INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

PROTEIN KINASE C (PKC) regulates a number of fundamental properties of epithelial cells including vectorial transport and barrier function. PKC profoundly alters epithelial Cl secretion, the physiological process that accounts for mucosal surface hydration (13, 30), through activation or inhibition of specific ion transporters and channels (8, 9, 12). In previous reports, we and others have shown that extended activation of PKC by phorbol 12-myristate 13-acetate (PMA) progressively inhibits cAMP-regulated Cl secretion in human intestinal T84 cells, a well-characterized model of epithelial Cl secretion (1, 8, 20). In addition, PKC influences the barrier function of epithelial cells. In T84 cells, we have shown that PMA impairs junctional integrity and slowly but profoundly decreases transepithelial electrical resistance (TER) (7, 20). Hecht and coworkers (10) showed that extended PMA treatment induces disassembly of T84 epithelial monolayers and causes cellular multilayering.

The PKC family of serine/threonine kinases plays a crucial role in diverse cellular responses such as membrane trafficking, cytoskeletal organization, ion transport, cell growth, and differentiation. At least 11 isoforms of PKC are known, and these are usually categorized into three distinct subtypes: conventional (cPKC) isozymes (, I, II, and ), novel (nPKC) isozymes (, , , µ, and ), and atypical (aPKC) isozymes ( and /) (23, 31). These three subtypes vary in their sensitivity to activators and cofactors: the cPKC isozymes are dependent on phosphatidylserine and the second messengers diacylglycerol (DAG) and Ca²⁺, and they can also be activated by PMA. The nPKC isozymes are similar to the cPKC isozymes in sensitivity to activators, except they obtain full catalytic activity in the absence of Ca²⁺. The aPKC isozymes are independent of DAG or Ca²⁺ and, as a general rule, cannot be directly activated by PMA. The PKC isoforms are widely distributed to varying degrees in mammalian tissue- and cell-specific patterns. Moreover, PKC isoforms exhibit distinct subcellular localizations within individual cell types.

A hallmark of activation of PKC family members is translocation from one biological compartment in the inactive state (e.g., cytosol) to another in the activated state (e.g., plasma or organellar membrane). However, translocation is not an absolute requirement; examples of changes in kinase activity without changes in subcellular localization (and vice versa) are known. Because there is considerable overlap in substrate specificity of the individual PKC isozymes, the precise subcellular localization of inactive and active forms probably confers isozyme specificity in regulating biological processes (23, 31).

The role of specific PKC isoform(s) in the alteration of T84 cell epithelial transport and barrier function by PMA has not been clearly defined. While PMA can have non-PKC cellular targets (e.g., -chimaerin), PKC-inactive phorbol esters do not affect T84 monolayer TER or Cl secretion. Interestingly, we have found that several non-phorbol PKC agonists exert some but not all of the effects of PMA on epithelial phenotype. For example, bryostatin-1 has minimal effect on barrier function despite its ability to inhibit transepithelial Cl secretion (7). In fact, bryostatin-1 is able to partially antagonize the effect of PMA on barrier function (7). We hypothesized that different PKC agonists are capable of activating selective subsets of PKC isoforms that differentially affect the cellular properties of transport and barrier function in T84 epithelia. To approach this hypothesis, we exploited the differential effects of three PKC agonists and three isozyme-selective PKC inhibitors to elucidate the major isoforms involved in downregulation of transport and barrier function in this model system.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cell culture. T84 human intestinal epithelial cells obtained from Dr. Kim Barrett (University of California, San Diego) were grown to confluence at pH 7.4 in 162-cm² flasks (Corning Costar, MA) with a 1:1 mixture of Dulbecco's modified Eagle's medium and Ham's F-12 nutrient mixture supplemented with 6% fetal bovine serum (FBS), 15 mM HEPES, 14.3 mM NaHCO₃, and antibiotics/antimycotic. Flasks were passaged weekly and fed every 3 days. Cell monolayers for experiments were grown to confluence on collagen-coated Transwell inserts (Corning Costar). Monolayers were fed every 3 days and used after stable TER was achieved, ~7-14 days postplating.

Electrophysiology. Under a wide variety of experimental conditions, short-circuit current (I_sc) has been shown to approximate net Cl secretion in T84 monolayers bathed in HEPES-phosphate-buffered Ringer solution (5). I_sc was therefore used as a general assay of transepithelial transport function and was measured in confluent monolayers grown on 0.33-cm² permeable supports using a dual voltage-current clamp and Ag-AgCl and calomel electrodes interfaced via "chopstick" KCl-agar bridges, as previously described (20, 21). In the absence of agonist stimulation, the basal I_sc of T84 monolayers is near zero, and the TER is generally in excess of 1,000  · cm². Thus, in this "tight" model epithelium, TER in electrically quiescent monolayers is a convenient measure of paracellular permeability and barrier function (5). It has previously been validated, for example, that the decrease in TER evoked by long-term treatment of T84 monolayers with PMA represents, specifically, an increase in paracellular permeability (10). TER is calculated by Ohm's law from the voltage deflection induced by a 25-µA external current pulse, as previously described.

