Role of the CCAAT/Enhancer-binding Protein NFATc2 Transcription Factor Cascade in the Induction of Secretory Phospholipase A2*

Inflammatory cytokines such as interleukin-1 and tumor necrosis factor-α modulate a transcription factor cascade in the liver to induce and sustain an acute and systemic defense against foreign entities. The transcription factors involved include NF-κB, STAT, and CCAAT/enhancer-binding protein (C/EBP). Whether the NFAT group of transcription factors (which was first characterized as playing an important role in cytokine gene expression in the adaptive response in immune cells) participates in the acute-phase response in hepatocytes is not known. Here, we have investigated whether NFAT is part of the transcription factor cascade in hepatocytes during inflammatory stress. We report that interleukin-1 or tumor necrosis factor-α increases expression of and activates NFATc2. C/EBP-mediated NFATc2 induction is temporally required for expression of type IIA secretory phospholipase A2. NFATc2 is also required for expression of phospholipase D1 and the calcium-binding protein S100A3. Thus, a C/EBP-NFATc2 transcription factor cascade provides an additional means to modulate the acute-phase response upon stimulation with inflammatory cytokines.

The inflammatory response is a part of a protective means to restore cellular homeostasis and to maintain normal physiological processes after invasion by foreign entities (reviewed in Refs. [1][2][3]. In reaction to tissue injury and infection, inflammatory cytokines such as interleukin-1 (IL-1) 2 and tumor necrosis factor-␣ (TNF-␣) are released to mediate an acute and systemic defense to restore cellular homeostasis. Both immune and nonimmune cells are actively involved in these processes. For example, differentiation of CD4 ϩ T cells into a specific T-helper cell lineage and increased production of plasma proteins in hepatocytes are part of these protective changes.
Upon stimulation with inflammatory cytokines, the expression of plasma proteins, proteases, and procoagulants in hepatocytes is increased in part by pre-existing and newly synthesized transcription factors (4,5). Proteins involved in this transcription factor cascade include NF-B, STAT (signal transducer and activator of transcription), and CCAAT/ enhancer-binding protein (C/EBP) family members. Cooperative action mediated by NF-B and C/EBP is also important for the induction of early phase inflammatory-responsive genes (6 -8).
In addition to the transcription factors that induce plasma proteins and proteases involved in the complement and coagulation cascades, early phase responsive genes also include transcription factors C/EBP␤ and C/EBP␦ (9). The induction of C/EBP␤ and C/EBP␦ suggests an autoregulation in gene expression among the members of the C/EBP family. The newly synthesized C/EBP proteins may then sustain and/or induce additional anti-inflammatory target genes to uphold transcription activation initiated by the inflammatory cytokines. Whether the expression of other transcription factors, in addition to C/EBP proteins, is induced and participates in the later stage of the inflammatory response to provide necessary metabolic changes to re-establish cellular homeostasis remains elusive. Understanding the basic mechanism of activation of the transcription factor cascade mediated by inflammatory cytokines will shed new light on the treatment of chronic inflammation such as sepsis and atherosclerosis.
The NFAT (nuclear factor of activated T cell) group of proteins was first characterized to play an important role in cytokine gene expression in immune cells (reviewed in Refs. 10 -12). Subsequent studies demonstrated that NFAT participates in multiple physiological processes. Recently, NFAT activity has been demonstrated in cardiac hypertrophy, adipocyte differentiation, and learning and memory (13)(14)(15)(16)(17)(18). Despite the critical function of NFAT in the adaptive immune response in T and B cells (reviewed in Refs. 19 and 20), the role of NFAT in the acute-phase response in hepatocytes is not known.
Four distinct genes encoding closely related, calcium/calcineurin-regulated NFAT proteins (NFATc1/NFATc/NFAT2, NFATc2/NFATp/ NFAT1, NFATc3/NFAT4/NFATx, and NFATc4/NFAT3) have been identified (reviewed in Refs. 21 and 22). Alternative mRNA splicing of these four genes further generates multiple NFAT polypeptides. The function of these alternatively spliced NFAT isoforms remains elusive. However, all NFAT family members contain a highly conserved N-terminal NFAT homology domain for calcium/calcineurin regulation and a C-terminal Rel homology region for DNA binding. The presence of conserved regulatory domains suggests functional redundancy among different NFAT family members, although specific regulation mediated by individual NFAT family members on selective targets remains elusive.
The purpose of this study was to examine the role of NFAT in hepatocytes during inflammatory stress. Because inflammatory cytokines mediate acute and systemic responses by regulating transcription factor expression/activation, we ask whether NFAT is part of the transcription factor cascade during inflammatory stress. Here, we report the induction and activation of NFATc2, but not other NFAT family members, by the inflammatory cytokine IL-1 or TNF-␣. Mechanistically, IL-1 induces C/EBP binding to NFATc2 transcription loci and mediates gene expression. C/EBP-mediated NFATc2 induction is temporally required for the expression of type IIA secretory phospholipase A 2 (sPLA 2 -IIA). NFATc2 is also required for the expression of phospholipase D 1 (PLD 1 ) and the calcium-binding protein S100A3. The C/EBP-NFATc2 transcription factor cascade provides an additional means to modulate the acute-phase response in the liver.

EXPERIMENTAL PROCEDURES
Reagents-The human NFATc2 (Ϫ1 to Ϫ2000 bp) and sPLA 2 -IIA (Ϫ1 to Ϫ1200 bp) gene promoters were amplified from human genomic DNA and subcloned into the pGL3-luciferase reporter plasmid (Promega) using the KpnI and XhoI sites and the MluI and XhoI sites, respectively. Deletions and mutations in the NFATc2 and sPLA 2 -IIA promoters were generated by PCR. The expression vectors for C/EBP␣, C/EBP␤, C/EBP␥, and NFATc2 have been described (13,23). The human NFATc2-specific short hairpin RNA (shRNA) was obtained from Open Biosystems (GenBank TM accession numbers NM_012340). Anti-sPLA 2 -IIA polyclonal antibody and a human sPLA 2 -IIA enzymelinked immunosorbent assay kit were purchased from Cayman Chemical Co. Anti-NFATc2 antibody was generously provided by Dr. Nancy Rice (antibody 1777) (24). Control rabbit and mouse IgG and anti-NFAT, anti-C/EBP, and anti-YY1 antibodies were purchased from Santa Cruz Biotechnology, Inc., and/or Affinity BioReagents. Anti-␤-actin Monoclonal antibody was purchased from Sigma. Anti-acetylated histone H3 polyclonal antibody was obtained from Upstate.
