Bone metastases are a frequent clinical complication associated with breast cancer and are observed in 70% to 80% of patients with advanced-stage breast cancer.1–4 The major clinical manifestations of bone metastases are increased morbidity associated with bone pain, skeletal fractures, hypercalcemia, and spinal cord compression.1, 4 Most bone metastases observed in breast cancer patients are osteolytic, resulting from invading malignant tumor cells releasing a variety of factors that enhance bone cell activity and lead to excessive bone loss.1, 4, 5 Insulin-like growth factor-1 (IGF-1) is a polypeptide systemic hormone with 40% to 50% sequence homology to insulin.6, 7 IGF-1 binds to a tetrameric transmembrane receptor, IGF-1R, which has high homology to insulin receptor (IR), particularly in the β subunit.8 The IGF-1R is a tyrosine kinase–containing heterotetramer that has potent antiapoptotic and mitogenic activities in a variety of cell types including bone and breast cancer cells.9–12 IGF-1R is activated by IGF-1, IGF-2, and to a lesser extent by insulin.9 Upon activation by IGF-1, the IGF-1R undergoes tyrosine phosphorylation, subsequently signaling through a number of downstream pathways including phosphatidylinositol 3-kinase (PI3K)/protein kinase B (Akt) and mitogen-activated protein kinase (MAPK).9, 12–19 In the skeleton, IGF-1 and its receptor play a key role in the regulation of bone remodeling and bone mass.11, 12, 18, 20, 21 Bone-derived IGF-1 released during bone resorption stimulates osteoclast formation and bone resorption.20, 22 Studies in transgenic animals demonstrated that genetic inactivation of IGF-1R increases trabecular bone mass and connectivity and reduces ovariectomy-induced bone loss mainly due to an osteoclast defect.12, 21, 23–25 Studies in osteoblasts have shown that IGF-1R activation stimulates receptor activator of NF-κB ligand (RANKL) production and enhances osteoblast differentiation, migration, and function in vitro and in vivo.12, 18, 21, 26, 27
IGF-1 also plays a central role in the regulation of mammary gland biology and the development and progression of breast cancer.11, 14, 28, 29 IGF-1R overexpression and enhanced activity are typical features of most primary breast cancers and were found to be essential for neoplastic transformation of epithelial cells in the breast.28, 30, 31 Moreover, IGF-1R activity is known to promote breast cancer cell proliferation, survival, transformation, and invasion in vitro and in vivo,13, 32–34 and blockage of IGF-1R has been shown to reduce tumor growth and progression.20, 35–38 A number of studies have shown that elevated circulating levels of IGF-1 are linked to increased risk of multiple aspects of breast cancer progression, including bone metastasis and resistance to chemotherapy.14, 15, 28, 30–32 Altogether, these studies strongly suggest that IGF-1R is essential for the development of breast cancer and plays a key role in bone metabolism. The suppression of the tyrosine kinase activity downstream of the IGF-1 receptor is therefore likely to be effective against bone metastasis associated with breast cancer. In view of this, we studied the effects of the selective inhibitor of IGF-1 receptor tyrosine kinase activity cis-3-[3-(4-methyl-piperazin-l-yl)-cyclobutyl]-1-(2-phenyl-quinolin-7-yl)-imidazo[1,5-a]pyrazin-8-ylamine (PQIP)39, 40 on human and mouse breast cancer–induced osteoblastic and osteoclastic changes in models of osteolytic bone disease. Previous studies have shown that PQIP inhibits proliferation and induces apoptosis in breast cancer cells in vitro and reduces tumor progression in vivo.39, 40 In this study, we show that PQIP prevents human and mouse breast cancer–induced osteolysis by reducing bone turnover through effects on both osteoclasts and osteoblasts.
Materials and Methods
IGF-1 receptor kinase inhibitor
The IGF-1 receptor tyrosine kinase, inhibitor PQIP was provided by OSI Pharmaceuticals (Farmingdale, NY, USA) and its synthesis and structure has been described.39 PQIP was dissolved in dimethyl sulfoxide (DMSO) or L-tartaric acid for in vitro and in vivo studies, respectively.
All solvents and reagents were purchased from Sigma-Aldrich (Dorset, UK) unless otherwise stated. Tissue culture medium was obtained from Invitrogen (Paisley, UK). All primary antibodies were purchased from Cell Signaling Biotechnology (Danvers, MA, USA). Human IGF-1 and macrophage colony-stimulating factor (M-CSF) were purchased from R&D Systems (Abingdon, UK). RANKL was a gift from Patrick Mollat (Galapagos SASU, Paris, France).
Breast cancer cell lines and conditioned medium
Human MDA-MB-231 and mouse 4T1 breast cancer cells were purchased from American Type Culture Collection (ATCC) (Manassas, VA, USA). MDA-MB-231 and 4T1 cells were cultured in standard α modified essential medium (α-MEM) (α-MEM supplemented with 10% fetal calf serum (FCS), penicillin, and streptomycin). For studies involving MDA-MB-231 conditioned medium, MDA-MB-231 cells were cultured in standard α-MEM until 80% confluent. Cells were then cultured in serum-free medium and were allowed to grow for a further 24 hours before conditioned medium was removed, filtered (0.22 µm), and stored.
