Regulation of Cyclooxygenase-2 Expression in Monocytes by Ligation of the Receptor for Advanced Glycation End Products*

Cyclooxygenase-2 (COX-2) enzyme and its inflammatory products such as prostaglandin E 2 (PGE 2 ) have been implicated in the pathogenesis of several inflammatory diseases. However their role in diabetic vascular disease is unclear. Advanced glycation end products (AGEs) act via their receptor, RAGE, to play a major role in diabetic complications. In this study, we investigated the effect of AGEs and S100b, a specific RAGE ligand, on the expression of COX-2 and the molecular mechanisms involved in cultured THP-1 monocytes and human peripheral blood monocytes. S100b treatment of THP-1 cells led to a significant 3–5-fold induction of COX-2 mRNA ( p < 0.001). COX-2 protein and its product PGE 2 were also increased, whereas COX-1 expression was un-affected . In vitro prepared AGE also induced COX-2 mRNA. S100b-induced COX-2 mRNA was blocked by an anti-RAGE antibody and by inhibitors of NF- (cid:1) B (Bay11-7082), oxidant stress, protein kinase C, ERK, and p38 MAPKs. S100b (4-h treatment) significantly increased transcription from a human COX-2 promoter-luciferase construct (4-fold, p < 0.001). Promoter deletion analyses and inhibition of transcription by an NF- (cid:1) B superre-pressor mutant confirmed NF- (cid:1) B involvement. This was further supported by inhibition of S100b-induced PGE 2 by Bay11-7082. glucose treatment of endothelial cells Serial deletion mutants demonstrated the the (cid:2) 860-bp flanking region containing the cis-acting elements cAMP-re- sponse element, two NF-IL-6 sites, and two NF- (cid:3) B sites for basal promoter activity ( ). Co-transfection of a I (cid:3) B mutant plasmid, along with the (cid:2) 860 deletion mutant, showed a significant inhibition in S100b-induced COX-2 promoter activation. This, along with the loss of promoter activity in the (cid:2) 360 deletion construct, indicates the importance the distal NF- (cid:3) (at (cid:2) (cid:2) in S100b-induced COX-2

induce COX-2 expression (12). Although a recent study demonstrated that high glucose can induce IL-1␤ in human pancreatic islets (32), the effects on COX-2 were not examined. High glucose (HG) was shown to enhance IL-1␤-induced COX-2 expression in vascular smooth muscle cells (33), and a very recent report showed that high glucose treatment of endothelial cells increased COX-2 expression and decreased nitric oxide availability (34). The administration of the selective COX-2 inhibitor NS-398 could prevent the onset of diabetes in mice (35). Elevated levels of COX-2 protein and its product PGE 2 were found in the spinal cord of diabetic rats (36). However, very little is known regarding the potential involvement of COX-2 in diabetic vascular complications, diabetic atherosclerosis, or the regulation of COX-2 in monocytes under diabetic conditions. This is a significant issue, especially because evidence now indicates the importance of inflammation in the pathogenesis of atherosclerosis and diabetes. Furthermore, the mechanisms of altered monocyte prostaglandin production in diabetes are largely unknown. Our present studies provide new information in this connection.
Studies with inflammatory cells such as monocytes demonstrate that simulated diabetic conditions in vitro, such as high glucose culture conditions or treatment with advanced glycation end products (AGEs), can induce the expression of inflammatory cytokine and related genes via activation of specific signaling pathways and transcription factors such as NF-B (37)(38)(39)(40)(41). This could then result in increased monocyte activation, migration, and adhesion to the endothelium. In the present study, we evaluated the hypothesis that AGEs acting via their receptor, RAGE, can lead to increased COX-2 gene expression and activity and subsequent monocyte activation.
AGEs are products of non-enzymatic glycation/oxidation of proteins/lipids that accumulate during natural aging and are also greatly augmented in disorders such as diabetes, renal failure, and Alzheimer's disease (42)(43)(44)(45). Formation of AGEs or glycooxidation products is related to circulating high glucose concentrations in diabetes. Several receptors for AGEs have been identified on vascular, renal, and other cells (46). The well studied cell surface receptor for AGE, namely RAGE, is a multiligand member of immunoglobulin superfamily (45)(46)(47)(48). Ligands for RAGE include AGEs, EN-RAGE, the S100/calgranulin family of proteins, amphoterin, amyloid ␤Ϫpeptide (49,50), and carboxymethyl lysine adducts of protein (47). Several short peptides including EN-RAGE and S100b belonging to S100/calgranulin family signal through RAGE and can produce key pro-inflammatory mediators in endothelium, mononuclear phagocytes, and lymphocytes (49,50) These peptides now serve as valuable tools in the study of RAGE signaling.
Interaction of these ligands with RAGE can lead to the generation of oxidant stress, production of growth factors and cytokines, chronic inflammatory responses, and cellular and vascular dysfunction associated with diabetic complications (49,50). Blockade of RAGE can suppress the inflammatory response in murine models, diabetic vascular hyperpermeability, and diabetes-induced accelerated atherosclerosis in apoE null mice (50 -52). The proximity of cells expressing RAGE to lesional areas rich in AGEs suggest that AGE-RAGE interaction can trigger key cell signal transduction pathways and thereby lead to chronic cellular activation and dysfunction associated with diabetes. AGE ligation of RAGE activates oxidant stress, p21 ras, and downstream targets such as mitogen-activated protein kinases (MAPKs) and leads to the activation of transcription factors such as NF-B (37,45,48,53).
Although AGE-RAGE interaction has been implicated in inflammatory responses, the effects on COX-2 expression and regulation are not known. In the present studies, we demonstrate for the first time that RAGE ligation by exogenous AGEs or by S100b can lead to potent increases in COX-2 expression and activity in THP-1 monocytic cells, as well as in primary human blood monocytes. COX-2 induction at 4 h was regulated transcriptionally and involved key NF-B elements on the COX-2 promoter. These results suggest that simultaneous production of cytokines and COX-2 products by HG and AGEs in a diabetic environment can trigger an amplifying inflammatory loop and lead to accelerated vascular complications.
Preparation of Methylglyoxal Modified Albumin (AGE)-Methylglyoxal-modified albumin (40) (AGE) was prepared as follows: bovine serum albumin (fraction V; fatty acid-free, low endotoxin) (10 mg/ml) was incubated with 100 M methylglyoxal (Sigma) in PBS containing 0.1% sodium azide and 1 mM phenylmethylsulfonyl fluoride at 37°C for 7 days. After incubation, the free methylglyoxal was removed by dialysis against PBS at 4°C for 24 h. The methylglyoxal modified albumin was then vacuum-dried and dissolved in PBS at a concentration of 10 mg/ml. Aliquoted samples were stored at Ϫ70°C.
