Macrophages as potential targets for zoledronic acid outside the skeleton—evidence from in vitro and in vivo models

Abstract

Purpose

Multiple cell types of the tumour microenvironment, including macrophages, contribute to the response to cancer therapy. The anti-resorptive agent zoledronic acid (ZOL) has anti–tumour effects in vitro and in vivo, but it is not known to what extent macrophages are affected by this agent. We have therefore investigated the effects of ZOL on macrophages using a combination of in vitro and in vivo models.

Methods

J774 macrophages were treated with ZOL in vitro, alone and in combination with doxorubicin (DOX), and the levels of apoptosis and necrosis determined. Uptake of zoledronic acid was assessed by detection of unprenylated Rap1a in J774 macrophages in vitro, in peritoneal macrophages and in macrophage populations isolated from subcutaneously implanted breast cancer xenografts following ZOL treatment in vivo.

Results

Exposure of J774 macrophages to 5 μM ZOL for 24 h caused a significant increase in the levels of uRap1A, and higher doses/longer exposure induced apoptotic cell death. DOX (10 nM/24 h) and ZOL (10 μM/4 h) given in sequence induced significantly increased levels of apoptotic cell death compared to single agents. Peritoneal macrophages and macrophage populations isolated from breast tumour xenografts had detectable levels of uRap1A 24 h following a single, clinically achievable dose of 100 μg/kg ZOL in vivo.

Conclusion

We demonstrate that macrophages are sensitive to sequential administration of DOX and ZOL, and that both peritoneal and breast tumour associated macrophages rapidly take up ZOL in vivo. Our data support that macrophages may contribute to the anti-tumour effect of ZOL.

Introduction

The cells of the tumour microenvironment, including macrophages, are increasingly recognised to play a key role both in cancer development and in response to therapy. As a result, several anti-cancer drugs have been specifically developed to target the microenvironment, including the potent bisphosphonate zoledronic acid that inhibits osteoclast-mediated bone resorption and is widely used to treat cancer-induced bone disease [1]. Due to the high bone affinity, the main cellular targets of bisphosphonates are bone-resorbing osteoclasts. However, a number of studies have shown that macrophages also take up zoledronic acid, both in vitro and in vivo, through their high phagocytic activity. As a consequence, much of the early work that revealed the molecular mechanism of action of bisphosphonates was carried out using macrophages [2–6]. Subsequent studies have demonstrated that macrophages are sensitive to the most potent of these agents, zoledronic acid, and undergo apoptotic cell death following exposure in vitro [7]. The effects of bisphosphonates on macrophages are not restricted to in vitro model systems. We have demonstrated that peritoneal macrophages do take up zoledronic acid following high dose in vivo administration in mice, supporting that cells outside the skeleton also are affected by this agent [8].

The accumulating literature on the effects of bisphosphonates on macrophages has been the topic of a recent comprehensive review, highlighting the potential for an anti-tumour effect through targeting of tumour-associated macrophages (TAMs) [9]. TAMs have been demonstrated to have numerous pro-tumourigenic effects, supporting a range of processes essential for tumour progression, including angiogenesis [10–12]. A high level of TAM infiltration in tumours correlates with decreased survival, especially in breast cancer, and is associated with high tumour grade, extent of tumour necrosis and large tumour size [13]. Modifying the number and/or activity of TAMs is therefore considered a potential way of altering the tumour microenvironment to cause inhibition of tumour growth, and agents that specifically target TAMs may complement existing anti-cancer therapies. The properties of bisphosphonates that limit their action to particular cell types either in the skeleton (osteoclasts) or with high phagocytic activity (macrophages) may make them ideally suited to modify the tumour microenvironment and to inhibit tumour growth.

Nanotechnology approaches have been developed in order to increase the bioavailability of zoledronic acid (ZOL) outside the skeleton, including ZOL-containing self-assembly PEGylated nanoparticles (NPs) or ZOL-encapsulating PEGylated liposomes (LIPO-ZOL). Administration of NPs was shown to reduce the number of TAMs, but not macrophage numbers in the liver, in a prostate cancer xenograft model [14]. Administration of free ZOL (3×/week for 3 weeks) had no effect on either tumour growth or TAM number in this study. The same group has shown that administration of liposome encapsulated ZOL (3× week for 3 weeks) also significantly reduces tumour growth and tumour vascularisation compared to the free drug in this extra-skeletal prostate cancer model [15].

