Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The ErbB family, particularly EGFR and ErbB2, are overexpressed in various tumors and have been correlated with an adverse prognosis for cancer patients (Prigent and Lemoine, 1992). One major limitation of current EGFR tyrosine kinase (TK) inhibitors is the high intracellular concentration of ATP that represents a major barrier to sustained inhibition of EGF-stimulated signal transduction in tumor cells. Where they cannot induce apoptosis, EGFR TK inhibitors are cytostatic agents that induce reversible antitumor effects. Recently, irreversible inhibitors of the EGFR family have been synthesized that showed greater potency than their reversible predecessors (Fry, 1998, 1999; Smaill et al., 2000). Unfortunately, irreversible inhibition of EGFR may not suffice to induce sustained antitumor activity if the cells possess alternative growth mechanisms, and combination with cytotoxic drugs to potentiate the action of the EGFR TK inhibition is now being considered a useful alternative (Ciardiello et al., 2000).

Our novel Combi-Targeting strategy seeks to combine the signal transduction inhibitory mechanisms of receptor TK inhibitors with the cytotoxic effects of DNA-damaging fragments into one single agent termed C-molecule. The latter is designed to 1) inhibit the receptor TK on its own and 2) be converted into another inhibitor of the same receptor TK plus a DNA-damaging species upon hydrolysis. This principle simply leads to a receptor-affinity TZ-I (Scheme 1) capable of generating a cytotoxic molecule (TZ) plus another inhibitor, I. 

View larger version (11K): [in this window] [in a new window]   Scheme 1.   A receptor-affinity TZ-I capable of generating a cytotoxic molecule (TZ) plus another inhibitor, I.

As depicted in Scheme 2, in EGFR-overexpressing cells, we predict that the extracellular TZ-I' (TZ representing the cytotoxic element and I or I', the EGFR inhibitory components) can diffuse through the membrane and in the cytosol (see TZ-I) either directly bind to the EGFR ATP-binding site to provide the TZ-I-EGFR complex (path 3) or degrade into a cytotoxic TZ molecule plus an EGFR TK inhibitor I (path 2). Although the TZ will exert cytotoxic activity by damaging DNA (see DNA-TZ), the generated inhibitor I is designed to inhibit EGFR-induced growth (path 5). The combined cytostatic and cytotoxic effects may lead to enhanced antiproliferative activity of the TZ-I in EGFR-overexpressing cells. More importantly, the TZ-I may directly alkylate the EGFR as outlined in path 4 wherein an inactivated (covalently modified) receptor (TZ-EGFR) may be formed, leaving an irreversibly inhibited receptor. If I loses affinity for the damaged receptor, we surmise that it will be released and subsequently bind to undamaged EGFR molecules (see I-EGFR).

View larger version (16K): [in this window] [in a new window]   Scheme 2.   The target-mediated selectivity component of our approach is based on the strong affinity of the C-molecule for the cytosolic domain of EGFR that may influence the equilibria TZ-I' (extracellular compartment)/TZ-I (cytosolic) (path 1) and TZ-I (cytosolic)/TZ-I-EGFR.

Recently, we reported preliminary evidence of the feasibility of a mixed EGFR/DNA-directed molecule termed SMA41 that contained a 3-methyl-1,2,3-triazene moiety appended to position 6 of a 4-anilinoquinazoline moiety (Matheson et al., 2001). It was shown to possess mixed EGFR/DNA-targeting properties on its own (IC₅₀ competitive binding = 0.2 µM) and to degrade into another inhibitor SMA52 (IC₅₀ competitive binding = 1 µM). The choice of the 3-alkyl-1,2,3-triazene as a DNA-alkylating moiety was inspired by its small size and its ability to heterolyze to an aromatic amine and an alkyldiazonium species that kill cells by alkylating DNA at position 6 of guanine (Cameron et al., 1985; Baig and Stevens, 1987). In comparison with temozolomide (TEM), a cyclic prodrug of the monoalkyltriazene 5-(3-methyltriazen-1-yl)imidazole-4-carboxamide (MTIC) (Stevens et al., 1984, 1987), we demonstrated that the potency of the C-molecule was superior to that of a classical combination of SMA52 and TEM at equitoxic doses. Although SMA41 possessed significant EGFR TK inhibitory activity, the stable molecule that it released (SMA52) was a weak EGFR TK inhibitor (IC₅₀ = 1 µM). Thus, we surmised that because the degradation of the TZ-I will lead to a stable inhibitor I that will remain longer in the cell medium, the potency of the latter may play a critical role in the overall antiproliferative activity. Therefore, we sought for a TZ-I capable of generating an I with stronger affinity than SMA52. Herein, we replaced the anilino methyl group by a less bulky chloro substituent to produce BJ2000, a C-molecule with 2-fold stronger affinity than SMA41 and capable of generating an I with 5-fold stronger affinity than SMA52. In this article, we demonstrated that the TZ-I BJ2000 was able to release the I termed FD105 (yield 87%) in cell culture medium supplemented with serum. This novel model exhibited 1) a mixed DNA-damaging and EGFR phosphotyrosine inhibitory activity, confirming path 1, 2, and 3; 2) a preferential inhibition of EGF-induced growth over PDGF and serum; and 3) a potent inhibition of EGFR-induced gene expression. More importantly, we demonstrated that BJ2000 induced irreversible inhibition of autophosphorylation, suggesting a pathway involving covalent binding of the alkylating product with EGFR (path 4) and partially irreversible inhibition of cell growth. This TZ-I or C-molecule has proven an invaluable tool for testing our Combi-Targeting postulates.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Drug Treatment

