Suppressive effect of ursodeoxycholic acid on type IIA phospholipase A2 expression in HepG2 cells


  • Presented in part at the annual meeting of the American Gastroenterology Association, in May 2004.

  • Potential conflict of interest: Nothing to report.


Phospholipase A2 IIA (PLA2IIA), which plays a crucial role in arachidonic acid metabolism and in inflammation, is upregulated under various pathological conditions, including in the gallbladder and gallbladder bile from patients with multiple cholesterol gallstones, in the liver and kidney of rats with cirrhosis, as well as in the colonic tissue of animals treated with a chemical carcinogen. The administration of ursodeoxycholic acid (UDCA) partially attenuated the PLA2IIA expression level in these different models. The aim of this study was to investigate the modulatory effect of UDCA on the PLA2IIA expression level at the cellular level. The HepG2 cells were selected to investigate the direct inhibitory effect of UDCA on PLA2IIA expression level. The proinflammatory cytokines (interleukin-6 and tumor necrosis factor α) -induced PLA2IIA expression in HepG2 cells was partially inhibited by the presence of UDCA in a dose-dependent fashion. The effect of UDCA on proinflammatory cytokines-induced PLA2IIA expression occurred at the transcriptional level. In addition, among the bile acids tested, this inhibitory effect was UDCA-specific. In conclusion, this study supports the possible alteration of arachidonic acid metabolism and PLA2IIA expression level, in particular, as the protective action of UDCA in patients with chronic liver disease. (HEPATOLOGY 2005;41:896–905.)

Changes in prostaglandin levels derived from arachidonic acid have been linked to hemodynamic alterations in patients with cirrhosis.1–4 One of the rate-limiting steps in the synthesis of prostaglandins is the release of arachidonic acid from membrane phospholipids by the phospholipase A2 (PLA2) enzyme. It has been reported that a secreted type of PLA2, PLA2 IIA, which plays a crucial role in inflammation, was upregulated in the liver and kidney of rats with cirrhosis.5 Although the mechanism accounting for the enhanced release of PLA2IIA in cirrhosis is unknown, several possible hypotheses have been described. The PLA2IIA enzyme is activated by endotoxins and proinflammatory cytokines such as interleukin (IL)-1, IL-6, and tumor necrosis factor alpha (TNF-α).6, 7 Chronic endotoxemia, caused by increased intestinal absorption or decreased hepatic clearance of lipopolysaccharide, has been documented in patients with cirrhosis8, 9 and may account for the upregulation of PLA2IIA. This would agree with existing evidence that leukotrienes and thromboxane play a major role in the pathogenesis of cirrhosis and portal hypertension.10 Thus, the demonstration of an increased leukotriene excretion in patients with cirrhosis is compatible with PLA2 activation. This hypothesis is further supported by the presence of elevated prostacyclin levels in patients with cirrhosis, which correlates with portal hypertension.11 Several reports indicated that the inhibition of key enzymes in the arachidonic acid metabolic pathway could block the progression of the disease and decrease oxidative DNA damage in the animal model of cirrhosis.12, 13

Ursodeoxycholic acid (UDCA) has been widely used for the dissolution of both cholesterol gallstones14 and gallstone fragments after extracorporeal shock-wave lithotripsy.15, 16 Furthermore, recent reports have indicated a variety of characteristics of UDCA in immunomodulation17 and anti-apoptosis.18 Despite its long history as a safe drug for patients with various hepatobiliary disorders, such as primary biliary cirrhosis19, 20 and chronic viral hepatitis,21 the mechanism of action of UDCA has not been fully understood. In addition, although the improvement of biochemical data in patients with these chronic hepatic diseases has been reported, the benefit of its administration for patients in advanced stage has not been determined.

Previously, we have reported that PLA2IIA was upregulated in the gallbladder and gallbladder bile from patients with multiple cholesterol stones.22 Furthermore, we hypothesized that the marked change in biliary composition by long-term administration of UDCA attenuated the exposure of gallbladder mucosa to crystalline cholesterol and reduced the PLA2IIA expression level induced by mucosal inflammation.23 In addition, the expression level, as well as the catalytic activity of PLA2IIA, was upregulated in the colonic tissue of azoxymethane-treated rats.24 Oral administration of UDCA suppressed the marked upregulation of PLA2 in this model.24 One possible explanation for these findings is that the UDCA-induced alteration of the biliary lipid composition contributes to attenuating the inflammatory response in both the gallbladder and colonic epithelia. Conversely, the direct intracellular mechanism of PLA2IIA suppression by UDCA can also be postulated, because an increasing number of studies have revealed the direct modulation of intracellular signaling by UDCA.25

