Flavivirus NS1 possesses NTPase and RNA helix unwinding activities

Abstract

Flaviviruses are a group of arthropod-borne positive-sense RNA viruses, including Zika virus (ZIKV), dengue virus 2 (DENV2) and Japanese encephalitis virus (JEV). For RNA viruses, RNA helicases play critical roles in viral replication by unwinding double-stranded RNA intermediates generated in viral life cycles by remodeling viral RNA structures. Here, we report that NS1 proteins from ZIKV, DENV2 and JEV, possesse NTPase and RNA helix unwinding activities. ZIKV NS1 unwinds dsRNA in an NTP-dependent manner with 5′ → 3′ directionality, and its unwinding activity is broad-spectrum, targeting dsRNA, dsDNA and RNA/DNA hybrids. Moreover, both DENV2 and JEV NS1 also display NTPase and helicase activities, but they differ in substrate preference and directionality: JEV NS1 unwinds in the 5′ → 3′ direction like ZIKV NS1, whereas DENV2 NS1 unwinds only shorter duplexes and shows 3′ → 5′ directionality. Mutagenesis identified the basic residue at position 211 (arginine in ZIKV, lysine in DENV2/JEV) as critical for the unwinding activity of ZIKV and JEV NS1, but dispensable for DENV2 NS1. These findings indicate that flavivirus NS1 represents a previously unrecognized non-canonical RNA helicase, suggesting that NS1 may have the capacity to participate in RNA remodeling processes.

Highlights

• •Flavivirus NS1 proteins (ZIKV, DENV2, JEV) possess intrinsic NTPase and RNA helicase activities.
• •ZIKV NS1 unwinds dsRNA in a 5′→3′ direction and acts on dsRNA, dsDNA and RNA/DNA hybrids.
• •DENV2 and JEV NS1 show helicase activity but differ in substrate preference and directionality.
• •Residue 211 is critical for unwinding activity of ZIKV and JEV NS1, but dispensable for DENV2 NS1.

Introduction

RNA helicases are a class of proteins that harness the energy from ATP hydrolysis to unwind RNA helix (Singleton et al., 2007) and participate in various RNA metabolic processes. RNA virus-encoded helicases can be broadly divided into two classes. The first class includes canonical helicases carrying conserved motifs, e.g., coronavirus nsp13 and enterovirus 2C (Chen et al., 2022; Shu et al., 2020; Xia et al., 2015). The second class comprises non-canonical helicases devoid of typical motifs but with demonstrated RNA helix unwinding activity, such as Ebola virus VP35 and respiratory syncytial virus P (Shu et al., 2019, 2025). During the viral life cycle, RNA performs diverse functions. Proper folding of RNA into correct structures is critical for executing accurate biological functions. However, due to the flexibility of RNA, it is prone to misfolding. RNA helicases represent the largest class of proteins that influence RNA structure and regulate RNA function (Anantharaman et al., 2002; Linder and Jankowsky, 2011). For RNA viruses, multiple cis-acting elements within the non-coding regions and open reading frames of the viral genomic RNA need to adopt correct secondary or tertiary structures to execute proper functions. At this point, RNA helicases exert their unwinding activity to open misfolded RNA duplexes, allowing them to refold into correct structures (Musier-Forsyth, 2010). For positive-sense single-stranded RNA viruses, during replication, the genome serves as a template to produce the antisense negative-strand RNA, which is then used as a template to synthesize new viral genomes (Griffiths et al., 2017). During this process, double-stranded RNA (dsRNA) replication intermediates are inevitably generated (Lescar et al., 2018). Therefore, the timely unwinding of replication intermediate dsRNA by RNA helicases is essential for successful viral genome synthesis and replication efficiency (Shi et al., 2021). Processes such as viral RNA transcription, replication, and packaging all require virus-encoded RNA helicases (Kadare and Haenni, 1997). The critical roles of RNA helicases in the viral life cycle have made them popular targets for antiviral drugs and important subjects of virology research (Fang et al., 2021; Kwong et al., 2005).

Flaviviruses are a group of arthropod-borne positive-sense RNA viruses belonging to the family Flaviviridae, including ZIKV, DENV, JEV, Yellow fever virus (YFV), Tick-borne encephalitis virus (TBEV), and West Nile virus (WNV) (Cody et al., 2025; Pierson and Diamond, 2020). Flaviviruses are important human pathogens capable of causing a range of severe diseases, including hemorrhagic fever, vascular shock syndrome, encephalitis, acute flaccid paralysis, congenital abnormalities, and fetal death (Guzman et al., 2010). Over the past decades, flaviviruses have spread widely and caused multiple outbreaks around the world, including epidemics of DENV and WNV, as well as the 2015–2016 ZIKV outbreak in the Americas. Flaviviruses infect up to 400 million people annually, posing a serious threat to global human health (Pierson and Diamond, 2020). Currently, although preventive vaccines are available for a few flaviviruses such as YFV and TBEV, no antiviral drugs against these viruses have entered the market (Miazga et al., 2023; Srivastava et al., 2026).

