Differential Expression of Intracellular and Secreted Osteopontin Isoforms by Murine Macrophages in Response to Toll-like Receptor Agonists*

Osteopontin (OPN), expressed by various immune cells, modulates both innate and adaptive immune responses. Different immune cells have shown differential expression of the two isoforms of OPN: secreted form of OPN (sOPN) and intracellular form of OPN (iOPN). However, the molecular mechanisms that control opn gene expression and the OPN isoforms produced by immune cells remain largely unknown. In this study, we demonstrate that OPN mRNA and protein expression are significantly up-regulated upon stimulation with TLR agonists in macrophages. Interestingly, we find that macrophages constitutively express the secreted form of OPN (sOPN), while the intracellular form of OPN (iOPN) is induced following the stimulation with TLR agonists. Phosphoinositide 3-kinase (PI3K), extracellular signal-regulated kinase (ERK), and c-Jun NH2-terminal kinase (JNK) that are activated by LPS stimulation were shown to upregulate OPN expression. In addition, chromatin immunoprecipitation (CHIP) assays showed that AP-1 binds to the proximal AP-1 site in the OPN promoter from LPS-stimulated macrophages. Mutation of the AP-1 site in OPN promoter completely ablates LPS-induced OPN promoter activation. Knockdown of c-Jun and c-Fos expression by small interfering RNA (siRNA) significantly decreases LPS-induced OPN expression. Stable cell lines with iOPN overexpression and knockdown showed that TLR-induced iOPN expression is a negative regulator for interferon-β (IFN-β) production. Our findings provide new insight into the transcriptional regulation of opn gene and further clarify the isoforms and functions of OPN produced by macrophages.

Osteopontin (OPN) 2 is a secreted glycophosphoprotein that is involved in many physiological and pathological processes including immune responses, inflammation, tumor growth, and metastasis, bone formation, and remodeling (1,2). In the immune system, OPN is expressed by various immune cells including macrophages, dendritic cells and T-lymphocytes and has been implicated in many inflammatory autoimmune diseases (3)(4)(5). It is widely accepted that OPN acts as a cytokine with both pro-inflammatory and anti-inflammatory functions. Two isoforms of OPN have been identified. One is the full length OPN with the signal peptide that targets OPN for secretion (sOPN), whereas another is the intracellular form of OPN (iOPN) lacking the OPN signal sequences by alternative translation (6). It was proposed that different immune cells show differential expression of sOPN and iOPN. Antigen-presenting cells (APCs) including dendritic cells (DCs) and macrophages express high levels of iOPN but low levels of sOPN, whereas T cells display a reversed expression pattern (7). Secreted intact OPN or its thrombin-cleaved fragment interacts with receptors, integrins and CD44 variants, to activate PI3K/Akt, NIK/NF-B, or IKK␤/NF-B signal pathways that activate distinct patterns of cytokine/chemokine expression and specific immune responses (8 -12). It has been reported in macrophages that interaction between sOPN and integrin receptors stimulated IL-12 expression, whereas interaction between sOPN and CD44 inhibited IL-10 expression (4). It was also demonstrated that interaction between CD44 and integrins initiates Langerhans cell/DC emigration from the epidermis and attracts them to draining lymph nodes (13). But, the expression and function of iOPN in immune cells are still ill-defined. Recently, TLR9-induced T-bet-dependent expression of iOPN in plasmacytoid dendritic cells (pDCs) was found to mediate IFN-␣ expression by selectively coupling TLR9 signaling with activation of IRF7 (14), whereas iOPN expression in conventional dendritic cells (cDCs) has a key role in promoting T H 17cell responses by suppressing the expression of IL-27 (15).
Macrophages express and secret OPN constitutively. LPS can further up-regulate OPN expression in macrophages (16). Previously, we demonstrated OPN transcription and promoter activity are significantly up-regulated in response to nitric oxide (NO) in LPS-stimulated RAW264.7 murine macrophages through heterogeneous nuclear ribonucleoprotein proteins (hnRNP)-A/B and hnRNP-U proteins (16,17). The stimulation of TLR4 by LPS induces IB kinase (IKKs), mitogen-activated protein kinase (MAPK), and PI3K, ultimately leading to activa-tion of NF-B and AP-1 resulting in the production of various proinflammatory mediators (18,19). A consensus AP-1 binding site (TGACACA between nt Ϫ69 and nt Ϫ75) has been identified in the murine OPN promoter and is responsible for OPN expression in several cell lines (20,21), suggesting that LPSinduced AP-1 activation may play an important role in LPSstimulated OPN expression in macrophages.