Isolation of Human Peripheral Blood Monocytes (PBMC)-50 -60 ml of blood from adult volunteers with established Type 1 diabetes and also from normal healthy donors were collected in the presence of anticoagulant in accordance with an approved Institutional Review Board protocol. The blood was diluted with equal volumes of PBS. An equal volume of diluted blood was overlaid on Ficoll-Paque-plus in 1:1 ratio and centrifuged at 400 ϫ g for 20 -30 min at 18 -20°C. The leukocyte population was collected from the interface and washed with PBS several times to remove plasma and Ficoll. About 50 million washed cells in 10 ml of RPMI medium containing 10% FCS were plated in 100-mm culture dishes to allow monocytes to adhere on the surface of the dish for 2-3 h. The non-adherent cells (mainly lymphocyte population) were removed, washed with fresh medium, cultured in RPMI medium, and labeled as PBLC. Attached monocytes were washed twice with warm RPMI medium containing 10% FCS and allowed to remain in the dish overnight at 37°C in 5% CO 2 . During this period the monocytes detach from the dish. They were collected and washed in fresh RPMI medium and labeled as PBMC. Then, about 1 ϫ 10 5 cells from normal volunteers per well in 6-well plates were treated with S100b for 4 h, and total RNA was isolated as described below. Monocytes from Type 1 diabetic patients were directly processed for RNA extraction.
RNA Preparation and Relative RT-PCR-THP-1 cells (2 ϫ 10 6 /sample) in 4 ml of medium containing 5.5 mM (NG) with or without 6.5 g/ml S100b were cultured in duplicate in 6-well dishes for various time intervals. Total RNA was isolated by the RNA-STAT method, and 1 g was used for the RT reaction using a Gene Amp RNA PCR kit (41). cDNA corresponding to 0.05 g of RNA was then used in multiplex PCR reactions containing gene-specific primers paired with Quantum RNA 18 S internal standards, and the multiplex PCR reactions were performed for 25-35 cycles in a GeneAmp9700 machine (Applied Biosystems, Inc., Foster City, CA).
RNase Protection Assay (RPA)-RPA was performed using RPA III TM kit according to instructions of the manufacturer (Ambion Inc., Austin, TX) with some modifications. COX-2 antisense riboprobe was prepared by cloning RT-PCR amplified COX-2 fragment into pGEM-T vector (Promega) and verified by sequencing. Antisense template was amplified using SP6 promoter primer as the 5Ј primer 5ЈGCTATTTA-GGTGACACTATAGAA3Ј and COX-2 primer 5ЈCACCAGGCAAATTG-CTGGCAG3Ј as the 3Ј primer. COX-2 antisense riboprobe was generated using SP6 RNA polymerase, [ 32 P]UTP, and PCR-amplified COX-2 DNA fragments as template. Antisense riboprobe of 18 S was generated using SP6 RNA polymerase and linearized plasmid pTRIS RNA 18 S (Ambion). Total RNA (150 g) isolated from control and S100b-treated THP-1 was hybridized with 32 P-labeled antisense riboprobes of human COX-2 (321 bp) and 18 S (165 bp). The hybridization product was then treated with RNase A and T to remove unhybridized probes. The protected and hybridized mRNAs were resolved in 6%-8 M urea polyacrylamide gel. The gel was dried and exposed on x-ray film at Ϫ80°C. The bands were visualized by autoradiography.
Western Blot Analysis-Cells were washed twice with 5 ml of Hanks' salt solution and then lysed with 0.1 ml of lysis buffer. The cell lysates were assayed for protein by the Bio-Rad Dc protein assay kit (Bio-Rad). Proteins were separated by electrophoresis on SDS-PAGE gels with 10% acrylamide. Proteins were transferred to Immobilon P membranes by semi-dry transfer and then subjected to immunoblotting as described earlier (38) using relevant antibodies. Detection of immunoreactive bands was by chemiluminescence. PGE 2 Enzyme Immunoassay-THP-1 cells (5 ϫ 10 5 cells/ml) were incubated in 6-well tissue culture plates in RPMI 1640 medium with 0.2% BSA. Cells were treated with or without S100b for 8 h. The supernatant conditioned medium was then harvested and assayed for PGE 2 levels using a specific EIA kit according to the manufacturer's instructions (Cayman Chemical, Ann Arbor, MI). Medium alone without cells was incubated under the same conditions and used as blank control for the EIA.
Plasmid Construction-Construction of the deletion mutants containing specific regions of the human COX-2 gene promoter in the luciferase reporter vector pGL3 Basic (Promega) was accomplished by PCR amplification. Plasmid hCOX-2 (Ϫ1437/ϩ127) was generated by deleting a 5673-bp DNA fragment from upstream of the recombinant plasmid containing firefly luciferase gene under the control of the ϳ7273-bp promoter region of the human COX-2 gene (generous gift from Dr. Thomas McIntyre, University of Utah, Salt Lake City, UT) by digesting with restriction enzyme KpnI. The resulting plasmid was re-ligated using T4-DNA ligase, yielding the luciferase gene under the control of Ϫ1437/ϩ127 human COX-2 promoter region. Deletion constructs containing various promoter regions of hCOX-2, namely from Ϫ860 to ϩ127, Ϫ360 to ϩ127, Ϫ218 to ϩ127, Ϫ123 to ϩ127, and from Ϫ52 to ϩ127, were generated using the recombinant plasmid containing an ϳ7-kb promoter region of the human COX-2 gene as template.
The following primers were used: upstream primers, from Ϫ860, 5ЈG-GTACCCACATTAACTATTTACAG3Ј; from Ϫ360, 5Ј GGTACCCCAAG-GCGATCAGTCCAG3Ј; from Ϫ218, 5ЈGGTACCTACCCCCTCTGC TCC-CAA3Ј; from Ϫ123, 5ЈGGTACCTTTTTTAAGGGGAGAGG3Ј; from Ϫ52, 5Ј GGTACCCATGGGCTTGGTTTTC3Ј; downstream primer ϩ127, 5Ј AAGCTTCGGGCAGGGCGCGGCGC 3Ј. All upstream PCR primers contained KpnI restriction sites (underlined), and the downstream primer contained a HindIII recognition site (underlined), which forced cloning of the fragments in the desired orientation into the pGL3 Basic vector. Orientation and sequence of all constructs were verified by direct sequencing using the ABI PRISM 377 DNA sequencer.
