Mechanical regulation of the helicase activity of Zika virus NS3

Abstract

Zika virus (ZIKV) is a positive-sense single-stranded RNA virus that infects humans and can cause birth defects and neurological disorders. Its non-structural protein 3 (NS3) contains a protease domain and a helicase domain, both of which play essential roles during the viral life cycle. However, it has been shown that ZIKV NS3 has an inherently weak helicase activity, making it unable to unwind long RNA duplexes alone. How this activity is stimulated to process the viral genome and whether the two domains of NS3 are functionally coupled remain unclear. Here, we used optical tweezers to characterize the RNA-unwinding properties of ZIKV NS3—including its processivity, velocity, and step size—at the single-molecule level. We found that external forces that weaken the stability of the duplex RNA substrate significantly enhance the helicase activity of ZIKV NS3. On the other hand, we showed that the protease domain increases the binding affinity of NS3 to RNA but has only a minor effect on unwinding per se. Our findings suggest that the ZIKV NS3 helicase is activated on demand in the context of viral replication, a paradigm that may be generalizable to other flaviviruses.

Significance

The non-structural protein 3 (NS3) is a critical component of the genome replication complex of Zika virus (ZIKV) and an attractive target for antiviral drugs. However, the development of potent NS3 inhibitors is hampered by a lack of mechanistic characterization of NS3’s operation. Specifically, it remains puzzling how ZIKV NS3—which harbors a weak RNA helicase activity—can nonetheless process the long RNA genome during the viral replication cycle. In this study, we employ optical tweezers to apply pNs of force to destabilize the duplex RNA substrate and find that such mechanical stimulation dramatically enhances the RNA unwinding activity of ZIKV NS3. This finding provides a biophysical basis for understanding the regulation of ZIKV NS3’s activity in the cell.

Introduction

Zika virus (ZIKV) is a positive-strand RNA virus that belongs to a group of mosquito-borne flaviviruses of the family Flaviviridae also including dengue virus, yellow fever virus, and West Nile virus (1,2). ZIKV infection has been associated with birth defects and neurological disorders such as microcephaly (3) and Guillain-Barré syndrome (4). The 11-kilobase single-stranded RNA (ssRNA) genome of ZIKV encodes a polyprotein that is processed into three structural proteins (C, prM/M, and E) for viral particle assembly as well as seven non-structural proteins (NS1, NS2A, NS2B, NS3, NS4A, NS4B, and NS5) for viral replication inside the host cell (5). The viral non-structural proteins, together with host proteins, form the replication complex at the endoplasmic reticulum membrane (6). Among the non-structural proteins, NS3 is a critical component of the viral genome replication complex and consists of two domains, an N-terminal serine protease domain that requires the NS2B cofactor for its activity, and a C-terminal helicase domain that belongs to the superfamily 2 group of helicases and uses energy from NTP hydrolysis to unwind nucleic acid duplexes (7,8,9). NS2B-NS3 protease is responsible for all cytoplasmic cleavages including at the junctions between NS2A/NS2B, NS2B/NS3, NS3/NS4A, NS4B/NS5, and, within the capsid, NS2A and NS4A proteins (5). During viral RNA amplification, NS5, the RNA-dependent RNA polymerase, first uses the positive-sense genomic RNA as the template to generate a double-stranded RNA (dsRNA) intermediate. NS3 then separates the duplex RNA to generate the negative-sense ssRNA that serves as template for the synthesis of new positive-sense genomic RNA.

