Silica nanocarrier-mediated intracellular delivery of rapamycin promotes autophagy-mediated M2 macrophage polarization to regulate bone regeneration

Abstract

Targeting macrophages to regulate the immune microenvironment is a new strategy for bone regeneration with nano-drugs. Nano-drugs have achieved surprising anti-inflammatory and bone-regenerative effects, however, their underlying mechanisms in macrophages remain to be clarified. Macrophage polarization, immunomodulation, and osteogenesis are governed by autophagy. Rapamycin, an autophagy inducer, has shown promising results in bone regeneration, but high dose-mediated cytotoxicity and low bioavailability hinder its clinical application. This study aimed to develop rapamycin-loaded virus-like hollow silica nanoparticles (R@HSNs) which are easily phagocytosed by macrophages and translocated to lysosomes. R@HSNs induced macrophage autophagy, promoted M2 polarization, and alleviated the degree of M1 polarization as indicated by the downregulation of inflammatory factors IL-6, IL-1β, TNF-α, and iNOS, and upregulation of anti-inflammatory factors CD163, CD206, IL-1ra, IL-10, and TGF-β. These effects were nullified by cytochalasin B-induced inhibition of R@HSNs uptake in macrophages. The conditioned medium (CM) collected from R@HSNs-treated macrophages promoted osteogenic differentiation of mouse bone marrow mesenchymal stromal cells (mBMSCs). In a mouse calvaria defect model, free rapamycin treatment was inhibited, but R@HSNs robustly promoted bone defect healing. In conclusion, silica nanocarrier-mediated intracellular rapamycin delivery to macrophages effectively triggers autophagy-mediated M2 macrophage polarization, further enhancing bone regeneration by triggering osteogenic differentiation of mBMSCs.

Graphical abstract

1. Introduction

Recent research advances have shown the great application potential of nano drugs in bone tissue engineering [1]. Various nanomaterials have been designed to directly promote the osteogenic differentiation of precursor cells during bone defect repair [2,3]. However, recent studies have shown that the vast majority of nanomaterials are more likely to be phagocytosed by the mononuclear immune cells, including macrophages resulting in immunomodulation [4]. Endocytosis of nanomaterials could trigger immune cell differentiation into proinflammatory and anti-inflammatory phenotypes. Although proinflammatory immune cells are required in a very early stage of bone defect healing, the middle and later stages of bone defect repair are facilitated by anti-inflammatory immune cells [5]. Macrophages are the key immune cells that participate in the immunomodulation of the bone microenvironment (osteoimmunology) under both physiological and pathological conditions [6]. Macrophages differentiate into pro-inflammatory phenotype M1 and anti-inflammatory phenotype M2 to regulate osteoimmunology. Based on the chemical composition and physical structure, nanomaterials could trigger the M1 or M2 phenotype. Our previous studies have shown that a low or mild level of M1 macrophage polarization is required in the early stage of bone defect repair, but that a high level and long-term M1 macrophage polarization inhibits bone defect healing [7]. M2 macrophages release several anti-inflammatory factors including IL10, IL4, and TGF-β that promote osteogenic differentiation of precursor cells and bone regeneration [7]. Therefore, the development of nano-drugs that can be easily endocytosed by macrophages and promote M2 macrophage polarization will be a desirable approach to induce osteoimmunology regulation-mediated bone regeneration.

Autophagy is a lysosome-mediated catabolic process to maintain intracellular homeostasis via the disposal of damaged, non-functional, or unnecessary proteins and organelles [8], and is an essential cellular process and a primary mechanism for maintaining cellular homeostasis [9,10]. Our previous research illustrated that nanomaterial-based activation of macrophage autophagy promotes osteogenic differentiation of bone marrow mesenchymal stromal cells (BMSCs) [11]. Autophagy activation induces macrophage phenotype switching from proinflammatory M1 to anti-inflammatory M2 phenotype that facilitates bone repair and regeneration [5,12]. Rapamycin, an FDA-approved immunosuppressant drug, is currently available as a standard and effective drug for injury repair via inducing autophagy [13]. Rapamycin promotes autophagy by suppressing the mammalian target of the rapamycin (mTOR) pathway [[14], [15], [16]]. Rapamycin targets mTOR signaling in BMSCs to modulate bone formation [17] and in macrophages to inhibit the production of inflammatory cytokines IL-1β and IL-18 [13]. However, rapamycin is a poorly soluble drug, unstable in gastric acid, and has only a 14% oral bioavailability [18]. Moreover, to improve the bioavailability at the target site, high doses are required, which in turn brings systemic severe adverse effects, such as hypertriglyceridemia, hypertension, hypercholesterolemia, and increased creatinine [19]. Rapamycin inhibits proliferation and promotes apoptosis of BMSCs and osteoblasts [20,21]. To overcome these challenges, rapamycin is usually injected directly into the focal area, such as the articular cartilage. However, the lymphatic clearance system can still rapidly reduce the drug concentration to sub-therapeutic levels [22]. Subsequently, frequent rapamycin administration are rquired which may lead to pain, patient non-compliance, and potential infection [23]. Therefore, the key to the clinical transformation of rapamycin is to effectively reduce cytotoxicity and improve bioavailability.

Current research is focused on biomedical materials such as magnetic mesoporous silica nanospheres [24], poly (ethylene glycol)-shelled nanoparticles [25], and PLGA microparticles [23] as rapamycin carriers to reduce cytotoxicity and improve bioavailability. Although these strategies have acquired some good results for rapamycin application in vitro and in animal models, recent studies have shown that most of the nanomaterials entering the body are phagocytic and cleared by immune cells, especially macrophages [4]. Therefore, it is wise to utilize macrophage endocytosis of nanomaterials as a key strategy for focal site-injected nanocarrier-based intracellular delivery of rapamycin in macrophages to induce autophagy-mediated M2 macrophage polarization during bone defect repair. We presume that such a strategy may reduce the dose of rapamycin applied in vivo, and most nanocarriers will be endocytosed by macrophages, hence, the BMSCs and osteoblasts remain safe from the adverse effects of rapamycin.

Nanoparticles mimicking virus surface topology and roughness can easily enter immune cells [[26], [27], [28]], showing potential in drug delivery [29] and antibacterial applications [30]. In addition, Silica nanoparticles are biocompatible, safe, stable, and currently undergoing clinical trials for biomedical applications such as oral drug delivery, diagnosis, and photothermal ablation therapy [31]. Hence, this study aimed to develop virus-like surface-structured (virus-like) hollow silica nanoparticles (HSNs with a specific surface morphology) as a carrier of rapamycin, in order to stimulate preferential endocytosis and subsequent intracellular release of rapamycin by macrophages. Rapamycin-loaded hollow silica nanocarriers (R@HSNs) were synthesized and characterized, and their effects on macrophage uptake, polarization, and subsequent osteogenic induction in vitro and in the repair of bone defects in mice calvaria were studied (Fig. 1).

