Gastrin stimulates expression of plasminogen activator inhibitor (PAI)-1 in gastric epithelial cells

Plasminogen activator inhibitor (PAI-)1 is associated with cancer progression, fibrosis and thrombosis. It is expressed in the stomach but the mechanisms controlling its expression there, and its biological role, are uncertain. We sought to define the role of gastrin in regulating PAI-1 expression and to determine the relevance for gastrin-stimulated cell migration and invasion. In gastric biopsies from subjects with elevated plasma gastrin, the abundances of PAI-1, uPA and uPAR mRNAs measured using qPCR, were increased compared to subjects with plasma concentrations in the reference range. In patients with hypergastrinemia due to autoimmune chronic atrophic gastritis, there was increased abundance of PAI-1, uPA and uPAR mRNAs that was reduced by octreotide or antrectomy. Immunohistochemistry revealed localisation of PAI-1 to parietal cells and ECL cells in micronodular neuroendocrine tumors in hypergastrinaemic subjects. Transcriptional mechanisms were studied using a PAI-1-luciferase promoter-reporter construct transfected into AGS-GR cells. There was timeand concentrationdependent increase of PAI-1-luciferase expression in response to gastrin that was reversed by inhibitors of the PKC and MAPK pathways. In Boyden chamber assays, recombinant PAI-1 inhibited gastrin-stimulated AGS-GR cell migration and invasion, and siRNA treatment increased responses to gastrin. We conclude that elevated plasma gastrin concentrations are associated with increased expression of gastric PAI-1 which may act to restrain gastrinstimulated cell migration and invasion. (213 words) Nørsett et al., page 3 In addition to its well known roles in stimulating gastric acid secretion and gastric epithelial cell proliferation, the gastric hormone gastrin also regulates the expression of a variety of genes that appear to be involved in control of mucosal organisation (7). These include both matrix metalloproteinases (MMP) such as MMP-7 and MMP-9, and serine protease inhibitors including tissue inhibitors of matrix metalloproteinases (TIMPS) and plasminogen inhibitor (PAI)-2 (33; 34; 38; 39). The latter is part of the urokinase plasminogen activator (uPA) system that in many tissues is associated with thrombolysis, fibrosis, tissue remodelling and the progression to cancer. This system consists of the serine proteinase uPA which converts plasminogen to plasmin, several plasminogen activator inhibitors (PAI-1, -2, and -3), and the uPA receptor (uPAR) which binds uPA at the advancing edge of migrating cells (30). In addition to activation of plasmin and consequent degradation of fibrin, uPA is associated with degradation of extracellular matrix, stimulation of angiogenesis, mitogenesis, cell migration and adhesion. The mechanisms controlling the expression of different members of the uPA system in gastric mucosa are still far from clear. Recent studies have indicated that Helicobacter pylori infection is associated with increased expression of PAI-2 and other members of the uPA system in gastric mucosa (18; 19; 35). Members of the uPA system are also expressed in gastric cancer and there is evidence that they are associated with poor clinical outcome in these tumours, as well as others (9; 11; 16; 23; 24). Together these observations suggest that the uPA system might have a role in controlling gastric mucosal organisation and function. Nørsett et al., page 4 Gastric neuroendocrine tumours may arise as a consequence of transformation of enterochromaffin-like (ECL) cells on a background of elevated plasma gastrin concentrations secondary to autoimmune atrophic gastritis leading to destruction of parietal cells, pernicious anemia (PA) and profound hypochlorhydria, or as a consequence of gastrinoma particularly in the multiple endocrine neoplasia (MEN)-1 syndrome (4; 5; 10). Although the association of some gastric neuroendocrine tumors with hypergastrinemia is well recognized, the mechanisms by which gastrin determines tumor progression are not well understood. A previous study noted that gastrin treatment of AGS-GR cells increased the abundance of PAI-1 detected in DNA arrays, and more recently uPA was shown to be upregulated in patients treated with the acid inhibitor esomeprazole who exhibited hypergastrinemia (25; 33). In the present study, we asked whether there was up-regulation of members of the uPA system in hypergastrinemia and examined the consequences for cell migration and invasion in the case of PAI-1. Nørsett et al., page 5 MATERIALS AND METHODS Cells, plasmids and reagents. AGS-GR cells were maintained as previously published (36). A plasmid encoding 4.5 