In the present study, we investigated the molecular mechanisms of TLR-induced OPN expression in macrophages. The results indicate that TLR stimulation can induce OPN expression through TLR-induced PI3K, JNK, ERK, and AP-1 activation. Importantly, the intracellular form of OPN (iOPN), but not the secreted form of OPN (sOPN), is induced following TLR stimulation in macrophages in vivo and in vitro. In addition, TLR-induced iOPN expression negatively regulates interferon-␤ production in murine macrophages.

EXPERIMENTAL PROCEDURES
Mice and Reagent-C57BL/6J mice were obtained from Joint Ventures Sipper BK Experimental Animal (Shanghai, China). All animal experiments were undertaken in accordance with the National Institute of Health Guide for the Care and Use of Laboratory Animals, with the approval of the Scientific Investigation Board of Shandong University, Jinan, Shandong, P. R. China. LPS (Escherichia coli, 055:B5), LTA and poly(I: C) were purchased from Sigma, and LPS was re-purified as described (22). LY294002, a specific inhibitor of PI3K, SP600125, a specific inhibitor of JNK, PD98059, a specific inhibitor of extracellular signal-regulated kinase (ERK/MEK1), and SB203580, a specific inhibitor of p38 kinase, were purchased from Calbiochem. Antibodies specific to c-Jun, c-Fos, actin, and horseradish peroxidase-coupled secondary antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). An osteopontin antibody (AF808) was obtained from R&D Systems (Minneapolis, MN). OPN 2A1 antibody (SC-21742) was obtained from Santa Cruz Biotechnology.
Plasmid Constructs-The mouse OPN reporter plasmid (Ϫ882/ϩ79) was kindly provided by Dr. David T. Denhardt (Rutgers University, New Brunswick, NJ). The mouse OPN promoter plasmid with the AP-1 binding site mutation was constructed by two-step PCR and described previously (20). The ISRE cis-reporting plasmid was kindly provided by Dr. Hongbing Shu (Wuhan University, China). The expression plasmids for the secreted form of OPN (sOPN) was constructed by RT-PCR with primers: OPN-F, 5Ј-CGCGAATTCCAT-GAGATTGGCAGTGATTTG-3Ј, and OPN-R, 5Ј-CGCGG-ATCCTTAGTTGACCTCAGAAGATG-3Ј; the 885-bp fragment was then inserted into the mammalian expression vector pcDNA3.1/HisB (Invitrogen). The expression plasmid for intracellular form of OPN (iOPN) deleting the codons from 1 to 15 was generated from sOPN expression plasmid by PCR-mediated mutagenesis with the QuikChange II XL Site-directed Mutagenesis kit (Stratagene).
Cell Culture-Female C57BL/6J mice (5-6 weeks old) were used for the preparation of primary mouse macrophages, and thioglycolate-elicited mouse peritoneal macrophages were prepared as described (27). The cells were cultured in endotoxinfree Dulbecco's modified Eagle's medium (DMEM) with 10% fetal calf serum (FCS) (Invitrogen). After 1 h, nonadherent cells were removed. On the next day, the cells were treated with TLR agonists. Mouse macrophage cell line RAW 264.7 was obtained from the American Type Culture Collection (Manassas, VA) and cultured at 37°C under 5% CO 2 in DMEM supplemented with 10% heat-inactivated FCS, 100 units/ml penicillin, and 100 g/ml streptomycin. LPS, poly(I: C), and LTA were used at a final concentration of 100 ng/ml, 10 g/ml, and 1 g/ml, respectively.
Treatment of Animals and Isolation of Peritoneal Lavage Fluid and Serum-Female C57BL/6J mice (4 weeks old) were intraperitoneally injected with thioglycolate to elicit peritoneal macrophages. After 3 days, the mice were treated with PBS or 1.8 mg/kg LPS intraperitoneal administration for indicated time period. At an appropriate time after treatment, mice were anesthetized with halothane and exsanguinated, and the contents of the peritoneal cavity were sampled by peritoneal lavage. To obtain samples for cytokine assessment, 1 ml of PBS was injected intraperitoneal, the abdominal area was massaged to distribute the fluid, the skin over the peritoneal cavity (but not the peritoneal lining) was removed to allow visualization of the fluid, and a sample (ϳ0.7 ml) from the peritoneal cavity was removed using a needle (25 gauge) and syringe. After centrifugation (300 ϫ g for 5 min), the cell pellet was saved, and the supernatant was stored at Ϫ20°C until needed for cytokine assay. If analysis of cells was to be part of the experiment, an additional 7 ml of PBS was injected into the peritoneal cavity, and the steps outlined above were repeated to obtain the remaining cells. The cell pellets from the 1-and 7-ml lavages were pooled and used for subsequent analysis.