In vitro kinase assay. Confluent T84 monolayers grown on 4.7-cm² permeable supports were treated with various PKC agonists and washed twice with cold PBS. Proteins were extracted by 30-min incubation in ice with 500 µl of apical lysis buffer (LB) containing 50 mM Tris · HCl, pH 7.5, 150 mM NaCl, 1% (vol/vol) Triton X-100 (TX-100), 2 mM EDTA, 1 mM EGTA, 30 mM sodium pyrophosphate, 50 mM NaF, 100 µM Na₃VO₄, and Complete protease inhibitor cocktail tablets. The protein concentration of each sample was measured and adjusted to contain 500 µg in 400 µl of LB. Polyclonal antibodies against cPKC (2 µg), nPKC (4 µg), or nPKC (2 µg) were added to each sample for overnight rotation at 4°C. After incubation, immune complexes were precipitated using protein A-agarose beads, washed, resuspended in 20 µl of kinase buffer (35 mM Tris · HCl, pH 7.5, 10 mM MgCl₂, 0.5 mM EGTA, 10 µCi [-³²P]ATP, 60 µM cold ATP, and 1 mM Na₃VO₄) with or without PKC inhibitors and incubated with 10 µg of myelin basic protein as a substrate at 30°C for 30 min. After incubation, the reaction was terminated by adding 5× Laemmli sample buffer to the samples, and the samples were boiled for 5 min. The supernatants were subjected to SDS-PAGE (12% gels), and the gel was dried and subjected to autoradiography.

Immunofluorescence and microscopy. Monolayers grown on 0.33-cm² permeable supports were treated with PMA in medium for the specified time and washed three times with cold PBS. Cells were then fixed in 4% paraformaldehyde for 1 h at room temperature, washed with PBS twice, permeabilized with 0.1% (vol/vol) TX-100 in PBS for 7 min, and rinsed with PBS twice. Filter membranes were cut out in rectangular shape from the Transwell plastic assembly, placed between 50 µl of blocking buffer (1% normal goat serum, 3% BSA in PBS) at both the top and bottom of the monolayers, and incubated for 30 min at room temperature. Polyclonal antibodies against either PKC or PKC were diluted to 10 µg/ml in the blocking buffer containing 0.1% TX-100, and 50 µl of each antibody were placed at both the top and bottom of the monolayers. After overnight incubation in a moisture chamber at 4°C, monolayers were washed in PBS three times for 10 min and incubated in rhodamine-conjugated goat anti-rabbit polyclonal IgG (1:100 dilution) for 1 h at room temperature along with FITC-phalloidin for F-actin staining. Monolayers were then washed three times in PBS and mounted on the microscope slide with Vectashield mounting medium. Confocal images were acquired using a Zeiss inverted microscope equipped with MRC-1024 and Lasersharp software (Bio-Rad).

Subcellular fractionation. T84 cells grown to confluence on collagen-coated permeable supports (4.7 cm²) were washed with ice-cold PBS three times and scraped into 400 µl of the cold homogenization buffer (HB) containing 20 mM Tris · HCl, pH 7.5, 250 mM sucrose, 4 mM EDTA, 2 mM EGTA, and Complete protease inhibitor cocktail tablets. The cells were homogenized on ice with 25 strokes of a glass tissue homogenizer. The resulting homogenate was ultracentrifuged at 86,000 g for 50 min at 4°C (TLA 45 rotor, TL-100 Ultracentrifuge; Beckman). The supernatant was designated the cytosolic fraction. The pellet was resuspended in 400 µl of the HB containing 0.5% (vol/vol) TX-100 by brief sonication and incubated in ice for 30 min. At the end of the incubation period, the samples were centrifuged at 14,000 g for 20 min at 4°C. The resulting supernatant was designated the membrane fraction.

Gel electrophoresis and Western blotting. Equal amounts (~50 µg/sample) of protein, as determined by the Bradford assay, were combined with Laemmli's Sample Buffer containing 5% (vol/vol) -mercaptoethanol and boiled for 5 min. Proteins were separated by electrophoresis on 7.5% SDS-PAGE gels and transblotted to nitrocellulose membranes. The protein-bound nitrocellulose sheets were first incubated for overnight at 4°C in a blocking buffer containing 20 mM Tris, pH 7.5, 500 mM NaCl, and 5% nonfat dry milk. Nitrocellulose sheets were then incubated with the polyclonal antibodies to different PKC isoforms diluted in the blocking buffer (PKC 1:10,000, PKC 1:100, PKC 1:100, and PKC 1:100) for 1 h at room temperature and rinsed for 30 min with a wash buffer containing 20 mM Tris, pH 7.5, 500 mM NaCl, and 0.2% Tween 20. Finally, the membranes were incubated with horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG antibody (1:3,000 dilution) for 1 h at room temperature and washed for 30 min with agitation, during which the wash buffer was changed every 5 min. PKC bands were visualized with ECL (enhanced chemiluminescence) detection reagents.

