Introduction

Yellow fever virus (YFV) is a member of the genus Flavivirus, family Flaviviridae. In addition to YFV, flaviviruses that are significant threats to human health include dengue virus (DENV), Saint Louis encephalitis virus (SLEV), West Nile virus (WNV), Japanese encephalitis virus (JEV), and tick-borne encephalitis virus (TBEV). YFV is endemic in the central regions of Africa and South America, where it exists primarily as a zoonosis in monkeys. Occasionally, the virus is transmitted to human populations, resulting in epidemics with mortality rates of up to 60% [3]. The YFV genome is a single positive-stranded RNA molecule of approximately 11 kb in size with a type I cap structure at the 5’-end but lacking a 3’-polyadenylate tail [38]. The single open reading frame, flanked by a 5′ untranslated region (UTR) and a 3′ UTR, encodes a polyprotein that is processed by viral and cellular proteases into three structural proteins (C, prM and E) and seven nonstructural proteins (NS1, NS2a, NS2b, NS3, NS4a, NS4b, and NS5) [21].

The NS5 protein, with a molecular weight of 103 kDa, is the largest and most highly conserved of the flaviviruses proteins. NS5 has two separate activities encoded in distinct domains containing nuclear localization sequences. The N-terminal 300 amino acids of NS5 represent the methyltransferase domain, which methylates the cap structure, and the C-terminal 605 amino acids contain the RNA-dependent RNA polymerase (RdRp) domain, which plays an important role in amplification of viral RNAs [reviewed in ref. 6].

RNA replication occurs in an endoplasmic reticulum membrane-associated replication complex that comprises both viral and cellular protein(s). The two major components of this complex are the NS3 and NS5 proteins, providing vital enzymatic properties for RNA replication [22]. During its replication, the viruses also recruit cellular factors that are normal components of the cellular RNA processing or translation machineries [18]. Most of these factors are not identified, so the identification of cellular proteins that interact with NS proteins would provide insights into the mechanism of YFV replication and pathogenesis. In recent years, it has been reported that the flavivirus NS5 protein interacts with a range of host proteins for virus replication. For example, cyclophilins and the La protein bind to NS5 and are involved in RNA transcription and translation [10, 35, 43]. NS5 can interact with cellular factors related to interferon pathways, cell proliferation, and junction permeability, suggesting that NS5 may be involved in viral pathogenesis [1, 8, 20, 30, 42].

In this study, we investigated protein-protein interactions between proteins from a HeLa cell cDNA library and the NS5-RdRp domain of YFV using the yeast two-hybrid assay. The screen resulted for the first time in the isolation of an YFV-NS5-interacting partner, the small nuclear ribonucleoprotein (snRNP)-specific polypeptide A (U1A).

U1A is a 282-amino-acid protein associated with the U1 snRNP, a small RNA-protein complex involved in pre-mRNA splicing [24]. U1A can be found in a snRNP-free form called SF-A that is complexed with a set of non-snRNP proteins such as polypyrimidine tract binding protein (PTB), polypyrimidine-tract binding protein-associated factor (PSF), and non-POU-domain-containing, octamer-binding (NONO) and DEAD (Asp-Glu-Ala-Asp) box polypeptide 5 (DDX5), which are critical for RNA 3’-end processing, specifically cleavage, splicing and polyadenylation [19, 25, 26, 32]. Furthermore, our results confirmed the NS5-U1A interaction, both in vitro by GST-pulldown, and in vivo by co-immunoprecipitation assays. The domain responsible for this interaction was found to be conserved among some flaviviruses of medical importance.

Materials and methods

Cells and virus

Vero E6 cells were grown in Eagle’s minimum essential medium (MEM) supplemented with 10% of fetal bovine serum, 100 U/ml of penicillin, and 100 mg/ml of streptomycin. These cells were maintained at 37°C in a humidified atmosphere containing 5% CO2. A stock of YFV vaccine strain 17DD (Fiocruz, Brazil) was propagated and titrated (107 PFU/ml) using standard protocols.

