FAM20C, also known as Golgi casein kinase (G-CK), is proposed to be the archetype for a family of secreted kinases that phosphorylate target proteins in the Golgi and in extracellular matrices, but FAM20C serving an extracellular function is controversial. FAM20C phosphorylates secretory calcium-binding phosphoproteins (SCPPs), which are associated with the evolution of biomineralization in vertebrates. Current models of biomineralization assume SCPP proteins are secreted as phosphoproteins and their phosphates are essential for protein conformation and function. It would be a radical departure from current theories if proteins in mineralizing matrices were dephosphorylated as part of the mineralization mechanism and rephosphorylated in the extracellular milieu by FAM20C using ATP. To see if such mechanisms are possible in the formation of dental enamel, we tested the hypothesis that FAM20C is secreted by ameloblasts and accumulates in the enamel extracellular matrix during tooth development. FAM20C localization was determined by immunohistochemistry in day 5 mouse incisors and molars and by Western blot analyses of proteins extracted from pig enamel organ epithelia (EOE) and enamel shavings. FAM20C localized intracellularly within ameloblasts and odontoblasts in a pattern consistent with Golgi localization. Western blots detected FAM20C in the EOE extracts but not in the enamel matrix. We conclude that FAM20C is not a constituent of the enamel extracellular matrix and functions intracellularly within ameloblasts. © 2013 American Society for Bone and Mineral Research.
Biomineralization in vertebrates is associated with the evolution of the secretory calcium-binding phosphoprotein (SCPP) gene family. SCPP proteins generally have one or more Golgi casein kinase (G-CK) phosphorylation sites that are recognized by their distinctive Ser-x-Glu/pSer target motif. Recently, it was discovered that G-CK is encoded by FAM20C and potentially secreted. Mutations in FAM20C cause Raine syndrome (OMIM #259775), an autosomal recessive disorder characterized by major defects in biomineralization, including dentin and enamel. The causes of the mineralization phenotype are multifaceted, because SCPP proteins function directly and indirectly in biomineralization, and G-CK phosphorylation motifs are found in proteins expressed in nonmineralizing systems. The SCPP protein DMP1, for instance, functions in phosphate homeostasis and Fam20c null mice have hypophosphatemic rickets.
Recently it was proposed that FAM20C is the archetypal member of a group of secreted kinases that phosphorylate target proteins in the Golgi and in extracellular matrices. If confirmed, this would force major revisions in our understanding of the biomineralization in vertebrates. Current models of biomineralization assume that SCPP proteins are secreted as phosphoproteins and the phosphates are essential for protein conformation and function. The secretion of G-CK into the matrix raises the question of what it is doing there. Are the phosphates on SCPP proteins consumed during the deposition of calcium phosphate solid phases and replenished by secreted FAM20C and ATP, perhaps repeating such cycles many times during the deposition of hydroxyapatite?
Dental enamel formation is highly dependent upon G-CK activity. Secretory stage enamel extracellular matrix is comprised mainly of three SCPP proteins: amelogenin (AMEL), enamelin (ENAM), and ameloblastin (AMBN), as well as enamelysin (MMP20), a low-abundance matrix metalloprotease (MMP). Serines phosphorylated by G-CK are strictly conserved in amelogenin and enamelin. An ENAM missense mutation that altered a G-CK phosphorylation site (p.Ser216Leu) caused inherited enamel defects. Because FAM20C is critical for proper dental enamel formation, its expression by ameloblasts should be readily detected. Because enamel is completely acellular and can be readily separated from the soft tissue that forms it, enamel formation is an ideal system for distinguishing between FAM20C that localizes in intracellular and extracellular compartments. In this study we test the hypothesis that FAM20C is secreted by ameloblasts (enamel-forming cells) and is a constituent of the enamel extracellular matrix during tooth development.
Materials and Methods
All experimental procedures involving the use of animals were reviewed and approved by the Institutional Animal Care and Use Program at the University of Michigan.
