Main

Nucleotides are crucial signalling molecules in all domains of life, but cyclic dinucleotides seem to be produced solely by Bacteria and Archaea. For example, c-di-GMP is a ubiquitous second messenger that regulates biofilm formation, motility and virulence in a diverse range of bacterial species8. Recently, cyclic diadenylate monophosphate (c-di-AMP) was discovered to be a bacterial regulatory molecule14, although its role remains to be fully characterized. Because they are unique to microorganisms, cyclic dinucleotides are appropriate targets for immune recognition15. Indeed, the induction of IFN production by Listeria monocytogenes depends on bacterial secretion of cyclic-di-AMP12. However, it remains unclear how cyclic dinucleotides are sensed in mammalian cells.

To address the mechanism by which mammalian cells sense cyclic dinucleotides, we first confirmed that cyclic dinucleotides are detected in the host cell cytosol10 by expressing RocR, a c-di-GMP-specific phosphodiesterase from Pseudomonas aeruginosa, in the cytosol of macrophages. In these cells, IFN induction in response to c-di-GMP (but not other stimuli) is tenfold lower than that in control, vector-transduced, cells (Fig. 1a), confirming that the cytosolic presence of c-di-GMP is important for inducing IFN.

Figure 1: STING is sufficient to restore responsiveness to cyclic dinucleotides.
figure 1

a, Immortalized Myd88−/−Trif−/− macrophages were transduced with retrovirus expressing rocR and then stimulated for 6 h. IFN induction was measured by quantitative PCR with reverse transcription (quantitative RT–PCR) and normalized to ribosomal protein 17 (Rps17) expression levels. b–d, HEK293T cells were transfected as indicated, together with an IFN-luciferase reporter, and luciferase activity was measured 6 h after stimulation. e, Myd88−/−Trif−/− macrophages were stimulated for 6 h, and IFN induction was measured as in a. a–e, Data are presented as mean ± s.d. (n = 3) and are representative of at least three independent experiments. ***, P < 0.001. SeV, Sendai virus; TMEV, Theiler’s murine encephalomyelitis virus; unstim, unstimulated; VV70mer, a stimulatory dsDNA oligonucleotide derived from vaccinia virus.

PowerPoint slide

Full size image

To identify candidate cyclic dinucleotide sensors, we sought to identify molecules that could reconstitute the IFN response to cyclic dinucleotides in HEK293T cells, which do not respond to c-di-GMP10. Because STING is essential for the IFN response to cyclic dinucleotides11 and because STING expression is low or undetectable in HEK293T cells (Supplementary Fig. 1 and data not shown), we first expressed STING in HEK293T cells. The overexpression of STING spontaneously induces an IFN reporter3,6,7, so we transfected a small amount of Sting-encoding vector that by itself was insufficient to induce IFN. Low levels of STING protein were sufficient to reconstitute the responsiveness of HEK293T cells to c-di-GMP (Fig. 1b) and c-di-AMP (Fig. 1c). By contrast, the non-functional goldenticket (gt) allele of Sting (which results in a STING protein in which asparagine has been substituted for isoleucine, I199N)11 did not restore responsiveness to c-di-GMP (Fig. 1b). Interestingly, the expression of wild-type STING did not reconstitute the responsiveness of HEK293T cells to double-stranded DNA (dsDNA) oligonucleotides (for example, a 70-base-pair oligonucleotide from vaccinia virus (VV70mer) or IFN-stimulatory DNA (ISD)) that had previously been shown to induce type I IFNs in macrophages through STING4,16 (Fig. 1b and Supplementary Fig. 2a). By contrast, the induction of IFN by poly(dA-dT)˙poly(dT-dA) DNA (denoted poly(dAT:dTA)) was identical in cells that were transfected with wild-type Sting and those transfected with gt, demonstrating that the RNA polymerase III DNA-sensing pathway17,18 is intact in these cells and is not responsible for the detection of c-di-GMP (Fig. 1d). As a positive control, Myd88−/−Trif−/− immortalized macrophages, which express STING, responded similarly to c-di-GMP, poly(dAT:dTA), VV70mer and ISD (Fig. 1e and Supplementary Fig. 2b). Together, our results show that STING expression is sufficient to restore the responsiveness of HEK293T cells to cyclic dinucleotides but not to DNA.

