A ΩXaV motif in the Rift Valley fever virus NSsprotein is essential for degrading p62, formingnuclear filaments and virulenceNormand Cyra, Cynthia de la Fuenteb, Lauriane Lecoqa, Irene Guendelb, Philippe R. Chabota, Kylene Kehn-Hallb,1,and James G. Omichinskia,1

aDepartment of Biochemistry and Molecular Medicine, Université de Montréal, Montréal, QC, Canada H3C 3J7; and bNational Center for Biodefense andInfectious Diseases, School of Systems Biology, George Mason University, Manassas, VA 20110

Edited by G. Marius Clore, National Institutes of Health, Bethesda, MD, and approved March 31, 2015 (received for review February 22, 2015)

Rift Valley fever virus (RVFV) is a single-stranded RNA virus capa-ble of inducing fatal hemorrhagic fever in humans. A key compo-nent of RVFV virulence is its ability to form nuclear filamentsthrough interactions between the viral nonstructural protein NSsand the host general transcription factor TFIIH. Here, we identifyan interaction between a ΩXaV motif in NSs and the p62 subunitof TFIIH. This motif in NSs is similar to ΩXaV motifs found in nu-cleotide excision repair (NER) factors and transcription factorsknown to interact with p62. Structural and biophysical studiesdemonstrate that NSs binds to p62 in a similar manner as theseother factors. Functional studies in RVFV-infected cells show thatthe ΩXaV motif is required for both nuclear filament formationand degradation of p62. Consistent with the fact that the RVFVcan be distinguished from other Bunyaviridae-family viruses dueto its ability to form nuclear filaments in infected cells, the motif isabsent in the NSs proteins of other Bunyaviridae-family viruses.Taken together, our studies demonstrate that p62 binding to NSsthrough the ΩXaV motif is essential for degrading p62, formingnuclear filaments and enhancing RVFV virulence. In addition, theseresults show how the RVFV incorporates a simple motif into theNSs protein that enables it to functionally mimic host cell proteinsthat bind the p62 subunit of TFIIH.

Rift Valley Fever Virus | TFIIH | NSs | NMR spectroscopy | nuclear filaments

Rift Valley fever virus (RVFV) is a single-stranded RNA virusand the causative agent of the zoonotic vector-borne disease

Rift Valley fever (1). Although RVFV is considered a tropicalvirus, like other tropical hemorrhagic fever viruses includingEbola, there is increasing concern that it will become moreprevalent in other regions of the world due to vector migration asa result of rising global temperatures (2, 3). The virus can betransmitted to both humans and livestock most frequently bymosquitoes, and infection with RVFV is initially characterizedby mild flu-like symptoms. However, the infection has the ca-pacity to progress to encephalitis, hepatitis and ultimately fatalhemorrhagic fever depending on the virulence of the strain (3–5). At this time, there are no approved drug therapies or vaccinesto treat RVFV infection (3, 6). Given its potential as both ahuman and agricultural pathogen (7), the United States Govern-ment currently classifies the virus as a category A priority pathogenbecause of its capacity to be used as a bio-terrorism agent (3, 6).RVFV is a member of the genus Phlebovirus family Bunyaviridae

and like other Bunyaviridae, it contains a tripartite genome com-prised of a large (L) and medium (M) negative-sense RNA seg-ment, as well as a small (S) ambisense RNA segment (4). Despitehaving no defined structural domains, the nonstructural protein(NSs) encoded by the S-segment is the major virulence factor as-sociated with the RVFV, in part through its ability to enter thenucleus of the host cells (8–10). Once in the nucleus, NSs alters anumber of processes in the host cell including suppressing the in-duction of type I interferons (11, 12), promoting the degradationof the antiviral factor protein kinase R (PKR; refs. 13–15) and

inducing p53-dependent apoptotic and DNA-damage signalingpathways (16, 17). One key element to these multiple functions isthe ability of NSs to form nuclear filaments at early stages of theinfection by sequestering the general transcription factor IIH(TFIIH; refs. 18–21).Given the functional importance of the NSs nuclear filaments

and their association with increased virulence of RVFV, severalstudies have examined their formation following RVFV infection(4). Initial studies demonstrated that nuclear filament formationrequired a direct interaction between the amino-terminal regionof NSs and the p44 subunit of TFIIH (21). These studies alsoshowed that the p44 subunit was an important component of thenuclear filament, and that several other subunits of the TFIIHcomplex (10 in total) were degraded following filament forma-tion including the p62 and XP (Xeroderma pigmentosum group)Dsubunits (21). Subsequently, it was determined that NSs formeda direct interaction with both the p62 subunit of TFIIH (22)and the F-box ubiquitin ligase FBXO3 (23), and that NSsbinding to FBXO3 was required for ubiquitin-mediated deg-radation of p62 following RVFV infection (23). Interestingly,these studies also demonstrated that p62 is not targeted byFBXO3 in noninfected cells, but that following RVFV infection,the degradation of p62 resulted in a suppression of the IFNresponse (23).

Significance

Infection with the Rift Valley fever virus (RVFV) has the ca-pacity to cause fatal hemorrhagic fever in humans. A uniquecharacteristic of RVFV infection is the presence of nuclear fila-ments whose formation is linked to synthesis of the viral NSsprotein. We identify a crucial interaction between a ΩXaVmotif present in the NSs protein and the p62 subunit of thehost TFIIH. This interaction is required for nuclear filamentformation, NSs-dependent degradation of p62 and for viru-lence. This ΩXaV motif is also found in human proteins thatbind p62 and our results are an example of how viruses in-corporate simple motifs into their protein sequences to mimichuman proteins and enhance their functional capabilities inhost cells during infections.

