Antiapoptotic Effects of Progastrin on Pancreatic Cancer Cells Are Mediated by Sustained Activation of Nuclear Factor-KB

Research Article

William Rengifo-Cam,1 Shahid Umar,2 Shubhashish Sarkar,1 and Pomila Singh1

Departments of 1Neuroscience and Cell Biology and 2Internal Medicine-Gastroenterology, University of Texas Medical Branch, Galveston, Texas

Abstract pancreatic cancer (6), lung cancer (7), and ovarian cancer (8) also Progastrin (PG) exerts proliferative and antiapoptotic effects express the gastrin gene and PGs. It is possible that PG-like on intestinal epithelial and colon cancer cells via Annexin II peptides play an equally important role in the growth and survival (ANX-II). In here, we show that ANX-II similarly mediates of many of these PG-expressing cancers as reported for PG- proliferative and antiapoptotic effects of PG on a pancreatic dependent CRCs (2, 9). Therefore, understanding the mechanisms cancer cell line, AR42J. The role of several signaling molecules by which PG-like peptides mediate growth and survival effects on was examined in delineating the biological activity of PG. PG cancer cells will be extremely useful in developing a targeted (0.1–1.0 nmol/L) caused a significant increase (2- to 5-fold) in approach for treating PG-dependent cancer as previously discussed the phosphorylation of phosphatidylinositol 3-kinase (PI3K), (2, 9, 10). The high-affinity PG/gastrin-binding sites are distinct Akt (Thr308), p38 mitogen-activated protein kinase (MAPK; from CCK1R and CCK2R and bind PG-like peptides in the order of Thr180/Tyr182), extracellular signal-regulated kinases (ERK; PG>Gly gastrin>G17>CCK8 (3). Annexin II (ANX-II) was recently Thr202/Tyr204), IKB kinase A/B (IKKA/B; Ser176/180), IKBA discovered as one such molecule and is required for mediating (Ser32), and p65 nuclear factor-KB (NF-KB; Ser536). Inhibition growth factor effects of PG/gastrin peptides on intestinal epithelial of p44/42 ERKs (PD98059), p38 MAPK (SB203580), Akt, and and colon cancer cells (11). Although high-affinity binding sites for PI3K (LY294002), individually or combined, partially reversed PG-like peptides are present on normal IECs, and on gastrin- antiapoptotic effects of PG. The kinetics of phosphorylation of responsive CRCs, it is not known if high-affinity PG-binding sites IKKA/B in response to PG matched the kinetics of phosphor- are present on other cancer cell types as well. We used a rat ylation and degradation of IKBA andcorrelatedwith pancreatic cancer cell line, AR42J, as the in vitro pancreatic cancer phosphorylation, nuclear translocation, and activation of cell model. AR42J cells, unlike human pancreatic cancer cell lines, p65 NF-KB. NF-KB essential modulator–binding domain do not express autocrine growth factors and/or elevated levels of n n peptide (an inhibitor of IKKA/B) effectively blocked the growth-promoting factors [such as nuclear factor- B (NF- B) or activity of p65 NF-KB in response to PG. Activation of p65 mutant Ras] and are responsive to exogenous G17/Gly gastrin NF-KB, in response to PG, was 70% to 80% dependent on (reviewed in ref. 2). In initial studies, we confirmed if AR42J cells phosphorylation of MAPK/ERK and PI3K/Akt molecules. were positive for high-affinity PG-binding sites and if the cells Down-regulation of p65 NF-KB by specific small interfering expressed ANX-II. Next, we confirmed if AR42J cells respond to RNA resulted in the loss of antiapoptotic effects of PG on proliferative and antiapoptotic effects of PG and if ANX-II mediates AR42J cells. These studies show for the first time that the growth effects of PG on AR42J cells. canonical pathway of activation of p65 NF-KB mediates Activation of c-Src kinase is required for mediating proliferative antiapoptotic effects of PG. Therefore, targeting PG and/or effects of PG on IECs (12). In mice overexpressing PG (hGAS p65 NF-KB may be useful for treating cancers, which are mice), significant increases in the relative levels of phosphorylated dependent on autocrine or circulating PGs for their growth. c-Src, phosphatidylinositol 3-kinase (PI3K)/Akt, Janus-activated [Cancer Res 2007;67(15):7266–74] kinase (JAK) 2, signal transducers and activators of transcription (STAT) 3, extracellular signal-regulated kinases (ERK), and trans- Introduction forming growth factor-a in colonic mucosa were reported (13), suggesting that one or more of these signaling molecules may be The full-length progastrin peptide (PG) is the precursor 1-80 involved in mediating proliferative and/or antiapoptotic effects of peptide of gastrin (G1-G17). G17 plays an important role in gastric PG. In fact, several of these molecules, including p38 mitogen- acid secretion via CCK R (1). Additionally, PG and gastrin peptides 2 activated protein kinase (MAPK)/ERK p44/42 and PI3K/Akt, have exert growth effects on normal and cancerous intestinal epithelial been shown to play an important role in proliferation and cells (IEC) via PG-preferring receptors (2–5). We now know that survival of pancreatic cancer cells (14–16). Additionally, activation colorectal cancers (CRC) express autocrine PG (2). Besides CRCs, of NF-nB has been reported to be one of the most frequent molecular alterations in pancreatic cancer cells (14) and is required for the proliferation and survival of the cells (14, 17). Note: Supplementary data for this article are available at Cancer Research Online Importantly, down-regulation of NF-nB sensitizes the pancreatic (http://cancerres.aacrjournals.org/). cancer cells to apoptotic death via anticancer agents (14, 18, 19). Requests for reprints: Pomila Singh, Department of Neuroscience and Cell Biology, University of Texas Medical Branch, 10.104 Medical Research Building, Route We therefore focused on examining the role of the above-listed 1043, Galveston, TX 77555-1043. Phone: 409-772-4842; Fax: 409-772-3222; E-mail: signaling molecules in mediating the observed survival effects of [email protected]. I2007 American Association for Cancer Research. PG on AR42J cells. As a result of our studies, we report for the doi:10.1158/0008-5472.CAN-07-1206 first time that PG exerts potent antiapoptotic effects on AR42J

