Matrix Regulation of Skeletal Cell Apoptosis

Abstract

Previously, we noted that inorganic phosphate (Pi), a major component of bone extracellular matrix, induced osteoblast apoptosis (Meleti, Z., Shapiro, I. M., and Adams, C. S. (2000) Bone (NY) 27, 359–366). Since Ca2+ along with Pi is released from bone during the resorption process, we advanced the hypothesis that Ca2+ modulates Pi-mediated osteoblast apoptosis. To test this hypothesis, osteoblasts were incubated with both ions, and cell death was determined. We noted that a modest increase in the medium Ca2+ concentrations ([Ca2+]e) of 0.1–1 mm caused a profound and rapid enhancement in Pi-dependent death of cultured osteoblasts. An elevation in [Ca2+]e alone had no effect on osteoblast viability, whereas Ca2+ channel blockers failed to inhibit killing of ion pair-treated cells. These results indicated that Pi-mediated cell death is not dependent on a sustained increase in the cytosolic Ca2+ concentration. Terminal dUTP nick-end labeling analysis and measurement of caspase-3 activity of the ion pair-treated cells suggested that death was apoptotic. Apoptosis was confirmed using caspase-3 and endonuclease inhibitors. The mitochondrial membrane potential and cytosolic Ca2+ status of the treated cells were evaluated. After incubation with [Ca2+ ]e and Pi, a decrease in mitochondrial fluorescence was noted, suggesting that the ions decreased the mitochondrial transmembrane potential. Subsequent to the fall in mitochondrial membrane potential, there was a transient elevation in the cytosolic Ca2+ concentration. Results of the study suggest that the ion pair conspire at the level of the plasma membrane to induce intracellular changes that result in loss of mitochondrial function. The subsequent increase in the cytosolic Ca2+ concentration may trigger downstream events that transduce osteoblast apoptosis. Previously, we noted that inorganic phosphate (Pi), a major component of bone extracellular matrix, induced osteoblast apoptosis (Meleti, Z., Shapiro, I. M., and Adams, C. S. (2000) Bone (NY) 27, 359–366). Since Ca2+ along with Pi is released from bone during the resorption process, we advanced the hypothesis that Ca2+ modulates Pi-mediated osteoblast apoptosis. To test this hypothesis, osteoblasts were incubated with both ions, and cell death was determined. We noted that a modest increase in the medium Ca2+ concentrations ([Ca2+]e) of 0.1–1 mm caused a profound and rapid enhancement in Pi-dependent death of cultured osteoblasts. An elevation in [Ca2+]e alone had no effect on osteoblast viability, whereas Ca2+ channel blockers failed to inhibit killing of ion pair-treated cells. These results indicated that Pi-mediated cell death is not dependent on a sustained increase in the cytosolic Ca2+ concentration. Terminal dUTP nick-end labeling analysis and measurement of caspase-3 activity of the ion pair-treated cells suggested that death was apoptotic. Apoptosis was confirmed using caspase-3 and endonuclease inhibitors. The mitochondrial membrane potential and cytosolic Ca2+ status of the treated cells were evaluated. After incubation with [Ca2+ ]e and Pi, a decrease in mitochondrial fluorescence was noted, suggesting that the ions decreased the mitochondrial transmembrane potential. Subsequent to the fall in mitochondrial membrane potential, there was a transient elevation in the cytosolic Ca2+ concentration. Results of the study suggest that the ion pair conspire at the level of the plasma membrane to induce intracellular changes that result in loss of mitochondrial function. The subsequent increase in the cytosolic Ca2+ concentration may trigger downstream events that transduce osteoblast apoptosis. Dulbecco's modified eagle's medium phosphonoformic acid 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide terminal dUTP nick-end labeling Asp-Glu-Val-Asp-aldehyde aurintricarboxylic acid Bone adapts to mechanical and physiological stress by a unique form of tissue replacement contained within discrete structures defined as basic multicellular units (2Frost H.M. N. Eng. J. Med. 1973; 289: 864-865PubMed Google Scholar). Within each of these units, the actual process