Search

Dentin phosphoprotein (DPP) is the most abundant noncollagenous protein in the dentin, where it plays a major role in the mineralization of dentin. However, we and others have shown that in addition to being present in the dentin, DPP is also present in nonmineralizing tissues like the kidney, lung, and salivary glands, where it conceivably has other functions such as in calcium transport. Because annexins have been implicated as calcium transporters, we examined the relationships between DPP and annexins. In this report, we show that DPP binds to annexin 2 and 6 present in a rat ureteric bud cell line (RUB1). Immunofluorescence studies show that annexin 2 and DPP colocalize in these cells. In addition, DPP and annexin 2 colocalize in the ureteric bud branches of embryonic metanephric kidney. In the RUB1 cells and ureteric bud branches of embryonic kidney, colocalization was restricted to the cell membrane. Studies on calcium influx into RUB cells show that in the presence of anti-DPP, there was a 40% reduction of calcium influx into these cells. We postulate that DPP has different functions in the kidney as compared with the odontoblasts. In the odontoblasts, its primary function is in the extracellular mineralization of dentin, whereas in the kidney it may participate in calcium transport.

Studies of mineralization of embryonic spicules and of the sea urchin genome have identified several putative mineralization-related proteins. These predicted proteins have not been isolated or confirmed in mature mineralized tissues. Mature Lytechinus variegatus teeth were demineralized with 0.6 N HCl after prior removal of non-mineralized constituents with 4.0 M guanidinium HCl. The HCl-extracted proteins were fractionated on ceramic hydroxyapatite and separated into bound and unbound pools. Gel electrophoresis compared the protein distributions. The differentially present bands were purified and digested with trypsin, and the tryptic peptides were separated by high pressure liquid chromatography. NH2-terminal sequences were determined by Edman degradation and compared with the genomic sequence bank data. Two of the putative mineralization-related proteins were found. Their complete amino acid sequences were cloned from our L. variegatus cDNA library. Apatite-binding UTMP16 was found to be present in two isoforms; both isoforms had a signal sequence, a Ser-Asp-rich extracellular matrix domain, and a transmembrane and cytosolic insertion sequence. UTMP19, although rich in Glu and Thr did not bind to apatite. It had neither signal peptide nor transmembrane domain but did have typical nuclear localization and nuclear exit signal sequences. Both proteins were phosphorylated and good substrates for phosphatase. Immunolocalization studies with anti-UTMP16 show it to concentrate at the syncytial membranes in contact with the mineral. On the basis of our TOF-SIMS analyses of magnesium ion and Asp mapping of the mineral phase composition, we speculate that UTMP16 may be important in establishing the high magnesium columns that fuse the calcite plates together to enhance the mechanical strength of the mineralized tooth.

After implantation in the exposed pulp, some molecules of the den-tin extracellular matrix induce the formation of a reparative dentinal bridge in the coronal pulp. In some cases, total occlusion of the root canal also is observed. This is the case for bone sialoprotein, bone morphogenetic protein-7, Dentonin (a fragment from matrix extracellular phosphoglycoprotein), and two small amelogenin gene splice products (A+4 and A-4). Cells implicated in the reparative process are recruited, proliferate, and differentiate into osteoblast-like and odontoblast-like cells. The same results may be obtained by direct implantation of odontoblast progenitor cell into the pulp.

