Differential Expression Proling of microRNAs in hPMSCs Co-culture With a Novel Porous Hydroxyapatite Scaffold

Background and objective: Scaffold materials used for bone defect repair are often limited by the osteogenic ecacy. Meanwhile, microRNAs (miRNAs) have been shown to be involved in regulating the expression of osteogenic related genes. In previous studies, we have veried that the enhancement of osteogenesis using a novel porous hydroxyapatite scaffold (HAG). In this study, we analyzed the contribution of HAG to osteogenic differentiation of Human placenta-derived mesenchymal stem cells (hPMSCs) from the perspective of miRNA differential expression. Methods: The properties of hPMSCs were identied by ow cytometry, including CD44, CD90 and CD45 surface marker. The expression of osteogenic differentiation related genes mRNA and protein were detected by quantitative real-time PCR (qRT-PCR) and western blotting. The mineral depositions were measured by Alizarin red S (ARS) staining. The miRNA proles were performed by microarray assay, and then further summarized through target, gene ontology and pathway analysis. The expression of differential miRNA were veried by qRT-PCR. Results: The results showed that HAG promoted the osteogenic differentiation of hPMSCs. Meanwhile, sequencing results showed that 16 miRNAs were signicantly up-regulated and 29 miRNAs were down-regulated with HAG. In addition, bioinformatics analyses showed that the differentially expressed miRNAs are involved in a variety of biological processes including signal transduction, cell metabolic, cell junction, cell development and differentiation and connect to osteogenic differentiation through axon guidance, MAPK, and TGF-beta signaling pathway. Furthermore, multiple potential target genes of miRNA are closely related to osteogenic differentiation. Conclusion: The work rst conrmed that differential expression of miRNAs in the process of osteogenic differentiation promoted by HAG scaffold, which also lays the foundation for the further construction of bone scaffolds loaded with miRNAs.


Introduction
Bone is a mineralized mesenchymal tissue that resizes mechanical forces and regulates mineral homeostasis and energy metabolism 1 . However, bone defects due to trauma, disease and other causes seriously impair people's quality of life and cause a huge social and economic burden. Although bone has a certain ability of regeneration, serious bone injuries such as large bone defects still need to be solved by bone grafting 2,3 . Autografts is the most ideal way to avoid the occurrence of graft immune response, but it is limited by multiple surgeries and donor site injury. Therefore, bone tissue engineering, a scaffold material loaded seed cells and growth factors, has become a more promising method for clinical therapy. Importantly, biological scaffolds material is a key part of it. After the scaffold material is implanted in the affected area, the cells attach to the scaffold surface according to the properties of the material's surface including microstructure, charge, chemical composition, and so on 4,5 . As a result, there is a growing demand for bone biomaterials with superior biological properties in recent years 6 . MicroRNA (miRNA) is endogenous small noncoding ~ 22-nt RNA, which regulates a series of physiological and pathological processes including bone regeneration and metabolism by targeting mRNA sequences of related genes 7,8 . For example, miR-335-5p has been shown to be highly expressed in mouse embryo osteogenesis 9 . Similarly, two consecutive studies demonstrated that both miR-2861 and miR-3960 are high expression in original mouse osteoblasts 10,11 . Additionally, in different tissue sources, miR-140-5p and miR-140-3p are both con rmed to highly expressed, including human bone-marrowderived stem cells (hBMSCs) 12 . Li et al. showed that miR-29b expression peaks at the mineral deposition stage of MC3T3 cell osteogenic differentiation 13 . Bhushan and his colleagues found that miR-181a expression was upregulated during BMP6-induced osteogenic differentiation of MC3T3 cells 14 . Also, miR-210 was increased in ST2 cell line with BMP-4 dependent manner 15 . Furthermore, let-7, miR-20, miR-34a and miR-199 were con rmed to be increased during hBMSCs osteoblast differentiation 16 . As mentioned above, different miRNAs are involved in the complex process of osteogenic differentiation. However, the mechanism of scaffold material-induced speci c osteogenic associated miRNAs has not been seen.
In this study, we rst observed that the differential expression of miRNA in the process of cell osteogenic differentiation induced by HAG scaffolds in hPMSCs. Hydroxyapatite is the most commonly used scaffold material for bone regeneration, due to its mineral composition with natural bone and HAG is a porous hydroxyapatite scaffold with microchannel structure developed by our team, the osteogenic response of which had be veri ed both in vitro and vivo 17,18 . Here, we compared the expression pro les of miRNAs in hPMSCs with or without HAG by microarray assays. Moreover, we further explore the biological functions that these differential miRNAs might be associated with osteogenesis through bioinformatics analysis and revealed that they are closely related to osteogenic differentiation related signaling pathways, including axon guidance, MAPK, and TGF-beta signaling pathway. The work may lay the foundation for the further construction of bone scaffolds loaded with miRNAs.

