Structural basis for Marburg virus VP35–mediatedimmune evasion mechanismsParameshwaran Ramanana,b, Megan R. Edwardsc, Reed S. Shabmanc, Daisy W. Leunga, Ariel C. Endlich-Frazierc,Dominika M. Borekd, Zbyszek Otwinowskid, Gai Liua, Juyoung Huha, Christopher F. Baslerc,and Gaya K. Amarasinghea,1

aDepartment of Pathology and Immunology, Washington University School of Medicine, St. Louis, MO 63110; bBiochemistry Graduate Program, Roy J. CarverDepartment of Biochemistry, Biophysics and Molecular Biology, Iowa State University, Ames, IA 50011; cDepartment of Microbiology, Mount Sinai Schoolof Medicine, New York, NY 10029; and dDepartments of Biochemistry and Biophysics, University of Texas Southwestern Medical Center, Dallas,TX 75390

Edited by Robert A. Lamb, Northwestern University, Evanston, IL, and approved November 2, 2012 (received for review August 8, 2012)

Filoviruses, marburgvirus (MARV) and ebolavirus (EBOV), are causa-tive agents of highly lethal hemorrhagic fever in humans. MARVand EBOV share a common genome organization but show impor-tant differences in replication complex formation, cell entry, hosttropism, transcriptional regulation, and immune evasion. Multifunc-tional filoviral viral protein (VP) 35 proteins inhibit innate immuneresponses. Recent studies suggest double-stranded (ds)RNA seques-tration is a potential mechanism that allows EBOV VP35 to antago-nize retinoic-acid inducible gene-I (RIG-I) like receptors (RLRs) thatare activated by viral pathogen–associated molecular patterns(PAMPs), such as double-strandedness and dsRNA blunt ends. Here,we show that MARV VP35 can inhibit IFN production at multiplesteps in the signaling pathways downstream of RLRs. The crystalstructure of MARV VP35 IID in complex with 18-bp dsRNA revealsthat despite the similar protein fold as EBOV VP35 IID, MARV VP35IID interacts with the dsRNA backbone and not with blunt ends.Functional studies show that MARV VP35 can inhibit dsRNA-de-pendent RLR activation and interferon (IFN) regulatory factor 3(IRF3) phosphorylation by IFN kinases TRAF family member-associ-ated NFkb activator (TANK) binding kinase-1 (TBK-1) and IFN kBkinase e (IKKe) in cell-based studies. We also show that MARVVP35 can only inhibit RIG-I and melanoma differentiation associ-ated gene 5 (MDA5) activation by double strandedness of RNAPAMPs (coating backbone) but is unable to inhibit activation ofRLRs by dsRNA blunt ends (end capping). In contrast, EBOV VP35can inhibit activation by both PAMPs. Insights on differentialPAMP recognition and inhibition of IFN induction by a similar filo-viral VP35 fold, as shown here, reveal the structural and functionalplasticity of a highly conserved virulence factor.

type I IFN | viral immune antagonist | RNA binding protein

The Filoviridae family of viruses, which includes marburgvirus(MARV) and ebolavirus (EBOV), can cause intermittent out-

breaks that often result in high fatality rates (1). The familyconsists of five species of EBOV, Zaire, Reston, Sudan, Taï Forest,and Bundibugyo; one species of MARV; and a proposed genusCuevavirus possessing a single species Lloviu cuevavirus (2). De-spite overall similarities in genome size and organization, virionstructure, and disease characteristics (3), EBOV and MARV ex-hibit important differences, including their strategies for immuneevasion (4). For example, although EBOV viral protein (VP) 24and MARVVP40 counter IFN signaling, neither MARVVP24 norEBOV VP40 appears to function similarly to its correspondingcounterparts with regard to immune evasion (5–10).Filoviruses also counteract innate immunity through the mul-

tifunctional VP35 proteins, which perform critical roles in viralRNA synthesis, virus assembly, and virus structure (reviewed inrefs. 11 and 12). EBOV VP35 interacts with several componentsof innate antiviral defenses, including retinoic-acid inducible gene-I (RIG-I)–like receptor (RLR) pathways that lead to IFN pro-duction (13–24). These include inhibition of IFN productionthrough double-stranded (ds)RNA sequestration and inhibition of

IFN regulatory factor (IRF) 3/IRF7 phosphorylation by directassociation with kinases that activate IRF3/IRF7, IKKe/TBK-1(13, 15, 21, 25). Structural and biochemical studies on ZaireEBOV (ZEBOV) and Reston EBOV (REBOV) VP35 IFN in-hibitory domains (IID) (termed zVP35 and zIID or rVP35 andrIID, for VP35 protein and IID, respectively) in free anddsRNA-bound forms identified a number of functionally criticalbasic residues (21, 26–28). These are located in the central basicpatch (CBP) in the β-sheet subdomain and the first basic patch(FBP) in the α-helical subdomain (21, 26–28). Based on thedsRNA-bound structures of VP35 IIDs, it was suggested thatthese basic patches are important for protein–protein and protein–RNA interactions (21, 27, 28). Consistent with this, mutation ofCBP residues abrogates the dsRNA-binding and IFN-inhibitoryactivities of zVP35 and greatly attenuated virus replication inIFN-competent cells and in vivo (15, 21, 27, 29). These studiesalso demonstrated that zIID/rIID proteins end cap dsRNA, po-tentially shielding this blunt-end dsRNA pathogen–associatedmolecular pattern (PAMP) from detection by RLRs (21).RIG-I and melanoma differentiation associated factor gene 5

