Attenuation of Peroxisome Proliferator-activated Receptor γ (PPARγ) Mediates Gastrin-stimulated Colorectal Cancer Cell Proliferation*

Peroxisome proliferators-activated receptor γ (PPARγ) has been shown to suppress cell proliferation and tumorigenesis, whereas the gastrointestinal regulatory peptide gastrin stimulates the growth of neoplastic cells. The present studies were directed to determine whether changes in PPARγ expression might mediate the effects of gastrin on the proliferation of colorectal cancer (CRC). Initially, using growth assays, we determined that the human CRC cell line DLD-1 expressed both functional PPARγ and gastrin receptors. Amidated gastrin (G-17) attenuated the growth suppressing effects of PPARγ by decreasing PPARγ activity and total protein expression, in part through an increase in the rate of proteasomal degradation. G-17-induced degradation of PPARγ appeared to be mediated through phosphorylation of PPARγ at serine 84 by a process involving the biphasic phosphorylation of ERK1/2 and activation of the epidermal growth factor receptor (EGFR). These results were confirmed through the use of EGFR antagonist AG1478 and MEK1 inhibitor PD98059. Furthermore, mutation of PPARγ at serine 84 reduced the effects of G-17, as evident by inability of G-17 to attenuate PPARγ promoter activity, degrade PPARγ, or inhibit the growth suppressing effects of PPARγ. The results of these studies demonstrate that the trophic properties of gastrin in CRC may be mediated in part by transactivation of the EGFR and phosphorylation of ERK1/2, leading to degradation of PPARγ protein and a decrease in PPARγ activation.

Peroxisome proliferators-activated receptor ␥ (PPAR␥) has been shown to suppress cell proliferation and tumorigenesis, whereas the gastrointestinal regulatory peptide gastrin stimulates the growth of neoplastic cells. The present studies were directed to determine whether changes in PPAR␥ expression might mediate the effects of gastrin on the proliferation of colorectal cancer (CRC). Initially, using growth assays, we determined that the human CRC cell line DLD-1 expressed both functional PPAR␥ and gastrin receptors. Amidated gastrin (G-17) attenuated the growth suppressing effects of PPAR␥ by decreasing PPAR␥ activity and total protein expression, in part through an increase in the rate of proteasomal degradation. G-17-induced degradation of PPAR␥ appeared to be mediated through phosphorylation of PPAR␥ at serine 84 by a process involving the biphasic phosphorylation of ERK1/2 and activation of the epidermal growth factor receptor (EGFR). These results were confirmed through the use of EGFR antagonist AG1478 and MEK1 inhibitor PD98059. Furthermore, mutation of PPAR␥ at serine 84 reduced the effects of G-17, as evident by inability of G-17 to attenuate PPAR␥ promoter activity, degrade PPAR␥, or inhibit the growth suppressing effects of PPAR␥. The results of these studies demonstrate that the trophic properties of gastrin in CRC may be mediated in part by transactivation of the EGFR and phosphorylation of ERK1/2, leading to degradation of PPAR␥ protein and a decrease in PPAR␥ activation.
Colorectal cancer (CRC) 2 is the third most common cause of cancerrelated death in the United States, accounting for 10% of all deaths caused by malignancy. The American Cancer Society estimated that in 2005, ϳ145,000 Americans would be diagnosed with CRC, of whom 56,000 would die. CRC generally arises from benign adenomas, which progress into malignant adenocarcinomas (2,3). The development of CRC appears to involve a multistep process of genetic mutations combined with largely undefined environmental factors whereby normal epithelial cells undergo dysplastic transformation, followed by proliferation and eventual histological progression to neoplasia (2). This transition appears to require multiple genetic alterations, such as mutations in the adenomatous polyposis coli (APC) gene. After the initial mutations have established a neoplastic phenotype, other factors appear to promote proliferation and neoplastic progression (2).
One such potential growth factor is the polypeptide hormone gastrin. In addition to its recognized role in the physiological regulation of acid secretion (4 -7), another biological property attributed to gastrin is its trophic effect on the gastrointestinal (GI) mucosa (6 -8). Numerous studies have provided evidence that gastrin peptides play an integral role in promoting colorectal tumor growth (3), as well as other malignancies throughout the GI tract (8,9). In a well controlled, elegantly designed epidemiologic study of nearly 130,000 individuals, Thorburn et al. (10) found that prolonged hypergastrinemia comprised a risk factor for the development of CRC. This study found a 3.9-fold increased risk of CRC due to elevated circulating gastrin levels, prompting the authors to conclude that 8.9% of all CRCs may be attributed to hypergastrinemia (10). Studies in transgenic mice overexpressing gastrin have demonstrated increased proliferation of the gastric and colonic epithelium after eight months. When these mice were followed for longer periods (ϳ20 months), an increased tendency to develop neoplasia was observed (11).
Although the trophic effects of gastrin are well recognized, the molecular and intracellular mechanisms by which gastrin modulate cell growth in the GI tract have not been fully elucidated. Previous studies have reported that stimulation of the CCK-2R by gastrin activates various signal transduction pathways implicated in cell proliferation, such as the mitogen-activated protein kinases (MAPK), which include ERK, JNK, and p38 kinase (8). One potential downstream target of the MAPKs is peroxisome proliferator-activated receptor ␥ (PPAR␥) (19 -24). The possibility of a functional relationship between gastrin and PPAR␥ has not been previously evaluated.
PPAR␥, a member of the nuclear hormone receptor family, functions as a transcription factor that regulates several biological processes, including growth and differentiation (25,26). In addition to its recog-nized role in adipogenesis (31), PPAR␥ has been shown to modulate the growth of cells in various organs. This trophic effect is most evident in the colon, where normal human colonic mucosa, colon adenocarcinoma, and cultured CRC cells express levels of PPAR␥1 equivalent to that detected in adipocytes (27,28). Activation of PPAR␥ in cultured colon cells inhibits growth and induces differentiation, reverses the malignant phenotype, and promotes apoptosis (27, 29 -32). In CRC cells, PPAR␥ activation results in both an increase in the cyclin-dependent kinase inhibitors, p21 and p27 (33), which repress cell cycle progression, leading to a decrease in cell growth and an increase in the differentiation of cancer cells, and up-regulation of caspase activity (34,35), resulting in DNA fragmentation and apoptosis. Moreover, a recent study demonstrated that 8% of primary colorectal tumors harbor a functional mutation in one allele of the PPAR␥ gene, further supporting the role of PPAR␥ as a tumor suppressor in humans (36).
