Sustained inhibition of tumor growth and prolonged survival following sequential administration of doxorubicin and zoledronic acid in a breast cancer model

Authors

  • Penelope D. Ottewell,

    1. Academic Unit of Clinical Oncology, School of Medicine and Biomedical Sciences, University of Sheffield, United Kingdom
    Search for more papers by this author
  • Diane V. Lefley,

    1. Academic Unit of Clinical Oncology, School of Medicine and Biomedical Sciences, University of Sheffield, United Kingdom
    Search for more papers by this author
  • Simon S. Cross,

    1. Academic Unit of Clinical Oncology, School of Medicine and Biomedical Sciences, University of Sheffield, United Kingdom
    Search for more papers by this author
  • C. Alyson Evans,

    1. Academic Unit of Clinical Oncology, School of Medicine and Biomedical Sciences, University of Sheffield, United Kingdom
    Search for more papers by this author
  • Robert E. Coleman,

    1. Academic Unit of Clinical Oncology, School of Medicine and Biomedical Sciences, University of Sheffield, United Kingdom
    Search for more papers by this author
  • Ingunn Holen

    Corresponding author
    1. Academic Unit of Clinical Oncology, School of Medicine and Biomedical Sciences, University of Sheffield, United Kingdom
    • Academic Unit of Clinical Oncology; School of Medicine and Biomedical Sciences; University of Sheffield, Beech Hill Road; Sheffield, S10 2RX, United Kingdom
    Search for more papers by this author
    • Fax: +44-0-114-271-1711


  • Conflict of interest: R.E. Coleman: Research funding, consultancy and speakers fee from Novartis. I. Holen: Research funding and speakers fee from Novartis

Abstract

Combination therapy, using agents that target the microenvironment as well as the cancer cells, is common in the treatment of advanced breast cancer. Here, we show that a 6-week course of weekly sequential administration of the cytotoxic drug doxorubicin (2 mg/kg), followed 24 hr later by the antiresorptive agent zoledronic acid (100 μg/kg), causes substantial inhibition of subcutaneous MDA-MB-436 breast tumor growth in immunocompromised mice, leading to significantly increased survival. Tumor growth did not resume following withdrawal of treatment after 6 weeks, with 60% of the animals in this group surviving for more than 160 days. In comparison, animals receiving single-agent therapy all died within 50 days. Molecular analysis of the tumors showed no effect on cell cycle or apoptosis following administration of 100 μg/kg zoledronic acid or 2 mg/kg doxorubicin alone. When doxorubicin was administered 24 hr before zoledronic acid, tumors displayed decreased expression of CYCLINS E1, B, D1 and D3 as well as CDK2, CDC2, CDK4 and CDK7, indicative of cell-cycle inhibition. Tumors from animals receiving sequential treatment also showed induction of both intrinsic- and extrinsic-apoptotic pathways, with increased expression of BAX, decreased expression of BCL-2 and activation of CASPASE 3, 8 and 9. Accumulation of the unprenylated form of RAP1a, a surrogate marker for uptake of zoledronic acid, was only detected in tumors from animals treated with doxorubicin 24 hr before zoledronic acid. Our data are the first to show a sustained antitumor effect in vivo following a limited course of sequential administration of doxorubicin followed by zoledronic acid.

Patients with advanced breast cancer may receive chemotherapeutic agents to reduce tumor growth as well as antiresorptive drugs to control cancer-induced bone disease, but the optimal way of combining these agents remains to be established. In vivo studies from different tumor types have reported increased antitumor effects in bone when cytotoxic drugs are combined with bisphosphonate zoledronic acid (reviewed in Ref.1). This has been shown by combining zoledronic acid with UHT in breast cancer models,2 with imatinib mesylate or paclitaxel in a prostate cancer model3 and with ifosfamide in an osteosarcoma model.4 However, these studies used high doses of zoledronic acid ranging from a single injection of 250–100 μg/kg administered 2× per week. In the B02 model of breast cancer bone metastases, we have shown that a clinically relevant dose of doxorubicin (2 mg/kg), combined with a single administration of zoledronic acid (100 μg/kg), reduces tumor growth in bone.5 There is emerging evidence that combination treatment with zoledronic acid may also affect tumors outside the skeleton. A recent clinical trial in breast cancer has reported that adding zoledronic acid to standard endocrine therapy leads to a reduction in recurrence at 5 years, including at visceral and locoregional sites.6 In support of this, in vivo studies of subcutaneously implanted tumors have reported that zoledronic acid combined with cytotoxic agents may reduce tumor growth depending on dosing regimens.1 This has been shown for combinations of zoledronic acid with paclitaxel/Gefitinib in prostate cancer3, 7 with paclitaxel in Ewing sarcoma8 and with Imatinib in leukemia.9 We have found that weekly treatment with 2 mg/kg doxorubicin followed 24 hr later by 100 μg/kg zoledronic acid eliminated subcutaneous breast tumor growth in mice and that sequential administration was more effective than giving the 2 drugs simultaneously.10

