Secreted Protein Acidic, Rich in Cysteine (SPARC), Mediates Cellular Survival of Gliomas through AKT Activation
2004; Elsevier BV; Volume: 279; Issue: 50 Linguagem: Inglês
10.1074/jbc.m409630200
ISSN1083-351X
AutoresQing Shi, Shideng Bao, Jill Maxwell, Elizabeth D. Reese, Henry S. Friedman, Darell D. Bigner, Xiao-Fan Wang, Jeremy N. Rich,
Tópico(s)Bone health and treatments
ResumoSecreted protein acidic, rich in cysteine (SPARC), is an extracellular matrix protein expressed in many advanced cancers, including malignant gliomas. We and others have previously shown that human glioma cell lines engineered to overexpress SPARC adopt an invasive phenotype. We now show that SPARC expression increases cell survival under stress initiated by serum withdrawal through a decrease in apoptosis. Phosphatidylinositol 3-OH kinase/AKT is a potent pro-survival pathway that contributes to the malignancy of gliomas. Cells expressing SPARC display increased AKT activation with decreased caspase 3/7 activity. Exogenous SPARC rapidly induces AKT phosphorylation, an effect that is blocked by a neutralizing SPARC antibody. Furthermore, AKT activation is essential for the anti-apoptotic effects of SPARC as the decreased apoptosis and caspase activity associated with SPARC expression can be blocked with dominant-negative AKT or a specific AKT inhibitor. As tumor cells face stressful microenvironments particularly during the process of invasion, these results suggest that SPARC functions, in part, to promote tumor progression by enabling tumor cells to survive under stressful conditions. Secreted protein acidic, rich in cysteine (SPARC), is an extracellular matrix protein expressed in many advanced cancers, including malignant gliomas. We and others have previously shown that human glioma cell lines engineered to overexpress SPARC adopt an invasive phenotype. We now show that SPARC expression increases cell survival under stress initiated by serum withdrawal through a decrease in apoptosis. Phosphatidylinositol 3-OH kinase/AKT is a potent pro-survival pathway that contributes to the malignancy of gliomas. Cells expressing SPARC display increased AKT activation with decreased caspase 3/7 activity. Exogenous SPARC rapidly induces AKT phosphorylation, an effect that is blocked by a neutralizing SPARC antibody. Furthermore, AKT activation is essential for the anti-apoptotic effects of SPARC as the decreased apoptosis and caspase activity associated with SPARC expression can be blocked with dominant-negative AKT or a specific AKT inhibitor. As tumor cells face stressful microenvironments particularly during the process of invasion, these results suggest that SPARC functions, in part, to promote tumor progression by enabling tumor cells to survive under stressful conditions. Malignant gliomas remain essentially lethal cancers despite maximal therapy because of resistance to all conventional therapies (1Scott C.B. Scarantino C. Urtasun R. Movsas B. Jones C.U. Simpson J.R. Fischbach A.J. Curran Jr., W.J. Int. J. Radiat. Oncol. Biol. Phys. 1998; 40: 51-55Abstract Full Text Full Text PDF PubMed Scopus (282) Google Scholar). Gliomas, like all cancers, share a restricted set of characteristics essential to tumor development and progression (reviewed in Ref. 2Hanahan D. Weinberg R.A. Cell. 2000; 100: 57-70Abstract Full Text Full Text PDF PubMed Scopus (22628) Google Scholar). The ability to resist apoptotic stimuli is prominent among these characteristics. Cancer cells face many noxious stimuli that may induce cell death, including nutrient restriction, hypoxia, acidosis, loss of cell attachment, and genomic instability. Gliomas commonly develop mechanisms through which they resist cell death either through disruption of apoptotic processes or activation of survival signals (3Bogler O. Weller M. Front. Biosci. 