Insulin-like Growth Factor-I Is an Autocrine Regulator of
`Chromogranin A Secretion and Growth in Human
`Neuroendocrine Tumor Cells
`(cid:160)
`Götz von Wichert, Peter M. Jehle, Andreas Hoeflich, et al.
`Cancer Res(cid:160)(cid:160)
`
`2000;60:4573-4581.
`
`(cid:160)
`(cid:160)
`(cid:160)
`(cid:160)
`
`(cid:160)
`(cid:160)
`
`Updated version
`(cid:160)
`
`Access the most recent version of this article at:
` http://cancerres.aacrjournals.org/content/60/16/4573
`
`(cid:160)
`
`Cited Articles
`(cid:160)
`Citing articles
`(cid:160)
`
`This article cites by 47 articles, 21 of which you can access for free at:
`http://cancerres.aacrjournals.org/content/60/16/4573.full.html#ref-list-1
`(cid:160)
`This article has been cited by 31 HighWire-hosted articles. Access the articles at:
`
` http://cancerres.aacrjournals.org/content/60/16/4573.full.html#related-urls
`(cid:160)
`
`
`
`E-mail alerts
`(cid:160)
`Reprints and
`Subscriptions
`(cid:160)
`Permissions
`(cid:160)
`
`Sign up to receive free email-alerts
`
` related to this article or journal.
`
`To order reprints of this article or to subscribe to the journal, contact the AACR Publications
`.
`pubs@aacr.org
`Department at
`(cid:160)
`To request permission to re-use all or part of this article, contact the AACR Publications
`.
`permissions@aacr.org
`Department at
`(cid:160)
`
`Roxane Labs., Inc.
`Exhibit 1007
`Page 001
`
`

`
`[CANCER RESEARCH 60, 4573– 4581, August 15, 2000]
`
`Insulin-like Growth Factor-I Is an Autocrine Regulator of Chromogranin A
`Secretion and Growth in Human Neuroendocrine Tumor Cells1
`
`Go¨tz von Wichert, Peter M. Jehle, Andreas Hoeflich, Stefan Koschnick, Henning Dralle, Eckhard Wolf,
`Bertram Wiedenmann, Bernhard O. Boehm, Guido Adler, and Thomas Seufferlein2
`Departments of Internal Medicine I [G. v. W., S. K., B. O. B., G. A., T. S.] and Internal Medicine II [P. M. J.], University of Ulm, D-89081 Ulm, Germany; Department of
`Veterinary Medicine, Genzentrum, D-80539 Munich, Germany [A. H., E. W.]; Department of Internal Medicine/Division of Hepatology and Gastroenterology, Charite, Humboldt
`University of Berlin, D-13086 Berlin, Germany [B. W.]; and Department of Surgery, Martin Luther University, D-06099 Halle, Germany [H. D.]
`
`ABSTRACT
`
`Carcinoid tumors are predominantly found in the gastrointestinal tract
`and are characterized by hypersecretion of various substances, including
`bioamines and neuropeptides, leading to functional tumor disease. Here,
`we demonstrate that human BON carcinoid tumor cells express function-
`ally active insulin-like growth factor-I (IGF-I) receptors and secrete
`IGF-I, suggesting an autocrine action of this growth factor. The IGF-I
`receptor was functionally active. IGF-I stimulated phosphatidylinositol
`3-kinase (PI3-kinase), p70 S6 kinase (p70s6k), and extracellular signal-
`regulated kinase 2 activity in BON cells. Furthermore, immunoneutral-
`ization of endogenously released IGF-I markedly reduced the high basal
`activity of p70s6k and extracellular signal-regulated kinase 2 in serum-
`starved BON cells. Exogenously added IGF-I induced a marked increase
`in chromogranin A secretion, a marker protein for neuroendocrine secre-
`tion, by a process that was largely dependent on PI3-kinase activity. In
`addition, immunoneutralization of endogenously released IGF-I markedly
`reduced basal chromogranin A release by BON cells. Thus, the autocrine
`IGF-I loop regulates basal neuroendocrine secretion in BON cells. Next,
`we investigated the role of IGF-I as a growth promoting agent for BON
`cells. Our data demonstrate that IGF-I stimulates anchorage-dependent
`and anchorage-independent growth of BON cells by a pathway that
`involves PI3-kinase, mammalian target of rapamycin/p70s6k, and mitogen-
`activated protein kinase kinase 1 activity. Interestingly, mitogen-activated
`protein kinase kinase 1 activity was less important for anchorage-inde-
`pendent growth of BON cells. Endogenously released IGF-I was found to
`be largely responsible for autonomous growth of BON cells in serum-free
`medium and for the constitutive expression of cyclin D1 in these cells. In
`conclusion, IGF-I is a major autocrine regulator of neuroendocrine secre-
`tion and growth of human BON neuroendocrine tumor cells. Because our
`data also demonstrate that a significant proportion of neuroendocrine
`tumors express the IGF-I receptor and its ligand, interference with this
`pathway could be useful in the treatment of hypersecretion syndromes
`and growth of human neuroendocrine tumors.
