Characterization of HNPs

HNPs were purchased from Zhongkekeyou Nano Technology (Beijing, China). The size and morphology were determined by scanning electron microscopy (SEM, TESCAN MIRA LMS, Czech Republic) and transmission electron microscopy (TEM, FEI Tecnai G2 F20, USA). The elemental composition was identified through energy-dispersive X-ray spectroscopy (EDS) analysis. The zeta potential was evaluated by a Zetasizer Nano-ZS system (Malvern Zetasizer Nano ZS90, UK). The structural characteristics of the HNPs were revealed through X-ray diffraction (XRD) using SmartLab SE (Rigaku SmartLab SE, Japan). The functional and vibrational characteristics of the HNPs were identified through Fourier transform infrared spectroscopy (FTIR, Nicolet iN10, USA) and Raman spectroscopy (LabRam HR Evolution, France).

In vivo studiesRat cranial defect model

Gelatine methacryloyl (GelMA, EFL-GM-90, Suzhou Intelligent Manufacturing Research Institute, China) was prepared at 10% (w/v) and then mixed with HNPs at different ratios (0.0001 wt%, 0.0005 wt%, 0.001 wt%, and 0.005 wt%). Each mixture was transferred to a circular mould measuring 1 mm in depth and 5 mm in diameter, then exposed to visible light at 405 nm for 1 min for curing.

Guangdong Huawei Micro Detection Co., LTD. (China) provided 6-week-old male Sprague-Dawley rats for in vivo studies. 48 rats were randomly divided into four groups, each containing 6 rats (n = 6): the control group (not implanted), the GelMA group (implanted with GelMA), the HNPs-GelMA group (implanted with 0.0001/0.0005/0.001/0.005 wt% HNPs-GelMA) and the 0.001 wt% HNPs-GelMA + ML141 (10 µM, Cat# S7686, Selleck, USA) group. Under pentobarbital sodium anaesthesia, circular full-layer bone defects of critical size were created on both sides of the skull using a ring bone drill with an internal diameter of 5 mm. The implants were sutured. The subsequent experiments were conducted two weeks later. The Animal Ethics Committee of Guangdong Huawei Micro Detection Co., Ltd. approved all procedures conducted in this study (202304011).

Three-dimensional (3D) images and microcomputed tomography (micro-CT) analysis

After the rats were anaesthetized, rhodamine B (Cat# R8881, Sigma, Germany) was injected intravenously through the inner canthus. A multiphoton laser scanning microscope (FV1200MPE, Olympus, Japan) was used to observe vascular formation in the cranial defect area of the rats, and assessment was conducted using Imaris 9.0.1. Subsequently, the rats were euthanized by being placed in sealed cages and continuously filling the cages with carbon dioxide.

For micro-CT analysis, the cranial defect area was dissected along with the hydrogel. After 4% paraformaldehyde was added for 3 days, the 3D morphology was assessed by micro-CT (BrukerSkyScan1275x, Germany) at a current of 78 µA and a voltage of 48 kV. The scanning accuracy was 10 microns. Images of the bone defect area were reconstructed based on the micro-CT findings using MIMICS software.

Histological analysis

The cranial defect area samples were collected, fixed with 4% paraformaldehyde (Cat# BL539A, Biosharp, China) for 24 h, decalcified in 10% EDTA (Cat# DD0002, Leagene, China) for 1 month, and embedded in paraffin wax. The sample was sliced into continuous sections with a thickness of 4 μm for subsequent histological evaluation.

Following the manufacturer’s instructions, consecutive cross-sectional slices of the decalcified samples were subjected to haematoxylin and eosin (HE) staining and immunohistochemistry for endothelial cell adhesion molecule 1 (CD31, 1:20, Cat# AF3628, R&D, USA). Additionally, immunofluorescence costaining experiments were performed on tissue sections using primary antibodies against CD31 (1:20, Cat# AF3628, R&D, USA) and Emcn (1:100, Cat# sc-65495, Santa Cruz, USA), as well as secondary antibodies Cy3-conjugated donkey anti-rabbit IgG (1:200, Cat# GB21403, Servicebio, China) and FITC-conjugated donkey anti-goat IgG (1:200, Cat# GB22404, Servicebio, China). Images of the immunofluorescence costained samples were taken via confocal microscopy. The proportion of H-type vessels’ area was then assessed using ImageJ in randomly selected areas.