2002; 7: E339-E353Crossref PubMed Google Scholar). Microenvironmental cues, including interactions between cells and the extracellular matrix, control cellular survival and resistance to apoptosis. Cancers display alterations of the normal cell-matrix interactions linked to increased proliferation, invasion, and angiogenesis (4Bissell M.J. Radisky D. Nat. Rev. Cancer. 2001; 1: 46-54Crossref PubMed Scopus (1745) Google Scholar). One component of the extracellular matrix is SPARC, 1The abbreviations used are: SPARC, secreted protein acidic, rich in cysteine; PI3K, and phosphoinositide 3-kinase; FBS, fetal bovine serum; DMEM, Dulbecco's modified Eagle's medium; DPBS, Dulbecco's phosphate-buffered saline; PBS, phosphate-buffered saline; PVDF, polyvinylidene difluoride; VEC, vector; BrdUrd, bromodeoxyuridine; EGFR, epidermal growth factor receptor; PDGFRβ, platelet-derived growth factor receptor-β; GSK3, glycogen synthase kinase 3-α/β; IGF1R, insulin-like growth factor-1 receptor. also known as osteonectin or BM-40, which is a 43-kDa, secreted extracellular glycoprotein that plays important roles in development, tissue healing and remodeling, and angiogenesis (5Bradshaw A.D. Sage E.H. J. Clin. Investig. 2001; 107: 1049-1054Crossref PubMed Scopus (524) Google Scholar, 6Framson P.E. Sage E.H. J. Cell. Biochem. 2004; 92: 679-690Crossref PubMed Scopus (223) Google Scholar). SPARC was originally discovered as a component of bone (7Termine J.D. Kleinman H.K. Whitson S.W. Conn K.M. McGarvey M.L. Martin G.R. Cell. 1981; 26: 99-105Abstract Full Text PDF PubMed Scopus (866) Google Scholar) but is also expressed in epithelia exhibiting high rates of turnover (8Porter P.L. Sage E.H. Lane T.F. Funk S.E. Gown A.M. J. Histochem. Cytochem. 1995; 43: 791-800Crossref PubMed Scopus (195) Google Scholar, 9Sage H. Vernon R.B. Decker J. Funk S. Iruela-Arispe M.L. J. Histochem. Cytochem. 1989; 37: 819-829Crossref PubMed Scopus (164) Google Scholar). Targeted disruption of SPARC expression in mice is associated with several phenotypes as follows: early cataract formation, increased wound healing and adipogenesis, and osteopenia (5Bradshaw A.D. Sage E.H. J. Clin. Investig. 2001; 107: 1049-1054Crossref PubMed Scopus (524) Google Scholar). SPARC induces an intermediate stage of cellular adhesion with ablation of focal adhesions, regulates matrix deposition, and induces growth inhibition in endothelial cells (5Bradshaw A.D. Sage E.H. J. Clin. Investig. 2001; 107: 1049-1054Crossref PubMed Scopus (524) Google Scholar, 6Framson P.E. Sage E.H. J. Cell. Biochem. 2004; 92: 679-690Crossref PubMed Scopus (223) Google Scholar, 10Murphy-Ullrich J.E. Lane T.F. Pallero M.A. Sage E.H. J. Cell. Biochem. 1995; 57: 341-350Crossref PubMed Scopus (148) Google Scholar). In addition to its normal physiological role, SPARC has been linked to cancer progression as many cancer types express increased SPARC levels upon invasion or metastasis (5Bradshaw A.D. Sage E.H. J. Clin. Investig. 2001; 107: 1049-1054Crossref PubMed Scopus (524) Google Scholar, 6Framson P.E. Sage E.H. J. Cell. Biochem. 2004; 92: 679-690Crossref PubMed Scopus (223) Google Scholar, 11Rempel S.A. Golembieski W.A. Ge S. Lemke N. Elisevich K. Mikkelsen T. Gutierrez J.A. J. Neuropathol. Exp. Neurol. 1998; 57: 1112-1121Crossref PubMed Scopus (120) Google Scholar, 12Loging W.T. Lal A. Siu I.M. Loney T.L. Wikstrand C.J. Marra M.A. Prange C. Bigner D.D. Strausberg R.L. Riggins G.J. Genome Res. 2000; 10: 1393-1402Crossref PubMed Scopus (88) Google Scholar). Malignant gliomas rarely metastasize but are highly invasive leading to a failure of curative surgery (13Giese A. Bjerkvig R. Berens M.E. Westphal M. J. Clin. Oncol. 