Up-regulation of Prostaglandin E2 Synthesis by Interleukin-1β in Human Orbital Fibroblasts Involves Coordinate Induction of Prostaglandin-Endoperoxide H Synthase-2 and Glutathione-dependent Prostaglandin E2 Synthase Expression*

Prostaglandin E2(PGE2) production involves the activity of a multistep biosynthetic pathway. The terminal components of this cascade, two PGE2 synthases (PGES), have very recently been identified as glutathione-dependent proteins. cPGES is cytoplasmic, apparently identical to the hsp90 chaperone, p23, and associates functionally with prostaglandin-endoperoxide H synthase-1 (PGHS-1), the constitutive cyclooxygenase. A second synthase, designated mPGES, is microsomal and can be regulated. Here we demonstrate that mPGES and PGHS-2 are expressed at very low levels in untreated human orbital fibroblasts. Interleukin (IL)-1β treatment elicits high levels of PGHS-2 and mPGES expression. The induction of both enzymes occurs at the pretranslational level, is the consequence of enhanced gene promoter activities, and can be blocked by dexamethasone (10 nm). SC58125, a PGHS-2-selective inhibitor, could attenuate the induction of mPGES, suggesting a dependence of this enzyme on PGHS-2 activity. IL-1β treatment activates p38 and ERK mitogen-activated protein kinases. Induction of both mPGES and PGHS-2 was susceptible to either chemical inhibition or molecular interruption of these pathways with dominant negative constructs. These results indicate that the induction of PGHS-2 and mPGES by IL-1β underlies robust PGE2 production in orbital fibroblasts.

The past decade has witnessed dramatic advances in our understanding of the prostanoid biosynthetic pathways in mammalian cells. Of particular importance was the discovery and molecular characterization of two cyclooxygenase isoforms and the realization that each protein might possess distinct functions and patterns of expression and regulation (1,2). Prostaglandinendoperoxide H synthase-1 and -2 (EC 1.14.99.1, PGHS) 1 are membrane-associated and contain a heme prosthetic group. The PGHS isoforms have received substantial attention, in large part because they are the targets of aspirin and a large series of nonsteroidal anti-inflammatory compounds (3). PGHS-1 is constitutively expressed in most tissues and is thought to generate prostaglandins involved in housekeeping activities (1,2,4,5). In contrast, PGHS-2 is expressed in most tissues only following cell activation by cytokines, growth factors, and mitogens (6 -10). Prostanoids generated through the activities of PGHS-2 are thought to represent those produced under circumstances of inflammation and tissue disruption.
A very recent and significant advance has resulted from the identification of two prostaglandin E 2 (PGE 2 ) synthase enzyme isoforms (EC 5.3.99.3) (11)(12)(13). Both are glutathione-dependent and catalyze the terminal conversion reaction of PGH 2 to PGE 2 . One enzyme is a constitutively expressed cytosolic protein, designated cPGES (12). Evidence has been advanced indicating that cPGES is identical to p23, a putative chaperone of hsp90 that stabilizes steroid hormone receptor-hsp90 complexes (14). It exhibits substantial constitutive expression in many tissues. Moreover, this expression was reported to be invariant in several cell lines in vitro with regard to treatment with IL-1␤ or TNF-␣ (12). cPGES protein was found to be induced modestly in rat brain 48 h after injection of lipopolysaccharide (12). The other isoform, mPGES, is a 16-kDa microsomal protein that can be regulated (11,13). This enzyme may be expressed and regulated in a cell type-specific manner. mPGES is induced by lipopolysaccharide in rat macrophages, and this up-regulation can be blocked by the glucocorticoid, dexamethasone (13). Neither treatment altered the levels of cPGES in macrophages. Thus mPGES represents a regulated enzyme and a potentially important drug target. Very recent studies imply a preferential, functional association of each PGES isoform with a particular PGHS enzyme (12,13). The studies demonstrated that cPGES is linked to PGHS-1 and that mPGES utilizes PGH 2 generated by PGHS-2. Further studies have extended the notion that functional, stimuli-dependent linkage exists between specific PGHS isoforms and cell typespecific downstream enzymes (15). These observations were made in transfected cells where one or more of the relevant enzymes had been over-expressed. Importantly, nothing is known currently about whether the expression of endogenous PGHS isoforms might in some way exhibit physiological coordination with that of the PGES enzymes. Such a coupling FIG. 1. IL-1␤ up-regulation of PGE 2 synthesis in human fibroblasts is associated with the coordinate induction of PGHS-2 and mPGES proteins. Orbital fibroblasts, in this case from a patient with severe thyroid-associated ophthalmopathy, were allowed to proliferate to confluence in medium supplemented with 10% FBS. A, monolayers were shifted to medium without serum for 24 h and then treated with IL-1␤ (10 ng/ml) for the times indicated along the abscissa. For the PGE 2 determinations, medium was removed 30 min before the cultures were harvested and replaced with PBS with the respective additives. PBS was then collected and subjected to the assay described under "Experimental Procedures." For Western blot analysis of PGES and PGHS protein levels, monolayers were harvested and analyzed as described. B, cultures were treated as described for panel A, but IL-1␤ concentrations were graded as indicated along the abscissa, and the treatment time was uniformly 16 h. Western blots were scanned; the relative densities are represented as columns underneath the images. Data from the PGE 2 assay are expressed as the mean Ϯ S.D. of triplicate determinations. The data derive from a single representative experiment of the three performed. would be consistent with a model of PGE 2 production that exhibited functional cellular compartmentalization.