Author contributions: N.C., C.d.l.F., L.L., K.K.-H., and J.G.O. designed research; N.C., C.d.l.F.,L.L., I.G., and P.R.C. performed research; N.C., C.d.l.F., and P.R.C. contributed new reagents/analytic tools; N.C., C.d.l.F., L.L., I.G., P.R.C., K.K.-H., and J.G.O. analyzed data; and N.C.,C.d.l.F., K.K.-H., and J.G.O. 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 2N0Y). The NMR chemical shifts have beendeposited in the BioMagResBank, www.bmrb.wisc.edu (accession no. 25540).1To whom correspondence may be addressed. Email: jg.omichinski@umontreal.ca orkkehnhal@gmu.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1503688112/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1503688112 PNAS | May 12, 2015 | vol. 112 | no. 19 | 6021–6026

BIOCH

EMISTR

Y

Despite its key role in RVFV virulence, there is currently littlemechanistic understanding of the interaction between the NSsprotein and the p62 subunit of TFIIH. In this manuscript, weidentify a short motif within the NSs that is homologous to theΩXaV motif (where Ω is Trp or Phe, X is any amino acid, a isAsp or Glu, and V is Val) that has been shown to mediate theinteraction of the p62/Tfb1 (human/yeast) subunit of TFIIH witha number of transcription and DNA repair factors in humansand yeast (24–27). Using biophysical, functional, and structuralstudies, we demonstrate that this ΩXaVmotif is required for NSsbinding to p62 as well as for formation of NSs nuclear filamentsand NSs-induced degradation of p62 in RVFV-infected cells.Taken together, our studies demonstrate that p62 binding to NSsis essential for forming nuclear filaments and suggest that thisinteraction plays a key role in p62 degradation and virulenceof RVFV.

ResultsThe NSs of RVFV Contains a p62-Binding Motif. It has been reportedthat infection with RVFV leads to proteasome-dependent deg-radation of the p62 subunit of TFIIH (22, 23). In addition, it wasshown that p62 degradation was dependent on the presence ofthe NSs protein and the two proteins can be coimmunoprecipi-tated from RVFV-infected cells (22). Given the fact that thepleckstrin homology (PH) domain of the p62/Tfb1 (human/yeast) subunit of TFIIH (p62PH/Tfb1PH) binds to a numberof human/yeast proteins involved in transcription activation andnucleotide excision repair (NER) that contain acidic/hydropho-bic motifs (ФXXФФ or ΩxaV; where Ф is hydrophobic) (24–31),we were interested to determine whether the NSs protein con-tains such a motif. Interestingly, we identify a region locatedat the very C terminus of NSs (NSsCT; residues 247–265) thatcontains a ΩXaV motif identified previously in TFIIEα (26, 27),XPG/Rad2 (24), and XPC/Rad4 (25) (Fig. 1A). Bioinformaticanalysis indicates that the presence of the ΩXaV motif is uniqueto the NSs of the RVFV among the NSs proteins of theBunyaviridae family (Fig. 1B and Fig. S1), and this is consistentwith the fact that the NSs proteins from these related viruseshave not been reported to either bind or degrade p62 (22, 23).

The ΩXaV Motif of NSsCT Binds to p62 and Tfb1. To verify that theregion containing the ΩXaV motif at the C terminus of NSs iscapable of binding to p62/Tfb1, isothermal titration calorimetry(ITC) studies were performed to determine dissociation con-stants (KD) for NSsCT binding to p62PH and Tfb1PH. The ITCstudies showed that NSsCT binds to both p62PH and Tfb1PHwith KD values of 10 nM and 90 nM, respectively (Fig. 2 A andB). Given that previous studies have shown both the aromaticand valine residues of the ΩXaV motif are crucial for binding ofp62PH/Tfb1PH to TFIIEα (26, 27), Rad2 (24), and Rad4 (25)(Fig. 1A), we prepared mutants (F261P, F261S, and V264S) ofNSsCT to test the importance of these residues for binding top62PH/Tfb1PH (Fig. 2B). Under the ITC conditions, mutatingF261 to serine (F261S) decreases binding by an order of mag-nitude, and we fail to observe any significant heat change witheither the F261P or the V264S mutant of NSsCT, suggestingthat their affinity for p62PH and Tfb1PH has decreased by atleast 2–3 orders of magnitude.

NSsCT Interacts with the PH Domain of p62 and Tfb1. Given thatNSsCT binds with high affinity to p62PH and Tfb1PH, NMRchemical shift perturbation studies were performed to determinethe region of NSsCT that interacts with p62PH/Tfb1PH. 1H-15Nheteronuclear single quantum coherence (HSQC) spectra wererecorded with 15N-labeled NSsCT both in the absence andpresence of either unlabeled p62PH (Fig. 2C, Left) or unlabeledTfb1PH (Fig. 2C, Right). A number of 1H and 15N chemical shiftchanges are observed for signals of NSsCT indicating the formationof complex with either p62PH or Tfb1PH and the exchange occurson the intermediate NMR time scale. The signals exhibiting the mostsignificant chemical shift changes {δΔ > 0.1; δΔ = [(0.17 ×ΔNH)

