| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Surgery, The University of Texas Medical Branch, Galveston, Texas 77555
Correspondence: Address all correspondence and requests for reprints to: B. Mark Evers, M.D., Department of Surgery, The University of Texas Medical Branch, 301 University Boulevard, Galveston, Texas 77555-0536. E-mail: mevers{at}utmb.edu
 |
Abstract |
|---|
 Top Abstract I. IntroductionII. GI HormonesIII. Proliferation and Repair...IV. Proliferation of Neoplastic...V. Future Perspectives and...References |
|---|
|
I. Introduction |
|---|
|
TopAbstractI. IntroductionII. GI HormonesIII. Proliferation and Repair...IV. Proliferation of Neoplastic...V. Future Perspectives and...References |
|---|
|
II. GI Hormones |
|---|
|
TopAbstractI. IntroductionII. GI HormonesIII. Proliferation and Repair...IV. Proliferation of Neoplastic...V. Future Perspectives and...References |
|---|
A. GI hormone distribution, synthesis, and secretion
1. Gastrin.
The presence of gastrin was postulated in 1906 by Edkins (1), but it was not until 1938 that Komarov (2) prepared useful, histamine-free antral extracts of this gastric stimulant. Gastrin is now firmly established as the physiological regulator of gastric acid secretion and an important regulator of gastric mucosal cell proliferation (3). The major site of gastrin synthesis and secretion is the gastrin-containing cell (G cell) in the antropyloric mucosa (4, 5, 6). Other minor sites of gastrin production include the endocrine cells of the pancreas (7), pituitary (8), and extraantral G cells (9). Gastrin release is stimulated by food components, particularly aromatic amino acids and amine derivatives of amino acids, and is inhibited by luminal acid (10).
Human progastrin, the precursor of gastrin, consists of a 21-amino-acid signal peptide, a 37-amino-acid N-terminal extension, the gastrin-34 sequence, and a 9-amino-acid C-terminal extension (3, 10). In antral G cells, progastrin is stored and processed into secretory granules, and N-terminal and C-terminal extensions are removed by prohormone convertases. The C-terminal basic amino acids are sequentially removed by a carboxypeptidase, which results in the formation of glycine-extended G34 (G34-gly). G34-gly is amidated by peptidyl glycine 
-amidating monooxygenase to form G34-NH2 (G-34) or cleaved at internal lysine/lysine residues to form G-17-gly (referred to as G-gly) (10). Recent studies have demonstrated that the majority of G-17-NH2 (also known as gastrin or G-17) arises from the conversion of G34-NH2 to G-17-NH2, rather than by amidation of G-gly, suggesting that the conversion of G-gly to amidated G-17 is blocked and that G-gly is a terminal, secondary end-product of progastrin processing in normal antral G cells (10, 11). This is contrasted by evidence in colon cancer cells that gastrin bypasses the processing machinery and is expressed in larger, unprocessed forms that are transported in secretory vesicles and continuously fused with the plasma membrane and released (11).
2. CCK.
CCK was described in 1928 by Ivy and Oldberg (12) as a contaminant in impure secretin preparations. These impurities were noted to produce gallbladder contraction in dogs and cats. In 1943, Harper and Raper (13) identified a single agent extracted from intestinal samples that possessed pancreatic-stimulating activity. After complete purification and sequencing, it was determined that this single compound, CCK, was responsible for gallbladder contraction and pancreatic enzyme secretion (14). Other actions of CCK include inhibition of gastric emptying, stimulation of bowel motility, potentiation of insulin secretion, and trophic effects on the pancreas and GI mucosa (3). CCK release is stimulated by fats, proteins, and amino acids.
CCK is produced by endocrine cells of the gut (primarily duodenum and jejunum), the neurons of the brain, and the peripheral nervous system of the GI tract (15, 16, 17, 18). Ultrastructural studies demonstrate that the cells are identical to I cells of the human intestine (19). CCK neurons are found in the myenteric plexus, submucosal plexus, and circular muscle layers of the distal intestine and colon (9). Postganglionic CCK nerve fibers are found in the pancreas surrounding the islets of Langerhans (20). Furthermore, although CCK has not been found in the intrinsic neurons of the stomach and duodenum, CCK cells are in the celiac plexus and can be found in the vagus nerve, especially after injury (21, 22).
The molecular forms of CCK are diverse and appear to be tissue-specific (23, 24). The most abundant form in the brain appears to be CCK-8, although significant amounts of larger carboxy-amidated forms like CCK-33, CCK-58, and CCK-83 have been isolated.
3. BBS/GRP.
BBS, a tetradecapeptide originally isolated from the skin of the frog Bombina bombina, is analogous to mammalian GRP (24). In general, BBS/GRP may be thought of as a universal on-switch with predominantly stimulatory effects. BBS/GRP stimulates the release of all GI hormones, intestinal and pancreatic secretion, and motility (24). The most important functions of BBS/GRP are antral gastrin release and stimulation of gastric acid secretion (25, 26). This peptide also stimulates growth of the GI mucosa and pancreas (24).
BBS/GRP-like immunoreactivity is widely distributed throughout the GI tract of rats, guinea pigs, dogs, and humans (27, 28, 29, 30, 31), predominantly in the neuronal populations of the gut. GRP is most abundant in the stomach, with GRP-positive cells and fibers innervating the oxyntic and antral mucosa and circular muscle.
4. NT.
NT, a tridecapeptide originally isolated from bovine hypothalamus (32), is localized mainly in the central nervous system (predominantly hypothalamus and pituitary) and in endocrine cells (N cells) of the jejunal and ileal mucosa (33, 34). NT is released in response to increased intraluminal fats (35, 36) and has numerous functions in the GI tract, including stimulation of pancreatic secretion (37), inhibition of gastric and small bowel motility (38), facilitation of fatty acid translocation from the intestinal lumen (39), and growth stimulation of various GI tissues (34, 40, 41, 42, 43).
NT is produced from a single precursor, preproneurotensin, that contains both NT 1–13 and the related peptide, neuromedin N. Neuromedin N comprises the C-terminal portion of the proneurotensin peptide after NT is removed. NT 1–13, the major product of proneurotensin in the ileum and brain (44), is cleaved at or soon after secretion at the dibasic arg8-arg9 residues, resulting in the inactive degradation product NT 1–8 and the biologically active NT 9–13. Substance P, muscarinic agonists (carbachol), catecholamines, and BBS/GRP stimulate NT release, suggesting that a complex interplay of neural, endocrine, and luminal mechanisms is responsible for NT release and physiological functioning (45, 46, 47).
5. PYY.
PYY, a 36-amino-acid peptide, is homologous to two other regulatory peptides, pancreatic polypeptide (PP) and neuropeptide Y (NPY) (3). Both the rat and human PYY genes have been isolated and characterized (48, 49). The conserved structural organization of the genes encoding PYY, NPY, and PP suggests that each gene is derived from duplication of a common ancestral gene. The biological actions of PYY include inhibition of pancreatic bicarbonate secretion and contraction of the gallbladder (50). In addition, PYY inhibits gastric emptying and intestinal transit and has been postulated as an agent that contributes to the ileal brake phenomenon, a negative-feedback mechanism that promotes intestinal absorption (3). PYY can also stimulate growth of the GI mucosa (51).
PYY, NPY, and PP are synthesized as prepropeptides consisting of a signal peptide followed by the 36-residue active peptide, a cleavage-sequence Gly-Lys-Arg, and a carboxy-terminal flanking peptide (52, 53). During processing, the precursor is cleaved at the carboxy terminal by a specific prohormone convertase. The Lys-Arg sequences are then lysed by a carboxypeptidase, and the carboxy-terminal tyrosine is amidated. Typical enteroendocrine cells, with long basal processes, are visualized by immunohistochemistry in the mucosa of the ileum, colon, and rectum (54).
