Part of this work was presented at the autumn meeting of the Bone and Tooth Society, London, U.K. September 1996.
Clodronate, alendronate, and other bisphosphonates are widely used in the treatment of bone diseases characterized by excessive osteoclastic bone resorption. The exact mechanisms of action of bisphosphonates have not been identified but may involve a toxic effect on mature osteoclasts due to the induction of apoptosis. Clodronate encapsulated in liposomes is also toxic to macrophages in vivo and may therefore be of use in the treatment of inflammatory diseases. It is generally believed that bisphosphonates are not metabolized. However, we have found that mammalian cells in vitro (murine J774 macrophage-like cells and human MG63 osteosarcoma cells) can metabolize clodronate (dichloromethylenebisphosphonate) to a nonhydrolyzable adenosine triphosphate (ATP) analog, adenosine 5′-(β,γ-dichloromethylene) triphosphate, which could be detected in cell extracts by using fast protein liquid chromatography. J774 cells could also metabolize liposome-encapsulated clodronate to the same ATP analog. Liposome-encapsulated adenosine 5′-(β,γ-dichloromethylene) triphosphate was more potent than liposome-encapsulated clodronate at reducing the viability of cultures of J774 cells and caused both necrotic and apoptotic cell death. Neither alendronate nor liposome-encapsulated alendronate were metabolized. These results demonstrate that the toxic effect of clodronate on J774 macrophages, and probably on osteoclasts, is due to the metabolism of clodronate to a nonhydrolyzable ATP analog. Alendronate appears to act by a different mechanism.
CLODRONATE (dichloromethylenebisphosphonate) is a synthetic analog of pyrophosphate1 and is widely used in the treatment of metabolic bone diseases that involve excessive bone resorption such as hypercalcemia of malignancy and osteolytic disease resulting from bone metastases1–3 and Paget's disease.4,5 Like alendronate and other bisphosphonates (BPs) that all have a P-C-P backbone,6 clodronate has high affinity for bone mineral and inhibits bone resorption in vitro and in vivo through effects on osteoclasts.1,7 Liposome-encapsulated clodronate is also toxic to macrophages in vivo following phagocytic uptake of the liposomes and intracellular release of clodronate8,9 and can inhibit the release of nitric oxide and proinflammatory cytokines from activated macrophages in vitro.10,11 Liposome-encapsulated clodronate is effective in the treatment of animal models of arthritis and may therefore eventually be used to treat inflammatory diseases such as rheumatoid arthritis.12–16
The exact molecular mechanisms of action of BPs have not yet been identified.17,18 BPs may have direct effects on osteoclasts that have internalized BPs during resorption of BP-coated bone mineral19–21 through mechanisms that can lead to osteoclast cell death by apoptosis.22 In addition, BPs have also been suggested to have effects on hemopoietic osteoclast precursors or on other cells in the bone microenvironment such as osteoblasts and macrophages.23–27 There is also evidence that alendronate and other recently developed, potent antiresorptive BPs may act by a different or additional mechanism to that of clodronate, one of the first BPs to be used clinically.28,29 For example, alendronate is a much more potent inducer of apoptosis than clodronate in J774 macrophages.30
We and others have previously shown that clodronate, and other BPs that closely resemble pyrophosphate in structure, can be metabolized by amebae of the slime mold Dictyostelium discoideum to methylene-containing, nonhydrolyzable analogs of ATP.31–34 In the case of clodronate, the metabolite is adenosine 5′-(β,γ-dichloromethylene) triphosphate (AppCCl2p).32–34 The more potent BPs such as alendronate, which have a nitrogen-containing side chain attached to the geminal carbon of the P-C-P group, are not metabolized.33
Incorporation of BPs such as clodronate into adenine nucleotides appears to occur by replacement of pyrophosphate with BP in a back-reaction catalyzed by cytoplasmic Class II aminoacyl-tRNA synthetase enzymes.33 Cell-free extracts from human cells can also metabolize the same BPs that are metabolized by Dictyostelium.35 Apart from one report suggesting that clodronate can be modified by calvarial cells in vitro following cellular uptake36 there is no other evidence that intact mammalian cells can metabolize clodronate. Indeed, it is generally considered that BPs are metabolically inert in vivo.
