Induction of 15-Lipoxygenase Expression by IL-13 Requires Tyrosine Phosphorylation of Jak2 and Tyk2 in Human Monocytes*

The enzyme 15-lipoxygenase (15-LO) participates in the dioxygenation of polyenoic fatty acids. This activity leads to the degradation of mitochondrial membranes during reticulocyte differentiation, the production of pro- and anti-inflammatory mediators by a variety of cell types, and the oxidation of lipids in atherosclerotic lesions. The cytokines, IL-4 and IL-13, are reported to induce the expression of 15-LO in human peripheral blood monocytes. In this report we explore the signaling mechanisms involved in the IL-13-mediated induction of 15-LO expression. First we demonstrate that the delayed induction of 15-LO requires continuous stimulation of monocytes for a minimum period of 12 h. We also found that tyrosine kinase inhibitors blocked the induction of 15-LO in a dose-dependent manner. By immunoprecipitation and antiphosphotyrosine blotting experiments, IL-13 was shown to induce tyrosine phosphorylation of Jak2 and Tyk2, but not Jak1 or Jak3, within 5 min of treatment in human monocytes. To investigate whether the early induction of tyrosine phosphorylation of both Jak2 and Tyk2 was ultimately involved in 15-LO expression, we generated antisense oligodeoxyribonucleotides (ODNs) against Tyk2 and Jak2. We employed a cationic lipid-mediated delivery technique to transfect the monocytes and found that both antisense ODNs inhibited expression of their target proteins by 75–85%. The treatments were specific and did not affect the expression of each other. Furthermore, the antisense ODNs to Jak2 and Tyk2 both inhibited the induction of expression of 15-LO in monocytes treated with IL-13. Parallel experiments with sense ODNs to Jak2 and Tyk2 did not affect their protein levels or the induction of 15-LO by IL-13, and down-regulation of Jak1 also did not affect expression of 15-LO. Our results suggest the novel finding that IL-13 can induce tyrosine phosphorylation of both Jak2 and Tyk2 in primary human monocytes. This occurs as an early and essential signal transduction event for the IL-13-mediated induction of 15-LO expression. These data represent the first characterization of upstream kinases involved in the induced expression of 15-LO.

Products of 15-LO-mediated oxidation of linoleic acid have been identified in human and rabbit atherosclerotic lesions and likely result from the action of 15-LO on lipoprotein lipids (6,14). It has been reported that 15-LO is expressed at a high level in lipid-loaded macrophages of atherosclerotic lesions (14 -16) and to a lesser extent in subtypes of arterial smooth muscle cells (17); however, the role of this enzyme in the pathogenesis of cardiovascular disease is far from clear. Because circulating blood monocytes of normal individuals do not express , the enzyme must be induced during activation or macrophage differentiation in the tissue.
Numerous lymphokines are known to modulate the inflammatory response through their actions on monocytes (19,20). The involvement of the local environment of the monocyte in influencing 15-LO expression is supported by studies showing that IL-4 and IL-13 are specific inducers of the enzyme (18,21). Attempts to study the molecular signaling mechanisms using monocytic cell lines have not been successful, as the cell lines failed to show similar induction of 15-LO when stimulated with interleukins (22). Of several permanent human hematopoietic and epithelial cell lines, only the lung epithelial cell line, A549, showed induction of 15-LO, even though all the cell lines tested were positive for the presence of cell surface receptors for both IL-4 and IL-13 (22).
Several reports support the view that interleukins exert their effects through the mediation of different combinations of receptor subunits in different cell lines and thus activate different combinations of Jaks and signal transducers and activators of transcription (STATs) (23)(24)(25)(26). Although IL-4R␣ (140 kDa) was shown to be a substrate for both IL-4-and IL-13-dependent tyrosine phosphorylation, the other constituents of the receptor complex in different cells and cell lines are not yet clear (27). Two different IL-13 receptors are reported, one with lower affinity toward IL-13 (K D ϭ 3-10 nM) having a molecular mass of 56 -68 kDa (28), termed IL-13R␣, and the other with higher affinity toward IL-13 (K D ϭ 20 -90 pM) having a molecular mass of 45-50 kDa (29), termed IL-13R␣Ј. Many other studies have provided evidence that IL-4 and IL-13 share receptor components (23,30,(32)(33)(34). Cross-competition by IL-4 and IL-13 for receptors was also observed in certain cell types (35)(36)(37)(38). Although IL-2R␥c is required for IL-4-mediated signal transduction in some cells, other IL-4-responsive cell lines (e.g. plasmacytoma B9 and renal cell lines) exist that do not express IL-2R␥c (39). Expression of IL-2R␥c on human monocytes has been shown to be very low (40) or absent (24), thus leading to the suggestion that the functional composition of IL-4 and IL-13 receptor complexes may vary from cell type to cell type and that IL-2R␥c is not absolutely necessary for receptor signaling.