`
`INTRODUCTION
`
`Carcinoid tumors are neuroendocrine neoplasms that are derived from
`neuroectodermal cells of the neural crest (1). These tumors are predom-
`inantly found in the gastrointestinal tract, although they may arise in
`various organs throughout the body. Carcinoids are characterized by
`hypersecretion of various substances, including bioamines and neuropep-
`tides, leading to functional tumor disease. BON cells have been estab-
`lished from a human pancreatic carcinoid tumor and are a useful model
`to study the biology of neuroendocrine tumors in vitro (2).
`Neuropeptides (3) and polypeptide growth factors, such as nerve
`growth factor and fibroblast growth factor (4, 5), have been implicated
`
`Received 4/13/00; accepted 6/19/00.
`The costs of publication of this article were defrayed in part by the payment of page
`charges. This article must therefore be hereby marked advertisement in accordance with
`18 U.S.C. Section 1734 solely to indicate this fact.
`1 This work was supported by Deutsche Forschungsgemeinschaft Grant SFB 518/B3
`(to T. S.) and the Landesforschungsschwerpunkt Baden-Wu¨rttemberg Grant “Modulation
`von Wachstumsfaktoren als Therapieprinzip” (to P. M. J).
`2 To whom requests for reprints should be addressed, at Abteilung Innere Medizin I,
`Medizinische Universita¨tsklinik Ulm, Robert Koch Strasse 8, D-89081 Ulm, Germany.
`Phone: 49-731-50201; Fax: 49-731-5024302; E-mail: thomas.seufferlein@medizin.uni-
`ulm.de.
`
`in the regulation of neuroendocrine tumor cell growth. In addition, the
`presence of IGF-I3 and the IGF-I receptor has been reported in
`neuroendocrine tumors, such as midgut carcinoids (6, 7). However,
`the signaling pathways induced by IGF-I and its precise role for
`secretion and/or growth in human neuroendocrine tumors are largely
`unknown. IGF-I is a 70-amino acid peptide closely related to insulin
`that binds to distinct high affinity receptors with intrinsic tyrosine
`kinase activity. Upon binding to the IGF-I receptor, IGF-I stimulates
`cell cycle progression and growth in various cell lines (8 –10). In
`addition, IGF-I and the IGF-I receptor have been implicated in mul-
`tistage carcinogenesis: overexpression of the IGF-I receptor or its
`ligand IGF-I causes abnormal growth, cellular transformation, inhibi-
`tion of apoptosis, and spontaneous tumor formation in transgenic mice
`(11, 12). Furthermore, expression of the IGF-I receptor is crucial for
`tumorigenesis in athymic mice (10, 13).
`Here, we report that human BON carcinoid cells express function-
`ally active IGF-I receptors and secrete IGF-I. Exogenously added
`IGF-I stimulated PI3-kinase, p70s6k, and ERK activity in BON cells.
`The endogenously released IGF-I was found to be largely responsible
`for the high basal activity of p70s6k and ERK2 in serum-starved BON
`cells. Exogenously added IGF-I markedly increased chromogranin A
`secretion by BON cells by a PI3-kinase-dependent pathway. This
`kinase also mediated autocrine IGF-I secretion. Immunoneutralization
`of endogenously released IGF-I substantially reduced basal chro-
`mogranin A release. These data demonstrate, for the first time, the
`existence of an autocrine IGF-I loop regulating neuroendocrine secre-
`tion in a carcinoid tumor cell line. In addition, both exogenously
`added and endogenously released IGF-I stimulated growth of BON
`cells by a PI3-kinase, p70s6k, and MEK-1-dependent signaling path-
`way. At the level of the cell cycle, endogenously released IGF-I was
`found to selectively regulate the expression of cyclin D1 and p27kip1.
`Thus, neuroendocrine secretion and autonomous growth of human
`BON tumor cells are largely regulated by endogenously released
`IGF-I. Because the presence of IGF-I and the IGF-I receptor can be
`demonstrated in various neuroendocrine tumors, the IGF-I signaling
`pathway is a potential novel target for the treatment of hypersecretion
`syndromes and growth of these tumors.