(MDA5) are RLRs that trigger innate immune signaling in re-sponse to viral infection. RLRs recognize PAMPs, such asdsRNA, dsRNA-containing 5′-triphosphate (5′PPP), and 5′OHblunt-end dsRNAs. Moreover, RLRs can detect the methylationstatus of the mRNA 5′ cap and specific secondary structuralfeatures or non–O-methylated RNA (30–34). RIG-I, in partic-ular, is thought to bind short dsRNA. In contrast, MDA5 is ac-tivated by long(er) dsRNA ligands, including poly I:C (pI:C), andhas the potential to form long filamentous signaling structures ondsRNA ligands (35, 36). RLRs are targeted by many virusencoded proteins to antagonize IFN induction (37), mechanismsby which viral proteins such as VP35 antagonize RLRs are notcompletely understood.To better define how MARV VP35 (mVP35) inhibits host in-

nate-immune responses and to develop filoviral VP35s as a po-tential panfiloviral therapeutic target, we performed structural andfunctional studies of the mVP35 protein. Our data support theability of mVP35 to antagonize IFN production through multiplemechanisms, including inhibition of RLR activation throughdsRNA sequestration and by direct targeting of IFN kinases IKKe/TBK-1. Consistent with this model, we observe that mVP35 IID

Author contributions: P.R., M.R.E., R.S.S., D.W.L., A.C.E.-F., C.F.B., and G.K.A. designedresearch; P.R., M.R.E., D.W.L., A.C.E.-F., D.M.B., G.L., and G.K.A. performed research;Z.O. and J.H. contributed new reagents/analytic tools; P.R., M.R.E., D.W.L., A.C.E.-F.,D.M.B., Z.O., G.L., and G.K.A. analyzed data; and P.R., M.R.E., R.S.S., D.W.L., A.C.E.-F.,D.M.B., C.F.B., and G.K.A. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Data deposition: The atomic coordinates and structure factors have been deposited in theProtein Data Bank, www.pdb.org (PDB ID code 4GHL).1To whom correspondence should be addressed. E-mail: gamarasinghe@path.wustl.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1213559109/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1213559109 PNAS | December 11, 2012 | vol. 109 | no. 50 | 20661–20666

MICRO

BIOLO

GY

Dow

nloa

ded

by g

uest

on

Nov

embe

r 2,

202

0

(mIID) binds 18-bp dsRNA through contacts with the phospho-diester backbone. Mutation of dsRNA-binding residues reducesdsRNA binding and correspondingly increases IRF3 phosphory-lation and IFN-β promoter activity. We also show that mVP35may use a distinct set of residues to inhibit IFN kinases IKKe/TBK-1, upstream of IFN kinase activity. zIID and mIID can alsoinhibit MDA5 ATPase activation by pI:C. However, only zIID canalso inhibit RIG-I ATPase activation by short (8- to 30-bp)blunt-end dsRNA. In contrast, both mVP35 and zVP35 inhibitRIG-I activation by 5′ overhang dsRNA, 3′-overhang dsRNA,and pI:C. Altogether, these data support a model where filo-viral VP35 antagonizes host IFN induction at multiple levels bydifferential recognition of viral PAMPs.

ResultsmIID Binds dsRNA Independent of Ends. dsRNA binding to zVP35 iscritical to its ability to fully inhibit type I IFN production (10, 15,20, 21). To examine the role of mVP35 in IFN antagonism, wetested its ligand-binding properties. mIID did not bind 8-bpdsRNA, as monitored by isothermal titration calorimetry (ITC).In contrast, zIID bound with high affinity in the same assay witha KD = 0.5 ± 0.1 μM (Fig. S1A) (21). mIID can bind longerdsRNA, where 18-bp dsRNA binds with a KD = 6.6 ± 0.6 μM and30-bp dsRNA binds with a KD = 1.3 ± 0.1 μM (Fig. S1B),whereas zIID binds 18- and 30-bp dsRNA with similar affinities(KD = 2.2 ± 0.2 μM) (Fig. S1C). We also tested 5′- or 3′-over-hang dsRNA to determine whether dsRNA ends are importantfor mIID binding. mIID binds 25-bp dsRNA with blunt, 5′ and 3′overhangs with similar affinity (KD = 1.3 ± 0.1 μM, KD = 1.7 ±0.1μM, and KD = 2.2 ± 0.1 μM, respectively) (Fig. S1D). In contrast,zIID shows a 3- and 15-fold decrease in binding to 5′- and 3′-overhang dsRNA, respectively, compared with blunt-end dsRNA(Fig. S1E). These results (Table S1) suggest that the dsRNA endsare important for zIID recognition of ligand and not for mIID.

mIID Binds the dsRNA Backbone. To explore the structural basis fordsRNA binding and specificity of mVP35, we solved the crystalstructure of mIID in complex with 18-bp dsRNA to 2.01-Å

resolution (see Fig. 1 and Table S2 for structure statistics). In thestructure, we observe four molecules of mIID (chains A, B, C,and D) and one 18-bp dsRNA (chain E and F). These inter-actions result in a configuration where dsRNA interacts withmultiple mIID molecules that appear to coat the backbone ofdsRNA with a binding “footprint” of 4–5 bp (Fig. 1 A and B).Moreover, the 18-bp dsRNAs are stacked end to end with aslight offset, creating a pseudocontinuous A–form dsRNA helix.Side chains from several CBP residues, including R294, K298,R301, and K328, form a positively charged surface that contactsthe negatively charged phosphodiester backbone of the RNA(Fig. 1C). For example, the side-chain NH1 of R294 and Nζ ofK298 interact with U9 base O1P atom through water-mediatedH bonds, and the side-chain Nζ of K328 and NH1 and NH2 ofR301 also form water-mediated H bonds with O2′ of C13 (Fig. 1D and E). These MARV residues correspond to R305, K309,R312, and K339 in zIID and R294, K298, R301, and K328 inrIID, which form critical interactions with the RNA backbone(Fig. S2 A and B) (26). Alignment of the mIID structure with zIIDand rIID structures shows that the overall fold of the α-helical andβ-sheet subdomains are similar, with backbone rmsds of 0.82 Å(chain B of PDB ID code 3L25 with chain D of PDB ID code4GHL) and 0.72 Å [chain A of PDB ID code 3KS8 with chain Dof PDB ID code 4GHL (Fig. S2 A and B)].Despite these striking similarities in the protein fold, there are

