https://orcid.org/0000-0003-2647-9200
Abstract
Glucocorticoid-induced osteoporosis (GIOP), the most common cause of secondary osteoporosis, is characterized by significant bone loss, decreased bone quality, and increased fracture risk. The current treatments for GIOP have several drawbacks. Exosomes are vital for cellular processes. However, very few studies have focused on using human umbilical cord mesenchymal stem cell-derived exosomes (hUCMSC-EXOs) for GIOP treatment. In vitro and in vivo dexamethasone was used to evaluate the therapeutic effects of hUCMSC-EXOs on GIOP. CCK-8 and EdU assays were used to evaluate cell viability and proliferation, respectively. We conducted an alkaline phosphatase activity assay, alizarin red staining, Western blotting, and real-time PCR to detect the effect on osteogenesis. TMT-labeled quantitative proteomic and bioinformatic analyses were performed. Furthermore, we performed Western blotting, immunofluorescence, reactive oxygen species assays, and lipid peroxidation assays to investigate the regulatory mechanism by which hUCMSC-EXOs affect cell proliferation and osteogenic differentiation. The in vivo effects of hUCMSC-EXOs were evaluated using micro-CT, hematoxylin, and eosin staining, and immunohistochemical staining. We found that hUCMSC-EXOs reversed the inhibitory effects of glucocorticoids on human bone marrow stromal cell (hBMSC) proliferation and osteogenic differentiation and demonstrated that hUCMSC-EXOs reversed GIOP via the PI3K/AKT signaling pathway, inhibiting lipid peroxidation in vitro and in vivo. HUCMSC-EXOs promote hBMSC osteogenesis through the PI3K/AKT signaling pathway, inhibit ferroptosis, and have therapeutic potential for GIOP in mice.
We showed that hUCMSC-EXOs promoted the proliferation and osteogenic differentiation of hBMSC under the inhibition of glucocorticoids (GCs). Notably, hUCMSC-EXOs reduce GCs-induced ferroptosis by inhibiting reactive oxygen species accumulation and lipid peroxidation in hBMSCs, improving bone microstructure in GIOP via PI3K/AKT signaling pathway. For the first time, the results systematically and completely demonstrate that the PI3K/AKT pathway-ferroptosis-osteogenesis regulatory mechanism exists in the process of hUCMSC-EXOs curing GC-induced osteoporosis. This study provides a novel method for GIOP management.
Introduction
Glucocorticoid (GC)-induced osteoporosis (GIOP) is the most common form of secondary osteoporosis1 characterized by significant bone loss, decreased bone quality, and increased fracture risk.2-4 GIOP can lead to disability and mortality, causing burdens to people and society. GIOP prevalence in the general population is 0.5%-1%5,6 and has remained unchanged from 1999-2016,7 according to a nationwide Denmark sales database. GIOP management focuses on prevention and early treatment. However, clinical agents remain limited owing to limited safety data and can cause unexpected side effects during the process of glucocorticoid (GC) therapy.8 For instance, although denosumab effectively prevents bone loss and improves bone mass in patients treated with glucocorticoids, rapid bone loss may occur when denosumab is discontinued, thereby increasing fracture risk.9 Teriparatide is a bone anabolic agent approved for treating and preventing GIOP over the past 10 years. Nevertheless, it is primarily associated with the patients with the highest fracture risk owing to its cost and inconvenience.1
Various studies have demonstrated that mesenchymal stem cell (MSC)-derived exosomes are beneficial in tissue regeneration.10,11 Exosomes are approximately 40-100 nm-diameter extracellular vesicles secreted by cells and tissues. These lipid bilayer membrane particles encapsulate biologically active substances, including microRNAs, proteins, and lipids,12 which mediate intercellular communication. Exosomes have the advantages of low immune rejection, high permeability, low toxicity, and good targeting ability.13 Wang et al have reported that bone marrow MSC (BMSC)-derived exosomes exhibit excellent regeneration ability, whereas umbilical cord MSC (UCMSC)-derived exosomes were prominent in tissue damage repair.14 However, access to BMSC is hampered by limited sources, invasive and painful procedures, and safety hurdles.15,16 In contrast, UCMSCs have the advantages of easy collection,17 low immunogenicity,18 and perfect proliferation and differentiation capacity.19 Therefore, human UCMSC-derived exosomes (hUCMSC-EXOs) represent a novel and promising method for treating GIOP.
The PI3K/AKT axis is a classic signaling pathway involved in regulating cell proliferation, differentiation, and apoptosis.20,21 Several studies have confirmed that PI3K/AKT pathway plays an important role in bone regeneration.22-24 Ferroptosis is a novel cell death mode characterized by the abnormal accumulation of iron-mediated lipid peroxidation products. Dixon et al. first proposed the definition of ferroptosis, which is an iron-dependent non-apoptotic cell death mode characterized by lipid reactive oxygen species (ROS) accumulation.25,26 Distinct from other regulated cell death processes, such as apoptosis, pyroptosis, autophagy, and necroptosis, ferroptosis is mainly characterized by decreased glutathione peroxidase 4 (Gpx4) activity, iron accumulation,27 and cellular lipid composition (ROS and LPO: ferroptosis indicators and proximate executioners, respectively28,29) by positive Acsl4 regulation.25,30 Recent studies have indicated that iron overload can partially inhibit osteoblast differentiation and mineralization, while activating osteoclasts, thus inhibiting osteogenesis and eventually causing osteoporosis.31,32 Jin et al33 have demonstrated that VK2 weakens ferroptosis induced by long-term high glucose and strengthens the osteogenic capacity of BMSCs by the AMPK/SIRT1 pathway, both in vivo and in vitro. Lu et al34 found the ferroptosis of osteoblasts in GIOP mice model. Recently, Li et al35 have indicated that melatonin could inhibit ferroptosis via the PI3K/AKT/mTOR axis, thus suppressing GIOP. However, they34,35 did not show that inhibiting ferroptosis promotes osteogenesis.
