| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 280, Issue 36, 31936-31948, September 9, 2005
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
-ARRESTINS*
¶||
From the
Departments of Surgery and Physiology, University of California, San Francisco, California 94143 and the ¶Intestinal Disease Research Program, McMaster University, Hamilton, Ontario L8S 4L8, Canada
Received for publication, June 10, 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
-arrestins, which recruit ERK1/2 to PAR2 in endosomes and retain ERK1/2 in the cytosol, on PAR2-mediated alterations in permeability. An ERK1/2 inhibitor abolished the effects of PAR2 agonist on permeability and redistribution of F-actin. Down-regulation of -arrestins with small interfering RNA inhibited PAR2-induced activation of ERK1/2 and suppressed PAR2-induced changes in permeability. Thus, mast cells signal to colonocytes in a paracrine manner by release of tryptase and activation of PAR2. PAR2 couples to -arrestin-dependent activation of ERK1/2, which regulates reorganization of perijunctional F-actin to increase epithelial permeability. These mechanisms may explain the increased epithelial permeability of the intestine during stress and inflammation. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
Mast cells control permeability of the intestinal epithelium by unknown mechanisms. Mast cells infiltrate the inflamed intestine, and the increased permeability of the intestinal epithelium during chronic stress, allergic inflammation, parasitic infection, and chemically induced inflammation is reduced in mast cell-deficient animals and by mast cell stabilizers (5, 6, 9-11). However, the mast cell products that mediate this effect and their mechanism of action are unknown.
Mast cells are replete with proteases that may induce alterations in epithelial permeability. The major protease of human mast cells is tryptase (12). Upon release, tryptase may remain active in tissues for prolonged periods as it is poorly diffusible and resistant to endogenous inhibitors (12). Antigen challenge of sensitized subjects promotes release of mast cell proteases into the intestinal lumen and vasculature, which correlates with increased permeability (13, 14). In the inflamed intestine there is also increased trypsin activity and activation of coagulation proteases (15-18). However, the effects of these proteases on permeability of the intestinal epithelium and their mechanisms of action are unknown.
Proteases can signal to cells by cleaving protease-activated receptors (PARs), a family of four G-protein-coupled receptors (reviewed in Ref. 19). PAR2 is expressed at the apical and basolateral membrane of the intestinal epithelial cells (20). Tryptase, trypsins, and coagulation factors VIIa and Xa cleave PAR2 within the extracellular N terminus at Ser-Lys-Gly-Arg-Ser34
Leu-Ile-Gly-Lys-Val (human, = cleavage site) to expose a tethered ligand domain (SLIGKV) that binds to and activates the cleaved receptor (21-25). Injection of trypsin and PAR2-activating peptide (AP, corresponding to the tethered ligand) into the intestinal lumen increases epithelial permeability and promotes infiltration of bacteria (26-28). However, the mechanisms by which PAR2 activation couples to assembly of TJs and regulates interactions of TJ proteins with the cytoskeleton are unknown. In enterocytes, PAR2 interacts with the multi-adaptor protein -arrestin (ARR), which mediates receptor endocytosis and recruits activated extracellular signal-regulated kinases 1/2 (ERK1/2) to endosomes (29). Cytosolically retained ERK1/2 may regulate the cytoskeleton to control paracellular permeability (30, 31).
We hypothesized that mast cells signal to colonocytes by release of tryptase and activation of PAR2, and that PAR2 couples to ERK1/2 by a ARR-dependent mechanism to regulate TJ assembly, perijunctional filamentous actin (F-actin), and paracellular permeability. Our goals were as follows: 1) to examine expression of PAR2 by colonocytes; 2) to investigate the effects of PAR2 activation on assembly of TJs and permeability; 3) to determine whether mast cells signal to colonocytes to increase permeability by release of tryptase and activation of PAR2; and 4) to investigate the role of ERK1/2 and ARR in this process.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
Antibodies—Rabbit antibody to human IgE was from Bethyl Laboratories (Montgomery, TX); mouse antibody to human tryptase was from Serotec Inc. (Raleigh, NC); rabbit ZO-1, mouse occludin, and mouse claudin1 antibodies were from Zymed Laboratories Inc.; mouse E-cadherin antibody was from Upstate%20Biotechnology">Upstate Biotechnology, Inc. (Lake Placid, NY). Antibodies to phosphorylated ERK1/2 (pERK1/2) and ERK2 were from Santa Cruz Biotechnology (Santa Cruz, CA). ARR1 antibody was from BD Transduction Laboratories, and ARR2 antibody was from R. Lefkowitz (Duke University). Antibody to 
-tubulin was from Convince (Berkeley, CA). Mouse primary antibody isotype control and chromatographically purified rabbit IgG were from Zymed Laboratories Inc., and rabbit IgG isotype control was from Bethyl Laboratories. Fluorescein isothiocyanate- and peroxidase-conjugated goat anti-rabbit and anti-mouse IgG were from Jackson ImmunoResearch (West Grove, PA). Texas Red-, Alexa 488-, and phycoerythrin-conjugated goat anti-rabbit and anti-mouse IgG were from Molecular Probes (Eugene, OR).