Fibroblasts have been shown to express components of the prostanoid biosynthetic pathways, and cultures derived from particular anatomic regions and tissues can generate large amounts of PGE 2 when provoked (16). These sentinel cells are critical to the orchestration of inflammatory responses, tissue remodeling, and wound healing and are key participants in the evolution of fibrosis (17). Diversity among human fibroblasts has only recently been appreciated as being potentially important to normal tissue function and disease manifestation. It is now clear that fibroblast subsets represent highly specialized cells that participate in reactive processes in the context of both normal function and pathology (18,19). Fibroblasts function in various aspects of immunity, the recruitment of bone marrowderived cells, tissue repair, and remodeling and provide the molecular and structural infrastructure supporting cellular cross-talk (17,18,20). Human orbital fibroblasts exhibit a phenotype that sets them apart from fibroblasts derived from other anatomic regions. They are composed of discrete subsets (18), one of which represents pre-adipocytes (21); possess a characteristic morphology (22); and display a distinct pattern of gangliosides (23) and surface receptors (24). Of particular relevance to their proposed participation in orbital inflammation are the robust responses to a variety of disease mediators such as cytokines, growth factors, and bioactive lipids (25)(26)(27)(28)(29). These fibroblasts have been implicated in the pathogenesis of thyroid-associated ophthalmopathy (TAO), an autoimmune process. Two hallmarks of the tissue remodeling observed in TAO are the accumulation of hyaluronan and an often intense inflammatory reaction (30). We hypothesize that it is this set of attributes that renders connective tissues in the human orbit susceptible to the remodeling associated with TAO.
Among the characteristics that set apart orbital fibroblasts is the dramatic up-regulation of PGE 2 synthesis observed following exposure of these cells in culture to proinflammatory cytokines (25,26). These increases in PGE 2 can be blocked by specific inhibitors such as SC58125 and NS-398, suggesting a dominant role for PGHS-2 in mediating the up-regulation (25). In fact, orbital fibroblasts when provoked by IL-1␤ or leukoregulin or through engagement of surface-displayed CD40 by CD154, exhibit a particularly robust induction of PGHS-2 (25,26,31).
Here, we report that the treatment of orbital fibroblasts with IL-1␤ results in dramatic increases in PGE 2 production and is associated with coordinate induction of both PGHS-2 and mPGES expression. These responses are mediated through elevations in the steady-state levels of their respective mRNAs and involve the use of overlapping intracellular signaling pathways, including the p38 and ERK mitogen-activated protein (MAP) kinases. Inhibiting PGHS-2 activity results in the blockade of mPGES induction by IL-1␤, indicating some involvement of the products of the former enzyme in the expression of the latter. These latest findings provide insights into the complex interactions between PGHS-2 and mPGES in orbital fibroblasts that culminate in the generation of PGE 2 .
Cell Culture-Orbital fibroblast cultures were initiated from tissue explants obtained as surgical waste during decompression surgery from severe TAO or were from normal appearing orbital tissues in patients undergoing surgery for non-inflammatory conditions. These activities have been approved by the Institutional Review Boards of Albany Medical College and Harbor-UCLA Medical Center. Some of the fibroblast strains were kindly provided by Dr. Rebecca Bahn (Mayo Clinic, Rochester, MN). Tissue fragments were generated by mechanical disruption of explants, and fibroblasts were then allowed to adhere to plastic culture plates. They were covered with Eagle's medium to which 10% fetal bovine serum (FBS), glutamine (435 g/ml), and penicillin/ streptomycin were added as described previously (33). Medium was changed every 3-4 days, and monolayers were maintained in a 5% CO 2 , humidified incubator at 37°C. Culture strains were utilized between the second and twelfth passage from initiation. All experimental manipulations were conducted after a state of confluence had been reached. We have already established the purity of these cultures and found them to be essentially free of contamination by endothelial and smooth muscle cells (18).
