Liver Biology and Pathobiology
Protein kinase Cδ mediates cyclic adenosine monophosphate–stimulated translocation of sodium taurocholate cotransporting polypeptide and multidrug resistant associated protein 2 in rat hepatocytes†
Article first published online: 19 DEC 2007
Copyright © 2008 American Association for the Study of Liver Diseases
Volume 47, Issue 4, pages 1309–1316, April 2008
How to Cite
Schonhoff, C. M., Gillin, H., Webster, C. R. L. and Anwer, M. S. (2008), Protein kinase Cδ mediates cyclic adenosine monophosphate–stimulated translocation of sodium taurocholate cotransporting polypeptide and multidrug resistant associated protein 2 in rat hepatocytes. Hepatology, 47: 1309–1316. doi: 10.1002/hep.22162
Potential conflict of interest: Nothing to report.
- Issue published online: 25 MAR 2008
- Article first published online: 19 DEC 2007
- Accepted manuscript online: 19 DEC 2007 12:00AM EST
- Manuscript Accepted: 28 NOV 2007
- Manuscript Received: 10 AUG 2007
- National Institutes of Health. Grant Numbers: DK-33436, DK-65975
Cyclic adenosine monophosphate (cAMP) stimulates translocation of Na+-taurocholate (TC) cotransporting polypeptide (Ntcp) and multidrug resistant associated protein 2 (Mrp2) to the plasma membrane. Because cAMP activates phosphoinositide-3-kinase (PI3K) and protein kinase C (PKC) activation is PI3K-dependent, the aim of the current study was to determine whether cAMP activates conventional and novel PKCs in hepatocytes and whether such activation plays a role in cAMP-stimulated Ntcp and Mrp2 translocation. The effect of cAMP on PKCs, TC uptake, and Ntcp and Mrp2 translocation was studied in isolated rat hepatocytes using a cell-permeable cAMP analog, CPT-cAMP. The activity of PKCs was assessed from membrane translocation of individual PKCs, and phospho-specific antibodies were used to determine PKCδ phosphorylation. TC uptake was determined from time-dependent uptake of 14C-TC, and a cell surface biotinylation method was used to determine Ntcp and Mrp2 translocation. CPT-cAMP stimulated nPKCδ but not cPKCα or nPKCϵ, and induced PI3K-dependent phosphorylation of nPKCδ at Thr505. Rottlerin, an inhibitor of nPKCδ, inhibited cAMP-induced nPKCδ translocation, TC uptake, and Ntcp and Mrp2 translocation. Bistratene A, an activator of nPKCδ, stimulated nPKCδ translocation, TC uptake, and Ntcp and Mrp2 translocation. The effects of cAMP and bistratene A on TC uptake and Ntcp and Mrp2 translocation were not additive. Conclusion: These results suggest that cAMP stimulates Ntcp and Mrp2 translocation, at least in part, by activating nPKCδ via PI3K-dependent phosphorylation at Thr505. (HEPATOLOGY 2008.)
Protein kinase Cs (PKCs) comprise a family of at least 12 isozymes.1 These include conventional (cPKCα, cPKC βI, cPKC βII, and cPKCγ), novel (nPKCδ, nPKCϵ, nPKCη, and nPKCθ), atypical (aPKCζ and aPKCλ) isoforms, and PKCμ. These isoforms differ in their dependency on Ca2+ and phospholipids, such that cPKCs are dependent on Ca2+ and diacylglycerol, nPKCs are Ca2+-independent, and aPKCs are independent of both Ca2+ and diacylglycerol. PKCs present in rat hepatocytes include cPKCα, nPKCδ, nPKCϵ, aPKCζ, and probably cPKCβII.2–5 Activation of most PKCs, if not all, is phosphoinositide-3-kinase (PI3K) dependent.1
Membrane transporter translocation and activity appear to be PKC-isoform specific.6 Tauroursodeoxycholate-induced activation of cPKCα may be involved in Mrp2 translocation to the canalicular plasma membrane.7 Taurocholate (TC) activates nPKCδ4 and translocates Bsep to the canalicular membrane.8, 9 cPKCα, but not nPKCϵ phosphorylate Bsep; this phosphorylation appears to increase transport activity without affecting translocation.10 Studies in adipocytes and skeletal muscles suggest a role for nPKCδ and aPKCζ in insulin-mediated glucose transporter 4 translocation.11, 12
cAMP stimulates TC uptake in hepatocytes by translocating Ntcp to the plasma membrane, an effect mediated via cAMP-induced PI3K-dependent activation of aPKCδ in hepatocytes.13 However, the role of other PKCs in Ntcp translocation is unknown. Cyclic AMP also stimulates translocation of multidrug resistant associated protein 2 (Mrp2) to the canalicular membrane in hepatocytes.8, 14 Because activation of most PKCs is PI3K dependent, it is likely that PKCs other than aPKCδ are also activated by cAMP. The aim of the current study was to determine whether cAMP activates PKCs other than aPKCδ in hepatocytes and whether such activation plays a role in cAMP-induced Ntcp and Mrp2 translocation. Results show that nPKCδ is activated by cAMP and is involved in cAMP-mediated Ntcp and Mrp2 translocation in hepatocytes.
