Poly(I-C)-induced Toll-like receptor 3 (TLR3)-mediated activation of NFkappa B and MAP kinase is through an interleukin-1 receptor-associated kinase (IRAK)-independent pathway employing the signaling components TLR3-TRAF6-TAK1-TAB2-PKR .

Recent studies show that a member of the interleukin-1 (IL-1)/Toll receptor superfamily, Toll-like receptor 3 (TLR3), recognizes double-stranded RNA (dsRNA). Because of the similarity in their cytoplasmic domains, IL-1/Toll receptors share signaling components that associate with the IL-1 receptor, including IL-1 receptor-associated kinase (IRAK), MyD88, and TRAF6. However, we find that, in response to dsRNA, TLR3 can mediate the activation of both NFkappaB and mitogen-activated protein (MAP) kinases in IL-1-unresponsive mutant cell lines, including IRAK-deficient I1A and I3A cells, which are defective in a component that is downstream of IL-1R but upstream of IRAK. These results clearly indicate that TLR3 does not simply share the signaling components employed by the IL-1 receptor. Through biochemical analyses we have identified an IRAK-independent TLR3-mediated pathway. Upon binding of dsRNA to TLR3, TRAF6, TAK1, and TAB2 are recruited to the receptor to form a complex, which then translocates to the cytosol where TAK1 is phosphorylated and activated. The dsRNA-dependent protein kinase (PKR) is also detected in this signal-induced TAK1 complex. Kinase inactive mutants of TAK1 (TAK1DN) and PKR (PKRDN) inhibit poly(dI.dC)-induced TLR3-mediated NFkappaB activation, suggesting that both of these kinases play important roles in this pathway.

The interleukin-1 (IL-1) 1 /Toll receptors play essential roles in inflammation and innate immunity. The defining feature of members of the superfamily is a TIR (Toll/IL-1 receptor) domain on the cytoplasmic side of the receptors (1, 2). The first described members of this superfamily are in the IL-1R family. These receptors contain three Ig domains in their extracellular regions (3)(4)(5). The second group includes only one receptor so far, the single Ig IL-1R-related molecule (SIGIRR), which has only a single extracellular Ig domain (6). The last group in the superfamily is the recently identified pattern recognition receptors, the Toll-like receptors (TLRs), 10 members of which contain two major domains characterized by extracellular leucine-rich repeats and an intracellular Toll-like domain (2,(7)(8)(9)(10). TLR4 has been genetically identified as a signaling molecule essential for the responses to LPS, a feature of Gramnegative bacteria (11). Mice with targeted disruption of the TLR4 gene are LPS-unresponsive. Unlike TLR4, TLR2 responds to mycobacteria, yeast, and Gram-positive bacteria (12)(13)(14)(15). TLR9 has been shown to recognize bacterial DNA (10), whereas TLR5 mediates the induction of the immune response by bacterial flagellins (16). Recent studies show that TLR3 recognizes dsRNA (17).
Much progress has been made in understanding the IL-1Rmediated signaling. Upon IL-1 stimulation, the cytosolic proteins MyD88 (18 -20) and Tollip (21) are recruited to the receptor complex, which then recruits serine-threonine kinases IRAK4 (IL-1 receptor-associated kinase 4) 2 (22,23) and IRAK (24). Although IRAK is hyperphosphorylated, mediating the recruitment of TRAF6 to the receptor complex (25), IRAK4 is probably the kinase that functions upstream of and phosphorylates IRAK (22,23). IRAK-TRAF6 then leaves the receptor complex to interact with TAK1, a member of the MAP kinase kinase kinase family, and the two proteins that bind to it, TAB1 and TAB2, on the membrane (25)(26)(27). TAK1 and TAB2 are phosphorylated on the membrane followed by the formation and translocation of TRAF6-TAK1-TAB1-TAB2 from the membrane to the cytosol (25), where TAK1 is activated (25). Although genetic studies show that IRAK is required for the IL-1-induced activation of TAK1 (28), in vitro biochemical analyses reveal that TRAF6-mediated ubiquitination may also play an important role in TAK1 activation (29). The activation of TAK1 eventually leads to the activation of IB kinase (IKK) by an unknown mechanism. Activated IKK phosphorylates IB proteins, which are degraded, releasing NFB to activate transcription in the nucleus (30 -33). Activated TAK1 has also been implicated in the IL-1-induced activation of MKK6 and JNK (26), leading to the phosphorylation and activation of activating transcription factor (ATF) and activating protein 1 (AP1), which thereby also activates gene transcription.