RNA Isolation and Northern Hybridization-Fibroblasts were cultivated in 100-mm-diameter plates to a confluent state and were then treated with the test agents specified in the figure legends. Cellular RNA was extracted from rinsed monolayers by the method of Chomczynski and Sacchi (34) with an RNA isolating system purchased from Biotecx (Houston, TX). The nucleic acid was subjected to electrophoresis through denaturing 1% agarose, formaldehyde gels. Integrity of the RNA was established by determining the 260/280 spectroscopic ratios and by staining the electrophoresed samples with ethidium bromide and inspecting them under UV light. The RNA was transferred to Zeta-probe membrane (Bio-Rad), and immobilized samples were hybridized with [ 32 P]dCTP-labeled PGHS-1, PGHS-2, p23, and mPGES cDNA probes generated by the random primer method. Hybridization was conducted in a solution containing 5ϫ SSC, 50% formamide, 5ϫ Denhardt's solution, 50 mM phosphate buffer (pH 6.5), 1% SDS, and 0.25 mg/ml salmon sperm at 48°C overnight. Membranes were washed under high stringency conditions, and then the RNA/DNA hybrids were visualized by autoradiography on X-Omat film (Kodak, Rochester, NY) following exposure at Ϫ70°C. Bands resulting from radioactive hybrids were scanned by densitometry. Membranes were then stripped according to the instructions of the manufacturer and rehybridized with a GAPDH cDNA probe, and the band densities were normalized to this signal.
For mPGES and PGHS-2 mRNA stability studies, cultures were treated with IL-1␤ for 3 h as a pretreatment. Cells were washed and incubated in growth medium for 4 h. At time 0, DRB (20 g/ml), an inhibitor of gene transcription, was added to the medium of all plates without or with IL-1␤ (10 ng/ml) for the intervals indicated in Fig. 4. Abundance of mRNAs for the two enzymes was quantified by Northern blot hybridization. mPGES and PGHS-2 mRNA signals were normalized to their respective GAPDH levels.
Western Blot Analysis of Fibroblast Proteins-Cellular proteins were solubilized from rinsed fibroblast monolayers following the treatments indicated in the figure legends. The ice-cold harvest buffer contained 0.5% Nonidet P-40, 50 mM Tris-HCl (pH 8.0), and 10 M phenylmethylsulfonyl fluoride. Lysates were taken up in Laemmli buffer and subjected to SDS-PAGE, and the separated proteins were transferred to polyvinylidene difluoride membrane (Bio-Rad). Primary monoclonal antibodies directed against PGHS-1 and PGHS-2 (10 g/ml, Cayman) were incubated with the membranes for 2 h at room temperature. Following washes, membranes were reincubated with secondary peroxidase-labeled antibodies. In other experiments, primary antibodies directed against human mPGES and cPGES were utilized for the detec- FIG. 3. A, Northern blot analysis of the effects of IL-1␤ on the steadystate levels of mRNAs encoding mPGES, cPGES, PGHS-2, and PGHS-1 in orbital fibroblasts. Confluent fibroblast cultures were shifted to medium without FBS to which IL-1␤ (10 ng/ml) was added for the time intervals indicated along the abscissa. Monolayers were rinsed and cellular RNA harvested as indicated under "Experimental Procedures." RNA was then subjected to Northern blot hybridization with the respective 32 P-labeled cDNA probes. The radioactive RNA/DNA hybrids were visualized by exposing membranes to X-Omat film. The relative densities of the PGHS-2 and mPGES hybridization are displayed as columns underneath the images following their normalization with GAPDH signals. B, induction of mPGES mRNA by IL-1␤ in orbital fibroblasts is independent of intermediate protein synthesis. Cultures were shifted to medium without FBS to which nothing (control), IL-1␤ (10 ng/ml), cycloheximide (Cyclo) (10 g/ml), or a combination of the test compounds was added for 6 h. Cellular RNA was extracted and subjected to Northern blot analysis. (OD of GAPDH-corrected PGHS-2 signal in IL-1␤-treated sample is 0.98 and that in the IL-1␤ ϩ cycloheximide sample is 0.21). tion of these enzyme proteins. The ECL (Amersham Biosciences) chemiluminescence detection system was used to generate signals, and the resulting bands were analyzed with a densitometer. With regard to assessing the activation of p38 and ERK MAP kinases, cell lysates from untreated orbital fibroblast cultures and those treated with IL-1␤ (10 ng/ml) were subjected to SDS-PAGE; the proteins were transferred to membrane and then probed with phospho-specific antibodies against the two kinases. PGE 2 Assay-Fibroblasts were grown to confluence in 24-well plastic cluster plates in medium containing 10% FBS. Monolayers were shifted to serum-free medium for the final 24 h of incubation. IL-1␤ and the other test compounds were added at the times and concentrations indicated in the figure legends. Medium was removed from the cultures, and the monolayers were covered with phosphate-buffered saline (PBS) in the presence of the respective agents for the final 30 min of the treatment period. PBS was collected quantitatively, clarified by centrifugation, and subjected to PGE 2 radioimmunoassay using a commercially available kit (Amersham Biosciences).
