Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Apr 1.
Published in final edited form as: J Cell Physiol. 2017 Oct 27;233(4):3476–3486. doi: 10.1002/jcp.26200

FAM20C regulates osteoblast behaviors and intracellular signaling pathways in a cell-autonomous manner

Chao Liu 1,2,*, Hua Zhang 2, Priyam Jani 2, Xiaofang Wang 2, Yongbo Lu 2, Nan Li 1, Jing Xiao 1, Chunlin Qin 2,*
PMCID: PMC5741497  NIHMSID: NIHMS908544  PMID: 28926103

Abstract

Recent studies indicate that Family with sequence similarity 20 member C (FAM20C) catalyzes the phosphorylation of secreted proteins, and participates in a variety of biological processes, including cell proliferation, migration, mineralization and phosphate homeostasis. To explore the local influences of FAM20C on osteoblast, Fam20c-deficient osteoblasts were generated by treating the immortalized Fam20cf/f osteoblasts with CMV-Cre-IRES-EGFP lentivirus. Compared with the normal Fam20cf/f osteoblasts, the expression of Bone sialoprotein (Bsp), Osteocalcin (Ocn), Fibroblast growth factor 23 (Fgf23) and transcription factors that promote osteoblast maturation were up-regulated in the Fam20c-deficient osteoblasts. In contrast, the expression of Dental matrix protein 1 (Dmp1), Dentin sialophosphoprotein (Dspp), Osteopontin (Opn), type I Collagen a 1 (Col1a1) and Alkine phosphatase (Alp) were down-regulated in the Fam20c-deficient cells. These alterations disclosed the primary regulation of Fam20c on gene expression. The Fam20c-deficient osteoblasts showed a remarkable reduction in the ability of forming mineralized nodules. However, supplements of extracellular matrix proteins extracted from the normal bone failed to rescue the reduced mineralization, suggesting that FAM20C may affect the biomineralization by the means more than local phosphorylation of extracellular matrix proteins and systemic phosphorus homeostasis. Moreover, although Fam20c deficiency had little impact on cell proliferation, it significantly reduced cell migration and lowered the levels of p-Smad1/5/8, p-Erk and p-p38, suggesting that the kinase activity of FAM20C might be essential to cell mobility and the activity of BMP ligands. In summary, these findings provide evidences that FAM20C may regulate osteoblast maturation, migration, mineralization and BMP signaling pathways in a cell-autonomous manner.

Keywords: FAM20C, mineralization, secretory proteins, phosphorylation, osteoblast

Introduction

FAM20C, previously known as Dentin matrix protein 4 (DMP4), was once regarded as an ECM component through immunohistochemical assay performed on mineralized tissues, but further investigations proved that FAM20C is an intracellular component enriched in Golgi apparatus (Hao et al., 2007; Ishikawa et al., 2012; Tagliabracci et al., 2012). At 2012, FAM20C was identified as the kinase phosphorylating the members of the small integrin-binding ligand, N-linked glycoproteins family (SIBLINGs, including DMP1, DSPP, BSP, OPN and MEPE [Matrix extracellular phosphoglycoprotein]), the major extracellular components regulating mineralization (Tagliabracci et al., 2012). The following studies indicated that FAM20C could specifically phosphorylate the serine residue at the Ser-X-Glu motif of secreted proteins (George and Veis, 2008;Tagliabracci et al., 2012; Tagliabracci et al., 2013; Tagliabracci et al., 2015). Latest study also predicted that in addition to the SIBLING proteins, the substrates of FAM20C mayinclude growth factors, components of lipid, calcium and phosphorus metabolism, neuropeptides, protease inhibitors and metalloprotease and so on, which make up the majority of secreted phosphoproteins (Tagliabracci et al., 2015). Therefore, the kinase activity of FAM20C was suggested to play crucial roles in various biological and pathological processes.

Inactivating mutations in FAM20C in humans lead to Raine Syndrome characterized by lethal or non-lethal skeletal disorders (Simpson et al., 2007; Fradin et al., 2011). Fam20c knock-out mice phenocopy the non-lethal cases of human Raine Syndrome, which manifests hypophosphataemic rickets (Wang et al., 2012a; Rafaelsen et al., 2013). Similar to the human Raine Syndrome with hypophosphataemic rickets, Fam20c-deficient mice exhibit a retarded growth, widened growth plates, thin and porous cortex of long bones, which are believed to be partially attributed to the elevated FGF23 and reduced serum phosphorus (Wang et al., 2012a). Recent study revealed the regulatory role of FAM20C in the degradation of circulating FGF23 (Tagliabracci et al., 2014). During the degradation process of FGF23, the proteolytic cleavage occurs at the peptide bond between Arg179 and Ser180 of FGF23 (Bergwitz and Jṻppner, 2010). The phosphorylation of Ser180 (located in the Ser180-Ala181-Glu182 motif) by FAM20C prevents the GalNac-T3-catalyzed O-glycosylation of Thr178. The O-glycosylation at Thr178 protects the proteolytic site of FGF23 from being exposed to the protease Furin and thus, prevents the degradation of FGF23 (Kato et al., 2006; Bergwitz et al., 2009). Therefore, the lack of phosphorylation at Ser180 of FGF23 in Fam20c-deficient models promotes the O-glycosylation at Thr178, which prevents FGF23 from being degraded and in turn, elevates plasma FGF23 concentration (Tagliabracci et al., 2014). Circulating FGF23 promotes the excretion of phosphorus by binding FGFR1 in the aid of Klotho to reduce sodium-phosphate co-transporters NaPi-2a/2c on renal proximal tubular cells (Kurosu et al., 2006; Farrow and White, 2010; Bhattacharyya et al., 2012). Although this study proposed a mechanism for the in vivo elevation of circulating FGF23, it could not explain the mechanism of Fgf23 mRNA elevation in the Fam20c-deficient osteoblasts and osteocytes (Wang et al., 2012a). It is also unclear whether the altered expression of bone related genes, such as Osterix (OSX), DMP1 and OCN, directly results from Fam20c inactivation or is secondary to the systemic feedbacks, such as the reduction of serum phosphorus and elevation of serum parathyroid hormone (Wang et al., 2012a). Moreover, since BMP ligands were speculated to be the substrates of FAM20C (Tagliabracci et al., 2015), it implies that FAM20C may directly regulate cell behaviors in a cell-autonomous manner by modifying paracrine or autocrine factors.

