The role of specific PKC isoforms in the regulation of epithelial Cl− secretion by Ca2+-dependent secretagogues remains controversial. In the developing rabbit distal colon, the bile acid taurodeoxycholate (TDC) acts via intracellular calcium to stimulate Cl− transport in adult, but not in young, animals, whereas the PKC activator phorbol dibutyrate (PDB) stimulates Cl− transport at all ages. We tested the hypothesis that specific PKC isoforms account for the age-specific effects of TDC. The effects of conventional (cPKC) and novel (nPKC) PKC-specific inhibitors on TDC- and PDB-stimulated Cl− transport in adult and weanling colonocytes were assessed by using 6-methoxy-quinolyl acetoethyl ester. In adult colonocytes, the cPKC inhibitor Gö-6976 inhibited PDB action but not TDC action, whereas the cPKC and nPKC inhibitor Gö-6850 blocked both TDC and PDB actions. Additionally, rottlerin and the PKC-δ-specific inhibitor peptide (δV1-1) inhibited TDC- and PDB-stimulated Cl− transport in adult colonocytes. Rottlerin also decreased TDC-stimulated short-circuit current in intact colonic epithelia. Only Gö-6976, but neither rottlerin nor δV1-1, inhibited PDB-stimulated transport in weanling colonocytes. Colonic lysates express PKC-α, -λ, and -ι protein equally at all ages, but they do not express PKC-γ or -θ at any age. Expression of PKC-β and PKC-ε protein was newborn>adult>weanling, whereas PKC-δ was expressed in adult but not in weanling or newborn colonocytes. TDC (1.6-fold) and PDB (2.0-fold) stimulated PKC-δ enzymatic activity in adult colonocytes but failed to do so in weanling colonocytes. PKC-δ mRNA expression showed age dependence. Thus PKC-δ appears critical for the action of TDC in the adult colon, and its low expression in young animals may account for their inability to secrete in response to bile acids.
- epithelial chloride transport
- signal transduction
neurohumoral agents, which act via intracellular calcium, are critical in fine-tuning epithelial fluid transport, and they are often linked to the protein kinase C (PKC) cascade (18, 38). The pathophysiological consequence of aberrations in this signaling is diarrhea. One class of Ca2+-dependent secretagogues is bile acids, which are essential for fat digestion and absorption in the intestinal lumen. Under physiological conditions, the mammalian intestine recycles the majority of the bile acids (≈95%) back to the liver. While the colon can effectively handle the normal levels of bile acids that are excreted, excess bile acids in the colon cause fluid loss, and prolonged exposure to bile acids promotes colon cancer (3, 17).
Conjugated or unconjugated dihydroxy bile acids, such as chenodeoxycholic acid or deoxycholic acid, stimulate Cl− secretion in a variety of mammalian colonic preparations (20–22, 27, 44, 50, 69), in pancreatic ducts (46), in gallbladder-derived epithelium (12), and in cholangiocytes (60). Bile acids have a plethora of other actions, ranging from carcinogenesis to mucin secretion, and many of these functions involve PKC. There are at least 12 PKC isoforms, which are grouped into three major classes: conventional (cPKC; α, βI, βII, and γ), novel (nPKC; δ, ε, η, θ, μ, and υ, which are viewed as a separate subclass or as PKD), and atypical (aPKC; ζ and λ/ι) (29, 45). A role for different PKC isoforms in bile acid action has been implicated in cholangiocarcinoma growth (1), in mucin secretion in gallbladder (12), in colon carcinogenesis models (33), and in the feedback inhibition of cholesterol 7α-hydroxylase in hepatocytes (64). However, it is unclear whether bile acids use the PKC pathway to stimulate Cl− secretion. Interestingly, the pharmacological activators of PKC, phorbol esters, have been shown to regulate Cl− secretion either by activating it (5, 38, 47, 57) or by inhibiting it (5, 8, 14, 72). These regulatory differences appear to be species, cell type, and isoform specific. Interpretation is further complicated by the fact that the times of exposure to phorbol esters can have differing effects since prolonged exposure to phorbol esters downregulates the action of some, but not all, PKC isoforms. Thus, in primary rabbit and human colonocytes (57) and in the human colon carcinoma cell line HT-29cl.19A (2), short-term exposure (5 min) to phorbol esters increases Cl− secretion. In contrast, in the T-84 cell line, phorbol esters do not affect basal Cl− secretion, but they attenuate cAMP-stimulated Cl− secretion by a mechanism linked to the PKC-ε-dependent internalization of the basolateral Na+-K+-2Cl− cotransporter (NKCC) (19). In rats, estradiol inhibits female, but not male, rat colonic Cl−secretion via the action of PKC-δ (14). In pancreatic ductal cells, short-term exposure increases the magnitude of cAMP-stimulated CFTR currents, whereas prolonged exposure (consistent with PKC downregulation) slowed CFTR current run-down, which suggests that PKC affects the stability of CFTR (72). Thus, in an examination of bile acid-mediated Cl− secretion, the contrasting effects of different PKC isoforms have to be considered.
We (69) and others (21, 22) showed that, in colonic epithelia, bile acids activate phospholipase C (PLC) to increase inositol 1,4,5-trisphosphate (IP3) and [Ca2+]i. We also showed that the secretagogue action of bile acids is segment specific, occurring in the rabbit distal, but not proximal, colon (69). Moreover, this effect is age specific, with bile acids evoking a response only in colonocytes of adult animals but not in those of weanling or newborn animals (20, 69). Interestingly, the recycling of bile acids is also age dependent, with the neonatal mammal showing little and only passive absorption in the ileum (67). Therefore, the refractoriness of the young animal to the secretory action of bile acids may serve as a protective mechanism against potentially deleterious luminal levels of this secretagogue.
In elucidating the basis of this refractoriness, we identified that the age dependency shown by bile acids extends to other Ca2+-dependent neurohumoral agents but that the Ca2+ ionophore A23187 (69) and phorbol esters (20) stimulate Cl− transport at all ages. The refractoriness in the young animal to all tested Ca2+-dependent secretagogues and bile acids was due to the inability to generate IP3 and to increase [Ca2+]i (69). Interestingly, phorbol esters, but not bile acids, activate cPKC enzyme activity in colonocytes in vitro (69). This led us to conclude that at least cPKCs may not be involved in the secretagogue action of bile acids (69). However, there is increasing evidence that the nPKCs and aPKCs play substantial roles in several biological processes, including in the actions of bile acids in other tissues (24, 29, 64).