Evidence supporting that zoledronic acid modifies TAMs outside the skeleton is provided by a recent in vivo study using a murine mammary carcinoma model where BALB-neuT mice received 100 μg/kg zoledronic acid weekly for 4 weeks followed by 3 weeks rest. Zoledronic acid increased both tumour-free and overall survival, as well as caused a significant reduction in tumour growth rate and mammary tumour multiplicity, compared to control [16]. Detailed histological analysis revealed that tumours isolated from mice receiving zoledronic acid had impaired TAM recruitment and infiltration, as well as reduced neo-vascularisation. This was the first demonstration that zoledronic acid treatment caused a reversal of the polarity of tumour-associated macrophages from the pro-tumourigenic M2 to the anti-tumourigenic M1 phenotype. These results suggest that it is not only changes in the total number of TAMs that determine the effects of zoledronic acid in tumours, but also their specific phenotype.

Although zoledronic acid caused a substantial delay in tumour progression, all animals eventually developed advanced disease, highlighting the need for combination therapy to further improve outcome [16]. In support of this, we have previously shown that sequential administration of zoledronic acid and doxorubicin (2 mg/kg doxorubicin followed 24 h later by 100 μg/kg zoledronic acid, weekly for 6 weeks) causes significant, prolonged growth inhibition of subcutaneously implanted breast tumour xenografts, compared to the individual agents [17, 18]. This sequence has also been shown to inhibit the development of spontaneous mammary carcinomas in the PyMT mouse model, demonstrating that zoledronic acid can affect extra-skeletal tumour growth when combined with chemotherapy [19]. We hypothesise that, in addition to causing tumour cell apoptosis, the combination of doxorubicin and zoledronic acid may also affect tumour macrophages, thereby inhibiting tumour angiogenesis and growth. However, no study to date has conclusively shown that tumour associated macrophages take up sufficient amounts of zoledronic acid to inhibit protein prenylation following a single, clinically achievable dose, either alone or combined with chemotherapy agents. In order to address this, we have investigated the effects of clinically relevant doses of zoledronic acid, alone and in combination with doxorubicin, on macrophages in vitro and in vivo. Our results support that macrophages are targeted by zoledronic acid, and that this may contribute to the observed anti-tumour effects in vivo.

Materials and methods

Reagents

Zoledronic acid (ZOL) ([(1-hydroxy-2-(1H-imidazol-1-yl) ethylidene] bisphosphonic acid) was supplied as the hydrated di-sodium salt by Norvartis Pharma AG, Basel, Switzerland) Switzerland. Doxorubicin (DOX) was supplied by TEVA, Leeds, UK and Geranylgeraniol (GGOH) (all trans-3,7,11-15-Tetramethyl-2,6,10,14-hexadecatetraen-1-ol) by Sigma-Aldrich.

Cell lines and tissue culture

J774.2 macrophages were obtained from the European Collection of Animal cell Cultures and cultured in DMEM (Gibco, Invitrogen Corporation, Paisley UK), supplemented with 10 % Foetal Calf Serum (FCS) and 1 % penicillin-streptomycin.

MDA-G8 cells were generated by transfection of MDA-MB-436 breast cancer cells (European Collection of Animal cell Cultures) with enhanced green fluorescent protein (eGFP) as previously described [16]. Cells were cultured in RPMI medium supplemented with 4,500 ml/L glucose (Invitrogen-Gibco, Paisley, UK) and with 10 % FCS.

Isolation of TAMs from MDA-G8 xenografts

5 × 10⁵ MDA-G8 cells were inoculated subcutaneously into 10-week old female Balb/C nu/nu mice (n = 15, Harlan, UK). Experiments were carried out with the UK Home Office approval under project licence 40/3531, University of Sheffield, UK. Tumour size was measured twice per week using callipers and mice were sacrificed by cervical dislocation once a tumour measured 1,000 mm³. Tumours were excised and minced using a scalpel blade and transferred into pre-warmed dissociation buffer (20 ml IMDM Invitrogen-Gibco, Paisley, UK, 40 mg dispase, Invitrogen-Gibco, Paisley, UK, 0.2 mg/ml collagenase Sigma, UK, 1 ml DNase, Sigma, UK) for 30 min at 37 °C with agitation. 10 % FCS was then added to the mixture before filtering, followed by 2 washes with PBS. PBS containing 7.5 μl of each antibody (anti-CD45, cat no 558108 from Becton Dickinson, anti-CD11b cat no 301317, anti-Ly6G cat no 127615 and F4/80 cat no 123113, all from Biolegend) was added to the cell pellet and left at 4 °C for 30 min. The sample was washed with PBS and resuspended and the populations sorted using a BD FACS Aria flow cytometer (BD Bioscience). Tumour-associated macrophages (TAMs) were identified according to the gating strategy shown in Fig. 7. First, the pan-leukocyte marker CD45 was used to identify immune cells. Second, expression of CD11b was used to subdivide the CD45+ leukocytes into non-myeloid and myeloid populations. This myeloid subpopulation was further subdivided using the granulocyte marker Ly6G. Finally, macrophages were identified on the basis of their expression of F4/80. The complete marker profile for the identification and isolation of TAMs was CD45+ CD11b+ Ly6G—F4/80+. After sorting, 5 % FCS was added to the cells, they were then pelleted and resuspended in lysis buffer for subsequent detection of unprenylated Rap1A by western blot analysis.

