Up-regulation of protein kinase C-epsilon promotes the expression of cytokine-inducible nitric oxide synthase in RAW 264.7 cells.

Stimulation of the murine macrophage RAW 264.7 cell line with phorbol esters fails to promote nitric oxide synthesis as occurs in rat hepatocytes or peritoneal macrophages. Transfection of RAW 264.7 cells with plasmids harboring protein kinase C (PKC) -ε isotype but not with PKC-α, -β1, -δ, or constitutively active -α and -β1 isotypes resulted in the expression of nitric oxide synthase type II (iNOS), as reflected by the synthesis of nitric oxide measured in the culture medium of transfected cells. cotransfection of RAW 264.7 cells with the −1592 to +121-base pair promoter region of the murine iNOS gene and PKC isotypes specifically induced the transactivation of this promoter in the case of the plasmids containing the PKC-ε isotype. The mechanism by which PKC-ε induced iNOS expression involved the activation of nuclear factor binding to κB sites (NF-κB) as deduced by the suppressive effect of pyrrolidine dithiocarbamate on nitric oxide synthesis, an inhibitor of NF-κB activation, and by the activation of κB sites in cells transfected with a vector containing a κB motif linked to a chloramphenicol acetyltransferase reporter gene. These results suggest that PKC-ε can regulate a pathway that promotes iNOS expression in macrophages in response to phorbol ester activation.

The knowledge of the mechanisms involved in the control of NO synthesis by different cell types is a subject of current interest because of the multiple physiological and pathological effects elicited by this molecule on different cell targets (1)(2)(3). At least three distinct but functionally and structurally related genes involved in the biosynthesis of NO have been identified in mammalian cells, and they are referred to as type I NOS, 1 constitutively expressed in cells of neural origin (3,4); type II NOS or iNOS, an activity which is mainly controlled at the transcriptional level in response to a wide array of pro-inflammatory cytokines, bacterial cell wall products, and through the engagement of signaling pathways involved in host defense against pathogens and tumor cells (1,5); and type III NOS that is constitutively expressed in cells of endothelial origin and participates in the control of the basal vascular tone (6). The activity of type I and III NOS is mainly regulated through changes in calcium/calmodulin concentration, whereas iNOS is constitutively active because of its tight binding to calcium/ calmodulin, and therefore, the transcriptional control of this gene is of great importance for the regulation of the enzyme activity (1,(7)(8)(9).
The transcription of the iNOS gene is switched off in most cells under physiological conditions, but after triggering with bacterial products or inflammatory cytokines, the gene is transcribed in many cell types including macrophages, hepatocytes, smooth muscle cells, astrocytes, and chondrocytes among others (9 -11). The specificity of the iNOS induction depends on the nature of the cells and on the animal species, being more sensitive in rodent cells than in their human counterparts, in which a cooperative interaction between various signals is required for iNOS expression (1,2,12).
The ability of lipopolysaccharide (LPS) and proinflammatory cytokines to induce iNOS expression has been widely documented. However, the existence of a phorbol ester-mediated iNOS induction mechanism has been observed in a more reduced number of cell types, as for example rat peritoneal macrophages and hepatocytes (13,14). In this regard, a role for PKC activation in the control of iNOS expression has been reported in the synergistic action of phorbol esters and IFN-␥ in murine peritoneal macrophages and in the cell line J774, as well as in astrocytes and microglia cells (15)(16)(17)(18). Also, priming with taxol of RAW 264.7 cells favors iNOS induction via a PKC-dependent mechanism, but treatment with phorbol esters alone has no effect on NO synthesis (19). This cellular diversity in iNOS expression in response to phorbol esters might be due to multiple mechanisms, among them the requirement of a precise pattern of expression of PKC isotypes (20). The view emerging from these studies is that PKC engagement might be involved as part of the intracellular signaling that promotes iNOS expression and, in some cases, constitutes a self-sufficient pathway to accomplish NO synthesis. Indeed, the structure of the murine iNOS gene has been analyzed, a 1.8-kb fragment of the promoter region has been characterized, and it contains most of the transcriptional motifs involved in the transactivation of iNOS including two critical B sites among other consensus binding sequences (21)(22)(23). Because some PKC isotypes have been reported to activate B motifs (24,25), we have investigated in this work whether ectopic expression of various phorbol ester-responsive PKC isotypes in RAW 264.7 cells can promote iNOS induction. Our results show that expression of PKC-⑀ is sufficient to induce iNOS synthesis in RAW 264.7 cells through a mechanism that involves the transactivation of response elements located in the 1.8-kb region of the iNOS promoter.

