AMP …AMPK 2 / MEFs, and resulted in decreased induction of IFN- mRNA at 6 h post-infection in AMPK -deficient MEFs compared with WT MEFs (Fig. 6A). Reduced IFN- induction is crucial

AMP-activated Kinase (AMPK) Promotes Innate Immunity andAntiviral Defense through Modulation of Stimulator ofInterferon Genes (STING) Signaling*

Received for publication, October 13, 2016, and in revised form, November 15, 2016 Published, JBC Papers in Press, November 22, 2016, DOI 10.1074/jbc.M116.763268

Daniel Prantner, Darren J. Perkins, and Stefanie N. Vogel1

From the Department of Microbiology and Immunology, University of Maryland, School of Medicine, Baltimore, Maryland 21201

Edited by Luke O’Neill

The host protein Stimulator of Interferon Genes (STING) hasbeen shown to be essential for recognition of both viral andintracellular bacterial pathogens, but its regulation remainsunclear. Previously, we reported that mitochondrial membranepotential regulates STING-dependent IFN-� induction inde-pendently of ATP synthesis. Because mitochondrial membranepotential controls calcium homeostasis, and AMP-activatedprotein kinase (AMPK) is regulated, in part, by intracellular cal-cium, we postulated that AMPK participates in STING activa-tion; however, its role has yet to be been defined. Addition of anintracellular calcium chelator or an AMPK inhibitor to eithermouse macrophages or mouse embryonic fibroblasts (MEFs)suppressed IFN-� and TNF-� induction following stimulationwith the STING-dependent ligand 5,6-dimethyl xanthnone-4-acetic acid (DMXAA). These pharmacological findings werecorroborated by showing that MEFs lacking AMPK activity alsofailed to up-regulate IFN-� and TNF-� after treatment withDMXAA or the natural STING ligand cyclic GMP-AMP(cGAMP). As a result, AMPK-deficient MEFs exhibit impairedcontrol of vesicular stomatitis virus (VSV), a virus sensed bySTING that can cause an influenza-like illness in humans. Thisimpairment could be overcome by pretreatment of AMPK-defi-cient MEFs with type I IFN, illustrating that de novo productionof IFN-� in response to VSV plays a key role in antiviral defenseduring infection. Loss of AMPK also led to dephosphorylation atSer-555 of the known STING regulator, UNC-51-like kinase 1(ULK1). However, ULK1-deficient cells responded normally toDMXAA, indicating that AMPK promotes STING-dependentsignaling independent of ULK1 in mouse cells.

Pathogen sensing by innate immune cells is essential for hostdefense in response to invading microorganisms. Recognitionby pattern recognition receptors typically activates signal trans-duction pathways, leading to up-regulation of cytokines likeTNF-� and IFN-�. Although IFN-� has a well established rolein defense against viral infection, intracellular bacterial infec-

tions also induce this cytokine. In the past decade, our under-standing of the molecular basis for these signaling pathways hasexpanded greatly and it has been discovered that there is someoverlap between proteins responsible for recognition of bacte-ria and virus alike. One particular example is the mitochondrialand endoplasmic reticulum resident protein stimulator ofinterferon genes (STING)2 (1–3). STING has been shown to becrucial for recognition of both DNA viruses (4 – 6) and intracel-lular bacterial pathogens (7–12). STING senses DNA virusesindirectly, requiring the host enzyme cyclic GMP-AMP syn-thase (cGAS) to convert the viral double stranded DNA into thecompound cGAMP (13–15). This metazoan second messenger,in turn, binds with high affinity to STING dimers, changing theprotein conformation (16), such that it leads to activation of thekinase TBK1 and subsequent phosphorylation of the transcrip-tion factor IRF3, which is pivotal for IFN-� induction (17).Mouse STING-dependent signaling can also be triggered by thebacteria-specific metabolites cyclic di-AMP (8) and cyclic di-GMP (18), albeit with lesser potency than the endogenouslyproduced cGAMP (16, 19). Although the literature suggeststhat mouse STING is essential for up-regulation of IFN-� dur-ing a number of bacterial infections, it remains unclear whetherthis is through direct sensing of bacterial-derived dinucle-otides or indirect means through cGAS-mediated produc-tion of cGAMP.

To study STING signaling in the absence of infection, ligandsthat activate the STING pathway, such as dsDNA or cGAMP,are routinely delivered into the cytosolic by transfection or dig-itonin permeabilization, adding additional variables into thestudy. A synthetic STING agonist, 5,6-dimethyl xanthenone-4-acetic acid (DMXAA), is a useful tool with which to studySTING-dependent signaling because it is cell permeable andbinds mouse, but not human, STING with rapid kinetics (20),leading to TBK1 activation, IRF3 phosphorylation, and subse-quent cytokine induction (21) that is entirely STING-depen-

* This work was supported, in whole or in part, by National Institutes of HealthGrant AI018797 (to S. N. V.) and Ruth L. Kirschstein National Research Ser-vice Award F32AI098167 (to D. P.). The authors declare that they have noconflicts of interest with the contents of this article. The content is solelythe responsibility of the authors and does not necessarily represent theofficial views of the National Institutes of Health.

1 To whom correspondence should be addressed: 685 W. Baltimore St., Rm.380, Baltimore, MD 21201. Tel.: 410-706-4838; Fax: 410-706-8607; E-mail:[email protected].

2 The abbreviations used are: STING, stimulator of interferon genes; AICAR,5-aminoimidazole-4-carboxamide ribonucleotide; AMPK, AMP-activatedkinase; CaMKK, calcium/calmodulin-dependent protein kinase kinase;cGAMP, cyclic GMP-AMP; cGAS, cyclic GMP-AMP synthase; DMXAA, 5,6-dimethyl xanthnone-4-acetic acid; IRF, interferon regulatory factor; ISG,interferon-stimulated genes; MEFs, mouse embryonic fibroblasts; MTT,methylthiazolyldiphenyl tetrazolium bromide; ULK, UNC-51-like kinase;VSV, vesicular stomatitis virus; m.o.i., multiplicity of infection; ANOVA, anal-ysis of variance; qRT, quantitative RT; BAPTA-AM, 1,2-bis(2-aminophenoxyl)-ethane-N,N,N�,N�-tetraacetic acid-acetoxy methylester.

crossmarkTHE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 292, NO. 1, pp. 292–304, January 6, 2017

© 2017 by The American Society for Biochemistry and Molecular Biology, Inc. Published in the U.S.A.

292 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 292 • NUMBER 1 • JANUARY 6, 2017

by guest on January 28, 2020http://w

ww

.jbc.org/D

ownloaded from

http://crossmark.crossref.org/dialog/?doi=10.1074/jbc.M116.763268&domain=pdf&date_stamp=2016-11-22

http://www.jbc.org/

dent (22). Thus, DMXAA mimics the effect of cGAMP inmouse cells. Recently, the small molecule 10-carboxymethyl-9-acridanone was reported to activate mouse STING in a species-specific manner similar to DMXAA (23).

Sensing of self-DNA has been linked to the developmentof autoimmunity in humans and in animal models (24, 25).Humans possessing a loss of function mutation of the cyto-solic DNase Trex1 develop the severe neurological conditionAicardi-Goutières syndrome (26), and Trex1-deficient micesuccumb to autoimmune myocarditis in an IFN-dependentfashion (27). Together, these data indicate that recognition ofcytosolic DNA by STING could be a major contributor to thedevelopment of autoimmunity, and stresses the need for tightregulation of signaling through this pathway. Interestingly, in acellular model of mitochondrial stress, IFNs and interferon-stimulated genes (ISGs) were found to be enhanced by stimu-lation with cytosolic nucleic acids (28), suggesting that themitochondria might contribute to this regulation. In additionto being a prominent source of cellular reactive oxygen speciesand ATP generation, the negatively charged mitochondrialmatrix also plays a role in homeostasis of cellular cations suchas calcium (29).

AMPK is a heterotrimeric host protein named for its centralrole in regulating host cellular metabolism through sensing ofAMP levels (30). AMPK is composed of three distinct subunits:the catalytic �-subunit, the AMP binding regulatory �-subunit,and the scaffolding �-subunit (reviewed in Ref. 31). All threeof these subunit proteins are encoded by at least two differ-ent isoforms in mammals. AMPK possesses a diverse setof substrates ranging from transcription factors to otherkinases, forming an expansive regulatory network (31).Although AMPK is regulated post-translationally at a numberof different amino acid residues, phosphorylation of the Thr-172 residue in the catalytic loop of the � subunit is required forkinase activity (32), making it the canonical marker for AMPKactivation. AMP binding to the �-subunit creates an allostericchange, allowing greater access to this Thr-172 residue to theupstream AMPK-activating protein liver kinase B1 (33–35). Inaddition to monitoring energy levels of the cell through sensingof AMP, the activity of AMPK is also controlled by intracellularcalcium levels downstream of calcium/calmodulin-dependentprotein kinase kinase 2 (CaMKK�) (36 –38).

