- Email: [email protected]
Evidence for the involvement of bone morphogenetic protein-2 in phenytoin-stimulated osteocalcin secretion in human bone cells
Archives of Oral Biology 45 (2000) 647±655
www.elsevier.com/locate/archoralbio
Evidence for the involvement of bone morphogenetic protein-2 in phenytoin-stimulated osteocalcin secretion in human bone cells H. Koyama a, O. Nakade a,*, T. Saitoh b, T. Takuma c, T. Kaku a a
Department of Oral Pathology, School of Dentistry, Health Sciences University of Hokkaido, Ishikari-Tobetsu, 061-0293, Japan Department of Operative Dentistry and Endodontology, School of Dentistry, Health Sciences University of Hokkaido, IshikariTobetsu, 061-0293, Japan c Department of Oral Biochemistry, School of Dentistry, Health Sciences University of Hokkaido, Ishikari-Tobetsu, 061-0293, Japan b
Accepted 14 March 2000
Abstract Recent work has shown that the actions of phenytoin on bone cell proliferation and dierentiation are, in part, mediated through the upregulation of transforming growth factor-b1 (TGF-b1). The present study was undertaken to examine the eect of phenytoin on bone morphogenetic proteins (BMP)-2 and -4, which are well-recognized osteoinductive proteins of the TGF-b superfamily, in osteoblastic cells. Treatment with 5±50 mM of phenytoin increased the amount of mRNA for BMP-2 after a 0.5±24 h incubation in normal human mandible-derived bone cells (HOB-M cells), but failed to aect the mRNA for BMP-4. Phenytoin treatment for 48 h signi®cantly increased the secretion of BMP-2 by approx. four-fold, at an optimal concentration of 10 mM. While TGF-b1 inhibited osteocalcin secretion from HOB-M cells, both phenytoin and BMP-2 signi®cantly stimulated it. Importantly, the stimulatory eects of phenytoin on osteocalcin release were completely blocked by the neutralizing antihuman BMP2 monoclonal antibody. These results indicate that the stimulatory action of phenytoin on osteocalcin secretion in normal human bone cells is mediated, at least partly, through the upregulation of BMP-2, rather than that of TGFb1. 7 2000 Elsevier Science Ltd. All rights reserved. Keywords: Phenytoin; Bone morphogenetic proteins (BMP); Normal human bone cells; Osteocalcin secretion; Transforming growth factor-b (TGF-b)
Abbreviations: ANOVA, analysis of variance; BMP, bone morphogenetic protein; BSA, bovine serum albumin; DMEM, Dulbecco's modi®ed Eagle's medium; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; 1,25(OH)2D3; 1,25-dihydroxyvitamin D3; PICP, procollagen type I c-peptide; RT± PCR, reverse transcription±polymerase chain reaction; TBS, tris-buered saline; TGF-b, transforming growth factor-b. * Corresponding author. Tel.: +81-1332-3-1390; fax: +811332-3-1390. E-mail address: [email protected] (O. Nakade).
1. Introduction Phenytoin is a widely used therapeutic agent for seizure disorders. It is eective but has serious sideeects on bone and calcium metabolism (Richens and Rowe, 1970; Hahn and Avioli, 1975). Of the known side-eects, two are relevant to oral biology. First, many patients treated with phenytoin develop gingival hyperplasia, which is caused by phenytoin-
0003-9969/00/$ - see front matter 7 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 0 0 3 - 9 9 6 9 ( 0 0 ) 0 0 0 3 6 - 4
648
H. Koyama et al. / Archives of Oral Biology 45 (2000) 647±655
induced increases in the proliferation of gingival ®broblasts and collagen synthesis (Ebadi and Scott, 1971). Secondly, chronic phenytoin therapy increases the thickness and density of maxillary and calvarial bones (Kattan, 1970). Patients on phenytoin therapy show signi®cantly less alveolar bone loss than untreated controls (Seymour et al., 1985). Phenytoin treatment also accelerates the healing of mandibular bone fractures in the rabbit (Sklans et al., 1967). Lau and co-workers recently showed that phenytoin treatment increases bone cell proliferation, dierentiation, and collagen synthesis in vitro (Nakade et al., 1995), and histomorphometric measures of bone formation in vivo (Ohta et al., 1995a,b), and that it also increases serum biochemical markers of bone formation in patients (Lau et al., 1995). Together, these observations strongly indicate that phenytoin has anabolic actions, particularly on bone. Understanding the mechanism by which phenytoin stimulates bone formation is not only essential to our understanding of the pathogenesis of its skeletal side-eects, but also may allow us to develop a rationale basis in designing eective osteogenic drugs. Thus, we are interested in the molecular mechanism of the osteogenic actions of phenytoin. We have recently provided compelling evidence that its osteogenic actions are, in part, mediated through the upregulation of TGF-b1 synthesis. Accordingly, phenytoin increased the synthesis and release of active TGF-b1 in human bone cells and the stimulatory eects of phenytoin on bone cell proliferation and alkaline phosphatase activity were completely blocked by a neutralizing antibody against TGF-b1 (Nakade et al., 1996). On the other hand, while phenytoin stimulates osteocalcin secretion in bone cells (Nakade et al., 1995; Lau et al., 1995), TGF-b1 inhibits, rather than stimulates, osteocalcin secretion in human bone cells in vitro (Bonewald et al., 1992; Wergedal et al., 1992; Pirskanen et al., 1994). Thus, some of the osteogenic actions of phenytoin, such as the stimulation of osteocalcin secretion, may not be mediated through TGFb1. Because BMP-2 and -4, members of the TGF-b superfamily (Wozney et al., 1988), are known to stimulate osteocalcin secretion in the osteoblastic cell lineage (Hiraki et al., 1991; Rickard et al., 1994; Suzuki et al., 1995), we postulate that phenytoin stimulates the synthesis of BMP-2 and/or -4 in normal human osteoblastic cells, and that this is responsible for the phenytoininduced synthesis of osteocalcin. We have now sought to test this hypothesis by determining the eects of phenyotin on the synthesis of BMP-2 and -4, and by evaluating whether BMP-2 and/ or -4 is responsible for phenytoin-induced osteocalcin secretion in normal human osteoblasts.
2. Materials and methods 2.1. Materials Tissue-culture supplies were obtained from Iwaki Glass Co. (Funabashi, Japan) or Falcon (Oxnard, CA, USA). DMEM, 0.5% trypsin±5.3 mM EDTA solution, and Trizol1, a reagent for the single-step isolation of total RNA, were from GIBCO/BRL Life Technologies (Grand Island, NY, USA). Iron-supplemented bovine calf serum was purchased from JRH Biosciences (Lenexa, KS, USA). BSA and Triton X-100 were from Sigma Chemical Co. (St. Louis, MO, USA). 1,25(OH)2D3 was obtained from Biomol Research (Plymouth Meeting, PA, USA), and a Gla-type human osteocalcin enzyme immunoassay kit was purchased from Takara (Ohtsu, Japan). Sense and antisense primers for GAPDH and all chemicals for RT-PCR, unless otherwise stated, were purchased from Clontech Laboratories Inc. (Palo Alto, CA, USA). Recombinant human TGF-b1 was from R&D System (Minneapolis, MN, USA). Recombinant human BMP-2 and monoclonal antibody speci®c for mouse anti-human BMP-2 antibody were generous gifts of the Genetic Institute (Boston, MA, USA). Non-immune mouse IgG was obtained from Seikagaku Kogyo Co. (Tokyo, Japan). Horseradish peroxidase-conjugated anti-mouse IgG antibody and dot-blot equipment (Bio Dot1) were purchased from Cosmo Bio (Tokyo, Japan). Enhanced chemiluminescence Western-blotting detection reagents and nitrocellulose membrane were products of Amersham Life Science (Buckinghamshire, England). All other chemicals were of reagent grade, either from Sigma or from Kanto Chemical Co. (Tokyo, Japan). 2.2. Cell cultures Normal human bone cells were isolated from mandibular samples by collagenase digestion as previously described (Wergedal et al., 1984). The resulting bone cells (HOB-M) were shown to be of osteoblastic nature, based on their responsiveness to parathyroid hormone for increased cAMP production and to 1,25(OH)2D3 for alkaline phosphatase activity and osteocalcin secretion (Wergedal and Baylink, 1984; Nakade et al., 1995). The cells were maintained in DMEM supplemented with 10% bovine calf serum and passaged every week at a 1:4 dilution ratio. Cells from passages 3±7 were examined in this study. Phenytoin stocks were dissolved in 95% ethanol and diluted in DMEM containing 0.01% BSA immediately before use. The ®nal ethanol concentration in each assay was 0.35%. We realize that this concentration of ethanol is relatively high. There is evidence that high concentrations of ethanol could have an adverse eect on bone cell proliferation and dierentiation (Farley et
H. Koyama et al. / Archives of Oral Biology 45 (2000) 647±655
