Title: Transforming Growth Factor-β Stimulates Parathyroid Hormone-related Protein and Osteolytic Metastases via Smad and Mitogen-activated Protein Kinase Signaling Pathways
Abstract: Transforming growth factor (TGF)-β promotes breast cancer metastasis to bone. To determine whether the osteolytic factor parathyroid hormone-related protein (PTHrP) is the primary mediator of the tumor response to TGF-β, mice were inoculated with MDA-MB-231 breast cancer cells expressing a constitutively active TGF-β type I receptor. Treatment of the mice with a PTHrP-neutralizing antibody greatly decreased osteolytic bone metastases. There were fewer osteoclasts and significantly decreased tumor area in the antibody-treated mice. TGF-β can signal through both Smad and mitogen-activated protein (MAP) kinase pathways. Stable transfection of wild-type Smad2, Smad3, or Smad4 increased TGF-β-stimulated PTHrP secretion, whereas dominant-negative Smad2, Smad3, or Smad4 only partially reduced TGF-β-stimulated PTHrP secretion. When the cells were treated with a variety of protein kinases inhibitors, only specific inhibitors of the p38 MAP kinase pathway significantly reduced both basal and TGF-β-stimulated PTHrP production. The combination of Smad dominant-negative blockade and p38 MAP kinase inhibition resulted in complete inhibition of TGF-β-stimulated PTHrP production. Furthermore, TGF-β treatment of MDA-MB-231 cells resulted in a rapid phosphorylation of p38 MAP kinase. Thus, the p38 MAP kinase pathway appears to be a major component of Smad-independent signaling by TGF-β and may provide a new molecular target for anti-osteolytic therapy. Transforming growth factor (TGF)-β promotes breast cancer metastasis to bone. To determine whether the osteolytic factor parathyroid hormone-related protein (PTHrP) is the primary mediator of the tumor response to TGF-β, mice were inoculated with MDA-MB-231 breast cancer cells expressing a constitutively active TGF-β type I receptor. Treatment of the mice with a PTHrP-neutralizing antibody greatly decreased osteolytic bone metastases. There were fewer osteoclasts and significantly decreased tumor area in the antibody-treated mice. TGF-β can signal through both Smad and mitogen-activated protein (MAP) kinase pathways. Stable transfection of wild-type Smad2, Smad3, or Smad4 increased TGF-β-stimulated PTHrP secretion, whereas dominant-negative Smad2, Smad3, or Smad4 only partially reduced TGF-β-stimulated PTHrP secretion. When the cells were treated with a variety of protein kinases inhibitors, only specific inhibitors of the p38 MAP kinase pathway significantly reduced both basal and TGF-β-stimulated PTHrP production. The combination of Smad dominant-negative blockade and p38 MAP kinase inhibition resulted in complete inhibition of TGF-β-stimulated PTHrP production. Furthermore, TGF-β treatment of MDA-MB-231 cells resulted in a rapid phosphorylation of p38 MAP kinase. Thus, the p38 MAP kinase pathway appears to be a major component of Smad-independent signaling by TGF-β and may provide a new molecular target for anti-osteolytic therapy. transforming growth factor parathyroid hormone-related protein extracellular signal-regulated kinase c-Jun N-terminal kinase TGF-β-activated kinase 1 mitogen-activated protein MAP kinase/ERK kinase monoclonal antibody MAP kinase kinase Substantial data support major roles for bone-derived TGF-β1 and tumor-derived parathyroid hormone-related protein (PTHrP) in the vicious cycle of local bone destruction that characterizes osteolytic metastases. Tumor-produced PTHrP stimulates osteoclastic bone resorption to result in the bone destruction associated with breast cancer metastases (1Bundred N.J. Ratcliffe W.A. Walker R.A. Coley S. Morrison J.M. Ratcliffe J.G. Br. Med. J. 1991; 303: 1506-1509Crossref PubMed Scopus (73) Google Scholar,2Powell G.J. Southby J. Danks J.A. Stillwell R.G. Hayman J.A. Henderson M.A. Bennett R.C. Martin T.J. Cancer Res. 1991; 51: 3059-3061PubMed Google Scholar). Neutralizing antibodies to PTHrP not only decreased osteoclastic bone resorption but also inhibited the development of metastases to bone by the human breast cancer cell line, MDA-MB-231 (3Guise T.A. Yin J.J. Taylor S.D. Kumagai Y. Dallas M. Boyce B.F. Yoneda T. Mundy G.R. J. Clin. Invest. 