In the companion article (Umar S, Scott J, Sellin JH, Dubinsky WP, and Morris AP, Am J Physiol Gastrointest Liver Physiol 278: 753–764, 2000), we have shown that transmissible murine colonic hyperplasia (TMCH) increased cellular cystic fibrosis transmembrane conductance regulator (CFTR) mRNA and protein expression, relocalized CFTR within colonocytes, and enhanced mucosal cAMP-dependent Cl−secretion. We show here that these changes were dependent on elevated cellular levels of membrane-bound Ca2+- and diacylglycerol-sensitive protein kinase C (PKC) activity (12-fold), induced by selective (3- to 4-fold) rises in conventional PKC (cPKC) isoform expression and membrane translocation. Three cPKC isoforms were detected in isolated crypts: α, β1, and β2. cPKC-β1 rises preceded and those of cPKC-α and cPKC-β2 paralleled cellular hyperproliferation and its effects on CFTR expression and cAMP-dependent Cl− current secretion. Only cPKC-β1 and cPKC-β2 were membrane translocated during TMCH. Furthermore, only cPKC-β1 trafficked to the nucleus, whereas cPKC-β2 remained partitioned among cytosolic, membrane, and cytoskeletal subcellular fractions. Modest increases in novel PKC-ε (nPKC-ε) expression and subcellular membrane partitioning were recorded during TMCH, but no changes were seen for PKC-δ or -η. No nPKC isoform nuclear partitioning was detected. The orally bioactive cPKC inhibitor Ro-32–0432 reversed both TMCH and elevated cellular CFTR mRNA levels, whereas a pharmacologically inert analog (Ro-31–6045) failed to inhibit either response. On the basis of these facts, we present a new hypothesis whereby PKC-dependent cellular proliferation promotes endogenous cellular CFTR levels. PKC-β1 was identified as a candidate regulatory PKC isoform.
- protein kinase C
- cystic fibrosis transmembrane conductance regulator
- anion transport
protein kinase c (PKC) consists of a family of serine/threonine kinases shown to be involved in the regulation of many aspects of cell growth, differentiation, and function (6). In the context of epithelial Cl− secretion, PKC has been implicated in the regulation of both apical and basolateral plasma membrane ion transport events (9, 15, 37). The cell biology of epithelial cAMP-stimulated anion transport in vitro is dominated by reports demonstrating a negative role of phorbol ester-sensitive PKC signaling in cystic fibrosis transmembrane conductance regulator (CFTR) expression. Chronic PKC stimulation by phorbol ester treatment is believed to lead to downregulation of CFTR mRNA and possibly protein levels (1, 12, 36, 38). However, phorbol esters can also mediate positive effects on the cellular function of CFTR. They have been shown to potentiate cAMP-responsive Cl− current generation in epithelial cells through phosphorylation-dependent effects on channel gating (4, 19 ). They are also predicted to enhance the apical plasma membrane accumulation of this anion channel through stimulatory effects within the cellular biosynthetic pathway (25).
However, a recognized caveat of phorbol ester use both in vitro and in vivo is their ability to promote both negative and positive effects on cellular signaling, depending on their time of application and dose. The broad stimulatory effects of these artificial agents in cells may also be accompanied by nonspecific actions on multiple cellular targets. Given that we had found that proliferation in vivo promoted native colonocyte CFTR mRNA and protein levels (see companion paper, Ref. 38a), we measured phorbol ester-sensitive conventional PKC (cPKC) and novel PKC (nPKC) activity during transmissible murine colonic hyperplasia (TMCH) to determine whether individual or multiple isoforms were downregulated during periods of enhanced CFTR expression in vivo. Our studies found that native tissue PKC activation does not mimic the in vitro effects of phorbol esters. cPKC activity increased in the TMCH model. Furthermore, pharmacological inhibition of cPKC activity in vivo prevented both TMCH and enhanced cellular CFTR expression levels. Three participating cPKC isoforms were identified.
Polyclonal rabbit anti-PKC-α, PKC-β1 anti-peptide, and monoclonal anti-PKC-γ antibodies were purchased from Sigma Immunochemicals (St. Louis, MO). PKC-β2, -δ, -η, and -ε anti-peptide antibodies were procured from Santa Cruz Biotechnology (Santa Cruz, CA). Ro-32–0432 and Ro-31–6045 were purchased from Alexis Biochemicals (San Diego, CA). Succinylated gelatin carrier (Gelofusine) was a kind gift from B. Braun Medical (Aylesbury, UK).
PKC activity assay.