2 +

(ΔHN)2]1/2} are virtually identical following the addition of either

p62PH or Tfb1PH and they correspond to the residues in theΩXaV motif NSsCT (Fig. S2). To identify the NSsCT bindingsite on p62PH/Tfb1PH, 1H-15N HSQC spectra were recorded with15N-labeled Tfb1PH both in the absence and presence of un-labeled NSsCT. A number of 1H and 15N chemical shift changesare observed for signals of Tfb1PH and when mapped on thestructure of Tfb1PH, the signals exhibiting the most significantchanges are associated with residues in β-strands β5, β6, and β7and helix H1 of Tfb1PH (Fig. S2). The pattern of chemical shiftchanges is similar to the ones observed previously in the presenceof the NER factors Rad2 (24) and Rad4 (25), and NMR com-petition studies indicate that NSsCT competes with Rad4 for acommon binding site on Tfb1PH (Fig. S3).

Structure of the NSsCT–Tfb1PH Complex. To structurally charac-terize the complex between NSsCT and p62PH/Tfb1PH, we per-formed NMR structural studies on an NSsCT–Tfb1PH complex.Tfb1PH was chosen because it is structurally and functionallyhomologous to p62PH (32) and we have previously determinedstructures of Rad2–Tfb1PH (24) and Rad4–Tfb1PH (25) com-plexes. The structure of the NSsCT–Tfb1PH complex (PDB code2N0Y) is well defined by the NMR data (Table S1). The 20lowest-energy structures (Fig. 3A) are characterized by goodbackbone geometry, no significant restraint violation and lowpairwise RMSD values. In complex with NSsCT, the structure ofTfb1PH is virtually identical to its free form (ref. 33; PDB code1Y5O) showing a typical PH domain fold consisting of sevenβ-strands assembled in two sets of anti-parallels β-sheets (β1–4and β5–7) forming a sandwich, and an α-helix (H1) (Fig. 3A).NSsCT binds to Tfb1PH in an extended conformation with theregion between amino acids 259 and 265 of NSsCT formingthe binding interface (Fig. 3). This interface is consistent with theresidues of NSsCT exhibiting the most significant chemical shiftsupon addition of either Tfb1PH or p62PH (Fig. S2).

NSsCT–Tfb1PH Binding Interface. The NSsCT binding interface onTfb1PH contains two shallow clefts (Fig. 3 B and C) covering asurface area of 485 ± 25 Å2. The first cleft accommodates thephenyl ring of F261 from NSsCT, which makes contacts with thealiphatic side chains of residues Q49, T51, P52, M59, R61, andM88 from Tfb1PH (Fig. 3B). A second, larger cleft is formedproximal to the helix H1 of Tfb1, where V264 of NSsCT inserts.In this cleft, contacts are made between the side chains of resi-dues K101 and Q105 from Tfb1 and the side chain of V264 at thevery C-terminal end of the viral protein (Fig. 3C). In additionto the interactions involving the aromatic residue and the Valresidue of the ΩXaV motif, the structure also suggests that

Fig. 1. The ΩXaV motif at the C terminus of the NSs protein. (A) TheC-terminal region of NSs shows high homology to the one found in the NERfactors XPC/Rad4, and XPG/Rad2 as well as in TFIIEα. (B) The ΩXaV motif isunique to the NSs of RVFV and it does not appear to be in related Phlebo-viruses and this is consistent with the fact that p62 degradation is unique toRVFV. PTPV, Punta Toro virus; RVFV, Rift Valley fever virus; SFSV, Sandflyfever Sicilian virus; TOSV, Toscana virus.

6022 | www.pnas.org/cgi/doi/10.1073/pnas.1503688112 Cyr et al.

intermolecular electrostatic interactions are involved in facili-tating the interaction between NSsCT and Tfb1PH. These in-clude the possible pairing of the carboxyl group of the acidicresidue from the ΩXaVmotif of NSs (E263) with the ammoniumgroup of either K47 or K112 from Tfb1PH and the possiblepairing of the carboxyl group of D259 from NSsCT with theguanidinium group of either R61 or R86 from Tfb1PH (Fig. S4).Overall, the binding interface of the NSs–Tfb1PH complex is verysimilar to the binding interfaces of the Rad2–Tfb1PH (24), Rad4–Tfb1PH (25) and TFIIEα-p62PH complexes (26), and clearlydemonstrates how the ΩXaV motif in NSs is able to mimic thebinding of these factors to the p62/Tfb1 subunit of TFIIH(Fig. S5).

Mutation of the ΩXaV Motif Attenuates NSs Function During RVFVInfection. To examine what effect disruption of the ΩXaV motifhas on NSs function in RVFV-infected cells, recombinant(r)MP12 viruses containing mutations at F261 of NSs were gener-ated. A triple tandem Flag tag (3xFlag) is incorporated at the Cterminus of the NSs protein to allow for monitoring NSs ex-pression and nuclear filament formation. In agreement with aprior report (14), the C-terminal tag is well tolerated; the rMP12virus containing wild-type NSs-3xFlag achieves infectious virustiters (as determined by plaque assay) similar to the untaggedNSs. However, when F261 of NSs is mutated to either a serine(F261S) or proline (F261P) residue the resulting viruses retaininfectivity but lose their ability to form plaques (Table S2).To examine this defect further, HSAECs cells (an IFN com-