Based on these findings, we speculated that hUCMSC-EXOs could reverse the inhibitory osteogenic effects of GCs and ferroptosis. Further analyses investigated the underlying mechanisms. In this study, we demonstrated that hUCMSC-EXOs could be transferred into human BMSCs (hBMSCs), inhibit ferroptosis through the PI3K/AKT pathway, promote osteogenesis in hBMSCs in vitro, and treat GIOP mice in vivo.
Material and methods
Cell culture
HUCMSCs and hBMSCs were purchased from Wuhan Sunncell Co. Ltd. and cultured in a-MEM supplemented with 10% FBS (Gibco) and 100 mg/mL penicillin/streptomycin (Gibco) for 3 days at 37 °C with 5% CO2.36 Once the effect of the hUCMSC-EXOs was tested, FBS was replaced with exo-free FBS.
Exosome isolation and identification
First, we removed the exosomes from FBS by ultracentrifugation (Thermo Fisher Scientific) as previously reported.36 The culture supernatant of hUCMSCs cultured with exo-free FBS from 3 to 6 generations was collected and centrifuged at 300 × g at 4 °C for 10 minutes to pellet the cells. Then, the supernatant was filtered with a 0.22-μm filter to remove cell debris, followed by a final ultracentrifugation at 120 000 × g for 70 minutes at 4 °C. The isolated exosomes were subdivided and frozen at −80 °C until use. Transmission electron microscopy (TEM) and nanoparticle tracking analysis (NTA) were used to identify hUCMSC-EXOs. Meantime, CD63, CD9, and TSG101 were detected by Western blotting.
Exosome labeling with PKH-26
The PKH-26 red membrane dye (Solarbio) was used according to the manufacturer’s instructions. Briefly, the isolated exosomes were mixed with PKH-26 for 30 minutes in dark. Then, we used 10% BSA to stop the staining reaction, followed by ultracentrifugation at 120 000 × g for 1 hour at 4 °C. After aspirating the supernatant, the exosomes were resuspended in 100 µL cold PBS (Gibco).
Cell counting kit-8 (CCK-8) assay
The CCK-8 (Solarbio) assay was used to detect cell proliferation. First, hBMSCs were seeded into 96-well plate and cocultured with different treatments at 37 °C with 5% CO2. On days 0-4, a 96-well plate was obtained and used according to the manufacturer’s instructions. A microplate reader (Thermo Fisher Scientific) was used to measure the OD at 450 nm.
5-Ethynyl-2ʹ-deoxyuridine (EdU) staining assay
An EdU Apollo567 in vitro assay kit (RiboBio) was used to evaluate DNA replication activity. Briefly, 3000 hBMSCs per well were seeded in 96-well plate and allowed to attach to the bottom for 24 hours. Different treatments were administered in vitro, and subsequent experiments were conducted according to the manufacturer’s instructions. The images were acquired using an inverted fluorescence microscope.
Alkaline phosphatase activity assay
After seeding in 12-well plates, hBMSCs were cultured in an osteogenic differentiation medium37 (OriCell) using different interventions at 37 °C with 5% CO2 for 24 hours. The osteogenic differentiation medium was replaced thrice a week. After fixation in 4% paraformaldehyde for 30 minutes, a BCIP/NBT alkaline phosphatase color development kit (Beyotime) was used to detect alkaline phosphatase (ALP) activity according to the manufacturer’s instructions. The dye was removed once the color reached the desired level. After rinsing with PBS, the cells were imaged under a microscope. The results were analyzed using the ImageJ software.
Alizarin red staining assay (ARS)
After seeding in 12-well plates, hBMSCs were cultured in osteogenic differentiation medium with different treatments at 37 °C with 5% CO2 for 24 hours. The osteogenic differentiation medium was replaced every other day for 3 weeks. After fixation in 4% paraformaldehyde for 30 minutes, 1 mL alizarin red solution (Aladdin) was added, followed by incubation for 5 minutes at room temperature in dark. The cells were then gently washed twice with PBS and observed under a microscope. The results were analyzed using the ImageJ software.
Oil red O staining assay
HBMSCs were cultured in adipogenic differentiation medium38 (Oricell) in 12-well plates for 3 weeks. First, we aspirated the complete adipogenic differentiation medium from the wells and washed it with PBS for 3 times. After fixation with 4% paraformaldehyde, the wells were washed 3 times with PBS. Subsequently, Oil red O staining solution (Cyagen) was added according to the manufacturer’s instructions and stained for 30 minutes. Finally, the dried samples were examined under a microscope.
