FUNCTIONAL ANATOMY

The stomach consists of three anatomic (fundus, corpus or body, and antrum) and two functional (oxyntic and pyloric gland) areas (Fig. 49-3). The oxyntic gland area, the hallmark of which is the oxyntic (oxys, Greek for acid) or parietal cell, comprises 80% of the organ (fundus and corpus). The pyloric gland area, the hallmark of which is the G or gastrin cell, comprises 20% of the organ (antrum). The human stomach contains approximately 1 × 10⁹ parietal cells and 9 × 10⁶ gastrin cells.²⁰ There is debate as to whether the cardia, a transition zone of 0 to 9 mm between the squamous mucosa of the esophagus and the oxyntic mucosa of the stomach, exists as a normal anatomic structure or develops as a result of abnormal reflux. Autopsy and endoscopic studies suggest that cardiac mucosa is absent in more than 50% of the general population.²¹ Gastric anatomy is discussed in greater detail in Chapter 47.

Figure 49-3. Functional gastric anatomy. The stomach consists of three anatomic (fundus, corpus or body, and antrum) and two functional (oxyntic and pyloric gland) areas. The hallmark of the oxyntic gland area is the parietal cell. The hallmark of the pyloric gland area is the G or gastrin cell.

The glandular area is organized in vertical tubular units that consist of an apical pit region, an isthmus, and the actual gland region that forms the lower part of the unit; the latter consists of a neck and a base (Fig. 49-4). The progenitor cell of the gastric unit, located in the isthmus, gives rise to all gastric epithelial cells. In the oxyntic gland area, the mucus-producing pit cells migrate upward from the progenitor cell toward the gastric lumen. Acid-secreting parietal cells migrate downward to the middle and lower regions of the gland; those at the bases are less active acid secretors. The turnover time for parietal cells is 54 days in mice and 164 days in rats.²⁰ Chief (zymogenic) cells predominate at the base and secrete pepsinogen and leptin²²; the latter is also present in parietal cells.²³ Several distinct neuroendocrine cell types are contained within the gland but only some of their products have been assigned physiologic functions (see Chapter 1). The cells include enterochromaffin (EC) cells (atrial natriuretic peptide [ANP], serotonin, and adrenomedullin), enterochromaffin-like (ECL) cells (histamine), D cells (somatostatin and amylin), and A-like or Gr cells (ghrelin and obestatin).^24–29 ECL cells constitute 66% of the neuroendocrine cell population in rats and 30% in humans. Somatostatin-containing D cells possess neural-like cytoplasmic processes that terminate in the vicinity of parietal and ECL cells (see Fig. 49-4). The functional correlate of this anatomic coupling is a tonic paracrine restraint exerted by somatostatin on acid secretion directly as well as indirectly by inhibiting histamine secretion (Fig. 49-5).^30–32

Figure 49-4. Gastric gland anatomy. Somatostatin-containing D cells contain cytoplasmic processes that terminate in the vicinity of acid-secreting parietal and histamine-secreting enterochromaffin-like cells in the oxyntic gland area (fundus and corpus) and gastrin-secreting G cells in the pyloric gland area (antrum). The functional correlate of this anatomic coupling is a tonic paracrine restraint on acid secretion by somatostatin that is exerted directly on the parietal cell as well as indirectly by inhibiting histamine and gastrin secretion. ANP, atrial natriuretic peptide.

(From Schubert ML, Peura DA. Control of gastric acid secretion in health and disease. Gastroenterology 2008; 134:1842-60.)

Figure 49-5. Model illustrating the inhibitory actions of somatostatin (SST) on gastric acid secretion in the oxyntic gland area (fundus and body) and the pyloric gland area (antrum). SST-containing D cells are structurally and functionally coupled to their target cells: parietal, enterochromaffin-like (ECL), and gastrin cells. SST, acting via SSTR2 receptors, tonically restrains acid secretion. This restraint is exerted directly on the parietal cell as well as indirectly by inhibiting histamine secretion from ECL cells and gastrin secretion from G cells.

Somatostatin-containing D cells are also present in the pyloric gland area (see Figs. 49-4 and 49-5); in this region they exert a tonic paracrine restraint on gastrin secretion from G cells.^33,34 The pyloric gland also contains EC cells (ANP and serotonin), A-like or Gr cells (ghrelin and obestatin), and endocrine cells containing orexin.^26,35,36

The stomach is innervated by a neural network, the enteric nervous system (ENS), that contains intrinsic neurons, that is, neurons whose cell bodies are contained within the gastric wall (Figs. 49-6 to 49-8). The ENS, the third division of the autonomic nervous system (the other two being the sympathetic and parasympathetic), is often referred to as the “little brain” because it contains as many neurons as the spinal cord, approximately 10⁶, and can function autonomous of central input (see Fig. 49-6).³⁷ Nevertheless, the ENS receives input from and sends input to the central nervous system via sympathetic and parasympathetic neurons. In rats and guinea pigs, most of the intrinsic neural innervation of the stomach originates in the myenteric plexus, located between the circular and longitudinal muscle layers; the submucosal plexus in these species, adjacent to the mucosal layer, contains only a small number of neurons. Humans, in contrast, have a clearly defined submucosal plexus that regulates gastric secretion and contains a variety of neurotransmitters (see Figs. 49-7 and 49-8). It should be noted that the vagus nerve is predominantly afferent, containing 80% to 90% afferent fibers and 10% to 20% efferent fibers. The efferent fibers arise from the dorsal motor nucleus of the brainstem. They are preganglionic and do not directly innervate parietal or neuroendocrine cells but rather synapse with postganglionic neurons of the ENS. The postganglionic neurons contain a variety of transmitters including ACh, GRP, nitric oxide (NO), VIP, and pituitary adenylate cyclase–activating polypeptide (PACAP) (see Fig. 49-8).³⁸ In the stomach, afferent nerve fibers containing calcitonin gene–related peptide (CGRP) are of extrinsic origin, that is, the cell bodies are located outside the stomach wall.³⁹ Postganglionic neurons of the ENS regulate acid secretion directly, as is the case for ACh, and/or indirectly by modulating the secretion of gastrin from G cells, somatostatin from D cells, and possibly histamine from ECL cells (see Fig. 49-8).

Figure 49-6. The autonomic nervous system consists of three divisions: sympathetic, parasympathetic, and enteric. The enteric division consists of the myenteric plexus, which regulates motility and the submucosal plexus, which regulates secretion. Although the enteric division can function autonomously, it receives input from and sends projections to the other divisions.

