Nitrogen-Containing Biphosphonates Inhibit the Mevalonate Pathway and Prevent Post-Translational Prenylation of GTP-Binding Proteins, Including Ras

Authors

  • Steven P. Luckman,

    1. Department of Human Metabolism and Clinical Biochemistry, University of Sheffield, Sheffield, United Kingdom
    2. Department of Medicine and Therapeutics, University of Aberdeen, Aberdeen, United Kingdom
    Search for more papers by this author
  • David E. Hughes,

    1. Department of Pathology, University of Sheffield, Sheffield, United Kingdom
    Search for more papers by this author
  • Fraser P. Coxon,

    1. Department of Human Metabolism and Clinical Biochemistry, University of Sheffield, Sheffield, United Kingdom
    2. Department of Medicine and Therapeutics, University of Aberdeen, Aberdeen, United Kingdom
    Search for more papers by this author
  • R. Graham G. Russell,

    1. Department of Human Metabolism and Clinical Biochemistry, University of Sheffield, Sheffield, United Kingdom
    Search for more papers by this author
  • Dr. Michael J. Rogers

    Corresponding author
    1. Department of Medicine and Therapeutics, University of Aberdeen, Aberdeen, United Kingdom
    • Department of Medicine and Therapeutics University of Aberdeen Polwarth Building Foresterhill Aberdeen AB25 2ZD, U.K.
    Search for more papers by this author

Abstract

Bisphosphonates are currently the most important class of antiresorptive drugs used for the treatment of metabolic bone diseases. Although the molecular targets of bisphosphonates have not been identified, these compounds inhibit bone resorption by mechanisms that can lead to osteoclast apoptosis. Bisphosphonates also induce apoptosis in mouse J774 macrophages in vitro, probably by the same mechanisms that lead to osteoclast apoptosis. We have found that, in J774 macrophages, nitrogen-containing bisphosphonates (such as alendronate, ibandronate, and risedronate) inhibit post-translational modification (prenylation) of proteins, including the GTP-binding protein Ras, with farnesyl or geranylgeranyl isoprenoid groups. Clodronate did not inhibit protein prenylation. Mevastatin, an inhibitor of 3-hydroxy-3-methylglutatyl (HMG)-CoA reductase and hence the bio-synthetic pathway required for the production of farnesyl pyrophosphate and geranylgeranyl pyrophosphate, also caused apoptosis in J774 macrophages and murine osteoclasts in vitro. Furthermore, alendronate-induced apoptosis, like mevastatin-induced apoptosis, could be suppressed in J774 cells by the addition of farnesyl pyrophosphate or geranylgeranyl pyrophosphate, while the effect of alendronate on osteoclast number and bone resorption in murine calvariae in vitro could be overcome by the addition of mevalonic acid. These observations suggest that nitrogen-containing bisphosphonate drugs cause apoptosis following inhibition of post-translational prenylation of proteins such as Ras. It is likely that these potent antiresorptive bisphosphonates also inhibit bone resorption by preventing protein prenylation in osteoclasts and that enzymes of the mevalonate pathway or prenyl protein transferases are the molecular targets of the nitrogen-containing bisphosphonates. Furthermore, the data support the view that clodronate acts by a different mechanism.

INTRODUCTION

Bisphosphonates are a class of nonhydrolyzable analogs of pyrophosphate that have high affinity for bone mineral and the ability to inhibit osteoclast-mediated bone resorption.(1) Several bisphosphonates have become the drugs of choice for the treatment of bone diseases that involve excessive bone resorption, including osteoporosis(2) metastatic bone disease,(1) and Paget's disease.(3) Although bisphosphonates have been used as antiresorptive drugs for three decades, their exact molecular mechanisms of action have not been identified. One of the first bisphosphonates to be developed for clinical use, clodronate, can be metabolized to a cytotoxic, nonhydrolyzable analog of ATP by mammalian cells.(4) However, the more potent nitrogen-containing bisphosphonates, such as pamidronate, alendronate, ibandronate, and risedronate, are not metabolized(4) and probably act by a different mechanism that can lead to osteoclast apoptosis(5) These bisphosphonates also cause apoptosis of J774 macrophages in vitro(6) Furthermore, bisphosphonate-induced apoptosis in osteoclasts and J774 macrophages probably occurs by a similar mechanism, since the structure-activity relationships of bisphosphonates for inhibition of bone resorption match those for induction of macrophage apoptosis (S.P. Luckman et al., manuscript in preparation).(7,8) Given the considerable difficulty in isolating large numbers of pure osteoclasts for biochemical studies, we are using J774 macrophages as a model with which to identify potential molecular targets for bisphosphonates and hence mechanisms by which bisphosphonates cause apoptosis.

