- Email: [email protected]
Transforming growth factor-α-mediated epithelial-mesenchymal interaction promotes fetal gastric epithelial cell growth
April 1998 which corresponded to a 19-fold decrease in PHAS-I expression. This observation suggests a role for putrescine in negative feedback control of polyamine biosynthesis by regulation of PHAS-I expression. The distribution of the PHAS-I message was determined in acutely isolated rat small intestinal epithelial cells. The PHAS-I transcript was identified in cells isolated from mid-villus of duodenum, jejenum and ileum. The crypt-villus axis was also examined. The PHAS-I transcript was expressed to a similar magnitude in all fractions from crypt to villus tip. This data demonstrate that PHAS-I is expressed in gut epithelium and may be involved in polyamine metabolism. • G4686
TRANSFORMING GROWTH FACTOR-a-MEDIATED EPITHELIALMESENCHYMAL INTERACTION PROMOTES FETAL G A S T R I C EPITHELIAL CELL GROWTH. RE Glasgow, BC Visser, SJ Mulvihill. Department of Surgery, UCSF. San Francisco, California, Epithelial-mesenchymal interactions play an important role in the regulation of epithelial development. Transforming growth factor-alpha (TGF-ct) is a potent growth factor for adult gastrointestinal epithelium. Although TGF-a and its receptor, epidermal growth factor receptor (EGF-R), have been identified in the fetal stomach, their role in development is not known. We hypothesized that TGF-a is a mitogen for fetal gastric epithelial cells and mediates epithelial-mesenchymal interactions in the developing fetal stomach. Fetal rabbit gastric tissue sections, cultured fetal rabbit gastric epithelial cells, and, separately, cultured fetal gastric mesenchymal cells from day 21 (term is 31 days) gestation New Zealand White rabbit fetuses were assayed for TGF-a and EGF-R by immunofiuorescence. Fetal gastric epithelial cell culture was used to evaluate EGF-R binding and proliferation in response to TGF-a. A co-culture model was used to identify mesenchymal cell influence on epithelial cell proliferation. Data was collected in quadruplicate using tissue pooled from 8-10 fetuses from a single litter. Each experiment was repeated in at least 3 separate litters. Statistical analysis is by Student's t test. Primary fetal gastric epithelial cell culture was established with 98% purity as measured by immunochemistry using antibodies for cytokeratin AEI:AE3 and epithelial specific antigen. Immunochemistry showed TGF-a and EGF-R expression in the epithelium of gastric tissue sections and gastric epithelial cells in culture. TGF-a immunofluorescence was also seen in mesenchymal ceils. Specific EGF-R binding was seen in epithelial cells by 125I-TGF-a receptor binding studies. Graded concentrations (10-13 to 10.8 M) of TGF-a caused a dose-dependent increase in 3H-thy uptake (basal, 394 _+31 cpm, vs. 10-8 M, 1796_+ 69 cpm, p<0.001). Similar effects were shown by MTS proliferation assay and cell count. These increases were blocked by an antiEGF-R antibody (Ab). Epithelial ceils co-cultured with gastric mesenchymal cells showed increased 3H-thy uptake (basal, 1.8_+0.1x104 cpm, vs. coculture, 4.0 _+0.2x104 cpm, p<0.05). This was attenuated by 59% with a neutralizing anti-TGF-a Ab and 36% with an anti-EGF-R Ab (p<0.05). Similar effects were shown with cell counts. We conclude that TGF-a is expressed in both fetal rabbit gastric epithelial and mesenchymal cells. TGF-a is a mitogen for fetal rabbit gastric epithelial cells and may act as an autocrine growth factor and/or by paracrine release from mesenchymal cells during development. • G4687
