J Oral Tissue Engin 2017;15(2):109-118

ORIGINAL ARTICLE Biochemical Mechanism of Titanium Fixation into Living : Acid Soluble Phosphoproteins in Bone Binds with Titanium and Induced Endochondral Ossification in vivo

Yoshinori KUBOKI1, 6, Kimitoshi YAGAMI2, Michiko TERADA-NAKAISHI3, Toshitake FURUSAWA4, Yasuko NAKAOKI5, and Masaaki KURASAKI6

1Professor Emeritus, Hokkaido University, Sapporo, Japan 2Department of Oral Implantology, Matsumoto Dental University, Shiojiri, Japan 3Department of Metallic Biomaterials, Institute of Biomaterials and Bioengineering, Tokyo Medical and Dental University, Tokyo, Japan 4School of Engineering, Yamagata University, Yamagata, Japan 5School of Dental Medicine, 6Graduate School of Environmental Science, Hokkaido University, Sapporo, Japan

SYNOPSIS In 2014, we discovered that bone phosphoproteins, which are collectively called SIBULING family of proteins, are equipped with titanium-binding ability. Further- more, the titanium implant devices which were coated with titanium-bound SIBLING induced more than 100 times faster bone formation of early stage when implanted into rat calvaria. These findings led us to an explanation why titanium implants could be fixed into living bone. Several other phosphoproteins including, phosvitin, caseins and phosphophoryn (a dentin phosphoprotein) were also found to bind with titanium by use of a chromatographic column packed with titanium beads. In this study we demonstrated that a typical phosphoprotein, phosvitin lost its titanium-binding ability in a time-dependent manner by the reaction with λ-protein phosphatase. The fact confirmed that certain specific phosphoserines λresidues in this protein were responsible for the titanium-protein interaction. For an additional confirmation of SIBLING-titanium interaction, we extracted bone and dentin proteins with a new and simple method of acidic condition and applied them to the chromatographic column packed with titanium beads. The results showed that definite portions of the acid soluble proteins from both bone and dentin were retained in the column. Electrophoretic analysis showed the retained fractions were Stains-all positive, indicating that both bone and dentin contain multiple phosphoproteins which have affinity with titanium. The titanium-bound fraction of acid extract was again coated on the titanium device and implanted into rat calvaria. After one week, histology showed that in addition to definite pattern of bone formation, process of endochondral ossification was clearly observed. In the control implant of uncoated titanium device, only collagenous tissues were observed, without any cartilage nor bone formation. Based upon these findings we reconfirmed that the core biochemical mechanisms underlying the strong bond between the titanium and living bone is based upon the interaction between the implanted titanium and multiple bone phosphoproteins in the host tissue.

Key words: Titanium implants, titanium-binding proteins, titanium- chromatography, phosphoproteins, phosvitin, bone phosphoproteins