The present study investigated whether bile acid-stimulated Cl− secretion in the rabbit colon involves any of the PKC isoforms and whether the regulation of this cascade is age dependent. The findings, for the first time, indicate that PKC-δ is involved in bile acid-stimulated colonic Cl− secretion and that the expression of PKC-δ protein is a critical step in the age-related refractoriness to bile acid signaling in the mammalian colon.
MATERIALS AND METHODS
Tissue culture media, Ham's F-12 nutrient mixture, fetal bovine serum, TRIzol reagents, SuperScript II RNase H− reverse transcriptase, oligo(dT)12-18 primer, RNAseOUT, and recombinant ribonuclease inhibitor were obtained from Invitrogen (Carlsbad, CA); sterile lactated Ringer (LR) solution from Baxter International (Deerfield, IL); iQ SYBR Green supermix and gel electrophoresis supplies from Bio-Rad (Hercules, CA); RNA-later from Ambion (Austin, TX); 6-methoxy-quinolyl acetoethyl ester (MQAE) from Molecular Probes (Eugene, OR); diphenylamine-2 carboxylate (DPC) from Aldrich (Milwaukee, WI); and taurodeoxycholate (TDC), Gö-6983, Gö-6850, and Gö-6976 from Calbiochem (San Diego, CA). Furosemide, phorbol dibutyrate (PDB), carbachol, chelerythrine, rottlerin, RedTaq polymerase, oligonucleotide primers, 10 mM dNTP mix, and DNase kits were obtained from Sigma-Aldrich (St. Louis, MO). Mouse monoclonal anti-PKC-α, -β, -γ, -δ, -ε, -θ, -ι, and -λ were purchased from Transduction (Lexington, KY). Goat anti-mouse IgG secondary antibodies were from Jackson ImmunoResearch (West Grove, PA). SuperSignal West Pico chemiluminescent substrate kit was from Pierce Biotechnology (Rockford, IL). PKC activity assay kit was purchased from Upstate Cell Signaling Solutions (Waltham, MA). PKC-δ substrate was from Calbiochem; [γ-32P]ATP was obtained from MP Biomedicals (Irvine, CA). PKC-δ-specific peptide inhibitor (δV1-1) and nonspecific (scrambled sequence) peptide were a kind gift from Karen Ridge, Northwestern University Chicago, Illinois. These peptides, which were linked (via an NH2-terminal Cys-Cys bond) to the Drosophila Antennapedia homeodomain-derived carrier peptide, were synthesized and purified at the Stanford Protein and Nucleic Acid Facility. All other reagents were obtained from Sigma-Aldrich or Fisher Scientific (Hanover Park, IL) and were of analytical grade.
Tissue procurement and cell isolation.
New Zealand White adult (≥6 mo old), weanling (25–28 days old), and newborn (7–9 days old) rabbits were purchased from New Franken Research Rabbits (New Franken, WI). Animals were housed in the Biological Resources Laboratory, University of Illinois, an accredited Association for Assessment and Accreditation of Laboratory Animal Care facility. The care and handling of animals and tissue processing for our experiments were approved by the Institutional Animal Care Committee. Distal colon, from splenic flexure to rectum, was excised, and mucosal sheets were separated from underlying muscle by use of blunt dissection. Colonocyte isolation and culture (18–24 h) from adult and weanling animals were carried out as previously described (6, 53, 57, 69). Briefly, tissues were collected and isolated in oxygenated LR solution with 5% dextrose (LRG) and antibiotics (ABX: 25 μg/ml ampicillin, 125 μg/ml penicillin, 270 μg/ml streptomycin, and 1.25 μg/ml amphotericin B). For the isolation, colonic mucosa was incubated at 37°C in LRG + ABX, which contained 0.03% collagenase, 0.1% pronase, and 0.07% dithiothreitol (DTT) for adult samples (90 min) and 0.015% collagenase, 0.05% pronase, and 0.023% DTT for weanling samples (60 min) (6, 53). The resulting cell suspension was enriched for crypt colonocytes by sequential centrifugation twice at 4,000 g for 5 min and twice at 400 g for 15 min. The final pellet was resuspended in 25 ml Ham's F-12 media containing 20% fetal bovine serum, 0.5 U/ml insulin, 4 mM l-glutamine, 1 mM hydrocortisone, 500 μM selenium, 1 mM sodium butyrate, and ABX antibiotics, was transferred to a T75 flask, and was incubated 18–24 h at 37°C.
The membrane-permeable halide-sensitive fluorescent probe MQAE was used to assess Cl− transport in isolated colonocytes as previously described (20, 57, 69). Isolated colonocytes were loaded with 10 mM MQAE (5 min at room temperature and 90 min on ice) in buffer A containing (in mM) 110 NaCl, 1 MgCl2, 1 CaCl2, 5 dextrose, 50 mannitol, 1 K2SO4, and 5 HEPES, pH 7.4. Cells were then Cl− depleted for 30 min in buffer B containing (in mM) 110 Na+ isethionate, 1 MgSO4, 5 dextrose, 50 mannitol, 1 K2SO4, 1 CaSO4, and 5 HEPES, pH 7.4. An Alphascan spectrofluorometer (Photon Technologies International, Princeton, NJ) was used to measure fluorescence at 350-nm excitation and 460-nm emission. Each test condition was run in triplicate, with ∼104 cells/assay cuvette. Cl− influx was measured in triplicate under basal conditions ± secretagogues ± Cl− transport inhibitors (50 μM DPC + 10 μM furosemide) ± various PKC inhibitors. We previously showed that DPC, a Cl− channel inhibitor, and furosemide, an NKCC inhibitor, can be used in this system to determine the combined contributions of CFTR and NKCC to the Cl− flux in the presence and absence of secretagogues (20, 57). The concentrations of PKC inhibitors were based on their known efficacy in T-84 (62) and Caco-2 cell systems (58, 59). The following PKC inhibitors were used: chelerythrine (2 μM), a general PKC inhibitor; Gö-6983 (100 nM), an inhibitor of most PKC isoforms, with the exception of PKC-μ; Gö-6850 (5 μM), an inhibitor of cPKCs and nPKCs; Gö-6976 (1 μM), an inhibitor of cPKCs; rottlerin (10 μM), commonly used as an inhibitor of PKC-δ; and peptide inhibitor δV1-1 (0.5 μM, cross-linked via an NH2-terminal Cys-Cys bond to carrier peptide), a specific inhibitor of PKC-δ. As a negative control, a nonspecific control peptide (scrambled sequence cross-linked to carrier peptide) was included in the δV1-1 studies.
