Expression of the NF-κB Target Gene X-Ray-Inducible Immediate Early Response Factor-1 Short Enhances TNF-α-Induced Hepatocyte Apoptosis by Inhibiting Akt Activation¹

Tumor necrosis factor-α is a multifunctional cytokine which plays a role in inflammation, immunity, antiviral responses, and a variety of diseases. In the liver, TNF-α modulates hepatocyte responses, depending on the physiological circumstances. TNF-α is particularly important in the pathophysiology of hepatocytes, inducing viral hepatitis, alcoholic liver disease, and fulminant hepatitis (1). TNF-α activates a variety of components implicated in cellular signal transduction. Binding of TNF-α to the TNFR-1 (p55) results in trimerization of its C-terminal cytoplasmic “death domain” and recruitment of some intracellular proteins involved in apoptotic signal transduction (2, 3, 4). However, hepatocytes are normally resistant to the cytotoxicity of TNF-α. Administration of TNF-α alone does not induce hepatocyte apoptosis in mouse liver in vivo (5, 6, 7) or in cultured hepatocytes (8), suggesting that TNF-α also activates molecules which protect cells from apoptosis. For example, TNF-α transmits antiapoptotic signals via NF-κB and phosphatidylinositol 3-kinase (PI3K)³/Akt. Blockage of these signaling pathways results in sensitization of hepatocytes to apoptosis induced by TNF-α (8). NF-κB regulates the expression of antiapoptotic gene products (9, 10, 11), such as antiapoptotic members of the Bcl-2 family (12) and the inhibitor of apoptosis proteins (13). However, the molecules which locate at downstream of NF-κB have not been fully elucidated.

In the present study, to gain insight into the mechanism of hepatocyte sensitization toward TNF-α-induced apoptosis, we have used cDNA microarrays, containing of 1081 human cDNA clones, and monitored gene expression profile which was induced by TNF-α and also associated with NF-κB in Hc human normal hepatocytes. Among several NF-κB-dependent genes, we focused on x-ray-inducible immediate early response factor-1 (IEX-1), which is a unique immediate-early gene that was initially identified following exposure of human squamous carcinoma cells to X-irradiation (14). Recent studies indicate that IEX-1 is implicated in apoptotic signaling and cell cycle control (15, 16, 17, 18, 19, 20, 21, 22). However, IEX-1 appears to exert apparently contradictory effects, depending on the type of cells, stimuli, and its expressed forms. For example, the longer IEX-1 transcript with 37 amino acid insertion, called IEX-1L, is regarded as an apoptosis inhibitor in NF-κB-mediated cell survival (15, 16), whereas IEX-1S, a shorter original IEX-1 transcript promotes apoptosis (17, 18, 19). Interestingly, it has been proposed by two independent investigators that IEX-1S may promote cell proliferation under favorable growth conditions but facilitate apoptosis under unfavorable conditions (17, 18).

In this report, we describe observations regarding the role of IEX-1 in TNF-α-induced signaling of Hc human hepatocytes. First, only IEX-1S is induced by TNF-α in human hepatocytes. Second, IEX-1S promotes TNF-α-induced hepatocyte apoptosis through blockage of the PI3K/Akt survival signaling. Third, TNF-α-induced IEX-1S mRNA expression is positively regulated by NF-κB and negatively controlled by the PI3K/Akt pathway.