petent cell line) were infected at a multiplicity of infection(MOI) of 3 and infectious viral titers determined by immuno-staining (focus formation assay). By 24 h post infection (hpi), boththe F261S and F261P mutant viruses display about 1 log decreasein titers similar to rMP12 with deletion of NSs (ΔNSs; Fig. 4).Although NSs has been shown to be dispensable for viral repli-cation, loss of NSs as an IFN antagonist can result in decreased

infectious titers (34). The similar decrease in infectious viral titersfor both the mutants and the ΔNSs control is also consistent withthe mutations in the ΩXaV motif disrupting NSs function. Toassess whether the observed decrease in viral titers is due to analteration of NSs localization, nuclear and cytoplasmic extractswere isolated and probed for NSs expression (Fig. S6). Althoughwild-type NSs demonstrates slightly higher proportion within thenuclear fraction, a near equal distribution between the fractions isobserved for both NSs mutants, indicating a significant portion ofthe mutant NSs proteins translocate to the nucleus. However,nuclear filament formation is abrogated with diffuse stainingbeing observed with both NSs mutants (F261S/P; Fig. 5). In ad-dition, p62 and PKR protein levels in the NSs mutant infectedcells are similar to mock or rMP12 ΔNSs infected cells (Fig. S7and SI Methods). Collectively, these data indicate that mutationof the F261 within the ΩXaV motif is sufficient to disrupt NSsfunction in viral-infected cells including the formation of NSsnuclear filaments, NSs induced p62 degradation and reduction ofthe overall level of virulence.

DiscussionOne crucial concern associated with outbreaks of RVFV in-fections is that, like Ebola Virus infections, they can progress tofatal hemorrhagic fever in both humans and livestock (3–5). Akey viral component in determining the virulence of RVFV in-fection is the NSs protein, due to its ability to enter the nucleusof infected cells, induce nuclear filament formation and disruptthe function of the general transcription factor TFIIH (18–21).In this manuscript, we structurally and functionally characterizethe interaction between the p62 subunit of TFIIH and the RVFVNSs protein. Based on sequence alignments with other factorsknown to bind to either p62 or its yeast homolog Tfb1 (24-31), weidentify a ΩXaV motif at the C-terminal region of NSs (NSsCT).Using ITC and NMR studies, we demonstrate that NSsCT bindswith high affinity to both Tfb1 and p62, and that the structure

Fig. 2. The acidic domain of NSs binds p62PH and Tfb1PH with high affinity. (A) Representative ITC thermogram obtained by successive addition of NSsCT toTfb1PH. Experiments are performed at 25 °C and the results fit to a single-binding site model with 1:1 stoichiometry. (B) Comparison of the KD values for thebinding of NSsCT and its mutants (F261S, F261P, and V264S) to Tfb1PH and p62PH. ND indicates that no heat of interaction was detected under the conditionstested for these mutants. (C) Overlay of the 1H-15N HSQC spectra for a 0.2 mM sample of 15N-labeled NSsCT in the absence (black) or presence of either 0.3 mMunlabeled p62PH (Left, cyan) or Tfb1PH (Right, marine).

Cyr et al. PNAS | May 12, 2015 | vol. 112 | no. 19 | 6023

BIOCH

EMISTR

Y

of an NSsCT–Tfb1PH complex is very similar to structures ofRad2–Tfb1PH (24), Rad4–Tfb1PH (25), and TFIIEα–p62PH(26) complexes. In addition, functional studies in RVFV-infectedcells show that the ΩXaV motif of NSs is essential both fornuclear filament formation, degradation of p62 and virulence.These results suggest that binding of NSs to the p62 subunit ofTFIIH is an essential step in nuclear filament formation, dis-ruption of the TFIIH complex, and the virulence associated withRVFV infection.One crucial step in the formation of nuclear filaments in

RVFV-infected cells appears to be the correlation between NSs

synthesis and the subsequent decrease in cellular p62 levels (21–23). In a recent study, FBXO3 was identified as the E3 ligaseresponsible for ubiquitin-mediated degradation of p62 in RVFV-infected cell expressing NSs (23). Interestingly, it was also de-termined that p62 was not a substrate for FBXO3 in uninfectedcells, but that FBXO3 targeted p62 specifically in RVFV-infectedcells. In addition, it was demonstrated that NSs recruits FBXO3to nuclear filaments through a direct protein-protein interaction,but NSs fails to induce p62 degradation in FBXO3 deficientcells. Importantly, p62 was detected in the nuclear filaments ofthe RVFV-infected cells depleted of FBXO3, but absent in thefilaments of infected cells with normal FBXO3 levels (23). Incombination with our results demonstrating that the ΩXaVmotifof NSs is required for both p62 degradation and nuclear filamentformation, this FBXO3 requirement suggests that NSs recruitsboth p62 and FBXO3 to nuclear filaments through direct pro-tein–protein interactions and that leads to FBXO3-dependentdegradation of p62 via the ubiquitin–proteosome system.RVFV-infected cells can be distinguished from cells infected

by other Bunyaviridae-family viruses due to the presence of nu-clear filaments and their formation correlates with the onset ofNSs synthesis (18–20). Our results demonstrating that the ΩXaVmotif present in the NSs protein is required for nuclear filamentformation following RVFV infection are consistent with the factthat this motif is absent in the NSs proteins of other Bunyavir-idae-family viruses. In previous studies, it was shown that NSsdirectly interacts with the p62 subunit of TFIIH, but only at timepoints shortly after the initiation of viral infection (22). Thisearly effect is thought to be due to the fact that p62 levelsdecrease substantially following RVFV infection in an NSs-dependent manner. In combination with the results in FBXO3-depleted cells (23), our results demonstrating that the ΩXaVmotif is required for nuclear filament formation strongly suggestthat NSs binding to the p62 subunit is the initial step leading tothe disruption of TFIIH following RVFV infection. Such amechanism would help explain earlier studies examining thesequestration of the p44 subunit of TFIIH in nuclear filamentformation (21). In these studies, it was determined that NSs