Colonic Epithelial Cells—T84 colonic epithelial cells (American Type Culture Collection, Manassas, VA) were cultured in Dulbecco's modified Eagle's medium/F-12 with 5% FCS, 14 mM sodium bicarbonate, 15 mM HEPES, 100 units/ml penicillin, and 100 µg/ml streptomycin (5% CO2, 37 °C). NCM460 cells (R. Benya, University of Illinois), a nontransformed human colonocyte line (35), were cultured in Ham's F-12 medium with 20% FCS, 0.5 units/ml insulin-transferrin, 0.4 µg/ml hydrocortisone, and 0.58 µg/ml glutamine (5% CO2, 37 °C). T84 and NCM460 cells were plated on plastic for RT-PCR and ERK1/2 assays and on glass coverslips for measurement of [Ca2+]i. T84 cells were plated on transwell polycarbonate filters (0.4 µm pore, 4.7 cm2, 0.5-1.0 x 106 cells/well) for study of permeability and redistribution of junctional proteins (7). Studies commenced when transepithelial resistance (TER) exceeded 1,000 
·cm-2, determined using Millicell ERS apparatus (Millipore, Bedford, MA). When used, T84 cells were treated with the 10 µM MEK inhibitor UO126 (Calbiochem) for 30 min prior to adding AP or RP, and TER and F-actin localizations were assessed after 24 h.
Mast Cells—Cultured human mast cells (CHMC, Rigel Pharmaceuticals, South San Francisco, CA) were derived from CD34+ precursor cells from umbilical cord blood. Cells were cultured for 5 days in Stem-Pro medium (Invitrogen) containing 200 ng/ml stem cell factor, 20 ng/ml flt3 ligand, 200 ng/ml interleukin 6, and 100 ng/ml IgE (all from Cortex Biochem. Inc., San Leandro, CA) to induce differentiation into mucosal type mast cells containing tryptase and not chymase. The human mast cell line HMC1 (J. H. Butterfield, Mayo Clinic, Rochester, MN (36)) was cultured in Iscove's modified Dulbecco's medium containing 10% iron-supplemented FCS, 1.2 mM 
-thioglycerol, 100 units/ml penicillin, and 100 µg/ml streptomycin (5% CO2, 37 °C). To induce degranulation, CHMC (5 x 106 cells/ml HBSS, 0.1% BSA, 20 mM HEPES, 25 µg/ml heparin, pH 7.4) were challenged with rabbit IgE antibody (1:1,000, 30 min, 37 °C) or vehicle (control). HMC1 (106 cells/ml Dulbecco's modified Eagle's medium) were challenged with compound 48/80 (5 µg/ml, 15 min, 37 °C) to induce degranulation. After treatments, CHMC and HMC1 were pelleted (14,000 x g, 5 min, 4 °C), and supernatants were assayed for tryptase and effects on [Ca2+]i and permeability of colonocytes.
Co-culture of T84 and HMC1 Cells—T84 cells (106 cells/well) were cultured on filters until confluent. HMC1 cells were added to the basal compartment (106 cells in 1.5 ml). Cells were co-cultured in Iscove's modified Dulbecco's medium. To degranulate mast cells, compound 48/80 (5 µg/ml) was added to the basal compartment. Permeability of the T84 monolayers was determined after 24 h.
RT-PCR—Total RNA from T84 and NCM460 cells was reverse-transcribed. PCRs used primers specific for PAR2 (forward, 5'-ccctttgtatgtcgtgaagcagac-3'; reverse, 5'-ttcctggagtgtttctttgaggtg-3'), trypsinogens (forward, 5'-gccaagcttcgcgtgtccaccatctgtgtcg-3'; reverse, 5'-cagcggccgcttagctgttggcagctatggtg-3'), ARR1 (forward, 5'-ggacaagaagcccctgacgcg-3'; reverse, 5'-agcctcttccatggcaacagg-3'), and ARR2 (forward, 5'-tcaactcgaacaagatgacca-3'; reverse 5'-gtggcatagttggtatcaaat-3'). Reverse transcriptase was omitted from controls. PCR products were separated by electrophoresis and sequenced.
Tryptase Assay—Samples were incubated with 200 µl of 50 mM Tris/HCl, pH 7.6, 120 mM NaCl, 20 µg/ml heparin containing 0.5 mM tosyl-Gly-Pro-Arg-p-nitroanilide (Sigma) for 15 min at 37 °C. Absorbance was measured at 405 nm to assess cleavage. Inhibitors were preincubated with samples for 15 min on ice before assay. Results are expressed as nanomoles of cleaved substrate/min/106 cells.