How doxorubicin and zoledronic acid elicit these combined antitumor effects remains to be established, but the individual mechanism of action of the drugs is well established. Doxorubicin is an anthracyclin that inhibits topoisomerase II and induces DNA double-strand breaks,11 interferes with DNA unwinding,12 induces differentiation and generates oxygen-free radicals.13 Doxorubicin does not appear to induce apoptosis in breast cancer cells in vitro, but induces growth arrest and nonapoptotic cell death accompanied by DNA damage.14–18 Zoledronic acid is a nitrogen-containing bisphosphonate, which inhibits FPP synthase, a key enzyme of the mevalonate pathway.19, 20 This pathway is ubiquitously expressed and responsible for cholesterol synthesis in eukaryotic cells as well as for posttranslational prenylation of a number of molecules including small GTPases.21 Zoledronic acid inhibits protein prenylation and induces apoptotic cell death in a range of tumor cell types in vitro (reviewed in Ref.21). In addition, zoledronic acid has been shown to reduce tumor cell proliferation by affecting genes regulating the cell cycle. In osteosarcoma cells, Ory et al.22 showed that zoledronic acid causes cell-cycle arrest in S and G2/M phases, with an increase in P-ATR, P-CHK1, WEE1 and P-CDC2 levels and a decrease in CDC25c, independent of P53 and Rb. This was accompanied by increased levels of apoptosis characterized by nuclear alterations, increased BAX expression and reduced BCL-2. Using HCT-116 colon carcinoma cells, Sewing et al.23 showed that zoledronic acid inhibited proliferation, causing G1 arrest accompanied by an induction of apoptosis via activation of caspases 3, 7, 8 and 9. Cleavage of PARP and cytochrome C release was detected as well as translocation of BAX into the mitochondria, BID activation and a decrease of overall BCL-2 expression.23 In an in vivo model of small cell lung cancer, Li et al.24 reported that zoledronic acid caused arrest of line-1 tumor cells at S/G2/ M-phases, at doses that did not induce apoptosis.

The majority of studies of antitumor effects of bisphosphonates have used short-term protocols lasting 3–6 weeks. This is the first report of a long-term (24 weeks) study showing that subcutaneous tumor growth does not resume following a 6-week course of weekly sequential treatment with doxorubicin and zoledronic acid and that this is associated with substantially increased survival. We have used pathway-specific microarrays, in combination with real-time PCR and western blotting, to identify the molecular mechanisms involved in the antitumor effects caused by the sequential treatment. Our data are the first to show that in vivo administration of clinically achievable doses of doxorubicin followed by zoledronic acid causes changes in gene expression in tumors outside the skeleton. The results suggest that sequential administration of doxorubicin followed by zoledronic acid should be evaluated in a clinical study of patients with organ-confined breast cancer.

Materials and Methods

Subcutaneous MDA-MB-436 tumor growth in vivo

MDA-MB-436 cells (0.5 × 106) were inoculated subcutaneously into 6-week-old female CD1 nu/nu mice n = 110; Charles River, Kent, UK). A total of 98 (89%) of mice inoculated developed tumors. Mice were maintained in a 12-hr light–dark cycle and given free access to food and water. Experiments were carried out with UK Home Office approval under project license 40/2343.

Mice were divided into groups of equal tumor burden (mean tumor volume =196.7 mm3) and injected once per week with (i) saline (intraperitoneal, i.p.) (n = 10), (ii) 2 mg/kg doxorubicin (intravenous. i.v., PHARMACHEMIE B.V., The Netherlands) (n = 10); (iii) 100 μg/kg zoledronic acid [i.p, (1-hydroxy-2-(1H-imidazol-1-yl) ethylidene] bisphosphonic acid)] supplied as the hydrated di-sodium salt by Novartis Pharma AG, Basel, Switzerland) (n = 10) or (iv) doxorubicin followed 24 hr later by zoledronic acid (n = 20). Following 6 weeks of treatment animals from group D were subdivided into 2 groups of equal tumor burden, groups E and F. Animals in group E continued to receive weekly administration of doxorubicin followed 24 hr later by zoledronic acid, whereas animals in group F received no further treatment for the duration of the experiment. Mice were sacrificed by cervical dislocation once tumor measured 1,000 mm3 or at the end of the experimental protocol (day 169).

For analysis of tumor gene and protein expression, a separate experiment was performed, where mice were weekly for 6 weeks with (i) saline, (ii) 2 mg/kg doxorubicin, (iii) 100 μg/kg zoledronic acid, (iv) doxorubicin and zoledronic acid at the same time, (v) doxorubicin followed 24 hr later by zoledronic acid or (vi) zoledronic acid followed 24 hr later by doxorubicin. Mice were killed by cervical dislocation 24 hr after their final treatment and the tumors excised. Half of each tumor was stored in RNAlater (Ambion, Huntingdon, UK) at −20°C before RNA extraction; the other half was placed in the cell lysis buffer [MCL1-1KT mammalian cell lysis kit (Sigma, Poole, UK)], and protein was extracted according to the manufacturer's instructions.

