Micellar LYTAC-engineered bionic coatings abrogate RANK-dependent osteocatabolism and boost fibrillogenesis-mimetic mineral accretion for empowering osteoporotic osseointegration

Abstract

Patients with osteoporosis are highly susceptible to fractures that often require implant fixation and stabilization. However, excessive osteoclast activity under osteoporotic conditions accelerates bone resorption at the bone-implant interface, severely compromising implant osseointegration. RANKL-RANK axis is a central regulator of osteoclast function and emerges as a promising target for osteoporosis treatment, yet current therapeutic options primarily rely on reversible ligand neutralization and can be followed by rebound bone loss after treatment discontinuation. Here, we developed a biomimetic titanium implant coating incorporating RANK-targeted micellar lysosome-targeting chimeras (LYTACs) to enhance osseointegration under osteoporotic conditions. Collagen-like peptides (CLPs) were conjugated to bispecific aptamers targeting RANK and the cation-independent mannose-6-phosphate receptor (CI-M6PR) through acid-responsive linkers. The resulting conjugates self-assembled into supramolecular LYTACs (CLP-Apt) and were integrated onto hydroxyapatite-modified titanium surfaces. Hydroxyapatite facilitated mesenchymal stem cell (MSC) recruitment, whereas acidic microenvironmental cues triggered CLP-Apt release. Simultaneous engagement of RANK and CI-M6PR on osteoclasts directed RANK toward lysosomal degradation, thereby suppressing osteoclast-mediated bone resorption while promoting MSC osteogenic differentiation and type H vessel formation. Following lysosomal processing, the released CLPs assembled extracellularly into bone matrix-mimetic fibers that chelated Ca²⁺ and accelerated new bone formation through a fibrillogenesis-mimetic process. By directly degrading RANK while promoting vascularized bone regeneration, this coating provides a localized strategy to overcome osteoclast-driven implant failure and achieve rapid, robust osseointegration under osteoporotic conditions.

Graphical abstract

Highlights

• •Hyperactivated RANK signaling promotes osteocatabolism of osteoclasts.
• •A micellar LYTAC system is developed for deleting osteoclast-intrinsic RANK.
• •The micelles could rearrange into ECM-like fibers to promote new bone formation.

1. Introduction

Orthopedic implants are commonly used medical devices for replacing or restoring damaged bone tissues, which demonstrate significant capacity for stabilizing fractures, reconstructing bone defects and restoring skeletal function [[1], [2], [3], [4]]. From a clinical perspective, the long-term therapeutic success of orthopedic implants fundamentally depends on their effective osseointegration, while inadequate osseointegration often leads to suboptimal mechanical anchorage, eventually resulting in implant failure [5,6]. Notably, the failure rate of implants is particularly problematic under osteoporotic conditions, where the overwhelming bone catabolism severely compromises new bone formation and remodeling capability. Patients with osteoporosis are highly susceptible to fragility fractures that often require implant-based treatment. However, the osteoporotic microenvironment, characterized by low bone mass, deteriorated trabecular architecture, and excessive bone resorption, impairs peri-implant bone formation and bone-implant contact, thereby compromising osseointegration and implant stability. Therefore, enhancing osseointegration in osteoporotic bone represents a critical unmet clinical need for improving implant stability and long-term therapeutic outcomes. Clinical evidence increasingly reveals that the RANK/RANKL signaling axis plays a critical role in osteoporosis pathology, which are key regulators of osteocatabolism [[7], [8], [9], [10], [11]]. Osteoporosis is universally accompanied with overactivation of osteoclast-intrinsic RANK signaling, which potently drives excessive osteoclastogenesis and bone matrix degradation [12,13]. Therefore, RANK signaling emerges as a potential therapeutic target for osteoporosis treatment and implant stabilization [14,15]. However, RANKL inhibition using conventional small molecule inhibitors such as Denosumab may not provide sustained control of osteoclast activity during implant osseointegration under osteoporotic conditions, as they predominantly suppress osteoclastogenesis and bone resorption by neutralizing RANKL but leave its receptor RANK intact [16,17], based on which the patients are highly susceptible to rebound bone turnover, rapid bone loss, and increased vertebral fracture risk after treatment discontinuation [7,18,19]. Thus, directly eliminating RANK rather than reversibly blocking RANKL may offer a more durable strategy to improve osteoporotic osseointegration via suppressing local osteoclast function [18,20].

Interestingly, the recent advances in targeted protein degrader technology offer new avenues to address the challenges associated with precise and effective RANK regulation for osteoporosis treatment [21,22]. In contrast to conventional pharmacological inhibition modalities, targeted protein degradation directly reduces the abundance of disease-driving proteins rather than blocking their activity or extracellular ligands. Thus, RANK depletion may directly diminish osteoclast responsiveness to RANKL, offering a mechanistically distinct and potentially more durable alternative to reversible RANKL neutralization [23,24]. Of note, lysosome-targeting chimeras (LYTACs) are a novel class of protein degraders capable of selectively deleting extracellular and membrane proteins by hijacking the endogenous lysosomal degradation pathway [25]. From a chemical perspective, LYTACs are essentially bispecific compounds that could simultaneously engage the target protein and a membrane-bound lysosomal trafficking receptor such as CI-M6PR or ASGPR, thereby enabling the sequential internalization and lysosomal degradation of the tethered target protein [[26], [27], [28]]. Interestingly, LYTAC technology offers distinct advantages for RANK inhibition compared with existing RANK/RANKL inhibitor designs, as it could directly eliminate RANK from the osteoclast surface rather than transiently blocking its activation via active site occupation, allowing for more durable suppression of RANKL-dependent osteoclastogenesis at the disease site. Importantly, osteoporotic osseointegration requires a coordinated balance between bone resorption, formation, mineralization, and vascular adaptation at the bone-implant interface, emphasizing the need for localized suppression of excessive osteoclast activity without disrupting physiological bone remodeling [29]. Implant surface-anchored RANK-targeting LYTACs may therefore offer a distinct advantage over systemic RANKL inhibition by confining receptor degradation to the bone-implant interface. However, reports in this area remain limited due to challenges associated with developing RANK-specific LYTACs and integrating them into functional Ti implant surfaces.

For this purpose, we first synthesized RANK–CI-M6PR-bispecific aptamers and conjugated them onto collagen-like peptides (CLPs) via an acidity-sensitive boronate ester linker, which could further undergo spontaneous self-assembly in aqueous solution to yield multifunctional micellar RANK-targeting LYTACs (MRTLs). Meanwhile, prior to HA mineralization, the Ti surface was functionalized with a phase-transitioned lysozyme interlayer generated using tris(2-carboxyethyl)phosphine (TCEP) as a reducing agent. TCEP-mediated reduction of the intramolecular disulfide bonds of lysozyme promoted protein unfolding and phase transition on the Ti surface, thereby generating a functional interlayer enriched with amino, carboxyl, and hydroxyl groups that facilitated Ca²⁺ coordination and subsequent mineral nucleation. A hydroxyapatite (HA) layer was then deposited on the functionalized Ti surface via simulated body fluid (SBF)-mediated in situ mineralization. Owing to the high porosity and strong DNA affinity of the HA layer, the MRTLs could be efficiently loaded into the HA coating to yield the final bionic coating (Ti/HA/CLP-Apt) (Scheme 1). After implantation into osteoporotic bone fracture site, the HA coating could facilitate the recruitment of pro-osteogenic MSCs to the implant interface, while the MRTLs are gradually released from the bionic coating and specifically bind to both RANK and CI-M6PR receptors on the osteoclast surface in a synchronized manner via aptamer-mediated targeting effect, thereby priming osteoclast-intrinsic RANK for lysosomal proteolysis. Of note, the micellar LYTAC-mediated osteoclast-specific RANK deletion markedly inhibit osteoclast maturation, which could not only suppress their osteolytic function but also promote osteoblastic differentiation of recruited MSCs as well as driving the peri-implant type H vessel formation, thus effectively ameliorating bone resorption. The acidic lysosomal environment would also cleave the boronate ester linkage within CLP-Apt to generate freestanding CLPs, which could further self-assemble into multi-helical bone ECM-like fibers and be excreted into the extracellular compartment. Owing to the phosphate group moieties in CLP, the CLP-based fibers could chelate microenvironmental Ca²⁺ ions and act as templates to direct new bone formation at bone-implant interface through a fibrillogenesis-mimetic process. These effects could act in a cooperative manner to remodel the osteoporotic bone microenvironment into a pro-healing state, providing new avenues for improving the therapeutic performance of orthopedic implants in the clinic.

