Research Article

Phosphorylation of innate immune adaptor proteins MAVS, STING, and TRIF induces IRF3 activation

See allHide authors and affiliations

Science  13 Mar 2015:
Vol. 347, Issue 6227, aaa2630
DOI: 10.1126/science.aaa2630

Innate immune receptor signaling, united

Innate immune receptors such as RIG-I, cGAS, and Toll-like receptors bind microbial fragments and alert the immune system to an infection. Each receptor type signals through a different adapter protein. These signals activate the protein kinase TBK1 and the transcription factor IRF3, which tells cells to secrete interferon proteins (IFNs) important for host defense. Liu et al. now report a common signaling mechanism used by all three types of innate immune receptor-adaptor protein pairs to activate IRF3 and generate IFNs. This is important because cells must regulate their IFN production carefully to avoid inflammation and autoimmunity.

Science, this issue 10.1126/science.aaa2630

Structured Abstract

INTRODUCTION

Sensing of pathogenic microbes and tissue damage by the innate immune system triggers immune cells to secrete cytokines that promote host defense. Viral RNA, cytosolic DNA, and the bacterial cell wall component lipopolysaccharide activate signaling cascades through a number of pattern recognition receptor (PRR)–adaptor protein pairs, including RIG-I–MAVS, cGAS-STING, and TLR3/4-TRIF (TLR3/4, Toll-like receptors 3 and 4). Activation of these signaling modules results in the production of type I interferons (IFNs), a family of cytokines that are essential for host protection. The adaptor proteins MAVS, STING, and TRIF each activate the downstream protein kinase TBK1, which then phosphorylates the transcription factor interferon regulatory factor 3 (IRF3), which drives type I IFN production. Although much progress has been made in our understanding of PRR and adaptor protein activation, the mechanism by which the adaptor proteins activate TBK1 and IRF3 remains unclear.

RATIONALE

Other signaling pathways besides the RIG-I–MAVS, cGAS-STING, and TLR3/4-TRIF pathways activate TBK1. However, IRF3 phosphorylation by TBK1 is observed only in the IFN-producing pathways that use MAVS, STING, or TRIF as the adaptor protein. The discrepant activation of TBK1 and IRF3 implies the existence of a kinase-substrate specification mechanism exclusive to the IFN-producing pathways. Specification of TBK1-mediated IRF3 activation is essential for the tight regulation of IFN production, which would otherwise lead to autoimmune diseases.

RESULTS

Using biochemical and mouse cell– and human cell–based assays, we found that both MAVS and STING interacted with IRF3 in a phosphorylation-dependent manner. We show that both MAVS and STING are phosphorylated in response to stimulation at their respective C-terminal consensus motif, pLxIS (p, hydrophilic residue; x, any residue; S, phosphorylation site). This phosphorylation event then recruits IRF3 to the active adaptor protein and is essential for IRF3 activation. Point mutations that impair the phosphorylation of MAVS or STING at their consensus motif abrogated IRF3 binding and subsequent IFN induction.

We found that MAVS is phosphorylated by the kinases TBK1 and IKK, whereas STING is phosphorylated by TBK1. Phosphorylated MAVS and STING subsequently bind to conserved, positively charged surfaces of IRF3, thereby recruiting IRF3 for its phosphorylation and activation by TBK1. Point mutations at IRF3’s positively charged surfaces abrogated IRF3 binding to MAVS and STING and subsequent IRF3 phosphorylation and activation. We further show that TRIF-mediated activation of IRF3 depends on TRIF phosphorylation at the pLxIS motif commonly found in MAVS, STING, and IRF3. These results reveal that phosphorylation of innate immune adaptor proteins is an essential and conserved mechanism that selectively recruits IRF3 to activate type I IFN production.

CONCLUSION

We uncovered a general mechanism of IRF3 activation by the innate immune adaptor proteins MAVS, STING, and TRIF, which functions in three distinct pattern recognition pathways. Following its activation, each adaptor protein recruits and activates downstream kinase TBK1, which phosphorylates the cognate upstream adaptor protein at a consensus motif. Phosphorylated MAVS, STING, or TRIF in turn recruits IRF3 through its conserved, positively charged phospho-binding domain, allowing IRF3 phosphorylation by TBK1. Phosphorylated IRF3 subsequently dissociates from the adaptor protein and dimerizes though the same phospho-binding domain before translocating into the nucleus to induce IFN. These results elucidate how IRF3 activation and IFN production are tightly controlled and explain why TBK1 is necessary but not sufficient to phosphorylate IRF3: Phosphorylation of IRF3 by TBK1 occurs only with the assistance of an adaptor protein such as MAVS, STING, or TRIF, which also must be phosphorylated.

Phosphorylation of innate immune adaptor proteins licenses IRF3 activation.

MAVS, STING, and TRIF—which are activated by viral RNA, cytosolic DNA, and bacterial lipopolysaccharide (LPS), respectively—activate the kinases IKK and TBK1. These kinases then phosphorylate the adaptor proteins, which in turn recruit IRF3, thereby licensing IRF3 for phosphorylation (P) by TBK1. Phosphorylated IRF3 dissociates from the adaptor proteins, dimerizes, and then enters the nucleus to induce IFNs.

Abstract

During virus infection, the adaptor proteins MAVS and STING transduce signals from the cytosolic nucleic acid sensors RIG-I and cGAS, respectively, to induce type I interferons (IFNs) and other antiviral molecules. Here we show that MAVS and STING harbor two conserved serine and threonine clusters that are phosphorylated by the kinases IKK and/or TBK1 in response to stimulation. Phosphorylated MAVS and STING then bind to a positively charged surface of interferon regulatory factor 3 (IRF3) and thereby recruit IRF3 for its phosphorylation and activation by TBK1. We further show that TRIF, an adaptor protein in Toll-like receptor signaling, activates IRF3 through a similar phosphorylation-dependent mechanism. These results reveal that phosphorylation of innate adaptor proteins is an essential and conserved mechanism that selectively recruits IRF3 to activate the type I IFN pathway.

The innate immune system employs germline-encoded pattern recognition receptors to detect common pathogenic molecular features (1). Cytosolic DNA and viral RNA are detected by cGAS and RIG-I–like receptors, respectively, to activate several convergent signaling pathways to produce type I interferons (IFNs) (26). After ligand binding, cGAS and RIG-I signal through respective adaptor proteins STING and MAVS to recruit the kinases IKK and TBK1, which then activate the transcription factors NF-κB and interferon regulatory factor 3 (IRF3), respectively. Recent studies on the RIG-I pathway have provided mechanistic insights into innate immune signaling. Specifically, activated RIG-I forms oligomers to convert MAVS into prion-like polymers, which then recruit ubiquitin E3 ligases TRAF2, TRAF5, and TRAF6 to synthesize polyubiquitin chains (710); in turn, these ubiquitin chains recruit and activate IKK and TBK1 to trigger IFN production (9, 11).

The critical role of TRAF2, TRAF5, and TRAF6 in MAVS downstream signaling closely resembles that of TRAF recruitment in other NF-κB activating pathways such as those emanating from interleukin-1β (IL-1β), tumor necrosis factor–α (TNF-α), T cell receptor, and CD40 (12). However, stimulation of cells with IL-1β or TNF-α activates only IKK and TBK1, but not IRF3 (fig. S1A, lanes 1 to 5) (13). In contrast, activation of the RIG-I–MAVS and cGAS-STING pathways through vesicular stomatitis virus (VSV) infection and herring testis DNA (HT-DNA) transfection, respectively, leads to activation of IKK and TBK1, as well as IRF3 (fig. S1A, lanes 6 to 8). Similarly, Toll-like receptors 3 and 4 (TLR3 and TLR4) signal through the adaptor protein TRIF to active TBK1 and IRF3 (1). However, other TLRs that do not signal through TRIF can activate TBK1 but are unable to activate IRF3. How MAVS, STING, and TRIF possess the ability to activate both TBK1 and IRF3 is unknown.

MAVS and IRF3 form a ubiquitination- and phosphorylation-dependent complex

To understand how MAVS activates IRF3, we used cell-free assays that recapitulate IRF3 and IκBα activation by different upstream activators (8, 11, 14). We have previously shown that a recombinant MAVS protein, MAVSΔTM (MAVS lacking the C-terminal transmembrane domain) (8), spontaneously forms prion-like fibers to activate IKK, TBK1, and IRF3 in HeLa cytosolic extracts (S100), which could be detected by immunoblotting with phospho-specific antibodies or by native gel electrophoresis that reveals IRF3 dimerization (Fig. 1A, top). However, recombinant TRAF6 protein led to the activation of IKK and TBK1, but not IRF3 (Fig. 1A, bottom), suggesting that TBK1 activation alone is insufficient to activate IRF3.

Fig. 1 IRF3 and MAVS form a ubiquitination- and phosphorylation-dependent complex.