Plasmids

The plasmid pGBKT7-RdRp, a bait for yeast-two hybrid screening, was created by cloning the region encoding the RdRp domain of YFV NS5 into the yeast expression vector pGBKT7 (Clontech) as a fusion to the GAL4 DNA-binding domain (DNA-BD). The fragment EcoRI-PstI RdRp of YFV NS5 was obtained from pACNR/FLYF-17Dx (provided by Dr. Charles Rice, Rockfeller University), a plasmid that contains the full-length infectious cDNA clone of the vaccine strain YF-17D, by PCR using the forward primer 5’-TATGAATTCAATGACAACCCTACAGGACC-3’ and the reverse primer 5’-TATCTGCAGTCAGATAAGCTCACCCAGTTGC-3’. The deleted constructions of RdRp, used for mapping the binding domain, were cloned into pGBKT7 in the same way as for pGBKT7-RdRp construction, but using different primers (available upon request from the authors). ID-GST, a plasmid used in the GST pulldown assay, was constructed by cloning a 222 bp fragment from pACNR/FLYF-17Dx into pGEX-4T-1 (Addgene), in a fusion with the GST protein. The fragment EcoRI-PstI ID from RdRp of YFV-NS5 was obtained by PCR using the forward primer 5’-TATGAATTCGCGGGAACTAGGAAGATC-3’ and reverse primer 5’-TATCTGCAGTCAACACCTGCCTTGTTGGTG-3’. The fragment ID from RdRp of DENV3-NS5 was obtained by PCR using the forward primer 5’- TATGGATCCTATGGCTTTGGAGAACCCTGGG-3’ and the reverse primer 5’-TCCTGCGAACACTTGCCCAATTTGTGGAGTTC-3’ for cloning into pGBKT7. The plasmids pGBKT7-IDDENV4-NS5 and pGBKT7-IDSLEV-NS5 were constructed in the same way but using the primers 5’-TATGAATTCTGGCTGTGGGCCCTCCTTGG-3’ (forward) and 5’-TTACTGCAGACATTTCCCTTCTTGGTGCAGAGCC-3’ (reverse) for ID DENV4-NS5 and 5’-TATGGATCCTATGGCTGTGGGACTTCGTTGC-3’ (forward) and 5’-CTCCTGCAGACACTCTCCTTTAAGATGGGCTTC-3’ (reverse) for ID SLEV-NS5. The plasmid NS5-pEGFP was created by cloning the YFV NS5 gene into pEGFP-C3 (Clontech). The fragment HindIII-XhoI NS5 of YFV was obtained from pACNR/FLYF-17Dx by PCR using the forward primer 5’-TATAAGCTTTCAGATAAGCTCACCCA-3’ and the reverse primer 5’-TATAAGCTTTCAGATAAGCTCACCCAG-3’. U1A-pRSET, the plasmid encoding the histidine tag, and U1A-pcNMS2FLAG were provided by Carol S. Lutz, New Jersey Medical School.

Protein expression and purification

E. coli BL(DE3)21 cells (Novagen) transformed with ID-GST, GST and U1A-pRSET plasmids were treated with isopropyl-1-thio-β-D-galactopyranoside at a final concentration of 1 mM at 37°C for 3 h to induce protein expression. A crude extract was prepared using lysozyme (Sigma) treatment and sonication. The GST proteins were purified on a glutathione Sepharose 4B column (Amersham Biosciences), and the HIS-fusion protein was purified on a HisTrap HP column (GE Healthcare). Fractions containing the protein were concentrated using a Centricon filter device (Millipore), quantified by the Bradford method using the Bio Rad Protein Assay (Bio Rad) and stored at −80°C.

Yeast two-hybrid assay

To identify proteins that interact with YFV NS5, a yeast two-hybrid screen was performed using MATCHMAKER GAL4 Two-Hybrid System 3 (Clontech). The bait plasmid pGBKT7-RdRp was introduced into Saccharomyces cerevisiae strain AH109 (HIS3, ADE2 and lacZ reporter genes) by transformation and then mated with the prey AD vector (pACT2)/HeLa cell cDNA library according to the manufacturer’s protocol (Clontech). Candidates for two-hybrid interaction were selected on synthetic dropout (SD) medium lacking His, Leu, Trp, and further on SD medium lacking Ade, His, Leu, Trp. Plasmid DNA was isolated from the positive clones and sequenced. To confirm the interaction, the recovered candidate clones were reintroduced into the same yeast strain and again selected on SD medium.