Given the difficulty in isolating large numbers of osteoclasts, we have used anion-exchange fast protein liquid chromatography (FPLC) to examine whether other mammalian cells, intact murine J774 macrophage-like cells, and human osteoblast-like MG63 cells, can metabolize clodronate to AppCCl2p in vitro. Both J774 and MG63 cells, like osteoclasts and Dictyostelium amebae, can internalize fluorescently labeled BP (Ref. 37 and unpublished results) and would therefore be expected to be capable of metabolizing clodronate. We also compared the ability of the phagocytic J774 cells to metabolize liposome-encapsulated clodronate and liposome-encapsulated alendronate.
The effect of accumulation of AppCCl2p within cells was examined by using AppCCl2p encapsulated within liposomes, which are internalized by J774 macrophages but not by MG63 cells. Since liposome-encapsulated clodronate has recently been reported to cause apoptosis in macrophages in vitro,38 we also investigated whether liposome-encapsulated clodronate and liposome-encapsulated AppCCl2p caused the characteristic morphological and biochemical features of apoptosis in J774 cells (chromatin condensation, nuclear fragmentation, and internucleosomal DNA cleavage; reviewed in Refs. 39 and 40).
MATERIALS AND METHODS
The disodium salts of clodronate (dichloromethylene-1,1-bisphosphonate) and alendronate (4-amino-1-hydroxybutylidene-1,1-bisphosphonate) were a kind gift from Gentili S.p.a. (Pisa, Italy). AppCCl2p was synthesized according to Blackburn et al.41 Medronic acid (methylene-1,1-bisphosphonic acid) and all other reagents were purchased from Sigma Chemical Co. (Poole, U.K.) unless stated otherwise.
J774.2 cells were obtained from the European Collection of Animal Cell Cultures (Porton Down, U.K.), and MG63 cells were obtained from the American Type Culture Collection (Rockville, MD, U.S.A.). Cultures were grown at 37°C in Dulbecco's modified Eagle's medium (DMEM) containing 10% heat inactivated fetal calf serum, 100 U/ml penicillin, 100 μg/ml streptomycin, and 1 mM glutamine in 5% CO2 atmosphere.
Preparation of bisphosphonate solutions and liposome-encapsulated bisphosphonates
Stock solutions of free clodronate, medronate, and alendronate at 50 mM were prepared by dissolving the BPs in phosphate-buffered saline (PBS), the pH was adjusted to 7.4 using 5 M NaOH, and the solutions were filter-sterilized using a 0.22 μm filter.
Clodronate and alendronate were encapsulated in distearoylphosphatidylglycerol liposomes using 100 mM solutions of the BPs in deionized water as described previously.42 Liposome encapsulation of AppCCl2p was carried out using a solution of 50 mM AppCCl2p in deionized water. Stock suspensions of clodronate- or alendronate-containing liposomes contained approximately 10 mM clodronate or alendronate, while AppCCl2p-containing liposomes contained approximately 2.5 mM AppCCl2p, i.e., disruption of the liposomes gave rise to these concentrations of compounds in solution. In different preparations, the molar drug:phospholipid ratio of clodronate- or alendronate-containing liposomes was always approximately 1.0, while that of AppCCl2p-containing liposomes varied between 0.2 and 0.4.42 Nonloaded liposomes (phospholipid concentration 10.6 mM) were used as a control. Liposomes were diluted to the required concentrations in DMEM immediately prior to use.