Full-length cDNA coding for dentin matrix protein 1 (DMP1) has been cloned and sequenced, but the corresponding complete protein has not been isolated. In searching for naturally occurring DMP1, we recently discovered that the extracellular matrix of bone contains fragments originating from DMP1. Shortened forms of DMP1, termed 37K and 57K fragments, were treated with alkaline phosphatase and then digested with trypsin. The resultant peptides were purified by a two-dimensional method: size exclusion followed by reversed-phase high performance liquid chromatography. Purified peptides were sequenced by Edman degradation and mass spectrometry, and the sequences compared with the DMP1 sequence predicted from cDNA. Extensive sequencing of tryptic peptides revealed that the 37K fragments originated from the NH2-terminal region, and the 57K fragments were from the COOH-terminal part of DMP1. Phosphate analysis indicated that the 37K fragments contained 12 phosphates, and the 57K fragments had 41. From 37K fragments, two peptides lacked a COOH-terminal lysine or arginine; instead they ended at Phe173 and Ser180 and were thus COOH termini of 37K fragments. Two peptides were from the NH2 termini of 57K fragments, starting at Asp218 and Asp222. These findings indicated that DMP1 is proteolytically cleaved at four bonds, Phe173-Asp174, Ser180-Asp181, Ser217-Asp218, and Gln221-Asp222, forming eight fragments. The uniformity of cleavages at the NH2-terminal peptide bonds of aspartyl residues suggests that a single proteinase is involved. Based on its reported specificity, we hypothesize that these scissions are catalyzed by PHEX protein. We envision that the proteolytic processing of DMP1 plays a crucial role during osteogenesis and dentinogenesis.

Peritubular dentin (PTD) is a hypermineralized phase within the dentinal tubules in some vertebrate teeth as an interface between the intertubular dentin (ITD) and the cell processes. Our aim has been to understand the composition, structure and role of PTD as a mineralized tissue. We have utilized the technique of time of flight secondary ion mass spectrometry (TOF-SIMS) to map the distribution of positive and negative inorganic ions as well as organic components in the fully mineralized, intact PTD structure in bovine tooth cross-sections, and correlated these with scanning electron microscopy (SEM) in standard and backscatter modes. In recent work, we developed a procedure to freeze fracture the teeth and separate PTD from the less dense ITD by the use of aqueous sodium phosphotungstate step density gradients, after degrading the ITD collagen with NaOCl. Here, PTD-containing fragments were characterized by SEM and TOF-SIMS surface structure analysis. The TOF-SIMS data show that the isolated PTD does not contain collagen, but its surface is rich in glutamic acid-containing protein(s). The TOF-SIMS spectra also indicated that the intact PTD fragments contain phospholipids, and chemical analyses showed phosphatidylserine, phosphatidylinositol and phosphatidylcholine as the principal lipid components. In SEM sections, untreated PTD shows as a smooth collar around the tubule, but after digestion with ethylenediamine to remove all organic components, the porous nature of the mineral phase of small, thin platy apatite crystals becomes evident. Thus, the organic matrix of PTD appears to be a proteolipid-phospholipid complex. Copyright © 2008 S. Karger AG, Basel [ABSTRACT FROM AUTHOR]

Amelogenin synthesis is initiated in a restricted time frame during odontogenesis. Polypeptides translated from several alternatively spliced isoforms of amelogenin mRNA have been identified in ameloblasts and odontoblasts. Recent studies suggest that the isoforms deleting exons 6a, 6b, and 6c produce polypeptides that might exert regulatory functions governing the late stages of ameloblast and odontoblast differentiation. Herein, the spatial and temporal expression of mouse amelogenin mRNA isoforms M194, M180, M73, and M59 have been determined around the perinatal development period using splice form-specific probes. Expression levels and distribution patterns varied with developmental stage and cell location. Amelogenin mRNA expression was most prominent within the enamel organ at boundaries between cell layers, beginning at the newborn stage (PN0.5). Odontoblasts supported the expression of M73 and M59 mRNA from developmental stages PN0.5 to PN1.5 (1 d of age). In contrast, ameloblasts expressed predominantly the M180 mRNA isoform with full exon 6 but devoid of exon 4. In the enamel organ, the stratum intermediun cells supported expression of the full-length isoform, M194, including the full exon 6 and exon 4 sequences, and strikingly, expression of M180 message was inhibited. In conclusion, ameloblasts, odontoblasts, and stratum intermedium cells demonstrate selective alternative splicing patterns of the amelogenin pre-mRNA transcript. [ABSTRACT FROM AUTHOR]

