Effect of the Specific Src Family Kinase Inhibitor Saracatinib on Osteolytic Lesions Using the PC-3 Bone Model

Prostate cancer skeletal involvement is the most common metastatic site in advanced-stage disease, observed in up to 80% of patients (1). Unlike other solid tumor metastases that cause bone loss, prostate cancer bone metastases result in an overall increase in bone mass or osteoblastogenesis. It is not entirely clear why prostate tumor cells localize to the bone environment. The enriched blood source and complex growth factors in the bone marrow provide a good soil for circulating prostate tumor cells to seed and multiply. Growth factors released from tumor cells support increase of both bone resorption and formation (2). Increased osteoclastogenesis has been detected in prostate cancer bone metastases (3). The initial osteoclast-mediated osteolytic process is critical to the subsequent osteoblastic lesions, and a metastatic phenotype consists of a combination of both bone resorption and formation (4–6). A vicious cycle in the tumor cell–bone interface drives the metastasis, whereby growth factors secreted by tumor cells stimulate osteoclast formation and matrix turnover. This osteoclastic resorption causes the local release of more growth factors that in turn activate tumor growth. Src, a non–receptor tyrosine kinase, is implicated in the cross-talk in this tumor-bone interface through its central role in bone metastasis and tumor progression.

The importance of Src kinase in osteoclasts was first discovered in Src knockout mice in which defective osteoclastic resorption gave rise to the phenotype of osteopetrosis and toothlessness (7, 8). In bone formation and remodeling, Src is activated in osteoclasts upon attachment of cells to bone matrix through integrin receptors, and activated Src recruits several signaling proteins for motility and cytoskeletal rearrangement (9, 10). Src is also important in the receptor activator of NF-κB ligand (RANKL)/RANK/osteoprotegerin signaling pathway, which contributes to osteoclast differentiation. Membrane-bound or soluble RANKL binds to its receptor RANK to initiate the signaling cascade leading to osteoclast activation. Upon ligand binding, RANK recruits the tumor necrosis factor receptor–associated factors to its cytoplasmic domain (11–13). This crucial step leads to activation of NF-κB (14), mitogen-activated protein kinase (MAPK; refs. 15–17), and Src (10, 18–20) signaling pathways in osteoclasts. In addition to osteoclast activation, Src recruits phosphatidylinositol-3 kinase and activates the Akt/mammalian target of rapamycin pathway to promote osteoclast survival (21, 22).

Our previous studies showed that Src kinase plays an important role in promoting androgen-independent prostate tumor growth (23–26). Upregulation of Src family kinase (SFK) activities has been reported in 28% of androgen-independent prostate cancer patients (27). Moreover, this enhanced SFK activity correlated with prostate cancer bone metastasis and a significant decrease in survival. The specific SFK inhibitor saracatinib (AZD0530) inhibited the activities of Src and focal adhesion kinase (FAK) kinases and their downstream substrates. Phosphorylation of Src and FAK was inhibited by saracatinib in the nanomolar range. As a result, tumor metastasis was profoundly inhibited in an animal study using castration-resistant LNCaP expressing neuropeptide gastrin-releasing peptide (LNCaP-GRP cells; ref. 26).

We proposed to use saracatinib not only as an adjunct to hormone ablation therapy and to target patients with upregulation in SFK activity but also for patients with bone metastasis. In this study, we use PC-3 cells for their ability to steadily induce osteolysis when implanted in the mouse tibia. In addition to downregulation of cell proliferation and migration as reported before, saracatinib inhibited the expression of interleukin-8 (IL-8), urokinase plasminogen activator (uPA), and matrix metalloprotease-9 (MMP-9), and PC-3 invasion, which altogether suggests that saracatinib may inhibit prostate tumor cell invasion into the bone matrix. Induced differentiation of macrophage cells into osteoclasts by RANKL was inhibited by saracatinib, as was the activation of Src, P38, and IκB-α in PC-3 cells—events generally mediated through RANKL stimulation. Our animal study further shows the ability of saracatinib to retard osteolytic lesions, a potential therapeutic advantage in treating patients with prostate cancer.