RNA Interfering-For transient transfection, siRNA specific for c-Jun, c-Fos, and control siRNA were purchased from Santa Cruz Biotechnology. siRNA duplexes were transfected into RAW 264.7 macrophages using Lipofectamine 2000 Transfection Reagent (Invitrogen) according to the standard protocol.
Overexpression and Knockdown of iOPN in Macrophages-To stably overexpress or knockdown iOPN in macrophages, RAW 264.7 cells were transfected with the iOPN expression plasmid or OPN shRNA plasmid and then selected with 800 g/ml G418 (Invitrogen) for 2-3 weeks, and then the cells were pooled, expanded, and used for the following experiments. RAW 264.7 cells stably transfected with empty vector pcDNA3.1 were used as control cells.
Quantitation of Cytokines by ELISA-2 ϫ 10 5 macrophages were seeded into 24-well plates and incubated overnight. The cells were stimulated for indicated time periods. The concentrations of secreted osteopontin and TNF-␣ in culture super-natants were measured by ELISA kits (R&D Systems, Minneapolis, MN). For cytokine quantitation from peritoneal lavage fluid and serum, peritoneal lavage fluid and serum were prepared as above, and secreted osteopontin and TNF-␣ were measured by ELISA kits (R&D Systems, Minneapolis, MN).
Western Blot-Cells were lysed with M-PER Protein Extraction reagent (Pierce) supplemented with a protease inhibitor 'mixture'; protein concentrations in the extracts were measured with a bicinchoninic acid assay (Pierce). OPN dephosphorylation and deglycosylation were performed using bovine alkaline phosphatase and peptide N-glycosidase F, respectively as described previously by Christensen et al. Equal amounts of extracts were separated by SDS-PAGE, then were transferred onto nitrocellulose membranes for immunoblot analysis as described previously (22).
Chromatin Immunoprecipitation (ChIP) Assay-RAW264.7 macrophages were stimulated with 100 ng/ml LPS for 2 h or left unstimulated. Chromatin from macrophages was fixed and immunoprecipitated using the ChIP assay kit as recommended by the manufacturer (Upstate Biotechnology, Inc.). The puri-fied chromatin was immunoprecipitated using 2 g of anti-c-Jun and c-Fos or 2 g of irrelevant antibody (anti-actin). The input fraction corresponded to 0.1 and 0.05% of the chromatin solution before immunoprecipitation. After DNA purification, the presence of the selected DNA sequence was assessed by PCR. The primers were 5Ј-CTCATGGTAGTTCGTTGC-TTTA-3Ј, 5Ј-TCTCATCCTT-AGCAAGGAAAAG-3Ј for OPN promoter (Ϫ156ϳϩ6) and 5Ј-GTC-CAAATAGAACATCTTACTC-3Ј, 5Ј-TAAAGCAACGAACTACCA-TGAG-3Ј for OPN promoter (Ϫ309ϳϪ136). The PCR program was: 94°C ϫ 4 min; followed by 94°C ϫ 45 s, 55°C ϫ 45 s, and 72°C ϫ 45 s for a total of 28 cycles; and then 72°C ϫ 7 min. PCR products were resolved in 10% acrylamide gels. The average size of the sonicated DNA fragments subjected to immunoprecipitation was 500 bp as determined by ethidium bromide gel electrophoresis.
Assay of Luciferase Reporter Gene Expression-Assay of luciferase reporter gene expression was performed as previously described (27). Briefly, RAW 264.7 cells (1.5 ϫ 10 4 cells/well) were seeded onto 96-well plates 24 h before co-transfection with 100 ng of various reporter plasmids and 25 ng of Renilla-TK plasmid using Jet-PEI transfection reagent (Polyplus). In some cases, iOPN expression plasmid was cotransfected; the total amount of transfected plasmid was equalized by empty vector pcDNA3.1. Twentyfour hours after transfection, the cells were left untreated or treated with LPS or IFN-␥ for 6 h. Luciferase activities were measured using Dual-Luciferase Reporter Assay system (Promega) on a LMAXII luminometer (Molecular Devices) according to the manufacturer's instructions. Firefly luciferase activity was normalized against Renilla luciferase activity.
Statistical Analysis-All data are presented as means Ϯ S.E. of three or four experiments. Analysis was performed using a Student's t test. Values of p Ͻ 0.05 were considered significant.