Scheme 1.

Construction of the bionic coating modified Ti/HA/CLP-Apt implants and their therapeutic mechanism for promoting osteoporotic osseointegration. The spontaneous activation of the bionic coating releases in osteoporotic bone microenvironment releases abundant CLP-Apt micellar LYTACs that are subsequently selectively taken up by osteoclasts. The micellar LYTACs not only delete membrane RANK to abolish the osteolytic function of osteoclasts but also transform to bone ECM-mimetic fibers and are excreted to the extracellular compartment, cooperatively improving new bone formation at the bone-implant interface.

2. Experimental section

2.1. Materials

Titanium alloy substrates were provided by the Northwest Institute for Nonferrous Metal Research (Shaanxi, China). Peptide synthesis precursors were obtained from Anhui Guoping Pharmaceutical Co., Ltd., and the aptamer was synthesized by Biotechnology (Shanghai) Co., Ltd. (Chongqing, China). All other chemicals were purchased from Chongqing Chuandong Chemical Co., Ltd. (China).

2.2. Synthesis and characterization of CLP-Apt micelles

The peptide sequence COOH-YYYPKGPKGSPTSPSp-NH₂ was synthesized via solid-phase peptide synthesis (SPPS) using the Fmoc protection strategy. MBHA resin was used as the solid support. The resin was first swollen in dichloromethane (DCM) and subsequently washed with N,N-dimethylformamide (DMF) to remove residual solvents. Fmoc deprotection was carried out using 20% piperidine in DMF to expose the amine group. Amino acid residues were sequentially coupled from the C-terminus to the N-terminus. Fmoc-protected amino acid was activated in DMF with suitable coupling reagents, and the activated solution was introduced into the reaction column. The coupling reaction continued at room temperature for 1 h, allowing amide bond formation between the added amino acid and the resin-bound peptide chain. Afterwards, the peptide-resin complex was deprotected using a mixture of 95% trifluoroacetic acid (TFA), 2.5% triisopropylsilane (TIS) and 2.5% H₂O at room temperature for 2 h. The crude peptide was then precipitated in ice-cold diethyl ether, purified by high-performance liquid chromatography (HPLC), and lyophilized to yield a powdered product, designated as CLP. ESI-MS calculated for C76H111N17O26P [M + H]+ 1708.75, found 1708.77; ESI-MS calculated for C76H111N17O26P [M + 2H]2 + 854.88, found 854.89.

Meanwhile, 4-Nitrophenyl 4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)benzyl carbonate (NBC) was synthesized following previous reports. Briefly, p-nitrophenyl chloroformate (0.47 g) was dissolved in 20 mL of anhydrous tetrahydrofuran (THF) under ice-cooling conditions, followed by the addition of 4-(hydroxymethyl)phenylboronic acid pinacol ester (0.5 g), 4-dimethylamino pyridine (DMAP, 0.04 g), and triethylamine (0.6 mL). The ice bath was then removed, and the reaction mixture was stirred at room temperature for 3 h. Upon completion, THF was removed under reduced pressure using a rotary evaporator. The resulting residue was dissolved in dichloromethane (DCM), and the organic layer was sequentially washed with 1 M HCl and saturated NaCl once each, followed by washing with saturated NaHCO₃ until the aqueous phase became clear. The crude product was then purified by silica gel column chromatography to afford the desired compound. 1H NMR (400 MHz, CDCl3) δ 8.27 (d, J = 9.2 Hz, 2H), 7.85 (d, J = 8.0 Hz, 2H), 7.44 (d, J = 8.4 Hz, 2H), 7.38 (d, J = 9.2 Hz, 2H), 5.31 (s, 2H), 1.35 (s, 12H).

Apt was prepared through the complementary pairing of the sticky ends of Cl - M6PR and RANK aptamers. Specifically, the Cl - M6PR aptamer was hybridized with the RANK aptamer by heating at 95°C for 5 min and slowly cooling to room temperature over 3 h (approximately at a cooling rate of 0.5°C min⁻¹). The sequence of Cl-M6PR aptamer: 5′- TT TTTGGGCGCGTAGATGACGAGCAGTCCTAACATCGTTTAGGAC-3’. The sequence of RANK aptamer: 5′-GATTCGTATGGGTGGGATCGGGAAGGGCTACGAACACCGAAA AAU-3’.

CLP (17 mg) was dissolved in 2 mL of NaHCO₃ solution (pH 8.5), and NBC (5 mg) was dissolved in 0.6 mL of dimethyl sulfoxide (DMSO). The two solutions were then mixed and allowed to react at room temperature for 10 h. The resulting product was subsequently dialyzed and freeze-dried to obtain CLP–NBC. Next, CLP–NBC (1.5 mg) was dissolved in 1 mL of aptamer (Apt) solution (1 mM) and reacted for 12 h at room temperature to yield the CLP–Apt complex.

CLP (20 mg) was dissolved in 1 mL of DMSO, and 3 mL of deionized water was added dropwise. The mixture was stirred for 3 h to form CLP micelles, which were purified by ultrafiltration centrifugation (2000 rpm, 5 min × 3) to obtain the purified CLP micelles.

CLP-Apt (50 mg) was dissolved in 1 mL of DMSO, and 3 mL of deionized water was added dropwise. The mixture was stirred for 3 h to form CLP-Apt micelles, which were purified by ultrafiltration centrifugation (2000 rpm, 5 min × 3) to obtain the purified CLP-Apt micelles.

2.3. Characterization of CLP-Apt micelles

The CLP–Apt compounds were analyzed using fluorescent amine assay and Fourier transform infrared spectroscopy (FTIR). The resulting CLP-Apt nanomicelles were further characterized by transmission electron microscopy (TEM), energy-dispersive X-ray spectroscopy (EDS), multi-angle dynamic light scattering (DLS), and zeta potential measurements to assess their morphology, elemental composition, particle size distribution, and surface charge characteristics.

2.4. Construction of HA coatings on Ti surfaces

Titanium disks were mechanically polished and ultrasonically cleaned prior to use. The disks were then immersed in a lysozyme buffer and incubated at room temperature. Subsequently, they were sequentially immersed first in a 20 mM calcium chloride (CaCl₂) aqueous solution for 2 days, followed by incubation in simulated body fluid (SBF) for 10 days. After thorough rinsing with ultrapure water and vacuum drying, a layer of white hydroxyapatite (HA) was observed on the titanium surface, designated as Ti/HA.

2.5. Surface loading of CLP-Apt micelles on Ti/HA implants

Ti/HA was immersed in a solution of CLP-Apt micelles (400 nM) for 12 h. The sample was then washed three times with deionized water for 5 min each to remove unbound micelles. The final product was designated as Ti/HA/CLP-Apt.

2.6. Physicochemical characterization of Ti/HA/CLP-Apt

The morphologies and roughness of different samples were characterized using scanning electron microscopy (SEM), environmental scanning electron microscope (ESEM) and atomic force microscopy (AFM). The composition and crystal structure of the samples were analyzed by X-ray spectroscopy (EDS), X-ray photoelectron spectroscopy (XPS), and X-ray diffraction (XRD). Surface energy and wettability were evaluated through contact angle measurements.

2.7. Biocompatibility assessment of the Ti/HA/CLP-Apt implants

RAW 264.7 cells, HUVECs, or MSCs were seeded onto the surfaces of different samples at a density of 2 × 10⁴ cells/cm². After 3 days of culture, the cytotoxicity was assessed using a CCK-8 working solution (UElandy, China) and a lactate dehydrogenase (LDH) assay kit (Beyotime, China).

For CLSM analysis, RAW 264.7 cells, HUVECs, or MSCs were seeded onto the surfaces of different samples at a density of 2 × 10⁴ cells/cm². After 3 days of culture, the cells were stained using Calcein-AM and propidium iodide (PI) to distinguish live and dead cells. The staining was performed by incubating the cells with the dyes at 37°C for 15 min. Cell survival on the material surfaces was finally visualized via CLSM.