Fig. 1.

The schematic illustration of intracellular delivery of rapamycin via hollow silica nanocarriers for bone regeneration.

2. Materials and methods

2.1. Materials

Chemicals: Tetraethyl orthosilicate (TEOS), anhydrous ethanol, aqueous ammonia (NH₃H₂O), and sodium carbonate (Na₂CO₃) were purchased from Aladdin Holdings Group Co. Ltd. (Hangzhou, China). Resorcinol, formaldehyde, ascorbic acid, fluorescein isothiocyanate (FITC), 4ʹ,6-dimidazole-2-phenylindole (DAPI), β-glycerophosphate, dexamethasone, pentobarbital sodium, and lipopolysaccharide (LPS) were bought from Sigma (USA). Rapamycin was obtained from the Energy Chemical Company (China, E080201). PLGA-PEG-PLGA was purchased from MeloPEG (China). Cytochalasin B was obtained from Merck KGaA (250,233-5MGCN, Germany). The 3-methyladenine (3-MA) was purchased from Selleck (s2767, USA). Hoechst 33,342 staining kit was from Beyotime (China). Mitochondria tracker (Mito-Tracker), Lysosomes tracker (Lyso-Tracker), and Endoplasmic reticulum tracker (Er-Tracker) were purchased from Biyuntian (China). Celltracker CM-DiI was purchased from MKbio (USA).

Cells and culture medium: RAW264.7 murine macrophages were purchased from Shanghai Cell Bank (Shanghai, China). The cell culture medium Dulbecco's modified Eagle's medium (DMEM) and other culture reagents including 0.25% EDTA, fetal bovine serum (FBS), and Penicillin-Streptomycin were purchased from Gibco (USA).

Antibodies: Rabbit anti-Osteopontin antibody (OPN) (bs-0026R, Bioss, China), rabbit anti-Collagen I antibody (COL-I) (bs-10423R, Bioss, China), rabbit anti-Osteocalcin antibody (OCN) (DF12303, Affinity, USA), rabbit anti-Mannose receptor antibody (CD206) (ab64693, Abcam, UK), rabbit anti-IL1 beta antibody (IL-1β) (AF5103, Affinity, USA), rabbit anti-iNOS antibody (ab178945, Abcam, UK), rabbit anti-mTOR (phospho S2448) antibody (p-mTOR) (ab109268, Abcam, UK), rabbit anti-mTOR antibody (ab32028, Abcam, UK), rabbit anti-LC3B antibody (ab192890, Abcam, UK), rabbit anti-Beclin 1 antibody (ab210498, Abcam, UK), rabbit anti-beta Actin antibody (β-actin) (ab8227, Abcam, UK), and goat anti-rabbit IgG H&L (ab6721, Abcam, UK) were used.

2.2. HSNs fabrication and characterization

HSNs were prepared by the double-layer template method. The inner template controlled the hollow size and the outer template controlled the virus structure on the surface. HSNs were synthesized by a one-pot approach of phenolic resin/silica nanocomposites preparation through controlled hydrolysis of tetraethyl orthosilicate (TEOS) [32]. Briefly, 0.1 ​g resorcinol and 0.14 ​mL formaldehyde were added to 9.64 ​mL water and 70 ​mL ethanol solution and dispersed evenly. Then 3.36 ​mL of aqueous ammonia (25% NH₃H₂O) was added, stirred at 27 ​°C for 6 ​h, and 0.5 ​mL TEOS was added. After 6 ​min reaction, 0.2 ​g resorcinol and 0.28 ​mL formaldehyde were added for another 18 ​h. After centrifugation, the sediment was washed three times with ethanol/water. The resultant HSNs were obtained after calcination at 550 ​°C for 2 ​h. Transmission electron microscopy (TEM, FEI Tecnai G2, 200 ​kV), scanning electron microscopy (JSM-7800F, 15 ​kV), and Dynamic Light Scattering (DLS, Nano-ZS90) were performed to characterize the size and morphology of the HSNs. Brunner-Emmett-Teller (BET, TriStar II 3flex) was used to measure the pose size and surface area of the HSNs.

2.3. HSNs uptake by mBMSCs and macrophages co-culture system

FITC binding in HSNs: FITC (20 ​mg) and APTES (0.1 ​μL) were added to anhydrous ethanol (20 ​mL) and stirred for 24 ​h in the dark to obtain APTES-FITC. After that, 20 ​μg HSNs were added and stirred at 40 ​°C for 2 ​h in the dark. After centrifugation, the sediment was washed with PBS to obtain FITC-binding HSNs (HSNs-FITC).

Visualization of HSNs’ internalization in mBMSCs and macrophages: Cell tracker CM-DiI pre-stained mBMSCs were seeded in a glass disk at the density of 9000 ​cell/mL overnight and then 18,000 RAW264.7 macrophages were added per well. After culturing for 6 ​h, the culture medium was replaced with a fresh culture medium with HSNs-FITC at a concentration of 5 ​μg/mL and slowly shaken at 37 ​°C for 1 ​h. After incubation for 6 ​h, the cells were washed with PBS three times, then the nucleus was stained with Hoechst 33,342 (Blue) for 5 ​min. The fluorescence microscope (DM8, Leica) was used to acquire images. Image J software was used for fluorescence quantification.

2.4. Visualization of HSNs’ localization in macrophages

Macrophages were treated with 20 ​μg/mL HSNs-FITC. After incubation for 6 or 24 ​h, the cells were washed with PBS three times and stained by Mito-Tracker, Lyso-Tacker, and Er-Tracker respectively. After staining for 30 ​min, the cells were washed with PBS and then stained by Hoechst 33,342 for 10 ​min. The stained cells were imaged by confocal laser scanning microscopy (CLSM). The fluorescence colocalization was calculated using the Image J colocalization analysis plugin.

2.5. Drug loading and release

HSNs (5 ​mg) and rapamycin (2 ​mL, 1 ​mg/mL) were subsequently added to a 3 ​mL toluene solution. After ultrasonic dispersion, rotating shaker for 24 ​h, centrifugal separation, and washing, all supernatant was collected to measure the remaining uploaded rapamycin concentration by ultraviolet spectrophotometer (UV–vis, Mapada, UV-6300). The drug loading amount of rapamycin in R@HSNs was calculated according to the formula: Drug loading amount ​= ​W_RAPA/W_R@HSNs, where W_RAPA is the weight of rapamycin loaded into the R@HSNs nanoparticles and W_R@HSNs is the mass of R@HSNs. Rapamycin-loaded HSNs (R@HSNs) was placed in a dialysis bag in PBS containing 1% sodium dodecyl sulfate aqueous solution, in a sealed beaker, at 37 ​°C in a shaking water bath. The 200 ​μL of dialysate was taken at the set times (6, 12, 24, 48, 96, and 144h) to detect the concentration, and the same volume of fresh dialysate was added simultaneously. The concentration of rapamycin in the collected dialysate was measured by UV–vis.

