Evidence for Common Mechanisms in the Transcriptional Control of Type II Nitric Oxide Synthase in Isolated Hepatocytes REQUIREMENT OF NF- (cid:107) B ACTIVATION AFTER STIMULATION WITH BACTERIAL CELL WALL PRODUCTS AND PHORBOL ESTERS*

Incubation of primary cultures of rat hepatocytes with lipopolysaccharide (LPS), S -[2,3-bis(palmitoyloxy)-(2- R,S )-propyl]- N -palmitoyl-( R )-Cys-Ser-Lys 4 (TPP), a synthetic lipopeptide present in bacterial cell wall li-poproteins, or with phorbol 12,13-dibutyrate (PDBu) in- duced an increase in nitric oxide synthesis through the expression of type II nitric oxide synthase (iNOS). Transfection of hepatocytes with a Hin dII fragment corresponding to the promoter region of the murine iNOS gene (from nucleotide (cid:50) 1588 to (cid:49) 165) resulted in the expression of the reporter gene when cells were stimu- lated with these factors. The transcription factors activated by these stimuli involved an increase in the nu- clear content of proteins that bind to (cid:107) B, AP-1, GAS, and SIE sequences. Inhibition of NF- (cid:107) B activation with pyrrolidine dithiocarbamate eliminated the expression of iNOS in hepatocytes stimulated with LPS, TPP, or PDBu. In addition to this, transfection of hepatocytes with promoter mutants in which a sequential 2-base pair change within the (cid:107) B sites was introduced (position (cid:50) 971 to (cid:50) following the ECL technique dimers in the nuclear extracts was determined by EMSA using the proximal NF- (cid:107) B oligonucleotide as target. The nature of the proteins present in the retarded bands was confirmed by supershift analysis with anti-p50 and anti-p65 antibodies (human origin). Quantitation of the amount of p50/p65 complexes was done by densitometric analysis of the corresponding unsaturated films of EMSA. Results show the of three

The involvement of nitric oxide in different pathophysiological pathways is a subject of current research (1,2). Three different nitric oxide synthase species have been identified in mammalian tissues, each exhibiting an important degree of tissue-specific expression as well as significant differences in their regulatory properties (1)(2)(3). The species recognized as type II nitric oxide synthase (iNOS) 1 is induced in a number of different cell types in response to cytokines involved in inflammation and host defense as well as by bacterial cell wall products and some pharmacological agents (3)(4)(5)(6). The promoter region of iNOS has been characterized in different species, including humans and mice (7,8). A 1753-base pair fragment of the promoter region of iNOS has been cloned and characterized from the murine RAW 246.7 macrophage cell line (8 -10). Sequence analysis of this promoter revealed the presence of at least 24 consensus motifs for binding of transcription factors, including 2 copies for nuclear factor B (NF-B), 2 copies for activator protein-1 (AP-1), 10 copies of IFN-␥ response elements (␥-IRE), 3 copies of the ␥-activated site (GAS), 2 copies of the IFN-stimulated response element (ISRE), and 2 copies of the tumor necrosis factor-␣ responsive element, among others (8 -11). Functional analysis in RAW 264.7 cells using deletional mutants of the iNOS promoter revealed the presence of two important regulatory regions, each one containing a B binding site for NF-B and exhibiting an important interaction of their roles in activation of iNOS transcription (8,9). These regions cover 200 base pairs upstream of the start site of transcription and positions Ϫ913 to Ϫ1020 and have been referred to as the proximal and distal regulatory regions, respectively (8,9). In addition to the cooperation between these two regions, a concerted synergism among nuclear factor binding sites exists in each region. This is the case for macrophages, where the interaction elicited by suboptimal concentrations of LPS and IFN-␥ on NO synthesis has been demonstrated to be the result of a cooperative interaction between the IFN regulatory factor 1 and NF-B sites in the distal region (12,13).
