantum dots (GQCDs) were synthesized via a hydrothermal method using anhydrous citric acid (CA) as the carbon source. Briefly, 0.4 g of CA and 0.2 g of polyhexamethylene guanidine (PHMG) were dissolved in 16 mL of ultrapure water under stirring. Then, 4 mL of diallyldimethylammonium chloride (DMDAAC) was added, and the mixture was stirred for an additional 10 min to obtain a homogeneous solution. The solution was transferred into a 50 mL Teflon-lined autoclave and reacted at 180 °C for 6 h, followed by natural cooling to room temperature. The resulting product was filtered through a 0.22 μm microporous membrane. The filtrate was collected and purified by dialysis (500–1000 Da molecular weight cut-off) against ultrapure water for 72 h. The dialyzed solution was aliquoted into centrifuge tubes, lyophilized to obtain GQCDs as a solid powder, and stored sealed at 4 °C for further use. Characterization of GQCDs The morphology and size distribution of GQCDs were examined using transmission electron microscopy (TEM). The average height of GQCDs was determined by atomic force microscopy (AFM). The graphitic structure was verified by X-ray powder diffraction (XRD). Surface functional groups were analyzed by Fourier-transform infrared spectroscopy (FTIR). Structural features were confirmed by X-ray photoelectron spectroscopy (XPS). The relative molecular weight was assessed by gel permeation chromatography (GPC). Zeta potentials of GQCDs, six bacterial strains, and bacteria treated with GQCDs were measured using a zeta potential analyzer. Bacterial culture and suspension preparation Staphylococcus epidermidis (ATCC 12228), Staphylococcus aureus (ATCC 25923), methicillin-resistant S. aureus (MRSA, ATCC 43300), Pseudomonas aeruginosa (ATCC 27853), Escherichia coli (ATCC 25922), and Enterobacter cloacae (ATCC 700323) were streaked onto agar plates and incubated at 37 °C to obtain single colonies. Isolated colonies were picked and suspended in sterile saline (0.9% NaCl) to prepare bacterial suspensions. The turbidity was adjusted to a 0.5 McFarland standard using a densitometer to achieve a final concentration of approximately 1.5 × 108 colony-forming units (CFU) per mL for each strain. Determination of minimum inhibitory concentration (MIC) Mueller–Hinton (MH) broth containing serial dilutions of GQCDs was prepared. Sterile tubes were labeled, and 1 mL of MH broth with varying GQCDs concentrations was mixed with 1 mL of bacterial suspension (1.5 × 106 CFU/mL). Negative control tubes contained 2 mL of MH broth only, while positive control tubes contained 1 mL of MH broth and 1 mL of bacterial suspension. Tubes were incubated at 37 °C for 24 h. The lowest concentration showing no visible bacterial growth was recorded as the MIC for each strain. The MICs of GQCDs precursors—anhydrous citric acid (CA), diallyldimethylammonium chloride (DMDAAC), and polyhexamethylene guanidine (PHMG)—were similarly determined against MRSA as an example. Bacterial resistance evaluation Serial passage assays were performed using Staphylococcus aureus (S. aureus) and methicillin-resistant Staphylococcus aureus (MRSA). The minimum inhibitory concentration (MIC) of GQCDs against both strains was first determined as described in Section 4.5. Bacterial suspensions from wells containing subinhibitory concentrations (one dilution below the MIC) were inoculated into 10 mL of fresh Mueller–Hinton (MH) broth supplemented with the same concentration of GQCDs and incubated at 37 °C for 24 h. After incubation, bacterial cells were harvested by centrifugation at 8000 rpm for 10 min at 4 °C, washed twice with sterile PBS to remove residual GQCDs, and resuspended in fresh MH broth. The MIC of GQCDs against the passaged strains was then re-determined. This procedure was repeated for 30 consecutive passages, and the MIC values were recorded at each passage to monitor resistance development. Vancomycin was used as a positive control under identical conditions. Time–kill curves Following the MIC assay setup, 200 μL aliquots were sampled from each tube at 2 h intervals for the first 24 h, then every 12 h up to 48 h. The optical density at 600 nm (OD600) was measured using a microplate reader, and time–kill curves were plotted. Bacterial colony-forming unit (CFU) assay