2.6. Mice calvarial bone defect healing

All the animal experiments were approved by the Experimental Animal Ethics Committee of Guangdong Huawei Testing Co., Ltd. (Approval number: 20210603). All animal experiments complied with the ARRIVE guidelines and were carried out in accordance with the U.K. Animals (Scientific Procedures) Act 1986 and associated guidelines. The Specific-pathogen-free (SPF) C57BL/6 male mice (7 weeks old, 20–23 ​g body weight) were purchased from Zhuhai BesTest Bio-Tech Co. Ltd. The animals were kept in cages with free access to food and water and were subjected to a 12 ​h dark/12 ​h light cycle in an air-conditioned room (22.5 ​± ​0.5 ​°C, 50 ​± ​5% humidity). After 7 days of acclimatization, at the age of 8 weeks, mice were anesthetized by intraperitoneal injection of pentobarbital sodium. The skin was incised from the midline of the occipital to the frontal bone to remove the underlying periosteum. Full-thickness skull defects (2 ​mm in diameter) were created respectively in unsutured bilateral parietal bones using a slow machine and a 1.8 ​mm drill under saline irrigation. Thermosensitive hydrogel (Poly (lactide-co-glycolide)-b-Poly (ethylene glycol)-b-Poly (lactide-co-glycolide), (PLGA-PEG-PLGA)) (4 ​μL) containing 1 ​mg of sterilized R@HSNs (total load rapamycin 50 ​μg), rapamycin (50 ​μg), or HSNs (1 ​mg) were placed into the defect site as experimental groups, hydrogel group was used as control group. The skin was closed with 4–0 surgical sutures. The mice were euthanized by inhalation of isoflurane at 3 and 28 days post-op. The head was taken and fixed in 4% paraformaldehyde (PFA) for 24 ​h at room temperature. Then the skulls were scanned using micro-CT (Skyscan 1172, Bruker) with an X-ray voltage 60 ​kV, 10 ​μm pixel size, and a 0.5 ​mm aluminum filter. After scanning, the datasets were standardized and reconstructed by NRecon software (Bruker, Belgium), and the three-dimensional reconstruction was performed by SkyScan CTvox software (Bruker, Belgium). The newly formed bone parameters i.e., the bone volume/total volume (BV/TV) and trabecular thickness (Tb·Th) were analyzed using CTan software (Bruker, Belgium). After that, the skulls were demineralized in 10% EDTA solution for 10 days and dehydrated in gradient ethanol. Afterward, the samples were embedded in paraffin and sectioned (4 ​μm) for hematoxylin and eosin (H&E) staining and immunochemistry staining.

2.7. Cell culture

RAW264.7 murine macrophage cell line was maintained in culture medium (DMEM supplemented with 10% FBS and 1% penicillin/streptomycin) at 37 ​°C with a 5% CO₂ humidified atmosphere. Mouse BMSCs were isolated as described previously [33]. The isolated cells were cultured medium and incubated at 37 ​°C with 5% CO₂. The unattached hematopoietic cells were washed away with fresh medium. Attached cells were digested when reaching around 90% confluency, and passaged using 0.25% trypsin.

2.8. CCK-8 ​cell proliferation assay

The CCK8 (BestBio, China) assay was performed to evaluate the cytotoxicity of R@HSNs. Similarly, to the in vivo study, this experiment contained 4 groups: Control, R@HSNs, HSNs, and rapamycin. RAW 264.7 and mBMSCs (2000 ​cells/well) were seeded into 96-well plates. The control group was treated with PBS. R@HSNs and HSNs group were treated with 100, 200, 400, 800, or 1600 ​ng/mL of R@HSNs or HSNs. The rapamycin group was treated with 5, 10, 20, 40, or 80 ​ng/mL of rapamycin dissolved in a culture medium to match the concentrations administered with the R@HSNs, since the 100, 200, 400, 800, and 1600 ​ng/mL R@HSNs group contained 5, 10, 20, 40, and 80 ​ng/mL of rapamycin, respectively. After 1 and 7 days of incubation, the culture medium was changed to a fresh medium containing CCK-8 solution (0.5 ​mg/mL) and cultured for 2 ​h at 37 ​°C. The absorbance of the samples was then read at 450 ​nm by a microplate reader (Molecular Devices, LLC, San Jose, CA).

2.9. RT-qPCR analysis

Total RNA was extracted from RAW264.7 and mBMSCs cells using an RNA extraction column kit (YEASEN, China). Subsequently, complementary DNA (cDNA) was synthesized from 1 ​μg of total RNA using a cDNA synthesis kit (YEASEN, China). Quantitative RT-PCR was performed using an SYBR Green Master mix (Takala, Japan) on LightCycler 480 (Roche, USA). The primers used are shown in Table 1. All experiments followed the MIQE guidelines and were replicated three times. The β-actin gene was used to normalize the gene expression of ATG5/7, LC3A/B, CD163, CD206, IL-1ra, IL-10, TGF-β, IL-6, IL-1β, TNF-α, and iNOS, whereas the GAPDH gene was used to normalize target gene expression of OPN, OCN, and ALP. The values of the experimental groups were compared to the normalized control group.

Table 1.

The primer sequences used for qRT-PCR.

Gene	Forward (5′-3′)	Reverse (5′-3′)
ATG5	GATGCGGTTGAGGCTCAC	CTGTCATTCTGCAGTCCCATC
ATG7	AGCCTGTTCACCCAAAGTTC	CATGTCCCAGATCTCAGCAG
LC3A	ACAGCATGGTGAGCGTCTC	AGGTTTCTTGGGAGGCGTAG
LC3B	GATAATCAGACGGCGCTTGC	TCTCACTCTCGTACACTTCGG
Beclin-1	AATCTAAGGAGTTGCCGTTATAC	CCAGTGTCTTCAATCTTGCC
CD163	ACATCATGGCACAGGTCAC	TGAGGAAACTGTAAGTCGCTG
CD206	AGACGAAATCCCTGCTACTG	CACCCATTCGAAGGCATTC
IL-1ra	CTCCAGCTGGAGGAAGTTAAC	CTGACTCAAAGCTGGTGGTG
IL-10	GAGAAGCATGGCCCAGAAATC	GAGAAATCGATGACAGCGCC
TGF-β	CAGTACAGCAAGGTCCTTGC	ACGTAGTAGACGATGGGCAG
IL-6	ATAGTCCTTCCTACCCCAATTTCC	GATGAATTGGATGGTCTTGGTCC
IL-1β	TGGAGAGTGTGGATCCCAAG	GGTGCTGATGTACCAGTTGG
TNF-α	CTGAACTTCGGGGTGATCGG	GGCTTGTCACTCGAATTTTGAGA
iNOS	CAGAAGTGCAAAGTCTCAGACAT	GTCATCTTGTATTGTTGGGCT
OPN	ATCTCACCATTCGGATGAGTCT	TGTAGGGACGATTGGAGTGAAA
OCN	CCGGGAGCAGTGTGAGCTTA	AGGCGGTCTTCAAGCCATACT
ALP	CCAACTCTTTTGTGCCAGAGA	GGCTACATTGGTGTTGAGCTTTT
GAPDH	TGACCACAGTCCATGCCATC	GACGGACACATTGGGGGTAG
β-actin	CATACCCAAGAAGGAAGGCTGG	GCTATGTTGCTCTAGACTTCGAGC