In vitro experimentsCell culture

Human umbilical vein endothelial cells (HUVECs) were obtained from Cyagen Biosciences, Inc. (Cat. #HUVEC-20,001, Cyagen, China), cultured in endothelial cell medium (ECM, Cat# 1001, ScienCell, USA) supplemented with 5% foetal bovine serum (FBS, Cat. #0025, ScienCell, USA), 1% penicillin/streptomycin (P/S, Cat. #0503, ScienCell, USA) and 1% endothelial cell growth supplement (ECGS, Cat. #1052, ScienCell, USA), and then incubated in incubators (5% CO2, 37 °C). HUVECs from the third to sixth generations were used in the subsequent experiments.

Cell proliferation

HUVECs were inoculated in 96-well plates and incubated with HNPs at concentrations of 0, 1, 10, 100 or 1000 ng/mL for 12, 24 or48 h. 10 µL of Cell Counting Kit-8 (CCK-8, Cat# CK04, Dojindo Laboratories, Japan) solution was added to each well, and the cultures were incubated at 37 °C for 2.5 h. After centrifugation, the supernatant was gathered and the absorbance at 450 nm was assessed using VersaMax Molecular Devices (USA).

Cell migration

For the scratch assay, HUVECs were inoculated in a 6-well plate. After the HUVECs had grown to 100% confluence, a sterile 200 µL pipette tip was used to make cell-free scratches in the middle of each well. PBS was used to rinse off the nonadherent cells, and 0 or 10 ng/mL HNPs prepared from serum-free ECM were added to each well. Images were taken at 0, 12 and 24 h, and the remaining gap areas were quantitatively analysed using ImageJ.

In the Transwell experiment, the treated HUVECs were inoculated into the upper cavity of Transwell cell culture inserts (Cat# 3422, Corning, China) at 2 × 104 cells per well and cultured with 200 µL of ECM (serum free). Then, 600 µL of ECM was added to the lower cavity. 24 h later, the cells that migrated to the lower cavity were fixed with 4% paraformaldehyde for 20 min and then subjected to crystal violet staining (Cat# DZ0055, Leagene, China) for 25 min. The images of each group were obtained by inverted microscope and the counts of cells migrating into the lower chamber were quantified by ImageJ software.

Tube formation

HUVECs were inoculated in 12-well plates and treated with 0 or 10 ng/mL HNPs. Matrigel (Cat# 356,231, BD Bioscience, USA) was added to the ibidi slides (Cat# 81,506, Ibidi, Germany), followed by a 30-minute incubation at room temperature. Then, the slides were transferred to a 37 °C incubator for an additional 30 min. Subsequently, 50 µL of 5 × 103 HUVECs was added to both groups of media and seeded into the ibidi plates coated with Matrigel. After 4 h, the cells were photographed with a stereomicroscope (Leica, Germany), and the numbers of nodes, meshes and the master segment length were quantitatively analysed with ImageJ.

Immunofluorescence staining

HUVECs were fixed with 4% paraformaldehyde (Cat# DF0135, Leagene, China) for 20 min, and the membranes were disrupted with 0.5% Triton X-100 (Cat# T6200G, Biotopped, China) for 5 min and then blocked with 5% immunostaining blocking solution (Cat# P0260, Beyotime, China) for 1 h. After washes with PBS, the HUVECs were incubated overnight with primary antibodies against EphrinB2 (1:100, Cat# AF6343, Affinity, USA) and VAV2 (1:100, Cat# 21924-1-AP, Proteintech, USA) at 4 °C. Subsequently, the HUVECs were incubated with FITC-labelled goat anti-rabbit IgG (1:200, Cat# GB22303, Servicebio, China) and Cy5-labelled goat anti-mouse IgG (1:200, Cat# GB27301, Servicebio, China) secondary antibodies at room temperature for 1 h. The cells were then incubated with FITC-labelled phalloidin (1:150, Cat# CA1620-300T, Solarbio, China) at 37 °C for 30 min. Finally, the cells were incubated with DAPI staining solution containing antifade mounting medium (Cat# G1407-25ML, Servicebio, China) for 20 min and imaged using laser confocal microscopy (Stellaris 5, Leica, Germany).