MTT assay for cell viability
To determine the concentrations of BPs that affected cell viability and hence concentrations that could be used to treat large-scale cultures of cells, we assessed the ability of viable cells to reduce MTT reagent (3-(4,5-dimethylthiazol-2yl)–2,5-diphenyltetrazolium bromide) following BP treatment. J774 and MG63 cells were plated at a density of 5 × 103 cells/well and 2 × 103 cells/well, respectively, in 96-well tissue culture plates (Costar, Cambridge, MA, U.S.A.). Eight wells were left without cells as a blank. Approximately 18 h later the medium was replaced with 100 μl of fresh medium containing BPs (5–1000 μM for free BPs; 0.1–500 μM for liposome-encapsulated compounds) in replicates of four or five wells. Control wells contained either the equivalent volume of PBS or the equivalent amount of phospholipid in the form of nonloaded liposomes. After incubation for 48 h, 10 μl of MTT reagent (0.5 mg/ml) were added to each well followed by incubation for a further 4 h at 37°C. The insoluble MTT product was dissolved by addition of 100 μl 20% (w/v) sodium dodecyl sulfide (SDS), 50% (v/v) dimethylformamide, pH 4.7, and incubation for 1 h at 37°C. The absorbance of the wells was measured at 570 nm using an Anthos Labtech 2001 plate reader (Anthos Labtech, Salzburg, Austria). The mean value of absorbance from the blank wells was subtracted from the values of the other wells. Results were then expressed as percentage of control (mean ± SEM).
Preparation of cell extracts
Four 162 cm2 flasks containing a total of approximately 107 MG63 or J774 cells were treated for 48 h with the lowest concentration of BPs that significantly reduced cell viability in the MTT assay, or with an equivalent volume of PBS. After treatment, the medium was removed and the adherent cells were washed with PBS. The cells were scraped from the flasks in 10 ml of ice-cold 7% (v/v) perchloric acid and left at 4°C for 1 h.32,33 The extracts were neutralized with a saturated solution of KHCO3, left for 30 minutes on ice and then centrifuged (5 minutes, 220g) to remove insoluble potassium perchlorate. Supernatants were then lyophilized and the dried extracts dissolved in 1 ml of distilled water. Adenosine triphosphate (ATP) and adenosine diphosphate (ADP) were converted to adenosine monophosphate (AMP) by incubating the extracts with 10 U of Grade V apyrase35 for 2 h at 37°C. Since nucleotides containing a β,γ-methylene group, such as AppCCl2p, are resistant to apyrase digestion, this allows easier detection of metabolites of BPs in cell extracts.35
Identification of BP metabolites by FPLC analysis
Two hundred microliter samples of the extracts were analyzed by FPLC using a 1.0 ml Pharmacia MonoQ anion-exchange column (Pharmacia, Uppsala, Sweden) eluted in a gradient of NH4HCO3 (120–360 mM NH4HCO3 over 17 minutes followed by 360–120 mM NH4HCO3 over 3 minutes, at a flow rate of 1.5 ml/minute) as described previously.32,33,35 Eluted nucleotides were detected by their absorption at 254 nm and identified by comparison with the retention times of authentic standards (AMP, ADP, ATP, AppCCl2p, and AppCH2p) and by spiking extracts with the standards.
Assessment of nuclear morphology and DNA fragmentation
J774 cells were seeded into 12-well plates at a density of 2 × 105 cells/well. After 24 h, the cells were treated with BP, liposome-encapsulated BP, liposome-encapsulated AppCCl2p, PBS, or nonloaded liposomes in replicates of at least three wells. After incubation at 37°C for either 6, 24, or 48 h, the nonadherent cells from each well were pooled with the adherent cells. The proportion of cells that excluded trypan blue was assessed before dividing the samples for analysis of nuclear morphology or for DNA extraction.30 In each case, cells were then centrifuged at 3200g for 5 minutes.
To examine nuclear morphology following treatment, the cells were fixed by resuspending the pellets in 400 μl of 4% (w/v) formaldehyde for 10 minutes. Two hundred microliters were cytospun onto glass slides using a Shandon cytospin (5 minutes, 500 rpm) before staining nuclei with 1 μg/ml DAPI for 10 minutes and mounting in Citifluor (Agar Scientific, Stansted, U.K.). Slides were then examined under a 40× objective using a Leitz DMRB fluorescence microscope equipped with a Photometrics CCD camera and SmartCapture image analysis software (Vysis U.K. Ltd., Richmond, U.K.). The percentage of apoptotic nuclei was determined by counting five random fields, each of 100 nuclei, per slide. Photographic images were also taken under a 100× oil immersion objective.