Previously, we reported the generation of immortalized Fam20cf/f osteoblast cell line (Liu et al., 2015). Taking the advantages of osteoblasts carrying Fam20c-floxed allele and Cre recombinase, we generated osteoblast cells devoid of Fam20c, which provides a tool excluding the systemic influences on the individual cell. By exploring the cell behaviors and gene expression of the Fam20c-deficient osteoblasts, we can distinguish the cell-autonomous regulation of Fam20c on osteoblast from the systemic influences in the in vivo models.

Materials and methods

Lentivirus infection and establishment of OB Fam20cKO osteoblast line

The immortalized Fam20cf/f osteoblast cell line (designated as “OB Fam20cf/f”) was generated in our previous study (Liu et al., 2015). When grown to 80% confluence, the OB Fam20cf/f osteoblasts were infected by the CMV-Cre-IRES-EGFP lentivirus following the manufacture’s protocol (GenTarget Inc. San Diego, CA), to create Fam20c-deficient osteoblasts (referred to “OB Fam20cKO”). After 24 hours of infection, the Fam20cf/f osteoblasts were re-plated in a low density to form separated single cell-derived clones. The selected clones were examined under the Olympus microscope for EGFP expression and DAPI counter staining. Only the clones in which all cells were EGFP positive after 10 passages were characterized in this study.

Polymerase Chain Reaction (PCR) and primers

Genotyping PCR was performed to confirm the removal of the floxed Fam20c alleles by Cre recombinase. The protocols and primer sequences distinguishing the floxed Fam20c allele from the wild type (WT) allele were described previously (Wang et al., 2012a). The null allele recombined from the floxed Fam20c as well as the Cre transgene were also identified as the previously described (Wang et al., 2012a).

Reverse transcription-PCR (RT-PCR) was performed to detect the transcripts from the floxed and recombined Fam20c alleles. Total mRNAs were extracted with the RNA extraction kit and reversely transcribed into cDNA with the RT-PCR Kit (Life Technologies, Inc., Grand Island, NY). Two set of primers were used to distinguish the floxed Fam20c alleles from the recombined Fam20c alleles. The first set of primers (Forward: 5’-TGCGGAGATCGCTGCCTTCC-3’; Backward: 5’-GCCACT GTCGTAGGGTGGCA-3’) that were designed to amplify a region from exon5 to exon 8 produced a band of 388 bp for the mRNA transcript from floxed Fam20c allele, and gave rise to no PCR band for the transcript of the recombined null allele. The second set of primers (Forward: 5’-GAGAGCAGGAGACGCCGCCT-3’; Backward: 5’-CCACCACACTGCTCAGCCCG-3’) amplifying a fragment from exon 5 to exon 11 gave rise to a 820-bp product for the mRNA of floxed Fam20c allele, and a 431-bp band for that of the recombined null allele.

Quantitative-PCR (Q-PCR) was employed to evaluate the effects of Fam20c loss on gene expression. The protocols and primers for the amplification of activating transcriptional factor 4 (Atf4), distal-less homeobox transcription factor 3 (Dlx3), runt-related transcription factor 2 (Runx2), osterix (Osx), dentin matrix protein 1 (Dmp1), dentin sialophosphateprotein (Dspp), bone sialoprotein (Bsp), matrix extracellular phosphoglycoprotein (Mepe), osteopotin (Opn), alkaline phosphatase (Alp), Type I Collagen, alpha-1 (Col1a1), fibroblast growth factor 23 (Fgf23), osteocalcin (Ocn), osteonectin (Osn) and glyceraldehyde-3-phosphate dehydrogenase (Gapdh) were described in the previous reports (Wu et al., 2010; Wang et al., 2012a; Liu et al., 2015). Q-PCR was also applied to compare the relative expression of Bmp2 with primers (Forward: 5’-TGACTGGATCGTGGCACCTC -3’; Backword: 5’- CAGAGTCTGCACTATGGCATGGTTA -3’), of Bmp4 with primers (Forward: 5’- ACAATGTGACACGGTGGGAAAC -3’; Backword: 5’- TGTGGGTGATGCTTG GGACTAC-3’) and of Bmp7 with primers (Forward: 5’-ACATCCGGGAGCGATTT GAC-3’; Backword: 5’-TCCTCAGAAGCCCAGATGGTC-3’)

Assay of cell morphology and cell proliferation

The morphologies of immortalized OB Fam20cf/f and OB Fam20cKO osteoblasts were examined under the Olympus phase contrast microscope. The 5-bromo -2’-deoxturidine (BrdU) was supplemented into the culture medium in the concentration of 30µM to label the nuclei of proliferating cells (Life Technologies, Inc., Grand Island, NY). After 4 hours of incorporation of BrdU, the BrdU positive nuclei of the Fam20cf/f and Fam20cKO osteoblasts were detected by the BrdU assay Kit (Life Technologies, Inc., Grand Island, NY).

Immunocytochemistry

The antibody against BSP was a gift from Dr. Larry Fisher at National Institute of Dental and Craniofacial Research (Liu et al., 2014; Liu et al., 2015; Xie et al., 2016). The anti-DMP1 and anti-DSP antibodies were applied as previously described (Baba et al., 2004; Huang et al., 2008; Liu et al., 2014, Liu et al., 2015). The polyclonal rabbit IgG antibodies against MEPE and OPN were obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA) and Abcam, Inc. (Cambridge, MA), respectively. The anti-mouse and anti-rabbit IgG conjugated with biotin were applied as the secondary antibodies (Vector Laboratories, Inc., Burlingame, CA).

