Inorganic Polyphosphate in Human Osteoblast-like Cells



Significant amounts of inorganic polyphosphates and of polyphosphate-degrading exopolyphosphatase activity were detected in human mandibular-derived osteoblast-like cells. The amount of both soluble and insoluble long-chain polyphosphate in unstimulated osteoblast-like cells was higher than in human gingival cells, erythrocytes, peripheral blood mononuclear cells, and human blood plasma. The cellular content of polyphosphate in osteoblast-like cells strongly decreased after a combined treatment of the cells with the stimulators of osteoblast proliferation and differentation, dexamethasone, β-glycerophosphate, epidermal growth factor, and ascorbic acid. The amount of soluble long-chain polyphosphate, but not the amount of insoluble long-chain polyphosphate, further decreased after an additional treatment with 1α,25-dihydroxyvitamin D3(1,25(OH)2D3). The decrease in polyphosphate content during treatment with dexamethasone, β-glycerophosphate, epidermal growth factor, and ascorbic acid was accompanied by a decrease in exopolyphosphatase, pyrophosphatase, and alkaline phosphatase activity. However, additional treatment with 1,25(OH)2D3 resulted in an increase in these enzyme activities. Osteoblast-like cell exopolyphosphatase activity and exopolyphosphatase activity in yeast, rat tissues, and human leukemia cell line HL60 were inhibited by the bisphosphonates etidronate and, to a lesser extent, clodronate and pamidronate. From our results, we assume that inorganic polyphosphate may be involved in modulation of the mineralization process in bone tissue.


INORGANIC LINEAR POLYPHOSPHATES (polyPs) are long polymers of orthophosphate (Pi) ranging in size of up to one thousand Pi residues.1 The occurrence and metabolism of polyPs have been studied mostly in microorganisms (reviewed in Refs. 2 and 3), but lower levels of these polymers are present also in animal cells and tissues.4–9 More recently, significant amounts of long-chain polyPs were also detected in human cells.6,9–11 Furthermore, polyP was found extracellularly in human blood plasma.11

The biological function(s) of inorganic polyP is not well understood. It may serve as storage molecule for energy-rich phosphate,2 as chelator for Ca2+ or other divalent cations,12 as counterion for basic amino acids in vacuoles,13 as regulator of the intracellular levels of adenylate nucleotides via inhibition of adenylate kinase,7 as donor of phosphate for certain sugar kinases,14 or it may play a role in apoptosis8 or the response of cells to osmotic stress15 and pH stress.16 Moreover, polyP is able to form complexes with poly-β-hydroxybutyrate and Ca2+ in cell membranes.17 In addition, it possesses antibacterial18 and antiviral activity.11

In bacteria, polyP is synthesized from ATP by the polyphosphate kinase19; in yeast and higher eukaryotes, synthesis of polyP may occur from Pi without prior formation of ATP by a mechanism not yet determined.10 The degradation of polyP is catalyzed by several endo- and exopolyphosphatases.7,20–24 The activities of these enzymes and the intracellular concentration and size of polyP have been shown to change in dependence on the availability of phosphate, the period of growth cycle,2 and during development25 and aging.8

Inorganic pyrophosphate (PPi) and polyP have been shown to inhibit both the formation26,27 and dissolution of calcium phosphate crystals28 and might therefore act as physiological regulators of calcification and decalcification.28 In the present report, the occurrence of polyP in human bone-forming osteoblasts was studied. In addition, the effect of bisphosphonates on polyphosphatase activity in osteoblast-like cells was examined. These compounds are synthetic analogs of PPi containing a methylene bridge between the phosphate groups instead of the phosphoanhydride bond, and are therefore resistant to chemical and enzymatic hydrolysis.29 The bisphosphonates, which differ in the structure of the side chains attached to the methylene bridge, are effective inhibitors of bone resorption. They are used clinically in the treatment of disorders characterized by excessive bone loss,29,30 including Paget's disease,31 tumor-induced hypercalcemia,32 metastatic bone disease,33 and osteoporosis.34,35 They bind strongly to hydroxyapatite crystals and inhibit, like PPi and polyP, both precipitation36 and dissolution of calcium phosphate.37,38 Our results show that human osteoblast-like cells contain a polyP-splitting exopolyphosphatase activity that responds to modulators of osteoblast growth and function, and this enzyme is inhibited by bisphosphonates, indicating that polyP metabolism in bone tissue may be a further target for these drugs.