EXPRESSION OF FUNCTIONAL GABAA RECEPTORS IN ISOLATED HUMAN INSULINOMA CELLS. G. Glassmeier 1, M. H~pfner t, H. Buhr2, S. Niedrig 1, E.-O. Riecken l, H. Stein 3, H.-J. Quabbe4, C. Rancso 3, B. Wiedenmann 1, H. Schertibl l. 1Abt. Innere Medizin/Gastroenterologie, 2 Abt. Allgemein-, Gef'~- und Thoraxchirurgie, 3 Abt. Allgemeine Pathologie, 4Abt. Innere Medizin/Endokrinologie, Universit~itsklinikum Benjamin Franklin, Freie Universit~it Berlin, Berlin, Germany. Pancreatic islets contain and release high concentrations of ~'-aminobutyic acid (GABA). GABA is thought to play a paracrine role in 13-cells. Searching for a paracrine function of GABA in neoplastic I~-cells we performed patchclamp studies in freshly isolated human insulinoma cells. We show that human insulinoma cells can express functional GABAA receptors. Application of GABA or the specific GABA A receptor agonist muscimol (10 -6 to 10-3 M) dose-dependently activated inward currents at negative holding potentials and outward currents at positive membrane potentials. The amplitudes of the GABA (0.5 mM) induced currents ranged between -35 and -55 pA (n=7) at a holding potential of -80 mV. The reversal potential of the GABA-induced current was determined by digital subtraction of current traces elicited in the absence or presence of GABA with a voltage ramp command. The current reversed close to the CI- equilibrium potential under the recording conditions used. Bath application of the GAB AA receptor antagonist picrotoxin (100 [aM) reversibly reduced the amplitude of GABA-evoked currents by 65 _+6% (n=4). To study voltage-gated calcium channel currents in the isolated ceils, the standard bath solution was replaced by a sodium- and potassium-free solution and barium (10.8 mM) was used as charge carrier. Under these conditions, dihydropyridine-sensitive inward currents were elicited in all cells in response to 100 ms pulses to various test potentials starting from a holding potential of -80 mV (n = 11). Using the perforated-patch technique, activation of GABA A receptors caused a reversible membrane depolarization in the insulinoma ceils. Membrane depolarization resulted in transmembraneous calcium influx through voltage-gated calcium channels. In accordance with the electrophysiological findings activation of GABA A receptors by muscimol increased insulin secretion. Insulin secretion was increased by 50 laM muscimol by about 280%. Thus, GABA A receptors can be expressed in human insulinoma cells and may regulate their insulin release.
Hormones, Transmitters, Growth Factors, and their Receptors Al145
The project was supported by the Mildred Scheel Stiftung and the Deutsche Forschungsgemeinschaft (Sche 326/3-1). G4688 EFFECTS OF PROTEIN ON GASTRIC INHIBITORY POLYPEPTIDE
(GIP) EXPRESSION IN THE RAT. K.D. Glazier, K.-B. Zhan, L.A. Jarboe, C.-C. Tseng, and M.M. Wolfe. Section of Gastroenterology, Boston University School of Medicine and Boston Medical Center, Boston, MA. GIP is a 42-amino acid regulatory peptide originally named on the basis of its ability to inhibit gastric acid secretion. The release of GIP into the circulation has been demonstrated primarily after the ingestion of two major nutrient stimuli, carbohydrates and fat. The present studies were directed to examine nutrient protein regulation of GIP expression in vivo in the rat. Rats fasted overnight were submitted to midline laparotomy, at which time they were perfused with 0.9% NaCI (control) or 10% beef protein hydrolysate (Peptone). Animals were then sacrificed at various time points, and GIP mRNA was measured by Northern blot hybridization and tissue and circulating GIP levels determined by RIA. In response to peptone infusion, duodenal mucosal GIP concentration increased from 8.4 -+ 1.5 (mean -+ SE) to 19.8 -+3.2 ng GIP per mg protein at 120 rain (p<0.01). Plasma GIP levels also increased from 288 _ 20 to 736 -+ 191 at 60 min (p<0.05), to 693 _+ 135 at 120 min (p<0.01), and to 505-+ 62 pg/ml at 240 min (p<0.01). In contrast to the effects on mucosal and circulating GIP, no significant changes in GIP mRNA levels were detected in response to peptone. To determine whether effects of protein on GIP were due to stimulation of gastric acid secretion, rats were pretreated with 25 mg/kg omeprazole, a dose previously shown to abolish acid output in the rat. Following treatment with omeprazole, peptone-stimulated mucosal and plasma GIP concentrations were partially attenuated at 120 min: 12.0 _+3.8 ng GIP per mg tissue protein and 487 _+ 108 pg/ml, respectively. To further examine the effects of luminal acid, anesthetized rats were administered 0.1 M HCI by intraduodenal infusion for 120-min. In