The effect of melatonin on the mouse ameloblast-lineage cell line ALCs

Melatonin plays a critical role in promoting the proliferation of osteoblasts and the growth and development of dental papilla cells. However, the effect and mechanism of melatonin on the growth and development of ALCs still need to be explored. CCK8 assay was used for the evaluation of cell numbers. qRT-PCR was used to identify the differentially expressed genes in ALCs after melatonin treatment. The number and morphology of ALCs were investigated by confocal microscopy. Alkaline phosphatase assay and Alizarin red S staining were used for measuring mineralization. Then, we focused on observing the crucial factors of the signaling pathway by RNA-seq and qRT-PCR. Melatonin limited the cell number of ALCs in a dose-dependent manner and promoted the production of actin fibers. A high concentration of melatonin significantly promoted the mRNA levels of enamel matrix proteins and the formation of mineralized nodules. RNA-seq data showed that Wnt signaling pathway may be involved in the differentiation of ALCs under the influence of melatonin. This study suggests that melatonin plays a regulatory role in the cell number, differentiation, and mineralization of the ALCs, and then shows the relationship between the Wnt signaling pathway with the ALCs under melatonin.

Amelogenesis imperfecta (AI), a heterogeneous group of genetic dental rare diseases cause patients symptoms of eating difficulties and feeling pain, which is difficult to treat 1 . AI includes deficiencies in the amount of enamel and a hypoplastic, and/or hypomineralized enamel phenotype, which is caused by mutations in numerous genes 2 . For example, amelogenin X-linked (AMELX) contributing ~ 90% of the enamel matrix proteins (EMPs), is essential for enamel crystallite growth and prismatic or rod architecture on the microscale 3 . It is known that there are many allelic mutations in the AMELX gene causing different alterations in the protein and resulting in different phenotypes 4 . AMELX alone is a weak promoter of nucleation but the addition of enamelin (ENAM) promotes nucleation rates in a highly nonlinear, nonmonotonic manner, which reaches a sharp maximum at a ratio of 1:50 ENAM/AMELX 5 . Possibly only isolated ENAM but not ENAM oligomers can act as sites of enhanced nucleation 6 . Mutations in the ENAM gene result in generalized thin enamel hypoplasia or bands of pitted enamel 2 . Odontogenic, ameloblast associated (ODAM) is considered one of the regulators of enamel mineralization and crystal elongation 7 , and its expression in the dental junctional epithelium helps maintain the integrity of the junctional epithelium attachment 8,9 . As above, enamel formation is a highly regulated process, which may be affected by various negative effects 2 . Therefore, it is urgent and necessary to study the influencing factors of enamel formation.
Melatonin (MT) is mainly secreted by the pineal gland (PT) in the brain and is regulated by the suprachiasmatic nucleus (SCN). It is one of the neuroendogenous hormones affecting the biological cycle 10,11 . It not only regulates the periodic rhythm, body temperature 12 , and blood pressure 13 , but also plays a role as an anti-cancer agent 14 , and regulates endocrine system activities 15 . Melatonin produces a marked effect by activating the highaffinity melatonin receptor (MTNR) in target tissues. MTNR1A and MTNR1B are two subtypes of MTNR in humans and mammals.
