Abstract
Using monolayers of intestinal (Caco-2) cells, we showed that oxidants disassemble the microtubule cytoskeleton and disrupt barrier integrity (permeability) (Banan et al., 2000a). Because exposure of ourparental cells to oxidants causes protein kinase C (PKC)-δ to be translocated to particulate fractions, we hypothesized that PKC-δ activation is required for these oxidant effects. Monolayers of parental Caco-2 cells were incubated with oxidant (H2O2) ± modulators. Other cells were transfected with an inducible plasmid to stably overexpress PKC-δ or with a dominant negative plasmid to stably inhibit the activity of native PKC-δ. In parentalcells, oxidants caused translocation of PKC-δ to the particulate (membrane + cytoskeletal) fractions, activation of PKC-δ isoform, increases in monomeric (S1) tubulin and decreases in polymerized (S2) tubulin, disruption of the microtubule cytoarchitecture, and loss of barrier integrity (hyperpermeability). In transfected cells, induction of PKC-δ overexpression by itself (3.5-fold over its basal level) led to oxidant-like disruptive effects. Disruption induced by PKC-δ overexpression was potentiated by oxidants. Overexpressed PKC-δ resided in particulate fractions, indicating its activation. Stable inhibition of native PKC-δ activity (98%) by dominant negative transfection substantially protectedagainst all measures of oxidative disruption. We conclude that 1) oxidants induce loss of intestinal epithelial barrier integrity by disassembling the microtubules in large part through the activation of the PKC-δ isoform; and 2) overexpression and activation of PKC-δ is by itself a sufficient condition for disruption of these cytoskeleton and permeation pathways. Thus, PKC-δ activation may play a key role in intestinal dysfunction in oxidant-induced diseases such as inflammatory bowel disease.
A fundamental property of the gastrointestinal (GI) epithelium is its ability to function as a highly selective permeability barrier that normally permits the absorption from the lumen of nutrients, water, and electrolytes but prevents the passage of proinflammatory molecules into the mucosa. Loss of GI barrier integrity, in contrast, can allow the penetration of normally excluded luminal substances (e.g., endotoxin) into the mucosa and can lead to the initiation or continuation of inflammatory processes and mucosal damage (Hollander, 1998; Banan et al., 1999; Keshavarzian et al., 1999). Indeed, loss of mucosal barrier integrity has been implicated in the pathogenesis of multiple organ system dysfunction, inflammatory bowel disease, ethanol- and nonsteroidal anti-inflammatory drug-induced chemical injury, and other GI disorders as well as systemic disorders (e.g., alcoholic liver disease) (Unno et al., 1996; Hollander, 1998; Keshavarzian et al., 1999). The underlying difficulty in managing these inflammatory disorders is due in large part to our limited understanding of their pathophysiology and lack of effective preventive strategies.
Although the pathophysiology of inflammation in mucosal barrier dysfunction remains poorly understood, several studies have shown that chronic gut inflammation is associated with high levels of oxidants (e.g., H2O2), and that oxidative damage is a key contributor to loss of barrier integrity and injury (Keshavarzian et al., 1992; McKenizie et al., 1996; Banan et al., 2000a,b,c, 2001a). Oxidative disruption is of clinical importance not only because oxidants are common in inflammation but also because they can lead to mucosal barrier hyperpermeability and, in turn, lead to the initiation and/or perpetuation of mucosal inflammation and dysfunction. A major advance in recent years in GI inflammation (inflammatory bowel disease) research was recognition that a leaky gut barrier can cause intestinal inflammation and that oxidants can cause this hyperpermeability in the intestinal tract (Yamada et al., 1993;Hermiston and Gordon, 1995). Thus, characterizing how gut barrier integrity is lost under oxidative, proinflammatory conditions is of fundamental clinical and biological importance.
Using monolayers of human intestinal (Caco-2) cells exposed to oxidants as a model of cytoskeletal and barrier disruption, we showed previously that oxidants (e.g., H2O2) induce loss of intestinal barrier integrity in part by disrupting the assembly of the microtubule cytoskeleton (Banan et al., 1999, 2000a,b). We also showed that the instability of microtubules is key in mucosal damage under in vivo (Banan et al., 1998a) and in vitro conditions (Banan et al., 1998b, 1999, 2000a,b, 2001c,d, 2002a). Damage is based on the inability of cellular polymeric tubulin pools to resist disassembly, and the ability of the monomeric tubulin pools to increase, leading to microtubule instability. Despite the critical importance of the microtubule cytoskeleton in the maintenance of intestinal barrier integrity, the intracellular signaling mechanism through which oxidants destabilize the microtubules and lead to gut barrier dysfunction remains poorly understood.
In previous studies using Caco-2 cell clones, we reported that specific PKC isoforms (PKC-β1 and PKC-ζ) are crucial in the protection of mucosal epithelial barrier and microtubule integrity (Banan et al., 2002a, 2001b,c). The PKC family, which includes at least 12 known isoenzymes, can be classified into three subfamilies based on differences in sequence homology and cofactor requirement (Boner et al., 1992; Goodnight et al., 1995; Cho et al., 1998; Banan et al., 2001c, 2002a): classical PKC isoforms (α, β1, β2, and γ), novel PKC isoenzymes (δ, ε, θ, η, and μ), and atypical PKC isoforms (λ, τ, and ζ). Intestinal epithelial (e.g., Caco-2) cells express at least six isoforms of PKC: PKC-α, PKC-β1, PKC-β2, PKC-δ, PKC-ε, and PKC-ζ (Wang et al., 1996); Abraham et al., 1998;Banan et al., 2001c, 2002a). Our pilot and exploratory observations of parental (wild-type) Caco-2 cells (Banan et al., 2002b) suggest that oxidants induce the membrane association of an abundant isoform of PKC, namely, δ, and therefore consider this isoform as a possible contributor to oxidant-mediated disruption of the microtubule cytoskeleton. To address this possibility, in the current study, we tested the hypothesis that oxidant-induced loss of both microtubule integrity and barrier permeability of epithelial monolayers depends on translocation and activation of the δ-isoform of PKC.
