The Ano5^KI/KI mouse model carrying a Han GDD mutation was generated using CRISPR Cas9 genomic editing technology and by introducing a transformation of cysteine into tyrosine at codon 360 of Ano5. C57BL/6 female mice and KM mouse strains were used as embryo donors and pseudo-pregnant foster mothers, respectively, and were purchased from the Beijing Vital River Laboratory Animal, Co., Ltd. Genotyping of the Ano5 knock-in mice was verified by PCR amplification and Sanger sequencing (forward primer: 5’-GCTTAGGTCTTCTACATCGGGCTGT-3’ and reverse primer: 5’-ATCCCCATGAAGAGCGCAAAGAACA-3’). Details related to generation and genotype identification of the knock-in mouse model were previously published (29). Mice were housed in a pathogen-free environment under 12 hours light-dark cycles and given standard food ad libitum. All animal experimentation protocols were approved by the Institutional Animal Care and Use Committee of the Beijing Stomatological Hospital (the approval number: KQYY-201611-001).

mCOBs from Ano5^+/+ and Ano5^KI/KI mice were isolated from postnatal 24 hour old littermates and cultured in DMEM after adding 20% fetal bovine serum (FBS; Gibco, USA) until reaching 80% confluence. mCOBs at the third passage were used for osteoblast differentiation in osteogenic medium, as previously described. Mature osteoblasts represent mCOBs after 14 days of osteoblast differentiation. Cultivation medium was exchanged every two days.

2×10⁷ mCOBs/well from Ano5^+/+ (n=6) and Ano5^KI/KI mice (n=8) after 14 days of osteogenic induction were collected into centrifuge tube after washing three times with PBS buffer and all subsequent operations were carried out on ice. Cell extracts (500 μl comprising 80% methanol and internal standards, 20% H₂O) were added and vortexed for 3 minutes to achieve complete sample suspension. Samples were placed on liquid nitrogen for 5 minutes to achieve rapid freezing, then thawed on dry ice and ice for 5 min each, and then mixed by vortex for 2 minutes. The entire procedure was repeated in triplicate. Centrifugation was subsequently conducted for 10 min at 12,000 rpm/min and 4°C. Supernatants from samples (300 μl) were transferred into another centrifuge tube and incubated at -20°C for 30 min. Insoluble fragments were discarded by centrifugation at 12,000 rpm/min for 3 min and 4°C for 3 min and then 200 μl of supernatant was removed into the liner column of the sampling bottle for LC-MS analysis.

Chromatographic separation was firstly performed in a ThermoUltimate 3,000 system equipped with a Waters ACQUITY UPLC HSS T3 C18 column (100 × 2.1 mm, 1.8 μm, Waters) maintained at 40°C. Gradient elution analysis was carried out with 0.1% formic acid in water (A) and 0.1% formic acid in acetonitrile (B) at a flow rate of 0.4 ml/min. The injection volume of each sample was 2 μl. A gradient of water/solvent B (v/v) was used as follows: 0 min, 95%/5%; 11.0 min, 10%/90%; 12.0 min, 10%/90%; 12.1 min, 95%/5%; 14 min: 95%/5%. Liquid chromatography was then accomplished with a Waters ACQUITY UPLC BEH Amide column (100 × 2.1 mm, 1.7 μm, Waters) maintained at 40°C. Gradient elution was proceeded in water with 20 mM ammonium formate and 0.4% ammonia solution (A) and acetonitrile (B) at a flow rate of 0.4 ml/min. The injection volume of each sample also was 2 μl. A gradient of water/acetonitrile (v/v) was used as follows: 0 min, 10%/90%; 9.0 min, 40%/60%; 10.0 min, 60%/40%; 11.0 min, 60%/40%; 11.1 min, 10%/90%; 15.0 min, 10%/90%. The mass spectroscopy acquisition conditions included an electrospray ionization (ESI) temperature of 500°C, positive voltage of 5,500 V, a negative voltage of -4,500 V, the ion source gas I (GS I) at 55 psi, the GS II at 60 psi, the curtain gas (CUR) at 25 psi; and with the high collision-activated ionization (CAD) parameter. Full scan detection was performed in triple quadrupole (Qtrap) mode according to the optimized declustering potential (DP) and collision energy (CE).