The role of AMPK in innate immune signaling is not as wellunderstood as its role in metabolism and reports of its role inSTING signaling are conflicting. Recently, phosphorylation ofUNC-51-like kinase 1 (ULK1) by AMPK was shown to inhibitsignaling downstream of STING through phosphorylation atSer-366 on STING (39). Conversely, a different study demon-strated that phosphorylation of this same Ser-366 site plays acrucial role in activating STING-dependent signaling by facili-tating binding of IRF3 to facilitate its phosphorylation by TBK1(40). Previously, our lab identified mitochondrial membranepotential as a requirement for STING-mediated cytokineexpression (22). This suggested the possibility that host mito-chondria regulates STING, but our studies did not clarify themechanism by which this may occur. Overall, the role of AMPKin events linked to the mitochondria led us to hypothesize thatAMPK is the effector protein that enhances STING signaling in

response to host metabolic changes. The goal of this study wasto test this hypothesis so to clarify unambiguously the role ofAMPK in STING-dependent signaling.

Results

Inhibition of AMPK Disrupts STING-dependent CytokineInduction in Primary Macrophages—Innate signaling pathwayscan be regulated at many levels and we have shown previouslythat the cytokine response to the STING-activating ligandDMXAA is regulated by mitochondrial membrane potential,independent of ATP generation (22). In addition to cellularmetabolism, host mitochondria also regulate calcium homeo-stasis (29), leading us to hypothesize that calcium concentra-tions alter the response to DMXAA. In support of this asser-tion, pretreatment of primary murine macrophages with theintracellular calcium chelator BAPTA-AM dose-dependentlydecreased expression of IFN-� (Fig. 1A) and TNF-� mRNA(Fig. 1B) induced by DMXAA. Because AMPK activity is posi-tively enhanced by calcium downstream of CaMKK� (37), ourfindings suggested that AMPK might be the effector proteinthat mediates this calcium-dependent response. This hypothe-sis was initially tested by using the AMPK inhibitor CompoundC. Compound C dose dependently decreased DMXAA-in-duced IFN-� (Fig. 2A) and TNF-� (Fig. 2B) in primary macro-phages, indicating that AMPK activity is required for this sig-naling pathway.

FIGURE 1. Intracellular calcium concentration regulates STING-depen-dent cytokine induction. Peritoneal mouse macrophages were pretreatedwith the indicated dose of the intracellular calcium chelator BAPTA-AM orvehicle alone for 30 min. The cells were then treated for an additional 2 h withDMXAA (100 �g/ml). Expression of IFN-� (A) and TNF-� (B) mRNA were deter-mined by qRT-PCR. The data represents the mean fold-induction calculatedfrom 3 independent experiments. The data points represent the individualmean from each experiment. **, p � 0.01; ***, p � 0.001.

Role of AMPK in STING Signal Transduction

JANUARY 6, 2017 • VOLUME 292 • NUMBER 1 JOURNAL OF BIOLOGICAL CHEMISTRY 293


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Inhibition of Cytokine Response by Compound C Is Indepen-dent of Host Mitochondria and Is Associated with DecreasedAMPK Activation in Primary Macrophages—That the mito-chondrial dehydrogenase activity of Compound C-treatedmacrophages was not disrupted at the doses used in our studyas measured by the methylthiazolyldiphenyl tetrazolium bro-mide (MTT) assay (Fig. 3A) argues against the possibility thatthe inhibitory effect of Compound C is due to the toxicity orsome other effect on mitochondrial physiology. If Compound Cwere acting at the level of AMPK, the question then becomeswhether it is blocking basal levels or ligand-dependent activa-tion of AMPK activity. Treatment of macrophages with theAMP analog AICAR (positive control) increased AMPK phos-phorylation at Thr-172, as expected; however, neither DMXAAnor LPS stimulation of macrophages was sufficient to increaseAMPK activity in primary macrophages (Fig. 3B). This findingindicates that Compound C possibly influences basal activity ofthis kinase complex. In agreement with this possibility, additionof Compound C alone decreased basal phosphorylation ofAMPK (Fig. 3, C and D).

Mouse Embryonic Fibroblasts (MEFs) Lacking AMPK Activ-ity Exhibit Impaired STING-dependent Signaling—To corrob-orate the finding that the activity of Compound C on macro-phages correlated with suppression of AMPK activity, a genetic

approach was employed (Fig. 3). Mice deficient for both genecopies of the catalytic AMPK� subunit are embryonic lethal(46), but MEFs from these mice are fully viable (42). Like theprimary macrophages examined above, DMXAA-dependentIFN-� expression in WT MEFs was also inhibited in the pres-ence of either BAPTA-AM or Compound C (Fig. 4, A and B).In response to the STING agonist, DMXAA, AMPK�1�/�

AMPK�2�/� MEFs exhibit a significantly impaired ability toup-regulate IFN-� and TNF-� mRNA expression (Fig. 4, C andD), showing a remarkably similar phenotype to primary murinemacrophages treated with Compound C (Fig. 2). AlthoughDMXAA is an artificial inducer of STING-dependent signaling,AMPK�1�/�AMPK�2�/� MEFs also exhibit impaired induc-tion of IFN-� following transfection of the natural ligandcGAMP (Fig. 4E). Transfection of cells with cGAMP led to sig-nificantly less robust induction of IFN-� mRNA compared withDMXAA in AMPK-deficient MEFs (compare Fig. 4, E to C).Transfection of cells with the RNA helicase activating ligandpoly(I:C) (47) was also decreased in AMPK�1�/�AMPK�2�/�

MEFs (Fig. 4F), albeit to a lesser extent than seen with STINGligands (Fig. 4E).

AMPK has numerous substrate targets in the cell, so it wasinitially unclear where in the STING signaling pathway inhibi-tion was taking place. Upon STING activation by DMXAA inWT MEFs, TBK1 is rapidly activated by phosphorylation, lead-ing to recruitment and activation of the transcription factorIRF3 (21), which then translocates to the nucleus. TBK1 phos-phorylation was detectable by 15 min post-DMXAA treatmentand IRF3 phosphorylation by 30 min (Fig. 5), with both increas-ing significantly by 60 min. Both of these protein modificationswere significantly blunted in the AMPK�1�/�AMPK�2�/�

MEFs (Fig. 5). This indicates that AMPK promotes the earliestevents within the STING-dependent signaling pathway.

AMPK-deficient MEFs Exhibit Impaired VSV-dependentIFN-� Expression and Impaired Control of Viral Growth—Defective STING signaling in AMPK-deficient MEFs would bepredicted to render these cells more permissive to viruses thatare recognized in a STING-dependent fashion. VSV, a STING-dependent virus (1), was used to infect WT and AMPK�1�/�

AMPK�2�/� MEFs, and resulted in decreased induction ofIFN-� mRNA at 6 h post-infection in AMPK�-deficient MEFscompared with WT MEFs (Fig. 6A). Reduced IFN-� inductionis crucial because VSV growth in vitro is restricted in an IFN-�-dependent fashion (48). Thus, reduced IFN-� induction inAMPK-deficient MEFs would be predicted to result inincreased susceptibility to virus infection. To test this possibil-ity, we infected WT and AMPK�1�/�AMPK�2�/� MEFs andmonitored cell survival. Pretreatment of MEFs with DMXAA toinduce IFN-� protected WT MEFs, but not AMPK�1�/�

AMPK�2�/� MEFs from VSV-induced cellular cytotoxicity(Fig. 6B). In the absence of DMXAA pretreatment, there was astatistically insignificant difference in survival between WT andAMPK�1�/�AMPK�2�/� MEFs after overnight infection withVSV (Fig. 6B), but the AMPK�1�/�AMPK�2�/� MEFs exhib-ited signs of cell death as early as 8 h compared with the infectedWT cells (Fig. 6C). To confirm that this was caused by anincrease in viral replication, VSV-G protein was monitored byWestern blot analysis. By 6 h post-infection, VSV-G protein

FIGURE 2. Inhibition of AMPK activity blocks induction of STING-depen-dent cytokines. Peritoneal macrophages were treated with the indicateddoses of the AMPK inhibitor Compound C or vehicle alone for 30 min. The cellswere then treated for an additional 2 h with DMXAA (100 �g/ml). Expressionof IFN-� (A) and TNF-� (B) mRNA were determined by qRT-PCR. The datarepresents the mean fold-induction calculated from 3 independent experi-ments. The data points represent the individual mean from each experiment.NS, not significantly different. *, p � 0.05; **, p � 0.01; ***, p � 0.001.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

was greatly increased in AMPK�1�/�AMPK�2�/� MEFs cellscompared with WT MEFs (Fig. 6D). To verify that this pheno-type was not due to impaired IFN-�-mediated signaling intrin-sic to the AMPK-deficient cells, MEFs were treated withexogenous IFN-� in the absence of infection. AMPK�1�/�

AMPK�2�/� MEFs exhibited decreased basal expression of theISGs IRF7 and ISG56 compared with WT MEFs (Fig. 7, A andB), but showed similar, but somewhat delayed up-regulation ofIRF7 and ISG56 gene expression in response to exogenousIFN-� (Fig. 7, C and D). Consistent with this result, pre-treat-ment of cells with IFN-� protected both WT and AMPK�1�/�

AMPK�2�/� MEFs from VSV-induced cell cytotoxicity (Fig.7E). In toto, our data indicate that in AMPK�1�/�AMPK�2�/�

MEFs, the STING signaling defect causes impaired IFN induc-tion leading to loss of control of viral infection. This illustratesa crucial role for AMPK in host antiviral defense mediated bySTING signaling.