al., 1985). Unfortunately, due to the relatively low solubility of phenytoin even in ethanol, we were unable to reduce the ®nal concentration further. However, we should emphasize that the solvent vehicle controls in this study also contained 0.35% ethanol, and that we were able to detect osteogenic eects of phenytoin in spite of this high ethanol concentration (Nakade et al., 1995). 2.3. Measurement of BMP-2 and -4 mRNA BMP-2 and -4 mRNA were assessed with a semiquantitative RT-PCR. In brief, total RNA from control and phenytoin-treated osteoblasts was isolated using Trizol1 and quanti®ed by spectrophotometry (Chomczynski and Sacchi, 1987). Reverse transcription was carried out in a 20 ml volume containing 1.0 mg total RNA, 2.5 U RNAsin, 5 mM MgCl2, 50 mM KCl, 10 mM Tris (pH 8.3), 2.0 M oligo (dT) 11±18 primer, and 10 U of reverse transcriptase. The samples were incubated at 378C for 60 min. Reverse transcriptase was then heat-inactivated by incubating at 958C for 7 min. For each ampli®cation reaction (PCR), 2 ml cDNA (equivalent to 100 ng total RNA) was added to a 100 ml reaction mixture, consisting of 10 mM Tris± HCl, 50 mM KCl, 0.2% Triton X-100, 1.5 mM MgCl2, and 0.2 mM each of dATP, dCTP, dGTP, and dTTP. Appropriate sense and antisense primers for BMP-2 and BMP-4, respectively, and 0.2 U of Thermophime plus DNA polymerase were then added to the reaction mixture to initiate the PCR reaction. Sense and antisense BMP-2 and BMP-4 primers (Ogose et al., 1996) were synthesized by Takara Ltd. (BMP-2 forward primer: 5 '-GCTGTACTAGCGACACCCCAC-3 '; BMP-2 reverse primer: 5'-TCATAAAACCCTGCAACAGCCAACTCG-3 '; BMP-4 forward primer: 5 '-ACTGGTCCACCACAATGTGACACG-3; BMP-4 reverse primer: 5 '-GCTGAAGTCCACATAGAGCGAGTG-3. Ampli®cation was for 25 cycles for BMP-4 and 30 cycles for BMP-2. Each ampli®cation cycle consisted of denaturation at 948C for 45 sec, primer annealing at 608C for 45 sec, and primer extension at 728C for 2 min. For normalization of the RNA loading, RT-PCR of GADPH was also included in each RT-PCR reaction as an internal control. [Sense and antisense primers for GAPDH: forward primer: 5 '-TGAAGGTCGGAGTCAACGGATTTGGT-3 '; reverse primer: 5 '-CATGTGGGCCATGAGGT CCACCAC-3 ']. The ampli®cation products were electrophoresed in 2% agarose gel, stained with ethidium bromide, illuminated with ultraviolet light, photographed, and analysed in a computerized FMBIO1 ¯uorescent image-analyser (Hitachi Software Engineering Co., Ltd., Yokohama, Japan). The relative intensity of the ¯uorescence of the BMP-2- and BMP4-speci®c bands was expressed as percentage of the
649
intensity of the GAPDH band (Bodamyali et al., 1998). 2.4. Dot-blot immunoassay of BMP-2 The amount of BMP-2 protein released into the conditioned medium of HOB-M cells was estimated by a semiquantitative dot-blot immunoassay (Matikainen et al., 1999). In brief, HOB-M cells were treated with the indicated concentrations of phenytoin or the solvent vehicle for 48 h. The conditioned medium of each treatment group was then separately collected and concentrated by SpeedVac vacuum centrifugation. The ®nal volume of the concentrated medium was adjusted with TBS (pH 7.5) to normalize for cellular protein content so that it would yield the same volume per cellular protein content in each well. This was necessary to account for dierences in the cell number, as phenytoin stimulates the proliferation of human bone cells. As a positive control, recombinant human BMP-2 (®nal concentration 30 ng/ml) in DMEM containing 0.01% BSA was included in each assay. The adjusted medium was serially diluted with TBS (pH 7.5), and the samples (500 ml/well) were blotted by vacuum on a nitrocellulose paper wetted with TBS. After blotting, each well was blocked with 1% BSA in 0.05% Tween 20-containing TBS (TTBS) for 30 min. Mouse antihuman BMP-2 monoclonal antibody (1:400 dilution) was used as the primary antibody, and horseradish peroxidase-conjugated anti-mouse IgG antibody raised in sheep was used as the second antibody. All steps were performed inTTBS. Enhanced chemiluminescence Western-blotting detection reagents were used for visualization of the dots (Whitehead et al., 1979; Matikainen et al., 1999). This dot-blot assay, albeit semiquantitative, allowed us to estimate and compare the relative amount of BMP-2 in each test conditioned medium. 2.5. Measurement of osteocalcin secretion Osteocalcin secretion in the conditioned medium was assayed using a speci®c Gla-type human osteocalcin enzyme immunoassay kit (Nakade et al., 1999). In brief, HOB-M cells were plated at a density of 30,000 cells/well in DMEM supplemented with 10% bovine calf serum in 48-well culture plates. After plating for 24 h, the culture medium was changed to fresh serumfree DMEM, supplemented with 0.01% BSA and 10ÿ8 M 1,25(OH)2D3 as previously described (Nakade et al., 1995). Twenty-four hours later, eectors (BMP-2, TGF-b1, or phenytoin) or solvent control were added, and the cells incubated for an additional 48 h. The amount of osteocalcin in the conditioned medium was determined by enzyme immunoassay (Wiechelman et al., 1988) and normalized against cellular protein con-
650
H. Koyama et al. / Archives of Oral Biology 45 (2000) 647±655
tent to adjust for the cell number. For the anti-BMP-2 antibody blocking studies, 1 h after changing the culture medium to fresh serum-free medium, either the neutralizing anti-human BMP-2 speci®c monoclonal antibody (2 mg/ml) or non-immune normal mouse IgG (2 mg/ml) were added to the culture well (Hatakeyama et al., 1993). Phenytoin (5±50 mM) or solvent vehicle were added to the cells 0.5 h later. The cells were then incubated for an additional 48 h, and the amount of osteocalcin secretion was determined as described above.
2.6. Statistical methods
Results are shown as mean 2SEM (n = 6). The statistical signi®cance of the dierence from the solvent vehicle controls was analysed by two-tailed Student ttest. The statistical signi®cance of the dierence between the groups was determined by one-way ANOVA. The dierence was considered signi®cant when p < 0.05.
Fig. 1. RT-PCR analyses of the eects of short-term (0.5 h) exposure to phenytoin on the BMP-2 and -4 mRNA. (A) Agarose gel electrophoresis of BMP-2 and -4 mRNA ampli®ed by RT-PCR; (B) mRNA of BMP-2 (a) and BMP-4 (b) calculated as a percentage of endogenous GAPDH mRNA and expressed as a relative value. The data are shown as a percentage of the corresponding control (mean2SEM, n = 4 for each group).
H. Koyama et al. / Archives of Oral Biology 45 (2000) 647±655
3. Results 3.1. Eect of phenytoin on mRNA levels of BMP-2 and -4 As both BMP-2 and -4 are well-established growth factors that stimulate osteocalcin secretion in the osteoblastic cell lineage, we evaluated the eects of phenytoin on the mRNA of BMP-2 and -4. Fig. 1 shows that exposure to osteogenic doses of phenytoin rapidly (within 0.5 h) increased the BMP-2 mRNA (shown as percentage of GAPDH transcript) with an optimal concentration between 10 and 50 mM ( p < 0.001, ANOVA). The stimulation was seen even after
651
the 12-h incubation (Fig. 2; p < 0.001, ANOVA) and was still evident after a 24-h incubation (data not shown). This indicates that phenyotin treatment increased the mRNA of BMP-2 and that the stimulatory eect of phenytoin on BMP-2 mRNA was sustained at least for 24 h in normal human bone cells. In contrast, the same osteogenic doses of phenytoin failed to aect the amount of BMP-4 mRNA in either shortterm (0.5 h) or long-term (12±24 h) treatments (data not shown). 3.2. Eect of phenytoin on the BMP-2 secretion We next tested whether phenytoin increased the se-
Fig. 2. RT-PCR analyses of the eects of long-term (12 h) exposure of phenytoin on the BMP-2 and -4 mRNA. (A) Agarose gel electrophoresis of BMP-2 and -4 mRNA ampli®ed by RT-PCR; (B) mRNA of BMP-2 (a) and BMP-4 (b) calculated as a percentage of endogenous GAPDH mRNA and expressed as a relative value of each mRNA. Data are shown as a percentage of the corresponding control (mean2SEM, n = 4 for each group).
652
H. Koyama et al. / Archives of Oral Biology 45 (2000) 647±655
cretion of BMP-2 protein from human HOB-M into conditioned medium, with a semiquantitative dot-blot assay using a monoclonal antibody speci®c for human BMP-2. Fig. 3 shows that the osteogenic doses of phenytoin markedly increased the release of BMP-2 into the conditioned medium by approx. four-fold, with an optimal phenytoin concentration of 10 mM. 3.3. Eect of BMP-2 on osteocalcin secretion Because phenytoin treatment increases secretion of osteocalcin by human bone cells (Nakade et al., 1995; Lau et al., 1995), and because TGF-b1, which, we believe, mediates the stimulatory actions of phenytoin on bone cell proliferation and alkaline phosphatase activity (Nakade et al., 1996), is known to inhibit osteocalcin secretion by human bone cells, we next examined the eect of recombinant human BMP-2 on osteocalcin secretion in HOB-M cells and compared that to the eects of phenytoin and TGF-b1. Fig. 4 shows that both BMP-2 and phenytoin signi®cantly stimulated osteocalcin secretion to a similar concentration ( p < 0.01 and p < 0.05, ANOVA, respectively). Conversely, the osteogenic dose of TGF-b1 markedly inhibited its secretion by HOB-M ( p < 0.001, ANOVA).