`
`MATERIALS AND METHODS
`
`Cell Culture. Human BON carcinoid tumor cells were maintained in DNM
`medium supplemented with 10% (v/v) fetal bovine serum in a humidified
`atmosphere of 5% CO2: 95% air at 37°C and passaged every 4 days.
`MiaPaCa-2, COS, LCC18, and HEK 293 cells were maintained in DMEM
`supplemented with 10% (v/v) fetal bovine serum in a humidified atmosphere
`of 5% CO2:95% air at 37°C and passaged every 3 days.
`125I-IGF-I Binding Assays. IGF-I
`equilibrium-competition-inhibition
`binding studies were performed as described previously (14). Briefly, BON
`
`3 The abbreviations used are: IGF, insulin-like growth factor; DNM, Dulbecco’s-F12
`nut mix; ERK, extracellular signal-regulated kinase; IGFBP, IGF-binding protein; IGF-I,
`type 1 insulin-like growth factor; IRS, insulin receptor substrate; MEK, mitogen-activated
`protein kinase kinase; PDB, phorbol 12,13-dibutyrate; PKC, protein kinase C; PI3-kinase,
`phosphatidylinositol 3-kinase.
`4573
`
`Roxane Labs., Inc.
`Exhibit 1007
`Page 002
`
`

`
`IGF-I IS AN AUTOCRINE REGULATOR OF CHROMOGRANIN A
`
`cells (1–2 3 106/dish) were washed twice with PBS and once with binding
`o-phosphoric acid, immersed in acetone, and dried before scintillation count-
`buffer and then incubated for 1 h atroom temperature with 125I-IGF-I (10 pM),
`ing. The average radioactivity of two blank samples containing no immune
`complex was subtracted from the result of each sample.
`and different concentrations of unlabeled IGF-I, r3-IGF-I, and insulin (each at
`10 pM to 0.1 mM). To assess whether binding of IGF-I changes during cell
`Measurement of Chromogranin A. To determine chromogranin A secre-
`proliferation, binding assays with 125I-IGF-I or 125I-rIGF-I were performed in
`tion, BON cells were incubated in Krebs-Ringer-HEPES buffer for 2 h at37°C.
`Cells were subsequently treated with IGF-I for 25 min at 37°C in the absence
`BON cells cultured for 24, 36, 48, and 72 h. At the end of each experiment,
`or presence of various inhibitors as indicated in the legend to Fig. 4. The
`cells were washed three times with PBS containing 0.1% BSA. In all exper-
`supernatant was then aspirated and stored at – 80°C until assayed. Chromogra-
`iments, degradation of tracer was less than 10%, thus excluding differences in
`nin A was subsequently determined in the supernatants using a specific
`binding caused by hormone degradation. Cell bound and free intact activity
`chromogranin A ELISA. Release is expressed as fold stimulation above
`were counted in an automatic gamma counter (Berthold, Munich, Germany)
`untreated controls.
`with 70% efficiency. Specific binding was determined by subtracting the
`amount of 125I-IGF-I or 125I-rIGF-I unspecifically bound (,0.5%) in the
`Determination of IGF-I Secretion. To characterize the autocrine release
`presence of 0.1 mM IGF-I and ranged between 5 and 10% in all experiments.
`of IGF-I, cells were washed three times in serum-free medium and incubated
`in serum-free medium for up to 7 days in the presence or absence of various
`A computer-assisted curve fitting program was used to determine the concen-
`tration of unlabeled peptide yielding a 50% inhibition (IC50) of 125I-tracer
`compounds as indicated. Medium was aspirated on specific days as indicated
`in the legend to Figs. 2 and 4 and stored at – 80°C until assayed. IGF-I
`binding (15). As described by Scatchard (16), the ratio of bound to free
`concentration was determined using a specific RIA. Values are expressed as ng
`hormone was plotted as a function of total hormone, and the number of binding
`of IGF-I/ml of supernatant.
`sites per cell and binding affinity (Kd values) were calculated.
`Analysis of IGFBP-2 Expression in BON Cell Conditioned Media.