Two potent nitrogen-containing bisphosphonates that cause apoptosis in J774 cells, YM175 and ibandronate,(6) have recently been reported to be inhibitors of sterol biosynthesis in J774 cells.(9,10) Furthermore, the concentrations of the bisphosphonates that inhibit sterol biosynthesis in these cells are similar to the concentrations of bisphosphonates that cause J774 apoptosis.(6,9,10) It is likely that YM175 and ibandronate inhibit sterol biosynthesis by inhibiting squalene synthase, an enzyme in the mevalonate pathway.(9,10) However, other bisphosphonates (including alendronate and pamidronate) that are less potent inhibitors of squalene synthase can also inhibit sterol biosynthesis in a cell-free system,(9) suggesting that these bisphosphonates may inhibit enzymes of the mevalonate pathway other than squalene synthase.

The pathway of sterol biosynthesis from mevalonate includes the synthesis by prenyl transferases of isoprenyl-pyro-phosphate intermediates, such as farnesyl pyrophosphate (FPP) and geranylgeranyl pyrophosphate (GGPP).(11) These isoprenyl groups can be transferred to a cysteine residue within carboxy-terminal motifs present in several classes of proteins, including the family of GTP-binding Ras, Rho, Rac, and Rab proteins and nuclear lamins,(12–15) in a reaction catalyzed by at least three distinct cytoplasmic prenyl protein transferases.(15) Post-translational modification of proteins with C15 farnesyl or C20 geranylgeranyl groups appears to be essential for the localization of these proteins to membranes and hence for their biological function.(15–17) Inhibition of protein prenylation by substrate inhibitors of prenyl protein transferases or by inhibitors of mevalonate or isopentenyl pyrophosphate synthesis (such as lovastatin, mevastatin,(18) and phenylacetate(19)) has a profound effect on cell morphology(20) cell replication,(21,22) and intracellular signal transduction(23) and can lead to induction of apoptotic cell death.(19,24) We therefore investigated whether mevastatin, like bisphosphonates, inhibits bone resorption and causes apoptosis in osteoclasts and J774 macrophages. In addition, we examined whether apoptosis induced by mevastatin or bisphosphonates is associated with the inhibition of post-translational prenylation of proteins such as Ras in J774 cells and whether the effects of bisphosphonates on apoptosis and bone resorption could be suppressed by the addition of components of the mevalonate pathway such as mevalonic acid, FPP, and GGPP.

MATERIALS AND METHODS

Chemicals

Clodronate, alendronate, ibandronate, YM175, and risedronate were provided by Proctor and Gamble Pharmaceuticals (Cincinnati, OH, U.S.A.). The bisphosphonates were dissolved in phosphate-buffered saline (PBS), the pH was adjusted to 7.4 with 1 N NaOH, then filter-sterilized by using a 0.2 μm filter. Mevastatin (also known as compactin) was purchased from Sigma Chemical Co. (Poole, U.K.), and converted from the lactone form by dissolving 5 mg of mevastatin in 100 μl ethanol and 100 μl of 1 N NaOH. After the addition of 1 ml of PBS, the pH of the solution was adjusted to approximately pH 8 using 1 N HCl before filter sterilization. A stock solution of 10 mM mevalonic acid lactone was prepared by dissolving the solid in dry ethanol. FPP and GGPP, purchased from Sigma, were dried to remove solvent then resuspended in culture medium immediately before use. [35S]methionine and [14C]mevalonolactone were from Amersham (Aylesbury, U.K.). All other chemicals were from Sigma Chemical Co. unless stated otherwise.