(A) HeLa cell extracts were incubated with MAVSΔTM or TRAF6 in the presence of ATP at 30°C for the indicated time. IRF3 dimerization (which indicates activation) in this and other panels was analyzed by native gel electrophoresis followed by immunoblotting (IB). Protein phosphorylation was detected by immunoblotting with the indicated antibodies. (B) After incubation of MAVSΔTM or TRAF6 in MEF extracts, Flag-IRF3 or Flag-IκBα was added before immunoprecipitation (Flag-IP). Coprecipitated proteins were detected by immunoblotting. (C) Twelve hours after Flag-IRF3-S385A/S386A (2A) transfection, HEK 293T cells were infected with Sendai virus as indicated. Flag-IP was carried out in whole-cell lysates, and coprecipitated proteins were detected by immunoblotting. (D) Indicated enzymes (vOTU and phosphatase) were added during or after (indicated by asterisks) incubation of recombinant MAVSΔTM in HEK 293T extracts. GST-IRF3 2A was then added as indicated, followed by GST-pull down. MAVS-IRF3 interaction was examined by immunoblotting. The data presented in this and all subsequent figures were reproduced in at least two independent experiments.

Next, to examine potential interactions between downstream effectors (IRF3 or IκBα) and upstream activators (MAVSΔTM or TRAF6), HeLa S100 was incubated with MAVSΔTM or TRAF6 at 0°C (control) or 30°C. Flag-tagged IRF3 (Flag-IRF3) or IκBα (Flag-IκBα) was then added to the reaction mixtures for immunoprecipitation (IP) (Fig. 1B). After an in vitro assay with MAVSΔTM, a smear of protein bands in complex with Flag-IRF3 was detected with a MAVS antibody. TBK1 was also present in the MAVS-IRF3 complex (Fig. 1B, lane 3). In contrast, in cell extracts incubated with TRAF6, Flag-IRF3 was unable to pull down TBK1 (Fig. 1B, lane 9), suggesting that MAVS, but not TRAF6, induced interaction between IRF3 and TBK1.

To determine if IRF3 and MAVS form a complex in virus-infected cells, we expressed IRF3 Ser385→Ala385 (S385A)/S386A (15) (Flag-IRF3 2A), which is unable to form a homodimer and may therefore associate with TBK1 or MAVS more tightly, in human embryonic kidney 293T (HEK293T) cells. Infection of these cells with Sendai virus led to the association of Flag-IRF3 2A with endogenous MAVS, TRAF2, TRAF6, and TBK1 (Fig. 1C, lane 12). These results suggest that MAVS may serve as a scaffold to bring IRF3 and TBK1 into proximity, thereby facilitating IRF3 phosphorylation by TBK1.

Because multiple E3 ubiquitin ligases are involved in MAVS downstream signaling (9), we next examined the role of ubiquitination in MAVS-IRF3 interaction. A deubiquitination enzyme containing the ovarian tumor type domain of the Crimean Congo hemorrhagic fever virus (vOTU) completely blocked MAVS-IRF3 interaction when it was added to the cell-free reaction (Fig. 1D, lane 3). However, vOTU no longer affected MAVS-IRF3 complex formation when added after the reaction, suggesting that ubiquitination is required only for initiating but not maintaining MAVS-IRF3 interaction (Fig. 1D, lane 4). In contrast, treatment with lambda protein phosphatase after the reaction abolished MAVS-IRF3 interaction, suggesting that MAVS-IRF3 interaction is dependent on phosphorylation (Fig. 1D, lane 5). Consistently, MAVS bound to IRF3 appeared on SDS–polyacrylamide gel electrophoresis (PAGE) as a slower-migrating smear, which was sensitive to phosphatase but not vOTU treatment (fig. S1B). These results suggest that the initial ubiquitination and subsequent phosphorylation on MAVS are necessary for MAVS-IRF3 interaction.

MAVS serine-rich clusters containing Ser442 are essential for IRF3 binding and activation

To map the MAVS phosphorylation site(s) essential for downstream signaling, we tested a panel of MAVS truncation mutants for their ability to bind and activate IRF3 (fig. S1C). We have previously shown that whereas the MAVS polymerization domain CARD is essential, the middle proline-rich (residues 94 to 153) and C-terminal transmembrane (510 to 540) regions of MAVS are dispensable for recombinant MAVS to activate IRF3 in the cell-free assay (8). Through a series of deletion analyses, we found that a truncated MAVS harboring the first 130 amino acids (containing CARD) and a C-terminal 61–amino acid fragment (MAVS-N460 Δ131-398) was sufficient for both IRF3 binding and activation (fig. S1, C to E). Further mutagenesis revealed two MAVS C-terminal serine/threonine clusters, S426/S430/S433 (3S) and S442/S444/T445/S446 (4T/S), to be essential for IRF3 activation (fig. S1, F to H, and fig. S2, A and B). An unbiased structure-guided alignment of full-length IRF3, MAVS, and STING across species revealed sequence similarity among the MAVS 4T/S cluster, STING C terminus and the IRF3 5T/S cluster (396 to 405), whose phosphorylation is known to be essential for IRF3 dimerization (Fig. 2, A and B). Specifically, not only are MAVS S442/S444 aligned with IRF3 S396/S398, the essential IRF3 phosphorylation sites, but the charged and hydrophobic residues surrounding MAVS 4T/S (such as D438, L439, and I441) also aligned well with those around IRF3 S396/S398, suggesting that these two regions take on a similar structural fold by sharing a DLxIS (where x is any amino acid) consensus motif. Mutagenesis of the hydrophobic resides L439 and I441 to Asp or Ala abolished the ability of MAVS to activate IRF3 (fig. S2C). Furthermore, the location of MAVS S426/S430/S433 (3S) relative to MAVS S442 resembles the position of IRF3 S385/S386 with respect to IRF3 S396 (fig. S2A); phosphorylation of both IRF3 serine clusters is essential for its dimerization and activation (16, 17).

Fig. 2 MAVS C terminus harbors a conserved serine-rich region containing Ser442 essential for IRF3 binding and activation.

(A) A structure-guided cross-species sequence alignment of full-length IRF3, STING, and MAVS using PROMALS3D (32) revealed a cLxIS (c, charged residues; x, any amino acid) consensus motif in the C-terminal regions. Conservation index score: 9 is the highest, ≥5 is significant (46). IRF3 structures were used to assist alignment by the software. (B) Diagrams of MAVS domains and its 4T/S site compared to IRF3 5T/S site. TM, transmembrane domain. (C) Recombinant MAVS WT and point mutants at the 4T/S site were tested for their ability to activate IRF3 in the cell-free assay. MAVS protein level was analyzed by immunoblotting. (D and E) Mavs−/− MEFs reconstituted with MAVS WT or 4T/S mutants were infected with VSV for the indicated time. IRF3 dimerization, IκBα phosphorylation, and protein expression were analyzed by immunoblotting. IFN-β mRNA levels were measured by q-RT-PCR. Error bars in this and other figures represent SDs of triplicates. (F) HA-tagged IRF3 2A was further stably expressed in the reconstituted cells described in (D) and (E). After VSV infection, HA-IP was carried out using a HA antibody to examine the interaction between IRF3 2A and the MAVS proteins.

To examine whether phosphorylation of the MAVS 4T/S cluster is important for IRF3 activation, we mutated the residues to alanine (4T/S→4A) or to aspartic acid (4T/S→4D). Recombinant MAVS 4T/S→4A with the transmembrane domain deletion (MAVS-N460 4T/S→4A) failed to bind or activate IRF3 (fig. S2, B and D). MAVS-N460 4T/S→4D, however, resulted in weak IRF3 interaction and activation, probably due to incomplete phosphate mimic by aspartic acid, analogous to the loss of function of IRF3 S385D/S386D (18, 19). Similar to the MAVS wild type (WT), the MAVS 4A mutant still triggered phosphorylation of TBK1 and IκBα, and it also directly interacted with TRAF2 and TRAF6 (fig. S2, B and E), indicating that the observed defect in IRF3 binding and activation is specific and not due to misfolding of MAVS.

Subsequent mutation of each T/S residue revealed that MAVS S422A alone abolished IRF3 dimerization (Fig. 2C). To examine the possibility that S442 may be important structurally (e.g., for hydrogen bonding or polarity), we mutated S442 into other amino acids. Mutating S442 into cysteine (S442C) or asparagine (S442N), which retains serine’s hydrogen bonding ability or polarity but can no longer be phosphorylated, abolished MAVS’ ability to activate IRF3 (fig. S2F). In contrast, serine-to-threonine (S442T) or serine–to–glutamic acid (S442E) mutants largely retained the activity, suggesting that MAVS S442 phosphorylation is essential for IRF3 activation in vitro.

To investigate the role of MAVS S442 during virus infection, we reconstituted MAVS WT or mutants into Mavs−/− mouse embryonic fibroblasts (MEFs). Mavs−/− MEFs expressing MAVS S442A failed to activate IRF3 or produce IFN-α or -β in response to VSV, whereas IκBα phosphorylation was unaffected (Fig. 2, D and E, and fig. S3A). Virus infection, however, induced comparable TBK1 phosphorylation at serine 172 in both MAVS WT and S442A-expressing cells, indicative of normal TBK1 activation (fig. S3B). Point mutation of S442 to Cys or Asn, but not Thr or Asp, abolished MAVS’ ability to induce IRF3 dimerization in cells upon virus infection, suggesting that MAVS S442 is essential due to its phosphorylation, but not the serine’s other structural roles (fig. S3, D to F). Both MAVS 4A and S442A mutations also diminished the ability of MAVS to interact with IRF3 2A in response to VSV infection (Fig. 2F and fig S3C). Altogether, these results indicate that MAVS phosphorylation at S442 is critical for IRF3 activation but dispensable for TBK1 and IKK activation.