Western Blotting Assay

To compare the secreted FGF23 concentration between the culture media of Fam20cf/f and Fam20cKO osteoblasts, α-MEM with 0.2% fetal bovine serum (FBS) was applied to replace the α-MEM supplemented with 10% FBS, when the osteoblasts became confluent. After 24 hours of culture, the α-MEM with 0.2% FBS was mixed with the protease inhibitor cocktail and centrifuged. The supernatant was collected and then further concentrated with Amicon Ultra Centrifugal Filters (EMD Millipore Company, Merck KGaA, Darmstadt, Germany). The protein concentration in the supernatants was determined by the Bicinchoninic Acid assay (Pierce Biotechnology). Protein samples containing equal amount of β-actin were loaded onto 5–15% SDS-PAGE for Western blotting assay following the protocol as previously described (Xie et al., 2016). The monoclonal mouse anti-human FGF23 antibody (anti-FGF23-79) was used as the primary antibody (Jani et al., 2016). To evaluate the influence of FAM20C on BMP signaling pathways, the antibodies against the phosphorylated Smad1/5/8 (p-Smad1/5/8), phosphorylated Erk (p-Erk), phosphorylated p-38 (p-p38), phosphorylated JNK (p-JNK), phosphorylated and non-phosphorylated Smad1/5/8 (pan-Smad1/5/8), phosphorylated and non-phosphorylated Erk (pan-Erk), phosphorylated and non-phosphorylated p-38 (pan-p38), and phosphorylated and non-phosphorylated JNK (pan-JNK) were all polyclonal rabbit IgG purchased from Santa Cruz Biotechnology. The mouse IgG against mouse β-actin was obtained from Sigma-Aldrich, St. Louis, MO. The secondary antibodies were goat anti-mouse and goat anti-rabbit IgG conjugated with Alkaline Phosphatase (Vector Laboratories, Inc., Burlingame, CA). The CD star detection system was applied for the detection of chemilumimescent bands (Sigma-Aldrich, St. Louis, MO).

Assay of alkaline phosphatase (ALP) activity and mineralization

The ALP activity assay and the mineralization assay followed the protocol previously described (Liu et al., 2015). The non-collagenous proteins in the bone matrix were extracted from the WT rat long bones as described in the previous reports (Qin et al., 2006). To rescue the incapacitated mineralization of Fam20cKO osteoblasts, the lyophilized non-collagenous extracts from the bone matrix were supplemented at the final concentration of 80 µg/ml into the mineralization-inducing medium (α-MEM supplemented with 10% fetal bovine serum, 100 unit/ml penicillin, 100ug/ml streptomycin, 50 µg/ml ascorbic acid and 10 µM sodium β-glycerophosphate). After 1 week of induction, the cells and their secreted matrix were fixed with 4% PFA for 20 min, rinsed with phosphate buffered Saline (PBS) twice and then, stained with Alizarin Red S (Sigma–Aldrich, St. Louis, MO).

Cell migration assay

When the OB Fam20cf/f and OB Fam20cKO osteoblasts were grown to form a monolayer, a scratch of 500 µm wide was made on the monolayer with a pipette tip. The images were taken at 0 and 24 hours after scratch was made. To further confirm the influence of FAM20C on cell migration, 1.0×107 osteoblasts were concentrated into a suspension of 100 µl and dropped onto each well of a 24-well plate. At 2 hours after dropping the cell suspension, α-MEM with 10% fetal bovine serum was added to cover the surface of each well. The diameter of each cell mass was measured after 24 hours to evaluate the migration capability.

Statistical Analysis

All values were presented as the mean with standard deviation (SD). The p-value of <0.05 from the unpaired student’s t-test was considered to be statistically significant.

Results

Generation of Fam20c-deficient osteoblast cell line

The OB Fam20cf/f osteoblasts infected by CMV-Cre-IRES-EGFP lentivirus were re-plated in a very low density to produce the single cell-derived colonies. Only the colonies in which all the offspring cells were EGFP positive (i.e., with Cre integrated into the genomes of the infected osteoblasts) were selected and expanded for further investigation (Fig. 1A). Genotyping PCR confirmed that the genomic DNA extracted from OB Fam20cf/f only gave rise to a 400-bp band for the floxed Fam20c allele (Fig. 1B). In contrast, the genomic DNA of the infected EGFP-expressing osteoblasts were devoid of the bands for both the WT and floxed Fam20c alleles, but only gave rise to the bands for the recombined Fam20c null allele as well as Cre recombinase gene (Fig. 1B). RT-PCR with Set 1 primers (covering exon 5to exon8) detected a 381-bp band only in the transcriptome of the OB Fam20cf/f (Fig. 1C). Meanwhile, Set 2 primers (from exon 5 to exon 11) were able to amplify an 820-bp band from the transcriptome of OB Fam20cf/f, and a 431-bp band matching the truncated transcript from the mRNAs of the EGFP-expressing osteoblasts (Fig. 1C). These results indicated the removal of the fragment from exon 6 to exon 9 of Fam20c allele in the EGFP-expressing osteoblasts, and confirmed that the floxed Fam20c allele was successfully transformed into null allele by the transfected Cre in the EGFP-expressing osteoblasts.

Fig 1. Removal of the floxed Fam20c allele by CMV-Cre-IRES-EGFP lentivirus.

