Constitutive androstane receptor (CAR) ligand, TCPOBOP, attenuates Fas-induced murine liver injury by altering Bcl-2 proteins


  • Potential conflict of interest: Nothing to report.


The constitutive androstane receptor (CAR) modulates xeno- and endobiotic hepatotoxicity by regulating detoxification pathways. Whether activation of CAR may also protect against liver injury by directly blocking apoptosis is unknown. To address this question, CAR wild-type (CAR+/+) and CAR knockout (CAR−/−) mice were treated with the CAR agonist 1,4-bis[2-(3,5-dichloropyridyloxy)] benzene (TCPOBOP) and then with the Fas agonist Jo2 or with concanavalin A (ConA). Following the administration of Jo2, hepatocyte apoptosis, liver injury, and animal fatalities were abated in TCPOBOP-treated CAR+/+ but not in CAR−/− mice. Likewise, acute and chronic ConA-mediated liver injury and fibrosis were also reduced in wild-type versus CAR−/− TCPOBOP-treated mice. The proapoptotic proteins Bak (Bcl-2 antagonistic killer) and Bax (Bcl-2-associated X protein) were depleted in livers from TCPOBOP-treated CAR+/+ mice. In contrast, mRNA expression of the antiapoptotic effector myeloid cell leukemia factor-1 (Mcl-1) was increased fourfold. Mcl-1 promoter activity was increased by transfection with CAR and administration of TCPOBOP in hepatoma cells, consistent with a direct CAR effect on Mcl-1 transcription. Indeed, site-directed mutagenesis of a putative CAR consensus binding sequence on the Mcl-1 promoter decreased Mcl-1 promoter activity. Mcl-1 transgenic animals demonstrated little to no acute liver injury after administration of Jo2, signifying Mcl-1 cytoprotection. In conclusion, these observations support a prominent role for CAR cytoprotection against Fas-mediated hepatocyte injury via a mechanism involving upregulation of Mcl-1 and, likely, downregulation of Bax and Bak. (HEPATOLOGY 2006;44:252–262.)

Hepatocyte apoptosis contributes to hepatocellular injury in diverse human liver diseases.1 Apoptosis can be triggered by activation of either an intrinsic, mitochondria-mediated, pathway, or an extrinsic, death receptor–mediated, pathway.2 Both pathways occur in the liver, but the death receptor–mediated pathway appears to be more common because the liver is exquisitely sensitive to cytotoxicity by the Fas receptor.3 The importance of Fas in liver injury is illustrated by the fatal hepatic injury to mice that occurs following administration of an agonistic Fas antibody4 or concanavalin A (ConA).5 Fas-mediated apoptosis has been implicated in human liver diseases, including viral hepatitis,6 autoimmune hepatitis,7 fulminant hepatic failure,8 acute liver allograft rejection,9 and nonalcoholic steatohepatitis.10 Collectively, these observations suggest Fas is an important death receptor mediating acute and chronic liver injury.

Fas (CD95/APO-1) is a type I transmembrane protein belonging to the tumor necrosis factor receptor superfamily and is the only death receptor known to be constitutively expressed in every type of hepatic cell.11 A key event in Fas signaling is the activation of caspase 8 via formation of a cytoplasmic, receptor-based, multimeric protein complex, referred to as the death-inducing signaling complex (DISC).12 In hepatocytes, caspase 8 activation by the DISC is inefficient, and propagation of the apoptotic signal depends on its amplification via the mitochondrial apoptosis pathway. Thus, regulation of this pathway is critical in the expression of Fas-mediated liver injury.

The proteins that make up the Bcl-2 family are critical regulators of the mitochondrial pathway and can be divided into anti- and proapoptotic proteins. Various antiapoptotic family members, such as myeloid cell leukemia factor–1 (Mcl-1), A1, Bcl-2, Bcl-w, and Bcl-xL, stabilize mitochondria and counteract mitochondrial dysfunction.13 Indeed, inhibition of the mitochondrial pathway ameliorates Fas-mediated liver injury, as observed with Bcl-2 transgenic animals.14 The proapoptotic members of this family can be further subdivided into multidomain and BH3-only proteins. Proapoptotic Bcl-2 multidomain members Bak (Bcl-2 antagonistic killer) and Bax (Bcl-2-associated X protein) permeabilize the mitochondrial outer membrane and are obligate for mitochondrial dysfunction during Fas-mediated apoptosis.15 Proapoptotic BH3-only proteins of the Bcl-2 family, such as Bid, trigger apoptosis by either indirectly or directly activating Bak and/or Bax.16 Consistent with these concepts, Bid knockout animals are resistant to Fas-induced injury.17 Thus, all classes of Bcl-2 proteins regulate Fas-mediated apoptosis in the liver.