In contrast to the S442 point mutants, individual mutation of the other three Ser/Thr at the 4T/S site (S444A, T445A, or S446A) did not impair IRF3 dimerization or IFN-α or -β induction (Fig. 2, D and E, and fig. S3A). However, simultaneous mutations of these residues (S444A/T445A/S446A) strongly diminished IRF3 dimerization and IFN-α or -β induction, indicating that residues in the 4T/S site function cooperatively in IRF3 activation.

To examine the role of the first serine cluster (3S) of MAVS in IRF3 activation, we reconstituted Mavs−/− MEFs with MAVS harboring simultaneous (S426A/S430A/S433A, or 3S→3A) or single mutations. MAVS 3S→3A, but not individual mutants, abolished virus induction of IFN-β (fig. S3G). Additionally, similar to MAVS 4T/S→4A, the MAVS 3S→3A mutation also completely abolished MAVS’ ability to interact with IRF3 upon VSV infection (fig. S3C). Hence, the MAVS 3S site, in addition to the 4T/S site, is also important for virus-induced IRF3 activation. This pattern of two key serine/threonine clusters on MAVS is reminiscent of those on IRF3, where phosphorylation at S385 and S386, in addition to S396/S398-containing 5T/S phosphorylation, is required for IRF3 dimerization.

MAVS 4T/S site, including Ser442, is phosphorylated redundantly by TBK1 and IKK

To monitor Ser442 phosphorylation, we generated a polyclonal antibody specific for human MAVS phospho-serine 442 (p-S442). After IRF3 activation in the cell-free assay, the MAVS p-S442 antibody recognized only MAVS-N460 WT, but not S442A (fig. S4A). MAVS S442 phosphorylation was not observed in assays with MAVS E26A, a polymerization-defective mutant unable to activate downstream kinases (fig. S4B, lane 3) (9). MAVS was also not phosphorylated in cell extracts treated with vOTU or deficient in TRAF2/5 or NEMO, which were unable to support IRF3 activation (fig. S4B, lanes 2, 6, and 7). However, S442-phosphorylated MAVS was strongly enriched in the IRF3 immunoprecipitates after the cell-free assay (fig. S4A, lane 3, top). Hence, MAVS S442 phosphorylation depends on MAVS polymerization and TRAF-mediated polyubiquitin synthesis and correlates with MAVS-IRF3 complex formation.

We then examined whether MAVS S442 phosphorylation could be detected in virus-infected cells. The p-S442 antibody was unable to detect a clear MAVS signal in whole-cell lysates before or after Sendai virus infection. However, IP with the p-S442 antibody revealed a MAVS smear that appeared only after virus infection, indicating that MAVS S442 phosphorylation was induced by the virus (Fig. 3A). Using targeted quantification by mass spectrometry, we observed a robust induction of phosphorylated peptides containing the MAVS 4T/S site, including S442 after virus infection (Fig. 3B and fig. S4, D to F; see also Materials and methods section on Mass spectrometry). Our mass spectrometry results also revealed that MAVS became phosphorylated at many other Ser/Thr residues after virus infection (table S2), which may explain its formation of a smear on SDS-PAGE (Fig. 3A, lane 4).

Fig. 3 The MAVS 4T/S cluster including Ser442 is redundantly phosphorylated by TBK1 and IKK.

(A) U2OS cells were infected with Sendai virus (SeV) for 12 hours, then whole-cell extracts were immunoprecipitated with a MAVS p-S442 specific antibody under denaturing conditions. The whole-cell extracts and IP products were then immunoblotted with a MAVS antibody. IRF3 dimerization was monitored by native gel electrophoresis and immunoblotting. (B) Targeted MS2 quantification of the IS442ASTSLGR peptide containing MAVS 4T/S site before or after SeV infection. The relative abundance of nonphosphorylated 4T/S (4T/S), singly phosphorylated 4T/S (p-4T/S), or phosphorylated S442 (p-S442) peptides were selectively monitored using ions specific for each species (also see table S1 and Materials and methods). The intensity ratio (in red) = MA+SeV/MA-SeV (MA, mass area). (C and D) Similar to (A) and (B), except that the cells were treated with a TBK1 inhibitor (BX795), an IKK inhibitor (TPCA-1), or both 1 hour before virus infection. In (D), the relative intensity of each peak is shown in red (normalized with peptide MA of DMSO-treated samples). (E) Recombinant TBK1 or IKKβ was incubated with purified MAVS N460 in the presence of ATP at 30°C. The reaction products were analyzed by immunoblotting with an antibody specific for p-S442 MAVS or total MAVS.

Virus infection recruits TBK1 to MAVS. However, in Tbk1−/− extracts, MAVS S442 phosphorylation and interaction with IRF3 were unaffected, whereas Ikkα/Ikkβ-deficient extracts showed reduced IRF3 activation and binding to MAVS (fig. S4B and fig. S5A). MAVS S442 phosphorylation in Tbk1−/− extracts was completely blocked by an IKK inhibitor, TPCA-1, but not by a TBK1 inhibitor BX795 (fig. S5B). Subsequent analysis indicated that the combination of IKK and TBK1 inhibitors, but not each alone, blocked MAVS-IRF3 interaction and MAVS S442 phosphorylation in WT cell extracts (fig. S5, C and D).

In U2OS cells infected with Sendai virus, the combination of IKK and TBK1 inhibitors, but not each alone, abolished MAVS phosphorylation, as revealed by IP with the p-S442 antibody (Fig. 3C). Mass spectrometry analysis indicated that the abundance of singly phosphorylated MAVS 4T/S peptide (including p-S442) was severely diminished only when both inhibitors were added to virus-infected HEK 293T cells (Fig. 3D, and fig. S5E).

In vitro kinase assays revealed that recombinant TBK1 and IKKβ directly phosphorylated MAVS S442 (Fig. 3E). TBK1 also induced MAVS-IRF3 interaction, which was abolished by phosphatase treatments (fig. S5F). Altogether, these results indicate that TBK1 and IKK are capable of directly phosphorylating MAVS C terminus upon virus infection. Phosphorylated MAVS may then recruit IRF3 for its phosphorylation by TBK1 (see below).

IRF3 C terminus harbors positively charged surfaces important for MAVS-IRF3 interaction and IRF3 dimerization

To map the IRF3 region responsible for binding to phosphorylated MAVS, we made a series of recombinant IRF3 truncation mutants and tested their MAVS binding ability (fig. S6, A and B). In vitro IRF3 IP revealed that the entire IRF3 C terminus (190 to 427, IRF3-C), containing the IRF3 association domain and the serine-rich region, is necessary and sufficient for MAVS binding (fig. S6, A and B). IRF3-C crystal structures revealed a similar fold to the Mad homology 2 (MH2) domain of the SMAD family of proteins (20, 21). As a well-known phospho-binding domain, the MH2 domain contains positively charged surfaces composed of conserved basic residues that are important for both phosphorylated transforming growth factor–β (TGF-β) receptor binding and subsequent dimerization of the phosphorylated SMADs (22). Similarly, IRF3-C contains positively charged patches composed of conserved basic resides (Fig. 4A, and fig. S7), mutations of which were shown to abolish phosphorylation and dimerization of IRF3 in virus-infected cells (21).

Fig. 4 IRF3 positively charged surfaces and MAVS aggregation are important for the interaction between IRF3 and phosphorylated MAVS.

(A) Diagrams of IRF3 domains and conserved, positively charged residues. DBD, DNA binding domain; IAD, IRF3 association domain; SR, serine-rich region. (B) Purified Flag-IRF3 2A or 2A containing positively charged surface mutations was examined for the ability to form a complex with MAVS and TBK1 in cell extracts. (C) Irf3−/−Irf7−/− (IRF DKO) or Mavs−/− MEF cells stably expressing HA-tagged IRF3 WT, 2A, or 2A containing mutations at the positively charged surfaces were infected with VSV, then HA-IP was carried out to examine the interaction between MAVS and IRF3 WT or mutants. (D) Flag-IRF3 WT and 2A proteins were either incubated with active recombinant MAVS in HEK 293T extracts at 30°C (lanes 5 and 6) or added after the reaction and then incubated at 4°C (lanes 7 and 8; indicated by asterisks). Flag-IP was then carried out to examine IRF3-MAVS-TBK1 complex formation. IRF3 dimerization after the reactions was monitored by immunoblotting. (E) MAVS WT and CARD mutants were incubated with recombinant TBK1 in the presence of ATP followed by the addition of phosphatase (CIP) when indicated. Flag-IRF3 2A was then added followed by Flag-IP to examine the interaction between MAVS and IRF3. MAVS proteins and their phosphorylation at S442 after the reaction were analyzed by immunoblotting.