2.8. Molecular dynamics simulation of CLP self-assembly

The self-assembly process of CLP molecules within the acidic microenvironment of lysosomes were smilated utilizing the molecular dynamics simulation software GROMACS 2023.5. Unconstrained molecular dynamics simulations were executed on the protein - protein molecular docking complex by means of the Amber14sb_parmbsc1 force field and the TIP3P water model. All simulations utilized a cubic solvent box with periodic boundary conditions set at 1 nm. The initial stabilization of the system entailed 500 ps of NVT equilibration, followed by 1000 ps of NPT equilibration. Under isothermal and isobaric conditions, the Nose - Hoover thermostatic coupling algorithm and the Parrinello - Rahman pressure coupling method were applied to maintain the system at 308.15 K and 1 bar. Subsequently, a 100 ns molecular dynamics simulation was conducted on the complex. Long - range electrostatic interactions were computed using PME with a Coulomb radius of 1.2 nm.

2.9. In vitro osteoclastogenesis and fusion assay

RAW 264.7 cells were seeded onto the surfaces of different samples at a density of 2 × 10⁴ cells/cm² and cultured for 3 days in medium supplemented with RANKL and M-CSF to induce osteoclast differentiation. After fixation, tartrate-resistant acid phosphatase (TRAP) activity was evaluated using a commercial TRAP staining kit. The cytoskeleton and nuclei were stained with rhodamine-conjugated phalloidin and DAPI, respectively, and the number of multinucleated cells on each material surface was observed using CLSM. Actin rings of osteoclasts were visualized by CLSM after staining with FITC-conjugated phalloidin.

For the osteoclast fusion assay, RANKL/M-CSF-induced RAW 264.7 cells were incubated on different samples for 3 days and then labeled with Hoechst, CM-DiO or CM-DiI at room temperature for 20 min. Two groups of differently labeled cells were subsequently co-cultured for 12 h, and cell–cell fusion events were visualized by CLSM.

2.10. Isolation and culture of MSCs

MSCs were isolated from the bone marrow cavity of Sprague Dawley rats (female, body weight 100 g) under sterile conditions. The MSCs were cultured in low-glucose DMEM medium supplemented with 10% fetal bovine serum (FBS), and the medium was replaced every other day. When the cell confluence reached 90%, the cells were detached and re-seeded into new culture flasks. Cells at passage 3 were used for all experiments.

2.11. HUVEC/RAW264.7 and MSC/RAW264.7 in vitro co-culture systems

For the HUVEC/RAW264.7 co-culture system, RAW 264.7 cells were seeded onto different titanium-based material surfaces and induced with culture medium containing RANKL and M-CSF to promote osteoclast differentiation. The resulting osteoclast-conditioned medium (OC-CM) was collected and mixed with fresh basal medium at a 1:1 ratio to obtain the conditioned medium used for culturing HUVECs.

For the MSC/RAW264.7 co-culture system, the same procedure was applied. RAW 264.7 cells were inoculated on titanium substrates and induced with RANKL and M-CSF, and the resulting osteoclast-conditioned medium was similarly mixed 1:1 with fresh medium to prepare the conditioned medium for MSC culture.

2.12. Transwell migration assay

HUVECs were seeded into the upper chambers at a density of 1.5 × 10⁴ cells per well and cultured with conditioned media from different experimental groups. After 24 h of incubation, the chambers were carefully removed and gently rinsed with PBS. Non-migrated cells on the upper surface of the membrane were removed using a cotton swab, while the migrated cells in the bottom chamber were fixed with 4% paraformaldehyde for 30 min and stained with 0.1% crystal violet for 20 min. The stained cells were subsequently visualized and quantified under an optical microscope.

2.13. Cell scratch assay

HUVECs were seeded into culture dishes and maintained in conditioned media from different experimental groups until a confluent monolayer was formed. A straight scratching wound was then created in the monolayer using a sterile pipette tip held perpendicular to the dish surface. Detached cells were removed by gently rinsing with PBS, and the initial wound area (0 h) was imaged under a microscope. The cultures were then returned to the incubator and maintained for an additional 24 h, after which the same regions were imaged again to assess cell migration and wound closure.

2.14. Angiogenesis assay

Matrigel was thawed overnight at 4°C prior to the experiment. Using pre-cooled pipette tips, the Matrigel was added to a pre-cooled multi-well plate and incubated at 37°C for 30–60 min to allow adequate gelation. HUVECs were then seeded onto the surface of the solidified gel at a density of 2 × 10⁴ cells per well. The culture plate was maintained in a 37°C, 5% CO₂ incubator for further cultivation. After 6 h, the formed tubular structures were imaged under a microscope and quantitatively analyzed using ImageJ software.

2.15. Morphology of MSCs

The morphology of MSCs cultured on different implant surfaces was examined using SEM. MSCs were seeded onto the surfaces and cultured in conditioned medium for 3 days. The samples were fixed with 2.5% glutaraldehyde for 12 h, rinsed twice with PBS, and sequentially dehydrated in ethanol solutions of increasing concentrations (30%, 50%, 70%, 80%, 90%, 5 min each), followed by two 5-min incubations in 100% ethanol. Finally, the samples were dried at 37°C prior to imaging. Cell spreading area and focal adhesion formation were further evaluated using CLSM. The cytoskeleton and nuclei were stained with rhodamine-conjugated phalloidin and DAPI, respectively, and immunofluorescence staining for vinculin was performed to visualize focal adhesions. Samples were then imaged using a confocal fluorescence microscope to assess cell morphology and adhesion.

2.16. Immunofluorescence staining and microscopy

Cells were washed once with culture medium and twice with PBS, followed by fixation with 4% paraformaldehyde for 15 min at room temperature. Samples were then rinsed three times with ice-cold PBS. Cells were permeabilized with 0.25% (v/v) Triton X-100 in PBS for 10 min at room temperature and washed three times with PBS. Non-specific antibody binding was blocked by incubation with 3% (w/v) bovine serum albumin (BSA) in PBST (0.1% v/v Triton X-100 in PBS) for 10 min at room temperature. Subsequently, samples were incubated with primary antibodies overnight at 4°C. After primary antibody incubation, samples were washed twice with PBST and three times with PBS. Samples were then incubated with secondary antibodies and DAPI for 60 min at room temperature, followed by two washes with PBST and three washes with PBS. Immunofluorescence images were acquired using CLSM.

2.17. ALP activity and mineralization assays of MSCs in vitro

MSCs were seeded onto different titanium-based material surfaces and cultured for 7 days. The expression and activity of ALP in MSCs were analyzed quantitatively and qualitatively using an ALP detection kit (Nanjing Jiancheng Bioengineering Institute, China), BCA protein assay kit (Beyotime, China), and BCIP/NBT staining kit (Beyotime, China). After 14 days of culture, cells were fixed with 4% paraformaldehyde and stained with Alizarin Red S. The samples were then observed and imaged under an optical microscope. For quantitative analysis, the stained cells were incubated with 10% acetic acid at 80°C for 10 min, followed by 10% ammonium hydroxide. The absorbance at 405 nm was measured using a microplate reader to assess mineralization levels.

2.18. Quantitative real - time PCR assay

Total RNA was extracted using the TRIzol method. First-strand cDNA was synthesized by reverse transcription using an RT mix with DNase kit (US EVERBRIGHT INC.). The resulting cDNA served as the template for quantitative real-time PCR (qRT-PCR), which was performed using the 2× SYBR Green qPCR Master Mix kit (US EVERBRIGHT INC.) on a CFX Connect Real-Time PCR system (Bio-Rad, USA). GAPDH was employed as the internal reference gene for normalization. All primers used in this study were synthesized by Sangon Biotech (Shanghai) Co., Ltd. (Table S1).

2.19. Western blotting analysis

Protein expression levels were assessed by Western blotting. Briefly, cells were lysed using RIPA buffer and centrifuged at 12,000 rpm for 10 min at 4°C to collect the supernatant. The supernatant was then mixed with 5× SDS-PAGE loading buffer at a 4:1 ratio (protein sample:loading buffer) and boiled for 5–10 min to ensure complete denaturation. The proteins were separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto a polyvinylidene fluoride (PVDF) membrane. The membrane was blocked with 5% skim milk for 2 h and then incubated with primary antibodies at 4°C overnight. After washing, the membrane was incubated with secondary antibodies for 1 h at room temperature. Protein signals were finally detected and analyzed using a multi-functional imaging system (Box Chemi XX9, Syngene G, UK).

