Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Cholecystokinin (CCK) is biologically active in the central nervous system and periphery (Mutt and Jorpes, 1971). In the central nervous system, CCK alters neuronal activity, alters behavior, and causes satiety (Gibbs et al., 1973; Crawley 1988; Wang et al., 1988). In the periphery, CCK is localized to intrinsic neurons of the brain, myenteric plexus, and endocrine cells (Buchan et al., 1978; Beinfeld et al., 1989). When secreted, CCK has potent effects on the gallbladder, pancreas, and stomach (Jensen et al., 1980; Steigerwalt et al., 1984; Moran et al., 1985). CCK is derived from a 115 amino acid precursor protein to 58, 39, 33, and 8 amino acid forms, which have a sulfated tyrosine (Dockray, 1977; Takahashi et al., 1985; Chang and Lotti, 1986). CCK₁ receptors, which are present in the pancreas, certain brain regions, and gallbladder, bind sulfated but not nonsulfated forms of CCK as well as the antagonist L-364,718 with high affinity (Makovec et al., 1985; Chang et al., 1989; Freidinger, 1989; Jensen et al., 1989). CCK₂ receptors, which are widely distributed in the central nervous system, bind gastrin, L-365,260, as well as sulfated and nonsulfated CCK with high affinity (Innes and Snyder, 1980; Saito et al., 1980; Lotti and Chang, 1984).

The CCK₂ receptor is a 390 amino acid protein that is a member of the heptahelical superfamily (Wank, 1995). It interacts with G proteins such as Gq causing phosphatidylinositol turnover and G_12,13 causing focal adhesion kinase tyrosine phosphorylation (Sethi and Rozengurt, 1991). Focal adhesion kinase interacts with adaptor proteins such as paxillin causing cytoskeletal reorganization, whereas phosphatidylinositol is metabolized to inositol-1,4,5-trisphosphate and diacylglycerol. The inositol-1,4,5-trisphosphate causes release of Ca²⁺ from intracellular organelles such as the endoplasmic reticulum, whereas the diacylglycerol activates protein kinase C. Protein kinase C phosphorylates protein substrates such as mitogen-activated protein kinase kinase, which phosphorylates mitogen-activated protein kinase (Whitmarsh and Davies, 1996). Phosphorylated mitogen-activated protein kinase enters the nucleus, leading to increased expression of the nuclear oncogene c-fos. After translation of the mRNA, c-fos protein forms heterodimers with c-jun binding to activating protein-1 (AP-1) sites in growth factor genes.

Previously, we found that CCK-8 bound with high affinity to SCLC cells and caused elevation of cytosolic Ca²⁺ (Yoder and Moody, 1987; Staley et al., 1990). CCK-8 or gastrin increased cytosolic Ca²⁺ and stimulated SCLC growth (Staley et al., 1990). The increase in cytosolic Ca²⁺ caused by CCK-8 or gastrin was blocked by L-365,260 (Staley et al., 1990). In the present study we report that CCK-8 increases mitogen-activated protein kinase and focal adhesion kinase tyrosine phosphorylation, and the effects of CCK-8 were antagonized by the CCK₂ receptor antagonist ([R-(R*,R*)]-4[[2-[[3-(1H-indole-3-yl)-2-methyl-1-oxo-2-[[tricyclo[3.3.1.1^3,7]dec-2-yloxy)carbonyl[amino]propyl]amino]-1-phenylethyl]amino]-4-oxobutanoic acid) (CI-988; PD134308) (Hughes et al., 1990). Also, CCK-8 increased c-fos and VEGF mRNA and the increase caused by CCK-8 was inhibited by CI-988. CCK-8 increased SCLC proliferation and the effects of CCK-8 were inhibited by CI-988. Also, CI-988 inhibited basal SCLC growth in vitro and in vivo. These results suggest that biologically active CCK₂ receptors that are present in SCLC cells are coupled to multiple second messenger cascades whose activation results in proliferation.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture. NCI-H82, H209, and H345 cells were cultured in RPMI-1640 containing 10% heat-inactivated fetal bovine serum (FBS; Invitrogen, Carlsbad, CA) (Carney et al., 1985). The cells, which were nonadherent, were split weekly by 1:1 dilution. The cells were mycoplasma free and were used when they were in exponential growth phase after incubation at 37°C in 5% CO₂, 95% air.

