The preparation and properties of gastrin

Evolution of Neuroendocrine Tumor Therapy

To better understand developments in treatment of neuroendocrine tumors of the gastroenteropancreatic system, and the pivotal roles of native somatostatin and its long-acting analogues play in normal peptide regulation and neuropeptide excess associated with neuroendocrine tumors (NETs), this article delineates and defines distinct eras in the history and discovery of gastrointestinal endocrinology. We highlight the collaboration between academia and industry in basic science and the clinical research that advanced Lu-177-DOTATATE to approval as standard of care therapy for low-grade NETs. Examples of new radioisotopes and therapy compounds currently in development for diagnosis and therapy for high-grade NETs are also discussed.

Gastric Peptides-Gastrin and Somatostatin

Gastric acid secretion (i) facilitates digestion of protein as well as absorption of micronutrients and certain medications, (ii) kills ingested microorganisms, including Helicobacter pylori, and (iii) prevents bacterial overgrowth and enteric infection. The principal regulators of acid secretion are the gastric peptides gastrin and somatostatin. Gastrin, the major hormonal stimulant for acid secretion, is synthesized in pyloric mucosal G cells as a 101-amino acid precursor (preprogastrin) that is processed to yield biologically active amidated gastrin-17 and gastrin-34. The C-terminal active site of gastrin (Trp-Met-Asp-Phe-NH₂ ) binds to gastrin/CCK₂ receptors on parietal and, more importantly, histamine-containing enterochromaffin-like (ECL) cells, located in oxyntic mucosa, to induce acid secretion. Histamine diffuses to the neighboring parietal cells where it binds to histamine H₂ -receptors coupled to hydrochloric acid secretion. Gastrin is also a trophic hormone that maintains the integrity of gastric mucosa, induces proliferation of parietal and ECL cells, and is thought to play a role in carcinogenesis. Somatostatin, present in D cells of the gastric pyloric and oxyntic mucosa, is the main inhibitor of acid secretion, particularly during the interdigestive period. Somatostatin exerts a tonic paracrine restraint on gastrin secretion from G cells, histamine secretion from ECL cells, and acid secretion from parietal cells. Removal of this restraint, for example by activation of cholinergic neurons during ingestion of food, initiates and maximizes acid secretion. Knowledge regarding the structure and function of gastrin, somatostatin, and their respective receptors is providing novel avenues to better diagnose and manage acid-peptic disorders and certain cancers. Published 2020. Compr Physiol 10:197-228, 2020.

Upper gastrointestinal physiology and diseases

Nordic research on physiology and pathophysiology of the upper gastrointestinal tract has flourished during the last 50 years. Swedish surgeons and physiologists were in the frontline of research on the regulation of gastric acid secretion. This research finally led to the development of omeprazole, the first proton pump inhibitor. When Swedish physiologists developed methods allowing the assessment of acid secretion in isolated oxyntic glands and isolated parietal cells, the understanding of mechanisms by which gastric acid secretion is regulated took a great step forward. Similarly, in Trondheim, Norway, the acid producing isolated rat stomach model combined with a sensitive and specific method for determination of histamine made it possible to evaluate this regulation qualitatively as well as quantitatively. In Lund, Sweden, the identification of the enterochromaffin-like cell as the cell taking part in the regulation of acid secretion by producing and releasing histamine was of fundamental importance both physiologically and clinically. Jorpes and Mutt established a center at Karolinska Institutet in Stockholm for the purification of gastrointestinal hormones in the 1960s, and Danes followed up this work by excelling in the field of determination and assessment of biological role of gastrointestinal hormones. A Finnish group was for a long period in the forefront of research on gastritis, and the authors' own studies on the classification of gastric cancer and the role of gastrin in the development of gastric neoplasia are of importance. It can, accordingly, be concluded that Nordic researchers have been central in the research on area of the upper gastrointestinal physiology and diseases.