addition to stimulatory effects on mucosal and circulating GIP, duodenal acidification increased the GIP/actin mRNA ratio from 1.44±0.12 to 3.22_+0.36 at 30 min (p<0.0001). The results of these studies indicate that in addition to lipids and carbohydrates, nutrient protein provides a potent stimulus for GIP expression in the rat, an effect that occurs at the posttranslational level and may be mediated in part through the acid stimulatory properties of protein. Furthermore, the stimulation of GIP by duodenal acidification may involve pretranslational mechanisms. The effects of acid on GIP expression are consistent with a role for GIP as an enterogastrone in the rat. This research was funded by grants DK48042 and DK53158 from the NIH. G4689
SERUM CHROMOGRANIN A AS A MARKER OF DISEASE PRESENCE, EXTENT OR ACTIVITY IN PATIENTS WITH GASTRINOMA. S.U. Goebel, F. Gibril, F. Yu, J. Serrano, R.T. Jensen, NIH, Bethesda, MD. Chromogranin A (CgA) is a peptide that is released from the neuroendocrine granules Of many normal ceils and neuroendocrine tumors. Studies show that the serum values correlate with tumor presence and burden in patients with pheochromocytomas and carcinoids. Determining the extent and progression of the tumor in patients with pancreatic endocrine tumors (PET's) such as gastrinomas can be difficult and expensive. Hence, the purpose of the present study was to evaluate prospectively whether CgA levels correlated with these parameters in a large number of patients with gastrinoma and to compare CgA results with those of the specific tumor marker, gastrin. In 114 consecutive patients with gastrinoma, serum CgA levels and fasting serum gastrin levels were measured by well-characterized radioimmunoassays. Each was correlated with extent of disease, disease activity, the presence of multiple endocrine neoplasia-type 1 (MEN-I) or gastric carcinoid tumors. Overall, there was a good correlation between CgA levels drawn on 2 consecutive days in a representative sample of patients (n=51, r=0.88, p=0.0001). Disease-free patients post resection had significantly lower CgA and gastrin values than patients not disease-free (p <0.001). A significantly higher proportion of patients not disease-free had elevated CgA levels [>50 ng/ml (72/78=92%)] than patients disease-free [12/36=33% (p <0.0001)]. In comparison, 1/36 (3%) of patients disease-free had an elevated fasting gastrin level whereas 60/78 (77%) that were not disease-free had an elevated level. To facilitate correlations with disease extent, patients were divided into 4 categories consisting of patients with no liver metastases and stable disease (stable, localized) (n=54), no liver metastases but extrahepatic tumor growth (localized with growth) (n=7), liver metastases without growth (stable liver metastases) (n=6) and liver metastases with growth (liver metastases with growth) (n=10). Serum gastrin (p=0.04), but not CgA levels (p=0.96) differed significantly between patients with stable localized disease, compared to localized with growth. Neither serum gastrin nor CgA levels differed between patients with stable localized disease, stable liver metastases or with liver metastases with growth. Patients with gastric carcinoid tumors (n=8) had significantly higher CgA (p=0.04) and gastrin levels (p=0.001) compared to patients without gastric carcinoid tumors (n=104). Patients with or without MEN-I did not differ in serum gastrin (p=0.34) or CgA levels (p=0.29). Five patients were studied before and after gastrinoma resection and there was a significant decrease in serum gastrin (p=0.047) but not in CgA post resection (p=0.07). These data demonstrate that CgA levels and gastrin levels are a good measure of the presence or absence of ZES; however, both poorly reflect the extent of the gastrinoma or the growth of the gastrinoma. The interpretation of the CgA is further complicated by the observation that the level correlates with the presence or absence of a gastric carcinoid tumor and therefore may not solely reflect gastrinoma changes. These data support the conclusion that CgA levels have limited utility in patients with gastrinomas.