Colonocytes were preincubated with δV1-1 or with control peptide for 30 min and with the other PKC inhibitors for 5 min each before the start of the transport assay. As described in detail previously, transport was calculated as JCl = [Fo/(KCl × F2)] × (dF/dt) (57), where Fo and F2 are initial and final fluorescence intensities and KCl is the Stern-Volmer constant. The rate of transport was expressed as transport inhibitor (DPC + furosemide)-sensitive Cl− influx (in mM/s). The secretagogue doses (50 μM TDC and 1 μM PDB) had previously been shown to elicit maximal Cl− transport in rabbit distal colonocytes (20, 69).
The distal colon of adult rabbits was excised, opened along the mesenteric line, and placed in cold oxygenated LR solution with 10 mM glucose. The mucosal layer was stripped of underlying muscle, mounted in modified Ussing chambers (area: 0.33 cm2; Physiologic Instruments, San Diego, CA), and bathed with oxygenated (95% O2-5% CO2) buffer C (5 ml/reservoir) of the following composition (in mM): 141.8 Na+, 5.4 Cl−, 3.0 PO43−, 1.2 Ca2+, 1.2 Mg2+, 21.0 HCO3− (pH 7.4), 5.4 K+, and 10 d-glucose, at 37°C. Tissues from the same intestinal segment and animal were paired according to comparable tissue resistance (Rt; Ω·cm2, ±25%). Tissue conductance (Gt; mS/cm2) and Rt were monitored throughout the experiment. Transmural short-circuit currents (Isc; μA/cm2) were measured by using the automatic voltage clamp apparatus (VCC-MC6; Physiologic Instruments) as described previously (71). Tissues were allowed to stabilize for ≈30 min until the Isc and Rt reached plateau levels. Four tissues were mounted from each animal, two control (DMSO) and two treated with rottlerin (10 μM). Reagents were added to both the mucosal and serosal bathing solutions 30 min before the bilateral addition of TDC (100 μM). For dose-response experiments, increasing concentrations of TDC were added sequentially (1–500 μM). Maximal increases in Isc were observed at 50–100 μM TDC (data not shown). At the end of each experimental run, carbachol (100 μM) was added to assess tissue responsiveness. This was confirmed in three experiments by measuring the Isc response to forskolin (10 μM) added at the end of the experimental run.
Protein preparation and assay.
With the use of previously published methods (71), colonic mucosal epithelia were homogenized and sonicated in buffer D containing (in mM) 1 EDTA, 2 MgCl2, 5 β-mercaptoethanol, 1 DTT, 25 Tris·HCl, pH 7.4, and 1 μg/ml each of leupeptin, pepstatin, and aprotinin. The homogenate was centrifuged at 3,000 × g for 10 min to obtain a postnuclear supernatant (total lysate), which was stored at −80°C until used. For positive controls, rabbit and rat brain homogenates were similarly prepared. Total protein was measured via the Bradford method (Bio-Rad).
SDS-PAGE and Western blot analysis.
As described previously (69), the method of H. Towbin was adapted for SDS-PAGE and Western blot analysis. Briefly, samples (100 μg protein/lane for rabbit colonic epithelial samples, 30 μg/lane each for rabbit brain and Jurkat cell lysates, and 10 μg/lane for rat brain) were separated on 7.5% polyacrylamide gels and were transferred to an Immobilon polyvinylidene difluoride transfer membrane (Millipore, Bedford, MA) at 250 mA for 2.5 h in transfer buffer containing 25 mM Tris, 192 mM glycine, 20% methanol, and 0.1% SDS. All subsequent procedures were carried out in Tris-buffered saline (TBS-T) containing 50 mM Tris·HCl (pH 7.4), 150 mM NaCl, and 0.05% Tween 20. The membranes were blocked with Blotto (5% Carnation nonfat dry milk) for 1 h at room temperature and were then incubated with the primary antibody in the blocking agent overnight at 4°C. The antibodies used were as follows: mouse anti-PKC-α (1:1,000), mouse anti-PKC-β (1:250), mouse anti-PKC-γ (1:1,000), mouse anti-PKC-δ (1:500), mouse anti-PKC-ε (1:1,000), mouse anti-PKC-θ (1:250), mouse anti-PKC-ι (1:250), and mouse anti-PKC-λ (1:250). After incubation, the blots were washed 3× for 15 min in TBS-T and were subsequently incubated with horseradish peroxidase-conjugated secondary antibody (1:1,000) at room temperature for 1.5 h. After the secondary antibody, the blots were washed 3× 15 min in TBS-T at room temperature, and the specific antibody-antigen interaction was visualized by using a SuperSignal West Pico chemiluminescent substrate kit. As reported previously (51), to ensure equal loading and transfer, the blots were stained with Ponceau S (10%), and the gels were stained with Coomassie blue after electrotransfer. The gels posttransfer revealed no proteins, and the blots showed uniform staining with Ponceau S. In addition, for each set of protein samples, the blots were stripped and reprobed with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibody, and these blots did not exhibit age-specific differences in expression, which suggests that the loading was uniform and differences in PKC isoform expression were specific.
PKC-δ activity assay.