1996; 98: 1544-1549Crossref PubMed Scopus (727) Google Scholar). TGF-β, stored in bone matrix (4Hauschka P.V. Mavrakos A.E. Iafrati M.D. Doleman S.E. Klagsbrun M. J. Biol. Chem. 1986; 261: 12665-12674Abstract Full Text PDF PubMed Google Scholar) and released locally in active form during osteoclastic resorption (5Pfeilschifter J. Mundy G.R. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 2024-2028Crossref PubMed Scopus (369) Google Scholar), stimulates PTHrP production by tumor cells (6Kiriyama T. Gillespie M.T. Glatz J.A. Fukumoto S. Moseley J.M. Martin T.J. Mol. Cell. Endocrinol. 1993; 92: 55-62Crossref PubMed Scopus (86) Google Scholar, 7Merryman J.I. DeWille J.W. Werkmeister J.R. Capen C.C. Rosol T.J. Endocrinology. 1994; 134: 2424-2430Crossref PubMed Scopus (68) Google Scholar, 8Southby J. Murphy L.M. Martin T.J. Gillespie M.T. Endocrinology. 1996; 137: 1349-1357Crossref PubMed Scopus (59) Google Scholar). A dominant-negative TGF-β type II receptor (TβRIIΔcyt) stably expressed in the MDA-MB-231 breast cancer line rendered the cells unresponsive to TGF-β and inhibited TGF-β-induced PTHrP secretion and the development of bone metastases in a mouse model. This dominant-negative type II blockade was reversed by a constitutively active TGF-β type I receptor (TβRI(T204D)). Furthermore, transfection of the cDNA for PTHrP into the dominant-negative MDA-MB-231 line also increased PTHrP production and accelerated bone metastases (9Yin J.J. Selander K. Chirgwin J.M. Dallas M. Grubbs B.G. Wieser R. Massague J. Mundy G.R. Guise T.A. J. Clin. Invest. 1999; 103: 197-206Crossref PubMed Scopus (843) Google Scholar). These published data establish that TGF-β in bone can promote osteolysis by increasing PTHrP secretion from breast cancer cells. They do not, however, exclude contributions from other TGF-β-responsive tumor factors. Here we demonstrate that PTHrP is the central mediator of TGF-β-induced osteolytic metastasis. We also show that TGF-β increases PTHrP secretion from MDA-MB-231 cells by signaling through both Smad and p38 MAP kinase pathways. First, to determine whether PTHrP is the major mediator of TGF-β-induced osteolysis, mice were inoculated with an MDA-MB-231 clonal line overexpressing the constitutively active type I TGF-β receptor, TβRI(T204D), and treated with neutralizing PTHrP antibody or control IgG. The mice treated with PTHrP antibody had a significantly lower tumor burden than the control mice, suggesting that the major downstream effector of TGF-β in the development and progression of bone metastases was PTHrP. TGF-β increases PTHrP expression by stabilizing mRNA as well as by transcriptional mechanisms (6Kiriyama T. Gillespie M.T. Glatz J.A. Fukumoto S. Moseley J.M. Martin T.J. Mol. Cell. Endocrinol. 1993; 92: 55-62Crossref PubMed Scopus (86) Google Scholar, 7Merryman J.I. DeWille J.W. Werkmeister J.R. Capen C.C. Rosol T.J. Endocrinology. 1994; 134: 2424-2430Crossref PubMed Scopus (68) Google Scholar, 8Southby J. Murphy L.M. Martin T.J. Gillespie M.T. Endocrinology. 1996; 137: 1349-1357Crossref PubMed Scopus (59) Google Scholar, 10Werkmeister J.R. Blomme E.A. Weckmann M.T. Grone A. McCauley L.K. Wade A.B. O'Rourke J. Capen C.C. Rosol T.J. Endocrine. 