In recent years, the biological role of melatonin in the oral cavity has been widely concerned 16,17 . Melatonin increases the stability of implant-bone osseointegration, strengthen the density of bone area, and reduces alveolar bone loss 18 . Meanwhile, melatonin promoted the differentiation of dental papilla cells and the expression of www.nature.com/scientificreports/ dentin sialoprotein (DSP) 19 , and induced the osteogenic differentiation of dental pulp mesenchymal stem cells 20,21 . According to the research report, MTNR is expressed in human tooth germ and mouse tooth epithelial cell line HAT-7 22 . When the SCN with high melatonin receptor expression in experimental animals was removed, the circadian growth and development of teeth immediately disappeared 23 . Therefore, we put forward the hypothesis for the first time that the rhythm change of melatonin and its receptor may be one of the mechanisms of tooth dysplasia 24 . To test the hypothesis, we found that after changing the circadian rhythm of BALB/c pregnant mice, the protein levels of MTNR1A and MTNR1B in neonatal mouse tooth germs changed in varying degrees. We injected a melatonin receptor antagonist into the peritoneum of BALB/c pregnant mice. It was found that the ameloblasts of mice at 7 and 10 days after birth were widely denatured, the degree of enamel mineralization was low, and the content of AMELX was significantly reduced 25 . The previous experimental results showed that melatonin and its receptor system were related to the periodic growth and development of enamel. However, we have not conducted in vitro experiments on melatonin-treated ameloblasts.
The ameloblast-lineage cell line (ALCs) was initially obtained from the organ of mandibular molar tooth germ of newborn (C57BL/6 J) mice by Akira Nakata et al. 26 . ALCs are a spontaneously immortalized ameloblast cell line and are thought to be in the early stages of maturation. They consist of enamel epithelioid structure formation, present ameloblast characteristics, and express specific enamel genes, such as Amelx, Enam and Odam. In addition, ALCs in the standard medium could form calcified nodules 26,27 .Thus, ALCs were proved to be a mature tool to study melatonin-medicated effect.
To explore the influencing factors of enamel development from the cellular level and the possible factors of enamel hypoplasia, we studied the effects of melatonin on the proliferation, differentiation, and mineralization in ALCs in vitro. The effects of Luzindole, a melatonin receptor antagonist, on ALCs were also studied. Then, we focused on observing the crucial factors of the signaling pathway.

Results
Effect of melatonin on the proliferation of ALCs. To evaluate the effect of melatonin on the cell number of ALCs, concentration gradient of 0, 10 -10 , 10 -8 , 10 -6 , 10 -4 and10 −3 M were set up, and the absorbance value at 450 nm was detected by the CCK-8 method. Melatonin inhibited the increase in ALCs' cell number in a timedependent (Fig. 1A) and dose-dependent manner (Fig. 1B). Effect of melatonin on the differentiation of ALCs. qRT-PCR was used to analyze the mRNA levels of EMPs in different concentrations of melatonin (Fig. 3). In the previous experiment, we explored the optimal concentration of melatonin on ALCs' differentiation at gradient of 10 -3 , 10 -4 , 10 -6 , 10 -8 , 10 −10 M, and found that 10 −4 M and 10 −3 M were suitable for the study of melatonin in ALCs. Therefore, 10 −4 M and 10 −3 M were selected for further study. In the experimental group, 10 −4 M melatonin concentration had no significant effect on the specific enamel genes and MTNRs after 5 days. However, the expression of Mtnr1a and Mtnr1b in 10 −3 M melatonin group showed an evident and similar change trend, both of which reached the peak on the third day. The www.nature.com/scientificreports/ expression of MTNRs was 70 times and 20 times higher than that of the control group (0 M MT), respectively. Amelx, Enam and Odam expression levels increased significantly in cells cultured in the medium containing 10 −3 M melatonin (P < 0.05). The expression of Amelx and MTNRs showed the similar trend, reaching a peak on the third day, and the expression level of Amelx was 50 times higher than that of the control group. The expression levels of Enam and Odam were 5 times and 1-2 times higher than those in cells in the standard medium, but the changing trend was different from that of MTNRs. As a result, melatonin has been shown to increase the mRNA expression of these specific enamel genes.