Recently, a number of the newer members of the nuclear receptor superfamily have been reported to have profound effects on hepatic physiology.18 In particular, the constitutive androstane receptor (CAR) has been implicated in modulating hepatic detoxification pathways, thereby protecting the liver from bile acid–mediated injury,19, 20 which is dependent on Fas.21 CAR is normally sequestered in the cytoplasm but, in response to specific xenobiotic or endobiotic stimuli, translocates to the nucleus and forms a heterodimer with the retinoid X receptor (RXR).22 This transcription factor complex activates the phenobarbital-responsive enhancer module, which is responsible for the induction of cytochrome (CYP) genes. However, CAR also regulates a diverse set of genes that go beyond drug and steroid metabolism.22 Because of its potent effects on hepatic detoxification pathways and cytoprotection against bile acid–mediated liver injury, we postulated that CAR may also provide more generalized hepatoprotection against Fas-mediated hepatic injury.

The overall objective of the current study was to ascertain whether CAR activation by the CAR agonist and nongenotoxic hepatocarcinogen 1,4-bis[2-(3,5-dichloropyridyloxy)] benzene (TCPOBOP) modulates Fas-induced hepatocyte apoptosis and liver injury. Two fundamental questions were formulated about TCPOBOP-treated livers: i) Does CAR activation inhibit Fas-mediated hepatocyte apoptosis and acute or chronic liver injury and improve survival? ii) Does CAR regulate specific apoptotic proteins, especially Bcl-2 family members? The results indicate that CAR activation in mice depletes hepatocytes of the proapoptotic proteins Bak and Bax and increases expression of the antiapoptotic protein Mcl-1 by directly promoting Mcl-1 transcription. These CAR-dependent processes render the liver resistant to Fas-mediated liver injury. An understanding of these mechanisms may allow the development of CAR agonists as therapeutic strategies for reducing liver injury from Fas-mediated insults.


CAR, constitutive androstane receptor; ConA, concanavalin A; Bak, Bcl-2 antagonistic killer; Bax, Bcl-2-associated X protein; Mcl-1, myeloid cell leukemia factor–1; TCPOBOP; 1,4-bis[2-(3,5-dichloropyridyloxy)] benzene; ALT, alanine aminotransferase.

Materials and Methods

Animal Models.

The care and use of the animals for these studies were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC). These studies employed C57/BL6 wild-type, CAR knockout,23 and Mcl-1 transgenic mice24 (20-25 g body weight). The mice were maintained in a temperature-controlled, pathogen-free environment and fed a standard rodent chow diet and water ad libitum. To assess the effect of CAR activation in the acute experiments, the mice were injected intraperitoneally (i.p.) 3 mg/kg daily for 3 days with either vehicle (corn oil) or TCPOBOP (Sigma-Aldrich, St. Louis, MO).19 In the acute studies, mice were injected i.p. with 10 μg of Jo2 antibody (BD Pharmingen, San Diego, CA) per mouse10 or via the tail vein with ConA (Sigma-Aldrich, St. Louis, MO) at a concentration of 25 mg/kg. In the chronic ConA model of liver fibrosis, mice were given ConA (8 mg/kg) dissolved in 200 μL of pyrogen-free phosphate-buffered saline (PBS) weekly for 6 weeks.25 In these chronic experiments, TCPOBOP was administered (3 mg/kg, i.p.) once weekly, 24 hours before the ConA injection. At selected times after TCPOBOP, Jo2, and/or ConA treatment, the animals were anesthetized with ether, and a hepatectomy was performed prior to euthanasia via exsanguination. Blood was drawn via the portal vein for alanine aminotransferase (ALT) determination.26 Liver tissue sections were placed in fixative for subsequent microscopic analyses. Liver sections were also subjected to RNA extraction using the Trizol reagent (Invitrogen, Carlsbad, CA) or were stored at −80°C for subsequent protein analysis.

Plasmid Transfection.

Huh7 cells, cultured as previously described,27 were transfected with 100 ng of expression vector (mCAR or empty vector) using 1 mL of OptiMEM-1 medium (GIBCO-BRL, Gaithersburg, MD) containing 6 μL of Plus reagent supplied with Lipofectamine Plus (Invitrogen, Carlsbad, CA).28 Cells were cotransfected with 10 ng of Renilla luciferase reporter vector (pRL; Promega, Madison, WI) and 0.5 μg of pGL3-Basic (Promega, Madison, WI) and the human genes Mcl-1 −140, Mcl-1 −294, and Mcl-1 −3893, which were cloned proximal to the luciferase gene in the firefly luciferase-based reporter vector, as previously described.29 TCPOBOP (1 μmol/L) was added at the time of transfection. Cells were then incubated for an additional 24 hours. The cell lysate was then assayed for firefly/Renilla luciferase activity using the Dual-Luciferase Reporter Assay System following the manufacturer's instructions (Promega, Madison, WI).29

Site-Directed Mutagenesis.