Alcian blue staining assay
HBMSCs were cultured in chondrogenic differentiation medium (Oricell) for 3 weeks. First, we aspirated the complete chondrogenic differentiation medium from the wells and washed with PBS for 3 times. After fixation with 4% paraformaldehyde, the wells were washed 3 times with PBS. Then, alcian blue staining solution (Cyagen)39 was added following the manufacturer’s instructions. Finally, we examined the effect of staining on the cartilage under a microscope.
Western blotting assay
The protein samples were extracted from cells as described before.40 First, we used 5 × loading buffer to dilute the cell lysates, which were subsequently boiled at 100 °C for 5 minutes. Then, electrophoresis was conducted with 20 µg protein mixture in each channel. The electrophoresed products were transferred to PVDF membranes and blocked using 5% BSA for 2 hours. After washing with Tris-buffered saline containing 0.1% Tween 20 (TBST) thrice, the membranes were incubated overnight at 4 °C with primary antibodies. They were then washed with TBST thrice, incubated with horseradish peroxidase-labeled secondary antibodies at room temperature for 1 hour, and finally examined using Amersham Imager 680.
The primary antibodies used were anti-CD9 (CST, 98327, 1:1000), anti-TSG101 (Abcam, ab30871, 1:1000), anti-CD63 (Abcam, ab59479, 1:1000), anti-Total-AKT (Abcam, ab8805, 1:500), anti-p-AKT(Thr308) (Cell Signaling Technology, 13038, 1:1000), anti-p-AKT(Ser473) (Cell Signaling Technology, 4060, 1:2000), anti-GAPDH (Proteintech, 10494-1-AP, 1:5000), anti-PI3K (Abcam, ab302958, 1:1000), anti-p-PI3K (p55) (Cell Signaling Technology, 11889, 1:1000), anti-BMP2 (Abcam, ab284387, 1:1000), anti-Osteopontin (Abcam, ab214050, 1:1000), anti-col1a1 (Cell Signaling Technology, 72026, 1:1000), and anti-GPX4 (Abcam, ab125066, 1:1000).
Real-time reverse-transcriptase polymerase chain reaction
The RT-PCR was performed as reported previously.41 Total RNA was extracted using Trizol Reagent (Invitrogen) and reverse transcribed using a PrimeScript RT Reagent Kit (TaKaRa). The SYBR Premix Ex Taq II kit (TaKaRa) was used for RT-PCR analysis. Finally, a Roche LightCycler 480 sequence detection system was used for detection. GAPDH was used as a loading control for the RT-PCR analysis. The primers used for RT-PCR were: ALP-F: 5ʹ-CTGGTACTCAGACAACGAGATG-3ʹ, ALP-R: 5ʹ-GTCAATGTCCCTGATGTTATGC-3ʹ, BMP2-F: 5ʹ-TCCACTAATCATGCCATTGTTCAGA-3ʹ, BMP2-R: 5ʹ-GGGACACAGCATGCCTTAGGA-3ʹ, RUNX2-F: 5ʹ-TCCACACCATTAGGGACCATC-3ʹ, and RUNX2-R: 5ʹ-TGCTAATGCTTCGTGTTTCCA-3ʹ.
Tandem mass tag technology for quantitative proteomic analysis
In this study, cells treated with dexamethasone in the absence or presence of hUCMSC-EXOs (Dex and Dex + Exo groups, respectively) were used for proteomic analysis. Sample lysis and protein extraction were performed using the SDT buffer composed of 4% sodium dodecyl sulfate (SDS), 100 mM Tris-HCl, and 1 mM DTT. The BCA Protein Assay Kit (Bio-Rad) was employed for protein quantification. Trypsin was used for protein digestion according to the filter-aided sample preparation procedure described by Matthias Mann.42 The digest peptides were desalted on C18 Cartridges (Empore SPE Cartridges C18 (standard density), bed I.D. 7 mm, volume 3 mL, Sigma), followed by vacuum centrifugation for concentrating and reconstituted in 40 µL 0.1% (v/v) formic acid.
The proteins in each sample were diluted with 5× loading buffer. After boiling for 5 minutes, 20 μg protein was loaded on 12.5% SDS polyacrylamide gel and electrophoresed for 90 minutes at 14 mA constant current. The protein bands were visualized using Coomassie Blue R-250 staining. Then, 100 μg peptide mixture of each sample was labeled using Tandem Mass Tag (TMT) reagent (Thermo Scientific) according to the manufacturer’s instructions. A high pH Reversed-Phase Peptide Fractionation Kit (Thermo Scientific) was used to fractionate the labeled peptides. A Q Exactive mass spectrometer (Thermo Scientific) coupled to an Easy nLC (Thermo Fisher Scientific) for 60/90 minutes was used for the LC-MS/MS analysis. The raw MS data for each sample were searched using the MASCOT engine (Matrix Science; version 2.2) embedded in Proteome Discoverer 1.4 software for identification and quantitative analysis.