Monosodium [32P]phosphate (specific activity, 600 Ci/mmol) was obtained from ICN Biomedicals (Meckenheim, Germany); tetrasodium pyrophosphate (PPi), pentasodium tripolyphosphate (polyP3), hexaammonium tetrapolyphosphate (polyP4), sodium polyP glass type 35 (with average chain length 34), 1-hydroxyethylidene-1,1-bisphosphonic acid (etidronic acid), inorganic pyrophosphatase (yeast, I 1891), epidermal growth factor (EGF; E 1257), 1α,25-dihydroxyvitamin D3 (1,25(OH)2D3), p-nitrophenyl phosphate, and 4-methylumbelliferyl phosphate were purchased from Sigma Chemical Co. (St. Louis, MO, U.S.A.); inorganic pyrophosphatase (yeast), deoxyribonuclease I (DNAse I; bovine pancreas), and ribonuclease A (RNAse A; bovine pancreas) were from Boehringer Mannheim (Mannheim, Germany); alkaline phosphatase (ALP; calf intestine) was from MBI Fermentas (Vilnius, Lithuania); and methylene-1,1-bisphosphonate (trisodium salt) was from Aldrich (Steinheim, Germany).

Dichloromethylene-1,1-bisphosphonate (clodronate, disodium salt) was kindly provided by Boehringer Mannheim, and 3-amino-1-hydroxypropylidene-1,1-bisphosphonate (pamidronate, disodium salt) was a gift from Novartis (CIBA) (Wehr, Germany).

Exopolyphosphatase was purified from yeast as described previously.23 Fraction 4 (phenyl-Sepharose) was used; the specific activity was 5.4 × 104 U/mg of protein. [32P]PolyP was synthesized as described.23 The specific radioactivity was 0.7–0.9 × 107 dpm/μmol of Pi. Commercial polyP35, which is a heterogeneous mixture of polyP molecules with varying chain lengths,39 was separated from Pi, PPi, and short-chain polyP by size exclusion chromatography on a Sephadex G-25 column (Pharmacia, Uppsala, Sweden). Aliquots were taken from the eluate fractions and assayed for polyP (metachromatic reaction) or Pi (ammonium molybdate reaction), as described below; polyP chain lengths were analyzed by electrophoresis on urea/polyacrylamide gel (see below). PolyP90 (size 70–110 Pi residues) was isolated by preparative gel electrophoresis of polyP35 as described.7

Cell culture

Primary cultures of human osteoblast-like cells were prepared from mandible biopsy material by treatment of the bone fragments with collagenase (4 mg/ml) in alpha modified essential medium (α-MEM) growth medium (α-MEM [Biochrom, Berlin, Germany] supplemented with 2 g/l of NaHCO3, 2 mM L-glutamine, 100,000 IU/l of penicillin, and 100 mg/l of streptomycin) at 37°C for 1 h. The collagenase-containing medium was then removed and centrifuged at 300g (at room temperature for 10 minutes). The resulting cell pellet was resuspended in fresh α-MEM growth medium supplemented with 2.5 mg/l of amphotericin B and 15% fetal calf serum (FCS), and maintained in 25-cm2 culture flasks at 37°C in a humidified atmosphere at 5% CO2 in air. The medium was changed after 1 day and then every 2 days. The cells were characterized histochemically by determination of ALP activity. Cells were frozen in fluid nitrogen in the presence of 8% dimethylsulfoxide. After thawing, medium was changed at days 1 and 4. Cells were trypsinated at day 6 and seeded in α-MEM growth medium in 25-cm2 cell culture flasks (5 × 104 cells/flask). Osteoblast-like cells were grown either 14 days in α-MEM growth medium or 8 days in α-MEM growth medium, then 6 days in α-MEM growth medium supplemented with 10−7 M dexamethasone, 10−7 M β-glycerophosphate, 10 ng/ml of EGF, and 50 μg/ml of ascorbic acid (α-MEMplus), or 8 days in α-MEM growth medium, then 3 days in α-MEMplus, then 3 days in α-MEMplus supplemented with 10−8 M 1,25(OH)2D3. For determination of polyP content and exopolyphosphatase activity, cell monolayers (106–107 cells per flask) were washed in Ca2+ and Mg2+-free phosphate-buffered saline (PBS; pH 7.3) and detached without trypsin using 0.02% EDTA in PBS.

