Downregulation of the long non-coding RNA MEG3 promotes osteogenic differentiation of BMSCs and bone repairing by activating Wnt/β-catenin signaling pathway

Juan Liu Shanghai First People's Hospital: Shanghai Jiaotong University First People's Hospital Xin Qi (  qixin19871012@163.com ) Shanghai Pudong Hospital Hong-Sheng Miao Shanghai First People's Hospital: Shanghai Jiaotong University First People's Hospital Zi-Chao Xue Qingdao Municipal Hospital Group San-Hu Zhao Shanghai First People's Hospital: Shanghai Jiaotong University First People's Hospital Guo-Yi Gao Shanghai First People's Hospital: Shanghai Jiaotong University First People's Hospital Mei-Qing Lou Shanghai First People's Hospital: Shanghai Jiaotong University First People's Hospital Cheng-Qing Yi Shanghai Pudong Hospital


Background
In the clinic, skull defects caused by trauma, severe infection, tumor resection, and decompressive craniectomy due to refractory high intracranial pressure are very common in neurosurgery. Although autologous bone transplantation, allogeneic bone transplantation, titanium mesh or polyether-etherketone (PEEK) cranioplasty were widely used in the clinical treatment of skull defects, complications such as donor site morbidity, bone resorption/loosening graft infection, implant exposure and high costs remain unresolved (1)(2)(3). Bone marrow mesenchymal stem cells (BMSCs) retain their self-renewal capability and have the potential to differentiate into a variety of cell types; hence, they are widely used for tissue repairing. Multiple studies have demonstrated combinations of biomaterials, growth factors or gene modi ed BMSCs-based tissue engineering are promising alternative approach to facilitate bone regeneration, and have acquired satis ed results in repairing critical-sized bone defects(4-6).
Maternally expressed gene 3 (MEG3) is a maternally expressed long non-coding RNA, and increasing evidences have revealed that MEG3 emerged as a key regulator in the process of development and tumorigenesis (7,8). We have previously reported that downregulation of MEG3 could promotes angiogenesis after ischemic brain injury (9). However, the function of MEG3 in osteogenic differentiation of MSCs and bone regeneration remains largely unknown. Recent studies have revealed MEG3 inhibits the osteogenic differentiation of periodontal ligament cells (10), human dental pulp stem cells (11) and BMSCs from postmenopausal osteoporosis (12). Down-regulated MEG3 promotes osteogenic differentiation of human dental follicle stem cells (13). While Zheng et al. has proved that down-regulation of MEG3 suppresses osteogenic differentiation of human adipose-derived stem cells (14). Upregulation of MEG3 promotes osteogenic differentiation of MSCs from multiple myeloma patients (15). In addition, Liu et al. has showed that MEG3 is up-regulated in non-union bone fracture, and silencing MEG3 could accelerate tibia fraction healing in mice(16). These contradictory ndings indicate that dysregulation of MEG3 might play an important role in osteogenic differentiation of MSCs and bone remodeling.
In this study, we found the expression of MEG3 increased during the process of osteogenic differentiation. Base on previous studies and our results, we hypothesize that MEG3 may participate in osteogenic differentiation of BMSCs and bone regeneration. The function of MEG3 during osteogenic differentiation of BMSCs was demonstrated by assessing the expression levels of osteogenic genes, alkaline phosphatase activity and calcium deposition. Notably, a further mechanism study revealed that the pro-osteogenic effects of MEG3 may be partially attributed to wnt/β-catenin signaling pathway. A composite poly (3-hydroxybutyrate-co-3-hydroxyhexanoate, PHBHHx)-mesoporous bioactive glass (MBG) scaffold (PHMG), composed of biodegradable PHBHHx and MBG, was selected as the vehicle in this study. Owing to the advantages of its good biocompatibility, favorable loading behavior of its mesoporous structure. Our previous studies have shown that composite PHMG scaffolds has no cytotoxicity and improved cellular a nity, furthermore, PHMG possesses hierarchical mesoporous structure with a comparatively large speci c surface area for cellular adhesion and spread (5). PHMG scaffold adherent with MEG3 modi ed BMSCs were implanted in a critical-sized skull defects of rat model for 8 weeks. Bone regeneration was determined using micro-CT, and histologic analyses. Our results show downregulation of MEG3 could enhanced osteogenic differentiation of BMSCs and promoting bone repairing via Wnt/β-catenin signaling.