a number of differences between the mIID-dsRNA complexstructure and the zIID/rIID-dsRNA structures (21, 26). ResiduesK319 and R322 in zIID (corresponding to K308 and R311 inrIID) have been shown previously to form part of a criticalnetwork of basic residues that contact dsRNA, as well as VP35protein–protein contacts (21, 26, 38). The sequence equivalentresidues in mVP35, T308, and K311(Fig. S2 C and D), however,are solvent-exposed and have no apparent role in RNA binding.Interestingly, a number of polar residues in mVP35 form criticalcontacts with the RNA backbone, including N261, Q263, andT267. Atom Nδ2 of N261 and atom Oe1 of Q263 form hydrogenbonds with O3′ and O2′ of base U11, respectively (Fig. S2 E andF). Furthermore, mIID does not appear to participate in end-

A 1 204 329OLIGOMERIZATION DOMAIN IID

B

18 bpC D E

U9,O1P/F

R294

K298

A8,O2’/F

S299

C13,O1P/EC13,O2’/ER301

K328R271

T267

Q263

U11,O2’/E

N

R301Q263N261

R271T267

K328

S299R294

C

90°

K298

Fig. 1. Crystal structure of mIID bound to 18-bpdsRNA. (A) Domain organization of mVP35 protein.(B) The crystallographic asymmetric unit containsfour molecules of mIID (chains A, B, C, and D col-ored in cyan, purple, green, and pink, respectively)binding to one 18-bp dsRNA (chains E and F, shownin surface representation in red). (Right) View ofthe complex structure down the dsRNA axis. (C)mIID (chain D, yellow) contacts the phosphodiesterbackbone of dsRNA through key basic and polarresidues. (D) Residues Q263, T267, R271, R301, andK328 form H bonds with chain E of dsRNA. (E) R294,K298, and S299 forms H bonds with chain F of dsRNA.

20662 | www.pnas.org/cgi/doi/10.1073/pnas.1213559109 Ramanan et al.

Dow

nloa

ded

by g

uest

on

Nov

embe

r 2,

202

0

capping interactions in the context of our crystal structure (Fig.2A). This is in direct contrast to the structures of zIID and rIIDbound to dsRNA, where IIDs interact with dsRNA blunt endsthrough hydrophobic contacts (F239, Q274, and I340 in zIID andF228, I267, and I329 in rIID) (21, 26) (Fig. 2 B and C). Mutation ofF239 in zIID resulted in loss of dsRNA binding, suggesting thatF239 (F228 in mIID) is important for both modes of dsRNAbinding (21). Moreover, mIID CBP residues are only important fordsRNA binding and not protein–protein contacts, as observed inthe zIID-dsRNA structure (21). Limited protein–protein contactsand corresponding low buried surface areas at this interface, 1,400Å2 for zIID-dsRNA compared with 555 Å2 for mIID-dsRNA,suggest that the mIID may function differently from zIID/rIID.

In Vitro Assays Validate the Protein–RNA Interface. In vitro filter-binding assays were used to assess the importance of residuesat the protein–RNA interface to bind 18-bp dsRNA (Fig. 3A),30-bp dsRNA (Fig. 3B), and pI:C (Fig. 3C). Alanine sub-stitutions of R294, K298, S299, R301, and K328 in the CBPresulted in decreased RNA binding (<40% of WT), whereasdouble mutants R294A/K298A and K298A/S299A further di-minished dsRNA binding (<5% of WT). In addition to basicresidues in the CBP that have been shown to be critical fordsRNA binding, F228A, N261A, and Q263A also resulted in lossof binding (<40% of WT). These mutations, when tested in thecontext of full-length mVP35 proteins in 293T cells, displayeda good correlation between mutants that disrupt RNA binding invitro with those that were impacted in the pI:C pull-downs. Theexceptions were R294A, K298A, and S299A, which show appre-ciable pI:C binding. This may reflect dsRNA length–dependentbinding because these mutants were less impaired for binding 30-bp RNA and show a statistically significant preference for longerdsRNA (P values of 0.007, 0.08, and 0.006, respectively) (Fig.3C). T267A and R271A also exhibited variable pull-down overseveral experiments. Interestingly, a fraction of some of themVP35 mutants exhibit retarded migration during SDS/PAGE

(Fig. 3C). The basis for this aberrant migration is unknown (21).Of note, N261A/Q263A mutant was either not well expressed ornot detected by our antibodies in the pI:C pull-down studies.

Residues Critical for dsRNA Binding Are also Important for IFNInhibition. To assess the effect of dsRNA-binding mutations onthe ability of mVP35 to suppress induction of type I IFN responses,reporter gene assays were carried out to measure the activation of

90°

90°

90°

A

C

B

Fig. 2. mIID binds dsRNA in a mode that is distinct from zIID and rIID. Surfacerepresentations of mIID (yellow) in complex with 18bp dsRNA (red) (PDB ID code4GHL) (A), zIID (orange) in complex with 8bp dsRNA (green) (PDB ID code 3L25)(B), and rIID (magenta) in complexwith 18bpdsRNA (cyan) (PDB ID code 3KS8) (C).