ZIKV NS3 binds to the 3′ overhang of double-stranded nucleic acids and unwinds the duplex in a 3′-to-5′ direction. Curiously, bulk biochemical assays have shown that ZIKV NS3 is a weak helicase, unable to separate RNA duplexes longer than ∼20 base pairs (bp) (10). This raises the question of how its helicase activity can be enhanced to a level necessary for efficient viral genome replication. Moreover, even though both enzymatic activities of NS3 (namely protease and helicase) are essential for viral assembly and replication, it remains unclear whether these two domains with disparate functions influence each other’s activities for ZIKV and for flaviviruses in general. Regarding the effect of NS3’s protease domain on its helicase activity, previous studies on different Flaviviridae members reached inconsistent conclusions. For example, a structural study of the dengue virus 4 NS3 suggested that its protease domain upregulates the activity of the helicase domain (11). Similarly, the hepatitis C virus (HCV) NS3 requires an intact protease domain to unwind duplex RNA efficiently (12). In contrast, West Nile virus NS3 exhibits a similar RNA-unwinding activity to its helicase domain alone (13). The structure of Murray Valley encephalitis virus NS3 shows minimal interaction between two well-separated domains (14). As for ZIKV NS3, it was shown that the helicase domain alone has a lower RNA-unwinding efficiency compared with the full-length protein but only under certain concentration regimes (10). It is difficult to measure from bulk biochemical and structural studies the intrinsic helicase characteristics of NS3 (e.g. velocity, processivity, step size), which require tools such as single-molecule techniques to follow the enzymes in action.

In this work, we used an optical tweezers assay (15) to visualize RNA unwinding by ZIKV NS3 in real time. We showed that the unwinding velocity and processivity of NS3 significantly increased when an external force on the order of 12–18 piconewton (pN) was applied to destabilize the dsRNA substrate. NS3 was observed to unwind RNA in 3-bp steps, each preceded by an ATP-concentration-dependent dwell. We also found that the full-length NS3 and the helicase domain alone displayed similar unwinding properties. Together, the single-molecule results presented in this study provide a biophysical framework for understanding the regulation of ZIKV NS3 activity.

Materials and methods

Protein purification

ZIKV NS3 constructs (NS3F and NS3h) were cloned into an MBP-tagged expression vector in a pET30a backbone. Transformed Escherichia coli Rosetta TM (DE3) cells (Novagen) were grown at 37°C until optical density 600 reached 1.2. Cells were induced with 0.2 mM IPTG at 18°C for 4 h and then harvested and resuspended in a Ni-binding buffer (20 mM Tris-HCl (pH 8.0), 250 mM NaCl, 10 mM imidazole) supplemented with DNase and protease inhibitors. After sonication, soluble protein was purified from cell lysate by a HiTrap column (GE Healthcare, Chicago, IL, USA), followed by TEV protease cleavage overnight at 4°C to remove the MBP tag, a Ni-NTA purification step, and a gel-filtration purification step using a Superdex-200 column (GE Healthcare). The final product was stored in 20 mM Tris-HCl (pH 8.0) and 250 mM NaCl.

Bulk helicase assay

The bulk helicase activities of ZIKV NS3F and NS3h were monitored by a fluorescence-based dsRNA unwinding assay, which was conducted at room temperature in a 20-μL volume with a substrate composed of a Cy5-labeled 22-nt RNA strand (Cy5-5′-GGUUCUGCGGGUGGCGGUACUA-3′) annealed to a 36-nt strand (5′-UAGUACCGCCACCCGCAGAACCAAAAAAAAAAAAAA-3′). The reaction mixture contained 25 mM MOPS (pH 6.5), 20 mM NaCl, 2 mM MgCl₂, 25% glycerol, 0.1% Tween-20, 1 mM DTT, 20 nM RNA substrate, 200 nM “trap” oligo complementary to the short strand (5′-UAGUACCGCCACCCGCAGAACC-3′), 5 μM helicase, and 5 mM ATP. The reactions were performed at room temperature for 2 h and then stopped by proteinase K. The reaction products were separated from the substrate on a non-denaturing 10% polyacrylamide gel in TBE buffer, scanned with a Typhoon imager, and analyzed using ImageJ.

Bulk binding assay

NS3-RNA interaction was evaluated in bulk by a fluorescence polarization assay using 50 nM ssRNA (5′-UCAAUAUGCUGAAA-3′-FAM) and varying concentrations of protein in a buffer containing 25 mM MOPS (pH 6.5), 20 mM NaCl, 2 mM MgCl₂, 0.1% Tween-20, 1 mM DTT, and 0.9% glycerol. Fluorescence readings were taken after a 30-min incubation at room temperature using a plate reader (BioTek Synergy H1). K_D values were extracted by fitting the data to ΔmP = mP_max [NS3]/(K_D + [NS3]).

