Interleukin-22 (IL-22) activates the JAK/STAT, ERK, JNK, and p38 MAP kinase pathways in a rat hepatoma cell line. Pathways that are shared with and distinct from IL-10.

IL (interleukin)-22 is an IL-10-related cytokine; its main biological activity known thus far is the induction of acute phase reactants in liver and pancreas. IL-22 signals through a receptor that is composed of two chains from the class II cytokine receptor family: IL-22R (also called ZcytoR11/CRF2-9) and IL-10Rbeta (CRF2-4), which is also involved in IL-10 signaling. In this report, we analyzed the signal transduction pathways activated in response to IL-22 in a rat hepatoma cell line, H4IIE. We found that IL-22 induces activation of JAK1 and Tyk2 but not JAK2, as well as phosphorylation of STAT1, STAT3, and STAT5 on tyrosine residues, extending the similarities between IL-22 and IL-10. However our results unraveled some differences between IL-22 and IL-10 signaling. Using antibodies specific for the phosphorylated form of MEK1/2, ERK1/2, p90RSK, JNK, and p38 kinase, we showed that IL-22 activates the three major MAPK pathways. IL-22 also induced serine phosphorylation of STAT3 on Ser(727). This effect, which is not shared with IL-10, was only marginally affected by MEK1/2 inhibitors, indicating that other pathways might be involved. Finally, by overexpressing a STAT3 S727A mutant, we showed that serine phosphorylation is required to achieve maximum transactivation of a STAT responsive promoter upon IL-22 stimulation.

In this report, we show that binding of IL-22 to its surface receptor on rat hepatoma cell line H4IIE induced the rapid activation of JAK1 and Tyk2, leading to phosphorylation of STAT1, STAT3, and STAT5. IL-22 also activated the three major MAPK pathways: the MEK-ERK-RSK, the JNK/SAPK, and the p38 kinase pathways. In addition, IL-22 induced phosphorylation of STAT3 on a serine residue. This further STAT modification was necessary for maximum transactivation and depended only marginally on the ERK pathway.
Plasmid Construction, Cell Transfection, and Luciferase Assay-The human IL-22R cDNA was amplified by reverse transcription PCR from the HepG2 hepatoma cell line before cloning into the pEF-BOSpuro expression vector (21). Mouse IL10R␣ cDNA was amplified by reverse transcription PCR from MC9 cells and also cloned into the pEF-BOSpuro expression vector. The pSG5-STAT3 wild-type and the pSG5-STAT3 S727A vectors encoding wild-type or mutated forms of human STAT3 were obtained as described (22). Expression of JAK1 and JAK2 was driven by the pRK5 vector as described by Dr. J. Ihle's group (23).
12.10 6 H4IIE cells were electroporated (250 V, 200 ohms, 1200 microfarads) with 50 g of pGRR5-luc (provided by Dr. P. Brennan, Imperial Cancer Research Fund) or 30 g of pSRE-luc plasmid (Stratagene, La Jolla, CA). The first construct contains five copies of the STAT-binding site of the Fc␥RI gene inserted upstream from a luciferase gene controlled by the TK promoter, and the second one contains repeats of the serum responsive element of the c-fos promoter. 5 g of another reporter plasmid, pRL-TK (Promega, Madison, WI) encoding renilla luciferase, was co-electroporated as an internal control of the transfection process. Cells were seeded in 12-wells plates at 10 6 /ml. The next day, cells were stimulated for 3 h with 2000 units/ml IL-22 before lysis. When indicated, cells were preincubated 1 h with 50 M PD98059 or 10 M U0126 MEK inhibitors. Luciferase assays were performed using the dual luciferase reporter assay kit (Promega).
Transient HEK293 cells transfections were carried out as described previously (24). Briefly, cells were seeded in 12-well plates at 4 ϫ 10 5 cells/well 1 day before transfection. Transfection was carried out using the LipofectAMINE method (Invitrogen), according to the manufacturer's recommendations, with 2 g of plasmid DNA in a total of: 500 ng of hIL-22R or mIL-10R cDNA, 100 ng of pGRR5, 100 ng of pRL-TK, 1 g of vector encoding wild-type or mutated forms of STAT3, and empty vector to 2 g. 5 h after transfection, cells were stimulated with control medium, with IL-22 (2000 units/ml) or human IL-10 (10 ng/ml) for 24 h. Luciferase assays were performed using the dual luciferase reporter assay kit (Promega). The same protocol was used for transient transfection of 2C4, U4C, and ␥2A fibrosarcoma cells with 500 ng of hIL-22R cDNA, 40 ng of pRK5-JAK1, pRK5-JAK2, or empty vector, 100 ng of pGRR5, and 100 ng of pRL-TK.  (Fig. 1A). This phosphorylation was transient, decreasing to barely detectable levels for STAT1 and STAT5 after 30 min. However, phosphorylated STAT3 could still be detected at least 1 h after IL-22 stimulation (data not shown). To confirm that IL-22-induced STAT phosphorylation correlated with transcriptional activation, we used luciferase assays with pGRR5-luc reporter plasmid. The pGRR5-luc construct is regulated by five copies of a STAT-binding se-quence recognizing at least STAT1, STAT3, and STAT5. As shown in Fig. 1B, when H4IIE cells were electroporated with pGRR5-luc, IL-22 stimulation induced a 35-fold increase in luciferase activity.