We next tested whether STING, or perhaps another protein in HEK293T cells, binds to c-di-GMP. We used HEK293T cell lysates in an in vitro ultraviolet radiation crosslinking assay to identify putative sensor proteins that interact directly with radiolabelled c-di-GMP (c-di-[32P]GMP). We expected to identify directly interacting proteins because only molecules within bond-length proximity are efficiently crosslinked by ultraviolet radiation19. We detected a prominent ∼40-kDa radiolabelled protein, which corresponds to the predicted molecular weight of monomeric STING, in the lysates of cells transfected with a vector encoding haemagglutinin (HA)-tagged STING (STING–HA) but not in the lysates of cells transfected with a vector encoding STING(I199N)–HA or vector only (Fig. 2a). The ∼40-kDa band did not appear when the same lysates were crosslinked in the presence of [32P]GTP, implying that crosslinking to c-di-[32P]GMP was specific (Fig. 2a). We also observed an ∼80-kDa species, which might correspond to a previously reported STING dimer6 (Fig. 2b). To test the hypothesis that STING crosslinks with c-di-[32P]GMP, we immunoprecipitated STING from transfected HEK293T cells and performed the c-di-[32P]GMP crosslinking assay on the immunoprecipitates. Bands corresponding to the molecular weight of the STING monomer and dimer were identified only in immunoprecipitates from lysates overexpressing STING and not in mock immunoprecipitates from lysates of vector-only-transfected cells (Fig. 2b). Thus, STING seems to bind to c-di-GMP.

Figure 2: STING binds cyclic dinucleotides.
figure 2

a, HEK293T cells were transfected as indicated, and the cell lysates were subjected to an in vitro ultraviolet radiation crosslinking assay with c-di-[32P]GMP. Samples were separated by SDS–PAGE and visualized by autoradiography or western blotting. b, HEK293T cells were transfected as in a, and anti-HA immunoprecipitates were stained with colloidal blue or subjected to the crosslinking assay with c-di-[32P]GMP. c, d, HEK293T cells were transfected as in a, and the cell lysates were crosslinked with ultraviolet radiation to c-di-[32P]GMP or [32P]GTP in the presence of cold competing nucleotides in tenfold serial dilutions beginning at 1 mM (as indicated by the wedges), except for guanosine (beginning at 0.1 mM), VV70mer (500 μg ml−1), poly(dAT:dTA) (50 μg ml−1) and poly(I)•poly(C) (denoted poly(I:C)) (50 μg ml−1). The arrow indicates STING, and the asterisk indicates a nonspecific band. nt, nucleotides; ppGpp, guanosine-3′,5′-bisdiphosphate. e, His6–STING (amino acids 138–378) (1 µg) was separated by SDS–PAGE and stained with Coomassie blue. f, His6–STING 138–378 was analysed as in c. g, The binding of c-di-GMP to purified His6–STING (10 µM) was measured by equilibrium dialysis. Bmax, maximum number of binding sites; h, hill coefficient; Kd, dissociation constant. a–g, Data are representative of three independent experiments.

PowerPoint slide

Full size image

To confirm that the binding of c-di-[32P]GMP to STING is specific, we performed the c-di-GMP crosslinking assay in the presence of unlabelled nucleotides. Unlabelled c-di-GMP and c-di-AMP specifically competed with c-di-[32P]GMP for binding to STING (Fig. 2c, d). By contrast, GTP, other guanosine derivatives and nucleic acids (including dsDNA) competed away nonspecific binding (Fig. 2c, d, asterisk); however, under our specific assay conditions, these molecules could not compete efficiently with c-di-[32P]GMP for binding to STING (Fig. 2c, d, arrow). Because the cell cytosol contains high concentrations of GTP (0.1–1 mM), a putative c-di-GMP sensor must have a high degree of specificity for c-di-GMP over GTP. We found that c-di-GMP efficiently crosslinked to STING even in the presence of 1 mM GTP (Fig. 2c).