The purpose of this study is to clarify the direct intracellular mechanism of action of UDCA in PLA2IIA expression. The endotoxin and proinflammatory cytokines upregulate the PLA2IIA during inflammation processes. Therefore, the direct involvement of UDCA in association with the cellular PLA2IIA expression level was investigated in vitro, using HepG2 cells, a human hepatoblastoma-derived cell line. HepG2 cells were selected because in this model PLA2IIA expression can be stimulated by proinflammatory cytokines.6


PLA2, phospholipase A2; IL, interleukin; TNF-α, tumor necrosis factor alpha; UDCA, ursodeoxycholic acid; RT-PCR, reverse transcription polymerase chain reaction; CAT, chloramphenicol acetyltransferase; siRNA, small interfering RNA; GR, glucocorticoid receptor; CA, cholic acid; TCA, taurocholic acid; TUDCA, tauroursodeoxycholic acid; DEX, dexamethasone.

Materials and Methods


UDCA and TUDCA were supplied by Mitsubishi Pharma Corporation (Osaka, Japan) with 98% purity by high-performance liquid chromatography. Recombinant human IL-6 and TNF-α were purchased from R&D Systems (Minneapolis, MN). Cytokine stock solutions (10 μg/mL) were prepared in sterile phosphate-buffered saline containing 0.1% bovine serum albumin. Dexamethasone (DEX) was obtained from Wako Chemicals (Tokyo, Japan). RU-486, a potent glucocorticoid receptor antagonist, was supplied by Dr. Sitruckware (Exelgen, Paris, France). The toxicity of these cytokines and reagents were determined by the MTT (3-(4,5-dimethylthiazol-2-yl)–2,5-diphenyl tetrazolium bromide) assay.

Cell Culture.

HepG2 cells (ATCC, Rockville, MD) were grown in minimal essential medium (GIBCO-Invitrogen Japan, Tokyo) supplemented with 10% heat-inactivated fetal calf serum (Hyclone, Logan, UT), 100 units/mL penicillin G, and streptomycin (GIBCO BRL, Grand Island, NY), L-glutamine (GIBCO), and nonessential amino acids (GIBCO). Cells were cultured at 37°C in a humidified incubator containing 5% CO2. In the experiments using either UDCA or DEX, the intrinsic steroids were removed from fetal calf serum according to the method previously reported by Weinstein.26

Determination of Protein Mass of PLA2IIA.

The protein mass of PLA2IIA levels in both cell culture supernatant and human serum were immunoradiometrically assayed as previously described,22, 23 with a minor modification using a combination of two monoclonal antibodies against purified human splenic PLA2II. Assay kits were supplied by Pharmaceuticals Research & Development Division, Shionogi & Co. Ltd. (Osaka, Japan). All assays were performed in triplicate. The quantitative range of PLA2IIA in this assay system was 25,000 to 20,000 ng/dL.

Reverse Transcription Polymerase Chain Reaction and Northern Blotting.

Cellular total RNA was extracted by using RNA-Zol B reagent (Tel-Test, Friendswood, TX) according to the method of Chomczynski and Sacchi.27 Complementary DNAs were synthesized from total RNA with Moloney murine leukaemia virus (M-MLV) reverse transcriptase (Toyobo, Tokyo, Japan) by the random primer method. Expression of PLA2 group IIA and group V were first examined by reverse transcription polymerase chain reaction (RT-PCR). PCR was performed as previously described.22 Aliquots of the reaction mixture were electrophoresed on a 2% agarose gel. Sequences of the PCR products were verified using an ABI PRISM 310 sequencer (Applied Biosystems, Foster City, CA).

The probe for the Northern blot determination of PLA2IIA mRNA was prepared by PCR by using a designated primer set for human PLA2IIA and HepG2 cells cDNA as a template. Twenty micrograms total RNA isolated from HepG2 cells was electrophoresed on a denatured 2% agarose gel, transferred to a nylon membrane (Hybond N+, Amersham-Pharmacia, Buckinghamshire, UK), and hybridized with PCR-amplified partial human PLA2IIA cDNA labeled with [α-32P] adenosine triphosphate (Amersham-Bioscience, Piscataway, NJ) by the random primer method. Expression levels were quantified with imaging analysis by using bio-imaging analyzer (BAS 5000 system, Fuji-Film, Tokyo, Japan) and its software (MacBAS, Fuji-Film).

Immunoprecipitation and Western Blotting.

