Rapid mitogen-activated protein kinase activation by transforming growth factor α in wounded rat intestinal epithelial cells

GASTROENTEROLOGY 1998;114:697–705

Rapid Mitogen-Activated Protein Kinase Activation by Transforming Growth Factor a in Wounded Rat Intestinal Epithelial Cells ¨ KE, MICHIYUKI KANAI, KATHRYN LYNCH–DEVANEY, and DANIEL K. PODOLSKY MICHAEL GO Gastrointestinal Unit, Department of Medicine, and Center for the Study of Inflammatory Bowel Disease, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts

Background & Aims: To define signaling events initiating healing after intestinal epithelial injury, activation of mitogen-activated protein kinase (MAPK) pathways was assessed after wounding using an in vitro model. Methods: Proteins isolated from wounded monolayers of nontransformed intestinal epithelial cells (IEC-6) were analyzed for tyrosine phosphorylation and MAPK expression by Western blot. Extracellular signalregulated kinase (ERK) 1, ERK2, and Raf-1 activities were assessed by immune complex kinase assays. Results: Tyrosine phosphorylation of several proteins including ERK1 was substantially increased 5 minutes after injury. Another MAPK, c-Jun-N-terminal protein kinase (JNK), was also activated after wounding. Conditioned medium from wounded but not intact IEC-6 monolayers resulted in increased activity of ERK1, ERK2, and Raf-1 kinase. Wound-conditioned medium stimulated proliferation of subconfluent IEC-6 cells compared with conditioned medium from intact IEC-6 cultures and contained higher amounts of transforming growth factor (TGF)-a than supernatants of confluent IEC-6 cultures. Activation of ERK1 and ERK2 was partially inhibited by neutralizing anti–TGF-a. Conclusions: Wounding of intestinal epithelial cells results in activation of Raf-1, ERK1, ERK2, and JNK1 MAPKs and subsequent cell proliferation in vitro. Activation of ERK1 and ERK2 is mediated in part by TGF-a.

he surface epithelium of the gastrointestinal tract forms a barrier that interfaces with a broad spectrum of noxious agents present within the lumen. Rapid repair is essential after various forms of injury, e.g., gastroduodenal erosions/ulcerations, inflammatory bowel disease, enteropathogenic infections, ischemia, or radiation. In vitro and in vivo studies have shown that repair of mucosal defects in the gastrointestinal tract involves at least two processes: rapid migration (also designated restitution) to reestablish epithelial continuity and proliferation to replace destroyed intestinal epithelial cells.1–4 Various growth factors and cytokines present in the epithelium or lamina propria of the mucosa modulate

T

intestinal epithelial cell migration and growth in vitro.5–13 Among these peptides, transforming growth factor (TGF)-a and -b have also been shown to be expressed in intestinal epithelial cells.14 TGF-a is known to increase proliferation of intestinal epithelial cells, whereas TGF-b inhibits intestinal epithelial cell growth. Despite their differential effects on epithelial cell growth, both peptides enhance intestinal epithelial cell migration in the important restitution phase early after epithelial injury.5,6 In contrast to accumulating knowledge about the effects of peptide growth factors and cytokines on intestinal epithelial wound healing, the intracellular events that initiate these epithelial responses have not been well characterized. In recent years, the mitogen-activated protein kinase (MAPK) pathways have been recognized as a major signaling system by which cells transduce extracellular signals. Three distinct MAPK cascades have been identified: (1) extracellular signal-regulated kinases (ERKs) ERK1 (p44) and ERK2 (p42),15 (2) c-Jun-N-terminal protein kinases ( JNKs)/stress-activated protein kinase,16–19 and (3) p38/RK/HOG1.20,21 Activation of ERK1 and ERK2 MAPKs is thought to play an important role in mediating cell proliferation in response to growth factors such as insulin, epidermal growth factor (EGF), TGF-a, acidic and basic fibroblast growth factor, and hepatocyte growth factor. In contrast, JNK/stressactivated protein kinases are activated by cellular stress including UV irradiation, osmotic challenge, heat shock, translational inhibitors (e.g., anisomycin), TNF-a, and interleukin 1b and, to a lesser extent, by mitogens. Members of the third type of MAPKs, p38 and HOG1, Abbreviations used in this paper: BSA, bovine serum albumin; DMEM, Dulbecco’s modified Eagle medium; EGF, epidermal growth factor; ERK, extracellular signal-regulated kinase; FBS, fetal bovine serum; IEC, intestinal epithelial cell; JNK, c-Jun-N-terminal protein kinase; MAPK, mitogen-activated protein kinase; MBP, myelin basic protein; SDS, sodium dodecyl sulfate; TBS, Tris-buffered saline; TGF, transforming growth factor. r 1998 by the American Gastroenterological Association 0016-5085/98/$3.00

698

¨ KE ET AL. GO

are activated in response to high osmolarity, endotoxic lipopolysaccharide, and many chemicals. Signaling events after intestinal wounding have not been well characterized. Because cytokines and regulatory peptides that promote epithelial wound repair can activate both ERK1, ERK2, and JNK MAPKs, it is reasonable to hypothesize that repair mechanisms after intestinal epithelial wounding are mediated by activation of MAPK signal transduction pathways. In the present study, an in vitro model of intestinal epithelial injury was used to define signal transduction pathways after wounding.