of bone removal is carried out by osteoclasts; replacement bone matrix is synthesized and mineralized by cells of stromal origin, osteoblasts. Examination of resorbing sites in developing skeletal tissues indicates that many of the cells are apoptotic (3Bronckers A.L.J.J. Goei W. Luo G. Karsenty G. D'Souza R.N. Lyaruu D.M. Burger E.H. J. Bone Miner. Res. 1996; 11: 1281-1291Crossref PubMed Scopus (136) Google Scholar). Thus, there is evidence of DNA fragmentation in osteoclasts, osteoblasts, and osteocytes (4Hughes D.E. Dai A. Tiffee J.C. Li H.H. Mundy G.R. Boyce B.F. Nat. Med. 1996; 2: 1132-1136Crossref PubMed Scopus (705) Google Scholar, 5Jilka R.L. Weinstein R.S. Bellido T. Parfitt A.M. Manolagas S.C. J. Bone Miner. Res. 1998; 13: 793-802Crossref PubMed Scopus (472) Google Scholar). In contrast, in mature skeletal tissues, only about 1–2% of all bone cells are dying or dead. In both the developing and mature skeleton, most of the apoptotic cells are confined to bone remodeling sites or locales of high bone turnover (5Jilka R.L. Weinstein R.S. Bellido T. Parfitt A.M. Manolagas S.C. J. Bone Miner. Res. 1998; 13: 793-802Crossref PubMed Scopus (472) Google Scholar, 6Jilka R.L. Weinstein R.S. Bellido T. Roberson P. Parfitt A.M. Manolagas S.C. J. Clin. Invest. 1999; 104: 439-446Crossref PubMed Scopus (896) Google Scholar, 7Noble B.S. Stevens H. Loveridge N. Reeve J. Bone. 1997; 20: 273-282Crossref PubMed Scopus (199) Google Scholar, 8Weinstein R.S. Jilka R.L. Parfitt A.M. Manolagas S.C. J. Clin. Invest. 1998; 102: 274-282Crossref PubMed Scopus (1428) Google Scholar). How osteoclasts communicate with and regulate the life history of other cells of the basic multicellular unit is a topic of intense debate. It is clear that osteoclast differentiation and activation are dependent on paracrine signals received from stromal cells in the multicellular unit (9Takahashi N. Akatsu T. Udagawa N. Sasaki T. Yamaguchi A. Moseley J.M. Martin T.J. Suda T.,. Endocrinology. 1988; 123: 2600-2602Crossref PubMed Scopus (856) Google Scholar). Within the past four years, chemical modulators of these processes have been identified, and recent evidence indicates that growth factors provide survival signals that result in bone cell proliferation and depression of the apoptotic process (5Jilka R.L. Weinstein R.S. Bellido T. Parfitt A.M. Manolagas S.C. J. Bone Miner. Res. 1998; 13: 793-802Crossref PubMed Scopus (472) Google Scholar, 6Jilka R.L. Weinstein R.S. Bellido T. Roberson P. Parfitt A.M. Manolagas S.C. J. Clin. Invest. 1999; 104: 439-446Crossref PubMed Scopus (896) Google Scholar, 10Hill P.A. Tumber A. Meikle M.C. Endocrinology. 1997; 138: 3849-3858Crossref PubMed Scopus (186) Google Scholar). In addition, it has been demonstrated that a number of pharmacological agents can induce osteoblast apoptosis in vitro (5Jilka R.L. Weinstein R.S. Bellido T. Parfitt A.M. Manolagas S.C. J. Bone Miner. Res. 1998; 13: 793-802Crossref PubMed Scopus (472) Google Scholar, 10Hill P.A. Tumber A. Meikle M.C. Endocrinology. 1997; 138: 3849-3858Crossref PubMed Scopus (186) Google Scholar, 11Kitajima I. Nakajima T. Imamura T. Takasaki I. Kawahara K. Okano T. Tokioka T. Soejima Y. Abeyama K. Maruyama I. J. Bone Miner. Res. 1996; 11: 200-210Crossref PubMed Scopus (64) Google Scholar, 12Ihbe A. Baumann G. Heinzmann U. Atkinson M.J. Calcif. Tissue Int. 1998; 63: 208-213Crossref PubMed Scopus (11) Google Scholar). Surprisingly, however, little is known of events that promote osteoblast death in situ. Although the resorption process may generate agents that stimulate osteoblast proliferation, it is probable that products of the resorbing bone may also stimulate bone cell death. Recent work from this laboratory has clearly demonstrated that one of the ions present in the bone matrix, inorganic phosphate (Pi), induces apoptosis of cultured human osteoblasts and chondrocytes (1Meleti Z. Shapiro I.M. Adams C.S. Bone ( NY ). 2000; 27: 359-366Crossref PubMed Scopus (210) Google Scholar, 13Mansfield K. Rajpurohit R. Shapiro I.M. J. Cell. Physiol. 