BJ2000 and FD105 were synthesized in our laboratories according to known procedures (Cameron et al., 1985; Manning et al., 1985; Rewcastle et al., 1995). Temozolomide was provided by Schering Plough (Kenilworth, NJ). In all assays, drug was dissolved in DMSO and subsequently diluted in RPMI 1640 medium containing 10% fetal bovine serum (FBS) (Wisent Inc., St-Bruno, QC, Canada) or in Dulbecco's modified Eagle's medium containing 10% bovine calf serum (Invitrogen, Burlington, ON, Canada) immediately before the treatment of cell cultures. In all assays, the concentration of DMSO never exceeded 0.2% (v/v).

Cell Culture

The cell line used in this study, the human epidermoid carcinoma of the vulva A431, was obtained from the American Type Culture Collection (Manassas, VA). The mouse fibroblasts NIH3T3 and NIH3T3HER14 (NIH3T3 cells stably transfected with EGFR gene) were generous gifts from Dr. Moulay Aloui-Jamali (Montreal Jewish General Hospital, Montreal, Canada). The A431 cell line was maintained in RPMI 1640 medium supplemented with 10% FBS and antibiotics as described previously (Matheson et al., 2001). NIH3T3 and NIH3T3HER14 cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% FBS and antibiotics. All cells were maintained in an atmosphere of 5% CO₂.

Degradation

The half-life of BJ2000 under physiological conditions was studied by UV-spectrophotometer. BJ2000 was dissolved in a minimum volume of DMSO, diluted with RPMI 1640 medium supplemented with 10% FBS, and absorbances read at 340 nm in a UV cell maintained at 37°C with a circulating water bath. The half-life was estimated by a one-phase exponential decay curve-fit method using the GraphPad software package (GraphPad Software, San Diego, CA).

The study of the conversion of BJ2000 into FD105 by HPLC was performed by adding BJ2000 (625 µM) to RPMI 1640 medium with 10% FBS (2 ml) and incubating it for different periods at 37°C. Thereafter, proteins were precipitated by addition of acetonitrile (3.5 ml) and the supernatant collected by centrifugation. The concentration of FD105 resulting from the degradation of BJ2000 was calculated using a standard curve obtained from the serial dilution of independently synthesized FD105 incubated in serum-containing medium under identical conditions. HPLC analyses were performed on a 1090 liquid chromatograph (Hewlett Packard, Palo Alto, CA) using a C₄ reverse phase column (15 µm, 300 × 3.9 mm; Waters, Milford, MA) to characterize and quantitate the products resulting from the degradation of BJ2000. The operating mode was isocratic and two solutions A (53% acetonitrile) and B (47% water) were used with a 0.5-ml/min flow rate and 10-µl injection volume. The peaks were detected at 254 nm. Under these conditions, independently synthesized FD105 and BJ2000 showed retention times around 10.5 and 15.2 min, respectively.