kb of the human PAI-1 promoter coupled to luciferase (PAI-1-luc) was generated as previously reported (19). Gastrin-17 was obtained from Bachem (St Helens, Merseyside, UK). Phorbol12-myristate-13-acetate (PMA) was obtained from Sigma (Poole, Dorset, UK); LY294002, PD98059, and U0126 were obtained from Promega (Southampton, UK). Human recombinant PAI-1 and Ro-32-0432 were obtained from Calbiochem (Nottingham, UK). Small interfering RNA (siRNA) for PAI-1 and control scrambled oligonucleotides were obtained from Santa Cruz (Santa Cruz, CA, US). Other chemicals were obtained from Sigma (Poole, Dorset, UK). Patients. Endoscopic pinch biopsies were obtained from dyspeptic patients with normal endoscopy. Two antral biopsies were taken for urease test and 2 – 4 antral and 2 – 4 corpus biopsies were taken for histology; an additional 6 corpus biopsies were taken for RNA extraction. The present studies describe patients that were H.pylori negative by serology, antral urease test and antral and corpus histology. Some patients were on proton pump inhibitors and these were discontinued one week before gastroscopy. On the basis of plasma gastrin concentrations, 16 subjects were selected with gastrin <30 pM (upper limit of the reference range) and 12 with gastrin >30 pM. In the former group 8 had been on PPIs and in the latter 6. In addition biopsies were taken from ten patients with PA. Five patients with PA who exhibited macroscopic ECL cell neuroendocrine tumours were subjected to octreotide (25 μg/h) Nørsett et al., page 6 suppression test for 72 h and based on their response underwent antrectomy (13). Gastric biopsies were obtained at endoscopy before and immediately after octreotide, and before and 12 months after antrectomy. In patients with macroscopic ECL cell nodules, the biopsies used for this study were taken from adjacent mucosa. The study was approved by the Ethics Committees of the Royal Liverpool and Broadgreen and Aintree University Hospitals NHS Trusts. All patients gave informed consent. Real-time PCR. Real Time PCR was performed using Taqman chemistry with Taqman Gene Expression 2X master mix (Applied Biosystems, Warrington, UK) and a 7500 real time PCR system (Applied Biosystems) as previously reported (19). The following primer and probe sets were used: human uPA forward primer 5'-ggaaaacctcatcctacacaagga, reverse primer 5'cggatcttcagcaaggcaat, probe 5'-ctgacacgcttgctcaccacaacga; PAI-1 forward primer 5'-ggctgacttcacgagtctttcag, reverse primer 5'-cgttcacctcgatcttcactttc, probe 5'-aagagcctctccacgtcgcgca; uPAR forward primer 5’cgaggttgtgtgtgggttagac, reverse primer 5’-ggcttcgggaataggtgaca, probe 5’tgcaaccagggcaactctggcc; GAPDH forward primer 5'-gctcctcctgttcgacagtca, reverse primer 5'-accttccccatggtgtctga, probe 5'-cgtcgccagccgagccaca, and an 18S rRNA control kit (Eurogentec, Southampton, UK). uPA, uPAR and PAI-1 mRNA abundances were determined relative to 18S and/or GAPDH in the same sample as appropriate. Gastrin radioimmunoassay. Plasma samples were assayed for total amidated gastrin concentrations by radioimmunoassay using antibody L2 which reacts Nørsett et al., page 7 at the C-terminus of G17 and measures G17, G34 and minor components such as G14 with similar affinity (8). Immunohistochemistry. Formalin-fixed, paraffin-embedded tissue sections were processed for immunohistochemical detection of PAI-1 (American Diagnostica Inc. Greenwich, CT, USA), chromogranin A (Dako UK Ltd, Ely, UK) and synaptophysin (Leica Microsystems, Milton Keynes, UK) as previously described (19). In control sections, the primary antibody was omitted. Luciferase promoter-reporter assays. Gastric glands were prepared from gastroscopic biopsies of control subjects as previously described (38) and transfected with PAI-1-luc according to the method used previously (19). AGS-GR cells (2 X 10) were plated in medium containing 10% bovine serum (full medium). The following day, medium was removed and cells were cotransfected with PAI-1-luc (0.25 μg/well) together with a constitutively active Renilla luciferase reporter (phRL-SV40, 0.5 ng/well, Promega) by use of TransFast (Promega) in serum-free medium for 1 h. After transfection, 2 ml of full medium was added and 20-24 h later cells were transferred to serum-free medium with G17 and other compounds as indicated in the text for a further 6 h. Luciferase activity was then measured by dual luciferase assay (Promega) according to the manufacturer’s instructions in a Lumat LB9507 luminometer (Berthold Technologies, Redbourne, Herts, UK). Results are presented as fold increase over unstimulated control, so the value of 1.0 signifies no change in luciferase activity. Nørsett et al., page 8 Migration and invasion