The bile acid taurochenodeoxycholate activates a phosphatidylinositol 3-kinase-dependent survival signaling cascade.

Liver injury during cholestasis reflects a balance between the effects of toxic and nontoxic bile acids. However, the critical distinction between a toxic and nontoxic bile acid remains subtle and unclear. For example, the glycine conjugate of chenodeoxycholate (GCDC) induces hepatocyte apoptosis, whereas the taurine conjugate (TCDC) does not. We hypothesized that the dissimilar cellular responses may reflect differential activation of a phosphatidylinositol 3-kinase (PI3K)-dependent signaling pathway. In the bile acid-transporting McNtcp.24 rat hepatoma cell line, TCDC, but not GCDC, stimulated PI3K activity. Consistent with this observation, inhibition of PI3K rendered TCDC cytotoxic, and constitutive activation of PI3K rendered GCDC nontoxic. Both Akt and the atypical protein kinase C isoform zeta (PKCzeta) have been implicated in PI3K-dependent survival signaling. However, TCDC activated PKCzeta, but not Akt. Moreover, inhibition of PKCzeta converted TCDC into a cytotoxic agent, whereas overexpression of wild-type PKCzeta blocked GCDC-induced apoptosis. We also demonstrate that TCDC activated nuclear factor kappaB (NF-kappaB) in a PI3K- and PKCzeta-dependent manner. Moreover, inhibition of NF-kappaB by an IkappaB super-repressor rendered TCDC cytotoxic, suggesting that NF-kappaB is also necessary to prevent the cytotoxic effects of TCDC. Collectively, these data suggest that some hydrophobic bile acids such as TCDC activate PI3K-dependent survival pathways, which prevent their otherwise inherent toxicity.

Bile acids are hydrophobic, potentially cytotoxic compounds synthesized from cholesterol in the liver and secreted into the bile canaliculus, where they promote bile flow. In man, bile acids are conjugated to glycine or taurine, with the glycine conjugates predominating (1). The conjugation decreases their hydrophobicity and renders the molecules less cytotoxic at physiologic concentrations. However, hepatic accumulation of bile acids is a salient pathophysiologic consequence of cholestasis (a syndrome of bile flow impairment) due to the failure to secrete these compounds into the bile canaliculus (2). Elevated concentrations of bile acids within the liver promote liver injury and the development of liver cirrhosis and liver failure. For example, children lacking the canalicular transport protein for bile acid secretion develop a progressive liver disease due to the inability to excrete bile acids from the hepatocyte (3). Numerous studies have now shown that bile acid concentrations that occur during cholestasis induce hepatocyte apoptosis, thus providing a cellular mechanism for bile acid-mediated liver injury (4). Not all bile acids are toxic, however, and minor changes in bile acid structure dramatically alter their potential cytotoxicity. For example, the glycine conjugate of chenodeoxycholate induces hepatocyte apoptosis in vitro, whereas the taurine conjugate does not (5). Previous concepts suggested that bile acid toxicity correlated with relative hydrophobicity, with hydrophobic bile acids being cytotoxic and hydrophilic bile acids being nontoxic. However, we could not establish a relationship between bile acid-induced apoptosis and relative hydrophobicity (5).
Recent studies demonstrated that bile acids activate cytoplasmic protein kinase cascades and function as ligands for the nuclear receptor farnesoid X receptor (6), suggesting that they may mediate their effects by altering cell signaling pathways. Indeed, the nontoxic bile acid taurocholate has been found to activate phosphatidylinositol 3-kinase (PI3K) 1 (7), a potent activator of survival signals (8), raising the possibility that nontoxic, yet hydrophobic bile acids do not trigger apoptosis because they activate a PI3K-dependent survival signaling pathway.