Receptor Binding. The ability of a CCK₁ receptor antagonist, L-364,718, and two CCK₂ receptor antagonists, L-365,260 and CI-988, to inhibit specific ¹²⁵I-CCK-8 binding to SCLC cells was investigated (Yoder and Moody, 1987). NCI-H209 or NCI-H345 cells (10⁶) were incubated with ¹²⁵I-CCK-8 (2200 Ci/mmol; PerkinElmer Life Sciences, Boston, MA) in 100 µl of SIT medium (RPMI-1640 containing 3 × 10⁸ M sodium selenite, 5 µg/ml bovine insulin, and 10 µg/ml transferrin) with 0.25% bovine serum albumin and 250 µg/ml bacitracin. After incubation at 25°C for 60 min, free ¹²⁵I-CCK-8 was removed and the cells that contained bound ¹²⁵I-CCK-8 were counted in an LKB gamma counter.

Cytosolic Ca²⁺. Previously, we found that CCK caused elevation of cytosolic Ca²⁺ in NCI-H209 cells, which was antagonized by L-365,260 (Staley et al., 1990). NCI-H209 cells were harvested (2.5 × 10⁶/ml) and incubated with 5 µM Fura-2 AM at 37°C for 30 min in SIT medium. The cells, which contained loaded Fura-2 (Calbiochem, La Jolla, CA), were centrifuged at 1500g for 10 min and resuspended at the same concentration in new SIT medium containing 20 mM HEPES · NaOH, pH 7.0. The fluorescence intensity was continuously monitored using a PerkinElmer LS2 spectrofluorometer equipped with a magnetic stirring mechanism and temperature (37°C)-regulated cuvette holder before and after the addition of CCK-8 in the presence or absence of CI-988, and the Ca²⁺ concentrations were calculated as described (Staley et al., 1990).

Western Blot. NCI-H209 cells were put in SIT containing 0.5% FBS overnight. Three hours before treatment cells were preincubated in ST media (RPM1-1640 containing 3 × 10⁸ M Se₂O₃ and 10 ng/ml transferrin). For the focal adhesion kinase and paxillin assays, cells were treated with CCK-8 in the presence or absence of CI-988 for 1 min, washed with saline, lysed with 1 ml of buffer containing 50 mM Tris/HCl, pH 7.5; 150 mM NaCl; 1% (w/v) Triton X-100; 1% (w/v) deoxycholate; 1% (w/v) NaN₃; 1 mM EGTA; 0.4 mM EDTA; 2.5 µg/ml aprotinin; 2.5 µg/ml leupeptin; 1 mM phenylmethylsulfonyl fluoride; and 0.2 mM Na₃VO₄ (Sigma Chemical, St. Louis, MO), sonicated for 5 s at 4°C, and centrifuged at 10,000g for 15 min. Protein concentration was measured by Bio-Rad protein assay reagent, and the volume was adjusted such that the cell lysates contained the same amount of protein (150 µg/ml). The lysates were incubated with 4 µg of anti-phosphotyrosine monoclonal antibody (PY20), 4 µg of goat anti-mouse IgG, and 30 µl of protein A-agarose overnight at 4°C.