DNA Transfection and Luciferase Assays-THP-1 cells plated in 6-well plates (1.2 ϫ 10 6 per well) were transfected with 1 g of the indicated COX-2 promoter plasmids, the control pGL3-Luc plasmid (Invitrogen), or the pCMV-mIB plasmid (41, 55) (generous gift from Dr. E. Zandi, University of Southern California) using LipofectAMINE 2000 in RPMI medium with serum according to manufacturer's protocols. Following the overnight recovery period, the transfected cells were cultured in medium containing 5.5 mM (NG) with or without 6.5 g/ml S100b. Cells were then washed with PBS, lysed with 100 l of lysis buffer, and stored overnight at Ϫ70°C. Samples were thawed, brought to room temperature, and 20 l of each lysate was used to analyze luciferase activity by the luciferase assay system according to the manufacturer's instructions.
THP-1 Cell Binding Assay-After 4 h of S100b treatment, THP-1 cells (10 6 cells/ml) were washed with Hanks' balanced salt solution twice and incubated with 10 g/ml of BCECF/AM fluorescent tag in PBS for 30 min at 37°C to label the cells and washed twice with serum-free medium. In some experiments, cells were pretreated for 1 h with various inhibitors as indicated before treatment with or without S100b. About 80% confluent PVSMC or human VSMC in 24-well plates were washed with serum-free medium, and 5 ϫ 10 4 fluorescently labeled THP-1 cells were added to each well and incubated at 37°C for 1 h. Non-specifically bound cells were removed by carefully washing several times with serum-free medium. Specifically bound cells were lysed with lysis buffer (100 mM Tris-Cl, pH 8, and 0.1% Triton X-100) at room temperature for 1 h. Fluorescence densities in the lysates were determined at 485 nm with a fluorescence multi-well plate reader, f-max (Molecular Devices), and quantitated using SOFT-max PRO-f software.

S100b
Induces COX-2 and IL-1␤ mRNAs but Not COX-1 mRNA in THP-1 Cells-Our recent study showed that high glucose (HG) could induce the transcriptional regulation of chemokine and cytokine genes in monocytes via activation of specific signaling pathways (38,41). Because AGEs accumulate under diabetic conditions and contribute to the progression of diabetic complications, we hypothesized that ligation of RAGE by AGEs or S100b can lead to the expression of the inflammatory gene, COX-2. This could unravel additional pro-inflammatory consequences of RAGE ligation. We also evaluated whether AGE-induced COX-2 expression occurs directly or via increases in the expression of other proteins or cytokine such as IL-1␤.
We initially carried out experiments to evaluate whether S100b could lead to the induction of COX-2 and COX-1 mRNAs in THP-1 cells and the optimal conditions required for the same. THP-1 cells were treated with or without S100b protein for various time intervals from 0.5 to 24 h as shown in the Fig.  1. RNA extracted from these experiments was subjected to relative RT-PCR analyses. In this method, specific primers for human COX-2, COX-1, and IL-1␤ genes were paired with 18 S rRNA primers as internal standards in the multiplex RT-PCR reactions. Amplification of 18 S rRNA, in addition to the mRNA of interest, allows one to normalize samples for differences in loading across several samples. S100b-induced changes in gene expression were evaluated as -fold over control NG samples after normalizing to 18 S rRNA internal control. Results showed that S100b treatment led to a marked increase in COX-2 mRNA expression by 1 h, peaking at 2 to 4 h and declining by 16 to 24 h (Fig. 1A). In contrast, COX-1 mRNA levels showed no delectable change during the 24-h time period (Fig. 1B). This suggests that S100b specifically up-regulates the inducible COX-2 isoform and not COX-1 mRNA. (Fig. 1, A  and B).
We noted that S100b also increased IL-1␤ mRNA expression that was evident by 30 min and peaking at 2 h (Fig. 1C). Because IL-1␤ is a potent inducer of COX-2, to determine whether S100b-induced COX-2 mRNA at 4 h is via IL-1␤ expression, we evaluated the effects of pretreatment with an IL-1␤ neutralizing antibody or control IgG added to the medium. We noted that the IL-1␤ antibody only slightly attenuated the effects of S100b at 4 h, indicating that the effect of S100b, at least at 4 h, is not primarily mediated via IL-1␤. Because cytokines such as IL-1␤ and IL-1␣ have been shown in many instances to mediate COX-2 mRNA stability rather than transcription (24,26), we hypothesized that S100b could be inducing COX-2 mRNA via both transcription and mRNA stabilization via de novo protein synthesis, the latter possibly via a protein such as IL-1␤. We therefore examined this further at both 2-h (time of peak appearance of IL-1␤) and 4-h time points.
Increased COX-2 Expression by S100b at 4 h, but Not 2 h, Is Because of Transcriptional Regulation-Time course analyses showed that S100b-induced expression of COX-2 mRNA is evident by 1 h, peaking at 2 to 4 h and then declining to control levels by 24 h (Fig. 1). To determine whether S100b-induced COX-2 mRNA expression at 2 and 4 h is because of increases in transcription, THP-1 cells were pretreated for 1 h with actinomycin-D an inhibitor of transcription or with cycloheximide, a inhibitor of de novo protein synthesis, and then treated with S100b for 2 or 4 h. RT-PCR analyses in Fig. 2A showed that at the 2-h time point, actinomycin-D did not block the effects of S100b, but cycloheximide was clearly inhibitory. In contrast, at 4 h, actinomycin-D completely blocked S100b-induced COX-2 mRNA expression whereas cycloheximide had no inhibitory effect, in fact being slightly stimulatory. (Fig. 2B). These interesting results indicate that increased COX-2 expression by S100b at 4 h, but not 2 h, is because of increases in transcription. Furthermore, new protein synthesis is involved at the 2-h time point but not at 4 h. Interestingly, a similar kind of biphasic dual mode of COX-2 regulation involving transcription and de novo protein synthesis at different time periods has been reported recently (56) with another agonist. Our data were further confirmed by luciferase reporter assays as shown later.
It is possible that the increased COX-2 mRNA by S100b at 2 h may be because of cytokines that are initially produced by S100b prior to COX-2 mRNA induction, for example, by increased IL-1␤ expression. IL-1␤ is a well known inducer of COX-2, and quite often by mRNA stabilization and post-transcriptional mechanisms (23)(24)(25)(26). Cycloheximide inhibited S100b-induced COX-2 mRNA expression only at 2 h and not 4 h (Fig. 2). Furthermore the IL-1␤ neutralizing antibody had only a slight inhibitory effect at 4 h (Fig. 1D). These new results indicate that the effects of S100b could be via both an IL-1␤Ϫdependent (early) and a unique (later) IL-1␤-independent mechanism involving transcription. S100b Induces Significant Increases in COX-2 mRNA and Protein Expression and Formation of Its Product PGE 2 in THP-1 Cells-Because S100b effects on COX-2 expression were maximal at 4 h and also appeared to be transcriptionally regulated at this time point, we performed additional experiments at this time period. The bar graph in Fig. 3A shows data from multiple experiments normalized to 18 S internal control. This indicates that the effects of S100b at 4 h are statistically significant (p Ͻ 0.001). Fig. 3B shows that the actions of S100b were dose-dependent with maximum effects on COX-2 mRNA induction occurring at a concentration of 5 to 7.5 g/ml.
