Chick embryo anchored alkaline phosphatase and mineralization process in vitro Influence of Ca 2+ and nature of substrates Eva Hamade, Ge ´ rard Azzar, Jacqueline Radisson, Rene ´ Buchet and Bernard Roux Laboratoire de Physico-Chimie Biologique, UMR CNRS 5013, Universite ´ Claude Bernard, Lyon I, Villeurbanne, France Bone alkaline phosphatase with glycolipid anchor (GPI- bALP) from chick embryo femurs in a medium without exogenous inorganic phosphate, but containing calcium and GPI-bALP substrates, served as in vitro model of min- eral formation. The mineralization process was initiated by the formation of inorganic phosphate, arising from the hydrolysis of a substrate by GPI-bALP. Several minerali- zation media containing different substrates were analysed after an incubation time ranging from 1.5 h to 144 h. The measurements of Ca/P i ratio and infrared spectra permitted us to follow the presence of amorphous and noncrystalline structures, while the analysis of X-ray diffraction data allowed us to obtain the stoichiometry of crystals. The hydrolysis of phosphocreatine, glucose 1-phosphate, glucose 6-phosphate, glucose 1,6-bisphosphate by GPI-bALP pro- duced hydroxyapatite in a manner similar to that of b-glycerophosphate. Several distinct steps in the mineral formation were observed. Amorphous calcium phosphate was present at the onset of the mineral formation, then poorly formed hydroxyapatite crystalline structures were observed, followed by the presence of hydroxyapatite crys- tals after 6–12 h incubation time. However, the hydrolysis of either ATP or ADP, catalysed by GPI-bALP in calcium- containing medium, did not lead to the formation of any hydroxyapatite crystals, even after 144 h incubation time, when hydrolysis of both nucleotides was completed. In contrast, the hydrolysis of AMP by GPI-bALP led to the appearance of hydroxyapatite crystals after 12 h incubation time. The hydroxyapatite formation depends not only on the ability of GPI-bALP to hydrolyze the organic phosphate but also on the nature of substrates affecting the nucleation process or producing inhibitors of the mineralization. Keywords: anchored bone alkaline phosphatase; minerali- zation; Ca 2+ ; hydroxyapatite. Alkaline phosphatase (bALP) (EC 3.1.3.1) is one of the most frequently used biochemical markers of osteoblast activity [1–4]. The role of bALP in mineral formation was evidenced in the case of hypophosphatasia, an inheritable disorder leading to a defective bone formation and a deficiency of bALP [5]. Mice deficient in the tissue nonspecific bALP gene mimic a severe form of hypophos- phatasia, indicating the importance of bALP in initiating mineral formation [6]. It has been suggested that bALP could be involved in the mineralization process by hydro- lyzing organic phosphates to release free inorganic phos- phate at sites of mineralization [7]. However, it is still not clear which organic phosphates are hydrolyzed by bALP. Osseous bALP, localized in the matrix vesicles, exist as a phosphatidylinositol-glycolipid (GPI) anchored protein [8,9]. Mineralization is initiated within and at the surface of extracellular matrix vesicles derived from osteoblasts [10]. Bone is constantly destroyed or resorbed by the osteoclasts and then replaced by the osteoblasts [11]. Poorly crystal- line hydroxyapatite [HA/Ca 10 (PO 4 ) 6 (OH) 2 ;Ca/P i molar ratio ¼ 1.67] is the major component of bone and other calcified tissues [12]. The maturation of the initial deposits of a solid phase of calcium phosphate in bone and other changes that occur with time, were investigated by using in vitro models, matrix vesicles or osteoblastic cell-cultures [13–34,34–40]. The most common precursors of HA are amorphous calcium phosphate [ACP/Ca 3 (PO 4 ) 2 , XH 2 O; Ca/P i molar ratio ¼ 1.5] [13–15], brushite [(CaHPO 4 ) 3 Æ 2H 2 O; Ca/P i molar ratio ¼ 1] [16] and octacalcium phos- phate [OCP/Ca 8 (PO 4 ) 6 H 2 ;Ca/P i molar ratio ¼ 1.33] [17], although the presence of OCP, brushite [18,19] was not reproducible and appeared most likely due to preparative artefacts [20]. Despite the remaining difficulties in charac- terizing the composition of mineral samples that initiate mineralization within osteoblast cell-cultures [25–27,32,33, 35,38,40] and matrix vesicles [28–31,34], these models permitted a better delineation of the roles of various proteins such as type III sodium-dependent phosphate transporter [26], type I collagen [28], type II and X collagen Correspondence to G. Azzar, Laboratoire de Physico-Chimie Biologique, UMR CNRS 5013, Universite ´ Claude Bernard, Lyon 1, 6 rue Victor Grignard, Baˆ timent Euge ` ne Chevreul, 69622 Villeurbanne Cedex, France. Fax: + 33 4 72 43 15 43, Tel.: + 33 4 72 44 83 24, E-mail: Gerard.Azzar@univ-lyon1.fr Abbreviations: ACP, amorphous calcium phosphate; bALP, bone alkaline phosphatase; GPI, phosphatidylinositol-glycolipid; GPI-bALP, bone alkaline phophatase with glycolipid anchor; HA, hydroxyapatite; OCP, octacalcium phosphate; pNPP, para- nitrophenylphosphate; b-GP, b-glycerophosphate. Enzyme: Alkaline phosphatase (EC 3.1.3.1). (Received 4 February 2003, accepted 19 March 2003) Eur. J. Biochem. 