AP23846, a novel and highly potent Src family kinase inhibitor, reduces vascular endothelial growth factor and interleukin-8 expression in human solid tumor cell lines and abrogates downstream angiogenic processes

Introduction

The Src family kinases are composed of nine structurally related, membrane-associated, nonreceptor protein tyrosine kinases (1, 2). Src family kinases are overexpressed and/or aberrantly activated in a variety of human tumors, including colon, breast, pancreatic, bladder, head and neck, ovarian, and brain (1). The degree of overexpression and/or activation of Src family kinases correlates strongly with metastatic potential in colon and breast cancers (3–9) and with survival of colon cancer patients (10). In recent years, work by several groups has shown that Src family kinases are important for multiple aspects of tumor growth and progression, including proliferation, migration, invasion, survival, and angiogenic factor production, depending on the tumor type (11–16). For these reasons, interest in development of Src family kinase inhibitors has increased sharply over the past several years both for their use as research tools, delineating functions specific to these kinases, and as potential anticancer therapeutics (17).

For nearly the past decade, the pyrazolopyrimidines PP1 and PP2 have been the standards for Src family kinase inhibition in vitro and in vivo. Duxbury et al. recently showed a decrease in pancreatic tumor growth and metastasis in nude mice treated with PP2, in combination with gemcitabine, relative to controls and either drug alone (18). Although PP1 and PP2 inhibit Src family kinase activity with IC₅₀s of ∼5 nmol/L in vitro, concentrations to 10 μmol/L are often necessary to achieve complete Src family kinase inhibition in cell culture (19). When used at these concentrations, “off-target” kinases, including CSK and p38 mitogen-activated protein kinase, are often inhibited as well, thus limiting the ability of these inhibitors to delineate Src family kinase–specific effects in the absence of supporting data from dominant negatives, small interfering RNA (siRNA), or gene knockout studies (20).

Recently, significant advances in structure-based drug design, computational chemistry, high-throughput biological screening, and synthetic chemistry have contributed to the development of a new generation of small-molecule Src family kinase inhibitors (21) improved in both potency and selectivity over PP1 and PP2. Several Src family kinase inhibitors, exemplifying preclinical lead compounds or clinical candidates, have shown encouraging results both in vitro and in vivo. Examples of such Src family kinase inhibitors include the quinoline SKI-606 (22), the pyridopyrimidine PD180970 (23), the quinazoline AZM-475271 (24), the thiazole BMS-354825 (25), and the purines AP23464 and AP23848 (26, 27). BMS-354825 is currently in clinical trial in chronic myelogenous leukemia due to its ability to inhibit Abl and Src and, more specifically, Bcr-Abl mutants that lead to imatinib resistance (25).

Using AZM-475271 in combination with gemcitabine, Yezhelyev et al. recently showed reduced vascularity, tumor size, and metastasis as well as increased apoptosis in human pancreatic cancer cells grown orthotopically in nude mice. These results suggest that Src inhibitors may be efficacious in treating pancreatic adenocarcinomas, but which functions are directly attributable to Src inhibition has not been evaluated carefully (24).

In this study, we describe for the first time a new purine analogue, AP23846 (Fig. 1 ), which shows excellent potency for inhibition of c-Src kinase and selectivity relative to several other oncogenic protein kinases [e.g., Abl, Kit, vascular endothelial growth factor (VEGF) receptor-2/KDR, and HER-2]. Although this inhibitor exhibits in vivo toxicity that renders it unsuitable for human studies, its potency and selectivity make it ideal to examine tumorigenic functions mediated by Src family kinases. We show inhibition of both interleukin-8 (IL-8) and VEGF expression, suggesting specific roles for Src inhibitors as antiangiogenic agents.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Structure of AP23846. Src inhibitors were based on the generic structure, a 2,6,9-trisubstituted purine template. IC₅₀ of various analogues for Src.

Materials and Methods

AP23846 and PP2

AP23846 and two structurally related analogues, AP23848 and AP23980, were designed and synthesized using methods as described previously (28). Details of the chemistry and structure-activity relationships of these and additional purine template-based analogues will be described elsewhere. PP2 was obtained from Calbiochem (Calbiochem/EMD Biosciences, San Diego, CA).

Src Kinase Inhibition and Selectivity Assays

Src kinase inhibition was measured using a LANCE assay that uses a biotinylated substrate (PKS1) that is quantitated via a europium-labeled anti-phosphotyrosine antibody and allophycocyanin-streptavidin. In the kinase assay, Src is first activated with ATP and then mixed with a biotinylated substrate peptide in the presence or absence of ARIAD Pharmaceuticals (Cambridge, MA) compounds (AP compounds; i.e., AP23846). The final concentrations of Src kinase (Upstate Biological, Charlottesville, VA) and PKS1 were 165 pmol/L and 50 nmol/L, respectively. The incubation buffer is composed of 20 mmol/L sodium HEPES (pH 7.4), 0.1 mg/mL bovine serum albumin, 1 mmol/L ATP, 10 mmol/L MgCl₂, and 0.41 mmol/L DTT. Src kinase was completely activated (Y419 phosphorylation) after 1.5 hours of preincubation. Incubation with substrate and inhibitor proceeded for 2 hours at 37°C. After the kinase assay incubation period, excess Src inhibitor was added to stop the kinase reaction along with europium-labeled anti-phosphotyrosine antibody and allophycocyanin-streptavidin. The biotinylated substrate peptide (phosphorylated or unphosphorylated by Src kinase in the absence or presence of inhibitors, respectively) binds to allophycocyanin-streptavidin as a function of biotin-avidin interaction. However, the europium-labeled anti-phosphotyrosine antibody binds only to substrate that has been phosphorylated by Src kinase. When the solution is excited at 615 nm, there is an energy transfer from the europium to the allophycocyanin when they are in close proximity (i.e., attached to the same molecule of biotinylated and phosphorylated substrate peptide). The allophycocyanin then fluoresces at a wavelength of 665 nm. Quantitation of europium-labeled anti-phosphotyrosine antibody binding to phosphorylated, biotyinlated substrate is accomplished by measurement of fluorescence emission at 665 nm following excitation at 615 nm using a Wallac Victor² V plate reader (Perkin-Elmer, Boston, MA).