Mouse osteoblast culture
Primary osteoblasts were isolated from the calvarial bones of 2-day-old mice by sequential collagenase digestion. For bone nodule assay, osteoblasts were seeded into 12-well plates at 10 × 105 cells per well in standard α-MEM, supplemented with β-glycerol phosphate (10 mM) and L-ascorbic acid (50 µg/mL). The cells were cultured for up to 21 days with replacement of the culture medium in the presence or absence of test substances every 48 hours. Osteoblast number, differentiation, and bone nodule formation were determined by AlamarBlue assay, alkaline phosphatase assay, and alizarin red (ALZ) staining, respectively.
RANKL and M-CSF mouse osteoclast culture
Osteoclast formation, survival, and activity were studied using RANKL- and M-CSF–generated mouse osteoclasts. Briefly, bone marrow (BM) cells were flushed from the long bones of 3- to 5-month-old mice and cultured for 48 hours in standard α-MEM supplemented with mouse M-CSF (100 ng/mL). For osteoclast generation, the resulting M-CSF–generated BM macrophages (BM-derived osteoclast precursor) were plated into tissue culture plates (96-well plates, 15 × 103 cells/well and 12-well plates, 150 × 103 cells/well) in standard α-MEM supplemented with M-CSF (25 ng/mL) and RANKL (100 ng/mL) for the desired period of time.
Breast cancer cell–BM-derived osteoclast precursor coculture
BM-derived osteoclast precursors were plated into 96-well plates (10 × 103 cells/well) in 150 µL of standard α-MEM supplemented with M-CSF (25 ng/mL) and RANKL (100 ng/mL) for 6 hours prior addition of MDA-MB-231 or 4T1 breast cancer cells (300 cells/well). For studies involving conditioned medium, medium from breast cancer cells was added to osteoclast cultures (10% vol/vol) in standard α-MEM.
Mouse osteoblast–BM cell cocultures
Osteoblasts and bone marrow cells were plated together at 8 × 103 cells per well and 2 × 105 cells per well, respectively, in 96-well plates in standard α-MEM supplemented with 10% FCS, antibiotics, and 1,25-(OH)2-vitaminD3 (10 nM) for 4 days. Cultures were incubated with vehicle (0.1% DMSO) or PQIP (200 nM) for 60 minutes prior to stimulation with conditioned medium from human MDA-MB-231 breast cancer cells (10% vol/vol). Cultures were terminated on day 6.
Tartrate-resistant acid phosphatase staining
Tartrate-resistant acid phosphatase (TRAcP) staining was used to identify multinucleated osteoclasts. Osteoclast cultures were fixed in 4% paraformaldehyde and incubated with naphtol-AS-BI-phosphate, pararosanilin, and tartrate in acetate buffer (30 µM) at 37°C for 45 minutes. TRAcP-positive cells with three or more nuclei were considered to be osteoclasts and were manually counted on a Zeiss Axiovert light microscope (Carl Zeiss, Oberkochen, Germany) using a 10× objective lens.
Quantification of resorption area
RANKL- and M-CSF–stimulated BM-derived osteoclast precursors were plated on Corning Osteo-Assay Surface multiple-well plates (Corning, USA) for 48 hours. Human MDA-MB-231 conditioned medium (10% vol/vol) was added to mature osteoclast culture in the presence and absence of treatment. Adherent osteoclasts were incubated in 20% bleach (Clorox-Ultra, USA) for 10 minutes and then resorption pits were visualized by phase contrast microscopy on an Olympus ScanR microscope. Resorbed area was quantified by Image Analysis using ImageJ.41
AlamarBlue assay was used to measure the number of mature osteoclasts, osteoblasts, and breast cancer cell.42 Cells were incubated in AlamarBlue reagent (10%, vol/vol) for a further 3 hours and fluorescence was measured (excitation, 530 nm, emission 590 nm) using a Biotek Synergy HT plate reader.
Alkaline phosphatase assay
Alkaline phosphatase (ALP) activity was used to assess mouse calvarial osteoblast differentiation. Osteoblasts were homogenized in ALP lysis buffer (1 M diethanolamine, 1 mM MgCl2, 0.05% Triton X100) and cell lysate was mixed with an equal volume of paranitrophenol-phosphate (20 mM). Absorbance was measured at 414 nm using a Biotek Synergy HT plate reader.
Alizarin red staining
Alizarin red (ALZ) staining was used to assess bone nodule formation in calvarial osteoblast cultures. Briefly, osteoblast cultures were fixed and immersed in 40 mM ALZ solution (pH 4.2) for 20 minutes at room temperature on an orbital rotator. Cultures were incubated in destaining solution (10%, wt/vol) cetylpyridinium chloride in 10 mM sodium phosphate for 15 minutes. Absorbance of the extracted stain was then measured at 562 nm using a Biotek Synergy HT plate reader and compared to an ALZ standard curve.