To introduce point mutations into a putative CAR-binding site on the human Mcl-1 promoter (AGGTCA CTTGAGGCCA [989-974]), PCR-based site-directed mutagenesis (Quickchange II XL site-directed mutagenesis kit; Stratagene, La Jolla, CA) was performed according to the manufacturer's instructions.30 This site was subjected to mutagenesis, with the GG sequence in the first half of the binding site mutated to TT (mutant I), and the GG in the second half of the binding site mutated to TT (mutant II). The primers used for the mutagenesis were: Mcl-1 mutant I with substitution of TT for GG—forward 5′-CCGAGACAGGCATTTCACTTGAGGCC-3′, reverse 5′-GGCCTCAAGTGAAATGCCTGTCTCGG-3′; Mcl-1 mutant II with substitution of GG for TT—forward 5′-CAGGTCACTTGATTCCATGAGTTCGAG-3′, reverse 5′-CTCGAACTCATGGAATCAAGTGACCTG-3′. The sequences of the engineered reporter constructs were confirmed by direct sequence analysis using an ABI Prism 377 fluorescent DNA sequencer (Perkin-Elmer Life Science, Norwalk, CT).

The two mutated and their parent reporter gene constructs (0.5 μg of each) were cotransfected with pRL (10 ng) to correct for variations in transfection efficiency. Twenty-four hours after transfection, the cell lysates were prepared; both firefly and Renilla luciferase activity was quantitated using the Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer's instructions.

Histology, Electron Microscopy, TUNEL Assay, and Immunohistochemical Identification of Activated Caspases 3/7.

Histology, electron microscopic analyses, the TUNEL assay, and immmunohistochemical analysis for activated caspase 3/7 were performed as previously described by us in detail.31 To accurately quantitate TUNEL-positive and caspase 3/7–positive cells, slides were then viewed by digital microscopy (Axioplan 2, Carl Zeiss, Inc. Oberkochen, Germany). Digital pictures were captured through a video archival system using a digital TV camera system (Axiocam High Resolution color, Carl Zeiss, Inc., Oberkochen, Germany). With an automated software analysis program (KS400, Zeiss, Inc., Oberkochen, Germany), the percent fluorescence/field area of the digital photomicrographs was quantified.

Immunoblot Analysis.

Livers were homogenized in lysis buffer.31 Homogenates were centrifuged at 13,000g at 4°C for 30 minutes in order to obtain cytosolic extracts. To obtain mitochondrial extracts to assess tBid organelle translocation, liver homogenates were prepared as previously described.32 Mitochondrial and cytosolic extracts were then blotted as previously described.32 In cytosolic extracts, expression of the proteins Bax, Bak, Fas, and Bcl-xL was assessed using mouse anti-Bax 1:500, rabbit anti-Bak 1:500, rabbit anti-Fas 1:500 (Santa Cruz Biotechnology Inc., Santa Cruz, CA), and mouse anti-Bcl-xL (Exalpha Biologicals, Maynard, MA). The association of tBid with mitochondria was assessed using goat anti-Bid 1:1,000 (R & D Systems, Minneapolis, MN). The membranes were washed and then incubated with the secondary antibodies antigoat 1:3,000, antirabbit 1:3,000, and antimouse IgG 1:8,000 (Biosource, Camarillo, CA).

Real-Time Polymerase Chain Reaction.

Total RNA was obtained from whole liver as previously described by us.26 After the reverse-transcription reaction, the cDNA template was amplified by PCR with Taq polymerase (Invitrogen), and mRNA was quantitated for the following genes as previously described.33 The primers used were:CYP2B10—forward 5′-TACT CCTATTGCATGTCTCCAAA-3′, reverse 5′-TCCAGAAGTCTCTTTTCACATGT-3′ (118 bp); Mcl-1—forward 5′-GCTCCGGAAACTGGACATTA-3′, reverse 5′-CCCAGTTTGTTACGCCATCT-3′ (82 bp); and αSMA—forward 5′-ACTACTGCCGAGCGTGAGAT-3′, reverse 5′-AAGGTAGACAGCGAAGCCAA-3′ (452 bp). Data were expressed as a ratio of product copies/μL to copies/μL of the housekeeping gene 18S from the same RNA (respective cDNA) sample and PCR run.