1998; 8: 291-299Crossref PubMed Scopus (33) Google Scholar, 11Benitez-Verguizas J. Loarte D. de Miguel F. Esbrit P. Life Sci. 1999; 65: 1807-1816Crossref PubMed Scopus (15) Google Scholar, 12Lindeman R.K. Ballschmieter P. Nordheim A. Dittmer J. J. Biol. Chem. 2001; 276: 46661-46670Abstract Full Text Full Text PDF PubMed Scopus (86) Google Scholar). The signaling pathways through which TGF-β increases PTHrP secretion in breast cancer cells have not been defined, although intracellular mediators known as Smads are indispensable for many of the responses to TGF-β (12Lindeman R.K. Ballschmieter P. Nordheim A. Dittmer J. J. Biol. Chem. 2001; 276: 46661-46670Abstract Full Text Full Text PDF PubMed Scopus (86) Google Scholar, 13Miyazono K. ten Duke P. Heldin C.H. Adv. Immunol. 2000; 75: 115-157Crossref PubMed Google Scholar, 14Zimmerman C.M. Padgett R.W. Gene (Amst.). 2000; 249: 17-30Crossref PubMed Scopus (152) Google Scholar, 15Wrana J.L. Attisano L. Cytokine Growth Factor Rev. 2000; 11: 5-13Crossref PubMed Scopus (232) Google Scholar). However, there is also accumulating evidence that TGF-β signals through other pathways (16Yamaguchi K. Shirakabe K. Shibuya H. Irie K. Oishi I. Ueno N. Taniguchi T. Nishida E. Matsumoto K. Science. 1995; 270: 2008-2011Crossref PubMed Scopus (1188) Google Scholar, 17Hanafusa H. Ninomiya-Tsuji J. Masuyama N. Nishita M. Fujisawa J. Shibuya H. Matsumoto K. Nishida E. J. Biol. Chem. 1999; 274: 27161-27167Abstract Full Text Full Text PDF PubMed Scopus (390) Google Scholar, 18Mulder K.M. Cytokine Growth Factor Rev. 2000; 11: 23-35Crossref PubMed Scopus (389) Google Scholar). TGF-β activates MAP kinase pathway components Ras, ERK1/2, and JNKs (19Frey R.S. Mulder K.M. Cancer Res. 1997; 57: 628-633PubMed Google Scholar, 20Frey R.S. Mulder K.M. Cancer Lett. 1997; 117: 41-50Crossref PubMed Scopus (61) Google Scholar, 21Zhou G. Lee S.C. Yao Z. Tan T.H. J. Biol. Chem. 1999; 274: 13133-13138Abstract Full Text Full Text PDF PubMed Scopus (74) Google Scholar, 22Hocevar B.A. Brown T.L. Howe P.H. EMBO J. 1999; 18: 1345-1356Crossref PubMed Google Scholar, 23Mulder K.M. Morris S.L. J. Biol. Chem. 1992; 267: 5029-5031Abstract Full Text PDF PubMed Google Scholar, 24Hartsough M.T. Mulder K.M. J. Biol. Chem. 1995; 270: 7117-7124Abstract Full Text Full Text PDF PubMed Scopus (277) Google Scholar, 25Hartsough M.T. Frey R.S. Zipfel P.A. Buard A. Cook S.J. McCormick F. Mulder K.M. J. Biol. Chem. 1996; 271: 22368-22375Abstract Full Text Full Text PDF PubMed Scopus (115) Google Scholar, 26Yue J. Buard A. Mulder K.M. Oncogene. 1998; 17: 47-55Crossref PubMed Scopus (35) Google Scholar, 27Kivinen L. Laiho M. Cell Growth Differ. 1999; 10: 621-628PubMed Google Scholar). TGF-β-activated kinase 1 (TAK1) and its upstream activator TAK1-binding protein mediate some responses to TGF-β family members (16Yamaguchi K. Shirakabe K. Shibuya H. Irie K. Oishi I. Ueno N. Taniguchi T. Nishida E. Matsumoto K. Science. 1995; 270: 2008-2011Crossref PubMed Scopus (1188) Google Scholar, 17Hanafusa H. Ninomiya-Tsuji J. Masuyama N. Nishita M. Fujisawa J. Shibuya H. Matsumoto K. Nishida E. J. Biol. Chem. 1999; 274: 27161-27167Abstract Full Text Full Text PDF PubMed Scopus (390) Google Scholar). Activation by TGF-β of p38 MAP kinase, which is downstream of TAK1, has also been reported (28Wang W. Zhou G., Hu, M.C.T. Yao Z. Tan T.H. J. Biol. Chem. 1997; 272: 22771-22775Abstract Full Text Full Text PDF PubMed Scopus (165) Google Scholar, 29Ravanti L. Hakkinen L. Larjava H. Saarialho-Kere U. Foschi M. Han J. Kahari V.M. J. Biol. Chem. 1999; 274: 37292-37300Abstract Full Text Full Text PDF PubMed Scopus (191) Google Scholar). Recently, MAP kinase kinase-independent activation of p38α by a TAK1-binding protein-dependent mechanism was demonstrated (30Ge B. Gram H., Di Padova F. Huang B. New L. Ulevitch R.J. Luo Y. Han J. Science. 