1999; 179: 276-286Crossref PubMed Scopus (100) Google Scholar). Since Ca2+ as well as Pi are released from the bone apatite lattice during the resorption process, the possibility exists that Ca2+ may influence Pi-mediated bone cell apoptosis. To test the hypothesis that this ion pair may trigger the death program, we examine the effect of Pi and Ca2+ on human osteoblast-like cells. We ask the questions, Can Ca2+ modulate Pi-induced cell death, and, if so, is death mediated by apoptosis? Using a cell culture system, we demonstrate that Ca2+ accentuates the apoptogenic effect of Pi. In addition, we provide evidence for the involvement of mitochondria and intracellular Ca2+ in the apoptotic pathway activated by the ion pair. Specimens of human bone were obtained during dental surgery performed at the Hospital of the University of Pennsylvania and during spinal surgery performed at the Children's Hospital of Philadelphia (Philadelphia, PA). Ages of the samples ranged from 9 to 32 years. The bone was chopped into very small pieces using a pair of rongeurs. The pieces were then digested in 10 ml of bacterial collagenase (4.6 mg/ml) (Sigma) in Ca2+- and Mg+-free Hanks' balanced salt solution for 1 h at 37 °C in a shaker bath. The supernatant was discarded, and the bone fragments were placed into cell culture dishes (Corning Glass, Corning, NY) containing Dulbecco's modified eagle's medium (DMEM)1 (Life Technologies, Inc.) supplemented with 10% fetal bovine serum, 2 mml-glutamine, and antibiotics. The Piconcentration of this medium was 0.9 mm, and the Ca2+ concentration ([Ca2+]e) was 1.8 mm. The cultures were maintained at 37 °C in a sterile incubator, and the medium was changed every day. Cell migration from the bone fragments was monitored daily by light microscopy. After a period of 3–6 weeks, osteoblast-like cells grew out from the explant. When confluent, cells were released from the tissue culture dishes by a brief treatment with 0.25% trypsin and 0.1% bacterial collagenase (Sigma) in Hanks' balanced salt solution. Cells were collected and replated at a density of 140 cells/mm2. Secondary cultures were maintained in DMEM supplemented with 10% fetal bovine serum, 2 mml-glutamine, 5 mm β-glycerophosphate, and antibiotics. After 48 h, fresh ascorbate (10 μg/ml) was added to the medium. Cells were fed fresh ascorbate at every media change. The osteogenic characteristics of the cells were confirmed by reverse transcription-polymerase chain reaction using probes for Cbfa-1, osteocalcin, osteonectin, and type I collagen (as described in Meletiet al. (1Meleti Z. Shapiro I.M. Adams C.S. Bone ( NY ). 2000; 27: 359-366Crossref PubMed Scopus (210) Google Scholar)). We also examined the effect of the ion pair on MC-3T3-E1 cells, a cell line that, in culture, recapitulates each of the major steps in osteoblast maturation (14Franceschi R.T. Iyer B.S. J. Bone Miner. Res. 1992; 7: 235-246Crossref PubMed Scopus (480) Google Scholar). These cells were grown to confluence in DMEM supplemented with 10% fetal bovine serum, 2 mml-glutamine, and 25 μg/ml ascorbate as described above. After in culture, the osteogenic characteristics of the cells were confirmed by reverse transcription-polymerase chain reaction using probes for Cbfa-1, osteocalcin, osteonectin, and type I collagen was as a of the Pi concentration and [Ca2+]e as well as treatment added to the medium in the form of The of and mm phosphate in a medium Pi concentration of and mm. Ca2+ was added as The of and 1 mm Ca2+ in medium Ca2+ concentrations of and In a the effect of the phosphonoformic acid on ion cell death was also evaluated. In addition, we the Ca2+ channel blockers and modulate Ca2+- and Pi-dependent apoptosis. of these agents were from In each cell death was by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide The is on the of mitochondria in cells to a salt to Cells were treated with the agents for the indicated and then incubated with μg/ml) at 37 °C for The was and of in was added to each The density of the solution was at in Martin A. Y. Cell PubMed Scopus Google Scholar). Since the of the is to the mitochondrial the decrease in the at a measurement of cell death. Although we have demonstrated that Pi induced osteoblast apoptosis (1Meleti Z. Shapiro I.M. Adams C.S. Bone ( NY ). 