Nerve growth factor-loaded biomimetic prussian blue nanocomplexes for reversing osteoporosis via promoting osteoblast precursor cell proliferation and differentiation

Abstract

The fundamental issue of osteoporosis (OP) is osteoblast decrease due to oxidative stress and the subsequent disruption of the osteogenic and osteoclastic dynamic balance. How to promote the proliferation and osteogenic differentiation of osteoblast precursor cells (MC3T3-E1) in an oxidative microenvironment is a great challenge for improving OP. In this study, Prussian blue nanoparticles (PB NPs) were initially functionalized with a polydopamine (PDA) coating. Nerve growth factor (NGF) was subsequently immobilized on the PDA layer, followed by the hybrid membrane coating composed of red blood cell membrane (RBCm) and MC3T3-E1 cell membrane (MC3T3m), thereby constructing a biomimetic Prussian blue nanocomplex loaded with NGF (MPDN NPs). In vitro studies have shown that the nanodrug restored the impaired proliferation viability of MC3T3-E1 cells and inhibit their apoptosis by scavenging reactive oxygen species (ROS), and further cooperate with NGF to promote osteogenic differentiation. In vivo studies have demonstrated that the nanodrug significantly inhibited bone loss and promote bone regeneration in osteoporotic mice. Moreover, this nanodrug with excellent safety both in vitro and in vivo showed the long half-life in the bloodstream and high accumulation in the bone. In summary, this strategy addresses the fundamental issue of decreased osteoblast in OP and offers a novel approach for preventing and treating OP.

Graphical abstract

The preparation process and anti-osteoporosis effect of MPDN NPs.

Highlights

• •Solving the fundamental problem of decreased osteoblast in osteoporosis.
• •Confirming the anti-osteoporotic potential of nerve growth factor.
• •Achieving the nanoformulation of nerve growth factor.
• •Preparing a multifunctional nanodrug which regulates the behavior of MC3T3-E1 cells.

1. Introduction

Osteoporosis (OP), is a systemic bone disorder that characterized by a loss of bone mass and microstructural damage to bone tissue, which leads to increased fragility and fracture risk [1]. According to statistics, the global prevalence of OP is 19.7 % [2]. With the accelerating pace of global aging, the social and economic burdens of OP and its complication, osteoporotic fractures, are increasing [3]. Currently, clinical medications and scientific research for OP primarily focus on anti-resorptive therapy targeting osteoclasts, such as bisphosphonates, which are the most studied and widely used. However, these strategies did not address the fundamental issue of impaired osteoblast function in OP and may even lead to atypical long bone fractures and bisphosphonate-related osteonecrosis of the jaw [4]. Recently, more and more studies have shown the excess of reactive oxygen species (ROS) at osteoporotic sites [5]. Excessive ROS can trigger the apoptosis of osteogenic precursor (MC3T3-E1) cells, inhibit their proliferation and osteogenic differentiation, and hinder osteoblast formation and maturation [6]. It is greatly beneficial to develop safe and effective anti-osteoporotic drugs, which can fundamentally address the issue of decreased osteoblast.

Nerve growth factor (NGF) is one of the earliest discovered neurotrophins, which has been widely used in research and clinical trials for neurological diseases. Meanwhile, clinical observations and recent scientific studies suggest that NGF is involved in bone regeneration [7,8]. Meyers and Li have both confirmed the important role of NGF in bone healing [9,10]. Zhang et al. constructed an engineered sensory nerve loaded with NGF to treat bone defects through the synergistic effect of innervation and osteogenesis [11]. Wang et al. also doped NGF into nanofibrous scaffolds for promoting bone formation [12]. As a water-soluble protein with a molecular weight of approximately 13 kDa, NGF can be easily hydrolyzed by enzymes in vivo, making it difficult to accumulate at the osteoporotic sites, accordingly limits the medical applications [13]. The emergence of nanodrug brings hope for addressing this issue. Nanomaterials such as nanofiber, mesoporous silica nanoparticles and poly(lactic-co-glycolic acid) have been used to load NGF [[14], [15], [16]]. However, these nanocarriers without antioxidant capability cannot eliminate excessive ROS in the osteoporotic microenvironment.

As an antidote approved by the Food and Drug Administration, Prussian blue nanoparticles (PB NPs) with good biosafety showed superior antioxidant activity [17]. Moreover, PB NPs have been proved to normalize the bone tissue microenvironment by inhibiting the generation of intracellular ROS, thereby delaying OP [18,19]. However, naked nanoparticles are easily recognized as foreign substances and rapid clearance by human immune system, which was embodied in the short half-life. Recently, encapsulating natural cell membranes on nanoparticles has been widely adopted to, help nanocomplexes evade clearance by the body's immune system, prolong their circulation time in the bloodstream, and increase the accumulation of drugs at osteoporotic sites [20].

Therefore, targeting the "high apoptosis, low osteogenesis" characteristics of osteoporotic sites and the application defects of NGF, this study loaded NGF onto the polydopamine (PDA) layer of PB NPs, combined with a hybrid membrane coating [21], to construct a NGF-loaded biomimetic Prussian blue nanocomplex (MPDN NPs). This achieves extended nanodrug half-life and enrichment of PB NPs and NGF at the osteoporotic sites, thereby fully exerting the role of scavenging ROS and promoting the proliferation and osteogenic differentiation of MC3T3-E1 cells, and ultimately obtaining significant anti-osteoporotic effects (see Scheme 1).

Scheme.

Synthesis process of MPDN NPs and the mechanism of anti-osteoporosis.

2. Materials and methods

2.1. Materials and reagents

K₃[Fe(CN)₆]•3H₂O was provided by Fuchen Chemical Reagent Factory (Tianjing, China). Nerve growth factor was provided by MedChemexpress (USA). Fetal bovine serum and α-MEM basal medium were purchased from Procell (Wuhan, China). Alkaline phosphatase staining kit was provided by Beyotime (Shanghai, China)). Alkaline phosphatase activity assay kit was purchased from Jiancheng (Nanjing, China). Alizarin red staining solution was purchased from Cyagen (Shanghai, China).

2.2. Cells and animals

Osteoblast precursor cells (MC3T3-E1) were provided by Procell (Wuhan, China). Four-month-old male SAMP6 and SAMR1 mice were provided by Peking University Health Science Center. The animal use protocol has been approved by the Institutional Animal Care and Use Committee, The Second Xiangya Hospital, Central South University, China (No. 20241067), and none of the animal experiments violated the aims of the Declaration of Helsinki of the Ministry of Health.

2.3. Preparation of MPDN NPs

Red blood cell membrane (RBCm): First, EDTA-coated tubes were used to collect whole blood from BALB/c mice. After plasma removal, the red blood cells were washed with 1 × PBS and lysed in 0.25 × PBS for 6h at 4 °C. After centrifugation and washing them with 0.25 × PBS, aliquots of the RBCm were prepared and stored at −80 °C for later use.