2003; 21: 1624-1636Crossref PubMed Scopus (947) Google Scholar). Gliomas express SPARC at sites of invasion and neoangiogenic blood vessels at the brain-tumor interface (11Rempel S.A. Golembieski W.A. Ge S. Lemke N. Elisevich K. Mikkelsen T. Gutierrez J.A. J. Neuropathol. Exp. Neurol. 1998; 57: 1112-1121Crossref PubMed Scopus (120) Google Scholar, 14Menon P.M. Gutierrez J.A. Rempel S.A. Int. J. Oncol. 2000; 17: 683-693PubMed Google Scholar). We and others have shown that malignant glioma cell lines engineered to overexpress SPARC adopt an invasive phenotype both in vitro and in vivo associated with an increased expression of specific matrix metalloproteinases (15Rich J.N. Shi Q. Hjelmeland M. Cummings T.J. Kuan C.T. Bigner D.D. Counter C.M. Wang X.F. J. Biol. Chem. 2003; 278: 15951-15957Abstract Full Text Full Text PDF PubMed Scopus (143) Google Scholar, 16Golembieski W.A. Ge S. Nelson K. Mikkelsen T. Rempel S.A. Int. J. Dev. Neurosci. 1999; 17: 463-472Crossref PubMed Scopus (80) Google Scholar, 17Schultz C. Lemke N. Ge S. Golembieski W.A. Rempel S.A. Cancer Res. 2002; 62: 6270-6277PubMed Google Scholar). SPARC thus represents a potentially important contributor to glioma malignancy. The mechanisms through which SPARC functions in cancer progression remain complex and depend on tumor cell type and the microenvironment. Mice with disrupted SPARC expression are less sensitive to the development of squamous cell carcinomas induced by ultraviolet radiation (18Aycock R.L. Bradshaw A.C. Sage E.H. Starcher B. J. Investig. Dermatol. 2004; 123: 592-599Abstract Full Text Full Text PDF PubMed Scopus (22) Google Scholar), but some tumor xenografts grow more aggressively in SPARC knockout mice with a lack of tumor capsule (19Brekken R.A. Puolakkainen P. Graves D.C. Workman G. Lubkin S.R. Sage E.H. J. Clin. Investig. 2003; 111: 487-495Crossref PubMed Scopus (172) Google Scholar, 20Puolakkainen P.A. Brekken R.A. Muneer S. Sage E.H. Mol. Cancer Res. 2004; 2: 215-224PubMed Google Scholar). Primary endothelial cells may be growth-suppressed in response to SPARC (21Sage E.H. Bassuk J.A. Yost J.C. Folkman M.J. Lane T.F. J. Cell. Biochem. 1995; 57: 127-140Crossref PubMed Scopus (62) Google Scholar), but we have previously shown that glioma cell lines expressing SPARC do not exhibit significant differences with control lines in either cellular proliferation or apoptosis in 10% serum cell culture conditions (15Rich J.N. Shi Q. Hjelmeland M. Cummings T.J. Kuan C.T. Bigner D.D. Counter C.M. Wang X.F. J. Biol. Chem. 2003; 278: 15951-15957Abstract Full Text Full Text PDF PubMed Scopus (143) Google Scholar). However, cells grown in vivo experience much greater restriction of the growth factors found in serum as many growth factors are reversibly bound to the matrix. Glioma cells undergoing invasion may be faced with even more stressful conditions relative to adherent cells as they move along brain white matter tracts (13Giese A. Bjerkvig R. Berens M.E. Westphal M. J. Clin. Oncol. 2003; 21: 1624-1636Crossref PubMed Scopus (947) Google Scholar). Recently, Rempel and co-workers (22Vadlamuri S.V. Media J. Sankey S.S. Nakeff A. Divine G. Rempel S.A. Neuro-oncol. 2003; 5: 244-254Crossref PubMed Scopus (11) Google Scholar) have shown that a glioma cell line engineered to overexpress SPARC exhibits differential cell number and morphology relative to parental control cells in serum-restricted (0.1%) conditions. Although the proliferation of cells expressing SPARC varied from parental lines with different extracellular matrices, no matrix preferentially modified the effect of SPARC (22Vadlamuri S.V. Media J. Sankey S.S. Nakeff A. Divine G. Rempel S.A. Neuro-oncol. 2003; 5: 244-254Crossref PubMed Scopus (11) Google Scholar). As SPARC may directly interact with growth factors or indirectly regulate growth factor receptor pathways (5Bradshaw A.D. Sage E.H. J. Clin. Investig. 