2002; 295: 1291-1294Crossref PubMed Scopus (470) Google Scholar). To examine the signaling pathways by which the TGF-β increases the PTHrP production, wild-type and dominant-negative Smad2, Smad3, and Smad4 were stably overexpressed in MDA-MB-231 breast cancer cells, and the changes in PTHrP production were measured. The data supported both Smad-dependent and independent mechanism for the TGF-β stimulation of PTHrP production by breast cancer cells. Specific protein kinase inhibitors were used to determine the Smad-independent signaling of TGF-β to increase PTHrP production. This study indicated that the MAP kinase pathway, and specifically p38 MAP kinase, is a major component of this Smad-independent signaling by TGF-β and provides new molecular targets for anti-osteolytic therapy. MDA-MB-231 cells (American Type Culture Collection) were cultured in Dulbecco's modified Eagle's medium (Invitrogen) containing 10% heat-inactivated fetal calf serum (HyClone Laboratories, Logan, UT), 1% penicillin/streptomycin (Invitrogen), and 1% nonessential amino acids (Invitrogen). The Smad2 wild-type (pcDNA3-FLAG-MADR2) and dominant-negative (pcDNA3-FLAG-MADR2(3S-A)) cDNA expression plasmids were provided by Drs. Wrana and Attisano (University of Toronto, Toronto, Canada) (31Macias-Silva M. Abdollah S. Hoodless P.A. Pirone R. Attisano L. Wrana J.L. Cell. 1996; 87: 1215-1224Abstract Full Text Full Text PDF PubMed Scopus (653) Google Scholar); Smad4 wild-type (pCMV5-DPC4-HA), Smad3 wild-type (pCS2-FLAG-Smad3), Smad3 dominant-negative (pCS2-FLAG-Smad3(3S-A)), and Smad4 dominant-negative (pCMV5-FLAG-DPC4 (1–514)) plasmids were provided by Dr. Massagué, (Memorial Sloan-Kettering Cancer Center, New York, NY) (32Lagna G. Hata A. Hemmati-Brivanlou A. Massague J. Nature. 1996; 383: 832-836Crossref PubMed Scopus (815) Google Scholar), and the dominant-negative pcDNA3-FLAG-Smad3D407E by Dr. Kato, Tokyo, Japan (33Goto D. Yagi K. Inoue H. Iwamoto I. Kawabata M. Miyazono K. Kato M. FEBS Lett. 1998; 430: 201-204Crossref PubMed Scopus (58) Google Scholar). The Smad wild-type and dominant-negative cDNAs were recloned into the pcDNA3 expression vector (Invitrogen). The Smad plasmids and pcDNA3 (empty vector) were transfected into MDA-MB-231 cells using LipofectAMINE PLUSTM reagent (Invitrogen). Single clones were isolated by limiting dilution in the presence of neomycin analogue G418 (Sigma). The clones were assessed by transient transfection with the TGF-β-responsive plasminogen activator inhibitor promoter linked to firefly luciferase (3TP-lux), as well as PTHrP production in response to 5 ng/ml recombinant human TGF-β1 (R & D Systems Inc.) in serum-free Dulbecco's modified Eagle's medium after 24 h of incubation. More than 50 clones were characterized for each construct. The MDA-MB-231 clone overexpressing dominant-negative TGF-β type II receptor (TβRIIΔcyt), which is truncated at the intracytoplasmic domain (34Wieser R. Attisano L. Wrana J.L. Massague J. Mol. Cell. Biol. 1993; 13: 7239-7247Crossref PubMed Google Scholar), was constructed as described previously (9Yin J.J. Selander K. Chirgwin J.M. Dallas M. Grubbs B.G. Wieser R. Massague J. Mundy G.R. Guise T.A. J. Clin. Invest. 1999; 103: 197-206Crossref PubMed Scopus (843) Google Scholar). This clone, MDA-MB-231/TβRIIΔcyt, is unresponsive to TGF-β. Stable expression of a constitutively active TGF-β type I receptor (TβRI(T204D)) (35Wieser R. Wrana J.L. Massague J. EMBO J. 1995; 14: 2199-2208Crossref PubMed Scopus (602) Google Scholar) into MDA-MB-231/TβRIIΔcyt reversed the dominant-negative TGF-β blockade. This stable clone, MDA-MB-231/TβRIIΔcyt + TβRI(T204D), demonstrated constitutive TGF-β activity, even in the absence of ligand (9Yin J.J. Selander K. Chirgwin J.M. Dallas M. Grubbs B.G. Wieser R. Massague J. Mundy G.R. Guise T.A. J. Clin. Invest. 