TIPE2 Controls Innate Immunity to RNA by Targeting the Phosphatidylinositol 3-Kinase–Rac Pathway

Toll-like receptor 3 and retinoid acid-inducible gene I (RIG-I)/melanoma differentiation-associated gene 5 (MDA5) have been identified as receptors for dsRNAs (1–3). RNA ligation of its receptors induces type I IFN and proinflammatory cytokines via a MyD88-independent pathway, which involves Toll/IL-1R domain-containing adapter inducing IFN-β or IFN-β promoter stimulator (2, 4). Toll/IL-1R domain-containing adapter inducing IFN-β relays signals to the kinases TANK-binding kinase 1 (TBK1) and IκB kinase, which in turn activate the downstream transcription factors IFN regulatory factor 3 (IRF3), NF-κB, and AP-1, leading to the production of type I IFN and proinflammatory cytokines (4–7). Recent studies indicate that the PI3K–AKT pathway and small GTPases may also be involved in TLR signaling (8–12). Small GTPases are enzymes that hydrolyze GTP. They are active when bound to GTP and inactive when bound to GDP and therefore serve as molecular on-and-off switches of signaling pathways that control a wide variety of cellular processes including growth, motility, vesicle trafficking, gene transcription, and death (13, 14). However, whether and how the small GTPases are involved in innate immunity to dsRNAs is not clear.

TIPE2, or TNF-α–induced protein 8 (TNFAIP8)-like 2 (TNFAIP8L2), is a newly described immune regulator of the TNFAIP8 family (15, 16). It is preferentially expressed in hematopoietic cells and significantly downregulated in patients with infectious or autoimmune disorders (17). The mammalian TNFAIP8 family consists of four members: TNFAIP8 (TIPE), TIPE1, TIPE2, and TIPE3, the functions of which are largely unknown. We report in this study that TIPE2 regulates Poly (I:C)-induced innate immune responses by targeting Rac GTPases in a PI3K-dependent manner.

Wild-type (WT) C57BL/6 (B6) mice were purchased from The Jackson Laboratory. The Tipe2^−/− B6 mice were generated by backcrossing Tipe2^−/− 129 mice to B6 mice for 12 generations. Age- and sex-matched WT and Tipe2-deficient mice were used in all experiments. Mice were housed in the University of Pennsylvania Animal Care Facilities under specific pathogen-free conditions. All animal procedures were preapproved by the Institutional Animal Care and Use Committee of the University of Pennsylvania.

Poly (I:C), CpG, imiquimod, and rhodamine-labeled Poly (I: C) were purchased from InvivoGen. LPS and peptidoglycan (PGN) were from Sigma-Aldrich. PI3K inhibitor LY294002 was obtained from Cell Signaling Technology. The p38 inhibitor SB203580, ERK kinase I inhibitor PD98059, JNK inhibitor SP600123, and NF-κB inhibitor Bay were purchased from Promega. Rac inhibitor NSC23766 was obtained from Tocris. GM-CSF was from R&D Systems. ELISA reagents were purchased from BD Biosciences, which include purified and biotinylated rat anti-mouse IL-6 and TNF-α. Abs used were as follows: rabbit anti-Rac1/2/3, rabbit anti-Na/K ATPase, rabbit anti–phospho-P21–activated kinase (PAK) 1 (Ser¹⁴⁴)/PAK2 (Ser¹⁴¹), rabbit anti-PAK1/2/3, rabbit anti–p-AKT, rabbit anti-total AKT, rabbit anti-IRF3 (Cell Signaling Technology), rabbit anti-Tiam (Santa Cruz Biotechnology), and HRP-conjugated anti-mouse or anti-rabbit Ig (GE Healthcare). Quantitative real-time PCR primers for IL-6, TNF-α, IL-1β, IFN-β1, and IFN-α4 were purchased from Qiagen.