Fig. 3. NMR Structure of the NSsCT–Tfb1PH complex. (A) Overlay of the 20lowest-energy structures of the complex between NSsCT (pink) and Tfb1PH(marine). (B) Ribbon representation of NSsCT (pink) and Tfb1PH (marine)highlighting the side chains (in sticks) of Tfb1PH (Q49, T51, P52, M59, R61,and M88) that interact with the aromatic ring of Phe261 (F261) of NSsCT.(C) Ribbon representation of NSsCT (pink) and Tfb1PH (marine) highlightingthe side chains (in sticks) of Tfb1PH (K101 and Q105) that interact with theside chain of Val264 (V264) of NSsCT.

Fig. 4. The ΩXaV motif of NSs is important for RVFV virulence. HSAECs wereinfected at an MOI 3 with rMP12 expressing wild-type NSs-3xFlag (WT),deletion of NSs (ΔNSs), or NSs-3xFlag mutants (F261S or F261P). Infectiousvirus titers were determined by focus formation assay. Data from three in-dependent experiments was plotted as a mean ± SD of focus forming unitsper mL (ffu/mL). Statistical significance determined by one-way analysis ofvariance (ANOVA) between wild-type NSs containing virus versus ΔNSs,F261S and F261P was observed (****P < 0.0001).

6024 | www.pnas.org/cgi/doi/10.1073/pnas.1503688112 Cyr et al.

interacts directly with the p44 subunit of TFIIH and that p44 isan important component of the nuclear filaments. However,the interaction between NSs and p44 only occurred when p44was isolated from the TFIIH complex. This result seemed toindicate that the NSs–p44 interaction formed subsequent todisruption of the TFIIH complex. Together, the results suggesta mechanism where NSs binding to p62 leads to its degradationby FBXO3 in nuclear filaments. The degradation of p62 causesthe disruption of the TFIIH complex, and leads to the se-questration of the p44 subunit in nuclear filaments. This disruptionresults in a global suppression of transcription in RVFV-infectedcells due to decreased levels of transcriptionally competent TFIIH.Our structural and functional studies demonstrate how NSs

binds to p62 through an interaction with a short intrinsicallydisordered protein region containing a ΩXaV motif and how thisinteraction is required for nuclear filament formation, p62 degra-dation and virulence associated with RVFV infection. Interestingly,a very similar motif has also been identified in transcription regu-latory factors (TFIIEα) and NER factors (Rad2 and Rad4) thatbind to either p62 or Tfb1 and regulate TFIIH function. Given thatTFIIH plays key roles in regulating both transcription activation andNER (35), this would indicate that the RVFV NSs proteins hasevolved to specifically inhibit the function of TFIIH. To accomplishthis, a simple motif has been incorporated into its sequence thatenables NSs to mimic host cell proteins that bind the p62 subunit ofTFIIH. Many mammalian proteins that regulate transcription andDNA repair use intrinsically disordered protein regions to partici-pate in highly dynamic protein–protein complexes (36), and thisenables them to interact with multiple partners. Our results suggestthat NSs induces nuclear filaments using similar mechanisms withintrinsically disordered protein regions. Future studies will be per-formed to determine the dynamic interchange between NSs, p62,and FBXO3 in nuclear filaments, and how their interactions leadto p62 degradation, disruption of the TFIIH complex and enhancevirulence in RVFV-infected cells.

MethodsPlasmids and Cloning. The DNA coding sequences for p62PH, Tfb1PH, andRad476–115 were cloned as described (25, 30, 33, 37). The DNA sequencecoding for the C-terminal of RVFV NSs (NSsCT, residues 247–265) was synthe-sized and ligated between the BamHI and EcoRI restriction sites of pGEX-2T.Point mutations were made using QuikChange II site-directed mutagenesis kit(Stratagene). For expression of Flag-tagged NSs, a triple tandem repeat Flagtag with serine and alanine spacer (SA-DYDDDDK) was cloned between thePstI and BsaA1 sites within the pProT7-S(+) plasmid. The resulting plasmid,

designated pProT7-S(+) NSs-3xFlag, was used to generate NSs mutations (F261Sand F261P). All coding sequences were verified by DNA sequencing.

Recombinant Proteins Expression and Purification. p62PH, Tfb1PH, NSsCT, andRad476–115 were expressed and purified as described (25, 33, 37). NSsCT andmutants were expressed as GST-fusion proteins in Escherichia coli, purifiedover GSH-Sepharose resin (GE Healthcare), and cleaved with thrombin(Calbiochem). Following cleavage, NSsCT peptide was further purified over aQ-Sepharose (GE Healthcare) column and dialyzed into appropriate buffersfor isothermal titration calorimetry (ITC) and NMR studies. 15N-labeled and15N/13C-labeled proteins were expressed in M9-minimal media containing15NH4Cl and/or

13C6-glucose (Sigma-Aldrich) as the sole nitrogen and carbonsources and purified as described above.