1999; 103: 197-206Crossref PubMed Scopus (843) Google Scholar). The inhibitors were dissolved in Me2SO (cell culture grade; Sigma) and used in the following concentrations: wortmannin (20 μm), herbimycin A (35 nm), KT5720 (100 nm), KT5823 (150 nm), SB203580 (10 μm), SB202190 (10 μm), and PD98059 (20 μm). All of the inhibitors except wortmannin (Sigma) were obtained form Calbiochem. The inhibitors diluted in normal growth medium were added to wells containing confluent MDA-MB-231 cells and incubated for 3 h. TGF-β (5 ng/ml) was added to serum-free medium containing the respective inhibitors and incubated for 24 h, after which time the conditioned media were collected and the cells were counted. The samples were stored at −80 °C until measured. Me2SO (0.05% v/v) diluted in medium was used as a negative control. The optimal inhibitor concentration and incubation time for MAP kinase inhibitors SB203580, SB202190, and PD98059 were determined by time course and dose-response studies. Concentrations for p38 inhibitors SB203580 and SB202190 (1, 5, 10, 20, and 40 μm) and for MEK inhibitor PD98059 (2, 10, 20, 40, and 80 μm) were tested. The concentrations of 10 and 20 μm were selected for p38 inhibitors and MEK inhibitor, respectively, based on the capacity to inhibit PTHrP production without affecting cell viability. Subsequently, incubation times of 6, 12, 24, 30, and 48 h after 3 h of preincubation were used to determine the optimal inhibitor incubation time without affecting cell viability. The results indicated that the 24-h incubation time was optimal for all of the MAP kinase inhibitors used. The animal protocols were approved by the Institutional Animal Care and Use Committee at the University of Texas Health Science Center at San Antonio and were in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Female nude (BALB/c) mice 4 weeks of age were housed in laminar flow isolated hoods. Water supplemented with vitamin K and autoclaved mouse chow were provided ad libitum. Whole blood samples for ionized calcium concentration were obtained by retro-orbital puncture under metofane anesthesia, and the radiographs were taken under mouse mixture (30% ketamine and 20% xylazine in 0.9% NaCl). Tumor inoculation into the left cardiac ventricle was performed as described previously (3Guise T.A. Yin J.J. Taylor S.D. Kumagai Y. Dallas M. Boyce B.F. Yoneda T. Mundy G.R. J. Clin. Invest. 1996; 98: 1544-1549Crossref PubMed Scopus (727) Google Scholar). The mice were inoculated with a tumor cell suspension of clonal MDA-MB-231/TβRIIΔcyt + TβRI(T204D) cells into the left cardiac ventricle on day 0, after base-line radiographs, body weight, and blood Ca2+ were determined. The mice were treated with neutralizing PTHrP mAb directed against PTHrP(1–34) (antibody-producing hybridoma, 3F5, obtained from Dr. T. J. Martin, St. Vincent's Institute, Melbourne, Australia) or isotype matched control IgG (Sigma) (n = 8/group; 1 mg subcutaneously, at the time of tumor inoculation and 1 week later). Ca2+, body weight, and radiographs were monitored weekly for 6 weeks, at which time the mice were sacrificed. All of the bones and soft tissues were fixed in formalin for histological analysis. The radiographs were analyzed as described later in this paper. Neutralizing PTHrP mAb was produced as ascitic fluid in BALB/c mice primed with Pristane (2,6,10,14-tetramethly pentadecane; Sigma) and purified by protein G-agarose chromatography (AmershamBiosciences). Response to TGF-β was measured by transient transfection of the 3TP-lux reporter construct, which contains three copies of a TGF-β-responsive element from the plasminogen activator inhibitor-1 promoter linked to a firefly luciferase-coding sequence. The 3TP-lux reporter is activated by Smad2, Smad3, and Smad4 (31Macias-Silva M. Abdollah S. Hoodless P.A. Pirone R. Attisano L. Wrana J.L. Cell. 1996; 87: 1215-1224Abstract Full Text Full Text PDF PubMed Scopus (653) Google Scholar, 32Lagna G. Hata A. Hemmati-Brivanlou A. Massague J. Nature. 