MC3T3-E1 cell membrane (MC3T3m): Initially, MC3T3-E1 cells were lysed for 30 min at 4 °C, following their resuspension in cold pre-extraction reagent supplemented with PMSF. Next, the cell suspension underwent ultrasonic disruption for a duration of 10 min, which was succeeded by four cycles of freezing at −80 °C and thawing at 37 °C. The supernatant was first collected through centrifugation at 800 rpm for a duration of 10 min, which subsequently underwent further centrifugation at 13000 rpm for 30 min, resulting in the precipitation of MC3T3m. Prior to aliquotting, we conducted an analysis of the total protein content within MC3T3m cells and subsequently stored them at −80 °C for later use.

Hybrid membrane: MC3T3m and RBCm solutions were blended together in a 1 : 1 mass ratio, subjected to ice bath ultrasonication for 3 min (80 W), and subsequently stirred at 37 °C for 1 h.

PB NPs: PB NPs were produced following a previous study [22]. For Solution A, FeCl₃•6H₂O (54.1 mg) was dissolved in 2 mL of deionized water and subsequently diluted to a concentration of 1 mM using a 25 mM citric acid solution. This diluted solution was then heated to 60 °C through an oil bath method. In parallel, Solution B was prepared by dissolving K₄[Fe(CN)₆]•3H₂O (8.5 mg) in a 25 mM citric acid solution to achieve a 1 mM concentration of K₄[Fe(CN)₆]. This solution was likewise heated to 60 °C in an oil bath. Following this, Solution B was dribbled into Solution A and the mixture was stirred for 30 min. Ultimately, the PB NPs was obtained by centrifugation, water washing and lyophilization.

PD NPs: To 500 μL of PB solution (4 mg/mL), 1.5 mg of dopamine hydrochloride was added. The mixture underwent sonication in a water bath for 15 min, subsequently added 2 mL of a 10 mM Tris-HCl solution (pH = 8.5). The solution was agitated at 35 °C and 800 rpm for 1 h. Following centrifugation and washing with deionized water, the precipitates were obtained as polydopamine-coated PB NPs (PB@PDA, PD).

PDN NPs: NGF solution (0.1 mL, 100 μg/mL) and PD solution (2 mL, 1 mg/mL) was mixed and agitated at 4 °C and 600 rpm for 16 h. After centrifugation, the precipitates were NGF-loaded PD NPs (PB@PDA@NGF, PDN).

MPDN NPs: PDN solution (1 mL, 1 mg/mL) was mixed with biomimetic hybrid membranes solution (1 mL, 1 mg/mL). The mixture underwent sonication for 3 min in a water bath, followed by stirring at 37 °C and 800 rpm for 2 h. After centrifugation, the precipitates were biomimetic hybrid cell membrane-coated PDN NPs (M@PB@PDA@NGF, MPDN).

2.4. Characterization of nanocomplexes

The microstructure of the NPs was observed under a transmission electron microscope (TEM). The encapsulation rate of NGF under different conditions was determined by Elisa. The elemental content of the NPs was detected via energy dispersive spectroscopy (EDS). The size and surface potential of the NPs were detected via Dynamic light scattering (DLS). MC3T3m was labeled with 3,3′-dioctadecyloxacarbocyanine perchlorate (DiO), and RBCm was labeled with 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate (DiI). The fluorescence of RBCm, MC3T3m, and hybrid membranes was detected through a confocal laser scanning microscope (CLSM). Protein samples of RBCm, MC3T3m, hybrid membranes, and MPDN NPs were prepared according to the instructions of the membrane protein extraction kit. Following separation by SDS-PAGE, the total protein map was stained with coomassie brilliant blue and photographed.

2.5. In vitro cellular uptake behavior

RAW264.7 and MC3T3-E1 cells were incubated with PD^Ce6 and M@PD^Ce6 for different time periods (0, 1, 4 h). CLSM was used to observe intracellular fluorescent signals.

2.6. In vivo plasma pharmacokinetics and biodistribution

Pharmacokinetics: After intravenous injection of PD^Ce6 and M@PD^Ce6 (Ce6, 2.5 mg/kg), eyelid blood was collected from adult male BALB/c mice at 0, 0.5, 2, 4, 8, 12 and 24 h. Blood fluorescence intensity was measured by small animal living imaging system, and the circulation half-life (t1/2) was calculated.

Biodistribution: Four-month-old male SAMP6 mice were intravenously injected with PD^Ce6 and M@PD^Ce6 (Ce6, 2.5 mg/kg). After 24 h, the lower limb bones and main organs of the mice were collected and subjected to fluorescence detection under small animal living imaging system.

2.7. In vitro anti-oxidation and pro-proliferation effect

ROS Scavenging Experiment: After pretreating MC3T3-E1 cells with MPDN NPs for 2 h, the cells were treated with H₂O₂ (400 μM) for 4 h, then incubated with DCFH-DA probe (10 μM) in the dark for 30 min. Fluorescent images were observed and captured by CLSM. Additionally, cells with the same treatment were digested with trypsin without EDTA, collected and detected by flow cytometry.

2,2-Diphenyl-1-picrylhydrazyl (DPPH) scavenging experiment: The capacity of MPDN NPs to scavenge radicals was evaluated using the DPPH scavenging assay. Briefly, a DPPH working solution in anhydrous ethanol was prepared and mixed thoroughly with NGF, MPD, and MPDN NPs. After incubation in the dark for 30 min, the absorbance at 517 nm was measured.

Flow cytometry: Drug treatment was the same as in the ROS scavenging experiment. The cells were digested, collected, and stained using the Annexin V-FITC/PI kit. Finally intracellular apoptosis was detected by flow cytometry.

Western blot analysis: Drug treatment is the same as above. The cells were collected to extract total protein. Proteins underwent separation via SDS-PAGE, were subsequently transferred onto a membrane, and then incubated with specific antibodies targeting Bax and Bcl2.

MTT assay: After pre-treating with MPDN NPs for 2 h, MC3T3-E1 cells were treated with H₂O₂ (400 μM) for 24 h. Cell viability was detected using an MTT kit. The absorbance at 490 nm was measured.

EdU staining: Drug treatment was the same as above. Staining was performed using the EdU kit, and fluorescent images were observed and captured under CLSM.

Immunofluorescence detection (Ki67): Drug treatment was the same as above. The cells were fixed with 4 % paraformaldehyde (PFA) for 30 min, incubated with 0.5 % Triton X-100 for 20 min, blocked with 5 % goat serum for 1 h, and incubated with the primary antibody (Ki67) at 4 °C overnight. On the next day, the cells were incubated with the secondary antibody, stained with DAPI, and finally detected by CLSM.

2.8. In vitro promotion of osteogenic differentiation

Alkaline phosphatase (ALP) staining and activity detection: After 7 days of osteogenic induction, with medium changes every 2–3 days and re-addition of H₂O₂ and therapeutic drugs during each medium change. The cells were observed and photographed via an inverted microscope after staining for 30 min. Additionally, cells from the same 7-day osteogenic induction were lysed on ice with 0.5 % Triton X-100, sonicated, centrifuged, and the supernatant was collected. ALP activity and total protein concentration were detected according to the instructions of the BCA and the ALP assay kit, respectively.