The flavivirus genome is a positive-sense (+) single-stranded RNA of approximately 10–11 ​kb, consisting of a single long open reading frame (ORF) encoding three structural proteins (C, prM, and E) and seven nonstructural proteins (NS1, NS2A, NS2B, NS3, NS4A, NS4B, and NS5) (Guzman et al., 2010). These proteins play major roles in virion assembly, cell receptor binding and entry, viral polyprotein processing, and viral replication (Barnard et al., 2021). Among them, NS1 is an important multifunctional protein composed of 352 amino acids, functioning as dimers or hexamers, and has been reported to participate in multiple critical processes of flavivirus infection and replication (Perera et al., 2023). Structurally, NS1 is organized around a central β-sheet domain, and residues 30 to 180 form a wing-like domain protruding from the β-sheet domain, termed the Wing domain. Two residues within the connecting region of the Wing domain are critical for the binding of NS1 to the NS4A-2K-4B polyprotein (Akey et al., 2014). The Wing domain contains an α/β subdomain consisting of a four-stranded sheet, two α-helices, and a disordered distal tip. Interestingly, structural comparison revealed that the α/β subdomain of the NS1 Wing domain resembles the helicase region of RIG-I (Akey et al., 2014). RIG-I belongs to the RLR family and contains a DExD/H-box helicase domain essential for RNA binding. When pathogenic dsRNA appears in the cytoplasm, the RIG-I helicase domain binds to it, utilizes ATP to complete conformational changes, forms oligomeric complexes, and becomes activated, triggering downstream signaling pathways and inducing antiviral and inflammatory gene expression (Yoneyama et al., 2004). However, despite the structural resemblance, the helicase-like features of NS1 have not been experimentally characterized, and it remains unknown whether NS1 possesses RNA unwinding activity, which limits our understanding of its potential roles in the viral life cycle.

In this study, we systematically characterized the helicase functions of NS1s from ZIKV, DENV2, and JEV. Our results show that although all three NS1s lack the conserved motifs of canonical helicases, they universally possess NTPase and RNA helix unwinding activities. However, their unwinding characteristics display clear interspecies divergence: ZIKV and JEV NS1 unwind RNA in the 5′ → 3′ direction, can process longer duplex substrates, and residue 211 (R or K) is critical for their unwinding activity. In contrast, DENV2 NS1 unwinds in the 3′ → 5′ direction, acts only on shorter duplexes, and mutation of residue 211 does not affect its activity. Nonetheless, the biochemical features of the unwinding activities of the three NS1s are highly similar to those of canonical helicases, including dependence on divalent metal cations and NTP hydrolysis, as well as clear directionality. These findings establish flavivirus NS1 as a novel helicase with conserved yet distinct properties, suggesting its potential role in viral replication.

Results

ZIKV NS1 possesses NTP-binding and NTPase activities

To explorer the potential enzymatic activities of flaviviral NS1, we firstly fused the N-termini of ZIKV NS1 to maltose-binding protein (MBP–NS1) and purified this recombinant protein (Supplementary Fig. S1B). Many viral RNA-remodeling proteins also exhibit NTPase activity (Li et al., 2018; Shu et al., 2019, 2020), we therefore examined whether ZIKV NS1 possesses the ability to hydrolyze NTPs (ATP, GTP, CTP, UTP) and dNTPs (dATP, dGTP, dCTP, dTTP) using a sensitive colorimetric assay that measures the total free phosphate released after ATP hydrolysis. Our data show that ZIKV NS1 hydrolyzes all four NTPs and four dNTPs, with a preference for ATP, GTP, and UTP over CTP, and for dATP, dGTP, and dTTP over dCTP (Fig. 1A). Moreover, increasing concentrations of ZIKV NS1 led to efficient ATP hydrolysis in a dose-dependent manner (Fig. 1B). Given that ZIKV NS1 exhibits NTPase activity, we next asked whether it also possesses NTP-binding activity. Our data show that 5′-ATP-agarose readily pulled down ZIKV NS1 but not MBP alone (Fig. 1C), indicating that NS1 indeed binds ATP. Notably, the MBP protein as a control was expressed and purified in exactly the same manner as MBP–NS1, yet its NTP hydrolysis and binding activities were negligible.

Fig. 1.

ZIKV NS1 has NTP-binding and NTPase activities. A 10 ​pmol MBP-NS1 was reacted with the indicated NTPs (2.5 ​mM) or dNTPs (2.5 ​mM). The NTPase activity was measured as nanomoles of released inorganic phosphate (pi) by using a sensitive colorimetric assay. B 2.5 ​mM ATP was incubated with MBP or MBP-NS1 at the increasing concentrations, respectively. C MBP or MBP-NS1 was pulled down by using maltose- or 5′-ATP agarose, followed by SDS-PAGE and immunoblotting using anti-MBP antibody. D–F 10 ​pmol MBP-NS1 was reacted with 2.5 ​mM ATP at 2 ​mM indicated divalent metal ions (D), at the indicated concentrations of MgCl₂ (E) or the indicated pH (F). Error bars represent standard deviation (SD) values from three separate experiments.