The present studies were directed to determine whether changes in PPAR␥ expression might mediate the effects of gastrin on the proliferation of CRC. We have demonstrated that gastrin-stimulated proliferation of CRC cells is associated with a significant concomitant decrease in cellular PPAR␥ levels. Moreover, gastrin attenuates the inhibition of cell growth induced by PPAR␥ agonists. Finally, our studies demonstrate that the trophic properties of gastrin may be mediated in part by transactivation of the epidermal growth factor receptor (EGFR) and phosphorylation of ERK1/2, leading to degradation of PPAR␥ protein and a decrease in PPAR␥ activation.

EXPERIMENTAL PROCEDURES
Materials-DLD-1 human adenocarcinoma cells were obtained from the American Type Culture Collection, SW48 human colon adenocarcinoma cells were kindly donated by Dr. B. Vogelstein (Baltimore, MD), and MC-26 cells, a transplantable mouse CRC cell line, were obtained from Dr. K. K. Tanabe (Boston, MA). The PPAR␥ monoclonal (E-8, no. sc-7273) antibody was purchased from Santa Cruz Biotechnology (Santa Cruz, CA), and ␣-amidated gastrin-17 (G-17) and Gly-G were purchased from Bachem (King of Prussia, PA) and Anspep (Parkville, Victoria, Australia), respectively. The phosphospecific PPAR␥ polyclonal antibody (clone AW504) and the phosphospecific monoclonal antibody to the epidermal growth factor receptor (pTyr 1173 ) were purchased from Upstate Biotechnology (Waltham, MA). The phosphospecific monoclonal antibody to ERK1/2 (Thr 202 /Tyr 204 ) was purchased from Cell Signaling Technologies (Beverly, MA), and the proteasome inhibitor, MG132 (N-carbobenzoxyl-Leu-Leu-Leucinal), cycloheximide, and thymidine were purchased from Sigma. [ 3 H]Thymidine was purchased from PerkinElmer Life Sciences (Boston, MA). FuGENE 6 and Complete Protease Inhibitor mixture tablets were purchased from Roche, and the thiazolidinediones, ciglitazone and rosiglitazone, which stimulate PPAR␥ activity, were purchased from Cayman Chemicals (Ann Arbor, MI). The pHD(x3)Luc vector was a kind gift from Dr. John Capone of McMaster University (Hamilton, ON). This vector contains three tandem repeats of the peroxisome proliferators response element from the promoter of the rat hydratase-dehydrogenase gene subcloned into the BamHI site of pCPSluc located immediately upstream of the carbamoyl-phosphate synthetase promoter. L-365,260 was generously provided by Dr. L. Iverson (Oxford, UK), pCMV-␤-gal was purchased from Invitrogen, and PD98059, SB203580, and lactacystin were purchased from Calbiochem. The RNAeasy Mini kit and the SYBR green quantitative PCR master mix were purchased from Qiagen (Valencia, CA), and pTracerA/Bsd plasmid and ThermoScript TM Reverse Transcriptase were purchased from Invitrogen. The Dual-Luciferase Reporter assay system, the ␤-Galactosidase Enzyme assay system, and the CellTiter 96 AQ ueous One Solution Cell Proliferation (MTT) assay kit were purchased from Promega. The QuikChange XLII site-directed mutagenesis kit was purchased from Stratagene (La Jolla, CA). KpnI, XbaI, and the Quick Ligase kit were purchased from New England Biolabs (Ipswich, MA).
Site-directed Mutagenesis and Generation of PPAR␥ Phosphorylation Mutants-Site-directed mutagenesis was performed utilizing the QuikChange XLII kit with the forward primer, 5-GTG-GAGCCTGCAGCTCCACCTTATTATTCTG-3, and the reverse primer, 3-CACCTCGGACGTCGAGGTGGAATAATAAGAC-5, and the pcDNA3-FLAG-wtPPAR␥ vector as a template to generate a substitution of Ser 84 to Ala. An initial denaturation step was performed at 95°C for 2 min. and followed by 20 cycles at 95°C for 1 min, annealing at 57°C for 1 min, and extension at 68°C for 7.5 min. A final extension phase was performed at 68°C for 7 min. The DNA sequence was confirmed by the Tufts University Core Facilities. PPAR␥ sequence inserts were double-digested with KpnI and XbaI and subcloned into the pTracerA/Bsd vector utilizing the Quick Ligase Kit according to the manufacturer's instructions, at a 3:1 ratio of PPAR␥ insert to pTracer. Stable cell lines expressing pTracer, pTracerFLAG-wtPPAR␥, or pTracerFLAG-mutPPAR␥ were generated by blasticidin selection (5 g/ml) and confirmed by Western analysis using the mouse monoclonal anti-PPAR␥ antibody and the mouse monoclonal anti-green fluorescent protein (GFP) antibody, and by fluorescent light microscopy.
Cell Culture-The human CRC cell line DLD-1 and the mouse CRC cell line MC-26 were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (FBS) (Invitrogen) and 1% penicillin/streptomycin (Invitrogen). The human CRC cell line SW48 was maintained in McCoy's medium (Invitrogen) supplemented with 10% FBS and 1% penicillin/streptomycin. All cell lines were kept in a humidified, 5% CO 2 environment. 24 h prior to experimentation, serum-containing media were replaced with serum-free media.
[ 3 H]Thymidine Incorporation-Equal amounts of cells (3 ϫ 10 5 cells/ well) were plated in 6-well plates. After 24 h of serum starvation in 0.5 mM thymidine-containing medium, cells were gently washed with phosphate-buffered saline (PBS) and then incubated for 22 h in fresh serum-free medium in the presence of G-17 or vehicle alone. In addition, to examine the involvement of the CCK-2R in DNA synthesis, 500 nM L-365,260, a specific CCK-2/gastrin receptor antagonist, was added to the culture medium. 4 h prior to termination of the experiment, cells were pulsed with [ 3 H]thymidine at a final concentration of 1 Ci/well. At the end of the incubation period, the radioactive medium was aspirated, and cell monolayers were gently rinsed with PBS at 4°C. Ice-cold 10% trichloroacetic acid was then added, and cells were gently rocked at 4°C for 30 min. Cell monolayers were washed again with PBS, followed by incubation for 30 min at 25°C in 1 N NaOH. Scintillation fluid was added, and samples were transferred into scintillation vials for measurement of [ 3 H]thymidine using an automatic Beckman liquid scintillation counter.