A

B

C

dnuoB

noitcarFdezila

mroN

0.0

0.2

0.4

0.6

0.8

1.0

1.2

WT

F228A

N261A

Q263A

R294AK29

8AS29

9A

R301A

N261A

Q263A

R294A

K298A

K298A

S299A

R285AK28

7A

N306A

Q321AK32

8A

dnuoB

noitcarFdezila

mroN

0.0

0.2

0.4

0.6

0.8

1.0

1.2

WT

F228A

N261A

Q263A

R294AK29

8AS29

9A

R301A

N261A

/Q26

3A

R294A

/K29

8A

K298A

/S29

9A

R285AK28

7A

N306A

Q321AK32

8A

β-tubulin

VP35

VP35

VP35

Pull-down: no pI:C

WCE

40

40

30

50

40

50

50Pull-down: pI:C beads

EmptyF22

8AN26

1A

Q263A

R294A

K298AS29

9AR30

1A

K328A

R294A

/K29

8A

K298A

/S29

9A

R285A

K287A

N306A

Q321A

N261A

/Q26

3A

WT

T267AR27

1A

T308K

T308K

/K31

1R

Fig. 3. Structure-based mutations disrupt mIID binding to dsRNA in vitro. (Aand B) 32P 5′-end–labeled 18-bp dsRNA (A) and 30-bp dsRNA (B) were ana-lyzed for binding to mIID by double-membrane filter-binding experiments.Fractional binding of the mutants (black bar) was normalized relative to WTprotein (gray bar). Error bars represent the standard deviation from twoindependent experiments. (C) Western blot analysis of pI:C pull-downs, us-ing full-length WT or mutant mVP35 proteins expressed in 293T cells andanti-mVP35 mAb. Empty is a transfected empty vector control. The uppertwo images are the pull-downs using beads without pI:C and beads con-taining poly I:C. The lower two images show the overall VP35 and β-tubulinprotein expressions in whole-cell extracts (WCE).

Ramanan et al. PNAS | December 11, 2012 | vol. 109 | no. 50 | 20663

MICRO

BIOLO

GY

Dow

nloa

ded

by g

uest

on

Nov

embe

r 2,

202

0

the IFN-β promoter upon Sendai virus (SeV) infection in the ab-sence or presence of mVP35 (13, 15). Our results suggest that themutation of residues that led to reduced dsRNA binding in vitroalso show attenuated function as an IFN antagonist in vivo (Fig. 4A and B). In particular, double mutants R294A/K298A, K298A/S299A, and N261/Q263 show near-complete loss of function,whereas mutants F228A, K298A, S299A, R301A, and K328A showdiminished ability to suppress IFN-β induction. In contrast, muta-tion of residues involved in protein–protein contacts in the crystal,R285, K287, N306, and Q321, were functional in this assay, sup-porting that these interactions are not important for IFN antago-nism. mVP35 constructs that inhibit activation of the IFN-βpromoter also inhibit SeV-induced phosphorylation of IRF3 (Fig.4C). These results support the ability of mVP35 to inhibit IFNinduction and further establish a correlation between dsRNA-binding and IFN-inhibitory functions of mVP35.

dsRNA-Independent Inhibition of IKK3/TBK-1 by mVP35. Previousstudies with zVP35 have shown that one mechanism underlyingthe ability of EBOV to suppress IFN-β production is throughinhibition of IRF3/IRF7 phosphorylation by the IFN kinasesIKKe/TBK-1 (11, 13, 25). Similar to zVP35, mVP35 is also ableto block IFN-β promoter activity in a dose-dependent mannerupon overexpression of IKKe or TBK-1 in 293T cells (Fig. S3 Aand B). Inhibition at the level of the kinases does not appearto be dsRNA-dependent because dsRNA-binding mutants im-paired in the SeV-based assays inhibited comparably to WTmVP35 (Fig. 4). In contrast, when a constitutively active form ofIRF3 is overexpressed, IRF3 5D, in 293T cells (Fig. S3C),mVP35 was unable to significantly inhibit the IFN-β promoteractivity, suggesting that the inhibitory effect of mVP35 is up-stream of IRF3 phosphorylation.

Differential Recognition and Inhibition of RLR PAMPs by mIID. In thestructure of the mIID-dsRNA complex, we observe mIID bind-ing only to the dsRNA backbone and not the blunt ends. Lack ofend capping may potentially be attributable to competing crystal-packing contacts. To test the functional relevance of our struc-ture, we assessed the ability of mIID to inhibit RIG-I and MDA5activation by different RNA PAMPs by an in vitro ATPase assay.As shown in Fig. S4A, 8-bp dsRNA is able to enhance the ATPaseactivity of RIG-I. RIG-I ATPase activity is reduced markedlyupon addition of zIID, which recognizes both double-stranded-ness and blunt-ended PAMPs. In contrast, mIID does not inhibitRIG-I activation by 8-bp dsRNA because mIID does not bind 8-bp dsRNA (Fig S1 A and B). RIG-I activation by 30-bpdsRNA is also inhibited by zIID, but mIID is unable to inhibitRIG-I ATPase activity despite its ability to bind 30-bp dsRNAbecause the blunt ends of 30-bp dsRNA are presumably availablepresumably to activate RIG-I (Fig. S4B). These findings areconsistent with our crystal structure because mIID is unable toinhibit RIG-I activation by blunt-ended dsRNA PAMP.Next, we tested MDA5 activation in ATPase assays with 30-bp

dsRNA, low-molecular-weight (LMW) pI:C, and high-molecu-lar-weight (HMW) pI:C, which show that mIID can inhibitMDA5 activation by 30-bp dsRNA (Fig. S4C). Moreover, mIIDand zIID can also inhibit the activation of MDA5 and RIG-I(Fig. 5A and Fig. S5) by HMW and LMW pI:Cs. mIID-con-taining mutations of residues involved in dsRNA contacts in thecrystal structure of mIID-dsRNA complex, such as F228A andR294A/K298A, are defective in their ability to inhibit MDA5ATPase activation (Fig. 5B). In contrast, mutants of residuesinvolved in protein–protein contacts, such as K287A and N306A,can inhibit MDA5 activity to levels similar to WT.Because mIID did not differentiate between blunt-end dsRNA