Measurement of cell migration
MC3T3-E1 cells were seeded into 12-well plates (500 × 103 cells/well) and allowed to grow to confluence over a period of 48 hours. Cell migration was assessed using a wound healing assay. Briefly, confluent monolayers of cells were scored with a fine pipette tip, and then the cells were treated with either vehicle or PQIP (200 nM) for 1 hour prior to exposure to IGF-1 (100 ng/mL) for 16 hours. Migration of cells was visualized on an Olympus ScanR microscope and percentage of wound closure was calculated using TScratch.
Measurement of cell spreading
The spreading of mouse BM–derived osteoclast precursors and MC3T3-E1 cells were assessed by using the real-time cell analyzer xCELLigence (Roche Applied Science, UK).43 Mouse MC3T3-E1 (10 × 103 cells/well) cells were plated into specialized 16-well plates (Roche Applied Science, UK) in standard α-MEM supplemented with IGF-1 (100 ng/mL) in the presence and absence of PQIP (200 nM) for 2 hours. For studies involving osteoclast precursor cells, mouse BM–derived osteoclast precursors were plated into 16-well plates (10 × 103 cells/well) in standard α-MEM supplemented with M-CSF (25 ng/mL) and RANKL (100 ng/mL) for 48 hours prior to addition of IGF-1 (100 ng/mL) in the presence and absence of PQIP (200 nM) for 2 hours. Data were analyzed using the xCELLigence software package, expressing changes in cell electrode impedance as changes in spreading as described.43 The kinetics shown represents mean ± SD for at least three measurements.
Western blot analysis was used to detect protein expression and phosphorylation in cultured bone cells. Briefly, cells were seeded in 12-well plates and maintained in standard α-MEM until confluent. Prior to stimulation with test agents or vehicle, cells were incubated in serum-free α-MEM medium for 60 minutes in the presence of test agents or vehicle. The cells were then scraped in standard lysis buffer (0.1% (wt/vol) SDS, 0.5% (wt/vol) sodium deoxycholate, 1% Triton X-100, 1 mM EDTA, 2% (vol/vol) protease inhibitor cocktail, 10 µM of sodium fluoride and 2% (vol/vol) phosphatase inhibitor cocktail. The lysate was incubated on ice for 10 minutes and centrifuged at 14,000g at 4°C for 5 minutes. The supernatant was collected and protein concentration was determined using BCA assay (Pierce, USA). Total protein (30–70 µg) was resolved by SDS-PAGE on 12% polyacrylamide SDS gels, transferred onto polyvinylidene fluoride (PVDF) membranes (BioRad, UK) and immunoblotted with antibodies according to the manufacturer's instructions. The immunocomplexes were visualized by an enhanced chemiluminescence detection kit (Pierce, USA) using horseradish peroxidase–conjugated secondary antibody (Jackson Labs, UK), and then visualized using chemiluminescence (Amersham, UK) on a Syngene GeneGnome imaging system.
Transcription factor assays
Osteoclasts and M-CSF–dependent macrophages were cultured in six-well tissue culture plates. The culture medium was replaced with serum-free medium containing vehicle (0.1% DMSO) or PQIP (200 nM) for 60 minutes prior to stimulation with RANKL (100 ng/mL) for 45 minutes. Nuclear extracts were prepared using a nuclear extract kit (Active Motif, Rixensart, Belgium) and DNA binding was measured using TRANSAM ELISA kits for NF-κB, cFos, and nuclear factor of activated T cells cytoplasmic 1 (NFATc1) (Active Motif, Rixensart, Belgium), according to the manufacturer's instructions.
Human MDA-MB-231 breast cancer cell–mouse calvaria coculture system
We studied the effects of the IGF-1 receptor kinase inhibitor PQIP on osteolysis induced by human MDA-MB-231 breast cancer cells and their derived factors ex vivo using an adaptation of the mouse calvarial organ culture as described.44 Neonatal mouse calvarias were isolated from 7-day-old mice and incubated in standard α-MEM for 24 hours. Each mouse calvaria was divided into equal halves along the medium sagittal suture and each half was placed into culture on stainless steel rafts in 48-well plates containing standard α-MEM with MDA-MB-231 (1 × 104 cells/well) (Fig. 5A). Tissue culture medium containing test agents was changed every 48 hours and the cultures were terminated after 7 days. Bone volume was assessed by using micro–computed tomography (µCT) (Skyscan 1172 scanner; Skyscan, Belgium) at a resolution of 5 µm.
Human MDA-MB-231 conditioned medium in vivo model of osteolytic bone disease
We studied the effects of the IGF-1 receptor kinase inhibitor PQIP on osteolysis induced by breast cancer derived factors using the human MDA-MB-231 conditioned medium in vivo model of osteolytic bone disease. All experimental protocols were approved by the Ethics Committee at the University of Edinburgh and were conducted in accordance with the UK Home Office regulations. Briefly, 3-week-old C57Bl'6 female mice were injected subcutaneously over the calvarial bones with 50 µL of human MDA-MB-231–conditioned medium on 3 consecutive days. Animals were divided into two groups and received daily intraperitoneal injection of either vehicle (L-tartaric acid) or PQIP (2 mg/kg) for 5 days. All animals were euthanized on day 5 and the calvarial bone was assessed using µCT (Skyscan 1172 scanner) at a resolution of 8 µm. No mice injected with human MDA-MB-231–conditioned medium in this study exhibited any obvious physical signs of illness or inflammatory response.
