TLR4 recruits TRIF-related adaptor molecule (TRAM, also known as TICAM2) as a sorting adaptor to facilitate the interaction between TLR4 and TRIF and then initiate TRIF-dependent IRF3 activation. However, the mechanisms by which TRAM links downstream molecules are not fully elucidated. In this study, we show that TRAM undergoes tyrosine phosphorylation upon TLR4 activation and that is required for TLR4-induced IRF3 activation. Protein tyrosine phosphatase nonreceptor type 4 (PTPN4), a protein tyrosine phosphatase, inhibits tyrosine phosphorylation and subsequent cytoplasm translocation of TRAM, resulting in the disturbance of TRAM–TRIF interaction. Consequently, PTPN4 specifically inhibits TRIF-dependent IRF3 activation and IFN-β production in TLR4 pathway. Therefore, our results provide new insight into the TLR4 pathway and identify PTPN4 as a specific inhibitor of TRIF-dependent TLR4 pathway. Targeting PTPN4 would be beneficial for the development of new strategy to control TLR4-associated diseases without unwanted side effects.

Pattern recognition receptors (PRRs) mediate the recognition of pathogen-associated molecular patterns and trigger innate immune responses against pathogens invasion (1, 2). TLRs are the best characterized PRRs, which possess leucine-rich repeats, transmembrane domains, and intracellular Toll-IL-1R (TIR) domains (2, 3). Unique among the large family of TLRs, TLR4 engages two distinct adaptor proteins: MyD88 and TRIF, which activates two separate pathways, the MyD88-dependent and TRIF-dependent pathway (also called MyD88-independent pathways) (3). In the case of TLR4 signaling, TLR4 uses two adaptors: TIR domain–containing adapter protein (TIRAP, also known as Mal) and TRIF-related adaptor molecule (TRAM, also known as TICAM2) as sorting adaptors to facilitate signal transduction (3). TIRAP recruits MyD88 to facilitate the activation of the MyD88-dependent pathway at the plasma membrane, resulting in the activation of NF-κB and MAPK pathways, and subsequent secretion of proinflammatory cytokines, such as TNF-α and IL-6. As for the TRIF-dependent pathway, TRAM is needed for the recruitment of TRIF (46). Following LPS stimulation, TRAM dissociate from the membrane to endosomes, where the TRAM–TRIF pathway is initiated (7, 8). This leads to the activation of transcription factor IRF3 and then induces the expression of type I IFN (IFN-α/β) (3). Besides TLR4, TLR3 also uses TRIF to activate downstream pathways, while it recruits TRIF directly, without the need for TRAM.

Reversible phosphorylation of proteins, which is catalyzed by kinases and phosphatases, is a key regulatory mechanism for numerous important aspects of physiology. Phosphorylation status of receptors or adaptors involved in PRRs signaling directly affects the transduction of activation signals and is crucial for optimal innate immune responses. For example, a dynamic balance between phosphorylation and dephosphorylation of RIG-I and MDA5 is essential for their signal transduction (9). Serine phosphorylation of NLRC4 (a cytosolic member of the NOD-like receptor family) is critical for inflammasome activation (10). In the setting of TLR pathways, TLR2, TLR3, TLR4, TLR5, TLR8, and TLR9 undergo tyrosine phosphorylation (11), and mutations of tyrosines within their TIR domains suppress TLRs activation. In addition to the receptors, other components of the PRRs signaling pathways (e.g., the adaptors, also undergo tyrosine phosphorylation upon ligand engagement) (11). Tyrosine phosphorylation of TIRAP is required for the recruitment of MyD88 to TLR2 and TLR4 (12, 13). TRIF and MyD88 have been shown to be tyrosine phosphorylated causing downregulation of TLR signaling (14). However, the phosphorylation modification of key proteins in PRRs signaling has not been fully elucidated. Protein tyrosine phosphatases (PTPs), which dephosphorylate tyrosine residues of target substrates (15, 16), have been reported to be involved in the regulation of innate immune responses, such as Src homology region 2 domain–containing phosphatase-1 (17), Src homology region 2 domain–containing phosphatase-2 (18), PTPN22 (19), PTP with proline-glutamine-serine-threonine-rich motifs (20), PTP1B (21), and so on. However, the PTPs identified to be PRRs regulators possessed broadly regulatory effects and that limited their potential implication for immunotherapy. Thus, more specific regulator remains to be identified.

PTP nonreceptor type 4 (PTPN4, also known as PTP-MEG1) functions in TCR cell signaling and apoptosis (22, 23). However, to our knowledge, the roles of PTPN4 in innate immune responses have never been investigated. In this study, we show that PTPN4 specifically inhibits TRIF-dependent TLR4 pathway by suppressing tyrosine phosphorylation of TRAM. TRAM undergoes tyrosine phosphorylation at Tyr167 upon TLR4 activation and that is required for TLR4-induced IRF3 activation and IFN-β secretion. PTPN4 attenuates TRAM tyrosine phosphorylation and inhibits its translocation and subsequent TRAM–TRIF interaction. These results indicate that tyrosine phosphorylation of TRAM is critical for TLR4 activation and also identify PTPN4 as a specific inhibitor of TRIF-dependent pathway triggered by TLR4 engagement.

