Report

Role of Adaptor TRIF in the MyD88-Independent Toll-Like Receptor Signaling Pathway

See allHide authors and affiliations

Science  01 Aug 2003:
Vol. 301, Issue 5633, pp. 640-643
DOI: 10.1126/science.1087262

Abstract

Stimulation of Toll-like receptors (TLRs) triggers activation of a common MyD88-dependent signaling pathway as well as a MyD88-independent pathway that is unique to TLR3 and TLR4 signaling pathways leading to interferon (IFN)-β production. Here we disrupted the gene encoding a Toll/IL-1 receptor (TIR) domain-containing adaptor, TRIF. TRIF-deficient mice were defective in both TLR3- and TLR4-mediated expression of IFN-β and activation of IRF-3. Furthermore, inflammatory cytokine production in response to the TLR4 ligand, but not to other TLR ligands, was severely impaired in TRIF-deficient macrophages. Mice deficient in both MyD88 and TRIF showed complete loss of nuclear factor kappa B activation in response to TLR4 stimulation. These findings demonstrate that TRIF is essential for TLR3- and TLR4-mediated signaling pathways facilitating mammalian antiviral host defense.

TLRs recognize specific patterns of microbial components and are critical in provoking innate immune responses through activation of signaling cascades via Toll/IL-1 receptor (TIR) domain-containing adaptors, such as MyD88 and TIRAP (1, 2). MyD88 is common to all the TLRs, whereas TIRAP is specifically involved in TLR2- and TLR4-mediated signaling pathways (35). In addition to the common MyD88-dependent pathway, TLR3 and TLR4 utilize a MyD88-independent signaling pathway that leads to the activation of IRF-3 and induction of IFN-β (68). TIR domain-containing adaptor inducing IFN-β (TRIF) was recently identified as a third adaptor and was shown to activate IFN-β expression via TLR3 (9, 10).

To assess the physiological role of TRIF, we generated mice lacking the Trif gene [supporting online material (SOM) Text and fig. S1]). Mutant mice homozygous for the disrupted Trif allele were born at the expected mendelian ratio and grew to be healthy in specific-pathogen–free conditions.

Previous in vitro studies have suggested that TRIF is involved in production of IFN-β in response to double-stranded RNA and its analog poly(I:C), both of which are recognized by TLR3 (911). Therefore, we first analyzed poly(I:C)-induced messenger RNA (mRNA) expression of IFN-β and several IFN-inducible genes, such as RANTES, IP-10, and MCP-1, in peritoneal macrophages (Fig. 1A). Macrophages from TRIF–/– mice showed impaired expression of IFN-β and IFN-inducible genes in response to poly(I:C), which is consistent with results seen in TLR3–/– mice (11, 12). In addition, splenocytes from TRIF–/– mice showed severely defective proliferation in response to poly(I:C), but not to the TLR9 ligand CpG DNA (Fig. 1B). TRIF–/– B cells were also severely impaired in poly(I:C)-induced, but not anti-IgM Ab (antibody to immunoglobulin M)–induced, augmentation of surface expression of CD69, CD86, and major histocompatibility complex (MHC) class II (Fig. 1C). Thus, TRIF–/– mice showed defective responses to poly(I: C), indicating that TRIF is essential for TLR3-mediated signaling pathways.

Fig. 1.

Impaired responses to poly(I:C) in TRIF-deficient cells. (A) Peritoneal macrophages were stimulated with 50 μg/ml poly(I:C) for the indicated periods. Total RNA (5 μg) was extracted and subjected to Northern blot analysis for expression of IFN-β, IP-10, RANTES, and MCP-1. The same membrane was rehybridized with a β-actin probe. (B) Proliferation of poly(I:C)- or CpG DNA–stimulated splenocytes. Splenocytes were cultured with the indicated concentration of poly(I:C) or CpG DNA for 24 hours. [3H]-thymidine (1 μCi) was pulsed for the last 12 hours. [3H]-thymidine incorporation was measured by a scintillation counter. (C) Splenic B220+ cells were cultured with 50 μg/ml poly(I:C) or 10 μg/ml anti-IgM antibody. At the 36-hour culture period, cells were harvested and stained with biotin-conjugated antibodies to CD69 or CD86, or antibody to I-Ab followed by streptavidin-phycoerythrin (PE). Stained cells were analyzed on FACS Calibur using Cell Quest software. Area under dashed line represents cells cultured in medium; under solid line, α-IgM; shaded in gray, poly(I:C).