assays. Migration and invasion assays were performed in 24-well plates containing either 8-μm pore Biocoat control inserts (migration assay) or Matrigel-coated inserts (invasion assays) according to the manufacturer’s instructions (Becton Dickinson, Bedford, MA). AGS-GR cells (10,000 cells/well for migration assays; 20,000 cells/well for invasion assays) in 0.5 ml serum-free medium were plated in the insert with or without addition of G17 for 18 h (migration) or 24 h (invasion). Cells invading the lower surface of the membrane were stained with Reastain Quick-Diff reagents (Reagena, Finland). The total cells in 5 fields per membrane were counted, and the mean of 3 membranes per experiment was taken. Adhesion assays. AGS-GR cells (1 X 10) were plated with full media in 24well plates and incubated with or without G17 for 45 min at 37°C. Media and non-adherent cells were then removed, cells were washed three times with PBS, and adherent cells were stained with 0.02% crystal violet. Adherent cells were then washed and solubilized with 2 mM Na2HPO4-50% ethanol, and absorbance was measured at 550 nm using a SpectraCount plate reader (Packard BioSience, Pangbourne, UK). Knockdown of PAI-1 by siRNA. Transfection of siRNA for PAI-1, or control siRNA, was performed on cells in suspension using an Amaxa Nucleofection Apparatus (Amaxa, Koln, Germany), solution V, program B023, according the manufacturer’s instruction. Knockdown was verified by qPCR and Western blotting. Nørsett et al., page 9 Western Blots. Protein extracts of AGS-GR cells were prepared and Western blotting was performed as previously described (22). Samples were probed with antibodies to PAI-1 (GeneTex Inc. Irvine, CA, USA) followed by HRPconjugated secondary antibody and detection by incubation with SuperSignal® West Pico Chemiluminescent Substrate (Pierce, Illinois, USA) and HyperFilm (Amersham) as previously described (22). Samples were reprobed for GAPDH (Biodesign, Saco, Maine, USA). Statistics. Results are presented as mean ± SE; comparisons were made by Student’s t-test or Anova as appropriate and were considered significant at P < 0.05. Nørsett et al., page 10 RESULTS Increased gastric PAI-1 in hypergastrinemia. In subjects with plasma gastrin concentrations that exceeded the reference range (>30pM), PAI-1 mRNA abundance was increased approximately 2-fold compared to subjects with plasma concentrations within the reference range (Fig. 1A,B). These subjects were H. pylori negative and so the increase was not attributable to Helicobacter infection which has previously been associated with increased expression of members of the uPA system (19). Because PAI-1 interacts with both uPA and uPAR we also analysed these in the same samples, and in both cases there were increases suggesting that the system as a whole was gastrin-regulated (Fig. 1B). To further explore the idea that increased gastrin might control expression of this system, we then examined mRNA abundance of PAI-1 in patients with hypergastrinemia due to chronic atrophic gastritis. Circulating gastrin in these patients was typically 10 to 100-fold elevated compared with normal subjects, and PAI-1 mRNA abundance was elevated approximately 10-fold (Fig. 1C,D). Again, there were also significant increases in uPA and uPAR mRNA (Fig. 1D). Decreased gastric PAI-1 mRNA after octreotide or antrectomy. In PA patients with ECL cell neuroendocrine nodules, 3-day infusion of octreotide was employed to establish suitability for subsequent antrectomy as previously described (13). After octreotide, there was decreased plasma gastrin. In biopsies taken from corpus mucosa that did not include ECL cell nodules, there was also decreased HDC mRNA abundance used as a marker for ECL Nørsett et al., page 11 cell function (Fig. 2A,B). In addition, octreotide was associated with a decrease of approximately 70% in PAI-1 mRNA in these biopsies and with a significant but slightly smaller decrease in uPA mRNA, compatible with acute endocrine regulation of these transcripts (Fig. 2C,D). Moreover, 12 months after antrectomy, plasma gastrin was reduced to concentrations within or close to the reference range and HDC mRNA was also significantly reduced (Fig. 2A,B). In these circumstances there was a 70% reduction in PAI-1 mRNA abundance (Fig. 2D). There was also decreased abundance of uPA mRNA which although statistically significant was less than that seen for PAI1 and HDC mRNAs (Fig. 2C). Cellular localisation of PAI-1. In immunohistochemistry, PAI-1 was found to be localised to parietal cells in hypergastrinemic patients without chronic atrophy confirming the localisation reported previously (Fig. 3A)(19). In PA patients with micronodular ECL