Materials And Methods
Cells and reagents human bone marrow stromal cells (hBMSCs) were obtained from four donors who gave their written informed consent. Brie y, marrow was extracted from the femoral midshaft and then suspended in minimum essential medium containing 10% fetal bovine serum (Hyclone; GE Healthcare, Little Chalfont, UK), 100 U/mL penicillin and 100 mg/L streptomycin. Subsequently, the non-adherent cells were discarded; the adherent cells converged to 80-90% con uence and were then replated as passage one (P1) cells. P3 cells were used for experiments. A density of 1×10 5 cells/mL was used in the cellular tests. Recombinant DKK1 was purchased from PeproTech (Rocky Hill, NJ, USA). In accordance with a previous study, the applied concentration of DKK1 was 0.5 μ g/mL (20,25).

Lentiviral packaging and cell infection
Lentivirus knockdown MEG3 particles and lentiviral RFP particles were described as previously (9). The lentiviral RFP particles were used as control group in this study. For infections, hBMSCs were incubated with lentiviral particles and polybrene (5 μ g/mL) in growth medium. After 6 h, the infection medium was discarded. After 3 days, the cells were screened using puromycin (4 μ g/mL; Sigma, Shanghai, China) and then passaged for use in subsequent experiments. The expression of MEG3 was quanti ed by quantitative real-time polymerase chain reaction (qPCR) and immuno uorescence.
Cell Counting Kit-8 (CCK-8) To assess the effect of MEG3 downregulation on the proliferation of hBMSCs, the cells were seeded into a 96-well plate (5000/well) and allowed to adhere for 24 h. After 24 h, the medium was removed, and the cells were treated with 10% CCK-8 (Dojindo, Kumamoto, Japan) in 100 μ L low-sugar Dulbecco's modi ed Eagle's medium (L-DMEM) without fetal bovine serum (FBS) for 2 h at 37 °C. Absorbance at 450 nm, which is directly proportional to cell proliferation, was measured using a microplate reader (ELX808; BioTek, Winooski, VT, USA).

Measurement of alkaline phosphatase (ALP) activity
For the measurement of ALP activity, cells were lysed in radioimmunoprecipitation assay (RIPA) lysis buffer (Beyotime, Shanghai, China), and the lysate (10 μ L) was incubated with 90 μ L fresh solution containing p-nitrophenyl phosphate substrate at 37 °C for 30 min. The reaction was stopped by the addition of 0.5N NaOH (100 μ L), and the absorbance was measured at 405 nm using a microplate reader (ELX808; BioTek). The total protein concentration was measured using a BCA protein assay kit (KeyGen BioTECH, Nanjing, China). The relative ALP activity is expressed as the percentage change in optical density (OD) per unit time per milligram protein: (OD/15 min/mg protein) × 100.

Alizarin red staining (ARS)
After the induction of osteogenic differentiation, mineral deposition was assessed by ARS (Cyagen Biosciences). Cells were xed in 4% paraformaldehyde (Sangon Biotech, Shanghai, China) for 15 min at room temperature and then washed with distilled water. A 1% solution of alizarin red was added and incubated for 30min at room temperature, followed by rinsing with distilled water. The solution was collected, and 200 μ L were plated on 96-well plates, which were read at 560 nm using a microplate reader (ELX808; BioTek). The readings were normalized to the total protein concentration.