Materials and Methods Cell Culture Intestinal epithelial cells (IEC-6) derived from rat small intestinal epithelium22 were cultured in Dulbecco’s modified Eagle medium (DMEM; Cellgro; Mediatech Inc., Herndon, VA) containing 4.5 g/L glucose, 4 mmol/L L-glutamine, 5 µg/mL insulin, 100 IU/mL penicillin, 100 µg/mL streptomycin, and 5% heat-inactivated fetal bovine serum (FBS; Sigma Chemical Co., St. Louis, MO). IEC-6 cells were used at 15–20th passage and grown in 100-mm plates for the study of protein expression and phosphorylation, in vitro kinase assays, or collection of conditioned medium. Cells were cultured in 24-well plates (Corning-Costar Corp., Cambridge, MA) for assessment of cell proliferation.

Peptides and Neutralizing Antibodies Used To define the effects of TGF-a on ERK1, ERK2, and Raf-1 activity, 5 and 10 ng/mL recombinant human TGF-a (R&D Systems, Minneapolis, MN) were added to the cell culture medium. For immunoneutralization of TGF-b and TGF-a, 50 µg/mL neutralizing chicken anti-human TGF-b1 antibodies (Collaborative Biomedical Products/Becton Dickinson, Bedford, MA), 15 µg/mL rabbit anti-human pan–TGF-b antibodies (R&D Systems), or 15 µg/mL goat anti-human TGF-a antibodies (R&D Systems) were added to serum-starved IEC-6 cells 3 minutes before scraping the monolayer and immediately after wounding for 5 minutes, or to conditioned medium collected from wounded IEC-6 monolayers. Bovine serum albumin (BSA; Sigma) and normal rabbit immunoglobulin (Ig) G (R&D Systems) served as controls for peptides and neutralizing antibodies, respectively.

In Vitro IEC-6 Wound Assay IEC-6 cell wound assays were performed by modification of a previously described technique.3,5,23 IEC-6 cells were grown in 100- or 60-mm polystyrene dishes (Corning, New York, NY) to confluency, then washed and cultured in serum-deprived (0.1% FBS containing) medium for 24 hours. To generate a high surface of wounded cells, multiple standard linear wounds were created with a razor blade; aggregate

GASTROENTEROLOGY Vol. 114, No. 4

wounds encompassed approximately 50% of the monolayer surface area. Immediately after scraping the monolayers, supernatants were aspirated to remove cellular debris, followed by addition of fresh serum-deprived medium to the wounded monolayers. At several time points (0, 1, 5, 15, 30, 60, and up to 360 minutes) after wounding, cells were lysed and proteins extracted. Confluent unwounded monolayers subjected to the same medium changes served as controls. For evaluation of paracrine effects on MAPK phosphorylation and activity after epithelial wounding, confluent IEC-6 monolayers maintained in 0.1% FBS containing DMEM were cultured for 5 minutes in the presence of conditioned medium collected from other IEC-6 cell cultures. Wound-conditioned medium was obtained by harvesting the medium 5 minutes after wounding; the conditioned medium was then centrifuged to remove all cellular debris. Conditioned medium from intact confluent IEC-6 monolayers was also collected for comparison.

Western Blotting of Tyrosine Phosphorylated Proteins in Intestinal Epithelial Cells IEC-6 were washed with phosphate-buffered saline, and proteins were extracted by lysis buffer (1 mmol/L ethylenediaminetetraacetic acid [EDTA]–10 mmol/L Tris-HCl [pH 7.4], 150 mmol/L NaCl, 1% NP-40, 0.05% deoxycholic acid containing 2 mg/mL aprotinin, 10 µg/mL leupeptin, 2 mmol/L phenylmethylsulfonamide, 100 mmol/L sodium fluoride, 200 µmol/L sodium orthovanadate, and 10 mmol/L tetrasodium pyrophosphate). Extracts were centrifuged at 10,000g for 15 minutes, and the protein concentration in each supernatant was determined by colorimetric Bradford protein assay (Bio-Rad Laboratories, Hercules, CA). Equal amounts of proteins (50 µg/lane) were subjected to electrophoresis in 7.5% polyacrylamide gels under reducing conditions (5%, vol/vol, 2mercaptoethanol) according to Laemmli24 and electroeluted onto polyvinylidene difluoride–Immobilon P transfer membranes (Millipore, Bedford, MA) in transfer buffer (50 mmol/L Tris, 0.38 mol/L glycine, and 10%, vol/vol, methanol) for 12 hours at 30 mA and stained (0.1% Ponceau red, 1% acetic acid). After washing, membranes were blocked in 13 Trisbuffered saline (TBS), 0.05% Tween 20, and 1% BSA (blocking buffer) at room temperature for 1 hour. Blots were then incubated with antiphosphotyrosine antibody (monoclonal mouse PY-20 IgG2b from Transduction Laboratories, Lexington, KY; diluted 1:20,000 in blocking buffer) for 3 hours at room temperature. After washing with TBS and 0.05% Tween 20, hybridization with secondary antibody (sheep anti-mouse Ig, horseradish peroxidase–linked antibody from Amersham Life Science, diluted 1:15,000 in TBS, 0.05% Tween 20, and 5% nonfat dry milk) was performed for 45 minutes at room temperature. After washing, phosphorylation signals were detected using Renaissance enhanced chemiluminescent reagents (DuPont NEN, Boston, MA) and autoradiography.