Conventional Ca2+- and diacylglycerol-sensitive PKC (cPKC) activity was measured in cytosolic and membrane fractions of isolated crypt, crypt denuded colon, and whole proximal and distal colon samples from normal and day 12 post-Citrobacter-infected animals using a PKC assay system (GIBCO BRL, Grand Island, NY). Subcellular fractions were prepared by homogenization of both isolated crypts and whole distal colon tissue in 1.5 ml of ice-cold homogenization buffer (50 mM Tris ⋅ HCl, pH 7.5, containing 5 mM EDTA, 10 mM EGTA, 0.3% mercaptoethanol, 10 mM benzamidine, 50 μg/ml phenylmethylsulfonyl fluoride, and 50 μg/ml leupeptin) using a Tekmar Tissuemizer (Tekmar, Cincinnati, OH) rotating at 20,000 rpm for five 40-s periods. Cytosolic and membrane fractions were separated by sedimentation for 30 min at 100,000 g. The resulting pellet was then detergent extracted in buffer (1% Triton X-100, 10 mM Tris ⋅ HCl, pH 7.5, 140 mM NaCl, 25 mM KCl, 5 mM MgCl2, 2 mM EDTA, and 2 mM EGTA) containing protease inhibitors, and the solubilized membrane fraction was collected by centrifugation. Protein content was determined using the Bio-Rad DC protein assay kit. PKC activity was quantified by the transfer of the terminal phosphate of [γ-32P]ATP (Amersham, Arlington Heights, IL) to myelin basic protein synthetic peptide (41). cPKC specificity was determined by using a cPKC pseudosubstrate inhibitor peptide [PKC-(19–36)] provided by the manufacturer. Incorporated radioactivity was determined by scintillation counting, and activities were expressed as picomoles of phosphate transferred per minute per milligram of protein.
Tissue preparation for Western blot analysis.
Male Swiss Webster mice (15–20 g; Harlan Sprague Dawley, Houston, TX) were killed by cervical dislocation 0, 1, 3, 6, 9, 12, and 15 days after Citrobacter inoculation. The entire distal colon was removed, flushed with ice-cold PBS, and cut open longitudinally. Crypts were then isolated by attachment of the colonic sheet to a paddle followed by immersion in Ca2+-free standard Krebs-buffered saline (in mmol/l: 107 NaCl, 4.5 KCl, 0.2 NaH2PO4, 1.8 Na2HPO4, 10 glucose, and 10 EDTA) maintained at 37°C and gassed with 5% CO2-95% O2. The crypts were then separated from the surrounding connective tissue/muscle layers by mechanical vibration for 30 s into ice-cold KCl HEPES saline (in mmol/l: 100 potassium gluconate, 20 NaCl, 1.25 CaCl2, 1 MgCl2, 10 HEPES, 10 glucose, and 5 sodium pyruvate) and 0.1% BSA, resembling the intracellular medium. The remaining crypt-denuded colon was then saved for biochemical analysis of PKC activity. Control crypts were similarly isolated from the entire (∼3 cm long) proximal colon of TMCH animals for cPKC activity measurements.
Crude cellular homogenates were prepared from isolated crypts taken from three normal and Citrobacter-infected animals per experimental observation by homogenization in detergent containing buffer (in mM: 50 Tris ⋅ HCl, 250 sucrose, 2 EDTA, 1 EGTA, pH 7.5, and 10 2-mercaptoethanol, with 0.5% Triton X-100, plus protease inhibitors) followed by a low-speed spin (15,000 g for 15 min). The clear supernatant was saved as total cell extract. Subcellular membrane and cytosolic fractions were prepared from similar crypts by omitting detergent and by centrifugation at 100,000g. The resulting pellet was then detergent extracted (1% Triton X-100, 10 mM Tris ⋅ HCl, pH 7.5, 140 mM NaCl, 25 mM KCl, 5 mM MgCl2, 2 mM EDTA, 2 mM EGTA, and protease inhibitors), and the solubilized membrane fraction was collected by centrifugation. The detergent-insoluble pellet was solubilized in RIPA buffer (PBS, 1% NP-40, 0.1% SDS, and 0.5% sodium deoxycholate) for estimates of cytoskeletal-associated protein.
Nuclear extracts were prepared from freshly isolated mucosal tissue by the method of Zhang and colleagues (42). Protein concentration was measured in cytosolic, membrane, cytoskeletal, and nuclear fractions before electrophoresis. Mouse brain homogenates acted as positive controls for PKC isoform immunoblotting assays. Total cell extracts or subcellular fractions (30 μg protein/lane) were subjected to 10% SDS-PAGE and electrotransferred to nitrocellulose membranes. The efficiency of electrotransfer was checked by backstaining gels with Coomassie blue and/or by reversible staining of the electrotransferred protein directly on the nitrocellulose membrane with ponceau S solution. No variability in transfer was noted. Destained membranes were blocked with 5% nonfat dried milk in 20 mM Tris ⋅ HCl and 137 mM NaCl, pH 7.5 (TBS), for 1 h at room temperature and then overnight at 4°C. Immunoantigenicity was detected by incubating the membranes for 2 h with PKC isoform/CFTR antibodies (0.5–1.0 μg/ml in TBS containing 0.1% Tween 20; Sigma). After washing, membranes were incubated with horseradish peroxidase-conjugated goat anti-rabbit IgG or rat anti-mouse IgG and developed using the ECL detection system (Amersham) according to the manufacturer's instructions.