2.10. Western blot assay

Total protein was extracted from RAW264.7 ​cells or mBMSCs using RIPA lysis buffer (RIPA: PMSF ​= ​100:1, Biyuntian, China). Protein concentration was measured using the BCA Protein Assay Kit (Biyuntian, China). Western blot was performed as previously described [34]. Primary antibodies (rabbit-originated) against Beclin-1 (1:1000), LC3A/B (1:1000), p-mTOR/mTOR (1:1000), CD206 (1:1000), iNOS (1:1000), IL-1β (1:1000), COL-1 (1:1000), OCN (1:1000) were used, GAPDH (1:1000) or β-actin (1:1000) were used as loading controls. The anti-rabbit IgG (1:5000, Abcam) was used as the secondary antibody. The membrane was developed for 1 ​min with an enhanced chemiluminescence developer (Biyuntian, China) and visualized with Bio-Rad ChemiDoc XRS (Bio-Rad, USA). The relative intensity of protein bands was quantified using Image J software (National Institutes of Health, USA).

2.11. Enzyme-linked immunosorbent assay (ELISA)

The supernatant was collected from RAW264.7 ​cells cultured in the three groups (Control/R@HSNS/rapamycin) for 24 ​h and centrifuged. The cytokine level of interleukin-10 (IL-10) was quantified using an ELISA kit (Cloud-Clone Corp.) according to the manufacturer's instructions. The IL-10 concentration was determined by correlation with a standard curve and the results were expressed as the amount (pg) of IL-10/mL supernatant.

2.12. Immunofluorescence staining

Immunofluorescence staining for CD206 was used to detect M2 macrophage polarization. RAW264.7 ​cells (2 ​× ​10⁵ ​cells/well) were seeded in a glass disk and cultured for 24 ​h. Subsequently, the culture medium was respectively replaced with a fresh medium containing PBS (Control), R@HSNs (400 ​ng/mL containing 20 ​ng/mL rapamycin), and rapamycin (20 ​ng/mL). After incubation for 24 ​h, the RAW264.7 ​cells were fixed with 4% PFA for 10 ​min at room temperature, blocked with blocking buffer for 30 ​min, and incubated with a rabbit anti-CD206 (1:200) antibody overnight at 4 ​°C. The next day, after washing five times, the cells were stained with a goat anti-rabbit Alexa Fluor 594-conjugated antibody (1:200, Abcam, USA) for 1 ​h at room temperature in the darkroom. After washing five times, DAPI was used to stain the nucleus for 5 ​min. The fluorescence microscope (DM8, Leica) was used to acquire images.

2.13. Alkaline phosphatase staining and alizarin red staining (ARS)

After osteogenic differentiation of mBMSCs for 5 days, mBMSCs were fixed with 4% PFA for 10 ​min and stained with the Alkaline Phosphatase Staining Kit (Biyuntian, China). After osteogenic differentiation of mBMSCs for 10 days, mBMSCs were fixed in 4% paraformaldehyde and stained with an ARS staining solution ((Biyuntian, China). Image J software was used for the quantitative analysis of ARS staining.

2.14. Analysis of macrophage phagocytosis of R@HSNs

Scanning electron microscopy analysis: RAW264.7 ​cells were fixed with 4% paraformaldehyde overnight, then rinsed two times with PBS. the samples were sequentially dehydrated with graded ethanol (50%, 70%, 80%, 90%, 95%, and 100%, 10–15 ​min/each). After coating with gold-palladium sputtering, the samples were then examined under scanning electron microscopy (S–3400 ​N, Hitachi).

TEM analysis: RAW264.7 ​cells were fixed with 2.5% glutaraldehyde overnight, then rinsed two times with 0.1 ​M cacodylate buffer and postfixed with 1% osmium tetraoxide for 1 ​h at 4 ​°C. Subsequently, the cells were washed with distilled water and then stained with 1% uranyl acetate for 1 ​h. The cells were sequentially dehydrated with graded ethanol (50%, 70%, 80%, 90%, 95%, and 100%). The cells were then treated with a series of epoxy resin concentrations in ethanol (33.3%, 50%, 66.7%, and 100%). The cells were soaked in resin at 70 ​°C for 24 ​h and then sectioned with an ultramicrotome. The sections were analyzed by TEM (Hitachi 7800).

Inductively coupled plasma source mass spectrometer (ICP-MS) assay: The RAW264.7 ​cells were seeded in a six-well plate and divided into two groups. After adherence, the first group was pretreated with cytochalasin B (Cyto-B, 10 ​μM) for 30 ​min to inhibit endocytosis, and then both two groups were treated with R@HSNs (400 ​ng/mL) for 6 ​h, then washed three times with PBS. The samples were centrifuged at 1000 ​g for 10 ​min, the supernatant was discarded, and 1 ​mL of 10% Na₂CO₃ solution was added to each tube. The mixture was dissolved at 100 ​°C for 1 ​h, the dissolved liquid was then transferred to a 15 ​mL centrifuge tube, and pure water was added to maintain the final volume to 10 ​mL. The concentration of silicon ion was analyzed by ICP-MS (Agilent 720 ​ES, USA).

2.15. Indirect co-culture of R@HSNs-treated macrophages and mBMSCs

Macrophages were induced to the pro-inflammatory M1 phenotype by LPS (100 ​ng/mL) treatment for 24 ​h. The LPS-containing medium was removed, and the cells were washed three times with PBS. Subsequently, the cells were treated with the complete medium containing PBS (control group), R@HSNs (400 ​ng/mL), and rapamycin (20 ​ng/mL). After 24 ​h of incubation, the culture medium was removed and the cells were washed twice with PBS, after which fresh serum-free DMEM was added to the culture. The conditioned medium (CM) derived from macrophages was collected after 6 ​h of culture incubation and centrifuged at 1000 ​g for 10 ​min. The supernatants were stored at −80 ​°C for follow-up experiments.