GST-pulldown assay

A 50% slurry of glutathione Sepharose 4B beads (Amersham Biosciences) equilibrated with PBS (140 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, pH 7.9 and protease inhibitor [Roche]) was incubated with 0.5 μg of ID-GST or GST in 0.5 ml of PBS for 1 h at 4ºC with gentle agitation. The beads were washed five times with NETN+ buffer (100 mM NaCl, 20 mM Tris pH 7.4, 1 mM EDTA, 0.5% Nonidet P-40, 1 mM DTT and protease inhibitor) and mixed with 10.2 μg of U1A-pRSET for 2 h at 4ºC with gentle agitation. The beads were washed five times with NETN+ buffer and treated with 10 μg of RNase A (Invitrogen) in NETN+ buffer at room temperature for 30 min with gentle agitation. After washing, the beads were eluted by boiling in SDS-PAGE loading buffer, separated by 10% SDS-PAGE, and transferred to a PVDF membrane. After blocking with 10% skim milk, the membrane was incubated either with anti-HIS antibody (Invitrogen) or with anti-GST antibody (GE Healthcare) followed by probing with secondary horseradish-peroxidase-conjugated antibody (Zymed). The protein bands were detected using SuperSignal West Pico Chemiluminescent Substrate (Pierce).

Co-immunoprecipitation assay

Vero cells that were 95% confluent were transfected with U1A-pcNMS2FLAG using Lipofectamine 2000 (Invitrogen) as instructed by the manufacturer. After 24 h, the cells were infected with YFV-17DD at an MOI of 3. Forty-eight hours postinfection, the cells were washed with PBS and scraped. Detached cells were collected by centrifugation and incubated with lysis buffer (150 mM NaCl, 50 mM Tris-HCl, pH 7.4, 1 mM EDTA, 1% Triton X-100 and protease inhibitors) for 30 min on ice. Cell debris was removed by centrifugation at 12,000g for 10 min at 4°C. Forty microliters of a 50% slurry of anti-FLAG M2 affinity gel (Sigma) was incubated with the lysate for 3 h at 4°C with gentle agitation. Subsequently, the beads were precipitated by centrifugation at 5,000g for 30 sec and washed four times with TBS (150 mM NaCl, 50 mM Tris-HCl, pH 7.4). Lastly, the bound proteins were eluted by boiling in SDS-PAGE loading buffer, separated by SDS-PAGE, and transferred to a PVDF membrane. After blocking with 10% skim milk, the membrane was incubated with either anti-FLAG (Sigma) or anti-YFV-NS5 [4], followed by probing with horseradish-peroxidase-conjugated antibody. A similar experiment was carried out using the plasmid NS5-pEGFP. In this case, Vero cells were co-transfected with U1A-pcNMS2FLAG and, instead of being infected with YFV-17DD, they were transfected with NS5-pEGFP. The interaction was detected using anti-GFP antibody (Sigma) in a western blot assay.

Results

Identification of NS5-RdRp-binding proteins

In this study, we investigated protein-protein interactions between proteins from a HeLa cell cDNA library and the NS5-RdRp domain of YFV using the yeast two-hybrid assay. There were more than 104 putative positive clones. In order to identify the genes encoding the interacting proteins, plasmids were isolated from 35 of these positive clones, amplified in Escherichia coli cells, and sequenced. One of these clones contained a sequence identical to the U1A gene. Then, a two-hybrid assay was repeated using pGBKT7-RdRp and AD-U1A plasmids to confirm the interaction, which was observed by HIS3 and ADE2 reporter gene activation (Fig. 1). This result identifies U1A as a potential novel binding partner for YFV-NS5.

Fig. 1
figure 1

Domain RdRp of YFV NS5 interacts with human U1A protein in a yeast two-hybrid assay. The yeast was co-transformed with the bait (pGBKT7-RdRp or the empty BD-vector) and the prey (pACT2-U1A or the empty AD-vector) to confirm the interaction observed in the two-hybrid screening. A specific interaction activated the reporter genes HIS3 and ADE2 on plates containing SD medium lacking His, Leu, and Trp, and SD medium lacking Ade, His, Leu, and Trp. SD medium without Leu and Trp was used as a transformant growth control