2000; 27: 359-366Crossref PubMed Scopus (210) Google it was to the ion pair bone cells by apoptosis. were The of the that during DNA into fragments of of The cell cultures were treated with the ion and the were using a osteoblasts were as Cells were then treated with mg/ml) at for We have that the of treatment not of DNA K. Rajpurohit R. Shapiro I.M. J. Cell. Physiol. 1999; 179: 276-286Crossref PubMed Scopus (100) Google Scholar). activity was by cells to in Cells were in a for and then incubated in a reaction containing and the of DNA at 37 After the reaction was and the were to These were then using to To the cells were not is a downstream of the apoptotic in osteoblasts. To that the ion pair induced osteoblast we the activity of the in treated cells using a with the activated and the increase in fluorescence is to the in caspase-3 activity A. M.J. P.A. J. Med. 2000; PubMed Scopus Google Scholar). Cells were treated with the ion pair for from to each the were and incubated with 10 for 1 h at 37 was the cells were and the fluorescence was by the Apoptosis is dependent on activation of a number of of these apoptosis. To that the ion pair induces we the and downstream To osteoblasts were incubated with a caspase-3 for 2 h treatment with 5 mm Pi and After h, cell death was by the To endonuclease aurintricarboxylic acid (Sigma) was A.M. 1999; PubMed Scopus Google Scholar). Cells were incubated with for 2 h and then treated with 5 mm Pi and mm cell death was by the were treated with a number of Ca2+ channel blockers to if it was for [Ca2+]e to the cell to apoptosis. cells were incubated for 2 h with the and U. S. A. PubMed Scopus Google Scholar). the osteoblasts were treated with 5 and mm Ca2+ for a The was using the Ca2+ channel W. Cell Res. 1999; PubMed Scopus Google and a Ca2+ channel R.L. Calcif. Tissue Int. 2000; PubMed Scopus Google Scholar). In all cell death was using the of probes was to examine the of inorganic ions on the cytosolic Ca2+ status and mitochondrial membrane potential (as indicated in al. T. 1998; PubMed Scopus Google Scholar)). Cells were treated with 5 mm Pi and mm and h, as described the of each of these the medium was with DMEM containing 1 or 5 for medium was then and with fresh Cells were then with the NY) with a a was to of the cells the To the of fluorescence was and the was at that for the These were then to the fluorescence of the treated cells. was using the each of the of cells was from the were to well and as a of the The were using a analysis of the effect of treatment of each well on cell When the were by the of the The University Scholar). When that not a analysis of on was of cells from the cultures was treated as a The of was When treated with Ca2+ and Pi, osteoblast-like cells a decrease in It noted that in the of 1 mm Pi, elevation in the extracellular Ca2+ concentration ([Ca2+]e) not induce cell death. if [Ca2+]e is to mm, the medium Pi concentration is at 1 mm, there is no loss of 1 the medium Pi level is at mm, a small in [Ca2+]e a increase in cell death 1 this Pi osteoblast-like cell by concentrations of Pi and [Ca2+]e increase the of the cells to Pi these and of osteoblasts, also indicates that the of MC-3T3-E1 cells to the ion pair is not to that of osteoblast-like cells. Thus, in with osteoblast-like cells, MC-3T3-E1 cells a to both mm Pi and and 5 mm Pi and mm In the in 1 indicates that a modest elevation in [Ca2+]e Pi-mediated cell death in osteoblasts and in MC-3T3-E1 cells. The Ca2+ activation of Pi-mediated cell death is rapid 2 h, there is a loss of cell and by h, of the osteoblasts are dead. rapid loss of is also MC-3T3-E1 cells are treated with medium containing concentrations of [Ca2+]e and These results with in 1 that [Ca2+]e a sustained elevation in the of of death processes in osteoblasts. Results of of that the osteoblast by apoptosis treatment with Ca2+ and Pi. analysis that cells treated with mm Pi and mm that the osteoblast-like cells have from the matrix and These changes are of cells apoptosis. of the caspase-3 indicates that in the of the ion there is a in caspase-3 activity there is elevation in cells. Thus, by a cells are by 140 many of the cells are and cells a high