270, 2082–2090 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03585.x [30,32] bALP [29], annexin II [30], annexin V [28,30] and annexin VI [30]. Very often an organic phosphate com- pound, b-glycerophosphate (b-GP) was added in the medium (in vitro models, matrix vesicles and cell cultures) to stimulate the mineralization process [25,31–34,34–40]. This exogenous organic phosphate compound is a substrate for bALP, producing an increase in phosphate content in the medium. To our knowledge, no other organic phosphate than b-GP was supplemented in the mineralization medium, with the exception of ATP [41]. The aim of this work was to examine whether other organic phosphate compounds, present in the cells at much higher concentrations than b-GP, could be the substrates for phosphatidylinositol-glycolipid anchored alkaline phosphatase (GPI-bALP), producing inorganic phosphate and initiating the mineralization process. Our results demonstrate that crystal formation obtained in the presence of GPI-bALP (extracted from chick embryo femurs), without supplementing exogenous P i , depended on the type of substrates and incubation conditions. Determination of Ca/P i molar ratio, IR spectroscopy and X-ray analysis revealed that the hydrolysis of substrates, such as AMP, b-GP, phosphocreatine, glucose 1-phos- phate, glucose 6-phosphate and glucose 1,6-bisphosphate, by GPI-bALP, produced inorganic phosphate and initi- ated the formation of HA. In contrast to these results, the hydrolysis of other substrates such as ADP and ATP by GPI-bALP, was not accompanied by the formation of HA. The nature of the substrates hydrolyzed by GPI- bALP, under in vitro conditions, can influence the onset and the formation of HA. Whether this property or age- related change of organic phosphate content could influence the first steps of matrix vesicle-induced miner- alization under in vivo conditions remains to be investi- gated. Materials and methods Materials Chick embryo femurs were isolated from 17-day-old-eggs obtained from a local producer. HA, N-octyl b- D -gluco- pyranoside, para-nitrophenylphosphate (pNPP), phospho- creatine, glucose 6-phosphate, glucose 1-phosphate, glucose 1,6-bisphosphate, nucleotides, protease inhibitors (leupep- tin, pepstatin, phenylmethylsulfonyl fluoride, benzamidin), were purchased from Sigma Chemical Co., St. Louis, MO, USA. AcA 202 and Octyl Sepharose were obtained from Pharmacia San Diego, CA, USA. All others reagents were of the highest purity commercially available. Preparation and purification of femur chick embryo GPI anchored alkaline phosphatase Membrane bound femur chick embryo alkaline phospha- tase (GPI-bALP) was prepared according to the procedure described by Radisson et al. [42]. Briefly, femurs were removed immediately, rinsed in phosphate buffer saline containing 8 m M Na 2 HPO 4 , 1.5 m M KH 2 PO 4 pH 7.4, 120 m M NaCl and 1.35 m M KCl. They were homogenized in 0.1 M Tris/HCl, pH 8.5, with 1 m M MgCl 2 ,1l M leupeptin, 0.72 l M pepstatin, 1 m M benzamidin and 0.25 m M phenylmethanesulfonyl fluoride, using a Virtis homogenizer for 5 min at 4 °C. The homogenate obtained was centrifuged at 90 000 g, for 180 min at 4 °C. GPI- bALP was extracted from the pellet, with 60 m M N-octyl b- D -glucopyranoside for 60 min at 4 °Cin0.1 M Tris/HCl, pH 8.5, 1 m M MgCl 2 ,1l M leupeptin, 0.72 l M pepstatin and 1 m M benzamidin. After centrifugation (90 000 g, 30 min), in the first step the surfactant-solubilized enzyme was further purified by gel filtration chromatography (ACA 202) equilibrated in 0.1 M Tris/HCl pH 8.5, 0.1 M NaCl and 1m M MgCl 2 . In the second step, GPI-bALP was applied on octyl Sepharose (CL4B) column containing 25 m M glycine pH 9 and 1.5 M Na 2 S0 4 . Elution was followed by a gradient of decreasing sodium sulfate concentration (1.5–0 M )and increasing 1-propanol concentration (0–40%, respectively). The GPI anchored enzyme fraction was dialysed against 100 m M Tris/HCl pH 7.4 and 1 m M MgCl 2 overnight. Protein determination Protein concentration was determined by Lowry’s method [43] after precipitation of proteins by 10% trichloracetic acid. It was followed by washing the precipitated material with methanol and methylal/methanol mixture (4 : 1, v/v) and dissolution in 1 M NaOH [44] Bovine serum albumin was used as a standard. A second method was used for small quantities of proteins based on Bradford’s method [45] modified by Read and Northcote [46]. Alkaline phosphatase activity GPI-bALP activity was measured according to the method of Cyboron and Wuthier [47], at 37 °C by means of a Uvikon 810 spectrophotometer. The reaction mixture contained 10 m M pNPP and 25 m M glycine buffer, pH 10.4. One unit of GPI-bALP activity (U) is defined as the amount of enzyme that is required to hydrolyze 1 lmol of pNPP per min at 37 °C. Mineralization process Purified GPI-bALP (0.8 U) was added in a mineralization solution containing Ca 2+ (2.5, 6, 7.5 and 24 m M , respect- ively), 62.5 l M ZnCl 2 , 62.5 l M MgCl 2 , 37.5 m M NaCl, 0.05% NaN 3 and 25 m M Tris, pH 7.3. It was incubated at 37 °C for up to 6 days according to Golub et al. [31] and Harrrison et al. [48]. However, the mineralization medium did not contain inorganic phosphate to initiate the mineralization process. Instead of inorganic phosphate, 12.5 m M of substrate for GPI-bALP (either b-GP, glucose 6-phosphate, glucose 1-phosphate, glucose 1,6-bisphos- phate, AMP, ADP, ATP, or phosphocreatine) was added in the mineralization medium. Controls without GPI- bALP were performed under the same conditions. After incubation at the indicated time, samples were centrifuged (2000 g, 5 min). The mineral pellet was washed twice with deionized water, and centrifuged under the same condi- tions. Another washing was performed with a chloroform/ methanol (2 : 1, v/v) mixture. After centrifugation, pellets (mineral samples) were dried under N 2 flow. Their calcium/ phosphate ratio, IR spectra and X-ray diffraction were measured. Ó FEBS 2003 Bone GPI-bALP, roles in mineralization (Eur. J. Biochem. 