1996; 383: 832-836Crossref PubMed Scopus (815) Google Scholar, 36Zhang Y. Feng X., We, R. Derynck R. Nature. 1996; 383: 168-172Crossref PubMed Scopus (762) Google Scholar, 37Nakao A. Imamura T. Souchelnytskyi S. Kawabata M. Ishisaki A. Oeda E. Tamaki K. Hanai J. Heldin C.H. Miyazono K. ten Dijke P. EMBO J. 1997; 16: 5353-5362Crossref PubMed Scopus (926) Google Scholar). Co-transfected pRL-TK (Promega, Madison, WI) was used as an internal control reporter expressingRenilla luciferase. The Renilla and firefly luciferase activity in cell lysates was measured using the Dual-Luciferase® reporter assay system (Promega) and a Turner TD-20e luminometer (Turner Designs Inc., Sunnyvale, CA). PTHrP concentrations were measured in 24-h conditioned media using a two-site immunoradiometric assay (Nichols Institute Diagnostics, San Juan Capistrano, CA) as reported previously (3Guise T.A. Yin J.J. Taylor S.D. Kumagai Y. Dallas M. Boyce B.F. Yoneda T. Mundy G.R. J. Clin. Invest. 1996; 98: 1544-1549Crossref PubMed Scopus (727) Google Scholar). The standard curve was generated by using Dulbecco's modified Eagle's medium as a standard diluent; 0.3pmol/l was the limit of detection. MDA-MB-231 breast cancer cells were cultured on Petri plates without or with 5 ng/ml of TGF-β for various times (0, 10, 20, and 30 min and 1, 2, 4, 6, 8, 24, 48, and 72 h). After incubation the cells were lysed using cell lysis buffer (New England Biolabs, Beverly, MA) according to the manufacturer's suggestions. Western blotting was performed according to the manufacturer's directions (New England Biolabs) using phospho-p38 MAP kinase (Thr180/Tyr182) and p38 MAP kinase-specific antibodies to detect the phosphorylated and total p38 MAP kinase. Relative phosphorylation was determined using intensity values quantitated with a Kodak Digital ScienceTMelectrophoresis documentation and analysis system 120 (Eastman Kodak Co.). Ca2+concentrations were measured in whole blood using a Ciba Corning 634 ISE Ca2+/pH analyzer (Corning Medical and Scientific, Medfield, MA) as described previously (3Guise T.A. Yin J.J. Taylor S.D. Kumagai Y. Dallas M. Boyce B.F. Yoneda T. Mundy G.R. J. Clin. Invest. 1996; 98: 1544-1549Crossref PubMed Scopus (727) Google Scholar). The animals were x-rayed using a 43855A x-ray system (Faxitron, Buffalo Grove, IL) as described previously (3Guise T.A. Yin J.J. Taylor S.D. Kumagai Y. Dallas M. Boyce B.F. Yoneda T. Mundy G.R. J. Clin. Invest. 1996; 98: 1544-1549Crossref PubMed Scopus (727) Google Scholar). All of the radiographs were evaluated without knowledge of the treatment groups. Quantitation of lesion area was performed using image analysis software (Java, Jandel Video Analysis; Jandel Scientific, Corte Madera, CA). Forelimb and hind limb bones were removed from mice at the time of sacrifice, fixed in 10% buffered formalin, decalcified in 14% EDTA, and embedded in paraffin. The sections were stained with hematoxylin, eosin, orange G, and phloxine. Total tumor area and osteoclast number/mm of tumor/bone interface were measured in midsections of tibiae and femora without knowledge of experimental groups. Histomorphometric analysis was performed using OsteoMeasure System program (Osteometrics Inc., Atlanta, GA). The