Downstream effectors of PI3K-dependent survival signals include Akt and the atypical protein kinase C (PKC) isoforms, especially PKC (9,10). Akt, the cellular homolog of the viral oncoprotein v-Akt, suppresses apoptotic cell death in a number of cell types (11)(12)(13). One substrate for Akt is BAD, a proapoptotic member of the Bcl-2 family. Akt phosphorylates BAD, thereby blocking it from binding and inactivating Bcl-2 and Bcl-x L , two anti-apoptotic Bcl-2 family members (14). Another substrate for Akt is the transcription factor NF-B, a potent regulator of a number of anti-apoptotic gene products (15)(16)(17). Akt activates NF-B by phosphorylating IB kinase-␣ (18). Active IB kinase-␣ then phosphorylates IB, resulting in dissociation from NF-B, allowing this transcription factor to enter the nucleus. PKC is another downstream effector of PI3K (19) and can also mediate activation of NF-B (20,21). Activation of NF-B by PKC also involves IB phosphorylation (22).
In this study, we demonstrate that the nontoxic bile acid taurochenodeoxycholate (TCDC) activates PI3K, whereas the toxic glycine conjugate (GCDC) does not activate this lipid kinase. Moreover, we show that genetic or pharmacologic inhibition of PI3K converts TCDC to a cytotoxic bile acid and that constitutive activation of PI3K blocks GCDC-induced apoptosis. PI3K mediated its anti-apoptotic effects by activating PKC and NF-B. Correspondingly, inhibition of PKC and NF-B converted TCDC to a potent inducer of apoptosis. Thus, the present data suggest that TCDC is not cytotoxic because it activates PI3K-dependent cell survival signaling pathways.

Kinase Assays
PI3K Assay-Cells were grown to 50 -75% confluence in 60-mm dishes and made quiescent by culture in serum-free DMEM for 8 -12 h. PI3K activity was measured using modifications of a published technique (24). After serum deprivation, cells were treated with diluent, 50 M GCDC, or 50 M TCDC for 15 min. Cells were lysed in ice-cold lysis buffer containing 1 mM phenylmethylsulfonyl fluoride and 1% Nonidet P-40 and rocked for 20 min on ice. Clarified protein (1.5 mg/sample) was incubated with 5 l of anti-PI3K p85 antibody (Upstate Biotechnology, Inc., Lake Placid, NY) overnight at 4°C. Immune complexes were precipitated by incubation with 60 l of protein A/G-agarose (Santa Cruz Biotechnology, Santa Cruz, CA) for 2 h. The immunoprecipitates were washed three times with buffer containing 137 mM NaCl, 20 mM Tris-HCl, 1 mM CaCl 2 , 1 mM MgCl 2 , 0.1 mM Na 3 VO 4 , and 1% Nonidet P-40 (pH 7.4); three times with buffer containing 0.1 M Tris-HCl, 5 mM LiCl, and 0.1 mM Na 3 VO 4 ; and twice with buffer containing 150 mM NaCl, 10 mM Tris-HCl, 5 mM EDTA, and 0.1 mM Na 3 VO 4 . Assays were then performed in a reaction mixture containing 0.88 mM ATP, 100 mM MgCl 2 , 30 Ci of [␥-32 P]ATP, and 20 g of phosphatidylinositol (Sigma) and incubated with agitation for 15 min at 37°C. The reactions were stopped with 20 l of 6 M HCl. The organic layer was extracted with 160 l of CH 3 OH/MeOH (1:1) and separated on a silica gel thin-layer chromatography plate (J. T. Baker Inc.). Thin-layer chromatography plates were developed in CHCl 3 /CH 3 OH/H 2 O/NH 4 OH (60:47:11.3:2) and dried. Radiolabeled phosphatidylinositol phosphates were visualized by autoradiography on X-Omat film (Eastman Kodak Co.).
