Intracellular osteopontin stabilizes TRAF3 to positively regulate innate antiviral response

Osteopontin (OPN) is a multifunctional protein involved in both innate immunity and adaptive immunity. However, the function of OPN, especially the intracellular form OPN (iOPN) on innate antiviral immune response remains elusive. Here, we demonstrated that iOPN is an essential positive regulator to protect the host from virus infection. OPN deficiency or knockdown significantly attenuated virus-induced IRF3 activation, IFN-β production and antiviral response. Consistently, OPN-deficient mice were more susceptible to VSV infection than WT mice. Mechanistically, iOPN was found to interact with tumor necrosis factor receptor (TNFR)-associated factor 3 (TRAF3) and inhibit Triad3A-mediated K48-linked polyubiquitination and degradation of TRAF3 through the C-terminal fragment of iOPN. Therefore, our findings delineated a new function for iOPN to act as a positive regulator in innate antiviral immunity through stabilization of TRAF3.

The innate immunity is the first line of defense against invading pathogens, which functions to respond to infection directly and relays signals for the activation of the adaptive immunity 1 . During viral infection, multiple signaling pathways in the innate immune system are triggered to promote the production of cytokines to suppress viral replication 2 . Central to the host antiviral response is the production of type I interferons (IFNs), which include IFN-α and IFN-β . Several classes of germline-encoded pattern-recognition receptors (PRRs) have been linked to the production of type I interferons during viral infection. These PRRs include Toll-like receptors (TLRs), retinoic acid-inducible gene I (RIG-I) like receptor (RLRs) and intracellular DNA sensors [3][4][5] . For example, TLR3 recognizes viral double-stranded RNA in endosomes and triggers a signaling pathway mediated by Toll/ IL-1R (TIR) domain-containing adaptor that induces IFN-β (TRIF) 6 . TLR4 also uses TRIF as adaptor to induce IFN-β production 7 . RIG-I is an important cytoplasmic PRR for the detection of positive-and negative-stranded RNA viruses, including Sendai virus (SeV), vesicular stomatitis virus (VSV), hepatitis C virus (HCV), and influenza A virus (IAV) 8,9 . The recognition of viral RNA by RIG-I leads to the RIG-I conformation change and the recruitment of the downstream mitochondrial antiviral signaling protein (MAVS) (also called IPS-1, Cardif or VISA) through the CARDs [10][11][12][13] . After recruitment of TRIF and MAVS, TLR3/4 and RIG-I activate convergent pathways composed of tumor necrosis factor receptor (TNFR)-associated factor 3 (TRAF3) 14 , TANK-binding kinase 1 (TBK1)/Iκ -B kinase ε (IKK-ε ) and IFN regulatory factor 3 (IRF3), leading to the production of IFN-β 15 . Intracellular DNA from invading pathogens could also induce IFN-β production through a very similar pathway composed of cyclic-GMP-AMP (cGAMP) synthase (cGAS), stimulator of interferon genes protein (STING), TBK1/IKK-ε and IRF3 16 .
TRAF3 is a member of the cytoplasmic signaling protein family called tumor necrosis factor receptor (TNFR)-associated factors (TRAFs), which are composed of 7 members and used by a large and diverse group of receptors including TLR, TNFR and RLR. TRAF3 was first identified to directly associate with CD40 and inhibit CD40-mediated NF-κ B activation in B cells 17,18 . Subsequent studies demonstrated that TRAF3 is crucial for TLR3/4-induced type I IFN production by macrophages and DCs 19,20 . Later on, it was shown that TRAF3 was also involved in the regulation of RLR-induced IFN production 14,21 . Activation of TLR3/4 and RLR signaling OPN protein and mRNA expression in peritoneal macrophages (Fig. 1A). ELISA analysis showed that sOPN was also increased upon virus infection (Fig. S1A). Similar to SeV infection, activation of the intracellular DNA receptor signaling by ISD (interferon-stimulating DNA) and cGAMP (cyclic GMP-AMP) induced OPN expression (Fig. S1B). Consistently, infection with a DNA virus Herpes simplex virus-1 (HSV-1) also increased OPN protein expression ( Fig. S1C) All together, these data demonstrated that OPN expression is induced by virus infection in murine peritoneal macrophages.