2001; 107: 1049-1054Crossref PubMed Scopus (524) Google Scholar), the cellular impact of SPARC may vary with the withdrawal of growth factors and other survival factors in serum. We have now shown that the expression of SPARC by gliomas induces cellular survival in serum-free conditions associated with AKT activation. Thus, SPARC-mediated cell survival may represent a novel mechanism by which PI3K-AKT activity may be induced by malignant gliomas. Furthermore, the ability of tumor cells to survive in stressful conditions may represent a critical aspect of glioma invasion as invading cells may face a loss of external survival signals from the microenvironment. Therefore, SPARC may represent an important therapeutic cancer target as it can modulate critical advanced tumor phenotypes, including invasion and resistance to apoptosis. Cells and Culture—The genetically defined glioma cell line THR was described previously (23Rich J.N. Guo C. McLendon R.E. Bigner D.D. Wang X.F. Counter C.M. Cancer Res. 2001; 61: 3556-3560PubMed Google Scholar). Briefly, normal human astrocytes were transformed with a combination of retroviruses encoding simian virus 40 early T antigen, the human telomerase catalytic subunit hTERT, and an oncogenic Harvey-ras (23Rich J.N. Guo C. McLendon R.E. Bigner D.D. Wang X.F. Counter C.M. Cancer Res. 2001; 61: 3556-3560PubMed Google Scholar). The well characterized human malignant glioma xenograft D54MG is the Duke University subline of A-172 (24Giard D.J. Aaronson S.A. Todaro G.J. Arnstein P. Kersey J.H. Dosik H. Parks W.P. J. Natl. Cancer Inst. 1973; 51: 1417-1423Crossref PubMed Scopus (1869) Google Scholar). U87MG cells were purchased from the American Type Culture Collection (Manassas, VA). THR, D54MG, and U87MG human glioma cell lines that were infected with either empty control retrovirus (VEC) or a retrovirus encoding a 1.5-kb cDNA fragment of SPARC (a generous gift from Sandra Rempel, Henry Ford Hospital) were described previously (15Rich J.N. Shi Q. Hjelmeland M. Cummings T.J. Kuan C.T. Bigner D.D. Counter C.M. Wang X.F. J. Biol. Chem. 2003; 278: 15951-15957Abstract Full Text Full Text PDF PubMed Scopus (143) Google Scholar). Early passage polyclonal cultures selected for antibiotic resistance were used for all experiments. Glioma cultures were maintained in culture in 10-cm tissue culture dishes in DMEM (THR) or Zinc Option (D54MG and U87MG) media containing 10% fetal bovine serum (Invitrogen) and glutamine (Invitrogen) until ready for use. All other tissue culture reagents were purchased from Invitrogen unless otherwise described. Reagents—Human platelet osteonectin (SPARC) and bovine osteonectin (SPARC) were purchased from Hematech (Essex Junction, VT). Recombinant human SPARC was a kind gift of E. Helene Sage (Hope Heart Institute) (25Bradshaw A.D. Bassuk J.A. Francki A. Sage E.H. Mol. Cell. Biol. Res. Commun. 2000; 3: 345-351Crossref PubMed Scopus (32) Google Scholar). The neutralizing anti-SPARC antibody AON33 (αSP303) was a kind gift of Rolf Brekken (University of Texas Southwestern) (26Sweetwyne M.T. Brekken R.A. Workman G. Bradshaw A.D. Carbon J. Siadak A.W. Murri C. Sage E.H. J. Histochem. Cytochem. 