A Previously Identified Regulator of STING Signaling, ULK1,Is Constitutively Dephosphorylated in AMPK-deficient Cells—ULK1 and ELF4 have been demonstrated recently to regulateSTING-dependent signaling in opposing fashions: ELF4enhances STING-dependent signaling by binding to regulatorysites within the IFN-� promoter, thereby increasing the bind-ing affinity of IRF3 and IRF7 to interferon-sensitive responseelements within the IFN-� promoter (49). Conversely, theULK1 kinase has been reported to inhibit STING-dependentIRF3 phosphorylation by phosphorylating the Ser-366 residueof STING to prevent recruitment of IRF3 (39). Conversely,

others have reported that this same Ser-366 residue promotesSTING-dependent signaling by facilitating interaction of IRF3with its kinase TBK1 (40). Because AMPK has been shown toassociate physically with ULK1 and to regulate the activity ofULK1 through direct phosphorylation (50), we initially hypoth-esized that ULK1 mediates repression of STING-dependentsignaling in AMPK-deficient MEFs. In untreated WT MEFs,ULK1 Ser-555 was basally phosphorylated and remained sothroughout 60 min of DMXAA treatment (Fig. 8). However, inthe AMPK�1�/�AMPK�2�/� MEFs, ULK1 Ser-555 remaineddephosphorylated throughout the 60-min incubation withDMXAA (Fig. 8). In contrast to ULK1, basal ELF4 protein levelswere equivalent in the WT and AMPK-deficient MEFs (datanot shown).

The ULK1 Inhibitor SBI-0206965 Represses Activation ofSTING-dependent Signaling, Whereas ULK1�/� MEFs BehaveLike WT—Considering the conflicting role(s) of ULK1 inSTING-dependent signaling, we sought to clarify its role in ourmodel system. SBI-0206965 was recently identified as a selec-tive small molecule inhibitor of ULK1 (51). Addition of theULK1 inhibitor at a concentration used in the prior study (51)led to a decrease in IFN-� expression in WT MEFs treated withDMXAA. Thus, these data indicate that ULK1 is unlikely to bea negative regulator of STING-dependent signaling (Fig. 9A).This decrease in STING-mediated signaling was not associatedwith a corresponding impairment of mitochondrial dehydroge-nase (MTT) activity (Fig. 9B), suggesting that it was not toxic tothe cells. To decipher further the role of ULK1 in STING-de-

FIGURE 3. AMPK inhibition lowers basal and inducible activities of the AMPK signaling complex. Peritoneal macrophages were treated with the indicateddoses of the AMPK inhibitor Compound C for 2 h. At the end of this incubation, mitochondrial dehydrogenase activity (A) was quantified by using the MTT assaydescribed under “Experimental Procedures.” B, peritoneal macrophages were treated with AICAR (0.5 mM), DMXAA (100 �g/ml), LPS (100 ng/ml), or media onlyfor 30 min. C, peritoneal macrophages were pre-treated with Compound C (20 �M) for 30 min prior to stimulation with AICAR (0.5 mM) or solvent alone for anadditional 30 min. Phosphorylation of the Thr-172 amino acid residue of AMPK� was determined by an ELISA-based assay as described under “ExperimentalProcedures” in B and C. The data represents the mean � S.D. for experimental conditions performed in triplicate (A) or duplicate (B and C). A representative of3 independent experiments is shown. **, p � 0.01; ***, p � 0.001. D, peritoneal macrophages were stimulated for the indicated times with Compound C (20 �M).Lysates were separated by SDS-PAGE. After transfer, proteins were detected by Western blot analysis using antibodies specific for phospho-AMPK (Thr-172) ortotal AMPK.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

pendent signaling, WT and ULK1�/� MEFs (Fig. 10A) weretreated with two direct agonists of the STING signaling path-way. In stark contrast to the results derived using the ULK1inhibitor, ULK1-deficient MEFs were not impaired for induc-tion of IFN-� in response to cGAMP (Fig. 10B) or DMXAA(Fig. 10C). ULK1�/� ULK2�/� MEFs also responded like WTMEFs after treatment with DMXAA or cGAMP (data notshown). To investigate this discrepancy further, STING-depen-dent signaling was induced by DMXAA treatment of WT andULK1�/� MEFs in the absence or presence of increasing dosesof the ULK1 inhibitor. Both the WT and ULK1�/� MEFs exhib-ited similar dose-dependent decreases in IFN-� induction inresponse to the ULK1 inhibitor (Fig. 10D). This indicates that

SBI-0206965 exerts an off-target effect in the case of STING-dependent signaling. Overall, our data indicate that AMPKplays a crucial role in the regulation of STING-dependent sig-naling (Fig. 11), but STING signaling is independent of thekinase ULK1.

Discussion

AMPK has been previously recognized as being crucial formaintaining metabolic homeostasis of cells. More recently,however, it was also proposed to be involved in immunity toviruses (52, 53). In these more recent studies, the effect ofAMPK on metabolism contributed directly to the antiviraleffect. However, in our study, we have identified a potentially

FIGURE 4. Cells deficient for AMPK signaling exhibit impaired cytokine induction mediated through cytosolic pathogen recognition receptors. A, WTMEFs were treated with the indicated doses of the intracellular calcium chelator BAPTA-AM or solvent alone for 30 min. The cells were then treated for anadditional 2 h with DMXAA (100 �g/ml). B, cells were treated as in A except with the AMPK inhibitor Compound C or solvent alone prior to DMXAA treatment.Expression of IFN-� (A and B) was determined by qRT-PCR. C and D, WT or AMPK�1�/�AMPK�2�/� MEFs were stimulated for 2 h with DMXAA (100 �g/ml).Expression of IFN-� (C) and TNF-� (D) mRNA were determined by qRT-PCR. E, WT and AMPK�1�/�AMPK�2�/� MEFs were transfected with 1 �g of cGAMP or1 �g of poly(I:C) and incubated for 6 h. Expression of IFN-� mRNA (E and F) was determined by qRT-PCR. For A–E, the data represents the mean fold-inductionfrom 2 or 3 independent experiments. *, p � 0.05; **, p � 0.01; ***, p � 0.001.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

new role for AMPK in the host antiviral defense that involvesregulation of host innate immune signaling pathways that leadto up-regulation of IFN-�.

The involvement of AMPK in innate immune signaling hasbeen studied less extensively than its role in metabolism, andthe results have been conflicting. AMPK activators such asAICAR, metformin, and resveratrol, have been linked to anti-inflammatory activities through inhibition of NF-�B signaling(54 –56) and AMPK activation causes up-regulation of the neg-ative regulatory protein Small Heterodimer Partner (alsoknown as NR0B2) (57). From a clinical standpoint, AICAR maybe able to mitigate allergy or other lung injuries by reducingairway inflammation (58, 59) and it has also been shown toattenuate the severity of experimental autoimmune encephali-tis in mouse models of multiple sclerosis (60, 61). Yet, despitethese studies, the role of AMPK in immunity is not exclusivelyanti-inflammatory. Addition of the AMPK inhibitor Com-pound C suppressed IFN-� induction in response to cytosolicnucleic acids (39) and IL-6 secretion in response to adiponectin(62), showing that AMPK can also induce a pro-inflammatoryresponse in certain instances.

Some of these discrepancies may be attributable to off-targeteffects of the AMPK activator AICAR and the AMPK inhibitorCompound C for AMPK. There are published examples ofAMPK-independent roles for both of these compounds (63,64), stressing the importance of using genetic models to corrob-orate pharmacologic data. Even after taking this caveat intoaccount, the genetic evidence for the role of AMPK is alsonot always clear-cut, sometimes even when the samepathways are being scrutinized. Macrophages lacking theAMPK�1 subunit have been shown to exhibit an increasedresponse to LPS (65). Yet, human monocytes in whichAMPK�1 was knocked down exhibit decreased cytokine induc-tion of NF-�B-dependent genes when treated with LPS (66).Additionally, mice and macrophages that lack one of the acti-vating kinases of AMPK, CaMKK�, have a diminished responseto LPS (67).

The focus of our study was to pinpoint the role of AMPK inthe STING signaling network. It was previously demonstratedthat the AMPK inhibitor Compound C inhibited cytokine

induction in response to transfected dsDNA (39), but therewas no corroboration of this finding using a genetic model.In our study, we have shown that perturbations to the AMPKpathway by chelation of calcium, treatment with the AMPKinhibitor Compound C, or absence of AMPK signaling(using AMPK�1�/�AMPK�2�/� MEFs) all lead to a failure toinduce cytokine gene expression in response to DMXAA treat-ment (Fig. 11). This implicates decreased signaling immediatelydownstream of STING.