To assay for purity of cytosolic, membrane, cytoskeletal, and nuclear fractions, the following non-PKC-related marker proteins were detected by Western blotting (data not shown): 1) Lamin B was used to assay nuclear purification. Lamin B does not partition into other cellular compartments, and, by this test, both membrane and cytosol were free of nuclear contamination. 2) The transmembrane protein E cadherin was used as a marker for the membranous fraction. E cadherin does not translocate into the nucleus and is not freely cytosolic. Both nuclear and cytosolic fractions were free of this marker, indicating negligible cytoplasmic contamination. 3) Monomeric β-actin was immunodetected to assay purity of the cellular cytosolic fraction. This molecule does not enter the nucleus and is not membrane bound. By this test, nuclear, cytoskeletal, and membrane fractions were not contaminated by cytosolic protein. 4) E cadherin association with the cytoskeleton was also assayed. This adhesion molecule was detected in both membranous and cytoskeletal fractions but not in cytosolic and nuclear fractions.
Cellular mRNA extraction.
Total mRNA was isolated from whole distal colon of normal andCitrobacter-infected mice using TRIzol reagent (GIBCO BRL) according to the manufacturer's instructions. For Northern blot analysis, each preparation (10 μg total RNA) was denatured and fractionated on a 1% agarose gel containing formaldehyde. RNA was then transferred to a GeneScreen Plus nylon membrane (DuPont NEN), and the blot was hybridized at 60°C in 10% dextran sulfate, 1 M NaCl, 1% SDS, and 100 μg/ml denatured salmon testes DNA with the use of an [α-32P]dCTP-labeled probe encompassing the R domain of CFTR (bases 1773–2654; 2 × 106 cpm/ml) and subsequently with a probe against glyceraldehyde 3-phosphate dehydrogenase (GAPDH; bases 163–608; 1 × 106cpm/ml). The latter signal was used to normalize the mRNA in each lane. The probe for CFTR detection was generated by PCR of full-length CFTR cDNA, and the GAPDH probe was generated by RT-PCR from mouse colonic RNA. Both were confirmed by oligonucleotide sequencing before random primed labeling.
Development of TMCH.
TMCH was developed in male Swiss Webster mice by oral inoculation with 16-h culture of Citrobacter freundii(biotype 4280, ATCC). Age-matched control mice received sterile culture medium only. Additional methodological details are given in the companion article (38a).
Analysis of PKC activity in subcellular fractions of isolated crypts, crypt-denuded colon, and whole colon tissue.
Total conventional Ca2+- and diacylglycerol-sensitive cPKC activity in cytosol and membrane fractions prepared from purified distal colonic crypts, crypt denuded distal colon, or whole distal colon tissue were measured before (day 0), and during (day 12 post-Citrobacterinfection) the peak mucosal hyperproliferatory response (38a). Enzymatic activity, recorded as phosphate incorporation into a synthetic substrate peptide (see methods), was detected in both fractions (Fig. 1 A).
cPKC activity in the cytosolic fractions obtained from freshly isolated crypts from normal and TMCH conditions were similar (Fig. 1 A;P < 0.01, n = 4 animals in duplicate). In contrast, membrane-bound cPKC activity in isolated crypts fromCitrobacter-infected animals exhibited a large 12-fold increase compared with uninfected controls, which was significant (Fig.1 A; P < 0.001, n = 6 animals in duplicate). Thus increased membrane-specific cPKC activity correlated with high levels of cellular CFTR mRNA and protein expression (38a). Cytosolic PKC activity levels remained constant, suggesting that recruitment of new enzyme through increased production had occurred.
cPKC activity was also measured in detergent-extracted crude cellular homogenates taken from whole distal colon, crypt-denuded whole distal colon, and whole proximal colon of normal and day 12post-Citrobacter-infected mouse colon (Fig. 1 B). TMCH was associated with a nearly twofold increase in specific cPKC activity in extracts from whole distal colon (Fig. 1 B). Extracts from crypt-denuded whole distal colon (lamina propria and musculature) also exhibited a similar magnitude change (Fig. 1 B) but started with, and achieved, much lower values than those recorded in the whole tissue. In comparison, cPKC activity in nonhyperproliferative proximal colon extracts from the same TMCH colon exhibited no detectable change (Fig. 1 B), and overall values were similar in magnitude to those recorded in crypt-denuded distal colon. Thus uninvolved colon from TMCH mice, which did not exhibit any changes in epithelial cell CFTR expression (38a), likewise failed to exhibit any change in specific cPKC activity. These experiments were duplicated four times, and reported changes in whole distal colon and crypt-denuded distal colon were significant (P < 0.001, Student'st-test).
This analysis did not, however, provide information regarding which phorbol ester-sensitive PKC isoforms were activated in hyperproliferating crypt colonocytes overexpressing CFTR. Therefore, an immunological approach was used to examine the subcellular distribution of both cPKC and nPKC in purified isolated crypt extracts from normal and TMCH distal colonic mucosa.