Cell Culture-HepG2 hepatoma cells were cultured in minimum Eagle's medium. HEK293 and Huh7 cells were cultured in Dulbecco's modified Eagle's medium. Primary hepatocytes from C57BL/6 mouse liver were prepared by collagenase digestion and cultured in Dulbecco's modified Eagle's medium in the presence of 1 M dexamethasone and 10 g/ml insulin as described (25). All media were also supplemented with 10% fetal calf serum, 2 mM L-glutamine, 100 units/ml penicillin, and 100 g/ml streptomycin (Invitrogen). Cells were transfected using Lipofectamine (Invitrogen). Cells were stimulated with IL-1 (4 ng/ml) or TNF-␣ (5 ng/ml) for the indicated times before harvest. To generate NFATc2 knockdown HepG2 cells, NFATc2 shRNA was transfected and selected under puromycin (0.1 mg/ml) for stably transfected clones. A similar protocol was followed to generate HepG2 cells expressing a dominant-negative (dn) NFAT inhibitor, except that cells were selected for neomycin resistance (1 mg/ml). Huh7 cells stably integrated with NFAT-luciferase reporter plasmids encoding triple repeats of the peroxisome proliferator-activated receptor-␥2 (PPAR␥2) proximal or distal NFAT-binding elements were similarly selected for neomycin resistance. Results were obtained with both pooled and single-cell clones. Representatives of isolated single-cell clones are presented.
Microarray Analysis-Microarray analysis was designed to profile gene expression upon administration of the inflammatory cytokine IL-1. The role of calcineurin/NFAT in IL-1-regulated genes was also investigated by pretreatment with the calcineurin inhibitor cyclosporin A (CsA). Glass oligonucleotide arrays manufactured in our genomic facility were used. These glass oligonucleotide arrays contain 34,580 70-mer probe sets with controls (Qiagen Human Genome Oligo Set Version 3.0). Total RNA isolated from IL-1-treated (12 h) or untreated HepG2 cells in the presence and absence of CsA was labeled with Cy5or Cy3-labeled dUTP and hybridized to glass oligonucleotide arrays. Gene annotation and function were determined using the Database for Annotation, Visualization and Integrated Discovery at NIAID, National Institutes of Health (available at david.niaid.nih.gov/david/) with assist-ance from our microarray facility. Candidate genes were selected with a difference of Ͼ1.5-fold and are reported (see supplemental material).
Binding reactions for gel mobility shift assays were carried out at room temperature in gel shift buffer (1 mM CaCl 2 , 1 mM MgCl 2 , 10 mM HEPES (pH 7.9), 50 mM NaCl, 15 mM ␤-mercaptoethanol, 10% glycerol, 0.1 mg/ml bovine serum albumin, and 1 mg/ml poly(dI-dC) for 30 min. Protein⅐DNA complexes were separated on 5% nondenaturing polyacrylamide gels in 25 mM Tris, 200 mM glycine, and 1 mM EDTA and visualized by autoradiography. For supershift analysis, antibody was preincubated with a nuclear extract for 30 min at before addition of the labeled probe. For competition analysis, an excess amount of the unlabeled oligonucleotide (50 fmol) was incubated together with the labeled probe before addition of the nuclear extract.
Luciferase Assays-The NFATc2 or sPLA 2 -IIA promoter plasmid (0.3 g) was cotransfected with the pRSV-␤-galactosidase control plasmid (0.1 g) or the C/EBP or NFATc2 expression plasmid (0.3 g) as indi-cated. Luciferase and ␤-galactosidase activities were measured 48 h after transfection. The data are presented as relative luciferase activity calculated as the ratio of luciferase activity to ␤-galactosidase activity (means Ϯ S.E., n ϭ 4).

RESULTS
Inflammatory Cytokine IL-1 or TNF-␣ Induces NFATc2 Expression-To investigate the role of NFAT during inflammatory stress in hepatocytes, we tested whether inflammatory cytokines such as IL-1 and TNF-␣ regulate NFAT expression. Stimulation of HepG2 hepatoma cells with IL-1 or TNF-␣ increased NFATc2 expression (Fig. 1A). Accumulation of NFATc2 protein was evident after 8 h of IL-1 or TNF-␣ stimulation, although NFATc2 induction was detected as early as 4 h. However, the expression of other NFAT family members (NFATc1, NFATc3, and NFATc4) was similar in treated and untreated cells. It is interesting that NFATc2 exhibited a modest increase in electrophoretic mobility, FIGURE 1. The inflammatory cytokine IL-1 or TNF-␣ induces NFATc2 expression. A, time course analysis of NFATc2 induction by IL-1 or TNF-␣ stimulation. Extracts prepared from HepG2 cells stimulated with IL-1 or TNF-␣ for various times (4, 8, and 16 h) were subjected to immunoblot analysis. The expression of various NFAT proteins (NFATc1-4) and C/EBP␦ was assessed. The expression of ␤-actin was used as a loading control. B, the induction of NFATc2 requires de novo mRNA transcription and protein synthesis. HepG2 cells were pretreated with the transcription inhibitor actinomycin D (ActD) or the protein synthesis inhibitor cycloheximide (CHX) for 1 h before stimulation with IL-1. The prepared extracts were subjected to immunoblot analysis to determine the expression of NFATc2 and ␤-actin. C, determination of IL-1-mediated NFATc2 induction. Semiquantitative RT-PCR was performed to determine the mRNA levels of NFATc2 and NFATc3 in IL-1-treated and untreated HepG2 cells. The levels of PCR products were determined using ImageQuant software. The expression level of GAPDH was used as a control. D and E, induction of NFATc2 by IL-1 in primary hepatocytes. Semiquantitative RT-PCR was performed to determine the mRNA levels of NFATc2 and GAPDH in primary hepatocytes upon IL-1 stimulation for various times (1, 2, and 4 h) (D). The expression of NFATc2 proteins in IL-1-treated (2, 4, and 12 h) primary hepatocytes was also determined by immunoblot analysis (E). A similar amount of ␤-actin was used as a control.
whereas the electrophoretic mobility of other NFAT family members was similar. As reported previously, the expression of C/EBP␦ was elevated upon IL-1 or TNF-␣ stimulation. The expression level of ␤-actin was used as a control. In addition to HepG2 cells, a similar induction of NFATc2 (but not other NFAT family members) upon IL-1 stimulation was observed in Huh7 hepatoma cells, HEK293 embryonic kidney cells, Caco-2 colorectal carcinoma cells, and human dermal microvascular endothelial cells (data not shown). These data demonstrate that inflammatory cytokines such as IL-1 and TNF-␣ increase NFATc2 expression.