After treatment with IL-6, the HepG2 cells were harvested by scraping and lysed with a lysis buffer containing 50 mmol/L Tris (pH 7.5), 100 mmol/L NaCl, 50 mmol/L NaF, 3 mmol/L sodium orthovanadate, 0.1% NP40, and protease inhibitors (soy bean trypsin inhibitor, aprotinin, and leupeptin). The cell lysate was incubated at 4°C for 18 hours and with specific antibodies against IL-6 receptor (Santa Cruz Biochemistry, Santa Cruz, CA), gp130 (Upstate Biotechnology, Lake Placid, NY) and STAT 3 (Transduction Laboratories, Lexington, KY) and immunoprecipitated with protein-A coupled to agarose beads (Immunopure Immobilized protein A, Pierce, Rockford, IL). Solubilized proteins were separated by SDS-PAGE and transferred onto polyvinylidene fluoride membrane as previously described. Phosphorylation of gp130 and STAT3 was detected by Western blotting by using an anti-phospho-tyrosine antibody (PY-20, Transduction Laboratory; 4G10, Upstate Biotechnology).

Electrophoretic Mobility Shift Assays.

Extracts of HepG2 nuclei were prepared as described by Osborn et al.28 Double-stranded probes or competitors were produced by diluting the pairs of complementary single-stranded synthesized oligonucleotides (5 μg each) in 20 μL 10 mmol/L Tris HCl (pH 8)–10 mmol/L MgCl2− 50 mmol/L NaCl, heating them to 95°C for 5 minutes, and allowing them to hybridize for 3 hours at room temperature. Single-stranded or double-stranded probes were labeled using T4 kinase and 50 μCi [α-32P] deoxycytidine triphosphate. Free nucleotides were separated from the labeled probes on a Sephadex G-50 column. The specific activity was estimated by spotting 1 μL of the probe (before G-50 chromatography) onto filter paper and counting the respective labeled probes and free nucleotide spots. Extracts of HepG2 nuclei were preincubated for 15 minutes at 4°C in 18 μL 25 mmol/L 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (pH 7.6), 40 mmol/L KCl, 5 mmol/L MgCl2, 1 mmol/L dithiothreitol, 3 μg poly (dI-dC). When competitor oligonucleotides were used, they were added in 100- to 250-fold excess over the concentration of the probe, and the reaction mixture was further incubated for 15 minutes at 4°C. Free DNA and protein-DNA complexes were resolved in a nondenaturing PAGE in buffer containing 6.7 mmol/L Tris-HCl, 3.3 mmol/L sodium acetate, and 1 mmol/L ethylenediaminetetraacetic acid (pH 7.9). The gels were dried and exposed to an imaging plate for BAS3000 (Fuji Film).

Transfection and CATAssays.

HepG2 cells were transfected by the calcium phosphate–DNA coprecipitation method, and harvested 53 hours later. The cells were resuspended in 100 μL 0.1 mol/L Tris-HCl (pH 7.8)–1 mmol/L ethylenediaminetetraacetic acid–150 mmol/L NaCl, and collected by centrifugation at 4°C for 5 minutes. The cells were lysed by incubation in 50 μL 100 mmol/L Tris (pH 7.8)–0.7% Nonidet P-40 for 15 minutes at 4°C, and the nuclei and cell debris were pelletted by low-speed centrifugation. The supernatants were stored at −80°C until analysis. CAT (chloramphenicol acetyltransferase) activity was determined by using a CAT assay kit (Promega, Madison, WI) according to the manufacture's instructions. The β-galactosidase activities of the cell lysates were used to normalize the variations in transfection efficiency. Each transfection was repeated 2 to 4 times with 2 to 3 different preparations of plasmid as indicated in the figure legends.

Glucocorticoid Receptor mRNA Silencing.

The specific small interfering RNA (siRNA) for human glucocorticoid receptor (GR)-alpha was obtained from Santa Cruz Biochemistry. The siRNAs were introduced into HepG2 cells at a concentration of 20 nmol/L by transient transfection with Lipofectamine 2000 transfection reagent (Invitrogen, Carlsbad, CA), using guidelines provided by the manufacturer. The effect of siRNA was evaluated by measuring the relative expression of GR protein versus beta-actin.


Results are expressed as mean ± SEM. ANOVA was followed by Student t test. All experiments were repeated at least 3 times.


Effect of UDCA on Proinflammatory Cytokine-Induced PLA2IIA Expression Level in HepG2 Cells.