Materials And Methods
Cell culture and transfection hPMSCs were cultured as before 18 . In details, cells were cultured in complete medium (DMEM with 10% FBS and 1% PS). When the cell density fusion was 80-90%, the digestive passage culture was conducted with 0.25% trypsin. The surface markers of hPMSCs were analyzed by ow cytometry, including CD44, CD90 and CD45. For cell transfection, 100pM miRNA mimics and NC were transfected to cells followed by the protocol of Lipofectamine™ 2000 Transfection Reagent (Thermo Fisher Scienti c, USA), when the cell density approached 80 percentage. The osteogenic differentiation of hPMSCs were cultured with conditional medium (DMEM with 5%FBS, 20 mmol/L β-glycerophosphate, 50 g/mL vitamin C and 10 mol/L dexamethasone). Alizarin Red staining was performed to detect calcium deposits after 21 days of induction culture. The GFP-mimics were synthesised from Genepharma. The DAPI (Sigma-Aldrich) were stained at 48 hours after transfection, then the pictures were captured by ZEISS AXIO Observer.A1.

Preparation of scaffolds
The HAG scaffold was produced and cleaned by referring to previous description [18]. The surface micromorphology of HAG was con rmed by scanning electron microscopy (SEM). The scaffolds were incubated overnight in DMEM prior to the cell experiment at 37°C.
Osteogenic differentiation of hPMSCs with HAG and quantitative real-time PCR (qRT-PCR) 20µL ell concentrate (4 × 10 6 /mL) was seeded in a HAG. Then, the cells co-culture with the scaffold in incubator. After osteogenic induction for different time point, extracted the total RNAs (TRIzol™ Reagent, Invitrogen), and reversed to cDNAs (Thermo, USA). Quantitative uorescence detection was performed using 2X ChamQ Universal SYBR Master Mix (Vazyme, China). ABI 7500 was applied for qRT-PCR.

Western blotting
The proteins were isolated by 10% SDS-PAGE gel, which transferred with PVDF membranes (Millipore). After incubation with GAPDH, ALP, OCN (Huabio, China) primary antibodies overnight, the membranes were washed for 5 times with TBST (2‰ Tween). Then, the HRP-conjugated secondary antibody was incubated for 1h. Finally, an Super ECL kit (biosharp, China) was used to detect the protein bands. The Tanon-5200 was applied for exposure.

Microarray Assays
Total RNA were extracted using Trizol™ reagent and used to build a miRNA library. The library quality was determined by Agilent 2100 Bioanalyzer. It was eventually sequenced on a Illumina NextSeq 500 sequencer.

Bioinformatics Analysis
Edge R was used to analyze the differential expression of miRNA between groups. Based on the database of TargetScan7.1 and mirdbV6, target genes were screened for the Top 10 miRNA up and down regulated in signi cant difference. Then the function enrichment analysis of target genes was performed by Gene Ontology, including Molecular function, Cellular component and Biological process. The miRNA-mRNA networks were constructed using Cytoscape.

Statistical Analysis
Three independent repeated experiments were used to perform and analysis related results (mean ± SD). The T test was used to analyze differences between groups. P values < 0.05 were considered signi cant.