Cell culture

J774.2 cells were obtained from the European Collection of Animal Cell Cultures (Porton Down, U.K.). Cultures were grown at 37°C in Dulbecco's modified Eagle's medium (GIBCO, Paisley, U.K.) containing 10% heat-inactivated fetal calf serum, 100 U/ml penicillin, 100 μg/ml streptomycin, and 1 mM glutamine in 5% CO2 atmosphere.

Assessment of osteoclast number and bone resorption in isolated murine calvariae

Osteoclast number and bone resorption were measured in isolated neonatal murine calvariae by vital staining using nitroblue tetrazolium (NBT), according to the method described by Garrett et al.(25) Osteoclasts reduce NBT to insoluble blue formazan with greater efficiency than other cell types present in bone or bone marrow. This, combined with their larger size and multinuclearity, allows them to be identified and counted in fixed whole calvariae when examined at low power by phase contrast microscopy.(25) This approach also allows identification of resorption lacunae which appear as discrete areas of increased lucency.

The calvariae of 1-day-old mice were cultured in 24-well plates for 24 h, after which time parathyroid hormone (PTH) was added to a final concentration of 10−8 M. Ten micromolar alendronate, 1–100 μM mevastatin, or vehicle (PBS) was added and calvariae were cultured for a further 24 h. Calvariae were washed twice with PBS, and 0.5 ml of 2 mg/ml NBT in PBS was added, followed by incubation for 30 minutes, after which time the calvariae were washed once with PBS and fixed by the addition of 0.5 ml of 4% formaldehyde in PBS. Osteoclast numbers were counted over the entire area of the parietal bone in each half calvaria, and resorption was estimated by point counting using an eyepiece graticule, sampling the most central field of the parietal bone at ×63 magnification.

Assessment of apoptosis

Osteoclast apoptosis was assessed in vitro using murine bone marrow–derived osteoclasts, as described previously.(5) Briefly, osteoclasts were generated by culturing un-fractionated murine bone marrow in 24-well plates with alpha modified essential medium, 10% fetal calf serum, and 10−8 M 1,25-dihydroxyvitamin D3 for 7 days. Cultures were then treated for 24 h with 1–100 μM mevastatin or 100 μM alendronate. Osteoclasts undergoing apoptosis detach from the plate and can be identified and counted in cytocentrifuge preparations of the culture supernatants.(5) Tartrate-resistant acid phosphatase/hematoxylin staining was used to identify these cells, apoptosis being identified by classical nuclear changes, namely chromatin condensation and nuclear fragmentation.(5)

Apoptosis in J774 macrophages was also identified on the basis of characteristic changes in nuclear morphology, i.e., condensation of chromatin and fragmentation of nuclei into apoptotic bodies, which distinguish apoptotic cells from cells undergoing necrotic cell death.(6) J774 cells were seeded into 12-well plates (Costar, Cambridge, MA, U.S.A.) at a density of 105 cells/well. After 24 h, the medium was replaced with fresh medium containing either 1–100 μM mevastatin, 100 μM alendronate, or 30 μM mevastatin with or without 200 μM FPP, 200 μM GGPP, or 200 μM mevalonic acid lactone. After treatment for 48 h, both adherent and nonadherent cells were collected, fixed with 4% (v/v) formaldehyde, then cytospun onto slides and visualized as described by Rogers et al.(6) In addition, DNA was extracted from ˜5 × 105 J774 cells following treatment with 20 μM mevastatin for 48 h and analyzed for the presence of oligonucleosome-sized fragments by agarose gel electrophoresis.(6) The effect of mevastatin, alendronate, and risedronate on loss of total cell viability was also assessed by measuring the ability of cells to metabolize 3-(4,5-dimethylthiazol-2yl)-2,5-diphenyltetrazolium bromide (MTT), as described previously.(6)