After confirming that ZIKV NS1 has NTPase activity, we used ATP as a substrate to determine the optimal biochemical reaction conditions, including different salt content, Mg²⁺ concentration, and pH. Our data show that 2 ​mM Mg²⁺, Mn²⁺, Ca²⁺, or Zn²⁺ could support the ATPase activity of NS1, with an efficiency order of Mg²⁺ ​> ​Zn²⁺ ​> ​Mn²⁺ ​> ​Ca²⁺ (Fig. 1D). ZIKV NS1 exhibited optimal ATPase activity in the presence of 2 ​mM Mg²⁺, while higher Mg²⁺ concentrations had some inhibitory effects. Importantly, the optimal ATPase activity of ZIKV NS1 required the presence of Mg²⁺, and NS1 had no ATPase activity in the absence of divalent metal ions (Fig. 1E), which is similar to other NTPases (Shu et al., 2020). Furthermore, we observed that ZIKV NS1 achieved optimal ATPase activity at pH 8.5 (Fig. 1F). Together, our data indicate that ZIKV NS1 possesses both NTP-binding and NTPase activities.

ZIKV NS1 possesses dsRNA-binding activity

The dsRNA-binding activity is essential for helicase function. Therefore, before characterizing the unwinding activity of ZIKV NS1, we used EMSA to confirm whether NS1 binds dsRNA. Considering that viral helicases function during viral genome replication, we hypothesized that ZIKV NS1 should be able to bind the ZIKV viral genome. To test this, we used in vitro transcribed ZIKV 200-nt 3′ UTR as a binding substrate. As shown in Fig. 2A, ZIKV NS1 bound the 3′ UTR, and the amount of bound dsRNA increased with increasing concentrations of ZIKV NS1, whereas the negative control MBP did not bind dsRNA. Furthermore, using 200-nt dsEGFP as a substrate, ZIKV NS1 still exhibited dsRNA-binding activity (Fig. 2B).

Fig. 2.

ZIKV NS1 has dsRNA binding activity. A Gel mobility shift assays were performed to evaluate the dsRNA binding activity of MBP-NS1. Protein was incubated with 0.1 ​pmol DIG-labeled 200 ​nt ZIKV 3′UTR dsRNA substrate, and the complex was analyzed by gel electrophoresis, transferred to membranes and then incubated with anti-DIG antibody conjugated with alkaline phosphatase. Lane 1, no protein supplemented; Lane 2, 20 ​pmol MBP supplemented; Lane 3–5, the indicated increasing concentrations (2.5–10 ​pmol) of MBP-NS1. B Gel mobility shift assays were performed by incubating the indicated increasing concentrations (2.5–10 ​pmol) of MBP-NS1 with 0.1 ​pmol 200 ​nt dsEGFP.

ZIKV NS1 possesses NTP-dependent RNA helix unwinding activity

Having established that ZIKV NS1 has NTPase and dsRNA-binding activities, we next asked whether ZIKV NS1 possesses unwinding activity using a standard helicase assay with a helix substrate containing a 42-nt complementary region flanked by 6-nt overhangs at both ends (Fig. 3A). As shown in Fig. 3B, ZIKV NS1 successfully unwound the dsRNA, releasing the labeled single-stranded RNA; the positive control protein EV71 2C exhibited unwinding activity as expected, whereas the negative control MBP did not.

Fig. 3.

ZIKV NS1 has NTP-dependent RNA helix unwinding activity. A Schematic illustration of the standard RNA helix substrate (R/R∗). Asterisks indicate the HEX-labeled strands. B The standard RNA helix substrate (0.1 ​pmol) was reacted with 20 ​pmol MBP, MBP-NS1, MBP-2C (lane 3–5). The unwinding activity was assessed via gel electrophoresis and scanning on a Typhoon 9500 imager. Non-boiled reaction mixture (lane 1) and reaction mixture with MBP (lane 3) were used as negative controls, and boiled reaction mixture (lane 2) and reaction mixture with MBP-2C (lane 4) were used as positive controls. C The RNA helix unwinding assay was performed in the presence of each indicated NTPs (5 ​mM). D MBP–NS1 (20 ​pmol) was reacted with the standard RNA helix substrate (0.1 ​pmol) in reaction mixture in the absence (lane 3) or presence (lane 4) of 5 ​mM ATP, or in the presence of 5 ​mM AMP-PNP (lane 5), or 5 ​mM ATP with 5 ​mM AMP-PNP (lane 6), followed by unwinding assay. E The RNA helix unwinding assay was performed in the presence of increasing concentrations of ATP.

Canonical helicases are NTP-dependent (Singleton et al., 2007). The energy provided by NTP hydrolysis is essential for helicase activity. To determine whether flavivirus NS1 is NTP-dependent, we first examined the utilization of four different NTPs by ZIKV NS1. As shown in Fig. 3C, in the absence of NTPs, NS1 failed to unwind, whereas in the presence of different NTPs, NS1 exhibited unwinding activity, with a preference order of ATP ​= ​GTP ​> ​UTP ​> ​CTP. Furthermore, we used the ATP analogue AMP-PNP as a control to re-examine the ATP dependence of NS1. AMP-PNP is structurally similar to ATP but cannot hydrolyze the terminal γ-phosphate and thus does not release energy, making it a common control for studying ATP importance. As shown in Fig. 3D, NS1 only exhibited unwinding activity in the presence of ATP; on unwinding was detected in the absence of ATP or in the presence of AMP-PNP alone. When both ATP and AMP-PNP were present, AMP-PNP inhibited the unwinding activity of NS1. Moreover, increasing ATP concentration enhanced the RNA unwinding activity of NS1 (Fig. 3E). Collectively, these results indicate that ZIKV NS1 functions as an NTP-dependent RNA helicase, strictly requiring NTP hydrolysis particularly ATP and GTP, for dsRNA unwinding.