Measurement of [Ca2+]i—T84 and NCM460 cells were incubated in HBSS with 0.1% BSA, 20 mM HEPES, pH 7.4, and 2.5-5 µM Fura-2AM (Molecular Probes, Eugene, OR) for 30-45 min at 37 °C. Coverslips were mounted in an open chamber (500 µl volume) in HBSS with 0.1% BSA, 20 mM HEPES, pH 7.4, at 37 °C, and fluorescence of individual cells was measured at 340 and 380 nm excitation and 510 nm emission (22). A Fura-2 calcium imaging calibration kit (Molecular Probes) was used to calculate [Ca2+]i according to the manufacturer's directions. A standard curve was generated (0.017-39.8 µM) and used to convert the emission ratio at 340/380 nm excitation to estimate the free [Ca2+]i.AP or proteases were added directly to the chamber in a volume of 50 µl, and maximal increases in [Ca2+]i were determined. In assays of mast cell products, the supernatant was obtained from 5 x 106 CHMC or HMC1, and 50 µl was added to the chamber to a final volume of 500 µl. When used, cells were preincubated with RWJ337842 (10 µM) for 2 min. To assess the contribution of extracellular Ca2+ ions to changes in [Ca2+]i, cells were incubated in Ca2+-free HBSS with 0.1% BSA, 20 mM HEPES, pH 7.4. For concentration-response analyses, cells were exposed to a single concentration of agonists and then 5 µM ionomycin. Results were normalized and are expressed as percent responses to 5 µM ionomycin. The EC50 was calculated by using Origin Microcal software (Northampton, MA). For desensitization experiments, cells were preincubated with proteases or vehicle (control) for 2-5 min and then challenged with AP, without an intervening wash. Observations were repeated in >4 separate experiments.
TER and Paracellular Flux of Macromolecules—Test substances were added to the apical or basal compartments of transwell with confluent T84 cells on filters. TER was determined as described at 0-48 h. Results are expressed as ·cm-2 and are normalized as percentage of base-line values. To measure paracellular flux of macromolecules, horseradish peroxidase (HRP, 44 kDa, type II, Sigma, 0.01 mM) was added to the apical compartment at 24 h. After 2 h, samples were obtained from medium in the apical and basolateral compartments and assayed for HRP using a kinetic enzymatic assay as described (9, 10). HRP flux is expressed as percentage of apically added HRP recovered in basolateral compartment over 2 h. Measurements of TER and HRP flux were made from four wells per experiment, and experiments were repeated four times.
Immunofluorescence—T84 cells grown to confluence on filters were fixed in 4% paraformaldehyde in 100 mM PBS, pH 7.4, for 20 min at 4 °C. Cells were washed and incubated in PBS with 5% normal goat serum, 0.3% Triton X-100 with primary antibodies to junctional proteins: rabbit ZO-1, mouse occludin, and mouse claudin1 (1:50-200, overnight, 4 °C). Filters were washed and incubated with fluorescein isothiocyanate-conjugated goat anti-IgG or anti-mouse (1:200, 2 h, room temperature). To detect F-actin, fixed, permeabilized cells were incubated with Texas Red-X-phalloidin (5 units/ml, Molecular Probes) for 20 min at room temperature. Cells were mounted in Prolong with DAPI (Molecular Probes). To localize IgE receptor and tryptase in CHMC, fixed and permeabilized cells were incubated with rabbit anti-human IgE (2 µg/ml) and mouse anti-human tryptase (2 µg/ml) antibodies (overnight, 4 °C), washed, incubated with Texas Red- and Alexa 488-conjugated goat anti-rabbit and mouse IgG (1:100, 2 h, room temperature), and mounted. In control experiments, the primary antibodies were replaced with isotype control antibodies at the same dilutions.
|
|
Flow Cytometry—CHMC, which had been cultured with 100 ng/ml IgE for 5 days, were suspended in 100 mM PBS, pH 7.4, containing 2% fetal bovine serum (0.5 x 106 cells/ml). Cells were incubated with rabbit anti-human IgE antibody (5 µg/ml) or rabbit IgG isotype control (5 µg/ml) for 15 min at 4 °C. Cells were washed and incubated with phycoerythrin-conjugated goat anti-rabbit IgG (2 µg/ml) for 15 min at 4 °C. Cells were washed and suspended in 1 ml of PBS/fetal bovine serum. Cells were analyzed with a FACSCalibur flow cytometer (BD Biosciences). Fluorophores were excited at 488 nm, and emission was collected at 575/25 nm for phycoerythrin.
Cellular Fractionation and Western Blotting—T84 cells were lysed in 1% Triton X-100, 100 mM NaCl, 10 mM HEPES, pH 7.6, 2 mM EDTA with protease inhibitors (Roche Applied Science) and passed 10 times through a 21-gauge needle. Lysates were centrifuged (15,000 x g, 30 min, 4 °C), and the supernatant formed the Triton-soluble or cytosolic fraction. The pellet was ultrasonicated in Triton buffer containing 1% SDS and cleared (15,000 x g, 5 min, 4 °C) to give the Triton-insoluble or TJ fraction. Samples (5-20 µg of protein determined by Bradford method) were fractionated on 7-15% SDS-polyacrylamide gel and transferred to PVDF membranes (Millipore, Bedford, MA). Membranes were blocked in Tris-buffered saline, 0.1% Tween 20, 5% nonfat dry milk for 1 h, and blots were incubated with primary antibodies for 2 h at room temperature: Z-O1, occludin, claudin-1, and E-cadherin (1:100-1:200). Membranes were washed and incubated with goat anti-mouse or -rabbit IgG conjugated to peroxidase (1 h, room temperature), and proteins were detected by chemiluminescence (ECL Plus, Amersham Biosciences). Experiments were repeated four times.