ULK1 and ELF4 were recently implicated in the regulationof STING-dependent signaling (39, 49). As AMPK has beenshown to interact with ULK1 (50), we hypothesized that, inthe absence of AMPK, ULK1 would repress STING-depen-dent signaling. In AMPK-deficient cells, ULK1 was constitu-tively dephosphorylated at Ser-555, consistent with theliterature that AMPK regulates ULK1 activity through phos-phorylation (69 –71).

ULK1 possesses at least 16 phosphorylation sites and is thetarget of multiple kinases (72). Of these different phosphoryla-tion sites, Ser-555 seems to be important for mitochondrialhomeostasis (50). AMPK can also phosphorylate ULK1 at Ser-317 and Ser-777 (73). A S555A mutation in ULK1 preventsrecruitment of ULK1 to mitochondria during hypoxic condi-tions, whereas the S555D mutant that mimics constitutive phos-phorylation can traffic to the mitochondria even in the absenceof AMPK (74). Most importantly, equivalent mutations inhuman ULK1 (at Ser-556) had a disparate effect on STING-de-pendent signaling: S556A ULK1 inhibited STING-dependentsignaling, whereas S556D ULK1 had no inhibitory effect (39).From these results, the authors reasoned that ULK1 inhibitedSTING-dependent signaling through direct phosphorylation ofSTING on Ser-366, whereas concluding that AMPK-depen-dent phosphorylation of ULK1 repressed its kinase activity (39).This runs counter to the role AMPK exerts on ULK1 kinaseactivity during autophagy and conflicts with a recent reportshowing that phosphorylation of STING on Ser-366 helps acti-vate STING-dependent signaling (40). An alternative interpre-tation is that the S556A mutation in ULK1 prevents its associ-ation with STING, preventing a phosphorylation eventrequired for STING activation from occurring. Given that theabsence of the autophagic protein ATG9a has been shown tocause excessive activation of STING-dependent signaling (75),it could be theorized that ULK1 initially activates STING-de-pendent signaling (possibly through direct phosphorylation),but enforces a short window of signal transduction by promot-ing the eventual degradation of STING protein throughautophagy. Our data involving the ULK1 inhibitor initially sup-ported this possibility, but this drug appears to exert an off-target effect in our system. Although the inhibitor-treated cellsdo appear healthy, as monitored by mitochondrial activity, it ispossible that there is a toxic effect that is not reflected withinthe first 2 h of treatment by our viability assay. Using a geneticmodel of ULK1 deletion, we showed that ULK1 does not play amajor regulatory role in STING-dependent signaling, contraryto previously published findings (39). In that study, ULK1knockdown was performed in a human fibroblast cell line, withULK1 knockdown cells exhibiting a modest 2-fold increase inIFN-� protein secretion in response to dsDNA (39). This dis-

FIGURE 5. Cells deficient for AMPK signaling exhibit impaired STING-de-pendent signaling. WT or AMPK�1�/�AMPK�2�/� MEFs were stimulatedfor the indicated times with DMXAA (100 �g/ml). Lysates were separated bySDS-PAGE. After transfer, proteins were detected by Western blot analysisusing antibodies specific for phospho-TBK1, total TBK1, phospho-IRF3, IRF3,and �-Tubulin. A representative of 3 independent experiments is shown.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

crepancy is not easily remedied, but it should be noted that thesignaling stimulus used in that study (39) was done by transfec-tion of dsDNA (14). Considering that STING-dependent sig-naling has already exhibited a species-specific differencerelated to agonist specificity (20), it is theoretically possible thatULK1 may only play a major regulatory role in human cells orthat its role is limited in some way to the secretory pathway. Inour study, we have shown that perturbations to the AMPKpathway lead to decreased STING-dependent signaling andthat AMPK regulates the phosphorylation of ULK1, but weconclude from our data that the role of AMPK in activation ofthe STING signaling pathway is independent of ULK1 in mousecells. It remains to be determined whether AMPK promotesactivation of this signaling pathway through direct or indirectmeans (Fig. 11).

We have shown that the lack of AMPK has a drastic effect onhost defense, increasing susceptibility to infection with VSV,whose control during in vitro infection has been shown torequire both STING (1) and IFN-� (48). RNA viruses like VSVare also thought to be primarily recognized directly by RNAhelicases (76, 77), independently of STING. This likely meansthat the role of STING in recognition of certain RNA viruses isindirect. Constitutive, low level activation of STING signalingthrough host or mitochondrial DNA could be required formaintaining expression of the signal transduction machinerythat contributes to activation of many pathways, includingcytosolic RNA helicases. We predict that this viral susceptibil-

ity in the absence of AMPK would be extrapolated to otherviruses that signal directly through STING such as HSV-1 (5),adenovirus (6), HIV (4), and Kaposi’s sarcoma-associated her-pesvirus (78). This pathway might even be a target for virusestrying to circumvent the innate immune system. In line withthis assertion, the Epstein-Barr virus-encoded protein latentmembrane protein 1 inhibits liver kinase B1 and inactivates theAMPK signaling network (68). In the future, it will be interest-ing to study whether viruses may exploit this pathway by mod-ulating the activity of AMPK to enable their growth or to evadethe innate immune system.

Experimental Procedures

Antibodies and Reagents—Compound C (BML-EI369-0005)and BAPTA-AM (BML-CA411-0025) were purchased fromEnzo Life Sciences (Farmingdale, NY), dissolved in DMSO, andstored in small volume aliquots. The ULK1 inhibitor SBI-0206965 (number 18477) was purchased from Cayman Chem-ical Company (Ann Arbor, MI), dissolved in DMSO, and storedin small volume aliquots. Recombinant mouse IFN-� (number12405-1) was purchased from PBL Interferon Source (Piscat-away, NJ). DMXAA (catalog number D5817) and MTT (num-ber 2128) was purchased from Sigma. Protein-free Escherichiacoli K235 LPS was isolated by methods previously described(41). The cytosolic immune receptor activators cGAMP (num-ber tlrl-cga) and poly(I:C)/LyoVec (number tlrl-picwlv) werepurchased from InvivoGen (San Diego, CA). The AMPK acti-

FIGURE 6. Decreased IFN-� response to VSV infection increases susceptibility of AMPK-deficient cells. WT and AMPK�1�/�AMPK�2�/� MEFs wereinfected with VSV (m.o.i. � 1) for 6 h. Prior to onset of cell death at 6 h post-infection, IFN-� mRNA (A) expression was quantified by qRT-PCR. For A, the datarepresents the mean fold-induction from 3 independent experiments. B, WT and AMPK�1�/�AMPK�2�/� MEFs were pretreated with DMXAA (100 �g/ml) orvehicle only for 6 h. Immediately thereafter, cells were infected with VSV (m.o.i. � 10) for 18 h. Viability was determined using the MTT assay and normalized tomock-infected cells to give a cell survival index. C, WT and AMPK�1�/�AMPK�2�/� MEFs were infected with the indicated m.o.i. of VSV. At 8 h post-infection,viability was calculated as in B. For B and C, the data represents the mean � S.D. for experimental conditions performed in duplicate in the same experiment.*, p � 0.05; **, p � 0.01. D, WT MEFs and AMPK�1�/�AMPK�2�/� MEFs were infected with VSV (m.o.i. � 1). At the indicated time points, cell lysates wereharvested and separated by SDS-PAGE. After transfer, proteins were detected by Western blot analysis using antibodies specific for VSV G protein or �-tubulin.A representative of 3 independent experiments is shown.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

vator AICAR and the following antibodies were purchasedfrom Cell Signaling Technology (Danvers, MA): anti-�-Tubu-lin (number 2146), anti-phospho-AMPK� (Thr-172) (number2535), anti-AMPK� (number 5832), anti-IRF-3 (number 4302),anti-phospho-IRF-3 (number 4947), anti-TBK1 (3504), anti-

phospho-TBK1 (5483), anti-ULK1 (number 8054), and anti-phospho-ULK1 (Ser-555) (number 5869). Anti-VSV-G anti-body (number sc-66180) was purchased from Santa CruzBiotechnology, Inc. (Dallas, TX). All primary antibodies wereused at 1:1000 dilution.