Kinase Assays

EGFR Kinase. This assay is similar to the one described previously (Matheson et al., 2001). Maxisorp 96-well plates (Nalge Nunc International, Naperville, IL) were incubated overnight at 37°C with 100 µl/well of 25 ng/ml PGT in PBS. Excess PGT was removed and the plate was washed three times with wash buffer (0.1% Tween 20 in PBS). The kinase reaction was performed by using 4.5 ng/well EGFR affinity-purified from A431 cells (Moyer et al., 1997; Vincent et al., 2000). The compound was added and phosphorylation initiated by the addition of ATP (20 µM). After 8 min at room temperature with constant shaking, the reaction was terminated by aspiration mixture and by rinsing the plate four times with wash buffer. Phosphorylated PGT was detected after a 25-min incubation with 50 µl/well of horseradish peroxidase-conjugated PY20 anti-phosphotyrosine antibody (Santa Cruz Biotechnology, Santa Cruz, CA) diluted to 0.2 mg/ml in blocking buffer (3% bovine serum albumin and 0.05% Tween 20 in PBS). Antibody was removed by aspiration and the plate washed four times with wash buffer. The signals were developed by the addition of 50 µl/well of 3,3',5,5'-tetramethylbenzidine peroxidase substrate (Kirkegaard and Perry Laboratories, Gaithersburg, MD), and after blue color development, 50 µl of H₂SO₄ (0.09 M) was added per well, and plates were read at 450 nm using an ELISA reader (model 2550; Bio-Rad, Hercules, CA).

c-src and Insulin Kinases. Conditions for the c-src and insulin kinase assays were the same as those for the EGFR kinase assay, except for the addition of 1 mM manganese chloride to the assay buffer and a final ATP concentration of 100 µM. The reaction was terminated by the addition of 50 µl of 250 mM EDTA before aspiration. For experiments comparing inhibition of EGFR with c-src-kinase or insulin receptor, Baculovirus-expressed human c-src (1.2 units/well; Upstate Biotechnology, Lake Placid, NY) or baculovirus-expressed cytoplasmic domain of the insulin receptor  subunit (15 ng/ml; BIOMOL Research Laboratories, Plymouth Meeting, PA) was substituted for the EGFR. As positive controls, PP1 (BIOMOL Research Laboratories), a selective inhibitor of c-src (IC₅₀ = 170 nM), and HNPMA-(AM)₃ (BIOMOL Research Laboratories), an inhibitor of insulin receptor (IC₅₀ = 100 µM), were used.

Autophosphorylation Assay

Inhibition of receptor autophosphorylation in viable cells was determined by anti-phosphotyrosine Western blots. Cells were grown to confluence in six-well plates, washed twice with PBS, and exposed to serum-free medium for 18 h. They were subsequently treated with the compounds for 90 min and then with EGF (100 ng/ml) for 10 min (A431 cells), or with PDGF (100 ng/ml) for 10 min (NIH3T3 cells). Thereafter, they were washed with PBS and resuspended in cold lysis buffer (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 1% Nonidet P-40, 1 mM EDTA, 5 mM NaF, 1 mM Na₃VO₄, and protease inhibitor tablet; Roche Applied Science, Laval, QC, Canada). The lysates were kept on ice for 30 min and collected by centrifugation at 10,000 rpm for 20 min at 4°C. Protein concentrations were determined against a standardized control using a protein assay kit (Bio-Rad). Equal amounts of protein from each cell lysate were added to an 8% SDS-polyacrylamide gel electrophoresis and transferred to a polyvinylidene difluoride membrane (Millipore, Bedford, MA). Nonspecific binding on the membrane was minimized with a blocking buffer containing 3% nonfat dry milk in PBS. The membrane was blotted for 1 h with anti-phosphotyrosine antibody PY20 (NeoMarkers, Fremont, CA) or anti-EGFR antibodies (NeoMarkers) and anti--tubulin antibodies (NeoMarkers) for the detection of equal loading. It was subsequently incubated with horseradish peroxidase-goat anti-mouse antibody (Bio-Rad) and the bands visualized with an enhanced chemiluminescence system (Amersham Biosciences UK, Ltd., Little Chalfont, Buckinghamshire, UK). Band intensities were measured using the GeneTools software package (SynGene, Cambridge, UK).

For the study of inhibition of mitogen-activated protein kinase (MAPK) activation by BJ2000, protein lysates were obtained as described above and Western blot was performed as reported by Tari and Lopez-Berestein (2000). The membrane was incubated with anti-phosphorylated MAPK (Erk2) antibodies or antibodies specific for Erk2 (Cell Signaling Technology Inc., Beverly, MA).