OPN positively regulates IFN-β production. To investigate the function of OPN in innate antiviral immune response, peritoneal macrophages were prepared from WT and OPN-deficient (Spp1 −/− ) mice and infected with SeV for various times. Then, the expression of IFN-β was measured. SeV infection induced the expression of IFN-β mRNA in WT macrophages. However, SeV-induced IFN-β mRNA expression was greatly decreased in OPN-deficient macrophages compared to that in WT macrophages (Fig. 1B). Consistently, OPN-deficient macrophages secreted less IFN-β protein than WT macrophages after SeV infection (Fig. 1C). The expression of CXCL10, Mx1 and CCL5, which are downstream genes of IFN-β signaling pathway, also decreased in SeV-infected OPN-deficient macrophages (Fig. 1B). Similar to SeV infection, VSV infection-induced expression of IFN-β , CXCL10, Mx1 and CCL5 also greatly decreased in OPN-deficient macrophages (Fig. 1D). LPS and poly(I:C) (polyinosinic: polycytidylic acid) stimulation, which activate TLR4 and TLR3 signaling respectively, induced less IFN-β production in OPN-deficient macrophages compared to that in WT macrophages (Fig. S2A). ISD transfection, which was shown to activate intracellular DNA receptors signaling, also led to a decreased IFN-β production in OPN-deficient macrophages compared to that in WT macrophages (Fig. S2B).
We further investigated the function of OPN on IFN-β expression using overexpression experiments. Transfection of iOPN and full length OPN expression plasmids into HEK293 cells increased SeV-and VSV-induced IFN-β expression (Fig. 1E). Similar to IFN-β mRNA expression, transfection of iOPN and full length OPN expression plasmids also increased SeV-induced IFN-β promoter activation in a dose-dependent manner (Fig. 1F). Further, we found that addition of OPN antibody into the culture medium could not inhibit full length OPN transfection-mediated IFN-β activation induced by SeV infection (Fig. 1G). Overexpression of iOPN also increased RIG-I-, melanoma differentiation-associated gene 5 (MDA5)-, TRIF-and cGAS+ STING-induced IFN-β promoter activation in a dose-dependent manner (Fig. 1H). In all circumstances, intracellular OPN seemed more potent to induce IFN-β expression than the full length OPN. Taken together, these data indicated that OPN, especially iOPN, positively regulates IFN-β production downstream of various innate immune signaling pathways including TLR3/4, RLRs and intracellular DNA receptor signaling.
OPN potentiates antiviral response. IFN-β plays an essential role in antiviral immune response 25 . To investigate the role of OPN in antiviral response, VSV was used to infect cells. Plaque assays showed that VSV replication greatly increased in peritoneal macrophages prepared from OPN-deficient mice compared to that from WT mice in the presence or absence of poly(I:C) ( Fig. 2A). Consistently, VSV RNA was also increased in OPN-deficient macrophages compared to that in WT macrophages ( Fig. 2A). In contrast, overexpression of iOPN and the full length OPN in HEK293 cells greatly attenuated VSV replication in the presence or absence of poly(I:C) (Fig. 2B). Taken together, these data indicated that OPN positively regulates antiviral immune response.