Although these data imply that STING directly and specifically binds to cyclic dinucleotides, they do not address whether other host proteins might also be required. STING is predicted to encode an amino-terminal domain with multiple transmembrane segments, followed by a globular carboxy-terminal domain (CTD). Because the CTD contains the amino acid substitution that abolishes STING function in gt mice (I199N)11, we suspected that the CTD might be involved in binding to cyclic dinucleotides. Thus, we subjected purified recombinant His6-tagged STING CTD (amino acids 138–378) (Fig. 2e) to the c-di-[32P]GMP crosslinking assay. We found that the recombinant CTD of STING bound to c-di-[32P]GMP and that this binding was specifically competed away with cold (unlabelled) c-di-GMP or c-di-AMP but not with cold GTP or ATP (Fig. 2f). We used equilibrium dialysis to obtain an estimate of the affinity (dissociation constant, Kd) of c-di-GMP binding to the STING CTD, which was ∼5 µM (Fig. 2g). In its native membrane-bound form, or in complex with other host factors, STING may have a stronger affinity for c-di-GMP; nevertheless, a 5 µM affinity is consistent with the dose response that has previously been observed in macrophages12. Consistent with the ability of STING to dimerize6, the binding data suggest a stoichiometry of one molecule of c-di-GMP per two molecules of STING.

To identify the amino acids involved in c-di-GMP binding and/or IFN induction, we introduced point mutations into STING. Focusing on clusters of conserved and charged residues, we mutated 67 amino acids, either individually or in groups, and we classified these mutants into five categories (Fig. 3, Supplementary Table 1 and Supplementary Figs 3 and 4). Class I consists of mutants that abolish both binding and IFN induction (Fig. 3a–c, red, and Supplementary Table 1). Class II mutants bind to c-di-GMP but fail to induce IFN (Fig. 3c, purple). Class III comprises ‘hyperactive’ mutants, which spontaneously induce IFN at low levels of transfection (Fig. 3a–c, green, and Supplementary Table 1). Class IV mutants induce IFN when overexpressed but do not respond to c-di-GMP (Fig. 3a–c, blue, and Supplementary Table 1). Class V consists of mutants that have no effect on c-di-GMP binding or IFN induction (Fig. 3c, yellow, and Supplementary Table 1). Although mutating STING can result in diverse phenotypes, a key finding is that all mutants that failed to bind to c-di-GMP also lost the ability to induce IFN in response to c-di-GMP. Consistent with our observation that the CTD is sufficient for binding to c-di-GMP (Fig. 2f), all mutations that affected c-di-GMP binding were located within the CTD.

Figure 3: Mutational analysis of STING.
figure 3

a, HEK293T cells were transfected with vectors encoding STING and STING mutants as indicated, together with an IFN-luciferase reporter, and luciferase activity was measured 6 h after stimulation. Data are presented as mean ± s.d. (n = 3). b, HEK293T cells were transfected as in a, except that the lysates were subjected to the crosslinking assay with c-di-[32P]GMP as in Fig. 2a. c, Organization of STING based on the membrane topology prediction programs SOSUI, TMHMM, HMMTOP and TMpred. Strongly predicted transmembrane domains are boxed; weakly predicted transmembrane domains have dashed boxes. Coloured residues indicate mutant classes (see text for description of mutants): class I (red), class II (purple), class III (green), class IV (blue) and class V (yellow). Bracketed mutations were made in combination. a–c, Data are representative of at least three independent experiments.