Single-molecule helicase assay

Substrate preparation

The substrate consists of two 1.5-kbp biotinylated dsDNA handles annealed to an RNA hairpin containing a 180-bp stem and a tetraloop. A 22-nt single-stranded region at the 5′ flank of the hairpin serves as the NS3 loading site. The RNA was synthesized by in vitro transcription of a DNA template previously described (16) using the MEGAscript T7 Transcription Kit (Thermo Fisher Scientific) and then purified using the MEGAclear Kit (Thermo Fisher Scientific). The complete sequence of the RNA hairpin is as follows:

• 5′ half: 5′-GUAGUAGUAGUAGUAGUGGUUGUUGUGGUAGUCGUUGUUGUGGUGGUCGUAGUCGUGGUCGUGGUGGUUGUUGUAGUUGUUGUUGUUGUUGUAGUGGUUGUGGUAGUGGUAGUCGUGGUGGUUGUUGUAGUGGUUGUUGUAGUGGUGGUGGGCCACGCGGCGAAAGCCUCAUUGGUCUCGUU

• 3′ half: 3′-CAUCAUCAUCAUCAUCACCAACAACACCAUCAGCAACAACACCACCAGCAUCAGCACCAGCACCACCAACAACAUCAACAACAACAACAACAUCACCAACACCAUCACCAUCAGCACCACCAACAACAUCACCAACAACAUCACCACCACCCGGUGCGCCGCUUUCGGAGUAACCAGAGCUU

The DNA handles were generated by PCR using modified primers to generate the overhang for annealing to the hairpin-containing RNA. The forward primer for the 5′ overhang handle contains an inverted base that terminates Phusion DNA polymerase (5′-GGATCCTAGGATTCGTGCTCATGGTCTC/iInvdT/GCATAACCCCTTGGGGCCTCTAAACG-3′). The forward primer for the 3′ overhang handle contains inverted bases at its 5′ end that terminate Taq polymerase (/5InvdG//iInvdC//iInvdA//iInvdA//iInvdA//iInvdT//iInvdC//iInvdT//iInvdC//iInvdC//iInvdG//iInvdG//iInvdG//iInvdG//iInvdT//iInvdT//iInvdC//iInvdC//iInvdC//iInvdC//iInvdA//iInvdA//iInvdT//iInvdA//iInvdC//iInvdG/TAGTCTAGAGAATTCATTGCGTTCTGTACA/3ddC/). The reverse primers for the 5′ and 3′ overhang handles each contain a biotin at their respective 5′ ends for bead attachment. To assemble the 5′ overhang handle complex, 1 μM 5′ overhang handle was mixed with 400 nM RNA hairpin in a total volume of 20 μL in TL buffer (50 mM HEPES (pH 7.5), 55 mM KOAc, 6 mM Mg(OAc)₂, and 1 mM DTT), incubated at 65°C for 90 s, and cooled to 4°C. For bead conjugation, a 10-μL reaction containing 3.5 μL of the annealed complex and 1.6 μL of 0.5% streptavidin-coated polystyrene beads (2 μm in diameter, Spherotech) was incubated on ice for 30 min. Then, the mixture was diluted with 1 mL TL buffer. A separate reaction containing 50 nM 3′ overhang handles and 1.6 μL of beads was incubated and diluted in the same way.

Optical tweezers experiments

All single-molecule measurements were made on a MiniTweezers instrument (17). The sample buffer consisted of 25 mM MOPS (pH 6.5), 20 mM NaCl, 2 mM MgSO₄, 0.9% glycerol, 0.1% Tween-20, and 1 mM DTT. To form a tether, a bead conjugated to the 5′ overhang handle complex was immobilized on a pipette via suction and brought into close proximity to a second bead conjugated to 3′ overhang handles held in an optical trap. Upon hybridization of the 3′ overhang handle with the 5′ overhang handle/RNA hairpin complex, a tether was formed, and its force-extension behavior was obtained by moving the optically trapped bead at a speed of 70 nm/s away from the bead held in the pipette. Once a single tether was confirmed by the shape of the force-extension curve, NS3F (1 nM)/NS3h (3 nM) and ATP (varying concentrations) were injected into the chamber, and unwinding data were collected in a force-feedback mode to maintain a constant tension in the tether. Data were collected at 200 Hz. Experiments were performed at room temperature (23°C ± 1°C).