Mouse 4T1 intratibial injection in vivo model of osteolytic bone disease
We investigated the effects of the IGF-1 receptor kinase inhibitor PQIP on osteolytic bone metastases using the mouse 4T1 model of osteolytic bone disease. All experimental protocols were approved by the Ethics Committee at the University of Edinburgh and were conducted in accordance with the UK Home Office regulations. Briefly, 14 female Balb/c 10-week-old mice received intratibial injection of mouse 4T1 breast cancer cells (5 × 104 cells) in the left leg or a sham injection of PBS into their right leg. Animals were divided into two groups and received oral dosing of either vehicle (L-tartaric acid) or PQIP (100 mg/kg). Dosages and treatment regimes used for PQIP were chosen based on previous in vivo studies, which demonstrated that this agent has antitumor effects.39 Animals were euthanized 14 days postinjection and both tibias and femurs were analyzed by µCT (Skyscan 1172 scanner). In order to study the effects on tumor growth within bone, the whole tibial metastases area was measured on 2D µCT images using Image J (1.34s; NIH, Bethesda, MD, USA) and results were expressed as a percentage of total metaphyseal area. Following scanning, the limbs were embedded and processed for bone histomorphometry.
Comparison between groups was done by analysis of variance (ANOVA) followed by Bonferroni posttest using SPSS for Windows version 11. A p value 0.05 or below was considered statistically significant. The concentration that produced 50% of response IC50 was calculated using GraphPad Prism 4 for Windows.
The IGF-1 receptor tyrosine kinase inhibitor PQIP inhibits human and mouse breast cancer–induced osteoclastogenesis
A number of studies have shown that inhibition of IGF-1R activity reduces the proliferation of bone and breast cancer cells,11, 12, 21, 28 but the effects of inhibitors of IGF-1R tyrosine kinase activity on breast cancer induced osteoclastogenesis have not been studied. In this study, we used PQIP, a novel IGF-1 receptor tyrosine kinase inhibitor that exhibits minimal activity toward a wide panel of other protein kinases.39, 45 First, we examined the effects of PQIP on breast cancer–induced osteoclast formation using breast cancer cell–BM–derived osteoclast precursor cocultures. As shown in Fig. 1A, the human MDA-MB-231 and mouse 4T1 breast cancer cells markedly increased osteoclast number in RANKL- and M-CSF–stimulated breast cancer cell–BM cocultures and these effects were significantly inhibited by PQIP (200 nM) (p < 0.05). PQIP had no effects on the growth and viability of MDA-MB-231, 4T1, or BM osteoclast precursors at the concentration that inhibited osteoclast formation in breast cancer cell–BM cocultures (Fig. 1B), thereby excluding inhibition of IGF-I receptor growth-related functions.46 PQIP (200 nM) exerted similar effects in cocultures of BM and human MCF7 or mouse MC57G cancer cells, which both are known to cause osteolysis in mice47 (Supporting Fig. S1).
PQIP inhibits osteoclast formation induced by MDA-MB-231–derived factors
To gain insight into the mechanism by which PQIP suppressed breast cancer–induced osteoclastogenesis, we studied the effects of this agent on osteoclast formation induced by factors derived from human MDA-MB-231 breast cancer cells. As shown in Fig. 1C, conditioned medium from MDA-MB-231 significantly enhanced RANKL- and M-CSF–stimulated osteoclast formation in BM cultures. It is worth noting that this increase was significantly less than that observed in BM–breast cancer cell cocultures (Fig. 1A). PQIP (200 nM) completely abolished conditioned medium–induced osteoclast formation, indicating a strong inhibitory effect on osteoclastogenesis induced by breast cancer–derived factors. We also studied the effect of PQIP on osteoclast formation by counting the number of osteoclasts with 10 or more nuclei. Conditioned medium from MDA-MB-231 significantly increased osteoclast size and nuclearity, such that the proportion of cells with more than 10 nuclei increased by 100% (Fig. 1D). Treatment with PQIP prevented this increase and also reduced the number of osteoclasts to 50% of the number obtained in vehicle-treated control cultures (Fig. 1D, E). To determine whether PQIP exerted inhibitory effects on mature osteoclast survival and activity, we conducted further studies in mature osteoclast cultures that were treated with PQIP in the presence and absence of MDA-MB-231–conditioned medium. We found that PQIP had no inhibitory effects on osteoclast survival (Supporting Fig. S2A) or bone resorption (Supporting Fig. S2B and C) at a concentration that inhibited osteoclast formation.
PQIP inhibits IGF-1 and breast cancer–induced PI3K/Akt pathway in osteoclasts and their precursors
In view of the fact that PI3K/Akt and MAPK are important components of signaling pathways downstream of the IGF-1 receptor,19 we next examined the effects of PQIP on IGF-1 and MDA-MB-231–conditioned medium–induced Akt and MAPK activation in osteoclasts. These experiments showed that PQIP treatment (200 nM) inhibited IGF-1–induced and conditioned medium–induced Akt phosphorylation at both threonine and serine sites (Fig. 1F). Interestingly, PQIP had no effects on conditioned medium–induced extracellular signal-regulated kinase 1/2 (ERK1/2) phosphorylation (Fig. 1G) but completely abolished ERK1/2 activity induced by IGF-1 (Fig. 1G). Similarly, PQIP failed to inhibit conditioned medium–induced c-Jun N-terminal kinase (JNK) and P38 MAPK phosphorylation (Fig. 1G). These findings indicate the selectivity of this compound toward IGF-1 receptor–mediated signaling and suggest that inhibition of PI3K/Akt signaling in osteoclasts is responsible, at least in part, for the inhibitory effects of PQIP on osteoclastogenesis.
