Digestive Diseases and Sciences, Vol. 47, No. 5 (May 2002), pp. 1008 –1014 (© 2002)

Migration of Primary Cultured Rabbit Gastric Epithelial Cells Requires Intact Protein Kinase C and Ca2⫹/Calmodulin Activity TUULA RANTA-KNUUTTILA, MD,* TUULA KIVILUOTO, MD,* HARRI MUSTONEN, MSc,* PAULI PUOLAKKAINEN, MD,* SUMIO WATANABE, MD,† NOBUHIRO SATO, MD,† and EERO KIVILAAKSO, MD*

Superficial gastric mucosal injury is rapidly repaired by epithelial cell migration. This study aims to characterize the intracellular signal transduction pathways underlying the repair process. Primary monolayer cultures of rabbit gastric epithelial cells were wounded. The measured spontaneous cell migration speed at the edge of the wound was 457 ⫾ 89 ␮m/24 hr. Epidermal growth factor stimulated and genistein (receptor tyrosine protein kinase inhibitor) inhibited cell migration significantly. Down-regulation of protein Kinase C (PKC) with long-term phorbol 12-myristate 13-acsetate or inhibition with calphostin-C significantly inhibited cell migration. Blocking of Ca2⫹ channels with verapamil and endogenous Ca2⫹ release with TMB-8 or inhibition of the Ca2⫹/calmodulin complex with calmidazolium likewise significantly inhibited migration speed and also abolished the rise of [Ca2⫹]i, which was measured in migrating cells. Modulation of the cAMP-PKA pathway or prostaglandin synthesis had no influence on cell migration. Gastric epithelial cell migration implies activation of receptor tyrosine kinase. It is associated with increased [Ca2⫹]i and requires an intact Ca2⫹/calmodulin complex. Intact PKC activity also is needed. KEY WORDS: migration; cell culture, gastric epithelial cells; Ca2⫹/calmodulin; protein kinase C.

Damage to the gastric surface epithelium is followed by a rapid repair called restitution. Uninjured epithelial cells migrate from viable adjacent areas such as gastric pits and glands over the denuded basal lamina. Manuscript received February 5, 2001; revised manuscript received July 25, 2001; accepted August 1, 2001. From the *Department of Gastroenterologic Surgery, Helsinki University Central Hospital, Helsinki, Finland; and †Department of Gastroenterology, Juntendo University School of Medicine, Tokyo, Japan. Supported by grants from the Research Foundation of Helsinki University Central Hospital, Sigrid Juselius Foundation, and Academy of Finland Address for reprint requests: Dr. Eero Kivilaakso, Department of Gastroenterologic Surgery, Helsinki University Central Hospital, Haartmaninkatu 4, 00290 Helsinki, Finland.

1008

The migrating cells change from normal columnar to a flattened ameboid shape, forming pseudopodiumlike structures (lamellepodia), which reach to the center of the wound. When the cell-free area is gradually covered with migrated cells, the lamellepodia will disappear. This process is independent of cell proliferation (1–3). The role of epidermal growth factor (EGF) and its receptor in the healing of mucosal lesions has been studied extensively. Gastric ulceration triggers an overexpression of EGF-receptor (EGF-R) in epithelial cells at the ulcer margin (proliferative zone) (4 – 6) and EGF synthesis is started around the ulcer, which probably contributes to cell migration and proDigestive Diseases and Sciences, Vol. 47, No. 5 (May 2002)