To determine whether the observed IRF3 phosphorylation defect was due to impaired MAVS binding, we tested IRF3 2A proteins containing positively charged surface mutations for their ability to bind MAVS. Compared to IRF3 2A, all of the positively charged surface mutants had reduced interaction with MAVS and TBK1 in our cell-free assay (Fig. 4B). Additionally, when reconstituted into Irf3−/−Irf7−/− MEF cells, only IRF3 2A but not 2A proteins containing positively charged surface mutations interacted with endogenous MAVS in response to VSV infection (Fig. 4C and fig. S6C). These results indicate that the positively charged surfaces on IRF3-C are important for two interactions: the binding to phosphorylated MAVS and to a second phosphorylated IRF3 molecule (i.e., dimerization of phosphorylated IRF3).

To uncouple the steps of MAVS-IRF3 binding and IRF3 dimerization, Flag-IRF3 WT or 2A (S385A/S386A, which cannot dimerize) was added to cell extracts with active MAVS either before the 30°C incubation, so that the reaction could go to completion, or afterward at 4°C where MAVS had been phosphorylated but IRF3 remained unphosphorylated (Fig. 4D). When added to the reaction at 30°C, IRF3 WT dimerized exclusively, whereas IRF3 2A stably interacted with MAVS (Fig. 4D, lane 5 and 6). However, when added at 4°C after the 30°C assay, both IRF3 WT and 2A interacted with MAVS comparably (lanes 7 and 8), suggesting that IRF3 first binds to phosphorylated MAVS (lane 7) before dissociating to form a dimer (lane 5). To further validate this model, Flag-IRF3 WT or 2A was added to Tbk1−/− cell extracts before the 30°C incubation, in which MAVS was still phosphorylated at S442 by IKK but IRF3 could not be phosphorylated (fig. S4B and fig. S6D, bottom). Here, unphosphorylated IRF3 exclusively interacted with MAVS without forming a dimer (fig. S6D, lane 2). This is in contrast to IRF3 WT in WT cell extracts, which was phosphorylated and formed a dimer (fig. S6D, lane 1). The phosphorylation-defective IRF3 2A mutant, however, exclusively interacted with MAVS (fig. S6D, lane 3). Hence, IRF3 phosphorylation causes its dissociation from MAVS. Consistently, compared with IRF3 2A, IRF3 WT pulled down less MAVS in response to VSV infection from cells, suggesting that the majority of IRF3 WT proteins had dissociated from MAVS and undergone phosphorylation-induced dimerization (Fig. 4C, lanes 2 and 7). Taken together, these results indicate that IRF3 directly binds to phosphorylated MAVS through conserved, positively charged surfaces within the IRF3-C domain, which also mediate interaction with another phosphorylated IRF3 monomer after its dissociation from MAVS.

Both MAVS polymerization and phosphorylation are required for MAVS-IRF3 interaction

In vitro kinase assay revealed that recombinant MAVS 361-460 could be directly phosphorylated by recombinant TBK1 and IKK, but the phosphorylated fragment failed to interact with IRF3, suggesting that phosphorylation alone is insufficient for IRF3 recruitment. Given that recombinant MAVS containing CARD and a C-terminal fragment (399 to 460) interacted with IRF3 in the cell-free assay (fig. S1E), we tested whether MAVS-CARD–mediated polymerization is important for IRF3 binding. Active recombinant TBK1 induced S442 phosphorylation in both MAVS-N460 WT and polymerization-defective mutants (Fig. 4E). However, only MAVS WT, but not the CARD polymerization mutants, interacted with IRF3 in a MAVS phosphorylation–dependent manner (Fig. 4E). These results indicate that MAVS-CARD polymerization is not only important for TRAF and kinase recruitment (9) but also required for subsequent IRF3 binding and activation.

STING phosphorylation at Ser366 is required for IRF3 binding and activation in the DNA-sensing pathway

STING is an essential adaptor protein in the cytosolic DNA sensing pathway that also activates IRF3 (23). Recently, STING Ser366 was shown to be important in IRF3 binding and activation because a mutation to alanine (S366A) abolished DNA induced IRF3 activation (24). In contrast, another study suggested that Ser366 phosphorylation by ULK1 negatively regulates STING because an aspartic acid (S366D) mutation renders the protein inactive (25). Thus, the mechanism of STING-mediated IRF3 activation and the role of Ser366 need to be clarified.

Structure-guided sequence alignment of IRF3, MAVS, and STING revealed that STING S366 is positioned within the cLxIS (c, charged residue; x, any residue) consensus motif that is also found around MAVS S442 and IRF3 S396 (Figs. 2A and 5A). Moreover, STING also harbors another serine cluster (S353/S358 in human STING; S354/S357 in mouse STING) upstream of S366, positioned similarly to IRF3 S385/S386 and MAVS 3S with respect to IRF3 S396 and MAVS S442 (fig. S8A). These observations suggest that STING and MAVS may activate IRF3 through similar mechanisms. Consistently, L929 cells in which endogenous STING was depleted by short hairpin RNA (shRNA) and replaced with mouse STING (mSTING) S365A (corresponding to hSTING S366A) failed to activate IRF3 after HT-DNA transfection (fig. S8, B and C). mSTING S357A (corresponding to hSTING S358A) also had diminished ability to activate IRF3, whereas mutations of other Ser or Thr residues had little effect. mSTING S365A and S357A mutations did not affect DNA-induced activation of TBK1 and IKK (fig. S8C). Contrary to a previous report (25), the S365A mutation of mSTING also did not inhibit DNA-induced degradation of STING (fig. S8D). When reconstituted in Sting−/− Raw264.7 macrophages or L929 cells depleted of endogenous STING, hSTING S366A, S366C, or S366N failed to activate IRF3 in response to DNA, whereas S366D retained weak activity (Fig. 5B and fig. S8E). These results suggest that, like MAVS S442, STING S366 is important because of its phosphorylation rather than other structural roles. Similar to the MAVS 4T/S→4D and IRF3 S385D/S386D mutants, the weak activity of S366D may be due to incomplete phosphate mimic by aspartic acid.

Fig. 5 STING Ser366 phosphorylation by TBK1 recruits IRF3.

(A) Diagrams of the STING domains and its Ser366-containing motif compared with the IRF3 C-terminal 5T/S region. (B) Sting−/− Raw264.7 macrophages were reconstituted with C-terminal Flag-tagged human STING WT and S366 mutants. Three hours after HT-DNA transfection, IRF3 dimerization and STING expression were analyzed by immunoblotting. (C) Targeted MS2 quantification of the LLIS365GMDQPLPLR peptide containing mouse STING Ser365 from DNA-stimulated (+HT-DNA), unstimulated (-HT-DNA), or DNA-stimulated cells treated with the TBK1 inhibitor (+HT-DNA+BX795). Phosphorylated (p-S365) or nonphosphorylated peptides (S365) were calculated by selectively using ions specific for each species (also see table S1 and Materials and methods). (D) L929 cells depleted of endogenous STING were reconstituted with hSTING WT or S366A. STING S366 phosphorylation after HT-DNA transfection was detected with the p-S366 antibody (fig. S8E). (E) S366 phosphorylation of endogenous STING in THP-1 cells after HT-DNA transfection was examined by immunoblotting with the p-S366 antibody. (F) WT or Tbk1−/− HEK293T cells stably expressing hSTING were stimulated by cytosolic delivery of cGAMP, then IRF3 dimerization and STING phosphorylation at Ser366 were analyzed by immunoblotting. (G) HA-tagged IRF3 WT, 2A, or 2A proteins containing mutations at the positively charged surfaces were stably expressed in the L929 STING-reconstituted cells, as described in (D). After HT-DNA transfection, HA-IP was carried out to examine the interaction between IRF3 and phosphorylated STING. (H) L929 cells stably expressing hSTING-Flag and HA-IRF3 2A as described in (G) were transfected with HT-DNA for 2 hours or mock transfected. Interaction between IRF3 2A and p-S366-STING was visualized under confocal microscopy, using PLA with an anti-HA monoclonal antibody for IRF3 and an anti–p-S366 rabbit polyclonal antibody (top). The cell in the top right was further subjected to a z-stack collection, which obtains and combines multiple images at different focal distances along the z axis (bottom). The image was shown as three-dimensional, maximum projection mode in the Zen software. No PLA-positive cells were observed without DNA transfection, and ~50% of the HT-DNA–stimulated cells are PLA-positive (red). Blue, DAPI; green, phalloidin labeled actin filaments. The images are representative of >100 cells examined.

Quantitative mass spectrometry analysis confirmed that phosphorylation of mSTING S365 was induced by more than 200-fold after HT-DNA transfection and that this phosphorylation was abolished by the TBK1 inhibitor BX795 (Fig. 5C and fig. S8F; see also Materials and methods section on Mass spectrometry). We also generated a rabbit polyclonal antibody specific for S366-phosphorylated hSTING (p-S366); this antibody detected phosphorylated hSTING in HT-DNA–transfected cells but not in unstimulated cells (Fig. 5, D and E). The S366A hSTING mutant was not detected by the antibody. Additionally, the hSTING p-S366 antibody detected a perinuclear punctate structure only in DNA-stimulated cells (fig. S8G, top), similar to the previously observed STING foci (fig. S8G, bottom) (26). To rule out any nonspecific staining by the phospho-specific antibody, we used proximity ligation assay (PLA) in L929-hSTING-Flag cells (Fig. S8H), where positive signals (red immunofluorescence) occur only when the hSTING p-S366 antibody is in close proximity to the Flag-antibody. We observed peri-nuclear signals only in cells after DNA transfection but not after poly(I:C) or mock transfection, indicating that our antibody specifically recognizes hSTING p-S366. After reconstitution of hSTING into WT and Tbk1−/− HEK293T cells (which lack endogenous STING), cGAMP stimulation induced STING p-S366 and IRF3 dimerization in WT but not Tbk1−/− cells, suggesting that TBK1 is essential for STING S366 phosphorylation in cells (Fig. 5F). These results indicate that STING undergoes TBK1-dependent phosphorylation at Ser366 only in DNA-stimulated cells.