Reverse EGFR Autophosphorylation

This assay was performed as described by Fry et al. (1998). A431 cells were grown to confluence in six-well plates and then incubated in serum-free medium for 18 h. Duplicate sets of cells were then treated with 30 µM of each compound for 90 min. One set of cells was then stimulated with EGF (100 ng/ml) for 10 min, and extracts were made as described under the Western blotting procedure described above. The other set of cells was washed free of the compound with warmed serum-free media and incubated for 2 h. Thereafter, the cells were washed, incubated for another 2 h, washed again, and then incubated for a further 4 h. This set of cells was then stimulated with EGF, and extracts were prepared as for the first set.

RT-PCR for c-fos Expression

A431 cells were grown to confluence in six-well plates and then incubated in serum-free medium for 18 h. Cells were exposed to the indicated concentrations of drug before stimulation with EGF (50 ng/ml) for 30 min. Total RNA was isolated using the High Pure RNA Isolation kit (Roche Applied Science, Mannheim, Germany), following the manufacturer's instructions. Quantitative analysis of c-fos mRNA and GAPDH mRNA (2 µg of RNA for each sample) was preformed by Titan One Tube RT-PCR kit (Roche Applied Science), following the manufacturer's instructions and using the following primers: 5'ATGATGTTCTCGGGCTTC3' (sense) and 5'CTCCTGCCAATGCTCTGC3' (antisense) for c-fos, and 5'CCATGGAGAAGGCTGGGG3'(sense) and 5'CAAAGTTGTCATGGATGACC3' (antisense) for GAPDH.

In Vitro Growth Inhibition Assay

To study the effect of our compounds on growth factor-stimulated proliferation, cells were grown to 70% of confluence in 48-well plates and washed twice with PBS after which they were exposed to serum-free medium for 18 h. Cells were exposed to each drug and growth factors (EGF, TGF, PDGF, or serum) for 72 h and cell growth measured using the sulforhodamine B (SRB) assay (Skehan et al., 1990). Briefly, after drug treatment, cells were fixed using 50 µl of cold trichloroacetic acid (50%) for 60 min at 4°C, washed five times with tap water, and stained for 30 min at room temperature with SRB (0.4%) dissolved in acetic acid (0.5%). The plates were rinsed five times with 1% acetic acid and allowed to air dry. The resulting colored residue was dissolved in 200 µl of Tris base (10 mM) and optical density read for each well at 540 nm using a microplate reader (model 2550; Bio-Rad). Each point represents the average of at least two independent experiments run in triplicate.

To study the reversibility of the antiproliferative effects of our compounds, cells were grown to 70% confluence in 96-well plates and subsequently washed twice with PBS after which they were exposed to serum-free medium for 18 h. Under continuous exposure, cells were exposed to different concentrations of each drug for 120 h. Under short exposure, they were exposed to each drug for 2 h, after which they were allowed to recover for 120 h in drug-free medium. Growth inhibitory activities were evaluated using the SRB assay as described above. Each point represents the average of at least two independent experiments run in triplicate.

Alkaline Comet Assay for Quantitation of DNA Damage

The alkaline comet assay was performed as described previously (Matheson et al., 2001). The cells were exposed to drugs (BJ2000, FD105, or TEM) for 30 min, harvested with trypsin-EDTA, subsequently collected by centrifugation, and resuspended in PBS. Cell suspensions were diluted to approximately 10⁶ cells and mixed with agarose (1%) in PBS at 37°C in a 1:10 dilution. The gels were cast on Gelbond strips (Mandel Scientific, Guelph, ON, Canada) using gel casting chambers as described previously (McNamee et al., 2000) and then immediately placed into a lysis buffer [2.5 M NaCl, 0.1 M tetra-sodium EDTA, 10 mM Tris-base, 1% (w/v) N-lauryl sarcosine, 10% (v/v) DMSO, and 1% (v/v) Triton X-100, pH 10.0]. After being kept on ice for 30 min, the gels were gently rinsed with distilled water and immersed in a second lysis buffer (2.5 M NaCl, 0.1 M tetrasodium EDTA, and 10 mM Tris-base) containing 1 mg/ml proteinase K for 60 min at 37°C. Thereafter, the gels were rinsed with distilled water, incubated in alkaline electrophoresis buffer for 30 min at 37°C, and electrophoresed at 300 mA for 60 min. The gels were subsequently rinsed with distilled water and placed in 1 M ammonium acetate for 30 min. Thereafter, they were soaked in 100% ethanol for 2 h, dried overnight, and stained with SYBR Gold (1/10, 000 dilution of stock supplied from Molecular Probes, Eugene, OR) for 20 min. Comets were visualized at 330× magnification and DNA damage was quantitated using the Tail Moment parameter (i.e., the distance between the barycenter of the head and the tail of the comet multiplied by the percentage of DNA within the tail of the comet). A minimum of 50 cell comets was analyzed for each sample, using ALKOMET, version 3.1, image analysis software.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Degradation of BJ2000