cell tumors, there was clear localisation of PAI-1 to ECL cell micronodules (Fig. 3B), characterised by chromogranin A and synaptophysin staining (Fig. 3C,D). In addition, weak staining for PAI-1 was observed in the stroma (Fig. 3B) but staining in control sections was negative (Fig. 3E). Gastrin increases PAI-1-luc expression via Protein Kinase C and MitogenActivated Protein Kinase pathways. In glands cultured from gastroscopic biopsies and transfected with PAI-1-luc, there is known to be expression in parietal and ECL cells (19). Treatment of transfected glands with G17 (1nM, 16 h) significantly increased PAI-1-luc expression (control: 1.0 ± 0.1; G17: Nørsett et al., page 12 2.97 ± 1.1, n=3 independent experiments, p<0.05) and there was comparable stimulation by IGF-II (control: 1.0 ± 0.1; IGF-II 100 ng/ml: 2.46 ± 0.7, n=4 independent experiments, p<0.05). Previous DNA array studies identified increased PAI-1 mRNA in AGS-GR cells treated with gastrin (33) and in the present study this was confirmed by qPCR (see below) indicating that these cells are an appropriate model for studies of the cellular mechanisms by which gastrin might induce PAI-1 gene transcription. In AGS-GR cells that express the CCK2 receptor and were transfected with PAI-1-luc, gastrin induced a concentration dependent increase in luciferase expression over the range 0.3 to 10 nM that was maximal 6 h after stimulation (Fig. 4A,B). The response to 1nM gastrin was inhibited by Ro-32-0432 which is an inhibitor of PKC; moreover, stimulation of PKC with PMA produced a robust increase in PAI-1-luc expression that was inhibited by Ro-32-0432 (Fig. 4C,D). In addition, the response to both gastrin and PMA was partially reversed by U0126, which inhibits the activation of MAPkinasekinase (MEK)-1 and -2 (Fig. 4C,D) indicating that the activation of MAPkinase was downstream of PKC. No significant effect was seen with an inhibitor of MEK1 (PD98059) pointing to the involvement of MEK2 in the AGS-GR cell response. The PI-3 kinase inhibitor LY204002 had no effect on PAI-1-Luc expression (Fig. 4C). PAI-1 inhibits AGS-GR cell migration and invasion. Since the uPA system is implicated in control of cell migration and invasion we examined the role of the system on G17-stimulated migration and invasion of AGS-GR cells in Boyden chambers. There was stimulation of AGS-GR cell migration and invasion in response to G17 as previously noted (35); in the presence of PAI-1 (40nM) Nørsett et al., page 13 the migration and invasion responses to G17 were 56.6 ± 11.3 % and 34.3 ± 1.7 %, respectively, of those to G17 alone (p<0.05 in both cases). We considered the possibility that induction of PAI-1 might also influence gastrin stimulated cell adhesion (26); however, while acute treatment of AGS-GR cells with 1 nM G17 increased adhesion in a dose-dependent manner, in conditions that increased PAI-1 and approximated to those used for migration and invasion assays (ie pretreatment with G17, 1nM, 18h) there was no change in the adhesion response compared with acute stimulation suggesting that gastrin does not stimulate migratory and invasion responses as a consequence of long term changes in cell adhesion (Fig. 5). To address the significance of gastrin-stimulated PAI-1 expression in AGS-GR cell motility, we used siRNA to depress PAI-1 expression. After treatment with siRNA, gastrin-stimulated increases in PAI-1 mRNA abundance were virtually abolished (Fig. 6A). There was also substantial reduction in PAI-1 detected by Western blot, but no change in GAPDH abundance indicating the specificity of the knockdown (Fig. 6B). In cells with PAI-1 knockdown, there were no differences in basal rates of migration and invasion. However there was significantly increased migration and invasion in response to an optimal concentration of G17 compared with cells treated with control siRNA (Fig. 7A,B). Nørsett et al., page 14 DISCUSSION The present study shows that increased plasma gastrin concentrations are associated with increased expression of PAI-1 in the stomach. There are also changes in expression of other members of the uPA system in gastric mucosa suggesting that the system as a whole is up-regulated. The cellular control mechanisms involved in PAI-1 expression include the PKC and MAPkinase pathways. There is increased cell migration and invasion in CCK2-expressing AGS cells in response to gastrin, and this is increased by siRNA knockdown of PAI-1 expression indicating that induction of PAI-1 plays a role in restraining cell migration and invasion and may therefore be considered to have a protective role in hypergastrinemia. There are multiple different serpine inhibitors of plasminogen activators (PAI-1, PAI-2, PAI-3 and protease nexin). Previous work has shown gastric PAI-2 to be increased by gastrin and H.pylori infection (33; 35); however, PAI-2 possesses a relatively weak signal sequence and at least some effects eg inhibition of apoptosis, appear to be mediated intracellularly. Although we previously found PAI-3 in stomach (19), we found no evidence of regulated expression. In contrast, PAI-1 mRNA abundance increases with raised plasma gastrin in Hpylori-negative subjects and in PA subjects with profound hypergastrinemia, and is decreased in the latter after treatments that lower plasma gastrin (octreotide infusion and antrectomy). Together with the results of promoterreporter experiments these observations indicate gastrin regulates PAI-1 expression in the stomach. Nørsett et al., page 15 It is well recognised that members of the uPA system are expressed in a wide variety of cell types which in the case of hollow organs can include both epithelial and stromal cells. In the stomach, the CCK2-receptor is expressed by both parietal and ECL cells. The localisation of PAI-1 and uPA to parietal cells is compatible with a direct role for gastrin in stimulating their expression. In the case of ECL cells, PAI-1 expression was clearest in micronodular ECL cell neuroendocrine tumors, again compatible with a direct effect of gastrin (albeit in the high concentrations characteristic of patients with these tumors). Previous studies have shown that gastrin stimulates the expression of several genes in gastric epithelial cells including HDC (14; 28), PAI-2 (33) VMAT2 (37), chromogranin A (6; 15) and trefoil factor-1 (20). In all these cases there is evidence for the involvement of PKC and MAPkinase pathways; the present data suggest activation of similar pathways leads to PAI-1 expression in gastric epithelial cells, and this is also compatible with studies of PAI-1 expression in a variety of other cell types including neurons (27) and vascular smooth muscle cells (32). However, in addition to expression in gastric epithelial cells, there is also PAI-1 expression in gastric stromal cells. These are not thought to express the CCK2 receptor under physiological conditions (2). Instead, it seems possible that gastrin exerts indirect effects on these cells. Studies of gastrin-activated paracrine mechanisms in AGS cell populations have identified both IL-8 and prostaglandins as mediating expression of PAI-2 (1; 35). In addition, there is evidence that gastrin activates epithelial-mesenchymal signalling pathways. Thus gastrin stimulates gastric epithelial MMP-7 expression and release that in Nørsett et al., page 16 turn cleaves insulin-like growth factor binding protein (IGFBP)-5 produced by myofibroblasts to liberate IGF which acts as an autocrine stimulant of myofibroblasts or as a paracrine stimulant of epithelial cells (12; 22). In support of this suggestion it should be noted that IGF-I has also been shown to stimulate PAI-1 expression in HepG2 cells (29) and IGF-II stimulated PAI-1-luc expression in gastric glands. The expression of PAI-1 in cancer, including gastric cancer, is associated with poor prognosis (3; 9; 16; 17). This is sometimes referred to as paradoxical given that uPA is also associated with poor prognosis and it might be expected that inhibition of uPA by PAI-1 would be beneficial. However, it is recognised that PAI-1 has both uPA dependent and independent effects, and that it can exert different actions with both beneficial and adverse consequences (21). Thus in addition to inhibition of uPA, PAI-1 also binds vitronectin and disrupts the binding of vitronectin to integrins eg αvβ3 (31; 40). These actions lead to decreased cell adhesion and inhibition of migration. Although gastrin acutely stimulates cell adhesion, we did not find evidence for changes in gastrinstimulated adhesion over a time course compatible with induction of PAI-1. In contrast, PAI-1 inhibited gastrin-stimulated cell migration and invasion, and suppression of PAI-1 synthesis by siRNA treatment led to increases in gastrinstimulated cell migration and invasion, suggesting that induction of PAI-1 by gastrin normally restrains its actions on cell motility. The progression to ECL cell neuroendocrine tumors in the stomach is driven by gastrin and enhanced by mutations of the MEN-1 gene, or mucosal Nørsett et al., page 17 inflammation. In general these tumors are not thought to be particularly aggressive and, for example, metastasis is a feature only of advanced stages of the disease. The factors governing progression remain poorly understood. However, the present study raises the possibility that some effects of gastrin may delay tumor progression. In particular, induction