Crypt hyperproliferation was associated with selective changes in PKC isozyme abundance.
To date, 12 PKC isozymes (α, β1, β2, γ, δ, ε, η, θ , ζ, ι, λ, and μ) with differing structure, substrate specificity, cofactor requirements, tissue expression, and localization have been identified. This diversity is believed to create the varied consequences of PKC activation in the same cell and are the reason why individual isozymes are thought to possess distinct and specialized functions in cell signaling (6).
cPKC expression in isolated crypts from normal and TMCH mice.
Individual cPKC isozyme expression was measured in Triton X-100-solubilized crypt extracts prepared from the distal colon of animals at 0–15 days after Citrobacter infection by Western blot analysis. As shown in Fig. 2A, mean cellular PKC-α expression increased 1.6-fold during crypt hyperproliferation. The mean relative cellular optical density of PKC-β1 increased even more dramatically (3.7-fold; n = 3 animals in duplicate). Changes in PKC-β1 expression were significantly different by the third day after inoculation and continued to be so for all subsequent days (Fig. 2 B; atday 3 = 0.6 ± 0.06 and normal = 0.38 ± 0.1; P < 0.001, Student's t-test; n = 6). This increase occurred before the onset of marked crypt cell hyperplasia at day 6 (2, 35). Large twofold increases in crypt-specific were also observed (n = 3 animals in duplicate; Fig.2 C). However, like PKC-α, statistically significant changes were not recorded until nine days after inoculation ( at day 9 . = 0.76 ± 0.07 and normal = 0.38 ± 0.01; P < 0.01; atday 9 = 1.2 ± 0.18 and normal = 0.73 ± 0.08; P < 0.001; n = 6). Thus, unlike PKC-β1, crypt-specific changes in PKC-β2 and PKC-α cellular expression paralleled the mucosal hyperproliferatory response. PKC-γ was undetectable in isolated crypt extracts. Antibody specificity was confirmed by using the relevant antigenic peptide for competition at a 1:2 (wt/wt) peptide-to-antibody ratio (data not shown). β-Actin housekeeping protein antisera were used to normalize variations in protein loading.
Thus, even though in vitro heterologous expression studies had previously reported that cellular PKC-α depletion enhanced cellular CFTR expression (12), we found opposite effects in vivo. PKC-α expression increased in TMCH colonocytes previously shown to express elevated cellular CFTR mRNA levels (38a). To determine whether this increase in protein abundance reflected functional inactivation during hyperproliferation, we measured subcellular cPKC isoform partitioning between cytosolic and membrane fractions in normal and hyperproliferative crypts.
Crypt colonocyte hyperproliferation was associated with enhanced cytosol-to-membrane translocation of all cPKC isoforms.
It is widely accepted that PKC exists in an inactive conformation in the cytosol and that activation of the enzyme results in its translocation from the cytosol to membranes (6). Thus an indication of the native activation status of PKC can be obtained by assessing its distribution among subcellular compartments of the cell.
Figure 3 shows a representative Western blot and corresponding mean densitometric ratio for the membrane-to-cytosolic partitioning (Rm:c) of all three cPKCs detected in isolated crypt cellular extracts from normal and hyperproliferative TMCH mucosa.
PKC-α exhibited a modest change in membrane association during TMCH. The PKC-α Rm:c increased 1.3 ± 0.2-fold compared with normal values but was not significantly different (Fig. 3 B). When cytosolic PKC-α levels were quantified by reprobing the same Western blots (Fig. 3 A) with antibody against β-actin, cytosolic PKC-α levels fell by 27% (decrease = 1.36 ± 0.26-fold). Thus crypt hyperproliferation modestly increased the number of membrane-bound activated PKC-α isozymes and proportionally decreased the size of the inactive cytosolic PKC-α pool. The Rm:cvalues themselves were not fully quantified because of the difficulty of identifying an appropriate housekeeping protein for membrane-bound estimation. We found no evidence that cPKC-α could act as a negative regulator of cellular CFTR mRNA or protein expression in vivo.
PKC-β1 crypt hyperproliferation was associated with enhanced translocation of PKC-β1 to membrane fractions. The PKC-β1 Rm:c increased 3.6 ± 0.5-fold compared with controls and was significantly different (Fig. 3 B; P < 0.01 in duplicates from three animals). When the relative cytosolic abundance of PKC-β1 was estimated on the same blots, the inactive cellular pool of PKC-β1 fell to 40% of control values (cytosolic normal colon = 0.56 ± 0.08 and day 12post-Citrobacter-infected colon = 0.34 ± 0.06). Thus the increase in cellular PKC-β1 isoform abundance recorded before and during Citrobacter-induced mucosal hyperplasia (Fig.2 B) was paralleled by a dramatic rise in its subcellular activation status.