Src kinase selectivity was determined at Upstate Discovery Laboratories (Charlottesville, VA) for AP23846 (Table 1 ) using a radiolabeling method consisting of [γ-³³P]ATP and slightly different assay conditions relative to some of the protein kinases tested. In Src, Abl, epidermal growth factor receptor, Flt3, platelet-derived growth factor receptor (PDGFR)-α, and PDGFR-β kinase inhibition, the assay buffer consisted of 20 mmol/L MOPS (pH 7.0), 1 mmol/L EDTA, 0.1% β-mercaptoethanol, 0.01% Brij-35, 5% glycerol, and 1 mg/mL bovine serum albumin. Details for Src and Abl kinase assays are described as follows. For Src kinase, in a final reaction volume of 25 μL, human c-Src (5–10 milliunits) is incubated with 8 mmol/L MOPS (pH 7.0), 0.2 nmol/L EDTA, 250 μmol/L peptide substrate (Cdc2 fragment KVEKIGEGTYGVVYK), 10 mmol/L magnesium acetate, and [γ-³³P]ATP (specific activity ∼500 counts/min/pmol). For Abl kinase, in a final reaction volume of 25 μL, human Abl (5–10 milliunits) is incubated with 8 mmol/L MOPS (pH 7.0), 0.2 mmol/L EDTA, 50 μmol/L peptide substrate (EAIYAAPFAKKK), 10 mmol/L magnesium acetate, and [γ-³³P]ATP-containing buffer solutions. After incubation for 40 minutes at room temperature, the reactions were stopped by the addition of 5 μL of a 3% phosphoric acid solution. The reaction solutions (10 μL) were then spotted onto a P30 filter mat and washed (thrice for 5 minutes in 75 mmol/L phosphoric acid and once in methanol) before drying and scintillation counting.

Cell Culture and Reagents

The L3.6pl cell line was a generous gift from Dr. Lee Ellis (University of Texas M.D. Anderson Cancer Center, Houston, TX). Cells were maintained in MEM supplemented with 10% FCS (Hyclone, Logan, UT), 2 mmol/L l-glutamine (Life Technologies, Grand Island, NY), and 0.6% penicillin/streptomycin (Life Technologies). For inhibition experiments, 1 × 10⁶ cells were plated in 10-cm dishes and maintained in MEM with 10% fetal bovine serum. At 70% to 80% confluence, the cell culture medium was replaced with serum-free MEM (5 mL) supplemented with the desired concentration of inhibitor or a corresponding volume of DMSO. Cell lysates were harvested after 1, 2, 4, 8, 12, 24, and 48 hours for Western blot analysis. The cells and supernatants were harvested after 24 hours for quantitation of VEGF and IL-8 protein levels by ELISA. Cells were photographed at ×100 magnification using a Nikon Coolpix 4500 digital camera (Nikon, Melville, NY).

Creation of siRNA Expression Plasmids Silencing c-Src Gene Expression

siRNA expression plasmids were generated using the Ambion pSilencer 1.0-U6 (Ambion, Austin, TX) system according to the manufacturer's protocol. Target sequences from the c-Src gene sequence were designed using the Ambion siRNA Web design tool. Two target sequences were used: 5′-AACAAGAGCAAGCCCAAGGAT-3′ (52–71 bp) and 5′-AAGCTGTTCGGAGGCTTCAAC-3′ (226–244 bp). Corresponding oligonucleotides harboring flanking Apa1 (5′) and R1 (3′) ends were obtained from Invitrogen Life Technologies (Carlsbad, CA). These oligonucleotides were ligated into the siRNA expression plasmid at compatible sites. Constructs were confirmed by DNA sequencing. The siRNA plasmids (0.5 ng each) were transfected into L3.6pl cells, along with 10 ng pcDNA G418 resistance promoter-less plasmid, via the Transfast transfection kit (Promega Corp., Madison, WI). Cells were grown in medium containing 600 μg/mL G418 (Mediatech, Inc., Herndon, VA) for selection of positive clones. Negative controls were transfected with an empty vector pcDNA plasmid at identical concentrations. Total c-Src expression in siRNA clones was determined by Western blot analysis.

