Molecular analysis of dibasic endoproteolytic cleavage signals.

Biologically active peptide hormones are synthesized from larger precursor proteins by a variety of post-translational processing reactions. To characterize these processing reactions further we have expressed preprogastrin in two endocrine cell lines and examined the molecular determinants involved in endoproteolysis at dibasic cleavage sites. The Gly93-Arg94-Arg95 carboxyl-terminal processing site of progastrin must be processed sequentially by an endoprotease, a carboxypeptidase, and an amidating enzyme to produce bioactive gastrin. For these studies the dibasic Arg94-Arg95 residues that serve as signals for the initiation of this processing cascade were mutated to Lys94-Arg95, Arg94-Lys95, and Lys94-Lys95. In the GH3 cells the Lys94-Arg95 mutation slightly diminished synthesis of carboxyl-terminally amidated gastrin, whereas in the MTC 6-23 cells this mutation had no effect on amidated gastrin synthesis. In contrast, both Arg94-Lys95 and Lys94-Lys95 mutations resulted in significantly diminished production of amidated gastrin in both cell lines. A specific hierarchy of preferred cleavage signals at this progastrin processing site was demonstrated in both cell lines, indicating that cellular dibasic endoproteases have stringent substrate specificities. Progastrins with the Lys94-Arg95 mutation in GH3 cells also demonstrated diminished processing at the Lys74-Lys75 dibasic site, thus single amino acid changes at one processing site may alter cleavage at distant sites. These studies provide insight into the post-translational processing and biological activation of not only gastrin but other peptide hormones as well.     

Hypergastrinaemia induced by acid blockade evokes enterochromaffin-like (ECL) cell hyperplasia in chicken,hamster and guinea-pig stomach

Summary Treatment of chickens, hamsters and guinea-pigs with large doses of the long-acting antisecretory agent omeprazole for 10 weeks resulted in elevated serum gastrin levels and in increased stomach weight and mass of oxyntic mucosa. Also the antral gastrin cell density was increased. Another striking effect was the hyperplasia of the histamine-producing enterochromaffin-like (ECL) cells — a prominent endocrine cell population with unknown function — in the oxyntic mucosa. Accordingly, the gastric mucosal histamine concentration and rate of histamine formation were increased in all three species. The results suggest that marked and long-lasting suppression of acid secretion leads to elevated serum gastrin levels and diffuse ECL cell hyperplasia not only in the rat, as previously seen, but also in the chicken, hamster and guinea-pig; this hyperplasia is associated with accelerated histamine formation in all three species. The following sequence of events is suggested to occur in mammalian as well as submammalian vertebrates: suppression of acid secretion — hypergastrinaemia — ECL cell hyperplasia.     

Synergistic action of gastrin and ghrelin on gastric acid secretion in rats

Gastrin and ghrelin are secreted from G cells and X/A-like cells in the stomach, respectively, and respective hormones stimulate gastric acid secretion by acting through histamine and the vagus nerve. In this study, we examined the relationship between gastrin, ghrelin and gastric acid secretion in rats. Intravenous (iv) administration of 3 and 10 nmol of gastrin induced transient increases of ghrelin levels within 10 min in a dose-dependent manner. Double immunostaining for ghrelin and gastrin receptor revealed that a proportion of ghrelin cells possess gastrin receptors. Although (iv) administration of gastrin or ghrelin induced significant gastric acid secretion, simultaneous treatment with both hormones resulted in a synergistic, rather than additive, increase of gastric acid secretion. This synergistic increase was not observed in vagotomized rats.These results suggest that gastrin may directly stimulate ghrelin release from the stomach, and that both hormones may increase gastric acid secretion synergistically.     

A cytochemical and immunofluorescence study of endocrine cells in the gut of the ascidian Styela clava

Summary Strong secretin-like immunofluorescence has been demonstrated in endocrine-like cells from the gastric epithelium of Styela. These cells also stain with lead haematoxylin and exhibit a brilliant formaldehyde-induced fluorescence, but do not show any other cytochemical features characteristic of the mammalian APUD series. Tests with antisera to glucagon, gastrin and somatostatin all proved negative. In the oesophagus tests with all four antisera proved negative. The significance of these results is discussed in relation to the phylogeny of vertebrate gastro-intestinal hormones.     