Hc cells (normal human hepatocytes) and cell culture media (CS-C complete) were obtained from Applied Cell Biology Research Institute (Kirkland, WA) and Cell Systems (Kirkland, WA), respectively. Recombinant human TNF-α was from Genzyme (Cambridge, MA). LY 294002 was from Alexis (San Diego, CA). Hoechst 33258 (bisbenzimide) staining dye was from Wako (Osaka, Japan). Cy3 and Cy5 monoreactive dye pack, QuickPrep Micro mRNA purification kit, anti-mouse, -rabbit, and -goat IgG HRP-coupled secondary Abs, and ECL Western blot detection system were from Amersham Pharmacia Biotech (Buckinghamshire, U.K.). Abs against Akt and phosphorylated Akt (Ser⁴⁷³) were from Cell Signaling Technology (Beverly, MA). Abs against IEX-1S/L (N-17), Mcl-1 (S-19), Bcl-XS/L (L-19), Bfl-1 (FL-175), and PI3K p110β (S-19) were from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-actin Ab was from Calbiochem (La Jolla, CA). Antiphosphotyrosine Ab (PY20) was from BD Transduction Laboratories (Lexington, KY). High performance thin-layer chromatography (HPTLC) plates were from Merck (Darmstadt, Germany). Isogen was from Nippon gene (Tokyo, Japan). Myc-tagged constitutively active and kinase-dead Akt plasmids (pUSE-myc-active Akt and pUSE-myc-dominant negative Akt) were obtained from Upstate Biotechnology (Lake Placid, NY). Atlas glass fluorescent labeling kit, Atlas glass human 1.0 microarray, and pIRES1neo vector were from Clontech Laboratories (Palo Alto, CA). Lipofectamine Plus, pcDNA3.1/HisA, and anti-Express Ab were from Invitrogen (Carlsbad, CA). [α-³²P]dCTP was from ICN Biomedicals (Costa Mesa, CA). [γ-³²P]ATP was from NEN Life Science Products (Boston, MA). Adenovirus expressing the mutant IκB (Ad5IκB) and control adenovirus Ad5LacZ were provided as previously reported (9). Due to missense mutations at phosphorylation sites where serines 32 and 36 are replaced with alanines, the mutant IκB irreversibly binds to NF-κB, preventing its activation (9). All other reagents used were of the highest analytical grade available.

Hc cells were cultured in CS-C complete medium supplemented with antibiotics (penicillin, streptomycin) (8). Cells (3 × 10⁵, 5 × 10⁵, and 3 × 10⁶) were plated on 35-, 60-, and 100-mm dishes, respectively. After a 12-h incubation in the medium, the cells were washed twice with PBS and the medium was changed to serum-free RPMI 1640. Then, cells were incubated for another 24 h in the presence or absence of recombinant adenoviruses (Ad5IκB, Ad5LacZ) at a multiplicity of infection of 25. Before stimulation with 20 ng/ml TNF-α, the cells were washed twice with PBS and, if necessary, incubated for 1 h in serum-free RPMI 1640 containing 25 μM LY 294002.

Nuclear extracts were prepared as described previously (8). NF-κB-binding consensus single-strand oligonucleotide (5′-TAGTTGAGGGGACTTTCCCAGG-3′) was first annealed with the complement oligonucleotide (5′-TGCCTGGGAAAGTCCCCTCAACTA-3′). The annealed DNA fragment was labeled with [α-³²P]dCTP by Klenow DNA polymerase. Nuclear proteins (20 μg) were incubated with 2.5 ng of ³²P-labeled double-strand oligonucleotide probe for 30 min at room temperature. The mixture was electrophoresed on 4% polyacrylamide gels with 0.5× Tris-borate-EDTA buffer at 4°C. Gels were dried and exposed to film for autoradiography.

An Atlas glass array was used according to the manufacturer’s instructions to compare mRNA expression patterns between cells stimulated with Ad5LacZ + TNF-α and with Ad5LacZ, and between cells stimulated with Ad5IκB + TNF-α and with Ad5LacZ + TNF-α. Total RNA was isolated using Isogen according to the manufacturer’s instructions, and mRNA was purified with a QuickPrep Micro mRNA purification kit. mRNA was reverse-transcribed and labeled with either Cy3 or Cy5 fluorescent probes using Atlas glass fluorescent labeling kit. The cDNA was hybridized on a glass (Atlas Glass Human 1.0 Microarray; Clontech Laboratories) containing 1081 supplied clones, human expression sequences tagged, according to the established protocols. Expression analyses were performed byGenePix 4000A and GenePix Pro 3.0 softwares (Axon Instruments, Union City, CA).

The differential gene expression observed by the microarray analysis was confirmed by Northern blotting. Total RNA (20 μg) was subjected to electrophoresis in 1% agarose/formaldehyde gel. RNA was transferred onto nylon membrane and hybridized with ³²P-labeled probes. IEX-1 probe for Northern blot analysis was prepared by RT-PCR using the primers based on reported cDNA sequence (14); sense, 5′-TCCAGAGGACGCCCCTAACG-3′ and antisense, 5′-GTTCACAGAACATACTAGGC-3′. Primers for detecting IEX-1L and IEX-1S transcripts by PCR were: sense, 5′-TCCGGTCCTGAGATCTTCAC-3′ and antisense, 5′-CTCTTCAGCCATCAGGATCT-3′. Total RNA isolated from TNF-α treated Hc cells was reverse-transcribed using random hexamer-mixed primers. cDNAs were amplified with the primers listed above at an annealing temperature of 58°C for 30 cycles.