`Western Blotting. BON cells were washed twice in serum-free DNM
`Conditioned media were analyzed by Western ligand blot analysis according to
`and incubated in fresh serum-free medium for further 24 h. Cells were then
`the method of Hossenlopp (17). Briefly, media were concentrated as described
`treated with factors as indicated in the legends to Figs. 3 and 6 and lysed
`in 50 mM Tris-HCl, 5 mM EDTA, 100 mM NaCl, 40 mM b-glycerophos-
`previously (18), diluted 1:5 with sample buffer [50 mM Na2HPO4, pH 7.0, 1%
`(w/v) SDS, 50% (w/v) glycerin] and boiled (5 min), and proteins were
`phate, 50 mM NaF, 1 mM Na3VO4, 1% Triton X-100, 1 mM phenylmeth-
`ylsulfonyl fluoride, 10 mg/ml aprotinin, 10 mg/ml leupeptin, pH 7.6 (lysis
`separated by SDS-PAGE. Separated proteins were transferred to a nitrocellu-
`lose membrane. The blots were blocked with 1% fish gelatin and incubated
`buffer). For detection of the IGF-I receptor in HEK 293, BON, COS,
`with 125I-IGF-II (106 cpm per blot). Binding proteins were visualized on a
`MiaPaCa-2, and LCC18 cells, serum-starved cells were lysed in lysis
`buffer, and protein content was determined. Proteins were subsequently
`PhosphorImager Storm (Molecular Dynamics, Krefeld, Germany). All hybrid-
`extracted in 53 SDS-PAGE sample buffer. Equal amounts of proteins were
`ization and washing steps were performed at 4°C. IGFBP-2 in the conditioned
`further analyzed by SDS-PAGE and Western blotting with a polyclonal
`media was identified by Western immunoblot analysis using antiserum to
`anti-IGF-I receptor antibody, followed by ECL detection. For p70s6k mo-
`human IGFBP-2 (kindly provided by Dr. M. Elmlinger, Universita¨ts-Kinder-
`bility shift assays and detection of cyclin D1, cyclin E, and p27kip1, cells
`klinik Tu¨bingen, Tu¨bingen, Germany). Membranes were prepared as described
`above and incubated with human IGFBP-2 antiserum (1/1000) for 1 h and with
`were treated as indicated in the legends to Figs. 3 and 6 and lysed in
`peroxidase-coupled antirabbit IgG antibody. Signals were generated using
`SDS-PAGE sample buffer, and samples were further analyzed by SDS-
`diaminobenzidine.
`PAGE and Western blotting using specific antisera to these proteins.
`Immunostaining of Cells and Cryostat Sections. Serum-starved cultures
`PI3-kinase Assays. Serum-starved BON cells were incubated with IGF-I
`of BON cells were washed with PBS and fixed in 4% formaldehyde at room
`for 10 min as indicated, subsequently lysed in lysis buffer, and incubated
`temperature for 20 min. Cells were then permeabilized with 0.2% Triton X-100
`with an antiphosphotyrosine antibody for 2 h with antimouse IgG agarose
`and stained with specific anti-IGF-receptor or anti-IGF-I antibodies for 1 h
`added for the second h. Immunoprecipitates were subsequently washed
`followed by detection with an Alexa-labeled secondary antibody. Samples
`three times with lysis buffer, twice with Buffer A (0.5 M LiCl, 0.1 M Tris,
`were further analyzed by confocal microscopy. Cryostat sections of neuroen-
`pH 7.4) and once with Buffer B (10 mM Tris-HCl, pH 7.4, 100 mM NaCl,
`docrine tumor specimens obtained endoscopically or surgically were fixed in
`1 mM EDTA). Immunoprecipitates were subsequently incubated in the
`kinase reaction mix containing Buffer B, 5 mM MgCl2, 50 mM ATP, 5
`4% ethanol-free formaldehyde at room temperature for 30 min. Tissue sections
`mCi/ml [g-32P]ATP, and 20 mg of phosphatidylinositol in 5 mM HEPES,
`were then permeabilized with 0.1% Triton X-100 for 10 min and preblocked
`with 2% BSA (w/v) for 2 h atroom temperature. Sections were subsequently
`pH 7.4, per condition for 25 min at 25°C. Reactions were stopped by adding
`100 ml of 1 M HCl and 200 ml of a mixture of CHCl3 and methanol (1:1,
`incubated overnight in 2% BSA (w/v) with a monoclonal anti-IGF-I antibody
`or a polyclonal anti-IGF-I receptor antibody. Tissues were then incubated with
`v/v). Samples were mixed by vortexing and subsequently briefly centri-
`Alexa-labeled antirabbit or antimouse antibodies and further analyzed by
`fuged. The CHCl3 phase was further purified by mixing with HCl/methanol
`immunofluorescence microscopy.
`(1:1, v/v). Finally the CHCl3 phase was transferred into a new tube, dried
`Growth Assay. BON cells, 3 days postpassage, were washed in serum-free
`under N2, and subsequently run on TLC plates in a mixture containing H2O,
`DNM, trypsinized, and resuspended in serum-free DNM. Cells were plated at
`CHCl3, methanol, acetone, and glacial acetic acid (16:30:26:30:24, v/v).