DNA microarray analysis

Three tumors from mice in each treatment group were pooled before mRNA extraction using a SuperArray, ArrayGrade mRNA isolation kit (tebu-bio, Peterborough, UK) and the manufacturer's instructions. Two milligrams of mRNA were used to produce biotin-labeled cRNA riboprobes with a SuperArray TrueLabeling-AMP 2.0 kit (tebu-bio) and biotinylated-UTP (Perkin Elmer, Boston, MA). Riboprobe from each treatment group was hybridized separately to an Oligo GEArray Human Cell Cycle DNA microarray (OHS-020) and to an apoptosis pathway-specific DNA microarray (OHS-012) (tebu-bio). Gene expression was analyzed using GEArray Expression Analysis Suite (http://geasuite.superarray. com/). Pathway Architect software (Stragagene, Amsterdam, The Netherlands) was subsequently used to generate gene maps linking genes to specific pathways.

Real-time PCR

tRNA was extracted with Trizol (Invitrogen AB, Stockholm, Sweden), before reverse transcription using Superscript II (Invitrogen), and the resulting cDNA was used as a template for real-time quantitative PCR. Three tumors per treatment group were analyzed separately for relative mRNA expression compared with the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH; Hs99999905_m1, Applied Biosystems, Warrington, UK) using an ABI 7900 PCR System (Perkin Elmer, Applied Biosystems, Foster City, CA) and Taqman universal master mix (Applied Biosystems). Relative levels of apoptosis and cell-cycle-related genes were assessed using the following Taqman gene expression assays (Applied Biosystems): Caspase 2 (Hs00154240_m1), Caspase 2 and RIPK1 domain-containing adaptor with death domain (CRADD) (Hs00388731_m1), Fas-associated protein with Death Domain (FADD) (Hs00538709_m1) tumor protein 53 (TP53) (Hs99999147_m1), B-cell CLL/lymphoma 2 (Bcl-2) (Hs00236808_s1), Bcl-2-associated X protein (Bax) (Hs99999001_m1) cyclin-dependent kinase inhibitor 1A (p21) (Hs00355782_m1), cyclin-dependent kinase inhibitor 1B (p27) (Hs00153277_m1), Cyclin D1 (Hs99999004_m1), Cyclin D3 (Hs 01017690_g1), Cyclin E (Hs01051894_m1), cell-division cycle 2, G1 to S and G2 to M (Cdc2) (Hs00364293_m1), cyclin-dependent kinase 7 (Cdk7) (Hs00387062_m1), cyclin-dependent kinase 2 (Cdk2) (Hs01548894_m1) and cyclin-dependent kinase 4 (Cdk4) (Hs00364847_m1). Relative mRNA was determined using the formula 2−ΔCT (CT;cycle threshold) where ΔCT = CT (target gene) − CT (GAPDH).

Western blotting

Tumor lysates (n = 3/treatment group) were pooled and protein quantified with bicinchoninic acid and copper sulphate as previously described.25 Twenty micrograms of protein were run on a 10% polyacrylamide gel and transferred onto imibilon-P nitrocellulose membrane (Millipore, Hertfordshire, UK). Nonspecific antibody binding was blocked with 1% casein (Vector Laboratories, Peterborough, UK). Proteins were detected using mouse monoclonal P53 (ab1101) 1:1,000, rabbit polyclonal CASPASE 9 (ab25758) 1:1,000, rabbit monoclonal (E6) CASPASE 8 (ab32125) 1:750, mouse monoclonal BCL-2 (ab692) 1:100, rabbit polyclonal BAX (ab10813) 1:1,000 rabbit monoclonal (E83-77) CASPASE 3 (ab32042) 1:500, mouse monoclonal p21 (ab54562) 1:10,000, rabbit monoclonal P27 KIP 1 (ab62364) 1:5,000, mouse monoclonal CYCLIN E (ab3927) 1:250, mouse monoclonal CYCLIN D1 (ab6152) 1: 500, rabbit monoclonal CYCLIN D3 (ab52598) 1:5,000, mouse monoclonal CYCLIN B 1:1,000 (BD Biosciences), mouse monoclonal CDK2 (ab6433) 1:2,500, mouse monoclonal CDK4 (ab6315) 1:1,000 and goat polyclonal RAP1a (C17) 1:200 (Santa Cruz). All primary antibodies were purchased from AbCam, Cambridge, UK, unless otherwise stated. Antibodies were made up in 1× PBS/1% casein and incubated with the membrane for 16 hr at 4°C. Secondary antibodies used were sheep anti-mouse-HRP (1:15,000) and donkey anti-rabbit-HRP (1:30,000) (GE Healthcare Life Sciences, Buckinghamshire, UK). HRP was detected with Supersignal chemiluminescence detection kit (Pierce, UK).

Histological analysis of tumors

Tumors were excised, fixed and paraffin-embedded following standard protocols. For detection of tumor macrophage infiltration, histological sections were stained using a rat anti-F4/80 monoclonal antibody (cat.no. MCA497R from AbDSerotec, Oxford, UK) at 1:50, followed by a biotinylated anti-rat antibody (Vectastain ABC staining from Vector labs, cat.no. PK-6104). Tumor Ki67 was determined by staining with an antibody specific for Ki67 (Vector, cat.no. VP-K4520 at 1:100) followed by the secondary biotinylated anti-mouse antibody (Vectastain ABC staining from Vector labs, cat.no. PK-6104). Slides were examined by an experienced histopathologist (SS Cross).