For expression of IEX-1S, the full-length cDNA was amplified by PCR with the primers: sense, 5′-CGGAATTCCGATGTGTCACTCTCGCAGCTG-3′ and antisense, 5′-CGGAATTCCGTTAGAAGGCGGCCGGGTGTT-3′. The cDNA fragment was subcloned into a pCR-II-Topo vector and the nucleotide sequence was determined. The plasmid was digested with EcoRI, and the resultant IEX-1S cassette was subcloned into pIRES1neo or pcDNA3.1/His by digestion with EcoRI. To overexpress constitutively active or dominant-negative Akt, expression plasmids pUSE-activated Akt or dominant-negative Akt were used. The expression plasmids were transfected into Hc cells using Lipofectamine Plus according to the manufacturer’s instructions. Stable transfectants were selected in growth medium containing 0.5 mg/ml G418.

For quantitation of apoptotic cells, cells were stained with Hoechst 33258 (Wako) and nuclear morphological changes were examined. Briefly, harvested cells were fixed at 4°C for 30 min with 1% glutaraldehyde in PBS, stained with 1 mM Hoechst 33258 for 30 min, and examined under a fluorescent microscope (Olympus BX60, Tokyo, Japan) with excitation at 360 nm.

Total cellular protein extracts were used for the Western blot analysis of IEX-1, Akt, phospho-Akt, Mcl-1, Bcl-X, Bfl-1, and actin. For the preparation of total cell proteins, cells were sonicated in radioimmunoprecipitation assay buffer (50 mM Tris-HCl, pH 8.8, 150 mM NaCl, 10 mM EGTA, 1% Triton-X, 0.1% SDS, 1% deoxycholic acid, 0.3 mM PMSF, 30 μg/ml (l-3-trans-carboxyoxirane-2-carboryl)-l-leucyl-agmatine (E64), 1 mM sodium orthovanadate, 10 mM sodium fluoride, 0.1 mM sodium molybdate, 0.5 mM 4-deoxypyridoxine). The proteins were separated by SDS-PAGE and were electrophoretically transferred onto polyvinylidene difluoride membranes. The membranes were first incubated with the primary Ab, and then incubated with the anti-mouse, -rabbit, or -goat IgG HRP-coupled secondary Abs. Detection was performed with an ECL system.

PI3K activity was measured as previously described (23). Briefly, Hc cells were washed twice with PBS and lysed in lysis buffer (1% Nonidet P-40, 20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 5 mM EDTA, 1 mM PMSF, 1 mM sodium orthovanadate). The lysate was incubated with anti-phosphotyrosine Ab at 4°C for 2 h. The immunocomplex was precipitated with anti-mouse IgG Ab conjugated with protein A-Sepharose. The immunocomplex was incubated with 200 μg/ml phosphatidylinositol and 10 μM ATP including [γ-³²P]ATP (1.0 μCi/assay) for 30 min at 25°C in 50 μl of reaction buffer (20 mM Tris-HCl, pH 7.5, 100 mM NaCl, 0.5 mM EDTA, 20 mM MgCl₂). Labeled lipids were extracted and separated on HPTLC plates in a solvent system of chloroform:methanol:25% ammonia:water (43:38:5:7, v/v) (23). After autoradiography, the radioactive phosphatidylinositol 3-phosphate spot was scraped off the plate and the radioactivity was measured.