Cell Counting-Equal amounts of cells (3 ϫ 10 5 cells) were plated onto 10-cm plates. After 24 h of serum starvation in 0.5 mM thymidine-containing medium, cells were gently washed with PBS, and the medium was replaced with 1% FBS-containing medium with or without G-17. After a 6-day incubation period, cells were trypsinized, harvested, and resuspended in equal volumes. Cell number was determined by counting with the aid of a hemacytometer under an inverted light microscope.
MTT Assays-MTT (3-(4,5-dimethylthiazoyl)-2,5-diphenyltetrazolium bromide) assays (Promega) were also performed to measure cell growth. This method measures the quantity of the formazan product, as measured by the ratio of 490/630 nm absorbance, which is proportional to the number of living cells in culture. Cells were seeded onto 96-well plates at a density of 5 ϫ 10 3 cells per well and incubated overnight in 10% FBS medium. The medium was then replaced with serum-free medium the following day and incubated for 24 h. To determine the effect of gastrin on cell growth, the medium were then replaced with medium containing 1% FBS and increasing concentrations of G-17 or Gly-G, and cells were grown for 4 days. To determine the effect of PPAR␥ ligands on cell growth, ciglitazone and rosiglitazone, two PPAR␥ agonists, were added after serum starvation in medium containing 1% FBS, and cells were grown for an additional 3 days. To determine the effects of gastrin on cell growth in the presence of PPAR␥ ligands, DLD-1 cells were preincubated with gastrin peptides for 12 h prior to addition of ciglitazone or rosiglitazone. At the end of each experiment, 20 l of CellTitre96 Aqueous Solution Reagent were added to each well, and the plate was incubated for 30 min. The absorbance ratio (490/630 nm) was recorded using a 96-well El x 800 universal plate reader (BIO-TEK Instruments, Inc., Winooski, VT).
Transient Transfection and Luciferase Assays-DLD-1 cells (4 ϫ 10 4 cells/well) were plated in 24-well plates. After an overnight incubation, cells were transiently transfected with the pHD(x3)luc vector in Opti-MEM for 16 h. To normalize for transfection efficiency, the cells were co-transfected with a pCMV-␤-gal reporter construct. FuGENE 6 was used according to the manufacturer's instructions, and a FuGENE6 to DNA ratio of 3:1 was used in each transfection experiment. After a 16-h transfection period, the medium was replaced with medium containing 1% FBS in the presence or absence of 200 nM gastrin for 12 h. The cells were then treated with various concentrations of ciglitazone for an additional 16 h, after which cell lysates were collected and PPAR␥ reporter activity measured using the luciferase assay system. Values were normalized to ␤-galactosidase activity.
RNA Extraction and RT-PCR Analysis-DLD-1 cells (1.50 ϫ 10 6 per plate) were plated onto 10-cm diameter plates and incubated overnight. After 24 h of serum starvation, DLD-1 cells were incubated in the presence or absence of G-17. At the indicated time points, RNA was extracted using the RNAeasy Mini kit. Total RNA was measured, and 1 g of total RNA was reverse-transcribed using the ThermoScript reverse transcriptase. The reverse transcriptional reaction was carried out at 50°C for 60 min and 85°C for 5 min. To quantify the amount of PPAR␥ cDNA, all samples were subjected to PCR amplification using the QuantiTect SYBR Green PCR Kit (Qiagen). The forward primer PPAR␥-F, 5-TCTCTCCGTAATGGAAGACC-3 and reverse primer PPAR␥-R, 5-GCATTATGAGACATCCCCAC-3, were used according to the method of Terashita et al. (37). The PCR protocol was as follows: initial denaturation at 95°C for 15 min, followed by 35 cycles at 94°C for 15 s, annealing at 55°C for 30 s, and extension at 72°C for 30 s. The PCR product was quantified by the intensity of SYBR Green I fluorescence at 83°C.
Western Blot Analysis-Cell monolayers were rinsed twice with 1ϫ PBS, directly lysed in the plate on ice with radioimmune precipitation assay buffer containing Tris-HCl (50 mM, pH 7.4), NaCl (150 mM), Nonidet P-40 (1%), sodium deoxycholate (0.5%), SDS (0.1%), 1 M phenylmethylsulfonyl fluoride, and complete Protease Inhibitor Mixture. Cell debris was pelleted by centrifugation at 14,000 rpm for 15 min, and the supernatant was collected for protein quantification. The bicinchoninic acid protein assay (Pierce) was used to estimate protein concentration according to the manufacturer's instructions. 50 g of protein were diluted with 4ϫ SDS sample loading buffer, boiled for 5 min, and separated by SDS-polyacrylamide gels. Following electrophoresis, separated proteins were transferred onto nitrocellulose membranes.
The membranes were then blocked with 5% milk/PBS and incubated with the indicated primary antibodies. After incubation with the primary antibodies, membranes were washed thoroughly in TBS-Tween buffer (25 mM Tris, pH 8.0, 125 mM NaCl, 0.1% Tween 20). Appropriate secondary antibodies conjugated with horseradish peroxidase were used to detect the primary antibodies. Immunoreactive bands were visualized by chemiluminescence in signaling solution (Pierce).
ERK1/2 Activation, EGF Receptor Activation, PPAR␥ Phosphorylation-To investigate the effects of gastrin on ERK1/2 activation, EGFR transactivation, or PPAR␥ phosphorylation, DLD-1 cells were incubated in the presence or absence of 200 nM G-17. At different time points, whole cell lysates were collected, and Western analysis was performed with specific antibodies. For identification of the active form of ERK1/2, an antibody specifically recognizing ERK1 and ERK2 phosphorylated at Thr 202 and Tyr 204 , respectively, was utilized. To determine the involvement of the EGFR in ERK1/2 activation, DLD-1 cells were preincubated for 30 min with the EGFR antagonist AG1478 prior to the addition of G-17. To evaluate EGFR transactivation by gastrin, Western analysis was performed with a monoclonal antibody specifically immunoreactive with the EGFR phosphorylated at Tyr 1173 to detect the activated form of the EGFR. To examine PPAR␥ phosphorylation, Western analysis was performed using a rabbit polyclonal antibody specifically immunoreactive with PPAR␥ phosphorylated at Ser 84 . Protein loading was normalized by the measurement of ␤-actin.