We have previously shown that a STING C-terminal fragment (residues 281 to 379) could activate IRF3 in cytosolic cell extracts (24). This STING fragment was phosphorylated at S366 in the cell extracts, as revealed by immunoblotting with the p-S366 antibody (fig. S9A). Phosphorylation of STING depended on TBK1 but not on NEMO (fig. S9A). To test whether STING phosphorylation by TBK1 is important for IRF3 association with TBK1, we incubated recombinant TBK1 with WT or S366A mutant STING (281 to 379) and then added Flag-IRF3 for IP (fig. S9B). WT but not S366A STING formed a complex with IRF3 and TBK1; the formation of this complex was abolished by phosphatase treatment (fig. S9B, lanes 5, 6, and 11). Similar results were obtained after incubating the recombinant STING (281 to 379) with crude cell extracts (fig. S9C). These results demonstrate that TBK1 directly phosphorylates STING at Ser366 and phosphorylated STING recruits IRF3, thereby facilitating IRF3 phosphorylation by TBK1.

We next examined whether full-length STING interacts with IRF3 in response to DNA stimulation in cells. Upon DNA transfection, IRF3 2A interacted with hSTING WT but not S366A, suggesting that hSTING S366 is essential for DNA-induced interaction between STING and IRF3 (Fig. 5G, lane 2 and 3). Additionally, IRF3 WT and IRF3 2A, but not IRF3 2A bearing positively charged surface mutations, interacted with S366-phosphorylated STING, suggesting that the phospho-binding domain of IRF3 is crucial for binding to phosphorylated STING. Notably, compared with IRF3 2A, IRF3 WT pulled down substantially less STING. This suggests that analogous to MAVS-IRF3 interaction, phosphorylated IRF3 dissociates from STING and forms a dimer (Fig. 5G, lane 8).

Using the PLA assay, we examined the interaction between phosphorylated STING and IRF3 in L929 cells in which endogenous STING was depleted by shRNA and replaced by hSTING. These cells also stably expressed HA-IRF3 2A. The PLA assay using the HA antibody and the p-S366 STING antibody revealed positive signals (red immunofluorescent dots) in the peri-nuclear regions only after DNA stimulation (Fig. 5H; also see Materials and methods). In contrast, no PLA-positive signals were observed in DNA-stimulated cells expressing STING S366A or IRF3 2A R211/R213A (a positively charged surface mutant) (fig. S9D). Altogether, these results indicate that analogous to the role of MAVS S442 phosphorylation, STING phosphorylation at S366 is critical for direct IRF3 recruitment and activation through IRF3’s phospho-binding domain.

Phosphorylation of TRIF recruits and activates IRF3

Besides the cytosolic nucleic acid–sensing pathways, stimulation of certain TLRs (namely TLR3 and TLR4) also activates IRF3 and induces type I IFNs. TLR3 and TLR4 can each activate the adaptor protein TRIF, which in turn activates IRF3 (27). TRIF contains an N-terminal domain that includes the TIR domain important for interaction with TLRs and a C-terminal RIP homotypic interaction motif, which can activate cell death pathways (fig. S10A). The N-terminal fragment containing ~540 amino acids of TRIF (TRIF-N540) has been shown to bind IRF3 and activate the interferon promoter when transiently expressed in HEK 293T cells (28, 29). To avoid triggering cell death, we chose TRIF-N540 to investigate the mechanism by which IRF3 is activated in the TLR3/4 pathways.

To examine whether TRIF interacts with IRF3 in a phosphorylation-dependent manner, we transiently expressed Flag-tagged IRF3 2A and TRIF-N540 in HEK293T cells. IP of IRF3 revealed an interaction between IRF3 2A and TRIF-N540 that was greatly diminished after pretreatment of cells with the TBK1 inhibitor BX795, suggesting that TBK1 is important for inducing TRIF-IRF3 interaction (Fig. 6A, lane 2 and 3). Moreover, phosphatase treatment of cell lysates using calf intestinal phosphatase (CIP) or Lambda phosphatase (Lambda PP) before Flag-IRF3 IP completely abolished TRIF binding to IRF3-2A, suggesting that a phosphorylation event induced the TRIF-IRF3 interaction (Fig. 6A, lanes 4 and 5). IRF3 WT and IRF3 2A containing mutations in the positively charged surface (described above) all failed to pull down TRIF-N540 (Fig. 6B). Collectively, these data indicate that, similar to MAVS and STING, TRIF also recruits IRF3 through a mechanism that depends on the kinase TBK1, its phosphorylation, and the phospho-binding domain of IRF3.

Fig. 6 TRIF recruits and activates IRF3 through phosphorylation at the pLxIS motif.

(A) TRIF-N540 and Flag-tagged IRF3 2A were transiently expressed in HEK293T cells or cells that were treated with or without the TBK1 inhibitor BX795. Cell extracts were treated with different phosphatases, as indicated, before they were subjected to Flag-IP to examine the interaction between TRIF and IRF3 2A. (B) Similar to (A), except that IRF3 WT or IRF3 2A proteins containing mutations at the positively charged surfaces were examined for their ability to bind TRIF. (C) A cross-species sequence alignment between TRIF, IRF3, MAVS, and STING showing the consensus phosphorylation motif in TRIF. (D) Similar to A, except that different TRIF mutants were examined for their ability to bind IRF3 2A. (E) Similar to (A), except that Flag-tagged TRIF WT or mutant, HA-IRF3 2A, and Flag-TBK1 WT or kinase-dead mutant (Mut) were transiently expressed in HEK293T cells. HA-IP was carried out to examine the TRIF-IRF3 interaction. (F and G) After transient expression of TRIF mutants in HEK293T cells, IRF3 dimerization, TBK1 phosphorylation, and TRIF expression were analyzed by immunoblotting.

Through examination of the TRIF-N540 sequence, we found a conserved Ser/Thr cluster (S210/S212/T214) whose surrounding sequence pLEIS (p, hydrophilic amino acid) shares marked similarity to the consensus cLxIS motif (c, charged; x, any) found in IRF3, MAVS, and STING (Fig. 6C). TRIF S210 is positioned similarly to IRF3 S396, MAVS S442, and STING S366 (Fig. 6C). On the basis of this sequence analysis, we expressed a panel of TRIF mutants, including S210A/S212A/T214A (TRIF 3A), in HEK293T cells and tested their ability to interact with IRF3 2A. Only TRIF 3A, but not other TRIF mutants, failed to interact with IRF3 2A (Fig. 6D). These results suggest that the TRIF S210/S212/T214 cluster probably functions similarly to MAVS 4T/S and STING S366 in IRF3 binding and activation.

Quantitative mass spectrometry analysis identified phosphorylated peptides containing S210 in TRIF protein that was transiently expressed in HEK293T cells (fig. S10B). The intensity of phosphorylated peptides containing S210 increased by >11-fold after TBK1 coexpression (fig. S10B), suggesting that TBK1 mediates TRIF phosphorylation at the S210/S212/T214 cluster. Overexpression of WT, but not kinase-dead mutant of TBK1 (TBK1 Mut), induced a robust and phosphatase-sensitive interaction between TRIF-N540 and HA-IRF3 2A (Fig. 6E, lanes 3 to 6). In contrast, TBK1 failed to induce interaction between TRIF 3A and HA-IRF3 2A, suggesting that TRIF phosphorylation at S210/S212/T214 by TBK1 probably induced the interaction between TRIF and IRF3 (Fig. 6E, lane 7). Taken together, these data indicate that TRIF is phosphorylated at its consensus motif S210/S212/T214 by the kinase TBK1, which leads to IRF3 recruitment.

When transiently expressed in HEK293T cells, only TRIF 3A, but not other TRIF mutants, failed to induce IRF3 dimerization (Fig. 6F). TRIF 3A led to normal TBK1 phosphorylation at S172, indicating that the observed IRF3 dimerization defect is specific and not due to misfolding of TRIF (Fig. 6F). Moreover, mutagenesis of S210 to alanine alone in TRIF-N540 or full-length TRIF abolished its ability to induce IRF3 dimerization, suggesting that, like MAVS S442 and STING S366, TRIF S210 is the critical phosphorylation site that mediates IRF3 activation (Fig. 6G). Altogether, our data indicate that TRIF binds and activates IRF3 through a phosphorylation-dependent mechanism that is similar to that of MAVS and STING.