PowerPoint slide

Full size image

DNA and cyclic dinucleotides induce indistinguishable transcriptional responses in macrophages10, and STING seems to be essential for both responses4,11. However, we found that STING expression is insufficient to restore the responsiveness of HEK293T cells to DNA, in contrast to cyclic dinucleotides (Fig. 1b and Supplementary Fig. 2a). Moreover, our competition assays indicate that DNA does not compete with cyclic-di-GMP for binding to STING under the conditions tested (Fig. 2d). Thus, although our data indicate that STING functions as a direct immunosensor of cyclic dinucleotides, it seems likely that additional host proteins are involved in IFN induction by DNA. Indeed, two candidate DNA sensors, DAI (also known as ZBP1) and IFI16, have been identified16,20, neither of which seems to be essential for the response to cyclic dinucleotides (ref. 10 and data not shown). To determine whether the responsiveness to cyclic dinucleotides and DNA are separable functions of STING, we sought to identify STING mutants that fail to respond to cyclic dinucleotides but still respond to DNA. We identified a STING mutant (R231A) that was unresponsive to c-di-GMP (Fig. 4a), although it induced IFN when overexpressed (Fig. 4a) and bound to c-di-GMP (Fig. 4b). Interestingly, STING R231A was able to restore the responsiveness of gt bone marrow macrophages to DNA but not to cyclic-di-GMP (Fig. 4c). Thus, cyclic dinucleotide sensing and DNA sensing can be uncoupled, suggesting that these two pathways are discrete but share STING as a common signalling molecule. It is unexpected that STING would function both as a direct immunosensor (of cyclic dinucleotides) and as a signalling adaptor (in the response to DNA). One possibility is that STING initially evolved as a cyclic dinucleotide sensor and was subsequently co-opted for DNA sensing.

Figure 4: The IFN response to DNA and c-di-GMP can be uncoupled.
figure 4

a, HEK293T cells were transfected as indicated, together with an IFN-luciferase reporter, and IFN reporter activity was measured 6 h after stimulation. b, HEK293T cells were transfected as in a, except that the lysates were subjected to the crosslinking assay with c-di-[32P]GMP as in Fig. 2a. c, Bone-marrow-derived macrophages from Sting-deficient (gt) mice were transduced with the indicated constructs. IFN induction in response to transfected cyclic-di-GMP or VV70mer was measured by quantitative RT–PCR and normalized to Rps17. ***, P < 0.001; NS, not significant, P = 0.1205. a, c, Data are presented as mean ± s.d. (n = 3). a–c, Data are representative of at least three independent experiments.

PowerPoint slide

Full size image

We previously used mouse mutagenesis to identify STING as an essential molecule in the in vivo IFN response to cyclic dinucleotides11. The requirement for STING can now be rationalized by our proposal that STING functions as a direct sensor of cyclic dinucleotides. Interestingly, STING does not share homology with any known immunosensor and therefore seems to represent a novel category of microbial detector. Although a BLAST search of the mouse proteome for homologues of the L. monocytogenes diadenylate cyclase (lmo2120; also known as DacA) identifies STING as the top hit, the homology is limited to a short region of the STING CTD (amino acids 311–358). STING does not seem to share homology with PilZ-domain-containing proteins, which function as c-di-GMP receptors in bacteria8. Structural studies are required to better characterize the interaction of STING with cyclic dinucleotides and to determine whether STING resembles any known protein in mammals or bacteria.

Numerous studies have demonstrated that cyclic dinucleotides are potent immunostimulatory compounds that may be valuable as novel immunotherapeutics or adjuvants9,13. The therapeutic development of cyclic dinucleotides will be greatly facilitated by an improved understanding of the mechanism by which they are sensed. Furthermore, our finding that STING is a direct detector of cyclic dinucleotides provides insight into the fundamental mechanisms by which the innate immune system can detect bacterial infection.

Methods Summary

Transfections

Transfections were carried out using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s instructions. Digitonin permeabilization was used to introduce c-di-AMP into cells as described previously12.

Recombinant STING

DNA encoding the CTD of mouse STING (nucleotides 414–1,137) was cloned into the vector pET28a for recombinant protein expression in Escherichia coli.

Ultraviolet radiation crosslinking

c-di-[32P]GMP was enzymatically synthesized using recombinant WspR and was used in an ultraviolet radiation crosslinking assay as described previously21. Briefly, 50 μg HEK293T cell lysate at a final concentration of 2 μg μl−1, or 1 μg recombinant His6-tagged STING, was incubated with 2 μCi c-di-[32P]GMP in binding buffer (20 mM Tris-HCl, pH 7.4, 200 mM NaCl and 1 mM MgCl2) for 15 min at 25 °C. The reactions were irradiated at 254 nm, and the proteins in the samples were then separated by SDS–PAGE.

Statistical analysis

Statistical differences were calculated with an unpaired two-tailed Student’s t-test using Prism 5.0b software (GraphPad).