6. GLP.
GLPs are a group of peptides known as the enteroglucagons. The enteroglucagons are products of the same gene that produces glucagon in the pancreatic -cell (3). The intestinal L cell produces two elongated glucagons, glicentin and oxyntomodulin, and both of the enteroglucagons, GLP-1 and GLP-2 (55). Both GLP-1 and GLP-2 have effects on nutrient absorption and GI tract physiology; however, these two peptides have distinct roles. GLP-1 has a significant effect on blood glucose levels, lowering blood glucose levels via stimulation of insulin secretion (55), thus suggesting that GLP-1 may provide some therapeutic benefit to patients with diabetes. GLP-2 displays minimal effects on glucose levels, but demonstrates potent trophic effects on intestinal epithelia (55).
Both GLP-1 and GLP-2 are released when the L cell is exposed luminally to the products of a mixed meal (carbohydrate or fat) (55). L cells are most abundant in the mucosa of the ileum and colon; they are the second most numerous population of endocrine cells in the human intestine, after enterochromaffin cells (56). Both GLP-1 and GLP-2 are rapidly cleaved by the exopeptidase dipeptidyl peptidase IV (55). Current concepts of L cell regulation involve integration of hormonal messages from peptides, such as GRP and glucose-dependent insulinotropic polypeptide, and neuronal control (55).
7. Somatostatin.
Somatostatin was isolated and characterized from ovine hypothalamic tissue during a search for a GH-releasing factor (57). Since the identification and purification of somatostation-14, precursor forms of greater molecular weight, including somatostatin-28, with somatostatin-14 making up the C terminus, and larger precursor forms of 120 or more amino acids have been identified (58). All of these peptides exert biological activity but differ in their relative potency.
Somatostatin has been detected in the nerves and cell bodies of the central and peripheral nervous systems, including the autonomic nervous system of the GI tract and the endocrine-like D cells of the pancreatic islets and mucosa of the stomach and intestine (3). More than 90% of the somatostatin immunoreactivity in the human gut is located within the mucosal endocrine D cells (58). In addition, somatostatin is located in the nerves of the myenteric plexus. Somatostatin in the pancreas is located in the D cells at the periphery of the islets closely associated with the -cells (59).
Somatostatin is a regulatory-inhibitory peptide, which, in contrast to BBS/GRP, may be considered as the universal endocrine off-switch. Somatostatin inhibits the release of GH and somatomedin C and all known GI hormones (3). Somatostatin also inhibits gastric acid secretion and motility, intestinal absorption, and pancreatic bicarbonate and enzyme secretion, and selectively decreases splanchnic and portal blood flow (60). In addition, somatostatin can inhibit the growth of normal and neoplastic tissues (61, 62, 63, 64, 65, 66, 67).
B. GI hormone receptors and signal transduction pathways
1. Receptors.
GI hormone-stimulated signal transduction occurs with the binding of hormones to their cognate cell surface receptor, the G protein-coupled receptor (GPCR) (68). The GI hormone-GPCRs have the typical structural features of G protein binding seven-transmembrane receptors (Fig. 1
). The receptors for gastrin, CCK, BBS/GRP, NT, PYY, GLP-2, and somatostatin, which are respectively, the gastrin/CCK-B receptor, CCK-A receptor, GRP receptor, the NT receptor (NTR), NPY receptor, GLP-2 receptor, and the somatostatin receptor (five subtypes), are all GPCRs (3, 55). GPCRs regulate a number of physiological processes, including proliferation, growth, and development (68). It was originally thought that, in order for GPCR signaling to occur, specific interactions between the GI hormone and the receptor were necessary to produce conformational changes in the receptor and stimulate intracellular signal transduction networks. However, recent studies suggest a more complex regulation of the GPCRs through: 1) dimerization with themselves and other receptors; 2) activation of differing G proteins; 3) internalization and desensitization; and 4) ability to change in conformation and interactions with empty, or inactive, receptors (69). It is suggested that this complicated mechanism of regulation allows peptides to interact with GPCRs to stimulate diverse intracellular signaling pathways and ultimately affect multiple physiological functions, depending on cell type.
|
-helical domains function as ligand-regulated guanine nucleotide exchange factors for the intracellular heterotrimeric G proteins (68). Heterotrimeric G proteins are composed of the products of three gene families encoding -, ß-, and 
-subunits (68). The agonist-activated GPCR catalyzes the exchange of GTP for GDP bound to the G-subunit, as well as the dissociation of the GTP-G from its cognate Gß dimer (Fig. 2A) (68). The activated GTP-G- and Gß-subunits, in turn, regulate the activity of various intracellular effector proteins such as phospholipases, adenylyl cyclases, protein kinases, membrane ion channels, and members of the Ras family of GTP-binding proteins (68). In addition, based on structural similarities, the 20 identified G-subunits have been divided into four subfamilies and assigned an effector pathway based on current evidence. The four are: 1) the cholera toxin-sensitive () subunits that stimulate adenyl cyclase and increase cAMP levels; 2) the pertussis toxin-sensitive (i/o) subunits that inhibit adenylyl cyclase activity; 3) the pertussis toxin-insensitive (q/11/14) subunits that stimulate membrane phospholipases; and 4) the 12/13 subfamily that links GPCR to the Ras-related GTP-binding protein, Rho (68). Additionally, 12 G- and 6-Gß subunits have been identified; these ß-dimers have been linked to the signaling molecules phosphatidylinositol 3-kinase (PI3K) and select forms of adenylyl cyclase and receptor kinases (68).
|
) (72). The mechanism by which this occurs involves a complex interplay of several known nonreceptor kinases and receptor kinases. The ability of tyrosine kinase inhibitors to reduce the activation of MAPK by GPCR (73) and the rapid tyrosine phosphorylation of Shc (Src homology and collagen) after GPCR stimulation with the consequent formation of Shc-Grb2 (growth factor receptor-bound 2) complexes (74) provide evidence that tyrosine kinases link GPCRs to the Ras-MAPK pathway.
Additionally, GPCRs link to the Jun N-terminal kinase (JNK), p38 MAPK, and the big mitogen-activated kinase 1 or ERK5 pathways (Fig. 2C) (68). JNK, also termed stress-activated protein kinase (SAPK), is structurally related to MAPK, but the pathways used by GPCRs to activate these kinases are different. Activated Rac and Cdc42 affect JNK activation through stimulation from free ß-dimers and G12 and G13.
Four p38 MAPKs have been described, p38 (CSBP-1), p38 ß, p38 (ERK6 or SAPK3), and p38
(SAPK4) (75). Yamauchi et al. (76) demonstrated that Gq and ß dimers activate p38. Two nonreceptor tyrosine kinases, Btk and Src (77, 78), have been associated with this process. ERK5 can be activated by oxidative stress and plays a role in early gene expression (79). GPCRs can potently stimulate ERK5 through a mechanism that involves Gq and G13, independent of Rho, Rac1, and Cdc42 (80, 81). Furthermore, ERK5 regulates early gene expression through the phosphorylation of the transcription factor, myocyte enhancer factor 2 (80).
The molecular mechanisms through which GPCRs transduce signals are complex and likely involve multiple signal pathways. In addition, the signaling pathways are likely cell-specific, which may explain the diverse physiological functions controlled by GI hormones ranging from regulation of secretion, mobility, and, in some instances, growth depending upon the target tissue.
|
III. Proliferation and Repair of Nonneoplastic Tissues by GI Hormones |
|---|
|
TopAbstractI. IntroductionII. GI HormonesIII. Proliferation and Repair...IV. Proliferation of Neoplastic...V. Future Perspectives and...References |
|---|
A. Gastrin
Gastrin is the GI hormone that has been best characterized for its trophic effects. In addition to stimulating acid secretion from gastric parietal cells, gastrin is the single most important trophic hormone of the stomach (82).