Materials. Tissue culture reagents and protein A-agarose beads were purchased from Life Technologies, and gel electrophoresis and Western blotting reagents were from Bio-Rad, with the exception of ECL detection reagent, which was purchased from Amersham. Complete protease inhibitor cocktail tablets were from Boehringer Mannheim, and FITC-phalloidin was from Molecular Probes. Anti-PKC for Western blotting was obtained from Sigma, and anti-PKC, anti-PKC, and anti-PKC for immunofluorescent staining and in vitro kinase assay were purchased from Santa Cruz Biotechnology. Secondary antibodies conjugated with various fluorescent dyes were from Jackson Laboratories, and Vectashield mounting medium was from Vector Laboratories. Secondary antibodies conjugated with HRP were obtained from Bio-Rad. The PKC inhibitors Gö-6976, Gö-6850, and rottlerin were obtained from Calbiochem. [-³²P]ATP with a specific activity of 3,000 Ci/mmol was purchased from NEN. All other chemicals were from Sigma.

Statistical analysis. Data are reported as means ± SE. Data were analyzed by one-way ANOVA with Bonferroni/Dunn's post hoc test for comparison with control. Statistical significance is indicated where P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effect of PMA on T84 cell transport function. Our previous results (7, 8, 20, 22) and the results of others (30) indicate that treatment of confluent T84 monolayers grown on permeable supports with PMA progressively inhibits cAMP-elicited Cl secretion. Figure 1A shows the peak I_sc achieved in response to stimulation by 10 µM forskolin at varying times after exposure to 100 nM PMA. Under control conditions, forskolin markedly stimulated I_sc, reaching the peak (baseline I_sc = 4.0 ± 0.2 µA/cm², peak I_sc = 135 ± 16 µA/cm², P < 0.05) within 15 min. Inhibition of peak I_sc was seen well within 1 h of PMA exposure. The IC₅₀ for I_sc inhibition by PMA was ~70 nM (Fig. 1B).

View larger version (11K): [in this window] [in a new window]   Fig. 1.   Phorbol 12-myristate 13-acetate (PMA) dose-dependently inhibits cAMP-elicited short-circuit current (Isc) in T84 cells. Isc, which equals net Cl secretion in T84 cells, was measured by dual voltage-current clamp. Forskolin was used to maximally activate cAMP-dependent Isc. A: monolayers were pretreated with 100 nM PMA for the time indicated and exposed to 10 µM forskolin for 15 min to obtain peak Isc. PMA inhibited cAMP-elicited Isc within 1 h after PMA addition and sustained for at least 4 h (n = 3 experimental and control monolayers at each time point). *P < 0.05. B: monolayers were exposed to different concentrations of PMA for 1 h before forskolin stimulation to obtain IC50 values. The PMA concentration of ~70 nM gave 50% inhibition of peak Isc elicited by forskolin.

Effect of PMA on PKC isozymes in T84 cells. We previously showed that PMA increases total PKC activity in T84 cells (20) and translocates both cPKC and nPKC (35) but not nPKC or aPKC. To further delineate the action of PMA on PKC isozymes in T84 cells, we performed an in vitro kinase assay. The time course for kinase activation of PKC and PKC by PMA closely correlated with membrane translocation of these isozymes (Figs. 2A and 6A). Unexpectedly, despite the absence of translocation in Western blot experiments, PKC kinase activity was increased within 30 min of exposure to PMA (Fig. 2A). Thus PMA increases the kinase activity of PKC, PKC, and PKC in T84 cells, but translocation is only evident by our methods for PKC and PKC. We next examined which of these isoforms was involved in regulation of transport and barrier function.

View larger version (35K): [in this window] [in a new window]   Fig. 2.   Effects of PMA and protein kinase C (PKC) inhibitors on activity of selective PKC isoforms. Monolayers were subjected to in vitro kinase assay to determine the effects of PMA and the selective PKC inhibitors on activity of the 3 PKC isoforms (, , and ) in T84 cells. A: monolayers were stimulated with 100 nM PMA for 30, 60, and 120 min, and PKC, -, and - were immunoprecipitated with isoform-specific antibodies to assess their kinase activities on myelin basic protein (MBP) (blots). PMA time-dependently increased the activities of the two novel isoforms, nPKC and nPKC, as shown by the steady increase in intensity of the 19-kDa MBP band. Activation of the conventional isoform cPKC was not evident until 2 h after PMA addition. Intensities of the bands were measured by densitometric scanning (bar graphs). The data represent the average percentage of the controls (100%) from 3 independent experiments (means ± SE). B: after 2-h stimulation with 100 nM PMA, PKC, -, and - were subjected to kinase reaction in the presence of the 3 selective PKC inhibitors Gö-6976, Gö-6850, and rottlerin. Both Gö-6976 (cPKC specific) and Gö-6850 (cPKC and nPKC specific) inhibited PMA activation of cPKC at 5 µM concentration. Rottlerin (10 µM; PKC specific) partially inhibited cPKC activation. Activation of nPKC by PMA was inhibited by Gö-6850 but not by Gö-6976 or rottlerin. All 3 inhibitors showed strong inhibitory action on nPKC activation by PMA. Intensities of the MBP bands measured by densitometric analysis are expressed in bar graphs (n = 3 for each condition).*P < 0.05.

