Introduction

Mesenchymal stem cells (MSCs) are non-hematopoietic multipotent cells and used for bone tissue regeneration [1, 2]. MSCs can differentiate into osteoblastic, chondrogenic, and other lineages through different stimulating factors [3, 4]. Bone morphogenetic proteins (BMPs) are a group of transforming growth factor-β (TGF-β) and play an important role in embryonic development, growth, and differentiation [5]. BMP-9 (also known as growth/differentiation factor-2) has been shown to play a pivotal role in many physiological processes including neuronal and adipocyte differentiation [6]. In recent years, a series of experiments have demonstrated that BMP-9 is more osteogenic than BMP-2 and is not antagonized by noggin, which is a BMP antagonist [7, 8]. BMP-9 was reported to have a positive role in promoting the MSCs differentiation through Smad signaling pathway [9]. However, the mechanism for BMP-9-induced osteogenic differentiation was unclear. Gene expression microarray used modern research method and has been used to explore the gene network and identify the potential target gene.

The purpose of our study was to compare the gene expression of BMP-9-induced osteogenic differentiation of MSCs in the GEO database. Subsequently, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment pathway analyses were performed to identify the potential pathway of BMP-9-induced osteogenic differentiation of MSCs. Furthermore, protein-protein interaction (PPI) was performed to identify relevant genes and pathways and molecular mechanisms.

Materials and methods

Gene expression microarray data

The microarray data GSE48882 used in our study was downloaded from Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/). GSE48882 was based on the Affymetrix GPL339 platform (Affymetrix Mouse Expression 430A Array; Affymetrix; Thermo Fisher Scientific, Inc., Waltham, MA, USA). The GSE48882 dataset contained six samples, including three GFP1-induced MSC samples and three BMP-9-induced MSC samples.

Data processing and identification of DEGs

The original array data were performed background correction and quartile data normalization [10]. The raw data was downloaded and then put into the U-EDIT software to further analyses. Then, all of the differentially expressed genes (DEGs) between GFP-induced MSCs and BMP-9-induced MSCs were identified based on analyses GEO2R (https://www.ncbi.nlm.nih.gov/geo/geo2r/) and bioconductor package limma [11]. The DEGs between the GFP-induced MSC samples and three BMP-9-induced MSC samples were selected (P < 0.05).

GO enrichment and KEGG pathway analysis of the DEGs

The DEGs was entered into the online software Database for Annotation, Visualization and Integrated Discovery (DAVID, https://david.ncifcrf.gov/) [12]. The DEGs were classified into three functional groups: molecular function (MF) group, biological process (BP) group, and cellular component (CC) group.

DEGs functional and signaling pathway enrichment were conducted using online websites of KEGG (http://www.genome.jp/) and P < 0.05 was considered to indicate a statistically significant difference [13].

Construction of the PPI Network of DEGs

We downloaded the comprehensive interaction information of BMP-9-induced osteogenic differentiation of MSCs from the Search Tool for the Retrieval of Interacting Genes (STRING) database (http://www.string-db.org/) [14]. A combined score > 0.4 was identified as experimentally validated interactions. Then, we generate network and node file and analyzed using Cytoscape software (version 3.5.1; www.cytoscape.org). P < 0.05 was considered to be a statistically significant difference [15].

Results

Identification of DEGs

P < 0.05, logFC (fold control) > 2.0 or logFC < − 2.0 was used in current study, and we listed the first tenth DEGs in Table 1. And the samples were normalized and results were shown in Fig. 1. Among the DEGs, 1029 genes (52.31%) and 938 genes (47.69%) were upregulated and downregulated, respectively. Then, we draw a volcano plot and can be seen in Fig. 2. The top 50 upregulated and downregulated DEGs were selected to generate the heatmap and is shown in Fig. 3.

Table 1 The differentially expressed genes (DEGs)
Full size table
Fig. 1
figure 1

Box plots for expression data before normalization (a) and after normalization (b). The horizontal axis is the name of samples while the vertical axis stands for the values of expression. The black line in the cassette is the median of data in each group, which represents the degree of normalization. The black line in the right figure was almost on the same line, indicating an excellent degree of normalization

Full size image
Fig. 2
figure 2

The volcano plot of differentially expressed genes. The abscissa is − log10 (P.Value) and the ordinates is logFC. The red dots stand for the up-expressed genes, green dots stand for down-expressed gene while the black dots represent genes not differentially expressed

Full size image
Fig. 3
figure 3

The heat map of the differentially expressed genes (top 50 upregulated and downregulated genes). Red, upregulation; blue, downregulation

Full size image

GO term enrichment analysis

DAVID was used to carry out a gene ontology (GO) function enrichment for DEGs. Results show that the DEGs were significantly enriched in CC, including cytoplasm, nucleus, and extracellular exosome (Fig. 4). For MF, the DEGs were enriched in protein binding and poly(A) RNA binding (Fig. 4).

Fig. 4
figure 4

Go term enrichment analysis of the different genes

Full size image

KEGG pathway analysis

Results show that the DEGs were enriched in protein processing in endoplasmic reticulum, protein export, RNA transport, apoptosis, Alzheimer’s disease, N-Glycan biosynthesis, Valine, leucine and isoleucine degradation, TNF signaling pathway, Toxoplasmosis, p53 signaling pathway, and AGE-RACE signaling pathway in diabetic complications (Figs. 5 and 6).