Specific PKC-δ enzymatic activity was measured by custom designing a modification of the Upstate Cell Signaling Solution PKC activity assay kit. The original kit is composed of a PKC lipid activator, PKA/CaMK inhibitors, and a generic PKC substrate. We substituted the generic PKC substrate with a specific PKC-δ peptide substrate, which has been shown to be phosphorylated by PKC-δ, but not by PKC-ε and -η, and only slightly phosphorylated by PKC-α, -β, and -γ (37). We therefore also added Gö-6976 to inhibit any cPKC activities. Adult and weanling distal colonocytes were isolated and incubated with 100 μM TDC or 10 μM PDB for 5 min. After the incubation, the buffer was immediately replaced with buffer D, and the cells were snap frozen in liquid nitrogen. The samples were thawed on ice, sonicated, and centrifuged at 2,000 g for 5 min to obtain a postnuclear supernatant. Membrane and cytosolic fractions were obtained from this supernatant by ultracentrifugation at 100,000 g for 30 min at 4°C as described previously (69). Equal amounts of protein from control and treated samples (10–50 μg) were incubated with PKC lipid activator, PKA/CaMK inhibitors, Gö-6976, PKC-δ substrate, and [γ-32P]ATP (10 μCi) for 10 min at 30°C. Samples (25 μl) were then spotted on P81 phosphocellulose paper and were washed three times with 0.75% phosphoric acid and once with acetone. 32P incorporated into the PKC-δ substrate was determined by scintillation spectrometry using a Tri-Carb liquid scintillation analyzer (Perkin Elmer, Boston, MA) and Opti-Fluor scintillation cocktail (Perkin Elmer, Shelton, CT).
RNA isolation and real-time RT-PCR.
Total RNA was isolated from rabbit colonic mucosa by using TRIzol reagent according to the manufacturer's instructions. Total RNA (1 μg) was treated with DNase and was reverse transcribed by using oligo(dT)12–18 primer and Superscript II reverse transcriptase. For each real-time PCR reaction (25 μl), cDNA synthesized from 50 ng of RNA was mixed with 2× iQ SYBR Green supermix containing 600 nM specific primers. The oligonucleotide primer sequences are listed in Table 1.
The PCR reactions were carried out in duplicate or triplicate, and the samples were analyzed on a Bio-Rad MyiQ single color real-time PCR detection system. The PCR products were sequenced at the Marine DNA Sequencing Center (Mount Desert Island Biological Laboratory, Salisbury Cove, ME) to verify amplification of the desired target gene. The target genes examined were PKC-δ, CFTR, NKCC, and, in preliminary studies, PKC-ζ. GAPDH was used as the internal reference. Melting curve analyses were done for target genes and GAPDH to confirm the product specificity. The 2−ΔΔCT method, according to Pfaffl (48), was used to compare the relative mRNA level of target genes, where −ΔΔCT = −[ΔCT,sample − ΔCT,calibrator], and ΔCT = CT,target gene − CT,GAPDH. For each real-time PCR experiment, a newborn sample was selected as the calibrator sample (also called reference sample), and all other samples were normalized to this when relative quantification was performed. The mRNA expression in this newborn sample was set as 1 (2°), while the expressions in the other newborn, weanling, and adult samples in the same PCR run were calculated as n-fold of this newborn sample. Real-time PCR amplification efficiencies for target genes and GAPDH were calculated to make sure that the amplification efficiencies were approximately equal.
Data are expressed as means ± SE. For Cl− transport and PKC-δ assay experiments, an n = 1 represents samples from either one adult rabbit or two pooled weanlings. For Western blot analysis, an n = 1 represents samples from one adult rabbit, one weanling, or two pooled newborns. For the real-time PCR experiments, each n value depicts a separate animal. Student's t-test was used to determine statistical significance between means within an experiment, and analysis of variance followed by Tukey's test was used to determine significance for experiments with >2 means. In all cases, P ≤ 0.05 was considered significant.
Effect of PKC inhibitors on TDC action in adult rabbit distal colon.
We previously showed that, in rabbit colonocytes, PDB, but not TDC, increases cPKC activity in vitro (69) and that the effects of PDB are not age specific (20). However, neither of these studies examined whether perturbation of the PKC signaling pathway would interfere with TDC action. We first examined whether inhibition of PKC activity by the broad-specificity PKC inhibitor chelerythrine or Gö-6983 affected TDC (50 μM)-stimulated Cl− transport in adult distal colonocytes. In all the inhibitor studies, PDB (1 μM) was used as a positive control. Chelerythrine (IC50 ≈ 0.66 μM) (32) inhibits all PKC isoforms, while Gö-6983 (IC50 ≈ 7–60 nM) (30) inhibits c-, n- and aPKCs but not PKC-μ (IC50 > 20 μM). As shown in Fig. 1, when isolated adult colonocytes are preincubated (5 min) with either chelerythrine (2 μM; Fig. 1A) or with Gö-6983 (100 nM; Fig. 1B), both TDC- and PDB-stimulated Cl− transport are inhibited to basal levels. These compounds do not affect basal transport, which confirms that they specifically affect furosemide- and DPC-sensitive transport stimulated by TDC or PDB. The above data indicate that activation of PKC(s), but not PKC-μ, is involved in the Cl− secretory response to TDC and, not surprisingly, in that of phorbol esters.
To dissect the roles of c-, n- and aPKCs, isoform-specific inhibitors were used. The compound Gö-6850 inhibits c- and nPKCs but not aPKCs (68). As seen in Fig. 2A, Gö-6850 (5 μM) inhibited both TDC- and PDB-stimulated Cl− transport to basal levels. In marked contrast, Gö-6976 (1 μM), an inhibitor that specifically affects cPKCs (42), had no effect on TDC action but inhibited PDB-stimulated Cl− transport to basal levels (Fig. 2B). Again, basal transport was not affected by either of these inhibitors. These data imply that, whereas PDB action involves both c- and nPKCs, the action of TDC may involve nPKCs but not cPKCs.