Measurement of protein prenylation in J774 macrophages

The ability of bisphosphonates to affect protein prenylation in J774 cells was investigated by measuring incorporation of [14C]mevalonolactone into proteins post-translationally modified with farnesyl and geranylgeranyl groups.(12,14) Cells were seeded into 6-well plates at a density of 5 × 105 cells/well. After 24 h, the medium was replaced and cells were depleted of mevalonate by incubation with 5 μM (final concentration) mevastatin for 4 h. The medium was then replaced with 1.0 ml of fresh medium containing 7.5 μCi/ml [14C]mevalonolactone (specific activity 57 mCi/mmol) and either 100 μM bisphosphonate (clodronate, alendronate, risedronate, ibandronate, or YM175) or vehicle (PBS). After 24 h (the approximate time at which apoptotic cells appear in bisphosphonate-treated cultures), the adherent and nonadherent cells were collected and centrifuged at 1000g for 5 minutes. The cells were then lysed in 0.5 ml of RIPA buffer (PBS, 0.1% [w/v] sodium dodecyl sulfate [SDS], 0.5% [w/v] sodium deoxycholate, 10 μg/ml phenyl-methylsulfonyl fluoride). The protein concentration of the lysates was determined using the BCA protein assay (Pierce Chemical Co., Rockford, IL, U.S.A.). Equal quantities of protein (usually 50 μg) of each lysate were then electrophoresed on 12% polyacrylamide-SDS gels under reducing conditions. After electrophoresis, the gels were dried then visualized after exposure to a high sensitivity phosphoimaging screen (Bio-Rad, Hemel Hempstead, U.K.) for 3 days.

Immunoprecipitation of Ras from J774 macrophages

Approximately 107 cells in 75-cm2 flasks were treated for 16 h with 100 μM alendronate. Control cells were treated with vehicle (PBS). The cells were then incubated with 8 ml of methionine-free medium (containing 100 μM alendronate) for 1 h, before the addition of 80 μCi of [35S]methionine (specific activity 1000 Ci/mmol) and further incubation for 24 h. Adherent and nonadherent cells from each flask were then harvested and lysed in 1.0 ml of RIPA buffer. Ras was immunoprecipitated by overnight incubation, at 4°C, of 1 ml of lysate with 20 μl pan-Ras antibody Y13–259 conjugated to agarose beads (Oncogene Science, Manhasset, NY, U.S.A.). Immunoprecipitates were washed five times with 1 ml of RIPA buffer, then bound proteins were solubilized by boiling for 4 minutes in 30 μl of Laemmli sample buffer. Finally, samples were electrophoresed on 12.5% polyacrylamide-SDS gels under reducing conditions and detected as described above.

Statistical analysis

Differences in cell viability or the proportion of apoptotic cells following treatments were analyzed using the Mann–Whitney U-test.

RESULTS

An inhibitor of the mevalonate pathway causes macrophage and osteoclast apoptosis

Concentrations of 1–100 μM mevastatin, an inhibitor of HMG-CoA reductase(14,18) which catalyzes the synthesis of mevalonate, caused a dose-dependent increase in the proportion of J774 cells with the morphological and biochemical features typical of apoptosis, i.e., chromatin condensation and formation of apoptotic bodies (Fig. 1A and 2) and oligonucleosomal DNA fragmentation (Fig. 2). Mevastatin appeared to be more potent at inducing apoptosis than alendronate or risedronate, since the EC50 for loss of cell viability (using an MTT assay) after a 48 h treatment with mevastatin, alendronate, or risedronate was 12, 25, and 31 μM, respectively.

Figure FIG. 1..

Mevastatin (Mev) causes a dose-dependent increase in apoptosis in (A) J774 macrophages and (B) murine osteoclasts in bone marrow cultures. Apoptosis was quantitated on the basis of nuclear morphology.

Figure FIG. 2..

Nuclear morphology of J774 macrophages (A) in control cultures, (B) after treatment for 48 h with 10 μM mevastatin. Nuclei were stained with DAPI. (C) Electrophoretic analysis of internucleosomal DNA fragmentation in J774 macrophages following treatment for 48 h with 100 μM alendronate or 20 μM mevastatin.