TRANSFORMING GROWTH FACTOR-a-MEDIATED EPITHELIALMESENCHYMAL INTERACTION PROMOTES FETAL G A S T R I C EPITHELIAL CELL GROWTH. RE Glasgow, BC Visser, SJ Mulvihill. Department of Surgery, UCSF. San Francisco, California, Epithelial-mesenchymal interactions play an important role in the regulation of epithelial development. Transforming growth factor-alpha (TGF-ct) is a potent growth factor for adult gastrointestinal epithelium. Although TGF-a and its receptor, epidermal growth factor receptor (EGF-R), have been identified in the fetal stomach, their role in development is not known. We hypothesized that TGF-a is a mitogen for fetal gastric epithelial cells and mediates epithelial-mesenchymal interactions in the developing fetal stomach. Fetal rabbit gastric tissue sections, cultured fetal rabbit gastric epithelial cells, and, separately, cultured fetal gastric mesenchymal cells from day 21 (term is 31 days) gestation New Zealand White rabbit fetuses were assayed for TGF-a and EGF-R by immunofiuorescence. Fetal gastric epithelial cell culture was used to evaluate EGF-R binding and proliferation in response to TGF-a. A co-culture model was used to identify mesenchymal cell influence on epithelial cell proliferation. Data was collected in quadruplicate using tissue pooled from 8-10 fetuses from a single litter. Each experiment was repeated in at least 3 separate litters. Statistical analysis is by Student's t test. Primary fetal gastric epithelial cell culture was established with 98% purity as measured by immunochemistry using antibodies for cytokeratin AEI:AE3 and epithelial specific antigen. Immunochemistry showed TGF-a and EGF-R expression in the epithelium of gastric tissue sections and gastric epithelial cells in culture. TGF-a immunofluorescence was also seen in mesenchymal ceils. Specific EGF-R binding was seen in epithelial cells by 125I-TGF-a receptor binding studies. Graded concentrations (10-13 to 10.8 M) of TGF-a caused a dose-dependent increase in 3H-thy uptake (basal, 394 _+31 cpm, vs. 10-8 M, 1796_+ 69 cpm, p<0.001). Similar effects were shown by MTS proliferation assay and cell count. These increases were blocked by an antiEGF-R antibody (Ab). Epithelial ceils co-cultured with gastric mesenchymal cells showed increased 3H-thy uptake (basal, 1.8_+0.1x104 cpm, vs. coculture, 4.0 _+0.2x104 cpm, p<0.05). This was attenuated by 59% with a neutralizing anti-TGF-a Ab and 36% with an anti-EGF-R Ab (p<0.05). Similar effects were shown with cell counts. We conclude that TGF-a is expressed in both fetal rabbit gastric epithelial and mesenchymal cells. TGF-a is a mitogen for fetal rabbit gastric epithelial cells and may act as an autocrine growth factor and/or by paracrine release from mesenchymal cells during development. • G4687
EXPRESSION OF FUNCTIONAL GABAA RECEPTORS IN ISOLATED HUMAN INSULINOMA CELLS. G. Glassmeier 1, M. H~pfner t, H. Buhr2, S. Niedrig 1, E.-O. Riecken l, H. Stein 3, H.-J. Quabbe4, C. Rancso 3, B. Wiedenmann 1, H. Schertibl l. 1Abt. Innere Medizin/Gastroenterologie, 2 Abt. Allgemein-, Gef'~- und Thoraxchirurgie, 3 Abt. Allgemeine Pathologie, 4Abt. Innere Medizin/Endokrinologie, Universit~itsklinikum Benjamin Franklin, Freie Universit~it Berlin, Berlin, Germany. Pancreatic islets contain and release high concentrations of ~'-aminobutyic acid (GABA). GABA is thought to play a paracrine role in 13-cells. Searching for a paracrine function of GABA in neoplastic I~-cells we performed patchclamp studies in freshly isolated human insulinoma cells. We show that human insulinoma cells can express functional GABAA receptors. Application of GABA or the specific GABA A receptor agonist muscimol (10 -6 to 10-3 M) dose-dependently activated inward currents at negative holding potentials and outward currents at positive membrane potentials. The amplitudes of the GABA (0.5 mM) induced currents ranged