Fig 1

Open in a new tab

(A) The immortalized Fam20cf/f osteoblasts were infected with CMV-Cre-IRES-EGFP lentivirus and cultured after 10 passages; note that all of the cells were positive for DAPI (blue) and EGFP (green), indicating an uniform infection of CMV-Cre-IRES-EGFP lentivirus in these osteoblasts. (B) The genomic DNA from Fam20cf/+ mouse tail (f/+), OB Fam20cf/f (f/f; without infection by lentivirus) and OB Fam20cKO (KO; after treatment with lentivirus) were genotyped with specific primers for the floxed Fam20c allele, recombined Fam20c allele and Cre, respectively. (C) RT-PCR was performed to identify the successful removal of the floxed Fam20c allele. The Set 1 primers for exon 5–8 gave a 388-bp band for OB Fam20cf/f cells (f/f), but produced no band for OB Fam20cKO cells (KO); the Set 2 primers for exon 5–11 of Fam20c transcript produced an 820-bp band for OB Fam20cf/f (f/f), and a truncated 421-bp band for OB Fam20cKO osteoblasts (KO). (Std: DNA ladder; Scale bar:100 um)

Fam20c-deficient osteoblasts (OB Fam20cKO) had no alteration in the morphology and proliferation

We did not find significant morphological differences between the OB Fam20cf/f and OB Fam20cKO osteoblasts under the bright filed microscope (Fig. 2A). Although in the OB Fam20cKO group, the median value for the ratio of BrdU positive cell to total cells was slightly lower than that in the OB Fam20cf/f, statistical analysis indicated that the difference between the proliferation ratio of these two lines was insignificant (p>0.05; Fig. 2B&C), which was in agreement with the findings obtained from the Fam20c-deficient mice (Wang et al., 2012a).

Fig 2. Cell morphology and proliferation of the OB Fam20cKO cell line.

Fig 2

Open in a new tab

(A) The micrographs OB Fam20cf/f and OB Fam20cKO osteoblasts were taken under the Olympus reverse phase-contrast microscope. (B) BrdU-labeling images of OB Fam20cf/f and OB Fam20cKO were photographed under the Olympus reverse phase-contrast microscope. (C) The statistical assay showed that there were no significant difference (p>0.05) in the proliferation rate between OB Fam20cf/f (Mean=45.24%, SD=8.701%) and OB Fam20cKO (Mean=37.83%, SD=5.814%). Scale bar in (A) and (B) equals to100 µm).

Altered gene expression in Fam20c-deficient osteoblasts

Q-PCR was performed to compare the gene expression profile between OB Fam20cKO and OB Fam20cf/f cells. Compared with OB Fam20cf/f, the transcription of Runx2 and Dlx3 were up-regulated, while the Atf4 and Osx were down-regulated in the OB Fam20cKO (Fig. 3A). The reduced transcription of Dmp1, Dspp and Opn were detected in the OB Fam20cKO, but the elevated Bsp and unaltered Mepe transcription were observed in these Fam20c-deficient osteoblasts (Fig. 3B). The mRNA levels of Alp and Col1a1 in the OB Fam20cKO were lower than in OB Fam20cf/f (Fig. 3C). In contrast, the mRNA levels of Fgf23 and Ocn in the OB Fam20cKO were higher than OB Fam20cf/f (Fig. 3C). Immunocytochemistry analyses further revealed the effects of FAM20C loss on the expression of SIBLINGs. Consistent with the Q-PCR, the immunostaining for DMP1, DSP and OPN in OB Fam20cKO were weaker than in OB Fam20cf/f (Fig. 4A). The staining of MEPE in the Fam20cKO was weaker than that in the OB Fam20cf/f, and the BSP staining of Fam20cKO was comparable to that in the OB Fam20cf/f (Fig. 4A).Western blotting assay indicated that the level of FGF23 secreted by OB Fam20cKO into culture medium was much higher than that of OB Fam20cf/f, which was consistent with the in vitro Q-PCR and in vivo serum biochemistry data (Fig. 4B).

Fig 3. The mRNA levels of bone-related genes in OB Fam20cf/f and OB Fam20cKO.

Fig 3

Open in a new tab

Total mRNAs from OB Fam20cf/f and OB Fam20cKO were extracted and reversed transcribed into cDNAs. Then, the cDNAs were amplified by Q-PCR to calculate the relative expression levels of bone-related transcription factors: Atf4, Dlx3, Osx and Runx2 (A), the SIBLING proteins: Dmp1, Dspp, Bsp, Mepe and Opn (B), and other bone-related genes: Alp, Fgf23, Col1a1, Ocn and Osn (C).(*: p<0.05; **:p<0.01; ***: p<0.001.)

Fig 4. The levels of bone-related gene expression detected by immunocytochemistry and Western blot.

Fig 4

Open in a new tab

(A) Immunocytochemistry with the antibodies against DMP1, DSP, BSP, MEPE and OPN were performed to assess the cultured OB Fam20cf/f and Fam20cKO cells. (B) Western blotting with FGF23 antibody showed that the FGF23 concentration in the serum free medium of the OB Fam20cKO was higher than in OB Fam20cf/f.

Impaired mineralization of Fam20c-deficient osteoblasts

In the mouse models, impaired mineralization in bone and tooth is the most evident defect caused by Fam20c deficiency (Wang et al., 2012a; Wang et al., 2012b). The capability of forming mineralized nodules of OB Fam20cKO was evaluated to examine the influence of FAM20C loss on mineralization. The ALP activity in OB Fam20cKO was remarkably lower than OB Fam20cf/f (Fig. 5A), which was consistent with the Q-PCR analyses showing that OB Fam20cKO had a lower level of Alp mRNA. As expected, the number of mineralized nodules in OB Fam20cKO was fewer than that of OB Fam20cf/f (Fig. 5B), indicating that Fam20c deficiency impaired the mineralization in a cell autonomous manner. To address if the phosphorylation of the SIBLINGs and/or other non-caollagenous proteins in ECM was essential for osteoblast mineralization, the non-collagenous proteins extracted from wild type rat long bone were supplemented into the culture medium. The addition of non-collagenous proteins extracted from the rat bone did not improve the formation of mineralized nodules in OB Fam20cKO (Fig. 5D), whereas in OB Fam20cf/f, supplementing the proteins extracted from the bone matrix enhanced the mineralized nodule formation (Fig. 5C). Thus, it implicates that Fam20c regulates mineralization by means more than phosphorylating SIBLINGs and/or ECM components.

Fig 5. Impaired mineralization of OB Fam20cKO.