Western Blot Analysis

Cells were washed twice with ice-cold TBS followed by lysis for 10 minutes in radioimmunoprecipitation assay buffer [20 mmol/L sodium phosphate buffer, 150 mmol/L NaCl, 5 mmol/L EDTA, 1% Triton X-100, 0.5% sodium deoxycholate (pH 7.4)] supplemented with one tablet of Complete Mini-EDTA protease inhibitor cocktail (Roche Diagnostic, Mannheim, Germany) and sodium orthovanadate (1 mmol/L; pH 7.4). Cells were scraped from plates, and cell lysates were clarified by centrifugation for 15 minutes at 13,000 rpm and 4°C. Total protein concentration was determined via the Bio-Rad D_c protein assay (Bio-Rad Laboratories, Hercules, CA) followed by spectrophotometric analysis using the TECAN Genios plate reader and Magellan version 4.0 software. Total cell protein (50 μg) was separated via 8% SDS-PAGE, transferred to polyvinylidene difluoride membrane (Millipore, Bedford, MA), and blocked for 30 minutes at room temperature with 5% nonfat milk/TBS-Tween 20. Primary antibodies were diluted 1:1,000 in blocking buffer and incubated overnight at 4°C with gentle rocking. Horseradish peroxidase–conjugated secondary antibodies (Bio-Rad goat anti-mouse and sheep anti-rabbit) were diluted 1:2,000 in 5% milk/TBS-Tween 20 and incubated for 1 hour at room temperature with gentle rocking. Proteins were visualized by incubation with enhanced chemiluminescence detection reagents (Perkin-Elmer, Boston, MA) and exposure to film (Kodak Biomax MR, Rochester, NY). The anti-Src monoclonal antibody (327) was obtained from Calbiochem/Oncogene Research Products (Calbiochem-Novabiochem, La Jolla, CA). The anti-phospho-Src^Y418, the anti-phospho-paxillin^Y118, and anti-paxillin polyclonal antibodies were obtained from Cell Signaling Technology (Beverly, MA).

Immunoprecipitations

Total cell protein (500 μg) was normalized to a total volume of 650 μL with radioimmunoprecipitation assay buffer and rotated with 6 μL anti-Src monoclonal antibody overnight at 4°C. The next day, 50 μL of 1:1 slurry of protein G agarose (Upstate Biological) in radioimmunoprecipitation assay buffer was added and incubated for an additional hour with rotating at 4°C. Bound proteins were pelleted by centrifugation, washed thrice with radioimmunoprecipitation assay buffer, and eluted by boiling in 1× Laemmli sample buffer. Bound proteins were then subjected to Western blot analysis as described above.

Proliferation Assays

L3.6pl cells (5 × 10⁵) were plated in triplicate in 100-mm tissue culture dishes in complete medium. Twenty-four hours after plating, cell medium was replaced with complete medium containing either 1 μmol/L AP23846 or an equal volume of DMSO. Viable cells were determined via trypan blue exclusion and counted after 24 and 48 hours of incubation in the presence or absence of inhibitor.

Cell Cycle Analysis

Cells were fixed in 70% ethanol and stained with propidium iodide as described previously. Stained cells were analyzed by fluorescence-activated cell sorting using a flow cytometer (Coulter Epics XL-MCL, Coulter Corp., Miami, FL) as described elsewhere (15).

Migration Assays

Migration assays were carried out via the modified Boyden chamber assay according to the manufacturer's protocol and as described previously (29). Briefly, L3.6pl cells (2 × 10⁵) or hepatic endothelial cells (1 × 10⁵) derived from the ImmortoMouse (ImmortoMice, Wilmington, MA, CBA/ca × C57BL/10 hybrid; Charles River Laboratories; ref. 30) were suspended in the upper well of the migration chamber (control inserts, 8 μm pore size, Becton Dickinson, Bedford, MA) in 0.5 mL MEM supplemented with 1% fetal bovine serum and 1 μmol/L AP23846 or an equal volume of DMSO for L3.6pl cells or 2% fetal bovine serum for endothelial cells. For L3.6pl cells, the lower chamber was filled with 0.75 mL MEM supplemented with 10% fetal bovine serum. For hepatic endothelial cells, the lower chamber was filled with 0.75 mL L3.6pl cell conditioned medium removed from cells that had been treated with or without the indicated concentration of AP23846. L3.6pl cells were treated for 24 hours with 0, 0.5, or 1.0 μmol/L AP23846. The medium was then replaced with complete medium lacking inhibitor. After an additional 24 hours, this medium was removed and used to examine effects of migration of endothelial cells. After 48 hours of incubation for L3.6pl cells or 72 hours of incubation for hepatic endothelial cells, the nonmigratory cells on the upper filter surface were removed with a cotton swab and cells that had migrated to the lower filter were fixed and stained with HEMA3 (Biochemical Sciences, Swedesboro, NJ) according to the manufacturer's instructions. The migratory cells were counted under a microscope at ×100 magnification. Cells were counted in five fields per insert in triplicate. Cell images were obtained using a Sony DXC-990 3 CCD color video camera (Sony Corp. of America, New York, NY). Cells were photographed at ×200 magnification.

ELISAs

Quantitative measurements of VEGF and IL-8 levels in cell culture supernatants were determined by ELISA. Cells were prepared for VEGF or IL-8 assays as described above. Cell culture supernatant (1 mL) was removed from each sample, centrifuged for 1 minute at 12,000 rpm to pellet any floating cells, and then transferred to a fresh microcentrifuge tube. Supernatants not assayed immediately were frozen at −80°C. The remaining cells were lysed and total protein was quantitated as described above. Cell culture supernatants were subjected to VEGF (Biosource, Camarillo, CA) or IL-8 (Quantikine Human IL-8 Immunoassay, R&D Systems, Minneapolis, MN) ELISA according to the protocols. Final VEGF and IL-8 concentrations (pg/mL) were determined using the TECAN Genios plate reader and Magellan version 4.0 software. VEGF and IL-8 concentrations (pg/mL) were normalized to the concentration of total protein obtained from the cells on the dish (mg/mL) and expressed as picogram VEGF per milligram protein.