The amino (N-telo) and carboxyl (C-telo) telopeptides of type I collagen play crucial roles, in vivo and in vitro, in the assembly of collagen fibrils, regulating the axial alignment of the molecules within a fibril, azimuthal orientations of neighboring molecules, and cross-link formation. High-resolution structures of the telopeptides are not available from X-ray diffraction studies, but computational methods permitted prediction of the N-telo structure within a fibril. Here, using a suite of molecular modeling software, the more complex heterotrimeric C-telo of human type I collagen has been built from the correct sequences and energy minimized and the energy minimum confirmed by molecular dynamics. The receptor triple helix was modeled on the basis of the Protein Data Bank coordinates of a collagen-like sequence. Docking of the heterotrimeric C-telopeptide to its receptor showed that hydrophobic interactions involving the short eQ C-telopeptide are crucial determinants of its azimuthal orientation within the docked structure. A docked C-telo can interact with only one neighboring helix. The two α1(I) C-telo chains in the €11- (A)-α2-α1(B) chain stagger do not have identical docked conformations, and one of the α1(I) C-telo chains appears to be favored for formation of a cross-link between its K16C and a helix K87. Prior studies showed that a docked N-telopeptide can interact with two adjacent collagen monomers, forming a tightly packed region. A recent X-ray analysis showed the N- and C-telo regions pack differently, with the C-telo region being less densely packed than N-telo regions. This difference between N- and C-telopeptide docked structures demonstrates how unique and specific packing can occur in the fibril at each boundary of the type I collagen gap region. [ABSTRACT FROM AUTHOR]

In bone and dentin the formation and mineralization of the extra cellular matrix structure is a complex process highly dependent on intermolecular interactions. In dentin, the phosphophoryns (PP) and type I collagen (COL1) are the major constituents implicated in mineralization. Thus, as a first step in understanding the tissue organization, we have initiated a study of their interaction as a function of pH, ionic strength, and relative concentrations or mixing ratios. Complex formation has been analyzed by dynamic light scattering to detect aggregate formation and by rotary shadowing electron microscopy (EM) to determine aggregate shape. The EM data showed that at the pH values studied, the PP-COL1 interaction leads to the formation of large fibrillar aggregates in which the PP are present along the fibril surfaces. The quantitative phase distribution data showed a 1/1molar equivalence at the maximum aggregation point, not at electrostatic PP-COL1 equivalence. As the ionic strength was raised, the PP-COL1 aggregates became smaller but the binding and asymmetric fibrillar aggregation persisted. In EM, the PP appear as dense spheres. Along the surfaces of the collagen aggregates, the PP are larger and more open or extended, suggesting that COL1-bound PP may undergo a conformational change, opening up so that a single PP molecule might interact with and electrostatically link several COL1 molecules. This might have important implications for dentin structure, stability, and mineralization. [ABSTRACT FROM AUTHOR]

Recombinant proteins have been produced from cDNAs corresponding to alternatively spliced transcripts comprised from exons 2,3,4,5,6d,7 and 2,3,5,6d,7 of the rat amelogenin gene. These peptides, designated as [A + 4] and [A - 4], respectively, induce embryonic muscle fibroblasts in culture in vitro to express proteins characteristic of the chondrogenic and osteogenic phenotypes, and in matrix-supported implants into rat muscle, in vivo, induce typical bone matrix proteins. The aim of the present work was to examine the potential role of these proteins on the development of odontogenic tissue. The lower first molars were collected from Charles River CD-1 mice at postnatal days 1 and 2 and were grown on semisolid, serum-free medium supplemented with ascorbic and retinoic acids and transferrin. The peptides were added to the serum-free media at 10 ng/ml. As controls, the medium was either 20% fetal bovine serum or the supplemented serum-free medium without either amelogenin peptide. The tooth germs were cultured for 6 days, then fixed and paraffin embedded by standard procedures. The tissue blocks were serially sectioned and stained with hematoxylin-eosin (H&E), or antibodies to collagen 1 (Col1), phosphophoryn (DMP2), or cementum attachment protein (CAP). CAP, DMP2, and Col1 expression was enhanced by the addition of the amelogenin peptides, as compared to the 0% fetal bovine serum (FBS) controls, but the peptides showed different effects. Expression of DMP2, characteristic of dentin matrix, was upregulated by [A + 4], whereas CAP, characteristic of cementum, was upregulated by [A - 4]. Since the recombinant peptides are active, their corresponding tissue forms may be important in the stimulation of mesenchymal tissue differentiation. Thus, these specific amelogenin proteins may be involved in tooth morphogenesis. [ABSTRACT FROM AUTHOR]