To investigate the effects of PBS, R@HSNs, and rapamycin-treated macrophages on osteogenesis, the CM was mixed with an osteogenic medium (DMEM, 20% FBS, 2% P/S, 20 ​mM β-glycerophosphate, 100 ​mM ascorbic acid, and 200 ​nM dexamethasone) at a 1:1 ratio, and the mixture was used to culture mBMSCs. After differentiation for 5, 7, and 10 days, mBMSCs were harvested for subsequent experiments. Three types of osteogenic induction conditioned medium were collected: Control-CM, R@HSNs-CM, and RAPA-CM.

2.16. Statistical analysis

Statistical analysis of the comparison between the two groups was carried out by Student's t-tests. Statistical comparison between multiple groups was performed by one-way analysis of variance followed by the Student/Newman/Keuls test at α ​= ​0.05 using Prism 7.0 (GraphPad Software, CA), and considered significantly different when p ​< ​0.05. Data are presented as mean ​± ​Standard deviation (SD) or standard error of the mean (SEM).

3. Results

3.1. HSNs preparation, characterization, and cellular internalization in the coculture system

To develop a nanomaterial that becomes more easily engulfed by macrophages, we designed a HSNs with special virus-like surface structure. The TEM images indicated that the HSNs have an outer shell and inner shell diameters of 149 ​± ​38 ​nm and 99 ​± ​17 ​nm respectively (Fig. 2A and Fig. S1A&B). The scanning electron microscopy image revealed the virus-like bulge on the surface of the HSNs (Fig. 2B). The DLS analysis of the HSNs showed an average diameter of 172 ​nm (Fig. 2C). BET assay showed the specific surface area to be 132.493 ​m²/g (Fig. 2D and Fig. S1C). Fluorescence images showed that the HSNs were more readily phagocyted by macrophages compared with mBMSCs in the co-culture system (Fig. 2E). Quantitative analysis of fluorescence intensity showed that the grey value of the green fluorescence in macrophages was 5.4 times higher compared to that in mBMSCs (Fig. S2) (∗P ​< ​0.05).

Fig. 2.

Characterization of rapamycin-loaded hollow silica nanoparticles (R@HSNs). (A) Transmission electron microscope (TEM) image, (B) Scan electron microscope image, (C) dynamic light scattering analysis (DLS), and (D) N2 adsorption-desorption isotherms of HSNs. (E) Fluorescence images illustrating HSNs uptake by macrophages and mBMSCs in a co-culture treated with FITC-HSNs for 24 ​h, blue: Hoechst, nuclei; red: pre-stained mBMSCs (cell tracker CM-Dil); green: FITC-HSNs nanoparticles. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

3.2. HSNs mainly localized in lysosomes of macrophages

LSCM images showed the colocalization of HSNs (Green) with lysosomes at 6 or 24 ​h, but not with the mitochondria and endoplasmic reticulum of macrophages (Fig. 3A–C). According to the profile analysis, we observed a prominent overlap between the HSNs-FITC green signals and Lyso-Tracker red signal, but not with the Mito-tracker and Er-tracker red signal (Fig. 3D), suggesting that HSNs localized in the lysosomes of RAW264.7 ​cells, but not in mitochondria and endoplasmic reticulum.

Fig. 3.

Cellular distribution and trafficking of HSNs in macrophages. CLSM images after internalization show the colocalization of HSNs (Green) with different organelles (Red, A: lysosomes, B: endoplasmic reticulum, and C: mitochondria). (D) Fluorescence signals analysis based on the yellow light in figures A, B, and C. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

3.3. Free rapamycin exerts cytotoxicity to mBMSCs and inhibits osteogenic differentiation and bone regeneration

Free rapamycin treatment inhibited bone regeneration in mice with calvarial bone defect (Fig. S3A), as indicated by the decreased BV/TV (Fig. S3B) and Tb·Th (Fig. S3C) compared with the control group. Moreover, rapamycin significantly inhibited mBMSCs proliferation even at a low concentration of 5 ​ng/mL (Figs. S3D and S3E). Rapamycin (20 ​ng/mL) treatment reduced the ALP production and matrix mineralization during the osteogenic differentiation of mBMSCs (Figs. S3F–I). These results indicate the inhibiting effect of free rapamycin on bone regeneration.

3.4. R@HSNs preparation, characterization, and cytotoxicity

The rapamycin loading amount in R@HSNs was determined by UV–vis absorbance at 276 ​nm and calculated to be 5.0 ​μg rapamycin/mg R@HSNs (Fig. S4), which indicates the concentration of rapamycin loaded in 100, 200, 400, 800, and 1600 ​ng/mL R@HSNs is 5, 10, 20, 40, and 80 ​ng/mL, respectively. The drug release profile of R@HSNs displayed a continuous release for 96 ​h in PBS (Containing 1% SDS) (Fig. S5A). R@HSNs did not exert cytotoxicity toward mBMSCs at a dose range of 100–1600 ​ng/mL (Figs. S5B and S5C). However, direct treatment of free rapamycin equivalent to the amount loaded in R@HSNs exerted cytotoxicity towards mBMSCs at day 1 and 7 of the culture (Figs. S5D and S5E). Both R@HSNs (400 ​ng/mL) and free rapamycin (20 ​ng/mL) did not exert cytotoxicity towards RAW264.7 ​cells (Fig. S5F). These results indicated that R@HSN can reduce the cytotoxicity of rapamycin towards mBMSCs.

3.5. R@HSNs degraded within lysosomes and induced macrophage autophagy

R@HSNs treatment for 24 ​h upregulated mRNA expression of autophagy markers ATG5, ATG7, and LC3A in macrophages compared with the control group and rapamycin group (Fig. 4A–C). R@HSNs upregulated mRNA expression of ATG7 at the low concentration (100 ​ng/mL R@HSNs containing 5 ​ng/mL rapamycin), but free rapamycin at the same concentration failed to induce ATG7 expression in macrophages (Figs. S6A and S6B). The highest ATG7 expression occurred at the concentration of 20 ​ng/mL rapamycin and the 400 ​ng/mL R@HSNs (containing 20 ​ng/mL rapamycin) group (Fig. S6A), and concentration was therefore selected for the next experiments. HSNs itself did not affect the expression of autophagy markers in macrophages (Figs. S7A and S7B). The protein levels of Beclin-1 and LC3-II/I were increased and p-mTOR/mTOR were decreased in R@HSNs treated macrophages compared to those of control and rapamycin groups (Fig. 4D–G). The TEM images indicated R@HSNs within Raw264.7 ​cells at 24 ​h treatment (Fig. 4H and I). Meanwhile, R@HSNs were taken up into lysosomes and degraded (Fig. 4J). In addition, the process of autophagosome formation (precursors, engulfment, and autophagosomes) was observed in TEM images (Fig. 4k). These results indicated a superior autophagy-inducing potential of R@HSNs compared with the equivalent concentration of free rapamycin. These results indicated that R@HSNs are taken up by lysosomes causing nanoparticle degradation and intracellular rapamycin release that effectively induces macrophage autophagy.