Effect of isozyme-selective PKC inhibitors on PMA inhibition of T84 transport function. To elucidate the PKC isoform involved in regulation of T84 transport function, we first examined the effect of three structurally unrelated PKC inhibitors, Gö-6976, Gö-6850, and rottlerin, which have well-established but different activity profiles against cPKC and nPKC isoforms. The substituted indolocarbazole compound Gö-6976 has been shown to inhibit cPKC isoforms exclusively (IC₅₀  2 nM against PKC in vitro) (19), with no demonstrable in vitro inhibitory activity against novel Ca²⁺-independent or aPKC isoforms even at high micromolar concentrations. We examined this selectivity for T84 cells by in vitro kinase assay. As expected, 5 µM Gö-6976 had no inhibitory effect on PKC in T84 cells but clearly inhibited activity of PKC in vitro (Fig. 2B). However, Gö-6976 did show some inhibitory activity against nPKC at the 5 µM concentration in T84 cells. The bisindolmaleimide compound Gö-6850 is known to inhibit both cPKC and nPKC isoforms (IC₅₀  8 nM and 132 nM in vitro for PKC and PKC, respectively) (19). In T84 cells, 5 µM Gö-6850 completely inhibited PMA activation of cPKC, nPKC, and nPKC in the in vitro kinase assay as shown in Fig. 2B. At 10 µM, rottlerin is rather specific for the nPKC isoform (IC₅₀  3-6 µM), weakly active against cPKC isoforms (IC₅₀  30 µM), and inactive against PKC (IC₅₀  100 µM) (46). We also confirmed this for T84 cells. Rottlerin showed a selective inhibition of PKC in vitro at 10 µM concentration without affecting PKC while weakly inhibiting PKC in T84 cells (Fig. 2B).

View larger version (13K): [in this window] [in a new window]   Fig. 3.   PMA inhibits cAMP-elicited Isc via activation of PKC. A: monolayers were pretreated with PKC isoform-selective inhibitors for 1 h, followed by 100 nM PMA for 2.5 h in the continued presence of inhibitors. The peak forskolin-stimulated Isc was subsequently measured and compared with control (n = 3 for each condition). The cPKC isoform inhibitor Gö-6976 (5 µM) as well as rottlerin had no effect on PMA-induced inhibition of cAMP-elicited Isc. However, the cPKC and nPKC inhibitor Gö-6850 (5 µM) prevented inhibition of Isc by PMA. B: monolayers were treated with varying concentrations of Gö-6850 for 1 h, followed by 1-h stimulation with 100 nM PMA (n = 3 for each condition). The peak forskolin-stimulated Isc was subsequently measured and compared with PMA. IC50 of Gö-6850 for inhibition of PMA effect on cAMP-elicited Isc is ~2 µM.

Effect of PMA and isozyme-selective inhibitors on T84 cell-barrier function. Figure 4A shows that treatment of T84 monolayers with 100 nM PMA results in a decrease in TER to 28 ± 5% of control within 4 h (baseline TER 1,142 ± 21  · cm² vs. 4-h PMA TER 318 ± 32  · cm²). However, closer examination of the time course of this change indicates that the earliest evidence of a change in basal TER occurs well after the observed inhibition of cAMP-stimulated I_sc, which was prominently evident within 1 h. The IC₅₀ of PMA for inhibition of TER after 2 h exposure was 300-500 nM as shown in Fig. 4B. In contrast to results with cAMP-stimulated I_sc, both Gö-6976 and Gö-6850 at 5 µM inhibited the PMA-induced decline of the TER (Fig. 5A). IC₅₀ values for Gö-6976 and Gö-6850 in vivo were ~2 and ~1 µM, respectively (Fig. 5B). Rottlerin (10 µM) partially reversed the effect of PMA. This partial inhibition could be due to the ability of rottlerin to weakly inhibit PKC at 10 µM (Fig. 2C). Rottlerin at a higher concentration inhibited both cPKC and nPKC (data not shown) and thus is not a valid tool for the inhibitor study. The strong sensitivity to Gö-6976 and partial inhibition by rottlerin suggest that activation of the conventional isoform PKC is associated with downregulation of junctional integrity.

