Interleukin (IL)-32β-mediated CCAAT/Enhancer-binding Protein α (C/EBPα) Phosphorylation by Protein Kinase Cδ (PKCδ) Abrogates the Inhibitory Effect of C/EBPα on IL-10 Production*

Background: IL-32β promotes IL-10 production in myeloid cells. Results: IL-32β-mediated C/EBPα serine 21 phosphorylation by PKCδ induced the dissociation of C/EBPα from IL-10 promoter, thereby promoting IL-10 production. Conclusion: IL-32β suppressed the inhibitory effect of C/EBPα on IL-10 production by mediating C/EBPα serine 21 phosphorylation by PKCδ. Significance: Our data suggest that IL-32β functions as an intracellular regulator of IL-10 production. We previously reported that IL-32β promotes IL-10 production in myeloid cells. However, the underlying mechanism remains elusive. In this study, we demonstrated that IL-32β abrogated the inhibitory effect of CCAAT/enhancer-binding protein α (C/EBPα) on IL-10 expression in U937 cells. We observed that the phosphorylation of C/EBPα Ser-21 was inhibited by a PKCδ-specific inhibitor, rottlerin, or IL-32β knockdown by siRNA and that IL-32β shifted to the membrane from the cytosol upon phorbol 12-myristate 13-acetate treatment. We revealed that IL-32β suppressed the binding of C/EBPα to IL-10 promoter by using ChIP assay. These data suggest that PKCδ and IL-32β may modulate the effect of C/EBPα on IL-10 expression. We next demonstrated by immunoprecipitation that IL-32β interacted with PKCδ and C/EBPα, thereby mediating C/EBPα Ser-21 phosphorylation by PKCδ. We showed that IL-32β suppressed the inhibitory effect of C/EBPα on IL-10 promoter activity. However, the IL-10 promoter activity was reduced to the basal level by rottlerin treatment. When C/EBPα serine 21 was mutated to glycine (S21G), the inhibitory effect of C/EBPα S21G on IL-10 promoter activity was not modulated by IL-32β. Taken together, our results show that IL-32β-mediated C/EBPα Ser-21 phosphorylation by PKCδ suppressed C/EBPα binding to IL-10 promoter, which promoted IL-10 production in U937 cells.

(TNF-␣), and IL-8 (1). Multiple studies have shown the association of IL-32 with various inflammatory diseases, such as rheumatoid arthritis, inflammatory bowel disease, and chronic obstructive pulmonary disease (2,3). In addition to those inflammatory diseases, viral and bacterial infections and even cancers induce IL-32 expression. However, the majority of protein is detected in cellular lysates rather than in the supernatants (2, 4 -6). Recently, the interactions of IL-32 with intracellular proteins, such as focal adhesion kinase 1 (FAK 1), paxillin, integrins, protein kinase C␦ (PKC␦), protein kinase C⑀ (PKC⑀), and STAT3, have been demonstrated, which implies that IL-32 may participate in inflammatory responses as an intracellular mediator (7)(8)(9).
The CCAAT/enhancer-binding protein (C/EBP) 3 family of transcription factors is known to play important roles in cell proliferation and differentiation. The C/EBP family comprises six members: C/EBP␣, C/EBP␤, C/EBP␥, C/EBP␦, C/EBP⑀, and C/EBP. Each member, except C/EBP, which lacks a canonical basic region, contains a similar basic region and leucine zipper sequence at its C terminus that mediate DNA binding and dimerization, respectively (10,11). C/EBP␣ dimerizes with other members of the C/EBP family or with itself (12) and interacts with other proteins such as transcription factor II B (TFIIB), TATA-binding protein (TBP), retinoblastoma protein (Rb), p300/CREB-binding protein (CBP), p21, and members of the SWI/SNF complex (13)(14)(15)(16)(17). C/EBP␣ is known to up-regulate the expression of granulocytic lineage-specific genes (18 -21), both directly and synergistically with other key genes such as core-binding factor (CBF) complex genes and PU.1 in myeloid development. C/EBP␣ also represses E2F-regulated genes by direct interaction with E2F, which induces cell cycle arrest (19,22).
We previously reported that IL-32␤ promotes IL-10 production in myeloid cells (28). In this study, we further verified our results and demonstrated an intracellular mediatory role of IL-32␤ for IL-10 production by showing that IL-32␤ promoted IL-10 production in a PKC␦-dependent way, where it mediated the phosphorylation of C/EBP␣ Ser-21 by PKC␦. The phosphorylation of C/EBP␣ Ser-21 suppressed its inhibitory role in IL-10 induction. Our data suggest that IL-32␤ may function as an intracellular regulator of inflammatory response as well as an inducer of inflammation.
