A Third-Generation Bisphosphonate, Minodronic Acid (YM529), Augments the Interferon α/β-Mediated Inhibition of Renal Cell Cancer Cell Growth Both In vitro and In vivo

Renal cell cancer (RCC) is the most lethal of the urologic malignancies and its incidence is currently on the increase (1). At the time of initial diagnosis, one third of patients with RCC exhibit visceral metastasis, and half of the remainder eventually develop distant metastases (2, 3). Currently, the only effective therapeutic and preventive agents for distant metastases and local recurrence are IFN and interleukin 2, although these agents have achieved response rates of only 15% (2). Therefore, the discovery of a novel, more effective therapeutic agent is urgently needed.

Bisphosphonate is an inhibitor of bone resorption and has been shown to directly and indirectly prevent proliferation and inhibit metastasis of various types of cancer cells (4–9). Bisphosphonates inhibit farnesyl pyrophosphate synthase, which serves in the mevalonate pathway (9). Consequently, they inhibit the activation of small G-proteins such as Ras, Rap1, and Rho, reduce the signals they mediate, and thereby prevent the growth, adhesion/spreading, and invasion of cancer cells (4–9). In addition, bisphosphonate-induced inhibition of the mevalonate pathway increases the levels of the middle metabolic products of isopentenyl pyrophosphate, which stimulate γ/δ T lymphocytes (10). γ/δ-T lymphocytes exhibited marked cytotoxicity against various tumor cells including RCC (11). Moreover, bisphosphonates inhibit angiogenesis by reducing the levels of vascular endothelial growth factor (VEGF) that are produced by the cancer cells (12).

RCC is an attractive target for both angiogenesis inhibitors and immunomodulators because angiogenesis is crucial for tumor development and RCC responds well to immunotherapy. The third-generation bisphosphonate zoledronic acid delays the onset of skeletal-related events and the progression of skeletal disease in patients with advanced RCC (13). However, the antitumor effect of bisphosphonates against RCC remains unknown. In this study, we show that minodronic acid (also known as YM529), another third-generation bisphosphonate, has anticancer activity and synergistically augments the growth inhibitory effects of IFN against RCC both in vitro and in vivo.

YM529 (3-amino-1-hydroxy-propylidene-1.1-bisphosphonate) was obtained from Yamanouchi (Tokyo, Japan). Natural human IFN-α and mouse IFN-α/β were kindly provided by Hayashibara Biochemical Laboratories, Inc. (Okayama, Japan). The human RCC cell lines 293, ACHN, CCFRC-1, CCFRC-2, CAKI-1, CAKI-2, NC65, and RPMI-SE were obtained from the American Type Culture Collection (Rockville, MD). The murine RCC cell line RENCA was stably transfected with luciferase (Luc) to produce RENCA^Luc cells, as described previously (8). Approval for these studies was obtained from the institutional review board at Kyoto University Hospital. Specific pathogen-free 6- to 8-week-old BALB/c mice were used in this study (SLC, Kyoto, Japan).

Western blot analysis was done as described previously (5). Goat polyclonal anti-unprenylated Rap1A antibody (diluted 1:1,000; Santa Cruz Biotechnology, Santa Cruz, CA) and mouse monoclonal anti-Ras antibody (diluted 1:1,000; Becton Dickinson, San Jose, CA) were used as the primary antibodies.

Cell proliferation was determined by a modified 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide(MTT) assay, as described previously (5). The IC₅₀ was obtained using the nonlinear regression program CalcuSyn (Biosoft, Cambridge, United Kingdom).

Two-color flow cytometric analyses were conducted by fluorescence-activated cell-sorting analysis (BD Biosciences, Mountain View, CA), as described previously (5). The antibodies used were FITC-conjugated anti-mouse TCR γ/δ (Immunotech, Marseille, France), Cychrome-conjugated anti-mouse TCRβ (BD PharMingen, San Diego, CA), and phycoerythrin-conjugated anti-mouse CD3 (BD PharMingen).

The concentrations of VEGF of the culture supernatants or the sera of the mice were determined by using ELISA (R&D Systems, Minneapolis, MN) according to the manufacturer's protocol. To investigate the serum concentrations of VEGF, the sera of treated mice (see below) were collected and analyzed.