Raw UPLC-MS/MS data were processed using Analyst (version 1.6.3). Integration and correction of chromatographic peaks were conducted using the MultiaQuant software package. Principal component analysis (PCA) was carried out using the base R software package (version 3.5.1). Intracellular metabolites significantly differential between groups were determined based on variable importance in projection (VIP) values ≥1 and p-values < 0.05. VIP values extracted from the orthogonal projection to latent structures discriminant analysis (OPLS-DA) results and associated score plots and permutation plots, as generated with the R package MetaboAnalystR (version 1.0.1). In order to avoid overfitting, a permutation test with 200 permutations was performed. Hierarchical cluster analysis (HCA) was conducted using the R package heatmaply (version 1.2.1) and ComplexHeatmap (version 2.7.1.1009). Pearson correlation coefficients between sample profiles were calculated using the cor function of R (version 3.5.1) and were visualized as heatmaps. Identified metabolites were annotated using the KEGG compound database and then mapped to the KEGG pathway database. Significantly enriched pathways were identified using the p-value from hypergeometric tests for a given set of metabolites.

Total RNAs of 5×10⁶ mCOBs from Ano5^+/+ and Ano5^KI/KI mice (n=3 per group) cultured after 14 days of osteogenic induction were extracted using the Trizol reagent (Ambion, Thermo Fisher Scientific, USA). The integrity of the total RNA was evaluated with an Agilent 2100 Bioanalyzer and with agarose gel electrophoresis (28S:18S values ≥1, RNA integrite number (RIN) values ≥7), along with purity and quantification analysis with a Nanodrop spectrophotometer (concentration ≥ 50 ng/μl, 260/280 absorbance ≥1.8, 28S/18S ≥1). mRNA was enriched using magnetic beads with Oligo (dT) and then fragmented. One-strand cDNA was synthesized from the mRNA templates using reverse transcription with random hexamers. Double-stranded cDNA was then synthesized by adding dNTPs and DNA polymerase I. AMPure XP beads were used to purify and select desired fragment size ranges of double-stranded cDNA, and finally PCR amplification was performed to construct cDNA libraries. Subsequently, qPCR was used to measure RNA concentrations (> 4 nM). Paired-end sequencing with a read length of 200-300 bp was then conducted on the Illumina HiSeqTM2500/4000 platform (Illumina Inc., San Diego, CA, United States).The experiments were carried out in triplicates and mCOBs cultures from at least three different mice.

Transcript expression levels are expressed as fragments per kilobase of exon model per million mapped reads (FPKM). Genes or transcripts with mean FPKM values > 1.0 are considered to be expressed in the group for statistical analysis. The HTSeq software package was used to analyze the expression levels in each sample using the UNION model. The Trimmomatic software program (version 0.33) was used to perform quality control of reads based on RNA-seq. Sequence reads were mapped to the reference genome using the STAR program (version 2.5.2b) software package. The DEGSeq 1.12.0 and DESeq package for R (version 1.10.1) were used to identify differentially expressed genes (DEGs) in Ano5^KI/KI mCOBs compared to Ano5^+/+ mCOBs. The negative binomial distribution and Benjamini–Hochberg methods were used to calculate p and FDR values, respectively. A total of 407 DEGs were filtered out under a reasonable threshold and effective criteria of fold change >1.2 and p _adj < 0.05. The GOseq (version 1.22) and KOBAS (version 2.0) packages for R were used to conduct Gene ontology (GO) and KEGG pathway enrichment to identify the physiological associations of DEGs. Significantly enriched items and pathways are identified using q values (corrected p-values) that were calculated by hypergeometric tests with BH correction.

mCOBs from Ano5^+/+ and Ano5^KI/KI mice were cultured on 24-well glass coverslips in osteogenic induction medium for 21 days. Samples (n=5 per group) were fixed with 2.5% glutaraldehyde and dehydrated with an ethanol gradient. The micromorphology of the nodules was then examined by SEM (TESCAN S9000X), while calcium and phosphate contents were analyzed by EDS using a MERLIN VP Compact system (Zeiss, Germany) in the UH-resolution scan mode under an acceleration voltage of 5 kV.