and dsRNA containing overhangs in our binding assays (Fig.S1D), we also tested the ability of mVP35 to inhibit RIG-I ac-tivation by a 25-bp dsRNA with blunt ends, 5′ overhang, and 3′overhang. Blunt-end dsRNA and 5′-overhang dsRNA both ac-tivated RIG-I to comparable levels (Fig. 5C), and each wasinhibited by zIID. In contrast, mIID was only able to inhibit 5′-overhang dsRNA. As previously shown, 3′ overhang dsRNA didnot robustly activate RIG-I (3, 32), and, therefore, the variationin RIG-I activity upon VP35 IID addition are not significant.Altogether, these results, summarized in Table S3, suggest thatmIID can inhibit RIG-I activation by double-stranded PAMPsbut cannot prevent RIG-I activation through dsRNA blunt ends,presumably because of lack of dsRNA end capping by mIID.

DiscussionStructural, biochemical, and cell-based studies of mVP35 revealedseveral important findings. mIID forms a fold similar to zIID andrIID, despite large sequence differences (9 residues between zIIDand rIID compared to 52 residues between zIID and mIID). How-ever, mIID can sequester dsRNA and inhibit IFN induction througha mechanism that appears to be independent of end capping (Figs.S2 and S6). In vitro studies show that mIID can only inhibit RIG-Iactivation by non–blunt-end RNA or pI:C. mVP35 also displays anability to inhibit IFN kinases IKKe/TBK-1 via a RNA-indepen-dent mechanism. Together, these findings highlight how a com-mon structural fold facilitates multiple intermolecular bindingmodes and functional outcomes, providing mechanistic insightsinto MARV antagonism of host IFN responses.We observe few VP35-VP35 contacts in the mIID-dsRNA

complex structure, with only 555 Å2 of buried surface area. Incontrast, VP35-VP35 interactions account for about 1,400 Å2 ofburied surface area in the zIID-dsRNA complex (21). We alsoobserve that whereas zIID-dsRNA and mIID-dsRNA complexeshave 7 VP35-VP35 contacts, there are no common residues that areshared between ZEBOV/REBOV and MARV (Fig. S6). In-terestingly, residues shown to be involved inVP35-VP35 interactionshave no functional impact on dsRNA-binding and IFN-inhibitoryfunctions of mVP35, whereas VP35-VP35 contact residues in zIID-dsRNA complex are important for IFN inhibition (21).

A B

C

--

-*

125

100

75

50

25

0% IF

N-β

repo

rter a

ctiv

ity

125

100

75

50

25

0% IF

N-β

repo

rter a

ctiv

ity

E

F228

AN

261A

Q26

3A

R29

4AK

298A

S29

9AR

301A

R28

5AK

287A

N30

6A

Q32

1AK

328AE

WT

T267

AR

271A

T308

K E

N26

1A/Q

263A

R29

4A/K

298A

K29

8A/S

299AE

WT

T308

K/K

311R

+SeV +SeV

N26

1A/Q

263A

R29

4A/K

298A

K29

8A/S

299A

T308

K/K

311R

F228

AN

261A

Q26

3A

R29

4AK

298A

S29

9AR

301A

R28

5AK

287A

N30

6A

Q32

1A

K32

8A

T267

A

R27

1A

T308

K

E +

IRF3

E WT

EE NP

+SeV+IRF3

5555

30

α-P-IRF3(Ser396)α-IRF3

α-VP35

Fig. 4. Mutations to the RNA-binding domain of mVP35 attenuate IFN-βinhibition. (A and B) IFN-β promoter activity induced by SeV infection of 293Tcells transfected with either WT or mutant mVP35 proteins (black bar) wereassayed and normalized relative to the empty vector control (gray bar). Errorbars represent standard error of the mean for triplicate experiments. (C)Western blot analysis of the IRF3 phosphorylation state in 293T cells afterSeV activation of the IFN-β promoter using anti–phospho-IRF3 (Ser396) (Top)and anti-IRF3 (Middle) antibodies. Cells were transfected with MARV nu-cleoprotein (NP) or indicated VP35 construct and IRF3. The expression of WTand mutant mVP35 proteins was assessed using anti-mVP35 mAb (Bottom). Erefers to empty vector transfection control. *Nonspecific band.

20664 | www.pnas.org/cgi/doi/10.1073/pnas.1213559109 Ramanan et al.

Dow

nloa

ded

by g

uest

on

Nov

embe

r 2,

202

0

Residues in the mIID CBP region that are important fordsRNA binding in the structure are also important for dsRNAbinding in vitro and in pI:C binding in cell extracts (Fig. 3).Based on structural comparisons, we observe that all 14 VP35-dsRNA contacts are shared by zIID and rIID, but only 11 arecommon among zIID, rIID, and mIID (Fig. S6A). Differences atpositions corresponding to K319/R322 in zVP35 may be impor-tant for intermolecular interactions (Fig. S6B). Although it is notclear why mIID is unable to recognize dsRNA blunt ends in thecrystal structure, variations at these positions in the mIID se-quence may contribute, at least in part, to observed differencesin structure and affinity.The abilities of mVP35 and zVP35 to inhibit RIG-I and