cDNA microarray analysis was performed to search for genes that were up-regulated by TNF-α through a NF-κB-dependent pathway in Hc hepatocytes. TNF-α (20 ng/ml) induced NF-κB activation within 5 min in Hc cells (data not shown). However, its activation was almost abolished when cells were infected with adenovirus expressing the mutant IκB (Ad5IκB) (25 multiplicity of infection) (Fig. 1), as previously reported (8). The genes, whose expression levels increased >2-fold by TNF-α (Ad5LacZ + TNF) compared with Ad5LacZ alone, and also were reduced by >25% by Ad5IκB (Ad5IκB + TNF) treatment compared with Ad5LacZ + TNF, were regarded as positive and are summarized in Table I. These 10 genes account for ∼1% of the total 1081 genes profiled. Among the genes, some are known for their roles in cell death and survival. For example, FLIP is up-regulated through NF-κB activation, resulting in increased resistance to TNF-α (24). Bcl2 and p53 binding protein (BBP/53BP2) is a DNA damage-inducible protein which functions to promote apoptosis upon DNA damage (25, 26). IEX-1L antideath protein has been recognized as an apoptosis inhibitor involved in NF-κB-mediated cell survival (15, 16). In contrast, a proapoptotic role of IEX-1 has also been presented (17, 18, 19). Thus, the role of IEX-1 in apoptotic signaling remains controversial. In this context, we focused on this molecule and examined its function in hepatocytes. To confirm the mRNA expression observed by microarray, Northern blot analysis was performed (Fig. 2,A) using a PCR-derived IEX-1 cDNA probe, which hybridized to both IEX-1S and IEX-1L transcripts. The mRNA expression of IEX-1 was up-regulated by TNF-α and reached a peak at 1 h. Infection of Hc cells with Ad5IκB resulted in a marked inhibition of TNF-α-mediated induction of IEX-1 mRNA, confirming NF-κB-dependent expression. The protein level of IEX-1 was analyzed by Western blot analysis. Consistent with the results of Northern blots, IEX-1 was up-regulated within 1 h after TNF-α treatment and reached a peak at 3 h. (Fig. 2 B). Similar increases of IEX-1 protein after TNF-α treatment were observed in Huh-7 human hepatoma cells (data not shown).

It has been proposed that IEX-1L is merely a transdominant-negative mutant of IEX-1S and is not expressed in native cells (27). To examine expression of IEX-1S and IEX-1L mRNA in Hc cells, PCR analysis was conducted using oligonucleotide primers which permit the detection of both transcripts. Analysis of PCR products revealed that TNF-α-treated Hc cells expressed only IEX-1S mRNA but not IEX-1L mRNA (Fig. 3,A). The Ab against IEX-1 used in the present study, which can recognize IEX-1S and IEX-1L at around 28 and 32 kDa, respectively, detected only a band at 28 kDa (Fig. 3 B), corresponding to the expected size of IEX-1S. Therefore, Northern blot and Western blot analyses reflect the changes of IEX-1S expression.

To investigate the role of IEX-1S in hepatocyte apoptosis induced by TNF-α, Hc cells were stably transfected with pIRES1neo plasmids expressing IEX-1S. pIRES1neo contains the internal ribosome entry site of the encephalomyocarditis virus, which permits the translation of two open reading frames from one mRNA. After selection with G418, nearly all surviving colonies stably express the gene of interest and they were used as IEX-1S-overexpressing cells. TNF-α treatment induced apoptosis of 13.5 ± 2.4% (n = 9) Hc-vector cells infected with Ad5IκB at 24 h, as inferred by Hoechst 33258 staining (Table II). In contrast, the proportion of apoptotic cells was 43.8 ± 3.0% in IEX-1S transfected cells (n = 9). The activation of executor caspase-3, as assessed by the processing of procaspase-3 and hydrolysis of its substrate poly(ADP ribose) polymerase, was enhanced in Hc cells overexpressing IEX-1S (data not shown). However, the levels of TNFR-1 and TNFR-associated death domain protein in IEX-1S-transfected cells were almost identical to those in Hc-vector cells (data not shown). These results suggest that IEX-1S plays a promotive role in TNF-α-induced hepatocyte apoptosis. Fas-mediated apoptosis was also augmented in IEX-1S-transfected cells (data not shown), suggesting that IEX-1S augments apoptosis not in a TNF-α-specific manner, but probably in a general manner.

IEX-1S overexpression was not lethal in Hc hepatocytes, which implies that the enhancement of TNF-α-induced apoptosis by IEX-1S overexpression is not due to promotion of proapoptotic signals. We then hypothesized that IEX-1S inhibits the antiapoptotic signaling induced by TNF-α. Activation of Akt via PI3K has been reported to protect cells from apoptosis induced by TNF-α (8, 28, 29). Bcl-2 family members have been documented to play a central role in regulating the apoptotic process (30). To investigate whether TNF-α affects the regulation of these antiapoptotic factors, activation of Akt, and expression of antiapoptotic members of Bcl-2 family; Bcl-2, Mcl-1, Bcl-X, and Bfl-1, were examined by Western blot analysis. The phosphorylated (activated) form of Akt was detected from 1 to 8 h after administration of TNF-α (Fig. 4,A). The expression level of Mcl-1 was up-regulated at 3 h following TNF-α treatment, although Bcl-X and Bfl-1 levels showed no significant changes (Fig. 4 A). Bcl-2 expression was not detected in Hc cells (data not shown).