Expression and post-translational processing of gastrin in heterologous endocrine cells

The biosynthesis of gastrin involves a complex series of post-translational processing reactions that result in the formation of a biologically active secretory product. To study the mechanisms for two specific reactions in gastrin processing, namely dibasic cleavage and amidation, we infected AtT-20, GH3, and Rin5-f cells with the retroviral expression vector, pZip-NeoSV(X), containing human gastrin cDNA. We detected gastrin and its glycine extended post-translational processing intermediates (G-gly) in the media and cell extracts of successfully infected cells. Characterization of the molecular forms of gastrin in these cell lines revealed that GH3 and Rin5-f processed gastrin in a manner similar to antral G-cells but the cleavage of the Lys74-Lys75 bond that converts G34 to G17 appeared to be suppressed in AtT-20 cells. Even after conversion of this site to Arg74-Arg75 via site-directed mutagenesis, the At-20 cells synthesized G34 predominantly. All of the infected cells amidated gastrin but the gastrin/G-gly ratio, a reflection of amidation within the cells, was enhanced in GH3 and Rin5-f cells but diminished in AtT-20 cells upon treatment with dexamethasone (10(-4) M) for 3 days. The dibasic cleavage of gastrin was uneffected by dexamethasone. Our data suggest that the activities of post-translational processing reactions responsible for the synthesis of biologically active gastrin exhibit considerable tissue and substrate specificity.     

Ultrastructural and Fluorescence Microscopical Characterization of the Intestinal Endocrine Cells in a Cyclostome,Myxine glutinosa

Endocrine cells of so-called basal-granulated-open type in the intestinal epithelium of a cyclostome, the Atlantic hagfish (Myxine glutinosa), are characterized ultrastructurally and fluorescence microscopically. These cells regularly extend from the basal lamina to the gut lumen, ending in an apical process with microvilli and a filamentous surface coat. Fasting results in an accumulation of secretion granules in all cytoplasmic portions, except for the terminal web area. A similarity is recorded between the distribution of secretion granules and the finely granular fluorescamine-induced fluorescence, suggesting that the fluorescence is associated to some component(s) of the secretory granules. Granule release may take place at the base after an adequate stimulus (presence of food?) at the luminal portion of the cells. The formaldehyde condensation technique shows that insulin-containing hagfish islet parenchymal cells, but not intestinal endocrine cells, store dopamine after intestinal supply of the amine precursor. Acidification of formaldehyde vapour-fixed intestinal epithelium induces fluorescence in the granules of zymogen cells but not of endocrine cells, indicating a low concentration of tryptophyl-peptide(s) in the secretory granules of hagfish intestinal endocrine cells.     

Leptin, gastrointestinal and stress hormones in response to exercise in fasted or fed subjects and before or after blood donation.

Leptin, an ob gene product of adipocytes, plays a key role in the control of food intake and energy expenditure but little is known about leptin response to strenuous exercise in fasted and fed subjects or before and after blood donation. This study was designed to determine the immediate effects of strenuous exercise in healthy volunteers under fasting or fed conditions and before and one day after blood donation (450 ml) on plasma levels of leptin and gut hormones [gastrin, cholecystokinin (CCK), pancreatic polypeptide (PP) and insulin], as well as on "stress" hormones (cortisol, catecholamines and growth hormone. Two groups (A and B) of healthy non-smoking male volunteers were studied. All subjects performed incremental exercise tests until exhaustion (up to maximal oxygen uptake--VO2max), followed by 2 h of rest session. Group A perfomed the tests on a treadmill, while group B on a cycloergometer. In group A, one exercise was performed under fasting conditions and the second following ingestion of a standard liquid meal. In group B, one exercise test was performed as a control test and the second 24 h after blood donation (450 ml). Blood samples were withdrawn 5 min before the start of the test, at the VO2max, and 2 h after finishing the exercise. No significant change in plasma teptin were observed both immediately and 2 h after the exercise in fasted subjects, but after the meal the plasma leptin at VO2max and 2 h after the test was significantly higher, while after blood donation was significantly reduced. The postprandial rise in plasma leptin was accompanied by a marked increment in gut hormones; gastrin, CCK and PP and stress hormones such as norepinephrine, cortisol and GH. These hormonal changes could contribute to the postprandial rise in plasma leptin concentrations, while the fall of leptin after blood donation could be attributed to the inadequate response of stress hormones and autonomic nervous system to exhausting exercise. We conclude that strenuous physical exercise; 1) fails to affect plasma leptin level but when performed after meal but not after blood withdrawal it results in an increase and fall in plasma leptin, and 2) the release of gut hormones (gastrin, CCK and PP) and stress hormones (norepinephrine, cortisol, GH) increase immediately after exercise independently of feeding or blood donation and 3) following blood donation the strenuous exercise resulted in a marked reduction in the plasma leptin, cortisol and GH concentrations, possibly due to the impairment in the autonomic nervous control of these hormones.     