Human primary fibroblasts were cultured from biopsies of the attached gingiva and maintained in RPMI 1640 medium (Biochrom) supplemented with 2 g/l of NaHCO3, 2 mM L-glutamine, 100,000 IU/l of penicillin, 100 mg/l of streptomycin, 2.5 mg/l of amphotericin B, and 10% FCS at 37°C in a humidified 5% CO2 in air atmosphere. Cells were incubated for 13 days, then washed in PBS (Ca2+- and Mg2+-free; pH 7.3), and detached without trypsin as above.

Human promyelocytic leukemia HL-60 cells were grown in RPMI 1640 medium (GIBCO BRL, Grand Island, NY, U.S.A.) supplemented with 10% FCS at 37°C in 5% CO2/air.

Peripheral blood mononuclear cells (PBMCs) were isolated from human blood by density gradient centrifugation on a polysucrose-sodium metrizoate medium. Citrate plasma was prepared from citrate-treated blood by centrifugation. Human erythrocytes were obtained from the blood bank of the University of Mainz.

Saccharomyces cerevisiae (stem A364 A) were cultivated in medium consisting of 1% yeast extract (GIBCO), 2% peptone (casein hydrolysate; GIBCO), and 2% glucose at 30°C for 24–48 h. Cells were harvested at the end of the logarithmic phase.

Preparation of cell extracts

Osteoblast-like cells, gingival fibroblasts, and HL60 cells were washed in PBS, centrifuged, and the resulting cell pellets were resuspended in extraction buffer (50 mM Tris-HCl, pH 7.5, 5 mM MgCl2, 150 mM NaCl, 1 mM dithiothreitol, 1% Triton X-100) and sonicated on ice (5 × 10 s). The liver from male Wistar rats (6 months old) was rapidly taken from the animals and homogenized in 50 mM Tris-HCl buffer (pH 7.5) containing 10 mM MgCl2, 0.5 mM EDTA, 150 mM NaCl, 100 μM leupeptin, and 100 nM pepstatin. Yeast cells were washed with water and homogenized by grinding with glass beads under liquid nitrogen and suspension in 50 mM Tris-HCl buffer (pH 7.5, 1 mM MgCl2, 0.5 mM EDTA, 100 μM leupeptin, 100 nM pepstatin). The supernatants obtained after centrifugation (14,000 rpm for 10 minutes in an Eppendorf microfuge) were stored at −80°C.

Exopolyphosphatase activity

The reaction mixture for determination of exopolyphosphatase activity consisted of 50 mM Tris-HCl buffer (pH 7.5, containing 5 mM MgCl2, 150 mM NaCl, 1 mM dithiothreitol, and 0.1% Triton X-100) in a final volume of 100 μl. If not stated otherwise, polyP35 (Sigma type 35, further purified as described above; 0.485 μmol of Pi residues/assay) was used as substrate. Incubations were performed at 37°C for various time periods. Reactions were stopped by chilling to 0°C, followed by addition of 120 mM sodium acetate, pH 5.0 (final), and 3 vol of ethanol, and incubation at –80°C for 30 minutes. The supernatant obtained after centrifugation in an Eppendorf centrifuge (14,000 rpm for 15 minutes at 4°C) was dried at 95°C then dissolved in 300 μl of distilled water. The Pi formed during the reaction was determined spectrophotometrically using the ammonium molybdate method.7,25

In some assays, unlabeled polyP was replaced by [32P]polyP (1.6 nmol of Pi residues/assay). The degradation products were analyzed either by gel electrophoresis (see below) or by chromatography on cellulose thin-layer plates, which were developed with isobutyric acid/NH4OH/water (25:3:12) containing 0.8 mM EDTA.16

Pyrophosphatase activity

Inorganic pyrophosphatase activity was determined essentially as described.40 Briefly, the assay mixture (final volume 100 μl) contained 50 mM Tris-HCl (pH 7.2), 10 mM MgCl2, 150 mM NaCl, and 0.2 mM sodium pyrophosphate. After incubation for 0–30 minutes at 25°C, 1 ml of 0.35 M H2SO4 was added for determination of Pi by the method of Fiske and Subbarow,41 except that ascorbic acid was used as reducing agent.