ERK1/2 Assays—T84 cells were serum-deprived for 12-16 h. Cells were lysed with 100 µl of RIPA (10 mM Tris/HCl, 200 mM NaCl, 50 mM NaF, 1 mM NaVO4, 0.5% deoxycholate, 0.1% SDS, 1% Nonidet P-40, 1 mM EDTA, pH 7.4) with protease inhibitors. Samples (5-20 µg of protein determined by Bradford method) were fractionated by 10% Tris-glycine SDS-polyacrylamide gels and transferred to PVDF membranes. Membranes were incubated with antibody to pERK1/2 (1:5,000 overnight, 4 °C) (29). Membranes were stripped and reblotted for ERK2 (1:5,000 overnight, 4 °C). Signals were detected using peroxidase-conjugated secondary antibodies and chemiluminescence. Experiments were repeated four times.
ARR siRNA—siRNA duplexes of 21-nucleotide sense and 21-nucleotide antisense strands with 2-nucleotide 3'-dTdT overhangs were designed to target rat ARR1 (GenBankTM accession number NM_012910
[GenBank]
) and ARR2 (GenBankTM accession number NM_012911
[GenBank]
.1) by searching for the motif AA(N19) (where N indicates any nucleotide) and selecting unique sequences (determined by blast of NCBI data base) with 
50% G/C content (Dharmacon Research, Lafayette, CO). Chosen sequences were ARR1 5'-gagcgacucaucaagaagcdTdT-3' (304-322 from start codon) and ARR2 5'-cucaagcacgaagacaccadTdT-3' (880-898 from start codon), and control siRNA 5'-aacgaagcaacuaagcucgdTdT-3' did not target any gene. T84 cells (30-50% confluence for ARR Western blotting and ERK1/2 assays and TER >1,000 ·cm-2 for permeability assays) were transfected with siRNA (5 µl of 20 µM stock added apically) using Oligofectamine (Invitrogen) according to the manufacturer's directions. For ARR Western blotting, cells were lysed at various times after transfection with 100 µl of RIPA (10 mM Tris/HCl, 200 mM NaCl, 50 mM NaF, 1 mM NaVO4, 0.5% deoxycholate, 0.1% SDS, 1% Nonidet P-40, 1 mM EDTA, pH 7.4) with protease inhibitors. Samples (5-20 µg of protein determined by Bradford method) were fractionated by 10% Tris-glycine SDS-polyacrylamide gels and transferred to PVDF membranes. ARR1 and -2 were detected by Western blotting of whole cell lysates using antibodies to ARR1 (1:300, overnight, 4 °C) or ARR2 (1:2,500, overnight, 4 °C). Membranes were stripped and re-blotted for -tubulin (1:3,000, overnight, 4 °C). Signals were detected using peroxidase-conjugated secondary antibodies and chemiluminescence. Experiments were repeated 3-4 times. The effects of AP on activation of ERK1/2 and on TER and HRP flux were examined 3-4 days after transfection.
|
|
Statistics—Results are expressed as mean ± S.E. and are compared by analysis of variance and Student-Newman-Keul's test, with p < 0.05 considered significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
To determine whether colonocytes express functional PAR2, we measured changes in [Ca2+]i in response to PAR2 agonists. The basal [Ca2+]i was 4.63 ± 0.80 nM in NCM460 cells (n = 23 cells) and 5.32 ± 0.95 nM in T84 cells (n = 27 cells, n > 4 experiments). Trypsin (10 nM) increased [Ca2+]i in NCM460 cells by 3.10 ± 0.63 nM (n = 10 cells) and T84 cells by 3.72 ± 1.11 nM (n = 8 cells) above basal values (Fig. 2a). AP (100 µM) increased [Ca2+]i in NCM460 cells by 4.31 ± 0.61 nM (n = 13 cells) and T84 cells by 4.49 ± 0.52 nM (n = 19 cells) above basal values (Fig. 2b). RP (100 µM) had no effect on [Ca2+]i in NCM460 or T84 cells, confirming specificity of AP (Fig. 2c).
|
|
Because trypsin may signal by cleaving receptors other than PAR2, we determined whether preincubation with trypsin desensitized responses to AP. Cells were incubated with 10 or 100 nM trypsin or vehicle (control) for 5 min and challenged with 100 µM AP. Preincubation with trypsin markedly desensitized responses to AP. When compared with control cells, responses to AP in NCM460 were reduced by 25 and 83%, after trypsin 10 and 100 nM, respectively. The response to AP in T84 was reduced by >70% after both concentrations or trypsin (Fig. 3b).
Together, these results indicate that colonocytes express PAR2 and respond to the PAR2 agonists AP and trypsin, and trypsin increases [Ca2+]i by cleaving PAR2.
