Abstract
c-Src is a proto-oncogene, belonging to the nonreceptor protein kinases family, which plays a prominent role in carcinogenesis. In this study, we tested the hypothesis that c-Src could promote breast cancer metastasis acting on several cell types and that pharmacological disruption of its kinase activity could be beneficial for the treatment of metastases. Female BALB/c-nu/nu mice were subjected to intracardiac injection of the human breast cancer cells MDA-MB-231 (MDA-231), which induced prominent bone and visceral metastases. These were pharmacologically reduced by treatment with the c-Src inhibitor [7-{4-[2-(2-methoxy-ethylamino-ethoxy]-phenyl}-5-(3-methoxy-phenyl)-7H-pyrrolo[2,3-d]pyrimidin-4-ylamine] CGP76030 (100 mg/kg/day p.o.), resulting in decreased morbidity and lethality. Metastases were more severe in mice injected with MDA-231 cells stably transfected with wild-type c-Src (MDA-231-SrcWT), whereas transfection in injected cells of a c-Src kinase-dead dominant-negative construct (MDA-231-SrcDN) resulted in reduced morbidity, lethality, and incidence of metastases similar to the mice treated with the inhibitor. An analogous beneficial effect of c-Src inhibition was observed in subcutaneous and intratibial implanted tumors. In vitro, c-Src suppression reduced MDA-231 cell aggressiveness. It also impaired osteoclast bone resorption both directly and by reducing expression by osteoblasts of the osteoclastogenic cytokines interleukin-1β and interleukin-6, whereas parathyroid hormone-related peptide was not implicated. c-Src was also modestly but consistently involved in the enhancement of endothelial cell proliferation in vitro and angiogenesis in vivo. In conclusion, we propose that c-Src disruption affects the metastatic process and thus is a therapeutic target for the treatment of breast cancer.
Breast cancer is a relatively common tumor, with an estimated incidence of 1.2 million new cases diagnosed world-wide every year (Henderson et al., 1993). The outcome of patients mainly depends on the development of distant metastases (Greenberg et al., 1996). Bone is the principal metastatic site in patients with mammary carcinomas (James et al., 2003), of whom approximately 20% survives for more than 5 years, whereas those with minor metastases in the bone can survive up to 10 years or more. In contrast, visceral metastases, although less common, are more likely to be fatal with a higher risk of early death.
Consistent evidence suggests the involvement of the protooncogene c-Src in the development and progression of many human cancers, including breast carcinomas (Otthenoff-Kalff et al., 1992; Verbeek et al., 1996; Dehm and Bonham, 2004; Ishizawar and Parsons, 2004). c-Src is a nonreceptor tyrosine kinase whose deficiency in mice affects only bone cell function, with no effects in other organs (Soriano et al., 1991; Marzia et al., 2000). Our previous data demonstrated the ability of c-Src inhibitors belonging to the pyrrolopyrimidine class to reduce the malignant activities of prostate cancer cells in vitro (Recchia et al., 2003). c-Src kinase activity is significantly increased in human breast cancer tissues compared with benign breast tumors or adjacent normal breast tissues, and this elevated c-Src activity is correlated with poor metastasis-free survival (Hennipman et al., 1989; Verbeek et al., 1996). A role for c-Src in the development of breast cancer metastases has been elegantly demonstrated by Myoui (2003). However, the cellular mechanisms underlying its involvement in the metastatic disease are poorly understood. As a result, whether or not c-Src is an appropriate target for pharmacological therapy remains to be established.
In this study, we tested the hypothesis that c-Src could promote breast cancer metastatic disease affecting various cell types and that, as a consequence, pharmacological disruption of its kinase activity could be useful for the development of novel therapies for the treatment of both bone and visceral metastases arising from breast malignancy.
Materials and Methods
Materials. DMEM, fetal bovine serum, penicillin, streptomycin, and trypsin were from GIBCO (Uxbridge, UK). Sterile plasticware was purchased from Falcon Becton-Dickinson (Cowley, Oxford, UK) or Costar (Cambridge, MA). ECL kit, Hybond nitrocellulose, [3H]thymidine, and [γ-32P]ATP were from Amersham Pharmacia Biotech (Little Chalfont, Buckinghamshire, UK). Anti-v-Src, anti-pTyr527 Src, and anti-pTyr416 Src antibodies and c-Src wild-type (WT) and c-Src kinase-dead dominant-negative (DN) (c-SrcK296R/Y528F) constructs were obtained from Upstate Biotechnology (Lake Placid, NY). Ultra-Vision detection system anti-Polyvalent HRP/diaminobenzidine kit was from Lab Vision (Scaffold, UK). The Brilliant SYBR Green QPCR master mix was from Stratagene (La Jolla, CA). The anti-pan-phosphotyrosines, anti-actin, anti-Ki-67, and anti-PTHrP antibodies were from Santa Cruz Biotechnology (Heidelberg, Germany). The monoclonal mouse anti-human Factor VIII-related antigen was from BiØmeda (Foster City, CA). TRIzol, lipofectamine, and PLUS reagent were purchased from Invitrogen (Carlsbad, CA). Rabbit muscle enolase and all chemicals of the purest grade were from Sigma-Aldrich (St. Louis, MO).