Transient Transfection of Orbital Fibroblasts with Plasmids Containing mPGES and PGHS-2 Promoters and DN Mutant p38 and ERK-For studies involving the transient transfection of human fibroblasts, cultures were allowed to proliferate to 80 -90% confluence in medium containing 10% FBS. With regard to assessment of promoter activities, a 510-bp fragment spanning Ϫ538 to Ϫ28 of the putative mPGES promoter was cloned with the Human GenomeWalker kit (CLONTECH, Palo Alto, CA) according to the instructions of the supplier. Two reverse primers used for the PCR reactions included 5Ј-CGCAGCTCAACTGTGGGTGTGATC-3Ј and 5Ј-GTGATCAGCTCGA-CAGAGGAGCAG-3Ј. The amplified fragment was sequenced and subcloned from pCR2.1-TOPO vector (Invitrogen) into a promoter-less pGL2-luciferase vector (Promega, Madison, WI). With regard to the human PGHS-2 promoter, a plasmid designated Ϫ1800pGL2, containing Ϫ1840 to ϩ123 and thus located five base pairs upstream from the ATG of the human PGHS-2 promoter, was used. Promoter constructs were transiently transfected into fibroblasts using the LipofectAMINE PLUS system (Invitrogen). 0.75 g of pGL2 promoter DNA and 0.1 g of pRL-TK vector DNA (Promega), serving as a transfection efficiency control, were mixed with PLUS reagent for 15 min before being combined with LipofectAMINE PLUS for another 15 min. The DNA-lipid mixture was added to culture medium of 80% confluent cells for 3 h at 37°C. Dulbecco's modified Eagle's medium containing 10% FBS replaced the transfection mixture overnight. Transfected cultures were then serum-starved, and some received either IL-1␤ (10 ng/ml) for 2 h or nothing (control) as indicated in the figure legends. Cellular material was harvested in buffer provided by the manufacturer (Promega) and stored at Ϫ80°C until assayed. Luciferase activity was monitored with the Dual-Luciferase Reporter Assay System (Promega) in an FB12 tube luminometer (Zylux). Values were normalized to internal controls, and each experiment was performed at least three times.
To interrupt the expression of potentially relevant signaling pathway components, DN constructs for p38 and ERK1 were ligated into pcDNA3.1 (Invitrogen). These were transiently transfected into cells as described above. Control cultures received a constant amount (2 g) of empty vector DNA. The diminished levels of the kinases were documented by Western blotting an aliquot of the lysate with relevant antibodies.

IL-1␤ Up-regulates PGE 2 Production in Orbital Fibroblasts; This Is Associated with Induction of PGHS-2 and mPGES
Proteins-Confluent orbital fibroblast monolayers were shifted to serumless medium without or with IL-1␤ (10 ng/ml) for up to 48 h. As the data in Fig. 1A suggest, the cytokine elicits a dramatic increase in PGE 2 levels, which are 90-fold above control levels at 6 h, remain near peak values for 24 h, and then begin to decline at 48 h. Western blot analysis of cellular proteins from fibroblasts treated under identical conditions reveal a large, time-dependent induction of mPGES and PGHS-2. Neither is expressed at detectable levels under basal conditions. With regard to PGHS-2, the protein is detectable at  4. A, effect of IL-1␤ on the activities of human mPGES and PGHS-2 gene promoters in orbital fibroblasts. Cells, in this case from an individual with severe TAO, were seeded in 6-well plastic culture plates and allowed to proliferate to 70% confluence. They were then shifted to a mixture of LipofectAMINE PLUS with empty vector or the PGHS-2 or mPGES promoter/reporter constructs described under "Experimental Procedures" for 3 h. Some wells received IL-1␤ (10 ng/ml) for 2 h. Cellular material was harvested and luciferase activity determined. Data are expressed as the mean Ϯ S.D. of three replicates. B, effect of IL-1␤ on the disappearance of PGHS-2 and mPGES mRNAs in orbital fibroblasts. Cultures were allowed to proliferate to confluence and then they were treated with IL-1␤ (10 ng/ml) for 3 h. Monolayers were washed and incubated in complete growth medium for 4 h, the cultures were then shifted to medium containing DRB (20 g/ml) without or with IL-1␤ for the duration of time indicated along the abscissas. Cultures were harvested and the RNA subjected to Northern blot hybridization with cDNA probes for PGHS-2 or mPGES. These signals were normalized with their respective GAPDH levels. and do not reach an apparent maximum until 24 h. Unlike PGHS-2, mPGES protein levels remain close to their peak at 48 h. In contrast, levels of cPGES and PGHS-1 are not altered as a consequence of IL-1␤ treatment.
The effects of IL-1␤ on PGE 2 synthesis, the expression of PGHS-2 and mPGES are dose-dependent, as the data in Fig.  1B suggest. A near maximal effect on PGHS-2 expression and PGE 2 production is achieved at a concentration of 1 ng/ml. That concentration yields a suboptimal induction of mPGES. IL-1␤ at 10 ng/ml, the highest concentration of the cytokine used, results in a considerably higher level of mPGES.
The induction of both mPGES and PGHS-2 by IL-1␤ was susceptible to blockade by the synthetic glucocorticoid, dexamethasone (10 nM, Fig. 2). That concentration of dexamethasone is associated with a high fractional occupancy of the nuclear glucocorticoid receptor and near maximal effects on human fibroblast metabolism (35). The steroid had no effect on basal enzyme expression. The blockade of IL-1␤-dependent enzyme expression was accompanied by a substantial inhibition of cytokine-provoked PGE 2 production.