Apart from macrophages, iNOS is induced in a variety of other cell types such as neural cells, keratinocytes, myocytes, mesangial cells, tumor cells, and hepatocytes in response to a wide array of either physiological or pathological cellular stresses (1). In the case of the liver, both hepatocytes and Kupffer cells express iNOS in response to different cell stimuli such as septic shock, cirrhosis, hyperdynamic circulation, or after partial hepatectomy (14 -16). Moreover, primary cultures of hepatocytes retain their ability to express iNOS after exposure to LPS, TPP, or after treatment with phorbol esters that are pharmacological activators of protein kinase C (5,6). All of these molecules induce by themselves the expression of a functional iNOS as deduced by the large amounts of NO released to the medium (5, 6, 14 -16). Hepatocytes also constitute an interesting experimental model because these cells have an extremely high transcriptional rate in combination with a very low proliferative capacity, and, therefore, most of the transcriptional factors required for commitment to cell growth are switched off (17). In this regard, we have investigated the factors involved in the transcriptional activation of hepatic iNOS in response to three defined effectors that, because of its unrelated chemical structure, act on the hepatocyte through different transduction pathways; LPS acts through the engagement of CD14 (18), whereas the TPP receptor remains elusive (6,19). Phorbol esters are potent activators of protein kinase C subspecies containing a tandem of two zinc-finger domains and whose expression pattern varies between tissues (20). The ability of cells to express iNOS in response to phorbol esters appears to be restricted to some cell types, including rat hepatocytes and peritoneal macrophages (5,21). For this reason we investigated the pattern of transcriptional factors relevant to iNOS expression in hepatocytes incubated with these three stimuli.
Our results show that treatment of primary cultures of hepatocytes with LPS, TPP, and phorbol esters trigger a similar pattern of transcription factor activation, including members of the NF-B, AP-1, GAS, and SIE binding proteins. This transactivation was also evident in hepatocytes transfected with plasmids encoding the murine iNOS promoter or consensus sequences for the binding of NF-B and AP-1 linked to a CAT reporter gene. These results suggest that the expression of iNOS in hepatocytes in response to LPS, TPP, and phorbol esters is mediated through the engagement of a similarly regulated transcriptional mechanism.

MATERIALS AND METHODS
Reagents-Cytokines and biochemicals were from Sigma. Materials and chemicals for electrophoresis were from Bio-Rad. S-[2,3-bis(palmitoyloxy)-(2-R,S)-propyl]-N-palmitoyl-(R)-Cys-Ser-Lys 4 (TPP) was from Boehringer Mannheim. LPS was from Salmonella typhimurium. Plasmids were purified by extensive washing using Qiagen columns (Hilden, FRG). The endotoxin content in cytokines and plasmid preparations was determined using the Limulus polyphemus test (Sigma) and was below 0.1 ng/mg protein and 30 pg/ml plasmid dilution, respectively, at the dose used for transfection. Serum and media were from BioWhittaker (Walkersville, MD).
Isolation and Culture of Hepatocytes-Isolation of hepatocytes was carried out from 3-month-old male rats by perfusion with collagenase in Krebs-bicarbonate buffer under continuous gassing with carbogen (O 2 / CO 2 , 19:1) and following the classic recirculation protocol (5). The hepatocyte suspension was washed twice with sterile Dulbecco's modified Eagle's medium and then resuspended in this medium supplemented with 50 g/ml gentamicin, 50 g/ml penicillin G, and 50 g/ml streptomycin (incubation medium). The hepatocytes (3 ϫ 10 6 ) were plated in 6-cm tissue culture dishes in a medium containing 2.5 ml of Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum. After 4 h of incubation to facilitate cell attachment to the matrix, the medium was aspirated, and the plates were washed twice with PBS to remove the nonadherent cells and filled with 2.5 ml of Dulbecco's modified Eagle's medium lacking serum. Additions were made so that the changes in the total incubation volume were less than 2%.
Transfection of Hepatocytes and CAT Assays-Freshly isolated hepatocytes were transfected by electroporation (22). Cells (4 ϫ 10 6 in 0.8 ml of PBS) were kept at 4°C and electroporated at 300 V, 960 F in a BTX electroporator (Electrocell Manipulator 600) in the presence of 10 g of plasmid. Electroporated cells were maintained for 10 min at 4°C and then transferred to 6-cm dishes filled with 3 ml of Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum and antibiotics at 37°C. After 6 -8 h of culture, the medium was aspirated and the dishes washed twice with PBS to remove nonadhered cells. The hepatocytes were maintained overnight with incubation medium supplemented with 2% of heat-inactivated fetal calf serum. Twenty-four hours after electroporation the medium was replaced by fresh medium and the cells were stimulated with different factors. To determine CAT activity, the cells were incubated with the stimuli for 36 h. After two washes of the plates with ice-cold PBS, the cells were scraped off the dishes. After centrifugation, the cell pellets were resuspended in 100 l of 100 mM potassium phosphate (pH 7.8) and 1 mM dithiothreitol and disrupted by three cycles of freezing and thawing followed by centrifugation at 12,000 ϫ g for 10 min. The supernatant was heated at 65°C for 10 min, and CAT activity was measured by synthesis of acetylated [ 14 C]chloramphenicol following the TLC method (22,23). CAT activity was expressed as the percentage of the activity measured in cells electroporated with a kSV 2 CAT plasmid (23). As an additional control to ensure that the plasmids were free of bacterial products involved in NO synthesis, electroporated cells in the absence of plasmids received an equivalent dose of plasmid upon culture.
