Insight into the maintenance of odontogenic potential in mouse dental mesenchymal cells based on transcriptomic analysis

Cell culture

All procedures involving animals described in the present study were reviewed and approved by the Animal Care and Use Committee of Guangzhou Institutes of Health and the Peking University in China (permit Number: CMU-B20100106). The lower first molar tooth germs were dissected from the mandibles of E14.5 mice with fine needles and treated with 0.75 mg/ml dispase (Becton, Dickinson and Co., Franklin Lakes, NJ, USA) for 40 min at 37 °C to facilitate separation from dental epithelium. The mesenchymal tissues were then incubated in 0.25% trypsin (Sigma-Aldrich, St. Louis, MO, USA) at 37 °C for 1 min and dissociated with gentle pipetting. The cells were then filtered through a 40 µm cell sieve, resuspended, and cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, Grand Island, NY, USA) plus 10% fetal bovine serum (FBS; Gibco) (Duailibi et al., 2004; Jiang et al., 2014; Yamazaki et al., 2007; Zhao et al., 2014). To maintain the odontogenic potential of mDMCs, supplementation of VEGFα (Sino Biological Inc, Beijing, China), BMP2 (Sigma-Aldrich), and BMP7 (Sigma-Aldrich) were added into the medium.

Quantitative reverse transcription PCR (qRT-PCR)

Total RNA was extracted with TRIzol (Invitrogen, Carlsbad, CA, USA) and reverse transcribed using an RT-PCR kit (TaKaRa Bio, Otsu, Japan) (Zheng et al., 2014). Quantitative PCR was performed in a Thermal CyclerDice7™ Real Time System with SYBR Green Premix EXTaq™ (TaKaRa Bio). The primers used are listed in Table 1. The relative expression of genes was calculated using the 2^−ΔCt method.

Tissue recombination and subrenal culture

The mDMCs were harvested at indicated time points using 0.25% trypsin (Sigma) and spun down to make cell pellets. The cell pellets were cultured at 37 °C for 2–3 h and then recombined with freshly isolated E14.5 dental epithelium. The recombinants were further cultured in vitro for 24 h prior to subrenal culture in adult ICR male mice. The host mice were sacrificed 3 weeks later to harvest the grafted tissue. Grafts were then fixed and subjected to H&E staining for histological analysis.

RNA isolation and sequencing

To determine the transcriptional regulation following in vitro culture, mDMCs from the developing molars in E14.5 mouse embryos were isolated and designated as P0. The cells were then subcultured in standard medium and passaged once they reached 90% confluence, with the first-passage culture designated as P1 and the second-passage culture as P2. P0, P1 and P2 cells were harvested using 0.25% trypsin (Sigma-Aldrich). Total RNA was extracted using the RNeasy Mini Kit and RNase-Free DNase Set according to the manufacturer’s protocol (Qiagen GmbH, Hilden, Germany). The purity and quantity of RNA were assessed using a spectrophotometer (model 8453; Agilent, Santa Clara, CA, USA). RNA libraries for samples were prepared according to instructions for the Illumina TruSeq™ RNA Sample Prep Kit. Sequencing was performed on an Illumina Hiseq™ 2000 (Illumina, San Diego, CA, USA) in duplicate.

Bioinformatics analysis

Sequenced reads were mapped to the mouse transcriptome (mm10, Ensembl v73) and then aligned using bowtie (v1.0.1) and RSEM (v1.2.12), as described previously (Hutchins, Takahashi & Miranda-Saavedra, 2015; Li & Dewey, 2011). EDASeq (v1.11.0) was used for GC normalization of samples, and differential expression was called using DESeq2 (v1.12.0). The fold change cut-off was set at twofold and p-value < 0.05 were considered to be statistically significant. The Database for Annotation, Visualization and Integrated Discovery (DAVID) was used to determine overrepresented GO categories and KEGG pathways using the entire mouse transcriptome as the background gene set. Protein-protein interaction network of DEGs was constructed using the STRING database (http://string-db.org/). Ingenuity pathway analysis (Ingenuity Systems, Redwood City, CA, USA) was used to analyze the upstream regulators of the differential expression between the samples (Razali et al., 2015).

Statistical analysis

Statistical analyses were performed using SPSS PASW version 18. Experiments with cell cultures were performed at least in triplicate and data are expressed as mean ± standard deviation (SD). The differences among groups were analyzed using one-way ANOVA. A two-tailed p-value < 0.05 was considered to be statistically significant.

Odontogenic potential of cultured mouse dental mesenchymal cells

mDMCs are typically cultured in Eagle’s minimum essential medium plus fetal bovine serum and antibiotics (Jiang et al., 2014; Keller et al., 2011; Zhao et al., 2014). mDMCs showed an atypical round- or spindle-shaped fibroblast-like morphology with a higher nuclear-to-cytoplasmic ratio, indicating their primitive character. Although the cells retained fibroblastic features up to the second passage, they exhibited a senescent phenotype with increased cytoplasm and augmented volume (Fig. 2A).

The odontogenic potential of P0, P1, and P2 cells were examined to identify the impact of the culturing. Although freshly isolated mDMCs developed into well-structured teeth when recombined with embryonic dental epithelium, cultured mDMCs failed to support tooth development (Fig. 2B). Moreover, the expression of genes that are essential for tooth development were analyzed using qRT-PCR. The expression of Msx1, Pax9, and Lhx6 was significantly reduced in P1 and P2 cells compared with P0 cells (Fig. 2C). The expression of Fgf3 and Bmp2 in cultured mDMCs was reduced compared with the P0 cells, but the expression of Bmp4 was not significantly different between the P0 cells and cultured mDMCs (Fig. 2D).