Because the receptors in the IL-1/Toll receptor superfamily are similar in their cytoplasmic domains, they share some signaling components that function with the IL-1 receptor, including MyD88, IRAK4, IRAK, and TRAF6, leading to the activation of NFB (22, 34 -36). Although studies with MyD88null mice have demonstrated that MyD88 functions as a general adaptor for the IL-1/Toll receptor superfamily (13,20,37), these studies also revealed a MyD88-independent pathway that leads to NFB and JNK activation in TLR4-dependent signaling. MyD88 adaptor-like protein (Mal/TIRAP), which was implicated in the activation of the MyD88-independent signaling pathway (38,39), has now clearly been shown to function for the MyD88-dependent pathway via TLR2 and TLR4 (40,41). PKR-deficient mice exhibit reduced responses to different TLR ligands and reduced production of proinflammatory cytokines in response to LPS, suggesting that PKR is an intermediary in TLR signaling (42). These studies indicate that the IL-1/Toll receptors probably mediate much more complex signaling pathways than simply sharing the same signaling components of the IL-1 pathway.
TLR3-deficient mice show reduced responses to poly(dI⅐dC) and reduced production of inflammatory cytokines, indicating that TLR3 mediates the induction of immune response by dsRNA. Although TLR3 induces cytokine production through a signaling pathway dependent on MyD88, poly(dI⅐dC) can still induce the activation of NFB and MAP kinases in MyD88deficient macrophages, suggesting that both MyD88-dependent and -independent pathways may be involved in TLR3-mediated signaling. However, the detailed molecular mechanism by which TLR3 transduces the signal is unclear. In this study, we took both biochemical and genetic approaches to the study of TLR3-mediated signaling. We report here that TLR3-mediated activation of NFB and MAP kinases is through an IRAKindependent pathway employing the signaling components TLR3-TRAF6-TAK1-TAB2-PKR.
Recombinant Plasmids and Stable Transfection-pE-selectin-luc, an NFB-dependent E-selectin-luciferase reporter plasmid, was described by Schindler and Baichwal (44). Dominant negative TAK1 (TAK1DN-K66W) was a kind gift from Dr. Kunihiro Matsumoto (Nagoya University, Japan). PKRDN (K296R) was previous l y described (45). For stable transfections, 2 ϫ 10 5 cells were seeded onto a 10-cm plate and cotransfected the following day by the calcium phosphate method with 10 g of each expression vector and 1 g of pBabePuro. After 48 h, the cells were selected with 1 g/ml puromycin until clones appeared.
In Vitro Phosphorylation Assay-TAK1 and TAB2 immunoprecipi-tates were incubated with 1 g of bacterially expressed MKK6 in 20 l of kinase buffer containing 10 mM HEPES, pH 7.4, 1 mM dithiothreitol, 5 mM MgCl 2 , and 5 Ci of ␥ 32 P-labeled ATP (3000 Ci/mmol) at 25°C for 2 min. Samples were resolved by SDS-PAGE, transferred to Immobilon-P membranes, and visualized by autoradiography. The membranes were also analyzed by immunoblotting. Subcellular Fractionation-The method as described previously (46) was used with minor modifications. Cells were lysed in a hypotonic buffer (10 mM HEPES, pH 7.4, 1.5 mM MgCl 2 , 10 mM KCl, 0.2 mM phenylmethylsulfonyl fluoride, and 0.5 mM dithiothreitol) and homogenized on ice with a Dounce homogenizer. Unlysed cells, cell debris, and nuclei were removed by centrifugation at 1000 ϫ g for 5 min. Soluble (supernatant, S-100) and particulate (pellet, P-100) fractions were generated by centrifugation at 100,000 ϫ g for 1 h at 4°C. Particulate fractions were solubilized in the Triton-containing lysis buffer followed by centrifugation at 10,000 ϫ g for 10 min.