Although S100b is a physiological ligand for RAGE, we wanted to determine whether AGEs prepared in vitro could also elicit similar effects on COX-2 expression as this would also be physiologically significant and relevant to diabetic complications. We therefore examined the effects of in vitro prepared methyl glyoxylated albumin (40). THP-1 cells were treated with this AGE (400 g/ml) for 1 to 4 h. Fig. 3C shows that this methyl glyoxal AGE could also increase COX-2 expression that reached a maximum by 2 h and to a lesser extent at 4 h. This demonstrates that physiological AGEs can also induce COX-2 in THP-1 cells (Fig. 3C). As a negative control we treated the cells with unmodified BSA (400 g/ml) for 4 h (Fig.  3D). This BSA treatment, unlike AGE, did not induce COX-2 expression thereby demonstrating that the AGE-induced COX-2 mRNA expression is specific to AGE-modified BSA.
Because both S100b and in vitro prepared AGE could induce COX-2 mRNA expression, we wanted to confirm the involvement of RAGE. THP-1 cells were pretreated with a specific anti-RAGE antibody (generous gift from Dr. David Stern, Medical College of Georgia) for 1 h prior to S100b treatment. RT-PCR analyses of RNA from these cells showed complete blockage of S100b-induced COX-2 mRNA levels in antibody-treated cells (lane 4 in Fig. 3E). This suggests that S100b-induced COX-2 mRNA is via RAGE activation.
FIG. 1. Analysis of S100b-induced gene expression by relative RT-PCR. Relative RT-PCRs were performed with total RNA isolated from THP-1 cells treated with or without S100b for 30 min to 24 h, using gene-specific primers. 18 S RNA primers were included in each PCR reaction as internal control. PCR products were analyzed on 2.5% agarose gels. The figure shows ethidium bromide-stained agarose gels of RT-PCR products. Symbols Ϫ and ϩ above the lanes indicate control and S100b-treated cells, respectively. A shows RT-PCR for COX-2; B, COX-1; C, IL-1␤. In D, cells were pretreated either with a neutralizing antibody to IL-1␤ or control (Ctrl) IgG and then stimulated with S100b. COX-2 mRNA expression was then determined by RT-PCR as described above.
FIG. 2. S100b treatment induces COX-2 mRNA via increased transcription, as well as protein synthesis. THP-1 cells were pretreated with actinomycin D (0.1 g/ml) or cycloheximide (0.5 g/ml) and then stimulated with S100b. Aliquots of cells were taken at 2 and 4 h, and RT-PCRs were performed as described for Fig. 1A. Panels A and B show the effects of actinomycin (Act D) and cycloheximide (Cyclo) on S100b-induced COX-2 mRNA at 2 and 4 h, respectively.
To further confirm the RT-PCR data on S100b-induced COX-2 mRNA regulation, we further performed sensitive RPA using 32 P-labeled COX-2 and 18 S antisense riboprobes. Fig. 4 shows a representative autoradiograph of the results. A clear protected COX-2 RNA (254-nt) band was seen in the S100b (6.5 g/ml)-treated sample (Fig. 4, lane 3), whereas no protected band was seen in the untreated control (lane 2). 18 S was used as an internal control. The 18 S protected bands of the predicted sizes are seen in both the control and treated lanes. Lane 1 shows probe whereas lane 4 is experiment run without RNA. These results substantiate the RT-PCR data.
The RT-PCR and RPA data clearly showed significant induction of COX-2 mRNA in S100b-treated cells. We therefore evaluated whether the COX-2 protein levels were also regulated in S100b-treated THP-1 cells. Western blot analysis with a specific COX-2 antibody was carried out using total protein prepared from control and S100b-treated cells at various time points as shown in Fig. 5A. It is seen that COX-2 protein appeared by 4 h and peaked at 8 h after S100b treatment. This is consistent with the mRNA data, and it demonstrates that COX-2 protein appears shortly after mRNA induction by S100b, which peaked by 2-4 h. Equal loading of protein in each lane was confirmed by probing with an anti-actin antibody as internal control (lower panel of Fig. 5A).
To determine whether the induction of COX-2 mRNA and protein levels were also associated with increase in COX-2 enzyme activity, we examined the levels of the COX-2 product PGE 2 in S100b-treated versus control THP-1 cells. PGE 2 released into the medium was measured in culture supernatants FIG. 3. A-D, S100b-treated THP-1 cells show specific and significant increase in COX-2 mRNA levels. The bar graph in A shows significant induction of COX-2 mRNA levels in S100b-treated versus control cells. Values are normalized to 18 S internal control and shown as mean Ϯ S.E. of five-nine independent experiments (*, p Ͻ 0.001). B shows dose response effects of S100b on COX-2 mRNA expression in THP-1 cells. Total RNA isolated from the THP-1 cells, treated with 0, 2.5, 5, 7.5, and 10 g/ml S100b for 4 h, were used to perform RT-PCR using genespecific primers and 18 S RNA primers. C shows agarose gel of RT-PCR products of COX-2 mRNA amplification from THP-1 cells treated with methyl glyoxal modified bovine serum albumin (AGE) for 1, 2, and 4 h. Symbols Ϫ and ϩ indicate without and with AGE treatment, respectively. D, control for experiment in C run with cells treated with or without unmodified BSA or AGE. Results show that, unlike AGE, unmodified BSA does not increase COX-2 mRNA expression. E shows the effect of anti-RAGE antibody on COX-2 induction by S100b. THP-1 cells were pretreated with 2 g/ml of anti-RAGE antibody for 1 h followed by 6.5 g/ml of S100b for 4 h. COX-2 mRNA levels were analyzed by RT-PCR using specific primers for COX-2 and 18 S internal control. Results shown in B-E are representative of two-four similar experiments. FIG. 5. S100b stimulates COX-2 protein and its product PGE2 in THP-1 cells. A, Western blot analysis of COX-2 protein in S100btreated THP-1 cells. Total protein isolated from control untreated THP-1 cells or cells treated with S100b for 0.5, 1, 2, 4, 8, and 24 h were resolved by SDS-PAGE, transferred onto Immobilon membranes, and probed with an anti COX-2 polyclonal antibody. Symbols Ϫ and ϩ indicate without (control) and with S100b treatment, respectively. The upper panel shows the COX-2 immunoblot, whereas the lower panel shows the same blot stripped and probed with actin, an internal control. B, S100b stimulates PGE 2 production in THP-1 cells. PGE 2 , a COX-2 enzyme product, was measured by EIA in the culture supernatants of THP-1 cells treated with or without S100b. In some experiments cells were pretreated with NS-398 or Bay11-7082 for 30 min and then treated with S100b for 4 h. Values shown are mean Ϯ S.E. of three independent experiments. *, p Ͻ 0.001 versus S100b; U.D, undetectable. by a specific EIA. Fig. 5B shows that S100b (6.5 g/ml) treatment for 8 h led to a significant increase in PGE 2 levels (340 Ϯ 48 fg/ml, p Ͻ 0.001). In contrast, there were no detectable levels of PGE 2 (Fig. 5B, U.D., undetectable) in supernates of untreated control cells. Pretreatment of the cells with the specific COX-2 inhibitor (NS-398) led to complete abrogation of S100binduced PGE 2 production confirming the source of PGE 2 . It was also blocked by an NF-B inhibitor (Bay11-7082), suggesting that COX-2 regulation by S100b is mediated to a large extent by NF-B activation. Overall, these results demonstrate that S100b and AGE can induce the expression and activity of COX-2, and this could be relevant to monocyte activation and dysfunction associated with atherosclerosis and other diabetic vascular and inflammatory complications.