Next, we tested whether the induction of NFATc2 by IL-1 requires de novo mRNA transcription and protein synthesis. Administration of either the transcription inhibitor actinomycin D or the protein synthesis inhibitor cycloheximide blocked the induction of NFATc2 by IL-1 (Fig.  1B). These data demonstrate that IL-1 modulates NFATc2 at least at the mRNA level.
To determine the amount of NFATc2 induction, we performed semiquantitative RT-PCR analysis using various PCR amplification cycles (Fig. 1C). IL-1 stimulation increased NFATc2 expression by ϳ5-fold in HepG2 cells. However, the expression of NFATc3 and GAPDH was similar in the presence and absence of IL-1 stimulation. These data demonstrate that IL-1 selectively increases NFATc2 expression in HepG2 cells.
Next, we confirmed the induction of NFATc2 by IL-1 in primary hepatocytes. Stimulation of primary hepatocytes with IL-1 increased NFATc2 mRNA expression (Fig. 1D). However, the expression of GAPDH was similar in the presence and absence of IL-1 stimulation.
We also confirmed the induction of NFATc2 protein expression in primary hepatocytes by immunoblot analysis (Fig. 1E). Similar to what was observed in HepG2 cells, NFATc2 protein accumulated upon IL-1 treatment in primary hepatocytes. A similar amount of ␤-actin was used as a control. Together, these data demonstrate that inflammatory cytokines such as IL-1 increase NFATc2 expression in primary hepatocytes as well as in hepatoma cell lines.
NFAT Activity Is Not Required for IL-1-induced NFATc2 Expression-An increase in NFAT electrophoretic mobility has been used as a hallmark for NFAT nuclear accumulation and activation (13,26,27). Next, we confirmed that the modest increase in electrophoretic mobility ( Fig. 1) promoted NFATc2 nuclear accumulation. We performed subcellular fractionation and determined the amount of NFATc2 in nuclear extracts upon IL-1 stimulation ( Fig. 2A). Immunoblot analysis demonstrated increased NFATc2 accumulation in the nuclear extracts upon IL-1 stimulation. However, the amount of NFATc3 and YY1 in the nuclear extracts remained similar upon IL-1 stimulation. These data demonstrate that IL-1 stimulation specifically increases and activates NFATc2.
A previous study demonstrated that the expression of NFATc1 is induced upon NFAT activation (28). Autoregulation in the expression of NFAT proteins may be analogous to the induction of C/EBP␤ and C/EBP␦ expression upon C/EBP activation during inflammatory stress. Thus, we examined whether the IL-1-mediated induction of NFATc2 expression requires NFAT activation using the immunosuppressant drug CsA or FK506 analog FK520, which inhibit calcineurin phosphatase and block NFAT-mediated transcription. Administration of CsA or FK520 had a minimal effect on NFATc2 induction upon IL-1 stimulation (Fig. 2B), although IL-1-mediated NFAT activation was CsA-sensitive, as there was a modest decrease in NFATc2 electrophoretic mobility. In addition, luciferase reporter assay demonstrated that IL-1-induced NFAT transcription activity was blocked by CsA (Fig.  2C). These data demonstrate that IL-1 increases the expression of and activates NFATc2. Unlike the induction of NFATc1, that of NFATc2 is not NFAT-dependent.
C/EBP Binds to and Regulates the NFATc2 Promoter-Sequence analysis indicated that the NFATc2 promoter encodes sequences resembling the consensus DNA-binding element for C/EBP (Fig. 3A).
To investigate whether C/EBP is recruited to the NFATc2 loci upon IL-1 stimulation, we performed chromatin immunoprecipitation assays using anti-C/EBP antibodies (Fig. 3A). In resting untreated HepG2 cells, there was minimal C/EBP association with the NFATc2 loci. Stimulation with IL-1 increased the association of C/EBP␣ or C/EBP␤ with the NFATc2 promoter. However, neither C/EBP␣ nor C/EBP␤ was associated with the GAPDH promoter. These data demonstrate that C/EBP associates with the NFATc2 promoter upon IL-1 stimulation in vivo.
To investigate binding of C/EBP to the NFATc2 promoter in vitro, we performed DNA affinity binding assays (Fig. 3B). Using biotinylated oligonucleotides encoding the putative C/EBP-binding site of the NFATc2 promoter, we found both C/EBP␣ and C/EBP␤ in the DNA precipitates. Specific binding of C/EBP was demonstrated by successful competition by the consensus C/EBP sequence, but not by the NF-Bbinding elements. In addition, C/EBP was not present in the precipitates in the absence of oligonucleotides. These data demonstrate that C/EBP binds to the NFATc2 promoter.

FIGURE 2. NFAT activity is not required for IL-1-induced NFATc2 expression. A,
induced NFATc2 is located in the nucleus. HepG2 cells were treated with IL-1 for the indicated times (0, 2, 4, and 8 h). NFATc2 accumulation in a prepared nuclear extract was examined by immunoblot analysis. The presence of NFATc3 and YY1 in an IL-1-treated nuclear extract was also examined. B, IL-1-increased NFATc2 expression is independent of calcineurin. IL-1-mediated induction of NFATc2 was examined by immunoblot analysis. The effect of the calcineurin inhibitor CsA or the FK506 analog FK520 on NFATc2 expression is shown. ns, nonspecific cross-reaction. C, IL-1-increased NFAT activity is calcineurin-dependent. Huh7 cells stably transfected with NFAT-luciferase reporter plasmids encoding triple repeats of either the distal or proximal NFAT site of the PPAR␥2 promoter were treated with IL-1 for 16 h or left untreated. Cell extracts were harvested, and the measured luciferase activity was normalized to protein concentration and is presented (n ϭ 4). The effect of CsA on NFAT activity is also shown.