Mast Cells Signal to Colonocytes by Tryptase and PAR2—Tryptase cleaves and activates PAR2 on many cell types, including intestinal epithelial cells (19, 21, 22).2 However, it is not known if mast cells signal to colonocytes by release of tryptase and activation of PAR2 or by other mechanisms. Therefore, we determined whether mast cells release tryptase upon stimulation and whether a supernatant from degranulated mast cells signals to colonocytes by tryptase activation of PAR2.
The mast cell line HMC1 is thoroughly characterized and contains and releases tryptase upon stimulation (e.g. with compound 48/80) (22, 36). We also characterized CHMC and determined whether they contain tryptase and release the protease after appropriate stimulation. When CHMC were cultured with IgE, immunoreactive IgE was detected at the plasma membrane by immunofluorescence and confocal microscopy (Fig. 4a, arrowheads). Immunoreactive tryptase was detected in granules that were densely packed in the cytosol (Fig. 4a, arrows). There was no staining in cells incubated with isotype control antibodies (Fig. 4a). The surface expression of IgE receptor was also confirmed by flow cytometry, which revealed specific surface binding of anti-IgE but not of isotype control antibody (Fig. 4b). Thus, CHMC bind IgE and therefore express IgE receptors and also contain tryptase in secretory granules. We cross-linked IgE receptors by incubating CHMC with IgE followed by IgE antibody (anti-IgE), and we measured enzymatic activity in the supernatant from degranulated cells using the tryptase substrate tosyl-Gly-Pro-Arg-p-nitroanilide. Incubation of CHMC with anti-IgE for 30 min induced an 10-fold increase in enzymatic activity in the supernatant over vehicle-treated cells (Fig. 4c). This activity was strongly suppressed by the tryptase inhibitors BABIM (10 µM) and RWJ337842 (10 µM) but was unaffected by SBTI (50 µg/ml), which does not inhibit tryptase but does inhibit trypsins that can also cleave this substrate. Thus, CHMC, like HMC1, contain and release tryptase.
We also determined whether supernatant from degranulated mast cells increased [Ca2+]i in colonocytes by tryptase activation of PAR2. Supernatant from undegranulated CHMC (vehicle-treated) had a small effect on [Ca2+]i in NCM460 and T84 cells, but cells still responded in the expected manner to AP (Fig. 5a, upper panels). Supernatant from degranulated CHMC (anti-IgE-treated) induced a rapid increase in [Ca2+]i in NCM460 and T84 cells (Fig. 5a, middle panels). Addition of the tryptase inhibitor RWJ337842 (10 µM, injected 2 min before supernatant) markedly inhibited the effects of supernatant from degranulated CHMC by 50% in both NCM460 and T84 cells (Fig. 5, a, lower panels, and b). The effects of BABIM on this response could not be examined because BABIM or a metabolite interfered with measurement of [Ca2+]i (results not shown). These results suggest that tryptase is a major component of the supernatant obtained from degranulated mast cells that signals to colonocytes. To determine whether the effect of supernatant from degranulated CHMC on [Ca2+]i was because of activation of PAR2, we assessed desensitization of responses to AP. Cells were either untreated (control), preincubated with supernatant from undegranulated CHMC, or degranulated CHMC for 2 min and then challenged with AP (100 µM). Preincubation with supernatant from undegranulated CHMC did not affect AP-induced increases in [Ca2+]i in colonocytes (Fig. 5, a, upper panels, and c), whereas supernatant from degranulated CHMC desensitized the response to AP by 45% in NCM460 and 61% in T84 cells (Fig. 5, a, middle panels, and c). These results suggest that the effects of the supernatant on [Ca2+]i are mediated in part by activation of PAR2. We similarly determined whether the stable mast cell line HMC1 could signal to colonocytes through release of tryptase and activation of PAR2. Compound 48/80 degranulated HMC1 to release tryptase-like activity that was inhibited by RWJ337842 and BABIM but not SBTI (not shown). Also, supernatant from degranulated HMC1 increased [Ca2+]i in NCM460 cells (Fig. 5d), and RWJ337842 inhibited this effect by >50% (Fig. 5e).
|
PAR2 Activation Increases Paracellular Permeability—To assess the effects of PAR2 activation on permeability of T84 cells, we measured TER and transepithelial flux of HRP. AP and trypsin applied to the basolateral membrane induced a time-dependent decrease in TER within 2 h that was maximal at 24 h (100 µM AP, 55.2 ± 2.7% basal; 10 nM trypsin, 61.3 ± 3.5% basal) and sustained for 48 h (Fig. 6a for AP, trypsin not shown). RP (Fig. 6a) and heat-inactivated trypsin (not shown) were ineffective. AP and trypsin applied to the apical membrane of T84 cells induced a smaller decline in TER (24 h: 100 µM AP, 78.3 ± 2.8% basal; 10 nM trypsin, 76.0 ± 3.5% basal). Thus, in subsequent experiments agonists were applied to the basolateral membrane. AP induced a concentration-dependent decrease in TER at 24 h (Fig. 6b) and an increase in HRP flux from 24 to 26 h (Fig. 6c). Thus, activation of PAR2 at the basolateral membrane of T84 cells increases permeability.