Female C57BL/6 mice (5–6 wk) were obtained from Joint Ventures Sipper BK Experimental Animal (Shanghai, China). All animal experiments were undertaken in accordance with the National Institutes of Health’s 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, China). LPS (Escherichia coli, 055:B5) and lipothechoic acid (LTA) were from Sigma-Aldrich (St. Louis, MO); peptidoglycan (PGN), polyinosinic:polycytidylic acid (poly[I:C]), and imidazoquinoline compound R848 were purchased from InvivoGen (San Diego, CA). The concentration of agonists were used as below: 100 ng/ml LPS, 10 μg/ml poly(I:C), 10 μg/ml PGN, 1 μg/ml R848, and 5 μg/ml LTA. The Abs specific to p-tyr (9416), anti-Myc (2272), IRF3 (4302), STAT1 (9175), anti-(p)-IRF3(Ser396) (4947), anti–p-STAT1(Tyr701) (7649), anti–p-JNK (4668), anti–p-p38 (4511) anti–p-ERK (4370), anti-JNK (9258), anti-p38 (8690), anti-ERK1/2 (4695), and anti–p-IκBα (2859) were from Cell Signaling Technology (Beverly, MA). Anti-Flag (F3165) was from Sigma-Aldrich. Anti-PTPN4 (PAB4081) was from Abnova. Anti-TRAM (sc-34748), anti–β-actin (sc-81178), and protein G–agarose (sc-2002) used for immunoprecipitation (IP) and HRP-conjugated secondary Abs were from Santa Cruz Biotechnology (Santa Cruz, CA). Sendai virus (SeV) was purchased from China Center for Type Culture Collection (Wuhan University, Wuhan, China).

To obtain mouse primary peritoneal macrophages, C57BL/6J mice were injected i.p. with 3% Brewer’s thioglycollate broth. Three days later, peritoneal exudate cells were harvested and incubated. Two hours later, nonadherent cells were removed, and the adherent monolayer cells were used as peritoneal macrophages. Mouse macrophage cell line RAW264.7 and human embryonic kidney (HEK)293 cells were obtained from American Type Culture Collection (Manassas, VA). HEK293–TLR3 and HEK293–TLR4 cell lines were obtained from InvivoGen. The cells were cultured at 37°C under 5% CO2 in DMEM supplemented with 10% FCS (Invitrogen Life Technologies), 100 U/ml penicillin, and 100 μg/ml streptomycin (24, 25).

Flag-tagged PTPN4 wild-type (WT) and DA plasmids were provided by Dr. N. S. van Oers (University of Texas Southwestern Medical Center, Dallas, TX) and subcloned in pCMV-Myc plasmids (Promega). PTPN4 DA plasmid is a phosphatase activity–disrupted mutant in which an Asp to Ala point mutation was introduced in the PTPase domain (22). PTP1B plasmid was provided by Dr. N. K. Tonks (Cold Spring Harbor Laboratory, Cold Spring, NY) (26). Flag-TRAM expression plasmid was provided by Dr. L. A. O’Neill (Trinity College Dublin, Dublin, Ireland). The TRAM Y154F and Y167F mutations were generated using the KOD-Plus-Mutagenesis kit (Toyobo, Osaka, Japan). NF-κB reporter plasmid was purchased from Stratagene. IFN-β and IRF3 reporter plasmids were gifts from Dr. X. Cao (Second Military Medical University, Shanghai, China). Expression plasmids for RIG-I, MAVS, TRIF, and TBK1 were described previously (27).

For transient transfection of plasmids into RAW264.7 cells, jetPEI reagents were used (Polyplus-transfection). For stable selection of RAW264.7 cell lines overexpressing PTPN4, transfected RAW264.7 macrophages were selected with G418 (800 μg/ml) and pooled for further experiments (24, 25). For transient silencing, duplexes of small interfering RNA (siRNA) were transfected into cells with the Geneporter 2 Transfection Reagent (GTS, San Diego, CA), according to the standard protocol (23, 24). Target sequences for transient silencing were 5′-GGAUUUGUUAAACCAUUAA-3′ (siRNA 1) and 5′-GGUUUGGAUUCAAUGUAAA-3′ (siRNA 2) for PTPN4, and “scrambled” control sequences were 5′-UUCUCCGAACGUGUCACGU-3′. siRNA sequences for TRIM38 were described previously (25).

The concentration of IFN-β was measured with ELISA kits (BioLegend, San Diego, CA). The concentrations of TNF-α and IL-6 were measured with ELISA kits (Dakewe Biotech, Shenzhen, China).

Total RNA was extracted with TRIzol reagent, according to the manufacturer’s instructions (Invitrogen). A LightCycler (ABI PRISM 7000) and a SYBR RT-PCR kit (Takara) were used for quantitative real-time RT-PCR analysis. Specific primers used for RT-PCR assays were 5′-ATGAGTGGTGGTTGCAGGC-3′ and 5′-TGACCTTTCAAATGCAGTAGATTCA-3′ for IFN-β and 5′-TGTTACCAACTGGGACGACA-3′ and 5′-CTGGGTCATCTTTTCACGGT-3′ for β-actin. Data are normalized to β-actin expression in each sample.