The formaldehyde-fluorescamine method

Summary Formaldehyde reacts with primary amino groups to derivatives which are unable to react with the fluorogenic primary amino group probe, fluorescamine. Paradoxically, however, certain specific cell systems continue to display strong fluorescamine-induced fluorescence after formaldehyde pretreatment. Among such formaldehyde-fluorescamine (FF) positive cell systems are certain peptide- and protein-secreting cells as well as all hitherto investigated types of cancer cells. We have now optimized the cytochemical FF method by using microfluorometry in combination with systematically varied reaction conditions. In addition, the quantitative data indicate that in FF positive cells, formaldehyde pretreatment causes a paradoxical increase in the fluorescence yield with fluorescamine. This has tentatively been ascribed to quenching phenomena, associated with closely spaced primary amino groups. Work with alternative fluorogenic amino group probes (MDPF and OPT) show that these display the same spectrum of tissue selectivity as fluorescamine, but that the latter remains the reagent of choice for the cytochemical FF reaction.     

Histochemically demonstrable monoamines in the pituitary gland and median eminence of the female rat during the postnatal development

Summary The postnatal development of formaldehyde induced fluorescence (FIF) was studied in the pituitary glands of female rats. The effects of 3,4-dihydroxyphenylalanine (L-dopa), D,L-5-hydroxytryptophan (DL-5-HTP) and dopamine (DA) treatments on the FIF were followed during the postnatal period.The appearance of specifically fluorescing monoamines into the cells of the pars intermedia occurred postnatally and the level of the adult fluorescence was reached at 4–5 weeks' age. The intensity of the fluorescence was independent on the density of the fluorescing nerves. Among the fluorescing nerves droplet fibres were regularly observed from the age of 3 weeks, which confirms the theory that these fibres are caused by toxic factors when the blood-brain barrier is not functioning.There was no change postnatally in the number of fluorescing cells in the pars distalis.The fluorescing innervation of the median eminence, developed most rapidly at the age of 1–3 weeks and the level of the adult fluorescence was reached at the age of 4–5 weeks.The first specifically fluorescing cells after L-dopa treatment were observed at 6 days age. A remarkable increase in the number of fluorescing cells was seen between 12 and 18 days. After DL-5-HTP treatment fluorescent cells were seen but at later stages. These observations suggest that the cells in the pituitary gland, which store amine-precursors and monoamines developmentally differ from the APUD-cells. The rapid increase of the fluorescing cells between 12 and 18 days and the simultaneous development of the fluorescing innervation of the median eminence suggest the following correlations: the development of dopaminergic innervation of the median eminence — the secretion of releasing hormones — the activity of PAS-positive cells (FSH, LH and TSH secretion) — the uptake of L-dopa and DL-5-HTP into the PAS-positive cells.Dopamine was not uptaken into the cells of pars distalis. The walls of the blood vessels began to show fluorescence suggesting a barrier mechanism, which prevents the DA-uptake into the PAS-positive cells.This work was supported by the Grant for Young Research Workers, University of Helsinki.     

In vivo and in vitro studies on sex steroid binding protein (SBP) regulation in rainbow trout (oncorhynchus mykiss): Influence of sex steroid hormones and of factors linked to growth and metabolism

The respective roles of sex steroids and hormones related to growth and metabolism, on SBP regulation have been studied in rainbow trout. In vivo, oestradiol (E2) supplementation induces a slow but significant increase of plasma SBP concentration. Testosterone or cortisol injections have no effect. In vitro, the steroid binding protein that accumulates in incubation medium of hepatic cell primary cultures has been characterized and found to be similar to blood SBP. Its production is increased by addition of E2 (maximum: + 300%). This effect develops slowly over several days of culture and is dose dependent; as little as 1–10 nM E2 is effective.