For analysis of internucleosomal DNA fragmentation, a characteristic feature of cells undergoing apoptosis, DNA was extracted from the cell pellets as described by Rogers et al.30 The DNA was electrophoresed, together with molecular weight markers of 123 base pair multiples (GIBCO BRL, Paisley, U.K.), for 3 h at 70 V on a 1.2% agarose gel containing 1 μg/ml ethidium bromide. Bands were visualized and photographed with ultraviolet transillumination.
Effects of bisphosphonates on cell viability
Free clodronate, medronate, and alendronate significantly decreased total viability of cultures of both J774 and MG63 cells after 48 h of treatment, although J774 cells were more susceptible to all three BPs (Figs. 1A and 1B). Alendronate was more potent than clodronate or medronate in both cell types. Liposome-encapsulated clodronate and liposome-encapsulated alendronate decreased total viability of J774 cells but did not affect MG63 cells, even at 500 μM (Figs. 1C and 1D). Encapsulation of the BPs increased the potency of clodronate and alendronate toward J774 cells by 300-fold and 33-fold, respectively (Table 1).
Table Table 1. THE POTENCY OF FREE CLODRONATE AND ALENDRONATE COMPARED WITH LIPOSOME-ENCAPSULATED CLODRONATE AND ALENDRONATE FOR REDUCTION OF CELL VIABILITY IN J774 AND MG63 CELLS
Free clodronate and medronate, but not alendronate, are metabolized by J774 cells
J774 cells were treated with the lowest concentrations of BPs that were found to have a significant effect on total cell viability after treatment for 48 h, i.e., 250 μM clodronate, 50 μM alendronate, or 100 μM medronate. Since alendronate was more potent at reducing cell viability than clodronate (shown in Table 1), it was not possible to incubate cultures with the same concentration of alendronate as that of clodronate (250 μM) because this would have led to insufficient numbers of cells with which to prepare cell extracts.
FPLC analysis of perchloric acid extracts from J774 cells that had been treated for 48 h with 250 μM clodronate demonstrated the presence of an apyrase-resistant peak eluting at 7.5 minutes (Fig. 2B) that was not present in extracts prepared from PBS-treated cells (Fig. 2A). This peak coeluted with authentic AppCCl2p. Extracts from medronate-treated cells contained two apyrase-resistant peaks eluting at 7 minutes and 11 minutes, which coeluted with AppCH2p and AppCH2ppA, respectively (data not shown). By contrast, extracts from alendronate-treated cells did not contain any new peaks that were not present in the control (Fig. 2C).
Free clodronate and medronate, but not alendronate, are metabolized by MG63 cells
MG63 cells were treated with the lowest concentrations of BPs that were found to have a significant effect on cell viability after 48 h of treatment, i.e., 750 μM clodronate, 100 μM alendronate, or 100 μM medronate.
Extracts from clodronate-treated MG63 cells contained an apyrase-resistant peak that eluted at 7.5 minutes (Fig. 3B) and that was not present in extracts prepared from PBS-treated cells (Fig. 3A). This peak coeluted with authentic AppCCl2p. Extracts from medronate-treated cells contained two minor apyrase-resistant peaks eluting at 7 minutes and 11 minutes, which coeluted with AppCH2p and AppCH2ppA, respectively (data not shown). Extracts from alendronate-treated cells did not contain any new, apyrase-resistant peaks that were not present in the control (Fig. 3C).
Liposome-encapsulated clodronate, but not liposome-encapsulated alendronate, is metabolized by J774 macrophages
J774 cells were treated for 24 h with 15 μM liposome-encapsulated clodronate, 1.5 μM liposome-encapsulated alendronate, or nonloaded liposomes containing an equivalent amount of phospholipid to that of the drug-encapsulated liposomes.