To this end, we used both pharmacological manipulations and transfected intestinal cell lines we have developed: in some clones the 75-kDa isoform PKC-δ was reliably overexpressed by induction; in the other clones, PKC-δ activity was inhibited. Herein, we report mediation of oxidant-induced barrier hyperpermeability and microtubule cytoskeletal disassembly and disruption by a member of the novel subfamily of PKC isoforms in intestinal cells.
Materials and Methods
Cell Culture.
Caco-2 cells were obtained from American Type Culture Collection (Rockville, MD) at passage 15. This cell line was chosen for our studies because they form monolayers that morphologically resemble small intestinal cells, with defined apical brush borders, tight junctions, and a highly organized microtubule network upon differentiation (Gilbert et al., 1991; Banan et al., 1998b). Caco-2 cells also form monolayers that can be studied for weeks, rather than just days, as is typical of most in vitro preparations. This allowed us to measure alterations in intestinal barrier integrity. In addition, Caco-2 cells closely resemble normal intestinal cells in that they express intestinal hydrolases such as sucrase-isomaltase and alkaline phosphatase. Furthermore, these cells are similar to native intestinal epithelial cells in that they have receptors for prostaglandins, growth factors, vasoactive intestinal peptide, low-density lipoprotein, insulin, and specific substrates such as dipeptides, fructose, glucose, hexoses, and vitamin B12 (Gilbert et al., 1991). Cells were maintained at 37°C in complete Dulbecco's modified Eagle's medium (DMEM) in an atmosphere of 5% CO2 and 100% relative humidity. Parental cells or stably transfected cells (see below) were split at a ratio of 1:6 upon reaching confluence, and set up in either six- or 24-well plates for experiments, or T-75 flasks for propagation. Cells grown for barrier function experiments were split at a ratio of 1:2 and seeded at a density of 200,000 cells/cm2into 0.4 μM Biocoat Collagen I Cell Culture Inserts (0. 3-cm2 growth surface; BD Biosciences, Franklin Lakes, NJ), and experiments were performed at least 7 days postconfluence. The media were changed every 2 days. The utility and characterization of this cell line have been reported previously (Gilbert et al., 1991; Banan et al., 1998b).
Plasmids and Stable Transfection.
The sense and dominant negative plasmids of PKC-δ were constructed as described previously (Gossen and Bujard, 1992; Banan et al., 2001c,d). A unique tetracycline-responsive expression (TRE) system was used to overexpress the native PKC-δ. cDNA encoding the entire reading frame of PKC-δ was subcloned into the TRE vector, creating TRE PKC-δ. The dominant negative PKC-δ plasmid was also constructed (Cho et al., 1998; Banan et al., 2001d).
Cultures of Caco-2 cells grown to 50 to 60% confluence were cotransfected with hygromycin resistance plasmid (pβ-hygro) and expression plasmids encoding either PKC-δ or dominant negative PKC-δ by Lipofectin (Lipofectin reagent; Invitrogen, Carlsbad, CA) as described previously (Banan et al., 2001c). Briefly, cells were incubated for 16 h at 37°C with the plasmid DNA in serum-free media in the presence of LipofectAMINE (25 μl/25-cm2 flask). Subsequently, the DNA-containing solution was removed and replaced by fresh media containing 10% fetal bovine serum to relieve cells from the shock of exposure to serum-free media. After transfection, cells were subjected to hygromycin selection (1 mg/ml) over 4 weeks. Resistant cells were maintained in DMEM/FBS and 0.2 mg/ml hygromycin (selection medium). More specifically for inducible overexpression of PKC-δ, Caco-2 cells were transfected with a plasmid expressing the tetracycline-responsive transactivator (tTA or so-called pTEToff because it encodes a tetracycline-regulated transcription factor that “represses” in the presence of tetracycline) along with a second plasmid conferring resistance to G-418. After selection in 0.6 mg/ml of G-418 (selection media), one such clone (i.e., parental tTA or pTEToff) was then itself transfected with the TRE PKC-δ expression system. pβ-hygro was included to confer resistance to hygromycin (selection marker, 1 mg/ml). Control conditions included vector alone (TRE-z). PKC-δ protein expression and activity were verified, respectively, by Western blot analysis of cell lysates or activity assay (see below). Clones were subsequently plated on cell culture inserts and allowed to form confluent monolayers and then used for experiments. When tetracycline was present in the medium, its concentration was 1 μg/ml.
Experimental Design.
In the first series of experiments, postconfluent monolayers of parental Caco-2 cells were preincubated with oxidant (H2O2, 0–0.5 mM) or vehicle (isotonic saline) for 30 min. As we have shown previously (Banan et al., 2000a,c, 2001a), H2O2 at 0.5 mM disrupts microtubules and barrier integrity in these cells. These experiments were then repeated using monolayers composed of cells either overexpressing PKC-δ (i.e., TRE PKC-δ) or almost completely lacking PKC-δ activity (dominant negative). Reagents were applied on the apical side of monolayers unless otherwise indicated. Because our previous studies (Banan et al., 2000a,c) showed that regardless of whether apical or basolateral exposure of oxidants was used the results were qualitatively similar, all current studies used apical application. In all experiments, microtubule cytoskeletal stability (cytoarchitecture, assembly, and disassembly), tubulin assembly, PKC-δ subcellular distribution (membrane, cytoskeletal, and cytosolic fractions), PKC-δ activity (immunoprecipitation and in vitro assay), and barrier integrity (clearance) were assessed.
In the second series of experiments, cell monolayers that were overexpressing PKC-δ were incubated (30 min) with oxidant (H2O2) or vehicle. Outcomes measured were as described above.