Fig. 3. Dot-blotting analyses of the release of BMP-2 into medium induced by phenytoin in HOB-M cells (48-h treatment). Recombinant human BMP-2 (®nal concentration 30 ng/ml) in DMEM containing 0.01% BSA was included as a positive control in each assay. (Note: the intensity of BMP-2 immunoreaction from the 4-diluted medium with 10 mM phenytoin treatment is as strong as that from the control medium (1), indicating that phenytoin enhanced the BMP-2 release by approx. four-fold; this is only an estimate.)
3.4. Eect of anti-BMP-2 antibody on phenytoinmediated increase in osteocalcin secretion We reasoned that, if the stimulatory action of phenytoin on osteocalcin secretion was mediated through the local bone-cell secretion of BMP-2, the stimulatory eect of phenytoin would be completely blocked by a neutralizing antibody against BMP-2. Thus, ®nally we examined whether the stimulatory eects of phenytoin on osteocalcin secretion were abolished by anti-BMP-2 antibody. Fig. 5 shows that the treatment of HOB-M cells with 5±10 mM of phenytoin in the presence of 2 mg/ml of non-immune normal mouse IgG signi®cantly stimulated osteocalcin secretion ( p < 0.05, ANOVA). However, the addition of 2 mg/ml of antiBMP-2 antibody completely abolished the stimulatory eects of phenytoin. 4. Discussion There is circumstantial evidence that at least some of the anabolic actions of phenytoin may be mediated through the upregulation of growth-factor genes in target cells and tissues. It has been suggested that phenytoin increases the synthesis and release of epithelial growth factor in gingival ®broblasts (Brown et al., 1991) and platelet-derived growth factor-B in macrophages and monocytes (Dill et al., 1993). The release of these growth factors has been suggested to be responsible for the gingival hyperplasia. However, we believe we were the ®rst to provide conclusive evidence that phenytoin acts through local upregulation of growth-factor genes to exert its anabolic eects, by showing that the stimulatory actions of phenytoin on bone cell proliferation and alkaline phosphatase activity were mediated through upregulation of TGF-b1 expression (Nakade et al., 1996). The present study, which shows that osteogenic doses of phenytoin also upregulate BMP-2 gene expression in human bone cells, is further consistent with the premise that phenytoin exerts its anabolic actions through upregulation of growth-factor genes. Although we have compelling evidence that the bone-cell proliferative eect of phenytoin is mediated through the upregulation of TGF-b1 (Nakade et al., 1996), the signi®cance of the phenytoin-mediated upregulation of BMP-2 gene expression in human bone cells cannot be underestimated. Accordingly, this study provides three pieces of strong evidence that the phenytoin-mediated increase in osteocalcin secretion is mediated through the upregulation of the expression of the BMP-2 gene. First, phenytoin markedly stimulated the synthesis and release of BMP-2. Second, phenytoin and BMP-2 each signi®cantly increased osteocalcin secretion from HOB-M to similar concentrations, while
H. Koyama et al. / Archives of Oral Biology 45 (2000) 647±655
653
Fig. 4. Comparison among the eects of BMP-2, TGF-b1, or phenytoin on osteocalcin secretion from normal human mandible-derived bone cells (HOB-M cells). HOB-M cells were exposed to BMP-2, TGF-b1, phenytoin, or vehicle control for 48 h.
TGF-b1 signi®cantly reduced its secretion by the same cells. Third, and more importantly, the stimulatory eects of phenytoin on osteocalcin secretion were completely blocked by a neutralizing antibody against BMP-2. Osteocalcin is an osteoblast-speci®c protein and a well-recognized dierentiation marker of mature osteoblasts (Lian and Gundberg, 1988). Therefore, ®ndings of this and our past studies lead us to speculate that phenytoin may stimulate dierent aspects of bone formation, i.e., osteoblast proliferation as compared to osteoblast maturation (e.g. osteocalcin secretion) through upregulation of dierent bone growth-factor genes (i.e. TGF-b1 and BMP-2, respectively). In addition, recent studies suggest that human osteoblasts may possess a speci®c receptor for osteocalcin (Bodine
Fig. 5. Eects of antihuman BMP-2 antibody on osteocalcin secretion stimulated by phenytoin. HOB-M cells were pretreated with anti-human BMP-2 monoclonal antibody (2 mg/ ml) (solid bars) or with non-immune mouse IgG (2 mg/ml) (open bars) for 0.5 h before the addition of 0, 5, 10, or 50 mM phenytoin for 48 h.
and Komm, 1999) and that osteocalcin may play an important part in the regulation of bone formation and resorption (Ducy et al., 1996; Boskey et al., 1998). Therefore, it is possible that the phenytoin-induced osteocalcin secretion through upregulation of BMP-2 could have additional biological functions other than being a marker of osteoblast maturation. Bone morphogenetic proteins were originally identi®ed as compounds that induce bone and cartilage formation in ectopic extraskeletal sites (e.g. muscles) in vivo Wozney et al., 1988; Wang et al., 1988). Extensive studies have demonstrated that bone morphogenetic proteins, including BMP-2, are potent bone cell-dierentiating factors as well as bone-formation stimulators (Wozney et al., 1988; Harris et al., 1995). Therefore, upregulation of BMP-2 expression in response to phenytoin could have more important biological consequences than just to stimulate osteocalcin secretion. More speci®cally, there is evidence that phenytoin in the rabbit accelerates the healing of mandibular bone fractures (Sklans et al., 1967) and promotes wound healing in both hard and soft tissues (Swann et al., 1975; Frymoyer, 1976; Anstead et al., 1996; DaCosta et al., 1998) as well as in the skin. Because both TGFb1 and BMP-2 are essential in bone-fracture healing (Si et al., 1997) and the healing of bone wounds in vivo (Toriumi et al., 1991; Bolander, 1992) as well as of soft tissues (Stelnicki et al., 1998), it is possible that enhanced wound healing in both hard and soft tissues by phenytoin is mediated through upregulation of the expression of these two growth factors and/or the interaction between them. Much additional work is needed to con®rm this hypothesis. Finally, one of the most intriguing ®ndings in this study, perhaps, is the ®nding that, while osteogenic doses of phenytoin markedly increased the expression of BMP-2 in HOB-M, the agent at the same doses had no eect on the expression of the BMP-4 gene. This
654
H. Koyama et al. / Archives of Oral Biology 45 (2000) 647±655
observation is puzzling, as among all bone morphogenetic proteins, BMP-2 and BMP-4 share the highest sequence similarity and many of the same biological functions (Wozney and Rosen, 1998). It may be speculated that BMP-2 and BMP-4 might have dierent functional roles in bone formation and/or wound healing. Understanding the reason why phenytoin upregulates the expression of BMP-2 but not BMP-4 might provide important insights into potential functional dierences between these two. In summary, we demonstrate, to the best of knowledge for the ®rst time, that osteogenic concentrations of phenytoin increase the synthesis and release of BMP-2 in human mandible-derived bone cells, and that the stimulatory eects of phenytoin on osteocalcin secretion are mediated, in part, by the upregulation of BMP-2.
Acknowledgements We wish to thank Dr K-HW. Lau, Loma Linda University, for helpful discussions.