To generate s.c. or renal orthotopic models of RCC, RENCA^Luc cells (1 × 10⁵ per 100 μL PBS) were injected into the s.c. tissue of the lateral flank or the renal subcapsule after exposing the kidney, respectively. Tumor growth was monitored by using an in vivo imaging system (IVIS, Xenogen, Alameda, CA) with an aqueous solution of luciferin (150 mg/kg i.p., Xenogen), as described previously (8).

Mice were given the indicated number of RENCA^Luc cells on day 0. One week later, the mice were observed by IVIS and the bioluminescence from the implanted cancer cells was measured. The mice were then divided into four groups of seven mice in such a way that each group had almost the same average bioluminescence. The mice were then either left untreated or were treated thrice a week for 2 weeks with 80 μg/kg YM529 and/or 1 × 10⁴ units of mouse IFNα/β. The YM529 and mouse IFN were given s.c. Three weeks after RENCA^Luc cell inoculation, all mice were killed humanely and their sera were collected. Serum levels of the following were then determined: aspartate aminotransferase, alanine aminotransferase, lactate dehydrogenase, total protein, creatinine, blood urea nitrogen, calcium, and VEGF.

Statistical significance was determined by the Student's ttest (P < 0.05) using Excel (Microsoft, Redmond, OR).

We investigated whether YM529 prevented the prenylation of Rap1A (which was activated after geranylgeranylation) and Ras (which was mainly activated after farnesylation). YM529 clearly inhibited the prenylation of Rap1A as it resulted in a dose- and time-dependent increase in unprenylated Rap1A levels in both human and murine renal cancer cell lines (ACHN and RENCA, respectively; Fig. 1A). The prenylation status of Ras was slightly inhibited in YM529-treated ACHN cells but its prenylation status was not altered in treated RENCA cells. Thus, similar to other cancer cells, bisphosphonates can inhibit prenylation in RCC cells, although they prevented geranylgeranylation more than farnesylation.

To examine the indirect effects of YM529 and IFN on RCC, we investigated the concentrations of VEGF in the culture supernatant of YM529-treated and untreated RENCA cells. The treatment reduced the elaboration of VEGF in a dose-dependent manner, with 50% reduction being achieved with 28.0 μmol/L YM529 (Fig. 1B). In contrast, IFN at concentrations below 10,000 units had no effect on the production of VEGF by the RCC cells (Fig. 1C).

Next, we investigated the ability of YM529 and human and mouse IFNs to inhibit the growth of eight human renal cancer cell lines and RENCA by the MTT assay. YM529 inhibited the growth of these cells in a time- and dose-dependent manner (Fig. 2A). The IC₅₀ values of YM529 with the nine RCC cell lines are summarized in Table 1. These results indicate that bisphosphonates can inhibit the growth of RCC cells as well as other cancer cells.

Bisphosphonates are known to augment the effects of several anticancer drugs (6, 7) but it has not yet been reported whether bisphosphonates enhance the anticancer effect of IFN. Thus, we investigated the effect on in vitro RCC cell growth of combining YM529 and IFN. The data are plotted as combination index (CI) versus the fraction affected (Fa). At growth inhibition levels exceeding Fa 0.2 and 0.4, YM529 plus IFN acted synergistically (CI <1.0) with two renal cancer cell lines, ACHN and RENCA, and acted additively with CCFRC-1 cells (Fig. 2B). At Fa 0.50 and 0.80, the CIs of ACHN, CCFRC-1, and RENCA were 0.681 ± 0.172 and 0.880 ± 0.161, 0.806 ± 0.209 and 0.910 ± 0.422, and 0.481 ± 0.0763 and 0.604 ± 0.0596, respectively. These results suggest that at least, combining YM529 and IFN does not reduce their respective anticancer effects and that they can act synergistically to inhibit the growth of RCC cells.

In vivo growth of the RENCA^Luc cells can be monitored by their extracorporeal bioluminescence, as we have observed an excellent correlation between the cancer cell bioluminescense and the tumor volume (8). We used this model to investigate the effect of YM529 alone or in combination with IFN on the in vivo growth of the tumor cells. The photon counts of the mice were measured every week and the average of the groups was plotted (Fig. 3A). On day 21 after cell implantation, when we sacrificed all of the mice humanely, we found that combining IFN and YM529 significantly inhibited the growth of the RENCA^Luc cells compared with the growth of these cells in the untreated mice (Fig. 3B). Furthermore, this combination significantly reduced the serum VEGF levels (Fig. 3C).