Determination of Liver Fibrosis.

Liver fibrosis was quantified using sirius red as previously described.26 The percent red/field area of digital photomicrographs was captured and quantitated as described for the TUNEL assay and immunohistochemistry for caspase 3/7.

Immunohistochemistry for PCNA and αSMA.

Unstained slides of liver tissue sections were deparaffinized and hydrated. Antigen retrieval was performed as previously described.34 Thereafter, the catalyzed signal amplification system (DAKO, Carpinteria, CA) was used for both PCNA and αSMA staining according to the manufacturer's instructions. The primary antibodies used were anti-PCNA (Santa Cruz Biotech, Santa Cruz, CA) and anti-αSMA (Ab1; NeoMarkers, Fremont, CA) mouse monoclonal antibodies, both used at a dilution of 1:400. The percent positive cells/field area of digital photomicrographs was captured and quantitated as described for the TUNEL assay and immunohistochemistry for caspase 3/7.


All data represent at least three independent experiments and are expressed as the mean ± SD unless otherwise indicated. The Student t test was used to calculate P values. P less than .05 was considered significant.


Is Jo2-Mediated Apoptosis Reduced by TCPOBOP?

TCPOBOP, a potent and specific mouse CAR agonist,20 was employed for these experiments. We first verified that the dose, route, and frequency of administration of TCPOBOP was effective in enhancing CAR-dependent expression of the well known CAR target gene CYP2B10.20CYP2B10 mRNA expression was increased 150-fold in CAR+/+ mice treated with TCPOBOP compared to untreated CAR+/+ animals (Fig. 1), but, as expected, TCPOBOP treatment had no effect on CYP2B10 mRNA expression in CAR−/− mice. Next, we examined the effects of TCPOBOP on Jo2-mediated liver cell apoptosis. Liver cell apoptosis (TUNEL- and active caspase 3/7–positive cells/low-power field) were markedly reduced 6 hours after administration of Jo2 in TCPOBOP-treated mice compared with vehicle-treated mice. In contrast, TCPOBOP-treated CAR−/− mice were not protected from Jo2-mediated apoptosis (Fig. 2A,B). Electron microscopic analysis revealed classic morphologic features of apoptosis, including nuclear chromatin marginalization, plasma membrane blebbing, and cytoplasmic condensation in hepatocytes of Jo2-treated animals; these were observed to be reduced in liver specimens from TCPOBOP-treated CAR+/+ animals (Fig. 2C). Thus, the xenobiotic TCPOBOP protects against Fas-induced hepatocyte apoptosis in a CAR-dependent manner.

Figure 1.

A CAR target gene was overexpressed in mice treated with TCPOBOP. CAR activation was assessed by measuring expression of CYP2B10 (a known primary CAR target gene) in whole liver from vehicle- and TCPOBOP-treated CAR+/+ and CAR−/− mice. Expression was measured by real-time PCR and normalized as a ratio using 18S mRNA as housekeeping genes. A value of 1 for this ratio was arbitrarily assigned to the data obtained from vehicle-treated CAR+/+ mice. CYP2B10 mRNA expression was 150-fold higher in TCPOBOP-treated (3 mg/kg, i.p., × 3 d) CAR+/+ mice than in vehicle-treated CAR+/+ mice (*P < .0001). CYP2B10 expression was 1,500-fold higher in TCPOBOP-treated CAR+/+ mice than in similarly treated CAR−/− mice (**P < .0001; n = 5 each group).

Figure 2.

Hepatocyte apoptosis in the Fas-treated mice was attenuated by TCPOBOP. (A and B) Fixed liver specimens were analyzed by TUNEL and immunofluorescence for active caspase 3/7 in order to identify apoptotic liver cells. The percent field area of TUNEL- and active caspase 3/7–positive cells was significantly higher in vehicle-treated mice and 6 hours after treatment in Jo2-treated (10 μg/mouse, i.p.) mice than in TCPOBOP-treated (3 mg/kg, i.p., × 3 d) mice (n = 5 each group; *P < .001). (C) Electron microscopic analysis of CAR+/+ mouse liver tissue specimens—a, basal conditions; b, Jo2 treatment; c, TCPOBOP treatment; d, TCPOBOP and Jo2 treatment—revealed apoptotic nuclei of hepatocytes in livers from vehicle- and Jo2-treated animals (b); this was not appreciated in liver specimens from mice treated with TCPOBOP. Note the typical apoptotic features of chromatin condensation, convolution of nuclear outlines, cell shrinkage, and bleb formation in (B), indicated by arrows (5,000× magnification).

Is Jo2-Mediated Liver Injury Reduced and Survival Increased by TCPOBOP?