Plasmids-The 1753-base pair HindII fragment corresponding to the 5Ј-flanking region of iNOS fused to a promoterless CAT reported gene (p1iNOS-CAT) (8) was a generous gift from Dr. Q.-w. Xie and Dr. C. Nathan. Plasmids (B) 3 ConACAT, (AP-1) 2 ConACAT, and ConACAT have 3 copies of the B motif from the human immunodeficiency virus long terminal repeat enhancer, two copies of the AP-1 motif from the collagenase promoter or the minimal promoter with no enhancer element of the conalbumin A promoter (used as a control), respectively, and have been previously described (24). Mutated B promoter plasmids were generated by polymerase chain reaction using oligonucleotide primers in which 2 GG bases of the B motif were substituted by a CC pair, kindly given by Dr. T. J. Evans (25). A scheme of these promoter mutants is shown in Fig. 9. Deletions of the p1iNOS vector were generated by digestion with EspI, SmaI, and SacI, respectively, which generated the plasmids p1NOS(⌬1) (nucleotides Ϫ1029 to ϩ165), p1NOS(⌬2) (nucleotides Ϫ725 to ϩ165), and p1NOS(⌬3) (nucleotides Ϫ333 to ϩ165), respectively (see Fig. 9). The vectors were sequenced to ascertain their fidelity. A kSV 2 CAT plasmid in which the CAT gene is driven by the simian virus 40 early promoter and enhancer as a control in transfection assays (22).
Determination of NO-NO release was determined spectrophotometrically by the accumulation of nitrite and nitrate in the medium (phenol red-free) as follows: 250 l of culture medium were transferred to 1.5-ml Eppendorf tubes and the nitrate was reduced to nitrite with 0.5 units of nitrate reductase (Boehringer Mannheim) in the presence of 50 M NADPH and 5 M FAD (6,16). After oxidation of the NADPH excess (which interferes with the chemical determination of nitrite) with 0.2 mM pyruvate and 1 g of lactate dehydrogenase, nitrite was determined with Griess reagent (16) by adding 1 mM sulfanilic acid and 100 mM HCl (final concentration). After incubation for 5 min the tubes were centrifuged, and 200 l of supernatant were transferred to a 96-well microplate. After a first reading of the absorbance at 595 nm, 50 l of naphthylenediamine (1 mM in the assay) were added. The reaction was completed after 15 min of incubation, and the absorbance at 595 nm was compared with a standard of NaNO 2 .
RNA Extraction and Analysis-Total RNA (3-4 ϫ 10 6 cells) was extracted following the guanidinium thiocyanate method (26). After electrophoresis in a 0.9% agarose gel containing 2% formaldehyde the RNA was transferred to a Nytran membrane (NY 13-N; Schleicher & Schuell) with 10 ϫ SSC (1.5 M NaCl, 0.3 M sodium citrate, pH 7.4). The membrane was prehybridized and the levels of iNOS or IB-␣ mRNA were determined using EcoRI-HindII or HindII-NotI fragments from the iNOS and IB-␣ cDNA, respectively (24,27), labeled with [␣-32 P]dCTP using the random primed labeling kit (Boehringer Mannheim). The membranes were washed with 0.1 ϫ SSC and 0.1% SDS at room temperature for 10 min and twice at 50°C for 30 min, followed by exposure to x-ray film (Hyperfilm, Amersham Corp.). Different exposition times of the x-ray films were used to ensure that bands were not saturated. Quantification of the films was performed by laser densitometry (Molecular Dynamics) using the hybridization with a ribosomal 18 S probe as an internal standard.