109

Kuboki et al., Titanium-binding Phosphoproteins in Bone and Dentin

Introduction which eventually create a strong bond The phenomenon of the strong affinity between titanium and living bone. between titanium and living bone was To confirm above theory, in this accidentally discovered by Brånemark study we investigated whether the 1-2 and is now applied worldwide in Ti-binding SIBLING proteins are in a clinical applications in dental and relatively free soluble state or in a bound orthopedic fields3-5. However, the bio- state with insoluble matrix, as shown in chemical mechanism behind this unique our previous report 7. The acid soluble phenomenon has not been fully eluci- proteins in bovine bone and dentin were dated. We hypothesized that the core extracted by a new and simple method biochemical mechanisms may reside in using 0.01 M HCl (decalcifying solution). the interaction between certain proteins It was found that a small fraction of acid in the host tissues and the implanted extracted proteins bound to the titanium titanium. In previous studies to confirm in column chromatography. SDS poly- the interaction between titanium and acrylamide gel electrophoresis (PAGE) proteins, we chose the technique of analysis of the titanium-binding fractions chromatography using spherical beads from both bone and dentin indicated that of titanium packed into a column. We they contained phosphoproteins judging showed that whilst most of the albumin from their staining behavior with and lysozyme eluted with the break- Stains-all 8. Furthermore, the acid solu- through peak, indicating practically no ble titanium-binding proteins (AS-TiBP) affinity for titanium, on the other hand, from bone showed bone enhancing -casein, phosvitin, phosphophoryn and function, when implanted with TW into bone phosphoproteins, which are rat calvaria. Different from previous collectively called SIBLING family report, histology showed that in addition proteins exhibited a distinct retained to definite bone formation, local endo- peak separate from the breakthrough chondral ossification was clearly peak 6-7. When we coated the titanium observed. In the control implants of implant devices, a titanium web (TW) uncoated titanium device, only colla- (Zellez, Hi-Lex-Funakoshi, Japan) with genous tissues were observed, without the titanium-bound SIBLING fractions any cartilage nor bone formation. and implanted them into rat calvaria, it was found that more than 100 times 2. Materials and Methods faster bone formation occurred within a 1) Extraction of extracellular matrix week after implantation 7. proteins from bovine bone and From these findings, we concluded dentin that phosphate groups (mainly those of Total soluble components of bone and phospho-serine) in the phosphoproteins dentin matrix were extracted by a modi- is the key factor in the binding of these fied method reported previously8-10. proteins with titanium. Furthermore, Briefly, fresh bovine metatarsal bone phosphoproteins (SIBLINGS) are were cleaned of soft tissues and well known to contain a cell adhering pulverized in liquid nitrogen to obtain amino acid sequence (Arg-Gly-Asp) that powders, with a particle size smaller recruits osteoblast precursors. Together than the opening of a 60-mesh sieve8, 9. with the potent hydroxyapatite inducing Bovine dentin powders were prepared ability of numerous phosphate groups, as described10. Both bone and dentin SIBLING proteins immobilized on the powders were decalcified with 0.01 M surface of implanted titanium may HCl, maintained at pH 2 by continuous induce rapid bone formation there, addition of 6 M HCl at 5ºC, and decalci-