Alizarin Red S (ARS) staining: After 21 days of osteogenic induction, the cells were fixed with 4 % PFA for 30 min. ARS staining solution was added to each well. The cells were observed and photographed via an inverted microscope after 30 min.

Immunofluorescence detection: After 3 days (Runx2) and 7 days (Col Ⅰ, OCN) of osteogenic induction, the cells were fixed with 4 % PFA before incubation with 0.5 % Triton X-100. Next, the cells were blocked with 5 % goat serum, and incubated with the primary antibody (Runx2, Col Ⅰ, OCN) at 4 °C overnight. On the next day, the cells were incubated with the secondary antibody, stained with DAPI, and finally detected by CLSM.

2.9. In vivo anti-osteoporotic effects

Animal Experiment: The senescence accelerated mouse (SAM) is an inbred strain model selected by Kyoto University in Japan during the breeding of AKR/J strain mice [23]. The SAMP6 mice, characterized by decreased bone density, reduced bone mass, and vascular calcification, serve as an ideal model for studying senile osteoporosis [[24], [25], [26], [27], [28]]. In contrast, the SAMR1 mice belonging to the senescence-resistant series represent mice with normal aging. Their physiological indicators and lifespan are similar to those of normal animals, making them often used as controls for the SAMP strain mice. SAMR1 mice served as the control group, while SAMP6 mice were randomly divided into four groups: Model, NGF (10 μg/kg), MPD (PB NPs, 2.5 mg/kg), and MPDN (NGF, 10 μg/kg; PB NPs, 2.5 mg/kg), with six mice in each group. After one week of acclimation, drugs were administered via tail vein injection every two days. The mice were euthanized six weeks later.

Micro-CT and histological evaluation: The left femurs of mice from each group were collected, fixed in 4 % PFA for 24 h, and stored in 75 % ethanol. Micro-CT was performed to assess changes in bone microstructure. For histological analysis, the left tibiae were collected and fixed in 4 % paraformaldehyde for 24 h, decalcified in 10 % EDTA for 30 days, embedded in paraffin, and sectioned at 5 μm thickness. As for histological evaluation, H&E, Masson, Trap, and immunohistochemical and immunofluorescence staining were all included.

Biomechanical evaluation: The biomechanical properties of the right tibiae were assessed using a three-point bending test. The right tibiae from each group were placed on a materials mechanical performance tester (uTS-100N/1.6 KN) with a span of 5 mm. The loading speed was set at 0.155 mm/s, and the load was gradually increased until the femur fractured. The maximum load and break load were calculated from the load-deflection curve obtained from the instrument.

2.10. Proteomic analysis of mouse bone

The femurs and tibiae of mice from each group were promptly immersed in liquid nitrogen for 15 min to achieve rapid freezing and were subsequently stored at −80 °C for proteomic analysis.

2.11. In vitro biocompatibility and in vivo safety assay

MTT assay: MC3T3-E1 cells and human umbilical vein endothelial cells (HUVECs) were co-cultured with varying concentrations of MPDN NPs for 24 h. Subsequently, the cell viability was assessed utilizing an MTT assay kit.

Coagulation test: Plasma was incubated with PBS, thrombin, NGF (NGF, 100 ng/mL), MPD NPs (PB NPs, 25 μg/mL), and MPDN NPs (PB NPs, 25 μg/mL) at 37 °C for 6 h. The absorbance at 540 nm was measured.

Hemolysis Test: Red blood cells were co-cultured with varying concentrations of MPDN NPs at 37 °C for 6 h. Afterward, a low-speed centrifugation process was applied. A microscope was utilized to observe morphological alterations of red blood cells present in the precipitate. Additionally, the absorbance at 540 nm of the supernatants were determined to assess the rate of hemolysis.

Zebrafish development experiment: NGF, MPD, and MPDN NPs were co-cultured with zebrafish embryos for 5 days. Heartbeat and body length were recorded daily under an inverted microscope and photographed.

H&E staining and hematological examination: H&E staining was employed to evaluate subtle structures of the main organs. Blood specimens were obtained from mice for the purpose of conducting biochemical and hematological assessments.

2.12. Statistical analysis

Data analysis was performed using GraphPad Prism (version 8.3.0). The one-way ANOVA analysis was employed, and the results are expressed as mean ± standard deviation. Representative data are presented. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

3. Results and discussion

3.1. Drug concentration screening based on the "high apoptosis, low osteogenesis" microenvironment induced by ROS at osteoporotic sites

The pathogenesis of OP is related to the high apoptosis and low osteogenic differentiation of osteoblast precursor cells induced by ROS [29,30]. Therefore, we first observed the levels of ROS, cell apoptosis, and osteogenic differentiation in the tibia of osteoporotic mice. The metaphysis is richly populated with trabecular bone, and osteoblasts are located on the surface of these trabecular bone. H&E staining showed that the trabecular bone was sparser in the Model group compared to the Control group (Fig. 1A). Dihydroethidium (DHE) staining revealed stronger red fluorescence in the bone marrow and the surface of the trabecular bone in the tibial metaphysis of Model group compared to the Control group (Fig. 1B). This result suggested overall elevation of ROS level in the microenvironment of the osteoporotic area, encompassing both osteoblasts and bone marrow cells. This phenomenon is also corroborated by published studies [31]. Immunohistochemical staining showed downregulation of the anti-apoptotic protein Bcl2 and upregulation of the apoptotic protein Bax in the Model group (Fig. 1C). Further OCN and TUNEL dual staining, showed more colocalization of red and green fluorescence in the Model group, confirming increased apoptosis of osteoblasts at osteoporotic sites (Fig. 1D). These results suggested that the abnormal accumulation of ROS in the osteoporotic microenvironment leads to increased osteoblasts apoptosis and weakened osteogenic differentiation ability, inhibiting bone regeneration [32].

Fig. 1.

The "high apoptosis, low osteogenesis" microenvironment induced by ROS at osteoporotic sites and drug concentration screening. (A) Representative images of H&E staining of tibia tissue. (B) Representative images of DHE staining (red) of tibia tissue. (C) Representative immunohistochemistry staining of Bcl2 and Bax in tibia sections. (D) Representative immunofluorescence double staining of OCN (red) and TUNEL (green) positive cells in tibia sections. (E–F) Cell viability of the MC3T3-E1 cells incubated with NGF or PB NPs for 24 h. (G) Flow cytometry identification of ROS in MC3T3-E1 cells incubated with PB NPs (20, 25, 30, 35 μg/mL). (H) ALP and ARS staining of MC3T3-E1 cells treated with NGF (0, 50, 100 ng/mL). (I) ALP activity of MC3T3-E1 cells treated with NGF (0, 50, 100 ng/mL). n = 3.