Transfection and Reporter Assays-For stable transfections, 2 ϫ 10 5 cells were seeded onto a 10-cm plate and cotransfected the following day by the calcium phosphate method with 10 g of each expression vector and 1 g of pBabePuro vector. After 48 h, the cells were selected with 1 g/ml puromycin until clones appeared. For reporter assays, 2 ϫ 10 5 cells were transfected by the same procedure with 1 g of pE-selectinluc, 1 g of pSV 2 -␤-gal, and 100 ng of each expression construct. After 48 h, the cells were split onto two 35-mm plates and the next day were stimulated with poly(dI⅐dC) for 4 h before harvest. Luciferase and ␤-galactosidase activities were determined by using the luciferase assay system and chemiluminescent reagents from Promega (Madison, WI).
Gel Shift Assays-An NFB binding site (5Ј-GAGCAGAGGGAAAT-TCCGTAACTT-3Ј) from the IP-10 gene was used as a probe (47). Complementary oligonucleotides, end-labeled with polynucleotide kinase (Roche Applied Science) and ␥ 32 P-labeled ATP, were annealed by slow cooling. Approximately 20,000 cpm of probe were used per assay (48). Whole cell extracts were used for the assay. The binding reaction was carried out at room temperature for 20 min in a total volume of 20 l containing 20 mM HEPES buffer, pH 7.0, 10 mM KCl, 0.1% Nonidet P-40, 0.5 mM dithiothreitol, 0.25 mM phenylmethylsulfonyl fluoride, and 10% glycerol.

IRAK Is Not Involved in TLR3-mediated Signaling-We
have previously taken a genetic approach to the study of IL-1dependent signaling pathways, through random mutagenesis, generating IL-1-unresponsive cell lines lacking specific components of the pathways. Mutant cell line I1A lacks both IRAK protein and mRNA (43,49). Neither NFB nor JNK is activated in IL-1-treated I1A cells, but these responses are restored in I1A-IRAK cells, indicating that IRAK is required for both (49). Mutant I3A, also deficient in both IL-1-induced NFB and JNK activation, is defective in a component upstream of IRAK but downstream of the IL-1 receptor (43,49). I1A and I3A mutant cells were used to examine whether these upstream signaling components of the IL-1 pathway are also required for TLR3mediated signaling. Transfection of TLR3 into the parental 293-TK/Zeo (293-TLR3) cells rendered them sensitive to poly(dI⅐dC)-induced NFB activation (49). Interestingly, poly(dI⅐dC) also induced NFB activation in I1A and I3A mutant cells transfected with TLR3 (I1A-TLR3 and I3A-TLR3), indicating that IRAK and the component deficient in I3A cells are not required for TLR3-mediated NFB activation (Fig. 1A). Although IRAK is necessary for IL-1-induced JNK, MAP kinases p42/44, and p38 activation (49), poly(dI⅐dC) can efficiently induce the activation of these kinases in I1A-TLR3 (Fig.  1, B and C), showing that IRAK is not required for TLR3mediated JNK, MAP kinases, p42/44, and p38 activation. Moreover, poly(dI⅐dC) induced IL-8 gene expression in both 293-TLR3 and I1A-TLR3 cells, indicating that TLR3-mediated gene expression is also IRAK-independent (Fig. 1D). Although the above described results clearly show that IRAK is not necessary for TLR3-mediated signaling, it is important to examine whether IRAK is at all involved in the pathway. Whereas IRAK is hyperphosphorylated and ubiquitinated upon IL-1 stimulation (49), IRAK is not modified upon poly(dI⅐dC) TLR3-mediated Signaling stimulation, indicating that IRAK does not participate in the TLR3-mediated dsRNA signaling pathway (Fig. 1E).