Data analysis

Single-molecule data were analyzed by custom-written MATLAB (Natick, MA, USA) scripts (16). The raw readout in nanometers was converted to bp using the extensible worm-like-chain model of ssRNA with the following parameters: persistence length = 1 nm, contour length = 0.59 nm per RNA nucleotide, and stretch modulus = 1000 pN (18). Unwinding velocities were determined by linear fitting to the time-distance data (from start to end of a continuous event). Unwinding processivities were determined as the maximum distance traveled prior to enzyme detachment or reversal. For step-detection analysis, time series data (200 Hz) were first filtered with a zero-phase digital filter using a 10-data-point window and run through a t-test-based algorithm to detect transition points with a significance threshold of 0.005.

Pairwise distance analysis

The extension data were processed to convert the units to bp and filtered as described above. The data were organized by segments where each segment contains data between the start point of each translocation event to its peak location. Distances between each pair of data points within the segment were computed, and only positive values were included in the final data and plotted.

Statistical analysis

ANOVA tests were performed across entire groups when the overall trend was not known a priori. Kolmogorov-Smirnov tests were performed to compare individual pairs of conditions. Significance levels are as follows: ^∗∗∗∗p ≤ 0.0001; ^∗∗∗0.0001 < p ≤ 0.001; ^∗∗0.001 < p ≤ 0.01; ^∗0.01 < p ≤ 0.05; ns, p > 0.05.

Results

Direct visualization of RNA unwinding by ZIKV NS3

To characterize the helicase activity of ZIKV NS3, we first expressed and purified a fusion construct NS3F in which the hydrophilic region of the NS2B cofactor (residues 47–89)—required for the proper folding and protease activity of NS3 (19)—is juxtaposed with the full-length NS3 (residues 1–617) (Fig. S1). We then used an optical tweezers assay to monitor NS3-driven RNA unwinding at the single-molecule level. A 180-bp RNA hairpin was tethered through two 1.5-kbp DNA handles between a 2-μm bead held in an optical trap and another 2-μm bead immobilized on a micropipette by suction. The RNA bears a 22-nt single-stranded region on the 3′ flank of the hairpin, which serves as the NS3 loading site (Fig. 1 a). We first collected force-induced unfolding trajectories of the RNA hairpin in the absence of NS3, which exhibited a characteristic pattern (Fig. 1 b) consistent with previous observations (20). No unfolding event was observed when the force applied to the tether was below 20 pN. Therefore, in the following experiments with NS3, we kept applied forces below 20 pN so that any observed unwinding must result from the enzyme’s helicase activity.

Figure 1.

Optical tweezers assay for studying RNA unwinding by ZIKV NS3. (a) Experimental geometry. A biotin ("B") in each DNA handle is used to attach the DNA to a streptavidin (“S”)-coated bead. (b) Representative force-extension trajectory of a tethered substrate depicted in (a) showing unfolding (magenta) and refolding (green) of the RNA hairpin in the absence of NS3. (c and d) Control experiments showing minimal NS3-driven RNA unwinding without any nucleotides (c) or with 1 mM AMP-PNP (d). (e) Representative extension-time trace with 1 nM NS3F and 1 mM ATP showing multiple rounds of RNA-unwinding events. Data in (c)–(e) were taken under 18 pN of external force. To see this figure in color, go online.