Bone marrow-derived dendritic cells (BMDCs) were generated as described (18). Briefly, 2 × 10⁶ bone marrow precursors from B6 and Tipe2-deficent mice were seeded in complete IMDM supplemented with 3.3 ng/ml GM-CSF in six-well plates. Two milliliters medium was added to the culture on day 3, and half of the medium was replaced with new medium on days 5, 6, and 7. To generate bone marrow-derived macrophages (BMMs), bone marrow cells were cultured for 7 d in DMEM supplemented with 10% FCS, 1% penicillin/streptomycin, 1% glutamine, and 30% L-929 cell culture supernatant. At the end of culturing, the cells were washed twice with cold PBS and rested overnight in complete DMEM before assays. BMDCs were 80% CD11c⁺, and BMMs were >95% CD11b⁺ and F4/80⁺ as determined by flow cytometry.

TLR2 ligand PGN (10 μg/ml), TLR3 ligand Poly (I:C) (20 μg/ml), TLR9 ligand CpG (5 μM), TLR4 ligand LPS (100 ng/ml), and TLR7 ligand imiquimod (5 μg/ml) were added to BMDC or BMM cultures for various times. For inhibitor assays, all of the inhibitors were added to the culture 30 min before TLR ligand stimulation.

The total RNA was isolated using TRIzol reagent (Invitrogen) and purified using RNeasy Mini kits (Qiagen) according to the manufacturers’ description. After treatment with RNase-free DNase I (Invitrogen), RNA samples were reversely transcribed with oligo(dT) and SuperScript II transcriptase (Invitrogen). Real-time quantitative PCR analysis was performed using specific Quantitect Primers for mouse GAPDH, TIPE2, IL-6, IFN-β1, IFN-α4, IL-1β, and TNF-α (Qiagen) in an Applied Biosystems 7500 system using Power SYBR Green PCR Master Mix (Applied Biosystems). Relative levels of gene expression were determined using GAPDH as the control.

The cell-culture supernatants were collected and stored at −80°C. Quantitative ELISA was performed using paired mAbs specific for corresponding cytokines according to the manufacturer’s protocols (BD Biosciences).

To assess Rac activation, the cell extracts were incubated with PAK–glutathione-S-transferase fusion protein beads (Cytoskeleton) at 4°C for 60 min. The collected beads were then washed three times and resuspended in SDS protein sample buffer. The bound proteins and total cell lysates were analyzed by SDS-PAGE and blotted with anti-Rac (Cell Signaling Technology).

Whole-cell lysate was prepared by lysing cells in a buffer containing 150 mM NaCl, 50 mM HEPES (pH 7), 1 mM EDTA, 1% Nonidet P-40 (NP-40), 1× complete protease inhibitors mixture (Roche), and 1× phosphatase inhibitor mixture (Roche). In certain experiments, cell membrane proteins and cytoplasmic proteins were prepared using a Subcellular Protein Fractionation Kit (Pierce) according to the manufacturer’s protocols. Protein concentration was determined by BCA assay (Pierce). Equal quantities of proteins were separated by SDS-PAGE, transferred to a nitrocellulose membrane, and blotted with specific Abs.

Whole-cell extracts were prepared in NP-40 lysis buffer (50 mM Tris-HCl [pH 8], 1% NP-40, 150 mM NaCl, 1× protease inhibitor mixture [Roche], and 1× phosphatase inhibitor mixture [Roche]), then were subjected to electrophoresis on 8% native acrylamide gels, which were prerun at 40 mA for 30 min at 4°C. The electrophoresis buffer was composed of a upper chamber buffer (25 mM Tris-HCl [pH 8.4], 192 mM glycine, and 0.2% sodium doxycholate) and a lower chamber buffer (25 mM Tris-HCl [pH 8.4] and 192 mM glycine). Protein for each sample was mixed with 2× loading buffer (125 mM Tris-HCl [pH 6.8], 30% glycerol, and bromophenol blue), subjected to electrophoresis by native PAGE at 25 mA for 90 min at 4°C, then transferred to a nitrocellulose membrane, and blotted with IRF3 Abs.

