Pancreatic secretion is regulated by highly integrated neural and hormonal influences that involve the brain, enteric nervous system, and gastrointestinal tract. Although these processes are complex they illustrate the finely regulated nature that is needed for maintaining sufficient secretion of pancreatic enzymes that are essential for adequate ingestion and digestion of nutrients.
KeywordsPatterns of secretion, Basal secretion, Integrated response to meals, Phases of the meal response, Cephalic phase, Gastric phase, Intestinal phase, Absorbed nutrient phase, Neurohormonal regulators, Neural mechanisms, Hormonal mechanisms, Other regulators, Feedback regulation of pancreatic secretion, Inhibition of pancreatic secretion, Glucagon, Glucagon-like peptide, Somatostatin, Pancreatic polypeptide, PYY, Pancreastatin, Leptin, Ghrelin, Pancreatic function testing
Pancreatic secretion is regulated by highly integrated neural and hormonal influences that involve the brain, enteric nervous system, and gastrointestinal tract. Although these processes are complex they illustrate the finely regulated nature that is needed for maintaining sufficient secretion of pancreatic enzymes that are essential for adequate ingestion and digestion of nutrients. Another source of information about the regulation of exocrine pancreatic physiology can be found in the The Pancreapedia: Exocrine Pancreas Knowledge Base ( www.pancreapedia.org ).
Patterns of Secretion
Basal secretion of pancreatic enzymes, bicarbonate, and fluid occurs when food has emptied from the stomach and the small intestine and is associated with fasting. Meal-induced pancreatic secretion occurs following the ingestion of the meal and is associated with the ingestion, digestion, and absorption of food. Meal-induced secretion is thought to be the most important aspect of pancreatic exocrine function because lack of pancreatic enzyme secretion results in malabsorption and maldigestion of nutrients and general poor nutrition. Interestingly, however, the amount of pancreatic secretions that are present in the intestine even during the basal condition may be sufficient to facilitate substantial enzymatic degradation of ingested foods and prevent malnutrition. These observations suggest that the exocrine pancreas functions at a level that is considerably greater than the minimum necessary for complete digestion of food. Despite this finding, most studies of pancreatic function have been devoted to studying stimulated secretion rather than basal secretion.
The rate of basal pancreatic exocrine secretion is generally expressed as a percent of the maximal amount that the pancreas can secrete when stimulated by a secretagogue such as cholecystokinin (CCK). When expressed as a percentage of the maximal pancreatic capacity to secrete enzymes, values for basal enzyme secretion range from 10% of maximal in cats to 20% in humans and 30% of maximal in rats. Basal secretion of bicarbonate, however, is often only 1%–2% of the maximal secretory rate compared to administration of exogenous secretin and there is considerably less species variation with the exception being the rat where basal bicarbonate secretion is 25% of maximal secretion. Therefore, the basal secretory rate of pancreatic enzymes may be adequate to prevent malabsorption of ingested nutrients since frank malabsorption generally appears only when pancreatic enzyme secretion is reduced to 10% or less.
Several different mechanisms are responsible for basal pancreatic exocrine secretion and could be due to (1) an automaticity of the gland, (2) regulation by low levels of gastrointestinal hormones such as CCK or secretin, or (3) due to the release of neurotransmitters such as acetylcholine. In vitro, pancreatic acinar cells demonstrate basal enzyme release, although it is unknown what is responsible for this basal exocytosis. In experimental animals such as dogs and rats, basal secretion is due primarily to cholinergic innervation as atropine blocks basal secretion and CCK receptor antagonists have repeatedly been shown to have no effect on basal pancreatic secretion. The inhibitory effects of anticholinergic drugs are probably due to the blockade of the muscarinic receptor on pancreatic acinar cells that block the effects of acetylcholine released from postganglionic pancreatic nerves. Moreover, basal levels of CCK are not sufficiently high to stimulate pancreatic exocrine secretion. Therefore, in contrast to secretin which appears to play a role in basal pancreatic secretion, there is strong evidence that CCK is not important in regulating basal pancreatic secretion in rats. In humans, however, varying reports have suggested that CCK and cholinergic inputs may contribute to basal pancreatic enzyme secretion. Profound inhibition of pancreatic secretion has been seen after atropine administration in humans and some reports indicate that CCK receptor antagonists may reduce pancreatic secretion. These findings indicate that low levels of circulating CCK or CCK acting locally on vagal afferent nerves are sufficient to stimulate human pancreatic secretion or that CCK released as a peptidergic transmitter from pancreatic neurons contributes to basal secretion. Recently, an alternative pathway for CCK-induced pancreatic secretion involving pancreatic stellate cells that lie in close proximity to acinar cells has been proposed. Human pancreatic stellate cells express CCK1 receptor and in primary culture it was shown that reasonably low levels of CCK (20 pM) stimulated acetylcholine release which induced amylase secretion from acini. This finding raises the possibility that CCK may regulate cholinergic stimulation of the pancreas through both neural and nonneural pathways.
Pancreatic bicarbonate secretion is largely regulated by secretin. Basal bicarbonate secretion correlates with plasma secretin levels, however, cholinergic inputs also affect bicarbonate release as atropine decreases both the basal and the secretin-stimulated bicarbonate secretion. It is likely that basal bicarbonate release is augmented by acetylcholine released from nerves locally in the pancreas.
Interestingly, the patterns of pancreatic secretion appear to be more complex if one examines secretory rates over time. Even under basal conditions, over several minutes, the secretory rate of pancreatic secretion varies. There are brief increases in bicarbonate and enzyme secretion that occur every 60–120 min. These bursts of pancreatic secretory activity coincide with periods of increased motor activity of the stomach and duodenum that are associated with the migrating motor complex (MMC) (25–27). Simultaneous with increases in pancreatic secretion, there are increases in gastric acid secretion and biliary flow into the duodenum. All of these actions are associated with increases in motilin levels in the blood. In addition, pancreatic polypeptide levels correlate well with the antral phase II motor activity and pancreatic enzyme secretion. These activities appear to be cholinergically mediated since atropine administration or ganglionic blockers abolish the periodic spikes in basal enzyme secretion. The administration of motilin prematurely initiates pancreatic secretion that is seen during the basal period and shortens the time between peaks of secretory activity. Conversely, immunoneutralization of motilin with specific antiserum abolishes the cyclic pattern of pancreatic secretion. The administration of pancreatic polypeptide (PP) inhibits basal pancreatic secretion and immunoneutralization with PP antiserum augments the peak in pancreatic secretion that is seen in the basal period. These findings are consistent with the overall belief that the MMC functions as a housekeeper to eliminate chyme, debris, and other secretions during the interdigestive period and to keep microbial populations in check. Although it is reduced, the periodic basal pattern of pancreatic secretion persists despite duodenectomy or autotransplantation of the pancreas indicating that independent of any obvious hormonal or neural influences the pancreas is able to generate an endogenous periodicity that is most likely due to activation of postganglionic cholinergic neural activity. These findings are consistent with electrophysiologic observations documenting spontaneous ganglionic neural activity. A circadian rhythm to pancreatic secretion has been described in rats where there is a peak in secretion during the dark phase of the day-night cycle.
Integrated Response of Meals
The major function of the exocrine pancreas is to facilitate the efficient digestion of food and absorption of micronutrients. The function of pancreatic enzymes is to break down macronutrients such as proteins to small peptides and amino acids, triglycerides to fatty acids and monoglycerides, and carbohydrates to sugars that can be easily absorbed from the small intestine. Pancreatic bicarbonate secretion is important for neutralizing gastric acid and creating an intraluminal environment with a pH that is hospitable to the action of enzymes especially in fat digestion. Unfortunately, very little is known about the actual pancreatic secretory process involved in the normal digestion of meals in humans. This problem is due in large part to difficulties in sampling intestinal secretions with and without food in the lumen without altering the normal physiologic function of the pancreas, biliary system, and intestine that together constitute an integrated response to a meal. Attempting to measure pancreatic juice free of biliary secretions requires cannulation of the pancreatic duct or diversion of pancreatic juice flow from the intestine. Diversion of pancreatic juice from the intestine disturbs the normal milieu by removing one or more factors that are either involved in neutralizing gastric acid or necessary for maintaining the proper intestinal environment in which intestinal releasing factors may function (mentioned later).
When pancreatic juice is diverted from the intestine, pancreatic bicarbonate is no longer present and the gastric acid entering the duodenum may avoid neutralization. Gastric acid is also a potent stimulus of secretin release which, in turn, stimulates pancreatic fluid and bicarbonate secretion. Interestingly, in dogs, the flow of pancreatic juice rich in bicarbonate increases following the diversion of pancreatic juice from the intestine. Although this fluid is rich in bicarbonate the amount of enzymes in the juice does not increase. In contrast, in rats diverting pancreatic juice stimulates pancreatic enzyme secretion to near-maximal amounts similar to that produced by exogenous CCK.
Quantifying pancreatic secretion associated with meals in humans is extremely difficult; therefore, there is very little quantitative information available. Sampling from the duodenum is fraught with hazards. Concentration measurements are notoriously unreliable because the volume of duodenal contents may vary. Moreover, duodenal bicarbonate is produced not only by the pancreas, but is secreted from the biliary tract and intestinal mucosa. There are also important species differences because pancreatic bicarbonate secretion in the pig is much less than biliary bicarbonate production, whereas pancreatic bicarbonate is the major source for neutralization of acid chyme in dogs. Maximum bicarbonate concentration in humans may be as high as 150 mM which is approximately twice as that in the rat. Measurement of trypsin is also difficult because it requires enzymatic activation of trypsinogen and the active form binds avidly to food making its quantification complex.
The physical nature of food is an important factor in the regulation of meal-stimulated pancreatic responses. When comparing a solid form of a meal to the same food homogenized to a liquid, the total pancreatic trypsin output is the same, however, the secretory response is prolonged, which is consistent with solid food emptying from the stomach at a slower rate than liquids. The maximal pancreatic enzyme response to fat occurs at low rates of fat delivery to the intestine. When the fat content of the meal was increased there was no further stimulation of pancreatic enzyme secretion; however, when the protein content was increased there was nearly a twofold increase in meal-stimulated pancreatic enzyme secretion.
The effects of food on gastric emptying rates appear to be important in regulating pancreatic secretory responses. The duration of pancreatic secretion correlates with the time required for the stomach to empty. As long as food was emptying from the stomach, pancreatic secretion was maintained at a high level; when the upper small intestine was free of food, pancreatic secretion declined. Importantly, delivery of food further down the intestine has an inhibitory effect on pancreatic secretion. Overall, meal-stimulated pancreatic secretion is 50%–60% of the maximal secretory capacity of the organ.