The proliferative ability of mCOBs was tested using a cell counting kit-8 (CCK-8) assay (Dojindo, Tokyo, Japan) according to the manufacturer’s instructions. Cells were plated at a density of 5 × 10³ cells/well in 96-well plates. Complete exchange of the medium was conducted at days 1, 3, and 4 using serum-free medium containing the CCK-8 reagent, followed by incubation with the cells at 37°C for 2 hours and OD measurements at 450 nm using a microplate reader. Meanwhile, the viability of mCOBs after 14 and 21 days of osteoblast differentiation was also measured. Cell proliferation rate was calculated as follows: cell proliferation rate = ((OD_treat-OD_Blank) - (OD_12h-OD_Blank))/(OD_12h-OD_Blank) x 100%.

Ano5^+/+ and Ano5^KI/KI mCOBs were seeded at a density of 3×10⁵ cells/well into 6-well plates and then cultured with α-MEM for 24 hours or with 14 day osteoblast differentiation. Cells were then digested and resuspended in a centrifuge tube and the supernatant was discarded after centrifugation at 1,000 g/minfor 5 min. After washing twice with cooled PBS, cells were fixed in 70% methyl alcohol at 4°C overnight. Cells were then incubated with 500 μL propidium iodide (PI) staining buffer for 30 min in the dark at 37°C. A BD Accuri C6 flow cytometer was applied to detect red fluorescence at an excitation wavelength of 488 nm, while ModFitLT V3.2 was utilized to analyze cell cycle distributions (G1, S, and G2).

Serum was collected from the retroorbital veins of 16 week old Ano5^+/+ and Ano5^KI/KI male mice (n=10) using a glass capillary. After allowing to naturally solidify at room temperature, supernatants were collected after centrifugation at 1,000 g for 20 min. In addition, cell-free extracts and corresponding culture supernatants from Ano5^+/+ and Ano5^KI/KI mCOBs were acquired at days 0 and 14. Calcitriol levels were determined using a calcitriol ELISA Kit (Sango, Shanghai, China) and FGF23 levels were examined with Mouse FGF23 ELISA Kit (Beyotime, Shanghai, China) according to the manufacturer’s instructions. OD values were immediately measured at 450 nm with a microplate reader.

mCOBs after 14 days of osteogenic induction were collected and lysed in RIPA buffer, as previously described (29). The protein concentrations in each sample were measured using Bradford assays with Coomassie brilliant blue G-250 (Bio-Rad, California, USA). A total of 20 μg protein was subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS‐PAGE). The SDS membranes were blocked with 5% nonfat milk for 1 hour and incubated with primary OCN antibodies overnight (DF12303, Affinity Biosciences, Beijing, China). ACTB (Abclonal, Wuhan, China) was examined as the housekeeping reference protein. After incubation for 1 hour with horseradish peroxidase (HRP)-conjugated anti-rabbit secondary antibody at a dilution of 1:5,000, band signals were detected using a Bio-Rad imaging system (Bio-Rad, USA) with NcmECL High (NCM Biotech, Jiangsu, China). The Image lab software program (Bio-Rad, USA) was utilized to perform relatively quantitative analysis of protein levels.

Total RNA was extracted using the TRIzol reagent (Ambion, Life Technologies, USA), and RNA quality was subsequently assessed using the Infinite M200 PRO NanoQuant absorbance microplate reader (TECAN, Chapel Hill, NC, USA). Reverse transcription was conducted with a SuperRT cDNA Synthesis Kit (CWbio, Beijing, China) and real-time quantitative reverse transcriptase-polymerase chain reactions (qRT-PCR) were performed using the Ultra SYBR Mixture with low ROX (CWbio), as previously described (29). Gene expression was calculated using the 2^−ΔΔCt method and Actb expression was used as the internal control. PCR primer sequences used in qPCRs are listed in Supplementary Table 1 and all assays were performed in triplicate.

