Abstract
Overexpression of the epidermal growth factor receptors (EGFRs) and human epidermal growth factor receptor 2 occurs frequently in human cancers and is associated with aggressive tumor behavior and poor patient prognosis. We have investigated the effects in vitro and in vivo of a new class of inhibitor molecules on the growth of several human cancer cell lines. BIBX1382 [N8-(3-chloro-4-fluoro-phenyl)-N2-(1-methyl-piperidin-4-yl)-pyrimido[5,4-d]pyrimidine-2,8-diamine] and BIBU1361 [(3-chloro-4-fluoro-phenyl)-[6-(4-diethylaminomethyl-piperidin-1yl)-pyrimido[5,4-d]pyrimidin-4-yl]-amine] are two new selective EGFR kinase inhibitors that do not block the activity of other tyrosine kinases. BIBU1361 blocked epidermal growth factor-induced phosphorylation of EGFR and also prevented downstream responses such as mitogen-activated protein kinase kinase (MAPK/extracellular signal-regulated kinase kinase) and MAPK activation in cells. In accordance with these observations thymidine incorporation into EGFR-expressing KB cells was selectively and potently inhibited by BIBX1382 and BIBU1361 with half-maximally effective doses in the nanomolar range. Oral administration of these compounds inhibited the growth of established human xenografts in athymic mice, including vulval and head and neck squamous cell carcinomas. Tumor growth inhibition by BIBX1382 coincided with reduced pEGFR and Ki-67 levels in vivo, which is in accordance with the expected effect of EGFR inhibitors. Collectively, these results show that the structural class of pyrimidopyrimidines, exemplified here by BIBX1382 and BIBU1361, represents an interesting scaffold for the design of EGFR inhibitors.
The control of proliferation, differentiation, migration, and apoptosis is essential for the development and homeostasis of multicellular organisms. Receptor tyrosine kinases, located in the plasma membrane, in concert with their cognate ligands control many of these processes. These receptors share basic structural elements and exist in distinct forms and have been classified in 18 different families (Ullrich and Schlessinger, 1990; Shawver, 1999). The human EGF receptor (HER1, c-erb-B1) is a member of the class I trans-membrane growth factor receptor family comprising HER2, for which no ligand has been described, HER3 and HER4. It is noteworthy that a large group of ligands, comprising EGF-related peptides and heregulins, differentially binds to these receptors. The ligands are encoded by more than 10 different genes, and alternative splicing as well as post-translational processing adds to the complexity of this signaling system. Upon binding to the extracellular domain, the ligands induce either homo- or heterodimerization (Carraway and Cantley, 1994). The preferred dimerization partner is the HER2 molecule, which is thought to act as a common subunit (Karunagaran et al., 1996). The close vicinity of the intracellular domains results in the activation of the kinases through transphosphorylation of the receptor subunits. The phosphorylated tyrosine residues in the cytoplasmic tail of the respective receptor will act as competent docking sites for an array of downstream signaling molecules. The signal transduction pathways include the mitogenic Ras/Raf/MAPK cascade, the phosphoinositide 3-kinase/Akt, the phospholipase C, and the Janus tyrosine kinase/signal transducer and activator of transcription pathways (van der Geer et al., 1994; Fernandes et al., 1999; Carpenter, 2000; Arteaga, 2001). Ultimately, the nature of the cellular response to class I receptor activation will depend on the type of signaling cascades used, the amplitude of the downstream signals, and the cellular signaling environment.