ALP activity

The activity of ALP was measured using the fluorogenic substrate, 4-methylumbelliferyl phosphate.42 Cells grown in 96-well plates were washed twice with PBS, then 200 μl of reaction mixture (100 mM Tris buffer, pH 8.9, 2 mM MgSO4, 1 mg/ml of 4-methylumbelliferyl phosphate) was added to each well. After incubation at 37°C for 3 h, the fluorescence was measured using a Cytofluor 2350 (Millipore, Bradford, MA, U.S.A.) (excitation wavelength, 360 nm; emission wavelength, 460 nm). The assay for ALP activity using p-nitrophenyl phosphate as substrate was performed as described.43

Extraction and determination of polyP content

PolyP was extracted from cells or tissues as described.15,39 Residual amounts of DNA and RNA in step 2 and step 3 extracts39 were removed from the extracts by treatment with RNAse A and DNAse I (250 μg/ml each) in the presence of 1 mM MgCl2 for 1 h at room temperature; reactions were stopped by the addition of 2 mM EDTA (final). Protein was then removed by one extraction with phenol/chloroform (1:1, v/v), followed by three successive extractions with chloroform.

The polyP content in step 2 extracts containing the easily extractable, “soluble” portion of long-chain polyP and step 3 extracts containing the remaining “insoluble” portion of long-chain polyP39 was determined by measuring the metachromatic effect of polyP on the absorbance spectrum of toluidine blue at 530 nm and 630 nm, as described.15 Commercial polyP with an average chain length of 34 Pi residues was used as standard. Absorbance spectra were recorded with a Beckman DU-64 spectrophotometer.

Determination of chain length of polyP

The size of the polyP was determined by electrophoresis on 7 M urea/16.5% polyacrylamide gels as described.7 [32P]PolyP was detected by autoradiography and unlabeled polyP by staining with o-toluidine blue O. PolyP standards of defined chain lengths were run in parallel.

Protein concentration

Protein concentrations were determined by the method of Lowry,44 with bovine serum albumin as a standard.


Changes in polyP content in osteoblast-like cells

The polyP content of human mandibular-derived osteoblast-like cells was determined by measuring the amount of the polymer in the soluble and insoluble long-chain polyP fraction and was compared with the polyP contents of human gingival cells and blood constituents. As summarized in Table 1, the amount of both soluble long-chain polyP (mainly consisting of polymers of 10–50 Pi residues) and insoluble long-chain polyP (mainly consisting of polymers of >50 Pi residues) was markedly higher in osteoblast-like cells than in gingival fibroblasts, PBMCs, erythrocytes, and blood plasma. The concentration of soluble long-chain polyP exceeded in general that of insoluble long-chain polyP; in osteoblast-like cells, it was about 3-fold higher than that of insoluble long-chain polyP.

Table Table 1. Concentration of Long-Chain PolyP in Different Human Cells and Blood Plasma
original image

The total content of long-chain polyP in osteoblast-like cells (528 μM [Pi]; calculated from the values given in Fig. 1) strongly decreased (by 61%) after a combined treatment of the cells with dexamethasone, β-glycerophosphate, EGF, and ascorbic acid. These compounds have been shown to stimulate proliferation and differentation of human bone-derived cells.45,46 The decrease in total polyP content was mainly caused by a decrease in the amount of soluble, long-chain polyP (by 68%); the decrease in insoluble, long-chain polyP amounted to 41% (Fig. 1). An additional treatment of the cells with 1,25(OH)2D3 resulted in a further decrease in the amount of soluble long-chain polyP (by 39%) but not in the amount of insoluble long-chain polyP which remained essentially constant (Fig. 1).

Figure FIG. 1.

Changes in polyP content of human osteoblast-like cells after treatment with different modulators. The amount of long-chain polyP was determined in step 2 extract (soluble polyP; open bars) and step 3 extract (insoluble polyP; filled bars) from unstimulated osteoblast-like cells (Obl) and from osteoblast-like cells subjected to combined treatment with dexamethasone (Dex), β-glycerophosphate (GP), EGF, and ascorbic acid (Asc) for 6 days, or to combined treatment with the same additives for 3 days in the absence and then for 3 days in the presence of 1,25(OH)2D3 (VitD3). For comparison, the polyP content of gingival fibroblasts (Gfib) is shown. PolyP35 was used as substrate. Results are means (± SD) of three independent experiments.