2.20. Establishment of osteoporotic SD rat models and treatments

8-week-old female SD rats were provided by Chongqing Ensiwei Experimental Animal Sales Co., Ltd., and all animal experiments were reviewed and approved by the Animal Ethics Committee of Army Medical University (Approval Number: AMUWEC20255450). The rats were anesthetized by intraperitoneal injection of 2% sodium pentobarbital. Bilateral ovariectomy was then performed, and the incisions were sutured. After the operation, the wounds were disinfected daily for 3 days, and the rats received an intramuscular injection of 100 μL of penicillin (5000 U/mL) each day to prevent infection.

Approximately 3 months later, the implantation surgery was conducted. The rats were anesthetized again by intraperitoneal injection of 2% sodium pentobarbital. A cylindrical defect with a diameter of 1.2 mm and a length of 10 mm was created along the femor in the epiphysis using an electric drill. A sterile titanium implant (diameter: 1.2 mm; length: 10 mm) was immediately inserted into the defect. Finally, the incision was sutured. Afterwards, the wounds were disinfected daily for 3 days, and the rats received an intramuscular injection of 100 μL of penicillin (5000 U/mL) each day to prevent infection.

2.21. Micro-CT scanning and analysis

At 8 weeks post-implantation, the rats were euthanized, and femoral bone samples containing different titanium-based implants were collected. The samples were fixed in 4% paraformaldehyde for 48 h. Subsequently, they were scanned using a micro-CT scanner (model Scanco vivaCT40) to quantitatively analyze new bone formation within a 1 mm region surrounding the implant in the rat epiphysis. The parameters analyzed included bone volume/total volume (BV/TV; %), trabecular number (Tb.N; 1/mm), trabecular thickness (Tb.Th; mm), and trabecular separation (Tb.Sp; mm).

2.22. Histological and immunofluorescence analysis

The fixed femoral samples were decalcified with 10% EDTA for 6 weeks. Subsequently, a portion of the bone was resected to allow for the retrieval of the titanium implant. The processed tissue was then embedded in paraffin to prepare sections for H&E staining, Masson staining, TRAP staining, immunohistochemical (IHC) staining, and immunofluorescence (IF) staining. Finally, samples were imaged using an optical microscope and CLSM.

2.23. Angiography of peri-implant bone tissues

Rats were euthanatized and the thoracic cavity was opened. A needle was inserted into the left ventricle, and the right atrial appendage was incised for sequential perfusion using heparinized saline (100 U/mL), 4% paraformaldehyde and heparinized saline (100 U/mL) again, and finally Microfil MV-120 contrast agent (Flow Tech). The procedure was terminated when the lower limbs exhibited distinct blue color and the contrast agent flowed out from the right atrial appendage. The perfused rats were refrigerated overnight at 4°C. The lower limbs were then dissected and fixed in 4% paraformaldehyde. Femoral vascular parameters were acquired using Micro-CT imaging.

2.24. Biocompatibility assessment

At the experimental endpoint, rats were euthanized, and their hearts, livers, spleens, lungs, and kidneys were collected. The tissues were fixed in 4% paraformaldehyde, dehydrated through treatment with graded alcohol series, embedded in paraffin, and finally sectioned. The sections were stained with hematoxylin and eosin (H&E) and mounted with neutral balsam. The prepared slides were observed and imaged under an optical microscope to evaluate histological responses.

2.25. Statistical analysis

All data were statistically analyzed using Origin software (Origin 2022). All data are expressed as mean ± standard deviation. Statistical analysis was performed using one-way analysis of variance (ANOVA). The thresholds of statistical significance are *P < 0.05, **P < 0.01, ***P < 0.001.

3. Results

3.1. Preparation and characterization of CLP-Apt micellar LYTACs

The construction process of the multifunctional CLP-Apt micellar LYTACs was illustrated in Fig. 1A. Specifically, the RANK-CI-M6PR-bispecific aptamers were synthesized by conjugating commercially available RANK-targeting aptamers to CI-M6PR-targeting aptamers. Of note, the RANK aptamer features an A-containing sticky end at its 3′ terminus, which could readily complex with the T-containing sticky end at the 5′ terminus of CI-M6PR aptamer. Moreover, the 5’ end of CI-M6PR aptamer is modified with three C-containing RNA bases for enabling boronate ester ligation with CLP-linker. The dual-targeting capability of CLP-Apt was also supported by reciprocal pull-down assay results (Fig. S1A). Anti-CI-M6PR pull-down efficiently co-enriched RANK, whereas anti-RANK pull-down reciprocally enriched CI-M6PR, with negligible signals detected in the bead-only and IgG controls. These results provide direct biochemical evidence that CLP-Apt facilitates the association of CI-M6PR and RANK, supporting its simultaneous engagement of both targets. Further, the capacity of the bispecific aptamer to induce lysosomal RANK degradation was tested on RANKL/M-CSF co-stimulated RAW264.7 cells, which were treated with CLP-Apt micellar at concentrations ranging from 50 to 1000 nM. The results revealed that the CLP-Apt micellar treatment induced an obvious dose-dependent decrease of RANK levels (Fig. 1B), for which the RANK inhibition efficacy reached a maximum at the CLP-Apt micellar dose of 400 nM (Fig. 1D). Time-dependent WB and flow cytometry analysis further showed that approximately 50% of RANK proteins were degraded after 12 h of CLP-Apt micellar treatment, and the maximal degradation effect was observed at 24 h (Fig. 1C–E). These data collectively confirmed that the bispecific Apt readily induced efficient RANK degradation in osteoclast-like cell models. Meanwhile, CLPs were synthesized by solid-phase peptide synthesis (SPPS) and their composition was validated via electrospray ionization mass spectrometry (ESI-MS), exhibiting characteristic peaks at 854.89 (m/z = 2) and 1708.77 (m/z = 1) (Fig. 1F). We then investigated the self-assembly behavior of the CLP components via molecular dynamics simulation using Amber14sb_parmbsc1 force field and TIP3P model, and the results evidently supported that the tyrosine moieties in CLPs could undergo π-π stacking in acidic environment to potentiate self-assembly into fibrous matrix (Fig. 1G). NBC was synthesized by referring to the existing literature [30], and its structure was confirmed by Proton NMR (1H NMR). As shown in Fig. S1B, the characteristic peaks at δ = 8.2,7.85,7.45,7.3,5.42 and 1.26 ppm correspond to benzene ring protons, methylene (-CH₂-) and methyl (-CH₃) hydrogen atoms in NBC, respectively, confirmation of NBC synthesis. Subsequently, the CLP-NBC complex was synthesized by the covalent reaction between the carbonate bond at the end of NBC and the primary amino group at the end of CLP. The results of fluoroamine analysis (Fig. S1C) showed that the fluorescence intensity of CLP-nbc was significantly reduced compared with that of free CLP, which was due to the fact that fluoroamines were only bound to the unreacted primary amino group, and the fluorescence intensity of CLP-nbc was significantly decreased compared with that of free CLP, the fluorescence intensity of CLP-nbc decreased significantly, indicating that the primary amino group of CLP was involved in the grafting reaction, which further verified the successful synthesis of CLP-nbc. The covalent connection of CLP and Apt is realized by the reaction of the boronic acid group at the end of NBC molecule with the ortho-hydroxyl group at the 3′ end of the aptamer to form a reversible borate ester bond. The pro-osteogenesis capacity of the generated CLP-based fibers was further studied under osteoporosis-like environment, where they showed marked biomineralization with freestanding calcium ions to form bone matrix-like hybrid structures, which was attributed to the abundant phosphate groups in the CLP segments. The CLPs were sequentially modified with the NBC linker and connected to the Apt via a reversible boronate ester ligation, leading to the formation of the self-assemblable CLP-Apt conjugates. The chemical structures of the CLP-Apt components were validated via Fourier-transform infrared spectroscopy (Fig. 1H) [31,32]. Dynamic light scattering (DLS) measurements revealed a hydrodynamic diameter of around 180 ± 8.7 nm for CLP-based micelles, which increased to 210 ± 10.9 nm after Apt modification (Fig. 1I). Zeta potential analysis revealed that the surface charge shifted from +2.2 ± 0.03 mV for CLP micelles to −9.8 ± 0.16 mV for CLP-Apt (Fig. 1J). Notably, incubating the CLP-Apt assembly at pH 7.4 induced only negligible Apt release, of which the accumulative release ratio was less than 10% after 24 h of incubation. In contrast, the accumulative Apt release under lysosome-mimetic pH of 5.0 reached around 84%, which was in line with the intrinsic acidity susceptibility of the NBC linkers for triggering Apt removal under lysosome-like conditions (Fig. 1K).These data collectively confirmed successful aptamer conjugation. It is also important to note that the DLS and zeta potential results of the micelles showed no obvious changes over five consecutive days of incubation in aqueous environment, supporting their excellent colloidal stability thereof (Fig. S1D and E). Owing to the balanced hydrophobic and hydrophilic moieties in the CLP components, they could readily self-assemble into micelles in aqueous environment. Transmission electron microscopy (TEM) images revealed that both CLP and CLP-Apt micelles exhibited uniform spherical morphology and good dispersibility in aqueous environment, indicating that Apt modification did not obviously affect their assembly capability (Fig. S1F and Fig. 1L). Energy-dispersive X-ray spectroscopy (EDS) confirmed the homogeneous distribution of C, N, O, and P elements within CLP-Apt micelles (Fig. 1M). To investigate if the incorporation of the NBC linker could facilitate the bioresponsive activation of CLP-Apt components in osteoporotic microenvironment, CLP-Apt self-assemblies were further incubated under different pH conditions to observe their compositional changes. It is also of interest to note that the particle-like self-assemblies readily switched to fiber-like morphology after incubation at pH 5.0, again validating the successful detachment of the Apt moieties as well as the bone ECM-mimetic self-assembly properties of CLP (Fig. S1G and F). Under osteoporosis-mimetic conditions (10 mM Ca²⁺, pH 4.0), these short fibers further assembled into bone-like matrices with strong resemblance to natural mineralized collagen (Fig. 1N). EDS elemental mapping of the generated fibrous assemblies showed marked presence of Ca²⁺ in the matrices (Fig. 1O), confirming the crucial role of Ca²⁺ ions in regulating the osteogenesis-mimetic biomineralization process.