PAR2 Activation Induces Redistribution of TJ Proteins and F-actin—Alterations in the localization and expression of TJ proteins and perijunctional F-actin are associated with changes in paracellular permeability of epithelial cells. Therefore, we examined the effects of PAR2 activation on localization of ZO-1, occludin, claudin-1, and F-actin in T84 cells by confocal microscopy. Optical sections in the x-y plane revealed that immunoreactive ZO-1 (Fig. 7a), occludin (Fig. 7b), and claudin-1 (Fig. 7c) were prominently localized to the perimeter at the apical pole in cells treated with RP (100 µM, 24 h). Projections in the z plane of stacks of optical sections confirmed the prominent localization of ZO-1, occludin, and claudin-1 in the apical region of the cell, consistent with the presence of these proteins in TJs. Lower levels of these proteins were detected in the basolateral membrane beneath the tight junctions and in the cytosol. Similar distributions were observed in cells treated with vehicle (not shown). In cells treated with AP (100 µM, 24 h), there was diminished intensity of staining for ZO-1 and occludin in the apical TJ region of cells. Both ZO-1 and occludin were more diffusely distributed, although it was not possible to discern whether there was diminished expression or redistribution of these proteins. The distribution of claudin-1 was not discernibly altered by exposure to AP. Staining was not observed when primary antibodies were replaced with isotype control antibodies, confirming specificity (Fig. 7d). F-actin was prominently detected in ring-like perijunctional structures in the perimeter of the apical pole of cells treated with RP (Fig. 7e) or vehicle (not shown). Exposure to AP resulted in a marked redistribution of F-actin from the perijunctional ring to particles within the cytosol (Fig. 7e).
|
Mast Cell Tryptase Increases Paracellular Permeability—To determine whether mediators from degranulated mast cells regulate permeability, we exposed the basolateral surface of T84 cells to supernatant from HMC1 treated with compound 48/80. The supernatant induced a decrease in TER that was detected within 1 h after stimulation and that was maximal after 4 h (30% reduction) (Fig. 9a). This effect was sustained and detected after 24 and 48 h from the initial stimulation. HMC1 supernatant also stimulated HRP flux, and after 24 h the HRP flux was elevated 4-fold in supernatant-treated cells (Fig. 9b). The tryptase inhibitor BABIM strongly inhibited the effects of supernatant on both TER and HRP flux, suggesting a prominent contribution of tryptase to the increased permeability (Fig. 9, a and b). However, BABIM did not affect TER or HRP flux in untreated T84 cells.
To examine paracrine regulation of colonocytes by mast cells, we co-cultured T84 cells on transwell filters with HMC1 cells in the basal compartment. The co-culture of HMC1 with T84 cells did not alter TER or HRP flux measured after 24 h (Fig. 9, c and d). However, addition of compound 48/80 to the basal compartment resulted in a 20% decline in TER and a 2-fold increase in HRP flux after 24 h, and BABIM strongly inhibited these effects. Compound 48/80 added to the basal compartment also increased tryptase activity in the medium (7.1 ± 2.3 nmol/min/106 cells after 24 h). However, compound 48/80 had no effect on TER or HRP flux of T84 cells culture in the absence of HMC1 cells. Thus, degranulated mast cells can signal to colonocytes in a paracrine manner by release of tryptase and probable activation of PAR2 at the basolateral membrane to increase permeability.
ARRs and ERK1/2 Mediate PAR2-induced Paracellular Permeability—Multiple signaling pathways regulate the assembly of TJs and the distribution of perijunctional F-actin to control permeability, including those that activate ERK1/2 (31). Because PAR2 couples to activation of ERK1/2 (29, 30), we hypothesized that PAR2 regulates permeability by an ERK1/2-mediated mechanism. AP (100 µM) induced a rapid phosphorylation and thus activation of ERK1/2 in T84 cells that was maximal after 10-15 min, assessed by Western blotting (Fig. 10a). The MEK1/2 inhibitor UO126 (10 µM) prevented AP-induced phosphorylation of ERK1/2 (not shown). UO126 also strongly inhibited the AP-induced decrease in TER observed after 24 h (Fig. 10b) and prevented AP-stimulated redistribution of perijunctional F-actin (Fig. 10c). Together, these results suggest that ERK1/2 mediates AP-induced increases in permeability.
|
ARRs recruit Raf-1, MEK1/2, and ERK1/2 to activated PAR2 in endosomes (29). This mechanism retains activated ERK1/2 in the cytosol where they may regulate cytosolic targets such as cytoskeletal and junctional proteins (30). To determine the importance of this pathway in PAR2-induced alterations in permeability, we down-regulated ARRs using siRNA. Transcripts corresponding to the predicted size of ARR1 (498 bp) and ARR2 (381 bp) were amplified from T84 cells and identified by sequencing, suggesting that T84 cells express both isoforms of ARRs (Fig. 11a). Because PAR2 can interact with ARR1 and ARR2 (37), we used siRNA to down-regulate both isoforms in T84 cells. siRNA to ARR1 and -2 down-regulated both ARRs by >50% within 3-4 days as determined by Western blotting (Fig. 11b) and strongly reduced mRNA levels as determined by RT-PCR (not shown). Control siRNA had no effect on expression of ARRs. Down-regulation of ARR1 and -2 resulted in a >2-fold inhibition of AP-induced activation of ERK1/2 (Fig. 11c).