Changes in exopolyphosphatase activity in osteoblast-like cells

Extracts from osteoblast-like cells were found to contain a polyP-degrading activity when incubated with polyP35 (free of Pi, PPi, and polyP < 10 Pi residues) or polyP90 (size range: 70–110 Pi residues; Fig. 2A) as substrate. Degradation of polyP resulted in the production of Pi, as revealed by analysis of the degradation products of [32P]polyP by thin-layer chromatography (data not shown). This result indicates that the enzyme is an exopolyphosphatase.

Figure FIG. 2.

Analysis of products obtained after incubation of long-chain polyP with osteoblast-like cell extract, inorganic pyrophosphatase, or ALP. PolyP90 (20 μg; chain length, 70–110 Pi residues) was incubated with osteoblast-like cell extract (53 μg of protein/assay) (A), purified inorganic pyrophosphatase (yeast, 20 ng/assay; Boehringer Mannheim) (B), or purified ALP (calf intestine, 23 ng/assay; MBI Fermentas) (C) in 50 mM Tris-HCl buffer (pH 7.5; 150 mM NaCl, 1 mM MgCl2) at 37°C for 0 (lane a), 1 (lane b), 2 (lane c), and 3 h (lane d). Aliquots were taken at the end of the incubation period and analyzed by gel electrophoresis on 16.5% polycrylamide/7 M urea gels (4 μg of polyP per lane). The gels were stained with toluidine blue. The migrations polyP standards (polyP100 and polyP80) are indicated. The specific exopolyphosphatase activity (substrate, polyP90) of osteoblast-like cell extract was 9.8 nmol Pi/minute/mg of protein (ammonium molybdate reaction).

As shown in Fig. 3A, the decrease in polyP content during treatment with dexamethasone, β-glycerophosphate, EGF, and ascorbic acid was accompanied by a decrease (to 27%) in exopolyphosphatase activity. However, additional treatment with 1,25(OH)2D3 resulted in a significant increase (by 79%, compared with the cultures without 1,25(OH)2D3) in the enzyme activity.

Figure FIG. 3.

Changes in exopolyphosphatase, inorganic pyrophosphatase, and ALP activities in human osteoblast-like cells after treatment with different modulators. (A) Exopolyphosphatase activity (substrate, polyP35), (B) inorganic pyrophosphatase activity, and (C) ALP activity were determined in total cell extracts from unstimulated osteoblast-like cells and gingival fibroblasts, and from osteoblast-like cells subjected to different treatments. For further details, see legend to Fig. 1. Results are means (± SD) of three independent experiments.

The changes in exopolyphosphatase activity were compared with those of inorganic pyrophosphatase and ALP, two well-characterized further enzymes involved in production of Pi. It was found that the activities of these enzymes were markedly reduced to 61% (Fig. 3B) and 62% (Fig. 3C), respectively, during the 6-day incubation period in α-MEM+, paralleling the decrease in exopolyphosphatase activity. Again, additional treatment with 1,25(OH)2D3 caused an increase in enzyme activity by 35% (pyrophosphatase; Fig. 3B) and by 48% (ALP; Fig. 3C), respectively.

To examine the possibility that production of Pi from polyP is due to inorganic pyrophosphatase or ALP activities present in osteoblast-like cell extracts, the ability of the purified enzymes to degrade long-chain polyP was determined. It has been reported that inorganic pyrophosphatase is quite specific for PPi.47 By using high-perfomance liquid chromatography (HPLC), Yoza et al.48 showed that the yeast enzyme does not degrade oligomeric phosphates (3–11 PPi residues). We found that inorganic pyrophosphatase from yeast (Sigma; HPLC purified) is also unable to hydrolyze longer polyP chains (Fig. 2B). Analysis of the Pi released by ammonium molybdate reaction revealed that the specific activity of the enzyme toward polyP90 (<15 nmol Pi/minute/mg of protein) is more than 70,000- to 80,000-fold lower than that toward PPi (1.14 ± 0.06 mmol Pi/minute/mg; mean ± SD, n = 4). ALP (bovine intestine) has been shown to catalyze hydrolysis of PPi and polyP3 at pH 7.2,49 in contrast to cleavages of P-O-C and P-F bonds of phosphate monoesters in the alkaline pH range.50 However, under the conditions used, no significant degradation of long-chain polyP by ALP (calf intestine) could be detected (Fig. 2C). It should be noted that the assay showing degradation of polyP (osteoblast-like cell extract; Fig. 2A) contained approximately the same amount of PPi-hydrolyzing activity as the assays showing no degradation of polyP (inorganic pyrophosphatase and ALP; Figs. 2B and 2C), as determined in parallel (24.9 ± 1.3, 22.8 ± 1.2, and 21.9 ± 2.8 nmol Pi/minute/assay, respectively; n = 3). In addition, the activities of purified ALP toward PPi (952 ± 39 μmol Pi/minute/mg of protein; at pH 7.5) and toward its specific phosphate monoester substrate p-nitrophenyl phosphate (3467 ± 102 μmol/minute/mg) were not inhibited by 1 mM NaF (hydrolysis of PPi: 891 ± 47 μmol Pi/minute/mg; hydrolysis of p-nitrophenyl phosphate: 3309 ± 144 μmol/minute/mg), while the exopolyphosphatase activity in osteoblast-like cell extracts measured in the standard assay with ammonium molybdate (9.8 ± 0.7 nmol Pi/minute/mg) was strongly reduced by 32% in the presence of 0.1 mM NaF (6.7 ± 0.2 nmol Pi/minute/mg) and by 81% in the presence of 1 mM NaF (1.9 ± 0.3 nmol Pi/minute/mg; values are means ± SD, n = 4; all reactions were performed at 37°C). These results indicate that hydrolysis of polyP by osteoblast-like cell extracts (Fig. 2A) cannot be due to inorganic pyrophosphatase or ALP activities.