A widely targeted metabolomics analysis was conducted using UPLC-MS/MS to explore critical biochemical compounds involved in GDD due to the Ano5^Cys360Tyr mutation. The stability and precision of the data were verified based on the typical base peak intensity chromatograms of the Ano5^+/+ and Ano5^KI/KI mCOBs cultured in osteogenic medium for 14 days ( Figure 1A ). The influence of the Ano5^Cys360Tyr mutation on the metabolic patterns of mature mCOBs was analyzed using OPLS-DA model and the result revealed that Ano5^KI/KI mCOBs metabolite profiles were clearly distinguished from the wildtype profiles ( Figure 1B ).

Significantly altered metabolites were identified via VIP values from the OPLS-DA analysis using the strict threshold of VIP > 1 and p-values < 0.05 from independent Student’s t-tests to identify significantly different metabolites. A total of 20 down-regulated and 22 up-regulated metabolites were screened out in Ano5^KI/KI mCOBs compared with wildtype cultures ( Figure 2A and Table 1 ). The 20 most differential metabolites based on |Log₂FoldChange| (|Log₂Fc|) values were further scrutinized and comprised calcitriol, carnitine C4:DC, 17a-estradiol, 17β-estradiol, N-acetyl-D-phenylalanine, cytosine, cytidine, DL-phenylmercapto uric acid, and cyclo (Pro-Leu) ( Figure S1 ). To visualize strengths of correlations across differential metabolites, Pearson correlation analysis was conducted and indicated that 31 metabolites were closely associated with calcitriol (|r| > 0.5) and 30 with carnitine C4:DC (|r| > 0.5), suggesting that these two metabolites considerably affected metabolic disturbances in Ano5^Cys360Tyr osteoblasts ( Figure S2 ).

The 42 differential metabolites belonged to ten classes including amino acid (21.4%), nucleotide (14.3%), organic acid (14.3%), benzene and substituted derivatives (9.5%), carbohydrates (9.5%), hormones and hormone related compounds (9.5%), alcohol and amines (7.1%), fatty acid (FA, 4.8%), glycerophospholipid (GP, 4.8%) and heterocyclic compound ( Figure 2B ). Collagen structure stability is closely associated with amino acid composition and sequence and thus, the most up-regulated amino acids potentially take part in enhanced bone formation in GDD (30). The metabolic pathways possibly influenced by Ano5^Cys360Tyr mutation were explored using metabolite enrichment analysis. A total of 44 metabolic pathways were altered, including pyrimidine metabolism, steroid biosynthesis, parathyroid hormone synthesis, secretion and action, endocrine and other factor-regulated calcium reabsorption, and mineral absorption ( Figure 2C ). The five most enriched metabolic pathways were highlighted in Table 2 , in which the activated metabolism of pyrimidine is positively associated with bone cell growth due to its capacity to provide energy and most metabolic substrates for living organisms (31). Additionally, obvious elevation of calcitriol, belonging to alcohol and amines class of metabolites, involved in endocrine and factor-regulated calcium reabsorption, which also plays a vital role in bone regeneration.

In order to explore the deep root of impaired Ano5^KI/KI osteoblast metabolism, the gene expression profiles of mCOBs after 14 days of osteogenic induction from Ano5^+/+ and Ano5^KI/KI mice were evaluated with RNA-seq. PCA revealed a clear distribution in gene expression profiles between Ano5^+/+ and Ano5^KI/KI mCOBs ( Figure 3A ). A total of 407 differentially expressed genes (DEGs) were screened out based on the criteria of fold change >1.2 and p-adj < 0.05. Volcano plots indicated that 239 DEGs were up-regulated and 168 were down-regulated compared to the wild type group ( Figure 3B ). In addition, the relative levels of DEGs in Ano5^+/+ and Ano5^KI/KI mCOBs were visualized with a heat map, which also revealed an obviously different clustering ( Figure 3C ). Based on log₂Fc, the top 25 up-regulated DEGs were highlighted in Table 3 . Mki67 is a standard marker of cellular proliferation and exhibited an obvious elevation. Cell Division Cycle 25C (Cdc25c) participates in regulating the G2/M transition and mediating DNA damage repair through activating the cyclin B1 (CACNB1)/CDK1 complex (27, 32, 33). Gad2 encodes glutamic acid decarboxylase that is responsible for catalyzing γ-aminobutric acid (GABA) production that has a positive effect on proliferation and osteogenic differentiation of mesenchymal stem cells (34).Simultaneously, Table 4 illustrated the top 25 down-regulated DEGs. Notably, Cytokine-like 1 (Cytl1) depletion exhibited a high bone mass phenotype due to enhanced osteogenesis and inhibited osteoclast activity (35). Sost is a potent negative regulator of bone formation by means of competitive interaction with the low‐density lipoprotein receptor‐related protein (LRP) 5/6 to antagonize WNT signaling (36, 37).