Fig. 4.

R@HSNs induced autophagy in macrophages. mRNA expression pattern of autophagy-related markers ATG5 (A), ATG7 (B), LC3A (C), and (D) representative western blot images of Beclin-1, LC3-II/I, and p-mTOR/mTOR of macrophages with PBS (Control group), R@HSNs and rapamycin treated for 24 ​h. Relative quantification of Beclin (E), LC3-II/I (F), and p-mTOR/mTOR (G) from Western blot images. (H, I) TEM images of macrophages with PBS and R@HSNs treated for 24 ​h. (J) The internalization and degradation of R@HSNs in macrophages. (K) Induction of autophagosome in macrophages by rapamycin released from degraded R@HSNs. Data are mean ​± ​SD or SEM, n ​= ​3, statistical significance between the groups, ∗∗P ​< ​0.01 and ∗∗∗P ​< ​0.001; ^#compared to the control group, ^###P ​< ​0.001. (H, I) TEM images of macrophages with PBS and R@HSNs treated for 24 ​h. (J) The internalization and degradation of R@HSNs in macrophages.

3.6. R@HSNs prompted M0 to M2 macrophage polarization

R@HSNs treatment upregulated mRNA expression of M2 markers CDl63 and CD206 and anti-inflammatory cytokines IL-10, and TGF-β compared to control and rapamycin groups (Fig. 5A–D). Similarly, the mRNA expression of anti-inflammatory IL-ra was higher in R@HSNs treated macrophages compared with the control group but not with the free rapamycin group (Fig. 5E). R@HSNs treatment remarkably promoted IL-10 protein secretion by macrophages compared with the control and free rapamycin groups (Fig. 5F). Immunofluorescence staining confirmed the higher expression of CD206 in R@HSNs treated macrophages compared with the control and free rapamycin groups (Fig. 5G and S8). Scanning electron microscopy images revealed the wider elongation and spindle-shaped (M2 macrophage-like [35]) morphology of R@HSNs treated macrophages (Fig. 5H), while the morphology of macrophages in the control group and free rapamycin-treated group was spherical. These results indicate the macrophage M0 to M2 polarization potential of R@HSNs. However, HSNs did not affect the expression of M2 markers (Figs. S7C and S7D). These results indicate that the R@HSNs-mediated intracellular release of rapamycin is responsible for the M0-M2 polarization of macrophages.

Fig. 5.

R@HSNs promoted macrophage M2 polarization and alleviated the degree of M1 polarization. mRNA expression patterns of M2 macrophage markers CD163 (A), CD206 (B), IL-10 (C), TGF-β (D), and IL-1ra (E). (F) Protein level expression of IL-10 analyzed by ELISA. (G) Immunofluorescence images for CD206 expression. (H) Representative scanning electron microscopy images of RAW cells treated with PBS (Control group), R@HSNs (400 ​ng/mL), and rapamycin (20 ​ng/mL) for 24 ​h. Expression patterns of M1 macrophage markers iNOS (I), IL-1β (J), IL-6 (K), TNF-α (L), and (M) IL-10. Data are mean ​± ​SD or SEM, n ​= ​3, statistical significance between the groups, ∗P ​< ​0.05, ∗∗P ​< ​0.01, and ∗∗∗P ​< ​0.001; ^#compared to the control group, ^#P ​< ​0.05. RAPA: free rapamycin.

3.7. R@HSNs mitigated LPS-induced M1 macrophage phenotype

We further tested the effect of R@HSNs treatment in LPS-induced M1 macrophages. LPS treatment induced the expression of M1 macrophage markers iNOS, IL-1β, IL-6, TNF-α (Fig. 5I–L), and, M2 marker IL-10 (Fig. 5M). R@HSNs showed a prominent effect in mitigating LPS-induced expression of M1 macrophage markers iNOS, IL-1β, IL-6, and TNF-α compared with the free rapamycin (Fig. 5I–L). Simultaneously, R@HSNs upregulated the mRNA expression of M2 maker IL-10 (Fig. 5M). These results indicate that R@HSNs not only promote M0 to M2 macrophage polarization but also alleviate the M1 polarization of macrophages.

3.8. Macrophage endocytosis of R@HSNs is crucial to induce autophagy and M2 macrophage polarization

Macrophage endocytosis inhibitor cytochalasin-B treatment did not affect macrophage autophagy and M2 macrophage polarization (Fig. S9). Cytochalasin-B treatment inhibited macrophage endocytosis of R@HSNs by 32% as indicated by the intracellular silica concentration (Fig. 6A). Inhibition of macrophage endocytosis of R@HSNs reduced the mRNA expression of autophagy markers LC3A, LC3B, ATG5, and ATG7, as well as M2 macrophage markers IL-10 and CD206 (Fig. 6B–G). Compared with untreated macrophages, LC3A was not increased in macrophages treated with cytochalasin-B or R@HSNs (Fig. S10A) alone, but ATG5 was still upregulated (Fig. S10B), while no difference was detected in the expression of CD206 and IL-10 (Figs. S10C–D). Western blot analysis further confirmed that inhibition of macrophage endocytosis inhibited the R@HSNs-induced expression of LCII/I, Beclin, and CD206 as well as induced the R@HSNs-suppressed mTOR expression (Fig. 6H–L). These results indicate that macrophage endocytosis of R@HSNs is necessary for M0 to M2 polarization.

Fig. 6.