MATERIALS AND METHODS
Chemicals-Materials and chemicals for electrophoresis were from Bio-Rad or from Amersham Corp. Metabolites and biochemicals were from Sigma or from Boehringer Mannheim. Serum and media were from BioWhittaker, Inc. (Walkersville, MD).
Cell Culture-RAW 264.7 cells were obtained from ATCC and correspond to a murine macrophage cell line. The cells were cultured in RPMI 1640 medium supplemented with 2 mM glutamine, 10% FCS, and antibiotics (50 g/ml of penicillin, streptomycin, and gentamicin). Peritoneal macrophages were obtained from Swiss mice following a published method (13). Rat hepatocytes, murine splenocytes and B cells, and human neutrophils were isolated and purified as described previously (14,26).
Plasmid Constructs and Preparation-The 1,749-bp HincII fragment, corresponding to the 5Ј-flanking region of iNOS (21,22) fused to a promoterless CAT reporter gene (p1.NOS-CAT), was a generous gift from Dr. Q.-w. Xie and C. Nathan (Cornell University, NY). A (B) 3 -CAT plasmid construct that contains three copies of the B motif from the human immunodeficiency virus long terminal repeat enhancer with the conalbumin promoter was used to measure B binding activity (27). kSV 2 -CAT plasmid was used as a reference for maximal efficiency of the transfection (28). The pCO 2 vector (Dr. S. Goodbourne, St. Georges Hospital, London) was used to express the PKC isotypes (24), and they were referred to as pPKC-␣, -␤ 1 , -␦, and -⑀. Plasmids directing expression of constitutively active isotypes were referred to as pPKC-␣* and pPKC-␤ 1 *, respectively (see Table I). Plasmids were purified using Qiagen columns (Hilden, FRG), and only those preparations that once diluted contained less than 30 pg/ml of endotoxin using the Limulus polyphemus test were used. Additionally, to evaluate the stimulatory capacity of the plasmid preparations to mediate iNOS expression due to bacterial product contamination, cells were electroporated in the absence of plasmids and challenged with an equivalent amount of plasmid upon culture. Under these conditions PKC plasmids never induced an NH 4 Ϫ release higher than 10% over the basal value. Transfection of RAW 264.7 Cells-The cell layer was trypsinized, and macrophages were collected by centrifugation at 200 ϫ g for 5 min. The cell pellet was resuspended in ice-cold PBS (3 ϫ 10 6 cells/ml) and kept at 4°C. Plasmids were added to the cells at the indicated concentration and incubated at 4°C for at least 5 min prior to electroporation (0.4 kV, 500 microfarad of capacitance) in a BTX electroporator. Cells were maintained for a further 5 min at 4°C and then transferred to RPMI 1640 medium containing 5% FCS. After two h of seeding to select adherent cells, the dishes were washed twice with PBS and maintained for 24 h in culture medium (phenol red-free medium plus 5% FCS). To follow NO release, the medium was changed, and stimuli (usually 50 ng/ml PDBu) were added. Cell incubation was continued for up to 48 h. When cells were transfected by lipofection (DOTAP reagent from Boehringer Mannheim), it was done following the recommendations of the supplier manufacturer.
Characterization of iNOS Expression by Northern Blot-Total RNA (2-4 ϫ 10 6 cells) was extracted using the guanidinium thiocyanate method (29). After electrophoresis in a 0.9% agarose gel containing 2% formaldehyde, the RNA was transferred to a Nytran membrane (NY 13-N; Schleicher & Schuell, FRG), and the levels of iNOS mRNA were determined using an EcoRI-HindII fragment from the iNOS cDNA (1, 5) labeled with [␣-32 P]dCTP using the Readyprime labeling kit (Amersham). The membranes were exposed to x-ray films (Hyperfilm, Amersham), and the intensity of the bands was measured by laser densitom-etry (Molecular Dynamics). Hybridization with an 18 S ribosomal probe was used as an internal standard.