Given the microenvironment characteristics of "high apoptosis, low osteogenesis" in OP, we determined the optimal cellular doses of PB NPs and NGF by focusing on the ROS-scavenging capability of PB NPs and the osteogenic differentiation-promoting effect of NGF. Firstly, the safe concentrations of PB NPs (0–50 μg/mL) and NGF (0–150 ng/mL) within 24 h were determined by MTT assay (Fig. 1E and F). Additionally, long-term cytotoxicity assay (72 h) for PB NPs and NGF revealed that the viability of MC3T3-E1 cells in all groups remained above 80 %, indicating that PB NPs (0–50 μg/mL) and NGF (0–150 ng/mL) exhibited minimal long-term cytotoxicity towards MC3T3-E1 cells (Fig. S1 and 2). Next, to screen for the optimal concentration of PB NPs, we co-cultured MC3T3-E1 cells with PB NPs and used 400 μM H₂O₂ as the inducer of ROS [19]. Flow cytometry was employed to assess the ROS-scavenging ability of different concentrations of PB NPs. The results indicated that PB NPs at concentrations ranging from 20 to 35 μg/mL significantly reduced the ROS levels in MC3T3-E1 cells, demonstrating remarkable ROS-scavenging capabilities of PB NPs. Besides, the ROS scavenging capacity was saturated at 25 μg/mL (Fig. 1G). As for NGF, we co-cultured MC3T3-E1 cells with NGF for 7 days. ALP staining and activity assay revealed that NGF promoted the secretion of ALP in MC3T3-E1 cells, with the most significant effect at the concentration of 100 ng/mL (Fig. 1H and I). After 21 days of osteogenic differentiation induction, the results of ARS staining were consistent with those of ALP staining, showing that 100 ng/mL of NGF enhanced bone mineralization most effectively (Fig. 1H). Therefore, based on the above experimental results, we determined the concentration of PB NPs to be 25 μg/mL, and the concentration of NGF to be 100 ng/mL in MPDN NPs.

3.2. Preparation and characterization of MPDN NPs

First, according to our previously reported methods [22,33,34], we prepared PB NPs (P), PB NPs coated with polydopamine (PD), PD loaded with NGF (PDN), and PDN coated with hybrid membranes (MPDN), with the preparation process shown in Fig. 2A. TEM images showed successful synthesis of uniform cubic PB, PD and MPDN NPs. Additionally, after membrane coating modification, a double-layer membrane structure appeared on the outermost layer of MPDN NPs (Fig. 2B). As shown in Fig. 2C, when the mass ratio of PB to NGF is 200 : 1 and the stirring time is 20 h, the NGF encapsulation rate measured by Elisa exceeds 80 %. Dynamic light scattering (DLS) analysis measured the average particle size of MPDN NPs to be 202.4 ± 0.8 nm (Fig. 2D) and the zeta potential to be −19.20 ± 1.57 mV (Fig. 2E). We also evaluated the stability of MPDN NPs in different solvents (Water, PBS, α-MEM with 10 % FBS) using DLS. It can be seen that the particle sizes of MPDN NPs in three solvents consistently stayed within the range of 200–500 nm, demonstrating the excellent stability of MPDN NPs within 72 h (Fig. 2F).

Fig. 2.

Characterization of MPDN NPs. (A) Schematic representation of the synthesis of MPDN NPs. (B) TEM images of P, PD and MPDN NPs. (C) NGF entrapment efficiency of MPDN at the different ratio and stirring time. (D–E) Diameter and potential of P, PD, PDN and MPDN NPs. (F) Stability of MPDN NPs in water, PBS, total medium with 10 % FBS (0, 24, 48, 72 h). (G-I)The element content of C, N, S, Fe in P, PD, PDN NPs. (J) CLSM images of hybrid membrane. (red: RBCm, green: MC3T3m). (K) SDS-PAGE of the membrane proteins (1: MC3T3m, 2: RBCm, 3: Hybrid membrane, 4: MPDN NPs).

To further confirm the successful preparation of MPDN NPs, EDS was employed to characterize the elemental composition of NPs. PB, PD and PDN NPs all contained the characteristic element Fe of PB NPs. Compared to PB and PD NPs, the content of the S element (a specific element of NGF) in PDN increased from 0.54 % and 0.53 % to 2.82 %, respectively, indicating successful NGF loading (Fig. 2G–I). In addition, we labeled RBCm and MC3T3m with red and green fluorescence, respectively, and used CLSM to observe membrane fusion efficiency. A large amount of yellow light overlapping from red and green light was observed (Fig. 2J), confirming successful synthesis of the hybrid membrane. SDS-PAGE detection showed that MPDN NPs and hybrid cell membranes simultaneously have similar protein profiles including RBCm and MC3T3m (Fig. 2K). The results indicate successful fusion of the hybrid cell membranes and their successful coating on PDN NPs.

3.3. In vitro cellular uptake behavior and in vivo biodistribution

The presence of biomimetic hybrid membranes (RBCm and MC3T3m) in nanocomplexes enables them to evade the immune system, prolong blood circulation time, and facilitate the enrichment of NGF and PB NPs at osteoporotic sites to exert their respective functions [20].

Firstly, we utilized the phagocytosis of macrophages to simulate the in vivo immune environment and assess the ability of biomimetic hybrid membrane-coated nanomaterials to escape immune system surveillance. The Fig. 3A shows that compared to PD^Ce6, M@PD^Ce6 group exhibited almost no red fluorescence, indicating that the hybrid membrane coating can effectively reduce the uptake of nanomaterials by macrophages, endowing the nanomaterials with significant immune escape capability. It is noteworthy that the capability of immune escape is mainly played by RBCm rather than MC3T3m (Fig. S3), which is consistent with previous studies [[35], [36], [37], [38], [39]]. Besides evading immune system surveillance, nanocomplexes also need to be internalized by MC3T3-E1 cells to exert their pharmacological effects [21]. Therefore, we evaluated the uptake efficiency of MC3T3-E1 cells for the nanocomplexes at different time points. It was observed that as the incubation time increased, the red fluorescence within MC3T3-E1 cells became increasingly intense, indicating that the nanocomplexes could be internalized by MC3T3-E1 cells, and within a certain time range, the uptake efficiency of MC3T3-E1 cells for the nanocomplexes was time-dependent (Fig. 3B). Furthermore, to clarify the intake mode of MPDN NPs by MC3T3-E1 cells, we synthesized PDN NPs encapsulated by these DiI-labeled hybrid membranes (DiI-M@PDN NPs), and found that the DiI-M@PDN NPs were internalized by MC3T3-E1 cells in the form of entire vesicles (DiI-M@PDN NPs), rather than as PDN nucleus (Fig. S4). This phenomenon can also be verified in previous studies [[40], [41], [42]].

Fig. 3.