TLR-induced OPN Gene Expression in Macrophages-We
first examined OPN gene expression in macrophages stimulated with LPS (TLR4 ligand) for various times by RT-PCR and real-time quantitative PCR. As shown in Fig. 1A, constitutive expression of OPN mRNA was found in unstimulated control RAW264.7 murine macrophages. Upon stimulation with LPS, a significant increase in expression of OPN mRNA was detected in RAW264.7 macrophages. Expression of OPN mRNA in RAW264.7 macrophages reached the peak level after stimulated with LPS for 4 h (Fig. 1A). To further confirm LPS-induced OPN mRNA expression in macrophages, thioglycolate-elicited mouse primary peritoneal macrophages were used. In a similar fashion, expression of OPN mRNA was up-regulated by LPS treatment (Fig. 1B); however, OPN mRNA expression reached the peak level in these cells 8 h after LPS treatment (Fig. 1B).
To confirm whether other TLR agonists can also induce OPN mRNA expression in macrophages, LTA (TLR2 ligand) and poly(I:C) (TLR3 ligand) were used to stimulate macrophages. Similarly, stimulation with both LTA and poly (I:C) greatly enhanced OPN mRNA expression 4 h after stimulation in both RAW 264.7 macrophages and peritoneal macrophages (Fig. 1, C and D). Taken together, these data indicate that TLR signaling is associated with significantly increased OPN mRNA expression in macrophages.
OPN Protein Expression and Secretion in TLR-stimulated Macrophages-Two isoforms of OPN, secreted form of OPN (sOPN) and intracellular form of OPN (iOPN), have been identified. Immunoblot analysis and ELISA were used to determine iOPN protein expression in the cell lysate and sOPN secretion in TLR-stimulated macrophages, respectively. Unstimulated RAW264.7 macrophages expressed and secreted detectable levels of OPN (Fig. 2, A and D), consistent with the constitutive expression of OPN mRNA as shown in Fig. 1A. Consistent with the LPS-induced expression of OPN mRNA as shown in Fig. 1A, OPN protein level in the cell lysate was greatly increased by LPS stimulation (Fig. 2D). In contrast, the level of secreted OPN remained the same as that of unstimulated macrophages during a 24-h period of LPS stimulation ( Fig. 2A). As a control, TNF-␣ secretion was greatly increased by LPS stimulation ( Fig. 2A), indicating LPS stimulation and the secretory machinery are not impaired in these LPS-stimulated macrophages.
Similarly, primary peritoneal macrophages expressed and secreted low detectable levels of OPN, and LPS further increased OPN protein expression in the cell lysate (Fig. 2E). But, the level of secreted OPN remained unchanged and showed the same expression as that of unstimulated macrophages during a 24-h period of LPS stimulation (Fig. 2B). As a control, TNF-␣ secretion was greatly increased by LPS stimulation (Fig. 2B) in peritoneal macrophages.
TLR2 ligand LTA and TLR3 ligand poly (I:C) were used to stimulate primary peritoneal macrophages, and OPN protein expression in the cell lysate and OPN secretion were measured as above. Immunoblot showed that OPN protein expression in the cell lysate was greatly induced in both RAW264.7 macrophages (Fig. 2D) and peritoneal macrophages (Fig. 2E), consistent with the increased OPN mRNA after stimulation (Fig.  1, C and D). Similar to the LPS stimulation, the level of secreted OPN displayed a comparable level to that of unstimulated control peritoneal macrophages (Fig. 2C). TNF-␣ secretion was also greatly increased by LTA and poly(I:C) stimulation (Fig. 2C).
Finally, immunoblot analysis was performed using OPN 2A1 Ab that has been used successfully to detect both forms of OPN (6). As shown in Fig. 2F (upper panel), two protein bands were detected in LPS-stimulated peritoneal macrophages. Importantly, the upper band (ϳ60 kDa), corresponding the secreted OPN, remained unchanged after LPS stimulation, whereas, the lower band (ϳ54 kDa), corresponding to the intracellular form of OPN, was greatly increased by LPS stimulation. OPN is a glycosylated phosphoprotein. OPN from control and LPS-stimulated cells was dephosphorylated and deglycosylated using bovine alkaline phosphatase and peptide N-glycosidase F, respectively. After dephosphorylation and deglycosylation, two protein bands (ϳ50 kDa and ϳ44 kDa) can still be detected with OPN 2A1 Ab (Fig. 2F, lower panel), suggesting that the two protein bands are not the full-length protein whose post-translational modifications have been altered by TLR signaling.