1. Stomach.
The trophic effect of gastrin was initially described more than 30 yr ago in two separate reports. Johnson et al. (83) and Crean et al. (84) demonstrated that pentagastrin, a synthetic gastrin analog containing the active carboxyl-terminal tetrapeptide, increased protein synthesis and parietal cell mass in rats. These results were further confirmed using the natural amidated gastrins, G-17 and G-34. G-17 and G-34 produced maximal stimulation of DNA synthesis in the oxyntic mucosa, duodenum, and colon at doses of 13.5 and 6.75 nmol/kg, respectively.
In the stomach, the oxyntic, acid-secreting mucosa and enterochromaffin-like cells (cells that produce histamines critical to parietal cell acid production) are particularly sensitive to the trophic actions of gastrin (85). The removal of endogenous gastrin, by antral resection, results in mucosal atrophy that can be prevented by administering exogenous gastrin.
Overexpression of either unprocessed gastrins or the amidated gastrins (G-17 and G-34) in transgenic mice results in a 2-fold elevation in serum-amidated gastrin and produces marked thickening of the oxyntic mucosa with increased bromodeoxyuridine (BrdU) labeling, representing an 85% increase in cells undergoing proliferation (86, 87). These findings are further supported by results in athymic nude mice bearing xenografts of a transplanted human gastrinoma (PT) demonstrating gastric and duodenal mucosal hyperplasia (Fig. 3) (Ref.88).
|
2. Small intestine.
The role of gastrin as a stimulant for small bowel mucosal growth is less clear. Johnson et al. (83) showed that [3H]thymidine incorporation and DNA synthesis of rat duodenal tissues were significantly increased by pentagastrin, G-17, and G-34 administration. Schwartz and Storozuk (91) found that exogenous gastrin (13.5 nmol/kg·d) enhanced growth of jejunoileal segments obtained from 19- to 20-d gestation fetal rats transplanted sc in adult rats. In addition, administration of gastrin enhanced the absorption of galactose and glycine, suggesting that gastrin increases functional activity as well as growth. Subsequent studies, however, using both pentagastrin and natural G-17 have failed to demonstrate significant trophic effects of gastrin in the small bowel (jejunum and ileum) mucosa of the adult rat as measured by vincristine-arrested metaphase (92). An explanation for these differing results is not entirely clear but may pertain to different experimental measures of proliferation. Therefore, although the trophic effects in the oxyntic and duodenal mucosa are well established, the trophic effect of gastrin in the remaining small intestine is currently considered minimal, at best.
3. Colon.
Similar to the small bowel, evidence for a trophic effect of gastrin in the colon is limited. In earlier studies by Johnson (93), administration of natural porcine gastrins stimulated DNA synthesis in the colon. In contrast, hypergastremia, induced through acid blockade by either omeprazole administration (94) or fundectomy (95), failed to produce proliferation of either normal colon or the colon carcinoma cells transplanted from an original 1,2-dimethylhydrazine-induced adenocarcinoma.
As opposed to results using amidated gastrin, recent experiments using G-gly demonstrate colonic proliferation, sparking renewed interest in a role for gastrin precursor products in colonic growth. Koh et al. (96) generated mice that overexpress progastrin truncated at glycine-72 (MTI/G-GLY), which demonstrates elevated serum and mucosal levels of G-gly compared with wild-type mice. MTI/G-GLY mice display a 43% increase in colonic mucosal thickness and a 41% increase in the percentage of goblet cells per crypt. Furthermore, administration of G-gly to gastrin-deficient mice resulted in a 10% increase in colonic mucosal thickness and an 81% increase in colonic proliferation (as measured by BrdU) when compared with control mice. Thus, these data, using exogenous administration and genetic mouse models, demonstrate that G-gly is a potent stimulator of colonic proliferation.
4. Pancreas.
The role of gastrin in normal pancreatic growth is also believed to be stimulatory. In two separate studies, investigators in our laboratory have demonstrated small, albeit significant, increases in pancreatic growth. In the first study, young adult rats were given pentagastrin, NT, or BBS in combination with an elemental diet (ED) (33). Pentagastrin at a maximal dose, 250 µg/kg, increased pancreatic weight compared with control rats. In the second study, pentagastrin (100 µg/kg) was given to young (3-month-old), adult (12-month-old), and aged (24-month-old) rats (97). In young rats, pentagastrin increased pancreatic weight, DNA, RNA, and protein content. In adult rats, however, pentagastrin only produced increases in RNA content, whereas aged rats showed no trophic effect. These results suggest that the effects of gastrin on pancreatic growth are dependent upon age. Further support for a stimulatory effect of gastrin on pancreatic cells is provided by in vitro studies using the rat pancreatic acinar cancer cell line, AR42J, demonstrating that gastrins can induce cellular proliferation through stimulation of the c-fos transcription factor via protein kinase C-dependent and independent mechanisms (98).
In contrast, Chen et al. (99) induced hypergastrinemia in male Sprague-Dawley rats by continuous infusion of human Leu 15-gastrin-17, by fundectomy, or by treatment with omeprazole. Gastrin infusion or omeprazole treatment did not affect pancreatic weight and DNA content, whereas fundectomy increased pancreatic weight and DNA content. These findings suggest that endogenous gastrin level variations do not directly play a role in pancreatic growth. Further studies are needed to conclusively establish the differences in gastrin-related hormonal signaling affecting the growth of the pancreas.
B. CCK
CCK, acting through CCK-A receptors, is one of the most potent secretagogues regulating pancreatic acinar cells (100). Additionally, CCK stimulates the growth of the pancreas in experimental studies (101, 102). However, evidence for CCK playing a major role in the growth of the GI mucosa (stomach, small bowel, and colon) is limited.
1. GI mucosa.
Administration of CCK and secretin prevented atrophy in the jejunum and ileum of dogs given total parenteral nutrition (TPN) as a sole nutrient source (103). In addition, administration of these peptides increased galactose absorption, suggesting that CCK is a positive enterotrophic factor for the gut. Weser et al. (104) provided additional evidence that CCK alone, or in combination with secretin, could prevent TPN-associated jejunal and ileal atrophy in rats. In subsequent studies, Fine et al. (105), using intestinal bypass models, demonstrated that the trophic response noted by CCK and secretin in the small bowel was the indirect result of increased pancreatobiliary secretion, as opposed to a direct stimulatory effect of these peptides on the gut mucosa. Confirmation of the lack of a direct effect of CCK on small bowel growth was provided by Stange et al. (106) using cultured rabbit jejunum and ileum preparations.
2. Pancreas.
In contrast, the trophic effects of CCK on the pancreas have been demonstrated by a number of experimental models (107, 108, 109, 110, 111, 112, 113). In one study, the effects of camostate (400 mg/kg), a potent inhibitor of trypsin (which blocks serine proteases), were compared with the effects of chronic exogenous administration of CCK-8 with or without administration of the CCK-receptor antagonist, CR 1409 (114). Chronic (10-d) camostate feeding increased pancreatic weight, protein, and DNA content, which was associated with increased CCK plasma levels. The administration of exogenous CCK produced similar increases in pancreatic growth. The combination of camostate and CCK-8 produced an additive stimulatory effect on the pancreas. The CCK-receptor antagonist, CR 1409, completely abolished the trophic effects of exogenous CCK-8 and inhibited the effects of chronic camostate feeding. Furthermore, CR 1409 alone decreased pancreatic weight, DNA, and protein content of rats. Additionally, other studies have substantiated and confirmed a stimulatory effect of CCK on the pancreas (110, 111, 112, 113). Taken together, these studies provide clear evidence that CCK is a potent stimulant for pancreatic growth.