Preparation of Cytosolic and Nuclear Extracts-A modified procedure based on the method of Schreiber et al. was used (28). The cell layers (3 ϫ 10 6 ) were washed twice with ice-cold PBS, and the hepatocytes were collected in PBS by centrifugation at 50 ϫ g for 5 min. The cell pellets were homogenized with 0.4 ml of buffer A (10 mM Hepes, pH 7.9, 1 mM EDTA, 1 mM EGTA, 10 mM KCl, 1 mM dithiothreitol, 0.5 mM phenylmethanesulfonyl fluoride, 2 g/ml aprotinin, 10 g/ml leupeptin, 2 g/ml 1-chloro-3-tosylamido-7-amino-2-heptanone, 5 mM NaF, 1 mM NaVO 4 , 10 mM Na 2 MO 4 ). After 10 min at 4°C Nonidet P-40 was added to reach a 0.5% concentration. The tubes were gently vortexed for 15 s, and nuclei were sedimented by centrifugation at 8000 ϫ g for 15 s. The supernatant was aliquoted and stored at Ϫ80°C (cytosolic extract), and their pellet was resuspended in 100 l of buffer A supplemented with 20% glycerol and 0.4 M KCl. Incubation was continued for 30 min at 4°C with gentle vortexing. Nuclear proteins were extracted by centrifugation at 13,000 ϫ g for 15 min and aliquots of the supernatant were stored at Ϫ80°C. Proteins were measured using the Bio-Rad protein reagent following the recommendations of the supplier. All steps of cell fractionation were carried out at 4°C.
The oligonucleotides were annealed after incubation for 5 min at 85°C in 10 mM Tris-HCl, pH 8.0, 50 mM NaCl, 10 mM MgCl 2 , and 1 mM dithiothreitol. Aliquots of 50 ng of these annealed oligonucleotides were end-labeled with the Klenow enzyme fragment in the presence of 50 Ci of [␣-32 P]dCTP and the other unlabeled dNTPs in a final volume of 50 l. The oligonucleotides were precipitated in ethanol, extracted with phenol/chloroform to remove the unincorporated nucleotides, and 5 ϫ 10 4 dpm of the DNA probe were used for each binding assay of nuclear extracts as follows: 5 g of protein extract were incubated for 30 min at 4°C with the DNA and 1 g of poly(dI⅐dC)/ml, 5% glycerol, 1 mM EDTA, 100 mM KCl, 5 mM MgCl 2 , 1 mM dithiothreitol, 10 mM Tris-HCl, pH 7.8, in a final volume of 20 l. The incubation mixture was applied to a 6% polyacrylamide gel which had been previously electrophoresed for 30 min at 100 V. Gels were run at 0.8 V/cm 2 in 45 mM Tris-borate, followed by transfer to 3MM Whatman paper, drying under vacuum at 80°C, and exposure at Ϫ80°C to an x-ray film (Hyperfilm, Amersham) using an intensifying screen. Analysis of competition with unlabeled oligonucleotides was performed using a 20-fold excess of double stranded DNA in the binding reaction and adding the nuclear extracts as the last step in the binding assay. Supershift assays were carried out after addition of the antibody (0.5 g) to the binding reaction and incubation for 1 h at 4°C (22,28).
Western Blot Analysis-Cytosolic and nuclear extracts were obtained as described previously for the EMSAs. After determining the protein content, samples were boiled in 250 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, and 2% ␤-mercaptoethanol. Proteins (30 g and 10 g/lane for cytosolic and nuclear extracts, respectively) were size-separated in 10% SDS-polyacrylamide gel electrophoresis. The gels were processed as recommended by the supplier of the antibodies, anti-iNOS from Transduction Laboratories, and anti-IB-␣, anti-IB-␤, anti-p52 (murine) from Santa Cruz Laboratories. Anti-p50 (human), anti-p65 (human), and anti-c-Rel (human) were a generous gift from Dr. N. R. Rice (31). After blotting onto a polyvinylidene fluorescein membrane (Millipore), proteins were revealed following the ECL technique (Amersham).

LPS, TPP, and PDBu Activate Hepatocytes Transfected with
an Heterologous iNOS Promoter-Incubation of primary cultures of hepatocytes with LPS, TPP, and PDBu was found to promote iNOS expression and NO synthesis when assayed at 6 and 18 h after incubation with these factors (Fig. 1, A and B).