`a density of 1 3 104 cells in 1 ml of serum-free DNM in the presence or
`TLC plates were then developed by autoradiography. Autoradiographs
`absence of 20 ng/ml rapamycin in duplicate. At the times indicated in the
`were scanned using the UMAX Vistascan program (version 3.1).
`p70s6k and ERK2 Immune Complex Kinase Assays. Serum-starved
`figure legends, cell number was determined using a cell counting chamber.
`Clonogenic Assay. BON cells were washed, trypsinized, and resuspended
`BON cells were incubated with IGF-I in the presence or absence of rapamycin
`or PD 098059 as indicated in the legend to Fig. 3. Controls received an
`in DNM. Cell number was determined using a cell counting chamber. Cells
`(3 3 104) were mixed with serum-free DNM containing 0.3% agarose in the
`equivalent amount of solvent. Cells were then lysed at 4°C in 1 ml of lysis
`buffer. Immunoprecipitations were performed at 4°C using an anti-p70s6K
`presence or absence of rapamycin at the concentrations indicated and layered
`antibody or an anti-ERK2 antibody for 2 h, with protein A-agarose added for
`over a solid base of 0.5% agarose in serum-free DNM in the presence or
`the second hour. Immune complexes were washed three times in lysis buffer
`absence of rapamycin at the same concentrations in 33-mm dishes. The
`and once with p70s6k kinase buffer (20 mM HEPES, pH 7.4, 10 mM MgCl2, 1
`cultures were incubated in humidified 5% CO2:95% air at 37°C for 14 days
`and then stained with the vital stain nitroblue tetrazolium. Colonies of .120
`mM DTT, and 10 mM b-glycerophosphate) or ERK kinase buffer (15 mM
`mm in diameter (20 cells) were counted using a microscope.
`MgCl2, 15 mM Tris-HCl, pH 7.4). Kinase reactions were performed by resus-
`pending the protein A-Sepharose pellets in 25 ml of kinase assay mixture
`Statistical Analysis. Cerenkov counts of the ERK and p70s6k immune
`complex kinase assays (n 5 6 for each condition) were tested for normal
`containing the appropriate kinase buffer with 0.2 mM S6 peptide (RRRLSS-
`LRA) or myelin basic protein, 20 mM ATP, 5 mCi/ml [g-32P]ATP, 2 mM
`distribution and extreme values. The levels of significance were determined by
`cAMP-dependent protein kinase inhibitor peptide, and 100 nM microcystin LR.
`Pearson correlation.
`Materials. The monoclonal antibody against Tyr(P) (clone 4G10), the
`Incubations were performed under linear assay conditions at 30°C for 20 min
`and terminated by spotting 25 ml of the supernatant onto Whatman p81
`monoclonal anti-IGF-I antibody used for the immunoneutralization experi-
`chromatography paper. Papers were washed four times for 5 min in 0.5%
`ments, and the polyclonal antibodies against IRS-2, the p85 subunit of PI3-
`4574
`
`Roxane Labs., Inc.
`Exhibit 1007
`Page 003
`
`

`
`IGF-I IS AN AUTOCRINE REGULATOR OF CHROMOGRANIN A
`
`kinase, and the N-terminally directed anti-p70s6k polyclonal antibody used to
`determine p70s6k activity were obtained from Upstate Biotechnology, Inc.
`(Lake Placid, NY). The polyclonal antibodies directed against the IGF-I
`receptor, IRS-1, p70s6k, p27kip1, cyclin D1, and cyclin E were from Santa Cruz
`Biotechnology (Santa Cruz, CA). Peroxidase-coupled antirabbit IgG antibody
`was from Dianova (Hamburg, Germany). The polyclonal anti-ERK2 antibody
`was the kind gift of Dr. Jo van Lint (Katholieke Universiteit Leuven, Leuven,
`Belgium). Rapamycin, LY 294002, and GF 109203X were from Calbiochem-
`Novabiochem (Schwalbach, Germany). PD 098059 was from New England
`Biolabs (Schwalbach, Germany). Protein A-Sepharose was obtained from
`Roche Molecular Biochemicals (Mannheim, Germany). Human IGF-I, ECL
`reagent, and [g-32P]ATP were obtained from Amersham Pharmacia Biotech
`(Freiburg, Germany). The Alexa red-labeled and Alexa green-labeled anti-
`mouse and antirabbit IgG antibodies were from Molecular Probes (Leiden, the
`Netherlands). The ELISA kit to determine chromogranin A in the supernatants
`was from DAKO Diagnostica GmbH (Hamburg, Germany). The RIA to
`determine IGF-I in the supernatants was from Nichols Institute Diagnostics
`(San Juan, CA). Nitrocellulose membranes and polyvinylidene difluoride
`membranes for Western blotting were obtained from Millipore (Eschborn,
`Germany). All other reagents were of the purest grade available.