To investigate the physiological role of OPN in antiviral response in vivo, Spp1 −/− mice and WT mice were infected with VSV, and the antiviral immune responses were examined. The amount of IFN-β protein induced by VSV infection was much less in sera of VSV-infected Spp1 −/− mice than that of WT mice (Fig. 2C). In accordance with reduced IFN-β production, VSV replication in the livers, spleens, and lungs was much higher in OPN-deficient (Spp1 −/− ) mice than in WT controls (Fig. 2D). Importantly, Spp1 −/− mice were more susceptible to VSV infection than WT mice (Fig. 2E). Spp1 −/− mice all died, while 50% of WT mice were alive 5 days after infection. These data suggested that OPN is an important positive regulator of IFN-β production and antiviral immune responses.
iOPN positively regulates IRF3 activation. IRF3 is the main transcription factor responsible for IFN-β transcription during the early phase of viral infection 26 . To investigate the function of OPN on IRF3 activation, series of experiments were performed. First, IFN-β PRD I/III reporter, which harbors only IRF3 binding site in IFN-β promoter, was used 27 . RIG-I, TRIF-and cGAS+ STING-induced-IFN-β PRD I/III activation was increased by iOPN overexpression in a dose-dependent manner (Fig. 3A). IRF3 activation requires the phosphorylation of conserved serine and theronine residues at the c-terminal region 28 . SeV infection induced IRF3 phosphorylation in macrophages from WT mice (Fig. 3B). While, SeV-induced IRF3 phophorylation was greatly decreased in macrophages from OPN-deficient (Spp1 −/− ) mice (Fig. 3B). Similarly, VSV infection-induced IRF3 phosphorylation was also greatly decreased in macrophages from Spp1 −/− mice compared to that from WT mice (Fig. 3B). LPS-and poly(I:C)-induced IRF3 phosphorylation was similarly decreased in macrophages from Spp1 −/− mice compared to that from WT mice ( Fig. S3A and B). In contrast, overexpression of iOPN in HEK293 cells substantially increased SeV-and VSV-induced IRF3 phosphorylation (Fig. 3C). After phosphorylation, IRF3 dimerizes and translocates into nuleus to initiate IFN-β transcription 28 . IRF3 dimerization was greatly decreased in OPN-deficient (Spp1 −/− ) macrophages compared to that in WT macrophages after SeV infection (Fig. 3D). Western blot analysis of cytoplasmic fraction and nuclear fraction showed that more IRF3 was translocated into nucleus in macrophages from WT mice compared to that from OPN-deficient (Spp1 −/− ) mice after virus infection (Fig. 3E). All together, these data demonstrated that OPN positively regulates IRF3 activation to regulate IFN-β production and antiviral response.
To directly identify iOPN targets, immunoprecipitation (IP) and western blotting (WB) were performed in HEK293 cells transfected with expression plasmids for RIG-I, MAVS, TRAF3, TNF receptor-associated factor 6 (TRAF6), STING, TBK1 and IRF3 together with iOPN. As shown in Fig. 4C, iOPN was shown to interact with TRAF3, but not with RIG-I, MAVS, STING, TBK1 and IRF3. TRAF3 has been shown to be required for the IFN-β expression downstream of TLR3/4 and RLR signaling 19,20 . Interestingly, iOPN could not interact with TRAF3 homologue TRAF6, which has been shown to activate NF-κ B, leading to the production of proinflammatory cytokines (Fig. 4D). Interaction between endogenous TRAF3 and OPN also detected in macrophages after SeV infection (Fig. 4E).
To confirm iOPN interacts with TRAF3 directly, iOPN and TRAF3 were expressed in an in vitro protein expression system, then mixed together and followed by pull-down assays with anti-OPN antibody. As shown in Fig. 4F, TRAF3 could coimunoprecipitate with OPN, indicating a direct interaction between iOPN and TRAF3.
The interactions were further supported by the colocalization studies. iOPN-GFP was found to diffuse in the cytoplasm and nucleus without SeV infection (Fig. 4G). TRAF3 was present in the cytoplasm exclusively (Fig. 4G). iOPN-GFP and TRAF3 showed less or no colocalization without SeV infection (Fig. 4G). SeV infection induced translocation of large amount of iOPN from nucleus into cytoplasm, where colocalization between iOPN and TRAF3 was greatly increased (Fig. 4G). Taken together, these data suggested that iOPN interacts with TRAF3 to positively regulate IFN-β production and antiviral response.