results are expressed as the means ± S.E. The data were analyzed by analysis of variance followed by Tukey-Kramer post-test. p < 0.05 was considered statistically significant. Six weeks after tumor inoculation with the MDA-MB-231 stable clone expressing the constitutively active TGF-β type I receptor (MDA-MB-231/TβRIIΔcyt + TβRI(T204D)), the number and the area of osteolytic lesions were significantly less in mice treated with PTHrP mAb compared with those treated with control IgG (Fig. 1,A and B). The tumor area was less, and osteoclasts were fewer in tumor-bearing mice treated with PTHrP mAb (Fig. 1, C and D). The body weights were greater in mice treated with PTHrP mAb compared with the control IgG (21.7 ± 0.7 g versus 17.7 ± 0.9 g; p = 0.009). To test whether the major effect of TGF-β on PTHrP production is via the Smad proteins, we stably expressed wild-type and dominant-negative mutants of Smad2, Smad3, and Smad4 in MDA-MB-231. Wild-type Smad2, Smad3, and Smad4 clones, which had enhanced responses to TGF-β, and dominant-negative Smad clones, which had reduced response to TGF-β in the 3TP-lux reporter assay, were further characterized. Fig. 2demonstrates that MDA-MB-231 stable clones that express dominant-negative Smad2, Smad3, or Smad4 did not respond to TGF-β in the 3TP-lux reporter assay. Stable MDA-MB-231 clones that express wild-type or dominant-negative Smad forms were characterized for PTHrP production in response to TGF-β using a two-site PTHrP immunoradiometric assay (Fig. 3). Compared with parental cells and empty vector controls, the MDA-MB-231/TβRIIΔcyt cell line was unresponsive to TGF-β, as reported previously (9Yin J.J. Selander K. Chirgwin J.M. Dallas M. Grubbs B.G. Wieser R. Massague J. Mundy G.R. Guise T.A. J. Clin. Invest. 1999; 103: 197-206Crossref PubMed Scopus (843) Google Scholar). In contrast, overexpression of the dominant-negative Smads (Smad2(3S-A), Smad3(3S-A), Smad3(D407E), or Smad4(1–514)) failed to completely suppress TGF-β-stimulated PTHrP secretion by MDA-MB-231 cells. On the other hand, overexpression of wild-type Smad2, Smad3, and Smad4 enhanced PTHrP production in response to TGF-β. These data suggest that TGF-β signaling to induce PTHrP production is mediated only partially through the Smad pathway. The data suggested that both Smad-dependent and -independent pathways mediated the effect of TGF-β to stimulate PTHrP production by MDA-MB-231 breast cancer cells. We investigated the Smad-independent pathways by treating parental MDA-MB-231 with specific signaling pathway inhibitors. Protein kinase A (KT5720), protein kinase G (KT5823), tyrosine kinase (herbimycin A), and phosphatidylinositol 3-kinase (wortmannin) inhibitors did not affect basal or TGF-β-induced PTHrP secretion (Fig. 4). However, the p38 MAP kinase inhibitors, SB203580 and SB202190, and the MEK1/2 inhibitor, PD98059, significantly reduced both basal and TGF-β-stimulated PTHrP production dose-dependently (Fig. 4). Therefore, these inhibitors were tested in the context of the Smad dominant-negative blockade. In empty vector MDA-MB-231 clones, TGF-β-stimulated PTHrP was effectively decreased by both p38 inhibitors; SB202190 was more potent (Fig. 5 A). The reduction of TGF-β-stimulated PTHrP production by the MEK inhibitor, PD98059, was likely due to decreased basal PTHrP production because the ratio of TGF-β-stimulated PTHrP to untreated was similar to the control. The cells expressing the dominant-negative TGF-β type II receptor (MDA-MB-231/TβRIIΔcyt) did not respond to TGF-β and secreted a low amount of PTHrP that was not affected by either p38 or MEK1/2 inhibitors (Fig. 5 B). p38 inhibition by SB202190 significantly reduced basal and TGF-β-induced PTHrP production by an