April 1998

Assessment of ERK1 and ERK2 Tyrosine Phosphorylation by Immunoprecipitation and Western Blotting Cells were lysed and proteins extracted as described above. Preclearing of lysates (1 mL/tube with a protein concentration of 1 mg/mL as assessed by colorimetric protein assay) was performed using 1 µg/mL normal rabbit IgG (R&D Systems) and 100 µL protein A–Sepharose (Pharmacia, Piscataway, NJ) on a rocking table for 30 minutes at 4°C. After brief centrifugation, supernatants were subjected to immunoprecipitation using 1.5 µg/mL polyclonal IgG rabbit anti-rat ERK1 antibody (sc-93; Santa Cruz Biotechnology, Santa Cruz, CA), 1.5 µg/mL polyclonal IgG rabbit anti-rat ERK2 antibody (sc-154; Santa Cruz), and protein A–Sepharose beads (100 µL/mL added 3 hours after addition of antibody) on a rocking table at 4°C overnight. Beads were washed two times with lysis buffer, 50 µL of 23 Laemmli sample buffer was added to each tube, and samples were boiled for 6 minutes. Electrophoresis of immunoprecipitates in 9% polyacrylamide gels was performed under reducing conditions. Proteins were electroeluted onto polyvinylidene difluoride–Immobilon P transfer membranes in transfer buffer as described above. Membranes were blocked in blocking buffer at room temperature for 1 hour. Blots were then incubated with an antiphosphotyrosine antibody (PY-20 diluted 1:20,000) for 3 hours at room temperature. After washing three times (5 minutes each) in 13 TBS and 0.05% Tween 20, hybridization with the secondary antibody (sheep anti-mouse, horseradish peroxidase–linked Ig antibody from Amersham Life Science, diluted 1:15,000 in 13 TBS, 0.05% Tween 20, and 5% nonfat dry milk) was performed for 45 minutes at room temperature. After probing with antiphosphotyrosine antibody (PY-20), membranes were stripped using 62.5 mmol/L Tris-HCl (pH 6.8), 2% sodium dodecyl sulfate (SDS), and 100 mmol/L 2-mercaptoethanol for 30 minutes at 50°C, subsequently washed in TBS and 0.05% Tween 20, and reprobed with anti-ERK1 and anti-ERK2 antibodies, respectively. After washing, incubation with the secondary antibody (donkey anti-rabbit Ig and horseradish peroxidase–linked antibody, diluted 1:15,000) was performed for 45 minutes at room temperature. Detection was carried out as described above.

Assessment of ERK1, ERK2, Raf-1, and JNK Activity by Immune Complex In Vitro Kinase Assays Cell harvesting, lysis, and protein extraction and immunoprecipitation with 1.5 µg/mL anti-ERK1 (sc-93; Santa Cruz), anti-ERK2 (sc-154; Santa Cruz), or anti–Raf-1 (polyclonal IgG rabbit anti-human Raf-1 antibody, sc-133; Santa Cruz) antibodies were performed as described above. After the second wash with lysis buffer, beads were washed with cold kinase buffer (40 mmol/L HEPES [pH 7.5], 10 mmol/L MgCl2, and 3 mmol/L MnCl2) and kept on ice. Subsequently, 50 µL of kinase buffer including 150 µg/mL

MAPK ACTIVATION IN WOUNDED IEC–6 CELLS

699

bovine brain myelin basic protein (MBP; Sigma) as substrate for ERK1, ERK2, or Raf-1 kinase and 10 µCi [g-32P]adenosine triphosphate (ATP) (DuPont NEN) was added to each tube. Kinase reactions were performed for 20 minutes in a waterbath at 28°C, then 50 µL of 23 Laemmli sample buffer was added to each tube. After boiling, the products of the kinase reactions were separated on 12.5% SDS-polyacrylamide gels. After electrophoresis, gels were dried and exposed to film. A similar method was used for JNK immune complex kinase assays, except different buffers were used,25 including a lysis buffer that consisted of 10 mmol/L Tris (pH 7.4), 50 mmol/L NaCl, 50 mmol/L NaF, 5 mmol/L EDTA, 1 mmol/L sodium vanadate, 0.1% BSA, 1% Triton X-100, 20 µg/mL aprotinin, and 20 µg/mL leupeptin; a wash buffer including 50 mmol/L Tris (pH 8.0), 0.1 mmol/L ethylene glycol-bis(b-aminoethyl ether)-N,N,N8,N8-tetraacetic acid, 0.5 mmol/L sodium vanadate, and 0.1% (vol/vol) b-mercaptoethanol; a kinase buffer containing 20 mmol/L HEPES (pH 7.6), 20 mmol/L MgCl2, 20 mmol/L b-glycerophosphate, 0.1 mmol/L sodium vanadate, and 2 mmol/L dithiothreitol and 1 µg of substrate. Anti-JNK1 antibody (sc-474) and glutathione S-transferase-c-Jun (amino acids 1–79) substrate (sc-4113) were obtained from Santa Cruz Biotechnology.

Measurement of TGF-a Concentration in IEC-6 Culture Supernatants TGF-a concentrations in cell culture media were assessed by a commercial EGF/TGF-a radioassay (Biomedical Technologies Inc., Stoughton, MA). This assay is based on competitive binding between endogenous TGF-a and exogenously added 125I-EGF to EGF/TGF-a receptors in A-431 cell membranes. Measurements were performed according to the manufacturer’s instructions.

Assessment of IEC-6 Cell Proliferation IEC-6 cells were grown in 24-well plates to approximately 50% confluency, washed, and cultured for 24 hours in fresh serum-deprived medium. Cells were cultured for an additional 24 hours at 37°C in the presence of conditioned medium collected from intact confluent control cells or wounded monolayers (wound-conditioned medium). After 20 hours, [3H]thymidine was added (1.8 µCi/well). After 4 hours, the incorporation of radiolabeled thymidine was determined as described.5 Briefly, cells were washed with phosphate-buffered saline and fixed with methanol–acetic acid (3:1, vol/vol). Acid-insoluble material was then lysed with NaOH, and radioactivity was counted using a liquid scintillation counter.