In a third series of experiments, monolayers of dominant negative transfected cells lacking PKC-δ activity were treated with oxidants. In all experiments, PKC-δ activity was determined in immunoprecipitated samples (see below). In corollary experiments, we investigated the effects of PKC-δ activation or inactivation on the state of tubulin assembly and disassembly and on stability of the cytoarchitecture of the microtubule cytoskeleton. Monomeric and polymerized fractions of tubulin (the structural protein subunit of microtubules) were isolated and then analyzed by quantitative immunoblotting (Banan et al., 2000a, 2001c). Microtubule integrity was assessed by 1) immunofluorescent labeling and fluorescence microscopy to determine the percentage of cells with normal microtubules, 2) detailed analysis by high-resolution laser scanning confocal microscopy (LSCM), and 3) immunoblot analysis of monomeric (S1) and polymerized (S2) tubulin pools.
Fractionation and Western Immunoblotting of PKC.
Differentiated cell monolayers grown in 75-cm2flasks were processed for the isolation of the cytosolic, membrane and cytoskeletal fractions as we described previously (Banan et al., 2001b,c). Briefly, after treatments, postconfluent monolayers were scraped and ultrasonically homogenized in Tris-HCl buffer (20 mM Tris-HCl pH 7.5, 0.25 mM sucrose, 2 mM EDTA, 10 mM EGTA, 2 μg/ml aprotinin, 2 μg/ml pepstatin, 2 μg/ml leupeptin, and 2 μg/ml phenylmethylsulfonyl fluoride). The homogenates were then ultracentrifuged (100,000g for 40 min at 4°C), and the supernatant was removed and used as a source of the cytosolic fraction. Next, pellets were washed with 0.2 ml of Tris-HCl buffer and resuspended in 0.8 ml of buffer containing 0.3% Triton X-100 and maintained on ice for 1 h. The samples were then centrifuged (100,000g for 1 h at 4°C), and the supernatant was used as the source of the membrane fraction. To this remaining pellet, 0.3 ml of cold (4°C) lysis buffer (150 mM NaCl, 50 mM Tris-HCl, 1 mM EDTA, 1 mM EGTA, 1% Nonidet P-40, 0.1% sodium deoxycholate, 0.1% SDS, 2 μg/ml aprotinin, 2 μg/ml pepstatin, 2 μg/ml leupeptin, and 2 μg/ml phenylmethylsulfonyl fluoride) was added. The samples were then placed on ice for 1 h and ultracentrifuged as described above. The remainder of the lysate or Triton-insoluble cytoskeletal fraction was then removed. Protein content of the various cell fractions was assessed by the Bradford method (Bradford, 1976). For total PKC extraction, which provides the fraction used to confirm total PKC-δ, scraped monolayers were placed directly into 1.5 ml of cold lysis buffer and subsequently ultracentrifuged as described above. The supernatant was used for bulk protein determination.
For immunoblotting, samples (75 μg of protein/lane) were added to SDS buffer (250 mM Tris-HCl pH 6.8, 2% glycerol, and 5% mercaptoethanol), boiled for 5 min, and then separated on 7.5% SDS-PAGE (Banan et al., 2001c). Subsequently, proteins were transferred to nitrocellulose membranes (0.2-μm pore size) and then blocked in 3% bovine serum albumin for 1 h followed by several washes with Tris-buffered saline. The immunoblotted proteins were incubated for 2 h in Tween 20, Tris-buffered saline, 1% bovine serum albumin, and the primary mouse monoclonal anti-PKC-δ (Santa Cruz Biotechnology, Santa Cruz, CA) at 1:1000 dilution for 1 h at room temperature. A horseradish peroxidase-conjugated goat anti-mouse antibody (Molecular Probes, Eugene, OR) was used as a secondary antibody at 1:3000 dilution. Proteins on membranes were visualized by enhanced chemiluminescence (Amersham Biosciences, Piscataway, NJ) and autoradiography, and subsequently analyzed by densitometry. The identity of the PKC-δ band was assessed by 1) using the PKC-δ blocking peptide (Santa Cruz Biotechnology) in combination with the anti-PKC-δ antibody that prevents the appearance of the corresponding “major” band in Western blots. 2) Additionally, in the absence of the primary antibody to PKC-δ, no corresponding band for PKC-δ was observed. 3) The PKC-δ band ran at the expected molecular mass of 75 kDa as confirmed by a known positive control for PKC-δ (from rat brain lysates). 4) Prestained molecular weight markers (Mr 67,000 and 93,000) were run in adjacent lanes. In preliminary studies using total PKC extracts, we confirmed that overexpression of PKC-δ or negative dominant inhibition of PKC-δ did not affect the relative expression levels of other PKC isoforms.
Immunoprecipitation and PKC-δ Activity Assay.
Immunoprecipitated PKC-δ was collected and processed for its ability to phosphorylate a synthetic peptide (Banan et al., 2001b,d; Vancurova et al., 2001). Briefly, after treatments, confluent cell monolayers were lysed by incubation for 20 min in 500 μl of cold-lysis buffer (20 mM Tris-HCl pH 7.4, 150 mM NaCl, 10 μg/ml anti-protease cocktail, 10% glycerol, 1 mM sodium orthovanadate, 5 mM NaF, and 1% Triton X-100). The lysates were clarified by centrifugation at 14,000g for 10 min at 4°C. For immunoprecipitation, the lysates were incubated for 90 min at 4°C with monoclonal anti-PKC-δ (1:2000 dilution, in excess). The extracts were then incubated with protein A/G plus agarose for 1 h at 4°C. The immunocomplexes were collected by centrifugation (2500g for 5 min) in a microfuge tube and washed three times with immunoprecipitation buffer containing 5 mM Tris-HCl pH 7.4 and 0.2% Triton X-100. They were then washed one time with kinase buffer (20 mM HEPES pH 7.5, 10 mM MgCl2, 2 mM MnCl2, and 20 μM ATP) and resuspended in 20 μl of kinase buffer and 5 μl of 5× reaction buffer (1 mg/ml histone H1 and 0.25 mg/mll-α-phosphatidyl-l-serine) plus 5 μCi of [γ-32P]ATP and subsequently incubated for 5 min at 30°C. Reactions were then stopped by the addition of 8 μl of 5× sample buffer, and the samples were boiled at 95°C for 5 min before separation by 7.5% PAGE. The extent of histone H1 phosphorylation was determined by scintillation counting of excised Coomassie blue histone polypeptide bands. Counts for blanks were subtracted from the sample activity. Sample activity was also corrected for protein concentration (Bradford, 1976), and PKC-δ activity was reported as picomoles per minute per milligram of protein.