Signal Transduction Mechanisms Involved in S100b-induced COX-2 mRNA Expression-To determine the key signal transduction pathways involved in S100b-induced COX-2 mRNA in THP-1 cells, we evaluated the effects of inhibitors of pathways known to be activated by AGEs. We pretreated THP-1 cells with Bay11-7082 (NF-B inhibitor), SB202190 (p38MAPK inhibitor), PD-98059 (MEK/ERK MAPK inhibitor), AG-490 (JAK inhibitor), GFX (PKC inhibitor), and NAC (antioxidant). Results in Fig. 6, A and B shows that S100b-induced COX-2 mRNA expression was significantly blocked by PD-98059 (PD), NAC, and Bay11-7082 (Bay) as well as by GFX and SB202190 (SB). However, AG-490, the JAK inhibitor, had no effect at all on S100b-induced COX-2 mRNA expression (Fig. 6B). These results implicate the involvement of multiple pathways including the MAPK pathway, oxidant stress, PKC, and the transcription factor NF-B in S100b-induced COX-2 mRNA expression.
We next determined whether these pathways were directly activated by S100b. Thus Fig. 7 shows that S100b increases the activities of p38MAPK, ERK1/2 MAPK, and PKC as indicated by increased levels of the phosphorylated forms of these (pPKC, pERK, pp38). Total p38MAPK (p38) was not altered under these conditions. S100b also led to IB-␣ phosphorylation and degradation (indicators of NF-B activation) as shown later in Fig. 10. S100b Treatment Activates Transcription from the COX-2 Promoter-To examine whether S100b at 4 h can induce transcription from the COX-2 promoter, we transfected THP-1 cells with promoter reporter construct plasmids, pCOX2(Ϫ7146/ ϩ127) and pCOX-2(Ϫ1430/ϩ127), that expresses the firefly luciferase (Luc) gene under the control of human hCOX-2 gene promoter. Luciferase analyses of 4-h S100b-stimulated THP-1 cells transfected with these plasmids showed significantly elevated luciferase activity in pCOX2(Ϫ1430/ϩ127) transfected samples (Fig. 8A) (p Ͻ 0.001). However, the longer promoter construct phCOX2 (Ϫ7146/ϩ127) did not yield any luciferase activity over that seen by the control pGL3-Luc construct (Fig.  8A). These results suggest that COX-2 up-regulation at 4 h is mediated by key promoter elements present within the 1430 region, but it appears that certain repressive elements could be present more upstream resulting in a silencing effect by the longer Ϫ7146 construct of the 5Ј end of the COX-2 promoter.
The human COX-2 promoter region (Ϫ1430/ϩ127) contains several cis-acting elements. We were interested in five cisacting elements near the 5Ј end of the coding region of COX-2, namely two NF-B binding sites, two NF-IL-6 binding sites, and a cAMP-response element binding site (Fig. 9A), all of which have been shown to be involved in the regulation of COX-2 gene transcription by various agonists. Our results in Fig. 5B and Fig. 6 showed clear inhibition of S100b-induced COX-2 activity and mRNA expression by the pharmacological NF-B inhibitor Bay11-7082, suggesting the involvement of NF-B transcription factor. To confirm this result at the molecular level and to further evaluate the key promoter elements involved in S100b transcriptional activation, we co-transfected the cells with the phCOX2 (Ϫ1430/ϩ127) and an NF-B supersuppressor IB (mutant) plasmid that suppresses NF-B activation. Fig. 8B shows that co-transfection of this mutant IB plasmid significantly inhibited S100b-induced COX-2 promoter activity. The mutant IB plasmid had no significant effect on FIG. 6. Inhibition of S100b-induced COX-2 mRNA by kinase inhibitors and anti-oxidant. THP-1 cells were pretreated with vehicle (0.01% Me 2 SO) or various inhibitors, AG-490 (AG; JAK inhibitor); Bay11-7082 (Bay; NF-B inhibitor); PD-98059 (PD; ERK1/2 MAPK inhibitor); SB202190 (SB; p38 MAPK inhibitor); GFX (GFx; PKC inhibitor); or NAC (antioxidant), for 1 h prior to S100b stimulation. Total RNA was isolated post-S100b stimulation, and COX-2 mRNA levels were analyzed by RT-PCR as described for Fig. 1. Symbols Ϫ and ϩ indicate the absence and presence of S100b, respectively. A shows a representative gel of COX-2 and 18 S PCR products, and B shows a bar graph of data quantitated from three to four experiments expressed as -fold over respective control. *, p Ͻ 0.001 versus S100b; **, p Ͻ 0.05 versus S100b; #, p Ͻ 0.01 versus S100b. the control pGL3-Luc plasmid.
Our earlier data indicated that at the 2-h time point, S100b effects on COX-2 mRNA were not transcriptionally regulated. We tested this further with promoter-luciferase assays. Fig. 8C shows that when cells were treated for 2 h with S100b, unlike at 4 h, there was no increase in luciferase activity with the pCOX-2 (Ϫ1430/ϩ127) construct. This provides additional support that at 2 h, COX-2 expression is not transcriptionally regulated.