MDA5 activation by a variety of dsRNA ligands were tested incell-based assays. Like its EBOV counterparts, mVP35 is able tobind dsRNA and pI:C, and suppress SeV-induced IFN-β in-duction in a dose-dependent manner (Figs. 3 and 4). Overall, weobserve a correlation between mIID residues that bind dsRNAand their ability to antagonize IFN responses. However, R294A,K298A, and S299A mutations, which show largely diminisheddsRNA binding to 18- and 30-bp dsRNA, were functional in thepI:C pull-downs. These mutants also show higher binding affin-ities for the longer dsRNA in vitro. Therefore, the longer lengthof pI:C may explain the retention of binding in this assay. Ad-ditionally, because full-length VP35 is used in these assays, thepresence of an oligomerization domain may promote multivalentbinding with enhanced affinity (Fig. 1A) (39, 40). Interestingly,double mutants R294A/K298A and K298A/S299A show a near-

complete loss of binding. The enhanced impact of the doublemutants may be attributable to the loss of potential compensa-tory role played by these residues, where loss of either residuecan be tolerated but not both. We see near-uniform correspon-dence between impact on dsRNA binding and ability to antag-onize IFN-β promoter activation by SeV infection by mutants(Fig. 4). As was seen with zVP35, mVP35 mutants defective indsRNA binding do retain some capacity to inhibit the IFN-βpromoter compared with empty vector–transfected controls.mVP35 can also inhibit IFN-β induction when IFN kinases IKKe/TBK-1 are expressed. Inhibition is lost when a constitutivelyactive IRF3-5D is used, suggesting that mVP35 targets thesekinases (25). Because all of the mVP35 dsRNA-binding mutantsare functional in this assay, suppression of RLRs vs. the kinaseslikely requires different residues.dsRNA binding by the C-terminal domain and the helicase

domains of RLRs results in a conformational switch that triggersdownstream signaling (reviewed in ref. 34). Although the natureof the exact ligands that activate RLRs in vivo have not yet beendefinitively identified, dsRNA-bound RIG-I and MDA5 struc-tures show that multiple PAMPs, including double-strandedness,dsRNA blunt ends, and short dsRNA with 5′OH or 5′PPP, canactivate RLRs (34). A correlation between dsRNA-bindingability of zVP35 and its IFN-inhibitory effect on RLR functionhas been documented in several studies (15, 16, 19, 21), and itsbiological relevance is supported by the observations that pre-activation of RIG-I dramatically decreases ZEBOV yield andthat recombinant ZEBOVs with mutations at critical RNA-binding residues are attenuated in vitro and avirulent in vivo (29,41). Our structural findings are consistent with the in vitrodsRNA-binding studies, suggesting that the crystal structurereflects a potentially physiologically relevant complex. The dsRNAin the crystallographic unit cell is stacked coaxially in an end-to-end fashion, forming a pseudocontiguous helix that is coated bymIID molecules along its backbone. Although this configurationmay be influenced by crystal packing, similar structural organi-zations have been observed previously for other viral IFN antag-onists, such as influenza A virus NS1 (PDB ID code 2ZKO),Tombusvirus P19 (PDB ID code 1RPU), and Flock house virusB2 (PDB ID code 2AZO). The impact on dsRNA PAMP rec-ognition in many of these instances remains to be defined.

LMW pI:CHMW pI:C

MDA5

mIID F228AmIID

mIID R294A/K298AmIID K287AmIID N306A

0.0

0.2

0.4

0.6

0.8

1.0

zIID

A B

CmIIDzIID

3’ overhang dsRNA5’ overhang dsRNA

blunt end dsRNA

RIG-I

Nor

mal

ized

ATP

ase

activ

ity

Nor

mal

ized

ATP

ase

activ

ity

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Fig. 5. Differential recognition of RNA blunt ends by mIID and zIID. ATPaseactivation of MDA5 by HMW and LMW pI:C in the absence and presence ofWTmIID or zIID proteins (A) or mutant mIID proteins (B) were assayed by TLC.The amount of free phosphate in each lane was quantitated and normalizedto activation of MDA5 by LMW poly I:C (gray bars) and reported as normal-ized ATPase activity. (C) RIG-I ATPase activation by 25-bp blunt-end dsRNA, 5′overhang, and 3′ overhang in the absence and presence of WT mIID or zIIDproteins were assayed by TLC. The normalized ATPase activity in each lane isplotted relative to RIG-I activation by blunt-end dsRNA (gray bar).

IPS-1

TBK-1/IKK

IRF-3 IRF-3P

viral PAMPs

MARV VP35 EBOV VP35

double-strandedness

blunt ends5’ppp

MDA5 / RIG-I

double-strandedness

5’ppp ?

Fig. 6. Model for RLR inhibition by filoviral VP35 proteins. A workingmodel, based on the current study for mVP35 and previous work for zVP35and rVP35, suggests that differences in dsRNA PAMP recognition by mVP35and zVP35 may result in different levels of MDA5 and RIG-I antagonism (seeDiscussion). Viral PAMPs activate the RLRs MDA5 or RIG-I, which then signalthrough IFN-β promoter stimulator 1 and TBK-1/IKKe to activate IRF3 phos-phorylation. Both MARV and EBOV VP35s are proposed to block RLR acti-vation at multiple steps in the RLR pathways. However, the data in this studysuggest that antagonism at the level of PAMP sequestration by MARV VP35occurs through its ability to bind dsRNA, whereas EBOV VP35 masks dsRNAand dsRNA blunt ends possessing RIG-I–activating 5′PPPs.