Cell Lines. The human breast cancer cell line MDA-MB-231 (MDA-231) was obtained from the American Tissue Culture Collection (ATCC, Manassas, VA) and grown in DMEM supplemented with 10% fetal bovine serum (FBS), 100 IU/ml penicillin, 100 μg/ml streptomycin, and 2 mM l-glutamine.
The EAHy926 endothelial cell line (Marr et al., 1997) was the kind gift of Dr. Cora-Jean S. Edgell (Department of Pathology, University of North Carolina, Chapel Hill, NC). Cells were cultured as above plus 1× hypoxanthine/aminopterin/thymidine. Immunocytochemical detection of von Willebrand factor, αVβ3 integrin, and endothelial-specific antigen P1H12 (Solovey et al., 1997) confirmed the endothelial phenotype.
c-Src Inhibitor. The c-Src inhibitor CGP76030 was a substituted 5,7-diphenyl-pyrrolo[2,3d]pyrimidine synthesized in the Chemistry Research Laboratories of Novartis Pharma (Table 1) (Missbach et al., 1999, 2000; Šuša et al., 2005). In vivo administration was performed by oral gavage (Missbach et al., 1999) (100 mg/kg/day) from the day after the intracardiac injection until the end of the experiment. For in vitro treatments, CGP76030 was dissolved in dimethyl sulfoxide at 10 mM and diluted in cell culture medium before use. Controls were carried out with a dimethyl sulfoxide concentration corresponding to the highest dose of test compound (0.2% v/v).
Cell Transfection. The MDA-231 cell line was transfected with the pUSEamp expression vector containing wild-type c-Src (MDA-231-SrcWT) or kinase-dead dominant-negative c-Src (MDA-231-SrcDN) carrying a double mutation in the catalytic site and in regulatory tyrosine-phosphorylation site 528 (K296R/Y528F). Transfected cells were selected for resistance to geneticin, with no clonal selection to avoid clonal variability. In preliminary experiments, empty vector-transfected cells were indistinguishable from nontransfected parental cells; therefore, this latter variant was used as control cell line.
Animals. Four-week-old female-immunocompromised BALB/c-nu/nu mice (Charles River, Milan, Italy) were maintained under sterile conditions and used for all in vivo experiments. Procedures involving animals and their care were conducted in conformity with national and international laws and policies (NIH Guide for the Care and Use of Laboratory Animals, NIH Publication 85-23, 1985; EEC Council Directive 86/609, OJ L 358, 1; 1987 Dec 12; Italian Legislative Decree 116/92, Gazzetta Ufficiale della Repubblica Italiana n. 40; 1992 Feb 18) and were approved by our Institutional Review Board.
In Vivo Experimental Metastases. MDA-231, MDA-231-SrcWT, and MDA-231-SrcDN cells (1 × 105/0.1 ml of PBS) were injected into the left ventricle of BALB/c-nu/nu mice anesthetized with i.p. injection of pentobarbital (60 mg/kg) (eight mice per group) as described by Arguello et al. (1988) and Yoneda et al. (1997). A group of eight mice injected with MDA-231 cells was treated with 100 mg/kg/day of c-Src inhibitor CGP76030 or with vehicle alone. A group of eight mice was injected with PBS as control. Animals were monitored daily for body weight, behavior, and survival. Cachexia was evaluated as body weight decreased. Mice were also weekly subjected to deep anesthesia and X-ray analysis (36 kilovoltage per amperage for 10 s) using a Cabinet X-ray system (Faxitron model 43855A; Faxitron X-Ray Corp., Buffalo Grove, IL) to follow the onset and progression of osteolytic lesions. At the end of the experiment (38 days), mice were sacrificed and subjected to final X-ray analysis and to anatomical dissection for evaluation of bone and visceral metastases, respectively.
Subcutaneous Xenograft Implants. BALB/c-nu/nu mice were anesthetized as described above, and cells (1 × 106/0.1 ml of PBS) were subcutaneously injected in the right flank using a tuberculin syringe with an 18G needle. Xenografts were monitored daily by measuring the average tumor diameter (two perpendicular axes) using a caliper. After 32 days, mice were sacrificed, and tumor mass was excised and weighed.
Xenograft Histology. Subcutaneous tumors were fixed in 4% formaldehyde in 0.1 M phosphate buffer, pH 7.2, and embedded in paraffin. Sections were cut using a Reichert-Jung 1150/Autocut microtome. Slide-mounted tissue sections (4 μm thick) were deparaffinized in xylene and hydrated serially in 100, 95, and 80% ethanol. Endogenous peroxidases were quenched in 3% H2O2 in PBS for 1 h, and then slides were incubated with the anti-Ki-67 or anti-Factor VIII-related antigen primary antibodies for 1 h at room temperature. Sections were washed three times in PBS, and antibody binding was revealed using the Ultra-Vision detection system anti-Polyvalent HRP/diaminobenzidine kit according to the manufacturer's instructions.