Meal-stimulated bicarbonate and enzyme responses are consistently found to be lower than maximal rates of secretion regardless of the species studied including dogs, rats, and humans. Although the reasons for this are not entirely clear they may be related to slow rates of gastric emptying. With delayed delivery of nutrients to the duodenum there may be submaximal pancreatic stimulation from the gastrointestinal hormones CCK and secretin. There also may be simultaneous or subsequent release of inhibitors of pancreatic secretion as food travels further down the intestine. Furthermore, as absorption of nutrients occurs they are no longer present in the lumen of the intestine to stimulate pancreatic responses. All of these possibilities may contribute to postprandial pancreatic secretion which is less than that can be induced by CCK alone.
Phases of the Meal Response
There are four major physiologic digestive processes that are used to describe pancreatic secretion. These include cephalic , gastric , intestinal , and absorbed nutrient phases which describe the sites at which signals to the pancreas originate. Each of the phases involves both secretory and inhibitory inputs although the overall effect is overwhelmingly stimulatory. The secretory processes involve multiple levels of regulation including neurohormonal and hormonal-hormonal interactions. Although many of the steps that stimulate pancreatic secretion seem redundant, the system ensures that adequate enzymes are available for digestion. In general, following a meal there is a temporal relationship in which the cephalic phase contributes to pancreatic secretion prior to initiation of the gastric and intestinal phases, respectively. The absorbed nutrient phase includes inputs from each of the other three phases and involves the effects of nutrients absorbed into the blood to affect pancreatic secretion. Although it is important to understand the contribution of each of these phases to pancreatic secretion, it is more important to recognize that with normal feeding there is considerable overlap. Therefore, the integrated response to a meal results from the combination of all phases for physiologic regulation of pancreatic secretion.
The cephalic phase of secretion results from inputs including the sight, smell, taste, and act of eating food and can account for up to 25% of the pancreatic exocrine secretion of a meal. Although we commonly think of these processes as stimulating pancreatic secretion they may also generate inhibitory signals when eating is associated with unpleasant features such as unattractive, malodorous, or bad tasting food. The cephalic phase of secretion has been produced in humans by presenting them with food that they see, smell, and taste but not swallow (a process known as modified sham feeding). In animals, food can be diverted from the esophagus by a surgically prepared esophageal or gastric fistula, and sham feeding can occur by allowing these animals to eat and swallow while preventing food from entering the stomach. In both dogs and humans, sham feeding stimulates low volumes of pancreatic secretions that are rich in enzymes but low in bicarbonate. The total pancreatic secretory response to sham feeding is approximately 25%–50% of maximal. Secretion of the islet hormone, pancreatic polypeptide, increases with sham feeding and has been used as an indicator of vagal innervation of the pancreas. In humans, the duration of pancreatic response to modified sham feeding is brief, lasting approximately 60 min, and ceases at the conclusion of sham feeding. If swallowing is included in sham feeding, the pancreatic secretory response is much greater. In contrast, in dogs, the pancreatic enzyme response to sham feeding lasts > 4 h.
There is substantial experimental data to support the concept that cephalic stimulation of pancreatic secretion is mediated by the vagus nerve. First, cholinergic agonists produce a pancreatic secretory response similar to that of cephalic stimulation. Second, the vagus nerve is the major source of cholinergic neurotransmitters to the pancreas. Third, electrical nerve stimulation to the vagus nerve or administration of 2-deoxyglucose (2-DG) which causes hypoglycemia (and initiates a vagal response) stimulated pancreatic juice flow similar to that of sham feeding. Finally, vagotomy blocked these responses. In anesthetized rats, pancreatic fluid and protein output following electrical nerve stimulation or 2-DG was partially blocked by atropine. Thus, although the vagus nerve carries fibers that bear peptidergic transmitters as well as acetylcholine, these data indicate that acetylcholine is the dominant neurotransmitter. The role of peptidergic efferent fibers in sham feeding is largely unknown.
Sham feeding is also a major stimulus of gastric secretion which may contribute to stimulation of pancreatic secretion through the release of secretin. Interestingly, it has been shown that mental stress produced by intense problem-solving can also stimulate pancreatic enzyme secretion in humans.
The regions of the dorsal and ventral anterior hypothalamus, including the medial hypothalamus, dorsomedial and ventromedial nuclei, and mammillary bodies, appear to generate signals for pancreatic secretion. Determination of the neurotransmitters and peptides that are involved in regulating these processes has been approached by examining effects of substances administered into the central nervous system. In rats, central administration of beta-endorphin, CGRP, and CRF inhibit pancreatic secretion. In contrast, TRH stimulates pancreatic secretion through the vagus nerve and involves both muscarinic and VIP receptors.
The gastric phase of pancreatic secretion has been difficult to study in unanesthetized intact animals and humans. Other than testing the effects of gastric distention by installing inert substances or balloon dilation of the stomach, it has been problematic to examine the effects of foods or other nutrients on pancreatic secretion because of the chemical properties of the nutrients themselves that stimulate neural reflexes and cause the release of hormones.
Gastrin is the best-studied gastrointestinal hormone and is a major regulator of gastric acid secretion. Although early reports suggested that gastrin was also a potent stimulus to pancreatic secretion these conclusions have been shown to be incorrect since plasma gastrin levels that are required for stimulation of pancreatic secretion are considerably higher than those that occur after a meal.
The gastric phase is responsible for about 10% of the pancreatic secretory response to a meal. This phase of pancreatic secretion is mediated primarily by gastropancreatic reflexes. Balloon distention of the stomach stimulates pancreatic secretion that is rich in pancreatic enzymes which is blocked by atropine or truncal vagotomy. Therefore, it appears that gastric contributions to pancreatic secretion are mediated by vagovagal cholinergic reflexes that originate in the stomach and terminate in the pancreas.
The stomach is also important in preparing food for delivery to the intestine where nutrients can stimulate the intestinal phase of pancreatic secretion. By the action of pepsin and gastric lipases, proteins are digested to peptides and triglycerides to fatty acids and monoglycerides, respectively. More extensive degradation of protein by enzymes other than pepsin does not further increase the pancreatic stimulatory activity of pepsin digests, suggesting that gastric digestion of protein is sufficient to produce protein products that initiate the intestinal phase of pancreatic secretion. In clinical situations, release of pancreatic enzymes is reduced in humans who have had gastric operations that alter gastric digestion or emptying. Of course, gastric emptying rates are critical for the delivery of nutrients to the intestine that is involved in the stimulation of neural reflexes and hormones that regulate pancreatic secretion.
The intestinal phase of pancreatic secretion begins when food and chyme empty from the stomach into the intestine. Under normal conditions, the pancreas is already primed by cephalic and gastric influences that have increased blood flow to the pancreas and initiated secretion. The intestinal phase which accounts for 50%–80% of pancreatic exocrine secretion is easier to study than the other phases since solutions can be instilled directly into the intestinal lumen. The interactions of various food components such as fats, proteins, carbohydrates, and their breakdown products with neural and hormonal factors are complex. In the intestine, pancreatic secretions serve two major purposes. First, pancreatic bicarbonate neutralizes gastric acid delivered to the duodenum. Second, pancreatic enzymes break down proteins, fats, and carbohydrates to their constituent components that are ultimately absorbed but during the process can initiate processes that influence pancreatic secretion. The intestinal phase of pancreatic secretion can contribute as much as 70% to the postprandial secretory response.
Role of Gastric Acid
Most studies investigating the role of acid on pancreatic bicarbonate secretion have been performed by instilling acid solutions into various regions of the small intestine. Instillation of HCl into the duodenum is a very potent stimulus of pancreatic bicarbonate secretion. However, gastric acid that is delivered to the intestine after a meal is strongly buffered by food, primarily proteins. The pH of the duodenum is 2.0–3.0 in the first few centimeters, however, there is a steep gradient and the pH rises to 5.0–6.0 in the mid-duodenum. The increase in pH in the more distal portion of the duodenum is due to pancreatic bicarbonate secretion stimulated in large part by the release of secretin from the intestinal mucosa. Although instilling acid solutions into the duodenum is not the same as delivery of gastric acid normally produced by the stomach, there is substantial experimental evidence that gastric acid-induced release of secretin following a meal stimulates pancreatic secretion. It has been shown that the pancreatic bicarbonate response to a meal is twofold greater in dogs in which the pancreatic juice has been diverted from the intestine, indicating that pancreatic juice in the intestine is necessary to neutralize the intestinal contents and that the lack of this neutralization results in greater bicarbonate secretion. In addition, administration of the histamine H2 receptor blocker, cimetidine, that blocks gastric acid production, has been found to substantially reduce pancreatic bicarbonate response to a meal. Interestingly, the pancreatic bicarbonate response to a liquid meal is related to the amount of free, unbuffered H + entering the duodenum rather than the total amount of buffered acid. Finally, in dogs, maintaining the pH of a liquid gastric meal above 4.5 resulted in little pancreatic bicarbonate secretion, however, secretion increased substantially as the pH values were lowered. Therefore, it appears that there is a pH threshold of 4.5 in the intestine that is important for the stimulation of pancreatic secretion and that the bicarbonate response is proportional to the load of acid entering the intestine. The pancreatic response to acid is also dependent on the length of small intestine that is exposed to a pH below 4.5.
Contributions of Proteins, Peptides, and Amino Acids to Pancreatic Secretion
There appears to be some species differences in the ability of dietary protein to stimulate pancreatic secretion. In dogs, intact proteins such as casein, albumin, and gelatin did not stimulate pancreatic secretion. However, enzymatic digestion of proteins into small peptides and amino acids converts them into effective stimulants of pancreatic enzyme secretion. Amino acids and peptides are only weak stimulants of pancreatic fluid and bicarbonate secretion but are more potent stimulants of pancreatic enzymes. The aromatic amino acids phenylalanine and tryptophan appear to be the most potent in dogs and humans. Moreover, only L-amino acids can stimulate pancreatic secretion which is consistent with the overall metabolic importance of these stereoisomers.
Although under experimental conditions amino acids can stimulate pancreatic secretion, overall, peptides may be the more physiologically relevant secretagogues since small peptides are much more abundant than amino acids in the lumen of the intestine after a meal. Di- and tri-peptides containing phenylalanine and tryptophan are effective stimulants of pancreatic secretion as are longer peptides generated by pepsin digestion of proteins. Despite expression of the intestinal oligopeptide transporter PepT1 in CCK cells, recent studies suggest that protein hydrolysates do not stimulate CCK release directly through this mechanism.
The amount of pancreatic secretion produced by intraluminal administration of amino acids or peptides is much less than that produced by maximal doses of exogenously administered CCK. This finding indicates that either (1) intraluminal amino acids or peptides are incapable of stimulating maximal release of hormones or neural signals that stimulate secretion or (2) inhibitors of secretion are also produced along with the stimulatory signals. It has recently been demonstrated that the aromatic amino acids phenylalanine and tryptophan act directly on intestinal CCK cells to stimulate CCK secretion. Importantly, this action is mediated by amino acid activation of the calcium-sensing receptor (CaSR) which is highly expressed in intestinal I cells. It is not yet known whether amino acids interact with CaSR on the apical surface of the cell or with receptors on the basolateral surface following nutrient absorption.