Different groups have utilized antisense oligodeoxyribonucleotides (ODNs) to inhibit the endogenous level of expression of different enzymes, including the Jaks and STATs, in various cells and cell lines (41)(42)(43). Jak2 antisense ODN treatment resulted in a reduction of expression of up to 46% in human and murine cell lines and normal human progenitor cells (41). The level of expression of STAT1␣ was reduced in the astroglioma cell line (CH235-MG) as well as in glomerular mesangial cells using antisense ODN specific to STAT1␣ (42). Our attempts to inhibit the activity and expression of classical protein kinase C (44) and cytosolic phospholipase A 2 (45) using antisense ODNs met with an even higher level of inhibition (up to 80%) as this approach worked more efficiently in nonproliferating, rapidly pinocytosing monocytes.
In the current study, we examined the signaling intermediates triggered by IL-13 in human monocytes and assessed their involvement in the induction of expression of 15-LO. Our data indicate that the induction of 15-LO was delayed and dependent on tyrosine kinases. Using immunoprecipitation studies, we demonstrated that IL-13 induced tyrosine phosphorylation of Jak2 and Tyk2. The induction of expression of 15-LO could be inhibited by using antisense ODNs against Jak2 and Tyk2, suggesting their involvement in the IL-13-signaling pathway in human monocytes.

Chemicals
Genistein, daidzein, tyrphostin 23, and tyrphostin 1 were purchased from Biomol Research Laboratories (Plymouth, PA). All the drugs were dissolved in Me 2 SO. All the reagents were made as 1000-fold concentrated stock solutions and stored at Ϫ20°C before use.
Recombinant human IL-13 was purchased from Upstate Biotechnology, Lake Placid, NY. Antibody against rabbit reticulocyte 15-LO crossreacting with human 15-LO was raised in sheep and was a kind gift of Dr. Joseph Cornicelli, Parke-Davis.
Rabbit antisera against Jak1, Jak2, Tyk2, and Jak3 were purchased from either Upstate Biotechnology or Santa Cruz Biotechnology, Inc., CA. Each one of the Jak antibodies was essentially noncross-reactive with the other members of the Jak family and recognized corresponding antigens under native (in 1% Triton X-100 extracts) as well as denaturing conditions.

Methods
Isolation of Human Monocytes-Human blood monocytes were isolated from heparinized whole blood by sequential centrifugation over a Ficoll-Paque solution and adherence to serum-coated tissue culture flasks as described previously (46). Nonadherent cells were removed by gentle washing, and adherent cells were collected after releasing with 5 mM EDTA and plated in tissue culture plates (Costar, Cambridge, MA) at 1.0 ϫ 10 6 cells/ml. The cell population typically contained more than 95% monocytes and was maintained in Dulbecco's modified Eagle's medium (DMEM) (Life Technologies, Inc.) supplemented with 10% bovine calf serum (BCS) (HyClone, Logan, UT) at 37°C in the presence of 10% CO 2 .