iOPN inhibits K48-linked polyubiquitination and degradation of TRAF3. TRAF3 activation is tightly regulated by protein ubiquitination. K63-linked TRAF3 polyubiquitination is responsible for the activation of downstream signaling 29 . While, K48-linked ubiquitination leads to the degradation of TRAF3 and deactivation of TRAF3-mediated downstream signaling 22,29 . To investigate the molecular mechanism of iOPN in the regulation of IFN-β production, TRAF3 polyubiquitination was investigated. TRAF3 was transfected into HEK293 cells together with WT HA-ubiquitin plasmid and iOPN expression plasmid. IP and WB showed that To further confirm OPN stabilizes TRAF3 protein through inhibition of K48-linked ubiquitination in physiological conditions, peritoneal macrophages from WT and Spp1 −/− mice were prepared and infected with SeV. IP and WB showed that K48-linked polyubiquitination of TRAF3 was greatly increased in the macrophages from Spp1 −/− mice compared to that in macrophages from WT mice after SeV infection (Fig. 5D, left). Increased K48-linked TRAF3 polyubiquitination in OPN-deficient macrophages was further confirmed with proteins immunoprecipitated with anti-TRAF3 under stringent conditions (Fig. 5D, right). Consistent with more TRAF3 ubiquitination, TRAF3 was degraded more rapidly in OPN-deficient macrophages (Fig. 5E, 5 h vs. 3 h). All together, these data indicated that OPN prevents TRAF3 from K48-linked polyubiquitination and degradation.
Another possibility is that iOPN prevents an E3 ligase from binding to TRAF3. Triad3A has been reported to be an E3 ligase involved in TRAF3 ubiquitination and degradation after virus infection 22 . To investigate whether iOPN inhibits Triad3A-mediated TRAF3 ubiquitination, TRAF3 was transfected into HEK293 cells together with Triad3A and iOPN. IP and WB showed that Triad3A promoted TRAF3 ubiquitination (Fig. 6B, lane 4 vs. lane 2). Overexpression of iOPN greatly decreased TRAF3 ubiquitination mediated by Triad3A (Fig. 6B, lane 5). In vitro ubiquitination assays with in vitro expressed proteins also confirmed that Triad3A-induced K48-linked TRAF3 polyubiqutination was greatly attenuated by iOPN (Fig. 6C). Triad3A binding to TRAF3 was also decreased by iOPN in a dose-dependent manner (Fig. 6D,E). In vitro pull-down assays confirmed the binding Triad3A to TRAF3 was gradually decreased with the increasing binding of iOPN to TRAF3 (Fig. 6F). Consistent with the inhibition of Triad3A-induced TRAF3 ubiqutination, Triad3A-induced degradation of TRAF3 was reversed by iOPN expression (Fig. 6B, input, lane 5 vs. lane 4). The Y residue and Q residue at position 441 and 443 of TRAF3 have been reported for Triad3A binding 22 . To confirm the importance of these two residues, TRAF3 mutant was constructed by mutating Y441 and Q443 to A (Fig. 6G). Mutation of YQ to AA ablated Triad3A binding to TRAF3 (Fig. 6H, lane 6 vs. lane 3). Notably, iOPN binding to TRAF3 mutant was also ablated, indicating iOPN binding to the same sites in TRAF3 as the Triad3A (Fig. 6H, lane 5 vs. 2). All together, these data demonstrated that iOPN compete with Trida3A for the binding to TRAF3, which prevents TRAF3 from K48-linked polyubiquitination and degradation promoted by Triad3A.

C-terminal fragment of iOPN binds to TRAF3.