Further Analysis of the Role of Specific NF-B Elements in the COX-2 Promoter-The COX-2 promoter contains two NF-B binding elements flanked by two NF-IL-6 elements (Fig.  9A). Data in Fig. 8 and earlier figures demonstrated the involvement of NF-B activation in COX-2 expression but did not provide information on which cis-acting NF-B elements are involved in this activation. So in the next step we examined the relative contribution of the two NF-B sites on the COX-2 promoter. We constructed several COX-2 promoter luciferase gene fusion constructs containing luciferase gene under the control of various regions of the hCOX-2 promoter. This was performed by PCR amplification as described under "Experimental Procedures." As shown in Fig. 9, A and B, COX-2 promoter deletion constructs such as pCOX2(Ϫ860/ϩ127), pCOX2(Ϫ360/ϩ127), pCOX2(Ϫ216/ϩ127), pCOX2(Ϫ123/ ϩ127), and pCOX2(Ϫ52/ϩ127) were transfected into THP-1 cells. After transfection, the cells were stimulated with S100b, and luciferase activities were measured. Fig. 9B shows that transfection of THP-1 cells with the pCOX2(Ϫ860/ϩ127) construct, which contains all five cis-acting elements, showed over a 2-fold increase in luciferase activity over control cells transfected similarly. In this case, basal activity was also increased markedly. As noted earlier in Fig. 8, the pCOX2(Ϫ1437/ϩ127) construct also showed a significant increase in activity over control. On the other hand, removal of NF-IL-6 (Ϫ530/Ϫ543) and the NF-B (Ϫ439/Ϫ448) (namely the pCOX2(Ϫ360/ϩ127) FIG. 8. S100b stimulates transcription from the COX-2 promoter in THP-1 cells. A, analysis of the promoter region responsible for S100b effects. THP-1 cells were transfected with either a control plasmid containing the promoter-less luciferase gene pGL3-Luc or plasmids containing luciferase gene under the control of human COX-2 promoter sequences (Ϫ7146/ϩ127) or (Ϫ1430/ϩ127). After a 24-h recovery period, cells were treated with S100b for 4 h, lysed, and then luciferase activities were determined. Results show that the Ϫ1430 flanking region is responsive to S100b in increasing COX-2 promoter activation. Values shown are mean Ϯ S.E. of three independent experiments (*, p Ͻ 0.001 versus control). B, involvement of NF-B in S100bstimulated COX-2 promoter activation in THP-1 cells. THP-1 cells were transfected with either a control plasmid (pGL3-Luc) or plasmid with COX-2 promoter pCOX-2 (Ϫ1430/ϩ127) as described for A. In addition some cells were co-transfected with plasmid containing a mutant IB expressed from the CMV promoter (pCMVmIB) that can repress NF-B activation. After a 24-h recovery period, cells were treated with S100b for 4 h, and luciferase activities were determined. Values shown are mean Ϯ S.E. of three independent experiments. (*, p Ͻ 0.001 versus control without S100b). C, involvement of both transcriptional dependent and independent expression of COX-2 mRNA at two different time points by S100b treatment. THP-1 cells were transfected with pCOX-2 (Ϫ1430/ϩ127), pCOX-2 (Ϫ7146/ϩ127), or control plasmid pGL3-Luc as described for A. After a 24-h recovery period, cells were treated with S100b. Aliquots of cells were taken at 2 and 4 h, and luciferase activities were determined. Values shown are mean Ϯ S.E. of three independent experiments (*, p Ͻ 0.001 versus control without S100b).
FIG. 9. The specific NF-B site involved in COX-2 promoter activation by S100b. The promoter activities of a series of 5Ј deletion mutants made in the COX-2 promoter flanking region were analyzed by transient transfection into THP-1 cells followed by treatment with or without S100b (6.5 g/ml). Deletion mutants of COX-2 promoter constructs are named by the length of the regulatory region. The TATA box (TATA) and several enhancer sites are indicated in A. Mutant IB-␣ plasmid is indicated as pIB (mut). Results are expressed as the mean Ϯ S.E. of three independent experiments (*, p Ͻ 0.001 versus control). Serial deletion mutants demonstrated the importance of the Ϫ860-bp flanking region containing the cis-acting elements cAMP-response element, two NF-IL-6 sites, and two NF-B sites for basal promoter activity (B). Co-transfection of a IB mutant plasmid, along with the Ϫ860 deletion mutant, showed a significant inhibition in S100b-induced COX-2 promoter activation. This, along with the loss of promoter activity in the Ϫ360 deletion construct, indicates the importance of the distal NF-B (at Ϫ439/Ϫ448) in S100b-induced COX-2 expression. construct), led to abrogation of the luciferase activity. Furthermore, co-transfection of the IB mutant plasmid with pCOX2(Ϫ860/ϩ127) blocked luciferase activity induced by this plasmid (Fig. 9B). These data confirm the key role of the distal NF-B binding site (Ϫ439/Ϫ448) in the COX-2 regulation by S100b.
S100b Induces IB-␣ Phosphorylation and Degradation in THP-1 Cells-The results in Fig. 5B, Fig. 6, Fig. 8B, and Fig. 9 all clearly indicate the involvement of NF-B in S100b-induced COX-2 mRNA expression and its product PGE 2 synthesis. To further confirm that S100b can lead to the activation of NF-B, we examined whether IB-␣ phosphorylation and degradation in the cytoplasm is induced by S100b. Cytoplasmic fractions from S100b treated THP-1 cells were prepared as described earlier (38) and analyzed by immunoblotting with a phosphospecific IB-␣ antibody and an anti-IB-␣ antibody. Fig. 10 (upper panel) showed that S100 treatment steadily decreased the levels of the IB-␣ protein with increasing time. This decrease reached a maximum at 60 min post-S100b treatment induction after which the levels began rising again by 2 and 4 h most likely because of new synthesis of the IB-␣ protein. At the same time, Fig. 10 (middle panel) shows that S100b treatment also increased levels of phosphorylated IB-␣ protein as early as 15 min and remained high up to 240 min. Equal protein loading was confirmed by stripping the blot and reprobing with an anti-actin antibody (Fig. 10, lower panel). These data confirm that S100b can directly lead to IB-␣ phosphorylation and degradation and hence NF-B activation in THP-1 cells.