PPAR␥ Stability-To examine the effects of gastrin on PPAR␥ protein stability, DLD-1 cells were incubated with the protein synthesis inhibitor cycloheximide (50 g/ml) in the presence or absence of 200 nM G-17. At different time points, whole cell lysates were collected, and Western analysis was performed using the anti-PPAR␥ antibody to detect PPAR␥, with ␤-actin measured as a loading control. Linear regression analysis was performed to determine the half-life of PPAR␥ in the presence and absence of gastrin. Values were normalized to ␤-actin and plotted on a log versus time scale. To ascertain whether PPAR␥ protein was degraded through the proteasomal pathway, cells were preincubated with the proteasomal inhibitor, MG132 (20 M) or lactacystin (5 M), for 30 min prior to the addition of 200 nM G-17. To determine whether ERK1/2 and EGFR activation might affect PPAR␥ half-life, DLD-1 cells were preincubated with 40 M PD98059, a potent MEK1 inhibitor, to inhibit the downstream activation of ERK1/2, with 100 nM AG1478, or with 10 M SB20358, the p38 MAPK inhibitor, for 30 min prior to addition of cycloheximide.
Statistical Analysis-All results are expressed as the mean Ϯ S.D. Statistical analysis was performed using analysis of variance and Student's t test. A p value Ͻ0.05 was considered to be statistically significant. In separate experiments, DLD-1 cells were incubated in the presence of increasing concentrations of two known PPAR␥ agonists, ciglitazone and rosiglitazone, to determine the effect of PPAR␥ activation on cellular proliferation. Both ciglitazone and rosiglitazone treatment suppressed cell growth in a concentration-dependent manner. Rosiglitazone treatment decreased cellular proliferation at concentrations as low as 1 M (by 18.4 Ϯ 6.7%, p Ͻ 0.05), and maximal inhibition began to plateau from 5 M (69.4 Ϯ 10.2% of control, p Ͻ 0.01) to 10 M (64.8 Ϯ 8.6% of control, p Ͻ 0.01) (Fig. 2a). Similarly, ciglitazone treatment decreased cellular proliferation at concentrations as low as 1 M (by 13.8 Ϯ 6.6%, p Ͻ 0.05), with maximal inhibition achieved at a concentration of 10 M (71.3 Ϯ 3.4% of control, p Ͻ 0.01) (Fig. 2b). In the presence of 200 nM G-17, PPAR␥ growth suppression induced by 10 M ciglitazone and 10 M rosiglitazone was significantly attenuated to 93.8 Ϯ 5.3% of control ( p Ͻ 0.05) (Fig. 2b) and 89.9 Ϯ 6.8% of control ( p Ͻ 0.05) (Fig. 2a), respectively.

Effects of PPAR␥ and Gastrin on Cellular Proliferation-To
Effects of Gastrin on PPAR␥ Activity-Because gastrin attenuated the inhibitory effects of PPAR␥ activation on the growth of DLD-1 cells, we hypothesized that PPAR␥ activity might likewise be affected. To examine this possibility, we measured PPAR␥ transcriptional activity using the pHD(x3)Luc vector, a construct that has been previously used to assess PPAR␥ activity (38). DLD-1 cells were transiently transfected with the luciferase reporter construct and the pCMV-␤-galactosidase vector as a control. Following transfection, DLD-1 cells were incubated with ciglitazone in the presence or absence of 200 nM G-17. As shown in Fig.3,ciglitazonetreatmentenhancedPPAR␥activityinaconcentrationdependent manner. A 1.6-fold increase ( p Ͻ 0.001) in PPAR␥ transcriptional activity was observed in the presence of 1 M ciglitazone, and maximum activity was achieved with 10 M ciglitazone (2.9-fold, p Ͻ 0.001). This ligand-dependent PPAR␥ activation correlated with the growth inhibition observed with ciglitazone treatment. However, when cells were preincubated with G-17, PPAR␥ activation induced by ciglitazone was markedly attenuated at all concentrations. In the presence of 200 nM G-17, PPAR␥ activation by 10 M ciglitazone was significantly attenuated by ϳ50% to a 1.42-fold increase ( p Ͻ 0.01) (Fig. 3).
Effects of Gastrin on PPAR␥ Expression-We next examined whether gastrin might affect PPAR␥ protein expression. DLD-1 cells were incubated in the presence or absence of G-17 (50 nM or 200 nM) for 12, 24, and 48 h, at which time PPAR␥ protein levels were measured by Western analysis. As shown in Fig. 4a, in response to the incubation of DLD-1 cells in the presence of G-17, PPAR␥ protein levels decreased when compared with control levels at all indicated time points. In addition, the decrease in PPAR␥ protein levels was concentration-dependent, with more pronounced reductions detected in the presence of 200 nM G-17. To determine whether the gastrin-promoted decrease in PPAR␥ protein concentrations occurred as a result of a decrease in PPAR␥ gene expression, DLD-1 cells were incubated in the presence of 50 nM and 200 nM G-17 or vehicle alone for 3, 6, 12, and 24 h, at which time total RNA was extracted and quantitative PCR analysis performed. No significant differences in PPAR␥ gene expression were detected in the presence of G-17 when compared with vehicle treatment at any of the above time points (data not shown).
Two additional CRC cell lines, SW48 and MC-26, were employed to determine whether the reduction in PPAR␥ protein levels induced by G-17 was cell-specific. As shown in Fig. 4b, 24-h incubation with 200 nM G-17 significantly diminished PPAR␥ protein levels in both SW48 and MC-26 CRC cells. Therefore, the effects of gastrin on PPAR␥ expression observed in DLD-1 cells, a human CRC cell line possessing an APC   mutation, appear to extend to a murine CRC cell line (MC-26) and to a human CRC cell line possessing the wild-type APC phenotype (SW48).