Intratibial Implants. BALB/c-nu/nu mice were anesthetized, a syringe with a 27\m=1/2G needle was inserted in the proximal end of the tibia, and 5 × 104 tumor cells suspended in 25 μl of PBS were injected into the intramedullary space. Radiographs were taken as described above at 20 and 32 days after injection.
Evaluation of Osteolytic Lesions. Radiographs were scanned using the Bio-Rad scanning densitometer (Hercules, CA), model GS800, and quantification of the area of interest was done using the Bio-Rad Quantity One image analysis software. For histological examination, tibias were dissected, cleared of soft tissue, and fixed in 4% formaldehyde in 0.1 M phosphate buffer, pH 7.2. Samples were then decalcified in EDTA and embedded in paraffin. Sections were cut and stained with trichrome stain Masson (Sigma-Aldrich kit number HT15-1KT) or, for the typical osteoclast marker, TRAcP (Sigma-Aldrich kit number 85) according to the manufacturer's instructions.
Western Blotting. For protein extraction, cells or tissues were lysed in RIPA buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, and 0.1% SDS) containing protease inhibitors. Proteins were resolved by 10% SDS-PAGE and transferred to nitrocellulose membranes. Blots were probed with the primary antibody for 1 h at room temperature, washed, and incubated with the appropriate HRP-conjugated secondary antibodies for 1 h at room temperature. Protein bands were revealed by ECL detection.
In Vitro c-Src Kinase Assay. One milligram of protein was extracted with RIPA buffer. Fifty microliters of protein G suspensions were incubated for 2 h at 4°C with 5 μg of anti-Src antibody. The agarose beads were subsequently washed five times with ice-cold immunoprecipitation washing buffer (50 mM HEPES, pH 7.3, 1 mM EDTA, 1% Nonidet P-40, 0.25% sodium deoxycholate, 1 mM phenyl-methylsulfonyl fluoride, 9 μg/ml leupeptin, and 0.2 mM aprotinin) and incubated overnight at 4°C with cell lysates. Beads were then rewashed five times with washing buffer, and then immunoprecipitates were washed and resuspended in kinase assay buffer (50 mM HEPES, pH 7.5, and 0.1 M EDTA) and incubated for 20 min at 30°C in the same buffer plus 1:3 ATP mix (0.15 mM ATP, 30 mM MnCl2, and 200 μCi/ml [γ-32P]ATP, specific activity 3000 Ci/mmol, in kinase assay buffer) in the presence of rabbit muscle enolase. Reducing sample buffer was then added, and samples were subjected to 10% SDS-PAGE. Electrophoretic gels were dried and exposed for two days to autoradiography films.
Conventional and Real-Time RT-PCR. Total RNA was extracted using the TRIzol procedure. One microgram of RNA was reverse-transcribed using Moloney murine leukemia virus reverse transcriptase, and the equivalent of 0.1 μg was used for the PCR reactions. For real-time PCR, the Brilliant SYBR Green QPCR master mix was used. PCR conditions and primer pairs used are listed in Tables 2 and 3.
Conditioned Media. MDA-231, MDA-231-SrcWT, and MDA-231-SrcDN cells were allowed to grow in DMEM plus 10% FBS until 80% confluence. The medium was then replaced with serum-free medium, and after 48 h, supernatants were collected and stored at -20°C until use.
Osteoclast Primary Cultures. Primary osteoclasts were differentiated from the bone marrow of 7-day-old CD1 mice. Bone marrow was flushed from the bone cavity of the long bones and minced in DMEM. Cells were recovered, plated in DMEM plus 10% FBS, and cultured up to 7 days in the presence of MDA-231, MDA-231-SrcWT, and MDA-231-SrcDN cell-conditioned media (dilution 1:4). Cultures were fixed in 3% paraformaldehyde in 0.1 M cacodylate buffer, and TRAcP activity was detected histochemically as described above.
Osteoclasts were also differentiated onto bone slices and fixed in 3% paraformaldehyde in 0.1 M cacodylate buffer. Cells were then removed by ultrasonication in 1% sodium hypochlorite, and slices were stained with 0.1% toluidine blue. Pits were counted, and the pit index computed according to Caselli et al. (1997).
Osteoblast Primary Cultures. Calvaria from 7-day-old CD1 mice were removed, cleaned free of soft tissues, and digested three times with 1 mg/ml Clostridium histolyticum type IV collagenase and 0.25% trypsin for 20 min at 37°C, with gentle agitation. Cells from the second and third digestions were plated and grown in DMEM plus 10% FBS. At confluence, cells were trypsinized by standard procedures and plated according to the experimental protocol. These cells expressed the osteoblast markers alkaline phosphatase, Runx-2, parathyroid hormone/PTHrP receptor, type I collagen, and osteocalcin (Marzia et al., 2000).
Cell Proliferation Assay. Cells were plated in 24-well multiplates (8000 cells/well) and grown for 24 h in DMEM plus 10% fetal bovine serum. They were then serum-starved for 24 h in DMEM plus 0.2% BSA and incubated overnight with 2 μCi/ml [3H]thymidine (specific activity 25 Ci/mmol). Cells to be treated with CGP76030 were serum-starved for 24 h in DMEM plus 0.2% BSA and then incubated with the test compound or vehicle alone for an additional 24 h in DMEM plus 0.2% BSA, during which 2 μCi/ml [3H]thymidine was added after 12 h from the beginning of treatment.