Determination of NO x Ϫ Synthesis-NO was measured as the accumulation of nitrite and nitrate in the incubation medium. Nitrate was reduced to nitrite with nitrate reductase (30). Nitrite was determined spectrophotometrically with Griess reagent (30) by adding 100 l of 1 mM sulfanilic acid and 100 mM HCl (final concentration) to 850 l of culture medium. After incubation for 5 min, the absorbance at 548 nm was measured, and 50 l of naphthylenediamine (1 mM in the assay) were added. The reaction was completed after 15 min of incubation, and the absorbance at 548 nm was compared with a standard of NaNO 2 . The amount of nitrate produced from NO release was below 20% of the nitrite measured.
Assay of PKC Activity-The cell layers were washed twice with ice-cold PBS and homogenized in a Dounce homogenizer using 20 mM Tris-HCl, pH 7.5, 5 mM EDTA, 5 mM EGTA, 10 mM ␤-mercaptoethanol, 200 M phenylmethylsulfonyl fluoride, 20 g/ml leupeptin, and 0.5% Nonidet P-40 (lysis buffer). The extract was centrifuged in an Eppendorf centrifuge for 15 min, and the supernatant was applied to a DEAE column (500-l gel) to partially purify the enzyme (31). The activity was assayed in a final volume of 250 l using myelin basic protein peptide (MBP) or the peptide ERMRPRKRQGSVRRRV (␤-PSS) as substrates following previously described methods (31,32).
Characterization of PKC and iNOS by Western Blot-Cultured RAW 264.7 cells (3-4 ϫ 10 6 ) were washed twice with PBS and homogenized in lysis buffer (see above), and after centrifugation in a microcentrifuge for 15 min, the proteins present in the supernatant were size separated in 10% SDS-polyacrylamide gel electrophoresis. The gels were blotted onto a polyvinylidene difluoride membrane (Millipore) and incubated with several anti-PKC-specific antibodies (33 and references herein) or with an iNOS antibody (Transduction Laboratories). PKC isotypes or iNOS were revealed by ECL following manufacturers instructions (Amersham).
Determination of CAT Activity-In experiments investigating the effect of PKC isotypes on the expression of the transfected p1.NOS-CAT promoter (21) or (B) 3 -CAT construct (27), cells were cotransfected with both plasmids and incubated with the indicated stimuli for 24 -48 h. After two washes of the cell layer with ice-cold PBS, the plates were treated with 0.5 ml of 0.25 M Tris-HCl, pH 7.8, at 4°C, and the cells were scraped off the dishes. The cell extract was submitted to three cycles of freezing and thawing followed by centrifugation at 12,000 ϫ g for 10 min, and the soluble protein was measured. Aliquots of the supernatant (150 -200 l) were normalized for protein (200 g) and heated at 65°C for 10 min. CAT activity was measured in a final volume of 250 l by the synthesis of acetylated [ 14 C]chloramphenicol following the thin layer chromatography method (28). To quantify the amount of acetylated substrate, the silica spots were scraped off the matrix, and the radioactivity was determined by liquid scintillation counting.
Data Analysis-Statistical differences (p Ͻ 0.05) between mean values were determined by one-way analysis of the variance followed by Student's t test. In experiments using x-ray films (hyperfilm), different exposure times were used to ensure that bands were not saturated.