Mevastatin, 1–100 μM, also caused a dose-dependent increase in the number of apoptotic osteoclasts in murine bone marrow cultures (Fig. 1B). Mevastatin appeared to be even more potent than alendronate at causing osteoclast apoptosis (Fig. 1B). Other cells in the marrow cultures (apart from macrophages) did not appear to be affected by mevastatin treatment (data not shown).

Mevastatin-induced J774 apoptosis could be partially prevented by coincubating J774 cells with 30 μM mevastatin and 200 μM FPP or 200 μM GGPP (Fig. 3). FPP or GGPP reduced the proportion of apoptotic cells from 80% after treatment with mevastatin alone to ˜50% (p < 0.05). Mevalonic acid lactone was even more effective at reducing the proportion of apoptotic cells (from 80% to about 10%, p < 0.05) (Fig. 3). J774 apoptosis caused by 100 μM alendronate could be partially prevented by coincubation for 48 h with 200 μM FPP or GGPP (the proportion of apoptotic cells was reduced from 70% to ˜35%, p < 0.05) (Fig. 3). Although 200 μM mevalonic acid lactone also decreased alendronate-induced J774 apoptosis (Fig. 3) the effect was not statistically significant.

Figure FIG. 3..

Inhibition of mevastatin (Mev)-induced apoptosis or alendronate (ALN)-induced apoptosis in J774 cells by intermediates of the mevalonate pathway: farnesyl pyrophosphate (FPP), geranylgeranyl pyrophosphate (GGPP), or mevalonic acid lactone (MVA). Values are the mean ± SEM (n = 4). *Significantly different from treatment with mevastatin or alendronate alone (p < 0.05).

Bisphosphonates and mevastatin inhibit bone resorption by affecting the mevalonate pathway

Concentrations ≥ 1 μM mevastatin caused a dose-dependent inhibition of PTH-stimulated bone resorption and a dose-dependent reduction in osteoclast number in murine calvariae in vitro after 24 h (Fig. 4). As expected, 10 μM alendronate also inhibited PTH-stimulated bone resorption and reduced osteoclast number in murine calvariae (Fig. 4 and 5). The inhibitory effect of 10 μM alendronate and 10 μM mevastatin on bone resorption could be overcome by coincubating calvariae with 1 mM mevalonic acid (Fig. 5). Similarly, the decrease in osteoclast number that occurred following treatment of calvariae with alendronate or mevastatin could be overcome by addition of 1 mM mevalonic acid (Fig. 5). Mevalonic acid alone did not affect PTH-stimulated bone resorption or osteoclast number (Fig. 5). FPP (200 μM) and GGPP were also tested for the ability to overcome the effects of bisphosphonates on cultures of calvariae but were found to be too toxic (data not shown).

Figure FIG. 4..

Mevastatin (Mev) causes a dose-dependent decrease in osteoclast number and bone resorption in murine calvariae in vitro. Cultures were also treated with 10 μM alendronate (ALN) as a comparison with mevastatin. Values are the mean ± SEM (n = 4).

Figure FIG. 5..

The effect of mevastatin (Mev) and alendronate (ALN) on (A) bone resorption (A) and (B) osteoclast number in murine calvariae in vitro can be overcome by the addition of 1 mM mevalonic acid (MVA). Values are the mean ± SEM (n = 4). *Significantly different from treatment with mevastatin or alendronate alone (p < 0.05).