At the end of incubation, cells were dissolved in 0.1% SDS, precipitated with 100% trichloroacetic acid, and centrifuged at 3000 rpm for 15 min at 4°C. Pellets were dissolved in 0.1% SDS, and the [3H]thymidine incorporation was measured in a β-counter.
Cell Adhesion. MDA-231 cells were treated for 30 min in suspension with different concentrations of CGP76030 or with vehicle alone, and then cells were plated on 96-well multiplates and allowed to adhere in the presence of the same treatments for 2 h. At the end of incubation, cultures were washed three times with PBS to remove nonadherent cells, fixed with 80% methanol, and stained with 0.5% crystal violet. Crystal violet was then dissolved in 0.1 N sodium citrate, and absorbance was evaluated at 595 nm in an enzyme-linked immunosorbent assay plate reader.
Migration and Invasion Assays. Migration was performed by the modified Boyden-chamber method (Albini et al., 1987). Cells were added to 12-μm polycarbonate filters coated with 4.5 μg/cm2 gelatin in the upper compartment of the Transwell chambers (Corning Life Sciences, Acton, MA). After 6 h (for MDA-231, MDA-231-SrcWT, and MDA-231-SrcDN cells) or 12 h (for EAHy926 cells), filters were stained with hematoxylin/eosin. Cell migration ability was evaluated by counting cells migrated to the lower side of the filters in five randomly chosen fields/filter at 200×. Invasion assays were performed in a similar manner, with the exception that i) the filters were coated with reconstituted Matrigel (35 μg/cm2) and ii) the evaluation was performed after 12 h (for MDA-231, MDA-231-SrcWT, and MDA-231-SrcDN cells) or 18 h (for EAHy926 cells). In both migration and invasion assays, the chemoattractants employed were i) NIH3T3-conditioned media for MDA-231, MDA-231-SrcWT, and MDA-231-SrcDN cells, and ii) MDA-231-, MDA-231-SrcWT-, and MDA-231-SrcDN-conditioned media for EAHy926 cells.
Statistics. All experiments were repeated at least three times. Data are expressed as the mean ± S.E.M. Statistical analysis was performed by one-way analysis of variance, followed by unpaired Student's t test. A p value <0.05 was conventionally considered statistically significant.
Results
Pharmacological Treatment with the c-Src Inhibitor
Myoui et al. (2003) have shown that reduced c-Src activity in breast cancer cells results in reduced tumor development both locally and in distant organs. c-Src is also known for its pivotal role in enhancing osteoclast activity (Miyazaki et al., 2004; Recchia et al., 2004) and maintaining the osteoblasts in a predifferentiation status (Marzia et al., 2000). These latter cell types are involved in the osteolytic vicious circle in association with the breast cancer cells (Roodman, 2004; Yoneda and Hiraga, 2005) and the endothelial cells that supply angiogenesis and regulatory factors (Guise and Mohammad, 2004; Roodman, 2004; Yoneda and Hiraga, 2005). Therefore, given that all of these cell types could be affected by disruption of c-Src tyrosine kinase activity, we tested whether the pharmacological treatment with an anti-c-Src compound could effectively counteract progression of experimental metastatic disease. We had previously demonstrated in an animal model of bone loss the therapeutic efficacy of the c-Src inhibitor GCP76030 (Table 1), which prevents c-Src from binding to ATP (Missbach et al., 2000; Šuša et al., 2005). Therefore, a similar regimen (100 mg/kg/day p.o.) was employed in mice intracardially injected with MDA-231 cells starting the day after cell inoculation. Indeed, in mice treated with CGP76030, we observed that cachexia (measured as body weight wasting; Fig. 1A) was delayed. Specifically, after 27 days from injection, we found 85% incidence of cachexia in control animals versus 12.5% in the treated group, the latter reaching only 37.5% incidence at the end of experiment (Fig. 1A). Accordingly, lethality (Fig. 1B) was also delayed, and its incidence had reduced in treated littermates relative to controls, since at day 31 from cell injection, 62.5% of treated animals were still alive versus 28.5% of control mice. Post mortem examination of visceral organs in mice injected with the tumor cells showed detectable development of lymph node and lung metastases, with a significant reduction of lung metastasis and a trend toward decrease for lymph node lesions in treated mice relative to the control group (Fig. 1C). In control mice, osteolytic bone metastases appeared at day 26 postinoculation and progressively increased to 57% (Fig. 1D). In contrast, in mice receiving the treatment with the c-Src inhibitor, no osteolytic lesions were detected until day 31, and the final incidence at sacrifice (day 38) was not higher than 25%. Western blot of tissue proteins extracted at sacrifice from tibias showed a marked reduction of activating c-Src Tyr416 autophosphorylation, which is indispensable for full c-Src kinase activity, in mice treated with CGP76030 compared with controls (Fig. 1E).