Statistical Analysis Shown experiments were performed in triplicate. For statistical analysis of proliferation and EGF/TGF-a assays, data are expressed as mean 6 SD. Statistical significance between different groups was evaluated using Wilcoxon’s (unpaired) signed-rank test.26

700

¨ KE ET AL. GO

GASTROENTEROLOGY Vol. 114, No. 4

Results Previous studies have shown the similarity between the nontransformed small intestinal epithelial crypt–derived IEC-6 cell line and normal rat small intestinal crypt epithelial cells including growth characteristics, morphology, extracellular matrix synthesis, and the presence of cell-specific plasma membrane antigens.22,23 In addition, TGF-a and TGF-b, peptide growth factors that regulate intestinal epithelial cell growth and cell migration in intestinal epithelial restitution,5,6 are both produced by IEC-6 cells.14 For the study of signaling events after intestinal epithelial wounding, standard wounds were made in confluent IEC-6 cell monolayers maintained in serumdeprived medium for 24 hours. Cells were wounded by scraping the monolayer followed by washing and further incubation in fresh serum-deprived medium. The use of serum-deprived medium avoids confounding effects of growth factors present in serum. Tyrosine Phosphorylation in Wounded Intestinal Epithelial Cells Overall phosphorylation of proteins in IEC-6 monolayers harvested at different time points after epithelial injury and from unwounded confluent control monolayers was assessed by Western blotting using an antiphosphotyrosine antibody. As shown in Figure 1, a marked increase in tyrosine phosphorylation of proteins with molecular masses of 46, 44, and 42 kilodaltons was observed in wounded IEC-6 monolayers within 5 minutes after injury. This was apparent in monolayers that were washed immediately after scraping of cells to remove cellular debris (shown) and in wounded monolayers that were not washed after injury (data not shown). Use of medium entirely devoid of serum did not change the phosphorylation pattern compared with serumdeprived (0.1% FBS containing) medium (data not shown). No bands were observed in blots incubated with control IgG instead of antiphosphotyrosine antibody. To identify the proteins that are tyrosine phosphorylated after wounding and to evaluate activation of major stress-mediated signal transduction pathways, lysates from wounded and unwounded IEC-6 monolayers were subjected to immunoprecipitation with antibodies specific for key signal transduction proteins and consecutive Western blot analysis of immune complexes using antiphosphotyrosine antibody. As shown in Figure 2A, marked tyrosine phosphorylation of ERK1 protein with a molecular mass of 44 kilodaltons was found within 5 minutes after injury. Stripping of the blot shown in Figure 1 and reprobing with anti-ERK1 antibody

Figure 1. Time course for tyrosine phosphorylation in total cell lysates from wounded IEC-6 cells. Twenty-five micrograms of proteins from wounded (1, 5, 15, 30, 60, and 360 minutes after wounding; w1–w360) and intact confluent monolayers (con 0) were subjected to SDS-polyacrylamide gel electrophoresis (PAGE) analysis and Western blot transfer. Hybridization with an anti–PY-20 antibody, detection, and autoradiography were performed as described in Materials and Methods.

showed a strong signal at 44 kilodaltons 5 minutes after wounding (not shown). For evaluation of total ERK1 protein expression after epithelial wounding, the blot shown in Figure 2A was stripped and reprobed with anti-ERK1 antibody, respectively. As shown in Figure 2B, overall ERK1 protein before and after injury was unchanged. Thus, changes in ERK1 content are not responsible for the increased ERK1 tyrosine phosphorylation in wounded IEC-6 cells. Expression of ERK-2 protein was also unchanged after wounding as assessed by Western blotting (data not shown). Assessment of ERK1, ERK2, and Raf-1 Kinase Activity in Wounded IEC-6 Monolayers The results of the phosphorylation studies suggested that the p44-p42 MAPK signaling pathway may be involved in signaling events after intestinal epithelial wounding. Subsequently, the kinase activities of key MAPK proteins including ERK1, ERK2, and Raf-1 were assessed in wounded and intact confluent IEC-6 monolayers by immune complex in vitro kinase assays using MBP as a substrate. Paralleling the tyrosine phosphorylation

April 1998

MAPK ACTIVATION IN WOUNDED IEC–6 CELLS

701

Figure 2. Tyrosine phosphorylation of ERK1 in wounded IEC-6 monolayers. Lysates containing equal amounts of protein (300 mg/tube) obtained from wounded (5, 15, 30, and 360 minutes after wounding; w5–w360) and intact confluent monolayers (con 0) were immunoprecipitated with anti-ERK1 antibody. Immune complexes were subjected to SDS-PAGE analysis and Western blot transfer. (A) Phosphorylation of ERK1 protein was assessed by hybridization with anti–PY-20 antibody. (B) For evaluation of ERK1 protein expression, the same blot was stripped and reprobed with anti-ERK1 antibody.

changes, a substantial increase of ERK1 and ERK2 kinase activity was observed within 5 minutes after scraping IEC-6 cells as shown in Figure 3A and B. In contrast, ERK2 protein expression was unchanged after wounding as evaluated by Western blotting (data not shown). Activation of Raf-1 protein occurred in IEC-6 cells within 1 minute after wounding reflected by increased phosphorylation of MBP in an in vitro kinase assay (Figure 3C).

Figure 3. Assessment of ERK1, ERK2, and Raf-1 kinase activity in wounded intestinal epithelial cells. Lysates containing equal amounts of protein (300 mg/tube) from wounded (1, 5, and 60 minutes after wounding; w1–w60) and intact confluent monolayers (con 0) were immunoprecipitated with (A) anti-ERK1, (B) anti-ERK2, and (C) anti– Raf-1 antibodies, respectively. Immune complexes were washed and subjected to an in vitro kinase reaction for 20 minutes at 307C using [g-32P]ATP and MBP as a substrate, and analyzed by SDS-PAGE and autoradiography as described in Materials and Methods.