Bioinformatic analysis
Hierarchical clustering analysis was conducted using Cluster 3.0 and Java Treeview software. To identify homologous sequences, NCBI BLAST + client software and InterProScan were used to search the selected differentially expressed protein sequences. Next, the Blast2GO software was employed to map the gene ontology (GO) terms and annotate the sequences. Finally, the studied proteins were blasted against the online KEGG database to retrieve their KEGG orthology identifications with subsequent pathway mapping in KEGG.
Immunofluorescence
HBMSCs were seeded in 10% FBS-supplemented α‐MEM for 24 h in 12‐well plates, after which different agents were added for a certain period. The cells were fixed with 4% paraformaldehyde at room temperature for 15 minutes, permeabilized using 0.2% Triton X-100 at room temperature for 30 minutes, and blocked with 10% normal goat serum (Beyotime) at room temperature for 1 hour. After incubation with primary antibodies against GPX4 and p-AKT (Ser 473) at 4 °C overnight, the cells were incubated with the corresponding fluorescein isothiocyanate (FITC)‐conjugated fluorescent secondary antibodies at room temperature for 1 hour in dark. DAPI was used to stain the cell nuclei. The cells were observed under a microscope. The results were analyzed using the ImageJ software.
ROS assay
Intracellular ROS levels were determined using a Reactive Oxygen Species Assay Kit (Solarbio) following the manufacturer’s protocol. After different interventions, hBMSCs were incubated with a serum‐free a-MEM containing ROS assay kit for 30 minutes at 37 °C in the dark. After washing, the fluorescence was observed using a fluorescence microscope.
Lipid peroxidation assay
BODIPY 581/591 C11 (Thermo Fisher Scientific, D3861) was used as a lipid peroxidation sensor following the manufacturer’s protocol. The hBMSCs were incubated with a serum-free medium mixed with 2 μM BODIPY 581/591 C11 at 37 °C in the dark for 30 minutes after different treatments. The fluorescence images were acquired using a fluorescence microscope. The probe fluorescence was determined by simultaneous red (581/610 nm) and green (484/510 nm) signal acquisition. The lipid peroxidation ratio index was calculated using the ImageJ software.
Animal model
All animal experiments were approved by the Animal Research Ethics Committee of Shandong Provincial Hospital and performed in accordance with the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals. Thirty 8-week-old male C57BL/6J mice were randomly categorized into six groups: (1) a control group, (2) dexamethasone group (Dex), (3) dexamethasone + exosomes group (Dex + Exo), (4) exosomes + LY294002 group (Exo + LY294002), (5) exosomes + Erastin group (Exo + Erastin), and (6) dexamethasone + Ferr-1 group (Dex + Ferr-1). Dexamethasone (500 μg/day) was injected intramuscularly for 8 weeks to induce GIOP. Then, 100 μg exosomes were injected intramuscularly every 2 days for 8 weeks. Moreover, 25 mg/kg LY294002 or 40 mg/kg Erastin were intraperitoneally injected twice a week for 8 weeks. Ferr-1 (5 mg/kg per mouse) was intraperitoneally injected thrice a week for 8 weeks. The animals were excluded if they died prematurely. Eight weeks later, the mice were euthanized, and then the tibial plateaus and femoral heads were isolated for further investigation.
Micro-CT analysis
A micro-CT imaging system (Scanco) was used for bone scanning. First, we scanned the tibial plateau. For further quantitative analysis, the cancellous bone of the tibial plateau was selected as the region of interest. Then, μCT Evaluation Program was employed to reconstruct the 3-dimensional structure and calculate the bone volume over total volume (BV/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th), and trabecular spacing (Tb.Sp).
Hematoxylin and eosin and immunohistochemical staining
The decalcified tibial plateau and femoral head were embedded in paraffin and cut into 5 µm sections. After deparaffinizing in xylene, rehydrating in a graded series of ethanol solutions, and rinsing in distilled water, hematoxylin and eosin (HE) staining was performed for histomorphometric analysis. Immunohistochemical (IHC) staining was performed to evaluate the underlying mechanism using an anti-p-AKT (Thr308) antibody and an anti-GPX4 antibody.
Statistical analysis
At least 3 independent experiments were performed. The data were analyzed with GraphPad Prism 8.0 and shown as means ± SD. More than 2 groups were compared using one-way analysis of variance (ANOVA) combined with Tukey’s multiple comparison test. Data with P < .05 were considered statistically significant. *P < .05, **P < .01, ***P < .001, ****P < .0001. NS denotes no statistical significance.
Results
HBMSC and hUCMSC characterization
HBMSCs (Supplementary Figure S1A) and hUCMSCs (Supplementary Figure S1F) have a whirlpool shape and each cell was spindle-shaped. In the 3 lineage differentiation assays, induced hBMSCs (Supplementary Figure S1B–E) and hUCMSCs (Supplementary Figure S1G–J) displayed stained areas, mineralization nodules, cartilage tissue, and lipid droplets. The osteogenic capacity of the hBMSCs was higher than that of the hUCMSCs. Moreover, trilinear differentiation induction indicated that the obtained cells were mesenchymal stem cells.