Effect of melatonin on the mineralization of ALCs. Alkaline phosphatase (ALP) activity was used to evaluate the mineralization of ALCs in different concentrations of melatonin medium (0, 1 × 10 -4 , 2 × 10 -4 , 5 × 10 -4 , 1 × 10 −3 M). The ALP activity of ALCs in the 1 × 10 −3 M and 5 × 10 −4 M melatonin mediums was significantly higher than that in the standard medium (P < 0.05). With the increase in melatonin concentration, the ALP activity of ALCs gradually increased, indicating an enhancement of mineralization potential (Fig. 4A). Alizarin red staining was used to evaluate mineralized nodule formation in ALCs exposed to different concentrations of melatonin medium and melatonin receptor antagonist. Only a few calcified nodules were stained in the control group (0 M MT). There were more calcified nodules in the 1 × 10 -4 , 2 × 10 -4 and 5 × 10 −4 M melatonin groups. After repeated experiments, dense and obvious calcified nodules were produced in the 1 × 10 −3 M melatonin medium. Compared with the control and low-concentration melatonin groups, the red staining became  www.nature.com/scientificreports/ melatonin receptor Mtnr1a and Mtnr1b. The results showed that melatonin promoted the mineralization of ALCs, and Luzindole inhibited the mineralization of ALCs (Fig. 4C).
Signaling pathway of melatonin in ALCs. Since we found that 10 −3 M melatonin exerted the most obvious impact on gene expression in ALCs after 3-day melatonin treatment, which was confirmed by qRT-PCR, we then performed high-throughput transcriptome sequencing of ALCs treated with or without 3-day melatonin treatment to reveal the mechanism of melatonin-medicated gene expression. The Venn diagram (Fig. 5A) showed 11,667 overlapping expressed genes between two groups, 626 differentially expressed genes in melatonin treatment (MT) group, and 772 differentially expressed genes in control (CTRL) group, respectively. After comparing the two groups according to the requirements of "Fold Change ≥ 2, significance level P < 0.001", the volcano map showed a total of 1354 differentially expressed genes (DEGs) were identified, of which there were 679 up-regulated genes and 675 down-regulated genes (Fig. 5B). These results indicated significant alterations in gene expression during differentiation of ALCs after melatonin treatment. However, we found the expression of Amelx and Enam was very low in ALCs, below the detection level when analyzed by RNA-seq. KEGG pathway enrichment analysis 28 showed 20 statistically significantly enriched DEG pathways (Fig. 5C). DEGs were largely enriched in the FoxO signaling pathway, Hedgehog signaling pathway, Wnt signaling pathway, NOD-like receptor signaling pathway and signaling pathways regulating pluripotency of stem cells.
The Wnt signaling pathway, closely related to tooth growth and development, attracted our attention. According to the results of RNA-seq, melatonin induced the expression of Wnt signaling pathway factors in ALCs at least twice as much as that in the control group (Fig. 5D). These genes were Wnt genes (Wnt1, Wnt4, Wnt5b, Wnt6, Wnt7b, Wnt10a and Wnt11), Frizzled family genes (Fzd1, Fzd3, Fzd4 and Fzd5), secreted frizzled-related genes (Sfrp2 and Sfrp4), and the cAMP response element-binding protein (CBP)-encoding gene. Among them, the expression of Sfrp2, Sfrp4, Fzd1, Fzd3, Fzd4 and Fzd5 genes in the melatonin group increased significantly, while that of Wnt4, Wnt5b, Wnt6, Wnt7b and Wnt10a genes decreased significantly.
The regulation of the Wnt signaling pathway factors was confirmed by qRT-PCR analysis (Fig. 5E), including the Wnt genes, conduction factor β-catenin, the transcription factor Lef, and the downstream factors c-Jun N-terminal kinase (c-Jun) and Cyclin D1. Compared with the control group, the expression levels of Wnt4 and Wnt6 in 10 −3 M melatonin were significantly decreased, which was consistent with the results of RNA-seq. Although the expression of β-catenin and Lef did not change significantly in RNA-seq, we also detected the expression of β-catenin and Lef because they are critical factors in the Wnt canonical pathway. However, the expression of β-catenin and Lef increased significantly in the experimental group, 27 times and 12 times higher, respectively. The downstream factor Cyclin D1 had no significant difference in expression level, while the expression level of the c-Jun gene was approximately 2 times that of the control group. The change in expression of these factors confirmed that the critical factors of the Wnt/β-catenin signaling pathway might play a role in the effect of melatonin-treated ALCs.