View larger version (11K): [in this window] [in a new window]   Fig. 4.   PMA decreases transepithelial resistance (TER) in a delayed fashion. The effect of PMA on basal TER was measured by dual voltage clamp. A: monolayers were treated with 100 nM PMA for the time indicated and TER was measured. PMA had no effect on basal TER during the initial 60 min but progressively and profoundly diminished TER over the subsequent 3 h. B: monolayers were exposed to varying concentrations of PMA for 2 h and TER was measured. IC50 of PMA for TER inhibition was ~400 nM.

View larger version (14K): [in this window] [in a new window]   Fig. 5.   PMA decreases TER via activation of PKC. The effect of selective PKC inhibitors on PMA-induced decline of TER was determined. A: monolayers were pretreated with 5 µM Gö-6976 or Gö-6850 for 1 h, followed by 100 nM PMA for 2.5 h in the presence of inhibitors. In contrast to Isc, both Gö-6976 and Gö-6850 inhibited the PMA-induced decline of TER (n = 3 for each condition). *P < 0.05. B: IC50 values for the effects of Gö-6976 and Gö-6850 on PMA-inhibited TER were determined by pretreating monolayers with varying concentrations of each inhibitor for 1 h, followed by 2.5-h exposure to PMA in the continued presence of inhibitors. TER was subsequently measured and compared with PMA (n = 3 for each condition).

Spatiotemporal characteristics of PKC and PKC translocation in response to PMA. Given the finding from the inhibitors study that nPKC and cPKC are the major isoforms responsible for regulation of transport and barrier function in T84 cells, we further characterized the time course for their membrane translocation upon PMA addition to substantiate the finding. With Western blot analysis, both PKC and PKC were found predominantly associated with the cytosolic fraction in control monolayers (Fig. 6A). Upon PMA addition, PKC promptly translocated to the membrane as early as 15 min after PMA addition. Translocation of PKC, however, did not begin to become evident until 60 min after treatment. Activation of both PKC and PKC continued for at least 4 h without significant degradation. The sequential activation of PKC followed by PKC is consistent with our functional data obtained using selective PKC isoform inhibitors that correlate early activation of PKC with I_sc inhibition (Fig. 1A) and later activation of PKC with inhibition of TER (Fig. 4A).

View larger version (26K): [in this window] [in a new window]   Fig. 6.   PMA induces membrane translocation of PKC and PKC. Cytosolic (C) and membrane (M) fractions of T84 homogenates were subjected to SDS-PAGE and blotted with PKC isoform-specific antibodies to examine subcellular redistribution of PKC isoforms in response to PMA. A: of the 4 PKC isoforms detected in T84 cells (, , , and ), only nPKC and cPKC were translocated from the cytosol to the membrane fraction by PMA. PKC was found predominantly associated with the membrane fraction as early as 15 min after PMA (100 nM) addition. Translocation of PKC, however, did not begin to be evident until 60 min after treatment. Activation of both PKC and PKC continued for at least 4 h without significant degradation. Similar experiments were confirmed in triplicate. B: distribution of PKC and PKC between cell fractions was determined by densitometric analysis, and the amount of membrane PKC measured after each treatment is expressed as a percentage of the total PKC found in membrane fractions. *P < 0.05.

View larger version (42K): [in this window] [in a new window]   Fig. 7.   PMA translocates PKC to the basolateral membrane. T84 monolayers were treated with 100 nM PMA and then fixed, permeabilized, and incubated with anti-PKC, followed by incubation with rhodamine-conjugated secondary antibody and FITC-phalloidin. The representative vertical (x-z) sections of each monolayer obtained by confocal microscopy are shown. AP, apical membrane; BA, basal membrane; LA, lateral membrane (open arrowhead). Red staining represents PKC and green staining represents F-actin. The control monolayers (A) show a diffuse distribution of PKC throughout the cytoplasm (arrows). As early as 15 min after PMA addition, PKC clears from the cytoplasm and moves toward the cell periphery (B, arrows). At 30 min after PMA addition (C), red staining becomes more sharply defined along the cell boundary (LA) and less prominent at the subapical region. By 1 h of treatment with PMA (D), PKC mostly associates with the basolateral membrane and sharply outlines the individual cells (arrows).

View larger version (42K): [in this window] [in a new window]   Fig. 8.   PMA translocates PKC to the apical membrane. Monolayers were incubated with 100 nM PMA for the time indicated, fixed, permeabilized, and fluorescently labeled with anti-PKC (red) and FITC-phalloidin (green). Confocal images of the vertical sections of each monolayer are shown. The control monolayer (A) shows the immunolocalization of PKC at the basal cytoplasm (arrows). PKC mostly remains at the basal cytoplasm (B, arrow) even after 15-min exposure to PMA. However, after 30 min (C), PKC begins to redistribute toward the apical region in many cells (arrows), and after 60 min (D) PKC becomes clearly localized to the apical membrane and subapical cytoplasmic domain (arrow). In some cells, less pronounced localization of PKC along the basolateral membrane was occasionally detected.