Animals

Pathogen-free female C3H/HeN mice were purchased from the Animal Production Area of the National Cancer Institute-Frederick Cancer Research and Development Center (Frederick, MD). Mice were housed and maintained in pathogen-free conditions. These facilities have been approved by the American Association for Accreditation of Laboratory Animal Care and meet all regulations and standards of the U.S. Department of Agriculture, U.S. Department of Health and Human Services, and NIH. Mice used in these experiments were between 8 and 12 weeks old and were used in accordance with institutional guidelines.

Gel Foam Angiogenesis Assay

The development and optimization of the gel foam assay has been described previously (31). Briefly, gel foam sponges (Pharmacia & Upjohn, Peapack, NJ) were cut into ∼0.5 × 0.5–cm fragments and soaked in PBS overnight at 4°C. The saturated sponges were then placed on sterile filter paper to allow for removal of excess PBS. Sponges were then soaked in a 50:50 mixture of 1% Ultrapure agarose/PBS (Life Technologies) and conditioned medium from untreated L3.6pl cells, conditioned medium from L3.6pl cells treated with 1 μmol/L AP23846, serum-free medium alone, serum-free medium supplemented with recombinant VEGF (rVEGF) and recombinant IL-8 (rIL-8; 50:50 mixture), or serum-free medium supplemented with rVEGF and IL-8 (50:50 mixture) and AP23846 (1 μmol/L). The final working concentration of cytokines was 2 μg/mL. After hardening at room temperature for ∼1 hour, the sponges were implanted into mice s.c. as described previously (31). After 2 weeks, the gel foam sponges were harvested and frozen in OCT (Sakura Fineter, Torrance, CA). The frozen samples were later sectioned, probed for CD31 (PharMingen, San Diego, CA), and stained via immunoperoxidase techniques using 3,3′-diaminobenzidine and hematoxylin as chromagens.

Statistical Analyses

The significance of differences in cell proliferation, VEGF and IL-8 expression, and cell migration between treatment groups was determined using a Student's t test (two-tailed). The significance of differences in the gel foam assay was determined via the Mann-Whitney U test. P < 0.05 was deemed significant.

Results

Chemical Structure and Src Kinase Selectivity of AP23846

AP23846 is a novel small-molecule inhibitor of Src family kinase (Fig. 1), with ≤0.5 nmol/L IC₅₀ for Src kinase (relative to two methods, the fluorescence-based LANCE and [γ-³³P]ATP-based UDL kinase assays as described above). This represents ∼10-fold improvement over the most frequently used commercially available Src family kinase inhibitors, PP1 and PP2, which inhibit Src family kinases with IC₅₀s of ∼5 nmol/L (19). The structure-activity relationships of AP23846 to inhibit Src kinase are shown relative to two 2,6,9-trisubstituted purine analogues, AP23848 and AP23980, which vary only with respect to one functional group (R1 in the generic structure). Although not detailed in this report, AP23846 was determined to be superior to AP23848 and AP23980 in terms of metabolic stability and related pharmacokinetic activities. AP23846 has similar potency against c-Yes, Fyn, Lyn, Lck, and Hck and is significantly less potent against Fgr, epidermal growth factor receptor, PDGFR-α, PDGFR-β, Kit, KDR/VEGF receptor-2, insulin receptor kinase, and Flt1. Interestingly, AP23846 was comparably potent against Flt3 as against Src family kinases (Table 1). Relative to Abl, AP23846 exhibits lower potency and increased selectivity for Src compared with recently reported dual Src/Abl kinase inhibitors, such as BMS-354825 and AP23464 (25, 27).