Absence of iNOS mRNA levels and NO synthesis resulted when the hepatocytes were incubated with ␣-phorbol 12,13-didecanoate, an inactive phorbol ester (data not shown). To investigate whether the iNOS expression elicited by TPP and PDBu is mediated by the regulatory elements identified in the murine iNOS promoter, hepatocyte suspensions were transfected by electroporation with a plasmid (p1iNOS-CAT) containing a 1.8kilobase fragment of the iNOS promoter region linked to a CAT gene, and CAT activity was measured after treatment with LPS, TPP, or PDBu. As Fig. 2A shows, hepatocytes transfected with this promoter construct exhibited a significant capacity to express CAT activity in response to these stimuli (7.2-and 14-fold increase in cells treated with PDBu and TPP or with LPS, respectively). Stimulation of hepatocytes transfected with a kSV 2 CAT vector with PDBu, TPP, or LPS did not affect CAT activity (Fig. 2B), suggesting a specificity in the response of the iNOS promoter to these factors. The release of NO and the expression of CAT activity by rat hepatocytes transfected with p1iNOS-CAT and stimulated with LPS, TPP, and PDBu suggest that the organization of the analyzed region of the iNOS promoter in murine and rat cells might be equivalent and therefore makes it possible to investigate the mechanism by which iNOS is induced by LPS, TPP, and PDBu on the basis of the known structure of the murine promoter.
Characterization of Transcriptional Activation by Electrophoretic Mobility Shift Assays-The 1.8-kilobase murine iNOS promoter region contains at least 24 consensus sequences for the binding of transcriptional factors, and some of them have an important role in the control of iNOS expression (8 -12). We have investigated whether the sequences corresponding to the distal and proximal B motifs of the iNOS promoter, as well as GAS, SIE, and AP-1 motifs, bind nuclear factors in response to hepatocyte activation by LPS, TPP, and PDBu. As Fig. 3A shows, when the proximal NF-B sequence was used for EMSA, an enhanced binding of nuclear proteins was observed in hepatocytes at 1 h after stimulation. In samples assayed at 6 h, the band corresponding to the b complex (p50/p65, see below) was significantly lower in hepatocytes incubated with PDBu but not when they were treated with LPS or TPP. Similar results were obtained when the distal B sequence was assayed. The nature of the proteins that bind to the B sequence were characterized using supershift assays. As Fig. 3B shows, the protein complexes retained in hepatocytes stimulated with 5 g/ml TPP for 1 h corresponded to p50/p50 dimers (band a) and p50/p65 heterodimers (band b), respectively. Treatment of the nuclear extracts with anti-p52 or anti-c-Rel antibodies did not modify the pattern of bands, indicating the absence of the corresponding NF-B/c-Rel proteins in the nuclei of control or stimulated hepatocytes. Identical supershift results were obtained using the distal B sequence or when nuclear extracts where prepared from cells treated for 1 h with LPS or PDBu. In addition to B, binding to AP-1, GAS, and SIE sequences reflected a similar pattern of engagement of these response elements (1 h after activation) in the mechanism of action of LPS, TPP, and PDBu (Fig. 4).
To ensure that the nuclear factors that bind B and AP-1 sequences in EMSA operate effectively in intact cells and were not the result of nonspecific binding or cytosolic contamination of the nuclear extracts, hepatocytes were transfected with plasmids harboring a tandem of three B ((B) 3 ConACAT vector) or two AP-1 motifs ((AP-1) 2 ConACAT vector) linked to a minimal promoter (corresponding to the conalbumin A gene). As Fig. 5, A and B show, CAT activity increased at least 6-fold upon stimulation with PDBu, TPP, or LPS of hepatocytes transfected with (B) 3 ConACAT or (AP-1) 2 ConACAT vectors. Transfection with a ConACAT plasmid failed to respond to these stimuli (data not shown).
Characterization of the Proteins Involved in NF-B Activation-In order to characterize the proteins involved in the activation of B motifs, the amount of p50 and p65 in the cytosolic and nuclear extracts from activated hepatocytes was assessed by Western blot. As Fig. 6 shows, p65 was translocated from the cytosol to the nuclear compartment in stimulated cells. The amount of p65 present in nuclear extracts of untreated cells was negligible, confirming the results observed by EMSA (Fig. 3). Regarding p50, the protein was immunodetected in the nuclear extracts of control cells and the levels increased upon translocation from the cytosol to the nucleus after cell stimulation with either LPS, TPP, or PDBu.