Because not all paradigms that attenuate hepatocyte apoptosis reduce liver injury or improve animal survival,35 we next determined whether TCPOBOP treatment ameliorated Fas-induced hepatic injury. Serum ALT values, an index of hepatocyte injury, were significantly lower in TCPOBOP-treated CAR+/+ mice than in vehicle-treated mice after administration of Jo2 (Fig. 3A). Histopathologic examination of liver specimens demonstrated pericentral hemorrhagic hepatitis and apoptosis in vehicle-plus-Jo2-treated mice, which was markedly reduced in TCPOBOP-plus-Jo2-treated CAR+/+ mice (Fig. 3B). TCPOBOP treatment also increased animal survival by 47% after 48 hours (Fig. 3C). In contrast, TCPOBOP did not decrease liver injury or improve animal survival in Jo2-treated CAR−/− animals (Fig. 3A-C). These data suggest that TCPOBOP-mediated CAR activation ameliorates Fas-induced lethal liver injury.

Figure 3.

Liver damage was reduced by TCPOBOP. (A) Six hours after Jo2 treatment (10μg/mouse, i.p.), serum ALT decreased in CAR+/+ animals treated with TCPOBOP (3 mg/kg, i.p. × 3 d) and with Jo2 compared to in vehicle-treated (corn oil) mice (*P < .01). (B) Fixed liver specimens from CAR+/+ and CAR−/− mice 6 hours after Jo2 administration (10 μg/mouse, i.p.) stained by conventional H&E: a, CAR+/+ mice treated with Jo2 plus vehicle (corn oil); b, CAR+/+ pretreated with TCPOBOP (3 mg/kg, i.p., × 3 d) plus Jo2; c, CAR−/− mice treated with Jo2 plus vehicle; d, CAR−/− mice pretreated with TCPOBOP (3 mg/kg, i.p., × 3 d) plus Jo2. Pericentral hemorrhagic hepatitis, including, inflammation, apoptosis, necrosis, and cellular swelling, was absent in liver sections from CAR+/+ mice treated with TCPOBOP and Jo2 (b). (C) CAR+/+ and CAR−/− mice (15 per group) were pretreated with vehicle or TCPOBOP and then treated with Jo2 (25 μg/mouse). Dead mice were counted 48 hours after Jo2 administration. Moribund mice were euthanized according to IACUC standards and counted as dead (*P < .01).

Because TCPOBOP is a known hepatic mitogen,36 we performed immunohistochemical staining for PCNA, which is central to both DNA replication and repair,37 to ascertain if a proliferative response was pivotal in TCPOBOP cytoprotection. TCPOBOP-treated animals showed an increased number of PCN-positive cells/field area compared with that in the controls but did not show a further increase in the number of PCNA-positive cells/field area compared with that in animals treated with Jo2 (data not shown). These data support the conclusion that a reduced apoptosis, not increased proliferation, was responsible for TCPOBOP-mediated, CAR restricted cytoprotection.

Does TCPOBOP Diminish Liver Damage and Fibrosis Induced by ConA Administration?

A well-described mouse model of acute immune-mediated liver injury is the intravenous (i.v.) injection of high doses of ConA.38 ConA-induced hepatic injury is mediated by multiple death ligands including the Fas ligand (Fas L), which is expressed by natural killer T cells (NK-T),5 TNF-α, and TRAIL.5, 39, 40 ConA-induced hepatic injury was markedly decreased in mice treated with TCPOBOP (Fig. 4). There was a significant reduction in serum ALT noted in CAR+/+ mice but not CAR−/− mice treated with ConA plus TCPOBOP versus ConA plus vehicle (Fig. 4A). Histopathologic examination also revealed less periportal and pericentral hepatocyte necrosis, inflammatory cell infiltrates, and cytoplasmic swelling in TCPOBOP-treated CAR+/+ animals than in CAR−/− -treated animals (Fig. 4B). These data suggest CAR agonists may be beneficial in protection against hepatic injury mediated by multiple death receptors.

Figure 4.

TCPOBOP reduced liver damage induced by ConA. TCPOBOP reduced ConA-induced liver injury. (A) Serum ALT decreased in CAR+/+ animals treated with TCPOBOP and ConA compared to in those treated with vehicle (*P < .01). (B) Fixed liver specimens from CAR+/+ and CAR−/− mice treated with ConA (25 mg/kg, i.v., × 16 hr). Mice treated with ConA and vehicle, a and c, and with ConA and TCPOBOP (3 mg/kg, i.p., × 3 d) mice, b and d, were stained by conventional H&E. Liver sections from CAR+/+ mice treated with ConA and vehicle and CAR−/− mice showed extensive liver damage with confluent hepatocyte apoptosis and necrosis with bridging inflammatory-cell infiltrates surrounding the portal and central veins (a, c, d). This was nearly absent in liver specimens from CAR+/+ mice receiving ConA and TCPOBOP pretreatment (b).