GO enrichment analysis was used to enrich the function of differentially expressed genes. Only four down-regulated categories were statistically enriched, including skeletal system development, tissue development, extracellular region, and extracellular matrix development ( Figure 4A ). While 251 significantly up-regulated terms were identified and mainly involved in system development, cell cycle process, and nuclear division ( Figure 4B ). Functional pathway analysis was also conducted based on KEGG enrichment, although only part of the ECM-receptor interaction signaling pathways exhibited significant down-regulation ( Figure 4C ). It is worth noting that prominent signaling pathways with elevated tendency were closely associated with osteogenic alteration of Ano5^KI/KI mCOBs, including ECM-receptor interaction, protein digestion and absorption, and calcium signaling pathway ( Figure 4D ).

In order to better understand the correction patterns of among significantly differentiated genes and metabolites, cluster analysis was performed ( Figure 5A ). Functional enrichment analysis revealed 26 shared pathways of DEGs and differential metabolites as illustrated by the Veen diagram ( Figure 5B ). The ten most enriched KEGG pathways were particularly scrutinized to reveal critical metabolic processes differing between Ano5^KI/KI and Ano5^+/+ mCOBs ( Figure 5C ). The analysis indicated that Slc8a1, accompanied by 17β-estradiol and calcitriol that participate in endocrine and other factor-regulated calcium reabsorption. Furthermore, β-alanine metabolism, biosynthesis of amino acids, and mineral absorption may be key processes associated with enhanced osteogensis in GDD.

GO analysis implied that cell cycle and nuclear division were abnormally activated in the Ano5^KI/KI mCOBs. Therefore, subsequent studies focused on the effects of the Ano5^Cys360Tyr mutation on cell proliferation. In addition to Mki67, cyclin A2 (Ccna2), an essential regulator of the G1/S and G2/M transition mediating binding and activating CDK2 (38) was up-regulated in Ano5^KI/KI mCOBs based on RNA-seq analysis. Likewise, Ccnb1 that forms a complex with CDK1 to promote the transition from the G2 phase of cell cycle to mitosis was also increased. qRT-PCR further confirmed that the p.Cys360Tyr mutation in Ano5 enhanced expression of those genes in both preosteoblasts at day 0 and mature mCOBs at day 14 after osteogenic induction ( Figure 6A ). Interestingly, the upward trend was more obvious at day 14. Subsequently, we are driven to disclose the proliferation ability of GDD-related osteoblasts and CCK8 indicated that either preosteoblasts or mature mCOBs from Ano5^KI/KI mice grew faster than those from wildtype mice ( Figures 6B, C ). To explore whether GDD-induced hyperproliferation was associated with cell cycle alternation, we detected the cell cycle distribution of mCOB from Ano5^+/+ and Ano5^KI/KI mice using flow cytometry to analyze cellular DNA content. mCOBs from Ano5^KI/KI mice displayed greater numbers of cells in the G2 phase compared with the wildtype group at day 0 ( Figure 6D ). Furthermore, Ano5^KI/KI mature osteoblasts exhibited decreased cell proportions in the G1 phase and significantly increased abundances of cells in the G2 phase (31.56 ± 0.94 vs. 21.74 ± 0.61) ( Figure 6E ). In conclusion, these observations suggested that increased osteogenesis in GDD can be partially explained by enhanced Ano5^KI/KI osteoblast proliferation.