We further established binding of C/EBP to the NFATc2 promoter by gel mobility shift assays. Similar C/EBP⅐DNA complexes were detected using either a 32 P-radiolabeled NFATc2 C/EBP-binding site or a consensus C/EBP DNA-binding element as a probe (Fig. 3C). Competition assays demonstrated that the C/EBP⅐DNA complex from the NFATc2 C/EBP-binding site was competed by the unlabeled NFATc2 C/EBP-binding site (i.e. self-competition) or the consensus C/EBP-binding element, but not by the mutant C/EBP-binding site. The C/EBP⅐DNA complex from the NFATc2 C/EBP-binding site was also supershifted by anti-C/EBP antibodies. A similar C/EBP⅐DNA complex was also detected using the consensus C/EBP-binding element as a probe. Together, these data demonstrate that C/EBP binds to the NFATc2 promoter.
We also examined the role of C/EBP in NFATc2 promoter activity. Administration of IL-1 increased transcription activity mediated by the NFATc2 promoter (Ϫ1 to Ϫ2000 bp) (Fig. 3D). Similarly, the expression of C/EBP␤ increased NFATc2 promoter activity (Fig. 3E). However, mutation or deletion (Ϫ1 to Ϫ1000 bp) of the C/EBP-binding site reduced transcription activity mediated by the NFATc2 promoter upon IL-1 stimulation or C/EBP␤ coexpression. The expression of C/EBP␤ caused minimal activation of the promoterless pGL3-Basic control. Together, these data demonstrate that IL-1 stimulation promotes C/EBP association with the NFATc2 promoter. Recruitment of C/EBP is likely to play a role in NFATc2 induction by IL-1 stimulation.
C/EBP Activity Is Required for IL-1-mediated NFATc2 Induction-Next, we investigated the requirement for C/EBP in the induction of NFATc2. Among the C/EBP family members, at least C/EBP␣ and C/EBP␤ bind to the NFATc2 promoter (Fig. 3). To inhibit C/EBP-mediated gene transcription, we expressed the transcription repressor C/EBP␥ (also known as Ig/enhancer-binding protein) in cells. C/EBP␥ lacks a transcription activation domain (29), although it contains a similar basic leucine zipper DNA-binding motif and interacts with other C/EBP family members. Thus, C/EBP-mediated transcription activation is abrogated by C/EBP␥. The expression of C/EBP␥ blocked IL-1-mediated NFATc2 promoter activity (Fig. 4A). Notably, the expression of C/EBP␥ reduced IL-1-mediated NFATc2 expression (Fig. 4B). A similar reduction in NFATc2 promoter activity and expression was also found upon expression of a dominant-negative C/EBP inhibitor (A-C/EBP) (data not shown). These data demonstrate that C/EBP regulates NFATc2 promoter activity and gene transcription. Activation of the C/EBP- Chromatin immunoprecipitations were performed to precipitate C/EBP␣-or C/EBP␤-bound promoters. The isolated DNA was amplified by PCR using primers specific for the NFATc2 or GAPDH promoter. Rabbit IgG was used as a control. A comparison of the NFATc2 and consensus C/EBP-binding site sequences is also shown. B, C/EBP binds to the NFATc2 promoter in vitro. Streptavidin-agarose-precleared HepG2 cell extracts were incubated with a double-stranded biotinylated oligonucleotide encoding the NFATc2 C/EBPbinding site. A protein⅐DNA complex was precipitated with streptavidin-agarose, separated by SDS-PAGE, and immunoblotted with anti-C/EBP␣ or anti-C/EBP␤ antibody. Competition was performed using a 5-fold excess of the non-biotinylated DNA-binding oligonucleotide encoding the consensus C/EBP-or NF-B-binding sequence. Binding with streptavidin-agarose alone (No oligo) was used as a control. C, binding of C/EBP to the NFATc2 promoter. IL-1-treated HepG2 nuclear extracts were incubated with 32 P-labeled oligonucleotide encoding the NFATc2 or consensus C/EBP-binding sequence. Formation of the C/EBP⅐DNA complex was examined by gel mobility shift assays. Competition was performed using an excess of the unlabeled wild-type or mutant NFATc2 C/EBP-binding site. Competition with the consensus C/EBP oligonucleotide is also shown. The presence of C/EBP in the protein⅐DNA complex was confirmed by supershift analysis using anti-C/EBP␣ or anti-C/EBP␤ antibody. D, IL-1 stimulation increases NFATc2 promoter activity. Huh7 cells stably transfected with the wild-type or C/EBP mutant NFATc2 promoter-luciferase reporter plasmid were treated with IL-1 for 16 h or left untreated. Cell extracts were harvested, and the measured luciferase activity was normalized to protein concentration and is presented (n ϭ 4). E, expression of C/EBP increases NFATc2 promoter activity. The wild-type or C/EBP mutant NFATc2 promoter-luciferase reporter plasmid was cotransfected with the expression vector for C/EBP␤ in HepG2 cells. Cell extracts were harvested after 36 h of transfection, and the measured luciferase activity was normalized to ␤-galactosidase activity. The relative fold induction of luciferase activity is presented (n ϭ 4). The effect of deletion of the C/EBP-binding site (Ϫ1 to Ϫ1000 bp) on NFATc2 promoter activity was also examined. A promoterless luciferase reporter plasmid (pGL3-Basic) was used as a control.

NFATc2 Regulates sPLA 2 -IIA Expression
NFATc2 transcription factor cascade and subsequent changes in the gene expression profile may be critical in mediating inflammatory stress.
IL-1-induced Transcription Regulation of sPLA 2 -IIA-What are the targets regulated by the C/EBP-NFATc2 transcription factor cascade? We performed microarray analysis and determined the gene expression profile that was regulated by IL-1. Candidate genes with a difference of Ͼ1.5-fold were considered significant and are reported (see supplemental material). The ranges of fold induction and suppression are 1.51-19.47 and 0.70 to 0.05 in the absence of CsA and 1.52-4.38 and 0.70 to 0.04 in the presence of CsA. Microarray analysis indicated that the expression of 245 genes were induced and that of 820 genes was reduced after 12 h of IL-1 stimulation. Upon administration of CsA, 32 induced genes and 312 reduced genes elicited by IL-1 were affected. These data demonstrated that calcineurin/NFAT signaling contributes to IL-1-mediated transcription regulation.