2004; 52: 723-733Crossref PubMed Scopus (38) Google Scholar). AKT inhibitor SH-5 and phosphoinositide 3-kinase (PI3K) inhibitors LY294002 and wortmannin were purchased from Calbiochem, dissolved in Me2SO at 10 mm concentration, and stored at –70 °C before use. All other chemicals were purchased from EMD Chemicals (Gibbstown, NJ) unless otherwise described. Cell Number Density Measurements—Glioma cultures were plated in quadruplicate 6-well dishes at various densities for each cell line and allowed to attach overnight in medium with 10% fetal bovine serum (FBS). The following morning, the medium was removed; cells were washed twice with Dulbecco's phosphate-buffered saline (DPBS), and serum-free medium was added. Cultures were visually inspected daily. Cell numbers were determined after trypan blue staining of viable tumor cells in parallel plates. Annexin V Assays—Cells were plated in triplicate in 6-well plates at variable densities in medium containing 10% serum overnight. Cultures were then washed twice with DPBS and then fed with serum-free media. At various incubation times, media were collected and combined with adherent cells harvested by trypsinization. After centrifugation, cells were washed, resuspended, and stained for annexin V and propidium iodide per the manufacturer's instructions (Pharmingen). Samples were analyzed on a BD Biosciences flow cytometer and analyzed with manufacturer's software. Bromodeoxyuridine (BrdUrd) Staining—Cells were plated in individual wells on slides in media containing 10% FBS overnight. Cells were washed twice with DPBS and allowed to grow in serum-free media. Cells were pulsed with BrdUrd for 4 h, fixed with methanol, and stained with diaminobenzidine. The ratio of cells incorporating BrdUrd was calculated using the number of brown cells relative to the total number of cells. Assays were performed in triplicate. Western Analysis—For each assay, a 100-mm plate was lysed in lysis buffer (50 mm Tris-HCl, pH7.5, 150 mm NaCl, 0.5% Nonidet P-40, and 50 mm NaF plus proteinase inhibitors and phosphatase inhibitors) and centrifuged at 14,000 rpm for 5 min at 4 °C. An equal amount of protein was run on polyacrylamide gels (SDS-PAGE), transferred to PVDF membrane (Millipore, Billerica, MA), hybridized, and detected by using an enhanced chemiluminescence system (Pierce). All antibodies were used according to the manufacturer's instructions. Phospho-specific antibodies for AKT (Ser-473), epidermal growth factor receptor (EGFR), platelet-derived growth factor receptor-β (PDGFRβ), and insulin-like growth factor-1 receptor (IGF1R), as well as total AKT antibodies were purchased from Cell Signaling Technology (Beverly, MA). Caspase3/CPP32 and poly(ADP-ribose) polymerase antibodies were purchased from BIOSOURCE International (Camarillo, CA). Phospho-specific ERK1/2 and total ERK1/2 antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Osteonectin (SPARC) antibody was purchased from Hematologic Technologies. An anti-tubulin antibody was purchased from Sigma. Secondary antibodies were goat anti-rabbit from Zymed Laboratories Inc. Flow Cytometric Analysis—Cells were plated into 100-mm plates at a density of 5 × 105 cells per plate, serum-starved overnight after attachment, and then fed with media containing inhibitors for 24 h. Cells were then trypsinized, fixed in 70% ethanol, washed once in DPBS, and resuspended in RNaseA (Sigma, 100 μg/ml) and propidium iodide (Sigma, 50 μg/ml). Samples were analyzed on a FACScan (BD Biosciences) flow cytometer. Each experiment was performed in triplicate. Transfection—Cells were maintained in medium containing 10% FBS until they reached 50–60% confluence. Plasmid DNAs were incubated with FuGENE 6 (Roche Diagnostics) for 30 min and added to cells per the manufacturer's instructions. Transfected cells were maintained for 24 h and then split for subsequent experiments. Caspase-3/7 Activity Assay—Apo-ONE Caspase-3/7 Activity Kits were purchased from Promega (Madison, WI). Assays were performed under the manufacturer's instructions in a 96-well black wall clear bottom tissue culture plate. 