Bisphosphonates inhibit post-translational prenylation of proteins

J774 cells metabolically labeled with [14C]mevalonolactone for 24 h contained radiolabeled proteins that could be separated by electrophoresis on 12% polyacrylamide gels into proteins of molecular weight 21–26 kDa (mostly geranylgeranylated GTP-binding proteins, but also farnesylated Ras proteins),(12,13,20) 60–70 kDa (farnesylated lamin B and prelamin A),(14,26) 17 kDa, and 46 kDa. A broad band at the migrating front of the gels (which did not stain with Coomassie blue and was not affected by prior treatment of cell lysates with proteinase K or RNAse)(12) was most likely radiolabeled, pyrophosphate-containing intermediates of the mevalonate pathway, such as FPP and GGPP. Treatment of J774 cells with 100 μM alendronate, ibandronate and risedronate during the 24 h labeling period markedly reduced the incorporation of radiolabel into the 21–26 kDa proteins and the broad band of compounds at the dye front (Fig. 6). The labeling of the 17 kDa and 46–70 kDa was also slightly reduced. YM175 (100 μM) almost completely inhibited the incorporation of radiolabel into all of the bands. By contrast, 750 μM clodronate (a concentration that reduces the viability of J774 cells but causes less than 5% apoptosis)(6) did not affect protein prenylation or synthesis of the radiolabeled dye-front compounds. Hence, there was a similar rank order of bisphosphonates in their ability to inhibit protein prenylation in J774 cells and their antire-sorptive potency (ibandronate > risedronate > YM175 > alendronate > > clodronate).(27) Inhibition of protein prenylation was not the result of an inhibitory effect of the bisphosphonates on de novo protein synthesis, since 24 h treatment with 100 μM alendronate (conditions similar to those used for metabolic labeling with [14C]mevalonolactone) did not inhibit incorporation of [35S]methionine into protein in J774 cells.(28) Furthermore, newly synthesized, [35S]-labeled Ras could be immunoprecipitated from alendronate-treated and mevastatin-treated cells (see below).

Figure FIG. 6..

The effect of bisphosphonates on incorporation of [14C]mevalonolactone into prenylated proteins. Cell lysates were from control cells (Ctrl) or cells treated with 750 μM clodronate (CLOD), 100 μM alendronate (ALN), 100 μM ibandronate (IBA), 100 μM risedronate (RIS), 100 μM YM175. Wells contain equal amounts of protein. The major [14C]-labeled bands are shown by arrowheads. Molecular size markers (kD) are also shown.

Inhibition of protein prenylation by risedronate was dose dependent. After incubating J774 cells for 24 h with 1 μM, 10, 50, or 100 μM risedronate in the presence of [14C]mevalonolactone, it was clear that 1 μM risedronate did not affect the incorporation of [14C]mevalonolactone into prenylated proteins. However, 10 μM risedronate markedly reduced protein prenylation, while 50 μM or 100 μM almost completely inhibited the incorporation of radiolabel into prenylated proteins (Fig. 7).

Figure FIG. 7..

Dose-dependent inhibition of protein prenylation by risedronate. J774 cells were incubated for 24 h with [14C]mevalonolactone in the absence or presence of 1, 10, 50, or 100 μM risedronate. Wells contain equal amounts of protein. The major [14C]-labeled bands are shown by arrowheads.

Bisphosphonates inhibit prenylation of Ras

To demonstrate further that prenylation of Ras was affected by bisphosphonates, we immunoprecipitated Ras from cell lysates of J774 macrophages that had been metabolically labeled with [35S]methionine. Immunoprecipitation of Ras from cell lysates of untreated J774 cells, using the pan-Ras antibody Y13–259, gave rise to two bands of around 21 kDa following electrophoresis of immunoprecipitates on 12.5% polyacrylamide gels. These represented the nonfarnesylated form of Ras (the minor upper band) and farnesylated Ras (the predominant lower band, which migrated faster owing to removal of the terminal tripeptide following prenylation).(16) After treatment of J774 cells for 41 h with 100 μM alendronate (by which time about 80% of the remaining cells were apoptotic) then immunoprecipitation of Ras with the Y13–259 antibody, the nonfarnesylated form (the upper band) became predominant and the farnesylated form was barely detectable (Fig. 8). Densitometric analysis of the prenylated and nonprenylated bands showed that the ratio of nonprenylated to prenylated forms of Ras increased from 0.4 in the control to 2.7 in alendronatetreated cells. Similar results were obtained after treatment with 15 μM mevastatin for 41 h.