Dissociation and degradation of the IB moieties of the cytosolic NF-B complexes is an important mechanism for the control of the translocation to the nucleus of members of the Rel family (32). Since the amount of NF-B/Rel proteins that become free to access to the nucleus is supposed to be proportional to the degradation of IB, we measured the IB-␣ and IB-␤ protein levels in the cytosol by Western blot. As Fig. 7 shows, the levels of IB-␣ 1 h after stimulation of hepatocytes with LPS, TPP, or PDBu decreased by 64, 73, or 42%, respectively. FIG. 4. Binding of nuclear proteins to GAS, SIE, and AP-1 motifs. Nuclear extracts were prepared and assayed as described in Fig. 3. The oligonucleotide sequences were incubated with samples of 5 g of nuclear proteins from cells treated for 1 h with 1 g/ml LPS, 5 g/ml TPP, or 40 nM PDBu. A 50-fold excess of unlabeled oligonucleotide completely displaced the binding in all cases. Results show one representative experiment of four.

FIG. 5. Transfection of hepatocytes with (B) 3 ConACAT or (AP-1) 2 ConACAT plasmids resulted in increased CAT activity upon stimulation with PDBu, TPP, and LPS.
Hepatocytes were transfected with 10 g of (B) 3 ConACAT (upper panel) or (AP-1) 2 ConACAT plasmids (lower panel), and after seeding and stimulation with 40 nM PDBu, 5 g/ml TPP, or 1 g/ml LPS CAT activity was determined after 24 h of incubation. CAT activity was expressed as a percentage of the activity of cells transfected with a kSV 2 CAT plasmid. Results correspond to the mean Ϯ S.E. of three experiments. P values Ͻ0.01 were obtained for stimulated versus control cells.
FIG. 3. EMSA of B motifs using nuclear extracts from activated hepatocytes. Nuclear extracts were prepared from hepatocytes (3 ϫ 10 6 cells) incubated for 1 or 6 h with 1 g/ml LPS, 5 g/ml TPP, or 40 nM PDBu. EMSAs were carried out using the labeled (5 ϫ 10 4 dpm) B p sequence of the iNOS promoter as a probe. Controls to ensure the specificity of the interaction were performed using a 50-fold excess of the unlabeled probe to displace the binding (A). To analyze the proteins involved in the binding to the B p sequence, supershift assays were carried out using 1 g of the indicated antibodies. Competition assays (50-fold excess of unlabeled oligonucleotide) and an unrelated competitor oligonucleotide (other motifs) were used to ensure band specificity. Similar results were obtained in hepatocytes stimulated with PDBu or LPS (B). The results show one representative experiment of four.
However, the levels of IB-␣ at 6 h fully recovered in the cytosol independently of the treatment of the cells. Regarding IB-␤, this protein only decreased in cells treated with LPS for 1 h, and control levels were restored 6 h after stimulation. At the mRNA level, an important up-regulation of IB-␣ was observed in cells stimulated for 6 h with LPS and TPP and to a lesser extent in cells treated with PDBu (Fig. 8). In hepatocytes stimulated for 18 h with LPS an important increase of the IB-␣ mRNA levels was still evident.
Role of NF-B Activation on iNOS Expression-To investigate further the importance of nuclear binding of proteins from the Rel family to the B sites, experiments were undertaken in which cells were treated for 1 h prior to exposure to LPS, TPP, or PDBu with pyrrolidine dithiocarbamate (25 M), an inhibitor of IB degradation. As Table I shows, pyrrolidine dithiocarbamate abrogated the NO synthesis elicited by hepatocytes incubated with LPS, TPP, or PDBu to the same extent. Under these conditions, an important reduction in the nuclear content of p50/p65 complexes was observed, reflecting the fact that NF-B activation is required for iNOS expression in response to either LPS, TPP, or PDBu in hepatocytes.
The relative contribution of each B site to the activity of the iNOS promoter was analyzed using constructs in which the proximal and/or distal sequences were mutated to abolish the binding of NF-B/Rel proteins (substitution of the GGG motif by the CCG sequence). As Fig. 9 shows, mutants of the proximal site (pNOS-B pϪ vector) still retained 27-35% of the original promoter activity, whereas mutation of the distal B site (pNOS-B dϪ vector) resulted in 82-84% inhibition of the promoter activity, regardless of treatment of hepatocytes with LPS, TPP, or PDBu. As predicted on the basis of the data reported in Table I, simultaneous mutation of both B sites (pNOS-B pϪ,dϪ vector) eliminated the promoter activity.