FIGURE 7. Cells deficient for AMPK signaling exhibit normal responses to exogenous type I IFN. Expression of the ISGs IRF7 (A) and ISG56 (B) mRNA inuntreated WT MEFs or AMPK�1�/�AMPK�2�/� MEFs were examined. Each data point represents the pooled results from 3 independent experiments.Induction of the ISGs IRF7 (C) and ISG56 (D) in WT or AMPK�1�/�AMPK�2�/� MEFs were monitored for 6 h following stimulation with IFN-� (100 units/ml). E,WT and AMPK�1�/�AMPK�2�/� MEFs were pretreated with IFN-� (100 units/ml) or vehicle only for 6 h. Immediately thereafter, cells were infected with VSVat the indicated m.o.i. for 18 h. Viability was calculated using the MTT assay. The data represents the mean � S.D. for experimental conditions performed intriplicate for panels C–E and a representative of 3 independent experiments is shown. *, p � 0.0016; **, p � 0.0001.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Primary Macrophage Isolation and in Vitro Cell Culture—The Institutional Animal Care and Use Committee at the Uni-versity of Maryland, Baltimore, MD, approved the protocolsdescribed below using animals. Primary peritoneal macro-phages were isolated and cultured as previously described (22).Briefly, 4 days after i.p. injection of C57BL/6J mice (JacksonLaboratories, Bar Harbor, ME) with 2 ml of 3% sterile fluidthioglycollate (catalog number R064710, Remel Products,

Lenexa, KS), cells were isolated by peritoneal lavage with freshculture medium (RPMI 1640 supplemented with 10% FBS, 10mM HEPES buffer, 1% nonessential amino acids, 2 mM L-gluta-mine, 1 mM sodium pyruvate, and 1% penicillin/streptomycin).After 2 h of culture, non-adherent cells were removed by aspi-ration and plate-bound macrophages were replenished withfresh culture medium for overnight incubation. Experimentswere performed the day after isolation. Macrophages weretreated with DMXAA at a concentration of 100 �g/ml, whichwas previously demonstrated to be the optimal dose for exam-ining up-regulation of IFN-� (21). Transfection of cGAMP wasaccomplished using the reagent SuperFect (Qiagen, Valencia,CA) according to the manufacturer’s instructions. In brief, 4 �gof cGAMP was incubated with 10 �l of reagent in serum-freemedia for 10 min at room temperature. One �g of this mastermixture was added to each recipient well. Poly(I:C) conjugatedto LyoVec as supplied facilitated its transfection without theuse of any additional reagent. In both cases, treated cellswere harvested 6 h after transfection for expression analysis.AMPK�1�/�AMPK�2�/� MEFs, which lack both isoforms ofthe genes that encode the catalytic �-subunit of AMPK, havebeen described previously and were kindly provided by Dr. SaraCherry (U. Pennsylvania) (42, 43). ULK1�/�, ULK1�/�, andULK1�/� ULK2�/� MEFs were kindly provided by Dr. SharonTooze (44). MEFs were routinely cultured in DMEM contain-ing 5% FBS. Cells were plated in tissue culture dishes the nightprior to the experiment.

Western Blotting and AMPK ELISA—Cell lysates were pre-pared in cell lysis buffer (Cell Signaling Technology) and frozenin single use aliquots at �70 °C. For the ELISA-based assay todetect phosphorylated AMPK (Cell Signaling Technology), cell

FIGURE 8. Cells deficient for AMPK signaling have constitutively dephos-phorylated ULK1. WT and AMPK�1�/�AMPK�2�/� MEFs were stimulatedfor the indicated times with DMXAA (100 �g/ml). Lysates were separated bySDS-PAGE. After transfer, proteins were detected by Western blot analysisusing antibodies specific for phospho-ULK1 (Ser-555), total ULK1, and �-Tu-bulin. A representative of 3 independent experiments is shown.

FIGURE 9. The ULK1 inhibitor SBI-0206965 represses activation of STING-dependent signaling. WT MEFs were treated for 30 min with the ULK1 inhib-itor SBI-0206965 (50 �M) or vehicle alone prior to treatment with DMXAA (100�g/ml) for 2 h. Expression of IFN-� mRNA (A) was determined by qRT-PCR. Thedata in A represents the mean fold-induction calculated from 3 independentexperiments. The data points represent the individual means from eachexperiment. B, WT MEFs were treated with the ULK1 inhibitor SBI-0206965 (50�M) or vehicle alone or 2 h. At the end of this incubation, mitochondrial de-hydrogenase activity was quantified by using the MTT assay as describedunder “Experimental Procedures.” The data in B represents the mean � S.D.for experimental conditions performed in triplicate.

FIGURE 10. Absence of ULK1 does not impair STING-dependent signaling.A, WT and ULK1�/� MEFs were genotyped by probing cellular lysates withantibodies specific for ULK1 and �-tubulin. B, WT and ULK1�/� MEFs werestimulated by transfecting 1 �g of cGAMP into the cells and incubated for 6 h.C, WT and ULK1�/� MEFs were stimulated for 2 h with DMXAA (100 �g/ml). D,cells were treated as in C, but were also pretreated for 30 min with the indi-cated dose of SBI-0206965. In B–D, expression of IFN-� was determined byqRT-PCR. The data in B–D, represents the mean fold-induction calculatedfrom 3 independent experiments.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

lysates were diluted 1:10 with sample diluent and added to wellscontaining antibodies specific for AMPK� subunit. The detec-tion antibody was specific for AMPK� phosphorylated at Thr-172. The secondary antibody was conjugated to horseradishperoxidase and the amount of target protein was monitoredcolorimetrically. Prior to immunoblotting, lysates weremixed with an equal volume of 2� Laemmli sample buffer(Bio-Rad) and heated to 95 °C for 5 min. Samples were sep-arated by SDS-PAGE in a continuous 10% acrylamide pre-cast gel (Bio-Rad). Proteins were transferred from the gel toa nitrocellulose membrane in transfer buffer for 1 h at 4 °C.Afterward, the membrane was washed with TBS and blockedwith TBS containing 0.1% (v/v) Tween 20 buffer and 5%(w/v) blotting grade blocker.

RNA Purification and qRT-PCR—RNA was isolated as previ-ously described (22) using the TriPure reagent (Roche AppliedScience). The purified RNA was then treated with RNase-freeDNase (Promega, Madison, WI) at 37 °C for 30 min to removeDNA contamination and reverse transcribed using the Super-Script III first strand synthesis system for real-time PCR (Invit-rogen) according to the manufacturer’s supplied instructionswith oligo(dT) and random hexamers to prime the reaction. To

analyze expression, quantitative PCR were carried out in the7900HT Fast Real-time PCR system (Applied Biosystems)using the 2� Power SYBR Green PCR master mix (LifeTechnologies Corp.), and the following primer sequences:Hprt sense: 5�-GCTGACCTGCTGGATTACATTAA-3�; Hprtantisense: 5�-TGATCATTACAGTAGCTCTTCAGTCT-3�;Ifnb1 sense: 5�-CACTTGAAGAGCTATTACTGGAGGG-3�,Ifnb1 antisense: 5�-CTCGGACCACCATCCAGG-3�; Tnfsense: 5�-GACCCTCACACTCAGATCATCTTCT-3�; Tnfantisense: 5�-CCACTTGGTGGTTTGCTACGA-3�. The totalamount of mRNA present was calculated using the compara-tive Ct method (45) and expressed as fold-induction comparedwith untreated or uninfected cells.

MTT Assay—A colorimetric assay to measure mitochondrialsuccinic dehydrogenase activity, a surrogate measure for celldeath, was performed as previously described (22). Briefly, atthe end of stimulation with the AMPK inhibitor or at timepoints following viral infection, culture supernatants wereremoved and replaced with a filter-sterilized solution ofMTT (1 mg/ml) dissolved in phosphate-buffered saline(number 21-040-CV, Mediatech, Inc., Herndon, VA). Themacrophage cultures were allowed to incubate for 30 min at37 °C and 5% CO2 in the presence of MTT. The MTT solu-tion was removed and replaced with isopropyl alcohol todissolve the water-insoluble formazan products (if mito-chondrial enzymes were active), yielding a purple solutionwhose absorbance was quantified at 595 nm on a Biotekmicroplate reader.

ViralInfections—Wild-type(WT)andAMPK�1�/�AMPK�2�/�

MEFs were infected with VSV (Indiana strain; m.o.i. � 1 or 10)for 1 h in serum-free DMEM. At the end of this incubation, themedia was removed by aspiration and the cells washed threetimes with PBS before being cultured further in growthmedium that contained 5% FBS.

Statistical Analysis—Experiments were performed in dupli-cate for AMPK phosphorylation and MTT assays. The numer-ical values obtained from these assays were analyzed for signif-icance using GraphPad Prism 4 (GraphPad Software Inc., LaJolla, CA) using either a one-way ANOVA or a two-wayANOVA depending on the number of experimental variables.Significance was accepted if p � 0.05. For the one-wayANOVA, if the data met this standard, a Tukey’s post hoc testwas used to test for significance between experimental groupsin the data set. When only two groups were being compared, anunpaired, two-tailed Student’s t test was performed to deter-mine significance. Significance for gene expression was ana-lyzed from the fold-induction values calculated from two orthree independent experiments. The resulting p values betweengroups are listed in the figure legends for each of the individualexperiments.

Author Contributions—D. P. designed and coordinated the study,performed and analyzed the experiments, and wrote the manuscript.D. J. P. was integral for studies involving VSV for Figs. 5 and 6 anddesign of experiments for Fig. 6. S. N. V. contributed to the designand conception of experiments, analysis of the data, and in draftingof the manuscript. All authors reviewed the results and approved thefinal version of the manuscript.