of PAI-1 may in this case limit cell migration and invasion, and so serve a protective function. We suggest that this system provides a useful model for further studies of the role of PAI-1 in cancer progression. Nørsett et al., page 18 Acknowledgements This work was supported in part by grants from the Medical Research Council and the NIHR-funded Liverpool Biomedical Research Centre for Microbial Diseases. JS was supported by the Wolfson Foundation and KGN by the Faculty of Medicine, Norwegian University of Science and Technology, Trondheim, Norway. We thank David Berry for assistance with immunohistochemistry and members of the Mersey Neuroendocrine Tumour Multidisciplinary team and Gastroenterology Unit at Royal Liverpool University Hospital for clinical assistance. Nørsett et al., page 19 References1. Almeida-Vega S, Catlow K, Kenny S, Dimaline R and Varro A.Gastrin activates paracrine networks leading to induction of PAI-2 viaMAZ and ASC-1. Am J Physiol Gastrointest Liver Physiol 296: G414-G423, 2009. 2. Ashurst HL, Varro A and Dimaline R. Regulation of mammaliangastrin/CCK receptor (CCK2R) expression in vitro and in vivo. ExpPhysiol 93: 223-236, 2008. 3. Beyer BC, Heiss MM, Simon EH, Gruetzner KU, Babic R, JauchKW, Schildberg FW and Allgayer H. Urokinase system expression ingastric carcinoma: prognostic impact in an independent patient seriesand first evidence of predictive value in preoperative biopsy andintestinal metaplasia specimens. Cancer 106: 1026-1035, 2006. 4. Bordi C, D'Adda T, Azzoni C, Pilato FP and Caruana P.Hypergastrinemia and gastric enterochromaffin-like cells. Am J SurgPathol 19 Suppl 1: S8-19, 1995. 5. Burkitt MD and Pritchard DM. Review article: Pathogenesis andmanagement of gastric carcinoid tumours. Aliment Pharmacol Ther 24:1305-1320, 2006. Nørsett et al., page 20 6. Cramer T, Juttner S, Plath T, Mergler S, Seufferlein T, Wang TC,Merchant J and Hocker M. Gastrin transactivates the chromogranin Agene through MEK-1/ERKand PKC-dependent phosphorylation ofSp1 and CREB. Cell Signal 20: 60-72, 2008. 7. Dockray G, Dimaline R and Varro A. Gastrin: old hormone, newfunctions. Pflugers Arch 449: 344-355, 2005. 8. Dockray GJ, Hamer C, Evans D, Varro A and Dimaline R. Thesecretory kinetics of the G cell in omeprazole-treated rats.Gastroenterology 100: 1187-1194, 1991. 9. Farinati F, Herszenyi L, Plebani M, Carraro P, De PM, Cardin R,Roveroni G, Rugge M, Nitti D, Grigioni WF, D'Errico A andNaccarato R. Increased levels of cathepsin B and L, urokinase-typeplasminogen activator and its inhibitor type-1 as an early event ingastric carcinogenesis. Carcinogenesis 17: 2581-2587, 1996. 10. Gilligan CJ, Lawton GP, Tang LH, West AB and Modlin IM. Gastriccarcinoid tumors: the biology and therapy of an enigmatic andcontroversial lesion. Am J Gastroenterol 90: 338-352, 1995. 11. Heiss MM, Babic R, Allgayer H, Gruetzner KU, Jauch KW, LoehrsU and Schildberg FW. Tumor-associated proteolysis and prognosis:new functional risk factors in gastric cancer defined by the urokinase-type plasminogen activator system. J Clin Oncol 13: 2084-2093, 1995. Nørsett et al., page 21 12. Hemers E, Duval C, McCaig C, Handley M, Dockray GJ and VarroA. Insulin-like growth factor binding protein-5 is a target of matrixmetalloproteinase-7: implications for epithelial-mesenchymal signaling.Cancer Res 65: 7363-7369, 2005. 13. Higham AD, Dimaline R, Varro A, Attwood S, Armstrong G,Dockray GJ and Thompson DG. Octreotide suppression test predictsbeneficial outcome from antrectomy in a patient with gastric carcinoidtumor. Gastroenterology 114: 817-822, 1998. 14. Hocker M, Henihan RJ, Rosewicz S, Riecken EO, Zhang Z, Koh TJand Wang TC. Gastrin and phorbol 12-myristate 13-acetate regulatethe human histidine decarboxylase promoter through Raf-dependentactivation of extracellular signal-regulated kinase-related signalingpathways in gastric cancer cells. J Biol Chem 272: 27015-27024, 1997. 15. Hocker M, Raychowdhury R, Plath T, Wu H, O'Connor DT,Wiedenmann B, Rosewicz S and Wang TC. Sp1 and CREB mediategastrin-dependent regulation of chromogranin A promoter activity ingastric carcinoma cells. J Biol Chem 273: 34000-34007, 1998. 16. Ito H, Yonemura Y, Fujita H, Tsuchihara K, Kawamura T, Nojima N,Fujimura T, Nose H, Endo Y and Sasaki T. Prognostic relevance ofurokinase-type plasminogen activator (uPA) and plasminogen activatorinhibitors PAI-1 and PAI-2 in gastric cancer. Virchows Arch 427: 487-496, 1996. Nørsett et al., page 22 17. Kaneko T, Konno H, Baba M, Tanaka T and Nakamura S.Urokinase-type plasminogen activator expression correlates with tumorangiogenesis and poor outcome in gastric cancer. Cancer Sci 94: 43-49, 2003. 