Endogenous OPN can be cleaved by thrombin at position 168 into two fragments 32 . In order to investigate the OPN fragment involved in the binding and regulation of TRAF3 ubiquitination, two OPN truncations were constructed and expressed in vitro (Fig. 7A). In vitro pull-down assays demonstrated that full length and the C-terminal fragment of iOPN, but not the N-terminal fragment, bound to TRAF3 (Fig. 7B). Consistent with the C-terminal fragment binding to TRAF3, Trida3A-induced TRAF3 ubiqutination was inhibited by the C-terminal fragment (lane 6), but not the N-terminal fragment (Fig. 7C,  lane 5). These data indicated that the C-terminal fragment of OPN is responsible for the binding and inhibition of ubiquitination of TRAF3. To investigate the inhibition of TRAF3 ubiquitination by WT and the C-terminal fragment of iOPN has a physiological role on IFN-β production, lentiviral expression plasmids for WT, N-terminal fragment and C-terminal fragment of iOPN were constructed and used to infect WT and OPN-deficient macrophages. Infection of lentivirus containing WT iOPN plasmid increased OPN expression in WT macrophages and restored iOPN expression in OPN-deficient (Spp1 −/− ) macrophages (Fig. 7D). Consistent with positive function of iOPN on IFN-β production, lentiviral infection of WT iOPN expression plasmid into WT macrophages further increased SeV-induced expression of IFN-β , CXCL10, Mx1 and CCL5 (Figs 7E and S4). Lentiviral infection of WT iOPN expression plasmid into OPN-deficient macrophages restored SeV-induced expression of IFN-β , CXCL10, Mx1 and CCL5 to the same level as that in WT macrophages (Figs 7E and S4). Consistent with ability to inhibit TRAF3 ubiquitination by the C-terminal fragment, lentiviral infection of the C-terminal fragment of iOPN increased SeV-induced expression of IFN-β , CXCL10, Mx1 and CCL5 in WT macrophages (Figs 7E and S4). SeV-induced expression of IFN-β , CXCL10, Mx1 and CCL5 in OPN-deficient macrophages was also restored upon infection with lentivirus containing the C-terminal fragment of iOPN (Figs 7E and S4). But, infection of lentivirus containing the N-terminal fragment of iOPN could not increase or restore SeV-induced expression of IFN-β , CXCL10, Mx1 and CCL5 in WT and OPN-deficient macrophages, respectively (Figs 7E and S4). Taken together, these data demonstrated that iOPN binds to TRAF3 through the C-terminal fragment, preventing TRAF3 from K48-linked ubiquitination and degradation and leading to increased IFN-β production and innate antiviral response.

Discussion
Osteopontin (OPN) is a multifunctional protein involved in both innate immunity and adaptive immunity.   35 . In the present study, we provided evidence to demonstrate that OPN plays essential roles in regulating innate antiviral immunity. Knockdown of OPN expression attenuated virus-induced IFN-β production and enhanced VSV replication, while overexpression of OPN increased virus-induced IFN-β production and attenuated VSV replication. Furthermore, OPN-deficient mice showed less IFN-β production, increased VSV replication and more susceptible to VSV infection.
OPN has two isoforms, sOPN and iOPN, which are initiated from one single mRNA but from different start codon 24 . The function of sOPN on the immune regulation has been studied extensively. But, the function of iOPN is still largely unknown. Recently, iOPN has been demonstrated to play essential roles in TLR signaling. For example, TLR9 stimulation promotes association between iOPN and MyD88, leading to IRF7 activation and IFN-α production in pDCs 36 . iOPN is also reported to be involved in TLR2 and dectin-1 pathways 37 . iOPN was identified to work as an adaptor molecule to facilitate formation of a receptor cluster composed of TLR2, dectin-1 and mannose receptor that are involved in anti-fungal responses. At the same time, iOPN associates with signaling molecules IRAK1 and Syk downstream of the TLR2 pathway and dectin-1 pathway, respectively, resulting in the MAPK activation.