PKC Assay-PKC activity was measured using modifications of a published technique (25). Briefly, cells were serum-deprived for 12 h and stimulated with diluent, 250 nM insulin, or 200 M TCDC for 90 min. Cells were then washed twice with ice-cold phosphate-buffered saline and lysed in 50 mM Tris (pH 7.5), 150 mM NaCl, 2 mM EDTA, 1 mM EGTA, 1% Triton X-100, and protease inhibitor mixture (Roche Molecular Biochemicals) by rocking for 30 min on ice. Cell lysates were centrifuged (14,000 ϫ g) for 15 min at 4°C, and 1 mg of the cytosolic protein was incubated with 10 l of anti-PKC antibodies (Santa Cruz Biotechnology) overnight at 4°C. Immune complexes were precipitated with 100 l of protein A-Sepharose (Zymed Laboratories Inc., South San Francisco, CA) overnight at 4°C and then washed seven times with lysis buffer modified to contain 500 mM NaCl. Washed immunoprecipitates were incubated with 2 g of myelin basic protein (Upstate Biotechnology, Inc.) and 10 Ci of [␥-32 P]ATP for 30 min at 37°C in 35 mM Tris (pH 7.5), 10 mM MgCl 2 , 5 mM EGTA, 1 mM CaCl 2 , and 1 mM phenyl phosphate. Proteins were separated by SDS-polyacrylamide gel electro-phoresis (10%) and transferred to nitrocellulose. Radiolabeled myelin basic protein was detected by autoradiography on BiomaxMR film (Kodak). At least three independent experiments of all PI3K and PKC assays were performed.

Quantitation of Apoptosis
Apoptosis was quantitated by assessing the characteristic nuclear changes of apoptosis using the DNA-binding dye 4,6-diamidino-2-phenylindole dihydrochloride and fluorescence microscopy (26).

Electrophoretic Mobility Shift Assay (EMSA)
Cells were stimulated with diluent (DMEM) or different concentrations of TCDC and GCDC. Nuclear protein extracts were then prepared as described by Dignam et al. (31) and used immediately or stored at Ϫ80°C. For EMSA, 6 g of nuclear proteins and 3 g of the nonspecific competitor poly(dI⅐dC) were incubated in binding buffer (100 mM HEPES, 300 mM KCl, 20% Ficoll, 0.05% Nonidet P-40, and 0.5 mg/ml bovine serum albumin) with 3.5 pmol of double-stranded DNA oligonucleotide containing an NF-B consensus binding sequence (5Ј-AGT TGA GGG GAC TTT CCC AGG C-3Ј) that was labeled with [␥-32 P]ATP using T4 polynucleotide kinase (Promega). Binding reactions were performed by incubating the samples for 15 min at 22°C. Protein-DNA complexes were separated from the unbound DNA probe by electrophoresis through 5% native polyacrylamide gels containing 0.5ϫ Tris borate/ EDTA. The gel was dried and exposed to BiomaxMR films. Specificity of binding was certified by competition with a 40-fold molar excess of unlabeled double-stranded consensus oligonucleotide and by supershifting with 1 l of anti-NF-B p65 antibody (Santa Cruz Biotechnology). At least three independent experiments of all EMSAs were performed.

Luciferase Reporter Gene Assay
McNtcp.24 cells were cotransfected with 0.2 g of TK-Renilla-CMV and 1.5 g of either p105 or p106. Forty-eight hours later, the cells were cultured in serum-free DMEM for 18 -24 h. Both firefly and Renilla luciferase activities were quantitated using the dual-luciferase reporter assay system (Promega) according to the manufacturer's instructions. Background expression of luciferase, as determined in cells transfected with the p105 vector, was substracted from p106 values.

Adenoviral Infection
The recombinant replication-deficient adenovirus Ad5-IB, containing an IB in which serines 32 and 36 are mutated to alanine (generous gift of D. A. Brenner, University of North Carolina, Chapel Hill, NC), and Ad-⌬E1, an empty adenovirus for control experiments, were grown and purified by banding twice in CsCl gradients as described previously (32). For adenoviral infection, McNtcp.24 cells were grown to 50 -75% confluence. The medium was replaced with DMEM containing 2% fetal bovine serum, and Ad5-IB or Ad-⌬E1 viral stock solutions were added at a multiplicity of infection of 100. Culture dishes were rocked every 15 min for 2 h, and the culture medium was supplemented with 20% serum. Cells were then cultured for an additional 12-16 h before they were used for experiments.