Fig. 5
figure 5

Barplot analysis of the Kegg pathway

Full size image
Fig. 6
figure 6

Dotplot analysis of the Kegg pathway

Full size image

PPI network of the DEGs and core genes in the PPI network

As shown in Fig. 7, the PPI network contained 197 nodes and 272 edges. The top 10 high-degree hub nodes included Hspa5, P4hb, Sec61a1, Smarca2, Pdia3, Dnajc3, Hyou1, Smad7, Derl1, and Surf4 (Fig. 8). Further analyses of the DEGs, Hspa5’s node degree score was 108 and identified as the highest node degree (Figs. 8 and 9).

Fig. 7
figure 7

PPI network of BMP-9 induced osteogenic MSCs genes. The circle stands for genes and the line indicates the interactions among genes. The interior of the circle represents the structure of proteins. The color of the line provides evidence of the different interactions among proteins. (A red line indicates the presence of fusion evidence; a green line, neighborhood evidence; a blue line, concurrence evidence; a purple line, experimental evidence; a yellow line, text mining evidence; a light blue line, database evidence; a black line, coexpression evidence)

Full size image
Fig. 8
figure 8

Histogram of numbers of genes adjacent to interaction network

Full size image
Fig. 9
figure 9

Interaction network of the proteins

Full size image

Discussion

In the clinical application of rhBMP-2, adverse events such as tissue inflammation and bone resorption perplexed the physician [16,17,18]. Therefore, there is a crying need for the development of an equivalent effective and more safe bone regenerative therapies for bone regeneration. Recently, we attached importance to the rhBMP-9 since rhBMP-9 possesses different properties than rhBMP-2 [19,20,21]. Its precursor protein shares 50–55% amino acid sequence identity with BMP 2, 4, 5, 6, and 7. However, the present knowledge of the molecular mechanism of BMP-9-induced osteogenic differentiation of MSCs remains insufficient.

The present study performed a comprehensive analysis and built a gene interaction network based on the gene expression profiles (GSE48882) comprising three GFP-MSC samples and three BMP-9-MSC samples. The results of the analysis demonstrated that 1967 genes were identified as DEGs. Moreover, GO and KEGG pathway analyses were performed to find the interactions of DEGs. Combining with the PPI network, key potential genes and pathways may be associated with BMP-9-induced osteogenic differentiation of MSCs.

The results of GO analyses indicated cytoplasm, nucleus and extracellular exosome, protein binding and poly(A) and RNA binding. Nakamura et al. [22] revealed that BMP-9 provided comparable new bone formation and with less adipose tissues compared to BMP-2 in calvarial critical-size defects model. Recently, Sreekumar et al. [23] found that rhBMP9 induced the highest levels of osteoblast activity when compared to the clinically utilized rhBMP2 and rhBMP7. Compared with rhBMP2 and rhBMP7, rhBMP9 significantly increased osteogenic activity (ALP activity and Smad nuclear translocation).

Lin et al. [24] also found that BMP9-induced osteogenic differentiation through inhibition of the Wnt/β-catenin and P38 pathway in MSCs. And adding Wnt antagonist Dickkopf-1 (Dkk1) or P38 inhibitor SB203580 could negatively regulate BMP9-induced osteogenic differentiation. In current study, gene array indicated that BMP-9 has no effect on the Wnt/β-catenin pathway.

Chen et al. [25] found that BMP-9 induce osteogenic differentiation of MSCs through Smad pathway. When there was an overexpression of the BMP-9 in MSCs, Smad 1, Smad 7, and Smad 9 was significantly increased than controls. And insulin-like growth factor (IGF-1) could enhance BMP9-induced osteogenic differentiation in MSCs through Smad signaling pathway. The above result was similar with our results and found that BMP-9 could also increase the expression of Smad7. PPI network revealed that Smad 7 ranked the eighth of the high-degree hub nodes. Daigang et al. [26] found that BMP-9 stimulates the phosphorylation of Smad1/5/8 at a dose response manner. What is more, BMP-9 could significantly increase osteoblast differentiation genes such as Runx2, osteocalcin, and ALP.

Hspa5 was the heat shock protein and plays an important role in protein folding and transport processes. Hspa5 plays a role in facilitating the assembly of multimeric protein complexes inside the endoplasmic reticulum. And thus, BMP-9 may through protein folding and transport and promote osteogenic differentiation of MSCs. Sharff et al. [27] revealed that Hey1, through its interplay with Runx2, may play an important role in regulating BMP9-induced osteoblast lineage differentiation of MSCs. Zhou et al. [28] revealed that BMP9 reduces bone loss in ovariectomized mice by dual regulation of bone remodeling. BMP9 exerts on bone remodeling by promoting bone anabolic activity and inhibiting osteoclast differentiation in ovariectomy (OVX) mice. Hey1 and Hspa5 interact with each other and promote the expression of MSCs. And from the PPI analysis, we found that HIF-1a also plays a central role in the osteogenic differentiation of MSCs. As we know, HIF-1a can stimulate the VEGF and stimulate the angiogenesis [29,30,31].

Several limitations existed in our study: (1) we did not verify the key gene and pathways through PCR and Western-blot method; (2) key miRNA and lncRNA were not further analyzed and should be performed in future research; and (3) optimal dose of BMP-9 was not explored in this gene expression microarray. Future studies should be focused on the specific mechanism BMP-9-induced osteogenic differentiation of MSCs. What is more, administration with small interference RNA to further confirms our conclusion.

Conclusion

In conclusion, our study provides a comprehensive bioinformatics analysis of key genes and pathways of BMP-9 induced osteogenic differentiation of MSCs. Their related GO terms found that cytoplasm and nucleus may play important roles in BMP-9-induced osteogenic differentiation of MSCs and have the potential to be used as an potential agent for MSCs osteogenic differentiation. Further genetic and experimental studies with larger sample size are needed to confirm our results.