HUCMSC-EXO identification
TEM, NTA, and Western blotting were performed to identify hUCMSC-EXOs. TEM indicated that hUCMSC-EXOs have a 100 nm-diameter round lipid bilayer membrane (Figure 1A). NTA showed that the overall size distribution of the exosomes was 50-100 nm (Figure 1B). Western blotting revealed that the extracted hUCMSC-EXOs contain specific exosomal biomarkers including TSG101, CD9, and CD63 (Figure 1C). In addition, hUCMSC-EXOs were labeled with PKH26 and endocytosis was detected in hBMSCs (Figure 1D). These results indicated that the isolated samples were exosomes that could be internalized by hBMSCs and qualified for further investigation.
HUCMSC-EXOs counteracts the decreased proliferation induced by dexamethasone in hBMSCs in vitro. (A) HUCMSC-EXO morphology identification using transmission electron microscopy (TEM; scale bar100 nm). (B) HUCMSC-EXO particle size and concentration measurement using nanoparticle tracking analysis (NTA). (C) Western blotting assay of the surface biomarkers CD9, CD63, and TSG101. (D) HUCMSC-EXOs were labeled with PKH26 and hUCMSC-EXO endocytosis detection using an inverted fluorescence microscope. (E) HBMSC proliferation detection using CCK-8 assay. (F) The quantitative calculation of HBMSC proliferation. (G) DNA replication monitoring in hBMSCs using 5-ethynyl-2ʹ-deoxyuridine (EdU) staining (scale bar 100 μm). (H) EdU-positive percentage calculated using ImageJ.
HUCMSC-EXOs counteracts the decreased proliferation induced by dexamethasone in hBMSCs in vitro
HBMSCs were cultivated in a conditioned medium supplemented with 100 μM dexamethasone with or without 50 μg/mL hUCMSC-EXOs. CCK-8 and EdU assays were used to investigate the effects of hUCMSC-EXOs on hBMSC proliferation. The results of the CCK-8 and EdU assays showed that from day 2, the proliferation rate of hBMSCs in the Dex group decreased than that in the control group, whereas hUCMSC-EXOs attenuated the Dex-mediated suppression of proliferation (Figure 1E–H). These results indicated that even under the inhibition using high GC concentrations, hUCMSC-EXOs enhanced hBMSC proliferation.
Effects of hUCMSC-EXOs on osteogenic differentiation of GC-treated hBMSCs in vitro
To investigate the effects of hUCMSC-EXOs on the osteogenic differentiation of hBMSCs under GC conditions, Western blotting, ALP staining, and ARS were performed. Then, hBMSCs were cultivated in an osteogenic differentiation medium supplemented with 100 μM Dex with or without 50 μg/mL hUCMSC-EXOs. Western blot analysis revealed that osteogenic protein expression (BMP2, OPN, and Col1a1) after Dex treatment was suppressed than that in the control, which was ameliorated by hUCMSC-EXOs (Figure 2A–D), while the expression of GPX4, a ferroptosis marker,43 was similar to that of osteogenic capacity (Figure 2A and E). ALP activity assay and ARS results revealed that the stained area of the Dex group was the smallest (Figure 2F–I). Collectively, these findings indicate that hUCMSC-EXOs strongly promote the osteogenic differentiation of hBMSCs.
Effects of hUCMSC-EXOs on osteogenic differentiation of GC-treated hBMSCs in vitro. Western blotting, alkaline phosphatase (ALP) activity assay, and alizarin red staining (ARS) were used to evaluate the osteogenic differentiation of hBMSCs after different treatments. (A) After osteogenic differentiation induction, the effect of hUCMSC-EXOs on Col1a1, BMP2, OPN, and Gpx4 expression levels were determined by Western blotting. (B-E) Quantitative analysis of Western blots using ImageJ software. (F) ALP activity assay (scale bar 500 μm). (G) Relative ALP-positive area was calculated using ImageJ software. (H) ARS (scale bar 500 μm). (I) Relative Alizarin red-positive areas were calculated using ImageJ software.
Analysis of functional attributes related to gene sets and signaling pathway
For further investigation, TMT-labeled quantitative proteomic and bioinformatics analyses were performed on the Dex + Exo and Dex groups. We found that 300 proteins were upregulated and 180 were downregulated. The results are illustrated using column plots (Supplementary Fig. S2A and B), volcano plots (Figure 3A), and heat maps (Supplementary Fig. S2C). GO enrichment analysis revealed that the biological processes involved in positive osteoblast differentiation regulation and positive ossification regulation were significantly upregulated in the Dex + Exo group than those in the Dex group (Figure 3B and Supplementary Fig. S2D). In addition, KEGG pathway enrichment analysis was performed to evaluate differentially expressed genes. The results indicated PI3K/AKT pathway probably being regulated and ferroptosis downregulation (Figure 3C and D). However, the relationship between osteogenesis, PI3K/AKT signaling pathway, and ferroptosis remains unclear.
Functional attributes related to hUCMSC-EXO-induced changes in proteomic and signaling pathways. Proteomic comparison between the cells treated with dexamethasone in the presence or absence of hUCMSC-EXOs (Dex + EXOs and Dex groups). (A) Volcano plot showing differentially expressed protein, red upregulated, and blue downregulated. (B) Gene ontology (GO) biological process classification of genes. (C and D) KEGG functional enrichment analysis.
HUCMSC-EXOs promote GC-treated hBMSC osteogenesis by activating PI3K/AKT signaling pathway in vitro.