Among the IL-1-induced genes, the expression of sPLA 2 -IIA was reduced upon CsA treatment (see supplemental material). We elected to further investigate sPLA 2 -IIA transcription regulation because sPLA 2 -IIA contributes to arachidonic acid release upon phospholipid hydrolysis (30 -34). Released arachidonic acid plays a critical role in the production of leukotrienes and prostaglandins through the lipoxygenase and cyclooxygenase pathways, respectively. Our focus on sPLA 2 -IIA induction was further catalyzed by the role of NFAT in inducible cyclooxygenase COX2 expression (35)(36)(37)(38)(39), which may constitute an inflammatory cascade in leukotriene and prostaglandin production through transcription regulation of the sPLA 2 -IIA and COX2 genes by NFAT. In addition, sPLA 2 -IIA exhibits antibacterial properties (9) and characteristics of receptor ligands (42). Thus, understanding the regulation of sPLA 2 -IIA expression during inflammatory stress is important.
To examine the role of calcineurin/NFAT in IL-1-induced sPLA 2 -IIA expression, we performed immunoblot analysis using cell extracts prepared from IL-1-stimulated HepG2 cells pretreated or not with the calcineurin inhibitor CsA or FK520. IL-1 stimulation increased sPLA 2 -IIA protein expression (Fig. 5A). Pretreatment with CsA or FK520 abolished IL-1-mediated sPLA 2 -IIA induction. Similarly, administration of the calcineurin inhibitor reduced sPLA 2 -IIA mRNA levels upon IL-1 stimulation (Fig. 5B). IL-1-regulated sPLA 2 -IIA induction at the mRNA level was further confirmed using the transcription inhibitor actinomycin D or the protein synthesis inhibitor cycloheximide. The induction of sPLA 2 -IIA by IL-1 was sensitive to pretreatment with actinomycin D or cycloheximide (Fig. 5C). Together, these data demonstrate that NFATc2 regulates sPLA 2 -IIA expression at least at the mRNA level.
We also ascertained the role of the calcineurin/NFAT signaling pathway in sPLA 2 -IIA induction using HepG2 cells stably expressing the dnNFAT inhibitor (Fig. 6A). We have demonstrated previously that dnNFAT blocks calcineurin targeting (40). Therefore, NFAT nuclear accumulation and subsequent transcription are abolished. Stimulation with IL-1 or TNF-␣ increased sPLA 2 -IIA expression in parental control HepG2 cells (Fig. 6, B and C). However, the expression of sPLA 2 -IIA FIGURE 4. C/EBP activity is required for IL-1-mediated NFATc2 induction. NFATc2 promoter activity (A) and NFATc2 protein induction (B) upon IL-1 stimulation were examined by coexpression with C/EBP␥ in HEK293 cells. Cell extracts were harvested after 36 h of transfection, and the measured luciferase activity was normalized to ␤-galactosidase activity. The prepared extracts were also subjected to immunoblot analysis to detect NFATc2 and ␤-actin expression.

NFATc2 Regulates sPLA 2 -IIA Expression
protein and mRNA was attenuated in dnNFAT-expressing cells. Together, these data demonstrate that the calcineurin/NFAT signaling pathway contributes to sPLA 2 -IIA induction.
Temporal Requirement for the Calcineurin/NFAT Signaling Pathway in IL-1-mediated sPLA 2 -IIA Induction-Next, we examined the kinetic profile of sPLA 2 -IIA induction. Administration of IL-1 increased sPLA 2 -IIA expression and subsequent secretion into the media (Fig.  5D). Notably, a marked increase in sPLA 2 -IIA expression was observed in the later phase (after 12 h) of IL-1 stimulation, although induction was detected as early as 6 h after stimulation. To delineate the temporal requirement for the calcineurin/NFAT pathway in IL-1-mediated sPLA 2 -IIA induction, we administered the calcineurin inhibitor CsA at various times during IL-1 stimulation (Fig. 7A). Pretreatment with CsA (i.e. administration of CsA 1 h before stimulation with IL-1 (i.e. Ϫ1 h)) reduced sPLA 2 -IIA protein (Fig. 7B) and mRNA (Fig. 7C) expression after 24 h of IL-1 stimulation. Secretion of sPLA 2 -IIA into the media was also reduced (Fig. 7B). Administration of CsA after 8 or 12 h of IL-1 stimulation reduced sPLA 2 -IIA expression to a lesser extent. Administration of CsA in the last 8 h (i.e. 16 h after the initial IL-1 stimulation), before determining sPLA 2 -IIA expression and secretion after 24 h of IL-1 stimulation, exhibited a minimal inhibitory effect. These data demonstrate that the calcineurin/NFAT pathway is temporally required for the expression of sPLA 2 -IIA upon inflammatory stimulation.
NFAT Binds to and Regulates the sPLA 2 -IIA Promoter-What is the molecular basis of NFAT in sPLA 2 -IIA expression? Sequence analysis indicated that there are three putative NFAT-binding sites in the sPLA 2 -IIA promoter. These three putative NFAT-binding sites are located at Ϫ200 (proximal), Ϫ250 (middle), and Ϫ500 (distal) bp from the sPLA 2 -IIA transcription start region (Fig. 8A). To examine whether NFAT binds to these sites, we performed gel mobility assays using nuclear extracts prepared from NFATc2-expressing COS cells (Fig. 8B) or IL-1-treated HepG2 cells (Fig. 8C). Distinct NFAT⅐DNA complexes were detected in all three sPLA 2 -IIA NFAT-binding sites. The specificity of the NFAT⅐DNA complexes was determined by competition analysis using consensus NFAT-binding elements from the IL-2 and PPAR␥2 promoters. Supershift analysis further demonstrated the presence of NFAT in the protein⅐DNA complexes. In addition, we performed chromatin immunoprecipitation assays to demonstrate binding of NFATc2 and acetylated histone H3 (but not NFATc3) to the sPLA 2 -IIA promoter (Fig. 8D). Neither NFATc2 nor NFATc3 was present in

NFATc2 Regulates sPLA 2 -IIA Expression
the GAPDH promoter. These data demonstrate that NFAT binds to the sPLA 2 -IIA promoter.