Northern Blot. The ability of CCK to stimulate nuclear oncogene expression was investigated (Weber et al., 2001). For the c-fos experiments, NCI-H209 cells were cultured with SIT medium containing 0.5% fetal bovine serum. After 1 h, the cells were treated with 10 nM CCK-8 in the presence or absence of CI-988 for 60 min. Total RNA was isolated using guanidinium isothiocyanate (Fluka, Buchs, Switzerland). Ten micrograms of denatured RNA was separated in a 0.66 M formaldehyde 1% agarose gel. The gel was treated with ethidium bromide to assess RNA integrity. The RNA was blotted onto a nytran membrane overnight and the membrane hybridized with DNA probes labeled with [³²P]dCTP by using an Invitrogen random priming kit. The membrane was exposed to Kodak XAR-2 film at 80°C for 1 day and the autoradiogram developed. The autoradiograms were analyzed using a Molecular Dynamics (Sunnyvale, CA) densitometer. For the VEGF experiments, the reaction was stopped 8 h after addition of CCK-8 (Casibang et al., 2001).

Proliferation Assays. The ability of CI-988 to alter the growth of SCLC cells was investigated in vitro and in vivo by using (4,5 dimethylthiazol-2-yl)-2.5-diphenyl-2H-tetrazolium bromide] (MTT) colorimetric assays. SCLC cells (10⁴/well) were placed in SIT medium and varying concentrations of CI-988 added. After 4 days, MTT (Sigma Chemical) was added. After 4 h, 150 µl of dimethyl sulfoxide was added. After 16 h, the optical density at 540 nm was determined using an enzyme-linked immunosorbent assay reader. Also, the effects of CI-988 were investigated using a clonogenic assay (Mahmoud et al., 1991). The base layer consisted of 3 ml of 0.5% agarose (FMC Bioproducts, Rockland, ME) in SIT medium containing 5% fetal bovine serum in six-well plates. The top layer consisted of 3 ml of SIT medium in 0.3% agarose, CI-988, and 5 × 10⁴ lung cancer cells. Triplicate wells were plated and after 2 weeks, 1 ml of 0.1% p-iodonitrotetrazolium violet was added and after 16 h at 37°C, the plates were screened for colony formation. The number of colonies larger than 50 µm in diameter were counted using an Omnicon (Fairfax, VA) image analysis system.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

CI-988 Binds with High Affinity to SCLC Cells. The binding of CCK receptor antagonists was investigated. Figure 1 shows that 0.1 nM CI-988 had little effect on ¹²⁵I-CCK-8 binding to NCI-H209 cells, whereas 100 nM CI-988 inhibited almost all specific ¹²⁵I-CCK-8 binding. The IC₅₀ was 5 nM for CI-988 and 2 nM for CCK-8. Table 1 shows that L-365,260, CI-988, and L-364,718 inhibited specific ¹²⁵I-CCK-8 binding to NCI-H209 cells with IC₅₀ values of 2, 5, and 200 nM. Similar data were obtained using NCI-H345 cells. These results indicate that CCK-8, L-365,260, and CI-988 bind with high affinity to SCLC cells.

View larger version (10K): [in this window] [in a new window]   Fig. 1.   CCK-8 binding. The ability of CI-998 () and CCK-8 () to inhibit specific binding of 125I-CCK-8 to NCI-H209 cells was determined. The mean value ± S.D. of three determinations each repeated in quadruplicate is indicated.

CI-988 Is an SCLC CCK₂ Receptor Antagonist. The ability of CI-988 to antagonize the actions of CCK was investigated. In a cytosolic Ca²⁺ assay, 10 nM CCK-8 increased the cytosolic Ca²⁺ in Fura-2 AM-loaded NCI-H209 cells from 150 to 180 nM (Fig. 2) within 15 s. The response to CCK-8 slowly declined and returned to baseline after 2 min. CI-988 (10 nM) and 100 nM CI-988 weakly and strongly antagonized the increase in cytosolic Ca²⁺ caused by CCK-8. These results indicate that CI-988 antagonizes the actions of CCK-8 in a concentration-dependent manner. Similar results were obtained using NCI-H345 cells (data not shown).

View larger version (9K): [in this window] [in a new window]   Fig. 2.   Cytosolic Ca2+. The ability of 10 nM CCK-8 to cause increased cytosolic Ca2+ was determined in the absence or presence of 10 or 100 nM CI-988 by using Fura-2 AM-loaded NCI-H209 cells. This experiment is representative of three others.