Substrate specificity of ZIKV NS1 unwinding activity

RNA virus-encoded helicases can unwind not only dsRNA but also dsDNA and DNA/RNA hybrid duplexes (Li et al., 2018). To investigate whether ZIKV NS1 exhibits substrate specificity for different types of helical duplexes, we prepared four different helix substrates: dsRNA, dsDNA, and RNA/DNA hybrids with either an RNA overhang or a DNA overhang. All four helix possessed single-stranded overhangs at both ends (Fig. 4A–D, upper panels). ZIKV NS1 was able to unwind dsRNA (Fig. 4A), dsDNA (Fig. 4B), and both types of hybrid duplexes (Fig. 4C, D). These results indicate that ZIKV NS1 has broad-spectrum unwinding capability, not only unwinding dsRNA but also dsDNA and RNA/DNA hybrid duplexes.

Fig. 4.

ZIKV NS1 has nucleic acid helix unwinding activity. A–D 20 ​pmol MBP-NS1 was reacted with 0.1 ​pmol of standard RNA helix (R/R∗) (A), DNA helix (D∗/D) (B), or DNA/RNA hybrid with longer RNA strand (D/R∗) (C) or longer DNA strand (R∗/D) (D). The schematic illustrations of the helix substrates are illustrated in the upper panels. Asterisks indicate the HEX-labeled strand.

The optimal biochemical conditions for ZIKV NS1 RNA helix unwinding activity

Flaviviruses constitute a large group of arthropod-borne viruses transmitted mainly by mosquito bites (Cannac and Nisole, 2024). Considering that mosquitoes are poikilothermic, with their body temperature varying according to the ambient environment (typically around 37 ​°C for laboratory-reared mosquitoes), whereas humans maintain a constant body temperature of approximately 37 ​°C, we tested whether the unwinding activity of NS1 differs at different temperatures. The unwinding activity of ZIKV NS1 was significantly stronger at 37 ​°C than at 27 ​°C (Fig. 5A). In addition, it is well established that divalent metal cations, especially Mg²⁺, are indispensable for the unwinding reaction of canonical helicases. DEAD-box helicases often require binding to divalent metal cations to induce conformational changes of the two helicase domains, thereby binding ssRNA and initiating unwinding. Our previous results had already shown that the NTPase activity of ZIKV NS1 is dependent on divalent metal cations, so we further examined the effect of divalent metal cations on the unwinding activity of NS1. We supplemented the reaction with 2 ​mM Mg²⁺, Mn²⁺, Ca²⁺, or Zn²⁺ to assess their effects. As shown in Fig. 5B, NS1 could utilize Mg²⁺, Mn²⁺, and Zn²⁺, with the strongest unwinding activity observed in the presence of Mg²⁺. In the absence of divalent metal cations or in the presence of Ca²⁺, NS1 failed to unwind. To further confirm the Mg²⁺ dependence of NS1 unwinding activity, we set up reactions with increasing Mg²⁺ concentrations (0–3.5 ​mM). As shown in Fig. 5C, NS1 unwinding activity was enhanced with increasing Mg²⁺ concentration. In addition to divalent metal cations, pH also plays an important role in enzymatic activity. Therefore, we performed unwinding assays at various pH values. As shown in Fig. 5D, the unwinding activity of NS1 was strongest at approximately pH 8. Together, we demonstrate that ZIKV NS1 unwinding activity is significantly stronger at 37 ​°C than at 27 ​°C, depends on divalent metal cations with a preference for Mg²⁺, and is optimal at approximately pH 8.

Fig. 5.

Optimal biochemical reaction conditions for the RNA helix unwinding activity of ZIKV NS1. A–D MBP-NS1 (20 ​pmol) was reacted with 0.1 ​pmol standard RNA helix substrate in the presence of each indicated temperature (A), indicated divalent metal ion (2 ​mM for each) (B), increasing concentrations of Mg²⁺(C), or indicated pH values (D).

ZIKV NS1 unwinding activity exhibits directionality

Canonical helicases possess directionality (5′ → 3′ or 3′ → 5′) (Shu et al., 2020). We therefore synthesized three helix substrates with different overhangs: a 42-nt complementary dsRNA with a 3′-overhang of 6 ​nt (Fig. 6A), a 42-nt complementary dsRNA with a 5′-overhang of 6 ​nt (Fig. 6B), and a blunt-ended 42-nt complementary dsRNA (Fig. 6C). Unwinding assays revealed that ZIKV NS1 was unable to unwind the 3′-overhang or blunt-ended dsRNAs (Fig. 6D, E), but efficiently unwound the 5′-overhang dsRNA (Fig. 6D). Moreover, increasing ATP concentrations did not enable NS1 to unwind the 3′-overhang substrate (Fig. 6F), whereas the unwinding activity on the 5′-overhang substrate was enhanced with increasing ATP concentration (Fig. 6G). These results demonstrate that ZIKV NS1, like canonical helicases, exhibits directionality and unwinds helix in the 5′ → 3′ direction.

Fig. 6.