Immunofluorescent Staining and High-Resolution Laser Scanning Confocal Microscopy of Microtubules.
Cells from monolayers were fixed in cytoskeletal stabilization buffer and then postfixed in 95% ethanol at −20°C as we described previously (Banan et al., 1998b,1999, 2000a,b, 2001b,c,d). Cells were subsequently pressed for incubation with a primary antibody, monoclonal mouse anti-β-tubulin (Sigma-Aldrich, St. Louis, MO), 1:200 dilution for 1 h at 37°C, and then incubated with a secondary antibody (fluorescein isothiocyanate-conjugated goat anti-mouse; Sigma-Aldrich), 1:50 dilution for 1 h at room temperature. Slides were washed thrice in Dulbecco's phosphate-buffered saline and subsequently mounted in Aquamount. After staining, cells were observed with an argon laser (λ = 488 nm) using a 63× oil immersion plan-apochromat objective, numerical aperture 1.4 (Carl Zeiss GmbH, Jena, Germany). Single cells and/or a clump of two to three cells from desired areas of monolayers were processed using the image processing software on an ultra high-resolution LSCM (Carl Zeiss GmbH). The cytoskeletal elements were examined in a blinded manner for their overall morphology, orientation, and disruption as we have described previously (Banan et al., 1999, 2000a,b, 2001b,c,d). At least 1200 cells/group (200 × six slides) were examined in four different fields by LSCM, and the percentage of cells displaying normal microtubules was determined. The identity of the treatment groups for all slides was decoded only after examination was complete.
Microtubule (Tubulin) Fractionation and Quantitative Immunoblotting of Tubulin Assembly and Disassembly.
Polymerized (S2) and monomeric (S1) fractions of tubulin were isolated using a unique method we described previously (Banan et al., 1998b, 1999,2001c). Cells were gently scraped and pelleted with centrifugation at low speed (700 rpm for 7 min at 4°C) and resuspended in microtubule stabilization-extraction buffer (0.1 M 1,4-piperazinediethanesulfonic acid pH 6.9, 30% glycerol, 5% dimethyl sulfoxide, 1 mM MgSO4, 10 μg/ml anti-protease cocktail, 1 mM EGTA, and 1% Triton X-100) at room temperature for 20 min. Tubulin fractions were separated after a series of centrifugation and extraction steps. Specifically, cell lysates were centrifuged at 105,000g for 45 min at 4°C and the supernatant containing the soluble monomeric pool of tubulin (S1) was gently removed. The remaining pellet was then resuspended in 0.3 ml of Ca2+-containing depolymerization buffer (0.1 M 1,4-piperazinediethanesulfonic acid pH 6.9, 1 mM MgSO4, 10 μg/ml anti-protease cocktail, and 10 mM CaCl2) and incubated on ice for 60 min. Subsequently, samples were centrifuged at 48,000g for 15 min at 4°C, and the supernatant (S2 fraction or cold/Ca2+-soluble fraction) was removed. To ensure the complete removal of the S2 fraction, the remaining pellet was treated with the Ca2+-containing depolymerization buffer twice more by resuspension and centrifugation. The “microtubules” were recovered by separately incubating (at 37°C for 30 min) the S1 and S2 fractions with stabilizing agents Taxol (10 μM) and GTP (1 mM) in microtubule stabilization buffer (0.1 M 1,4-piperazinediethanesulfonic acid pH 6.9, 30% glycerol, 5% dimethyl sulfoxide, 10 μg/ml anti-protease cocktail, 1 mM EGTA, 1 mM MgCl2, and 1 mM GTP) to promote polymerization of tubulin. Tubulin was then recovered by centrifugation and resuspended in the above-described stabilization buffer. Fractionated S1 and S2 samples were then flash frozen in liquid N2 and stored at −70°C until immunoblotting. For immunoblotting, samples (5 μg of protein/lane) were placed in a standard SDS sample buffer, boiled for 5 min, and then subjected to PAGE on 7.5% gels. Procedures for Western blotting were performed as described previously (Banan et al., 1998b, 1999, 2001c). To quantify the relative levels of tubulin, the optical density (O.D.) of the bands corresponding to immunoradiolabeled tubulin was measured with a laser densitometer.
Determination of Barrier Permeability by Fluorometry.
Status of the integrity of monolayer barrier function was assessed by a widely used and validated technique that measures the apical-to-basolateral paracellular flux of fluorescent markers such as fluorescein sulfonic acid (FSA; 200 μg/ml, 0.478 kDa) as we (Banan et al., 1999,2000a,b,c, 2001a,b,c,d, 2002a) and others (Sanders et al., 1995; Unno et al., 1996) have described. Briefly, fresh phenol-free DMEM (800 μl) was placed into the lower (basolateral) chamber and phenol-free DMEM (300 μl) containing probe (FSA) was placed in the upper (apical) chamber. Aliquots (50 μl) were obtained from the upper and lower chambers at zero time and at subsequent time points and transferred into clear 96-well plates (Clear bottom; Costar, Cambridge, MA). Fluorescent signals from samples were quantitated using a fluorescence multiplate reader (FL 600; BIO-TEK Instruments, Boulder, CO). The excitation and emission spectra for FSA were excitation 485 nm and emission 530 nm. Clearance (Cl) was calculated using the following formula: Cl (nl/h/cm2) = Fab/([FSA]a × S), where Fab is the apical-to-basolateral flux of FSA (light units/h), [FSA]a is the concentration at baseline (light units/nl), and S is the surface area (0.3 cm2). Simultaneous controls were performed with each experiment.
Statistical Analysis.