For IP, whole-cell extracts were lysed in IP buffer containing 1% (v/v) Nonidet P-40, 50 mM Tris-HCl (pH 7.4), 50 mM EDTA, 150 mM NaCl, and a protease inhibitor “mixture” (Merck). After centrifugation for 10 min at 14,000 × g, supernatants were collected and incubated with protein G Plus–Agarose IP reagent together with specific Ab. After 6 h of incubation, beads were washed five times with IP buffer. Immunoprecipitates were eluted by boiling with 1% (w/v) SDS sample buffer. For Western blot, cells were lysed with M-PER Protein Extraction Reagent (Pierce, Rockford, IL) supplemented with a protease inhibitor “mixture,” and then protein concentrations in the extracts were measured with a bicinchoninic acid assay (Pierce). Equal amounts of extracts were separated by SDS-PAGE and then were transferred onto nitrocellulose membranes for immunoblot analysis (24, 25).

Membrane and cytoplasmic protein extraction was performed using a Membrane and Cytosol Protein Extraction kit (Beyotime, Shanghai, China). Briefly, 4 × 106 cells were resuspended in 200 μl membrane protein isolation solution A and homogenized on ice. The homogenate was centrifuged at 700 × g for 10 min at 4°C. The supernatant was then centrifuged at 14,000 × g for 30 min at 4°C. The resulting supernatant was the cytoplasmic protein fraction. The pellet was resuspended in 40 μl membrane protein isolation solution B. The homogenate was incubated on ice for 10 min, vortexed for 5 s, and then centrifuged at 14,000 × g for 5 min. The resulting supernatant was the membrane protein fraction.

Luciferase activities were measured with Dual-Luciferase Reporter Assay System (Promega), according to the manufacturer’s instructions (24, 25). Data are normalized for transfection efficiency by subtracting Firefly luciferase activity with that of Renilla luciferase.

ChinaPeptides (Shanghai, China) immunized a rabbit with a synthetic peptide corresponding to aa 163–174 of TRAM (RQHKYNSVIPMR) with a phosphotyrosine incorporated instead of the tyrosine. The phosphospecific Abs were then affinity purified from the blood by using the phosphopeptide.

The Protein Data Bank file (PDB code: 2M1W) of TIR domain of human TRAM is downloaded from Protein Data Bank (http://www.rcsb.org). Structural figures were generated using PyMol (http://www.pymol.org).

All experiments were independently performed three times. Data are presented as means ± SD of three or four experiments. Analysis was performed using a Student t test or ANOVA. The p values < 0.05 were considered to be statistically significant.

To identify possible roles for PTPN4 in innate immune responses, we transfected HEK293-TLR4, HEK293-TLR3, or HEK293 cells with an IFN-β or NF-κB luciferase reporter, as well as PTPN4 expression plasmid, and then treated the cells with LPS (TLR4 ligand) or poly(I:C) (TLR3 ligand) or infected with SeV (a type of ssRNA virus recognized by RIG-I). PTPN4 specifically inhibited LPS-induced IFN-β activation and had no effects on IFN-β activation induced by poly(I:C) or SeV (Fig. 1A). In addition, PTPN4 had no effects on NF-κB activation induced by TLR4, TLR3, or RIG-I (Fig. 1B). As a parallel control, TLR4/3- and RIG-I–induced NF-κB activation was greatly inhibited by PTP1B, which was consistent with a previous report (21).

To further confirm that PTPN4 specifically attenuates TLR4-induced IFN-β activation, knockdown experiments were performed. The expression of PTPN4 was greatly decreased with transfection of PTPN4-specific siRNA in mouse peritoneal macrophages (Fig. 1C). PTPN4 siRNA 1, which has a higher efficiency to knockdown PTPN4 protein expression (Fig. 1C), has a greater potential to increase the LPS-induced IFN-β production (Fig. 1D). Therefore, PTPN4 siRNA 1 was used in the following experiments. Although PTPN4 knockdown significantly increased LPS-induced IFN-β production, it had no effects on poly(I:C), PGN (a TLR2 ligand), R848 (TLR7/8 ligand), or SeV-induced IFN-β production (Fig. 1E). In addition, PTPN4 knockdown could not promote TNF-α or IL-6 production induced by LPS, poly(I:C), PGN, R848, or SeV (Fig. 1F). As a parallel control, TNF-α or IL-6 production induced by LPS, poly(I:C), PGN, R848, or SeV was enhanced by TRIM38 knockdown, which was consistent with previous report (25). The effect of PTPN4 knockdown was also analyzed by quantitative RT-PCR. PTPN4 siRNA treatment significantly increased LPS-induced IFN-β mRNA expression (Fig. 1G) and had no effects on LPS-induced TNF-α or IL-6 mRNA expression (Fig. 1G), demonstrating that PTPN4 inhibits LPS-induced IFN-β expression at both mRNA and protein levels in macrophages. Collectively, these data indicate that PTPN4 specifically inhibits TLR4-induced IFN-β production.