4000 cells/well were plated in the assay. Fluorescent intensity was measured at an excitation wavelength of 485 nm and emission wavelength of 520 nm on SpectraMax Gemini plate scanner. Nonradioactive Cell Proliferation Assay—CellTiter96 cell proliferation assay kit was purchased from Promega (Madison, WI). Assays were performed, following the manufacturer's instructions, on a 96-well tissue culture plate. 4000 cells/well were plated in the assay. Plates were read for absorbance on VESAMax plate scanner at wavelengths 570 nm (measure) and 650 nm (background). Nonradioactive AKT Kinase Assay—Cells were plated in 100-mm plates and subjected to serum starvation. After 24 h of incubation with either PBS or human platelet SPARC (2 μg/ml), cell lysates were prepared. Equal protein was incubated with a glycogen synthase kinase 3-α/β (GSK3) fusion protein per the manufacturer's instructions (Cell Signaling). The lysates were resolved by SDS-PAGE, transferred to a PVDF membrane, and hybridized. Chemiluminescence was used to detect the phosphorylated GSK3α/β protein. Statistical Analysis—Wilcoxon Rank Sum Test was used in all analyses (27Kaplan E.L. Meier P. J. Am. Stat. Assoc. 1958; 53: 457-481Crossref Scopus (48517) Google Scholar). SPARC Expression Is Associated with Increased Cellular Survival in Serum-free Conditions—Glioma cells that express SPARC exhibit differences relative to parental cells in cellular morphology and growth on different substrates in low serum conditions (22Vadlamuri S.V. Media J. Sankey S.S. Nakeff A. Divine G. Rempel S.A. Neuro-oncol. 2003; 5: 244-254Crossref PubMed Scopus (11) Google Scholar). We have characterized previously the impact of SPARC expression in several glioma cell lines, and we found a consistent increase in invasion measures both in immunocompromised rodents and in Matrigel invasion assays but little impact of SPARC expression on cellular proliferation or apoptosis in standard culturing conditions (15Rich J.N. Shi Q. Hjelmeland M. Cummings T.J. Kuan C.T. Bigner D.D. Counter C.M. Wang X.F. J. Biol. Chem. 2003; 278: 15951-15957Abstract Full Text Full Text PDF PubMed Scopus (143) Google Scholar). We serendipitously found that glioma cell lines expressing SPARC survived better relative to control cell lines when left unfed for prolonged periods (data not shown). Therefore, we examined the response of glioma cultures expressing SPARC to withdrawal of serum relative to vector controls. In each cell line tested (the genetically defined human glioma cell line (THR) as well as cell lines derived from patient specimens (D54MG and U87MG)), we consistently observed greater numbers of cells in SPARC-expressing cell lines in serum-free conditions relative to vector control cell lines (Fig. 1). The increase in cell number associated with SPARC expression may be due to increased cellular proliferation or resistance to cell death. Pulsed incorporation of bromodeoxyuridine was used to determine the percentage of cells undergoing DNA replication. All cell lines had very low proliferative indices upon serum starvation, but we found no relative difference between SPARC and vector control cell lines (data not shown). In contrast, flow cytometric measurement of cellular apoptosis by annexin V levels in serum-free conditions revealed a consistent decrease in apoptosis with SPARC expression across each glioma line tested (Fig. 2). Cell cycle analysis of glioma cell lines expressing SPARC or vector control was performed using propidium iodide labeled flow cytometric analysis. We confirmed the decrease in apoptotic index with SPARC expression in each cell line as measured by the sub-G0 fraction, but the remaining cell cycle fractions displayed no relative differences between one another (data not shown). In sum, these results suggest that glioma cells expressing SPARC are resistant to apoptosis induced upon withdrawal of growth factors contained in serum.Fig. 2Glioma cell lines expressing SPARC display resistance to apoptosis induced by serum withdrawal. 