To further investigate the role of PI3K/AKT pathway on osteogenic differentiation regulation by exosomes, we used LY294002 as a PI3K pathway inhibitor, and performed Western blot analysis. The results showed that PI3K and AKT phosphorylation at both Thr 308 and Ser 473 sites was downregulated in the Dex group, which notably increased after adding hUCMSC-EXOs. However, PI3K and AKT (Thr 308 + Ser 473) phosphorylation level in the Exo + LY294002 group was lower than that in the Dex + Exo group (Figure 4A and C–G). The expression of osteogenic differentiation indicators (Col1a1, BMP2, and OPN) increased in the Dex + Exo group than that in the Dex group. In contrast, Exo + LY294002 downregulated the expression of osteogenic differentiation indices than Dex + Exo (Figure 4B and H–J). The alteration in GPX4 expression agreed with the alteration in osteogenic differentiation indices (Figure 4B and K). The results of ALP activity assay (Figure 4L and M) and ARS (Figure 4N and O) were consistent with those of Western blotting. Furthermore, immunofluorescence of AKT phosphorylation at Ser473 was performed in different treatment groups. The fluorescence intensity of the Dex + Exo group was higher than that of the Dex and Exo + LY294002 groups (Figure 4P and Q). These results revealed that hUCMSC-EXOs promoted hBMSC osteogenesis through the PI3K/AKT signaling pathway in vitro.
HUCMSC-EXOs promote GC-treated hBMSC osteogenesis by activating PI3K/AKT signaling pathway in vitro. (A and B) Identification of the underlying correlation between PI3K/AKT signaling pathway and osteogenic differentiation using western blotting. (C–K) Quantitative analysis of Western blotting images using ImageJ. (L) Alkaline phosphatase (ALP) activity assay staining (scale bar 500 μm). (M) Relative ALP-positive area as calculated by ImageJ. (N) Alizarin red staining (scale bar 500 μm). (O) Relative Alizarin red-positive area as calculated by ImageJ. (P) p-AKT (Ser 473) expression detection using immunofluorescence. (Q) Quantitative analysis of immunofluorescence by ImageJ.
HUCMSC-EXOs inhibit GC-treated hBMSC ferroptosis via PI3K/AKT signaling pathway in vitro.
To elucidate the mechanism by which hUCMSC-EXOs inhibit hBMSC ferroptosis, we performed an EdU assay to assess hBMSC proliferation. HBMSC proliferation was inhibited after Exo + LY294002 treatment, which blocked the PI3K/AKT pathway, than after Dex + Exo treatment. Ferrostatin-1(Ferr-1), a ferroptosis inhibitor,44 attenuated hBMSC proliferation inhibition. Erastin44 was used as a ferroptosis inducer to inhibit system Xc- activity. Statistical analysis revealed no differences between the Dex + Exo, Dex + Erastin, and Dex + Ferr-1 groups (Figure 5A and B). We performed immunofluorescence and Western blotting to assess Gpx4 expression. The fluorescence intensity of the Dex group was lower than that of the control group, which was reversed in the Dex + Exo group. Fluorescence intensity in the Exo + Ly294002 group decreased when the PI3K/AKT pathway was blocked. The fluorescence intensity in the Dex + Ferr-1 group was higher than that in the Dex group. No statistically significant difference was found between the Dex + Exo and Dex + Ferr-1 groups (Figure 5C and D). The results of Western blotting agreed with those obtained by immunofluorescence. In contrast, no statistically significant difference was observed between the Dex + Exo and Exo + Erastin groups (Figure 5E and F). In addition, a ROS assay was conducted to assess the ferroptosis level in different groups. The fluorescence intensity of the Dex group was significantly higher than that of the Dex + Exo and control groups, and ROS accumulated again when the PI3K/AKT signaling pathway was blocked. The fluorescence intensity of the Dex + Ferr-1 group was lower than that of the Dex group. No statistical differences were found between the Dex + Exo, Exo + Erastin, and Dex + Ferr-1 groups (Figure 5G and H). The results of the lipid peroxidation assay (Figure 5I and J) were in accordance with the ROS findings. Therefore, these findings imply that hUCMSC-EXOs inhibit hBMSC ferroptosis via the PI3K/AKT signaling pathway.
HUCMSC-EXOs inhibit GC-treated hBMSC ferroptosis via PI3K/AKT signaling pathway in vitro. (A) Detection of DNA replication of hBMSCs after different treatments using 5-ethynyl-2ʹ-deoxyuridine (EdU) staining (scale bar 100 μm). (B) Quantitative analysis of EdU by ImageJ. (C) Gpx4 expression detection using immunofluorescence (scar bar 25 µm). (D) Quantitative analysis of immunofluorescence by ImageJ. (E) Western blotting was performed to evaluate Gpx4 expression with different interventions. (F) Quantitative analysis of Western blotting using ImageJ. (G) Oxidative stress level monitoring using reactive oxygen species (ROS) staining (scale bar 100 μm). (H) Quantitative analysis of the number of ROS-positive cells per field. (I) Detection of lipid peroxidation using the C11 BODIPY™ 581/591 fluorescent probe (scale bar 100 μm). (J) Quantitative analysis of oxidized/non-oxidized C11 ratio using ImageJ.
