Search

Bisphosphonates are potent inhibitors of osteoclast function widely used to treat excessive bone resorption associated, e.g., with bone metastases. They have also antitumor activity. However, it is unclear whether this reflects an indirect effect via inhibition of bone resorption or a direct antitumor effect. Nitrogen-containing bisphosphonates (N-BPs), including zoledronic acid (ZOL), act by inhibiting farnesyl pyrophosphate synthase (FPPS). The mevalonate pathway is blocked and the accumulation of isopentenyl pyrophosphate (IPP) consequently occurs. IPP is conjugated to AMP to form a novel ATP analog (ApppI). The present study was undertaken to clarify whether IPP and/or ApppI has a direct involvement in apoptosis caused by ZOL in different cancer cell lines. There are marked differences in ZOL-induced ApppI formation between different cancer cell lines. On this basis, we selected three cancer cell lines that differ significantly from each other in their ZOL-induced IPP and ApppI accumulation: human estrogen-dependent (MCF7) and estrogen-independent (MDA-MB 436) breast cancer cell lines and a human myeloma cell line (RPMI 8226). The amount of IPP/ApppI correlated with the capacity of cells to undergo apoptosis. Geranylgeraniol (GGOH), an intermediate of mevalonate metabolism, blocks both IPP and ApppI formation and to some degree ZOL-induced apoptosis in a cell line-dependent manner. In addition, lovastatin (LOV), an inhibitor of the enzyme HMGCoA reductase, completely blocks IPP/ApppI formation as determined by mass spectrometry analysis, but enhances apoptosis. In conclusion, the current data suggest that ZOL-induced IPP/ApppI formation can contribute to ZOL-induced apoptosis. This mechanism and the inhibition of protein prenylation, both outcomes of FPPS inhibition in mevalonate pathway, seem to act in concert in ZOL-induced apoptosis in cancer cells.

Bisphosphonates have become an important addition to the pharmacological armamentarium against postmenopausal osteoporosis. One of the major side effects of oral therapy with some nitrogen-containing bisphosphonates appears to be gastrointestinal (GI) intolerability, particularly esophageal irritation and ulceration. Because nitrogen-containing bisphosphonates can cause apoptosis in a variety of cell types in vitro, by inhibiting the mevalonate pathway, we hypothesized that the effect of these agents on the GI tract may be due to apoptosis or inhibition of growth of gut epithelial cells. A comparison between clodronate, etidronate, pamidronate, alendronate, and risedronate demonstrated that only the nitrogen-containing bisphosphonates were effective at inducing apoptosis or inhibiting proliferation of Caco-2 human epithelial cells in vitro, at concentrations of between 10 and 1000 μmol/L. The ability of nitrogen-containing bisphosphonates to cause apoptosis and inhibit Caco-2 cell proliferation was due to inhibition of the mevalonate pathway, because the addition of farnesol, oxidized low-density lipoprotein (LDL) cholesterol, or especially geranylgeraniol suppressed the effects. Furthermore, pamidronate, alendronate, and risedronate inhibited protein prenylation in Caco-2 cells, as determined by analysis of the processing of Rap1A, a prenylated small GTPase. These studies suggest that the effects of nitrogen-containing bisphosphonates observed in the GI tract may be due to inhibition of proliferation or apoptosis of gut epithelial cells, following loss of prenylated proteins and sterols.