IRF3 is the key transcription factor that mediates the expression of IFN-β in TLR3-, TLR4-, and RIG-I–mediated signal transduction. We then observed the effect of PTPN4 on IRF3 activation. IRF3 phosphorylation was greatly increased by PTPN4 knockdown in LPS-stimulated macrophages (Fig. 2A) but not in poly(I:C)- or SeV-treated macrophages (Fig. 2B, 2C). LPS induces phosphorylation of STAT1 at Tyr701, which is IFN-β dependent (28). To investigate whether the effect of PTPN4 is consistent for IFN-mediated signaling, we examined the effects of PTPN4 on STAT1 phosphorylation. PTPN4 knockdown substantially increased LPS-induced phosphorylation of STAT1 at Tyr701 (Fig. 2A). We next observed the effect of PTPN4 on IFR3 activation by detecting IRF3 luciferase reporter gene expression. PTPN4 attenuated TLR4-induced IRF3 activation and had no effects on TLR3- or RIG-I–induced IRF3 activation (Fig. 2D). Accordingly, PTPN4 greatly inhibited LPS-induced IRF3 phosphorylation in HEK293–TLR4 cells (Fig. 2E). However, PTPN4 knockdown had no effects on LPS-induced phosphorylation of ERK, JNK, p38, and IκBα (Fig. 2F), indicating that PTPN4 bears no regulatory effects on MAPK and NF-κB activation induced by TLR4 engagement. Taken together, these data suggested that PTPN4 specifically attenuated TLR4-induced IRF3 activation.

PTPN4 is an intracellular protein tyrosine phosphatase, and its phosphatase activity is important for its role in dephosphorylation of downstream substrates. We then investigated the role of phosphatase activity in its inhibitory effects. PTPN4 WT greatly inhibited TLR4-induced IFN-β and IRF3 activation, whereas phosphatase inactive mutant (PTPN4 DA) lost the inhibitory effects on TLR4-induced IFN-β and IRF3 activation in HEK293–TLR4 cells (Fig. 3A). Next, we established RAW264.7 cell lines that stably expressed Myc-tagged PTPN4 WT plasmid, PTPN4 DA plasmid, or empty vector. Overexpression of PTPN4 was confirmed by Western blotting with both PTPN4 and Myc Abs (Fig. 3B). PTPN4 WT greatly inhibited IFN-β production in LPS-induced RAW264.7 cells, whereas PTPN4 DA lost the inhibitory effects (Fig. 3C). The phosphatase inactive mutant PTPN4 DA bears no effects on TLR4-induced IFN-β activation, suggesting that PTPN4 inhibited TRIF-dependent TLR4 signaling via its phosphatase activity.

To determine the molecular targets of PTPN4 in TLR4 pathway, the effects of PTPN4 on IFN-β or IRF3 promoter activation mediated by various molecules were examined in reporter assays. PTPN4 significantly inhibited TRAM-induced IFN-β or IRF3 activation (Fig. 4A). However, TRIF-, RIG-I–, MAVS-, and TBK1-induced IFN-β or IRF3 activation remain unchanged by PTPN4 overexpression (Fig. 4A). These data indicate that PTPN4 may target TRAM or its upstream receptors. Our previous data indicated that PTPN4 specifically inhibited TRIF-dependent IRF3 activation (Fig. 2A, 2E) and had no regulatory effects on MAPK or NF-κB activation in TLR4 pathway (Figs. 1B, 2F). Furthermore, TLR3 and RIG-I pathways, whose signals do not use TRAM, were not affected by PTPN4 (Figs. 1A, 1D, 2B–D). Therefore, we speculated that PTPN4 targeted TRAM to inhibit IFN-β production. Consistent with our speculation, PTPN4 inhibited TRAM-induced IFN-β activation in a dose-dependent manner, and the phosphatase mutant PTPN4 DA lost the inhibitory effects (Fig. 4B). TRAM–TRIF pathway could trigger late NF-κB activation in TLR4 pathway. We then investigated whether PTPN4 could affect the TRAM-dependent NF-κB activation. PTPN4 inhibited TRAM-induced NF-κB activation, whereas PTPN4 DA lost the inhibitory effects (Fig. 4B). Taken together, these data indicated that PTPN4 inhibited TRIF-dependent TLR4 signaling by targeting TRAM.

TRAM predominately locates in the plasma membrane and translocates to cytoplasm upon TLR4 activation. To investigate whether PTPN4 possess the potential to target TRAM, we first detected the subcellular localization of PTPN4. PTPN4 was presented predominantly at the plasma membrane and translocated to cytoplasm following TLR4 activation in macrophages (Fig. 4C). Similarly, TRAM was presented at the plasma membrane, and the amount of TRAM in the membrane fraction was decreased upon LPS stimulation (Fig. 4C), suggesting that TRAM is disappearing from the membrane, which was consistent with previous report (7). We then investigated whether endogenous PTPN4 interacted with TRAM. An association between PTPN4 and TRAM could be detected in resting macrophages and was enhanced by LPS stimulation (Fig. 4D). To further confirm the interaction between PTPN4 and TRAM, Myc-tagged PTPN4 together with Flag-tagged TRAM were transfected into HEK293 cells, and TRAM was precipitated with specific Flag Ab. PTPN4 coimmunoprecipitated with Flag-tagged TRAM (Fig. 4E). As a negative control, TRIF could not interact with PTPN4 (Fig. 4E). Collectively, these data indicated that PTPN4 could interact with TRAM.