RESULTS
Transfection of RAW 264.7 Cells with pPKC-⑀ Promotes the Expression of iNOS-Incubation of cultured RAW 264.7 cells with LPS promoted a large release of NO to the medium, whereas exposure of cells to phorbol esters failed on their own to trigger an effective NO synthesis, and only a low response (6% of the effect of LPS) was observed (Fig. 1). Contrary to this behavior and in agreement with previous work (13,14), stimulation of primary cultures of rat peritoneal macrophages or isolated hepatocytes with PDBu produced a NO x Ϫ release that was 83 and 71%, respectively, of the response elicited after LPS treatment. This PDBu-dependent NO x Ϫ synthesis was absent in other cells such as isolated splenocytes, B cells, or human neutrophils. The different behavior observed between RAW 264.7 cells and peritoneal macrophages in the expression of iNOS in response to phorbol esters suggests the existence of cell-specific signaling pathways in these two cell types, either because of a distinct distribution of the PKC isotypes prevailing in each cell type or because of a divergent signaling mechanism downstream of the PKC network. Taking advantage of this inability of RAW 264.7 cells to display a significant PDBu-dependent NO x Ϫ synthesis, we studied whether a transient expression of PKC isotypes after transfection with plasmids harboring distinct PKC isoenzyme constructs might restore a PKCdependent induction of iNOS. To do this, RAW 264.7 cells were transfected by electroporation with plasmids encoding the phorbol ester responsive wild-type ␣, ␤1, ␦, and ⑀ isoforms of PKC, and the release of NO x Ϫ was followed from 24 to 48 h after transfection (Fig. 2). Also, cells were transfected with plasmids encoding constitutively active forms of PKC (␣* and ␤1* isotypes), and the release of NO x Ϫ was followed either in the absence or in the presence of 50 ng/ml of PDBu. As Fig. 2 shows, transfection with pPKC-⑀ resulted in a substantial NO x Ϫ synthesis that was barely affected by addition of PDBu. To ensure that this NO x Ϫ synthesis effectively corresponded to the expression of iNOS, cells were collected at the end of the incubation period, and the presence of iNOS was assessed by Western blot (Fig. 2, inset). Also, the nature of the NOS present in PKC-transfected cells was determined following the mRNA levels corresponding to iNOS. Only cells transfected with PKC-⑀ exhibited a substantial increase in the mRNA levels of iNOS in samples collected at 18 and 24 h after transfection (Fig. 3). Similar results in NO x Ϫ synthesis were obtained when cells were transfected by lipofection (DOTAP reagent) with pPKC-␣, -␣*, and -⑀ (Table II).
As Fig. 2 shows, the extent of NO x Ϫ release from cells transfected with pPKC-⑀ was not further stimulated by the addition of phorbol esters. One explanation of this result is that there may be a high over-expression of the enzyme in the transfected cells. To test this idea, cells were transfected with increasing amounts of pPKC-⑀ plasmid followed by the addition of PDBu. As Fig. 4 shows, when cells were transfected with low amounts of plasmid, a gain in sensitivity to PDBu stimulation was observed; however, this was not the case when cells were trans-  fected with concentrations of plasmid higher than 10 g/ml. Fig. 4 also shows that addition of the plasmid to the cells after electroporation was unable to trigger NO synthesis, discounting a potential contribution of bacterial cell wall products from the plasmid preparation to the induction of iNOS.
The previous results suggest that some PKC isotypes are key regulators for the control of iNOS expression, at least in RAW 264.7 cells. Therefore, it can be proposed that the absence of response of RAW 264.7 cells to PDBu might be attributed to a low content of PKC-⑀ or perhaps to an inappropriate intracellular localization of the enzyme. On analysis, the level of the PKC-⑀ in RAW 264.7 cells was practically undetectable by Western blot when compared with other cell types such as 3T3 cells (Fig. 5A). When cells were transfected with the panel of pPKC plasmids used, PKC-⑀ was only evident in cells transfected with pPKC-⑀ (Fig. 5B). Moreover, this overexpressed PKC-⑀ enzyme was active since the total PKC activity with MBP as substrate was notably increased as was the phosphorylation of a more "preferred" substrate of PKC-⑀, ␤-PSS (Fig. 5C).
Transfection with pPKC-⑀ Induces CAT Activity of Cells Cotransfected with a p1.NOS-CAT Plasmid Encoding a 1.8-kb Fragment of the iNOS Promoter-The preceding results suggest that some isotypes of PKC are sufficient to trigger an efficient expression of iNOS. To further investigate the mechanisms involved in the response of iNOS to PKC activation, we used a cotransfection assay in which both PKC isotypes and a 1.8-kb fragment of the promoter region of iNOS (21) linked to a CAT reporter gene were expressed. As Fig. 6A shows, CAT activity was present in cells transfected with either 10 or 20 g/ml of plasmid harboring PKC-⑀. Moreover, these results also show the absence of transactivation when iNOS promotertransfected cells were stimulated with phorbol esters, whereas treatment with LPS elicited a marked increase in the reporter activity, therefore, reflecting the inability of the endogenous PKC to mediate a significant transactivation of this promoter region of the iNOS gene. The ability of the different pPKC plasmids to express CAT activity, upon cotransfection with the iNOS promoter, is shown in Fig. 6B. CAT activity was negligible when cells were transfected with pPKC-␣ and -␦, whereas a small effect was observed in cells harboring the ␤ 1 isotype (Fig. 6B).