Fig 5

Open in a new tab

(A) In situ histochemistry assay showed a decrease in ALP activity in OB Fam20cKO compared with OB Fam20cf/f. (B) After 3 weeks of induction, Alizarin red staining showed that the formation of mineralized nodules was severely impaired in OB Fam20cKO, compared to OB Fam20cf/f. (C) After one week of culture in the osteogenic medium, the addition of non-collagenous proteins extracted from normal bone matrix of rat legs enhanced the formation of mineralization nodules in OB Fam20cf/f. (D) There was no improvement on the formation of mineralization nodules by the addition of non-collagenous proteins extracted from normal bone matrix into OB Fam20cKO osteoblasts.

Altered canonical and non-canonical BMP signaling pathways in Fam20c-deficient osteoblasts

As a class of secreted proteins with typical Ser-X-Glu motif, bone morphogenetic proteins (BMPs), especially BMP4, were also predicted to be substrates of FAM20C (Tagliabracci et al., 2015). However, up to date, there was no study clarifying that the secreted BMPs were phosphorylated and if phosphorylation played a role in BMP activity. In our study, Western blot analyses showed that both the pan-Smad1/5/8 and the p-Smad1/5/8 were significantly attenuated in the OB Fam20cKO, though the ratios of p-Smad1/5/8 to pan-Smad1/5/8 changed insignificantly between OB Fam20cKO and OB Fam20cf/f (Fig. 6A). The pan-Erk showed no difference between OB Fam20cKO and OB Fam20cf/f, however, the ratio of p/pan-Erk in OB Fam20cKO was dramatically decreased, because of the decreased p-Erk (Fig. 6B). In spite of the down-regulation in both the pan- and p-p38 levels, the ratio of p/pan-p38 increased significantly in OB Fam20cKO (Fig. 6C). These results suggested that Fam20c deficiency attenuated both the canonical and non-canonical BMP signaling pathway. Intriguingly, the elevated transcription of Bmp2 and Bmp7, along with the unaltered Bmp4 transcription were detected in OB Fam20cKO (Fig. 6E), which implied that the phosphorylation was essential for the activity of BMP ligands.

Fig 6. The altered BMP signaling pathways in OB Fam20cKO.

Fig 6

Open in a new tab

Canonical and non-canonical BMP signaling pathways were analyzed by Western blotting with antibodies against pan-Smad1/5/8 and p-Smad1/5/8 for canonical BMP signaling (A), with antibodies against pan-Erk and p-Erk for BMP/p38 signaling (B), and with antibodies against pan-p38 and p-p38 (C) for BMP/p38 signaling. β-actin was used as the internal control to normalize the loading volume of OB Fam20cf/f and OB Fam20cKO cell lysates (D). Q-PCR was applied to compare the relative transcription of Bmp2, Bmp4 and Bmp7 between OB Fam20cf/f and Fam20cKO cells (E). (*: p<0.05; **: p<0.01.)

The attenuated migration of Fam20c-deficient osteoblasts

In the scratch tests, OB Fam20cf/f almost covered the scratch area in 24 hours, whereas only a few OB Fam20cKO osteoblasts migrated into the scratch (Fig.7A). The OB Fam20cf/f suspension developed colonies with larger diameters than OB Fam20cKO suspension containing equal number of cells in 24 hours (Fig.7B). The statistical analysis on the colony diameters confirmed that the diameter difference was statistically significant (Fig. 7C; p<0.001).

Fig 7. Reduced migration rate of OB Fam20cKO.

Fig 7

Open in a new tab

(A) After 24 hours of culture, the scratch in OB Fam20cf/f was nearly fully covered by the cells. In contrast, there were only a few OB Fam20cKO cells that had migrated into the scratch. (B) After 24 hours of dropping a suspension containing the 1×107 of cells into a culture dish, the diameter of OB Fam20cKO colony was notably smaller than that of Fam20cf/f osteoblasts. (C) The statistical assay showed a significant difference between the diameters of OB Fam20cf/f colonies (Mean=5.025mm, SD=0.359mm) and OB Fam20cKO colonies (Mean=3.98mm, SD=0.37mm). (***: p<0.001).

Discussion

The Fam20c-deficient mice manifest rickets/osteomalacia along with the elevation of serum FGF23, reduction of serum phosphorus and increased level of serum parathyroid hormone (Wang et al., 2012a; Liu et al., 2014). However, it is still unclear whether these skeletal abnormalities in Fam20c-deficient mice are primarily attributed to hypophosphatemia or mainly due to the intrinsic defects of osteoblasts. The availability of OB Fam20cf/f allows us to assess the local effects of Fam20c deficiency on the osteoblasts, by excluding the influence of systemic factors, such as hypophosphatemia and altered serum level of parathyroid hormone.

The effects of Fam20c deficiency on the gene expression of osteoblasts

The reduced transcriptions of Dmp1, Dspp and Opn, and their relatively weaker immunocytochemical staining in OB Fam20cKO indicate that the suppressed expression of these ECM molecules in the Fam20c knock-out mice results primarily from the loss of Fam20c (Wang et al., 2012a). Although the elevated Bsp transcription in OB Fam20cKO suggests a negative regulation of FAM20C on Bsp transcription, the unchanged BSP staining in OB Fam20cKO implicates that the phosphorylation on BSP may regulate its concentration post-transcriptionally, most likely by protein degradation. Moreover, the unchanged BSP staining in OB Fam20cKO is controversial to the elevated BSP staining in Fam20c knock-out mouse tooth (Wang et al., 2012b), revealing a possible systemic feedback to Fam20c deficiency on BSP concentration. Similarly, the comparable Mepe transcription and the weaker MEPE staining in OB Fam20cKO also suggested that the phosphorylation of MEPE could regulate its concentration post-translationally.