Effects of Wound-Conditioned Medium on Raf-1, ERK1, and ERK2 Kinase Activity in Intestinal Epithelial Cells To determine whether paracrine effects contribute to the rapid increase in ERK1 and ERK2 MAPK activities, conditioned medium obtained 5 minutes after wounding of IEC-6 monolayers or conditioned medium from intact confluent control monolayers was added to confluent serum-starved IEC-6 cells for 5 minutes. Kinase activities were again assessed by phosphorylation of the substrate MBP by ERK1, ERK2, and Raf-1 kinases, respectively. As shown in Figure 4, conditioned medium collected from wounded IEC-6 monolayers induced increased ERK1 and ERK2 activity. MAPK activation was paralleled by increased tyrosine phosphorylation and unchanged expression of ERK1 and ERK2 proteins in IEC-6 cells cultured in the presence of wound-

Figure 4. Effect of wound-conditioned medium on Raf-1, ERK1, and ERK2 kinase activity of intestinal epithelial cells. Conditioned medium was collected from wounded IEC-6 cultures 5 minutes after wounding or from unwounded confluent monolayers and then added to serumstarved intact confluent IEC-6 monolayers. After culture of IEC-6 cells in the presence of either wound conditioned (w) or control conditioned medium (con) for 5 minutes, cell lysates were harvested. Equal amounts of protein (300 mg/tube) were immunoprecipitated with anti–Raf-1, anti-ERK1, and anti-ERK2 antibodies, respectively. Immune complexes were subjected to an in vitro kinase assay with MBP as a substrate, then analyzed by SDS-PAGE and autoradiography.

702

¨ KE ET AL. GO

GASTROENTEROLOGY Vol. 114, No. 4

conditioned medium (data not shown). As shown in Figure 4, an increase in Raf-1 kinase activity was also observed in IEC-6 cells incubated with wound-conditioned medium. Effects of Neutralizing Anti–TGF-a and Anti–TGF-b Antibodies on ERK1 and ERK2 Kinase Activity in Intestinal Epithelial Cells To identify factors that might mediate the rapid activation of the p44-p42 MAPK pathway after intestinal epithelial wounding, ERK1 and ERK2 activities were assessed in unwounded IEC-6 cells that were incubated with conditioned medium from unwounded monolayers, wounded cultures, or wounded cultures supplemented with control IgG, anti–TGF-a, or anti–TGF-b neutralizing antibodies. As shown in Figure 5, wound-conditioned medium containing 15 µg/mL anti–TGF-a antibody resulted in reduced stimulation of ERK1 and ERK2 kinase activity. In contrast, wound-conditioned medium containing either control IgG or anti–TGF-b antibodies

A

B

Figure 5. Effects of neutralizing anti–TGF-a and anti–TGF-b antibodies on ERK1 and ERK2 activity of intestinal epithelial cells. Confluent serum-starved IEC-6 cells were cultured for 5 minutes with conditioned medium collected from unwounded confluent monolayers (con), untreated wounded monolayers (w), or wounded monolayers containing 15 mg/mL normal rabbit IgG (NR IgG), anti–TGF-a antibody, or anti–TGF-b antibody. Equal amounts of proteins in lysates were immunoprecipitated with (A) anti-ERK1 and (B) anti-ERK2 antibodies, respectively. ERK1 and ERK2 in vitro kinase activity was assessed using MBP as a substrate and analyzed as in Figures 3 and 4.

Figure 6. Effects of wound-conditioned medium on IEC-6 cell proliferation. Confluent serum-starved monolayers were cultured for 20 hours in the presence of fresh serum-deprived medium (DMEM), conditioned medium from intact confluent control monolayers (con CM), or wound-conditioned medium in the absence or presence of 15 mg/mL neutralizing anti–TGF-a antibody (w CM and w CM 1 anti–TGF-a, respectively). Cells were fixed and cell proliferation assessed by measurement of [3H]thymidine incorporation (in cpm) as detailed in Materials and Methods. *P 5 0.029 vs. con CM.

did not diminish ERK1 or ERK2 kinase activity in IEC-6 cells. Relevance of TGF-a–Mediated ERK1/ERK2 MAPK Activation in Wounded Intestinal Epithelial Cells Because the above described experiments suggested that TGF-a is involved in activation of the p44-p42 MAP-kinase pathway after intestinal epithelial wounding, the concentration of TGF-a in woundconditioned medium compared with control-conditioned medium was assessed using a competitive EGF/TGF-a radioassay. TGF-a concentrations were significantly higher in conditioned medium collected from wounded IEC-6 cultures 5 minutes after injury (5.6 6 1.9 ng/mL) than in confluent control conditioned medium (0.7 6 0.2 ng/ mL; P 5 0.041). After demonstration of increased TGF-a concentrations in wound-conditioned medium, the physiological relevance of this finding was evaluated. Proliferation of subconfluent IEC-6 cells cultured in the presence of fresh serum-deprived medium or conditioned medium from either wounded or intact confluent monolayers was assessed by measurement of [3H]thymidine incorporation. As shown in Figure 6, IEC-6 cells incubated with wound-conditioned medium had significantly higher [3H]thymidine incorporation rates (10810 6 1621 cpm) compared with IEC-6 cells cultured with controlconditioned medium (4349 6 397 cpm; P 5 0.029) or serum-deprived medium (3924 6 577 cpm). Of note, addition of neutralizing anti–TGF-a antibody to woundconditioned medium substantially blocked the stimulatory effect on IEC-6 cell proliferation (5173 6 457 cpm).