Data are presented as mean ± S.E.M. All experiments were carried out with a sample size of at least six observations per treatment group that were run in triplicate on two to three different days. Statistical analysis comparing treatment groups was performed using analysis of variance followed by Dunnett's multiple range test (Harter, 1960). Correlational analyses were done using the Pearson test for parametric analysis or, when applicable, the Spearman test for nonparametric analysis. P values <0.05 were deemed statistically significant.
Results
We initially confirmed our previous findings (Banan et al., 2000a,c, 2001a) that incubation of parental Caco-2 monolayers with oxidant (H2O2) substantially causes barrier hyperpermeability (increases in FSA clearance) without any cell death (ethidium homodimer probe), indicating loss of cell monolayer barrier integrity. We also confirmed our pilot and exploratory findings in these parental cells (Banan et al., 2002b) that oxidant (H2O2) causes the translocation (shift) in the distribution of the PKC-δ isoform from the cytosol-to-membrane-bound fractions. In the current investigation, using both pharmacological and targeted molecular interventions (transfection), the possible role of PKC-δ isoform in the underlying cause of oxidant-induced barrier dysfunction was investigated.
Stable Overexpression of PKC-δ Isoform after Transfection of Intestinal Cells.
Parental Caco-2 cells [tTA Parental(or pTEToff)] were cotransfected with cDNA encoding both hygromycin resistance (for selection) and a TRE PKC-δ. In this TRE, overexpression of PKC-δ is achieved in the absence of tetracycline (Fig. 1A), whereas its presence reduces expression to the levels seen in the parental cell line (Fig. 1B). Cell lysates of confluent monolayers were prepared from these transfected cells and then analyzed by Western immunoblotting. Figure 1A shows overexpression of the PKC-δ isozyme in these transfected cells. The PKC-δ isolated from transfected cells comigrated with a known standard (∼75 kDa) for PKC-δ from rat brain lysates. The identity of the PKC-δ band was further ascertained by using the PKC-δ blocking peptide in combination with the anti-PKC-δ antibody that prevented the appearance of the corresponding major band in the Western blots. As expected, exclusion of the primary antibody also resulted in the disappearance of the corresponding PKC-δ band. The immunoblot shown in Fig. 1B demonstrates that total PKC-δ levels were elevated by 3.5-fold compared with parental cells. Overexpression of PKC-δ at this level caused neither any cellular toxicity (0% cell death assessed by ethidium homodimer probe) nor any changes in Caco-2 cell growth (assessed by bromodeoxyuridine assay) (Banan et al., 2000a,c).
Disruptive Effects of Overexpression of PKC-δ Isoform on Cell Monolayer Barrier Integrity and Potentiation of Oxidant-Induced Damage.
In exploratory experiments, multiple clones of intestinal Caco-2 cells transfected with 1, 2, 3, 4, or 5 μg of TRE PKC-δ cDNA showed a dose-dependent loss of barrier paracellular integrity in monolayers as assessed by increased FSA clearance. The clone transfected with 4 μg of TRE PKC-δ cDNA provided disruption at a level comparable with that of oxidant (H2O2) in parental cell line. Accordingly, we used this clone for the experiments described below.
In Caco-2 cells, PKC-δ overexpression by itself, in the absence of added oxidant, deleteriously affected monolayer barrier function (Figs.2) and the microtubule cytoskeleton (Fig.3). For example, in cells stably overexpressing PKC-δ (TRE PKC-δ and exposed to vehicle) monolayer barrier integrity was disrupted as determined by increased FSA clearance (∼60% greater clearance; Fig. 2). Incubation of these same transfected cells overexpressing PKC-δ with tetracycline (i.e., TRE PKC-δ + tetracycline), as expected, maintained monolayer barrier integrity at near normal levels. Similarly, parental type cells (those not overexpressing PKC-δ) exposed to vehicle (with or without tetracycline) also showed normal and intact barrier integrity. These parental type cells, on the other hand, had their barrier integrity disrupted by oxidant (H2O2; 0.5 mM). Moreover, incubation with oxidant potentiated loss of monolayer barrier integrity in the transfected cells overexpressing PKC-δ (TRE PKC-δ). As expected, this potentiation was inhibited in the presence of tetracycline. Furthermore, as expected, transfection of only the TRE-z vector by itself did not cause barrier damage (FSA clearance = 18 ± 6 nl/h/cm2 for vector-transfected cells exposed to vehicle and 19 ± 5 for parental cells exposed to vehicle; 821 ± 19 for vector-transfected cells exposed to H2O2 alone and 825 ± 22 for parental cells incubated in H2O2). Indeed, both vector-transfected cells and parental cells responded in a similar manner to either vehicle or H2O2.
PKC-δ overexpression by itself also disrupted the microtubule cytoskeleton as demonstrated by the low percentage of intestinal cells displaying normal microtubules (Fig. 3). As expected, this overexpression-induced disruption was prevented when tetracycline was present (i.e., TRE PKC-δ + tetracycline). Incubation with oxidant potentiated loss of microtubule integrity in these transfected cells. Tetracycline also substantially prevented this potentiating effect. Inparental cells (i.e., tTA parental), as for disruption of barrier integrity, microtubules were damaged by oxidant (H2O2; 0.5 mM). Furthermore, transfection of TRE-z vector alone did not damage the microtubules (percentage of normal microtubules = 97 ± 3% for vector transfected cells exposed to vehicle and 98 ± 2 forparental cells exposed to vehicle). Both vector-transfected cells and parental cells also responded in a comparable manner to oxidant (40 ± 4% for vector-transfected cells exposed to H2O2 and 42 ± 5% forparental cells exposed to H2O2).