15-LO Induction and Detection-IL-13 (250 pM) was added in tissue culture plates (100 mm) containing 10.0 ϫ 10 6 cells/plate with 1.0 ϫ 10 6 cells/ml of DMEM-BCS medium and incubated for up to 36 h. Subsequently, the cells were washed three times with PBS to remove the traces of DMEM, 10% BCS. The plates were placed on ice, and the cells were lysed using 200 l of lysis buffer (1% Triton X-100, 150 mM NaCl, 50 mM Tris-HCl, pH 7.4, 1 mM phenylmethylsulfonyl fluoride and 10 l of protease inhibitor mixture (Sigma)/1 ml of lysis buffer). After 30 min, the lysate was centrifuged for 15 min at 9300 ϫ g. The supernatant was collected, and the protein concentration was determined using the Bio-Rad protein assay kit and loaded on a 7.5% SDS-PAGE gel (50 g of lysate/well). The proteins were transferred to a PVDF membrane (0.2 m) (Bio-Rad) using a Trans-Blot SD electrophoretic transfer cell (Bio-Rad). The membrane was blocked in 3% nonfat milk in 20 mM Tris-HCl, pH 7.4, 150 mM NaCl, and 0.1% Tween 20 for one h at room temperature and was probed with antibody to rabbit reticulocyte 15-LO (diluted 1:2000 in 20 mM Tris-HCl, pH 7.4, 150 mM NaCl, and 0.1% Tween 20 for 1 h at room temperature. This antibody was shown to be specific for 15-LO and does not cross-react with 5-LO (47). A horseradish peroxidase-labeled secondary antibody (ICN Biochemicals, Cleveland, OH) diluted 1:5000 in 20 mM Tris-HCl, pH 7.4, 150 mM NaCl and 0.1% Tween 20 was added for 1 h, and the hybridization signal was detected using Enhanced Chemiluminescence (ECL) detection reagents (Pierce) according to the manufacturer's guide and followed by autoradiography.
Immunoprecipitation and Western Blotting-Freshly isolated monocytes were pretreated with activated sodium vanadate solution (5 mM final concentration) for 30 min followed by treatment with IL-13 (250 pM) for 5 min. The treated cells were immediately lysed at 50 ϫ 10 6 /ml in lysis buffer (1% Triton X-100, 150 mM NaCl, 50 mM NaF, 50 mM Tris, pH 7.4, 5 mM sodium pyrophosphate, 1 mM sodium orthovanadate, 500 M phenylmethylsulfonyl fluoride, and 1:100 diluted protease inhibitor mixture). After 30 min on ice, the extracts were centrifuged at 9300 ϫ g for 15 min at 4°C, and the post-nuclear extract (500 g/ml) was incubated with the relevant antibody for 2 h at 4°C with constant rotation. Immune complexes were collected using Sepharose-protein A beads (20 l packed volume). After washing 3 times (10 min each) with lysis buffer, the immune complexes were released in SDS sample buffer and analyzed by SDS-PAGE followed by electrophoretic transfer to PVDF membranes (Bio-Rad). The membranes were then probed with either antibody to phosphotyrosine or individual antibodies to Jak/Tyk kinases and developed using ECL. For immunoprecipitation experiments using anti-phosphotyrosine, PY-99 (Santa Cruz, CA), no sodium pyrophosphate was included in the lysis buffer.
Studying the Half-lives of Jak2 and Tyk2-Fresh monocytes were isolated and adhered to 6-well plates (5 ϫ 10 6 cells/well) in the presence of BCS (10%) and DMEM for 2 h. The plates were washed once with fresh DMEM (without methionine) containing no BCS and subsequently replaced with DMEM (also without methionine) containing radioactive methionine ([ 35 S]methionine, 100 Ci/ml and 2 ml/well). The plates were incubated for 4 h and washed with fresh DMEM containing 10% BCS, and subsequently, the cells were lysed after 0, 4, 8, 12, and 24 h of further incubations. Each one of the Jak/Tyk kinases were immunoprecipitated from the lysates (10 ϫ 10 6 /group), run on a 7.5% SDS-PAGE gel, transferred onto PVDF membranes, and exposed on a PhosphorImager screen (Molecular Dynamics). The half-life of the protein was calculated as the time necessary for the dpm of incorporated, radioactive methionine in that protein to decrease by 50%.
Treatment of Monocytes with Oligodeoxyribonucleotides-The antisense sequences for human Jak2, Tyk2, and Jak1 were selected after studying the predicted secondary structural conformations of their mRNAs using the software program Mulfold©. The regions lacking major predicted secondary structures, e.g. loops, were mainly targeted to generate the antisense ODNs. Before final selection of the target region, the sequences were screened for uniqueness in all nonredundant GenBank CDS translations ϩ PDB ϩ SwissProt ϩ PIR using Blast© and also were tested for lack of internal secondary structure or pairing using Mulfold© (48).
The antisense oligomer selected against Jak2 was complementary to nucleotides 59 -78 of the human Jak2 sequence (accession number AA453345), whereas the antisense oligomer against Tyk2 was complementary to nucleotides 481-500 of the human Tyk2 sequence (accession number X54637). The antisense ODN to Jak1 was complementary to bases 462-480 of the human Jak1 sequence (accession number M64174).