The regulation of gastric acid secretion - clinical perspectives

The purpose of this review, based upon 40 years of research, is to clear old controversies. The gastric juice is a strong acid with active enzymes (pepsin and lipase); ideal for killing swallowed microorganisms. Totally isolated rat stomach and histamine determination. Human gastric carcinomas were examined for ECL cell differentiation because tumours found in rodents after dosing with inhibitors of acid secretion were reclassified to be of ECL cell origin. The gastrin receptor is localized to the ECL cell only, where gastrin stimulates the function and growth. Drug-induced hypo-acidity induces hypergastrinaemia and ECL cell hyperplasia responsible for rebound acid hypersecretion. Every condition with long-term hypergastrinaemia disposes to ECL cell neoplasia. In man, both atrophic gastritis and gastrinoma lead to ECL cell carcinoids. Proton pump inhibitors induce hypergastrinaemia with ECL cell hyperplasia and ECL cell carcinoids that disappear when stopping treatment. The gastrin antagonist netazepide induces regression of ECL cell carcinoids due to atrophic gastritis. Human gastric carcinomas of diffuse type, particularly the signet-ring subtype, show ECL cell differentiation, suggesting involvement of gastrin in the carcinogenesis. Helicobacter pylori (Hp) causes gastritis and peptic ulcer, and when infecting the antrum only gives a slight hypergastrinaemia with acid hypersecretion predisposing to duodenal ulcer, but protecting from gastric cancer. When Hp infection spreads to oxyntic mucosa, it induces atrophy, reduced acid secretion and marked hypergastrinaemia and cancer.It is remarkable that the interaction between Hp and gastrin may explain the pathogenesis of most diseases in the upper gastrointestinal tract.

Evolution of gastrointestinal hormones: the cholecystokinin/gastrin family

PURPOSE OF REVIEW

To describe recent advances in our understanding of the evolution of gastrointestinal hormones, with the gastrin/cholecystokinin (CCK) family as a model.

RECENT FINDINGS

The release of 11 genomic sequences in the last year has provided a wealth of additional information on peptide hormone sequences. The alternative approach of reverse genetics has identified a separate class of CCK receptor ligands in the nematode Caenorhabditis elegans.

SUMMARY

Three classes of ligands, insect sulfakinins, nematode neuropeptide-like proteins and vertebrate gastrins/cholecystokinins, have now been described for the family of CCK receptors. Although all terminate in an amidated phenylalanine, similarity between the three classes is minimal elsewhere in the sequences. The occurrence of separate gastrin and CCK genes in the dogfish Squalus acanthias dates the divergence of gastrin and CCK to at least 528 +/- 56 Myr ago. The presence of a polyglutamate sequence in marsupial gastrins suggests that the ability to bind ferric ions, which is a critical determinant of biological activity for nonamidated gastrins, was acquired at least 173 +/- 12 Myr ago. Comparison of gastrin or CCK sequences between species suggests that, apart from the C-terminal tetrapeptide amide that is required for receptor binding, conservation is largely restricted to the dibasic processing sites and to the C-terminal flanking peptides of gastrin and CCK. The independent conservation of the latter peptide may be either a consequence of a requirement for precise processing, or may indicate a separate function.

Sporadic versus hereditary gastrinomas of the duodenum and pancreas: Distinct clinico-pathological and epidemiological features

Gastrinomas are defined as gastrin secreting tumors that are associated with Zollinger-Ellison syndrome (ZES). ZES is characterized by elevated fasting gastrin serum levels, positive secretin stimulation test and clinical symptoms such as recurrent peptic ulcer disease, gastroesophageal reflux disease and occasional diarrhea. Genetically, nonhereditary (sporadic) gastrinomas are distinguished from hereditary gastrinomas, which are associated with multiple endocrine neoplasia type 1 (MEN1) syndrome. In general, duodenal gastrinomas are small and solitary if they are sporadic and multiple as well as hereditary. The sporadic gastrinomas occur in the duodenum or in the pancreas while the hereditary gastrinomas almost all occur in the duodenum. Our series of 77 sporadic duodenal neuroendocrine tumors (NETs) includes 18 patients (23.4%) with gastrinomas and ZES. Of 535 sporadic NETs in the pancreas collected from the NET archives of the departments of pathology in Zürich, Switzerland, and Kiel, Germany, 24 patients (4.5%) suffered from sporadic pancreatic gastrinomas and ZES. These NETs have to be distinguished from tumors with immunohistochemical positivity for gastrin but without evidence of ZES. An additional 19 patients suffered from MEN1 and ZES. These patients showed exclusively duodenal gastrinomas, but not pancreatic gastrinomas. The prognosis of sporadic and MEN1-associated duodenal gastrinomas is better than that of pancreatic gastrinomas, since they progress slowly to liver metastasis. In summary, sporadic and MEN1-associated gastrinomas in the duodenum and pancreas show different clinico-pathological and genetic features. The incidence of sporadic duodenal gastrin-producing tumors is increasing, possibly due to optimized diagnostic procedures. In contrast, pancreatic MEN1-associated gastrinomas seem to be extremely rare. A considerable subset of tumors with immunohistochemical expression of gastrin but without evidence of ZES should be designated as functionally inactive NETs expressing gastrin, but not as gastrinomas.