High-resolution laser scanning confocal microscopy of immunofluorescently stained microtubule cytoskeleton also shows (Fig.4) that Caco-2 cells overexpressing PKC-δ (i.e., TRE PKC-δ without tetracycline), exhibit an abnormal cytoskeleton in cell monolayers exposed to vehicle (C). This abnormality is shown by the intracellular appearance of a fragmented, disrupted, and collapsed microtubule network. In the presence of tetracycline (D) these cells, exhibited an intact microtubule network as shown by a normal stellate and radial cytoarchitecture of the cytoskeleton originating from the perinuclear region. This normal cytoarchitecture is indistinguishable from the parentalcells exposed to vehicle (A). Parental type cells (B) exhibit microtubule damage when exposed to oxidant alone as shown by a disrupted microtubule cytoskeleton in the cytosol, which is comparable with that of the PKC-δ-overexpressing cells (C).
To determine the effects of PKC-δ overexpression on the dynamic alterations in the polymerization and depolymerization states of the microtubule cytoskeleton, we performed immunoblotting analysis of tubulin, the structural protein of microtubules. To this end, the polymerized tubulin fraction (S2, an index of microtubule stability) and the monomeric tubulin (S1, an index of microtubule disruption) were isolated and analyzed by a SDS-PAGE fractionation technique we developed for this purpose. Immunoblotting analysis of tubulin (Fig.5) corroborated the microtubule studies noted above. PKC-δ-overexpressing cells (TRE PKC-δ, vehicle-treated) showed an abnormal tubulin assembly that was comparable with oxidant exposed parental cells as shown by a reduction in the polymerized S2 tubulin and an increase in the monomeric S1 tubulin. In tetracycline-incubated TRE PKC-δ cells (where overexpression of δ is prevented), in contrast, neither any decreases in polymerized S2 tubulin nor any increases in monomeric S1 tubulin were observed, indicating normal assembly of the microtubule cytoskeleton. This was comparable with the normal tubulin polymerization seen in parental cells exposed to vehicle. Inparental cells, treatment with oxidant resulted in increased tubulin depolymerization. Transfection of vector alone, similar to its lack of effects on microtubules and barrier integrity, did not affect tubulin assembly (e.g., percentage of tubulin assembly = 65.5 ± 0.4% for vector-transfected Caco-2 cells exposed to vehicle and 66 ± 0.6% for parental cells exposed to vehicle; 47 ± 1.0% for vector-transfected cells exposed to H2O2 and 46 ± 0.5% for parental cells exposed to H2O2).
Figure 6 shows a representative Western blot of the alterations in tubulin assembly, demonstrating that PKC-δ overexpression by itself decreases the stable polymerized tubulin band density to a level comparable with that of oxidant-exposedparental cells; this overexpression-induced disruptive effect was prevented by tetracycline. These findings parallel the injurious effects of PKC-δ overexpression on intestinal microtubule integrity and on barrier permeability.
Intracellular Distribution and Constitutive Activation of Overexpressed PKC-δ in Transfected Intestinal Monolayers.
Western immunoblotting assessment (Fig.7, A–D) of the cytosolic, membrane and cytoskeletal-associated fractions from transfected cells overexpressing PKC-δ showed that the δ (75-kDa)-isoform of PKC is found mostly in the membrane and cytoskeletal fractions of transfected cells with only a small distribution to the cytosolic fractions (Fig. 7C). Inparental cells (Fig. 7A), in contrast, we found a mostly cytosolic distribution of PKC-δ (indicating inactivity) with smaller pools in the membrane and cytoskeletal (i.e., particulate) fractions. Figure 8 shows the intracellular distribution of the overexpressed PKC-δ in various Caco-2 cell fractions as a fraction of total distribution (expressed in arbitrary units). Finding PKC-δ in particulate pools indicates that the overexpressed PKC-δ isoform is “constitutively active” because achieving this distribution by PKC-δ did not require any stimulus. Treatment of transfected cells with oxidant (Figs. 7D and 8), however, further increased the fraction of PKC-δ isoform in the membrane and cytoskeletal fractions, reaching near total activation of PKC-δ. As expected, parental type cells exposed to oxidant (Figs. 7B and 8) also show increased membrane and cytoskeletal distribution of PKC-δ. On the other hand, parental type cells exposed to vehicle (Figs. 7A and 8) show a mostly cytosolic distribution of PKC-δ, indicating inactivity.
Intracellular Distribution and Constitutive Activation of PKC-δ Isoform in Intestinal Cells Correlates with Several Different Indices of Monolayer Barrier Disruption.
Using data across all experimental conditions, there was a significant (p < 0.05) correlation (r = 0.95) between PKC-δ levels (optical density from the particulate fraction) and increased monolayer FSA clearance, suggesting that constitutive activation of the δ-isoform may be important in disruption of intestinal barrier permeability. Similarly, we found other robust (positive) correlations when either microtubule instability (i.e., disruption) or tubulin disassembly (i.e., increased S1 pool) was correlated with the PKC-δ levels (r = 0.92 and 0.93, respectively,p < 0.05 for each). When another marker of disruption, reduced tubulin assembly (i.e., decreased S2 pool) was used against PKC-δ an additional robust correlation was observed (r = 0.90, p < 0.05), further suggesting that activation of the PKC-δ isoform is critical in disruption by oxidant.
Dominant Negative Inhibition of PKC-δ to Inactivate δ and Its Prevention of Oxidant-Induced Disruptive Effects.
The aforementioned findings indicate that PKC-δ might, by itself, play a key role in cell monolayer barrier disruption and possibly in oxidant-induced barrier hyperpermeability. To show that PKC-δ is required for oxidant-induced monolayer hyperpermeation, we used a dominant negative approach to stably decrease the steady-state activity of the PKC-δ isoform. Figure 9 shows activity levels of PKC-δ isoform from immunoprecipitated particulate fractions of parental type Caco-2 cells that were transfected with PKC-δ dominant negative cDNA (negative dom. PKC-δ) and plasmid encoding hygromycin resistance. These data show a substantial reduction (−98%) in the activity of PKC-δ isoform in these dominant negative-transfected cells, which were exposed to vehicle. In comparison, in parental cells exposed to oxidant, PKC-δ activity is increased, whereas in dominant negative-transfected cells oxidant can no longer increase δ-isoform activity. Figure 9 further shows that TRE PKC-δ-overexpressing cells have substantial increases in δ-activity in the particulate fractions. This activation in these cells overexpressing PKC-δ is further increased in the presence of oxidant, paralleling data in Figs.7 and 8. As expected, parental cells exposed to vehicle or TRE PKC-δ-transfected cells treated with tetracycline (the latter expressing almost native levels of δ-protein) show low activation levels for this isoform. These data further parallel our findings on tubulin dynamics, microtubule integrity, and intestinal barrier permeability. Furthermore, as expected, we did not observe any affects of the dominant negative transfection on the total expression levels of PKC-δ isoform. Specifically, densitometry analysis of Westerns from whole cell extracts probed for total PKC-δ levels indicated that the O.D. for δ-bands equals 4578 ± 137 O.D. units for parental cells (not transfected) and 4622 ± 175 O.D. units for dominant negative-transfected cells, indicating no differences in protein levels.