110

J Oral Tissue Engin 2017;15(2):109-118

fied and extracted in the same at pH 2. yolk, Sigma-Aldrich), 10 mg of the same The acid extracts from bone and dentin protein which had been treated with were immediately desalted by dialysis λ-protein phosphatase, and 40 mg each against distilled water at 5ºC until the pH of the lyophilized acid extracts from of the aliquot reached 5, and then bovine bone and bone. lyophilized. The acid extracts and the samples eluted following titanium 4) SDS polyacrylamide gel electro- chromatography were analyzed by phoresis (PAGE) SDS-PAGE electrophoresis. Ninety micrograms of each sample were dissolved in SDS sample buffer and 2) De-phosphorylation of phosvitin applied to a 15% polyacrylamide gel by λ-protein phosphatase containing 0.1% SDS or a 5-20% Ten milligrams each of phosvitin pre-cast gradient gel (Atto). Pre-stained (chicken egg yolk, Sigma-Aldrich, Tokyo, standard molecular weight markers Japan) were incubated with 4,000 units (Odyssey, Protein Molecular Weight of λ-protein phosphatase (Cell Signaling Marker, Li-Cor, NE, USA) and a stan- Technology, Tokyo, Japan) in 0.05M dard pepsin-treated type I from Tris-HCl/1M NaCl/2mM MnCl2 at 30 ºC, bovine skin (I-PC, Koken Co., Tokyo, for 2.5 and 6 h. Japan) were also applied on the gel. The electrophoresed gel was stained 3) Titanium-beads chromatography with 0.0025% Stains-all11, 0.25% Method for titanium-beads chroma- Coomassie Brilliant Blue (CBB) R-250 tograhy was reports previously6,7. Briefly, (Takara Bio Co., Ltd., Shiga, Japan). For pure titanium beads with an average Western blotting of the gel was done diameter 45 μm were obtained from a using anti-N-terminal (90-111) of DMP1 commercial supplier (Osaka Titanium antibody (Takara Bio). Technologies, Hyogo, Japan). Finer par- ticles were carefully removed by 5) Implantation into rat calvaria repeated decantation from the suspen- Discs form of (2 x 3 mm) of titanium sion in distilled water. The beads were webs (TW), (Zellez®, Funakoshi, Japan) packed into a commercial glass chro- were immersed in a solution of acid matography column (XK16/20, GE soluble titanium-binding proteins Health Care, Tokyo, Japan,) to obtain a (AS-TiBP) in PBS (1 mg/500 mL). The bed volume of 16 x 50 mm. The column TW/ AS-TiBP composites were air dried was eluted with 2 M urea in Dulbecco’s in a clean room and store at 5ºC. physiological buffered saline (PBS) at a Eight-week-old male rats (Wistar, flow rate of 120 ml/h using a peristaltic 200-230) were anesthetized with pen- pump (SJ-1211, Atto, Tokyo, Japan). tobarbital solution (3.6 mg body weight, Samples for analysis dissolved in 2 M Nembutal R, Dainippon Sugimitomo urea/PBS were applied to the column, Pharma, Osaka, Japan). The TW/AC- which were first eluted with 60 ml of 2 M TiBP composite were implanted into the urea in PBS, then with a straight gradi- holes of calvaria created by diamond ent system made by using 2 M bur and trepan bur. After implantation, urea/PBS and 25 mM NaOH/2 M urea periostea and skin were carefully placed (100 mL of each). Elution was monitored back and sutured. The protocol of the by spectroscopy using an automatic UV animal experiments was approved by monitor system at 254 nm (AC 5100, the Committee for Animal Experiments Atto). Samples applied to the column in Kanagawa Dental College. It was were: 10 mg of phosvitin (chicken egg carried out in accordance with guide-

111

Kuboki et al., Titanium-binding Phosphoproteins in Bone and Dentin

lines proposed by the Institutional 2) Binding of bone and dentin pro- Animal Care and Use Committee. One teins on titanium column piece of either the TW/AC-TiBP com- Results of titanium chromatography of posite or TW alone was implanted into the bone extracts, SDS PAGE profiles one rat calvaria. After one week and one their fractions stained with Stains-all and month, the samples were removed for the same profiles stained with CBB are examination as described previously4,7. shown in figures 2A, 2B and 2C, respectively. Most of the acid soluble 6) Histological observation proteins from bone applied to the Excised pieces of cranial bone, which column eluted at the breakthrough contained implant, were fixed in 10% fraction, indicating that they were not neutral formaldehyde, embedded in poly- adsorbed to the titanium. However, ester resin (Rigolac, Showa Denko, Tokyo, small secondary peaks were observed Japan), cut and polished into 80 m sec- after the gradient was elevated to about tion by Maruto Systems (Rigolac). Sec- 10 mM NaOH with the proteins from tions were stained with Cole’s hematoxylin 4, 7 both tissues, which indicates that and eosin and analyzed histologically . these protein fraction adsorbed to tita- nium with a certain affinity. RESULTS The SDS PAGE profile of the crude 1) Effect of de-Phosphorylation of extracts of bone which was stained with phosvitin Stains-all showed a broad band around Figure 1 shows titanium-beads chro- 50-60 KDa (lane 2 in figure 2B). The matography of intact phosvitin (10 mg) breakthrough fraction (lane 3 in figure and the same amount of the proteins 2B) similarly showed broad bands which were de-phosphorylated by ranging from 10 to 70 kDa. But in the λ-phosphatase for 2.5 h and 6 h. It was titanium-binding fraction (lane 3 in figure clearly shown that the second peaks 2B), two faint bands which were stained which had been retained in the titanium blue by Stains-all of about 40 to 70 kDa column were decreased as the reaction and several blue-stained bands of periods were increased, while the smaller molecular weight less than 20 breakthrough fractions which were not kDa were discernible. bound to the titanium beads under this As shown in figure 2C, the pattern chromatographic condition increased as of SDS PAGE of the crude extracts of shown by thinner lines in the chroma- bone (lane 2 in Fig. 2C) which was togram. stained with CBB showed a distinct double band at an approximate molecular weight of 100 kDa, which coincide with the -chains of standard bovine collagen (lane 5 in figure 2C), and diffused bands starting from 50 kDa to 20-25 kDa were observed (lane 2 in figure 2C). The breakthrough fraction likewise shows a double band at 100 kDa and diffused bands starting from 50 kDa to 20-25 kDa (lane 3 in figure 2C). However, in the titanium-binding fraction (lane 4 in figure 2C), a new clear band appeared at a slightly higher molecular Figure 1 Comparative titanium chromatograms of 10 weight than 50 kDa. mg of phosvitin (a), the same protein treated with λ-protein phosphatase for 2 h (b) and 6 h(c).