We examined the proportion of blood cells that consists of α/β and γ/δ T lymphocytes in the RENCA tumor-bearing mice after 2 weeks of treatment. The α/β T lymphocytes increased in the IFN and IFN plus YM529 treatment groups but this was not statistically significant (Fig. 3D). The γ/δ T-lymphocyte levels also did not alter between the different groups (Fig. 3E). Thus, it seems that these agents do not alter the proportion of α/β or γ/δ T lymphocytes in the blood, which suggests that YM529 augments the direct growth inhibitory effect of IFN as well as indirect effects by reducing the elaboration of VEGF by the tumor.

After administering bisphosphonates, their concentrations in the kidney are substantial because it is the only eliminating organ for the drugs (14). Although the pharmacokinetics and pharmacodynamics of bisphosphonates may differ between humans and mice, this observation suggests that bisphosphonates may be of value in treating the primary renal lesion. To investigate this, we established a renal orthotopic model by implanting 1 × 10⁵ RENCA^LUC cells into the left renal subcapsule. Shortly after the implantation, bioluminescence was detected in all of the mice. The photon emissions were not detectable the following day but were detected again 1 week later, after which the photon emissions increased substantially over time (Fig. 4A).

Using this orthotopic mouse model, we investigated the effect of YM529 alone or in combination with IFNα/β on the in vivo growth of the RENCA^Luc cells. As with the s.c. model, YM529 did not significantly inhibit the tumor cell growth on its own. However, it did augment the growth inhibitory effect of IFN (Fig. 4B). Figure 4C shows that by day 21, the inhibitory effect of combining IFNα/β and YM529 was statistically significant compared with the untreated mice.

To investigate the safety of this combination, we examined the body weight during the treatment period and the serum concentrations of aspartate aminotransferase, alanine aminotransferase, lactate dehydrogenase, total protein, creatinine, blood urea nitrogen, and calcium. Combining the treatments did not have an adverse effect on body weight or any of the serum values tested (data not shown).

This study shows that the third-generation bisphosphonate YM529 has a direct effect on the in vitro proliferation of RCC cells and that it augments the effects of IFN on RCC cells both in vitro and in vivo. IFN has been widely used in immunotherapy for RCC (2, 3), but there is still a paucity of effective medical treatments for RCC, which is one of the reasons RCC is the most lethal urologic malignancy (1, 2). To improve the current therapies for this disease, clinical trials with several new attractive agents such as anti-VEGF antibody bevacizumab (15), the small molecule inhibitor of VEGF-mediated signaling SU5416 (16), and thalidomide (3, 17) have been carried out. However, a well-established medical therapy for patients with RCC with or without IFN is still lacking.

We have focused on bisphosphonate as a therapeutic partner for IFN because, as with other cancer cells, it may exert a direct anticancer effect on RCC as well as act as an immunomodulator that induces γ/δ T lymphocytes (10) and prevents angiogenesis (12). It is known that bisphosphonates inhibit protein prenylation in various cancer cells (5–9), but the effects of bisphosphonates on RCC have not yet been reported. Here we found that YM529 blocked the prenylation of Rap1A in ACHN and RENCA cells (human and murine RCC cell lines, respectively; see Fig. 1) and inhibited the in vitro growth of RCCs (Fig. 2A) in a dose- and time-dependent manner. This indicates that YM529 may be a potent anti-RCC agent.

When we examined a broader range of RCC lines, we found that YM529 inhibited the in vitro growth of five of eight human RCC lines and the one mouse RCC line that we evaluated with IC₅₀ values ranging from 7 to 50 μmol/L (Table 1). However, three of the human lines required more than 50 μmol/L YM529 to inhibit their in vitro growth. One multidrug resistance mechanism of RCC is the cellular overproduction of P glycoprotein, which acts as an efflux pump for various anticancer drugs (18). Differences in the sensitivity of the RCC lines to YM529 may depend on their bcl-2, P glycoprotein, or farnesyl pyrophosphate synthase activities. However, our previous studies indicated that there is no correlation between sensitivity to zoledronic acid and bcl-2 expression in small-cell lung cancer cell lines and that the growth inhibitory effect of zoledronic acid does not involve the P glycoprotein–related multidrug resistance system (5, 6). Salmo et al. have reported that bisphosphonate-resistant cells show increased farnesyl pyrophosphate synthase activity without up-regulation of its gene transcription (19). Therefore, it may be that the differences in RCC line sensitivity to YM529 that we observed is due to varying farnesyl pyrophosphate synthase activity.