View larger version (32K): [in this window] [in a new window]   Fig. 3.   Mitogen-activated protein kinase and CI-988. The ability of 10 nM CCK-8 to cause tyrosine phosphorylation of mitogen-activated protein kinase was determined after 2 min in the absence or presence of 10 or 100 nM CI-988. This experiment is representative of four others.

View larger version (29K): [in this window] [in a new window]   Fig. 4.   Focal adhesion kinase (FAK) and CI-988. The ability of 100 nM CCK-8 to cause tyrosine phosphorylation of FAK and paxillin was determined after 1 min in the absence or presence of 1000 nM CI-988. This experiment is representative of four others.

View larger version (35K): [in this window] [in a new window]   Fig. 5.   C-Fos mRNA and CI-988. Top, ability of 10 nM CCK-8 to cause elevated c-fos mRNA was determined after 60 min in the absence or presence of 100 nM CI-988. Bottom, equal amounts of actin mRNA were loaded onto the gel based on Northern analysis. This experiment is representative of three others.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   VEGF mRNA. Top, ability of 10 nM CCK-8 to cause elevated VEGF mRNA was determined after 8 h in the absence or presence of 100 nM CI-988. Bottom, equal amounts of mRNA were loaded onto the gel based on ethidium bromide staining of the 28S and 18S rRNA bands. This experiment is representative of three others.

CI-988 Inhibits Proliferation of SCLC. Figure 7 shows that 3000 nM but not 100 or 300 nM CI-988 significantly inhibited the proliferation of NCI-H82, H209, or H345 cells in vitro using the MTT assay. Also, 1000 nM CI-988 significantly inhibited the proliferation of NCI-H82 cells. Table 2 shows that 1000 nM CI-988 or L-365,260 but not L-364,718 significantly inhibited NCI-H209 proliferation in the clonogenic assay. In contrast, 10 nM CCK-8 increased colony number 2-fold. CI-988 (1000 nM) reversed the increase in colony number caused by CCK-8.

View larger version (10K): [in this window] [in a new window]   Fig. 7.   MTT assay. Ability of CI-988 as a function of concentration to inhibit NCI-H82 (), H345 (), and H209 () proliferation was determined after 3 days. The mean value ± S.D. of eight determinations is indicated; p < 0.05 using Student's t test. *, experiment is representative of two others.

View larger version (10K): [in this window] [in a new window]   Fig. 8.   Xenograft proliferation. The ability of CI-988 to inhibit NCI-H209 xenograft proliferation in nude mice was determined in the absence () or presence () of CI-988 (10 µg by gavage daily in polyethylene glycol-400). The mean value ± S.D. of four determinations is indicated; p < 0.01 using Student's t test. **, experiment is representative of two others.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

SCLC cells have CCK₂ receptors. CCK-8 and gastrin inhibit ¹²⁵I-CCK-8 binding to NCI-H209 cells with high affinity (IC₅₀ = 2 and 6 nM, respectively; Yoder and Moody, 1987). CCK-8 elevates cytosolic Ca²⁺ and the increase caused by CCK-8 was inhibited by L-365,260 but not L-364,718 (Staley et al., 1990). Here the effects of the potent CCK₂ receptor antagonist CI-988 were investigated on SCLC cells.

CI-988 inhibited specific ¹²⁵I-CCK-8 binding to NCI-H209 cells with high affinity (IC₅₀ = 5 nM). Similarly, L-365,260 (IC₅₀ = 2 nM) but not L-364,718 (IC₅₀ = 200 nM) inhibited ¹²⁵I-CCK-8 binding to NCI-H209 cells with high affinity. Using Chinese hamster ovary cells transfected with CCK₂ receptors, L-365,260 and CI-988 had similar potency to inhibit the ability of CCK to elevate cytosolic Ca²⁺ (Kopin et al., 2000). These results suggest that CCK₂ receptors are present on NCI-H209 cells. Surprisingly, L-365,260 binds with 20-fold lower affinity to canine CCK₂ than human CCK₂ receptors. Subsequent studies showed that if canine Leu³⁵⁵ was changed to Val, L-365,260 bound with high affinity to the CCK₂ receptors (Dunlop et al., 1996). These results indicate that species polymorphisms can occur in CCK₂ receptors, which affect receptor antagonist binding. In contrast, agonists such as gastrin and CCK-8 bind with high affinity to both human and canine CCK₂ receptors. These results suggest that CCK agonists and antagonists bind to different sites on CCK₂ receptors.