Measurement of IL-10 Expression Levels and Enzyme-linked Immunosorbent Assay (ELISA)-IL-10 mRNA expression was detected by reverse transcription-polymerase chain reaction (RT-PCR) of total RNAs extracted from U937 cells after the various treatments with 10 nM PMA, 1 g/ml LPS, or 10 g/ml poly(I:C) for 24 h. U937 cells were also treated with increasing amounts of the PKC␦-specific inhibitor rottlerin for 24 h. The culture media were collected for ELISA. The IL-10 primer set used for PCR was as follows: sense 5Ј-AACCTGCCTAACAT-GCTTCGA-3Ј; antisense 5Ј-CTCATGGCTTTGTAGATGC-CT-3Ј. Human IL-10, TNF␣, and IL-8 ELISAs were performed using ELISA kits from R&D systems ( Minneapolis, MN).
Electroporation-U937 promyelomonocytic cells were transfected with 1.5 g of IL-32 siRNA or nontargeting siRNA using a Neon TM transfection system (Invitrogen). PKC␦ siRNA was transfected in the same way. PKC␦ and IL-32 siRNA and nontargeting siRNA were purchased from Dharmacon (Lafayette, CO). The transfected cells were incubated overnight, and 10 nM PMA was then applied for the indicated time, after which cell lysates were prepared for Western blotting.
Chromatin Immunoprecipitation (ChIP) Assay-We used the commercially available ChIP assay kit (Millipore) according to the manufacturer's instructions. Briefly, U937 cells were electroporated with IL-32 siRNA or nontargeting RNA. After overnight incubation, cells were treated with PMA for 90 min. The cells were fixed with 1% formaldehyde, lysed in kit lysis buffer, and sonicated with 5 pulses for 5 s each. After centrifugation at 13,000 rpm for 20 min, the supernatants were precleared with 45 l of protein A-agarose/salmon sperm DNA (50% slurry) for 60 min. After a brief centrifugation, the supernatants were mixed with 3 g of C/EBP␣ antibody and maintained overnight. Sixty microliters of protein A-agarose/ salmon sperm DNA (50% slurry) was added to each sample, and the pulled-down DNA fragments were eluted. PCR amplification using the eluted DNA as the template was performed for 35 cycles at an annealing temperature of 61°C. The primers for PCR amplification of the IL-10 promoter were as follows: sense (from Ϫ577 to Ϫ557) 5Ј-CTTTGAGGATATTTAGCC-CAC-  Immunofluorescence staining was performed to localize the endogenous IL-32␤ in U937 cells. Cells were treated with 50 nM PMA or 1 g/ml LPS for 40 min. Cells were fixed, permeabilized, and labeled for nuclei with DAPI and endogenous IL-32 with FITC. Fluorescence signals were analyzed by confocal microscopy. The same amount of mock IgG (1 g) as KU32-52 antibody was used for control.
Statistical Analysis-Statistical analysis was performed using the unpaired two-tailed Student's t test. Differences were considered statistically significant at p Ͻ 0.05.

IL-32␤ Promotes IL-10 Production in U937 Cells upon PMA
Stimulation-We demonstrated previously that IL-32␤ promoted the production of the anti-inflammatory cytokine IL-10 in myeloid cells (28). To verify that IL-32␤ is involved in IL-10 production, we knocked down IL-32␤ by using siRNA in U937 cells (Fig. 1A). Only the production of IL-10 was decreased upon PMA treatment after IL-32␤ silencing, whereas the levels of TNF-␣ and IL-8 were not changed (Fig. 1). The ectopic introduction of IL-32␤ into U937 cells further increased IL-10 production (data not shown). Therefore, these data confirmed that IL-32␤ up-regulated the production of IL-10 by PMA treatment in U937 cells.