Of the four nPKC isoforms, -δ, -ε, -θ, and -η, PKC-δ has been implicated in a variety of functions. Rottlerin, which inhibits at the ATP binding site (31), has been extensively used as a fairly specific inhibitor of PKC-δ action (59). As shown in Fig. 3, rottlerin (10 μM) had no effect on basal levels of Cl− transport. However, rottlerin inhibited both TDC- and PDB-stimulated Cl− transport, which suggests a role for PKC-δ in TDC action. Although our cumulative evidence suggests that rottlerin specifically inhibits PKC-δ in rabbit colonocytes (see discussion), rottlerin has been implicated to have non-PKC-δ effects in a few cell types. Therefore, we examined the effects of a specific PKC-δ inhibitor, δV1-1. The specifically designed δV1-1 peptide blocks PKC-δ translocation and activation in a variety of cells, including alveolar epithelial cells (55). As shown in Fig. 3, δV1-1 had no effect on basal levels of Cl− transport, but it inhibited both TDC- and PDB-stimulated Cl− transport. A control scrambled peptide had no effect on basal, TDC-, or PDB-stimulated Cl− transport. These studies confirmed a role for PKC-δ in TDC and PDB actions. The observation that PDB action was completely inhibited by δV1-1 and rottlerin is intriguing, and its possible implications are explored in discussion.
To determine whether the role of PKC-δ seen in isolated colonocytes can be observed in intact epithelia, adult colonic mucosa stripped of underlying muscle were mounted in Ussing chambers, and the effects of TDC ± rottlerin on Isc were assessed. Shown in Fig. 4 are the effects of 100 μM TDC on tissues (n = 5–6 animals, each representing the mean of duplicates) pretreated with 10 μM rottlerin (30 min) or with DMSO (vehicle). This concentration of TDC caused a threefold increase in Isc (P < 0.05). In data not shown, this concentration yielded the maximal Isc in a dose-response study. While rottlerin alone had no effect on basal Isc (P > 0.05), it caused a significant inhibition of TDC-stimulated Isc (P < 0.05 compared with TDC alone); equally important, there was no significant difference in Isc of rottlerin-treated tissues ± TDC (P > 0.05). To assess tissue responsiveness, carbachol or forskolin were added at the end of the experiment. The Isc responses to these secretagogues in TDC- and TDC + rottlerin- treated tissues were similar. These results, in the intact epithelium, further support the notion that TDC appears to act via PKC-δ to stimulate Cl− transport.
Effects of PKC inhibitors on the weanling colon.
We next examined whether the differences in PKC signaling employed by TDC (nPKC) and PDB (cPKC and nPKC) could account for the fact that TDC, but not PDB, exhibits age-specific effects. Therefore, we compared the effects of the cPKC and nPKC inhibitor Gö-6850 with the effects of the cPKC inhibitor Gö-6976 and the PKC-δ inhibitors rottlerin and δV1-1 on PDB action in weanling colonocytes. As shown in Fig. 5, A and B, both Gö-6850 and Gö-6976 inhibited PDB action. In marked contrast, although rottlerin and δV1-1 inhibited PDB action in the adult colonocytes (Fig. 3), they failed to inhibit PDB action in the weanling colonocytes (Fig. 5C). None of the inhibitors affected basal transport. These results were provocative and suggested that the observed refractoriness to TDC in the weanling animal was perhaps due to reduced PKC-δ activity in the weanling animal.
Expression of PKC-δ transcript.
We examined the expression of PKC-δ in distal colonic mucosa of adult, weanling, and newborn rabbits. Only some of the PKC isoforms have been cloned from the rabbit. GeneBank lists only PKC-α, -β, and -ζ, and there is an accession number (M19338) listed as PKC-δ. However, this is misleading because a comparison of this sequence with that of PKC isoforms cloned from other species shows that the GeneBank M19338 is >90% identical to human PKC-γ (GeneBank accession no. BC047876) and shares low identity with human PKC-δ (GeneBank accession no. NM_006254.3) or rat PKC-δ (GeneBank accession no. BC076505.1). We designed degenerate oligonucleotide primers on the basis of the conserved regions of published PKC-δ mRNA sequences of human, mouse, and rat, and we obtained a partial (1423 bp) sequence of rabbit PKC-δ. A GeneBank search using Basic Local Alignment Search Tool software showed that the predicted partial amino acid sequence of rabbit PKC-δ has 90% identity to human PKC-δ and has 89% identity to rat PKC-δ. We registered this rabbit PKC-δ sequence in GeneBank (accession no. AY847776). It is interesting that, in contrast with the rat brain, in which expression of PKC-δ is robust, expression of PKC-δ is very low in adult rabbit brain (≈1.5% of that in adult distal colon).
As shown in Table 2, by real-time PCR, we observed a significant age-dependent expression of PKC-δ, with the adult having 1.6-fold greater transcript level than the weanling and 2.3-fold (P < 0.05; n ≥ 3) greater than the newborn. In contrast with PKC-δ, the expression of the transport proteins CFTR and NKCC1 did not vary with developmental age (Table 2). In preliminary experiments, we examined PKC-ζ, the only other nonconventional rabbit PKC that has been cloned. Expression of PKC-ζ mRNA was adult≃weanling>>newborn. Since inhibitor studies imply that atypical PKCs may not be involved, an examination of these isoforms was not further pursued in the present study. These data show that, at the level of the mRNA transcript, PKC-δ specifically shows an age-related expression.
Age-dependent expression of PKC protein.