Careful examination indicated that the animals' behavior was unremarkable, and no obvious detrimental effects were noted elsewhere in the body for the time frame of the treatment, in line with the notion that c-Src disruption in mice affects only the skeleton (Soriano et al., 1991).
MDA-MB-231 Stable Transfectants
To investigate the cellular mechanisms affected by c-Src in the metastatic process of MDA-231 cells, we prepared stable transfectants carrying WT or DN c-Src constructs evaluating transfection efficiency by Western blot analysis. Figure 2A (third row) shows that c-Src was abundantly expressed in MDA-231 parental cells and that both transfectants had higher levels of the protein. MDA-231-SrcDN cells, which carried the c-Src Y528F mutant, showed levels of phosphorylation of inactivating Tyr528 and activating Tyr416 comparable to those of the MDA-231 parental cell c-Src (Fig. 2A, second and first rows, respectively). In MDA-231-SrcWT cells, phosphorylation of both Tyr528 and Tyr416 was higher than in MDA-231 and MDA-231-SrcDN cells (Fig. 2A, second and first rows, respectively). Western blots with antipan-phosphotyrosine antibody of the same cell lysates showed a higher phosphorylation pattern in MDA-231-SrcWT cells and a proportionally lower degree of tyrosine phosphorylation in MDA-231-SrcDN cells compared with parental cells (Fig. 2B, top). In vitro kinase assay showed a higher c-Src kinase activity in MDA-231-SrcWT cells relative to MDA-231 (Fig. 2C, top), whereas no activity was found in MDA-231-SrcDN.
In Vivo Cellular Mechanisms
Effect of c-Src on Development of MDA-231 Cell Metastases in Nude Mice. To test whether c-Src modulation directly affects the ability of MDA-231 cells to develop experimental metastases, mice were injected in the left ventricle with the parental cells or with cells carrying the c-Src WT and DN variants. In mice injected with MDA-231 and MDA-231-SrcWT cells, a similar progression and incidence of cachexia was noticed, whereas none of the mice injected with MDA-231-SrcDN cells showed weight wasting during the time frame of the observations (Fig. 2D). Death started slightly earlier, and its incidence was higher in mice injected with MDA-231-SrcWT cells relative to those injected with MDA-231 cells (70 versus 57%, respectively). Noticeably, all animals injected with MDA-231-SrcDN cells survived for the whole length of the experiment (Fig. 2E).
Interestingly, a significant reduction of lung and lymph node metastases was noticed in MDA-231-SrcDN-injected mice relative to the other two groups (Fig. 2F). In mice injected with MDA-231 or MDA-231-SrcWT cells, osteolytic bone metastases appeared at days 20 to 25 postinoculation and progressively increased to 57 and 70% incidence, respectively (Fig. 2G). In contrast, in mice receiving MDA-231-SrcDN cells, no osteolytic lesions were detected until day 30, with a 25% incidence at sacrifice.
Effect of c-Src on Development of Osteolytic Lesions. To investigate the role of MDA-231 cell c-Src in the osteolytic lesions induced by tumor cells, we performed intratibial inoculations of MDA-231, MDA-231-SrcWT, and MDA-231-SrcDN cells. Radiographs taken at 20 and 32 days from injection revealed the lowest frequency of tumor growth and osteolysis in tibias receiving MDA-231-SrcDN cells, whereas the frequency was similarly higher in tibias inoculated with MDA-231 and MDA-231-SrcWT cells (Table 4). Slightly smaller osteolytic areas were observed in those tibias inoculated with MDA-231-SrcDN cells that developed the lesion (50% of the mice) compared with the other two groups (Table 4; Fig. 3a). Histological examination of tibias injected with MDA-231 and MDA-231-SrcWT cells showed wide tumor burden and trabecular bone erosion in the osteolytic areas (Fig. 3b). Tibias endowed with MDA-231-SrcDN cells had smaller tumor mass confined within the medullary cavity (Fig. 3b). Histomorphometric analysis in sections histochemically stained for the osteoclast marker TRAcP showed a significant increase of osteoclast surface/bone surface and osteoclast number/bone surface in mice injected with MDA-231 cells compared with PBS-injected controls. This increase was even greater in mice receiving MDA-231-SrcWT cells, whereas it was smaller in tibias injected with MDA-231-SrcDN cells (Fig. 3, c-d). Similar results were observed in the tibias of mice developing bone metastases by intracardiac injection of the three cell lines (data not shown).
Effect of c-Src on Subcutaneous Growth of MDA-231 Cells. c-Src is a tyrosine kinase known to stimulate cell proliferation. Therefore, we sought to test whether manipulation of c-Src in our transfectant cells changed their ability to grow in vivo when injected subcutaneously. Subcutaneous xenografts of MDA-231 cells grew at a rate appreciable by gross observations. The growth rate of MDA-231-SrcWT cells was significantly higher than that of parental cells, resulting in heavier and bigger tumors (Fig. 4A, a and b). In contrast, subcutaneous tumors formed by MDA-231-SrcDN cells were approximately 40% the weight and the size of those formed by MDA-231 cells (Fig. 4A, a and b), suggesting a lower proliferation rate. In agreement with this hypothesis, immunohistochemical analysis showed increased levels of the Ki-67 proliferation marker in MDA-231-SrcWT tumors (Fig. 4B). Consistently, MDA-231-SrcDN tumors showed the lowest Ki-67 expression among the three groups (Fig. 4B).