Fig. 7.

Macrophage content in the subcutaneous tumour xenograft and gating strategy for isolation of Tumour-Associated macrophages. Tumour macrophage content (a): 5 × 10⁵ MDA-G8 cells were injected subcutaneously into 10-week old female Balb/C nu/nu mice and left to grow for 4 weeks. Histological tumour sections were prepared and stained using an antibody specific for the macrophage marker F4/80 (brown stain). Gating strategy (b): First, the pan-leukocyte marker CD45 was used to identify immune cells. Second, expression of CD11b was used to subdivide the CD45+ leukocytes into non-myeloid and myeloid populations. This myeloid subpopulation was further subdivided using the granulocyte marker Ly6G. Finally, macrophages were identified on the basis of their expression of F4/80. The complete marker profile for the identification and isolation of TAMs was CD45+ CD11b+ Ly6G—F4/80+

Effects of zoledronic acid on macrophages in vivo

To determine the effects of ZOL on protein prenylation in vivo, female Balb/c nude mice were treated with 100 μg/kg ZOL (s.c.), either alone or 24 h after 2 mg/kg DOX (i.v.). Mice were sacrificed either 24 h or 2 weeks after ZOL administration and peritoneal cells obtained by peritoneal flushing. In separate experiments, animals bearing subcutaneous breast tumour xenografts were given a single administration of 100 μg/kg ZOL (n = 10) or saline (n = 5) 24 h prior to sacrifice and tumour macrophage isolation. Due to the low number of tumour-associated macrophages (TAMs), samples from each group were pooled to yield sufficient cells for detection of uRap1A by western blotting.

Effects of zoledronic acid on macrophages in vitro

J774.2 macrophages were exposed to 5–100 μM ZOL for 24 h; 10–100 nM DOX for 24 h; or 1–10 nM DOX for 24 h, followed by 48 h incubation in drug free media. For combination experiments, macrophages were exposed to either PBS control or the following for 24 h; 10 μM ZOL, 10 nM DOX, 10 μM ZOL and 10 nM DOX simultaneously, 10 μM ZOL followed by 10 nM DOX, 10 nM DOX followed by 10 μM ZOL. Following each treatment, media were removed and replaced with subsequent drug (or drug-free medium) to give a total incubation time of 72 h. In separate experiments the ZOL exposure time was reduced to 4 h and the DOX concentration to 2.5 nM. For reversal experiments using GGOH, J774.2 cells were exposed to 10 μM ZOL (4 h), 2.5 nM DOX (24 h), alone or in combination in the absence or presence of 10 μM GGOH (4 h). Following treatment completion cells were incubated in fresh media for the remainder of the 72 h.

F4/80 immunohistochemistry

Macrophage content in subcutaneous G8 xenografts was assessed on histological slides. Antigen retrieval was done with trypsin (Menarini Diagnostics; MP-TRP25) for 10 min at room temperature followed by a serum block (NGS; Vector S-1000) overnight at 4 °C. Slides were incubated with the primary antibody (mouse anti-F4/80, Serotec MCA487R) for 1 h at ambient temperature. After washing in TBS for 5 min, slides were incubated with 3 % H₂O₂ for 30 min at room temperature to block endogeneous peroxidase. Slides were incubated with the secondary antibody (biotinylated anti-rat antibody, Vector Laboratories PK-6100) for 1 h. Slides were washed for 5 min in TBS and then incubated with SA-HRP (Dako Cytomation; DK-2600) for 30 min at ambient temperature followed by DAB chromagen (Vector; SK-4100). The slides were rinsed in tap water and counterstained with Gill’s haematoxylin.