The cellular localization of the cholecystokinin 2 (gastrin) receptor in the stomach

The role of the gastric acid secretagogues acetylcholine, gastrin and histamine has been debated for decades. Initially, the mast cell was considered the source of acid stimulatory histamine. Later, Håkanson & Owman (1969) showed that the entero-chromaffinlike (ECL) cell produces and stores histamine in several species, including rat and man. Kahlson et al. (1964) showed that food and gastrin stimulated oxyntic mucosal histamine synthesis and release, Berglindh et at. (1976) that histamine and cholinergics but not gastrin induced acid secretion in isolated oxyntic glands and parietal cells, and Rangachari (1995) that acetylcholine or gastrin released histamine in isolated mucosa. These findings suggested that gastrin stimulates acid secretion through release of ECL cell histamine. Studying simultaneous histamine release and acid secretion in isolated oxyntic mucosal cells, we found that gastrin stimulated acid secretion only in preparations releasing histamine. Moreover, in the isolated rat stomach, gastrin stimulated both histamine release and acid secretion. Maximal acid output was higher with histamine than with gastrin, and augmented by acetylcholine but not by gastrin. These findings strongly suggested that gastrin acts by releasing histamine. Finally, a fluorescein-labelled gastrin analogue bound to the ECL cell, not to the parietal or stem cell regions. This is interesting, recalling that gastrin has a potent and specific trophic effect on the ECL cell and only a general effect on all other oxyntic cell types. In conclusion, physiological observations are best explained by localising the CCK2 receptor only to the ECL cell, the other effects of gastrin on the gastric mucosa being secondary to the release of mediators from the ECL cell.

Neural hormonal regulation of exocrine pancreatic secretion

Exocrine pancreatic secretion is regulated by hormone-hormonal and neural-hormonal interactions involving several regulatory peptides and neurotransmitter from the gut, the pancreas and the vagus nerve. The roles of the gastrointestinal peptides including secretin, CCK, neurotensin, motilin, PYY and pancreatic islet hormones including insulin, pancreatic polypeptide and somatostatin have been established. Interactions among secretin, CCK and neurotensin produce synergistic stimulatory effect. Motilin modulates the cyclic pattern of pancreatic secretion while local insulin provides a permissive role for the action of secretin and CCK at physiological concentration. Somatostatin, PYY and pancreatic polypeptide are inhibitory regulators, acting either on the release of secretin and CCK or on the action of the two stimulatory hormones. The vagal afferent-efferent pathway mediates the actions of many of these regulatory peptides, particularly of secretin and CCK. Acetylcholine and nitric oxide are the neurotransmitters known to mediate the actions of secretin and CCK. Serotonin (5-HT) released from enterochromaffin cells in the intestinal mucosa and nerve terminals of the enteric nervous system and intrapancreatic nerves may be involved in both stimulatory and inhibitory mechanism through its various receptor subtypes. 5-HT also mediates the action of secretin and CCK. The regulatory roles of neuropeptides, PACP and GRP, are now established, whereas those of others are being uncovered. Pancreatic juice provides both positive and negative feedback regulation of pancreatic secretion through mediation of both secretin- and CCK-releasing peptides. Three CCK-releasing peptides have been purified: monitor peptide from pancreatic juice, diazepam-binding inhibitor from porcine intestine, and luminal CCK-releasing factor from rat intestinal secretion. All have been shown to stimulate CCK release and pancreatic enzyme secretion. Pancreatic phospholipase A2 from pancreatic juice and intestinal secretion appears to function as a secretin-releasing peptide. However, the detailed map of neurohormonal regulatory pathways of exocrine pancreatic secretion is yet to be constructed.