Bacterial suspensions prepared as in Section 4.4 were treated with GQCDs at specified concentrations. Then, 100 μL of each mixture was spread evenly on agar plates using a sterile spreader. Plates were incubated inverted at 37 °C for 24 h, after which colony growth was observed and recorded. Stability assessment of GQCDs MIC Stability: Lyophilized GQCDs samples were stored under dry conditions. At weeks 0, 2, 4, 6, 8, 10, 12, 14, 16, 18, and 20, MICs against all six bacterial strains were determined to evaluate antimicrobial stability over time. Storage Stability: GQCDs stored for 4, 8, and 12 weeks were tested in bacterial culture assays to confirm retained antibacterial activity. Batch-to-Batch Consistency: GQCDs synthesized in three separate batches (months 1–3) were subjected to bacterial culture assays to verify the reproducibility of the synthesis method. Anti-biofilm activity of GQCDs Biofilm Inhibition Assay: In a 96-well plate, 50 μL of bacterial suspension (1.5 × 108 CFU/mL) was mixed with 50 μL of MH broth containing various concentrations of GQCDs. After 48 h incubation at 37 °C, non-adherent cells were removed by washing three times with saline. Biofilms were fixed with methanol for 15 min, stained with 0.1% crystal violet for 15 min, washed again, and dissolved in 33% glacial acetic acid. Biofilm biomass was quantified by measuring OD570. Mature Biofilm Eradication Assay: Bacterial biofilms were pre-formed by incubating 100 μL of bacterial suspension (1.5 × 108 CFU/mL) in 96-well plates for 48 h at 37 °C. After washing, fresh MH broth containing GQCDs at different concentrations was added, and plates were incubated for an additional 24 h. Biofilm biomass was quantified using the crystal violet staining method as above. Validation of antibacterial performance SEM: Bacterial suspensions (1.5 × 106 CFU/mL) of each strain were prepared. A 20 μL aliquot of each suspension was placed on a glass coverslip in a 24-well plate. Controls received 1 mL of plain MH broth, while test groups received MH broth containing 60 μg/mL GQCDs. After 24 h incubation, bacterial morphology on the coverslips was examined by SEM. Live/Dead Staining Assay: Bacterial suspensions (100 μL, 1.5 × 106 CFU/mL) were added to confocal dishes and incubated for 24 h. After washing with PBS to remove non-adherent cells, fresh MH broth with different concentrations of GQCDs was added and incubated for another 24 h. Cells were then stained with DAPI (0.5 μg/mL final concentration) and propidium iodide (PI, 5 μg/mL final concentration) for 30 min, washed, and imaged using confocal laser scanning microscopy (CLSM). Effect of GQCDs on bacterial ATP levels Bacterial cultures were grown overnight in MH broth at 37 °C, harvested by centrifugation (3000 rpm, 15 min), and washed three times with saline. The pellet was resuspended in saline to an OD600 of 0.5. Then, 100 μL of GQCDs solution at various concentrations was added to 900 μL of bacterial suspension and incubated at 37 °C for 4 h. Cells were pelleted by centrifugation (3000 rpm, 15 min), and ATP levels in the supernatant were measured using an ATP assay kit. Bacterial transcriptomics Each of the six bacterial strains was cultured in 15 mL centrifuge tubes. For each strain, three replicates were prepared for both control (plain MH broth) and treatment groups (MH broth containing 30 μg/mL GQCDs). After 24 h co-incubation, cultures were centrifuged at 9000 rpm for 20 min. The supernatant was discarded, and pellets were flash-frozen in liquid nitrogen before being shipped on dry ice for transcriptomic sequencing. Molecular docking analysis Based on the structural characterization of GQCDs, their molecular structure was drawn using Chem3D and optimized with MOPAC to calculate PM3 atomic charges. The ligand structure was prepared using AutoDock Tools 1.5.6. Global molecular docking between GQCDs and bacterial target proteins was performed with AutoDock 4.2.6. The docking grid was centered on the protein with dimensions 100 × 100 × 100 points, encompassing the entire protein structure. Fifty docking runs were performed for each complex, with other parameters set to default values. Validation of antibacterial mechanisms Glutathione Assay: Each bacterial strain was cultured in 1.5 mL tubes with