Discussion

Through in vitro reconstitution and cell-based assays, we reveal that phosphorylation of a consensus motif in MAVS, STING, and TRIF is essential for IRF3 recruitment and subsequent phosphorylation by TBK1. Based on previously published results and those presented here, we propose the following model for MAVS/STING/TRIF-mediated IRF3 activation (Fig. 7): (i) Adaptor protein activation: after ligand binding, RIG-I, cGAS, and TLR3/4 activate downstream adaptor proteins MAVS, STING, and TRIF, respectively. (ii) Kinase activation: active MAVS polymers or TRIF proteins recruit TRAF family ubiquitin E3 ligases (i.e., TRAF2, 5 and 6) to synthesize polyubiquitin chains to activate IKK and TBK1, whereas active STING directly recruits and activates TBK1. (iii) Adaptor protein phosphorylation: the recruited kinases then phosphorylate MAVS, STING, or TRIF at their conserved pLxIS motif. (iv) IRF3 recruitment: IRF3 binds to phosphorylated MAVS, STING, or TRIF through IRF3’s conserved, positively charged surface. (v) IRF3 phosphorylation: IRF3 is efficiently phosphorylated by TBK1 once they are in close proximity. (vi) IRF3 self-association: phosphorylated IRF3 dissociates from the adaptors and dimerizes though the same positively charged surfaces. IRF3 dimer then enters the nucleus, where it functions together with NF-κB to turn on type I interferons and other cytokines.

Fig. 7 A model of kinase-substrate specification by phosphorylated adaptors.

(A) After its binding to viral RNA, RIG-I induces the polymerization of MAVS on the mitochondrial outer membrane. MAVS then recruits TRAF proteins to activate IKK and TBK1, which in turn phosphorylate MAVS at the consensus pLxIS motif. Phosphorylated MAVS binds to the C-terminal positively charged region of IRF3, thereby recruiting IRF3 for phosphorylation (P) by TBK1 through induced proximity. Phosphorylated IRF3 then forms a homodimer that enters the nucleus to turn on transcription. See the Discussion section for a more detailed description of the steps in the pathway. (B) Similar to (A), except that STING is phosphorylated by TBK1 in response to stimulation by DNA. DNA in the cytosol activates cGAS to produce the second messenger cGAMP, which then binds and activates STING. (C) lipopolysaccharide (LPS) stimulation activates TLR4, which in turn activates the adaptor proteins MyD88 and TRIF. TRIF activates TBK1, which in turn phosphorylates TRIF at the consensus motif. Phosphorylated TRIF then recruits IRF3 to facilitate IRF3 phosphorylation by TBK1.

The recruitment of IRF3 to phosphorylated MAVS, STING, and TRIF for its subsequent activation by TBK1 is markedly similar to the mechanism of SMAD activation by the TGF-β receptor (22). Analogous to TGF-β receptor I (TβR-I) phosphorylation by recruited TβR-II, MAVS, STING, and TRIF are phosphorylated at their conserved consensus serine-rich clusters by recruited TBK1 and/or IKK. Phosphorylated TβR-I and MAVS/STING/TRIF then recruit SMAD and IRF3, respectively, through structurally similar phospho-binding domains on SMAD and IRF3 that also mediate their respective homodimerization. This allows phosphorylated IRF3 or SMAD to dimerize and then enter the nucleus (30). In contrast, IL-1β, TNF-α, and other TLRs (e.g, TLR2) only activate IKK and TBK1, but not IRF3, due to the lack of a phosphorylated adaptor protein that can recruit IRF3.

Sequence profiles of IRF3, MAVS, STING, and TRIF across mammalian species revealed a consensus IRF3 binding motif: pLxIS (p, hydrophilic; x, nonaromatic) (fig. S11), which is always located in a disordered region of the respective protein. In addition, the presence of another conserved serine(s) can be found within 15 amino acids upstream of the pLxIS motif and serves as a second phosphorylation site (i.e., S385/S386 in IRF3 and S426/S430/S433 in MAVS) for IKK-related kinases. By taking these features into account, we performed a mammalian proteome-wide bioinformatic search to identify other proteins that harbor a similar motif (31, 32). Twenty-one mammalian gene candidates were found to contain such a motif, and they were grouped based on their functional similarity (fig. S12, and table S3, B and C; also see Materials and methods section on Computational biology). Among these 21 candidates, proteins related to innate immunity form the largest cluster, suggesting that the consensus motif found in MAVS, STING, TRIF, and IRF3 is mostly related to innate immune responses, possibly as a conserved mechanism to regulate the IRF family of transcription factors. In this regard, it is interesting to note that IRF5, a key transcription factor that regulates inflammation, also contains the pLxIS motif. The serine residue within this motif (S445 in mouse IRF5 and S446 in human IRF5) has recently been shown to be phosphorylated by IKKβ for IRF5 activation (33, 34). In summary, phosphorylation of innate immune adaptor proteins (e.g, MAVS, STING, and TRIF) and their recruitment of kinase substrates (e.g, IRF3) may be a general and conserved mechanism that provides signaling specificity in innate immunity.

Materials and methods

Antibodies

Rabbit antibodies against human IRF3, TRAF2, TRAF6, and IKKα/β and mouse antibody against human MAVS (residues 1 to 130) were obtained from Santa Cruz Biotechnology; Flag antibody (M2), M2-conjugated agarose, and tubulin antibody were purchased from Sigma; HA antibody and anti-HA–conjugated agarose were from Covance; antibodies against p-IRF3 Ser396, p-TBK1 Ser172, p-IκBα Ser32/36, and p-IKKα Ser176/p-IKKβ Ser177 were from Cell Signaling; and mouse IRF3 antibody was from Invitrogen. Anti-TBK1 monoclonal antibody was from IMGENEX. Polyclonal antibody against human TRIF was from Cell Signaling. Mouse immunoglobulin G TrueBlot ULTRA was from Rockland. The rabbit antibodies against human MAVS and STING were generated as described before (24, 35). Rabbit antibodies against MAVS p-Ser442 and STING p-Ser366 were generated by immunizing rabbits with KLH-conjugated, chemically synthesized peptides “acetyl-EDLAIS(phospho)ASTSC” and “acetyl-PELLIS(phospho)GMEKC,” respectively. The antibodies were affinity purified with corresponding phospho-peptide columns.

Expression constructs, recombinant proteins, and RNAi

For expression in mammalian cells, human cDNAs encoding N-terminal Flag-tagged IRF3 and IκBα were cloned into pcDNA3. After overexpression of these constructs in HEK293T cells, the encoded Flag-tagged proteins were further purified with M2 agarose, followed by Flag peptide elution. Human cDNA encoding MAVS WT and mutants were cloned into pTY-EF1A-puroR-2a lentiviral vector. Flag- or HA-tagged human IRF3-S385AS386A (2A) and other mutants were cloned into pTY-EF1A-HygromycinR-2a lentiviral vector. A human MAVS shRNA sequence (5′-GGAGAGAATTCAGAGCAAG-3′) containing U6 promoter, along with Flag-tagged human MAVS A440RM449R (2R), was cloned into pTY-U6-shRNA-EF1A-puroR-2a. STING lentiviral vectors pTY-U6-sh-mSTING-EF1A-puroR-2a-STING were cloned as described previously (24). These lentiviruses were transduced into Mavs–/– MEF cells, HEK293T cells, or L929 cells as described previously (24). HA- or Flag-tagged TRIF-N540 and mutants were cloned into pcDNA3. Flag-tagged TRIF-FL was in pEF-BOS vector. Mutants were constructed with the QuikChange Site-Directed Mutagenesis Kit (Stratagene; also see table S4 for the primer information). For expression in Escherichia coli, pET23a-His6-MAVS ΔTM [amino acids (aa) 1 to 510] and pET28a-His6-SUMO-MAVS-N460 (aa 1 to 460; WT and mutants) were transformed and expressed in E. coli BL21(DE3)-pLysS strain. These His-tagged proteins were purified as described previously (8). Sumo protease was subsequently used to cut off His6-SUMO tag, yielding nontagged proteins. In addition, His8-IRF3 (from E. coli), His6-TRAF6 (from insect Sf9 cells), GST-TBK1 (Sf9 cells; GST, glutathione S-transferase), and GST-IKKβ (Sf9 cells) were purified as described previously (8, 11, 14).

Viruses, cell culture, and transfection

Sendai virus (Cantell strain, Charles River Laboratories) was used at a final concentration of 100 hemagglutinating units/ml. VSV (ΔM51)-GFP virus was from J. Bell (Univ. of Ottawa) and was propagated in Vero cells. Plasmids and HT-DNA were transfected into cells using lipofectamine 2000 (Life Technologies). Digitonin permeabilization was used to deliver cGAMP into cultured cells as previously described (36). A lentiviral system for stable gene expression and shRNA knockdown were used as described before (24). All cells were cultured at 37°C in an atmosphere of 5% (v/v) CO2. HEK293T and Raw264.7 macrophages were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% (v/v) cosmic calf serum (Hyclone), penicillin (100 U/ml), and streptomycin (100 μg/ml). HeLa, MEF, U2OS, L929, and BJ cells were cultured in DMEM supplemented with 10% (v/v) fetal bovine serum (Atlanta) and antibiotics. THP1 cells were cultured in RPMI 1640 supplemented with 10% fetal bovine serum, 2 mM β-mercaptoethanol, and antibiotics. Sting–/– Raw264.7 macrophages were purchased from InvivoGen.