April 1998

Effects of TGF-a on ERK1, ERK2, and Raf-1 Activity in Intestinal Epithelial Cells After showing that neutralizing anti–TGF-a antibody blocked much of the increase in ERK1 and ERK2 kinase activities in wounded IEC-6 cells and unwounded IEC-6 cells incubated with wound-conditioned medium, the effects of TGF-a on ERK1 and ERK2 kinase activities were studied directly using concentrations approximating those observed after monolayer wounding. As shown in Figure 7, stimulation of confluent serum-starved IEC-6 cells with TGF-a at a final concentration of 5 and 10 ng/mL, but not addition of BSA control protein, resulted in a marked increase in ERK1, ERK2, and Raf-1 activities within 5 minutes as assessed by immune complex in vitro kinase assays. The level of activation was comparable with that observed when cells were treated with conditioned medium from wounded but not unwounded monolayers.

A

B

MAPK ACTIVATION IN WOUNDED IEC–6 CELLS

703

A

B

Figure 8. Effects of wounding on JNK1 activity in intestinal epithelial cells. (A) Lysates containing equal amounts of protein were harvested from wounded monolayers (1, 5, 15, 30, and 60 minutes after wounding; w1–w60), and confluent control (con 0) IEC-6 monolayers were immunoprecipitated with anti-JNK1 antibody. Immune complexes were subjected to an in vitro kinase assay using [g-32P]ATP and c-Jun–glutathione-S-transferase fusion protein substrate as described in Materials and Methods. (B) Intact confluent serum-starved IEC-6 monolayers were cultured in the presence of either wound-conditioned (w CM) or control-conditioned (con CM) medium for 5 minutes and harvested. In parallel, a wounded monolayer 5 minutes after injury and confluent monolayer were harvested. JNK1 in vitro kinase activity was assessed as described above.

Effects of Wounding on JNK1 Activity in Intestinal Epithelial Cells

Figure 7. Effect of TGF-a on ERK1, ERK2, and Raf-1 activity in intestinal epithelial cells. Lysates from confluent serum-deprived IEC-6 monolayers were harvested 5 minutes after stimulation with TGF-a (5 or 10 ng/mL) or BSA (dissolved in the same carrier as that for TGF-a). In parallel studies, serum-deprived IEC-6 monolayers were exposed to conditioned medium (CM) obtained from wounded or unwounded monolayers for 5 minutes. The conditioned medium was collected from dishes with or without change in medium (‘‘wash’’ or ‘‘no wash’’). (A) Raf-1 and (B) ERK1 and ERK2 were determined after immunoprecipitation with relevant antibody followed by in vitro kinase assay as described in the text. Activity of ERK1, ERK2, and Raf-1 proteins was determined by an in vitro kinase assay and analyzed as described in Materials and Methods.

After examining the activation of the p44-p42 MAPK pathway, the activity of JNK1, an important stress-activated signal transduction pathway, was assessed in wounded IEC-6 monolayers. Confluent serum-starved cells were wounded and JNK1 activity was evaluated in the time course by an in vitro kinase assay using c-Jun–glutathione-S-transferase fusion protein as a substrate. A marked increase in phosphorylation of c-Jun was present within 1 minute after wounding and peaked after 5 minutes as shown in Figure 8A. To evaluate whether paracrine effects contribute to the rapid increase in p44-p42 and Raf-1 MAPK activity, conditioned medium collected from wounded or intact confluent control monolayers was added to confluent serum-starved IEC-6 cells. As shown in Figure 8B, JNK1 activity was only slightly higher in IEC-6 cells when incubated with woundconditioned medium compared with control-conditioned medium. Addition of control-conditioned medium confluent IEC-6 monolayers also resulted in increased JNK1 activity

704

¨ KE ET AL. GO

compared with parallel monolayers cultured in fresh serum-deprived medium.

Discussion In vivo and in vitro studies have identified growth factors and cytokines that promote repair after wounding of the intestinal epithelium. In the nontransformed rat small intestinal epithelial crypt–derived IEC-6 cell line, TGF-a, EGF, TGF-b, hepatocyte growth factor, acidic and basic fibroblast growth factor, and interleukin 1b have been shown to increase cell migration and/or proliferation.5–7 Many of these factors exert similar effects in other models of wound repair. The studies described in this report show that wounding of intestinal epithelial cells in vitro results in rapid activation of ERK1, ERK2, and JNK1 MAPKs. The observation of increased ERK1 and ERK2 kinase activity is consistent with a recent preliminary report of enhanced phosphorylation of these proteins after intestinal epithelial wounding.27 Because several regulatory peptides that are known to enhance epithelial wound repair also activate ERK1, ERK2, and (to a lesser degree) JNK1 MAPKs,28–30 the present data suggest that these factors might promote wound healing through activation of MAPK cascades. This hypothesis is supported by the finding that wound-conditioned medium activates ERK1, ERK2, and Raf-1 MAPKs in IEC-6 cells, as shown by enhanced incorporation of radiolabeled phosphate into the substrate MBP. In contrast, wound-conditioned medium only slightly increased JNK activity. Taken together, the conditioned medium experiments support the concept that the ERK1 and ERK2 signaling pathways might, at least in part, be activated by paracrine mechanisms. The observed activation of JNK1 after wounding of intestinal epithelial cells does not seem to be mediated through a paracrine mechanism and might be caused by cellular stress during wounding similar to activation of stress-activated protein kinase in mechanically stressed cardiac myocytes.31 Because ERK1 and ERK2 MAPKs can be activated through Raf-1–dependent and Raf-1–independent mechanisms,28 it is noteworthy that wounding of intestinal epithelial monolayers resulted in moderate but consistent increases in Raf-1 activity. This was apparent in lysates obtained from wounded IEC-6 cells and lysates from intact IEC-6 monolayers cultured in the presence of wound-conditioned medium, suggesting that activation of ERK1 and ERK2 MAPKs results from Raf-1 activation. This is consistent with recent observations of Raf-1–dependent activation of ERK1 and ERK2 in mechanically stretched rat cardiac myocytes.32 ERK1/2