Inhibition of c-Src Autophosphorylation with AP23846

To determine the effectiveness of AP23846 in inhibiting c-Src in human cancer cells, the L3.6pl human pancreatic cancer cell line and HT29 colon cancer cells, chosen because both express high levels of activated Src and because Src inhibition has been shown to affect tumorigenic properties in vitro and in vivo (24, 32), were treated with 0, 0.25, 0.5, or 1.0 μmol/L concentrations of AP23846. c-Src activity was assessed via immunoprecipitation of total c-Src followed by Western blot analysis against phosphorylated Tyr⁴¹⁸, the autophosphorylation site of c-Src and a hallmark of c-Src activation as described in Materials and Methods. As seen in Fig. 2 , AP23846 inhibited c-Src autophosphorylation in L3.6pl and HT29 cells with IC₅₀s of 375 and 125 nmol/L, respectively (Fig. 2A and B). Maximum inhibition of c-Src autophosphorylation was detected at 1 μmol/L AP23846 (Fig. 2A). Blots were stripped and reprobed for total c-Src. Although some decreases in Src expression were observed, suggesting that the inhibitor might have a secondary effect on Src expression, the primary change was in Src phosphorylation. We next sought to determine the duration of this inhibitory response. Cells were incubated with 1 μmol/L AP23846 for 1, 2, 4, 6, 8, 12, 24, and 48 hours. c-Src autophosphorylation was assayed via anti-phospho-Src^Y418 Western blot analysis at each of these time points. Inhibition of c-Src autophosphorylation was detected after 1-hour incubation with 1 μmol/L AP23846 and was maintained throughout the 48-hour time course (Fig. 2C). Thus, AP23846 induces a potent and stable inhibition of c-Src kinase activity. As an additional determinant of the effectiveness of Src family kinase inhibition by AP23846, the phosphorylation of paxillin, a known focal adhesion–associated Src substrate, was assessed by phosphospecific Western blot analysis. As seen in Fig. 2, treatment of L3.6pl cells for 24 hours with 1 μmol/L AP23846 resulted in a complete loss of tyrosine phosphorylation of paxillin at Tyr¹¹⁸ (Fig. 2D). Similar results were obtained with p130 CAS, an additional focal adhesion–associated Src substrate (data not shown). Further, similar results were observed in cells grown in serum, suggesting that the presence of serum does not affect inhibition of Src activity (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Inhibition of c-Src autophosphorylation via AP23846. L3.6pl cells (A) or HT29 cells (B) were plated as described in Materials and Methods. Twenty-four hours after plating, medium was replaced with serum-free medium supplemented with the indicated concentrations of AP23846 in a volume of 5 μL/5 mL medium (or an equal volume of DMSO for 0 nmol/L concentrations) for an additional 24 h. Cells were lysed, and lysates were prepared for immunoprecipitation as described in Materials and Methods. Total c-Src was immunoprecipitated from cell lysates, and Western blot analysis was done using an anti-phospho-Src^Y418 antibody. Blots were stripped and reprobed for total c-Src. C, L3.6pl cells were treated as described above, with the exception that only one concentration of AP23846 was used (1 μmol/L) over a 48-h time course. Cells were lysed after incubation with AP23846 for 1, 2, 4, 8, 12, 24, and 48 h. Cell lysates were immunoprecipitated for total c-Src, and Western blot analysis was done using an anti-phospho-Src^Y418 antibody. Blots were stripped and reprobed for total c-Src. D, L3.6pl cells were treated for 24 h with 1 μmol/L AP23846 or an equal volume of DMSO. Whole-cell lysates (50 μg) were resolved via SDS-PAGE, and Western blot analysis was done using an anti-phospho-paxillin^Y118 polyclonal antibody (1:1,000). Blots were stripped and reprobed with an anti-total paxillin polyclonal antibody (1:1,000).

L3.6pl Cell Proliferation/Cytotoxicity

We next investigated potential cytotoxic effects of AP23846. L3.6pl cells were photographed after 1 and 24 hours of incubation with or without 1 μmol/L AP23846. As seen in Fig. 3 , cells treated with AP23846 displayed a more rounded phenotype after 1-hour incubation in the presence of the drug (Fig. 3A and B); however, this effect was no longer evident after 24 h of incubation (Fig. 3C and D). These data suggest that Src family kinases have short-term effects on cell adhesion, consistent with the recent work of Jones et al. (33). The effect of AP23846 on cell proliferation was assessed by counting viable cells daily in control and treated cells (1 μmol/L AP23846) described in Materials and Methods. As seen in Fig. 4A , at this concentration, AP23846 inhibited cell growth, consistent with previous studies suggesting that Src family kinases are important in proliferation (Fig. 4A). To examine whether these changes were due to decreased cell growth and/or apoptosis, L3.6pl cells treated with or without 1.0 μmol/L AP23846 were stained with propidium iodide and subjected to flow cytometric analysis to determine the percentage of cells at sub-G₀-G₁. As seen in Fig. 4B, no major difference was observed in the percent sub-G₀-G₁ cells between treated (G₁, 63%; G₂, 6%; S, 25%; sub-G₀-G₁, 6%) and untreated (G₁, 66%; G₂, 6%; S, 24%; sub-G₀-G₁, 4%) groups (Fig. 4B). As Src family kinases are involved in dynamic regulation of the actin cytoskeleton and cellular processes dependent on cytoskeletal remodeling, such as cell migration, the effect of AP23846 on serum-induced L3.6pl cell migration was determined. Cell migration was quantitated using the modified Boyden chamber assay as described in Materials and Methods. As seen in Fig. 4, L3.6pl cell migration was virtually completely inhibited in the presence of 1 μmol/L AP23846 (Fig. 4C). Thus, AP23846 decreases L3.6pl cell growth and migration without inducing demonstrable cytotoxic or apoptotic effects.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Morphology of L3.6pl cells treated with AP23846. L3.6pl cells were plated as described in Materials and Methods. Cells were serum starved overnight, and the following day, cell culture medium was replaced with medium containing 1 μmol/L AP23846 (B and D) or an equal volume of DMSO (A and C). The cells were photographed after 1 h (A and B) and 24 h (C and D) at ×100 magnification using a Nikon Coolpix 4500 digital camera.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Effects of AP23846 on cell proliferation, cell cycle, and migration. A, cell proliferation analysis. Cells were plated as described in Materials and Methods. Twenty-four hours after plating, cell medium was changed to MEM containing 1.0 μmol/L AP23846 or an equal volume of DMSO. Viable cells were determined via trypan blue exclusion and counted at 0, 24, and 48 h. B, cell cycle analysis. L3.6pl cells were treated with 1 μmol/L AP23846 or an equal volume of DMSO for 24 h. Cells were then trypsinized, fixed with 70% ethanol, stained with propidium iodide, and subjected to flow cytometric analysis as described in Materials and Methods. C, cell migration. Cell migration was determined via the modified Boyden chamber assay as described in Materials and Methods. L3.6pl cells (2 × 10⁵) were seeded in the top chamber in the presence or absence of 1.0 μmol/L AP23846. Cells were allowed to migrate for 48 h, at which point migratory cells on the bottom half of the insert membrane were stained with HEMA3 and counted under ×100 magnification. Migratory cells were graphed as the number of migrated cells per high-power field (HPF). Five fields were counted in triplicate from each insert. Cell images were obtained using a Sony DXC-990 3 CCD color video camera. Cells were photographed at ×200 magnification (n = 3). *, P < 0.005.