The pivotal role of John S. Edkins in the discovery of gastrin

John Sydney Edkins was born in London in 1863. After gaining two open scholarships, he attended Caius College, Cambridge and studied physiology under the tutelage of J.N. Langley and M. Foster. During this period he published on the chemical nature of pepsinogen with Langley. After qualifying as a medical doctor, he worked in Manchester before returning to London, where he succeeded E. Klein as Head of Physiology at St. Bartholomew's. For financial reasons he also worked part time at Bedford College for Women. In 1902 Bayliss and Starling overturned Pavlov's doctrine of the nervous regulation of gastrointestinal function by discovering the pancreatic secretagogue secretin-the first identifiable chemical messenger. Edkins applied a similar rationale to the stomach and in a classic series of experiments noted that injection of a pyloric mucous membrane extract resulted in gastric acid and pepsin secretion in anesthetized cats. In 1905 he named this putative active agent "gastrin." Although his ideas were initially accepted, the discovery of histamine in 1910 and the identification that extracts from other tissues had a similar physiologic effect raised serious questions regarding the validity of the existence of gastrin. Somewhat discouraged, Edkins pursued the teaching and training of women physiologists at Bedford College, where he later became Chairman of Physiology. Outside of science he successfully pursued other interests and became President of the British Croquet Association. The demonstration that gastrin was a unique antral stimulant of acid secretion by Komarov in 1938 was followed by the purification and elucidation of its chemical structure by Gregory and Tracy in 1964. Their work allowed final validation of Edkins' original hypothesis. Edkins died in London in 1940, not only fated to predecease the vindication of his hypothesis but unable to witness the evolution of his discovery into a paradigm for the hormonal regulation of secretory and proliferative cellular activity.

The G cell

The study of gastrin continues to serve as an excellent model for gastrointestinal regulatory processes. This review highlights some recent advances in the field by outlining gastrin biosynthesis, summarizing current understanding of gastrin receptors, describing the regulation of gastrin release, and discussing the clinical implications of gastrin in the pathogenesis of peptic ulcer disease. Emphasis is on three emerging areas of gastrin research: the novel finding that one of gastrin's posttranslational processing intermediates has biological activity distinct from that of the mature peptide; elucidation of gastrin's signal transduction mechanisms that mediate the trophic effects of the peptide; and the role of gastrin in peptic ulcer disease pathogenesis secondary to Helicobacter pylori infection.

Review article: the use of gastric acid-inhibitory drugs--physiological and pathophysiological considerations

All vertebrates secrete gastric acid. Acid denatures the proteins in the food and thus makes them more accessible to proteolytic enzymes, and it kills swallowed micro-organisms. Gastric acid plays an important pathogenetic role in peptic ulcer disease and reflux oesophagitis. In these diseases, drugs that inhibit secretion of gastric acid will heal the lesions and suppress the symptoms. However, both reflux oesophagitis and peptic ulcer tend to recur when the acid-inhibitory treatment is stopped. Therefore, these patients often require long-term treatment with acid-inhibitors. In this overview the potential risks of long-term profound inhibition of acid secretion, raising the pH above 4 for a considerable time, resulting in reduced killing of micro-organisms and secondary hypergastrinaemia, are discussed. Gastrin regulates both the function (production and release of histamine) and growth of the enterochromaffin-like (ECL) cell. Hitherto, the role that this cell plays in gastric carcinogenesis appears to have been underestimated.

Biologic and immunologic gastrin activity in serum of patients with gastrinoma. Bioassay of gastrin activity in serum

The biologic gastrin activity in serum from 14 patients with the Zollinger-Ellison syndrome was assessed by the stimulation of histamine release and acid secretion from the isolated vascularly perfused rat stomach and compared with the immunologic activity as determined by radioimmunoassay using an antibody directed towards the active site of the gastrin molecule. Biologic gastrin activity assessed by the stimulation of histamine release was more closely correlated to immunologic gastrin activity than biologic activity assessed by the stimulation of gastrin acid secretion. This study does not contradict the concept that gastrin stimulates acid secretion at least partly by releasing histamine and also shows that the immunologic gastrin activity determined with the help of an antibody directed towards the active site reflects biologic activity.