References Anstead, G.M., Hart, L.M., Sunahara, J.F., Liter, M.E., 1996. Phenytoin in wound healing. Ann. Pharmacother. 30, 768±775. Bodamyali, T., Bhatt, B., Hughes, F.J., Winrow, V.R., Kanczler, J.M., Simon, B., Abbott, J., Blake, D., Stevens, C., 1998. Pulsed electromagnetic ®elds simultaneously induce osteogenesis and upregulate transcription of bone morphogenetic proteins 2 and 4 in rat osteoblasts in vitro. Biochem. Biophys. Res. Com. 250, 458±461. Bodine, P.V.N., Komm, B.S., 1999. Evidence that conditionally immortalized human osteoblasts express an osteocalcin receptor. Bone 25, 535±543. Bolander, M.E., 1992. Regulation of fracture repair by growth factors. Proc. Soc. Exp. Biol. Med. 200, 165±170. Bonewald, L.F., Kester, M.B., Schwarty, Z., Swain, L.D., Khare, A., Johnson, T.L., Leach, R.J., Boyan, B.D., 1992. Eects of combining transforming growth factor-b1 and 1,25-dihydroxyvitamin D3 on dierentiation of a human osteosarcoma (MG-63). J. Biol. Chem. 267, 8943±8949. Boskey, A.L., Gadaleta, S., Gundberg, C., Doty, S.B., Ducy, P., Karsenty, G., 1998. Fourier transform infrared microspectroscopic analysis of bones of osteocalcin-de®cient mice provides insight into the function of osteocalcin. Bone 23, 187±196. Brown, R.S., Beaver, W.T., Bottomley, W.K., 1991. On the mechanism of drug-induced gingival hyperplasia. J. Oral Pathol. Med. 20, 201±209. Chomczynski, P., Sacchi, N., 1987. Single-step method of RNA isolation by acid guanidium thiocyanate±phenol± chloroform extraction. Anal. Biochem. 162, 156±159. DaCosta, M.L., Regan, M.C., al Sader, M., Leader, M.,
Bouchier-Hayes, D., 1998. Diphenylhydantoin sodium promotes early and marked angiogenesis and results in increased collagen deposition and tensile strength in healing wounds. Surgery 123, 287±293. Dill, R.E., Miller, E.K., Weil, T., Lesley, S., Farmer, G.R., Iacopino, A.M., 1993. Phenytoin increases gene expression for platelet-derived growth factor B chain in macrophages and monocytes. J. Periodontol. 64, 169±173. Ducy, P., Desbois, C., Boyce, B., Pinero, G., Story, B., Dunstan, C., Smith E, Bonadio, J., Goldstein, S., Gundberg, C., Bradley, A., Karsenty, G., 1996. Increased bone formation in osteocalcin-de®cient mice. Nature 382, 448±452. Ebadi, M.S., Scott, P.M., 1971. Increase in collagen level by diphenylhydantoin as possible mechanism of drug-induced gingival hyperplasia. Clin. Toxicol. 4, 39±46. Farley, J.R., Fitzsimmons, R., Taylor, A.K., Jorch, U.M., Lau, K.-H.W., 1985. Direct eects of ethanol on bone resorption and formation in vitro. Arch. Biochem. Biophys. 238, 305±314. Frymoyer, J.W., 1976. Fracture healing in rats treated with diphenylhydantoin (dilantin). J. Trauma 16, 368±370. Hahn, T.J., Avioli, L.V., 1975. Anticonvulsant osteomalacia. Arch. Intern. Med. 135, 997±1000. Harris, S.E., Feng, J.Q., Harris, M.A., Ghosh-Choudhury, N., Dallas, J.M., Wozney, J.M., Mundy, G.R., 1995. Recombinant bone morphogenetic protein 2 accelerates bone cell dierentiation and stimulates BMP-2 mRNA expression and BMP-2 promoter activity in primary fetal rat calvarial osteoblast cultures. Mol. Cell. Dier. 3, 137±155. Hatakeyama, S., Ohara-Nemoto, Y., Yakumoto, S., Satoh, M., 1993. Expression of morphogenetic protein in human adenocarcinoma cell line. Biochem. Biophys. Res. Com. 190, 695±701. Hiraki, Y., Inoue, H., Shigeno, C., Sanma, Y., Bentz, H., Rosen, D.M., Asad, A., Suzuki, F., 1991. Bone morphogenetic proteins (BMP-2 and BMP-3) promote growth and expression of the dierentiated phenotype of rabbit chondrocytes and osteoblastic MC3T3-E1 cells. J. Bone Miner. Res. 6, 1373±1385. Kattan, K.R., 1970. Calvarial thickening after dilantin medication. Am. J. Roentgenol. Radium Ther. Nucl. Med. 110, 102±106. Lau, K.-H.W., Nakade, O., Baylink, D.J., 1995. Phenytoin increases markers of osteogenesis for the human species in vitro and in vivo. J. Clin. Endocrinol. Metabol. 80, 2347± 2353. Lian, J.B., Gundberg, C.M., 1988. Osteocalcin: biochemical considerations and clinical applications. Clin. Orthopaed. Rel. Res. 226, 267±291. Matikainen, T., Kakonen, S.S., Pettersson, K., Karp, M., Lovgren, T., Vaahanen, H.K., Hellman, J., 1999. Demonstration of the predominant urine osteocalcin fragments detectable by two-site immunoassays. J. Bone Miner. Res. 14, 431±438. Nakade, O., Baylink, D.J., Lau, K.-H.W., 1995. Phenytoin at micromolar concentrations is an osteogenic agent for human mandible-derived bone cells in vitro. J. Dent. Res. 74, 331±337. Nakade, O., Baylink, D.J., Lau, K.-H.W., 1996. Osteogenic
H. Koyama et al. / Archives of Oral Biology 45 (2000) 647±655 actions of phenytoin in human bone cells are mediated in part by TGF-b1. J. Bone Miner. Res. 11, 1880±1888. Nakade, O., Koyama, H., Ariji, H., Yajima, A., Kaku, T., 1999. Melatonin stimulates proliferation and type I collagen synthesis in human bone cells in vitro. J. Pineal. Res. 27, 106±110. Ogose, A., Motoyama, T., Hotta, T., Watanabe, H., 1996. Expression of bone morphogenetic proteins in human osteogenic and epithelial tumor cells. Pathol. Int. 46, 9±14. Ohta, T., Wergedal, J.E., Gruber, H.E., Baylink, D.J., Lau, K.-H.W., 1995a. Low dose phenytoin is an osteogenic agent in the rat. Calcif. Tissue Int. 56, 42±48. Ohta, T., Wergedal, J.E., Gruber, H.E., Baylink, D.J., Lau, K.-H.W., 1995b. Phenytoin and ¯uoride act in concert to stimulate bone formation and to increase bone volume in adult male rats. Calcif. Tissue Int. 56, 390±397. Pirskanen, A., Jaaskelainen, T., Maenpaa, P.H., 1994. Eect of transforming growth factor b1 on the regulation of osteocalcin synthesis in human MG-63 osteosarcoma cells. J. Bone Miner. Res. 9, 1635±1642. Richens, A., Rowe, D.J.F., 1970. Disturbance of calcium metabolism by anticonvulsant drugs. Br. Med. J. 4, 73±76. Rickard, D.J., Sullivan, T.A., Shenker, B.J., Leboy, P.S., Kahdan, I., 1994. Induction of rapid osteoblast dierentiation in rat bone marrow stromal cell cultures by dexamethasone and BMP-2. Dev. Biol. 161, 218±228. Seymour, R.A., Smith, D.G., Turnbull, D.N., 1985. The eects of phenytoin and sodium valproate on the periodontal health of adult epileptic patients. J. Clin. Periodontol. 12, 413±419. Si, X., Jin, Y., Yang, L., Tipoe, G.L., White, F.H., 1997. Expression of BMP-2 and TGF-b1 mRNA during healing of the rabbit mandible. Eur. J. Oral Sci. 105, 325±330. Sklans, S., Taylor, R.G., Shklar, G., 1967. Eect of diphenylhydantoin sodium on healing of experimentally induced fractures in rabbit mandibles. J. Oral Surg. 25, 310±319. Stelnicki, E.J., Longaker, M.T., Holmes, D., Vanderwall, K., Harrison, M.R., Largman, C., Homan, W.Y., 1998. Bone morphogenetic protein-2 induces scar formation and skin
655
maturation in the second trimester fetus. Plast. Reconstr. Surg. 101, 12±19. Suzuki, R., Kobayashi, M., Takiguchi, T., Yamaguchi, A., Nagumo, M., Hasegawa, K., 1995. Bone morphogenetic protein-2 enhances the expression of the osteoblast phenotypes in human periodontal ligament ®broblasts and osteoblasts. J. Bone Miner. Res. 10 (Suppl. 1), T216. Swann, W.P., Swenson, H.M., Shafer, W.G., 1975. Eects of dilantin on the repair of gingival wounds. J. Periodontol. 46, 302±305. Toriumi, D.M., Kotler, H.S., Luxenberg, D.P., Holtrop, M.E., Wang, E.A., 1991. Mandibular reconstruction with a recombinant bone-inducing factor. Arch. Otolaryngol. Head Neck Surg. 17, 1101±1112. Wang, E.A., Rosen, V., Cordes, P., Hewick, R.M., Kriz, M.J., Luxenberg, D.P., Sibley, B.S., Wozney, J.M., 1988. Puri®cation and characterization of other distinct boneinducing factors. Proc. Natl. Acad. Sci. USA 85, 9484± 9488. Wergedal, J.E., Baylink, D.J., 1984. Characterization of cells isolated and cultured from human bone. Proc. Soc. Exp. Biol. Med. 176, 60±69. Wergedal, J.E., Matsuyama, T., Strong, D.D., 1992. Dierentiation of normal human bone cells by transforming growth factor-b1 and 1,25 (OH)2 vitamin D3. Metabolism 41, 42±48. Whitehead, T.P., Kricka, L.J., Carter, T.J., Thorpe, G.H., 1979. Analytical luminescence: its potential in the clinical laboratory. Clin. Chem. 25, 1531±1546. Wiechelman, K., Braun, R., Fitzpatrick, J., 1988. Investigation of bicinchoninin acid protein assay: identi®cation of the groups responsible for color formation. Anal. Biochem. 175, 231±237. Wozney, J.M., Rosen, V., 1998. Bone morphogenetic protein and bone morphogenetic protein gene family in bone formation and repair. Clin. Orthopaedics 346, 26±37. Wozney, J.M., Rosen, V., Celeste, A.J., Mitsock, L.M., Whitters, M.J., Kriz, R.W., Hewick, R.M., Wang, E.A., 1988. Novel regulators of bone formation: molecular clones and activities. Science 242, 1528±1533.