FIGURE 11. Multiple regulatory pathways are utilized by the cell to con-trol activation of the STING signaling pathway. A, our data support amodel in which the presence of AMPK signaling is required for optimal acti-vation of STING-dependent signaling. Conditions that interfere with AMPKsignaling, including chelation of intracellular calcium (1), pharmacologic inhi-bition of AMPK signaling (2), or genetic ablation of AMPK signaling (3) inhibitSTING-dependent signaling. As ULK1�/� and ULK1�/�ULK2�/� MEFs exhibitno defect in response to STING agonists, we conclude that the regulatory roleof AMPK on STING is independent of ULK1 and that the ULK1 inhibitor SBI-0206965 has off-target effects leading to signal impairment. B, the exact roleof AMPK in the regulation of STING is unknown at present. AMPK mightdirectly promote STING signaling by substrate level phosphorylation (1).Alternatively, it might act indirectly by activating a positive regulator (PR) orinhibiting a negative regulator (NR) of STING-dependent signaling (2). Analternate possibility is that the STING signaling pathway operates efficientlyin a narrow metabolic window and is indirectly inhibited in the absence of thecatabolic pathways promoted by AMPK (3). Future studies will delineatewhich of these mechanisms is operative.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Acknowledgments—We thank Dr. Sara Cherry for graciously provid-ing the AMPK�1�/�AMPK�2�/� MEFs and also Dr. Sharon Toozefor graciously providing the ULK1�/� and ULK1�/�ULK2�/� MEFsthat were used in our study.

References1. Ishikawa, H., and Barber, G. N. (2008) STING is an endoplasmic reticulum

adaptor that facilitates innate immune signalling. Nature 455, 674 – 6782. Zhong, B., Yang, Y., Li, S., Wang, Y. Y., Li, Y., Diao, F., Lei, C., He, X.,

Zhang, L., Tien, P., and Shu, H. B. (2008) The adaptor protein MITA linksvirus-sensing receptors to IRF3 transcription factor activation. Immunity29, 538 –550

3. Sun, W., Li, Y., Chen, L., Chen, H., You, F., Zhou, X., Zhou, Y., Zhai, Z.,Chen, D., and Jiang, Z. (2009) ERIS, an endoplasmic reticulum IFN stim-ulator, activates innate immune signaling through dimerization. Proc.Natl. Acad. Sci. U.S.A. 106, 8653– 8658

4. Gao, D., Wu, J., Wu, Y. T., Du, F., Aroh, C., Yan, N., Sun, L., and Chen, Z. J.(2013) Cyclic GMP-AMP synthase is an innate immune sensor of HIV andother retroviruses. Science 341, 903–906

5. Ishikawa, H., Ma, Z., and Barber, G. N. (2009) STING regulates intracel-lular DNA-mediated, type I interferon-dependent innate immunity. Na-ture 461, 788 –792

6. Lam, E., Stein, S., and Falck-Pedersen, E. (2014) Adenovirus detection bythe cGAS/STING/TBK1 DNA sensing cascade. J. Virol. 88, 974 –981

7. Prantner, D., Darville, T., and Nagarajan, U. M. (2010) Stimulator of IFNgene is critical for induction of IFN-� during Chlamydia muridarum in-fection. J. Immunol. 184, 2551–2560

8. Sauer, J. D., Sotelo-Troha, K., von Moltke, J., Monroe, K. M., Rae, C. S.,Brubaker, S. W., Hyodo, M., Hayakawa, Y., Woodward, J. J., Portnoy, D. A.,and Vance, R. E. (2011) The N-ethyl-N-nitrosourea-induced Goldenticketmouse mutant reveals an essential function of Sting in the in vivo inter-feron response to Listeria monocytogenes and cyclic dinucleotides. Infect.Immun. 79, 688 – 694

9. Jin, L., Hill, K. K., Filak, H., Mogan, J., Knowles, H., Zhang, B., Perraud,A. L., Cambier, J. C., and Lenz, L. L. (2011) MPYS is required for IFNresponse factor 3 activation and type I IFN production in the response ofcultured phagocytes to bacterial second messengers cyclic-di-AMP andcyclic-di-GMP. J. Immunol. 187, 2595–2601

10. Jones, J. W., Kayagaki, N., Broz, P., Henry, T., Newton, K., O’Rourke, K.,Chan, S., Dong, J., Qu, Y., Roose-Girma, M., Dixit, V. M., and Monack,D. M. (2010) Absent in melanoma 2 is required for innate immunerecognition of Francisella tularensis. Proc. Natl. Acad. Sci. U.S.A. 107,9771–9776

11. de Almeida, L. A., Carvalho, N. B., Oliveira, F. S., Lacerda, T. L., Vascon-celos, A. C., Nogueira, L., Bafica, A., Silva, A. M., and Oliveira, S. C. (2011)MyD88 and STING signaling pathways are required for IRF3-mediatedIFN-� induction in response to Brucella abortus infection. PLoS ONE 6,e23135

12. Manzanillo, P. S., Shiloh, M. U., Portnoy, D. A., and Cox, J. S. (2012)Mycobacterium tuberculosis activates the DNA-dependent cytosolic sur-veillance pathway within macrophages. Cell Host Microbe 11, 469 – 480

13. Ablasser, A., Goldeck, M., Cavlar, T., Deimling, T., Witte, G., Röhl, I.,Hopfner, K. P., Ludwig, J., and Hornung, V. (2013) cGAS produces a 2�-5�-linked cyclic dinucleotide second messenger that activates STING. Na-ture 498, 380 –384

14. Sun, L., Wu, J., Du, F., Chen, X., and Chen, Z. J. (2013) Cyclic GMP-AMPsynthase is a cytosolic DNA sensor that activates the type I interferonpathway. Science 339, 786 –791

15. Diner, E. J., Burdette, D. L., Wilson, S. C., Monroe, K. M., Kellenberger,C. A., Hyodo, M., Hayakawa, Y., Hammond, M. C., and Vance, R. E. (2013)The innate immune DNA sensor cGAS produces a noncanonical cyclicdinucleotide that activates human STING. Cell Rep. 3, 1355–1361

16. Gao, P., Ascano, M., Zillinger, T., Wang, W., Dai, P., Serganov, A. A.,Gaffney, B. L., Shuman, S., Jones, R. A., Deng, L., Hartmann, G., Barchet,W., Tuschl, T., and Patel, D. J. (2013) Structure-function analysis of

STING activation by c[G(2�,5�)pA(3�,5�)p] and targeting by antiviralDMXAA. Cell 154, 748 –762

17. Wathelet, M. G., Lin, C. H., Parekh, B. S., Ronco, L. V., Howley, P. M., andManiatis, T. (1998) Virus infection induces the assembly of coordinatelyactivated transcription factors on the IFN-� enhancer in vivo. Mol. Cell 1,507–518

18. Burdette, D. L., Monroe, K. M., Sotelo-Troha, K., Iwig, J. S., Eckert, B.,Hyodo, M., Hayakawa, Y., and Vance, R. E. (2011) STING is a direct innateimmune sensor of cyclic di-GMP. Nature 478, 515–518

19. Wu, J., Sun, L., Chen, X., Du, F., Shi, H., Chen, C., and Chen, Z. J. (2013)Cyclic GMP-AMP is an endogenous second messenger in innate immunesignaling by cytosolic DNA. Science 339, 826 – 830

20. Conlon, J., Burdette, D. L., Sharma, S., Bhat, N., Thompson, M., Jiang, Z.,Rathinam, V. A., Monks, B., Jin, T., Xiao, T. S., Vogel, S. N., Vance, R. E.,and Fitzgerald, K. A. (2013) Mouse, but not human STING, binds andsignals in response to the vascular disrupting agent 5,6-dimethylxanthe-none-4-acetic acid. J. Immunol. 190, 5216 –5225

21. Roberts, Z. J., Goutagny, N., Perera, P. Y., Kato, H., Kumar, H., Kawai, T.,Akira, S., Savan, R., van Echo, D., Fitzgerald, K. A., Young, H. A., Ching,L. M., and Vogel, S. N. (2007) The chemotherapeutic agent DMXAApotently and specifically activates the TBK1-IRF-3 signaling axis. J. Exp.Med. 204, 1559 –1569

22. Prantner, D., Perkins, D. J., Lai, W., Williams, M. S., Sharma, S., Fitzgerald,K. A., and Vogel, S. N. (2012) 5,6-Dimethylxanthenone-4-acetic acid(DMXAA) activates stimulator of interferon gene (STING)-dependentinnate immune pathways and is regulated by mitochondrial membranepotential. J. Biol. Chem. 287, 39776 –39788

23. Cavlar, T., Deimling, T., Ablasser, A., Hopfner, K. P., and Hornung, V.(2013) Species-specific detection of the antiviral small-molecule com-pound CMA by STING. EMBO J. 32, 1440 –1450

24. Yasutomo, K., Horiuchi, T., Kagami, S., Tsukamoto, H., Hashimura, C.,Urushihara, M., and Kuroda, Y. (2001) Mutation of DNASE1 in peoplewith systemic lupus erythematosus. Nat. Genet. 28, 313–314

25. Napirei, M., Karsunky, H., Zevnik, B., Stephan, H., Mannherz, H. G., andMöröy, T. (2000) Features of systemic lupus erythematosus in DNase1-deficient mice. Nat. Genet. 25, 177–181