Effects of AP23846 on VEGF and IL-8 Expression

Previous studies from several laboratories implicated Src activity as critical to VEGF and IL-8 expression (34, 35). Therefore, we examined the effects of AP23846 on production of the proangiogenic molecules VEGF and IL-8. L3.6pl cells were incubated for 24 hours in the presence of 0, 0.25, or 0.5 μmol/L AP23846. At the end of this period, cell lysates and supernatants were harvested. Cell culture supernatants were assayed for VEGF and IL-8 by ELISA as described in Materials and Methods. Lysates were used to normalize IL-8 and VEGF levels to total protein. As seen in Fig. 5 , Src family kinase inhibition via AP23846 induced a dose-dependent decrease in both IL-8 and VEGF expression (Fig. 5A). VEGF expression levels were reduced to 77% and 47% of control values in the presence of 0.25 and 0.5 μmol/L AP23846, respectively, whereas IL-8 levels were reduced to 55% and 24% of controls (Fig. 5A).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Effects of AP23846 on VEGF and IL-8 production. A, effects of increasing concentrations of AP23846 on VEGF and IL-8 expression in L3.6pl cells. Cells were plated and maintained as described in Materials and Methods. Twenty-four hours after plating, cells were serum starved overnight, after which medium was replaced with serum-free MEM containing 0, 0.25, or 0.5 μmol/L AP23846. After an additional 24 h, cell culture supernatants and cell lysates were harvested. VEGF and IL-8 levels were quantitated by ELISA and normalized to the amount of total cellular protein. VEGF and IL-8 levels were graphed relative to DMSO controls. B, comparison of the effects of AP23846 and PP2 on IL-8 expression. Cells were treated as described in A, with the exception that cells were inhibited with 0, 0.25, 0.5, or 1.0 μmol/L AP23846 or 0, 2.5, 5.0, or 10.0 μmol/L PP2. IL-8 levels were quantitated and expressed as described in A. C, effects of AP23846 on VEGF and IL-8 expression in PC3 and HT29 cells. PC3 and HT29 cells were plated and maintained as described in Materials and Methods. Cells were treated as in A, with the exception that only a single concentration of AP23846 (1 μmol/L) was used. VEGF and IL-8 levels were quantitated and graphed as described in A. Significance of differences in VEGF and IL-8 expression were determined by t test (n = 3). *, P < 0.005.

To verify that these results were specific for Src inhibition, plasmids encoding c-Src-targeted siRNA were transfected into L3.6pl cells. Stable L3.6pl clones expressing c-Src siRNA were obtained in which c-Src expression levels were reduced by 80% (36). As reported in Table 2 , relative to AP23846-mediated Src family kinase inhibition, a similar reduction in VEGF expression and an even greater reduction in IL-8 expression were detected in L3.6pl cell lines stably expressing c-Src-targeted siRNA (Table 2). These results suggest that Src is a critical target of AP23846 in affecting expression of IL-8, although potential off-target effects of the inhibitor cannot be excluded.

To underscore the increased potency of AP23846 relative to one of the most widely used commercially available Src family kinase inhibitors, PP2, IL-8 protein expression from cell culture supernatants was assayed in the presence of increasing concentrations of either AP23846 or PP2. A dose-dependent decrease in IL-8 expression was observed with increasing concentrations of both AP23846 and PP2 (Fig. 5B). However, AP23846 was more effective at decreasing IL-8 expression than PP2. AP23846 reduced IL-8 expression levels to ∼25% of control levels at a concentration of 0.5 μmol/L, whereas IL-8 expression remained over 50% of control levels in the presence of 5.0 μmol/L PP2 (Fig. 5B). Comparable results were obtained at the highest concentrations used of either drug, as 1.0 μmol/L AP23846 reduced IL-8 expression to 14% of control levels, whereas PP2 reduced IL-8 expression to 34% of control levels at a concentration of 10 μmol/L (Fig. 5B). VEGF expression was similarly inhibited, with comparable differences observed between AP23846 and PP2 (data not shown).

To determine if these results were more broadly applicable to human tumor cells and not specific for L3.6pl pancreatic cancer cells, the effect of AP23846 on VEGF and IL-8 expression was assessed in HT29 colon and PC3 prostate cancer cells. Fig. 5 shows that, in PC3 cells, Src family kinase inhibition via AP23846 resulted in >80% and 60% decreases in VEGF and IL-8 expression, respectively, relative to untreated controls (Fig. 5C). In HT29 cells, AP23846 caused a 40% reduction in VEGF expression and a >90% reduction in IL-8 expression (Fig. 5C). These results indicate that AP23846 is capable of reducing expression of the proangiogenic factors VEGF and IL-8 in multiple human tumor cell lines and further suggest a potential antiangiogenic application of drugs of this class for human solid tumors.