The observation that the class I receptor pathway is often implicated in the development and progression of many common human epithelial cancers makes the blockade of this growth pathway an attractive target for anticancer therapy (Sainsbury et al., 1987; Gullick, 1991; Helmout and Dean, 1994; Klapper et al., 2000; Grunwald and Hidalgo, 2003; Mendelsohn and Baselga, 2003; Ross et al., 2003). Several approaches have been used to block class I receptor function in cancer cells. Among these, monoclonal antibodies directed to the EGFR (e.g., IMC-C225) or HER2 receptor (Herceptin), as well as small molecules inhibiting the tyrosine kinase activity show promising activity as anticancer agents in preclinical models and clinical trials (Sridhar et al., 2003). In recent years, several classes of small molecules, inhibiting the function of EGFR or HER2 molecules, have been identified (Traxler, 2003). All these small molecules bind competitively to the ATP binding pocket of the receptor kinase moiety and thus inhibit class I receptor activation. Although similar in their inhibition mode, these compounds differ with regard to potency and selectivity. The quinazolines Iressa and erlotinib, which are the most advanced compounds, are selective and reversible EGFR inhibitors. Irreversible inhibitors such the quinazoline canertinib (Smaill et al., 2000; Slichenmyer et al., 2001) and the cyanoquinoline EKB-569 (Wissner et al., 2003) covalently bind into the catalytic cleft of the EGFR molecule and display long-lasting inhibition that is only overcome by new receptor synthesis (Fry et al., 1998). Some inhibitors such as lapatinib and canertinib are dual kinase inhibitors that simultaneously inhibit the EGFR and HER2 kinase activities. Here, we report about two new EGFR inhibitors belonging to the structural class of pyrimido [5,4-d]-pyrimidines comprising a number of molecules displaying strong inhibitory activity against EGFR-dependent processes in vitro and showing good antitumor efficacy in vivo. Whereas BIBX1382 failed to achieve plasma exposure in humans due to a first path metabolism (Dittrich et al., 2002) BIBU1361 is expected to be unaffected by this metabolism.
Materials and Methods
Materials
BIBX1382, BIBU1361, and Iressa. BIBX1382 and BIBU1361 were synthesized at Boehringer Ingelheim GmbH (Ingelheim, Germany) (Himmelsbach et al., 1997), and Iressa [4-(3-chloro-4-fluoro-anilino-7-methoxy-6-(3-morpholinopropoxy) quinazoline] was synthesized as described previously (Gibson, 1996).
Cell Lines. The human cancer cell lines KB (oral epidermoid carcinoma), A431 (vulval squamous cell carcinoma), and FaDu (hypopharyngeal squamous cell carcinoma) were obtained from the American Type Tissue Culture Collection (Manassas, VA). The HN5 cell line (head and neck squamous cell carcinoma) was kindly provided by Michael J. O'Hare (Ludwig Institute for Cancer Research, London, UK). All cell lines were maintained at 37°C in a 5% CO2 humidified atmosphere in Dulbecco's modified Eagle's medium (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum (Sigma Chemie, Deisenhofen, Germany), 1 mM Na-pyruvate, 1% nonessential amino acids, 2 mM l-glutamine, 100 units/ml penicillin, and 100 μg/ml streptomycin.
Methods
Kinase Assays. The cytoplasmic tyrosine kinase domains of EGFR, HER2, VEGFR2, HGFR, IGF1R, β-InsRK, and c-src were cloned as such or as fusion proteins into pFastBac (Invitrogen, Renfrew, Renfrewshire, UK). The protein was expressed in Sf9 insect cells using the baculovirus expression system according to the manufacturer's protocol. Sf9 cells were harvested 72 h postinfection, washed twice with PBS, and extracted with HEPEX (20 mM HEPES, pH 7.4, 100 mM NaCl, 10 mM β-glycerophosphate, 10 mM para-nitro-phenylphosphate, 30 mM NaF, 5 mM EDTA, 5% glycerol, 1% Triton X-100, 1 mM Na3VO4, 0.1% SDS, 0.5 μg/ml pepstatin A, 2.5 μg/ml 3,4-dichloroisocoumarin, 2.5 μg/ml trans-epoxysuccinyl-l-leucyl-l-amido butane, 20 KIU/ml aprotinin, 2 μg/ml leupeptin, 1 mM benzamidine, and 0.002% phenylmethylsulfonyl fluoride). The extracts were used for the determination of IC50 values. The dilution of enzymes was set so that incorporation of phosphate was linear with respect to time and amount of enzyme. Enzyme activities were assayed in the presence or absence of serial dilutions of the inhibitor performed in 50% dimethyl sulfoxide. Each microtiter plate contained internal controls such as blank, maximum reaction (no inhibitor added), and historical reference compound (substance with known IC50 value).