In vitro cellular uptake behavior and in vivo biodistribution. (A) Fluorescence images of cellular uptake of MPDN NPs by MC3T3-E1 cells. (B) Fluorescence images of cellular uptake of MPDN NPs by RAW264.7 cells. (C) Representative fluorescence images of blood samples collected from mice after administration of Ce6 or M@PD^Ce6 at preset time point (0, 0.5, 2, 4, 8, 12, 24 h). (D) The pharmacokinetics curves of plasma samples. (E–F) Fluorescence distribution and quantitative fluorescence analysis of lower limbs bone after 24 h post-injection. (G–H) Fluorescence distribution and quantitative fluorescence analysis of main organs after 24 h post-injection. n = 3.

Next, we investigated the effects of the hybrid cell membrane coating and nanoformulation strategy on the half-life and biodistribution of free drugs. The pharmacokinetic results in osteoporotic mice were obtained by measuring the plasma fluorescence intensity at different time points after intravenous injection of Ce6 and M@PD^Ce6. The fluorescence signal intensity in the collected mouse plasma samples gradually decreased over time (Fig. 3C). Compared to the free Ce6 group, the rate of decrease in fluorescence signal intensity over time was slower in the M@PD^Ce6 group. Meanwhile, quantitative fluorescence analysis revealed that the half-life of M@PD^Ce6 in BALB/c mice was 1.91 times longer than that of free Ce6 (0.576 h vs 1.099 h) (Fig. 3D). In previous studies, the PB nanocomplex after the modification of the hybrid membrane coating (HM@PB^Ce6 NPs) had a half-life of up to 1.2 h, approximately 3 times longer than that of PB^Ce6 (0.41 h) [22]. This indicates that the nanoformulation and cell membrane coating modification strategy can effectively prolong the free drug half-life, providing a good guarantee for the enrichment of MPDN NPs at osteoporotic sites.

Lower limb bones and main organs were collected for fluorescence imaging 24 h after injection of Ce6 and M@PD^Ce6. The results reflected that the fluorescence signal intensity in the lower limb bones of the M@PD^Ce6 group was obviously higher than that of the Ce6 group (Fig. 3E). Quantitative analysis showed that the fluorescence intensity in the M@PD^Ce6 group was 3.90 times higher than that of the Ce6 group (Fig. 3F). Fluorescence imaging of main organs also showed higher fluorescence intensity in the heart, liver, spleen, lungs, and kidneys of the M@PD^Ce6 group (Fig. 3G and H). We speculate that free Ce6 is mostly captured and cleared by the blood system before reaching the main organs and bones, so that the fluorescence intensity at above-mentioned sites of the Ce6 group is all lower than the M@PD^Ce6 group. Taken together, the hybrid cell membrane coating and nanoformulation strategy slow down the metabolic rate of free drugs in vivo, thereby enhancing drug enrichment in bone sites.

3.4. In vitro anti-oxidant, pro-proliferation and anti-apoptotic properties

Previous studies have confirmed that H₂O₂ causes a surge in ROS, triggering a series of functional disorders such as cell proliferation, apoptosis, and differentiation [43,44]. In this study, we verified that H₂O₂ increases ROS levels in MC3T3-E1 cells, inhibiting proliferation, triggering apoptosis, and impairing osteogenic differentiation. PB NPs in nanocomplexes can significantly scavenge ROS, thereby promoting proliferation, reducing apoptosis, and providing a foundation for osteogenic differentiation of MC3T3-E1 cells.

Based on previous studies, we used H₂O₂ (400 μM) to simulate the oxidative microenvironment of osteoporotic sites in vitro [19,43,44]. Firstly, the DCFH-DA probe was applied to detect ROS in MC3T3-E1 cells. The ROS fluorescence image shows that the green fluorescence was strongest in the Model and NGF groups, while the green fluorescence intensity in the MPD and MPDN groups both decreased significantly (Fig. 4A). Subsequently, we used flow cytometry to detect intracellular ROS levels quantitatively, and found that the addition of MPD and MPDN NPs effectively reversed the H₂O₂-induced ROS surge (Fig. 4B and C). DPPH serves as a stable free radical indicator for assessing the antioxidant capacity of substances. As shown in Fig. 4D, compared to the Model group, the color of the solution was lighter in the MPDN and MPD groups, and the absorbance at 517 nm in the MDPN group was significantly reduced to 50.7 % of the Model group. The DPPH radical scavenging experiment further confirmed the superior antioxidant capacity of MPDN NPs.

Fig. 4.

MPDN NPs scavenge ROS in vitro thereby promoting proliferation of MC3T3-E1 cells and inhibiting their apoptosis. (A) Representative fluorescence images of ROS in MC3T3-E1 cells. (B–C) Flow cytometry assay of ROS in MC3T3-E1 cells. (D) Anti-oxidant abilities of NGF, MPD, MPDN by DPPH assay. (E) Cell viability of the MC3T3-E1 cells incubated with MPDN and H₂O₂ for 24 h. (F–G) Representative fluorescence images and quantitative analysis of EdU positive MC3T3-E1 cells. (H–I) Representative fluorescence images and quantitative analysis of Ki67 in MC3T3-E1 cells. (J–K) Flow cytometry analysis of apoptosis in H₂O₂-treated MC3T3-E1 cells and quantitative analysis of the total apoptotic rate (early and late apoptotic cells in MC3T3-E1 cells. (L–M) Western blot assay of Bcl2 and Bax apoptosis-related proteins and quantitative analysis of the ratio of Bcl2/Bax. n = 3.

After assessing ROS scavenging properties of MPDN NPs, we further assessed the effects of MPDN NPs on the proliferation of MC3T3-E1 cells stimulated by H₂O₂ from three aspects: cell metabolism, DNA replication, and the proliferation-related protein marker Ki67. MTT assay showed significantly decreased cell viability in the Model group, but this situation was significantly improved after treatment with MPDN and MPD NPs, with cell viability almost approaching that of the Control group (Fig. 4E). The EdU method accurately reflects cell proliferation by directly detecting DNA synthesis. We observed EdU-labeled cells under a fluorescence microscope and found that there was almost no red fluorescence in the Model and NGF groups, while the red fluorescence intensity in the MPDN and MPD groups was close to that of the Control group (Fig. 4F). Quantitative analysis of the above fluorescence images showed that the proportion of EdU-positive cells in the Control group was 28.60 %, and the proportion of EdU-positive cells in the MPDN group (26.06 %) was 6.82 times that of the Model group (3.82 %) (Fig. 4G). Ki67 is a protein located in the cell nucleus, which is closely related to cell proliferation activity. Immunofluorescence staining showed the weak red fluorescence in the Model and NGF groups, while the red fluorescence in the MPDN and MPD groups was very strong, comparable to that in the Control group (Fig. 4H). Quantitative analysis also showed a consistent trend (Fig. 4I).