0163-2116/02/0500-1008/0 © 2002 Plenum Publishing Corporation

GASTRIC EPITHELIAL CELL MIGRATION

liferation in a healing ulcer. EGF has also been shown to promote epithelial restitution in cell cultures (7–9) and in in vitro rabbit duodenum(10). The early cellular events and signal transduction pathways underlying cell migration have been studied in detail in various cell systems. It has been shown that tyrosine kinase activity associated with EGF-R is significantly elevated within 30 min after gastric mucosal injury, suggesting that activation of this enzyme is important in the initiation of the reparative processes (11). Phospholipase C-␥1 (PLC-␥1) is a substrate for EGF-R tyrosine kinase, and its phosphorylation subsequent to EGF-R tyrosine kinase activation has been demonstrated in the early phase of gastric mucosal injury repair (12, 13). Most studies investigating the cellular events underlying cell migration and epithelial restitution have been performed with cell lines originating from neoplastic or transformed cells. Therefore, the relevance of these findings to normal tissue remains somewhat uncertain. In the present study we have used primary rabbit gastric epithelial cell culture in order to explore the signal transduction pathways underlying the migration process in normal epithelial cells. According to Watanabe et al, hardly any cell proliferation is seen during the first 24 hr, the initial recovery of the wound occurring solely by cell migration (14). MATERIALS AND METHODS Cell Culture Technique. Gastric epithelial cells were isolated and cultured as described by Watanabe et al.(14). Fasting male white rabbits, weighing 2.0 –3.0 kg, were anesthetized and sacrificed with intraperitoneal administration of Nembutal (pentobarbital sodium, 50 mg/kg). The stomach was removed, opened along the greater curvature, and rinsed carefully with ice-cold isotonic saline. The corpus part of the mucosa was quickly separated with a razor blade and minced into small pieces (2–3 mm2). The pieces were incubated in a medium containing 0.07% collagenase (type I, Wako Chemical Inc.), 130 mM NaCl, 12 mM NaHCO3, 3 mM NaH2PO4, 2 mM MgSO4, 1 mM CaCl2, 0.1% bovine serum albumin, and 0.2% glucose for 15 min in a shaker bath at 37°C. After incubation, the minced tissue was washed with Ca2⫹- and Mg2⫹-free Hanks’ balanced salt solution (HBSS) with 1 mM EDTA. These procedures were repeated twice before the tissue was filtered through a metal mesh (pore size 300 ␮m). The cells were then washed in Ca2⫹- and Mg2⫹-free HBSS containing 1 mM EDTA and 0.1% bovine serum albumin. The isolated gastric mucosal cells were inoculated (5 ⫻ 106 cells/dish) on collagen type I coated plastic culture dishes (60 mm, diam. Biocoat Collagen I Cellware, Falcon, Becton Dickinson Labware, Bedford, Massachusetts, USA) and cultured in Coon’s modified Ham’s F-12 medium supplemented with inactivated 10% fetal bovine serum, 100 Digestive Diseases and Sciences, Vol. 47, No. 5 (May 2002)

units/ml penicillin, 100 ␮g/ml streptomycin, and 0.25 ␮g/ml amphotericin. Cells were incubated at 37°C in humidified atmosphere containing 5% CO2–95% O2. Artificial Wounding and Assay of Restoration of Monolayer. The cultured gastric epithelial cells formed a complete monolayer sheet within 48 hr after cell inoculation. As reported earlier and confirmed by us, more than 90% of the cells were periodic acid–Schiff (PAS) -positive mucous cells, 5% were succinic dehydrogenase-positive parietal cells, and 4% were Nile blue-positive chief cells. Five artificial wounds, each with a cell-free area of constant size (3 mm2), were made in an epithelial cell sheet without damaging the coated dish surface using a modified pencil-type mixer with a rotating silicone tip. The test agents were added 10 –15 min before the wounding. After wounding, the cell migration process was monitored using a phase-contrast microscope in serum-free free conditions by taking micrographs at 0, 24, and 48 hr and measuring the cell-free area on the epithelial cell sheet. The results are expressed as migration speed of the cells (microns per/24 hr). Measurement of Intracellular Free Calcium. Measurement of intracellular free calcium concentration [Ca2⫹]i was performed by using a fluorescent Ca2⫹ indicator dye, fura-2, which binds Ca2⫹ tightly. By measuring the ratio of fluorescence intensity at two excitation wavelengths, the concentration ratio of the Ca2⫹-bound indicator to the Ca2⫹-free indicator can be determined. The gastric epithelial cells were cultured on collagen (type I) glass cover slips (Biocoat, Benecton-Dickinson) and loaded with the acetoxymethyl ester of fura-2 (10 ␮M, Molecular Probes, Eugene, Oregon, USA) for 30 – 45 min at 37°C. Emission intensity (510 nm) was measured with cooled digital 12-bit camera (Sensicam, PCO, Kelheim, Germany) at 340 and 380 nm excitation and the 340/380 ratio image was calculated. After the experiment, the calcium measurements were calibrated (about 30 min) with calcium-free and 2.5 mM calcium solutions containing 10 ␮M ionomycin (Sigma Chemical Co., St. Louis, Missouri, USA). Thereafter the monolayer was exposed to 20 ␮M Digitonin (Fluka, Buchs, Switzerland) to measure background intensities, and these measurements were subtracted from the intensity measurements before any calculations. [Ca2⫹]i was calculated according to the Grynkiewicz equation(15). Pharmacological Agents Used in Cell Cultures. Before wounding the cell cultures were washed twice with 37°C Ham’s F-12 and the medium (Ham’s F-12) that had been removed was replaced with a serum-free medium containing one of the following test agents: EGF (human recombinant, Promega, Madison, Wisconsin, USA; 30 ng/ml), genistein (inhibitor of tyrosine kinase; 100 nM), verapamil (calcium channel blocker; 100 nM, 3,4,5-trimethoxybenzoic acid 8-(diethylamino)octyl ester (TMB-8, intracellular Ca2⫹-antagonist, 10 ␮M), calmidazolium (Ca2⫹/calmodulin inhibitor; 5 ␮M), thapsigargin (endogenous calcium releaser; 100 and 500 nM), ionomycin (calcium ionophore; 100 and 500 nM), phorbol 12-myristate 13-acetate (PMA; 100 nM), calphostin-C (100 nM; protein kinase C inhibitor), cAMP (100 and 500 ␮M), forskolin (adenyl cyclase agonist; 10 ␮M), and 2,3-dideoxyadenosine (DDA, c-AMP inhibitor; 10, 100 ␮M). All these agents were purchased from Sigma Chemicals (St. Louis, Missouri, USA) and dissolved