In exploratory inhibition studies, we observed a dose-dependent effect of varying amounts (1, 2, 3, 4, or 5 μg) of PKC-δ dominant negative cDNA on inhibition of oxidant-induced disruption in intestinal monolayers. The clone transfected with 4-μg plasmid for PKC-δ dominant negative provided maximum inhibition of oxidant-mediated barrier dysfunction, and it was thus used for subsequent inhibition studies (Figs.10-12).
PKC-δ inactivation by itself did not deleteriously affect Caco-2 monolayer barrier integrity (FSA clearance; Fig. 10). Dominant negative inhibition of the PKC-δ activity did, however, substantially and significantly attenuate the barrier hyperpermeability induced by 0.5 mM oxidant. Indeed, a large percentage (∼64%) of oxidant-induced monolayer hyperpermeation seems to be PKC-δ-dependent.
In parallel, analysis of the percentage of dominant negative transfected cells with a normal microtubule cytoskeleton demonstrates (Fig. 11) that dominant negative inhibition of PKC-δ activity prevented injury to microtubules by a disruptive dose of oxidant. PKC-δ isoform inactivation by itself did not injure the microtubules.
Immunoblotting analysis of tubulin from these dominant negative transfected cells further demonstrates (Fig. 12) that in the absence of PKC-δ isoform activation, oxidant does not elicit any decreases in the stable S2 tubulin fraction (nor any increases in monomeric S1 tubulin), indicating prevention of (or protection against) microtubule disassembly and disruption.
Discussion
These studies, which use monolayers of intestinal epithelial cells as a model of gut barrier integrity, demonstrate that translocation and activation of the δ-isoform of PKC is required for oxidant-induced loss of microtubule cytoskeletal assembly and cytoarchitecture and of barrier integrity. They also demonstrate that PKC-δ by itself is sufficient to induce barrier hyperpermeability. The mechanism for the effects of PKC-δ isoform seems to be destabilization of the dynamic alterations in polymerized (S2) tubulin and monomeric (S1) tubulin-based cytoskeletons in the intestinal epithelium. These conclusions are supported by several independent lines of evidence as discussed below.
First, incubation of parental intestinal monolayers with oxidant activates PKC-δ and evokes a cascade of alterations that are consistent with the proposed mechanism. Oxidant activates a specific PKC isoform, δ, increases the levels of unstable monomeric tubulin pool while reducing the size of stable polymerized tubulin pool, decreases the percentage of Caco-2 cells with intact microtubule architecture, and reduces monolayer barrier integrity. Second, overexpression of PKC-δ induces an oxidant-like disruption. In these transfected cells, PKC-δ evokes a similar and consistent cascade of alterations. This disruption seems to require overexpression and activation PKC-δ. More specifically, as in parental cells, loss of integrity required activation through the translocation of PKC-δ from the cytosolic to the particulate fraction. Third, transfected cells that overexpress PKC-δ are more sensitive to the loss of the microtubules and barrier integrity by oxidant. Indeed, in these stably transfected cells, induction of PKC-δ overexpression potentiates the disruptive effects of the exogenously added oxidant. The enhanced sensitivity to oxidant in transfected Caco-2 cells seems to require not only overexpression but also enhanced activation of PKC-δ isoform. Fourth, cells that are transfected with dominant negative to PKC-δ and that show inactivation of PKC-δ (98% decrease from normal levels) were rendered less sensitive to all the disruptive effects of oxidant. In fact, these cells were substantially protected against oxidative insult.
Finally, quantitative considerations such as robust correlations between several outcome measures (e.g., microtubules, tubulin pools, and monolayer barrier integrity versus PKC-δ isoform activation) further support our conclusions. The high strength of these correlations, which explains 90 to 95% of the variance, is consistent with the idea that increased PKC-δ activation is critical to the disruptive effects of oxidant on the intestinal epithelial microtubules and on barrier permeability. Other PKC isoforms may also be disruptive. However, our data indicate that a substantially large portion (∼60%) of disruption by oxidant insult is mediated through PKC-δ.
Our findings using transfected cells are not only consistent with the known properties of PKC isoforms but also with published reports, including those in non-GI models. All PKCs consist of N-terminal regulatory domains and C-terminal catalytic domains, which are separated by a flexible hinge region (Gopalakrishna and Jaken, 2000). In resting cells, PKC is mainly found in an inactive conformation. In this inactive phase, PKC is mainly distributed in the soluble (cytosolic) fraction and only loosely bound to membrane components. Regulatory domains of PKC isoforms vary from one subfamily to the next as well as among individual isoforms within a given subfamily (Cho et al., 1998; Mullin et al., 1998; Gopalakrishna and Jaken, 2000). Not surprisingly, differences among isozymes with respect to activation conditions, and subcellular location indicates that individual PKC isoforms have distinct activation mechanisms as well as mediate distinct biological processes (Abraham et al., 1998; Cho et al., 1998;Banan et al., 2001c, 2002a). Additionally, previous pharmacological reports (Boner et al., 1992; Saxon et al., 1994; McKenna et al., 1995;Wang et al., 1996), including our own (Banan et al., 2001b), have shown that general PKC activation (translocation) is necessary for the observed effects of PKC. More specifically, the translocation of PKC from the cytosolic to the particulate fraction of the cell is a key step in its activation (Goodnight et al., 1995; Wang et al., 1996). An immunofluorescent study suggested that PKC seems to be involved in tumor necrosis factor-α-induced disruption in intestinal (IEC-18) cells (Chang and Tepperman et al., 2001). Also, PKC-δ can cause disruption (hyperpermeation) of pig kidney (LLC-PK1) cell monolayers (Mullin et al., 1998). However, the effects of PKC activation in cellular models can sometimes be complex and may vary with different experimental settings and cell types. For example, we recently reported original findings (Banan et al., 2001c, 2002a) that the classical β1 (78-kDa) isoform of PKC and the atypical ζ (72-kDa)-isoform of PKC are required for growth factor (EGF)-induced protective effects on the intestinal epithelial barrier integrity. Thus, it seems that activating or mimicking just different isoforms of PKC will have distinct effects on the GI epithelium, including protection.