Since both the in vitro and in vivo data showed that the transcriptions of Col1α1 are reduced in the context of Fam20c deficiency, it is believed that the primary regulation of FAM20C on Col1α1 expression is negative. The reduced Alp transcription and ALP activity in OB Fam20cKO consistently indicate a direct dependence of Alp expression on FAM20C activity. Both the mRNA and protein of Fgf23 in the OB Fam20cKO are remarkably increased, which is in agreement with the observation in Fam20c-deficient mice. In contrast, the transcription of Ocn was up-regulated in OB Fam20cKO, which disagreed with the in vivo findings. The in vitro up-regulation of Ocn may result from the elevated Runx2 and Atf4, which promotes Ocn expression in vivo (Ferron et al., 2010; Fulzele et al., 2010; Oury et al., 2011; Oury et al., 2013). Additionally, it also suggests that the decreased transcription of Ocn in Fam20c knock-out mice may be secondary to Fam20c deficiency and caused by systemic feedbacks.

The transcriptions of Runx2 and Dlx3 promote matrix production during the whole process of osteoblast maturation, while Atf4 fulfills the same role only in late stage of osteoblast maturation (Hartmann, 2009). In OB Fam20cKO, the transcriptions of Runx2, Dlx3 and Atf4 were all up-regulated. The transcription of Osx only ensures the differentiation from osteochondrogenic progenitors into immature osteoblasts (Hartmann, 2009), and was down-regulated in OB Fam20cKO, which is consistent with the in vivo result (Wang et al., 2012a). Therefore, we speculate that the elevated transcriptions of Runx2, Dlx3 and Atf4 may be a compensatory response to the impaired osteoblast maturation caused by Fam20c deficiency. This speculation was also supported by the unaltered transcriptions of Mepe which is an inhibitor to osteogenic differentiation and mineralization (Gowen et al., 2003), and the unchangded Osn which works as an extracellular matrix down-regulated in mature bone in the OB Fam20cKO (Rosset and Bradshaw, 2016).

The effects of Fam20c deficiency on in vitro mineralization

In this study, we observed that the capability of OB Fam20cKO to form mineralized nodules was severely impaired. Since the influence of systemic factors, such as serum phosphorus, was excluded in these in vitro experiments, the impaired mineralization must be solely attributed to the intrinsic defects of the Fam20c-deficient osteoblasts. We once assume that the disabled phosphorylation of SIBLINGs and/or other non-collagenous components in Fam20c-deficient osteoblasts results in the impaired mineralization, and the phosphorylated non-collagenous components were sufficient to rescue the Fam20c-deficientosteoblast mineralization. To our surprise, the non-collagenous extracts from the normal bone matrix failed to rescue the mineralization defects of the Fam20c-deficient osteoblasts, suggesting that the loss or reduction of phosphorylation of the non-collagenous proteins was not the only cause responsible for the incapacitated mineralization of Fam20c-deficient osteoblasts. Since reduced ALP activity results in the accumulation of pyro-phosphate on the hydroxyapatite (HA) surface, which suppresses the further growth of HA crystals (Meyer, 1984), the significantly reduced Alp transcription and activity are likely to be one of the factors responsible for the impaired mineralization of the Fam20c-deficient osteoblasts. Obviously, further studies are needed to detect the concentration of pyro-phosphate in the Fam20c-deficient osteoblast cultured medium. Secondly, the non-collagenous proteins extracted in this manner do not include growth factors, such as BMP2 and BMP4. Thus, if the failure in phosphorylation for these growth factors contributes to the defective differentiation and maturation of Fam20c-deficient osteoblasts, and then the lack of such growth factors in our protein extract might be a factor for the rescue failure. Moreover, the reduced collagen expression was also supposed to impair the mineralization of OB Fam20cKO.

Attenuated BMP signaling pathways in Fam20c-deficient osteoblasts

FAM20C catalyzes the attachment of phosphate to Ser residues in the Ser-X-Glu motifs of secretory proteins in the Golgi apparatus. The presence of Ser-X-Glu motifs in BMP2 (Ser46-Asp47-Glu48, Ser117-Leu118-Glu119 and Ser147-Ala148-Glu149), BMP4 (Ser50-His51-Glu52, Ser91-Gly92-Glu93 and Ser155-Ala156-Glu157) and BMP7 (Ser48-Gln49-Glu50, Ser219-Glu220-Glu221 and Ser248-Val249-Glu250) implies that these proteins might be the potential substrates of FAM20C. In this study, we observed that both the canonical BMP signaling, and the non-canonical BMP/p38 and BMP/Erk signaling were attenuated in the OB Fam20cKO osteoblasts, suggesting thatFAM20Cmay enhance the BMP signaling pathways by phosphorylating the ligands, through which this kinase regulates osteoblast differentiation and maturation. Our further studies are warranted to provide more direct evidence as whether or not FAM20C phosphorylates BMPs and if the phosphorylation of BMPs by this kinase is indeed critical for their activities.

Since BMP ligands possess various biological functions, it implies that FAM20C can exert regulatory roles not only in mineralization, but also in many other biological processes. Although many studies revealed that BMP ligands promoted osteoblast differentiation and maturation by activating osteogenic transcription factors (Chen et al., 2012; Wu et al., 2016; Salazar et al., 2016;), there are few reports concerning the role of BMP signaling in the expression of ECM. In the OB Fam20cKO osteoblasts, if the altered expression of SIBLINGs and other ECM genes resulted from the attenuated BMP signaling pathways requires further investigation.

The effects of Fam20c deficiency on osteoblast migration

Latest study demonstrated that ablation of Fam20c by CRISPR/Cas9 technique in breast cancer cell migration revealed that cell migration relies on the FAM20C-dependent phosphorylation of insulin-like growth factor binding protein-7 (IGFBP-7) in an insulin-like growth factor-1 (IGF-1)-independent manner (Tagliabracci et al., 2015). In our study, the Fam20c-deficient osteoblasts are originated from the Fam20c knock-out mice generated by homologous recombination in embryonic stem cells (Wang et al., 2012a), which provides a non-tumor model and circumvents the off-target effects in CRISPR/Cas9 system. The OB Fam20cKO osteoblasts exhibit a noticeably retarded capability of migration in both scratch wound healing and suspension drop expanding experiments. These data demonstrated that FAM20C may play a critical role in the migration capability and the phosphorylation of secreted proteins may be essential for cell migration.