Phosphophoryns (PPs) are unique aspartic acid and phosphoserine-rich proteins present in all species of dentin. Rat incisor odontoblast cDNA libraries contain messages encoding several acidic phosphorylated, serine-rich proteins. At least two of these share a common C-terminal domain coding region sequence. The polypeptide sequences in the N-terminal direction immediately adjacent to the conserved C-terminal domains of these two proteins (DMP2, DMP3) are distinctly different. In this domain, the DMP2 has extensive sequences of (DSS)n repeats with n as large as 24. DMP3 has fewer and shorter triplet sequences, n = 3, 4. The major rat incisor PPs (90-95 kDa) probably have the (DSS) n ≫ 3. We propose that the name phosphophoryn be reserved for the extracellular matrix proteins with these extended repeats. DMP1, although strongly acidic, does not fit this category. If the S residues are phosphorylated and n > 3. conformational energy minimization computations show the (DSS)n sequence to assume a unique extended structure with parallel arrays of carboxylate and phosphate groups which may function as Ca2+ ion interaction edges. The phosphorylation of recombinant DMP2 C-terminal domain by various kinases has been examined. The repeat domains are not direct substrates for the CK2-like kinases but the kinases act in concert, so that the phosphorylation is hierarchical, apparently controlled by the presence of specific interruptions between the triplet domains. [ABSTRACT FROM AUTHOR]

Collagens are modular triple-helical proteins that constitute the major structural components of the extracellular matrix of all animals. They occur as diverse superstructures such as fibrils, and networks that play functional roles in cell adhesion, cell differentiation, and structural integrity of organs. The suprastructures are assembled from a family of 43 distinct gene-products called a-chains that are classified as 28 types of collagens. A key self organizing step, common to all collagen types, is trimerization that selects, binds and registers cognate a-chains for assembly of triple-helical protomers that subsequently oligomerize into suprastructures. How a-chains recognize each other is a fundamental question in matrix biology that remains largely unanswered. In this presentation, we review recent findings on the chain-selection mechanism. We infer that terminal noncollagenous domains function as recognition modules in trimerization, and are therefore key determinants of specificity in the assembly of suprastructures. [ABSTRACT FROM AUTHOR]

Assembly of the type I procollagen molecule begins with interactions among the C-pro α1(1) and C-pro α2(I) domains. The C-propeptide domains themselves have subdomains of distinct structures. The important questions are where chain association begins and the basis of the chain selectivity which leads to the preferential formation of the [C-pro α1(I)]2[C-pro α2(I)] heterotrimer. These questions are addressed by energy minimization modeling of the individual C-propeptide structures, study of their docking interactions, and comparison of the heterotrimeric and homotrimeric C-pro structures and stability. The comparisons show the remarkable impact of the C-pro α2 chain on the structure of the assembled trimeric C-propeptide. In the modeling, the three chains were anchored and registered by a short C-terminal collagen triple-helical segment followed by the C-telopeptides in their docked conformation, and then the remaining C-propeptide chains were allowed to interact and dock. Surprisingly, propeptide trimerization did not proceed through the previously proposed N-terminal "oligomerization domain" of the C-propeptide [McAlinden et al. (2003) J. Riot Chem. 278, 42200] but rather in the most C-terminal domains of type I procollagen chains. Molecular dynamics showed heterotrimer assembly to begin with dimer formation between globular G2α2 and the G2α12 domains followed by trimerization at the G1 domains. Assembly initiation in the putative oligomerization coiled-coil domain is not possible because of the Pro residues at positions 3, 7, and 1.1. at the N-terminus of the α2 C-propeptide chain. To confirm the computations and proposed assembly pathway, the G2α1 and G2α2 domains were prepared recombinantly as the maltose binding protein constructs, and their interactions were studied by dynamic light scattering and gel filtration chromatography. Under the conditions examined MBP remained as monomer, MBP-G2α1 and MBP-G2α2 alone formed dimers, but a 2:1 mixture of MBP-G2α1 and MBP-G2α2 favored trimer formation. Thus, the C-terminal globular domains (G2) of the type I collagen C-propeptides play a crucial role in the initiation of intermolecular assembly and heterotrimer selectivity. [ABSTRACT FROM AUTHOR]