Detection of unprenylated Rap1a

Unprenylated Rap1A was detected by western blotting using a goat polyclonal antibody (SC-1482, Santa Cruz, Biotechnology Inc, USA) at 1:200. Secondary antibody was a polyclonal rabbit anti-goat (P0449, Dako Cytomation, Glostrup, DK). GAPDH mouse monoclonal antibody (AB9485, Abcam, Cambridge, UK) and RNA Polymerase mouse monoclonal antibody (Abcam) were followed by HRP-linked sheep anti-mouse IgG (HP979NA, GE Healthcare UK Limited, UK).

Apoptosis assay

Following completion of treatment, cells were incubated with Hoechst 33342 (Sigma-Aldrich, Poole, UK) and PI (propidium iodide, Molecular Probes, Cambridge, UK). The percentage of viable, apoptotic and necrotic cells was quantified by counting five fields using an inverted Leica DMIRB microscope fitted with a Whipple graticule. Each experiment was repeated 3 times in triplicate.

Statistical analysis

Statistical analysis was performed using Graph Pad Prism software, version 5. For comparison between means a one-way analysis of variance (ANOVA) was used as well as Tukey’s multiple comparison post-hoc analysis for comparison between treatment groups. P < 0.05 was considered a significant result.

Results

The effects of doxorubicin and zoledronic acid on J774.2 macrophage apoptosis

We first determined the effects of ZOL and DOX alone on J774.2 macrophages in order to establish the appropriate doses and incubation times for use in subsequent combination studies. Cells were treated with 5–100 μM ZOL for 24 h, and the percentage of viable, apoptotic and necrotic cells quantified. Figure 1 shows that 10–100 μM ZOL induced significantly increased levels of apoptosis compared to control. Only the two highest doses of ZOL used, 50 and 100 μM, induced a significant increase in necrosis compared to control (10.3 % and 13.9 % v 5.4 % in control, p < 0.05, data not shown).

Fig. 1.

Effect of zoledronic acid on apoptosis in J774.2 macrophages. J774.2 macrophages were treated with 5–100 μM zoledronic acid for 24 h. The percentage of apoptotic cells was determined by evaluation of nuclear morphology following staining with Hoechst and PI. The experiment was repeated 3 times in triplicate. The data are expressed as % cells ± SEM (n = 3), and analysis was done using 1-way ANOVA with Tukey’s post-hoc test. Asterisks indicate the statistical significance when compared with the control (***p < 0.001, **p < 0.01, *p < 0.05)

To determine the effects of doxorubicin, J774.2 macrophages were incubated either with 10–100 nM DOX for 24 h (Fig. 2a) or with 1–10 nM DOX for 24 h, followed by 48 h in drug-free media (Fig. 2b). Compared to control, all doses of DOX induced significant increases in apoptosis at 24 h following continuous drug exposure (5.9 %, 10.7 %, 15.4 % v 1.1 %, p < 0.001). Lower doses of DOX (1, 2.5 and 5 nM for 24 h) also induced significant increases in apoptosis detected at 72 h compared to control (2.2 %, 4.3 % and 7.3 % v 0.5 %, p < 0.0001). Exposure to 5 and 10 nM DOX caused significant increases in macrophage necrosis (5.2 % and 6.9 % v 2.7 %, p < 0.001, data not shown). Based on these results, 10 μM ZOL for 4 h/24 h and 10/2.5 nM DOX/24 h were used in subsequent combination therapy experiments.

Fig. 2.

Effect of doxorubicin on apoptosis in J774.2 macrophages. J774.2 macrophages were treated with 10–100 nM doxorubicin for 24 h (a) or 1–10 nM doxorubicin for 24 h and then incubated in drug-free media for 48 h (b). The percentage of apoptotic macrophages was determined by evaluation of nuclear morphology following staining cells with Hoechst and PI. The experiment was repeated 3 times in triplicate. The data are expressed as % cells ± SEM (n = 3), and analysis was done using 1-way ANOVA with Tukey’s post-hoc test, asterisks indicate the statistical significance when compared with the control (**p < 0.01, ***p < 0.001)

Effect of doxorubicin and zoledronic acid in combination

Initial combination experiments used 10 μM ZOL and 10 nM DOX, the doses at which each drug was found to induce low but significant levels of apoptosis (Fig. 3a). To investigate whether the drugs induced increased levels of apoptotic cell death when used together, J774.2 macrophages were incubated with 10 μM ZOL or 10 nM DOX for 24 h, alone, in combination or in sequence, and apoptosis levels were assessed at 72 h. DOX followed 24 h later by ZOL caused the highest levels of macrophage apoptosis (43.4 %) compared to the reverse sequence (36.1 % p < 0.001), the drugs given together (23.2 %, p < 0.001) and ZOL or DOX alone (19.4 % and 5.5 %, p < 0.001). These results show that the order in which the drugs are given determines the level of cell death.