www.elsevier.com/locate/archoralbio
Evidence for the involvement of bone morphogenetic protein-2 in phenytoin-stimulated osteocalcin secretion in human bone cells H. Koyama a, O. Nakade a,*, T. Saitoh b, T. Takuma c, T. Kaku a a
Department of Oral Pathology, School of Dentistry, Health Sciences University of Hokkaido, Ishikari-Tobetsu, 061-0293, Japan Department of Operative Dentistry and Endodontology, School of Dentistry, Health Sciences University of Hokkaido, IshikariTobetsu, 061-0293, Japan c Department of Oral Biochemistry, School of Dentistry, Health Sciences University of Hokkaido, Ishikari-Tobetsu, 061-0293, Japan b
Accepted 14 March 2000
Abstract Recent work has shown that the actions of phenytoin on bone cell proliferation and dierentiation are, in part, mediated through the upregulation of transforming growth factor-b1 (TGF-b1). The present study was undertaken to examine the eect of phenytoin on bone morphogenetic proteins (BMP)-2 and -4, which are well-recognized osteoinductive proteins of the TGF-b superfamily, in osteoblastic cells. Treatment with 5±50 mM of phenytoin increased the amount of mRNA for BMP-2 after a 0.5±24 h incubation in normal human mandible-derived bone cells (HOB-M cells), but failed to aect the mRNA for BMP-4. Phenytoin treatment for 48 h signi®cantly increased the secretion of BMP-2 by approx. four-fold, at an optimal concentration of 10 mM. While TGF-b1 inhibited osteocalcin secretion from HOB-M cells, both phenytoin and BMP-2 signi®cantly stimulated it. Importantly, the stimulatory eects of phenytoin on osteocalcin release were completely blocked by the neutralizing antihuman BMP2 monoclonal antibody. These results indicate that the stimulatory action of phenytoin on osteocalcin secretion in normal human bone cells is mediated, at least partly, through the upregulation of BMP-2, rather than that of TGFb1. 7 2000 Elsevier Science Ltd. All rights reserved. Keywords: Phenytoin; Bone morphogenetic proteins (BMP); Normal human bone cells; Osteocalcin secretion; Transforming growth factor-b (TGF-b)
Abbreviations: ANOVA, analysis of variance; BMP, bone morphogenetic protein; BSA, bovine serum albumin; DMEM, Dulbecco's modi®ed Eagle's medium; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; 1,25(OH)2D3; 1,25-dihydroxyvitamin D3; PICP, procollagen type I c-peptide; RT± PCR, reverse transcription±polymerase chain reaction; TBS, tris-buered saline; TGF-b, transforming growth factor-b. * Corresponding author. Tel.: +81-1332-3-1390; fax: +811332-3-1390. E-mail address: [email protected] (O. Nakade).
1. Introduction Phenytoin is a widely used therapeutic agent for seizure disorders. It is eective but has serious sideeects on bone and calcium metabolism (Richens and Rowe, 1970; Hahn and Avioli, 1975). Of the known side-eects, two are relevant to oral biology. First, many patients treated with phenytoin develop gingival hyperplasia, which is caused by phenytoin-
0003-9969/00/$ - see front matter 7 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 0 0 3 - 9 9 6 9 ( 0 0 ) 0 0 0 3 6 - 4
648
H. Koyama et al. / Archives of Oral Biology 45 (2000) 647±655
induced increases in the proliferation of gingival ®broblasts and collagen synthesis (Ebadi and Scott, 1971). Secondly, chronic phenytoin therapy increases the thickness and density of maxillary and calvarial bones (Kattan, 1970). Patients on phenytoin therapy show signi®cantly less alveolar bone loss than untreated controls (Seymour et al., 1985). Phenytoin treatment also accelerates the healing of mandibular bone fractures in the rabbit (Sklans et al., 1967). Lau and co-workers recently showed that phenytoin treatment increases bone cell proliferation, dierentiation, and collagen synthesis in vitro (Nakade et al., 1995), and histomorphometric measures of bone formation in vivo (Ohta et al., 1995a,b), and that it also increases serum biochemical markers of bone formation in patients (Lau et al., 1995). Together, these observations strongly indicate that phenytoin has anabolic actions, particularly on bone. Understanding the mechanism by which phenytoin stimulates bone formation is not only essential to our understanding of the pathogenesis of its skeletal side-eects, but also may allow us to develop a rationale basis in designing eective osteogenic drugs. Thus, we are interested in the molecular mechanism of the osteogenic actions of phenytoin. We have recently provided compelling evidence that its osteogenic actions are, in part, mediated through the upregulation of TGF-b1 synthesis. Accordingly, phenytoin increased the synthesis and release of active TGF-b1 in human bone cells and the stimulatory eects of phenytoin on bone cell proliferation and alkaline phosphatase activity were completely blocked by a neutralizing antibody against TGF-b1 (Nakade et al., 1996). On the other hand, while phenytoin stimulates osteocalcin secretion in bone cells (Nakade et al., 1995; Lau et al., 1995), TGF-b1 inhibits, rather than stimulates, osteocalcin secretion in human bone cells in vitro (Bonewald et al., 1992; Wergedal et al., 1992; Pirskanen et al., 1994). Thus, some of the osteogenic actions of phenytoin, such as the stimulation of osteocalcin secretion, may not be mediated through TGFb1. Because BMP-2 and -4, members of the TGF-b superfamily (Wozney et al., 1988), are known to stimulate osteocalcin secretion in the osteoblastic cell lineage (Hiraki et al., 1991; Rickard et al., 1994; Suzuki et al., 1995), we postulate that phenytoin stimulates the synthesis of BMP-2 and/or -4 in normal human osteoblastic cells, and that this is responsible for the phenytoininduced synthesis of osteocalcin. We have now sought to test this hypothesis by determining the eects of phenyotin on the synthesis of BMP-2 and -4, and by evaluating whether BMP-2 and/ or -4 is responsible for phenytoin-induced osteocalcin secretion in normal human osteoblasts.
2. Materials and methods 2.1. Materials Tissue-culture supplies were obtained from Iwaki Glass Co. (Funabashi, Japan) or Falcon (Oxnard, CA, USA). DMEM, 0.5% trypsin±5.3 mM EDTA solution, and Trizol1, a reagent for the single-step isolation of total RNA, were from GIBCO/BRL Life Technologies (Grand Island, NY, USA). Iron-supplemented bovine calf serum was purchased from JRH Biosciences (Lenexa, KS, USA). BSA and Triton X-100 were from Sigma Chemical Co. (St. Louis, MO, USA). 1,25(OH)2D3 was obtained from Biomol Research (Plymouth Meeting, PA, USA), and a Gla-type human osteocalcin enzyme immunoassay kit was purchased from Takara (Ohtsu, Japan). Sense and antisense primers for GAPDH and all chemicals for RT-PCR, unless otherwise stated, were purchased from Clontech Laboratories Inc. (Palo Alto, CA, USA). Recombinant human TGF-b1 was from R&D System (Minneapolis, MN, USA). Recombinant human BMP-2 and monoclonal antibody speci®c for mouse anti-human BMP-2 antibody were generous gifts of the Genetic Institute (Boston, MA, USA). Non-immune mouse IgG was obtained from Seikagaku Kogyo Co. (Tokyo, Japan). Horseradish peroxidase-conjugated anti-mouse IgG antibody and dot-blot equipment (Bio Dot1) were purchased from Cosmo Bio (Tokyo, Japan). Enhanced chemiluminescence Western-blotting detection reagents and nitrocellulose membrane were products of Amersham Life Science (Buckinghamshire, England). All other chemicals were of reagent grade, either from Sigma or from Kanto Chemical Co. (Tokyo, Japan). 2.2. Cell cultures Normal human bone cells were isolated from mandibular samples by collagenase digestion as previously described (Wergedal et al., 1984). The resulting bone cells (HOB-M) were shown to be of osteoblastic nature, based on their responsiveness to parathyroid hormone for increased cAMP production and to 1,25(OH)2D3 for alkaline phosphatase activity and osteocalcin secretion (Wergedal and Baylink, 1984; Nakade et al., 1995). The cells were maintained in DMEM supplemented with 10% bovine calf serum and passaged every week at a 1:4 dilution ratio. Cells from passages 3±7 were examined in this study. Phenytoin stocks were dissolved in 95% ethanol and diluted in DMEM containing 0.01% BSA immediately before use. The ®nal ethanol concentration in each assay was 0.35%. We realize that this concentration of ethanol is relatively high. There is evidence that high concentrations of ethanol could have an adverse eect on bone cell proliferation and dierentiation (Farley et
H. Koyama et al. / Archives of Oral Biology 45 (2000) 647±655