Cancer Res 2007; 67: (15). August 1, 2007 7266 www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 2007 American Association for Cancer Research. NF-kB Mediates PG Effects on Pancreatic Cancer Cells cells, which are mediated by the phosphorylation of InB kinase (Ser536; 7F1), anti–phosphorylated InBa (Ser32/36; 5A5), polyclonal anti– (IKK) a/h, resulting in the degradation of InBa and activation of phosphorylated stress-activated protein kinase/c-Jun NH2-terminal kinase n n (JNK; Thr183/Tyr185), anti-MAPK p38, anti-ERK p44/42, anti-phosphotyr- p65 NF- B; activation of NF- B pathway seems to be 70% to 80% 180 182 dependent on the phosphorylation (activation) of p38 MAPK/ERK osine p85 PI3K, anti–phosphorylated MAPK p38 (Thr /Tyr ), anti– phosphorylated ERK p44/42 (Thr202/Tyr204), anti-IKKa/h, and anti– and PI3K/Akt. phosphorylated IKKa/h (Ser176/180) antibodies were from Cell Signaling Technology. Monospecific polyclonal anti-PG antibodies were generated in house and used as described previously (3, 20, 21). Kinase inhibitors, Materials and Methods LY294002, Akt inhibitor, SB203580, PD98059, and NF-nB essential Materials. Leupeptin, aprotinin, benzamidine, phenylmethylsulfonyl modulator (NEMO)-binding domain peptide were from Calbiochem. fluoride, sodium, orthovanadate, EDTA, NP40, octyl-D glucoside, h- Stock solutions (1 mmol/L) of all the inhibitors were prepared in 100% mercaptoethanol, Tris, HEPES, sodium chloride, sodium fluoride, glycerol, DMSO (Sigma) and stored at À70jC. Recombinant human PG (rhPG) was camptothecin, and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium generated in our laboratory as described previously (3) and radiolabeled bromide (MTT) were obtained from Sigma Chemical Co. Polyclonal with Na125I by our published methods (3). G17 and CCK8 were purchased anti-caspase-3 and anti-caspase-9 antibodies were from BD PharMingen. from Bachem. Polyclonal anti-InBa (C-21), anti-p50 NF-nB (H-119), anti–phosphorylated Cell culture. The rat pancreatic cancer cell line AR42J was propagated in Akt1/2/3 (Thr308)-R, anti-Akt1/2 (H-136), anti-PI3K p85a (Z-8), anti-actin culture as described previously (22). In vitro (I-19), anti-CCK2R (C-18), and anti-ANX-II (H-50) antibodies were from growth assays. The effect of increasing concentrations of Santa Cruz Biotechnology. Monoclonal anti-lamin B (101-B7) antibody rhPG, in the presence or absence of specific antibodies against ANX-II, PG, was from Calbiochem. Monoclonal anti–phosphorylated p65 NF-nB or CCK2R, on the growth response of AR42J cells was measured either in a

Figure 1. A, dose-dependent (0.1 nmol/L to 1.0 Amol/L) growth effects of PG on AR42J cells. Growth in response to 1%FCS as a positive control ( C). Total number of cells/35-mm dish at the end of the assay is presented. Columns, mean of six to nine separate measurements from three experiments; bars, SE. The control dishes were treated with the vehicle alone. Bi and Bii, antiapoptotic effects of an optimal dose of PG (0.1 nmol/L) were measured on AR42J cells in terms of relative levels of activated caspase-3 (Bi) and caspase-9 (Bii) in control (nontreated) versus PG-treated cells. Cells grown in the presence or absence (control) of 0.1 nmol/L PG for 48 to 72 h were treated with a potent proapoptotic agent camptothecin (10 Amol/L) for 4 h. At the end of the assay, relative levels of activated caspase-3 (Bi) and caspase-9 (Bii) were measured by Western blot analysis. Top, Western blot data from a representative experiment of a total of three experiments for caspase-3 and caspase-9. The Western blot membranes were stripped and reprobed for h-actin levels as an internal control. The ratio of the densitometric readings for activated caspases versus h-actin in control samples was arbitrarily designated 100% in (Bi) and (Bii). The ratio of activated caspases versus h-actin in PG-treated samples is presented as a %of that measured in control samples. C, RBA of PG, G17, and CCK8 for displacing the binding of either 125I-rhPG (Ci) or 125I-BH-CCK8 (Cii) to AR42J cells. AR42J cells were incubated with the indicated radiolabeled ligand in the presence or absence of increasing concentrations of either CCK8 (cholecystokinin), G17, or rhPG. The RBA of the gastrin-like peptides for the specific binding sites for rhPG (Ci; ANX-II) and CCK8 (Cii) to CCK2R was calculated as described in Materials and Methods. Points, mean values of triplicate measurements from a single experiment and representative of two similar experiments. The intraexperimental variation for each data point was <5%to 10%. X axis, nanomolar concentration of each peptide used presented in a log scale. Parentheses, excess unlabeled peptide used for displacing the binding of the radiolabeled ligand. Y axis, percentage loss in the relative binding of the radiolabeled ligand in the presence of the increasing concentrations of the indicated nonlabeled peptide. D, the effect of anti-PG, anti-ANX-II, and anti-CCK2R antibodies (Ab) on the growth response of 0.1 nmol/L PG on AR42J cells was determined. Cells were treated with antibodies and PG and growth was determined in a MTT assay as described in Materials and Methods. At the end of the assay, the colorimetric reaction was read at 540 nm. IgG, nonimmune IgG. Columns, mean of six to eight measurements from an experiment measured in duplicate; bars, SE. SFM, serum-free medium.

www.aacrjournals.org 7267 Cancer Res 2007; 67: (15). August 1, 2007

Figure 2. A–D, PG treatment stimulates degradation of InBa and increases phosphorylation of NF-nB p65 in the cellular and nuclear compartments. AR42J cells were treated with 0.1 nmol/L PG for increasing times from 3 to 48 h. At the end of the treatment, cells were processed for Western blot analysis with specific antibodies. A, samples were probed with antibodies against total InBa and reprobed with anti-h-actin antibodies. Top, representative autoradiographs of Western blot data from one of three separate experiments with antibodies against InBa; bottom, densitometric readings at 0 h for total InBa versus h-actin was arbitrarily assigned a 100%value, and the ratio at all other time points is presented as a %change from control values. *, P < 0.05 versus 0 h values in (A). B and C, samples were processed for measuring phosphorylated NF-nB (Ser536) in either the cellular lysates (B) or the nucleus (C). For nuclear samples, relative levels of h-actin in the nucleus were measured as an internal control. A set of samples in (B) was additionally pretreated with or without NEMO (an inhibitor of IKKa/h) for 1 h as indicated in Fig. 2D. B–D, top, representative autoradiographs from one of three separate experiments. The signal intensity of the bands was determined densitometrically for each blot. For samples in (B), the ratio of the readings for phosphorylated NF-nB versus total NF-nB in the cellular lysate samples was arbitrarily assigned 100%value for 0 time point. The ratios for all other time points are presented as a %change from the 0 time point values. For samples in ( C), the densitometric readings for phosphorylated NF-nB (p65536) in the nucleus versus nuclear h-actin levels was arbitrarily assigned a 100%value at 0 time point. Ratios at all other time points are presented as a %change from 0 time point values and were significantly different ( P < 0.05) versus control (0 h) values at 6, 12, 24, 36, and 48 h. Columns, mean of four blots from two separate experiments; bars, SE. *, P < 0.05 versus 0 time point values. Cells in (D) were treated for 6 h with PG in the presence or absence of NEMO. The ratio of the densitometric readings for phosphorylated NF-nB versus total NF-nB for the untreated samples (first column) was arbitrarily assigned a 100%value, and the ratios of the readings for the samples treated with PG F NEMO are presented as a %change from control nontreated samples. cell count assay or in a MTT assay as described previously (4). To examine cells on the indicated days of cell culture using Nuclear Extract kit from the effect of anti-PG antibodies, 0.1 nmol/L rhPG was preincubated with Active Motif as per the instructions of the manufacturer. Cellular lysates the antibody for 1 h before adding to the cells. In preliminary studies, we and nuclear extracts were prepared from control and PG-treated AR42J first confirmed the presence of significant levels of ANX-II protein in AR42J cells as described previously (12). The samples were processed for Western cells by Western blot analysis; significant levels of ANX-II were measured immunoblot analysis with the indicated antibodies by our published (data not shown) as previously reported for IEC and colon cancer cells (11). methods (5, 12). All the primary antibodies were used at a dilution of To examine the effect of receptor antibodies (including ANX-II), cells plated 1:1,000. The antigen-antibody complexes were detected using chemilumi- in the wells were preincubated with either anti-ANX-II or anti-CCK2R nescence reagent kit (Amersham Pharmacia Biotech). To confirm antibodies for 1 to 2 h; control cells were treated with an equal amount equivalent loading, membranes were stripped and reprobed with either of nonimmune IgG followed by the addition of 0.1 nmol/L rhPG for anti-h-actin antibody or an antibody that detects the total amount of the 48 to 72 h. indicated kinase as internal controls. The relative density of the bands was Treatment of AR42J cells for measuring surrogate markers of densitometrically analyzed. apoptosis. Cells were cultured in 35-mm dishes as described previously Binding affinity and relative binding affinity of peptides for 125I (5). For measuring surrogate markers of apoptosis (relative levels of rhPG-binding sites. The full-length rhPG was radiolabeled with Na125I activated caspase-3 and caspase-9), cells were additionally treated for (Amersham), and intact 125I-rhPG was purified by high-performance liquid 5 h with the proapoptotic agent (camptothecin) as described (5). At the chromatography as described previously (3). The binding affinity was end of the treatment, cells were processed for measuring relative levels obtained from binding assay data as described previously (3, 4). The relative of activated caspase-3 and caspase-9 by Western immunoblot analysis as binding affinity (RBA) of gastrin-like peptides for displacing specific binding described (5). In a few experiments, cells were preincubated for 1 h with PP3 of rhPG was determined as described previously (3, 4). 125I-Bolton Hunter- (20 Amol/L), LY294002 (20 Amol/L), Akt inhibitor (20 Amol/L), SB203580 CCK8 (125I-BH-CCK8; Amersham Biosciences) was also used as the (10 Amol/L), PD98059 (20 Amol/L), NEMO (200 Amol/L), and caspase-3 radiolabeled ligand, and the RBA of gastrin-like peptides for displacing inhibitor I (200 Amol/L) before stimulating with rhPG as described in the the binding of 125I-BH-CCK8 to AR42J cells was examined by our published figure legends. methods (4). Preparation of cellular lysates and crude nuclear extracts for TransAM p65 NF-KB Chemi/Transcription Factor Assay to confirm Western blot analysis. Crude nuclear extracts were prepared from AR42J the binding of activated p65 NF-KB to DNA. Activation of p65 NF-nB