Overview of the mouse dental mesenchymal cells’ transcriptome

To obtain a global view of genes regulating the loss of odontogenic potential, total mRNA of P0, P1, and P2 cells was extracted and sequenced. After data correction, 11,340 transcripts could be matched exactly to known mouse Ensemble transcripts. A total of 9,815 genes were shared among P0, P1 and P2 cells, whereas 563 genes were expressed exclusively in P0 cells (Fig. 3A). P0 cells that were not exposed to in vitro culture conditions showed a striking separation from P1 and P2 cells (Fig. 3B; Fig. S1). The transcriptional disparity between freshly isolated and cultured mDMCs is consistent with their phenotypic differences. Differential expression analysis revealed that in vitro expansion of mDMCs promoted the selective overexpression of 859 genes, whereas 763 genes were downregulated in P1 cells (Fig. 3C). Comparison of the transcriptomes of P0 and P2 cells revealed that 1,004 genes were upregulated and 948 were downregulated (Fig. 3C). In contrast, 13 genes were upregulated and two genes were downregulated in P1 compared with P2 cells (Fig. 3C). These results suggested that the transcriptome of mDMCs was significantly influenced by in vitro culture conditions. In addition, the expression levels of Msx1, Lhx6, Pax9, Bmp4, Fgf10, Bmp2, and Fgf3 were comparable when analyzed with RNA-seq and qRT-PCR (Fig. 3D).

Gene ontology analysis of differentially expressed genes

Gene ontology (GO) analysis provides an intuitive and effective approach to understand the function of genes in three domains: biological processes, cellular components, and molecular functions. To understand the function of differentially expressed genes, GO analysis was conducted (Fig. S2) and a network diagram was created to illustrate the communication of differentially expressed genes in the enriched clusters of biological processes (Fig. 4). The network is composed of: (a) genes around the node Dnmt3b, namely Dnmt1, Ezh2, Uhrf1, and Rbl1, which participate in chromatin organization and transcriptional regulation; (b) genes around the node Runx2, namely Bmp2, Dlx1, Lhx6, and Sp7, involved in cell differentiation and skeletal system development; (c) genes around the node Fn1, namely Robo4, Tie1, Pecam1 and Col5a1, involved in cell motion and cell adhesion; and (d) genes that are not linked to any specific node, whose function is related to apoptosis (Tradd, Tnfrsf11b) and epithelial morphogenesis (Grem1). These various genes and their functions are expected because chromatin modification, cell motion, and cell adhesion are basic cellular functions involving the development of multiple organs.

Pathway analysis of the differentially expressed genes

Multiple pathways including fibroblast growth factor (FGF), bone morphogenetic protein (BMP), hedgehog (SHH), and Wnt signaling play critical roles in odontogenesis (Liu et al., 2013; Thesleff, Vaahtokari & Partanen, 1995); thus, the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis was used to determine the pathways involving the differentially expressed genes (Fig. 5A). Not surprisingly, genes showing a threefold or greater change in expression were related to the pathways regulating basic cellular activities (extracellular matrix-receptor interaction [ko04512], focal adhesion [ko04510], DNA replication [ko03030], and p53 signaling pathway [ko04115]). Notably, essential pathways regulating tooth development were also enriched, including the mitogen-activated protein kinase (MAPK) signaling pathway [ko04010], the transforming growth factor beta (TGFβ) signaling pathway [ko04350], and the Hedgehog signaling pathway [ko04350]. However, numerous members of these signaling pathways were reduced in cultured mDMCs (Fig. 5B), and it was difficult to identify the critical regulators that are responsible for the loss of odontogenic potential in cultured mDMCs.

An upstream analysis predicts which upstream regulators were most likely to be involved based on the current levels of gene expression detected by RNA-seq analysis. Numerous upstream regulators were related to the loss of odontogenic potential in cultured mDMCs. Specifically, the activities of Bdnf, Vegfα, Bmp2, and Bmp7 were predicted to be significantly inhibited, while Igf2, Fgf10, and Tgfb1 were more active in cultured mDMCs (Fig. 5C). Although Bdnf was slightly increased in cultured mDMCs, its receptor Ntrk2 was significantly decreased (Fig. 5B). The inhibition of Bdnf led to the downregulation of genes that promote odontogenesis, such as Egr1, Egr2, and Fos (File S1). Downregulation of Bmp2 and Bmp7 probably contributed to the loss of odontogenic potential via the inhibition of Runx2, Sp7, and Msx2 (File S1).

VEGFα, BMP2, and BMP7 restore the expression of downregulated genes

Given the pivotal roles of VEGFα, BMP2, and BMP7 in tooth development, we selected these proteins to increase the expression of odontogenic genes. Indeed, tooth mesenchyme-specific genes were upregulated including Msx1 and Pax9 in mDMCs treated with Bmp2 and Bmp7 (Figs. 6B and 6C). VEGFα slightly increased the expression of Msx1 and Pax9, but not that of Lhx6 (Fig. 6A). In parallel, VEGFα, BMP2, and BMP7 were adopted to maintain the odontogenic potential of mDMCs. Although well-structured tooth was not observed, dentin-like structures were present (Fig. 6D). Together, VEGFα, BMP2, and BMP7 promoted the expression of dental mesenchyme-specific genes and the formation of dentin-like structures.