1009

RANTA-KNUUTTILA ET AL in dimethyl sulfoxide (DMSO). The final concentration of DMSO in the culture medium was 0.01%, which was also added to the respective control cultures. No significant differences were observed between standard and 0.01% DMSO controls. 16,16-Dimethyl-prostaglandin E2 (1, 10, 100 ␮M) was dissolved in ethanol. The final ethanol concentration in culture medium was ⬍ 0.05%. Indomethacin (cyclooxygenase inhibitor; 1, 10, 100 ␮M) and piroxican (cyclooxygenase inhibitor; 10 and 100 ␮M) were also from Sigma Chemicals, and NS-389 (a specific cyclooxygenase-2 inhibitor; 10 and 50 ␮M) was from Cayman Chemicals (Ann Arbor, Michigan, USA). These inhibitors were added to the cell culture 1 and 3 hr before wounding. H-89 (inhibitor of protein kinase-A, 10 ␮M) was purchased from Seikagaku Co (Rockville, Maryland, USA). Determination of Cellular Disruption by Trypan Blue Exclusion. The vital dye trypan blue (100 ␮l of 1% w/v solution, Sigma) was added directly to the monolayer culture. After 5 min, the numbers of stained and nonstained cells were estimated in a random manner by counting 100 –200 cells from each culture under a microscope at 100⫻ magnification. In the monolayer cultures used in the experiments, more than 90% of the cells were vital (nonstained). Statistical Analysis. The results are expressed as mean ⫾ SD. Student’s unpaired t test for equal and unequal variances and analysis of variance were used for statistical analysis of the raw data. P ⬍ 0.05 was considered as statistically significant.

RESULTS In control cultures the mean size of the artificial wound was 3.06 ⫾ 0.5 mm2 (N ⫽ 100). The cell-free area was repopulated gradually by migration and proliferation by cells from the edges of the wound, and the monolayer sheet was completely restored within 48 hr after wounding. The rate of cell migration was linear up until confluence of the monolayer cell sheet, being 21.0 ⫾ 3.8 ␮m/hr and 457 ⫾ 89 ␮m/24 hr (N ⫽ 100). Receptor Tyrosine-Specific Protein Kinase Activation. In order to explore the role of receptor tyrosine protein kinase (RTPK) in the initiation and maintenance of cell migration, the activity of RTPK was modulated by stimulation and inhibition. Stimulation by EGF significantly accelerated migration. The migration speed in the EGF (30 ng/ml) -treated cultures was 621 ⫾ 144 ␮m/24 hr as compared to 457 ⫾ 89 ␮m/24 hr in the respective controls (P ⬍ 0.05; N ⫽ 100). Genistein, a nonspecific receptor tyrosine kinase inhibitor, retarded migration significantly. The migration speed in the genistein (100 ␮M) -treated cultures was 180 ⫾ 90 ␮m/24 hr as compared to 453 ⫾ 82 ␮m/24 hr in the respective controls (P ⬍ 0.05, N ⫽ 100). Genistein also inhibited the EGFprovoked stimulation of migration. The migration speed in the genistein ⫹ EGF-treated cultures was

1010

Fig 1. Effects of EGF and genistein on epithelial cell migration speed compared to the respective controls and effect of genistein on EGF-stimulated migration speed (*P ⬍ 0.05). Black column ⫽ agent, white column ⫽ respective control.