level of of cells It is that the cell to cell in fluorescence is to the of caspase-3 was performed to inhibit of downstream death When the ion pair concentration is to 5 mm Pi and mm cell death in a the of caspase-3 activity apoptosis 5 concentrations the of cell In a of [Ca2+]e and Pi induce cell death, is and and by the of These results are all with the that the ion pair induces osteoblast of caspase-3 by treatment with Pi and were treated with 5 mm Pi and mm the indicated Cells were then treated with the caspase-3 10 for 1 h at 37 of the cells were then using the the increase in activation of a apoptosis of apoptosis on Pi and Ca2+- mediated osteoblast were with or for 2 Cells were then cultured for h in the of 5 mm Pi and Cell was by the concentrations endonuclease osteoblast death. of at concentrations cell of these are downstream of the apoptotic the and of the with Pi with cells treated with 5 mm mm When Pi is the ion pair to trigger cell death Thus, of the osteoblasts to mm Ca2+ and mm Pi. The of of Pi osteoblast apoptosis if the Pi concentration is to mm, cell death. also MC-3T3-E1 cells from apoptosis. that if the medium the cells treated with high of the ion pair Ca2+ and mm alone has no effect on bone cell Results of this of suggest that or of Pi by the treated cells is a of the apoptotic Although Pi are for cell death, of Ca2+ channel little effect on ion osteoblast apoptosis. with or no in Pi-mediated bone cell death. I that at all concentrations these Ca2+ channel blockers to increase the of cells. a Ca2+ channel to from ion apoptosis. a concentration of this the number of cells. high are to cells, it is probable that the increase in is to the apatite out of solution I. A. P. J. 1999; PubMed Scopus Google Scholar). of the the activity of the ion pair as well as the concentration of to of the Ca2+ also no from Ca2+- and Pi-induced apoptosis not results of these that of a number of has a effect in Ca2+ dependent Pi-mediated bone cell of channel blockers on and osteoblast from from from from from from from from from from from from of the was treated with 5 and mm Ca2+ along with the indicated concentration of Ca2+ channel for The number of cells was using the and failed to the cells from apoptosis. a Ca2+ channel only at high are the and from from in a of the was treated with 5 and mm Ca2+ along with the indicated concentration of Ca2+ channel for The number of cells was using the and failed to the cells from apoptosis. a Ca2+ channel only at high are the and indicates that mitochondria are in the activation of apoptosis and When with a membrane osteoblasts a level of mitochondrial fluorescence h treatment with 5 mm Pi and mm is a transient increase in fluorescence by h, the fluorescence has to the cells are cell to cell are and many of the cells are The fluorescence of these cells is to the to membrane The in mitochondrial fluorescence is in A. with the for 2 h a increase in fluorescence to by the also that to the cell a small increase in fluorescence at on the other is in the cells In to mitochondrial a in fluorescence is 2 h then to intracellular Ca2+ and a is h treatment and h, apoptosis in intracellular fluorescence is The in is in increase in fluorescence is and are h When in and are it is that Ca2+ are loss of mitochondrial of osteoblast-like cells treated with 5 mm Pi and and with Cells were treated with Pi and Ca2+ and then incubated with The in fluorescence a period was by microscopy. the were to the level of cytosolic Ca2+ fluorescence in these cells. by h a increase in fluorescence was of mitochondrial membrane and intracellular Ca2+ The fluorescence of cells in and were with the cells from each were defined and the was and mitochondrial intracellular that the of of the Ca2+ was loss of Since of the mitochondrial membrane potential, it is probable that loss of mitochondrial other the and of the with with the Using bone cell human osteoblasts and MC-3T3-E1 we demonstrated that Pi-dependent apoptosis and that the of the ion pair was profound that by Pi alone (1Meleti Z. Shapiro I.M. Adams C.S. Bone ( NY ). 