al., 1985). Unfortunately, due to the relatively low solubility of phenytoin even in ethanol, we were unable to reduce the ®nal concentration further. However, we should emphasize that the solvent vehicle controls in this study also contained 0.35% ethanol, and that we were able to detect osteogenic eects of phenytoin in spite of this high ethanol concentration (Nakade et al., 1995). 2.3. Measurement of BMP-2 and -4 mRNA BMP-2 and -4 mRNA were assessed with a semiquantitative RT-PCR. In brief, total RNA from control and phenytoin-treated osteoblasts was isolated using Trizol1 and quanti®ed by spectrophotometry (Chomczynski and Sacchi, 1987). Reverse transcription was carried out in a 20 ml volume containing 1.0 mg total RNA, 2.5 U RNAsin, 5 mM MgCl2, 50 mM KCl, 10 mM Tris (pH 8.3), 2.0 M oligo (dT) 11±18 primer, and 10 U of reverse transcriptase. The samples were incubated at 378C for 60 min. Reverse transcriptase was then heat-inactivated by incubating at 958C for 7 min. For each ampli®cation reaction (PCR), 2 ml cDNA (equivalent to 100 ng total RNA) was added to a 100 ml reaction mixture, consisting of 10 mM Tris± HCl, 50 mM KCl, 0.2% Triton X-100, 1.5 mM MgCl2, and 0.2 mM each of dATP, dCTP, dGTP, and dTTP. Appropriate sense and antisense primers for BMP-2 and BMP-4, respectively, and 0.2 U of Thermophime plus DNA polymerase were then added to the reaction mixture to initiate the PCR reaction. Sense and antisense BMP-2 and BMP-4 primers (Ogose et al., 1996) were synthesized by Takara Ltd. (BMP-2 forward primer: 5 '-GCTGTACTAGCGACACCCCAC-3 '; BMP-2 reverse primer: 5'-TCATAAAACCCTGCAACAGCCAACTCG-3 '; BMP-4 forward primer: 5 '-ACTGGTCCACCACAATGTGACACG-3; BMP-4 reverse primer: 5 '-GCTGAAGTCCACATAGAGCGAGTG-3. Ampli®cation was for 25 cycles for BMP-4 and 30 cycles for BMP-2. Each ampli®cation cycle consisted of denaturation at 948C for 45 sec, primer annealing at 608C for 45 sec, and primer extension at 728C for 2 min. For normalization of the RNA loading, RT-PCR of GADPH was also included in each RT-PCR reaction as an internal control. [Sense and antisense primers for GAPDH: forward primer: 5 '-TGAAGGTCGGAGTCAACGGATTTGGT-3 '; reverse primer: 5 '-CATGTGGGCCATGAGGT CCACCAC-3 ']. The ampli®cation products were electrophoresed in 2% agarose gel, stained with ethidium bromide, illuminated with ultraviolet light, photographed, and analysed in a computerized FMBIO1 ¯uorescent image-analyser (Hitachi Software Engineering Co., Ltd., Yokohama, Japan). The relative intensity of the ¯uorescence of the BMP-2- and BMP4-speci®c bands was expressed as percentage of the
649
intensity of the GAPDH band (Bodamyali et al., 1998). 2.4. Dot-blot immunoassay of BMP-2 The amount of BMP-2 protein released into the conditioned medium of HOB-M cells was estimated by a semiquantitative dot-blot immunoassay (Matikainen et al., 1999). In brief, HOB-M cells were treated with the indicated concentrations of phenytoin or the solvent vehicle for 48 h. The conditioned medium of each treatment group was then separately collected and concentrated by SpeedVac vacuum centrifugation. The ®nal volume of the concentrated medium was adjusted with TBS (pH 7.5) to normalize for cellular protein content so that it would yield the same volume per cellular protein content in each well. This was necessary to account for dierences in the cell number, as phenytoin stimulates the proliferation of human bone cells. As a positive control, recombinant human BMP-2 (®nal concentration 30 ng/ml) in DMEM containing 0.01% BSA was included in each assay. The adjusted medium was serially diluted with TBS (pH 7.5), and the samples (500 ml/well) were blotted by vacuum on a nitrocellulose paper wetted with TBS. After blotting, each well was blocked with 1% BSA in 0.05% Tween 20-containing TBS (TTBS) for 30 min. Mouse antihuman BMP-2 monoclonal antibody (1:400 dilution) was used as the primary antibody, and horseradish peroxidase-conjugated anti-mouse IgG antibody raised in sheep was used as the second antibody. All steps were performed inTTBS. Enhanced chemiluminescence Western-blotting detection reagents were used for visualization of the dots (Whitehead et al., 1979; Matikainen et al., 1999). This dot-blot assay, albeit semiquantitative, allowed us to estimate and compare the relative amount of BMP-2 in each test conditioned medium. 2.5. Measurement of osteocalcin secretion Osteocalcin secretion in the conditioned medium was assayed using a speci®c Gla-type human osteocalcin enzyme immunoassay kit (Nakade et al., 1999). In brief, HOB-M cells were plated at a density of 30,000 cells/well in DMEM supplemented with 10% bovine calf serum in 48-well culture plates. After plating for 24 h, the culture medium was changed to fresh serumfree DMEM, supplemented with 0.01% BSA and 10ÿ8 M 1,25(OH)2D3 as previously described (Nakade et al., 1995). Twenty-four hours later, eectors (BMP-2, TGF-b1, or phenytoin) or solvent control were added, and the cells incubated for an additional 48 h. The amount of osteocalcin in the conditioned medium was determined by enzyme immunoassay (Wiechelman et al., 1988) and normalized against cellular protein con-
650
H. Koyama et al. / Archives of Oral Biology 45 (2000) 647±655
tent to adjust for the cell number. For the anti-BMP-2 antibody blocking studies, 1 h after changing the culture medium to fresh serum-free medium, either the neutralizing anti-human BMP-2 speci®c monoclonal antibody (2 mg/ml) or non-immune normal mouse IgG (2 mg/ml) were added to the culture well (Hatakeyama et al., 1993). Phenytoin (5±50 mM) or solvent vehicle were added to the cells 0.5 h later. The cells were then incubated for an additional 48 h, and the amount of osteocalcin secretion was determined as described above.
2.6. Statistical methods
Results are shown as mean 2SEM (n = 6). The statistical signi®cance of the dierence from the solvent vehicle controls was analysed by two-tailed Student ttest. The statistical signi®cance of the dierence between the groups was determined by one-way ANOVA. The dierence was considered signi®cant when p < 0.05.
Fig. 1. RT-PCR analyses of the eects of short-term (0.5 h) exposure to phenytoin on the BMP-2 and -4 mRNA. (A) Agarose gel electrophoresis of BMP-2 and -4 mRNA ampli®ed by RT-PCR; (B) mRNA of BMP-2 (a) and BMP-4 (b) calculated as a percentage of endogenous GAPDH mRNA and expressed as a relative value. The data are shown as a percentage of the corresponding control (mean2SEM, n = 4 for each group).
H. Koyama et al. / Archives of Oral Biology 45 (2000) 647±655
3. Results 3.1. Eect of phenytoin on mRNA levels of BMP-2 and -4 As both BMP-2 and -4 are well-established growth factors that stimulate osteocalcin secretion in the osteoblastic cell lineage, we evaluated the eects of phenytoin on the mRNA of BMP-2 and -4. Fig. 1 shows that exposure to osteogenic doses of phenytoin rapidly (within 0.5 h) increased the BMP-2 mRNA (shown as percentage of GAPDH transcript) with an optimal concentration between 10 and 50 mM ( p < 0.001, ANOVA). The stimulation was seen even after
651
the 12-h incubation (Fig. 2; p < 0.001, ANOVA) and was still evident after a 24-h incubation (data not shown). This indicates that phenyotin treatment increased the mRNA of BMP-2 and that the stimulatory eect of phenytoin on BMP-2 mRNA was sustained at least for 24 h in normal human bone cells. In contrast, the same osteogenic doses of phenytoin failed to aect the amount of BMP-4 mRNA in either shortterm (0.5 h) or long-term (12±24 h) treatments (data not shown). 3.2. Eect of phenytoin on the BMP-2 secretion We next tested whether phenytoin increased the se-
Fig. 2. RT-PCR analyses of the eects of long-term (12 h) exposure of phenytoin on the BMP-2 and -4 mRNA. (A) Agarose gel electrophoresis of BMP-2 and -4 mRNA ampli®ed by RT-PCR; (B) mRNA of BMP-2 (a) and BMP-4 (b) calculated as a percentage of endogenous GAPDH mRNA and expressed as a relative value of each mRNA. Data are shown as a percentage of the corresponding control (mean2SEM, n = 4 for each group).
652
H. Koyama et al. / Archives of Oral Biology 45 (2000) 647±655
cretion of BMP-2 protein from human HOB-M into conditioned medium, with a semiquantitative dot-blot assay using a monoclonal antibody speci®c for human BMP-2. Fig. 3 shows that the osteogenic doses of phenytoin markedly increased the release of BMP-2 into the conditioned medium by approx. four-fold, with an optimal phenytoin concentration of 10 mM. 3.3. Eect of BMP-2 on osteocalcin secretion Because phenytoin treatment increases secretion of osteocalcin by human bone cells (Nakade et al., 1995; Lau et al., 1995), and because TGF-b1, which, we believe, mediates the stimulatory actions of phenytoin on bone cell proliferation and alkaline phosphatase activity (Nakade et al., 1996), is known to inhibit osteocalcin secretion by human bone cells, we next examined the eect of recombinant human BMP-2 on osteocalcin secretion in HOB-M cells and compared that to the eects of phenytoin and TGF-b1. Fig. 4 shows that both BMP-2 and phenytoin signi®cantly stimulated osteocalcin secretion to a similar concentration ( p < 0.01 and p < 0.05, ANOVA, respectively). Conversely, the osteogenic dose of TGF-b1 markedly inhibited its secretion by HOB-M ( p < 0.001, ANOVA).