Cancer Res 2007; 67: (15). August 1, 2007 7268 www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 2007 American Association for Cancer Research. NF-kB Mediates PG Effects on Pancreatic Cancer Cells was examined using TransAM p65 NF-nB Chemi Transcription Factor studies, we have similarly reported biphasic effects of PG and Assay kit from Active Motif as per the instructions of the manufacturer. gastrin peptides on the growth of IECs (3). We therefore used A Briefly, 30 L of complete binding buffer were added into each well coated 0.1 nmol/L PG as an optimal dose for all further studies. n with immobilized oligonucleotide containing the NF- B consensus site Previously, we showed potent antiapoptotic effects of PG on IEC (5¶-GGGACTTTCC-3¶). Nuclear proteins (20 AL), prepared from cells cells (5) and colon cancer cells (25). In the current studies, PG was incubated with or without rhPG for different times, were added to the wells in 80 AL complete lysis buffer and incubated for 1 h at room found to similarly exert potent antiapoptotic effect on AR42J cells; temperature. The wells were washed and incubated with anti-p65 NF-nB camptothecin-induced activation of caspase-3 and caspase-9 was antibody (1:1,000 dilution) for 1 h and washed thrice again followed by significantly reduced by >70% to 80% in the presence of 0.1 nmol/L incubation with horseradish peroxidase–conjugated second antibody PG (Fig. 1Bi and Bii). (1:10,000; 50 AL) for 1 h. The wells were washed four times followed by Several investigators have shown that PG peptides exert growth the addition of the chemiluminescent working solution. The resulting effects on target cells via novel binding sites that are distinct n chemiluminescence, proportional to the levels of nuclear p65 NF- B, from CCK2R (2). Because CCK2R are expressed at high concen- bound to the consensus sequence, was measured with a luminometer trations on AR42J cells (4, 26), in the current study, we established (DinexTechnologies) by our published methods (23). the RBA of gastrin-like peptides (PG, G17, and CCK) for the high- Promoter-reporter assay to verify the transcriptional activity of p65 affinity CCK receptors and the high-affinity PG-binding receptors. NF-KB. A promoter-reporter plasmid (2 mg), which had three copies of the interleukin-8 (IL-8) promoter fragment containing RelA/NF-kB1–binding PG dose dependently displaced the binding of radiolabeled rhPG site, cloned upstream of an inert TATA boxdriving the expression of firefly (Fig. 1Ci) but was largely ineffective in displacing the binding 125 luciferase (IL-8nBE)3-p59rAT/LUC (Invitrogen Life Technologies), was of radiolabeled I-BH-CCK8 to CCK2R on AR42J cells (Fig. 1Cii). transfected in the cells by our published methods (23, 24). Twenty-four On the other hand, CCK and G17, both of which bind CCK2R hours after transfection, the cells were treated with or without rhPG with high affinity (4, 26), showed a negligible binding affinity for (0.1 nmol/L) for 6 h (optimal for measuring maximum activity of nuclear PG-binding sites (Fig. 1Ci) but completely displaced binding of NF-nB as determined from previous experiments). Cells were lysed to 125I-BH-CCK8 (Fig. 1Cii). Recently, we discovered that ANX-II measure luciferase levels by our published methods (23, 24). molecules serve the role of high-affinity PG-binding sites (11). To K Down-regulation of p65 NF- B by small interfering RNA for confirm the role of ANX-II in mediating growth effects of PG on examining the role of p65 NF-KB in mediating the antiapoptotic AR42J cells, we examined the effect of anti-ANX-II and anti- effects of PG on AR42J cells. p65 NF-nB small interfering RNA (siRNA; CCK R antibodies (Fig. 1D). Anti-PG antibodies and anti-ANX-II from Dharmacon) was used. AR42J cells were seeded in 35-mm dishes and 2 incubated at 37jC for 24 h followed by growth in serum- and antibiotic-free antibodies significantly reduced growth of PG-stimulated AR42J medium for 24 h. The cells were then transfected with either 100 nmol/L cells to control levels, confirming a direct role of ANX-II and PG p65 NF-nB siRNA construct or a negative control siRNA construct in the observed effects (Fig. 1D); treatment of cells with anti- (Dharmacon) using 4 AL DharmaFECT transfection reagent 4 (Dharmacon). CCK2R antibodies had statistically insignificant effects on PG- After 48 h, the siRNA-transfected cells were treated with or without stimulated growth of AR42J cells (Fig. 1D). The binding affinity of 0.1 nmol/L PG for 48 h. The cells were treated with 10 Amol/L camp- rhPG for AR42J cells was determined by Scatchard analysis to be tothecin for 4 h for inducing apoptosis. Total cellular proteins were in the range of 0.1 to 0.5 nmol/L (data not shown) as previously n extracted from camptothecin-treated cells, and relative levels of p65 NF- B reported for IEC cells (3). and activated caspase-3 were measured by Western immunoblot analysis as Phosphorylation of PI3K, Akt, p38 MAPK, and p44/42 ERK described above. in response to rhPG in AR42J cells. Specific antibodies against Cell death assay to measure the loss of antiapoptotic effects of PG in the presence of kinase inhibitors. A Cell Death Detection ELISAPLUS kit (Roche) was used for these studies as published (5). Briefly, AR42J cells were seeded in 35-mm dishes (0.5 Â 104) and incubated at 37jC for 24 h followed by growth in serum-free medium for 24 h. Specific kinase inhibitors were added either alone or in combinations for 1 h followed by the addition of 0.1 nmol/L PG for 48 h. The proapoptotic agent camptothecin (10 Amol/L) was added at the end of the incubation for 4 h. The cells were processed as per the instructions of the supplier. The relative levels of cell death in the various samples were obtained from spectrophotometric readings of the wells at 405 nm. Statistical analysis. Data are presented as the mean F SE of values obtained from four to eight samples/two to three experiments. To test for significant differences between means, the nonparametric Mann-Whitney test was used using StatView 4.1 (Abacus Concepts); P values were considered to be statistically significant if <0.05.