plain MH broth (control) or MH broth containing 30, 60, or 90 μg/mL GQCDs (treatment). After 48 h incubation, glutathione levels were measured using an ELISA kit. Reactive Oxygen Species Detection by Flow Cytometry: Following the same treatment protocol as above, bacterial cultures were incubated for 48 h, and intracellular ROS levels were quantified by flow cytometry. Cytotoxicity of GQCDs CCK-8 Assay: MC3T3-E1 cells were seeded in 96-well plates at 8 × 103 cells per well. After 24 h, the medium was replaced with fresh medium containing various concentrations of GQCDs. Cell viability was assessed after 1 and 4 days of co-incubation by measuring the absorbance at 450 nm using a microplate reader. Osteogenic activity of GQCDs RT-qPCR Analysis: MC3T3-E1 cells were seeded in 6-well plates at 5 × 105 cells per well. After 24 h, the medium was replaced with MEM containing different concentrations of GQCDs. Following 48 h of treatment, total RNA was extracted, and the expression of osteogenic markers Runx2 and OCN was quantified by RT-qPCR. ATP Assay: Cells were treated as described above. After 48 h, extracellular ATP levels in cultures treated with 0, 30, 60, and 90 μg/mL GQCDs were measured using an ATP assay kit. Immunofluorescence Staining: MC3T3-E1 cells were seeded on confocal dishes at 5 × 105 cells per dish. After 24 h, cells were treated with different concentrations of GQCDs for 24 h, then fixed and stained with antibodies against OCN and Runx2, along with an ROS probe. Stained cells were imaged by CLSM to evaluate the effects of GQCDs on osteogenic differentiation. Transcriptomics and metabolomics of osteoblasts MC3T3-E1 cells were cultured in 10 cm dishes until confluent, then subcultured at a 1:4 ratio. When cells reached approximately 80% confluence, the medium was replaced with osteogenic medium containing different concentrations of GQCDs. After 24 h, cells were scraped, pelleted by centrifugation (1000 rpm, 5 min), flash-frozen in liquid nitrogen, and shipped on dry ice for combined transcriptomic and metabolomic analysis. Mechanism of GQCDs-Mediated osteogenesis RT-qPCR for Mitochondrial Markers: MC3T3-E1 cells were seeded and treated as in Section 4.17. After 48 h, total RNA was extracted, and the expression of ACSL4、TFR1、COX4 and GPX4 was measured by RT-qPCR. Flow cytometry detection: Cells were cultured in 10 cm dishes and treated with GQCDs when approximately 70% confluent. After 24 h, cells were harvested and stained with JC-1 working solution for 20 min in the dark. Mitochondrial membrane potential was analyzed by flow cytometry. After staining with C11-BODIPY 581/591 for 30 min under the same conditions, the level of oxidized C11-BODIPY was determined by flow cytometry. Fluorescence Staining Assays:Osteoblasts were seeded into confocal 24-well plates at a density of 3 × 104 cells/well. For ROS detection, cells were treated with GQCDs for 12 h, followed by incubation with a diluted ROS probe (5 μM final concentration) for 30 min in the dark. Fluorescence images were acquired using a confocal microscope. For intracellular labile iron detection, osteoblasts were seeded as above and subjected to different treatments: (i) untreated control, (ii) positive control (infection with S. aureus for 12 h), and (iii) experimental group (co-incubation with S. aureus and GQCDs for 12 h). After treatment, cells were stained with FerroOrange for 30 min. Changes in intracellular labile iron were visualized under a confocal microscope, and fluorescence intensity was quantified using ImageJ software. Glutathione Assay: MC3T3-E1 cells seeded in 24-well plates (5 × 104 cells/well) were treated with GQCDs for 24 h, after which intracellular glutathione levels were measured using an ELISA kit. Fluorescence Staining: Cells seeded in 24-well plates (3 × 104 cells/well) were treated with GQCDs for 24 h, then incubated with a diluted ROS probe (5 μM final concentration) for 30 min in the dark. Fluorescence images were acquired using a fluorescence microscope, and signal intensity was quantified with a multifunctional plate reader. Mitochondrial colocalization assay To verify the mitochondrial targeting of GQCDs, a mitochondrial colocalization assay was performed using high-resolution laser-scanning confocal microscopy (LSCM). Osteoblasts were seeded into confocal culture dishes and allowed to adhere for 24 h. GQCDs, which possess intrinsic blue fluorescence, were then added and co-incubated with the cells for 12 h. Following incubation, the cells were stained with the mitochondria-specific fluorescent probe MitoTracker™ Deep Red FM at 37 °C for 30 min in the dark. After staining, the cells were washed three times with pre-warmed PBS to remove unbound probes and subsequently imaged using a super-resolution confocal microscope. The acquired images were analyzed using ImageJ software for colocalization analysis, and the Pearson correlation coefficient and Manders overlap coefficient were calculated to quantitatively assess the spatial colocalization between GQCDs and mitochondria. Zeta potential of osteoblasts MC3T3-E1 cells were cultured in 15 mL centrifuge tubes and grouped based on treatment with GQCDs, E. coli, or S. aureus. After 12 h of co-culture, cells were harvested by centrifugation (1000 rpm, 5 min), resuspended in ultrapure water, and zeta potential was measured. Transcriptomics of infected osteoblasts MC3T3-E1 cells were seeded in 6-well plates at 5 × 105 cells per well and cultured overnight in α-MEM supplemented with 10% FBS (without antibiotics). Cells were co-cultured with S. aureus or E. coli for 10 h, then treated with osteogenic medium containing 60 μg/mL GQCDs for an additional 12 h. Non-adherent bacteria were removed by washing, and infected osteoblasts were collected, flash-frozen, and shipped on dry ice for transcriptomic analysis. Differentially expressed genes were defined as those with a fold change ≥2 and a Q-value <0.05. Effect of GQCDs on mitochondrial function in infected osteoblasts A transwell co-culture system was used: MC3T3-E1 cells (5 × 104 cells/well, lower chamber) were co-cultured with S. aureus (5 × 103 cells/well, upper chamber). After 24 h, GQCDs were added to the medium. Following another 24 h incubation, cells were washed and stained with mitochondria-specific and lysosome-specific fluorescent probes for 30 min. After washing, cells were imaged using high-resolution CLSM. The intrinsic blue fluorescence of GQCDs, combined with red and green organelle signals, was used to assess mitochondrial localization and effects. Transmission electron microscopy of osteoblasts MC3T3-E1 cells (5 × 104 cells/well) were co-cultured with S. aureus (5 × 103 cells/well) in a transwell system. When cells reached approximately 50% confluence, they were treated with osteogenic medium containing different concentrations of GQCDs for 24 h. Cells were then scraped, pelleted by centrifugation (1000 rpm, 5 min), and fixed in glutaraldehyde overnight at 4 °C before processing for TEM observation. Competitive binding assay of osteoblasts and E. coli in vitro Osteoblasts were seeded into confocal culture dishes and cultured for 24 h prior to treatment. The experiment was divided into the following groups: (1) untreated control group; (2) bacterial infection group, co-incubated with E. coli for 3 h; and (3) GQCDs competitive co-culture group, in which E. coli and GQCDs were added simultaneously and co-incubated for 3 h or 6 h. Bacterial viability and membrane integrity were assessed using the SYTO 9/PI live/dead bacterial staining kit according to the manufacturer's instructions. Mitochondria were labeled with MitoTracker Green, and osteoblast plasma membranes were stained with CellMask™ (Invitrogen). After staining, cells were washed three times with pre-warmed PBS to remove unbound probes, and then imaged and photographed using a confocal microscope. Fluorescence intensity was quantitatively analyzed using ImageJ software. Western blot analysis The osteoblasts infected with Staphylococcus aureus were collected for WB analysis, which was consistent with the method in 2.21, total protein was obtained and homogenized in RIPA buffer (#P0013B, Beyotime), and protein concentrations were determined by a BCA assay kit (#P0012, Beyotime). Then, proteins were separated by 10% SDS polyacrylamide gel for electrophoresis and subsequently transferred electrophoretically toandpolyvinylidene difluoride membranes (Millipore Corp., Bedford, MA, USA). After blocking with 5% skim milk for 1 h, membranes were incubated overnight