In this study, we found transfection of iOPN expression plasmid could induce IFN-β production, IRF3 activation and attenuate VSV replication. At the same time, we found transfection of the full length OPN expression plasmid could also induce IFN-β production, IRF3 activation and attenuate VSV replication. However, full length OPN expression plasmid was found to be less potent to induce IFN-β production compared to iOPN expression plasmid. Furthermore, we found that addition of OPN antibody into the culture medium could not inhibit full length OPN transfection-mediated IFN-β activation induced by SeV infection, indicating that the secrted form OPN is not involved in the positive regulation of IFN-β signaling. Thus, the phenomenon that full length OPN expression plasmid induced IFN-β production may be caused by the translation and production of iOPN from the unknown translation start site in the full length OPN expression plasmid. Alternatively, it may be contributed to the C-terminal fragment of full length OPN cleaved by thrombin because we demonstrated that C-terminal fragment of iOPN could efficiently restored the OPN function in OPN-deficient macrophages. All together, our data suggested that iOPN is an important positive regulator of innate antiviral immunity.
We found that iOPN interacted with TRAF3 after virus infection. OPN is an extremely acidic protein. When it is overexpressed, OPN may bind non-specifically to some, especially basic, intracellular proteins. We found overexpressed iOPN could not interact with the TRAF6 protein, which is very similar to TRAF3, indicating the specific binding of iOPN to TRAF3. Similar to our study, Inoue et al. recently reported iOPN interacts with TNF receptor-associated factor 2 (TRAF2) in CD40 and TLR4 signaling pathways to regulate TNF production by LPS-stimulated macrophages and to control LPS-induced endotoxemia 38 .
We further demonstrated that iOPN inhibits K48-linked TRAF3 polyubiquitination and degradation. Based on these data, we proposed that binding of iOPN to TRAF3 preventing an E3 ligase from binding to and ubiquitnating TRAF3. Traid3A has been reported to be an E3 ligase involved in TRAF3 ubiquitination and degradation after virus infection 22 . Indeed, we found Triad3A-mediated K48-linked ubiquitination and degradation of TRAF3 was greatly attenuated by iOPN expression. Recently, ubiquitin specific protease 25 (USP25) has been reported to cleave K48-linked ubiquitination to stabilize TRAF3 30,31 . We found overexpression of USP25 indeed decreased TRAF3 K48-linked ubiquitination. But, iOPN could not further decrease K48-linked ubiquitination of TRAF3 (Fig. 6A). Thus, iOPN may mainly prevent Triad3A from binding to and ubiquinating TRAF3, leading to its stabilization.
Cantor and colleagues showed that intracellular osteopontin (iOPN) plays essential roles in the differentiation of Follicular helper T cells (T FH cells) and follicular regulatory T cells (T FR cells) 39 . Mechanistically, they demonstrated that iOPN translocates into nucleus and interacts with transcription factor Bcl-6 after activation of the signaling via the receptor ICOS. Binding of iOPN prevented Bcl-6 from ubiquitination-mediated degradation. Similarly, we found iOPN bound to TRAF3 and prevented its degradation mediated through K48-linked ubiqutiation. Therefore, iOPN may bind to and stabilize key molecules in various signaling pathways to regulate the immune responses.
In conclusion, we identified a novel function for iOPN to positively regulate the production of IFN-β and the antiviral response. Our study also delineated a new regulatory mechanism in innate antiviral signaling through iOPN-mediated stabilization of TRAF3. Therefore, iOPN is a very important regulatory component in the antiviral response and may represent a new target for drug design against virus infection.
Mice. WT C57BL/6J mice were obtained from Joint Ventures Sipper BK Experimental Animal (Shanghai, China). Spp1 −/− mice (B6.Cg-Spp1tm1blh/J; cat. no. 004936) in the C57BL/6J background are provided by Prof. Zhinan Yin (Jinan University, Guangzhou, China), who originally obtained the mice from The Jackson Laboratory. OPN-deficient mice were backcrossed with WT C57BL/6J mice for 7 generations. Spp1 +/− heterozygous mice were bred to generate age-matched Spp1 +/+ and Spp1 −/− mice for experiments in Fig. 1. For other experiments, Spp1 −/− homozygous mice were mated to generate OPN-deficient mice. Age-and sex-matched WT C57BL/6J littermates were used as controls. Mice were hosted in a pathogen-free facility under standard 12-hour light-dark cycle, fed standard rodent chow, and water ad libitum. 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 Medical School of Shandong University, Jinan, Shandong Province, China.