Induction of PGHS-2 and mPGES by IL-1␤ Is Mediated at the Pre-translational Level-Northern blot analysis was employed to determine the relationship in orbital fibroblasts between the induction of PGHS-2 and of mPGES mRNAs. Confluent cultures were shifted to medium without FBS overnight, and then IL-1␤ (10 ng/ml) was added at various times prior to monolayer harvest. As the Northern blot shown in Fig. 3A indicates, the transcript encoding PGHS-2 is not detectable under basal culture conditions but was induced as a single, 4.8-kb mRNA with the addition of IL-1␤ within 6 h. At 12 h, the mRNA levels were at least 100-fold above control levels. At 24 h, PGHS-2 mRNA levels had dropped substantially and were again undetectable at 48 h. When the membrane was rehybridized with an mPGES cDNA probe, a 2-kb transcript was apparent under control conditions and was strongly upregulated with IL-1␤, an effect that reached a maximum at 12 h, when it was ϳ7-fold above control levels. The induction was partially sustained for 48 h, the duration of the study, when it remained increased by 2.7-fold. Steady-state levels of cPGES and PGHS-1 mRNA were relatively constant following the addition of IL-1␤ to the culture medium (Fig. 3A). The former is expressed on Northern blot analysis as a single band. PGHS-1 mRNA migrates as an ϳ5.2-kb transcript, which is consistent with the pattern found previously in monocytes (36) and endothelial cells (37) and differs from the predominant 2.8-b mRNA found in some other human cells (5). Thus the relative levels of their respective mRNAs mirrored the pattern of protein induction provoked by cytokines.

The Up-regulation of mPGES by IL-1␤ Represents a Primary Induction That Is Not Dependent upon Intermediate Protein
Synthesis-We have reported previously that the induction of PGHS-2 mRNA by the proinflammatory cytokine leukoregulin is partially dependent on ongoing protein synthesis (25). We next determined whether the effects of IL-1␤ on mPGES and PGHS-2 mRNA levels were altered by an inhibition of protein synthesis created by cycloheximide (10 g/ml). This inhibitor concentration blocks Ͼ95% of protein synthesis in human fibroblasts (35). As the Northern blot in Fig. 3B clearly indicates, the inhibitor failed to attenuate the induction of mPGES after 6 h of IL-1␤ treatment. The inhibitor did down-regulate by ϳ78% the induction of PGHS-2 mRNA by IL-1␤. Thus, although the up-regulation by IL-1␤ of mPGES mRNA apparently represents a primary inductive event, the increase in PGHS-2 is dependent, at least in part, on the induction of an intermediate protein(s). genes were transiently transfected into orbital fibroblasts, and then cultures were incubated without or with the cytokine. As the data contained in Fig. 4A suggest, the PGHS-2 promoter exhibited considerable activity under basal conditions, at least 20-fold more luciferase activity than that found in controls transfected with the empty vector. IL-1␤ (10 ng/ml) increases the activity of PGHS-2 promoter by ϳ2-fold after 2 h of treatment. The up-regulation is consistent with our earlier finding that PGHS-2 gene transcription, as assessed by nuclear run-on assays, was enhanced 2-fold by leukoregulin (25). The mPGES promoter exhibits considerably less basal activity in orbital fibroblasts. IL-1␤ increases the activity of this promoter by ca 2.5-fold after 2 h of treatment. This response of the mPGES promoter is consistent with the modest effects of IL-1 observed in A549 human alveolar cells transfected with similar promoter constructs (38).

Induction of mPGES and PGHS-2 by IL-1␤ Involves the Modest Up-regulation of Their Respective Gene Promoters-The
The effect of IL-1␤ on the PGHS-2 and mPGES promoters was relatively modest. This fractionally small effect on PGHS-2 was unexpected because of the large induction of steady-state PGHS-2 mRNA levels observed (Fig. 3A). The 2-fold increase in mPGES promoter activity following IL-1␤ treatment was also less than anticipated, given the 7-fold increase in mPGES mRNA. We therefore next examined the impact of the cytokine on the decay of the respective hybridizable mRNAs. As the data from Northern blots in Fig. 4B indicate, under conditions without IL-1␤, PGHS-2 mRNA decays rapidly, and the cytokine dramatically retards the decay of the PGHS-2 transcript following transcriptional blockade with DRB (20 g/ml). The apparent t1 ⁄2 of PGHS-2 mRNA under culture conditions in the absence of IL-1␤ is around 1 h, and the mRNA levels have fallen 80% by 1.5 h. This instability is consistent with its behavior in a number of cell types (9). The disappearance of PGHS-2 mRNA was dramatically decreased following treatment with IL-1␤ (10 ng/ml), and levels remained constant for 5 h. mPGES mRNA was far more stable than that of PGHS-2 in control orbital fibroblasts not treated with the cytokine. The data in Fig. 4B demonstrate a t1 ⁄2 of ϳ6 h in untreated cultures. The addition of IL-1␤ failed to further enhance the survival of this relatively long-lived transcript.