Fig. 3.

Effect of zoledronic acid and doxorubicin in combination on macrophage apoptosis. J774.2 macrophages were treated with: 10 μM zoledronic acid for 24 h/10 nM of doxorubicin for 24 h (a); 10 μM zoledronic acid for 4 h/10 nM of doxorubicin for 24 h (b); 10 μM zoledronic acid for 24 h/2.5 nM of doxorubicin for 24 h (c), alone, in combination or in sequence. Cells were then incubated in fresh drug-free media up to 72 h and the percentages of apoptotic cells were quantified by assessment of nuclear morphology following staining with Hoechst and PI. 3 experiments were carried out in triplicate. The data are expressed as % cells ± SEM (n = 3), and analysis was done using 1-way ANOVA with Tukey’s post-hoc test, asterisks indicate the statistical significance between treatment groups (***p < 0.001, **p < 0.01, *p < 0.05 and ns p > 0.05)

As ZOL has a short plasma half-life, we also investigated the effects of short time exposure to ZOL in combination with DOX. J774.2 macrophages were exposed to 10 μM ZOL for 4 h, 10 nM DOX for 24 h, either alone, in combination or sequentially (Fig. 3b). Again, DOX followed by ZOL caused the highest levels of macrophage apoptosis (44.2 %), compared to the reverse sequence, the drugs given in combination, or DOX and ZOL alone (39 %, 30 % 26.7 % and 3.6 %, p < 0.001). This showed that macrophage apoptosis was induced even when cells were exposed to ZOL for a short time period (4 h), hence continuous drug exposure for 24 h is not required.

We also investigated whether combination therapy with DOX and ZOL remained effective if the concentration of DOX was lowered from 10 to 2.5 nM. J774.2 macrophages were exposed to 10 μM ZOL for 4 h or 2.5 nM DOX for 24 h, either alone, in combination or sequentially (Fig. 3c). Following treatment completion cells were incubated in fresh media for the remainder of the 72 h. As expected, DOX followed by ZOL induced the highest levels of macrophage apoptosis compared to the reverse sequence, DOX and ZOL together, or DOX and ZOL alone (16.6 %, 8.9 %, 6.6 %, 4.7 % and 2.5 %., respectively p < 0.001). These results demonstrate that even at lower concentrations there is significant additional induction of macrophage apoptosis caused by administration of DOX and ZOL in sequence.

Inhibition of protein prenylation contributes to the increased levels of J774.2 macrophage apoptosis induced by combination therapy

To confirm that the increased levels of macrophage apoptosis caused by sequential DOX and ZOL treatment involved ZOL-induced inhibition of protein prenylation, we used geranylgeraniol (GGOH), an intermediate metabolite in the mevalonate pathway that reverses the effects of ZOL. J774.2 macrophages were incubated with 2.5 nM DOX (24 h) followed 24 h later by 10 μM ZOL (4 h), with or without 10 μM GGOH. Combining ZOL with GGOH has been shown to reverse ZOL-induced apoptosis in breast cancer cells and prostate cancer cells [20, 21]. Figure 4 shows that DOX followed by ZOL caused 13.7 % apoptosis, significantly more than DOX or ZOL alone (5.3 % and 2.6 %, p < 0.001), as expected. The addition of GGOH reversed the level of apoptosis to the same as induced by DOX alone (6 % vs 5.3 %, ns). These results demonstrate that ZOL-induced inhibition of the mevalonate pathway is a major contributor to the combination effect of sequential DOX/ZOL treatment.

Fig. 4.

The effect of geranylgeraniol (GGOH) on zoledronic acid-induced apoptosis. J774.2 macrophages were treated with either 10 μM zoledronic acid alone for 4 h. 2.5 nM of doxorubicin alone for 24 h, 10 μM GGOH alone for 4 h or 2.5 nM doxorubicin followed by 10 μM zoledronic acid, either alone or in combination with GGOH. Following treatment completion, cells were incubated in fresh medium for the remainder of the 72 h. The percentages of apoptotic cells were quantified by assessment of nuclear morphology following staining with Hoechst and PI. 3 experiments were carried out in triplicate. The data are expressed as % cells ± SEM (n = 3), and analysis was done using 1-way ANOVA with Tukey’s post-hoc test, asterisks indicate the statistical significance between treatment groups (***p < 0.001)