Ramanan et al. PNAS | December 11, 2012 | vol. 109 | no. 50 | 20665

MICRO

BIOLO

GY

Dow

nloa

ded

by g

uest

on

Nov

embe

r 2,

202

0

To determine whether the mechanism of immune evasion bymVP35 is distinct from zVP35, we tested the effect of bothmVP35and zVP35 on RIG-I and MDA5 activation. Our data show thatshort dsRNA is able to activate RIG-I ATPase function. We alsoobserve thatMDA5 can be activated by 18- to 30-bp dsRNA, whenused at sufficiently high concentrations (Fig. S4C). However, onlyzVP35 is able to inhibit dsRNA-mediated RLR activation by shortdsRNAs in a dose-dependent manner. This correlates with theability of zVP35 to compete with RIG-I for dsRNA backbone andblunt-end binding because mutation of residues such as F239,R312, R319, and K322 leads to a loss of dsRNA binding anda corresponding loss of RIG-I inhibition (15, 16, 19, 21). We alsoshow that both MARV and ZEBOV can effectively inhibit RLRactivation by pI:C, presumably through interactions with thedsRNA backbone because the dominant PAMP in pI:C is thedouble-strandedness.Weobserve thatmIID, which was unable toinhibit RIG-I activation by blunt 25-bp dsRNA, can inhibit RIG-Iactivation by 5′-overhang dsRNA. These data strongly suggest thatMARV and ZEBOV can directly antagonize RLR activation bythe double-strandedness of RNA. A model consistent with ourresults described above is shown in Fig. 6, where mVP35 can onlyantagonize recognition of RNA double-strandedness by RLRs. Incontrast, ZEBOV (and likely REBOV) can compete and inhibitRLR activation by masking the blunt ends (with 5′OH or 5′PPP)and double-strandedness. The relative contributions of thesePAMPs toward RLR activation are currently unknown. However,the significant differences in PAMP recognition described forRIG-I and MDA5, coupled with our observations here, suggest thatthe type of PAMPs present during MARV and EBOV infec-tions may also be different.

Materials and MethodsStructure Determination. Diffraction quality crystals for mIID-dsRNA complexwas obtained formIID (204–329) and 18-bp dsRNA (AGACAGCAUAUGCUGUCU)(Integrated DNA Technologies) mixed in 2:1 molar ratio using hanging-dropvapor in well solution containing 0.1 M ammonium citrate (pH 6.3), 0.1 Mammonium citrate (pH 6.2), 0.25% ethylene glycol, 14% (vol/vol) PEG 3350,and 0.23 M ammonium sulfate. Diffraction data were collected at the Ad-vanced Photon Source (Beamline Structural Biology Center 19), processed,and refined as described previously using the zIID structure (PDB ID code 3FKE)as a search model (21). Collection and refinement statistics are in Table S2.

Cell-Based Functional Studies. IFN-β promoter studies, IRF3 phosphorylation,and pI:C pull-down studies were carried out for WT and mutant mIID andzIID proteins as described previously (13, 15, 21, 25).

RNA Binding and ATPase Studies. RNA-binding studies for mIID and zIID werecarried out as described previously under conditions indicated in the figurelegends, using dsRNA sequences (Table S4). MDA5 and RIG-I ATPase assayswere carried out in the presence or absence of IID proteins, and the hy-drolysis was measured on polyethyleneimine (PEI)-cellulose TLC using rela-tive signal-intensity measurements for inorganic 32P and [32P]ATP.

Detailed methods are described in the SI Materials and Methods.

ACKNOWLEDGMENTS. We thank Drs. B. Fulton, R. Honzatko, and T. Wang(Iowa State University); Drs. S. Ginell, N. Duke, M. Cuff, and J. Lazarz (StructuralBiology Center); Dr. J. Nix (Advanced Light Source 4.2.2); and Ms. J. Binningfor technical assistance in data collection and analysis. Use of the ArgonneNational Laboratory Structural Biology Center beamlines was supported byUS Department of Energy (DOE) Contract DE-AC02-06CH11357. This workwas supported, in part, by National Institutes of Health (NIH) Grants AI089547(to C.F.B. and G.K.A.), AI059536 and AI057158 (Northeast Biodefense Center-Lipkin) (to C.F.B.), F32AI084453 (to R.S.S.), GM053163 (to Z.O.), F32AI084324(to D.W.L.), and AI081914 (to G.K.A.).

1. Sanchez A, Wagoner KE, Rollin PE (2007) Sequence-based human leukocyte antigen-Btyping of patients infected with Ebola virus in Uganda in 2000: Identification of allelesassociated with fatal and nonfatal disease outcomes. J Infect Dis 196(Suppl 2):S329–S336.

2. Kuhn JH, et al. (2010) Proposal for a revised taxonomy of the family Filoviridae: Classi-fication, names of taxa and viruses, and virus abbreviations. Arch Virol 155(12):2083–2103.

3. Enterlein S, et al. (2009) The marburg virus 3′ noncoding region structurally andfunctionally differs from that of ebola virus. J Virol 83(9):4508–4519.

4. Mühlberger E, Weik M, Volchkov VE, Klenk HD, Becker S (1999) Comparison of thetranscription and replication strategies of marburg virus and Ebola virus by usingartificial replication systems. J Virol 73(3):2333–2342.

5. Valmas C, et al. (2010) Marburg virus evades interferon responses by a mechanismdistinct from ebola virus. PLoS Pathog 6(1):e1000721.

6. Reid SP, et al. (2006) Ebola virus VP24 binds karyopherin alpha1 and blocks STAT1nuclear accumulation. J Virol 80(11):5156–5167.

7. Prins KC, et al. (2010) Basic residues within the ebolavirus VP35 protein are requiredfor its viral polymerase cofactor function. J Virol 84(20):10581–10591.

8. MateoM, Reid SP, Leung LW, Basler CF, Volchkov VE (2010) Ebolavirus VP24 binding tokaryopherins is required for inhibition of interferon signaling. J Virol 84(2):1169–1175.

9. Valmas C, Basler CF (2011) Marburg virus VP40 antagonizes interferon signaling ina species-specific manner. J Virol 85(9):4309–4317.

10. Kash JC, et al. (2006) Global suppression of the host antiviral response by Ebola- andMarburgviruses: Increased antagonism of the type I interferon response is associatedwith enhanced virulence. J Virol 80(6):3009–3020.

11. Basler CF, Amarasinghe GK (2009) Evasion of interferon responses by Ebola andMarburg viruses. J Interferon Cytokine Res 29(9):511–520.

12. Leung DW, Prins KC, Basler CF, Amarasinghe GK (2010) Ebolavirus VP35 is a multi-functional virulence factor. Virulence 1(6):526–531.