The Magnitude of mPGES and PGHS-2 Induction in Orbital Fibroblasts by IL-1␣, IL-1␤, and CD154 Is Considerably Greater than That Elicited by Other Cytokines-We next determined whether the dramatic induction of mPGES and PGHS-2 by IL-1␤ in orbital fibroblasts could be seen following exposure to other cytokines implicated in human autoimmune disease. Confluent cultures were treated with TNF-␣ (10 ng/ml), TGF-␤ (5 ng/ml), IL-4 (10 ng/ml), IL-1␣ (10 ng/ml) as well as IL-1␤ for 16 h. The effects of IL-1␣, IL-1␤, and CD154 on both mPGES and PGHS-2 expression are substantially greater than those of the other cytokines tested (Fig. 5). Especially surprising is the apparent lack of response to TNF-␣, considering the relatively large induction of PGHS-2 and PGE 2 found in other cell types. IL-4 exerts modest effects on PGE 2 production in orbital fibroblasts (25,39) but the cytokine failed to up-regulate either mPGES or PGHS-2 in the current studies. These fibroblasts display high levels of CD40, the receptor for CD154 (40). Cao et al. (31) have demonstrated that PGHS-2 expression can be induced by activation of the CD40/CD154 bridge. Cultures pretreated with interferon ␥ (100 units/ml) respond to recombinant human CD154 with a sizable induction of both mPGES FIG. 6. A, IL-1␤ activates p38 and ERK 1/2 MAP kinases in orbital fibroblast cultures in a time-dependent manner. Confluent cultures were shifted to medium without FBS for at least 16 h and were then treated with IL-1␤ (10 ng/ml) for the indicated times. Cell lysates were harvested and subjected to SDS-PAGE, transferred to membranes, and Western blotted with phospho-specific antibodies against p38 and ERK. Films were subjected to densitometric analysis, the results of which are shown. B, effect of specific inhibitors of MAP kinases on the induction by IL-1␤ of PGHS-2 and mPGES. Cultures were shifted to medium without FBS for 24 h and then incubated for 16 h without or with IL-1␤, alone or in combination with PD98059 (10 M) or/and SB203580 (10 M). Cell lysates were collected and subjected to Western blot analysis with specific antibodies against PGHS-2 and mPGES. The resulting bands were analyzed for relative densities. and PGHS-2 proteins after 12 h (Fig. 5). The action of CD154 on prostanoid generation in orbital fibroblasts has been shown to result from an intermediate induction of IL-1␣ induction (31). FBS was also found to induce these enzymes when fibroblasts were incubated under reduced serum conditions and were then shifted to medium with 10% serum. A 20-fold induction of PGHS-2 and 4-fold up-regulation of mPGES expression resulted from serum treatment (data not shown). Thus, it would appear that agents found to provoke the expression of PGHS-2 also enhance the levels of mPGES in orbital fibroblasts. Moreover, mPGES, like PGHS-2, appears to represent a proximate target for proinflammatory signals derived from T cells and conveyed through the CD40/CD154 activational bridge.

IL-1␤ Activates ERK and p38 MAP Kinases in Human Orbital Fibroblasts: The Induction of PGHS-2 and mPGES Is Dependent on the Activities of ERK and p38 MAP Kinases-
Studies examining the signal transduction pathways utilized by proinflammatory cytokines in the induction of PGHS-2 suggest that multiple pathways are involved. Notable among the pathways thus far implicated are the MAP kinase pathways (41,42). Nothing has thus far been reported about the signaling involved in the activation of mPGES expression by cytokines. Thus, we next compared the signaling pathways utilized in the activation of PGHS-2 and mPGES by IL-1␤ in orbital fibroblasts. The cytokine rapidly activates both p38 and ERK 1/2 MAP kinases (Fig. 6A). These effects are time-dependent and are sustained for at least 48 h. We treated orbital fibroblasts with specific kinase inhibitors in combination with IL-1␤ to determine whether the MAP kinase pathway was utilized in the induction of mPGES. When the activity of the p38 MAP kinase pathway was interrupted with SB203580 (10 M), the induction of PGHS-2 and mPGES by IL-1␤ was attenuated by 55 and 80%, respectively (Fig. 6B). PD98059 (10 M), a specific inhibitor of MEK, which is immediately up-stream from ERK, also blocked the induction of PGHS-2 and mPGES by 33 and 65%, respectively. These inhibitor concentrations have been shown to specifically inhibit their respective target pathways (43,44). When the two inhibitors were added together, the induction of both PGHS-2 and mPGES was blocked further. The combination of compounds reduced the respective induction by 83 and 89%.