26. Crow, Y. J., Hayward, B. E., Parmar, R., Robins, P., Leitch, A., Ali, M., Black,D. N., van Bokhoven, H., Brunner, H. G., Hamel, B. C., Corry, P. C., Cowan,F. M., Frints, S. G., Klepper, J., Livingston, J. H., et al. (2006) Mutations inthe gene encoding the 3�-5� DNA exonuclease TREX1 cause Aicardi-Goutieres syndrome at the AGS1 locus. Nat. Genet. 38, 917–920

27. Stetson, D. B., Ko, J. S., Heidmann, T., and Medzhitov, R. (2008) Trex1prevents cell-intrinsic initiation of autoimmunity. Cell 134, 587–598

28. West, A. P., Khoury-Hanold, W., Staron, M., Tal, M. C., Pineda, C. M.,Lang, S. M., Bestwick, M., Duguay, B. A., Raimundo, N., MacDuff, D. A.,Kaech, S. M., Smiley, J. R., Means, R. E., Iwasaki, A., and Shadel, G. S.(2015) Mitochondrial DNA stress primes the antiviral innate immuneresponse. Nature 520, 553–557

29. Williams, G. S., Boyman, L., Chikando, A. C., Khairallah, R. J., and Lederer,W. J. (2013) Mitochondrial calcium uptake. Proc. Natl. Acad. Sci. U.S.A.110, 10479 –10486

30. Yeh, L. A., Lee, K. H., and Kim, K. H. (1980) Regulation of rat liver acetyl-CoA carboxylase: regulation of phosphorylation and inactivation ofacetyl-CoA carboxylase by the adenylate energy charge. J. Biol. Chem. 255,2308 –2314

31. Hardie, D. G. (2007) AMP-activated/SNF1 protein kinases: conservedguardians of cellular energy. Nat. Rev. Mol. Cell Biol. 8, 774 –785

32. Stein, S. C., Woods, A., Jones, N. A., Davison, M. D., and Carling, D. (2000)The regulation of AMP-activated protein kinase by phosphorylation.Biochem. J. 345, 437– 443

33. Hawley, S. A., Boudeau, J., Reid, J. L., Mustard, K. J., Udd, L., Mäkelä, T. P.,Alessi, D. R., and Hardie, D. G. (2003) Complexes between the LKB1tumor suppressor, STRAD �/� and MO25 �/� are upstream kinases inthe AMP-activated protein kinase cascade. J. Biol. 2, 28

34. Woods, A., Johnstone, S. R., Dickerson, K., Leiper, F. C., Fryer, L. G.,Neumann, D., Schlattner, U., Wallimann, T., Carlson, M., and Carling, D.(2003) LKB1 is the upstream kinase in the AMP-activated protein kinasecascade. Curr. Biol. 13, 2004 –2008




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

35. Shaw, R. J., Kosmatka, M., Bardeesy, N., Hurley, R. L., Witters, L. A.,DePinho, R. A., and Cantley, L. C. (2004) The tumor suppressor LKB1kinase directly activates AMP-activated kinase and regulates apoptosis inresponse to energy stress. Proc. Natl. Acad. Sci. U.S.A. 101, 3329 –3335

36. Hurley, R. L., Anderson, K. A., Franzone, J. M., Kemp, B. E., Means, A. R.,and Witters, L. A. (2005) The Ca2�/calmodulin-dependent protein kinasekinases are AMP-activated protein kinase kinases. J. Biol. Chem. 280,29060 –29066

37. Hawley, S. A., Pan, D. A., Mustard, K. J., Ross, L., Bain, J., Edelman, A. M.,Frenguelli, B. G., and Hardie, D. G. (2005) Calmodulin-dependent proteinkinase kinase-� is an alternative upstream kinase for AMP-activated pro-tein kinase. Cell Metab. 2, 9 –19

38. Woods, A., Dickerson, K., Heath, R., Hong, S. P., Momcilovic, M., John-stone, S. R., Carlson, M., and Carling, D. (2005) Ca2�/calmodulin-depen-dent protein kinase kinase-� acts upstream of AMP-activated proteinkinase in mammalian cells. Cell Metab. 2, 21–33

39. Konno, H., Konno, K., and Barber, G. N. (2013) Cyclic dinucleotides trig-ger ULK1 (ATG1) phosphorylation of STING to prevent sustained innateimmune signaling. Cell 155, 688 – 698

40. Liu, S., Cai, X., Wu, J., Cong, Q., Chen, X., Li, T., Du, F., Ren, J., Wu, Y. T.,Grishin, N. V., and Chen, Z. J. (2015) Phosphorylation of innate immuneadaptor proteins MAVS, STING, and TRIF induces IRF3 activation. Sci-ence 347, aaa2630

41. McIntire, F. C., Sievert, H. W., Barlow, G. H., Finley, R. A., and Lee, A. Y.(1967) Chemical, physical, biological properties of a lipopolysaccharidefrom Escherichia coli K-235. Biochemistry 6, 2363–2372

42. Moser, T. S., Jones, R. G., Thompson, C. B., Coyne, C. B., and Cherry, S.(2010) A kinome RNAi screen identified AMPK as promoting poxvirusentry through the control of actin dynamics. PLoS Pathog. 6, e1000954

43. Laderoute, K. R., Amin, K., Calaoagan, J. M., Knapp, M., Le, T., Orduna, J.,Foretz, M., and Viollet, B. (2006) 5�-AMP-activated protein kinase(AMPK) is induced by low-oxygen and glucose deprivation conditionsfound in solid-tumor microenvironments. Mol. Cell. Biol. 26, 5336 –5347

44. McAlpine, F., Williamson, L. E., Tooze, S. A., and Chan, E. Y. (2013)Regulation of nutrient-sensitive autophagy by uncoordinated 51-like ki-nases 1 and 2. Autophagy 9, 361–373

45. Schmittgen, T. D., and Livak, K. J. (2008) Analyzing real-time PCR data bythe comparative CT method. Nat. Protoc. 3, 1101–1108

46. Viollet, B., Athea, Y., Mounier, R., Guigas, B., Zarrinpashneh, E., Horman,S., Lantier, L., Hebrard, S., Devin-Leclerc, J., Beauloye, C., Foretz, M.,Andreelli, F., Ventura-Clapier, R., and Bertrand, L. (2009) AMPK: lessonsfrom transgenic and knockout animals. Front. Biosci. (Landmark Ed) 14,19 – 44

47. Gitlin, L., Barchet, W., Gilfillan, S., Cella, M., Beutler, B., Flavell, R. A.,Diamond, M. S., and Colonna, M. (2006) Essential role of mda-5 in type IIFN responses to polyriboinosinic:polyribocytidylic acid and encephalo-myocarditis picornavirus. Proc. Natl. Acad. Sci. U.S.A. 103, 8459 – 8464

48. Shirey, K. A., Nhu, Q. M., Yim, K. C., Roberts, Z. J., Teijaro, J. R., Farber,D. L., Blanco, J. C., and Vogel, S. N. (2011) The anti-tumor agent, 5,6-dimethylxanthenone-4-acetic acid (DMXAA), induces IFN-�-mediatedantiviral activity in vitro and in vivo. J. Leukocyte Biol. 89, 351–357

49. You, F., Wang, P., Yang, L., Yang, G., Zhao, Y. O., Qian, F., Walker, W.,Sutton, R., Montgomery, R., Lin, R., Iwasaki, A., and Fikrig, E. (2013) ELF4is critical for induction of type I interferon and the host antiviral response.Nat. Immunol. 14, 1237–1246

50. Egan, D. F., Shackelford, D. B., Mihaylova, M. M., Gelino, S., Kohnz, R. A.,Mair, W., Vasquez, D. S., Joshi, A., Gwinn, D. M., Taylor, R., Asara, J. M.,Fitzpatrick, J., Dillin, A., Viollet, B., Kundu, M., Hansen, M., and Shaw, R. J.(2011) Phosphorylation of ULK1 (hATG1) by AMP-activated protein ki-nase connects energy sensing to mitophagy. Science 331, 456 – 461

51. Egan, D. F., Chun, M. G., Vamos, M., Zou, H., Rong, J., Miller, C. J., Lou,H. J., Raveendra-Panickar, D., Yang, C. C., Sheffler, D. J., Teriete, P., Asara,J. M., Turk, B. E., Cosford, N. D., and Shaw, R. J. (2015) Small moleculeinhibition of the autophagy kinase ULK1 and identification of ULK1 sub-strates. Mol. Cell 59, 285–297

52. Moser, T. S., Schieffer, D., and Cherry, S. (2012) AMP-activated kinaserestricts Rift Valley fever virus infection by inhibiting fatty acid synthesis.PLoS Pathog. 8, e1002661

53. Nakashima, K., Takeuchi, K., Chihara, K., Hotta, H., and Sada, K. (2011)Inhibition of hepatitis C virus replication through adenosine monophos-phate-activated protein kinase-dependent and -independent pathways.Microbiol. Immunol. 55, 774 –782