270) 2083 Calcium and phosphate determination Mineral samples were treated with 1% HCl. Calcium was determined by atomic absorption spectroscopy (Perkins- Elmer 3110 spectrometer, atomization temperature: 1700– 3151 °C, wavelength: 422.7 nm, nitrous oxide-acetylene flame). To avoid refractory aggregates, EDTA was added to the solution to chelate calcium, thereby preventing a reaction with phosphate. Phosphorus assay was performed using the method of Chen et al. [49]. Fourier-transform infrared spectroscopic measurements KBr pellets (100 mg) containing 1–2 mg of mineral samples obtained during the mineralization process were analysed using IR spectroscopy. IR spectra were recorded, using a Nicolet 510 M FTIR spectrometer equipped with a DTGS detector. Two hundred and fiftysix interferograms were measured, Fourier transformed to yield infrared spectra at 4cm )1 resolution. Each IR spectrum is representative of at least three independent measurements. During data acqui- sition, the spectrometer was continuously purged with dried filtered air (Balston regenerating desiccant dryer, model 75-45 12 VDC). X-ray diffractometry Mineral samples were analysed with a Siemens D500 diffractometer using cooper Ka radiation and a highly crystalline mineral hydroxyapatite as a standard. Diffrac- tometer was equipped with a scintillation counter and a geometrical goniometer (Bragg–Breteno). Diffraction angle 2h is comprised between 10° and 70°. The voltage and current intensity were 40 kV and 30 mA, respectively. X-ray analysis was performed in the Henri Longchambon diffrac- tometer centre in the University of Lyon I, Villeurbanne, France. Results Evidence of different steps in the GPI-bALP mediated mineralization: effects of Ca 2+ concentration To assess the role of Ca 2+ in initiating the mineralization process, we modelled the biological process by using 0.8 U of purified GPI-bALP in a mineralization medium without exogenous phosphate but containing 12.5 m M b-GP as phosphate substrate. GPI-bALP was incubated at 37 °C in this medium at different concentrations of Ca 2+ for 1.5– 144 h. At various time intervals, the Ca/P i ratios were measured. A theoretical Ca/P i molar ratio of 1.67 (which is the highest among the various calcium-phosphate com- plexes) corresponds to HA. When 2.5 m M Ca 2+ was added in the incubation medium, no calcium-phosphate minerals were obtained during 72 h of incubation. The Ca/P i ratio was 0.8 at 144 h, characteristic of a poorly crystalline structure. When the Ca 2+ concentration increased from 2.5 m M to 6–7.5 m M , the mineralization began at 48 h. At 144 h the Ca/P i ratio was % 1.2–1.3. At 24 m M Ca 2+ , HA crystals were obtained at about 12 h with a Ca/P i ratio of 1.5 in the presence of anchored GPI-bALP (Fig. 1). The presence of HA in the mineral samples was confirmed by measuring their IR spectra and X-ray diffraction patterns (see below). In all cases, medium lacking GPI-bALP did not produce any mineral material. Influence of other phosphate substrates in the mineralization process To examine the effects of substrates catalysed by GPI-bALP in the production of inorganic phosphate and in the mineralization process, different phosphate compounds were used: b-GP, phosphocreatine, phosphate sugars and nucleotides. In all cases, the mineralization medium without exogenous phosphate contained 24 m M Ca 2+ , 12.5 m M phosphate substrate and 0.8 U of GPI-bALP. In the presence of phosphocreatine, the incubation time necessary to obtain a Ca/P i characteristic of HA was 24 h. Its Ca/P i ratio was the highest (Fig. 2A) of those obtained with different phosphate substrates. When glucose 1-phosphate was used as a substrate, HA mineral was formed within 12 h of incubation (Ca/P i ratio, 1.32). With glucose 6-phosphate, glucose 1,6-bisphosphate, the incubation time to produce HA was longer, i.e. 24 and 72 h, respectively; Ca/P i ratios were, respectively, 1.35 and 1.12 (Fig. 2B). Results obtained in the presence of nucleotides were different. AMP gave crystal identified as HA. However, the incubation time necessary for the mineralization process was at least 72 h and the Ca/P i ratio was 1.38. ATP and ADP did not produce any organized mineral structure and their Ca/P i ratio did not correspond to hydroxyapatite crystals (Fig. 2Ca,b). Although GPI-bALP from chick embryo femurs hydrolyzed ATP and ADP, the products of hydro- lysis did not lead to the formation of HA. No mineral material was formed in media lacking GPI-bALP. Mineral calcium phosphate identification by infrared spectra To identify the quality and the type of mineral formation obtained in the mineralization medium during the GPI- bALP-induced hydrolysis of different phosphate substrates Fig. 1. Ca/P i ratio (m M /m M ) during in vitro mineralization for different Ca 2+ concentrations in the presence of b-GP as substrate (12.5 m M ). Incubation containing equal activities of purified GPI-bALP (0.8 U) without PO 4 3– . Mineralization solution is described in Materials and methods. At each time point (0–144 h), an aliquot was taken and centrifuged. Ca 2+ and P i were determined in the pellets corresponding to the mineral structure. Symbols corresponding to various Ca 2+ concentrations are: n, Ca 2+ 2.5 m M ; e, Ca 2+ 6m M ; h, Ca 2+ 7.5 m M ; s, Ca 2+ 24 m M . 