Considering that ZIKV NS3 appears to be a weak helicase from bulk experiments, we first made single-molecule measurements of NS3’s helicase activity at 18 pN of external force—a force close to the threshold required for mechanically induced unfolding—to seek maximum unwinding efficiency. At the beginning of each recording, we verified that a single tether was held between the two beads by examining its characteristic unfolding and refolding pattern (Fig. 1 b). We then flowed 1 nM of NS3F into the reaction chamber, held the tether at a constant force (18 pN), and monitored the end-to-end distance change of the RNA hairpin. Based on the theoretical model of ssRNA elasticity (21), the inter-bead distance change can be converted, at a given force, into the number of RNA bp unwound. As expected for ATP-dependent helicases, no unwinding events were observed in a nucleotide-free buffer (Fig. 1 c) or in the presence of AMP-PNP, a non-hydrolyzable ATP analog (Fig. 1 d). By contrast, in the presence of ATP (1 mM), we observed multiple rounds of RNA-unwinding events—as reflected by an increase of tether extension over time—within the same recording (Fig. 1 e), clearly demonstrating the ATP-dependent helicase activity of NS3F. All observed unwinding events, however, terminated before reaching the full length (180 bp) of the hairpin. Two types of behavior were observed upon termination (Fig. 1 e, zoom-in views): an instantaneous re-zipping of the RNA hairpin, which likely signifies NS3 dissociation, and a gradual re-zipping of the hairpin, which indicates that NS3 may undergo backward slippage or strand switching.

Effect of [ATP] on NS3-driven RNA unwinding

Next, we investigated the effect of ATP concentration on the helicase activity of ZIKV NS3. Representative single-molecule unwinding trajectories at 1, 0.1, and 0.05 mM ATP are presented in Fig. 2 a. We observed slower unwinding at lower [ATP] (Fig. 2 b). The mean velocity-[ATP] curve can be fit to a Michaelis-Menten function (V = V_max [ATP]/(K_M + [ATP]), yielding V_max = 45 bp/s (95% confidence interval [CI]: 40 bp/s, 51 bp/s) and K_M = 156 μM (95% CI: 107 μM, 227 μM) at 18 pN of applied force (Fig. 2 c).

Figure 2.

Effect of ATP concentration on the unwinding activity of ZIKV NS3. (a) Representative extension-time traces of RNA unwinding by NS3F (1 nM) with 1 mM ATP (black), 0.1 mM ATP (red), and 0.05 mM ATP (blue) at 18 pN of external force. The decrease in extension corresponds to re-zipping of the RNA hairpin due to enzyme disengagement or reverse translocation. (b) Histograms of unwinding velocity at 1 mM ATP (black; 141 traces), 0.1 mM ATP (red; 154 traces), and 0.05 mM ATP (blue; 56 traces) at 18 pN of external force. (c) Unwinding velocity as a function of ATP concentration with Michaelis-Menten fit. Data are presented as mean ± standard deviation (SD). n = 18, 56, 154, 66, 79, and 141 events for 0.01, 0.05, 0.1, 0.2, 0.5, and 1 mM ATP, respectively. (d) Unwinding processivity for NS3F at different ATP concentrations for the data used in (c). Each event is shown as a dot. The mean processivity is 12, 20, 22, 25, 25, and 29 bp at 0.01, 0.05, 0.1, 0.2, 0.5, and 1 mM ATP, respectively. Error bars represent SD. Kolmogorov-Smirnov (KS) tests were conducted to compare between different conditions. To see this figure in color, go online.

The processivity of NS3 unwinding, which is defined as the number of RNA bp unwound before enzyme disengagement or reversal, shows no significant difference among ATP concentrations above K_M (0.2, 0.5, and 1 mM ATP) but decreases significantly once [ATP] drops below K_M (an average of 29 bp at 1 mM [ATP] compared with 12 bp at 0.01 mM [ATP]; Fig. 2 d). These results indicate that NS3 may detach from or back slide on the RNA while waiting for ATP binding and that ATP binding shifts the competition between termination and productive unwinding toward the latter. We note that sometimes partial hairpin re-zipping events were observed upon termination of unwinding (e.g., Fig. 2 a). This could be explained by two possible scenarios. (1) The leading NS3 dissociates from the ssRNA-dsRNA junction and additional helicases bound to the ssRNA prevent complete re-zipping of the hairpin. (2) The NS3 at the junction temporarily loses its grip on the RNA and re-engages after partial RNA re-annealing. We favor the latter scenario because the concentration of NS3 used in our experiments was low and its binding to RNA was expected to be infrequent, although a rigorous test requires NS3 titration experiments.