Statistical analysis

Statistical analysis was done by 1-way Kruskal–Wallis (nonparametric) test followed by Dunn's multiple comparisons test for tumor growth, 1-tailed Mantel–Haenszel test and log rank test for trend for survival curves and analysis of variance (ANOVA) followed by Dunnett's 2-sided multiple comparison test for gene expression. Statistical significance was defined as p less than or equal to 0.05. All p values are 2-sided.

Results

Sustained antitumor effect and increases survival following a 6-week course of sequential treatment with doxorubicin and zoledronic acid

As combination therapy has been shown to prevent the growth of tumors outside the skeleton in short-term experiments, we investigated whether sequential administration of doxorubicin followed by zoledronic acid caused a reduction in the volume of established subcutaneous breast tumors, comparing a 6-week schedule to continuous weekly treatment for 24 weeks. Following subcutaneous implantation of MDA-MB-436 cells, tumors reached approximately half-maximal permissible size (mean tumor volume 196.7 ± 13.2 mm3) before starting treatment on day 20. Mice were treated 1× per week with saline (control), 2 mg/kg doxorubicin, 100 μg/kg zoledronic acid, doxorubicin followed 24 hr later by zoledronic acid weekly for the duration of the experiment (24 weeks) or for a period of 6 weeks followed by 18 weeks of monitoring only. The doses of both agents are equivalent to those used in the clinical treatment of advanced breast cancer.

As shown in Figure 1a, there was no significant difference in tumor growth between control animals receiving saline and animals treated with either 2 mg/kg doxorubicin or 100 μg/kg zoledronic acid. All animals in these groups had to be sacrificed due to large tumor size (1,000 mm3). The 6-week course of weekly doxorubicin followed by zoledronic acid caused a decrease in mean tumor volume from 191.4 ± 36.95 mm3 to 123.5 ± 8.5 mm3 (p > 0.001 compared with control or single agents). Following withdrawal of sequential treatment, tumor volume did not increase over the next 107 days (mean tumor volume was 115.7 ± 60.9 mm3 on day 62 and 93.6 ± 47.3 mm3 on day 169). No significant differences in tumor volume were observed between animals treated with doxorubicin then zoledronic acid weekly for the duration of the experimental protocol, and those whose sequential treatment was stopped after 6 weeks.

Figure 1.

Effects of doxorubicin and zoledronic acid on subcutaneous MDA-MB-436 tumor growth and survival. (a) Mean tumor volume ± standard error of mean for subcutaneous tumor growth. Mice (n = 10 per group) were treated with saline (control), 2 mg/kg doxorubicin, 100 μg/kg zoledronic acid or doxorubicin followed 24 hr later by zoledronic acid 1× per week for the duration of the experiment. A fifth group was treated with doxorubicin followed 24 hr later by zoledronic acid 1× per week and treatment withdrawn after 6 weeks. Arrows indicate the commencement and cessation of treatment. Statistical analysis was by Kruksall–Wallis followed by Dunn's multiple comparison test. (b) Kaplan–Meier survival curves representing % survival of mice bearing subcutaneous MDA-MB-436 breast tumors. Statistical analysis is by 1-tailed Mantel–Haenszel test and log rank test for trend.

Animals treated sequentially with doxorubicin followed by zoledronic acid exhibited significant increases in survival; 70% of animals receiving continuous treatment and 60% of animals whose treatment was withdrawn after 6 weeks survived the entire duration of the experiment (169 days) (Fig. 1b). Tumors did not reach maximum volume in either of the sequential treatment groups; however, 7 animals were sacrificed as a result of skin blistering on the tumor surface. With this exception, all mice appeared healthy, and no significant weight loss was detected throughout the course of the study.

Doxorubicin and zoledronic acid in sequence affects apoptosis-related genes and proteins in subcutaneous breast tumors

To elucidate the molecular mechanisms by which sequential treatment reduced tumor growth, we used pathway-specific microarrays to examine alterations in the expression of 112 apoptosis-related genes in tumors isolated from animals following a 6-week course of treatment. Genes that showed at least a 2-fold change in expression in tumors isolated from animals treated with doxorubicin followed by zoledronic acid compared to controls (saline or the single agents) were clustered using GEAsuite software, and Pathway Architect software was subsequently used to link altered genes to specific pathways. This analytical method identified increased expression of 5 proapoptotic genes (p53, Bax, CRADD, FADD and Caspase 2) and a decrease in the antiapoptotic gene bcl-2, in tumors from animals receiving doxorubicin followed by zoledronic acid, compared to saline or single-agent treatment.

The changes in gene expression identified in the microarray analysis were subsequently confirmed by real-time quantitative RT-PCR (Table 1). Expression of Bcl-2 was decreased in tumors from animals receiving doxorubicin followed by zoledronic acid compared with controls (p < 0.001) and expression of proapoptotic Caspase 2, CRADD, FADD bax and p53 were increased, compared with all other treatment groups (p < 0.005). Caspase 2, FADD and Bax expression were also increased following administration of doxorubicin and zoledronic acid at the same time, compared with control or single agent treatment (p < 0.05). No alterations in expression of any of the apoptosis genes tested were detected following treatment with doxorubicin or zoledronic acid alone, or when zoledronic acid was administered 24 hr before doxorubicin (Table 1).