The effect of zoledronic acid and doxorubicin on protein prenylation in vitro

ZOL causes apoptotic cell death by inhibiting key enzymes of the mevalonate pathway, responsible for the prenylation of small GTP-binding proteins such as Rap1A. Accumulation of unprenylated Rap1A in the cytosol is therefore a commonly used surrogate measure of ZOL uptake. We investigated the effects of ZOL, alone or in combination with DOX, on Rap1a prenylation in J774.2 macrophages (Fig. 5). Cells were incubated with 10 μM ZOL (24 h) or 2.5 nM DOX (4 h) either alone, in combination or in sequence. Following treatment completion, cells were incubated in fresh media for the remainder of the 72 h. The presence of unprenylated Rap1A was assessed by western blot using an antibody specific for the unprenylated form of Rap1A. Treating the cells with DOX followed by ZOL did not appear to cause a greater inhibition of Rap1A prenylation than ZOL treatment alone, implying DOX does not increase the uptake of ZOL by J774.2 macrophages. However, of the combination schedules, DOX followed by ZOL 24 h later induced the greatest accumulation of unprenylated Rap1A

Fig. 5.

Effect of zoledronic acid, alone or in combination with doxorubicin, on Rap1A prenylation in J774.2 macrophages in vitro. J774.2 macrophages were treated with PBS (control), 10 μM zoledronic acid for 24 h alone or in combination/sequence with 2.5 nM doxorubicin for 24 h as indicated. Following treatment completion, cells were incubated in fresh media until the end of 72 h. Cells were lysed for detection of accumulation of unprenylated Rap1A by western blotting. Top panel shows unprenylated Rap1A (22 kD) and bottom panel GAPDH (36 kD)

Effect of zoledronic acid and doxorubicin on Rap1A prenylation in vivo

To establish that ZOL inhibits protein prenylation in macrophage populations in vivo, mice were treated with a single dose of PBS (control), 100 μg/kg ZOL, alone or 24 h after adminstration of 2 mg/kg DOX. Animals were sacrificed 24 h or 2 weeks after the last treatment and peritoneal cell populations collected by peritoneal flushing. Figure 6 shows that as expected, inhibition of protein prenylation was clearly detectable by accumulation of uRap1A in peritoneal cells 24 h after a single injection of ZOL, alone or following DOX, whereas uRap1A was undetectable in cells isolated from the saline treated control animals. Importantly, uRap1A remained detectable in peritoneal cells even 14 days following a single sequential administration of DOX and ZOL.

Fig. 6.

Effect of zoledronic acid, alone or in combination with doxorubicin, on Rap1A prenylation in peritoneal cells in vivo. Mice were treated with PBS (control) or 100 μg/kg zoledronic acid either alone, or 24 h after 2 mg/kg doxorubicin treatment. Animals were sacrificed 24 h or 2 weeks after treatment, peritoneal cell populations collected by flushing and lysed for detection of accumulation of unprenylated Rap1A (22 kDa) by western blotting using specific antibodies

Effect of zoledronic acid on Rap1A prenylation in tumour macrophages in vivo

We next aimed to establish whether tumour associated macrophages (TAMs) take up sufficient levels of ZOL for prenylation to be affected following administration of a single, clinically achievable dose in vivo. Animals bearing subcutaneous human breast cancer xenografts were treated with 100 μg/kg ZOL (n = 10) or saline (n = 5) 24 h prior to sacrifice. Histological analysis of the tumours showed a sigificant content of F4/80+ TAMs (Fig. 7a). TAMs isolated from the pooled tumours were sorted by flow cytometry as described in Section 2 and Fig. 7b, followed by detection of uRap1A by western blotting. Peritoneal macrophages were also collected to confirm uptake of ZOL. In previous studies we have been unable to detect increased levels of uRap1A in breast tumour xenograft lysates, even after prolonged and repeated ZOL treatment [17]. As shown in Fig. 8, uRap1A was detected both in the peritoneal macrophages and in the TAMs from ZOL treated animals, but not from saline treated controls. To our knowledge this is the first demonstration that a single administration of a clinically relevant dose of ZOL reaches peripheral tumours at a sufficient concentration to modify prenylation in cells of the tumour microenvironment.

Fig. 8.