After addition of CCK to NCI-H209 cells Gq may be activated, resulting in phosphatidylinositol turnover (Sethi and Rozengurt, 1991). The resulting inositol 1,4,5-trisphosphate causes release of Ca²⁺ from intracellular organelles such as the endoplasmic reticulum. Previously, we found that nanomolar concentrations of CCK-8 half maximally increased cytosolic Ca²⁺ in Fura-2 AM-loaded NCI-H209 cells within 15 s. Because the K_d for CCK-8 binding to NCI-H209 cells is 2 nM, occupation of approximately 50% of the CCK₂ receptors causes a strong Ca²⁺ response, whereas no response was observed using 0.1 nM CCK-8. Here 10 nM CCK-8 caused a strong Ca²⁺ response. Because an order of magnitude more CCK-8 was used relative to the K_d, 10 nM CI-988 had little effect, whereas 100 nM CI-988 blocked the increase in cytosolic Ca²⁺ caused by CCK-8. Thus, it is necessary to use approximately 1 order of magnitude more CI-988 than CCK-8 to block SCLC CCK₂ receptors.

When phosphatidylinositol is metabolized, diacylglycerol is released, which activates protein kinase C. Protein kinase C may phosphorylate protein substrates such as mitogen-activated protein kinase kinase, which in turn phosphorylates mitogen-activated protein kinase. CI-988 (100 nM) had little effect on basal mitogen-activated protein kinase phosphorylation but antagonized the increase in tyrosine phosphorylation of the 42- and 44-kDa bands caused by 10 nM CCK-8 after 2 min. Phosphorylated mitogen-activated protein kinase may enter the nucleus of SCLC cells and alter gene expression.

Ten nanomolar CCK-8 increased the c-fos mRNA 2-fold in NCI-H209 cells after 1 h. CI-988 (100 nM) had little effect on basal c-fos mRNA but antagonized the increase in c-fos mRNA caused by CCK-8. Preliminary data (T. Moody, unpublished data) indicate that CCK-8 increased c-jun mRNA. The c-fos and c-jun proteins may form heterodimers and activate AP-1 sites in growth factor genes. Previously, we showed that agents that cause elevated cAMP, such as prostaglandin E₂ and vasoactive intestinal peptide, increased VEGF mRNA in lung cancer cells. Also, prostaglandin E₂ and vasoactive intestinal peptide increase secretion of VEGF from lung cancer cells (Casibang et al., 2001). The VEGF may diffuse and activate receptors present on endothelial cells, facilitating angiogenesis of lung cancer tumors. The increase in VEGF mRNA after 8 h caused by 10 nM CCK-8 was antagonized by 100 nM CI-988. Preliminary data (T. Moody, unpublished data) indicate that 10 nM CCK-4 had little effect on VEGF mRNA; CCK-4 is a weak agonist for CCK₂ receptors, which binds with approximately 2 orders of magnitude lower affinity than does CCK-8. It remains to be determined whether CI-988 inhibits angiogenesis of lung cancer tumors.

CCK-8 (100 nM) caused increased tyrosine phosphorylation of focal adhesion kinase and paxillin 1 min after addition to NCI-H209 cells. The samples were immunoprecipated using an anti-phosphotyrosine antibody and blotted using anti-focal adhesion kinase and anti-paxillin antibodies, which recognized both phosphorylated and unphosphorylated forms of the protein. Preliminary data (T. Moody, unpublished data) indicate that the increases in p125^FAK and paxillin tyrosine phosphorylation caused by CCK-8 were accompanied by reorganization of the actin cytoskeleton and by the assembly of SCLC focal adhesion plaques where both p125^FAK and paxillin are localized.