Fig. 3. Dot-blotting analyses of the release of BMP-2 into medium induced by phenytoin in HOB-M cells (48-h treatment). Recombinant human BMP-2 (®nal concentration 30 ng/ml) in DMEM containing 0.01% BSA was included as a positive control in each assay. (Note: the intensity of BMP-2 immunoreaction from the 4-diluted medium with 10 mM phenytoin treatment is as strong as that from the control medium (1), indicating that phenytoin enhanced the BMP-2 release by approx. four-fold; this is only an estimate.)
3.4. Eect of anti-BMP-2 antibody on phenytoinmediated increase in osteocalcin secretion We reasoned that, if the stimulatory action of phenytoin on osteocalcin secretion was mediated through the local bone-cell secretion of BMP-2, the stimulatory eect of phenytoin would be completely blocked by a neutralizing antibody against BMP-2. Thus, ®nally we examined whether the stimulatory eects of phenytoin on osteocalcin secretion were abolished by anti-BMP-2 antibody. Fig. 5 shows that the treatment of HOB-M cells with 5±10 mM of phenytoin in the presence of 2 mg/ml of non-immune normal mouse IgG signi®cantly stimulated osteocalcin secretion ( p < 0.05, ANOVA). However, the addition of 2 mg/ml of antiBMP-2 antibody completely abolished the stimulatory eects of phenytoin. 4. Discussion There is circumstantial evidence that at least some of the anabolic actions of phenytoin may be mediated through the upregulation of growth-factor genes in target cells and tissues. It has been suggested that phenytoin increases the synthesis and release of epithelial growth factor in gingival ®broblasts (Brown et al., 1991) and platelet-derived growth factor-B in macrophages and monocytes (Dill et al., 1993). The release of these growth factors has been suggested to be responsible for the gingival hyperplasia. However, we believe we were the ®rst to provide conclusive evidence that phenytoin acts through local upregulation of growth-factor genes to exert its anabolic eects, by showing that the stimulatory actions of phenytoin on bone cell proliferation and alkaline phosphatase activity were mediated through upregulation of TGF-b1 expression (Nakade et al., 1996). The present study, which shows that osteogenic doses of phenytoin also upregulate BMP-2 gene expression in human bone cells, is further consistent with the premise that phenytoin exerts its anabolic actions through upregulation of growth-factor genes. Although we have compelling evidence that the bone-cell proliferative eect of phenytoin is mediated through the upregulation of TGF-b1 (Nakade et al., 1996), the signi®cance of the phenytoin-mediated upregulation of BMP-2 gene expression in human bone cells cannot be underestimated. Accordingly, this study provides three pieces of strong evidence that the phenytoin-mediated increase in osteocalcin secretion is mediated through the upregulation of the expression of the BMP-2 gene. First, phenytoin markedly stimulated the synthesis and release of BMP-2. Second, phenytoin and BMP-2 each signi®cantly increased osteocalcin secretion from HOB-M to similar concentrations, while
H. Koyama et al. / Archives of Oral Biology 45 (2000) 647±655
653
Fig. 4. Comparison among the eects of BMP-2, TGF-b1, or phenytoin on osteocalcin secretion from normal human mandible-derived bone cells (HOB-M cells). HOB-M cells were exposed to BMP-2, TGF-b1, phenytoin, or vehicle control for 48 h.
TGF-b1 signi®cantly reduced its secretion by the same cells. Third, and more importantly, the stimulatory eects of phenytoin on osteocalcin secretion were completely blocked by a neutralizing antibody against BMP-2. Osteocalcin is an osteoblast-speci®c protein and a well-recognized dierentiation marker of mature osteoblasts (Lian and Gundberg, 1988). Therefore, ®ndings of this and our past studies lead us to speculate that phenytoin may stimulate dierent aspects of bone formation, i.e., osteoblast proliferation as compared to osteoblast maturation (e.g. osteocalcin secretion) through upregulation of dierent bone growth-factor genes (i.e. TGF-b1 and BMP-2, respectively). In addition, recent studies suggest that human osteoblasts may possess a speci®c receptor for osteocalcin (Bodine
Fig. 5. Eects of antihuman BMP-2 antibody on osteocalcin secretion stimulated by phenytoin. HOB-M cells were pretreated with anti-human BMP-2 monoclonal antibody (2 mg/ ml) (solid bars) or with non-immune mouse IgG (2 mg/ml) (open bars) for 0.5 h before the addition of 0, 5, 10, or 50 mM phenytoin for 48 h.
and Komm, 1999) and that osteocalcin may play an important part in the regulation of bone formation and resorption (Ducy et al., 1996; Boskey et al., 1998). Therefore, it is possible that the phenytoin-induced osteocalcin secretion through upregulation of BMP-2 could have additional biological functions other than being a marker of osteoblast maturation. Bone morphogenetic proteins were originally identi®ed as compounds that induce bone and cartilage formation in ectopic extraskeletal sites (e.g. muscles) in vivo Wozney et al., 1988; Wang et al., 1988). Extensive studies have demonstrated that bone morphogenetic proteins, including BMP-2, are potent bone cell-dierentiating factors as well as bone-formation stimulators (Wozney et al., 1988; Harris et al., 1995). Therefore, upregulation of BMP-2 expression in response to phenytoin could have more important biological consequences than just to stimulate osteocalcin secretion. More speci®cally, there is evidence that phenytoin in the rabbit accelerates the healing of mandibular bone fractures (Sklans et al., 1967) and promotes wound healing in both hard and soft tissues (Swann et al., 1975; Frymoyer, 1976; Anstead et al., 1996; DaCosta et al., 1998) as well as in the skin. Because both TGFb1 and BMP-2 are essential in bone-fracture healing (Si et al., 1997) and the healing of bone wounds in vivo (Toriumi et al., 1991; Bolander, 1992) as well as of soft tissues (Stelnicki et al., 1998), it is possible that enhanced wound healing in both hard and soft tissues by phenytoin is mediated through upregulation of the expression of these two growth factors and/or the interaction between them. Much additional work is needed to con®rm this hypothesis. Finally, one of the most intriguing ®ndings in this study, perhaps, is the ®nding that, while osteogenic doses of phenytoin markedly increased the expression of BMP-2 in HOB-M, the agent at the same doses had no eect on the expression of the BMP-4 gene. This
654
H. Koyama et al. / Archives of Oral Biology 45 (2000) 647±655
observation is puzzling, as among all bone morphogenetic proteins, BMP-2 and BMP-4 share the highest sequence similarity and many of the same biological functions (Wozney and Rosen, 1998). It may be speculated that BMP-2 and BMP-4 might have dierent functional roles in bone formation and/or wound healing. Understanding the reason why phenytoin upregulates the expression of BMP-2 but not BMP-4 might provide important insights into potential functional dierences between these two. In summary, we demonstrate, to the best of knowledge for the ®rst time, that osteogenic concentrations of phenytoin increase the synthesis and release of BMP-2 in human mandible-derived bone cells, and that the stimulatory eects of phenytoin on osteocalcin secretion are mediated, in part, by the upregulation of BMP-2.
Acknowledgements We wish to thank Dr K-HW. Lau, Loma Linda University, for helpful discussions.