Upon examining the single-molecule trajectories, we found that ZIKV NS3 unwinds RNA in a stepwise fashion featuring alternating stationary waiting periods (dwell times) and translocation steps (Fig. 3 a). This pattern was observed across different ATP concentrations. The mean step size determined by our step-detection algorithm is 2.92 bp (95% CI: 2.85 bp, 3.00 bp). A pairwise distance distribution analysis of the unwinding trajectories revealed regularly spaced peaks at 3 bp, 6 bp, 9 bp, etc. (Fig. 3 b), again suggesting a step size of ∼3 bp for RNA unwinding by ZIKV NS3. The 3-bp periodicity in pairwise distance distribution is independent of the ATP concentration (Fig. S2). On the other hand, the mean dwell time shortens with increasing [ATP], from 0.64 s (95% CI: 0.50 s, 0.84 s) at 0.01 mM ATP to 0.13 s (95% CI: 0.12 s, 0.15 s) at 1 mM ATP, indicating that ATP binding takes place during the dwell between translocation steps. Notably, for all ATP concentrations tested, a single-exponential decay function can be fit to the dwell-time distribution (Fig. 3 c).

Figure 3.

ZIKV NS3 unwinds RNA in discrete steps. (a) Representative traces of RNA unwinding by NS3F at 0.01, 0.1, and 1 mM ATP under 18-pN force. Raw data (blue points) are overlaid with the idealized stepping trace (red line). (b) Pairwise distance distribution for the unwinding traces taken at 0.1 mM ATP (264 steps from 50 traces). (c) Dwell-time distributions for the stepping traces taken at 0.01 mM ATP (red), 0.1 mM ATP (magenta), and 1 mM ATP (green). The shaded areas represent 95% confidence intervals (CIs). The solid lines represent fits to single-exponential decay functions, yielding the time constants: 0.64 s (95% CI: 0.50 s, 0.84 s) for 0.01 mM ATP, 0.23 s (95% CI: 0.21 s, 0.26 s) for 0.1 mM ATP, and 0.13 s (95% CI: 0.12 s, 0.15 s) for 1 mM ATP. To see this figure in color, go online.

Previous bulk ATPase assay (conducted at zero force) reported k_cat of 1.5 s⁻¹ and K_M of 110 μM for ZIKV NS3 with a ssRNA substrate (10). Given the V_max of 45 bp/s and assuming 3-bp translocation per ATP hydrolyzed, our single-molecule assay using an RNA hairpin substrate suggests an ATP turnover rate of ∼15 s⁻¹ at 18 pN. Thus, notwithstanding the different substrates used, these measurements indicate that a hairpin-destabilizing force can increase the catalytic rate of ZIKV NS3.

Effect of external force on NS3-driven RNA unwinding

In the experiments described above, an external force of 18 pN was applied to the tether to destabilize the RNA hairpin and promote its unwinding by ZIKV NS3. To investigate how the magnitude of force impacts the enzyme’s helicase activity, we also collected single-molecule RNA-unwinding trajectories at 16, 14, and 12 pN of force (Fig. 4 a). The processivity of NS3 was observed to be force dependent (Fig. 4 b): at 18 pN, an average of 29 bp of dsRNA was unwound per event (at saturating [ATP]), with the maximum distance reaching ∼120 bp; at 12 pN, the enzyme only unwound 18 bp of dsRNA on average. Importantly, we found that the unwinding velocity of ZIKV NS3 is also sensitive to force, with the average velocity dropping from 39 bp/s at 18 pN to 13 bp/s at 12 pN (Fig. 4 c). This is consistent with previous bulk results showing that, at zero force, NS3 cannot unwind dsRNA substrates longer than 20 bp (10).

Figure 4.