IL-32␤
Associates with PMA-activated PKC␦-We localized IL-32␤ in U937 cells by using confocal microscopy. Under normal conditions, IL-32␤ was distributed in the cytosol, but it shifted to the membrane upon PMA stimulation. LPS treatment did not result in the movement of IL-32␤ in U937 cells (Fig. 2). This finding suggests that IL-32␤ played an indirect mediatory role rather than a role as a direct transcriptional activator for IL-10 expression. Because the effect of IL-32␤ on IL-10 production was PMAdependent, and its membrane localization upon PMA stimulation resembled the movement of PKC, we screened the relevant PKCs by immunoprecipitation, as we had previously performed with IL-32␣ (7). IL-32␤ specifically interacted with PMA-activated PKC␦ because the interaction was inhibited by a PKC␦-specific inhibitor, rottlerin, but was not inhibited by a classical PKC inhibitor, Gö6976 (Fig. 3, A and B). Meanwhile, several studies demonstrated that PKC␦ is involved in IL-10 production (29 -31). Thus, we examined whether PKC␦ was involved in IL-10 production in U937 cells as well. When U937 cells were treated with increasing doses of rottlerin, the production of IL-10 was decreased in a rottlerin dose-dependent manner (Fig. 3C). The TNF-␣ level was also affected by rottlerin, but IL-8 was not (Fig. 3, D and E). These data suggest that IL-32␤ may be inter-related with PKC␦ for IL-10 production.
Serine 21 of C/EBP␣ Is Phosphorylated by PKC␦ in an IL-32␤involved Way in U937 cells-The human promyelomonocytic U937 cells undergo monocytic differentiation by PMA stimu- lation. A tissue-specific transcription factor, C/EBP␣, blocks 12-O-tetradecanoylphorbol-13-acetate-induced monocytic differentiation of bipotential myeloid cells when ectopically expressed (32) and induces growth arrest by repressing proliferation genes (10). Hence, we investigated the status of C/EBP␣ in U937 cells. C/EBP␣ was expressed in U937 cells and was phosphorylated at serine 21 by PMA stimulation (Fig. 4A). The phosphorylation of C/EBP␣ Ser-21 was inhibited by the treatment of rottlerin, a PKC␦-specific inhibitor, which means that C/EBP␣ Ser-21 could be phosphorylated by PKC␦ as well. C/EBP␣ Ser-21 phosphorylation was also inhibited by PD98059, an ERK1/2 inhibitor, to a lesser extent than rottlerin, as consistent with the findings of a previous study (26) (Fig. 4B). To further verify that PKC␦ is involved in C/EBP␣ Ser-21 phosphorylation, PKC␦ was knocked down with 80% efficiency compared with GAPDH by using PKC␦-specific siRNA. As shown in Fig. 4C, phospho-C/EBP␣ at Ser-21 was significantly decreased by PKC␦ silencing, which means that PKC␦ is also involved in C/EBP␣ Ser-21 phosphorylation. We then examined whether IL-32␤ might affect the phosphorylation of C/EBP␣. When IL-32␤ was knocked down by siRNA, the phosphorylation of Ser-21 of C/EBP␣ was not induced upon PMA treatment (Fig. 4D), which suggests that IL-32␤ involved is in C/EBP␣ Ser-21 phosphorylation. To analyze the relevance of C/EBP␣ to IL-10 expression, we performed ChIP for the putative C/EBP␣-binding site on IL-10 promoter after silencing of IL-32␤. We observed that the binding of C/EBP␣ to IL-10 promoter was increased by IL-32␤ knockdown. However, C/EBP␣ binding to Bcl-2 or lactoferrin promoter was not affected by IL-32␤ silencing. We further verified the ChIP results by quantitative real-time PCR (Fig. 5). This finding implies that IL-32␤ suppressed C/EBP␣ binding to IL-10 promoter, thereby relieving the IL-10 promoter from C/EBP␣ repression. ERK1/2 is known to directly phosphorylate C/EBP␣ Ser-21 through an FXFP docking site. However, we speculated that PKC␦ might phosphorylate C/EBP␣ Ser-21 in a IL-32␤-mediated way, which inhibited the binding of C/EBP␣ to IL-10 promoter.