Since PCR is a highly sensitive method of detection and demonstration of a transcript does not necessarily imply protein expression, we examined whether PKC protein expression was different in adults and weanlings. There have been few thorough analyses of PKC isoform expression in primary preparations of colon as opposed to cell lines (13), and there is one report on expression of PKC isoforms in adult rabbit ileum (36). Using an array of commercially available antibodies, we screened rabbit colonic lysates for the expression of various PKC isoforms (Fig. 6 and Table 3); rat and rabbit brain and Jurkat cells were used as positive controls. As described in materials and methods, samples were checked for uniform loading and transfer. Of the cPKCs, the rabbit colon expresses PKC-α and -β but not -γ (Fig. 6A). Furthermore, PKC-β shows a biphasic age-dependent expression with high expression in the adult and newborn (2.1 ± 0.5-fold of adult) and low expression in the weanling (0.5 ± 0.5-fold of adult). However, it is unlikely that expression of PKC-β is responsible for the age-dependent effects of TDC since TDC does not affect cPKC activity (Fig. 2B) (69). Of the nPKC isoforms, the rabbit colon expresses only PKC-ε and -δ but not PKC-θ (Fig. 6B). Expression of both PKC-ε and PKC-δ was age dependent. Like PKC-β, PKC-ε shows a biphasic expression, with low expression in the weanling and high expression in the newborn (0.5- and 2.5-fold of the adult, respectively; Table 3). PKC-ε is an unlikely candidate as a mediator of TDC action; it is not sensitive to rottlerin or to δV1-1, and its biphasic age-dependent expression does not parallel the transport effects of TDC (see discussion). In marked contrast, there is a dramatic age-dependent increase in the expression of PKC-δ, with highest expression in the adult and barely detectable expression in the weanling and newborn. Confirming our PCR findings, there is no detectable expression of PKC-δ protein in the rabbit brain, whereas, as shown by others (52), we find high expression of PKC-δ in rat brain (Fig. 6). As shown in Fig. 6C, the rabbit colon expresses atypical PKC-ι and -λ, and there was no dramatic change in expression with age.
To relate protein expression to activity, we next examined PKC-δ enzymatic activity in response to TDC and PDB. We first attempted to measure activity by immunoprecipitating the protein. However, although the antibodies were good for immunodetection by Western blot analysis, neither the mouse monoclonal nor the goat polyclonal commercial antibodies yielded consistent immunoprecipitation (data not shown). Therefore, by using a PKC-δ-specific substrate and a series of selective inhibitors, we modified the commercially available PKC activity assay kit to specifically measure PKC-δ activity (see materials and methods). PKC-δ activity was measurable in both cytosol and membrane fractions, and both TDC (1.6-fold) and PDB (2.0-fold) caused significant increase over basal in membrane-associated PKC-δ activity (Fig. 7A). There was a concomitant, statistically significant, 11% decrease in PKC-δ activity in the cytosolic fraction of PDB-treated cells, and a similar (9%), but statistically insignificant, decrease in that of TDC-treated cells. Our data thus far suggested that, in the weanling, PDB stimulation of Cl− transport was PKC-δ independent (i.e., rottlerin and δV1-1 insensitive; Fig. 5C), and this may be due to low expression of PKC-δ protein (Fig. 6B). To confirm this further, we examined PKC-δ activity in weanling colonocytes treated with PDB or TDC. As shown in Fig. 7B, neither agent caused a stimulation of PKC-δ activity. These results strongly suggest that TDC acts via PKC-δ and that the age-specific effects of TDC are, in part, due to the lack of PKC-δ in the young animal.
The PKC cascade is often linked to the action of neurohumoral modulators acting via intracellular calcium. We had previously shown that Ca2+-dependent secretagogues, including bile acids, exhibited age refractoriness in the rabbit colon and that TDC does not alter the activity of conventional PKCs (69). The present study is the first to demonstrate a role for PKC-δ in the secretagogue action of bile acids and that a lack of PKC-δ protein and activity most likely accounts for the refractoriness of the colon of young animals to bile acid-stimulated Cl− secretion.
There are multiple PKC isoforms, and the availability of isoform-specific inhibitors allows the delineation of specific signaling pathways. Another useful tool for studying the PKC signaling cascade is the phorbol esters, general activators of PKC. While there are some inherent limitations in the use of such pharmacological manipulations, collectively, they help us to understand the mechanisms underlying hormonal signaling pathways. Our earlier results (69) indicated that, in contrast with neurohumoral agents, neither PDB nor A23187 showed age refractoriness, and in contrast with TDC, PDB stimulated conventional PKC enzymatic activity. The implications of these results were that the age-dependent effects of TDC did not involve cPKCs, were different from the actions of PDB, and occurred at a proximal step in signaling. This was confirmed by the finding that, in the weanling, secretagogues such as TDC and neurotensin failed to increase colonic IP3 production or [Ca2+]i or to stimulate Cl− transport. However, these studies did not examine the roles of the nPKCs and aPKCs in TDC action. Cognizant of the potential differences in TDC and PDB signaling, we compared the effect of PKC isoform-specific inhibitors on the actions of TDC and PDB in the developing rabbit colon to elucidate the basis of age refractoriness in TDC signaling.
The pan-specific PKC inhibitors chelerythrine and Gö-6983 completely inhibited the stimulatory action of both TDC and PDB. Chelerythrine acts on the phosphate acceptor site of the catalytic domain of PKC and selectively inhibits PKC compared with PKA, CaMK, and tyrosine kinases (TK) (32). Gö-6983 also acts on the catalytic site to inhibit several PKC isozymes but does not inhibit PKC-μ (30). Therefore, TDC-stimulated Cl− transport involved PKCs but perhaps not PKC-μ. Gö-6850 is a potent and selective inhibitor of cPKCs and nPKCs, but not of PKA and TKs (68). The IC50 of Gö-6850 for aPKCs (>5.8 μM) was much higher than those for cPKCs or nPKCs (2–210 nM) (42); therefore, at the concentration used in this study, 5 μM, Gö-6850 specifically inhibited cPKCs and nPKCs but not aPKCs. In contrast, Gö-6976, a compound that specifically inhibits cPKCs and PKC-μ with a similar potency (30), blocked PDB but not TDC action. Collectively, these data imply that TDC action did not depend on aPKCs, PKC-μ, or cPKCs and is consistent with our previous report that only PDB, but not TDC, activated cPKC enzymatic activity (69).
