Bisphosphonates are used with increasing frequency in the management of skeletal complications in patients with breast cancer. In this paper, we have investigated whether bisphosphonates, besides their known beneficial effects on tumor-associated osteoclastic resorption, are capable of inhibiting breast cancer cell adhesion to bone matrix. For that we used two in vitro models for bone matrix (cortical bone slices and cryostat sections of trabecular bone from neonatal mouse tails).Four bone matrix-bound nitrogen-containing bisphosphonates (pamidronate, olpadronate, alendronate, and ibandronate) inhibited adhesion and spreading of breast cancer cells to bone dose-dependently, whereas etidronate and clodronate had little or no effect. Strikingly, the relative order of potency of the bisphosphonates in inhibiting the adhesion of cancer cells to cortical and trabecular bone corresponded to their relative antiresorptive potencies in vivo as well as their ranking in in vitro bone resorption assays with predictive value for their clinical efficacy. It appears that nitrogen-containing bisphosphonates alter selectively the adhesive properties of the extracellular bone matrix preventing the attachment of breast cancer cells to it.Besides the beneficial effects of bisphosphonates on tumor-induced osteoclastic resorption, the previously unrecognized effect presented in this paper makes these agents suitable for earlier pharmacologic intervention in patients with breast cancer at risk of developing bone metastases.

No abstract

Bisphosphonates, synthetic compounds used in the treatment of skeletal disorders, suppress osteoclast-mediated bone resorption by a yet unidentified mechanism. Previous studies showed that some bisphosphonates can inhibit enzymes of the mevalonate pathway, and nitrogen-containing bisphosphonates inhibit protein prenylation in mouse macrophages. In the present study, we examined the involvement of the mevalonate pathway in basal and bisphosphonate-inhibited osteoclastic resorption in fetal mouse long bone explants, an experimental model representative of the in vivo action of bisphosphonates. Mevastatin inhibited bone resorption at concentrations similar to those of the potent bisphosphonate ibandronate. This effect could be totally reversed by the addition of mevalnate and geranylgeraniol but not farnesol. The first two intermediates but not the latter could also stimulate basal bone resorption. The inhibitory effect of ibandronate on bone resorption could be totally reversed by the addition of geranylgeraniol and to a small extent only by mevalonate and farnesol, indicating that the bisphosphonate acts at a level of the mevalonate pathway different from that of mevastatin. Histologic sections of ibandronate-treated bone explants showed further rescue of functioning osteoclasts during concomitant treatment with geranylgeraniol. Finally, the reversibility of bisphosphonate inhibited osteoclastic resorption by geranylgeraniol was also demonstrated for the potent nitrogen-containing bisphosphonates alendronate, olpadronate, and risedronate but not for the non-nitrogen-containing bisphosphonates clodronate and etidronate. These studies demonstrate that protein geranylgeranylation but not farnesylation is important for osteoclast-mediated bone resorption and that nitrogen-containing bisphosphonates exert their antiresorptive action probably by affecting enzymes of the mevalonate pathway involved in the generation of geranylgeranyl pyrophosphate.

Acquisition of an invasive phenotype by cancer cells is a requirement for bone metastasis. Transformed epithelial cells can switch to a motile, mesenchymal phenotype by epithelial-mesenchymal transition (EMT). Recently, it has been shown that EMT is functionally linked to prostate cancer stem cells, which are not only critically involved in prostate cancer maintenance but also in bone metastasis. We showed that treatment with the non-peptide α(v)-integrin antagonist GLPG0187 dose-dependently increased the E-cadherin/vimentin ratio, rendering the cells a more epithelial, sessile phenotype. In addition, GLPG0187 dose-dependently diminished the size of the aldehyde dehydrogenase high subpopulation of prostate cancer cells, suggesting that α(v)-integrin plays an important role in maintaining the prostate cancer stem/progenitor pool. Our data show that GLPG0187 is a potent inhibitor of osteoclastic bone resorption and angiogenesis in vitro and in vivo. Real-time bioluminescent imaging in preclinical models of prostate cancer demonstrated that blocking α(v)-integrins by GLPG0187 markedly reduced their metastatic tumor growth according to preventive and curative protocols. Bone tumor burden was significantly lower in the preventive protocol. In addition, the number of bone metastases/mouse was significantly inhibited. In the curative protocol, the progression of bone metastases and the formation of new bone metastases during the treatment period was significantly inhibited. In conclusion, we demonstrate that targeting of integrins by GLPG0187 can inhibit the de novo formation and progression of bone metastases in prostate cancer by antitumor (including inhibition of EMT and the size of the prostate cancer stem cell population), antiresorptive, and antiangiogenic mechanisms.