Figure 49-7. The enteric nervous system contains intrinsic neurons, the cell bodies of which are contained within the gastric wall. The myenteric plexus, which innervates the circular and longitudinal muscle layers, regulates motility. The submucosal plexus, which innervates the mucosa, regulates secretion.

Figure 49-8. Functional neural anatomy. The vagus nerve contains preganglionic neurons that synapse with postganglionic neurons within the wall of the stomach and that are part of the enteric nervous system. The postganglionic neurons contain a variety of transmitters including acetylcholine (ACh), gastrin-releasing peptide (GRP, or mammalian bombesin), nitric oxide (NO), vasoactive intestinal peptide (VIP), and pituitary adenylate cyclase–activating by the parietal cell polypeptide (PACAP). The postganglionic neurons regulate acid secretion by the parietal cell directly and/or indirectly by modulating the secretion of gastrin from G cells, somatostatin from D cells, and possibly histamine from enterochromaffin-like (ECL) cells.

ACID SECRETION: PARACRINE, HORMONAL, NEURAL, AND INTRACELLULAR REGULATION

Parietal cells secrete hydrochloric acid at a concentration of approximately 160 mM or pH 0.8. Acid is thought to gain access to the lumen via channels in the mucus layer created by the relatively high intraglandular hydrostatic pressures generated during secretion, about 17 mm Hg.⁴⁰

Acid facilitates the digestion of protein and absorption of iron, calcium, and vitamin B₁₂ as well as prevents bacterial overgrowth, enteric infection, and possibly community acquired pneumonia.^6–14,18,41 However, when levels of acid (and pepsin) overwhelm mucosal defense mechanisms, ulcers occur. To prevent such damage, gastric acid must be precisely regulated and produced according to need. This is accomplished by a highly coordinated interaction among a number of neural, hormonal, and paracrine pathways. These pathways can be activated directly by stimuli originating in the brain or reflexively by stimuli originating in the stomach after ingestion of a meal such as mechanical stimulation (e.g., distention) or chemical stimulation (e.g., protein and acid).

The principal stimulants of acid secretion are (1) ACh, released from postganglionic enteric neurons (neurocrine), (2) gastrin, released from antral G cells (hormonal), and (3) histamine, released from oxyntic ECL cells (paracrine) (see Fig. 49-1; Fig. 49-9). These agents interact with specific G protein–binding receptors (M₃, CCK-2, and H₂, respectively) that are coupled to two major signal transduction pathways: intracellular calcium in the case of gastrin and ACh, and adenylate cyclase or adenosine 3′,5′-cyclic monophosphate (cAMP) in the case of histamine (see Fig. 49-9). There is evidence for potentiation (or synergism) between histamine and either ACh or gastrin, probably as a result of postreceptor interaction between the two signaling pathways.⁴² The main inhibitor of acid secretion is somatostatin, released from oxyntic and antral D cells (paracrine) (see Figs. 49-1, 49-5, and 49-9). Each of these agents acts directly on the parietal cell as well as indirectly by modulating the secretion of neuroendocrine cells (Fig. 49-10).

Figure 49-9. Model illustrating parietal cell receptors and transduction pathways. The principal stimulants of acid secretion at the level of the parietal cell are histamine (paracrine), gastrin (hormonal), and acetylcholine (ACh; neurocrine). Histamine, released from enterochromaffin-like (ECL) cells, binds to H₂ receptors that activate adenylate cyclase (AC) and generate adenosine 3′,5′-cyclic monophosphate (cAMP). Gastrin, released from G cells, binds to cholecystokinin-2 (CCK-2) receptors that activate phospholipase C (not shown) to induce release of cytosolic calcium (Ca⁺⁺). Gastrin stimulates the parietal cell directly and, more importantly, indirectly by releasing histamine from ECL cells. ACh, released from intramural neurons, binds to M₃ receptors that are coupled to an increase in intracellular calcium. The intracellular cAMP- and calcium-dependent signaling systems activate downstream protein kinases, ultimately leading to fusion and activation of H⁺,K⁺-ATPase, the proton pump. Somatostatin, released from oxyntic D cells, is the principal inhibitor of acid secretion. Somatostatin, acting via the SSTR2 receptor, inhibits the parietal cell directly as well as indirectly by inhibiting histamine release from ECL cells. +, stimulatory; −, inhibitory.

Figure 49-10. Model illustrating the neural, paracrine, and hormonal regulation of gastric acid secretion. Efferent vagal fibers synapse with intramural gastric cholinergic (ACh) and peptidergic (gastrin-releasing peptide [GRP] and vasoactive intestinal peptide [VIP]) neurons. In the fundus (oxyntic mucosa), ACh neurons stimulate acid secretion directly as well as indirectly by inhibiting somatostatin (SST) secretion, thus eliminating its restraint on parietal cells and histamine-containing enterochromaffin-like (ECL) cells. In the antrum (pyloric mucosa), ACh neurons stimulate gastrin secretion directly as well as indirectly by inhibiting SST secretion, thus eliminating its restraint on gastrin-containing G cells. GRP neurons, activated by luminal protein, also stimulate gastrin secretion. VIP neurons, activated by low-grade gastric distention, stimulate SST and thus inhibit gastrin secretion. Dual paracrine pathways link SST-containing D cells to parietal cells and to ECL cells in the fundus. Histamine released from ECL cells acts via H₃ receptors to inhibit SST secretion. This serves to accentuate the decrease in SST secretion induced by cholinergic stimuli and thus augments acid secretion. In the antrum, dual paracrine pathways link SST-containing D cells to gastrin cells. Release of acid into the lumen of the stomach restores SST secretion in both the fundus and antrum; the latter is mediated via release of calcitonin gene-related peptide (CGRP) from extrinsic sensory neurons. Acute infection with Helicobacter pylori (HP) also activates CGRP neurons to stimulate SST and thus inhibit gastrin secretion. In duodenal ulcer patients who are chronically infected with HP, the organism or cytokines released from the inflammatory infiltrate inhibit SST and thus stimulate gastrin (and acid) secretion.

SOMATOSTATIN

The main inhibitor of acid secretion is somatostatin. Somatostatin is synthesized from a 92-amino acid preprosomatostatin precursor molecule that is processed to yield somatostatin-14 and somatostatin-28. Somatostatin-14 is predominantly found in stomach, pancreatic islets, and enteric neurons, whereas somatostatin-28 is the major form in small intestine. The half-life of somatostatin-14 is 1 to 3 minutes, whereas the half-life of somatostatin-28 is about 15 minutes.