Table 1. Effects of zoledronic acid and doxorubicin, alone and in sequence or combination, on apoptosis- and cell-cycle-related gene expression
inline image

We next investigated the effects of the different treatments on cleavage of CASPASE 8 and CASPASE 9, molecules involved in induction of the extrinsic- and intrinsic-apoptotic pathway, respectively. Western blot analysis showed that the administration of all combinations of doxorubicin and zoledronic acid caused increased CASPASE 8 cleavage, compared with control or single agents (Fig. 2). CASPASE 9 cleavage was only detected in tumors treated with doxorubicin and zoledronic acid, either sequentially or simultaneously. To confirm that apoptosis was occurring via the mitochondrial pathway, levels of BAX and BCL-2 protein were assessed: Tumors from animals receiving sequential treatment had increased levels of proapoptotic BAX and decreased levels of antiapoptotic BCL-2. BCL-2 was also reduced in tumors from animals receiving doxorubicin and zoledronic acid at the same time, compared with control or single agent. In addition, cleaved (active) CASPASE 3, indicating increased levels of apoptosis, was detected following sequential and simultaneous treatment (Fig. 2).

Figure 2.

Effects of zoledronic acid and doxorubicin, alone and in sequence or combination, on apoptosis-related protein expression. Tumor lysates were prepared from animals treated 1× per week for 6 weeks with saline, 2 mg/kg doxorubicin, 100 μg/kg zoledronic acid, doxorubicin and zoledronic acid together, zoledronic acid followed 24 hr later by doxorubicin or the reverse sequence. Histogram shows mean final tumor volume (± standard error of the mean) for each treatment group. Western blots show expression of the apoptosis-related proteins BAX and BCL-2 and cleavage of CASPASE 8, CASPASE 9 and CASPASE 3. RNA polymerase was used to show equal loading of protein samples.

Doxorubicin and zoledronic acid in sequence affects cell cycle-related genes and proteins in subcutaneous breast tumors

Administration of doxorubicin followed by zoledronic acid has been shown to decreased tumor cell proliferation by more than 91% compared with control or single agents, but the mechanisms involved were not addressed.10 Here, we used cell-cycle-specific microarrays to elucidate mechanisms by which tumor cell proliferation is suppressed. Analysis of gene expression showed increased levels of p53, p21WAF1/CIP1, GADD45A and p27KIP1 and decreased levels of Cyclin D3, Cyclin D1, Cyclin E, Cdk2, Cdk2, Cdk7, Cdk4 and Rb, in tumors following sequential treatment with doxorubicin then zoledronic acid, compared with control or single agents (Fig. 3). Subsequent real-time PCR analysis confirmed increased expression of p53 (p < 0.001), p21WAF1/CIP1 (p < 0.001), GADD45A (p < 0.001) and p27KIP1 (p < 0.001) compared with controls (Table 1). In addition, p21WAF1/CIP1 was increased following administration of doxorubicin alone (p < 0.05) and doxorubicin and zoledronic acid at the same time (p < 0.05).

Figure 3.

Effects of zoledronic acid and doxorubicin, alone and in sequence or combination, on cell cycle-related protein expression. Tumor lysates were prepared from animals treated 1× per week for 6 weeks with saline, 2 mg/kg doxorubicin, 100 μg/kg zoledronic acid, doxorubicin and zoledronic acid together, zoledronic acid followed 24 hr later by doxorubicin or the reverse sequence. Histogram shows mean (± standard error of the mean), final tumor volume for each treatment group. Western blots show expression of the cell cycle-related proteins P53, P21, P27, CYCLIN D3, CDK4, CYCLIN B, CDK2 and CYCLIN E. RNA polymerase was used to show equal loading of protein samples.

Sequential administration of doxorubicin and zoledronic acid also resulted in decreased tumor expression of the following genes encoding for cyclins and cyclin-dependent kinases: Cdc2, Cyclin D3, Cyclin D1, Cyclin E, Cdk2, Cdk7, Cdk4 and Rb (all p < 0.001 compared with controls, Table 1). Expression of Cyclins D3, D1, E and the cyclin-dependent kinases Cdk2, Cdk7 and Cdk4, was also decreased in tumors treated with doxorubicin and zoledronic acid simultaneously (all p < 0.05 compared with control, single agents or zoledronic acid followed by doxorubicin). It is important to note that when reversing the sequence, giving zoledronic acid before doxorubicin, no alterations in tumor gene expression were detected, and only minor changes in gene expression were detected in tumors from animals receiving the single agents.