ZIKV NS1 unwinds RNA helix in the 5 ′ to 3 ′ directionality. A–C Schematic illustrations of the RNA helix substrates with 3 ′ tailed (A), 5 ′tailed (B), and blunt ends (C). Asterisks indicate the HEX-labeled strand. D, E 20 ​pmol MBP-NS1 was reacted with 0.1 ​pmol 3′ tailed (lane 3) or 0.1 ​pmol 5′ tailed (lane 6) RNA helix substrate (D), or 0.1 ​pmol RNA helix substrate with blunt ends (E). F, G 20 ​pmol MBP-NS1 was reacted with 0.1 ​pmol 3 ′ tailed (F) or 0.1 ​pmol 5 ′ tailed (G) RNA helix substrate in the presence of increasing concentrations of ATP.

DENV2 and JEV NS1 possess NTPase and RNA helix unwinding activities

The NS1 proteins encoded by flaviviruses are highly conserved in sequence (Supplementary Fig. S1A). We hypothesized that NS1 from other flaviviruses might also possess NTPase and unwinding activities. Therefore, we expressed and purified MBP-tagged DENV2 NS1 and JEV NS1 under identical conditions (Supplementary Fig. S1B). We found that both DENV2 and JEV NS1s possess NTPase (Fig. 7A, B). Moreover, they also contain RNA helix unwinding activities, but exhibit specificity regarding the length of the dsRNA substrate. JEV NS1, unwound a 42-bp helix substrate (Fig. 7C), whereas DENV2 NS1 could not (Fig. 7D); instead, DENV2 NS1 only unwound shorter duplexes (30-nt complementary region, Fig. 7E).

Fig. 7.

DENV2 and JEV NS1 has NTPase and RNA helix unwinding activities. A,B 2.5 ​mM ATP was incubated with DENV2 NS1 (A) or JEV NS1 (B) at the increasing concentrations, respectively. C The RNA 42-nt complementary helix substrate (0.1 ​pmol) was reacted with 20 ​pmol MBP (lane 3), JEV NS1 (lane 4), EV71 2C (lane 5). D 20 pmol DENV2 NS1 was reacted with 0.1 pmol the 42-nt complementary RNA helix substrate. E The 30-nt complementary RNA helix substrate (0.1 pmol) was reacted with 20 pmol MBP (lane 3), DENV2 NS1 (lane 4), EV71 2C (lane 5). F, G 20 ​pmol DENV2 NS1 (F) or JEV NS1 (G) was reacted with 0.1 ​pmol 3 ′ tailed (lane 3) or 0.1 ​pmol 5′ tailed (lane 6) RNA helix substrate. Schematic illustrations of the helix substrates are illustrated in the upper panels. Asterisks indicate the HEX-labeled strand.

We next determined the unwinding directionality of DENV2 and JEV NS1s. First, using 30-nt complementary dsRNA substrates with different overhangs (Fig. 7F, top), we examined the directionality of DENV2 NS1. We found that DENV2 NS1 unwound the 3′-overhang helix but not the 5′-overhang helix (Fig. 7F, bottom). Next, using 42-nt complementary dsRNA substrates with different overhangs (Fig. 7G, top), we examined the directionality of JEV NS1. We found that JEV NS1 unwound the 5′-overhang helix but not the 3′-overhang helix (Fig. 7G, bottom). Therefore, both DENV2 and JEV NS1s exhibit unwinding directionality, but DENV2 NS1 unwinds in the 3′ → 5′ direction (opposite to ZIKV NS1), whereas JEV NS1 unwinds in the same 5′ → 3′ direction as ZIKV NS1.

Together, our findings demonstrate that both DENV2 and JEV NS1s possess NTPase and unwinding activities but exhibit distinct substrate length specificities and opposing directionalities: JEV NS1 unwinds a 42-bp duplex in the 5′ → 3′ direction like ZIKV NS1, whereas DENV2 NS1 only unwinds shorter duplexes and functions in the 3′ → 5′ direction.

Residue 211 is critical for the unwinding activity of ZIKV and JEV NS1

We sought to determine the key amino acids responsible for NS1's unwinding activity. ZIKV NS1 protein surface harbors multiple positively charged residues, including a cluster of arginine and lysine residues (R211, K213, R324) that are exposed to the solvent, representing potential RNA-interacting sites (Baez et al., 2017). Consistent with it, we found that mutation of the RKR amino acid, including arginine (R) at position 211, lysine (K) at position 213, and arginine (R) at position 324, abolished the unwinding activity of ZIKV NS1 (Fig. 8A, B). Subsequent single-residue mutagenesis identified residue R211 as critical for ZIKV NS1 unwinding activity (Fig. 8C). And R211A also dramatically reduces the dsRNA-binding activity of ZIKV NS1 (Fig. 8D). Moreover, sequence alignment revealed that the corresponding residue to ZIKV NS1 R211 is a K in JEV and DENV2 NS1 (Supplementary Fig. S1A); as both R and K are positively charged residues, this physicochemical property is conserved among the three viruses. Thus, we mutated the corresponding lysine in DENV2 and JEV NS1 to alanine (K211A). The K211A mutation did not affect the unwinding activity of DENV2 NS1 (Fig. 8E), but abolished the unwinding activity of JEV NS1 (Fig. 8F). Thus, the basic residue at position 211 is essential for the unwinding activity of ZIKV and JEV NS1, but is dispensable for DENV2 NS1.

Fig. 8.