Sequences, plasmid constructs, and transfection. Murine full-length OPN cDNA was cloned by standard RT-PCR and inserted into the pFLAG-CMV2 vector (Sigma-Aldrich) with the following primers: forward, 5′-CCGGAATTCAATGAGATTGGCAGTGATTTG-3′; reverse, 5′-CGCGGATCCTTAGTTGACCTCAG Scientific RepoRts | 6:23771 | DOI: 10.1038/srep23771 (Boston Biochem) following protocols recommended by the manufacturer. Recombinant proteins were mixing with 100 nM E1, 2 mM E2 and 200 mM Ub-K48 in a final volume of 20 ml reaction buffer (50 mM Hepes pH 8.0, 100 mM NaCl, 10 mM Mg 2+ − ATP, 0.5 mM DTT). The reaction was carried out at 37 °C for 1 h and products were analyzed by immunoblotting with anti-TRAF3 antibody. Native PAGE. The IRF3 dimerization assay was performed as described previously with modifications 43 . In brief, macrophages were harvested with 30 ml of ice-cold lysis buffer (50 mM Tris/HCl, pH 7.5, 150 mM NaCl, and 0.5% NP-40) containing protease inhibitor cocktail. After centrifugation at 13,000 g for 10 min, supernatants were quantified using a BCA assay (Thermo Fisher Scientific) and diluted with 2× native PAGE sample buffer (125 mM Tris/HCl, pH 6.8, 30% glycerol, and 0.1% Bromophenol blue), then 20 μg of total protein was applied to a pre-ran 7.5% native gel for separation. After electrophoresis, the proteins were transferred onto a nitrocellulose membrane for immunoblotting.
VSV infection of mice. For in vivo cytokine production studies, Spp1 −/− and WT mice (female, 6-8 weeks old) were intraperitoneally infected with VSV (1 × 10 8 pfu per mouse). The virus RNA in lung, spleen and liver were determined by qRT-PCR and by measurement of VSV V protein with VSV-G antibody. For the survival experiments, mice were intravenously infected with VSV (5 × 10 8 pfu per mouse) and then monitored for survival after VSV infection.
VSV plaque assay and detection of virus replication. VSV plaque assay was performed as previously described 40 . The HEK293 cells (2 × 10 5 ) were transfected with the indicated plasmids for 36 h before VSV infection (MOI of 0.1). At 1 h after infection, cells were washed with PBS three times and then medium was added. The supernatants were harvested 24 h after washing. The supernatants were diluted 1:10 6 and then used to infect confluent HEK293 cells cultured on 24-well plate. At 1 h after infection, the supernatant was removed, and 3% methylcellulose was overlaid. At 3 d after infection, overlay was removed, and cells were fixed with 4% formaldehyde for 20 min and stained with 0.2% crystal violet. Plaques were counted, averaged, and multiplied by the dilution factor to determine viral titer as LOG10 (pfu/ml). Total HEK293 cellular RNA was extracted and VSV RNA replicates were examined by qRT-PCR. Primers for VSV were as follows: 5′-ACGGCGTACTTCCAGATGG-3′ (sense) and 5′-CTCGGTTCAAGATCCAGGT-3′ (antisense).

Lentivirus preparation and infection. Lentiviral expression plasmids for iOPN-WT, iOPN-N, iOPN-C
were constructed by inserting the corresponding coding sequence into pWPXL vector (addgene). Lentivirus particles were produced through transfection of pWPXL-OPN, psPAX2 and pMD2.G plasmids (addgene) with a proportion of 20:15:7 into HEK293T cells, 3 days later the culture was harvested and enriched by PEG8000. The enriched lentivirus particle (MOI, 50) was used to infect WT and OPN-deficient macrophages for 4 days.