S100b Induces the Binding of THP-1 Monocytes to VSMC-To determine the functional significance of S100binduced COX-2 expression and increased PGE 2 levels, we evaluated a potential role in monocyte adherence to PVSMC. Monocyte adhesion to VSMC may play a key role in monocyte retention in vessel wall, a key step in the pathogenesis of atherosclerosis that has not been well studied. Results in Fig.  11A show that S100b-treated THP-1 cells displayed a 2-to 3-fold increase in adherence to PVSMC relative to the untreated control cells. Pretreatment with NS-398, a COX-2-specific inhibitor abolished the S100b-induced adherence, suggesting the involvement of COX-2 in S100b-induced monocyte activation and adhesion. To further evaluate the signal transduction pathways involved in S100b-induced monocyte binding, we performed the binding assays with cells that had been pretreated with various inhibitors, NS-398 (COX-2 inhibitor), Bay11-7082 (NF-B inhibitor), AG-490 (JAK inhibitor), PD-98059 (MEK inhibitor), SB202190 (p38 MAPK inhibitor), or GFX (PKC inhibitor). Fig. 11B shows that inhibitors of COX-2, NF-B, MEK, p38 MAPK, and PKC significantly inhibited binding activity induced by S100b, whereas AG-490 had no effect. These results are similar to their effects on COX-2 mRNA induction and suggest the involvement of NF-B, MAPK, and PKC signaling pathways in S100b-induced monocyte binding, as well as COX-2 induction. S100b Induces COX-2 mRNA Expression in PBMC-Because THP-1 cells may not fully represent the phenotype of blood monocytes, we next examined whether S100b can also induce COX-2 mRNA expression in PBMC isolated from normal healthy adult donors. These PBMC were treated with S100b for 4 h, total RNA was isolated, and COX-2 mRNA levels were analyzed by RT-PCR as described in Fig. 1. Fig. 12A shows that S100b could clearly induce COX-2 mRNA expression in the PBMC. However, PBLC obtained from the same donors and treated similarly did not respond to S100b in this manner. COX-2 induction in THP-1 cells is seen in the far right. These results clearly suggest that induction of COX-2 by S100b is evident even in primary blood monocytes and is specific to monocytes, because no induction was observed in lymphocytes.

COX-2 mRNA Expression Is Elevated in Peripheral Blood Monocytes Isolated from Diabetic Patients Relative to Normal
Healthy Non-diabetic Controls-We next obtained PBMC from Type 1 diabetic patients to determine whether they had increased levels of COX-2 mRNA. Fig. 12B shows that PMBC FIG. 10. S100b induces phosphorylation and degradation of IB-␣ protein. Cytoplasmic fractions of THP-1 cells were prepared after stimulating with S100b at various time points from 15 to 240 min. Proteins separated by SDS-PAGE electrophoresis were immunoblotted and probed with anti-phospho-IB-␣ antibody, anti-IB-␣ antibody, or anti-actin antibody as an internal control. Probing with the anti-IB-␣ antibody demonstrated increasing degradation of IB-␣ protein with maximum seen at 1 h post-S100b stimulation (A). The figure also shows the induction of phosphorylation of IB-␣ protein by S100b, which appears by 15 min and remains elevated up to the 240-min time period (B). The lowest band in C shows the stripped blot reprobed with actin as an internal control. Symbols Ϫ and ϩ indicate without and with S100b treatment, respectively.
FIG. 11. COX-2 activation is involved in S100b-induced binding of THP-1 cells to PVSMC. THP-1 cells were treated with S100b protein (6.5 g/ml) for 4 h and then labeled with the fluorescent tag BCECF/AM for 30 min. Fluorescently labeled THP-1 cells (5 ϫ 10 4 ) were allowed to adhere to a monolayer of PVMSC for 1 h. Non-specifically bound cells were removed by extensive washings with medium. The bound THP-1 cells were quantitated on a fluorescence plate reader. S100b-stimulated THP-1 cells showed greater than 2-fold increase in binding relative to control (*, p Ͻ 0.001 versus control) (A). The COX-2-specific inhibitor NS-398 significantly inhibited S100b-induced adherence (A). The bar graph in B shows that the JAK inhibitor AG-490 did not significantly block S100b effects, whereas Bay11-7082 (NF-B inhibitor), NS-398 (COX-2 inhibitor), PD-98059 (MEK inhibitor), SB (p38 MAPK inhibitor), and GFX (PKC inhibitor) all significantly blocked the S100b-induced THP-1 adherence to PVSMC (*, p Ͻ 0.001 versus S100b). isolated from two Type 1 diabetic patients (run in duplicate) had markedly elevated expression of COX-2 mRNA unlike the normal controls who did not have any COX-2 mRNA expression. These important results further establish the significance and pathological relevance of our in vitro data. DISCUSSION COX-2 and its products such as PGE 2 have been implicated in several inflammatory diseases including atherosclerosis (27)(28)(29)(30). COX-2 has also been implicated in the inhibition of insulin secretion and mediation of islet dysfunction related to the development of Type 1 diabetes (11)(12)(13)(14)(15). Evidence now clearly supports the role of inflammatory mediators in the development of vascular diseases such as atherosclerosis (57). Furthermore, diabetes is associated with significantly accelerated rates of atherosclerosis, hypertension, and inflammation. Although changes in vascular prostaglandin production are implicated in the derangement of vascular reactivity associated with diabetes (58), the role of COX-2 in diabetic monocytes and atherosclerosis is not known. In the present studies, we demonstrated for the first time that treatment of monocytes with in vitro prepared AGEs or the specific RAGE ligand, S100b, can significantly increase COX-2 mRNA and protein expression and production of the COX-2 product, PGE 2 , in THP-1 human monocytes. This occurred through the RAGE receptor, because an anti-RAGE antibody could block the increase in COX-2 mRNA expression. S100b-induced COX-2 mRNA expression appeared to be partly dependent on transcription (at 4 h) and partly related to new protein synthesis (at 2 h) depending on the time of treatment. Thus we have uncovered a dual mode of action in which S100b may mediate its effects partly through the production of another factor or cytokine such as IL-1␤ and partly through a novel IL-1␤Ϫindependent mechanism based on transcription. This kind of biphasic dual regulation of COX-2 based on transcription and de novo protein synthesis was observed recently (56) with another agonist. Importantly, we noted that S100b also increased COX-2 mRNA expression in normal primary human blood monocytes from non-diabetic volunteers but not in lymphocytes from the same individuals. Furthermore, we noted high levels of COX-2 mRNA expression in monocytes isolated from Type 1 diabetic patients, whereas monocytes from non-diabetic subjects do not have basal expression of COX-2 mRNA. These data support the physiological and pathological relevance of our findings. S100b-induced COX-2 mRNA expression at 4 h was transcriptionally regulated, because actinomycin D blocked expression, and furthermore, S100b at this time period could increase transcription from a minimal human COX-2 promoter luciferase construct. Promoter deletion analyses, as well as pharmacologic and genetic inhibitors of NF-B, demonstrated the key role of NF-B in S100b-induced COX-2 gene transcription. Furthermore, the specific S100b-responsive NF-B site in the hCOX-2 promoter was evaluated using promoter deletion analyses. The promoter region of the human COX-2 gene has two NF-B consensus sites (18), one located within Ϫ455 to Ϫ428 bases, and the other located within Ϫ232 to Ϫ205 bases from the transcriptional start site (Fig. 9A). In the present study, we were able to determine that the site located further away from the start site (at Ϫ439/Ϫ448) is essential for S100b-induced COX-2 transcription. S100b was also able to phosphorylate and degrade IB-␣, indicating NF-B activation. S100b also increased activities of PKC, p38, and ERK1/2 MAPKs. Furthermore, using signal transduction pathway-specific inhibitors, we were able to determine that activation of NF-B, ERK1/2, p38 MAPK, and PKC are all involved in S100b-induced COX-2 mRNA induction, thus indicating the operation of multiple pathways in COX-2 regulation in monocytes under diabetic conditions. There are reports that AGEs can activate the JAK-STAT pathway (59). However, because the JAK inhibitor, AG490, did block COX-2 mRNA induction, it appears that the JAK-STAT pathway does not mediate S100b effects on COX-2 mRNA expression.