EGFR kinase activity was assessed in a 50-μl reaction containing 5% dimethyl sulfoxide, 40 mM HEPES, pH 7.4, 10 mM Mg-acetate, 0.5 mg/ml poly (Glu-Tyr), 0.05% Triton X-100, 50 μM ATP, 1 μCi of [γ-33P]ATP, and 10 μl of enzyme preparation. Assays were carried out at room temperature for 30 min and terminated by the addition of 10 μl of 5% H3PO4. The entire reaction was trapped onto GF/B filters, and the incorporated radioactivity was determined by scintillation counting using a Microbeta counter.
The reaction conditions for the radioactive HER2 kinase assay have been described previously (Stratowa et al., 1999). The radioactive assays for VEGFR2 and HGFR were performed as described above with the following modifications: VEGFR2 assays contained 5% Me2SO, 40 mM HEPES, pH 7.4, 5 mM MgCl2, 5 mM MnCl2, 0.5 mg/ml poly(EY), 0.05% Triton X-100, 100 μM ATP, and 1 μCi of [γ-33P]ATP, and the reaction was carried out at room temperature for 20 min. The HGFR assays were performed like the VEGFR2 assays but in the absence of MnCl2.
The kinases assays for c-src, β-InsRK, and IGF1R were adapted to a nonradioactive setup. For the c-src kinase assays, a 100-μl reaction contained 10 μl of inhibitor in 50% Me2SO, 20 μl of substrate solution [200 mM HEPES, pH 7.4, 50 mM Mg-acetate, 2.5 mg/ml poly(EY), 5 μg/ml bio-poly(EY), and 1000 μM Na3VO4] and 20 μl of enzyme preparation. The enzymatic reaction was started by addition of 50 μl of a 1 mM ATP solution made in 10 mM Mg-acetate. For the β-InsR kinase assays, the 20-μl substrate solution contained 250 mM Tris, pH 7.4, 10 mM dithiothreitol, 2.5 mg/ml poly(EY), and 5 μg/ml bio-poly(EY), and the 2 mM ATP solution was made in 8 mM MnCl2 and 20 mM Mg-acetate. For the IGF1R kinase assays, the 20-μl substrate solution contained 200 mM HEPES, pH 7.4, 2.5 mg/ml poly(EY), and 5 μg/ml bio-poly(EY), and the 2 mM ATP solution was made in 10 mM MgCl2 and 20 mM MnCl2. The nonradioactive kinase reactions were carried out at room temperature for 30 min (20 min for IGF1R) and terminated by the addition of 50 μl of stop solution (250 mM EDTA in 20 mM HEPES, pH 7.4). Then, 100 μl was transferred to a streptavidin-coated microtiter plate, after an incubation time of 60 min at room temperature the plate was washed with 200 μl of wash solution (50 mM Tris and 0.05% Tween 20). Then, 100 μl of a horseradish peroxidase-labeled anti-PY antibody (PY20H anti-PY:horseradish peroxidase supplied by BD Transduction Laboratories, Lexington, KY) (250 ng/ml) was added. After a 60-min incubation the plate was washed three times with a 200-μl wash solution. The samples were then developed with a 100-μl substrate solution [1:1 (v/v), supplied by Bender MedSystems, Vienna, Austria; catalog BMS402 and BMS403]. The reaction was stopped after 10 min with 100 μl of 1 M phosphoric acid. The plate was transferred to an ELISA reader, and extinction was measured at OD450 nm.
The data generated with these assays were analyzed with the program Prism version 3.0 (GraphPad Software Inc., San Diego, CA). The inhibitor concentrations were transformed to logarithmic values, and the raw data were normalized. These normalized values were used to calculate the IC50 by a nonlinear regression curve fit (sigmoidal dose response, variable slope). All iteration data had a correlation coefficient above 0.9, and top and bottom values of the curves displayed a window of at least 5.