Residue 211 is critical for the unwinding activity of ZIKV and JEV NS1. A Schematic illustration for deletion mutations and points mutation. The RKR sites are replaced with alanine. B The RNA helix unwinding assays were performed either with 20 pmol RKR-to-AAA mutant protein at the indicated ATP concentrations (0–10 ​mM), or with 2.5 ​mM ATP at the indicated concentrations of the mutant protein (0–10 ​pmol). C The RNA helix unwinding assays were performed by incubating 0.1 ​pmol standard helix substrate with 20 ​pmol of the indicated ZIKV NS1 or mutant proteins. D Gel mobility shift assays were performed by incubating 20 ​pmol of the indicated ZIKV NS1 or mutant proteins with 0.1 ​pmol 200 ​nt dsEGFP. E The RNA helix unwinding assays were performed by incubating 0.1 ​pmol standard helix substrate with 20 ​pmol of the indicated DENV2 NS1 or mutant proteins. F The RNA helix unwinding assays were performed by incubating 0.1 ​pmol standard helix substrate with 20 ​pmol of the indicated JEV NS1 or mutant proteins.

Residue R211 is located on the β13 region (β-ladder) of NS1 (Xu et al., 2016). To further elucidate the role of R211 in the RNA-binding and helicase activities of NS1, we employed a domain-swapping strategy by replacing the RNA-binding motif V of the canonical helicase ZIKV NS3 with the β13 region of NS1 (Fig. 9A and B) (Jain et al., 2016). We then performed unwinding assays to determine whether the β13 region could compensate for the reduced unwinding activity caused by deletion of motif V. The RNA unwinding activity of these proteins was examined in vitro using a standard duplex substrate. The results showed that deletion of motif V modestly reduced the unwinding activity of NS3, while replacement of motif V with the NS1 β13 restored the unwinding activity of NS3 (Fig. 9C). Taken together, the β13 region of NS1 can functionally substitute for motif V of NS3, indicating the importance of R211 for NS1's unwinding activity.

Fig. 9.

Effect of domain swapping between ZIKV NS1 and NS3 on RNA unwinding activity. A Schematic of full-length ZIKV NS1 and NS3, along with ZIKV NS3 domain deletion or domain swapping mutants. B The Flag-fusion NS1, NS3, and mutant proteins were expressed using the 293F expression system. The purified recombinant proteins were subjected to 10% SDS-PAGE followed by Coomassie brilliant blue R250 staining. Lane 1, protein marker; lane 2, full-length Flag-ZIKV NS1; lane 3, full-length Flag-ZIKV NS3; lane 4, Flag-ZIKV NS3 with motif V deleted; lane 5, Flag-ZIKV NS3 with motif V replaced by the NS1 β13 region. C The RNA helix unwinding assays were performed by incubating 0.1 pmol standard helix substrate with 20 pmol of the indicated ZIKV NS1, ZIKV NS3 or mutant proteins. The unwinding efficiency (UE) was quantified as the percentage of RNA released from the total RNA helix substrate.

Discussion

This study demonstrates that flavivirus nonstructural protein NS1 possesses canonical RNA helicase activity, including broad-spectrum NTPase activity and NTP-dependent RNA helix unwinding activity. Through systematic functional characterization of NS1 from ZIKV, DENV2 and JEV, we found that although these proteins lack the conserved motifs of canonical helicases, they can efficiently unwind various helix substrates (dsRNA, dsDNA, and RNA/DNA hybrids) in a direction-dependent manner. These findings expand the known biochemical repertoire of NS1 beyond its established roles in immune evasion and membrane remodeling, and suggest the possibility that NS1 may participate in RNA structure remodeling during viral replication.

The flavivirus-encoded NS3 protein is a canonical DExD/H-box helicase (Du Pont et al., 2022). The finding that NS1 also possesses helicase activity, independent of canonical helicase motifs, suggests that flaviviruses may have evolved two functionally complementary RNA helicase systems. NS3, as a core component of the replication complex, has its helicase activity tightly coupled with RNA synthesis. In contrast, NS1 may play an auxiliary role in local RNA structure remodeling, such as unwinding misfolded cis-acting elements or facilitating the transient rearrangement of replication intermediates. The division of labor between these two proteins in terms of replication timing, substrate preference, and sites of action is an important direction for future research.

The three viral NS1s exhibit clear divergence in unwinding directionality: ZIKV and JEV NS1 show 5′ → 3′ directionality, whereas DENV2 NS1 shows 3′ → 5′ directionality. Moreover, ZIKV and JEV NS1 can unwind longer duplex substrates (42-nt complementary region), while DENV2 NS1 can only unwind shorter duplexes (30-nt complementary region). This functional divergence is highly consistent with phylogenetic relationships – ZIKV and JEV are more closely related to each other and share similar unwinding properties, whereas DENV2 represents a functional side branch (Gaunt et al., 2022). From an evolutionary perspective, the divergence in directionality may stem from differences in the genomic RNA secondary structures of different flaviviruses and their replication complex assembly modes. The 5′; → 3′ direction is common among many virus-encoded helicases (e.g., Hepatitis C virus NS3) (Kwong et al., 2000), facilitating forward progression along the genome. In contrast, the 3′ → 5′ directionality adopted by DENV2 NS1 may relate to its specific replication intermediate configuration or its mode of cooperation with NS3. Furthermore, the fact that DENV2 NS1 can only process short duplexes suggests that in vivo it may participate only in local, short-range RNA structural remodeling rather than the unwinding of full-length replication intermediates. These functional differences provide new perspectives for understanding the diversity of flavivirus replication strategies.