Materials and Methods
TC (Na+-salt) and rottlerin were purchased from Calbiochem, San Diego, CA. 8-Chlorophenylthio cAMP (CPT-cAMP), rottlerin, bistratene A, wortmannin, aprotinin, leupeptin, okadaic acid, phorbol myristate acetate (PMA), and collagenase were obtained from Sigma Chemical Co., St. Louis, MO. [24-14 C]TC (56 mCi/mmol), and [methoxy-3H]inulin (80 Ci/mmol) were purchased from PerkinElmer (Boston, MA). Anti-fusion protein antibodies to the cloned Ntcp were prepared as previously described.15 Rat Mrp2 antibody was a gift from Dr. Keppler (University of Heidelberg, Heidelberg, Germany). Antibodies for cPKCα, nPKCδ, nPKCϵ, and phospho-PKCδ (Thr505 and Tyr 311) were obtained from Cell Signaling Technology (Danvers, MA). Male Wistar rats (200-300 g) obtained from Charles River Laboratories served as liver donors, and the protocol for harvesting livers was approved by the Institutional Animal Care and Use Committee.
Hepatocytes were isolated from rat livers using a previously described collagenase perfusion method.16 Freshly prepared hepatocytes suspended (100 mg wet weight/ml) in a 94-(2-hydroxyethyl)-1-piperazine ethanesulfonic acid] (4-(2-hydroxyethyl)-1-piperazine ethanesulfonic acid) assay buffer (pH 7.4) containing 20 mM 94-(2-hydroxyethyl)-1-piperazine ethanesulfonic acid, 140 mM sodium chloride, 5 mM potassium chloride, 1 mM magnesium sulfate, 1.0 mM calcium chloride, 0.8 mM potassium phosphate monobasic, and 5 mM glucose were incubated for 30 minutes at 37°C under air before initiating studies. Details of these experiments are given in the legends of the figures. All studies were repeated in at least 3 different cell preparations.
PKC Activity and nPKCδ Phosphorylation.
The effect of various agents on PKC activity was assessed from membrane translocation of individual PKCs as described by others.7, 11 Briefly, hepatocytes treated with CPT-cAMP or PMA were homogenized followed by centrifugation at 400g. The resultant supernatant was centrifuged at 148,000g for 1 hour. The pellet was washed 3 times at 12,000g and then dissolved in the lysis buffer. The dissolved pellet was centrifuged at 12,000g for 15 minutes, and the resulting supernatant was saved as the membrane fraction. This membrane fraction and the original homogenate were subjected to immunoblot analysis for PKCs using isoform-specific antibodies. Membrane PKC was expressed as a fraction of total PKC present in the homogenate (membrane/total). The phosphorylation of nPKC at Thr505 and Tyr311 was evaluated by immunoblot analysis of cell lysates using phosphospecific antibodies. The blots were first probed for phospho-nPKCδ followed by total nPKCδ. To correct for loading variations, the result was expressed as a ratio of phospho/total nPKCδ with the control ratio set at 1.0.
TC Uptake and Translocation of Ntcp and Mrp2.
The initial uptake rate of TC (20 μM) in hepatocytes was calculated from the slope of the linear portion of time-dependent uptake curves, and was expressed in nmol/min/mg protein as previously described.17 A cell surface protein biotinylation method as previously described by us18, 19 was used to assess Ntcp and Mrp2 translocation in hepatocytes. Briefly, after various treatments, cell surface proteins were biotinylated by exposing hepatocytes to sulfo-NHS-LC-Biotin (0.5 mg/mL, Pierce) followed by preparation of cell lysate used to determine biotinylated and total Ntcp and Mrp2 mass using immunoblot analysis. The absence of biotinylated actin was routinely checked to assure that intracellular proteins were not biotinylated.18, 19
The Lowry method was used to determine cell protein.20 The blots were scanned in gray scale using Adobe Photoshop (Adobe Systems Inc., San Jose, CA), and the relative band densities were quantitated using Sigmal Gel (Jandel Scientific Software, San Rafael, CA). All values are expressed as mean ± standard error of the mean (SEM). Analysis of variance or Student t Test was used to statistically analyze data, with P < 0.05 considered significant.