PKC-β2 also exhibited increased membrane translocation during TMCH. PKC-β2 Rm:c increased 2.3 ± 0.26-fold (duplicates from three animals; Fig. 3B). However, variability in the membrane-bound β2 component did not make this change highly significant (P< 0.05; n = 3 animals in duplicate). Hyperproliferating crypts were found to contain four times more cytosolic (inactive) PKC-β2 isoforms on a per-cell basis than their normal counterparts ( normal colon = 0.106 ± 0.05 and day 12post-Citrobacter-infected colon =0.43 ± 0.13). Densitometric analysis revealed that the membrane-bound (nonquantified) PKC-β2 immunoreactive signal was also higher (see Fig. 3 A). PKC-β2 immunoreactivity also partitioned into a Triton X-100 detergent-resistant fraction, and crypt hyperproliferation slightly reduced the size of this pool. Thus crypt hyperplasia during TMCH was associated with both late increases in cellular PKC-β2 abundance (Fig. 2 C) and modest PKC-β2 membrane activation.
nPKC isoform expression during TMCH.
The expression of individual nPKC isozymes was likewise measured in Triton X-100-solubilized crypt extracts prepared from the distal colon of normal and TMCH animals. A panel of isozyme-specific polyclonal antibodies were used as probes for Western blot analysis. The specificity of each antibody was determined by competitive blotting with corresponding immunizing peptides, and their relative abundance was compared with mouse brain nPKC homologs (normalized to β-actin to account for differences in gel loading). Three novel PKC isoforms, δ, ε, and η, were expressed in isolated crypts (Fig.4 A).
PKC-η abundance levels were similar to those found in brain, whereas PKC-δ and -ε were slightly lower (data not shown). PKC-δ expression in isolated crypts remained fairly constant, exhibiting a slight increase throughout the time course of TMCH (1.2-fold atdays 1 and 3, reaching 1.6-fold at day 15; Fig.4 B). Interrupting this slow rise was a significant decrease in cellular expression at day 6 at onset of marked crypt cell hyperplasia (2, 35). PKC-ε relative cellular abundance started to increase at day 1 (1.4-fold) but was not significant untildays 6–12 (2-fold; Fig. 4 B). The relative abundance of PKC-η decreased dramatically during TMCH (levels were 2.1-fold lower at day 6 and thereafter continued to decline;P < 0.05; n = 3 animals in duplicate). Previous studies have shown that differentiated senescent but not undifferentiated proliferating epithelial cells express PKC-η (28). Thus decreases were expected during TMCH when crypt proliferatory rate had increased eightfold and the zone of cellular proliferation was extended throughout the longitudinal crypt axis (38a).
Crypt colonocyte hyperproliferation was associated with modest changes in cytosol-to-membrane translocation of specific nPKC isoforms.
The distribution of crypt-expressed nPKC isozymes was assessed in cytosol and membrane fractions from normal and day 12post-Citrobacter-infected (TMCH) mice (Fig.5). A representative Western blot (Fig.5 A) and corresponding mean densitometric Rm:c for nPKC (Fig. 5 B) demonstrated the following: PKC-δ did not exhibit significant increases in membrane association during TMCH. PKC-δ Rm:c was 1.12 ± 0.4-fold of normal values. However, cytosolic PKC-δ levels (Fig. 5 A), quantified by reprobing the same blots with antisera against β-actin, exhibited an 18% (1.27 ± 0.2-fold) decrease during crypt hyperproliferation, but low levels of expression of this protein made it difficult to determine whether these changes were real. PKC-ε expression changes in TMCH crypts (Fig. 4 B) were paralleled by a concomitant increase in the membrane translocation. The Rm:c increased 2.16 ± 0.15-fold above normal values. This correlated with a 32% decease in the relative cytosolic abundance of PKC-ε during crypt hyperproliferation ( normal colon = 0.48 ± 0.06 and day 12post-Citrobacter infected colon = 0.29 ± 0.05). Thus crypt hyperproliferation was also correlated with both early changes in colonocyte PKC-ε expression and biochemical activation, as measured by this isoform's translocation to membrane fractions. PKC-η detected in isolated crypts exhibited a modest but insignificant increase in membrane association during crypt proliferation. The Rm:c increased 1.3 ± 0.2-fold compared with control values. However, cytosolic expression of this isoform was barely detectable either in normal mice or during crypt hyperproliferation and again were so low as to hinder accurate measurement.
Isoform-specific nuclear partitioning occurs in hyperproliferating crypts.