Effect of bryostatin-1 and carbachol on T84 cell transport and barrier function. Like PMA, the non-phorbol ester PKC agonist bryostatin-1 rapidly translocated PKC to the membrane fraction of T84 cells (Fig. 9A). PKC in this fraction was sustained for at least 4 h after bryostatin-1 addition without any evidence of downregulation (degradation) of total PKC protein (Fig. 9B). We also confirmed similar activation of PKC by the in vitro kinase assay (data not shown). In parallel experiments, the peak I_sc elicited by forskolin was significantly inhibited by 100 nM bryostatin-1 (e.g., 25 ± 1% control at 2.5 h, Fig. 9C), similar to our earlier reported findings (7). Inhibition of I_sc was prevented by 5 µM Gö-6850 but not Gö-6976 or 10 µM rottlerin, consistent with the notion that PKC is the key isoform involved in inhibition of epithelial Cl secretion. Membrane translocation of PKC, on the other hand, occurred substantially later with bryostatin-1 than with PMA (Fig. 10A), and, also in contrast to PMA, the total level of both cytosolic and membrane PKC reduced to 39 ± 10% control, suggestive of PKC downregulation or degradation. Barrier function was only transiently (and relatively minimally) affected (Fig. 10B). The basal TER was 94 ± 5% control at 4 h after bryostatin-1 addition.

View larger version (22K): [in this window] [in a new window]   Fig. 9.   Bryostatin-1 promptly activates PKC and inhibits cAMP-elicited Isc. A: monolayers were treated with 100 nM bryostatin-1 for the time indicated, and redistribution of PKC from the cytosolic to membrane fraction was examined. PKC translocated from the cytosol to membrane as early as 20 min after bryostatin-1 addition. Activation of PKC continued for at least 4 h without significant degradation. B: densitometric analysis of data from A, expressed as a percentage of total PKC found in membrane fractions (memb PKC). Similar data were obtained in triplicate experiments. Values are means ± SE. *P < 0.05. C: bryostatin-1 also inhibited cAMP-elicited Isc. Representative inhibition by bryostatin-1 at 2.5 h is shown. Inhibition of Isc was prevented by pretreatment with 5 µM Gö-6850 but not by Gö-6976 or rottlerin. *P < 0.05. Bryo-1, bryostatin-1.

View larger version (20K): [in this window] [in a new window]   Fig. 10.   Bryostatin-1 downregulates PKC and exerts minimal effect on TER. A: in contrast to PMA, 100 nM bryostatin-1 caused a slow activation of PKC. Translocation of PKC from cytosolic to membrane fraction was only evident at the 4-h time point in bryostatin-1-treated monolayers. The amount of membrane-associated PKC as measured by densitometric analysis is shown as a percentage of total PKC found in membrane fractions (top bar graph). At 4 h, however, the total amount of PKC was markedly reduced compared with control (bottom bar graph). *P < 0.05. Similar results were obtained in 3 replicate experiments. B: 100 nM bryostatin-1 induced only a small and transient decrease in TER over 4 h of treatment (n = 3 compared with 3 control monolayers at each time point). *P < 0.05.

View larger version (23K): [in this window] [in a new window]   Fig. 11.   CCh inhibits cAMP-elicited Isc via activation of PKC. A: monolayers were treated with 100 µM CCh for the time indicated, and translocation of PKC was examined. PKC rapidly translocated to the membrane in response to CCh and remained active for at least 4 h. At 4 h, however, PKC began to return to the cytosol, and by 12 h, PKC was mostly associated with the cytosolic fraction. Densitometric analysis of PKC distribution (bar graph) is shown as a percentage of total PKC found in membrane fractions. *P < 0.05. B: monolayers were incubated with 100 µM CCh, and the changes in peak cAMP-elicited Isc were examined for 4 h. CCh caused a significant inhibition of cAMP-elicited Isc, followed by a subsequent recovery. *P < 0.05. C: representative inhibition by 100 µM CCh at 2.5 h is shown (n = 3 for each condition). Inhibition of Isc was prevented by pretreatment with 5 µM Gö-6850 but not with Gö-6976 or rottlerin. *P < 0.05.