The direct effect of UDCA on proinflammatory cytokine-induced PLA2IIA expression was studied. The PLA2IIA released by the HepG2 cells after cytokine addition was significantly increased above basal (6-fold by IL-6, P < .05; 1.5-fold by TNF-α, P < .05; and 13-fold by the combination of IL-6 + TNF-α, P < .005) (Fig. 1A). These results are consistent with a previous report by Crowl et al.6 UDCA had no effect by itself on the basal PLA2IIA expression level. However, pre-incubation with 100 μmol/L UDCA for 24 hours resulted in a significant inhibition of the proinflammatory cytokine-induced PLA2 IIA protein expression (Fig. 1A). The inhibitory effect of UDCA on the PLA2 IIA level was most prominent when stimulated by IL-6 (IL-6, 96.3%; TNF-α, 80%; IL-6/TNF-α, 78%). After pre-incubation of the cells with UDCA for 24 hours, IL-6/TNF-α–induced PLA2 IIA secretion was significantly reduced in a dose-dependent manner (Fig. 1B).

Figure 1.

Inhibitory effect of UDCA on proinflammatory cytokine-induced PLA2IIA expression in HepG2 cells. HepG2 cells were seeded onto 24-well culture plates. The cells at 80% confluent were incubated for 24 hours in the presence or absence of increasing concentrations 25-100 μmol/L UDCA, then incubated after additions of various cytokines (10 ng/mL IL-6, 10 ng/mL TNF-α, or a combination) for 48 hours. The culture medium was replaced once, then further incubated for an additional 48 hours. The culture medium was collected at the end of the incubation period, and the release of PLA2IIA protein into the medium was determined by radioimmunoassay. The effect of 100 μmol/L UDCA on cytokine-induced PLA2IIA release into the culture medium is shown in Fig. 1A. The dose-dependent effect of UDCA on IL-6/TNF-α –induced PLA2IIA expression is reported in Figure 1B. CTL, control; PLA2, phospholipase A2; IL-6, interleukin 6; TNF-α, tumor necrosis factor alpha; UDCA, ursodeoxycholic acid.

To determine whether this suppressive effect was UDCA-specific, HepG2 cells were incubated with different bile acids such as chenodeoxycholic acid (CDCA), cholic acid (CA), the taurine (T) conjugates, trichloroacetic acid (TCA), and tauroursodeoxycholic acid (TUDCA). The appropriate, nontoxic bile acid concentration was assessed by using the MTT assay. The cell viability remains unchanged after 96 hours of incubation of the HepG2 cells with 100 μmol/L of either UDCA, CA, or TCA. However, under the same conditions, significant toxicity was observed with CDCA concentrations greater than 30 μmol/L. Therefore, HepG2 cells were pre-incubated with 100 μmol/L of either UDCA, TUDCA, CA, or TCA, or 30 μmol/L CDCA for 24 hours before the addition of IL-6. IL-6–induced PLA2IIA secretion was significantly suppressed by approximately 58% (P < .005) and approximately 76% (P < .05) with UDCA and TUDCA, respectively (Fig. 2). In addition, although both TUDCA and UDCA have a suppressive effect, the extent of the effect was significantly less with TUDCA (P < .05). Conversely, 100 μmol/L TCA and CA did not significantly alter IL-6–induced secretion of PLA2IIA in HepG2 cells (Fig. 2). The IL-6–induced PLA2IIA expression level was rather increased when tested in the presence of 30 μmol/L CDCA (Fig. 2). Therefore, this study suggests that the inhibitory effect was bile acid specific and that unconjugated UDCA had the most potent inhibitory effect.

Figure 2.

Effect of various bile acids on IL-6/TNF-α-induced PLA2IIA expression in HepG2 cells. The cells were incubated in the presence of 100 μmol/L of either ursodeoxycholic acid (UDCA), taurine (T)-UDCA, cholic acid (CA), or taurocholic acid (TCA) for 24 hours before the addition of 10 ng/mL IL-6 and 10 ng/mL tumor necrosis factor α, and further incubated for 72 hours before PLA2IIA protein release was measured in the culture medium. Only CDCA was tested at a concentration of 30 μmol/L due to its cytotoxicity. PLA2IIA concentration was determined as described in the legend of Fig. 1. PLA2, phospholipase A2; IL-6, interleukin-6; CDCA, chenodeoxycholic acid.

In addition, the steady-state mRNA PLA2IIA expression level was determined by Northern blotting. As shown in Fig. 3A and B, the steady-state level of PLA2IIA mRNA was significantly upregulated by IL-6 treatment, but this increase was significantly reduced by the pre-incubation with 50 μmol/L UDCA (in densitometric analysis, the percentage of reduction of PLA2IIA mRNA by UDCA was 36.4 ± 10.5% from control). Therefore, the inhibition of PLA2IIA expression by UDCA may have occurred, at least in part, at the transcriptional level.