Effects of bisphosphonates on exopolyphosphatase activity

In the following experiments, we investigated the influence of bisphosphonates, which are analogs of PPi containing a P-C-P bond instead of a P-O-P bond, on exopolyphosphatase activity. These compounds cannot be degraded by the known enzymes, pyrophosphatase,51 ALP, and exopolyphosphatase (yeast; results not shown). Since bisphosphonates may chelate Mg2+ ions, the Mg2+ concentration in the assays was adjusted to 10 mM. As shown in Fig. 4, the enzyme activity in osteoblast-like cells and exopolyphosphatase activity in yeast, rat liver, and HL60 cells were inhibited by the bisphosphonates etidronate and, to a lesser extent, clodronate and pamidronate. At the concentration used (4.85 mM), the inhibition of the yeast enzyme by etidronate was nearly complete. Methylene-1,1-bisphosphonate (3.18 mM) displayed only a very weak effect on the enzyme in yeast and osteoblast-like cells.

Figure FIG. 4.

Effect of different bisphosphonates on exopolyphosphatase activity in human osteoblast-like cells and other cells. Exopolyphosphatase actvity was determined in total cell extracts from osteoblast-like cells (Obl, solid bars), yeast (hatched bars), rat liver (open bars), and HL60 cells, using polyP35 as substrate, in the presence of the following bisphosphonates (1 mg/ml): etidronate (Et, final concentration 4.85 mM), clodronate (disodium salt) (Cl, 2.79 mM), pamidronate (disodium salt) (Pa, 2.85 mM), and methylene-1,1-bisphosphonate (trisodium salt) (Me, 3.18 mM). Results are means (± SD) of three independent experiments. The specific activities in the absence of inhibitor were as follows: osteoblast-like cells, 3.51 ± 0.29 nmol/minute/mg of protein; yeast, 291 ± 13 nmol/minute/mg; rat liver, 598 ± 41 pmol/minute/mg; HL60, 413 ± 19 pmol/minute/mg. Significantly different from control, *p < 0.05, **p < 0.001.

Figure 5 shows the effect of various concentrations of etidronate on exopolyphosphatase activity in osteoblast-like cells; the IC50 was 2.04 mM. The inhibitory effect of this bisphosphonate on the enzyme from yeast and rat liver was similar (Fig. 5); the IC50 values amounted to 1.75 and 1.31 mM, respectively. Etidronate inhibited both the degradation of longer (high molecular weight) polyP chains and shorter polyP chains by the purified exopolyphosphatase (yeast) when analyzed by electrophoresis on urea/polyacrylamide gels (Fig. 6). The IC50 values for clodronate and pamidronate on the yeast exopolyphosphatase were 3.06 and 3.24 mM, respectively (data not shown).

Figure FIG. 5.

Inhibition of exopolyphosphatase activity by etidronate in human osteoblast-like cells, yeast, and rat liver. Exopolyphosphatase assays were performed in total cell extracts, using polyP35 as substrate. Results are means of three independent experiments; SD ≤ 15%.

Figure FIG. 6.