An alternate, molecular approach was then employed for interrupting the p38 and ERK MAP kinase pathways by transiently transfecting DN constructs for these kinases into orbital fibroblasts. The induction of mPGES protein by IL-1␤ is attenuated partially following transfection with either the p38 or ERK kinase DN constructs (Fig. 7). Control cultures were transfected with an empty vector. When both DN constructs are co-transfected into the same cultures, a near complete blockade of the mPGES induction can be appreciated. Thus, it would appear that both p38 and ERK kinase signaling pathways are involved in the induction of mPGES by IL-1␤ in orbital fibroblasts. This pathway utilization overlaps that involved in the induction of PGHS-2. Congruent results were obtained with pharmacological and molecular strategies for pathway interruption.
Inhibition of PGHS-2 Activity Is Associated with a Decrease in IL-1␤-provoked mPGES Expression-A functional link between PGHS-2 and mPGES was implied by the earlier results of studies in HEK293 cells, where cDNAs encoding the two enzymes were co-transfected (13). A marked increase in PGE 2 production occurred when a low concentration of exogenous arachidonate (2-5 M) was added to the culture medium. When PGHS-1 was co-transfected with mPGES, increased PGE 2 production occurred only at a relatively high concentration of arachidonate (10 M) (13). We thus determined whether a functional relationship between endogenous PGHS-2 and mPGES exists in orbital fibroblasts. To ascertain whether the activity of PGHS-2 was directly influencing the expression of mPGES, fibroblasts were treated with IL-1␤ (10 ng/ml) alone or in combination with SC58125 (5 M) for 16 h. As the data in Fig. 8A indicate, IL-1␤ increases the levels of mPGES protein by 13-fold. Addition of the PGHS-2 inhibitor resulted in a 92% attenuation of that induction. Moreover, SC58125 also partially blocked the induction by IL-1␤ of mPGES mRNA (Fig.  8B). The magnitude of this inhibition was 70%. We have reported previously that SC58125, at comparable concentrations to those used in the current studies, inhibits cytokine-dependent PGE 2 production in orbital fibroblasts (25,26). To determine whether SC58125 exerts its effect on mPGES expression directly through the inhibition of PGHS-2 activity, exogenous FIG. 7. The effect of transfected DN p38 and ERK1 expression vectors on the induction of mPGES by IL-1␤ in orbital fibroblasts. 80 -90% confluent cultures were shifted to serum-free medium containing LipofectAMINE plus and the plasmid DNA constructs (2 g each) for 3 h as described under "Experimental Procedures." Control cultures received empty vector DNA. Cultures were shifted to medium supplemented with 10% FBS for 24 h and then to serum-free medium for 16 h. Subsequently, some plates received IL-1␤ (10 ng/ml) for an additional 16 h. The resulting monolayers were solubilized and subjected to Western blot analysis with anti-mPGES primary antibody. The right panel demonstrates the analysis of cultures transfected with empty vector, DN p38, or DN ERK followed by Western blotting with anti-p38 or ERK. arachidonate (10 M) was added to IL-1␤-treated cultures also receiving SC58125. Exogenous arachidonate partially restored the cytokine-provoked mPGES expression (Fig. 8A). The downregulation of mPGES expression resulting from SC58125 treatment (92%) was reduced to a 36% inhibition. This result suggests strongly that SC58125 acts on mPGES expression through its effects on PGHS-2 activity.

DISCUSSION
The up-regulation of PGE 2 synthesis by IL-1␤ in human orbital fibroblasts involves the coordinate induction of multiple enzymes in the prostanoid biosynthetic pathway. The levels of PGHS-2, an enzyme that catalyzes two rate-limiting intermediate reactions that ultimately yield PGH 2 , and mPGES, which facilitates the terminal reaction converting PGH 2 to PGE 2 (1), are both increased. Induction of the two enzymes occurs sequentially, with PGHS-2 mRNA and protein becoming elevated above base line more rapidly and transiently than those of mPGES. The susceptibility of cytokine-provoked mPGES induction to a highly selective inhibitor of PGHS-2 activity suggests that up-regulation of the former enzyme by IL-1␤ is dependent, at least in part, on a product of PGHS-2. This is further supported by the partial restoration of IL-1␤-dependent mPGES expression with high concentrations of exogenous arachidonate. Thus, it would appear that mPGES expression is linked functionally in the orbital fibroblast with the activity of PGHS-2. The current findings regarding the functional interaction of endogenously expressed mPGES and PGHS-2 are entirely consistent with previous findings in transfected cells over-expressing both enzymes (13). Those earlier studies indicate a preferential coupling of mPGES to PGHS-2. In addition, the level of arachidonate generated and its distribution among cellular compartments might also influence the efficiency with which PGHS-2 and mPGES interact (15). The dramatic downregulation of PGHS-2 and mPGES in these fibroblasts by physiologically relevant concentrations of dexamethasone would suggest that both enzymes are targets of steroid regulation. Thus, the impact of glucocorticoids on prostanoid synthesis is complex and apparently involves multiple levels of control.