Effect of force on the unwinding activity of ZIKV NS3. (a) Representative RNA-unwinding traces of NS3F at different hairpin-destabilizing forces. (b) Distributions of the RNA-unwinding processivity from individual unwinding events measured at different forces. The mean processivity is 18, 19, 22, and 29 bp at 12, 14, 16, and 18 pN of force, respectively. (c) Distributions of the unwinding velocity measured at different forces. The mean velocity is 13, 20, 39, and 39 bp/s at 12, 14, 16, and 18 pN of force, respectively. n = 25, 26, 120, and 141 events for 12, 14, 16, and 18 pN of force, respectively. Error bars represent SD. Welch’s ANOVA tests were performed to compare among multiple force conditions in a group, and the results are shown on top of (b) and (c). KS tests were performed to compare between individual pairs of forces.

Role of the protease domain in the unwinding activity of ZIKV NS3

Next, we asked whether the helicase activity of ZIKV NS3 is affected by its protease domain. To this end, we purified the helicase-only construct NS3h (residues 168–617) (Fig. S1) and compared its RNA-unwinding behavior with NS3F. We first conducted fluorescence polarization assays to measure the RNA binding affinities of NS3h and NS3F using a 14-nt ssRNA substrate. We obtained a K_D value of 672 nM (95% CI: 555 nM, 817 nM) for NS3h and 268 nM (95% CI: 235 nM, 306 nM) for NS3F (Fig. S3). The 2.5-fold difference in binding affinity explains the weaker helicase activity displayed by NS3h in bulk compared with the full-length ZIKV NS3 when using a low concentration of proteins (10). We then conducted bulk helicase experiments using a dsRNA substrate and found that NS3F and NS3h exhibited comparable helicase activities at a saturating enzyme concentration (Fig. S4). Finally, we carried out single-molecule RNA-unwinding experiments at various external forces (12–18 pN) using 3 nM NS3h (Fig. 5 a). Similar to what was observed for NS3F, the processivity of NS3h was dependent on force (Fig. 5 b): at 18 pN, an average of 26 bp of RNA duplex was unwound per event; at 12 pN, the average value dropped to 18 bp per event. There is no significant difference in processivity between NS3h and NS3F at all forces examined. The unwinding velocity of NS3h was also sensitive to force, although the sensitivity was somewhat lower compared with NS3F, with the average velocity decreased from 35 bp/s at 18 pN to 26 bp/s at 12 pN (Fig. 5 c). Therefore, we concluded that the protease domain only has a minor effect on ZIKV NS3’s RNA-unwinding activity and that its two domains (helicase and protease) are largely uncoupled from one another, in contrast to the strong coupling reported for NS3 from some other flaviviruses such as HCV (22,23).

Figure 5.

Comparison between the unwinding activities of NS3h and NS3F. (a) A representative recording and zoom-in views of RNA unwinding events driven by NS3h with 1 mM ATP at 18 pN of force. (b and c) RNA-unwinding processivities (b) and velocities (c) for NS3h and NS3F under different forces. n = 18, 25, 40, and 92 events at 12, 14, 16, and 18 pN of force, respectively, for NS3h. n = 25, 26, 120, and 141 events at 12, 14, 16, and 18 pN of force, respectively, for NS3F. Error bars represent SD. Welch’s ANOVA tests were performed to compare among multiple force conditions in a group, and the results are shown on top of (b) and (c). KS tests were performed to compare between NS3h and NS3F at a given force.

Discussion

In this article, we took a single-molecule approach to study the helicase activity of ZIKV NS3, an essential enzyme in the life cycle of a viral pathogen whose human infection is an emerging public health threat. Our data reveal that ZIKV NS3 harbors an ATP-dependent RNA-unwinding activity that is highly sensitive to the external force applied to the RNA substrate. This behavior contrasts with the NS3 enzyme from HCV, which belongs to a different genus (Hepacivirus) of the family Flaviviridae from ZIKV (Flavivirus) and unwinds RNA insensitively to applied force (18). The strong force dependence of ZIKV NS3 is instead reminiscent of the severe acute respiratory syndrome coronavirus 2 nsp13 helicase that we recently characterized, a superfamily 1 helicase whose unwinding velocity can also be potently enhanced by force (16). The unwinding velocity of some hexameric ring-shaped viral replicative helicases, such as T7 gp4 (24) and T4 gp41 (25), is also force dependent. A useful theoretical framework to characterize helicases is to classify them into “active” or “passive” enzymes based on the free energy provided by enzyme to actively open the base pairs (26). For both active and passive helicases, the unwinding processivity is expected to increase with higher force (27). However, the unwinding velocity of a helicase becomes less sensitive to force when the interaction potential is higher (i.e., more active) (27). The strong dependence of ZIKV NS3’s unwinding velocity on force indicates that it operates more like a passive helicase. Future experiments to measure the unwinding velocity of ZIKV NS3 with different RNA sequences and its translocation velocity on ssRNA will further elucidate its unwinding mechanism.