Collectively, these data suggest that macrophages differentially express both forms of OPN, sOPN is expressed constitutively in macrophages, while iOPN is expressed in TLR-activated macrophages. TLR signaling can further increase iOPN protein expression.
Differential Expression of sOPN and iOPN in Vivo-To further confirm differential OPN expression in macrophages, we measured OPN secretion in blood and peritoneal fluid in a mouse model of endotoxemia by intraperitoneal injection of LPS. Female C57BL/6J mice (4 weeks old) were first injected with thioglycolate to elicit peritoneal macrophages. After 3 days, the mice were treated with PBS or LPS intraperitoneal administration (1.8 mg/kg) for indicated time period. Blood and peritoneal fluid were obtained at 4, 8, 24 h after LPS intraperitoneal injection, and OPN secretion was measured by ELISA. As shown in Fig. 3A, serum OPN level was greatly increased as early as 4 h after LPS injection, and continued to reach the peak level at 24 h after LPS injection. In contrast, OPN

. TLRs induce OPN expression in vivo.
A, female C57BL/6J mice (4 weeks old) were intraperitoneally injected with thioglycolate to elicit peritoneal macrophages. After 3 days, the mice were treated with PBS or 1.8 mg/kg LPS intraperitoneal administration for the indicated time period. Secreted OPN and TNF-␣ production in the serum were detected by ELISA. B, female C57BL/6J mice (4 weeks old) were treated as in A, Secreted OPN and TNF-␣ production in the lavage were detected by ELISA. C, female C57BL/6J mice (4 weeks old) were treated as in A, intracellular OPN expression in the peritoneal macrophages was detected by Western blot. Actin was used as a cytoplasmic protein loading control. The expression levels were quantitated by measuring band intensities and shown as fold increase relative to background. D and E, primary splenocytes from female C57BL/6J mice (4 weeks old) were treated with 100 ng/ml LPS as indicated. Secreted OPN in supernatants was measured by ELISA. Intracellular OPN expression was detected by Western blot. Actin was used as a cytoplasmic protein loading control. Data are shown as mean Ϯ S.D. (n ϭ 3) of one representative experiment. level in peritoneal fluid remained unchanged compared with that of PBS-simulated mouse (Fig. 3B). As a control, TNF-␣ secretion in both serum and peritoneal fluid was greatly increased in LPS-challenged mice (Fig. 3, A and B). To confirm OPN expression was induced by LPS intraperitoneal injection, the peritoneal macrophages at various time points after LPS administration were isolated and cell lysate was used to detect iOPN protein expression. Immunoblot showed that intracellular OPN protein level was significantly increased in the peritoneal macrophages at various time points with LPS intraperitoneal administration (Fig. 3C).
Shinohara et al. (23) has reported that activated T cells are a potent source of secreted OPN in vitro and in vivo. Consistent with their results, we found that primary splenocytes expressed and secreted large amounts of OPN with LPS stimulation (Fig.  3, D and E). This may account for the OPN in the blood from mice with LPS stimulation. Collectively, these results suggest that macrophages express sOPN constitutively, whereas iOPN is expressed from TLR-activated macrophages. Secreted OPN in the blood from LPS-stimulated mice may be from other immune cells.

PI3K Is Involved in LPS-induced iOPN Expression in Macrophages-
The stimulation of TLR4 by LPS induces PI3K activation, and the PI3K pathway plays a very important role in LPS signaling (24). To investigate the role of PI3K in LPS-induced OPN expression, the PI3K-specific inhibitor LY294002 and wortmannin were used to inhibit PI3K activation in the setting of LPS-induced iOPN expression. LY294002 and wortmannin pretreatment completely inhibited Akt phosphorylation, indicating PI3K activation in LPS-stimulated macrophages was inhibited by LY294002 and wortmannin treatment (data not shown). With the inhibition of PI3K activation by LY294002 and wortmannin pretreatment, LPS-induced OPN mRNA expression was significantly decreased compared with that of DMSO-treated control cells at 2 and 4 h after LPS stim-ulation (Fig. 4, A and B). LY294002 and wortmannin pretreatment resulted in a 7-and 6-fold decrease in OPN mRNA expression as measured by Q-PCR at 4 and 8 h after LPS stimulation, respectively (p Ͻ 0.01). Similarly, LPS-induced OPN protein expression was also significantly decreased in the presence of LY294002 and wortmannin at 4 and 8 h after LPS stimulation (Fig. 4C). These data indicate that LPS-induced PI3K activation is required for increased OPN mRNA and iOPN protein expression in macrophages.