The up-regulation by IL-1␤ of the steady-state levels of mPGES and PGHS-2 mRNAs are a consequence, at least in part, of modest increases in the activities of their respective promoters (Fig. 4A). With regard to mPGES, this small increase (2-3-fold) is consistent with previously reported experience in A549 cells (38), where the promoter activity was only minimally influenced. With regard to PGHS-2, Wang et al. (25) reported that the proinflammatory cytokine leukoregulin increases steady-state levels of PGHS-2 mRNA dramatically in orbital fibroblasts, but the increase in PGHS-2 gene transcription, as assessed by nuclear run-on assays, was modest. Of considerable potential importance to the induction of PGHS-2 by IL-1␤ in these cells is the impact that the cytokine appears to exert on the stability of the mRNA (Fig. 4B). Our data clearly demonstrate a dramatic enhancement of PGHS-2 mRNA stability and are consistent with the structure of the 3Ј-untranslated region of human PGHS-2. That sequence contains a number of AUUUA instability elements, and the transcript has been shown to exhibit rapid turnover in a number of cell types (9,45). mPGES mRNA exhibits considerably greater stability and treatment of orbital fibroblasts with IL-1␤ failed to alter the survival of this already long-lived mRNA. The sustained induction of mPGES mRNA following IL-1␤ treatment (Fig. 3) thus can be explained on the basis of a small increase in gene transcription in the setting of a relatively stable mRNA.
During the course of these studies, a report appeared from Thorén and Jakobsson (46) demonstrating that PGES activity in A549 cells could be inhibited by sulindac sulfide and the PGHS-2 selective agent, NS-398, with IC 50 values of 80 and 20 M, respectively. In contrast, the IC 50 for NS-398 on PGHS-2 is ϳ3.8 M (47). These findings are of considerable interest because they suggest structural similarities in the drug-binding sites on PGHS-2 and PGES. Coupled with our current observations that selective inhibition of PGHS-2 activity also interferes with mPGES expression, it is possible that many partic- The monolayers were then harvested and subjected to Western blot analysis with primary anti-mPGES antibody. B, confluent cultures were shifted to serumless medium overnight. Some plates were treated with IL-1␤ (10 ng/ml) alone or in combination with SC58125 (5 M) for 16 h. RNA was harvested and subjected to Northern blot hybridization. The membrane was rehybridized with a GAPDH probe, and the signals were used to normalize the levels.
ularly effective anti-inflammatory drugs may target multiple enzymes in the PGE 2 biosynthetic cascade.
The precise roles for PGE 2 in the initiation and evolution of the inflammatory response and in tissue fibrosis are uncertain, and the topics are open to considerable debate. Clearly the insinuation of dramatic PGE 2 production into the repertoire of the orbital fibroblast phenotype suggests a great potential for this prostanoid in conditioning the immune reactivity of the orbit, both in healthy and pathological states, such as those occurring in TAO. The capacity of resident fibroblasts to generate high levels of PGE 2 following activation by recruited bone marrow-derived cells could impact substantially on immune responses occurring there and could direct the pattern of tissue remodeling. This is particularly relevant with regard to the potential for fibrotic reactions in orbital muscles in late stage TAO, because that process, associated with substantial morbidity, has been thought to be the consequence of T H 2 predominance (48). Thus, our current finding that mPGES is also highly inducible in orbital fibroblasts defines an additional component of the inflammatory machinery that might be exploited as a potential therapeutic target.
From the current studies, it would appear that IL-1␤ utilizes overlapping signaling pathways with regard to the induction of PGHS-2 and mPGES expression in orbital fibroblasts. Although nothing had been reported previously concerning signal transduction related to the mPGES expression provoked by cytokines, an extensive literature has evolved suggesting a complex array of pathway utilization up-stream from PGHS-2 gene activation in several cell types. Prominent among these are the p38 and ERK MAP kinases (43,44,49). We have reported that CD40 engagement with CD154 leads to the induction of PGHS-2 and that interruption of the ERK pathway substantially attenuates this cellular response in orbital fibroblasts (31). In the current studies, the addition of specific inhibitors of both p38 and ERK could partially attenuate the induction of PGHS-2 by IL-1␤. This was also true with regard to the induction of mPGES. Similar results were obtained when each of the two kinase pathways was interrupted by the transfection of DN constructs. The co-transfection of both DN plasmids led to a further and additive blockade of the induction. Taken together, it would appear that signaling from IL-1␤ to mPGES in orbital fibroblasts involves multiple pathways. This insight defines potential therapeutic targets for the interruption of PGE 2 -related inflammation in the orbit. Moreover, an overlap in cell signaling that culminates in the activation of both PGHS-2 and mPGES expression suggests that potentially important coordination exists between the enzymes. This association is further supported by the finding that cytokines inducing PGHS-2 also up-regulate mPGES expression.
These results demonstrate a substantial capacity of IL-1␤ to induce multiple enzymes in the synthetic cascade for PGE 2 in orbital fibroblasts. Why these cultures should exhibit greater responses to proinflammatory cytokines than do other types of fibroblasts (25,26) is not clear, but it could be related to the normal function of orbital connective tissue. What emerges from our current findings is a further clarification of the molecular basis underlying the particularly robust involvement of orbital connective tissue in autoimmune inflammation, such as that associated with TAO.