Several posttranslational modifications are reported to be critical for the optimal function of TRAM, including myristoylation and serine phosphorylation (7, 29). The myristoylation of TRAM localizes it on the plasma membrane (29), where it undergoes phosphorylation at Ser16 and then translocates to the cytoplasm (7). TIRAP, the sorting adaptor for MyD88-dependent pathway, undergoes tyrosine phosphorylation at Tyr86, Tyr106, and Tyr159 residues within its TIR domain following LPS stimulation, and that is required for the activation of MyD88-dependent pathway in TLR2 and TLR4 signaling (12, 13). Similar to TIRAP, TRAM also bears conserved TIR domain and function as a sorting adaptor for bridging TLR4 and TRIF in TLR4 activation (46). PTPN4 inhibited TRAM-induced signal transduction via its phosphatase activity, and that promoted us to investigate whether TRAM could undergo tyrosine phosphorylation during TLR4 activation. We then detected the tyrosine phosphorylation of endogenous TRAM in macrophages following TLRs activation. LPS stimulation prompted an increase in tyrosine phosphorylation of TRAM, whereas poly(I:C) and LTA could not induce it (Fig. 5A). Next, we transfected with Flag-tagged TRAM plasmid into HEK293–TLR4 cells and examined the tyrosine phosphorylation of the ectopically expressed TRAM. Flag-tagged TRAM was tyrosine phosphorylated with LPS stimulation (Fig. 5B). These data indicate that TRAM undergoes tyrosine phosphorylation in TLR4 signaling.

To determine the tyrosine phosphorylation sites of TRAM, we first performed alignment of TRAM sequences from human, mouse, bovine, and rat. Only two tyrosine (Y) residues, which both located in the TIR domain, are conserved in the entire TRAM sequences from the four species (Fig. 5C), attesting to their possible importance in the functioning of TRAM. In humans, the two tyrosine residues are located at 154 aa (Y-154) and 167 aa (Y-167), respectively (Fig. 5D). The structure of TIR domain of TRAM from human has been determined (30). We analyzed this structure and noted that both Y-154 and Y-167 were surface exposed (Fig. 5E) and possessed the potential to be phosphorylated. Notably, Y-167 was located in the middle of a mobile randon coil (from 161 to 181 aa), which made it easier to be phosphorylated. To examine the role of these tyrosine residues in TRAM phosphorylation, we mutated the two tyrosine residues conservatively to phenylalanine (F) by using site-directed mutagenesis. The phosphorylation of TRAM in response to LPS was completely abolished when Y-167 was mutated (Fig. 5F). Mutation of Y-154 had no effect on TRAM phosphorylation (Fig. 5F), suggesting that Y-154 is not involved in LPS-induced phosphorylation of TRAM. It is therefore likely that TRAM is phosphorylated on Y-167. We then raised an Ab to a synthetic peptide comprising aa 163–174 of TRAM with a phosphotyrosine inserted instead of a tyrosine at Y-167 and applied to specifically detect the Y-167 phosphorylation of TRAM. LPS could induce TRAM Y-167 phosphorylation, and the phosphorylation peaks at 30 min in peritoneal macrophages (Fig. 5G), whereas poly(I:C), LTA, or SeV could not induce tyrosine phosphorylation of TRAM (Fig. 5G). We next tested the ability of the TRAM mutants to activate IFN-β, IRF3, and NF-κB reporter genes in HEK293 cells. Y-167 mutant TRAM (TRAM Y167F) greatly lost the ability to activate IFN-β, IRF3, or NF-κB (Fig. 5H). Taken together, these data indicate that TRAM undergoes tyrosine phosphorylation following TLR4 activation, and the Y-167 phosphorylation is critical for TRIF-dependent TLR4 pathway.

We then investigated whether PTPN4 could inhibit tyrosine phosphorylation of TRAM using specific Ab for Y-167 phosphorylation of TRAM. PTPN4 overexpression inhibited Y-167 phosphorylation of TRAM in HEK293–TLR4 cells (Fig. 6A), whereas PTPN4 knockdown had the opposite effects in mouse peritoneal macrophages (Fig. 6B). TRAM cytoplasm translocation is a key step for the activation of downstream pathway. We next investigated whether PTPN4 could regulate TRAM translocation. TRAM translocated from plasma membrane to cytoplasm following LPS stimulation and PTPN4 inhibited the dissociation of TRAM from the membrane in HEK293–TLR4 cells (Fig. 6C).

Next, we investigated whether tyrosine phosphorylation of TRAM was also required for the association between TRAM and TRIF. Hemagglutinin (HA)-tagged TRIF and Flag-tagged TRAM WT, TRAM Y154F, or TRAM Y167F were transfected into HEK293 cells. TRAM Y167F significantly reduced the ability of TRAM to interact with TRIF (Fig. 6D), indicating that Y-167 phosphorylation of TRAM is critical for TRAM–TRIF complex formation. To determine whether PTPN4 could disrupt the TRAM–TRIF complex formation, HA-tagged TRIF and Flag-tagged WT TRAM were transfected in HEK293 cells with or without Myc-tagged PTPN4. PTPN4 overexpression resulted in a reduction in the interaction between TRIF and TRAM (Fig. 6E). Altogether, these data indicated that Y167 phosphorylation of TRAM is required for the TRAM–TRIF complex formation and PTPN4 could suppress TRAM–TRIF interaction by inhibiting Y167 phosphorylation of TRAM.