Moreover, we quantitatively detected apoptosis in MC3T3-E1 cells via Annexin V-FITC/PI kit. Flow cytometry showed that compared to the Model group (20.2 %), the total apoptosis rate in the MPDN group (16.8 %) was significantly reduced, indicating that MPDN inhibits the apoptosis process of MC3T3-E1 cells induced by ROS (Fig. 4J and K). Western blotting showed that MPDN NPs upregulated the expression of the anti-apoptotic protein Bcl2 and downregulated the pro-apoptotic protein Bax (Fig. 4L). And the ratio of Bcl2/Bax in the MPDN group was 4.5-fold higher than that in the Model group (Fig. 4M). This confirmed that MPDN NPs reduced apoptosis of MC3T3-E1 cells by enhancing Bcl2 expression and inhibiting Bax expression.

It is noteworthy that the role of ROS scavenging to promote proliferation and inhibit apoptosis is mainly played by PB NPs rather than NGF, which is consistent with the expected results. In summary, MPDN NPs can scavenge ROS in MC3T3-E1 cells, restore their impaired proliferation vitality and inhibit apoptosis, hereby providing a solid foundation for subsequent osteogenic differentiation mediated by NGF of MC3T3-E1 cells.

3.5. Pro-osteogenic differentiation properties in vitro oxidative microenvironment

The positive effect of NGF on bone regeneration under normal conditions has been confirmed. However, the role of NGF and MPDN NPs in osteogenic differentiation in the oxidative microenvironment of OP remains unclear.

We used H₂O₂ (400 μM) to simulate the oxidative microenvironment of osteoporotic sites in vitro and investigated the effect of MPDN NPs on the osteogenic differentiation ability of MC3T3-E1 cells. After co-incubation of NGF, MPD NPs and MPDN NPs with MC3T3-E1 cells for 2 h followed with 400 μM H₂O₂ addition. ALP staining and activity assays were performed 7 days later, and ARS staining was performed 21 days later. ALP staining showed that the ALP content in the Model group was lower than that in the Control group, confirming that oxidative stress had a significant inhibitory effect on the osteogenic differentiation. The ALP content in the NGF group was slightly higher than that in the Model group, indicating that the osteogenic differentiation-promoting effect of NGF was inhibited by oxidative stress. However, the ALP content in the MPDN group was much higher than that in the H₂O₂ group, suggesting that MPDN NPs can resist the inhibitory effect of oxidative stress on osteogenic differentiation (Fig. 5A). The ALP activity assay also showed a trend consistent with the ALP staining experiment (Fig. 5C). ARS staining can detect calcium nodules formed by osteoblasts in the late stage of osteogenic differentiation, namely bone mineralization [45]. The staining images and their quantitative analysis showed that compared with the Model group, the number and area of calcium nodules in the NGF group increased slightly, while the number and area of orange-red calcium nodules in the MPDN group were significantly larger, indicating a higher overall level of bone mineralization in the MPDN group, even comparable to the blank control group. This suggests that MPDN NPs can fully reverse the impaired osteogenic differentiation ability under oxidative stress (Fig. 5B–D).

Fig. 5.

MPDN NPs promote osteogenic differentiation of MC3T3-E1 cells in vitro oxidative microenvironment. (A–B) Photographs of ALP and ARS staining of MC3T3-E1 cells after 7 and 21 days of various treatments. (C–D) Quantitative analysis of ALP and ARS staining. (E–G) Representative fluorescence images of early (Runx2) and late (Col Ⅰ and OCN) osteogenesis-related proteins. (H–J) Quantitative fluorescence analysis of Runx2, Col Ⅰ and OCN. (K) Schematic diagram of pro-osteogenic differentiation process.

Runx2 is a key transcription factor for osteoblast differentiation and the earliest and most critical biomarker for bone formation [46]. 5 days after osteogenic induction, the expression of Runx2 in MC3T3-E1 cells of different groups was detected by immunofluorescence. The immunofluorescence images and quantitative analysis showed that compared with the Control group, the green fluorescence intensity in the Model group was significantly reduced, the green fluorescence intensity in the NGF group increased slightly, and the expression of green fluorescence intensity returned to normal levels after the addition of MPDN NPs (Fig. 5E–H). This indicates that the addition of H₂O₂ significantly reduced the expression of Runx2, but treatment with NGF, MPD, and MPDN NPs can reverse this trend to varying degrees. Collagen Ⅰ (Col Ⅰ) and osteocalcin (OCN) are markers of the late stage of osteogenic differentiation [45]. 21 days after osteogenic induction, the expression of Col Ⅰ and OCN in MC3T3-E1 cells of different groups was also detected by immunofluorescence, with results consistent with the expression trend of Runx2 (Fig. 5F, G, I, J).

Taken together, oxidative stress suppressed the expression of osteoblast-specific proteins such as ALP, Runx2, OCN, and Col Ⅰ in MC3T3-E1 cells. Additionally, oxidative stress inhibited the bone mineralization of MC3T3-E1 cells. And the osteogenic differentiation-promoting activity of NGF is also inhibited under oxidative stress, while MPDN NPs can rescue the osteogenic differentiation-promoting activity of NGF in the oxidative microenvironment depend on the ROS scavenging ability of PB NPs, thereby synergistically promoting long-term osteogenic differentiation of MC3T3-E1 cells and restoring the normal expression of osteogenic differentiation markers such as Runx2, Col Ⅰ and OCN (Fig. 5K).

The above results showed that MPDN NPs inhibited apoptosis of MC3T3-E1 cells and promoting their proliferation and osteogenic differentiation.

3.6. In vivo anti-osteoporotic effect

This study further explored the therapeutic potential of MPDN NPs for OP. We used four-month-old male SAMP6 (senescence-accelerated mouse prone 6) mice as an animal model of OP and age-matched SAMR1 (senescence-accelerated mouse resistant 1) mice as normal controls [25]. The specific experimental protocol is shown in Fig. 6A.

Fig. 6.

MPDN NPs prevents bone loss in vivo. (A) Flowchart of animal experiments. (B) Representative micro-CT images of distal femur. n = 5. (C–G) BV/TV, Tb. Sp, Tb.N, SMI, Tb.Th of mouse tibia obtained from micro-CT data. n = 5. (H) Photograph of the three-point bending test. (I–J) Maximum load and break load of tibia. (n = 4). (K–M) H&E, Masson and Trap staining of tibia tissue. n = 3.

We performed micro-CT scanning on the femurs to investigate the inhibitory effect of MPDN NPs on the OP. The three-dimensional reconstruction showed that the Model group had significant bone loss. The NGF and MPD groups had a slight therapeutic effect on bone loss, while the bone loss in the MPDN group was significantly reversed, approaching the Control group (Fig. 6B). We performed quantitative analysis on trabecular bone parameters such as bone volume/tissue volume (BV/TV), trabecular separation (Tb.Sp), trabecular number (Tb.N), and structure model index (SMI), which can reflect changes in bone mass [19]. The results showed that BV/TV and Tb.N values increased, while Tb. Sp and SMI values decreased (Fig. 6C–G). The quantitative analysis more clearly confirmed that compared with the Model group, the MPDN group showed significantly more trabecular bone, and bone mass was restored and maintained at normal levels, suggesting that MPDN NPs has a significant therapeutic effect on the OP. The three-point bending experiments of tibia also confirmed the significant anti-osteoporotic effect of MPDN NPs from the perspective of bone strength (Fig. 6H–J).