Figure 3.

Effect of UDCA on steady-state PLA2IIA mRNA expression level. HepG2 cells were incubated with 10ng/mL IL-6 for 48 hours after 24-hour preincubation with 100μmol/L UDCA. Total RNA 20 μg was loaded onto a 2% denatured agarose gel, transferred onto a nylon membrane and hybridized with a PLA2IIA mRNA radiolabeled probe. The expression level of the housekeeping G3PDH mRNA was used as control. A representative of three independent experiments is shown (Fig. 3A). The result of densitometric analysis which was standardized with G3PDH mRNA expression level is shown in Fig. 3B. *P < .05, compared to IL-6- treated cells. PLA2, phospholipase A2; IL-6, interleukin 6; UDCA, ursodeoxycholic acid; mRNA, messenger RNA; CTL, control.

Effect of UDCA on IL-6 Receptor Signal Transduction.

The concentration of IL-6 in the culture medium determined by enzyme-linked immunosorbent assay (Endogen, Woburn, MA) was unchanged 24 hours after incubation of the cells with 100 μmol/L UDCA (data not shown). Therefore, the possible direct effect of UDCA on IL-6 expression level alteration can be excluded. To investigate the possible effects of UDCA on the cytokine response, gp130/STAT-3 protein tyrosine phosphorylation was studied. As shown in Fig. 4A, the IL-6 receptor-α expression level was unaffected when incubated for 24 hours with IL-6 in either control cells or the cells pre-treated with 50 μmol/L and 100 μmol/L UDCA for 24 hours. Then, whether IL-6 induced a time-dependent phosphorylation of either gp130 or STAT-3 was tested (Fig. 4B-C). The phosphorylation levels of these molecules were most prominent at 30 minutes after addition of IL-6. Therefore, the effect of UDCA on gp130 and STAT-3 phosphorylation was determined after incubation of the HepG2 cells with IL-6 for 30 minutes. However, pre-incubation of the cells with 100 μmol/L UDCA for 24 hours did not alter the IL-6–induced phosphorylation of either gp130 or STAT-3 (Fig. 4D-E). Taken together, these data suggest that the inhibitory effect of UDCA on PLA2 IIA expression exists downstream from IL6/gp130-STAT3.

Figure 4.

Effect of UDCA on IL-6-induced signal transduction pathway. HepG2 cells were incubated with 10 ng/ml IL-6 in the presence and absence of 100 μmol/L UDCA before the addition of IL-6, unless otherwise indicated. The soluble proteins were collected as described under Materials and Methods. (A) Effect of UDCA on IL-6 receptor expression. The cell lysate was immunoprecipitated with a specific IL-6 receptor antibody, and this precipitant was subjected to Western blotting with a IL-6 receptor antibody. (B, C) Time course of gp130 and STAT-3 phosphorylation. The cell lysates were incubated with an anti-phospho tyrosine antibody (PY-20) and precipitant was separated on a SDS-PAGE. Transferred proteins were blotted with either anti gp130 antibody (B) or anti STAT3 antibody (C). (D) Effects of UDCA on gp130 phosphorylation. The cells were incubated with IL-6 for 30 minutes in the presence and absence of 24-hour preincubation with 100 μmol/L UDCA. Cell lysate was incubated with PY-20 and the transferred precipitant was immunoblotted with an anti gp130 antibody. (E) Effects of UDCA on STAT3 phosphorylation. The cells were incubated with IL-6 for 30 minutes in the presence and absence of 24-hour preincubation with 100 μmol/L UDCA. Phosphorylation of STAT3 was also determined by immunoprecipitation as described. In addition, in this experiment, the effect of DEX was also studied. UDCA, ursodeoxycholic acid; DEX, dexamethasone; IL-6, interleukin 6.

Effect of UDCA on the PLA2IIA Gene Transcription Level.