The half-life of BJ2000 measured by UV-spectrophotometry at 340 nm was 34 min in RPMI 1640 medium supplemented with 10% FBS at 37°C; however, we noted that absorbance at this wavelength reached early saturation at high absorbance units (e.g., 0.6). Therefore, the study was performed with HPLC analysis whereby the decreasing and increasing peaks associated with the disappearance of BJ2000 and appearance of FD105, could be clearly monitored. As expected, an inverse relationship was observed between the areas of the peaks associated with these two species with observed half-life of 75 min for BJ2000. Calculations based upon area/concentration standard curve [Area = 2.796x (concentration)  38.65; R² = 0.99] indicated that BJ2000 was converted into FD105 in an 87% yield after a 24-h incubation in RPMI 1640 medium supplemented with 10% FBS at 37°C (Fig. 1).

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Degradation of BJ2000 to FD105 in RPMI 1640 medium supplemented with 10% FBS at 37°C. Experiments were carried out by reverse phase HPLC analysis: BJ2000 (625 µM) was added to RPMI 1640 medium with 10% FBS (2 ml) and incubated for different periods at 37°C. Thereafter, proteins were precipitated by addition of acetonitrile (3.5 ml) and the supernatant collected by centrifugation. The concentration of FD105 resulting from the degradation of BJ2000 was calculated using a standard curve obtained from the serial dilution of independently synthesized FD105 incubated in serum-containing medium under identical conditions.

Inhibition of EGFR TK Activity

In a competitive EGFR binding assay, BJ2000 (IC₅₀ = 0.1 µM) showed 2-fold greater binding affinity than its metabolite FD105 (IC₅₀ = 0.2 µM) for the ATP site of purified EGFR. Therefore, both the drug and its corresponding prodrugs showed significant affinity for EGFR. As previously reported (Matheson et al., 2001), TEM did not show any effect on the tyrosine kinase activity of this receptor (IC₅₀ > 100 µM) (Fig. 2A). Furthermore, the EGFR selectivity of these two agents was tested by comparing their EGFR inhibitory activities with those of other TKs such as c-src and the insulin receptor. In ELISA assays, BJ2000 and its metabolite FD105 did not block c-src TK activity nor did it exert any effect on insulin receptor TK in the 1 to 100 µM range, indicating the selectivity of these agents for EGFR (Fig. 2, B and C).

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Selectivity of BJ2000 and FD105 for EGFR TK (A) versus c-src kinase (B) and insulin receptor kinase (C). Inhibition of purified kinases was measured as described under Materials and Methods. Each point represents at least two independent experiments.

Inhibition of EGFR Autophosphorylation

Western blot analysis demonstrated that BJ2000 blocked EGF-induced EGFR autophosphorylation in A431 cells in a dose-dependent manner with an IC₅₀ value of ~6 µM without affecting the levels of EGFR (Fig. 3A). At concentrations as high as 30 µM, it had no effect on PDGF-induced PDGFR autophosphorylation in NIH3T3 cells (Fig. 3B). These results represent further evidence of BJ2000 selectivity for EGFR.

View larger version (39K): [in this window] [in a new window]   Fig. 3.   Selective inhibition of EGFR and PDGFR autophosphorylation in intact cells by BJ2000. Serum starved A431 cells (A) or NIH3T3 cells (B) were preincubated for 90 min with the indicated concentrations of BJ2000 before stimulation with EGF (A) for 10 min or PDGF (B) for 10 min. Equal amount of cell lysates was analyzed by Western blotting using anti-phosphotyrosine antibodies. Membranes were stripped of anti-phosphotyrosine and reprobed with anti-EGFR or anti--tubulin antibodies as a loading control. Band intensities were measured using the GeneTools software package (SynGene).