PQIP inhibits IGF-1–induced osteoclast formation without affecting RANKL and M-CSF signaling
In view of the important role that RANKL and M-CSF play in osteoclastogenesis, we next examined the effects of PQIP on IGF-1–induced osteoclast formation and signaling in the presence and absence of RANKL and M-CSF. Treatment of BM-derived osteoclast precursors with IGF-1 (100 ng/mL) induced Akt phosphorylation (Fig. 2A) and significantly enhanced RANKL and M-CSF–stimulated osteoclast formation (Fig. 2I). Pretreatment of these cultures with PQIP for 1 hour prior to stimulation with IGF-1 significantly inhibited Akt activity (Fig. 2A) and completely abolished osteoclast formation (Fig. 2I). In fact, PQIP reduced the number of osteoclasts below levels observed at cultures treated only with RANKL and M-CSF (Fig. 2I), implying nonspecific inhibitory effects on RANKL and M-CSF signaling. However, these were excluded because PQIP had no effects on RANKL-induced phosphorylation of I-κB (Fig. 2B), ERK1/2 (Fig. 2C), and P38 (Fig. 2D), and it also failed to suppress RANKL-induced NF-κB (Fig. 2F), NFATc1 (Fig. 2G), and cFOS (Fig. 2H) nuclear translocation and DNA binding. PQIP also failed to inhibit M-CSF–induced Akt phosphorylation (Fig. 2E) or the growth and viability of BM-derived osteoclast precursors (Fig. 2K). To test the effects of PQIP on mature osteoclast survival, we generated mature osteoclasts in RANKL- and M-CSF–stimulated BM cultures for 5 days and then exposed these cultures to PQIP (200 nM) for 48 hours. In these experiments, PQIP (200 nM) had no effects on mature osteoclast survival (Fig. 2J). To further explore the mechanism by which PQIP suppressed osteoclast formation, we studied the effects of this compound on IGF-1–induced motility and proliferation of osteoclast precursors. Treatment of BM-derived osteoclast precursors with PQIP (200 nM) for 1 hour prior to exposure to IGF-1 (100 ng/mL) significantly inhibited IGF-1–induced spreading of osteoclast precursors (Fig. 2L, M) without affecting cell viability (Fig. 2N). Our interpretation of these results is that PQIP inhibits osteoclast formation by selectively suppressing IGF-1–mediated motility of osteoclast precursors via a mechanism dependent on PI3K/Akt inhibition. However, further studies are needed to fully examine the effects of PQIP on fusion of mature osteoclasts and their precursors.
PQIP inhibits breast cancer–induced osteoblast–osteoclast coupling
IGF-1 regulates osteoblast support for osteoclast formation and activity.12, 18, 21, 26, 27 To investigate whether PQIP affects this process, we first examined the expression of IGF-1 receptor and signaling in calvarial osteoblasts. We have found that osteoblasts express the highest level of IGF-1R, significantly higher than osteoclasts (94%, p < 0.01) and BM-derived osteoclast precursors (97%, p < 0.01) (Supporting Fig. S3). Next, we went on to investigate the effect of PQIP on osteoclast formation in osteoblast–BM cell cocultures in the presence of MDA-MB-231–conditioned medium. As shown in Fig. 3A, B, conditioned medium from MDA-MB-231 significantly enhanced osteoclast formation in osteoblast–BM cell cocultures and PQIP (200 nM) completely abolished these effects. In view of the fact that osteoblasts are a major source of essential osteoclastic factors such as RANKL and OPG,48 we next tested the effects of PQIP on RANKL and OPG mRNA expression in osteoblasts. As shown in Fig. 3C, both IGF-1 (100 ng/mL) and MDA-MB-231–conditioned medium caused a significant increase in the RANKL/OPG ratio and these effects were completely inhibited by PQIP (200 nM), thereby indicating strong inhibitory effects on osteoblast support to osteoclastogenesis. The levels of Akt phosphorylation in these osteoblast cultures after exposure to IGF-1 or conditioned medium from MDA-MB-231 breast cancer cells are shown in Fig. 3D. As expected, PQIP treatment (200 nM) inhibited basal, IGF-1, and MDA-MD-231–conditioned medium–induced Akt phosphorylation (Fig. 3D). These results together imply that PQIP is likely to inhibit osteoblast support for osteoclastogenesis in the metastatic bone microenvironment.