CI-988 inhibited the proliferation of SCLC cells. In the MTT assay, CI-988 at 3000 nM significantly inhibited the proliferation of NCI-H82, H209, and H345 cells. In the clonogenic assay 10 nM CCK-8 stimulated the proliferation of NCI-H209 cells, which was antagonized by 1000 nM CI-988. The clonogenic assay may be more sensitive to CI-988 than the MTT assay due to the fact that the NCI-H209 cells are exposed to the CI-988 for 2 weeks in the clonogenic assay as apposed to 4 days in the MTT assay. These results indicate that CI-988 is a CCK₂ receptor antagonist whose action can be reversed by addition of exogenous CCK-8. Also, pancreatic cancer has CCK₂ receptors that are antagonized by CI-988; however, CI-988 caused gastric gland degeneration and mucosal atrophy consistent with its role of stimulating stomach mucosal proliferation (Smith and Watson, 2000). Also, CI-988 and L-365,260 but not L-364,718 inhibited the clonal growth of NCI-H209 cells. These results suggest that SCLC cells may synthesize and secrete endogenous CCK-8-like peptides that function as autocrine growth factors. Previously, we did not detect CCK-8-like immunoreactivity in extracts derived from SCLC cells (Yoder and Moody, 1987). Gastrin-like peptides have been detected in colon carcinoma cells (Dockray, 1977), however, and it remains to be determined whether SCLC cells make gastrin.

In peripheral cancers, gastrin expression is an early event in the adenoma-carcinoma sequence. Normal colonic mucosa does not express gastrin or CCK₂ receptors, whereas 78 and 81% of polyps expressed gastrin and CCK₂-R immunoreactivity. In particular, 91% of the human specimens had progastrin immunoreactivity, 80% had glycine extended gastrin, 47% had amidated gastrin (Hellmich et al., 2000), and 96% of the polyps had CCK₂ receptors. Also, colonic cancer has a splice variant of the CCK₂ receptor that is constitutively active (Goetze et al., 2000). In pancreatic cancer, amidated gastrin was detected in 75% of the specimens, whereas CCK₂ receptors were detected in 100% of the specimens (Caplin et al., 2000). Normal pancreas had little detectable gastrin or CCK₂ receptors (Dethloff et al., 1999). CCK-8 and amidated gastrin stimulated but L-365-260 inhibited the growth of many pancreatic cancer cell lines (Ohlsson et al., 1999). These results indicate that CCK₂ receptors may regulate the growth of colon and pancreatic cancer cells.

CI-988 inhibited the growth of NCI-H209 tumors in nude mice. Preliminary data (T. Moody, unpublished data) indicate the 10 µg/day but not 0.1 µg/day CI-988 in polyethylene glycol-400 by gavage inhibited NCI-H209 xenograft proliferation. Although CI-988 slowed xenograft proliferation by over 95%, tumor regression was not observed. The effects of CI-988 were reversible in that if CI-988 administration was discontinued the xenografts rapidly grew. These results suggest that CI-988 can be administered orally and absorbed from the gastrointestinal tract into the blood. Because CI-988 is an organic chemical, it is not readily degraded by stomach enzymes. Previously, it was found that i.p. administration of 1 mg/kg CI-988 prevented morphine tolerance in rats, enhanced the locomotor response to phencyclidine and dizocilpine maleate, and enhanced the reflex-depressive effect of morphine in axotomized rats (Xu et al., 1992, 1994). It remains to be determined whether CI-988 crosses the blood-brain barrier. Because the body weight of animals treated with CI-988 was similar to that of the control, whereas tumor weight was significantly reduced, there was little evidence of CI-988 toxicity in the nude mice bearing NCI-H209 xenografts (T. Moody, unpublished data).

In summary, CI-988 is a CCK₂ receptor antagonist for SCLC cells. It remains to be determined whether CI-988 will be a therapeutic agent for lung cancer.