In the current study, we reported that TRAM underwent tyrosine phosphorylation, and this modification was required for optimal activation of TRAM. Moreover, we identified that the Y167 residue located in TIR domain of TRAM was the phospho-accepting tyrosine, and Y167 mutation of TRAM lost the activity of triggering downstream signals. It has been reported that TIRAP underwent tyrosine phosphorylation at the residues within its TIR domain following LPS stimulation, and this modification initiated a conformational change in the TIR domain of TIRAP (12, 13), thereby leading to activation of downstream signals. Tyrosine phosphorylation results in the recruitment of the phosphorylase kinase, such as Ser/Thr kinase and PI3K, to the signaling complex, a step that is essential for the activation of downstream transcription factors (11). Tyrosine phosphorylation then provides docking sites for other proteins (11). Structural analysis indicates that both TRIF and TRAM TIR domains form a BB-loop–mediated homodimer (30). The dimerization of TRAM TIR presents an interaction surface for TRIF (30). Tyrosine phosphorylation within the TIR domain of TRAM may initiate the conformational change and then provide docking sites to facilitate the interaction between TRAM and TRIF.

Modulating the TRAM–TRIF interaction is an important mechanism for the specific regulation of TRIF-dependent pathway in TLR4 signaling. TRAM adaptor with GOLD (TAG) domain and transmembrane emp24 domain-containing protein 7 (TMED7) have been reported to be involved in the regulation of TRIF-dependent TLR4 pathway. Transmembrane emp24 domain-containing protein 7interacts with TRAM and displaces the TRIF from TRAM in endosomes (31). TMED7 colocalizes with TRAM and TLR4 in endosomes, where it encounters TAG and mediates the inhibitory effects of TAG on TRAM–TRIF association (32). Here we identified PTPN4 as another specific inhibitor by directly targeting TRAM via its phosphatase activity. PTPN4 attenuates TRAM tyrosine phosphorylation, inhibits its cytoplasm translocation and subsequent TRAM–TRIF interaction, resulting in the decrease of TLR4-mediated IRF3 activation and IFN-β production. Thus, our results provide new insight into the TLR4 signaling and mechanisms that specifically regulate the production of type I IFNs and proinflammatory cytokines in TLR4 signaling.

TLR4 engagement initiates a complex signaling pathway that culminates in the induction of many immune proteins including both proinflammatory cytokines and type I IFNs (2, 3, 31). So far, several diseases, such as endotoxin shock, rheumatoid, atherosclerosis, arthritis, asthma, and ischemia–reperfusion injury, were documented to be linked with TLR4. Inappropriate TLR4 activation and unbalanced production of type I IFNs and proinflammatory cytokines can promote the development of multiple diseases. Thus, elucidating the mechanisms by which differentially regulates the production of type I IFNs and proinflammatory cytokines in TLR4 signaling would be beneficial for the development of new strategy to control TLR4-associated diseases. Although multiple molecules have been identified as regulators of TLR4 signaling up to now, few of them were specific for TRIF-dependent TLR4 pathway. In the current study, we identified PTPN4 as a specific inhibitor of TRIF-dependent TLR4 pathway. Targeting PTPN4 would be beneficial for the development of new strategy to control TLR4-associated diseases without unwanted side effects.

We thank Drs. Nicolai S. van Oers, Nicholas K. Tonks, Luke A. O’Neill, and Xuetao Cao for providing plasmids.

This work was supported by grants from the National Natural Science Foundation of China (Grant 31370017), Shandong Provincial Nature Science Foundation for Distinguished Young Scholars (Grant JQ201420), and the National “973” Program of China (Grant 2011CB503906).

Abbreviations used in this article:

HA

hemagglutinin

HEK

human embryonic kidney

IP

immunoprecipitation

LTA

lipothechoic acid

PGN

peptidoglycan

poly(I:C)

polyinosinic:polycytidylic acid

PRR

pattern recognition receptor

PTPN4

protein tyrosine phosphatase nonreceptor type 4

SeV

Sendai virus

TAG

TRAM adaptor with GOLD

TIR

Toll-IL 1R

TIRAP

TIR domain–containing adapter protein

TRAM

TRIF-related adaptor molecule

WT

wild-type.