Recently, the cellular mechanisms regulating the proliferative effect of CCK have been examined. Similar to gastrin, CCK activates the MAPK cascade, leading to the activation of ERK, JNK, and p38 MAPK in the pancreas (115). In addition, there is evidence that other signaling pathways, such as the PI3K-mTOR (mammalian target of rapamycin)-p70s6k pathway, are involved in CCK-stimulated mitogenesis and cellular proliferation (115). p70s6k is physiologically important for phosphorylating the small ribosomal subunit protein S6, which was the first regulated phosphoprotein identified in pancreatic acinar cells (115), thereby facilitating the synthesis of RNA. Bragado et al. (116) demonstrated the phosphorylation and activation of p70s6k in rat pancreatic acini by CCK, carbachol, and BBS, but not by cAMP, phorbol ester, or calcium ionophore. This activation was blocked by rapamycin, which inhibits mTOR, and by wortmannin (a PI3K inhibitor). In addition, the PI3K-mTOR pathway is known to play a role in activating protein synthesis by phosphorylating the binding protein eIF4E, the translation initiation factor that binds to the 7-methyl guanosine cap at the 5' end of most eukaryotic mRNA molecules (117). This binding protein, known as eIF4E-BP or PHAS-I, possesses multiple phosphorylated sites and, when sufficiently phosphorylated, dissociates from eIF4E, which can then interact with eIF4G (a scaffold protein) together with eIF4A (a RNA helicase) to form a complex, known as eIF4F, and stimulate protein synthesis. CCK stimulation of rat pancreatic acini leads to PHAS-I phosphorylation and initiation of a complex formation (118).
C. BBS/GRP
BBS/GRP stimulates pancreatic, gastric, and intestinal secretion, and gut motility, smooth muscle contraction, and release of all gut hormones (29). In addition, BBS/GRP is a potent trophic factor for the GI tract and pancreas.
1. Stomach.
Lehy and colleagues (119, 120) reported that BBS, given orally or sc, stimulated the growth of gut mucosa and pancreas in neonatal rats. In this study, 7-d-old rats were injected sc with BBS (20 µg/kg) twice daily for 6 d. Gastric weight, fundic and antral mucosal height, and the density of parietal cells were increased in BBS-treated pups compared with saline-treated controls.
Dembinski et al. (121, 122) demonstrated that BBS, administered to rats for 7 successive days, significantly increased the weight and RNA and DNA contents of the oxyntic mucosa of the stomach and the duodenal mucosa; somatostatin attenuated the proliferative effect of BBS. Antrectomy, which removes the gastrin-secreting cells, and the CCK receptor inhibitor, L-364,718, partly reduced but did not abolish the proliferative effects of BSS, suggesting that the growth-enhancing mechanisms of BBS involved gastrin and CCK stimulation. More importantly, these studies confirmed that BBS had direct effects on the gastroduodenal mucosa.
Later, Dembinski et al. (122) extended their initial findings by assessing the effects of the BBS/GRP receptor antagonist, RC-3095, on BBS-mediated growth. BBS (10 µg/kg) administered three times daily for 2 d in fasted rats significantly increased the rate of DNA synthesis in the gastroduodenal mucosa and pancreas as measured by the incorporation of [3H]thymidine; this proliferative effect was abolished by RC-3095. These results provide further evidence that BBS/GRP stimulates the growth of the stomach and duodenum and that these effects are due predominantly to a direct effect of BBS/GRP stimulation.
2. Small intestine.
BBS also stimulates growth of the remainder of the small bowel mucosa. We examined the effects of BBS in the prevention of gut mucosal atrophy in Sprague-Dawley rats given liquid ED (33). Four groups of rats were given an ED and injected with saline (control), pentagastrin (250 µg/kg), NT (300 µg/kg), or BBS (10 µg/kg) sc every 8 h. A fifth group was fed a regular chow diet. Atrophy of the ileal mucosa was apparent on d 6 and 11, and atrophy of the jejunal mucosa was present by d 11. BBS prevented jejunal mucosal atrophy and significantly increased ileal mucosal growth (mucosal weight, RNA, DNA, protein content) compared with control.
Chu et al. (123) determined whether the trophic actions of BBS on the small bowel were mediated by nonluminal or luminal (i.e., pancreaticobiliary secretion) factors by construction of isolated small bowel loops [i.e., Thiry-Vella fistulas (TVFs)]. BBS increased mucosal weight, DNA, and protein content in both jejunal and ileal TVF compared with control animals, suggesting that BBS-mediated stimulation of small bowel mucosal growth is mediated by factors that are independent of luminal contents and pancreaticobiliary secretion.
In addition to its effects on gut mucosal growth, BBS exhibits protective effects in the gut after injury. Using a lethal enterocolitis model in rats induced by the chemotherapeutic agent methotrexate (MTX), BBS enhanced gut mucosal growth and significantly inhibited mortality in rats (124). The beneficial effect of BBS on survival was noted when BBS was given before or at the same time as MTX, which suggested that BBS may act through additional mechanisms other than gut mucosal growth alone. One possibility is that BBS may produce its beneficial effects through enhancement of the immune system, which is a known action of BBS (125).
3. Colon.
Studies assessing the effects of BBS/GRP on the colonic mucosa are limited but, similar to the small bowel, appear to support a trophic effect of this agent in the colon. Johnson and Guthrie (126) demonstrated that BBS (20 µg/kg, three times daily for 7 d) stimulated colonic mucosal growth. Puccio and Lehy (119) found that administration of BBS orally in the neonatal period stimulated colonic growth. In contrast, we have not detected a proliferative effect of BBS in the colon of chow-fed rats after 7 or 14 d (127). In rats given an ED, BBS produced a proliferative effect only in the proximal colon. Therefore, although BBS appears to exert a trophic effect on the colonic mucosa, the effects are less pronounced than in the stomach or small bowel.
4. Pancreas.
A number of studies have shown that BBS stimulates the growth of the pancreas (97, 119, 120, 128). For example, Lehy and colleagues (119, 120) demonstrated a trophic response for BBS in the pancreas as well as other GI tissues. In addition, electron morphometric analysis indicated that the increase in pancreatic weight was due to hypertrophy of the acinar cells. Increases in pancreatic chymotrypsin and trypsinogen content were noted, with little effect on lipase or colipase and amylase levels. Similar to its effects on the pancreatic acini, BBS stimulates endocrine cells of the pancreas (129).
Liehr et al. (130) assessed the role of CCK in BBS-induced growth in rats. Sprague-Dawley rats received sc injections of BBS every 8 h for 5 d, alone or in combination with the CCK receptor antagonist, L-364,718. BBS produced a dose-dependent increase in pancreatic weight, DNA and protein content, as well as amylase and chymotrypsinogen levels. L-364,718 significantly inhibited pancreatic growth induced by a high concentration (5 nmol/kg) of BBS but had minimal effects on growth induced by lower dosages of BBS. These results suggest that low doses of BBS stimulate pancreatic growth in a direct manner, independent of CCK, whereas high doses of BBS act in part through CCK release.