Discussion
AI is common in oral clinic, which usually causes the incidence of oral diseases such as caries and premature tooth loss and even seriously affect the self-esteem of patients 29 . Enamel development is a highly regulated process, which can be perturbed by many environmental conditions and genetic alterations 2 . It is of great clinical significance to further study the impacting factors in enamel development during which the pathological changes may cause or boost AI. In the field of molecular developmental biology, many literatures on melatonin in oral hard tissue have reported that melatonin is referred to tooth growth and development, such as regulating the cellular function of odontogenic cells 20 , affecting tooth regeneration by MT-induced DNA methylation machinery 30 . However, the effect of melatonin on ameloblasts remains unclear. This study aimed to research on the effect of melatonin on ameloblasts and explore the possible mechanisms.
Melatonin has been shown to modulate the proliferation of various cells effectively 31 . For example, 1 mM melatonin inhibits the proliferation of breast cancer MCF-7 32 and umbilical vein endothelial cells 33 in vitro. Melatonin (10 -6 to 10 −4 M) has no noticeable effect on the proliferation of human osteoblast cell line hFOB1. 19 34 , but melatonin (10 -12 to 10 −10 M) has an apparent inhibitory effect on the proliferation of human dental pulp cells or HDPCs 35 . All the evidence suggests that melatonin may have different sensitivity and specificity in inhibiting cell proliferation in vitro. Our study strongly confirmed the evidence that melatonin concentration of 10 −3 M inhibited the number of ALCs in a time-and concentration-dependent manner. Then, we found that ALCs were more evenly distributed and produced more pseudopodia in the melatonin medium than in the standard medium. To some extent, with the increase in melatonin concentration, myofilament proteins were more densely arranged and interwoven, but the number of cells had decreased simultaneously, which was similar to the results of the cell proliferation experiment. The changes in cell number and myofilament protein levels were not apparent  www.nature.com/scientificreports/ at 48 h, compared with 96 h. Therefore, we speculated that melatonin may promote actin fiber production and increase cytoskeleton construction time-and concentration-dependent. In our previous work, loss of circadian rhythm mainly inhibited enamel formation and Amelx expression in mice through the melatonin receptor 25 . Therefore, we hypothesized that melatonin might promote the amelogenic behavior of ALCs. According to the experimental results, melatonin receptors Mtnr1a and Mtnr1b are present and stably expressed in ALCs, suggesting that ameloblasts may be one of the targets of melatonin. We found that 10 −3 M melatonin significantly enhanced the expression of specific enamel genes in ALCs. The expression curve of Amelx mRNA was consistent with that of Mtnr1a and Mtnr1b, which peaked on the third day and then declined in the following days, suggesting the expression of Amelx in melatonin group may be associated with MTNRs. Melatonin stimulates the differentiation of enamel cells, which is consistent with previous experimental results in vivo. However, the expression curve of Enam and Odam was relatively stable within 5d. There was a significant difference compared with the control group, and it did not change with the melatonin receptor expression. This suggests that melatonin also promotes the expression of Enam and Odam, but it may not be related to the MTNRs. Basically, melatonin stimulates the differentiation of ALCs, which is consistent with previous experimental results in vivo.
In addition to regulating the expression of the specific enamel genes, melatonin can further manipulate the functional behaviors of ameloblasts. Given that ALCs can calcify spontaneously, melatonin medication will further induce ALCs calcification, especially high concentrations of melatonin (10 −3 M), as confirmed by microscopic observations of ALP synthesis and mineralized nodule formation. Melatonin can promote ALP activity in ALCs concentration-dependent. Meanwhile, the number of calcified nodules and the positive staining degree of Alizarin Red deepened with increased melatonin concentration. Luzindole, a melatonin receptor inhibitor, significantly inhibited the calcification of ALCs, which suggested that melatonin could regulate the calcification of ameloblasts through Mtnr1a and Mtnr1b.