EGFR and pEGFR ELISA Assays. EGF receptor phosphorylation was assessed in A431 cells using a commercially available ELISA kit (QIA95; Oncogene Research Products, San Diego, CA) according to the manufacturer's protocol with minor modifications. In brief, 1 × 104 A431 cells were transferred into each well of a 96-well microtiter plate in 90 μl of serum-free culture medium. Cells were incubated at 37°C and 5% CO2 in a humidified atmosphere overnight. The next day, 10 μl of each test compound concentration was added to duplicate wells. Each microtiter plate contained internal controls, such as blank, maximum reaction, and historical reference compound, for the qualification of the results. The microtiter plates were then incubated at 37°C and 5% CO2 in a humidified atmosphere for 1 h. EGF-stimulation was done at a final concentration of 100 ng/ml for 10 min at room temperature. Cells were washed two times with 200 μl/well ice-cold PBS before addition of 120 μl/well receptor extraction buffer provided in the ELISA kit and shaking for 1 h at room temperature. After extraction, 100 μl of cell extract was transferred to the assay plate provided in the kit. After 1-h incubation at room temperature, the plate was washed three times with water before addition of the horseradish peroxidase-conjugate. The bound antibody was then detected with TMB substrate solution. Coloring was stopped with 100 μl of stop solution. The extinction was measured at 450 nm. All incubation steps were performed on a microtiter plate shaker.
Expression of the EGF receptor was determined with a commercially available assay (QIA35; Oncogene Research Products), and the A431 extracts were diluted with the sample diluent included in the kit and processed as described in the manufacturer's protocol. Each dilution was measured in duplicate.
The data generated with both assays were analyzed with Prism as described above.
Thymidine Incorporation Assays. Subconfluent KB cultures were used for the assays. Cells were trypsinized, washed twice in serum-free culture medium, and seeded in 200 μl of medium at a density of 1500 cells/well in a flat bottom 96-well culture dish. A 10-μl aliquot of either 200 or 60 μM compound dilution made in PBS was added to the respective first well and serially transferred between the wells. All samples were prepared in triplicate. Compounds were tested between 3 and 10,000 nM. EGF-stimulated cells received 30 ng/ml EGF, which was determined to be the optimal dose in these assays; serum-stimulated cells received 0.5% FCS; unstimulated controls received neither EGF nor serum. After addition of the inhibitors and growth factors, the cells were incubated for 76 h, and then [3H]thymidine (0.1 μCi/well) was added. After an additional 16 h, the cells were harvested on glass fiber filters, and the incorporated radioactivity was determined by scintillation counting using a TopCount (PerkinElmer Life and Analytical Sciences, Boston, MA). To determine inhibition of thymidine incorporation by EGFR inhibitors, the difference between the mean counts of stimulated and unstimulated control wells was set to a 100%. Counts incorporated by unstimulated samples were set to 0%. The stimulation of compound-treated samples was expressed in percentage of the full stimulation. The inhibitor concentrations were transformed to logarithmic values. The percentage of stimulation values were used to calculate the IC50 by iteration using a nonlinear regression curve fit program (Prism; sigmoidal dose response with variable slope).
In Vivo Xenograft Experiments. Five- to six-week-old athymic NMRI-nu/nu female mice (21–31 g) were purchased from Harlan Winkelmann (Borchen, Germany) and maintained under specific pathogen-free conditions. All experiments complied with the Declaration of Helsinki and European Policy Legislations (FELASA and GV-SOLAS) on the Care and Use of Laboratory Animals. After acclimatization mice were inoculated s.c. with 1 × 106 (in 100 μl) A431, FaDu, or HN5 cells into the right flank of the animal. After 7 to 11 days, tumors reached a average volume of 40 to 130 mm3. The mice were randomized and treated daily p.o. with BIBX1382, BIBU1361, or vehicle control on the basis of individual weights. Tumors were measured three times a week with calipers, and tumor volumes were calculated by the formula π/6 × length × (width)2. Experimental compounds were prepared in 25% aqueous hydroxypropyl-β-cyclodextrin (33259-3; Sigma-Aldrich, St. Louis, MO) and administered by intragastral gavage. The administration volume was 10 ml/kg b.wt.