Extracts from cultures treated with liposome-encapsulated clodronate contained a new, apyrase-resistant peak of absorbance at 254 nm that coeluted with authentic AppCCl2p (Fig. 4A). This peak was not present in extracts from cultures that had been treated with liposome-encapsulated alendronate (Fig. 4B) or nonloaded liposomes (Fig. 4C). The size of the AppCCl2p peak in the extracts increased when the cells were treated with higher doses of liposome-encapsulated clodronate (data not shown). As expected, MG63 cells did not metabolize 15 μM liposome-encapsulated clodronate or 5 μM liposome-encapsulated alendronate, even after treatment for 48 h (data not shown). Furthermore, these concentrations did not affect total cell viability (Figs. 1C and 1D).
Effect of liposome-encapsulated AppCCl2p on viability of J774 macrophages
The effect of liposome-encapsulated AppCCl2p on the total viability of cultures of J774 cells following treatment for 48 h was determined using the MTT assay as described above. In three independent experiments, the potency of liposome-encapsulated AppCCl2p was higher than that of liposome-encapsulated clodronate. In the representative experiment shown in Fig. 5, the EC50 for liposome-encapsulated clodronate was 2 μM, while the EC50 for liposome-encapsulated AppCCl2p was 0.3 μM (after taking into account the molar drug:phospholipid ratio of the two liposome preparations, 1.0 and 0.2, respectively).
Cell death induced in J774 macrophages by liposome-encapsulated clodronate, AppCCl2p, or alendronate
Rounding and loss of cell adhesion began to occur approximately 6 h after commencing treatment with 15 μM liposome-encapsulated clodronate or 10 μM liposome-encapsulated AppCCl2p. By contrast, changes in the morphology of cells treated with 5 μM liposome-encapsulated alendronate or free alendronate30 only occurred after approximately 20 h.
To examine whether liposome-encapsulated clodronate or its metabolite could cause apoptosis, we examined the morphology of DAPI-stained nuclei following treatment of J774 macrophages with liposome-encapsulated clodronate or liposome-encapsulated AppCCl2p. Cell death by apoptosis can be identified on the basis of changes in nuclear morphology, which involve condensation of chromatin and nuclei and fragmentation of nuclei into apoptotic bodies.30,39 By contrast with control cells (Fig. 6A), some cells that had been treated with 15 μM liposome-encapsulated clodronate for 6 h or more had pale, diffusely stained chromatin and poorly defined, irregular nuclear membrane profile, as well as swollen, disintegrating nuclei (Fig. 6B) characteristic of necrosis.39 This was supported by an increase in the proportion of cells that failed to exclude trypan blue, a feature of necrotic cell death.39 After treatment with 15 μM liposome-encapsulated clodronate or 10 μM liposome-encapsulated AppCCl2p for 48 h, the proportion of trypan blue–positive cells was 22% and 35%, respectively, compared with 7% in cultures treated with nonloaded liposomes. In addition, the proportion of apoptotic cells increased slightly (2% at 24 h) compared with control cells that had been treated with nonloaded liposomes (<1% at 24 h). However, this difference was statistically significant (p < 0.05) only after treatment for 24 h but not after 48 h (Fig. 7A). The nuclei of cells that had been treated with 10 μM liposome-encapsulated AppCCl2p for 6 h or more showed similar morphological features to those treated with liposome-encapsulated clodronate (Fig. 6C). In addition to the appearance of necrotic nuclei, liposome-encapsulated AppCCl2p also caused an increase in the proportion of apoptotic nuclei compared with the control, although this only just reached statistical significance (p < 0.05) after 48 h of treatment (Fig. 7B). The presence of oligonucleosomal DNA fragments is also a characteristic feature of apoptosis.30,39 However, neither liposome-encapsulated clodronate nor liposome-encapsulated AppCCl2p caused sufficient apoptosis to detect internucleosomal DNA fragmentation after treatment for 6, 24, or 48 h (Fig. 8).
By contrast with liposome-encapsulated clodronate and liposome-encapsulated AppCCl2p, liposome-encapsulated alendronate caused a dose- and time-dependent increase in the proportion of apoptotic cells, which occurred after treatment for 24 h or more (Figs. 6D and 7A). Furthermore, liposome-encapsulated alendronate caused internucleosomal DNA fragmentation that was detectable after treatment for 24 h and more evident after 48 h (Fig. 8).