A series of experiments was performed to determine whether the effect of UDCA was at the level of the promoter for IL-6–induced PLA2IIA gene transcription. The pUC-[-326; +20] - PLA2-H-CAT plasmid was transfected into HepG2 cells, and the CAT activity induced by IL-6 in the presence and absence of 100 μmol/L UDCA was measured. According to previous reports, [-326; +20] is thought to be the most potent region in the PLA2IIA promoter.28 In this study, the addition of 10 ng/mL IL-6 for 30 hours induced the [-326; +20] CAT activity approximately 2.4-fold above basal, which is consistent with previous reports.28 The pre-incubation with 100 μmol/L UDCA for 24 hours significantly attenuated (P < .05) IL-6–induced [-326; +20] CAT activity (Fig. 5A). Previous studies showed the simultaneous presence of positive and negative regulatory elements in the [-326; +20] region of PLA2IIA (see Fig. 5B for details). The element B, spanning region [−125; −80] of the coding strand, has been postulated as the main positive element. In addition, the element C ([−210; −176]) as well as the element D ([−248; −210]) have been considered negative regulating elements of PLA2IIA expression, because the deletion of these elements from [-326; +20] resulted in greater transcriptional activity than that of the whole region.29 We tested the effect of UDCA on the binding ability of the proteins to these different promoter element sequences by electrophoretic mobility shift assays (Fig. 5C). Element B [-125; -80] forms two major complexes with nuclear extracts of HepG2 cells treated with IL-6, as shown in Fig. 5C. Element C [-210; -176] forms 3 major bands with it, and Element D [-248; -210] forms 2 major bands as well, as indicated in Fig. 5C. These associations are specific, because cold oligonucleotide used as competitor could abolish any formed bands (Fig. 5C, lane N). The 24-hour pre-incubation with 100 μmol/L UDCA resulted in the partial competition for the binding of HepG2 nuclear extracts to element B. Furthermore, the enhancement of the binding of 100 μmol/L UDCA pre-treated HepG2 nuclear extracts to element C and D was observed. These changes were also observed in the presence of 100 nmol/L nM DEX. However, the competition or enhancement of binding were less prominent than in the presence of UDCA (Fig. 5C).

Figure 5.

Transcriptional regulation of PLA2IIA by UDCA. (A) The effect of UDCA on the CAT activity of HepG2 cells transfected by the pUC-[-326;+20]-PLA2-SH-CAT construct. Results are the mean ± SEM of the three independent experiments. *P < .005; **P < .05. (B) Putative regulatory elements in the PLA2IIA promoter region according to the footprint assays. (C) HepG2 cells were pretreated with either 100 nmol/L DEX or 100 μmol/L UDCA for 24 hours,then further incubated with 10 ng/mL IL-6 for another 24 hours, respectively. The nuclear extracts were isolated from treated HepG2 cells and incubated with [α32P]-dCTP labeled double-strand oligonucleotid. The formed oligonucleotide-protein complexes were separated in a polyacrylamide gel. Dried gels were exposed onto imaging plate for BAS. The arrows indicate the major complexes formed between respective elements and HepG2 cell nuclear extracts. The formed complexes either between element B (lane B), element C (lane C), or element D (lane D) are shown. Lane N represents the formed complexes between cold competitor oligonuleotide and HepG2 extracts. UDCA, ursodeoxycholic acid; DEX, dexamethasone; IL-6, interleukin 6; CTL, control; CAT, chloramphenicol acetyltransferase.

Relationship Between Glucocorticoid Receptor and UDCA.

Intranuclear GR is thought to be one of the transcriptional factors that can modify the PLA2IIA expression level, because glucocorticoid is a potent inhibitor of PLA2IIA expression by HepG2 cells. In addition, previous reports have supported GR activation by UDCA in a GR ligand independent manner.29, 30 Therefore, the role of GR in the PLA2IIA inhibition by UDCA was investigated. One hundred nanomolar DEX showed a maximum suppressive effect on the PLA2IIA concentration (Fig. 6A). The addition of 50 μmol/L UDCA showed an additive inhibitory effect above that induced by DEX alone (Fig. 6B). Moreover, whereas the specific glucocorticoid receptor inhibitor, RU486, inhibited the suppressive effect of DEX, it could not reverse that of UDCA (Fig. 6C). To further ensure the role of GR, the expression of GR was genetically silenced by siRNA specifically designed for human GR-α. The transfection of siRNA specific for human GR-α resulted in a roughly 50% reduction of the protein expression level of GR (Fig. 6D), whereas non-silencing siRNA transfection did not lead to any significant change in GR expression. The GR-defective cell was stimulated with 10 ng/mL IL-6 after the 24-hour pre-incubation with 100 μmol/L UDCA. Compared with the control, the suppressive effect of UDCA was partially blocked (Fig. 6E), suggesting direct involvement of GR molecule in this mechanism(s).

Figure 6.