We found that the unwinding activity of ZIKV NS1 is significantly stronger at 37 ​°C (human host temperature) than at 27 ​°C (mosquito vector temperature), suggesting that this protein may play a more important unwinding role in the human host. This temperature dependence may correlate with flavivirus pathogenicity: at higher temperatures, the enhanced unwinding activity of NS1 could facilitate efficient viral replication in humans. This also provides an important reference for in vitro functional assays.

Because NS1 is highly conserved among flaviviruses and has no host homologue (Supplementary Fig. S1A), its unique non-canonical helicase active sites (e.g., R211 in ZIKV, K211 in JEV) represent promising targets for antiviral drug development. Targeting the helicase activity of NS1 may circumvent the toxicity and specificity issues associated with traditional drugs targeting the NS3 helicase, which often arise from host homologues. However, it is important to emphasize that the present study provides only in vitro biochemical evidence, and extensive validation using cell-based and animal models of flavivirus infection would be required before any realistic assessment of drugability can be made. The observed differences in NS1 helicase mechanisms among different flaviviruses represent an interesting biochemical observation, but whether these differences can be translated into broad-spectrum or virus-specific antiviral strategies remains entirely speculative at this stage.

Conclusions

In summary, this study establishes flavivirus NS1 as a previously unrecognized non-canonical RNA helicase possessing NTPase activity and direction-dependent unwinding activity, and reveals significant divergence in directionality, substrate length preference, and key amino acid function among different virus species. These findings expand the known functional repertoire of NS1 and suggest that NS1 may have the capacity to participate in RNA remodeling processes.

Materials and methods

Plasmid and recombinant baculovirus construction

The construction of pFastBac HTB-MBP, pFastBac HTB-MBP-ZIKV NS1, pFastBac HTB-MBP-DENV2 NS1, pFastBac HTB-MBP-JEV NS1, pFastBac HTB-MBP-EV71 2C and have been described previously (Wang et al., 2012a, 2012b). The cDNA fragment of ZIKV NS1 and ZIKV NS3 (GenBank accession no. KU820898.1), DENV2 NS1 (GenBank accession no. OM257419.1), JEV NS1 (GenBank accession no. AF221499.1), and EV71 2C (GenBank accession no. KC954662) were cloned into the vector pFastBac HTB-MBP, in which the maltose binding protein (MBP) was N terminally fused. The mutations were conducted as previously described (Qi et al., 2011). The resulting plasmid was subjected to Bac-to-Bac baculovirus system to express the recombinant protein. Additionally, ZIKV NS1 and ZIKV NS3 were subcloned into the pV7-Flag vector digested with XhoI and NotI. The primers used in this study are shown in Supplementary Table S1.

Expression and purification of recombinant fusion protein

The expression and purification of recombinant MBP-fusion ZIKV NS1, DENV2 NS1, JEV NS1, EV71 2C and negative control MBP from Bac-to-Bac system were conducted as previously described (Cheng et al., 2013; Wang et al., 2012a). Briefly, sf9 cells were infected with the recombinant baculoviruses and harvested at 72h post-infection. Cell pellets were re-suspended, lysed by sonication and subjected to centrifugation for 15 ​min at 12000 ​g to remove debris. The protein in the supernatant was purified using amylose affinity chromatography (New England BioLabs, Ipswich, MA), according to the manufacturer's protocol.

The expression and purification of recombinant Flag-fusion ZIKV NS1,ZIKV NS3 and NS3 mutant from 293F expression system. Briefly, 293F cells were transfected pV7-Flag-ZIKV NS1, pV7-Flag-ZIKV NS3 or mutants and harvested at 72 ​h post-transfection. Cell pellets were re-suspended, lysed by sonication and subjected to centrifugation for 20 ​min at 12000 ​g to remove debris. The protein in the supernatant was purified using Anti-DYKDDDDK G1 Affinity Resin (GenScript), according to the manufacturer's protocol.

Then the protein was concentrated using Ultra-15 filters (Millipore, Schwalbach, Germany), and the store buffer was exchanged to 50 ​mM2-[4-(2-hydroxyethyl)-1-piperaziny] ethanesulfonic acid (HEPES)-KOH (pH 7.5). All the proteins were quantified by the Bradford method and stored at −80 ​°C in aliquots. Proteins were separated on 10% SDS-PAGE and visualized by Coomassie blue.

Preparation of oligonucleotide helix substrates

In brief, the RNA helix, DNA helix and RNA-DNA hybris helix were prepared by annealing two complementary nucleic acid strands. One strand was labeled at 5′ end with hexachloro-fluorescein (HEX), and the other one was unlabeled. HEX-labeled oligonucleotide strands were purchased from TaKaRa (Dalian, China). Unlabeled DNA strands were synthesized by Invitrogen, and unlabeled RNA strands were in vitro transcribed using T7 RNA polymerase (Promega, Madison, WI). The transcribed RNA strands were purified by Poly-Gel RNA Extraction Kit (Omega Bio-Tek, Guangzhou, China) according to the manufacturer's instructions. The two strands were mixed in a proper ratio, and annealed through heating and gradually cooling as previously described (Cheng et al., 2013; Yang et al., 2014). The resulting duplexes were examined by 15% native-PAGE gel to make sure that all single-stranded RNA and DNA were annealed in a ratio of 1:1.