Generation of TBK1 knockout HEK293T cells by CRISPR/Cas9

Single-guide RNA (sgRNA) with the sequence 5′-CATAAGCTTCCTTCGTCCAG-3′ was designed for targeting exon 2 of the human TBK1 genomic locus. The sgRNA sequence driven by a U6 promoter was cloned into a lentiCRISPR vector that also expresses Cas9 as previously described (37). The lentiviral plasmid DNA was then packaged into a lentivirus for infection in HEK293T cells. Infected cells were selected in puromycin (2 μg/ml) for 2 weeks before single colonies were selected and tested for TBK1 expression by immunoblotting.

Biochemical assays for IRF3 activation and MAVS-IRF3 complex formation

Cell-free assays for IRF3 activation and phosphorylation of IκBα, IKKα/β, and TBK1 were preformed as described previously (8, 9, 11). Similarly, MAVS S442 and STING S366 phosphorylation was detected by immunoblotting with the p-S442 and p-S366 antibodies, respectively, after the cell-free assay. For a better MAVS p-S442 detection, His6-tagged MAVS-N460 in 100-μl reaction mixtures was pulled down with the Ni-NTA agarose in 8 M urea. After washing the agarose with buffer A [20 mM Tris-HCl (pH 7.0), 1 M NaCl, and 0.5% NP-40], MAVS phosphorylation was detected by immunoblotting with the p-S442 antibody.

To determine MAVS-IRF3 or STING-IRF3 complex formation in the cell-free assay, a reaction mixture (100 μl) containing buffer B [20 mM HEPES-KOH (pH 7.0), 2 mM adenosine triphosphate (ATP), 5 mM MgCl2, and 0.5 mM dithiothreitol (DTT)], 200 ng recombinant MAVSΔTM or MAVS-N460 or His6-STING (281 to 379), 200 μg cytosolic extracts (S5 or S100), and 200 ng Flag-IRF3 2A or GST-Flag-IRF3 2A was incubated at 30°C for 1 hour. In some experiments, Flag-IRF3 WT or 2A was added into the reaction mixture at 4°C right before IP as indicated. Flag-IP was then carried out using Flag antibody (M2) agarose at 4°C for 2 hours in the presence of buffer C [20mM Tris-HCl (PH 7.5), 150 mM NaCl, 0.5% NP-40] and the protease inhibitor cocktail (Roche). The agarose beads were washed three times with buffer C, and coprecipitated proteins were detected by immunoblotting.

For MAVS or STING phosphorylation by purified kinases TBK1 or IKKβ, a reaction mixture (10 μl) containing buffer A, 25 ng recombinant MAVS protein, and 40 ng GST-TBK1 or GST-IKKβ was incubated at 30°C for 40 min. Phosphorylation of MAVS and STING were detected by immunoblotting with the p-S442 and p-S366 antibodies, respectively.

To determine MAVS-IRF3 or STING-IRF3 complex formation in the purified kinase assay, a reaction mixture (100 μl) containing buffer A, 250 ng recombinant MAVS or STING protein, and 400 ng GST-TBK1 or GST-IKKβ was incubated at 30°C for 40 min. After stopping the reaction with 5 mM EDTA, Flag-IRF3 2A was added into the reaction mix at 4°C before IP. Flag-IP was then carried out using M2-agarose at 4°C for 1 hour in the presence of buffer C and 0.2 μg/μl bovine serum albumin. The agarose beads were washed three times with buffer C, and coprecipitated proteins were detected by immunoblotting.

IP assay for adaptor-IRF3 interactions in cells

To determine MAVS-IRF3 complex formation in cells, tagged IRF3 WT or 2A was stably expressed in the indicated cells. After virus infection, whole-cell lysates were prepared in the presence of buffer D [20 mM Tris-HCl (PH 7.5), 100 mM NaCl, 10% glycerol, 0.5% NP-40, and 0.5% DDM] and the protease inhibitor cocktail (Roche). Anti-Flag (M2) agarose or anti-HA agarose was added into the whole-cell lysate and incubated at 4°C for 2 hours. The agarose beads were washed three times with buffer D, and coprecipitated proteins were detected by immunoblotting.

To determine STING-IRF3 complex formation in cells, HA-tagged IRF3 WT or 2A was stably expressed in the indicated cells. After HT-DNA transfection, cells were homogenized in hypotonic buffer (10 mM Tris-Cl [pH 7.5], 10 mM KCl, 0.5 mM EGTA, 1.5 mM MgCl2, and Roche EDTA-free protease inhibitor cocktail). The homogenates were centrifuged at 1000 × g for 5 min to pellet nuclei and unbroken cells (P1). The supernatant (S1) was subjected to centrifugation at 5000 × g for 10 min to separate crude mitochondrial pellet from cytosolic supernatant (S5). Anti-HA agarose beads were added into the S5 and incubated at 4°C for 2 hours. The agarose beads were washed three times with buffer C, and coprecipitated proteins were detected by immunoblotting.

To determine TRIF-IRF3 complex formation in cells, tagged IRF3 and TRIF-N540 were transiently expressed in HEK293T cells. 24 hours after transfection, whole-cell lysates were prepared in the presence of buffer E [20 mM Tris-HCl (PH 7.5), 150 mM NaCl, 10% glycerol, 0.5% CHAPS] and the protease inhibitor cocktail (Roche). Anti-Flag (M2) agarose or anti-HA agarose beads were added into the whole-cell lysates and incubated at 4°C for 2 hours. The agarose beads were washed three times with buffer E, and coprecipitated proteins were detected by immunoblotting.

Quantitative reverse transcription polymerase chain reaction (q-RT-PCR)

Total RNA was isolated using TRIzol (Invitrogen). 0.1 μg total RNA was reverse-transcribed into cDNA using iScript Kit (Bio-Rad). The resulting cDNA served as the template for quantitative PCR analysis using iTaq Universal SYBR Green Supermix (Bio-Rad) and Real-Time PCR System (ABI). Primers for specific genes are listed as follows: mouse β-actin, 5′-TGACGTTGACATCCGTAAAGACC-3′ and 5′-AAGGGTGTAAAACGCAGCTCA-3′; mouse IFN-β, 5′CCCTATGGAGATGACGGAGA-3′ and 5′-CTGTCTGCTGGTGGAGTTCA-3′; mouse IFN-α, 5′-ATTTTGGATTCCCCTTGGAG-3′ and 5′-TATGTCCTCACAGCCAGCAG-3′.

Purification of MAVS, STING, and TRIF for phosphorylation site identification by mass spectrometry

For human MAVS, HEK293T cells were stably infected with a lentiviral vector, pTY-U6-shMAVS-EF1A-Flag-MAVS-2R, which depleted endogenous MAVS by shRNA and replaced it with a Flag-tagged MAVS containing two amino acids substitution (A440R/M449R or MAVS-2R), which permits trypsin digestion and therefore facilitates phosphorylation site mapping by mass spectrometry (see below). The cells were infected with Sendai virus to activate the RIG-I–MAVS pathway. For STING, HT-DNA was transfected into L929 cells with endogenous STING depleted by shRNA and replaced with mouse STING-Flag (pTY-U6-shSTING-EF1A-mSTING-flag). For TRIF, HA-TRIF-N540 was transfected alone or cotransfected with Flag-TBK1 in HEK293T cells. Cells were then harvested in PBS containing 3 mM EDTA, followed by addition of 1% SDS to denature proteins. The denatured cell lysates were sonicated to shear DNA and boiled at 95°C for 10 minutes. Excess SDS was removed with SDS-OUT precipitation kit from Pierce and Flag-IP for MAVS and STING or HA-IP for TRIF was carried out in the presence of buffer A [20 mM Tris-HCl (pH 7.0), 1M NaCl, 0.5% NP-40] at 4°C overnight. The anti-Flag M2 agarose beads were then washed three times with buffer A and buffer C. Bound STING or MAVS proteins were eluted with the Flag peptide. Eluted proteins were buffer-exchanged into 50 mM NH4HCO3, reduced in 5 mM DTT, and alkylated in 2.7 mM iodoacetamide. After trypsin digestion (1:20) at 30°C overnight, the peptide mixture was acidified with 1% formic acid, purified with C18 Zip-tip (Millipore), and then analyzed by nano–liquid chromatography–mass spectrometry (nano-LC-MS). For HA-TRIF, anti-HA agarose beads with bound proteins were directly boiled in SDS sample buffer containing 2% SDS and subjected to SDS-PAGE and Coomassie-blue staining. Gel slices near 75 kD were subsequently subjected to trypsin digestion and C18 Zip-tip purification, followed by nano-LC-MS analysis.

Mass spectrometry

Mass spectrometry analyses and targeted quantification of tryptic MAVS, STING, or TRIF peptides were conducted on a Dionex Ultimate 3000 nanoLC system coupled to a Q-Exactive mass spectrometer (Thermo Scientific). The LC conditions and ion source parameters have been described before (4, 9). For tandem MS/MS analyses, full-scan mass spectra were acquired in the range of mass/charge ratio (m/z) = 300 to 1500, with a resolution of 70,000 at m/z = 200 in the Orbitrap. MS/MS spectra (resolution: 17,500 at m/z = 200) were acquired in a data-dependent mode whereby the top 15 most abundant parent ions were subjected to further fragmentation by higher-energy collision dissociation (HCD). Phosphorylated and nonphosphorylated TRIF S210–containing peptides were directly analyzed and quantified with Xcalibar 2.2 (Thermo Scientific), according to the specific ions indicated in table S1.