Glucocorticoid receptor (GR) and effect of UDCA. (A) Dose-dependent inhibition of PLA2IIA expression by dexamethasone: HepG2 cells were incubated with increasing concentration of dexamethasone (DEX) (1 nmol/L-100 μmol/L) in the presence or absence of 10 ng/mL IL-6 for 72 hours. PLA2IIA release into the culture medium was determined by radioimmunoassay. (B) Effect of UDCA on the IL-6-induced inhibition of PLA2IIA expression by DEX. HepG2 cells were incubated with DEX and UDCA in the presence of 10 ng/mL IL-6 for 72 hours. Released PLA2IIA into culture medium was determined by radioimmunoassay. (C) Effect of RU486 on the inhibitory effect of UDCA. HepG2 cells were incubated with either 100nmol/L DEX or 100 μmol/L UDCA and in the presence and absence of 1 μmol/L RU486, a specific glucocorticoid receptor antagonist. The cells were further stimulated with 10 ng/mL IL-6 for 72 hours. The released PLA2IIA into culture medium was determined by radioimmunoassay. (D) GR-alpha silencing by small interference (si)RNA. siRNA specifically designed for human GR-alpha was transfected into HepG2 cell using Lipofectamine 2000 (Invitrogen). The expression level of GR-alpha was determined by Western blot. Approximately 50% of GR-alpha was suppressed by siRNA 48 hours after transfection. (E) Effect of UDCA on GR-suppressed HepG2 cells. The transfected HepG2 cells were incubated with 10 ng/mL IL-6 in the presence or absence of 100 μmol/L UDCA for 96 hours. The released PLA2IIA into cultre medium was determined by ELISA (Cayman Chemicals, Ann Arbor, MI). *P < .05, compared to IL-6-treated cells. PLA2, phospholipase A2; IL-6, interleukin 6; UDCA, ursodeoxycholic acid; CAT, chloramphenicol acetyltransferase.


UDCA has been recognized as a potential therapeutic agent in various chronic liver diseases.19, 21, 30 Although the mechanism(s) of the UDCA effect on the improvement of liver function are still unclear, it has been postulated that UDCA, a less hydrophobic bile acid, displaces the more hydrophobic, potentially toxic bile acids from the membranes, protecting the hepatocytes.31 Conversely, the immunomodulatory properties32–34 as well as an antiapoptotic effect of UDCA18 have been indicated. In our study, because the stimulation of proinflammatory cytokines did not induce apoptosis, the effect of UDCA was supposedly independent from its antiapoptotic action. Therefore, in the current findings, UDCA-attenuate proinflammatory cytokine-induced PLA2IIA expression is novel and may count as another antiinflammatory property of UDCA.

This study indicates that proinflammatory cytokine-induced PLA2IIA expression was inhibited by UDCA at the transcriptional level. Although there is limited information on the intracellular translocation and secretion machinery of PLA2IIA, this finding suggests that the inhibition of cytokine-induced PLA2IIA by UDCA is an intracellular event. Among the bile acids tested, only UDCA and TUDCA had an inhibitory effect, whereas other bile acids, such as CA and TCA, had no significant effect. This suppressive effect may be a specific characteristic of UDCA and TUDCA. Unconjugated UDCA can to a certain extent cross the plasma membrane because of its relatively higher hydrophobicity.35 Conversely, its conjugate TUDCA has to bind to a specific carrier protein at the plasma membrane level to be internalized, and cells, which lack this transporter, show little uptake.36 Therefore, although UDCA and TUDCA work intracellularly, their differential effect on PLA2IIA expression is probably attributable to differential uptake and signaling mechanisms.36

In this study, extrinsic cytokine stimulation did not affect the intracellular cytokine level, regardless of the presence or absence of UDCA. This indicated that the attenuation of the responsiveness against cytokine stimulation by UDCA is not attributable to the alteration of cytokine expression levels. Neuman et al.37 reported that the increase of these proinflammatory cytokines by ethanol in HepG2 cells was inhibited by UDCA and TUDCA. However, the concentration of cytokines employed in this study was far beyond those determined in their study.37 In our study, the given extracellular cytokine concentration for each group of cells was equivalent regardless of the presence or absence of UDCA. Although we could not exclude the possible alteration in plasma IL-6 level in the clinical study, the administration of UDCA does not affect the plasma IL-6 levels in patients with primary sclerosing cholangitis.38 Taken together, it is proposed that inhibition of PLA2IIA expression after IL-6 production is likely to be one of the major actions of UDCA.

The inhibitory effect of UDCA was most prominent on IL-6–induced PLA2IIA expression. However, the IL-6 receptor expression level was not altered by the presence of UDCA. In addition, neither phosphorylation of gp130 nor phosphorylation of STAT-3 was affected by the presence of UDCA. These findings suggested that there is no modulatory effect of UDCA on the IL-6/STAT3 signal transduction pathway, although the involvement of the cross-talk between the JAK/STAT-3 pathway and other signaling pathways cannot be excluded.