PQIP inhibits osteoblast spreading, migration, differentiation, and bone nodule formation
In view of the fact that IGF-1 regulates osteoblast differentiation and activity in vitro and in vivo,18, 21, 26, 27 we next studied the effects of PQIP on IGF-1–induced osteoblast growth, spreading, migration, differentiation, and bone nodule formation. As shown in Fig. 4A, IGF-1 (100 ng/mL) caused a significant increase in bone nodule formation after 21 days of continuous treatment. PQIP completely abolished IGF-1–induced bone nodule formation and significantly inhibited basal levels of bone nodule formation. PQIP (200 nM) also reduced osteoblast differentiation as measured by ALP activity (Fig. 4C) without affecting cell viability (Supporting Fig. S4). On the other hand, conditioned medium from human MDA-MB-231 breast cancer cells caused a significant reduction in bone nodule formation (Fig. 4B) and ALP activity (Fig. 4D) in calvarial osteoblast cultures and these effects were not significantly altered in the presence of PQIP (200 nM). Next we went on to examine the effects of PQIP on the spreading and migration of the osteoblast-like cells MC3T3-E1 using the real-time cell analyzer xCELLigence and wound healing assay, respectively. Treatment with IGF-1 (100 ng/mL) resulted in an increase in MC3T3-E1 spreading within 1 hour (Fig. 4E) and enhanced MC3T3-E1 migration after 16 hours (Fig. 4F), and these effects were inhibited by PQIP (200 nM). These effects were accompanied by a significant and dose-dependent inhibition of IGF-1–induced PI3K/Akt phosphorylation (Supporting Fig. S5). Altogether, these experiments clearly demonstrate that PQIP inhibits both basal and IGF-1–induced osteoblast motility, differentiation, and activity in vitro. More importantly, these results suggest that inhibition of IGF-1R tyrosine kinase activity by PQIP in the bone metastatic environment is likely to exacerbate the inhibitory effects of breast cancer cell–derived factors on osteoblasts.
PQIP reduces human breast cancer–induced osteolysis ex vivo and in vivo
To study the effects of PQIP on osteolysis associated with human breast cancer, we first tested the effects of PQIP on bone loss using our human MDA-MB-231–mouse calvarial coculture system (Fig. 5A). As shown in Fig. 5B and C, human MDA-MB-231 breast cancer cells caused osteolysis characterized by a significant loss in bone volume when cocultured with mouse calvaria for 7 days, and these effects were completely prevented by PQIP (500 nM). Indeed, treatment with PQIP (500 nM) caused a significant gain in bone volume in comparison to vehicle-treated cultures (34.4% ± 12% increase, p < 0.05). PQIP had no effect on the viability (Fig. 5D) or motility (Fig. 5E, F) of human MDA-MB-231 breast cancer cells, indicating a strong inhibitory effect on soluble factor–dependent osteolysis. In view of this, we next went on to examine if treatment with PQIP influences the ability of human breast cancer–derived factors to cause osteolysis in adult mice. Osteolysis was assessed by measuring bone loss after subcutaneous injection of conditioned medium from MDA-MB-231 over the calvarial bones (Fig. 5G). All mice that injected with conditioned medium developed significant bone loss in calvarial bones indicative of osteolytic bone damage, whereas these effects were significantly reduced in mice receiving daily intraperitoneal injection of PQIP (2 mg/kg) for 5 days (Fig. 5H, I).
PQIP inhibits mouse breast cancer–induced osteolytic bone disease and bone turnover in vivo
We next investigated the effects of oral PQIP treatment on the development of osteolytic bone disease in vivo using the mouse 4T1 breast cancer model. Oral daily administration of PQIP (100 mg/kg/d) was started 1 week after the mouse 4T1 cell inoculation (5 × 104) and continued until day 14. As shown in Fig. 6A, B, all mice which were treated with vehicle developed significant trabecular bone loss below the growth plate, indicative of osteolytic bone lesions. Osteolysis was accompanied by deterioration in the quality of the bone characterized by the disappearance of some trabeculae reflected in the significant reduction in trabecular number (Fig. 6C). These changes led to a loss of trabecular connectivity as indicated by an increased trabecular pattern factor (Tb.Pf) value (inverse connectivity index, Fig. 6D). Once-daily oral administration of PQIP (100 mg/kg) totally prevented trabecular bone loss and caused a significant gain in trabecular bone volume when compared to sham mice (8% increase; p < 0.05). A detailed bone histomorphometric analysis showed that osteoclast numbers and active resorption surfaces were increased following 4T1 cell inoculation in vehicle-treated mice, whereas these effects were totally prevented following treatment with PQIP (33% reduction in osteoclast number and 43% reduction in resorption; p < 0.01) (Table 1, Supporting Fig. S6A). Treatment with PQIP (100 mg/kg/d) also reduced osteoblast number (3%; p < 0.01), osteoid surface (30%; p < 0.01), mineral apposition rate (34%; p < 0.01) and bone formation rate (53%; p < 0.01) (Table 1, Supporting Fig. S6B). This is indicative of a significant reduction in osteoblast activity and bone formation. Analysis of tumor size in histological sections showed a nonsignificant trend toward reduction in tumor area within the metaphysis of the proximal tibia following treatment with PQIP (100 mg/kg/d) (Fig. 6E, F). Altogether, these results show that PQIP reduces breast cancer–induced bone turnover and osteolysis with a modest antitumor activity in this mouse model of breast cancer.