Full size image

Mapping the NS5-RdRp-binding domain of the U1A protein

The binding domain for the interaction between YFV NS5 and U1A was mapped using the two-hybrid assay. Initially, the strain AH109 of S cerevisiae was co-transformed with NS5-RdRp (amino acids 300-905) or with one of the 14 deletion constructs, followed by AD-U1A. Analysis of the results indicated that the region between amino acid residues 368 and 448 of YFV NS5, the Interaction Domain (ID), was responsible for the NS5-U1A interaction (Fig. 2a). In order to determine more precisely the site of protein interaction within the ID region (residues 368-448), plasmids expressing the smaller constructs ID1 (368-412), ID2 (385-431) and ID3 (431-448) were evaluated in the two-hybrid assay (Fig. 2b). In conformity with the HIS3 and ADE2 activation on plates containing SD medium lacking His, Leu, and Trp, and SD medium lacking Ade, His, Leu, and Trp, the minimum domain was localized in the C-terminal region, between residues 431 and 448 (Fig. 2b, ID3+U1A). A predicted structural model of YFV NS5 was constructed based on the crystallographic structure of DENV NS5 (PDB: 2J7U). The model shows an exposed structure with one α-helix between amino acid residues 431 and 448 that corresponds to ID 3 (Fig. 2c). Furthermore, amino acid sequences of other flavivirus NS5 proteins, including those from DENV2 (AF022437.1), DENV3 (AY858045), DENV4 (AY762085), ROCV (AY632542.4), SLEV (AY632544), JEV (AF080251), WNV (DQ211652.1) and YFV (DQ100292) were aligned and analyzed using the DS Gene 2.0 program (Accelrys, USA). Based on the ID 3 region from YFV NS5, sequences corresponding to ID 3 from these other flaviviruses were identified and shown to be highly conserved, with about 73% of identity with YFV-ID3 (residues 431-448) (Fig. 3a). ID 3 sequences from DENV3 (RPDen06/41), DENV4 (Boa Vista) and SLEV (BeH 355964) were cloned into pGBKT7 and assayed with U1A protein in the yeast two-hybrid system. Transformants were observed on plates containing SD medium without His, Leu, and Trp, and SD medium without Ade, His, Leu, and Trp, indicating that the U1A protein interacted with all of the flavivirus NS5 proteins studied (Fig. 3b).

Fig. 2
figure 2

The region of NS5-RdRp that interacts with U1A. (a) Schematic representation of the constructs tested with U1A protein for activation of HIS3 and ADE2 reporter genes in the two-hybrid assay. The fragment size is given in amino acids residues. ID is the region that interacts with U1A. (b) Schematic representation of the three ID segments (1-3) amplified and cloned into pGBKT7. The yeast was co-transformed with these constructs and pACT2-U1A (or the empty AD-vector). The interaction activated HIS3 and ADE2 on plates containing SD medium lacking His, Leu, and Trp, and SD medium lacking Ade, His, Leu, and Trp. SD medium without Leu and Trp was used as a growth control. (c) Predicted structural model for YFV NS5 constructed using the Modeller 6v2 program based on the DENV and WNV NS5 structures (PDB-2J7U and 2OY0, respectively). Model analysis, residue interactions, and comparisons with other related proteins were performed with the PyMOL and Coot programs. The region in green between amino acids residues 431 and 448 corresponds to ID 3

Full size image
Fig. 3
figure 3

The U1A protein interacts with NS5 from other flaviviruses. (a) Alignment of amino acid sequences from flavivirus NS5 of DENV2 (AF022437.1), DENV3 (AY858045), DENV4 (AY762085), ROCV (AY632542.4), SLEV (AY632544), JEV (AF080251), WNV (DQ211652.1) and YFV (DQ100292), using the DS Gene 2.0 program (Accelrys, USA). (b) Activation of HIS3 (SD medium -HIS, -LEU, -TRP) and ADE2 (SD medium -ADE, -HIS, -LEU, -TRP) by interaction of ID from DENV3, DENV4 and SLEV with U1A protein. AD is the empty vector

Full size image

NS5-RdRp interacts with U1A in vitro and in vivo

The YFV NS5-U1A interaction has not been reported previously, so it was further confirmed biochemically using a GST-pulldown assay with purified ID-GST, GST and U1A-pRSET proteins expressed in E. coli. YFV-ID bound to the U1A protein and not to GST alone (Fig. 4). Because U1A is an RNA-binding protein, RNase A was included in the binding reaction. The complex ID-U1A was stable to RNase treatment, indicating a true association between these two proteins that was not due to RNA bridging. These results correlated with those from the yeast two-hybrid assays, and they support that the interaction is direct and specific and does not require additional cellular factors.