The relative contribution of the two AP-1 binding sites to the promoter activity was investigated using deletional mutants. As Fig. 9 also shows, promoter activity of the vector p1iNOS(⌬1), which contains both the distal and proximal AP-1 sites, was roughly similar to the activity of the pNOS-B vector, which lacks the distal AP-1 site, independently of the stimulation with bacterial products or phorbol esters. The same behavior was observed when the promoter activity of p1iNOS(⌬2), which contains the proximal AP-1 and B sites, was compared with the activity of the vector p1iNOS(⌬3), which lacks the proximal AP-1 site (Fig. 9). These results agree with the data reported by other authors using different approaches and suggest that the activity of the promoter is totally independent of AP-1 activation. DISCUSSION The expression of iNOS in response to LPS and TPP is well established among different cell types; however, the ability of phorbol esters to induce iNOS seems to be restricted to some cell types such as hepatocytes, peritoneal rat macrophages, and astrocytes (5,6,33,34). For this reason, primary cultures of rat hepatocytes were used to investigate the transcriptional control mediating iNOS expression in response to bacterial products and phorbol esters.
The promoter region of the rat iNOS gene is still unknown, but it is likely to be closely related to the cloned murine sequence in view of the similarities in the transactivation mechanism of iNOS (8,9,25,33). We first confirmed that rat hepatocytes transfected with a plasmid harboring an heterologous murine iNOS promoter displayed a functional behavior that was equivalent to the response described in murine cells (8,9) and paralleled the release of NO by the endogenous enzyme, therefore suggesting that, as a first approach, this murine promoter region contains regulatory sequences activated upon treatment of rat hepatocytes with LPS, TPP, or phorbol esters.
The involvement of NF-B activation in the iNOS expression mechanism elicited by proinflammatory cytokines and LPS has been widely recognized as an absolute requirement (10,11), and in this regard, incubation of hepatocytes with pyrrolidine dithiocarbamate, an inhibitor of IB degradation (35), blocked the NO synthesis mediated in response to LPS, TPP, and phorbol esters. These results were confirmed using promoter mutants in which both B sites have been suppressed. On analysis, the distal B site contributed more than the proximal to the activity of the promoter, in agreement with the results reported by Spink et al. (25) in rat vascular smooth muscle cells. In hepatocytes transfected with plasmids containing B sequences we have confirmed the involvement of NF-B upon stimulation with both TPP and PDBu, in addition to the expected response to LPS. The characterization of the proteins involved in the formation of the NF-B complexes (36) in activated hepatocytes revealed the presence of p50/p65 heterodimers as detected by supershift assays and by Western blotting. The translocation of NF-B proteins from the cytosol to the nucleus is largely dependent on the phosphorylation and degradation of the IB subunits present in the cytosolic complexes (36 -38). In hepatocytes, treatment with LPS, TPP, or PDBu decreased the IB-␣ levels at 1 h, and only LPS was able to decrease the amount of IB-␤, which is in agreement with previous results (38). It is possible that the regulation of binding to B motifs in activated hepatocytes is even more complex, because NO seems to be an important modulator of IB-␣ function (39). However, NO has no effect on IB-␤ levels or over other transcription factors such as AP-1 or GATA (39). The importance of the role of IB-␣ in the regulation of NF-B functioning is underlined by the observation that the transrepression of NF-B by glucocorticoids is due to the control exerted by these hormones on IB-␣ levels (40).
The 1.7-kilobase murine iNOS promoter region that has been characterized contains two AP-1-related binding sequences (8,9), which prompted us to investigate the effect of this transcription factor upon hepatocyte stimulation. Fos and Jun form part of the dimeric complex recognized as AP-1, and this transcription factor participates in the regulation of the basal or the inducible activity of several genes (41). As expected from previous work (41)(42)(43), PDBu and LPS stimulated the binding of proteins to AP-1 consensus sequences at the time that induced CAT activity in hepatocytes transfected with a (AP-1) 2 ConA-CAT plasmid. The ability of TPP to activate AP-1 was previously unrecognized. However, deletion of the AP-1 binding sites in the iNOS promoter did not affect the reporter activity in hepatocytes stimulated with LPS, TPP, or PDBu. These results confirm the low contribution of the AP-1 sites to the activity of the iNOS promoter in macrophages stimulated with cytokines (8,9). Current view about transcriptional activation through the AP-1 complex stresses the importance of the phosphorylation at distinct residues of c-Jun as a critical requirement for the stability and transcriptional activity of AP-1 (41,43,44). Therefore, since several kinases can independently regulate AP-1 activity, the mechanisms that maintain a functionally active AP-1 complex in hepatocytes stimulated with LPS, TPP, or PDBu might be different.