409 ⫾ 107 ␮m/24 hr as compared to 456 ⫾ 72 ␮m/24 hr in the respective controls (N ⫽ 50) (Figure 1). Intracellular Signaling Pathways. In order to assess the roles of the different intracellular signaling pathways in the initiation and maintenance of epithelial cell migration, the activities of potential second messengers were modulated by specific stimulation and inhibition. Ca2ⴙ/Calmodulin Pathway. Inhibition of Ca2⫹ influx across the cell membranes with a Ca2⫹-channel blocker, verapamil, significantly retarded the migration of cultured epithelial cells. In verapamil (100 nM)-treated cultures the cells migrated 247 ⫾ 92 ␮m/24 hr, while the rate was 455 ⫾ 87 ␮m/24 hr in the respective controls (P ⬍ 0.05, N ⫽ 100). Inhibition of endoplasmic Ca 2⫹ release with TMB-8 retarded cell migration. TMB-8 (10 ␮M) -treated cells migrated 385 ⫾ 111 ␮m/24 hr, while the respective controls migrated 486 ⫾ 94 ␮m/24 hr (P ⬍ 0.05 N ⫽ 50). Inhibition of the Ca2⫹/calmodulin complex with calmidazolium likewise retarded the migration of cultured epithelial cells. Calmidazolium (5 ␮M) -treated cells migrated 306 ⫾ 78 ␮m/24 hr, while the speed in the respective controls was 459 ⫾ 69 ␮m/24 hr (P ⬍ 0.05, N ⫽ 70). Calmidazolium also partially abolished the EGF induced stimulation of cell migration. EGF (30 ng/ml) -treated cells migrated 619 ⫾ 179 ␮m/24 hr, while calmidazolium ⫹ EGFtreated cells migrated 378 ⫾ 109 ␮m/24 hr (P ⬍ 0.05, N ⫽ 58) and the respective controls migrated 459 ⫾ 69 ␮m/24 hr (N ⫽ 58) (Figure 2). An artificial increase of intracellular Ca2⫹ by ionomycin (a calcium ionophore, which allows Ca2⫹ to move from extracellular fluid into the cytosol) or by thapsigargin (a releaser of endoplasmic Ca2⫹ from intracellular endoplasmic stores) had no influence on the migration speed of epithelial cells. Ionomycin Digestive Diseases and Sciences, Vol. 47, No. 5 (May 2002)

GASTRIC EPITHELIAL CELL MIGRATION

Fig 2. Effects of verapamil, TMB-8, calmidazolium, ionomycin, and thapsigargin on cell migration speed during the first 24 hr compared to the controls and the effect of calmidazolium on EGFstimulated migration speed compared to the controls, (*P ⬍ 0.05). Black column ⫽ agent, white column ⫽ respective control, grey column ⫽ with EGF (#P ⬍ 0.05 as compared to EGF-stimulated monolayers).

(100 nM) -treated cells migrated 479 ⫾ 148 ␮m/24 hr and the respective controls 481 ⫾ 103 ␮m/24 hr (N ⫽ 76), while thapsigargin (100 nM) -treated cells migrated 445 ⫾ 93 ␮m/24 hr and the respective controls 491 ⫾ 98 ␮m/24 hr (N ⫽ 56) (Figure 2). PKC Pathway. The role of protein kinase C (PKC) in epithelial cell migration was studied by using the PKC activator PMA and inhibitor calphostin-C. In short-term exposure (5 min–2 hr) PMA (100 nM) did not stimulate cell migration compared to the controls. However, long-term exposure (24 hr) to PMA, which has been reported to down-regulate PKC expression in many cell lines (16, 17), significantly retarded cell migration. The migration speed decreased to 228 ⫾ 117 ␮m/24 hr, while it was 379 ⫾ 98 ␮m/24 hr in the respective controls (P ⬍ 0.05, N ⫽ 56) (Figure 3). Inhibition of PKC by calphostin-C (1 ␮M) significantly retarded cell migration to 315 ⫾ 97 ␮m/24 hr (P ⬍ 0.05, N ⫽ 50). It also completely abolished EGF-induced stimulation of cell migration, slowing it to 320 ⫾ 183 ␮m/24 hr (P ⬍ 0.05, N ⫽ 50). Migration speed of the respective controls was 424 ⫾ 87 ␮m/24 hr (Figure 3). Cyclic-AMP-Dependent Protein Kinase A (PKA) Pathway. cAMP, forskolin (cAMP agonist), and DDA (cAMP antagonist) had no influence on the migration speed of epithelial cells. Similarly, H-89 (PKA antagonist) had no effect on migration speed (Table 1). Intracellular Free Calcium Concentration. The Digestive Diseases and Sciences, Vol. 47, No. 5 (May 2002)