More recently, Fiorucci et al. (131) found that chronic BBS administration stimulated pancreatic regeneration after pancreatectomy in pigs. Furthermore, these studies examined the cellular processes that may be regulated by BBS. Three groups of pigs underwent sham operation, subtotal distal pancreatectomy, or subtotal pancreatectomy combined with BBS for 4 wk. After treatment, the pancreas was removed, weighed, and assayed for p42 and p44 MAPK, p46Shc, p52Shc, p66Shc, and Grb2. BBS administration resulted in a 100% increase in residual pancreatic tissue when compared with control animals and approximately a 3-fold increase in the rate of pancreatic acinar cell proliferation. Also, BBS administration significantly increased p46Shc/p52Shc and MAPK expression and/or activity in whole pancreas extracts. These results further demonstrate a proliferative effect of BBS in a large animal model and provide evidence that BBS can stimulate cell proliferation via the MAPK pathway.
Upp et al. (128) demonstrated that BBS had both direct and indirect actions in the pancreas. Polyamine synthesis, an essential part of DNA synthesis, was assessed with BBS administration over a time course. BBS produced significant pancreatic hyperplasia (increased pancreatic weight, protein, and DNA content) after 14 d of treatment. CR1409, the CCK receptor antagonist, inhibited only BBS-mediated increases in DNA content. BBS stimulated polyamine biosynthesis as early as 2 h after administration; CR1409 did not inhibit this increase in polyamines. These results suggested that the trophic actions of BBS are both direct and indirect and that the direct effects are mediated by polyamine synthesis.
D. NT
The physiological functions of NT in the GI tract include stimulation of pancreatic and biliary secretions (37) and inhibition of small bowel and gastric motility (38). In addition, NT stimulates growth of the gastric antrum, small bowel (33, 34, 42), colon (132), and pancreas (40, 133).
1. Stomach.
Feurle et al. (40) demonstrated that sc administration of high dosages of NT for 2 wk increased protein concentration and thickness of the gastric antrum in Wistar rats, whereas DNA content and weight were not affected. In contrast, Hoang et al. (134) noted that NT alone had no effect on the oxyntic gland area or the antrum, but it inhibited increases in antral weight, DNA, and protein induced by secretin. The effects of NT on antral growth are, therefore, minimal at best and may occur with prolonged, high dosages of NT.
2. Small bowel.
In contrast to the stomach, the trophic effects of NT in the small bowel are more pronounced. Wood et al. (42) first noted that NT stimulated the small bowel mucosa of rats fed a normal chow diet. We have shown that administration of NT prevents gut mucosal atrophy induced by feeding rats an ED (Fig. 4) (Ref.34) and stimulates mucosal growth in defunctionalized self-emptying jejunoileal loops or isolated TVFs, thus supporting a direct role for NT in the stimulation of gut mucosal growth. These findings demonstrate that NT is an important trophic hormone that can maintain and even augment gut mucosal structure by an increase in overall mucosal cellularity. Consistent with our findings, Vagianos et al. (135) reported that NT restores gut mucosal integrity in rats and prevents the translocation of indigenous bacteria after radiation-induced mucosal injury. We have extended the observation that NT increases proliferation of small bowel mucosa by evaluating the effects of NT in rats of different age groups and found that increases of growth measurements (DNA, RNA, and protein) were the greatest in the small bowel mucosa of aged rats. These findings demonstrate that, with aging, the ability of the small bowel mucosa to respond to the trophic stimulus of NT is retained and may, in fact, be greater when compared with the proliferative response in young rats given NT.
|
3. Colon.
We have shown that NT (300 µg/kg three times daily) stimulates proliferation of colonic mucosa in young and aged rats in a differential fashion (41). Colonic proliferation in young (2-month-old) rats was characterized by increases in cell number, whereas in aged (2-yr-old) rats, indices of hypertrophy were elevated significantly without a concomitant increase in cell number. This study shows that NT is not only important for small bowel growth and mucosal integrity, but also functions as an important hormone regulating colonic mucosal proliferation. Furthermore, Hoang et al. (134) demonstrated that NT (100 µg/kg) significantly increased protein content in the colon but did not affect weight or DNA content in F344 rats. The decreased effect noted in their study relates to the decreased dosage of NT compared with our previous study. It is clear that, although not as pronounced as the small bowel, NT can stimulate proliferation of colonic mucosa.
4. Pancreas.
Feurle et al. (40) demonstrated that administration of NT increased pancreatic weight, DNA, RNA, and protein contents as well as lipase and amylase concentrations. A proliferative effect for NT on the pancreas was likewise demonstrated by Wood et al. (133) by chronic administration of NT, which produced small but statistically significant trophic effects in the pancreas, with the highest dose of NT (300 µg/kg) increasing pancreatic weight by 16%, DNA content by 12%, and protein content by 17%. When NT was given in combination with either cerulein or secretin, it had minimal additive effects on the responses of these two peptides (134). Therefore, from these studies it can be concluded that NT stimulates pancreatic growth but not to the same extent as CCK or BBS/GRP.
E. PYY
PYY inhibits blood flow to the GI tract (140, 141), inhibits small bowel fluid and electrolyte secretion (142, 143, 144), and regulates GI motility. In addition, administration of PYY can stimulate proliferation of small bowel and colonic mucosa. The effects of PYY on the stomach and pancreas are not known.
1. Small bowel and colon.
Gomez et al. (51) first demonstrated a trophic effect for PYY (150–200 nmol/kg, three times daily) on the small bowel and colonic mucosa of both rat and mouse. Consistent with these findings, Chance et al. (145) found that PYY treatment (1 nmol/kg·h) in Sprague-Dawley rats given TPN produced significant increases in jejunal, ileal, and colonic protein contents.
The trophic effects noted for PYY were in contrast to work by Guan et al. (146) demonstrating that PYY, administered at 400 pmol/kg·h (continuous infusion) for 4 d, failed to affect duodenal weight, protein, or DNA/RNA contents in male Wistar rats. In addition, using either a small bowel resection model (147) or a gut atrophy model induced by TPN (148), administration of PYY failed to elicit a trophic effect on small bowel mucosa. These negative results were attributable to the smaller doses and different experimental designs.
F. GLP-2
A trophic effect for proglucagon-derived peptides (PGDPs) on the intestinal mucosa has been postulated since the description of a glucagon-secreting tumor of the kidney associated with small bowel mucosal hypertrophy. Drucker et al. (149) were the first to demonstrate that the intestinotrophic factor was GLP-2.
1. Small intestine.
To assess the potential consequences of increased proglucagon gene expression on intestinal growth, Brubaker et al. (150) created glucagon-SV40 T-antigen transgenic mice that developed proglucagon-producing tumors and elevated levels of the PGDPs. Small bowel mucosal hypertrophy was noted in the transgenic mice; however, it was not possible to ascertain which of the PGDPs was responsible for small bowel growth. To better ascertain which of the PGDPs stimulated gut growth, nude mice with three different sc proglucagon-producing tumor xenografts (InR1-G9, RIN1056A, and STC-1 cells) were analyzed (149). Furthermore, CD1 mice were treated with GLP-1, GLP-2, glicentin, and intervening peptide (IP)-1 (all are produced from the proglucagon gene). Mice with all three xenograft types demonstrated an increase in small bowel weight, crypt-villus height, and percentage of BrdU-positive cells; glicentin (43.75 µg) administered twice daily for 10 d produced modest increases in total small bowel weight. Neither GLP-1 nor IP-1 had an effect on the gut mucosa, whereas GLP-2 produced a 50% increase in small bowel weight. Furthermore, only GLP-2 produced a significant increase in mucosal thickness, with the increase attributable to increased villus height (149). Similarly, Ghatei et al. (151) demonstrated prominent trophic effects of GLP-2 in Wistar rats. Using varying doses of enteroglucagon, GLP-1, oxyntomodulin, and GLP-2, only GLP-2 produced a dose-dependent increase in the weight of stomach, small intestine, and colon of parenterally fed rats. Therefore, these studies confirmed that, of the PGDPs, GLP-2 was the major mediator of intestinal epithelial proliferation.