The PC-3 prostate cancer cell line was obtained from the American Type Culture Collection and cultured in RPMI 1640 supplemented with 5% fetal bovine serum (FBS). RAW264.7 cells, a murine osteoclast precursor, was from the American Type Culture Collection and kept in DMEM with 10% FBS.

PC-3 cells were seeded (2,000 per well) in triplicates in 96-well plates. Varied concentrations of saracatinib (0–5 μmol/L, AstraZeneca) were added, and cell growth was monitored by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT, Sigma) assay at 1, 3, and 5 days posttreatment.

Invasion assays were done according to procedures described by Lochter and colleagues (28) and modified as follows. PC-3 cells (1 × 10⁵) were suspended in growth medium containing dimethyl sulfoxide or various concentrations of saracatinib (8 nmol/L to 5 μmol/L in 5-fold increments) into the cell culture inserts containing 8 μmol/L pores for 24-well plates. The inserts were coated with Matrigel (2–3 mg/mL protein) and allowed to solidify before cell plating. The lower chambers were filled with 300 μL of growth medium containing control or drugs of corresponding concentrations. After 72 hours in culture, cells were fixed with 5% glutaraldehyde in PBS and stained with 0.5% toluidine blue in 2% Na₂CO₃. Only cells that penetrated the membrane were counted in five microscopic fields per filter.

Osteoclast formation assay was done by plating 5 × 10⁴ RAW264.7 cells per well in 24-well plates in DMEM with 10% FBS for 4 days. Soluble RANKL (100 ng/mL, Santa Cruz Biotechnology) was added to stimulate osteoclast differentiation, and saracatinib (1 μmol/L) was added for inhibition. At the end of incubation, the cells were fixed and stained with a tartrate-resistant acid phosphatase (TRAP) staining kit (Acid Phosphatase Kit 387-A, Sigma), and TRAP-positive staining osteoclasts (multinuclear cells with > 3 nuclei) were counted in five different microscopic fields.

Total RNA was isolated from PC-3 cells treated with or without 1 μmol/L of saracatinib at different time points (0, 6, and 24 hours). Two micrograms of RNA were reverse transcribed into cDNA followed by PCR using specific primers. Quantitative PCR was done in iCycler with SYBR Greener QPCR iCYLer (Invitrogen), and the output was normalized against the controlled amplification products with glyceraldehyde-3-phosphate dehydrogenase.

MMP-9 activity was analyzed with gelatin zymography as described below. Conditioned media harvested from PC-3 treated with or without various concentrations of saracatinib for 48 hours were concentrated through centrifugation using Microcon concentrators (Amicon). Concentrated conditioned medium was normalized to total cellular protein (equivalent to 100 μg per loading), then subjected to electrophoresis in gels containing 0.2% gelatin without denaturing. After electrophoresis, the gels were washed in 2% Triton X-100 for 30 minutes and incubated in 50 mmol/L Tris-HCl (pH 7.4), 200 mmol/L NaCl, and 10 mmol/L CaCl₂ at 37°C overnight. Gels were then stained with Coomassie brilliant blue R-250 and destained with methanol/HOAc/H₂O (4:1:5) until clear bands appeared.