ITC Experiments. ITC titrations were performed on a MicroCal VP-ITC (Microcal)at 25 °C in 20 mM Tris·HCl (pH 7.4). All titrations fit a single-binding sitemechanism with 1:1 stoichiometry and values are the average of two or moreseparate experiments.

NMR Experiments. All NMR experiments were carried out at 300 K on VarianUnity Inova 500 and 600 MHz spectrometers equipped with z-pulsed-fieldgradient units and triple resonance probes. All proteins were dissolved in 20 mMsodium phosphate pH 6.5, 2 mM DTT and either 10% (vol/vol) D2O/90% (vol/vol)H20, or 100%D2O. The

1H, 15N, and 13C resonances for NSsCT and Tfb1PH wereassigned as reported for free Tfb1PH (33). Interproton distance restraints weremeasured from 3D 15N-edited NOESY-HSQC and 13C-edited HSQC-NOESYspectra with a τm = 90 ms. Intermolecular distance restraints were obtainedfrom 3D 15N/13C F1-filtered, F3-edited NOESY experiment with a τm = 90 ms.The NMR data were processed with NMRPipe/NMRDraw (38) and analyzedwith Analysis from the CCPNMR suite (39). For details, see SI Methods.

Structure Calculations. The structure of the NSsCT–Tfb1PH complex was cal-culated using ARIA 2.3 (40) and refined with CNS (41). Backbone dihedralangles were derived with TALOS+ for NSsCT and TALOS-N for Tfb1PH (42). Thequality of the 20 lowest energy structures was analyzed using PROCHECK-NMR (43) and CNS. The structure was deposited in the PDB (PDB 2N0Y) andBMRB (BMRB 25540).

Cell Culture. BSR-T7/5, a BHK-21 cell clone stably expressing T7 RNA poly-merase (44) was cultured in MEM media containing 1% L-glutamine, 1%penicillin/streptomycin, 7.5% (vol/vol) heat inactivated FBS and 500 μg/mLgeneticin. Vero cells were maintained in DMEM containing 1% L-glutamine, 1%penicillin/streptomycin, 10% (vol/vol) FBS. The human small airway epithelial cells(HSAECs) were grown in Ham’s F-12 media supplemented with 1% L-glutamine,1% penicillin/streptomycin, 1% nonessential amino acids, 1% sodium pyruvate,10% (vol/vol) FBS, and 0.1% 55 mM β-mercaptoethanol (Life Technologies;ref. 45). All cells were maintained at 37°C in a humidified 5% (vol/vol) CO2

atmosphere.

Fig. 5. The ΩXaV motif of NSs is required for nuclearfilament formation in RVFV-infected cells. HSAECswere infected at an MOI 3 with rMP12 expressing wild-type NSs-3xFlag (WT), deletion of NSs (ΔNSs), or NSs-3xFlag mutants (F261S or F261P). Cells were probed forNSs-3xFlag (green) and DAPI (nuclear counterstain;blue). A merged twofold enlargement (far right panel)of a representative area is provided and white arrow-heads indicate the location of nuclear filaments withthe wild-type NSs.

Cyr et al. PNAS | May 12, 2015 | vol. 112 | no. 19 | 6025

BIOCH

EMISTR

Y

RVFV Viral Stocks. Recombinant MP12 viruses were rescued by transfection ofBSR-T7/5 cellswith the following plasmids: pProT7-M(+), pProT7-L(+), pProT7-S(+),pT7-IRES-vN, pT7-IRES-vL, and pCAGGS-vG (34, 46). To generate an initialseed stock, cells (seeded at 3 × 106 cells per 75cm2 flask) were transfectedwith pProT7-M(+), pProT7-L(+), pProT7-S(+), pT7-IRES-vN and pT7-IRES-vL,and pCAGGS-vG. For further details, see SI Methods. To generate a P1 viralstock, subconfluent Vero monolayers were infected at an MOI of 0.1. Twodays post infection, when cytopathic effect was observed within the culture,media supernatants were harvested twice and stored at 4 °C. After the lastcollection, supernatants were pooled together, filtered (0.2 μM), and storedat −80°C in aliquots.

Focus Formation Assay. Vero cells seeded in 12-well plates were infected in alimiting dilution assay. After 1-h incubation, an overlay (1:1 ratio) comprisedof 2.4% (wt/vol) Avicel (RC-591 NF; FMC BioPolymer) and 2× EMEM media[containing 1% L-glutamine, 2% (vol/vol) penicillin/streptomycin, 1% non-essential amino acids, 1% sodium pyruvate, and 5% (vol/vol) FBS] was addedto wells. After 24 h post infection (hpi), overlay was aspirated and cells,washed once with D-PBS, and then formalin [10% (vol/vol)] fixed. Cells werepermeabilized with D-PBS-0.2% Triton-X100 and blocked with 3% (wt/vol)BSA in D-PBS. The α-RVFV N mouse monoclonal antibody (clone 1D8; 1:1,000in blocking buffer) was incubated with cells for 2 h at room temperature.Afterward, cells were washed and secondary Alexa Fluor 488-conjugated

antibody (1:1,000) was used to label primary antibody signal. Using EVOS FLCell Imaging Station (Life Technologies), the number of foci containing morethan two cells was counted to determine average focus forming units (ffu)/mL.