View larger version (14K): [in this window] [in a new window]   Fig. 12.   CCh does not activate PKC and has no effect on basal TER. A: up to 4 h after addition of 100 µM CCh, PKC remained associated with the cytosolic fraction as shown by both Western blot (top) and densitometric analysis (bottom). B: CCh had no effect on basal TER (see text). TER remained unaltered up to 4 h after CCh addition.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Activation of PKC by phorbol esters exerts complex and time-dependent effects on transepithelial Cl secretion. In dog trachea, for example, PMA induces a transient activation of I_sc within minutes of exposure, followed by a progressively profound inhibition of cAMP-dependent electrogenic ion transport (2). In native tissue models, it is difficult to ascertain whether the effect of PMA (and, by extension, PKC) is exerted directly at the level of the Cl secretory epithelial cell or on altered neurohormonal regulatory input. However, PKC has been shown to acutely regulate several ion transporters and channels involved in Cl secretion. PKC has been shown to activate CFTR Cl channels at low intracellular Ca²⁺ concentrations (15), perhaps by facilitating protein kinase A-dependent phosphorylation of CFTR (12, 48). Liedtke and coworkers (16-18) showed that Na⁺-K⁺-2Cl cotransporter (NKCC1) in airway epithelia is acutely activated by PKC. The use of cultured epithelial cell lines as models of electrogenic Cl secretion has facilitated detailed mechanistic analysis of the effects of PKC on specific elements of the secretory apparatus. In certain intestinal lines (e.g., HT29cl.19A cells) (1, 42, 43), but not in others (T84 cells) (2, 20), PMA induces a transient, albeit small, increase in I_sc. However, the dominant effect of PMA appears to be inhibitory. Warhurst et al. (45) noted that phorbol ester markedly inhibited the T84 cell Cl secretory response to prostaglandin E₂ by a mechanism that involved receptor desensitization (45), although a variety of later studies indicated a profound inhibition of I_sc responses to all cAMP-mediated stimuli, including permeant analog (20, 22, 30). Initially, the mechanism of inhibition of Cl secretion was thought to involve downregulation of CFTR gene expression (39). However, later studies indicated that the inhibitory effect of PMA on Cl secretion in fact preceded effects at the apical membrane by at least several hours (20, 30). Rather, inhibition of transepithelial Cl secretion by PMA correlated most closely with inhibition of several basolateral membrane transport sites including K⁺ channels (1, 30) and NKCC1 (8, 20). However, the basis for PMA inhibition of multiple independent transport pathways at the basolateral membrane by PMA was not established. As one possible unifying mechanism, we postulated that PMA promoted the endocytic retrieval (and thereby reduced surface expression) of multiple transport pathways at the basolateral membrane (35).

In the current study, we have provided several lines of evidence linking activation of PKC to the progressive inhibition of cAMP-dependent Cl secretion by PMA in the T84 cell model. First, inhibition of I_sc by PMA was prevented by the cPKC and nPKC isoform inhibitor Gö-6850 but not the PKC-selective inhibitor Gö-6976 or the PKC-specific inhibitor rottlerin. The specificity of these inhibitors was confirmed by in vitro kinase assay. Second, early inhibition of forskolin-stimulated I_sc temporally correlates with immediate activation of PKC as shown by in vitro kinase assay as well as early translocation of PKC, which occurred within 15 min after PMA addition. Moreover, bryostatin-1 and CCh both inhibited I_sc and translocated PKC, and in the case of CCh, PKC is the only PKC isoform noted to be translocated and activated during this time period. Finally, immunolocalization studies showed that in response to PMA, PKC takes on a distribution associated with the basolateral membrane.

The roles of PKC in cell function, particularly in epithelial cells, remain poorly understood. A role in cytoskeletal organization was suggested on the basis of findings that F-actin can serve as an isozyme-selective RACK (receptor for activated C kinase) for PKC (29) and that PKC is a MARCKS (myristoylated alanine-rich C kinase substrate) kinase (3). In cardiac myocytes, PKC appears to play a major role in ischemic preconditioning and functions within the context of p42/p44 mitogen-activated kinase pathway in response to diverse cellular growth factors and forms of cell stress (27, 28). PKC has been shown to associate with caveolae in cardiac myocytes (33), suggesting that it may play a role in membrane traffic and in coordinating integrated signaling responses within these specialized membrane microdomains. Weller et al. (47) recently showed that activation of PKC in colon cancer cells may be a trigger for proliferative responses to PMA, and, indeed, overexpression of PKC promotes tumorigenicity (26). In T84 cells, Chow et al. (4) showed that PKC is activated in response to epidermal growth factor (EGF) and may participate in the negative regulation of Ca²⁺-dependent Cl secretion by EGF and CCh.

In addition to vectorial transport, epithelial cells also possess the property of barrier function. PKC appears to play an important role in junction formation after epithelial disassembly, such as in the Ca²⁺ "switch" model, although the precise mechanism whereby PKC regulates this process remains to be established (37). PKC-dependent junction assembly is initiated at the level of E-cadherin and the zonula adherens, rather than at the tight junctions, and thus this process may not necessarily be mediated by the same PKC isoenzyme(s) that influence paracellular permeability (which is largely determined at the level of the zonula occludens). PKC-dependent junctional hyperpermeability in confluent T84 monolayers is known to be induced simply by an elevation of cell Ca²⁺ (38), although whether the Ca²⁺ dependence of junctional permeability reflects a specific role for a cPKC is uncertain.