between -35 and -55 pA (n=7) at a holding potential of -80 mV. The reversal potential of the GABA-induced current was determined by digital subtraction of current traces elicited in the absence or presence of GABA with a voltage ramp command. The current reversed close to the CI- equilibrium potential under the recording conditions used. Bath application of the GAB AA receptor antagonist picrotoxin (100 [aM) reversibly reduced the amplitude of GABA-evoked currents by 65 _+6% (n=4). To study voltage-gated calcium channel currents in the isolated ceils, the standard bath solution was replaced by a sodium- and potassium-free solution and barium (10.8 mM) was used as charge carrier. Under these conditions, dihydropyridine-sensitive inward currents were elicited in all cells in response to 100 ms pulses to various test potentials starting from a holding potential of -80 mV (n = 11). Using the perforated-patch technique, activation of GABA A receptors caused a reversible membrane depolarization in the insulinoma ceils. Membrane depolarization resulted in transmembraneous calcium influx through voltage-gated calcium channels. In accordance with the electrophysiological findings activation of GABA A receptors by muscimol increased insulin secretion. Insulin secretion was increased by 50 laM muscimol by about 280%. Thus, GABA A receptors can be expressed in human insulinoma cells and may regulate their insulin release.
Hormones, Transmitters, Growth Factors, and their Receptors Al145
The project was supported by the Mildred Scheel Stiftung and the Deutsche Forschungsgemeinschaft (Sche 326/3-1). G4688 EFFECTS OF PROTEIN ON GASTRIC INHIBITORY POLYPEPTIDE
(GIP) EXPRESSION IN THE RAT. K.D. Glazier, K.-B. Zhan, L.A. Jarboe, C.-C. Tseng, and M.M. Wolfe. Section of Gastroenterology, Boston University School of Medicine and Boston Medical Center, Boston, MA. GIP is a 42-amino acid regulatory peptide originally named on the basis of its ability to inhibit gastric acid secretion. The release of GIP into the circulation has been demonstrated primarily after the ingestion of two major nutrient stimuli, carbohydrates and fat. The present studies were directed to examine nutrient protein regulation of GIP expression in vivo in the rat. Rats fasted overnight were submitted to midline laparotomy, at which time they were perfused with 0.9% NaCI (control) or 10% beef protein hydrolysate (Peptone). Animals were then sacrificed at various time points, and GIP mRNA was measured by Northern blot hybridization and tissue and circulating GIP levels determined by RIA. In response to peptone infusion, duodenal mucosal GIP concentration increased from 8.4 -+ 1.5 (mean -+ SE) to 19.8 -+3.2 ng GIP per mg protein at 120 rain (p<0.01). Plasma GIP levels also increased from 288 _ 20 to 736 -+ 191 at 60 min (p<0.05), to 693 _+ 135 at 120 min (p<0.01), and to 505-+ 62 pg/ml at 240 min (p<0.01). In contrast to the effects on mucosal and circulating GIP, no significant changes in GIP mRNA levels were detected in response to peptone. To determine whether effects of protein on GIP were due to stimulation of gastric acid secretion, rats were pretreated with 25 mg/kg omeprazole, a dose previously shown to abolish acid output in the rat. Following treatment with omeprazole, peptone-stimulated mucosal and plasma GIP concentrations were partially attenuated at 120 min: 12.0 _+3.8 ng GIP per mg tissue protein and 487 _+ 108 pg/ml, respectively. To further examine the effects of luminal acid, anesthetized rats were administered 0.1 M HCI by intraduodenal infusion for 120-min. In addition to stimulatory effects on mucosal and circulating GIP, duodenal acidification increased the GIP/actin mRNA ratio from 1.44±0.12 to 3.22_+0.36 at 30 min (p<0.0001). The results of these studies indicate that in addition to lipids and carbohydrates, nutrient protein provides a potent stimulus for GIP expression in the rat, an effect that occurs at the posttranslational level and may be mediated in part through the acid stimulatory properties of protein. Furthermore, the stimulation of GIP by duodenal acidification may involve pretranslational mechanisms. The effects of acid on GIP expression are consistent with a role for GIP as an enterogastrone in the rat. This research was funded by grants DK48042 and DK53158 from the NIH. G4689