It has been reported that the expression of Mcl-1 is controlled by PI3K/Akt (31) and that induction of Bcl-x and Bfl-1 is regulated by NF-κB (12, 32). LY 294002 (25 μM), a PI3K inhibitor, completely prevented TNF-α-induced Akt activation in Hc cells (Fig. 4,B), indicating that Akt is located downstream of PI3K, as previously reported (8). Both NF-κB and PI3K/Akt are involved in the antiapoptotic signaling in TNF-α-stimulated Hc cells (8). Less than 20% of TNF-α-treated cells were apoptotic when infected with Ad5IκB or pretreated with LY 294002, but nearly 50% of cells (50.9 ± 3.0%, n = 9) were apoptotic when pretreated with both Ad5IκB and LY 294002 (Table III), as previously reported (8). To elucidate the involvement of NF-κB and PI3K/Akt in expression of Bcl-2 family members, the effects of Ad5IκB and LY 294002 were examined in Hc cells at 3 h following TNF-α treatment (Fig. 4 B). LY 294002 suppressed TNF-α-induced up-regulation of Mcl-1, but Ad5IκB infection had little effect. Bcl-X and Bfl-1 levels showed no significant changes by LY 294002 and Ad5IκB. These results suggest that expression of Mcl-1 is regulated through PI3K/Akt activation in Hc hepatocytes.

IEX-1S enhanced TNF-α-induced hepatocyte apoptosis, probably through inhibition of the antiapoptotic signaling. Therefore, the changes in TNF-α-induced activation of Akt and expression of Mcl-1 were examined in IEX-1S overexpressing Hc cells (Fig. 5). Western blot analysis revealed that Akt activation and Mcl-1 induction were inhibited in Hc cells transfected with pIRES1neo-IEX-1S (Fig. 5,A). Interestingly, the total amount of Akt was also reduced to an appreciable extent in IEX-1S overexpressing cells. To assess whether the IEX-1S level correlates with the changes in Akt activation and Mcl-1 expression, Hc cells were transfected with pcDNA3.1/His plasmid-expressing IEX-1S with the N-terminal Xpress tag. After selection with G418, the colonies expressing different amounts of IEX-1S were screened (Fig. 5,B). The higher IEX-1S levels produced the greater inhibition of Akt activation and Mcl-1 up-regulation (Fig. 5 B). These results suggest that IEX-1S inhibits activation as well as expression of Akt by TNF-α, leading to suppression of Mcl-1 expression.

To further elucidate the inhibitory action of IEX-1S on the PI3K/Akt pathway, the effect of IEX-1S expression on PI3K activity was determined in TNF-α-treated Hc cells (Fig. 6). There was a 2.5-fold increase in PI3K activity in pIRES1neo-vector-transfected Hc cells for 1 h of treatment with TNF-α (Fig. 6,A). In contrast, the elevation of PI3K activity by TNF-α was not observed in pIRES1neo-IEX-1S-transfected Hc cells. It was to be noted that the level of PI3Kβ (p110) was also discernibly decreased in IEX-1S-overexpressing cells (Fig. 6 B). PI3Kα and PI3Kγ were undetectable in Hc cells by Western blot analysis (data not shown). These results suggest that IEX-1S inhibits both PI3K activation and PI3K expression induced by TNF-α, thereby resulting in blockage of Akt activation.