IFN-␥ has proved to be an important cytokine for iNOS induction, synergistically acting with LPS and tumor necrosis factor-␣ or with phorbol esters in a more exclusive fashion (1,11,12,34,45,46). These effects of IFN-␥ on iNOS transcription have been attributed to the presence in the murine promoter region of this gene of a sequence for binding of interferon regulatory factor-1 (IRF-1), and this motif acts in combination with the distal NF-B site, giving rise to IFN-␥-dependent potentiation when macrophages are activated with suboptimal doses of LPS (8 -11). In addition to this, two important targets of IFN-␥ action are the GAS and SIE motifs to which STAT proteins bind (29). GAS and SIE sequences are activated not only by IFN-␥ but also by other cytokines such as epidermal growth factor, platelet-derived growth factor, and interleukin-6, the latter being an important cytokine that participates in the hepatic acute phase response (14,47,48). Moreover, a GAS site has been identified in the promoter region of the IRF-1 gene, suggesting a feedback up-regulation of this impor-TABLE I Effect of PDTC on NO synthesis and p50/p65 nuclear complexes in activated hepatocytes Cells (6 ϫ 10 6 ) were treated with 25 M PDTC prior to stimulation with 1 g/ml LPS, 5 g/ml TPP, or 40 nM PDBu and incubated for 6 h. NO synthesis was achieved using Griess reagent. The presence of p50/p60 dimers in the nuclear extracts was determined by EMSA using the proximal NF-B oligonucleotide as target. The nature of the proteins present in the retarded bands was confirmed by supershift analysis with anti-p50 and anti-p65 antibodies (human origin). Quantitation of the amount of p50/p65 complexes was done by densitometric analysis of the corresponding unsaturated films of EMSA. Results 9. Role of NF-B and AP-1 on the activity of the iNOS promoter in stimulated hepatocytes. Primary cultures of hepatocytes were transfected with the indicated vectors. Cells were stimulated for 24 h with LPS (1 g/ml), TPP (5 g/ml), or PDBu (40 nM), and CAT activity was measured. The activity of the different promoters for each treatment was referred to the pNOS-B vector taken as 100%. The activity of the pNOS-B promoter was 17, 18, and 11% of the kSV 2 CAT vector in cells stimulated with LPS, TPP, and PDBu, respectively. The vectors containing mutated B sites are indicated by the 2-base substitutions introduced in the GGG motif. Results show the mean Ϯ S.E. of three experiments. tant transcription factor for iNOS expression (49). Using the GAS site of the Ly-6E gene promoter (29), we observed in EMSA the presence of a specific band 1 h after stimulation of hepatocytes with LPS, TPP, or PDBu. Although the nature of the proteins that participate in the shift have not yet been identified, they are reminiscent of those detected in FS2 cells stimulated with IFN-␥ (29). Regarding the SIE sequence, we detected one "constitutive" band in control cells (presumably STAT1 homodimers), but upon cell stimulation two additional bands were detected (possibly STAT1⅐STAT3 heterodimers and STAT3 homodimers). These results suggest that in hepatocytes stimulated with LPS, TPP, or PDBu, common transcription factors that bind to the GAS and SIE sequences are translocated to the nucleus.
The results reported extend previous work from other groups on the identification of the pathways involved in the transcriptional control of iNOS, taking advantage of a system where the enzyme can be induced in response to different stimuli such as LPS, TPP, and phorbol esters. Moreover, it cannot be ruled out that some of the agents used (phorbol esters or bacterial products) might induce the release to the medium or the synthesis of other factors (for example, tumor necrosis factor-␣ or interleukin-6 by the minimal presence of Kupffer cells in the hepatocyte culture) that could be the final mediators of the response in a way similar to that described for the control of T cell activation via interleukin-2 receptor, in which secreted lymphokines play a prominent role in the time course of T cell triggering (50,51).
Finally, throughout this work we have observed that the transcriptional control of iNOS expression in response to these three unrelated molecules, LPS, TPP, and PDBu, is driven by the use of common NF-B/Rel transcription factors. The knowledge of the transcriptional mechanisms that control iNOS expression in the hepatocyte may contribute to a better understanding of the role of NO in hepatic pathophysiology (47).