RNA isolation and qPCR
Total cellular RNA was isolated using RNAiso reagent (Takara, Dalian, China) and quanti ed by measuring the absorbance at 260 nm (NanoDrop 2000; Thermo Fisher Scienti c, Waltham, MA, USA). Total RNA (≤1000 ng) was reverse-transcribed into cDNA in a reaction volume of 20 μ L using the Double-Strand cDNA Synthesis Kit (Takara, Dalian, China). One microliter of cDNA was used as the template for the qPCR reaction. All gene transcripts were quanti ed by qPCR using the Power SYBR Green PCR Master Mix (Takara) on the ABI StepOnePlus System (Applied Biosystems, Warrington, UK). The mRNAs of the target genes and GAPDH were quanti ed in separate tubes. All primers were synthesized by Sangon Biotech (Shanghai, China). The primer sequences used are shown in Table 1. The cycle conditions were as follows: 95 °C for 30 s and then 40 cycles of 95 °C for 5 s and 60 °C for 30s. The relative target gene expression levels were calculated using the 2 −△△Ct method.

Western blot analysis
Cells were lysed in RIPA lysis buffer supplemented with a proteasome inhibitor (Beyotime). Total proteins were separated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis and then transferred to a polyvinylidene uoride membrane (Millipore, Shanghai, China). After blocking in 5% non-fat milk for 2 h, the membranes were incubated overnight at 4 °C with antibodies speci c to β-actin (1:1000, Abcam Inc., USA), or β-catenin (1:1000, Cell Signaling Technology, USA). Horseradish peroxidase (HRP)conjugated goat anti-rabbit IgG (1:1500; Cell Signaling Technology, USA) was applied as a secondary antibody for 2 h at room temperature. The immunoreactive bands were detected using an enhanced chemiluminescent detection reagent (Millipore, Shanghai, China). Signal intensity was measured using a Bio-Rad XRS chemiluminescence detection system (Bio-Rad, Hercules, CA, USA).

Cell seeding
Prior to seeding cells, the prefabricated PHMG scaffolds were sterilized using gamma irradiation. Cell suspension (100 μ L) was added to four groups at a density of 1×10 4 cells/scaffold. After 4 h, 100 μ L of culture medium was carefully added to the base of the culture plate until the scaffold was covered with su cient culture medium.

In vivo evaluation in animals
Animal experiments were approved by the Research Ethics Committee of the Shanghai General Hospital, and performed in accordance with the Care and Use of Laboratory Animals protocols. Brie y, mature Sprague Dawley (SD) male rats (mean body weight 250-300 g) were provided with sterilized food and water and housed in a barrier facility with a 12-h light/dark cycle. These rats were randomly divided into three groups, each containing six rats: PHMG, PHMG + sh-Ctrl, and PHMG + sh-MEG3 group. For the surgical procedure, as previously described, the animals were anesthetized by intraperitoneal injection of chloral hydrate (4%; 9 mL/kg body weight) and all operations were performed under sterile conditions. A 1.5-cm sagittal incision was made in the scalp and the calvarium was exposed by blunt dissection. Two critical-sized calvarial defects with a bilateral diameter of 5 mm were created using a dental trephine, and the scaffolds were then implanted into the defects. Following the operation, the animals received intramuscular antibiotic injections, were allowed free access to food and water and were monitored daily for potential complications. Eight weeks after the operation, the rats were killed by an overdose of anesthetics and their craniums were harvested and xed in a 4% paraformaldehyde solution buffered with 0.1 M phosphate solution (pH 7.2) overnight before further analysis.

Micro-computed tomography (CT) evaluation
All the harvested specimens were examined using the mCT-80 system to evaluate new bone formation within the defect region. Brie y, the undecalci ed samples were scanned at a resolution of 18 μ m and decalci ed samples perfused with Micro l® were scanned at a resolution of 9 μ m. After 3D reconstruction, the bone mineral density (BMD) and bone volume fraction (bone volume/total volume [BV/TV]) in the defect regions were used to calculate new bone formation using the auxiliary software of the mCT-80 system.