We have demonstrated that intact mammalian cells in vitro can metabolize clodronate to a dichloromethylene-containing analog (AppCCl2p) of ATP, which could be identified in cell extracts from J774 and MG63 cells by using anion-exchange FPLC. Although J774 cells were incubated with a concentration of clodronate that was 3-fold lower than that used to treat approximately the same number of MG63 cells, the size of the peaks of ultraviolet absorbance owing to the metabolite in the FPLC profiles were similar. Hence, J774 cells appear to be more efficient at metabolizing clodronate than MG63 cells, probably as a result of a greater capacity of J774 macrophages to internalize clodronate by pinocytosis.
Medronate, like clodronate, was metabolized by both J774 and MG63 cells to a methylene-containing analog (AppCH2p) of ATP. In addition, medronate was also incorporated into a methylene-containing analog (AppCH2ppA) of diadenosine tetraphosphate. By contrast, no metabolites of alendronate could be detected by using FPLC. We have obtained the same results following incubation of Dictyostelium discoideum amebae32,33 or incubation of cell-free extracts of mammalian cells with these three BPs.35 Both J774 and MG63 cells were incubated with lower concentrations of alendronate than clodronate (owing to the higher cytotoxicity of alendronate), raising the possibility that low levels of an alendronate metabolite could have been produced but which would not have been detected in our assay. However, we have previously shown that cell-free extracts of mammalian cells can metabolise 500 μM clodronate but not 500 μM alendronate.35 These observations indicate that alendronate is not metabolized.
Liposome-encapsulated clodronate could also be metabolized to AppCCl2p by phagocytic J774 cells following endocytosis of the liposomes, but not by nonphagocytic MG63 cells. Furthermore, as expected, clodronate was internalized by J774 macrophages more efficiently when encapsulated in liposomes, since the size of the ultraviolet absorbance peak owing to the metabolite was similar in extracts from cells treated for 24 h with 15 μM liposome-encapsulated clodronate and in extracts from cells treated for 48 h with 250 μM free clodronate. As with free alendronate, no metabolite could be detected in extracts from J774 macrophages that had been treated with liposome-encapsulated alendronate.
The incorporation of clodronate and medronate was most likely the result of a pyrophosphate-utilizing back reaction catalyzed by several Class II aminoacyl-tRNA synthetases.33 Since clodronate and medronate closely resemble pyrophosphate, these compounds could replace pyrophosphate in a condensation reaction with aminoacyl-adenylate. The resulting metabolite of medronate, AppCH2p, may undergo a further condensation with aminoacyl-adenylate to form AppCH2ppA.33 This mechanism is supported by the observation that the BPs that are metabolized (including clodronate and medronate) inhibit the forward reaction catalyzed by several human Class II but not Class I aminoacyl-tRNA synthetases in vitro.32 Alendronate, which does not appear to be incorporated into adenine nucleotides, does not inhibit the activity of aminoacyl-tRNA synthetase enzymes.32,35
The incorporation of clodronate into AppCCl2p by aminoacyl-tRNA synthetases could have several consequences for cell metabolism, including depletion of intracellular ATP, inhibition of protein synthesis, and accumulation of a toxic ATP analog.31 Since clodronate does not affect the level of ATP in macrophage-like RAW264 cells (J. Mönkkönen, personal observation), we investigated whether accumulation of intracellular AppCCl2p could account for the cytotoxic effects of clodronate by using liposome-encapsulated AppCCl2p. Internalization of these liposomes by J774 macrophages should lead to intracellular accumulation of AppCCl2p following disruption of the liposomes by phospholipases.43 We found that the potency of liposome-encapsulated AppCCl2p for reducing total viability of J774 cell cultures was even greater than that of liposome-encapsulated clodronate, after taking into account the drug:phospholipid ratio of the two liposome preparations. Both liposome-encapsulated clodronate and liposome-encapsulated AppCCl2p caused changes in cell morphology after approximately 6 h, together with cell death mostly by necrosis. By contrast, liposome-encapsulated alendronate