at 4 °C with primary antibodies against -xCT (1:1000,ab307601, Abcam, UK), GPX4(1:1000,ab125066,Abcam, UK), and GAPDH (1:1000,#AF7021, AffinityBiosciences). The membranes were incubated with HRP-conjugated anti-rabbits econdary antibody (1:10000, #7074, Cell Signaling Technology) or anti-mouse secondary antibody (1:10000, #ab6789, Abcam) for 1 h. The visualization was performed by an ECL system (GE ImageQuant 350, GE Healthcare, Waukesha, WlUSA). Validation of osteogenic mechanisms of GQCDs under infected conditions MC3T3-E1 osteoblasts were cultured in 10 cm dishes until confluent, then subcultured at a 1:4 ratio into new dishes. When cells reached approximately 70% confluence, they were treated with GQCDs, liproxstatin-1 (Lip-1), or Staphylococcus aureus as indicated. After 24 h, cell viability was assessed using a CCK-8 assay. Intracellular reactive oxygen species (ROS) levels were quantified by flow cytometry, and malondialdehyde (MDA) and glutathione (GSH) levels were measured using commercial assay kits according to the manufacturers’ instructions. Seahorse assay for mitochondrial respiration Oxygen consumption rate (OCR), basal respiration, maximal respiration, and ATP production were measured using a Seahorse XF96 Pro Analyzer (Agilent). Osteoblasts were seeded at 1 × 104 cells per well in a Seahorse microplate. After 24 h, cells were infected with S. aureus (3 × 104 CFU/well) and treated with GQCDs (60 μg/mL) for an additional 12 h. Before analysis, the microplate was equilibrated for 15 min at 37 °C in a non-CO2 incubator. Mitochondrial respiration was monitored in real-time by sequential injection of mitochondrial modulators: oligomycin (5 μM), carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone (FCCP, 1 μM), and rotenone/antimycin A (2 μM). After the assay, cells were stained with crystal violet (Sigma-Aldrich) for normalization. OCR-derived parameters were calculated using Wave software (version 2.6.4.24). Surface modification of implant materials with GQCDs Polyvinyl butyral (PVB) was dissolved in a mixture of ethyl acetate and ethanol (1:9 v/v) to prepare a 5 mg/mL PVB solution. GQCDs were dispersed in this solution at different concentrations. Sterilized implant discs (Ti, Ta, Ni, Mg, PEEK; 11 mm × 2 mm) were placed in 24-well plates, immersed in 1 mL of the GQCDs-PVB solution, and dried at 60 °C in an oven. The coating process was repeated three times to obtain a uniform GQCDs–coated surface. Cytotoxicity of modified implant materials Osteoblasts were seeded onto implant discs modified with 50, 100, or 200 μg/mL GQCDs in 24-well plates. Cell viability was evaluated using a CCK-8 assay on days 1 and 4 to assess the cytotoxicity of each modified material (Ti, Ta, Ni, Mg, PEEK). Characterization of modified implants Surface Morphology: Surface topography was examined using field-emission scanning electron microscopy (FE-SEM, Zeiss Sigma HD). Water Contact Angle: Hydrophilicity was evaluated by contact angle measurements (KRÜSS DSA100) on three independent replicates per group. Release Profile: The intrinsic fluorescence of GQCDs (excitation 310 nm, emission 393 nm) was used to quantify their release from Ti-GQCDs, Ta-GQCDs, Ni-GQCDs, Mg-GQCDs, and PEEK-GQCDs into simulated body fluid over time. Mechanical Properties: Stress–displacement curves of bare and GQCDs–coated implants were recorded using an electronic universal testing machine (MTS E45.105). Antibacterial performance of modified implants Quantitative Antibacterial Assay: Bacterial suspensions (1.5 × 106 CFU/mL) of S. aureus, S. epidermidis, MRSA, P. aeruginosa, E. coli, and E. cloacae were prepared. Each modified disc was placed on an agar plate, inoculated with 60 μL of bacterial suspension, and incubated for 24 h at 37 °C. Discs were then transferred into 5 mL of sterile saline, vortexed for 2 min, and the resulting suspensions were serially diluted. Aliquots (100 μL) were spread on agar plates, incubated for 16 h, and colony-forming units (CFU) were counted. Morphological Observation by SEM: Discs were placed in 24-well plates, inoculated with 60 μL of bacterial suspension (1.5 × 106 CFU/mL), and covered with 1 mL of broth. After 24 h, discs were washed with PBS, fixed with 2.5% glutaraldehyde at 4 °C overnight, dehydrated