Histological and Immunohistochemical (IHC) analysis
The one part of calvarias were decalci ed in 10% EDTA for 14 days, dehydrated with graded ethanol solutions, embedded in para n and sectioned at 5 μ m at the central area of the defect. Sections were stained with hematoxylin and eosin (HE) to observe new bone formation.
The other part of each cranium was decalci ed for approximately 2 weeks, dehydrated using a graded alcohol series, embedded in para n and sectioned into 5 μ m sections. Osteocalcin (OCN) IHC was performed to evaluate osteogenesis in specimens.

Statistical analysis
Statistical analysis was performed using SPSS 17.0 software (IBM, Armonk, NY, USA). All experiments were performed at least in triplicate, and the data are presented as means ± standard deviation.
Statistical signi cance was determined using a two-tailed Student's t-test when comparing two groups, and one-way ANOVA followed by Bonferroni's post hoc test when comparing more than two groups. P < 0.05 was considered to indicate statistical signi cance.

Results
MEG3 was increased during osteogenic differentiation of BMSCs.
To determine whether Meg3 involved in osteogenic differentiation of BMSCs, we examined the expression Meg3 in BMSCs at day 0, 7, and 14 after osteoblastic induction. Compared with undifferentiated BMSCs, the expression of Meg3 was signi cantly increased at day 7 and 14 during the process of osteogenic differentiation (P < 0.05, Figure 1A).
To understand the role of Meg3 during osteogenic differentiation, we rst performed Meg3 knockdown in the BMSCs using a lentivirus vector. Immuno uorescence showed that the RFP marker was stably expressed in the BMSCs ( Figure 1B). PCR results showed the expression of Meg3 in the lenti-Meg3 treated BMSCs (sh-Meg3 group) were knockdown (0.7-fold) when compared with the mock treated BMSCs (blank group) and lenti-control treated BMSCs (sh-Ctrl group) (P < 0.05, Figure 1C). CCK8 results showed that no signi cant difference was detected in the cell proliferation rate between sh-Meg3 group and sh-Ctrl group at day 1, 3, and 7, which indicate Meg3 knockdown did not affect BMSCs proliferation (Figure supplement).

MEG3 knockdown promotes osteogenic differentiation of BMSCs in vitro.
To further explore the function of Meg3 knockdown during osteogenic differentiation, the levels of osteospeci c genes, including ALP, RUNX2, and osteocalcin (OCN) were detected by qPCR. qPCR analysis revealed that ALP, RUNX2, and OCN mRNA levels were signi cantly higher in sh-Meg3 group at day 7, and 14 than in blank group and sh-Ctrl group (P < 0.05, Figure 2A-C).
We evaluated ALP activity, an early marker of osteogenesis, at day 14 during osteogenic differentiation.
Compared with the blank group and sh-Ctrl group, higher ALP activity was observed in sh-Meg3 group (P < 0.05, Figure 2D and 2F). Calcium deposits were also examined by ARS, and the staining areas were quanti ed by measuring the absorbance at 560 nm. More calcium deposits appeared in the sh-Meg3 group than in blank group and sh-Ctrl group at day 28 (P < 0.05, Figure 2D and 2E).