Because the activation of the B sequences located in the 1.8-kb fragment of the iNOS promoter plays an important role in the transactivation of the gene (21-23), we investigated whether this was the case in cells cotransfected with PKC isotypes and an expression vector carrying a tandem of 3 B consensus sequences linked to a CAT reporter gene (27). As Fig. 7 shows, 24 h after transfection with vectors directing the expression of PKC-␣* and -⑀ isotypes, a clear increase in CAT activity was observed in cells expressing PKC-⑀ (10.2-fold increase), whereas CAT activity was minimal in cells transfected with PKC-␣*. When cells transfected with PKC expression vectors were incubated with PDTC, an inhibitor of IB degradation (34), CAT activity and NO x Ϫ synthesis were abrogated in agreement with the expected results using this NF-B activation inhibitor. As a control, stimulation with LPS of cells transfected with the (B)3-CAT vector induced a large CAT activity confirming the effect of LPS on B activation in these cells (21)(22)(23). DISCUSSION The study of the induction of iNOS in murine peritoneal macrophages and in macrophage cell lines has constituted the

cells transfected by lipofection
Cells (2 ϫ 10 6 ) were transfected by lipofection (DOTAP) for 4 h with the indicated plasmids (10 g/ml of medium) following the instructions of the manufacturere. Transfected cells were washed twice and maintained in culture medium for 12 h. After replacing the culture medium, cells were stimulated with or without 50 ng/ml of PDBu, and the release of NO reference model for the characterization of the mechanisms responsible for the transcriptional control of this enzyme (1, 5, 7, 10 -11, 21-23). The murine promoter region of iNOS has proved to be quite complex, containing at least 24 consensus sequences for the binding of transcription factors; among them, proteins of the NF-B family appear to be essential components for the transactivation of iNOS (21)(22)(23). Protein kinase C stimulation by phorbol esters is able to activate NF-B and AP-1, two transcription factors that bind to the iNOS promoter (25,35,36). However, treatment of RAW 264.7 cells with different phorbol esters and using a broad range of concentrations of these molecules to prevent side-effects in the course of PKC activation (such as the down-regulation of some PKC isotypes) failed to promote iNOS expression and, therefore, NO synthesis. For this reason, RAW 264.7 cells were transfected with expression vectors encoding different phorbol ester-responsive PKC isotypes as an approach to assess whether any of these enzymes might provide an appropriate signal to trigger the expression of iNOS. Our results clearly show that among the PKC isotypes tested, PKC-⑀ is a likely candidate to mediate the stimulatory effect on NO synthesis observed in other cell types in response to PKC activation (13)(14)(15)(16)(17)(18). The complete absence of PKC-⑀ in RAW 264.7 cells (undetectable by Western blot when compared with 3T3 fibroblasts as a murine reference) was unexpected; however, it provided a plausible explanation for the lack of response of NO synthesis after incubation of these cells with phorbol esters. Indeed, overexpression of the ␤ 1 isotype barely induced NO release to the medium, whereas expression of the ␣ and ␦ isotypes failed to induce iNOS. An additional conclusion from these results is that expression of iNOS in response to LPS or proinflammatory cytokines does not require PKC-⑀ engagement, at least in these cells. However, it is of interest to note that in this cell model, several authors have described an inhibitory effect in NO synthesis using more or less specific PKC inhibitors (19,37,38). The identification of biological effects dependent on the specific activation of a particular PKC isotype is a matter of current research interest since differences among the PKC isotypes exists in the requirements of cofactors for activation, phospholipid-dependence, and subcellular distribution (39). To provide a few examples, it has been reported that overexpression of PKC-␣ in Chinese hamster ovary cells specifically inhibited the signaling of several members of the insulin receptor family (40). Regarding PKC-⑀, this enzyme, and to a lesser extent the ␣ isotype, has been reported as a regulator of the transcription factors AP-1 and nuclear factor of activated T cells (NF-AT-1) in human T cell lymphoma, whereas