Fig 3. Effects of PMA (long- and short-term treatment) on epithelial cell migration speed and effects of calphostin-C on normal and EGF-stimulated cell migration speed during the first 24 hr compared to the controls (*P ⬍ 0.05). Black column ⫽ agent, white column ⫽ respective control, grey column ⫽ with EGF (#P ⬍ 0.05 as compared to EGF stimulated monolayers).

[Ca2⫹]i in nonmigrating cells (confluent monolayer) was 58 ⫾ 48 nM. In migrating cells at the edge of the wound, it was significantly (P ⬍ 0.05, N ⫽ 10) higher (236 ⫾ 30 nM) than in intact cells of the same epithelium (47 ⫾ 8 nM). In epithelia treated with verapamil (100 nM), calmidazolium (5 ␮M), or TMB-8 (10 ␮M), [Ca2⫹]i in cells at the wound edge was 120 ⫾ 17 nM (N ⫽ 6), 77 ⫾ 12 nM (N ⫽ 10) and 61 ⫾ 20 nM (N ⫽ 6), respectively. These values were not significantly different from those of intact nonmigrating cells of the same epithelia (63 ⫾ 21, 52 ⫾ 4 and 47 ⫾ 15 nM for verapamil, calmidazolium and TMB-8, respectively), but they were significantly (P ⬍ 0.05) smaller than [Ca2⫹]i of migrating control cells (see above) (Figure 4). All [Ca2⫹]i measurements were performed 2 hr after wounding. Prostaglandin Synthesis. Tissue and cellular damage is usually associated with stimulation of prostaTABLE 1. EFFECTS OF COX-1 AND COX-2 INHIBITORS AND dm-PGE2 ON EPITHELIAL CELL MIGRATION SPEED IN 24 HOURS AFTER WOUNDING COMPARED TO THE CONTROLS Concentration (␮M) Indomethacin NS-389 Piroxicam PGE2 PGE2 ⫹ Indomethacin

100 10 50 100 1 1 ⫹ 100

Agent

Control

N

P

52 ⫾ 58 53 ⫾ 48 46 ⫾ 91 47 ⫾ 93 32 ⫾ 80 36 ⫾ 194

484 ⫾ 61 520 ⫾ 79 508 ⫾ 74 456 ⫾ 86 355 ⫾ 133 355 ⫾ 133

36 24 30 42 24 9

NS NS NS NS NS NS

1011

RANTA-KNUUTTILA ET AL

Fig 4. Intracellular Ca2⫹ concentration of nonmigrating intact epithelial cells and migrating epithelial cells (edge of the wound) and the effects of verapamil, calmidazolium, and TMB-8 treatment on intracellular Ca2⫹ concentration 2 hr after wounding, (*P ⬍ 0.05).

glandin synthesis, which is assumed to have protective and preparative functions. We therefore assessed also the potential role of this process in the initiation and maintenance of epithelial cell migration. It turned out that exogenous PGE2 had no influence on epithelial cell migration. Similarly inhibition of endogenous prostaglandin synthesis nonspecifically by indomethacin and piroxicam or COX-2 specifically by NS-389 had no influence on epithelial cell migration speed compared to the controls (Table 2). DISCUSSION EGF has been shown to increase intestinal epithelial cell monolayer restitution by stimulating EGF receptor tyrosine phosphorylation and PLC activity (13). The present studies also showed that activation of EGF-R is an important event in gastric epithelial cell migration and repair. We showed that EGF significantly stimulates gastric epithelial cell migration, while genistein significantly retarded it. Genistein also blocked EGF-induced stimulation of the migration speed. This is in accordance with previous studies, which showed that EGF promoted electrophysiological restoration and morphological restitution of erosive lesions of rabbit duodenum in vitro (10) and TABLE 2. EFFECTS OF DDA, FORSKOLIN, cAMP AND H-89 ON EPITHELIAL CELL MIGRATION SPEED IN 24 HOURS AFTER WOUNDING COMPARED TO CONTROLS

cAMP Forskolin DDA H-89

1012

Concentration (␮M)