The low basal helicase activity of ZIKV NS3 has been reported by bulk biochemical assays (10). However, it was not clear whether this is due to weak binding, slow unwinding, or low processivity of the enzyme. Our single-molecule study unambiguously describes the intrinsic helicase properties of ZIKV NS3. The diminished unwinding processivity and velocity at low forces explain why the enzyme is unable to separate duplexes longer than 20 bp in bulk (i.e., at zero force) (10). Therefore, in order for NS3 to efficiently process the entire viral RNA genome, there must exist an external factor that exerts a dsRNA-destabilizing force or switches NS3 into a more potent helicase. These mutually non-exclusive effects may be exerted by the RNA-dependent RNA polymerase NS5 (28). The interaction between NS3 and NS5 has been shown to be important for viral replication in HCV, Dengue virus, and ZIKV (10,23,29). Interestingly, ZIKV NS5-NS3 interaction was found to mainly promote NS3’s RNA-unwinding velocity rather than processivity (10). This may provide a clue on the magnitude of destabilizing force generated by NS5, given the non-equivalent dependence of NS3’s processivity and velocity on force (Fig. 4 b and c). Future single-molecule experiments using the NS5-NS3 complex will directly characterize the mechanical effect of NS5 on NS3 operation.

Our data show that ZIKV NS3 unwinds RNA in discrete 3-bp steps separated by [ATP]-dependent dwells. The step size is a fundamental helicase property that constrains the possible models for enzyme operation. The E. coli UvrD helicase also exhibits a step size averaging 3 bp but has a smaller substep size (∼1 bp) per ATP hydrolyzed (30,31). This phenomenon (i.e., apparent larger steps composed of smaller substeps) can be rationalized by a strand sequestration and delayed release model (31), which was also proposed for HCV NS3 (32,33). While a similar model involving substeps smaller than 3 bp may be applicable to ZIKV NS3, the single-exponential distribution of dwell times preceding the observed 3-bp steps—regardless of ATP concentration—indicates that each dwell contains only one ATP binding event. Alternatively, each dwell could contain multiple ATP binding events, but one of them occurs much slower than the others (i.e., rate limiting) (34). Delineating the precise movement of the RNA strand relative to the enzyme’s power stroke awaits high-resolution single-molecule and structural data.

ZIKV NS3 shares mostly identical sequences with NS3 from other flaviviruses such as Dengue virus, yellow fever virus, West Nile virus, and Japanese encephalitis virus (notably, not with HCV NS3) (Fig. S5). Therefore, we speculate that the characteristics of ZIKV NS3 learned from this study can be applied to other viral systems as well. In particular, the “on-demand” helicase activity based on mechanical stimulation may be a general strategy to minimize spurious unwinding but still enable efficient viral genome replication when needed. This is likely an important design principle for maintaining stability of the viral genome and integrity of the viral replication machinery (35). These mechanistic insights hold promise to facilitate the development of antivirals inhibiting the actions of NS3 to combat infection by ZIKV and other flaviviruses that still lack vaccine or specific treatment (36).

Author contributions

T.J. and S. Liu conceived the study and oversaw the experiments. X.C. and Y.L. prepared the protein samples. X.C. performed the bulk biochemical assays. X.C. performed the single-molecule experiments with help from S.Y. and S. Li. K.L. and X.C. analyzed the single-molecule data. X.C., K.L., S.Y., T.J., and S. Liu wrote the manuscript with input from all authors.

Editor: Susan J. Schroeder.

Data availability

Experimental materials, source data files, and data analysis scripts related to this work are available upon request.