Immunofluorescence Analysis. HSEACs were grown on coverslips in a six-wellplate, infected with rMP12 viruses at an MOI 3. After 24 hpi, cells were washedwith PBS (−Ca and −Mg) then fixed with 4% (wt/vol) paraformaldehyde. Cellswere permeabilized, blocked and probed with primary Flag antibody (1:1,000diluted M2 clone; Sigma Aldrich). After 24 h, primary antibody was probedwith Alexa Fluor 488 donkey anti-goat secondary antibody (1:1,000; Invi-trogen). DAPI (4’,6-diamidino-2-phenylindole), diluted 1:1,000, was used to vi-sualize nuclei. Coverslips were mounted to glass slides using Fluoromount G(Southern Biotech). Images were viewed using an oil-immersion 60× objectivelens on a Nikon Eclipse TE 2000-U. Each sample was subjected to at least fiveimages, with a representative image shown. Images were subjected to four-lineaveraging and processed using Nikon NIS-Elements AR Analysis 3.2 software.

ACKNOWLEDGMENTS. We thank Drs. Ursula Buchholz, Connie Schmaljohn,and Shinji Makino for providing BSR-T7/5 cells, RVFV N antibody, and re-combinant MP12 reverse genetics system, respectively. This work was supportedby the Canadian Institutes for Health Research (130414; to J.G.O.) and the US-National Institutes of Health (R15 AI100001; to K.K.-L.). L.L. is a postdoctoralfellow of the NSERC CREATE program.

1. Weber F, Elliott RM (2002) Antigenic drift, antigenic shift and interferon antagonists:How bunyaviruses counteract the immune system. Virus Res 88(1-2):129–136.

2. Soldan SS, González-Scarano F (2005) Emerging infectious diseases: The Bunyaviridae.J Neurovirol 11(5):412–423.

3. Pepin M, Bouloy M, Bird BH, Kemp A, Paweska J (2010) Rift Valley fever virus(Bunyaviridae: Phlebovirus): An update on pathogenesis, molecular epidemiology,vectors, diagnostics and prevention. Vet Res 41(6):61.

4. Bouloy M, Weber F (2010) Molecular biology of Rift Valley fever virus. Open Virol J 4:8–14.5. Bird BH, Ksiazek TG, Nichol ST, Maclachlan NJ (2009) Rift Valley fever virus. J Am Vet

Med Assoc 234(7):883–893.6. Hartley DM, Rinderknecht JL, Nipp TL, Clarke NP, Snowder GD; National Center for Foreign

Animal and Zoonotic Disease Defense Advisory Group on Rift Valley Fever (2011) Potentialeffects of Rift Valley fever in the United States. Emerg Infect Dis 17(8):e1.

7. Ikegami T, Makino S (2011) The pathogenesis of Rift Valley fever. Viruses 3(5):493–519.8. Bouloy M, et al. (2001) Genetic evidence for an interferon-antagonistic function of

Rift Valley fever virus nonstructural protein NSs. J Virol 75(3):1371–1377.9. Bird BH, et al. (2008) Rift Valley fever virus lacking the NSs and NSm genes is highly

attenuated, confers protective immunity from virulent virus challenge, and allows fordifferential identification of infected and vaccinated animals. J Virol 82(6):2681–2691.

10. Vialat P, Billecocq A, Kohl A, Bouloy M (2000) The S segment of Rift Valley feverphlebovirus (Bunyaviridae) carries determinants for attenuation and virulence inmice. J Virol 74(3):1538–1543.

11. Billecocq A, et al. (2004) NSs protein of Rift Valley fever virus blocks interferon pro-duction by inhibiting host gene transcription. J Virol 78(18):9798–9806.

12. Le May N, et al. (2008) A SAP30 complex inhibits IFN-beta expression in Rift Valleyfever virus infected cells. PLoS Pathog 4(1):e13.

13. Habjan M, et al. (2009) NSs protein of rift valley fever virus induces the specific degra-dation of the double-stranded RNA-dependent protein kinase. J Virol 83(9):4365–4375.

14. Ikegami T, et al. (2009) Rift Valley fever virus NSs protein promotes post-transcrip-tional downregulation of protein kinase PKR and inhibits eIF2alpha phosphorylation.PLoS Pathog 5(2):e1000287.

15. Ikegami T, et al. (2009) Dual functions of Rift Valley fever virus NSs protein: Inhibitionof host mRNA transcription and post-transcriptional downregulation of proteinkinase PKR. Ann N Y Acad Sci 1171(Suppl 1):E75–E85.

16. Baer A, et al. (2012) Induction of DNAdamage signaling upon Rift Valley fever virus infectionresults in cell cycle arrest and increased viral replication. J Biol Chem 287(10):7399–7410.

17. Austin D, et al. (2012) p53 Activation following Rift Valley fever virus infection con-tributes to cell death and viral production. PLoS ONE 7(5):e36327.

18. Struthers JK, Swanepoel R (1982) Identification of a major non-structural protein inthe nuclei of Rift Valley fever virus-infected cells. J Gen Virol 60:381–384.

19. Struthers JK, Swanepoel R, Shepherd SP (1984) Protein synthesis in Rift Valley fevervirus-infected cells. Virology 134(1):118–124.

20. Swanepoel R, Blackburn NK (1977) Demonstration of nuclear immunofluorescence inRift Valley fever infected cells. J Gen Virol 34(3):557–561.