Effect of etidronate on degradation of polyP by purified exopolyphosphatase (yeast). The32P-labeled polyP preparation (lane b) was incubated with the enzyme for 10 minutes in the presence of 1000 μg/ml (4.85 mM; lane c), 600 μg/ml (2.91 mM; lane d), 300 μg/ml (1.46 mM; lane e), 100 μg/ml (0.49 mM; lane f), 30 μg/ml (0.15 mM; lane g), 10 μg/ml (0.05 mM; lane h), or 0 μg/ml of etidronate (0 mM; lane i), as described in the Materials and Methods. The chain lengths of the degradation products of the [32P]polyP preparation were analyzed by gel electrophoresis on a 16.5% polyacrylamide/7 M urea gel. Lane a, thermal degradation of [32P]polyP (10 mM HCl, 70°C, 3 minutes). Shown is the autoradiogram of the gel. The migration positions of high-molecular-weight (HMW; approximate size, 35–40 Pi residues) polyP, medium-sized polyPn, orthophosphate (Pi), and trimetaphosphate (mP3) are indicated. The chain lengths of polyPs (in number of residues) can be determined by counting bands on the gel (from bottom to top) starting with Pi.

The inhibitory profile of bisphosphonates on exopolyphosphatase activity differed from that found for inorganic pyrophosphatase activity as determined by using a commercial enzyme preparation from yeast (Boehringer Mannheim). In contrast to the exopolyphosphatase, which was mostly sensitive to etidronate, clodronate was the most potent inhibitor of pyrophosphatase (Fig. 7). The following IC50 values were determined: clodronate, 0.28 mM; etidronate, 0.84 mM; methylene-1,1-bisphosphonate, 3.04 mM; and pamidronate, >5 mM.

Figure FIG. 7.

Inhibition of pyrophosphatase activity by different bisphosphonates. The effect of various concentrations of etidronate, clodronate, pamidronate, and methylene-1,1-bisphosphonate on the purified enzyme (yeast) was determined, using PPi as substrate. Results are means of three independent experiments; SD ≤ 10%.


In this report, we demonstrate that bone-forming human osteoblast-like cells possess comparably high levels of inorganic polyP. The total concentration of long-chain polyP in unstimulated osteoblast-like cells (= 528 μM, in terms of Pi residues), which mostly consisted of soluble polymers (394 μM), was 3.4- to 6.2-fold higher that in human gingival cells, erythrocytes, and PBMCs, and 10.3-fold higher than that in human blood plasma. The polyP content found in osteoblast-like cells was also higher than that determined in different tissues of rat (middle-aged animals), e.g., liver, 70 μM, and brain, 114 μM.8 Osteoblast-like cells also possess a polyP-degrading exopolyphosphatase activity. In unstimulated cells, the specific activity of the enzyme amounted to 0.56 μmol of Pi liberated/h/mg of protein. The intracellular localization of polyP and of exopolyphosphatase activity in osteoblast-like cells is at present not known; from studies using other cells and tissues, polyP is localized predominantly in the nuclei,4,6,8 but an enrichment in the mitochondria and the plasma membrane has also been reported.52

Osteoblasts are known to respond to glucocorticoids, 1,25(OH)2D3, and growth factors. They possess high levels of ALP, which participates in mineralization; β-glycerophosphate induces changes in ALP activity. We found that simultaneous treatment of osteoblast-like cells with dexamethasone, β-glycerophosphate, EGF, and ascorbic acid results in a strong decrease in cellular polyP content and exopolyphosphatase activity. The amount of soluble long-chain polyP, but not the amount of insoluble long-chain polyP, further decreased after an additional treatment with 1,25(OH)2D3. However, additional treatment with 1,25(OH)2D3 caused an increase in the exopolyphosphatase activity. 1,25(OH)2D3 is known as a stimulator of bone resorption; it increases the expression of genes that are associated with the mineralization process, such as osteocalcin in mature osteoblasts, resulting in an enhanced mineralization. However, expression of osteocalcin is suppressed by dexamethasone. However, combined treatment of osteoblasts with 1,25(OH)2D3 and dexamethasone has been shown to result in a synergistic increase in transcriptional activity.53

The finding that polyP content and exopolyphosphatase activity decrease at the same time (following treatment of the cells with dexamethasone, β-glycerophosphate, EGF, and ascorbic acid) might be surprising. However, polyP molecules are dynamic polymers, displaying a rapid turnover rate.6 The level of polyP is expected to be regulated both by polyP anabolic and polyP catabolic enzyme activities. In the present study, only the activity of polyP-degrading exopolyphosphatase has been determined. An enzyme for the synthesis of polyP, like polyP kinase in bacteria and yeast, has not been identified as yet in mammalian cells,6 and such an enzyme could influence the cellular polyP content, too.