In conclusion, melatonin reduces the number of ALCs and promotes the differentiation of ALCs. Therefore, it is informative to explore molecular mechanisms of melatonin in the growth and development of ALCs 36 . The Wnt signaling pathway can be divided into the β-catenin-dependent canonical and β-catenin-independent pathways. Previous studies have shown that Wnt/β-catenin signaling pathways are critical for the formation and development of hard tissue, and are involved in various stages of tooth morphogenesis, such as epithelial-mesenchymal interaction, tooth-specific cell proliferation, migration, and differentiation 37 . Various signaling factors in Wnt signaling pathway are expressed in the inner enamel epithelium and ameloblast 38 . Several studies have reported an association between the Wnt signaling pathway and melatonin. For example, melatonin was shown to enhance osteogenic differentiation of BMSCs via the HGF/PTEN/Wnt/β-catenin axis 39 or H19/miR-541-3p/Wnt/β-catenin axis 40 and the hypertrophic differentiation of MSC-derived chondrocytes through the Wnt signaling pathway 41 .
Thus, RNA-seq was applied to screen possible signaling pathway, and data showed that genes associated with Wnt signaling pathway changed significantly, indicating the melatonin-medicated activation of Wnt signaling pathway. Interestingly, in our research, the expression of the FZD receptor was significantly increased, and WNT ligand was generally decreased. This finding was consistent with the results of qRT-PCR, in which Wnt4 and Wnt6 were 7-and 11-fold down-regulated, respectively. Scholars believed that the inhibition of Wnt5a family members, including Wnt4, Wnt5a, and Wnt11, effectively activated the canonical pathway 42 . Therefore, we speculated that melatonin activates the canonical pathway by inhibiting the expression of Wnt5a family factors. In the canonical pathway, when the concentration of β-catenin increases, LEF in the nucleus is activated, which then transmits information through downstream target genes Cyclin D1, C-myc and c-Jun 43,44 , to promote the growth and development of the tooth enamel 45 .We then used qRT-PCR to detect the critical signaling molecules involved in the Wnt/β-catenin signaling pathway. Among them, the levels of β-catenin and LEF-1 were significantly upregulated to activate downstream factors. The expression of c-Jun decreased, and Cyclin D1 did not change significantly. These results indicated that Wnt signaling pathway could be significantly activated during melatonin treatment of ALCs. The signaling pathways activated by melatonin in enamel development need to be further explored and verified. At the same time, cell lines cannot simulate the environment in vivo, so future studies will focus on validating these hypotheses using vivo and animal models for pre-clinical trials.

Conclusion
Melatonin can inhibit the cell number of ALCs and promote differentiation and mineralization. Wnt/β-catenin signaling pathway may be associated with the differentiation of ALCs under melatonin induction. The findings of this study provide an experimental basis for the study of tissue engineering of tooth germ and the factors influencing enamel development abnormality.

Materials and methods
Cell culture. ALCs, were obtained from the Ninth People's Hospital Affiliated to Shanghai Jiaotong University School of Medicine (Shanghai, China). The appearance of epithelial cells indicates heterogeneous cell changes in ameloblasts, including differentiation, secretion, transition, and maturation 27 . ALCs were cultured in a standard Dulbecco's modified Eagle's medium (DMEM, high glucose, no phosphates, Invitrogen, USA) containing 10% fetal bovine serum (FBS, Invitrogen) and 1% penicillin/streptomycin complex (Invitrogen) at 37 °C in a humidified atmosphere under 5% CO 2 in the air. The medium was changed twice a week. When the cells reached 70% confluence, they were digested, counted, and transferred to 10 cm culture dishes.