Figure 3. PG induces activation of NF-nB p65 as measured in a DNA binding assay. AR42J cells in culture were treated with 0.1 nmol/L PG for the indicated Results times. At the end of the treatment period, the cells were processed for the Growth and antiapoptotic effects of PG on AR42J cells: preparation of nuclear protein extracts. The presence of activated NF-nB p65 in the nuclear extracts was examined with the TransAM NF-nB p65 Chemi presence of high-affinity binding sites (ANX-II) on AR42J cells. Transcription Factor Assay kit. The readout for the DNA binding assay was Dose-dependent effects of PG (0.1–1,000 nmol/L) were examined on measured in chemiluminescence units, and the readings for control nontreated samples at 0 time point was arbitrarily assigned a 100%value. As a positive the growth of AR42J cells in vitro. A biphasic response was measu- control, nuclear extracts were also prepared from AR42J cells stimulated with red, wherein the growth of AR42J cells was significantly increased (by 25 ng TNF-a/mL medium for 30 min. The chemiluminescence units measured for >60–70%) in response to low concentrations (0.1–1.0 nmol/L) of PG, samples treated with either TNF-a or with PG for increasing time points are presented as a %change from the control 0.0 values. Columns, mean of data whereas a loss or no change in growth was measured at higher from three separate experiments; bars, SE. *, P < 0.05 versus untreated concentrations of 10 nmol/L to 1.0 Amol/L (Fig. 1A). In previous 0.0 values in the first lane. www.aacrjournals.org 7269 Cancer Res 2007; 67: (15). August 1, 2007

Figure 4. Phosphorylation and activation of NF-nB p65 is downstream of phosphorylation (activation) of MAPK/ERK and PI3K/Akt. AR42J cells in culture were pretreated with the indicated inhibitor followed by stimulation with 0.1 nmol/L PG as described in Materials and Methods. At the end of the treatment, the cells were either processed for preparation of extranuclear (cellular) extracts (A and B) or processed for the preparation of nuclear extracts (C). The samples in (A) and (B) were further processed for Western blot analysis using specific antibodies against either phosphorylated p65 NF-nB (Ser536) or total NF-nB p65. A and B, top, representative autoradiographs from one of three separate experiments. The signal intensity of the bands was densitometrically analyzed, the ratio of the readings for phosphorylated NF-nB p65 (Ser536) versus the corresponding relative levels of total NF-nB was arbitrarily assigned a 100%value, and the ratios for all other treated samples are presented as a %of the control values in ( A) and (B). SB, SB203580; LY, LY294002; PD, PD98059. Columns, mean of data from three separate experiments; bars, SE. *, P < 0.05 versus control values (first column); c, P < 0.05 versus PG alone treated samples (second column). The abbreviation and the concentration of the indicated inhibitors are the same as described in the legend of Supplementary Fig. S2. For data presented in (C), a promoter-reporter assay was used for measuring the relative activation of NF-nB in response to 0.1 nmol/L PG in the presence or absence of the indicated kinase inhibitors. Cells were transiently transfected with the promoter-reporter plasmids as described in Materials and Methods. The transfected cells were then pretreated with the indicated kinase inhibitors followed by stimulation with 0.1 nmol/L PG. Control samples were not treated with either PG or the inhibitors. At the end of the treatment, cells were processed for measuring the relative levels of luciferase, and the luciferase units measured in the control samples were arbitrarily assigned a 0 value. The luciferase levels measured in all other treated samples are presented as a %of control values. A–C, columns, mean of data from two to three separate experiments. *, P < 0.05 versus control values (first column); c, P < 0.05 versus PG alone treated samples (second column). the indicated phosphorylated kinases were used. In the case of survival effects of PG in the presence of the indicated inhibitors in PI3K, we used a phosphotyrosine p85 PI3K-binding motif antibody response to the proapoptotic agent camptothecin (Supplementary that binds to tyrosine-phosphorylated YXXM motif, is recognized Fig. S2). Treatment with inhibitors of ERKs, MAPK, Akt, and PI3K by p85, and docks PI3K to the membrane resulting in activation of significantly reversed the antiapoptotic effects of PG by f70%, the kinase. As can be seen from Supplementary Fig. S1A to E, f60%, f45%, and f30%, respectively, indicating that MAPK/ERK phosphorylation of all the indicated kinases examined in this study and PI3K/Akt kinases are involved in mediating antiapoptotic was significantly increased by approximately 2- to 5-fold in AR42J effects of PG. Interestingly, inhibition of MAPK or ERK activity was cells in response to PG. In all cases, the relative levels of more effective than inhibition of PI3K/Akt in reversing the phosphorylated kinases were corrected for the total levels of the antiapoptotic effects of PG (Supplementary Fig. S2). corresponding kinase. The phosphorylation moieties examined are PG stimulates phosphorylation of IKKA/B,IKBA, NF-KB, detailed in the legend of Supplementary Fig. S1. and translocation of NF-KB into the nucleus of AR42J cells. Loss of antiapoptotic effects of PG on AR42J cells in Because activation of MAPK/ERKs and PI3K/Akt is reported to response to kinase inhibitors. To assess the role of phosphor- result in activation of NF-nB (27), we next examined if NF-nBis ylated (activated) kinases in mediating the antiapoptotic effects of activated in response to PG via the canonical pathway involving PG, specific inhibitors of the kinases were used as detailed in phosphorylation of IKKa/h and degradation of InBa (27–30). Materials and Methods and in the legend of Supplementary Fig. S2. Treatment of AR42J cells with 0.1 nmol/L PG resulted in a Activation of caspase-3 was used as a surrogate marker for loss of significant increase in the phosphorylated levels of IKKa/h, after