Cyclic AMP Activates nPKCδ.
To determine whether cAMP activates conventional and novel PKC, the concentration-dependent effect of CPT-cAMP on PKC (α, δ, and ϵ isoforms) translocation was determined. The fraction of total PKCs in membrane under control conditions was 55% ± 13.5%, 30% ± 3%, and 28% ± 7% for cPKCα, nPKCδ, and nPKCϵ, respectively. PMA, used as a positive control, increased membrane translocation of all 3 PKC isoforms (Fig. 1). These values are comparable to those reported earlier for rat hepatocytes.5 In contrast to PMA, CPT-cAMP increased membrane PKCδ by 42%, 65%, and 97% at 10, 50, and 100 μM, respectively, without affecting membrane cPKCα or nPKCϵ (Fig. 1). Total PKCs in homogenates were not affected by CPT-cAMP or PMA. Thus, cAMP activates nPKCδ, but not cPKCα and nPKCϵ.
PI3K-Dependent Phosphorylation of nPKCδ at Thr505 by cAMP.
Activation of nPKCδ involves phosphorylation at various sites,21, 22 including phosphorylation at Thr50523 and Tyr311.22 Thus, whether phosphorylation at these sites is involved in cAMP-mediated activation of nPKCδ was studied. PMA, used as a positive control, increased phosphorylation at both sites (Figs. 2, 3), as it has been reported in cardiomyocytes.24, 25 CPT-cAMP increased phosphorylation at Thr505 by 1.7-fold to 2.0-fold in a time-dependent manner, with significant increases observed as early as 5 minutes (Fig. 2A,C). To determine whether the cAMP-induced phosphorylation at Thr505 was PI3K dependent, the effect of wortmannin, a specific PI3K inhibitor, was studied. Wortmannin completely inhibited CPT-cAMP–induced phosphorylation (Fig. 2B,C), indicating that the effect of cAMP is PI3K-dependent. The effect of PMA on Thr505 phosphorylation was not affected by wortmannin (Fig. 2B,C). PMA did not activate PKB and wortmannin-inhibited PI3K-dependent activation of PKB by CPT-cAMP (data not shown), as previously reported.18 Thus, the effect of PMA is not mediated via the PI3K pathway.
In contrast to Thr505 phosphorylation, phosphorylation at Tyr311 was inconsistently observed in control cells (Fig. 3), and only faint unquantifiable bands can be detected in some blots (Fig. 3B). Similar findings are reported for unstimulated glioma cells26 and cardiomyocytes.24 Similarly to control cells, either no band (Fig. 3A) or only faint bands (Fig. 3B) could be detected in CPT-cAMP–stimulated cells (and only if observed in control cells) for up to 60 minutes. Although the PMA effect on Tyr311 phosphorylation was readily detectable, no such difference was apparent between control and CPT-cAMP–treated cells. Thus, it is highly unlikely that CPT-cAMP induces Tyr311 phosphorylation.
Rottlerin Inhibits cAMP-Induced TC Uptake, and Ntcp and Mrp2 Translocation.
Activation of nPKCδ by cAMP raises the possibility that cAMP-induced translocation of Ntcp and Mrp2, similar to insulin-mediated glucose transporter 4 translocation,11 may be mediated via nPKCδ. We thus studied the effect of rottlerin, a specific inhibitor of nPKCδ,27 on TC uptake and translocation of Ntcp and Mrp2. Rottlerin, at 5 and 10 μM, inhibited CPT-cAMP–induced increases in membrane nPKC by 54% and 77%, respectively (Fig. 4A), indicating inhibition of cAMP-induced nPKCδ activation. Rottlerin did not affect cPKCα or nPKCϵ translocation (data not shown). The same concentrations of rottlerin did not affect basal TC uptake but did decrease cAMP-induced increases in TC uptake by 66% and 88%, respectively (Fig. 4B). Rottlerin, at 5 and 10 μM, also inhibited CPT-cAMP–induced increases in plasma membrane Ntcp by 41% and 87%, respectively, without affecting basal values (Fig. 4C). CPT-cAMP, as expected based on previous studies,8, 14 increased plasma membrane Mrp2 by 90%, and this effect was inhibited by 52% and 85% by 5 and 10 μM rottlerin, respectively (Fig. 4D). These results suggest that nPKCδ may be involved in cAMP-induced Ntcp and Mrp2 translocation in hepatocytes.