MDA-MB-231 clone that expressed the constitutively active TGF-β type I receptor (TβRIIΔcyt + TβRI(T204D)), whereas MEK inhibition by PD98059 reduced the basal PTHrP synthesis (Fig. 5 B,inset).Figure 5Effect of MAP kinase inhibitors on PTHrP production in MDA-MB-231 cells expressing dominant-negative Smads. A, empty vector. B, dominant-negative TGF-β type II receptor (TβRIIΔcyt). Inset, constitutively active TGF-β type I receptor, (TβRIIΔcyt + TβRI(T204D)).C, Smad2(3S-A). D, Smad3(3S-A). E, Smad3(D407E). F, Smad4(1–514). The clonal lines were grown to near confluence in 48-well plates, washed, and treated with inhibitors (10 μm SB203580, 10 μm SB202190, and 20 μm PD98059) as described in Fig. 4. The PTHrP concentrations in conditioned media were corrected for cell number. The values represent the means ± S.E. (n= 3/group). ***, the statistical difference between the TGF-β-treated and untreated control; ♦♦♦, represent the statistical difference between the inhibitor and Me2SO control (for *** and ♦♦♦, p < 0.001; for ** and ♦♦,p < 0.01; for * and ♦, p < 0.001).NS, not significant.View Large Image Figure ViewerDownload Hi-res image Download (PPT) The effects of MAP kinase inhibition on PTHrP production by Smad dominant-negative clones are shown in Fig. 5 (C–F). In the presence of Smad2 dominant-negative blockade, p38 inhibition decreased the basal and TGF-β-induced PTHrP production. MEK inhibition reduced basal PTHrP production, but the ratio of TGF-β-stimulated PTHrP to control was similar to no inhibition (Fig. 5 C). The combination of p38 inhibition and Smad2 dominant-negative blockade totally inhibited the TGF-β induction of PTHrP, comparable with that observed in the dominant-negative type II TGF-β receptor. Similar results were obtained with the Smad3 and Smad4 dominant-negative clones (Fig. 5, D–F). However, the clone that expressed the dominant-negative Smad3(D407E) displayed less responsiveness to TGF-β in the presence of p38 inhibition than the Smad3(3S-A) clone (Fig.5 E). The data suggested that TGF-β signals via p38 MAP kinase to induce PTHrP production in MDA-231 cells. We confirmed the involvement of p38 MAP kinase in TGF-β-signaling in parental MDA-MB-231 cells using Western blotting and antibodies specific for nonphosphorylated and phosphorylated p38 MAP kinase. The relative phosphorylation of p38 MAP kinase was increased in MDA-MB-231 cells treated with 5 ng/ml of TGF-β compared with control. The TGF-β-induced phosphorylation of p38 MAP kinase occurred as early as 10 min post-treatment and peaked at 1 h with a 5.5-fold increase. (Fig.6). The role of TGF-β in malignancy is multifunctional (38Massague J. Blain S.W. Lo R.S. Cell. 2000; 103: 295-309Abstract Full Text Full Text PDF PubMed Scopus (2116) Google Scholar, 39Massagué J. Nat. Rev. 2000; 1: 169-178Crossref Scopus (1671) Google Scholar). TGF-β inhibits growth of epithelial and some tumor cells. Certain carcinomas have inactivated or deleted TGF-β receptors or Smad proteins (40Hahn S.A. Schutte M. Hoque A.T. Moskaluk C.A. da Costa L.T. Rozenblum E. Weinstein C.L. Fischer A. Yeo C.J. Hruban R.H. Kern S.E. Science. 1996; 271: 350-353Crossref PubMed Scopus (2189) Google Scholar). In contrast, TGF-β can also promote invasion and metastases (41Kretzschmar M. Doody J. Timokhina I. Massague J. Genes Dev. 1999; 13: 804-816Crossref PubMed Scopus (860) Google Scholar, 42Oft M. Heider K.H. Beug H. Curr. Biol. 1998; 8: 1243-1252Abstract Full Text Full Text PDF PubMed Google Scholar, 43Welch D.R. Fabra A. Nakajima M. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 7678-7682Crossref PubMed Scopus (381) Google Scholar, 44McEarchern J.A. Kobie J.J. Mack V., Wu, R.S. Meade-Tollin L. Arteaga C.L. Dumont N. Besselsen D. Seftor E. Hen