There are other reported effects of PKC-δ in cellular models. These include PKC-δ tyrosine phosphorylation by oxidants, which in turn increases its enzymatic activity and promotes oxidative processes in variety of cell lines, including stable clones of Caco-2 cells. The consequences of this increased PKC-δ activity also include loss of mitochondrial membrane potential, increased apoptosis, up-regulation of ornithine decarboxylase, loss of anchorage, and decreased cell growth (Konishi et al., 1999; Otieno and Kensler, 2000; Sun et al., 2000;Cerda et al., 2001; Majumder et al., 2001). For example, in several cells oxidant stress induces targeting of PKC-δ to mitochondria, causing the loss of mitochondrial transmembrane potential and release of cytochrome c, and the induction of apoptosis (Majumder et al., 2001). In murine papiloma PE cells, PKC-δ seems to transiently regulate the expression of genes that control cell growth such as ornithine decarboxylase under oxidative stress (Otieno and Kensler, 2000).
One study reported (Cerda et al., 2001) that a 4- to 4.5-fold increase in the levels of PKC-δ protein in transfected Caco-2 cells seems to cause toxic effects, including cell cycle arrest, loss of anchorage dependence, and apoptosis. We did not find such toxicity, possibly because PKC-δ was only expressed 3.5-fold in our cells. Our finding is more similar to PKC-δ overexpression in pig kidney epithelial cells, which causes monolayer hyperpermeability without any toxic side effects (Mullin et al., 1998). Overall, our current study on the δ-isoform of PKC, we believe, suggests a unique pathophysiological role, namely, mediation of oxidant-induced loss of both barrier permeability and microtubule cytoskeletal assembly, among the “novel” subfamily of PKC isoforms in intestinal cells.
Despite the critical importance of PKC signal transduction in disruption, the role of the PKC-δ isoform in cell dysfunction, especially in loss of epithelial monolayer barrier function, is not understood. More specifically, the mechanism through which PKC-δ isoform disrupts and disassembles the microtubules (tubulin-based) cytoskeleton and leads to barrier hyperpermeability is unknown. Our previous reports showed that inducible nitric-oxide synthase activation (Banan et al., 2000b, 2001a) and intracellular calcium accumulation (Banan et al., 2001b) can lead to oxidative disruption of the cytoskeleton in our monolayers. Therefore, up-regulation of either of these mechanisms can underlie PKC's effects on monolayer barrier hyperpermeability.
Finally, our previous studies indicate that microtubules (Banan et al., 1999, 2000a) and actin (Banan et al., 2000c, 2001a) components of the cytoskeleton are essential to the maintenance of an intact monolayer barrier in Caco-2 cells. Disruption of either microtubules or actin is sufficient to cause loss of monolayer barrier integrity. Conversely, stabilization of either one of these components protects barrier integrity. For example, using colchicine, a known microtubule-disruptive agent, we showed that disassembling the microtubules causes barrier dysfunction, whereas Taxol, a microtubule stabilizer, protects monolayer barrier integrity against oxidant insult (Banan et al., 1999, 2000a). Similarly, using cytochalasin-D, an actin destabilizer, we showed that depolymerizing actin leads to barrier disruption, while stabilizing actin by phalloidin protects monolayer integrity (Banan et al., 2000c). In this intestinal model, we have no evidence that microtubules and/or actin components affect each other directly or indirectly. Indeed, our studies support an independent role for each of these cytoskeletal elements in the protection of intestinal barrier integrity.
In summary, our findings strongly support the idea that PKC-δ can substantially disrupt the dynamics of microtubule cytoskeletal assembly and intestinal barrier integrity and that it is key for oxidant-induced hyperpermeability in intestinal cell monolayers. We recognize that Caco-2 cells are a transformed cell line and that tumor cells may respond differently to changes in barrier integrity than do nontransformed cells, including enterocytes in native tissue. Nonetheless, this new knowledge may prove useful because decreasing or suppressing the activity of disruptive PKC isoform through inactivation of endogenous PKC may lead to new therapeutic strategies for the treatment of a variety of GI inflammatory disorders (e.g., inflammatory bowel disease) that are associated with oxidative disruption.
Footnotes
-
This work was supported in part by a grant from Rush University Medical Center, Department of Internal Medicine, and by a grant from the American College of Gastroenterology. Portions of this work were presented in abstract form as an oral presentation: “Research Forum”, during the annual meeting of the American Gastroenterological Association (Digestive Disease World), May 2002.
-
DOI: 10.1124/jpet.102.037218
- Abbreviations:
- GI
- gastrointestinal
- IBD
- inflammatory bowel disease
- PKC
- protein kinase C
- DMEM
- Dulbecco's modified Eagle's medium
- TRE
- tetracycline-responsive expression
- tTA
- tetracycline-responsive transactivator
- LSCM
- laser scanning confocal microscopy
- PAGE
- polyacrylamide gel electrophoresis
- O.D.
- optical density
- Received April 10, 2002.
- Accepted May 23, 2002.
- The American Society for Pharmacology and Experimental Therapeutics