In the stomach, somatostatin cells are closely coupled to their target cells (gastrin cells in the antrum, or parietal and ECL cells in the fundus/body) either directly via cytoplasmic processes or indirectly via the local circulation.^28,79 The functional correlate of this anatomic coupling is a tonic restraint exerted by somatostatin on acid secretion from the parietal cell, histamine secretion from the ECL cell, and gastrin secretion from the G cell (see Figs. 49-5 and 49-10).^{30,31,34,80,81} Removing this restraint (i.e., disinhibition or elimination of the influence of an inhibitor), by activation of cholinergic neurons, is an important physiologic mechanism for stimulating acid secretion (see Fig. 49-10). In the stomach, the actions of somatostatin are mediated primarily via the somatostatin subtype 2 receptor (SSTR2).^82–84 Gastrin, GRP, VIP, PACAP, β2/β3-adrenergic agonists, secretin, ANP, adrenomedullin, amylin, adenosine, and CGRP stimulate, whereas ACh and interferon-γ inhibit somatostatin secretion. As mentioned, an increase in luminal acidity acts to attenuate acid secretion via a pathway involving release of somatostatin in the antrum and the fundus. The change in gastric somatostatin secretion can be demonstrated over a range of pH 3 to pH 5, which is within the range observed after ingestion of a meal (Fig. 49-11).⁸⁵

Figure 49-11. Relationship between luminal pH and gastric somatostatin secretion. In isolated mouse stomach, addition of bicarbonate (HCO₃⁻) to neutralize basal acid secretion or HCl to augment luminal acidity causes a corresponding change in somatostatin secretion.

(From Schubert ML, Edwards NF, Makhlouf GM. Regulation of gastric somatostatin secretion in the mouse by luminal acid: A local feedback mechanism. Gastroenterology 1988; 94:317-22).

PARIETAL CELL INTRACELLULAR PATHWAYS

In parietal cells, acid secretion is increased by activation of intracellular cAMP- and calcium-dependent signaling pathways that activate downstream protein kinases, ultimately leading to fusion and activation of H⁺,K⁺-ATPase (the proton pump) with concomitant activation of luminal membrane conductances for K⁺ and Cl⁻ (see Fig. 49-9; Fig. 49-12). The H⁺,K⁺-ATPase actively pumps out H⁺ against a tremendous concentration gradient (cell interior pH 7.4 or 40 nM; acid secreted at pH 0.8 or 160 million nM) in exchange for luminal K⁺. The energy required comes from adenosine triphosphate (ATP) produced by the parietal cell’s extensive mitochondrial network. H⁺,K⁺-ATPase consists of an α–subunit that carries out the catalytic and transport function of the enzyme and contains sequences responsible for apical membrane localization¹⁰⁶ as well as a β-subunit, which is heavily glycosylated, and protects the enzyme from degradation and is necessary for trafficking to and from the luminal membrane.¹⁰⁷

Figure 49-12. Model illustrating ion transport in the parietal cell. Acid secretion requires a functional H⁺,K⁺-ATPase as well as apical K⁺ and Cl⁻ channels and basolateral transporters and/or channels for K⁺, Cl⁻, and HCO₃⁻. Acid is produced from the hydration of CO₂ to form H⁺ and HCO₃⁻, a reaction catalyzed by cytoplasmic carbonic anhydrase (CA). In the presence of luminal K⁺, H⁺,K⁺-ATPase pumps H⁺ into the lumen in exchange for K⁺. Luminal K⁺ channels (KCNE2/KCNQ1 and ROMK [not shown]) recycle K⁺ across the luminal membrane. The source of intracellular K⁺ is the basolateral Na⁺,K⁺-ATPase and the sodium-2 chloride potassium-cotransporter-1 (NKCC1). For each H⁺ secreted, a HCO₃⁻ exits the cell across the basolateral membrane via the anion exchanger (AE2, or Slc4a2). Concurrently with H⁺, Cl⁻ is extruded across the luminal membrane via an apical chloride channel. The sources of intracellular Cl⁻ are AE2, NKCCl, and the SLC26A7 channel.

In the resting unstimulated state, H⁺,K⁺-ATPase activity is sequestered within cytoplasmic tubulovesicles. On stimulation, there is a dramatic morphologic transformation as these vesicles fuse with the apical plasma membrane, resulting in a 6- to 10-fold increase in the membrane and the formation of the canalicular system (Fig. 49-13). Translocation of the H⁺,K⁺-ATPase into the canalicular membrane together with the presence of luminal K⁺ activates the enzyme.¹⁰⁸ On cessation of secretion, the H⁺,K⁺-ATPase is retrieved from the apical membrane and the tubulovesicular compartment is reestablished. The precise mechanisms regulating trafficking are not known, but data suggest that it involves actin-based microfilaments, small GTPases, docking/fusion proteins, ezrin, and clathrin.^109–111

Figure 49-13. Model illustrating translocation and activation of H⁺,K⁺-ATPase. In the resting state, H⁺,K⁺-ATPase is sequestered within cytoplasmic tubulovesicles and is inactive. On stimulation, the tubulovesicles move to and fuse with the apical membrane, forming an extensive canalicular system. Translocation of H⁺,K⁺-ATPase into the canalicular membrane together with the presence of luminal K⁺ activates the enzyme.

Acid secretion requires not only a functional H⁺,K⁺-ATPase but also apical K⁺ and Cl⁻ channels and basolateral HCO₃⁻ and Cl⁻ exchangers. Acid is produced from the hydration of CO₂ to form H⁺ and HCO₃⁻, a reaction catalyzed by cytoplasmic carbonic anhydrase (see Fig. 49-12). Because the H⁺,K⁺-ATPase is unable to pump H⁺ into the lumen without a parallel uptake of K⁺, sufficient quantities of K⁺ must be delivered to the lumen. This K⁺ recycling is accomplished by luminal KCNE2/KCNQ1 and ROMK (KCNJ1) potassium channels. KCNQ1 is a voltage-activated K⁺ channel, which, when modified by the small regulatory subunit, KCNE2, becomes voltage insensitive, constitutively open, and acid activated.^108,112,113 ROMK may be regulated by the cystic fibrosis transmembrane conductance regulator (CFTR).^{108,112–114} The concentration of K⁺ in gastric juice (8 to 12 mM) exceeds plasma K⁺ by two- to four-fold.

For each H⁺ secreted, an HCO₃⁻ ion exits the cell across the basolateral membrane via the anion exchanger 2 (AE2), Slc4a2 (see Fig. 49-12).¹¹⁵ As a result of this HCO₃⁻/Cl⁻ exchange, the pH within the parietal cell remains only slightly alkaline during acid secretion.¹¹⁶ Rapid entry of HCO₃⁻ from parietal cells into blood has been referred to as the alkaline tide. Some of this HCO₃⁻ may be taken up and secreted by surface epithelial cells.