through a graded ethanol series (30%, 50%, 75%, 90%, 95%, 100%), dried with hexamethyldisilazane (HMDS), sputter-coated with gold, and imaged by SEM. Coating thickness characterization and interfacial stability assessment of modified implants The thicknesses of unmodified and GQCDs-coated implants were measured using a vernier caliper to evaluate coating-induced thickness changes. In parallel, the surface loading of GQCDs on the coated implants was quantitatively analyzed based on the in vitro drug release profile of GQCDs. To evaluate the mechanical and interfacial stability of the coating under physiologically relevant conditions, GQCDs–coated implants were subjected to the following treatments: (1) immersion in simulated body fluid (SBF) for 24 h, and (2) high-power ultrasonication in distilled water for 15 min. Following treatment, the antibacterial activity of the coated implants was assessed by zone-of-inhibition assays and bacterial colony culture assays to verify the post-treatment antibacterial stability of the coating. Osteogenic performance of modified titanium (Ti-GQCDs) Cell Morphology by SEM: Osteoblasts and macrophages were seeded onto bare Ti and Ti-GQCDs discs. After 24 h, samples were washed, fixed with 2.5% glutaraldehyde, dehydrated, and observed by SEM. Immunofluorescence Staining: Osteoblasts cultured on modified discs for 24 h were fixed, permeabilized, and stained with primary antibodies against OCN and Runx2, followed by fluorescent secondary antibodies. Fluorescence intensity was quantified to evaluate osteogenic differentiation. Sample size estimation and ethical statement The sample size calculation was based on the study design and was performed using online software (Experimental design assistant; https://eda.nc3rs.org.uk/eda/login/auth). The power of the experiment was set to 80%. Two-tailed student's t-tests with a significance threshold of 0.05 were used. To minimize circadian influence, all experiments were performed between 8:00 a.m. and 12:00 a.m. after 7 days of acclimatization. The experimental protocol was approved by the Local Ethical Committee for Animal Experiments at the Air Force Medical University (Ethics Committee Approval Number: 20250032) and complied with the European Communities Council Directive 2010/63/EU. All experimental procedures were performed in accordance with the current guidelines for the care of laboratory animals (including the use of the 3Rs procedures) and were in line with the ARRIVE guidelines. Rat model of implant-associated bone infection All experimental animals were purchased from the Laboratory Animal Center of Air Force Medical University. Male Sprague-Dawley rats (8 weeks old, 250 ± 20 g) were anesthetized by intraperitoneal injection of chloral hydrate. After skin preparation and disinfection, the knee joint was exposed, and a drill hole (2 mm diameter) was created in the distal femur between the medial and lateral condyles. A 30 μL suspension of MRSA (1 × 104 CFU/mL) was injected into the bone cavity. Implants (bare Ti, Ta, Ni, Mg, PEEK, or their GQCDs–coated counterparts) were inserted into the defect. Post-operatively, rats were housed under standard conditions. Post-implantation evaluation in rats Radiography: Femoral X-rays were taken at 0, 2, and 4 weeks post-surgery to monitor implant stability, infection control, and bone integration. Bone marrow smear: Bone marrow aspirates were collected at 1, 3, 10, and 30 days post-implantation, plated onto agar plates for bacterial culture, and the colonies were counted to determine the number of colony-forming units (CFU). Gait Analysis: At 8 weeks, footprint analysis was performed to assess functional recovery of the operated limb. Micro-CT, Histology, and Immunohistochemistry: At 8 weeks, animals were euthanized, and femur samples were harvested for micro-CT scanning, hard-tissue sectioning (stained with van Gieson or similar), and immunohistochemical analysis of osteogenic markers. Transmission Electron Microscopy (TEM): Bone tissue samples were processed for TEM to examine mitochondrial ultrastructure within osteoblasts. In vivo biodistribution assessment by fluorescence imaging GQCDs were labeled with Cy5.5 and used for implant modification. Briefly, Cy5.5 was dissolved in anhydrous