only PKC-⑀ transactivated B constructs (24). In this system, activation via overexpression of PKC-⑀ required the use of constitutively active mutants, whereas wild-type PKC-⑀ was ineffective. These data contrast with our results where a gain of function (NO synthesis) was exclusively mediated by overexpression of this isotype and in the absence of PKC-⑀ activators. However, incubation of the transfected cells with phorbol esters produced a leftward shift in the dose dependence of PKC-⑀ cDNA required to induce iNOS (see Fig. 4). Moreover, it is possible that specific subcellular locations of PKC isotypes play an important role in the regulation of the activation pathway (e.g. PKC-⑀ has been reported to be localized to the Golgi via its Zn-finger domain, see Ref. 41). This interaction with subcellular structures has been reported for other PKC isotypes; for example, changes in the subcellular localization of PKC-␣, but not in the total amount of enzyme, are closely associated with the transformation of rat embryo fibroblasts (42). Recently, a role for PKC-⑀ has been reported in the signaling mechanism elicited by platelet-derived growth factor involving two independent pathways, phospholipase C-␥ and phosphatidylinositol-3-kinase activation (43). In this case, PKC-⑀ would appear to integrate signals released by redundant, but independent, input stimuli.
Deletion analysis of the promoter region of iNOS upon transfection of RAW 264.7 cells with a 1.8-kb region revealed the requirement for the integrity of the B sequences for the transactivation after triggering with LPS or when LPS was acting synergistically with IFN-␥ (5,(21)(22)(23). Indeed, treatment with PDTC of RAW 264.7 cells transfected with PKC-⑀ blocked iNOS expression, revealing the necessity of NF-B activation in this process. Cotransfection with the Ϫ1592 to ϩ121 promoter region of iNOS and distinct PKC isotypes also showed a specific response in the reporter activity when PKC-⑀ was expressed, confirming an important degree of PKC specificity in the process of iNOS induction. In agreement with these results, NF-B was activated in cells transfected with a pPKC-⑀ vector. These data agree with results obtained in nontransfected cells stimulated with PDBu in which a lack was observed in the binding of nuclear proteins of the rel family to oligonucleotide se- quences corresponding to the B sites of the iNOS promoter (nucleotides extending from position Ϫ85 to Ϫ76 and Ϫ971 to Ϫ962; see Refs. 21-23) when assayed following the electrophoretic mobility shift of these oligonucleotides (results not shown). Moreover, the occurrence in RAW 264.7 cells of autocrine mechanisms switched on through PKC-⑀ activation (i.e. release of proinflammatory cytokines) that could contribute to the process of iNOS expression cannot be excluded. Indeed, this is the case for LPS that promotes the synthesis of cytokines such as TNF-␣ in activated macrophages in such a way that the control of iNOS expression relies on the signaling of more than one cytokine (1,7,44). In addition to these potential regulatory mechanisms, a cell-specific modulation of the promoter activity of iNOS can be proposed. This is the case for vascular smooth muscle cells in which the activity of the 1.8-kb murine iNOS promoter appears to be differentially regulated when compared with the macrophage counterparts (45). This situation has been interpreted in terms of a local and beneficial response that facilitates the delivery of immunomodulatory effectors intended to restore the normal function of the injured area.
In summary, evidence is presented for the selective and specific effect of PKC-⑀ in inducing iNOS in RAW 264.7 cells. All the PKC isotypes tested are phorbol ester responsive and can induce biological effects in other contexts (see Ref. 24). The specific ability of PKC-⑀ to induce iNOS and to confer PDBu responsiveness to the RAW 264.7 cells indicates a particular role for PKC-⑀ that cannot be fulfilled by these other phorbol ester responsive PKC isotypes. Consistent with the iNOS promoter analysis, it is only the PKC-⑀ isotype that is capable of inducing NF-B. Thus the specificity of this response suggests an important role for the PKC-⑀/NF-B/iNOS pathway in host defense, inflammation and cytotoxic responses. While apparently not involved in the LPS response, it will be of future interest to establish those agonists that trigger iNOS via PKC-⑀.