Agent

Control

N

P

100 10 100 10

471 ⫾ 132 394 ⫾ 78 397 ⫾ 19 541 ⫾ 75

437 ⫾ 161 355 ⫾ 133 415 ⫾ 44 421 ⫾ 96

69 36 9 33

NS NS NS NS

genistein inhibited EGF-stimulated Caco-2 cell migration (7, 8). The presence of Ca2⫹ is necessary for gastric reepithelization to occur. The absence of Ca2⫹ in the incubation medium prevents restitution in isolated frog gastric mucosa by an unknown mechanism(18). The present study demonstrates that intracellular Ca2⫹ has an important role in the migration process of gastric surface epithelial cells. Initiation of cell migration after wounding the monolayer culture is associated with a significant increase in cytosolic Ca2⫹ concentration. If this increase is abolished by blocking Ca2⫹ channels in the cell membrane (with verapamil) or by inhibiting Ca2⫹ release from intracellular stores (with TMB-8), cell migration is inhibited. On the other hand, the artificial increase of free cytosolic Ca2⫹ concentration by a Ca2⫹ ionophore, ionomycin, or a liberator of Ca2⫹ from intracellular stores, thapsigargin, did not have any influence on cell migration either in normally healing or in genistein-inhibited wounds. This indicates that even though cytosolic Ca2⫹ is involved in intracellular signaling of epithelial cell migration, its concentration is not the critical determinant or rate limiting factor of migration speed. Obviously, other concomitant signaling mechanisms are also involved, which may mutually interact with cytosolic Ca2⫹ or Ca2⫹/calmodulin complex. For example, activation of PKC, which likewise contributes to epithelial cell migration (see below), requires the presence of Ca2⫹. Calmodulin is a Ca2⫹ binding protein present in all eukaryotic cells so far studied. It functions as a multipurpose intracellular Ca2⫹ receptor mediating Ca2⫹-regulated processes. Calmodulin has four highaffinity Ca2⫹ binding sites. The Ca2⫹/calmodulin complex has no enzymatic activity itself, but acts by binding to other proteins, thereby altering their activity. The present study shows that inhibition of the formation of Ca2⫹/calmodulin complex with calmidazolium significantly retards gastric epithelial cell migration and, further, abolishes the EGF-induced stimulation of cell migration, thus suggesting that the Ca2⫹/ calmodulin complex acts as a second messenger in EGF-R activation promoted cell migration. The mechanism of action of Ca2⫹/calmodulin complex in epithelial cell migration and restitution may be related to the actin polymerization necessary for epithelial restitution. Cytochalasin B, which prevents actin polymerization and causes actin filaments to retract, prevents lamellepodium formation in cells at the edge of the wound and completely inhibits wound repair and restitution (18, 19). Myosin light-chain Digestive Diseases and Sciences, Vol. 47, No. 5 (May 2002)