GASTROENTEROLOGY Vol. 114, No. 4

activation has also been observed in shear stressstimulated human endothelial cells.33 Previous studies have shown that TGF-a and TGF-b, as well as specific receptors for TGF-a and TGF-b, are expressed by intestinal epithelial cells.14 In this context, it is noteworthy that neutralizing anti–TGF-a but not anti–TGF-b antibodies added to wound-conditioned medium resulted in a diminished increase in kinase activity of both ERK1 and ERK2 proteins in IEC-6 cells when compared with culture in the presence of untreated wound-conditioned medium. Increased production of TGF-a by the wounded monolayers was also found. These observations suggest that activation of ERK1 and ERK2 MAPKs is in part mediated by TGF-a, a peptide growth factor that is not only expressed in IEC-6 cells but has also been shown to promote both intestinal epithelial cell migration and proliferation, important mechanisms for healing of epithelial lesions. The finding that TGF-a activates ERK1 and ERK2 MAPKs in intestinal epithelial cells is consistent with previous observations.29 In addition, the present data suggest that activation of the p44-p42 MAPK pathway is mediated, at least in part, by TGF-a activation of Raf-1. In vivo, restitution can reestablish epithelial continuity within minutes to hours, a much shorter time frame than that needed for cell proliferation. Proliferation of intestinal epithelial cells as a mechanism for mucosal wound repair is thought to begin 12–16 hours after injury and takes one to several days to complete. The finding that neutralizing anti–TGF-a but not anti– TGF-b antibodies inhibit activation of ERK1 and ERK2 MAPKs suggests that TGF-a–mediated stimulation of intestinal epithelial cell proliferation and enhancement of TGF-b–mediated restitution may be regulated by different signaling cascades. Because previous experiments indicated that TGF-a enhances intestinal epithelial restitution through a TGF-b–dependent pathway,6 TGF-a may mediate its stimulatory effects on proliferation and migration of intestinal epithelial cells through different mechanisms. Recently, it has been shown that Ras-like guanosine triphosphatase Rho is required for endogenous and EGF-induced migration of IEC-6 cells.34 The physiological relevance of these findings is shown by the observation that ERK1 and ERK2 MAPK activation is followed by increased proliferation of IEC-6 cells cultured in the presence of wound-conditioned medium compared with conditioned medium from intact confluent monolayers. The concept of TGF-a–mediated proliferation of wounded intestinal epithelial cells is supported by the observation that the stimulatory effect of wound-conditioned medium on IEC-6 cell proliferation was substantially diminished in the presence of

April 1998

neutralizing anti–TGF-a antibody, an effect that was even more pronounced than the inhibitory effect of anti–TGF-a on ERK activity. However, the exact mechanism resulting in increased TGF-a bioactivity remains to be elucidated. A possible explanation is release of surface bound TGF-a from mechanically wounded IEC-6 cells. Growth factor activation of p42-p44 signal transduction cascade may be pivotal for initiation of proliferation, which is essential for mucosal wound repair after epithelial injury in the gastrointestinal tract.

MAPK ACTIVATION IN WOUNDED IEC–6 CELLS

18.

19.

20.

21.

References 1. Feil W, Wenzl E, Vattay P, Starlinger M, Sogukoglu T, Schiessel R. Repair of rabbit duodenal mucosa after acid injury in vivo and in vitro. Gastroenterology 1987;92:1973–1986. 2. Lacy ER. Epithelial restitution in the gastrointestinal tract. J Clin Gastroenterol 1988;10(Suppl 1):72–77. 3. McCormack SA, Viar MJ, Johnson LR. Migration of IEC-6 cells: a model for mucosal healing. Am J Physiol 1992;263:G426–G435. 4. Silen W. Gastric mucosal defense and repair. In: Johnson LR, ed. Physiology of the gastrointestinal tract. Volume 2. 2nd ed. New York: Raven, 1987:1055–1069. 5. Ciacci C, Lind SE, Podolsky DK. Transforming growth factor b regulation of migration in wounded rat intestinal epithelial monolayers. Gastroenterology 1993;105:93–101. 6. Dignass AU, Podolsky DK. Cytokine modulation of intestinal epithelial cell restitution: central role of transforming growth factor b. Gastroenterology 1993;105:1323–1332. 7. Dignass AU, Tsunekawa S, Podolsky DK. Fibroblast growth factors modulate intestinal epithelial cell growth and migration. Gastroenterology 1994;106:1254–1262. 8. Dignass A, Lynch-Devaney K, Kindon H, Thim L, Podolsky DK. Trefoil peptides promote epithelial migration through a transforming growth factor b–independent pathway. J Clin Invest 1994;94: 376–383. 9. Basson MD, Modlin IM, Madri JA. Human enterocyte (Caco-2) migration is modulated in vitro by extracellular matrix composition and epidermal growth factor. J Clin Invest 1992;90:15–23. 10. Yanaka A, Muto H, Fukutomi H, Ito S, Silen W. Role of transforming growth factor b in the restitution of injured guinea pig gastric mucosa in vitro. Am J Physiol 1996;271:G75–G85. 11. 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 1996;111:28–36. 12. Go¨ke M, Podolsky DK. Regulation of the mucosal epithelial barrier. Baillieres Clin Gastroenterol 1996;10:393–405. 13. Dieckgraefe BK, Stenson WF, Alpers DH. Gastrointestinal epithelial response to injury. Curr Opin Gastroenterol 1996;12:109– 114. 14. Koyama S, Podolsky DK. Differential expression of transforming growth factors a and b in rat intestinal epithelial cells. J Clin Invest 1989;83:1768–1773. 15. Boulton TG, Nye SH, Robins DJ, Ip NY, Radziejewska E, Morgenbesser SD, DePinho RA, Panayotatos N, Cobb MH, Yancopoulos G. ERKs: a family of protein-serine/threonine kinases that are activated and tyrosine phosphorylated in response to insulin and NGF. Cell 1991;65:663–675. 16. Bogoyevitch MA, Ketterman AJ, Sugden PH. Cellular stresses differentially activate c-Jun-N-terminal protein kinases and extracellular signal-regulated protein kinases in cultured ventricular myocytes. J Biol Chem 1995;270:29710–29717. 17. Sanchez I, Hughes RT, Mayer BJ, Yee K, Woodgett JR, Avruch J, Kyriakis JM, Zon LI. Role of SAPK/ERK kinase-1 in the stress-