Our immunological studies identified four phorbol ester-sensitive PKC isoforms with increased cellular expression and intracellular membrane activation status during TMCH: cPKC-α, cPKC-β1, cPKC-β2, and nPKC-ε. To determine whether these PKC isoforms accumulated in other subcellular locations, purified isolated nuclei from normal and TMCH colon were probed by Western blotting (Fig.6). Low levels of nuclear PKC-β1 immunoreactivity were detected in normally proliferating crypts, whereas hyperproliferating crypts exhibited dramatically increased levels of this kinase (Fig. 6). In the nuclei from hyperproliferating mucosa, PKC-β2 levels were too low to quantify effectively (Fig. 6). Densitometric analysis revealed that the PKC-β1 nuclear signal (Fig.6, top) was ∼18.5 ± 4.5% of the total immunoreactivity detected in the pooled Triton X-100-extractable membrane and cytosolic fractions in TMCH crypts (n = 2 animals in duplicate). Nuclear PKC-β2 levels accounted for <2% of this fraction (Fig. 6, top). Nuclear PKC-β1 immunoreactivity was competed by immunizing peptide (data not shown). Purity of the nuclear fraction was estimated by the presence of nuclear lamin B (Fig. 6,middle) and exclusion of monomeric β-actin immunoreactivity (Fig. 6, bottom). The latter was detected only in total solubilized cell extract and in cytoplasmic cellular extracts but not in purified nuclear extract. PKC-α, -β2, and -ε were not detected in normally proliferating crypt nuclei. Although we cannot rule out the possibility that antibody affinity may account for the lack of nuclear signal recorded by these isoforms, the results clearly demonstrated that a significant proportion of PKC-β1 was translocated into the nucleus of proliferating colonocytes.
The PKC-β2 homolog of PKC-β1 remained predominantly cytoskeletal-associated in both normal and hyperproliferative crypt mucosa.
Triton X-100-insoluble RIPA-solubilized extracts from normal andday 12 post-Citrobacter-infected mouse whole distal colon revealed significant PKC-β2 immunoreactivity (Fig.7). In the cytoskeletal fraction of both normal and hyperproliferating crypts, PKC-β2 holoenzyme (80 kDa) accounted for <5% of immunodetectable densitometric signal (Fig. 7). The majority (>95%) of PKC-β2 was detected at ∼50 kDa (Fig. 7). Both bands were competed away with the PKC-β2 COOH-terminal peptide (amino acids 657–673; peptide-to-antibody ratio = 1:1 wt/wt) against which the antibody was raised (Fig. 7, middle). The 50-kDa immunoselective band was occasionally detected in Triton X-100 solubilized fractions (3 out of 12 blots), where it comprised between 1 and 12% of the 80-kDa PKC-β2 signal (Fig. 7; n = 12), which suggests it is a proteolytic product of the holoenzyme. When densitometric analysis of the total cytoskeletal and membrane-bound PKC-β2 immunoreactivity were compared, modest increases in Triton X-100-extractable signal during TMCH (Fig. 7; see also Fig. 3) were reflected by modest decreases in cytoskeletal signal (Fig. 7). This movement may account in part for the modest increases in cellular expression levels (Fig. 2) and cytosol-to-membrane partitioning (Fig.3) seen for PKC-β2 during TMCH. No PKC-β1 immunoreactivity was detected in the cytoskeletal fractions of either normal or hyperproliferating mucosa (Fig. 7, bottom). These findings were confirmed in four uninfected/infected animals in duplicate. Cytoskeletal levels of nPKC-ε were not quantified.
Citrobacter-induced crypt extension and elevated cellular CFTR mRNA levels were inhibited by Ro-32–0432 but not by its pharmacologically inactive analog Ro-31–6045.
The possibility that nuclear translocated PKC-β1 was responsible for TMCH-dependent increases in cellular CFTR message levels was investigated by dosing mice with the highly selective cPKC antagonist bisindolylmaleimide XI ⋅ hydrochloride (Ro-32–0432). Ro-32–0432 was given periorally one day before Citrobacter inoculation, and booster doses were administered daily over the following two weeks. Ro-32–0432 has previously been shown to be an effective in vivo antiarthritic agent capable of inhibiting phorbol ester-induced paw edema in rats (ED50 = 11 mg/kg; Ref. 5) but had not before been used in mice. This agent inhibits endogenous cPKCs with 100- to 10,000-fold or more selectivity over a variety of other serine/threonine kinases in vitro (inhibition constant =10–30 nM). However, higher doses have been found to be required in vivo because endogenous ATP competes with this agent for the ATP hydrolysis site of PKC (5). Groups of five mice were dosed at 10 mg/kg body wt with Ro-32–0432 plus carrier or with carrier alone (4% succinylated gelatin; see Table1).
Cellular CFTR mRNA levels in day 12post-Citrobacter-infected mice exposed to Ro-32–0432 plus carrier were not statistically different from uninfected litter mates, whereas Citrobacter-infected mice treated with carrier alone exhibited three- to fivefold higher message levels (Table 1). Ro-32–0432 clearly inhibited the rise in cellular CFTR message levels recorded during crypt hyperproliferation as well as inhibiting the increase in mucosal crypt length associated with TMCH (P< 0.001; n = 5 animals; Table 1 and Fig.8).