Endocytosis of R@HSNs by macrophages was mandatory to induce autophagy-mediated anti-inflammatory responses. (A) Intracellular silicon concentration after 6 ​h treatment of R@HSNs or Cyto-B ​+ ​R@HSNs. mRNA expression patterns of autophagy-related factors LC3A (B), LC3B (C), ATG5 (D), and ATG7 (E). mRNAs expression patterns of M2 macrophage markers IL-10 (F) and CD206 (G). (H) Representative Western blots of autophagy markers LC3-I, LC3-II, Beclin-1, and p-mTOR, and M2 macrophage marker CD206. Relative quantification of LC3-II/I (I), Beclin-1 (J), p-mTOR (K), and CD206 (L) from Western blots images. Data are mean ​± ​SD or SEM, n ​= ​3, statistical significance between the groups, ∗P ​< ​0.05, ∗∗P ​< ​0.01, and ∗∗∗P ​< ​0.001Cyto-B: cytochalasin B (10 ​μM). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

3.9. R@HSNs prompted M0 to M2 macrophage polarization and mitigated LPS-induced M1 macrophage phenotype by inducing autophagy

Blocking autophagy showed an inhibitory trend in the R@HSNs-induced mRNA expression of autophagy marker ATG5 but it was not statistically significant (Fig. 7A). Inhibition of autophagy mitigated the R@HSNs-induced expression of autophagy marker ATG7 and M2 macrophage markers CD163 and IL-10 (Fig. 7B–D). Similarly, inhibition of autophagy mitigated the R@HSNs-induced protein level expression of M2 macrophage markers Beclin-1, LCII/I, and CD206 (Fig. 7E–H). Then, LPS was used to induce the polarization of macrophages towards M1. Compared with the control group, the autophagy gene Beclin-1 showed no statistical difference, ATG7 expression was down-regulated, while LC3A expression was up-regulated in M1 macrophages (Figs. S11A–C). The expressions of M1 macrophage markers (iNOS and IL-6) were significantly increased (Figs. S11D–E), while the expression of M2 maker (TGF-β) was decreased (Fig. S11F). We further tested the role of autophagy in R@HSNs’ reduced LPS-induced M1 macrophage phenotype. R@HSNs treatment induced mRNA expression of autophagy markers ATG5 and ATG7 in M1 macrophages and autophagy inhibition reversed this effect (Fig. 7I & J). Furthermore, R@HSNs treatment inhibited the mRNA expression of proinflammatory cytokines IL-1β and IL-6 in M1 macrophages and autophagy inhibition reversed this effect (Fig. 7K and L). However, R@HSNs treatment or autophagy inhibition did not alter mRNA expression of anti-inflammatory cytokines CD206 and IL-10 in M1 macrophages (Figs. S12A–B). Similarly, R@HSNs treatment induced the protein expression of Beclin-1 and inhibited the protein level expression of iNOS and IL-1β in M1 macrophages and autophagy inhibition reversed these effects of R@HSNs (Fig. 7M–P). These results indicate that autophagy in R@HSNs promoted M0 to M2 macrophage polarization and R@HSNs-mitigated LPS-induced M1 macrophage phenotype.

Fig. 7.

Autophagy inhibition nullified the R@HSNs-induced M2 macrophage polarization or the R@HSNs-inhibited M1 macrophage polarization. mRNA expression patterns of autophagy markers ATG5 (A) and ATG7 (B) and M2 macrophage markers CD163 (C) and IL-10 (D). (E) Representative Western blots of autophagy markers Beclin-1, LC3-I, and LC3-II and M2 macrophage marker CD206. Relative quantification of Beclin (F), LC3-II/I (G), and CD206 (H) from Western blot images. mRNA expression patterns of autophagy markers ATG5 (I) and ATG7 (J), and M1 macrophage markers IL-1β (K) and IL-6 (L). (M) Representative Western blots of autophagy markers Beclin-1 and M1 macrophage markers iNOS and IL-1β. Relative quantification of Beclin-1 (N), iNOS (O), and IL-1β (P) from Western blot images. Data are mean ​± ​SD or SEM, n ​= ​3, statistical significance between the groups, ∗P ​< ​0.05, ∗∗P ​< ​0.01, and ∗∗∗P ​< ​0.001; compared to the control group, ^#P ​< ​0.05 and ^###P ​< ​0.001.3-MA: 3-Methyladenine.

3.10. R@HSNs-mediated macrophage immunomodulation promoted osteogenic differentiation of mBMSCs

Free rapamycin-treated macrophage-CM did not induce the osteogenic differentiation of mBMSCs compared with the control macrophage-CM (Fig. 8A–J). However, R@HSNs-treated macrophage-CM robustly promoted mRNA expression of osteogenic differentiation markers ALP, OPN, and OCN as well as protein level expression of Col1α1, OCN, and ALP in mBMSCs compared with the control macrophage-CM and free rapamycin-treated macrophage-CM (Fig. 8A–H). Furthermore, R@HSNs-treated macrophage-CM induced the matrix mineralization in mBMSCs culture compared with the control macrophage-CM and free rapamycin-treated macrophage-CM (Fig. 8I & J). These results indicate that R@HSNs-mediated macrophage immunomodulation promoted osteogenic differentiation of mBMSCs.

Fig. 8.

R@HSNs-treated macrophage-conditioned medium (CM) promoted osteogenic differentiation of mBMSCs. mRNA expression patterns of osteogenic markers ALP (A), OPN (B), and OCN (C) in mBMSCs. Representative Western blots of COL-Ⅰ and OCN (D). Relative quantification of COL-I (E) and OCN (F) from western blot images. (G) ALP staining and (H) quantification of ALP staining in mBMSCs. (I) Alizarin Red staining and (J) quantification of alizarin red staining in mBMSCs. Data are mean ​± ​SD or SEM, n ​= ​3, statistical significance between the groups, ∗P ​< ​0.05, ∗∗P ​< ​0.01, and ∗∗∗P ​< ​0.001; compared to the control group, ^#P ​< ​0.05. RAPA: free rapamycin, ARS: alizarin red staining. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

3.11. In vivo evaluation of bone defect repair

Upregulated autophagy marker Beclin-1 was detected in the group with R@HSNs treatment for 3 days compared with that in the control or RAPA treatment group (Fig. S13A). Correspondingly, compared with the control and RAPA treatment groups, the expression of the M1 marker (iNOS) was the lowest in the R@HSNs treatment group (Fig. S13C). In contrast, the expression of the M2 marker CD206 was the highest (Fig. S13C) in the R@HSNs treatment group. Local application of HSNs and free rapamycin in calvaria bone defects did not promote bone regeneration (Fig. S14). However, local delivery of R@HSNs robustly promoted bone regeneration in calvaria bone defects as indicated by a higher BV/TV, BS/TV, and Tb·Th (Fig. 9A–D). H&E staining of histological tissue sections also indicated a higher area of newly formed bone in the R@HSNs-treated group compared with the control, and free rapamycin group (Fig. 9E). Immunohistochemical staining showed higher expression of OPN in the R@HSNs group (Fig. 9F). These results indicate the osteoinductive effect of R@HSNs in bone regeneration.

Fig. 9.

Local application of R@HSNs promoted bone regeneration in mice calvarial bone defects. (A) Micro-CT images. Quantification of newly formed bone parameters BV/TV (B), BS/TV (C), and Tb·Th (D). (E) H&E staining and (F) OPN immunohistochemistry. Data are mean ​± ​SD or SEM, n ​= ​3, statistical significance between the groups, ∗P ​< ​0.05, ∗∗P ​< ​0.01, and ∗∗∗P ​< ​0.001. RAPA: free rapamycin.