Because gastrin decreased PPAR␥ protein levels in the absence of exogenous PPAR␥ ligands, we next examined the effect of gastrin on PPAR␥ protein levels in the presence of the PPAR␥ ligand, rosiglitazone. When DLD-1 cells were incubated with 5 M rosiglitazone, PPAR␥ protein levels slightly increased when compared with untreated cells (Fig. 4c). The addition of 200 nM G-17 to the culture medium containing 5 M rosiglitazone significantly diminished PPAR␥ protein levels, whereas PPAR␥ levels decreased to a greater extent when cells were incubated in the presence of 200 nM G-17 alone (Fig. 4c).
Proteasomal Degradation of PPAR␥-Recent studies have demonstrated that the treatment of 3T3-L1 adipocytes with interferon ␥ (IFN␥) decreases PPAR␥ half-life (39,40). In addition, Hauser et al. (71) have demonstrated that PPAR␥ is degraded by the proteasome in adipocytes. To further define the mechanisms mediating the gastrin-stimulated decrease in PPAR␥ protein levels, we investigated the possibility that gastrin might promote PPAR␥ proteasomal degradation. DLD-1 cells were incubated in the presence of the protein synthesis inhibitor, cycloheximide (50 g/ml), with or without G-17. As depicted in Fig. 5a, after a 12-h incubation, G-17 induced PPAR␥ degradation, with the maximal effect observed using 200 nM G-17. Furthermore, the half-life of PPAR␥ in DLD-1 cells incubated in the presence of 200 nM G-17 decreased from ϳ11.3 h to ϳ7.1 h (p Ͻ 0.05) (Fig. 5b).
To further investigate the role of the proteasome in PPAR␥ degradation, DLD-1 cells were incubated in the presence of two 26 S proteasomal inhibitors, lactacystin or MG132. In the presence of 5 M lactacystin or 20 M MG132, both basal and gastrin-induced PPAR␥ degradation were markedly inhibited (Fig. 6). These results are consistent with the hypothesis that a decrease in PPAR␥ protein levels following gastrin treatment is mediated by targeting PPAR␥ for proteasomal degradation.
Gastrin Transactivation of the EGF Receptor-Earlier studies have demonstrated the involvement of many growth factors in the transactivation of the EGFR (41)(42)(43)(44)(45)(46)(47)(48)(49). However, gastrin transactivation of the EGFR has not been previously examined in CRC. To evaluate the role of EGFR transactivation by gastrin in CRC, DLD-1 cells were incubated in the presence of 200 nM G-17 for 2 h, and whole cell lysates were collected at various time points. The lysates were evaluated by Western analysis using an antibody against the phosphorylated, active form of the EGFR. As shown in Fig. 7, G-17 promoted a biphasic activation of the EGFR. An initial increase in EGFR phosphorylation was observed after 5 min, which then declined after 30 min to a level slightly above control. The second component of the biphasic activation of the EGFR was detected at 60 min and remained steadily up-regulated for up to 2 h.

Gastrin-stimulated Activation of ERK1/2 and the Involvement of EGF
Receptor Transactivation-The Ras/Raf/Mek/ERK1/2 pathway comprises a known signaling pathway for the promotion of cell proliferation. To investigate the possibility that gastrin activates the MAPK pathway in CRC, we incubated DLD-1 cells in the presence or absence of 200 nM G-17 for 2 h, and whole cell lysates were collected at various time points. Western analysis was subsequently performed using antibodies specific for the active, phosphorylated forms of the MAPK extracellular signalregulated protein kinases, ERK1 and ERK2. As seen in Fig. 8, ERK1 and ERK2 (ERK1/2) were dramatically activated in the presence of 200 nM G-17 by 5 min, as determined by ERK1/2 phosphorylation. This activation declined to baseline levels after 15 min, which was followed by a second rise in ERK1/2 activation after 60 min treatment. ERK1/2 activation by 200 nM G-17 remained elevated at 120 min. ERK1/2 activation by G-17 exhibited a similar profile to gastrin-induced EGFR transactivation (Fig. 7).
Past studies have demonstrated that ERK1/2 activation may occur, in part, through transactivation of the EGFR by G-protein-coupled receptors (10 -18). Because ERK1/2 activation by gastrin nearly paralleled gastrin-stimulated EGFR transactivation in our study, we hypothesized that gastrin may promote activation of ERK1/2 through transactivation of the EGFR. To examine the role of gastrinstimulated transactivation of the EGFR in ERK1/2 activation, DLD-1 cells were pretreated with the EGFR kinase inhibitor AG1478, which inhibited gastrin-stimulated EGFR phosphorylation (Data not shown). Moreover, in the presence of AG1478, activation of ERK1/2 by G-17 was nearly abolished at all observed time points when compared with the activated ERK1/2 levels in cells treated with G-17 alone (Fig. 8). A slight up-regulation of ERK2 activation still remained after 5 min and 120 min when compared with the levels evaluated at 0 min (Fig. 8). These results suggest that gastrin-induced EGFR transactivation may play a role in ERK1/2 activation. (19,20,22,23). Because gastrin promoted MAPK activation, the peptide may have, in turn, also promoted PPAR␥ phosphorylation. To examine this possibility, DLD-1 cells were incubated in the presence of 200 nM G-17 for 1 h, after which Western analysis was performed using a phosphospecific antibody recognizing Ser 84 phosphorylation of PPAR␥. As depicted in Fig. 9a, G-17 promoted PPAR␥ phosphorylation starting at 5 min and persisting up to 60 min. Furthermore, G-17-induced PPAR␥ phosphorylation coincided with ERK1/2 activation. ␤-Actin was used as a loading control as previous studies have shown that the total PPAR␥ protein levels are not significantly affected over a 60-min time period.