21. Le May N, et al. (2004) TFIIH transcription factor, a target for the Rift Valley hem-orrhagic fever virus. Cell 116(4):541–550.

22. Kalveram B, Lihoradova O, Ikegami T (2011) NSs protein of rift valley fever virus promotesposttranslational downregulation of the TFIIH subunit p62. J Virol 85(13):6234–6243.

23. Kainulainen M, et al. (2014) Virulence factor NSs of Rift Valley fever virus recruits theF-box protein FBXO3 to degrade subunit p62 of general transcription factor TFIIH.J Virol 88(6):3464–3473.

24. Lafrance-Vanasse J, et al. (2012) Structural and functional characterization of in-teractions involving the Tfb1 subunit of TFIIH and the NER factor Rad2. Nucleic AcidsRes 40(12):5739–5750.

25. Lafrance-Vanasse J, Arseneault G, Cappadocia L, Legault P, Omichinski JG (2013)Structural and functional evidence that Rad4 competes with Rad2 for binding to theTfb1 subunit of TFIIH in NER. Nucleic Acids Res 41(4):2736–2745.

26. Okuda M, et al. (2008) Structural insight into the TFIIE-TFIIH interaction: TFIIE and p53share the binding region on TFIIH. EMBO J 27(7):1161–1171.

27. Di Lello P, et al. (2008) p53 and TFIIEalpha share a common binding site on theTfb1/p62 subunit of TFIIH. Proc Natl Acad Sci USA 105(1):106–111.

28. Okuda M, Nishimura Y (2014) Extended string binding mode of the phosphorylatedtransactivation domain of tumor suppressor p53. J Am Chem Soc 136(40):14143–14152.

29. Langlois C, et al. (2012) Structure-based design of a potent artificial transactivationdomain based on p53. J Am Chem Soc 134(3):1715–1723.

30. Langlois C, et al. (2008) NMR structure of the complex between the Tfb1 subunit ofTFIIH and the activation domain of VP16: Structural similarities between VP16 andp53. J Am Chem Soc 130(32):10596–10604.

31. Chabot PR, et al. (2014) Structural and functional characterization of a complex be-tween the acidic transactivation domain of EBNA2 and the Tfb1/p62 subunit of TFIIH.PLoS Pathog 10(3):e1004042.

32. Gervais V, et al. (2004) TFIIH contains a PH domain involved in DNA nucleotide ex-cision repair. Nat Struct Mol Biol 11(7):616–622.

33. Di Lello P, et al. (2005) NMR structure of the amino-terminal domain from the Tfb1subunit of TFIIH and characterization of its phosphoinositide and VP16 binding sites.Biochemistry 44(21):7678–7686.

34. Ikegami T, Won S, Peters CJ, Makino S (2006) Rescue of infectious Rift Valley fevervirus entirely from cDNA, analysis of virus lacking the NSs gene, and expression of aforeign gene. J Virol 80(6):2933–2940.

35. Compe E, Egly JM (2012) TFIIH: When transcription met DNA repair. Nat Rev Mol CellBiol 13(6):343–354.

36. Liu J, Faeder JR, Camacho CJ (2009) Toward a quantitative theory of intrinsicallydisordered proteins and their function. Proc Natl Acad Sci USA 106(47):19819–19823.

37. Di Lello P, et al. (2006) Structure of the Tfb1/p53 complex: Insights into the interactionbetween the p62/Tfb1 subunit of TFIIH and the activation domain of p53.Mol Cell 22(6):731–740.

38. Delaglio F, et al. (1995) NMRPipe: A multidimensional spectral processing systembased on UNIX pipes. J Biomol NMR 6(3):277–293.

39. Vranken WF, et al. (2005) The CCPN data model for NMR spectroscopy: Developmentof a software pipeline. Proteins 59(4):687–696.

40. Rieping W, et al. (2007) ARIA2: Automated NOE assignment and data integration inNMR structure calculation. Bioinformatics 23(3):381–382.

41. Brünger AT, et al. (1998) Crystallography & NMR system: A new software suite for mac-romolecular structure determination. Acta Crystallogr D Biol Crystallogr 54(Pt 5):905–921.

42. Shen Y, Bax A (2013) Protein backbone and sidechain torsion angles predicted fromNMR chemical shifts using artificial neural networks. J Biomol NMR 56(3):227–241.

43. Laskowski RA, Rullmannn JA, MacArthur MW, Kaptein R, Thornton JM (1996) AQUAand PROCHECK-NMR: Programs for checking the quality of protein structures solvedby NMR. J Biomol NMR 8(4):477–486.

44. Buchholz UJ, Finke S, Conzelmann KK (1999) Generation of bovine respiratory syn-cytial virus (BRSV) from cDNA: BRSV NS2 is not essential for virus replication in tissueculture, and the human RSV leader region acts as a functional BRSV genome pro-moter. J Virol 73(1):251–259.

45. Popova TG, et al. (2010) Reverse-phase phosphoproteome analysis of signalingpathways induced by Rift Valley fever virus in human small airway epithelial cells.PLoS ONE 5(11):e13805.

46. Kalveram B, Lihoradova O, Indran SV, Ikegami T (2011) Using reverse genetics tomanipulate the NSs gene of the Rift Valley fever virus MP-12 strain to improve vaccinesafety and efficacy. J Vis Exp (57):e3400.

6026 | www.pnas.org/cgi/doi/10.1073/pnas.1503688112 Cyr et al.