Acknowledgments

This study was supported by NIH Grant R01 DE022549 (to CQ) and the Start-up Fund by Dalian Medical University (201679; to CL).

Footnotes

Conflict of interest

All the authors have no conflict of interest.

References

  1. Bergwitz C, Banerjee S, Abu-Zahra H, Kaji H, Miyauchi A, Sugimoto T, Jüppner H. Defective O-glycosylation due to a novel homozygous S129P mutation is associated with lack of fibroblast growth factor 23 secretion and tumoral calcinosis. J Clin Endocrinol Metab. 2009;94:4267–4274. doi: 10.1210/jc.2009-0961. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bergwitz C, Jüppner H. Regulation of phosphate homeostasis by PTH, vitamin D, and FGF23. Annu Rev Med. 2010;61:91–104. doi: 10.1146/annurev.med.051308.111339. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bhattacharyya N, Chong WH, Gafni RI, Collins MT. Fibroblast growth factor 23: state of the field and future directions. Trends Endocrinol Metab. 2012;23:610–618. doi: 10.1016/j.tem.2012.07.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Chen G, Deng C, Li YP. TGF-β and BMP signaling in osteoblast differentiation and bone formation. Int J Biol Sci. 2012;8:272–288. doi: 10.7150/ijbs.2929. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Farrow EG, White KE. Recent advances in renal phosphate handling. Nat Rev Nephrol. 2010;6:207–217. doi: 10.1038/nrneph.2010.17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Ferron M, Wei J, Yoshizawa T, Del Fattore A, DePinho RA, Teti A, Ducy P, Karsenty G. Insulin signaling in osteoblasts integrates bone remodeling and energy metabolism. Cell. 2010;142:296–308. doi: 10.1016/j.cell.2010.06.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Fradin M, Stoetzel C, Muller J, Koob M, Christmann D, Debry C, Kohler M, Isnard M, Astruc D, Desprez P, Zorres C, Flori E, Dollfus H, Doray B. Osteosclerotic bone dysplasia in siblings with a Fam20C mutation. Clin Genet. 2011;80:177–183. doi: 10.1111/j.1399-0004.2010.01516.x. [DOI] [PubMed] [Google Scholar]
  8. Fulzele K, Riddle RC, DiGirolamo DJ, Cao X, Wan C, Chen D, Faugere MC, Aja S, Hussain MA, Brüning JC, Clemens TL. Insulin receptor signaling in osteoblasts regulates postnatal bone acquisition and body composition. Cell. 2010;142:309–319. doi: 10.1016/j.cell.2010.06.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. George A, Veis A. Phosphorylated proteins and control over apatite nucleation, crystal growth, and inhibition. Chem Rev. 2008;108:4670–4693. doi: 10.1021/cr0782729. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Gowen LC, Petersen DN, Mansolf AL, Qi H, Stock JL, Tkalcevic GT, Simmons HA, Crawford DT, Chidsey-Frink KL, Ke HZ, McNeish JD, Brown TA. Targeted disruption of the osteoblast/osteocyte factor 45 gene (OF45) results in increased bone formation and bone mass. J Biol Chem. 2003;278:1998–2007. doi: 10.1074/jbc.M203250200. [DOI] [PubMed] [Google Scholar]
  11. Hao J, Narayanan K, Muni T, Ramachandran A, George A. Dentin matrix protein 4, a novel secretory calcium-binding protein that modulates odontoblast differentiation. J Biol Chem. 2007;282:15357–15365. doi: 10.1074/jbc.M701547200. [DOI] [PubMed] [Google Scholar]
  12. Hartmann C. Transcriptional networks controlling skeletal development. Curr Opin Genet Dev. 2009;19:437–443. doi: 10.1016/j.gde.2009.09.001. [DOI] [PubMed] [Google Scholar]
  13. Ishikawa HO, Xu A, Ogura E, Manning G, Irvine KD. The Raine syndrome protein FAM20C is a Golgi kinase that phosphorylates bio-mineralization proteins. PLoS One. 2012;7:e42988. doi: 10.1371/journal.pone.0042988. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Jani PH, Gibson MP, Liu C, Zhang H, Wang X, Lu Y, Qin C. Transgenic expression of Dspp partially rescued the long bone defects of Dmp1-null mice. Matrix Biol. 2016;52–54:95–112. doi: 10.1016/j.matbio.2015.12.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kato K, Jeanneau C, Tarp MA, Benet-Pagès A, Lorenz-Depiereux B, Bennett EP, Mandel U, Strom TM, Clausen H. Polypeptide GalNAc-transferase T3 and familial tumoral calcinosis. Secretion of fibroblast growth factor 23 requires O-glycosylation. J Biol Chem. 2006;281:18370–183777. doi: 10.1074/jbc.M602469200. [DOI] [PubMed] [Google Scholar]
  16. Kurosu H, Ogawa Y, Miyoshi M, Yamamoto M, Nandi A, Rosenblatt KP, Baum MG, Schiavi S, Hu MC, Moe OW, Kuro-o M. Regulation of fibroblast growth factor-23 signaling by klotho. J Biol Chem. 2006;281:6120–6123. doi: 10.1074/jbc.C500457200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Liu C, Wang X, Zhang H, Xie X, Liu P, Liu Y, Jani PH, Lu Y, Chen S, Qin C. Immortalized Mouse Floxed Fam20c Dental Papillar Mesenchymal and Osteoblast Cell Lines Retain Their Primary Characteristics. J Cell Physiol. 2015;230:2581–2587. doi: 10.1002/jcp.25008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Liu P, Zhang H, Liu C, Wang X, Chen L, Qin C. Inactivation of Fam20C in cells expressing type I collagen causes periodontal disease in mice. PLoS One. 2014;9:e114396. doi: 10.1371/journal.pone.0114396. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Meyer JL. Can biological calcification occur in the presence of pyrophosphate? Arch Biochem Biophys. 1984;231:1–8. doi: 10.1016/0003-9861(84)90356-4. [DOI] [PubMed] [Google Scholar]