The standard RNA helix substrate (R/R∗) was annealed with a 42-nt HEX-labeled single-stranded RNA1 and a 24-nt non-labeled single-stranded RNA2. The 3′-tailed and 5′-tailed RNA substrates were constructed by annealing RNA1 and a 48-nt non-labeled single-stranded RNA3 (3′-tailed) and RNA4 (5′-tailed). The blunt-ended RNA substrate was prepared by annealing RNA1 and a 42-nt non-labeled single-stranded RNA5. (D∗/D) was prepared by annealing a 28-nt HEX-labeled single-stranded DNA1 and a 49-nt non-labeled single-stranded DNA2. (D/R∗) was prepared by annealing RNA1 and a 30-nt non-labeled single-stranded DNA3. (R∗/D) was prepared by annealing RNA1 and a 54-nt non-labeled single-stranded DNA4. The 30-nt complementary RNA helix substrate was prepared by annealing RNA1 and a 30-nt non-labeled single-stranded RNA6. The 30-nt complementary 3′-tailed and 5′-tailed RNA substrates were constructed by annealing RNA1 and a 30-nt non-labeled single-stranded RNA7 (3′-tailed) and RNA8 (5′-tailed). All oligonucleotides used in this study are listed in Supplementary Table S2.

NTPase assay

NTPase activities were determined via measuring the inorganic phosphate released during NTP hydrolysis using a direct colorimetric assay according to standard procedure as previously described (Cheng et al., 2013). All of the results obtained from this quantitative assay represent the average of three repeated experiments.

Gel mobility shift assay

Gel mobility shift assay was performed in 50 ​mM HEPES-KOH (pH 7.5), 100 ​mM NaCl, 2 ​mM MgCl₂, 1 ​mM tRNA, 2 ​mM DTT, 20 U RNase inhibitor (Promega), in a total volume of 10 ​μL reaction with indicated amount of protein and 0.1 ​pmol of dsRNA. The dsRNA was labeled with DIG-UTP (Roche) by in vitro transcription and derived from 200-nt ZIKV 3′UTR or EGFP. Reactions were incubated for 30 ​min at 25 ​°C and terminated by the addition of 2.5 ​μL 5 ​× ​sample buffer [20 ​mM Tris-HCl (pH 7.5), 36% glycerol and 0.1% bromophenol blue]. The nucleic acid-protein complexes were separated by electrophoresis on protein 1.5% agarose gel and transferred to Hybond-A nylon membrane (GE Healthcare). After that, membrane was subjected to cross-linking at 120 ​°C and was incubated with anti-DIG-alkaline phosphatase antibody (Roche), followed by incubation with CDP-STAR (Roche) for 15 ​min at 37 ​°C. The signals were then detected by X-ray film (Fujifilm, Tokyo, Japan).

Nucleic acid helix unwinding assay

The standard helix destabilizing assay was performed as previously described (Yang et al., 2014). Briefly, 20 ​pmol of recombinant protein and 0.1 ​pmol of HEX-labeled helix substrate were added to a mixture containing 50 ​mM HEPES-KOH (pH 7.5), 1 ​mM MgCl₂, 100 ​mM NaCl and 20 U RNase inhibitor (Promega). Reactions were incubated for 60 ​min at 37 ​°C and terminated by adding 5 ​× ​loading buffer [100 ​mM Tris-HCl, 1% SDS, 50% glycerol, and bromophenol blue (pH 7.5)]. The mixtures were then electrophoresed on 15% native-PAGE gels, followed by scanning with a Typhoon 9500 imager (GE Healthcare, Piscataway, NJ).

Data availability

The data that support the findings of this study are available within the article and its supplementary information files.

Ethics statement

This article does not contain any studies with human or animal subjects performed by any of the authors.

Author contributions

Ting Shu: Conceptualization, Data curation, Funding acquisition, Investigation, Methodology, Validation, Writing-original draft. Mengyu Hu: Investigation. Yiyang Li: Investigation. Mengge Han: Investigation. Xingyue Wang:Investigation. Qi Cheng: Funding acquisition. Xiqiu Xu: Funding acquisition. Yang Qiu: Project administration, Supervision, Writing-review & editing. Xi Zhou: Conceptualization, Project administration, Supervision, Writing-review & editing.

Conflict of interest

The authors declare that they have no conflict of interest. Prof. Zhou Xi is editorial board member for Virologica Sinica and was not involved in the editorial review or the decision to publish this article.

Appendix A. Supplementary data

The following are the Supplementary data to this article.

Supplementary Figure S1.

Flavivirus NS1 amino acid sequence alignment and protein expression. A Amino acid sequence alignment results of ZIKV NS1, JEV NS1, and DENV2 NS1. B The MBP-fusion NS1 was expressed using a eukaryotic (baculovirus) system. The purified recombinant proteins were subjected to 10% SDS-PAGE followed by Coomassie brilliant blue R250 staining. Lane 1, protein marker; lane 2, the purified MBP alone; lane 3, the purified eukaryotically expressed MBP-ZIKV NS1; lane 4, the purified eukaryotically expressed MBP-JEV NS1; lane 3, the purified eukaryotically expressed MBP-DENV2 NS1.