Control ODN for Jak2 and Tyk2 consisted of sense ODN. The sense ODN sequence for Jak1 was predicted to possibly serve as an antisense for human ␥ adaptin mRNA. We therefore chose to use a scrambled ODN as a control for the Jak1 antisense. The scrambled Jak1 ODN did not display internal secondary structure, did not pair with itself, and was unique in the nucleic acid data bases listed above. All ODN contained phosphorothioate-modified oligonucleotides to limit DNA degradation, and all were HPLC-purified before use (Genosys Biotechnology Inc., Woodlands, TX). The sequences of the ODN are as follows. Cationic lipids were used to aid the delivery of oligonucleotides (49). Lipids were prepared from chloroform stock solutions of DDAB and DOPE (250 mg/ml). DDAB and DOPE were added to yield a ratio of 2:5 (w/w), DDAB:DOPE. This mixture was dried under nitrogen and then sonicated in sterile deionized water until clear to yield a concentration of 5 mM DDAB. The lipid preparation was used as a 100ϫ stock solution.
Before treating the monocytes with ODNs, the cationic lipid preparation was diluted 100-fold in DMEM without serum or antibiotics, and ODNs were added to yield a final concentration of 1.0 M sense or antisense ODN. This mixture was incubated for 30 min to form lipid: ODN complexes. Medium was removed from the monocytes, they were rinsed with DMEM without serum or antibiotics, and the lipid:ODN mixture was added. After 1 h of incubation, BCS was added to 10% (v/v), and the incubation was continued for 24 h. After this treatment, monocytes were either lysed, and the level of target proteins was analyzed, or the monocytes were exposed to IL-13 for another 24 h to study the expression of 15-lipoxygenase. Numerous studies have reported the involvement of Jak/Tyk kinases and the STAT molecules in interleukin-mediated signaling pathways (23)(24)(25)(26). At the same time, the reports indicate that these activation and tyrosine phosphorylation events take place immediately, within a few min of treatment. But in our case, we found that for the expression of 15-LO, the cells needed to be stimulated with the ligand (IL-13) for a minimum of 12 h. We were therefore interested in examining the role of Jak/Tyk involvement in 15-LO expression.

IL-13-mediated 15-LO Induction Requires Continuous Expo
Effect of Kinase Inhibitors on IL-13-mediated Induction of 15-LO Expression-Previous studies have shown IL-13-mediated tyrosine phosphorylation of numerous proteins in human monocytes (50). To examine whether IL-13-mediated induction of 15-LO expression involved signaling by tyrosine phosphorylation events, we examined the effects of potent tyrosine kinase inhibitors on this pathway. First, the monocytes were pretreated with tyrosine kinase inhibitors like genistein and tyrphostin 23 for 30 min at different dosages followed by treatment with IL-13 for 24 h, and the level of 15-LO was studied by Western blotting (Fig. 2). Whereas 10 g of tyrphostin 23/ml of medium caused only slight inhibition, 25 g/ml had a more profound effect (Ϸ50% inhibition). At 50 g/ml, tyrphostin 23 inhibited detectable 15-LO expression by approximately 98%. Genistein also substantially inhibited the induction of 15-LO when applied at 25 g/ml. In contrast, the negative structural analog controls for both genistein (daidzein) and tyrphostin 23 (tyrphostin 1) had little effect (Ͻ10% inhibition) on the expression of 15-LO when tested at identical concentrations. These results indicate that both of the tyrosine kinase inhibitors were able to substantially inhibit IL-13-mediated induction of 15-LO.
IL-13 Induces Tyrosine Phosphorylation of Jak2 and Tyk2-To determine whether the Jak family of kinases was involved in IL-13-mediated signaling pathways, cell lysates from both untreated and IL-13-treated fresh human monocytes were immunoprecipitated with antibodies to Jak1, Jak2, Tyk2, and Jak3, electrophoresed, blotted on PVDF membranes, and immunoblotted with antibody to phosphotyrosine (Fig. 3A). We found that IL-13 induced phosphorylation of Tyk2 kinase within 5 min of treatment. Jak1 and Jak2 kinases were constitutively phosphorylated at low levels, and IL-13 treatment enhanced the level of phosphorylation of Jak2 considerably, whereas Jak1 phosphorylation was not induced. Monocytes expressed very low levels of Jak3, and no increased phosphorylation of Jak3 was observed in response to IL-13.