18. Keates AC, Tummala S, Peek RM, Jr., Csizmadia E, Kunzli B,Becker K, Correa P, Romero-Gallo J, Piazuelo MB, Sheth S, KellyCP, Robson SC and Keates S. Helicobacter pylori infection stimulatesplasminogen activator inhibitor 1 production by gastric epithelial cells.Infect Immun 76: 3992-3999, 2008. 19. Kenny S, Duval C, Sammut SJ, Steele I, Pritchard DM, AthertonJC, Argent RH, Dimaline R, Dockray GJ and Varro A. Increasedexpression of the urokinase plasminogen activator system byHelicobacter pylori in gastric epithelial cells. Am J Physiol GastrointestLiver Physiol 295: G431-G441, 2008. 20. Khan ZE, Wang TC, Cui G, Chi AL and Dimaline R. Transcriptionalregulation of the human trefoil factor, TFF1, by gastrin1.Gastroenterology 125: 510-521, 2003. 21. Lijnen HR. Pleiotropic functions of plasminogen activator inhibitor-1. JThromb Haemost 3: 35-45, 2005. 22. McCaig C, Duval C, Hemers E, Steele I, Pritchard DM, Przemeck S,Dimaline R, Ahmed S, Bodger K, Kerrigan DD, Wang TC, Dockray Nørsett et al., page 23 GJ and Varro A. The role of matrix metalloproteinase-7 in redefiningthe gastric microenvironment in response to Helicobacter pylori.Gastroenterology 130: 1754-1763, 2006. 23. Nakamura M, Konno H, Tanaka T, Maruo Y, Nishino N, Aoki K,Baba S, Sakaguchi S, Takada Y and Takada A. Possible role ofplasminogen activator inhibitor 2 in the prevention of the metastasis ofgastric cancer tissues. Thromb Res 65: 709-719, 1992. 24. Nekarda H, Schmitt M, Ulm K, Wenninger A, Vogelsang H, BeckerK, Roder JD, Fink U and Siewert JR. Prognostic impact of urokinase-type plasminogen activator and its inhibitor PAI-1 in completelyresected gastric cancer. Cancer Res 54: 2900-2907, 1994. 25. Norsett KG, Laegreid A, Kusnierczyk W, Langaas M, Ylving S,Fossmark R, Myhre S, Falkmer S, Waldum HL and Sandvik AK.Changes in gene expression of gastric mucosa during therapeutic acidinhibition. Eur J Gastroenterol Hepatol 20: 613-623, 2008. 26. Pagliocca A, Wroblewski LE, Ashcroft FJ, Noble PJ, Dockray GJand Varro A. Stimulation of the gastrin-cholecystokinin(B) receptorpromotes branching morphogenesis in gastric AGS cells. Am J Physiol283: G292-G299, 2002. 27. Presta M, Ennas MG, Torelli S, Ragnotti G and Gremo F. Synthesisof urokinase-type plasminogen activator and of type-1 plasminogen Nørsett et al., page 24 activator inhibitor in neuronal cultures of human fetal brain: stimulationby phorbol ester. J Neurochem 55: 1647-1654, 1990. 28. Raychowdhury R, Zhang Z, Hocker M and Wang TC. Activation ofhuman histidine decarboxylase gene promoter activity by gastrin ismediated by two distinct nuclear factors. J Biol Chem 274: 20961-20969, 1999. 29. Schneider DJ and Sobel BE. Augmentation of synthesis ofplasminogen activator inhibitor type 1 by insulin and insulin-like growthfactor type I: implications for vascular disease in hyperinsulinemicstates. Proc Natl Acad Sci U S A 88: 9959-9963, 1991. 30. Smith HW and Marshall CJ. Regulation of cell signalling by uPAR.Nat Rev Mol Cell Biol 11: 23-36, 2010. 31. Stefansson S and Lawrence DA. The serpin PAI-1 inhibits cellmigration by blocking integrin alpha V beta 3 binding to vitronectin.Nature 383: 441-443, 1996. 32. Takeda K, Ichiki T, Tokunou T, Iino N, Fujii S, Kitabatake A,Shimokawa H and Takeshita A. Critical role of Rho-kinase andMEK/ERK pathways for angiotensin II-induced plasminogen activatorinhibitor type-1 gene expression. Arterioscler Thromb Vasc Biol 21:868-873, 2001. Nørsett et al., page 25 33. Varro A, Hemers E, Archer D, Pagliocca A, Haigh C, Ahmed S,Dimaline R and Dockray GJ. Identification of plasminogen activatorinhibitor-2 as a gastrinregulated gene: Role of Rho GTPase andmenin. Gastroenterology 123: 271-280, 2002. 34. Varro A, Kenny S, Hemers E, McCaig C, Przemeck S, Wang TC,Bodger K and Pritchard DM. Increased gastric expression of MMP-7in hypergastrinemia and significance for epithelial-mesenchymalsignaling. Am J Physiol Gastrointest Liver Physiol 292: G1133-G1140,2007. 35. Varro A, Noble PJ, Pritchard DM, Kennedy S, Hart CA, Dimaline Rand Dockray GJ. Helicobacter pylori induces plasminogen activatorinhibitor 2 (PAI-2) in gastric epithelial cells through NF-kB and RhoA:implications for invasion and apoptosis. Cancer Res 64: 1695-1702,2004. 36. Varro A, Noble PJ, Wroblewski LE, Bishop L and Dockray GJ.Gastrin-cholecystokinin(B) receptor expression in AGS cells isassociated with direct inhibition and indirect stimulation of cellproliferation via paracrine activation of the epidermal growth factorreceptor. Gut 50: 827-833, 2002. 37. Watson F, Kiernan RS, Deavall DG, Varro A and Dimaline R.Transcriptional activation of the rat vesicular monoamine transporter 2 Nørsett et al., page 26 promoter in gastric epithelial cells: regulation by gastrin. J Biol Chem276: 7661-7671, 2001. 