Bistratene A Stimulates cAMP-Induced TC Uptake, and Ntcp and Mrp2 Translocation.
If the effect of cAMP is mediated via nPKCδ, then an activator of nPKCδ other than cAMP should also stimulate Ntcp and Mrp2 translocation. This hypothesis was tested by using bistratene A, a specific activator of nPKCδ.28, 29 Bistratene A stimulated nPKCδ translocation by 16%, 35%, and 50% at 50, 100, and 200 nM (Fig. 5A) and did not affect translocation of cPKCα and nPKCϵ (data not shown). Bistratene A has been shown to induce translocation of nPKCδ without affecting other isoforms in HL60 cells and melanoma cell lines.28, 29 Bistratene A stimulated TC uptake by 25%, 53%, and 64% at 50, 100, and 200 nM (Fig. 5B), Ntcp translocation by 15%, 31%, and 70% (Fig. 5C), and Mrp2 translocation by 20%, 108%, and 192% (Fig. 5D) at 50, 100, and 200 nM, respectively. Significant stimulation of nPKCδ, TC uptake, and the translocation of Ntcp and Mrp2 was observed at 100 and 200 nM bistratene A. CPT-cAMP stimulated TC uptake and translocation of Ntcp and Mrp2 by 108%, 82%, and 105%, respectively, in these experiments, and bistratene A did not further increase these effects of CPT-cAMP (Fig. 5B-D). These results indicate that the effects of cAMP and bistratene A on Ntcp and Mrp2 translocation are not additive.
The aim of the current study was to determine whether cAMP activates conventional and novel PKCs in hepatocytes and whether such activation plays a role in cAMP-induced Ntcp and Mrp2 translocation. Our studies showed that cAMP activates nPKCδ, but not cPKCα and nPKCϵ (Fig. 1); cAMP activates nPKCδ in human leukemia HL60 cells.30 Although TC, but not tauroursodeoxycholate and TLC, has been shown to activate nPKCδ,4, 5, 31 this is the first demonstration of cAMP activating nPKCδ in hepatocytes. Our studies also suggest that cAMP-induced activation of nPKCδ may involve PI3K-dependent phosphorylation of PKCδ-Thr505, and the effect of cAMP on Ntcp and Mrp2 translocation may be mediated via nPKCδ, as discussed further.
The mechanism by which cAMP activates nPKCδ in hepatocytes is unknown. Recent studies suggest that the activation of PKCs are PI3K dependent.1 The activation of PKCϵ by TLC in hepatocytes is PI3K-dependent.32 Although the activation of nPKCδ involves phosphorylation at different sites,33 the PI3K-dependent activation of nPKCδ by serum in HEK293 cells23 and by vascular endothelial growth factor in human umbilical vein endothelial cells34 involves phosphorylation of Thr505 in the activation loop. In the current study, CPT-cAMP induced PI3K-dependent phosphorylation of nPKCδ-Thr505 in hepatocytes (Fig. 2). Thus, it is likely that cAMP-induced activation of nPKCδ in hepatocytes involves PDK1 mediated phosphorylation at Thr505. Because nPKCδ activation involves phosphorylation at other sites as well,35 the possibility that cAMP activates nPKCδ by phosphorylating other sites cannot be ruled out.
In the current study, the role of nPKCδ was studied using an inhibitor (rottlerin) and an activator (bistratene A) of nPKCδ. Rottlerin, a selective inhibitor of nPKCδ,27 has been shown to inhibit insulin-induced nPKCδ activation in myocytes11 and has been used by other investigators to study the role of nPKCδ.36–39 In the current study, rottlerin inhibited cAMP-induced nPKCδ translocation without affecting translocation of cPKCα or nPKCϵ. We also studied the effect of bistratene A, a specific activator of nPKCδ. In the current study, bistratene A induced translocation of nPKCδ without affecting the translocation of cPKCα and nPKCϵ in hepatocytes, as has been reported in other cells.28, 29, 40 In addition, activation of nPKCδ by bistratene A was associated with increased Ntcp and Mrp2 translocation. Furthermore, the effects of cAMP and bistratene A on Ntcp and Mrp2 translocation are nonadditive; a nonadditive effect is consistent with a common mediator. Taken together, these results would suggest that the effects of rottlerin and bistratene A are due to their inhibitory and stimulatory effects, respectively, on nPKCδ in hepatocytes. Thus, it is highly likely that cAMP stimulates Ntcp and Mrp2 translocation, at least in part, by activating nPKCδ. Further studies using a molecular approach to activate and inhibit nPKCδ would provide stronger evidence in support of this conclusion. Unfortunately, Ntcp is rapidly down-regulated in cultured primary hepatocytes,41 and this limits the use of a molecular approach, which takes 48 hours or more for adequate expression. In addition, further studies are needed to confirm that nPKCδ targets Ntcp and Mrp2 to sinusoidal and canalicular membranes, respectively.