Results
The osteogenic differentiation of hPMSC was promoted by HAG Compared with autograft or allograft, biomaterials have become an indispensable bone material source in the eld of bone regeneration due to their non-immunogenicity and accessibility. Among various bone materials, hydroxyapatite (HA) is the most commonly used one because of the following advantages, the same mineral composition as natural bone, good biocompatibility, bone conductivity and mechanical stability. However, the interaction between scaffolds and cells is in uenced by various characteristics and morphology of scaffold surface, including chemical composition, electric charge, microstructure, and so on 4,5 . In previous work, on the basis of porous hydroxyapatite scaffold, we prepared a new scaffold material HAG through integrates microchannel structure. Moreover, the HAG promoted the osteogenic response of hPMSCs than HA 18 . Consistently, in this study, we repeated the osteogenic effect of HAG in hPMSCs (Fig. 1). The properties of hPMSCs were detected by ow cytometry. The positive surface marker CD44 and CD90 were highly expressed (97.08%), while CD45 (0.10%) was hardly expressed in the cells (Fig. 1A). The porous structure (as the arrow shown, left) and a 25-30 µm groove structure (as the arrow shown, right) on HAG scaffolds surface as shown in Fig. 1B. After osteogenesis induced for 7, 14 days, the expression of ALP, BMP2, Ocn mRNA were quanti ed by qPCR, and the protein levels of ALP, OCN were detected at 7d (Fig. 1C,D) Alizarin Red staning was used to detect the mineralized nodules 3 weeks after induction (Fig. 1E). The results reproduced the positive effect of HAG scaffolds osteogenic differentiation in hPMSC.
The miRNA expression in HAG-hPMSCs and control hPMSCs were differential In recent years, several studies revealed that miRNAs participated the process of osteogenic differentiation through targeting mRNA sequences of different genes related to osteogenesis [19][20][21][22] . However, roles of miRNAs in bone formation by HAG scaffolds mediated remain unclear. Thus, in order to uncover that how HAG scaffolds promoted osteogenic differentiation of hPMSCs, microarray assay was used to analyse the miRNA expression pro les between HAG-hPMSCs and hPMSCs. We observed 45 mature miRNAs signi cantly differentially expressed in HAG-hPMSCs group (FC ≥ 2, Fig. 2A). Red represented highly expressed genes, while green represents the down-regulated genes, and the darker the color, the more distinct the expression difference. Compared with differentiated hPMSCs, the expressions of 16 miRNAs in HAG-hPMSCs were up-regulated and 29 were down-regulated, as shown in the volcano plot (FC ≥ 2, Fig. 2B).
Pathway and functional analysis of differential miRNAs The differential expression of miRNAs in HAG-hPMSCs group may indicate that miRNA plays an important role in the process of osteogenic differentiation promoted by HAG. Therefore, the gene ontology enrichment analyses, including BP (biological process), CC (cellular component), and MF (molecular function) by cluster Pro ler in R software, were performed to explore the biological functions may be regulated by differential miRNAs. The statistically signi cant results (p values < 0.05) showed that these miRNAs participated in the regulation of cell metabolic (p value = 2.58E-12), cell junction (p value = 6.45E-08), cell development (p value = 1.82E-11), cell differentiation (p value = 4.04E-07), and signal transduction (p value = 8.12E-12) (Fig. 3A). Importantly, the process of osteogenic differentiation is closely related to these functions.
Furthermore, KEGG pathway analysis was used to understand the signaling pathways in which these differentially expressed miRNAs were involved in regulation. The results indicated that genes in multiple signaling pathways, including axon guidance (p value = 5.02E-05), MAPK signaling (p value = 6.65E-07), and the TGF-β signaling pathway (p value = 5.7E-04), highly associated with osteogenic differentiation are potential target genes for these differentially expressed miRNAs (Fig. 3B). It suggested that HAG may achieve its osteogenic effect by promoting differential expression of these miRNAs.

qRT-PCR validation of miRNA expression and miRNA-mRNA network analysis
According to the results of microarray analyses, the 6 miRNAs (miR-210-3p;miR-146a-5p; miR-483-5p; miR-3615 miR-125b-2-3p miR-145-5p), with the most signi cant expression differences were veri ed by qPCR. Consistently, the expression of miR-210-3p, miR-146a-5p, miR-483-5p were obviously increased, while miR-3615, miR-125b-2-3p and miR-145-5p were signi cantly decreased (FC > 2, p < 0.05) (Figu.4A). As known, miRAN achieve its regulatory effect by pairing with the complementary sequence of the target gene and degrading it or preventing its translation. More importantly, a miRNA can target multiple mRNA and the same mRNA can be regulated by multiple miRNAs 23 . Thus, TargetScan7.1 and mirdbV6 were used to predict the target genes of the above 6 miRNAs. Moreover, the miRNA-mRNA networks were constructed by Cytoscape (Fig. 4B). As shown in Fig. 4B, the miRNAs can target hundreds of mRNA, while the mRNA of BRCC3 can be recognized by different miRNA. More importantly, a great quantity of target mRNA were to be involved in osteogenic differentiation related pathways. For instance, the Brain-Derived Neurotrophic Factor (BDNF), a well-known growth factor of the neurotrophin family, is involved in regulating the growth, survival of neurons and angiogenesis 24  Overexpression of miR146a-5p promoted the osteogenic differentiation of hPMSCs To verify the effect of differentially expressed miRNA on osteoblastic differentiation, we transfected the miR146a-5p mimics into the hPMSCs. As shown in Fig. 5A and B, the mimics were transfected into cells successfully. Furthermore, we observed that the expression of osteogenic genes were increased when miR146a-5p were overexpressed (Fig. 5C). Meanwhile, following osteogenic differentiation, the hPMSCs transfected with miR-146a-5p mimics showed a higher intensity of ALP staining and formed a larger number of calci ed nodules (Fig. 6D, E). Therefore, the differentially expressed miRNAs may play an important role in the process of HAG promoting osteogenic differentiation.