Recombinant rainbow trout GH (rtGH)—0.01 to 1 μg/ml—also increases SBB accumulation as compared to control cells and seems to maintain SBP production over culture duration. In preliminary experiments, (1) insulin-like growth factor (IGF) and SBP concentrations were found to change inversely after a 4 days stimulation with increasing concentrations of GH; (2) recombinant human IGF₁ (250 ng/ml) tended to be inhibitory when SBP production was expressed per mg of total cellular protein, and a micromolar concentration of bovine insulin was clearly inhibitory.

Other hormones tested in vitro: triiodothyronine (10–1000 nM), thyroxine (100 nM), 17,20β-dihydroprogesterone (10–2000 nM), and testosterone (1–1000 nM) did not influence SBP concentration in hepatic cells culture media.     

Identification ultrastructurale et étude immunocytochimique des cellules à somatostatine de la muqueuse antrale du lapin et de la souris

Résumé Les techniques d'immunofluorescence, d'immunoenzymatique, d'imprégnation ou de coloration sélective des cellules endocrines digestives, et de microscopie électronique ont été combinées pour permettre l'identification et l'étude des cellules à somatostatine chez des animaux normaux ou ayant reçu de la L-Dopa. Ces cellules n'ont pas de fluorescence induite à la formaldéhyde et ne sont que très faiblement argyrophiles en technique de Grimélius; elles ne réagissent pas en technique de Sevier-Munger, Hellerström-Hellman, Mac Conaill. Après administration de L-Dopa elles acquièrent und fluorescence (cellules GIC) et deviennent nettement argyrophiles en technique de Grimélius, plus faiblement en technique de Sevier-Munger.En microscopie électronique après identification immunologique sur coupe semi-fine on peut analyser les mêmes cellules sur coupes fines. Il s'agit de cellules basigranuleuses présentant, outre des grains denses difficiles souvent à distinguer de ceux des cellules G, trois structures fines caractéristiques par leur constance et leur coexistence; ce sont: des grains peu denses et homogènes—des graïns centrés par un matériel filamenteux grossier—des microfilaments nombreux. Les cellules à somatostatine semblent différentes des cellules X et D₁.

---

Ultrastructural identification and immunocytochemical study of somatostatin cells in antral mucosa of the rabbit and the mouse

Summary In normal and L-Dopa treated rabbits and mice, combined immunochemical methods, photonic histological methods for endocrine cells and ultrastructural methods were used to elucidate ultrastructure and properties of somatostatin cells of the antral mucosa. In normal rabbits, immunoreactive cells giving no fluorescence with Falck's technic, they corresponded neither to serotonin cells nor gastrin cells; they were unreactive with Fontana, Hellerström-Hellmann, Sevier-Munger and Mac Conaill methods but very slightly stained with Grimelius methods. In L-Dopa treated animals somatostatin cells gave formaldehyde induced fluorescence (they were included in GIC cells, thus in Apud group), exhibited a good reaction with Grimelius and Sevier-Munger methods.In order to carry out the alternate semi-thin/thin section procedure (semi-thin sections for immunofluorescence or immunoenzymatic detection and serial thin sections counterstained for conventional ultrastructure studies), immunological treatment were performed on M.F.F.—glutaraldehyde fixed small fragments of mucosa before inclusion in Epon 812 or, after inclusion, on semi-thin sections. We succeeded in identifying ultrastructurally somatostatin cells. They displayed round or ovoïd shaped secretory granules, and three constant typical structures: numerous microfilaments—light and homogenous granules, often seeming like lipids—granules made up by coarsely filamentous cores surrounded by a large empty halo. Somatostatin cells seemed different of X cells because of their predominant localisation in the antral mucosa (in the rabbit X cells were predominantly in the fundus) and because of the lack of nuclear microfilaments; they also seemed ultrastructuraly different of D₁ cells.

---

---

Travail effectué avec l'aide de la D.G.R.S.T.-Contract n^o 72.7.0043.     