The mechanisms by which intestinal factors stimulate pancreatic secretion are incompletely understood, however, the release of CCK into the circulation and the stimulation of cholinergic reflexes are thought to be most important. Only recently are the precise cellular and molecular processes by which amino acids or peptides interact with cells of the intestinal mucosa being defined. For example, it has been proposed that there are specific receptors or transporters on enterocytes that bind amino acids or peptides and generate intracellular signals stimulating hormone release or a neural reflex. Recent in vitro studies on isolated intestinal cells indicate that proteins and peptides have little effect on CCK release and other factors are probably involved (see releasing factors).
The pancreatic response to intraluminal infusion of amino acids is concentration dependent. It appears as though amino acid concentrations of 3–8 mM are necessary to stimulate pancreatic secretion. In addition, the pancreatic response to amino acids is dependent on the entire load of nutrients, not just the concentration of nutrients. Only the proximal small intestine is involved in the stimulatory actions of pancreatic secretion. Once nutrients are introduced into the distal jejunum and ileum other hormones and neural reflexes are activated which inhibit pancreatic secretion and gastric function. This inhibitory action is known as the “ileal brake.” Addition of acid to amino acid or peptide preparations potentiates the pancreatic bicarbonate response but does not affect pancreatic enzyme.
The pancreatic response to dietary protein differs in the rat. Intestinal or intragastric administration of certain proteins such as casein or soy protein potently stimulates pancreatic enzyme secretion. However, hydrolyzed casein does not stimulate pancreatic secretion and amino acids are much less effective than in other species. The effects of proteins on pancreatic secretion appear to be mediated by the release of CCK.
Contributions of Dietary Fat to Pancreatic Secretion
Fatty acids with > 8 carbons stimulate both enzyme and bicarbonate release in dogs, rats, and humans. It also appears that fatty acids are effective stimulants of pancreatic secretion only when in a micellar form. Following lipase digestion of fatty acids in humans, monoglycerides stimulate the pancreas but glycerol does not. Not all fatty acids are equal in their ability to stimulate pancreatic secretion. In humans, fatty acids of 8, 12, and 18 carbon atoms are effective stimulants and the order of potency for stimulating enzyme output is C18 > C12 > C8. The explanation for differences in potency is not entirely clear but it does not appear to be due to rates of fatty acid absorption. In STC-1 cells, which have been used as a model of CCK cells, in vitro fatty acids of medium chain length were shown to stimulate CCK release, raising the possibility that the effects of fatty acids in the intestine are mediated through the release of CCK. However, oleate did not stimulate CCK release from isolated rat intestinal cells containing CCK without producing nonspecific effects on LDH release (a sign of cell toxicity). Delivery of fatty acids in micellar form to rat intestinal cells in vitro was not performed.
Recent evidence suggests that fat-stimulated CCK release may be mediated by one or more G protein-coupled receptors that are activated by long-chain fatty acids. On their discovery, GPR40 and GPR120 (now known as free fatty acid receptors 1 and 4, respectively) were designated orphan receptors and only later were they localized on endocrine, “nutrient-sensing” cells including CCK cells. GPR40-deficient mice were found to exhibit reduced CCK release following long-chain fatty acid exposure and this effect appeared to be mediated through CCK cells. However, the mechanisms resulting in fat-stimulated CCK release may be more complicated than previously realized. It has recently been demonstrated that a non-G protein-coupled receptor, immunoglobulin-like domain containing receptor 1 (ILDR1), is expressed by CCK cells and mediates intestinal fatty acid-induced elevation of blood CCK levels in wild-type mice but not in Ildr1 -deficient mice. Also, the uptake of fluorescently labeled lipoproteins in ILDR1-transfected CHO cells and release of CCK from isolated CCK cells required a unique combination of fatty acid plus high-density lipoprotein (HDL). Although full characterization of gut endocrine cells has not been completed, it is likely that fatty acids in the gut may stimulate hormone secretion through specific free fatty acid receptors and ILDR1.
Intestinal fatty acids also produced a robust pancreatic bicarbonate response that was as high as 70% of the maximal response to exogenous secretin in dogs. In contrast to the effects seen on enzyme secretion, the order of fatty acid chain length on bicarbonate secretion is the reverse with shorter chain fatty acids (e.g., C8) being more potent than longer chain fatty acids (e.g., C18). Fatty acid stimulation of bicarbonate secretion occurs at a neutral or alkaline pH and fatty acids do not interact with acid to potentiate bicarbonate secretion.
Role of Bile Acids in Regulating Pancreatic Secretion
Under normal conditions, bile acids are secreted into the intestinal lumen where they form micelles with fatty acids, triglycerides, and phospholipids. Consequently, free bile acids are in low concentrations in the intestine. It is possible that bile acids interact with the intestinal mucosa to elicit some response. Alternatively, bile acids may solubilize triglycerides and their digestion products that could affect pancreatic secretion. However, the overall importance of bile acids in regulating pancreatic secretion is not well understood. In humans, intraduodenal infusion of bile and the bile salt sodium-taurodeoxycholate stimulated secretin release, therefore, it is possible that any effects of bile on pancreatic bicarbonate and fluid secretion may be due to release of secretin. The effects of bile on pancreatic enzyme secretion was inhibited by atropine indicating that a cholinergic mechanism is involved. Bile does not modify the pancreatic response to exogenously administered CCK or secretin .
In rats, diversion of bile from the intestine stimulates pancreatic secretion. This may be due to the destruction of intraluminal pancreatic enzymes in the absence of bile, which leads to increased levels of endogenously produced releasing factors that exert a positive effect on pancreatic secretion through the release of CCK. This phenomenon of feedback regulation is easily demonstrated in the rat and differs somewhat in other species (see discussion below).
Distention of the intestine with a balloon inhibits pancreatic secretion in conscious but not anesthetized animals. It was reported that distention-induced inhibition of pancreatic secretion could be blocked by topical applications of lidocaine or intravenous atropine indicating that neural nonadrenergic pathways were involved.
Osmolality may also influence pancreatic secretory responses. Infusions of mannitol at concentrations up to 520 mOsm/kg were shown to stimulate pancreatic secretion to levels approximately 20% of maximal. However, higher levels of osmolality may actually inhibit pancreatic secretion.
Calcium and magnesium salts perfused into the intestine of dogs or humans stimulated pancreatic secretion to levels similar to those of maximal doses of CCK. It is possible that magnesium stimulates CCK release since gallbladder contraction was also observed. It is also conceivable that absorbed calcium and magnesium stimulate the pancreas directly.
Absorbed Nutrient Phase
The absorbed nutrient phase refers to the concept that nutrients once absorbed from the intestine may directly stimulate pancreatic secretion. Such an action would represent a direct effect of nutrients on the pancreas or an indirect effect through the ability of absorbed nutrients to stimulate the release of hormones or neurotransmitters that may affect pancreatic exocrine secretion.
Amino Acids and Fats
The effects of intravenous infusion of amino acids and fatty acids have been controversial. Infusion of a mixture of amino acids has been shown to stimulate pancreatic enzyme secretion in some studies and inhibit secretion in others.
Infusion of glucose to produce hyperglycemia actually inhibited pancreatic enzyme and bicarbonate secretion. Large amounts of calcium infused intravenously have been shown to stimulate pancreatic enzyme secretion.
Overall, the evidence that nutrients absorbed after a meal may have significant effects on pancreatic exocrine secretion are weak. However, the studies described above are provocative and illustrate the need to correlate levels of amino acids, lipids, and glucose that occur postprandially with those achieved after intravenous infusion in order to accurately assess the effects on pancreatic secretion.
Neural and Hormonal Regulators
For many decades it was thought that the vagus nerve and the hormones CCK and secretin were the only regulators of pancreatic secretion. However, it is now clear that multiple neural pathways involving a number of different neurotransmitters and neuropeptides as well as a variety of hormones can influence pancreatic secretion in either a stimulatory or inhibitory manner. These pathways will be discussed in the following section.
The gastrointestinal tract and the pancreas are the only peripheral organs that have a well-developed neural plexus. The intrinsic nerve plexus of the pancreas receives inputs from both the parasympathetic and sympathetic nervous systems. Parasympathetic nerves innervating the pancreas originate primarily in the dorsal vagal nucleus of the brain. Most of these fibers run through the posterior vagal nerve trunk and terminate as preganglionic vagal nerves on pancreatic ganglia. Sympathetic innervation of the pancreas is supplied by neurons with cell bodies in the celiac and superior mesenteric ganglia. These nerve cells send fibers to the nerve cell bodies in pancreatic ganglia, acinar and duct cells, blood vessels, and islets. However, the sympathetic innervation of the exocrine cells of the pancreas is less than that of blood vessels and islets.
Pancreatic ganglia integrate neuronal inputs of both the exocrine and the endocrine pancreas. These ganglia are innervated by vagal preganglionic, sympathetic postganglionic, enteric neurons, and sensory fibers. Postganglionic nerves containing cholinergic, noradrenergic, peptidergic, and nitrergic transmitters surround most acinar clusters and regulate exocrine secretion.
A major role for intrapancreatic cholinergic neurons being involved in stimulation of pancreatic secretion has evolved from numerous experimental studies. These neurons receive input from both the central nervous system through the cephalic phase of pancreatic secretion and via vagovagal reflexes that involve both the afferent and efferent limbs of the vagus nerve. Vagovagal enteropancreatic reflexes play a major role in the control of the intestinal phase of pancreatic secretion. These reflexes involve afferents originating in the duodenal mucosa, and efferents mediating central input on the pancreatic ganglia, activate intrapancreatic postganglionic neurons. There is considerable experimental evidence for these pathways. Vagovagal reflexes are initiated by gastric and intestinal stimulation and converge on cholinergic intrapancreatic neurons that release acetylcholine. Acetylcholine, in turn, binds to muscarinic receptors on pancreatic acinar and duct cells to stimulate both enzyme and bicarbonate secretion. The effects of acetylcholine potentially are aided in most species by the hormones secretin, VIP, and perhaps gastrin releasing peptide (GRP). Furthermore, VIP released from peptidergic nerves in the pancreas also regulates pancreatic secretion.
Electrical stimulation of the vagus nerve trunks activates efferent vagal nerve fibers causing pancreatic secretion. In most species, with the pig being an exception, electrical stimulation of the vagus nerve causes greater release of pancreatic enzymes relative to fluid and bicarbonate. Similar to electrical stimulation, insulin-induced hypoglycemia or administration of 2-deoxyglucose also potently stimulates pancreatic secretion. Approximately 50% of the maximal CCK-induced response can be produced by vagal stimulation. Vagotomy eliminates most of the effects of insulin or 2-DG on pancreatic secretion. It is believed that each of these methods to stimulate the vagus nerve causes the release of acetylcholine which activates muscarinic receptors on the pancreas as each action can be blocked by atropine. Conversely, treatment of animals with acetylcholine or other muscarinic cholinergic agonists stimulates pancreatic enzyme secretion. A number of other observations have clarified the role of cholinergic innervation in the pancreas. Intrapancreatic nerves and ganglia, identified by histological examination and electron microscopy, have been found to lie in close approximation to both pancreatic ducts and acinar cells. Moreover, the pancreas contains the enzymes choline acetyltransferase and acetylcholinesterase which are involved in the synthesis and inactivation, respectively, of acetylcholine. Stimulation of vagal nerve trunks even stimulates the isolated pancreas. Using radiolabeled antagonists, specific muscarinic receptors have been identified on pancreatic acinar cells and choline acetyltransferase and vesicular acetylcholine transporter have been described in stellate cells. All of these data provide strong support for cholinergic stimulation of pancreatic enzyme and bicarbonate secretion. At the level of the brainstem, modulation of vagovagal reflexes may occur by inputs from higher brain centers.