Tumor Vascularization. One of the relevant steps for tumor growth and invasion is the capacity of cancer cells to interact with the endothelium and stimulate angiogenesis. Based on this characteristic, we tested the hypothesis that c-Src activity in MDA-231 cells could contribute to stimulation of tumor angiogenesis. Therefore, we evaluated blood vessel development in the subcutaneous tumor xenografts. Immunohistochemical detection of the endothelial marker Factor VIII showed well developed capillaries in MDA-231 tumors (Fig. 4C). A trend toward an increase in microvascular density was observed in the MDA-231-SrcWT tumors (Fig. 4C). Basal stimulation was not c-Src-dependent, as no changes relative to MDA-231 cells were observed in MDA-231-SrcDN tumors.
In Vitro Cellular Mechanisms
Having demonstrated that c-Src tyrosine kinase pharmacological inhibition can delay the development of metastases in vivo, we sought to unravel the underlying cellular mechanisms in vitro.
MDA-231 Cells. c-Src tyrosine kinase activity is involved in many critical cellular functions, and metastatic cells could be affected by c-Src manipulation at different levels. Therefore, we first investigated whether changes in c-Src activity could modulate MDA-231 cell proliferation. The in vitro [3H]thymidine incorporation assay exhibited a significant decrease of MDA-231-SrcDN proliferation rate compared with MDA-231 cells, with a trend toward increased proliferation in MDA-231-SrcWT cell cultures (Fig. 5A).
Malignant cells exhibit the enhanced motility and invasive capacity indispensable for the metastatic process, and c-Src is known to modulate motility by its role in cytoskeletal remodeling. Therefore, we tested whether c-Src manipulation could affect migration and invasion ability in MDA-231 cells. We observed that both were significantly increased in MDA-231-SrcWT cells, whereas they were reduced in MDA-231-SrcDN cells relative to MDA-231 cells (Fig. 5, B and C). Similar to down-regulation of c-Src activity in MDA-231-SrcDN cells, in vitro treatment of MDA-231 parental cells with the c-Src inhibitor CGP76030 caused a concentration-dependent decrease of cell proliferation, migration, and adhesion (Fig. 5, D-F).
Paracrine Stimulation of Osteoclasts. We and others had shown that paracrine factors released by breast cancer cells induce osteoclastogenesis and activate bone resorption (Roodman, 2004; Rucci et al., 2004; Yoneda and Hiraga, 2005). Because of the relevance of this paracrine activity for osteolysis at the bone metastatic site, we evaluated whether the manipulation of c-Src in MDA-231 cells could affect their ability to influence osteoclastogenesis. Indeed, this was not the case, because similar osteoclast formation and bone resorption rates were observed in bone marrow cultures challenged with conditioned media from the three cell lines, independent of their wild-type or mutant c-Src expression (Fig. 6, A and B).
In the context of bone metastasis, however, tumor cells could induce the expression of osteoclastogenic factors in osteoblasts (Roodman, 2004; Rucci et al., 2004; Yoneda and Hiraga, 2005). Therefore, we tested whether c-Src manipulation in MDA-231 cells could modulate their ability to influence osteoblast paracrine activities. Mouse calvarial osteoblast cultures treated with the conditioned media from our MDA-231 and MDA-231-SrcWT cultures showed an increase of IL-1β and IL-6 mRNAs relative to untreated cells. This increase was not noticed in osteoblasts incubated with conditioned media from MDA-231-SrcDN cells (Fig. 6, C and D). M-CSF, GM-CSF, PTHrP, TNF-α, TGF-β, IL-12, IL-18, and Rankl/Opg transcripts were unremarkable with no changes among the three groups of cultures (data not shown). These results suggest an indirect role of MDA-231 cell c-Src in the paracrine stimulation of osteoclast formation via the osteoblast route, which could be selectively mediated by the osteoclastogenic cytokines interleukin-1β and interleukin-6 (Perez et al., 2001). Remarkably, none of the above cytokines was changed in the MDA-231-SrcWT and MDA-231-SrcDN cells relative to MDA-231 parental cells (Fig. 7A). Starting from the results obtained by Myoui et al. (2003), who demonstrated a direct effect of c-Src on PTHrP production and activity, we evaluated PTHrP mRNA (Fig. 7B, left panels) and protein (Fig. 7B, right panels) expression in the MDA-231 cell lines but failed to observe any modulation. Therefore, the MDA-231 paracrine factors that influence osteoblasts and are under c-Src control remain to be elucidated.
Bone Cells. We had previously demonstrated (Recchia et al., 2004) and confirmed in this study (data not shown) that the CGP76030 c-Src inhibitor reduced osteoclast formation and bone resorption and induced osteoclast apoptosis in vivo and in vitro. These results suggest that a direct inhibition of osteoclast activity could contribute to the reduced incidence of osteolytic lesions upon pharmacological inhibition of c-Src.