Cancer Res 2007; 67: (15). August 1, 2007 7270 www.aacrjournals.org

6 to 36 h of stimulation, with a peak at 6 h (Supplementary Table p65 NF-nB, binding of nuclear proteins to nB sites in an in vitro S1A). The relative levels of phosphorylated InBa were also DNA-binding assay was conducted as described in Materials and significantly increased by approximately 2- to 3-fold after 6 h of Methods. A significant increase in the binding of nuclear protein PG stimulation, and these levels remained elevated until 36 h after (NF-nB) was measured at 6 h after PG stimulation, which was f4- stimulation (Supplementary Table S1B). As expected, increased fold higher than that in control samples and f1/2 of that phosphorylation of InBa was associated with a significant decline in measured with the potent stimulator tumor necrosis factor-a the relative levels of total InBa by 6 h of PG stimulation, and the (TNF-a; Fig. 3). Importantly, activation of NF-nB in terms of levels of total InBa remained significantly low until f48 h after PG DNA binding remained elevated until f36 h after PG stimulation stimulation (Fig. 2A). In association with the degradation of InBa, (Fig. 3), once again providing evidence that PG stimulation results a significant increase in phosphorylation of cellular p65 NF-nB was in sustained activation of NF-nB. In transient transfection assays, measured in response to PG, which once again peaked at 6 h after using a promoter-reporter construct, a significant increase in the PG stimulation (Fig. 2B). A significant increase in total nuclear expression of luciferase was once again measured in AR42J cells NF-nB (Supplementary Table S1C) and in relative levels of phos- stimulated by PG for 6 to 36 h (data not shown). phorylated NF-nB(Fig.2C) was measured in the nuclear fraction of Phosphorylation (activation) of p65 NF-KB is downstream of AR42J cells after 6 h of stimulation with 0.1 nmol/L PG; the increase PI3K/Akt/MAPK/ERKs. In the next set of experiments, we in nuclear NF-nB remained elevated until f48 h after PG treatment, examined the role of individual kinases in mediating phosphory- suggesting a sustained phosphorylation of p65 NF-nB in response lation of the indicated kinases/transcription factor in response to to PG (Fig. 2C). In the presence of NEMO-binding peptide, which PG. Surprisingly, inhibition of either Akt (via Akt inhibitor) or PI3K inhibits the activation of IKKa/h (31), phosphorylation of p65 (via LY294002 compound) had no effect on the PG-stimulated NF-nBinresponsetoPGat6hwasreducedtocontrollevels(Fig.2D). phosphorylation of p44/42 ERKs (Supplementary Fig. S3A), or the PG stimulates the activation of p65 NF-KB as measured in phosphorylation of p38 MAPK (Supplementary Fig. S3B), suggest- DNA-binding assays. To confirm functional activation of nuclear ing that phosphorylation of MAPK/ERK in response to PG

Figure 5. A and B, p65 NF-nB siRNA reverses antiapoptotic effects of PG on AR42J cells. AR42J cells in culture were transiently transfected either with a control negative (Neg) siRNA or with a specific NF-nB p65 siRNA construct as described in Materials and Methods. The transfected cells were treated with or without 0.1 nmol/L PG for 48 to 72 h as indicated in (A) and (B) followed by treatment with camptothecin (10 Amol/L) for 4 h before the end of the treatment. The cells were processed for Western blot analysis with specific antibodies against either total p65 NF-nB(A) or activated caspase-3 (B). The blots were reprobed with anti-h-actin antibodies as an internal control. Treatment with the control (negative) siRNA or PG (0.1 nmol/L) or camptothecin (10 Amol/L) had no effect on the relative levels of total NF-nB(A, lanes 1–4), but treatment with specific NF-nB p65 siRNA (A, lane 5) resulted in approximately 70%to 75%loss in the relative levels of total NF-nB. Similarly, the antiapoptotic effects of PG in response to camptothecin (lanes 3 and 4) were not affected in cells transfected with the negative (control) siRNA (B, lane 4) but were completely reversed in the presence of specific NF-nB p65 siRNA (B, lane 5). C, effect of kinase inhibitors on reversing the antiapoptotic effects of PG in response to camptothecin. AR42J cells in culture were treated with the indicated kinase inhibitors followed by treatment with PG as described in Materials and Methods. Apoptosis was induced by camptothecin. In one set of experiments (lane 9), the cells were treated with the caspase-3 inhibitor (Caspase-3-I;1Amol/L for 2 h). The explanation for the abbreviations for the inhibitors and concentrations used are the same as described in the legend of Supplementary Fig. S2. At the end of the treatment, the cells were processed for measuring the relative levels of cell death using a cell death assay kit as described in Materials and Methods. An increase in absorbance (A) readings at 405 nm reflects the relative increase in cell death of the samples being examined. Columns, mean of data from four separate observations from two experiments; bars, SE. c, P < 0.05 versus lane 1 values; *, P < 0.05 versus lane 2 values. www.aacrjournals.org 7271 Cancer Res 2007; 67: (15). August 1, 2007

mediating protective (survival) effects of PG on pancreatic cancer cells. Relative role of MAPK/ERKs/PI3K/Akt and p65 NF-KBin mediating antiapoptotic effects of PG on AR42J cells. In the final set of experiments, rather than using surrogate markers of apoptosis, we used the cell death assay kit for measuring the role of various kinases in mediating antiapoptotic effects of PG as described in the legend of Fig. 5C. Inhibiting the activation of NF- nB (via NEMO peptide) was most effective in completely reversing protective effects of PG on cell death of AR42J cells in response to camptothecin (Fig. 5C, lane 4 versus lane 3). Inhibitors of MAPK and ERKs were also quite effective (70–80%) in reversing protective effects of PG (Fig. 5C, lanes 5 and 6 versus lane 3); inhibitors of PI3K and Akt were somewhat less effective (55–65%; Fig. 5C, lanes 7 and 8 versus lane 3). Inhibiting the activation of caspase-3 (Fig. 5C, lane 9) significantly reduced proapoptotic effects of camptothecin (Fig. 5C, lane 2). Various combinations of kinase inhibitors as shown in Fig. 5C (lanes 10–16) were not any n Figure 6. Diagrammatic representation of receptor subtypes and intracellular more effective than inhibiting activation of NF- B with NEMO kinases that are likely to be involved in mediating antiapoptotic effects of PG on peptide alone (Fig. 5C, lane 4); these results suggest that AR42J cells. Solid lines, pathways that were confirmed in the current study; activation of NF-nB may represent the ultimate pathway for broken arrows, pathways that need to be confirmed. The results in Fig. 1D confirm a role for ANX-II (PG/gastrin receptor) in mediating the growth effects mediating antiapoptotic effects of PG on pancreatic cancer cells. of PG. Other receptor subtypes, other than CCK2R, may also be involved in Based on the results presented in Figs. 1–5, a hypothetical model mediating the growth factor effects of PG as stipulated previously (49). Arrows up within the boxes, an increase in the phosphorylation (activation) status of the is presented in Fig. 6, describing the upstream versus downstream indicated kinase; arrows down in the boxes for caspase-3 and caspase-9, loss effects of PG on activation of various kinases. in the activation of these proapoptotic enzymes; sideway arrow in the box for p38 MAPK and ERK, possible coregulatory role of these kinases in mediating n the up-regulation of the phosphorylation (activation) status of NF- B. Discussion In the current study, we show for the first time that NF-nB stimulation may not be dependent on phosphorylation of PI3K/Akt plays a critical role in mediating antiapoptotic effects of PG on a at the time point examined. On the other hand, inhibition of either pancreatic cancer cell line. There are five members of the MAPK or ERKs resulted in a significant inhibition in relative levels mammalian NF-nB family, RelA (p65), RelB, c-Rel, p50/p105 (NF- of phosphorylated PI3K (Supplementary Fig. S3C) and phosphor- nB1), and p52/p100 (NF-nB2), which exist in unstimulated cells ylated Akt (Supplementary Fig. S3D) to control levels, suggesting as homodimers or heterodimers bound to InB family proteins that phosphorylation of PI3K/Akt molecules may be downstream (27, 28). Binding of InB to NF-nB prevents NF-nB-InB complex of phosphorylation of MAPK/ERK kinases in response to PG at from translocating to the nucleus, thereby maintaining NF-nBin time points examined in this study. an inactive state. In the classic (canonical) pathway of NF-nB The role of MAPK/ERKs and PI3K/Akt in the phosphorylation activation, on stimulation by a cytokine/growth factor, the and activation of NF-nB in response to PG was next examined as signaling pathways lead to activation of h subunit of IKK, which described in the legend of Fig. 4. Inhibition of either MAPK or ERKs phosphorylates InB proteins, leading to their polyubiquitination resulted in an almost complete loss in the relative levels of and degradation (28). The freed NF-nB dimers translocate to the phosphorylated p65 NF-nB(Fig.4A), whereas inhibition of either nucleus, where they bind to specific sequences in the promoter or PI3K or Akt was less effective (Fig. 4B versus Fig. 4A). In transient enhancer regions of target genes. In AR42J cells, PG stimulated transfection studies, using a promoter-reporter assay, a similar phosphorylation of IKKa/h, resulting in phosphorylation and pattern was obtained in response to the specific kinase inhibitors degradation of InBa followed by phosphorylation, nuclear (Fig. 4C). Inhibition of either MAPK or ERKs was more effective translocation, and activation of NF-nB p65. than inhibition of either PI3K or Akt in reversing the activation of The IKK complexcontains two kinase subunits (IKK a and IKKh) NF-nB in response to PG (Fig. 4C). and a regulatory subunit, NEMO (also called IKKg; ref. 31). In the NF-KB is required for mediating antiapoptotic effects of PG classic NF-nB signaling pathway, IKKh is both necessary and on AR42J cells. We next examined if NF-nB was required for sufficient for phosphorylating InBa and InBh. We confirmed a mediating antiapoptotic effects of PG on AR42J cells. NF-nB siRNA critical role of IKKh and NEMO in experiments with NEMO peptide was used for down-regulating expression of NF-nB. As can be seen (a selective inhibitor of IKKh). NEMO peptide attenuated PG- in Fig. 5A, cells transfected with NF-nB siRNA showed a loss of stimulated phosphorylation of NF-nB p65 and abrogated the f75% of NF-nB protein (Fig. 5A, lane 5), whereas treatment of cells antiapoptotic effects of PG on AR42J cells; this critical role of with a negative control siRNA in the presence or absence of PG or NF-nB was further confirmed in experiments with specific siRNA camptothecin had no effect on relative levels of NF-nB(Fig.5A, against NF-nB p65 transcripts. Results of our studies thus suggest lanes 1–4). Loss of NF-nB as a result of treatment of cells with NF- that NF-nB activation by the canonical pathway is primarily nB siRNA resulted in a complete loss of antiapoptotic effects of PG involved in mediating growth effects of PG on AR42J cells. (in terms of relative levels of activated caspase-3) in response to Multiple signaling pathways lead to phosphorylation and proapoptotic agent camptothecin (Fig. 5B, lane 5). Results in Fig. 5 activation of IKK complexes. IKK kinases include members of thus confirmed for the first time that NF-nB was required for the MAPK kinase family, such as MAPK/ERK kinase kinase (MEKK)