IL
IL-22 Activates JAK1 and Tyk2-As JAK kinases are known to be responsible for STAT phosphorylation in response to cytokines, we next investigated which JAK kinase is activated by IL-22. H4IIE rat hepatoma cells were stimulated with control medium or IL-22, and a Western blot analysis was performed with an anti-phospho-Tyk2 antibody, while immunoprecipitations were used for JAK1 and JAK2. As shown in Fig.  2, A and B , with 250 units/ml IFN-␥, or with control supernatant for 5 min before lysis and immunoprecipitation with antibodies directed against JAK1 or JAK2. Western blot analysis was performed with an anti-phosphotyrosine antibody, and the membranes were then reprobed with anti-JAK1 or anti-JAK2 antibody. Similar results were obtained when H4IIE cells were stimulated with E. coli-derived IL-22 (data not shown). C, 400,000 2C4 parental, U4C JAK1-deficient, or ␥2A JAK2-deficient cells were seeded in 12-well plates. The next day, cells were transfected with a vector coding for IL-22R together with pGRR5-luc and pRL-TK reporter plasmids and, when mentioned, with a vector coding for JAK1 or JAK2 cDNA. Cells were stimulated with IL-22 (2000 units/ml) or with control medium for 4 h before a luciferase assay was performed. phosphorylation of Tyk2 and JAK1 but not of JAK2. As a control for JAK2 immunoprecipitation, H4IIE cells were stimulated with IFN-␥, which is known to activate this JAK family member as well as JAK1 (25). To further assess the functional role of JAK1 in IL-22 signaling, we transfected JAK1-deficient U4C cells with the IL-22R cDNA together with pGRR5-luc reporter plasmid. As shown in Fig. 2C, IL-22 failed to induce a luciferase activity in U4C cells unless these cells were transfected with JAK1 cDNA. By contrast, parental cells (2C4) or JAK2-deficient cells (␥2A) were both able to respond to IL-22.
IL-22 Activates MAPK Pathways-We next analyzed the ability of IL-22 to activate the MAP kinase pathways. As shown in Fig. 3A, IL-22 induced a sustained phosphorylation of ERK1/2. Two MEK inhibitors, PD98059 and U0126, totally blocked this phosphorylation, suggesting that ERK1/2 phosphorylation results from MEK activation. In line with this result, IL-22 induced the phosphorylation of MEK1/2 (Fig. 3B). p90RSK, a well known substrate of ERK, was also phosphorylated in response to IL-22. To confirm the functional activation of this pathway, we electroporated H4IIE cells with the pSREluc reporter plasmid regulated by the serum-responsive element from the c-fos promoter. As shown in Fig. 4, IL-22 stimulation induced a 2.25-fold increase in luciferase activity, which was completely abolished when cells were preincubated with any of the MEK inhibitors. In contrast with IL-22, IL-10 did not activate the ERK MAPK pathway in H4IIE cells transfected with IL-10R␣ cDNA (Fig. 5). In addition to this MEK-ERK-RSK cascade, IL-22 also induced a delayed phosphorylation of JNK/SAPK and p38 MAP kinases, detectable after 30 -40 min (Fig. 3B).

IL-22 Induces STAT3 Serine Phosphorylation by a MAPKindependent
Mechanism-In addition to tyrosine phosphorylation, STAT3 can be phosphorylated on a serine residue in response to cytokines such as IL-6 (22). In our system, IL-22 stimulation of H4IIE cells induced rapid serine phosphorylation of STAT3. Although MAPKs have often been described as mediating STAT3 Ser phosphorylation (22,26,27), preincubation of H4IIE cells with MEK inhibitors only slightly delayed but did not inhibit this effect (Fig. 6, A and B).
IL-22-induced STAT3 Serine Phosphorylation Is Necessary for Maximal Transactivation-To test the functional significance of STAT3 serine phosphorylation, we transfected H4IIE cells with the pGRR5-luc and pRL-TK reporter plasmids together with a plasmid encoding wild-type or a mutated form of STAT3, in which the serine 727 is mutated into alanine, thus preventing phosphorylation (28). As shown in Fig. 7, mutation of Ser 727 of STAT3 reduced luciferase induction from an 8-fold to a 4-fold increase upon IL-22 stimulation, strongly suggesting that STAT3 serine phosphorylation is required to achieve maximal transactivation.