1
Janeway
C. A.
 Jr.
,
Medzhitov
R.
.
2002
.
Innate immune recognition.
Annu. Rev. Immunol.
20
:
197
216
.
2
Kawai
T.
,
Akira
S.
.
2010
.
The role of pattern-recognition receptors in innate immunity: update on Toll-like receptors.
Nat. Immunol.
11
:
373
384
.
3
O’Neill
L. A.
,
Golenbock
D.
,
Bowie
A. G.
.
2013
.
The history of Toll-like receptors—redefining innate immunity.
Nat. Rev. Immunol.
13
:
453
460
.
4
Fitzgerald
K. A.
,
Rowe
D. C.
,
Barnes
B. J.
,
Caffrey
D. R.
,
Visintin
A.
,
Latz
E.
,
Monks
B.
,
Pitha
P. M.
,
Golenbock
D. T.
.
2003
.
LPS-TLR4 signaling to IRF-3/7 and NF-κB involves the Toll adapters TRAM and TRIF.
J. Exp. Med.
198
:
1043
1055
.
5
Yamamoto
M.
,
Sato
S.
,
Hemmi
H.
,
Uematsu
S.
,
Hoshino
K.
,
Kaisho
T.
,
Takeuchi
O.
,
Takeda
K.
,
Akira
S.
.
2003
.
TRAM is specifically involved in the Toll-like receptor 4-mediated MyD88-independent signaling pathway.
Nat. Immunol.
4
:
1144
1150
.
6
Oshiumi
H.
,
Sasai
M.
,
Shida
K.
,
Fujita
T.
,
Matsumoto
M.
,
Seya
T.
.
2003
.
TIR-containing adapter molecule (TICAM)-2, a bridging adapter recruiting to Toll-like receptor 4 TICAM-1 that induces interferon-β.
J. Biol. Chem.
278
:
49751
49762
.
7
McGettrick
A. F.
,
Brint
E. K.
,
Palsson-McDermott
E. M.
,
Rowe
D. C.
,
Golenbock
D. T.
,
Gay
N. J.
,
Fitzgerald
K. A.
,
O’Neill
L. A.
.
2006
.
Trif-related adapter molecule is phosphorylated by PKCε during Toll-like receptor 4 signaling.
Proc. Natl. Acad. Sci. USA
103
:
9196
9201
.
8
Kagan
J. C.
,
Su
T.
,
Horng
T.
,
Chow
A.
,
Akira
S.
,
Medzhitov
R.
.
2008
.
TRAM couples endocytosis of Toll-like receptor 4 to the induction of interferon-beta.
Nat. Immunol.
9
:
361
368
.
9
Wies
E.
,
Wang
M. K.
,
Maharaj
N. P.
,
Chen
K.
,
Zhou
S.
,
Finberg
R. W.
,
Gack
M. U.
.
2013
.
Dephosphorylation of the RNA sensors RIG-I and MDA5 by the phosphatase PP1 is essential for innate immune signaling.
Immunity
38
:
437
449
.
10
Qu
Y.
,
Misaghi
S.
,
Izrael-Tomasevic
A.
,
Newton
K.
,
Gilmour
L. L.
,
Lamkanfi
M.
,
Louie
S.
,
Kayagaki
N.
,
Liu
J.
,
Kömüves
L.
, et al
.
2012
.
Phosphorylation of NLRC4 is critical for inflammasome activation.
Nature
490
:
539
542
.
11
Chattopadhyay
S.
,
Sen
G. C.
.
2014
.
Tyrosine phosphorylation in Toll-like receptor signaling.
Cytokine Growth Factor Rev.
25
:
533
541
.
12
Gray
P.
,
Dunne
A.
,
Brikos
C.
,
Jefferies
C. A.
,
Doyle
S. L.
,
O’Neill
L. A.
.
2006
.
MyD88 adapter-like (Mal) is phosphorylated by Bruton’s tyrosine kinase during TLR2 and TLR4 signal transduction.
J. Biol. Chem.
281
:
10489
10495
.
13
Piao
W.
,
Song
C.
,
Chen
H.
,
Wahl
L. M.
,
Fitzgerald
K. A.
,
O’Neill
L. A.
,
Medvedev
A. E.
.
2008
.
Tyrosine phosphorylation of MyD88 adapter-like (Mal) is critical for signal transduction and blocked in endotoxin tolerance.
J. Biol. Chem.
283
:
3109
3119
.
14
Han
C.
,
Jin
J.
,
Xu
S.
,
Liu
H.
,
Li
N.
,
Cao
X.
.
2010
.
Integrin CD11b negatively regulates TLR-triggered inflammatory responses by activating Syk and promoting degradation of MyD88 and TRIF via Cbl-b.
Nat. Immunol.
11
:
734
742
.
15
Mustelin
T.
,
Vang
T.
,
Bottini
N.
.
2005
.
Protein tyrosine phosphatases and the immune response.
Nat. Rev. Immunol.
5
:
43
57
.
16
Rhee
I.
,
Veillette
A.
.
2012
.
Protein tyrosine phosphatases in lymphocyte activation and autoimmunity.
Nat. Immunol.
13
:
439
447
.
17
An
H.
,
Hou
J.
,
Zhou
J.
,
Zhao
W.
,
Xu
H.
,
Zheng
Y.
,
Yu
Y.
,
Liu
S.
,
Cao
X.
.
2008
.
Phosphatase SHP-1 promotes TLR- and RIG-I‑activated production of type I interferon by inhibiting the kinase IRAK1.
Nat. Immunol.
9
:
542
550
.
18
An
H.
,
Zhao
W.
,
Hou
J.
,
Zhang
Y.
,
Xie
Y.
,
Zheng
Y.
,
Xu
H.
,
Qian
C.
,
Zhou
J.
,
Yu
Y.
, et al
.
2006
.
SHP-2 phosphatase negatively regulates the TRIF adaptor protein-dependent type I interferon and proinflammatory cytokine production.
Immunity
25
:
919
928
.
19
Wang
Y.
,
Shaked