To determine the importance of ARR-mediated activation of ERK1/2 to PAR2-induced changes in permeability, we examined the effects of AP on TER and HRP flux. Oligofectamine or control siRNA had no effect on AP-induced inhibition of TER or stimulation of HRP flux (Fig. 12, a and b). However, siRNA to ARR1 and -2 strongly inhibited the AP-induced decrease in TER and increase in HRP flux at 24 h (Fig. 12, a and b). Thus, ARRs are required for PAR2-induced activation of ERK1/2 and increased permeability.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
ARR- and ERK1/2-dependent process. This mechanism increases paracellular permeability to macromolecules and may explain the critical role of mast cells in the increased permeability of the intestine during stress and inflammation (5, 6, 9-11). In this manner, proteases and PAR2 play a critical role in disrupting the intestinal barrier and promoting the ingress of macromolecules and bacteria that cause inflammation and infection (26-28).
|
1,000-fold greater than the potency of AP. Agonists had similar potencies in both NCM460 and T84 cells that are comparable with potencies reported for these agonists in transfected cells, enterocytes, and tumor cell lines (38). Moreover, preincubation of cells with trypsin strongly desensitized responses to AP, as we have observed previously in transfected cells and enterocytes (39). Together, these results suggest that trypsin and AP increased [Ca2+]i in colonocytes by activating PAR2. However, we cannot exclude the possibility that trypsin also activates other receptors in these cells, such as PAR4 (19). In support of our results, other intestinal epithelial cell lines also express PAR2 and respond to trypsin and AP (20, 29, 40).
Supernatants from two mast cell lines degranulated by cross-linking IgE receptors or by treatment with compound 48/80 triggered increased [Ca2+]i in NCM460 and T84 cells. The supernatants contained enzymatic activity that was suppressed by two tryptase inhibitors, but unaffected by a trypsin-selective inhibitor, and is thus likely to be tryptase. Moreover, a tryptase inhibitor reduced the effects of the supernatant on the [Ca2+]i by 50%. Thus, tryptase plays a major role in mediating the effects of mast cell supernatant on [Ca2+]i in colonocytes. Other factors released by degranulated mast cells, such histamine, serotonin, and prostaglandins, may also contribute to the activity of the supernatant that was unaffected by the tryptase inhibitor. Preincubation of colonocytes with mast cell supernatant reduced the responses to AP by 50%, suggesting that tryptase in the supernatant cleaved PAR2, resulting in homologous desensitization of this receptor. In support of these results, we have reported previously that supernatants from degranulated mast cells increase [Ca2+]i in colonic myocytes and ileal myenteric neurons by a tryptase-dependent mechanism and desensitize responses of these cells to PAR2-selective AP (22, 41). However, the observation that supernatants desensitized responses to AP does not unequivocally prove activation of PAR2, because certain mast cell products may activate other receptors and signaling pathways that result in heterologous desensitization of PAR2.
By studying monolayers of T84 cells that form TJs and maintain a high TER, we determined the effects of selectively activating PAR2 at the basolateral or apical membranes on permeability. When applied to the basolateral membrane, AP and trypsin decreased TER and increased transepithelial flux of macromolecules from the apical to the basolateral compartments, consistent with an increased permeability. The magnitude of these effects is similar to what we have observed with cytokines such as interleukin 4 and atopic serum (42). The PAR2 agonists were more effective when applied basolaterally than apically, which is consistent with reports that serosally applied PAR2 agonists promote electrolyte secretion from intestinal tissue, whereas agonists applied to the mucosal surface are less effective (43-45). However, PAR2 agonists increase permeability when injected into the colonic lumen of rats and mice (26-28), and physiological concentrations of trypsin in the intestinal lumen can signal to enterocytes by cleaving apical PAR2 (20). Thus, activation of PAR2 at the basolateral and apical membranes of colonocytes may increase epithelial permeability.
Supernatants from degranulated mast cells, when applied to the basolateral membranes of colonocytes, decreased TER and increased macromolecular flux, indicative of increased epithelial permeability. Furthermore, when mast cells were co-cultured with colonocytes, degranulation of mast cells resulted in a marked increase in permeability of colonocytes. Tryptase inhibitors strongly reduced the effects of mast cell supernatant and of mast cell degranulation on the permeability of colonocytes. Thus, upon degranulation, mast cells increase paracellular permeability of colonocytes by a tryptase-dependent mechanism that is probably due to activation of PAR2 on the basolateral membrane. We propose that this mechanism accounts for the observation that the increased permeability of the intestine during chronic stress, allergic inflammation, parasitic infection, and chemically induced inflammation is reduced in mast cell-deficient animals and by mast cell stabilizers (5, 6, 9-11). Experiments with tryptase inhibitors and PAR2-deficient animals are underway to investigate directly this hypothesis.