PC-3 cells were cultured to 70% confluency and treated with various concentrations of saracatinib for 1 hour. Cells were then lysed in ice-cold buffer (Tris 25 mmol/L, MgCl₂ 10 mmol/L, β-glycerolphosphate 25 mmol/L, Na₃VO₄ 5 mmol/L, and protease inhibitor cocktail) and ruptured through three freeze-thaw cycles. Cell lysates were separated from debris by centrifugation, and protein concentrations were determined by the bicinchoninic acid assay (Pierce). Thirty micrograms of protein were separated on SDS-PAGE and electrotransferred onto polyvinylidene difluoride membrane. Immunoblotting was done by incubating membranes with the primary antibodies overnight, followed by 1 hour of secondary antibody incubation. Signals were detected by SuperSignal West Pico CL (Pierce) coupled with X-ray film exposure. For inhibition of the RANKL/RANK pathway, PC-3 cells were serum-starved for 48 hours before stimulation by RANKL (100 ng/mL) in 2% FBS medium for 0, 2, 5, 15, 30, and 60 minutes. SFK inhibitor saracatinib (1 μmol/L) was added to the last three time points together with RANKL. At the end of treatments, cells were collected as described, and lysates were subjected to gel electrophoresis. Antibodies against p-Src (Tyr416), total Src (36D10), Akt, p-Akt (Ser473), P38, p-P38 (Thr180/Tyr182, 28B10), extracellular signal-regulated kinase 1/2 (Erk1/2; p44/42MAPK), IκBα (44D4), and p-IκBα (Ser32, 14D4) were purchased from Cell Signaling; anti–p-FAK (Y397) was from BD Biosciences; and anti–p-FAK (Y861) was from Affinity BioReagents.

PC-3 cells (2 × 10⁵) mixed with equal amounts of Matrigel in a final volume of 10 μL were injected into 24 male severe combined immunodeficient (SCID) mice tibia. Intratibial injections were done using a 29-gauge, 1/2-inch needle inserted through the tibial plateau of the flexed knee. After animals recuperated from the surgery (7 days later), they were randomly divided into two groups and given oral saracatinib (25 mg/kg) or buffer (0.5% hydroxypropyl methylcellulose and 0.1% Tween 80) daily for 8 weeks from implantation. Anesthetized mice had bone lesions monitored weekly using Faxitron (Faxitron X-ray Corp.). At the end of 8 weeks, animals were sacrificed with collection of urine, sera, and tibias for biochemical and pathologic analyses. The tibias were fixed in 10% buffered formalin for 2 days, followed by decalcification in 10% EDTA solution for 2 weeks at room temperature with occasional stirring and solution changes. Specimens were paraffin embedded, sectioned, and stained with H&E and TRAP. Histomorphometric analysis was done on an Olympus system. The number of large active osteoclasts (TRAP-positive osteoclasts with three or more nuclei) per millimeter of tumor-bone interface was measured from 4-week control and treatment bone sections in five different fields at 200× magnification. Total tumor percentage in the medullary cavity of each bone from 8-week controls and treatment samples was measured at 40× magnification.

The pyridinoline cross-links and helical peptides representing the degree of bone resorption were measured by EIA assays (Metra Serum pyridinoline assay and Metra Helical Peptide EIA Kit, Quidel).

Statistical analyses were done using unpaired Student's t tests (StatView, SAS Institute). Significant results were determined as P ≤ 0.05.

We have reported that SFK inhibitor saracatinib efficiently inhibits the growth of most prostate cancer cells, LNCaP, PC-3, DU-145, and CWR22Rv1, with IC₅₀ values in the low micromolar range (29). PC-3 cells are especially sensitive to saracatinib with an IC₅₀ of 0.7 μmol/L (Fig. 1A). Inhibition of phosphorylation of Src kinase and its immediate substrate FAK was achievable with submicromolar levels of saracatinib, as was the case for MAPK-P38. Inhibition of Akt required higher doses of saracatinib, implying an indirect relationship of Src and Akt (Fig. 1B).

Saracatinib is very potent in inhibiting PC-3 cell migration with an IC₅₀ value of ∼50 nmol/L. This SFK inhibitor also blocks PC-3 penetration through Matrigel in invasion assay (Fig. 2A). After 3 days, cell invasion was reduced to 57% and 25% of control in the presence of 40 and 200 nmol/L of saracatinib, respectively. PC-3 cells remained 78% and 73% viable using the same concentrations of inhibitor over the same period of time; thus, the reduction was indeed the effect of saracatinib on cell invasion. The expression of invasion-related molecules IL-8 and uPA was inhibited by saracatinib in a timely manner (Fig. 2B). Treatment of PC-3 cells with 1 μmol/L of saracatinib reduced IL-8 transcripts to 30% in 6 hours and to 18% in 24 hours (Fig. 2C). The same treatment halved uPA expression in 6 hours and further decreased to 44% after 24 hours. The level of uPA receptor was only slightly reduced (to around 85%) after 24 hours of treatment. MMP-9 activity in PC-3 cultured medium was inhibited by saracatinib in a dose-responsive manner (Fig. 2D).