I.
,
Stanford
S. M.
,
Zhou
W.
,
Curtsinger
J. M.
,
Mikulski
Z.
,
Shaheen
Z. R.
,
Cheng
G.
,
Sawatzke
K.
,
Campbell
A. M.
, et al
.
2013
.
The autoimmunity-associated gene PTPN22 potentiates toll-like receptor-driven, type 1 interferon-dependent immunity.
Immunity
39
:
111
122
.
20
Zhang
P.
,
Liu
X.
,
Li
Y.
,
Zhu
X.
,
Zhan
Z.
,
Meng
J.
,
Li
N.
,
Cao
X.
.
2013
.
Protein tyrosine phosphatase with proline-glutamine-serine-threonine-rich motifs negatively regulates TLR-triggered innate responses by selectively inhibiting IκB kinase β/NF-κB activation.
J. Immunol.
190
:
1685
1694
.
21
Xu
H.
,
An
H.
,
Hou
J.
,
Han
C.
,
Wang
P.
,
Yu
Y.
,
Cao
X.
.
2008
.
Phosphatase PTP1B negatively regulates MyD88- and TRIF-dependent proinflammatory cytokine and type I interferon production in TLR-triggered macrophages.
Mol. Immunol.
45
:
3545
3552
.
22
Young
J. A.
,
Becker
A. M.
,
Medeiros
J. J.
,
Shapiro
V. S.
,
Wang
A.
,
Farrar
J. D.
,
Quill
T. A.
,
Hooft van Huijsduijnen
R.
,
van Oers
N. S.
.
2008
.
The protein tyrosine phosphatase PTPN4/PTP-MEG1, an enzyme capable of dephosphorylating the TCR ITAMs and regulating NF-κB, is dispensable for T cell development and/or T cell effector functions.
Mol. Immunol.
45
:
3756
3766
.
23
Babault
N.
,
Cordier
F.
,
Lafage
M.
,
Cockburn
J.
,
Haouz
A.
,
Prehaud
C.
,
Rey
F. A.
,
Delepierre
M.
,
Buc
H.
,
Lafon
M.
,
Wolff
N.
.
2011
.
Peptides targeting the PDZ domain of PTPN4 are efficient inducers of glioblastoma cell death.
Structure
19
:
1518
1524
.
24
Huai
W.
,
Zhao
R.
,
Song
H.
,
Zhao
J.
,
Zhang
L.
,
Zhang
L.
,
Gao
C.
,
Han
L.
,
Zhao
W.
.
2014
.
Aryl hydrocarbon receptor negatively regulates NLRP3 inflammasome activity by inhibiting NLRP3 transcription.
Nat. Commun.
5
:
4738
.
25
Zhao
W.
,
Wang
L.
,
Zhang
M.
,
Yuan
C.
,
Gao
C.
.
2012
.
E3 ubiquitin ligase tripartite motif 38 negatively regulates TLR-mediated immune responses by proteasomal degradation of TNF receptor-associated factor 6 in macrophages.
J. Immunol.
188
:
2567
2574
.
26
Flint
A. J.
,
Tiganis
T.
,
Barford
D.
,
Tonks
N. K.
.
1997
.
Development of “substrate-trapping” mutants to identify physiological substrates of protein tyrosine phosphatases.
Proc. Natl. Acad. Sci. USA
94
:
1680
1685
.
27
Zhang
M.
,
Wang
L.
,
Zhao
X.
,
Zhao
K.
,
Meng
H.
,
Zhao
W.
,
Gao
C.
.
2012
.
TRAF-interacting protein (TRIP) negatively regulates IFN-β production and antiviral response by promoting proteasomal degradation of TANK-binding kinase 1.
J. Exp. Med.
209
:
1703
1711
.
28
Platanias
L. C.
2005
.
Mechanisms of type-I‑ and type-II-interferon‑mediated signalling.
Nat. Rev. Immunol.
5
:
375
386
.
29
Rowe
D. C.
,
McGettrick
A. F.
,
Latz
E.
,
Monks
B. G.
,
Gay
N. J.
,
Yamamoto
M.
,
Akira
S.
,
O’Neill
L. A.
,
Fitzgerald
K. A.
,
Golenbock
D. T.
.
2006
.
The myristoylation of TRIF-related adaptor molecule is essential for Toll-like receptor 4 signal transduction.
Proc. Natl. Acad. Sci. USA
103
:
6299
6304
.
30
Enokizono
Y.
,
Kumeta
H.
,
Funami
K.
,
Horiuchi
M.
,
Sarmiento
J.
,
Yamashita
K.
,
Standley
D. M.
,
Matsumoto
M.
,
Seya
T.
,
Inagaki
F.
.
2013
.
Structures and interface mapping of the TIR domain‑containing adaptor molecules involved in interferon signaling.
Proc. Natl. Acad. Sci. USA
110
:
19908
19913
.
31
Palsson-McDermott
E. M.
,
Doyle
S. L.
,
McGettrick
A. F.
,
Hardy
M.
,
Husebye
H.
,
Banahan
K.
,
Gong
M.
,
Golenbock
D.
,
Espevik
T.
,
O’Neill
L. A.
.
2009
.
TAG, a splice variant of the adaptor TRAM, negatively regulates the adaptor MyD88-independent TLR4 pathway.
Nat. Immunol.
10
:
579
586
.
32
Doyle
S. L.
,
Husebye
H.
,
Connolly
D. J.
,
Espevik
T.
,
O’Neill
L. A.
,
McGettrick
A. F.
.
2012
.
The GOLD domain‑containing protein TMED7 inhibits TLR4 signalling from the endosome upon LPS stimulation.
Nat. Commun.
3
:
707
.

The authors have no financial conflicts of interest.