Binding of NFATc2 may regulate sPLA 2 -IIA gene transcription. Using luciferase reporter assays, we tested whether NFAT regulates the sPLA 2 -IIA gene promoter. The expression of NFATc2 increased sPLA 2 -IIA promoter activity (Fig. 8E). Deletion of NFAT-binding sites to Ϫ150 bp abrogated increases in sPLA 2 -IIA promoter activity. Together, these data demonstrate that NFAT binds to and regulates the sPLA 2 -IIA gene promoter.
Induction of NFATc2 Is Required for IL-1-increased sPLA 2 -IIA Expression-Because NFAT binds to and regulates sPLA 2 -IIA expression (Fig. 8), we hypothesized that the induction of NFATc2 by IL-1 is required for sPLA 2 -IIA expression. One approach to study sPLA 2 -IIA expression in the absence of NFATc2 is to utilize NFATc2 null mice, especially because NFATc2 null mice in a C57BL/6 background seem to exhibit hypersensitivity upon lipopolysaccharide stimulation. 3 However, sPLA 2 -IIA is not expressed in mouse liver (41,42). It is expressed only in mouse small intestine. Indeed, sPLA 2 -IIA is expressed only in the small intestine in a certain mouse background (43). Thus, the available NFATc2 null mouse (mainly in mouse strains C57BL/6 and 129/SvJ, which are defective in sPLA 2 -IIA expression) is not appropriate for analyzing the induction of sPLA 2 -IIA upon inflammatory cytokine challenge.
To circumvent the lack of sPLA 2 -IIA expression in mouse liver and to address the role of NFATc2 in sPLA 2 -IIA expression, we exploited RNA interference technology using shRNA targeting NFATc2 in HepG2 cells (Fig. 9A). Cells stably transfected with shRNA targeting NFATc2 exhibited reduced expression of NFATc2 compared with the parental control HepG2 cells. The expression of NFATc3 and NFATc4 was similar, however. Notably, targeted reduction of NFATc2 abolished IL-1-induced NFATc2 expression. These data demonstrate that shRNA specifically reduces NFATc2 expression in resting and IL-1-stimulated HepG2 cells.
Next, we examined the expression of sPLA 2 -IIA in NFATc2 knockdown and parental control HepG2 cells (Fig. 9, B-E). Stimulation with IL-1 (Fig. 9B) or TNF-␣ (Fig. 9C)   . NFAT binds to and regulates the sPLA 2 IIA promoter. A, sequence comparison of NFAT-binding sites (proximal, middle, and distal) in the sPLA 2 -IIA promoter with consensus NFAT-binding sites in the IL-2 and PPAR␥2 promoters. The consensus binding sequences for NFAT partners (AP-1 and C/EBP) are also indicated. B and C, NFAT binds to the sPLA 2 -IIA promoter in vitro. Nuclear extracts prepared from COS cells expressing NFATc2 (B) or IL-1-induced HepG2 cells (C) were incubated with 32 P-labeled sPLA 2 -IIA NFAT-binding oligonucleotides (proximal, middle, and distal). Formation of the NFAT⅐DNA complex was examined by gel mobility shift assays and visualized by autoradiography. The specificity of the NFAT⅐DNA complex was examined by competition with consensus NFAT-binding elements in the IL-2 or PPAR␥2 promoter. The presence of NFATc2 in the protein⅐DNA complex was demonstrated by antibody supershift analysis (anti-NFATc2 antibody; see asterisks). D, association of NFATc2 with the sPLA 2 -IIA promoter. Nuclear proteins associated with chromatin in HepG2 cells were cross-linked with 1% formaldehyde. Chromatin was sheared and immunoprecipitated with anti-NFATc2, anti-NFATc3, or anti-acetylated histone H3 (Ac-H3) antibody. DNA in the NFATc2, NFATc3, or acetylated histone 3 precipitate was purified, and the presence of the sPLA 2 -IIA and GAPDH promoters was determined by PCR. Rabbit IgG was used as a control. E, NFAT positively regulates sPLA 2 -IIA promoter activity. The sPLA 2 -IIA promoter (Ϫ1 to Ϫ1200 bp) was cotransfected with the NFATc2 expression vector (gray bars) or empty vector (white bars) in HepG2 cells. Cell extracts were harvested after 36 h of transfection, and the measured luciferase activity was normalized to ␤-galactosidase activity and is presented (n ϭ 4). The effect of deletion of NFAT-binding elements (Ϫ1 to Ϫ150 bp) on sPLA 2 -IIA promoter activity was also examined. A promoterless luciferase reporter plasmid (pGL3-Basic) was used as a control.

NFATc2 Regulates sPLA 2 -IIA Expression
in parental control HepG2 cells. The expression of sPLA 2 -IIA in NFATc2 knockdown cells was abolished, however. Targeted reduction of NFATc2 expression also decreased sPLA 2 -IIA mRNA expression (Fig. 9D). Notably, kinetic analysis demonstrated that the induction and subsequent secre-tion of sPLA 2 -IIA were abolished in NFATc2 knockdown cells (Fig. 9E). Together, these data demonstrate that the induction of NFATc2 by IL-1 is required for sPLA 2 -IIA expression.
Induction of NFATc2 Is Required for IL-1-increased PLD 1 and S100A3 Expression-Our microarray analysis revealed that, in addition to sPLA 2 -IIA, the IL-1-mediated induction of PLD 1 and the calcium-binding protein S100A3 was sensitive to CsA inhibition (see supplemental material). In conjunction with sPLA 2 -IIA, PLD 1 hydrolyzes phospholipids (e.g. phosphatidylcholine) to lysophosphatidic acid (LPA), arachidonic acid, and choline (44,45). LPA is a potent G-protein-coupled receptor agonist that may further modulate inflammatory signals (46,47). In association with annexins, S100A3 may modulate membrane dynamics and intracellular trafficking (e.g. exocytosis, endocytosis, and membrane architecture and remodeling) (48). Unlike the role of sPLA 2 -IIA, that of PLD 1 and S100A3 during inflammatory stress has yet to be demonstrated.