IL-32␤ Interacts with PKC␦ and C/EBP␣ and Mediates C/EBP␣ Ser-21 Phosphorylation-We next investigated how IL-32␤ mediated C/EBP␣ Ser-21 phosphorylation by PKC␦. IL-32␤ association with PMA-activated PKC␦ was shown in Fig. 3, A and B. We then examined whether C/EBP␣ interacted with these molecules together. We cloned human C/EBP␣ cDNA into 6ϫMyc-tagged expressing vector and then performed immunoprecipitations after co-transfection of HEK293 with different combinations of IL-32␤, C/EBP␣, and PKC␦. IL-32␤ co-immunoprecipitated with C/EBP␣ and PKC␦, and this interaction was inhibited by rottlerin, but not by PD98059, an ERK1/2 inhibitor (Fig. 6, A and B). The interaction of C/EBP␣ with IL-32␤ was also confirmed in U937 cells. The interaction was inhibited by the treatment of rottlerin, but not by PD98059 (Fig. 6C). We then proved that IL-32␤ mediated the phosphorylation of C/EBP␣ Ser-21 by PKC␦. We transfected HEK293 cells with 6ϫMyc-tagged C/EBP␣ with or without IL-32␤ and then immunoprecipitated C/EBP␣ with Myc antibody. C/EBP␣ Ser-21 was not phosphorylated in the absence of IL-32␤. The phosphorylation of Ser-21 was inhibited by rottlerin, but the phospho-C/EBP␣ Ser-21 was still detected even in the presence of a ERK1/2 inhibitor, PD98059 (Fig. 6D), which means that PKC␦ also phosphorylated C/EBP␣ Ser-21 in a IL-32␤-mediated way. When C/EBP␣ serine 21 was mutated to glycine, it was no longer phosphorylated (Fig. 6E). However, the interaction of C/EBP␣ with PKC␦ and IL-32␤ was not inhibited by the mutation of serine 21 to glycine (S21G). We reconfirmed the interaction of IL-32␤ with mutant C/EBP␣ S21G in A549 lung carcinoma cells, which means that the phos- The phosphorylation of C/EBP␣ at serine 21 was detected by phospho-C/EBP␣ (S21)-specific antibody. Quantitation is shown below the blots, normalized to 0 min (for 10, 30, and 60 min). B, U937 cells were treated with 10 M rottlerin (Rott) or 10 M PD98059 (PD) for 1 h; thereafter, they were treated with 10 nM PMA for a further 60 min. Quantitation is shown below the blots, normalized to nontreated control for the others. C, U937 cells were transfected with 2 g of PKC␦ siRNA and nontargeting siRNA (NT). Cells were treated with 10 nM PMA for 30 min after overnight incubation, and then Western blotting was performed. D, 1.5 g of IL-32 siRNA and nontargeting siRNA (NT) was introduced into U937 cells. After overnight incubation, cells were treated with 10 nM PMA for the indicated time. IL-32␤ band (25 kDa) was detected with KU32-52 antibody. After an overnight incubation following transfection, 10 nM PMA was treated for 90 min. ChIP was carried out with 3 g of C/EBP␣ antibody. Quantitation is shown below the PCR data, normalized to control (con) nontargeting siRNA for the others. The consensus binding sequence of C/EBP and a putative C/EBP␣-binding site and its location on IL-10 promoter are presented. ChIP results from Bcl-2 and lactoferrin promoters are presented as controls. IL-10 ChIP products were also measured by quantitative real-time PCR. The results are given as -fold change of the IL-10/lactoferrin ChIP ratio as compared with the value of nontreated nontargeting siRNA. All values are mean Ϯ S.E. (n ϭ 3).
phorylation of C/EBP␣ Ser-21 did not influence its association with IL-32␤ and PKC␦ (Fig. 6, F and G). We also characterized the interaction domains of C/EBP␣ and IL-32␤. IL-32␤ inter-acted with the full-length C/EBP␣ (FL) as well as F2 and F4 (Fig.  7A). F2 contains the DNA-binding domain and the basic leucine zipper domain, and F4 contains only the basic leucine zip- After co-transfection of HEK293 cells with 6ϫMyc-tagged IL-32␤ and C/EBP␣, cells were treated with inhibitors and PMA in the same way as in A. D, cell lysates were immunoprecipitated with 1 g of Myc tag antibody; the phosphorylated C/EBP␣ at serine 21 was detected using phospho-C/EBP␣ Ser-21-specific antibody. E and F, a mutant C/EBP␣ S21G was co-transfected with IL-32␤ and PKC␦ into HEK293 cells, and then immunoprecipitations were performed with Myc tag antibody (E) or KU32-52 antibody (F). G, A549 lung carcinoma cells were co-transfected with a mutant C/EBP␣ S21G, IL-32␤, and PKC␦. Each plasmid was also transfected as controls. Immunoprecipitation was carried out with Myc tag antibody. The expression levels of the introduced genes were determined by Western blotting with 30 g of whole cell lysates (WCL). HEK293 cells were co-transfected with 6ϫMyctagged C/EBP␣ mutants and 5ϫFLAG-tagged IL-32␤. After overnight incubation, cells were treated with 20 nM PMA for 90 min. Immunoprecipitation was performed with 1 g of FLAG antibody. B, a schematic illustration of IL-32␤ deletion mutants is presented. HEK293 cells were co-transfected with 5ϫFLAGtagged C/EBP␣ and 6ϫMyc-tagged IL-32␤ mutants. Immunoprecipitation was performed as in A with 1 g of FLAG antibody. The pulled down D1 is marked by arrow. The expression levels of the transfected genes were determined by Western blotting with 30 g of whole cell lysates (WCL). Immunoglobulin G light chain (IgG L) is indicated by an arrow. per domain. Therefore, these data mean that IL-32␤ binds to the basic leucine zipper region of C/EBP␣. C/EBP␣ bound to the N-terminal region of IL-32␤ (Fig. 7B). Most isoforms of IL-32 contain the same C terminus (D2), but the N terminus (D1) is diverse because of alternative splicing. Meanwhile, the two mutants of IL-32␤ had no effect on IL-10 production, which means that only the intact form of IL-32␤ is functional (data not shown).