COX-2 products have been associated with several cellular properties including vasodilation, vasoconstriction, cellular adhesion, and migration (3)(4)(5). Products such as PGI 2 are protective and vasodilatory, whereas others including PGE 2 are inflammatory and are related to the proinflammatory properties of COX-2 (3)(4)(5). Although PGI 2 is a major COX-2 product in endothelial cells, PGE 2 is the major product in monocytes. Hence, our observation of increased PGE 2 production under diabetic conditions in monocytes demonstrates that diabetes induces a proinflammatory environment. The overall beneficial effects of specific COX-2 inhibitors in the treatment of various inflammatory diseases without the gastrointestinal side effects of aspirin indicate the pathogenic role of COX-2 (3-5, 17, 31). However, some reports have indicated potential drawbacks of long term use of COX-2 inhibitors (60). Based on our data, it is possible that COX-2 inhibitors could be beneficial for diabetic vascular complications. This is further supported by the new report showing that high glucose could increase COX-2 expression in endothelial cells and thromboxane production and decreased PGI 2 release (34).
To determine the functional relevance of COX-2 induction by S100b in monocytes, we examined whether COX-2 activation plays a role in the adhesion of THP-1 cells to PVSMC. This assay models monocyte retention in the vascular wall, a key FIG. 12. COX-2 regulation in human PBMC. A, S100b induces COX-2 mRNA expression in PBMC from normal volunteers. PBMC were isolated from normal healthy non-diabetic adult donors as described under "Experimental Procedures." 2 ϫ 10 5 isolated monocytes in 4 ml of RPMI medium were stimulated with S100b (6.5 g/ml) for 4 h. COX-2 mRNA levels were measured by RT-PCR analysis as described for Fig. 1. Isolated lymphocyte fractions (PBLC) from the same donors were also similarly stimulated with S100b protein. THP-1 cells are seen in the far right. As seen in the figure, there is no basal expression of COX-2 mRNA in any of the cells. S100b leads to a marked stimulation in COX-2 expression in PBMC and THP-1 cells but not in PBLC. Results shown are representative of two experiments each from two separate donors. B, monocytes from Type 1 diabetic patients have elevated COX-2 mRNA expression relative to non-diabetic subjects. PBMC were isolated from Type-1 diabetic patients (DM) or normal subjects (Nor) as described for A. 2 ϫ 10 5 isolated monocytes were directly processed for total RNA isolation. RT-PCR analyses showed markedly elevated COX-2 mRNA expression in two diabetic patients (DM) run in duplicate. In contrast, RNA from two non-diabetic subjects (Nor) run in duplicate showed no basal expression of COX-2 mRNA at all. step in the pathogenesis of atherosclerosis. Although monocyte adhesion to endothelial cells is a well documented and studied key early step in atherogenesis, we are interested in evaluating the less studied mechanisms of monocyte retention in the vascular wall. We observed that S100b treatment of THP-1 cells could significantly increase their adhesion to a monolayer of VSMC. This increased adhesion was significantly inhibited by the COX-2 inhibitor, NS-398, as well as inhibitors of NF-B, PKC, p38, and ERK1/2 MAPK pathways. Because these inhibitors also blocked S100b-induced COX-2 mRNA expression, our data suggest that the activation of these signaling pathways by S100b (or AGEs) can lead to monocyte adhesion, retention, and foam cell formation. Overall these new results indicate that AGEs acting via RAGE can lead to COX-2 expression and monocyte activation and thereby accelerate the progression of atherosclerosis.
Several reports have shown that COX-2 is a highly inducible gene. It is particularly responsive to growth factors and mediators of inflammation such as TNF-␣, IL-1␤, IL-6, and lipopolysaccharide (1)(2)(3)(4)(5)(6)(7)(8)(9). Regulation by growth factors such as platelet-derived growth factor suggest that COX-2, like the immediate early genes, c-fos and c-myc, may also play some general role in mitogenesis. Thus, although COX-2 is induced rapidly as also seen in the present study, it can have sustained cellular effects. COX-2 products, once formed, can further propagate cellular effects via their own receptors.
Recent studies have shown that HG culture of monocytes can lead to the transcriptional regulation of the inflammatory cytokine, TNF-␣, and the chemokine monocyte chemoattractant protein-1 via oxidant stress-dependent mechanisms, as well as key signaling pathways (38,41). Furthermore, NF-B activation by HG seemed to play a key role in the transcription of the TNF-␣ and monocyte chemoattractant protein-1 genes in monocytes (38,41). HG culture of THP-1 monocytes also significantly increased their adherence to human aortic endothelial cells (41). Very recently, gene profiling with DNA arrays demonstrated that key cytokine, chemokine, and related genes are regulated by HG in THP-1 monocytes (41). Most of the genes induced by HG were regulated at least in part by NF-B, thereby indicating a central role played by this transcription factor in leading to diabetic vascular and other complications. The present data showing the importance of NF-B in COX-2 regulation by AGEs, and RAGE ligation further underscores this and also adds to the growing list of inflammatory genes regulated under diabetic conditions. Furthermore, it indicates the operation of a vicious loop in diabetes in which various cells within the vessel wall and in circulation can respond to HG and AGEs by producing inflammatory mediators that operate in autocrine and paracrine fashions to amplify the overall inflammatory response. Thus, although the role of COX-2 and its products in leading to inhibition of insulin secretion and islet destruction is known, this report describes for the first time the regulation of COX-2 by diabetic conditions in monocytes that could have far reaching consequences.