|
|
increases the permeability of intestinal epithelial cells, which may contribute to the loss of the epithelial barrier in inflammatory bowel disease (46). Mast cell supernatant desensitized AP-induced Ca2+ signaling in colonocytes and mimicked the effects of AP on permeability. Thus, PAR2 probably mediates the effects of mast cell proteases on permeability. However, mast cell proteases may regulate permeability by additional mechanisms, possibly by degrading junctional proteins, as suggested for mouse mast cell protease-1 (6). In addition to tryptase, there are elevated levels of trypsins and activation of coagulation proteases in the inflamed intestine, which may also regulate permeability by cleaving PARs (15-18). Although the precise form of trypsin that is up-regulated in the inflamed intestine is unknown, one possibility is trypsin IV, which is prominently expressed in the human colon and activates PAR2 (24). These proteases may regulate permeability by cleaving PAR2 or other PARs, such as PAR1 and PAR4, which are also expressed in the colon (47, 48). Activation of PAR1 increases permeability by inducing apoptosis of epithelial cells (47).
Mechanisms of PAR2 Regulation of Paracellular Permeability—PAR2 agonists induced redistribution of ZO-1 and occludin from TJs and caused translocation of perijunctional F-actin to the cytosol, as determined by confocal microscopy and subcellular fractionation. We have observed similar effects of PAR2 agonists on the localization of ZO-1 in enterocytes,2 and others have reported (28) that AP induces redistribution of ZO-1 from tight junctions to the cytoplasm of SCBN cells in culture. A similar redistribution of TJ proteins and perijunctional actin is observed in the inflamed intestine (2-6) and in epithelial cells treated with other agents that increase permeability (7, 8). Together, they would be expected to increase paracellular permeability.
Several signaling pathways may account for PAR2-mediated alterations in paracellular permeability, including calmodulin-mediated activation of myosin light chain kinase and the resultant phosphorylation of myosin light chain, which occurs in the colon of mice treated with PAR2-AP (28). The observation that a MEK1/2 inhibitor prevented the effects of AP on TER and F-actin redistribution suggests that activation of ERK1/2 is required for PAR2-induced increases in permeability. Supporting this possibility, PAR2 activates ERK1/2 in colonocytes and enterocytes (29, 40), and MEK1/2 inhibition suppresses AP-induced formation of stress fibers in fibroblasts (30). ERK1/2 also regulate permeability in other epithelial cell lines possibly through phosphorylation of cytoskeletal or TJ proteins (31). To regulate cytosolic or cytoskeletal targets, activated ERK1/2 must be retained in the cytosol rather than translocating to the nucleus. Many pathways may couple activated PAR2 to ERK1/2 in intestinal epithelial cells, including protein kinase C (29) and transactivation of the epidermal growth factor receptor (40). However, these pathways lead to nuclear translocation of ERK1/2 and stimulation of proliferation, rather than regulation of cytosolic targets. We have previously reported that ARRs are scaffolds that recruit Raf-1, MEK1/2, and ERK1/2 to activated PAR2 in endosomes, thereby preventing the nuclear trafficking of these kinases (29). We found that down-regulation of ARRs strongly inhibited AP-induced activation of ERK1/2 and prevented increased permeability and reorganization of F-actin in response to AP. These results suggest that cytosolically retained ERK1/2 regulates the cytoskeleton to control the permeability of TJs.
Proteases that are released and generated during intestinal inflammation may regulate paracellular permeability by directly activating PAR2 on epithelial cells or by indirect mechanisms that are secondary to PAR2-induced intestinal inflammation. Our observations on colonic epithelial cells in culture indicate that PAR2 agonists can directly regulate paracellular permeability, and our observations support previous studies (28) showing direct effects of PAR2 agonists on the permeability of cultured enterocytes. Furthermore, the intracolonic injection in mice of low doses of PAR2 AP, which do not cause detectable inflammation, increases permeability, presumably due to a direct action on colonocytes (28). However, at higher doses PAR2 agonists also trigger intestinal inflammation by mechanisms that depend on the release of proinflammatory peptides substance P and calcitonin gene-related peptide from sensory nerves and on the generation of nitric oxide (27, 49). This PAR2-induced inflammation results in elevated tissue levels of tumor necrosis factor and interferon (28), which can in turn increase paracellular permeability (46, 50).
In summary, upon degranulation mast cells release tryptase, which cleaves PAR2 on the basolateral membrane of colonocytes, thereby activating ERK1/2. ARRs retain ERK1/2 in the cytosol, where they induce redistribution of TJ proteins and perijunctional F-actin, to increase paracellular permeability. Because TJs between colonocytes are an essential component of the barrier to the ingress of luminal bacteria and macromolecules, this mechanism may be of critical importance to the control of inflammation and infection in the intestine. Additional experiments with tryptase inhibitors and PAR2-deficient animals are required to define the importance of this mechanism in inflammation and stress-induced alterations in permeability of the colonic mucosa.
|
FOOTNOTES |
|---|