In addition to the TLR3 ligand, the TLR4 ligand LPS has been shown to induce IFN-β and subsequent expression of IFN-inducible genes in a MyD88-independent manner (68). We analyzed LPS-induced mRNA expression of IFN-inducible genes such as RANTES, IP-10, and MCP-1 in embryonic fibroblast cells from TRIF-deficient mice (Fig. 2A). In TRIF–/– cells, LPS-induced expression of IFN-inducible genes was severely reduced, demonstrating that TRIF–/– mice were also defective in MyD88-independent responses to LPS.

Fig. 2.

Impaired responses to LPS in TRIF-deficient cells. (A) Embryonic fibroblasts were stimulated with 10 μg/ml LPS for the indicated periods. Total RNA (10 μg) was extracted and subjected to Northern blot analysis for expression of IP-10, RANTES, and MCP-1. The same membrane was rehybridized with a β-actin probe. (B) Peritoneal macrophages from TRIF-deficient mice or wild-type mice were left unstimulated or stimulated with 10 μg/ml peptidoglycan (PGN), 100 ng/ml LPS, 100 nM R-848, or CpG DNA (1 μM) in the presence of 30 ng/ml IFN-γ. Supernatants were collected for IL-12 p40 analysis by enzyme-linked immunosorbent assay (ELISA) 24 hours later. Indicated values are means ±SD of triplicates. N.D., not detected. (C) Peritoneal macrophages were stimulated with the indicated concentrations of LPS in the presence of 30 ng/ml IFN-γ for 24 hours, and the supernatants were subjected to measurement of TNF-α, IL-6, and IL-12 p40 by ELISA. Indicated values are means ±SD of triplicates. (D) Proliferation of LPS- or CpG DNA–stimulated splenocytes. Splenocytes were cultured with the indicated concentrations of LPS or CpG DNA for 24 hours. [3H]-thymidine (1 μCi) was pulsed for the last 12 hours. [3H]-thymidine incorporation was measured by a scintillation counter. (E) Splenic B220+ cells were cultured with 10 μg/ml LPS or 10 μg/ml anti-IgM antibody. At the 36-hour culture period, cells were harvested and stained with biotin-conjugated antibodies to CD69 or CD86 antibody followed by streptavidin-PE. Stained cells were analyzed on FACS Calibur using Cell Quest software. Areas under dashed and solid lines are as in Fig. 1C. Area under shaded line is for LPS.

We next analyzed inflammatory cytokine production in response to several TLR ligands, which is induced in a MyD88-dependent manner (Fig. 2B). Both wild-type and TRIF–/– macrophages produced similar levels of interleukin-12 (IL-12) p40 in response to the TLR2 ligand peptidoglycan, the TLR7 ligand R-848, and the TLR9 ligand CpG DNA (13, 14). However, LPS-induced production of tumor necrosis factor–α (TNF-α), IL-6, and IL-12 p40 was abolished in TRIF–/– macrophages (Fig. 2, B and C). All wild-type animals injected with a lethal dose of LPS upon D-GalN (D-galactosamine hydrochloride) sensitization produced high concentrations of TNF-α and IL-6 and died within 24 hours of injection. (SOM Text and fig. S2). By contrast, TRIF–/– mice showed defective production of inflammatory cytokines and resistance to LPS-induced septic shock. In addition, splenocytes from TRIF–/– mice showed profoundly impaired proliferative responses to LPS, although these cells proliferated normally in response to R-848 (Fig. 2D) (13). Furthermore, LPS-induced augmentation of CD69 and CD86 expression was severely reduced in TRIF–/– B cells, although anti-IgM Ab–induced expression was unaffected (Fig. 2E). Thus, TRIF–/– mice showed defective responses to LPS in terms of both MyD88-dependent and MyD88-independent responses.

MyD88-independent signaling through TLR3 and TLR4 has been reported to activate the transcription factor IRF-3 (6, 8). Indeed, native–polyacrylamide gel electrophoresis (PAGE) analysis showed poly(I: C)- and LPS-induced formation of homophylic IRF-3 dimers in wild-type lung fibroblast cells and peritoneal macrophages, respectively (Fig. 3, A and B). However, poly(I:C)- or LPS-induced dimerization of IRF-3 was not observed in TRIF–/– cells. These findings demonstrate that TRIF is essential for MyD88-independent activation of IRF-3 in response to TLR3 and TLR4 signaling. In TLR3–/– and TRIF–/– cells, poly(I:C)-induced NF-κB activation was also severely impaired, indicating that TRIF plays a major role in TLR3-mediated NF-κB activation (Fig. 3C). In sharp contrast, LPS stimulation led to almost normal activation of NF-κB and mitogen-activated protein (MAP) kinase JNK in TRIF–/– embryonic fibroblast cells (Fig. 3D). Moreover, LPS-induced autophosphorylation of IRAK-1, which is activated just downstream of MyD88, was intact in TRIF–/– macrophages (SOM Text and fig. S3). In the case of TLR4 signaling, the MyD88-dependent pathway leads to an early-phase activation of NF-κB and MAP kinases, whereas the MyD88-independent pathway induces a late-phase activation of NF-κB and MAP kinases, as demonstrated by a delayed NF-κB and MAP kinase activation in MyD88–/– cells (6, 15). Therefore, we hypothesized that the impairment in the TRIF-dependent late NF-κB and JNK activation was masked by the MyD88-dependent early activation in TRIF–/– mice. To test this directly, we generated mice lacking both TRIF and MyD88. In embryonic fibroblast cells from TRIF/MyD88 double deficient mice, LPS-induced activation of NF-κB and JNK was completely inhibited (Fig. 3E). In addition, LPS induction of IFN-inducible genes such as IP-10, MCP-1, and RANTES was completely abolished in TRIF/MyD88 double-deficient cells (Fig. 3F). These findings clearly demonstrate that TRIF is essential for TLR4-mediated activation of the MyD88-independent signaling pathway.