Fig. 4
figure 4

NS5-ID binds directly to the U1A protein. The GST-pulldown assay was performed with ID-GST+U1A-HIS (ID-GST+U1A) and GST alone + U1A-HIS (GST+U1A). Western blotting (WB) was carried out using anti-HIS (αHIS) or anti-GST antibody (αGST). The U1A-HIS purified protein expressed in E. coli was used as input. Addition of RNase A did not affect the interaction (ID-GST+U1A+Rnase)

Full size image

We next examined the in vivo association of U1A-FLAG with NS5 after infection of mammalian cells by YFV. Analysis by immunoblotting using anti-YFV/NS5 antibodies revealed a protein band corresponding to YFV NS5 that was precipitated from a virus-infected cell lysate, but no protein was detected with the controls. The reactivity with anti-FLAG antibodies shows specific U1A precipitation (Fig. 5a). These results were confirmed when a similar experiment was carried out using cell transfection with NS5-pEGFP instead of infection with YFV-17DD (Fig. 5b). As a negative control, we transfected cells with GFP and infected them with YFV. The extracts were subjected to co-immunoprecipitation using GFP as antibody. NS5 was not pulled down in this assay, showing that the interaction is specific (Fig. 5c). These results show that YFV-NS5 interacts with U1A protein in virus-infected cells, suggesting a physiological role for NS5-U1A complex formation.

Fig. 5
figure 5

In vivo protein-protein interaction of YFV-NS5 and U1A. (a) Vero E6 cells were infected with YFV and transfected with U1A-FLAG, and nondenatured extracts were incubated with an anti-FLAG affinity gel. The resulting immunoprecipitates were washed and detected by western blotting with anti-YFV-NS5 (αNS5) or anti-FLAG (αFLAG) antibodies. YFV-infected or non-infected, transfected or non-transfected cells with U1A-FLAG cell lysate were also examined by immunoprecipitation. The YFV-infected cell lysate and U1A-FLAG transfected-cell lysate were used as input. IP, immunoprecipitation; WB, western blotting. (b) A similar experiment was performed using an NS5-pEGFP and U1A-FLAG co-transfected-cell lysate. The co-immunoprecipitated proteins were detected by western blotting with anti-GFP (αGFP) or anti-FLAG (αFLAG) antibodies. (c) Co-immuoprecipitation was performed in GFP-transfected and YF-infected cells. The cell extracts were subjected to immunoprecipitation using GFP antibody and western blot using GFP and NS5 antibody

Full size image

Discussion

Flavivirus RNA synthesis occurs at a perinuclear ER-membrane-associated replication complex that involves viral NS proteins. NS5 is one of the most important proteins of this complex, and cellular proteins are believed to be recruited to assist in viral replication. The investigation of NS5-host cell interactions offers insights into viral and cellular function in addition to providing antiviral targets.

This study shows for the first time that YFV NS5 is able to interact with the U1A protein. The results clearly demonstrate that U1A interacts with YFV NS5 in a region of about 17 amino acids residues within the interaction domain. The ID was predicted to be a pendulous structure, making interaction with a protein easier. Notably, the ID sequence was conserved among some flaviviruses of medical importance, especially at the region of interaction. Therefore, it is possible that the flavivirus ID is a site of interaction with a range of viral or cellular proteins other than U1A that are involved in virus replication. In this case, this site could be a candidate for the development of drugs to combat flaviviruses infection. However, other studies are necessary to test this hypothesis.