Litvak et al. (152, 153) demonstrated that GLP-2 (1.75 µg/kg twice a day) significantly increased the weight of jejunum, ileum, and colon of athymic nude mice compared with both control mice and mice treated with NT (600 µg/kg three times a day). These studies confirmed the previous findings of a trophic effect for GLP-2 and demonstrated that the proliferative effects were equal to or greater than those of NT. In another study, GLP-2-producing tumor cells (STC-1 cells) were implanted in athymic mice and then randomized to receive either NT or saline (152, 153). The mice treated with NT and GLP-2 (from STC-1) displayed significant increases in jejunal and ileal weight and protein content over either NT or GLP-2 alone. The administration of NT appeared to enhance the enterotrophic effects of GLP-2, suggesting that the combination may be useful to enhance intestinal growth in patients with short bowel syndrome.
In addition to the effects of GLP-2 on normal mucosa, the effects of this agent during periods of gut injury or atrophy have also been assessed. Mice treated with the nonsteroidal agent indomethacin develop small bowel enteritis associated with significant mortality at 48–72 h after administration; treatment with human [Gly2]GLP-2 before, during, or after indomethacin administration resulted in reduced mortality and decreased mucosal injury (154). The protective effects were attributed to significantly increased crypt cell proliferation and decreased crypt compartment apoptosis. The effect of GLP-2 on chemotherapy-induced intestinal mucositis has also been assessed. Pretreatment of mice with human [Gly2]GLP-2 before administration of the topoisomerase inhibitor irinotecan resulted in reduced bacterial infection, intestinal damage, and mortality (155). Histological and biochemical analyses revealed significant reductions in crypt compartment apoptosis and reduced caspase-8 activation. Consistent with these reports, Tavakkolizadeh et al. (156) noted decreased intestinal damage in rats given GLP-2 in combination with the chemotherapeutic agent 5-fluorouracil. Finally, repeated cyclical administration of both human [Gly2]GLP-2 and irinotecan resulted in decreased mortality in groups of BALB/C mice implanted with sc CT-26 colon carcinomas. These data suggest that GLP-2 may have therapeutic advantages for cancer patients by supporting bowel integrity without diminishing the effectiveness of the chemotherapy.
2. Colon.
Although the small bowel is significantly more sensitive to the effects of GLP-2, studies have shown that GLP-2 and GLP-2 analogs can stimulate the growth of colonic mucosa. As previously noted, Litvak et al. (152) demonstrated a trophic effect of GLP-2 on the colonic mucosa of athymic nude mice. Drucker et al. (157) demonstrated an increase in colonic growth using dipeptidyl peptidase IV-resistant GLP-2 analog, human [Gly2]GLP-2 in CD1 mice. Furthermore, the combination of this agent with IGF-I, GH, or long [Arg3]IGF-I produced a greater increase in large bowel mass than mice treated with [Gly2]GLP-2 alone.
Treatment with GLP-2 has also been shown to reduce colonic mucosal injury, similar to the findings in the small bowel (51). In a dextran sulfate-induced colitis model, concomitant administration of sc human [Gly2]GLP-2 and oral dextran sulfate for 10 d resulted in markedly reduced colonic damage and decreased weight loss in CD1 and BALB/C mice. Treatment with [Gly2]GLP-2 preserved colonic length, decreased intestinal histological damage, reduced IL-1 expression, and increased crypt cell proliferation.
G. Somatostatin
Somatostatin has predominantly inhibitory effects, which have been used to advantage in certain clinical scenarios. Somatostatin inhibits pancreatic and GI secretion and GI motility and inhibits the release of GH and all known GI hormones (3). The active peptide analog of somatostatin (octreotide) has been effectively used as an agent to ameliorate symptoms associated with endocrine-overproducing tumors, pancreatic fistulas, and enterocutaneous fistulas (158, 159). In experimental studies, somatostatin has also been shown to inhibit the growth of the GI mucosa and normal pancreas (158, 159). These effects may be through either an indirect mechanism such as inhibition of other trophic hormones or a direct effect through interaction with the somatostatin receptor subtype 2 (160, 161). This somatostatin receptor subtype associates and stimulates tyrosine phosphatase src homology 2-containing tryosine phosphatase 1 activity, which in turn arrests cells in the Go/G1 phase of the cell cycle associated with up-regulation of the cyclin-dependent kinase inhibitor p27kip1 and an increase in hypophosphorylated retinoblastoma protein levels (160, 161). Although the major portion of this review has focused on the stimulatory effects of various GI hormones, it is important to discuss the inhibition of growth using somatostatin and its analogs.
1. Stomach.
The effects of somatostatin on gastric mucosal growth have been assessed. Treatment of Wistar rats with somatostatin (50 µg/kg·h) reduced nuclear uptake of [3H]thymidine and cell division in both fundic and antral progenitor cells (162). Furthermore, the combination of somatostatin and gastrin reduced gastrin-stimulated DNA synthesis in the gastric mucosa. In contrast, in the duodenum and jejunum, the effects of somatostatin were less consistent, with nocturnal somatostatin producing slight decreases in DNA synthesis. These findings suggested that somatostatin could inhibit cell proliferation in the mucosa of the normal GI tract and, furthermore, could antagonize the trophic activity of gastrin, particularly in the fundus and antrum of the stomach. Similar results were noted by Rivard et al. (61), demonstrating inhibitory effects on rat mucosal growth of the duodenum using the somatostatin analog, Sandostatin. The administration of Sandostatin in this model was associated with a decrease in plasma CCK and IGF-I levels (61), suggesting an induced effect of somatostatin.
2. Small intestine.
There is also evidence to support an inhibitory effect of somatostatin in the jejunum and ileum. Bass et al. (62) demonstrated that the normal adaptive hyperplasia noted in rats after 40% small bowel resection could be attenuated by administration of octreotide, as noted by assessment of villus height and residual bowel weight. Using a rabbit model of an ileal mucosal defect, Thompson et al. (163) demonstrated that octreotide inhibited normal but not EGF-stimulated cell migration. In addition, octreotide decreased EGF-stimulated proliferation. In contrast, Vanderhoof and Kollman (164) failed to show significant differences in mucosal weight, protein, and sucrase levels in rats treated with octreotide after an 80% small bowel resection. The differences noted in this study, compared with that of Bass et al. (62), may simply represent the fact that a more extensive resection model was used, which may have provided an enhanced stimulation of mucosal regeneration that could not be inhibited by administration of somatostatin.
The question remains, however, as to the role endogenous somatostatin plays in the inhibition of normal intestinal growth and proliferation. Parekh et al. (63) demonstrated that somatostatin appears to function as an endogenous inhibitory factor. Male F344 rats were given cysteamine (CYS; an agent known to deplete endogenous somatostatin). The administration of CYS stimulated growth of the proximal and distal small intestine as assessed by increases in weight, DNA, and RNA content; this effect was assumed to be secondary to a depletion of endogenous somatostatin, although other indirect effects cannot be entirely ruled out.