MAP Kinases JNK and ERK Are Involved in LPS-induced iOPN Expression in Macrophages-
The stimulation of TLR4 by LPS induces the activation of MAP kinases including JNK, ERK, and p38 leading to activation of NF-B and AP-1. To investigate the role of MAP kinases in LPS-induced OPN expression, the JNK-specific inhibitor SP60012, ERK specific inhibitor PD98089 and p38 inhibitor SB203580 were used in the setting of LPS-induced OPN expression. As shown in Fig.  5, A and B LPS-induced OPN mRNA expression at 2 and 4 h after LPS stimulation was significantly decreased by JNK inhibitor SP60012 treatment and ERK inhibitor PD98089 treatment (p Ͻ 0.01), but not by P38 kinase inhibitor SB203580 treatment (p Ͼ 0.05). Similarly, LPS-induced OPN protein expression at 4 and 8 h after LPS stimulation was also greatly decreased by JNK inhibitor SP60012 treatment and ERK inhibitor PD98089 treatment, but not by P38 kinase inhibitor SB203580 treatment (Fig.  5C). Taken together, these results demonstrate that LPS-induced MAP kinases JNK and ERK activation is associated with significantly increased OPN mRNA and iOPN protein expression in macrophages.

AP-1 Is Involved in LPS-induced iOPN Expression in
Macrophages-AP-1 is often activated in LPS-stimulated macrophages downstream of the MAP kinase. Moreover, a consensus AP-1 binding site has been identified in the OPN promoter and is responsible for OPN expression in several cell lines (20,21). We investigated the functions of AP-1 in the LPS-induced OPN expression upon binding to the OPN promoter.
To show that AP-1 proteins bind to the consensus AP-1 site (nt Ϫ69 and nt Ϫ75) directly, CHIP experiments were performed. PCR analysis showed that c-Jun and c-Fos antibody precipitated the OPN promoter region (nt Ϫ156 to nt ϩ6) from RAW264.7 cells activated with LPS for 2 h (Fig. 6A), whereas unstimulated controls did not demonstrated this DNA binding. Furthermore, PCR analysis showed that c-Jun and c-Fos did not bind the OPN promoter region from nt Ϫ309 to Ϫ136 (Fig. 6A). Finally, control antibodies (anti-actin antibody) did not exhibit this DNA binding activity (Fig. 6A). These data suggest that AP-1 binds specifically to the AP-1 cis-regulatory region of the OPN promoter in LPS-stimulated macrophages.
To confirm the function of AP-1 in LPS-induced OPN expression, OPN promoter luciferase reporter construct with point mutations in the AP-1 site was used (Fig. 6B). WT OPN promoter construct demonstrated a 4-fold increase in luciferase activity by LPS stimulation. In contrast, mutation of the AP-1 binding site completely ablated this increase in luciferase activity (p Ͻ 0.01), suggesting the AP-1 site is essential for LPSinduced OPN promoter activation.
To assess the functional relevance of AP-1 in the setting of LPS-induced OPN expression, c-Jun and c-Fos siRNA was designed to inhibit c-Jun and c-Fos expression. C-Jun and c-Fos siRNA greatly decreased both c-Jun and c-Fos expression in RAW264.7 macrophages as measured by immunoblot analysis (Fig. 6C, left). Then, OPN protein expression in LPS-stimulated RAW264.7 macrophages was measured by immunoblot analysis. Transfection of c-Jun and c-Fos siRNA significantly decreased LPS-induced OPN expression (Fig. 6C, right). These results suggest that the AP-1 site (nt Ϫ69 and nt Ϫ75) of the OPN promoter is a binding site for AP-1, which is essential for OPN promoter activation and OPN protein expression with LPS stimulation.
TLR-induced iOPN Expression Negatively Regulates Interferon-␤ Production-TLR9-induced expression of iOPN was found to mediate IFN-␣ expression in pDCs (14). To investigate the role of TLR-induced iOPN expression in Type I interferon production in macrophages, RAW264.7 macrophage stable cell lines with high expression and knockdown of iOPN were constructed. The overexpression and knockdown of iOPN was confirmed by immunoblot analysis (Fig. 7, A and C). IFN-␤ production from OPN knockdown stable cells was greatly increased by ϳ50% compared with that of control siRNA-transfected cells (p Ͻ 0.01) after LPS stimulation (Fig. 7B). Exogenous addition of recombinant OPN protein to the culture medium of OPN knockdown stable cells did not reverse the increase of IFN-␤ production (data not shown). In contrast, as shown in Fig. 7D, after overexpression of His-tagged iOPN in RAW264.7 cells, LPS-induced IFN-␤ production was significantly decreased by ϳ50% compared with that of control plasmid transfected cells (p Ͻ 0.01). Then cotransfection experiments with IFN-␤ promoter reporter plasmid and iOPN expression plasmid were performed (Fig. 7E). Cotransfection of iOPN expression plasmid inhibited LPSinduced IFN-␤ reporter gene activity by ϳ40% (p Ͻ 0.01). These data indicate that iOPN is a negative regulator for LPS-induced IFN-␤ expression and secretion in macrophages.