Replication of RNA viruses involves specific RNA-RNA, RNA-protein, and protein-protein interactions. The flavivirus UTRs contain cis-acting elements and trans-acting factors that may interact with cellular and viral proteins, including NS5, which directs the viral replication machinery to initiate replication at the 3’ end of the viral genome (reviewed in ref. 6). Since the genomic RNAs of flaviviruses lack a poly(A) tail, it is intriguing that an internal poly(A) tail was identified at the 3’ end of some replicon genomes of hepatitis C virus (HCV is a member of the family Flaviviridae) during 3’-terminal sequence analysis of serum samples from a patient with chronic hepatitis [41]. A poly (A) tail has also been observed in some strains of TBEV [29]. One possible mechanism for polyadenylation is non-templated nucleotide addition to the 3’ end by terminal nucleotidyl transferase mediated by NS5-RdRp [39]. Some plus-strand RNA viruses of plants use a polyadenylation signal in the 3’NTR for binding host polyadenylation factors, with subsequent polyadenylation of the RNA genomes influencing viral replication [5, 11]. There is no polyadenylation signal in the 3’NTRs of flaviviruses, and the mechanism of addition of the poly(A) tail is unknown. A recent study showed that the poly(A)-binding protein (PABP) interacts with the DENV 3’UTR in an A-rich region and modulates in vitro translation efficiency [33]. PABP is a protein that simultaneously interacts with terminal poly(A) tails of cellular mRNAs and eukaryotic initiation factors (eIFs) to enhance translation [13]. Since flaviviruses compete with the cellular translation machinery for the production of their viral proteins, they have developed novel mechanisms to ensure efficient translation. In fact, under conditions that inhibit cellular cap-dependent translation, the DENV genome could be translated without a functional cap structure by a novel non-IRES-mediated mechanism that requires both the DENV 5’ and 3’ UTR [7]. Thus, it would not be surprising if the YFV NS5-U1A interaction were involved in polyadenylation of the viral RNA genome in vivo to enhance virus translation.

Furthermore, as for cellular mRNAs, a poly(A) tail could protect the YFV RNA against degradation by cellular ribonuclease. Genome 3’-end repair is a mechanism used by some viruses, such as DENV, HCV and Sindbis virus to protect the integrity of their genome [36, 39, 41]. Deletion of the 3’ nucleotides could prevent flavivirus RNA synthesis, since this region is involved in initiation of replication. Analysis of the restored HCV-RNA 3’ end revealed that, in the repair process, several genomes were polyadenylated at the 3’ end [41]. Thus, 3’-end polyadenylation of YFV RNA by the NS5-U1A interaction could be a consequence of a repair mechanism to contribute to RNA stability and consequently to virus replication.

NS5 is found in the ER-membrane-associated replication complex, free in the cytoplasm or in the nuclei of flavivirus-infected cells [2, 17, 34, 40]. It is known that the DENV2 NS5 protein can shuttle between the nucleus and the cytoplasm by interacting with nuclear transporters (Imp-α/β, Imp-β and CRM1), and this is mediated by nuclear localization signals (NLS) or nuclear export signals (NES). The role of NS5 in the nucleus is presently unclear, but it has the potential to alter host processes or play a role in viral RNA synthesis [15, 34, 37].

There are different pathways by which proteins can be transported to the nucleus in addition to the well-characterized importin- α/β pathway [14]. In fact, NS5-DENV-2 has been shown to be transported to the nucleus by direct interaction with importin-β, independently of importin-α, and the region of NS5 responsible was not the previously characterized NLS of NS5 recognized by importin- α/β heterodimer [15]. NS5 requires an NLS to enter the nucleus unless it uses an unknown host- or virus-derived factor [9].

U1A contains two RNA-recognition domains (RRM1 and RRM2). RRM1 interacts specifically with stem-loop II of U1 RNA, and RRM2 does not appear to associate with any RNAs [23, 27, 28]. Therefore, it appears to be a site that is free for binding with other RNAs or proteins. U1A is predominantly a nuclear protein, but it can shuttle between the nucleus and the cytoplasm independently of interactions with U1 snRNA. The concentration of the protein in the nucleus and cytoplasm can be perturbed by introducing RNA sequences that can specifically bind U1A in either the nuclear or cytoplasmic compartment [16]. Proteins that shuttle between the nucleus and the cytoplasm are good candidates for coupling nuclear and cytoplasmic processes to each other, and they are also cofactors that are important for the nucleocytoplasmic trafficking of other macromolecules [31]. Thus, the presence of YFV NS5 in the cytoplasm could alter U1A distribution, which would promote NS5-U1A interaction, which could be an alternative form of NS5 to shuttle between the nucleus and the cytoplasm.

On the other hand, similar to the YFV NS5, the U1A protein has a NLS in the central domain that is responsible for its nuclear import [16], and it is dependent on the importin- α/β and Ran pathways [12]. It is possible that the YFV NS5-U1A interaction can modulate NS5 nuclear transport.

In summary, we first report in this work the interaction between a flavivirus NS5 protein and U1A, which may have an important role in virus replication, although its functional significance remains to be determined.