Histological staining further confirmed the reversal effect of MPDN NPs on OP. H&E (Fig. 6K) and Masson staining (Fig. 6L) showed that the trabecular bone below the epiphysis of the tibia in the Model group were significantly sparser than those in the Control group, while the trabecular bone in the MPDN group were significantly denser than those in the Model group. Osteoclasts are the cells responsible for bone resorption during bone remodeling and tartrate resistant acid phosphatase (Trap) is a specific enzyme for them [47]. Trap staining showed that the number of osteoclasts in the epiphysis of the tibia in the Model group increased apparently, while the MPDN group showed the significant disappearance of the mature osteoclasts (Fig. 6M). These results collectively confirm that MPDN NPs has significant advantages in the treatment of OP.

3.7. Effects and mechanisms of MPDN NPs in reversing OP

To further confirm the anti-osteoporotic effects and potential mechanisms of MPDN NPs, we employed data-independent acquisition (DIA) proteomics to identify differentially expressed proteins (DEPs) in bone tissues (femur and tibia) among the normal (Control) group, OP (Model) group and OP + MPDN NPs (MPDN) group. Principal component analysis (PCA) revealed a clear separation among the Control group, MPDN group and Model group, indicating that MPDN NPs effectively restored the osteoporotic state in model mice (Fig. 7A). Based on the significance criteria of DIA (FC > 1.2, p < 0.05), this study identified 1254 upregulated and 1194 downregulated proteins in the comparison between the Model and Control group, 931 upregulated and 1081 downregulated proteins in the comparison between the MPDN group and the Model group (Fig. 7B). Furthermore, 1103 proteins coincide in the DEPs of Control-vs-Model and Model-vs-MPDN (Fig. 7C).

Fig. 7.

MPDN NPs exert anti-osteoporotic effects by regulating the expression of multiple proteins in bones. (A) PCA plot of proteomic data of Control, Model, MPDN NPs groups. (B) Histogram of differentially expressed proteins. (C) Venn map of differentially expressed proteins. (D) KEGG pathway analysis of differentially expressed proteins (Control-vs-Model). (E) KEGG pathway analysis of differentially expressed proteins (Model-vs-MPDN). (F–I) GSEA plots of significant biological process associated with osteogenesis and ROS metabolism (Model-vs-MPDN). (J–K) GSEA plots of Chemokine and Hedgehog signaling pathway (Model-vs-MPDN). (L) Heat map of antioxidation-related proteins. (M) Heat map of cell growth-related proteins. (N) Heat map of osteogenesis-related proteins.

Next, we applied KEGG pathway annotation analysis to the DEPs and found that the DEPs between the Control and Model group were mainly enriched in multiple pathways including chemokine signaling, MTOR signaling, JAK-STAT signaling, Hedgehog signaling, IL-17 signaling (Fig. 7D), suggesting that the pathogenesis of OP may be related to these pathways. Furthermore, we found that the DEPs between the Model and MPDN group were also enriched in the above pathways, including chemokine and Hedgehog signaling pathway, indicating that MPDN NPs may exert anti-osteoporotic effects through these pathways (Fig. 7E). Based on this, we conducted Gene set enrichment analysis (GSEA) on the Model and MPDN group. The results showed that after MPDN NPs treatment, some biological processes were upregulated, including osteoblast differentiation, positive regulation of osteoblast differentiation and bone mineralization (Fig. 7F–H), while ROS metabolic processes were downregulated (Fig. 7I), which also confirmed the positive effects of MPDN NPs on osteoblast differentiation, bone mineralization and ROS scavenging. At the same time, GSEA also showed that the chemokine signaling pathway and Hedgehog signaling pathway were downregulated under the action of MPDN NPs (Fig. 7J and K), which overlaps with the KEGG enrichment analysis results of DEPs mentioned above, suggesting that MPDN NPs are likely to exert anti-osteoporotic effects through these pathways. The downregulation of the chemokine signaling pathway confirmed the anti-inflammatory properties of MPDN NPs, which are beneficial for the treatment of OP [48]. In addition, previous study demonstrated that the activity of Hedgehog signaling pathway gradually decreased as the osteoblasts continued to mature after birth [49,50]. Selective upregulating the activity of Hedgehog signaling causes an increase in bone formation, but the formed bone is very fragile, loose, and also significant bone resorption, eventually resulting in severe bone deficiency in mutant mice. The possible reason is that the activated Hedgehog signaling pathway in mature osteoblasts, closely related to the proliferation and differentiation of osteoclasts, and accelerates bone resorption by promoting the expression of RANKL. This also suggests that the complex regulatory mode of Hedgehog signaling pathway is bidirectional, and the specific effects and mechanism still need to be further verified.

Additionally, we identified certain DEPs related to oxidative stress, cell growth and osteogenic differentiation. Compared with the Model group, the expression of antioxidant proteins such as Foxo1, Foxo3, Sod1, Sod3, and Oxr1 was upregulated in the MPDN group (Fig. 7L). It has been reported that forkhead box O (Foxo) proteins are activated under oxidative stress conditions, promoting the expression of a series of antioxidant genes, to enhance cellular antioxidant capacity [51]. Superoxide dismutase (SOD) is an important antioxidant enzyme in organisms, and when its level increase can improve cellular antioxidant capacity [52]. Oxidation resistance gene 1 (Oxr1) is also a protein that protects cells from oxidative stress [53]. These results directly confirmed that the antioxidant activity in the bone of mice was significantly elevated after MDPN NPs treatment.

We also found that compared with the Model group, the expression of pro-proliferation protein (Mki67) and anti-apoptotic proteins (Triap1and Aatf) was upregulated, while the expression of pro-apoptotic proteins such as Aifm1 and Aifm2 was downregulated in the MPDN group (Fig. 7M). The levels of MKi67 (marker of proliferation Ki67) protein locating in the nucleus is closely related to cell proliferation activity [54]. TP53 Regulated Inhibitor of Apoptosis 1 (Triap1) protein is located in the mitochondrial intermembrane space and nucleoplasm, participating in negative regulation of apoptosis [55]. Apoptosis Antagonizing Transcription Factor (Aatf) has antioxidant, DNA damage, and anti-apoptotic effects [56]. Apoptosis-inducing factor, mitochondrial-associated 1/2 (Aifm1/Aifm2) mediate non-caspase-dependent apoptosis, causing chromatin condensation and large-scale DNA degradation [57]. The significant changes of the above proteins related to proliferation and apoptosis indicated that MPDN NPs improved the pro-proliferation and anti-apoptotic capacity in mice.