The studies with Gö-6850 and Gö-6976 imply that TDC action may involve only nPKCs. Of the nPKCs, rottlerin inhibits PKC-δ (31, 62) (IC50 = 3–6 μM) and PKC-θ (70) with greater selectivity. Rottlerin has significantly reduced inhibitory activity on PKC-ε (IC50 = 80–100 μM). The latter concentrations are 8 times the concentrations used in the present study, and because PKC-θ was not expressed in the rabbit distal colon (Fig. 6B), it is likely that any action of rottlerin reflects involvement of PKC-δ. While some reports have disputed the selectivity of rottlerin in inhibiting PKC-δ (16), our results cumulatively show that the effects of rottlerin are specific for PKC-δ in the rabbit colon. The most compelling evidence is that the effects of rottlerin and those of the specific peptide inhibitor δV1-1 are similar. In contrast, a scrambled sequence control peptide had no effect on TDC-stimulated transport. In addition, rottlerin shows age specificity affecting only the adult and not weanling colonocytes. If it had nonspecific effects, this would be unlikely. Thus, in other systems, rottlerin has been reported to uncouple mitochondrial oxidative phosphorylation and to inhibit the ubiquitous elongation factor 2-specific CaMK III (31). If rottlerin were acting by either of these mechanisms in the rabbit colon, then it should have inhibited Cl− transport in both adult and weanling, and this is not the case (Figs. 3 and 5C).
Rottlerin and the specific peptide inhibitor δV1-1 inhibited TDC-stimulated Cl− transport in isolated adult colonocytes (Fig. 3). Equally important, in intact colonic epithelia, rottlerin inhibited TDC-stimulated, but not basal Isc, a measure of Cl− secretion (Fig. 4). It is intriguing that rottlerin and δV1-1 completely inhibited PDB-stimulated transport in the adult colonocytes. If PDB activated different PKC isoforms, which then converged on a common target (e.g., a transporter), then one might have predicted a partial inhibition. However, these data imply that multiple PKC isoforms may act at different steps of the signaling cascade resulting in Cl− secretion (see below).
In contrast with the adult, neither rottlerin nor δV1-1 had an effect on PDB-stimulated Cl− transport in weanling colonocytes. More important, the absence of function attributable to PKC-δ in the weanling colon was associated with reduced PKC-δ transcript (Table 2), low PKC-δ protein expression (Fig. 6), and undetectable PKC-δ activity (Fig. 7). These activity assays were conducted under conditions in which all second messenger-regulated kinases, including cPKCs, PKA, and CaMK, were inhibited. Furthermore, a PKC-δ-specific peptide was used as the substrate. We are intrigued by the marked differences in PKC-δ protein expression in the different age groups and by the modest twofold difference in PKC-δ mRNA expression. While a more detailed comparison of the transcript vs. proteins of the various isoforms is warranted, at this juncture, we can only speculate that there is perhaps greater PKC-δ protein stability in adult colonocytes than in weanling colonocytes. Taken together, the inhibitor (rottlerin and δV1-1) studies, the age dependence, and the activity measurements show that PKC-δ is critical for stimulation of Cl− transport by TDC in the rabbit distal colon.
What then are the targets for PDB and TDC in the signal transduction cascade leading to Cl− secretion? The simplest canonical sequence is as follows: Ca2+ dependent secretagogues→receptor→PLC activation→IP3 and diacylglycerol production→intracellular calcium release and PKC activation→activation of a combination of transporters, including Cl− channels, NKCC cotransporter, K+ channels, and Na+-K+-ATPase. This scenario is compounded by the ability of key signaling molecules to affect multiple steps in the cascade (cross talk), which results in a complex signaling network. It is clear from the age-dependence studies that, while PDB and TDC may both activate PKC-δ, PDB affects other PKCs and, therefore, perhaps other targets. Furthermore, PKC-δ may be affecting more than one step. The age dependency of PKC-δ expression and activity, together with our earlier data (69) that TDC cannot increase IP3 and [Ca2+]i in the weanling, suggests that PKC-δ may be affecting at least one early step in signal transduction. One possible site is the activation of PLC-γ rather than its expression since our preliminary data indicate that PLC protein expression is not age dependent (R. Prasad et al., unpublished observation). Such a role for PKC was previously reported in the regulation of bradykinin-stimulated phosphoinositide breakdown in astrocytes (10).
On the basis of the lack of age specificity of PDB action, we speculate that it is unlikely that PDB activates these initial steps. In addition, the evidence for PDB to stimulate [Ca2+]i in colonocytes is spotty at best. Thus, in T-84 cells, it does not increase [Ca2+]i but attenuates the ability of agents such as carbachol to increase it (54). However, in oligodendrocytes, PDB increases [Ca2+]i (73). Our evidence that rottlerin and the δV1-1 peptide completely rather than partially inhibit PDB action leads us to the intriguing speculation that PKC-δ may be acting at another step in the cascade and this step supersedes the activation of the cPKCs by PDB. Albeit highly speculative, we provide one example of many cross-talk signaling mechanisms we envisaged. It is conceivable that, associated with the increased expression of PKC-δ during development from weanling to adult, two steps are introduced: first, PKC-δ plays a role in an early step in signal transduction (as discussed above) and second, PKC-δ may be involved in increasing transporter recruitment to the membrane. In contrast, cPKCs cause direct activation of the transporter in the weanling and adult, but in the latter, owing to the emerging role of PKC-δ, the action of cPKC could be superseded by that of PKC-δ. Thus, if there are insufficient transporters in the membrane, activation will be attenuated. This might explain our data wherein PKC-δ inhibition results in a complete attenuation of TDC and PDB stimulated Cl− transport in the adult.
What then might be the physiological role of the TDC-PKC-δ refractoriness? In contrast with PKC-δ, there was no age dependency in the expression of the Cl− transporters NKCC1 and CFTR in the rabbit distal colon. This supports the observation that the colon of young animals is able to transport Cl− in response to some secretagogues, including cAMP and PDB. Many neurohumoral agents use Ca2+ signaling, and the lack of a fully operative cascade may be a protection against excess loss of fluid, with baseline fluid secretion occurring in response to cAMP.