Murine macrophage RAW264.7 precursors for osteoclasts may be stimulated by RANKL (100 ng/mL) to undergo osteoclast differentiation. Addition of 1 μmol/L saracatinib drastically inhibited osteoclast differentiation by limiting the number of TRAP-stained multinuclear osteoclasts to less than 18% of control (Fig. 3A). Treating RAW264.7 cells with this concentration of saracatinib alone did not cause massive cell death as shown in the staining (Fig. 3B); hence, the reduction in TRAP staining is likely due to inhibition of osteoclast differentiation. It has been reported that functional RANK is expressed in prostate cancer cell lines such as PC-3 and DU-145 (30). The receptors render these prostate cancer cells under the control of bound or soluble RANKL from osteoblasts or stroma to enhance tumor growth in the bone environment. We therefore tested if saracatinib would block the RANKL/RANK pathway in PC-3 cells. Activation of Src, P38, and IκBα (a few representatives from the RANKL/RANK pathway) was detectable at 2 minutes from RANKL addition and peaked within 15 minutes (Fig. 3C). Saracatinib completely blocked phosphorylation of these kinases at 15, 30, and 60 minutes.

Based on our in vitro observations, we then implanted PC-3 cells into the tibia of SCID mice to study the effect of the Src inhibitor in vivo. PC-3 cells cause osteolysis when injected directly into mouse bones (31–33). In our study, osteolysis started to appear in the control group 3 weeks after injection. Radiographs taken at 4 and 8 weeks (end points) showed that saracatinib significantly inhibited the osteolytic lesion in the treatment group compared with controls (Fig. 4A). At 5 weeks from surgery, 9 of 12 control mice showed osteolytic lesions compared with only 4 in the treatment group. TRAP staining of the tibial samples taken at 4 weeks showed growth of tumor cells leading to loss of bone architecture in the control tibias with characteristic TRAP staining signifying the presence of osteoclasts (Fig. 4B). Most of the tibias from the treatment animals retained bone integrity, bearing no signs of tumor cells or osteoclasts. The osteoclast perimeters from the 4-week bone samples were counted. Saracatinib decreased TRAP-positive osteoclasts in tibias of tumor-bearing mice by >10 fold when compared with controls. At the end of 8 weeks, all but one mouse in the control group developed severe lesions, but only half of the treatment animals had milder lesions. The areas of tumor versus bone were measured from histologic staining and plotted to show that saracatinib reduced this ratio in the treatment group to less than half of controls (Fig. 4C). In addition to bone histomorphometric measurements, the effect of saracatinib on released mouse bone metabolites, such as pyridinoline cross-links in serum and α-helical peptides in urine, were measured. Compared with the control group, treated mice had a 46% reduction in pyridinoline (P = 0.07, not statistically significant) and 60% reduction in α-helical peptide (P < 0.01; Fig. 4D).

When standard therapy for advanced prostate cancer using androgen ablation fails, there are limited treatment options for patients with prostate cancer metastatic to bone. At present, bone health is fostered by use of calcium, vitamin D, and zolendronic acid. Unbalanced bone remodeling mediated by prostate tumor cells replaces normal bone tissue with highly disorganized osteoblasts. This leads to predominantly osteoblastic lesions causing significant pain in patients with advanced prostate cancer (34). New therapeutic strategies such as small-molecule inhibitors of signaling pathways responsible for cancer progression are desirable. Src kinase has been implicated in prostate tumor proliferation, migration, survival (23, 35, 36), and progression to castration resistance (26). Many recently developed small-molecule SFK inhibitors, including bosutinib (SKI-606; ref. 37), dasatinib (BMS-354825; refs. 38, 39), and saracatinib are able to reduce the proliferation, migration, and invasion of prostate cancer cell lines in vitro and are currently in clinical trials. These inhibitors also decrease prostate cancer growth and metastasis in mouse xenograft studies (26, 29, 39). Recent studies using saracatinib showed that Src has a pivotal role in the formation and activation of human osteoclasts (40) and is involved in breast cancer bone metastasis regardless of the estrogen receptor status (41). A phase I study showed a dose-dependent decrease in bone resorption markers in men receiving saracatinib daily for 2 weeks (42). There were no significant adverse effects. Dasatinib alone and in combination therapy with docetaxel improved bone mineral density in the LNCaP C4-2B tumor–implanted tibias (43).