112

J Oral Tissue Engin 2017;15(2):109-118

Titanium chromatography, SDS PAGE profiles stained with Stains-all and the same profiles stained with CBB of acid soluble proteins from bovine dentin were shown in Figures 3A, 3B and 3C respectively. Result of the chromatography was similar to that of bone proteins (Fig. 2A), but in the dentin extract, relative amount of the retained fraction (titanium-binding) to break- through (non-binding) fraction seems to be larger than that of bone.

Figure 2 Titanium chromatography of acid soluble protein Figure 3 Titanium chromatography of acid soluble protein fractions from bovine bone (A), their SDS PAGE from bovine dentin (A) and their SDS PAGE pro- profiles stained with Stains-all (B), the same files stained with Stains-all (B) and the same profiles stained with CBB (C). DMP-1 was de- profiles stained with CBB (C). tected on nitrocellulose membrane with Lane 1 in Fig. 3B and C, molecular mass anti-DMP1 antibody by Western blotting (D, red standard, with their kDa in the left side; lane 2, box). the extracted crude proteins; lane 3, the break- Lane 1 in Fig. 2BC, molecular mass standard, through indicated by bar in A; lane 4, titanium- with their kDa in the left side; lane 2, the binding fractions indicated by bar in A; lane 5, extracted crude proteins; lane 3, the break- standard bovine skin collagen. through fraction indicated by bar in A; lane 4, titanium-binding fractions indicated by bar in A; lane 5, standard bovine type I collagen. 113