For preparation of FITC-HNPs, BSA-FITC solution (Cat# SF063, Solarbio, China) was mixed with HNPs at a ratio of 1:10, followed by incubation for 8 h in a dark incubator at 37 °C. After the mixture was centrifuged at 14,000×g at 4 °C for 35 min, the collected FITC-HNPs samples were resuspended in the ECM.

RNA-seq assay

Total RNA was extracted from HUVECs exposed to 0 or 10 ng/mL HNPs using TRIzol reagent (Cat# 15596026, Thermo Fisher, USA). The purity of the extracted samples was evaluated using a NanoDrop spectrophotometer (IMPLEN, CA, USA). Once the samples were deemed suitable for detection, 1 to 3 µg of total RNA was extracted from each sample to construct the transcriptome sequencing library. Transcriptome sequencing and data analysis were conducted on an Illumina NovaSeq 6000 S4 platform.

Reverse transcription‒polymerase chain reaction (RT-PCR)

Total RNA was extracted using the SteadyPure Universal RNA Extraction Kit (Cat# AG21017, Accurate Biology, China). Followed by reverse transcription using the Evo M-MLV Reverse Transcription Kit (Cat# AG11706, Accurate Biology, China). RT-qPCR was performed using the SYBR Green Pro Taq HS Premix qPCR Kit (Cat# AG11701, Accurate Biology, China). The primer sequences utilized in this experiment are depicted in Table S1. The relative expression of VEGF, CD31 and EFNB2 mRNA were normalized to the expression level of glyceraldehyde-3-phosphate dehydrogenase (GAPDH).

Western blotting (WB)

Cells were seeded at a density of 6 × 105 cells per well in a 6-well plate and treated with 10 ng/mL HNPs or 10 µmol/mL holmium solution (Cat#H117331, Aladdin, China). Cell proteins were extracted from cells that reached 90% confluence using RIPA lysis buffer (Cat# P0013B, Beyotime, China) with PMSF (Cat# ST506, Beyotime, China) and phosphatase inhibitor (Cat#CW2383S, CWBIO, China). The protein concentration extracted from the samples was determined using the BCA Protein Assay kit (Cat# P0010, Beyotime, China). Then, protein samples from each group were separated by SDS‒polyacrylamide gel electrophoresis (Cat# P0014D, Beyotime, China) and transferred onto a PVDF membrane, which was blocked with 5% skim milk powder (Cat# P0216, Beyotime, China) at 25℃ for 1 h. Antibodies against CD31 (1:1000, Cat# 11265-1-AP, Proteintech, USA), VEGF (1:1000, Cat# 19003-1-AP, Proteintech, USA), EphrinB2 (1:1000, Cat# sc-398,735, Santa, USA), phosphorylated EphrinB2 (Tyr304) (1:1000, Cat# AF7321, Affinity, China), VAV2 (1:1000, Cat# 21924-1-AP, Proteintech, USA), phosphorylated VAV2 (Tyr142) (1:1000, Cat# AF4446, Affinity, China), CDC42 (1:1000, Cat# 10155-1-AP, Proteintech, USA), and GAPDH (1:1000, Cat# 10494-1-AP, Proteintech, USA) were then incubated with the membranes overnight at 4 °C. Following TBST washes, the membrane was incubated with HRP-conjugated anti-rabbit IgG (1:5000, Cat# 7074P2, CST, USA) or anti-mouse IgG (1:5000, Cat# 7076P2, CST, USA) at 25℃ for 1 h. The proteins on PVDF membranes were visualized using a high-sensitivity chemiluminescence kit (Cat# P0018AS, Beyotime, China).