TRAF6, TAK1, and TAB2 Are Recruited to TLR3 upon Poly(dI⅐dC) Stimulation-To investigate the molecular mechanism by which TLR3 mediates the activation of NFB and MAP kinases, we first identified the signaling components that are recruited to TLR3 upon poly(dI⅐dC) stimulation. Cell extracts from 293-TLR3 cells, stimulated with poly(dI⅐dC) or left untreated, were immunoprecipitated with monoclonal anti-Flag (M2) antibody to pull down the Flag-tagged TLR3, followed by Western analyses with antibodies against signaling components that are implicated in pathways mediated by IL-1/Toll receptors, including MyD88, Mal/TIRAP, IRAK4, IRAK, TRAF6, TAK1, and TAB2. Among these components, TRAF6, TAK1, and TAB2 form a complex with TLR3 upon poly(dI⅐dC) stimulation, indicating that TLR3 recruits these proteins to mediate signaling ( Fig. 2A and Table I). Although MyD88, IRAK4, and IRAK are recruited to IL-1R (Fig. 2B), these IL-1R-proximal components are not recruited to TLR3, suggesting that they are probably not utilized by TLR3 to mediate signaling. Because Mal/TIRAP has been shown to play a role in the TLR2-and TLR4-mediated pathway, we also examined whether this adapter molecule is recruited to TLR3 upon poly(dI⅐dC) stimulation. However, we fail to detect any interaction between TLR3 and Mal/TIRAP before or after poly(dI⅐dC) stimulation ( Fig. 2C and Table I). Taken together, the results show that TLR3 recruits TRAF6, TAK1, and TAB2 upon poly(dI⅐dC) stimulation but not MyD88, MAL/TIRAP, IRAK4, or IRAK, forming a unique receptor complex that differs from the other members of the IL-1/Toll receptor superfamily. This conclusion is consistent with the above described genetic experiments in IL-1-unresponsive mutants, in which IL-1R-proximal signaling components including IRAK are shown to be dispensable for TLR3-mediated signaling.
To confirm the interactions between TRAF6, TAK1, and TAB2 and TLR3, extracts from 293-TLR3 cells untreated or

TLR3-mediated Signaling
stimulated with poly(dI⅐dC) were immunoprecipitated with anti-TRAF6 followed by Western analyses with anti-Flag (M2), anti-TAK1, anti-TAB2, and anti-TRAF6. The results show that TRAF6, TAK1, and TAB2 indeed form a complex at TLR3 upon stimulation ( Fig. 2D and Table I).
TLR3-mediated Signaling Leads to TAK1 Activation-TAK1, a member of the MAP kinase kinase kinase family, and two proteins that bind to it, TAB1 and TAB2 (26,27), have recently been shown to play a critical role in the IL-1 pathway. TAK1 is activated upon IL-1 stimulation, which in turn leads to the activation of IKK complex and MKK6 by an unknown mechanism, resulting in NFB and MAP kinase activation (29). The fact that TAK1 is recruited to TLR3 upon poly(dI⅐dC) stimulation suggests that this kinase may play an important role in TLR3-mediated pathway. To investigate the function of TAK1 in poly(dI⅐dC)-induced TLR3-mediated NFB activation, we examined TAK1 kinase activity upon poly(dI⅐dC) stimulation. Whole cell extracts from 293-TLR3 cells untreated or stimulated with poly(dI⅐dC) were immunoprecipitated with anti-TAK1 antibody followed by an in vitro kinase assay using bacterially expressed MKK6 as a substrate. TAK1 is highly activated in both wild type (293-TLR3) and IRAK-deficient (I1A-TLR3) cells upon poly(dI⅐dC) treatment (Fig. 3A), indicating that TAK1 is likely to be involved in this IRAK-independent TLR3-mediated signaling. In support of this conclusion, we found that a kinase-inactive TAK1 mutant (DN-TAK1) can effectively inhibit TLR3-mediated NFB activation, suggesting that TAK1 is probably involved in this pathway (Fig. 3B). As expected, DN-IRAK failed to inhibit poly(dI⅐dC)-induced NFB activation, serving as a negative control for DN-TAK1.

TLR3-mediated Signaling
Poly(dI⅐dC) Induces the Translocation of TRAF6-TAK1-TAB2 from the Membrane to the Cytosol, Interaction with PKR, and Activation of TAK1 in the Cytosol-Previous studies showed that TAB2, associated with TAK1-TAB1, is localized exclusively on the membrane before stimulation and translocates to the cytosol upon IL-1 stimulation (25,27,46). To investigate the fate of TAB2 upon poly(dI⅐dC) stimulation, membrane (P-100) and cytosolic (S-100) fractions from 293-TLR3 cells, untreated or stimulated with poly(dI⅐dC) for different lengths of time, were examined by the Western procedure with anti-TAB2. The data demonstrate that TAB2 translocates from the membrane to the cytosol upon poly(dI⅐dC) stimulation (Fig. 4A). Actin was used as a cytosolic marker and a loading control.