Kuboki et al., Titanium-binding Phosphoproteins in Bone and Dentin

The results of SDS PAGE of dentin pied by bone and cartilage to the total proteins stained with Stained-all (Fig. 3B) TW area in the AS-TiBP-coated group show that the crude extract and break- (65%) was 6.5 times higher than those through fraction show quite broad distri- of the uncoated TW group (10%). bution of blue-staining zone, the typical electrophoretic patterns observed in DISCUSSION dentin phosphoproteins. In the tita- Phosvitin is a highly phosphorylated egg nium-binding fraction (lane 4 in figure 3B) yolk protein, with 40 amino acid resi- however, two or three discrete bands dues, nearly half of which are phos- around 50 kDa were observed. phorylated serine12. We have previously The results of the SDS PAGE of den- shown that phosvitin binds to the tita- tin proteins stained with CBB (Fig. 3C) nium column and assumed that the show that the crude extract and break- phosphate groups of this protein were through fractions exhibit a similar pattern responsible for the binding. In order to with a double band at about 100 kDa confirm the mechanism, we attempted which corresponds with collagen (lane 5 in de-phosphorylation of this protein in this figure 3C), a distinct band at slightly study. Figure 1 indicates almost cer- higher molecular weight than 50 kDa, and tainly that the binding can be attributed a lower molecular weight band at 20-25 to phosphate groups of phosvitin, kDa. In the titanium-binding fraction (lane although direct chemical confirmation 4 in figure 3C) however, collagen bands may be still necessary. disappeared, but the distinct band at Mature bone tissue of vertebrate slightly higher molecular weight than 50 animals contains various non- kDa and a diffused band at 20-25 kDa collagenous matrix proteins, in addition bands still remained. to bone collagen and hydroxyapatite as the two major chemical components. It 3) Bone-enhancing effect in the is well established that many of these AS-TiBP by histological observation matrix proteins are phosphorylated and As shown in Figs 4A-F, the titanium play important roles in bone formation 3,9,13 webs (TW) which were coated with acid and metabolism . Since our soluble titanium-binding proteins hypothesis is that phosphoproteins in (TW/AS-TiBP) induced active bone and bone may bind to titanium and function cartilage in the whole areas of TW, to promote bone formation on the sur- except a corner part of TW. In enlarged face of the implanted titanium biomate- images (Fig. 4C-E), clear distinction rial, the first step of the verification may between cartilage (left side) and bone be detection of titanium-binding (right side) were observed. At the high- phosphoprotein in bone. For this est magnification (Fig. 4F), many longi- purpose, we chose an easily extractable tudinal sections of capillaries were seen fraction of the proteins, which is the 0.01 near the titanium fibers. It was noted M HCl soluble fraction, because they that cartilage formations were limited in may be more easily interact with the area where the density of titanium implanted titanium biomaterials, rather fibers were relatively low, while the bone than more insoluble bone matrix formation with capillaries are located in proteins, which are only solubilized with the area where density of titanium wires dissociating solvents such as 4 M were constantly higher density. Effect of guanidine HCl. So, we have previously bone enhancement by the AS-TiBP- shown the presence of titanium-binding coating on TW was estimated histomet- proteins which enhanced bone rically. Average ratio of the areas occu- formation in 2M urea soluble fraction from bone 7.

114

J Oral Tissue Engin 2017;15(2):109-118

Figures 4 Histological observation of the composites of TW coated with acid soluble titanium-binding proteins (TW/AS-TiBP) (Figs. 4A-F) and their uncoated control TW (Figs. 4G-J) retrieved one week after implantation TW/AS-TiBP composites (Figs. 4A-F) induced active new bone (indicated by the ellipse at right side of Fig. 4D) and cartilage formation (indicated by the ellipse at left side in Fig. 4D). In contrast, uncoated TW control (Fig. 4G and H) induced no bone nor cartilage formation. Titanium fibers (50 micron in width) in the TW were seen black structures. Burs indicate 1 mm in Figs. 4 A, B, G and H; and 100 m in Figs. C, D, I and J, and 50 m in Figs. E and F.