Immunoblot Analysis
Immunoblot analysis of Akt was performed on McNtcp.24 whole cell lysates. Cells were lysed in 100 l of SDS-polyacrylamide gel electrophoresis sample buffer. Proteins were then separated by SDS-polyacrylamide gel electrophoresis (12.5%) and transferred to nitrocellulose. The membrane was blocked with 5% nonfat dried milk in 20 mM Tris, 137 mM NaCl, and 0.05% Tween 20 (pH 7.0) for 60 min and then incubated overnight with a 1:1000 dilution of rabbit anti-Akt or rabbit antiphospho-Akt(Ser 473 ) antibody (New England Biolabs, Beverly, MA). After washing, membranes were incubated for 60 min with a 1:3000 dilution of peroxidase-conjugated goat anti-rabbit IgG (New England Biolabs) and washed again. Bound antibody was visualized using chemiluminescent substrate (ECL, Amersham Pharmacia Biotech) and exposed to X-Omat film. Membranes were then stripped of antibodies in 100 mM mercaptoethanol, 62.5 mM Tris, and 2% SDS for 30 min at 50°C and reblotted overnight with a 1:1000 dilution of goat anti-actin antibody (Santa Cruz Biotechnology) to demonstrate equal protein loading. Blots were washed as described above, incubated for 60 min with a 1:5000 dilution of peroxidase-conjugated swine anti-goat IgG (BIO-SOURCE, Camarillo, CA), and visualized as described above. Antibodies against HA (Santa Cruz Biotechnology) were used in a 1:1000 dilution for immunoblots demonstrating expression of transfected Akt and wild-type and dominant-negative PKC, which all have an HA tag. Blots were washed as described above, incubated for 60 min with a 1:10000 dilution of peroxidase-conjugated goat anti-rabbit IgG (BIO-SOURCE), and visualized as described above. At least three independent experiments of Akt immunoblots and anti-HA immunoblots were performed.

RESULTS
Does PI3K Activity Modulate Bile Acid Cytotoxicity?-To ascertain a putative role for PI3K in modulating bile acid cytotoxicity, we first asked if toxic and nontoxic bile salts activate PI3K. Serum-deprived McNtcp.24 cells were stimulated with 50 M TCDC, 50 M GCDC, or diluent for 15 min. PI3K was immunoprecipitated, and the activity of the immunopurified protein was measured. The nontoxic bile acid TCDC readily activated PI3K, whereas no kinase activity was observed in quiescent cells (Fig. 1A). Wortmannin, a potent PI3K inhibitor (33), blocked the TCDC-induced stimulation of PI3K, demonstrating the specificity of the assay. In marked contrast, the cytotoxic glycine conjugate of chenodeoxycholate did not stimulate PI3K activity, demonstrating that PI3K is differentially activated by structurally similar, but not identical bile acids.
Because the nontoxic bile acid TCDC activated PI3K and the cytotoxic bile acid GCDC did not, we next tested the possibility that PI3K activity might be protective in this setting. Mc-Ntcp.24 cells were pretreated with wortmannin or LY294002 and then incubated with 50 M TCDC for 4 h. Neither PI3K inhibitor alone caused cell death. Additionally, as a single agent, TCDC also did not kill cells. However, TCDC effectively induced apoptosis in cells pretreated with the PI3K inhibitors (Fig. 1B). To further confirm that PI3K activity protected the cells from TCDC-induced cell death, we transiently expressed dominant-negative PI3K and treated the cells with TCDC (Fig.  1C). Consistent with the results seen with the pharmacologic inhibitors, genetic blockade of PI3K activation also sensitized the cells to TCDC-induced apoptosis. These findings predicted that constitutive activation of PI3K would prevent normally toxic bile salts from killing cells. We tested this possibility by expressing constitutively active PI3K, which effectively blocked apoptosis induced by the cytotoxic bile salt GCDC (Fig.  1D). Collectively, these results demonstrate that bile acid-induced PI3K activation can modulate bile acid cytotoxicity. Additionally, they suggest that TCDC is inherently toxic; however, its cytotoxicity is blocked by activation of a PI3Kdependent survival signal.
Which PI3K-activated Effectors Mediate Cell Survival following TCDC Stimulation?-Both Akt and PKC are downstream PI3K effectors that, in some settings, prevent apoptosis. Therefore, we next investigated if these two kinases are involved in downstream signaling of PI3K-mediated survival signaling.