Figure FIG. 8..

Inhibition of post-translational prenylation of Ras in J774 cells following treatment with alendronate. Ras was immunoprecipitated from cells metabolically labeled with [35S]methionine. In contrast to lysates from control cells, the nonprenylated form of Ras is more abundant than the prenylated form in lysates from alendronate-treated cells.

DISCUSSION

Bisphosphonates are capable of causing apoptosis of osteoclasts in vivo, while in vitro bisphosphonates can also cause apoptosis of osteoclasts, J774 macrophages, and myeloma cell lines.(5,6,29,30) Owing to the difficulty in studying osteoclasts in vitro, we are using the J774 macrophage cell line to help identify the molecular mechanism of bisphosphonate-induced apoptosis.(6) Our observations demonstrate that potent, nitrogen-containing bisphosphonate drugs such as risedronate, YM175, ibandronate, and alendronate can inhibit post-translational modification of proteins with isopreyl (farnesyl and geranylgeranyl) groups in J774 macrophages. Furthermore, the incorporation of [14C]mevalonolactone into prenylated proteins, including Ras, was inhibited by concentrations of the bisphosphonates that also cause apoptosis (10–100 μM).(6) Risedronate very effectively inhibited protein prenylation even at 10 μM, a concentration 3-fold lower than the EC50 for reducing total cell viability. Furthermore, another inhibitor of protein prenylation, mevastatin, was even more potent than alendronate at causing apoptosis of macrophages and murine osteoclasts and inhibited bone resorption in vitro. Hence, inhibition of protein prenylation is a likely route by which bisphosphonates cause apoptosis in J774 macrophages and osteoclasts. The fact that clodronate did not inhibit the incorporation of [14C]mevalonolactone into prenylated proteins in J774 cells, docs not inhibit sterol biosynthesis in vitro(9) and is much less potent at causing J774 apoptosis(6) also supports the view that this bisphosphonate affects cells by a mechanism that is different from that of the more potent bisphosphonates and that involves the formation of a cytotoxic metabolite.(4)

Induction of J774 apoptosis by nitrogen-containing bisphosphonates and by mevastatin could result from incorrect assembly of the nuclear lamina following loss of prenylation of lamins, thus allowing endonucleolytic digestion of chromatin.(31) Accumulation of nonprenylated GTP-binding proteins such as Ras could also lead to apoptosis, possibly as a result of intracellular acidification following loss of Ras- or Rho-dependent pH homeostasis(32,33) Indeed, we have recently found that bisphosphonate-induced and mevastatin-induced apoptosis in J774 macrophages is associated with cytoplasmic acidification.(34)

The exact enzymes of the mevalonate pathway that are inhibited by bisphosphonates remain to be identified. Mevastatin is an inhibitor of HMG-CoA reductase and thus prevents synthesis of mevalonate. Hence, mevastatin-induced apoptosis in J774 cells was prevented by the addition of mevalonic acid lactone or the mevalonate-derived compounds FPP and GGPP. The inhibitory effect of mevastatin on bone resorption and the decrease in osteoclast number in calvariae in vitro was also overcome by the addition of mevalonic acid lactone. Alendronate-induced apoptosis in J774 cells could also be partially overcome by the addition of components of the mevalonate pathway, especially FPP and GGPP. As with mevastatin-induced apoptosis, FPP and GGPP did not completely suppress alendronate-induced apoptosis, probably owing to the poor cellular uptake of such negatively charged compounds. The inhibitory effect of alendronate on bone resorption and the decrease in osteoclast number in calvariae in vitro could also be overcome by the addition of mevalonic acid lactone (which bypasses the rate-limiting step in the pathway and would increase the intracellular levels of intermediates required for synthesis of FPP and GGPP). These observations indicate that alendronate may inhibit several enzymes of the mevalonate pathway, such as FPP synthase or GGPP synthase. However, since FPP is converted to GGPP by GGPP synthase,(11) it is possible that the inhibitory effects of bisphosphonates on both farnesylation and geranylgeranylation of proteins could be the result of inhibition of FPP synthase alone. Indeed, bisphosphonates have recently been reported to inhibit FPP synthase from amoebae of the eukaryotic microorganism Dictyostelium discoideum,(35) the growth of which is inhibited by bisphosphonates.(27) Alternatively, bisphosphonates could prevent the transfer of prenyl groups to proteins by prenyl protein transferases. It is possible that bisphosphonates (like α-hydroxyfamesylphosphonate) actually inhibit several enzymes of the mevalonate pathway that contain similar prenyl-pyrophosphate binding sites. For example, YM175 and ibandronate (but not alendronate) are potent inhibitors of squalene synthase,(9,10) while we have demonstrated that YM175 and ibandronate probably also inhibit enzymes involved in FPP or GGPP synthesis. The potency of bisphosphonates could therefore depend on the combination of enzymes that are inhibited.