We had also demonstrated that c-Src inhibition stimulated osteoblast differentiation and bone formation (Marzia et al., 2000). Because of the role of osteoblasts in osteoclastogenesis, we asked whether c-Src inhibition by CGP76030 could directly affect the osteoclastogenic ability of osteoblasts. While reducing osteoblast proliferation and stimulating differentiation, treatment with CGP76030 had no effect on the expression of pro-osteoclastogenic cytokines, including Rankl, IL-1β, IL-6, M-CSF, PTHrP, TGF-β, TNF-α, or anti-osteoclastogenic factors, such as Opg, IL-12, IL-18, and GM-CSF (data not shown).
Endothelial Cells. Due to the marginal effect shown in vivo on tumor vascularization, we next tested in vitro whether c-Src manipulation in tumor cells could affect endothelial activity by paracrine factors. EAHy926 endothelial cells were allowed to grow in the presence of conditioned media from the three tumor cell cultures. EAHy926 proliferation was slightly but consistently higher in the presence of MDA-231-SrcWT-conditioned medium relative to MDA-231 and MDA-231-SrcDN-conditioned media (Fig. 8A), mirroring the trend toward an increased microvascular density showed in vivo in the MDA-231-SrcWT xenografts. Again, basal stimulation by MDA-231 cell-conditioned medium was not dependent on c-Src because no changes were noticed in endothelial cultures incubated with MDA-231-SrcDN cell-conditioned medium.
EAHy926 cells were also allowed to migrate and invade a Matrigel substrate in the presence of tumor cell-conditioned media. MDA-231 cell-conditioned medium showed a potent chemoattractant activity with a significant increase of EAHy926 cell migration and invasion in comparison with cells incubated with nonconditioned medium. However, c-Src modulation in tumor cells did not influence this activity (Fig. 8, B and C), suggesting the release of selective c-Src-dependent paracrine factors slightly stimulating cell growth but not motility and invasion. MDA-231 cells expressed VEGF and βFGF, which could account for the potent paracrine effect on endothelial cells. However, these mRNAs were not modulated by overexpression of SrcWT or SrcDN (Fig. 8D). Therefore, the c-Src-dependent paracrine activity influencing endothelial proliferation remains to be elucidated. Inhibition of c-Src tyrosine kinase activity in the endothelial cell line by the c-Src inhibitor CGP76030 had no effect on proliferation, migration, and invasion, suggesting resistance of the endothelium to direct c-Src inhibition (Fig. 8, E-G).
Discussion
Development of metastases is responsible of poor outcome in patients with breast cancers and other carcinomas. Visceral metastases are fatal and known to cause remarkable morbidity and eventually mortality. Bone metastases are frequent in breast cancers and typically result in extensive painful osteolytic lesions and hypercalcemia (Cifuentes and Pickren, 1979; Elte et al., 1986; Coleman and Rubens, 1987), which severely affect the quality of life of the patient. However, bone metastases can be dormant for many years, and patients carrying only this type of peripheral disease may survive significantly longer than those affected by visceral metastases (Greenberg et al., 1996). Many cell types are involved in the development of the osteolytic lesions, including cancer cells themselves, osteoclasts, osteoblasts, and endothelial cells, which altogether establish a “vicious circle” (Roodman, 2004; Yoneda and Hiraga, 2005). In this study, we have demonstrated that manipulation of c-Src tyrosine kinase activity by various means affects all of these cell types, albeit to a different extent and with variable potency.
Significantly, blockade of c-Src tyrosine kinase activity by a specific pharmacologic inhibitor proved effective in reducing the incidence of metastases both in bone and in visceral organs, thus setting a background for the use of c-Src ATP binding antagonists to prevent or retard this severe complication of breast cancer. CGP76030 is a substituted 5,7-diphenyl-pyrrolo[2,3d]pyrimidine that acts as a potent inhibitor of the c-Src tyrosine kinase activity (Šuša et al., 2005). It shows limited selectivity versus all tested c-Src family members, except c-Yes, for which it has a fairly similar IC50 in enzymatic assays. Lack of selectivity was also shown versus the platelet-derived growth factor receptor and c-Kit, whereas the IC50 was severalfold higher for members of the receptor and nonreceptor tyrosine kinase families as well as for the serine/threonine protein kinases (Missbach et al., 1999; 2000; reviewed in Šuša et al., 2005; summarized in Table 1). Therefore, it is clear that not only inhibition of c-Src but also blockade of a few other tyrosine kinases may contribute to the improvement observed in our experimental models. However, it is worth mentioning that systemic administration of the c-Src inhibitor shared a common trait with the action of c-Src kinase-dead dominant-negative mutant transfected into the tumor cells, suggesting that c-Src is central to the mechanism of action of CGP76030. This latter condition appeared to induce more potent effects than the treatment with the c-Src inhibitor. However, the two situations cannot be fully compared because the drug pharmacokinetics (Table 1) prevents to reproduce steady inhibition as that caused by the constitutive expression of a c-Src dominant-negative transgene.