2000; 27: 359-366Crossref PubMed Scopus (210) Google Scholar). The of the ions on cell to the by these apatite the death [Ca2+]e alone had no effect on osteoblast the ion pair was a of the death When Pi was there was from apoptosis a number of Ca2+ channel blockers failed to inhibit apoptosis the results suggest that Ca2+ may with Pi at the level of the cell membrane and that this intracellular Pi-induced osteoblast death. In of the in of these it is probable that at sites of bone there is a increase in both al. Cell Res. 1988; PubMed Scopus Google that in to osteoclasts, extracellular Ca2+ can this we suggest that, at resorption a in the ion pair concentration apoptosis in osteoblasts as well as other cells. this many of the osteoblasts present at sites apoptotic death (4Hughes D.E. Dai A. Tiffee J.C. Li H.H. Mundy G.R. Boyce B.F. Nat. Med. 1996; 2: 1132-1136Crossref PubMed Scopus (705) Google Scholar). Results of the and that of Pi caused and osteoblast apoptosis (1Meleti Z. Shapiro I.M. Adams C.S. Bone ( NY ). 2000; 27: 359-366Crossref PubMed Scopus (210) Google Scholar, 13Mansfield K. Rajpurohit R. Shapiro I.M. J. Cell. Physiol. 1999; 179: 276-286Crossref PubMed Scopus (100) Google Scholar). The that was on Ca2+ indicated that the concentration of this ion was from 1.8 to mm, there was a increase in Pi-mediated osteoblast death It noted that the in [Ca2+]e was The Ca2+ concentration was the level present in and the at sites Cell Res. 1988; PubMed Scopus Google Scholar). death was apoptosis was confirmed the the of the cells, and Thus, in the of the ion the cells characteristics of of the cells, of osteoblasts from the and loss of with cells as well as in the mitochondrial membrane potential. In addition, the activity of both caspase-3 and was there was of apoptosis. Since activation of these is a of it is that elevation in the Ca2+ and Pi concentration as a We the by Ca2+ Pi-mediated cell death by that into osteoblasts. was by the that at Pi elevation in [Ca2+ ]e not cell death. Thus, of this on osteoblast When the medium Pi was a in [Ca2+]e caused a rapid of killing The of Pi was by the that that and Pi into cells, ion osteoblast death for a for Ca2+ in osteoblast we the possibility that increase in [Ca2+]e the intracellular Ca2+ concentration and promote the death osteoblasts were treated with a of ion apoptosis was not Thus, cell death was dependent on the of both ions, apoptosis was not activated by a of [Ca2+]e into the treated cells. The channel indicated that bone cell apoptosis was mediated by to a in mitochondrial function. of other apoptotic that mitochondrial results in of Ca2+ from intracellular a in intracellular Ca2+ and loss of to downstream events in the apoptotic process A. G. Cell 1999; PubMed Scopus Google Scholar, T. G. J. 1999; PubMed Scopus Google Scholar). In bone cells, the mitochondrial was to the of mitochondria to the ion of the mitochondrial was h the had that of the cells In the there was elevation in intracellular Ca2+ fluorescence h, the cells clear evidence of both mitochondria and intracellular had decreased and Thus, the of the mitochondria was reaction to the there was a in Ca2+ in the treated cells. It is to that the increase in Ca2+ fluorescence was to the loss of mitochondrial membrane potential and the subsequent of the mitochondria and to intracellular Ca2+ When this is the transient in the cytosolic Ca2+ concentration may have to downstream of the apoptotic from the there are a number of other in [Ca2+]e Pi-mediated osteoblast apoptosis. [Ca2+]e and Pi may at the cell membrane and induce death as ion ion or The ions to sites on the membrane and trigger that the apoptotic if are at the plasma into the cell and trigger apoptosis In of is the that phosphate can W. 1997; PubMed Scopus Google Scholar, A. Calcif. Tissue Int. PubMed Scopus Google Scholar, Calcif. Tissue Int. PubMed Scopus Google Scholar). and A. J. 1998; PubMed Google out that in osteocytes to this recent that if the is then of can induce osteoblast apoptosis J. Res. 1997; PubMed Scopus Google Scholar, J. Res. 1999; PubMed Scopus (64) Google Scholar). the other the that osteoblast apoptosis to the that to induce Pi to the cell a plasma membrane of the the results that major products of the process as a