Because liver injury, specifically hepatocyte apoptosis, has been linked to liver fibrogenesis,26 we examined the effects of TCPOBOP on hepatic fibrosis during chronic administration of ConA (Fig. 5). Weekly injections of low doses of ConA, a well-established model for assessing liver fibrosis,38 were employed for these studies. TCPOBOP-treated CAR+/+ mice demonstrated decreased liver fibrosis, as assessed by sirius red staining and α-smooth muscle actin (αSMA) immunohistochemistry, an established marker for hepatic stellate cell activation41 (Fig. 5A,B). αSMA mRNA expression also showed a twofold reduction in TCPOBOP-treated animals (Fig. 5C). Thus, CAR activation by TCPOBOP is also effective in reducing hepatic fibrosis.

Figure 5.

TCPOBOP administration reduces liver fibrosis. (A) Quantitation of sirius red–stained liver sections from CAR+/+ mice treated with ConA (8 mg/kg, i.v.) weekly for 6 weeks with and without a once weekly pretreatment of TCPOBOP (3 mg/kg, i.p.). Sirius red staining was quantitatively greater in vehicle/ConA-treated mice than in TCPOBOP-treated animals. (B) αSMA immunostained fixed-liver specimens from CAR+/+ mice treated with ConA with and without TCPOBOP pretreatment. Immunoreactivity for αSMA was greater in vehicle/ConA-treated mice (a) than in TCPOBOP ConA-treated animals (b). The immunoreactivity was sinusoidal in location, consistent with staining of hepatic stellate cells or myofibroblasts. (C) Liver tissue was isolated and total RNA was extracted from TCPOBOP- and vehicle-treated mice after weekly ConA injections, as described in the Materials and Methods section. αSMA mRNA expression was quantitated by real-time PCR. Liver tissue from TCPOBOP-treated mice showed reduced expression of αSMA compared to that in vehicle-treated mice.

Do TCPOBOP-Treated Animals Modulate Expression of Antiapoptotic and/or Proapoptotic Bcl-2 Proteins?

Fas-mediated liver injury requires activation of the proapoptotic Bcl-2 family protein Bid.42 Active caspase 8 is generated in the Fas receptor complex, which in turn cleaves Bid, forming truncated Bid (tBid). tBid activates Bax and Bak, triggering the mitochondrial pathway of apoptosis.43 The effect of the stimulation of CAR on the expression of these apoptotic effector molecules was examined. In TCPOBOP-treated mice, Fas (data not shown) and Bid expression was unchanged, and mitochondrial-associated tBid was identified following administration of Jo2 (Fig. 6A). These observations demonstrate TCPOBOP does not prevent Fas-mediated caspase-8 generation and cleavage of Bid. Thus, TCPOBOP appears to block liver injury by interfering with events downstream of Bid at the level of the mitochondria, an organelle whose participation in apoptosis is regulated by members of the Bcl-2 protein family. Indeed, expression of the multidomain proapoptotic proteins Bak and Bax was markedly reduced in TCPOBOP-treated CAR+/+ mice compared with that in vehicle-treated or CAR−/− mice (Fig. 6B). Thus, TCPOBOP prevents Fas-mediated apoptosis in a CAR-restricted manner at the level of the mitochondria. Although this effect could certainly be a result of Bak and Bax depletion, upregulation of a protective Bcl-2 protein is not excluded as a mechanism of this cytoprotection. The antiapoptotic Bcl-2 proteins expressed in hepatocytes are Mcl-1 and Bcl-xL, as murine hepatocytes do not express Bcl-2.44 Expression of Mcl-1 mRNA increased fourfold in CAR+/+ mice treated with TCPOBOP compared with that in vehicle-treated mice (Fig. 6C). In contrast, an increase of Bcl-xL protein was not observed (data not shown). Taken together, these data suggest CAR cytoprotection is mediated by loss of the multidomain Bcl-2 proapoptotic proteins Bak and Bax and by an increase in the antiapoptotic Bcl-2 protein Mcl-1.

Figure 6.