54. Zheng, L., Yang, W., Wu, F., Wang, C., Yu, L., Tang, L., Qiu, B., Li, Y., Guo,L., Wu, M., Feng, G., Zou, D., and Wang, H. (2013) Prognostic significanceof AMPK activation and therapeutic effects of metformin in hepatocellu-lar carcinoma. Clin. Cancer Res. 19, 5372–5380

55. Li, L., Hai, J., Li, Z., Zhang, Y., Peng, H., Li, K., and Weng, X. (2014)Resveratrol modulates autophagy and NF-�B activity in a murine modelfor treating non-alcoholic fatty liver disease. Food Chem. Toxicol. 63,166 –173

56. Katerelos, M., Mudge, S. J., Stapleton, D., Auwardt, R. B., Fraser, S. A.,Chen, C. G., Kemp, B. E., and Power, D. A. (2010) 5-Aminoimidazole-4-carboxamide ribonucleoside and AMP-activated protein kinase inhibitsignalling through NF-�B. Immunol. Cell Biol. 88, 754 –760

57. Yuk, J. M., Shin, D. M., Lee, H. M., Kim, J. J., Kim, S. W., Jin, H. S., Yang,C. S., Park, K. A., Chanda, D., Kim, D. K., Huang, S. M., Lee, S. K., Lee,C. H., Kim, J. M., Song, C. H., et al. (2011) The orphan nuclear receptorSHP acts as a negative regulator in inflammatory signaling triggered byToll-like receptors. Nat. Immunol. 12, 742–751

58. Park, C. S., Bang, B. R., Kwon, H. S., Moon, K. A., Kim, T. B., Lee, K. Y.,Moon, H. B., and Cho, Y. S. (2012) Metformin reduces airway inflamma-tion and remodeling via activation of AMP-activated protein kinase.Biochem. Pharmacol. 84, 1660 –1670

59. Park, S. J., Lee, K. S., Kim, S. R., Chae, H. J., Yoo, W. H., Kim, D. I., Jeon,M. S., and Lee, Y. C. (2012) AMPK activation reduces vascular permeabil-ity and airway inflammation by regulating HIF/VEGFA pathway in a mu-rine model of toluene diisocyanate-induced asthma. Inflamm. Res. 61,1069 –1083

60. Nath, N., Giri, S., Prasad, R., Salem, M. L., Singh, A. K., and Singh, I. (2005)5-Aminoimidazole-4-carboxamide ribonucleoside: a novel immuno-modulator with therapeutic efficacy in experimental autoimmune en-cephalomyelitis. J. Immunol. 175, 566 –574

61. Prasad, R., Giri, S., Nath, N., Singh, I., and Singh, A. K. (2006) 5-Aminoim-idazole-4-carboxamide-1-�-4-ribofuranoside attenuates experimentalautoimmune encephalomyelitis via modulation of endothelial-monocyteinteraction. J. Neurosci. Res. 84, 614 – 625

62. Tang, C. H., Chiu, Y. C., Tan, T. W., Yang, R. S., and Fu, W. M. (2007)Adiponectin enhances IL-6 production in human synovial fibroblast viaan AdipoR1 receptor, AMPK, p38, and NF-�B pathway. J. Immunol. 179,5483–5492

63. Kuo, C. L., Ho, F. M., Chang, M. Y., Prakash, E., and Lin, W. W. (2008)Inhibition of lipopolysaccharide-induced inducible nitric oxide synthaseand cyclooxygenase-2 gene expression by 5-aminoimidazole-4-carboxamide riboside is independent of AMP-activated protein kinase.J. Cell Biochem. 103, 931–940

64. Kim, Y. M., Kim, M. Y., Kim, H. J., Roh, G. S., Ko, G. H., Seo, H. G., Lee,J. H., and Chang, K. C. (2011) Compound C independent of AMPK inhib-its ICAM-1 and VCAM-1 expression in inflammatory stimulants-acti-vated endothelial cells in vitro and in vivo. Atherosclerosis 219, 57– 64

65. Carroll, K. C., Viollet, B., and Suttles, J. (2013) AMPK�1 deficiency ampli-fies proinflammatory myeloid APC activity and CD40 signaling. J. Leukoc.Biol. 94, 1113–1121

66. Kim, S. Y., Jeong, S., Jung, E., Baik, K. H., Chang, M. H., Kim, S. A., Shim,J. H., Chun, E., and Lee, K. Y. (2012) AMP-activated protein kinase-�1 asan activating kinase of TGF-�-activated kinase 1 has a key role in inflam-matory signals. Cell Death Dis. 3, e357

67. Racioppi, L., and Means, A. R. (2012) Calcium/calmodulin-dependentprotein kinase kinase 2: roles in signaling and pathophysiology. J. Biol.Chem. 287, 31658 –31665

68. Lo, A. K., Lo, K. W., Ko, C. W., Young, L. S., and Dawson, C. W. (2013)Inhibition of the LKB1-AMPK pathway by the Epstein-Barr virus-en-coded LMP1 promotes proliferation and transformation of human naso-pharyngeal epithelial cells. J. Pathol. 230, 336 –346

69. Bach, M., Larance, M., James, D. E., and Ramm, G. (2011) The serine/threonine kinase ULK1 is a target of multiple phosphorylation events.Biochem. J. 440, 283–291




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

70. Mack, H. I., Zheng, B., Asara, J. M., and Thomas, S. M. (2012) AMPK-de-pendent phosphorylation of ULK1 regulates ATG9 localization. Au-tophagy 8, 1197–1214

71. Shang, L., Chen, S., Du, F., Li, S., Zhao, L., and Wang, X. (2011) Nutrientstarvation elicits an acute autophagic response mediated by Ulk1 dephos-phorylation and its subsequent dissociation from AMPK. Proc. Natl. Acad.Sci. U.S.A. 108, 4788 – 4793

72. Dorsey, F. C., Rose, K. L., Coenen, S., Prater, S. M., Cavett, V., Cleveland,J. L., and Caldwell-Busby, J. (2009) Mapping the phosphorylation sites ofUlk1. J. Proteome Res. 8, 5253–5263

73. Kim, J., Kundu, M., Viollet, B., and Guan, K. L. (2011) AMPK and mTORregulate autophagy through direct phosphorylation of Ulk1. Nat. Cell.Biol. 13, 132–141

74. Tian, W., Li, W., Chen, Y., Yan, Z., Huang, X., Zhuang, H., Zhong, W., Chen,Y., Wu, W., Lin, C., Chen, H., Hou, X., Zhang, L., Sui, S., Zhao, B., Hu, Z., Li, L.,and Feng, D. (2015) Phosphorylation of ULK1 by AMPK regulates transloca-tion of ULK1 to mitochondria and mitophagy. FEBS Lett. 589, 1847–1854

75. Saitoh, T., Fujita, N., Hayashi, T., Takahara, K., Satoh, T., Lee, H.,Matsunaga, K., Kageyama, S., Omori, H., Noda, T., Yamamoto, N.,Kawai, T., Ishii, K., Takeuchi, O., Yoshimori, T., and Akira, S. (2009)Atg9a controls dsDNA-driven dynamic translocation of STING andthe innate immune response. Proc. Natl. Acad. Sci. U.S.A. 106,20842–20846

76. Crill, E. K., Furr-Rogers, S. R., and Marriott, I. (2015) RIG-I is required forVSV-induced cytokine production by murine glia and acts in combinationwith DAI to initiate responses to HSV-1. Glia 63, 2168 –2180

77. Kato, H., Sato, S., Yoneyama, M., Yamamoto, M., Uematsu, S., Matsui, K.,Tsujimura, T., Takeda, K., Fujita, T., Takeuchi, O., and Akira, S. (2005)Cell type-specific involvement of RIG-I in antiviral response. Immunity23, 19 –28

78. Ma, Z., Jacobs, S. R., West, J. A., Stopford, C., Zhang, Z., Davis, Z., Barber,G. N., Glaunsinger, B. A., Dittmer, D. P., and Damania, B. (2015) Modu-lation of the cGAS-STING DNA sensing pathway by gammaherpesvi-ruses. Proc. Natl. Acad. Sci. U.S.A. 112, E4306 – 4315




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Daniel Prantner, Darren J. Perkins and Stefanie N. Vogelthrough Modulation of Stimulator of Interferon Genes (STING) Signaling

AMP-activated Kinase (AMPK) Promotes Innate Immunity and Antiviral Defense

doi: 10.1074/jbc.M116.763268 originally published online November 22, 20162017, 292:292-304.J. Biol. Chem.

10.1074/jbc.M116.763268Access the most updated version of this article at doi:

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here

http://www.jbc.org/content/292/1/292.full.html#ref-list-1

This article cites 78 references, 29 of which can be accessed free at


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/lookup/doi/10.1074/jbc.M116.763268

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;292/1/292&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/292/1/292

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=292/1/292&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/292/1/292

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/292/1/292.full.html#ref-list-1

http://www.jbc.org/

Documents

AMP …AMPK 2 / MEFs, and resulted in decreased induction of IFN- mRNA at 6 h post-infection in AMPK -deficient MEFs compared with WT MEFs (Fig. 6A). Reduced IFN- induction is crucial