Determination of pEGFR Level in Tumor Samples. Excised A431 tumors were snap-frozen in liquid nitrogen immediately after dissection. The tumors were lyophilized over night and extracted with HEPEX (20 mM HEPES, pH 7.4, 100 mM NaCl, 10 mM β-glycerophosphate, 10 mM para-nitro-phenylphosphate, 30 mM NaF, 5 mM EDTA, 5% glycerol, 1% Triton X-100, 1 mM Na3VO4, 0.1% SDS, 0.5 μg/ml pepstatin A, 2.5 μg/ml 3,4-dichloroisocoumarin, 2.5 μg/ml trans-epoxysuccinyl-l-leucyl-l-amido butane, 20 KIU/ml aprotinin, 2 μg/ml leupeptin, 1 mM benzamidine, and 0.002% phenylmethylsulfonyl fluoride). The EGFR and pEGFR levels were determined using the ELISA procedures described above [QIA95 (pEGFR) and QIA35 (EGFR); Oncogene Research Products]. To asses the PY-EGFR ratio (PY-EGFR units/EGFR molecule), an in-house standard of EGF-stimulated HN5 cells was used. The EGFR concentration of this standard was 8560 fmol/ml as determined by QIA35. The PY-EGFR concentration was arbitrarily assigned a concentration of 2.8 × 1015 PY units/ml. The PY-EGFR /EFGR ratio was calculated from the results of the PY-EGFR QIA95 ELISA (expressed in units per milliliter), and the EGFR QIA35 ELISA (expressed in femtomoles per milliliter) and expressed as arbitrary PY-units/EGFR molecule.
Histology and Immunohistochemistry. The proliferation of tumor cells in samples was assessed by Ki-67 immunostaining. Frozen tumor tissues were paraffin embedded, sectioned to 3 to 5 μmin thickness, deparaffinized, and subjected to immunohistochemistry protocol as described previously (Heider et al., 1993a,b). For Ki-67 detection an affinity-purified rabbit anti-human Ki-67 serum was applied (DakoCytomation Denmark A/S, Glostrup, Denmark). As secondary antibody either a biotinylated swine anti-rabbit or rabbit anti-mouse antibody was used, followed by incubation with the avidin-biotin-peroxidase complex (DakoCytomation Denmark A/S). After color development in aminoethylcarbazole, the slides were counterstained with hematoxylin and mounted with aqueous mounting medium. For evaluation of histology, H&E-stained serial sections were used.
Results
Potency and Selectivity of BIBX1382 and BIBU1361 in Molecular Assays. The data presented in Table 1 demonstrate that BIBX1382 and BIBU1361 are both potent and selective submicromolar inhibitors of the EGFR kinase activity. An IC50 value of 3 nM was determined for both compounds. The potency of these two compounds compares with the one obtained with Iressa, which is a leading EGFR inhibitor in the field. Inhibition of the closest family member, HER2, was 100- to 1000-fold less potent. Furthermore, BIBX1382 and BIBU1361 did not inhibit a number of other related tyrosine kinases.
Inhibition of EGFR Autophosphorylation in Cells. The inhibitory activity of these two compounds was confirmed at the cellular level. A431 cells, which highly express the EGF receptor, were used to examine the effect of BIBU1361 and BIBX1382 on EGF-induced EGFR phosphorylation. A 1-h pretreatment of the cells with either compound inhibited EGF-induced phosphorylation of EGFR in a dose-dependent manner. For BIBU1361, the EC50 was 122 nM. As a control, the EGF receptor expression was monitored and showed no variation in this experimental setup (Fig. 1). Under the same conditions, the drug concentrations required for half-maximal inhibition by BIBX1382 and Iressa were 141 and 86 nM, respectively (data not shown). Thus, these data demonstrate that both pyrimidopyrimidines reach their target in the cell and that the potency compares to that of Iressa.
Effects of BIBU1361 on Downstream Signaling. After showing that BIBU1361 is capable of inhibiting the activation of the EGF receptor, we further evaluated whether this compound would also abrogate downstream signaling events. Several signaling pathways that control the G1/S transition are activated in response to EGF in A431 cells. Therefore, the effect of BIBU1361 on one such pathway (activation of the MAPK signaling) was analyzed in more details by means of electrophoretic analyses. An aliquot of the extracts used for the autophosphorylation assays was loaded on SDS-polyacrylamide gel electrophoresis and blotted onto polyvinylidene difluoride membranes for Western blot analyses. In line with its ability to prevent EGFR activation, BIBU1361 also suppressed the activation of MEK, a central molecule in the MAPK cascade (Fig. 2). Pretreatment of A431 cells with 313 nM resulted in nearly complete inhibition of phospho-MEK signal. Although MAPK is still activated in the starved A431 cells, pretreatment for 1 h with 1250 nM BIBU1361 induced complete dephosphorylation of this molecule.