Detection of unprenylated of Rap1A peritoneal cells and tumour macrophages following a single dose of zoledronic acid in vivo. MDA-G8 breast tumour bearing mice were administered PBS (control, n = 5) or 100 μg/kg zoledronic acid (n = 10) and sacrificed 24 h after treatment. Macrophages were obtained via peritoneal flushing and tumour macrophages isolated by flow cytometry. Cells were then lysed for detection of accumulation of unprenylated Rap1A (22 kDa) and GAPDH (37 kDa) by western blotting using specific antibodies

Discussion

To our knowledge, this is the first study to demonstrate that macrophages take up sufficient amounts of zoledronic acid for protein prenylation to be significantly reduced, after administration of a single, clinically relevant, dose in vivo. We show inhibition of protein prenylation, in both peritoneal cell populations and tumour-associated macrophages, 24 h after drug injection, supporting that this bone-targeted agent is rapidly taken up by phagocytic cells outside the skeleton. When zoledronic acid was given in combination with the chemotherapy agent doxorubicin, inhibition of protein prenylation remained detectable in peritoneal cells 2 weeks later. Our data support that following uptake by macrophages, zoledronic acid continues to inhibit prenylation for a considerable period, despite the short half-life of the agent in the circulation. This is the first study to directly demonstrate inhibition of prenylation in tumour-associated macrophages by a single dose of zoledronic acid in vivo, providing preliminary evidence for a potential anti-tumour effect via this mechanism.

Cells of the host microenvironment, including macrophages, play a significant part in both tumour development and response to therapy [10, 11]. This may be particularly important in breast cancer, where macrophages often constitute a significant proportion (up to 50 %) of the tumour mass [13]. The complex interplay between tumour cell and the host-derived stroma during tumour progression is the focus of intense research, and it is increasingly accepted that successful anti-cancer therapy will involve combinations of agents that target the host microenvironment as well as the cancer cells. This has long been the strategy in treatment of breast cancer bone metastasis, using a combination of chemotherapy, endocrine therapy and biological agents that target the tumour cells, with bisphosphonates that target the bone microenvironment [1]. However, whether bone-targeted agents (in particular zoledronic acid) have direct or indirect anti-tumour effects also outside the skeleton has been the subject of much debate. Due to the rapid homing to bone, the main target of the bisphosphonates is the osteoclast, which undergoes apoptotic cell death resulting in decreased bone resorption [22]. There is little evidence that other cell types, including tumour cells, are exposed to sufficient amounts of zoledronic acid to be affected by the standard 4 mg clinical administration every 3–4 weeks. A recent elegant study by Chinault et al. used a bioluminescence reporter to demonstrate that protein prenylation was not inhibited in breast cancer cells grown as xenografts, in either the mammary fat pad or in bone, following administration of 30 μg/kg/day zoledronic acid [23]. This clearly supports that tumour cells are not directly targeted through uptake of zoledronic acid in vivo, and the authors suggest that future studies should focus on identifying whether host cells, in particular those of the myeloid lineage that includes macrophages, mediate the observed anti-tumour effects. We agree that the key point to establish in relation to the anti-tumour effect of zoledronic acid outside bone is whether cells of the tumour microenvironment, including macrophages, are exposed to sufficient concentrations of zoledronic acid for protein prenylation to be inhibited. Due to their high phagocytic activity and key role in tumour progression, macrophages are frequently suggested as potential cellular targets of zoledronic acid [11]. In contrast to tumour cells (with low endocytic capacity), macrophages may take up significant amounts of zoledronic acid even during relatively short exposure to the drug. In the present study, we have therefore investigated the effects of short term/low dose zoledronic acid on macrophages in vitro, and whether a single, clinically relevant dose of this agent reaches macrophage populations outside bone in vivo. We show that exposing J774.2 macrophages to 10 μM zoledronic acid for 4 h in vitro was sufficient to increase levels of apoptosis and induce accumulation of unprenylated Rap1A compared to control, supporting that the cells do take up significant amounts of the drug even during a short pulse of treatment. This is highly relevant, as zoledronic acid is only given once every 3–4 weeks in the treatment of cancer-induced bone disease, and remains in the circulation at around 1–2 μM for only 2 h.

In marked contrast to previous studies that have used high doses and/or prolonged and often repeated exposures to zoledronic acid [19, 24, 25], we are the first to investigate the effects on macrophage populations of a single in vivo administration of 100 μg/kg, equivalent to the 4 mg dose used to treat cancer-induced bone disease [18]. We found that this dose was sufficient to cause accumulation of unprenylated Rap1A in peritoneal cell populations at 24 h, and this remained easily detectable 2 weeks later. Our results demonstrated that similarly to the in vitro setting, there is rapid uptake of zoledronic acid by peritoneal cell populations (consisting of 15–20 % F4/80 positive cells assumed to be macrophages) following a single dose in vivo, and that prolonged/repeated exposure to the drug is not required to inhibit prenylation as the effect persisted for at least 2 weeks.