References Anstead, G.M., Hart, L.M., Sunahara, J.F., Liter, M.E., 1996. Phenytoin in wound healing. Ann. Pharmacother. 30, 768±775. Bodamyali, T., Bhatt, B., Hughes, F.J., Winrow, V.R., Kanczler, J.M., Simon, B., Abbott, J., Blake, D., Stevens, C., 1998. Pulsed electromagnetic ®elds simultaneously induce osteogenesis and upregulate transcription of bone morphogenetic proteins 2 and 4 in rat osteoblasts in vitro. Biochem. Biophys. Res. Com. 250, 458±461. Bodine, P.V.N., Komm, B.S., 1999. Evidence that conditionally immortalized human osteoblasts express an osteocalcin receptor. Bone 25, 535±543. Bolander, M.E., 1992. Regulation of fracture repair by growth factors. Proc. Soc. Exp. Biol. Med. 200, 165±170. Bonewald, L.F., Kester, M.B., Schwarty, Z., Swain, L.D., Khare, A., Johnson, T.L., Leach, R.J., Boyan, B.D., 1992. Eects of combining transforming growth factor-b1 and 1,25-dihydroxyvitamin D3 on dierentiation of a human osteosarcoma (MG-63). J. Biol. Chem. 267, 8943±8949. Boskey, A.L., Gadaleta, S., Gundberg, C., Doty, S.B., Ducy, P., Karsenty, G., 1998. Fourier transform infrared microspectroscopic analysis of bones of osteocalcin-de®cient mice provides insight into the function of osteocalcin. Bone 23, 187±196. Brown, R.S., Beaver, W.T., Bottomley, W.K., 1991. On the mechanism of drug-induced gingival hyperplasia. J. Oral Pathol. Med. 20, 201±209. Chomczynski, P., Sacchi, N., 1987. Single-step method of RNA isolation by acid guanidium thiocyanate±phenol± chloroform extraction. Anal. Biochem. 162, 156±159. DaCosta, M.L., Regan, M.C., al Sader, M., Leader, M.,
Bouchier-Hayes, D., 1998. Diphenylhydantoin sodium promotes early and marked angiogenesis and results in increased collagen deposition and tensile strength in healing wounds. Surgery 123, 287±293. Dill, R.E., Miller, E.K., Weil, T., Lesley, S., Farmer, G.R., Iacopino, A.M., 1993. Phenytoin increases gene expression for platelet-derived growth factor B chain in macrophages and monocytes. J. Periodontol. 64, 169±173. Ducy, P., Desbois, C., Boyce, B., Pinero, G., Story, B., Dunstan, C., Smith E, Bonadio, J., Goldstein, S., Gundberg, C., Bradley, A., Karsenty, G., 1996. Increased bone formation in osteocalcin-de®cient mice. Nature 382, 448±452. Ebadi, M.S., Scott, P.M., 1971. Increase in collagen level by diphenylhydantoin as possible mechanism of drug-induced gingival hyperplasia. Clin. Toxicol. 4, 39±46. Farley, J.R., Fitzsimmons, R., Taylor, A.K., Jorch, U.M., Lau, K.-H.W., 1985. Direct eects of ethanol on bone resorption and formation in vitro. Arch. Biochem. Biophys. 238, 305±314. Frymoyer, J.W., 1976. Fracture healing in rats treated with diphenylhydantoin (dilantin). J. Trauma 16, 368±370. Hahn, T.J., Avioli, L.V., 1975. Anticonvulsant osteomalacia. Arch. Intern. Med. 135, 997±1000. Harris, S.E., Feng, J.Q., Harris, M.A., Ghosh-Choudhury, N., Dallas, J.M., Wozney, J.M., Mundy, G.R., 1995. Recombinant bone morphogenetic protein 2 accelerates bone cell dierentiation and stimulates BMP-2 mRNA expression and BMP-2 promoter activity in primary fetal rat calvarial osteoblast cultures. Mol. Cell. Dier. 3, 137±155. Hatakeyama, S., Ohara-Nemoto, Y., Yakumoto, S., Satoh, M., 1993. Expression of morphogenetic protein in human adenocarcinoma cell line. Biochem. Biophys. Res. Com. 190, 695±701. Hiraki, Y., Inoue, H., Shigeno, C., Sanma, Y., Bentz, H., Rosen, D.M., Asad, A., Suzuki, F., 1991. Bone morphogenetic proteins (BMP-2 and BMP-3) promote growth and expression of the dierentiated phenotype of rabbit chondrocytes and osteoblastic MC3T3-E1 cells. J. Bone Miner. Res. 6, 1373±1385. Kattan, K.R., 1970. Calvarial thickening after dilantin medication. Am. J. Roentgenol. Radium Ther. Nucl. Med. 110, 102±106. Lau, K.-H.W., Nakade, O., Baylink, D.J., 1995. Phenytoin increases markers of osteogenesis for the human species in vitro and in vivo. J. Clin. Endocrinol. Metabol. 80, 2347± 2353. Lian, J.B., Gundberg, C.M., 1988. Osteocalcin: biochemical considerations and clinical applications. Clin. Orthopaed. Rel. Res. 226, 267±291. Matikainen, T., Kakonen, S.S., Pettersson, K., Karp, M., Lovgren, T., Vaahanen, H.K., Hellman, J., 1999. Demonstration of the predominant urine osteocalcin fragments detectable by two-site immunoassays. J. Bone Miner. Res. 14, 431±438. Nakade, O., Baylink, D.J., Lau, K.-H.W., 1995. Phenytoin at micromolar concentrations is an osteogenic agent for human mandible-derived bone cells in vitro. J. Dent. Res. 74, 331±337. Nakade, O., Baylink, D.J., Lau, K.-H.W., 1996. Osteogenic
H. Koyama et al. / Archives of Oral Biology 45 (2000) 647±655 actions of phenytoin in human bone cells are mediated in part by TGF-b1. J. Bone Miner. Res. 11, 1880±1888. Nakade, O., Koyama, H., Ariji, H., Yajima, A., Kaku, T., 1999. Melatonin stimulates proliferation and type I collagen synthesis in human bone cells in vitro. J. Pineal. Res. 27, 106±110. Ogose, A., Motoyama, T., Hotta, T., Watanabe, H., 1996. Expression of bone morphogenetic proteins in human osteogenic and epithelial tumor cells. Pathol. Int. 46, 9±14. Ohta, T., Wergedal, J.E., Gruber, H.E., Baylink, D.J., Lau, K.-H.W., 1995a. Low dose phenytoin is an osteogenic agent in the rat. Calcif. Tissue Int. 56, 42±48. Ohta, T., Wergedal, J.E., Gruber, H.E., Baylink, D.J., Lau, K.-H.W., 1995b. Phenytoin and ¯uoride act in concert to stimulate bone formation and to increase bone volume in adult male rats. Calcif. Tissue Int. 56, 390±397. Pirskanen, A., Jaaskelainen, T., Maenpaa, P.H., 1994. Eect of transforming growth factor b1 on the regulation of osteocalcin synthesis in human MG-63 osteosarcoma cells. J. Bone Miner. Res. 9, 1635±1642. Richens, A., Rowe, D.J.F., 1970. Disturbance of calcium metabolism by anticonvulsant drugs. Br. Med. J. 4, 73±76. Rickard, D.J., Sullivan, T.A., Shenker, B.J., Leboy, P.S., Kahdan, I., 1994. Induction of rapid osteoblast dierentiation in rat bone marrow stromal cell cultures by dexamethasone and BMP-2. Dev. Biol. 161, 218±228. Seymour, R.A., Smith, D.G., Turnbull, D.N., 1985. The eects of phenytoin and sodium valproate on the periodontal health of adult epileptic patients. J. Clin. Periodontol. 12, 413±419. Si, X., Jin, Y., Yang, L., Tipoe, G.L., White, F.H., 1997. Expression of BMP-2 and TGF-b1 mRNA during healing of the rabbit mandible. Eur. J. Oral Sci. 105, 325±330. Sklans, S., Taylor, R.G., Shklar, G., 1967. Eect of diphenylhydantoin sodium on healing of experimentally induced fractures in rabbit mandibles. J. Oral Surg. 25, 310±319. Stelnicki, E.J., Longaker, M.T., Holmes, D., Vanderwall, K., Harrison, M.R., Largman, C., Homan, W.Y., 1998. Bone morphogenetic protein-2 induces scar formation and skin
655
maturation in the second trimester fetus. Plast. Reconstr. Surg. 101, 12±19. Suzuki, R., Kobayashi, M., Takiguchi, T., Yamaguchi, A., Nagumo, M., Hasegawa, K., 1995. Bone morphogenetic protein-2 enhances the expression of the osteoblast phenotypes in human periodontal ligament ®broblasts and osteoblasts. J. Bone Miner. Res. 10 (Suppl. 1), T216. Swann, W.P., Swenson, H.M., Shafer, W.G., 1975. Eects of dilantin on the repair of gingival wounds. J. Periodontol. 46, 302±305. Toriumi, D.M., Kotler, H.S., Luxenberg, D.P., Holtrop, M.E., Wang, E.A., 1991. Mandibular reconstruction with a recombinant bone-inducing factor. Arch. Otolaryngol. Head Neck Surg. 17, 1101±1112. Wang, E.A., Rosen, V., Cordes, P., Hewick, R.M., Kriz, M.J., Luxenberg, D.P., Sibley, B.S., Wozney, J.M., 1988. Puri®cation and characterization of other distinct boneinducing factors. Proc. Natl. Acad. Sci. USA 85, 9484± 9488. Wergedal, J.E., Baylink, D.J., 1984. Characterization of cells isolated and cultured from human bone. Proc. Soc. Exp. Biol. Med. 176, 60±69. Wergedal, J.E., Matsuyama, T., Strong, D.D., 1992. Dierentiation of normal human bone cells by transforming growth factor-b1 and 1,25 (OH)2 vitamin D3. Metabolism 41, 42±48. Whitehead, T.P., Kricka, L.J., Carter, T.J., Thorpe, G.H., 1979. Analytical luminescence: its potential in the clinical laboratory. Clin. Chem. 25, 1531±1546. Wiechelman, K., Braun, R., Fitzpatrick, J., 1988. Investigation of bicinchoninin acid protein assay: identi®cation of the groups responsible for color formation. Anal. Biochem. 175, 231±237. Wozney, J.M., Rosen, V., 1998. Bone morphogenetic protein and bone morphogenetic protein gene family in bone formation and repair. Clin. Orthopaedics 346, 26±37. Wozney, J.M., Rosen, V., Celeste, A.J., Mitsock, L.M., Whitters, M.J., Kriz, R.W., Hewick, R.M., Wang, E.A., 1988. Novel regulators of bone formation: molecular clones and activities. Science 242, 1528±1533.