Concurrently with H⁺, Cl⁻ is extruded across the luminal membrane via an apical chloride channel, the precise identity of which is not known. The sources of intracellular Cl⁻ are the basolateral anion exchanger 2 (AE2), sodium-2 chloride potassium-cotransporter-1 (NKCC1), and the SLC26A7 channel (see Fig. 49-12).^117,118

PROTON PUMP INHIBITORS

Current PPIs (e.g., omeprazole, lansoprazole, rabeprazole, pantoprazole, and esomeprazole) consist of two heterocyclic moieties, a pyridine and a benzimidazole ring, connected by a methylsulphinyl group (Fig. 49-14). They are weak bases (pKa 4 or 5) that concentrate in acidic spaces within the body that have a pH less than 4. The pKa of a molecule refers to the degree of willingness of the compound to accept or donate a proton and is based on a logarithmic scale such that a compound with a pKa of 5 is 10-fold more basic than a compound with a pKa of 4. When a compound is in an environment with a pH equal to its pKa, half the molecules will be protonated and half will be non-protonated. If a compound with a pKa of 4 is placed in an environment with a pH less than 1, more than 99.9% of the molecules will be protonated. PPIs are membrane permeable in the nonprotonated form and relatively impermeable in the protonated form. In blood (pH 7.4), PPIs are predominantly nonprotonated and thus pass readily into cells (time to reach peak plasma concentration, ≈2 hours; elimination half-life, ≈1 hour), diffuse through the cytoplasm, and then become protonated and trapped, probably as sulfenamides or sulfenic acid, in the acid environment of the secretory canaliculus (Fig. 49-15).¹¹⁹ As a consequence of the basic pKa of PPIs and “ion trapping,” the concentration of PPIs in the secretory canaliculus is 100,000- to 1,000,000-fold higher than in the blood. The sulfenamide rapidly reacts with cysteines on the luminally exposed α-subunit of the H⁺,K⁺-ATPase to form a covalent (electrochemical) disulfide bond (see Fig. 49-15).¹²⁰ Whereas all PPIs bind to cysteine 813, omeprazole also binds to cysteine 892, lansoprazole to cysteine 321, and pantoprazole to cysteine 822. Because only the apical-membrane inserted H⁺,K⁺-ATPase is susceptible to blockade by PPIs and an acid environment (pH < 4) is necessary for trapping and converting the PPI to its reactive species, the potency of PPIs is decreased when administered during the basal (fasting) state or when acid secretion is inhibited.^121,122 Because most pumps are inserted with breakfast, it is recommended that PPIs be taken 30 minutes to 1 hour before the first meal. If greater inhibition is needed, an additional dose should be taken before dinner. Recovery from inhibition of acid secretion occurs by de novo synthesis of pump protein (54 hours in rat). It has been postulated that reduction of the cysteine disulfide bonds by reducing agents such as glutathione (15 hours in rat) could also play a role.¹²³

Figure 49-14. Structure of proton pump inhibitors (PPIs). PPIs consist of two heterocyclic moieties, a benzimidazole ring, and a pyridine, connected by a methylsulphinyl bridge. PPIs are weak bases (pKa 4-5) that accumulate and activate in acidic spaces within the body that have a pH less than 4. Once activated within the parietal cell canaliculus, the PPI binds covalently with certain cysteine residues within the α-subunit of the inserted H⁺,K⁺-ATPase.

Figure 49-15. Model illustrating the mechanism of action of proton pump inhibitors (PPIs). PPIs reach the parietal cell from the bloodstream, diffuse through the cytoplasm, and accumulate in the acid environment of the secretory canaliculus. In the canaliculus, the PPI becomes protonated and trapped as a sulfenic acid followed by dehydration to a sulfenamide. The sulfenamide binds covalently by disulfide bonds to one or more cysteines of the H⁺,K⁺-ATPase to inhibit the enzyme. Whereas all PPIs bind to cysteine 813, omeprazole also binds to cysteine 892, lansoprazole to cysteine 321, and pantoprazole to cysteine 822.

HELICOBACTER PYLORI–INDUCED PERTURBATIONS IN ACID SECRETION (see also Chapter 50)

Acute infection with H. pylori (HP) results in hypochlorhydria,^149–152 whereas chronic infection results in either hypo- or hyperchlorhydria (Fig. 49-16). Appreciation of the pathways discussed earlier provides some insight into the mechanisms whereby HP colonizes the stomach and may lead to ulceration.

Figure 49-16. Model illustrating the consequences of Helicobacter pylori (HP) infection on gastric acid secretion. Acute HP infection is associated with an increase in somatostatin (SST) and thus a decrease in gastrin and acid secretion. Chronic infection may be associated with either decreased or increased acid secretion depending on the severity and distribution of gastritis. Most patients chronically infected with HP manifest a pan-gastritis and exhibit decreased acid secretion. A minority of chronically infected patients manifest an antral-predominant gastritis; these patients, who are predisposed to duodenal ulcer, produce increased amounts of acid as a result of reduced antral SST secretion and a subsequent reciprocal increase in gastrin secretion.

(From Schubert ML, Peura DA. Control of gastric acid secretion in health and disease. Gastroenterology 2008; 134:1842-60.)

The decrease in acid secretion during acute HP infection is thought to facilitate survival of the organism and its colonization of the stomach.¹⁵³ The mechanism whereby HP inhibits acid secretion is multifactorial and includes (1) direct inhibition of the parietal cell (and perhaps ECL cell) by a constituent of the bacterium (e.g., vacuolating cytotoxin, lipopolysaccharide, or acid-inhibitory factor) and (2) indirect inhibition of parietal cell function as a result of changes in cytokines as well as hormonal, paracrine, and neural regulatory mechanisms.^154–157 HP itself inhibits human H⁺,K⁺-ATPase α-subunit gene expression.¹⁵⁸ It also elicits secretion of at least two cytokines, IL-1β and tumor necrosis factor-α, that directly inhibit parietal cell secretion.¹⁵⁹ In preliminary studies we have shown that HP activates CGRP sensory neurons coupled to stimulation of somatostatin and thus inhibition of gastrin, histamine, and acid secretion.¹⁶⁰