`
`RESULTS
`Human BON Carcinoid Tumor Cells Express Functional IGF-I
`Receptors and Secrete IGFBP-2. The expression of the IGF-I re-
`ceptors in BON cells was examined by Western blotting. A specific
`anti-IGF-I receptor antibody detected a single band migrating with a
`Mr of 90,000 in BON cells corresponding to the b chain of the IGF-I
`receptor. Comparison of IGF-I receptor expression in BON cells with
`other human (HEK 293, MiaPaCa-2, and LCC18) or monkey (COS)
`cell lines by Western blotting revealed that IGF-I receptor expression
`in BON cells was comparable to cell lines that exhibit high levels of
`IGF-I receptor expression (Fig. 1A, top panel). IGF-I receptor immu-
`noreactivity was mainly detected at the plasma membrane and around
`the nucleus (Fig. 1A, bottom panels).
`Binding of 125I-IGF-I to the IGF-I receptor was assessed using
`increasing concentrations of IGF-I, r3-IGF-I, and insulin as compet-
`itive ligands. Consistent with typical IGF-I receptors (Fig. 1B, top
`panel), IGF-I competed for 125I-IGF-I binding well within the nano-
`molar range (IC50, 0.72 6 0.1 nM), whereas insulin showed a 100-fold
`
`Fig. 1. BON cells express IGF-I receptors. A, top panel, serum-starved cultures of 293, COS, BON, LCC18 (LCC), and MiaPaCa-2 (Mia) cells were lysed in lysis buffer, and proteins
`were extracted in 53 SDS-PAGE sample buffer. Equal amounts of proteins were separated by SDS-PAGE and further analyzed by Western blotting using a specific polyclonal
`anti-IGF-I receptor antibody. A, bottom panels, subcellular localization of the IGF-I receptor was assessed by immunocytochemistry. Staining of the plasma membrane is indicated by
`an arrow (right panel). No specific signal could be detected when cells were stained using only the secondary antibody (left panel). B, top panel, 125I-IGF-I competitive binding. BON
`cells (1.2 3 106/dish) cultured for 48 h were incubated with 10 pM 125I-IGF-I and increasing concentrations (10 pM to 0.1 mM) of recombinant IGF-I, r3-IGF-I, or insulin in assay buffer
`for 1 h at20°C. Consistent with typical IGF-I receptors, IGF-I and r3-IGF-I competed for 125I-IGF-I binding with approximately 100-fold higher affinity than insulin. B, middle panel,
`5 0.77 6 0.1 nM) could be demonstrated by Scatchard analysis.
`Scatchard plot of 125I-IGF-I binding. One single class of cell membrane IGF-I receptors (72700 6 2000 sites/cell; Kd
`B, bottom panel, time course of IGF-I receptor number. IGF-I receptor number was determined as described in “Materials and Methods.” The number of IGF-I receptors significantly
`increased during maximal BON cell proliferation (49,700 versus 72,700 sites/cell at 36 and 48 h, respectively; P , 0.05). C, secretion and identification of IGFBP-2 in the conditioned
`media of BON cells by Western ligand analysis (top panel) and immunoblot analysis (bottom panel). Cells were kept up to 6 days under serum-free conditions before IGFBP expression
`was analyzed as described in “Materials and Methods” using rabbit anti-hIGFBP-2 antiserum.
`4575
`
`Roxane Labs., Inc.
`Exhibit 1007
`Page 004
`
`

`
`IGF-I IS AN AUTOCRINE REGULATOR OF CHROMOGRANIN A
`
`Fig. 2. BON cells store and secrete IGF-I. Left and middle panels, serum-starved BON cells were fixed, and IGF-I immunoreactivity was further analyzed by immunocytochemistry
`using a specific anti-IGF-I antibody (1). No specific signal was obtained when cells were stained using only the secondary antibody (–). Right panel, the IGF-I concentration in the
`supernatants of serum-starved BON cells was determined over several days after plating as indicated using a specific ELISA.
`
`connective tissue, suggesting an autocrine and/or paracrine mecha-
`nism of action.
`IGF-I Activates PI3-kinase, p70s6k, and ERKs in BON Cells.
`The IGF-I receptors in human BON carcinoid cell lines were func-
`tionally active: incubation of BON cells with IGF-I led to a marked
`increase in tyrosine phosphorylation of the IGF-I receptor and of two
`major IGF-I receptor substrates, IRS-1 and IRS-2 (data not shown).