HUCMSC-EXOs promote GC-treated hBMSC osteogenesis by activating PI3K/AKT pathway-mediated ferroptosis inhibition in vitro.
Based on the above results, we performed Western blotting, RT-PCR, ALP activity assay, and ARS to prove that inhibiting ferroptosis could enhance osteogenesis. According to the results of Western blotting, BMP2, Col1a1, and OPN expression in Dex + Ferr-1 group was substantially increased than those in the Dex group. However, BMP2 and Col1a1 expression were not statistically significantly different between the Dex + Exo, Exo + Erastin, and Dex + Ferr-1 groups (Figure 6A-D). Furthermore, we chose BMP2, ALP, and RUNX2 as target genes to assess osteogenic capacity. The results of RT-PCR in the different groups were consistent with those of Western blotting (Figure 6E-G). These findings were further supported by ALP activity assay (Figure 6H and I) and ARS (Figure 6J and K) results, which were in accordance. In general, we showed that hUCMSC-EXOs promoted osteogenesis via the PI3K/AKT pathway, inhibiting ferroptosis in GIOP cells in vitro.
HUCMSC-EXOs promoted GC-treated hBMSC osteogenesis by activating PI3K/AKT pathway-mediated ferroptosis inhibition in vitro. (A) Col1a1, BMP2, and OPN expression evaluation after different interventions using Western blotting. (B–D) Quantitative analysis of western blotting using ImageJ. (E–G) BMP2, ALP, and Runx2 expression levels in different groups were verified using real-time PCR. (H) Alkaline phosphatase (ALP) activity assay staining (scale bar 500 μm). (I) Relative ALP-positive area as calculated by ImageJ. (J) Alizarin red staining (scale bar 500 μm). (K) Relative Alizarin red-positive area as calculated by ImageJ.
HUCMSC-EXOs promote osteogenesis through PI3K/AKT pathway-mediated ferroptosis inhibition in GIOP mice model
For further in vivo investigations, we established a GIOP mouse model. Micro-CT analysis was conducted to quantitatively detect the bone content within the tibial plateau (Fig. 7A). Relative parameters, including BV/TV, Tb.N, and Tb.Th, were lower in the Dex group than in the control and Dex + Exo groups, indicating that bone loss caused by long-term exposure to high GC concentrations was remarkably attenuated by exosomes. These parameters were worse in the Exo + LY294002 group than in the Dex + Exo group, implying that blocking the PI3K/AKT signaling pathway attenuated osteogenesis. Meanwhile, the above parameters in the Dex + Ferr-1 group were significantly higher than those in the Dex group (Figure 7B–D). Furthermore, the Tb.Sp of the Dex group was significantly higher than that of the control group, and this increase was profoundly attenuated by hUCMSC-EXOs and Ferr-1. Conversely, Tb.Sp attenuation by exosomes was suppressed when the PI3K/AKT signaling pathway was inhibited (Figure 7E). HE staining indicated noticeable bone loss in the Dex group with sparse and thin trabeculae. Exosomes and Ferr-1 reversed the bone loss caused by high-dose GCs, whereas the protection against bone loss caused by exosomes was attenuated by blocking the PI3K/AKT signaling pathway (Figure 7F). The results of IHC staining for p-AKT indicated that AKT phosphorylation in the feromal head decreased in the Dex group and hUCMSC-EXOs could reverse the inhibitory effect. This reversal by hUCMSC-EXOs was attenuated by Ly294002 (Figure 7G). The results of IHC staining for GPX4 implied that GPX4 expression of hBMSCs in the tibial plateau decreased in the Dex group, while the expression of GPX4 was higher in the Dex + Exo group, Exo + Erastin group, and Dex + Ferr-1 group. LY294002 attenuated the promoting effect of hUCMSC-EXOs (Figure 7H). These data suggest that hUCMSC-EXOs regulate GIOP via the PI3K/AKT signaling pathway, inhibiting ferroptosis.
In vivo osteogenesis was regulated through PI3K/AKT pathway-mediated ferroptosis inhibition. (A) Micro-CT scan showing the trabecular levels of tibial plateau after different interventions and rectangle showing the region of interest. (B-E) Quantitative bone volume per tissue volume (BV/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th), trabecular separation (Tb.Sp) analysis after different treatments. (F) Hematoxylin and eosin (HE) staining showing the distribution and morphology of bone trabeculae in bone tissue (scar bar 500 µm). (G) Immunohistochemical staining to detect p-AKT (Thr 308) expression levels in the femoral head of mice after different treatments (scar bar 100 µm). (H) Immunohistochemical staining to detect GPX4 expression levels in the tibial plateau of mice after different treatments (scar bar 100 µm).
Discussion
The clinical use of GCs, particularly high-dose and long-term use, is the leading cause of medication-induced osteoporosis, with more than 30% of patients treated with GC experiencing a fragile fracture and over 10% developing osteonecrosis.45 Interestingly, GC-induced bone loss occurs rapidly in trabecular bone-rich sites and is slow and sustained in cortical bone sites, thus increasing fracture risk over time.46 Current agents for curing GIOP are usually expensive, inconvenient to use, and have side effects. Therefore, more suitable agents for treating GIOP should be identified. Excess GC generally directly affects bone cells, including osteoblasts, osteoclasts, and osteocytes, interfering with their proliferation, differentiation, apoptosis, and other activities, which may be a key event in GIOP pathogenesis. Our study revealed that GCs decreased hBMSC proliferation and osteogenic differentiation.