Table 1. Static Histomorphometry Following Intratibial Injection of Mouse 4T1 Breast Cancer Cells in the Presence and Absence of Treatment With the Selective IGF-1 Receptor Kinase Inhibitor PQIP (100 mg/kg/d)
Values are expressed as means ± SEM and are obtained from six different samples.
Previous studies have shown that selective inhibition of IGF-1R tyrosine kinase activity using PQIP inhibits proliferation and induces apoptosis in breast cancer cells in vitro and reduces tumor progression in vivo. The observations presented in this study demonstrate that PQIP prevents human and mouse breast cancer–induced osteolysis by reducing bone turnover through effects on both osteoclasts and osteoblasts. Experiment in cocultures of a variety of human and mouse breast cancer cells with BM-derived osteoclast precursors showed that PQIP significantly inhibited the increase in osteoclast formation. In this model, we excluded the effects of PQIP on the mitogenic effects of IGF-1 by examining the proliferation of MDA-MB-231, 4T1, and BM-derived osteoclast precursors. In these experiments, PQIP failed to exert any significant effects at concentrations that inhibited osteoclast formation (200 nM). The data presented here support the hypothesis that PQIP inhibits osteoclastogenesis in this model by disrupting breast cancer support for osteoclast formation. In this regard, we showed that early treatment of BM-derived osteoclast precursors with PQIP prior to exposure to conditioned medium from MDA-MB-231 cells inhibited conditioned medium–induced osteoclast formation, size, and nuclearity. PQIP was significantly more potent in inhibiting conditioned medium–induced osteoclast formation in this culture compared to breast cancer–BM cocultures (Fig. 1A). This was likely due to the fact that BM cultures exposed to conditioned medium had fewer osteoclasts than those cultured with breast cancer cells. Notwithstanding this, the data presented in this study clearly demonstrate that PQIP exerts a strong inhibitory effect on osteoclastogenesis induced by breast cancer–derived factors.
Experiments in mature osteoclasts showed that PQIP had no effects on osteoclast survival and bone resorption in MDA-MB-231–osteoclast cocultures. This raises the possibility that the inhibitory effect of PQIP in osteoclast formation is mainly due to an early effect on osteoclast precursor cell motility. To investigate this, we tested the effects of short-term PQIP treatment on the spreading of osteoclast precursors. These experiments showed that PQIP completely abolished IGF-1–induced spreading of bone-derived osteoclast precursors within 3 minutes of exposure. These data are consistent with our in vitro data that showed PQIP inhibited osteoclast size and nuclearity and also in agreement with the established role of IGF-1 in cell migration and spreading.12, 14–18 In addition to its direct effects on osteoclasts and their precursors, PQIP also inhibited IGF-1– and breast cancer–induced RANKL/OPG ratio in osteoblasts and completely abolished MDA-MB-231–conditioned medium–induced osteoclast formation in osteoblast–BM cell cocultures. Taken together, these findings are consistent with a model whereby PQIP inhibits breast cancer–induced osteoclastogenesis; directly, by inhibiting the motility and spreading of osteoclast precursors; and indirectly, by suppressing osteoblast support for osteoclastogenesis.
In view of the important role that RANK, M-CSF, and IGF-1 receptors play in the regulation of osteoclast formation,12, 21, 26, 27, 49 we studied the effects of PQIP on key signaling pathways downstream of these receptors. Analysis of the PI3K/Akt pathway showed that PQIP completely abolished both IGF-1– and conditioned medium–induced Akt phosphorylation in osteoclasts and their precursors. These effects were observed at concentrations similar to those that inhibited osteoclast formation. Studies by Ji and colleagues39 reported that PQIP inhibited IGF-1–induced ERK1/2 MAPK activation in tumor cells. In this study, we showed that PQIP inhibited IGF-1–induced ERK1/2 in osteoclasts and we also demonstrated for the first time that PQIP had no significant effects on conditioned medium–induced ERK1/2, JNK, and P38 MAPK activation. Although these results need to be interpreted with caution because of uncertainty about the concentrations of IGF-1 in breast cancer–conditioned medium, our studies clearly demonstrate that PQIP exerts strong inhibitory effects on both IGF-1– and conditioned medium–induced PI3K/Akt activation at similar concentrations to those that caused osteoclast inhibition. This raises the possibility that inhibition of PI3K/Akt but not the MAPK pathway may have contributed to the inhibitory effects of PQIP on breast cancer–induced osteoclastogenesis. ERK1/2 belongs to the MAPK family and lies downstream of IGF-1 and various other factors such as parathyroid hormone–related protein (PTHrP) and M-CSF, which act independently of the IGF-1 receptor but are known to be present in conditioned medium of human MDA-MB-231 breast cancer cells.50, 51 Therefore, it is not surprising that PQIP had more pronounced inhibitory effects on MDA-MB-231–conditioned medium–induced Akt activity than the MAPK pathway.