Bright-field images of hematoxylin and eosin-stained mats of crypts isolated from age-matched normal mice (Fig. 8 A), day 12post-Citrobacter-infected mice (Fig. 8 B), and day 12 post-Citrobacter-infected mice treated daily periorally with Ro-32–0432 (Fig. 8 C) taken at the same magnification. The >2-fold increase in crypt length seen during TMCH (38a) was largely prevented by Ro-32–0432 administration. Crypt length in the latter instance reverted to 110% of normal (n = 6 animals). Crypts were, however, thinner, and some disruption of the surface mucosa, consistent with Citrobacter infection, was always observed (compare Fig. 8, A and B). When specific cPKC activity was measured in crypt extracts from these three experimental conditions, the dramatic (in this instance 14-fold) increase in membrane-bound activity seen during TMCH was largely prevented (Fig. 8 D; a 55% inhibition of stimulated membrane-bound cPKC activity was recorded; n = 6 animals in duplicate). Similar doses of Ro-32–0432 (∼2.5 μM) failed to affect the growth rate of an overnight culture of Citrobacter in Luria-Bertani broth (data not shown). These results supported our hypothesis that phorbol ester-sensitive PKC activation underlies both TMCH and TMCH-dependent elevations in cellular CFTR expression.
The cell biology of CFTR expression in vitro is dominated by reports investigating the negative role of phorbol ester-sensitive PKC signaling (reviewed in Ref. 26). We therefore examined PKC expression and function in the TMCH model in which cellular CFTR levels increased (38a).
Cellular roles of individual PKC-β isoforms in proliferation-dependent changes in CFTR expression.
Although PKC activation has been implicated in colonic epithelial cell proliferation in response to various stimuli, very little information regarding the role of individual PKC isoforms exists. Rat and human colonic mucosa have been shown to express both mRNA and protein for six PKC isoforms (α , β , δ , ε, η, and ζ) (13, 20). In the present study we concentrated on phorbol ester-sensitive conventional and novel PKCs in which a distinct pattern of cPKC-α, -β1, -β2, -ε, -δ, and -η isozyme expression and activation status was recorded.
Although isolated crypt colonocytes increased their levels of cPKC-α during TMCH (Fig. 2), no significant changes in membrane translocation were recorded (Fig. 3) and PKC-α was undetectable in nuclear extracts from either control or hyperproliferating crypts (data not shown). Thus we found no evidence that PKC-α acts as either a negative regulator of cellular CFTR expression, as suggested by in vitro studies (12), or as positive regulator of CFTR expression in the in vivo TMCH gastrointestinal epithelium.
Various studies have shown that PKC-β can regulate epithelial cell cytokinetics. The role of PKC-β1 has been demonstrated in colonic tumor cell lines in which growth cessation correlated with a selective 3-fold decrease in PKC-β1 abundance, a 10-fold decrease in membrane-bound cPKC activity, and loss of mitogen-kinase cell signaling (22). PKC-β heterologous overexpression has been shown to cause dedifferentiation, proliferation, and an enhanced growth of colonocytes in athymic mice (34). Marked increases in both cPKC-β1 and -β2 abundance (Fig. 2) and membrane translocation (Fig. 3) indicated that these isoforms contributed the most to the 12-fold increase in particulate cPKC activity recorded in hyperproliferating crypts (Fig.1). Temporal differences in their cellular expression and subcellular organelle location allowed us to further distinguish roles for these two isoforms.
Control crypts normally exhibited low cytosolic and membrane levels of PKC-β1 (Figs. 2 and 3). However, colonocyte levels of PKC-β1 increased at the onset of TMCH before both PKC-α, PKC-β2, and gross changes in mucosal mass occurred. In hyperproliferative mucosa, ∼20% of PKC-β1 immunoreactivity was also relocalized into the nucleus (Fig. 6), whereas <2% was detected in nuclear extracts from normal crypts. These findings predict a role for PKC-β1 in early nuclear events involved in crypt proliferation and accompanying increases in colonocyte CFTR expression.
Increased crypt levels of PKC-β2 followed six days later than PKC-β1 in the TMCH model (Fig. 2), at a time when dramatic increases in PKC-α abundance and mucosal mass occurred. Hyperproliferating crypts at day 12 contained 2.3-fold higher levels of membrane-bound PKC-β2 (Fig. 3). Additionally, this isoform was partitioned into cytoskeletal fractions under both normal and hyperproliferative conditions (Fig. 7), whereas vanishingly low levels of nuclear immunoreactivity were detected in both normal and hyperproliferating crypts (Fig. 6). Given these findings, PKC-β2 appears less likely than PKC-β1 to mediate early proliferation-dependent nuclear events leading to enhanced cellular CFTR mRNA expression.
Previous studies relating the expression and cellular roles of PKC-δ have shown both positive and negative effects on cell division (17). In general, expression is seen predominantly in postmitotic cells of the upper crypt and surface mucosa (39). In our model, colonocytes increased their levels of PKC-δ slightly during TMCH (Fig. 4). However, no significant changes in membrane translocation were recorded during TMCH, and PKC-δ was undetectable in the purified nuclear extracts from either control or hyperproliferating mucosa (data not shown). Thus PKC-δ may not be critical to hyperproliferation-dependent changes in cellular CFTR expression.