Many studies have shown that hUCMSC-EXOs have substantial advantages in the treatment of kidney diseases, ocular diseases, Alzheimer’s disease, inflammatory bowel disease, nerve injury-induced pain, spinal cord injury, acute myocardial ischemia, cutaneous wound healing, and type 2 diabetes mellitus because of their excellent tissue regeneration ability.47 Ge et al reported that hUCMSC-EXOs significantly promote osteoblast differentiation, exhibiting a therapeutic function in an OVX mouse model.48 However, no studies have reported whether hUCMSC-EXOs are effective for treating GIOP. In our study, we verified the proliferative effect of hUCMSC-EXOs in the presence of Dex using CCK-8 and EdU assays (Figure 1E–H). In addition, the osteogenic differentiation-promoting impact of hUCMSC-EXOs in vitro was verified using Western blotting, ALP activity assay, and ARS (Figure 2A-I).
For further investigation, we conducted TMT-labeled quantitative proteomic and bioinformatics analyses of the Dex + Exo and Dex groups. GO enrichment analysis revealed that the biological processes involved in positive osteoblast differentiation regulation and positive ossification regulation were significantly upregulated by exosomes (Figure 3B and Supplementary S2D). We found that the number of differentially expressed protein involved in PI3K/AKT signaling pathway was the most, indicating that PI3K/AKT pathway needs to be investigated in depth (Figure 3C). The results of KEGG enrichment revealed ferroptosis downregulated in the Dex + Exo group. Several studies35,49 have shown that GCs can suppress the expression of GPX4, a ferroptosis marker, and enhance the accumulation of lipid peroxidation products in vitro, leading to ferroptosis activation. Li et al35 have demonstrated that melatonin activates the PI3K/AKT/mTOR signaling axis and inhibits the ferroptosis pathway in rat BMSCs, ultimately attenuating GIOP. Thus, we speculated that hUCMSC-EXOs inhibited GC-induced ferroptosis via the PI3K/AKT signaling pathway, promoting osteogenic differentiation. First, we conducted a pathway study using Western blotting, ALP activity assay, ARS, and immunofluorescence assays and demonstrated that hUCMSC-EXOs promote GC-treated hBMSC osteogenesis via PI3K/AKT pathway in vitro (Figure 4A-Q). EdU assay, immunofluorescence, Western blotting, and ROS and lipid peroxidation assays were conducted to study the relative ferroptosis level. These results suggest that hUCMSC-EXOs inhibit ferroptosis through the PI3K/AKT pathway, which is in accordance with our speculation (Figure 5A-J). In contrast to the results of other studies, we found that inhibiting ferroptosis promoted bone formation. The results of Western blotting, RT-PCR, ALP activity assay, and ARS were in agreement with our hypothesis (Figure 6A-K). An in vivo GIOP mouse model was used to demonstrate the benefits of hUCMSC-EXOs and the underlying mechanism. According to the results of micro-CT, HE, and IHC, hUCMSC-EXOs showed therapeutic potential in GIOP and regulated osteogenesis via the PI3K/AKT signaling pathway, inhibiting ferroptosis (Figure 7A-H).
In summary, we demonstrated the positive effects of hUCMSC-EXOs on cell proliferation and osteogenesis. The results, for the first time, systematically and completely showed that the PI3K/AKT pathway-ferroptosis-osteogenesis axis exists in the process of hUCMSC-EXOs curing GC-induced osteoporosis. Nevertheless, this study has some limitations. First, we did not investigate the specific exosome components that activate the PI3K/AKT signaling pathway. Second, the evidence of ferroptosis is insufficient. Further studies should focus on the changes in mitochondria and iron metabolism to clarify the evidence of ferroptosis.
Conclusion
hUCMSC-EXOs promote osteogenesis in hBMSCs in vitro and treat GIOP in vivo. Our research not only implicates PI3K/AKT-ferroptosis-osteogenesis as a potential regulatory mechanism in maintaining bone homeostasis but also provides a novel method for GIOP management.
Author contributions
Zhi-Meng Zhao: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Resources, Software, Writing. Jia-Ming Ding, Yu Li: Data curation, Formal analysis, Investigation, Methodology, Resources, Software, Visualization. Da-Chuan Wang: Conceptualization, Data curation, Formal analysis, Project administration, Supervision, Validation, Writing. Ming-Jie Kuang: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Software, Supervision, Writing
Funding
This study was supported by funding from the Natural Science Foundation of Shandong Province (NO. ZR2021QH235).
Conflict of interest
The authors declared no potential conflicts of interest.
Ethics approval
This study was approved by the Experimental Animal Ethics Committee of Shandong Provincial Hospital, Shandong University, China (NO.2020-071). All procedures involving animals were performed in accordance with the International Guiding Principles for Biomedical Research Involving Animals.
Data availability
All data generated in this study are available from the corresponding author upon reasonable request.