GASTRIC EPITHELIAL CELL MIGRATION

kinase (MLCK), a Ca2⫹/calmodulin-dependent protein kinase (Cam-kinase), is the enzyme that catalyzes phosphorylation of myosin light chains. The action of MLCK is dependent on the binding of the Ca2⫹/ calmodulin complex. Wortmannin, a potent selective inhibitor of myosin light-chain kinase, also inhibits epithelial cell migration in basal conditions (13, 20) as well as after EGF stimulation (13). Thus, the Ca2⫹/ calmodulin complex seems to have an important role in the initiation and maintenance of migration process of epithelial cells. Protein kinase C has been intimately implicated in intracellular signal transduction in various cellular processes, including cell migration. Polk (13) has shown that the initial steps in intestinal epithelial cell migration are tyrosine phosphorylation of the EGF receptor (probably triggered by ligand activation) and phosphorylation of phospholipase C-␥1 (PLC-␥1). Activation of PLC-␥1 leads to increased hydrolysis of phosphatidylinositol 4,5-biphosphate (PIP2) to inositol 1,4,5-triphosphate (IP 3 ) and diacylglycerol (DAG), which, in turn, promotes activation of PKC. It was speculated that PKC activity might be the downstream target of EGF-stimulated signaling, which initiates and regulates epithelial cell migration. Indeed, the studies indicated that inhibition of PKC activity directly by calphostin-C or indirectly by longterm exposure to phorbol esters (which downregulate PKC activity) (16, 21) completely abolishes EGF-induced stimulation of cell migration, whereas short-term exposure to phorbol esters stimulates PKC activity and reverses the effects of PLC-␥1 and IP3 inhibitors on migration. Our findings are largely in accordance with these observations. The present study also showed that long-term exposure to PMA inhibited cell migration. In contrast, short-term exposure to PMA did not stimulate migration speed. Calphostin-C treatment retarded cell migration and completely abolished EGF-induced stimulation of it. cAMP is synthesized from ATP by the plasmamembrane-bound enzyme adenyl cyclase. Forskolin directly activates and DDA inhibits the function of adenyl cyclase. cAMP exerts its effects in animal cells mainly by activating the enzyme protein kinase A (PKA). PGE2 has been shown to stimulate corneal endothelial cell migration, and this effect seems to be mediated via PGE2-induced activation of the intracellular cAMP signaling pathway (22). Further, the PKA inhibitor H-89 and adenyl cyclase inhibitor DDA reversed the stimulatory effects of PGE2. In the present study, activation of adenyl cyclase by forskolin or inhibition of it by DDA had no effect on epithelial Digestive Diseases and Sciences, Vol. 47, No. 5 (May 2002)

cell migration. Likewise, inhibition of PKA by H-89 did not display any effect on it. Prostaglandins are signaling molecules that often act in an autocrine fashion in animal tissues. They are synthesized from precursors (arachidonic acid) that are cleaved from membrane phospholipids and released to the cell exterior. Prostaglandin synthesis is activated, among other things, by tissue damage, such as epithelial wounding. In several cell types, PGE2 and/or increased levels of cAMP induce changes in actin microfilament organization, which results in alteration of cell shape (23, 24). Such cytoskeletal changes may contribute to the enhanced migratory movement by PGE2 as observed in corneal endothelial cells (22) and in fibroblasts (23). However, in the present study addition of exogenous PGE2 or inhibition of endogenous prostaglandin synthesis had no effect on gastric epithelial cell migration. In conclusion, the present results indicate that EGF-R activation has as an important role in gastric epithelial cell migration. The increase in [Ca2⫹]i and formation of the Ca2⫹/calmodulin complex are essential for gastric epithelial cell migration and EGFinduced stimulation of it. PKC seems also to have an important role in this process. Yet, the present study could not show any processes dependent on cAMP and PKA or prostaglandin that might have a role in gastric epithelial cell migration. REFERENCES 1. Svanes K, Ito S, Takeuchi K, Silen W: Restitution of the surface epithelium of the in vitro frog gastric mucosa after damage with hyperosmolar sodium chloride. Morphologic and physiologic characteristics. Gastroenterology 82:1409 –1426, 1982 2. Rutten MJ, Ito S: Morphology and electrophysiology of guinea pig gastric mucosal repair in vitro. Am J Physiol 244:G171– G182, 1983 3. Feil W, Klimesch S, Karner P, Wenzl E, Starlinger M, Lacy ER, Schiessel R: Importance of an alkaline microenvironment for rapid restitution of the rabbit duodenal mucosa in vitro. Gastroenterology 97:112–122, 1989 4. Tarnawski A, Stachura J, Durbin T, Sarfeh IJ, Gergely H: Increased expression of epidermal growth factor receptor during gastric ulcer healing in rats. Gastroenterology 102:695– 698, 1992 5. Wright NA, Pike C, Elia G: Induction of a novel epidermal growth factor-secreting cell lineage by mucosal ulceration in human gastrointestinal stem cells. Nature 343:82– 85, 1990 6. Lee H, Hansson HA, Norstrom E, Helander HF: Immunoreactivities for epidermal growth factor (EGF) and for EGF receptors in rats with gastric ulcers. Cell Tissue Res 265:211– 218, 1991 7. Basson MD, Beidler DR, Turowski G, Zarif A, Modlin IM, Jena BP, Madri, JA: Effect of tyrosine kinase inhibition on

1013

RANTA-KNUUTTILA ET AL

8.