Akt activation was investigated by immunoblot analysis in whole cell lysates using a phospho-specific anti-Akt antibody to identify active Akt (34). Although Akt was expressed in the McNtcp.24 cells, activated Akt was detected only in cells stimulated with insulin (a positive control), but not in those stimulated with TCDC ( Fig. 2A). To demonstrate that Akt does not participate in preventing bile acid-induced cell death, we overexpressed a constitutively active form of Akt in McNtcp.24 cells. Overexpression of active Akt did not decrease GCDCinduced apoptosis compared with control transfected cells; however, staurosporine-induced apoptosis was reduced in cells overexpressing Akt (Fig. 2B), demonstrating that constitutively active Akt was functional in these cells. Collectively, these results suggest that TCDC-mediated PI3K survival signals are unlikely to involve Akt activation.
We next determined if PKC, another downstream effector of PI3K, participates in mediating TCDC-induced activation of cell survival signaling. Analogous to the experimental approach used above for PI3K, McNtcp.24 cells were treated with TCDC or insulin, and PKC activity was measured. As demonstrated in Fig. 3A, TCDC activated PKC in a wortmannin-dependent manner, suggesting that PKC activation requires a PI3K-generated signal. This TCDC-induced activation of PKC was concentration-dependent (Fig. 3B). We next asked whether PKC is also required for the survival response. To demonstrate that we could manipulate PKC activity, McNtcp.24 cells were transfected with empty vector, dominant-negative mu- tant PKC, or wild-type PKC. Transfection of dominant-negative PKC markedly reduced PKC activity, whereas overexpression of wild-type PKC increased total PKC activity (Fig.  3C). We then used this dominant-negative PKC construct to assess whether PKC participates in the survival response. Correspondingly, expression of dominant-negative PKC increased TCDC-induced apoptosis ϳ4-fold compared with cells transfected with an empty plasmid (Fig. 3D). To further assess the possible protective effect of PKC in bile acid-induced apoptosis, McNtcp.24 cells were transfected with either a control expression vector or wild-type PKC. Cells were then treated with the toxic bile acid GCDC, and apoptosis was quantitated. In cells expressing wild-type PKC, GCDC-induced apoptosis was significantly reduced (Fig. 3E). Collectively, our data suggest that PKC is a critical downstream effector in a TCDCactivated, PI3K-dependent survival signaling pathway.
Does TCDC Activate an NF-B-dependent Survival Pathway?-We noted that pretreatment of McNtcp.24 cells with the transcription inhibitor actinomycin D also converted nontoxic TCDC into a cytotoxic bile acid (Fig. 4). This result reminded us of the effects of actinomycin D in TNF-␣ treated cells (35), where the coincident transcriptional activation of anti-apoptotic NF-B target genes prevented TNF-␣ from killing cells. Additionally, PI3K and PKC, both of which are activated by TCDC, can activate NF-B. Taken together, these observations suggest that the anti-apoptotic pathway activated by TCDC might include NF-B.
Activation of NF-B was first assessed by EMSA. Nuclear proteins were extracted from McNtcp.24 cells after treatment with 0 -200 M TCDC or with 28 ng/ml TNF-␣ as a positive control for 1 h (Fig. 5A). TCDC induced a concentration-dependent accumulation of an activity that bound to the NF-B consensus oligonucleotide. This TCDC-induced NF-B activation was maximal at 200 M (data not shown). To demonstrate that the binding activity was due to NF-B, we incubated binding reactions with a supershifting antibody to the p65 subunit of NF-B and with an excess of unlabeled oligonucleotide (Fig. 5A). As shown in Fig. 5B, TCDC-induced activation of NF-B was also time-dependent, starting at 30 min with a maximum at 60 -90 min. Thus, treatment of hepatocytes with TCDC, but not GCDC (Fig. 5D), activates NF-B-DNA binding in a concentration-and time-dependent manner. Next, we determined if TCDC-induced NF-B activation requires PI3K activity. McNtcp.24 cells were treated with 200 M TCDC in the absence or presence of the PI3K inhibitors LY294002 and wortmannin (Fig. 5C). Both of these structurally dissimilar PI3K inhibitors markedly reduced NF-B acti- vation by TCDC, demonstrating that TCDC activates NF-B by a mechanism dependent upon and downstream of PI3K activity.