Involvement of EGFR Transactivation and ERK1/2 Activation in Gastrin-stimulated Phosphorylation of PPAR␥ at Ser 84 -Numerous studies have demonstrated that MAPK phosphorylates PPAR␥ in adipocytes
Because gastrin stimulated ERK1/2 activation and appeared to promote EGFR transactivation, we examined the possibility that these two pathways may also play a role in gastrin-stimulated PPAR␥ phosphorylation by utilizing AG1478 and PD98059 to inhibit EGFR and ERK1/2 activity, respectively. DLD-1 cells were pretreated individually with these inhibitors for 45 min prior to addition of 200 nM G-17. As depicted in Fig. 9b, in the presence of AG1478, gastrin-stimulated PPAR␥ phosphorylation at Ser 84 was significantly diminished. Similarly, PPAR␥ phosphorylation was abolished by co-incubation with PD98059 (Fig.  9b). Moreover, gastrin-induced ERK1/2 activation was abolished by the co-incubation of DLD-1 cells with either AG1478 or PD98059 (Fig. 9b). These results suggest that gastrin-enhanced PPAR␥ phosphorylation at Ser 84 is mediated, in part, through EGFR transactivation and ERK1/2 activation. (40) have demonstrated that ERK1/2 activation is involved in IFN␥ -promoted PPAR␥ degradation in adipocytes. However, the role of MAPK phosphorylation in the regulation of PPAR␥ has not been evaluated in neoplasia. To examine the roles of gastrin-stimulated EGFR and ERK1/2 activation in PPAR␥ degradation, we employed PD98059 to inhibit ERK1/2 activation and AG1478 to inhibit EGFR activity and used Western analysis to measure PPAR␥ protein levels. We also used SB203580, a p38 MAPK inhibitor, as a   negative control because this MAPK has not been demonstrated to play a major role in PPAR␥ phosphorylation (20). DLD-1 cells were pretreated with the inhibitors individually for 45 min prior to the addition of cycloheximide and vehicle alone or 200 nM G-17. In the presence of PD98059, G-17-induced PPAR␥ degradation was abrogated when compared with G-17 treated cells alone (Fig. 10). In addition, EGFR inhibition by AG1478 also abolished gastrin-induced PPAR␥ degradation (Fig. 10). As predicted, SB203580 had no significant effect on gastrinstimulated PPAR␥ degradation (Fig. 10). These results suggest that gastrin may promote PPAR␥ degradation through activation of ERK1/2 and transactivation of the EGFR, without involvement of the p38 MAPK pathway.

The Roles of EGFR and ERK1/2 Activation in PPAR␥ Degradation by Gastrin-Floyd and Stephens
Mutation of the Ser 84 Phosphorylation Site Reverses the Attenuation of Ligand-dependent PPAR␥ Activity by Gastrin-Ser 84 phosphorylation of PPAR␥ has been demonstrated to inhibit ligand-dependent PPAR␥ activation in adipocytes (19). Because gastrin promoted PPAR␥ phosphorylation at Ser 84 , we examined the possibility that gastrin might attenuate ligand-dependent PPAR␥ activation through this mechanism. A phosphorylation mutant of PPAR␥ was generated by mutating the Ser 84 residue into an alanine residue, and stable lines of DLD-1 cells expressing GFP-tagged pTracer, pTracer-FLAG-wtPPAR␥, or pTracer-FLAG-mutPPAR␥ S84A were created. Protein expression was confirmed by Western analysis (Fig. 11a), and PPAR␥ transcriptional activity in these cell lines was measured using the pHD(x3)Luc vector. Cells were transiently transfected with the luciferase reporter construct and the pCMV-␤-galactosidase vector as a control. Following transfection, DLD-1 cells were incubated with 0, 5, or 10 M rosiglitazone in the presence or absence of 200 nM G-17. As shown in Fig. 11b, no significant differences were observed in basal PPAR␥ activity between the cell lines overexpressing wild-type PPAR␥ (wtPPAR␥) and those overexpressing mutant PPAR␥ S84A. Rosiglitazone incubation increased PPAR␥ activity in both cell lines either overexpressing wtPPAR␥ or the mutant PPAR␥ S84A. Maximal activity began to plateau at a rosiglitazone con-centration of 5 M in both the wild-type (2.4-fold versus 0 M rosiglitazone, p Ͻ 0.001) and mutant (2.2-fold versus 0 M rosiglitazone, p Ͻ 0.001) PPAR␥ overexpressing cell lines. However, whereas co-incubation with G-17 markedly attenuated PPAR␥ activation stimulated by rosiglitazone at all concentrations in DLD-1 cells overexpressing wtPPAR␥, gastrin had no effect in cells overexpressing the mutant PPAR␥ S84A (Fig. 11b).
Mutation of the Ser 84 Phosphorylation Site Attenuates Gastrin-promoted Degradation of PPAR␥ Protein-Because attenuation of liganddependent PPAR␥ activity by gastrin was abrogated by alteration of the Ser 84 phosphorylation site, we next determined whether this mutation might affect gastrin enhanced PPAR␥ protein degradation. DLD-1 cells stably overexpressing either wtPPAR␥ protein or mutPPAR␥ S84A protein were incubated in the presence of cycloheximide (50 g/ml), with or without G-17. After 12 h, PPAR␥ protein levels were measured by Western analysis. Whereas wtPPAR␥ protein levels were significantly diminished in the presence of gastrin, no significant effect of gastrin on mutPPAR␥ S84A protein levels was observed (Fig. 11c). These observations indicate that gastrin may promote PPAR␥ protein degradation through the phosphorylation of Ser 84 .
Mutation of the Ser 84 Phosphorylation Site Attenuates the Effects of Gastrin on Ligand-dependent PPAR␥ Growth Suppression-We next determined whether mutation of the Ser 84 phosphorylation site of PPAR␥ might also affect the attenuation of PPAR␥-induced growth suppression by gastrin. 5 M rosiglitazone decreased DLD-1 cell prolif- eration to 64% of control ( p Ͻ 0.001) in DLD-1 cells overexpressing wtPPAR␥, and to 66% ( p Ͻ 0.001) in cells overexpressing mutant PPAR␥ S84A, when compared with their respective controls (Fig. 11d ). Moreover, in response to gastrin incubation, rosiglitazone-induced growth suppression was diminished by ϳ64% ( p Ͻ 0.001 versus 5 M rosiglitazone alone) in wtPPAR␥-overexpressing cells. In contrast, in the presence of 200 nM G-17, PPAR␥ growth suppression was minimally attenuated in cells overexpressing mutPPAR␥ S84A (Fig. 11d). These results indicate that gastrin appears to attenuate rosiglitazone-induced PPAR␥ growth suppression, at least in part, through phosphorylation of PPAR␥ at Ser 84 . Because gastrin reversed the effects of rosiglitazone on PPAR␥ growth suppression in part by the phosphorylation of PPAR␥ at Ser 84 , we next examined whether the overexpression of mutPPAR␥ S84A might attenuate gastrin-stimulated DLD-1 cell proliferation. As seen in Fig. 11e, 200 nM G-17 promoted a 49.3 Ϯ 8.4% increase in DLD-1 cell proliferation when compared with control conditions, an effect that was diminished by ϳ38% in response to the overexpression of mutPPAR␥ S84A (Fig. 11e).