The murine RAW264.7 macrophages (American Type Culture Collection) and HEK293T cells (American Type Culture Collection) were cultured in DMEM containing 10% heat-inactivated FBS, 2 mM l-glutamine, and 100 U/ml penicillin/streptomycin. To prepare retroviruses, packaging cells (293T) were cultured in 10-mm culture dishes and transfected with nerve growth factor receptor murine stem cell virus-based retroviral vector containing WT Rac, Rac1 T17N, and Rac1 Q61L along with pVSVG and pCGP sequences using the CalPhos mammalian transfection kit (Clontech). The transfection medium was replaced with fresh medium 6 h posttransfection. At 24 and 48 h posttransfection, the culture medium containing recombinant retroviruses was harvested, filtered (with a 0.20-μm filter) and used to infect Raw cells. The infection efficiency was ∼70% as determined by flow cytometry.

Six- to 8-wk-old WT and Tipe2^−/− mice were i.p. injected with Poly (I:C) (40 mg/kg body weight). Mice were observed for sickness for 5 d. The serum was collected at 24 h after injection.

The differences in mRNA and proteins were analyzed by two-tailed Student t test. The differences in survival rate were analyzed by Mann–Whitney U test.

Myeloid cells play important roles in innate immunity to pathogens. TIPE2 is highly expressed in resting myeloid cells including dendritic cells and macrophages (15). To explore the roles of TIPE2 in dendritic cell-mediated innate immunity, TIPE2 expression was examined in murine BMDCs before and after stimulation with different TLR ligands/agonists. Upon stimulation with LPS (the TLR4 ligand), Poly (I:C) (the TLR3/MDA5 agonist), CpG (the TLR9 agonist), and PGN (the TLR2 ligand), TIPE2 mRNA expression was significantly reduced (Fig. 1A); by contrast, the mRNA (Fig. 1B, 1C) and/or protein levels (Fig. 1D) of cytokine genes (IFN-β1, IFN-α4, IL-6, and TNF-α) were significantly increased in BMDCs. Similar effects were observed in the murine BMMs (Fig. 1E) (16). This inverse correlation between TIPE2 and cytokine levels in the innate immune cells treated with TLR ligands/agonist suggests a role for TIPE2 in regulating innate immune responses.

Type I IFNs and inflammatory cytokines such as IL-6 produced by activated dendritic cells play crucial roles in antimicrobial immunity. To determine the potential roles of TIPE2 in Poly (I:C)-induced cytokine production, we compared cytokine gene expression in WT and Tipe2^−/− BMDCs. We observed significantly increased IFN-β1 and IL-6 expression in Tipe2^−/− dendritic cells (Fig. 2A). Transcription factor IRF3 mediates IFN-β1 production in innate immune cells. We observed constitutively active IRF3 in Tipe2^−/− BMDCs upon Poly (I:C) stimulation (Fig. 2B).

To establish the in vivo relevance of these findings, we injected Poly (I:C) intraperitoneally into WT and Tipe2^−/− mice. Remarkably, all Tipe2^−/− mice were very sick 24 h after Poly (I:C) injection and died after blood withdraw, whereas none of WT littermates did (Fig. 2C). The serum IL-6 and TNF-α levels were dramatically elevated in Tipe2^−/− mice (Fig. 2D), indicating the important role of TIPE2 in Poly (I:C)-mediated inflammatory cytokine expression.

To explore the mechanism of TIPE2 action in Poly (I:C)-induced cytokine expression, we first compared ligand uptake between WT and Tipe2^−/− BMDCs. The rhodamine-labeled Poly (I:C) was comparably taken up by BMDCs from the WT and Tipe2^−/− mice (Fig. 2E), indicating that increased cytokine expression in Tipe2^−/− cells was not due to enhanced uptaking of the TLR agonists. We recently found that endogenous TIPE2 can constitutively bind to the small GTPase Rac in immune cells (Z. Wang and Y. Chen, unpublished observations). We compared Rac activation between WT and Tipe2^−/− BMDCs. An elevated constitutive Rac activation was observed in Tipe2^−/− BMDCs (Fig. 2F), indicating that Rac was activated and involved in Poly (I:C)-mediated cytokine expression. PAK and AKT are downstream effectors of Rac and PI3K, respectively. As shown in Fig. 2G, PAK and AKT were also constitutively active in Tipe2^−/− BMDCs. Rac1 and PAK1 have been reported to act upstream of TBK1/IκB kinase-ε in the viral activation of IRF3 (19). Increased constitutive activation of TBK1 was also observed in Tipe2^−/− cells as compared with the WT cells (Fig. 2G), indicating that TIPE2 regulates TBK1 via the Rac/PAK pathway.