Phosphorylation of the Jak family of kinases was also studied by immunoprecipitating the tyrosine-phosphorylated proteins from untreated and IL-13-treated monocyte cell lysates using antibody to phosphotyrosine followed by blotting with different Jak/Tyk kinase antibodies. Data presented in Fig. 3B indicate that antibody to phosphotyrosine could immunoprecipitate both Jak2 and Tyk2 from the IL-13-treated cells, whereas no Tyk2 and a basal level of Jak2 were observed in the untreated lanes. No Jak1 or Jak3 kinases were observed in either treated or untreated immunoprecipitates (data not shown). These results suggest that both Jak2 and Tyk2, but not Jak1 and Jak3, were tyrosine-phosphorylated in response to IL-13 stimulation in fresh human monocytes.
Antisense ODN Inhibition of Expression of Jak2 and Tyk2-Earlier reports have demonstrated successful inhibition of Jak/ STAT protein expression by using antisense ODNs (41)(42)(43). The use of antisense oligomers is particularly useful in studies employing cells that are not conducive to expression of transfected genes. Nonproliferating cells, such as monocytes, have served as ideal hosts for in vitro knock out studies using antisense ODNs (44,45). An important criterion in such experiments is to make sure that the antisense ODNs are present for durations long enough that pre-existing levels of the protein of interest are degraded. We, therefore, examined the half-lives for both Jak2 and Tyk2 by following turnover of metabolically radiolabeled proteins. In both cases the loss of specific radioactivity was linear over time. In human monocytes, the half-life of Jak2 was determined to be 6.25 h, and Tyk2 was found to have a half-life of 7.3 h.
Freshly isolated monocytes were treated with either antisense or sense ODNs for 24 h following procedures described under "Methods" and lysed with lysis buffer. 50 g of cell lysates from each sample were loaded on each lane of a 7.5% SDS-PAGE gel, electrophoresed, blotted on PVDF membrane, and immunoblotted with the corresponding kinase antibodies. Fig. 4 shows the effect of treatment with antisense and sense ODNs designed to target Tyk2 in human monocytes. The results show that antisense ODNs to Tyk2 inhibited the expression of Tyk2 by 75-85% in the monocytes, whereas sense ODNs as well as the cationic lipid vehicle alone had no effect. The same blot was reprobed with antibody to Jak2 kinase after stripping to check the effect of antisense or sense ODNs to Tyk2 on the level of expression of Jak2 in monocytes. The bottom panel shows nearly equal levels of Jak2 protein in all the lanes, thus suggesting that the antisense oligomer for Tyk2 inhibited the expression of the target protein (Tyk2) while having no effect on the level of expression of Jak2.
Similar experiments were conducted using antisense ODNs against Jak2 (Fig. 5). Nearly 80% inhibition of Jak2 expression was noted when the monocytes were treated with antisense ODNs to Jak2. No such effect was observed when sense ODNs were used to treat the cells nor in cells treated with the cationic lipid vehicle. Stripping and reprobing experiments using Tyk2 kinase antibody to examine the effect of antisense Jak2 on the level of Tyk2 kinase expression showed that antisense to Jak2 specifically inhibited the level of expression of Jak2 while having no effect on Tyk2 expression (Fig. 5, bottom panel).
Jak2 and Tyk2 Are Required for IL-13 Induction of 15-LO Expression-Because we already established that IL-13 activated and induced tyrosine phosphorylation of Jak2 and Tyk2, we next investigated the necessity of each of these kinases for the IL-13-induced signaling pathways leading to expression of 15-LO in monocytes. For these studies, monocytes were treated with antisense or sense ODN against Jak2 or Tyk2 kinases for 24 h. After treatment, the cells were exposed to IL-13 for another 24 h. Finally the cells were lysed, run on SDS-PAGE, electroblotted, and probed with antibody directed against rabbit reticulocyte 15-LO, which cross-reacts with human 15-LO. The results are shown in Fig. 6. The antisense ODN against either the Jak2 or Tyk2 kinases inhibited the IL-13-mediated induction of expression of 15-LO, whereas the sense ODN and the vehicle controls had no effect on 15-LO expression. As we have already established that the Jak2 antisense has no effect on the level of expression of Tyk2 and vice versa, our results clearly established that both Jak2 and Tyk2 were absolutely necessary for the IL-13-mediated signaling, and depriving the monocytes of either of these enzymes caused the inhibition of expression of 15-LO.