TCPOBOP-treated mice overexpressed Mcl-1 and underexpressed Bak and Bax. (A) Representative immunoblot analysis for Bid and tBid from mitochondrial extracts of CAR+/+ mouse livers treated with Jo2 (10μg/mouse, i.p.) for 6 hours with and without TCPOBOP pretreatment (3mg/kg, i.p., × 3 d). Cytochrome oxidase was used as a loading control. (B) Representative immunoblot analysis for Bak and Bax from CAR+/+ and CAR−/− mouse livers treated with vehicle or TCPOBOP (3mg/kg, i.p., × 3 d). Actin was used as a loading control. (C) Liver tissue was isolated, and total RNA was extracted from TCPOBOP- and vehicle-treated mice as described in the Materials and Methods section. Mcl-1 m RNA expression was quantitated by real-time PCR. Liver tissue from TCPOBOP-treated mice showed enhanced expression of Mcl-1 compared to that from vehicle-treated mice (*P = .02).

Does CAR Have a Direct Effect on Mcl-1 Transcription, and Is Mcl-1 Up-Regulation Sufficient to Block Fas-Mediated Liver Injury?

The increase in Mcl-1 mRNA was likely mediated at the transcriptional level, as CAR transfection increased Mcl-1 promoter activity (Fig. 7). To identify a potential CAR-binding site in the human Mcl-1 promoter, deletion constructs were generated (Fig. 7A) and assessed in a reporter gene assay employing the human hepatoma cell line Huh-7. These constructs demonstrated that deletion of the promoter proximal to the 294-bp region reduced promoter activity (Fig. 7B). A computer search demonstrated a putative consensus CAR-binding site at position 989-974, which contains the hexamer direct repeat 4 sequence AGGTCA CTTG AGGCCA. Indeed, transfection of CAR with the administration of TCPOBOP did further increase Mcl-1 promoter activity in the construct that contained this region (3,893-bp region; Fig. 7C). To further assess the functional importance of this putative CAR-binding site, we generated two mutant constructs, mutants I and II, in which alteration of residues 988 and 987 and of residues 978 and 977 occurred. Both these point mutations significantly decreased reporter gene activity with and without administration of TCPOBOP (Fig. 7D). Administration of TCPOBOP further increased Mcl-1 promoter activity when cotransfected with the native construct and mCAR compared to without TCPOBOP (P = .01), consistent with a functional CAR-binding site.

Figure 7.

TCPOBOP-induced Mcl-1 expression was CAR dependent. (A) A series of human Mcl-1 promoter constructs used in the reporter gene assays are schematically illustrated. They were generated by restriction enzyme digestion and cloned proximal to the luciferase gene in the firefly luciferase-based reporter vector pGL3-Basic. The HindIII/SpeI, HindIII/XhoI, or HindIII/PvuII reporter gene constructs encoded 3,893, 294, and 140 bp of the Mcl-1 promoter, respectively. (B) Huh-7 cells were cotransfected with 10 ng of the Renilla luciferase reporter plasmid, 100 ng of expression vector (mCAR or empty vector), plus 0.5 μg of the Basic, −140, −294, or −3893 firefly luciferase-based constructs. Twenty-four hours after transfection, cell lysates were prepared to perform reporter gene assays. The −3893 Mcl-1 promoter construct showed increased activity when cotransfected with mCAR versus that of the empty vector (P = .0001). (C) Huh-7 cells transfected with the mCAR or empty vector were cotransfected with pRL-CMV promoter encoding the Renilla luciferase gene plus −3893 reporter gene construct with and without TCPOBOP 1 (μmol/L) administration. After 24-hour transfection, cell lysates were prepared to perform reporter gene assays. TCPOBOP increased promoter activity in the −3893 reporter gene construct. (D) Huh-7 cells were also transfected with or without TCPOBOP (1 μmol/L) and with a mutant construct in which residues 989-974 were changed from GG to TT, and prepared cell lysates were subjected to the reporter gene assay. Point mutations of a CAR-binding site in the 5′-flanking region of the Mcl-1 gene (Mutant I and Mutant II) abrogated Mcl-1 promoter activity as compared to the full length Mcl-1 construct (−3893). There was no significant difference in promoter activity of mutants cotransfected with mCAR with and without TCPOBOP. Cotransfection with the addition of TCPOBOP to the full-length construct and mCAR showd significantly increased (*P = .01) luciferase activity (149 ± 42.7) compared to that without the addition of TCPOBOP (75 ± 7.1). Data from four individual experiments are expressed as the ratio of firefly luciferase/Renilla luciferase luminescence intensity.

Finally, to determine if upregulation of Mcl-1 is sufficient to attenuate Fas-mediated liver injury, we examined Mcl-1 transgenic animals after they were administered Jo2. These animals showed little to no acute liver injury, as assessed by serum ALT (Fig. 8). These data suggest Mcl-1 upregulation is sufficient to ameliorate Fas-mediated liver injury and that CAR-directed cytoprotection can be explained in part by increased Mcl-1 expression.