We have shown that IEX-1S was induced by NF-κB and inhibited PI3K/Akt in TNF-α-stimulated Hc cells. To further clarify the relationship between IEX-1S and PI3K/Akt, the effect of PI3K/Akt on IEX-1S expression was examined. As demonstrated above, infection of Hc cells with Ad5IκB suppressed TNF-α-induced IEX-1S expression. In contrast, LY 294002, a PI3K inhibitor, increased TNF-α-induced IEX-1S expression, at both mRNA and protein levels, in Ad5LacZ- and Ad5IκB-infected cells (Fig. 7). To delineate the role of the PI3K/Akt pathway in IEX-1S expression, Hc cells were stably transfected with plasmids containing kinase-dead or constitutively active Akt. As shown in Fig. 8,A, the levels of IEX-1S mRNA expression by TNF-α were increased and decreased in the cells expressing kinase-dead Akt and the cells expressing constitutively active Akt, respectively. Moreover, the amounts of kinase-dead Akt were correlated with levels of IEX-1S expression (Fig. 8 B). These results indicate that the IEX-1S expression is negatively regulated by the PI3K/Akt pathway.

Our present study demonstrates that TNF-α induces IEX-1S expression through the NF-κB-dependent pathway in hepatocytes. The gene is superinduced by treatment of cells with cycloheximide, suggesting that it is an immediate-early gene (14, 33). This gene is also significantly induced by TNF-α in an NF-κB-dependent manner in other cell types; Huh-7 cells (data not shown), HepG2 cells (20), and HeLa cells (17). IEX-1S had a proapoptotic effect on TNF-α-induced hepatocyte apoptosis through blockage of the antiapoptotic PI3K/Akt pathway. These results are consistent with the recent observations that IEX-1S facilitates apoptosis (17, 18, 19).

IEX-1S itself is not an apoptosis-inducible molecule, because its overexpression was not mortal in Hc cells. IEX-1S promotes proliferation of HeLa cells (17) and HaCaT keratinocytes (18) overexpressing it, when growth conditions were favorable. However, the rate of apoptosis increased in these IEX-1S-overexpressing cells in response to apoptosis stimuli, such as Fas activation, etoposide, comptothecin, and UV B irradiation. In Hc cells transfected with IEX-1S, the growth rate, as assessed by [³H]thymidine incorporation increased under growth conditions (data not shown), whereas apoptosis was facilitated in response to TNF-α and Fas activation. Thus, it may be reasonable to speculate that the proapoptotic effect of IEX-1S is mediated by the inhibition of antiapoptotic signal(s) activated by TNF-α. In several types of cells such as HeLa cells and human endothelial cells, TNF-α appears to activate serine/threonine kinase Akt via PI3K (28, 34, 35, 36), which protects cells from apoptosis (28, 29, 36, 37). In hepatocytes, TNF-α activates the PI3K/Akt pathway which inhibits apoptosis mediated by TNF-α and this protective effect is independent of NF-κB (8). Mechanisms for the antiapoptotic effects of Akt have been reported previously. Pastorino et al. (36) reported that TNF-α induces phosphorylation of a proapoptotic Bcl-2 family member Bad through the PI3K/Akt pathway and that phosphorylated Bad loses the ability to bind to Bcl-x_L, which is known to act on mitochondria to block the apoptotic signaling cascade (38). However, phosphorylation of Bad was not observed in the TNF-α-treated Hc cells (data not shown). The possible implication of PI3K/Akt or NF-κB in the expression of antiapoptotic members of Bcl-2 family has been reported (12, 31, 32). The levels of Bcl-x and Bfl-1 in Hc cells were not affected by TNF-α. The mcl-1 gene, a bcl-2 family member, was originally identified as an early gene induced during differentiation of ML-1 myeloid leukemia cells (39). It has an antiapoptotic function when transfected and overexpressed in mammalian cells (40, 41, 42). Depression of Mcl-1 induction using Mcl-1 antisense oligonucleotide promoted PMA-mediated apoptosis in U937 cells (43). In Hep3B hepatocytes, Mcl-1 is up-regulated by the PI3K/Akt pathway and inhibits apoptosis induced by TGF-β (44). In the present study, it was demonstrated that TNF-α induced Mcl-1 expression via PI3K/Akt. The blockage of PI3K by LY 294002 was observed to enhance TNF-α-induced apoptosis. Taken together, these results suggest that the PI3K/Akt/Mcl-1 pathway plays an important role in regulating apoptosis of hepatocytes. In IEX-1S-overexpressed Hc cells, TNF-α-induced activation of Akt and up-regulation of Mcl-1 were abrogated. These findings demonstrated that the proapoptotic effect of IEX-1S is, at least in part, mediated by blockage of PI3K/Akt/Mcl-1 pathway.