IL-32␤ Suppressed the Inhibitory Effect of C/EBP␣ on IL-10 Expression-We further investigated the effect of C/EBP␣ on IL-10 expression using IL-10 promoter reporter assay. We observed that IL-10 promoter was not activated under C/EBP␣ expression without IL-32␤. However, in the presence of IL-32␤, the promoter was activated (Fig. 8A). The promoter activity was inhibited by rottlerin, but not by PD98059, which means that ERK1/2 was not involved in IL-10 expression in this system. The activity of C/EBP␣ is repressed by the phosphorylation of Ser-21 (26); as is consistent with this study, a mutant C/EBP␣ S21G, which could not be phosphorylated by PKC␦, repressed IL-10 promoter activity even under co-expression with IL-32␤ (Fig. 8B). When the cells were treated with rottlerin under mutant C/EBP␣ S21G expression, the promoter activity was further reduced, but was not modulated by PD98059 (Fig. 8C). These data therefore suggest that IL-32␤-mediated phosphorylation of C/EBP␣ Ser-21 by PKC␦ suppressed the inhibitory effect of C/EBP␣ on IL-10 production.

DISCUSSION
The initiation of the proinflammatory response is a prerequisite for the proper immune responses upon viral or bacterial infections. Although the proinflammatory reaction is essential, it should be resolved to prevent damage to the host. IL-32 has been studied for its proinflammatory function because multiple studies have shown its association with inflammatory diseases. IL-32␤ is known to induce TNF-␣, IL-8, IL-6, and macrophage inflammatory protein-1␣ (MIP-1␣), all of which are proinflammatory mediators. However, we showed that IL-32␤ induced an anti-inflammatory cytokine IL-10 in myeloid cells. In this study, we questioned the mechanism of action of IL-32␤ because IL-32␤ protein was detected only with cellular lysates, not in the culture medium. In this study, we solved our previous question by revealing the mechanism by which IL-32␤ promotes IL-10 production. Recently, several studies have shown the intracellular mediatory function of IL-32 (7,8,33). We demonstrated that IL-32␤ mediated the phosphorylation of serine 21 on C/EBP␣ by PMA-activated PKC␦, which resulted in IL-10 production. Phosphorylation of C/EBP␣ at Ser-21 inhibits its transcriptional activity through induction of conformational changes. The involvement of PKC␦ in IL-10 production is known, but there is little evidence of the effect of C/EBP␣ on IL-10 production. Although a previous study showed that C/EBP␣ induced IL-10 production (34), the effect was synergistic with cAMP-responsive element-binding protein/activating transcription factor (CREB/ATF). C/EBP␣ not only activates the granulopoiesis-related genes, but also represses cell cyclerelated genes, which means that C/EBP␣ acts as a transcriptional activator as well as a transcriptional inhibitor. There are a couple of models explaining the inhibitory function of C/EBP␣. C/EBP␣ may inhibit S-phage genes indirectly through an association with E2F, or by direct binding to E2F sites. Alternatively, C/EBP␣ may directly inhibit S-phage genes by binding to the C/EBP-binding site on promoters of those genes (10). On the grounds of these studies, we examined the effect of IL-32␤ on C/EBP␣ and showed that IL-32␤ inhibited C/EBP␣ binding to IL-10 promoter by mediating Ser-21 phosphorylation by PKC␦. However, whether the inhibition of the IL-10 gene by C/EBP␣ results from the binding of C/EBP␣ to IL-10 promoter or the interaction of other inhibitory proteins with C/EBP␣ located on the IL-10 promoter remains unclear. In any event, according to our results, IL-32␤ has an intracellular mediatory function in addition to its proinflammatory function.