Tissue preparation and immunohistochemistry
All procedures were carried out at 4°C, unless otherwise indicated. Day 5 mouse heads were dissected off skin, immersed in 4% paraformaldehyde fixative overnight, washed in PBS 4-5X (every 0.5–1 hour), and decalcified by immersion in 1 L of 4.13% EDTA (pH 7.3) with agitation. The EDTA solution was changed every other day for 8 to 9 days. After decalcification the tissues were immersed in 30% sucrose overnight for cryoprotection and embedded in OCT/Tissue Tek (Sakura Finetek, Torrance, CA, USA). The blocks were cryosectioned at 10-µm thickness at –20°C. The slides were rinsed with PBT buffer (0.1% Triton X-100 in PBS), blocked with 5% sheep serum (S-22; Chemicon, Billerica, MA, USA) in PBT for 30 minutes at room temperature, and serial sections were incubated overnight with anti-FAM20C (1:200, HPA019823; Sigma-Aldrich, St. Louis, MO, USA), anti-MMP20 (1:200, ab39038; Abcam, Cambridge, MA, USA), or anti-AMEL (1:500) antibodies. The sections were washed with PBT for 15 minutes and incubated for 30 minutes at room temperature in solutions containing anti-rabbit immunoglobulin G (IgG) secondary antibody conjugated with Alexa Fluor 594 (1:500, A11012; Invitrogen, Grand Island, NY, USA). Sections were rinsed in PBT for 15 minutes, mounted with ProLong Gold antifade reagent with DAPI (P-36931; Invitrogen), and examined using an Olympus BX51 with fluorescence attachments and photographed using an Olympus DP71 camera with DP controller and manager software.
Western blot analyses of enamel organ epithelia and enamel extracts
Tooth germs of unerupted second molars were surgically extracted to obtain enamel organ epithelia (EOE) and enamel shavings as described. Enamel shavings were sonicated in 25 mL of HF Buffer (pH 1; 0.17 N HCl, 0.95 N formic acid/1 mM Protease Inhibitor Cocktail Set III; Calbiochem, Billerica, MA, USA), and then centrifuged for 15 minutes at 15,000g. The pH was raised to 3.5 by the addition of 600 µL 6N NaOH, desalted 3× using a 10-kDa centrifugal filters (Amicon, Billerica, MA, USA) for 30 minutes at 3000g, frozen at –80°C for 3 hours and lyophilized for 48 hours. The enamel extract was raised in 1 mL of 0.1% formic acid and its concentration determined (6.5 mg/mL) by Bradford assay using an rP172 standard curve.
EOE (1 g stored at –80°C) was suspended in 10 mL of TU Buffer (6 M Urea, 25 mM Tris-HCl, 1 mM Protease Inhibitor Cocktail Set III; pH 7.27; Calbiochem), sonicated and centrifuged for 15 minutes at 15,000g. The supernatant concentration was determined by Bradford assay using a bovine serum albumin (BSA) standard curve. The final concentration was 5.5 mg/mL. The enamel and EOE extracts were raised in Laemmli sample buffer with 8% beta-mercaptoethanol, 82.5 µg of protein was applied per lane, and then separated on replica 4-12% Bis-Tris gels (Invitrogen). One SDS-PAGE was stained with Coomassie Brilliant Blue (CBB), the other was transblotted onto a nitrocellulose membrane (Invitrogen) and blocked for 1 hour in 1× TBS-T (0.01% Tween-20) with 5% blotting grade blocker (BioRad, Hercules, CA, USA). The membranes were incubated overnight at 4°C with the same antibodies used for immunohistochemistry (FAM20C 1:2000 and MMP20 1:5000). The membranes were washed and then incubated with anti-rabbit IgG horseradish peroxidase (HRP)-conjugated secondary antibody diluted to 1:10,000 in 5% blotting grade blocker (BioRad) for 1 hour at room temperature. The membranes were exposed to film for 5 minutes.