Mechanism of EGFR Inhibition

Unlike FD105, BJ2000 is a reactive molecule capable of alkylating nucleophiles. Thus, we surmised that it might inflict some covalent damage to the ATP site of EGFR, thereby inducing irreversible inhibition. To test this hypothesis, we used the reversibility assay described by Smaill et al. (2000) and Fry et al. (1998) according to which the cells are treated with drug for 90 min and the culture medium repeatedly removed and replaced three times after treatment, after which EGFR autophosphorylation is measured. As expected, BJ2000 and FD105 at 30 µM completely suppressed EGF-dependent EGFR autophosphorylation in A431 cells immediately after drug exposure. However, at 8 h post-treatment in drug-free medium (after repeated washouts), only 40% of the EGFR autophosphorylation activity was restored in cells treated with BJ2000, indicating that the latter is capable of inducing partially irreversible inhibition of EGFR autophosphorylation. In contrast, 96% of EGFR autophosphorylation activity was restored in cells treated with FD105 at the same dose. As expected, TEM did not show any inhibitory activity immediately after the 90-min treatment, nor did it induce any effect at 8 h post-treatment (Fig. 4).

View larger version (41K): [in this window] [in a new window]   Fig. 4.   Reversal of EGFR autophosphorylation in the presence of BJ2000, FD105, or TEM in A431 cells. Duplicate sets of cells were treated with 30 µM of designated compound to be tested as a reversible EGFR inhibitor for 90 min. One set of cells was then stimulated with EGF for 10 min, and extracts were made as described under the Western blotting procedure. The other set of cells was washed free of the compound with warmed serum-free media, incubated for 2 h, washed again, incubated for another 2 h, and incubated for a further 4 h after a subsequent wash. This set of cells was then stimulated with EGF, and extracts were made similar to the first set. The phosphotyrosine and EGFR were detected as described under Materials and Methods.

Antiproliferative Activity of BJ2000

Inhibition of Growth Factor-Stimulated Proliferation. SRB assays demonstrated that like FD105 (Fig. 5A), BJ2000 (Fig. 5B) was capable of selectively blocking EGF or TGF-induced proliferation in NIH3T3 cells stably transfected with the EGFR gene (NIH3T3HER14) (100% growth inhibition at 1 µM). This C-molecule and its metabolite were approximately 30-fold less effective in inhibiting PDGF-stimulated growth (Fig. 5, A and B) (100% inhibition at around 30 µM). Similarly, they exhibited a lesser effect on serum-stimulated growth in NIH3T3HER14 cells (100% growth inhibition at concentrations >30 µM) (Fig. 5, A and B). These EGFR-selective effects are in agreement with those observed from ELISA and whole cell autophosphorylation assays.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Effect of FD105 and BJ2000 on growth factor-stimulated proliferation in NIH3T3HER14 cells. A, FD105 + EGF (TGF, PDGF, or serum). B, BJ2000 + EGF (TGF, PDGF, or serum). Cells were exposed to each drug and growth factors (EGF, TGF, PDGF, or serum) for 72 h. Cell growth was measured using SRB assay. Each point represents at least two independent experiments run in triplicate.

Reversibility of Growth Inhibitory Activity. The A431 cells express TGF and overexpress its cognate receptor EGFR, leading to an aggressive autocrine cell growth. These cells have been shown to be sensitive to antiproliferative agents targeting the EGFR both in vitro or in vivo (Lanzi et al., 1997). Moreover, they also express the DNA repair enzyme O⁶-alkylguanine transferase (AGT) known to be responsible for resistance to monoalkyltriazenes of the same class as BJ2000 (Mitchel and Dolan, 1993). Thus, this cell line was found an appropriate model for testing the sustainability of the antiproliferative effects of our different agents. Under 120-h continuous exposure, the results obtained from SRB assay as illustrated by Fig. 6A, showed that BJ2000 was ~2-fold more potent (IC₅₀ = 15 µM) than its metabolite FD105 alone (IC₅₀ = 47 µM; Fig. 6B) in the AGT-proficient cell line A431. In contrast, TEM at concentrations as high as 200 µM did not show any significant antiproliferative activity in these cells (Fig. 6C). More importantly, in a short exposure assay (2 h) followed by 120-h recovery, an almost complete loss of activity was observed for FD105 in the A431 cell line (IC_{50 >} = 100 µM; Fig. 6B), indicating that it induced significantly reversible growth inhibitory activity. In contrast, BJ2000 showed significant retention of activity (IC₅₀ = 38 µM; Fig. 6A), indicating a more sustained effect of the latter.