Furthermore, we found that compared with the Model group, Runx2, ALP1, Col1a1, Col1a2, Spp1 and Omd were upregulated in the MPDN group (Fig. 7N). As for Runx2, ALP 1, Col1a1, Col1a2, these proteins are all protein markers for osteogenesis, which were also proved to promote osteogenesis in the in vitro experiments of this study. Besides, it was reported that Secreted Phosphoprotein 1 (Spp1), also known as osteopontin (OPN) exerts crucial functions in bone matrix mineralization and calcium deposition [58], and osteomodulin (Omd) is a proteoglycan involved in bone mineralization [59]. The upregulation of these osteogenesis-related proteins in the bone of mouse strongly confirmed the effective anti-osteoporotic effect of MPDN NPs.

Based on the above proteomics and in vitro results, we evaluated the expression of Foxo1, Ki67, Bax, Bcl2, Runx2, Col Ⅰ and OCN proteins in the tibia tissue of mice. Consistent with proteomics results, Fig. 8A and B showed that MPDN NPs upregulated the expression of Foxo1 and Ki67 in the tibia tissue. Immunohistochemical staining showed that MPDN NPs promoted the expression of Bcl2 and inhibited the expression of Bax in the tibia tissue of OP mice (Fig. 8C and D), which is consistent with in vitro results. Moreover, immunofluorescence staining showed that MPDN NPs upregulated the level of Runx2 in the tibia tissue (Fig. 8E). And immunohistochemical staining showed that the MPDN group secreted more Col Ⅰ and OCN proteins compared with the Model group (Fig. 8F and G).

Fig. 8.

MPDN NPs regulate the expression of antioxidation, cell growth, and osteogenesis-related proteins in bones of OP mice. (A) Immunohistochemistry staining of Foxo1 in the tibia tissue. (B) Immunofluorescence staining for Ki67 in the tibia tissue. (C–D) Immunohistochemistry staining of Bcl2 and Bax in the tibia tissue. (E) Immunofluorescence staining for Runx2 in the tibia tissue. (F–G) Immunohistochemistry staining of Col Ⅰ and OCN in the tibia tissue. n = 3.

In conclusion, the proteomic results confirmed the anti-osteoporotic effect of MPDN NPs, and suggested that the anti-osteoporotic effect acts through reducing oxidative stress, regulating cell growth and promoting osteogenic differentiation, which was consistent with the results of in vitro experiments.

3.8. In vitro biocompatibility and in vivo safety

Good biocompatibility is a prerequisite for the application of nanodrugs. Firstly, the MTT assay was employed to evaluate the cellular compatibility of MPDN NPs. MC3T3-E1 cells or HUVECs were co-incubated with NGF, MPD, MPDN for 24 h. The cell viability of MC3T3-E1 and HUVECs in the MPDN group was over 80 % compared to the Control group (Fig. 9A), indicating that MPDN NPs has little cytotoxicity to the cells. Next, the blood compatibility of MPDN was further confirmed through hemolysis and coagulation experiments. Compared with the water group, the morphology of RBCs co-incubated with MPDN NPs (25, 50, 100 μg/mL) remained normal under the microscope (Fig. 9B), and the hemolysis rates of RBCs for 4 h were all below 5 % (Fig. 9C). The coagulation experiment showed that the coagulation rates of MPDN NPs were similar to those of the PBS group (Fig. 9D), collectively indicating that MPDN NPs does not cause significant coagulation or hemolysis, which has good blood compatibility.

Fig. 9.

MPDN NPs have excellent safety in vitro and vivo. (A) Cytotoxicity of the MC3T3-E1 cells and HUVEC incubated with NGF, MPD and MPDN NPs for 24 h. (B) Morphological images of RBCs incubated with water, PBS, MPDN NPs (25, 50, 100 μg/mL) for 4 h. (C) The images and hemolysis rate of RBCs incubated with MPDN NPs (25, 50, 100 μg/mL) for 4 h. (D) Platelet aggregation assay of NGF, MPD and MPDN NPs. (E) Representative images of the zebrafish incubation process during 96 h. n = 6. (F–G) The heart rate and body length of zebrafish. n = 6. (H) H&E staining of main organs from different groups. (I) Complete blood count of WBC, RBC, HGB, and PLT levels. (J) Hepatotoxicity and nephrotoxicity by measuring plasma levels of CREA, UREA, ALT and AST. n = 3.

Regarding in vivo biocompatibility, we evaluated it from three aspects: zebrafish incubation experiment, H&E staining of mice organs and hematological examination in mice. The zebrafish incubation experiment showed that all zebrafish in each group hatched successfully without adverse events such as malformation or death (Fig. 9E), and the heart rate and body length of zebrafish in the NGF, MPD, and MPDN groups were not significantly different from those in the Control group (Fig. 9F and G). In addition, main organs were collected from mice in each group in the in vivo experiment and subjected to H&E staining. The results demonstrated that long-term injection of NGF, MPD, and MPDN NPs had no significant impairment to the heart, liver, spleen, lungs, and kidneys (Fig. 9H). Hematological examination showed that the blood routine (Fig. 9I) and biochemical parameters (Fig. 9J) in each group were within the normal reference range for mice [60], further confirming that MPDN NPs have good in vivo safety. Collectively, MPDN NPs have excellent safety.

4. Conclusions

In summary, we have developed a novel nanodrug (MPDN NPs) based on NGF and PB NPs, and evaluated its therapeutic effects and mechanisms on OP. The prepared MPDN NPs can not only resist oxidative stress to promote osteogenic precursor cells proliferation and prevent their apoptosis but also exert osteogenic differentiation-promoting effects. More importantly, MPDN NPs can effectively treat OP with excellent safety. Overall, as a novel multifunctional nanodrug, MPDN NPs have the impressive therapeutic efficacy and biosafety, which possess important implications for OP and other oxidative stress-related diseases.

CRediT authorship contribution statement

Yuyi Tian: Writing – original draft, Formal analysis, Data curation. Yihan Lin: Visualization, Formal analysis. Hao Liu: Visualization, Funding acquisition. Xiaona He: Visualization. Shang Zhu: Supervision. Luhong Dai: Supervision. Yuqi Lu: Supervision. Lihong Liu: Writing – review & editing, Funding acquisition, Conceptualization. Bin Liu: Writing – review & editing, Investigation, Conceptualization.

Ethics approval and consent to participate

The animal use protocol has been approved by the Institutional Animal Care and Use Committee, The Second Xiangya Hospital, Central South University, China (No. 20241067), and none of the animal experiments violated the aims of the Declaration of Helsinki of the Ministry of Health.

Consent for publication

All authors are consent for publication.

A vailability of data and materials

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

Fundings

This work was supported by grants from the National Natural Science Foundation of China [grant number 82171581], National Undergraduate Innovation and Entrepreneurship Training Program [grant number S202410533139], Hunan Province Basic Education Teaching Reform Research Project [grant number Y2024898], Central South University Educational Reform Research Project [grant number 2023jy100], Hunan Provincial Natural Science Foundation of China [grant number 2022JJ40698].

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:

Data availability

Data will be made available on request.