2084 E. Hamade et al. (Eur. J. Biochem. 270) Ó FEBS 2003 (b-GP, phosphocreatine, phosphate sugars and nucleo- tides), mineral deposits were analysed by IR spectroscopy as a function of incubation time ranging from 1.5 h to 144 h. The phosphate bands (m1, m3andm4) were used to monitor the formation of mineral deposits. The absorption in the 550–650 cm )1 region arises primarily from the antisymmet- ric P-O bending mode (m4), while the absorption around the 900–1200 cm )1 region is characteristic of symmetric (m1) and antisymmetric (m3) P-O stretching mode [12]. Correla- tions between IR spectral features and phosphate phases are presented in Table 1. These correlations facilitated the interpretation of IR spectra of mineral deposits. Distinct mineralization steps were observed during the GPI-bALP- induced hydrolysis of b-GP. At incubation times in the range 1.5–3 h, two bands located at 1053 cm )1 and 1040 cm )1 were observed (Fig. 3, top trace). The 1053 cm )1 band (Fig. 3, top trace) found at early incubation time suggested the presence of a disordered phosphate environment, characterized by a 1060 cm )1 band [50]. Alternatively, this band may correspond to ACP, usually identified by a m3 band at 1052–1060 cm )1 [51,52] (Table 1). The component band, located at 1040 cm )1 (Fig. 3, top trace) was tentatively assigned to a m3 vibrational mode that Fig. 2. Molar Ca/P i ratio during in v it ro mineralization in the presence of 12.5 m M phosphate substrates as indicated below and 24 m M Ca 2+ . Experimental conditions are identical to those presented in Fig. 1. (A) r, 12.5 m M creatine phosphate. (B) 12.5 m M Phosphate sugars. h, glucose 1-phosphate; r, glucose 6-phosphate; n, glucose 1,6-bis- phosphate. (C) (a) n, 12.5 m M AMP; r, 12.5 m M ADP; (b) r, 12.5 m M ATP. Table 1. Assignment of some spectral features in m1, m3 and m4 phosphate regions, used to interpret infrared spectra of mineral samples. Mineral complexes Wavenumber (cm )1 ) Origin Ref. HPO 2À 4 1125–1145 HPO 2À 4 stretch [57] HPO 2À 4 in OCP 918–924 HPO 2À 4 stretch [51] ACP 1052–1060 m3 [51,52] Amorphous OCP 587–600 m4 [56] 1038 m3 [53] Brushite 1127 PO 3 stretch in HPO 4 2– [53] H-bonded OH group 741 OH libration [59,60] HA 561–563 m4 [57] 575 m4 [57] 601–603 m4 [57] 960 m1 [57] 1032 m3 [57] 1092–1094 m3 [57] Poorly crystalline HA 1112–1116 m3 [57] Fig. 3. Time evolution of the m1, m3andm4 bands of the IR spectrum during mineralization process from the ACP (amorphous calcium phos- phate) to HA (hydroxyapatite), in the presence of b-glycerophosphate. For experimental conditions, see Materials and methods. Ó FEBS 2003 Bone GPI-bALP, roles in mineralization (Eur. J. Biochem. 270) 2085 was not part of the apatite structure as in the case of OCP, where a component band absorbs at 1038 cm )1 [53]. However, our X-ray data did not indicate the presence of crystalline OCP. After 12 h incubation time, HA was identified. The m3 bands shifted to 1031–1032 cm )1 and to 1092 cm )1 (Fig. 3, middle and bottom traces), typical for HA [54,55] (Table 1). In addition, m4 bands were resolved into two distinct peaks located at 563 cm )1 and 602 cm )1 , with a shoulder around 575 cm )1 (Fig. 3), confirming the presence of HA [50,56,57]. The appearance of a resolved m1 band located at 960 cm )1 (Fig. 3) is indicative of crystallized HA [50] but its intensity is not proportional to crystal size [50,58]. The 741 cm )1 band (top traces in Fig. 3) that disappeared once crystalline apatites were formed is prob- ably associated with hydrogen-bonded OH libration [59,60] (Table 1). The mineral formations induced by the GPI- bALP-induced hydrolysis of phosphocreatine, glucose 1-phosphate, glucose 6-phosphate, glucose 1,6-bisphosphate as a function of incubation time were similar to that produced by the GPI-bALP-induced hydrolysis of b-GP. In all cases, the appearance of characteristic 1108–1092 cm )1 , 1032–1031 cm )1 , 960 cm )1 , 602–601 cm )1 and 563– 561 cm )1 bands (Fig. 4A–D), confirmed the presence of HA [50,56,57]. Nucleotides as phosphorous substrates can also be hydrolyzed by alkaline phosphatase. There was no crystalline HA formation during the GPI-bALP-induced hydrolysis of ATP (Fig. 5A) or ADP (Fig. 5B) even after 144 h incubation and complete hydrolysis of nucleotide by GPI-bALP. However, formation of HA was observed at around 144 h during hydrolysis of AMP by GPI-bALP (Fig. 5C). The formation of HA induced by hydrolysis of AMP by GPI-bALP was relatively slow. ADP hydrolysis by GPI-bALP in the mineralization medium, produced inor- ganic phosphate which led to a mineral deposit. The IR spectrum of the mineral deposit was different from that of HA. Two bands located at 558 cm )1 and at 918 cm )1 (Fig. 5B) suggested the presence of HPO 4 2– group in OCP [61] (Table 1). The other bands may reveal different types of calcium-phosphate deposits. ATP hydrolysis by GPI-bALP in the mineralization medium produced a mineral deposit containing also HPO 4 2– . The IR spectrum of this mineral deposit exhibited two bands located at 1122 cm )1 and 918 cm )1 bands (Fig. 5A). As mentioned above, the 918 cm )1 band may correspond to HPO 4 2– group in OCP. The 1122 cm )1 band (Fig. 5A) may be associated with an Fig. 4. IR spectra of mineral samples obtained after 144 h incubation in the mineralization medium containing: (A) 12.5 m M creatine phosphate; (B) 12.5 m M glucose 6-phosphate; (C) 12.5 m M glucose 1-phosphate; (D) 12.5 m M glucose 1,6-biphosphate. For experimental conditions, see Materials and methods. Fig. 5. IR spectra of mineral samples obtained after 144 h incubation in the mineralization medium containing: (A) 12.5 m M ATP; (B) 12.5 m M ADP; (C) 12.5 m M AMP. For experimental conditions, see Materials and methods. 