Our data indicate that UDCA can directly attenuate the promoter activity of the human PLA2IIA gene. Furthermore, UDCA can modify the binding between unknown proteins and different promoter regions of PLA2IIA. The overall attenuation of the PLA2IIA promoter activity by UDCA seemed to be a consequence of the reduction or enhancement of several protein associations with different sequences of the promoter. Many other transcriptional factors have been postulated to be regulatory molecules for PLA2IIA expression.39 Recent investigation indicated that the nuclear farnesoid x receptor and oxysterol x receptor bind bile acids, resulting in a positive/negative regulation in the transcription of several gene products.40–44 However, there is no binding site for these nuclear receptors in the human PLA2IIA gene promoter.39 In addition, UDCA has been reported to have a limited affinity for these nuclear receptors.40 Taken together, these bile acid–specific nuclear receptors are unlikely to be associated with the current inhibitory effect of PLA2IIA gene expression.

Recent reports indicated that the exogenous addition of PLA2 potentiates the expression of PLA2IIA induced by proinflammatory cytokines.45 Beck et al.45 reported that PLA2IIA released from mesangial cells by TNF-α stimulates its own expression via an autocrine loop involving cytosolic PLA2 and peroxisome proliferators–activated receptor α.45 In this study, the contribution of exogenous secreted PLA2 in the upregulation of PLA2IIA was calculated to be approximately 50%, whereas the remaining 50% was attributable to the cytokine stimulatory action.45 Therefore, the secreted PLA2IIA released from HepG2 cells after proinflammatory cytokine stimulation may self-amplify the long-term total expression level of PLA2IIA. These phenomena may explain the discrepancies among 40% suppression on promoter activity, 40% suppression in mRNA expression level, and 96% suppression in protein expression level by UDCA in the current study.

Glucocorticoids have been known to inhibit PLA2IIA expression.46 In addition, there is a putative binding site for GR located in the [-208; -203] region of the human PLA2IIA promoter.39, 47 Furthermore, the activation of the GR by UDCA has been previously reported.32, 48 UDCA has been suggested to interact with GR in a distinct manner from the classic receptor–ligand interaction seen with DEX.48 Our findings that (1) UDCA has an additive inhibitory effect above that of DEX on PLA2IIA expression, and (2) RU486, a prototypic GR antagonist, can reverse the effect of DEX but not that of UDCA is compatible with this hypothesis. Furthermore, suppression of GR protein levels by siRNA reveals the requirement of GR for the inhibitory action of UDCA. Therefore, the effect of UDCA on the PLA2IIA expression was most likely mediated through GR. Previous reports also suggested that UDCA induces nuclear translocation and DNA binding activity of the wild-type GR, whereas UDCA-activated GR lacks transactivational potential.32 Furthermore, UDCA suppresses nuclear factor-κB–dependent transcription through the intervention of GR-p65 interaction at least in part via activation of GR.48 Therefore, although the current study shows support for a GR-mediated UDCA action, the detailed mechanisms(s) of UDCA-associated transcriptional regulation of PLA2IIA will require further elucidation.

In conclusion, this study provides evidence for the intracellular suppressive role of UDCA on PLA2IIA expression in human cells. Because arachidonic acid metabolites have been implicated in several important pathophysiological processes in the liver, the suppression of PLA2IIA in various inflammatory conditions may be beneficial for the improvement of the hepatic disorders. Indeed, clinical observations from our laboratory indicate a significantly elevated serum PLA2IIA level in patients with cirrhosis (Ikegami, unpublished findings, 2004).

Furthermore, this increase was partially suppressed by oral administration of UDCA in 12 patients with cirrhosis. However, this clinical study could not clearly pinpoint the clinical significance of this effect because of both the limited number of enrolled patients and the period of time of this clinical study. Nevertheless, a recent report indicates that the inhibition of PLA2IIA by specific inhibitors attenuated hepatic ischemic–reperfusion injury in a canine model.49 Therefore, administration of UDCA should be considered as a therapeutic option in the treatment of patients with chronic liver disease in advanced stage and may be preventive for the endotoxemia-induced systemic complication in patients with cirrhosis.


The authors thank Dr. H. Arita and Mr. S. Endoh (Shionogi Pharmaceutical Co., Ltd, Osaka, Japan) for providing the materials for IIA PLA2 radioimmunoassay. The authors also thank Dr. Sitruckware (Exlegen, Paris) for providing RU486.