`IGF-I markedly stimulated the activation of PI3-kinase, which could
`be prevented by treatment of cells with the selective PI3-kinase
`inhibitor LY 294002 (Fig. 3A, left panel). Similar results were ob-
`tained with a different PI3-kinase inhibitor, wortmannin (data not
`shown). The effect of LY 294002 on PI3-kinase activation was
`selective, as treatment of cells with the MEK-1 inhibitor PD 098059
`did not block IGF-I-induced PI3-kinase activation (data not shown).
`PI3-kinase is composed of two subunits, a Mr 110,000 catalytic
`subunit and a Mr 85,000 regulatory subunit, which possesses two
`Src-homology 2 domains. Both of these Src-homology 2 domains
`specifically associate with tyrosine-phosphorylated IRS-1 and IRS-2
`(19). To examine which IRS would mediate PI3-kinase activation in
`response to IGF-I, we determined the IRS-1- and IRS-2-associated
`PI3-kinase activity in BON cells stimulated with IGF-I. Cells were
`treated with 100 ng/ml IGF-I for 10 min and lysed. The lysates were
`immunoprecipitated with specific anti-IRS-1 or anti-IRS-2 antibodies
`and further analyzed by PI3-kinase assays. A 2- and 1.5-fold increase
`in PI3-kinase activity could be detected in IRS-1 and IRS-2 immu-
`noprecipitates, respectively (Fig. 3A, middle and right panels). Thus,
`
`lower affinity (IC50, 716 3 nM). Scatchard analysis (Fig. 1B, bottom
`panel) revealed a single class of IGF-I receptors (72700 6 2000
`5 0.77 6 0.1). As shown in Fig. 1B, bottom panel, the
`sites/cell, Kd
`number of IGF-I receptors markedly increased during BON cell
`proliferation, reaching a maximum after 48 h.
`Interestingly, r3-IGF-I, which does not cross-bind with IGFBPs,
`showed a slightly higher affinity (IC50, 0.4 6 0.1 nM) than IGF-I (Fig.
`1B, top panel), suggesting the existence of IGFBPs in the supernatant
`of BON cells. As demonstrated by ligand blotting, only a single class
`of IGFBPs could be detected in BON cell supernatants (Fig. 1C, top
`panel). This band was identified as IGFBP-2 by Western blotting (Fig.
`1C, bottom panel). The amount of IGFBP-2 increased in the super-
`natant of serum-starved cells with time, reaching a maximum 5 days
`after plating of the cells.
`To examine whether the increase in 125I-IGF-I binding with time
`could be attributable to cross-binding of 125I-IGF-I to IGFBPs, we
`performed binding assays using 125I-r3-IGF-I, which acts as a ligand
`for the IGF-I receptor but not for IGFBPs. These experiments revealed
`binding data very similar to those obtained with 125IGF-I: using
`125I-r3-IGF-I, a 1.7-fold increase in IGF-I receptor number/cell was
`detected after 48 h of incubation (data not shown). Thus, the increase
`in 125I-IGF-I binding was not caused by increased binding to IGFBPs
`but was indeed the result of an increase in receptor numbers.
`BON Cells Secrete IGF-I. It has been speculated that IGF-I could
`act on neuroendocrine tumors by an autocrine mechanism (6). Indeed,
`IGF-I could be demonstrated in the cytoplasm of human BON cells by
`immunocytochemistry using a specific anti-IGF-I antibody (Fig. 2,
`left and middle panels). To examine whether IGF-I was also secreted
`by BON cells, we determined the amount of IGF-I in the supernatants
`of serum-starved BON cells at various times. As shown in Fig. 2, right
`panel, IGF-I concentrations in the supernatant of serum-starved BON
`Tumor type (total n 5 11)
`cells increased in a time-dependent manner: a maximum concentra-
`Carcinoid (n 5 6)
`5/6
`5/6
`5/6
`3/6
`tion of 15 ng IGF-I/ml of conditioned medium was obtained 7 days
`A
`A
`Pheochromocytoma (n 5 2)
`2/2
`2/2
`after plating of the cells. These data suggest the existence of an
`Gastrinoma (n 5 2)
`2/2
`2/2
`1/2
`1/2
`Insulinoma (n 5 1)
`autocrine loop involving IGF-I in human BON cells. To substantiate
`0/1
`1/1
`0/1
`1/1
`these findings, we examined a panel of various human neuroendocrine
`Cryostat sections of six carcinoids tumors, two pheochromocytomas, two gastrinomas,
`and one insulinoma were stained as described in “Materials and Methods” using antibod-
`tumors by immunohistochemistry using specific antibodies against
`ies directed against chromogranin A, IGF-I, and the IGF-I receptor. All tumors exhibited
`IGF-I and the IGF-I receptor. As shown in Table 1, all but one
`positive chromogranin A staining. About 60% of the tumors (7 of 11) stained positive for
`neuroendocrine tumor exhibited positive staining for the IGF-I recep-
`IGF-I. Interestingly, 10 of 11 tumors exhibited IGF-I immunoreactivity within the tumor
`and/or in the surrounding connective tissue (ct). Ninety percent (10 of 11) of the tumors
`tor. IGF-I immunoreactivity was also detected in a large proportion of
`examined stained positive for the IGF-I receptor. The tissue sections of the pheochromo-
`tumor specimens either in the tumor itself and/or in the surrounding
`cytomas did not contain any connective tissue (indicated by A).