IEX-1S was up-regulated by the NF-κB-dependent pathway in the Hc hepatocyte in response to stimulation by TNF-α. The expression of IEX-1 is induced during cellular differentiation (45) and by a variety of stimuli; DNA-damaging stimuli such as X-irradiation (14) and UV radiation (18, 33); growth factors such as epithelial growth factor (33), tumor-promoting phorbol ester (33), peptide growth factors such as pituitary adenylate cyclase activating peptide (46); and steroid hormones such as 1α,25-dihydroxyvitamin D3 (21). Several regulatory factors such as p53 (20, 47), NF-κB (17, 20), and Sp1 (48) are involved in transcriptional control of the IEX-1 gene. However, the intracellular regulatory mechanism(s) of IEX-1 expression remains unclear. Thus, we have examined whether reciprocal interaction exists between IEX-1S and Akt. Inhibition of PI3K by LY 294002 resulted in the increase of IEX-1 up-regulation induced by TNF-α in both NF-κB-inhibited and control cells (Fig. 7). The same treatment with LY 294002 did not affect NF-κB activation (data not shown), as previously reported (8). Moreover, TNF-α-induced IEX-1 expression was increased by overexpression of dominant-negative Akt and decreased by overexpression of constitutively active Akt, suggesting that the activation of PI3K/Akt by TNF-α suppresses IEX-1 expression through the NF-κB-independent pathway. Thus, it can be considered that the antiapoptotic effect of Akt may be, in part, mediated by down-regulation of this proapoptotic molecule.

IEX-1 appears to exert apparently contradictory effects, depending on the type of cells, stimuli, and its expressed forms. IEX-1L has been reported to protect Jurkat cells expressing a mutant IκBα (15), p65-deficient 3T3 cells (15), respiratory epithelial cells (16) from TNF-α-induced apoptosis. Accordingly, apoptosis of activated T cells was impaired in IEX-1L transgenic mice (49). In contrast, IEX-1S does not prevent cell death but instead triggers apoptosis in HeLa cells (17). The rate of apoptosis in response to stresses (UV B irradiation, camptothecin treatment) is augmented in HaCaT keratinocytes overexpressing IEX-1S (18). In addition, functional disruption of IEX-1 expression by ribozymes decreases the sensitivity of 293 cells to apoptotic triggers, like Fas activation or anticancer drug treatment (19). These reports suggest that IEX-1L functions as an antiapoptotic factor and that IEX-1S facilitates apoptosis under unfavorable conditions. However, it has been proposed that IEX-1L is merely a transdominant-negative mutant of IEX-1S which is not expressed in native cells (27). In Hc hepatocytes, we were unable to detect IEX-1L. Therefore, we can at least speculate that in TNF-α-stimulated human hepatocytes, the IEX-1 transcript plays a proapoptotic role. It has been demonstrated that IEX-1S contributes to cell cycle progression under good growth conditions (14, 17, 18). However, it increases the sensitivity to apoptosis under unfavorable conditions; death receptor activation, chemotherapeutic agents, an x-ray or UV irradiation. This may be due to a signal conflict caused by inappropriate cell cycle progression (50, 51). Other cell cycle-promoting mediators, such as c-myc, E2F, and cyclin D1, also trigger apoptosis induced by various apoptotic stimuli. To further understand the roles of IEX-1 transcripts, it is important to clarify their downstream targets. Besides Akt, there must exist other downstream targets of IEX-1S. The elucidation of the NF-κB-dependent intracellular signaling including IEX-1S will provide a clearer picture of the mechanisms of hepatocyte apoptosis and fulminant hepatitis, and further investigation is under current progress.

In summary, we have shown in this study that TNF-α induces IEX-1S expression through the NF-κB-dependent pathway in human Hc hepatocytes. Despite being a NF-κB-dependent gene, IEX-1S exerted a proapoptotic effect in TNF-α-stimulated hepatocytes. Overexpression of IEX-1S resulted in acceleration of TNF-α-induced apoptosis through blockade of the PI3K/Akt survival pathway. Inversely, Akt activation inhibits IEX-1S expression. These findings lead us to propose the hypothesis that there are reciprocal interactions between IEX-1S located downstream of NF-κB and PI3K/Akt, which mutually regulate TNF-α-induced hepatocyte apoptosis. The hypothetical IEX-1 effects are schematically summarized in Fig. 9.

We thank Dr. Pann-Ghill Suh for AKT expression plasmids.