Glycine-extended Gastrin Promotes Proliferation by Decreasing PPAR␥ Protein Levels-Studies were next performed to determine whether, similar to ␣-amidated gastrin, glycine-extended gastrin might also promote DLD-1 cell proliferation by decreasing PPAR␥ protein levels. DLD-1 cells were incubated in the presence of 1 nM -100 nM Gly-G. As shown in Fig. 12a, Gly-G promoted a concentration-dependent increase in DLD-1 cell proliferation, as determined using the MTT proliferation assay. DLD-1 cell growth increased by 15.0 Ϯ 2.2% and 26.1 Ϯ 2.2% in the presence of 1 nM and 10 nM Gly-G, respectively, with maximal proliferation detected with 50 nM Gly-G (49.5 Ϯ 1.5%) and 100 nM Gly-G (50.1 Ϯ 1.8%). To determine whether the mitogenic properties of Gly-G might be mediated, in part, by decreasing PPAR␥ levels, DLD-1 cells overexpressing the empty vector, wild-type PPAR␥, or mutPPAR␥ S84A were incubated in the presence of 50 nM Gly-G for 24 h. As depicted in Fig. 12b, 50 nM Gly-G decreased PPAR␥ protein levels in the cell lines overexpressing the empty vector or wild-type PPAR␥. In contrast, no change in PPAR␥ protein levels was observed in the cells overexpressing mutPPAR␥ S84A. These observations suggest that Gly-G may likewise promote the growth of DLD-1 colorectal cancer cells, in part, by decreasing PPAR␥ protein levels through the promotion of PPAR␥ phosphorylation at Ser 84 .

DISCUSSION
Recent reports have indicated that both hypergastrinemia and decreased expression of PPAR␥ are associated with an increased risk for the devel-opment of CRC (10,37,50,51). However, the possibility of a link between gastrin and PPAR␥ has not been reported previously. In the present study, we have reported for the first time that ligand-dependent PPAR␥ growth suppression is attenuated by gastrin, at least in part, through attenuation of PPAR␥ activity and through an increase in PPAR␥ protein degradation. In addition, gastrin-induced EGFR transactivation and ERK1/2 activation appear to play a central role in this process.
Gastrin has been implicated as a trophic factor for various neoplasms of GI origin, including CRC and those of pancreatic, gastric, and esophageal origin (52)(53)(54)(55)(56)(57)(58)(59)(60)(61)(62). Gastrin and gastrin receptors are expressed aberrantly in colorectal adenomas and malignant tumors and have been implicated in malignant progression (17,18). While Smith and Watson (17) and Guo et al. (46) reported that a majority (up to 80%) of all colorectal adenomas express the CCK-2R, others have detected the receptor in only 11-38% of such tumors (15,63). These discrepancies may be explained, at least to some extent, by the detection of splice variants of the CCK-2R and by differences in methods utilized by various investigators.
The results of the present study demonstrate that gastrin and its precursor, Gly-G, stimulated the growth of DLD-1 cells in a concentration-dependent manner. Interestingly, the specific CCK-2 receptor antagonist L-365,260 only partially attenuated G-17-stimulated growth of DLD-1 cells. These findings suggest that gastrin-stimulated cell growth in these cells may be mediated, at least in part, by other receptors that bind gastrin peptides. In addition to the "classical" gastrin receptor binding site, several alternative binding sites on DLD-1 cells have been proposed. Ahmed et al. (64) recently identified two gastrin binding sites with nanomolar and micromolar affinities, respectively, on the cell membrane, both of which played key roles in cell proliferation stimulated by amidated and precursor gastrin peptides. Furthermore, Yang et al. (65) identified a micromolar affinity site for both G-17 and the non-amidated glycine-extended gastrin precursors on DLD-1 cells.
Thus, it appears that the activation of alternative gastrin receptors, in addition to the "classical" CCK-2R, might mediate gastrin-stimulated proliferation of DLD-1 cells.
The activation of ERK1/2 is a known pathway by which gastrin stimulates cell proliferation (42,46,66,67). Although several studies have demonstrated that gastrin promotes ERK1/2 activation through transactivation of the EGFR, many of them have been performed in cell lines ectopically overexpressing the CCK-2R (42,45,46,60,62). The present study represents the first to demonstrate that gastrin promotion of ERK1/2 activation involves EGFR transactivation in human CRC cells expressing endogenous gastrin receptors. Our findings corroborate an earlier study by Guo et al. (46) who reported that gastrin activates ERK1/2 in a biphasic manner in rat intestinal epithelial (RIE) cell lines ectopically expressing the CCK-2R (46). However, in their study, EGFR transactivation appeared to promote only the later phase of ERK1/2 activation. In our study, EGFR transactivation played a key role in both early and late ERK1/2 activation, as demonstrated by significant inhibition of both phases in the presence of the EGFR kinase inhibitor, AG1478. In addition, the biphasic activation of ERK1/2, although not identical, paralleled that of EGFR transactivation in the presence of gastrin, further suggesting the involvement of the EGFR in the early and late phases of ERK1/2 activation. Moreover, these results are consistent with the hypothesis that gastrin-stimulated EGFR transactivation promotes ERK1/2 activation, which in turn appears to play a critical role in mediating the trophic properties of gastrin in CRC.
Nevertheless, the possibility of the involvement of additional path- ways independent of the EGFR in the regulation of gastrin-enhanced phosphorylation of ERK1/2 cannot be excluded. As indicated above, gastrin-stimulated phosphorylation of the EGFR exhibited a similar profile to that of ERK1/2, but was not identical. For example, at 15 min, ERK1/2 phosphorylation was diminished while EGFR phosphorylation was increased. Furthermore, residual ERK1/2 activation was evident at 5 and 120 min the presence of AG1478. Thus, it is likely that gastrin promotes ERK1/2 phosphorylation by additional pathways not involving the EGFR, including direct stimulation of MAP kinase pathways (42,46,66,67).