Fig. 3.

Activation of signaling cascade in TRIF-deficient and TRIF/MyD88 double-knockout cells. Lung fibroblasts (A) or peritoneal macrophages (B) were stimulated with (A) 50 μg/ml poly (I:C) or (B) 1 μg/ml LPS for the indicated periods. Cell lysates were prepared and subjected to native-PAGE. Monomeric (arrow) and dimeric (arrowhead) forms of IRF-3 were detected by Western blot. Lung fibroblasts were stimulated with (C) 50 μg/ml poly(I:C) and (D) 10 μg/ml LPS for the indicated periods. Nuclear extracts were prepared, and NF-κB DNA binding activity was determined by electrophoretic mobility shift assay (EMSA) using an NF-κB–specific probe. Arrows and asterisks indicate induced NF-κB complex and nonspecific bands (N.S.), respectively (upper panel). For the LPS-stimulated cells, JNK1 activation was also determined by western blotting using anti–phospho-JNK specific antibody toward the cell extracts (lower panel). (E) Embryonic fibroblasts from wild-type and TRIF/MyD88 double-deficient (DKO) mice were stimulated with 10 μg/ml LPS or 10 ng/ml TNF-α for the indicated periods. Nuclear extracts were prepared and NF-κB DNA binding activity was determined by EMSA using an NF-κB–specific probe. Arrows and asterisks indicate induced NF-κB complex and nonspecific bands, respectively (upper panel). For the LPS stimulated cells, JNK1 activation was also determined by western blotting using anti–phospho-JNK specific antibody toward the cell extracts (lower panel). (F) Embryonic fibroblasts from wild-type and TRIF/MyD88 DKO mice were stimulated with 10 μg/ml LPS for the indicated periods. Total RNA (10 μg) was extracted and subjected to Northern blot analysis for expression of IP-10, RANTES, and MCP-1. The same membrane was rehybridized with a β-actin probe.

We report the physiological function of TRIF revealed by analysis of TRIF–/– mice, in which the MyD88-independent response induced by the TLR3 and TLR4 ligands was severely impaired. TRIF is an essential adaptor in TLR3 signaling because all the poly(I:C)-induced responses were abolished in TRIF–/– mice. In the case of TLR4 stimulation, TRIF–/– mice displayed normal LPS-induced MyD88-dependent activation of IRAK-1, NF-κB, and MAP kinase, indicating that TRIF is not involved in the LPS-induced activation of the MyD88-dependent signaling. However, TRIF–/– mice were impaired in the LPS-induced inflammatory cytokine production, which needs activation of the MyD88-dependent signaling. In this regard, we propose that cooperation of the MyD88-dependent and MyD88-independent (TRIF-dependent) signaling pathways are required for the TLR4-mediated inflammatory cytokine production. Recently, two noncanonical IκB kinases (IKKs), IKKϵ and TBK1, have been demonstrated to activate IRF-3 downstream of TRIF (16, 17). These IKKs have originally been shown to activate NF-κB (1821). Therefore, IKKϵ and TBK1 are likely to have a role in TRIF-dependent activation of both NF-κB and IRF-3. Elucidation of physiologic roles for these IKKs in the TLR-mediated signaling pathways will clarify the precise mechanism by which the TRIF-dependent (MyD88-independent) signaling leads to activation of NF-κB and IRF-3.

In summary, we have identified an essential adaptor that regulates the MyD88-independent pathway, which is considered to be important in mediating antiviral host response. A precise analysis on the involvement of TRIF in the viral response of TRIF–/– mice will provide a insight into the relation between TLRs and viral recognition.

Supporting Online Material

www.sciencemag.org/cgi/content/full/1087262/DC1

Materials and Methods

SOM Text

Figs. S1 to S3

References

References and Notes

View Abstract

Navigate This Article