Our studies also have to be examined in the wider context of what is known about the functions of PKC-δ and specifically the role of PKCs in Cl− secretion. Although roles ranging from a proapoptotic signal (39), to a tumor suppressor and to a regulator in stress, immune, and inflammatory responses have been ascribed to PKC-δ (29, 63), the PKC-δ knockout mice generally develop and grow normally. The knockout mice show a clear phenotype only in certain cell functions (63); examples include reduced migration of neutrophil to the injured site and loss of preconditioning-induced cardioprotection (63). Intestinal ion transport has not been examined in these animals. While a role for PKCs has been implicated in a variety of electrolyte transport processes, specific isoforms have been identified only in some systems, and it is clear that there are considerable cell-type and species differences. For example, PKC-δ mediates the stimulation of Na+/H+ exchange by phorbol esters in C6 glioma cells (11), and PKC-α mediates the inhibition of Na+/H+ exchange by serotonin in Caco-2 cells (28). In contrast, inhibition of Cl−/OH− exchange in Caco-2 cells by serotonin involves PKC-δ (59), and inhibition by phorbol esters involves PKC-ε (58). Activation of PKC-δ, but not PKC-ε or PKC-α, decreases IEC-18 epithelial integrity (66). In heterologous expression systems, the transporters involved in Cl− secretion are differentially regulated by PKC. For example, in baby hamster kidney cells, direct phosphorylation by PKC of CFTR, presumably on its PKC consensus site at S686, results in partial activation of the channel (9). However, phorbol esters inhibit NKCC1 activity in Xenopus oocytes overexpressing NKCC1 (25, 49). In human embryonic kidney 293 cells with stably expressed Kir3.1/3.2 channels, PKC-δ-mediated channel inhibition was caused by the activation of muscarinic receptors (7). Thus PKC isoform-specific modulation is dependent on cell type.
The most compelling evidence for specific roles for PKC isoforms in Cl− secretion comes from studies in airway epithelia by Liedtke and colleagues (40, 41). Thus, while PKC-ε is involved in CFTR activation, a complex of kinases, including PKC-δ, and protein phosphatases (PP2A) plays a role in the adrenergic activation of NKCC1 and thereby Cl− secretion. Recent evidence suggests that PKC-δ directly associates with actin (61) and that actin serves as a scaffold for PKCs, PP2A, and STE-20 related proline alanine-rich kinase to optimally modulate NKCC activity (41). With respect to colonic Cl− secretion, phorbol esters and, therefore, PKCs, stimulate it in rabbit (50, 57) and rat distal colon (23), in the human colonic cell line HT-29cl.19A (2), and in primary human colonocytes (57). In some studies, the effects of phorbol esters on epithelial cells have been shown to occur by activation of secretagogues, such as prostaglandins from the lamina propria (8). However, we showed that the action of TDC in isolated colonocytes (69) is not dependent on prostaglandin release. In contrast, PKC activation by phorbol ester is inhibitory in some cell types (5) (see introduction); For example, it inhibits Ca2+-dependent Cl− secretion and NKCC activity in T-84 cells, which appear to involve PKC-ε but not PKC-δ (5, 13, 19). The phorbol ester-induced internalization of NKCC1 in T-84 cells occurs via PKC-ε (19). Our present study on the developmental expression of PKC isoforms provides some insights into isoform-specific regulation. Only PKC-δ, -β, and -ε showed age-specific protein expression (Fig. 6 and Table 3), with reduced expression in weanlings compared with the adult. While we have shown that PKC-δ (Fig. 5), but not PKC-β (Fig. 2B), plays a role in the age-specific secretagogue action of TDC, the role, if any, of PKC-ε in TDC remains to be explored. On the basis of the T-84 studies, it is tempting to make a couple of speculations. First, in the adult, it is conceivable that PKC-ε may serve to temper the overall response to PDB or TDC by inhibiting NKCC. Second is the suggestion that the high expression of PKC-ε (and thereby reduced NKCC) in the newborn may account for the lack of TDC responsiveness in the newborn. While an attractive suggestion, this does not account for the weanling data (low PKC-ε and no TDC effect) and the fact that both PDB and cAMP stimulate Cl− transport in the newborn (20). With the caveat that there are tissue/cell line-specific differences, short-term PKC-δ activation appears to stimulate colonic Cl− transport.
Age-dependent expression of PKC-δ has been reported in other tissues, but to our knowledge, ours is the first study to link developmental expression of PKC-δ to function in a physiological model. For example, in developing rat brain, PKC-δ is involved in ventilatory regulation, and its expression of PKC-δ, -μ, and -β, but not -ι/λ, is low in 2-day-old and 10-day-old pups and is high in adult cortex (4). However, PKC-δ expression was greater in the fetal and neonatal rat and declined 2 wk postnatally (56). In the rat colon, PKC-δ mRNA expression was lower in the colonic mucosa of 10-day-old rats than in 25-day-old rats, and hypoxia increased PKC-δ translocation to the membrane (15). In none of these studies were the mechanisms regulating PKC-δ expression and signaling during development examined. Potential candidates include glucocorticoids and thyroxine that rise during weaning, which have been implicated in the developmental regulation of proteins involved in ileal bile acid transport (34, 35, 43). Analysis of the promoter region of PKC-δ has revealed several potential regulatory elements, including NF-κB, GATA, p53, and MyoD, but not any consensus sequence motifs for FXR, the putative bile acid receptor (65). However, the PKC-δ promoter contains a related motif, the androgen response element, and PKC-δ transcription increases in an androgen-dependent manner in prostate cells (26). It is conceivable that a combination of other hormones, diet, and intestinal flora could dictate PKC-δ expression.
In summary, we provide one of the first demonstrations wherein age-dependent expression of a signaling molecule like PKC-δ is correlated with functional changes. Our data strongly imply that the expression of PKC-δ protein may be a critical step in the age-related refractoriness to bile acid signaling in the rabbit colon. This signaling cascade may be used by several neurohumoral agents, such as histamine, serotonin, and neurotensin, and such Ca2+-dependent secretagogues are responsible for the minute-by-minute regulation needed in intestinal fluid homeostasis. In the young animal, in which excess fluid loss can have grave consequences, this refractoriness to Ca2+-dependent secretagogues may represent an overall protective mechanism.
This study was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-58135 (to M. C. Rao) and the Royal Golden Jubilee Program, Thailand (to P. Piyachaturawat and J. Kanchanapoo).
The authors are deeply indebted to Karen Ridge, Division of Pulmonary and Critical Care, the Feinberg School of Medicine, Northwestern University, for generous gifts of PKC-δ-specific peptide inhibitor (δV1-1) and nonspecific (scrambled sequence) peptide.
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. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society