During the cephalic phase of secretion, afferent input to the vagus nerve comes from olfactory and gustatory receptors as well as higher centers. The gastric and intestinal phases of the meal response activate both chemical and stretch receptors. Together these stimuli generate a response that is manifest by vagal efferent signaling. Interestingly, it is still incompletely understood how the specific chemical signals such as acid, amino acids, peptides, fatty acids, monoglycerides, and divalent cations actually activate the vagus nerve. These effects are partially mediated by the release of hormones released from the intestinal mucosa, but other factors also appear to be involved. Electrical activation of sites within the brain and the vagus has been recorded after gastric distention and intestinal perfusion with amino acids.
The effects of atropine on pancreatic secretion have been fairly uniform throughout all species tested. Atropine decreased pancreatic enzyme secretion more than it reduced fluid or bicarbonate. Surgical vagotomy also had a selective effect on reducing enzyme release and reduced both basal and stimulated secretion. The observation that atropine further reduced basal pancreatic secretion following vagotomy suggests that intrapancreatic neurons were still active following vagotomy.
Pancreatic secretion stimulated by intestinal perfusion with amino acids, fatty acids, and peptides can be inhibited by vagotomy and anticholinergic drugs. The secretory response of the autotransplanted pancreas was only 40%–50% of that of the intact pancreas. Although it has been proposed that cholinergic influences mediate the release of CCK there is little experimental evidence to suggest that CCK release is regulated directly or indirectly by cholinergic nerves. However, topical application of anesthetic drugs to the intestine blocks the pancreatic enzyme response to intestinal perfusion with amino acids. In the autotransplanted pancreas in which the gland is entirely denervated, the response to CCK is unaltered.
Furthermore, enzyme secretion in the autotransplanted pancreas in response to intestinal stimulants is unaltered by vagotomy or atropine, whereas the intact pancreas is strongly inhibited by each of these maneuvers. These findings indicate that the effects of vagotomy and atropine on the intact pancreas are not due to the inhibition of CCK release or other hormones. Rather the inhibitory effects of vagotomy and anticholinergic drugs on pancreatic enzyme secretion in response to intestinal stimulants are mediated by enteropancreatic reflexes rather than a direct effect on CCK secretion. Quantitatively, approximately one-half of the enzyme response to intestinal stimulants such as amino acids and fatty acids is mediated by neural pathways that are largely cholinergic and vagovagal. However, when the effects of different loads of intestinal stimulants are considered it appears that vagal cholinergic reflexes are the major mediators of pancreatic secretion in response to low intestinal loads but hormones may mediate pancreatic responses to high loads of intestinal stimulants.
Anatomic evidence of enteropancreatic innervation has been provided by histochemical studies in the rat demonstrating that enteric neurons actually project to pancreatic ganglia. Binding studies using isolated pancreatic acinar cells and specific muscarinic receptor antagonists demonstrate that M1 and M3 receptors reside on pancreatic acinar cells. These studies have been confirmed by demonstration of mRNAs for each receptor subtype. M1 receptors are also involved in regulating pancreatic secretion probably through a nonacinar presynaptic mechanism.
In summary, there is good evidence for direct cholinergic stimulation of pancreatic enzyme secretion that mediates the cephalic, gastric, and intestinal phases of pancreatic enzyme secretion in response to a meal. Moreover, cholinergic reflexes potentiate the bicarbonate response to secretin. Enteropancreatic vagovagal cholinergic reflexes are the primary mediators of pancreatic enzyme secretion in response to low loads of proteins, amino acids, fatty acids, and HCl. These data should not be taken as evidence, however, that hormonal influences, primarily those of CCK, do not also play an important role in regulating meal-stimulated pancreatic secretion. The CCK is the major mediator of pancreatic enzyme secretion when high loads of protein, amino acids, and fatty acids are instilled into the intestine. The interrelationship between CCK and cholinergic influences regulating pancreatic secretion will be discussed below. Sympathetic nerves of the pancreas do not appear to mediate pancreatic secretory responses to intestinal stimuli ( Fig. 40.1 ).
Vasoactive Intestinal Peptide
Electrical stimulation of the vagus nerve trunks stimulates pancreatic enzyme and bicarbonate secretion. This stimulation achieves pancreatic secretory levels approximately 50% of the response seen with exogenous secretin. In addition to effects on secretion, electrical stimulation of the vagus nerve also increases pancreatic blood flow and local vasodilation. A body of work has now implicated VIP as the transmitter most likely responsible for fluid and bicarbonate secretion and vasodilation.
Data to support that VIP as a neural peptide involved in regulating pancreatic fluid and bicarbonate secretion came from two primary observations. First VIP-containing neurons are found within the pancreas. These fibers surround intrapancreatic ganglia and pancreatic ducts. Second, VIP can directly stimulate pancreatic bicarbonate secretion from the isolated vascularly perfused the pancreas. Subsequent studies demonstrated that electrical stimulation of the vagus nerve following atropine treatment resulted in an increase in bicarbonate secretion and an increase in VIP in the venous effluent from the pancreas. In addition, the pattern of VIP release paralleled the secretion of fluid and bicarbonate and the amount of VIP in pancreatic venous effluent was similar to that required to stimulate pancreatic secretion with intra-arterial injection. Nevertheless, it is difficult to conclude that VIP plays a major role in regulating pancreatic fluid and bicarbonate secretion in vivo, and it is most likely that it serves as a neurotransmitter that is involved in mediating pancreatic vasodilation and in increasing blood flow. Further studies indicated that multiple mediators are involved in the pancreatic secretory response to electrical stimulation of the vagus nerve.
Gastrin Releasing Peptide
Gastrin releasing peptide (GRP) immunoreactivity has been identified in nerve fibers surrounding pancreatic acini and ducts and nerve cell bodies of intrapancreatic ganglia. The GRP strongly stimulates pancreatic enzyme, fluid and bicarbonate secretion in humans, pigs, dogs, and rats, and enzyme secretion in vitro in mice, rats, and guinea pigs. In pigs, the response to GRP is approximately 40% of the secretion produced by maximal amounts of secretin and nearly 100% of the enzyme response to CCK. Moreover, the GRP receptor has been identified in pancreatic cancer cells and mediates enzyme secretion. However, the effects of GRP appear to be species specific since the GRP analog, bombesin, does not stimulate pancreatic secretion in the dog. For many years, it was believed that the effects of GRP on the pancreas were mediated by the release of CCK, however, this does not appear to be true since GRP’s effects persist even in the presence of CCK receptor blockade.
It appears that GRP stimulates release of acetylcholine from postganglionic intrapancreatic nerves in addition to stimulating acinar cells directly. Electrical stimulation of the vagus nerve stimulated pancreatic enzyme secretion and increased GRP in venous effluent from the pancreas. Pancreatic enzyme secretion in response to vagal stimulation was also inhibited by GRP desensitization, GRP antagonists, or GRP immunoneutralization. The GRP and its analog bombesin also have been shown to stimulate acetylcholine release. In addition, tetrodotoxin inhibited GRP-stimulated amylase release by over 70% in rat pancreatic lobules.
Other Peptide Neurotransmitters
Immunohistochemical staining has demonstrated a number of gastrointestinal peptides in nerves of the pancreas. In vitro studies in isolated pancreatic acini have also shown that these peptides may have stimulatory or inhibitory activities. The CCK, GRP, neurotensin, peptide histidine isoleucine, and CGRP have all been found in cell bodies of intrapancreatic neurons and stimulate pancreatic secretion from acinar cells. Inhibitory peptides include substance P, enkephalin, and NPY. The NPY is released from the porcine pancreas following stimulation of the vagus nerve or splanchnic nerves innervating the pancreas. The NPY has no effects on basal pancreatic secretion and only modest inhibitory effects on secretin- or CCK-stimulated secretion. However, it does increase vascular resistance in blood vessels of the pancreas and it is believed that NPY is involved in the regulation of pancreatic blood flow.
Various molecular forms of CCK have been identified in the pancreas, however, it is unclear if there is an important role for CCK as an intrapancreatic neurotransmitter that regulates pancreatic exocrine secretion since there was no detectable CCK in the venous effluent of the isolated perfused pig pancreas following vagal stimulation.
Adrenergic nerves containing the neurotransmitter norepinephrine are believed to exert inhibitory influences on pancreatic secretion. However, studies to establish these effects have been difficult because of the many diverse effects of sympathetic nerve stimulation or nonspecific effects following administration of adrenergic agents. The cell bodies of norepinephrine-containing nerves that innervate the pancreas are found in the celiac ganglion. Fibers from these cells extend to nerves in the intrapancreatic ganglia and to blood vessels, ducts, and islets. Using immunohistochemical techniques it appears as though there is little noradrenergic innervation of acinar cells. Electrical stimulation of splanchnic nerves to the pancreas inhibited pancreatic secretion. Conversely, cutting splanchnic nerves was reported to increase the pancreatic secretion.
Celiac denervation has been shown to reduce pancreatic secretion by ~ 70% while increasing blood flow by 350%. This finding suggests that there was disruption not only of stimulatory fibers but also of sympathetic fibers that maintain tonic constriction of pancreatic vessels.
Functional studies examining the effect of adrenergic transmitters on pancreatic secretion have been fraught with hazard due to the wide-ranging effects of norepinephrine on multiple processes including blood pressure, blood flow, various neural reflexes, and release of hormones. Various reports describe norepinephrine stimulating, inhibiting, or having no effect on pancreatic secretion. Overall, the effects of alpha and beta receptor antagonists have not been helpful in increasing our understanding of adrenergic regulation of pancreatic secretion.
Norepinephrine is believed to be the primary transmitter released by intrapancreatic adrenergic nerves. In vitro studies are difficult to extrapolate to the in vivo situation and in vivo studies are complicated by the widespread effects of adrenergic agonists and antagonists. Nevertheless, in humans, a selective beta 2 adrenergic receptor agonist has a weak inhibitory effect on CCK-stimulated enzyme secretion. Although pharmacologic effects can be demonstrated in vitro, currently there is insufficient evidence to conclude that adrenergic regulation of pancreatic exocrine secretion is important other than for inhibiting pancreatic juice and bicarbonate secretion that likely occurs through direct and indirect effects on pancreatic blood flow.