To further dissect the pathways involved in Poly (I:C)-mediated cytokine production, we used inhibitors of several signaling pathways. As shown in Fig. 3A, PI3K inhibitor LY294002 and p38 inhibitor SB203580 significantly diminished the expression of IFN-β1 and IL-6, whereas PD98059 (an ERK inhibitor), SP600123 (a JNK inhibitor), and Bay (an NF-κB inhibitor) did not. Next, we examined whether LY294002 could also affect Rac activation. As shown in Fig. 3B, LY294002 significantly blocked Rac activation. To directly confirm the role of Rac in IFN-β1 and IL-6 production, the Rac inhibitor NSC 23766 was added to cells 20 min before Poly (I:C) stimulation. Similar to LY294002 (Fig. 3C), NSC23766 effectively inhibited IFN-β1 and IL-6 production in a dose-dependent manner (Fig. 3D). Similar results were obtained for BMMs. NSC23766 significantly inhibited IFN-β1, IL-6, IFN-α4, and TNF-α mRNA expression but not IL-1β expression (Fig. 4). The inhibitors tested in Figs. 3 and 4 did not significantly affect the apoptotic levels of the cells in the cultures as determined by flow cytometry (data not shown).

To directly examine whether Rac activation is involved in IFN-β1 and IL-6 production, we retrovirally transduced Raw macrophages with a dominant-negative Rac mutant (Rac17N) or a constitutively active Rac (Rac61L). The expression of Rac17N significantly suppressed Poly (I:C)-induced IFN-β1 and IL-6 production, whereas that of Rac61L had the opposite effect (Fig. 5). These results indicate that Rac activation is directly involved in Poly (I:C)-mediated cytokine expression.

Small GTPases are activated by guanine nucleotide exchange factors (GEFs). NSC23766 effectively blocks the binding of Rac1 to the Rac-specific GEF Tiam, thus inhibiting Rac1 activity. To determine whether Poly (I:C) also affects membrane Tiam levels via the PI3K pathway, we checked membrane Tiam and cytosolic PAK levels. As shown in Fig. 6A, Tiam membrane recruitment was enhanced by Poly (I:C) stimulation, but the enhancement was not affected by PI3K inhibitor LY294002. Similarly, the phosphorylation at serine 396 was not affected by inhibition of PI3K (Fig. 6C). However, the activation of Rac downstream effector PAK was significantly affected. Thus, the activation of PAK was markedly elevated upon Poly (I:C) stimulation, which was significantly blocked by LY294002 (Fig. 6B). Taken together, we have found that Poly (I:C) stimulated the recruitment of Rac GEF Tiam to the membrane but in a PI3K-independent manner, whereas Rac effector PAK was activated by Poly (I:C) in a PI3K-dependent manner.

Dendritic cells play important roles in innate antimicrobial immune responses and the initiation of adaptive immune responses. The activated dendritic cells trigger large quantities of type I IFN and inflammatory cytokine production (20). Poly (I:C) activates various immune cells through two major dsRNA sensors: TLR3 and RIG-I/MDA5. RIG-I/MDA5 are cytosolic dsRNA sensors detecting cytosolic Poly (I:C) via adaptor protein IFN-β promoter stimulator, whereas TLR3 is located in intracellular endosome recognizing dsRNA and triggering type I IFN and inflammatory cytokine production via a MyD88-independent pathway. However, signaling through both receptors converges on the induction of type I IFN. Results reported in this study indicate that TIPE2 serves as an important regulator of type I IFN and proinflammatory cytokine production in Poly (I:C)-stimulated dendritic cells by targeting Rac activation. TIPE2 is constitutively expressed at high levels in immune cells. Exposure to TLRs markedly downregulates TIPE2 levels, which removes its inhibitory effect on Rac GTPases; the activated Rac GTPases in turn activate their downstream effector targets in a PI3K-dependent manner. Thus, TIPE2 plays an important role in cytokine expression mediated by Poly (I:C) in innate immune cells.