LPS-induced IFN-␤ production can activate specific target genes through ISREs in macrophages. To confirm the function of iOPN mediated down-regulation of IFN-␤ production, transient transfection assays were performed using the ISRE cisreporting vector together with iOPN expression plasmid (Fig.  7E). LPS induced ISRE-dependent luciferase activity was significantly lower by ϳ80% in iOPN expression plasmid-transfected cells compared with that of control vector-transfected cells (p Ͻ 0.01). Taken together, these data indicate that LPS-induced iOPN expression negatively regulates INF-␤ production and subsequent INF-␤ signaling.

DISCUSSION
OPN is a ϳ298-amino acid secreted phosphoprotein and contributes to diverse physiological and pathological processes including immune responses, inflammation, tumor growth, and metastasis, bone formation, and remodeling (5,25). Its expression is tissue-specific and subject to regulation by many transcription factors (2). Macrophages constitutively express and secret low levels of OPN. Its expression can be further upregulated by LPS stimulation (16). But, the molecular mechanisms of LPS-mediated induction of OPN expression and whether other TLR agonists can induce OPN expression are not well defined.
Previously, we demonstrated a nitric oxide (NO)-dependent mechanism to control OPN transcription in LPS-stimulated macrophages through heterogeneous nuclear ribonucleoprotein proteins (hnRNP)-A/B and hnRNP-U proteins (16,17). However, macrophage NO production was initially detected following 8 h of LPS treatment; peak NO production was noted after 12 h of stimulation (data not shown). We detected OPN transcription as early as 1 h following LPS simulation with the peak expression after 4 h of stimulation (Fig. 1A). In addition, inhibition of NO synthesis by pretreatment with L-NAME can't ablate LPS-induced OPN expression (data not shown). Therefore, a NO-independent mechanism may be present to regulate LPS-induced OPN expression in macrophages.
TLR signaling through ligation with TLR agonists induces various kinases including IB kinase (IKKs), mitogen-activated protein kinase (MAPK), and PI3K, resulting in the production of various proinflammatory mediators and type I interferon (18,19). The transcription of proinflammatory mediators such as TNF-␣ and IL-6 is very rapid and transient after TLR stimulation. The nature of OPN rapid transcription indicates that TLR-induced kinase activation and subsequent activation of transcription factors may involved in TLR-induced OPN expression. We found that treatment with the JNK specific inhibitor SP60012, ERK specific inhibitor PD98089, and PI3K specific inhibitor LY294002 significantly decreased OPN expression to a level equivalent to that of unstimulated controls. PI3K involvement in OPN expression has been shown in melanoma and breast cancer cells (26,20). LPS-induced JNK and ERK activation can be inhibited by PI3K specific inhibitor LY294002, but not p38 activation (27). Consistently, we found that OPN expression was not decreased by P38 kinase inhibitor SB203580, indicating LPS-induced PI3K and subsequent JNK and ERK activation are involved in LPS-induced OPN expression. We speculate that NO-independent and PI3K/JNK/ERK-dependent mechanism mediate early expression of OPN (1-8 h of stimulation), whereas NO-dependent mechanism can mediate late expression of OPN (8 -24 h of stimulation). Consistent with this speculation, OPN transcription can be detected during a period to 24 h of LPS stimulation. A number of transcription factors have been found to control OPN expression in immune cells. For example, OPN expression in activated T cells and pDCs depends on the transcription factor T-bet (14,23), an essential factor in T H 1 lineage commitment. But, transcription factors controlling OPN expression in macrophages remain to be determined. AP-1 is often activated in TLR-stimulated macrophages downstream of the MAP kinase and PI3K. A variety of inflammatory genes are under the control of AP-1-dependent signaling pathways. CHIP assays revealed AP-1 binding to the consensus site in the OPN promoter after LPS stimulation in macrophages. The importance of this AP-1 site in LPS-induced OPN promoter activity was further demonstrated by a complete loss of inducible OPN promoter activity after site-directed mutagenesis of this AP-1 site. The transcription factor AP-1 consists of a variety of dimers composed of members of the Jun and Fos families of proteins (28). The Jun proteins can both homo-and heterodimerize with Fos members to form transcriptionally active