9.

10.

11.

12.

13.

14.

15.

basal and epidermal growth factor-stimulated human Caco-2 enterocyte sheet migration and proliferation. J Cell Physiol 160:491–501, 1994 Dignass AU, Podolsky DK: Cytokine modulation of intestinal epithelial cell restitution: Central role of transforming growth factor beta. Gastroenterology 105:1323–1332, 1993 Maehiro K, Watanabe S, Hirose M, Iwazaki, R, Miwa H, Sato N: Effects of epidermal growth factor and insulin on migration and proliferation of primary cultured rabbit gastric epithelial cells. Journal of Gastroenterology 32:573–578, 1997 Riegler M, Sedivy R, Sogukoglu T, Cosentini E, Bischof G, Teleky B, Feil W, Schiessel R, Hamilton G, Wenzl E: Epidermal growth factor promotes rapid response to epithelial injury in rabbit duodenum in vitro. Gastroenterology 111:28 –36, 1996 Relan NK, Fligiel SE, Dutta S, Tureaud J, Chauhan DPX, Majumdar AP: Induction of EGF-receptor tyrosine kinase during early reparative phase of gastric mucosa and effects of aging. Lab Invest 73:717–726, 1995 Majumdar AP, Fligiel SE, Jaszewski R, Tureaud J, Dutta, S, Chelluderai B: Inhibition of gastric mucosal regeneration by tyrphostin: Evaluation of the role of epidermal growth factor receptor tyrosine kinase. J Lab Clin Med 128:173–180, 1996 Polk DB: Epidermal growth factor receptor-stimulated intestinal epithelial cell migration requires phospholipase C activity. Gastroenterology 114:493–502, 1998 Watanabe S, Hirose M, Wang XE, Maehiro, K, Murai T, Kobayashi O, Mikami H, Otaka K, Miyazaki A, Sato N: A new model to study repair of gastric mucosa using primary cultured rabbit gastric epithelial cells. J Clin Gastroenterol 21(suppl 1):S40 –S44, 1995 Grynkiewicz G, Poenie M, Tsien RY: A new generation of

1014

16.

17.

18.

19.

20.

21. 22. 23. 24.

Ca2⫹ indicators with greatly improved fluorescence properties. J Biol Chem 260:3440 –3450, 1985 Derman MP, Toker A, Hartwig JH, Spokes K, Falck JRX, Chen CS, Cantley LC, Cantley LG: The lipid products of phosphoinositide 3-kinase increase cell motility through protein kinase C. J Biol Chem 272:6465– 6470, 1997 Rozengurt E, Rodriguez-Pena M, Smith KA: Phorbol esters, phospholipase C, and growth factors rapidly stimulate the phosphorylation of a Mr 80,000 protein in intact quiescent 3T3 cells. Proc Nat Acad Sci USA 80:7244 –7248, 1983 Critchlow J, Magee D, Ito S, Takeuchi K, Silen W: Requirements for restitution of the surface epithelium of frog stomach after mucosal injury. Gastroenterology 88:237–249, 1985 Watanabe S, Hirose M, Yasuda T, Miyazaki, A, Sato N: Role of actin and calmodulin in migration and proliferation of rabbit gastric mucosal cells in culture. J Gastroenterol Hepatol 9:325– 333, 1994 Watanabe S, Wang XE, Hirose M, Sato N: Effect of myosin light chain kinase inhibitor wortmannin on the wound repair of cultured gastric mucosal cells. Biochem Biophys Res Commun 199:799 – 806, 1994 Mellgren RL: Calcium-dependent proteases: An enzyme system active at cellular membranes? FASEB J 1:110 –115, 1987 Joyce NC, Meklir B: PGE2: A mediator of corneal endothelial wound repair in vitro. Am J Physiol 266:C269 –C275, 1994 Gotlieb AI: Prostaglandin induced shape changes in fibroblasts grown in cell culture. Prostaglandins 19:865– 871, 1980 Northover AM: The effects of indomethacin and verapamil on the shape changes of vascular endothelial cells resulting from exposure to various inflammatory agents. Agents Actions 24:351–355, 1988

Digestive Diseases and Sciences, Vol. 47, No. 5 (May 2002)