We next examined where TAK1 is activated in response to poly(dI⅐dC) stimulation. We examined whether TAK1 is activated upon its recruitment to TLR3 upon poly(dI⅐dC) stimulation. Cell extracts from 293-TLR3 cells, stimulated with poly(dI⅐dC) or left untreated, were immunoprecipitated with monoclonal anti-Flag (M2) antibody to pull down the Flagtagged TLR3 followed by an in vitro kinase assay using MKK6 as a substrate. As shown in Fig. 2A, TAK1 is not activated when it is bound to the receptor complex. We then examined whether TAK1 is activated on the membrane or in the cytosol.
Membrane and cytosolic fractions prepared from 293-TLR3 cells were immunoprecipitated with anti-TAK1 followed by an in vitro kinase assay using MKK6 as a substrate. As shown in Fig. 4B, similar to how TAK1 is activated in IL-1-mediated signaling, the kinase activity of TAK1 is not activated on the membrane but only in the cytosol upon poly(dI⅐dC) stimulation. In contrast, TAK1 and TAB2 are also phosphorylated only in the cytosol upon poly(dI⅐dC) stimulation, whereas they are phosphorylated on the membrane in response to IL-1 (25), suggesting that a different mechanism is likely to be involved in activating TAK1 in response to poly(dI⅐dC) and to IL-1.
To identify the signaling components associated with TAK1, extracts from 293-TLR3 cells treated with poly(dI⅐dC) were immunoprecipitated with anti-TAK1 followed by Western analyses with antibodies against candidate proteins, including TRAF6, TAB1, TAB2, IRAK4, and PKR. TRAF6 and PKR form a poly(dI⅐dC)-induced complex with TAK1 and TAB2, suggesting the involvement of these components in TAK1 activation (Fig. 4C). We also found that a kinase-inactive PKR mutant (DN-PKR) can efficiently inhibit TLR3-mediated NFB activation, suggesting that PRK is probably involved in this pathway (Fig. 4D). DISCUSSION On the basis of the results presented above, we have provided a working model for TLR3-mediated signaling pathway (Fig. 5). Upon binding of dsRNA to TLR3, TRAF6, TAK1, and TAB2 are recruited to the receptor to form a complex. The TRAF6-TAK1-TAB2 complex then leaves the receptor and translocates to the cytosol to interact with PKR, where TAK1 is phosphorylated and activated. Activated TAK1 then leads to the phosphorylation and activation of IKK and MKK6, resulting in the activation of NFB and MAP kinase.
Although both IL-1R and TLR3 activate TAK1, the upstream signaling events that lead to their activation are quite different (Fig. 5). In this study, we have clearly shown that IRAK is not required for TLR3-mediated signaling. Instead, TLR3 recruits TRAF6-TAK1-TAB2 to form a receptor complex. One important question is how TRAF6-TAK1-TAB2 are recruited to TLR3 upon poly(dI⅐dC) stimulation. TRAF6 is often recruited to receptors through TRAF binding sites that reside on the receptors (50). However, such TRAF binding sites are absent in the cytoplasmic domain of TLR3. Therefore, an adapter molecule(s) is likely to be required to mediate the interaction of TRAF6-TAK1-TAB2 with the TIR domain of TLR3. Although MyD88 and Mal/TIRAP are the two logical candidates, several lines of evidence suggest that they are probably not involved in TLR3mediated signaling. First, poly(dI⅐dC) can still induce activation of NFB and MAP kinases in MyD88-deficient macrophages (17). Second, we failed to detect any interaction of MyD88 and Mal/TIRAP with TLR3 upon poly(dI⅐dC) stimulation. Third, although MyD88 is the adapter that recruits IRAK in the signaling pathways mediated by other IL-1/Toll receptors, both MyD88 and Mal/TIRAP fail to activate NFB in IRAK-null cells 2 (49), suggesting that both MyD88 and Mal/ TIRAP mediate downstream signaling through IRAK. Importantly, our results clearly show that IRAK is not involved in TLR3-mediated signaling, implying that MyD88 and Mal/ TIRAP may also not be required for this pathway. Finally, although TLR3 fails to induce the production of some cytokines in MyD88-deficient macrophages (17), it could very well be because of the involvement of a poly(dI⅐dC)-induced secondary pathway. Therefore, novel adapter molecules are probably required for TLR3 to recruit TRAF6-TAK1-TAB2. TLR3-mediated Signaling ent activation gene expression, and TRIF associates with TLR3, suggesting that TRIF is a likely candidate adapter to recruit TRAF6-TAK1-TAB2 to TLR3.