Cancer Res 2007; 67: (15). August 1, 2007 7272 www.aacrjournals.org

1, 2, and 3 (27, 31). Other putative IKK kinases include members of of NF-nB has similarly been reported to be delayed but sustained in the atypical protein kinase C (PKC) family, Akt, and two proteins response to lipopolysaccharide (LPS; ref. 41). NF-nB activation related to IKKa/h, IKKI and NF-nB–activating kinase B (27). Our induced by cigarette smoke condensate is similarly persistent (42). studies show that NF-nB activation in response to PG is In the case of PG, we have measured activation of several signaling downstream of MAPKs/ERKs and PI3K/Akt signaling pathways. kinases, including MAPK/ERKs/JNK and PI3K/Akt (current studies) A potent inhibitor of histone deacetylase, trichostatin A was also and h-catenin.3 It is thus possible that activation of multiple reported to activate NF-nB via PI3K and MAPK signaling path- signaling pathways may have resulted in the observed delayed and ways (32). sustained activation of NF-nB in response to PG in AR42J cells as In many cells, Akt acts upstream of NF-nB (33). However, in a reported for LPS, etoposide, and cigarette smoke condensate. few cases, NF-nB activation was reported to be independent of Because IKK is one of the many kinases implicated in direct Akt function (34). Whereas one group has reported that PI3K/Akt phosphorylation of p65 at Ser536 (29), we examined Ser536 does not contribute toward constitutive NF-nB activation in the phosphorylation of p65 in response to PG. We have also measured pancreatic cancer cells (19), another group reported that Ser276 phosphorylation of p65 in AR42J cells in response to PG,3 constitutive NF-nB–mediated transcription in CRC cell lines was suggesting that NF-nB is likely phosphorylated at both Ser276 and dependent on activation of PI3K/Akt (35). These discrepancies Ser536 in response to PG stimulation in AR42J cells. Whereas almost may be due to noncoordinate expression of IKKs, which was all pancreatic cancers have high levels of constitutively active NF- postulated to play a role in determining the cell type–specific role nB (14), because AR42J cells are relatively differentiated, levels of of PI3K/Akt in NF-nB activation (33). Cells with high proportion activated NF-nB are very low in unstimulated AR42J cells (Fig. 3). of IKKa (activated by Akt) to IKKh are more sensitive to PI3K We therefore measured a significant increase in the proliferation of inhibitors and vice versa (33). The results of our studies suggest AR42J cells in response to 0.1 to 1.0 nmol/L PG, which correlated that inhibition of ERKs and/or p38 MAPK almost completely with potent antiapoptotic effects of PG (Fig. 1). attenuated activation of NF-nB in response to PG, whereas Several receptor subtypes mediate the growth factor effects of inhibition of PI3K/Akt was less effective. Similarly, adiponectin, an gastrin/PG peptides on target cells (2, 26). However, we and obesity-related factor, activated NF-nB via increased phosphory- others have shown that the growth factor effects of PG are lation of ERKs and p38 MAPKs (36). However, a cell-specific role mediated by novel receptor mechanisms that are distinct from of MAPKs versus PI3K/Akt has been shown. For example, acetyl- CCK2R (2, 43, 44). In the current study, we confirmed that PG had h 11-keto- -boswellic acid suppressed activation of IKK through a relatively poor binding affinity for CCK2R(Fig.1C). On the other inhibition of Akt and p38 MAPK but not p44/p42 MAPK (37), hand, the binding of radiolabeled PG to AR42J cells was not whereas both ERK1/2 and Akt pathways mediate inhibitory inhibited by either CCK or G17 (Fig. 1C). In cell lines that do not À effects of green tea polyphenol ( )-epigallocatechin-3-gallate on express CCK2R (such as IEC 18 cells), G17 showed a significant NF-nB activation in colon cancer cells (38). These studies show binding affinity for PG-binding sites (3), which led us to postulate that signaling pathways leading toward activation of NF-nB are that RBA of G17 for PG-binding sites is dictated by the presence or cell and/or stimulus specific. absence of CCK2R (45). The results of the binding studies (Fig. 1) NF-nB can be additionally activated by IKK-independent path- provide further evidence for our above hypothesis. Several years ago, ways. For example, protein kinase A binds p65 and phosphorylates we had identified a 33 to 36 kDa protein as a high-affinity binding p65 at Ser276 independent of InB degradation (31). Akt and PKC~ protein for PG/gastrin peptides, which was pharmacologically also phosphorylate p65 at distinct residues and increase trans- different from CCK1R and CCK2R (46). Recently, we discovered that criptional activation of NF-nB without involving InB degradation ANX-II represented the novel p36 receptor protein (11). Unlike (31). a2-Macroglobulin, bound to surface-associated GRP 78 on CCK2R antibodies, ANX-II antibodies completely attenuated growth prostrate cancer cells, was reported to activate NF-nB by both effects of PG on AR42J cells (Fig. 1D), confirming our previous canonical and noncanonical pathways (39). However, a role of findings with IEC and colon cancer cells (11). noncanonical pathways in response to PG in AR42J cells seems Besides PG, other gastrin-like peptides also activate NF-nBvia less likely because NEMO almost completely attenuated antiapop- either CCK1R (47) or CCK2R (48). CCK2R antagonist and a PKC totic effects of PG on AR42J cells. inhibitor completely attenuated gastrin-induced activation of In the current studies, we measured a delayed activation of NF-nB NF-nB (48); surprisingly, inhibition of p38 MAPK had no effect on in response to PG in AR42J cells. A small peak of activation was NF-nB activation (48). However, our studies suggest that p38 MAPK measured at 30 min (Fig. 2B) followed by a more robust peak at 6 h, plays a critical role in mediating activation of NF-nB in response to which was sustained for f36 h. NF-nB regulates its own activity by PG in AR42J cells. These findings once again show that, although transcriptional activation of InBa, which represents a negative many ligands, ligand receptors, and kinases activate NF-nB, the feedback loop and drives oscillations in levels of activated NF-nB specific kinases that signal to NF-nB activation are stimulus and/or (40). Etoposide (a topoisomerase inhibitor) induces lower amplitude cell specific. The findings thus far suggest that binding of PG/G17- oscillations than TNF-a, with a delayed peak of NF-nB activation at like peptides to their cognate receptors (such as ANX-II, CCK1R, and n 300 min (40). The pattern of peak timing and amplitude/oscillation CCK2R) can potentially result in activation of NF- B via many is not only stimulus specific but also cell specific (40). We measured different pathways, resulting in differential effects on different peak levels of phosphorylated InBa and phosphorylated NF-nB cell types. p65Ser536 (translocated to the nucleus) at 6 h compared with a rapid Binding of various ligands with ANX-II has been reported to induction of peak levels in response to TNF-a (Fig. 2C). Activation result in the activation of several signaling pathways, including NF- nB, JAK/STAT, p38 MAPK, and MEKK4, as recently reviewed (49). It is possible that PG binding to ANX-II may provide a similar platform for multimeric complexformation of various kinases 3 S. Umar, S. Cowey, S. Sarkar, P. Singh, unpublished data. resulting in the observed activation of several signaling molecules www.aacrjournals.org 7273 Cancer Res 2007; 67: (15). August 1, 2007

Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 2007 American Association for Cancer Research. Cancer Research in response to PG, including Src, PI3K/Akt, JAK2, STAT5/3, ERKs, Grant support: NIH grants CA97959 and CA114264 (P. Singh) and CA099121 n (S. Umar). p38 MAPK, and NF- B (current studies; refs. 12, 50), as diagram- The costs of publication of this article were defrayed in part by the payment of page matically presented in our recent review article (49). charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Acknowledgments We thank Dr. Sanjeev Choudhary (Department of Internal Medicine, University of Texas Medical Branch, Galveston, TX) for his help with some of the methods and Received 4/5/2007; revised 5/10/2007; accepted 5/30/2007. Cheryl Simmons and Lyn Schilling for the secretarial help.

References 18. Li Y, Ahmed F, Ali S, Philip PA, Kucuk O, Sarkar FH. expression in colorectal cancer by activating nuclear Inactivation of nuclear factor nB by soy isoflavone factor-nB and h-catenin. Oncogene 2005;24:1021–31. 1. Sandvik AK, Dockray GJ. Biological activity of carboxy- genistein contributes to increased apoptosis induced by 36. Ogunwobi OO, Beales ILP. Adiponectin stimulates terminal gastrin analogs. Eur J Pharmacol 1999;364:199– chemotherapeutic agents in human cancer cells. Cancer proliferation and cytokine secretion in colonic epithelial 203. Res 2005;65:6934–42. cells. Regul Pept 2006;134:105–13. 2. Rengifo-Cam W, Singh P. Role of progastrins and 19. Arlt A, Gehrz A, Muerkoster S, et al. Role of NF-nB 37. Takada Y, Ichikawa H, Badmaev V, Aggarwal BB. gastrins and their receptors in GI and pancreatic and Akt/PI3K in the resistance of pancreatic carcinoma Acetyl-11-keto-h-boswellic acid potentiates apoptosis, cancers: targets for treatment. Curr Pharm Des 2004; cell lines against gemcitabine-induced cell death. inhibits invasion, and abolishes osteoclastogenesis by 10:2345–58. Oncogene 2003;22:3243–51. suppressing NF-nB and NF-nB-regulated gene expres- 3. Singh P, Lu X, Cobb S, et al. Progastrin1-80 stimulates 20. Singh P, Velasco M, Given R, Wargovich M, Varro A, sion. J Immunol 2006;176:3127–40. growth of intestinal epithelial cells in vitro via high- Wang TC. Mice overexpressing progastrin are predis- 38. Peng G, Dixon DA, Muga SJ, Smith TJ, Wargovich MJ. affinity binding sites. Am J Physiol Gastrointest Liver posed for developing aberrant colonic crypt foci in Green tea polyphenol (À)-epigallocatechin-3-gallate Physiol 2003;284:G328–39. response to AOM. Am J Physiol Gastrointest Liver inhibits cyclooxygenase-2 expression in colon carcino- 4. Singh P,Owlia A, Espeijo R, Dai B. Novel gastrin receptors Physiol 2000;278:G390–9. genesis. Mol Carcinog 2006;45:309–19. mediate mitogenic effects of gastrin and processing 21. Cobb S, Wood T, Ceci J, Varro A, Velasco M, Singh P. 39. Misra UK, Deedwania R, Pizzo SV. Activation and intermediates of gastrin on Swiss 3T3 fibroblasts. Absence Intestinal expression of mutant and wild-type progastrin cross-talk between Akt, NF-nB, and unfolded protein of detectable cholecystokinin (CCK)-A and CCK-B recep- significantly increases colon carcinogenesis in response response signaling in 1-LN prostate cancer cells tors. J Biol Chem 1995;270:8429–38. to azoxymethane in transgenic mice. Cancer 2004;100: consequent to ligation of cell surface-associated 5. Wu H, Owlia A, Singh P. Precursor peptide progas- 1311–23. GRP78. J Biol Chem 2006;281:13694–707. trin(1-80) reduces apoptosis of intestinal epithelial cells 22. Singh P, Draviam E, Guo YS, Kurosky A. Molecular 40. Nelson DE, Ihekwaba AE, Elliott M, et al. Oscillations and upregulates cytochrome c oxidase Vb levels and characterization of bombesin receptors on rat pancre- in NF-nB signaling control the dynamics of gene synthesis of ATP. Am J Physiol Gastrointest Liver atic acinar AR42J cells. Am J Physiol 1990;258:G803–9. expression. Science 2004;306:704–8. Physiol 2003;285:G1097–110. 23. Dai B, Wu H, Holthuizen E, Singh P. Identification of 41. Covert MW, Leung TH, Gaston JE, Baltimore D. 6. Caplin M, Savage K, Khan K, et al. Expression and a novel cis element required for cell density-dependent Achieving stability of lipopolysaccharide-induced NF-nB processing of gastrin in pancreatic adenocarcinoma. Br J down-regulation of insulin-like growth factor-2 P3 activation. Science 2005;309:1854–7. Surg 2000;87:1035–40. promoter activity in Caco2 cells. J Biol Chem 2001;276: 42. Shishodia S, Potdar P, Gairola CG, Aggarwal BB. 7. Koh TJ, Field JK, Varro A, et al. Glycine-extended 6937–44. Curcumin (diferuloylmethane) down-regulates cigarette gastrin promotes the growth of lung cancer. Cancer Res 24. Shen Q, Singh P. Identification of a novel SP3 binding smoke-induced NF-nB activation through inhibition of 2004;64:196–201. site in the promoter of human IGFBP4 gene: role of SP3 InBa kinase in human lung epithelial cells: correlation 8. van Solinge WW, Odum L, Rehfeld JF. Ovarian cancers and AP-1 in regulating promoter activity in CaCo2 cells. with suppression of COX-2, MMP-9 and cyclin D1. express and process progastrin. Cancer Res 1993;53:1823–8. Oncogene 2004;23:2454–64. Carcinogenesis 2003;24:1269–79. 9. Singh P, Owlia A, Varro A, Dai B, Rajaraman S, Wood T. 25. Wu H, Rao GN, Dai B, Singh P. Autocrine gastrins in 43. Ahmed S, Budai B, Heredi-Szabo K, et al. High and Gastrin gene expression is required for the proliferation colon cancer cells up-regulate cytochrome c oxidase Vb low affinity receptors mediate growth effects of gastrin and tumorigenicity of human colon cancer cells. Cancer and down-regulate effluxof cytochrome c and activa- and gastrin-Gly on DLD-1 human colonic carcinoma Res 1996;56:4111–5. tion of caspase-3. J Biol Chem 2000;275:32491–8. cells. FEBS Lett 2004;556:199–203. 