Chronic infection with HP may be associated with either decreased or increased acid secretion depending on the severity and distribution of gastritis (see Fig. 49-16).¹⁶¹ Most patients chronically infected with HP manifest a pangastritis and produce less than normal amounts of acid.¹⁶² Reduced acid secretion, initially, is due to functional inhibition of parietal cells by either products of HP itself or, more likely, products of the inflammatory process, as discussed earlier for acute infection^163,164; this is usually reversible on eradication of the organism.^165–167 In such patients, HP may be protective against gastroesophageal reflux disease (GERD), Barrett’s esophagus, and esophageal adenocarcinoma, as well as augment the antisecretory effect of PPIs.^168,169 Conversely, rebound acid hypersecretion may occur in HP-eradicated patients when long-term (i.e., >8 weeks) use of PPIs is discontinued; this may unleash or exacerbate GERD, particularly in patients with large hiatal hernias.^170,171 Acid hypersecretion persists at least eight weeks and appears to be due to hypergastrinemia-induced increases in parietal and ECL cell masses.¹⁷² The reason the phenomenon does not occur in HP-positive individuals who stop taking PPIs may be due to the fact that HP as well as the cytokines produced by the inflammatory infiltrate inhibit acid secretion and thus mask the rebound. With time, atrophy of oxyntic glands with loss of parietal cells may occur in patients chronically infected with HP, resulting in irreversible achlorhydria.

Autoimmune gastritis is an inflammatory disorder of the oxyntic mucosa often associated with antiparietal cell autoantibodies directed against H⁺,K⁺-ATPase with subsequent loss of parietal cells.¹⁷³ H⁺,K⁺-ATPase is a major autoantigen in a subset of patients infected with HP and these antibodies may play a role in the subsequent development of atrophic gastritis. It is postulated that antibodies are acquired due to molecular mimicry between HP lipopolysaccharide and H⁺,K⁺-ATPase, both of which contain Lewis epitopes.¹⁷⁴

About 10% to 15% of patients chronically infected with HP have antral-predominant inflammation. These patients, who are predisposed to duodenal ulcer (see Chapters 50, 51, and 52), produce increased amounts of acid as a result of reduced antral somatostatin content and elevated basal and stimulated gastrin secretion (see Fig. 49-16).^175–177 Gastrin stimulates histamine secretion from fundic/body ECL cells and induces ECL hyperplasia.¹⁷⁸ The mechanism by which somatostatin secretion is decreased is not known but may involve cytokines induced by the inflammation and/or the production of N^a-methyl histamine, a selective H₃-receptor agonist, by HP.^179,180 One may speculate that the H₃-receptor agonist could diffuse across the antral mucosa to interact with H₃ receptors on antral somatostatin cells, causing inhibition of somatostatin secretion, and, thus, stimulation of gastrin secretion.⁴⁵ In addition, IL-8 and platelet activating factor are upregulated in HP-infected mucosa and are capable of stimulating gastrin release from isolated G cells.^181,182

MEASUREMENT OF GASTRIC ACID SECRETION

DISEASES ASSOCIATED WITH INCREASED GASTRIC ACID SECRETION

Duodenal ulcer patients, as a group, manifest increased basal and stimulated gastrin and acid production (see Table 49-1).¹⁹⁰ It is recognized that most cases of duodenal ulcer are due to infection with HP (Chapter 52) and that this infection is responsible for the perturbations in acid secretion observed in these patients. Pentagastrin-stimulated PAO, an indicator of functional parietal cell mass, is increased in HP-infected duodenal ulcer patients as is GRP-stimulated peak acid output, an indicator of the stomach’s functional response to endogenous gastrin.^177,191,192 Suppression of somatostatin secretion by the infection may be the root cause for these changes (see Fig. 49-10; Fig. 49-16). Eradication of HP restores somatostatin as well as basal and stimulated gastrin and acid secretion, over time, to normal in most individuals, thus providing a permanent cure for duodenal ulcer disease.^{176,177,191,193–195}

In contrast to duodenal ulcer, gastric ulcer patients, as a group, exhibit normal or decreased basal and stimulated acid production (see Table 49-1), even though they too are often infected with HP. This suggests that altered gastric mucosal defense may be of primary pathophysiologic importance. Gastric ulcers have been classified according to their location and concomitant association with duodenal ulcer.¹⁹⁶ Type I ulcers occur in the gastric body and are generally characterized by low acid secretion. These findings may reflect a greater degree and more generalized mucosal inflammation of the oxyntic mucosa with reduced functional parietal cell mass. Type II ulcers occur in the antrum and are characterized by low, normal, or high acid secretion. Type III ulcers occur within 3 cm of the pylorus, commonly accompany duodenal ulcer, and are characterized by high acid output. Type IV ulcers occur in the gastric cardia and are characterized by low acid secretion.¹⁹⁷ Accordingly, the more distant a gastric ulcer is from the pylorus the more likely acid secretion will be low.

A number of uncommon conditions are marked by gastric acid hypersecretion and subsequent ulceration (see Chapter 52). In patients with systemic mastocytosis, high histamine levels, as a consequence of increased numbers of mast cells, continuously stimulate parietal cells to secrete acid.¹⁹⁸ When a portion of gastric antrum is retained in the afferent remnant after antrectomy with Billroth II anastomosis, it is bathed in alkaline secretions leading to decreased somatostatin secretion, hypergastrinemia, increased acid production, and anastomotic ulceration.^65,85,199 Acid hypersecretion also can result from chronic hypercalcemia of any cause because calcium directly stimulates gastrin secretion from G cells and acid secretion from parietal cells.^200,201

The best characterized acid hypersecretory condition is ZES,^202–204 as discussed in Chapter 32. The BAO is almost always higher than 15 mmol/hr and the BAO/PAO ratio is usually 0.6 or greater (see Table 49-1). Gastrin, synthesized by the tumor, is secreted into the bloodstream, where it binds to CCK-2 receptors on acid-producing parietal and histamine-containing ECL cells to induce secretion as well as proliferation. The clinical correlate of the proliferation is rugal hypertrophy with prominent gastric folds.