115

Kuboki et al., Titanium-binding Phosphoproteins in Bone and Dentin

Figure 2A and 2B show there is a be noted that all of these phosphopro- small amount (10-20% of total acid teins have the cell-adhesive functions soluble proteins) of titanium-binding acid through their integrin-binding sequence soluble proteins in bone, which were RGD, and the functions in mineralization stained positively with Stains-all, indi- through their phosphate groups 6, 9. To cating the proteins are phosphorylated. our knowledge, there are little or no A similar situation was observed in acid systematic study of bone and dentin soluble dentin proteins. From these re- proteins from the view-point of sults it is reasonable to assume that the titanium-binding ability. This study gives titanium-binding protein at least partly the first evidence that there are consid- contain phosphoprotein, judging from erable amounts of titanium-binding their staining behavior with Stains-all. It phosphoproteins in the acid soluble should be noted that the distinct bands form (AS-TiBP) in bone and dentin. In with approximate molecular weight of 50 addition, when the AS-TiBP coated tita- kDa were found in both bone (lanes 4 in nium device (TW) was implanted into figure 2C) and dentin extracts (lanes 4 the rat calvaria, 6.5 times higher amount in figure 3B and 3C). of bone was observed than the control Dentin matrix protein 1 (DMP1) was implantation of uncoated TW. This first identified by cDNA cloning using rat bone-enhancing effect is comparably odontoblast mRNA14 and later detected high with the previous results when we in bone. Recent study on the post- observed on the TiBP which was translational proteolytic processing of extracted with 2 M urea from bone. DMP1 showed that there are two frag- ments of DMP1 with 37 and 57 kDa CONCLUSION fragments, the former being the Enzymatic dephosphorylation of phos- N-terminal and the latter C-terminal part vitin by using λ-protein phosphatase of DMP1, respectively15. Their study clearly reduced the titanium-binding suggested us that one of the candidates ability of this phosphoprotein in a for the titanium-binding proteins with reaction-time dependent manner. The approximate molecular weight of 50 kDa phenomenon clearly verifies the cause in lanes 4 in figures 2C, 3B and 3C may of titanium binding ability of phospho- include DMP1. Furthermore, immu- protein are phosphate group of the nological staining of this 50 kDa band phosphoserine in phosvitin. with anti-DMP antibody revealed clear Not only in the 2 M urea extracts as double bands, as shown in Fig. 2D, shown in the previous reports7, but also confirming the identification. Detailed in the acid decalcification solution, tita- identification of other titanium-binding nium-binding phosphoproteins were proteins in bone and dentin is still detected by titanium chromatography ongoing. and SDS-electrophoresis (notably The group of non-collagenous DMP1), showing the presence of rela- phosphoproteins was categorized as tively free state of titanium-binding members of a family known as SIBLING (AS-TiBP) which are ready to SIBLINGs (Small Integrin-Binding bind with implanted titanium. Further- Ligand N-linked Glycoproteins)16-17, more, AS-TiBP showed a remarkable which include osteopontin (OPN), bone bone- enhancing effect when implanted sialoprotein (BSP), dentin sialophos- with titanium device. Both results con- phoprotein (DSPP), dentin matrix pro- firmed the definite role of titanium- tein 1 (DMP1) and matrix extracellular binding phosphoproteins in titanium phosphoglycoprotein (MEPE). It should implant fixation in bone.