38. Wroblewski LE, Noble PJ, Pagliocca A, Pritchard DM, Hart CA,Campbell F, Dodson AR, Dockray GJ and Varro A. Stimulation ofMMP-7 (matrilysin) by Helicobacter pylori in human gastric epithelialcells: role in epithelial cell migration. J Cell Sci 116: 3017-3026, 2003. 39. Wroblewski LE, Pritchard DM, Carter S and Varro A. Gastrin-stimulated gastric epithelial cell invasion: the role and mechanism ofincreased matrix metalloproteinase 9 expression. Biochem J 365: 873-879, 2002. 40. Zhou A, Huntington JA, Pannu NS, Carrell RW and Read RJ. Howvitronectin binds PAI-1 to modulate fibrinolysis and cell migration. NatStruct Biol 10: 541-544, 2003. LEGENDSFIG. 1. Increased PAI-1, uPA and uPAR mRNA in the gastric corpus ofhypergastrinemic subjects. A, Subjects who were H.pylori negative weredivided into two groups based on fasting plasma gastrin concentrations eitherwithin the reference range (<30 pM) or above it. B, The relative abundance ofPAI-1, uPA and UPAR mRNAs determined by qPCR in patients with plasmagastrin within or above the reference range (*, p<0.05, n=25-28).C, Plasma gastrin was elevated in PA patients. D, and there were significantincreases in PAI-1, uPA and uPAR mRNA abundance (*, p<0.05, n=7-10). Nørsett et al., page 27 FIG. 2. Decreased PAI-1, uPA and uPAR mRNA after octreotide orantrectomy in the gastric corpus of PA patients. PA patients receivedoctreotide (25 μg/h, 3 days), plasma and endoscopic biopsies from corpusmucosa unaffected by macroscopic nodules , were taken before and at theend of the infusion period. Subsequently, the patients underwent antrectomyfor treatment of gastric neuroendocrine tumours and plasma and gastricbiopsies were taken 12 months post operatively. After both treatments therewere decreases in A, plasma gastrin, B, HDC, C, uPA and D, PAI-1 (*,p<0.05, vs the relevant control, n= 5). Fig 3. Immunohistochemical localisation of PAI-1 to parietal cells andmicronodular hyperplastic ECL cells. A, In a hypergastrinemic patient, PAI-1 islocalised to parietal cells (arrow) in a gastric biopsy from a region notexhibiting atrophy. B, In a hypergastrinemic patient with micronodular ECL cellhyperplasia, PAI-1 is localised to the nodules, and there is also light stainingof surrounding stroma (arrows). C,D, In the same patient there is localisationof chromogranin A and synaptophysin, respectively, to nodules (note B-D arenot consecutive sections). E, Control section without primary antibody.Magnification x20. Fig. 4. Gastrin increases PAI-1-luc expression in AGS-GR cells. A,Progressive increase in expression of 4.5 kb PAI-1-luc with concentrations ofG17 from 0.1 nM to 10 nM. B, The expression of PAI-1-luc in response to 1nM G17 was maximal after 6 hours. C, Inhibition of G17-stimulation of PAI-1Nørsett et al., page 28 luc by the MEK inhibitor U0126 (10 μM) and the PKC inhibitor Ro-320432 (1μM). In the presence of the PI-3 kinase inhibitor LY204002 (50 μM) and MEK1inhibitor PD98059 (20 μM) the relative change in expression following G17-stimulated was not significantly different to control. D, PMA (100nM)stimulated PAI-1-luc expression and this was inhibited by Ro-320432 andU0126. * P <0.05, n=3. Fig. 5. Increased AGS-GR cell adhesion by acute treatment with gastrin, butno additional effect with prolonged treatment. Cell adhesion over 45 min wasincreased in the presence of G17 (1nM). When cells were pretreated with 1nM G17 for 18 h to stimulate gene expression there was no change gastrin-stimulated adhesion. N=3, * P<0.05 vs the appropriate control. Fig. 6. Knockdown of PAI-1 mRNA by PAI-1 siRNA treatment. A, Increasedexpression of PAI-1 determined by qPCR in response to gastrin (1nM) in thepresence of control siRNA. In cells treated with PAI-1 siRNA there is profoundinhibition of the gastrin-stimulated increase in PAI-1 transcript abundance. B,Western blot of cells treated with G17 (1nM), show decreased abundance ofPAI-1 protein after siRNA treatment compared with control siRNA, but nochange in GAPDH. Fig. 7. Gastrin-stimulated AGS-GR cell migration and invasion is increasedafter PAI-1 knockdown. A, Migration of unstimulated AGS-GR cells throughBoyden chambers is stimulated by gastrin (0.3 nM; 18 hours) and theresponse is increased in PAI-1 siRNA-treated cells. B, Invasion of AGSNørsett et al., page 29 GR.cells in Matrigel-coated Boyden chambers is virtually undetectable incontrol conditions, but is stimulated by G17 (0.3 nM, 24 h) and this is furtherincreased by treatment of cells with PAI-1 siRNA. In each case, the sum ofthe migrating or invading cells in five fields is presented; horizontal doubleheaded arrows indicate significant differences between the bars indicated,p<0.05, n=3.