Sequence analysis indicated that there are nine and five putative NFAT-binding sites (GGAAA) in the PLD 1 and S100A3 promoters, respectively. In the PLD 1 gene, these NFAT-binding sites are located at Ϫ400, Ϫ420, Ϫ600, Ϫ1000, Ϫ1200, Ϫ1300, Ϫ2000, Ϫ2300, and Ϫ2350 bp upstream of the promoter. In the S100A3 gene, these NFAT-binding sites are located at Ϫ170, Ϫ200, Ϫ350, Ϫ450, and Ϫ680 bp upstream of the promoter. These observations further support that PLD 1 and S100A3 could be NFAT targets.
Next, we investigated the role of NFATc2 in IL-1-induced PLD 1 and S100A3 expression. RT-PCR analysis demonstrated that IL-1 increased PLD 1 and S100A3 expression (Fig. 10A). However, the induction of PLD 1 and S100A3 was abrogated by CsA. Targeted reduction of  . Induction of NFATc2 is required for IL-1-increased PLD 1 and S100A3 expression. A, calcineurin inhibitor cyclosporin A (CsA) abolishes IL-1-mediated PLD 1 and S100A3 induction. HepG2 cells were pretreated (ϩ) or not (Ϫ) with the calcineurin inhibitor CsA for 1 h before stimulation with IL-1 for 12 h (ϩ). The expression levels of PLD 1 and S100A3 mRNAs were determined by semiquantitative RT-PCR. The expression level of GAPDH was used as a control. B, targeted reduction of NFATc2 abolishes IL-1mediated PLD 1 and S100A3 induction. NFATc2 knockdown (shRNA NFATc2) and parental control HepG2 cells were stimulated (ϩ) or not (Ϫ) with IL-1. The expression levels of PLD 1 and S100A3 mRNAs were determined by semiquantitative RT-PCR. The expression level of GAPDH was used as a control.

DISCUSSION
Role of NFAT in the Inflammatory Response in Nonimmune Cells-In this study, we have demonstrated that inflammatory cytokines induce NFATc2 expression in hepatocytes. C/EBP-mediated NFATc2 induction plays a critical role in the expression of sPLA 2 -IIA, an important phospholipase in the regulation of arachidonic acid release and the subsequent leukotriene and/or prostaglandin pathway. These results provide a mechanism for NFATc2 expression and expand the repertoire of NFAT function to the acute-phase response in the liver. Previously, we demonstrated that IL-6 increases NFATc2 expression, which then promotes IL-4 expression and subsequent Th2 cell differentiation (49,50). Together, these results demonstrate that the transcription factor NFAT contributes to the restoration of cellular homeostasis in both nonimmune and immune cells upon the induction of inflammatory stress by the invasion of foreign entities.
Our microarray analysis demonstrated that 32 genes induced by IL-1 stimulation are sensitive to CsA administration. Among the CsA-sensitive IL-1-induced genes, we have confirmed the regulation of sPLA 2 -IIA and PLD 1 by NFATc2. In addition, the calcium-binding protein S100A3 is also regulated by NFATc2. These data indicate that calcineurin/NFAT signaling can contribute to feedback regulation of phospholipid hydrolysis and calcium mobilization upon IL-1 challenge.
Our microarray analysis also demonstrated that transcription repression mediated by IL-1 is sensitive to CsA. The expression of 312 genes suppressed by IL-1 stimulation is sensitive to CsA. These genes include transcription factors, signaling mediators, and secretory proteins. Further characterization of these targets is warranted to provide a comprehensive analysis of the role of calcineurin/NFAT during inflammatory stress in hepatocytes.
In addition to hepatocytes, adipocytes have also been implicated to play a role in the acute-phase response (51,52). How adipocytes contribute to the acute-phase response remains to be determined. Adipocytes may modulate glucose and lipid metabolism in response to invasion by foreign entities. Induction and subsequent secretion of adipose-specific peptides/hormones (adipokines) may also participate in part in antibacterial/antifungal function and macrophage recruitment. Indeed, analogous to the role of NFAT in cytokine gene transcription in immune cells, NFAT also regulates adipokine gene expression in adipocytes. 3 Hence, adipocytes may provide an additional means to enable necessary metabolic changes during inflammatory stress.
It is intriguing that the transcription factor C/EBP plays a critical role in gene transcription in both hepatocytes and adipocytes (4,53,54). C/EBP-mediated NFATc2 induction in hepatocytes may be extrapolated to adipocytes during inflammatory stress. Similar NFATc2 induction may also be required for the role of NFAT in adipocytes (13,14,18,55), especially because a C/EBP transcription factor cascade is important for the initiation of adipocyte differentiation (56 -61). In addition, the extent of adiposity and insulin resistance correlates with the degree of inflammation (62)(63)(64). Thus, the C/EBP-NFATc2 transcription factor cascade may provide a mechanism for gene transcription regulation in hepatocytes and adipocytes.
In addition to HepG2 hepatoma cells, Huh7 hepatoma cells, HEK293 embryonic kidney cells, Caco-2 colorectal carcinoma cells, and human dermal microvascular endothelial cells also exhibited NFATc2 induction upon IL-1 stimulation (data not shown). The induction of NFATc2 may be a conserved mechanism in nonimmune cells to combat invasion by foreign entities in response to IL-1 stimulation. It is interesting that components of the IL-1 signaling pathway are highly conserved in multiple species (65,66), whereas NFAT is expressed only in vertebrates (67). Integration of these two apparent evolutionarily independent pathways may provide diverse responses to restore cellular homeostasis in vertebrates.
NFAT Versus NF-B in the Inflammatory Response-Previous studies demonstrated that IL-1 activates NF-B and mediates immediate responses during inflammatory stress (68 -71). The induction of target genes such as IB␣ and MCP1 is mediated by NF-B within 1 h of stimulation with inflammatory cytokines (70,72,73). Activation of NF-B is subsequently down-regulated by the newly synthesized IB. Here, we have demonstrated that NFAT temporally modulates sPLA 2 -IIA expression upon stimulation with inflammatory cytokines. These data suggest that NF-B and NFAT may mediate distinct responses upon inflammatory cytokine stimulation.