2086 E. Hamade et al. (Eur. J. Biochem. 270) Ó FEBS 2003 acid-phosphate environment or brushite (Table 1). The 556 cm )1 band and the faint shoulder at 600 cm )1 (Fig. 5A) suggested the presence of octacalcium phosphate [56] in addition to other types of complexes. Based on the inter- pretation of its IR spectrum, the mineral deposit obtained after hydrolysis of ATP contains a mixture of distinct types of mineral formation such as brushite, OCP and other phosphate species. Indeed, its IR spectrum is different from the IR spectrum of a single chemical species such as brushite, OCP or amorphous phosphate [50,51,56,62,63]. X-ray diffraction analysis Mineral crystalline structures were also identified by X-ray diffraction, after 144 h incubation of one of the eight phosphate substrates (12.5 m M ), in mineralization medium containing calcium (24 m M ) in the presence of purified chick embryo GPI-bALP. X-ray spectra corresponding to 144 h incubation time are presented in Fig. 6. The data shown are in agreement with those obtained by IR spectra. Indeed, crystalline structures observed after 144 h incubation of one of these substrates, b-GP (Fig. 6A), phosphocreatine (Fig. 6A), glucose 1-phosphate (Fig. 6B), glucose 6-phos- phate (Fig. 6B), glucose 1,6-bisphosphate (Fig. 6B) were unambiguously identified as HA in comparison with HA standard. The 144 h incubation of the mineralization medium containing either ATP or ADP, as a putative phosphate donor, did not induce any organized crystalline structure, as evidenced by the lack of X-ray characteristic bands of HA (Fig. 6C). In contrast to this result, when AMP was present in the mineralization medium and after 144 h incubation, the X-ray spectrum revealed the presence of crystalline HA (Fig. 6C). Discussion Organic phosphate compounds that induce mineralization b-GP has been routinely supplemented in bone cells, chondrocytes and cultured cell types to initiate mineraliza- tion and to better monitor different aspects of mineral formation [25,31–40,58,64–68]. The mechanism of minerali- zation caused by the addition of b-GP is related to several factorssuchashydrolysisofb-GPbyGPI-bALPtoincrease the concentration of inorganic phosphate [31,40,58,66,68], blocking the activity of a phosphorylated phosphatase [58] or affecting other biological events [68]. Addition of exogenous phosphate was often supplemented [26,28– 31,40,58,59,69–71] to facilitate the mineralization. Hexose phosphate esters [7], phosphoethanolamine [72] and ATP [41] could induce mineral formation; however, they were usually not routinely supplemented in the mineralization medium. The hydrolysis of glucose 1-phosphate, glucose 6-phosphate, glucose 1,6-bisphosphate and phosphocrea- tine, by GPI-bALP in the mineralization medium produced HA, in a similar manner as during the hydrolysis of b-GP by GPI-bALP. These results suggest that supplementation of phosphate compounds other than b-GP, may be a valuable tool to better monitor the activity of different enzymes participating in the mineralization process. Whether these organic phosphates are substrates for GPI-bALP in vivo remains to be evaluated. Organic phosphate compounds that inhibit mineralization Nucleotides can be released in extracellular medium from chondrocytes and can be rapidly degraded [72]. Under our in vitro conditions, formation of HA was observed during hydrolysis of AMP by GPI-bALP, at a slower rate as compared with the other organic phosphate substrates. The relative rate of hydrolysis of AMP by the enzyme is identical to that of glucose 1-phosphate (value from mammalian bALP [73]), suggesting that the rate of hydrolysis is not the limiting factor for mineral formation. There was no formation of any HA after 144 h incubation time in the presence of either ADP or ATP in hydrolysing medium containing GPI-bALP and excess of calcium (to compen- sate the complexation of calcium by the phosphate within nucleotides). The relative rate of hydrolysis of ADP was Fig. 6. X-ray spectra of mineral samples obtained after 144 h incubation in the mineralization medium. Themediumcontained(A)toptrace: 12.5 m M b-GP; bottom trace: 12.5 m M creatine phosphate; (B) 12.5 m M phosphate sugars: top trace: glucose 6-phosphate (G6P); middle trace: glucose 1-phosphate (G1P); bottom trace: glucose 1,6- bisphosphate (G1–6P); (C) 12.5 m M nucleotides: top trace: AMP, middle trace: ADP, bottom trace: ATP. For experimental conditions, see Materials and methods. Ó FEBS 2003 Bone GPI-bALP, roles in mineralization (Eur. J. Biochem. 