The gastrin receptor assay

Gastrin is a major physiologic regulator of gastric acid secretion and growth of the oxyntic mucosa. Biologically active radiolabelled hormones may be used to characterize and localize receptors for peptide hormones. The cellular localization of the gastrin receptor in the fundic mucosa, however, is still a matter of great debate owing to difficulties in developing a gastrin receptor binding assay. Despite considerable work in several laboratories, the criteria for true receptor binding have not yet been fulfilled. The preparation of a suitable tissue receptor material (plasma membranes or isolated cells) from the heterocellular fundic mucosa seems to be the major problem. This problem may be related to the fact that the receptors are only present on the enterochromaffin-like cells (ECL), which constitute but a minor fraction of the cells in the oxyntic mucosa. Furthermore, the second messenger of gastrin is still not known, and the poor functional responsiveness of isolated cells and the oxyntic glands to gastrin further complicates the evaluation of the gastrin receptor. In this review the different steps in the gastrin receptor assay (the labelling of gastrin, preparation of the receptor, and the incubation and correlation of the binding and biologic effect of gastrin) are discussed.

Role of peptide radioimmunoassay in understanding peptide-peptide interactions and clinical expression of gastroenteropancreatic endocrine tumors

Peptide radioimmunoassay has become an important clinical and research tool in understanding the role of peptides in the pathophysiology of gut endocrine tumor syndromes. A gut peptide radioimmunoassay laboratory has been established for the diagnosis and clinical monitoring of endocrine tumors of the gastroenteropancreatic (GEP) system. Radioimmunoassay has enhanced our awareness that co-occurring peptide interactions may modify and ultimately influence the clinical expression of these tumors. Furthermore, it has helped develop a rationale for the use of prototype peptides such as somatostatin and its long-acting analogue Sandostatin (SMS 201-995) in the management of GEP tumors. This group's experience, as well as the experience of other investigators, is presented, and the clinical utility of peptide radioimmunoassay in the field of gut endocrinology is demonstrated.

Gastrins and gastrinomas

The past 20 years have seen gastrin attain true hormonal status. Its structure has been characterized, it has been synthesized, radioimmunoassays for its measurement in blood and tissues have been developed and its physiology and metabolism elucidated. Of much interest to clinicians has been the association between gastrin and tumours of the pancreas (gastrinomas) and atrophic gastritis. The advent of gastrin measurement has facilitated the diagnosis of gastrinoma and the availability of powerful acid suppressants has altered the therapy of gastrinoma.

The intestinal phase hormone

The existence of a stimulatory intestinal phase of gastric acid secretion has been suspected, but largely ignored, for many years. Recently, however, it has become clear that the intestinal phase plays an important role in acid production during digestion. The intestinal phase is of additional interest in relation to the profound gastric acid hypersecretion associated with portacaval shunt (PCS). Substantial evidence indicates that PCS-related gastric hypersecretion is due to unmasking of the intestinal phase by hepatic bypass of a humoral stimulant in portal blood that is normally degraded to a considerable extent by the liver. Studies in our laboratory during the past 12 years have provided strong physiologic evidence for humoral mediation of both the intestinal phase of gastric secretion and of PCS-related hypersecretion by a hormone that arises in the small intestine, particularly in the jejunum. Furthermore, our studies have demonstrated that this intestinal phase hormone (IPH) exists in humans as well as in dogs, rats, and pigs. Additionally, recent work by a number of investigators, as well as by our group, has provided convincing evidence that IPH is different from any of the known gastric stimulatory hormones. With these physiologic observations as a background, we have used a classical method for extracting acidic peptides to prepare a hog intestinal mucosa extract (HIME) that has all of the known physiologic properties of an IPH. Specifically, HIME contains a potent stimulant of gastric acid secretion that acts according to a linear dose-response relationship; that is not gastrin in any of its immunoassayable forms; that significantly augments the maximal acid secretory responses to pentagastrin, gastrin, CCK, and histamine; and that is substantially degraded by the liver, in contrast to gastrin and CCK. Efforts at isolating the gastric stimulatory substance in HIME suggest that it is a peptide of low molecular weight. Work directed at isolating IPH in pure form and identifying it is in progress.

Effects of atropine on secretion and motility in isolated gastric mucosa and attached muscularis externa from ferret and cat

1. A combined in vitro preparation of gastric mucosa and adjacent muscle from young ferrets and kittens has been used to study the effects of atropine on acid secretion and motility produced by acetylcholine (ACh) and pentagastrin.2. The minimal dose of atropine required to abolish a maximum secretory response to ACh also prevented the associated motility response. This dose of atropine also blocked the motility response to pentagastrin, but was without influence on the secretory effect of this agent. A 10(3) times larger dose of atropine reduced the secretory effect of pentagastrin by half, probably not by anti-muscarinic effect. The results exclude the possibility that the acid secretory response to pentagastrin necessarily involves a cholinergic receptor.3. The results support the view that the response of the fundic smooth muscle to pentagastrin depends on the excitation of cholinergic nerves.4. No evidence has been found of any cholinergic component in the acid secretory response to pentagastrin. In assessing the significance of this result, however, it must be remembered that the Auerbach plexus has been removed over the major part of the mucosa, and the Meissner plexus deprived of input and probably damaged.5. The results are compatible with the hypothesis that the depressant effect of atropine on acid secretion produced by gastrin and its derivatives is due to the elimination of a cholinergic potentiating influence arising in the intramural plexuses. The residual Meissner plexus elements in this in vitro preparation appear inadequate to sustain this effect.

Integration of biochemical functions of different cells of rat gastric mucosa for hydrochloric acid secretion

The regulation patterns of gastric acid secretion in rats were investigated. Pentagastrin and histamine stimulate gastric acid secretion, but the inhibitors of DNA-dependent synthesis of RNA and of proteins prevent only the pentagastrin action. It has been found that pentagastrin induces histidine decarboxylase in gastric mucosa, ensuring local accumulation of histamine. The latter activates adenylate cyclase and results in 3',5'-AMP accumulation in gastric tissues. The administration of pentagastrin, histamine or 3',5'-AMP enhances the activity of gastric carbonic anhydrase, the enzyme which takes part in HCl formation. The data suggest that these three compounds act sequentially (pentagastrin leads to histamine leads to3',5'-AMP) and the effect of the last one could be mediated through 3',5'-AMP dependent protein kinase. The experiments in vitro demonstrated that gastric carbonic anhydrase can be separated into two isoenzymes and thephosphorylation of one of them by the 3',5'-AMP dependent protein kinase sharply increases its activity. The findings raise the possibility that histamine and 3',5'-AMP, mediating gastrin action, form together with enzymes (histidine decarboxylase, adenylate cyclase, protein kinase, carbonic anhydrase) a caascade of amplifiers. Autoradiographic studies have shown that [3H]-pentagastrin is not bound by oxyntic cells but adheres preferentially to histamine-producing alpha-like endocrine cells and to the chief cells, while 3H-histamine adheres preferentially to oxyntic and to chief cells. Electron microscopy indicates that only pentagastrin (but not histamine) initiates in alpha-like endocrine cells ultrastructural changes characteristic for induction. Pentagastrin, histamine and 3',5'-AMP administration produces in oxyntic cells ultrastructural changes typical for the secretion processes. These results lead to assumption that pentagastrin (gastrin) induces histidine decarboxylase in alpha-like endocrine cells of gastric glands. Histamine which is secreted enhances adenylate cyclase activity in the neighbouring oxyntic cells where 3',5'-AMP dependent protein kinase activates carbonic anhydrase by means of phosphorylation. These different cells form, probably, a multicellular functional unit for gastric acid secretion.

Gastric acid secretion

Advances in the understanding of physiologic control of gastric secretion raise the hope that an effective nonsurgical therapy for peptic ulcer disease will be developed soon. This paper reviews these advances with special emphasis upon those aspects of hormonal control, cyclic nucleotide function, and histamine activity which may have therapeutic applications.

Control factors in the release of gastrin by direct electrical stimulation of the vagus

The release of gastrin by direct electrical stimulation of the vagus was studied together with the relative effects on the response of antral and duodenal acidification. As expected, gastrin levels increased to three times the normal simulated response following antral neutralization. In contrast, duodenal acidification failed to influence the vagal release of gastrin when the antrum was neutralized although it had a minor effect when the antrum was acidified. Thus the antral pH dominates over duodenal pH as a factor in controlling gastrin release. Surprisingly, atropine in doses which blocked acid release and produced marked cardiac effects failed to inhibit the release of gastrin from the antrum on vagal stimulation. This suggests that, using this model, vagal release of gastrin, if cholinergic, is highly resistant to atropine.

The Bayliss-Starling lecture 1973. The gastrointestinal hormones: a review of recent advances

Images