As administration of doxorubicin followed by zoledronic acid clearly altered tumor expression of several key genes associated with regulation of the cell cycle, we next investigated whether this was translated into corresponding changes in protein expression. Increased levels of the cell-cycle regulatory proteins P53, P21WAF1/CIP1 and P27KIP1, accompanied by decreased levels of CYCLINS D3, B and E, CDK4 and CDK2, were detected in tumors following sequential treatment, compared with any other treatment group (Fig. 3). Tumors from mice treated with doxorubicin and zoledronic acid simultaneously exhibited increased levels of P21WAF1/CIP1 and P27KIP1, as well as decreased levels of CYCLIN D3 CYCLIN E and CDK2, compared control, single agents or zoledronic acid followed by doxorubicin. In addition, levels of CDK2 protein were reduced following administration of doxorubicin, compared with control or zoledronic acid. These findings were all in agreement with the gene-expression studies.

Effects of doxorubicin and zoledronic acid on protein prenylation

Zoledronic acid is known to inhibit the action of several enzymes in the mevalonate pathway resulting in incomplete farnesylation and genranylgeranylation of signaling GTPases.20 This leads to intracellular accumulation of the unprenylated forms of these proteins and ultimately to the induction of apoptosis.21 Detection of the unprenylated form of Rap1a is commonly used as an indicator of zoledronic acid uptake in vitro, but due to the rapid homing of bisphosphonates to bone it has been considered unlikely that tumors outside the skeleton would be exposed to sufficient levels for protein prenylation to be affected. Whether pretreatment with doxorubicin affects tumor cell uptake of zoledronic acid and thereby causing an antitumor effect in vivo, is unknown. We therefore determined the levels of unprenylated Rap1a in subcutaneous MDA-MB-436 tumors in vivo, following administration doxorubicin and zoledronic acid, alone or in sequence. By western blot analysis, a 22KDa band indicative of unprenylated Rap1a was only detected in tumors from animals treated sequentially with doxorubicin followed by zoledronic acid (Fig. 4). Surprisingly, we were unable to detect unprenylated Rap1a in tumors following any of the other treatment schedules that included zoledronic acid, suggesting that pre-exposure to doxorubicin caused increased uptake of subsequently administered bisphosphonate.

Figure 4.

Effects of zoledronic acid and doxorubicin, alone and in sequence or combination, on prenylation of the small GTPase Rap1a. Tumor lysates were prepared from animals treated 1× per week for 6 weeks with saline, 2 mg/kg doxorubicin, 100 μg/kg zoledronic acid, doxorubicin and zoledronic acid together, zoledronic acid followed 24 hr later by doxorubicin or the reverse sequence. Western blots show expression of the unprenylated form of rap1a.

Analysis of residual tumor mass following sequential treatment

Although tumors failed to regrow even after extensive periods when sequential treatment was discontinued at 6 weeks, we could still detect palpable tumor masses in the animals. Histological analysis of these residual tumors was performed and assessed for viability, necrosis (Fig. 5) and macrophage infiltration (Fig. 6). Although tumors contained large necrotic areas, there were still significant areas with high numbers of viable, Ki67-positive tumor cells, indicating that the cells were actively proliferating (Fig. 5). Zoledronic acid has been reported to reduce infiltration and MMP9 expression of macrophages in cervical carcinoma, potentially resulting in an antiangiogenic effect in vivo.26 Staining the tumors from all treatment groups with an antibody specific for macrophages (F4/80), we found that in all cases the macrophages seemed to specifically cluster along the rim of the tumors (Fig. 6). Tumors from animals treated with doxorubicin followed by zoledronic acid appeared to have fewer macrophages in the central tumor core, but quantitative assessment of this would require further studies

Figure 5.

Histology showing viable and necrotic areas in the residual tumor mass. Histology of a representative tumor following 6 weeks of sequential treatment with doxorubicin followed by zoledronic acid and then sacrificed on day 169. (a) Low power view of the whole tumor showing that about half of the tumor is necrotic (n) but there are areas of viable tumor at the periphery (v). (b) High magnification view showing mitotic figures in the viable tumor cells. (c) Medium magnification view of Ki67-stained preparation showing that in this area of viable tumor around 50% of the cells have positive nuclear staining indicating that they are in the active phases of the cell cycle.

Figure 6.

Effects of sequential treatment on tumor macrophage infiltration. Representative examples of immunohistochemical staining with an antibody against F4/80 showing macrophages in tumors. (a) Tumor (day 169) following 6 weeks of treatment and (b) tumor (day 169) after continuous weekly treatment with doxorubicin followed by zoledronic acid. There are large numbers of macrophages clustered at the periphery of the tumor, whereas very few macrophages were detected in the centre.