Jak1 Is Not Involved in Regulating the Induction of 15-LO Expression in Response to IL-13-As an additional control, we also examined whether inhibition of Jak1 altered expression of 15-LO. As noted in Fig. 3A, Jak1, in contrast to Jak2 and Tyk2, was not phosphorylated on tyrosine in response to IL-13 treatment and therefore would not be predicted to regulate IL-13mediated events. We prepared an antisense ODN to Jak1 and a scrambled ODN control. Monocytes were treated with antisense or control ODN or with the cationic lipid vehicle alone for 24 h as indicated under "Methods" and either lysed to check the level of Jak1 (Fig. 7A) or incubated an additional 24 h with IL-13 to examine induction of 15-LO expression (Fig. 7B). Results indicated substantial inhibition of Jak1 expression (ϳ90%) in the monocytes treated with antisense to Jak1, whereas treatment with the control ODN or with the vehicle alone caused no inhibition of Jak1 expression (Fig. 7A). We then examined whether inhibition of Jak1 expression affected the IL-13-induced expression of 15-LO. Results displayed in Fig. 7B indicate that 15-LO expression was not inhibited in monocytes with lower levels of Jak1. In contrast, induction of 15-LO expression was substantially diminished by inhibition of either Jak2 or Tyk2 in confirmation of the results presented in Fig. 6. DISCUSSION IL-4 and IL-13 are the only cytokines that have been shown to induce both the expression and activity of 15-LO (20,21). In both cases, a substantial increase in expression of 15-LO was shown, but the minimum exposure to cytokine required for the effect had not been examined. Our studies indicate that a minimum exposure to IL-13 of 12 h is sufficient to induce the expression of 15-LO (as detected at 36 h), but maximal expression was observed when IL-13 was present for the full 36-h period. The delayed expression of 15-LO may be the result of a complex signaling cascade and likely involves, because of the lengthy time frame, new transcription/translation of an activator or degradation of a protein inhibitor. Inhibition of induction of 15-LO in the IL-13-treated monocytes by potent tyrosine kinase inhibitors suggest that activation and phosphorylation of tyrosine kinases are necessary for 15-LO induction.
Several earlier reports have predicted that the receptors for interleukins are differentially regulated depending on the cell type studied (23)(24)(25)(26). For example, although the level of expression of IL-4R was up-regulated in response to IL-4 in T-cells, IL-4 had no such effect on the IL-4R expression in monocytes (24). It is now quite well established that IL-4R plays a role in the IL-13-mediated signaling in conjunction with another receptor subunit (IL-13R) (28). In T-cells, IL-2R␥c is suggested to be a member of IL-4 and IL-13 cytokine receptor superfamily, but the level of IL-2R␥c is extremely low (40) or absent (24) in human monocytes. Thus, Jak3, which has been shown to be associated with the IL-2R␥c receptor subunit in T-and B-cells and several other cell lines (51) Monocytes were treated with antisense ODN to Jak1, control (Scramble) ODN, or with the cationic lipid vehicle alone by protocols described under "Methods." A, after a 24-h incubation with or without ODN, monocytes were lysed and loaded (50 g/lane) onto a 7.5% SDS-PAGE, transferred to a PVDF membrane, and immunoblotted with antibody to Jak1. The arrow indicates the predicted migration of Jak1 as determined by the migration of molecular weight standards in adjacent wells. B, the untreated or ODN-treated monocytes were treated for an additional 24 h with IL-13 as in Fig. 6, lysed, and loaded (50 g/lane) onto a 7.5% SDS-PAGE. The proteins were resolved, transferred to a PVDF membrane, and immunoblotted with antibody to 15-LO. The arrow indicates the predicted migration of 15-LO (68 kDa) based on the migration of molecular weight markers that were run in adjacent lanes. cytes. Moreover, the level of Jak3 in unactivated monocytes has been shown to be extremely low and is induced upon monocyte activation following IL-2, LPS and IFN-␥ treatment (51). Interestingly, IL-2R␥c expression also is up-regulated in activated monocytes, although nearly absent in freshly isolated monocytes (40). Thus, it seems possible that signaling intermediates in fresh, unactivated as compared with activated monocytes may vary depending upon the level of expression of both kinases and receptor molecules. In fact, Musso et al. (51) have shown that in LPS-activated monocytes, Jak3 is tyrosine-phosphorylated in response to IL-4; in contrast IL-4 did not induce tyrosine phosphorylation of Jak3 in unactivated monocytes. This finding is consistent with our results with IL-13 that clearly showed that phosphorylation of Jak3 was not augmented in fresh monocytes treated with IL-13. These data indicate the participation of tyrosine kinases other than Jak3 in IL-13-treated freshly isolated monocytes.