Effects of intraventricular injection of gastrin on release of LH,prolactin, TSH and GH in conscious ovariectomized rats

Conscious ovariectomized (OVX) rats bearing a cannula implanted in the 3rd ventricle were injected with 2 μl of 0.9% NaCl containing varying doses of synthetic gastrin and plasma gonadotropin, GH and TSH levels were measured by RIA in jugular blood samples drawn through an indwelling silastic catheter. Control injections of saline iv or into the 3rd ventricle did not modify plasma hormone levels. Intraventricular injection of 1 or 5 μg gastrin produced significant suppression of plasma LH and prolactin (Prl) levels within 5 min of injection. Injection of 1 μg gastrin had no effect on plasma GH, but increasing the dose to 5 μg induced a progressive elevation, which reached peak levels at 60 min. By contrast, TSH levels were lowered by both doses of gastrin within 5 min of injection and the lowering persisted for 60 min. Intravenous injection of gastrin had no effect on plasma gonadotropin, GH and TSH, but induced an elevation in Prl levels. $\text{In}$$\text{vitro}$ incubation of hemipituitaries with gastrin failed to modify gonadotropin, GH or Prl but slightly inhibited TSH release at the highest dose of 5 μg gastrin. The results indicate that synthetic gastrin can alter pituitary hormone release in unrestrained OVX rats and implicate a hypothalamic site of action for the peptide to alter release of a gonadotropin, Prl and GH. Its effect on TSH release may be mediated both via hypothalamic neurons and by a direct action on pituitary thyrotrophs.     

Development of specific and non-specific somatostatin analogs

Biological activity of six somatostatin analogs has been investigated. In these analogs, disulfide bond is replaced by ethylene bond cyclized with alpha-amino suberic acid. In addition, they contain unique D-configuration in both Trp8 and Cys14 moiety with dicarba substitution. An analog of the short chain length, C omega 7-cyclo (Phe6-Phe7-D-Trp8-Lys9-Thr10-Phe11-D-Asu14) (analog 4) has suppressive effect for GH, but not for other hormones. Analog 6, C omega 9-cyclo(Asn5-Phe6-Phe7-D-Trp8-Lys9-Thr10-Ph e11-Thr12-D-Asu14), has suppressed GH and insulin secretion, but not for gastrin and glucagon. Analog 1, C omega 11-cyclo (Lys4-Asn5-Phe6-Phe7-D-Trp8-Lys9-Thr10-Phe11- Thr12-Ser13-D-Asu14] and 5, C omega 9-cyclo (Lys4-Asn5-Phe6-Phe7-D-Trp8-Lys9-Thr10-Phe11-D-+ ++Asu14) have broad suppressive effect for GH, gastrin, insulin and glucagon release after arginine infusion. The shortest analog, analog 2, C omega 5-cyclo (Phe7-D-Trp8-Lys9-Thr10-D-Asu14) has weak suppressive effect of GH, insulin and glucagon secretion, and it is suggested that Phe6 and Phe11 are necessary for the appearance of suppressive effect of GH. Specific analog, analog 4, may be useful for the future treatment for acromegaly and diabetic retinopathy. Nonspecific analogs, 1 and 5 are candidates for the clinical application of wide variety.     

Cyclic AMP-independent, dual regulation of voltage-dependent Ca2+ currents by LHRH and somatostatin in a pituitary cell line.

Voltage-dependent Ca2+ currents appear to be involved in the actions of hormones that regulate pituitary secretion. In order to investigate modulation of Ca2+ currents by release-inducing and release-inhibiting hormones, we performed whole-cell clamp experiments in the pituitary cell line GH3. The resting potential was approximately -40 mV; spontaneous action potentials were observed in the majority of cells. Superfusion of cells with the stimulatory hormone, LHRH, depolarized the plasma membrane to approximately -10 mV, whereas the inhibitory hormone, somatostatin, caused hyperpolarization to approximately -60 mV; both hormones suppressed spontaneous action potentials. Under voltage clamp conditions, GH3 cells exhibited slowly and fast inactivating Ca2+ currents. LHRH increased whereas somatostatin decreased the slowly inactivating currents; fast inactivating currents were not affected by these hormones. The stimulatory effect of LHRH was not mimicked by intracellularly applied cAMP. In contrast to vasoactive intestinal peptide and forskolin, LHRH did not activate adenylate cyclase in membranes of GH3 cells, but rather appeared to cause inhibition of the enzyme. Hormonal stimulation and inhibition of inward currents were abolished by pretreatment of the cells with pertussis toxin. In membranes of GH3 cells, we identified a pertussis toxin-sensitive G-protein of the Gi-type and Go. We conclude that LHRH and somatostatin modulate voltage-dependent Ca2+ currents via cAMP-independent mechanisms involving pertussis toxin-sensitive G-proteins. The occurrence of both pertussis toxin-sensitive hormonal stimulation and inhibition of voltage-dependent Ca2+ currents in one cell type suggest that these opposite regulations are mediated by distinct G-proteins.     