In line with the previous results, treatment of cells with either BIBX1382 or BIBU1361 suppresses EGF-induced thymidine incorporation in KB cells in a dose-dependent manner. For BIBX1382, the EC50 value is 150 nM. Pretreatment of KB cells with 400 nM resulted in almost complete inhibition of new DNA synthesis. Similarly, the EC50 value for BIBU1361 was 225 nM (data not shown). In comparison, the FCS-dependent incorporation was inhibited with an EC50 value of 3195 nM (Fig. 3). Thymidine incorporation can also be stimulated by hepatocyte growth factor in this cell line. The EC50 value was 1815 nM (data not shown). Together, the cellular data show that BIBX1382 and BIBU1361 are potent and selective inhibitors of EGF-induced DNA synthesis.
Antitumor Activity of BIBX1382 and BIBU1361. The antitumor activity of BIBX1382 and BIBU1361 was evaluated in vivo in various xenograft models, including A431, HN5, and FaDu. In nude mice, oral once daily dosing at 10 mg/kg with either BIBX1382 or BIBU1361 completely suppressed tumor growth of human A431 xenografts with respective T/C values of 15 and 6% after 2 weeks of treatment (Fig. 4). The average tumor volume of the control group had reached at least 1000 mm3. On the last treatment day, plasma samples were obtained 4, 8, and 24 h after application. Both compounds reached their maximum 4 h after administration. The C4h values were 2222 (BIBX1382) and 755 nM for BIBU1361. The C24h with 244 and 356 nM, respectively, were in the range of the EC50 values in the different cellular assays. No weight loss was observed in treated animals over the course of the 2-week treatment period. Hence, these data demonstrate that both compounds at a daily dose of 10 mg/kg are well tolerated in mice and display antitumor activity as single agents.
BIBX1382 and BIBU1361 were concurrently evaluated in additional tumor models. The data in Table 2 shows that both compounds also suppress tumor growth of additional xenograft models thus confirming their antitumor efficacy. The effect of BIBX1382 is dose-dependent in A431, FaDu, and HN5 xenografts. Dose dependence was observed in the A431 model for BIBU1361.
The antineoplastic potential of BIBX1382 was further explored in a long-term oral administration schedule. A431-tumor bearing mice were treated once daily for 170 days at a dose of 50 mg/kg. The body weights were stable, and no adverse clinical signs could be observed in the treated animals for the whole treatment period. Tumor regressions were obtained in all treated individuals (n = 7) after 90 days. Although two of the animals were tumor free at the end of the experiment, and this even after the 100-day recovery period, two of the tumors started to regrow after cessation of treatment and three tumors did not significantly progress in size during the experiment (Fig. 5). Thus, these data demonstrate the feasibility of long-term treatment schedules with BIBX1382.
Effects of BIBX1382 on Tumor Cells in Vivo. To correlate the antitumor activity of these compounds with biomarker modulation, we investigated in vivo, the ability of BIBX1382 to inhibit EGFR phosphorylation in tumors derived from A431 xenograft-bearing mice treated daily p.o. for 14 days, using biochemical and immunohistochemical approaches. The data in Fig. 6A show that the EGF receptor is constitutively phosphorylated in control A431 tumors. Treatment of mice, carrying established A431 tumors, at a dose of 50 mg/kg/day results in dephosphorylation of the EGF receptor (Fig. 6A). Thus, the antitumor activity of BIBX1382 coincides with inhibition of the constitutive tyrosine phosphorylation of the EGF receptor molecule in A431 xenografts.