100,000 glioma cells infected with a control retroviral vector (VEC) or a vector encoding SPARC (SP) were plated into triplicate wells of 6-well tissue culture plates in DMEM containing 10% FBS and serum-starved after 24 h growth. Cells were collected at specific time points and stained with annexin V and propidium iodide per the manufacturer's suggestions (Pharmingen), and annexin V positive/propidium iodide negative or positive cells were quantified by flow cytometric analysis. A–C, representative plots and average annexin V positive-PI negative percentages are displayed. A, THR sublines at day 7 of serum starvation; B, U87MG sublines at day 7; C, D54MG sublines at day 4. *, p < 0.05 compared with vector control by Wilcoxon Rank Sum Test. BrdUrd labeling showed no differences between VEC and SP cell sublines (data not shown).View Large Image Figure ViewerDownload (PPT) SPARC Expression Leads to Increased AKT Phosphorylation in Serum-free Conditions—The underlying mechanisms by which serum may increase cell survival continue to be elucidated. Serum treatment induces activation of several growth factor receptors, recruitment of intracellular signaling mediators to the internal surface of the cell membrane, and initiation of pathways regulating apoptosis. Prominently, PI3K and the downstream effector protein kinase B/AKT are activated in response to serum and decrease the activity of pro-apoptotic proteins (28Kennedy S.G. Wagner A.J. Conzen S.D. Jordan J. Bellacosa A. Tsichlis P.N. Hay N. Genes Dev. 1997; 11: 701-713Crossref PubMed Scopus (980) Google Scholar, 29Andjelkovic M. Jakubowicz T. Cron P. Ming X.F. Han J.W. Hemmings B.A. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 5699-5704Crossref PubMed Scopus (431) Google Scholar). Like many cancers, malignant gliomas display increased AKT activation through activity of growth factor receptors or loss of the tumor suppressor gene PTEN (30Haas-Kogan D. Shalev N. Wong M. Mills G. Yount G. Stokoe D. Curr. Biol. 1998; 8: 1195-1198Abstract Full Text Full Text PDF PubMed Google Scholar, 31Knobbe C.B. Reifenberger G. Brain Pathol. 2003; 13: 507-518Crossref PubMed Scopus (208) Google Scholar, 32Choe G. Horvath S. Cloughesy T.F. Crosby K. Seligson D. Palotie A. Inge L. Smith B.L. Sawyers C.L. Mischel P.S. Cancer Res. 2003; 63: 2742-2746PubMed Google Scholar). The THR cell line expresses a wild type PTEN, whereas D54MG expresses a PTEN protein with deletion of exons 3–9, and U87MG does not express PTEN protein (data not shown). 2B. K. A. Rasheed, personal communication. AKT has been linked not only to apoptotic resistance but also to increased tumor cell invasion (33Kubiatowski T. Jang T. Lachyankar M.B. Salmonsen R. Nabi R.R. Quesenberry P.J. Litofsky N.S. Ross A.H. Recht L.D. J. Neurosurg. 2001; 95: 480-488Crossref PubMed Scopus (104) Google Scholar, 34Kim D. Kim S. Koh H. Yoon S.O. Chung A.S. Cho K.S. Chung J. FASEB J. 2001; 15: 1953-1962Crossref PubMed Scopus (436) Google Scholar). Thus, AKT appeared to be an excellent potential downstream target of SPARC as both SPARC and AKT mediate tumor invasion and resistance to apoptosis. Indeed, glioma cell lines expressing SPARC displayed increased phosphorylation of a key activating residue of AKT (Ser-473) relative to vector cells in the absence of serum (Fig. 3, A and B). In contrast, the phosphorylation of the proproliferative extracellular signal-regulated kinases-1 and-2 (ERK1/2) did not differ between the control and SPARC cell lines (Fig. 3A), suggesting a specific relationship between SPARC and AKT. These findings provide further explanation for the selective impact of SPARC expression on resistance to apoptosis with a lack of proliferative advantage that we have demonstrated. SPARC Expression Associates with Decreased Executor Caspase Activity—Caspases are cysteine proteases that play critical roles as effectors of apoptotic cell death (35Salvesen G.S. Abrams J.M. Oncogene. 