Fig. 1.

Fabrication, characterization, and RANK degradation activity of CLP -Apt micellar LYTACs. A) Schematic representation of the micelle formation of CLP - Apt components and their morphological transformation into bone ECM-mimetic fibers under different conditions. B) WB analysis of the changes in RANK protein expression in osteoclasts treated with varying concentrations of RANK-targeting Apt for 48 h. C) WB analysis of the changes in RANK expression in osteoclasts treated with 400 nM Apt at different time points. D) Flow cytometric analysis of the alterations in FITC - RANK fluorescence intensity on osteoclast surfaces following treatment with different concentrations of Apt. E) Flow cytometric analysis of the alterations in FITC - RANK fluorescence intensity on the cell surfaces after Apt treatment at different time points. F) Electrospray ionization - mass spectrometry (ESI - MS) spectrum of CLP. G) Molecular dynamics simulation of the CLP assembly triggered in the context of acidic microenvironment and Ca²⁺. H) Fourier transform infrared spectroscopy analysis of CLP and CLP - Apt. I) Size distribution of CLP and CLP - Apt nanomicelles via DLS analysis. J) Zeta potential of CLP and CLP - Apt. K) Release profile of Apt under different pH conditions. in the presence of Ca²⁺. L) TEM image of CLP-Apt nanomicelles. M) EDS images of CLP - Apt nanomicelles. N) TEM images of CLP-Apt self-assembly in the presence of Ca²⁺. O) EDS images of CLP - Apt self-assembled nanofibers.

3.2. Preparation of the bionic coating on Ti implants

To facilitate the functionalization of the Ti implants, an HA layer was first established on Ti implant surface via the well-established SBF biomimetic mineralization method. The surface morphology of samples was characterized using SEM. As shown in Fig. S2A, pure Ti surface exhibited a relatively smooth morphology, whereas the HA-modified samples (Ti/HA) displayed uniformly distributed needle-like nanostructures, indicating successful HA deposition. After anchoring the CLP-Apt micelles, numerous spherical nanostructures appeared on the implant surface, further confirming the immobilization of the micellar LYTACs thereof. ESEM was further employed to examine the cross-sectional microstructure, yielding results consistent with those described above. (Fig. 2A). Following the confirmation of C, O, N, P, and Ca enrichment on LYTAC-engineered Ti implants by EDS analysis (Fig. 2B), TEM imaging and elemental mapping (Fig. S2B) further validated the successful assembly of the HA/CLP-Apt hybrid nanostructure, with uniform distributions of O, P, and Ca elements indicating the preservation of HA/CLP composition after LYTAC modification. Atomic force microscopy (AFM) characterization (Fig. 2C) showed that the surface roughness of Ti/HA was significantly higher than that of pure Ti, primarily due to the nanoscale needle-like or lamellar structures generated via HA deposition. Following the modification of the micellar LYTACs, the surface roughness further increased to around 79.2 ± 4.2 nm, which was attributed to the attachment of the micelles that partially filled the microscopic gaps between HA crystals, resulting in a more complex surface topology. According to previous investigations, the moderate roughness of samples is beneficial for promoting osteoblast adhesion and proliferation [33,34]. The measured Ca/P molar ratio of the Ti/HA sample was 1.68 (Fig. S2C and D), closely matching the theoretical value of natural HA crystals (1.67), thereby confirming the successful formation of the HA coating. This is also supported by X-ray diffraction (XRD) analysis on Ti/HA samples, of which the diffraction peaks were highly consistent with the standard HA pattern in the ICDD database (Fig. 2D). X-ray photoelectron spectroscopy (XPS) analysis also detected characteristic Ca 2p and P 2p peaks in the Ti/HA group (Fig. 2E), further verifying the successful addition of the HA layer onto Ti surface after SBF treatment. Notably, Ti/HA/CLP showed pronounced N 1s peak due to the amide groups in CLP, whereas Ti/HA/CLP-Apt showed enhanced P 2p signal, which was assigned to the abundant phosphate groups in Apt molecules. Surface wettability assessment revealed that Ti/HA, Ti/HA/CLP and Ti/HA/CLP-Apt samples exhibited significantly lower water contact angles than pure Ti (Fig. 2F), indicating that surface modification effectively enhanced the hydrophilicity of Ti. The observations above collectively confirmed the successful construction of the bionic coating on Ti implant surface as well as its biological responsiveness. Moreover, the pull-off adhesion assay (Fig. S2E) revealed that Ti/HA/CLP and Ti/HA/CLP-Apt coatings maintained strong interfacial bonding strength above 15 MPa, which was comparable to Ti/HA. Meanwhile, the introduction of CLP and bispecific aptamer micelles did not obviously compromise the adhesion strength of the HA-based coating. These results collectively support the mechanical stability of the bionic coating under clinically relevant conditions.

Fig. 2.

Ti/HA/CLP-Apt fabrication, physicochemical properties, and in vitro LYTAC-mediated RANK degradation. A) ESEM images of different titanium-based materials. B) EDS images of Ti/HA/CLP-Apt. C) AFM images and surface roughness analysis of different titanium-based materials. D-F) XRD, XPS and contact angle analysis of different titanium-based materials. G) CLSM images of osteoclasts cultured for 24 h with Ti/HA/CLP-Apt-FITC-conditioned culture media in a time dependent manner. Blue: cell nuclei, red: cell membranes, and green: FITC-labeled nanomicelle. H) Flow cytometric analysis of CLP-Apt-FITC uptake. I) CLSM image of intracellular CLP assembly. J) WB analysis on the degradation of osteoclast-intrinsic RANK treated with 400 nM Apt for 72 h. The osteoclasts were co-incubated with 100 μM lysosomal inhibitor baffloramifene (Baf-A1) or 5 μM proteasome inhibitor MG132; K) Flow cytometric analysis of surface RANK expression after different treatments. L) Co-localization of CLP-Apt-FITC with lysosomes. M) RNAK protein expression levels in osteoclasts cultured on different Ti-based samples. N) Flow cytometry analysis of RANK expression in osteoclasts. O) CLSM images of RNAK distribution in osteoclasts.