Overall dopamine has a secretin-like effect on the pancreas and stimulates fluid and bicarbonate secretion with little effect on pancreatic enzyme output. In dog, the effect of dopamine is blocked by dopamine receptor antagonists but is unaffected by alpha or beta receptor antagonists. Pancreatic acinar cells have been shown to express receptors for dopamine. These receptors are linked to the activation of adenylate cyclase and stimulation of fluid and bicarbonate secretion. However, dopamine causes very little enzyme secretion from either acinar cells in vitro or the pancreas in vivo . In rats and dogs, high doses of dopamine stimulate secretion of pancreatic fluid. This effect is inhibited in rats by propranolol but not by the dopamine receptor antagonist haloperidol, indicating that the effects of dopamine are mediated through the beta adrenergic receptor rather than the dopamine receptor. Dopamine has been reported to inhibit pancreatic enzyme secretion in rats. In humans, dopamine has been reported to have little effect alone. Recently, dopamine containing neurons have been identified in pancreatic tissue.
Serotonergic nerves of the pancreas appear to originate in the intestine. Intrapancreatic nerves have been shown to take up serotonin and pancreatic acinar cells have been shown to remove serotonin from plasma and secrete serotonin suggesting that serotonin release occurs physiologically within the pancreas. In anesthetized rats, the 5-HT2 serotonin antagonist ketanserin and the 5-HT3 antagonist ondansetron each inhibited secretin-stimulated pancreatic volume and bicarbonate secretion, suggesting that serotonin may mediate secretin’s effects through both 5-HT2 and 5-HT3 receptor subtypes. Interestingly, both of these serotonin antagonists have been reported to also inhibit secretin release induced by duodenal acid. Nevertheless, despite these recent advances, the overall functional significance of serotonin in pancreatic secretion still remains largely unknown.
Nitric oxide is involved in the control of a number of physiologic and pathophysiologic functions in the gastrointestinal tract. In the pancreas, nitric oxide can affect secretory activity and pancreatic blood flow. As a volatile gas, it is not practical to directly measure nitric oxide in biological tissues, therefore, studies on the localization of nitric oxide have been done by examining nitric oxide synthase (NOS), the enzyme responsible for the production of nitric oxide. Pharmacologic tools have been used to study the physiologic actions of nitric oxide in animals, organ preparations, and cells. The actions of nitric oxide have been deduced by using NO donors, NOS inhibitors, agents that can inactivate nitric oxide (such as superoxide-generating compounds) or agents that stabilize nitric oxide (e.g., superoxide dismutase). Nitric oxide is unusual compared to other signaling molecules that bind to receptors in the cell membrane since nitric oxide can penetrate cells to interact directly with what is considered to be its primary target, guanylate cyclase. In this way, nitric oxide activates the enzyme to produce cGMP.
Immunohistochemical studies using antisera against NOS isoforms have shown that the enzyme responsible for the synthesis of nitric oxide is abundant in nerve fibers of the pancreas of all species including the mouse, rat, chicken, cat, monkey, hamster, guinea pig, and human. Neuronal NOS (nNOS) positive nerve fibers are found in intrapancreatic ganglion cells with many also containing other neuropeptides including VIP. These nerve fibers course through interlobular and interacinar spaces of the pancreas. Nerve fibers of either intra- or extra-pancreatic ganglion cells including nitrergic enteropancreatic neurons and viscerosensory fibers also contain nNOS. These nerve fibers are found in perivascular, periacinar, and periductal regions of the exocrine pancreas and surround and innervate pancreatic islets. Islet cells may co-store NOS and somatostatin. A constitutive form of NOS found in endothelial cells (eNOS) has been localized to vascular endothelium of the pancreas. Some acinar cells and duct endothelial cells have been shown to have nNOS-like immunoreactivity. The nitrergic supply of the exocrine pancreas and blood vessels is similar in every species examined.
Ascertaining the functional role of nitric oxide in regulating pancreatic secretion has been complicated by the wide distribution of nitric oxide-producing cells and the wide range of effects of nitric oxide on a number of different cell types and tissues. In vivo studies are often difficult to interpret because of the effects of nitric oxide on blood flow, neurotransmission, and release of other hormones and transmitters. In general, in vivo studies indicate that nitric oxide stimulates exocrine pancreatic secretion. In humans, the NOS inhibitor, L-NMMA, in a dose-dependent manner, reduced pancreatic enzyme but not bicarbonate or fluid secretion stimulated by secretin and caerulein. This effect was also reversed by L-arginine. In conscious dogs prepared with chronic pancreatic fistulae, it was found that pancreatic protein output induced by either sham feeding, a meal, or infusion of secretin plus CCK were all inhibited by the NOS inhibitor L-NOARG. Nitric oxide may exert tonic effects to inhibit intestinal motility and stimulate pancreatic secretion during the basal state. This concept is supported by studies in rats in which basal pancreatic secretion is attenuated by NOS inhibitors. In conscious rats, secretin-stimulated pancreatic secretion was inhibited by the NO synthase inhibitor, N-nitro-L-arginine, an effect that was reversed by L-arginine.
Nitric oxide donors have been used to test the effects of exogenous nitric oxide on pancreatic secretion. Both glyceryl trinitrate and sodium nitroprusside increased basal pancreatic secretion in conscious dogs and anesthetized cats but had no effect on meal-stimulated pancreatic responses. The mechanism by which nitric oxide may affect pancreatic exocrine secretion is still a matter of controversy. In rats and dogs, it has been shown in vivo that attenuation of pancreatic secretion with NOS blockade was associated with a reduction in pancreatic blood flow. However, additional studies indicate that other mechanisms are probably involved. It has been shown that the effects of NOS inhibitors on pancreatic secretion do not always parallel changes in blood flow.
Evidence of nitric oxide having a direct effect on pancreatic secretion comes from studies on isolated pancreatic acini and lobules. The treatment of pancreatic acini with the NOS substrate L-arginine increased nitrite and cGMP levels and amylase release. These effects were blocked by L-NMMA and L-NOARG. The effects of nitric oxide on pancreatic acinar cell secretion appear to be mediated by an increase in cGMP.
In summary, these findings indicate that nitric oxide may have direct stimulatory effects on pancreatic acinar cell secretion and indirect effects by increasing pancreatic blood flow or through modulation of intrapancreatic parasympathetic nerves.
Intestinal stimulants induce secretion from both intact pancreas and denervated, autotransplanted pancreas suggesting that pancreatic secretion is under the regulation of circulating hormones. To function as a hormone that stimulates pancreatic secretion several criteria must be met. First, if such a hormone plays a physiologic role in postprandial pancreatic secretion it must be released into the circulation following ingestion of the meal. Second, levels of the hormone in the blood that occur after a meal should be similar to the levels that are required for an exogenously infused hormone to affect pancreatic secretion. Third, strong evidence for the existence of a stimulatory regulator of pancreatic secretion would be the demonstration that administration of a specific hormone receptor antagonist or an antiserum that blocked the hormone in question reversed the stimulatory effects on pancreatic secretion ( Fig. 40.2 ).
Although we now appreciate that hormones interact with neural factors, hormonal transmitters provide a major stimulus to pancreatic secretion. Historically, secretin and CCK have been considered the major hormones regulating pancreatic secretion but other hormones may significantly modulate these effects.
Secretin is the most potent stimulant of pancreatic fluid and bicarbonate secretion. Produced by enteroendocrine cells, known as S cells, of the upper small intestine, secretin is released during the intestinal phase of a meal. Acid secreted from the stomach into the duodenum is the major stimulant of secretin although ingested fatty acids may also contribute to secretin release. One of the major difficulties in studying secretin physiology has been the lack of sufficiently sensitive and specific secretin radioimmunoassays. It is likely that many of the early reports on secretin levels in plasma were performed with insufficiently sensitive or specific assays. Consequently, it was difficult to measure the experimentally induced or postprandial secretin response. More recently, however, improved assays have led to the conclusion that postprandial secretin levels can be accurately measured in experimental animals and humans. It appears as though the postprandial secretin response is small in humans amounting to approximately a doubling of basal levels. It would be natural to ask if such small increases in plasma secretin levels are sufficient to affect pancreatic exocrine secretion. Multiple studies have demonstrated that the pancreas is very sensitive to secretin. Significant increases in pancreatic fluid and bicarbonate secretion have been demonstrated with continuous intravenous secretin infusions as low as 1–2.8 pmol/kg ∙ h. Bolus injections of secretin as low as 0.125 pmol/kg stimulated fluid and bicarbonate secretion in humans. Infusion of exogenous secretin that reproduces postprandial secretin blood levels stimulated pancreatic bicarbonate secretion equivalent to 10% of the maximal pancreatic secretory response indicating that other hormones or neurotransmitters contribute to postprandial pancreatic bicarbonate secretion in humans. However, interpretation of these studies may not be straightforward due to the specificities of secretin antisera and the molecular forms of endogenous secretin that may exist in blood. For example, many secretin antibodies do not recognize glycine-extended secretin which has been shown to be present in blood and can stimulate pancreatic secretion. These data indicate that secretin contributes to some but not all of the pancreatic bicarbonate secretion caused by intestinal acid or a meal.
The contribution of secretin to bicarbonate secretion should not be minimized. It is worthwhile noting that the pH threshold in the intestinal lumen for stimulating secretin release and pancreatic bicarbonate are the same. Moreover, the timing for delivery of acid to the duodenum, initiation of pancreatic bicarbonate secretion, and stimulation of secretin release is also similar. In humans who are achlorhydric or in whom gastric acid is neutralized by sodium bicarbonate, plasma secretin levels were low. In dogs, administration of antisecretin antiserum reduced the pancreatic secretory response to both luminal stimulation by sodium oleate and exogenous secretin ( Fig. 40.3 ).
The failure of exogenous secretin to reproduce the entire pancreatic bicarbonate response to intestinal acid or a meal may be due to the contributions of other neurotransmitters and hormones such as acetylcholine and CCK that act synergistically on the pancreas. In several species, electrical stimulation of the vagus nerve potentiated the effects of exogenous secretin. Potentiation of secretin-stimulated pancreatic secretion by basal cholinergic activity may explain why even low levels of secretin that occur after a meal as measured by radioimmunoassay are sufficient to stimulate bicarbonate output.
Insulin plays a permissive role on the action of secretin and CCK. Immunoneutralization of insulin with specific insulin antisera in rats substantially inhibited pancreatic secretion stimulated by either a meal or secretin plus CCK. These findings are likely to be physiologically relevant since acinar cells are exposed to high concentrations of insulin via a vascular portal network in which venous blood flows from islets and bathes pancreatic acini.
Cholecystokinin is released from specific enteroendocrine cells of the small intestine known as I cells. I cells are more abundant in the proximal small intestine and decrease in abundance farther down the small intestine. The major dietary components that stimulate CCK release are fats and protein. In certain species, such as the rat, CCK release is controlled by intraluminal proteases and an active feedback system participates in the release of a putative intestinally secreted CCK-releasing factor. One of the major problems that has plagued the CCK field has been the difficulty in developing sensitive and specific CCK assays. Several radioimmunoassays and a bioassay have been developed that have sufficient sensitivity to detect what are low basal levels of CCK in plasma. Another problem has been the appearance of various molecular forms of CCK in blood. Molecular forms ranging in size from CCK-8 to CCK-58 have been described in dogs, rats, and humans. It is important that assays detect all of the potentially biologically active molecular forms in order to accurately estimate relevant CCK concentrations in blood. More recently, using techniques to prevent CCK degradation in plasma, it was found that the large form of CCK, CCK-58, was the predominant form in dogs and humans and was the only form detected in the rat. . Although some studies suggested that CCK-58 may have activity distinct from shorter forms any differences are unlikely to be of major significance. All biologically active forms of CCK possess an intact carboxyl terminus. Therefore, bioassays that measure biologically active hormone or radioimmunoassays directed at the carboxyl terminus of CCK and can distinguish CCK from gastrin (which shares an identical pentapeptide carboxyl terminus) should provide similar estimates of CCK activity. Basal CCK levels are as low as 0.5–1 pM and increase to 10–16 pM after a mixed meal. Of the specific nutrients that stimulate CCK release, fats and protein or amino acids are the most potent. Although some groups have reported that HCl stimulates CCK release, this has not been found by other investigators. A transient increase in CCK has been observed after the ingestion of glucose but not water or saline.
In studies in which plasma CCK levels and pancreatic exocrine secretion were measured simultaneously with the infusion of exogenous CCK, it was found that pancreatic enzyme secretion could be stimulated by low levels of CCK. The physiologic relevance of this finding in humans has been assessed by comparing plasma CCK levels and pancreatic secretory responses after a meal. The investigators of this study concluded that because plasma CCK levels and pancreatic enzyme secretion after a meal were similar to those that occurred following the infusion of caerulein, endogenously released CCK is a major hormonal regulator of pancreatic exocrine secretion. However, one of the difficulties in interpreting the results of this study was that an average of CCK levels was used to estimate the total CCK responses and individual CCK levels could not be identically reproduced by infusion of exogenous hormone.
Further support for CCK playing an important role in normal pancreatic secretion has come from studies using specific CCK receptor antagonists. The CCK blocker, proglumide, administered prior to intestinal perfusion of either amino acids or a fat emulsion, or the infusion of exogenous CCK-8, significantly decreased pancreatic enzyme secretion. The specificity of this response was further confirmed by the observation that proglumide did not alter pancreatic secretion stimulated by a cholinergic agonist. The CCK antagonist loxiglumide has also been shown to inhibit pancreatic enzyme secretion stimulated by CCK plus secretin. In humans, blockade of CCK receptors with loxiglumide or devazepide decreased pancreatic secretion to 40%–60% of that following a meal.
In vitro, a number of studies have demonstrated that CCK potentiates secretin-stimulated enzyme secretion in pancreatic acini. In humans, pancreatic bicarbonate secretion in response to secretin is potentiated by exogenous CCK. However, studies have not been done to determine whether potentiation of secretin-stimulated pancreatic secretion occurs with physiologic levels of CCK in vivo. It did not appear as though CCK-stimulated pancreatic enzyme secretion was potentiated by secretin in dogs or humans. It should be noted, however, that potentiation of pancreatic secretion by a variety of hormones and transmitters which are thought to be important in the physiologic pancreatic secretion in other species do not appear to be operational in dogs.
It has been demonstrated in anesthetized rats that atropine and hexamethonium almost completely blocked the pancreatic enzyme response to low but not high doses of CCK. Furthermore, transection of vagal afferent nerves also blocked the effects of CCK. In this study, CCK did not increase pancreatic secretion above that produced by bethanechol, thus refuting the proposal that direct stimulation of pancreatic acinar cells by CCK requires a basal vagal cholinergic tone. CCK infusion producing plasma CCK levels similar to those that occur after a meal, stimulated pancreatic enzyme secretion that could be inhibited by atropine ( Fig. 40.4 ). These findings suggest that CCK acts through a neuronal pathway rather than directly on the pancreas to stimulate exocrine secretion. Earlier reports failed to demonstrate that vagotomy reduced CCK-stimulated pancreatic secretion, however, these may be explained by differences in the duration of time from vagotomy to experimentation. The pancreas has a unique ability to regain function following vagotomy and if this is not taken into consideration, differences in secretory responses may change over time which could alter experimental interpretation.
Capsaicin is a neurotoxin that targets small-diameter sensory neural fibers. Local application of capsaicin to the vagus nerve reduced pancreatic secretion in response to physiologic doses of CCK in a manner similar to that of vagotomy or atropine administration. In addition, gastroduodenal, but not jejunal, application of capsaicin blocked pancreatic secretion in response to physiologic doses of CCK. Together these findings indicate that at physiologic concentrations, CCK simulates pancreatic secretion through a capsaicin-sensitive afferent vagal pathway that originates in the gastroduodenal mucosa ( Fig. 40.5 ).
Several anatomical and electrophysiologic studies support a role for CCK acting on the vagus nerve. The CCK receptors have been identified in the rat vagus nerve by autoradiography. Moreover, CCK receptors reside on the presynaptic portion of vagal afferent nerve terminals. In agreement with the functional studies described above, perivagal capsaicin treatment reduced the binding of CCK to the vagus nerve. In rats, CCK-stimulated vagal afferent nerves as recorded by microelectrodes implanted in the nodose ganglia. Additional electrophysiologic and secretion studies used in combination with the CCK analog JMV-180 have characterized both high- and low-affinity vagal CCK receptors. JMV-180 acts as an agonist on high-affinity CCK receptors and as an antagonist on low-affinity CCK receptors. In rats, the CCK-1 receptor antagonist L-364,718 blocked pancreatic secretion stimulated by JMV-180. However, acute vagotomy in anesthetized rats and perivagal capsaicin in conscious rats blocked this response indicating that JMV-180 stimulated pancreatic secretion through a vagal afferent mechanism. Interestingly, JMV-180 did not block pancreatic secretion in response to physiologic doses of CCK. In order to assess the site of action of endogenous CCK, JMV-180 was found to enhance rather than block the pancreatic secretory response to intraduodenal administration of casein. These findings indicate that both endogenous and exogenous CCK-stimulated pancreatic secretion by acting on high-affinity CCK receptors.
Very recently it was shown that many enteroendocrine cells including CCK cells have cytoplasmic extensions named neuropods that project from the basal surface. Neuropods contain abundant mitochondria and possess many neuron-like features including neurofilaments, pre- and postsynaptic proteins, and both large dense and small clear secretory vescicles. Most neuropods extend below the mucosal epithelium and some even penetrate the basal lamina where they connect to subepithelial enteric neurons ( Fig. 40.6 ). Intrinsic primary afferent neurons reside in the subepithelial villus and synapse with neurons in the submucosal and myenteric plexi. It is possible that direct contact of CCK cells with enteric neurons provides an enteroendocrine-neuronal connection that explains the CCK-mediated vagal influence on pancreatic secretion.
Substantial insight into the mechanism by which CCK affects the human pancreas has been provided by the characterization of CCK receptor subtypes. CCK-1 receptors have been identified on rat pancreatic acinar cells using functional, biochemical, and molecular biological methods. Radioligand binding to isolated acini, receptor cross-linking, receptor autoradiography, Northern blot analysis, and in vitro secretion studies all confirmed the existence of CCK-1 receptors. In contrast to the rodent, it has been very difficult to demonstrate CCK-1 receptors in the human pancreas. Very little CCK-1 receptor mRNA was detectable by reverse-transcriptase polymerase chain reaction in the human pancreas. However, CCK-2 receptor was expressed in human pancreas and appeared to be the predominant CCK receptor subtype. Demonstration of CCK-2 receptors on acinar cells of the human pancreas has been difficult. Immunohistochemical studies using specific CCK receptor antisera found most staining on somatostatin-containing D cells of the islets of Langerhans and CCK-1 receptors were only identified on insulin-secreting cells in the fetal human pancreas. Some functional studies also failed to demonstrate CCK stimulation of enzyme secretion from human pancreatic acini although difficulties in obtaining viable tissue brought some results into question. Using isolated human pancreatic acini which demonstrated robust secretory responses, muscarinic cholinergic agonists stimulated increases in both intracellular Ca 2 + and enzyme secretion but CCK had no effect. Further evidence for the lack of functional CCK receptors was provided by studies in which the CCK-2 receptor was artificially inserted into human acinar cells by adenoviral-mediated transfer of CCK receptor DNA. The CCK-stimulated increases in intracellular Ca 2 + were seen after, but not before, CCK receptor gene transfer. Despite these negative findings, it was recently demonstrated that application of physiologic concentrations of CCK-8 and CCK-58 to human acinar cells caused intracellular Ca 2 + oscillations and normal exocytosis of pancreatic enzyme ( Fig. 40.7 ). This latter finding indicates that functional CCK receptors are expressed on human pancreas, thus, the proposed species differences in expression of CCK receptors between human and rodent pancreas may not be as great as previously predicted.
A series of studies have shown the CCK acts in concert with serotonin to stimulate postprandial pancreatic enzyme secretion. In rats, the CCK receptor antagonist, L-364,718 reduced postprandial pancreatic enzyme secretion by 54%, whereas the combination of L-364,718 and ketanserin, a 5-HT 2 antagonist, and ICS-205,930, a 5-HT 3 antagonist, reduced pancreatic secretion by 94%. Electrophysiologic studies have demonstrated that serotonin and serotonin analogs activate 5-HT 3 receptors on nerve terminals within the ferret gastric and duodenal mucosa and stimulate firing of vagal afferent fibers. Recordings from the nodose ganglia were activated by intraluminal osmotic loads or administration of carbohydrates, effects that were antagonized by the 5-HT 3 antagonist granisetron. Even intraluminal administration of serotonin stimulated afferent vagal firing through 5-HT 3 receptors and stimulated pancreatic protein output. These effects were blocked by the 5-HT 3 antagonist ondansetron, acute vagotomy, administration of methscopolamine, or perivagal application of capsaicin. Agents that deplete the mucosa of serotonin such as p-chlorophenylalanine, eliminated neuronal activation within the nodose ganglia normally stimulated by intraluminal factors. Thus, it appears that intraluminal chemical stimuli activate 5-HT 3 receptors on mucosal vagal afferent fibers. In contrast, mechanical stimulation of the mucosa stimulated both 5-HT 3 and 5-HT 2 receptors on vagal afferents to stimulate pancreatic secretion. Together these findings indicate that 5-HT and CCK both activate vagal afferent pathways that originate in the intestinal mucosa and stimulate pancreatic secretion. The ability of CCK to potentiate the actions of 5-HT were demonstrated in studies in which infusion of subthreshold doses of CCK-8 which by themselves did not stimulate pancreatic secretion, when accompanied by intraduodenal infusion of maltose or hyperosmotic saline (that activate 5-HT receptors), produced greater pancreatic secretion than either agent alone. Electrophysiologic studies have identified some neurons within the nodose ganglia possessing high-affinity CCK receptors that respond to intraluminal infusion of serotonin. In these studies, neurons were stimulated with 5-HT and subthreshold doses of CCK that by themselves produced no measurable electrophysiologic effect. The effects of 5-HT were blocked by the CCK receptor antagonist, CR-1409 demonstrating that CCK potentiated the actions of 5-HT. These potentiating actions of CCK and serotonin may explain why even small postprandial increases in plasma CCK levels are sufficient to stimulate pancreatic enzyme secretion. Both CCK-1 and 5-HT 3 receptors have been identified on the same dorsal vagal afferent neurons and it has been suggested that interaction of receptors may exist.
Neurotensin is produced by endocrine cells (N cells) of the intestine, primarily in the ileum, and in nerves, predominantly in the myenteric plexus. It has been demonstrated in several species including humans that neurotensin stimulates pancreatic secretion. Intestinal administration of fatty acids increased plasma neurotensin levels and the pattern of neurotensin-stimulated pancreatic secretion is similar to that stimulated by intestinal fat. However, it is unclear that neurotensin is a physiologic regulator of pancreatic secretion since the amounts of neurotensin infused produced higher plasma neurotensin levels than that occurred after a meal or after intestinal perfusion with fatty acids. In humans, neurotensin stimulated bicarbonate secretion but actually decreased enzyme secretion stimulated by secretin and caerulein. Another group of investigators have found that neurotensin accentuated the enzyme response but not the bicarbonate response to secretin and the bicarbonate response but not the enzyme response to caerulein. In rats, it was observed that neurotensin inhibited basal pancreatic secretion but did not affect secretin- or CCK-stimulated secretion. The conflicting results in humans and the lack of inclusive experimental data in animals, along with the high doses of neurotensin that were required for stimulating pancreatic secretion make it unlikely that neurotensin has a role in regulating meal-stimulated pancreatic secretion.
Xenin is a 25 amino acid peptide that is structurally related to neurotensin. It is produced by endocrine cells of the gastric mucosa and has biological actions that are similar to neurotensin. Xenin is released into the circulation after ingestion of the meal. In the gastrointestinal tract, xenin binds to the neurotensin receptor and inhibits acid secretion and pancreatic exocrine secretion. Neurotensin receptor antagonists block some but not all of the actions of xenin, indicating that these two peptides may interact differently with this receptor. Although the biological actions of xenin in pancreatic secretion have been demonstrated by the infusion of exogenous peptide, the physiologic role of xenin in pancreatic exocrine secretion is unknown.
Although the principal effect of histamine on the gastrointestinal tract is the regulation of gastric acid secretion, it has also been shown to have effects on pancreatic exocrine secretion. Histamine immunoreactive mast cells are distributed throughout the pancreas in many species including humans, dogs, fox, sheep, pigs, cattle, and rats. In the human pancreas, histamine is more abundant in the head portion compared to the tail. Mast cells are primarily located around blood vessels in the pancreas and are less abundant in the acinar and endocrine portions of the gland. Due to its location around blood vessels and its known vasodilatory effects, it was postulated that histamine may play a role in mediating pancreatic blood flow, particularly in the microcirculation.
Pancreatic acinar cells express both H1 and H2 histamine receptors. H1 receptors are coupled to inositol phosphate metabolism and to generate intracellular calcium as an intracellular signal. H2 receptors are coupled to Gs and upon stimulation, result in generation of cAMP. H3 receptors are located on presynaptic nerve terminals and inhibit the release of neurotransmitters.
Histamine is found in pancreatic juice of dogs following stimulation with secretin or pilocarpine. In anesthetized dogs and the isolated dog pancreas, histamine stimulated pancreatic juice secretion but appeared to be less effective in the conscious dog and in rats. Histamine directly stimulated pancreatic enzyme secretion from pancreatic lobules in vitro from the rat, rabbit, and guinea pig. The effects of histamine were dose dependent. However, the concentrations needed to achieve maximal stimulation of enzyme secretion (~ 10 − 3 M) were much higher than those of traditional hormonal secretagogues such as CCK or secretin.
Histamine is a powerful stimulant of gastric acid secretion by virtue of its ability to activate H2 receptors on parietal cells. Through its actions on acid secretion and the ability of acid to stimulate secretin release, histamine may indirectly affect pancreatic fluid and bicarbonate secretion. Therefore, in vivo studies of the effects of histamine and histamine receptor antagonists on pancreatic fluid and bicarbonate secretion should be interpreted with caution. The effects of specific histamine receptor antagonists on pancreatic exocrine secretion have varied depending on the animal species, the state of anesthesia, and the type of antagonist used. In humans, H2 receptor blockers do not appear to inhibit food- or secretin-stimulated pancreatic secretion.
In the anesthetized guinea pig, histamine, secretin, and CCK stimulated pancreatic juice flow and enzyme secretion. However, the effects of secretin and CCK-8 were much larger than those obtained with histamine. When histamine was infused with secretin there was marked potentiation in both juice flow and enzyme output. When histamine was administered with CCK there was only potentiation of CCK-stimulated fluid output. This finding suggests that histamine potentiates CCK-stimulated effects similar to those of secretin and may signal through a common intracellular mediator such as cAMP.
Taken together these findings indicate that histamine, which functions through paracrine actions in the pancreas, can exert secretory effects via activation of the H1 or H2 receptors on pancreatic acinar cells, or inhibitory effects through actions via the H3 receptor on presynaptic neurons. It is possible that under conditions in which mast cells are more abundant in the pancreas such as pancreatitis, pancreatic exocrine secretion could be altered by histamine. However, current evidence does not support a strong role for histamine as a regulator of pancreatic secretion under normal conditions.
Proteinase-Activated Receptor (PAR) Signaling
Proteinase-activated receptors (PARs) comprise a novel family of G protein-coupled membrane receptors that are activated by proteolytic cleavage. Four types of PAR receptors (PAR-1, -2, -3, and -4) have been identified. PAR-1, PAR-3, and PAR-4 are activated by thrombin while PAR-2 is activated by trypsin, tryptase, and possibly other proteinases. The activation of PARs occurs by proteolytic cleavage of the amino-terminal portion of the extracellular region of the receptor allowing a new amino terminus to bind the receptor as a tethered ligand. PAR-1, PAR-2, and PAR-4 can also be activated by applying synthetic peptides based on the receptor-activated sequence of each receptor (such as SLIGKV for the human PAR-2). The PARs transmit signals by activating the G protein, Gq/11 with subsequent activation of phospholipase C. The gastrointestinal tract is particularly rich in PAR-1 and PAR-2.
PAR-2 (but not the other PARs) is involved in the regulation of pancreatic exocrine secretion. The administration of the PAR-2 activating peptide stimulated pancreatic juice secretion in anesthetized rats. PAR-2 activation also stimulated amylase release from isolated pancreatic acini in vitro as well as in conscious mice. Therefore, the effect of PAR-2 activation appears to be directly on the acinar cell. In dog pancreatic duct cells, PAR-2 agonists activated ion channels that are likely involved in pancreatic juice secretion. In pancreatic duct cells, PAR-2 activation does not increase cAMP, but has prominent effects on intracellular Ca 2 + and protein kinase C to stimulate secretion fluid and mucin secretion.
Although it is difficult to determine the precise role of PAR-2 in the regulation of pancreatic exocrine secretion, the location of the receptor on acinar cells and the abundance of proteinases within the pancreas and inflammatory cells that can migrate to the pancreas, raise the possibility that PAR signaling may be important under physiologic or pathophysiologic conditions. Because of its location on acinar cells and its susceptibility to trypsin activation, it has been suggested that PAR-2 may play a role in the pathogenesis of acute pancreatitis. However, experimental models have been conflicting and it is possible that activation of PAR-2 primarily on duct cells may provide a protective mechanism against pancreatitis by stimulating fluid secretion.
Feedback Regulation of Pancreatic Secretion
Installation of trypsin inhibitor into the upper small intestine of the rat stimulates pancreatic enzyme secretion. Furthermore, surgical diversion of the bile pancreatic duct to remove mixed bile and pancreatic juice produced a potent stimulus for exocrine secretion, and this stimulation was suppressed by intraduodenal infusion of trypsin. These observations formed the basis for the hypothesis that a negative feedback mechanism existed by which signals for pancreatic secretion emanating from the gut are controlled by the presence or absence of proteases in the intestinal lumen.
Most studies examining feedback regulation of the pancreas have been conducted in rats. Diversion of pancreatic juice in unanesthetized animals produced pancreatic responses that were similar in magnitude to those of exogenous administration of CCK. During diversion of bile-pancreatic juice there was no additional increase in pancreatic secretion with the administration of exogenous CCK, carbachol, insulin, or 2-deoxyglucose. However, in anesthetized rats, exogenous CCK and 2-deoxyglucose were able to strongly stimulate pancreatic secretion even in the presence of bile-pancreatic juice diversion implying that anesthesia interfered with the generation of a signal coming from the intestine but did not interfere with the ability of the pancreas to respond to secretagogue stimulation.
Diversion of bile alone stimulated pancreatic secretion, however, this effect was only partially abolished by the infusion of bile acids. The effect of bile diversion was thought to be related to the degradation of intestinal proteases in the absence of bile, reducing protease activity and thus initiating the feedback response. Return of as little as 10% of the bile-pancreatic juice abolished the effects of diversion on pancreatic secretion. This implies that approximately 90% of pancreatic juice needs to be eliminated before an increase in pancreatic secretion is observed through the feedback mechanism.
Feedback regulation of pancreatic secretion is particularly dependent on protease activity. Whereas installation of trypsin, chymotrypsin, or elastase into the upper small intestine suppressed the pancreatic secretory response to diversion, amylase, lipase or inactive proteases had no effect. The ability of these enzymes to modify feedback regulation is confined to the upper third of the intestine.
Further evidence of the role of intraluminal proteases in feedback regulation comes from extensive studies testing different types of protease inhibitors. In the absence of pancreatic juice diversion, a variety of protease inhibitors including soybean trypsin inhibitor, potato chymotrypsin inhibitor, aprotinin, pancreatic secretory trypsin inhibitor, and some other synthetic trypsin inhibitors have been shown to be effective stimulants of pancreatic secretion.
The abilities of certain proteins to stimulate secretion correlate with their susceptibility to serve as substrates for trypsin and perhaps other proteases. It has been hypothesized that dietary proteins could behave as transient “protease inhibitors” when instilled into the gut since they serve as substrates for these enzymes and could temporarily bind proteases within the intestinal lumen.
The bile-pancreatic juice diversion not only stimulates pancreatic secretion in many species but also promotes pancreatic growth in rats and hamsters. It has been difficult to demonstrate a positive effect of pancreatic juice diversion on pancreatic secretion in the dog and intraduodenal trypsin inhibitors have not been shown to stimulate pancreatic secretion in this species. Accordingly, there is very little data to suggest that intestinal feedback regulation of pancreatic secretion exists in the dog. However, pancreatic juice diversion has been reported to stimulate pancreatic secretion in the cow.
The existence of feedback regulation of pancreatic secretion in humans has been controversial since these studies have been more difficult to interpret. Studies instilling trypsin inhibitors into the duodenum have not consistently demonstrated an effect on pancreatic secretion. Specifically, the protease inhibitors aprotinin and FOY-305, which inhibit trypsin but have little effect on chymotrypsin, did not significantly stimulate enzyme secretion in humans. Interestingly, human trypsin may be resistant to some trypsin inhibitors such as those prepared from soybean. Therefore, the type of trypsin inhibitor used may be critical for assessing the protease-sensitive component responsible for regulating pancreatic secretion. A Bowman-Birk soybean trypsin inhibitor which inhibits chymotrypsin and elastase caused significant stimulation of pancreatic secretion when instilled into the duodenum of human volunteers. Together these studies indicate that some combination of trypsin, chymotrypsin, and elastase participate in the regulation of pancreatic secretion and that inhibition of each of these may be necessary to effectively stimulate pancreatic secretion in humans.
In summary, it appears as though active proteolytic enzymes in the upper small intestine are central to a feedback mechanism regulating pancreatic secretion in rats, mice, hamsters, chickens, cows, pigs, and humans.
Role of Cholecystokinin in Feedback Regulation
Although it had been postulated that the effects of pancreatic diversion on pancreatic secretion were mediated by CCK it was not until the mid-1980s that assays sensitive and specific enough to actually measure blood levels of CCK became available. Since then studies have demonstrated that installation of trypsin inhibitor into the stomach of rats was a potent stimulus of CCK secretion and that dietary proteins that were potent stimulants of pancreatic secretion were also effective stimulants of CCK release. It was also shown that dietary proteins that effectively stimulated CCK secretion were ineffective with the coadministration of exogenous trypsin. Moreover, diversion of bile-pancreatic juice significantly increased plasma CCK levels.
Substantial evidence indicated that CCK mediated the effects of trypsin inhibitor feeding and bile-pancreatic juice diversion on pancreatic growth. This has been demonstrated in rodents, where not only did each of these interventions elevate plasma CCK levels but these effects were completely blocked by CCK receptor antagonists. In rats, a relationship was noted between pancreatic growth responses and CCK levels during the course of feeding a high-protein diet. During the first few days of feeding a high casein diet, plasma CCK levels were very high. Over the ensuing week pancreatic size increased substantially after which plasma CCK levels declined to basal levels. It appeared as though casein-stimulated CCK release caused pancreatic growth. It was proposed that pancreatic hypertrophy was accompanied by an increase in pancreatic enzyme output establishing a new steady state in which sufficient enzymes were produced to reduce the effects of high-protein diet on CCK release. These findings suggested that a dynamic interaction existed between dietary, hormonal, and pancreatic secretory and growth responses.
The mechanism by which intraluminal proteases modulated pancreatic secretion had long been a subject of speculation leading to the discovery that stimulation of pancreatic secretion and CCK release occurring in the absence of luminal trypsin was caused by a trypsin-sensitive CCK-releasing peptide that was secreted into the intestinal lumen. This “CCK-releasing factor” stimulated CCK release when introduced into the intestinal lumen and was sensitive to trypsin degradation. According to the hypothesis, in the rat, the CCK-releasing factor is secreted spontaneously into the lumen of the proximal small intestine. In the presence of proteolytic enzymes the releasing factor is degraded and inactive. However, when bile-pancreatic juice is diverted or protease inhibitors are present, the releasing factor remains intact and is able to stimulate CCK secretion. Proteins appear to stimulate pancreatic secretion by complexing with enzymes within the intestinal lumen and protecting the CCK-releasing factor from proteolytic degradation. The releasing factor is thereby preserved so that it can stimulate the CCK cell to secrete CCK. The CCK, in turn, stimulates pancreatic enzyme secretion. This hypothesis is supported by the observation that intact proteins stimulated CCK secretion in vivo but not from isolated mucosal cells in vitro .
The factors regulating production and secretion of the CCK-releasing factors are largely unknown. It is controversial whether cholinergic innervation is important for the regulation of CCK-releasing factor-mediated pancreatic secretion. In a model of vascularly perfused intestinal segments, in which CCK release could be measured after intraluminal administration of food and other stimulants, luminal nutrients were found to stimulate CCK release without activating intramural neurons. Somatostatin prevented elevations in plasma CCK levels accompanying bile-pancreatic juice diversion by inhibiting the release of CCK-releasing factor.
Two groups, at the same time, purified from rat intestinal juice or porcine intestinal extracts different substances that possessed CCK-releasing activities. A 72 amino acid protein named luminal CCK-releasing factor (LCRF) was isolated from washings of the rat small intestine. When introduced into the lumen of the upper small intestine, the amino-terminal 35 amino fragment of LCRF exhibited full biological activity for stimulating pancreatic exocrine secretion. The stimulatory activity of intestinal washings was eliminated by immunoneutralization with a specific LCRF antiserum. Histochemical staining demonstrated immunoreactive LCRF in intestinal mucosal cells and enteric nerves. It was proposed that LCRF released into the lumen of the intestine interacts with the apical surface of CCK cells to stimulate CCK secretion. This possibility has been supported by later studies demonstrating that LCRF stimulated CCK release from isolated human intestinal mucosal cells in vitro and from the CCK-containing cell line, STC-1. Therefore, LCRF had a direct effect on both murine and human CCK cells.
Diazepam binding inhibitor (DBI) is an 89 amino acid protein that is abundant in the central nervous system and peripheral nerves and serves as a natural inhibitor of diazepam binding. DBI was purified from porcine intestinal extracts using a bioassay method for screening CCK-releasing activity. It was proposed that DBI was responsible for the feedback regulation of pancreatic secretion and the postprandial release of CCK. To support this hypothesis it was demonstrated in rats that (i) intraduodenal administration of an 18 amino acid fragment of DBI, DBI 33–50 , increased pancreatic secretion and plasma CCK levels; (ii) increases in luminal DBI immunoreactivities paralleled changes in plasma CCK levels during diversion of pancreatic juice, and (iii) intraduodenal administration of DBI antiserum reduced pancreatic secretion and release of CCK during diversion of bile-pancreatic juice. The DBI immunostaining and in situ hybridization studies in the rat demonstrated that it was widely distributed in the villi of the upper intestine. Therefore, it is conceivable that the apical surface of CCK cells would be exposed to DBI secreted into the lumen of the intestine. In vitro studies demonstrated that DBI 33–50 stimulated CCK release from the murine CCK-containing cell line, STC-1, by eliciting Ca 2 + oscillations through a voltage-dependent L-type Ca 2 + channel.
Monitor peptide, also known as pancreatic secretory trypsin inhibitor-I (PSTI-I), is a 61 amino acid peptide that was originally purified from pancreatic juice. When instilled into the lumen of the intestine, monitor peptide stimulated pancreatic secretion by virtue of stimulating CCK release. Although monitor peptide possesses trypsin inhibitor activity, its ability to stimulate CCK secretion appears to be independent of this effect since monitor peptide could directly stimulate CCK release from isolated intestinal mucosal cells in vitro . Monitor peptide is produced by pancreatic acinar cells and is secreted into the pancreatic duct where it eventually reaches the duodenum. Therefore, the physiologic action of monitor peptide differs substantially from that of CCK-releasing factors that are produced by the intestine. Because monitor peptide is secreted in pancreatic juice, the levels in the intestine are not elevated unless pancreatic secretion is stimulated. Therefore, monitor peptide does not account for the negative feedback responses caused by diversion of bile-pancreatic juice, which appear to be mediated by the intestinal CCK-releasing factor. Although the physiologic role of monitor peptide is unknown it could either mediate protein-stimulated pancreatic secretion or potentiate CCK release and pancreatic secretion once the process of pancreatic secretion has been initiated. Therefore, monitor peptide could provide a positive feedback mechanism reinforcing meal-stimulated secretion. Experimentally, two lines of evidence suggest that monitor peptide may participate in a positive feedback loop. First, in rats fed a high-protein diet, monitor peptide gene expression in the pancreas was increased. Second, administration of exogenous CCK increased monitor peptide mRNA levels suggesting that the effects of diet were mediated by CCK.
Monitor peptide can interact directly with CCK cells. Radiolabeled monitor peptide has been shown to bind to intestinal mucosal cells. In vitro studies have shown that monitor peptide directly stimulated CCK secretion from isolated rat intestinal mucosal cells and did so in a dose- and calcium-dependent manner.
A putative monitor peptide receptor has been partially characterized in dispersed intestinal mucosal cells from rat jejunum. Specific monitor peptide binding was demonstrated to be reversible, temperature- and pH-dependent, properties that are consistent with receptor binding. Autoradiography of an affinity cross-linked complex using I-labeled monitor peptide revealed a potential receptor with a molecular mass of 33 or 53 kDa in the reduced or unreduced form, respectively. Because monitor peptide is present in the intestinal lumen it is believed that these receptors reside on the apical surface of mucosal CCK cells. In this position, monitor peptide receptors would be available to bind monitor peptide and initiate a cascade of signaling events leading to CCK secretion.
Feedback Regulation of Secretin
Secretin release and pancreatic bicarbonate secretion has also been proposed to be regulated by a feedback system involving intraluminal protease activity. Several observations support this hypothesis. Diversion of pancreatic juice from the duodenum increases blood secretin levels and pancreatic secretion in anesthetized rats. Effects on both secretin release and pancreatic exocrine secretion were inhibited by reinfusion of pancreatic juice. It was proposed that the effects of intraduodenal proteases on secretin release were due to degradation of a secretin-releasing peptide. This hypothesis was based on evidence that acid-induced release of secretin was mediated by a protease-sensitive substance in duodenal juice. In anesthetized rats, intestinal perfusate was collected in the presence of duodenal acid infusion. This material, when reinfused into the intestine, stimulated pancreatic bicarbonate secretion. A similar material has been found in dogs which when instilled into the duodenum of recipient rats stimulated pancreatic bicarbonate secretion. Moreover, bicarbonate secretion was inhibited by intravenous infusion of a specific antisecretin serum. Upon further purification, the amino-terminal sequence of 31 residues of the putative secretin-releasing peptide was found to be identical to canine pancreatic phospholipase A 2 . In support of the concept that phospholipase A 2 may be a functional secretin-releasing peptide, was the observation that acid infusion in the duodenum was shown to stimulate pancreatic PLA 2 immunoreactivity and increase pancreatic bicarbonate secretion. In addition, treatment of intestinal washings collected from rats following acid infusion of the duodenum with anti-PLA 2 serum abolished the stimulatory effects on secretin release and pancreatic secretion in recipient rats. Interestingly, pancreatic phospholipase A 2 has been shown to stimulate secretin release in secretin-producing cells in vitro . These findings suggest that pancreatic PLA 2 acts as a secretin-releasing factor. However, porcine pancreatic PLA 2 was unable to stimulate secretin release in rats raising the possibility that other factors may be involved.