Online Methods

Synthesis of c-di-[32P]GMP

The synthesis of c-di-[32P]GMP was carried out as described previously21. Briefly, recombinant His6-tagged WspR was incubated with [α-32P]GTP (3,000 Ci mmol−1, 10 μCi μl−1, Amersham Biosciences) for 2 h at 25 °C, followed by heat inactivation of WspR at 95 °C for 5 min. Residual 32P was removed by incubation with calf intestinal phosphatase (CIP) (New England Biolabs) for 10 min at 37 °C. CIP was heat inactivated at 95 °C for 5 min, followed by centrifugation at 16,000g for 5 min. The [32P]GTP used as a negative control was prepared identically except that His6-tagged WspR and CIP were omitted from the preparation. Radiolabelled nucleotides were quantified by separation by thin-layer chromatography on cellulose-PEI plates (Macherey-Nagel) using 1.5 M KH2PO4, pH 3.65 (Supplementary Fig. 5).

Cell lines and animals

C57BL/6 Myd88−/−Trif−/− mice were obtained from G. Barton, and immortalized macrophages were generated as described previously22. Immortalized bone marrow macrophages were maintained in RPMI 1640 (Invitrogen) supplemented with 10% FBS, penicillin-streptomycin and glutamine. HEK293T cells were maintained in DMEM supplemented with 10% FBS, penicillin-streptomycin and glutamine. Animal use was approved by the Animal Care and Use Committee at the University of California, Berkeley.

Plasmids

A construct encoding RocR (NP_252636) from P. aeruginosa23 was a gift from S. Lory. rocR was cloned into the MSCV2.2 retroviral expression construct upstream of an internal ribosome entry site (IRES)–green fluorescent protein (GFP). MSCV-rocR was transduced into immortalized macrophages from Myd88−/−Trif−/− mice, and cells were sorted for GFP expression. Mouse Sting and the gt (I199N) mutant allele of Sting were cloned into the vector pcDNA3 with a C-terminal HA tag as described previously11. DNA encoding the CTD of mouse Sting (nucleotides 414–1,137) was cloned into pET28a for recombinant protein expression in Escherichia coli.

Site-directed mutagenesis

Mutations in Sting were generated using the QuikChange Site-Directed Mutagenesis Kit (Stratagene) according to the manufacturer’s guidelines.

Reagents

c-di-GMP was synthesized as described previously24. Purified c-di-AMP was a gift from J. Woodward and D. Portnoy. Poly(dAT:dTA), GTP, ATP, GMP and guanosine were obtained from Sigma-Aldrich. Poly(I)•poly(C) (denoted poly(I:C)) was purchased from Invivogen. Guanosine-3′,5′-bisdiphosphate (ppGpp) was obtained from TriLink Biotechnologies. Sendai virus was purchased from Charles River Laboratories. Theiler’s murine encephalomyelitis virus (TMEV) strain GDVII was provided by M. Brahic and E. Freundt. DNA oligonucleotides corresponding to the VV70mer and ISD were purchased from Elim Biopharmaceuticals and were annealed as described previously2,16.

Cell stimulation

All transfections (excluding for c-di-AMP) were carried out using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s instructions. The VV70mer was transfected at a final concentration of 0.5 µg ml−1. Poly(dAT:dTA), poly(I:C) and c-di-GMP were transfected at a final concentration of 4 µg ml−1. c-di-AMP was used at a final concentration of 5.4 mM, and stimulation was performed using digitonin permeabilization as described previously12.

Luciferase assay

HEK293T cells were plated in TC-treated 96-well plates at 0.5 × 106 cells ml−1. The next day, the cells were transfected as indicated, together with IFN-β-firefly luciferase and TK-Renilla luciferase reporter constructs. Following stimulation for 6 h with the indicated ligands, the cells were lysed in passive lysis buffer (Promega) for 5 min at 25 °C. The cell lysates were incubated with firefly luciferase substrate (Biosynth) and the Renilla luciferase substrate coelenterazine (Biotium), and luminescence was measured on a SpectraMax L microplate reader (Molecular Devices). The relative Ifnb expression was calculated as firefly luminescence relative to Renilla luminescence.