Inhibition of Endothelial Cell Migration

As IL-8 and VEGF are both known to induce migration of endothelial cells (37, 38), we next sought to determine the effects of Src family kinase inhibition via AP23846 on the ability of cell culture supernatants from L3.6pl cells to induce endothelial cell migration in vitro. L3.6pl pancreatic cancer cells were incubated in serum-free medium for 24 hours in the presence of 0.0, 0.5, or 1.0 μmol/L AP23846, at which point the medium was replaced with fresh serum-free medium, and the cells were incubated for an additional 24 hours. Cell culture medium from each of these treatments was evaluated for their ability to stimulate migration of hepatic endothelial cells in a modified Boyden chamber assay. Migrating cells were stained and counted, and the results were graphed as number of endothelial cells migrated per high-power field. Conditioned medium from L3.6pl cells pretreated with AP23846 failed to support endothelial cell migration at levels that approached those induced by conditioned medium from untreated cells (Fig. 6 ). Pretreatment of L3.6pl cells with 0.5 and 1.0 μmol/L AP23846 caused 16- and 50-fold reductions in the ability of these cell culture supernatants to induce endothelial cell migration, respectively (Fig. 6). Again, by way of contrast, conditioned medium from L3.6pl cells pretreated with 10 μmol/L PP2 were only 1.4-fold reduced in their ability to induce endothelial cell migration (data not shown). These results indicate that Src family kinase inhibition by AP23846 induces a functionally relevant decrease in the production of endothelial cell migratory factors by human pancreatic cancer cells.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

Effects of AP23846 on endothelial cell migration in vitro. Endothelial cell migration assays were carried out as described in Materials and Methods. Briefly, murine hepatic endothelial cells were seeded in the upper well of a Boyden assay migration chamber. The bottom well was filled with conditioned medium removed from L3.6pl cells that had been treated for 24 h with 0, 0.5, or 1.0 μmol/L AP23846. Endothelial cells were allowed to migrate for 72 h, after which migratory cells were fixed and stained with HEMA3. Migratory cells were viewed and counted under ×100 magnification (Nikon Microphot FX). Migratory cells were graphed as the number of migrated cells per high-power field. Five fields were counted in triplicate from each insert. Cell images were obtained using a Sony DXC-990 3 CCD color video camera. Cells were photographed at ×200 magnification (n = 3). *, P < 0.00005.

Inhibition of Gel Foam Angiogenesis In vivo

Finally, having established that AP23846-mediated Src family kinase inhibition reduces production of the proangiogenic factors VEGF and IL-8 in vitro and that cell culture supernatants from AP23846-treated cells fail to support endothelial cell migration in vitro, we next sought to determine if the reduction in angiogenic factor production caused by AP23846 would affect the ability of conditioned medium from L3.6pl cells to support angiogenesis in vivo. To evaluate the ability of conditioned medium from untreated or AP23846-treated L3.6pl cells to support an angiogenic response in vivo, a gel foam assay was used as described in Materials and Methods. Briefly, fragments of surgical sponges were saturated with a solution containing a 50:50 mixture of 2% agarose and either L3.6pl cell conditioned medium, conditioned medium from L3.6pl cells treated with 1 μmol/L AP23846, serum-free medium, serum-free medium supplemented with rVEGF and IL-8, or serum-free medium supplemented with rVEGF, rIL-8, and exogenous AP23846. The saturated sponges were implanted s.c. into the flanks of mice and left unmolested for 2 weeks. After the 2-week incubation, the sponges were removed, sectioned, and stained for CD31 to visualize endothelial cell infiltration. Fig. 7 shows representative fields photographed from CD31-stained sections of gel foams removed from mice after 2 weeks. Robust endothelial cell infiltration (as determined by the presence of brown-stained cells) can clearly be observed in gel foams saturated with L3.6pl conditioned medium or serum-free medium supplemented with rVEGF and rIL-8, whereas gel foams saturated with serum-free medium alone were essentially devoid of infiltrating endothelial cells. Although conditioned medium from L3.6pl cells treated with 1 μmol/L AP23846 did support some endothelial cell infiltration into gel foam sponges, endothelial cell counts were significantly lower (P < 0.0001) than those obtained with conditioned medium from untreated cells (Fig. 7; Table 3 ). These results could not be explained by residual AP23846 in the medium directly inhibiting endothelial cell migration, as exogenous AP23846 did not affect the ability of serum-free medium supplemented with rVEGF and rIL-8 to support angiogenesis into gel foams (Fig. 7; Table 3). Interestingly, conditioned medium from L3.6pl cells was much more efficient at promoting gel foam angiogenesis than rVEGF and rIL-8 alone. These results suggest that, in addition to IL-8 and VEGF, AP23846 may be inhibiting other factors affecting endothelial cell migration. Overall, these results show that pretreatment of L3.6pl pancreatic cancer cells with the Src family kinase inhibitor AP23846 results in a biologically relevant decrease in the production of proangiogenic molecules by these tumor cells.

• Download figure
• Open in new tab
• Download powerpoint

Figure 7.