Phosphoproteins of the organic matrix of bone and dentin have been implicated as regulators of the nucleation and growth of the inorganic Ca-P crystals of vertebrate bones and teeth. One such protein identified in the dentin matrix is phosphophoryn (PP). It is highly acidic in nature because of a high content of aspartic acid and phosphate groups on serines. The 244-residue carboxyl-terminal domain of rat PP, predominantly containing the aspartic acid-serine repeats, has been cloned, and the corresponding protein has been expressed recombinantly in Escherichia coli. This portion of PP, named DMP2 (dentin matrix protein 2), is not phosphorylated by the bacteria and thus provided a means to study the function of the phosphate groups, the major post-translational modification of native PP. The recombinant DMP2 (rDMP2) possessed much lower calcium binding capacity than native PP. Small angle x-ray scattering experiments demonstrated that PP folds to a compact globular structure upon calcium binding, whereas rDMP2 maintained an unfolded structure. In vitro nucleation experiments showed that PP could nucleate plate-like apatite crystals in pseudophysiological buffer, whereas rDMP2 failed to mediate the transformation of amorphous calcium phosphate to apatite crystals under the same experimental conditions. Collagen binding experiments demonstrated that PP favors the formation of collagen aggregates, whereas in the presence of rDMP2 thin fibrils are formed. Overall these results suggested that the phosphate moieties in phosphophoryn are important for its function as a mediator of dentin biomineralization. [ABSTRACT FROM AUTHOR]

Full-length cDNA coding for dentin matrix protein 1 (DMP1) has been cloned and sequenced, but the corresponding complete protein has not been isolated. In searching for naturally occurring DMP1, we recently discovered that the extracellular matrix of bone contains fragments originating from DMP1. Shortened forms of DMP1, termed 37K and 57K fragments, were treated with alkaline phosphatase and then digested with trypsin. The resultant peptides were purified by a two-dimensional method: size exclusion followed by reversed-phase high performance liquid chromatography. Purified peptides were sequenced by Edman degradation and mass spectrometry, and the sequences compared with the DMP1 sequence predicted from cDNA. Extensive sequencing of tryptic peptides revealed that the 37K fragments originated from the NH[sub 2]-terminal region, and the 57K fragments were from the COOH-terminal part of DMP1. Phosphate analysis indicated that the 37K fragments contained 12 phosphates, and the 57K fragments had 41. From 37K fragments, two peptides lacked a COOH-terminal lysine or arginine; instead they ended at Phe[sup 173] and Ser[sup 180] and were thus COOH termini of 37K fragments. Two peptides were from the NH[sub 2] termini of 57K fragments, starting at Asp[sup 218] and Asp[sup 222]. These findings indicated that DMP1 is proteolytically cleaved at four bonds, Phe[sup 173]-Asp[sup 174], Ser[sup 180]-Asp[sup 181], Ser[sup 217]-Asp[sup 218], and Gln[sup 221]-Asp[sup 222], forming eight fragments. The uniformity of cleavages at the NH[sub 2]-terminal peptide bonds of aspartyl residues suggests that a single proteinase is involved. Based on its reported specificity, we hypothesize that these scissions are catalyzed by PHEX protein. We envision that the proteolytic processing of DMP1 plays a crucial role during osteogenesis and dentinogenesis. [ABSTRACT FROM AUTHOR]