Discussion

In this study, animals carrying subcutaneous MDA-MB-436 tumors were administered weekly doxorubicin followed by zoledronic acid to investigate the long-term effects of both continuous- and a 6-week course of treatment. In accordance with the previous studies,5, 10 we found that weekly treatment with either 2 mg/kg doxorubicin or 100 μg/kg zoledronic acid for 6 weeks had no effect on tumor growth, whereas administration of doxorubicin followed 24 hr later by zoledronic acid leads to a 35.5% reduction in final tumor volume. Although tumor volume was reduced following long-term sequential treatment, small (pinhead sized) tumors were still visible, indicating that this schedule does not eradicate all tumor cells. When the treatment was withdrawn after 6 weeks, tumor volume remained small but constant; no increases in tumor size were recorded during the subsequent 15.2 weeks until the termination of the experiment. No significant difference in tumor size was recorded between mice treated continuously and mice that received a 6-week course of treatment. These data indicate that subcutaneous breast tumor cells not killed by sequential treatment are rendered “dormant” for a long period of time and that continuous weekly treatment is not required to inhibit tumor growth. As histological analysis revealed that after treatment the residual tumors contained significant numbers of cells in active phases of the cell cycle as well as apoptotic tumor cells, tumor cell death and proliferation are both taking place but without a net increase in tumor volume. Potentially, this may be explained by reduced tumor vascularization reported following sequential treatment,10 limiting the tumor size. In agreement with this, we observed that the central cores of the sequentially treated tumors contained very low numbers of infiltrating macrophages, believed to be a prerequisite for tumor angiogenesis. The reduced tumor burden in animals receiving sequential treatment did result in substantially increased survival compared to giving the same agents as single therapy.

Both in vitro and in vivo studies have shown that sequential administration of doxorubicin followed by zoledronic acid increases apoptosis and reduces proliferation of breast cancer cells.5, 10, 27 Here, we have demonstrated the molecular pathways by which these 2 drugs interact to induce apoptosis, inhibit proliferation and ultimately inhibit MDA-MB-436 breast tumor growth in vivo. Administration of doxorubicin or zoledronic acid alone had no effect on expression of the apoptosis-related genes tested. Analysis of protein revealed small increases in CASPASE 8 cleavage and proapoptotic BAX expression in tumors treated with doxorubicin, but these increases were not sufficient to induce caspase 3 cleavage, which is required for apoptosis induction. No alterations in levels of proteins associated with apoptosis were detected in tumors from animals treated with zoledronic acid alone. In addition, treatment with a single drug had little effect on cell-cycle regulatory genes. These results are consistent with the lack of effect of any of the single agents on tumor burden. Although direct proapoptotic and antiproliferative effects of bisphosphonates have been demonstrated in vitro,28 the bioavailability of bisphosphonates in soft tissue, in vivo, is poor. Clinical administration of zoledronic acid is a 4-mg infusion every 3–4 weeks, resulting in a peak plasma concentration of 1–2 μM, which is cleared from the circulation within 1–2 hr.29 It is therefore likely that subcutaneous tumors are exposed to low concentrations of zoledronic acid for only a few hours. In this study, 100 μg/kg zoledronic acid was used, which is equivalent to the clinical 4 mg dose. We conclude that treating mice with a clinically relevant dose of zoledronic acid had no effect on subcutaneous MDA-MB-436 tumor apoptosis or proliferation, probably due to poor bioavailability of the bisphosphonate. It appears likely that an initial exposure to doxorubicin renders tumors more sensitive to subsequent treatment with zoledronic acid. Evidence from in vitro and in vivo models suggest that bisphosphonates exert synergistic antitumor activity when administered in combination with cytotoxic drugs, targeted molecular therapies or radiotherapy.30 The demonstration of these antitumor effects in vivo indicates that the low-serum concentrations of bisphosphonates may be sufficient to exert antitumor effects in peripheral tissues when combined with other drugs. However, none of these studies have addressed the mechanisms by which these synergistic antitumor effects occur. Suggested possibilities are that zoledronic acid by targeting small GTPases inhibits pathways that are highly active in malignant cells, and this antitumor effect is then enhanced by the addition of a more specific therapeutic agent, as has been shown for a combination with imatinib in models of leukemia and lung cancer.31 Other reports have suggested that enhanced inhibition of Ras-dependent survival pathways accounts for the increased antitumor effect seen when zoledronic acid is combined with farnesyltransferase inhibitors in prostate cancer,32 and that decreased levels of VEGF may be part of the effects of zoledronic acid in combination therapy.33 However, none of the reported studies have provided an explanation for the sequence-specific antitumor effect seen in this study with doxorubicin and zoledronic acid. We speculate that pretreatment with doxorubicin may increase subsequent uptake of zoledronic acid, and this is supported by the data showing increased levels of unprenylated Rap1a in tumors only following treatment with doxorubicin followed by zoledronic acid. Another possibility we are investigating is whether zoledronic acid enhances the retention of doxorubicin in tumors, thereby enhancing the antitumor effect.

Inhibition of the cell cycle by doxorubicin in a number of cancer cell lines is well documented, including breast,14–17 colon15, 34–37 and hepatocarcinoma.38 Doxorubicin inhibits cellular proliferation via P53-dependent and-independent mechanisms.15, 18, 36, 37 In P53 wild-type MCF7 breast cancer cells, doxorubicin treatment results p53 accumulation,14–16 P53 recruits P21 and GADD45A17 leading to decreased expression of CYCLIN B and CDC216 resulting in a cell-cycle block at G2/M.14 However, MDA-MB-436 cells express mutated P5339; thus, it is unlikely that increases in p53 expression seen in these tumors are functionally recruiting p21 and GADD45A. In the absence of functional P53, P73 and P63 have been shown to replace P53 inducing P21 and GADD45A in doxorubicin-treated breast cancer cells,18 and this is a possible mechanism by which these proteins accumulate following sequential treatment with doxorubicin then zoledronic acid. Experiments in colon cancer cells have shown that doxorubicin can inhibit CYCLIN D1,35 indicating that this drug may also block G1.