`4576
`
`Table 1 IGF-I and the IGF-I receptor are expressed in a variety of
`neuroendocrine tumors
`
`IGF-I
`
`IGF-IR
`
`Tumor
`
`ct
`
`Tumor
`
`ct
`
`Roxane Labs., Inc.
`Exhibit 1007
`Page 005
`
`

`
`IGF-I IS AN AUTOCRINE REGULATOR OF CHROMOGRANIN A
`
`both adapter proteins can directly mediate IGF-I-stimulated PI3-
`kinase activation in human BON cells.
`PI3-kinase regulates several downstream targets, one of which is
`p70s6k (20). To determine whether IGF-I-induced PI3-kinase activa-
`tion was sufficient to trigger the activation of downstream targets, we
`examined p70s6k activation in response to IGF-I. As shown in Fig. 3B,
`stimulation of cells with IGF-I induced a moderate but significant
`increase in p70s6k activity in immune complex kinase assays
`(P , 0.02). IGF-I-stimulated p70s6k activity could be prevented by
`treatment of cells with the PI3-kinase inhibitor LY 294002, as well as
`the specific inhibitor of p70s6k activation, rapamycin (21). Rapamycin
`and LY 294002 also strikingly inhibited basal p70s6k activity in BON
`cells, suggesting that this kinase was constitutively active in BON
`cells. In contrast, the MEK-1 inhibitor PD 098059 had no effect on
`basal and IGF-I-induced p70s6k activity (Fig. 3B, inset).
`IGF-I also stimulated activation of ERK2 in human BON cells: as
`shown in Fig. 3C, 100 ng/ml IGF-I induced a moderate but significant
`increase in ERK2 activity (P , 0.01). Similar results were obtained in
`ERK1 immune complex kinase assays (data not shown). IGF-I-induced
`ERK2 activation could be prevented by treatment of cells with the
`selective MEK-1 inhibitor PD 098059 (Ref. 22; Fig. 3C). PD 098059 also
`reduced basal ERK2 activation by 40%, suggesting that ERK2 is also
`constitutively active in these cells. Because IGF-I could activate PI3-
`kinase in BON cells, we examined whether this kinase could be an
`upstream regulator of ERK activation in BON cells. Treatment of cells
`with the selective PI3-kinase inhibitor LY 294002 reduced basal and
`IGF-I-stimulated ERK2 activation. Thus, IGF-I-stimulated ERK activa-
`tion requires the activation of MEK-1 and PI3-kinase (Fig. 3C).
`Constitutive Activation of ERK2 and p70s6k in Human BON
`Cells Is Caused by Autocrine IGF-I Secretion. As shown in Fig. 3,
`A and B, both p70s6k and ERK2 can be further activated by exogenously
`added IGF-I but exhibit a high basal activity in serum-starved BON cells
`(Fig. 3, B and C). Therefore, we examined whether basal activation of
`these kinases in BON cells could be attributable to stimulation by the
`endogenously released IGF-I. Serum-starved BON cells were incubated
`with the immunoneutralizing antibody directed against IGF-I, and sub-
`sequently, p70s6k and ERK2 activities were determined. Activation of
`p70s6k by mitogens can be determined by the appearance of slower
`migrating forms in SDS-PAGE attributable to phosphorylation of p70s6k
`on Thr229, Thr389, and Ser404, which are not basally phosphorylated in
`quiescent cells (23). As shown in Fig. 3D, left panel, incubation of cells
`with the IGF-I-blocking antibody reduced basal phosphorylation of
`p70s6k in BON cells as demonstrated by the increased electrophoretic
`mobility of the protein in SDS-PAGE. Incubation with the IGF-I-block-
`ing antibody was almost as efficient in inducing dephosphorylation of
`p70s6k as treatment with the selectiv