4. Discussion

During in vivo bone regeneration application, nanomaterials are mainly phagocytized and cleared by immune cells such as macrophages, thereby hindering the direct effect of nanomaterials on osteogenic precursor cells [36]. Macrophage autophagy-mediated osteoimmunology regulates bone regeneration [37]. Based on these aspects, this study designed a virus-like nanocarrier “R@HSNs” for specific intracellular delivery of the autophagy inducer rapamycin in macrophages to induce osteoimmunology favoring bone regeneration. R@HSN reduced the cytotoxicity of rapamycin towards mBMSCs. R@HSNs promoted macrophage M0 to M2 polarization and mitigated the inflammatory phenotype of LPS-induced M1 macrophages. R@HSNs failed to induce the osteogenic differentiation of mBMSCs directly but R@HSNs-treated macrophages robustly promoted the osteogenic differentiation of mBMSCs. Moreover, R@HSNs promoted bone regeneration in mice calvarial bone defects, indicating its possible application in osteoimmunology regulation-based bone tissue engineering.

Nanoparticles with virus-like surface structures mimic virus surface topology and roughness [29], which facilitate their endocytosis and internalization in immune cells [[26], [27], [28]]. Virus-like nanoparticles have demonstrated promising potential in drug delivery [29] and antibacterial [30] applications. Silica nanoparticles have good biocompatibility, biosafety, and stability properties, and have entered clinical trials for various biomedical applications, including oral drug delivery, diagnosis, and photothermal ablation therapy [31]. Silica-based mesoporous hollow R@HSNs showed surface morphology of virus-like spikes and a higher rate of endocytosis and internalization in macrophages than was the case in mBMSCs. R@HSNs were mainly captured in macrophage lysosomes. Macrophage lysosomes degraded the nanocarriers and may cause intracellular drug release. These carriers were mainly captured in macrophage lysosomes. These results indicate R@HSNs as a possible nanocarrier for intracellular rapamycin delivery in macrophages.

Although autophagy inducer rapamycin induces osteogenic differentiation of BMSCs and bone regeneration [38,39], it has dose-dependent adverse effects, including cytotoxicity to BMSCs, promotion of osteoclastogenesis [20,21], hypertriglyceridemia, hypertension, hypercholesterolemia, and increased creatinine [19]. Moreover, rapamycin as an inhibitor of the Akt/mTOR signaling attenuates mRNA translation as shown in osteoblasts [40,41] and free rapamycin equivalent to the dose loaded in R@HSNs inhibited the proliferation and osteogenic differentiation of mBMSCs in vitro as well as bone regeneration in mice calvarial bone defects. Rapamycin is a poorly soluble drug, unstable in gastric acid, and has an oral bioavailability of only 14% [18]. It has been reported that rapamycin ameliorates inflammasome formation and clears reactive oxygen species in macrophages inducing an anti-inflammatory phenotype of macrophages [42,43]. R@HSNs were cytocompatible to macrophages and mBMSCs and robustly induced autophagy in macrophages compared with the equivalent dose of free rapamycin. Macrophage autophagy has been reported to promote M2 polarization and reduce the pro-inflammatory properties of M1 macrophages [44]. R@HSNs induced M2 macrophage polarization and inhibited pro-inflammatory properties of M1 macrophages via autophagy activation. Moreover, these effects of R@HSNs were nullified during the inhibition of macrophage endocytosis of R@HSNs. These results indicate the importance of nanocarrier-mediated rapamycin intracellular delivery on autophagy-induced M2 macrophage polarization and immunomodulation. Our data indicate that reducing the in vivo application dose of rapamycin while maintaining bioavailability and targeted macrophage internalization could induce macrophage immunomodulation and eliminate free rapamycin-mediated local and systemic adverse effects.

M2 macrophages promote osteogenic differentiation and bone regeneration via the release of anti-inflammatory cytokines and growth factors [45]. In contrast, M1 macrophage-released pro-inflammatory cytokines inhibit osteoblast functions and bone regeneration [46]. It is well established that IL-10 is the key growth factor in the M2-macrophage conditioned medium that promotes osteogenic differentiation of BMSCs and bone regeneration in vivo [47,48]. In this study, R@HSNs-induced M2 macrophage polarization and R@HSNs-treated macrophages robustly promoted the osteogenic differentiation of mBMSCs compared with the free rapamycin-treated macrophages. Interestingly, free rapamycin inhibited, HSNs did not affect, but R@HSNs robustly promoted bone regeneration in mice calvarial bone defects. Moreover, the protein level expression of IL-10 was robustly upregulated in the conditioned medium of R@HSNs-treated M2 macrophages, indicating IL-10 as a possible key growth factor that promoted macrophage-mediated osteogenic differentiation of mBMSCs and bone regeneration in this study. These results indicate that R@HSNs-mediated intracellular delivery of rapamycin in macrophages promotes M2 polarization and regulates osteoimmunology to induce osteogenic differentiation and bone regeneration.

Our study successfully prepared a virus-like nanocarrier “R@HSNs” that promoted osteogenic differentiation mBMSCs via regulation of macrophage immunomodulation and induced bone regeneration. However, we did not investigate the role of local R@HSNs-mediated macrophage immunomodulation on bone regeneration in vivo. Further studies analyzing the bone defect repair potential of R@HSNs using more mice in each group, later time points of observation (8 and 12 weeks), and macrophage-ablated mice are recommended to clarify this issue.

5. Conclusions

This study designed a virus-like nanocarrier “R@HSNs” for intracellular delivery of rapamycin specific in macrophages. R@HSNs promoted M2 macrophage polarization and inhibited pro-inflammatory properties of M1 macrophages via autophagy activation. R@HSNs-induced M2 macrophages promoted osteogenic differentiation of mBMSCs and locally-grafted R@HSNs showed a higher bone regenerative effect in mice calvarial bone defects compared with free rapamycin. Our results indicated that R@HSNs-mediated intracellular delivery of rapamycin promotes M2 macrophage polarization via activation of autophagy, and regulates osteoimmunology to promote bone regeneration that could minimize the free rapamycin-mediated local and systemic adverse effects during bone defect repair.

Credit author statement

Qing Zhang: Conceptualization, Methodology, Investigation, Data curation, Formal analysis, Visualization, Writing – review & editing; Mengyu Xin: Methodology, Investigation, Data curation, Formal analysis, Writing – original draft & review; Shuang Yang: Methodology, Investigation, Data curation; Qiuyu Wu: Methodology, Investigation; Xi Xiang: Investigation; Tianqi Wang: Methodology; Wen Zhong: Investigation; Marco N Helder: Supervision; Richard T Jaspers: Supervision; Janak Lal Pathak: Conceptualization, Supervision, Writing – review & editing; Yin Xiao: Funding acquisition, Project administration, Supervision, Writing – review & editing.

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.