Quantitative real-time reverse transcription-polymerase chain reaction (qRT-PCR). qRT-PCR
was used to detect the expression of melatonin receptors, such as the Mtnr1a and Mtnr1b, along with specific enamel genes (Amelx, Enam, and Odam). The cells were cultured in a 12-well plate at the density of 1 × 10 5 cells/ well. After 24 h, the medium was removed, and cells were treated with a standard medium containing various concentrations (0, 10 -4 , and 10 −3 M) of melatonin, and cultured for 1, 2, 3, 4 and 5d. Following the manufacturer's instructions, total RNA was extracted and isolated with Invitrogen TRIZOL reagents (Thermo Fisher Scientific, Germany  Mtnr1a  NM_026642  GAG TCT GTA GTT GTC GCA GTTG  GGC CTT TTT GGA AGG AAG AGGA  246   Mtnr1b  NM_205822  GTT ATG GCC CTG TTT GTG AGA  CAA TAG TTC CGG CCT CAA ATCAA  125   Amelx  NM_001081978 GAT GGC TGC ACC ACC AAA TC CTG AAG GGT GTG ACT CGG G  65   Enam  NM_017468  TGC AGA AAT CCG ACT TCT CCT  CAT CTG GAA TGG CAT GGC A  114   Odam  NM_027128  ATC AAT TTG GAT TCG CAC CACC AGT TGG ATC TAT CCC AGA AGTGA  www.nature.com/scientificreports/ ferent concentrations of melatonin medium (the same grouping conditions as the fifth paragraph), and the other half of the samples were exposed to (L-MT−), (L + MT−), (L-MT+), or (L + MT+) for 28 days. Among them, the control group utilized the standard medium (L-MT−), and the three treatment groups involved the standard medium containing luzindole (L + MT−), 10 -3 M melatonin (L-MT+), or both (L + MT+). After 28 days, the cells were fixed with 4% paraformaldehyde for 30 min and stained with alizarin red solution for 10 min at room temperature. After staining, the cells were cleaned with distilled water and air-dried. Images were taken using an inverted phase-contrast microscope (Leica Microsystems, Germany).
RNA sequencing and qRT-PCR. Shanghai Majorbio Bio-pharm Technology was responsible for RNA isolation, cDNA library construction, sequencing, and transcriptome data analysis. ALCs were seeded into 60 mm culture dishes at a density of 2 × 10 5 cells/well, cultured in standard medium or melatonin medium supplemented with 10 -3 M for 3 days. RNA was extracted with TRIzol. ND-2000 (NanoDrop Technologies) and 2100 Bioanalyzer (Agilent) methods were used to determine the quality of RNA samples, to ensure the use of qualified samples (OD 260/280 = 1.8-2.2, OD 260/230 ≥ 2.0, Rin ≥ 6.5, 28S:18S ≥ 1.0, > 2 μg) for transcriptome sequencing library. The RNA library prepared by TruSeq RNA sample preparation Kit from Illumina (San Diego, CA) was sequenced with an Illumina HiSeqxten/NovaSeq 6000 sequencer. The reading length of the sequence was PE 150. The clean reads were then compared with the reference genome using TopHat software (http:// tophat. cbcb. umd. edu/, version 2.0.0) 47 . The R statistical package software EdgeR was used for differential expression analysis. qRT-PCR was performed using the same cDNA of RNA sequencing. The primers used are listed in Table 1.

Statistical analysis.
All experiments were repeated three times, and each experiment was performed in four parallel wells. All data were analyzed using SPSS ver. 19.0 (IBM, Armonk, NY, USA) and expressed as mean ± standard deviation (SD). One-way analysis of variance (ANOVA) and Dunn's multiple comparison test were used to analyze statistical differences. When P < 0.05, the difference was considered statistically significant.

Data availability
RNA-Seq datasets have been deposited on Sequence Read Archive (SRA) with accession number: PRJNA 774042.