13. Basler CF, et al. (2003) The Ebola virus VP35 protein inhibits activation of interferonregulatory factor 3. J Virol 77(14):7945–7956.

14. Basler CF, et al. (2000) The Ebola virus VP35 protein functions as a type I IFN antag-onist. Proc Natl Acad Sci USA 97(22):12289–12294.

15. CárdenasWB, et al. (2006) Ebola virusVP35proteinbindsdouble-strandedRNAand inhibitsalpha/beta interferon production induced by RIG-I signaling. J Virol 80(11):5168–5178.

16. Enterlein S, et al. (2006) VP35 knockdown inhibits Ebola virus amplification and pro-tects against lethal infection in mice. Antimicrob Agents Chemother 50(3):984–993.

17. Feng Z, Cerveny M, Yan Z, He B (2007) The VP35 protein of Ebola virus inhibits theantiviral effect mediated by double-stranded RNA-dependent protein kinase PKR.J Virol 81(1):182–192.

18. Haasnoot J, et al. (2007) The Ebola virus VP35 protein is a suppressor of RNA silencing.PLoS Pathog 3(6):e86.

19. Hartman AL, et al. (2008) Inhibition of IRF-3 activation by VP35 is critical for the highlevel of virulence of ebola virus. J Virol 82(6):2699–2704.

20. Hartman AL, Towner JS, Nichol ST (2004) A C-terminal basic amino acid motif of Zaireebolavirus VP35 is essential for type I interferon antagonism and displays highidentity with the RNA-binding domain of another interferon antagonist, the NS1protein of influenza A virus. Virology 328(2):177–184.

21. Leung DW, et al. (2010) Structural basis for dsRNA recognition and interferon an-tagonism by Ebola VP35. Nat Struct Mol Biol 17(2):165–172.

22. Mühlberger E, Lötfering B, Klenk HD, Becker S (1998) Three of the four nucleocapsid pro-teins of Marburg virus, NP, VP35, and L, are sufficient to mediate replication and tran-scription of Marburg virus-specific monocistronic minigenomes. J Virol 72(11):8756–8764.

23. Schümann M, Gantke T, Mühlberger E (2009) Ebola virus VP35 antagonizes PKR ac-tivity through its C-terminal interferon inhibitory domain. J Virol 83(17):8993–8997.

24. Shabman RS, et al. (2011) DRBP76 associates with Ebola virus VP35 and suppressesviral polymerase function. J Infect Dis 204(Suppl 3):S911–S918.

25. Prins KC, Cárdenas WB, Basler CF (2009) Ebola virus protein VP35 impairs the function ofinterferonregulatory factor-activatingkinases IKKepsilonandTBK-1. JVirol83(7):3069–3077.

26. Kimberlin CR, et al. (2010) Ebolavirus VP35 uses a bimodal strategy to bind dsRNA forinnate immune suppression. Proc Natl Acad Sci USA 107(1):314–319.

27. Leung DW, et al. (2009) Structure of the Ebola VP35 interferon inhibitory domain.Proc Natl Acad Sci USA 106(2):411–416.

28. Leung DW, et al. (2010) Structural and functional characterization of Reston Ebolavirus VP35 interferon inhibitory domain. J Mol Biol 399(3):347–357.

29. Prins KC, et al. (2010) Mutations abrogating VP35 interaction with double-strandedRNA render Ebola virus avirulent in guinea pigs. J Virol 84(6):3004–3015.

30. Kumar H, Kawai T, Akira S (2011) Pathogen recognition by the innate immune system.Int Rev Immunol 30(1):16–34.

31. Pichlmair A, et al. (2006) RIG-I-mediated antiviral responses to single-stranded RNAbearing 5′-phosphates. Science 314(5801):997–1001.

32. Schlee M, et al. (2009) Recognition of 5′ triphosphate by RIG-I helicase requires shortblunt double-stranded RNA as contained in panhandle of negative-strand virus.Immunity 31(1):25–34.

33. Züst R, et al. (2011) Ribose 2′-O-methylation provides a molecular signature for thedistinction of self and non-self mRNA dependent on the RNA sensor Mda5. NatImmunol 12(2):137–143.

34. Leung DW, Amarasinghe GK (2012) Structural insights into RNA recognition andactivation of RIG-I-like receptors. Curr Opin Struct Biol 22(3):297–303.

35. Berke IC, Modis Y (2012) MDA5 cooperatively forms dimers and ATP-sensitivefilaments upon binding double-stranded RNA. EMBO J 31(7):1714–1726.

36. Peisley A, et al. (2011) Cooperative assembly and dynamic disassembly of MDA5filaments for viral dsRNA recognition. Proc Natl Acad Sci USA 108(52):21010–21015.

37. Leung DW, Basler CF, Amarasinghe GK (2012) Molecular mechanisms of viral in-hibitors of RIG-I-like receptors. Trends Microbiol 20(3):139–146.

38. Leung DW, et al. (2010) Crystallization and preliminary X-ray analysis of Ebola VP35interferon inhibitory domain mutant proteins. Acta Crystallogr Sect F Struct Biol CrystCommun 66(Pt 6):689–692.

39. Möller P, Pariente N, Klenk HD, Becker S (2005) Homo-oligomerization of MarburgvirusVP35 is essential for its function in replication and transcription. J Virol 79(23):14876–14886.

40. Reid SP, Cárdenas WB, Basler CF (2005) Homo-oligomerization facilitates the in-terferon-antagonist activity of the ebolavirus VP35 protein. Virology 341(2):179–189.

41. Spiropoulou CF, et al. (2009) RIG-I activation inhibits ebolavirus replication. Virology392(1):11–15.

20666 | www.pnas.org/cgi/doi/10.1073/pnas.1213559109 Ramanan et al.

Dow

nloa

ded

by g

uest

on

Nov

embe

r 2,

202

0