The possibility that inhibition of RANKL and M-CSF signaling may contribute to PQIP effects in osteoclast formation was excluded because this compound was found to be completely ineffective in inhibiting RANKL and M-CSF signaling in osteoclasts and their precursors. It is important to note, however, that whereas PQIP had no effects on RANKL- and M-CSF-induced osteoclast formation in the absence of exogenous IGF-1 and conditioned medium, it did exert a modest and significant inhibitory effect on formation of osteoclasts with 10 or more nuclei. We cannot readily explain this except to note that tissue culture medium is known to contain a variety of growth factors, such as insulin, IGF-1, and IGF-2, that could potentially enhance osteoclast formation. Therefore, disruption of tyrosine kinase activity associated with these receptors by PQIP may have contributed to osteoclast inhibition observed in the absence of exogenous IGF-1, but further work will be required to explore the mechanisms responsible. Because PQIP inhibited Akt phosphorylation and led to a significant reduction in osteoclast number when activated by MDA-MB-231–conditioned medium, we hypothesized that that inhibition of tumor- and/or bone-derived IGF-1 signaling in osteoclasts may contribute to the effects of PQIP in this model. Breast cancer cells including MDA-MB-231 are known to produce IGF-1, to express the receptor, and to respond to IGF-1 action.38, 52 Both IGF-1 and its receptor were found to be highly expressed in osteoclasts and BM-derived osteoclast precursors used in this study (Supporting Fig. S7). This suggests that an IGF-1 autocrine/paracrine loop is likely operating in osteoclast/BM–breast cancer cell cocultures and it may therefore play a role in PQIP-mediated osteoclast inhibition. We cannot completely exclude the possibility that PQIP may also inhibit tyrosine kinase activity associated with receptors for insulin or inhibit the actions of other growth factors present in conditioned medium that activates the IGF-1 receptor such as IGF-2.
We have studied the effects of PQIP on high bone turnover and osteolysis associated with human and mouse breast cancer ex vivo and in vivo. Treatment with PQIP inhibited osteolysis induced by human MDA-MB-231 breast cancer cells in the cancer cell–calvarial coculture system and significantly reduced osteolytic bone damage induced by human breast cancer–derived factors in adult mice. This is consistent with the effects of this agent in osteoclasts in vitro and demonstrates that PQIP is capable of inhibiting human breast cancer–induced osteolysis in this model. In keeping with this, we found that oral administration of PQIP inhibited the development of osteolytic lesions induced by intratibial injections of mouse 4T1 cells. PQIP treatment in this model reduced osteoclast numbers and inhibited bone resorption, thereby leading to higher bone volume and a more preserved architecture. Consistent with the important role of IGF-1 on osteoblast activity and bone formation,12, 18, 21 we observed a significant reduction in osteoblast number and osteoid surface, MAR, and BFR, indicative of a significant reduction in osteoblast activity and bone formation. This is in agreement with previous studies that showed inhibition of pharmacological and genetic inactivation of IGF-1 reduces osteoblast activity,12, 21, 26, 27 and is entirely consistent with our in vitro data which showed that PQIP inhibited osteoblast spreading, migration, differentiation, and mineralization. In spite of significant inhibition of cancer-induced osteoblastic and osteoclastic changes by PQIP, we observed no significant reduction in tumor growth in the 4T1 model. These findings differ from previous studies that have shown that PQIP inhibits tumor growth in other in vivo models.39 We cannot readily explain this difference except to note that previous studies were carried out in different models using cell lines that are known to express high levels of IGF-1 and insulin receptors and have been shown to rely on signaling through the IGF-1/IGF-II/Insulin receptor.39
In conclusion, the results reported here show that pharmacologic inhibition of IGF-1 receptor tyrosine kinase activity using PQIP inhibits osteoclast and osteoblast changes induced by breast cancer cells in vitro and in vivo and prevent the development of bone metastases induced by injection of these cells in vivo. Although the work presented here indicates that PQIP and its novel derivatives which are now in advanced clinical development, have great potential as treatments for metastatic bone disease associated with breast cancer, our studies were restricted to the MDA-MB-231 ex vivo and 4T1 in vivo models. Therefore, further studies will be required to determine if PQIP and its novel analogues are effective in preventing the development and progression of metastases in other disease models. It is also important to note that the significant reduction in osteoblast differentiation and bone formation following treatment with PQIP may limit its long-term usefulness as bone-sparing drug. Therefore, further long-term studies are needed to ascertain whether, and to what extent, inhibition of osteoblast differentiation may limit the long-term usefulness of this class of agents for the treatment of osteolytic bone disease associated with breast cancer.
All authors state that they have no conflicts of interest.
This study was partly supported by Cancer Research UK grant C6088/A12063. JGL is supported by a PhD Studentship from the MRC Institute of Genetics and Molecular Medicine (University of Edinburgh). VGB is supported by Cancer Research UK program grant C157/A9148.
Authors' roles: Study design: AII, VGB, and JGL. Study conduct: AII, JGL, AS, SM, and MM. Data collection: AII, JGL, AS, SM, and MM. Data analysis: AII, JGL, AS, and SM. Data interpretation: AII, JGL, AS, and SM. Drafting manuscript: AII and VGB. Revising manuscript content: AII, VGB, JGL, AS, and SM. Approving final version of manuscript: AII, VGB, JGL, AS, MM, and SM take responsibility for the integrity of the data analysis.