complexes. Consistently, c-Jun and c-Fos were found to bind to OPN promoter in CHIP assays, and RNAi knockdown of c-Jun and c-Fos . Involvement of AP-1 in LPS-induced OPN expression. A, RAW264.7 macrophages were stimulated with 100 ng/ml LPS or vehicle (water) for 2 h, and the ChIP assay was used to assess the binding of c-Jun and c-Fos to the AP-1 binding sites within the Ϫ156 to ϩ6 region of the murine OPN promoter. Total extract was used as a loading control, and immunoprecipitation with irrelevant antibody (anti-actin) was used as a negative control. PCR products from the amplication of an AP-1 site-free region within Ϫ309 to Ϫ136 of the murine OPN promoter were used as specificity controls. B, RAW264.7 macrophages were transfected with 100 ng of OPN luciferase reporter construct or the AP-1 binding sites mutant reporter construct and 10 ng pTK-Renillaluciferase plasmid. After 24 h of culture, the cells were stimulated with 100 ng/ml LPS for 6 h. Luciferase activity was measured and normalized by Renilla luciferase activity. Data are shown as mean Ϯ S.D. (n ϭ 6) of one representative experiment (**, p Ͻ 0.01). C, RAW264.7 macrophages were transfected with control small RNA (Ctrl siRNA), c-Jun siRNA, or c-Fos siRNA. After 36 h, c-Jun or c-Fos expression in the cells was detected by Western blot. Similar results were obtained in three independent experiments. D, RAW264.7 macrophages were transfected with Ctrl siRNA, c-Jun siRNA or c-Fos siRNA and then treated with 100 ng/ml LPS for the indicated time period. The production of intracellular OPN was detected by Western blot with R&D OPN Ab. Similar results were obtained in three independent experiments. expression ablated LPS-induced OPN expression. These data indicate that heterodimers composed of c-Jun and c-Fos can bind to OPN promoter and regulate OPN expression in LPSstimulated macrophages.
Two forms of OPN have been identified. One is the fulllength OPN with the signal peptide that targets OPN for secre-tion (sOPN), whereas another is the intracellular form of OPN lacking the OPN signal sequences by alternative translation (6). It was proposed that different immune cells show differential expression of the two forms of OPN. APCs including DCs and macrophages express high levels of iOPN but low levels of sOPN, whereas T cells display a reversed expression pattern (7). In accordance with this proposal, we found that both forms of OPN are expressed in macrophages. Macrophages express a low level of secreted form of OPN (sOPN) constitutively. TLR signaling can further up-regulate OPN expression in macrophages and the main form of TLR-induced OPN is iOPN. The differential expression of sOPN and iOPN was further confirmed by measuring OPN secretion in blood and peritoneal fluid in a mouse model of endotoxemia by intraperitoneal injection of LPS. Differential expression of sOPN and iOPN in immune cells suggests sOPN and iOPN have different functions in the immune responses. Indeed, sOPN has been shown to interact with integrins and CD44 to regulate cell migration (29), cell activation and survival (30,31), apoptosis (32,33), and proinflammatory cytokine production (4,34). Whereas, iOPN has been shown to regulate cell motility (9,35) and cytokine expression (14,15).
TLR9-induced expression of iOPN was found to increase IFN-␣ expression in pDCs (14). Type I IFN has been shown to suppress iOPN expression in cDCs (15). Both type I IFN and iOPN are induced in macrophages early after exposure to TLR stimulation. But, the cross-regulation between iOPN and type I IFN production in macrophages is not clear. In contrast to the function of iOPN in pDCs, we found that TLR-induced iOPN is a negative regulator for INF-␤ production and subsequent INF-␤ signaling in macrophages. This feedback regulation may be very important to further increase iOPN expression through attenuation of the suppression effect of type I IFN in macrophages. The functions of iOPN especially the cross-regulation between iOPN and type I IFN in macrophages are currently being investigated.
In conclusion, our results demonstrate that TLR-induced OPN expression in macrophages requires TLR-induced PI3K, JNK, ERK and AP-1 activation. Macrophages express a low level of secreted form of OPN constitutively. TLR signaling can further up-regulate iOPN expression in macrophages and increased iOPN expression can attenuate IFN-␤ production in TLR-stimulated macrophages. Therefore, our findings delineate a new molecular mechanism involved in the transcriptional regulation of OPN and clarify the forms and functions of OPN produced by macrophages.