SERUM CHROMOGRANIN A AS A MARKER OF DISEASE PRESENCE, EXTENT OR ACTIVITY IN PATIENTS WITH GASTRINOMA. S.U. Goebel, F. Gibril, F. Yu, J. Serrano, R.T. Jensen, NIH, Bethesda, MD. Chromogranin A (CgA) is a peptide that is released from the neuroendocrine granules Of many normal ceils and neuroendocrine tumors. Studies show that the serum values correlate with tumor presence and burden in patients with pheochromocytomas and carcinoids. Determining the extent and progression of the tumor in patients with pancreatic endocrine tumors (PET's) such as gastrinomas can be difficult and expensive. Hence, the purpose of the present study was to evaluate prospectively whether CgA levels correlated with these parameters in a large number of patients with gastrinoma and to compare CgA results with those of the specific tumor marker, gastrin. In 114 consecutive patients with gastrinoma, serum CgA levels and fasting serum gastrin levels were measured by well-characterized radioimmunoassays. Each was correlated with extent of disease, disease activity, the presence of multiple endocrine neoplasia-type 1 (MEN-I) or gastric carcinoid tumors. Overall, there was a good correlation between CgA levels drawn on 2 consecutive days in a representative sample of patients (n=51, r=0.88, p=0.0001). Disease-free patients post resection had significantly lower CgA and gastrin values than patients not disease-free (p <0.001). A significantly higher proportion of patients not disease-free had elevated CgA levels [>50 ng/ml (72/78=92%)] than patients disease-free [12/36=33% (p <0.0001)]. In comparison, 1/36 (3%) of patients disease-free had an elevated fasting gastrin level whereas 60/78 (77%) that were not disease-free had an elevated level. To facilitate correlations with disease extent, patients were divided into 4 categories consisting of patients with no liver metastases and stable disease (stable, localized) (n=54), no liver metastases but extrahepatic tumor growth (localized with growth) (n=7), liver metastases without growth (stable liver metastases) (n=6) and liver metastases with growth (liver metastases with growth) (n=10). Serum gastrin (p=0.04), but not CgA levels (p=0.96) differed significantly between patients with stable localized disease, compared to localized with growth. Neither serum gastrin nor CgA levels differed between patients with stable localized disease, stable liver metastases or with liver metastases with growth. Patients with gastric carcinoid tumors (n=8) had significantly higher CgA (p=0.04) and gastrin levels (p=0.001) compared to patients without gastric carcinoid tumors (n=104). Patients with or without MEN-I did not differ in serum gastrin (p=0.34) or CgA levels (p=0.29). Five patients were studied before and after gastrinoma resection and there was a significant decrease in serum gastrin (p=0.047) but not in CgA post resection (p=0.07). These data demonstrate that CgA levels and gastrin levels are a good measure of the presence or absence of ZES; however, both poorly reflect the extent of the gastrinoma or the growth of the gastrinoma. The interpretation of the CgA is further complicated by the observation that the level correlates with the presence or absence of a gastric carcinoid tumor and therefore may not solely reflect gastrinoma changes. These data support the conclusion that CgA levels have limited utility in patients with gastrinomas.