270) 2087 comparable to that of b-GP (1.13 vs. 1.31 for the mammalian enzyme) while that of ATP was about 0.37 [73]. Although ADP and ATP were completely hydrolyzed by GPI-bALP and were not inhibitors of GPI-bALP, different types of mineral deposits such as OCP and HPO 4 2– complexes of calcium were obtained. The lack of HA formation is probably associated with the presence of the pyrophosphate group within the two nucleotides, as pyro- phosphate [36] as well as organic biphosphonates [74–76] are known inhibitors of HA formation. Our findings are consistent with the observation [77] that a specific ATPase, rather than alkaline phosphatase is responsible for the ATP dependent mineral deposits within matrix vesicles [77]. Possibly, the undefined Ca- and Pi- deposits [41,77] may contain complexes other than HA, which could result from ATP hydrolysis by GPI-bALP. Conclusions Our results indicate that the formation of HA does not depend uniquely on the ability of GPI-bALP to hydrolyze the organic phosphate substrate, suggesting that a nucle- ating mechanism or interactions with the products of hydrolysis are important factors that may affect the mineralization process. Acknowledgements The authors thank Professor Slawomir Pikula for helpful scientific discussions and for revision of the text. The authors gratefully acknowledge the careful X-ray diffraction measurements by Dr Ruben Vera. We also thank Dr John Carew for reviewing the English manuscript. This work is part of PhD Thesis of Mrs Eva Hamade who is supported by CNRS UMR 5013. References 1. Garnero, P. & Delmas, P.D. (1996) New developments in bio- chemical markers for osteoporosis. Calcif. Tissue Int. 59 (Suppl. 1), S2–S9. 2. Risteli, L. & Risteli, J. (1993) Biochemical markers of bone metabolism. Ann. Med. 25, 385–393. 3. Nawawi, H., Samson, D., Apperley, J. & Girgis, S. (1996) Biochemical markers in patients with multiple myeloma. Clinica Chimica Acta 253, 61–77. 4. Magnusson, P., Larsson, L., Magnusson, M., Davie, M.W.J. & Sharp, C.A. (1999) Isoforms of bone alkaline phosphatase: Characterization and origin in human trabecular and cortical bone. J. Bone Mineral Res. 14, 1926–1933. 5. Whyte, M.P. (1994) Hypophosphatasia and the role of alkaline phosphatase in skeletal mineralization. Endocr. Rev. 15, 439–461. 6. Narisawa, S., Fro ¨ hlander, N. & Millan, J.L. (1997) Inactivation of two mouse alkaline phosphatase genes and establishment of a model of infantile hypophosphatasia. Dev. Dynamics 208, 432–446. 7. Robison, R. (1923) The possible significance of hexosephosphoric esters in ossification. Biochem. J. 17, 286–293. 8. Noda, M., Yoon, K., Rodan, G.A. & Koppel, D.E. (1987) High lateral mobility of endogenous and transfected alkaline phospha- tase: a phosphatidylinositol-anchored membrane protein. J. Cell Biol. 105, 1671–1677. 9. Pizauro, J.M., Ciancaglini, P. & Leone, F.A. (1994) Osseous plate alkaline phosphatase is anchored by GPI. Brazilian. J. Med. Res. 27, 453–456. 10. Anderson, H.C. (1995) Molecular biology of matrix vesicles. Clin. Orthop. Res. 314, 266–280. 11. Ducy, P., Schinke, T. & Karsenty, G. (2000) The osteoblast: a sophisticated fibroplast under central surveillance. Science 289, 1501–1504. 12. Fowler, B.O. (1974) Vibrational assignments for calcium, stron- tium, and barium hydroxyapatites utilizing isotopic substitution. Inorganic Chem. 13, 194–207. 13. Eanes, E.D., Termine, J.D. & Posner, A.S. (1967) Amor- phous calcium phosphate in skeletal tissue. Clin. Orthop. 53, 223–235. 14. Termine, J.D. & Posner, A.S. (1966) Infrared analysis of rat bone age depending on amorphous and crystalline mineral fractions. Science 153, 1523–1525. 15. Termine, J.D. & Posner, A.S. (1966) Amorphous/crystalline interrelationship in bone mineral? Calcif. Tissue Res. 6, 335–342. 16. Francis, M.D. & Webb, N.C. (1971) Hydroxyapatite formation from hydrated calcium monohydrogen phosphate precursor. Calcif. Tissue Res. 6, 335–342. 17. Brown, W.E. (1966) Crystal growth of bone mineral. Clin. Orthop. 44, 205–220. 18. Munzenberg, K.J. & Gebhardt, M. (1972) Crystalline calcium phosphate of bone. Biomineralization 6, 91–95. 19. Roufosse, A.H., Landis, W.J., Sabine, N.K. & Glimcher, M.J. (1979) Identification of brushite in newly deposited bone mineral from embryonic chicks. J. Ultrastruct. Res. 68, 235–255. 20. Boner, L.C., Grypnas, M.C. & Glimcher, M.J. (1984) Failure to detect crystalline brushite in embryonic chick and bovine bone by X-ray diffraction. J. Ultrastruct. Res. 86, 93–99. 21. Legros, R., Balmain, N. & Bonel, G. (1986) Structure and com- position of mineral phase of periosteal bone. J. Chem. Res. Synop. 1, 8–9. 22. Pellegrino, E.D. & Blitz, R.M. (1972) Mineralization in chick embryo. I. Monohydrogen phosphate and carbonate relationship during maturation of the bone crystal complex. Calcif. Tissue Res. 10, 128–135. 23. Grypnas, M.D., Bauer, L.C. & Glimcher, M.J. (1984) Failure to detect amorphous calcium phosphate solid phase in bone mineral: a radial distribution function study. Calcif. Tissue Int. 36, 291–309. 24. Ashton, B.A. & Bromley, L. (1985) The influence of b-glycero- phosphate on osteoblastic-like cell lines. Bone 7, 305–306. 25. Kuhn,L.T.,Wu,Y.,Rey,C.,Gerstenfeld,L.C.,Grypnas,M.D., Ackerman, J.L., Kim, H.M. & Glimcher, M.J. (2000) Structure, Composition, and maturation of newly deposited calcium-phos- phate crystals in chicken osteoblast cell cultures. J. Bone Mineral Res. 15, 1301–1309. 