Diagnosis and treatment of gastrinoma are discussed in detail in Chapter 32. The basis of the secretin test to diagnose gastrinoma is that normally somatostatin cells in the antrum tonically restrain gastrin secretion from G cells. Secretin stimulates the G cell directly and, at the same time, inhibits the G cell indirectly by stimulating somatostatin secretion; the effect of the latter usually dominates and gastrin is not stimulated to a significant degree (Fig. 49-17). Because the gastrinoma does not contain functionally coupled somatostatin cells, the effect of secretin is solely stimulation of gastrin secretion from the tumor.^205–207 Almost all gastrinomas contain somatostatin receptors and somatostatin receptor scintigraphy (SRS) using [¹¹¹In-DPTA-Dphe1]-octreotide is considered the initial localization study of choice, with a 71% sensitivity and 86% specificity for primary tumors and 92% detection for metastatic disease.^208,209 PPIs are the antisecretory therapies of choice and are able to control acid secretion and prevent complications in most patients with ZES.²¹⁰

Figure 49-17. Zollinger-Ellison syndrome (ZES). Model illustrating the action of gastrin in oxyntic mucosa and secretin in pyloric mucosa of stomach in patients with ZES. Gastrin, synthesized and secreted by the gastrinoma into the bloodstream, acts via cholecystokinin-2 (CCK-2) receptors on acid-secreting parietal and histamine-secreting enterochromaffin-like (ECL) cells to increase acid secretion and induce cell proliferation. In the antrum, exogenous secretin (i.e., secretin stimulation test [see text and Chapter 32]) stimulates gastrin secretion directly and concomitantly inhibits gastrin secretion by stimulating somatostatin (SST) secretion resulting in little or no gastrin release. Because the gastrinoma does not contain functionally coupled SST cells, the effect of secretin in ZES patients is solely to stimulate gastrin secretion from the tumor.

(From Hung PD, Schubert ML, Mihas AA. Zollinger-Ellison syndrome. Curr Treat Options Gastroenterol 2003; 6:163-70.)

PEPSINOGEN SECRETION

Pepsinogens, which belong to a family of enzymes called gastric aspartic proteases, are inactive polypeptide proenzymes known as zymogens. They are synthesized primarily in chief cells but also in mucous neck cells. Pepsinogens are converted in the gastric lumen by gastric acid to pepsins, which contain two active-site aspartate residues. Once this reaction begins, pepsins can autocatalyze the conversion of pepsinogens to pepsins.²¹¹ Pepsins are optimally active at pH 1.8 to 3.5, reversibly inactivated at pH 5, and irreversibly denatured at pH 7. Gastric acid not only provides an optimum pH for peptic activity but itself denatures dietary protein, making it more susceptible to peptic hydrolysis. Thus, acid and pepsin work in concert to promote digestion of dietary protein. As discussed, partially digested protein stimulates gastrin and thus acid secretion.¹⁴² More recent data suggest that pepsins may also be important for killing ingested bacteria.^17,212

Pepsinogens have been electrophoretically separated into seven isozymogens. The five fractions (pepsinogen 1 to pepsinogen 5) that migrate toward the anode most rapidly at pH 5 are similar immunologically and are referred to as group I pepsinogens (PGI; old term, pepsinogen A).²¹³ PGI is expressed in chief and mucous cells of the oxyntic mucosa. Migrating slightly behind the PGIs are two immunologically similar isozymogens, pepsinogen 6 and pepsinogen 7, that are referred to as group II pepsinogens (PGII; old term, pepsinogen C). PGII, which represents approximately 20% of total pepsin content, is expressed in oxyntic and pyloric mucosa as well as in duodenal Brunner’s glands.

The most important physiologic stimulant for pepsinogen secretion is ACh released from intramural cholinergic neurons. ACh, acting via both M₁ and M₃ muscarinic receptors on chief cells, induces an increase in cytosolic calcium.²¹⁴ Calcium, in turn, activates cytosolic kinases, phosphatases, and nitric oxide synthase that induce pepsinogen secretion.²¹⁵ Other agents capable of stimulating pepsinogen secretion from chief cells via the calcium signaling pathway include CCK, gastrin, and GRP.^216–219 Agents that increase cAMP within chief cells, such as isoproterenol, secretin, VIP, and histamine also augment pepsinogen secretion as do agents that activate tyrosine kinase such as epidermal growth factor and transforming growth factor-α.²¹⁸ Inhibitors of pepsinogen secretion include somatostatin, neuropeptide Y, PYY, and IL-1β. Optimal pepsinogen secretion requires a functional sodium-2 chloride potassium cotransporter-1 (NKCC1), a basolateral cotransporter responsible for chloride uptake into secretory epithelia,¹¹⁸ including the chief cell (and the parietal cell; see Fig. 49-12). It has been postulated that loss of the flushing action of neutral gastric secretion may impair pepsinogen secretion.

Serum levels of PGI correlate with maximal acid output. A linear correlation exists between the loss of chief cells in patients with oxyntic atrophy and serum PGI^220,221; a serum PGI/PGII ratio of 2.5 or less has been used as a noninvasive test to detect gastric mucosal atrophy.^5,222 Serum PGI is also increased in humans treated with PPIs, but the precise mechanism is not known.²²³ Both PGI and PGII are filtered and metabolized by the kidney, but serum PGI concentration is increased more than PGII concentration in patients with renal insufficiency.^224,225

GASTRIC LIPASE SECRETION

Gastric lipase, secreted by chief cells of the oxyntic mucosa, helps initiate the digestion of dietary triglycerides by hydrolyzing them to free fatty acids, diglycerides, and 2-monoglycerides. The properties of gastric lipase are quite distinct from those of pancreatic lipase. Gastric lipase has a pH optimum of 4.5 to 5.5 (compared with 6.5 to 7.4 for pancreatic lipase) and does not require colipase. Furthermore, protection from peptic proteolysis by an N-glycosylated asparagine at residue 308 permits gastric lipase to retains its full activity in acidic gastric juice (pH 2) despite a high gastric juice peptic activity.^226,227

Stimulants and inhibitors of gastric lipase are similar to those for pepsinogen secretion. Aging has been reported to decrease gastric lipase secretion, but data are controversial.²²⁸ In humans, increasing the amount of lipid in the diet causes a corresponding increase in gastric lipase secretion into gastric juice.²²⁹ The amount of gastric lipase secreted after a meal is small relative to the amount of pancreatic lipase. However, the specific activity of gastric lipase is equal to or greater than that of pancreatic lipase. Thus, gastric lipase is capable of digesting 10% to 25% of dietary triglyceride.²³⁰ In patients with chronic pancreatitis, gastric lipase secretion is increased three- to four-fold and can partly, but incompletely, compensate for loss of pancreatic lipase (Fig. 49-18).²³¹ A feedback mechanism exists whereby fat in the small intestine inhibits gastric lipase secretion by a humoral mechanism, with GLP-1 a prime candidate to be the mediator.²³²

Figure 49-18. Relative outputs of human gastric lipase (left graph) human pancreatic lipase (middle graph), and lipolytic activity in duodenal contents (right graph) in healthy volunteers (violet columns) and patients with chronic pancreatitis with exocrine insufficiency (CP; beige columns). Note that despite the increase of human gastric lipase output in patients with CP, the overall magnitude of the rise is insufficient to fully correct the reduction in duodenal lipolytic activity.