Using saracatinib, we were able to show a dose-dependent inhibition of proliferation and migration in PC-3 and DU-145 cells (29). Western blot analysis showed that Src phosphorylation in saracatinib-treated cells was reduced in a dose- and time-dependent manner. Intracellular Src-activating proteins, including FAK, P130Cas, and paxillin, were also inhibited, supporting the efficacy of saracatinib in preventing prostate cancer progression. In addition to the previous prostate cancer lines that lack androgen receptor, LNCaP-GRP cells that contain androgen receptor but grow independent of androgen were also sensitive to saracatinib. With the importance of Src in both bone remodeling and prostate cancer tumor progression, the use of Src inhibitors has been proposed in the treatment of prostate cancer bone metastasis (44, 45). In our study, saracatinib has shown specific inhibition of FAK and P38 kinase activation downstream of Src in PC-3 cells. This block in signaling may prevent cytoskeletal rearrangement and migration through FAK (29). We again observed the discrepancy between the doses of saracatinib required for inhibiting Src phosphorylation and cell/tumor growth as discussed in our previous article (29). In addition to speculation of partial inhibition of Src kinase on growth and its downstream signaling, we recently reported that saracatinib induced macroautophagy in cells to spare them from apoptosis by inhibiting the phosphoinositide 3-kinase/Akt/mammalian target of rapamycin pathway (46). Blocking autophagy with drugs such as chloroquine significantly improved the efficacy of saracatinib in treating PC-3 subcutaneous tumors, which may also be applicable to bone treatment. Saracatinib reduced the synthesis of IL-8, uPA, and MMP-9, molecules involved in angiogenesis and basement membrane degradation (47), in PC-3 cells, which is consistent with cell invasion inhibition. RANKL secreted from tumor and stroma cells is responsible for osteoclast differentiation in the tumor-bone interface. Using RAW264.7 cells as a model for osteoclast differentiation induced by RANKL, saracatinib significantly inhibited the formation of osteoclasts. Furthermore, the RANKL-induced RANK/RANKL pathway in PC-3 cells was also inhibited by saracatinib along with P38 and IκBα. Finally, in vivo experiments showed that progression of PC-3 implants in the metastatic tibial site was inhibited by saracatinib. The inhibition not only prevented proliferation of PC-3 cells but also hindered multiplication of osteoclasts in the bone environment. Osteoclast perimeter counts were significantly lower at the 4-week time point, and the tumor/bone ratio was halved at the end of the in vivo study in the inhibitor-treated specimens. Reduction in secretion of bone turnover markers pyridinoline and helical peptides in mice receiving saracatinib was consistent with fewer osteolytic lesions and decreased osteoclast proliferation. Although the majority of advanced prostate cancer patients develop osteoblastic lesions, bone resorption occurs before new bone formation. Our report on the efficacy of saracatinib on inhibiting PC-3 osteolytic bone lesions is applicable to patients with prostate cancer metastatic to bone.

Results presented herein show the utility of saracatinib in advanced prostate cancer with mechanistic and in vivo animal studies supporting clinical trials using saracatinib as a treatment for prostate cancer patients with bone metastasis.

C.P. Evans is a consultant for AstraZeneca. No other potential conflicts of interest were disclosed.

We thank Maggie Chiu of Pathology for making sections and staining the animal study samples.

Grant Support: Department of Defense PC10520 and Prostate Cancer Foundation Competitive Award (C.P. Evans).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.