To confirm that TCDC transcriptionally activates NF-B, we transfected McNtcp.24 cells with a luciferase reporter construct and stimulated the cells with the bile acids TCDC and GCDC. To normalize for transfection efficiency, a control Renilla luciferase construct was cotransfected into the cells. Consistent with the EMSA results, TCDC, but not GCDC, increased expression of the luciferase construct (Fig. 5E). Moreover, the TCDC-induced transcriptional response was blocked by LY294002, whereas LY294002 did not reduce luciferase activity in control cells (data not shown). Thus, both the EMSA and luciferase assays demonstrated that TCDC activated NF-B in a PI3K-dependent manner.
PKC mediates NF-B activation in several cell types. To determine the role of PKC in NF-B activation in our model, we performed NF-B reporter gene assays in cells that were transfected with wild-type or dominant-negative PKC and stimulated with diluent or 100 M TCDC. As shown in Fig. 5F, luciferase activity was markedly increased in TCDC-stimulated cells expressing wild-type PKC, whereas only a minimal increase was observed in cells expressing dominant-negative PKC. These data strongly suggest that PKC regulates NF-B activation in the proposed TCDC-induced survival signaling cascade.
Does TCDC-induced NF-B Activation Participate in the Survival Response?-Unless NF-B is inhibited, TNF-␣-induced apoptosis usually does not occur in many cell types because the simultaneous activation of NF-B blocks this cell death signaling pathway. To determine the role of NF-B in TCDC-induced survival signaling, McNtcp.24 cells were transfected with an adenovirus that expresses the super-repressor of IB (Ad5-IB) or an empty control virus (Ad-⌬E1). Ad5-IB contains an IB that cannot be phosphorylated because serines 32 and 36 have been mutated to alanines (32). At a multiplicity of infection of 100, nearly all cells were infected under these conditions as confirmed by demonstrating a 95% rate of apoptosis in transfected cells treated with 28 ng/ml TNF-␣ for 4 h (data not shown). Control experiments also showed that Ad5-IB effectively blocked TNF-␣-induced activation of NF-B as demonstrated by EMSA, whereas Ad-⌬E1 had no effect on NF-B activation compared with uninfected cells (data not shown). Following infection with Ad5-IB, TCDC-induced apoptosis increased ϳ8-fold compared with cells infected with the empty virus (Fig. 6), thus demonstrating that NF-B plays a key role in the transduction of a TCDC-activated survival signal.

DISCUSSION
In this study, we have demonstrated that the nontoxic bile acid TCDC, but not the toxic bile salt GCDC, activated PI3K and initiated an anti-apoptotic signaling cascade in hepatocytes. Consistent with this observation, inhibition of PI3K transformed TCDC into a cytotoxic agent. Our data also suggest that the PI3K-dependent survival signal is mediated by the atypical PKC isoform PKC, but not by the protein kinase Akt, which is implicated in many PI3K-dependent survival signaling pathways. We also showed that TCDC activated NF-B in a PKC-dependent manner and that NF-B was a key participant in the anti-apoptotic response. Thus, these data demonstrate that some bile acids prevent their inherent cytotoxicity by simultaneously activating intrinsic cell survival signals. These observations are relevant to liver injury in cholestasis and suggest that liver injury can be attenuated by activation of PI3K-dependent survival pathways. Accumulating evidence suggests that bile acids modulate signal transduction pathways in hepatocytes (36). For example, signaling pathways affected by bile acids include protein kinase A-, protein kinase C-, and calcium-dependent signal transduction cascades (36). In addition, the bile acid taurocholate was recently identified as an activator of PI3K. However, these studies did not address the impact of PI3K activation on hepatocyte survival (7). The present studies demonstrate that a nontoxic bile acid selectively activates PI3K-dependent survival signals in hepatocytes. However, the mechanisms by which bile acids activate PI3K remain to be elucidated. Misra et al. (7) could not identify a direct effect of bile acids on PI3K activity, suggesting an indirect mechanism for activation. Bile acids likely stimulate PI3K activity by facilitating its association with receptor complexes known to activate this lipid kinase.