Immunoprecipitation (IP) and co-immunoprecipitation (CO-IP)

IP experiments were used to determine the expression of active CDC42. HUVECs were cultured in 10 cm culture dishes. Upon reaching 90% confluence, the cells were treated with corresponding reagents according to different groups. After 24 h, the cells were lysed with lysis buffer at 4 °C for 30 min, followed by centrifugation at 4 °C for 30 min to obtain the supernatant. Subsequently, active CDC42 antibody (Cat# 26905, WuHan NewEast Biosciences Co., Ltd., China) and protein G beads were added and incubated at 4 °C for 1 h. Afterwards, the beads were washed three times with lysis buffer. After the last wash, all the supernatant was carefully removed. The samples were resuspended in 20 µL of 2x loading buffer and boiled for 5 min at 100 °C, the supernatant was aspirated and retained. The final samples were analysed by protein blotting using an anti-CDC42 antibody.

CO-IP assays conducted with a magnetic immunoprecipitation kit (Cat# Bes3011, Biotech Group Co., Ltd., China), and the results were visualized via WB. Protein samples from HUVECs transfected with the EFNB2 and VAV2 overexpression plasmids were extracted in 10 cm culture dishes, and then, 500 µL of the supernatant was incubated with 8 µg of IP antibodies (EphrinB2 or VAV2) or control IgG at 4 °C for 12 h. In the mixture obtained from the previous step, add magnetic beads and incubate at 4 °C for 2 h. Wash the magnetic beads four times with lysis buffer and discard the supernatant. Add elution buffer and boil the mixture for 10 min. Retain the supernatant, then add 5x loading buffer and boil for 5 min. Finally, WB experiments were carried out.

Transfection

HUVECs were placed in 10 cm culture dishes at a density of 10 × 105 cells/well. Lipofectamine 3000 (Cat# L3000001, Thermo Fisher, USA) in Opti-MEM I reduced serum medium (Cat# 31,985,070, Thermo Fisher, USA) was used to transiently transfect siEFNB2, siVAV2 and the overexpression plasmids pcDNA3.1(+)-EFNB2-3xFLAG and pcDNA3.1(+)-VAV2-HA. The siRNAs used in this study were designed and constructed by OBiO Technology Co., Ltd. (China) (Table S2).

Analysis of ROS

HUVECs were inoculated in 6-well plates at a density of 5 × 105 cells/well and cultured with ECM containing 0, 1, 10, 100 or 1000 ng/mL HNPs for 24 h. Each sample was incubated with 10 µM DCFH-DA (Cat# S0033S, Beyotime, China) for 20 min, followed by three washes with PBS. The samples were resuspended in PBS and immediately analyzed for ROS levels using a flow cytometer (BD FACSAria III, USA).

In addition, the positive control group was treated with active oxygen positive control reagent for 30 min. One millilitre of serum-free ECM containing 10 µM DCFH-DA was added to each well and incubated in the cell incubator in the dark for 20 min. After washing the cells three times with PBS, 1 mL of ECM (serum free) was added to each well, followed by immediate image capture using a laser scanning confocal microscope.

Molecular dynamic (MD) simulations

This study obtained the structure files for EphrinB2 and VAV2 from the PDB database. Protonation under neutral conditions was initially carried out using an H + + 3 online server. Pymol software was then used to remove the heteroatoms and water molecules from the crystal structure, leaving only the protein structure. HawkDock was used for interconnection, and the configuration with the highest initial interconnection score was selected for further analysis. The configuration with the highest interconnection score was selected from among the interconnection configurations for calculation by the MMGBSA of Hawdock. PyMOL 2.04 and ChimeraX were used for 3D graphic analysis, and ligplus + was used for two-dimensional effect analysis [21, 22].

Statistical analyses

All the data are expressed as the mean ± standard deviation (SD) with a minimum sample size of ≥ 3. Statistical comparisons were performed using univariate analysis of variance (ANOVA) or t tests (if applicable) with GraphPad Prism software. P < 0.05 was considered to indicate statistical significance.