Several studies have reported effects of zoledronic acid on tumour-associated macrophages (TAMs) in different tumour types in vivo, but all used repeated dosing regimens that resulted in a high total cumulative dose of zoledronic acid being delivered. Coscia and colleagues reported that 100 μg/kg zoledronic acid given weekly for 4 weeks followed by 3 weeks rest (average of 16 injections) increased both tumour-free and overall survival, as well as caused a significant reduction in tumour growth rate and mammary tumour multiplicity, compared to control, in a BALB-neuT mouse mammary carcinoma model [16]. Tumours isolated from mice treated with zoledronic acid had impaired TAM recruitment and infiltration, reduced neo-vascularisation, and this was the first report that zoledronic acid treatment caused a change in the polarity of tumour-associated macrophages from the pro-tumourigenic M2 to the anti-tumourigenic M1 phenotype. In contrast, we found no effect of 6 weeks of weekly administration of 100 μg/kg on the number of macrophages infiltrating tumours in the PyMT mouse model of mammary carcinoma, but macrophage polarity was not assessed [19]. Reduced numbers of tumour macrophages have also been reported following repeated administration of zoledronic acid in models of hepatocellular carcinoma (100 μg/kg/day for 5 weeks) [25], mammary carcinoma (100 μg/kg/5 days per week for 16–24 weeks) [24] and mesothelioma (100 μg/kg/daily for 25 days) [26]. Taken together, these results suggest that macrophages may be modified by zoledronic acid, and thereby contribute to a potential anti-tumour effect. However, none of the studies published to date have demonstrated that zoledronic acid is taken up by tumour macrophages in vivo at sufficient concentrations to inhibit protein prenylation. We therefore carried out a pilot study aiming to detect the presence of unprenylated Rap1A in tumour macrophages isolated from breast cancer xenografts 24 h after administration of a single dose of 100 μg/kg zoledronic acid. By pooling TAMs isolated from 15 tumours we were able to demonstrate that the macrophage population did have detectable levels of uRap1A following zol treatment in vivo, providing the first direct evidence that this agent reaches macrophages in extra-skeletal tumours at sufficient concentrations to inhibit the mevalonate pathway. We were unable to separate M1 and M2 macrophages using this method, and future studies will clarify whether a single dose of zoledronic acid is sufficient to modify tumour macrophage polarisation. As human breast tumours have a high macrophage content, it would be of great interest to include zoledronic acid in a neo-adjuvant treatment protocol that would allow subsequent isolation of TAMs for detection of accumulation of uRap1a and to confirm the clinical relevance of our preliminary findings.

Zoledronic acid is commonly used in combination with other anti-cancer agents in the treatment of cancer-induced bone disease from solid tumours [1]. Studies using in vivo model systems have demonstrated that zoledronic acid inhibits breast tumour growth, both in bone and at extra-skeletal sites, when used in combination with chemotherapy agents (e.g. doxorubicin) [18, 19]. The order in which the agents were given determined the extent of the resulting anti-tumour effect, with doxorubicin followed 24 h later by zoledronic acid being the most effective schedule. The mechanisms behind this sequence-specific effect, and the cellular targets of zoledronic acid outside the skeleton, remain to be established. We show that as for tumour cells, sequential administration of doxorubicin, followed 24 h later by zoledronic acid, induced the highest level of macrophage apoptosis, compared to the reverse sequence, the drugs given in combination or alone. Induction of apoptosis was prevented by addition of an intermediary of the mevalonate pathway, demonstrating that zoledronic acid-mediated inhibition of protein prenylation contributes to the effect of sequential therapy. Pre-treatment of the macrophages with doxorubicin did not result in increased uptake of zoledronic acid, hence the mechanism behind the increase in macrophage apoptosis by this schedule remains to be established.

The exact clinical relevance of our findings will be determined in our future studies, but recent clinical trials in breast cancer have reported that addition of zoledronic acid to standard anti-cancer therapy does reduce recurrence outside bone, suggesting that bisphosphonates do modify tumour growth at peripheral sites [27]. These surprising results have renewed the interest in identifying the cellular and molecular targets modified by clinically relevant doses of zoledronic acid in vivo, including a potential effect on macrophages. Taken together, our data support that these highly versatile cells are potential targets for zoledronic acid, and that modification of tumour macrophages may contribute to the anti-cancer effect of this agent.