26. Guicheux, J., Palmer, G., Shukunami, C., Hiraki, Y., Bonjour, J.P. & Caverzasio, J. (2000) A novel in vitro culture system for analysis of functional role of phosphate transport in endochondral ossification. Bone 27, 69–74. 27. Wang, D., Canaff, L., Davidson, D., Corluka, A., Liu, H., Hendy, G.N. & Henderson, J.E. (2001) Alterations in the sensing and transport of phosphate and calcium by differentiating chon- drocytes. J. Biol. Chem. 276, 33995–34005. 28. Kirsch, T., Nah, H D., Shapiro, I.M. & Pacifi, M. (1997) Regu- lated production of mineralization-competent matrix vesicles in hypertrophic chondrocytes. J. Cell. Biol. 137, 1149–1160. 29. Kirsch, T., Harrison, G., Worch, K.P. & Golub, E.E. (2000) Regulatory roles of zinc in matrix vesicle-mediated minerali- zation of growth plate cartilage. J. Bone Mineral Res. 15, 261–270. 30. Kirsch, T., Harrison, G., Golub, E.E. & Nah, H D. (2000) The roles of annexins and types II and X collagen in matrix vesicle- mediated mineralization of growth plate cartilage. J. Biol. Chem. 275, 35577–35583. 2088 E. Hamade et al. (Eur. J. Biochem. 270) Ó FEBS 2003 31. Bellows, C.G., Aubin, J.E. & Heersche, J.N.R. (1991) Initiation and progression in mineralization of bone nodules formed in vitro: the role of alkaline phosphatase and organic phosphate. Bone Mineral 14, 27–40. 32. Coc, M.R., Summers, T.A., Parson, S.J., Boskey, A.L. & Bolian, J. (1992) Matrix mineralization in hypertrophic chondrocyte cul- tures. Beta glycerophosphate increases type X collagen messenger RNA and the specific activity of pp60 kinase. Bone Mineral 18, 91–106. 33. Berry, L. & Schutterworth, C.A. (1989) Expression of the chon- drogenic phenotype by mineralizing cultures of embryonic chick calvarial bone cell. Bone Mineral 7, 31–45. 34. Gerstenfeld, L.C. & Landis, W.J. (1991) Gene expression and extracellular matrix ultrastructure of a mineralizing chondrocyte cell culture system. J. Cell. Biol. 112, 501–513. 35. Sakamoto, S., Sakamoto, M., Goldberg, L. & Colarusso, L. & Gotah, Y. (1989) Mineralization induced by beta-glycerophos- phate in cultures leads to a marked increase in collagenase synthesis by mouse osteogenic MC3T3–E1 cells under subsequent stimulation with heparin. Biochem. Biophys. Res. Commun. 162, 773–780. 36. Tenebaum, H.C. (1987) Levamisole and inorganic pyrophosphate inhibit b-glycerophosphate induced mineralization of bone formed in vitro. Bone Mineral 3, 13–26. 37. Tenenbaum, H.C., McCulloch, C.A.G., Fair, C. & Birek, C. (1989) Stimulatory effects of phosphates on bone mineralization in vitro. Cell. Tissue Res. 257, 555–563. 38. Tennenbaum, H.C., Lineback, H., McCulloch, C.A.G., Mamuju, H., Sukku, B. & Tarantali, M. (1987) Osteogenic-phase speci- fic co-regulation of collagen synthesis and mineralization by b-glycerophosphate in chick periosteal culture. Bone 12, 129–138. 39. Iwamoto, M., Shapiro, I., Yagami, K., Boskey, A.L., Leboy, P., Adams, S. & Pacifici, M. (1993) Retinoic acid induces rapid mineralization and expression of mineralization-related genes in chondrocytes. Exp. Cell. Res. 207, 413–420. 40. Chang, Y.L., Stanford, C.M. & Keller, J.C. (2000) Calcium and phosphate supplementation promotes bone cell mineralization: implications for hydroxyapatite (HA) enhanced bone formation. J. Biomed. Mater. Res. 52, 270–278. 41. Hsu, H.H. (1992) Further studies on ATP-mediated CA deposi- tion by isolated matrix vesicles. Bone Mineral 17, 279–283. 42. Radisson, J., Angrand, M., Chavassieux, P., Roux, B. & Azzar, G. (1996) Differential solubilization of osteoblastic alkaline phos- phatase from human primary bone cell cultures. Int. J. Biochem. Cell. Biol. 28, 421–430. 43. Lowry, O.H., Rosebrough, N.I., Farr, A.L. & Randall, R.J. (1951) Protein measurement with the folin phenol reactant. J. Biol. Chem. 193, 265–273. 44. Frot Coutaz, J. & Got, R. (1971) Proprie ´ te ´ s de troisglucosyltrans- fe ´ rases de la muqueuse intestinale de rat. Biochimie 53, 597–601. 45. Bradford, M.M. (1978) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254. 46. Read, S.M. & Northcote, D.H. (1981) Minimization of variation in the response to different proteins of the Coomassie Blue G dye-binding assay for proteins. Anal. Biochem. 116, 53–64. 47. Cyboron, G.W. & Wuthier, R.E. (1981) Purification and initial characterization of intrinsic membrane bound alkaline phospha- tase from chicken epiphyseal cartilage. J. Biol. Chem. 256, 7262– 1268. 48. Harrison, G., Shapiro, I.R. & Golub, E.E. (1995) The phos- phatidylinositol-glycolipid anchor on alkaline phosphatase facil- itates mineralization initiation in vitro. J. Bone Mineral Res. 10, 568–573. 49. Chen, P.S. Jr, Toribara, T.Y. & Warner, H. (1956) Micro- determination of phosphorus. Anal. Chem. 28, 1756–1758. 50. Pleshko, N., Boskey, A. & Mendelsohn, R. (1991) Novel infrared spectroscopic method for the determination of cristallinity of hydroxyapatite minerals. Biophys. J. 60, 786–793. 51. Sauer, G.R. & Wuthier, R.E. (1988) Fourier transform infrared characterization of mineral phases formed during induction of mineralization by collagenase-released matrix vesicles in vitro. J. Biol. Chem. 263, 13718–13724. 52. Pleshko, N.L., Boskey, A.L. & Mendelsohn, R. (1992) An FT-IR microscopic investigation of the effects of tissue preservation on bone. Calcif. Tissue Int. 51, 72–77. 53.Gadaleta,S.J.,Paschalis,E.P.,Betts,F.,Mendelsohn,R.& Boskey, A.L. (1996) Fourier transform infrared spectroscopy of the solution-mediated conversion of amorphous calcium phos- phate to hydroxyapatite. New correlations between X-ray dif- fraction and infrared data. Calcif. Tissue Int. 58, 9–16. 