  20. Oury F, Khrimian L, Denny CA, Gardin A, Chamouni A, Goeden N, Huang YY, Lee H, Srinivas P, Gao XB, Suyama S, Langer T, Mann JJ, Horvath TL, Bonnin A, Karsenty G. Maternal and offspring pools of osteocalcin influence brain development and functions. Cell. 2013;155:228–241. doi: 10.1016/j.cell.2013.08.042. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Oury F, Sumara G, Sumara O, Ferron M, Chang H, Smith CE, Hermo L, Suarez S, Roth BL, Ducy P, Karsenty G. Endocrine regulation of male fertility by the skeleton. Cell. 2011;144:796–809. doi: 10.1016/j.cell.2011.02.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Qin C, Huang B, Wygant JN, McIntyre BW, McDonald CH, Cook RG, Butler WT. A chondroitin sulfate chain attached to the bone dentin matrix protein 1 NH2-terminal fragment. J Biol Chem. 2006;281:8034–8040. doi: 10.1074/jbc.M512964200. [DOI] [PubMed] [Google Scholar]
  23. Rafaelsen SH, Raeder H, Fagerheim AK, Knappskog P, Carpenter TO, Johansson S, Bjerknes R. Exome sequencing reveals FAM20c mutations associated with fibroblast growth factor 23-related hypophosphatemia, dental anomalies, and ectopic calcification. J Bone Miner Res. 2013;28:1378–1385. doi: 10.1002/jbmr.1850. [DOI] [PubMed] [Google Scholar]
  24. Rosset EM, Bradshaw AD. SPARC/osteonectin in mineralized tissue. Matrix Biol. 2016;52–54:78–87. doi: 10.1016/j.matbio.2016.02.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Salazar VS, Gamer LW, Rosen V. BMP signalling in skeletal development, disease and repair. Nat Rev Endocrinol. 2016;12:203–221. doi: 10.1038/nrendo.2016.12. [DOI] [PubMed] [Google Scholar]
  26. Schaeren S, Jaquiéry C, Wolf F, Papadimitropoulos A, Barbero A, Schultz-Thater E, Heberer M, Martin I. Effect of bone sialoprotein coating of ceramic and synthetic polymer materials on in vitro osteogenic cell differentiation and in vivo bone formation. J Biomed Mater Res A. 2010;92:1461–1467. doi: 10.1002/jbm.a.32459. [DOI] [PubMed] [Google Scholar]
  27. Simpson MA, Hsu R, Keir LS, Hao J, Sivapalan G, Ernst LM, Zackai EH, Al-Gazali LI, Hulskamp G, Kingston HM, Prescott TE, Ion A, Patton MA, Murday V, George A, Crosby AH. Mutations in FAM20C are associated with lethal osteosclerotic bone dysplasia (Raine syndrome), highlighting a crucial molecule in bone development. Am J Hum Genet. 2007;81:906–912. doi: 10.1086/522240. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Tagliabracci VS, Engel JL, Wen J, Wiley SE, Worby CA, Kinch LN, Xiao J, Grishin NV, Dixon JE. Secreted kinase phosphorylates extracellular proteins that regulate biomineralization. Science. 2012;336:1150–1153. doi: 10.1126/science.1217817. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Tagliabracci VS, Xiao J, Dixon JE. Phosphorylation of substrates destined for secretion by the Fam20 kinases. Biochem Soc Trans. 2013;41:1061–1065. doi: 10.1042/BST20130059. [DOI] [PubMed] [Google Scholar]
  30. Tagliabracci VS, Engel JL, Wiley SE, Xiao J, Gonzalez DJ, NidumandaAppaiah H, Koller A, Nizet V, White KE, Dixon JE. Dynamic regulation of FGF23 by Fam20C phosphorylation, GalNAc-T3 glycosylation, and furin proteolysis. Proc Natl Acad Sci U S A. 2014;111:5520–5525. doi: 10.1073/pnas.1402218111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Tagliabracci VS, Wiley SE, Guo X, Kinch LN, Durrant E, Wen J, Xiao J, Cui J, Nguyen KB, Engel JL, Coon JJ, Grishin N, Pinna LA, Pagliarini DJ, Dixon JE. A Single Kinase Generates the Majority of the Secreted Phosphoproteome. Cell. 2015;161:1619–1632. doi: 10.1016/j.cell.2015.05.028. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Wang X, Wang S, Li C, Gao T, Liu Y, Rangiani A, Sun Y, Hao J, George A, Lu Y, Groppe J, Yuan B, Feng JQ, Qin C. Inactivation of a novel FGF23 regulator, FAM20C, leads to hypophosphatemic rickets in mice. PLoS Genet. 2012a;8:e1002708. doi: 10.1371/journal.pgen.1002708. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Wang X, Wang S, Lu Y, Gibson MP, Liu Y, Yuan B, Feng JQ, Qin C. FAM20C plays an essential role in the formation of murine teeth. J Biol Chem. 2012b;287:35934–35942. doi: 10.1074/jbc.M112.386862. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Wu M, Chen G, Li YP. TGF-β and BMP signaling in osteoblast, skeletal development, and bone formation, homeostasis and disease. Bone Res. 2016;4:16009. doi: 10.1038/boneres.2016.9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Wu LA, Feng J, Wang L, Mu YD, Baker A, Donly KJ, Gluhak-Heinrich J, Harris SE, MacDougall M, Chen S. Immortalized mouse floxed Bmp2 dental papilla mesenchymal cell lines preserve odontoblastic phenotype and respond to BMP2. J Cell Physiol. 2010;225:132–139. doi: 10.1002/jcp.22204. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Xie X, Liu C, Zhang H, Jani PH, Lu Y, Wang X, Zhang B, Qin C. Abrogation of epithelial BMP2 and BMP4 causes Amelogenesis Imperfecta by reducing MMP20 and KLK4 expression. Sci Rep. 2016;6:25364. doi: 10.1038/srep25364. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES