Epidermal Growth Factor Receptor Signal Transduction Pathway with Potent inVivo Antitumor Activity

Deregulated signal transduction via the epidermal growth factor (EGF) receptor family of tyrosine protein kinase growth factor receptors is associated with proliferative diseases such as cancer and psoriasis. In an attempt to selectively block signal transduction from the EGF receptor, we have synthesized a new class of dianilino-phthalimide tyrosine protein kinase inhibitors with selectivity for the EGF receptor tyrosine protein kinase. 4,5-Dianilino-phthalimide (DAPH 1) was metabolized in vitro by mouse liver fractions and in vivo. The major metabolite has been identified as 4-(4-hydroxyanilino)-5-anilino-phthalimide. To specifically block this biotransformation (hydroxylation), we have synthesized 4,5-bis(4-fluoroanffino)phthalimide (DAPH 2), a potent and selective EGF receptor tyrosine protein kinase inhibitor. DAPH 2 inhibits the EGF receptor and protein Idnase C 12 enzymes with equal potency. In cells, DAPH 2 inhibits signal output from the EGF receptor, but not from other classes of receptor protein tyrosine kinases, such as the platelet-derived growth factor receptor, fibroblast growth factor receptor, insulin-like growth factor I receptor, and insulin receptor. Selective antitumor activity was demonstrated in vivo at well-tolerated doses in mice. This publication describes the biological profile of DAPH 2 and investigates its cellular and in vivo mechanism of action. INTRODUCTION The EGF2 receptor (1, 2) is a transmembrane glycoprotein that mediates the response of cells to the EGF family of mitogenic polypeptides which include EGF, transforming growth factor a (3), and amphiregulin (4). The EGF receptor and its Received 2/23/95; accepted 2/28/95. I To whom requests for reprints should be addressed, at Ciba-Geigy Limited, Pharmaceuticals Division, Oncology Research Department, K-12S.4.20, CH-4002 Basel, Switzerland. 2 The abbreviations used are: EGF, epidermal growth factor; DAPH, dianilino-phthalimide: DAPH 1 , 4,5-dianilino-phthalimide; DAPH 2, 4,S-bis(4-fluoroanilino)phthalimide; PDGF, platelet-derived growth factor; 50% inhibitory concentration; FGF, fibroblast growth factor; IGF-I. insulin-like growth factor I; BCS. bovine calf serum. ligands are involved in epithelial proliferation and have been strongly implicated in malignant tumor growth (5-7) and proliferative skin diseases such as psoriasis (8). Enzymatic activity of the intracellular kinase domain of the EGF receptor is essential for signal transduction via the EGF receptor (9, 10). The activated EGF receptor phosphorylates tyrosine residues in its C-terminal tail, as well as intracellular substrate proteins, many of which contain Src homology 2 domains (reviewed in Ref. 1 1). These intracellular phosphorylation events in turn mediate the intracellular signaling cascade which results in activation of nuclear events such as transcription of immediate early genes (12). The c-erbB-l gene has been found to undergo gene amplification or overexpression in a number of human tumors of epithelial and neuroepithelial origin (6). In addition, the closely related c-erbB-2 gene has been shown to be amplified or overexpressed at high frequency in human breast and ovarian carcinomas (13). The involvement of the EGF receptor tyrosine protein kinases in epithelial proliferation suggests that selective enzyme inhibitors could have therapeutic potential in the treatment of malignant and nonmalignant epithelial diseases. We have recently described a novel group of tyrosine protein kinase inhibitors with potent it, vitro and in vito activity (1, 2) from the DAPH class. DAPH I was metabolized to 4-(4-hydroxyanilino)S-anilino-phthalimide in liver extracts and in vito. In an attempt to block this metabolism, we have synthesized 4,S-bis(4-fluoroanilino)phthalimide. In the present report we describe the biological profile of DAPH 2, an analogue of DAPH 1 which differs from the parent compound in its in s’is’o potential for metabolism and in its biological profile. Furthermore, we describe mechanistic studies aimed at understanding the in vitro and in t’ivo activity of dianilino-phthalimides. MATERIALS AND METHODS 4,S-Dianilino-phthalimides were synthesized by Ciba Pharmaceuticals Division as previously described (2). For in vitro and cellular assays, a stock concentration of 10 msi DAPH 2 was prepared in Me2SO and stored at -20#{176}C.Dilutions for all assays were made up freshly prior to use. v-sis transfected BALB/c 3T3 cells (kindly provided by Dr. C. Stiles, DanaFarber Cancer Institute, Boston, MA) were maintained in DMEM, supplemented with 10% BCS (Hyclone, Logan, UT) and 0.5 mg/ml 0418 (GIBCO-BRL, Gaithersburg, MD). Culture conditions for the T24 human bladder carcinoma cells were as previously described (14). The human insulin receptor overexpressing Rat-I cells (kindly provided by Dr. D. A. McClain, on May 1, 2017. © 1995 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from 814 Inhibition of the EGF Receptor Kinase University of Alabama, Birmingham, AL; Ref. 15) were grown in DMEM supplemented with 10% BCS and 0.5 mg/ml G418. Transformed NIH 3T3 cells (NIH 3T3 L!SNc4 line) overexpressing the IGF-! receptor (kindly provided by Dr. M. Kaleko, Fred Hutchinson Cancer Research Center, Seattle, WA; Ref. 16) were cultured in DMEM supplemented with 10% BCS. FDC-P1 cells were grown in Iscove’s modified Dulbecco’s medium supplementd with 10% FCS and conditioned medium from WEHI-3B cells (17). NADP sodium salt was obtained from Boehringer Mannheim (Mannheim, Germany). DL-!socitrate and isocitrate dehydrogenase were from Sigma (St. Louis, MO). 3-Glucuronidase/arylsulfatase was from Calbiochem (La Jolla, CA). Other cell lines are as described (1, 2, 18), and reagents were analytical grade quality from either E. Merck (Darmstadt, Germany) or Fluka AG (Buchs SG, Switzerland). Lauroglycol was obtained from Gattefoss#{233} s.a. (Saint Priest, France). Kinase Assays. Purification of protein kinases and in vitro enzyme tests were performed as previously described (1, 19, 20). Inhibition of cdc2/cyclin B protein kinase was assayed as previously described (21). Cellular Tyrosine Phosphorylation. Inhibition of EGFand PDGF-stimulated total cellular tyrosine phosphorylation in A431 cells and BALB/c 3T3 cells, respectively, was measured using a microtiter ELISA assay according to a previous report (2). Western Blot Analysis. Serum-starved A431 cells or quiescent BALB/c 3T3 cells were prepared as previously described (18) and were incubated for 90 mm with the indicated concentrations of drug prior to stimulation with EGF (100 ng/ml) or PDGF BB (50 ng/ml) for 10 mm, respectively. Similarly, serum-starved Rat-i or NIH 3T3 USNc4 cells were treated with the drug and stimulated with insulin (5 i.g/ml) or IGF-I (1 p.g/ml) for 10 mm, respectively. Serum-starved SK-BR-3 human breast tumor cells were treated for 10 min with 10 ng/m! FWP 51, an anti-erbB-2 mAb which activates receptor autophosphorylation (22). Equal amounts of protein of cell lysates were analyzed by Western blotting using antiphosphotyrosine antibodies produced by hybridoma 4G10 (23). Bound antibodies were detected using the ECL Western blotting system from Amersham (Amersham, United Kingdom). Isolation of RNA and Northern Blot Analysis. c-fos induction assays were performed as previously described (1). In Vitro and in Vivo Biotransformation. Postmitochondrial liver fractions (S12) were prepared from human liver samples (liver KDL 18) provided by the University of Base! according to local regulatory guidelines. Liver pieces from male mice [Tif:MAGf (SPF)], male rats [Tif:RA!f (SPF)], a male mongrel dog, or human liver samples were suspended in 0.15 M KC1 (1:5 w/v) containing 0.1 mM phenylmethylsulfonyl fluoride, 0.02 mM butylated hydroxytoluene, and 1 mst EDTA, and homogenized in a Polytron homogenizer at 5000 rpm and subsequently with a Potter-Elvehjem homogenizer. The homogenate was centrifuged at 12,000 X g for 15 mm. The supernatant (S12) was recovered and adjusted to 3.4 mg protein/ml. Compounds (2 p.! of a stock solution in DMSO) were incubated with 178 i.l postmitochondrial liver fraction (S120 and 20 j.tl of a NADPH regenerating system containing isocitrate dehydrogenase (1 units/mi), NADP (1 mM), isocitrate (S mM), and MgCl2 (S mM) at 37#{176}C in a shaking water bath (100 rpm). Control incubations were made with heat-inactivated postmitochondria! liver fractions. Reactions were stopped and deproteinized by the addition of 400 p.1 acetonitrile followed by centrifugation for S mm at 10,000 x g and analyzed by HPLC (method 1). Pharmacokinetics. Mouse pharmacokinetic studies on DAPH 2 were performed in female BALB/c mice. Stock solutions of compound (40 mg/mI) were prepared in 100% Me2SO, containing 0.5% Tween 80 (polysorbate), and stirred at 60#{176}C until a clear solution was obtained. Stock solutions were diluted 1:20 (v/v) with sterile 0.9% NaCl and used for treatment. Two animals were used for each time point in pharmacokinetic studies. A single dose of 100 mg/kg animal weight was used. For rat pharmacokinetics, compound was administered p.o. as a 15% suspension in PEG 400 and Tween 80 (9:1 ratio) at a dose of 100 mg/kg animal weight; two rats were used per group. Analytical HPLC Method 1. A LiChrospher 60 RP select B column (Merck; S p.m. 250 X 4 mm) was used at a flow rate of 2 ml/min with a gradient of acetonitrile (25-50% in 20 mm) and sodium acetate buffer, pH 5.1 (0.13 M). The cornpounds were detected by UV absorbance at 273 nm and DAPH 1 and DAPH 2 quantified using external standards. Analytical HPLC Method 2. Analysis of deproteinized serum samples was carried out on a 125x 4.6-mm Nucleosil C18 (S .i.m) analytical column eluted with a mobile phase of 50% acetonitrile and 0.05% trifluoroacetic acid in water. The flow rate was 1 mI/mm. The compound was detected by UV absorbance at 310 nm. Concentrations were calculated from peak areas in comparison to a standard curve, which was calculated from HPLC analysis of sera spiked with known amounts of similarly processed compound. In Vivo Experiments. In vivo antiturnor activity was tested using the A431 human epithelial carcinoma, v-sis transfected BALB/c 3T3 cells, and T24 bladder carcinoma cells. A431 or T24 tumor pieces of approximately 25 rnm3 were transplanted into the left flank of female BALB/c nude mice (Bomholtgaard, Copenhagen, Denmark). Six animals were used per treated or control group. Drug was given p.o. once daily for 15 days, starting on day S or 6 after transplantation. v-sistransformed cells (1 X 10 ’ BALB/c 3T3 v-sis in 0.2 ml PBS) were injected into the left flank of each animal. Drug was given P.O. once daily for 18 days, starting 24 h after cell injection. Stock solutions were prepared as described for pharrnacokinetic experiments, diluted 1 :20 (v/v) with sterile 0.9% NaCI, and used for treatment. Some antitumor experiments were performed using compound formulated in lauroglycol, which resulted in similar antiturnor activity as compared to the Me2SO/Tween80/ NaCl formulation (data not shown). Solution and dilution were prepared daily prior to application. Tumor growth was followed by measuring perpendicular tumor diameters. Animals were sacrificed when the control tumor group had reached a volume of 1-1.5 cm3 (as required by local regulations). Tumor volumes were calculated as previously described (24) using the formula err X L x D2/6). RESULTS Metabolism of DAPH 1 and DAPH 2 in Vitro and in Vivo. The biotransformation of DAPH 1 was investigated in vitro using postmitochondrial liver fractions from mouse, rat, dog, and humans (Fig. 1). The main metabolite of DAPH 1 in on May 1, 2017. © 1995 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from