Although our studies show that cAMP-induced Ntcp and Mrp2 translocation in hepatocytes may be nPKCδ dependent, nPKCδ has been shown to be involved in various cellular functions, including transport functions.42 For example, the transporter regulatory effects of nPKCδ include stimulation of Na+-H+ exchanger in glial cells,43 α1-adrenergic activation of Na+-K+-2Cl− cotransport in tracheal epithelial cells,44 insulin-mediated glucose transporter 4 translocation in myocytes,11 and serotonin-mediated inhibition of Cl−/OH− exchange in Caco-2 cells.39 The mechanism by which nPKCδ regulates these transporters is unclear. In previous studies, we reported that cAMP-induced Ntcp translocation is aPKCζ dependent13 and aPKCζ is required for microtubule-based motility of Ntcp-containing vesicles isolated from rat hepatocytes.45 Phosphorylation of the γ chain of the high-affinity receptor for immunoglobulin E by nPKCδ correlates with endocytosis of the receptor,46 and nPKCδ promotes motility of Madin-Darby canine kidney cells.47 Thus, it can be speculated that nPKCδ, like aPKCζ, may stimulate vesicle motility and thereby stimulate transporter translocation.
PKCs have been implicated in cholestatic and anticholestatic effects of bile acids.6, 48 Thus, TLC-induced cholestasis (Mrp2 retrieval) may be mediated via nPKCϵ,32 and the reversal of TLC-induced Mrp2 retrieval by tauroursodeoxycholate may involve cPKCα.7 Cyclic AMP has been shown to reverse TLC-induced retrieval of Bsep49 and estradiol glucuronide-induced Mrp2 retrieval.50 Because cAMP does not stimulate cPKCα (Fig. 1), the anticholestatic effect of cAMP is unlikely to be mediated via cPKCα. Based on the results of the current study, it is possible that nPKCδ is involved in the anticholestatic effect of cAMP.
Recent studies in hepatocytes suggest that nPKCδ may be involved in allyl alcohol-induced hepatotoxicity37 and 4-hydroxynonenal-induced apoptosis51; rottlerin was used to define the role of nPKCδ in both studies. The apoptotic and toxic effects of nPKCδ are difficult to reconcile with the activation of nPKCδ by cAMP and the known anti-apoptotic effect of cAMP in hepatocytes.52, 53 The current study showed that CPT-cAMP did not stimulate nPKCδ-Tyr311 phosphorylation (Fig. 3), and this result may explain, at least in part, why cAMP-induced nPKCδ activation may not lead to apoptosis. nPKCδ has been reported to positively and negatively regulate apoptosis, depending on sites phosphorylated by various stimuli.33, 35 It has been suggested that the cleavage of activated nPKCδ to a catalytic fragment stimulates apoptosis.33 The cleavage of nPKCδ appears to be dependent on phosphorylation at Tyr311,22 and the phosphorylation at this site by c-Abl, a nonreceptor tyrosine kinase, promotes the apoptotic effect of nPKCδ in glioma cells.26 If Tyr311 phosphorylation is also involved in toxic effects of nPKCδ, then cAMP would not be expected to produce such effects, because activation of nPKCδ by cAMP does not appear to involve Tyr311 phosphorylation in hepatocytes.
In summary, the current study in hepatocytes showed that cAMP stimulates Ntcp and Mrp2 translocation, at least in part, by activating nPKCδ via PI3K-dependent phosphorylation at Thr505.
We thank Holly Jameson and Ariel Hobson for excellent technical assistance.
- 7Tauroursodeoxycholic acid inserts the apical conjugate export pump, Mrp2, into canalicular membranes and stimulates organic anion secretion by protein kinase C-dependent mechanisms in cholestatic rat liver. HEPATOLOGY 2001; 33: 1206–1216., , , , , , et al.