As discussed, many of the cell-cycle inhibitory effects seen in this study can be attributed to the action of doxorubicin, but there are also reports supporting that zoledronic acid may contribute to inhibition of cell proliferation.22, 40, 41 There is no evidence to suggest a role for doxorubicin in altering the expression of either CYCLIN E or CYCLIN D3, whereas zoledronic acid has been shown to reduce the expression of CYCLIN E in osteosarcoma cells22 and CYCLIN D3 in colon and leukemia cells40 in a p53-independent manner. Zoledronic acid has also been shown to reduce expression of CYCLIN D141 and CYCLIN B,40 however, not all cell-cycle inhibitory effects observed can be attributed to the actions of this bisphosphonate. There is evidence from leukemia, colon cancer40 and osteosarcoma models41 that zoledronic acid does not alter P53 or P21 expression but acts independently of these proteins. Weekly administration of zoledronic acid was unable to exert antiproliferative effects in subcutaneous MDA-MB-436 tumors, probably due to the short half-life in serum as discussed earlier. Regardless of the molecular mechanisms by which doxorubicin and zoledronic acid interact, it appears that both drugs contribute to inhibition of tumor cell proliferation in subcutaneous MDA-MB-436 tumors in vivo.

The roles of doxorubicin and zoledronic acid in the induction of apoptosis in peripheral breast tumors are less clear. Exposure to increasing concentrations of doxorubicin (1–2,500 nM) for 24 hr in vitro does not result MDA-MB-436 apoptosis (data not shown), and resistance of these cells to doxorubicin-induced apoptosis has been reported.42 In addition, we found that MDA-MB-436 cells do not undergo apoptosis following treatment with 10–100 μM zoledronic acid (data not shown). In agreement with this, no alterations in the expression profile of apoptosis-related genes were found in tumors treated with doxorubicin or zoledronic acid alone compared with control. Experiments carried out in osteosarcoma,22 pancreatic43 and colon23 cancer cells treated with 0.1–100 μM zoledronic acid as well as studies in rat cardiomyocytes,44 T-lymphocytes,45 and rat cortical neurons46 treated with 0.1–20 μM doxorubicin for 24 hr have shown that both doxorubicin and zoledronic acid are capable of executing apoptosis via the intrinsic and extrinsic apoptotic pathways. Both drugs have been shown to induce apoptosis by disturbing the BCL-2 family protein balance and disrupting the mitochondrial membrane potential,22, 30 leading to activation of CASPASE 942, 43 and subsequent cleavage of CASPASE 323, 45; the mechanism of, intrinsic, mitochondrion-dependent apoptotic signaling. In addition, both drugs can induce cleavage of CASPASE 823, 45 resulting in activation of the extrinsic, mitochondrion-independent pathway.

In vitro studies have shown that the antiproliferative and proapoptotic effects of zoledronic acid can be abolished by simultaneous addition of zoledronic acid and geranylgeraniol, an intermediate of the mevalonate pathway downstream of the enzyme targeted by zoledronic acid, farnesyl diphosphate synthase.17, 26 These data indicate that zoledronic acid exerts anticancer effects via inhibition of protein prenylation. Restoration of geranylgeranylation has been shown to rescue the G1 arrest caused by inhibitors of mevalonate synthesis resulting in increased expression of CYCLIN D1, CYCLIN E and CDK2.47 We detected accumulation of the unprenylated form on Rap1a in subcutaneous tumors from animals having received doxorubicin followed by zoledronic acid in vivo, indicating increased uptake of zoledronic acid following exposure of tumor cells to doxorubicin. These data suggest that antitumor effects observed following sequential treatment may in part be due to inhibition of the mevalonate pathway. However, coadministration of intermediaries of this pathway is not feasible in animal models, preventing direct confirmation that antitumor effects of sequential treatment are due to the disruption of prenylation caused by zoledronic acid.

Our data show that weekly, sequential administration of clinically achievable doses of doxorubicin (2 mg/kg) and zoledronic acid (100 μg/kg) for 6 weeks induces significant alterations in tumor apoptosis- and cell-cycle-regulatory genes/proteins, increased tumor cell apoptosis and reduced tumor cell proliferation, leading to reduced subcutaneous tumor burden and increased long-term survival. This is the first reported long-term study showing that a 6-week course of sequential treatment was sufficient to reduce tumor growth for at least another 18 weeks and that continuous weekly treatment therefore is not required to prevent tumor regrowth. This study suggests that sequential administration of doxorubicin followed 24 hr later by zoledronic acid may be of benefit to patients with organ-confined breast cancer.

Acknowledgements

We thank Professor NJ Brown (University of Sheffield, UK) who holds the UK Home Office project license and Dr Jonathan Green (Novartis Pharma, Switzerland) for the kind gift of zoledronic acid.

Ancillary