When we assessed the effect on in vitro RCC cell growth of combining YM529 with IFN, we found that it had synergic or at least additive antiproliferative effects on two human cell lines, ACHN and CCFRC-1, and the murine cell line, RENCA (Fig. 2B). Therefore, we investigated the growth inhibitory effect of YM529 in combination with IFN in two mouse models, namely, a s.c. implanted model and a renal orthotopic model. In both models, treatment with YM529 or IFN alone yielded a marginal anticancer effect. However, combining IFN and YM529 had a significant antiproliferative effect in both models (Fig. 3B  and 4C). The low efficacy of YM529 alone may be because it is difficult to achieve therapeutically effective serum concentrations of YM529 in vivo. Supporting this is the fact that bisphosphonate has a high affinity for mineralized bone and rapidly localizes to the bones. Moreover, a previous study evaluating the efficacy of zoledronic acid in treating osteoporosis found that the peak serum concentrations were in the range of 1 to 3 μmol/L and were maintained for only a few hours, which indicates that the serum concentrations needed for effective anticancer activity may be difficult to achieve (20). Nevertheless, after the bone and bone marrow, the kidney is the next major site of bisphosphonate accumulation, and consequently, the concentration of bisphosphonate in the kidney is higher than it is in serum (20). In addition, Tassone et al. (21) have reported that zoledronic acid–mediated apoptotic death in pancreatic cancer cells is induced after only a 30-minute pulse exposure, which indicates that continuous drug exposure is not required for this proapoptotic property of bisphosphonates. Because YM529 in combination with IFN is effective in reducing RCC cell growth in vitro and in vivo, it seems that the peak serum concentration that is achieved with our YM529 treatment regimen and the sustained period of YM529 exposure can be enough to augment the effect of IFN against RCC cell growth in vivo.

The cumulative YM529 concentration that we used to treat the in vivo RCC models (80 μg/kg thrice a week for 2 weeks) was several times higher than the recommended dose for osteoporosis, which is 4 mg given as a 15-minute i.v. infusion at intervals of 3 to 4 weeks (22). Treatment of patients with high levels of bisphosphonates can cause renal toxicity (23, 24). However, Chen et al. (25) have reported that the type, frequency, or severity of adverse events and changes from baseline vital signs and clinical laboratory variables are unrelated to the dose of zoledronic acid (4, 8, and 16 mg) or the pharmacokinetic parameters that they examined. In the present study, we examined the body weight of the mice and their serum variables. We did not detect any harmful effects, even in the mice that were treated with a combination of YM529 and IFN. Thus, we speculate that the dosage we used to treat the RCC-grafted mice may be safe for patients with RCC as well.

We also investigated the effect of YM529 on production of VEGF by the RCCs and the T-lymphocyte subset levels in the blood to determine how YM529 imposed its anticancer effects. We did not detect an inducing effect on γ/δ T lymphocyte numbers in YM529-treated mice, but we did find that combining YM529 and IFN depleted the VEGF levels in the serum. VEGF is an important therapeutic target, as described above (3, 15–17). These observations suggest that the growth inhibitory effect of YM529 plus IFN in vivo is due to direct synergism between the antiproliferative activities of the two drugs in combination with an indirect effect, namely, the reduction of VEGF levels.

In conclusion, YM529 is a third-generation bisphosphonate with anticancer activity against RCC that acts synergistically with IFN both in vitro and in vivo. This indicates that IFN plus YM529 may be a promising therapeutic strategy for RCC. Recently, the U.S. Food and Drug Administration approved several bisphosphonates for treating not only osteoporosis but also cancer-related bone complications (26). The efficacy and safety of IFN plus bisphosphonate as a therapy for RCC should be verified by early-phase clinical trials.

We thank Yoko Nakagawa for her skillful technical assistance.