PI3K is implicated as an activator of a variety of anti-apoptotic signaling effectors, including Akt and the atypical isoforms of PKC (PKC and PKC/) (9, 37). We could not find evidence for Akt activation in cells treated with TCDC. This was surprising, as Akt is activated by the PI3K product phosphatidylinositol 3,4,5-trisphosphate (38). Perhaps, bile acids alter the lipid-binding site of Akt, its cellular distribution, or other events necessary for Akt activation by phosphatidylinositol 3,4,5-trisphosphate. Our data did, however, implicate PKC as a critical component of a PI3K-induced signaling cascade. This interpretation is supported by the observation that the PI3K inhibitor wortmannin blocked TCDC-induced activation of PKC. Furthermore, a dominant-negative PKC mutant blocked the PI3K-dependent survival pathway and transformed TCDC to a cytotoxic bile acid. Others have shown that PKC activation resulted in NF-B-dependent transcriptional activity (39), and we confirmed this observation in our model. Two mechanisms have been proposed for PKC regulation of NF-B activity. PKC may activate IB kinase-␤, the kinase that phosphorylates IB␣, resulting in the dissociation of NF-B from its inhibitor and subsequent nuclear translocation of the transcriptionally active subunits (22). Alternatively, PKC activation promotes phosphorylation of the RelA subunit of NF-B, leading to enhanced transcriptional activity (39). The two mechanisms may also act in concert to promote NF-B transcriptional activity. Collectively, our studies implicate PKC as a likely link between TCDC-induced PI3K activation and NF-B activation in bile acid-treated hepatocytes.
We have previously shown that bile salt cytotoxicity both in vivo and in vitro is mediated by the death receptor Fas (40,41). Toxic bile acids induce Fas oligomerization and activate caspase-8, resulting in apoptosis (40). A recent study demonstrated that enhanced PI3K activity inhibits Fas-mediated apoptosis (42). Thus, bile acid-induced Fas activation appears to be inhibited by the simultaneous activation of a kinase-dependent, anti-apoptotic signaling pathway that blocks bile acid cytotoxicity. We identified the transcription factor NF-B as one of the downstream targets of the TCDC-stimulated PI3K activity. NF-B regulates expression of a large number of potential anti-apoptotic genes, including cIAP-1, XIAP, and IEX-IL (15)(16)(17). Interestingly, cIAP-1 inhibits apoptosis by suppressing activation of an apical caspase, most likely caspase-8 (15). Thus, NF-B may suppress bile acid-mediated Fas/caspase-8 activation by up-regulating cIAP-1 expression.
Our observations of TCDC stimulation of both pro-and antiapoptotic signaling processes are reminiscent of signaling by TNF receptor-1 (35). Although TNF receptor-1 signaling has been shown to activate PI3K (18), we observed TCDC-mediated activation of NF-B by EMSA in mouse hepatocytes obtained from TNF receptor-1 knockout animals (data not shown). These data exclude a role for TNF receptor-1 in TCDC-associated activation of NF-B.
The observations presented in this study have significant implications for human liver diseases. Hepatic retention of toxic bile acids is thought to play a key role in liver injury during cholestasis (43) and is, in part, caused by hepatocyte apoptosis (44). Our data suggest that some bile acids attenuate their inherent cytotoxic effects by activating a PI3K-dependent survival signal that is mediated by PKC and NF-B. This concept suggests that the net effect of bile acids in mediating liver injury reflects a balance between pro-and anti-apoptotic processes. Thus, factors that positively modulate the PI3K-dependent survival signaling cascade may attenuate liver injury. Administration of modified bile acids that activate survival pathways might represent a rational therapy for cholestatic liver injury. Such pharmacologic agents could also be useful in other liver diseases associated with Fas-mediated liver injury (e.g. viral hepatitis and alcohol-associated liver disease). Because bile acids can be administered orally and taken up by the liver with a high first pass clearance, these concepts deserve further investigation.