It is interesting that both NFAT and NF-B are Rel domain-containing transcription factors (67,74). NFAT interacts with NFAT partners (e.g. Fos⅐Jun⅐AP-1 complex, C/EBP, and GATA proteins), whereas NF-B binds as a homo-or heterodimer to mediate gene transcription. The distinct regulation mediated by NFAT and NF-B in gene transcription may account for the temporally discrete redundant role in response to inflammatory cytokine stimulation.
With respect to NFAT partners, the striking differences in various NFAT⅐DNA complexes formed in the sPLA 2 -IIA promoter (Fig. 8) imply that NFAT cooperates with distinct nuclear factors to mediate gene transcription during inflammatory stress. These nuclear factors may include C/EBP, STAT, and Forkhead members (4,5,75,76). Dimerization of NFAT may also contribute to transcription cooperation (77,78). The presence of distinct NFAT⅐DNA complexes is also found in the PPAR␥2 gene promoter upon adipocyte differentiation (13,14). Because dissociation of NFAT⅐DNA complexes is regulated by NFAT partners (14,55), interaction with unique nuclear factor is likely to modulate the extent of NFAT activation to provide a dose-dependent regulation of inflammation and adipocyte differentiation.
The dose-dependent regulation of NFAT activation may also be mediated by the co-localization of multiple but distinct NFAT⅐DNA complexes in the sPLA 2 -IIA and PPAR␥2 gene promoters upon transcription. Hence, a graded response, in part due to interaction with various NFAT partners and/or the number of recruited NFAT⅐DNA complexes in the active transcription loci, may arise and contribute to a threshold regulation of NFAT target genes. A goal for future research will be to identify NFAT targets regulated by the threshold mechanism.
sPLA 2 -IIA Induction and Disease-In this study, we have demonstrated that induction of NFATc2 regulates sPLA 2 -IIA expression. Two possible mechanisms may account for the induction of sPLA 2 -IIA by NFATc2. First, NFATc2 may selectively activate sPLA 2 -IIA. In this mechanism, NFAT DNA-responsive elements in the sPLA 2 -IIA gene may selectively bind NFATc2, but not other NFAT family members. Further characterization of these sPLA 2 -IIA NFAT elements may shed new light on the molecular basis of NFATc2-specific target genes. Alternatively, newly synthesized NFATc2 may be selectively activated and account for the sPLA 2 -IIA induction. This may be due in part to differential regulation by upstream NFAT kinases and/or phosphatases to modulate NFAT subcellular localization and DNA binding, transactivation, and/or degradation.
Second, accumulation of NFAT proteins per se may account for the induction of sPLA 2 -IIA. Hence, the purpose of increased NFATc2 expression is to provide sufficient proteins to be activated by and to interact with the NFAT-binding elements (proximal, middle, and distal) to mediate sPLA 2 -IIA gene transcription. This mechanism supports a possible functional redundancy among different members of the NFAT family. In addition, this mechanism implies a threshold requirement for the NFAT function. Indeed, a threshold requirement for NFAT-mediated gene transcription has been demonstrated upon targeted removal of the endogenous calcineurin inhibitor calcipressin-1 (79), also known as modulatory calcineurin-interacting protein or Down syndrome critical region protein (80 -83). Nonetheless, these two possible mechanisms (NFATc2-specific versus NFAT accumulation) for sPLA 2 -IIA expression upon NFATc2 induction are not mutually exclusive. However, both of these mechanisms require activation of the induced NFATc2 for sPLA 2 -IIA expression (Figs. 2 and 5). Increased NFATc2 expression without activation does not seem to be sufficient to mediate sPLA 2 -IIA induction. Future research will elucidate the molecular mechanism of NFATc2 regulation in different targets. sPLA 2 catalyzes hydrolysis of phospholipids at the sn-2 position to release arachidonic acid and lysophospholipids (30,31). Released arachidonic acid is metabolized to leukotrienes and prostaglandins by the lipoxygenase and cyclooxygenase pathways, respectively. Previous studies have demonstrated that the expression of the inducible form of cyclooxygenase (COX2) is regulated by NFAT (35)(36)(37)(38)(39). Thus, the transcription factor NFAT impinges on leukotriene and prostaglandin production through transcription regulation of the sPLA 2 -IIA and COX2 genes.
In addition to leukotriene and prostaglandin production, sPLA 2 -IIA may also participate in LPA generation. LPA is a potent G-protein-coupled receptor agonist that modulates mitogenic signaling through the endothelial differentiation gene receptor. Lysophospholipids released by sPLA 2 -IIA-mediated hydrolysis could be further converted to LPA by PLD 1 (46,47), the expression of which is also NFATc2-dependent (Fig. 10). The potent mitogenic signals from LPA may then play an important role in the recovery of the damaged cells during inflammatory stress. Together, the induction of sPLA 2 -IIA, PLD 1 , and COX2 by NFAT suggests that phospholipid hydrolysis could further amplify/antagonize the initial inflammatory signals.
Previous studies have demonstrated that the level of sPLA 2 -IIA expression/secretion correlates with various acute and chronic inflammatory conditions, including sepsis, asthma, rheumatoid arthritis, and atherosclerosis (84 -87). Indeed, increased levels of sPLA 2 -IIA have been demonstrated in the sera and synovial fluids of arthritis patients. In addition, accumulating evidence suggests that chronic inflammation promotes tumorigenesis (88,89), as the expression/secretion of sPLA 2 -IIA in many different types of cancer is elevated (90 -92). Together, these data indicate that therapeutic strategies to down-regulate sPLA 2 expression/activity are critical in decreasing the subsequent production of bioactive lipid mediators (e.g. leukotrienes, prostaglandins, and LPA). The reduced production of bioactive lipid mediators may then alleviate the inflammatory response. Because calcineurin/ NFAT regulates sPLA 2 -IIA, PLD 1 , and COX2 expression, calcineurin inhibitors (e.g. CsA and tacrolimus FK506) may be applicable to both acute and chronic inflammation stresses (93,94), in addition to their initial discovery as immunosuppressants.
In conclusion, we have demonstrated that inflammatory cytokines induce binding of C/EBP to the NFATc2 loci. C/EBP-mediated NFATc2 induction is temporally required for the expression of sPLA 2 -IIA in hepatocytes. These results provide a mechanism for NFATc2 expression and expand the repertoire of NFAT function to the acute-phase response in the liver.