116

J Oral Tissue Engin 2017;15(2):109-118

ACKNOWLEDGEMENTS 8) Kuboki Y, Takita H, Kobayashi D, Tsu- ruga E, Inoue M, Murata M, Nagai N, Authors appreciate valuable advice Dohi Y, Ohgushi H 1998 BMP-induced given by Dr. Arthur Veis, Professor osteogenesis on the surface of hy- Emeritus, North Western University, droxyapatite with geometrically feasible USA, and Associate Professor Rachel and nonfeasible structures: topology of osteogenesis J. Biomed. Mater. Res. 39 Sammons, Birmingham University, UK. 190-199. This study was supported in part by a 9) Kobayashi D, Takita M, Mizuno M, Tot- Grant-in-Aid for Challenging Exploratory suka Y, Kuboki Y. Time-dependent ex- Research (23659907) and a pression of bone sialoprotein fragments in osteogenesis induced by bone Grant-in-Aid for Research (C) morphogenetic protein. J Biochem. (15K05554) from the Ministration of 1996; 119: 475-481. Education, Culture, Sports and Tech- 10) Kuboki Y, Fujisawa R, Aoyama K, Sa- nology in Japan. saki S. Calcium-specific precipitation of dentin phosphoprotein: a new method of purification and the significance for the REFERENCES mechanism of calcification J Dent Res 1) Brånemark PI, Adell UR, Breine U, 1979; 58: 1926-1932. Hansson BO, Lindström J, Ohlsson A. 11) Campbell KP, MacLennan DH, Jorgen- Intra-osseous anchorage of dental sen AO 1983 Staining of the prostheses I Experimental studies Ca2+-binding proteins, , Scand. J Plast Reconstr Surg 1969; 3: , , and S-100, with 81-100 the cationic carbocyanine dye 2) Brånemark PI, Adell UR, Breine U, "Stains-all". J Biol Chem 1983; 258: Hansson BO, Lindström J. Repair of 11267-11273. defects in mandible. Scand J Plast Re- 12) Finn RN. Vertebrate yolk complexes constr Surg 1970: 4: 100-108. and the functional implications of phos- 3) Kuboki Y, Yagami K, Furusawa T. New vitin and other subdomains in vitel- principles of regenerative medicine: logenin. Biol Reproduc 2007; 76: 926– special reference to mechano-dynamic 935 factors. J Oral Tissue Engin 2017; 14: 13) He G, Ramachandran A, Dahl T, 128-152. George S, Schultz D, Cookson D, Veis 4) Kuboki Y, Iku S, Yoshimoto R, Kaku T, A, George A. Phosphorylation of phos- Takita H. Modification of titanium sur- phophoryn is crucial for its function as a faces based on the principles of ge- mediator of . J Biol ometry of the artificial extracellular ma- Chem 2005; 280: 33109-33114. trix (ECM), in: Tanaka J, Itoh S and 14) George A, Sabsay B, Simonian PA, Chen G (Eds) Surface Design and Veis A. Characterization of a novel den- Modification of the Biomaterials for tin matrix acidic phosphoprotein. Impli- Clinical Application, Transworld Re- cations for induction of biomineralization search Network, Kerala, India, 2008, pp J Biol Chem 1993; 268: 12624-12630. 1-27. 15) Qin C, Brunn JC, Cook RG, Orkiszewski 5) Brunette DM, Tengvall P, Textor M, RS, Malone JP, Veis A, Butler WT. Thomsen P. Titanium in Medicine Evidence for the proteolytic processing Springer, Berlin, 2001, pp 457-947. of dentin matrix protein 1. Identification 6) Kuboki Y, Furusawa T, Sato M, Sun Y, and characterization of processed Unuma H, Fujisawa R, Abe S, Akasaka fragments and cleavage sites. J Biol T, Watari F, Takita H, Sammons R. In- Chem 2003; 278: 34700-34708. teraction between titanium and phos- 16) Bellahcene A, Castronovo V, Ogbureke phoproteins revealed by chromatogra- KU, Fisher LW, Fedarko NS. Small in- phy column packed with titanium beads. tegrin-binding ligand N-linked glycopro- Biomed Mater Eng 2012; 22: 283-288. teins (SIBLINGs): multifunctional pro- 7) Kuboki Y, Furusawa T, Sato M, Sun Y, teins in cancer. Nat Rev Cancer 2008; Unuma H, Abe S, Fujisawa R, Akasaka 8: 212-226. T, Watari F, Takita H, Sammons R. 17) Huang B, Sun Y, Maciejewska I, Qin D, Bone enhancing effect of tita- Peng T, McIntyre B, Wygant J, Butler nium-binding proteins isolated from bo- WT, Qin C. Distribution of SIBLING vine bone and implanted into rat cal- proteins in the organic and inorganic varia with titanium scaffold. Bio-Med phases of rat dentin and bone Eur. J Mater Eng 2014; 24: 1539–1548. Oral Sci 2008; 116: 104-112.

117

Kuboki et al., Titanium-binding Phosphoproteins in Bone and Dentin

(Received, November 15, 2017/ Accepted, December 26, 2017)

Corresponding author: Yoshinori Kuboki, Ph.D. Professor Emeritus, Graduate School of Environmental Science, Hokkaido University, N-10, W-5, Kita-ku, Sapporo, 060-0810, Japan. Tel: +81 11 706 2252 Fax: +81 11 706 4864 E-mail: [email protected]

118