As mentioned above, we have proved definitively that IRAK is not at all involved in TLR3-mediated signaling. Now the question is whether any of the other IRAK family members are involved in this pathway. There are three more IRAK-like molecules, including IRAK2, IRAKM, and IRAK4. The expression of IRAK2 and IRAKM is extremely low or non-existent in 293-TLR3 cells where the TLR3-mediated signaling is intact, strongly suggesting that these two IRAKs are not required for the pathway. IRAK4 has been shown to be the kinase that phosphorylates IRAK upon IL-1 treatment, whereas IRAK is not phosphorylated upon poly(dI⅐dC) stimulation. Furthermore, IRAK4 fails to activate NFB in IRAK-null cells, confirming that IRAK4 functions through IRAK 2 (23). The fact that IRAK does not participate in TLR3-mediated signaling suggests that IRAK4 is probably not involved in this pathway either. Moreover, we could not detect any interaction of IRAK4 with TLR3 or with other signaling components upon poly(dI⅐dC) stimulation. Taken together, these results suggest that IRAK4 is probably not involved directly in TLR3-mediated signaling. Although it was shown that poly(dI⅐dC)-mediated cytokine production is reduced in IRAK4-deficient mouse embryonic fibroblasts (22), TLR3-mediated early signaling events have not been reported in IRAK4-deficient macrophages and mouse embryonic fibroblasts. The observed defect could be because of the effect of a poly(dI⅐dC)-induced secondary pathway.
PKR has been implicated in different stress-induced signaling pathways including dsRNA signaling to NFB activation (45,53), although the precise function of PKR in these signaling pathways remains controversial (53,54). PKR has also been shown to interact with Mal/TIRAP in TLR4-mediated signaling, although it is not required for LPS-induced interferon-␤ production (55). However, we did not detect any interaction of PKR with TLR3 (Table I). Instead, we detected signal-depend-PKR. WCE, whole cell extracts. IP, immunoprecipitation. D, PRK dominant negative mutant. Increasing amounts of DN-PKR (500 ng, 1 g, and 3 g) were co-transfected with E-selectin luciferase reporter construct (1 g) into 293-TLR3 cells. The transfected cells were either untreated or stimulated with poly(dI⅐dC) for 4 h followed by a luciferase reporter assay. Vector DNA (3 g) and E-selectin luciferase reporter construct (1 g) were transfected into 293-TLR3 cells as a control. Data are presented as the -fold induction of luciferase activity in the poly(dI⅐dC)-treated cells. Shown are the averages from three independent experiments.
FIG. 4. TAK1 is activated in the cytosol. A, poly(dI⅐dC)-induced translocation of TAB2. 293-TLR3 cells were untreated or stimulated with poly(dI⅐dC) (0.1 mg/ml) for the indicated times and fractionated into membrane (P-100) and cytosolic (S-100) fractions followed by Western analyses with anti-TAB2 and anti-actin. Whole cell extracts (WCE) from 293-TLR3 cells, untreated or stimulated with poly(dI⅐dC), were also examined by Western immunoblotting with anti-TAB2. B, TAK1 activation. 293-TLR3 cells were untreated or stimulated with poly(dI⅐dC) (0.1 mg/ml) for the indicated times and fractionated into membrane and cytosolic fractions followed by immunoprecipitation with anti-TAK1. The immunoprecipitates were subjected to in vitro kinase assay using His-MKK6 as a substrate and Western analyses with anti-TAB2 and anti-TAK1. C, TAK1 complex. Extracts of 293-TLR3 untreated or stimulated with poly(dI⅐dC) (0.1 mg/ml) for the indicated times were immunoprecipitated with anti-TAK1 followed by Western analyses with antibodies against TAK1, TAB2, TRAF6, and TLR3-mediated Signaling ent interaction of PKR with the TAK1 complex (TRAF6-TAK1-TAB2), providing evidence for the involvement of PKR in TLR3-mediated signaling. It is important to note that TAK1 is both phosphorylated and activated in response to poly(dI⅐dC) stimulation. Therefore, it is possible that PKR may contribute to the activation process of TAK1. Alternatively, PKR might function downstream of TAK1. The precise function of PKR in TLR3-mediated signaling will likely require the identification of all components of this pathway(s).