(From Carrière F, Grandval P, Renou C, et al. Quantitative study of digestive enzyme secretion and gastrointestinal lipolysis in chronic pancreatitis. Clin Gastroenterol Hepatol 2005; 3:28-38.)

Orlistat (tetrahydrolipstatin) is a lipophilic compound derived from lipstatin produced by Streptomyces toxitricini and is a potent inhibitor of all lipases.²³³ Orlistat reacts covalently with the catalytic serine residue of lipases.²³⁴ Orlistat decreases lipid digestion by about 35% and has been used clinically to induce modest weight loss.²³⁵ Adverse gastrointestinal events include loss of lipid-soluble vitamins, accelerated gastric emptying, fecal urgency, abdominal discomfort, flatus with discharge, and fecal urgency (see Chapter 6).^236,237

INTRINSIC FACTOR SECRETION

Intrinsic factor (IF), a 50-kd glycoprotein secreted by human parietal cells and, to a lesser degree, chief cells, is necessary for the absorption of cobalamin (vitamin B₁₂).²³⁸

Although all stimulants (e.g., gastrin, histamine, and acetylcholine) and inhibitors (e.g., somatostatin) of gastric acid secretion discussed previously have similar effects on IF secretion, the secretion of IF is not coupled to acid secretion. PPIs, for example, inhibit acid but have no significant effect on basal or stimulated IF secretion.²³⁹ Several reports indicate, however, that chronic use of PPIs may result in low serum levels of cobalamin, probably as a result of impaired acid-facilitated release of cobalamin from food.^240,241 The recommended daily allowance of cobalamin is 2 µg/day and total body stores are 2.5 mg. Cobalamin deficiency due to antisecretory therapy is therefore rare because acid secretion is not completely suppressed and because the body contains relatively large stores.^10,242,243

The delivery of cobalamin from food to tissues begins with release of cobalamin from dietary protein by the pH-dependent activity of pepsin, followed by binding of cobalamin to two binding proteins that are secreted into gastric juice: IF and haptocorrin (R binder).^244,245 Haptocorrin is also secreted in saliva and bile. Haptocorrin binds cobalamin more avidly than IF in the acidic stomach and, therefore, most cobalamin initially becomes attached to haptocorrin. In the duodenum, cobalamin is released from haptocorrin by pancreatic trypsin and the free cobalamin then binds to IF. IF-cobalamin complexes are resistant to pancreatic proteolysis and eventually bind to a specific receptor on the distal ileal mucosa. This receptor, cubilin, is expressed in clefts between microvilli and mediates endocytosis of the IF-cobalamin complex.²⁴⁶ Once within the ileal enterocyte, IF is degraded by lysosomal enzymes and cobalamin binds to transcobalamin. The cobalamin-transcobalamin complex is released into the circulation and from there enters cells by receptor-mediated endocytosis. Once within cells, cobalamin is dissociated from its transport protein and converted to its active forms, methylcobalamin and 5-deoxyadenosyl cobalamin. The active forms serve as coenzymes for methionine synthase and methylmalonyl-coenzyme A mutase, enzymes involved in methylation of homocysteine to methionine and the catabolism of branched-chain amino acids and odd-chain fatty acids in mitochondria, respectively.²⁴⁷

When radiolabeled cobalamin is administered orally after a large dose of nonradioactive cobalamin is given parenterally, patients with IF deficiency excrete much lower amounts of radioactive cobalamin in a 24-hour urine collection than do normal controls (Schilling test, part I). If IF is administered orally together with radioactive cobalamin to IF-deficient patients, urinary radioactive cobalamin excretion normalizes (Schilling test, part II). In addition to IF deficiency, cobalamin malabsorption may result from achlorhydria or hypochlorhydria (reduced peptic hydrolysis of cobalamin from food protein), bacterial overgrowth (cobalamin competed for by bacteria), pancreatic insufficiency (impaired tryptic cleavage of haptocorrin-cobalamin complex), ileal receptor defect (cubilin mutation), and ileitis or ileal resection (absent IF-cobalamin absorptive site).^248–250

Secretion of IF far exceeds the amount necessary for cobalamin absorption. Thus, in most patients with hypochlorhydria, continued IF secretion in low amounts is sufficient to prevent cobalamin deficiency. Patients with pernicious anemia, however, which affects 2% of individuals older than the age of 60, do develop cobalamin deficiency.²⁵¹ The pathology involves a chronic inflammatory, mainly lymphocytic, infiltrate of the oxyntic mucosa accompanied by loss of parietal and chief cells. The pathogenesis involves proinflammatory TH1 CD4 T cells directed toward the α- and β-subunits of the parietal cell H⁺,K⁺-ATPase, as well as circulating antibodies directed against H⁺,K⁺-ATPase or IF.²⁵² Chronic infection with HP may play a primary role in the immunologic response. It is proposed that HP-induced inflammation results in breakdown of tolerance for self antigens such as H⁺,K⁺-ATPase in genetically susceptible individuals or that antibodies are acquired due to molecular mimicry between HP lipopolysaccharide and H⁺,K⁺-ATPase, both of which contain Lewis epitopes.^173,174 Antibodies directed against H⁺,K⁺-ATPase are found in 90% of patients with pernicious anemia but the incidence of these antibodies may decrease to about 55% to 80% with progression of autoimmune gastritis presumably because of disappearance of HP and the loss of antigenic drive.^252–254

BICARBONATE SECRETION

Regulation of gastric bicarbonate (HCO₃⁻) secretion has been studied much less intensively than has duodenal HCO₃⁻ secretion. The precise function of gastric HCO₃⁻ secretion is uncertain given the overwhelming simultaneous secretion of H⁺. Nevertheless, HCO₃⁻ secretion has been implicated in the formation of a protective pre-epithelial alkaline layer (see Fig. 49-2).