Although not examined in the present study, several mechanisms responsible for gastrin-stimulated EGFR transactivation have been suggested in other cell lines overexpressing the CCK-2R. For example, CCK-2R activation by gastrin increases the shedding of several membrane-bound EGFR ligands, including HB-EGF, TGF␣, FGF, and amphiregulin (44,45,60,62,68). Shedding of these ligands may promote autocrine/paracrine stimulation of the EGFR. In RIE cells ectopically overexpressing the CCK-2R, gastrin transactivation of the EGFR was reported to be mediated through an intracellular signaling pathway involving activation of the adaptor protein Src (46). Src activation promotes the formation of a multiprotein complex containing EGFR, Src, and other adaptor proteins, which leads to activation of the EGFR (69). An alternative suggested mechanism involves gastrin-stimulated expression of EGFR ligands. In AGS human gastric carcinoma cells, gastrin stimulated the expression of HB-EGF through PKC-and EGFRdependent mechanisms (60). Future studies will be necessary to determine whether these pathways likewise mediate gastrin-stimulated EGFR transactivation in DLD-1 and other CRC cells.
Activation of ERK1/2 plays a key role in promoting the mitogenic effects of gastrin, and PPAR␥, a tumor suppressor, has been identified as a substrate for phosphorylation by the various MAPKs (19,20,22,23,70), including ERK1/2. Because gastrin is known to increase ERK1/2 activation, we sought to determine whether PPAR␥ phosphorylation might also be involved in promoting the mitogenic effects of gastrin. Our findings indicate that activation of ERK1/2 by gastrin was associated with the phosphorylation of PPAR␥1 at Ser 84 , a putative MAPK phosphorylation site located within the ligand-independent transactivation domain (AF-1). Previous studies have demonstrated that phosphorylation of PPAR␥ in the AF-1 domain down-regulates its activity (19,20,22,23,70). Interestingly, the EGFR kinase inhibitor AG1478 only partially attenuated gastrin-stimulated PPAR␥ phosphorylation, while the MEK inhibitor PD98059 abolished PPAR␥ phosphorylation. These observations provide additional evidence that both EGFR-dependent and -independent activation of ERK1/2 may promote gastrinstimulated PPAR␥ phosphorylation.
In the present studies, gastrin also decreased ligand-dependent PPAR␥ activity. The elimination of the Ser 84 phosphorylation site prevented the attenuation of ligand-dependent PPAR␥ activity by gastrin. Thus, gastrin-promoted phosphorylation at Ser 84 appears to play an integral role in the down-regulation of ligand-enhanced PPAR␥ activity. It has been previously shown that phosphorylation of PPAR␥ may diminish PPAR␥ activity by activating the proteasome. Floyd and Stephens (40) demonstrated that IFN␥ treatment of 3T3-L1 adipocytes enhanced ERK1/2 activation which coincided with the ubiquitination and subsequent proteasomal degradation of PPAR␥, thus suggesting a role for phosphorylation as a signal for PPAR␥ degradation. In our studies, gastrin promoted the phosphorylation of PPAR␥ at Ser 84 and enhanced the proteasomal degradation of PPAR␥. Moreover, the mutation of the Ser 84 phosphorylation site stabilized PPAR␥ protein in the presence of gastrin and was associated with attenuation of ligand-de-pendent PPAR␥ growth suppression. Therefore, gastrin-stimulated phosphorylation of PPAR␥ at Ser 84 may promote subsequent proteasomal degradation of PPAR␥, leading to a decrease in PPAR␥ protein expressionandligand-dependentactivity,therebyattenuatingtheliganddependent PPAR␥ growth suppression of DLD-1 cells.
Although Hauser et al. (71) previously demonstrated that PPAR␥ ligands target PPAR␥ for degradation, in the present study, PPAR␥ protein levels increased in response to the incubation of DLD-1 cells with rosiglitazone. This discrepancy may be explained in part by different cell types used during the course of the studies. Whereas Hauser et al. (71) employed NIH-3T3 fibroblasts and fibroblast-derived adipocytes, we performed our studies using colorectal cancer cell lines. Similar discrepancies have been observed in studies examining the estrogen receptor, in which incubation with estradiol led to the rapid degradation of the estrogen receptor ␣ in MCF-7 breast cancer cells, but stabilization in Ishikawa uterine-derived cells (72).
Proper maintenance of PPAR␥ activity and protein levels is critical for the suppression of tumorigenesis, as demonstrated by recent reports showing that PPAR␥ ϩ/Ϫ mice are more susceptible to CRC (1). Moreover, Sarraf et al. (36) recently reported that 8% of CRC tumors harbor a loss of function mutation in the PPAR␥ gene, which results in decreased functional PPAR␥ protein levels. As evident by the partial attenuation of gastrin-induced cell proliferation in cells overexpressing the PPAR␥ S84A mutant (Fig. 11e), our studies suggest that gastrin appears to exert its trophic effects in part through phosphorylation, inactivation, and degradation of PPAR␥.
In conclusion, the present study is consistent with the following proposed model (Fig. 13). Gastrin peptides appear to promote cell proliferation, in part, by inducing PPAR␥ degradation. Initially, gastrin stimulates ERK1/2 activation, at least in part, through transactivation of the EGFR, which, in turn, phosphorylates PPAR␥ at the Ser 84 residue. Although not examined in the present study, previous studies have demonstrated that gastrin also stimulates ERK1/2 activation directly (42,46,66,67). The phosphorylation of PPAR␥ at Ser 84 decreases PPAR␥ transactivational activity and signals PPAR␥ for proteasomal degradation. As a result, PPAR␥ protein levels are diminished, thereby attenuating the tumor suppressive effects of PPAR␥, resulting in enhanced cellular proliferation. Additional studies will be necessary to fully elucidate the relationship between gastrin peptides and PPAR␥ and to determine their precise roles in the pathogenesis CRC and other malignancies of the GI tract. FIGURE 13. Proposed model of gastrin-stimulated cell proliferation by enhancement of PPAR␥ protein degradation. Gastrin promotes activation of ERK1/2, in part, through activation of the EGFR. Although not examined in the present study, gastrin may also stimulate ERK1/2 activation directly. ERK1/2 activation leads to phosphorylation of PPAR␥ at Ser 84 . Ser 84 phosphorylation decreases ligand-dependent PPAR␥ activity and signals PPAR␥ for degradation. As a result, PPAR␥ levels are diminished, which, in turn, stimulates DLD-1 cell proliferation.