To further support this hypothesis, we studied the effect of this STAT3 S727A mutant on IL-10-induced transactivation because IL-10 has never been described to phosphorylate STAT3 on a serine residue. HEK293 cells were transiently transfected with a plasmid encoding either IL-22R or IL-10R cDNA together with a plasmid encoding the wild-type or S727A mutated form of STAT3, pGRR5-luc, and pRL-TK reporter plasmids. As shown in Fig. 8A, IL-22 stimulation of IL-22Rtransfected HEK293 cells induced a 6.5-fold increase in lucif-erase activity. Co-transfection of the STAT3 S727A mutant reduced this induction to 4-fold. By contrast, co-transfection of the STAT3 S727A mutant in cells expressing IL-10R had no effect on IL-10-induced transactivation (Fig. 8A). In line with this result, IL-22, but not IL-10, induced STAT3 serine phosphorylation in these cells, whereas both cytokines induced tyrosine phosphorylation of STAT3 (Fig. 8B). DISCUSSION IL-22, a new cytokine that is structurally related to IL-10, was originally identified as an IL-9-induced gene and shown to up-regulate the acute phase response in the liver (1, 6). In addition to their sequence homology, IL-22 and IL-10 share one receptor subunit, IL-10R␤, whereas their functional receptor complex also involves IL-22R for IL-22 and IL-10R␣ for IL-10 (7,8,13). In this paper, we have analyzed the signaling pathways activated by IL-22 in rat hepatoma cells. We showed that IL-22 induced the phosphorylation of JAK1 and Tyk2, but not JAK2, and that JAK1 is absolutely required for IL-22 signaling. Because Tyk2 has been shown to be associated with IL-10R␤ (13), these results suggest that JAK1 associates with IL-22R. This observation extends the similarities between IL-22 and IL-10, which also activates JAK1 and Tyk2. Moreover, both cytokines induce the phosphorylation of the same STAT factors (12)(13)(14). However, our results unravel some of the differences between IL-22 and IL-10 signaling. First, IL-22 induces the activation of the ERK, JNK, and p38 MAPK pathways that are not activated by IL-10 (2). Secondly, IL-22, but not IL-10, induces serine phosphorylation of STAT3. Thus, despite the structural relationship between IL-22 and IL-10, these cytokines activate overlapping but not identical signaling pathways.
IL-22 activity and signaling pathways are reminiscent of those of IL-6 on hepatocytes. Indeed, IL-22 has been shown to induce acute phase proteins in the liver (6), an effect that has been described to be regulated by IL-6 mainly through STAT3 activation (30 -32). Although the IL-6 receptor complex is less related to the IL-22 receptor than the IL-10 receptor complex, IL-6 binding also leads to JAK1 activation and the subsequent phosphorylation of STAT3, STAT1, and STAT5 (33). Moreover, IL-6 also activates the MAPK pathways (34,35). A synergistic effect between the JAK/STAT and MAPK pathways has been proposed to regulate acute phase protein expression in response to IL-6 (36). Our results suggest that such a synergy can also take place in the regulation of the acute phase response by  Accumulating data have stressed the importance of STAT regulation by serine phosphorylation. In this report, we have shown that STAT3 Ser 727 phosphorylation is induced upon IL-22 stimulation and is required for maximal transcriptional activation. Similar results were previously reported for IL-6 (22). By contrast, IL-10 signaling does not involve STAT3 Ser 727 phosphorylation, indicating that this process is not shared by all STAT3-activating cytokines.
In the case of IL-6, STAT3 serine phosphorylation results from the sequential activation of a Vav-Rac-1-SEK-1/MKK-4-PKC␦ cascade (22,37). However, we failed to show any IL-22induced PKC␦ activation (data not shown). STAT serine phosphorylation can also be mediated by MAPKs such as ERK (38), MEKK1 (39), JNK (27), or p38 kinase (40). Interestingly, IL-22-induced STAT3 serine phosphorylation was only slightly delayed but not inhibited by MEK inhibitors PD98059 and U0126, suggesting that this effect could result from several cooperating signaling pathways. This hypothesis is further supported by a recent report showing that both H7 (a PKC inhibitor) and PD98059 inhibitors were required to block STAT3 serine phosphorylation induced by high IL-6 concentrations (41). However, we have failed thus far to identify the kinases responsible for IL-22-induced STAT3 serine phosphorylation by combining signaling pathway inhibitors (data not shown).
This report is the first description of signal transduction by one of the recently described IL-10-related cytokines (IL-22, IL-19, IL-20, mda-7/IL-24, AK155/IL-26; reviewed in Refs. 3 and 4). Our data show both similarities (JAK1, Tyk2, and STAT3 tyrosine phosphorylation) and discrepancies (STAT3 serine phosphorylation, MAPKs activation) between IL-22 and IL-10 signaling. Little is known about the pathways activated by the other members of this family, except that they all induce STAT3 tyrosine phosphorylation (29). 2 Further studies will be needed to determine whether signaling pathways can be used to define subclassifications within this family.