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Reversibility of antiproliferative effect of BJ2000 (A), FD105 (B), and TEM (C) in A431 cells. Cells were exposed to each drug for 2 h, after which they were allowed to recover for 120 h in drug-free medium or continuously for 120 h. Cell growth was measured using SRB assay. Each point represents at least two independent experiments run in triplicate.

Inhibition of EGFR-Mediated Signaling

To determine whether blockade of EGFR autophosphorylation translates into inhibition of downstream signaling, we analyzed the effect of the C-molecule on EGF-induced phosphorylation of MAPK (Erk2) and c-fos expression in A431 cells. The results showed that BJ2000 induced complete inhibition of Erk2 phosphorylation at concentrations as low as 1 µM without affecting the levels of Erk2 (Fig. 7). Similarly, RT-PCR analysis showed that BJ2000 induced nearly 100% inhibition of EGF-mediated c-fos gene expression at low concentrations (0.3-3 µM) (Fig. 8), indicating that inhibition of EGFR phosphorylation by the C-molecule is accompanied by a significant blockade of EGFR-dependent downstream signaling.

View larger version (27K): [in this window] [in a new window]   Fig. 7.   Effect of BJ2000 on MAPK (Erk2) activation in A431 cells. Serum-starved cells were preincubated for 90 min with the indicated concentrations of BJ2000 before stimulation with EGF for 10 min. Protein lysates were obtained, and Western blot was performed as described by Tari and Lopez-Berestein (2000).

View larger version (35K): [in this window] [in a new window]   Fig. 8.   Effect of BJ2000 on c-fos gene expression in A431 cells. Serum-starved cells were preincubated for 90 min with the indicated concentrations of BJ2000 before stimulation with EGF for 30 min. Quantitative analysis of c-fos and GAPDH was preformed by RT-PCR, as described under Materials and Methods.

Quantitation of DNA Damage

Using the alkaline comet assay, we demonstrated that, in contrast to FD105 (Fig. 9) and like TEM (Matheson et al., 2001), BJ2000 induced a dose-dependent DNA damage in both A431 and NIH3T3HER14 cells after a 30-min drug exposure (Fig. 9), implying indirect evidence of the formation of metastable methyldiazonium species.

View larger version (16K): [in this window] [in a new window]   Fig. 9.   Quantitation of DNA damage using the alkaline comet assay. Tail moment was used as a parameter for the detection of DNA damage in A431 and NIH3T3HER14 cells exposed to BJ2000 and FD105 for 30 min. Each point represents at least two independent experiments.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Agents targeting EGFR and its closest homolog p185^neu, the erbB2 gene product, present two major advantages: 1) they induce target-selective antitumor activities and 2) they exhibit good toxicity profiles. However, where they cannot induce apoptosis, they are cytostatic agents that induce reversible antitumor effects. For sustained antitumor activity, combinations with other cytotoxic drugs (e.g., cisplatin, doxorobucin, and taxans) have proven a useful alternative (Ciardiello et al., 2000). However, the lack of selectivity of the latter agents may negatively alter the overall toxicity profiles of these regimens. We now propose a novel approach to this problem that seeks to combine EGFR TK inhibitors with the pharmacophores of known cytotoxic DNA-damaging properties into single chimeric molecules targeted to EGFR. As outlined in Scheme 2, the target-mediated selectivity component of our approach is based on the strong affinity of the C-molecule for the cytosolic domain of EGFR that may influence the equilibria TZ-I' (extracellular compartment)/TZ-I (cytosolic) (path 1) and TZ-I (cytosolic)/TZ-I-EGFR (Scheme 2). Both bound and unbound fractions of TZ-I will eventually degrade to generate I that may further block EGFR TK (path 5). Fractions of unbound TZ-I may diffuse through the nucleus where the generated methyldiazonium species (TZ) may alkylate and damage DNA (path 6). Bound fractions may react with amino acid residues of the active site of the receptor, thereby irreversibly inhibiting it. More importantly, unreacted fractions of the in situ generated TZ may diffuse away from the receptor. If the in situ-generated I loses affinity for the damaged receptor, it may bind to other nondamaged receptor molecules, leading to a more sustained EGFR TK inhibition (path 4). Using BJ2000 as a probe, the work discussed herein gave prima facie support to our postulates.