The antiproliferative effect of BIBX1382 was confirmed in vivo by assessing tumor cell proliferation as determined by Ki-67 immunostaining of tumor tissue. A431 xenografts were treated daily p.o. with vehicle control or 30 mg/kg BIBX1382 for 2 weeks. The formalin-fixed and paraffin-embedded tumor tissues were then subjected to Ki-67 immunohistochemistry (see Materials and Methods). The counts of proliferative tumor cells in treated tissues were significantly decreased compared with vehicle-treated control tumors. The microphotographs of Ki-67 immunostaining illustrating a reduction in proliferation are shown in Fig. 6, B and C. A similar reduction in Ki-67 staining was observed with BIBU1361 (data not shown). These data demonstrate that BIBX1382 and BIBU1361 inhibit tumor cell proliferation in vivo.
Discussion
Excess EGFR signaling is implicated in the development of a significant proportion of human carcinomas and glioblastomas. Therefore, inhibition of EGFR-signaling has been viewed as a promising approach for the treatment of EGFR-positive human tumors. Both, anti-EGFR antibodies and small-molecule EGFR tyrosine kinase inhibitors are being actively tested in clinical trials (Grunwald and Hidalgo, 2003).
Here, we report a new structural class of small-molecule EGFR tyrosine kinase inhibitors, namely, pyrimido-[5,4-d]-pyrimidines. BIBX1382 and BIBU1361 are two examples of this class and inhibit the EGFR tyrosine kinase activity in vitro in the low nanomolar range. This potency is comparable with other small-molecule inhibitors derived from other structural classes (Table 1; Traxler, 2003). Relative to four distinct members of the tyrosine kinase family, both compounds seem to be selective for the class I receptors. Furthermore, at a concentration of 10 μM BIBX1382 and BIBU1361 failed to inhibit the kinase activities of a number of serine/threonine kinases comprising MEK, phosphorylase kinase, protein kinase A, and stress-activated protein kinase 2α (data not shown), thus confirming the selectivity of BIBX1382 and BIBU1361 for the EGFR kinase. Because all the kinase assays were performed at saturating ATP concentrations, cellular activity for these compounds could be anticipated.
Indeed, BIBX1382 and BIBU1361 in the nanomolar range could interfere with the EGF-induced EGFR autophosphorylation in intact A431 cells (Fig. 1). In accordance with this observation, BIBU1361 also blocked downstream EGFR signaling events. Many intracellular proteins get phosphorylated in response to mitogenic signals such as EGF. Some examples are phospholipase Cγ, phosphoinositide 3-kinase, Src homology 2 domain-containing transforming protein, or MEK. In some cells, constitutive activation (phosphorylation) of these proteins has been linked to the transformed phenotype. Thus, a reduction in the phosphorylation status in one of these signal transducers induced by BIBU1361 could suggest an antiproliferative effect of this compound. Therefore, we analyzed the effect of BIBU1361, 10 min after EGF addition, on the activation of MEK and MAPK. As shown in Fig. 2, BIBU1361 at concentrations of 313 nM completely blocked activation of MEK. In mitogenic assays performed with KB cells (Fig. 3), BIBU1361 as well as BIBX1382 demonstrated antiproliferative activity. The EC50 values for the two compounds were in the nanomolar range when EGF was used as a mitogenic agent. The maintained capacity of KB cells for proliferation in response to hepatocyte growth factor or FCS in the presence of micromolar concentrations of either BIBU1361 or BIBX1382 confirms the selectivity of these compounds. Together, the cellular data show that BIBX1382 and BIBU1361 are potent and selective inhibitors of the EGFR pathway.
Both BIBX1382 and BIBU1361 in mice and rats show serum concentrations of the compounds that are sufficient to achieve therapeutic efficacy. The ability of BIBX1382 and BIBU1361 to inhibit either A431 or HN5 tumor growth is similar to that reported for other EGFR inhibitors currently in clinical trials (Fan et al., 1993; Modjtahedi et al., 1998). Treatment for 14 days, at doses ranging between 10 and 30 mg/kg/day, resulted in complete tumor growth suppression in these models (Fig. 4; Table 2).
To further explore the potential of these compounds as anticancer agents, a long-term treatment study with BIBX1382 was initiated. Mice bearing A431 tumors were treated daily for a period of 170 days. Tumor regressions were observed in all treated individuals after 90 days (Fig. 5). Therefore, these data could suggest that sustained long-term inhibition of EGFR signaling in A431 tumors is required for a curative effect. Indeed, in 5 of 7 individuals, tumors had either disappeared or remained growth suppressed even during the 100-day treatment-free period. It has to be acknowledged that two of the tumors did regrow after termination of treatment. Interestingly, one of these tumors when passaged into mice responded again to BIBX1382 treatment (data not shown). These data also clearly demonstrate that, in mice, BIBX1382 can be given for an extended period of time with good tolerability.