Specificity of the stimulatory effect of prostaglandins on hormone release in rat anterior pituitary cells in culture

Specificity of the effect of prostaglandins (PGs) on hormone release by the anterior pituitary gland was studied using cells in primary culture. Growth hormone (GH) release is stimulated by all eight PGs studied, PGE₁ and E₂ being 1000-fold more potent than the corresponding PGFs. The release of luteinizing hormone (LH), follicle-stimulating hormone (FSH), and prolactin (PRL) remains unchanged upon addition of PGEs. While the basal release of thyrotropin (TSH) is only slightly stimulated by concentrations of PGEs above 10⁻⁶M, an important potentiation of the stimulatory effect of thyrotropin-releasing hormone on TSH release is observed. The release of GH, TSH and LH is stimulated equally well by PGAs and PGBs at concentrations higher than 10⁻⁶M, 3 × 10⁻⁶M, and 10⁻⁵M, respectively. PGFs do not affect the release of any of the measured pituitary hormones at concentrations below 10⁻⁴M. The stimulation of GH release by PGE₂ can be inhibited by the PG antagonist 7-oxa-13-prostynoic acid, a half-maximal inhibition being found at a concentration of 4 × 10⁻⁵M of the antagonist in the presence of 10⁻⁶M PGE₂. In the presence of somatostatin (10⁻⁸M), the inhibition of GH release cannot be reversed by PGE₂ at concentrations up to 10⁻⁴M. 8-bromo-cyclic AMP-induced GH release is additive with that produced by PGE₂.The present data show that 1) of the five pituitary hormones measured, only GH release is stimulated by prostaglandins at relatively low concentrations, 2) the PGE-induced GH release can be competitively inhibited by 7-oxa-13-prostynoic acid, 3) the inhibition of GH release by somatostatin cannot be reversed by PGE₂ and 4) the PGEs increase the responsiveness of the thyrotrophs to TRH.     

Ultrastructural and cytochemical studies on the cytodifferentiation of duodenal endocrine cells

Summary The development and cytodifferentiation of endocrine cells that produce the gastrointestinal hormones gastrin, cholecystokinin and secretin have been studied by a combined fluorescence-cytochemical, immunocytochemical and ultrastructural approach. The results show that, during development, several ultrastructurally distinct cell types exhibit COOH-terminal gastrin and cholecystokinin immunoreactivity. Furthermore, some cells simultaneously contain both gastrin- and cholecystokinin-specific antigenic determinants. Studies on the time course of development of gastrin and cholecystokinin cells, together with the above-mentioned data, suggest that gastrin cells may be converted into cholecystokinin cells in development. During this period, gastrin, cholecystokinin and secretin cells store the biogenic monoamine, 5-hydroxytryptamine a feature not displayed by the adult counter-parts of these cells. In the adult duodenum, characteristic enterochromaffin (EC) cells store 5-hydroxytryptamin for which, evidence for a possible hormonal role has been presented. Taken together, our data indicate that the differentiation of duodenal endocrine cells occurs in distinct steps, each involving a restriction in the biosynthetic repertoire of the cell.     

Expression of the human insulin gene in the gastric G cells of transgenic mice

The goal of this study was to engineer gastrin-producing G cells of the gastric antrum to produce insulin. A pGas-Ins chimeric gene in which the gastrin promoter drives expression of the human insulin gene was constructed and was validated by transient transfection of GH4 and AGS cells. RT-PCR analysis and sequencing revealed three forms of differentially spliced insulin mRNA in GH4 cells transiently transfected by pGas-Ins. Gas-Ins transgenic mice were generated utilizing this chimeric gene. Northern blot analysis, in situ hybridization, and immunohistochemistry demonstrated expression of the human insulin gene specifically in antral G cells. Northern blot analysis demonstrated that the shortest of the insulin mRNA three forms is predominantly expressed in stomach tissue. RT-PCR analysis also showed expression of the transgene in colon, pancreas, and brain tissues that was undetectable by northern analysis. We conclude that gastrin promoter can be used for targeting expression of human insulin to antral G cells and that antral G cells can express human insulin. Further refining of the chimeric gene design is required to enhance expression.