54. Paschalis, E.P., DiCarlo, E., Betts, F., Shermann, P., Mendelsohn, R. & Boskey, A.L. (1996) FTIR microscopic analysis of human osteonal bone. Calcif. Tissue Int. 59, 480–487. 55. Rey, C., Shimizu, M., Collins, B. & Glimcher, M.J. (1991) Resolution enhanced Fourier transform infrared spectroscopy study of the environment of phosphate ions in the early deposits of a solid phase calcium phosphate in bone and enamel, and their evolution with age: 2. Investigations in the m3PO 4 domain. Calcif. Tissue Int. 49, 383–388. 56. Rey, C., Shimizu, M., Collins, B. & Glimcher, M.J. (1990) Resolution enhanced Fourier transform infrared spectroscopy study of the environment of phosphate ions in the early deposits of a solid phase of calcium-phosphate in bone and in enamel, and their evolution with age. I: Investigations in the m4PO 4 domain. Calcif. Tissue Int. 46, 384–394. 57. Ou-Tang, H., Pachalis, E.P., Boskey, A.L. & Mendelsohn, R. (2000) Two-dimensional vibrational correlation spectroscopy of in vitro hydroxyapatite maturation. Biopolymers (Biospectro- scopy) 57, 129–139. 58. Boskey, A.L., Guidon, P., Doty, S.B., Leboy, P. & Binderman, I. (1996) The mechanism of b-glycerophosphate action in mineral- izing chick limb-bud mesenchymal cell cultures. J. Bone Mineral Res. 11, 1694–1702. 59. Feki, H.E., Rey, C. & Vignoles, M. (1991) Carbonate ions in apatites: infrared investigations in the m4CO 3 domain. Calcif. Tissue Int. 49, 269–274. 60. Dahm, S. & Risnes, S. (1999) A comparative infrared spectro- scopic study of hydroxide and carbonate absorption bands in spectra of shark enameloid, shark dentin and a geological apatite. Calcif. Tissue Int. 65, 459–465. 61. Penel, G., Leroy, G., Rey, C. & Bres, E. (1998) MicroRaman spectral study of the PO 4 and CO 3 vibrational modes in synthetic and biological apatites. Calcif. Tissue Int. 63, 475–481. 62. Jikko, A., Aoba, T., Murakami, H., Takano, Y., Iwamoto, M. & Kato, Y. (1993) Characterization of the mineralization process in cultures of rabbit growth plate chondrocytes. Develpmental Biol. 156, 372–380. 63. Gadaleta, S.J., Geicke, A., Boskey, A.L. & Mendelsohn. R. (1996) Two-dimensional infrared correlation spectroscopy of synthetic and biological apatites. Biospectroscopy 2, 353–364. 64. Boskey, A.L., Camacho, N.P., Mendelsohn, R. & Binderman, I. (1992) FT-IR microscopic mappings of early mineralization in chick limb bud mesenchymal cell cultures. Calcif. Tissue Int. 51, 443–448. 65. Golub. E.E., Harisson, A.G., Taylor, S., Camper, S. & Shapiro, I.M. (1992) The role of phosphatase alkaline in cartilage miner- alization. Bone Mineral 17, 273–278. 66. Chung, C.H., Golub, E.E., Forbes, E., Tokuoka, T. & Shapiro, I.M. (1992) Mechanism of action of b-glycerophosphate on bone cell mineralization. Calcif. Tissue Int. 51, 305–311. Ó FEBS 2003 Bone GPI-bALP, roles in mineralization (Eur. J. Biochem. 270) 2089 67. Mehta, S., Reed, B. & Antich, P. (1995) Effects of high levels of fluoride on bone formation: an in vitro model system. Biomaterials 16, 97–102. 68. Anagnostou, F., Plas, C., Nefussi, J.R. & Forest, N. (1996) Role of b-GP-derived Pi in mineralization via ecto-alkaline phosphatase in cultured fetal calvaria cells. J. Cell. Biochem. 62, 262–274. 69. Rattner, A., Sabido, O., Le, J., Vico, L., Massoubre, C., Frey, J. & Chamson, A. (2000) Mineralization and alkaline phosphatase in collagen lattices populated by human osteoblasts. Calcif. Tissue Int. 66, 35–42. 70. Tenenbaum, H.C., McCulloch, C.A.G., Fair, C. & Birck, C. (1989) The regulatory effect of phosphates on bone metabolism in vitro. Cell. Tissue Res. 257, 555–563. 71. Skrtic, D. & Eanes, E.D. (1992) Effect of different phospholipid- cholesterol membrane compositions on liposome-mediated for- mation of calcium phosphates. Calcif. Tissue Int. 50, 253–260. 72. Hatori, M., Teixeira, C.C., Debolt, K., Pacifici, M. & Shapiro, I.M. (1995) Adenine nucleotide metabolism by chondrocytes in vitro: role of ATP in chondrocyte maturation and matrix mineralization. J. Cell. Phys. 165, 468–474. 73. Fernley, H.N. (1971) Mammalian alkaline phosphatase. In: The Enzymes, Vol. IV (ed. Boyer, P.D.). Academic Press, New York, pp. 417–447. 74. Skrtic, D. & Eanes, E.D. (1994) Effect of 1-hydroxyethylidienne- 1,1-biphosphonate on membrane-mediated calcium phosphate formation in model liposomal suspensions. Bone Mineral 26, 219–229. 75. Francis, M.D. (1969) The inhibition of calcium hydroxyapatite crystal growth by polyphosphonates and polyphosphates. Calcif. Tissue Res. 3, 151–162. 76. Fleisch, H., Russel, R.G.G. & Francis, M.D. (1969) Diphos- phonates inhibit hydroxyapatite dissolution in vitro and bone resorption in tissue culture in vivo. Science 165, 1262–1264. 77. Hsu, H.H.T. & Andersen, H.C. (1996) Evidence of the presence of a specific ATPase responsible for ATP-initiated calcification by matrix vesicles isolated from cartilage and bone. J. Biol. Chem. 271, 26383–26388. 2090 E. Hamade et al. (Eur. J. Biochem. 270) Ó FEBS 2003 . Chick embryo anchored alkaline phosphatase and mineralization process in vitro In uence of Ca 2+ and nature of substrates Eva Hamade,. France. Results Evidence of different steps in the GPI-bALP mediated mineralization: effects of Ca 2+ concentration To assess the role of Ca 2+ in initiating the mineralization process,