The decrease in exopolyphosphatase activity after treatment of osteoblast-like cells with dexamethasone, β-glycerophosphate, EGF, and ascorbic acid, and the increase in the activity of this enzyme after additional treatment with 1,25(OH)2D3 were accompanied by a decrease and increase, respectively, of inorganic pyrophosphatase and ALP activity.

Similar effects of dexamethasone, growth factors, and 1,25(OH)2D3 have also been observed for the ecto-nucleoside triphosphate pyrophosphatase activity in human osteoblast-like cells.54 This enzyme is located on the cell surface and is assumed to serve in the generation of extracellular PPi in bone.55 Extracellular PPi is most likely involved in the regulation of mineralization in bone tissue. Dexamethasone causes a decrease in ecto-nucleoside triphosphate pyrophosphatase activity, while 1,25(OH)2D3 induces an increase in enzyme activity.54

Our studies revealed that exopolyphosphatase activity present in osteoblast-like cells, and also in yeast, rat tissue, and human leukemia cell line HL60, is inhibited by the bisphosphonates etidronate and, to a lesser extent, clodronate and pamidronate. The bisphosphonate methylene-1,1-bisphosphonate did not display a significant effect on exopolyphosphatase activity at a concentration of 1 mg/ml. The clinically applied agent etidronate significantly inhibited exopolyphosphatase activity already at a concentration of 0.49 mM. A nearly complete inhibition of the yeast enzyme was found at an etidronate concentration of 2.91 mM. However, it should be noted that the potency of the studied bisphosphonates in inhibition of exopolyphosphatase activity is inversely correlated to their ability to inhibit bone resorption, which increases in the following order: etidronate, clodronate, pamidronate. However, clodronate was found to be the most potent inhibitor of inorganic pyrophosphatase activity, but of the bisphosphonates tested, pamidronate was the least effective again.

The mode of action by which bisphosphonates inhibit bone resorption is not clear but most likely more than one mechanism is involved.56 Bisphosphonates may act on bone-resorbing osteoclasts either directly or indirectly by inhibiting the formation and function of osteoclast. At least some bisphosphonates act as inhibitors of an osteoclast tyrosine phosphatase activity possibly involved in the regulation of osteoclast formation and function.57 Recent results, however, indicate that the inhibitory action of the bisphosphonates on bone resorption in part is mediated by osteoblasts.58 Bisphosphonates have been shown to induce osteoblasts to synthesize an osteoclast inhibitor.59 They also have direct effects on osteoblastic activity, by prevention of mineralization.60

In bone, extracellular PPi has been identified as a calcification inhibitor which is also present in plasma61 and urine.27 It is assumed that such inhibitors must be removed from the site of mineralization before calcification can occur.26 PPi is known to be hydrolyzed by ALP generating Pi for calcium phosphate precipitation.62 It may also be split by PPi present in bone tissue. These enzymes could locally hydrolyze PPi in bone, thus allowing local formation of hydroxyapatite crystals. The increase in both enzyme activities induced by 1,25(OH)2D3 would be expected to result in an increased production of Pi and a decrease in PPi concentration, thus modulating mineralization during bone formation. Bone mineralization indeed has been found to be associated with changes in pyrophosphatase activity.63

It is reasonable to assume that polyP may also be involved in modulation of the mineralization process in bone tissue, in addition to PPi. It has been shown that long-chain polyP inhibits hydroxyapatite precipitation more strongly than PPi.26 Therefore, the presence of an exopolyphosphatase might be necessary for a tissue to mineralize. Our results show that bone cells contain such an enzyme that may locally destroy polyP, in addition to PPi-degrading pyrophosphatase and ALP activities.


We thank Mrs. A. Sapotnick for excellent technical assistance. We also thank Dr. F. Bauss (Präklinische Forschung und Entwicklung, Boehringer Mannheim GmbH, Mannheim, Germany) for valuable discussion. This work was supported by a grant from the Deutsche Forschungsgemeinschaft (Schr 277/8–1).