10. Ferrand A, Wang TC. Gastrin and cancer: a review. 26. Dufresne M, Seva C, Fourmy D. Cholecystokinin and 44. Ahmed S, Murphy RF, Lovas S. Importance of N- and Cancer Lett 2006;238:15–29. gastrin receptors. Physiol Rev 2006;86:805–47. C-terminal regions of gastrin-Gly for preferential 11. Singh P, Wu H, Clark C, Owlia A. Annexin II binds 27. Hayden MS, Ghosh S. Signaling to NF-nB. Genes Dev binding to high and low affinity gastrin-Gly receptors. progastrin and gastrin-like peptides, and mediates 2004;18:2195–224. Peptides 2005;26:1207–12. growth factor effects of autocrine and exogenous 28. Karin M, Ben Neriah Y. Phosphorylation meets 45. Singh P, Cobb S. Role of gastrins in colon carcino- gastrins on colon cancer and intestinal epithelial cells. ubiquitination: the control of NF-nB activity. Annu Rev genesis. In: Merchant J, Buchan A, and Wang T, editors. Oncogene 2007;26:425–40. Immunol 2000;18:621–63. Gastrin in the new millennium. Los Angeles (CA): CURE 12. Brown D, Yallampalli U, Owlia A, Singh P. pp60c-Src 29. Sakurai H, Chiba H, Miyoshi H, Sugita T, Toriumi W. Foundation; 2004. p.319–327. kinase mediates growth effects of the full-length InB kinases phosphorylate NF-nB p65 subunit on serine 46. Chicone L, Narayan S, Townsend CM, Jr., Singh P. The precursor progastrin1-80 peptide on rat intestinal 536 in the transactivation domain. J Biol Chem 1999;274: presence of a 33-40 KDa gastrin binding protein on epithelial cells, in vitro. Endocrinology 2003;144:201–11. 30353–6. human and mouse colon cancer. Biochem Biophys Res 13. Ferrand A, Bertrand C, Portolan G, et al. Signaling 30. Sizemore N, Lerner N, Dombrowski N, Sakurai H, Commun 1989;164:512–9. pathways associated with colonic mucosa hyperprolif- Stark GR. Distinct roles of the InB kinase a and h 47. Han B, Logsdon CD. CCK stimulates mob-1 expres- eration in mice overexpressing gastrin precursors. subunits in liberating nuclear factor nB (NF-nB) from sion and NF-nB activation via protein kinase C and Cancer Res 2005;65:2770–7. InB and in phosphorylating the p65 subunit of NF-nB. intracellular Ca(2+). Am J Physiol Cell Physiol 2000;278: 14. Li L, Aggarwal BB, Shishodia S, Abbruzzese J, J Biol Chem 2002;277:3863–9. C344–51. Kurzrock R. Nuclear factor-nB and InB kinase are 31. Ghosh S, Karin M. Missing pieces in the NF-nB 48. Ogasa M, Miyazaki Y, Hiraoka S, et al. Gastrin constitutively active in human pancreatic cells, and puzzle. Cell 2002;109:S81–96. activates nuclear factor nB (NFnB) through a protein their down-regulation by curcumin (diferuloylmethane) 32. Kwon O, Kim KA, Kim SO, et al. NF-nB inhibition kinase C dependent pathway involving NFnB inducing is associated with the suppression of proliferation and increases chemosensitivity to trichostatin A-induced kinase, inhibitor nB(InB) kinase, and tumour necrosis the induction of apoptosis. Cancer 2004;101:2351–62. cell death of Ki-Ras-transformed human prostate factor receptor associated factor 6 (TRAF6) in MKN-28 15. Furukawa T, Kanai N, Shiwaku HO, Soga N, Uehara A, epithelial cells. Carcinogenesis 2006;27:2258–68. cells transfected with gastrin receptor. Gut 2003;52: Horii A. AURKA is one of the downstream targets of 33. Gustin JA, Ozes ON, Akca H, et al. Cell type-specific 813–9. MAPK1/ERK2 in pancreatic cancer. Oncogene 2006;25: expression of the InB kinases determines the signifi- 49. Singh P. Role of Annexin-II in GI cancers: 4831–9. cance of phosphatidylinositol 3-kinase/Akt signaling to interaction with gastrins/progastrins. Cancer Lett 16. Bose C, Zhang HL, Udupa KB, Chowdhury P. Activation NF-nB activation. J Biol Chem 2004;279:1615–20. 2007;252:19–35. of p-ERK1/2 by nicotine in pancreatic tumor cell line 34. Machuca C, Mendoza-Milla C, Cordova E, et al. 50. Hollande F, Choquet A, Blanc EM, Lee DJ, Bali JP, AR42J: effects on proliferation and secretion. Am J Physiol Dexamethasone protection from TNF-a-induced cell Baldwin GS. Involvement of phosphatidylinositol 3- Gastrointest Liver Physiol 2005;289:G926–34. death in MCF-7 cells requires NF-nB and is independent kinase and mitogen-activated protein kinases in 17. Greten FR, Weber CK, Greten TF, et al. Stat3 and from AKT. BMC Cell Biol 2006;7:9. glycine-extended gastrin-induced dissociation and NF-nB activation prevents apoptosis in pancreatic car- 35. Agarwal A, Das K, Lerner N, et al. The AKT/InB migration of gastric epithelial cells. J Biol Chem cinogenesis. Gastroenterology 2002;123:2052–63. kinase pathway promotes angiogenic/metastatic gene 2001;276:40402–10.

Cancer Res 2007; 67: (15). August 1, 2007 7274 www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 2007 American Association for Cancer Research. Antiapoptotic Effects of Progastrin on Pancreatic Cancer Cells Are Mediated by Sustained Activation of Nuclear Factor-κB

William Rengifo-Cam, Shahid Umar, Shubhashish Sarkar, et al.

Cancer Res 2007;67:7266-7274.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/67/15/7266

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2007/08/03/67.15.7266.DC1

Cited articles This article cites 49 articles, 18 of which you can access for free at: http://cancerres.aacrjournals.org/content/67/15/7266.full#ref-list-1

Citing articles This article has been cited by 4 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/67/15/7266.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/67/15/7266. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.