To quantify phosphorylated and nonphosphorylated MAVS and STING peptides, targeted SIM and targeted MS2 assays were developed on the Q-Exactive mass spectrometer. The precursor ions in the inclusion list for both of the assays are shown in table S1. Settings for targeted SIM were: resolution, 70,000; AGC target, 5E4; maximum injection time, 250 ms; isolation window, 0.5 m/z. The targeted MS2 settings were as follows: resolution, 35,000; AGC target, 2E4; maximum injection time, 120 ms; isolation window, 0.5 m/z; and normalized collision energy, 30. Data acquisition and analyses were performed with Xcalibar 2.2. The relative abundance of each peptide or site-specific phosphorylation on the same peptide was represented by the intensity of product ions that are specific to each phosphorylation site (see table S1).

Targeted mass spectrometry identifies MAVS phosphorylation sites in virus-infected cells

The MAVS C-terminal region containing the 3S and 4T/S sites (aa 420 to 460) lacks lysine and arginine (fig. S2A). To facilitate trypsin digestion and mass spectrometry analysis, a Flag-tagged MAVS mutant (A440R/M449R or MAVS-2R) was introduced into a HEK293T cell line that was depleted of endogenous MAVS by shRNA. The MAVS-2R mutant robustly rescued IRF3 dimerization after viral infection (fig. S4C), indicating that the substitutions do not alter the function of MAVS. Trypsin digestion of purified Flag-MAVS-2R from cells created a short peptide (“ISASTSLGR”) including the 4T/S site (4T/S peptide), which was identified by nano-LC-MS. Following viral infection, we observed substantially increased abundance of an ion with m/z of 486.23 (z = 2+), the fragmentation pattern of which matched to singly phosphorylated 4T/S peptides (fig. S4D). Furthermore, by using targeted quantification, we observed robust induction of singly phosphorylated 4T/S peptides from the virus-infected sample (Fig. 3B and fig. S4D). The singly phosphorylated peptides eluted as two peaks (peak a and b). By examining site-localizing fragment ions specific to each phospho-serine in the MS2 spectra, we found that peak a contains mainly p-S446 peptides, whereas peak b is a mixture of p-S442 and p-S444 peptides (fig. S4, E and F). Both phospho-S442 and total phospho-4T/S signals were induced by virus infection (Fig. 3B).

Targeted mass spectrometry reveals phosphorylation of mouse STING at Ser365 in DNA-stimulated cells

We generated L929 cells stably expressing a C-terminal Flag-tagged mouse STING (human STING C-terminus lacks arginine or lysine that can be cleaved by trypsin) in place of endogenous STING, which was depleted by shRNA. In these cells, HT-DNA transfection resulted in strong IRF3 activation, which was completely blocked by treatment with the TBK1 inhibitor, BX795 (fig. S8F). Tryptic digestion of purified mSTING-Flag from the DNA-transfected cells led to the identification of a peptide (“LLIS365GMDQPLPLR”) covering the S365 site (S366 in human STING) by nano-LC-MS. Following virus infection, we observed substantially increased abundance of an ion with m/z of 766.90 (z = 2+), which was confirmed to be phosphorylated S365 peptides by HCD fragmentation. Furthermore, by using targeted quantification (table S1), we observed robust induction of the phosphorylated S365 peptide only from the DNA-transfected sample (Fig. 5C).

Kinase inhibitors for cell-free assay or cell-based experiments

The kinase inhibitors were dissolved in dimethyl sulfoxide (DMSO) and used both in vitro and in cells at the following final concentrations: TBK1 inhibitor (BX795, Selleckchem), 4 μM; IKK inhibitor (TPCA-1, Sigma), 20 μM; and PLK1 inhibitor (BI2356, Selleckchem), 20 μM. For cell-free assays, the kinase inhibitors were incubated with the reaction mixtures at 30°C, and the final DMSO concentration was kept below 5% of total reaction volume. In cell-based experiments, the inhibitors were added 1 hour before viral infection, HT-DNA, or TRIF transfection.

Immunofluorescence

BJ cells with or without HT-DNA transfection were fixed and incubated with the STING p-S366–specific antibody, followed by a secondary antibody conjugated with Alexa Fluor 488. Green fluorescent signals were imaged by a Zeiss LSM710 confocal microscope (Carl Zeiss). Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI) in the mounting medium (Vector Labs).

Proximity ligation assay (PLA)

Proximity ligation assays to detect STING phosphorylation and STING-IRF3 interaction were conducted using the Duolink In Situ Red Kit Mouse/Rabbit (Sigma-Aldrich). L929 cells stably expressing huSTING-Flag and HA-IRF3 2A with or without HT-DNA transfection were fixed and incubated with primary antibodies against two proteins as indicated; secondary antibodies conjugated with plus and minus PLA probes were then added. After ligation, rolling circle amplification, and hybridization with fluorescently labeled oligonucleotides, red fluorescent dots (indicating close proximity of proteins recognized by two different primary antibodies) were imaged by a Zeiss LSM710 confocal microscope. Cell shapes were indicated by phalloidin-labeled actin filaments, and nuclei were stained with DAPI in the mounting medium.

Computational biology

Combining the experimental data and sequence profiles of IRF3, MAVS, STING, and TRIF (fig. S11), we derived that an IRF3 binding motif harbors the following sequence features: pLxIS (p, hydrophilic; x, nonaromatic; S, phosphorylated) and another serine surrounded by a hydrophobic residue within the upstream 15 amino acids. Starting from all of the 99,459 alternatively spliced isoforms of human proteins in the Ensembl database (38), we identified 4229 motifs with this pattern. Requiring this motif to be conserved among the 43 mammalian species (table S3A), we narrowed this list down to 403 motifs from 399 isoforms. We required 90% of sequences to follow the sequence pattern for the most conserved positions (the second and the last two positions), whereas for those more variable positions, we allowed 20% of sequences to be amino acids that were not suggested as permissible in experimental data and sequence alignments of IRF3, MAVS, STING, and TRIF.

The observation that this motif tends to be in the flexible linkers (usually predicted to be disordered) between domains helped us to further reduce the number of candidates. We detected domains in these proteins using HMMER (39) against Pfam domains (40) and predicted the disordered regions using the Espritz Web server (41) with parameters trained on disordered proteins in the Disprot database (42). We identified 72 candidate motifs from 72 protein isoforms that were largely in the linkers between domains and that were predicted to be largely disordered. All of these candidate protein isoforms were further mapped to the Uniprot entries (43), resulting in the final candidate list of 21 protein-coding genes (table S3B).

We extracted the gene ontology (GO) terms (44) in the category of “biological process” associated with these candidate proteins from Uniprot. The enriched GO terms associated with these proteins were detected with a binomial test, and the most significantly enriched GO terms are listed in table S3C. In addition, we used the number of shared GO terms to represent the functional similarity between proteins and clustered all of the candidates in CLANS (45) based on similarity in function; the result is shown in fig. S12. We further assigned a confidence level of these motifs based on: (i) how well they overlap with domain linkers and disordered regions and (ii) additional sequence features we observed in IRF3, MAVS, STING, and TRIF (i.e., the presence of multiple serine residues upstream of the motif and the preference of negatively charged versus positively charged residues in the motif). The confidence level of these motifs is reflected by the color code in fig. S12.

The largest cluster of proteins containing the consensus motif and other features described above is that related to innate immune responses (fig. S12). Statistical tests also support significant enrichment in several GO terms related to innate immune response. In addition to MAVS, STING, TRIF, and IRF3, the proteins identified from this analysis include IRF5, which is important for proinflammatory cytokine induction by multiple pathways; brain-specific angiogenesis inhibitor 1–associated protein 2 (BAIP2_HUMAN in Uniprot); and dual-specificity mitogen-activated protein kinase kinase 4 (MP2K4_HUMAN in Uniprot). The roles and regulations of these and other proteins identified from our computational analysis require further studies.

Supplementary Materials

www.sciencemag.org/content/347/6227/aaa2630/suppl/DC1

Figs. S1 to S12

Tables S1 to S4

References (4749)

References and Notes

  1. Single-letter abbreviations for the amino acid residues are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr.
  2. Acknowledgments: We thank Y. Yu, H. Mirzae, and C. Long for advice on mass spectrometry; L. Jia for advice on generating phospho-specific antibodies; and L. Sun, N. Varnado, and M. Xu for critically reading our manuscript. The data presented in this manuscript are tabulated in the main paper and in the supplementary materials. This work was supported by grants from the NIH (AI-93967 and GM-63692 to Z.J.C. and GM-094575 to N.G.) and the Welch Foundation (I-1389 to Z.J.C. and I-1505 to N.G.). X.Cai, J.W., and Q.C. were supported by International Student Fellowships from HHMI. Z.J.C. is an HHMI Investigator.
View Abstract

Subjects

Navigate This Article