Remarkably, the pharmacological treatment used in this study promises further developments because it did not induce any obvious detrimental effect in our animals, in line with the notion that targeted disruption of the c-Src gene in mouse strongly affects only the skeleton, with no deleterious effects elsewhere in the body (Soriano et al., 1991). In a recent work, Yezhelyev et al. (2004) demonstrated that c-Src inhibition, either alone or in combination with conventional chemotherapy, shows antitumoral and antimetastatic activity in an orthotropic nude mouse model for human pancreatic cancer. A closely related inhibitor proved effective at inhibiting osteoclast-mediated bone resorption in healthy male volunteers, without significant adverse effects (Hannon et al., 2005), opening up a new avenue for the use in human diseases.
The development of visceral metastases and the incidence of morbidity and mortality were alike in mice inoculated with MDA-231 and with MDA-231-SrcWT cells. In contrast, the onset of bone metastases occurred earlier, and the incidence was higher in mice receiving MDA-231-SrcWT. Our understanding of this difference points to the special role that the tyrosine kinase is likely to play for the homing of tumor cells to the osteomedullary site, which is probably not maximized by the “physiological” expression of the gene in the tumor cells.
Likewise, whereas in vitro migration and invasion of tumor cells were enhanced by overexpression of c-SrcWT, proliferation was not affected. This finding is in sharp contrast with the findings in subcutaneous implants of MDA-231-SrcWT cells, which were larger with a significantly higher expression of the proliferation marker Ki-67. This outcome suggests that in vivo other determinants are likely to affect tumor growth. These may include environmental factors that could converge on the c-Src pathway, thus affecting cell proliferation in a synergistic fashion (Brown and Cooper, 1996).
Much more striking and interesting were the effects observed using MDA-231-SrcDN cells, because all of the parameters evaluated were negatively affected by overexpression of this inactive, kinase-dead mutant, albeit with slightly variable potency. Most importantly, our work demonstrated that the same negative effects could be reproduced both in in vivo and in vitro models receiving the c-Src inhibitor CGP76030.
It is interesting to note that c-Src activity in MDA-231 cells also affected their ability to stimulate the other cells involved in the osteolytic “vicious circle.” Many important mediators are known to activate bone cells and endothelial cells, but none of those more commonly involved and investigated in this study was directly modulated in MDA-231 cells by changes in c-Src. Notably, in contrast with the data from Myoui et al. (2003), who used similar MDA-231 transfectants and injection strategy, in our study, PTHrP was not transcriptionally nor post-transcriptionally reduced upon c-Src inhibition. However, it is interesting to note that the lack of PTHrP changes did not prevent the beneficial effect of c-Src tyrosine kinase inhibition on the MDA-231 cell osteolytic lesions. However, it is clear that as-yet-unknown factors released by MDA-231 cells, seemingly under the control of c-Src, can affect osteoclastogenesis (only via the osteoblasts) and endothelial proliferation (modestly but consistently). It will be a challenge in the future to gain knowledge on the specific MDA-231-secreted factor(s) influenced by c-Src, for example, by the means of global gene expression profiling and/or proteomic analyses. Agents blocking these thus far unrecognized factors could be used in combination with antagonists of the cytokines involved and with c-Src inhibitors, strengthening the emerging concept of personalized, multiple low-dose treatments for cancer-induced complications (Ung et al., 1995; Blumenschein et al., 1997).
In conclusion, we provide compelling evidence that in vivo pharmacological treatment with a c-Src inhibitor could be very promising for the treatment of breast cancer and its metastatic complications.
Acknowledgments
We thank Dr. Rita Di Massimo for editing this manuscript.
Footnotes
-
This work was supported by the E.C. Grant METABRE (LSHM-CT-2003-503049) and by a grant from the Associazione Italiana per la Ricerca sul Cancro (AIRC).
-
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
-
doi:10.1124/jpet.106.102004.
-
ABBREVIATIONS: DMEM, Dulbecco's modified minimum essential medium; WT, wild type; DN, dominant-negative; MDA-231, MDA-MB-231; MDA-231-SrcWT, MDA-231 cells stably transfected with wild-type c-Src; MDA-231-SrcDN, CGP76030, [7-{4-[2-(2-methoxy-ethylamino-ethoxy]-phenyl}-5-(3-methoxy-phenyl)-7H-pyrrolo[2,3-d]pyrimidin-4-ylamine]; MDA-231 cells injected with c-Src kinase-dead dominant-negative construct; ECL, enhanced chemiluminescence; FBS, fetal bovine serum; GM-CSF, granulocyte macrophage colony-stimulating factor; HRP, horseradish peroxidase; M-CSF, macrophage colony-stimulating factor; OPG, osteoprotegerin; PBS, phosphate-buffered saline; RANKL, receptor activator of NF-kB ligand; TRAcP, tartrate-resistant acid phosphatase; VEGF, vascular endothelial growth factor; TGF-β, transforming growth factor-β; PTHrP, parathyroid hormone-related peptide; RIPA, radioimmune precipitation; βFGF, β-fibroblast growth factor; PCR, polymerase chain reaction; TNF-α, tumor necrosis factor-α.
- Received January 31, 2006.
- Accepted April 19, 2006.
- The American Society for Pharmacology and Experimental Therapeutics