GDD lesions are mainly characterized by disturbances in bone regeneration, wherein calcium is a structural inorganic component and plays a prominent role in bone matrix mineralization and osteoblast differentiation. Accordingly, subsequent studies are needed to evaluate calcium signaling in this context. Both transcriptomics and metabolomics analysis suggested that Ano5^Cys360Tyr mutation could influence calcium reabsorption and calcium signaling pathways. Further, RNA-seq data also indicated significant up-regulation of solute carrier family 8 member A1 (Slc8a1), calcium voltage-gated channel subunit alpha 1 C (Cacna1c), Htrt1, Grm1, Pde1b, Erbb3, Pdgfra, and Ptk2b combined with down-regulation of Chrm1, Adra1b, and Ptgfr; all resulting in disturbed calcium homeostasis ( Supplementary Table 2 ). Cacna1c is of particular interest, because a specific gain-of-function mutation (G406R) in its encoded protein is responsible for Timothy Syndrome (TS) that is clinically manifested by small teeth and dysmorphic facial features beyond classical cardiac arrhythmia (39). qRT-PCR indicated that the Ano5^Cys360Tyr mutation promoted the expression of Cacna1c at days 14 after osteoblast differentiation, although no significant difference was observed at day 0 ( Figure 7A ). In addition, Slc8a1 was upregulated in Ano5^Cys360Tyr mature mCOBs, which participates in calcium transport to the extracellular depending on sodium (Na⁺) concentration gradients. These trends were further verified by qRT-PCR analysis ( Figure 7B ). It has been reported that Slc8a1 is regulated by calcitriol also known as active 1,25(OH)₂D (1,25-dihydroxy vitamin D), was classically involved in bone homeostasis. An ELISA experiment was further performed and revealed obvious intracellular calcitriol elevation in Ano5^KI/KI mCOBs at days 0 and 14, even though the supernatants did not exhibit significant differences ( Figure 7C and Figure S3 ). Calcitriol also plays an important role in the systemic circulation of calcium and phosphate (40). Ano5^KI/KI mice manifested a massive up-regulation of serum calcitriol (217.523 ± 98.963 pmol/ml) compared with Ano5^+/+ mice (98.026 ± 53.298 pmol/ml) and this was combined with slight phosphate increases, which indicated high bone turnover ( Figures 7D, E ). Calcitriol is synthesized by the mitochondrial enzyme 25-hydroxyvitamin D-1a-hydroxylase encoded by CYP27B1 using hydroxylation of 25(OH)D₃ as the substrate (41, 42). Meaningfully, qRT-PCR showed that Cyp27b1 was significantly elevated in the kidneys of Ano5^KI/KI mice and a three-fold increase in Cyp27b1 mRNA accumulation was also observed in Ano5^Cys360Tyr mCOBs after 14 days of osteoblast differentiation, but not at day 0 ( Figures 7F, G ). Some studies have reported that calcitriol and Cyp27b1 cooperatively regulate OCN during osteogenic differentiation (43, 44). Consistently, the protein levels of OCN, the most abundant non-collagenous bone matrix protein that is specifically synthesized by osteoblasts, was elevated in differentiated Ano5^KI/KI mCOBs ( Figure 7H ).

The fact of calcium signaling alteration provides a reasonable explanation for lower levels of free calcium in the cytoplasm of Ano5^Cys360Tyr mCOBs compared to Ano5^+/+ mCOBs as we previously observed (29). In the present study, SEM-EDS were further performed to measure the contents of calcium and phosphorous in mineralized nodules. The ratio of calcium versus phosphorous was higher in Ano5^Cys360Tyr than Ano5^+/+ mCOBs ( Figure 7I ). This phenomenon implies that Ano5^Cys360Tyr mutation may perturb calcium balance between extracellular environments and the cytoplasm, leading to enhanced bone formation that can be attributed to up-regulation of calcitriol and Cyp27b1 expression.