3. Pancreas.
The effects of somatostatin on the normal pancreas have been extensively analyzed. The finding of D cells in the normal pancreas suggested a role for this peptide in not only physiological functions of the pancreas, but also growth of the normal pancreas. Rivard et al. (61) demonstrated that Sandostatin significantly decreased pancreatic weight, DNA, RNA, and protein content. Furthermore, results of Parekh et al. (63) demonstrated that somatostatin depletion with CYS stimulated pancreatic weight by 127% and DNA content by 141% compared with control animals. In addition, CYS augmented the trophic effect of BBS on the pancreas, thus suggesting that somatostatin may serve to limit normal pancreatic growth. Blocking endogenous somatostatin may release these inhibitory constraints and allow for increased proliferation of the normal pancreas. The specific mechanisms, however, as to how somatostatin acts in vivo are still unknown. The effects may be direct, acting through the somatostatin receptor subtype 2 receptor, or indirect through the inhibition of other proliferative agents.
|
IV. Proliferation of Neoplastic Tissues by GI Hormones |
|---|
|
TopAbstractI. IntroductionII. GI HormonesIII. Proliferation and Repair...IV. Proliferation of Neoplastic...V. Future Perspectives and...References |
|---|
In a manner analogous to breast and prostate cancers, the hypothesis was formulated that GI cancers may also possess receptors and likewise be responsive to various GI hormones. In this regard, receptors for various GI hormones have been identified in gastric, pancreatic, and colorectal cancers; these cancers are responsive to the effects of GI hormone treatment (169).
A. Gastric cancer
Although the incidence of gastric cancer has been decreasing in the United States, it is still estimated to be one of the more common cancers worldwide and is endemic in certain areas of the world such as Japan, Eastern Europe, and South America (170). The therapeutic options are limited, because radiation and chemotherapy are, for the most part, ineffective in increasing survival in patients with metastatic disease (170).
Certain gastric cancers possess both low- and high-affinity gastrin/CCK-B receptors (171, 172). The growth of these cancers is stimulated by pharmacological and, in certain instances, physiological concentrations of gastrin or gastrin analogs in vitro and in vivo (173, 174, 175, 176, 177, 178). Sumiyoshi et al. (174) identified gastrin receptors in the human gastric cancer, SC-6-JCK; treatment of this poorly differentiated adenocarcinoma with gastrin resulted in increased tumor weight, size, and labeling index of [3H]thymidine. Gastrin receptor antagonists blocked these trophic effects, demonstrating a direct effect through the gastrin/CCK-B receptor (174, 179). Other therapeutic strategies that decrease gastrin receptor-positive gastric cancer growth include inhibition of gastrin release by agents such as enprostil, a prostaglandin derivative that lowers serum gastrin levels in fasted and fed humans and mice. Administration of enprostil to mice bearing gastrin receptor-positive gastric cancers inhibited the growth of these cancers; however, the growth of gastrin receptor-negative cancers was not affected (180). Therefore, these findings indicate that certain gastric cancers possess gastrin receptors, and growth may be modulated by either blocking the receptors or inhibiting gastrin release.
The glycine-extended G-17 peptide, Gly-G, which is colocalized with amidated gastrin in antral G cells and cosecreted with G-17, was initially thought to be biologically inactive. However, Iwase et al. (181) have shown that Gly-G stimulated the growth of the human gastric cancers, AGS and SIIA. This increased growth was not blocked by either the CCK-B or the CCK-A receptor antagonists, suggesting the presence of a novel receptor that is different from the classic CCK-A or CCK-B receptor. In contrast, G-17 stimulated the growth of both AGS and SIIA cells. Consistent with these findings, Szabo et al. (182) demonstrated stimulation of AGS cell growth with both G-17 and pentagastrin. Pentagastrin was found to be 10 times less potent as a stimulator of AGS proliferation when compared with G-17, suggesting a specific, active CCK-B/gastrin receptor in these cells.
Although the effects of gastrin appear well established in tumor models, the effects of gastrin on carcinogen-induced gastric cancers have produced conflicting results. Tahara and Haizuka (183) noted that gastrin increases the incidence of N-methyl-N-nitro-N-nitroguanidine (MNNG)-induced cancers in rats, whereas Tasuta et al. (184) found that gastrin inhibits the incidence of stomach cancers induced by the same carcinogen. The timing and administration of gastrin appears critical in these two studies and may account for the conflicting results. For example, it was noted that gastrin administered during exposure to MNNG increased gastric cancer development, but late administration did not display the same growth-enhancing effect. Therefore, the effects of gastrin may be more pronounced in the early stages of tumor development rather than in the later stages of growth in which other peptide factors may play a more prominent role. Recent findings by Hagiwara et al. (185) further support a role for gastrin in carcinogen-induced stomach cancers. In this study, carbachol-induced glandular carcinomas in F344 rats were associated with increased endogenous gastrin levels.
In addition to experimental studies, clinical studies further link gastrin to the development and growth of gastric cancers. Recently, gastrin has been linked to Helicobacter pylori infection, a known risk factor for gastric cancer (186). Moreover, significantly higher plasma and luminal gastrin levels have been measured in gastric cancer patients compared with control patients (186). Watson and colleagues (173, 187) evaluated 90 archival samples of atrophic gastritis; intestinal metaplasia; mild, moderate and severe gastric epithelial dysplasia; and intestinal-type gastric adenocarcinoma. Increased expression of gastrin and the CCK-B/gastrin receptor was associated with the progression of atrophic dysplasia to adenocarcinoma (173). Together, these clinical studies provide further support, albeit circumstantial, for the role of gastrin in gastric carcinogenesis.
In addition to gastrin, GRP is thought to contribute to gastric cancer growth. Preston et al. (188) identified high-affinity GRP receptors in a single gastric cancer cell line, St42, and Kim et al. (189) identified expression of GRP receptors in two in vitro gastric cancer cell lines (SIIA and MKN-45) and in three of five human gastric cancer xenografts. The presence of GRP receptors in gastric cancer appears to play a functional role in the growth of these tumors, as noted by Qin et al. (190) who demonstrated that the GRP receptor antagonist, RC-3095, blocked the growth of the human gastric cancer, Hs746T, both in vitro and when placed as xenografts in nude mice. Pinski et al. (191) also found that RC-3095 significantly inhibited MKN-45 xenograft tumor weight and volume. As an assessment of signaling pathways regulating the mitogenic effect of GRP, Kim et al. (189) from our laboratory have shown that treatment of the gastric cancer cell line, SIIA, with BBS stimulates release of intracellular calcium and increases the expression of the activator protein (AP)-1-related proteins, c-Jun and c-Fos, that have been shown to play a role in cellular proliferation; this effect was blocked by specific GRP receptor antagonists. Clinical studies indicate that as many as 50% of gastric cancers display high-affinity binding for GRP (192). Similarly, Carroll et al. (193) identified expression of GRP receptor mRNA in eight of 20 (40%) nonantral gastric cancers. Of these eight, six of the GRP receptors were mutated; some of the mutations resulted in nonfunctional GRP receptors. Survival of patients whose tumors expressed functional GRP receptors was not statistically different from those that did not; however, this may represent the fact that only a small number of patients were analyzed. Future studies, assessing a larger patient population, are required before a definitive assessment can be made regarding the association of GRP receptors with gastric cancers.
Finally, in limited studies, the hormone NT has been shown to contribute to carcinogen-induced gastric cancer. Tatsuta et al. (194) found that prolonged alternate-day administration of NT significantly increased the incidence of MNNG-induced gastric cancer in Wistar rats compared with rats given MNNG alone. The effect of NT on human gastric cancers, however, has not been assessed.
B. Pancreatic cancer
Carcinoma of the pancreas is the fifth leading cause of cancer deaths in the United States (195). The prognosis from this disease remains dismal, with a mean survival time after diagnosis of about 4–6 months. The effects of both stimulatory and inhibitory hormones on pancreatic cancer growth have been assessed, and the signaling mechanisms regulating hormone-induced pancreatic cancer proliferation are currently being analyzed.