3.3. In vitro evaluation of micellar LYTAC uptake and RANK degradation

Extending from the sustainable micellar LYTAC release from the bionic coating in osteoporosis-mimetic environment, the interaction between micellar LYTACs and osteoclasts was comprehensively investigated in vitro using osteoclast-like RANKL/M-CSF co-stimulated RAW264.7 cell model. Herein, the osteoclast-like cells were incubated for 24 h in the soaking solutions of various samples constructed using fluorescently labeled micellar LYTACs. CLSM results revealed a gradual increase in green fluorescence intensity with increasing sample pre-immersion time, reaching a plateau after approximately 21 days of pre-immersion (Fig. 2G). The CLSM analysis results were consistently supported by the quantitative flow cytometry analysis (Fig. 2H), supporting the gradual buildup of the LYTAC components in induced RAW264.7 cells. To facilitate the in-situ tracking of the self-assembly of CLPs in induced RAW264.7 cells, the CLPs were fluorescently labeled with FITC for CLP-Apt preparation, and their subcellular distribution after uptake by induced RAW264.7 cells was visualized by CLSM. Notably, the CLP-associated green fluorescence was largely overlapped with lysosomes after 12 h of incubation, indicating efficient uptake of the CLP-Apt micelles through the endocytosis route. Contrastingly, obvious fiber-like structures with marked green fluorescence appeared in the intracellular compartment after 36 h, indicating successful micelle-fiber transformation of the self-assembled CLP components (Fig. 2I). Moreover, the generated fibrous structures were readily excreted to the extracellular compartment after 48 h of incubation. Notably, the Ti substrates in the Ti/HA/CLP and Ti/HA/CLP-Apt groups exhibited abundant ECM-like fibrous structures, which were absent in the Ti and Ti/HA groups. Consistent with the fluorescence imaging results, bio-TEM analysis (Fig. S2F) provided direct ultrastructural evidence of the intracellular micelle-to-fiber transition, showing the progressive formation of CLP-derived nanofibers within cells followed by their extracellular secretion. These findings were further corroborated by SEM imaging of the titanium implant surfaces after co-incubation with cells (Fig. S2G), which revealed similar fibrous structures. These observations collectively supported the proposed pro-osteogenesis activity of the bionic coating via sequential LYTAC-dependent RANK degradation and ECM-like fiber excretion. The impact of the bionic coating-released micellar LYTACs on osteoclast-intrinsic RANK expression and the underscoring mechanisms were further investigated via inhibitor-mediated rescue experiment. Notably, culturing induced RAW264.7 cells with micellar LYTACs caused marked downregulation of RANK levels, while treating the LYTAC-conditioned osteoclast-like cells with lysosomal inhibitor bafilomycin A1 (Baf-A1) induced obvious recovery of overall RANK abundance. Interestingly, treating LYTAC conditioned osteoclast-like cells with proteasome inhibitor MG132 caused no significant restoration of RANK expression. The Western blot and flow cytometry analyses collectively demonstrated that the micellar LYTACs attenuated osteoclast-intrinsic RANK levels via mediating their lysosomal degradation rather than the proteolysis pathways (Fig. 2J and K). Indeed, CLSM imaging revealed that the fluorescently labeled RANK-CI-M6PR-bispecific Apt constructs were highly overlapped with osteoclast lysosomes (Fig. 2L), indicating their efficient lysosomal trafficking after cellular uptake. Furthermore, Western blot analysis revealed that only the Ti/HA/CLP-Apt group induced effective RANK degradation in osteoclast-like cell populations, whereas no significant changes in RANK levels were observed in other groups (Fig. 2M), highlighting the Apt constructs as the effector for deleting osteoclast-intrinsic RANK while excluding the contribution of other coating components in this process. The RANK degradation efficacy of Ti/HA/CLP-Apt was also consistently supported by both flow cytometry and immunofluorescence analyses (Fig. 2N and O). Collectively, these results demonstrate that the RANK-CI-M6PR-bispecific Apt constructs mediates efficient osteoclast-intrinsic RANK degradation via hijacking the lysosomal proteolysis pathway, underscoring the molecular basis of the bionic coating to regulate osteoclast function for ameliorating osteocatabolism at bone-implant interface.

3.4. In vitro evaluation of micellar LYTAC-mediated osteoclast reprogramming

The impact of bionic coating delivered micellar LYTACs on the osteolytic function of osteoclasts was further investigated. Notably, the bionic coating showed no obvious toxicity to induced RAW264.7 cells according to CCK-8 assays, lactate dehydrogenase (LDH) release tests and live/dead staining (Fig. 3A, S3A, B), immediately confirming its biocompatibility under clinically relevant conditions. In line with these morphological and phenotypical observations, Western blot analysis revealed that the Ti/HA/CLP-Apt treatment significantly downregulated both phosphorylated TAK1 (p-TAK1) and phosphorylated NF-κB p65 (p-p65) levels (Fig. 3B) immediately confirming that the CLP-Apt micellar treatment blocked RANKL/RANK binding to inhibit the downstream TAK1-NF-κB signaling, eventually suppressing osteoclast maturation and their osteolytic function. Mechanistically, lysosomal blockade with bafilomycin A1 restored RANK abundance and reversed the CLP-Apt-mediated inhibition of TAK1/NF-κB signaling, evidenced by the marked recovery of p-TAK1 and p-p65 levels. These findings confirm that RANK degradation is a key upstream event responsible for the suppression of downstream signaling events (Fig. S3C). Of note, the Ti/HA/CLP-Apt treatment markedly inhibited the maturation of co-cultured osteoclasts, evidenced by the significantly reduced mRNA levels of multiple osteoclastic genes including Acp5, Ctsk, Nfatc1 and Dcstamp (Fig. 3C–F). Meanwhile, TRAP staining of the osteoclast-like cells confirmed that the CLP-Apt micellar treatment induced a marked decrease in the number of TRAP-positive cells (Fig. 3G and H). These data confirmed that the Apt substantially attenuated the osteolytic capability of osteoclasts. Microscopic imaging of the osteoclasts after various treatment further revealed that the CLP-Apt micellar treatment induced obvious morphological changes including diminished actin ring circumference (Fig. 3I and J), lower frequency of multinucleated osteoclasts (Fig. 3K and L) and reduced membrane fusion rate of preosteoclasts (Fig. 3M and N), suggesting that the aptamers blocked the formation of mature multinucleated osteoclasts despite RANKL/M-CSF co-stimulation.

Fig. 3.

Inhibition of osteoclast maturation and function by micellar LYTACs. A) Calcein/PI staining of RAW264.7 cells on different titanium-based material surfaces. B) WB analysis of TAK1, p65, IκBα, and phosphorylated TAK1 and p65 in osteoclasts cultured on different titanium-based materials. C-F) The relative mRNA levels of Acp5, Ctsk, Nfatc1, and Dcstamp, respectively. G, H) TRAP staining and quantitative analysis to assess the effects of different titanium-based materials on bone resorption. I, J) CLSM images and quantitative analysis of cellular F-actin rings on different titanium-based material surfaces. K, L) CLSM images and quantitative analysis of multinucleated cells on different titanium-based material surfaces. M, N) CLSM observation on osteoclast fusion and statistical analysis of cell membrane fusion rates on different titanium-based material surfaces. Data were presented as mean ± SD for n = 3, **p < 0.01, ***p < 0.001.