Although PKC-ε is expressed at very low levels in all normal tissues except for brain, it is expressed at high levels in several hemopoietic cell lines and tumors (24). During periods of mucosal hyperproliferation, PKC-ε expression in the crypt increased, with concomitant increase in membrane translocation and hence cellular activation status (see Figs. 4 and 5). However, despite these changes, nuclear translocation of PKC-ε was not observed. This was unexpected because PKC-ε has been established in other systems to possess oncogenic potential and, furthermore, the overexpression of PKC-ε had been implicated in ras-mediated signal transduction during neoplastic transformation of the colonic epithelium (31). However, TMCH does not cytologically correspond to neoplasia or tumorigenic transformation (3), and the lack of nuclear PKC-ε partitioning may reflect this fact. Therefore, despite being the PKC isoform with best-described oncogenic potential, it is difficult to postulate a direct role for PKC-ε in the regulation of TMCH-dependent nuclear CFTR gene transcription. However, studies in several nonepithelial cell lines have shown that both mitogenic and differentiating factors (27) may use PKC-ε signaling at the cellular plasma membrane to affect responses at the genome level (33, 40). Thus we cannot exclude the possibility that PKC-ε can activate protein kinase cascades within the nucleus without being physically present there. In vivo transgenic or epigenic approaches will be required to determine whether this isoform actively participates in the CFTR expression response.
No evidence was found for the participation of this isoform in proliferation-dependent changes in cellular CFTR expression during TMCH.
In vivo effects of PKC signaling on CFTR abundance differ from those proposed in vitro.
cPKCs are identified as the major cellular receptor for active phorbol esters (8). Phorbol esters have been shown to acutely increase electrogenic Cl− secretion in ex vivo colonic mucosa (9, 15), predominantly through cPKC activation and secondary PGE2 production (37), and to directly stimulate CFTR current generation in isolated epithelial cells (4, 19). Phorbol ester-dependent stimulation of cPKC activity in vivo also induces enhanced proliferation (7, 11) within the same crypt cell populations we have shown to overexpress CFTR protein (38a). However, phorbol esters and calcium ionophores are also known to downregulate cellular CFTR mRNA levels within 1–12 h [by as much as 80% within 2 h (36), but found to downregulate less by other groups (1, 38)] in vitro and, in a separate corroborating in vitro study, ras-induced PKC-α activation has been reported to both decrease cellular CFTR mRNA levels and inhibit cAMP-dependent Cl− secretion (12). How then can the negative effects of phorbol esters on cellular CFTR expression in vitro be reconciled with the positive effects of cPKC and nPKC activation recorded in the TMCH model in vivo?
One explanation is that phorbol esters may not only stimulate cellular PKC activity in vitro but also promote unrelated events that profoundly inhibit gene transcription. The CFTR promoter contains a bona fide cAMP-responsive element-binding (CREB)/ATF-binding cAMP-response element that regulates basal and cAMP-stimulated gene transcription (23, 32). Ras-overexpression in vitro has been shown to negatively affect this type of cAMP-response element (16), and both phorbol esters and calcium ionophores have been shown to induce both the cytoplasmic trapping and to alter the phosphorylation status/composition of nuclear CREB/ATF DNA-binding proteins interacting with this element to prevent gene expression (21, 30). This form of inhibitory cross talk has been reported for the human tissue-type plasminogen (t-PA) gene, whose 5′-promoter region contains an identical variant cAMP-response element to that found in CFTR (10). Phorbol ester effects on CFTR expression in vitro may be masked by similar inhibitory phenomena.
In summary, we have shown that cPKC activation during TMCH was correlated with enhanced CFTR gene expression and that pharmacological inhibition of cPKC function in vivo prevented both TMCH and its subsequent effects on cellular CFTR mRNA levels (Fig. 8, Table 1). We hypothesize that PKC-β1 is a prime candidate for cPKC-mediated effects on CFTR expression in this model and suggest that PKC-β1 may exert its effects within the nucleus, as has been implicated for positive PKC effects on cAMP-responsive genes (18, 29) and as originally proposed for PKC-dependent regulation of CFTR (23). Other candidate PKCs with roles originating within the cytoplasm (cPKC-β1 and nPKC-ε) may also participate in this response in yet-undefined ways. Further investigation into these three signaling molecules should provide important insights into how both CFTR expression and subcellular location (38a) are regulated in vivo and may lead to new approaches for the treatment of the disease cystic fibrosis, which is characterized by low levels of apical cellular plasma membrane CFTR expression (14).
This work was supported by funds from the Cystic Fibrosis Foundation, the American Institute for Cancer Research, and the Cancer Research Foundation of America.
Address for reprint requests and other correspondence: A. P. Morris, Dept. of Internal Medicine, Division of Gastroenterology, Hepatology, and Nutrition, The Univ. of Texas Health Science Center at Houston, Medical School, Houston, TX 77030 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
- Copyright © 2000 the American Physiological Society