DMSO to prepare a fresh solution, and GQCDs were dispersed in PBS. The pH of the GQCDs solution was adjusted to 8.3–8.5 to maximize the deprotonation of amino groups. Under dark conditions, the Cy5.5 solution was slowly added to the GQCDs solution and allowed to react at room temperature for 4 h. The reaction mixture was then dialyzed against ultrapure water using a dialysis membrane (molecular weight cut-off: 1000 Da) to remove unreacted dye and byproducts, yielding Cy5.5-labeled GQCDs. The Cy5.5-labeled GQCDs were subsequently applied to modify implants in a rat model of implant-associated infection. In vivo fluorescence imaging was performed at 1, 3, 10, 15, and 30 days post-implantation. At each time point, the implants, femurs, and major organs (heart, liver, spleen, lungs, and kidneys) were collected for ex vivo fluorescence imaging. Beagle canine large-animal model 10-month-old beagles were anesthetized and prepared as described above. A 4 mm drill hole was created in the distal femur, and 100 μL of MRSA suspension (1 × 105 CFU/mL) was injected. Bare Ti or Ti-GQCDs implants were placed into the defect. At 8 weeks, gait analysis was performed. At 12 weeks post-implantation, all animals were euthanized, and specimens were systematically harvested. Venous blood samples were collected for hematological profiling; gastrocnemius muscles were excised for Masson's trichrome staining; and femurs were retrieved for subsequent analyses, including micro-CT scanning, hard-tissue sectioning with histological staining, and transmission electron microscopy to evaluate cell ultrastructure. Bacterial Culture of Peri-Implant Tissues: Bone tissues within a 1-mm radius around the implant were collected and weighed (1 g), then thoroughly homogenized in 1 mL of sterile saline. The homogenate was serially diluted (1000-fold), and 100 μL of the diluted sample was spread onto appropriate agar plates. The bacterial load was calculated as colony-forming units per gram of tissue (CFU/g). Statistical analysis Data are presented as mean ± standard deviation (SD). Statistical analyses were performed using GraphPad Prism 8. For comparisons between two groups, a two-tailed unpaired Student's t-test was used. For comparisons involving multiple groups, one-way analysis of variance (ANOVA) was performed, followed by Tukey's multiple comparisons test. A P-value < 0.05 was considered statistically significant. Significance levels are indicated as *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. Ethics approval and consent to participate All animal studies were approved by the Animal Care and Use Committee of Fourth Military Medical University (251134), and all experimental protocols were approved by the Medical and Ethics Committee of Xijing Hospital, Fourth Military Medical University (20250032), Xi'an, China. CRediT authorship contribution statement Xinxin Bai: Writing – original draft, Validation, Methodology, Formal analysis, Data curation, Conceptualization. Anhui Qin: Formal analysis, Data curation, Conceptualization. Zhenguo Wang: Writing – original draft, Data curation, Conceptualization. Guang Kong: Data curation, Conceptualization. Menghan Zhang: Data curation, Conceptualization. Hongyu Li: Methodology, Data curation. Rui Zhao: Software, Data curation. Yue Yu: Data curation, Conceptualization. Yongfeng Zhang: Validation, Data curation. Qi Liu: Data curation, Conceptualization. Ziqing Zhu: Validation, Data curation. Bing Xia: Project administration, Investigation. Shengyou Li: Validation, Software. Yiming Hao: Supervision, Data curation. Mingze Qin: Data curation, Conceptualization. Dong Lei: Validation, Data curation. Lingli Guo: Data curation, Conceptualization. Meijia Su: Software, Data curation. Zhuojing Luo: Funding acquisition, Data curation. Shaohuang Weng: Writing – original draft, Visualization, Supervision, Conceptualization. Teng Ma: Writing – original draft, Visualization, Validation, Project administration. Jinghui Huang: Writing – review & editing, Supervision, Project administration, Funding acquisition. Declaration of competing interest The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Jinghui huang has patent #ZL2025 1446638.2 issued to Jinghui Huang. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.