Previous clinical studies collectively demonstrate that osteoclast precursors have significant pro-angiogenic capability via secreting platelet-derived growth factor-BB (PDGF-BB) and can enhance type H vessel formation to promote new bone formation, but will switch to a pro-osteolysis and anti-angiogenesis role after maturation into multinucleated osteoclasts[29,35,36]. Therefore, we investigated if the bionic coating could promote type H vessel formation via blocking RANK-dependent osteoclast maturation. Notably, HUVECs inoculated on surfaces of different implant materials showed no toxic responses after three days according to CCK-8 assays, LDH release tests and live/dead staining, affirming the non-toxicity of the bionic coating to HUVECs thereof (Fig. 4A and B, S4B). Subsequently, the HUVEC/RAW264.7 co-culture system was established and subjected to different treatment (Fig. S4A). Analysis results showed that both mRNA and protein expression levels of PDGF-BB were significantly upregulated in the Ti/HA/CLP-Apt group (Fig. S4C and D), suggesting that the CLP-Apt micellar treatment effectively blocked the maturation of osteoclast precursors for preserving their PDGF-BB secretion capacity. Immunofluorescence staining further revealed that HUVECs from the Ti/HA/CLP-Apt group showed elevated expression of CD31 and EMCN than the control groups (Fig. 4C–F), indicating that they have greater propensity to form H-type vessels. In line with the trends of PDGF-BB secretion levels after various treatment, transwell migration and wound healing assays demonstrated that HUVECs in the Ti/HA/CLP-Apt group displayed markedly higher migratory activity than all other groups (Fig. 4G–J). In vitro tube formation assays showed that HUVECs in the Ti/HA/CLP-Apt group developed more continuous and complex tubular networks with significantly increased total vessel length and branching points than the other groups (Fig. 4K–M). These findings demonstrate that the micellar LYTAC-mediated RANK deletion could preserve the PDGF-BB secretion capability of osteoclast precursors to promote type H vessel formation and contribute to osteogenesis thereof.

Fig. 4.

Analysis of type H vessel formation after different treatment. A) Calcein/PI staining of HUVEC cells on different titanium-based material surfaces. B) Viability of HUVECs on different Ti-based material surfaces. C, D) EMCN Immunofluorescence staining images and quantitative analysis of HUVECs on different titanium-based materials. Green: EMCN, red: CD31, blue: cell nuclei. E,F) CD31 Immunofluorescence staining images and quantitative analysis of HUVECs on different titanium-based materials. red: CD31, blue: cell nuclei. G-J) HUVEC migration ability assessed by transwell migration assay and scratch assay. K) Tubular network formation of HUVECs after incubation for 6 h on different implants. Dark blue dots indicated nodes, red circles around dark blue dots indicated junctions, light blue circles indicated meshes, yellow circles indicated master segments, blue lines indicated extremities, and green lines indicated branches. L, M) Quantitative analysis of node count in the total vessel length and tubular network of L. Data were presented as mean ± SD for n = 3, **p < 0.01, ***p < 0.001.

Importantly, the treatment-induced enhancement of type H angiogenesis may provide a mechanistic link between RANK-dependent regulation of osteoclast lineage cells and improved implant osseointegration. Accumulating evidence indicates that type H vessels constitute a specialized vascular niche closely associated with osteogenesis by facilitating the recruitment and expansion of osteoprogenitor cells and providing endothelial-derived angiocrine signals. The type H vessel-mediated coupling of angiogenesis and osteogenesis is particularly relevant to osseointegration, as the coordinated development of a functional vascular network and new bone is essential for promoting bone maturation and establishing a stable bone-implant interface. Of note, we observed that micellar LYTAC-mediated RANK degradation limits the maturation of osteoclast precursors into bone-resorbing osteoclasts while maintaining their PDGF-BB secretion, which may promote type H vessel formation at the bone-implant interface, thereby creating a pro-osteogenic vascular microenvironment that supports peri-implant bone formation and enhances osseointegration.

MSCs play pivotal roles in osseointegration by serving as progenitors of osteoblasts [37]. However, the excessive mature osteoclasts in osteoporotic microenvironment substantially suppress the osteogenic capacity of MSCs through promoting adipogenesis and inflammation [38]. Previous studies also confirm that HA contents could attract MSC on account of their intrinsic bioactivity [39,40]. To investigate if the bionic coating could ameliorate mature osteoclast-induced anti-healing stress and promote the osteoblastic differentiation of recruited MSCs, here bone-marrow derived MSCs were seeded onto the surfaces of different samples and cultured in RAW264.7 cell-conditioned media (Fig. 5A). After 3 days of culture, the BMSCs were double stained with calcein-AM/PI to indicate the biocompatibility of individual treatment (Fig. 5B), which showed no obvious cytotoxic effects for all groups. The favorable cytocompatibility of the bionic coating was also supported by results of CCK-8 assay and lactate dehydrogenase (LDH) release assay (Fig. 5C–S5A). Meanwhile, it is also detected that the HA layers significantly upregulated vinculin expression and enhanced BMSC adhesion (Fig. 5D), which was consistent with the observations in previous reports [39,40]. SEM images in Fig. 5E and F demonstrated that BMSCs in the Ti/HA/CLP-Apt group exhibited the largest spreading area, which is indicative of the coating-promoted focal adhesion formation and cytoskeletal extension that are favorable for promoting their osteogenic differentiation. To investigate whether integrin-mediated mechanotransduction contributes to MSC adhesion on the coatings, ROCK activity and αvβ3 integrin function were pharmacologically inhibited. Upon ROCK inhibition with Y27632, the spreading morphology and F-actin organization of MSCs were markedly altered across the different substrates, accompanied by a significant reduction in cell spreading area (Fig. S5B and C), indicating that ROCK-dependent cytoskeletal tension contributes to the maintenance of cell spreading. Similarly, blockade of αvβ3 integrin altered the F-actin organization of MSCs compared with the corresponding non-blocking condition (Fig. S5D and F), suggesting that αvβ3 participates in cell-coating interactions and cytoskeletal organization. These findings are consistent with previous work demonstrating that αvβ3 integrin contributes to adhesion-associated mechanotransduction and couples adhesive interactions to RhoA/ROCK signaling and intracellular tension [41]. We further examined the downstream adhesion-associated signaling and osteogenic response by assessing phosphorylated FAK and Osterix following αvβ3 blockade. The fluorescence intensity of P-FAK was significantly lower on the HA-containing coatings than on Ti (Fig. S5D and E), while a similar pattern was observed for Osterix expression (Fig. S5F and G), suggesting that HA incorporation modulates focal adhesion signaling and the associated osteogenic response. Furthermore, we have examined the expression of key osteogenic genes and functional markers in adherent BMSCs (Fig. 5G–J) and found that the Ti/HA/CLP-Apt treatment substantially upregulated the mRNA levels of Runx2, OCN, Col I and ALP compared with other groups. Consistent with the results above, ALP staining and Alizarin Red S (ARS) mineralization assays (Fig. 5K–M) revealed markedly enhanced ALP activity and calcium nodule formation in the Ti/HA/CLP-Apt group than the other groups and the differences were statistically significant (Fig. 5L–N). These data collectively demonstrated that the Ti/HA/CLP-Apt coating-mediated osteoclast inhibition could promote osteogenic differentiation of adherent MSCs, which may add to the new bone formation at implant interface.

Fig. 5.

Enhanced osteogenic differentiation of MSCs facilitated by Ti/HA/CLP-Apt. A) Schematic illustration showing the co-culture of BMSCs and RAW264.7 cells. B, C) Calcein/PI staining and viability of BMSCs on different Ti-based material surfaces. D) Immunofluorescence staining of vinculin (green) and cytoskeleton (red). E, F) SEM images of BMSCs on different Ti-based material surfaces and quantitative analysis results. G-J) Relative mRNA levels of Runx2, OCN, COI I, and ALP. K, L) ALP staining and quantitative analysis of BMSCs on different Ti-based materials. M, N) ARS staining and corresponding quantitative analysis of BMSCs on different Ti-based materials. Data were presented as mean ± SD for n = 3, *p < 0.05, **p < 0.01, ***p < 0.001.