BJ2000 was designed to possess two major components: an EGFR inhibitory component imprinted into the quinazoline moiety and a DNA-damaging methyldiazonium species masked by the appended 3-methyl-1,2,3-triazene moiety. The effects of the EGFR component were demonstrated by both enzyme and whole cell assays, whereby the ability of the C-molecule to block substrate (ELISA) and EGFR autophosphorylation (Western blotting) was clearly shown. Furthermore, BJ2000 exhibited selectivity for EGFR in both ELISA and growth factor-stimulated proliferation assays. The reversibility assay showed an only 40% recovery of the autophosphorylation activity, 8 h after cell exposure to BJ2000, despite multiple washouts. In contrast, almost complete recovery of initial EGFR autophosphorylation activity was observed in cells treated with FD105 at the same dose (Fig. 4). It is not clear at this point whether the partial inactivation of the EGFR TK occurred through alkylation of amino acid residues located at the active site. Nevertheless, Smaill et al. (2000) and Fry et al. (1998) demonstrated that acryloyl moieties attached to position 6 of quinazoline reacted with cysteine 773 of EGFR, leaving an irreversibly inhibited receptor. The triazene chain of our C-molecule being appended to the same position and possessing approximately the same length as the acryloyl moiety is likely to undergo a similar type of interaction.

As for the DNA-damaging component, we compared our activities with those of the clinical drug TEM, a cyclic triazene that generates an open chain monoalkyltriazene of the same type as BJ2000 (Baig and Stevens, 1987). Like TEM, BJ2000 induced significant DNA damage in A431 cells after a 30-min drug exposure (Fig. 9). This represents an indirect confirmation of the formation of the DNA-damaging methyldiazonium species (TZ) postulated by path 6. The complete confirmation of the latter path that predicts the hydrolytic conversion of TZ-I into TZ + I was provided by HPLC detection of I (FD105) and the kinetic analysis. The results clearly indicated an inverse relationship between the degradation of BJ2000 (t_1/2 = 75 min) and the formation of FD105 in an 87% yield after complete degradation (Fig. 1, 4-24 h). Thus far, the results in toto lend support to almost all the postulates depicted in Scheme 2 (paths 1-6), according to which a TZ-I (or C-molecule) can bind to EGFR (paths 1-3) or degrade to another high-affinity inhibitor I (paths 2 and 5) plus a reactive species that can damage DNA (path 6) and perhaps the receptor (path 4).

The principle by which the multiple properties of BJ2000 should culminate into sustained growth inhibition was demonstrated in comparison with its derived inhibitor FD105. In contrast to the latter, BJ2000 antiproliferative activity in A431 cells was partially retained as long as 5 days after a 2-h drug exposure. The complete loss of activity of FD105 and the marked inactivity of TEM, a known DNA-damaging agent, in the A431 cells, suggest that the sustained antiproliferative activity of our C-molecule may result from an interactive effect between its DNA-damaging component and its signal transduction inhibitory properties. This assumption may be further corroborated by our demonstration of the ability of the C-molecule to block MAPK phosphorylation and c-fos gene expression at low concentrations (1-3 µM). Because MAPK is a critical signal transduction protein known to induce c-fos gene expression and to mediate the mitogenic effects of EGF-activated signaling (Davis, 1993, 1995), blockade of EGFR autophosphorylation by our C-molecule and by its derived inhibitor I may translate into inhibition of downstream signaling associated with expression of genes required to rescue the cells. Indeed, Kaina et al. (1997) demonstrated that c-fos-deficient cells exhibited a more severe mutagen-induced block to DNA replication and were compromised in the abolition of replication blockage. Thus, the mechanism by which blockade of EGF-induced signal transduction by the TK inhibitory element down-regulates DNA repair enzymes (e.g., AGT and DNA glycosylases) or other critical proteins can be evoked to rationalize the EGFR-mediated sustained inhibition of proliferation induced by our C-molecule. We are now verifying this hypothesis by analyzing the regulatory DNA repair genes induced in response to cell exposure to our C-molecules.

In summary, the testing of our Combi-Targeting postulates has led to the development of a potent molecule capable of behaving as a masked form of multiple effects: 1) direct inhibition of EGFR autophosphorylation, 2) inhibition of the same target by its derived inhibitor I, 3) induction of DNA damage and perhaps covalent damage of EGFR, 4) sustained growth inhibitory activity, and 5) inhibition of EGFR-mediated MAPK activation and c-fos gene expression. Thus, our work may contribute to the development of novel binary tumor targeting strategies with the prospect of circumventing the adverse effects associated with the lack of selectivity and potency of classical chemotherapy.