22. 23.

24. 25.

26.

27.

28. 29.

30.

31.

32.

33.

34.

705

activated pathway regulating transcription factor c-Jun. Nature 1994;72:794–798. Derijard B, Hibi M, Wu IH, Barrett T, Su B, Deng T, Karin M, Davis RJ. JNK1: a protein kinase stimulated by UV light and Ha-Ras that binds and phosphorylates the c-Jun activation domain. Cell 1994;76:1025–1037. Kyriakis JM, Banerjee P, Nikolakaki E, Dai T, Rubie EA, Ahmad MF, Avruch JA, Woodgett JR. The stress-activated protein kinase subfamily of c-Jun kinases. Nature 1994;369:156–160. Rouse J, Cohen P, Trigon S, Morange M, Alonso-Liamazares A, Zamanillo D, Hunt T, Nebreda AR. A novel kinase cascade triggered by stress and heat shock that stimulates MAPKAP kinase-2 and phosphorylation of the small heat shock proteins. Cell 1994;78:1027–1037. Han J, Lee JD, Bibbs L, Ulevitch RJ. A MAP kinase targeted by endotoxin and hyperosmolarity in mammalian cells. Science 1994;265:808–811. Quaroni A, Wands J, Trelstad RL, Isselbacher KJ. Epithelioid cell cultures from rat small intestine. J Cell Biol 1979;80:248–265. Go¨ke M, Zuk A, Podolsky DK. Regulation and function of extracellular matrix in intestinal epithelial restitution in vitro. Am J Physiol 1996;271:G729–G740. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227:680–685. Mucsi I, Skorecki KL, Goldberg HJ. Extracellular signal-regulated kinase and the small GTP-binding protein, Rac, contribute to the effects of transforming growth factor b1 on gene expression. J Biol Chem 1996;271:16567–16572. Mendenhall W. Nonparametric statistics. In: Mendenhall W, ed. Introduction to probability and statistics. Boston: PWS Publishers, 1987:727–780. Dieckgraefe BK, Weems DM, Santoro SA, Alpers DH. Erk and p38 MAP kinase pathways are mediators of intestinal epithelial wound–induced signal transduction. Biochem Biophys Res Comm 1997;233:389–394. Cano E, Mahadevan LC. Parallel signal processing among mammalian MAPKs. Trends Biochem Sci 1995;20:117–122. Oliver BL, Sha’Afi RI, Hajjar JJ. Transforming growth factor a and epidermal growth factor activate mitogen-activated protein kinase and its substrates in intestinal epithelial cells. Proc Soc Exp Biol Med 1995;210:162–170. Brunet A, Pouyssegur J. Identification of MAP kinase domains by redirecting stress signal into growth factor responses. Science 1996;272:1652–1655. Komuro I, Kudo S, Yamazaki T, Zou Y, Shiojima I, Yazaki Y. Mechanical stretch activates the stress-activated protein kinases in cardiac myocytes. FASEB J 1996;10:631–636. Yamazaki T, Komuro I, Kudo S, Zou Y, Shiojima I, Mizuno T, Takano H, Hiroi Y, Ueki K, Tobe K, Kadowaki T, Nagai R, Yazaki Y. Mechanical stress activates protein kinase cascade of phosphorylation in neonatal rat cardiac myocytes. J Clin Invest 1995;96: 438–446. Takahashi M, Berk BC. Mitogen-activated protein kinase (ERK1/2) activation by shear stress and adhesion in endothelial cells. J Clin Invest 1996;99:2623–2631. Santos FM, McCormack SA, Guo Z, Okollcany J, Zheng Y, Johnson LR. Rho proteins play a critical role in cell migration during the early phase of mucosal restitution. J Clin Invest 1997;100:216– 225.

Received March 6, 1997. Accepted December 23, 1997. Address requests for reprints to: Daniel K. Podolsky, M.D., Gastrointestinal Unit, GRJ-719, Massachusetts General Hospital, 55 Fruit Street, Boston, Massachusetts 02114. e-mail: Podolsky.Daniel@ mgh.harvard.edu; fax: (617) 724-2136. Supported by grants DK 41557 and DK 43351 from the National Institutes of Health (to D.K.P.) and the Else Kro ¨ner-FreseniusStiftung (to M.G.).