In the 14-day A431 xenograft experiments, determination of the plasma concentrations of BIBX1382 and BIBU1361 on the last day of treatment, 24 h postapplication, revealed that the minimum plasma levels were 244 and 356 nM, respectively. It is noteworthy that in the cellular assays described herein, such concentrations resulted in nearly complete inhibition of EGFR-dependent processes. Therefore, this observation suggests that in vivo the EGF-dependent processes are completely and permanently inhibited by BIBX1382 and BIBU1361 over a 24-h period. The biochemical data shown in Fig. 6A is consistent with this observation. Indeed, a 2-week treatment of mice carrying established A431 tumors with BIBX1382-induced dephosphorylation of constitutively activated EGFR. Because of its ability to block early EGFR signaling in vivo, it was anticipated that BIBX1382 would also inhibit tumor cell proliferation in the tumor. Ki-67 immunostaining was used to visualize proliferating cells in vivo. In the treated samples (Fig. 6C), a reduction of stained cells was evident.
In conclusion, we have shown that pyrimidopyrimidines, exemplified here by BIBX1382 and BIBU1361, are potent inhibitors of EGFR-mediated signaling in vitro. In vivo, both compounds demonstrate antitumor activity in xenograft models such as A431, FaDu, and HN5. All these models have proven to be reliable predictors for class I receptor-targeted therapeutics such as erlotinib (Pollack et al., 1999), Iressa (Sirotnak et al., 2000), canertinib (Nelson and Fry, 2001), or lapatinib (Rusnak et al., 2001), which all have advanced to clinical trials. The results described above also show that BIBX1382 is well tolerated in long-term treatment regimens in animal models with sufficient exposure. Unfortunately, BIBX1382, which was tested in clinical phase I studies, did not show acceptable plasma exposure in humans. The reason for this behavior has been attributed to a first path metabolism (Dittrich et al., 2002), and the enzymes involved have been resolved and are the subject of another publication. Unlike BIBX1382, BIBU1361 is not a substrate for this enzyme and therefore plasma exposure in humans is expected. Thus, pyrimidopyrimidines, in principle, can be viewed as an attractive structural class for the development anticancer agents for patients whose cancers display aberrant EGFR signaling. Attention should be paid to first path metabolism effects as demonstrated for BIBX1382.
Acknowledgments
We thank L. Süly, T. Troels, U. Strobl, M. Leber, R. Meyer, and I. Kothbauer for excellent technical assistance. We also thank A. Greischel and S. Blech for providing preclinical pharmacokinetic data. The unequivocal support of T. Metz (Oncotest, Freiburg, Germany) for this program is gratefully acknowledged.
Footnotes
-
doi:10.1124/jpet.104.069138.
-
ABBREVIATIONS: EGF, epidermal growth factor; HER, human epidermal growth factor receptor; MAPK, mitogen-activated protein kinase; EGFR, epidermal growth factor receptor; BIBX1382, N8-(3-chloro-4-fluoro-phenyl)-N2-(1-methyl-piperidin-4-yl)-pyrimido[5,4-d]pyrimidine-2,8-diamine; BIBU1361, (3-chloro-4-fluoro-phenyl)-[6-(4-diethylaminomethyl-piperidin-1-yl)-pyrimido[5,4-d]pyrimidin-4-yl]-amine; VEGFR2, vascular endothelial growth factor receptor 2; HGFR, hepatocyte growth factor receptor; IGF1R, insulin-like growth factor 1 receptor; InsRK, insulin receptor kinase; PBS, phosphate-buffered saline; FCS, fetal calf serum; pEGFR, phosphorylated epidermal growth factor receptor; ELISA, enzyme-linked immunosorbent assay; MEK, mitogen-activated protein kinase kinase.
- Received March 30, 2004.
- Accepted June 15, 2004.
- The American Society for Pharmacology and Experimental Therapeutics