Experimental studies have demonstrated a role for CCK and gastrin in pancreatic cancers. CCK receptors have been identified in hamster and human pancreatic cancer cell lines. In vitro studies have demonstrated that CCK and its analog, JMV-180, can stimulate the growth of human pancreatic cancer cell lines, including MIA PaCa-2 (196, 197, 198, 199), PANC-1 (197, 199, 200), BxPC-3 (198, 199), Capan-2 (199), RWP-2 (199), SW-1990 (199, 201, 202), and MD PANC-3 (203). Similarly, the presence of CCK receptors can predict the responsiveness of human pancreatic cancer cells to CCK treatment (180, 204, 205). In addition to effects on pancreatic cancer proliferation, Hirata et al. (206) reported that CCK-8 increases the production of metalloproteinase-9 in pancreatic cancer cells via induction of protein kinase C, suggesting a role for CCK in pancreatic cancer invasiveness and possibly metastasis.
Cerulein, a CCK analog, significantly stimulates the growth of the CCK receptor-positive pancreatic cell line SKI (207); however, cerulein has no effect on the CAV pancreatic cancer cell line, which does not possess CCK receptors. Furthermore, Asperlin, a competitive, nonpeptide CCK receptor antagonist, inhibits growth of SKI xenografts (208). Townsend et al. (209) demonstrated that growth of the hamster pancreatic cancer cell line, H2T, was stimulated by cerulein and that growth could be further enhanced by concomitant treatment with secretin. Both gastrin and Gly-G can stimulate the proliferation of AR42J rat pancreatic tumor cells (210, 211); these growth-promoting effects can be blocked by L-365,260, a CCK-B receptor antagonist (210, 211, 212). Therefore, these studies provide strong support for a modulatory effect of CCK, secretin, and gastrin on certain pancreatic cancers. In contrast, Liehr et al. (213) found that CCK, at dosages of 10–12 to 10–6 M, had no effect on the growth of either MIA PaCa-2 or PANC-1 cells. Moreover, stable transfection of MIA PaCa-2 and PANC-1 cells with CCK-A or CCK-B receptors resulted in growth-inhibitory responses after activation of both receptor subtypes by their ligands. The reason for these discrepant results is not readily apparent.
In vivo studies support a role for CCK in the stimulation of pancreatic cancer growth. Pancreatic tumors in rats can be induced by endogenous CCK using CCK-releasing agents, such as trypsin inhibitors, by biliary diversion or using bile salt-binding drugs (214, 215, 216). For example, feeding rats a soybean diet, a natural trypsin inhibitor, induced the formation of preneoplastic lesions in the pancreas and, when treatment was prolonged for years, pancreatic cancer developed (217, 218, 219).
Evidence exists that CCK can enhance chemical carcinogen-induced pancreatic cancers, as well. When combined with azaserine (a carcinogenic DNA alkylating agent), the long-term administration of exogenous CCK or elevation of endogenous CCK levels (220, 221, 222) enhanced the development or shortened the latency period of the preneoplastic acinar lesions. These effects could be blocked by CCK-A receptor antagonists (223, 224, 225). Consistent with these results, Satake et al. (107) found that administration of cerulein enhanced the carcinogenic effect of N-nitrosobis (2-hydroxypropyl) amine in hamsters. Furthermore, Howatson and Carter (226) found that both CCK and secretin enhanced pancreatic carcinogenesis induced by N-nitrosobis (2-oxypropyl) amine (BOP) in hamsters. In contrast, Johnson et al. (227) reported that CCK-8 inhibited development of nitrosamine-induced pancreatic cancer in hamsters. The differences noted in these studies may relate to differences in the animal models. With the exception of this study, the majority of studies suggest that agents capable of increasing endogenous CCK and stimulating the growth of normal pancreas can also stimulate pancreatic carcinogenesis.
Similar to CCK, BBS/GRP, another pancreatic trophic factor, has been associated with pancreatic carcinogenesis. Douglas et al. (223) demonstrated that rats treated with azaserine (30 mg/kg) and BBS (10 mg/kg) for 16 wk developed preneoplastic lesions at a higher rate than rats treated with azaserine alone. Similar results were noted by Meijers et al. (228) and Lhoste and Longhecker (208). This effect was not mediated solely by induction of CCK because the development of preneoplastic lesions was not blocked by the CCK receptor antagonist CR-1409 (223). In contrast, another study demonstrated that administration of BBS was accompanied by a decrease in the number of preneoplastic lesions in BOP-treated hamsters. The differential effects of BBS in these studies may be attributed to species differences and also to the differences in the carcinogenic agents.
In another study, Szepeshazi et al. (229) demonstrated that administration of the BBS receptor antagonist, RC-3095, decreased BOP-induced pancreatic cancers in hamsters. RC-3095 significantly decreased EGF binding capacity, suggesting that indirect effects might be largely responsible for the reduction identified in BOP-associated hamster cancer. In a subsequent study (230), these investigators showed that the combination of GRP or BBS with RC-3095 was not able to overcome the effects of RC-3095 but, conversely, augmented the inhibitory effects of RC-3095. These data are more consistent with the observations of Meijer and Baak (231) in hamsters and show the complexity of BBS/GRP regulation in pancreatic cancer, which may be highly species-dependent.
More recently, the effects of a potent and specific GRP receptor antagonist, BIM 26226 [[D-F5 Phe 6, D-Ala 11] BBS (6–13) OMe], were evaluated in pancreatic cancers in vivo and in vitro (232). Chronic BIM 26226 administration significantly reduced tumor volume, protein, RNA, amylase, and chymotrypsin content in both GRP-responsive and GRP-unresponsive pancreatic cancers. These findings suggest that GRP receptor antagonists may work through either a direct or indirect effect to inhibit tumor growth. Furthermore, a recent study by Burghardt et al. (233) demonstrated that BBS increases expression of three transcription factors (c-fos, c-myc, and high-mobility group protein IY) that are associated with proliferation in the human pancreatic cell line HPAF. These results provide potential novel mechanisms for targeting the trophic effects of BBS/GRP in pancreatic cancers.
Similar to CCK, NT also stimulates pancreatic cancer proliferation. Our laboratory has shown that the human pancreatic cancer cell line MIA PaCa-2 possesses NTRs and mobilizes intracellular calcium (234). Iwase et al. (181) have shown that the NTR antagonist SR48692 inhibits intracellular calcium mobilization, IP-3 turnover, and MIA PaCa-2 in vitro cell growth induced by NT in a dose-dependent fashion. Furthermore, in vivo experiments demonstrated that NT significantly increased the size, weight, and DNA and protein contents of xenografted MIA PaCa-2 tumors; SR48692 blocked this effect. Recent findings by Reubi et al. (235) have shown that approximately 75% of pancreatic adenocarcinomas express NTRs. Consistent with these reports, Ehlers et al. (236) have shown NTR expression in approximately 90% of resected pancreatic cancers.
The signaling mechanisms regulating NT-mediated proliferation in NTR-positive cancers have been assessed. NT stimulates ERK and JNK activity in MIA PaCa-2 cells and increases AP-1 binding activity (237). Consistent with these results, Ryder et al. (238) reported that NT stimulated Ca2+ mobilization and activated ERK1 and ERK2 and DNA synthesis in the human pancreatic cancer cell line, PANC-1. Recently, Guha et al. (239) reported that NT induced a rapid activation of the PKC isoform, PKD, in PANC-1 cells, which was linked to the mitogenic effect of NT.
The effects of PYY on pancreatic cancer growth have been examined. Treatment of the human pancreatic cancer cell lines, MIA PaCa-2 and PANC-1, with the synthetic analog of PYY, PYY 22–36, inhibited the growth of these cells in vitro (240). A significant decrease in the growth of human pancreatic cancer xenografts was demonstrated in nude mice given PY