3.5. In vivo evaluation of treatment-mediated osseointegration efficacy

The osseointegration properties of bionic coating functionalized Ti implants were further investigated on an OVX osteoporosis rat model, in which the implants were inserted into the distal femur metaphysis of rats 3 months after OVX surgery, and the Ti implants were inserted into the distal femur metaphysis of rats 3 months after OVX surgery, and implant and tissue samples were collected at 2 and 3 months after implantation for subsequent analysis (Fig. 6A). Notably, histological analysis of major rat organs, including heart, lung, liver, spleen and kidney, showed no pathological changes after various treatments (Fig. S6A), indicating that the liver and spleen were not affected by the treatment, the good biocompatibility of the biomimetic coating was verified. Micro-CT analysis confirmed the successful establishment of an osteoporosis model. Compared with ctrl group, OVX rats showed impaired trabecular bone structure, BV/TV, TB. N and Tb.Th decreased, and TB. Sp increased (Fig. S6B–F), exhibiting a pronounced osteoporotic phenotype. Micro-CT three-dimensional reconstruction imaging (Fig. 6B) showed that the Ti/HA/CLP-Apt group significantly stimulated new bone formation in the peri-implant area at two and three months compared with the other groups. Quantitative analysis of bone tissue within 1 mm of the area around the implant (Fig. 6C–F) further showed the bone volume fraction (BV/TV) in the Ti/HA/CLP-Apt group, which was similar to that in the control group, trabecular thickness (Tb.Th) and trabecular number (Tb.N) were significantly higher in the Ti/HA/CLP-Apt group than in all other groups. Notably, Ti/HA/CLP treatment induced a modest amount of new bone formation around the implants, which was attributed to the osteogenic effect of HA and CLP components. Hematoxylin and eosin (H & E)(Fig. 6G) and Masson's trichrome (MT) staining (Fig. 6H) results confirmed these findings, and at 2-and 3-month planting times, there was no significant difference between the two groups, among them, abundant newly formed bone tissue was observed around the implants in the Ti/HA/CLP-Apt and Ti/HA/CLP groups, and the new bone thickness in the Ti/HA/CLP-Apt group was the largest among all groups. Notably, new bone thickness increased further at 3 months and remained maximal in the Ti/HA/CLP-Apt group (Fig. S6G). Whereas MT staining showed abundant collagen deposition around these implants. In addition, at two and three months, the Ti/HA/CLP and Ti/HA/CLP-apt groups exhibited greater osteoid (red) formation than the Ti and Ti/HA groups (Fig. 6I), suggesting that the Ti/HA/CLP-apt group is more susceptible to osteoid formation, supporting the osteogenic benefit of CLP-based ECM mimicking fibers. The in vivo results collectively indicate that Ti/HA/CLP-apt implants effectively promote new bone formation to improve osseointegration of functionalized Ti implants under osteoporotic conditions.

Fig. 6.

In vivo evaluation of osseointegration in OVX rat model. A) Schematic illustration of the animal model construction and operation procedures. B) Micro-CT images of new bone formation and trabecular thickness around different implants after implantation for 60 and 90 days. C-F) Quantitative analysis of BV/TV, Tb.N, Tb.Th, and Tb.Sp. G-I) H&E-stained images, masson-stained images and Goldner's Trichrome staining around titanium implant for 60 and 90 days. Green: mineralized bone; orange: represents osteoid; black arrows: multinucleated osteoclasts. Data were presented as mean ± SD for n = 3, **p < 0.01, ***p < 0.001.

On the basis of the above-mentioned peri-implant bone histomorphometric analysis, we further investigated the mechanism of the biomimetic coating in promoting osseointegration by detecting key osteogenic-related proteins and vascular responses by immunohistochemical staining. At two months, the expression of OCN, RUNX2 and Col I was significantly increased in the Ti/HA/CLP-Apt group (Fig. 7A–D), while TRAP staining showed a significant decrease in osteoclast activity in the Ti/HA/CLP-Apt group (Fig. 7E), indicating that the activity of osteoclasts in the Ti/HA/CLP-Apt group was significantly decreased, suggesting enhanced osteogenic differentiation and bone matrix synthesis. Importantly, the same trend was maintained at three months, with a persistent increase in OCN, RUNX2 and Col I expression and a decrease in trap-positive osteoclasts in the Ti/HA/CLP-Apt Group (Fig. S7A and B), the results indicate that the biomimetic coating can effectively promote the osteogenic differentiation of MSCs in the peri-implant area and enhance the synthesis of bone matrix in the long term. On the other hand, the treatment of CLP-Apt micelles at two and three months also greatly promoted the formation of CD31 + EMCN + type H vessels in the periimplantation area (Fig. 7F and G), suggesting that CLP-Apt micelles may play an important role in the formation of CD31 + EMCN + type H vessels in the periimplantation area, this is consistent with the in vitro results of the pro-h-type angiogenic ability of the biomimetic coating. Indeed, microangiographic analysis (Fig. 7H–J and S7C) further revealed that the vascular network around the Ti/HA/CLP-Apt implant exhibited typical morphological features of type H vessels, including greater vessel volume, and that the vascular network around the Ti/HA/CLP-Apt implant was similar to that of the Ti/HA/CLP-Apt implant, greater surface area, more complex branching patterns, and a greater number of interconnected nodes. Collectively, the above results collectively suggest that the biomimetic coating on the surface of titanium implants can block osteoclast-intrinsic RANK signaling to exert a multifaceted pro-osteogenic effect, and that the biomimetic coating on the surface of titanium implants can block osteoclast-intrinsic RANK signaling, leading to superior osteoporotic osseointegration in vivo.

Fig. 7.

In vivo evaluation of osteogenic and angiogenic capacities of the bionic coating in OVX rat model. A-D) Immunohistochemical staining of OCN, RUNX2 and Col Ⅰ positive regions and quantitative analysis results after implantation for 60 days. E) TRAP staining image of new bones around the implant after implantation for 60 days. F, G) Immunofluorescence staining of CD31 and EMCN around the implant and quantitative analysis of CD31⁺ EMCN + positive area after implantation for 60 and 90 days. Blue: DAPI, Red: CD31, Green: EMCN. H, I) Angiography images of epiphyseal and quantitative analysis of vascular volume in angiograms after implantation for 60 days. J) Quantitative analysis of vessel surface in angiograms of Fig. 7H. Data were presented as mean ± SD for n = 3, *p < 0.05, **p < 0.01, ***p < 0.001.

4. Conclusion

In this study, we have developed a multifunctional bionic coating for ameliorating the osteoporotic osseointegration of Ti implants, which was developed through anchoring micellar LYTACs based on aptamer engineered CLPs onto HA-modified Ti implants. The bionic coating could recruit MSCs to the bone-implant interface while releasing the micellar LYTACs. Notably, the LYTACs could tether RANK to adjacent CI-M6PR via bispecific aptamer binding, thus efficiently deleting osteoclast-intrinsic RANK proteins through hijacking the endogenous lysosomal pathway. The LYTAC-mediated RANK deletion yielded multifaceted pro-osteogenic benefit as it not only attenuated the osteolytic functions of hyperactivated osteoclasts in osteoporotic bone microenvironment but also promoted the osteoblastic differentiation of recruited MSCs and type H angiogenesis. Furthermore, the CLP-Apt components could be activated by the acidic lysosomal pH to free-standing CLPs and reassemble to nanofibers, which were excreted to the extracellular compartment via exocytosis and readily chelated with microenvironmental Ca ions to yield bone ECM-like matrices, thus accelerating bone mineral accretion through a fibrillogenesis-mimetic process. The bionic coating efficiently remodeled the osteoporotic microenvironment into a pro-healing state and substantially accelerated the osseointegration of Ti implants in vivo, providing new avenues for improving the therapeutic performance of orthopedic implants in the clinic [42].

CRediT authorship contribution statement

Youbo Zhao: Writing – original draft, Formal analysis, Data curation. Jie Zhou: Writing – original draft, Formal analysis, Data curation. Yi He: Writing – original draft, Formal analysis, Data curation. Yueting Ren: Writing – review & editing, Formal analysis, Data curation. Jiayi Zhang: Writing – review & editing, Methodology. Maohua Chen: Writing – review & editing, Formal analysis, Data curation, Conceptualization. Zhong Luo: Writing – review & editing, Writing – original draft, Supervision, Project administration, Conceptualization. Zipeng Gong: Writing – review & editing, Project administration, Funding acquisition, Conceptualization, Methodology, Writing – review & editing.

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Ethics approval and consent to participate

All animal experiments were reviewed and approved by the Animal Ethics Committee of Army Medical University (Approval Number: AMUWEC20255450).

Declaration of competing interest

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

Appendix A. Supplementary data

The following is the Supplementary data to this article: