Medical Physiology A Cellular and Molecular Approach, Updated 2nd Ed.


Christopher R. Marino and Fred S. Gorelick


The pancreas and major salivary glands are compound exocrine glands

The exocrine pancreas and major salivary glands are compound exocrine glands—specialized secretory organs that contain a branching ductular system through which they release their secretory products. The principal function of these exocrine glands is to aid in the digestion of food. The saliva produced by the salivary glands lubricates ingested food and initiates the digestion of starch. Pancreatic juice, rich in HCO3 and digestive enzymes, neutralizes the acidic gastric contents that enter the small intestine and also completes the intraluminal digestion of ingested carbohydrate, protein, and fat. Each of these exocrine glands is under the control of neural and humoral signals that generate a sequential and coordinated secretory response to an ingested meal. We discuss the endocrine pancreas in Chapter 51.

Morphologically, the pancreas and salivary glands are divided into small but visible lobules, each of which represents a subdivision of the parenchyma and is drained by a single intralobular duct (Fig. 43-1A). Groups of lobules separated by connective tissue septa are drained by larger interlobular ducts. These interlobular ducts empty into a main duct that connects the entire gland to the lumen of the gastrointestinal tract.


Figure 43-1 Acinus duct morphology. A, The fundamental secretory unit is composed of an acinus and an intercalated duct. Intercalated ducts merge to form intralobular ducts, which, in turn, merge to form interlobular ducts, and then the main pancreatic duct. B, The acinar cell is specialized for protein secretion. Large condensing vacuoles are gradually reduced in size and form mature zymogen granules that store digestive enzymes in the apical region of the acinar cell. C, The duct cell is a cuboidal cell with abundant mitochondria. Small microvilli project from its apical membrane.

Within the lobules reside the microscopic structural and functional secretory units of the gland. Each secretory unit is composed of an acinus and a small intercalated duct. The acinus represents a cluster of 15 to 100 acinar cells that synthesize and secrete proteins into the lumen of the epithelial structure. In the pancreas, acinar cells secrete ~20 different digestive zymogens (inactive enzyme precursors) and enzymes. In the salivary glands, the principal acinar cell protein products are α-amylase, mucins, and proline-rich proteins. Acinar cells from both the pancreas and salivary glands also secrete an isotonic, plasma-like fluid that accompanies the secretory proteins. In all, the final acinar secretion is a protein-rich product known as the primary secretion.

Each acinar lumen is connected to the proximal end of an intercalated duct. Distally, the intercalated ducts fuse with other small ducts to form progressively larger ducts that ultimately coalesce to form the intralobular duct that drains the lobule. Although the ducts provide a conduit for the transport of secretory proteins, the epithelial cells lining the ducts also play an important role in modifying the fluid and electrolyte composition of the primary secretion. Thus, the final exocrine gland secretion represents the combined product of two distinct epithelial cell populations, the acinar cell and the duct cell.

In addition to acini and ducts, exocrine glands contain a rich supply of nerves and blood vessels. Postganglionic parasympathetic and sympathetic fibers contribute to the autonomic regulation of secretion through the release of cholinergic, adrenergic, and peptide neurotransmitters that bind to receptors on the acinar and duct cells. Both central and reflex pathways contribute to the neural regulation of exocrine secretion. The autonomic nerves also carry afferent pain fibers that are activated by glandular inflammation and trauma. The vasculature not only provides oxygen and nutrients for the gland but also carries the hormones that help to regulate secretion.

Acinar cells are specialized protein-synthesizing cells

Acinar cells—such as those in the pancreas (Fig. 43-1B) and salivary glands—are polarized epithelial cells that are specialized for the production and export of large quantities of protein. Thus, the acinar cell is equipped with extensive rough endoplasmic reticulum (ER). However, the most characteristic feature of the acinar cell is the abundance of electron-dense secretory granules at the apical pole of the cell. These granules are storage pools of secretory proteins, and they are poised for releasing their contents after stimulation of the cell by neurohumoral agents. The secretory granules of pancreatic acinar cells contain the mixture of zymogens and enzymes required for digestion. The secretory granules of salivary acinar cells contain either α-amylase (in the parotid gland) or mucins (in the sublingual glands). Secretory granules in the pancreas appear uniform, whereas those in the salivary glands often exhibit focal nodules of condensation within the granules known as spherules.

The pancreatic acinar cell has served as an important model for elucidating protein synthesis and export through the secretory pathway. Synthesis of secretory proteins (see Chapter 2) begins with the cellular uptake of amino acids and their incorporation into nascent proteins in the rough ER (Fig. 43-2). Vesicular transport mechanisms then shuttle the newly synthesized proteins to the Golgi complex.


Figure 43-2 Movement of newly synthesized proteins through the secretory pathway. The cell model at the top illustrates the vectoral movement of nascent proteins through the compartments of the secretory pathway. The four records in the graph show the time course of secretory proteins moving through these compartments. To label newly synthesized proteins radioactively, the investigators briefly pulsed the pancreatic acinar cells with 3H-labeled amino acids. At specific times after the pulse, tissues were fixed, and the distribution of the radioactive amino acid was determined using autoradiography. Each of the four records shows the number of radiographic grains—as a fraction of all of the grains—found in each compartment at various times after the pulse. (Data from Jamieson J, Palade G: J Cell Biol 1967; 34:597-615; and Jamieson J, Palade G: J Cell Biol 1971; 50:135-158.)

Within the Golgi complex, secretory proteins are segregated away from lysosomal enzymes. Most lysosomal enzymes require the mannose 6-phosphate receptor for sorting to the lysosome (see Chapter 2). However, the signals required to direct digestive enzymes into the secretory pathway remain unclear.

Secretory proteins exit the Golgi complex in condensing vacuoles. These large membrane-bound structures are acidic and maintain the lowest pH within the secretory pathway.

Maturation of the condensing vacuole to a secretory or zymogen granule is marked by condensation of the proteins within the vacuole and pinching off of membrane vesicles. The diameter of a zymogen granule is about two thirds that of a condensing vacuole, and its content is more electron dense. Secretory proteins are stored in zymogen granules that are located in the apical region of the acinar cell. The bottom portion of Figure 43-2 shows the results of a pulsechase experiment that follows the cellular itinerary of radiolabeled amino acids as they move sequentially through the four major compartments of the secretory pathway.

Exocytosis, the process by which secretory granules release their contents, is a complex series of events that involves the movement of the granules to the apical membrane, fusion of these granules with the membrane, and release of their contents into the acinar lumen. Secretion is triggered by either hormones or neural activity. At the onset of secretion, the area of the apical plasma membrane increases as much as 30-fold. Thereafter, activation of an apical endocytic pathway leads to retrieval of the secretory granule membrane for recycling and a decrease in the area of the apical plasma membrane back to its resting value. Thus, during the steady state of secretion, the secretory granule membrane is simultaneously delivered to and retrieved from the apical membrane.

The cytoskeleton of the acinar cell plays an important role in the regulation of exocytosis. A component of the actin network appears to be required for delivery of the secretory granules to the apical region of the cell. A second actin network, located immediately below the apical membrane, acts as a barrier that blocks fusion of the granules with the apical plasma membrane. On stimulation, this second network reorganizes and then releases the blockade to permit the secretory granules to approach the apical plasma membrane. Fusion of the granules with the plasma membrane probably requires the interaction of proteins on both the secretory granules and the apical plasma membrane, as well as various cytosolic factors (see Chapter 2). After fusion, the granule contents are released into the acinar lumen and are carried down the ducts into the gastrointestinal tract.

Duct cells are epithelial cells specialized for fluid and electrolyte transport

Pancreatic and salivary duct cells are polarized epithelial cells specialized for the transport of electrolytes across distinct apical and basolateral membrane domains. Duct epithelial cells contain specific membrane transporters and an abundance of mitochondria to provide energy for active transport, and they exhibit varying degrees of basolateral membrane infolding that increases membrane surface areas of pancreatic duct cells (Fig. 43-1C) and salivary duct cells. Although some duct cells contain prominent cytoplasmic vesicles, or storage granules—an indication of an additional protein secretory function, the synthetic machinery (i.e., ER and Golgi complex) of the duct cell is, in general, much less developed than that of the acinar cell.

Duct cells exhibit a considerable degree of morphologic heterogeneity along the length of the ductal tree. At the junction between acinar and duct cells in the pancreas are small cuboidal epithelial cells known as centroacinar cells. These cells express very high levels of carbonic anhydrase and presumably play a role in HCO3 secretion. The epithelial cells of the most proximal (intercalated) duct are squamous or low cuboidal, have an abundance of mitochondria, and tend to lack cytoplasmic vesicles. These features suggest that the primary function of these cells is fluid and electrolyte transport. Progressing distally, the cells become more cuboidal columnar and contain more cytoplasmic vesicles and granules. These features suggest that these cells are capable of both transport of fluid and electrolytes and secretion of proteins. Functional studies indicate that the types of solute transport proteins within duct cells differ depending on the cell’s location in the ductal tree.

Ion transport in duct cells is regulated by neurohumoral stimuli that act through specific receptors located on the basolateral membrane. As is the case for cells elsewhere in the body, duct cells can increase transcellular electrolyte movement either by activating individual transport proteins or by increasing the number of transport proteins in the plasma membrane.

Goblet cells contribute to mucin production in exocrine glands

In addition to acinar and duct cells, exocrine glands contain varying numbers of goblet cells. These cells secrete high-molecular-weight glycoproteins known as mucins. When hydrated, mucins form mucus(see Chapter 2). Mucus has several important functions, including lubrication, hydration, and mechanical protection of surface epithelial cells. Mucins also play an important immunologic role by binding to pathogens and interacting with immune-competent cells. These properties may help to prevent infections. In the pancreas, mucin-secreting goblet cells are primarily found among the epithelial cells that line the large, distal ducts. They can account for as many as 25% of the epithelial cells in the distal main pancreatic duct of some species. In the salivary gland, goblet cells are also seen in the large distal ducts, although in less abundance than in the pancreas. However, in many salivary glands, mucin is also secreted by acinar cells.


The acinar cell secretes digestive proteins in response to stimulation

To study secretion at the cellular level, investigators use enzymes to digest pancreatic connective tissue and obtain single acini (small groups of 15 to 100 acinar cells), or they mechanically dissect single lobules (groups of 250 to 1000 cells). The measure of secretion is the release of digestive proteins into the incubation medium. The amount released over a fixed time interval is expressed as a percentage of the total content at the outset of the experiment. Because amylase is released in a fully active form, it is common to use the appearance of amylase activity as a marker for secretion by acinar cells.

When the acinar cells are in an unstimulated state, they secrete low levels of digestive proteins through a constitutive secretory pathway. Acinar cells stimulated by neurohumoral agents secrete proteins through a regulated pathway. Regulated secretion by acini and lobules in vitro is detected within 5 minutes of stimulation and is energy dependent. During a 30-to 60-minute stimulation period, acinar cells typically secrete 5 to 10 times more amylase than with constitutive release. However, during this period of regulated secretion, the cells typically secrete only 10% to 20% of the digestive proteins stored in their granules. Moreover, acinar cells are able to increase their rate of protein synthesis to replenish their stores.

The acinar cell may exhibit two distinct patterns of regulated secretion: monophasic and biphasic (Fig. 43-3A). An agonist that generates a monophasic dose-response relationship (e.g., gastrin-releasing peptide [GRP]) causes secretion to reach a maximal level that does not fall with higher concentrations of the agent. In contrast, a secretagogue that elicits a biphasic dose-response relationship (e.g., cholecystokinin [CCK] and carbachol) causes secretion to reach a maximal level that subsequently diminishes with higher concentrations of the agent. As discussed later, this biphasic response may reflect the presence of functionally separate high-affinity and low-affinity receptors and is related to the pathogenesis of acute pancreatitis (see the box entitled Acute Pancreatitis).


Figure 43-3 Neurohumoral agents elicit different secretory responses from the pancreatic acinar cell. (A, Data from Jensen RT: In Johnson LR [ed]: Physiology of the Gastrointestinal Tract, pp 1377-1446. New York: Raven Press, 1994; B, Data from Burnham DB, McChesney DJ, Thurston KC, Williams JA: J Physiol 1984; 349:475-482.)

Regulated secretion of proteins by pancreatic acinar cells is mediated through cholecystokinin and muscarinic receptors

Although at least a dozen different receptors have been found on the plasma membrane of the pancreatic acinar cell, the most important in regulating protein secretion are the CCK receptors and the muscarinic acetylcholine (ACh) receptors. These receptors have many similarities: both are linked to the Gαq heterotrimeric G protein, both use the phospholipase C (PLC)/Ca2+ signal transduction pathway (see Chapter 3), and both lead to increased enzyme secretion from the acinar cell.

Two closely related CCK receptors are distinguished by their structure, affinity for ligands, and tissue distribution (see Chapter 42). Although both CCK receptors may be activated by CCK or gastrin, the CCKA receptor has a much higher affinity for CCK than for gastrin, whereas the CCKB receptor has approximately equal affinities for CCK and gastrin. In some species, both forms of the CCK receptor are present on the acinar cell.

An important feature of both CCK receptors is their ability to exist in both a high-affinity and a low-affinity state. Low (picomolar) concentrations of CCK activate the high-affinity forms of the CCK receptors and stimulate secretion. Conversely, supraphysiological (10- to 100-fold higher) concentrations of CCK activate the low-affinity forms of the receptors and inhibit secretion. As we explain in the next section, these two affinity states (i.e., activated by different concentrations of CCK) of each of the two CCK receptors generate distinct second-messenger signaling patterns. It is likely that, under physiological conditions, only the high-affinity states of the CCK or muscarinic receptor are activated. Stimulation of the lower-affinity states by supraphysiological concentrations of either CCK or ACh not only inhibits enzyme secretion but also may injure the acinar cell (see the box titled Acute Pancreatitis).

The muscarinic receptor on the acinar cell is probably of the M3 subtype (see Chapter 14) found in many glandular tissues. Like the CCK receptor, the M3 receptor is localized to the basolateral membrane of the cell. Numerous other receptors, including those for GRP, somatostatin, and vasoactive intestinal polypeptide (VIP; see Chapter 42); calcitonin gene–related peptide (CGRP; see Chapter 52), insulin (see Chapter 51), and secretin are also found on the pancreatic acinar cell. Although these other receptors may also play a role in regulating secretion, protein synthesis, growth, and transformation, their precise physiological functions remain to be clearly defined.

Activation of receptors that stimulate different signal transduction pathways may lead to an enhanced secretory response. For example, as shown in Figure 43-3B, simultaneous stimulation of the high-affinity CCK receptor (which acts through [Ca2+]i) and the VIP receptor (which acts through cAMP) generates an additive effect on secretion. Alternatively, acinar cells that have previously been stimulated may become temporarily refractory to subsequent stimulation. This phenomenon is known as desensitization.

Ca2+ is the major second messenger for the secretion of proteins by pancreatic acinar cells

Ca2+ Much of the pioneering work on the role of intracellular Ca2+ in cell signaling has been performed in the pancreatic acinar cell (Fig. 43-4A). Generation of a cytosolic Ca2+ signal is a complex summation of cellular events (see Chapter 3) that regulates cytosolic free Ca2+ levels ([Ca2+]i). Even when the acinar cell is in the resting state, [Ca2+]i oscillates slowly. In the presence of maximal stimulatory (i.e., physiological) concentrations of CCK or ACh, the frequency of the oscillations increases (Fig. 43-4B), but little change in their amplitude is noted. This increase in the frequency of [Ca2+]i oscillations is required for protein secretion by acinar cells. In contrast, supramaximal (i.e., hyperstimulatory) concentrations of CCK or ACh generate a sudden, large spike in [Ca2+]i and eliminate additional [Ca2+]ioscillations. This [Ca2+]i spike and the subsequent absence of oscillations are associated with an inhibition of secretion that appears to be mediated by disruption of the cytoskeletal components that are required for secretion.


Figure 43-4 Stimulation of protein secretion from the pancreatic acinar cell. A, The pancreatic acinar cell has at least two pathways for stimulating the insertion of zymogen granules and thus releasing digestive enzymes. ACh and CCK both activate Gαq, which stimulates PLC, which ultimately leads to the activation of PKC and the release of Ca2+. Elevated [Ca2+]i also activates calmodulin (CaM), which can activate protein kinases (PK) and phosphatases (PP). Finally, VIP and secretin both activate Gαs, which stimulates adenylyl cyclase (AC), leading to the production of cAMP and the activation of PKA. B, Applying a physiological dose of CCK (i.e., 10 pM) triggers a series of [Ca2+]i oscillations, as measured by a fluorescent dye. However, applying a supraphysiological concentration of CCK (1 nM) elicits a single large [Ca2+]i spike and halts the oscillations. Recall that high levels of CCK also are less effective in causing amylase secretion. (B, Data from Tsunoda Y, Stuenkel EL, Williams JA: Am J Physiol 1990; 259:G792-G801.)

cGMP Physiological stimulation of the acinar cell by either CCK or ACh also generates a rapid and prominent increase in [cGMP]i levels. The increase in [cGMP]i has been linked to nitric oxide metabolism; inhibition of nitric oxide synthase (see Chapter 3) blocks the increase in [cGMP]i after secretagogue stimulation. Some evidence suggests that cGMP may be involved in regulating Ca2+ entry and storage in the acinar cell.

cAMP Secretin, VIP, and CCK increase cAMP production and thus activate protein kinase A (PKA) activity in pancreatic acinar cells. Low concentrations of CCK cause transient stimulation of PKA, whereas supraphysiological concentrations of CCK cause a much more prominent and prolonged increase in [cAMP]i and PKA activity. ACh, however, has little, if any, effect on the cAMP signaling pathway.

Effectors As summarized in Figure 43-4A, the most important effectors of intracellular second messengers are the protein kinases. Stimulation of CCK and muscarinic receptors on the acinar cell leads to the generation of similar Ca2+ signals and activation of calmodulin-dependent protein kinases and members of the protein kinase C (PKC) family (see Chapter 3). Activation of secretin or VIP receptors increases [cAMP]i and thus activates PKA. These second messengers probably also activate protein phosphatases, as well as other protein kinases not depicted in Figure 43-4A. The protein targets of activated kinases and phosphatases in the pancreatic acinar cell are largely unknown. However, some are involved in regulating secretion, whereas others mediate protein synthesis, growth, transformation, and cell death.

In addition to proteins, the pancreatic acinar cell also secretes a plasma-like fluid

In addition to protein, acinar cells in the pancreas secrete an isotonic, plasma-like fluid (Fig. 43-5). This NaCl-rich fluid hydrates the dense, protein-rich material that the acinar cells secrete. The fundamental transport event is the secretion of Cl across the apical membrane. For transcellular (plasma to lumen) movement of Cl to occur, Cl must move into the cell across the basolateral membrane. As in many other Cl-secreting epithelial cells (see Chapter 5), basolateral Cl uptake by the acinar cell occurs through an Na/K/Cl cotransporter. The Na-K pump generates the Na+ gradient that energizes the Na/K/Cl cotransporter. The K+ entering through the Na-K pump and through the Na/K/Cl cotransporter exits through K+ channels that are also located on the basolateral membrane. Thus, a pump, a cotransporter, and a channel are necessary to sustain the basolateral uptake of Cl into the acinar cell. (See Note: ClSecretion into Pancreatic Secretory Vesicles)


Figure 43-5 Stimulation of isotonic NaCl secretion by the pancreatic acinar cell. Both ACh and CCK stimulate NaCl secretion, probably through phosphorylation of basolateral and apical ion channels.

The rise in [Cl]i produced by basolateral Cl uptake drives the secretion of Cl down its electrochemical gradient through channels in the apical membrane. As the transepithelial voltage becomes more lumen negative, Na+ moves through the cation-selective paracellular pathway (i.e., tight junctions) to join the Cl secreted into the lumen. Water also moves through this paracellular pathway, as well as through aquaporin water channels on the apical and basolateral membranes. Therefore, the net effect of these acinar cell transport processes is the production of an isotonic, NaCl-rich fluid that accounts for ~25% of total pancreatic fluid secretion.

Like the secretion of protein by acinar cells, secretion of fluid and electrolytes is stimulated by secretagogues that raise [Ca2+]i. In the pancreas, activation of muscarinic receptors by cholinergic neural pathways and activation of CCK receptors by humoral pathways increase the membrane conductance of the acinar cell. A similar effect is seen with GRP. Apical membrane Cl channels and basolateral membrane K+ channels appear to be the effector targets of the activated Ca2+ signaling pathway. Phosphorylation of these channels by Ca2+-dependent kinases is one likely mechanism that underlies the increase in open-channel probability that accompanies stimulation.


The pancreatic duct cell secretes isotonic NaHCO3

The principal physiological function of the pancreatic duct cell is to secrete an HCO3-rich fluid that alkalinizes and hydrates the protein-rich primary secretions of the acinar cell. The apical step of transepithelial HCO3 secretion (Fig. 43-6) is mediated in part by a Cl-HCO3 exchanger, a member of the SLC26 family (see Chapter 5) that secretes intracellular HCO3 into the duct lumen. Luminal Cl must be available for this exchange process to occur. Although some luminal Cl is present in the primary secretions of the acinar cell, anion channels on the apical membrane of the duct cell provide additional Clto the lumen in a process called Cl recycling. The most important of these anion channels is the cystic fibrosis transmembrane conductance regulator (CFTR), a cAMP-activated Cl channel that is present on the apical membrane of pancreatic duct cells (see Chapter 5). Cl recycling is facilitated by the co-activation of CFTR and SLC26 exchangers through direct protein-protein interactions. In some species, such as the rat and mouse, pancreatic duct cells also contain a Ca2+-activated Cl channel on the apical membrane; this channel also provides Cl to the lumen for recycling. Apical Cl channels may also directly serve as conduits for HCO3 movement from the duct cell to the lumen.


Figure 43-6 HCO3 secretion by the cells of the pancreatic duct. Secretin activates the cAMP signaling pathway and opens the CFTR Cl channels through phosphorylation. Cl movement out of the cell leads to basolateral membrane depolarization, thus generating the electrical gradient that favors NaHCO3cotransport.

The intracellular HCO3 that exits the duct cell across the apical membrane arises from two pathways. The first is direct uptake of HCO3 through an electrogenic Na/HCO3 cotransporter (NBCe1), which presumably operates with an Na+/HCO3 stoichiometry of 1 : 2. The second mechanism is the generation of intracellular HCO3 from CO2 and OH, catalyzed by carbonic anhydrase (see Chapter 28 for the box on this topic). The OH in this reaction is derived, along with H+, from H2O. Thus, the H+ that accumulates in the cell must be extruded across the basolateral membrane. One mechanism of H+ extrusion is Na-H exchange. The second mechanism for H+ extrusion across the basolateral membrane, at least in some species, is an ATP-dependent H+ pump. Pancreatic duct cells contain acidic intracellular vesicles (presumably containing vacuolar-type H+ pumps) that are mobilized to the basolateral membrane of the cell after stimulation by secretin, a powerful secretagogue (see later). Indeed, H+ pumps are most active under conditions of neurohumoral stimulation. Thus, three basolateral transporters directly or indirectly provide the intracellular HCO3 that pancreatic duct cells need for secretion: (1) the electrogenic Na/HCO3 cotransporter, (2) the Na-H exchanger, and (3) the H+ pump. The physiological importance of these three acid-base transporters in humans has yet to be fully established. The pancreatic duct cell accounts for ~75% of total pancreatic fluid secretion. (See Note: Bicarbonate secretion by the pancreatic duct)

Secretin (through cAMP) and acetylcholine (through Ca2+) both stimulate HCO3 secretion by the pancreatic duct

When stimulated, the epithelial cells of the pancreatic duct secrete an isotonic NaHCO3 solution. The duct cells have receptors for secretin, ACh, GRP (all of which stimulate HCO3 secretion) and substance P (which inhibits it). Although some evidence indicates that CCK modulates ductular secretory processes, CCK receptors have not been identified on these cells.

Secretin is the most important humoral regulator of ductal HCO3 secretion. Activation of the secretin receptor on the duct cell stimulates adenylyl cyclase, which raises [cAMP]i. Because forskolin and cAMP analogues stimulate ductal HCO3 secretion, the secretin response has been attributed to its effect on [cAMP]i and activation of PKA. However, even low concentrations of secretin that do not measurably increase [cAMP]i can stimulate HCO3 secretion. This observation suggests that the secretin response may be mediated by (1) unmeasurably small increases in total cellular cAMP, (2) cAMP increases that are localized to small intracellular compartments, or (3) activation of alternative second-messenger pathways. Secretin acts by stimulating the apical CFTR Cl channel and the basolateral Na/HCO3 cotransporter, without affecting the Na-H exchanger.

HCO3 secretion is also regulated by the parasympathetic division of the autonomic nervous system (see Chapter 14). The postganglionic parasympathetic neurotransmitter ACh in-creases [Ca2+]i and activates Ca2+-dependent protein kinases (PKC and the calmodulin-dependent protein kinases) in pancreatic duct cells. The ACh effect is inhibited by atropine, a finding indicating that this neurotransmitter is acting through muscarinic receptors on the duct cell. Although ductular secretion in the rat is also stimulated by GRP, the second messenger mediating this effect is not known. Unlike the effect on the acinar cell, GRP does not increase [Ca2+]i in the duct cell. GRP also does not raise [cAMP]i.

In the rat, both basal and stimulated ductular HCO3 secretion is inhibited by substance P. The second messenger mediating this effect is also unknown. Because substance P inhibits HCO3 secretion regardless of whether the secretagogue is secretin, ACh, or GRP—which apparently act through three different signal transduction mechanisms—substance P probably acts at a site that is distal to the generation of second messengers, such as by inhibiting the Cl-HCO3 exchanger.

Apical membrane chloride channels are important sites of neurohumoral regulation

In the regulation of pancreatic duct cells by the neurohumoral mechanisms just discussed, the only effector proteins that have been identified as targets of the protein kinases and phosphatases are the apical Clchannels, basolateral K+ channels, and the Na/HCO3 cotransporter. CFTR functions as a low-conductance, apical Cl channel (see Chapter 5). CFTR has nucleotide-binding domains that control channel opening and closing as well as a regulatory domain with multiple potential PKA and PKC phosphorylation sites. Neurohumoral agents that control fluid and electrolyte secretion by the pancreatic duct cells act at this site. Agents that activate PKA are the most important regulators of CFTR function. PKC activation enhances the stimulatory effect of PKA on CFTR Cl transport, but alone it appears to have little direct effect on CFTR function. Thus, the CFTR Cl channel is regulated by ATP through two types of mechanisms: interaction with the nucleotide-binding domains and protein phosphorylation (see Fig. 5-10B).

In addition to CFTR, pancreatic duct cells in some species have an outwardly rectifying Cl channel (ORCC) on the apical membrane. This channel, which has been identified in a variety of epithelial cells, can be activated by increases in [cAMP]i or [Ca2+]i. Studies suggest that part of the effect of cAMP on ORCCs may be indirect and may occur through CFTR. The working hypothesis is that stimulation of CFTR somehow promotes ATP efflux from the cell to the lumen and that the ATP binds to an apical purinergic receptor to activate ORCCs in an autocrine/paracrine fashion (Fig. 43-6).

In rat pancreatic duct cells, Ca2+-sensitive basolateral K+ channels seem to be targets of neurohumoral stimulation. Activators of the cAMP pathway stimulate phosphorylation by PKA, thus enhancing the responsiveness of these channels to [Ca2+]i and increasing their probability of being open.

Pancreatic duct cells may also secrete glycoproteins

Although the primary function of the pancreatic duct cells is to secrete HCO3 and water, these cells may also synthesize and secrete various high-molecular-weight proteoglycans. Some of these proteins are structurally distinct from the mucin that is produced by the specialized goblet cells in the duct. Unlike the proteins that are secreted by acinar cells, the glycoproteins synthesized in duct cells are not accumulated in large secretion granules. Rather, they appear to be continuously synthesized and secreted from small cytoplasmic vesicles. Secretin increases the secretion of glycoproteins from these cells, but this action appears to result from stimulation of glycoprotein synthesis, rather than from stimulation of vesicular transport or exocytosis itself. The role of these proteins may be to protect against protease-mediated injury to mucosal cells.


Pancreatic juice is a protein-rich, alkaline secretion

Humans produce ~1.5 L of pancreatic fluid each day. The pancreas has the highest rates of protein synthesis and secretion of any organ in the body. Each day, the pancreas delivers between 15 and 100 g of protein into the small intestine. The level of pancreatic secretion is determined by a balance between factors that stimulate secretion and those that inhibit it.

The human pancreas secretes more than 20 proteins, some of which are listed in Table 43-1. Most of these proteins are either inactive digestive enzyme precursors—zymogens—or active digestive enzymes. The secretory proteins responsible for digestion can be classified according to their substrates: proteases hydrolyze proteins, amylases digest carbohydrates, lipases and phospholipases break down lipids, and nucleases digest nucleic acids. The functions of other secretory proteins—such as glycoprotein II (GP2), lithostathine, and pancreatitis-associated protein—are less well defined.

Table 43-1 Pancreatic Acinar Cell Secretory Products










Proprotease E


Procarboxypeptidase A


Procarboxypeptidase B





Carboxyl ester lipase











Trypsin inhibitor

Blockade of trypsin activity


Possible prevention of stone formation; constituent of protein plugs


Endocytosis?; formation of protein plugs

Pancreatitis-associated protein


Na+, Cl, H2O

Hydration of secretions



Cystic Fibrosis

CF is the most common lethal genetic disease in whites, in whom it affects ~1 in 2000. Approximately 1 in 20 whites carry the autosomal recessive genetic defect. Clinically, CF is characterized by progressive pancreatic and pulmonary insufficiency resulting from the complications of organ obstruction by thickened secretions. The disease results from mutations in the CF gene (located on chromosome 7) that alter the function of its product, CFTR (see Fig. 5-10). CFTR is a cAMP-activated Cl channel that is present on the apical plasma membrane of many epithelial cells. In the pancreas, CFTR has been localized to the apical membrane of duct cells, where it functions to provide the luminal Cl for Cl-HCO3 exchange (Fig. 43-6).

Most CF gene mutations result in the production of a CFTR molecule that is abnormally folded after its synthesis in the ER. The ER quality control system recognizes these molecules as defective, and most mutant CFTR molecules are prematurely degraded before they reach the plasma membrane. Subsequent loss of CFTR expression at the plasma membrane disrupts the apical transport processes of the duct cell and results in decreased secretion of HCO3 and water by the ducts. As a result, protein-rich primary (acinar) secretions thicken within the duct lumen and lead to ductal obstruction and eventual tissue destruction. Pathologically, the ducts appear dilated and obstructed, and fibrotic tissue and fat gradually replace the pancreatic parenchyma—hence the original cystic fibrosis designation. The subsequent deficiency of pancreatic enzymes that occurs leads to the maldigestion of nutrients and thus the excretion of fat in the stool (steatorrhea) by patients with CF. Before the development of oral enzyme replacement therapy, many patients with CF died of complications of malnutrition.

Now, the major cause of morbidity and mortality in CF is progressive pulmonary disease. The pathophysiology of lung disease in CF is more complex than that of pancreatic disease. A major finding is that the airway mucus is thick and viscous as a result of insufficient fluid secretion into the airway lumen. The pulmonary epithelium probably both secretes fluid (in a mechanism that requires CFTR) and absorbs fluid (in a mechanism that requires apical ENaC Na+ channels). In CF, the reduced activity of CFTR shifts the balance more toward absorption, and a thick mucous layer is generated that inhibits the ciliary clearance of foreign bodies (see Chapter 26). The results are increased rate and severity of infections and thus inflammatory processes that contribute to the destructive process in the lung.

The pulmonary symptoms most commonly bring the patient to the physician’s attention in early childhood. Cough and recurrent respiratory infections that are difficult to eradicate are usually the first indications of the illness. The child’s sputum is particularly thick and viscous. Pulmonary function progressively declines over the ensuing years, and patients may also experience frequent and severe infections, atelectasis (collapse of lung parenchyma), bronchiectasis (chronic dilatation of the bronchi), and recurrent pneumothoraces (air in the intrapleural space). In addition to the pancreatic and pulmonary manifestations, CF also causes a characteristic increase in the [NaCl] of sweat, which is intermediate in heterozygotes. Pharmacological approaches that bypass the Cl transport defect in a lung with CF are currently being evaluated, and considerable effort is being directed toward the development of in vivo gene transfer techniques to correct the underlying genetic defect.

GP2 is an unusual protein with an N-terminal glycosyl phosphatidylinositol moiety that links it to the inner leaflet of the zymogen granule membrane. GP2 has been implicated in the regulation of endocytosis. After exocytosis, luminal cleavage of the GP2 linkage to the zymogen granule membrane seems to be necessary for proper trafficking of the zymogen granule membrane back into the cell from the plasma membrane. Under certain circumstances, the released GP2—and also lithostathine—may form protein aggregates in the pancreatic juice. This finding is not surprising inasmuch as GP2 is structurally related to the Tamm-Horsfall protein, which is secreted by the renal thick ascending limb (see Chapter 33). The tendency of GP2 and lithostathine to form aggregates may have detrimental clinical consequences in that both proteins have been implicated in the pathologic formation of protein plugs that can obstruct the lumen of acini in patients with cystic fibrosis and chronic pancreatitis.

Pancreatitis-associated protein is a secretory protein that is present in low concentrations in the normal state. However, levels of this protein may increase up to several hundred-fold during the early phases of pancreatic injury. Pancreatitis-associated protein is a bacteriostatic agent that may help to prevent pancreatic infection during bouts of pancreatitis.

Pancreatic juice is also rich in Ca2+ and HCO3Calcium concentrations are in the millimolar range inside the organelles of the secretory pathway of the acinar cells. These high levels of Ca2+ may be required to induce the aggregation of secretory proteins and to direct them into the secretory pathway. Bicarbonate secreted by duct cells neutralizes the acidic gastric secretions that enter the duodenum and allows digestive enzymes to function properly; HCO3 also facilitates the micellar solubilization of lipids and mucosal cell function. The [HCO3] in pancreatic juice increases with increases in the secretory flow rate (Fig. 43-7). In the unstimulated state, the flow is low, and the electrolyte composition of pancreatic juice closely resembles that of blood plasma. As the gland is stimulated and flow increases, exchange of Clin the pancreatic juice for HCO3 across the apical membrane of the duct cells produces a secretory product that is more alkaline (pH of ~8.1) and has a lower [Cl]. Concentrations of Na+ or K+, however, are not significantly altered by changes in flow.


Figure 43-7 Flow dependence of the electrolyte composition of pancreatic fluid. In this experiment on a cat, increasing the level of secretin not only increases the rate at which fluids flow out of the pancreas but also changes the composition of the fluid. (Data from Case RM, Harper AA, Scratcherd T: J Physiol 1969; 201:563-596.)

In the fasting state, levels of secreted pancreatic enzymes oscillate at low levels

Pancreatic secretion is regulated in both the fasted and fed states. Under basal conditions, the pancreas releases low levels of pancreatic enzymes (Fig. 43-8). However, during the digestive period (eating a meal), pancreatic secretion increases in sequential phases to levels that are 5- to 20-fold higher than basal levels. The systems that regulate secretion appear to be redundant; if one system fails, a second takes its place. These mechanisms ensure that the release of pancreatic enzymes corresponds to the amount of food in the small intestine.


Figure 43-8 Time course of pancreatic secretion during fasting and feeding. The interdigestive output of secretory products (e.g., trypsin) by the pancreas varies cyclically and in rough synchrony with the four phases of motor activity (MMCs) of the small intestine, shown by colored vertical bands. During the fed state, one notes a massive and sustained increase in trypsin release by the pancreas, as well as a switch of small intestine motility to the fed state. (Data from DiMagno EP, Layer P: In Go VLW, DiMagno EP, Gardner JD, et al [eds]: The Pancreas: Biology, Pathobiology and Disease, 2nd ed, pp 275-300. New York: Raven Press, 1993.)

Like other organs in the upper gastrointestinal tract, the pancreas has a basal rate of secretion even when food is not being eaten or digested. During this interdigestive (fasting) period, pancreatic secretions vary cyclically and correspond to sequential changes in the motility of the small intestine (see Chapter 41). Pancreatic secretion is minimal when intestinal motility is in its quiescent phase (phase I); biliary and gastric secretions are also minimal at this time. As duodenal motility increases (phase II), so does pancreatic secretion. During the interdigestive period, enzyme secretion is maximal when intestinal motility—the migrating motor complexes (MMCs; see Fig. 41-6)—is also maximal (phase III). However, even this maximal interdigestive secretory rate is only 10% to 20% of that stimulated by a meal. The peak phases of interdigestive intestinal motor activity and pancreatic secretory activity are followed by a declining period (phase IV). Fluid and electrolyte secretion rates during the interdigestive phase are usually less than 5% of maximum levels.

The cyclic pattern of interdigestive pancreatic secretion is mediated by intrinsic and extrinsic mechanisms. The predominant mechanism of pancreatic regulation is through parasympathetic pathways. Telenzepine, an antagonist of the M1 muscarinic ACh receptor, reduces interdigestive enzyme secretion by more than 85% during phases II and III. Although cholinergic pathways are the major regulators of interdigestive pancreatic secretion, CCK and adrenergic pathways also play a role. CCK appears to stimulate pancreatic enzyme secretion during phases I and II. In contrast, basal α-adrenergic tone appears to suppress interdigestive pancreatic secretion. Although human and canine pancreas denervated during transplantation exhibits cyclic secretion, this secretion is no longer synchronous with duodenal motor activity. These observations support a role for the autonomic nervous system in regulating basal (resting) pancreatic secretion.

Cholecystokinin from duodenal I cells stimulates enzyme secretion by the acini, and secretin from S cells stimulates HCO3 and fluid secretion by the ducts

CCK plays a central role in regulating pancreatic secretion. CCK is released from neuroendocrine cells (I cells; see Table 41-1) present in the duodenal mucosa and acts on pancreatic acinar cells to increase protein secretion (Fig. 43-4). In response to a meal, plasma CCK levels increase 5- to 10-fold within 10 to 30 minutes. Three lines of evidence show that CCK is a physiological mediator of pancreatic protein secretion: (1) CCK levels increase in the serum in response to a meal, (2) administration of exogenous CCK at the same levels produced by a meal stimulates pancreatic protein secretion to higher levels than those generated by a meal (the meal may also stimulate the release of inhibitory factors in addition to CCK), and (3) a specific CCK inhibitor reduces pancreatic protein secretion by more than 50%.

The most potent stimulator of CCK release from I cells is lipid. Protein digestive products (i.e., peptones, amino acids) also increase CCK release, but carbohydrate and acid have little effect. CCK secretion may also be stimulated by CCK-releasing factors, which are peptides released by mucosal cells of the duodenum or secreted by the pancreas. The level of these releasing factors may reflect a balance between the relative amounts of nutrients and digestive enzymes that are present in the gut lumen at any one time, so the level of the factors reflects the digestive milieu of the duodenum. In the fasting state, luminal CCK releasing factors are degraded by digestive enzymes that accompany basal pancreatic secretion, so little releasing factor remains to stimulate the I cells. However, during a meal, the digestive enzymes are diverted to the digestion of ingested nutrients entering the gut lumen, and the CCK-releasing factors are spared degradation. Hence, the relative level of proteins to proteases in the small intestine determines the amount of CCK-releasing factor available to drive CCK release and thus pancreatic secretion.

CCK acts on the acinar cell through both direct and indirect pathways: it directly stimulates enzyme secretion through a CCKA receptor on the acinar cell (Fig. 43-4), and it may indirectly stimulate enzyme secretion by activating the parasympathetic (cholinergic) nervous system. As we see later, the parasympathetic pathway plays a major role in mediating the intestinal phase of pancreatic secretion. Vagal stimulation can drive pancreatic secretion to nearly maximum levels. Atropine, an antagonist of muscarinic ACh receptors (see Chapter 14), reduces the secretion of enzymes and HCO3 during the intestinal phase of a meal. Atropine also inhibits secretion in response to stimulation by physiological levels of exogenous CCK. Together, these findings suggest that CCK somehow stimulates the parasympathetic pathway, which, in turn, stimulates muscarinic receptors on the acinar cell.

Like CCK, GRP—which is structurally related to bombesin—may also be a physiological regulator of pancreatic enzyme secretion. Stimulation of acinar cells with GRP leads to enzyme secretion. In contrast to the hormone CCK, the major source of GRP appears to be the vagal nerve terminals.

Secretin is the most potent humoral stimulator of fluid and HCO3 secretion by the pancreas (Fig. 43-6). Secretin is released from neuroendocrine cells (S cells) in the mucosa of the small intestine in response to duodenal acidification and, to a lesser extent, bile acids and lipids. To stimulate secretin secretion, duodenal pH must fall to less than 4.5. Like CCK, secretin levels increase after the ingestion of a meal. However, when these levels are reached experimentally by administration of exogenous secretin, pancreatic HCO3 secretion is less than that generated by a meal. These findings suggest that secretin is acting in concert with CCK, ACh, and other agents to stimulate HCO3secretion.

In addition to hormones of intestinal origin, insulin and other hormones secreted by the islets of Langerhans within the pancreas (see Chapter 51) may also influence pancreatic exocrine secretion. Blood flow from the pancreatic islets moves to the exocrine pancreas through a portal system. This organization allows high concentrations of islet hormones to interact with pancreatic acinar cells. One result of this arrangement may be that insulin modifies the composition of digestive enzymes within the acinar cell and increases the relative levels of amylase.

Regulation of exocrine pancreatic secretion is complex, and understanding this process has been made difficult by the following: (1) tissue levels of an exogenously infused hormone may not match those generated physiologically; (2) because several neurohumoral factors are released in response to a meal, the infusion of a single agent may not accurately reflect its physiological role; (3) specific neurohumoral inhibitors are often unavailable; and (4) pancreatic responses may differ depending on the species.

A meal triggers cephalic, gastric, and intestinal phases of pancreatic secretion that are mediated by a complex network of neurohumoral interactions

The digestive period has been divided into three phases (Table 43-2) based on the site at which food acts to stimulate pancreatic secretion, just as for gastric secretion (see Chapter 42). These three phases (cephalic, gastric, and intestinal) are sequential and follow the progression of a meal from its initial smell and taste to its movement through the gastrointestinal tract (Fig. 43-9). These phases act in a coordinated fashion to maximize efficiency of the digestive process. For example, stimulation of secretion before the entry of food into the small intestine during the cephalic and gastric phases ensures that active enzymes are present when food arrives. Conversely, suppression of secretion during the late digestive phase suppresses the release of pancreatic enzymes when nutrients are no longer present in the proximal end of the small intestine.


Figure 43-9 Three phases of pancreatic secretion. A, During the cephalic phase, the sight, taste, or smell of food stimulates pancreatic acinar cells, through the vagus nerve and muscarinic cholinergic receptors, to release digestive enzymes and, to a lesser extent, stimulates duct cells to secrete HCO3 and fluid. The release of gastrin from G cells is not important during this phase. During the gastric phase, the presence of food in the stomach stimulates pancreatic secretions—primarily from the acinar cells—through two routes. First, distention of the stomach activates a vagovagal reflex. Second, protein digestion products (peptones) stimulate G cells in the antrum of the stomach to release gastrin, which is a poor agonist of the CCKA receptors on acinar cells. B, The arrival of gastric acid in the duodenum stimulates S cells to release secretin, which stimulates duct cells to secrete HCO3 and fluid. Protein and lipid breakdown products have two effects. First, they stimulate I cells to release CCK, which causes acinar cells to release digestive enzymes. Second, they stimulate afferent pathways that initiate a vagovagal reflex that primarily stimulates the acinar cells through M3 cholinergic receptors.

Table 43-2 The Three Phases of Pancreatic Secretion


The Cephalic Phase During the cephalic phase, the sight, taste, and smell of food usually generate only a modest increase in fluid and electrolyte secretion (Fig. 43-9A). However, these factors have prominent effects on enzyme secretion. In most animal species, enzyme secretion increases to 25% to 50% of the maximum rate evoked by exogenous CCK. In humans, the cephalic phase is short-lived and dissipates rapidly when food is removed. The cephalic phase is mediated by neural pathways. In the dog, stimulation of several regions of the hypothalamus (dorsomedial and ventromedial nuclei and the mammillary body) enhances pancreatic secretion. The efferent signal travels along vagal pathways to stimulate pancreatic secretion through ACh, an effect blocked by atropine. The cephalic phase does not depend on gastrin or CCK release, but it is probably mediated by the stimulation of muscarinic receptors on the acinar cell.

The Gastric Phase During the gastric phase (Fig. 43-9A), the presence of food in the stomach modulates pancreatic secretion by (1) affecting the release of hormones, (2) stimulating neural pathways, and (3) modifying the pH and availability of nutrients in the proximal part of the small intestine. The presence of specific peptides or amino acids (peptones) stimulates gastrin release from G cells in the antrum of the stomach and, to a much lesser extent, G cells in the proximal part of the duodenum. The gastrin/CCKB receptor and the CCKA receptor are closely related (see Chapter 42). Although in some species the gastrin/CCKB receptor is not present on the pancreatic acinar cell, gastrin can still act—albeit not as well—through the CCKA receptor. Although physiological concentrations of gastrin can stimulate pancreatic secretion in some species, the importance of gastrin in regulating secretion in the human pancreas remains unclear. As far as local neural pathways are concerned, gastric distention stimulates low levels of pancreatic secretion, probably through a vagovagal gastropancreatic reflex. Although the presence of food in the stomach affects pancreatic secretion, the most important role for chyme in controlling pancreatic secretion occurs after the gastric contents enter the small intestine.

The Intestinal Phase During the intestinal phase, chyme entering the proximal region of the small intestine stimulates a major pancreatic secretory response by three major mechanisms (Fig. 43-9B). First, gastric acid entering the duodenum and, to a lesser extent, bile acids and lipids stimulate duodenal S cells to release secretin, which stimulates duct cells to secrete HCO3 and fluid. The acid stimulates fluid and electrolyte secretion to a greater extent than it stimulates protein secretion. Second, lipids and, to a lesser degree, peptones stimulate duodenal I cells to release CCK, which stimulates acinar cells to release digestive enzymes. Finally, the same stimuli that trigger I cells also activate a vagovagal enteropancreatic reflex that predominantly stimulates acinar cells.

The pattern of enzyme secretion—mediated by the CCK and vagovagal pathways—depends on the contents of the meal. For example, a liquid meal elicits a response that is only ~60% of maximal. In contrast, a solid meal, which contains larger particles and is slowly released from the stomach, elicits a prolonged response. Meals rich in calories cause the greatest response.

The chemistry of the ingested nutrients also affects pancreatic secretion through the CCK and vagovagal pathways. For example, perfusion of the duodenum with carbohydrates has little effect on secretion, whereas lipids are potent stimulators of pancreatic enzyme secretion. As far as lipids are concerned, triglycerides do not stimulate pancreatic secretion, but their hydrolytic products—monoglycerides and free fatty acids—do. The longer the chain length of the fatty acid, the greater is the secretory response; C-18 fatty acids generate protein secretion that is near the maximum produced by exogenous CCK. Some fatty acids also stimulate pancreatic HCO3 secretion. Because fatty acids also reduce gastric acid secretion and delay gastric emptying, they may play an important role in modulating pH conditions in the proximal part of the small intestine. Protein breakdown products are intermediate in their stimulatory effect. Nonessential amino acids have little effect on protein secretion, whereas some essential amino acids (see Chapter 58) stimulate secretion. The most potent amino acid stimulators are phenylalanine, valine, and methionine. Short peptides containing phenylalanine stimulate secretion to the same extent as the amino acid itself. Because gastric digestion generates more peptides than amino acids, it is likely that peptides provide the initial pancreatic stimulation during the intestinal phase.

The relative potency of the different nutrients in stimulating secretion is inversely related to the pancreatic reserves of digestive enzymes. Thus, the pancreas needs to release only a small portion of its amylase to digest the carbohydrate in a meal and to release only slightly greater portions of proteolytic enzymes to digest the proteins. However, a greater fraction of pancreatic lipase has to be released to efficiently digest the fat in most meals. The exocrine pancreas has the ability to respond to long-term changes in dietary composition by modulating the reserves of pancreatic enzymes. For example, a diet that is relatively high in carbohydrates may lead to a relative increase in pancreatic amylase content.

The pancreas has large reserves of digestive enzymes for carbohydrates and proteins, but not for lipids

The exocrine pancreas stores more enzymes than are required for digesting a meal. The greatest pancreatic reserves are those required for carbohydrate and protein digestion. The reserves of enzymes required for lipid digestion—particularly for triglyceride hydrolysis—are more limited. Even so, nutrient absorption studies after partial pancreatic resection show that maldigestion of dietary fat does not occur until 80% to 90% of the pancreas has been removed. Similar reserves exist for pancreatic endocrine function. These observations have important clinical implications because they indicate that individuals can tolerate large pancreatic resections for tumors without fear of developing maldigestion or diabetes postoperatively. When fat maldigestion or diabetes does develop because of pancreatic disease, the gland must have undergone extensive destruction.

Fat in the distal part of the small intestine inhibits pancreatic secretion

Once maximally stimulated, pancreatic secretion begins to decrease after several hours. Nevertheless, the levels of secretion remain adequate for digestion. Regulatory systems only gradually return secretion to its basal (interdigestive) state. The regulatory mechanisms responsible for this feedback inhibition are less well characterized than those responsible for stimulating pancreatic secretion. The presence of fat in the distal end of the small intestine reduces pancreatic secretion in most animals, including humans. This inhibition may be mediated by peptide YY (PYY), which is present in neuroendocrine cells in the ileum and colon. PYY may suppress pancreatic secretion by acting on inhibitory neural pathways, as well as by decreasing pancreatic blood flow. Somatostatin (particularly SS-28; see Chapter 48), released from intestinal D cells, and glucagon, released from pancreatic islet α cells, may also be factors in returning pancreatic secretion to the interdigestive state after a meal.

Several mechanisms protect the pancreas from autodigestion

Premature activation of pancreatic enzymes within acinar cells may lead to autodigestion and could play a role in initiating pancreatitis. To prevent such injury, the acinar cell has certain mechanisms for preventing enzymatic activity (Table 43-3). First, many digestive proteins are stored in secretory granules as inactive precursors or zymogens. Under normal conditions, zymogens become activated only after entering the small intestine. There, the intestinal enzyme enterokinase converts trypsinogen to trypsin, which initiates the conversion of all other zymogens to their active forms (see Chapter 45). Second, the secretory granule membrane is impermeable to proteins. Thus, the zymogens and active digestive enzymes are sequestered from proteins in the cytoplasm and other intracellular compartments. Third, enzyme inhibitors such as pancreatic trypsin inhibitor are co-packaged in the secretory granule. Sufficient pancreatic trypsin inhibitor is present in the secretory granules to block 10% to 20% of the potential trypsin activity. Fourth, the condensation of zymogens, the low pH, and the ionic conditions within the secretory pathway may further limit enzyme activity. Fifth, enzymes that become prematurely active within the acinar cell may themselves be degraded by other enzymes or be secreted before they can cause injury.

Acute Pancreatitis

Acute pancreatitis is an inflammatory condition that may cause extensive local damage to the pancreas, as well as compromise the function of other organs such as the lungs. The most common factors that initiate human acute pancreatitis are alcohol ingestion and gallstones. However, other insults may also precipitate acute pancreatitis. Hypertriglyceridemia, an inherited disorder of lipid metabolism, is one such culprit. Less commonly, toxins that increase ACh levels, such as cholinesterase inhibitors (some insecticides) or the sting of scorpions found in the Caribbean and South and Central America, may lead to pancreatitis. Supraphysiological levels of ACh probably cause pancreatitis by overstimulating the pancreatic acinar cell.

Experimental models of pancreatitis suggest a primary defect in protein processing and acinar cell secretory function. More than 100 years ago, it was found that treating animals with doses of ACh 10 to 100 times greater than those that elicited maximal enzyme secretion caused “hyperstimulation” pancreatitis. The same type of injury can be generated by CCK. The injury in this model appears to be linked to two events within the acinar cell: (1) zymogens, in particular proteases, are pathologically processed within the acinar cell into active forms; in this model, the protective mechanisms outlined in Table 43-3 are overwhelmed, and active enzymes are generated within the acinar cell; and (2) acinar cell secretion is inhibited, and the active enzymes are retained within the cell. Although premature activation of zymogens is probably an important step in initiating pancreatitis, other events are important for perpetuating injury, including infl ammation, induction of apoptosis, vascular injury, and occlusion that results in decreased blood flow and reduced tissue oxygenation (ischemia).

Knowledge of the mechanisms of acute pancreatitis may lead to effective therapies. In experimental models, serine protease inhibitors that block the activation of pancreatic zymogens improve the course of the acute pancreatitis. In some clinical forms of pancreatitis, prophylactic administration of the protease inhibitor gabexate appears to reduce the severity of the disease.

Table 43-3 Mechanisms That Protect the Acinar Cell from Autodigestion

Protective Factor


Packaging of many digestive proteins as zymogens

Precursor proteins lack enzymatic activity

Selective sorting of secretory proteins and storage in zymogen granules

Restricts the interaction of secretory proteins with other cellular compartments

Protease inhibitors in the zymogen granule

Block the action of prematurely activated enzymes

Condensation of secretory proteins at low pH

Limits the activity of active enzymes

Nondigestive proteases

Degrade active enzymes

Degradation of prematurely active enzymes may be mediated by other enzymes that are present within the secretory granule or by mixing secretory granule contents with lysosomal enzymes that can degrade active enzymes. Three mechanisms lead to the combination of digestive proteases and lysosomal enzymes: (1) lysosomal enzymes may be copackaged in the secretory granule; (2) secretory granules may selectively fuse with lysosomes (a process called crinophagy); and (3) secretory granules, as well as other organelles, may be engulfed by lysosomes (a process called autophagy). Failure of these protective mechanisms may result in the premature activation of digestive enzymes within the pancreatic acinar cell and may initiate pancreatitis.


Different salivary acinar cells secrete different proteins

The organizational structure of the salivary glands is similar to that of the pancreas (Fig. 43-1A); secretory acinar units drain into progressively larger ducts. Unlike the pancreas, the salivary glands are more heterogeneous in distribution and contain two distinct acinar cell populations that synthesize and secrete different protein products. The acinar cells of the parotid glands in most species secrete a serous (i.e., watery) product that contains an abundance of α-amylase. Many acinar cells of the sublingual glands secrete a mucinous product that is composed primarily of mucin glycoproteins. The morphologic appearance of these two acinar cell populations differs as well. The submandibular gland of many species contains both mucus-type and serous-type acinar cells. In some species, these two distinct cell types are dispersed throughout the submandibular gland, whereas in other species such as humans, distinct mucus and serous acinar units are the rule. In addition to α-amylase and mucin glycoproteins, salivary acinar cells also secrete many proline-rich proteins. Like mucin proteins, proline-rich proteins are highly glycosylated, and like other secreted salivary proteins, they are present in the acinar secretory granules and are released by exocytosis.

Cholinergic and adrenergic neural pathways are the most important physiological activators of regulated secretion by salivary acinar cells

Unlike the pancreas, in which humoral stimulation plays an important role in stimulating secretion, the salivary glands are mostly controlled by the autonomic nervous system (see Chapter 14). The major agonists of salivary acinar secretion are ACh and norepinephrine, which are released from postganglionic parasympathetic and sympathetic nerve terminals, respectively (see Fig. 14-8Table 43-4). The cholinergic receptor on the salivary acinar cell is the muscarinic M3 glandular subtype. The adrenergic receptors identified on these cells include both the α and β subtypes. Other receptors identified in salivary tissue include those for substance P (NK1 receptors), VIP, purinergic agonists (P2z receptors), neurotensin, prostaglandin, and epidermal growth factor (EGF). However, some of these other receptors are found only on specific salivary glands and may be present on duct cells rather than acinar cells. Significant species variation with regard to surface receptor expression is also seen. Thus, for the salivary glands, it is difficult to discuss the regulation of acinar cell secretion in general terms. It is fair to say, however, that both cholinergic and adrenergic neurotransmitters can stimulate exocytosis by salivary acinar cells.

Table 43-4 Autonomic Control of Salivary Secretion


Both cAMP and Ca2+ mediate salivary acinar secretion

Protein secretion by the salivary acinar cell, as in the pancreatic acinar cell, is associated with increases in both [cAMP]i and [Ca2+]i. Activation of cAMP through the β-adrenergic receptor is the most potent stimulator of amylase secretion in the rat parotid gland. Activation of Ca2+ signaling pathways through the α-adrenergic, muscarinic, and substance P receptors also stimulates amylase secretion by the parotid gland. Increases in [Ca2+]i cause G protein–dependent activation of PLC and thus lead to the formation of inositol 1, 4, 5-trisphosphate (IP3) and diacylglycerol (DAG). IP3 releases Ca2+ from intracellular stores and stimulates Ca2+-dependent protein kinases such as PKC and calmodulin kinase, whereas DAG directly activates PKC (see Chapter 3). The repetitive spikes in [Ca2+]i in salivary acinar cells, as in pancreatic acinar cells, depend on Ca2+-induced Ca2+ release from intracellular stores (see Chapter 9) and on the influx of extracellular Ca2+. ATP co-released with norepinephrine (see Chapter 14) activates a P2z receptor, which is a receptor-gated cation-selective channel that allows Ca2+ to enter across the plasma membrane and thus increase [Ca2+]i.

Fluid and electrolyte secretion is the second major function of salivary acinar cells, accounting for ~90% of total salivary volume output under stimulatory conditions. The mechanisms in salivary acinar cells are similar to those in pancreatic acinar cells (Fig. 43-5). The primary secretion of the salivary acinar cell is isotonic and results largely from the basolateral uptake of Cl through Na/K/Cl cotransporters, working in conjunction with Na-K pumps and basolateral K+ channels. Secretion of Cl and water into the lumen is mediated by apical Cl and aquaporin water channels. Na+ and some water reach the lumen through paracellular routes. The salivary acinar cells in some species express carbonic anhydrase as well as parallel basolateral Cl-HCO3 and Na-H exchangers, a finding suggesting that other pathways may also contribute to the primary secretion.

Stimulation of fluid and electrolyte secretion by salivary acinar cells is largely mediated by cholinergic and α-adrenergic stimulation. Substance P, acting through its own receptor, also initiates conductance changes in the salivary acinar cell. All these effects seem to be mediated by rises in [Ca2+]i. Apical Cl channels and basolateral K+ channels appear to be the effector targets of the activated Ca2+ signaling pathway. Phosphorylation of these channels by Ca2+-dependent kinases may affect the probability that these channels will be open and may thus increase conductance.


Salivary duct cells produce a hypotonic fluid that is poor in NaCl and rich in KHCO3

In the salivary glands, as in the pancreas, the duct modifies the composition of the isotonic, plasma-like primary secretion of the acinar cells (Fig. 43-10). The active transport activity of these cells is reflected by numerous basolateral membrane infoldings and abundant mitochondria, which give the basal portion of the cells a characteristic striated appearance—hence the term striated duct epithelial cell (Fig. 43-10C). In general, salivary duct cells absorb Na+ and Cl and, to a lesser extent, secrete K+ and HCO3. Because the epithelium is not very water permeable, the lumen thus becomes hypotonic. However, significant differences are seen in the various types of salivary glands.


Figure 43-10 Salivary duct transporters.

Reabsorption of Na+ by salivary duct cells is a two-step transcellular process. First, Na+ enters the cell from the lumen through apical epithelial Na+ channels (ENaCs; see Chapter 5). Second, the basolateral Na-K pump extrudes this Na+. Elevated [Na+]i provides feedback inhibition by downregulating ENaC activity, presumably through the ubiquitin-protein ligase Nedd4 (see Chapter 2).

Reabsorption of Cl is also a two-step transcellular process. Entry of Cl across the apical membrane occurs through a Cl-HCO3 exchanger. To a certain extent, apical Cl channels, including CFTR, recycle the Cl that is absorbed by the Cl-HCO3 exchanger. Duct cells also have basolateral Cl channels that provide an exit pathway for Cl.

Secretion of HCO3 occurs through the apical Cl-HCO3 exchanger mentioned earlier. This process depends on functional CFTR, thereby confirming the coupling of CFTR to the Cl-HCO3 exchanger in salivary duct cells. HCO3accumulation inside the salivary duct cell may follow the same routes as in the pancreatic duct cell (Fig. 43-6). Indeed, Na/HCO3 cotransporters in the identification of rat and human salivary duct epithelial cells support this possibility. (See Note: Bicarbonate Secretion by the Salivary Duct)

Secretion of K+ occurs through the basolateral uptake of K+ through the Na-K pump. The mechanism of the apical K+ exit step is not well established, but it may involve K-H exchange or other pathways.

Parasympathetic stimulation decreases Na+ absorption, whereas aldosterone increases Na+ absorption by duct cells

Regulation of duct cell transport processes is less well understood in the salivary glands than in the pancreas. In the intact salivary gland (i.e., acini and ducts), secretion is stimulated primarily by parasympathetic input through ACh. In the duct cell, cholinergic agonists, acting through muscarinic receptors, increase [Ca2+]i and presumably activate Ca2+-dependent regulatory pathways. The effector targets of this Ca2+ signaling pathway are not known. The role played by duct cells in the increased saliva production that occurs in response to cholinergic stimulation is limited and may reflect decreased Na-Cl absorption more than increased K-HCO3 secretion.

The specific effects of adrenergic stimulation on duct cell transport activity are unclear. Nevertheless, activation of the β-adrenergic receptor increases [cAMP]i and activates the CFTR Cl channel.

Salivary duct cell function is also regulated by circulating hormones. The mineralocorticoid hormone aldosterone stimulates the absorption of Na-Cl and secretion of K+ by salivary duct cells in several species. Although its role has not been well examined in salivary duct cells, aldosterone in other Na+-absorbing epithelia (e.g., kidney and colon) stimulates Na+ transport by increasing both ENaC and Na-K pump activity (see Chapter 35). Salivary duct cells may also have receptors for certain neuropeptides such as VIP, although their physiological significance remains unknown.

Salivary duct cells also secrete and take up proteins

Duct cells handle proteins in three ways. Some proteins that are synthesized by duct cells are secreted into the lumen, others are secreted into blood, and still others are reabsorbed from the lumen to the cell.

Intralobular duct epithelial cells in rodent submandibular glands synthesize various proteins that are stored in intracellular granules and are secreted in response to neurohumoral stimuli. EGF, nerve growth factor, and kallikrein are among the most abundant proteins that are packaged for secretion by these cells. Salivary duct cells may also synthesize, store, and secrete some digestive enzymes (α-amylase and ribonucleases). Degranulation of intralobular duct cells occurs primarily in response to α-adrenergic stimulation, a finding suggesting that protein secretion by duct cells is regulated primarily by the sympathetic division.

Although regulatory peptides (i.e., glucagon and somatostatin) have also been detected in salivary duct cells, no evidence indicates that they are stored in granules or are secreted into the lumen (i.e., they may be basolaterally secreted as peptide hormones). In addition, duct cells synthesize polymeric IgA receptors that are responsible for the basolateral endocytosis of IgA, and they also synthesize a secretory component that facilitates the apical release of IgA.

Salivary duct cells can also remove organic substances from the duct lumen. Endocytosis of acinar proteins and other materials (e.g., ferritin) at the apical pole of the duct cell has been demonstrated immunocytochemically. In addition, salivary duct cells express the transferrin receptor (see Chapter 2) on the apical membrane, a finding indicating that some regulated endocytosis also occurs in these cells. The latter process may function to take up specific luminal substances or to traffic ion transporters to and from the apical plasma membrane.


Depending on protein composition, salivary secretions can be serous, seromucous, or mucous

Most saliva (~90%) is produced by the major salivary glands: the parotid, the sublingual, and the submandibular glands. The remaining 10% of saliva comes from numerous minor salivary glands that are scattered throughout the submucosa of the oral cavity. Each salivary gland produces either a serous, a seromucous, or a mucous secretion; the definition of these three types of saliva is based on the glycoprotein content of the gland’s final secretory product. In humans and most other mammals, the parotids produce a serous (i.e., low glycoprotein content) secretion, the sublingual and submandibular glands produce a seromucous secretion, and the minor salivary glands produce a mucous secretion.

Serous secretions are enriched in α-amylase, and mucous secretions are enriched in mucin. However, the most abundant proteins in parotid and submandibular saliva are members of the group of proline-rich proteins, in which one third of all amino acids are proline. These proteins exist in acidic, basic, and glycosylated forms. They have antimicrobial properties and may play an important role in neutralizing dietary tannins, which can damage epithelial cells. In addition to these protective functions, proline-rich salivary proteins contribute to the lubrication of ingested foods and may enhance tooth integrity through their interactions with Ca2+ and hydroxyapatite. Saliva also contains smaller amounts of lipase, nucleases, lysozyme, peroxidases, lactoferrin, secretory IgA, growth factors, regulatory peptides, and vasoactive proteases such as kallikrein and renin (Table 43-5).

Table 43-5 Major Organic Components of Mammalian Saliva


Saliva functions primarily to prevent dehydration of the oral mucosa and to provide lubrication for the mastication and swallowing of ingested food. The sense of taste and, to a lesser extent, smell depend on an adequate supply of saliva. Saliva plays a very important role in maintaining proper oral hygiene. It accomplishes this task by washing away food particles, killing bacteria (lysozyme and IgA activity), and contributing to overall dental integrity. Although α-amylase is a major constituent of saliva and digests a significant amount of the ingested starch, salivary amylase does not appear to be essential for effective carbohydrate digestion in the presence of a normally functioning pancreas. The same can be said for lingual lipase. However, in cases of pancreatic insufficiency, these salivary enzymes can partially compensate for the maldigestion that results from pancreatic dysfunction.

At low flow rates, the saliva is hypotonic and rich in K+, whereas at higher flow rates, its composition approaches that of plasma

The composition of saliva varies from gland to gland and from species to species. The primary secretion of the salivary acinar cell at rest is plasma-like in composition. Its osmolality, which is mostly the result of Na+ and Cl, is ~300 mosmol/kg. The only significant difference from plasma is that the [K+] of the salivary primary secretion is always slightly higher than that of plasma. In some species, acinar cells may help to generate a Cl-poor, HCO3-rich primary secretion after salivary gland stimulation. In most species, however, salivary gland stimulation does not significantly alter acinar cell transport function or the composition of the primary secretion. The leakiness of the tight junctions between acinar cells contributes to the formation of a plasma-like primary secretory product (see Chapter 5).

The composition of the primary salivary secretion is subsequently modified by the transport processes of the duct cell (Fig. 43-10). At low (basal) flow rates, Na+ and Cl are absorbed and K+ is secreted by the duct cells of most salivary glands (Table 43-6). These transport processes generate a K+-rich, hypotonic salivary secretion at rest. The tightness of the ductal epithelium inhibits paracellular water movement and therefore contributes to the formation of a hypotonic secretory product.

Table 43-6 Electrolyte Components of Human Parotid Saliva


Unstimulated or Basal State (mM)

Stimulated (Cholinergic Agonists) (mM)










Total CO2



Data from Thaysen JH, Thorn NA, Schwartz IL: Am J Physiol 1954; 178:155-159.

At higher flow rates, the composition of the final secretory product begins to approach that of the plasma-like primary secretion (Table 43-6). This observation suggests that, as in the case of the renal tubules, the ductular transport processes have limited capacity to handle the increased load that is presented to them as the flow rate accelerates. However, the extent to which the transporters are flow dependent varies from gland to gland and from species to species. Human saliva is always hypotonic, and salivary [K+] is always greater than plasma [K+]. In humans, increased salivary flow alkalinizes the saliva and increases its [HCO3]. This salivary alkalinization and net HCO3 secretion in humans neutralize the gastric acid that normally refluxes into the esophagus.

Parasympathetic stimulation increases salivary secretion

Humans produce ~1.5 L of saliva each day. Under basal conditions, the salivary glands produce saliva at a rate of ~0.5 mL/min, with a much slower flow rate during sleep. After stimulation, flow increases 10-fold over the basal rate. Although the salivary glands respond to both cholinergic and adrenergic agonists in vitro, the parasympathetic nervous system is the most important physiological regulator of salivary secretion in vivo.

Parasympathetic Control Parasympathetic innervation to the salivary glands originates in the salivatory nuclei of the brainstem (see Fig. 14-5). Both local input and central input to the salivatory nuclei can regulate the parasympathetic signals transmitted to the glands. Taste and tactile stimuli from the tongue are transmitted to the brainstem, where their signals can excite the salivatory nuclei and stimulate salivary gland secretion. Central impulses triggered by the sight and smell of food also excite the salivatory nuclei and can induce salivation before food is ingested. These central effects were best illustrated by the classic experiments of Pavlov, who conditioned dogs to salivate at the sound of a bell. For his work on the physiology of digestion, Ivan Pavlov received the 1904 Nobel Prize in Physiology or Medicine. (See Note: Ivan Petrovich Pavlov)

Preganglionic parasympathetic fibers travel in cranial nerve (CN) VII to the submandibular ganglia, from which postganglionic fibers reach the sublingual and submandibular glands (see Fig. 14-4). Preganglionic parasympathetic fibers also travel in CN IX to the otic ganglia, from which postganglionic fibers reach the parotid glands. In addition, some parasympathetic fibers reach their final destination through the buccal branch of CN V to the parotid glands or through the lingual branches of CN V to the sublingual and submandibular glands. Postganglionic parasympathetic fibers from these ganglia directly stimulate the salivary glands through their release of ACh. The prominent role of the parasympathetic nervous system in salivary function can be readily appreciated by examining the consequences of cholinergic blockage. Disruption of the parasympathetic fibers to the salivary glands can lead to glandular atrophy. This observation suggests that parasympathetic innervation is necessary for maintaining the normal mass of salivary glands. Clinically, some medications (particularly psychiatric drugs) have “anticholinergic” properties that are most commonly manifested as “dry mouth.” For some medications, this effect is so uncomfortable for the patient that use of the medication must be discontinued. Conversely, excessive salivation is induced by some anticholinesterase agents that can be found in certain insecticides and “nerve gases.”

Sjögren Syndrome

Sjögren syndrome is a chronic and progressive autoimmune disease that affects salivary secretion. Patients with Sjögren syndrome generate antibodies that react primarily with the salivary and lacrimal glands. Lymphocytes infiltrate the glands, and subsequent immunologic injury to the acini leads to a decrease in net secretory function. Expression of the Cl-HCO3 exchanger is lost in the striated duct cells of the salivary gland. Sjögren syndrome can occur as a primary disease (salivary and lacrimal gland dysfunction only) or as a secondary manifestation of a systemic autoimmune disease, such as rheumatoid arthritis. The disease primarily affects women; systemic disease usually does not develop.

Individuals with Sjögren syndrome have xerostomia (dry mouth) and keratoconjunctivitis sicca (dry eyes). Loss of salivary function causes these patients to have difficulty tasting, as well as chewing and swallowing dry food. They also have difficulty with continuous speech and complain of a chronic burning sensation in the mouth. On physical examination, patients with Sjögren syndrome have dry, erythematous oral mucosa with superficial ulceration and poor dentition (dental caries, dental fractures, and loss of dentition). Parotid gland enlargement is commonly present.

The proteins that are the targets of the immunologic attack in Sjögren syndrome are not known. Therefore, no specific therapy for the disorder is available. Until the underlying cause of Sjögren syndrome is discovered, patients will have to rely on eyedrops and frequent oral fluid ingestion to compensate for their deficiencies in lacrimal and salivary secretion. Various stimulants of salivary secretion (sialogogues), such as methylcellulose and sour candy, can also be helpful. Patients with severe involvement and functional disability are sometimes treated with corticosteroids and immunosuppressants.

Sympathetic Control The salivary glands are also innervated by postganglionic sympathetic fibers from the superior cervical ganglia that travel along blood vessels to the salivary glands (see Fig. 14-4). Although sympathetic (adrenergic) stimulation increases saliva flow, interruption of sympathetic nerves to the salivary glands has no major effect on salivary gland function in vivo. However, the sympathetic nervous system is the primary stimulator of the myoepithelial cells that are closely associated with cells of the acini and proximal (intercalated) ducts. These stellate cells have structural features of both epithelial and smooth muscle cells. They support the acinar structures and decrease the flow resistance of the intercalated ducts during stimulated secretion. Thus, the net effect of myoepithelial cell activation is to facilitate secretory flow in the proximal regions of the gland, thus minimizing the extravasation of secretory proteins that could otherwise occur during an acute increase in secretory flow.

The sympathetic division can also indirectly affect salivary gland function by modulating blood flow to the gland. However, the relative contribution of this vascular effect to the overall secretory function of the salivary glands is difficult to determine.

In addition to cholinergic and adrenergic regulation of salivary secretion, some autonomic fibers that innervate the salivary glands contain VIP and substance P. Although acinar cells in vitro respond to stimulation by substance P, the physiological significance of these neurotransmitters in vivo has not been established. Salivary secretion is also regulated, in part, by mineralocorticoids. The adrenal hormone aldosterone produces saliva that contains relatively less Na+ and more K+. The opposite effect on saliva is seen in patients with adrenal insufficiency caused by Addison disease. The mineralocorticoid effect represents the only well-established example of a humoral (i.e., non-neural) agent regulating salivary secretion.


Books and Reviews

Beger HG, Warshaw AL, Büchler M, et al. (eds): The Pancreas, 2nd ed. Cambridge, MA: Blackwell Publishing, 2008.

Dobrosielski-Vergona K (ed): Biology of the Salvary Glands. Boca Raton, FL: CRC Press, 1993.

Go VLW, DiMagno EP, Gardner JD, et al: The Pancreas: Biology, Pathobiology, and Disease, 2nd ed. New York: Raven Press, 1993.

Johnson LR, et al. (eds): Physiology of the Gastrointestinal Tract, 4th ed. New York: Raven Press, 2006.

Turner RJ, Sugiya H: Understanding salivary fluid and protein secretion. Oral Diseases 2002; 8:3-11.

Williams JA: Intracellular signaling mechanisms activated by cholecystokinin-regulating synthesis and secretion of digestive enzymes in pancreatic acinar cells. Annu Rev Physiol 2001; 63:77-97.

Journal Articles

Ishiguro H, Naruse S, Steward MC, et al: Fluid secretion in interlobular ducts isolated from guinea pig pancreas. J Physiol 1998; 511:407-422.

Jamieson J, Palade G: Synthesis, intracellular transport, and discharge of secretory proteins in stimulated pancreatic exocrine cells. J Cell Biol 1971; 50:135-158.

Petersen OH: Stimulus-secretion coupling: Cytoplasmic calcium signals and the control of ion channels in exocrine acinar cells. J Physiol 1992; 448:1-51.

Sohma Y, Gray MA, Imai Y, Argent BE: HCO3 transport in a mathematical model of the pancreatic ductal epithelium. J Membr Biol 2000; 176:77-100.

Thaysen JH, Thorn NA, Schwartz IL: Excretion of sodium, potassium, chloride, and carbon dioxide in human parotid saliva. Am J Physiol 1954; 178:155-159.

Zhao H, Xu X, Diaz J, Muallem S: Na+, K+, and H+/HCO3 transport in submandibular salivary ducts. J Biol Chem 1995; 270:19599-19605.