Discussion And Conclusion
How to solve different conditions of bone defects is a major clinical challenge 28 . Autograft is the ideal way to repair bone defects 29 . However, due to the limited source, the high infection rate in donor sites, and increases the pain burden of patients made its clinical application rate low. With the rapid development of the eld of biomaterials and the advantages of good biocompatibility and no immunogenicity, bone biomaterials are the most widely material for bone repair. Among them, hydroxyapatite, which has the same chemical composition as natural bone, long-term biocompatibility and interacts well with soft tissue or bone in vivo, is widely used to prepare various bone materials [30][31][32] . Furthermore, the micros structure of the material surface is essential for cell adhesion, protein release, and interaction between the material and the cell 33 . Therefore, hydroxyapatite scaffold materials with porous 750-900 µm and microchannel structure of 25-30 µm (HAG) were prepared in previous work. Importantly, the osteogenic capacity of HAG in vivo and vitro has been demonstrated 18 . However, the mechanism which the HAG promotes osteogenic differentiation remains unclear.
More and more studies have revealed that miRNAs expressed in the various stages of bone formation, including the differentiation of bone marrow stromal stem cells into osteogenic precursor and the further development of precursor cells into mature osteoblasts. During these processes, they may play either a positive or negative role, including miR-96, miR-124, and miR-199a and so on [34][35][36][37][38] . However, the role of speci c biological scaffold material-related miRNAs in bone formation remains unclear. In this work, we tried to explore the osteogenic mechanisms of HAG scaffolds, in particular the miRNA involved. Interestingly, we did detect differential expression of miRNAs in the HAG-hPMSCs group. Moreover, the integration of pathway analysis, functional analysis and miRNA-mRNA cross network analysis showed that the miRNAs differentially expressed in HAG scaffolds group were highly correlated with axon guidance, MAPK, and TGF-beta signaling pathway. These signaling pathways are closely related to osteogenic differentiation.
Consequently, our results indicated that the HAG scaffolds may promote the bone formation through regulating the differentially expressed of miRNAs. In details, we successfully uncovered the statistically signi cant miRNAs in HAG-hPMSCs and constructed the miRNA-mRNA network through different bioinformatics analysis. However, we did not elucidate the speci c mechanism of action involved in osteogenic differentiation promoted by HAG. That's what we're working on right now. We believe that guring out the key target miRNA for HAG promoting osteogenic differentiation will lay the foundation for the preparation of bone scaffolds loaded with miRNA. Therefore, in this study, we rst analyzed the differential expression pro le of miRNAs in the HAG-mediated osteogenic differentiation, which is an indispensable step to achieve the nal goal.

Consent for publication
Not applicable.

Competing interests
There is no con ict of interest among the authors.

Figures
of ALP, BMP2 and Ocn protein level in hPMSCs with or without HAG, Gapdh for loading control. E. Alizarin Red-S staining images of hPMSC with or without HAG scaffolds for 21 days.

Figure 2
The miRNA differential expression pro les in HAG-hPMSC and hPMSC A. Different miRNAs heat map in hPMSC with or without HAG scaffolds. B. The volcano plot of miRNAs in HAG-hPMSC and hPMSC. The red means signi cant upregulation, while the green means signi cant downregulation (fold change ≥2, p value < 0.05). Notes: control: hPMSCs without HAG; HAW: hPMSC with HAG scaffolds.  The expressions of selected miRNAs were veri ed and the miRNA-mRNA interaction network was constructed A. The expression of chosen miRNAs were detected by qPCR in hPMSCs with or without HAG, normalized to U6 expression. B. miRNA-mRNA cross network were made by Cytoscape. These mRNAs represent the potential downstream targets of chosen miRNAs.

Figure 5
Overexpression of miR146a-5p promoted the osteogenic differentiation of hPMSCs A. The transfection of miR146a-5p mimics were veri ed by GFP expression. B. The expression of miR146a-5p were detected by qPCR in hPMSCs, normalized to U6 expression. C. Expression of ALP, Osx and Runx2 mRNA level during