was much more potent at causing cell death with the characteristic features of apoptosis, namely chromatin condensation, nuclear fragmentation, and internucleosomal DNA laddering.30,39,40 In addition, these features only became apparent after approximately 20 h. These results suggest that the effects of clodronate, but not of alendronate, on J774 macrophages can be attributed to the intracellular accumulation of a nonhydrolyzable analog of ATP. This metabolite could have numerous inhibitory effects on biosynthetic, metabolic, or signal transduction pathways that involve the hydrolysis or synthesis of ATP, including DNA synthesis, glycolysis, and protein phosphorylation. Indeed, AppCCl2p is known to be a potent inhibitor of yeast hexokinase and rabbit muscle pyruvate kinase.44 This may explain the observations by Fast et al.45 that clodronate could inhibit glycolysis and lactate production by connective tissue cells in vitro. Furthermore, our demonstration that cell-free extracts can convert clodronate to detectable levels of AppCCl2p only after several hours35 may also explain why Felix and Fleisch46 could not detect inhibition of lactate production in cell-free extracts of calvarial cells after just 1 h of incubation with clodronate. Alendronate, by contrast with clodronate, appears to act by a mechanism that does not involve accumulation of a nucleotide metabolite and that leads to apoptosis in J774 cells.30
In vivo, osteoclasts are the cells most likely to be exposed to relatively high concentrations of BPs during resorption of bone and release of BP from hydroxyapatite in the acidic environment of the resorption space.19,20 It is therefore likely that osteoclasts also metabolize clodronate to AppCCl2p. This may account for the clodronate-induced effects on morphology and viability of osteoclasts both in clodronate-treated animals and in vitro.21,22,47–50 Whereas we found that J774 macrophages underwent apoptotic and necrotic cell death after treatment with liposome-encapsulated clodronate and liposome-encapsulated AppCCl2p, other cells (including osteoclasts, peritoneal macrophages, and peripheral blood monocytes) appear to be more susceptible to cell death by apoptosis following treatment with clodronate or liposome-encapsulated clodronate (J.C. Frith and M.J. Rogers, unpublished observations, and Refs. 22 and 38). This may reflect differences in the dependence of macrophages and osteoclasts on certain metabolic pathways51 or differences in the intracellular signaling pathways that control cell death in osteoclasts and other cell types.
Macrophages play a key role in inflammatory diseases such as rheumatoid arthritis, which is characterized by chronic joint inflammation, hyperplasia of the synovial lining, and increased numbers of macrophages in the joint cavity. Systemic as well as intra-articular injection of liposome-encapsulated clodronate into rats with adjuvant-induced or antigen-induced arthritis has been found to cause selective elimination of synovial lining cells and reduce inflammation.10–13 Our demonstration that liposome-encapsulated clodronate can be metabolized to a toxic metabolite, AppCCl2p, provides a mechanism by which macrophages, but not nonphagocytic cells, can be selectively affected or killed following ingestion of the liposomes.
Until now, the exact mechanisms of action of BPs have not been identified. It is clear that different BPs have different biological properties, since the potent, amino-containing BPs such as alendronate and pamidronate can stimulate cytokine production by mononuclear cells in vitro and cause a transient, acute-phase response in vivo, whereas clodronate does not.52–54 The finding that clodronate, but not alendronate, can be metabolized is an important step forward in our ability to account for such differences and it supports the view that different BPs may inhibit bone resorption by different molecular mechanisms. For alendronate, this mechanism remains to be identified, a subject that is currently under investigation.
This work was supported by a Medical Research Council (ROPA) project grant. We are also grateful to the MRC for a studentship for J.C.F. M.J.R. is supported by the J.G. Graves Medical Research Fellowship and J.M. by the Academy of Finland. We also thank Dr. Donald Watts for his encouragement and for many helpful suggestions and Mr. Markku Taskinen for preparing the liposomes.