MEG3 knockdown activates Wnt/β-catenin signaling pathway in BMSCs.
Wnt/β-catenin signaling pathway plays an important role in the osteogenic differentiation of BMSCs. To gain insights into the mechanism by which Meg3 regulates osteogenic differentiation of BMSCs, the expression changes of β-catenin in BMSCs were performed. Western blot results showed that the protein level of β-catenin increased in the sh-Meg3 group when compared with sh-Ctrl group (P < 0.05, Figure 3A and 3B). DKK1 was previously reported as an inhibitor of Wnt/β-catenin pathway.
To verify the relevance of the Wnt/β-catenin pathway and Meg3 knockdown, we evaluated the expression of β-catenin between sh-Ctrl group and sh-Meg3 group treated with or without DKK1. Western blot results showed that Meg3 knockdown resulted in an increase of β-catenin in sh-Meg3 group compared to the sh-Ctrl group, and the upregulation of β-catenin could inhibit by DKK1 (P < 0.05, Figure 3C and 3D). Taken together, these data indicate that Meg3 activates the Wnt/β-catenin signaling pathway in BMSCs.
Next, we determined whether the osteogenesis effect of Meg3 knockdown on BMSCs was mediated via the Wnt/β-catenin pathway. After treated with or without DKK1, osteo-speci c genes of OCN and Runx2 were detected by qPCR in sh-Ctrl group and sh-Meg3 group. Compared with the sh-Ctrl group, silencing of Meg3 increased the expression of OCN and Runx2 in sh-Meg3 group, whereas the increasing of expression of OCN and Runx2 could blocked by DKK1 in sh-Meg3 group (P < 0.05, Figure 4A and 4B).
Moreover, there was also less matrix mineralization at day 28 of osteogenic differentiation in the sh-MEG3 group treated with DKK1than in sh-MEG3 group ( Figure 4C and 4D). Blocking the Wnt/β-catenin pathway was able to reverse the osteogenesis activity induced by knockdown of Meg3 as indicated by ALP activity (P < 0.05, Figure 4E). These results indicate that the Wnt/β-catenin pathway could mediate the osteogenesis effect of MEG3.
MEG3 knockdown accelerated bone repairing in a rat critical-sized skull defect.
To verify the pro-osteogenesis effect of Meg3 in vivo, PHMG adherent with Meg3 knockdown BMSCs were implanted in a critical-sized skull defects of rat model. The 3D morphology and 2D slice images of the newly-formed calvarial bones of each group at week 8 are reconstructed by micro-CT ( Figure 5A-5C). As showed in the sagittal view, little bone growth was observed in the defect in the PHMG group ( Figure  5A3) the PHMG + sh-Ctrl group showed increased new bone formation than PHMG group ( Figure 5B3). Compared with the PHMG group and the PHMG + sh-Ctrl group, newly-formed bone was apparently augmented in the PHMG + sh-Meg3 group, as the defect was almost completely lled with new calvarium, and the interfaces between the scaffold and bone tissues were connected ( Figure 5C3). The local BMDs were markedly the highest at 0.497 ± 0.043 g/cm 3 , and there was a signi cant difference between the PHMG group and PHMG + sh-Meg3 groups (P < 0.05, Figure 5D). Moreover, BV/TV showed the same tendency as the BMD levels, there was a signi cant difference in the PHMG + sh-Meg3 group compared with the PHMG and PHMG + sh-Ctrl groups (P < 0.05, Figure 5E). These results indicate that PHMG scaffold with Meg3 knockdown BMSCs can synergistically improve bone regeneration compared in vivo.
HE staining clearly showed that barely any new bone formation was found in the PHMG group ( Figure  6A1). Only a small amount of new bone formation was observed in the PHMG + sh-Ctrl group ( Figure  6A2). While in the PHMG + sh-Meg3 group, the ingrowth of new bone formation was evident in the central area of the defects as well as in the peripheral area near the pre-existing bones ( Figure 6A3). The osteogenic marker OCN was also detected by IHC staining of decalci ed craniums of each group. The results showed that there was less positive staining for OCN in the PHMG group ( Figure 6B1). Positive brown staining for OCN was more apparent in the PHMG + sh-Ctrl ( Figure 6B2) and PHMG + sh-Meg3 groups ( Figure 6B3). HE and OCN IHC staining analysis of bone regeneration in calvarial defects indicated that Meg3 knockdown in BMSCs can increase bone regeneration.

Discussion
In this study, we found the expression of MEG3 was upregulated during osteogenesis in BMSCs. Furthermore, downregulation of MEG3 could promote the osteogenic differentiation of BMSCs. Moreover, we showed that decreasing the expression of MEG3 in BMSCs could markedly accelerate bone repairing with upregulated bone mineral density, bone volume, and increased new bone generation in rat criticalsized skull defects model. Mechanistically, we found that silencing of MEG3 could promoted osteogenic differentiation of BMSCs via the Wnt/β-catenin signaling pathway. To our knowledge, this is the rst report demonstrating that MEG3 enhances osteogenic differentiation of BMSCs, at least partly through activation of the Wnt/β-catenin signaling pathway.
BMSCs combined with biomedical materials, hold great promise for regenerative medicine, especially for treating the unmet critical-sized bone defects in clinics. Therefore, effectively enhancing BMSCs osteogenic differentiation and BMSCs-mediated bone regeneration are crucial in bone tissue regeneration. Recent evidences have revealed that dysregulation of MEG3 is closely associated with bone or bone degenerative diseases. For example Liu et al. found MEG3 is up-regulated in non-union fracture bone(16) while MEG3 is downregulated in Osteoarthritis(26,27). In this study, we found the expression of MEG3 was upregulated during osteogenesis in BMSCs. RUNX2 is a master transcription factor involved in osteogenic differentiation. The expression of RUNX2 was signi cantly increased following the downregulation of MEG3. The levels of an early marker of osteogenic differentiation (ALP) and late markers of osteogenic differentiation (OCN) were also increased due to MEG3 depression. In agreement with our results, Li et al. also found the expression of serum lncRNA MEG3 was increased in fracture patients and intervention with MEG3 siRNA could obviously promote the proliferation and differentiation of osteoblast cell line MC3T3-E1 in vitro (24). In addition, better bone healing was also observed when MEG3-modi ed BMSCs scaffold with PHMG used in a rat critical-sized skull defects model. Similarly Since MEG3 was signi cantly downregulated in cultured dental mesenchymal cells but were upregulated in odontogenic dental mesenchymal tissues (29). Therefore, different expression of MEG3 in different cell types, tissues or diseases indicate MEG3 may have speci cally function under certain states.
Wnt/β-catenin signaling is an essential pathway in the osteogenic differentiation of MSCs and bone maintenance. Wnt signaling results in cellular accumulation of Wnt/β-catenin, followed by nuclear translocation of β-catenin and activation of target genes (17,30). A previous study reported crosstalk between MEG3 and Wnt/β-catenin signaling in stem cells; that is, down-regulated MEG3 promotes osteogenic differentiation of human dental follicle stem cells by epigenetically regulating Wnt pathway (13). Gong et al. also found highly expressed MEG3 could weaken Wnt/β-catenin signaling in glioma (22). MEG3 was reported located in the cytoplasm (13) and nucleus(16). In this study, higher βcatenin accumulation was observed following downregulation of MEG3 during osteogenesis, suggesting that MEG3 downregulation activates β-catenin-mediated transcription. Furthermore, the increased osteogenesis of BMSCs by MEG3 downregulation was blocked by an inhibitor of Wnt/β-catenin. Since after the inhibition of canonical Wnt signaling pathway by DKK1, the osteogenic differentiation marker ALP, OCN and Runx2 were downregulated and less mineralized nodules were formed. Together, our data suggested that downregulated MEG3 might promote osteogenesis of BMSCs by partly activating the Wnt/β-catenin signaling pathway.
The present study has some limitations. First, although we indicated that down-regulation of MEG3 mediates the Wnt/β-catenin signaling pathway to promote osteogenic differentiation of BMSCs, it is probably involved in the activation of other signaling pathways as well. Second, the mechanisms of MEG3 interact with Wnt/β-catenin has not been clari ed completely, and thus further studies are needed.

Conclusion
Based on our data, we found that downregulation of MEG3 could promote osteogenic differentiation of BMSCs and bone repairing, partly through activation of the Wnt/β-catenin signaling pathway. MEG3modi ed BMSCs can be used as a novel promising strategy for skull defects.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. SupplementaryFigure.pdf