Immunohistochemistry of developing incisors localizes FAM20C intracellularly within ameloblasts and odontoblasts (enamel and dentin forming cells, respectively), in patterns consistent with localization in the Golgi (Fig. 1). No FAM20C signal is detected in the enamel or dentin extracellular matrices, although MMP20 (a low-abundance secreted protein) and amelogenin (a high-abundance secreted protein) are readily detected in positive controls. Similarly, no FAM20C signal is observed in enamel or dentin matrices in mouse molars (Fig. 2). Western blot analyses detect FAM20C in EOE extracts (containing ameloblasts), whereas FAM20C is not detected in enamel extracellular extracts that are strongly positive for MMP20. Because MMP20 is positive despite being a minor matrix constituent, FAM20C should have been detected in the extracellular matrix if it is secreted in any significant quantity. These results demonstrate that FAM20C localizes intracellularly and not in the extracellular matrix during enamel and dentin biomineralization.
Our finding that FAM20C is not secreted during enamel and dentin formation should force a reevaluation of the hypothesis that FAM20C is an archetype of the four-jointed family of secreted protein kinases that phosphorylate target proteins in extracellular matrices as well as in the Golgi. In humans there are three FAM20 members: FAM20A, FAM20B, and FAM20C. The family with sequence similarity 20 was first discovered in hematopoietic cells and was thought to be secreted, although it was not yet appreciated that they were kinases and associated with the Golgi. More recently, it was shown that FAM20A and FAM20B ectopically expressed in HEK293T cells localized to the Golgi but not the medium. Although the target sequence for its kinase activity is unknown, FAM20A is critical for dental enamel formation as well as other processes. Mutations in both alleles of FAM20A cause amelogenesis imperfecta and gingival fibromatosis syndrome (AIGFS; OMIM #614253) and enamel renal syndrome (ERS).[17, 18] FAM20B regulates glycosaminoglycan synthesis by phosphorylating xylose in the glycosaminoglycan–protein linkage region of proteoglycans, a function that requires localization to the Golgi, because subsequent steps in glycosaminoglycan synthesis occur in the Golgi. No human genetic diseases have been associated with FAM20B mutations.
HeLa cells overexpressing FLAG-tagged FAM20C retained some of the kinase in the Golgi, but most (∼90%) localized to the medium. However, an in vivo study demonstrated that the specific activity of G-CK in Golgi was almost 50× that found in milk, and a similar ratio was observed for galactosyltransferase, an enzyme with no putative extracellular function. Thus the in vivo findings from milk conflict with the in vitro overexpression data. We suspect that the observed secretion of recombinant FAM20C in cell culture may be an artifact of overexpression. Perhaps FAM20C, which does not have a transmembrane domain, is held in the Golgi by interactions with transmembrane proteins that are not present in sufficient quantity to retain the overexpressed recombinant protein.
In summary, the other members of the FAM20 family (FAM20A and FAM20B) localize to the Golgi and are not secreted. FAM20C is detectable in milk, but in very low amounts, and has not been associated with any extracellular function there. FAM20C was clearly identified in odontoblasts and ameloblasts in developing incisors by immunohistochemistry, but was not detected in the dentin and enamel matrices. FAM20C was detected in EOE extracts, but not in enamel extracts. We conclude that FAM20C functions in the Golgi during dental enamel formation and that a clear demonstration of significant FAM20C secretion in vivo should precede testing hypothesis concerning its function in extracellular matrices.
All authors state that they have no conflicts of interest.
This study was supported by NIDCR/NIH research grant DE015846. We thank Jennifer Dominguez, manager of the Michigan State University Meat Laboratory, and members of the Michigan State University Department of Animal Science for their kind assistance in obtaining fresh-developing molars from pigs.
Authors' roles: Experiments were performed by S-KW and ACS. All authors helped design the study and write the manuscript.