Although the effects of vitamin K2 and vitamin K1 on bone metabolism have been reported, the difference between them has not been investigated. We now show the effects of menatetrenone, one of the vitamin K2 homologues, and vitamin K1 on bone resorption. Menatetrenone at greater than 3 x 10(-6) M significantly inhibited the calcium release from mouse calvaria induced by 3 x 10(-10) M of 1,25(OH)2D3 or 10(-7) M of prostaglandin E2, and it also inhibited osteoclast-like multinucleated cell (MNC) formation induced by 10(-8) M of 1,25(OH)2D3 in co-culture of spleen cells and stromal cells at the same concentrations. In contrast, the same doses of vitamin K1 had no effects on bone resorption and MNC formation in these in vitro systems. The inhibitory effect of menatetrenone on the calcium release from calvaria was not affected by the addition of 3 x 10(-5) M of warfarin, an inhibitor of vitamin K cycle. The same concentration of geranylgeraniol, the side-chain component of menatetrenone at the 3-position of the naphthoquinone, inhibited tartrate-resistant acid phosphatase (TRACP) activity and MNC formation to the same degree as menatetrenone. Phytol, the side-chain component of vitamin K1, did not affect TRACP activity at all doses tested, but weakly inhibited MNC formation. Moreover, multi-isoprenyl alcohols of two to seven units, except geranylgeraniol which contains four units, did not effect MNC formation. These findings suggest that the inhibitory effect of menatetrenone on bone resorption is not due to gamma-carboxylation and that the side chain of menatetrenone may play an important role in this inhibitory effect.

Nitrogen-containing bisphosphonates (NBps) are taken up by osteoclasts and inhibit farnesyl pyrophosphate synthase, an enzyme of the mevalonate pathway. There is evidence, however, that cells other than mature osteoclasts, like osteoclast precursors and osteoblasts, are also involved in the action of Bps on bone resorption in vitro. To examine this issue further, we developed a new in vitro model, which allows the study of the effects of additives on early osteoclast precursors. In this model, osteogenic cells are essential for osteoclastogenesis. The model consists of 15-day-old fetal mouse metatarsals. At time of explantation, these bone rudiments do not yet contain a mineralized matrix or osteoclasts; only early osteoclast precursors are present in the perichondrium. During culture and after the addition of Nabeta-glycerolphosphate, the bones form a mineralized matrix that is consequently resorbed by osteoclasts that develop from their precursors. Short treatment of these explants with Bps, before the formation of a mineralized matrix, resulted in a subsequent dose-dependent inhibition of bone resorption. The relative potencies of eight Bps to suppress resorption were comparable with those observed after the addition of Bps after the formation of a mineralized matrix, the natural target of Bps. In addition, the effects of the NBp olpadronate, but not of clodronate, on osteoclastic resorption, could be partly reversed by geranylgeraniol. Results indicate that Bps can suppress osteoclastic resorption in vitro by a direct action on very early osteoclast precursors at the bone surface, and not by affecting the osteoclastogenic capacity of osteogenic cells. Moreover, the mechanism of action of the NBp olpadronate, but not clodronate, on early tartrate-resistant acid phosphatase-negative osteoclast precursors involves inhibition of protein geranylgeranylation, indicating a molecular mechanism similar to that established for mature osteoclasts.

Bisphosphonate inhibition of bone resorption was proposed to be due to osteoclast apoptosis. We tested this hypothesis for both the N-containing bisphosphonates alendronate and risedronate, which inhibit farnesyldiphosphate synthase and thus protein isoprenylation, and for clodronate and etidronate, which are metabolized to adenosine triphosphate (ATP) analogs. We found, in dose-response studies, that alendronate and risedronate inhibit bone resorption (in pit assays) at doses tenfold lower than those reducing osteoclast number. At an N-bisphosphonate dose that inhibited resorption and induced apoptosis, the antiapoptotic caspase inhibitor, Z-VAD-FMK, maintained osteoclast (Oc) number but did not prevent inhibition of resorption. Furthermore, when cells were treated with either alendronate alone or in combination with Z-VAD-FMK for 24 or 48 h, subsequent addition of geranylgeraniol, which restores geranylgeranylation, returned bone resorption to control levels. On the other hand, Z-VAD-FMK did block etidronate and clodronate inhibition of resorption. Moreover, in cells treated with etidronate, but not alendronate or risedronate, Z-VAD-FMK also prevented actin disruption, an early sign of osteoclast inhibition by bisphosphonates. These observations indicate that, whereas induction of apoptosis plays a major role in etidronate and clodronate inhibition of resorption, alendronate and risedronate suppression of bone resorption is independent of their effects on apoptosis.