Figure 8.

Mcl-1 upregulation alone was adequate to abrogate Fas-mediated liver injury. Serum ALT values were unchanged in Mcl-1 transgenic mice given Jo2 (10 μg per mouse) assessed 6 hours after administration.


The principal findings of this study relate to the mechanisms of CAR cytoprotection during Fas-mediated liver injury. The results demonstrated that the murine CAR agonist TCPOBOP: i) attenuates Fas- and ConA-mediated liver injury, ii) depletes the liver of proapoptotic proteins Bak and Bax, and iii) increases Mcl-1 expression in a CAR-dependent manner. These data suggest CAR activation can reduce hepatocellular injury induced by Fas and likely by other death ligands by regulating expression of Bcl-2 proteins.

Both the Jo2 and ConA models of acute hepatitis were employed for this study. The Jo2 model mimics soluble Fas ligand, whereas the ConA model duplicates liver injury by membrane-bound death receptors expressed by NK-T cells.5 In the current study, TCPOBOP treatment reduced serum ALT and improved liver histology following administration of either Jo2 or ConA. Collectively, these data suggest CAR activation is cytoprotective during acute death receptor–mediated liver injury. We also assessed chronic liver injury following weekly injections of ConA, a well-established model for liver fibrosis.38 TCPOBOP pretreatment reduced liver fibrosis as assessed by quantitative sirius red staining and αSMA expression. These data provide further evidence of a mechanistic link between hepatocyte apoptosis and liver fibrosis.26 The ability of activated CAR to block the sequelae of chronic liver injury also emphasizes how robust the cytoprotection is provided by CAR.

Consistent with its antiapoptotic effects, TCPOBOP CAR dependently modified expression of pro- and antiapoptotic Bcl-2 family protein members. Activation of CAR reduced cellular levels of the proapoptotic multidomain Bcl-2 family proteins Bax and Bak. In many cell types such as hepatocytes, death receptor–mediated cytotoxicity requires Bax and Bak activation for mitochondrial dysfunction and cell death.13 Moreover, fibroblasts derived from Bax and Bak double-knockout mice are resistant to Fas-mediated apoptosis.15 The down-regulation of Bax and Bak observed after TCPOBOP treatment therefore may account for its cytoprotection against Fas-mediated liver injury.

Mcl-1 is a potent multidomain antiapoptotic Bcl-2 family protein expressed in the liver.16 We observed a direct regulatory effect of CAR on Mcl-1 expression, and furthermore, Mcl-1 transgenic animals were protected from Fas-induced liver injury. Although, Mcl-1 can be regulated by translational and post-translational mechanisms, TCPOBOP stimulation of CAR directly regulates Mcl-1 promoter activity. Indeed, point mutations of a CAR-binding site in the 5′-flanking region of the human Mcl-1 gene abrogated Mcl-1 promoter activity. To our knowledge, our study is the first to characterize these specific binding sequences for CAR within the human Mcl-1 promoter. Thus, CAR appears to have a direct effect on Mcl-1 expression. Consistent with our observations, hepatocyte growth factor–mediated resistance to Fas cytotoxicity has been attributed to upregulation of Mcl-1.45 Taken together, our observations identify a novel mechanism for CAR cytoprotection that is independent of its well-established effects on detoxification pathways, namely, modulation of Bcl-2 family proteins.

Bile acids, especially lithocholic acid, activate CAR, which mitigates their toxicity.46 Other nuclear receptors such as pregnane X receptor are complementary to CAR in ameliorating bile acid cytotoxicity.47 CAR cytoprotection in these studies has been attributed to detoxification of lithocholic acid by its biotransformation via several pathways including sulfation.48 The sulfated form of lithocholic acid induces hepatocyte apoptosis by a Fas-dependent mechanism.49 The current studies demonstrating that CAR induces Mcl-1-expression and decreases cellular levels of Bax and Bak implicate additional pathways by which CAR may ameliorate lithocholic acid injury.

The role of CAR in modulating hepatotoxicity is dependent on the nature of the stimulus. For example, by increasing biotransformation of acetaminophen, CAR agonists increase acetaminophen hepatotoxicity.50 CAR agonists may also modulate bile acid synthesis in such a manner as to accentuate cholestatic liver injury.51 However, Fas-mediated apoptosis and injury contribute to a multitude of human liver diseases, and CAR activation confers protection against Fas-mediated cell death and hepatic injury. Therefore, CAR agonists may prove therapeutically useful in many human liver diseases characterized by death receptor–mediated injury.


The exquisite secretarial assistance of Erin Bungum is greatly appreciated.