Effects of Src inhibition on gel foam angiogenesis. L3.6pl cells were serum starved in the presence of either AP23846 (1.0 μmol/L) or an equal volume of DMSO for 24 h. Following this incubation, cell culture supernatants were harvested. Conditioned medium from AP23846 or DMSO-treated cells, serum-free medium, serum-free medium supplemented with a 50:50 mixture of rVEGF and IL-8 (2 μg/mL total), or serum-free medium supplemented with rVEGF, rIL-8 (50:50 mixture, 2 μg/mL total), and AP23846 (1.0 μmol/L) were used to saturate gel foam sponges. The sponges were implanted into C3H/HeN mice s.c. as described in Materials and Methods. Gel foam sponges were harvested after 2 wks, sectioned, stained for CD31, and visualized by 3,3′-diaminobenzidine and hematoxylin. Stained sections were photographed at ×100 magnification. White arrows, CD31⁺ vessel infiltration.

Discussion

Many small-molecule inhibitors of cellular tyrosine kinases have been developed over the past decade, targeting both receptor and nonreceptor tyrosine kinases (21, 39). Such inhibitors have been developed both as research tools, used to identify functions specific to their target kinases, or as potential therapeutic agents (17, 21, 39). For Src family kinases, the issues of potency and selectivity have been especially challenging for advancing both effective research tools and therapeutic agents (21, 40). Toward the development of more effective Src family kinase inhibitors, it is noteworthy to highlight the pyrazolopyrimidines PP1 and PP2, which have been used extensively in vitro and, in some cases, in vivo, to investigate Src family kinase–dependent cellular signal transduction pathways. PP1 and PP2 inhibit Src family kinases with IC₅₀s in the 5 nmol/L range in vitro, typically requiring a higher concentration to inhibit Src family kinases in cell culture (19). Although these inhibitors are more selective than the previous generation of inhibitors used to target Src family kinases, including herbimycin A and genistein, they still inhibit off-target kinases with sufficient potency to make interpretation of results as being due exclusively to Src family kinase inhibition difficult (19, 20). Thus, although PP2 is still used for in vivo preclinical studies (18), it is rarely used in vitro without supporting data from dominant-negative Src family kinases or siRNA (18, 41) and is not being evaluated clinically. Fortunately, a new generation of Src family kinase inhibitors, greatly improved in selectivity and potency over PP1 and PP2, has recently advanced into various stages of preclinical research or clinical development by multiple pharmaceutical companies (42). These compounds inhibit Src family kinases with IC₅₀s in the low to subnanomolar range and display increased selectivity for Src family kinases (22–26). The studies presented here are focused on AP23846, a new purine-based Src family kinase inhibitor lead compound with an IC₅₀ of ∼0.5 nmol/L against purified c-Src kinase. Not surprisingly, higher concentrations of the inhibitor were required to achieve maximum inhibition of c-Src autophosphorylation in cultured cells; however, AP23846 showed greater potency in achieving inhibition of c-Src activity and downstream functions than did PP2 at a 10-fold higher concentration. AP23846 also seems to be highly selective for Src, as it inhibits the structurally similar Abl tyrosine kinase with a 40-fold higher IC₅₀. AP23846 was shown to inhibit Flt3 with high potency in vitro, but Flt3 is not expressed in L3.6pl cells; thus, Flt3 inhibition is unlikely a factor in the results reported in these studies.³ Importantly, for cell culture studies, AP23846-mediated inhibition of c-Src kinase activity persisted over a 48-hour time course. Unfortunately, therapeutic levels of AP23846 have not been achieved in rodents, precluding in vivo studies (data not shown). However, the selectivity of AP23846 allows assessment of biological effects mediated by Src inhibitors in vitro that are relevant to future in vivo studies.

Recent studies using molecular approaches and less selective Src inhibitors suggest that Src family kinases regulate expression of proangiogenic molecules. Initiation of the angiogenic switch and formation of a tumor vasculature are essential for the progression of most solid tumors (43). Tumor angiogenesis has thus gained considerable interest as a potential target for therapeutic intervention. Many of these studies have focused on inhibition of the VEGF receptor or its ligand VEGF (44–46). Unfortunately, the VEGF/VEGF receptor axis, while an important mediator of tumor angiogenesis, is only one of several angiogenic factor/receptor pathways through which angiogenesis may be initiated (47). Recent evidence, including the data presented here, suggests that Src may be a mediator of expression of multiple proangiogenic molecules (35, 48). Similar results were obtained by our laboratory using c-Src-specific siRNA, providing further evidence that these effects are Src specific. Additionally, these results were not pancreatic cancer cell specific, as AP23846-mediated Src family kinase inhibition caused similar reductions in VEGF and IL-8 expression in PC3 prostate cancer cells and HT29 colon cancer cells. Although these studies focused exclusively on the ability of a Src family kinase inhibitor, acting directly on tumor cells, to decrease production of proangiogenic molecules, the antiangiogenic effect of Src family kinase inhibitors in vivo is likely to be even more pronounced. In addition to blocking production of proangiogenic factors by tumor cells, Src family kinase inhibitors in vivo have been shown to disrupt Src family kinase–mediated functions in host endothelial cells, including the ability of tumor cells to extravasate (49, 50). Thus, inhibition of Src family kinase activity by a highly potent and selective small-molecule inhibitor may emerge as an effective antiangiogenic strategy for human solid tumors. In this regard, AP23846 is a promising novel Src inhibitor and provides a proof-of-concept preclinical lead compound to optimize for such antiangiogenic activities as well as anti-invasive and antimetastatic activities that are also emerging therapeutic strategies for small-molecule inhibitors of Src family kinases.