Quantitative PCR

The analysis of Ifnb expression by bone marrow macrophages was conducted as described previously10.

Preparation of HEK293T cell lysates and immunoprecipitations

HEK293T cells were plated at a density of 1 × 106 cells well−1 in a 6-well plate. The following day, the cells were transfected with pcDNA3 or pcDNA3 expressing HA-tagged wild-type or mutant STING using Lipofectamine 2000 (Invitrogen). The day after that, the cells were rinsed once with PBS and transferred to Eppendorf tubes in PBS containing 1 mM EDTA. The cells were pelleted briefly by centrifugation at 1,000g at 4 °C. The cell pellet was lysed in an equal volume of digitonin lysis buffer (0.5% digitonin, 20 mM Tris-HCl, pH 7.4, and 150 mM NaCl) containing protease inhibitors (Roche) for 10 min on ice. The cell lysates were centrifuged at 10,000g for 10 min at 4 °C. The protein concentration in the resultant supernatant was measured using the Bradford reagent (Bio-Rad).The cell lysates were subjected to a c-di-GMP binding (crosslinking) assay (see below). The lysates were separated by SDS–PAGE, and the separated proteins were transferred to a nitrocellulose membrane, which was then probed with rat anti-HA antibodies (Roche), to confirm STING–HA expression, and mouse anti-β-actin antibodies (Santa Cruz Biotechnology). To immunoprecipitate HA-tagged STING, the cell lysates were prepared similarly in digitonin lysis buffer and incubated with anti-HA-antibody-conjugated agarose beads (Sigma) for 2 h at 4 °C. Washed beads were subjected to a c-di-GMP binding assay or separated by SDS–PAGE and stained with colloidal blue protein stain (Thermo Scientific).

c-di-GMP binding assays

The c-di-GMP binding assay (also called the crosslinking assay) was based on a method described previously21. Briefly, 50 μg HEK293T cell lysate at a final concentration of 2 μg μl−1, or 1 μg recombinant His6-tagged STING, was incubated with 2 μCi radiolabelled nucleotide in binding buffer (20 mM Tris-HCl, pH 7.4, 200 mM NaCl and 1 mM MgCl2) for 15 min at 25 °C. The reactions were irradiated at 254 nm for 20 min on ice at a 3-cm distance from a UVG-54 mineral light lamp (UVP). Immediately after crosslinking, the reactions were terminated by the addition of SDS sample buffer (40% glycerol, 8% SDS, 2% 2-mercaptoethanol, 40 mM EDTA, 0.05% bromophenol blue and 250 mM Tris-HCl, pH 6.8), boiled for 5 min and then separated by SDS–PAGE. The gels were dried, exposed to a phosphor screen and visualized using a Typhoon Trio imager (GE Healthcare).

Protein purification

The construct expressing a constitutively active form of WspR (pQE-WspR*)23,25 was a gift from S. Lory. The purification of His6-tagged WspR was carried out as described previously, using Ni-NTA affinity chromatography (Qiagen)26. DNA encoding the CTD of mouse STING (nucleotides 414–1,137) was cloned into the vector pET28a and purified by Ni-NTA affinity chromatography (Qiagen) according to the manufacturer’s instructions.

Equilibrium dialysis

The binding affinity of radioactive c-di-GMP was measured by equilibrium dialysis, using a 96-well equilibrium dialyser (Harvard Apparatus) with a 5,000 molecular weight cut-off membrane. One chamber contained 150 µl 10 µM purified His6-tagged STING(138–378) in assay buffer (25 mM Tris-HCl, pH 7.4, 100 mM NaCl, 1 mM MgCl2 and 10% glycerol), while the other was filled with 150 µl c-di-[32P]GMP at a range of concentrations (40–160 µM). Equilibrium was reached after 48 h at 25 °C, and three samples were drawn from each chamber and mixed with 2 ml Econo-Safe scintillation fluid. Samples were measured in an LS 6000 IC scintillation counter (Beckman). Data analysis was performed using Prism 5.0b software (GraphPad). The dissociation constant (Kd), the maximum number of binding sites (Bmax) and the Hill coefficient (h) were generated using nonlinear regression, allowing one-site specific binding with a Hill slope.