The present experiments closely link extended activation of PKC to impaired barrier function in model T84 epithelia. First, we observed that TER remains relatively constant after PMA treatment (and PKC translocation) until a time that follows the later translocation of PKC to the membrane fraction. Second, the PMA-associated fall in TER was prevented by Gö-6976, a PKC inhibitor that is highly selective for Ca²⁺-dependent cPKC isoforms. Third, in response to PMA, PKC translocated from the basal cytoplasm to the apical zone of T84 monolayers, in the vicinity of the junctional complexes and perijunctional actomyosin ring known to affect junctional integrity. Comparison of the effect of PMA with that of other PKC agonists provides indirect support for the hypothesis that PKC is the key isoform involved in junctional regulation. For example, CCh, unlike PMA, had no effect on TER and did not alter PKC subcellular distribution at any time point.

Bryostatin-1, compared with PMA, induced a delayed membrane translocation of PKC that was associated with a smaller and transient fall in TER. The return of TER toward control levels with extended bryostatin-1 treatment may reflect accelerated degradation (downregulation) of PKC. Indeed, bryostatin-1 has been shown to induce proteosome-mediated degradation of PKC through enhanced ubiquitinization (14). Bryostatin-1, which shares with PMA an affinity for the DAG binding site of PKC, is known to induce a subset of the cellular responses evoked by PMA and, interestingly, to antagonize many of the responses it does not share with PMA (11). Thus our earlier finding that bryostatin-1 is able to partially antagonize the effect of PMA on TER (7) is likely to reflect the ability of bryostatin-1 to downregulate PKC. We speculate that bryostatin-1-activated PKC can induce only minimal effects on TER before it is depleted and that the early downregulation of PKC by bryostatin-1 prevents extended PMA activation of PKC and thereby attenuates the fall in TER.

A role for PKC in junctional regulation in epithelia has previously been postulated (32). Notably, Mullin et al. (24) showed that overexpression of wild-type PKC in LLC-PK₁ cells renders the cells more sensitive to PMA-induced junctional disruption, whereas expression of a dominant-negative PKC construct renders them resistant. Other PKC isoforms may also influence junctional structure and permeability under certain conditions. For example, junctional permeability is increased by overexpression of PKC in cultured renal epithelial cells (24), and in Madin-Darby canine kidney and Caco-2 epithelial cells, the aPKC is the only isoform that specifically localizes near the tight junctional complex (6). Because the PMA-induced decrease in TER was partially inhibited by rottlerin, a role for PKC cannot be entirely excluded. However, at the concentration used, rottlerin also partially inhibited PKC. It is unknown, and our studies could not address, whether PMA can alter PKC activity, and thus a role for this aPKC also cannot be excluded.

Our localization data indicate that PMA induces an intracellular redistribution of PKC toward the apical zone of the cell, where it may potentially interact directly or indirectly with various components of the tight junction. The target of PKC that leads to altered junctional permeability remains to be elucidated but is likely to involve the cytoskeleton. Hecht et al. (10) showed that disruption of T84 monolayer integrity by PMA is associated with disruption of perijunctional F-actin (10). The permeability characteristics of tight junctions are known to be modulated by the tension of the perijunctional actin-myosin ring, which, in turn, is mediated by myosin light chain kinase (MLCK) (41). PKC is known to itself alter phosphorylation of both MLC and MLCK, although whether this mechanism can account for the observed effects of PMA in T84 cells remains speculative. Conflicting reports have appeared regarding this concept. Turner et al. (40) showed in a Caco-2 subclone that PMA acutely increased MLCK phosphorylation and decreased MLC phosphorylation; this response was associated with an acute increase in TER, presumably due to relaxation of the perijunctional actin-myosin ring. Other Caco-2 clones, however, have been shown to behave similarly to T84 cells with a progressive decrease in TER in response to PMA (36). Enhanced actin-myosin contractile activity through PKC-mediated regulation of MLCK also has been reported (25), but in one instance, enhanced junctional permeability due to phorbol ester was shown to be independent of MLCK (34). Other potential targets of PKC must also be considered. Interestingly, phorbol ester-induced barrier dysfunction in endothelial cells appears to involve extracellular signal-regulated kinase (ERK1/2) signaling via Ras (44).

In summary, by using multiple agonists and isozyme selective inhibitors, we have been able to dissociate PKC actions on transport function and barrier function in T84 model epithelia. Activation of PKC appears to inhibit electrogenic Cl secretion, whereas extended activation of PKC decreases TER. The present studies demonstrate that PKC-dependent stimuli can elicit divergent effects on epithelial cell function through differential activation of distinct PKC isoenzymes, which in turn act at distinct subcellular localizations. It thus may prove possible to selectively target specific PKC isoenzymes for activation, inhibition, or downregulation in the context of antidiarrheal and anticancer drug development.