Rac proteins constitute a subgroup of the Rho family of small GTPases, which includes Rac1, Rac2, Rac3, and Rac1b, the splice variant of Rac1 (13, 21). They function as molecular switches that control signaling pathways regulating cytoskeleton organization, gene expression, cell cycle progression, cell motility (22), and innate immune responses related to pathogen sensing, intracellular uptake, and destruction (23, 24). They can be activated through interaction with Dbl family GEFs (e.g., Tiam, Trio, and Dock180) and inactivated by GTPase-activating proteins (GAPs) (e.g., RacGAP). NSC23766 specifically blocks the binding of Tiam to Rac1. Tiam contains two pleckstrin homology (PH) domains flanking its Dbl homology domain. Dbl homology domain is responsible for GEF activity, whereas in general, the N-terminal PH domains have been implicated in directing the subcellular localization by binding to phosphatidylinositides and/or membrane-associated protein targets (25). In our study, NSC23766, a small molecule blocking the binding of Tiam to Rac1 (26), leads to downregulation of cytokine expression induced by Poly (I:C), indicating that the binding of Tiam and Rac1 plays crucial roles in the Poly (I:C)-mediated cytokine expression in dendritic cells. Although PI3K inhibitor LY294002 could inhibit Rac activation and block the cytokine production induced by Poly (I:C) as shown in our experiments, it does not block the Tiam membrane translocation, indicating an alternative mechanism for Tiam membrane translocation independent of PI3K. Rac activation could occur at several steps: removal of inhibitory molecules such as the GDP dissociation inhibitor, recruitment or activation of GEFs, and removal or inhibition of GAP; the current available evidence suggests that it depends largely on the activation of the GEF (27). However, the mechanism of Tiam activation remains unclear. Although the N-terminal PH domain of Tiam binds to PIP3, its membrane translocation, as reported by several laboratories, does not require PI3K activity (28, 29). The role of PI3K in Tiam activation is still controversial (30).

The activation of the transcription factor IRF3 is crucial for the production of type I IFNs. The dormant form of IRF3 is present as a monomer in cytosol. Viral infection and treatment with dsRNA induce phosphorylation of IRF3, leading to its dimerization and eventual activation (31). As reported in this study, the transcription factor IRF3 was constitutively activated in Tipe2^−/− cells, but the phosphorylation of IRF3 at serine 396 was not affected by Poly (I:C), which is consistent with a previous report (9). Although phosphorylation at serine 396 is not affected by PI3K, PI3K is necessary for the phosphorylation of additional residues of IRF3 and its full activation (9). Poly (I:C)-induced Rac activation and downstream signaling molecule PAK activation is PI3K dependent. PI3K inhibitor LY294002 prevented the majority of cytokine expression, whereas the dominant-negative Rac T17N only partially blocked the cytokine expression in Poly (I:C)-stimulated DCs, indicating that PI3K is most likely functioning at the receptor-proximal end, whereas Rac signaling is one of the PI3K downstream events.

Thus, upon ligation of TLR3 by Poly (I:C), TIPE2 level decreases, removing the inhibitory effects on the Rac–TBK1 pathway. The activated TBK1 in turn causes the activation of IRF3 and subsequent expression of cytokine genes. These findings raise additional questions. How does TIPE2 regulate PI3K–Rac pathway? Do they interact with each other directly? What are the roles, if any, of other TIPE2-interacting proteins in innate immunity? Answering these questions will broaden our knowledge of innate immune regulation and help develop strategies to manipulate innate immune responses in humans.

The authors have no financial conflicts of interest.