Our results indicate that Jak2 and Tyk2 were tyrosine-phosphorylated in response to IL-13. Although a basal level of phosphorylated Jak1 was detected, no induction of phosphorylation was observed in response to IL-13. There is an earlier report of Jak2 activation in response to IL-13 in a human ovarian carcinoma cell line (52). Jak1 and Tyk2 phosphorylation was also induced in these carcinoma cells. Jak2 was also phosphorylated in human endothelial cells in response to IL-13 (55). Another study on lymphohematopoietic cells (TF-1) showed that Tyk2 and Jak1 were tyrosine-phosphorylated in response to IL-13 (35), whereas IL-13-induced phosphorylation of Tyk2, but not Jak1 or Jak2, in an Epstein-Barr virus-transformed B cell line (53). In a recent study, Yu et al. (54) reported that NK cells and T cells respond to IL-13 by phosphorylating tyrosines on Jak3, and in murine plasmacytoma cell lines (B9), Tyk2 and Jak3 were shown to be phosphorylated when treated with either IL-4 or IL-13 (39). The disparity of these responses likely reflects the involvement of different IL-13 receptor components as well as variability in expression of the Jak kinases in these distinct cell types.
Further studies investigating the association of Jak2 and Tyk2 with the receptor molecules showed that Jak2 but not Tyk2 was associated and co-immunoprecipitated with the IL-4R, 140 kDa subunit 2 ; thus, the association of Tyk2 with an IL-13R molecule is predicted. Moreover, although in some cells and cell lines an association between Jak1 and the IL-4R␣ chain has been reported, we found that both Jak1 and Jak2 can associate with the IL-4R␣ in the IL-13-treated human monocytes. 2 To more specifically address the participation of Jak2 and Tyk2, both Jak2-and Tyk2-specific antisense and sense ODNs were developed. The sequences were carefully chosen from regions predicted to be without secondary structure and lacking substantial homology with other sequenced human mRNAs or genes. The oligonucleotides were phosphorothioate-modified to limit degradation and purified by HPLC before use to remove all incomplete synthesis products, thereby limiting nonspecific effects. We have found this latter step to be critical in rendering specificity to antisense ODN activity in human monocytes. We found by Western blot analysis that the antisense treatment, but not treatment with control, sense ODNs, resulted in decreased expression of both Jak2 and Tyk2. We also have established that the ODNs were very specific, showing that antisense ODN to Jak2 had no effect on the level of expression of Tyk2, and antisense ODN to Tyk2 did not inhibit Jak2 expression. The monocytes, pretreated with antisense ODN to either Jak2 or Tyk2, could not induce 15-LO expression when treated with IL-13 for 24 h, thereby leading us to believe that expression of both Jak2 and Tyk2 is absolutely necessary for the IL-13-mediated signaling pathway, resulting in the induction of expression of 15-LO. As per the proposed model of the mechanism of action of the Jak/STAT pathway (31), where Jak2 and Tyk2 would phosphorylate each other after ligand (IL-13)-mediated heterodimerization of the activated receptor molecules, our data shows that obliteration of either of the Jak/Tyk would abrogate the downstream effects in the IL-13-mediated signaling pathway.
We additionally inhibited the expression of a tyrosine kinase that was not activated by treatment of monocytes with IL-13, namely Jak1. Jak1 expression was substantially and selectively inhibited by treatment with antisense ODN to Jak1. The inhibition of Jak1 expression had no effect on the induction of 15-LO expression in response to IL-13 treatment. This finding was predicted because tyrosine phosphorylation of Jak1 was not induced by treatment with IL-13 and supports the involvement of select Jak kinases, namely Jak2 and Tyk2, in the induction of 15-LO expression by IL-13.
The results presented here suggest that IL-13-mediated induction of 15-LO expression in human monocytes needs a minimum stimulation of 12 h and requires tyrosine kinase activity. IL-13 treatment of the monocytes induced tyrosine phosphorylation of Jak2 and Tyk2, but not Jak1 or Jak3, within 5 min of treatment. Both of these tyrosine kinases appear to be required for the ultimate induction of 15-LO protein expression in IL-13-treated primary human monocytes.