In preliminary studies we have found that the addition of mevalonic acid lactone does not overcome the decrease in osteoclast number and inhibition of bone resorption in calvariae in vitro caused by treatment with clodronate. This adds credence to the hypothesis that clodronate and alendronate act by different molecular mechanisms.(4) Furthermore, although we found that mevastatin is a potent inhibitor of bone resorption in vitro, it is unlikely that this cholesterol-lowering drug would inhibit bone resorption in vivo since it lacks the bone-targeting property of bisphosphonates.

Several studies have suggested that bisphosphonates affect osteoclasts and other cells by interfering with cellular metabohsm.(36,37) Our observations clearly suggest that, for the nitrogen-containing bisphosphonates, this may be due to inhibition of the mevalonate pathway and loss of protein prenylation, since the effects of the bisphosphonates on macrophages and osteoclasts could be overcome by addition of intermediates of this pathway. Since bisphosphonates may inhibit bone resorption by affecting osteoblasts(38,39) or osteoclast precursors,(40,41) as well as mature osteoclasts, it is possible that the protective effect of mevalonic acid was also on these cells in the calvarial experiment.

Prenylated proteins such as Rho, Rac, Rab, and Ras play key roles in the regulation of cytoskeletal organization, cell morphology, membrane ruffling, endocytosis, and apoptosis.(42–45) Rho and Rab proteins have also been shown to be involved in bone resorption by osteoclasts.(46,47) It is therefore likely that many of the characteristic features of bisphosphonate-treated osteoclasts, including altered cytoskeletal rearrangement,(36) loss of ruffled borders,(36,37) decreased lysosomal enzyme production,(48) and induction of apoptosis,(5) could be accounted for by inhibition of protein prenylation. Identification of the inhibitory effect of bisphosphonates on protein prenylation in macrophages may therefore have provided an important new insight into the molecular mechanism by which nitrogen-containing bisphosphonates inhibit bone resorption.

Acknowledgements

This work was funded by a grant from the Medical Research Council (ROPA), U.K., with additional support from Proctor & Gamble Pharmaceuticals.

JBMR Anniversary Classic. Nitrogen-Containing Bisphosphonates Inhibit the Mevalonate Pathway and Prevent Post-Translational Prenylation of GTP-Binding Proteins, Including Ras

SP Luckman, DE Hughes, FP Coxon, RGG Russell, MJ Rogers

Originally published in Volume 13, Number 4 pp 581–589 (1998)

 

Figure  .

Michael J. Rogers

Bisphosphonates inhibit bone resorption by mechanisms that can lead to osteoclast apoptosis. They also induce apoptosis in mouse J774 macrophages in vitro, probably by the same mechanisms that lead to osteoclast apoptosis. Luckman et al found that, in J774 macrophages, nitrogen-containing bisphosphonates (such as alendronate, ibandronate, and risedronate) inhibit post-translational modification (prenylation) of proteins, including the GTP-binding protein Ras, with farnesyl or geranylgeranyl isoprenoid groups. These observations suggested the seminal conclusion that nitrogen-containing bisphosphonate drugs cause apoptosis following inhibition of post-translational prenylation of proteins such as Ras.

Ancillary