2004; 23: 2774-2784Crossref PubMed Scopus (213) Google Scholar). As SPARC expression is associated with a decrease in apoptosis, we expected that SPARC expression would be associated with decreased caspase activity. Indeed, glioma cell lines under serum-starved conditions that constitutively overexpress SPARC displayed significantly lower levels of cleaved (active) caspase 3 and one of its common substrates poly(ADP-ribose) polymerase on Western analysis (Fig. 3C). As further confirmation, glioma cell lines expressing SPARC in the absence of serum also exhibit decreased caspase activity of the executor caspases 3 and 7 using a fluorescent caspase 3/7 substrate assay (Fig. 3D). SPARC Treatment Acutely Induces AKT Phosphorylation— Although we have shown that glioma cell lines that constitutively express SPARC express increased AKT phosphorylation, we sought to define whether the impact of SPARC on AKT is a primary effect or is secondary to other cellular effects of long term SPARC expression. Therefore, we examined the effect of exogenous human SPARC purified from platelets on AKT phosphorylation of the parental glioma cell lines (THR, D54MG, and U87MG) from which our engineered cell lines were derived. SPARC induced a rapid phosphorylation of AKT in 2–10 min on Western analysis depending on which cell line was examined (Fig. 4A), suggesting that extracellular SPARC initiates the activity of signal transduction pathways that result in AKT activation. Corresponding to our findings with cell lines constitutively expressing SPARC, we found that exogenously administered SPARC had no impact on ERK phosphorylation (data not shown). AKT phosphorylation in response to exogenous SPARC was self-limited, returning to near base-line levels in 60–240 min (Fig. 4A and data not shown) in a manner reminiscent of that of ligand-receptor signaling. As SPARC can bind or influence several growth factor receptor pathways, we examined critical phosphorylating events in several mitogenic growth factor receptor pathways commonly active in malignant gliomas. We found that EGFR, IGF1R, and PDGFRβ were not phosphorylated on Western analysis in response to exogenous SPARC treatment rapidly enough or to a sufficient degree to explain the rapid induction of AKT phosphorylation (data not shown). Recent reports suggest that SPARC may activate key intracellular mediators of the transforming growth factor-β pathway, including SMAD3 (36Schiemann B.J. Neil J.R. Schiemann W.P. Mol. Biol. Cell. 2003; 14: 3977-3988Crossref PubMed Scopus (133) Google Scholar), but we found only a slow and modest induction of phosphorylation of SMAD2 and nuclear localization of SMAD3 in U87MG or D54MG in response to exogenous SPARC treatment (data not shown). Therefore, it is unlikely that SPARC activates AKT by direct activation of these growth factor receptors that are frequently linked to glioma pathophysiology. SPARC induces AKT phosphorylation at low doses, suggesting the specificity of this effect (Fig. 4B). Exogenous SPARC treatment also induced an increase in AKT kinase activity as measured by a nonradioactive kinase assay measuring the phosphorylation state of an AKT substrate, GSK3α/β, after AKT immunoprecipitation (Fig. 4C). As SPARC prepared from different tissues are differentially glycosylated (37Kaufmann B. Muller S. Hanisch F.G. Hartmann U. Paulsson M. Maurer P. Zaucke F. Glycobiology. 2004; 14: 609-619Crossref PubMed Scopus (43) Google Scholar) with potential functional differences as a result, we validated that human SPARC prepared ei
Referência(s)