Measurement of gastric HCO₃⁻ secretion has been impeded by the presence of considerable H⁺ secretion, necessitating the use of potent antisecretory compounds or measurement methods unaffected by the presence of acid. One means of measuring bulk HCO₃⁻ secretion is with the use of inline pH and CO₂ electrodes, in which HCO₃⁻ concentration is calculated using the Henderson-Hasselbach equation.^255,256 An electrophysiologic method has been used to measure alkaline secretion in the lumen of individual gastric glands isolated from the frog Rana esculenta.²⁵⁷

Gastric HCO₃⁻ secretion is an energy-dependent process. The finding that there is virtually no change in gastric electrical potential difference during HCO₃⁻ secretion suggests that HCO₃⁻ transport takes place via an electroneutral ion exchange mechanism, probably an exchange of HCO₃⁻ for Cl⁻ at the luminal surface.²⁵⁸ Although several candidate anion exchangers, solute carriers, and anion transporters have been localized to gastric surface cells including AE2, AE4, Slc4a2, Slc4a4, PAT1, and Slc26a6, there is little evidence that gastric surface cells actually secrete HCO₃⁻.^259–263 The cell responsible for HCO₃⁻ secretion is not known. The source of some of the HCO₃⁻ secreted during H⁺ secretion may actually be the parietal cell. As discussed previously, for each H⁺ secreted, a HCO₃⁻ ion exits the parietal cell across the basolateral membrane via the anion exchanger 2 (AE2), Slc4a2 (see Fig. 49-12).¹¹⁵ This HCO₃⁻ may alkalinize the blood that perfuses surface epithelial cells, be taken up by sodium bicarbonate cotransporters (NBC1 and NBC2), and then be secreted by epithelial cells in an effort to protect them from luminal acid.²⁶⁴ The parietal cell, however, may not be the only source of HCO₃⁻ for surface cells, as marked inhibition of gastric H⁺ secretion by PPIs does not significantly diminish gastric HCO₃⁻ secretion in patients with duodenal ulcer.²⁶⁵

Prostaglandin E₂ analogs stimulate gastric HCO₃⁻ and mucus secretion^266,267 and blockade of endogenous prostaglandin synthesis reduces gastric HCO₃⁻ secretion.²⁶⁸ In rats and mice, the prostaglandin E receptor subtype involved in stimulation of gastric HCO₃⁻ secretion is EP1.^269,270 Gastric mucosal prostaglandin synthesis and HCO₃⁻ secretion decline in older adults.^271,272

MUCUS SECRETION

A firmly adherent viscous mucus gel overlies the gastric surface. It is composed of 95% water and 5% extensively cross-linked mucin glycoproteins that are products of MUC genes.^273,274 Ultrastructural studies reveal alternating layers of two distinct mucin classes, MUC5AC (secreted by surface and pit area epithelial cells) and MUC6 (secreted from neck and gland cells).^275–278 Observations that the gastroprotective compound, geranylgeranylacetone (GGA), increases MUC6 expression in rat gastric mucosa²⁷⁹ and that lafutidine, a gastroprotective compound used clinically in Japan, increases mucus thickness and mucin content in humans²⁸⁰ suggest that the mucus gel may contribute to mucosal defense.^273,281,282 Furthermore, electron microscopy studies demonstrate that HP accumulates within and disrupts the MUC5AC-enriched gel layer.^277,283,284 MUC5AC secretion is increased in first-degree relatives of subjects with gastric cancer infected with HP, suggesting that mucus secretion may be a marker for a more severe inflammatory response to the organism.²⁸⁵

Mucus gel thickness can be measured continuously and noninvasively in living rodents by alternately focusing between fluorescently labeled surface epithelial cells and the mucus gel surface, as delineated by carbon particles or by fluorescent microspheres.^286–288

A pH gradient at the gastric mucosal surface has been observed in a variety of species including humans.^289–292 In most cases, the gradient is relatively alkaline at the tissue surface and gradually more acidic at distances further from the surface. The gradient is due to active bicarbonate secretion and is considered a defense against luminal acid.^290,293 Although most measurements using microelectrodes have reported values near neutrality at the gastric mucosal surface, more recent measurements using pH-sensitive fluorescent dyes with ex vivo confocal microscopy report a surface pH near pH 4 in guinea pigs and frogs.^294,295 The observation that the steady-state surface pH gradient extends beyond the thickness of the mucus gel layer suggests that the unstirred layer formed at the interface between the mucus and the aqueous lumen or the interface between the epithelial surface and the luminal contents may also play a role in mucosal defense by restricting mixing of molecules within the unstirred layer.^283,293,296

Although it has been suggested that gastric mucus may physically retard diffusion of protons, the evidence for this is equivocal. In vitro measurements of proton and bicarbonate diffusion have yielded diffusion coefficients that are up to 10 times slower in isolated mucus compared with saline solution (≈0.5 − 2 × 10⁻⁵ cm²/sec).^287,297 However, these diffusion coefficients in mucus are comparable to those found in saline for ions such as Na⁺, K⁺, and Cl⁻. Even accepting the slowest reported diffusion of protons, one-third of protons will theoretically diffuse 80 µm or greater within 10 seconds, a distance greater than the average mucus thickness in rats and mice.^281,290,296 There are also concerns regarding diffusion measurements performed in vitro as nondestructive removal of the tightly adherent mucus is extremely difficult, mucus structure and function is sensitive to environmental conditions, and measurement instruments may produce artifactual unstirred layers.^298,299

Trefoil factor (TFF) peptides are cosecreted with mucins.³⁰⁰ These are 7- to 12-kd peptides sharing a common structure of three internal disulfide bonds that yield the signature “trefoil” structure of three internal loops and are designated TFF1, TFF2, and TFF3. TFFs are markedly pepsin resistant and able to survive intact in the gastric lumen³⁰¹; the concentration of TFF2 in rat gastric mucus has been estimated at 10 µm.^302,303 The distribution of each trefoil peptide in the normal GI tract is distinct. TFF1 is stored in gastric pit and surface mucous cells, TFF2 is present in gastric gland mucous cells, and TFF3 is stored in intestinal goblet cells.^303–307 The relative abundance of TFF1 and TFF2 expression is reciprocally regulated by gastrin, H⁺,K⁺-ATPase, and inflammation.³⁰⁸

The localization and coordinated secretion of trefoil peptides with mucins suggest that they too may be involved in mucosal defense.^305,309,310 In support of this notion (1) increased TFF1 expression is observed after administration of gastroprotective agents³¹¹; (2) addition of TFF2 to mucin solutions significantly increases viscosity and elasticity in vitro and in vivo^312–313; (3) chronic treatment with PPIs increases TFF2 concentration in gastric secretions coincident with promoting repair and preventing injury in response to luminal noxious agents³¹⁴; and (4) TFF2 -/- knockout mice exhibit shortened gastric glands with decreased epithelial migration, increased net acid secretion, and a four-fold increase in number of lesions after a 12 hour-exposure to the nonselective cyclooxygenase (COX) inhibitor indomethacin.³¹⁵

ACKNOWLEDGMENT

The authors thank Mary Beatty-Brooks for the artwork and Mark Feldman, John Walsh, Andrew Soll, and Gabriel Makhlouf, pioneers in the field of gastric exocrine and endocrine secretion, for their encouragement, support, and friendship.